Department of Physics MSSEESA Conference on P.O. Box 32379 … · 2019-09-05 · Thin Films for...

1st Young Scientists’ MSSEESAConference on Materials Science and Solar Cell Technology 2013 MSSEESA CONFERENCE PROCEEDINGS

Held at United Kenya Club on 28-29 November 2013, Nairobi, Kenya

Copyright @ 2013 Materials Science and Solar Energy Network for Eastern and Southern Africa (MSSESA)and the authors.

No part of this material protected by this copy right maybe reproduced or utilized in any form or by any means, without the prior written permission of the copyright owners, unless the use is a fair dealing for the purpose of

private study, research or review. The authors reserve the right that their material can be used for purely education and research purpose.

1st Young Scientists’ MSSEESA Conference on Materials Science and Solar Cell Technology

MSSEESA Conference Proceedings iii

MSSEESA COORDINATING BOARD

Dr. Sylvester Hatwaambo MSSEESA Chief Coordinator

Department of Physics University of Zambia P.O. Box 32379-10101 Lusaka, Zambia

Prof. Julius M. Mwabora Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya

Prof. Rogath T. Kivaisi Department of Physics University of Dar es Salaam P.O. Box 35063 Dar es Salaam,

Tanzania

Dr. Onesmus Munyati Department of Physics University of Zambia P.O. Box 32379-10101 Lusaka, Zambia

Prof. Tom Otiti Department of Physics Makerere University P.O. Box 7062, Kampala, Uganda

Prof. Joseph Buchweishaija Department of Chemistry University of Dar es Salaam P.O. Box 35063 Dar es Salaam,

Tanzania

Prof. Bernard O Aduda Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya

Prof. Yusto Kaahwa Department of Physics Makerere University P.O. Box 7062, Kampala, Uganda

Prof. Maurice M. Mwamburi Department of Physics University of Eldoret P.O. Box 1125-30100 Eldoret, Kenya

Dr. Christopher Maghanga Department of Physics Kabarak University Private Bag 20157, Kabarak, Kenya


2013 MSSEESA Conference proceedings ii

WELCOME NOTE BY MSSEESA CHIEF COORDINATOR, DR SYLVESTER HATWAAMBO

It gives me great pleasure and honour to welcome you all to this first Young Scientist MSSEESA Conference on Materials Science and Solar Cell Technology. Recent years have seen a tremendous development in materials science giving rise to a wide range of new materials, fine tuning or re-engineering of existing ones. This rapid development has in turn resulted in many new applications in many diverse fields from medicine, engineering, electronics, smart fabrics to solar energy. Accelerated developments in materials science and emergency of new applications can be ascribed to high level of research activities in this area. It is without doubt that this has not only had positive health impacts but also how we live. Research and application of materials for solar energy is particular interest to MSSEESA as it recognizes the significance that energy plays to socio-economic development. However, it is a known factor that greater part of rural Africa remains without electricity due to high cost for grid extension and installation. While solar energy is seen as an alternative that has the potential for rapid penetration in provision of electrical energy to rural areas, costs associated with large scale solar system installation remain a barrier. These costs are related to costs associated with processing of silicon that is used in conventional inorganic solar cell. MSSEESA recognizes the need for research in materials for low-cost efficient solar cells that have superior properties. In order for the region to contribute towards achieving this goal there is need for cadre of high skilled human resource actively engaged in research materials for solar energy applications. This conference, therefore, is significant in the sense that it brings together young scientists from different countries working in the areas of materials science and solar energy whose research can have societal impact. The nature and diversity of papers to be presented clearly reflected the strong materials in solar energy technology linkage. It is the aspiration of MSSEESA that its contribution from the various Chapters will in future be recognized through its research outputs.

I would also like to thank the local organizing committee, the MSSEESA Board members, the International Science Program (ISP), the Director, the International Programme in the Physical Sciences (IPPS), Prof Ernst van Groningen, for their support materially, financially or otherwise to the hosting of this Conference.

Dr. Sylvester Hatwaambo

MSSEESA Chief Coordinator


2013 MSSEESA Conference proceedings i

WELCOME NOTE BY THE CHAIRMAN, LOCAL ORGANIZING COMMITTEE, PROF. JULIUS M. MWABORA

I wish to welcome you all to the 2013 Materials Science and Solar Energy Network for Eastern and Southern Africa (MSSEESA) Conference on Materials Science and Solar Cell Technology. Since its inception, this is the first MSSEESA Young Scientist conference of its kind where early career scientists, PhD students and Masters Students will be presenting their work.

In our midst today, there are prominent scientists in the area of material science, solar energy materials and solar cell as well as thin film experts, together with these students drawn from various universities within the MSSEESA network which includes Kenya (University of Nairobi and University of Eldoret), Uganda (Makerere University), Tanzania (University of Dar Es Salaam) and Zambia (University of Zambia).

For the two days of this conference, you will be able to see contributions from young scientists who will also benefit from the network’s senior scientists.

We welcome you to this exciting and beneficial conference.

Prof. Julius M. Mwabora

Chairman, Local Organising Committee


2013 MSSEESA Conference proceedings iv

MSSEESA CONFERENCE LOCAL ORGANISING COMMITTEE

Prof. Julius M. Mwabora, Chairman Local Organizing Committee

Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya

Prof. Maurice M. Mwamburi Department of Physics University of Eldoret P.O. Box 1125-30100 Eldoret, Kenya

Dr. Robinson J. Musembi Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya

Dr. Christopher Maghanga Department of Physics Kabarak University Private Bag 20157, Kabarak, Kenya

Mr. Bernard M. Kamau Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya


2013 MSSEESA Conference proceedings vii

14.15 – 14.30 Luminous Transmittance and Transition Temperature of VO2:Ce Thin Films Prepared by DC Reactive Magnetron Sputtering

MR. SULTAN AMBONISYE

14.45 – 14.45 Characterization of TiO2 Based Dye-Sensitized Solar Cell Prepared by Screen Printing Method

MR. M.N. MUENDO

14.45 – 15.00 Sizing a Standalone Photovolaitic Electrical Solar System for Domestic Application

MR. M. MUKERE

15.00 – 15.15 Optical Characterization of Compounds for Dye-Sensitized Solar Cell Applications

MS. PAULINE KABUGA

15.30 – 15.45 Projection for PV Sizing for a Stand-alone “all-direct current” Telecommunication System Using the Mismatch Factor Algorithms

MR. SYMON NYAGA

15.45 – 16.15 TEA BREAK 16.15 – 16.30 Optical Properties of Flash Evaporated

Thin Films: Effect of Film Thickness MR. AUSTINE MULAMA

DR. FRANCIS NYONGESA

16.30 – 16.45 Effect of Substrate Deposition Temperature on the Properties of SnxSey/ZnO:Sn

MS. SALLY NDONYE

16.45 – 17.00 Fabrication and Characterization of Germanium Doped Titanium Dioxide (Ge:TiO2) Thin Film for Photovoltaic Application

MR. SAMSON NJOGU

17.00 – 17.15 Optical and Electrical Characterization of CdxNil-xS and Sb2S3 Thin Films for Photovoltaic Applications

MR. S.R. TSISAMBO

29 Friday 2013 09.00 – 09.15 Influence of Films Thickness on Optical Properties of

Nb-doped TiO2 (NTO) Thin Films Deposited by DC Reactive Magnetron Sputtering

MR.E.R. OLLOTU DR. JAMES MDOE

09.15 – 09.30 Transparent and Conducting TiO2:Nb Thin Films Prepared by Spray Pyrolysis Technique

MR. MAXWELL MAGETO

09.30 – 09.45 A Density Functional Theory Study of Electronic Structure of TiO2 Rutile (110) Surfaces with Catechol Adsorbate

MR. V.K. MENGWA

09.45 – 10.00 Thermal and Photovoltaic Performance of a High Output Hybrid Photovoltaic/Thermal Concentrator System

MR. FRANCIS GAITHO

10.00 – 10.30 TEA BREAK 10.30 – 10.45 Hydrogen as an alternative fuel: An ab-initio study of

Lithium Hydride and Magnesium Hydride MR. D. MAGERO

DR. C. B. UISO

10.45 – 11.00 LEDs for Off-grid Electrification in Kenya MR. JAMES C. WAFULA

11.00 – 11.15 Effect of Sn Doping on the Electrical Properties of as Prepared and annealed ZnO thin films Prepared by Reactive Evaporation

MR. PETER KINYUA

11.15 – 11.30 Morphological and Structural Characterization of TiO2/Nb2O5 Composite Electrode Thin Films Synthesized by Electrophoretic Deposition (EPD) Technique

MR. JOHN NJAGI NGUU

MSSEESA Conference Proceedings vii


2013 MSSEESA Conference proceedings vi

The 1st YOUNG SCIENTIST MSSEESA CONFERENCE ON MATERIAL SCIENCE AND SOLAR CELL TECHNOLOGY, UNITED KENYA CLUB, NAIROBI, KENYA

PROGRAMME

TIME ACTIVITY PRESENTER CHAIRING 28 Thursday 2013 08.00 – 9.00 Registration of Participants 09.00 – 09.05 Welcome by Chief Coordinator and local

chairman MASTER OF

CEREMONY 09.05 – 10.00 Opening Ceremony speeches 10.00 – 10.30 TEA BREAK 10.30 – 10.45 Optimization of SnxSey Deposited by Reactive

Thermal Evaporation for Solar cell Application MR. LAWRENCE MUNGUTI

DR. ALEX OGACHO

10.45 – 11.00 Effects of Target Composition on the Optical Constants of DC Sputtered ZnO:Al Thin Films

MR. BERNARD SAMUEL

11.00 – 11.15 Optical and Electrical Properties of Pd Doped SnO2 Thin Films Deposited By Spray Pyrolysis

MR. VICTOR ODARI

11.15 – 11.30 Effects of Aluminum and Tungsten Co-doping on the Optical Properties of VO2 Based Thin Films

MR. CEPHAS LYOBA

11.30 – 11.45 Fabrication and Characterization of Aluminum and Gallium Mono- and Co-doped Zinc Oxide Thin Films by Radio Frequency Sputtering For photovoltaic Applications

MR. J.P. ENEKU

11.45 – 12.00 Efficiency Enhancement in P3HT:PCBM Blends using Squarylium III Dye

MR. M. TEMBO DR. JUSTUS SIMIYU

12.00 – 12.15 Structural and Electronic Properties of TiO2, Nb:TiO2 and Cr:TiO2: A First Principles Study

MS. WINFRED MULWA

12.15 – 12.30 Effect of Substrate Temperature on the Structural, Optical and Electrical Properties of DC Sputtered ZnO:B Films

MR. L. PHOLDS

12.30 – 12.45 Optoelectronic Properties of Palladium Doped Tin (IV) Oxide (Pd: SnO2) Thin Films Deposited by Spray Pyrolysis

MR. PATRICK MWATHE

12.45 – 13.00 Effects of Oxygen Partial Pressure and Substrate Temperature on Optical Properties of Sputter Deposited CuCrO2 Thin Films.

MR. ALEX ALFRED

13.00 – 14.00 LUNCH BREAK 14.00 – 14.15 Performance of TiO2/In(OH)xSy/Pb(OH)xSy

Composite eta Solar Cell Fabricated from Nitrogen Doped TiO2 Thin Film Window Layer

MR. CHARLES OPIYO

DR. NURU MLYUKA


2013 MSSEESA Conference proceedings v

MSSEESA CONFERENCE PROCEEDINGS EDITORIAL COMMITTEE

Prof. Julius M. Mwabora, Chairman Local Organizing Commiee Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya

Prof. Maurice M. Mwamburi Department of Physics University of Eldoret P.O. Box 1125-30100 Eldoret, Kenya Dr. Christopher Maghanga Department of Physics Kabarak University Private Bag 20157, Kabarak, Kenya

Dr. Robinson J. Musembi Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya

Dr. Sebastian Waita Department of Physics Kabarak University Private Bag 20157, Kabarak, Kenya

Dr. Elijah Ayieta Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya Dr. Silas Mureramanzi Department of Physics University of Nairobi P.O. Box 30197-00100 Nairobi, Kenya


2013 MSSEESA Conference proceedings viii

11.30 – 11.45 Optoelectronic Properties of F-co-doped PTO Thin Films Deposited by Spray Pyrolysis

MR. VICTOR ODARI

11.45 – 12.00 Aloe secundiflora Extract as a Green Corrosion Inhibitor for Carbon Steel in Portable Water Systems

MR. J. MUTASINGWA

DR. SILAS MURERAMANZI

12.00 – 12.15 Characterization of Indium doped Tin Selenide (In:SnxSey) Thin Films for Phase Change Memory Application

MR. BRADEN MUTANGE

12.15 – 12.30 Synthesis and Characterization of Niobium Oxide Thin Films for DSSC Application

MS. MIRIAM KINEENE

12.30 – 12.45 Optical Properties of TiO2 Based Multilayer Thin Films

MR. M. KITUI

12.45 – 13.00 Optical and Electrical Properties of Magnesium Doped Zinc Oxide for Photovoltaic Applications

MR. GODWIN ASIIMWE

13.00 – 14.00 LUNCH BREAK 14.00 – 14.15 Ab-initio Studies of Point Defects in TiO2: A Density

Functional Approach MR. P.V. MWONGA

DR. YAHYA MAKAME

14.15 – 14.30 Electrosynthesis of Organic Polymer Films using 1-methoxy-3-pentadecylbenzene Derived from Cashew nut Shell Liquid

MR. N. KOMBA

14.45 – 14.45 Preparation and Characterization of Transparent and Conducting Doped Tin Oxide

MR. MAXWELL MAGETO

14.45 – 15.00 Fabrication and characterization of TiO2/In(OH)xSy/SnS composite ETA solar cell

MR. R.M. ONCHURU

15.00 – 15.15 Characterization of SnxSey/SnO2:Co P-N Junction Deposited by Spray Pyrolysis for Photovoltaic Application

MR. R.G. GITONGA

15.30 – 15.45 Characterization of SnxSey/SnO2:Ni P-N Junction Prepared by Spray Pyrolysis for Photovoltaic Application

MR. N. MUGANMBI

15.45 – 16.15 TEA BREAK 16.15 – 16.30 Characterization of Tin doped Antimony Selenium

(Sn:SbxSe1.5x) Thin Film for Phase Change Memory (PCM) Applications

MR. ABEDNEGO MULU

DR. MARGARET SAMIJI

16.30 – 16.45 The Complementarity of Dye-Sensitized and Amorphous Silicon Photovoltaics in Field Application in the Tropics

MR. RAPHAEL OTAKWA

16.45 – 17.00 Deposition and Characterization of CuAlxB1-x Se2 Thin Film Deposited by DC-RF Co-sputtering for Photovoltaic Application

MR. CHARLES WANGATI

17.00 – 17.15 Nanoporous Ceramics for Water Filtration MR. JEFARASIO NDUNGU

17.15 – 18.15 PLENARY SESSION DR. ONESMUS MUNYATI

Effects of Oxygen Partial Pressure and Substrate Temperature on Optical Properties of Sputter Deposited

CuCrO2 Thin FilmsA. Alfred1, J. M. Mwabora1, R. J. Musembi1, S. M. Waita1, Th. Dittrich2, K. Ellmer2

1Department of Physics, University of Nairobi, P. O. Box 30197-00100, Nairobi, Kenya 2Helmholtz-Zentrum Berlin für Materialien und Energie, Hahn-Meitner-Platz 1, D-4109 Berlin, Germany

Abstract The effects of oxygen partial pressure and substrate temperature on optical properties of CuCrO2 thin films deposited on

float glass substrate by reactive direct current (DC) magnetron sputtering system using CuCr alloy targets have been studied. The sputtering was performed in Argon (Ar) and Oxygen ( ) atmosphere and the substrate temperature varied up to 263 °C. The optical constants: refractive index( n) extinction coefficient( k) at different oxygen partial pressure and substrate tem-peratures were determined from measured transmittance and reflectance data fitted in SCOUT software for wavelength range 200-2247 nm. The optical studies gave Urbach energy which ranges from 0.27-0.31 eV for samples prepared at Po2 between 0.153-0.187 bar and 0.81-1.45 eV for those prepared at substrate temperature of 263 °C and as-grown film, respectively.

Keywords: Delafossite, SCOUT, DC magnetron sputtering, partial pressure, Urbach energy.

1. Introduction Inorganic transparent conducting oxides (TCOs) have found wide applications in certain areas such as transparent elec-trodes in flat panel display [1], coating in low emissivity architecture glass [2], anti-static electric coatings [3], heat mirrors [4], anti-reflectance in solar cells [5] and liquid crystal displays (LCDs) [6]. The TCOs applied to transparent devices, such as transparent diodes, transistors and light-emitting diodes (LEDs) require high optical transpar-ency and electrical conductivity. Properties of these films (TCOs) are highly dependent on defect structure, method of deposition, substrate temperature and film thickness [7]. In a nut shell, most of TCOs such as ZnOIn : [8], [9], [10], (ITO) [11], [12],

[13] and [14] are n-type. P-type TCOslike [15], [16 - 18], CuCrO2 [19], CuYO2

[20] are an emerging area with little work reported in the last decade.

Delafossite films have been prepared by a number of vacuum based technologies, such as pulsed laser deposition (PLD) [16, 21], sputtering [22, 23] and electron beam evaporation method [24]. Another method which has been used is sol-gel techniques [25]. Among these techniques, DCmagnetron sputtering is one of the best techniques for pre-paring thin films on large area substrate at low substrate temperature [22].

The electrical conductivity of p-type materials is mainly due to the metal deficiency or oxygen rich defect within the crystalline sites [18, 24]. Thus, excess oxygen

atmosphere on fabrication process can induce p-type con-ductivity [18]. Defect chemistry plays an important role in making the metal oxides either p- or n-type. Nonstoichio-metric defect reaction for a metal deficient oxide may be represented by the following equation [18]:

, (1)

where Oo, VCu, VAl and H denote lattice oxygen, Copper (Cu) vacancy, Aluminum (Al) vacancy and hole, respectively. Superscript x, - and + denote effective neutral, negative and positive charge states, respectively. Similar nonstoichio-metric reaction mechanism as that shown in equation (1) could be found in other Cu-based delafossite structure ma-terials. Thus, subsequent improvement in the electrical conductivity and optical transmittance is possible through controlling the processing condition such as oxygen partial pressure, substrate temperature, sputtering power and working pressure [26].

Delafossite discovery in 1973 by Fridel initiated a lot of interest in these type of materials [16, 17, 21, 22, 24, 27]. Many compounds that share the delafossite structure are all of type where A can be Cu, Ag, Pt and B is Al, Ga, In, Sc, Cr, Fe or rare earth [20].

2. Experimental Procedure CuCrO2 thin films were deposited onto float glass substrates by reactive DC magnetron sputtering using a homemade sputtering system at the Helmholtz Centre for Materials and

1


2013 MSSEESA Conference proceedings ix

Table of Contents

Page No

Welcome note by the Chairman, Local Organizing committee i

Welcome note by MSSEESA Chief Coordinator ii

MSSEESA Coordinating Board iii

MSSEESA Conference Local Organising Committee iv

MSSEESA Conference Proceedings Editorial Committee v

MSSEESA Conference Programme vi

A. Alfred, J. M. Mwabora, R. J. Musembi, S. M. Waita, Th. Dittrich, K. EllmerEffects of Oxygen Partial Pressure and Substrate Temperature on Optical Properties of Sputter Deposited CuCrO2 Thin Films

1

S. Ambonisye, R.T. Kivaisi, M.E. Samiji and N.R. MlyukaLuminous Transmittance and Transition Temperature of VO2:Ce Thin Films Prepared by DC Reactive Magnetron Sputtering

6

A. A. Mulama, J. M. Mwabora, A. O. Oduor, and C. C. MuivaOptical Properties of Flash Evaporated Se100-xBix Thin Films: Effect of Film Thickness

11

H. Wafula, R. J. Musembi, A. Juma, R. Sáez-Araoz, Christian-Herbert Fischer, Th. Dittrich and E. Wendler

Role of Cl on Diffusion of Cu in In2S3 Layers Prepared by Ion Layer Gas Reaction Method

19

B. Samwel, N.R. Mlyuka, M. E. Samiji, and R. T. KivaisiEffects of Target Composition on the Optical Constants of DC Sputtered ZnO:Al Thin Films

23

J. Ndungu, F. W. Nyongesa, A. Ogacho and B.O. AdudaNanoporous Water Filtration Using Disc Ceramic Water Filters

29

P. K. Nyaga, R. J. Musembi and M. K. MunjiEffect of Sn Doping on the Electrical Properties of as Prepared and Annealed ZnO Thin films Prepared byReactive Evaporation

33

L. K. Munguti, R.J. Musembi, W.K. Njoroge 37


2013 MSSEESA Conference proceedings x

Optimization of SnxSey Deposited by Reactive Thermal Evaporation for Solar cell Application

C. J.Lyobha, R.T. Kivaisi, N. R. Mlyuka and M. E. SamijiEffects of Aluminium and Tungsten Co-doping on the Optical Properties of VO2 Based Thin Films

42

M. J. Mageto, C.M. Maghanga, M. Mwamburi, H. JafriTransparent and Conducting TiO2:Nb Thin Films Prepared by Spray Pyrolysis Technique

49

D. Magero, N. W. Makau, G. O. Amollo, S. Lutta, M. D. O. Okoth, J. M . Mwabora , R. J.Musembi, C. M. Maghanga, R. Gateru

Hydrogen as an Alternative Fuel: An ab-initio Study of Lithium Hydride and Magnesium Hydride

61

J. Mutasingwa, J. Buchweishaija, J. E.G. MdoeAloe Secundiflora Extract as a Green Corrosion Inhibitor for Carbon Steel in Potable Water Systems

64

J. N. Nguu, R. J Musembi, F. W. Nyongesa, & B.O. AdudaMorphological and Structural Characterization of TiO2/Nb2O5 Composite Electrode Thin Films Synthesized by Electrophoretic Deposition (EPD) Technique

70

B.V Odari, M.J Mageto, R.J Musembi, F.M Gaitho, V.W Muramba, H. OthienoOptical and Electrical Properties of PTO Thin Films Prepared by Spray Pyrolysis

78

E. R. Ollotu, M.E.Samiji, N.R.Mlyuka, R.T. KivaisiInfluence of Films Thickness on Optical Properties of Nb-Doped TiO2 (NTO) Thin Films Deposited by DC Reactive Magnetron Sputtering

86

C. O. Ayieko, R. J. Musembi, S. M. Waita, B. O. Aduda and P. K. JainImproved Performance of TiO2/In(OH)iSj/Pb(OH)xSy Composite ETA Solar Cell through Nitrogen doped TiO2 Thin Filmwindow layer.

91

G. Samukonga, S. Hatwaambo, G. ChinyamaConstruction and Evaluation of a PV/T Collector Using the Batch Method

98

R. T. Shikambe, M. K. Munji, R. J. Musembi, P. M. MwatheThe effect of surface passivation on optical properties of as-grown and annealed chemically deposited CdxNi1-xS thin films as a material for photovoltaic application

104

M. Tembo, O.s Munyati, S. Hatwaambo, M. MaazaEnhancement of Photo-Absorption in P3HT:PCBM Blends using Squarylium III Dye

118

Energy in Berlin, Germany. The targets used were 7.6 cm diameter and 0.6 cm thick CuCr (50% Cu and 50% Cr)from FHR Anlagenbau GmbH (Gasellschaft mit beschrankter Haftung), Germany. Before sputtering, the substrates (1 × 1 square inch) were cleaned in methanolmixed with in ratio of 1:1 in an ultra-sonic bath, dried with N2 gas and then loaded into the sputtering chamber. The sputtering system was then evacuated to a base pressure of about 6 x 10-7 mbar (6 10-5 Pa). The substrate was heated to the desired temperature up to 287 °C using an halogen lamp. Pre – sputtering for 10 minutes to remove possible contaminants on the target. The distance between the sputter target and the substrate was 66 mm for all depositions. The sum of O2 and Ar flow was set to 50 standard cubic centi-meter per minute (sccm). Oxygen partial pressure was var-ied from 0.153 to 0.187 µbar by varying O2 and Ar flow while keeping the sum O2 and Ar flow constant at 50 sccm. Deposition time for CuCrO2 films was 20 and 5 minutes for variation of Oxygen partial pressure and substrate tempera-ture, respectively. For both films, the sputter plasma was generated at a constant power of 100 W. Transmittance and reflectance data were measured using Varian Cary 5EUV- VIS / NIR spectrophotometer in the range 200 to 3200 nm.

3. RESULTS AND DISCUSSION3.1 Effect of O2 Partial Pressure on Optical Properties of

CuCrO2

3.1.1 TransmittanceFigure 3.1 shows the optical transmittance for CuCrO2 films measured over the wavelength range of 200-2500 nm at various oxygen partial pressures. The deposition temper-ature was fixed at 209 °C for various Po2 which resulted inthicknesses of 414, 495 and 602 nm for films deposited at Po2 of 0.187, 0.170 and 0.153 bar, respectively. One can see that the transmittance was less than 5% in the UV-Visible range of solar spectrum (200-750 nm). Between 750-1250 nm the transmittance increased rapidly for wave-length range to about 50-60%. Thereafter, the increment was gradual to values between 70 and 90%. The observed low optical transmittance of CuCrO2 at short wavelength might be attributed to absorption of ions which is in agreement with the ligand field theory [25].

The transmittance spectra are dominated by interference effects for wavelengths greater than 1200 nm, an observa-tion attributed to large film thickness [9]. Higher transmit-tance of 50-70% for rf-sputtered CuCrO2:N in the visible range has been reported [23]. The inset, in figure 3.1 shows the fitting of sample prepared at 0.153 bar as an example for the model pre-developed in SCOUT software for deter-mining optical constants; the solid represents simulated transmittance while dashed line is for experimental data.

Figure 3.1. Plot of Transmittance versus wavelength (nm) for CuCrO2 films deposited at different O2 partial pressures. Substrate temperature is 209 °C. The inset in fig. 4.1 is an example of CuCrO2 model (dotted line) and measured transmittance (solid line), film were deposited at oxygen partial pressure of 0.153 bar.

3.1.2 Effect of partial pressure on refractive index Dispersion analysis using model for dielectric suscepti-

bility of the film consisting of constant refractive index and Drude model was applied to model the experimental trans-mittance and reflectance spectra. Optical constants were extracted after the best fit between the model and measured transmission spectra in SCOUT. Generally the refractive index n( ) for all O2 partial pressures increased to maxi-mum at 376 nm then decreased gradually (figure 3.2).The peak at 376 nm for various Po2 is associated with resonant absorption by bound charge carries in the vicinity of resonance wavelength. The bound carriers are believed to perform forced oscillation upon influence of an alternating electric field (i.e light); this is in line with Lorentz postulate [28]. For film deposited at oxygen partial pressure of 0.153

bar the refractive index was 2.15 at 200 nm and increases to a maximum of 2.26 at 376 nm then drops gradually with increase in wavelength to a value of 2.05 at 2247 nm. On increasing Po2 to 0.170 μbar, the refractive index was found to shift to lower values with similar pattern. Further in-crease in Po2 to 0.187 μbar resulted in reversal of trend for the refractive index. The varying refractive index with oxy-gen partial pressure can be attributed to variation of the microstructure and defect chemistry of the materials [29].

Figure 3.6. Plot of ln versus photon energy ( h ) for CuCrO2 films deposited at various substrate temperatures. Oxygen partial pressure is fixed at 0.170 bar.

References 1. Li, D., X. Fang, A. Zhao, Z. eng, W. ong, R. Tao. Vacuum,

84 (2010) 851.2. Schaefer, C., G. Brave, C.Szczyrbowsk, Surface and

Coating Technology, 93 (1997) 39.3. Al-Dahoudi, N., H. Bisht, C. Gobbert, T. Krajewski, M.A.

Aegerter. Thin Solid Film, 392 (2001), p. 299.4. Georgi, H.D and M.M. R. Milka, Thin Solid Films, 515

(2007) 8465.5. Ramamoorthy, K., K. Kumar, K. Sankaranayanan. Mate-

rial Science and Engineering B, 126 (2006) 1.6. Yang, C., S. Lee, T. Lin, W. Zhuang, Material Science and

Engineering B, 134 (2006) 68.7. Ali, H. M. Phys.sta.sol, 204 (2005) 2742.8. Song, I., D. Kang, C. Kim, J. Park, Y. Park, R. Martins, P.

Barquinha. Thin Solid Films, 516 (2008), p. 1322.9. Choi, W., K. Jung, S. Yoon, H.J. Kim. J. Choi, Applied

Surface Science, 256, (2010a) 6219.10. Musat, V., B. Tiexeira, , E. Fortunato, R.C.C. Monteiro,

Thin Solid Films, 502 (2006) 219.11. D’Elia, S., N. Scaramuzza, F. Ciuchi, C. Versace, G.

Strangi, R. Bartolino, Applied Surface Science, 255 (2009),p. 7203.

12. Moholkar, A.V., S.M. Pawar, R.Y. Rajpure, C.H. Bhosale, R.H. Kim. Applied Surface Science, 255 (2009), p. 9358.

13. Gassert, T.A., T. Coutts, L. Xiaonan, Applied Surface Sci-ence, 223 (2004), p. 138.

14. Al-Mobarak, M.S., F.M. Amanullah, A.M. Al-Dhafiri, K. M. Al-Shibani, Material Chemistry and Physics, 59 (1999),p. 247.

15. Sato, H., T. Minami, S. Takata, T. Yamada. Thin Solid Films, 236 (1993), p. 27.

16. Kawazoe, H., M. Yasukawa, H. Hyodo, M. Kurita, H. Yanagi, H. Hosono.. Nature, 389 (1997), p. 939.

17. Yanagi, H., S. Inoue, K. Ueda, H. Kawazoe. Journal of Applied Physics, 88 (2000), p. 4159.

18. Banerjee, A.N., R.Maity, K.K. Chattopadhyay. Material Letters, 58 (2003), p. 10.

19. Nagarajan, R., A.D. Draeseke, A.W. Sleight, J. Tate.Journal of Applied physics, 89 (2001), p. 8022.

20. Jayaraj, M.K, A.D. Draeseke, J. Tate, W. Sleight. Thin Solid Films, 397(2001) , p.244.

21. Deng, Z., X. Fang, T. Ruhua, W. Dong, D. Li, Z. Xuebin, Journal of Alloy and Compound, 466 (2008), p. 408.

22. Reddy, A.S., H. Park, G.H. Rao, U. Uthanna, P.S. Reddy.Journal of Alloy and Compound, 474 (2009), p. 401.

23. Dong, G., M. Zhang, X. Zhao, C. Tian, Y. Ren. Applied Surface Science, 256 (2010), p. 4121.

24. Kim, D.S., S.J. Park, H.K. Lee, S.Y. Choi. Thin Solid Films, 515 (2007), p. 5103.

25. G tzend rfer, S., C. Polenzky, U. Stephan, L.Peer , Thin Solid Films, 518 (2009), p. 1153.

26. Banerjee, A.N., C.K.Ghosh, K.K. Chatopadhyay. Solar Energy Materials and Solar Cells, 89 (2005a), p. 75.

27. Marquardt, M.A., N.A. Ashmore, D.P. Cann, Thin Solid Films, 496 (2006), p. 146.

28. Hummel R.E. Electronic properties of materials, 3rd edi-tion, Springer-Verlage, , Berlin 1993.

29. Cui, H., V. Teixeira, L. Meng, R. Martin30. Fox, M (2001). Optical properties of solid, Oxford Uni-

versity Press, New York NY.E. Fortunato, Vacuum, 82 (2008), p. 1507.

31. Al-Kuhaili, F and S.M.A. Durreni. Optical Materials, 29(2007), p. 709.

32. Arushanov, E., S. Lovcenko, N. Syrbu, V. Tezlevan, M. Merino, M. Leon. Phys. Sta. Sol (a), 203 (2006), p. 2909.

33. Yakuphanoglu, F., M. Sekerci, O.F. Ozturk. Optics Communication, 239 (2004), p. 275.

34. O’Leary, S.K., S.R. Jonson, P.K. Lim. Journal of Applied Physics. 82 (1997), p. 3334.

35. Lim, S.H., S. Desu, A.C. Rastogi, Journal of physics and Chemistry of solids, 69 (2008) 2047.

1.6 1.8 2.0 2.2 2.4 2.6

14.4

14.6

14.8

15.0

15.2

15.4

15.6

15.8

16.0

ln(

)

Photon energy (eV)

263 °C 209 °C 143 °C as-grown

Po2 = 0.170 bar

5 2

Figure 3.2. Plot of refractive index and extinction coefficient versus wavelength for CuCrO2 deposited at different O2 partial pressure, deposition temperature is fixed at 209 °C.

The extinction coefficient, (k), decreases with increase in O2 partial pressure in the visible range of the solar spec-trum, while in the NIR the pattern is reversed (see figure3.2). The extinction coefficient drops to near zero in theNIR. This is associated with Lorentz oscillation of charge carriers, that have absorption strongly peaked at resonance wavelength and falls of as a 2 as one tunes away from resonance wavelength [30]. Thus, there is no significant absorption as one tunes far from resonance (see figure 3.2). It can be seen that the extinction coefficient k was found to be dependent on O2 partial pressure. The difference in k was notably large in the visible region and decreased as wave-length increases to the near infrared region and then far infrared. This could be ascribed to defect chemistry caused by oxygen vacancies [31]. The large values of extinction coefficients in the visible range of solar spectrum suggest that CuCrO2 prepared under the above sputtering conditions could be applied as a solar absorber material.

3.1.3 Effect of O2 partial pressure on Urbach energy In an amorphous semiconductor, the Urbach energy, Eu,

(also called absorption tail) is the width of the tail of local-ized states in the forbidden band arising due to smearing out of the valence and conduction band edges into the forbidden gap. Having determined the absorption co-efficient at par-ticular wavelength, we used it to obtain the Urbach’s energy from Urbach’s rule [32]:

(2)

Where and are material parameter and Urbach’s energy, respectively. From equation (2) a plot of

))(ln( h against gives a straight line with the inter-cept on the axis giving the material parameter .Urbach energy characterizes the slope of the exponen-tial-edge region and the inverse of the slope gives the width

of the localized state. The plot of ln against h is linear near the fundamental absorption edge thus absorption coefficient near the fundamental absorption edge is expo-nentially dependent on the photon energy and obey the Urbach rule [33, 34]. Figure 3.3 shows the plot of ln(α) vs photon energy ( ) for CuCrO2 thin films deposited at dif-ferent O2 partial pressure. The values of Urbach energy for films deposited at Po2 0.153, 0.170 and 0.187 μbar were found to be 0.27, 0.32 and 0.31 eV, respectively. Generally the values of Urbach energy increases with oxygen partial pressure. This implies that the more the oxygen content the more the total number of localized tail states and the distri-bution of tail states further encroaches into the Tauc gap

[34].

Figure 3.3. Plot of the variation of ln( ) versus energy ( ) for CuCrO2

thin film deposited at different O2 partial pressure. Substrate temperature is fixed at 209 C.

3.2 Effect of Substrate Temperature on Optical Proper-ties of CuCrO2 Films

3.2.1 TransmittanceFigure 3.4 depicts the variation of transmission spectra

with substrate temperature; where approximately 10-30% transparency in visible region (400-700 nm) was obtained as temperature increased from ambient (as-grown) to 263 °C. Generally transmittance increases with substrate temperatures. The maximum transmittance is about 65% (at

= 2247 nm for deposition temperature of 263 °C). The optical absorption edge of the films shifted toward lower wavelengths side with the increase of substrate tempera-tures. The shift of absorption edge and increase in optical transmittance with substrate temperature may be attributed to reduction in defects level and improvement in film crys-tallinity with increasing substrate temperature [35]. Higher transmittance of 50-70% for rf-sputtered NCuCrO :2 in the visible range was reported [23]. The inset [figure 3.4]shows an example of the optical transmittance spectra ex-perimental data fitted to the optical model pre-developed for dielectric function in SCOUT software which was used in determination of optical constants.

0 500 1000 1500 2000 25000.0

0.5

1.0

1.5

2.0

2.5

Extin

ctio

n co

effic

ient

(k)

R

efra

ctiv

e in

dex

(n)

Wavelength (nm)

n() at Po2 = 0.153 bar k() at Po2 = 0.153 bar n() at Po2 = 0.170 bar k() at Po2 = 0.170 bar n() at Po2 = 0.187 bar k() at Po2 = 0.187 bar

Ts = 209 °C

0.4 0.6 0.8 1.0 1.2 1.4 1.66.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

10.5

ln(

)

Photon energy (h

0.153 bar 0.170 bar 0.187 bar

Ts = 209 °C

)exp()( 0uE

hh (2)

Figure 3.4. Plot of transmittance versus wavelength (nm) for CuCrO2 thin films deposited at different substrate temperature. Oxygen partial is fixed at 0.170 bar.

3.2.2 Refractive index The spectral dispersion of the real and imaginary part

of refractive index )(n and for CuCrO2 thin films formed at different substrate temperatures is shown in fig. 3.5. The values of real refractive index at 263 °C and 143 °C overlap whereas at 209 °C the value of n was low ( as compared to other temperatures. FromFfigure.3.5 the difference in refractive index for the films prepared at 263 °C and 209 °C was very small. The refractive index values of about 2.08, 2.15, 2.06 and 2.14 for as-grown temperature, 143, 209 and 263 °C, respectively, were ob-tained as shown in Figure 3.5. Lower refractive index of the order of 1.29, for pulsed laser deposited CuCrO2 film have been reported [1].

On the other hand, the variation of extinction coeffi-cient )(k with substrate temperatures (Figure 3.5) in thevisible region is not much pronounced, but as you enter the NIR region there is a trend where the extinction co-efficient were found to increase with decrease in substrate tempera-ture. The increase of k for the as-grown and 143 °C film in the NIR region indicates the presence of more defects in these films compared to those films prepared at 209 °C and 263 °C. These defects may be attributed to oxygen vacan-cies [34]. However, the as-grown and 143 °C films are less deficient in oxygen, but totally amorphous, and thus contain more structural defects [25]. Typical extinction coefficient for as-grown film is seen to decrease exponentially from 0.8 at 200 nm to 0.3 at 2247 nm as wavelength increased

3.2.3 The Urbach energy. The Urbach energy for CuCrO2 thin films deposited at

different substrate temperatures was obtained from the plot of against versus photon energy ( ). By taking the inverse slope of the linear region Figure 3.6, the Urbach energy for films deposited at 263, 209, 143 °C and as-grown were found to be 0.819, 0.934, 1.19 and 1.45 eV, respectively.

The Urbachenergy increases with decrease in deposition temperature. This is ascribed to improvement in film quality by reducing

Figure 3.5. Plot of refractive index and extinction coefficient versus wavelength for CuCrO2 thin films deposited at different substrate temper-atures. Oxygen partial pressure is fixed at 0.170 bar.

the structural disorder [33]. Thus, the degree of disorder was decreasing with increase in substrate temperature. As-grown CuCrO2 thin films are normally amorphous [25], hence, more disordered than films deposited at elevated tempera-ture.

4. Conclusion The CuCrO2 thin films prepared at various oxygen par-

tial pressures were highly absorbing and opaque in the UV-Visible range of the solar spectrum. This is reflected in high values of refractive indices (n of the order of 2.2) and extinction coefficients, 0.5. The Urbach energy has been found to increase from 0.277 to 0.31 eV with increasing of oxygen partial pressure from 0.153 to 0.187 bar.The optical transmittance of CuCrO2 films has been found to increase with substrate temperature up to 38% in the visible range (400-700 nm). On the other hand, the Urbach energy increased from 0.81 (at 263 C) to 1.47 eV (at as-grown temperature),- an observation attributed to amorphous nature of the film deposited at low temperatures. The study showed that CuCrO2 films can be applied in shielding infrared sensor from electromagnetic interference, selective surface coating materials and anti-reflection coating in eye glass for night vision.

0 500 1000 1500 2000 2500

0.0

0.5

1.0

1.5

2.0

2.5

Ext

inct

ion

coef

fici

ent(

k)

R

efra

ctiv

e in

dex(

n)

Wavelength (nm)

n at 263 °C n at 209 °C n at 143 °C k at 263 °C k at 209 °C k at 143 °C n as-grown k as-grown

Po2 = 0.170 bar

0 500 1000 1500 2000 25000.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Tran

smitt

ance

Wavelength (nm)

Measured at 263 °C simulated at 263 °C

Po2 = 0.170 bar

0 500 1000 1500 2000 25000

10

20

30

40

50

60

70

80

90

100

Tran

smitt

ance

(%)

Wavelength (nm)

263 °C 209 °C 143 °C As-grown

Po2 = 0.170 bar

3 4

Luminous Transmittance and Transition Temperature of VO2:Ce Thin Films Prepared by DC Reactive Magnetron

Sputtering

S. Ambonisye1, R.T. Kivaisi2, M.E. Samiji and N.R. Mlyuka2

1 Physics Unit, Faculty of Science, Dar es Salaam University College of Education, P.O. Box 2329, Dar es Salaam, Tanzania.2Solar Energy Group, Physics Department, University of Dar es Salaam, P. O. Box 35063 Dar es Salaam, Tanzania.

AbstractVanadium dioxide (VO2) and cerium-doped vanadium dioxide (VO2:Ce) films were prepared by DC reactive magnetron sputtering from vanadium and V-Ce alloy targets. The Shimadzu SolidSpec-3700 DUV-VIS-NIR spectrophotometer was used to measure the transmittance and reflectance of the films. The phase transition temperature (c) of the films was ob-tained from both the transmittance-vs-temperature and sheet resistance-vs-temperature curves. A change in sheet resistance of 2 to 3 orders of magnitude was observed for both undoped and Ce-doped VO2 films. Comparison between undoped and doped VO2 films revealed that cerium inclusion alters both the visible and infrared transmittance of VO2 thin films. Luminous transmittance was slightly enhanced while the c was slightly depressed by cerium inclusion in VO2. A two-step increase in transmittance observed in the cooling loop in pure VO2 was found to be suppressed by cerium inclusion.

Keywords: Vanadium Dioxide, Luminous Transmittance, Transition Temperature

1. Introduction Discovery of novel behavior of vanadium dioxide where it undergoes a metal-to-insulator phase transition at a temper-ature (τc) of ~68 oC accompanied by dramatic changes in electrical and optical properties has attracted intensive re-search in this thermochromic material. The phase transition in VO2 has been tailored to suit various applications in de-vices such as smart windows, thermal sensors and switching devices [1]. However, the transition temperature is too high for smart window applications and it is therefore desirable for it to be reduced. Many studies have been conducted to lower τc to the vicinity of room temperature especially through doping with foreign atoms such as tungsten, mo-lybdenum, niobium and fluorine. Although tungsten (W) doping has shown incredible reduction in τc to room tem-perature [2], W-doped VO2 films are reported to have lower infrared transmittance at room temperature compared with the undoped films [3] and hence unsuitable for high per-formance optical switching applications.

Several techniques have been reported for depositing VO2-based thin films including, sputtering [4], reactive evaporation [5], chemical vapor deposition (CVD) [6], pulsed laser deposition (PLD) [7], molecular beam epitaxy (MBE) [8] and sol-gel process [9]. Despite the being ex-pensive, sputtering offers several advantages such as the

ability to produce uniform films and efficient deposition [10]. It is also one of the most promising techniques for large-area and large-scale coating applications [11]. This paper reports the luminous transmittance and τc of Ce-doped VO2 (VO2:Ce) thin films.

2. Experimental Details VO2 and VO2:Ce thin films were prepared by reactive DC magnetron sputtering of metallic V target and V-Ce alloy targets (99.9% purity) using the BALZERS BAE 250 coat-ing unit. Both VO2 and VO2:Ce thin films were deposited on normal microscope quartz glass substrates placed ~16 cm above the target. The alloy targets had varying percentage compositions of vanadium and cerium as V99%-Ce1%, V98%-Ce2% and V96%-Ce4%. Three sets of VO2:Ce films were prepared and designated as VO2:Ce1, VO2:Ce2 andVO2:Ce4, from V-Ce (1% Ce), V-Ce (2% Ce) and V-Ce (4% Ce) alloy targets, respectively. Prior to the deposition, the vacuum chamber was evacuated down to a base pressure of ~3.0 x10-5 mbar. The sputtering gas (Ar of 99.999% purity) and reactive gas (O2 of 99.999% purity) were introduced in the chamber at the rates of 75 and 4.4–4.7 mln/min respec-tively. The working pressure was maintained at about 3.8 x10-3 mbar and the sputtering power was 200 watts. The substrate temperature at which the films were deposited was

miss-match, intrinsic stress, presence of second phase (de-viation from stoichiometry) or addition of impurities (dop-ing). It was suggested in this case that inclusion of cerium induced a change in the V-O and V-V distances of the VO2

structure due to the differences in ionic radii between Ce3+

(0.102 nm) and V4+ (0.059 nm) . This caused phase transi-tions at lower temperatures than that for undoped VO2 thin films as it was also reported by Lu et al. [19]. This conforms to the findings by Kiri et al. [10] who reported that dopants with atomic radii larger than the V4+ ion cause a decrease in c whereas those with smaller ionic radii increase c. Table 1 summarises major properties of VO2 and VO2:Ce thin films.

VO2 VO2:Ce1 VO2:Ce2 VO2:Ce460

62

64

66

68

70

Phas

e Tr

ansit

ion

Tem

pera

ture

(o C)

Sample

SDUVs TPP Jandel

Figure 6. Transition temperature for VO2 and VO2:Ce thin films obtained from transmittance-vs-temperature and sheet resistance-vs-temperature curves. SDUVs = Shimadzu SolidSpec-3700 DUV spectrophotometer, TPP = two-point probe, Jandel = JANDEL four-point probe unit.

Table 1. Summary of the properties of undoped and Ce-doped VO2 thin films

Conclusion VO2 and cerium-doped VO2 thin films have been success-fully prepared by reactive DC magnetron sputtering. The

films had sheet resistance switching by 2 to 3 orders of magnitude. The maximum transmittances in the visible re-gion, ranging from 34 - 40% and 26 - 34% were observed for < c and > c respectively. The integrated luminous transmittance (Tlum) increased as the result of cerium doping in VO2. The integrated solar transmittance (Tsol) of the VO2:Ce films increased as cerium concentration increased. The hysteresis loop widths decreased as the result of cerium inclusion in VO2. A two-step increase in transmittance ob-served in the cooling loop in VO2 was suppressed by cerium doping. On the other hand, cerium doping in VO2 was ob-served to decrease its transition temperature.

AcknowledgementOne of the authors, Ambonisye, S., wishes to acknowledge the Dar es Salaam University College of Education (DUCE) for sponsoring the study. The partial sponsorship from the Solar Energy Group in the Department of Physics (UDSM) by which the deposition materials and equipment were ac-cessed is also acknowledged.

References [1] Miyazaki, H. and Yasui, I., J. Phys. D: Appl. Phys. 39,

(2006), pp. 2220-2223

[2] Batista, C., Ribeiro, R.M. and Teixeira, V., Nanoscale Re-search Letters, (2011), pp. 1-7

[3] Wang, H., Yi, X. and Li, Y., Optics Communications 256,(2005), pp. 305-309

[4] Kivaisi, R.T. and Samiji, M., Solar Energy Materials & Solar Cells 57, (1999), pp. 141-152

[5] Golan, G., Axelevitch, A., Sigalov, B. and Gorenstein, B., J. Optoelectronics and Advanced Mater. 6, (1), (2004), pp. 189-195

[6] Manning, T.D., Parkin, I.P., Blackman, C. and Qureshi, U., J. Mater. Chem. 15, (2005), pp. 4560-4566

[7] Lappalainen, J., Heinilehto, S., Jantunen, H. and Lantto, V., J. Electroceram 22, (2009), pp. 73-77

[8] Rata, A.D., Chezan, A.R., Presura, C. and Hibma, T., Surface Science 532–535, (2003), pp. 341-345

[9] Lu, S., Hou, L. and Gan, F., Thin Solid Films 353, (1999), pp. 40-44

[10] Kiri, P., Hyett, G. and Binions, R., Adv. Mat. Lett. 1 (2), (2010), pp. 86-105

[11] Sobhan, M.A., Kivaisi, R.T., Stjerna, B. and Granqvist, C.G., Solar Energy Materials & Solar Cells 44, (1996), pp. 451-455

[12] Béteille, F., Morineau, R., Livage, J. and Nagano, M., Mate-rials Research Bulletin 32, (8), (1997), pp. 1109-1117

[13] Jin, P. and Tanemura, S., Jpn. J. Appl. Phys. 34, (1995), pp.

9 6

~420 oC. Deposition rate was about 3.6 nm/min for VO2

films and about 3.4 nm/min for VO2:Ce films.

Electrical measurements were performed using a two-point probe and the JANDEL model RM3-AR test unit in combi-nation with the four-point probe. In order to determine temperature dependence of sheet resistance the films were placed on a temperature regulated hotplate whose tempera-ture was varied within the interval 25<Te<100 oC. The heating and cooling rates were ~10 oC/min and ~0.5 oC/min, respectively near 25 oC , and ~7 oC/min and ~5 oC/min, respectively near 100 oC. For optical characterization, the Shimadzu SolidSpec-3700 DUV-VIS-NIR was used. A heating cell was used to change the film surface temperature between ~25 oC and ~100 oC when measuring the transmit-tance of the films as a function of temperature. The rate of heating was ~10 oC/min near room temperature and 6oC/min near 100 oC while the cooling rate was ~5 oC/min near 100 oC and ~0.8 oC/min near 25 oC. Integrated luminous (lum) and solar (sol) transmittance values were obtained from the equation,

( )

( )

( )

, ,

,

b

lum solalum sol b

lum sola

T dT

d

(1)

where the integral is evaluated from a = 0.38 to b = 0.76 mfor luminous transmittance and from a = 0.25 to b = 2.5 mfor solar transmittance, lum is the spectral sensitivity of the light-adapted human eye, and sol is the solar irradiance spectrum ASTM G-173 for air mass 1.5 (corresponding to the sun standing 37o above the horizon).

3. Results and Discussion 3.1. Electrical Properties of VO2 and VO2:Ce Thin Films

Both VO2 and VO2:Ce thin films exhibited sheet resistance change of about three orders of magnitude across the phase transition. This is a typical behavior for VO2 films deposited on amorphous substrates [12]. The variation of sheet re-sistance with temperature for undoped and cerium-doped VO2 thin films measured using a two-point probe are shown in Figure 1.

20 40 60 80 100

100

1000

10000

100000

Shee

t Res

istan

ce (

sq)

Temperature (oC)

VO2

VO2:Ce1

VO2:Ce2

VO2:Ce4

Figure 1. Variation of sheet resistance with temperature for VO2,VO2:Ce1, VO2:Ce2 and VO2:Ce4 thin films (100 nm thick, deposited at 420 oC) measured using a two-point probe.

A two-step decrease in sheet resistance (i.e. with two points of inflexion) was observed in the heating loop for all the films, an observation that was also reported by Kivaisi and Samiji [4]. The sheet resistance determined by the JANDEL four point probe however did not register the step. The drop in sheet resistance exhibited by all the films was not very sharp (Figure 1). This could be caused by the presence of large number of microscopic crystals each characterized by its own c and hysteresis loop, whose totality blurred and broadened the switching. The figure also shows a decrease in both the transition temperature and the hysteresis loop width with increasing cerium concentration. The loop widths were 20.31 oC, 13.43 oC, 16.68 oC and 12.96 oC for VO2, VO2:Ce1, VO2:Ce2 and VO2:Ce4 thin films, respectively. Jin and Tanemura [13] attributed such a change in hysteresis loop width to improvement of crystallinity and microstructure. It also appeared that increase in Ce content made the high temperature phase of VO2 less conducting. Similar behavior was observed by Béteille et al. [12] for titanium-doped VO2

thin films. For VO2, VO2:Ce1 and VO2:Ce2 films, the tem-peratures at which the phase change occurred took a de-creasing order. However, it appeared to go up again for VO2:Ce4 film.

3.2. Optical Properties of VO2 and VO2:Ce Thin Films

3.2.1. Luminous Transmittance

The inclusion of cerium in VO2 influenced changes in lu-minous transmittance of VO2 films. While the low temper-ature phase of undoped VO2 had a maximum transmittance of about 34% at 710 nm, it was 40.6% at ~651 nm for VO2:Ce1, 35.6% at ~650 nm for VO2:Ce2 and 38.7% at ~638 nm for VO2:Ce4.

500 1000 1500 2000 25000

10

20

30

40

50

60

70

~100 oC

% T

rans

mitt

ance

Wavelength (nm)

VO2

VO2:Ce1

VO2:Ce2

VO2:Ce4

~25 oC

Figure 2. Spectral transmittance for single VO2, VO2:Ce1, VO2:Ce2 and VO2:Ce4 thin films (100 nm thick, deposited at 420 oC

Small transmittance values for the films in the visible region were considered to originate from the strong intra-band and inter-band absorption in the short-wavelength range for both the metallic and semiconductive phases as cited by Zhang et al. [14]. The increase in luminous transmittance for Ce-doped VO2 was thought to be attributed to replacement of V4+ by Ce3+ which resulted into oxygen vacancies. The ob-served slight shift of the maximum transmittance in the vis-

ible region to lower wavelengths justified the small change in color of the films. There was a significant modulation in visible transmittance for low and high temperature phases for both undoped and Ce-doped VO2 thin films owing to free career concentration differences. The high temperature phase of undoped VO2 had a maximum transmittance of about 26.5% at 663 nm while it was 33% at 643 nm for VO2:Ce1, 29.3 % at 642 nm for VO2:Ce2 and 34% at 650 nm for VO2:Ce4 (Figure 2).

Integrated luminous transmittance (Tlum) values (at ~25 oC) computed using equation 1 revealed a Tlum of 18.9%, 27.9%, 23.9% and 28% for VO2, VO2:Ce1, VO2:Ce2 and VO2:Ce4 respectively. For the high temperature phase (at ~100 oC), the Tlum values were 17.3%, 24.3%, 21.4% and 24.5% for VO2, VO2:Ce1, VO2:Ce2 and VO2:Ce4 respectively. The differences in Tlum values for the low- and high-temperature phases of VO2 compare with the results by Xu et al. [15] for 50 nm VO2 films on quartz glass. The Tlum values are shown in Table 1 and the general trend is presented in Figure 3. It could generally be suggested from Figure 3 that the inclusion of cerium increased the visible transmittance of VO2 thin films.


16

18

20

22

24

26

28

30

Tlum for Te < C

Tlum for Te > C

% T

lum

Sample

Figure 3. Integrated luminous transmittance (Tlum) for VO2, VO2:Ce1, VO2:Ce2 and VO2:Ce4 thin films (100 nm thick, deposited at 420 oC)

3.2.2. Solar and Infrared Transmittance

Integrated solar transmittance (Tsol) values at ~25 oC were respectively, 26.5%, 29.2%, 26.7% and 27.9% for VO2,VO2:Ce1, VO2:Ce2 and VO2:Ce4 thin films. Comparatively, Khan and Granqvist [16] obtained a Tsol = 31% for a 75 nm thick VO2 thin film at 25 oC. In this study, Tsol values at ~100 oC were 13.7%, 18.6%, 16.2% and 18.5% for VO2, VO2:Ce1, VO2:Ce2 and VO2:Ce4 thin films respectively. Tsol modula-tion shown in Figure 4 indicates that inclusion of cerium in VO2 slightly decreased its solar modulation.

As for the near infrared (NIR) transmittance, data extracted from Figure 2 gives transmittance contrasts of 51.4% for VO2, 54.3% for VO2:Ce1, 49.5% for VO2:Ce2 and 53.3% for VO2:Ce4 all at = 2500 nm. From Figure 2, VO2:Ce1 is observed to be the most transmitting at 2500 nm and has the highest NIR transmittance contrast for the Ce concentration values studied. These data are supported by the transmit-tance-versus-temperature curves at a wavelength of about

2500 nm shown in Figure 5. The infrared transmittance of the films dropped drastically to values below 3%, an indica-tion that the reflectivities of these films at 2500 nm were quite high as reported by Béteille et al. [12].


10

11

12

Tso

l Mod

ulat

ion

Sample

Figure 4. Integrated solar transmittance (Tsol) modulation for VO2,VO2:Ce1, VO2:Ce2 and VO2:Ce4 thin films (100 nm thick, deposited at 420 oC)

20 30 40 50 60 70 80 90 1000

10

20

30

40

50

60

% T

rans

mitt

ance

at 2

500

nm

Temperature (oC)

VO2

VO2:Ce1

VO2:Ce2

VO2:Ce4

Figure 5. NIR switching and hysteresis loops in VO2 and VO2:Ce thin films (100 nm thick, deposited at 420 oC) obtained using the Shimadzu SolidSpec-3700 DUV-VIS-NIR spectrophotometer

The transmittance hysteresis loop widths for VO2 films ap-peared to decrease with increase in cerium concentration. Atwo-step increase in transmittance was observed in the cooling loop in pure VO2. This could be due to the formation of an intermediate metastable phase called M2, between the usual monoclinic (M1) and rutile (R) phases [18]. However, the two-step increase in transmittance appeared to be sup-pressed by cerium inclusion.

3.2.3. Effect of Cerium Doping on Phase Transition Temperature of VO2 Thin Films

From the sheet resistance versus temperature curves for VO2

and VO2:Ce samples, it was noted that dramatic changes in sheet resistance for VO2 occurred at a temperature, c69 oC. This value is close to that of the bulk VO2 (c68 oC). The transition temperatures for VO2:Ce films were about 65 oC, 63 oC and 66 oC for VO2:Ce1, VO2:Ce2 and VO2:Ce4 thin films respectively. The VO2:Ce thin films had the c sslightly lower than that of undoped VO2 thin films (Figure 6). The reduction in c can be due to thermal expansion

7 8

2459-2460

[14] Zhang, Z., Gao, Y., Chen, Z., Du, J., Cao, C., Kang, L. and Luo, H., Langmuir 26, (13), (2010), pp.10738-10744

[15] Xu, G., Jin, P., Tazawa, M. and Yoshimura, K., Jpn. J. Appl. Phys. 43, (2004), pp. 186-187

[16] Khan, K.A. and Granqvist, C.G., Appl. Phys. Lett. 55, (1989), pp. 4-6

[17] Maaza, M., Bouziane, K., Maritz, J., McLachlan, D.S. Swa-nepool, R., Frigerio, J.M. and Every, M., Optical Materials 15, (2000), pp. 41-45

[18] Béteille, F. and Livage, J., J. Sol-Gel Sc. Tech. 13, (1998), pp. 915-921

[19] Lu, S., Hou, L. and Gan, F., Journal of Mater. Science Let-ters 15, (1996), pp. 856-857

3.1.3. Absorption Coefficient and Extinction Coefficient of the as-Deposited Thin FilmsThe optical absorption coefficient from figure 3 increases with an increase in photon energy and with increase in bis-muth content. For low energies, corresponding to high wavelengths, thin film interference effects are observed from figure 3. This may be due to the overlaying of light that is reflected on both sides of the as-deposited thin films ofSe100-xBix [15] .

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

2

4

6

8

10

12

14

Abso

rptio

n coe

fficien

t (×1

04 cm-1)

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 200nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

2

4

6

8

10

12

14

16

Abso

rption

coeff

icien

t (×10

4 cm-1 )

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 300nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

2

4

6

8

10

12

14

16

Abso

rption

coeff

icient

(×104 cm

-1 )

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 400nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

2

4

6

8

10

12

14

Abso

rption

coeff

ficien

t (×10

4 cm-1 )

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 500nm

Figure 3 (a-d). Absorption coefficient against photon energy (eV) for

Se100-xBix thin films at different thicknesses

Increase in extinction coefficient values with increase in bismuth concentration has been observed from figure 4.

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60.0

0.1

0.2

0.3

0.4

0.5

0.6

Extin

ctio

n co

effic

ient

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 200nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60.0

0.1

0.2

0.3

0.4

0.5

0.6

Extin

ction

coef

ficien

t

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 300nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60.0

0.1

0.2

0.3

0.4

0.5

0.6

Extin

ctio

n co

effic

ient

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 400nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60.0

0.1

0.2

0.3

0.4

0.5

0.6

Extin

ction

coef

ficien

t

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 500nm

Figure 4 (a-d). Extinction coefficient against photon energy (eV) for


3.1.4. Refractive Index and Real Part of Dielectric Con-stant of the as-Deposited Thin FilmsThe method used to determine the refractive index in this study was the Swanepoel envelope method [16]. The first step was to calculate the maximum and minimum transmit-

tance envelop functions, MT and mT , respectively. From

these functions, n can be found;

a

b

c

d

c

b

a

d

13 10

Optical Properties of Flash Evaporated Se100-xBix Thin Films: Effect of Film Thickness

A. A. Mulama1, 2*, J. M. Mwabora1, A. O. Oduor2, and C. C. Muiva3

1 Department of Physics, University of Nairobi, Nairobi, 00100, Kenya2 Department of Physics and Materials Science, Maseno University, Maseno, 40105, Kenya

3 Department of Physics, Botswana International University of Science and Technology, Botswana, 00741, Botswana* [email protected]

Abstract Amorphous thin films of Se100-xBix (x = 0.0, 1.5, 3.0, 4.5 and 6.0 at. %) deposited by flash evaporation technique, have been investigated in the wavelength range of 500nm-1000nm. It is found that the effect of increasing bismuth content on the as-deposited films led to increase in the absorption coefficient, reflectance, real and imaginary parts of dielectric constant while transmittance and optical band gap energy decreased. On the other hand, reflectance, absorption coefficient, extinc-tion coefficient, refractive index, band gap energy, real and imaginary parts of dielectric constant increased with increase in film thickness but transmittance decreased.

Keywords: Amorphous S100-xBix thin film, Optical property, Film thickness

1. Introduction Selenium is one of the chalcogen elements of group VI of the periodic table. These chalcogen elements combine with other elements like germanium (Ge), arsenide (As), bismuth (Bi), to form chalcogenide glasses [1]. These glasses have low phonon energies, are highly transparent from the visible to the mid- infrared wavelengths, high refractive indices and are optically highly nonlinear. [2]. Research on chalcogenide glasses has received little attention due to competition from silicon technology. The current high demand for sili-con-based devices has made chalcogenide glasses receive attention especially in optical switching devices [3].

Amorphous selenium semiconducting glass exhibits unique property of reversible transformation making it an attractive chalcogenide material for use in threshold switching and optical memory devices [4]. However, it has a short life span and is less sensitive to electromagnetic radiation. These problems have been overcome by alloying selenium with other elements like Ge, Te, Bi, and Sb, which improves the sensitivity, crystallization temperature and reduces aging effects of the amorphous selenium [5]. Selenium-bismuth glasses at higher concentration of bismuth reverses from p to n type [6].

Recently, considerable attention has been focussed on glasses of selenium-bismuth system as these materials have

found application in optoelectronics owing to their electrical, optical, dielectric and thermal properties. Work on the elec-trical [7, 8], dielectric [9], density of defect states [10, 11] and glass transition kinetics [12, 13], of selenium-bismuth alloy has been done. However, less work has been done on the optical characterization of this alloy which is important in commercial applications [14]. Further, limited work is reported on the effect of film thickness on the optical prop-erties of the as deposited thin films. Therefore, this focuses on the effect of film thickness and bismuth content on the optical properties of as-deposited, flash evaporated thin films of Se100-xBix (x = 0.0, 1.5, 3.0, 4.5 and 6.0 at. %).

2. Experimental Details The bulk samples of Se100-xBix (x = 0.0, 1.5, 3.0, 4.5 and 6.0 at. %) were prepared by melt quenching method. The ap-propriate amounts of high purity of selenium and bismuth (99.999%) were weighed according to their atomic per-centages in powder form on an electronic balance (LIBROR AEG-120, Japan), were sealed in quartz ampoules vacuumed to 6.0 10-5mbar (Edwards AUTO 306 Vacuum system, UK). The ampoules were heated in a rotating furnace to ensure a homogeneous mixture to a maximum of 800oC for 12 hours. The temperature was raised slowly at a rate of 3oC per minute. Subsequently, the ampoules were quenched in ice-cooled water. The ampoules were broken to obtain the

solid alloy that was crushed into powder for flash evapora-tion.

Thickness of the thin films has been measured by comput-erized KLA-Tencor Alpha-Step IQ surface profiler with a resolution of 10nm (KLA-Tencor Corporation, USA). The amorphous nature of the deposited thin films was confirmed by X-Ray Diffraction method (Phillips PW3710, UK: CuKα

radiation, λ = 1.5406Å). Thin films of glassy seleni-um-bismuth alloys were prepared by flash evaporation technique at a vacuum of 3.0 10-5mbar (Edwards AUTO 306 Vacuum system, UK). Optical study was done by Sol-idSpec.3700 DUV Spectrophotometer (Solidspec. 3700 DUV Spectrophotometer, Japan) in the spectral range from 500nm to 1000nm. The optical band gap energy was evalu-ated from the plot of (αhν)1/2 versus photon energy

3. Results and Discussion 3.1. Results The thicknesses of the deposited films were; 200±10nm, 300±10nm, 400±10nm and 500±10nm. These films were amorphous, homogeneous and uniform.

3.1.1. Nature of the as-Deposited Thin FilmsThe recorded X-Ray Diffraction (XRD) patterns for the studied as-prepared Se-Bi films as shown in figure 1 are characterized by the absence of diffraction peaks indicating amorphous and glassy nature of the films. The humps are due to the glass substrate [15].

0 10 20 30 40 50 600

100

200

300

400

500

600

Coun

ts pe

r sec

ond

Diffraction angle, (2

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0 : Blank slide

Figure 1. X- Ray Diffraction patterns for 500nm Se100-xBix thin films.

3.1.2. Transmittance and Reflectance of the as-Deposited Thin Films

Transmittance (%) and Reflectance (%) against wave-length (nm) for Se100-xBix thin films at different thicknesses is shown in figure 2.

500 600 700 800 900 10000

20

40

60

80

100

Tran

smitt

ance

/ R

efle

ctan

ce (%

)

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 200nm

T

R

500 600 700 800 900 10000

20

40

60

80

100

Tran

smitt

ance

/ R

efle

ctan

ce (%

)

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 300nm

T

R

500 600 700 800 900 10000

20

40

60

80

100

Tran

smitt

ance

/ R

efle

ctan

ce (%

)

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 400nm

T

R

500 600 700 800 900 10000

20

40

60

80

100

Tran

smitta

nce

/ Ref

lecta

nce

(%)

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 500nm

T

R

Figure 2(a-d). Transmittance (%) and Reflectance (%) against wavelength

(nm) for Se100-xBix thin films at different thicknessesThere were interfer-

ence patterns in the transmittance and reflectance spectra. The transmittance

was zero at 500nm. The reflectance values increased and decreased at

specific wavelengths as bismuth concentration increased due to effect of

transmission of light in the as-deposited thin films.

a

b

c

d

11 12

2/12/122 )( SNNn (1)

Where 2

122

S

TTTTSNmM

mM (2)

S (1.52) is the refractive index of the glass substrate

Real )( 1 and imaginary )( 2 parts of dielectric

constant are given by [17];

nkkn 2; 222

1 (3)

Figure 5 shows that the refractive index decreases with in-crease in the wavelength. In addition, the refractive index increases with increase in the bismuth concentration.

600 700 800 900 1000

2.7

2.8

2.9

3.0

3.1

3.2

Refra

ctive

inde

x

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 200nm

600 700 800 900 1000

2.6

2.7

2.8

2.9

3.0

3.1

3.2

3.3

3.4

Refra

ctive

inde

x

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 300nm

600 700 800 900 1000

2.6

2.7

2.8

2.9

3.0

3.1

3.2

3.3

Refra

ctive

inde

x

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 400nm

600 700 800 900 1000

2.6

2.7

2.8

2.9

3.0

3.1

3.2

3.3

3.4

Refra

ctive

inde

x

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 500nm

Figure 5(a-d). Refractive index against wavelength (nm) for Se100-xBix thin

films at different thicknesses

The real part of the dielectric constant decreased with in-crease in wavelength as shown in Figure 6.

600 700 800 900 1000

6

7

8

9

10

11

Real

part

of d

ielec

tric c

onsta

nt

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 200nm

600 700 800 900 1000

6

7

8

9

10

11

Rea

l par

t of d

iele

ctric

con

stan

t

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 300nm

600 700 800 900 1000

6

7

8

9

10

11

Real

par

t of d

iele

ctric

con

stan

t

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 400nm

a

b

c

d

c

b

a

600 700 800 900 1000

6

7

8

9

10

11

12

Real

part

of d

ielec

tric c

onsta

nt

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 500nm

Figure 6. Real part of dielectric constant against wavelength (nm) for


3.1.5. Optical Band Gap Energy of the as-Deposited Thin FilmsThe plots of (αhν)1/2 (eVcm-1)1/2 against photon energy (eV)(Figure 7) were used to get the optical band gap energy of the as-deposited thin films according to Tauc proposal for most amorphous semiconductors [18]. This was made possible by extrapolating the graphs to the energy axis. Other re-searchers have also found selenium rich binary and ternary glasses to obey the indirect transition rule [14, 15, 19]

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

200

400

600

(h

)1/2 (e

Vcm

-1)1/

2

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 200nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

100

200

300

400

500

600

(h

)1/2(e

Vcm-1

)1/2

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 300nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

100

200

300

400

500

600

(h

)1/2(e

Vcm-1

)1/2

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 400nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

100

200

300

400

500

600

(h

)1/2 (e

Vcm

-1)1/

2

(Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 500nm

Figure 7 (a-d). (αhν)1/2 against photon energy for Se100-xBix thin films at

different thicknesses

3.1.6. Effect of Film ThicknessFigure 8 (a-d) shows increase in reflectance, absorption coefficient, optical band gap energy and a decrease in transmittance with increase in film thickness.

200 250 300 350 400 450 500

0.20

0.25

0.30

0.35

0.40

0.45

Refle

ctanc

e

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

200 250 300 350 400 450 500

1.2

1.4

1.6

Abso

rptio

n co

effic

ient (

×104 cm

-1)

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d

a

b

c

d

a

b

transmittance decreased which was an indication of increase in light absorption by the films. The increase of reflectance with film thickness at specific wavelengths was due to the effect of decreased transmittance.Increase in the optical absorption coefficient with increase in photon energy and with increase in bismuth content may be an indication of increased photosensitivity of the as-deposited thin films with bismuth addition. This trend fits well with already published work [14, 15, 20]. The absorp-tion coefficient values found were in the order of 104cm-1.For low energies, corresponding to high wavelengths, thin film interference effects were observed from figure 3. This might have been due to the overlaying of light that is re-flected on both sides of the as-deposited thin Films [15]. The absorption coefficient increased with increasing film thick-ness as observed from Figure 8 (a-d). This effect may be explained by the fact that thicker films increases the absorption path length of the as deposited thin films decreasing the transmittance. There was increase in extinc-tion coefficient values with increase in bismuth concentra-tion and film thickness as shown in Figure 8 (a-d). This wasas a result of increased absorption coefficient of the as-deposited thin films. The rise and fall of extinction coef-ficient (Figure 4) in the forbidden gap region was directly related to the absorption of light.

Increase in refractive index as the bismuth content increased may be due to increased polarizability of the larger bismuth atoms compared to selenium atoms. The atomic radius of bismuth is 1.50Å and that of selenium is 1.20Å [21]. The high polarizability of the chalcogenide glasses causes them to exhibit the highest intrinsic nonlinear response. The re-fractive index generally increased with increase in film thickness which could be due to increase in the film density. The real part of the dielectric constant with increase in wavelength was observed to decrease (Figure 6). This is because the real part of the dielectric constant represents the in-phase component of the dielectric response function and results in dispersion. The values of real part of the dielectric constant were greater than those for the imaginary part. This is an indication that the absorption coefficient of the as de-posited thin films was high. In addition, both the real part and imaginary parts of the dielectric constant varied with film thickness due to the effect of the refractive index and ex-tinction coefficient.

Decreased band gap energy with bismuth addition is due to bismuth in Se100-xBix thin films that could have led to a dis-tortion of the material valence band, resulting in decreased band gap energy. When bismuth is added to selenium, it introduces defect states in the amorphous selenium glass structure and this reduces the optical band gap energy of the system [14, 22]. Furthermore, introduced bismuth atoms leads to the formation of the heteropolar Se-Bi bonds at the expense of the homopolar Se-Se bonds. The number of Se-Bi bonds increase with increase in bismuth concentration. In-creasing Bi concentration at the expense of Se leads to a shortage of Se-Se bonds, which may lead to lower average bond energy of the alloys. This may be responsible for the decrease of band gap energy with Bi concentration. In addi-tion, bismuth is said to partly break the Se8 ring structure and this may increase the chain fraction [23].

The optical band gap energy depends on the strength of the kind of bond that exists in the Se100-xBix thin films under consideration. The decrease in the optical band gap energy may have been due to decrease in cohesive energy of the system which is the stabilization energy of an infinitely large cluster of the material per atom determined by summing the bond energies of the consequent bonds expected in the as deposited films [24]. The expected bonds in the system under study are Se-Se and Se-Bi. The formation of Bi-Bi bonds is ruled out as it is argued that their bond energy is lower than that of the Se chains [25]. The average cohesive energy of all the bonds expected in the as deposited thin films of Se100-xBix is therefore low, leading to reduced op-tical band gap energy. This is because the optical band gap energy is sensitive to the bond energy [26]. The band gap energy of the as deposited Se100-xBix thin films varied with film thickness due to high density of dislocations resulting from bismuth impurity in the system. Since dislocation den-sity increases with increase in film thickness, it is expected that high dislocation density leads to increase in band gap energy as the film thickness increases. Greater deposition generally builds up more homogenous network which minimizes the number of defects and the localized states,thereby increasing the electronic band gap energy.

17 14

600 700 800 900 1000

6

7

8

9

10

11

12Re

al pa

rt of

diel

ectri

c con

stant

Wavelength (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 500nm

Figure 6. Real part of dielectric constant against wavelength (nm) for


3.1.5. Optical Band Gap Energy of the as-Deposited Thin FilmsThe plots of (αhν)1/2 (eVcm-1)1/2 against photon energy (eV)(Figure 7) were used to get the optical band gap energy of the as-deposited thin films according to Tauc proposal for most amorphous semiconductors [18]. This was made possible by extrapolating the graphs to the energy axis. Other re-searchers have also found selenium rich binary and ternary glasses to obey the indirect transition rule [14, 15, 19]

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

200

400

600

(h

)1/2 (e

Vcm

-1)1/

2

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 200nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

100

200

300

400

500

600

(h

)1/2(e

Vcm-1

)1/2

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 300nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

100

200

300

400

500

600

(h

)1/2(e

Vcm-1

)1/2

Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 400nm

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.60

100

200

300

400

500

600

(h

)1/2 (e

Vcm

-1)1/

2

(Photon energy (eV)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d = 500nm

Figure 7 (a-d). (αhν)1/2 against photon energy for Se100-xBix thin films at

different thicknesses

3.1.6. Effect of Film ThicknessFigure 8 (a-d) shows increase in reflectance, absorption coefficient, optical band gap energy and a decrease in transmittance with increase in film thickness.

200 250 300 350 400 450 500

0.20

0.25

0.30

0.35

0.40

0.45

Refle

ctanc

e

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

200 250 300 350 400 450 500

1.2

1.4

1.6

Abso

rptio

n co

effic

ient (

×104 cm

-1)

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

d

a

b

c

d

a

b

200 250 300 350 400 450 500

1.2

1.4

1.6

1.8

Band

gap e

nergy

(eV)

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

200 250 300 350 400 450 500

0.6

0.7

0.8

0.9

Tran

smitta

nce

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

Figure 8 (a-d). R, α, Eg and T against film thickness (nm) for Se100-xBix

thin films

Reflectance and absorption coefficient increased with increase in bismuth content while transmittance and optical band gap energy decreased with increase in bismuth con-tent as observed from figure 8 (a-d). An increase in extinction coefficient, refractive index, real and imaginary parts of the dielectric constant has been observed in figure 9 (a-d) with increase in film thickness.

200 250 300 350 400 450 500

0.06

0.08

0.10

0.12

0.14

Extin

ction

coeff

icien

t

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

200 250 300 350 400 450 500

2.8

2.9

3.0

3.1

Refra

ctive

inde

x

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

200 250 300 350 400 450 500

7.5

8.0

8.5

9.0

9.5

10.0

Real

part

of d

ielec

tric c

onsta

nt

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

200 250 300 350 400 450 5000.0

0.2

0.4

0.6

0.8

1.0

Imag

inary

par

t of d

ielec

tric c

onsta

nt

Film thickness (nm)

: x=0.0 : x=1.5 : x=3.0 : x=4.5 : x=6.0

Figure 9 (a-d). k, n, ε1 and ε2 against film thickness (nm) for Se100-xBix thin

films

It is observed from figure 9 (a-d) that the extinction coeffi-cient, refractive index, real and imaginary parts of the die-lectric constant increase with increase in bismuth content for the system under investigation.

3.2. DiscussionFlash evaporated thin films of Se100-xBix (x = 0.0, 1.5, 3.0, 4.5 and 6.0 at. %) were found to be amorphous due to absence of prominent peaks in the XRD patterns. Observed zero transmittance at 500nm was as a result of high light absorp-tion at this wavelength [14, 15, 20]. Transmittance decreased with increase in bismuth content due to bismuth defects in the as deposited thin films. As the film thickness increased,

c

d

d

c

b

a

15 16

4. ConclusionThe study investigated the effect of film thickness on the optical properties of flash evaporated, as-deposited, amor-phous Se100-xBix (x = 0.0, 1.5, 3.0, 4.5 and 6.0 at. %) thin films. It is found that transmittance and band gap energy decreased with increase in bismuth content, while absorption coefficient, reflectance, real and imaginary parts of dielectric constant increased with increase in bismuth content. Further, as the film thickness increased from 200±10nm to 500±10nm, transmittance decreased but band gap energy, reflectance, absorption coefficient, extinction coefficient, refractive index, real and imaginary parts of dielectric con-stant increased.

AcknowledgementThe authors wish to thank the National Council for Science and Technology (Now Commission for Science, Technology and Innovation (NACOSTI)) for the 2012-2013 Science, Technology and Innovation (ST & I) 010545 grant support. I also acknowledge University of Nairobi (Physics Depart-ment) for availing the research equipment and Mr. Boniface Muthoka from the same department for his help in setting up of the experiments and the technological advice.

.

REFERENCES [1] N. Mohd and M. Zulfequar, Journal of Inorganic

Non-Metallic Materials, 2, (2012), pp. 11-17[2] A. B. Seddon, W. J. Pan, D. Furniss, C. A. Miller, H. Rowe, D.

Zhan, E. McBrearty, Y. Zhang, A. Loni, P. Sewell, and T. M. Benson, Journal of Non-Crystalline Solids, 352, (2006), pp. 2515-2520

[3] A. V. Kolobov and K. Tanaka, (2001), Handbook of Ad-vanced Electronic and Photonic Materials and Devices, (Ed-ited by Nalwa H. S., 5), Chapter 2: Photo-induced phe-nomena in amorphous chalcogenides: From phenomenology to Nanoscale, Chalcogenide Glasses and Sol-Gel Materials, Academic Press, Japan

[4] M. M. Hafiz, O. El-Shazly, and N. Kinawy, Applied Surface Science, 171, (2001), pp. 231-241

[5] N. Kushwaha, V. S. Kushwaha, R. K. Shukla, and A. Kumar, Journal of Non-Crystalline Solids,351, (2005), pp. 3414-3420

[6] M. M. Malik, M. Zulfequar, A. Kumar, and M. Husain, Journal of Physics: Condensed Matter, 4, No. 43, (1992), pp. 8331-8338

[7] S. Yadav, S. K. Sharma, P.K. Dwivedi, and A. Kumar,Journal of Non-Oxide Glasses, 3, No. 1, (2011), pp. 17-24

[8] K. Ji, R. K. Pal, A. K. Agnihotri, R. K. Shukla, and A. Kumar, Chalcogenide Letters, 6, No. 2, (2009), pp. 77-81

[9] M. A. Majeed Khan, S. Kumar, M. Husain, and M. Zulfequar,Journal of Non-Oxide Glasses, 1, No. 1, (2009), pp. 71-80

[10] K. Ji, R. K. Pal, A. K. Agnihotri, R. K. Shukla, and A. Kumar,Journal of Ovonic Research, 5, No. 3, (2009), pp. 71-76

[11] M. A. Majeed Khan, M. Zulfequar, and M. Husain, Physica B, Netherland, 322, (2002), pp. 1-11

[12] S. Srivastava, M. Zulfequar, and A. Kumar, Chalcogenide Letters, 6, No. 8, (2009), pp. 403-414

[13] M. A. Majeed Khan, S. Kumar, M. Husain, and M. Zulfequar, Chalcogenide Letters, 4, No.12, (2007), pp. 147-153

[14] M. A. Majeed Khan, M. Zulfequar, and M. Husain, Optical Materials, 22, (2003), pp. 21-29

[15] R. Chauhan, A. Tripathi, A. K. Srivastava, and K. K. Sri-vastava, Chalcogenide Letters, 10, No. 2, (2013), pp. 63-71

[16] R. Swanepoel, Journal of Physics E: Science Instrum., 16,(1983), pp. 1214-1218

[17] M. Ohring, (2001), The Material Science of Thin Films,Academic Press, London

[18] J. Tauc, (1974), Amorphous and Liquid Semiconductors, Plenum Press, London, New York

[19] K. Kumar, P. Sharma, S. C. Katyal, and N. Thakur, Phys. Scr., 84, No. 045703 (2011), pp. 1-6

[20] A. Kumar, P. Heera, P. B. Barman, and R. Sharma, Journal of Ovonic Research, 8, No. 5, (2012), pp. 135-146

[21] K. Kumar, N. Thakur, S. C. Katyal, and P. Sharma, Defect and Diffusion Forum, 305/306, (2010), pp. 61-69

[22] N. F. Mott and E. A. Davis, (1979) Electronic Processes in Non-Crystalline Materials, 2nd ed., Clarendon Press, Oxford,

[23] J. Schottmiller, M. Tabak, G. Lucovsky, and A. Ward, Journal of Non-Crystalline Solids, 4, (1970), pp. 80-96

[24] S. R. Elliot, (2000), The Physics and Chemistry of Solids, 3rd

Ed., Wiley, Chichester [25] M. M. Hafiz, A. A. Othmana, M. M. El-Nahassb, and A. T.

Al-Motasema, Physica B, 390, No. 1-2, (2007), pp. 348-355 [26] J. Bicerano and S. R. Ovshinsky, Journal of Non-Crystalline

Solids, 74, No. 1, (1985), pp. 75-84

The shape of the peaks depicts a concentration gradient of Cu in In2S3 with higher concentration at the surface. Inclusion of Cu into In2S3 matrix results in an increase in the volume of the In2S3 which can be seen by increase in the width of the In and S peaks with increasing diffusion temperature.

Figure 2 shows the Cu peaks from Figure 1. The distribution of Cu in the In2S3 bulk is not similar for In2S3(acac):Cu and In2S3(Cl):Cu layers. The distribution in the other In2S3(Cl):Cu layers with varying Cl content was similar to figure 2(b). The Cu peaks increased systematically with diffusion temperature for In2S3(acac):Cu similar to the case for thermally evaporated In2S3 layers reported in reference [23]. The Cu peaks from In2S3(Cl):Cu increased slowly with diffusing temperature for temperatures below 200°C and then increased strongly above 200°C. The distribution of Cu tends to be more

constant in the In2S3(Cl):Cu bulk than in In2S3(acac):Cu. The difference in the nature of the concentration gradients can be attributed to the presence of Cl. Cl therefore plays a crucial role in the diffusion of Cu in In2S3 layers.

The depth profiles of Cu were extracted from the Cu peaks of the RBS spectra for the different samples using WiNDF software [30]. A recursion equation was developed from the Fick’s second law of diffusion and used to numerically calculate the Cu concentration profiles and compare the results to the measured data. Details can be found in reference [23]. The depth profiles were fitted numerically with minimum deviation and values for the diffusion coefficients were obtained. The measured and simulated depth profiles for In2S3(acac):Cu and In2S3(Cl):Cu are shown in figure 3.

The depth profiles show a thin surface layer with high Cu concentration independent of diffusion temperature. This means an interfacial layer with a large amount of Cu was formed upon deposition of CuSCN and annealing. A thin interfacial layer due to diffusion of Cu into In2S3 layer was also observed for Cu(In,Ga)Se2 /In2S3 layer system for Chalcopyrite solar cells [3]. This thin surface layer was therefore not used in fitting the profiles to obtain diffusion coefficient values.

Diffusion coefficient values were calculated for all the samples with different Cl contents and plotted against inverse temperature in an Arrhenius plot as shown in Figure 4.

The activation energy of Cu diffusion in the different In2S3 layers and exponential prefactors were obtained from the Arrhenius plots. The results for Cu diffusion

in thermally evaporated In2S3 layers [23] are included for comparison.

The values of the activation energies and exponentialprefactors for all the samples with different Cl concentrations are shown in Table 2.

Table 2

Acknowledgement The author wishes to acknowledge all the contributors for their commitment and especially the solid state physics lab and Helmholtz-zentrum Berlin where the work was done.

In2S3:Cl13.8at.%

In2S3:Cl11.3at.%

In2S3:Cl8.5at.%

In2S3:Cl7.8at.%

In2S3:acac

In2S3:PVD

EA (eV) 0.70 0.72 0.78 0.76 0.24 0.30

D0(cm2/s)

6.0x10-6 3.0x10-6 3.2x10-5 1.2x10-5 2.7x10-11 9.0x10-11

Figure 3: Simulated and measured depth profiles for Cu in (a) In2S3(acac):Cu and (b) In2S3 (Cl):Cu from which diffusion coefficients were determined

Figure 4: Arrhenius plot of diffusion coefficient against inverse temperature for different In2S3 layers

21 18

Role of Cl on Diffusion of Cu in In2S3 Layers Prepared by Ion Layer Gas Reaction Method

H. Wafula1*, R. J. Musembi2, A. Juma3, R. Sáez-Araoz3, Christian-Herbert Fischer3, Th. Dittrich3 and E. Wendler4 1Physics Department-Masinde Muliro University of Science and Technology, Kakamega-50100, Kenya 2Physics Department-University of Nairobi, Nairobi-00100, Kenya 3Helmholtz - zentrum Berlin für Materialien und Energie. Hahn-Meitner-Platz 1, 14109 Berlin, Germany 4Friedrich-Schiller-Universität Jena, Institute für Festkörperphysik, Max-Wien-Platz 1, 07743 Jena, Germany

*Corresponding author:[email protected]

Abstract Ion layer gas reaction (ILGAR) method allows for deposition of Cl-containing and Cl-free In2S3 layers from InCl3 and In(OCCH3CHOCCH3)3 precursor salts, respectively. A comparative study was performed to investigate the role of Cl on the diffusion of Cu from CuSCN source layer into ILGAR deposited In2S3 layers. The Cl concentration was varied between 7 and 14 at.% by varying deposition parameters. The activation energies and exponential pre-factors for Cu diffusion in Cl-containing samples were between 0.70 to 0.78 eV and between 6.0 x 10-6 and 3.2 x 10-5 cm2/s respectively. The activation energy in Cl-free ILGAR In2S3 layers was about 3 times less and the pre-exponential constant six orders of magnitude lower but comparable to values obtained from thermally evaporated In2S3 layers. The residual Cl-occupies S sites in the In2S3 structure leading to non-stoichiometry and hence different diffusion mechanism for Cu compared to stoichiometric Cl-free layers

Keywords: diffusion, ILGAR, In2S3, CuSCN

1.0 Introduction In2S3 has found increased attention in photovoltaics as a replacement for toxic CdS because of its suitable properties[1-4]. It has been effectively used as a buffer layer for both chalcopyrite [1-3] and nanocomposites [5, 6] solar cells and as an extremely thin absorber (eta) [7, 8] and nanocomposites [9] solar cells with ZnO nanorods ornanoporours-TiO2 electrodes, respectively. Diffusion of Cu from Cu(In,Ga)(S,Se)2 absorber [3, 10, 11] or CuSCN hole conductor [12] into In2S3 layers has been a major drawback to effective performance and stability of these solar cells.

In2S3 exists in three phases: β-phase which is stable at room temperature up to 4200 [13,14], the α-phase which is stable above 420°C [15] up to 754°C and the γ-phase which is stable between 754°C and its melting point 1090°C [16]. The most commonly used phase is the β-In2S3. β-In2S3 has a defect spinel superstructure in which In occupies all octahedral sites and a third of the tetrahedral sites [16, 17]. A third of the tetrahedral cationic sites remain vacant but ordered along 4I screw with alignment of three spinel blocks in c-direction [17]. β-In2S3 can accommodate foreign atoms for example Fe [18], Ag [19], Cl [20], Cu and Na [10].

The presence of foreign impurities in host In2S3 affects occupation and migration of new foreign atoms. Diffusion of Cu in In2S3 is inhibited by the presence of Na because both Na and Cu compete for the same cationic vacancies and substitution sites in In2S3 structure [21]. Different authors have reported diffusion of Cu in In2S3 to be by vacancy [22] or insertion/substitution [21] mechanism. A

better understanding and hence control of Cu diffusion in In2S3 layers is crucial for photovoltaic applications [23].

ILGAR is a sequential and cyclic method of depositing highly conformal In2S3 layers on different substrates [24,25]. In2S3 layers can be deposited from InCl3 or In(OCCH3CHOCCH3)3 to obtain Cl-free and Cl-containing layers [26]. The presence of Cl in In2S3 increased the optical band up to 2.4 eV compared to 2.0 eV for Cl-free layers. Higher conversion efficiency was achieved for Cl-free In2S3layers compared to Cl-containing layers both prepared by ILGAR [26]. The stoichiometry as well as opto-electronic properties of In2S3 layers also changes with inclusion of Cl. Crystallinity and photosensitivity increased with increasing Cl content [20]. It is therefore important to investigate the effect of Cl on diffusion of Cu in ILGAR deposited In2S3layers.

Rutherford backscattering spectroscopy (RBS) is an absolute method for determination of elemental composition and depth profiling of various materials with high accuracy [27, 28]. The diffusion coefficient of Cu in In2S3 can be obtained from Cu concentration profiles by solving the diffusion equation analytically or numerically with appropriate boundary conditions and comparing or fitting the resulting profiles with the experimental data. Activation energy and diffusion prefactor of 0.3 eV and 9 x 10-11 cm2s-1 determined from RBS depth profiling have been reported for Cu diffusion in thermally evaporated In2S3layers [23].

In this work, we report a comparative study of Cu diffusion in Cl-free and Cl-containing In2S3 layers. In2S3 layers were prepared by ILGAR method which was modified to obtain

layers with varying Cl content. The Cu source was CuSCN deposited on top of In2S3. The distribution of diffused Cu in In2S3 was profiled by RBS from which the diffusion coefficients were determined as a function of annealing temperature.

2.0 Experiments In2S3 layers were deposited by ILGAR [24, 25] onto c-Siwafers without additional treatment of the c-Si substrates. InCl3 or In(acac)3 precursor salt was dissolved into ethanol solvent to obtain a solution of 25mM. In(acac)3 precursor was used to deposit Cl-free layers while InCl3 was used for Cl-containing layers. The deposition temperature for deposition of all samples was 200°C. The standard ILGAR cycles were adjusted to produce In2S3 layers with varying Cl content. The details of the deposition parameters for the different samples are given in Table 1. CuSCN was deposited onto each of the c-Si / In2S3 samples by spray-spin coating method [23] from a solution of 50 mM CuSCN dissolved in propyl sulfide [29]. For each sample, ten spray-spin coating cycles were deposited, which corresponds to about 500 nm thickness as determined by Dektak profilometer. The c-Si / In2S3 / CuSCN samples were then cut into smaller sizes of about 7 x 7 mm. The smaller pieces were then annealed for 5 minutes at temperatures between 150 °C and 250 °C. A set of small samples annealed at these temperatures and one not annealed was obtained for each large c-Si / In2S3 / CuSCN sample.

CuSCN layer was then etched away in pyridine solution [23]. The sample was first dipped into pyridine solution for 2 seconds then rinsed in deionized water to wash away the pyridine. This was repeated until a shiny surface similar to the as-deposited In2S3 was obtained. RBS measurements were performed with 1.4 MeV He+ ion beam and the backscattered ions detected at a scattering angle of 168° [23].

3.0 Results The composition of In2S3(acac) and In2S3(Cl) was determined from RBS measurements on as-deposited samples. The integral of the respective peaks were used to calculate the areal densities, from which the layer

thicknesses were obtained. The number of atoms (Nt)) also called areal density of a given element in a layer is given by [27],

0

1cos)(EQ

ANt i

(1)

where A the integral of the peak, Q the number of ions incident on the sample, Ω the detector solid angle and σ(E0)the differential scattering cross-section. The concentration of each element was obtained by dividing the atomic density (Ni) of each element by that of In2S3 (4.53 x 1022cm-

3). The values of the atomic concentrations and layer thicknesses of the samples are given in table 1.

Table 1: Thickness and layer composition of In2S3 films

Sample S0 was In2S3(acac) while samples S8, S9, S11 and S12 were In2S3(Cl). Deposition parameters were adjusted to obtain In2S3(Cl) with different Cl content. The difference in layer thicknesses was due to a difference in deposition rate due to the changes in the deposition steps. The presence of large amounts of Cl in In2S3(Cl) layers led to a strong deviation from stoichiometry while In2S3(acac) layers were more stoichiometric with S/In ration of 1.56 which is very close the expected value of 1.5. The amount of In was almost constant for all In2S3(Cl) samples while S content dependent on the residual Cl. This means that the InCl3precursor layers were not completely sulfurized, leaving behind Cl residues.

RBS spectra for In2S3(acac) and In2S3(Cl) containing diffused Cu after annealing for 5 min at different temperatures are shown in figure 1. The various peaks corresponding to the elements detected in the samples are shown. The Cu peak increases in height with increasing diffusion (annealing) temperature.

Figure 2: Cu peaks for (a) In2S3(acac):Cu and (b) In2S3(Cl):Cu for different diffusion temperatures. The scatters represent measured data while the solid lines are from simulation.

Figure 1: RBS spectra for (a) In2S3(acac):Cu and In2S3(Cl):Cu after diffusion of Cu at different

temperatures. The respective peaks are indicated.

Name Thickness(nm)

In (at.%) S (at.%) Cl(at.%) S/In

S0 86 39 61 0 1.56

S8 152 37.5 52.2 11.3 1.41

S9 65 37.1 49.1 13.8 1.32

S11 64 37.8 54.4 7.8 1.44

S12 45 37.6 53.9 8.6 1.43

For films prepared from ceramic target, it is seen that the peak becomes more intense and sharper with decreasing Ar flow rate. This means that the film crystallinity is improved and the grain size of the crystallites becomes larger at a lower Ar flow rate. In this study, lower gas flow rate are expected to improve the crystallinity of the film by increasing the moment transfer of sputtered atoms to the film growing on the heated substrate. A weak diffraction peak of (201) was observed in films prepared using ceramic and alloy targets at an Ar and O2 flow rate of 30ml/min and 5.23 ml/min, respectively, which then disappeared with decrease in Ar flow rate. The diffraction peaks of (100) and (004) were also observed in both films but neither oxide Al2O3 nor metallic (Zn or A) phases were observed. This indicates that, the prepared ZnO:Al films were crystalline exhibit a strong c-axis preferred orientation.

3.2 Electrical Properties

Compared with films prepared from ceramic target having resistivity and mobility ranging from 3.421 x 10-3 to 4.023 x 10-4 Ωcm and 14.08 to 16.53 cm2/Vs, respectively, the electrical properties of films prepared using alloy target are effectively improved. The films exhibit relatively low resistivity ranging from (1.588 – 2.908)x10-4 Ωcm which is close to those reported by [22] and higher mobility ranging from 13.5 to 16 cm2/Vs. The carrier concentration for films prepared from alloy target ranged from 1.732 x 1021 cm-3 to 2.981 x 1021 cm-3 while those prepared from ceramic target ranged from 8.608 x 1020 cm-3 to 9.733 x 1020 cm-3 (Figures 2 &3).

Figure 2. Influence of O2 gas flow rate on carrier concentration and

mobility

Figure 3. Effects of Ar gas flow rate on carrier concentration and

mobility

Despite the fact that both films having the same thickness ~300 nm but they have quite different electrical properties. The good electrical properties of the films prepared from alloy target might be due to the incorporation of Al from the alloy target into zinc in the lattice and thus increasing carrier concentration and mobility of the films. However, with increasing O2 gas flow into the chamber, leads to the increase in resistivity and decrease mobility and carrier concentration. This observed increase in resistivity with increase in O2 gas flow rate can be caused by increased O atoms involved in the process of reactive sputtering, oxygen atom combined to metal oxide with metal, leading to the decreased Zn interstitial atoms. Furthermore, metallic Al atoms might be combined with oxygen atoms to oxides, and thus increasing resistivity and there after decreasing the mobility and carrier concentration. Similar results were reported by [23, 24].

The same trend was also observed for films prepared using ceramic target, where by the resistivity was found to decrease with the decrease of Ar gas flow rate. The lowest resistivity 4.023 x 10-4 was obtained at an Ar gas flow rate of 20 ml/min. The conduction mechanism of ZnO:Al films prepared using ceramic target, is affected by Ar flow rate.This is due to the variation in grain size which differs accordingly with Ar flow rate [25]. In this study, the mobility and carrier concentration were found to decrease with increase in Ar gas flow rate due to the fact that at increased Ar flow rate resulted in stronger collision between the sputtered target atoms with the argon atoms thus reducing kinetic energy of the sputtered atoms. This can hinder their migrating capability on the substrate, resulting in small grain size of the film which also can reduce the mobility of the film. Collectively (from figures), the films grown from the alloy targets exhibited

19 20

References 1. N. Barreau, J.C. Bernéde. H. El Maliki, S. Marsillac, X.

Castel and J. Pinel, Solid State Comm. 122, (2002), 445 2. M. Bär, N. Allsop, I. Lauermann and Ch.-H. Fischer,

Appl. Phys. Lett. 90, (2007), 132118. 3. D. Abou.Ras, G. Kostorz, A. Strohm, H.-W. Schock

and A.N. Tiwari, J. Appl. Phys. 98, (2005), 123512.4. N. Barreau, Solar Energy 83, (2009)363. 5. M. Nanu, J. Schoonman, and A. Goossens, Nano Lett.,

5, (2005), 1716. 6. M. Nanu, J. Schoonman, and A. Goossens, Adv. Mater.

16, (2004), 453. 7. A.Belaidi, Th.Dittrich, D.Kieven, J.Tornow,

K.Schwarzburg, M.Kunst, N.Allsop, M.-Ch. Lux-Steiner and S.Gavrilov, Solar Energy Materials & Solar Cells 93, (2009), 1033.

8. D. Kieven, T. Dittrich, A. Belaidi, J. Tornow, K. Schwarzburg, N. Allsop, and M. Lux-Steiner, Appl. Phys. Lett. 92, (2008), 153107.

9. C. Herzog, A. Belaidi, A. Ogacho and Th. Dittrich, Energy Environ. Sci., 2, (2009), 962.

10. D. Abou-Ras, D. Rudmann, G. Kostorz, S. Spiering, M. Powalla and A. N. Tiwari, J. Appl. Phys. 97, (2005),084908.

11. P. Pistor, N. Allsop, W. Braun, R. Caballero, C. Camus, Ch.-H. Fischer, M. Gorgoi, A. Grimm,B. Johnson, T. Kropp, I. Lauermann, S. Lehmann, H. Mönig, S. Schorr, A. Weber and R. Klenk, Phys. Status Solidi A 206, (2009), 1059.

12. T. Dittrich, D. Kieven, A. Belaidi, M. Rusu, J. Tornow, K. Schwarzburg, and M. Ch. Lux-Steiner, J. Appl. Phys. 105, (2009), 034509.

13. C.J.M. Rooyman, J. Inorg. Nucl. Chem. 11, 78 (1959). 14. G.S.D. King, Acta Cryst. 15, (1962), 512. 15. H.Hahn and W. Klinger, Z.Anorg. Chem. 260, (1949),

97.16. R. Diehl, C.-D. Carpentier and R. Nitsche, Acta Cryst.

B 32, (1976), 1257. 17. R. Diehl and R. Nitsche, J. Cryst. Growth 28, (1975),

306. 18. M. Womes, J. Olivier-Fourcade, J.-C. Jumas, F.

Aubertin, and U. Gonser, J. OF Solid State Chem. 97,(1992), 249.

19. M. Mathew, C. Sudha Kartha, K. P. Vijayakumar, J Mater Sci: Mater Electron 20, (2009), S294.

20. A. S. Cherian, M. Mathew, C. S. Kartha, K.P. Vijayakumar, Thin Solid Films 518, (2010), 1779

21. C. Guillot-Deudon, S. Harel, A. Mokrani, A. Lafond, N. Barreau, V. Fernandez, and J. Kessler, Phys. Rev. B 78, (2008), 235201.

22. F. Py, M. Womes, J.M. Durand, Olivier-Fourcade, J.-C. Jumas, J.m. Esteva, R.C. Karnatak, J. Alloys and Compounds 178, (1992), 297.

23. A. O. Juma, P. Pistor, Steffen Fengler, Thomas Dittrich and Elke Wendler, Thin Solid Films 520, (2012), 6740.

24. Ch.-H. Fischer, H.-J. Muffler, M. Bär, S. Fiechter, B. Leupolt and M.C. Lux-Steiner, Journal of Crystal Growth 241, (2002), 151.

25. S. Gledhill, R. Allison, N. Allsop, Y. Fu, E. Kanaki, R. Sáez-Araoz, M. Lux-Steiner and C.-H. Fischer, Thin Solid Films 519, (2011), 6413.

26. R. Sáez-Araoz, J. Krammer, S. Harndt, T. Koehler, M. Krueger, P. Pistor, A. Jasenek, F. Hergert, M. Ch. Lux-Steiner and C.-H. Fischer, Prog. Photovolt: Res. Appl. 20, (2012), 855.

27. C. Jeynes, N. P. Barradas, and E. Szilagyi, Anal. Chem. 84, (2012), 6061.

28. C. Jeynes, N. P. Barradas, H. Rafla-Yuan, B. P. Hichwa and R. Close, Surf. Interface Anal. 30, (2000), 237.

29. G.R.R.A. Kumara, A. Konno, G.K.R. Senadeera, P.V.V. Jayaweera, D.B.R.A. De Silva, K. Tennakone, Solar Energy Materials & Solar Cells 69, (2001), 195.

30. N. P. Barradas, C. Jeynes, and R. P. Webb, Appl. Phys. Lett. 71, (1997), 291.

Effects of Target Composition on the Optical Constants

of DC Sputtered ZnO:Al Thin Films B. Samwel*, N.R. Mlyuka1, M. E. Samiji1, and R. T. Kivaisi1.

* College of Natural and Mathematical Sciences, Physics Department, University of Dodoma, P. O. Box 338, Dodoma,

Tanzania 1Solar Energy Group, Physics Department, University of Dar es Salaam, P. O. Box 35063 Dar es Salaam, Tanzania.

.

Abstract Al-doped ZnO thin films were deposited from ZnO:Al ceramic and Zn:Al metal alloy targets. Deposition took place in Ar and Ar + O2 atmosphere for ZnO:Al and Zn:Al targets, respectively, using Direct Current (DC) magnetron sputtering. Transmittance (T) measurements showed T > 80% in visible region with good Near Infrared (NIR) shielding. The band gap energy ranged from 3.34 to 3.44 eV and 3.39 to 3.46 eV for films prepared from alloy and ceramic targets, respectively. The films with lowest electrical sheet resistance of 10 Ω/ and highest values of mobility and carrier concentration of 15.9 cm2/Vs and 2.98 x 1021 cm-3 respectively, were obtained using alloy-target at a substrate temperature of 200 oC. However, films prepared from ceramic target at a substrate temperature of 300 oCrevealed the lowest sheet resistance of 32 Ω/, with the highest values of mobility and charge carrier concentration of 14.1 cm2/Vs and 9.733 x 1020 cm-3 respectively. Optical spectra of the films were fitted to SCOUT software in order to determine the refractive index, n and extinction coefficient, k. Generally, the calculated n and k in the visible part of the solar spectrum for different samples, ranged from 1.59 to 2.2 and 0.00013 to 0.0194 respectively, which are in agreement with results calculated using other methods.

Key words: DC Magnetron Sputtering, Optical Constants, Transparent Conducting Oxides (TCO)

1. Introduction Al doped zinc oxide films (ZnO:Al) are wide-band-gap semiconductors with high visible transparency, high reflectance in the near infrared wavelength range along with good electrical conductivity of nearly semi-metallic regime. The combination of high visible transparency with good electrical conductivity makes ZnO:Al films widely applicable in a variety of optoelectronic devices where one makes use of either the optical or electrical properties, or both of them [1-4]. However, their properties are highly dependent on defect structure, deposition technique and the deposition parameters employed. Film thickness and the substrate used can also have great influence on the properties [5-6].

ZnO:Al films have recently gained much attention as a promising alternative material to tin oxide (SnO2) and indium oxide (In2O3) and their doped compound because

of its competitive electrical and optical properties as well as low-cost materials and abundance in natural resource [7-9]. Apart from that, ZnO has good chemical stability in hydrogen plasma, for example in silane (SiH4) which is used for the preparation of a-Si:H thin film solar cells ([10, 11). Moreover, ZnO:Al films have good behaviour concerning roughening and surface texturing which helps to enhance light scattering and absorption inside a solar cell and thus enhance the performance of a solar cell [11]. Various deposition techniques have been employed in preparing ZnO:Al films. Besides chemical vapour deposition (CVD), spray pyrolysis, pulsed laser deposition, sol-gel, magnetron sputtering is widely used [12-16]. This technique allows deposition of ZnO:Al films at low substrate temperatures down to room temperature and at high deposition rates. The sputtered films show good adhesion on substrates and the technique offers very good thickness uniformity and high density of the films ([15, 16]. Taking this into account, magnetron sputtering is

For films prepared from ceramic target, it is seen that the peak becomes more intense and sharper with decreasing Ar flow rate. This means that the film crystallinity is improved and the grain size of the crystallites becomes larger at a lower Ar flow rate. In this study, lower gas flow rate are expected to improve the crystallinity of the film by increasing the moment transfer of sputtered atoms to the film growing on the heated substrate. A weak diffraction peak of (201) was observed in films prepared using ceramic and alloy targets at an Ar and O2 flow rate of 30ml/min and 5.23 ml/min, respectively, which then disappeared with decrease in Ar flow rate. The diffraction peaks of (100) and (004) were also observed in both films but neither oxide Al2O3 nor metallic (Zn or A) phases were observed. This indicates that, the prepared ZnO:Al films were crystalline exhibit a strong c-axis preferred orientation.

3.2 Electrical Properties

Compared with films prepared from ceramic target having resistivity and mobility ranging from 3.421 x 10-3 to 4.023 x 10-4 Ωcm and 14.08 to 16.53 cm2/Vs, respectively, the electrical properties of films prepared using alloy target are effectively improved. The films exhibit relatively low resistivity ranging from (1.588 – 2.908)x10-4 Ωcm which is close to those reported by [22] and higher mobility ranging from 13.5 to 16 cm2/Vs. The carrier concentration for films prepared from alloy target ranged from 1.732 x 1021 cm-3 to 2.981 x 1021 cm-3 while those prepared from ceramic target ranged from 8.608 x 1020 cm-3 to 9.733 x 1020 cm-3 (Figures 2 &3).

Figure 2. Influence of O2 gas flow rate on carrier concentration and

mobility

Figure 3. Effects of Ar gas flow rate on carrier concentration and

mobility

Despite the fact that both films having the same thickness ~300 nm but they have quite different electrical properties. The good electrical properties of the films prepared from alloy target might be due to the incorporation of Al from the alloy target into zinc in the lattice and thus increasing carrier concentration and mobility of the films. However, with increasing O2 gas flow into the chamber, leads to the increase in resistivity and decrease mobility and carrier concentration. This observed increase in resistivity with increase in O2 gas flow rate can be caused by increased O atoms involved in the process of reactive sputtering, oxygen atom combined to metal oxide with metal, leading to the decreased Zn interstitial atoms. Furthermore, metallic Al atoms might be combined with oxygen atoms to oxides, and thus increasing resistivity and there after decreasing the mobility and carrier concentration. Similar results were reported by [23, 24].

The same trend was also observed for films prepared using ceramic target, where by the resistivity was found to decrease with the decrease of Ar gas flow rate. The lowest resistivity 4.023 x 10-4 was obtained at an Ar gas flow rate of 20 ml/min. The conduction mechanism of ZnO:Al films prepared using ceramic target, is affected by Ar flow rate.This is due to the variation in grain size which differs accordingly with Ar flow rate [25]. In this study, the mobility and carrier concentration were found to decrease with increase in Ar gas flow rate due to the fact that at increased Ar flow rate resulted in stronger collision between the sputtered target atoms with the argon atoms thus reducing kinetic energy of the sputtered atoms. This can hinder their migrating capability on the substrate, resulting in small grain size of the film which also can reduce the mobility of the film. Collectively (from figures), the films grown from the alloy targets exhibited

25 22

Effects of Target Composition on the Optical Constants

of DC Sputtered ZnO:Al Thin Films B. Samwel*, N.R. Mlyuka1, M. E. Samiji1, and R. T. Kivaisi1.

* College of Natural and Mathematical Sciences, Physics Department, University of Dodoma, P. O. Box 338, Dodoma,

Tanzania 1Solar Energy Group, Physics Department, University of Dar es Salaam, P. O. Box 35063 Dar es Salaam, Tanzania.

.

Abstract Al-doped ZnO thin films were deposited from ZnO:Al ceramic and Zn:Al metal alloy targets. Deposition took place in Ar and Ar + O2 atmosphere for ZnO:Al and Zn:Al targets, respectively, using Direct Current (DC) magnetron sputtering. Transmittance (T) measurements showed T > 80% in visible region with good Near Infrared (NIR) shielding. The band gap energy ranged from 3.34 to 3.44 eV and 3.39 to 3.46 eV for films prepared from alloy and ceramic targets, respectively. The films with lowest electrical sheet resistance of 10 Ω/ and highest values of mobility and carrier concentration of 15.9 cm2/Vs and 2.98 x 1021 cm-3 respectively, were obtained using alloy-target at a substrate temperature of 200 oC. However, films prepared from ceramic target at a substrate temperature of 300 oCrevealed the lowest sheet resistance of 32 Ω/, with the highest values of mobility and charge carrier concentration of 14.1 cm2/Vs and 9.733 x 1020 cm-3 respectively. Optical spectra of the films were fitted to SCOUT software in order to determine the refractive index, n and extinction coefficient, k. Generally, the calculated n and k in the visible part of the solar spectrum for different samples, ranged from 1.59 to 2.2 and 0.00013 to 0.0194 respectively, which are in agreement with results calculated using other methods.

Key words: DC Magnetron Sputtering, Optical Constants, Transparent Conducting Oxides (TCO)

1. Introduction Al doped zinc oxide films (ZnO:Al) are wide-band-gap semiconductors with high visible transparency, high reflectance in the near infrared wavelength range along with good electrical conductivity of nearly semi-metallic regime. The combination of high visible transparency with good electrical conductivity makes ZnO:Al films widely applicable in a variety of optoelectronic devices where one makes use of either the optical or electrical properties, or both of them [1-4]. However, their properties are highly dependent on defect structure, deposition technique and the deposition parameters employed. Film thickness and the substrate used can also have great influence on the properties [5-6].

ZnO:Al films have recently gained much attention as a promising alternative material to tin oxide (SnO2) and indium oxide (In2O3) and their doped compound because

of its competitive electrical and optical properties as well as low-cost materials and abundance in natural resource [7-9]. Apart from that, ZnO has good chemical stability in hydrogen plasma, for example in silane (SiH4) which is used for the preparation of a-Si:H thin film solar cells ([10, 11). Moreover, ZnO:Al films have good behaviour concerning roughening and surface texturing which helps to enhance light scattering and absorption inside a solar cell and thus enhance the performance of a solar cell [11]. Various deposition techniques have been employed in preparing ZnO:Al films. Besides chemical vapour deposition (CVD), spray pyrolysis, pulsed laser deposition, sol-gel, magnetron sputtering is widely used [12-16]. This technique allows deposition of ZnO:Al films at low substrate temperatures down to room temperature and at high deposition rates. The sputtered films show good adhesion on substrates and the technique offers very good thickness uniformity and high density of the films ([15, 16]. Taking this into account, magnetron sputtering is

considered as a favourable deposition technique in which ZnO:Al films with good optical properties can be deposited.

Among the factors that influence physical properties of DC sputtered ZnO:Al thin films are the target composition and deposition parameters used. It was reported that, using metallic Zn:Al targets in preparing sputtered ZnO:Al resulted in high quality films with good physical properties [17-20]. Other literature [10,12] reported that high quality ZnO:Al thin films with good physical properties were obtained by preparing the films from ceramic targets. This study therefore investigated the effects of the target compositions (ceramic against metallic targets) on the optical constants of ZnO:Al thin films. The idea here was to identify suitable type of targets for the preparation of good quality DC Magnetron sputtered ZnO:Al thin films for solar cell applications.

2. Experimental methods 2.1 Sample preparation

ZnO:Al films were deposited on soda lime glass (SLG) substrates by reactive DC magnetron sputtering using BALZER BAE 250 coating unit. DC magnetron sputtering using metallic targets consisting of a Zn:Al alloy (2% Al) and ceramic target containing ZnO:Al (2% Al) (both from AJA International Inc., USA) was employed. The sputtering chamber was pumped down to a base pressure of 4.3 x 10-6 mbar using a turbo-molecular pump backed by a rotary pump. The sputtering gas Ar (of 99.999% purity) and reactive gas (O2 of 99.999% purity) were introduced into the chamber separately, and were controlled through the gas mass flow rate controllers. The gas flow meters were controlled precisely to allow flow rates of oxygen and argon gases at about 5.0 – 5.5 and 60 ml/min, for the case of Zn:Al target. However, for ceramic target ZnO:Al, oxygen and argon gases flow rates were at about 0.0 ml/min and 10 – 70 ml/min, respectively. The substrate temperature was varied from 100 oC to 350 oC,however, in this study, the optimal values of substrate temperatures were 200 oC and 300 oC for films prepared from alloy and ceramic targets, respectively. The sputtering power of 70 W was used and pre-sputtering for10 minutes was done so as to make the plasma stable and to remove any unwanted material. Target to substratedistance was fixed at 15 cm for both targets.

2.2 Characterization techniques

A computerized, high sensitivity surface profiler, Alpha step IQ Surface profiler was used to measure the thickness of ZnO:Al films. The optical transmittance and reflectance

in the UV/VIS/NIR was measured in the wavelength range of 250 - 3000 nm at an interval of 1 nm by using Shimadzu SolidSpec-3700 DUV Spectrophotometer. Optical spectra of the films were fitted to SCOUT software [21] in order to determine the refractive index, n and extinction coefficient, k, whereby three layers; air, ZnO:Al and substrate were theoretically implemented. The electrical properties were investigated using a Hall effect Measurement System (Ecopia HMS 3000). X-ray diffractometer (XRD) with θ-2θ scanning mode was used for crystalline characterization. The diffraction angle 2θwas varied from 5.000o to 84.997o. Film characterization was carried out at room temperature.

3. Results and Discussion 3.1 Structural Characterization

The crystalline structure and orientation of the ZnO:Al films having thickness of 300 nm were investigated by XRD. Figure 1 shows the XRD spectra of AZO thin films deposited at a substrate temperature of 200 oC for alloy target and 300 oC for ceramic target, the O2 and Ar flow rate was kept at 5.17, 5.19 and 5.23 ml/min and 20, 30, 50 ml/min, respectively. The films exhibited a high degree of preferred orientation which depends on the various gas flow rates. In this study, the θ-2θ scan data of ZnO:Al films exhibit a strong 2θ peaks corresponding to the (002) peaks of ZnO. For the films prepared using alloy target, increasing O2 flow rates from 5.17 ml/min to 5.19 did not change the locations of the measured diffraction peaks but they become more intense. However, further increase in O2flow rate to 5.21 ml/min, the intensity of the peak decreases.

Figure 1. The XRD patterns of ZnO:Al films with thickness 300 nm

deposited under different gas flow rate in the sputtering chamber.

23 24

much preferable electrical characteristics, whereas lower sheet resistance and resistivity and higher mobility and carrier concentration have been obtained at a substrate temperature of 200 oC. This demonstrates that DC sputtering from alloy target is better suited for TCO since it can be easily obtained at low temperature with good electrical properties.

3.3 Optical Properties

The optical properties were controlled by adjusting Ar gas flow ratio and O2 flow ratio as a mixture of Ar+O2 gases for ceramic and alloy targets, respectively. The transmittance of the films prepared from both targets was higher than 80 % with oscillatory character due to interference effects in the visible region from 300 to 800 nm with a sharp fall in the transmittance at the absorption edge. Plasma frequency shifted to higher wavelength with increase in O2 flow ratio in the mixture of Ar+O2 gasesand Ar flow ratio for films prepared from alloy and ceramic targets, respectively, and showed excellent NIR shielding as can be seen in Figure 4a,b. This might be due to zinc adsorption and substitutional doping of an Al3+ at the Zn2+ site creating one extra free carrier in the process which occurs at a substrate at low gas flow ratios which leads to the increase in carrier concentration [13, 15, 20].As is generally known an increased carrier concentration can give increased absorption in the long wavelength region due to free electron gas. Similar results for the films prepared using alloy target were reported by [13, 15, 21,24]. The films from both targets show typical optical characteristics of TCOs whereas they have a transmission window in the 400 – 1500 nm wavelength range.However, the highest average optical transmittance was achieved using ceramic target and was calculated to be ~80.14 %. The difference arises in near normal reflectance in the NIR region where the films prepared from alloy target showed high NIR reflectance, higher than 80%. The band gap shift was also observed where the band gap was shifted from 3.49 to 3.43 eV with decrease in Ar gas flow rate, while for films prepared from alloy; the band gap was increased from 3.34 to 3.54 eV with decrease in O2 gas flow rate. This is possibly due to the Burstein-Moss shift which is attributed to an increase in the carrier concentration as it is seen section 3.2. This increase in carrier concentration blocks the lowest states (filled states) in the conduction band from absorbing the photons [24]resulting in effective widening of the band gap energy. The shift in absorption edge might also be due to the presence of residual compressive stress as reported by [26].

Figure 4a Optical transmittance and reflectance spectra of ZnO:Al films

grown at various O2 flow rates as a mixture of Ar + O2 gas at a substrate

temperature of 200 oC.

Figure 4b Optical transmittance and reflectance spectra of ZnO:Al films

grown at various Ar gas flow rates at a substrate temperature of 300 oC.

The optical constants (refractive index, n and extinction coefficient, k) as obtained by fitting the experimental spectral transmittance data using SCOUT software accord with the results of polycrystalline ZnO:Al thin films calculated by other methods which suggest the excellent fit of the models used. As can be seen from (Figure 5a), with variation of the oxygen gas flow rate the refractive index varies as well as the extinction coefficient but in a less pronounced way. It was with the increase of the oxygen flow rate, ZnO:Al thin films tend to be non-stoichiometric and the packing density increases which results in an

increase of the refractive index ([27, 28]. It is well known that the refractive index is closely related to the density of materials, being lower at lower density. With further argumentation of the oxygen gas flow rate, the number of oxygen atoms may be more than the number of zinc atoms and the film tends again to be non-stoichiometrical which exhibited a lower degree of crystallinity as depicted in Figure 1, which results in an increase of the refractive index [28]. Similar results were observed for films prepared from ceramic target, refractive indices were found to increase with the increase of Ar gas flow ratio (Figure 5b). The reason behind the decrease might be due to the increase in carrier density which results in the shift of both plasma frequency and the band gap to higher energies, similar to that reported for ITO films [32].

Figure 5a Dependence of refractive indices and extinction coefficients of

ZnO:Al films on O2 flow rates.

In the visible region, refractive index n decreases with wavelength which correlates with the corresponding increase in transmittance. The obtained values of n agree well with the values obtained using other methods reported by [28, 29, 30, 31]. This is a typical behavior for a semiconductor and thus shows a normal dispersion. The extinction coefficient k as a function of wavelength of the films is almost constant in the visible region for all the samples indicating weak absorption as evidenced by high transmittance in this region. These films exhibit high extinction coefficient in the plasma edge, suggesting absorption of photons in NIR regions, respectively. However, for both films, the change was small implying little absorption in the visible range, which is consistent with the transmittance spectra results, as shown in Figure

4a and 4b. The result is in agreement with those reported for ZnO:In [29].

Figure 5b Dependence of refractive indices and extinction coefficients of

ZnO:Al films on argon flow rates.

4. Conclusion In the present work ZnO:Al thin films were deposited onto soda lime glass (SGL) slides by reactive and non-reactive DC magnetron sputtering. The effects of target composition (Zn:Al (2% Al) metal alloy and ZnO:Al (2% Al) ceramic targets on optical constants of ZnO:Al thin films were studied. XRD scans of the ZnO:Al films indicated (002) crystals were predominantly grown and were more intense under lower O2 and Ar flow rate. The as deposited ZnO:Al films were highly conductive. However, films with lowest electrical resistivity of 2.04 × 10-4 and highest carrier concentration and mobility of 2.98 x 1021

cm-3 and 15.97 cm2/Vs, respectively, were obtained from ZnO:Al films prepared using alloy target. All the films showed more than 80% transparency in the entire visible region. The values of n and k were found to vary significantly with deposition parameters, whereby for different samples ranged from 1.59 to 2.2 and 0.00013 to 1.2194 respectively. Nevertheless, the obtained values are in range of the values reported for ZnO:Al films. Despite similar optical constants and being in range with those reported in literatures, these results confirm that the films with good optical properties can be achieved using metal alloy target, since good electrical together with structural properties were obtained using metal alloy target. This would also suggest that ZnO:Al thin films prepared from metal alloy can be used as an appropriate TCO material in CIGS solar cells since they are deposited at lower temperature 200 oC, recommended for CIGS solar cells.

a

b

Nanoporous Water Filtration Using Disc Ceramic Water Filters

J. Ndungu, F.W. Nyongesa, A. Ogacho, B. O. Aduda

Department of Physics, University of Nairobi, P.O. Box 30197-00100, Nairobi, Kenya.

Abstract

Over four billion cases of diarrhea that occur worldwide each year result in about 1.8 million deaths. In Africa, over 35% of the population suffer from diseases related to unsafe water supply and sanitation. Over 38% of Kenyans lack access to safe drinking water. One of the methods of water treatment is the Point of Use Water Treatment Systems (PUWTS), such as ceramic disc water filters, is one of the proven PUWTS methods and has been shown to reduce diarrhea by an average of 80%. Although ceramic water filters have been proven effective for improving water quality, users and implementers often express concern over their inability to produce a sufficient quantity of water due to their slow flow rate. If flow rate could be increased by altering the filter design, it would improve the ceramic filter’s viability as a scalable House Hold Water Treatment System option. In our study, we create a filter design that exploits physics of nanoscale for water purification with a faster flow rate than the traditional design, while maintaining high levels of bacterial reduction. In trying to achieve this goal, new the ratio of combustible material (sawdust) to clay was increased and then firing at different temperatures below 1000 oC. These new filters were tested for bacterial reduction and flow rate. Minor alterations in filter design or raw materials plus the firing temperature can affect the performance of locally produced ceramic disc water filters to the point where their ability to produce safe drinking water is compromised. The results of this research suggest that the mean flow rate for a properly functioning filter (50% sawdust) fired at 950 oC is 1.7 Litres/hour. This flow rate is more than 2.9 litres per day as recommended by World Health Organization on average conditions. This filter also removed more than 99.97% of E. coli.

Keywords: Disc Ceramic water filters; Porosity; Flow rate; permeability; E. coli filtration.

1.0 Introduction Access to safe drinking water and appropriate sanitation are both essential to life. The lack of safe water supply is one of the world’s major causes of preventable morbidity and mortality. The World Health Organization (WHO) also estimates that 884 million people (13% of the world population) live without access to an improved water source that provides 20 L of water per person per day within 1 km of the person’s residence [1].

In an effort to address health and mortality issues associated with the consumption of contaminated water, Dr. Fernado Mazariegos of the Central American Research Institute Guatemala (ICAITI) made the first frustum-shaped ceramic water filter with a colloidal silver coating [2]. The filter relies on gravity for flow of water, and size exclusion for the removal of bacteria and other pathogens for water, the colloidal silver is antimicrobial [3]. Although there are other methods for the treatment of water in

developing countries (boiling, pasteurization, chlorination, flocculation disinfection, solar disinfection, biosand filter [3, 4] point of use filtration is one of the most promising solutions available [3]. Ceramic water filters are most appealing because of their low cost easy to fabricate and use, and their ability to filter out bacteria from water effectively. About 500000 people in the developing world have adopted some form of porous ceramic filter technology [3].Clay based ceramic water filters are usually produced by mixing of clay, sawdust and water. Other combustible organic materials, such as rice husk, coffee husk, or flour can also be used [5].

The flow rate which may depend on the water turbidity conditions decreases gradually with increasing filter use during the 2-3 year recommended life time of the ceramic water filters. This is due to the accumulation of solids on the inner surface and gradual blockage of the pores by trapped contaminants [6]. During use, PFP recommends that the CWF be cleaned by fire heating, scrubbing and

29 26

much preferable electrical characteristics, whereas lower sheet resistance and resistivity and higher mobility and carrier concentration have been obtained at a substrate temperature of 200 oC. This demonstrates that DC sputtering from alloy target is better suited for TCO since it can be easily obtained at low temperature with good electrical properties.

3.3 Optical Properties

The optical properties were controlled by adjusting Ar gas flow ratio and O2 flow ratio as a mixture of Ar+O2 gases for ceramic and alloy targets, respectively. The transmittance of the films prepared from both targets was higher than 80 % with oscillatory character due to interference effects in the visible region from 300 to 800 nm with a sharp fall in the transmittance at the absorption edge. Plasma frequency shifted to higher wavelength with increase in O2 flow ratio in the mixture of Ar+O2 gasesand Ar flow ratio for films prepared from alloy and ceramic targets, respectively, and showed excellent NIR shielding as can be seen in Figure 4a,b. This might be due to zinc adsorption and substitutional doping of an Al3+ at the Zn2+ site creating one extra free carrier in the process which occurs at a substrate at low gas flow ratios which leads to the increase in carrier concentration [13, 15, 20].As is generally known an increased carrier concentration can give increased absorption in the long wavelength region due to free electron gas. Similar results for the films prepared using alloy target were reported by [13, 15, 21,24]. The films from both targets show typical optical characteristics of TCOs whereas they have a transmission window in the 400 – 1500 nm wavelength range.However, the highest average optical transmittance was achieved using ceramic target and was calculated to be ~80.14 %. The difference arises in near normal reflectance in the NIR region where the films prepared from alloy target showed high NIR reflectance, higher than 80%. The band gap shift was also observed where the band gap was shifted from 3.49 to 3.43 eV with decrease in Ar gas flow rate, while for films prepared from alloy; the band gap was increased from 3.34 to 3.54 eV with decrease in O2 gas flow rate. This is possibly due to the Burstein-Moss shift which is attributed to an increase in the carrier concentration as it is seen section 3.2. This increase in carrier concentration blocks the lowest states (filled states) in the conduction band from absorbing the photons [24]resulting in effective widening of the band gap energy. The shift in absorption edge might also be due to the presence of residual compressive stress as reported by [26].

Figure 4a Optical transmittance and reflectance spectra of ZnO:Al films

grown at various O2 flow rates as a mixture of Ar + O2 gas at a substrate

temperature of 200 oC.

Figure 4b Optical transmittance and reflectance spectra of ZnO:Al films

grown at various Ar gas flow rates at a substrate temperature of 300 oC.

The optical constants (refractive index, n and extinction coefficient, k) as obtained by fitting the experimental spectral transmittance data using SCOUT software accord with the results of polycrystalline ZnO:Al thin films calculated by other methods which suggest the excellent fit of the models used. As can be seen from (Figure 5a), with variation of the oxygen gas flow rate the refractive index varies as well as the extinction coefficient but in a less pronounced way. It was with the increase of the oxygen flow rate, ZnO:Al thin films tend to be non-stoichiometric and the packing density increases which results in an

increase of the refractive index ([27, 28]. It is well known that the refractive index is closely related to the density of materials, being lower at lower density. With further argumentation of the oxygen gas flow rate, the number of oxygen atoms may be more than the number of zinc atoms and the film tends again to be non-stoichiometrical which exhibited a lower degree of crystallinity as depicted in Figure 1, which results in an increase of the refractive index [28]. Similar results were observed for films prepared from ceramic target, refractive indices were found to increase with the increase of Ar gas flow ratio (Figure 5b). The reason behind the decrease might be due to the increase in carrier density which results in the shift of both plasma frequency and the band gap to higher energies, similar to that reported for ITO films [32].

Figure 5a Dependence of refractive indices and extinction coefficients of

ZnO:Al films on O2 flow rates.

In the visible region, refractive index n decreases with wavelength which correlates with the corresponding increase in transmittance. The obtained values of n agree well with the values obtained using other methods reported by [28, 29, 30, 31]. This is a typical behavior for a semiconductor and thus shows a normal dispersion. The extinction coefficient k as a function of wavelength of the films is almost constant in the visible region for all the samples indicating weak absorption as evidenced by high transmittance in this region. These films exhibit high extinction coefficient in the plasma edge, suggesting absorption of photons in NIR regions, respectively. However, for both films, the change was small implying little absorption in the visible range, which is consistent with the transmittance spectra results, as shown in Figure

4a and 4b. The result is in agreement with those reported for ZnO:In [29].

Figure 5b Dependence of refractive indices and extinction coefficients of

ZnO:Al films on argon flow rates.

4. Conclusion In the present work ZnO:Al thin films were deposited onto soda lime glass (SGL) slides by reactive and non-reactive DC magnetron sputtering. The effects of target composition (Zn:Al (2% Al) metal alloy and ZnO:Al (2% Al) ceramic targets on optical constants of ZnO:Al thin films were studied. XRD scans of the ZnO:Al films indicated (002) crystals were predominantly grown and were more intense under lower O2 and Ar flow rate. The as deposited ZnO:Al films were highly conductive. However, films with lowest electrical resistivity of 2.04 × 10-4 and highest carrier concentration and mobility of 2.98 x 1021

cm-3 and 15.97 cm2/Vs, respectively, were obtained from ZnO:Al films prepared using alloy target. All the films showed more than 80% transparency in the entire visible region. The values of n and k were found to vary significantly with deposition parameters, whereby for different samples ranged from 1.59 to 2.2 and 0.00013 to 1.2194 respectively. Nevertheless, the obtained values are in range of the values reported for ZnO:Al films. Despite similar optical constants and being in range with those reported in literatures, these results confirm that the films with good optical properties can be achieved using metal alloy target, since good electrical together with structural properties were obtained using metal alloy target. This would also suggest that ZnO:Al thin films prepared from metal alloy can be used as an appropriate TCO material in CIGS solar cells since they are deposited at lower temperature 200 oC, recommended for CIGS solar cells.

a

b Acknowledgements The University of Dodoma and Solar Energy Group, Physics Department, University of Dar es Salaam, are greatly acknowledged for full financial assistance to the current work. Prof. Julius Mwabora of the University of Nairobi is thanked for his total involvement in Optical measurement and research guidance when in Nairobi. The assistance in XRD measurements and characterization from Marques Hueso, Jose, School of Engineering & Physical Sciences, Heriot-Watt University, Edinburgh EH14 4AS, United Kingdom and Umakanth Dammalapati of Physics Department, The University of Dodoma, Tanzania are highly acknowledged.

References

1. Li L N, Zhao Y, Chen X L, Sun J and Zhang, X D, Physics Procedia 32, (2012), 687 – 695.

2. Minami T, Nanto H and Takata S, Japanese Journal of Applied Physics. 24 (8), (1985), L605-L607.

3. Schaefer C, Brave G and Szczyrbowsk C, Surface and Coating Technology. 93, (1997),39.

4. Granqvist C G 2007 Transparent Conductors as Solar Energy Materials: A Panoramic Review. Solar Energy Materials & Solar Cells. 91, (2007), 1529–1598.

5. Reddy R. S, Sreedhar A, Reddy S A, Uthanna S, Adv. Mat. Lett., 3 (3): (2012), 239-245.

6. Sittinger V, Dewald W, Werner W and Szyszka B, Photovoltaics International. 6: (2010),108-115.

7. Ehrmann N and Reineke-Koch R, Thin Solid Films.519: (2010), 1475-1485.

8. Kim S, Lee W and Lee C, Materials Science and Technology. 23 (3), (2007), 303–306.

9. Kuo C C, Liu C C, He S C, Chang J T and He J L, Journal of Nanomaterials., 562701, (2012), 1-8.

10. Huang C H, Cheng H L, Chang W E and Wong M S, Journal of the Electrochemical Society. 158 (5): (2011), H510-H515.

11. Beyer W, Hüpkes J and Stiebig H, Thin Solid Films.516: (2007), 147–154.

12. Wang H, Xu J, Ren M and Yang L., Journal of material science: Matter Elecron. 19:, (2009), 1135-1139.

13. Sittinger V, Ruske F, Pflug A, Dewald W, Szyszka B and Dittmar G, Thin Solid Films. 518:, (2010), 3115–3118.

14. Fernández S and Naranjo F B, Solar Energy Materials & Solar Cells. 94:, (2010), 157–163.

15. Hwang D, Ahn J, Hui K, Hui K N and Son Y,Journal of Ceramic Processing Research. 12 (2):,(2011), 150-154.

16. Alnajjar A A 2012 ZnO:Al Grown by Sputtering from two Different Target Sources: A comparison Study. Advances in Condensed Matter Physics. 2012:, (2012), 682125.

17. May C, Menner R, Strumpfel J, Oertel M and Sprecher B, Surface and Coatings Technology. 169-170:, (2003), 512-516.

18. Hong R J, Jiang X, Szyszka B, Sittinger V and Pflug A, Applied Surface Science. 207:,(2003), 341-350.

19. Suchea M, Christoulakis S, Katsarakis N, Kitsopoulos T and Kiriakidis G, Thin Solid Films.515:, (2007), 6562-6566.

20. Sittinger V, Dewald W, Werner W and Szyszka B, Photovoltaics International. 6:, (2009), 101-108.

21. Theiss, W. (2008). Hard and Soft Copy for Optical Spectroscopy, Aachen, Germany. (www.mtheiss.com).

22. Minami T, Nanto H and Takata S., Japanese Journal of Applied Physics. 24 (8):, (1985), L605-L607.

23. Li L N, Zhao Y, Chen X L, Sun J and Zhang, X D.,Physics Procedia 32., (2012), 687 – 695.

24. Liu H, Avrutin V, Izyumskaya N, Özgür Ü and Morkoç H., Superlattices and Microstructures. 48:,(2010), 458–484.

25. Dong-Joo Kwak, Kang-Il Park, Byung-Sub Kim and Su-Ho Lee, Journal of the Korean Physical Society.45(1):, (2004), 206-210.

26. Özgür Ü, Alivov Y I, Liu C, Teke A, Reshchikov M A, Doğan S, Avrutin V, Cho S J and Morkoç H.,Journal of Applied Physics. 98, (2005), 041301.

27. Yuste D G, Chen Y C, Yang T And Xie Y P, Adv. Mat. Lett. 255:, (2012), 9279.

28. Jung Y K, Seo J K, Lee D W, Jeo D Y, Thin Solid Films. 445:, (2003), 63

29. Li-Ping P, Liang F, Wei-Dong W, Xue-Min W and Li L., Chin. Phys. B. 21 (4):, (2012), 047305.

30. Bandyopadhyay S, Paul G K, Sen S K., Solar Energy Materials and Solar Cells, 71:,(2002), 103–113.

31. Xue S W, Zu X T, Zheng W G, Deng H X and Xiang X., Physica B: Condensed Matter. 381 (1-2):, (2006), 209-213.

32. Qiao Z, Agashe C and Mergel D, Thin Solid Films.496:, (2006), 520 – 525.

27 28

brushing to remove caked impurities that form with time [3]. The removal of chemical contaminants by the CWFs, such as arsenic, iron and fluoride have also been reported [3].

Although porous CWFs have been used successfully in the field [3, 5, 7, 8] for over a decade, scientific understanding of the effects of porosity on the water flow rate and microbial filtration efficiency is still very limited. There is therefore the need for scientific studies of the effects of porosity on the water filtration properties of Disc Ceramic Water Filters (DCWFs). This paper presents the results of experimental study of the effect of porosity on water flow and filtration characteristics of nanoporous ceramic disc water filter, without a coating of colloidal silver. 2.0 Experimental Method and Materials 2.1 Materials and processing Disc Ceramic Water Filters (DCWFs) were made by mixing red clay soil from Nyeri County composed of Iron and kaolinite clay with sawdust consisting of a mixture from cypress, rosewood and pine trees. In order to produce filters with different porosities, four different proportions of clay to sawdust by volume (45:55; 50:50; 55:45 and 60:40) were used to produce DCWFs. Before mixing the sawdust and the clay were sieved using 100-1000 mesh wire sieve. Chemical analysis was also done on the red clay soil at the Ministry of Roads (Materials Testing and Research Department) (Table1). The initial blending of sieved sawdust and clay was then done manually to ensure a thorough mixing and to avoid the formation of clustered pores. This was done for approximately 10 minutes before adding 2.85 L of water. This was done in 3 portions of 0.95 Litres. Wet mixing was done for approximately 10 minutes per addition of water before the dough was ready for pressing.

Approximately 0.75 kg of the blended mixture was then manually formed into ball, which was pressed tightly together so that no cracks were visible on its surface. This ball was compacted in a two piece aluminium mold that was smeared with oil to prevent the green ware from sticking to the walls of the mold during pressing. The mixture was placed into the female part of the mold before applying a pressure of 140 kPa to the male part of the mold using a 50 Ton hydraulic press. After pressing the green disc filters were dried in laboratory at a room temperature of 26 oC for 8-14 days depending on the mixture ratio of sawdust to clay. After drying, the green membranes

were sintered in an electric furnace at the Department of Physics, University of Nairobi. This involved the preheating of the green membranes to 450-550 oC for 3 hours (to burn off the sawdust), followed by heating to the sintering temperature 950 oC in the furnace. The initial heating rate of 50 oC per hour was increased to 100 oC per hour beyond a furnace temperature of 200 oC. The green wares were sintered for 5 hours at a peak temperature of about 950 oC. They were then left in the furnace to cool to room temperature. The disc shaped ceramic water filters had a radius of 15 cm and a thickness of 2.0 cm. After cooling, the porosities of the DCWFs were characterized using the Archimedes Immersion Technique. The flow rates and bacterial filtration characteristics of the filters were also determined.

2.2 Measuring of porosity The porosities of the DCWFs were characterized using the Archimedes Immersion technique. This technique involved boiling filters in water for three hours. The apparent porosity was then calculated using the relation in equation 1 below [9]:

(1)

Where P=apparent porosity; Ws= saturated filter weight; Wd=dry filter weight and ρw=density of water. 2.3 Flow Rate experiment Prior to the water flow experiment, the DCWFs were saturated by complete submersion in water for about 12 hours to remove internal air bubbles. The flow rates were then measured through the four DCWFs(45:55, 50:50, 55:45, and 60:40). The flow rates were obtained by measuring the volume of water discharged from the DCWFs as a function of time. Ahole approximately 13 cm in diameter was drilled at the bottom of the top bucket and then the DCWFs being tested was fitted at the bottom of a plastic bucket (see Figure 1 below) using Silicone clear paste (Hi-temp clear RTV Gasket maker 100% silicone rubber).

Figure 1. A Filter fixed at the bottom of the top bucket

Effect of Sn Doping on the Electrical Properties of as Prepared and Annealed ZnO Thin films Prepared by

Reactive Evaporation

P. K. Nyaga1, R. J. Musembi2 and M. K. Munji1

1Department of Physics, Kenyatta University, P.O. Box 43844-00100 GPO Nairobi- Kenya 2Department of Physics, University of Nairobi, P.O. Box 30197- 00100 GPO Nairobi-Kenya

Abstract

Layers of transparent and conductive (Tin) Sn-doped Zinc oxide (ZnO) have been prepared by reactive evaporation on glass substrates. The deposition has been done at various doping levels ranging from 1% to 8%. Annealing of samples was done using Rapid Thermal Processing (RTP). In this work, Nabertherm Programmable Furnace system was used and annealing done at 300 oC for one hour. Electrical characterization has been done for both as prepared and annealed samples using four point probe configuration at room temperature (25 oC) to obtain the sheet resistance. The sheet resistance for tin doped zinc oxide reduced with increase in tin doping to a minimum of 11.92 Ωcm at 4% tin doping for as prepared samples and 11.89 Ωcm for annealed samples.

Keywords Zinc oxide, doping effect, electrical properties, reactive evaporation, annealed sample, sheet resistance

1. Introduction Due to their excellent structural and optical properties, zinc oxide (ZnO) thin films have wide applications as solar cells.[1], gas sensors [2], light emitting diodes (LED’s), laser systems [3] and transparent electrodes [4]. Moreover, they can be prepared by different techniques such as magnetron sputtering [5], reactive evaporation [6], chemical vapour deposition (CVD) [7], pulsed laser deposition (PLD) [8] and spray pyrolysis [9]. The electrical properties of ZnO can be improved using different techniques. Doping is a very useful way for turning as ZnO to an electronic material [10]. Many dopants such as group in I, V and rare-earth elements have been used. Among the I, V and rare-earth group elements, doping with (Tin) Sn and (Antimony) Sb elements can result in interesting structural, optical and electrical properties. For example, Sn-doping can not only result in a p-type material through substitution of oxygen, but can also be a source of donor electrons which introduces deep states in the band gap and results in reducing the gap[10].

In the present study, we have investigated the effect of tin doping on the electrical properties of as prepared and

annealed ZnO thin films. The thin films were prepared using the thermal evaporation method.

2. Materials and methods 2.1. Doping of Zinc Oxide with Tin (ZnO:Sn)

Tin doped ZnO film was prepared from Zinc granules and Tin pellets. Zinc granules were 99.9% pure while the purity of Tin pellets was 99.5%. Zinc was Tin doped in various percentages by mass starting from 1% to 8%. This was arrived at by calculations from the relative molecular mass of the compound. The ratios of the various samples that were made are shown in table 2.1. Zinc granules were mixed with Tin pellets as per the various doping concentrations and then heated in presence of a constant flow of Argon gas in a glass tube. The resulting alloy was cooled naturally to room temperature forming ingots.

33 30

At the start of each experiment, the water saturated DCWF was filled completely with approximately 10 Liters of water and covered with a lid to prevent evaporation which may lead to reduction of mass. 2.3 Permeability The permeability of the porous DCWFs was calculated using Darcy’s equation which is given by[3]:

µ (2)

Where Q= flow rate; k= permeability of the material; A= surface area; L= thickness of the material; µ=dynamic viscosity of the fluid; and ΔP=pressure difference from the top to the bottom of the surface.

2.4 E-coli Removal Experiments Highly contaminated surface water was filtered using the DCWF membrane and was then tested for E-coli. All the equipment used in this technique were sterilized each and every time in use. 100 ml sample of the effluents from all the filters were put in various Petri dishes and placed in the incubator at 35 oC for 24 hours. The number of colonies forming unit (CFU) was counted under magnifying glass and expressed as CFU/100 mL. The indicator organisms level in water sample was expressed as he number per 100 ml the number of the indicator organism’s in water tested was determined using the relation:

У (3) Where ψ=number of indicator organisms per 100 ml; У=number of indicator organisms counted and M=milliliters of sample 3.0 Results and Discussion 3.1 Effect of volume fraction of sawdust on the porosity of the DCWF membrane The porosities determined from the Archimedes Immersion technique showed that the porosity of the DCWF increased with increasing volume fraction of sawdust (Figure 2).

Figure 2. Variation of porosity and the volume fraction of sawdust used in DCWF fabrication.

3.2 Effect of volume fraction of sawdust on the flow rate and permeability of the DCWF membrane The pore size and pore size distribution of a filter membrane are important in determining the filter’s efficiency at removing pathogens from water the porosity and permeability are important in determining the rate of fluid flow through the filter. The most porous of the filters studied, the 45:55 DCWF, exhibited the fastest discharge. The slowest filter in flow rate in the study was the 60:40 DCWF. Therefore, the rate of water flow by DCWF increases with the volume fraction of sawdust used in making the filter. The flow rates obtained for the 50:50 DCWF were on average of between 1.37 and 1.94 Litres/hour during the first few hours. This is the range that is typical of PFP filters. Flow rate for the 55:45 and 60:40 DCWFs were below this level on average. The flow rate for 45:55 DCWF was far much high than all other filters. It ranged between 2.83 and 5.80 Litres/hour. This is unusual flow rate and could not give good efficiency of the filter membrane. The amount of water through all the designs of the DCWF in the study decreased with time. This was attributed to the pressure head decreasing plus also the clogging during multiple flow experiments and possibly the effect of the degree of saturation of the DCWF prior to testing. The permeability of the filters studied were found to be between 1.8 x10-4 to 8.3x10-3 m2. Fig3 shows the variation of permeabilities of DCWFs and the percentage volumes of sawdust.

Figure 3. Variation of permeability and the volume fraction of sawdust used in DCWF fabrication. 3.3 Bacterial Filtration E. coli filtration experiments were performed on the two DCWFs with the fastest flow rates 45:55 DCWF had more than 10 coliforms bacteria detected in 100 mL of drinking water. The efficiency of these filters

was 86%. The 50:50 DCWF had an efficiency of 99.97% while 55.45 and 60:40 had 99.996 and 99.9998% respectively.

40 45 50 5515

20

25

30

35

Volume percentage of sawdust (%)

Percen

tag

e o

f p

oro

sity

(%

)

35 40 45 50 55 600

0.002

0.004

0.006

0.008

0.01

Volume percentage of sawdust(%)

Per

mia

bil

ity

(m

2)

Table: Chemical Analysis on Red Clay Soil (Soil pH=5.08)

Thus the removal efficiency of the 60:40 DCWF was slightly higher than that of the other filters. Only the three designs of the four types of DCWFs met the WHO standard for water treatment [3] that neither E.coli nor coliform bacteria should be detectable in 100 mL of drinking water. The filtration efficiency of the PFP filter has been reported to be approximately 99.99% [3, 5].

It implies that point- of –use DCWFs with different porosities can be used to filter out most of the bacterial pathogens in water. With E. coli removal rates approximately 99.9% for all the filters, the use of the DCWFs can contribute to the removal of microbial pathogens from drinking water in the developing world, where more than 4900 people die every day from the effects of consuming contaminated water.

3.4 Chemical Analysis The chemical analysis of the raw clay elements and the fired samples are presented in table1. The raw clay was rich in Silica (2.48%), Iron (18.04%) and Alumina (1.73%). There was an overall decrease in the composition of the elements and fluxes with increasing firing temperature (Table 1). This is due to transformation of various fluxes in the clay upon

firing at high temperatures. Iron is the major constituent of the red clay and it is responsible for the red color of the soil. The Al2O3 is also a major constituent of the clay mineral kaolinite (Al2O3.2SiO2.2H2O). 4.0 Conclusion This paper presents the results of an experimental study of the water flow and E. coli removal efficiency of porous DCWFs produced by the sintering of well controlled mixtures of red clay and sawdust. The porosity, permeability and overall flow rates increase with increasing volume fraction of sawdust. An average rate of 1.7 Litres/hour was obtained from the DWCFs with a sawdust volume fraction of 50%. This filter also removed more than 99.97% of E. coli.

Acknowledgments The author would like to thank the supervisors Dr. Nyongesa, Prof. Aduda and Dr. Ogacho for providing guidance throughout the study, Mr. Muthoka for his continued support in the laboratory, during collection and analysis of the data. References 1. World Health Organization (WHO). (2008).

“Drinking water.” http:// whqlibdoc.who.int/publications/2008/9789241563673_part3_eng.pdf (Apr. 17, 2009).

2. Potters for Peace (PFP). (2008). “Potters for Peace.” http://www. pottersforpeace.org(Aug. 28, 2008).

3. Yakub, Ismaiel; Anand Plappally; Megan Leftwich,; Karen Malatesta,;Katie C. Friedman; Sam Obwoya,; Francis Nyongesa,; Amadou H. Maiga,; Alfred B. O. Soboyejo; StefanosLogothetis; and Wole Soboyejo, American Society of Civil Engineers (2013).

4. Clasen, T., W. P. Schmidt, T. Rabie, I. Roberts & S. Cairncross, BMJ, 334, (2007), (7597), 782.

5. Oyanedel-Craver, V. & J. Smith, Environmental Science and Technology, (2008).

6 Brown, J., M. Sobsey & D. Loomis, Am J Trop Med Hyg, 79(3), (2008), 394- 400.

7. Albert, J., Luoto, J., and Levine, D, Environ. Sci. Technol.,(2010), 4426-4432.

8. Van Halen, D. (2006) Ceramic silver impregnated pot filters for household drinking water treatment in developing countries. Sanitary Engineering Section, Department ofWater Management, Faculty of Civil Engineering. Delft University of Technology, Delft.

9. Crimshaw W., The Chemistry and Physics of Clays. Ernest Ben Ltd, London,(1991);30-39.

lement Percentage in raw

clay

After Firing at

950 oC

Iron as Fe2O3,

%m/m

18.04 15.06

Silica as SiO2,

%m/m

2.48 1.92

Aluminum as

Al2O3 %m/m

1.73 1.49

Organic content,

%m/m

2.96 -

Potassium as

K2O, %m/m

2.95 -

Sodium Oxide as

Na2O, %m/m

4.04 -

Magnesia as

MgO, %m/m

3.17 3.08

Calcium as CaO,

%m/m

3.83 2.48

Others 9.90 8.76

Insoluble

Residue, %m/m

54.94 -

31 32

Table 1 Zinc oxide sample showing tin doping concentrations by percentage.

Sample name Tin doping level for ZnO (%)

F UndopedG 1H 2I 3J 4K 5L 6M 7N 8

2.2. Deposition of thin films

The films were prepared using resistive evaporation method in an Edwards vacuum deposition set up model type Auto 306. The Auto 306 uses a turbo molecular pump and its fully automatic in pumping procedures. The pressure is monitored using a compact gauge capable of recording atmospheric to sub atmospheric pressure. The deposition procedure is as follows:

The material to be deposited was loaded on to molybdenum boat. The substrate was placed on a sample holder which is connected to a rotator. This means as the material is being deposited; the substrate can be rotated so that there is uniformity of the coated film. The vacuum chamber was pumped down to base pressure of 4.0×10-5 mbars. After reaching the above pressure, deposition commenced by first applying a current of 3 A gradually and oxygen allowed at the flow rate of 20 standard cubic centimetres (sccm). After allowing few seconds for pre deposition the shutter was removed so that coating on the substrate could start. After deposition, the vacuum chamber was vented and the samples were removed. Figure 2.1 shows a schematic of evaporation process used to prepare tin doped zinc oxide thin films.

Figure 1. Schematic diagram of the deposition process

2.3. Electrical characterization

The sheet resistivity of the prepared and annealed films of Tin doped Zinc Oxide was measured using the four point probe method at room temperature (25oC). The measurements were made through four contact terminals at each corner of each corner of the thin film using a Keithley 2400 source meter. A high impedance current source was used to supply current through the two probes 1 and 2 and a voltmeter was to measure the voltage across the probes 3 and 4 as shown in Figure 2.2. The values of sourced current and measured voltages were used to determine the sample resistivity. Typical probe spacing between X1 and X2 is about 1mm.

2.4. Elemental composition

The elemental composition of the films was done using XRay Fluorescence (XRF) spectroscopy. This is a widely used method for environmental, geological, industrial and other types of samples. This method provides a wide range of concentrations from 100% to few parts per million. A main disadvantage is that analysis is generally restricted to heavier elements than fluorine.

Figure 2. A schematic diagram showing the arrangement used to measure sheet resistance.

2.4. Annealing of ZnO:Sn thin films

Samples were annealed using Naberthem Programmable furnace at 300 oC under high vacuum (10-5mbars) for 30 minutes. The samples were left to cool for 24 hours then their sheet resistivity was compared with the un annealed samples.

3.0. Results and Discussion3.1. Electrical resistivity of as prepared ZnO:Sn films

The electrical resistivity of ZnO:Sn for the as prepared films for various doping concentrations were measured and are presented in Table 2 below. Thickness of the films was in the range of 90-150 nm.

Table 2. Variation of resistivity with different doping level for as prepared samples

Sample name

Tin doping level in ZnO(%)

Resistivity at room temperature ±0.10 Ωcm

1 0 35.292 1 20.093 2 19.824 3 19.07

Optimization of SnxSey Deposited by Reactive Thermal Evaporation for Solar cell Application

L. K. Munguti1, R.J. Musembi2, W.K. Njoroge1

1Department of Physics, Kenyatta University, P.O. Box 43844, GPO, Nairobi- Kenya.2Department of Physics, University of Nairobi, P.O. Box 30197- 00100 GPO, Nairobi-Kenya.

AbstractIn this study, tin selenide was prepared at different ratios using tin and selenium pellets in glass tube filled with argon and then heated up to 350°C. The resulting materials were cut into ingots which were used in preparation of thin films by thermal evaporation. The evaporation was done using Edwards Auto 306 Coating Unit. The chamber pressure was maintained at 5.0 x 10-5mbars during the film deposition. Thin films of tin-selenide produced using various ratios of tin to selenide were characterized for optical properties and sheet resistance. The optical measurements were done using Solid State 3700 DUV UV-VIS-NIR spectrophotometer in the visible range (380-750nm) and the transmittance spectra data obtained were analyzed using the SCOUT software. The films with ratio of 1:1 showed the highest transmittance of 85% with a band gap energy obtained as 1.40 eV. The electrical characterization measurements were carried out using a Four Point Probe at room temperature (25°C) to obtain the sheet resistivity. The resistivity ob-tained for the films was 20.1Ω cm.

Keywords: Tin selenium, Optical properties, Band gap, Electrical properties, Sheet resistance.

1. Introduction Extensive attention has been paid in search of new semi-conducting materials for efficient solar energy conversion. Group IV-VI offer a range of optical band gaps suitable for various optoelectronic applications such as memory switching devices, solar cells, holographic recording sys-tems and gas sensors [1]. Tin selenide is a direct band gap group IV-VI semiconductor with a melting point of 860°C and an energy gap of about 1 eV [2]. A number of tech-niques have been employed in the formation of high quality thin films such as Chemical Vapor Deposition, Molecular Beam Epitaxy, Electrochemical Atomic Layer Epitaxy, Thermal Evaporation and Sputtering [3]. However, there is an interest to investigate other techniques, which could be new possibilities in terms of device properties and structure. Thermal Evaporation belongs to these alternative tech-niques that could also produce high quality films of IV–VI semiconductor materials [4]. It is an inexpensive, simple, low temperature method that could produce good quality films for device application. However, the effect of the ratio of tin to selenium on the optical and electrical properties has not been investigated. In this study, thin films of tin sele-nide consisting of various ratios of tin to selenium were deposited by reactive thermal evaporation. Optical and electrical characterization of films were carried out.

2. Materials and methods 2.1. Sample preparation

2.1.1. Tin selenide (SnxSey)

Tin selenide (SnxSey) compound was prepared from tin and selenium pellets. The purity given by the supplier was 99.9% for tin while that of selenium was 99.99%. The alloy was prepared by mixing tin and selenium in their specific ratios by mass as shown in Table 2.1 in a sealed glass tube containing argon gas and the mixture was heated to a tem-perature of 350°C. This is because tin selenide (SnxSey) can be formed by heating tin in contact with selenium either in vacuum or in inert atmosphere above 200°C [5]. The heating was continuous and gentle to prevent the glass from breaking. During the melting shaking was continuously done to en-sure homogeneity. The temperature was measured using a thermocouple thermometer. The compound formed was allowed to cool naturally to room temperature (25°C) thus solidifying into an ingot. The setup that was used in prepa-ration of the alloy ingots is shown in Figure 2.1.

Table 2.1. Tin selenide samples showing various tin to selenium ratios.

Sample name Sn:SeA 1:0.2B 1:0.4C 1:0.6D 1:0.8E 1:1.0

37 34

5 4 15.436 5 16.137 6 17.358 7 17.359 8 20.18

Table 2 shows a general drop in resistivity up to a doping level of 4%. This is attributed by increase in concentration caused by Sn ions substituting the Zn2+ ions [2]. Sheet resistance of the films also decreases with increasing doping concentration.

The high values of sheet resistivity noted are as a result of Tin dopant being oxidized during deposition process hence losing its doping effect. During deposition, variation of phase and structure as a result of changing deposition parameters like chamber pressure and temperature results in a phase mixing consisting of SnO and SnO2 in addition to the intended ZnO:Sn. These unintended phases influence electrical and structural properties in the film. More pronounced is the increase of sheet resistivity.

3.2. Electrical resistivity for annealed ZnO:Sn films

Table 3 below shows the resistivity of the annealed Tin doped ZnO at various doping levels.

Table 3. Variation of resistivity with different doping level for annealed samples

Sample name

Tin doping level in ZnO (%)

Resistivity at room temperature ±0.10 Ωcm

1 0 23.512 1 16.303 2 14.404 3 13.825 4 11.896 5 12.117 6 13.028 7 14.209 8 15.20

The surface morphology of ZnO:Sn thin films is strongly dependent on the annealing concentration of the dopant. At an annealing temperature of 300 oC, the sheet resistivity reduced with 4% Tin doping concentration recording the lowest resistivity of 11.89 Ωcm. The sheet resistance of a semiconductor depends on the surface morphology of material and annealing decreases the microstructure of the film and makes the material denser [11]. Comparing with the un annealed samples, annealed samples had low sheet resistivity. This shows that the crystal quality of the films is highly improved on annealing. This therefore makes a structure more crystalline. Sheet resistivity increased above 4% tin doping concentration reaching a maximum of 15.20 Ωcm at 8%.

Since the ionic radius of Tin (0.38Å) is smaller than of the Zinc ion (0.6Å), the Tin atoms doped into a ZnO lattice act as donors by supplying two free electrons when the Tin ions occupy Zn ions sites. This in turn increases the free carrier concentration and hence decreases the resistivity. In order to maintain electrical neutrality, two negative electrons are induced to compensate the excess positive charges. Hence

the resistivity decreases due to increasing free electrons in the film. In the concentration between 1% and 4% of Sn dopants, little decrease of resistivity is found. However, when more than 4% of Sn was added, the resistivity gradually increased.

3.3. Elemental composition of ZnO:Sn thin films

The analysis for elemental composition ZnO:Sn was carried out using an X-Ray Florescence (XRF) minipal machine which was computer controlled. Figure 3 shows the peak elemental composition for ZnO:Sn doping ratio.

Figure 3. Peak elemental composition for ZnO:Sn

The shaded peaks are the elemental composition. For Zn peaks were detected at energies of 1.18KeV, 8.42KeV and 9.6KeV. The peak X is the detector escape which is a spectrum for detector peak.

4. Conclusion Deposition of thin films of ZnO:Sn has been done by reactive thermal evaporation technique. The effect of Sn concentration on the electrical properties for as prepared and annealed samples was investigated. The sheet resistance for prepared films reduced from 20.09Ωcm at 1% tin doping reaching a minimum of 15.43Ωcm at 4% tin doping. More doping increased the sheet resistance. Annealing the films at 300 oC reduced the sheet resistance of the films from a maximum of 16.30 Ωcm at 1% Tin doping and minimum of 11.89 Ωcm at 4% Tin doping.

Acknowledgement The Authors of this work would wish to thank the Chief Technologists Department of Physics of both Kenyatta University and University of Nairobi for the assistance during deposition and characterization of the thin films. The Department of mines and geology ministry of environment and natural resources is thanked for the use of XRF spectrometer.

References

1. J.B.Baxter and E.S.Aydil, Solar Energy Materials and Solar Cells, 90, No.5, (2006), pp.607-622.

2. S.T Shishiyanu, T.S. Shishiyanu and O.I.Lupan,”Sensors and Actuators B:chemical,.107,No.1, (2005),pp. 379-386.

3. J.M.Szarko, J.K. Song, BC.W.Lackledge, I.Swart, S.R. Leone, S.Li, Chemical Physics Letters,404, No.1-3,(2005), pp.171-176.

4. T.Ootsuka, Z.Liu, M.Osamura, Y.Fukuzawa, R.Kuroda, Y.Suzuki, Thin Solid Films, 476, No.1, (2005),pp.30-34.

5. X.Yu, J. Ma, F.Ji, Y. Wang, X.Zhang and H.Ma, Thin Solid Films,.483,No.1-2, (2005),pp.296-300.

6. J.Ma,F.Ji, D.Zhang, H.Ma and S.Li, Thin Solid Films,.357,No2,(1999),pp.98-101.

7. D.A.Lamb and S.J.C.Irvine, Journal Crystal Growth,273, No.1-2, (2004),pp.111-117.

8. J.Zou,S.Zhou,C.Xia,Y.Hang,J.Xu,S.Gu,, Journal Crystal Growth, 280, No. 1-2, (2005), pp.185-190.

9. P.Nunes,E.Fortunato, P. Tonello, F. Braz Fernandez, P.Vilarinho and R.Martins, Thin Films,64,No.3-4,2004,pp.281-285.

10. L.K.Munguti, R.J.Musembi and W.K.Njoroge, Thin film Tech. 61,(2013) 17162-17165.

11. Nadia Chahmat, Ammar Haddad, Azzedine Ain-Souya, Rachid Ganfoudi, Nadir Attaf, Mouhamed Salah Aida, Mokhtar Ghers, Journal of Modern Physics,3, (2012), ,1781-1785.

35 36

Figure 2.1 Setup used for melting tin and selenium to form tin selenide ingot.

Five samples were prepared separately and then the glass tubes were sealed. The glass tubes were broken after they had cooled to room temperature and the ingots were removed. Then from each of the samples labeled A to E, 0.1 g was measured using an electronic balance and placed in a boat for evaporation.

Figure 2.2: Vacuum evaporation chamber.

2.2 Thermal Evaporation

The deposition of the compound material was done using an Edwards Auto 306 Coating Unit. The current was main-tained at a constant value of 3.5A for 3 minutes for all the samples. This ensured that the deposition rate was main-tained at 0.2 nm/sec for a uniform film thickness. Clean microscope glass slides were used as substrates. The glass slides were washed using a mixture of deionized water, liquid detergent and sodium hydroxide in the ratio of 3:2:1 and then dried with a spray of pressurized argon. A typical diagram of the vacuum evaporation chamber is shown in Figure 2.2.

2.3. Optical properties of tin selenide

The raw data for the optical transmittance of the deposited tin selenide thin films was measured using the solid state 3700 DUV UV-VIS-NIR spectrophotometer at normal in-cidence. The data for each of the films was analyzed using the SCOUT software and the respective graphs plotted using origin PRO7 software.

2.4. Electrical characterization

The sheet resistance was carried out using the four point configuration at room temperature (25ºC). The Four Point Probe set up usually consists of four equally spaced tungsten metal tips with finite radius. Each tip is supported by springs on the other end to minimize sample damage during probing. The four metal tips are part of an auto-mechanical stage which travels up and down during measurements. A high impedance current source was used to supply current through the outer two probes and a voltmeter used to measure the voltage across the inner two probes as shown in Figure 2.3.These values of sourced current and measured voltage were used to determine the sample resistivity. Typical probe spacing S was about 1 mm.

Figure 2.3:The Four Point Pprobe linear set up

2.5. Elemental composition of the films.

To determine the elemental composition of the films, XRF spectroscopy was used. XRF spectroscopy is widely used for the qualitative and quantitative elemental analysis of envi-ronmental, geological, biological, industrial and other types of samples. Compared to some competitive techniques such as Atomic Absorption Spectroscopy (AAS), Inductively Coupled Plasma Spectroscopy (ICPS) and Neutron Activa-tion Analysis (NAA), XRF has the advantage of generally being non destructive, multi-elemental, fast and cost effec-tive. It also provides a uniform detection limit across a large portion of the periodic table and is applicable to a wide range of concentrations, from 100% to few parts per million. A main disadvantage is that analyses are generally restricted to heavier elements than fluorine.

3.0. Results and Discussion 3.1. Introduction

This chapter presents the data obtained from all experi-mental procedures and the various analyses carried out on the results. The trend of various graphs is also accounted for in their respective experiments.

3.2. Optical spectra of the films

3.2.1. Optical transmittance spectra of tin selenide (Snx-

Sey) thin films.

A plot of transmittance against wavelength for the raw data of various samples is shown in the Figure 3.1. Gener-ally the samples of tin selenide had low transmittance (60%)

[5] K.Bindu and P. K. Nair, Semiconductor Science Technology,19 (2004), pp. 1348–1353.

[6] H. Abdulla Norazia, S. Shaari, M. Z. Nuawi and N. S. Mo-hamed, American J. of Engineering and Applied Sciences, 3(1) (2010) pp. 171-179.

[7] K. C. Park, D. Y. Ma and K. H. Kim, Thin Solid Films, 305 (1997), pp. 201-209.

[8] M. Vaezi and S. Sadrnezhaad, Material Science Engineer-ing, 141 (2007), pp. 23-27.

[9] N. Kumar, U. Parihar, R. Kumar, K. Patel, J. Panchal and N.Padha, American Journal of Material Science, 2(1) (2012), pp. 41-45.

[10] J. Poortmans and V. Arkhipov (2006). Thin film solar cells fabrication, characterization and applications. 1st ed., John Wiley and Sons. Chinchester.

41 38

in the visible range (380nm-780nm) except for the sample with ratio of 1:1 whose transmittance was 85% near 700nm. This was due to high absorption as a result of a narrow band gap. The fundamental absorption edge was between 350 nm and 650 nm although there was a shift of the absorption edge towards the shorter wavelength. The increase in carrier concentration from tin atoms blocks the lowest states in the conduction band [6]. This could be attributed to Burstein Moss effect. Burstein pointed out that an increase in Fermi level in the conduction band of degenerate semiconductor leads to the widening of the energy band (blue shift) [7].The undulating pattern in the transmittance curves are caused by interference of light in the thin film itself [8].

Figure 3.1: The variation of transmittance against wavelength for various ratios of tin to selenium. The transmittance increases as the ratio approaches 1:1.

3.2.2. Optical reflectance spectra for tin selenide (SnxSey)thin films.

Figure 3.2: Optical reflectance spectra of tin selenide. Low reflectance indicated that the films had high absorption making the films suitable for the absorber layer.

Reflectance data (Figure 3.2) shows that the average re-flectance obtained within the visible range was less than 40%. Therefore, tin selenide thin films exhibit high absorption behavior which makes it a good absorber material for solar cell applications.

3.3.3. Optical constants and band gap

The raw data was analyzed using the SCOUT software which uses various physical models to generate its own numerical values through simulations. The simulated data was fitted into the experimental data to obtain the optical constants such as refractive index (n), absorption coefficient (α) and the optical band gap (Eg). The data obtained was plotted using the Microcal Origin Lab™ software and the graphs such as the one in Figures 3.3 were obtained.

Figure 3.3: Simulated versus experimental graph for Sn:Se =1:0.6

The optical band gap dependence of the absorption coef-

ficient is given by the equation (3.1) where direct transitions were assumed.

α2(hν)=A(hν –Eg) (3.1)

Figure 3.4: Extrapolation for the linear part for Sn:Se = 1:0.8. The band gap obtained was 1.60eV.

where A and Eg are constant and optical band gap re-

spectively. The band gap values were obtained by plotting (αhv)2 against energy, E (hv) (eV). The linear part of the graph was extrapolated to the point (αhv)2 = 0 on the x-axis, this gives the energy gap Eg. A sample graph is shown in Figure 3.4.

200 400 600 800 1000 1200-10

0102030405060708090

1000 2 4 6 8 10

0

2

4

6

8

10

Tran

smitt

ance

(%)

Wavelength (nm)

Sn:Se 1:0.2 1:0.4 1:0.6 1:0.8 1:1.0

400 600 800 1000 12000

20

40

60

80

1000 2 4 6 8 10

0

2

4

6

8

10

Refle

ctan

ce %

Wavelength (nm)

Sn:Se 1:0.2 1:0.4 1:0.6 1:0.8 1:1.0

200 400 600 800 1000 1200

0

10

20

30

40

50

0 2 4 6 8 10

0

2

4

6

8

10

Tran

smitt

ance

(%)

Wavelength(nm)

SIMULATED EXPERIMENTAL

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.20.0

2.0x1012

4.0x1012

6.0x1012

8.0x1012

1.0x1013 0 2 4 6 8 10

0

2

4

6

8

10

(h

cm2 e

V

Energy (eV)

Eg = 1.60eV

The band gap for various samples of tin selenide were obtained as in Table 3.1.

Table 3.1: The variation of tin to selenium ratios with the band gap. The

band gap reduced as the ratio of tin to selenium approached 1:1.Sample name Sn:Se Ratio Band gap(eV) at room

temp(25°)A 1:0.2 2.04B 1:0.4 1.80C 1:0.6 1.75D 1:0.8 1.60E 1:1.0 1.40

The values of band gap obtained were the same as those obtained directly from the SCOUT software. Although the values of band gap energy obtained were high, they were found to be within the range of those obtained by other re-searchers [9] who deposited tin selenide by a similar method, (thermal evaporation), whose band gap energy values were 1.24 eV to 1.74 eV.

3.4. Absorption coefficient for tin selenide

A plot of absorption coefficient against wavelength from 300nm to 1200nm is shown in Figure 3.5. The absorption coefficient was high at lower wavelengths but decreased with increasing in wavelength. This shows that tin selenide has a high absorption coefficient in the visible range (380 -750 nm) which is an important property for forming a good absorber layer for solar cells. In addition the average ab-sorption was more than 2.0 x 105 cm-1which agrees with the values in literature of between 104-105 cm-1 [10]

Figure 3.5: The absorption coefficient against wavelength. There was a decrease in the absorption coefficient at longer wavelength.

3.5. XRF results

The elemental analysis was carried out using the pana-lytical XRF spectrometer and the data obtained direct from a computer using MiniPal software.

3.5.1. XRF spectrum for tin selenide (SnSe)

Figure 3.6: XRF spectrum for optimized SnSe (ratio 1:1)

The yellow colored spectrum in figure 3.6 shows the

composition of the elements present in the thin films at Kα,Kβ, Lα and Lβ lines, respectively. The elements present were Sn (51%) and Se (49%). The bottom line represents the Background Radiation which shows the radiations detected by the XRF machine but are not present in the thin film samples. These radiations may have originated from the machine itself, or and the room in which the measurements were taken. The highest colorless peak represents the de-tector escape which is a spectrum for detector peak (Rho-dium) which is the element that the machine is made of.

4.0. Conclusion Deposition of thin films of SnxSey was done by Thermal

Evaporation Technique. Although the transmittance of the films was generally low, it increased with increasing the amount of selenium in the films reaching a value of 85% for optimized ratio within the visible range. The band gap for the optimized films was 1.4 eV while the resistivity of the films also decreased as the ratio of tin to selenium approached 1:1 reaching the lowest level of 20.1Ω cm.

ACKNOWLEDGEMENTThe authors of this work want to thank the Chief Geologist

Department of Mines and Geology Ministry of Environment and Natural Resources for the use XRF spectrometer.

REFERENCES [1] G. H. Chandra, J. N. Kumar,N. M. Rao and S. Uthama,

Journal of Crystal Growth, 306 (2007) pp.68-78.

[2] R. Teghil, A. Santagata, V. Marotta, S. Orlando, G. Pizzella, A. Giardini-Guidoni and A. Mele. Applied Science, 90, No.9(1998), pp. 505-514.

[3] N. A. Okereke, and A. J. Ekpubi. Chalcogenide Letters, 7(2010), pp. 531-538.

[4] C. Wang, W. X. Qian, X. M. Zhang, Y. Xie and Y. T. Qian, Material Research Bulletins, 34 (1999), p.1637.

200 400 600 800 1000 1200-1x104

01x104

2x104

3x104

4x104

5x104

6x104

7x104

8x104

9x104

1x105

1x105

1x105

Sn:Se 1:1.0 1:0.8 1:0.6 1:0.4 1:0.2

Abs

orpt

ion

coef

fici

ent c

m-1

Wavelength (nm)

The low transmittance observed at wavelengths in the range of 250 – 330nm is due partly to absorption in the substrate. From the transmittance spectra in Figure 2, it is noted that the films also exhibits clear thermochromism especially in the near infrared region of the spectrum, where large contrast in transmittance between the two phases is observed. The transmittance peak values for the six samples at 25 oC and 100 oC as per Figure 2 is shown in Table 1.

Table 1. Peak transmittance for VO2 based thin films as derivedfrom Figure 2

Sample (nm) at which the transmittance (T)peak occur ( at 25 oC)

% Tpeak at 25oC

(nm)at which the transmittance (T)peak occur (at 100 oC)

% Tpeak at 100 oC

VO2 658 44.7 654 38VO2 : W 680 34.3 656 23.4VO2 : W

: Al1729 34.4 647 31.6

VO2 : W : Al2

729 54.1 729 49.7

VO2 : W : Al3

683 37.6 700 32.0

VO2 : W : Al4

638 25.4 652 23.8

The strong suppression of reflectance, R ( , T) as shown in Figure 3 could be another reason for the higher transmittance in the visible region. The plotted results, show R ( , T) in the low temperature semiconducting phase being suppressed from a peak value of 35 % for VO2 : W film at λ = 550 nm to about 16 and 14 % at the same wavelength for VO2 :W : Al and VO2 films, respectively. For the co-doped films, reflectance in the visible region varies from 23 % at = 475 to 6 % at = 640 nm. Generally, all the films had reflectance values monotonically increasing in the wavelength range 250 to approximately 500nm and decreasing in the wavelength range 500 < < ~ 650 nm. At λ =2500 nm, the reflectance of VO2 :W : Al is much lower, at approximately 41 % in the metallic phase compared to around 54 % for undoped VO2 films.

0

10

20

30

40

50

60

70

80

0 500 1000 1500 2000 2500

VO2

VO2 : W : Al

VO2 : W

Reflecta

nce (

%)

Wavelength (nm)

100 oC

100 oC25 oC

100 oC

25 oC25 o C

Figure 3: Reflectance spectra of VO2, VO2:W and VO2:W:Al thin films showing suppression in the visible region of the

solar spectrum as a result of co – doping.

3.2 Integrated Luminous, Solar Transmittance and Modulation

Integrated luminous transmittance (Tlum), solar transmittance (Tsol), luminous modulation, Tlum and solar modulation ( Tsol) values were of particular interest in this study for practical applications of VO2- based films, particularly for smart windows applications. Tlum and Tsol values of the optical properties were obtain from:

,

,,,i

ba

iba

iXX (1)

where the integral is evaluated from a = 0.385 to b =0.76 m for luminous transmittance and from a =0.25 to b = 2.5 m for solar transmittance, X is the average transmittance, (Ts + Tp)/2 or average reflectance, (Rs +Rp)/2 for s and p- polarized light for human eye in the wavelength range of 0.385 – 0.76

m , and sol is the solar irradiance spectrum for air mass 1.5 (corresponding to the sun standing 37° above the horizon) Transmittance modulation, Tlum and solar modulation ( Tsol) values are obtained from:

hlumllumlum TTT ,, (3)

and

39 40

Effects of Aluminium and Tungsten Co-doping on the Optical Properties of VO2 Based Thin Films

C. J.Lyobha1,2,a, R.T. Kivaisi2, N. R. Mlyuka2 and M. E. Samiji2

1College of Natural and Mathematical Sciences, Physics Department, The University of Dodoma, P. O. Box 338, Dodoma, Tanzania

2Solar Energy Group, Physics Department, University of Dar es Salaam, P. O. Box 35063 Dar es Salaam, TanzaniaaCorrespondence Author, Email: [email protected]

Abstract

Aluminum and tungsten co-doped vanadium oxide (VO2 : W : Al) thin films were deposited by DC reactive magnetron sputtering technique. In this work, we report on the effects of aluminum and tungsten co – doping on the optical properties of vanadium dioxide (VO2) based thin films with a view of combining both increased luminous transmittance (Tlum) and lowering the transition temperature (τc). The effect of aluminum and tungsten co-doping on semiconductor-metal transition of vanadium dioxide films was investigated, and is compared with tungsten doped and undoped films. Spectral transmittance of the films was obtained using Shimadzu SolidSpec-3700 DUV UV-VIS-NIR spectrophotometer. The results revealed that the transmittance of tungsten and aluminum co–doped vanadium dioxide using two Al pellets was enhanced to more than 54% in the visible spectral range with fairly good switching characteristics and reduced transition temperature to 61 oC.

Keywords: Transition temperature, luminous transmittance and Co-deposition

1. Introduction The excessive use of heating systems on cold

climate and air conditioning systems in warm climate results in extensive use of electricity in order to maintain such systems. This situation calls for new technologies for energy generation and energy conservation in industry, transportation, and building sectors. The building sector is of particular importance since, according to comprehensive study by the United Nations Environment Programme, it accounts for 30 - 40 % of the primary energy used in the world [1]. This energy is spent mainly on heating, cooling, lighting and ventilation.

Energy efficiency fenestration materials and devices have the potential to significantly decrease the energy expenditure in buildings [2, 3]. Chromogenic materials are of much interest for energy saving as they are able to change their optical and electrical properties when subjected to a change

in environment (temperature, light, pressure) [4-6].The most important chromogenic materials are electrochromic, photochromic and thermochromic . Electrochromic materials are the ones which, once incorporated in multi-layer devices, are able to vary the optical properties by electrical charging and discharging. Photochromic materials are the ones coloring under light intensity irradiation and bleaching in the dark. Materials whose optical, electrical and structural properties depend on temperature are called thermochromic materials [2]. The change in optical properties can be in the form of absorption, transmittance, reflectance, or scattering. The change can be either within the visible or beyond visible spectrum [6]. Windows with optical coatings that can adjust their optical properties in response to dynamic needs are called ‘smart windows’ [l, 7].

The low transmittance observed at wavelengths in the range of 250 – 330nm is due partly to absorption in the substrate. From the transmittance spectra in Figure 2, it is noted that the films also exhibits clear thermochromism especially in the near infrared region of the spectrum, where large contrast in transmittance between the two phases is observed. The transmittance peak values for the six samples at 25 oC and 100 oC as per Figure 2 is shown in Table 1.

Table 1. Peak transmittance for VO2 based thin films as derivedfrom Figure 2

Sample (nm) at which the transmittance (T)peak occur ( at 25 oC)

% Tpeak at 25oC

(nm)at which the transmittance (T)peak occur (at 100 oC)

% Tpeak at 100 oC

VO2 658 44.7 654 38VO2 : W 680 34.3 656 23.4VO2 : W

: Al1729 34.4 647 31.6

VO2 : W : Al2

729 54.1 729 49.7

VO2 : W : Al3

683 37.6 700 32.0

VO2 : W : Al4

638 25.4 652 23.8

The strong suppression of reflectance, R ( , T) as shown in Figure 3 could be another reason for the higher transmittance in the visible region. The plotted results, show R ( , T) in the low temperature semiconducting phase being suppressed from a peak value of 35 % for VO2 : W film at λ = 550 nm to about 16 and 14 % at the same wavelength for VO2 :W : Al and VO2 films, respectively. For the co-doped films, reflectance in the visible region varies from 23 % at = 475 to 6 % at = 640 nm. Generally, all the films had reflectance values monotonically increasing in the wavelength range 250 to approximately 500nm and decreasing in the wavelength range 500 < < ~ 650 nm. At λ =2500 nm, the reflectance of VO2 :W : Al is much lower, at approximately 41 % in the metallic phase compared to around 54 % for undoped VO2 films.

0

10

20

30

40

50

60

70

80

0 500 1000 1500 2000 2500

VO2

VO2 : W : Al

VO2 : W

Reflecta

nce (

%)

Wavelength (nm)

100 oC

100 oC25 oC

100 oC

25 oC25 o C

Figure 3: Reflectance spectra of VO2, VO2:W and VO2:W:Al thin films showing suppression in the visible region of the

solar spectrum as a result of co – doping.

3.2 Integrated Luminous, Solar Transmittance and Modulation

Integrated luminous transmittance (Tlum), solar transmittance (Tsol), luminous modulation, Tlum and solar modulation ( Tsol) values were of particular interest in this study for practical applications of VO2- based films, particularly for smart windows applications. Tlum and Tsol values of the optical properties were obtain from:

,

,,,i

ba

iba

iXX (1)

where the integral is evaluated from a = 0.385 to b =0.76 m for luminous transmittance and from a =0.25 to b = 2.5 m for solar transmittance, X is the average transmittance, (Ts + Tp)/2 or average reflectance, (Rs +Rp)/2 for s and p- polarized light for human eye in the wavelength range of 0.385 – 0.76

m , and sol is the solar irradiance spectrum for air mass 1.5 (corresponding to the sun standing 37° above the horizon) Transmittance modulation, Tlum and solar modulation ( Tsol) values are obtained from:

hlumllumlum TTT ,, (3)

and

45 42

Attention of many researchers since 1959, when F. J. Morin first observed its remarkable metal - to - insulator transitions upon cooling or heating through a critical temperature τс of about 68 oC [8]. VO2 is technologically important due to its ability to undergo a fully reversible metal-to-semiconductor phase transition. The conversion of the low temperature monoclinic phase VO2 (M) to the high temperature rutile phase VO2 (R) is associated with significant changes in electrical conductivity and optical properties especially in the near-infrared region [8-10]. The VO2 (M) is a semiconductor and infrared transparent at room temperature, but above τс VO2

(R) becomes metallic and infrared reflecting [11-13].Vanadium dioxide thin films have been revealed from many studies as one of the potential materials for fabrication of practical smart windows [14 -18].So far all these investigations did not result in producing VO2 films which sufficiently accomplish the demand for practical applications, particularly, demand for high transmittance in the visible spectral range and a transition temperature near room temperature. The challenge would be overcome if the thermochromic properties of doped VO2 thin films, namely, optical, electrical and structural are improved. The τс of VO2 films have been reported to be lowered by several techniques such as tungsten (W), molybdenum (Mo), niobium (Nb) or rhenium (Re) doping or by introduction of stress [19], but these dopants showed lowered optical contrast [20,21]. Doping with elements known to form wide band gap oxides such as magnesium (Mg),aluminum (Al),titanium(Ti), could yield improved transmittance [17,22,23].

This paper reports the recent results on determining the combined effect of Al and W doping of VO2 thin films geared to obtain both improved luminous transmittance and lowered transition temperature

2. Film Deposition and Characterization

Thin films of undoped vanadium dioxide (VO2), tungsten doped vanadium dioxide (VO2 : W) and aluminium and tungsten co-doped vanadium dioxide (VO2 : W : Al) were deposited by DC reactive magnetron sputtering in an argon/oxygen atmosphere using BALZERS BAE 250 coating unit. The respective thin films were made from metallic V and V – W alloy targets stuck with aluminium pellets. V - target was 99.9% pure, 5.1 cm in diameter and 0.6 cm thick. The alloy target had percentage

composition of vanadium and tungsten; V 99% - W1%. Previous to film deposition the sputter chamber was evacuated first by pumping down to a pressure less than 5.0 x10-2 mbar using rotary pump and then the turbomolecular pump was turned on. The chamber was initially evacuated down to a pressure of ~ 3 × 10-6 mbar before the substrates were heated to a temperature of about 450 oC. The gases of purities 99.999 % and 99.9 % for argon and oxygen respectively were introduced into the chamber at the rates of 75 and 6.6 – 7.2 mln/min respectively. The optimum oxygen rate was 7.0 mln/min for the best films. The deposition time and power were fixed at 25 minutes and 200 W respectively. The working pressure was about 5.8 - 6.1 × 10-3 mbar. VO2, VO2 :W and VO2 : W : Al thin films were deposited on well cleaned normal soda lime glass substrates. KLA - Tencor Alpha Step IQ surface profiler was used to measure film thicknesses, which were typically 150 nm.

To obtain VO2 : W : Al thin films, a number of high-purity, 99.99% aluminium pellets were cut into pieces with diameter, length and mass of 6 mm, 5 mm and 0.32 g respectively. Those sizes and mass were optimum for the switchable films. The metal pieces were placed at the centre, over the tungsten doped vanadium (V : W) target surface so that both elements could be co-sputtered allowing a homogeneous dispersion of the dopant elements in the film. In order to obtain films with different aluminium concentrations, the number of aluminium pellets was varied. The maximum number able to produce switchable film within one run was four.

Electrical measurements were performed on films in order to verify reversible temperature-dependent properties. Sheet resistance of samples in kilo - ohms per square (kΩ/) was measured by pressing a two point probe, connected to a Fluke 77III Multimeter, against the films. The probe was locally constructed and connected to the digital multimeter. A heating plate was used to heat the samples. The temperature was measured by using a digital thermometer whose tip was placed over against the film surface to sense its surface temperature. The change in resistivity was recorded for both heating and cooling cycles from around 25 oC to 100 °C. Figure 1 shows sheet resistance in kilo - ohms per square (kΩ/) for VO2

thin film with thickness 150 nm measured from an AJANDEL Model RM3-AR test unit in combination with a four point probe within the interval 25 < T < 100 oC. The figure indicates a reversible change with drop in the electrical resistance transition of about two orders of magnitude at τc = 68 C , which is similar to the one observed in high-quality

polycrystalline VO2 thin films [13] but less than that reported by [24,15]. The sheet resistance of VO2 film drawn in log scale shows a temperature hysteresis width approximately equal to 12 °C. This is comparable to results from rf -sputtered VO2 film [15] and other data from the literature [11,16].

1 x 103

1 x 104

1 x 105

1 x 106

20 40 60 80 100

VO2

Shee

t res

istan

ce (O

hm/sq

uare

)

Temperature (oC)

Figure 1 Sheet resistance versus temperature for a VO2 thin films ( 150 nm thick, deposited at 450 oC)

The optical/thermochromic properties have been evaluated using Shimadzu SolidSpec-3700 DUV UV-VIS-NIR spectrophotometer and Perkin Elmer Spectrum BX FT-IR spectrophotometers inserted with a locally made sample heating-cooling cell. It has been done by measuring the spectral normal transmittance and reflectance at the UV-VIS - NIRrange, from 250 to 2500 nm, below and above the transition temperature. The determination of the transition temperature was carried out by evaluating the optical transmittance change with temperature at a NIR wavelength of λ = 2500 nm. The phase transition temperatures τc, was then estimated by determining the average between temperature at mid point of the transmittance – temperature curve during heating and cooling cycles.

3. Results and Discussions 3.1. Transmittance and Reflectance of VO2 Based Thin Films

The analysis of the optical behavior of VO2, VO2 :W doped and VO2 : W : Al thin films was mainly on spectral transmittance T ( ) and reflectance R ( )

values obtained from the Shimadzu SolidSpec-3700DUV- VIS-NIR spectrophotometer. The optimum film thickness for all samples was found to be 150 nm. The transmittance and reflectance were measured in the 2500250 nm while temperature-dependence transmittance was done at = 2500 nm. For the high temperature metallic phases of VO2

based thin films, the samples were heated from room temperature of the order of 25 to 100 o C using a heating cell. Figure 2 shows the spectral transmittance curves of six samples, VO2, VO2 : W : Al1, VO2 : W : Al2, VO2 : W : Al3 and VO2 : W : Al4 for comparison corresponding to the two phases of VO2 based thin films, the semiconductor phase, at 25 oC and metallic phase at 100 oC

0

10

20

30

40

50

60

70

80

0 500 1000 1500 2000 2500 3000

VO2

VO2 : W

VO2

: W : Al1

VO2 : W : Al2

VO2 : W : Al3

VO2 : W : Al4

Tran

smitt

ance

(%)

Wavelength (nm)

25 oC

100 oC

100 oC

25 oC100 oC

Figure 2. Spectral transmittance for VO2, VO2:W, VO2:W:Al1,

VO2:W:Al2, VO2:W:Al3 and VO2:W:Al4 thin films at 25

and 100 oC ( 150 nm thick, deposited at 450 oC).

The analysis revealed that the co-doped samples had higher transmittance in the visible region compared to W- doped and undoped VO2 films. The best being the one co- doped with two aluminumpellets (VO2 : W : Al2). This film had the highest transmittance peak of 54 % at 729 nm. The result agrees with those reported by other researchers [3]. Generally, the transmittance measurements at

729250 nm in both below and above the transition temperature, show monotonous increase in transmittance with wavelength. At room temperature, the transmittance of undoped VO2 film rises sharply from ~ 0 % at 250 nm to a peak of 44 % at 658 nm; the VO2 : W reaches a peak of 34 % at 680 nm. The spectral transmittance data at room temperature (semiconducting state) and temperatures above transitions (metallic state) are comparable to those obtained by other researchers [16, 25].

43 44

polycrystalline VO2 thin films [13] but less than that reported by [24,15]. The sheet resistance of VO2 film drawn in log scale shows a temperature hysteresis width approximately equal to 12 °C. This is comparable to results from rf -sputtered VO2 film [15] and other data from the literature [11,16].

1 x 103

1 x 104

1 x 105

1 x 106

20 40 60 80 100

VO2

Shee

t res

istan

ce (O

hm/sq

uare

)

Temperature (oC)

Figure 1 Sheet resistance versus temperature for a VO2 thin films ( 150 nm thick, deposited at 450 oC)

The optical/thermochromic properties have been evaluated using Shimadzu SolidSpec-3700 DUV UV-VIS-NIR spectrophotometer and Perkin Elmer Spectrum BX FT-IR spectrophotometers inserted with a locally made sample heating-cooling cell. It has been done by measuring the spectral normal transmittance and reflectance at the UV-VIS - NIRrange, from 250 to 2500 nm, below and above the transition temperature. The determination of the transition temperature was carried out by evaluating the optical transmittance change with temperature at a NIR wavelength of λ = 2500 nm. The phase transition temperatures τc, was then estimated by determining the average between temperature at mid point of the transmittance – temperature curve during heating and cooling cycles.

3. Results and Discussions 3.1. Transmittance and Reflectance of VO2 Based Thin Films

The analysis of the optical behavior of VO2, VO2 :W doped and VO2 : W : Al thin films was mainly on spectral transmittance T ( ) and reflectance R ( )

values obtained from the Shimadzu SolidSpec-3700DUV- VIS-NIR spectrophotometer. The optimum film thickness for all samples was found to be 150 nm. The transmittance and reflectance were measured in the 2500250 nm while temperature-dependence transmittance was done at = 2500 nm. For the high temperature metallic phases of VO2

based thin films, the samples were heated from room temperature of the order of 25 to 100 o C using a heating cell. Figure 2 shows the spectral transmittance curves of six samples, VO2, VO2 : W : Al1, VO2 : W : Al2, VO2 : W : Al3 and VO2 : W : Al4 for comparison corresponding to the two phases of VO2 based thin films, the semiconductor phase, at 25 oC and metallic phase at 100 oC

0

10

20

30

40

50

60

70

80

0 500 1000 1500 2000 2500 3000

VO2

VO2 : W

VO2

: W : Al1

VO2 : W : Al2

VO2 : W : Al3

VO2 : W : Al4

Tran

smitt

ance

(%)

Wavelength (nm)

25 oC

100 oC

100 oC

25 oC100 oC

Figure 2. Spectral transmittance for VO2, VO2:W, VO2:W:Al1,

VO2:W:Al2, VO2:W:Al3 and VO2:W:Al4 thin films at 25

and 100 oC ( 150 nm thick, deposited at 450 oC).

The analysis revealed that the co-doped samples had higher transmittance in the visible region compared to W- doped and undoped VO2 films. The best being the one co- doped with two aluminumpellets (VO2 : W : Al2). This film had the highest transmittance peak of 54 % at 729 nm. The result agrees with those reported by other researchers [3]. Generally, the transmittance measurements at

729250 nm in both below and above the transition temperature, show monotonous increase in transmittance with wavelength. At room temperature, the transmittance of undoped VO2 film rises sharply from ~ 0 % at 250 nm to a peak of 44 % at 658 nm; the VO2 : W reaches a peak of 34 % at 680 nm. The spectral transmittance data at room temperature (semiconducting state) and temperatures above transitions (metallic state) are comparable to those obtained by other researchers [16, 25].

hsollsolsol TTT ,, (4)

where l and h denote low and high - temperature corresponding to, respectively, semiconductor and metallic phases of VO2 thin films.

Calculations for Tlum, and Tsol were done based on ASTM G173-03 Reference Spectra derived from SMARTS v. 2.9.2 taken in the wavelength range 385 760 and 280 2500 nm, respectively. The calculations show that the VO2 : W : Al2 film exhibit high integrated luminous transmittance (Tlum, l = 39.3 %, Tlum, h = 36.6 %) and luminous modulation ( Tlum

= 2.6 %, from Tlum, l = 39.3 % to Tlum, h = 36.6 %), compared to VO2 (Tlum, l = 30.3 %, Tlum, h = 26.8 %)and its luminous modulation ( Tlum = 3.4 %, fromTsol, l = 30.3 % to Tsol, h = 26.8 %). Solar transmittance, Tsol, of VO2 : W : Al2 film in both low and high phases are larger than those of VO2 thin film. The result signifies also that the solar modulation of VO2 : W : Al2 thin film is also larger compared to that of VO2 thin film. Tlum, l, Tlum, h, Tsol, l,

Tsol, h, Tlum and Tsol values for other compositions are also shown in the table.

In view of the fact that solar energy in the visible region has a peak at 550 nm, the Tlum across the metal-insulator transition, MIT, effectively influences the Tsol [27]. For instance, Tsol increased by 6.8 %, from 9.1 % for undoped VO2 films to 15. 9 % for VO2 : W : Al2 films. On the other hand, doping with W decreased the Tsol by 2.6 % from 9.1 % for undoped films to 6.5 % for W doped films. Furthermore, VO2 : W : Al2 films had the highest solar modulation compared to the other co-doped films, with Tsol being 15.9 % compared to 2.85 % for VO2 : W : Al1 films, 10.9 % for VO2 : W : Al3 and 6.6 % for VO2 : W : Al4 films.

3.3 Effect of Aluminum and Tungsten Co-doping on the Hysteresis and Transition Temperature of the VO2 Thin Films From Transmittance Measurements

Figures 4 (a) – (e) show the temperature- dependent transmittance at = 2500 nm for VO2:W, VO2:W : Al1, VO2:W : Al2, VO2:W : Al3 and VO2:W : Al4 thin films. From these Figures, transition temperatures, τc, were determined, defined as the average between temperature at mid point of the transmittance – temperature curve during heating

and cooling cycles. The VO2 film has the highest τcof approximately 68 oC compared to 32 oC, 42.6 oC, 61.3 oC, 61.1 oC and 47 oC for VO2:W, VO2:W:Al1, VO2:W:Al2, VO2:W:Al3 and VO2:W:Al4 thin films, respectively.

5

10

15

20

25

30

35

40

30 40 50 60 70 80 90 100

VO2 : W

Tra

nsm

ittance

(%

)

Temperature ( oC )

Figure 4 a: Transmittance as a function of temperature for tungsten doped vanadium dioxide thin film at =

2500 nm (~150 nm thick, deposited at 450 oC).

5

10

15

20

25

30

35

40

30 40 50 60 70 80 90 100

VO2 : W : Al1

Tra

nsm

ittan

ce (

%)

Temperature (oC)

Figure 4.a: Transmittance as a function of temperature for VO2 : W:Al1 thin film at = 2500 nm (~150 nm thick,

deposited at 450 oC)

Transparent and Conducting TiO2 : Nb Thin Films Prepared by

Spray Pyrolysis Technique

M. J. Mageto1,3*, C.M. Maghanga1, M. Mwamburi2, H. Jafri1, 1Department of Engineering Sciences, The Ångström Laboratory, P.O.Box 534, SE-75121 Uppsala,Sweden 2Department of Physics, Moi University, P.O. Box 1125, Eldoret, Kenya 3Department of Physics, Masinde Muliro University of Science and Technology, P.O. Box 190, 50100, Kakamega, Kenya.

[email protected] Abstract

Sputtering and Pulsed Laser Deposition (PLD) techniques have been employed successfully to fabricate highly conducting and transparent TiO2:Nb (TNO) films. In this article, we have demonstrated that transparent and conducting Nb:TiO 2 films can be made by the Spray Pyrolysis Technique. The films were deposited on Corning 7059 glass substrates at 500 15˚C using an alcoholic precursor solution consisting of titanium (iv) isopropoxide and 5NbCl . The influence of increasing Nbconcentration on the

electrical, optical and structural properties was investigated. The minimum resistivity, 3.36 -310 Ω cm, for 2xx-1 ONbTi film (x = 0.15) was obtained after 1 hour post deposition annealing in hydrogen atmosphere at 500˚C. The x-ray diffraction of hydrogen annealed films showed a polycrystalline anatase (004)-oriented phase without any second phases. The optical band gap for undoped and doped films lay in the range 3.38 – 3.47 eV. Using dispersion analysis, optical constants were determined from spectro-photometric measurements for films on glass.

Keywords: spray pyrolysis; titanium dioxide; transparent conductor; doping.

1. Introduction

Titanium dioxide ( 2TiO ) thin films have been investigated extensively in recent years owing to their potential applications in areas such as photo-catalysis [1], electro-chromism [2], solar cells [3], self cleaning windows[2], gas sensors[2], optical wave-guides[3] etc. Transparent conducting oxides (TCOs) [2] such as fluorine doped tin oxide (FTO), tin doped indium oxide (ITO), aluminum doped zinc oxide (AZO) have attracted much attention both in fundamental research and device applications such as flat panel displays (FPDs), touch panels, light emitting diodes (LEDs), and Si-based solar cells [4]. Since 2005, niobium doped titanium oxide (TNO) has joined the conventional class of TCOs mentioned above [2].

Anatase undoped 2TiO films have been deposited by various techniques [4,5] including sputtering, pulsed laser deposition (PLD),

chemical vapour deposition (CVD)[6], sol-gel[7,8] and spray pyrolysis[ 9]. Sputtering [10-14]and Pulsed Laser Deposition (PLD) [15-19]techniques have been employed successfully to fabricate high conducting and transparent TiO2:Nb (TNO) films [4]. TNO films with resistivity 2 - 3 10-4 Ω cm at dopant concentrations of 3 - 6 at. % have been achieved by PLD [4,18] although resistivity remained on the order of 10-4 Ω cm for dopant concentrations as high as 20 % [18]. These high conductive TNO films are typically degenerate semiconductors: the dopant Nb atoms exist as Nb5+ ions, and release conduction electrons with high efficiency[4]. The ionic radius of 5Nb is 0.64 Å. This is close to that of 4Ti radius (0.60 Å ). Thus Nb is substituted for Ti in the lattice making the doped 2TiO a n-type semiconductor. In this paper, we report the fabrication of low resistivity and transparent TNO films made by the low cost, vacuum-free spray pyrolysis combined with post deposition annealing. The

49 46

hsollsolsol TTT ,, (4)

where l and h denote low and high - temperature corresponding to, respectively, semiconductor and metallic phases of VO2 thin films.

Calculations for Tlum, and Tsol were done based on ASTM G173-03 Reference Spectra derived from SMARTS v. 2.9.2 taken in the wavelength range 385 760 and 280 2500 nm, respectively. The calculations show that the VO2 : W : Al2 film exhibit high integrated luminous transmittance (Tlum, l = 39.3 %, Tlum, h = 36.6 %) and luminous modulation ( Tlum

= 2.6 %, from Tlum, l = 39.3 % to Tlum, h = 36.6 %), compared to VO2 (Tlum, l = 30.3 %, Tlum, h = 26.8 %)and its luminous modulation ( Tlum = 3.4 %, fromTsol, l = 30.3 % to Tsol, h = 26.8 %). Solar transmittance, Tsol, of VO2 : W : Al2 film in both low and high phases are larger than those of VO2 thin film. The result signifies also that the solar modulation of VO2 : W : Al2 thin film is also larger compared to that of VO2 thin film. Tlum, l, Tlum, h, Tsol, l,

Tsol, h, Tlum and Tsol values for other compositions are also shown in the table.

In view of the fact that solar energy in the visible region has a peak at 550 nm, the Tlum across the metal-insulator transition, MIT, effectively influences the Tsol [27]. For instance, Tsol increased by 6.8 %, from 9.1 % for undoped VO2 films to 15. 9 % for VO2 : W : Al2 films. On the other hand, doping with W decreased the Tsol by 2.6 % from 9.1 % for undoped films to 6.5 % for W doped films. Furthermore, VO2 : W : Al2 films had the highest solar modulation compared to the other co-doped films, with Tsol being 15.9 % compared to 2.85 % for VO2 : W : Al1 films, 10.9 % for VO2 : W : Al3 and 6.6 % for VO2 : W : Al4 films.

3.3 Effect of Aluminum and Tungsten Co-doping on the Hysteresis and Transition Temperature of the VO2 Thin Films From Transmittance Measurements

Figures 4 (a) – (e) show the temperature- dependent transmittance at = 2500 nm for VO2:W, VO2:W : Al1, VO2:W : Al2, VO2:W : Al3 and VO2:W : Al4 thin films. From these Figures, transition temperatures, τc, were determined, defined as the average between temperature at mid point of the transmittance – temperature curve during heating

and cooling cycles. The VO2 film has the highest τcof approximately 68 oC compared to 32 oC, 42.6 oC, 61.3 oC, 61.1 oC and 47 oC for VO2:W, VO2:W:Al1, VO2:W:Al2, VO2:W:Al3 and VO2:W:Al4 thin films, respectively.

5

10

15

20

25

30

35

40

30 40 50 60 70 80 90 100

VO2 : W

Tra

nsm

ittance

(%

)

Temperature ( oC )

Figure 4 a: Transmittance as a function of temperature for tungsten doped vanadium dioxide thin film at =

2500 nm (~150 nm thick, deposited at 450 oC).

5

10

15

20

25

30

35

40

30 40 50 60 70 80 90 100

VO2 : W : Al1

Tra

nsm

ittan

ce (

%)

Temperature (oC)

Figure 4.a: Transmittance as a function of temperature for VO2 : W:Al1 thin film at = 2500 nm (~150 nm thick,

deposited at 450 oC)

25

30

35

40

45

30 40 50 60 70 80 90 100

VO2

: W : Al2

Tra

nsm

itta

nce

(%

)

Temperature ( oC)

Figure 4.b: Transmittance as a function of temperature for VO2 : W : Al2 thin film at = 2500 nm (~150 nm thick, deposited at 450 oC).

20

25

30

35

40

45

50

55

30 40 50 60 70 80 90 100

VO2 : W : Al3

Tran

smitt

ance

(%)

Temperature ( oC )

Figure 4.c:Transmittance as a function of temperature for VO2:W : Al3 thin film at = 2500 nm (~150 nm thick, deposited at 450 oC).

10

20

30

40

50

60

30 40 50 60 70 80 90 100

VO2 : W : Al4

Tra

nsm

ittance

(%

)

Temperature ( oC)

Figure 4.d: Transmittance as a function of temperature for VO2:W : Al4 thin film at = 2500 nm (~150

nm thick, deposited at 450 oC).

There is evidence that for the co-doped films, the transition temperature is lower compared to undoped VO2 film as seen in Figure 5. The reduction in transition temperature with increasing doping level of aluminum in VO2 - based thin films have been also reported before by other researchers [23] while other workers reported rising transition temperature with increasing Al doping level [17,28].

20

30

40

50

60

70

VO2

VO2 : W

VO2 : W

: Al1

VO2 : W

: Al2

VO2 : W

: Al3

VO2 : W

: Al4

Transition temperature

Tran

sition

temp

eratur

e(o C)

Thin film composition

Figure 5 Variation in transition temperature for 150 nm thick VO2based thin films deposited at 450 oC, showing decrease in τc in VO2 : W : Al films as compared to undoped VO2film. The data was obtained from temperature Vs transmittance measurements by Shimadzu UV-VIS-NIR SolidSpec-3700 DUV spectrophotometer.

4. Conclusion Thermochromic VO2 thin films were successfully

synthesized by DC reactive magnetron sputtering. From a practical point of view, aluminum and tungsten co-doped VO2 films exhibit promising characteristics with regard to optical transmittance and switching properties. Application of the VO2 : W : Al2 coating enhances the transmittance in the visible spectral range to more than 54 %. With VO2 :W : Al2 the reversible phase transition occurs at 61 °C, compared to 68 °C for undoped VO2film.

Acknowledgements The MSSEESA network, Tanzania Chapter is thanked for financial support which enabled the first author to do optical analysis using Shimadzu SolidSpec-3700 DUV UV-VIS-NIR spectrophotometer at Nairobi University. The University of Dodoma is appreciated for the two years MSc sponsorship.

References [1] United Nations Environment Programme (UNEP)

(2007), ‘‘Buildings and Climate Change: Status, Challenges and Opportunities’’, Paris, France

[2] C M. Lampert and C. G. Granqvist (1990) (Eds), ‘‘Large-Area Chromogenics: Materials and Devices for Transmittance Control’’, SPIE Optical Engineering Press, Bellingham, WA, USA.

[3] C.G. Granqvist. Thin Solid Films, 193 – 194 (1990), p. 730-741.

[4] C. B. Greenberg. Thin Solid Films, 110 (1983), p. 73 – 82.

[5] P. Jin and S. Tanemura. Journal .Appl. Phys, 33(1994) , pp. 1478-1483.

[6] C. G. Granqvist, N. R. Mlyuka, and G. A. Niklasson,.(2010), ‘‘Thermochromic Material and Fabrication Thereof’’; International Patent Publication Number WO 2010/039067 A1.

[7] C. B Greenberg. Thin Solid Films, 252 (1994), pp. 81 - 93.

[8] F. Morin. Physical Review Letters, 3(1) (1959), pp.33 - 34.

[9] W. H. Verleur, A. S. Barker and C. N. Berglund, ,Physics Review Letter, 172(3) (1968), pp. 788 - 798.

[10] K. D. Rogers. Nasa Astrophysic Data System, 8(1993), pp. 240 - 244.

[11] E. E. Chain. Applied Optics, 30(19) (1991), pp. 2782 – 2787.

[12] J. Livage. Coordination Chemistry Reviews, 190 –192 (1999) , pp. 391 - 403.

[13] V. Msomi, and O. Nemraoui. South African Journal of Sciences, 106(207) (2010) , p. 1 - 4

[14] C. G. Granqvist, (1991), ‘‘In Energy Efficiency Windows: Present and Forthcoming Technology’’, Edited by C G Granqvist, Materials Science for Solar Energy Conversion Systems, Sweden, Pergamon Press, p.140.

[15] R. T. Kivaisi, and M. E. Samiji, M. E. Energy Materials and Sol. Cells, 57 (1999) , pp. 141 – 152.

[16] N. R. Mlyuka, (2003), Structural Stability of Vanadium Dioxide Under Different Ambient Conditions, A Thesis for The Degree of Master of Science (Physics), University of Dar es Salaam, Tanzania.

[17] N. R. Mlyuka, (2010), Vanadium Dioxide Based Thin Films: Enhanced Performance For Smart Window Applications, Phd Thesis, University of Dar es salaam, Tanzania.

[18] Blackman, C. S. and Parkin, I. P. (2010), ‘‘Thermochromic Coatings II: Internation Patent number 2010/0270519.

[19] K. Kato, P. K. Song, Y. Shigesato, and H. Odaka, Proceedings of SPIE, 4458 (2001) , pp. 261 -267.

[20] F. Béteille, and J. Livage, J. Sol-Gel Sc. Tech. 13(1) (1998), pp. 915 - 921

[21] M. A. Sobhan, R. T. Kivaisi, B. Stijerna, and C. G Granqvist. Solar Energy Materials and Solar Cells,44 (1996), pp. 451 - 455.

[22] M. Soltani, M. Chaker, E. Haddad., R. V.Kruzelecky, and J. Margot. Applied Physics Letters, 85 (11) (2004) , pp. 1958- 1960

[23] A. Gentle and G. B. Smith J. Phys. D. Appl. Phys., 41 (2008), pp. 015402.

[24] D. Kucharczyk.and T. Niklewski. Journal of Applied Crystallography, 12 (1979), pp. 370 - 374

[25] V. Msomi (2008), Vanadium Dioxide Thermochromic Thin Films Correlation Between Microstructure, Electrical and Optical Properties, Thesis for degree of Master of Sciences, University of Zululand

[26] C. Batista, R. M. Ribeiro and V. Teixeira. Nanoscale Research Letters, 6 (2011), pp. 301 - 305.

[27] Y. Gao, L. Kang, Z. Chen, and H. Luo (2011), Nanofabrication. Solution to Processing of Nanoceramic VO2 Thin Films for Application to Smart Window. Edited by Masude, Y. ISBN 978 –953 – 307. InTech, p. 167 – 198.

[28] J. B. MacChesney and H. J. Guggenheim. J. Phys. Chem. Solids, 30 (1969) , pp. 225 – 234.

47 48

influence of increasing Nb concentration on the electrical, optical and structural properties was investigated. This paper is organized as follows: Section 2 describes the experimental setup for the preparation of the films including sample preparation, post deposition annealing and characterization techniques. Section 3 is devoted to experimental results including chemical composition, structural, optical and electrical properties. An analysis of the optical data is given; results are reported for optical constants and band gaps. This Section also covers film characterization with x-ray diffraction and scanning electron microscopy. Section 4, summarizes the main results.

2. Experimental Procedure

2.1 Sample Preparation

The substrates were cleaned in an ultrasonic bath containing soap detergent for 15 minutes and the procedure was repeated using pure ethanol instead of detergent. Undoped titanium oxide and niobium-doped titanium oxide films were deposited onto pre-cleaned 0.5 mm thick

2cm52.5 corning 7059 glass and silicon substrates by spray pyrolysis at a substrate temperature of 500 15˚C. The deposition apparatus consists of a spray chamber, hot plate (substrate heater), and temperature controller, atomizer (spray nozzle) of diameter ~ 1mm,input gas valve, gas compressor, gas flow meter, conduit tube and pressure gauge.

Figure 1: Spray Pyrolysis experimental setup.

Figure 1 shows schematic view of in-house-made spray pyrolysis system . The setup works inside a

fume chamber whose air venting velocity is 0.51 m/s. The undoped 2TiO films were produced from a precursor solution consisting of titanium (iv) isopropoxide (TPT) (97% Alfa Aesar) prepared by mixing 34.7g of TPT with 54ml of 2,4-pentanedione (99% Alfa Aesar). They react to form Titanium diisopropoxide bis(acetylacetonate) and isopropanol [21]. This reaction is exothermic. After cooling, the resulting solution was diluted with 810 ml pure ethanol and 10ml of 35% HCl under standard atmospheric conditions and then stirred at 60 rpm for 5 hours at room temperature. During this stirring process, plastic wrap was used to cover the top of the beaker to prevent reaction with humidity. The resulting solution (labeled T) was divided into several portions each of 200 ml per beaker. Nb - doped 2TiO films were prepared from solutions (labeled N) of (28.4g, 0.11 mol) Niobium (V) Chloride, 5NbCl ,(99% Alfa Aesar) dissolved in 500 ml pure ethanol as well as 10ml of acetic acid, stirred as above and certain different amounts of this solution Nadded to each of the 200 ml of solution T. Since solution N was responsible for the introduction of dopant, the amount of 5NbCl in the solution mixture of T and N was based on the required doping level. The mixture was also stirred as stated above for 30 minutes.

The spraying parameters were as follows: substrate temperature of 500 15˚C was monitored with a thermocouple fixed 1 mm deep in the interior of the substrate holder, directly under the substrate position. Carrier-gas: Nitrogen, Carrier-gas pressure: 2.1 bar, flow rate of solution 7ml/min, nozzle-to-substrate distance: 65 cm horizontally and 40 cm vertically. To prevent rapid cooling of the hot plate temperature, spraying was done in short bursts.

2.2 Post Deposition Annealing

After recording their transmittance, reflectance and X-Ray Diffraction (XRD) spectra, the sprayed samples were annealed in hydrogen ambient at a pressure of 1atm at heating ramp speed of 4°/min with a dwell time of 1 hour at a temperature of 500°C and thereafter cooling at a ramp speed of 4.4°/min, translating to a total of 4.8 hours.

3.2 Structural Studies

20 30 40 50 60 70

0

10000

20000

30000

(103) (213)(105)

(200)

(004)

6.5% Nb

5.7% Nb

2.0% Nb

1.0% Nb

Inte

nsity

(a. u

.)

ngle

Undoped

As deposited(101)

(116)

20 30 40 50 60 70

0

10000

20000

30000

40000

50000

(105)(103)

6.5% Nb

5.69% Nb

2.0% Nb

Inte

nsity

(a. u

.)

Angle (2 ))

Undoped

Hydrogen Annealed

1.0% Nb

(101)

(004)

(200) (213) (116)

Figure. 3: XRD pattern of undoped and Nb-doped

2TiO films on Corning 7059 glass substrates before and after post deposition annealing in hydrogen at 500°C Thickness of each film was 600 ± 10 nm.

20 30 40 50 60 700

20000

40000

60000

Inte

nsity

Angle (2)

15% Nb

9.3% Nb

9.7% Nb

22% Nb

Annealed in Hydrogen

(101)

(004)

(103)(105)

(213) (116)

Figure 4: XRD pattern of undoped and Nb-doped 2TiOfilms on Corning 7059 glass substrates at before and after post deposition annealing in hydrogen at 500°C Thickness of each film 336 ± 6 nm

Figure. 3 shows the XRD patterns of the undoped and Nb doped (low Nb concentrations) polycrystalline titanium oxide films on Corning 7059 glass substrates before and after post

annealing. The peaks in the spectra are identified as originating from reflections from the (101), (103), (004), (200), (105) planes of polycrystalline anatase titanium oxide. No peaks from starting materials or any residual species were found in the spectra, confirming the proper phase formation of the materials. Before annealing, the (101) peak increases with increasing Nb-doping, but after post annealing the peak is suppressed.

Figure 4 shows the XRD patterns of Nb doped (high Nb concentrations) amorphous and polycrystalline titanium oxide films on Corning 7059 glass substrates before and after post annealing respectively. Note that as Nb concentration was increased to 9.3 at. % Nb, the films turned from crystalline to amorphous structure before annealing. After post annealing in hydrogen, the films crystallized. The intensity of the (004) peak increased remarkably as the Nb concentration was increased to 15 at. % Nb, as shown in Figure 4.

20

21

22

23

24

25

26

27

28

0 5 10 15 20

Cry

stal

lite

Size

(nm

)

Nb content (at. %)

Figure. 5: Variation of grain size for Nb:TiO2 films

annealed at 500 C0 in hydrogen as a

function of doping concentration

53 50

2.3 Sample Characterization

The crystal structure and phase characterization were determined using a Siemens D5000 X-ray Diffractometer (XRD). The glazing incident angle was 01 in parallel beam geometry with the diffraction angle, 2θ 0, between 020 and 070 .The wavelength, λ , of the αCuK radiation was 1.540598 Å. A high resolution LEO 1550 Scanning Electron Microscope (SEM) with a field emission gun was used to examine grain size, surface and cross-sectional morphologies for different concentrations of Nb in the doped samples on silicon substrates. The elemental chemical compositions of the TiO2:Nb films deposited on Si substrates were determined by Energy Dispersive X-ray Spectroscopy (EDS).The transmittance and reflectance measurements were done at near normal angle of incidence in the solar wavelength range from 0.3 to 2.5 µm ona Perkin-Elmer Lamda 900 UV/VIS/NIR double beam spectrophotometer equipped with an integrating sphere. A barium sulphate film served as reflectance standard. Thickness d, of the film as-deposited and hydrogen annealed samples prepared with different Nb doping concentrations were estimated using the number of interference fringes, from the equation [22].

1221

21

nλnλ2λλd

,

(1)

where 1n and 2n are refractive indices at two adjacent maxima (or minima) at 1λand 2λ respectively and the refractive index, n was determined as described by Swanepoel [22].

22 SNNn

(2) where

21S

TTTT2SN

2

mM

mM

(3)

S denotes the refractive index of the substrate, MT and mT represents the maximum and minimum transmittance envelopes at the fringes. The estimated thickness was verified by fitting the experimental spectral data to theoretical spectral data based on dispersion analysis using the SCOUT software [23] in the wavelength range 0.3 - 2.5 µm and also from SEM cross sections. The thicknesses obtained from the three different methods is in agreement within a discrepancy of not more than 5% and are indicated in XRD spectra below. Dispersion analysis using a model for dielectric susceptibility of the film consisting of two parts: Drude [24] and Kim terms [25] was used to simulate the measured reflectance and transmittance. Drude free electron model accounts for intra-band transitions of the conduction electrons which contribute to the optical properties. This model has two adjustable parameters: plasma frequency, pΩ and damping constant, γ . The plasma frequency is proportional to the square root of the carrier density, the damping constant is proportional to the inverse of the mobility. The Drude dielectric susceptibility Drudeχ , expressed as a function of frequency ω, is given as [24]

iω

Ωωχ 2

2p

Drude

(4)

The one oscillator contribution developed by Kim contains four adjustable parameters: TOΩ resonance frequency, pΩ oscillator strength,

τΩ damping constant, and the so called Gauss–Lorentz-switch constant σ . σmay vary between zero and infinity. For

σ= 0, a Gaussian line shape is achieved. A large value of σ (larger than 5) leads to a Lorentzian line shape. The Kim oscillator models the weak broad interband absorption in the measured wavelength range. The interband dielectric susceptibility described by Kim is given by [25];

ωτωiωΩΩ

χ 22TO

2p

Oscillator Kim,

(5) where

2

τ

TO2τ Ω

Ωωσ1

1expΩωτ

(6)

Model parameters were determined from best fits between computed and experimental data using Scout software [23]. The best fit directly gave the optical constants of the film being studied as well as film thickness. The sheet resistance of the thin films was determined by a two-point square probe with 8 point electrodes, having Au tips in a linear arrangement and touching the coated surface. Since the two-point probe had equal length and width, resistance per unit square was obtained. By measuring the sheet resistance, sRand the average thickness, t , of the films, the electrical resistivity, ρ or conductivity, σ of the samples was calculated using the following relation.

tRσ1ρ s

(7)

Temperature dependent electrical resistivity measurements were done in the temperature range 10K to 300K in a physical property measurement system (PPMS-6000, quantum design) using the standard four-probe method.

3. Results and Discussion

3.1 Composition

0

5

10

15

0 10 20 30 40 50 60 70Nb atomic % in solution

Nb

atom

ic %

in th

e fil

m

Figure 2: Nb at. % in solution (calculated) versus Nb at. % as determined from EDS

Energy dispersive spectroscopy (EDS) was used to measure the percentage of niobium in the films. Figure 2 shows the niobium concentration in the film as a function of chemical solution concentration for films deposited under similar conditions. The composition of the as deposited

Nb:TiO 2 films were determined using Silicon substrates. The experimentally determined composition was proportional to the Nb content in the solution, as shown in Figure 2.This difference is attributed to low efficiency and a high consumption of chemical [26] due to the large separation between the nozzle and substrate to allow vaporization of the solution.

51 52

Figure 6: SEM image of films on silicon substrates as deposited (a) Undoped film (b) 5.7% Nb doped (c) 9.3%

Nb doped 2TiO

The effective grain size was determined from the full-width at half-maximum (FWHM) of the x-ray peak of the (004) plane using the well known Debye-Scherrer formula, i.e.

θcosLλ0.9D

(8)

Where L is the diameter of the crystallites forming the film, λ is the wavelength of the x-ray line, and D is the FWHM of the XRD peak. As displayed in Figure. 5, for samples post annealed in hydrogen atmosphere at 500 C0 for 1 hour, the estimated crystallite size increased with increasing Nb content up to 15 at. %.

From the SEM observations in Figure 6, it can be concluded that the average particle size of undoped titaniun oxide is 70-150nm. For Nb-doped titanium oxide, the particle size is larger in Nb-containing samples and the homogeneity is good as evidenced from the elemental chemical mapping Figure 7.

Figure 7: Elemental chemical mapping for 9.3 at % Nb

doped 2TiO

3.2 Optical Studies

In order to compare the transparency of hydrogen annealed Nb:TiO 2 thin films with various Nb-doping levels, their optical spectra in the UV-VIS–NIR region were measured. The optical transparency of Nb:TiO 2 thin films for various Nb-doping levels is shown in Figure 8.The transparency at 540 nm decreased with increasing Nb doping concentration. Near the infrared wavelength range, the transparency decrease with increasing Nb doping became much more pronounced for the hydrogen annealed films only.

0

20

40

60

80

100

500 1000 1500 2000 2500

01.03 % Nb2.0% Nb5.7% Nb6.5 % Nb

Tra

nsm

ittan

ce /

Ref

lect

ance

Wavelength (nm)

As deposited

609 ±10 nm

Complete map;9.3%Nb

Nb L

Ti L O K

9.3% Nb

0% Nb 5.7% Nb

9.3% Nb9.3% Nb

0% Nb0% Nb 5.7% Nb5.7% Nb

0.001

0.01

0.1

1

10

100

0 5 10 15 20 25

Ele

ctri

cal r

esist

ivity

(

cm

)

Nb atomic % in film

Figure 13: Variation in electrical resistivity with increasing Nb doping concentration in TiO2 for Hydrogen annealed films

Undoped anatase TiO2 exhibited no measurable conductivity. Nb-doping caused a marked decrease in resistivity, with a minimum values being 3.36 ×10-3 Ωcm at 300 K for 15 at. % Nb as shown in Figure . Xiao Wan Zheng [Ref 30] has documented a room temperature resistivity of ~1.3×10-3 Ωcm for 420 nm thick 15 at. % Nb doped TiO2 thin films deposited on (100) LaAlO3 substrates by rf sputtering method.

4.2

4.3

4.4

4.5

4.6

4.7

4.8

4.9

-1.2

-1.15

-1.1

-1.05

-1

-0.95

-0.9

-0.85

-0.8

0 20 40 60 80 100

2% Nb

9.3% Nb

1000 / T

Ln(C

ondu

ctiv

ity) f

or 9.

3% N

b

Ln(C

ondu

ctiv

ity) f

or 2%

Nb

Figure14: ln σ versus 1/T plots for Nb-doped TiO2 films.

The Arrhenius plots of natural logarithm of electrical conductivity, Inσ(T), versus reciprocal temperature, 1/T, for films having Nb/Ti = 2 and 9.3 in the temperature range from 10 to 300 K is depicted in figure 14. The films clearly do not follow Arrhenius behavior.

Three non-distinct regions of σ (T)corresponding to the low, intermediate and high temperature ranges are observed. In the high temperature region (190 – 300 K) for 9.3% Nb and (227 - 300 K) for 2% Nb the abrupt increase in σ means that the conduction is thermally activated and assumed to arise from the contribution of the conduction between the extended states. This behaviour of σ (T) is well described by the simple Arrhenius law,

σ(T)= σ0 exp.∆E/KBT,

(10)

where ∆E is the corresponding activation energy and σ0 the pre-exponential factor. From the slope, the activation energy for 2TiO was calculated to be ~ 1.0 eV which compares favourably with the value of 0.78 eV as already reported for sprayed undoped 2TiO films [31]. The calculated values of both ∆E and σ0 are given in Table 1. It is seen that uncertainties in the values of both parameters are high. The intrinsic electrical conductivity in n-type anatase

2TiO films has been measured. The value for the band gap Eg was calculated from the experimental data, assuming intrinsic conductivity of the form

T/2kEExp.constantσ Bg (11)

We used the intrinsic region in the figure to calculate the band gap of 2TiO samples. The intrinsic region is the straight line extrapolated with a dashed line.

constantT1

2KE

σnB

g

(12)

The slope,B

g

2KE

, directly gave the band gap

of intrinsic 2TiO as 2 eV which was somewhat narrower than the empirical value of 3.2 eV. This under estimation of the band gap is similar to the one reported for GGA calculations.

57 54

0

20

40

60

80

100

500 1000 1500 2000 2500

Tra

nsm

itta

nce

/ Ref

lect

ance

Wavelength (nm)

Hydrogen Annealed 0

Nb at. 1.03%

5.7%

2.00 %

6.50 %

609 ±10 nm

Figure 8: Experimental Spectral Transmittance and Reflectance of undoped and Nb-doped titanium oxide films prepared at 500 15°C before and after post-annealing in Hydrogen atmosphere at 500°C for 1 hour (low Nb concentrations). Thickness of films was 609 10 nm.

0

20

40

60

80

100

500 1000 1500 2000 2500

0.09.3 % Nb9.7 % Nb22 % Nb15 % Nb

Tran

smitt

ance

/ R

efle

ctan

ce

Wavelength (nm)

As deposited

0

20

40

60

80

100

500 1000 1500 2000 2500

Tra

nsm

itta

nce

/ Ref

lect

ance

Wavelength (nm)

336 ± 6 nm

0.0

22% at. Nb

9.3 %

9.7 %15%

Hydrogen Annealed

Figure 9: Experimental Spectral Transmittance and Reflectance of undoped and Nb-doped titanium oxide films prepared at 500 15°C before and after post-annealing in Hydrogen atmosphere at 500°C for 1 hour (higher Nb concentrations). Thickness of films was 336 6 nm

Transmittance and reflectance spectra for the TNO films before and after post annealing for low (Figure 8) and high Nb concentrations (Figure 9) demonstrate the high transparency of TNO. The Thickness of films displayed in Figure 8 is almost twice those shown in Figure 9 for the purpose of amplifying their optical behaviour in the NIR for ease of comparison. Doping transforms the optical properties of the films to a significant degree only after post annealing in hydrogen. Transmittance is approximately 65 - 83% throughout the visible wavelength region. With increasing 5Nb concentration, the transmittance decreased in the wavelength region over 1000 nm. The film with the lowest resistivity had the lowest transmission, especially in the near infrared. According to [12], the above can be expected for plasma oscillations of the conduction band electrons in a partially filled d band.

1,01,52,02,53,03,54,0

500 1000 1500 2000 2500-0,20,00,20,40,60,81,01,2

43

n

12

5

3

54

2

k

Wavelength (nm)

5) 6.5% Nb 3) 5.7% Nb4) 2.0% Nb 2) 1.0% Nb 1) Undoped

1

1,01,52,02,53,03,54,0

500 1000 1500 2000 2500-0,20,00,20,40,60,81,01,2

n

a)15% Nb b) 9.3% Nb c) 22% Nb d) undoped

Wavelength (nm)

dcba

k

a

b

cd

Figure 10: Spectral refractive index and extinction

coefficient for 2TiO and Nb:TiO2 films after

annealing in 1 atm 2H atmosphere at 500˚C for 1 hour.

The optical constants are reported in Figure 10showing spectral refractive index n (λ) and extinction coefficient k (λ) for undoped and doped films. The undoped film and the 22 at. % Nb doped film exhibit a dielectric behaviour with n≈2.1 irrespective of wavelength for λ>500 nm. For the Nb doped films, the optical constants are qualitatively different. In this case k(λ) increases for increasing λ, as expected for a metallic material, while n(λ) drops gently towards larger λ. These effects increase in magnitude with increasing doping level of Nb.

0

2000

4000

6000

8000

1 104

0

1 1015

2 1015

3 1015

4 1015

5 1015

6 1015

7 1015

8 1015

0 1 2 3 4 5Energy, h (eV)

h

eV

m

Dir

ect B

andg

ap

h

1/

eV1/

m

Indi

rect

Ban

dgap

Figure 11: 2αhv and 1/2αhv versus hv plots for determining the optical direct and indirect bandgaps of 9.3% Nb-doped titanium oxide films indicated in the figure

The optical bandgap, gE , was determined using the standard formula [27].

ngEhνααhν ,

(9)

Where α=2πk/λ is the absorption coefficient, hν, the photon energy, and n = 2 accounts for the fact that the indirectly allowed transitions across the band gap are expected to dominate. Figure 11

shows plots of 21

αhν versus photon energy, hν , in the high absorption region. Extrapolation of the curve to hν = 0 gave the indirect band gap of Nb:TiO 2 films in the range 3.38 eV – 3.47 eV for the undoped and heavily doped films; which is comparable with the values already reported [10,14 ,28]. The indirect band gap slightly widens due to the increase in the number of charge carriers with increase in Nb-doping. The direct band gap (n =

1/2) was determined to be 3.75 eV which compares well with 3.8 eV as reported [29].

3.1 Electrical studies

200

250

300

350

400

450

50 100 150 200 250 300

Res

ista

nce,

Temperature (Kelvin)

9.3 at. % NbHeating curve

6.5 104

7 104

7.5 104

8 104

8.5 104

9 104

9.5 104

1 105

50 100 150 200 250 300

Heating Curve 22 at. % Nb

Temperature (Kelvin)

Resis

tanc

e,

Figure12: Resistivity variation as a function of temperature for 2 films

To determine whether the TNO films exhibited metallic behaviour, we measured the resistivity variation with temperature from 10 to 300 K for 9.3 at. % Nb and 2 at.% Nb films and the results are presented in Figure. Surprisingly, the polycrystalline TNO film with room temperature ρ ~10-3 Ω cm did not exhibit metallic ρ – Tbehaviour. Instead, the ρ – T curve exhibited semiconducting behaviour. The semiconducting carrier transport with dρ/dT< 0 is a result of the formation of shallow Nb impurity states [4]. The literature values of resistivity variation with temperature from 10 to 300 K of 8 at. % Nb doped [10] and 15 at % Nb doped TiO2 [30] increases with increasing temperature depicting a metallic behavior.

55 56

-1

0

1

2

3

4

5

6

7

0.35 0.4 0.45 0.5 0.55 0.6

9.3%2%

T -1/4(K)-1/4

In[

T1/

2 ] (

cm

-1K

1/2)

Figure: 15 Plots of ln( σT1/2) versus (1/T)1/4 for Nb-doped TiO2 films.

We plotted Tσn versus 1/41/T for the low temperature (9.7 - 38.9 K ) regime shown in figure 15. The curves are well fitted with straight lines, satisfying Mott’s formula for variable range hopping [32]

1/40

/0 /TTexpT/σTσ

(13)

Where fB

30 EN/K16αT

Rearranging the above equation

constantT1TTTσn 1/4

1/40

(14)

Gradient = 1/40T = 1/4

fB3 EN/K16α

0 is the pre-exponential factor, fEN is the density of localized states at the Fermi level, α describes the spatial extent of the localized wave function and is assumed to be 0.124 Å-1 and KBis the Boltzmann constant. In accordance to Mott’s theory, the room temperature hopping distance, R, and hopping energy, W, is given by the following expressions

1/4fB ENTKαπ9/8R

(15)

f3 ENRπ43/W

(16)

Necessary conditions for Mott’s variable-range hopping process TKW B and 1Rα are satisfied.

Table 1: Calculated values of ∆E and associated parameters.

Nb/Ti

∆E(eV)

0σ1cm)(Ω

210

0T

(K)

/0σ

1/21Kcm)(Ω

fEN

31cmeV

2510

R(cm)

910

W(MeV)

410

2 1.2

1.77

731.16

3007.005

1.9 2.2

1.2

9.3

1.0

0.19

530.84

17.68117

2.6 2.2

1.2

3.2 Conclusion

Conducting and transparent thin films of Nb doped 2TiO on Corning 7059 glass substrates

were successfully synthesized at C15500 oby the low cost, vacuum-free spray pyrolysis technique for the first time. Prior to this, only sputtering and PLD had been used successfully to fabricate highly conducting and transparent TNO films. The Nb-doped 2TiO films were deposited with dopant concentrations ranging from 0 to 22 at. % in the film. The XRD of annealed films showed a polycrystalline anatase (004)-oriented phase without any second phases. EDS studies of chemical composition of the films showed that the exact amounts of Nb in the films are less than that taken in the solution. SEM studies showed that the particle sizes lie in the range 70–300 nm. In the 2TiO thin films,

5Nb acts as a donor so that, with increasing Nb doping up to 15 at. %, the resistivity of the films decreases systematically. However, increase in Nb doping beyond 15 at. %, the resistivity increases sharply. The as deposited films were non-conductive but after annealing in

Hydrogen as an Alternative Fuel: An ab-initio Study of Lithium Hydride and Magnesium Hydride

D. Magero1, N. W. Makau1, G. O. Amollo1 , S. Lutta2, M. D. O. Okoth2, J. M . Mwabora3 , R. J. Musembi3, C. M. Maghanga4, R. Gateru5

1Computational Material Science Group, Department of Physics, University of Eldoret, P.O BOX 1125-30100, Eldoret

2Department of Chemistry and Biochemistry, University of Eldoret, P.O BOX 1125-30100, Eldoret 3Department of Physics, College of Biological and Physical Sciences, University of Nairobi, Box 30197, Nairobi.

4Department of Physics, Kabarak University, Private Bag - 20157 KABARAK 5Kenya Methodist University, P.O. Box 45240 – 00100, Nairobi,

AbstractLimited energy resources and growing pollution associated with conventional energy production have stimulat-ed the search for cleaner, cheaper and more efficient energy technologies. Hydrogen as a fuel is seen as one of the promising energy technologies alternative to fossil fuel. Metal hydrides have been suggested as potential candidates for the bulk storage of hydrogen. In this study, ab-initio calculations of metal hydrides that are prom-ising candidates for hydrogen storage applications, that is, magnesium hydride (MgH2) and lithium hydride (LiH) was carried out using the Quantum Espresso computer code. The calculated quantities were the equilibri-um structural parameters as well as the thermodynamic properties. The calculated lattice parameters for MgH2

were a = 4.54 Å and c = 3.019 Å. Both values of a and c are in good agreement with experimental values of a = 4.501 Å and c = 3.01 Å. The calculated lattice parameters for LiH were a = b = c = 3.93 Å. The lattice parame-ter of LiH shows a correlation of approximately -3.79% with the experimental value of 4.083 Å. Thermodynam-ic properties of LiH and MgH2 were investigated by performing density functional theory within the quasi har-monic approximation. The temperature dependence of the heat capacity at constant volume CV, the Helmholtz free energy ∆F, the internal energy ∆E and the entropy ∆S were obtained. The thermodynamic properties are in good agreement with the experimental data.

Keywords: Hydrogen storage, metal hydrides, ab-initio studies.

1. Introduction There is a rapid increase in the threat from global

warming due to the consumption of fossil fuels. There is need to adopt new strategies to harness inexhaustible sources of energy. CO2 emission from burning fossil fuels has been clearly identified as the main source for global warming. There has been an increase in CO2 emission from fossil fuel combustion from 20.7 billion tonnes in 1990 to 32.5 billion tonnes in 2006 [1].

One of the most promising models for the clean energy-based economy is the so-called Hydrogen economy.Hydrogen is considered to be one of the leading candidates for clean energy sources in the future [2, 3]. Hydrogen is an

environmentally benign, safe and attractive medium; being the most abundant element in the universe, with a high energy density per unit weight (the chemical energy of hydrogen, that is, 142 MJ/kg is at least three times larger than that of other chemical fuels [3] and water as the major by-product of its combustion. A method of utilizing hydrogen is to dissociate water into hydrogen and oxygen, which is an energy emitting process. Therefore, hydrogen is not considered to be the energy source itself, but is described as an energy carrier. One of the concerns about the successful harnessing of hydrogen as energy source is generally associated with its storage and transportation. These problems are more and less related to the lightness and chemical activity of free hydrogen. As regards storage, hydrogen can be stored for use as energy source, in the following ways: gaseous storage, liquid storage, metal

61 58

2H at 1 atm. for 1 hour, doped films turned conductive. Films doped with 15 at.% Nb, 336 nm thick, exhibited a lowest room temperature resistivity of 3.36 -310 Ω cm and visible transmittance of over 65% after post deposition annealing in 2H . The electrical conductivity increases with the (004) peak intensity which is strongly correlated with the anatase crystallinity. The ρ - T curve for Nb:TiO 2 exhibited semiconducting behaviour. The optical band gap for undoped and doped films lay in the range 3.38 – 3.47 eV.

Acknowledgements

The authors would like to thank the International Program in the Physical Sciences (IPPS), Uppsala University for the financial and material support to the Department of Physics, Moi University. Financial support from the Swedish Research Council for this research is greatly appreciated.

References

1. Fujishima A. and Honda K., Nature, 238(1972) 37

2. C.G. Granqvist, Transparent conductors as solar energy materials: A panoramic review, Solar Energy Materials & Solar Cells 91(2007) 1529–1598

3. Siefering K.L. and Griffin G.L., J. Electrochem. Soc. 137 (1990) 1206.

4. T. Hitosugi, N. Yamada, S. Nakao and T. Hesegawa, Properties of 2TiO -based transparent conducting oxides, Phys. Status Solidi A 207, No. 7 (2010) 1529-1537.

5. D. P. Norton, Synthesis and properties of epitaxial electronic oxide thin-film materials, Mater. Sci. and Eng. R 43 (2004) 139–247

6. Kurtz S. R. and Gordon R. G., Chemical vapor deposition of doped TiO2, thin films, Thin Solid Films, 147 (1987) 167-176

7. Mattsson A., Leideborg M., Larsson K., Westin G., and Osterlund L., Adsorption and Solar Light Decomposition of Acetone on Anatase TiO2 and Niobium Doped TiO2 thin Films, J. Phys. Chem. B, 110, No. 3 (2006) 1212-1220

8. A. Tsuzuki, H. Murakami, K. Kani, S. Kawakami, Y. Torii, Preparation of Nb-doped TiO2 films by Sol-gel method, J. Mater. Sci. Lett. 9 (1990) 624 - 626

9. Conde-Gallardo A., Guerreroa M., Castillo N., Soto A.B., Fragoso R., Cabanas-Moreno J.G., TiO2 anatase thin films deposited by spray pyrolysis of an aerosol of titanium diisopropoxide, Thin Solid Films

473 (2005) 68– 7310. Y. Sato, Y. Sanno, C. Tasaki, N. Oka, T.

Kamiyama, Y. Shigesato, Electrical and optical properties of Nb-doped 2TiO films deposited by dc magnetron sputtering using slightly reduced Nb-doped x-2TiO ceramic targets, J. Vac. Sci. Technol. A 28(4), (2010) 851–855.

11. N. Yamada, T. Hitosugi, J. Kasai, N. L. H. Hoang, S. Nakao, Y. Hirose, T. Shimada, T. Hesegawa, Transparent conducting Nb-doped 2TiO (TNO) thin films sputtered from various targets, Thin Solid Films 518(2010) 3101–3104

12. M. A. Gillispie, M. F. A. M. van Hest, M. S. Dabney, J. D. Perkins, and D. S. Ginley, rfmagnetron sputter deposition of transparent conducting Nb-doped 2TiO films on

3SrTiO , J. Appl. Phys. 101, (2007) 033125 13. N. L. H. Hoang, N. Yamada, T. Hitosugi, J.

Kasai, S. Nakao, T. Shimada, and T. Hasegawa, Low-temperature Fabrication of Transparent Conducting Anatase Nb-doped

2TiO Films by Sputtering, Appl. Phys. Express 1 (2008) 115001

14. C.M. Maghanga, J. Jensen, G.A. Niklasson, C.G. Granqvist and M. Mwamburi, Transparent and conducting Nb:TiO 2films made by sputter deposition: Application to spectrally selective solar reflectors, Solar Energy Materials & Solar Cells 94 (2010) 75–79

15. T. Hitosugi, A. Ueda, S. Nakao, N. Yamada, Y. Furubayashi, Y. Hirose, S. Konuma, T. Shimada, T. Hasegawa, Transparent conducting properties of anatase

20.060.94 ONbTi polycrystalline films on glass substrate, Thin Solid Films 516 (2008) 5750–5753.

16. T. Hitosugia, A. Ueda, S. Nakao, N. Yamada, Y. Furubayashi, Y. Hirose, T. Shimada and T. Hasegawa, Fabrication of highly conductive 2xx-1 ONbTipolycrystalline films on glass substrates via

crystallization of amorphous phase grown by pulsed laser deposition, Appl. Phys. Lett 90,(2007) 212106–3.

17. T. Hitosugi, A. Ueda, Y. Furubayashi, Y. Hirose, S. Konuma, T. Shimada, and T. Hasegawa, Fabrication of 2TiO -Based Transparent Conducting Oxide Films on Glass by Pulsed Laser Deposition, Jpn. J. Appl. Phys. Vol. 46, No. 3, (2007), pp. L86–L88.

18. Y. Furubayashi, T. Hitosugi, Y. Yamamoto, K. Inaba, G. Kinoda, Y. Hirose, T. Shimada and T. Hasegawa, A transparent metal: Nb-doped anatase 2TiO , Appl. Phys. Lett. 86,(2005) 252101–3

19. Y. Furubayashi, T. Hitosugi, Y. Yamamoto, Y. Hirose, G. Kinoda, K. Inaba, T. Shimada, T. Hasegawa, Novel transparent conducting oxide: Anatase 2xx-1 ONbTi , Thin Solid Films 496 (2006) 157 – 159.

20. Karlson T., Roos A. and Ribbing C. G., Influence of spray conditions and dopants on highly conducting tin oxide films, Solar Energy Materials 11 (1985) 469 – 478

21. L. Kavan and M. Gratzel, Highly efficient semiconducting tio2, photoelectrodes prepared by aerosol pyrolysis, Electrochimica Acta, 40 No. 5 (1995) 643-652

22. Swanepoel R. Determination of the thickness and optical constants of amorphous siliconJ. Phys. E, Sci. Instrum. 16, 1214 (1983)

23. W. Theiss, in: W. Teiss (Ed.), Scout Thin Film Analysis Software Handbook, Hard- and Software, Aachen, Germany, 2001 <www.mtheiss.com> , pp. 54–57.

24. N. W. Ashcroft, N. D. Mermin, Solid State Physics, , CBS Publishing, Philadelphia, USA, 1976, p. 18.

25. C.C. Kim, J.W. Garland, H. Abad, P.M. Raccah, Modeling the optical dielectric function of semiconductors: extension of the critical-point parabolic-band approximation, Phys. Rev. B 45 (1992) 11749

26. J. C. Manifacier, Thin metallic oxides as transparent conductors, Thin Solid Films, 90(1982) 297 – 308

27. C.G. Granqvist, Handbook of Inorganic Electrochromic Materials, Elsevier, Amsterdam, The Netherlands, 1995, pp. 265–275.

28. R. Ayouchi, C. Casteleiro, R. Schwarz, J. R. Barrado, and F. Martín, Optical properties of TiO2 thin films prepared by chemical spray

pyrolysis from aqueous solutions, Phys. Status Solidi C 7, No. 3–4, (2010) 933– 936.

29. Nickolay Golego, Thin-film polycrystalline titanium-oxygen semiconductors prepared by spray spray pyrolysis, PhD thesis, The University of Guelph, Canada, August 1998.

30. Xiao Wan Zheng, Zhi Qing Li, (2010) Low-Temperature magnetoresistance of Nb –doped TiO2 transparent conducting films, Solid State Communications, Vol. 150 PP 1625 – 1628.

31 Ayouchi R. (2010), C. Casteleiro, R. Schwarz, J.R. Barrado, and F. Martín, Optical properties of TiO2 thin films prepared by chemical spray pyrolysis from aqueous solutions, Phys. Status Solidi C, volume 7, No. 3–4, pp. 933– 936.

32 Mott, N.F., Conduction in non-crystalline materials, III. Localized states in a pseudogap and near extremities of conduction and valence bands Phil. Mag. 19:(1969) 835.

59 60

where E0 (V) is the static contribution to the internal energy at volume (V) and can be easily obtained from standard DFT calculations. From the vibrational free energy, the heat capacity and the vibrational entropy as a function of temperature at zero pressure can be calculated within the framework of the qua-si harmonic approximation. Figures 1-2 show the variation of the calculated phonon contribution to the entropy (S) andheat capacity at constant volume CV as a function of tem-perature from 0 to 1000K for LiH and MgH2 compared to experimental result. The experimental value of CV cannot be found, but the experimental value of heat capacity is given at constant pressure (CP).

Figure 1a and 1b show the calculated phonon contribution to the entropy (S) compared to experimental values

Figure 1a: MgH2

Figure 1b: LiH

Figure 2a and 2b show the calculated phonon contribution to the heat capacity at constant volume (Cv) compared with experimental heat capacity values at constant pressure (Cp).

Figure 2a: LiH

Figure 2b: MgH2

For the calculated error in enthalpy (∆S) of LiH and MgH2,our values are somewhat underestimated compared to the experimental values, but the consistency between the theo-retical and the experimental data may change when the temperature is increased by every 25 K. The slightly large error at higher temperature suggests the limitation of the harmonic approximation, as it accounts only partially for the effects of anharmonicity. To improve the theoretical

secundiflora. Details pertaining to the preparation of each are given in the respective subheadings.

2.1.1. Working Electrode SpecimensSpecimens for working electrode were machined from the

parent material into discs of carbon steel electrodes. The lat-ter were in-turn carefully wet abraded with silicon carbide papers of different grits to 4000 grit surface finish by using LaboPol-5 grinding machine so as to ensure having the same surface finish for all working electrodes used for each elec-trochemical measurement. The prepared electrodes were then ultrasonically cleaned with acetone and rinsed with eth-anol. Finally, the prepared electrodes were stored in a desic-cator. The chemical composition in percentage weight of this carbon steel sample was: 0.172 C, 0.019 Si, 0.433 Mn, 0.210 Cr, 0.014 S, 0.070, 0.038 Co, 0.115 Cu, 0.002 Pb, 0.022 Sn and 98.905 Fe.

2.1.2. Test SolutionThe formulation water sample was prepared in the labora-

tory using distilled water into which chloride and sulphate ions were added according to the gathered technical specifi-cations [13]. All chemicals used throughout the work were prepared from fine analytical reagent grades and distilled wa-ter. The pH of the test solution was maintained at 7.0.

2.1.3. Preparation of Corrosion InhibitorThe corrosion inhibitor employed in this study, was an ex-

tract of Aloe secundiflora, which was prepared through a number of isolation processes.

2.2. Analysis of the Electrochemical Measurements

2.2.1. Experimental Set up for Corrosion Inhibition Study

The investigation was carried out in a three electrodes Py-rex glass cell assembly, schematically shown in “Figure 1”.A Rotating Disk Electrode (RDE) assembly with adjustable speed was used as working electrode.

Figure 1 A picture of the experimental setup that was employed in the corrosion inhibition study through electrochemical tech-niques.

The working electrode of interest was a carbon steel whose specifications are given in the materials section. The surface area of carbon steel electrode that was exposed to the test so-lution was 0.385 cm2. A platinum rod was used as a counter electrode and Ag/AgCl electrode was employed as a refer-ence, both of them being mounted in the Pyrex glass cell lid. The cell consisted also the thermostatic jacket whose temper-ature was constantly controlled by special temperature con-trol unit.

2.2.2. Electrochemical Impedance Meas-urements

Electrochemical impedance measurements were per-formed using a computer assisted Autolab PGSTAT20 Fre-quency Response Analyser (FRA) in a frequency range from 10 kHz to 10 mHz at a sweeping rate of 10 points per decade, logarithmic division. All measurements were performed at the open circuit potential and were taken after nine hours of exposure before performing the potentiodynamic measure-ment.

The electrochemical impedance observed data were ana-lysed by fitting them to Equivalent Circuits (EC). The EC parameters for the charge transfer resistance, Rct, solution re-sistance, Rs, and electrochemical double layer capacitance, Cdl were all obtained by using a computer assisted Autolab PGSTAT20.

2.2.3. Potentiodynamic MeasurementsPotentiodynamic measurements were performed using a

computer controlled PGSTAT20. Measurements were per-formed on the specimens by sweeping the potential at the rate of 1 mV/s in a range of 100 mV vs Ag/AgCl in both cathodic and anodic directions from the open circuit potential (Eocp).The cathodic and anodic polarization curves were taken sep-arately, starting with the cathodic sweep. All potentiody-namic measurements were taken after the electrochemical impedance measurements.

For analysis purpose, anodic and cathodic Tafel lines were performed on the respective Tafel regions of polarization curves so as to obtain the Tafel slopes.

The tests were performed at different desired experimental conditions using the test solution with and without corrosion inhibitor. The effects of inhibitor concentrations were inves-tigated.

3. Results and Discussion Inhibitor concentration is one among the key parameters

which ought to be tested in order to establish the performance of any inhibitor in question. This section, therefore, attempts to present the findings of laboratory investigations which were done aiming at finding the minimum inhibitor concen-tration required to attain optimum corrosion protection for carbon steel employed in fresh water distribution networks.

3.1. Results from Polarization Measure-ments

It is well established that polarization curves can help to understand how a certain inhibitor works. Inhibitors can

65 62

results, denser meshes are needed, which in turn means huge computational task, or better pseudo potential ap-proach should be used. We note that the deviation between calculated and experimental data in entropy confirms that our theoretical considerations are reasonable. Figure 2a and 2b shows the calculated phonon contribution to the heat capacity at constant volume CV for LiH and MgH2 com-pared with experimental heat capacity at constant pressure Cp. The relation between CP and CV can be given by the following equation:

……….(1)

where α is the volume thermal expansion coefficient, B0 is the bulk modulus, V is the volume and T is the absolute temperature. According to this equation, the difference be-tween CP and Cv increases with increasing T. The difference between the CP and CV values is however not significant at the low-temperature limit. At high temperatures, however, CV becomes smaller than CP as seen from the plot due to the contribution of high temperature as seen in equation (1). Our findings suggest that the calculated results can indeed be quite useful as a prediction tool for future investigations. 4.0 ConclusionA presentation of ab-initio calculations of the equilibrium structural parameters of LiH and MgH2 has been done. The relaxed lattice parameters agreed to within 2% of the experimental ones. From the calculated phonon contribution to the ∆S and CV, the results were in excellent agreement with the experimental findings. The accuracy achieved from the calculated values showed that experimental work can indeed be guided by the computational theory, thus tremen-dously increasing throughput and also cutting costs. AcknowledgementThis work was supported by the National Commission for Science, Technology and Innovation (NACOSTI), Kenya, and the authors are most grateful.

REFERENCES [1] B. Metz, Intergovernmental Panel on Climate Change:

Special Report on CO2 Capture and Storage, 2005.

[2] B. Sakintuna, F. Lamari-Darkrim, M. Hirscher. Int. J. Hydrogen Energy, 32 (2007), pp. 1121–1140..

[3] R. A. Varin, T. Czujko, Z. S Wronski, J. Alloys Compd, 393 (2005), pp.1–4.

[4] Zhigang W, Mark D.A and Jeffrey C. Grossman. J. AM. CHEM. SOC. 131 (2009), pp. 13918–13919.

[5] Novaković N, Radisavljević I, Colognesi D, Ostojić S, Ivanović N. J Phys Condensed Matter, 19 (2007), pp.406211.

[6] David E. J Mater Process Technology 162-163 (2005), pp.169-77.

[7] Giannozzi, P., Baroni, S., Bonini, N., Calandra, M., Car, R., Cavazzoni, C., ... & Wentzcovitch, R. M. Journal of Physics: Condensed Matter, 21(2009), p.395502. <http://www.quantum-espresso.org>.

[8] Noritake. Journal of Alloys and Compounds, 356 (2003), pp. 84-86.

[9] Principi G, Agresti F, Maddalena A, Russo SL. Energy, 34(2009), pp. 2087-2091.

[10] Weast RC, Astle MJ, Beyer WH. CRC handbook of chemistry and physics. 64th ed., Boca Raton, FL: CRC Press; 1983.

[11] Bouhadda Y, Rabeli A, Bezzari -Tahar-Chaouche S. Revue

Des Energies Renouvelables, 10(4) (2007),pp. 545-550. [12] Andreasen A. Predicting formation enthalpies of metal

hydrides. Ris national laboratory report, Ris-R-1484 (EN); 2004.

[13] H. Nakamura, D. Nguyen-Manh, G. Pettifor, J. Alloys Compd. 281 (1998), p. 81.

Aloe Secundiflora Extract as a Green Corrosion Inhibitor for Carbon Steel in Potable Water Systems

1J. Mutasingwa, 2J. Buchweishaija, 2J. E.G. Mdoe

1Physical Science Department, Open University of Tanzania, P.O.Box 23409, Dar es Salaam, Tanzania.2Chemistry Department, University of Dar es Salaam, P.O.Box 35061, Dar es Salaam, Tanzania.

Abstract A botanical species of Aloe secundiflora has been tested for the first time to ascertain its corrosion inhibition properties for

carbon steel in neutral and aerated soft water solutions. Kinetic results obtained from analysis of impedance plots and polar-ization curves for stationary carbon steel electrodes in uninhibited and inhibited aerated fresh water solution, have revealed profound ability of the botanical extract in addressing the corrosion challenge facing potable water networks. The investiga-tion was performed at different inhibitor concentrations under static and dynamic conditions using a Rotating Disk Electrode (RDE). Inhibitor efficiency of 98% was registered when optimum concentration of 200 ppm of Aloe secundiflora extract was employed. The relationship between the inhibition efficiency and the bulk concentration of the inhibitor at constant temper-ature has been shown to be well explained by Temkin’s adsorption isotherm.

Keywords: Corrosion, corrosion inhibitor, carbon steel, Aloe secundiflora

1. Introduction Carbon steel is one of the most important alloys which are

frequently used in wide industrial applications. In particular, it is used in petroleum pipe lines, pumping stations for do-mestic water supply as well as agricultural water irrigation schemes. Nevertheless, corrosion attack remains a challenge to such profound usefulness of carbon steel [1-5]. Problems caused by corrosion attack can be grouped into three catego-ries: health, aesthetic and economics [6, 4]. Internal corro-sion in fresh water distribution systems causes considerable consequences for municipalities [7, 2]. The consequences of internal corrosion are pipe breaks, overflow, clogging of pipes with corrosion products and the most serious problem is water quality deterioration which poses health problems to users [8, 3]. Carbon steel pipes have been widely used in fresh water distribution system for decades and for metallur-gical strength reasons it is difficult to substitute them with plastic pipes especially when employed from the main pump-ing stations. Thus, their usage in fresh water industry will still last for some years, and hence research about the corrosion protection of carbon steel is still of great importance [1, 5].

Many researches have been devoted to investigate hun-dreds of organic and inorganic compounds as inhibitors for carbon steel corrosion in fresh water systems [9, 2]. Environ-mental restrictions imposed on heavy-metal based corrosion inhibitors, have oriented scientific researches towards study-ing non-toxic and environmentally friendly corrosion inhibi-tors [10, 4]. It should be mentioned here that the first used corrosion inhibitors were naturally occurring substances ex-tracted from various parts of different plants [6, 3]. Besides

their safe handling, plant extracts are usually cheap and could be obtained by simple extraction process.

A botanical species of Aloe secundiflora has been previ-ously studied to ascertain its bioactivity on newcastle disease and fowl typhoid in local chickens in Tanzania [11]. Re-search findings do show that the plant possesses profound medicinal elements in the form of anthraquinones, and have been proven to be effective in treating the mentioned diseases [11]. Like other Aloe species, Aloe secundiflora is exten-sively grown in semi arid areas, Tanzania mainland being one of the areas where it is grown. In particular, it is culti-vated in Same district, in northern Tanzania. The chemical components of Aloe secundiflora extract reveal the existence of active groups which suggest a plant to be a potential source of green corrosion inhibitors. The chemical components pre-sent include the benzyl pyrone, barbaloin, isobarbaloin, alo-inside and isoaloesin, most of them resembling those of some well-known organic corrosion inhibitors [12, 3]. Hence, the present study has been done to establish the potential of these active groups in addressing the corrosion challenge facing potable water networks.

2. Experimental Key laboratory activities were undertaken in the course of

execution of this study as narrated in the following subhead-ings.

2.1. Materials A number of materials were employed in this study includ-

ing, working electrode specimens, test solution for corrosion study as well as the corrosion inhibitor extracted from Aloe

63 64

modify the anodic process, the cathodic process or both lead-ing to a decreased rate of the corrosion process. The inhibi-tion effect of Aloe secundiflora extract on carbon steel in a mentioned test solution was investigated by polarization technique in both stationary and dynamic systems. In all cases the polarization curves were recorded 9 hours after the injection of the Aloe secundiflora extract in the test solution at 30 oC. “Figure 2” shows polarization curves where the in-hibitor was applied in the solution under stationary condi-tions (0 rpm) while varying the inhibitor concentration. Onthe other hand, “Figure 3” depicts polarization curves where the inhibitor concentration was maintained at its optimum value of 200 ppm while varying the rotation speed of the RDE.

Figure 2 Selected polarization curves showing the performance of Aloe secundiflora extract at different inhibitor concentrations.

Figure 3 Selected polarization curves showing the performance of Aloe Secundiflora extract at different rotation speeds.

In “Figure 2”, two important trends are evident. Firstly that, Aloe secundiflora extract was found to block the electro-chemical processes taking place on the steel undergoing cor-rosion in aerated fresh water medium. It reduces both, the rate of cathodic and anodic reactions by reducing the current densities on both sides of the polarization curves in the stud-ied potential region. Secondly, it shifts the open corrosion potentials towards less negative values with reference to the blank, especially for inhibitor concentrations that are close to

optimum inhibitor concentration. These factors suggest that the Aloe secundiflora extract inhibitor acts as a mixed type corrosion inhibitor. As for “Figure 3”, the trend shows that as the rotation speed increases beyond 2000 rpm, the current densities seem to slightly increase, and the open circuit po-tentials shift towards more positive value. This suggests that the performance of Aloe secundiflora extract as corrosion in-hibitor tends to gradually diminish as the rotation speed in-creases.

The data obtained from polarization curves by Tafel ex-trapolations are presented in Table 1 and 2.

Table 1 Electrochemical parameters obtained from polarization measure-ments on stationary carbon steel electrodes in the presence of dif-ferent Aloe secundiflora extract concentrations.

Cinh ba bc Ecorr icorr IE

ppm (mV/dec) (mV/dec) (mV vs Ag/AgCl) (µA cm-2) (%)

0 27 -28 -668 0.99 -40 22 -41 -696 0.34 66%80 19 -30 -685 0.20 80%

160 17 -25 -682 0.08 92%200 12 -15 -655 0.03 97%250 15 -15 -654 0.03 97%

Table 2 Electrochemical parameters obtained from polarization measure-ments on rotating mild steel electrodes with different rotation speeds in the presence of 200 ppm of Aloe secundiflora extract.

Rotation ba bc Ecorr icorr IESpeed(rpm) (mV/dec) (mV/dec) (mV vs

Ag/AgCl)(µA

cm-2) (%)

0 12 -15 -655 0.03 97%1,000 20 -23 -654 0.07 93%2,000 19 -22 -648 0.08 92%3,000 26 -44 -662 0.09 91%4,000 24 -36 -671 0.11 89%5,000 21 -42 -683 0.12 88%

The two tables present the values of corrosion potential (Ecorr.), corrosion current density (icorr.), and Inhibitor Effi-ciency (IE) for the stationary (0 rpm) and varied Aloe secun-diflora extract concentrations, as well as dynamic system op-erating at a single optimum inhibitor concentration. It can be seen from Table 1 that the corrosion rates decrease signifi-cantly at all inhibitor concentrations under stationary condi-tion. Table 1 also gives IE for the Aloe secundiflora extract,being calculated from equation 1, that is,

wcorr,

icorr.,i

i1IE (1)

where icorr.,i and icorr.,w are the corrosion rates with and with-out inhibitor, respectively. It can be seen that 97 % efficiency is obtained when 200 ppm of Aloe secundiflora extract is ap-plied. Analysed data in Table 2, however, do show the grad-ual increase in corrosion rate as the rotation speed increases, especially beyond 3000 rpm, resulting into corresponding de-crease in IE of the Aloe secundiflora extract.

Polarization measurements results have indicated that Aloe secundiflora extract is a mixed type inhib-itor for carbon steel corrosion in fresh water system, with a significant reduction of current on the anodic reaction.

Of all the tested adsorption isotherms, Temkin ad-sorption isotherm has been confirmed to offer best explanation on the relationship between the inhibi-tion efficiency and the bulk concentration of the in-hibitor at 30 C. The calculated value of equilibrium constant, K, for the adsorption process, as well as the corresponding calculated value of standard free energy of adsorption, , were both found to fall in the range of 1.288 to 1.276 ppm-1 and -614.33 to -637.93 Jmol-1 respectively.

ACKNOWLEDGEMENTThe authors would like to acknowledge the financial sup-

port provided by the Open University of Tanzania, and the University of Dar es Salaam. Many thanks should also go to the technical staff, Chemistry Department, UDSM for the technical assistance which made this work possible.

REFERENCES [1] Teng, F., Guan, Y.T. and Zhu, W.P., 2008, Effect of biofilm

on cast iron pipe corrosion in fresh water distribution system: Corrosion scales characterization and microbial community structure investigation, Journal of Corrosion Science, 50, 2816–2823.

[2] Devarayan K., Mayakrishnan G. and Nagarajan S., 2012, Green inhibitors for corrosion of metals: A Review, Chemical Science Review and Letters, 1(1), 1-8.

[3] Fouda, A.S., Rashwan, S.M. and Abo-Mosallam, H.A., 2013, Fennel seed extract as green corrosion inhibitor for 304 stain-less steel in hydrochloric acid solutions, Journal of Desalina-tion and Water Treatment, 1-12.

[4] Denni, A. A., Norinsan, K. O., Azman, J., Abdul Razak, D.,Abdul Rahman, I. and Al-hardan, N. H., 2013, Nanosilicate extraction from rice husk ash as green corrosion inhibitor, In-ternational Journal of Electrochemical Science, 8, 1759 –1769.

[5] Pandian B. R., Mehran F., Ahmad K. Q., Afidah A. R., Hasnah O., Marc, L. and Khalijah A., 2013, Evaluation of green corrosion inhibition by alkaloid extracts of Ochrosia oppositifolia and isoreserpiline against mild steel in 1 M HCl medium, Journal of Industrial & Engineering Chemistry Re-search, 52, 10582−10593.

[6] El-Etre, A.Y., 2008, Inhibition of carbon steel corrosion in acidic solution using the aqueous extract of zallouh root, Jour-nal of Materials Chemistry and Physics, 108, 278–282.

[7] Sander, A., Berghult, B., Elfström Broo, A., Johansson, E.L and Hedberg, T., 1996, Iron corrosion in fresh water distribu-tion systems – The effect of pH, calcium and hydrogen car-bonate, Journal of Corrosion Science, 38, 443–455.

[8] Choi, Y., Shim, J. and Kim, J., 2004, Corrosion behavior of low alloy steel containing Cr, Co and W in synthetic fresh wa-ter, Journal of Materials Science and Engineering, 385, 148–156.

[9] Li, Y.-J., Wu, B., Zeng, X.-P., Liu, Y.-F., Ni, Y.-M., Zhou, G.-D. and Ge, H.-H., 2002, The voltammetry-photocurrent re-sponse study of passivation of carbon steel in slightly alkaline solutions containing the corrosion inhibitor phosphor-poly-maleic acid-ZnSO4, Journal of Thin Solid Film, 405, 153–161.

[10] Choi, D., You, S. and Kim, J., 2002, Development of an envi-ronmentally safe corrosion scale, and microorganism inhibi-tor for open recirculating cooling systems, Journal of Materi-als Science and Engineering, 335, 228–235.

[11] Waihenya, R., “Investigations on the bioactivities of Aloe sec-undiflora (Aloeaceae) on newcastle diseases and fowl typhoid in local chickens (Gallus Domesticus) in Tanzania,” PhD the-sis, University of Dar es Salaam, Tanzania, 2002.

[12] Buchweishaija, J., “Inhibiting Properties and Adsorption of an Amine Based Fatty Acid Corrosion Inhibitor on Carbon Steel in Aqueous Carbon dioxide Solutions,” PhD Thesis, Norwe-gian University of Science and Technology, Norway, 1997.

[13] Salasi, M., Shahrabi, T., Roayaei, E. and Aliofkhazraei, M., 2007, The electrochemical behaviour of environment-friendly inhibitors of silicate and phosphonate in corrosion control of carbon steel in soft water media, Journal of Materials Chem-istry and Physics, 104, 183–190.

[14] Philip, J.N.Y., Buchweishaija, J., and Mkayula, L.L., 2001, Cashew Nut Shell Liquid as an alternative corrosion inhibitor for carbon steels, Tanzania Journal of Science, 27, 9-19.

[15] Ali, Sk.A., Saeed, M.T. and Rahman, S.U.,2003, Journal of Corrosion Science, 45, 253.

69 66

3.2. Results from Impedance Measure-ments

Results of impedance measurements in the Nyquist format for carbon steel in fresh water test solution containing differ-ent concentrations of Aloe secundiflora extract at stationary and dynamic conditions are shown in “Figure 4” and “Figure 5” respectively.

Figure 4 Impedance spectra in Nyquist format showing the perfor-mance of Aloe secundiflora at different inhibitor concentra-tions.

Figure 5 Impedance spectra in Nyquist format showing the perfor-mance of 200 ppm Aloe secundiflora at different rotation speeds.

Analysis was done in the observed spectra by fitting them with equivalent circuit for coated metal/solution interface us-ing a computer assisted Autolab PGSTAT20, to estimate equivalent circuit parameters [14]. The latter includes charge transfer resistance (Rct), and double-layer capacitance (Cdl) values. Table 3 and 4 give the values of charge transfer re-sistance (Rct), double layer capacitance (Cdl) and inhibitor ef-ficiency (IE) obtained from “Figure 4” and “Figure 5”. Since, corrosion rate is inversely proportional to charge transfer Re-sistance, IE was calculated by using equation 1.

Table 3 Electrochemical parameters obtained from impedance measure-ments on stationary mild steel electrodes in the presence of dif-ferent Aloe Secundiflora concentrations.

Cinh Eocp Rct Cdl icorr IE

ppm (mV vs Ag/AgCl) (/cm2) (µF cm2) (µA cm-2) (%)

0 -664 2,599 654 2.401 -40 -690 8,675 101 0.749 69%80 -683 13,556 80 0.390 84%160 -682 27,349 11 0.168 93%200 -656 56,030 8 0.054 98%250 -645 56,315 7 0.050 98%

Table 4 Electrochemical parameters obtained from impedance measure-ments on rotating mild steel electrodes with different rotation speeds in the presence of 200 ppm of Aloe Secundiflora.

Rotation Eocp Rct Cdl icorr IE

Speed(rpm)

(mV vs Ag/AgCl) (/cm2) (µF cm2) (µA

cm-2) (%)

0 -656 56,030 8 0.054 98%1,000 -652 35,750 18 0.136 94%2,000 -647 30,180 21 0.153 94%3,000 -646 25,235 22 0.265 89%4,000 -666 24,463 24 0.267 89%5,000 -682 24,375 25 0.292 88%

It is seen from Table 3 that as Aloe secundiflora extract concentration increases, the Rct values increase while the double-layer capacitance decreases. This indicates that the extract is corrosion inhibitive in nature. The decrease in the Cdl values in the presence of Aloe secundiflora extract show that it adsorbs on the metal surface which results in decrease in double layer capacitance. The inhibition efficiencies ob-tained by impedance studies are in agreement with those ob-tained when employing potentiodynamic polarization tech-nique under stationary condition and varied inhibitor concen-tration. On the other hand, for the analysed kinetic data in Table 4, the trend shows that as the rotation speed is in-creased from 1000 rpm to as high as 5000 rpm, the Rct values decrease while the double-layer capacitance increases. This also suggests that the inhibitor film formed is less stable when subjected to high rotation speeds.

3.3. Adsorption Isotherms The relationship between the inhibition efficiency and the

bulk concentration of the inhibitor at constant temperature, which is known as isotherm, gives an insight into the nature of adsorption process. The latter shows the adsorption behav-iour of organic adsorbate on the metal surface. For organic inhibitors that have the ability to adsorb strongly on metal surface, thus impeding the dissolution reaction, the metal sur-face coverage () can be evaluated as the inhibition effi-ciency [15].

The results obtained from both polarization and electro-chemical impedance spectroscopy (EIS) measurements were analysed using equation 1 so as to evaluate the degree of sur-face coverage for different concentrations of the Aloe secun-diflora extract inhibitor. Table 5 presents the analysed kinetic data from both impedance and polarization results.

Table 5 Calculated values from both polarization and impedance data for the application of adsorption isotherms for Aloe secundiflora ex-tract inhibitor in aerated and neutral fresh water solution at 30 °C.

Cinh Log Ci Polarization results’ Impedance results’

ppm (ppm) Surface Coverage()

Surface Coverage()

40 1.6021 0.66 0.6980 1.9031 0.80 0.84

160 2.2041 0.92 0.93200 2.3010 0.97 0.98250 2.3979 0.97 0.98

The data in Table 5 were tested graphically by fitting them to various adsorption isotherms. Of all the tested adsorption isotherms, only Temkin adsorption isotherm has been shown

to give the best linear relationship between surface coverage () and bulk concentration of the inhibitor at constant tem-perature through relevant Temkin’s equation 2, that is,

= log K + log Ci (2)where, K is the equilibrium constant of the adsorption pro-

cess and Ci is the inhibitor concentration in the bulk solution. This implies that the adsorption of the Aloe secundiflora ex-tract inhibitor on carbon steel surface follows Temkin ad-sorption isotherm. Temkin’s adsorption isotherm is based on the assumption that solid surface is composed of small areas of equal size, at each of which Langmuir isotherm holds in-dependently. Corroding metal surface that has distributed an-odic and cathodic areas is akin to the solid surface assumed in Temkin’s isotherm [15].

The variation of with logCi at 30 °C is shown in “Figure 6” and “Figure 7” for polarization results and electrochemi-cal impedance results, respectively. Fairly straight line plots which are observed in the two figures are the result of the polarization and EIS data presented in Table 5, which are fit-ted to a straight line, giving a correlation coefficient of about 0.98.

Figure 6 Temkin’s adsorption isotherm for Aloe secundiflora extract inhibitor in aerated and neutral fresh water solution at 30 °C, plotted from polarization data.

Figure 7 Temkin’s adsorption isotherm for Aloe secundiflora extract inhibitor in aerated and neutral fresh water solution at 30 °C, plotted from impedance data.

Again, by making use of an expression for Temkin adsorp-tion isotherm (ie. Equations 2), values of equilibrium con-stant for the adsorption process have been calculated from the results of the linear regression, that is from the intercepts: logK at logCi=0, in “Figures 6 and 7”. But from thermody-namics point of view, the equilibrium constant is also related to standard free energy of adsorption by the equation 3,

(3)

which can be rewritten as equation4, (4) where, K is the equilibrium constant of adsorption process,

R is the molar gas constant (8.3145 J mol-1 K-1) and T is the absolute temperature of the test solution (which was 303.15 K).

Hence, by making use of equation 4, it was possible to ex-press the equilibrium constant values in terms of their corre-sponding values of standard free energy of adsorption. Table 6 presents both, the calculated values of equilibrium constant as well as their corresponding values of standard free energy of adsorption.

Table 6 Calculated values of equilibrium constants from the curve fitting of data for Aloe secundiflora extracts to the Temkin’s adsorption isotherm, as well as their corresponding values of standard free energy of adsorption.

Electrochemical Intercept KTechnique (log K) (ppm-1) (J mol-1)

Polarization results 0.110 1.288 -614.33

Impedance results 0.106 1.276 -637.93

It is well known from principles of thermodynamics that, the more negative the value of standard free energy of ad-sorption, the more spontaneous the adsorption process. Val-ues of standard free energy of adsorption in Table 6 are seen to be negative for both polarization and impedance results. Such an observation implies spontaneous inhibitor adsorp-tion process on metal surface by the Aloe secundiflora extract.

4. Conclusions Outlined below are the main findings obtained in this

study: Aloe secundiflora extract has been shown through

this study to be an excellent corrosion inhibitor for carbon steel in aerated soft water medium. The ex-tract performed well in both stationary and dynamic systems, though a better performance was registered under stationary condition. This suggests that the tested inhibitor might be a better choice for storage tanks than pipes.

In addition, the best quality films from visual inspection cor-respond to deposition time of 90 seconds.

50 100 150 200

20

40

% T

rans

mitta

nce

Deposition time (s)

% Transmittance

Figure 3 Variation of transmittance (at 1508nm) with deposition time for films electrophoretically deposited at35V from 0.01g/40mL suspension

3.3 Optimization of concentration

Figure 4 and figure 5 show the results for optimization of concentration at voltage of 35V ad deposition time of 90s. Figure 4(a) show the photographs from digital camera of the TiO2/Nb2O5 composite electrode thin films electrophoretical-ly deposited from powder concentration from 0.05g/40mL to 0.3g/40mL. The corresponding transmission spectra for var-ied concentrations are shown in figure 4 (b). The concentra-tion of 0.01g of metal oxide powders in 40mL of propan-2-ol (equivalent to 0.25g/L) produced electrophoretically deposit-ed films with a 55% transmittance, which was the highest obtained in the study. Visual inspection showed that the films corresponding to 0.01g/40ml concentration had the best quality films.

Figure 5 shows that for fixed applied voltage and deposition time, transmittance decreased with increased concentration. This decrease in transmittance resulted from development of thick films which hinder transmittance of light rays. Further, films deposited at high concentration would be non-porous and unsuitable for dye absorption for solar cell application.

The uncoated glass gave the highest transmittance value. It was difficult therefore to determine the minimum concentra-tion from transmittance values as shown in figure 5. Powder concentration of 0.01 g/40mL was therefore chosen as opti-mum EPD concentration value for fabrication of TiO2/Nb2O5composite electrode thin films.

.

0 1000 2000 3000

0

20

40

60

% (T

rans

mitt

ance

)

Wavelength (nm)

c35V,90s, 0.10g/40mL c35V,90s,0.015g/40mL c35V,90s,0.02g/40mL c.35,90,0.025 g/40mL c35V,90s,0.30g/40mL

Figure 4. (a) Photographs of and (b) Transmittance spectra for TiO2/Nb2O5composite electrode thin films electrophoretically deposited at 35 DCV for 90 s and from varying powder concentration.

0.01 0.02 0.03 0.04

20

30

40

50

% Transmittance

% T

rans

mitt

ance

concentration g/40mL

Figure 5 Variation of transmittance (at 1508nm) with concentration for films electrophoretically deposited at 35V from 0.01g/40mL suspension

Deposition time of 90s became the optimum EPD deposition time for fabrication of TiO2/Nb2O5 composite electrode thin films.

a

b

67 68

Morphological and Structural Characterization of TiO2/Nb2O5 Composite Electrode Thin Films Synthesized by Electrophoretic Deposition (EPD) Technique

J. N. Nguu, R. J Musembi, F. W. Nyongesa, & B.O. Aduda

Department of Physics, University of Nairobi, Nairobi, P. O. Box 30197-00100, Kenya

E-mail: [email protected] or [email protected]

Abstract Composite electrodes of titanium dioxide (TiO2) and niobium (V) oxide (Nb2O5) have been deposited onto glass substrates by electrophoretic deposition (EPD) technique. The mechanism of EPD involves charged particles in a liquid sus-pension being forced to move towards and deposit on an oppositely charged electrode upon application of electric field. In this study, TiO2 and Nb2O5 nano-sized powders were suspended in a Pyrex glass containing propan-2-ol. Magnesium nitrate hexa-hydrate (Mg (NO3).6H2O) pellets were added to the suspension to induce surface charges on the metal oxides. The structure of the thin films was characterized by X-ray Diffraction (XRD) while morphology was characterized by Scanning Electron Mi-croscope (SEM). The XRD spectra indicated that the films are of polycrystalline nature and that TiO2 and Nb2O5 particles were present in the deposited composite film. Transmittance of the deposited films was measured using the UV-Vis-NIR spectro-photometer spectra in the range 200 to 3200nm. The visual inspection of the films and the morphological investigations from SEM showed that porous films of high quality with well controlled morphology can be deposited using the EPD technique. Further work is needed to evaluate the potential of TiO2/Nb2O5 composite electrode thin films deposited by EPD for dye-sensitized solar cells applications.

Keywords Electrophoretic deposition (EPD), Transmittance, TiO2/Nb2O5 composite electrode thin films, Solar energy materials, Photovoltaic application,

1. Introduction

The development and application of renewable energy tech-nologies is the focus of increased scientific research to help mitigate against challenges of climatic change and the possi-ble depletion of the fossil fuels. Solar energy represents animportant renewable energy resource which is carbon free, and ubiquitous, with terrestrial potential about 600TW. It is also useful for off-grid utilizations. The commonly used pho-tovoltaic systems are crystalline silicon-based with positive attribute of high power conversion efficiencies (24.7%) [1]. However, these modules require vacuum-based techniques for fabrication resulting in high production costs. A lot of re-search is focused on developing solar cells that require low production costs and have adequate conversion efficiencies. Consequently, among the excitonic solar cells, the dye-sensitized solar cells (DSSC) have attracted widespread in-terest owing to their easy f1abrication processes, low produc-tion costs, less sensitivity to impurities, wide range of tem-perature for optimal operation, compatibility with glass and flexible substrates and fair solar–to–electrical conversion ef-ficiencies [2].

Whereas DSSCs have achieved conversion efficiencies of 11.6% [3], there are obstacles to overcome in order to in-crease efficiency and make the cells reproducible on large scale. Dye sensitized solar cells contain a porous, semicon-ductor material (photoelectrode) which absorb dye molecules and conduct injected photoelectrons. The dye molecules play the role of capturing light photons by extending the optical adsorption spectrum of the wide energy gap (Eg > 3eV) sem-iconductor. Titaniun dioxide (TiO2) is the most widely used semiconductor for preparing DSSC photoelectrodes because it provides a high energy conversion efficiency[2]. The high efficiency is the result of titanium dioxide films providing ease of electron transmission through them thus enhancing electron conduction to the conducting glass before recombi-nation. However, to further reduce recombination and in-crease efficiency, research efforts are geared towards modi-fication of the photoelectrode through use of other large band gap semiconductors and through use of compo-sites[4,5]. According to Lin et al., (2009)[6], the importance of thin film technology in fabricating solar cells is that thin films shortens the diffusion length of photo generated carri-ers and decreases the recombination. Use of thin films saves on material and simplifies the synthesis of solar cells. Modi-

fication of photoelectrode can also be achieved by use of composites [7,8].

Composites are regarded as engineered materials made from two or more constituent materials with significant different physical or chemical properties which remain separate and distinct at the microscopic scale within the finished structure. A number of authors [9,10,11] have proposed that niobium pentoxide (Nb2O5), a wide band gap (Eg=3.49eV), and n-type semiconductor, can be combined with titanium dioxide (TiO2) to fabricate composite thin films.

Thin films of titanium dioxide have been fabricated by varie-ty of methods like doctor blade [12], power screen printing [13], DC magnetron [10], and chemical vapor deposition (CVD) [14]. Some of these methods require high power con-sumption and vacuum chambers. In addition, electrophoretic deposition (EPD) technique has not been fully exploited in photoelectrode preparation for dye-sensitized solar cells de-spite being a relatively cheap technique and which could be scaled up for mass production if successfully developed. EPD is a colloidal technique in which charged particles are forced to move towards and deposit on to an oppositely charged electrode by the application of a DC electric field [15,16,23]. The advantages of EPD technique include good control of film thickness, short deposition time, high deposi-tion speed, simple equipment for its construction, wide ap-plicability to diverse materials and low manufacturing tem-perature. Further, EPD does not require highly purified mate-rials and vacuum chambers for film deposition [12, 14, 18]

Literature survey reveal that authors have used diverse val-ues for EPD process parameters and therefore there is need to optimize the same parameters for composite deposition [17,19,23,24,25]. As a result, EPD process parameters have been optimized and the technique used to deposit good quali-ty-nanocrystalline and nanoporous TiO2/Nb2O5 composite electrode thin films for application in the dye-sensitized solar cells.

2. Materials and methods 2.1 Materials

Commercial TiO2 nanopowder (Cas No. 13463-67-7 Al-drich), Nb2O5 nanopowder (Cas No. 1313-96-8 Acros Or-ganics BVBA, Belgium), 5x10-5M- Magnesium nitrate hexa-hydrate (Mg(NO3)2.6H2O 99.9%, Aldrich), and propan-2-ol (Isopropyl alcohol: IPA) (Scharlau chemie) were used in the study.

Glass substrates, 16x25x1 (mm), covered with a conducting layer of Fluorine doped tin oxide (FTO) (Pilkington, Hart-ford Glass Co. Inc., USA) having sheet resistances of 8Ω/square were used as electrodes in EPD setup. These glass

substrates were cleaned with soap detergent mixed with so-dium hydroxide and then sonicated for 10 minutes in dis-tilled water to remove any stains that could interfere with adhesion of glass surface deposited film.

2.2 Apparatus and instruments

The equipment used included; Topedo weighing balance (Japan) used to weigh amount of metal oxide powder, Pyrex glass beaker as EPD cell, Power sonic 405 used for cleaning FTO glass slides and to stir (10min) the suspension to ensure homogeneity, DC power supply (Thurlby Thadar T S30225,30V, 2A dual power supply), and UV-Vis-NIR spec-trophotometer

2.3 EPD deposition of the composite film

Film preparation involved electrophoretic deposition of the TiO2 and Nb2O5 nanoparticles on fluorine-doped tin oxide (FTO) glass slide. To achieve this deposition, TiO2 and Nb2O5 nanopowders were mixed with 40mL propan-2-ol in a Pyrex glass to form the EPD suspension. Magnesium nitrate hexahydrate pellets were added to suspension to provide magnesium ions to attach to semiconductors and thereby control the zeta potential of suspension as reported by [18, 19]. The pH of suspension was measured using digital pH meter. Experimental set up of electrophoretic deposition technique comprised of photoelectrode and counter electrode both of FTO coated glass slides partially immersed in sus-pension contained in EPD cell in a parallel configuration and with DC electric field applied across the electrodes as in fig-ure 1 [20].

Figure 1 A schematic drawing of the EPD setup showing the two FTO cov-ered glass slides partially immersed into TiO2/Nb2O5/propan-2-olsuspension after [21,22]

V

A

DC supply Cathode

Nb2O5/TiO2/propan-2-ol suspension

Anode

In addition, the best quality films from visual inspection cor-respond to deposition time of 90 seconds.

50 100 150 200

20

40

% T

rans

mitta

nce

Deposition time (s)

% Transmittance

Figure 3 Variation of transmittance (at 1508nm) with deposition time for films electrophoretically deposited at35V from 0.01g/40mL suspension

3.3 Optimization of concentration

Figure 4 and figure 5 show the results for optimization of concentration at voltage of 35V ad deposition time of 90s. Figure 4(a) show the photographs from digital camera of the TiO2/Nb2O5 composite electrode thin films electrophoretical-ly deposited from powder concentration from 0.05g/40mL to 0.3g/40mL. The corresponding transmission spectra for var-ied concentrations are shown in figure 4 (b). The concentra-tion of 0.01g of metal oxide powders in 40mL of propan-2-ol (equivalent to 0.25g/L) produced electrophoretically deposit-ed films with a 55% transmittance, which was the highest obtained in the study. Visual inspection showed that the films corresponding to 0.01g/40ml concentration had the best quality films.

Figure 5 shows that for fixed applied voltage and deposition time, transmittance decreased with increased concentration. This decrease in transmittance resulted from development of thick films which hinder transmittance of light rays. Further, films deposited at high concentration would be non-porous and unsuitable for dye absorption for solar cell application.

The uncoated glass gave the highest transmittance value. It was difficult therefore to determine the minimum concentra-tion from transmittance values as shown in figure 5. Powder concentration of 0.01 g/40mL was therefore chosen as opti-mum EPD concentration value for fabrication of TiO2/Nb2O5composite electrode thin films.

.

0 1000 2000 3000

0

20

40

60

% (T

rans

mitt

ance

)

Wavelength (nm)

c35V,90s, 0.10g/40mL c35V,90s,0.015g/40mL c35V,90s,0.02g/40mL c.35,90,0.025 g/40mL c35V,90s,0.30g/40mL

Figure 4. (a) Photographs of and (b) Transmittance spectra for TiO2/Nb2O5composite electrode thin films electrophoretically deposited at 35 DCV for 90 s and from varying powder concentration.

0.01 0.02 0.03 0.04

20

30

40

50

% Transmittance

% T

rans

mitt

ance

concentration g/40mL

Figure 5 Variation of transmittance (at 1508nm) with concentration for films electrophoretically deposited at 35V from 0.01g/40mL suspension

Deposition time of 90s became the optimum EPD deposition time for fabrication of TiO2/Nb2O5 composite electrode thin films.

a

b

73 70

Morphological and Structural Characterization of TiO2/Nb2O5 Composite Electrode Thin Films Synthesized by Electrophoretic Deposition (EPD) Technique

J. N. Nguu, R. J Musembi, F. W. Nyongesa, & B.O. Aduda

Department of Physics, University of Nairobi, Nairobi, P. O. Box 30197-00100, Kenya

E-mail: [email protected] or [email protected]

Abstract Composite electrodes of titanium dioxide (TiO2) and niobium (V) oxide (Nb2O5) have been deposited onto glass substrates by electrophoretic deposition (EPD) technique. The mechanism of EPD involves charged particles in a liquid sus-pension being forced to move towards and deposit on an oppositely charged electrode upon application of electric field. In this study, TiO2 and Nb2O5 nano-sized powders were suspended in a Pyrex glass containing propan-2-ol. Magnesium nitrate hexa-hydrate (Mg (NO3).6H2O) pellets were added to the suspension to induce surface charges on the metal oxides. The structure of the thin films was characterized by X-ray Diffraction (XRD) while morphology was characterized by Scanning Electron Mi-croscope (SEM). The XRD spectra indicated that the films are of polycrystalline nature and that TiO2 and Nb2O5 particles were present in the deposited composite film. Transmittance of the deposited films was measured using the UV-Vis-NIR spectro-photometer spectra in the range 200 to 3200nm. The visual inspection of the films and the morphological investigations from SEM showed that porous films of high quality with well controlled morphology can be deposited using the EPD technique. Further work is needed to evaluate the potential of TiO2/Nb2O5 composite electrode thin films deposited by EPD for dye-sensitized solar cells applications.

Keywords Electrophoretic deposition (EPD), Transmittance, TiO2/Nb2O5 composite electrode thin films, Solar energy materials, Photovoltaic application,

1. Introduction

The development and application of renewable energy tech-nologies is the focus of increased scientific research to help mitigate against challenges of climatic change and the possi-ble depletion of the fossil fuels. Solar energy represents animportant renewable energy resource which is carbon free, and ubiquitous, with terrestrial potential about 600TW. It is also useful for off-grid utilizations. The commonly used pho-tovoltaic systems are crystalline silicon-based with positive attribute of high power conversion efficiencies (24.7%) [1]. However, these modules require vacuum-based techniques for fabrication resulting in high production costs. A lot of re-search is focused on developing solar cells that require low production costs and have adequate conversion efficiencies. Consequently, among the excitonic solar cells, the dye-sensitized solar cells (DSSC) have attracted widespread in-terest owing to their easy f1abrication processes, low produc-tion costs, less sensitivity to impurities, wide range of tem-perature for optimal operation, compatibility with glass and flexible substrates and fair solar–to–electrical conversion ef-ficiencies [2].

Whereas DSSCs have achieved conversion efficiencies of 11.6% [3], there are obstacles to overcome in order to in-crease efficiency and make the cells reproducible on large scale. Dye sensitized solar cells contain a porous, semicon-ductor material (photoelectrode) which absorb dye molecules and conduct injected photoelectrons. The dye molecules play the role of capturing light photons by extending the optical adsorption spectrum of the wide energy gap (Eg > 3eV) sem-iconductor. Titaniun dioxide (TiO2) is the most widely used semiconductor for preparing DSSC photoelectrodes because it provides a high energy conversion efficiency[2]. The high efficiency is the result of titanium dioxide films providing ease of electron transmission through them thus enhancing electron conduction to the conducting glass before recombi-nation. However, to further reduce recombination and in-crease efficiency, research efforts are geared towards modi-fication of the photoelectrode through use of other large band gap semiconductors and through use of compo-sites[4,5]. According to Lin et al., (2009)[6], the importance of thin film technology in fabricating solar cells is that thin films shortens the diffusion length of photo generated carri-ers and decreases the recombination. Use of thin films saves on material and simplifies the synthesis of solar cells. Modi-

fication of photoelectrode can also be achieved by use of composites [7,8].

Composites are regarded as engineered materials made from two or more constituent materials with significant different physical or chemical properties which remain separate and distinct at the microscopic scale within the finished structure. A number of authors [9,10,11] have proposed that niobium pentoxide (Nb2O5), a wide band gap (Eg=3.49eV), and n-type semiconductor, can be combined with titanium dioxide (TiO2) to fabricate composite thin films.

Thin films of titanium dioxide have been fabricated by varie-ty of methods like doctor blade [12], power screen printing [13], DC magnetron [10], and chemical vapor deposition (CVD) [14]. Some of these methods require high power con-sumption and vacuum chambers. In addition, electrophoretic deposition (EPD) technique has not been fully exploited in photoelectrode preparation for dye-sensitized solar cells de-spite being a relatively cheap technique and which could be scaled up for mass production if successfully developed. EPD is a colloidal technique in which charged particles are forced to move towards and deposit on to an oppositely charged electrode by the application of a DC electric field [15,16,23]. The advantages of EPD technique include good control of film thickness, short deposition time, high deposi-tion speed, simple equipment for its construction, wide ap-plicability to diverse materials and low manufacturing tem-perature. Further, EPD does not require highly purified mate-rials and vacuum chambers for film deposition [12, 14, 18]

Literature survey reveal that authors have used diverse val-ues for EPD process parameters and therefore there is need to optimize the same parameters for composite deposition [17,19,23,24,25]. As a result, EPD process parameters have been optimized and the technique used to deposit good quali-ty-nanocrystalline and nanoporous TiO2/Nb2O5 composite electrode thin films for application in the dye-sensitized solar cells.

2. Materials and methods 2.1 Materials

Commercial TiO2 nanopowder (Cas No. 13463-67-7 Al-drich), Nb2O5 nanopowder (Cas No. 1313-96-8 Acros Or-ganics BVBA, Belgium), 5x10-5M- Magnesium nitrate hexa-hydrate (Mg(NO3)2.6H2O 99.9%, Aldrich), and propan-2-ol (Isopropyl alcohol: IPA) (Scharlau chemie) were used in the study.

Glass substrates, 16x25x1 (mm), covered with a conducting layer of Fluorine doped tin oxide (FTO) (Pilkington, Hart-ford Glass Co. Inc., USA) having sheet resistances of 8Ω/square were used as electrodes in EPD setup. These glass

substrates were cleaned with soap detergent mixed with so-dium hydroxide and then sonicated for 10 minutes in dis-tilled water to remove any stains that could interfere with adhesion of glass surface deposited film.

2.2 Apparatus and instruments

The equipment used included; Topedo weighing balance (Japan) used to weigh amount of metal oxide powder, Pyrex glass beaker as EPD cell, Power sonic 405 used for cleaning FTO glass slides and to stir (10min) the suspension to ensure homogeneity, DC power supply (Thurlby Thadar T S30225,30V, 2A dual power supply), and UV-Vis-NIR spec-trophotometer

2.3 EPD deposition of the composite film

Film preparation involved electrophoretic deposition of the TiO2 and Nb2O5 nanoparticles on fluorine-doped tin oxide (FTO) glass slide. To achieve this deposition, TiO2 and Nb2O5 nanopowders were mixed with 40mL propan-2-ol in a Pyrex glass to form the EPD suspension. Magnesium nitrate hexahydrate pellets were added to suspension to provide magnesium ions to attach to semiconductors and thereby control the zeta potential of suspension as reported by [18, 19]. The pH of suspension was measured using digital pH meter. Experimental set up of electrophoretic deposition technique comprised of photoelectrode and counter electrode both of FTO coated glass slides partially immersed in sus-pension contained in EPD cell in a parallel configuration and with DC electric field applied across the electrodes as in fig-ure 1 [20].

Figure 1 A schematic drawing of the EPD setup showing the two FTO cov-ered glass slides partially immersed into TiO2/Nb2O5/propan-2-olsuspension after [21,22]

V

A

DC supply Cathode

Nb2O5/TiO2/propan-2-ol suspension

Anode

2.4 Method

Various amounts (0.01g to 0.3g) of TiO2 and Nb2O5 powderwere added into 40 milliliters of propan-2-olin a glass beaker to form EPD suspension. In each case, the two metal pow-ders were mixed in ratio of 1:1. The suspension was then stirred using power sonic 405 vibrator for 10 minutes. The color of the suspension whose pH was 4.9 turned milky white after vibration. The fluorine doped tin oxide (FTO) glass slides were partially immersed in the suspension and various values of DC voltage (25V to 60V) applied across electrodes. Electrophoretic deposition was carried out using varied deposition times (60s to 180s). Film deposition oc-curred on the cathode showing that TiO2 and Nb2O5 particles acquired positive surface charges when suspended in propan-2-ol.

The fabricated films were annealed at 400 deg C for 30 minutes and then allowed to cool gradually to room tempera-ture (25 deg C).

The TiO2/Nb2O5 composite electrode thin films were as-sessed from visual inspection and using UV-Vis-NIR spec-trophotometer (transmission spectra) to judge the quality and morphology of the deposited films. Morphology of films was further determined using Scanning electron mi-croscope (SEM).

3.0 Theory

3.1 Theory of EPD parameters The mass of electrophoretically deposited film on conducting substrate is related to EPD parameter according to Hamak-er’s equation [15]

tlVCm r

132

0 (1)

Where m is deposit yield, C is particle mass concentration in the suspension, o is permittivity of free space, r is relative permittivity of the solvent, is the zeta potential of the parti-cles, is the viscosity of the solvent, V is the applied DC voltage, l as the distance between the electrodes, and t is the deposition time.

The parameters in Eq. (1) can be categorized into process pa-rameters and materials-related parameters. Process related parameters include concentration (C), electric field (E=V/L), and deposition time (t). The Hamaker’s equation reduces to,

tEcm (2)

Therefore concentration, voltage and deposition time are the process related parameter we sought to optimize for fabrica-tion of TiO2/Nb2O5 composite electrode thin films

3.2 Optimization of deposition time

Figure 2 and figure 3 show results for optimization of deposi-tion time. The photograph and corresponding transmittance spectra shown in figure 2 (a) and figure 2 (b) are for TiO2/Nb2O5 composite electrode thin films electrophoretically deposited at 35 DC Voltage, from 0.01g/40mL suspension and for deposition times from 60s to 180s.

Transmittance increased from 0% to 55% in case of 90s depo-sition time and then decreased with increased time (figure 3). As can be seen in figure 5, the maximum value of transmit-tance was for deposition time of 90s, which is therefore the optimum deposition time. For short deposition times (<90s), the transmittance decreased since time was too short for for-mation of appreciable layer of film from the low level of con-centration. Longer deposition times yielded thicker films, which led to reduction in the transmittance values. Thick films are generally non-porous and do not absorb much dye. They are undesirable in solar cells applications.

0 1000 2000 3000

0

20

40

60

% (T

rans

mitta

nce)

wavelength (nm)

0.01g/40mL, 35V,90s c.01g/40mL,35V,60s c.01g/40mL,35V,120s c.01g/40mL,35V,150s c.01g/40mL,35V,180s

Figure 2 (a) Photographs of and (b) Transmittance spectra for TiO2/Nb2O5composite electrode thin films electrophoretically deposited from 0.01 g/40mL, at 35DCV for varying deposition time

b

a

a

b

REFERENCES [1] R. Jose, V. Thavasi, and S. Ramakrishna, Journal of the

American Ceramic Society, 92(2) (2009) 289-301.

[2] B. O’Regan, and M. Gretzel, Nature, 353 (1991) 737–740.

[3] Chiba, et al., Japanese Journal of Applied Physics, 45 (25) (2006) L638-640.

[4] A. Taotao, Journal of Wuhan University of Technology,24 (5) (2009) 732-735.

[5] E. Palomares, N. Clifford, S.A. Haque, T. Lutz, and J. R. Durrant, Chemical communications, 14 (2002) 1464-1465,.

[6] H. Lin, W-L. Wang, Y-Z. Liu, X. Li, and J-B. Li, Front. Mater. Sci. China, 3(4) (2009) 345-352.

[7] A. Kitiyanan, and S. Yoshikawa, Material Letters, 59 (2005) 4038-4040.

[8] K. Tonooka, T. Chiu, and N. Kikuchi, Applied Surface Science, 255 (2009) 9695-9698,.

[9] J. Sancho-Parramon, V. Janicki, and H. Zorc, Thin Solid Films, 516 (16) ( 2008) 5478-5482.

[10] K. Eguchi, H. Hoga, K. Sekizawa, and K. Sasaki, Jour-nal of the Ceramic Society Japan, 108 (2000) 1067-1071.

[11] S. G. Chen, S. Chappel, Y. Diamant, and A. Zaban, Chemistry of Materials, 13 (12) (2001) 4629-4634,2001.

[12] Yamada, et al., Thin Solid Films, 518 (2010) 3101-3104.

[13] J. J. Van Tassel, and C. A. Randall, Key Engineering Materials, 314 (2006) 167-174.

[14] H. Dickerson, and A. Boccaccini (Eds.), Electrophoretic Deposition of nano-materials XII, (2012) 376, ISBN: 978-1-4419-9690-9.

[15] O. Van der Biest, and L. J. Vandeperre, Annual Reviews Material Science, 29 (1999) 327-352.

[16] T. Miyasaka, and Y. Kijitori, Journal of Electrochemical Society, 151 (11) (20040 A1767-1773.

[17] Lee et al., (2011) 182-210. Dr. Brahim Attaf (Ed.), ISBN: 978-953-307-235-7, In Tech,

[18] J. Bandy, Q. Zhang, and G. Cao, Materials Sciences and Applications, 2 (2011) 1427-1431.

[19] J-H. Yum, S-S. Kim, D-Y. Kim, and Y-E. Sung, J. Pho-tochemistry and Photobiology A: Chemistry, 173 (2005) 1-6.

[20] Nyongesa, F. W., Nyaga, W. G., and Aduda, B. O.,Poster presented at the 6th Edward Bouchet-Abdus Salam Institute International Conference, iThemba LABS, Cape Town, South Africa, 2007.

[21] P. Sarkar, and P. S. Nicholson, , Journal of American Ceramic Society. 79 (1996) 1987–2002. (doi:10.1111/j.1151- 2916.1996.tb08929.x)

[22] J. Will, M. K. M. Hruschka, L. Gubler, and L. J. Gauck-ler, Journal of the American Ceramic Society, 84 (2) (2001) 328-332.

[23] W. Jarernboon, et al., Thin Solid Films (2009), doi: 10.1016/j.tsf.2009.02.129

[24] L. Grinis, S. Dor, A. Ofir, A. Zaban, Journal of Photo-chemistry and Photobiology A: Chemistry (2007), doi: 10.1016/j.jphotochem.2008.02.015.

[25] J. Cho, S. Schaab, and J. A. Roether, Journal of Nano-particle Research, 10 (2008) 99-105 DOI 10.1007/s11051-007-9230-x.

[26] Chang, et al., Materials Transactions, 50 (12) (2009) 879-2884.

71 72

2.4 Method

Various amounts (0.01g to 0.3g) of TiO2 and Nb2O5 powderwere added into 40 milliliters of propan-2-olin a glass beaker to form EPD suspension. In each case, the two metal pow-ders were mixed in ratio of 1:1. The suspension was then stirred using power sonic 405 vibrator for 10 minutes. The color of the suspension whose pH was 4.9 turned milky white after vibration. The fluorine doped tin oxide (FTO) glass slides were partially immersed in the suspension and various values of DC voltage (25V to 60V) applied across electrodes. Electrophoretic deposition was carried out using varied deposition times (60s to 180s). Film deposition oc-curred on the cathode showing that TiO2 and Nb2O5 particles acquired positive surface charges when suspended in propan-2-ol.

The fabricated films were annealed at 400 deg C for 30 minutes and then allowed to cool gradually to room tempera-ture (25 deg C).

The TiO2/Nb2O5 composite electrode thin films were as-sessed from visual inspection and using UV-Vis-NIR spec-trophotometer (transmission spectra) to judge the quality and morphology of the deposited films. Morphology of films was further determined using Scanning electron mi-croscope (SEM).

3.0 Theory

3.1 Theory of EPD parameters The mass of electrophoretically deposited film on conducting substrate is related to EPD parameter according to Hamak-er’s equation [15]

tlVCm r

132

0 (1)

Where m is deposit yield, C is particle mass concentration in the suspension, o is permittivity of free space, r is relative permittivity of the solvent, is the zeta potential of the parti-cles, is the viscosity of the solvent, V is the applied DC voltage, l as the distance between the electrodes, and t is the deposition time.

The parameters in Eq. (1) can be categorized into process pa-rameters and materials-related parameters. Process related parameters include concentration (C), electric field (E=V/L), and deposition time (t). The Hamaker’s equation reduces to,

tEcm (2)

Therefore concentration, voltage and deposition time are the process related parameter we sought to optimize for fabrica-tion of TiO2/Nb2O5 composite electrode thin films

3.2 Optimization of deposition time

Figure 2 and figure 3 show results for optimization of deposi-tion time. The photograph and corresponding transmittance spectra shown in figure 2 (a) and figure 2 (b) are for TiO2/Nb2O5 composite electrode thin films electrophoretically deposited at 35 DC Voltage, from 0.01g/40mL suspension and for deposition times from 60s to 180s.

Transmittance increased from 0% to 55% in case of 90s depo-sition time and then decreased with increased time (figure 3). As can be seen in figure 5, the maximum value of transmit-tance was for deposition time of 90s, which is therefore the optimum deposition time. For short deposition times (<90s), the transmittance decreased since time was too short for for-mation of appreciable layer of film from the low level of con-centration. Longer deposition times yielded thicker films, which led to reduction in the transmittance values. Thick films are generally non-porous and do not absorb much dye. They are undesirable in solar cells applications.

0 1000 2000 3000

0

20

40

60

% (T

rans

mitta

nce)

wavelength (nm)

0.01g/40mL, 35V,90s c.01g/40mL,35V,60s c.01g/40mL,35V,120s c.01g/40mL,35V,150s c.01g/40mL,35V,180s

Figure 2 (a) Photographs of and (b) Transmittance spectra for TiO2/Nb2O5composite electrode thin films electrophoretically deposited from 0.01 g/40mL, at 35DCV for varying deposition time

b

a

a

b

3.4 Optimization of applied voltage

The results for optimization of applied DC voltage corre-

spond to figure 6 and figure 7.

The photographs of electrophoretically deposited

TiO2/Nb2O5 composite thin films show that the films were

deposited on the glass slides (figure 6). The corresponding

transmission spectra, for films electrophoretically deposited

from 0.01g per 40mL suspension and 90s, yielded the 55%

transmittance value corresponding to applied DC voltage of

35V (figure 6).

Figure 7 shows that transmittance decreased with increased

applied voltage. Maximum transmittance (55%) was ob-

tained for voltage of 35V. Higher voltages (<35V) produced

relatively thick films which yielded low transmittance val-

ues. Typically, thick films hinder transmission of light rays

through them resulting to low transmittance values. The

thick films resulting from deposition at high voltages were

non-porous and unsuitable for dye absorption. In addition,

thick films decrease the electron diffusion which result in

low conversion efficiency. It follows, that optimum applied

DC voltage for deposition of TiO2/Nb2O5 composite thin

films is 35V

0 1000 2000 3000

0

20

40

60

% T

rans

mitta

nce

Wavelength (nm)

0.01g/40mL, 90s, 35V c.01g/40mL,90s,25V c.01g/40mL,90s,45V c.01g/40mL,90s,55V c.01g/40mL,90s,60V

Figure 2 (a) Photographs of and (b) Transmittance spectra for TiO2/Nb2O5composite electrode thin films electrophoretically deposited for 90 s from 0.01g/40mL, and for varying voltages

20 40 60

20

30

40

50

% T

rans

mitt

ance

Applied DC Voltage ( (V)

% Transmittance

Figure 3 Variation of Transmittance (at 1508nm) with applied voltage for films electrophoretically deposited at 35V from 0.01g/40mL suspension

The optimum EPD parameters are presented in summary in Table 1. Table 2 gives comparison of different EPD processparameter, namely DC applied voltage, deposition times and concentrations. Even for electrophoretic deposition of titani-um dioxide alone, a great variation in the EPD process pa-rameters was observed. These results underscore the im-portance of EPD process optimization especially when coat-ing a composite film of TiO2/Nb2O5 as used in this study.

a b

a

Table 1. Optimized electrophoretic deposition parameters

Concentration of Applied DC voltage distance between Deposition TiO2 & Nb2O5 particles across electrodes electrodes time 0.01g/40mL (=0.25g/L) 35V 12 (mm) 90 sec

Table 2 Variety of EPD optimum process parameters

Applied Volt-age (V)

Electrode sepa-ration

Concentration Deposition time

Photoelectrode Reference

75 2cm 0.25g/L 180 s TiO2 on flexible DSSC Yum et al (2005) [19]

40 2cm 0.25g/L 15min TiO2 on FTO glass Bandy et al (2011) [7]

5-20 1cm 0.5g/50mL 1-8min TiO2 on FTO glass Jarernboon, et al., (2009) [23]6 0.3mm - 0.5-1min TiO2 on conductive glass Miyasaka, (2003) [16]

40 18mm 2.4g/L 2 min TiO2 FTO glass compressed Grinis et al (2008) [24]

55 1cm 0.6/100mL 4min CNT/TiO2 composites Cho, (2008) [25]

3.5 Morphological and XRD characterization of fabricat-ed TiO2/Nb2O5 composite thin films

3.5.1 Morphology of deposited films

The morphology of TiO2 & Nb2O5 composite thin films elec-trophoretically deposited using combination 0.01g/40mL, 90s, & 35DCV, was determined using SEM micrographs (Figures 8 to 10). SEM used had EHT=10.00KV. same mag-nification was used for the images as shown in figure 8 to 13.

The SEM images of deposited TiO2/Nb2O5 composite thin films, reveal that TiO2 and Nb2O5 nanoparticles in the com-posite films were of near uniform distribution with a small fraction of uncoated glass surface. The more connected the composite particles are to each other the more the electron transport to the conductive substrate. The surface coating appears to adhere well to the glass surface. This can be at-tributed to magnesium ions that attached to the surface of metal oxide particles. In addition, these films have the de-sired porosity for dye absorption and are therefore suitable for application in the dye-sensitized solar cells(DSSC)

Figure 4 SEM micrograph of TiO2/Nb2O5 composite thin film on a F:SnO2(mag x 200K)

Figure 5 SEM micrograph of micrograph of Nb2O5 semiconductor thin film on a F:SnO2 (mag = 201.12KX)

REFERENCES [1] R. Jose, V. Thavasi, and S. Ramakrishna, Journal of the

American Ceramic Society, 92(2) (2009) 289-301.

[2] B. O’Regan, and M. Gretzel, Nature, 353 (1991) 737–740.

[3] Chiba, et al., Japanese Journal of Applied Physics, 45 (25) (2006) L638-640.

[4] A. Taotao, Journal of Wuhan University of Technology,24 (5) (2009) 732-735.

[5] E. Palomares, N. Clifford, S.A. Haque, T. Lutz, and J. R. Durrant, Chemical communications, 14 (2002) 1464-1465,.

[6] H. Lin, W-L. Wang, Y-Z. Liu, X. Li, and J-B. Li, Front. Mater. Sci. China, 3(4) (2009) 345-352.

[7] A. Kitiyanan, and S. Yoshikawa, Material Letters, 59 (2005) 4038-4040.

[8] K. Tonooka, T. Chiu, and N. Kikuchi, Applied Surface Science, 255 (2009) 9695-9698,.

[9] J. Sancho-Parramon, V. Janicki, and H. Zorc, Thin Solid Films, 516 (16) ( 2008) 5478-5482.

[10] K. Eguchi, H. Hoga, K. Sekizawa, and K. Sasaki, Jour-nal of the Ceramic Society Japan, 108 (2000) 1067-1071.

[11] S. G. Chen, S. Chappel, Y. Diamant, and A. Zaban, Chemistry of Materials, 13 (12) (2001) 4629-4634,2001.

[12] Yamada, et al., Thin Solid Films, 518 (2010) 3101-3104.

[13] J. J. Van Tassel, and C. A. Randall, Key Engineering Materials, 314 (2006) 167-174.

[14] H. Dickerson, and A. Boccaccini (Eds.), Electrophoretic Deposition of nano-materials XII, (2012) 376, ISBN: 978-1-4419-9690-9.

[15] O. Van der Biest, and L. J. Vandeperre, Annual Reviews Material Science, 29 (1999) 327-352.

[16] T. Miyasaka, and Y. Kijitori, Journal of Electrochemical Society, 151 (11) (20040 A1767-1773.

[17] Lee et al., (2011) 182-210. Dr. Brahim Attaf (Ed.), ISBN: 978-953-307-235-7, In Tech,

[18] J. Bandy, Q. Zhang, and G. Cao, Materials Sciences and Applications, 2 (2011) 1427-1431.

[19] J-H. Yum, S-S. Kim, D-Y. Kim, and Y-E. Sung, J. Pho-tochemistry and Photobiology A: Chemistry, 173 (2005) 1-6.

[20] Nyongesa, F. W., Nyaga, W. G., and Aduda, B. O.,Poster presented at the 6th Edward Bouchet-Abdus Salam Institute International Conference, iThemba LABS, Cape Town, South Africa, 2007.

[21] P. Sarkar, and P. S. Nicholson, , Journal of American Ceramic Society. 79 (1996) 1987–2002. (doi:10.1111/j.1151- 2916.1996.tb08929.x)

[22] J. Will, M. K. M. Hruschka, L. Gubler, and L. J. Gauck-ler, Journal of the American Ceramic Society, 84 (2) (2001) 328-332.

[23] W. Jarernboon, et al., Thin Solid Films (2009), doi: 10.1016/j.tsf.2009.02.129

[24] L. Grinis, S. Dor, A. Ofir, A. Zaban, Journal of Photo-chemistry and Photobiology A: Chemistry (2007), doi: 10.1016/j.jphotochem.2008.02.015.

[25] J. Cho, S. Schaab, and J. A. Roether, Journal of Nano-particle Research, 10 (2008) 99-105 DOI 10.1007/s11051-007-9230-x.

[26] Chang, et al., Materials Transactions, 50 (12) (2009) 879-2884.

77 74

3.4 Optimization of applied voltage

The results for optimization of applied DC voltage corre-

spond to figure 6 and figure 7.

The photographs of electrophoretically deposited

TiO2/Nb2O5 composite thin films show that the films were

deposited on the glass slides (figure 6). The corresponding

transmission spectra, for films electrophoretically deposited

from 0.01g per 40mL suspension and 90s, yielded the 55%

transmittance value corresponding to applied DC voltage of

35V (figure 6).

Figure 7 shows that transmittance decreased with increased

applied voltage. Maximum transmittance (55%) was ob-

tained for voltage of 35V. Higher voltages (<35V) produced

relatively thick films which yielded low transmittance val-

ues. Typically, thick films hinder transmission of light rays

through them resulting to low transmittance values. The

thick films resulting from deposition at high voltages were

non-porous and unsuitable for dye absorption. In addition,

thick films decrease the electron diffusion which result in

low conversion efficiency. It follows, that optimum applied

DC voltage for deposition of TiO2/Nb2O5 composite thin

films is 35V

0 1000 2000 3000

0

20

40

60

% T

rans

mitta

nce

Wavelength (nm)

0.01g/40mL, 90s, 35V c.01g/40mL,90s,25V c.01g/40mL,90s,45V c.01g/40mL,90s,55V c.01g/40mL,90s,60V

Figure 2 (a) Photographs of and (b) Transmittance spectra for TiO2/Nb2O5composite electrode thin films electrophoretically deposited for 90 s from 0.01g/40mL, and for varying voltages

20 40 60

20

30

40

50

% T

rans

mitt

ance

Applied DC Voltage ( (V)

% Transmittance

Figure 3 Variation of Transmittance (at 1508nm) with applied voltage for films electrophoretically deposited at 35V from 0.01g/40mL suspension

The optimum EPD parameters are presented in summary in Table 1. Table 2 gives comparison of different EPD processparameter, namely DC applied voltage, deposition times and concentrations. Even for electrophoretic deposition of titani-um dioxide alone, a great variation in the EPD process pa-rameters was observed. These results underscore the im-portance of EPD process optimization especially when coat-ing a composite film of TiO2/Nb2O5 as used in this study.

a b

a

Table 1. Optimized electrophoretic deposition parameters

Concentration of Applied DC voltage distance between Deposition TiO2 & Nb2O5 particles across electrodes electrodes time 0.01g/40mL (=0.25g/L) 35V 12 (mm) 90 sec

Table 2 Variety of EPD optimum process parameters

Applied Volt-age (V)

Electrode sepa-ration

Concentration Deposition time

Photoelectrode Reference

75 2cm 0.25g/L 180 s TiO2 on flexible DSSC Yum et al (2005) [19]

40 2cm 0.25g/L 15min TiO2 on FTO glass Bandy et al (2011) [7]

5-20 1cm 0.5g/50mL 1-8min TiO2 on FTO glass Jarernboon, et al., (2009) [23]6 0.3mm - 0.5-1min TiO2 on conductive glass Miyasaka, (2003) [16]

40 18mm 2.4g/L 2 min TiO2 FTO glass compressed Grinis et al (2008) [24]

55 1cm 0.6/100mL 4min CNT/TiO2 composites Cho, (2008) [25]

3.5 Morphological and XRD characterization of fabricat-ed TiO2/Nb2O5 composite thin films

3.5.1 Morphology of deposited films

The morphology of TiO2 & Nb2O5 composite thin films elec-trophoretically deposited using combination 0.01g/40mL, 90s, & 35DCV, was determined using SEM micrographs (Figures 8 to 10). SEM used had EHT=10.00KV. same mag-nification was used for the images as shown in figure 8 to 13.

The SEM images of deposited TiO2/Nb2O5 composite thin films, reveal that TiO2 and Nb2O5 nanoparticles in the com-posite films were of near uniform distribution with a small fraction of uncoated glass surface. The more connected the composite particles are to each other the more the electron transport to the conductive substrate. The surface coating appears to adhere well to the glass surface. This can be at-tributed to magnesium ions that attached to the surface of metal oxide particles. In addition, these films have the de-sired porosity for dye absorption and are therefore suitable for application in the dye-sensitized solar cells(DSSC)

Figure 4 SEM micrograph of TiO2/Nb2O5 composite thin film on a F:SnO2(mag x 200K)

Figure 5 SEM micrograph of micrograph of Nb2O5 semiconductor thin film on a F:SnO2 (mag = 201.12KX)

Figure 6 SEM micrograph of micrograph of TiO2 semiconductor thin film on a F:SnO2 (mag = 200.00KX)

3.5.2 X-RD characterization

Figure 11-13 show the X-Ray Diffraction (X-RD) spectra for composite TiO2/Nb2O5 , TiO2 and Nb2O5 thin films electro-phoretically deposited at 35DCV, 0.01g/40mL, and 90 sec-onds. In figure 11, the dominant peak of TiO2 was at 2-theta of 25.5 deg while that of Nb2O5 was at 26.5 deg. However the peak (counts=1600) of TiO2 was longer than that of Nb2O5 (900 counts) which translates to ratio 1.78 to 1. The length of peaks corresponds to degree of crystallinity [26]. Itfollows therefore that the crystallinity of TiO2 is higher than that of Nb2O5.

Figure 11 XRD spectra of TiO2/Nb2O5 composite thin films annealed at 4000C for 30 minutes under atmospheric conditions.

Figure 7 XRD graph for electrophoretically deposited thin films of TiO2 in 2-propanol

Figure 8 XRD graphs for electrophoretically deposited thin films of Nb2O5 in 2-propanol

4 Conclusions EPD process parameters which include concentration, applied voltage and deposition time, were successfully optimized in this study. Values of 0.01g/40mL, 35DCV, and 90s respective-ly were obtained as optimum values for fabrication of TiO2/Nb2O5 composite electrode thin films. The deposited films were seen to be crack free and of good porosity. Moreo-ver, a pH of ≈ 4.9 was found to be optimal for propan-2-ol sus-pension used in EPD method. Both TiO2 and Nb2O5 nanoparti-cles were shown, by XRD graphs, to be present in the compo-site films in the ratio 1.78 to 1 (or 1:0.56). The EPD technique after optimization, has been used in the study to fabricate good-quality TiO2/Nb2O5 composite thin films, and of smooth mor-phology as shown by SEM images. Further work is needed to evaluate the potential of TiO2/Nb2O5 composite films deposited by EPD for solar cell applications.

75 76

Figure 6 SEM micrograph of micrograph of TiO2 semiconductor thin film on a F:SnO2 (mag = 200.00KX)

3.5.2 X-RD characterization

Figure 11-13 show the X-Ray Diffraction (X-RD) spectra for composite TiO2/Nb2O5 , TiO2 and Nb2O5 thin films electro-phoretically deposited at 35DCV, 0.01g/40mL, and 90 sec-onds. In figure 11, the dominant peak of TiO2 was at 2-theta of 25.5 deg while that of Nb2O5 was at 26.5 deg. However the peak (counts=1600) of TiO2 was longer than that of Nb2O5 (900 counts) which translates to ratio 1.78 to 1. The length of peaks corresponds to degree of crystallinity [26]. Itfollows therefore that the crystallinity of TiO2 is higher than that of Nb2O5.

Figure 11 XRD spectra of TiO2/Nb2O5 composite thin films annealed at 4000C for 30 minutes under atmospheric conditions.

Figure 7 XRD graph for electrophoretically deposited thin films of TiO2 in 2-propanol

Figure 8 XRD graphs for electrophoretically deposited thin films of Nb2O5 in 2-propanol

4 Conclusions EPD process parameters which include concentration, applied voltage and deposition time, were successfully optimized in this study. Values of 0.01g/40mL, 35DCV, and 90s respective-ly were obtained as optimum values for fabrication of TiO2/Nb2O5 composite electrode thin films. The deposited films were seen to be crack free and of good porosity. Moreo-ver, a pH of ≈ 4.9 was found to be optimal for propan-2-ol sus-pension used in EPD method. Both TiO2 and Nb2O5 nanoparti-cles were shown, by XRD graphs, to be present in the compo-site films in the ratio 1.78 to 1 (or 1:0.56). The EPD technique after optimization, has been used in the study to fabricate good-quality TiO2/Nb2O5 composite thin films, and of smooth mor-phology as shown by SEM images. Further work is needed to evaluate the potential of TiO2/Nb2O5 composite films deposited by EPD for solar cell applications.

Optical and Electrical Properties of PTO Thin Films Prepared by Spray Pyrolysis

B.V Odari1, M.J Mageto1, R.J Musembi2, F.M Gaitho1, V.W Muramba1,H. Othieno3

1Department of Physics, MasindeMuliro University of Science and Technology, P.O Box 190, 50100, Kakamega, Kenya2Department of Physics, University of Nairobi, P.O Box 30197-00100, Nairobi, Kenya

3Department of Physics, Maseno University, P.O Box Private Bag, Maseno, KenyaE-mail: [email protected]

Abstract SnO2: Pd thin films were deposited on glass substrate at 450±100C using Tin (IV) Chloride (SnCl4.5H2O) solution and Palladium Chloride (PdCl2) solution. Investigations were made on the effect of increasing Pd concentration on the optical and electrical properties of the films. The minimum resistivity of 5.14×10-2Ωcm for 3.68at% Pd film was obtained. The optical properties were studied in the UV/VIS/NIR region. The optical band gap for undoped SnO2 films lies at 3.93 eV and Palladium doped films lay in the range 3.86 – 3.99 eV. Simulation of experimental spectra using SCOUT software using Drude and Kim Oscillator terms gave the optical constants for films on glass.

Keywords Palladium doped Tin Oxide, Spray Pyrolysis, Drude and Kim Oscillator model

1. Introduction

Tin oxide has received a great scientific interest because of its wide range of applications which include the field of sensors, opacities, transparent electrodes in solar panels and other electrochromic devices, thin film magnetic recording media and as material for Li-ion batteries [1, 2]. The film is highly transparent in the visible region, chemically inert, mechanically hard and can resist high temperatures [3, 4, 5] as it is only attacked by hot concentrated alkalis [2]. It be-longs to a class of materials that combines high electrical conductivity with optical transparency, and therefore, it constitutes an important component for the optoelectronic applications [1]. The efficiency in these applications is usu-ally improved by suitably doping the tin oxide for example, doping with Sb and F increases the conductivity of tin oxide [2, 6, 7, 8]. Tin oxide is a crystalline solid with a tetragonal crystal lattice. It is a wide gap, non-stoichiometric semi-conductor and behaves as a degenerate n-type semiconductor with a low resistivity ( Ωcm) [8]. It can exist in two structures belonging to an indirect band gap of about 2.6 eV [9] and direct band gap that ranges from 3.6 eV to 4.6 eV at room temperature [8,10]. Some of the most popular methods of depositing Tin oxide thin films include the following: Reactive sputtering [11], Plasma enhanced Chemical Vapour Deposition [1], Sol-gel [12], Electronspun [13], Inkjet printing [14], Electron beam Evaporation [5] and Spray pyrolysis [3, 4, 6, 15]. Of these methods the spray pyrolysis method represents the less expensive method since it can produce large area, high-quality and low cost thin films [6]. It is also suitable for substrates with complex geometry and can be used for a variety of oxide materials [16]. The purpose

of this research was to improve the properties of the spray deposited SnO2: Pd thin films as a suitable window pane gas sensor.

2. Experimental

2.1.Sample preparation

The substrates used were ordinary float glass slides which were 1.2 mm thick 2.5cm by 7.6 cm wide. The cleaning procedure involved scrubbing the glass slides gently on both sides using a cotton swab soaked in foam made from a mixture of deionized water, liquid detergent and sodium hydroxide in the ratio of 3:2:1. They were then drag wiped using a lens cleaning tissue held at an angle of 45 degrees to the horizontal before being wiped with Isopropyl alcohol and acetone respectively. Lastly, the substrates were ultrason-ically cleaned in distilled water for 30 minutes before drying with spray of pressurized air. Spray pyrolysis tech nique was used to deposit the films at a substrate temperature of 450 100Cusing compressed air at 1 bar as the atomiza-tion gas. The experimental set up of the in-house-made spray pyrolysis system is as shown in Figure 1. It consisted of a fume chamber, hot plate, spray nozzle of diameter 1mm, input gas valve, gas compressor, gas flow meter, conduit tube, thermocouple and pressure gauge.

[6,15,19, 24]. The band gap for palladium doped films is found to be wider than the undoped tin oxide film for 1.88at% Pd and 3.68 at% Pd. The increase in the energy gap can be correlated with the Moss-Burstein effect since the absorp-tion edge of the films shifts to shorter wavelength. This indicates an optimum level palladium doping on tin oxide causes a widening effect on the band gap and this may be attributed to the gradual increase in carrier concentration and mobility [19].

Figure 4: Energy bandgap for undoped tin oxide and Pd-doped tin oxide

films prepared at 450 100C, (i) Undoped SnO2 (ii) 1.88 at%

SnO2:Pd (iii) 3.68 at% SnO2:Pd (iv) 5.42 at% SnO2:Pd (v) 7.10 at%

SnO2:Pd

Figure 5 shows spectral refractive index (η) and extinction coefficient k for SnO2 and SnO2:Pd. The refractive index (η), was estimated from the transmission and reflectance data and it was found to be around 1.95 at 500 nm for the un-doped SnO2 film of thickness 167nm, deposited at 450 100C. It was observed that the refractive index of all the films decreases with wavelength and then attains almost a constant value towards higher wavelengths [5].

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.50

1x1011

2x1011

3x1011

4x1011

(hv

)2 (eV2 c

m-2 )

Energy (eV)

(i)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.50

1x1011

2x1011

3x1011

4x1011

5x1011

6x1011

7x1011

(hv

)2 (eV2 c

m-2 )

Energy (eV)

(ii)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.50.0

5.0x1010

1.0x1011

1.5x1011

2.0x1011

2.5x1011

3.0x1011

(hv

)2 (eV2 c

m-2 )

Energy (eV)

(iii)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.50.0

5.0x1010

1.0x1011

1.5x1011

2.0x1011

2.5x1011

(hv

)2 (eV2 c

m-2 )

Energy (eV)

(iv)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.50.0

2.0x1010

4.0x1010

6.0x1010

8.0x1010

1.0x1011

1.2x1011

1.4x1011

1.6x1011

1.8x1011

(hv

)2 (eV2 c

m-2 )

Energy (eV)

(v)

81 78

Figure1: Spray pyrolysis setup

The undoped SnO2 film was produced from a precursor solution consisting of Tin (IV) chloride (99%) prepared by dissolving completely 6g of stannic chloride with 100ml of ethanol (99.9%) [3,5,6]. Pd-doped SnO2 films were prepared by dissolving 0.5g of PdCl2 (59-60%Pd) in 50ml of ethanol (99.9%) then added to the spraying solution at varying volumes ranging from 2ml to 8ml (1.88 at% - 7.10 at%).

The spraying was done under the following conditions: Substrate temperature: 450±100C Carrier-gas: Compressed air at 1 bar Solution flow rate: 6ml/min Nozzle-to-substrate distance: 33cm horizon-

tally 32+3cm vertically After spraying, the films were left to cool with the hot plate before removal for transmittance and reflectance measure-ments.

2.2 Sample characterization The transmittance and reflectance measurements were done at near normal incidence in the solar wavelength range from 300nm to 2500nm on a computerized double beam sol-id-spec 3700DUV Shimadzu Spectrophotometer equipped with 198851 Barium Sulphate (BaSO4) integrating sphere. Barium Sulphate plate was used as a reference for the cal-culation of optical properties such as band gap, absorption coefficient and refractive index. The thickness of the as-deposited samples with Pd doping concentrations were estimated by fitting the experimental spectral data to theo-retical spectral data based on Drude and Kim analysis using the SCOUT software [17] in the wavelength range 300nm –2500nm. Electrical resistivity of the films was calculated from two adjustable parameters of Drude: plasma frequency Ωp and damping constant γ.


3.1. Analysis of optical properties

In order to compare the transparency of SnO2 thin films with various Pd-doping levels, the optical spectra in the UV-VIS-NIR region of the samples was measured. The optical transparency of SnO2: Pd thin films for various Pd-doping levels for both experimental and computed spec-tral are shown in Figure 2. The films exhibited a transpar-ency near 88% (for undoped tin oxide) at wavelength of 695nm. The transparency at 390 – 440nm decreased with increasing Pd doping concentration which is probably due to the increase in fundamental absorption as photon strikingincreases with increase in carrier concentration [18].

Figure 2: Experimental and Computed Spectral Transmittance and Reflec-

tance of undoped tin oxide and Pd-doped tin oxide films

500 1000 1500 2000 25000

10

20

30

40

50

60

70

80

90

100

0% doped SnO2 167nm 1.88at% Pd:SnO2 140.7nm 3.68at% Pd:SnO2 186.5nm 5.42at% Pd:SnO2 185.6nm 7.10at% Pd:SnO2 223.7nm

Tra

nsm

ittan

ce/R

efle

ctan

ce (

%)

Wavelength (nm)

EXPERIMENTAL

500 1000 1500 2000 25000

10

20

30

40

50

60

70

80

90

100

Tra

nsm

ittan

ce/R

efle

ctan

ce (

%)

Wavelength (nm)

0% doped SnO2 167nm 1.88at% Pd:SnO2 140.7nm 3.68at% Pd:SnO2 186.5nm 5.42at% Pd:SnO2 185.6nm 7.10at% Pd:SnO2 223.7nm

THEORETICAL

The maximum transmittance observed in the undoped tin oxide thin film may be attributed to a decrease in diffuse and multiple reflections caused by the increase in grain size and a reduction in light-scattering effect [19]. A sharp fall in transmission at about 310 nm is due to the absorption of the glass substrate [12].

Dispersion analysis using a model for the dielectric suscep-tibility of the film consisting of Drude [20] and Kim terms [21] was used to model the measured reflectance and transmittance. For doped semiconductors, the charge carri-ers set free by the donors or acceptors can be accelerated by very little energies and hence do respond to applied electric fields with frequencies in the infrared region [17]. The Drude model is a free electron contribution which describes the intraband contributions to the optical properties. This model has two adjustable parameters: plasma frequency, p and damping constant, . The plasma frequency is pro-portional to the square root of the carrier density and the damping constant is proportional to the inverse of the mo-bility. The Drude dielectric susceptibility Drude, expressed as a function of frequency, ω, is given as [17, 20]

iω

Ωωχ 2

2p

Drude

The one oscillator contribution developed by Kim contains four adjustable parameters: TO resonance frequency, p

oscillator strength, damping constant and Gauss–Lorentz-switch constant, . may vary betweenzero and infinity. For = 0, a Gaussian line shape is achieved. A large value of (larger than 5) leads to a Lo-rentzian line shape. The Kim oscillator models the weak broad interband absorption in the measured wavelength range. The interband dielectric susceptibility described by Kim is given by [20];

ωτωiωΩΩ

χ 22TO

2p

Oscillator Kim,

(1)

where

2

τ

TO2τ Ω

Ωωσ1

1expΩωτ (2)

Model parameters were determined from the best fit be-tween computed and experimental data, using Scout soft-ware [17]. The best fit gives the optical constants of the film

under study. Figure 3 shows spectral absorption coefficient (α) for SnO2 and SnO2:Pd with four doping levels. For all the films, the absorption edge lies in the UV region and increases with increase in palladium concentration.

Figure 3: Absorption coefficient (α) against wavelength (i) and Absorption

coefficient (α) against Energy for undoped tin oxide and

Pd-doped tin oxide films

The optical band gap, Eg, was determined using the stand-ard formula [22,23]:

αhν (hν – Eg)n (3)

Where is the absorption coefficient, hν, the photon energy, and n = 1/2 accounts for the fact that the directly allowed transitions across the bandgap are expected to dominate. Figure 4 shows plots of versus photo energy, hν, in the high absorption region. Extrapolation of the curve to hν = 0 gave the direct band gap of SnO2:Pd films in the range 3.86 eV – 3.99 eV for the Pd doped tin oxide films and 3.93 eV for the undoped tin oxide film; which is comparable with the values already reported

500 1000 1500 2000 25000

50000

100000

150000

200000

Alph

a (1

/cm

)

Wavelength (nm)

0% doped SnO2 1.88at% Pd:SnO2 3.68at% Pd:SnO2 5.42at% Pd:SnO2 7.10at% Pd:SnO2

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.50

50000

100000

150000

200000

Alph

a (1

/cm

)

Energy (eV)


79 80

Figure 5: Spectral refractive index and extinction coefficient for SnO2 and

SnO2:Pd thin films.

Figure 6: Energy gap versus Pd concentration

Figure 6 shows the band gap energy of the undoped and Pddoped SnO2 thin films with different atomic percentage

of Pd. It is observed that the band gap energy increases with Pdpercentage up to 1.88at% then it decreases when Pdper-centage is more than 1.88at% which is in accordance with the findings of Magetoet.al., 2012 and Fatemaet.al., 2011 [16, 26]. This is due to the decrease in the number of charge carriers with increase in Pd doping.3.2. Analysis of Electrical Properties In most cases it is the electric field of the probing light wave that interacts with the sample. Hence excitations can be observed in optical experiments that are going along with a polarization. The polarization P induced by an ex-ternally applied electric field E in a homogeneous material is given by the electric susceptibility χ;

(4)The dielectric function ε which connects the dielectric dis-placement and the electric field vector is closely related to the susceptibility χ;

(5)The frequency dependence of the susceptibility is very characteristic for a material since itincorporates vibrations of the electronic system and the atomic cores as well as contributions from free charge carriers. The doping of sem-iconductors leads to free charge carriers which can be in-vestigated by IR spectroscopy. The response of the free car-riers to oscillating electric fields can be described to a good approximation by the Drude model. The parameters of that model relate the concentration of the charge carriers and their mobility to properties of the dielectric function. After a model parameter fit of the simulated spectrum to measured data the carrier concentration and the mobility or resistivity can be computed [17]. The Drude model relates the macro-scopic susceptibility to the microscopic quantities carrier concentration n and mobility μ;

Ω Ω (6)

andΩ Ω

(7)

where e is the elementary charge , is the permittivity of the free space and m the effective mass of the charge carriers ( )[27] and is the mass of an electron. Resistivity can also be computed using the formula [16];

ΩΩ

ΩΩ

(8)

Table 1 below gives the calculated values of carrier concen-tration (n), mobility (μ), resistivity (ρ) and conductivity (σ) of the undoped and palladium doped films.

500 1000 1500 2000 25001.85

1.90

1.95

2.00

2.05

2.10

2.15

2.20

2.25

n

Wavelength (nm)


500 1000 1500 2000 2500

0.0

0.1

0.2

0.3

0.4

0.5

k

Wavelength (nm)


-1 0 1 2 3 4 5 6 7 83.84

3.86

3.88

3.90

3.92

3.94

3.96

3.98

4.00

Ener

gy g

ap (e

V)Palladium Concentration (at%)

[28] Mohamed HA. Optoelectronics and advanced mate-rials – rapid communications. 3(9) (2009) 936 – 941

[29] Chatterjee K, Chatterjee S, Banerjee A, Raut M, Pal NC, Sen A, Maiti SH. Journal of Materials Chemis-try and Physics. 81(1)(2003) 33-38

[30] Elangovan E., Ramamurthi K. J. Optoelectronics and Advanced Materials. 5(1) (2003) 45-54

85 82

Table 1: Calculated values of carrier concentration (n), mobility (μ), resis-

tivity (ρ) and conductivity (σ) of the undoped tin oxide and palladium

doped tin oxide films

The relatively low resistivity ( Ω cm) observed for the un-doped films is attributed to the deviation from stoichiometry due to oxygen vacancies, which act as elec-tron donors and increase of the free carrier concentration [28] as shown in Figure 7. It can be seen here that initially with doping, electrical resistance decreases due to increase in number of charge carriers. But as we kept on increasing the doping concentration, at a point resistance increases again. This may be attributed to the fact that at higher con-centration, Palladium atoms incorporates at the interstitial sites and crystal structure of the films start to deteriorate, hence decreases the mobility of the free electrons and in-creases the electrical resistivity [18, 29, 30]. The resistivity of Pd doped SnO2 films are higher compared to fluorine and antimony doped tin oxide reported in other peoples’ work.

Figure 7: Resistivity versus Pd Concentration

This is because of the adsorbed oxygen ions on the surface of the Pd doped SnO2 films which inhibit intraband and interband transitions. Therefore, in the presence of a reduc-

ing gas, desorption of the oxygen ions is expected to take place setting free more charge carriers in the film hence increasing film conductivity. The electrical charge concen-tration as a function of palladium concentration is shown in Figure 8. Initially there is a rise in their values up to 3.68at% Pd, beyond which a consistent fall is observed. The changes in carrier mobility are comparatively small as shown in Ta-ble 1. Hence the electrical conductivity seems to be gov-erned by the carrier concentration. The rise in charge con-centration is because of the rise in palladium incorporation at regular and interstitial sites. Further increase in palladium concentration, the interstitial incorporation of palladium must be large enough so that due to the carrier compensa-tion phenomenon the carrier concentration drops. Such a compensation of donors has been observed in the case of excess tin doped indium oxide films and tin incorporation on the growth mechanism of sprayed SnO2 films [3].

Figure 8: Carrier concentration versus Pd/Sn concentration

4. Conclusion

In the present work, thin films of pure tin oxide and palla-dium doped tin oxide are prepared by Spray pyrolysis tech-nique from SnCl4 precursor. The resistivity of the undoped films decreases with initial doping of palladium to attain a minimum value and increases for higher level of doping. The resistivity achieved for the films doped with 3.68at % Pd is the lowest ( Ωcm) for these films from SnCl4 precursor. Initial doping of SnO2 with Palladium leads to the widening of the band gap which decreases on further doping which may be due to the change in structure of the films as the dopant concentration increases. Increase

0 1 2 3 4 5 6-1.0x10-1

0.0

1.0x10-1

2.0x10-1

3.0x10-1

4.0x10-1

5.0x10-1

6.0x10-1

7.0x10-1

8.0x10-1

9.0x10-1

1.0x100

1.1x100

Res

istiv

ity (

Palladium Concentration (at%)

0 2 4 6 8 10 12

-2.0x1022

0.0

2.0x10224.0x10226.0x10228.0x10221.0x10231.2x10231.4x10231.6x10231.8x10232.0x10232.2x1023

Car

rier C

once

ntra

tion

(n)

Pd/Sn Concentration (at%)

in dopant concentration also leads to decrease in transmit-tance in the visible wavelengths with high transmittance of the undoped SnO2 film found to be 88%.

5. Acknowledgement

The authors would like to thank the National Council of Science and Technology, Kenya, for financial support to-wards carrying out this research and Material Science and Solar Energy Network for Eastern and Southern Africa (MSSEESA) for providing a platform for sharing ideas in this work.

References

[1] Young SK, Ansari SG, Ansari ZA, Rizwan W, Hyung-Shik S. American Institute of Physics, Review of Scientific Instruments. 81 (2010) 113903

[2] R. Díaz. Tin oxide thin films: electronic properties and growth mechanism under electrochemical con-trol. University of Barcelona, Barcelona, Spain. (2002).

[3] Chitra A, Takwale MG, Bhide VG, Shailaja M, Kul-karni SK. American Institute of Physics, Journal of Applied Physics. 70 (1991) 7382

[4] Mishra RL, Sheo KM, Prakash SG. Journal of Ovonic Research. 5(4), (2009), 77 - 85

[5] Shamala KS, Murthy LCS, Narasimha KR. Indian Academy of Sciences, Bulletin of Material Science27(3) (2004) 295–301

[6] Shadia JI, Riyad NAB. American Journal of Applied Sciences. 5 (6) (2008) 672-677

[7] Jebbari N, Kamoun NT, Bennaceur R. International Renewable Energy Congress, Sousse, Tunisia. 2010

[8] Agashe C, Marathe RB, Takwale GM, Bhide GV. Thin Solid Films. 164 (1988) 261-264.

[9] Mohammad TM. Solar Energy Materials. 20 (1990)297-305

[10] Rakhshani EA, Makdisi Y, Ramazaniyan A.H. Journal Applied Physics. 83 (1998) 1049-1057.

[11] Stanislav R, Ekaterina R, Tamara S. Sensors & Transducers Magazine. 40 (2) (2004) 145-151.

[12] Sandipan R, Gupta PS, Gurdeep S. Journal of Ovonic Research. 6(1) 2010 63-74

[13] Joong-Ki C, In-Sung H, Sun-Jung K, Joon-Shik P, Soon-Sup P, UnyongJeongc et.al. Sensors and Actu-ators B. 150 (2010) 191–199

[14] Wan ZS, Muhamad MS, Ashkan S, Mohd AY. Sains Malaysiana. 40(3) (2011) 251–257

[15] Boshta M, Mahmud FA, Sayed MH. Journal of Ovonic Research. 6 (2) (2010) 93 - 98

[16] Mageto M, Mwamburi M. Elixir Chemical Physics.53 (2012) 11922-11927

[17] W. Theiss, in: W. Theiss (Ed.), Scout Thin Film Analysis Software Handbook, Hard and Software, Aachen, Germany, 2001 www.mtheiss.com, Pg. 54-57

[18] Yousaf SA, Ali S. Coden. Journal of Natural Sci-ences and Mathematics and Computer, 2009; 48(1 & 2), 43-50

[19] Sankara NS, Santhi B, Sundareswaran S, Venkata-krishnan KS. Metal-Organic, and Nano-Metal Chemistry. 36(1)(2006) 131-135

[20] Ashcroft NW, Mermin ND. Solid State Physics. CBS Publishing, Philadelphia, USA. 1976; Pg. 18

[21] Kim CC, Garland JW, Abad H, Raccah PM. Physical Reviews. B. 45 (1992)11749

[22] Maghanga CM, Jensen J, Niklasson GA, Granqvist CG, Mwamburi M. Solar Energy materials and Solar Cells. 94(1) 2010 75-79

[23] Mageto JM, Maghanga CM, Mwamburi M, Niklasson GA, Granqvist CG. Transparent and Conducting TiO2:Nb Films prepared by spray pyrolysis technique. Unpublished

[24] Brajesh N, Venu GB, Amirthapandian S, Sumita S, Panigrahi BK. et al. Structural and Optical studies of Pd doped tin oxide nanoparticles prepared by chemi-cal co-precipitation method. American Institute of Physics.2012; AIP Conf. Proc. 1447, 439

[25] Modaffer A. Mohammed, Engineering & Technology Journal, 27(6)(2009),

[26] Fatema RC, Shamima C, Firoz H, Tahmina B. Jour-nal of Bangladesh Academy of Sciences. 35(1)(2011) 99-111

[27] Roy GG, Criteria for Choosing Transparent Con-ductors, MRS bulletin/august 2000, www.mrs.org/publications/bulletin

83 84

Influence of Films Thickness on Optical Properties of Nb-Doped TiO2 (NTO) Thin Films Deposited by DC Reactive Magnetron Sputtering

E. R. Ollotu1, M.E.Samiji2, N.R.Mlyuka2, R.T. Kivaisi2

1 Department of Physics, Mkwawa University College of Education, P.O. Box 2513, Iringa, Tanzania2 Departments of Physics, University of Dar es Salaam, P.O. Box 35063, Dar es Salaam, Tanzania.

Abstract Transparent conducting oxide (TCO) thin films, particularly sputter-deposited Sn-doped In2O3 (ITO) have a wide application in optical coatings and optoelectronics. However, due to the scarcity and increasing cost of indium, efforts have been done to search for indium-free materials. Sputter-deposited Nb-doped TiO2 (NTO) is among recent materials which have shown great potential for TCO applications. In this study we report and discuss on the influence of films’ thickness on the optical properties of Nb-doped TiO2 (NTO) films deposited on glass substrate from an alloy target of Ti0.99

Nb0.01 (purity 99.9%). It was observed that luminous transmittance (Tlum), solar transmittance (Tsol) and the band gap energy (Eg) increased with increased films thickness while optical constant (n,k) decreased with films thickness. An average Tlum

and Tsol of above 60% and a relative increase of IR reflectance were achieved. The results implied a possibility for achieving low-emissivity (low-E) material upon improving the IR reflectance by raising the Nb-doping level.

Keywords Magnetron sputtering, Optical properties of semiconductors

1. Introduction Transparent conducting oxide (TCO) thin films have a wide application on optical coatings and transparent electrodes in devices such as flat-panel displays, light-emitting diodes, touch panels, defrosters, and solar cells [1]. In particular, sputter-deposited Sn doped In2O3 (ITO) has been established as a practical TCO material because of its excellent low resistivity ρ (~2×10−4 Ωcm) and visible transmittance (~80–90%) and simple preparation process [2]. However, efforts have been made to find materials that can replace indium because of its [3]. Therefore, the replacing materials must be abundant and non-toxic materials such as zinc, titanium and tin [4].

In search for varieties of materials alternative to ITO, Furubayashi et al. (2005) discovered Nb-doped TiO2 (NTO) as TCO materials [5]. However, to make NTO as a practical TCO, there should be sputtering procedures to fabricate the films. Normally, practical TCO films, including ITO, have been mostly manufactured by sputtering technique, which is characterized by low-cost production on a large-area uniform coating [6]. The general objective of the present work was to establish the sputtering procedures to fabricate NTO thin films for practical low-E on a window glasses. Toward achieving the objective, here we report sputtering procedures of NTO films on a glass slide and on the influence of films thickness on optical properties of the films.

2. Experimental Details Nb-doped TiO2 films were deposited on glass slide (76

mm x 26 mm x1 mm) by Dc magnetron sputtering in the argon-oxygen environment. A Ti0.99 Nb0.01 (purity 99.9 %) alloy disk 2" diameter x 0.250" thick, (supplied by Plasmaterials Inc.) was used as the target. Oxygen (purity 99.999 %) was used as reactive gas and argon (purity 99.999 %) was used as sputtering gas. BALZERS BAE 250 coating unit, was used to deposit the films. The unit chamber of the unit was evacuated in two stages; by rotary pump (DUO 016B) down to 5.0 x10-2 mbar then picked up by air-cooled turbo-molecular pump (TPH33O) to a base pressure of 1.7 x10-6 mbar. Then, heating of the chamber was carried out by means of a radiant heater, whose temperature controlled by a.c. variac transformer with output 0-260 V. The substrate temperature was recorded from digital Multi-meter (FLUKE 87 V) connected to the substrate through locally constructed and calibrated, heat resistant chromel-alumel thermocouple. Figure 1 represents a calibration of a constructed chromel-alumel thermocouple against a standard thermocouple. The constructed chromel-alumel was used to establish equilibrium substrate temperatures for several set points of the transformer with 75.4 mil/min argon. Figure 2 is a plot of equilibrium temperatures of the chamber as a function of several variacset points. The plot enabled to estimate substrate temperature during sputtering, thus replacing a need for using thermocouple in the actual sputtering process. Sputtering power was controlled by W. MDX 1.5 Magnetron drive (750 V/2 A/1.5 KW). The glass slides were ultrasonically cleaned in an Ultrasonic (Decon FS300)

Figure 5. Dependence of 21)( hv on photon energy hv for 200, 250, 300 and 350 nm TNO thin films deposited at 300 oC and 0.04 O2/Ar flow ratio.

3.3 Optical Constants

The optical constants (n,k) of the films were obtained after a model parameter fit of the simulated transmittance spectrum to the measured transmittance spectrums using commercial software SCOUT2. From the fit the complex dielectric permittivity hence, the complex refractive index n ik extracted. Based on the software was defined as [10]:

)4()( 0 DB

Where B is the inter-band dielectric susceptibility defined

as;

dxix

x foB

22

2

2

2

)2

)((exp21 (5)

where o is the resonance frequency, f is the oscillator

strength parameter and is the damping and is the width of distribution

Figure 5 and 6 represent dependence of optical constants (n, k) on the films thickness. Increased films thickness in interval of 50 nm from 200 nm to 350 nm decreased from 2.1 to 1.8 and 0.058 to 0.027 respectively. The decrease was related to improved compactness and uniformity of the films with increasing films thickness

Figure 6. Dependence of refractive index )(n on film’s thickness for 200, 250, 300 and 350 nm TNO films deposited at 300 oC and 0.04 O2/Ar flow ratio.

Figure 7. Dependence of extinction coefficient )(k on films thickness for 200, 250, 300 and 350 nm TNO films deposited at 300 oC and 0.04 O2/Ar ratio

3.4 IR reflectance Figure 8 depict the IR reflectance of the films as a

function of films thickness. The IR reflectance of the TNO thin films was characterized by valleys at around 8 m and 12 m and a peak at around 10 m . According to Tonooka et al., 2009 and Martin-Palma et al., (1998), the peak belongs to radiations of Si – O stretching from the glass. Increased films’ thickness indicated a relative increase of IR reflectance of the films. The observation implied a possibility for achieving low-e material upon improving the IR reflectance by raising the Nb-doping level.

89 86

frequency sweep for 25 minutes and then suspended in ethanol vapor before kept into desiccators ready for use. The target was clean-sputtered for 5 min in order to remove surface oxide layers and contamination and then pre-sputtering for 10 min under the same condition as the film deposition. Films in interval of 50 nm from 200 to 350 nm were fabricated at about 300 oC substrate temperature (Ts), 6.0 x 10-3 mbar working pressure and 200 W sputtering power and 0.04 O2/Ar flow ratio.

Figure 1. Heating and cooling curves for both locally constructed and a standard FLUKE 87 V cable chromel-alumel thermocouples.

Figure 2. Substrate equilibrium temperatures as a function of a.c. variac transformer set points.

Films’ thickness was determined using Alfa step IQ based on mechanical stylus. Films’ UV/VIS/NIS transmittance was determined using SHIMADZU SolidSpec-3700 DUV spectrophotometer (250-2500 nm) and IR reflectance using PERKIN ELMER spectrum BX FT-IR spectrophotometer in the range (2500-50,000 nm). Surface morphology was determined using Atomic Force Microscope (AFM) and films’ band gap energy (Eg) was estimated from Tauc plots developed from MATLAB 7.1. The optical constants (n,k) were worked out from commercial software SCOUT2.


3.1 Luminous and Solar Transmittance

Luminous transmittance (Tlum) and solar transmittance (Tsol) as functions of films thickness is represented in Figure 3. The Tlum (400 - 760 nm) and Tsol (300 - 2500) nm were computed based on the following respective formula [7]:

)1()(

)()(76.0

4.0

76.0

4.0

m

m

m

mlum

dQ

dTQT

and

)2()(

)()(5.2

3.0

5.2

3.0

m

m

m

msol

dG

dTGT

Where T() is transmittance in a given wavelength range,G() is Air Mass 1.5 solar irradiance ISO 9845 and G())is Air Mass 1.5 illuminant ISO/CIE 10526 [8]. The Integrated Tsol and Tlum are the calculated area under the curve and normalized using standard data that is ISO 9845 and ISO/CIE 10526 respectively.

Figure 3. Tsol and Tlum on films thickness for 200, 250, 300 and 350 nm TNO films deposited at 300 oC and 0.04 O2/Ar flow ratio.

Increased films thicknesses in interval of 50 nm from 200 to 350 nm improved Tsol and Tlum from 56 to 61 % and Tlum

from 59 to 65 % respectively as shown in Figure 3. The increased Tsol and Tlum were attributed to improved films formation with increased films thickness as suggested by the AFM surface morphology images in Figure 4. The images reveled that, the films developed from deprived grains shape for films of 200 nm (Figure 4.a) to a fairly developed grains with existing islands for films of 250 nm (Figure 4.b) into a more defined mix of small and large grain shape for films of 300 nm (Figure 4.c). The morphological improvement of the films with increased films thickness likely reduced films absorption as a result improved both Tsol and Tlum. (a)

(b)

(c)

Figure 4. AFM images for (a) 250 nm (b) 300 nm and (c) 350 nm TNO films deposited at 300 oC and 0.04 O2/Ar flow ratio.

3.2 Optical Band Gap Energy

Figure 5 represents dependence of optical band gap energy (Eg) films thickness. The Eg was computed using MATLAB m-file program based on the Tauc plot governed by the equation [9];

)3()( mg hvEAhv

where = 2k/ is the absorption coefficient, hv is the photon energy, m=2 account for indirect allowed transition across the allowed band gap and A depends on transition probability which can be taken as constant within the optical frequency range. Increased films thickness in interval of 50 nm from 200 to 350 nm increased band gap energy (Eg) from 3.35 to 3.40 eV. The increase (Eg ) was associated with improved films’ structure with increased films thickness.

2.2. Deposition of lead hydroxy sulphide

(Pb(OH)xSy) coating

Lead hydroxy sulphide (Pb(OH)xSy) coating was deposited by successive ion layer adsorption and reaction (SILAR). In this process, saturated lead acetate (Pb(CH3OOH)2) solution (97% pure) and sodium sulfide (Na2S) solution (97% pure) both commercially obtained from Sigma Aldrich Ltd. were used. To deposit lead hydroxy sulphide (Pb(OH)xSy) coating, the following four-stage process were followed; (a) the samples were submerged in lead acetate solution for 10 seconds for the metal ions to be adsorbed, (b) the samples were removed and then rinsed in distilled water, (c) the samples were submerged in sodium sulfide solution for 10 seconds to allow sulphur ions to react with metal ions, (d) finally the samples were rinsed in distilled water. These four steps constituted one cycle giving a particular Pb(OH)xSy deposit thickness. After all the layers (the TiO2 thin film window layer and doped with nitrogen, indium hydroxy sulphide buffer layer, lead hydroxy sulphide absorber layer and the silver paste back contact) were deposited, the schematic cross-section of the TiO2/In(OH)iSj/Pb(OH)xSy/Ag ETA solar cell appears as shown in Figure 1.

Figure 1. A schematic cross - section of the N:TiO2/In(OH)iSj/Pb(OH)xSycomposite (ETA) solar cell.

2.3. I-V Characterization

Current voltage (I-V) characterization was done using a 0.5kW Xenon arc lamp model 66902 Oriel instruments Ltd. solar simulator. The source of illumination simulated one sun obtained by use of AM 1.5 Global air mass filter. The cell was powered from the DC source meter (Keithly meter-2400).Values useful for I-V curve plot were keyed in a computerized system on a Visual Basic program platform. These values include start voltage (0 V), end voltage (5 V), cell area (0.1256 cm2), the number of data points (100) and the value of radiation (1000 W/m2).Characteristic I-V curves were plotted using Visual Basic software and relevant performance parameters such as open circuit voltage, Voc, and short circuit current density, Jsc, deduced from the curves.

3.0 Results and Discussion

3.1. Structural Analysis

Figure 2 shows a SEM micrograph of Nitrogen-doped TiO2window layer deposited by spray pyrolysis with indium hydroxy sulphide (Pb(OH)xSy) deposited on top of porous nitrogen-doped TiO2 window layer. Titanium dioxide deposited by spray pyrolysis method has a characteristic porous film towards the surface from the base of the substrate (FTO). The film is characterized by nano-scale particles and nano-scale pores of approximately 500 nm. This morphology is favourable for light scattering effect and increases the light optical path, thus enhancing light absorption which makes the films appropriate for solar cell applications.

Figure 2. SEM micrograph of nitrogen-doped TiO2 (N:TiO2) coating on FTO glass substrates and annealed at 400 oC with indium hydroxy sulphide (In(OH)iSj) deposited on top of porous nitrogen-doped TiO2.

3.2. Optical Characterization

3.2.1 Reflectance and Transmittance

Figure 3 shows absorption coefficient, α, of the various samples of nitrogen-doped and undoped TiO2 thin films, both annealed and unannealed, and it is observed that at all wavelengths the absorption coefficient, α, is higher for the nitrogen-doped films compared to the undoped films. This means that within the wavelength range of 300 – 1200 nm considered here there is enhanced absorption. Figure 4 shows transmittance and reflectance characteristics for undoped and nitrogen-doped samples both annealed and unannealed. Absorption, reflectance and transmittance should all add up to unity. Notable from Figure 4 is that the transmittance of the undoped films is higher than for the doped for all wavelengths, but reflectance is fairly constant.

This implies that only transmittance and absorptance of the film vary. Therefore, if transmittance is high then absorption is reduced and vice versa and we can infer from both figures that doping increases absorptance.

The graphs in Figure 3 and Figure 4 shift towards the right, that is, from the ultra violet region to the visible region on

doping the films whether annealed or unannealed. This is indicative of an improvement in photo-absorption. The improvement is attributed to the reduction of band gap upon nitrogen doping.

Considering the graphs for the doped samples in Figure 3 it isnoted that absorption coefficients are higher for the doped TiO2 thin films within the visible part of the spectrum. The

Silver paste back contact

FTO glass substrate

In(OH)iSj layerPb(OH)xSy layer

N:TiO2 coating

TiO2 matrix

In(OH)iSj

Nano-pores

87 88

Figure 8. Influence of films thickness on IR reflectance for 200, 250, 300 and 350 nm TNO films deposited at 300 oC and 0.04 O2/Ar ratio

4. Conclusion Thin films of thickness in interval of 50 nm from 200 nm to 350 nm were prepared from alloy target of Ti0.99

Nb0.01 by DC reactive magnetron sputtering at at 300 oC and 0.04 O2/Ar ratio. The influence of films thickness on the optical and IR reflectance was investigated. Analysis indicated an increase of both Tlum and Tsol and band gap energy (Eg), and a decrease of refractive index (n) with an increased films thickness. An average of above 60 % Tlum

about 70% and E of 87 % were achieved. The results implied a possibility for Low-E with the material upon lowering the emissivity by raising the Nb-doping level

ACKNOWLEDGEMENT The author acknowledges the Belgian Technical Cooperation (BTC) for the scholarship and Material Science and Solar Energy Network for Eastern and Southern Africa (MSSEESA) for facilitating travel to Nairobi and Moi University in Kenya for optical characterization.

REFERENCES

[1] K.L. Chopra, S. Major, D.K. Pandya, “Transparent conductors- a status review,” Thin Solid Films 102 (1983) 1.

[2] I. Hamberg, C. Granqvist, “Evaporated Sn doped In2O3 films: Basic optical properties and applications to energy efficient windows,” J. Appl. Phys. 60 (1986) R123.

[3] Y.Sato, H.Akizuki, T. Kamiyama, Y.Shigesato,“Transparent conductive Nb-doped TiO2 films deposited by direct-current magnetron sputtering using a TiO2 − x target,” Thin Solid Films, 516 (2008) 5758–5762.

[4] K. Tonooka, T. Chiu and N. Kikuchi, “Preparation of transparent conductive TiO2: Nb thin films by pulsed laser deposition” Applied Surface Science, 255 (2009), 9695–9698.

[5] Y. Furubayashi, T. Hitosugi, Y. Yamamoto, K. Inaba, G. Kinoda, Y. Hirose, T.Shimada and T. Hasegawa,” A transparent metal: Nb-doped anatase TiO2,” APPL. PHY. LETTERS 86, (2005) 252101.

[6] N. Yamada, T. Hitosugi, N. L.H. Hoang, Y. Furubayashi, Y.Hirose, S. Konuma, T. Shimada, and T. Hasegawa, “Structural, electrical and optical properties of sputter-deposited Nb-doped TiO2 (TNO) polycrystalline films”Thin Solid Films, 516 (2008), 5754–5757

[7] C.M. Maghanga, J.Jensen, G.A.Niklasson , C.G.Granqvist, M.Mwamburi, “Optical modeling of spectrally selective reflectors based on TiO2:Nb transparent conducting oxide films for silicon solar cell applications,” Sol.Energy Mater. Sol. Cells (2009), doi:10.1016/j.solmat.2009.02.023

[8] American Society for Testing and Materials (1992),Reference Solar Spectral Irradiance: Air Mass 1.5 (online) available: http://rredc.nrel.gov/solar/spectra/am1.5S

[9] C.G. Granqvist, Handbook of Inorganic Electrochromic Materials, Elsevier, Amsterdam, The Netherlands, 1995, pp. 265–275.

[10] C.M. Maghanga, J. Jensen, G.A. Niklasson, C.G. Granqvist, and M. Mwamburi, Sol. Energy Mater. Sol. Cells, 94(2009), 1.

[11] C.G. Granqvist, “Transparent conductors as solar energy materials: A panoramic review” Solar Energy Materials & Solar Cells 91 (2007) 1529–1598

[12] http://www.wtheiss.com/docs/sc2_tu1/index.

Improved Performance of TiO2/In(OH)iSj/Pb(OH)xSy Composite ETA Solar Cell through Nitrogen doped TiO2 Thin Film

window layer.C. O. Ayieko*1, R. J. Musembi1, S. M. Waita1, B. O. Aduda1 and P. K. Jain2

1 Department of Physics, University of Nairobi, P.O. Box 00100-30197, Nairobi, Kenya. 2 Department of Physics, University of Botswana, P/ Bag 0022, Gaborone, Botswana.

Abstract

TiO2 thin films were doped with nitrogen (N2) gas and the presence of nitrogen atoms was confirmed in the films by Energy Dispersive X-Ray (EDX) spectrum analysis. Undoped TiO2 thin films had an energy band gap of 3.25 eV while the doped films had 2.90 eV. The reduction in energy band gap was attributed to the introduction of N2 impurity states on the bands (conduction band and or valence). Effect of doping titanium dioxide window layer on the efficiency of the extremely thin absorber (ETA) TiO2/In(OH)iSj/Pb(OH)xSy solar cell was investigated using conventional current-voltage (I-V) technique. The photovoltaic conversion efficiency (η) was 1.06% for the solar cell with undoped films and 1.32% for the solar cell with doped films. The increase in photovoltaic conversion efficiency on doping was attributed to decrease in band gap due to nitrogen doping.

* Corresponding author: [email protected] (C. O. Ayieko)

Keywords Titanium dioxide, Conversion efficiency, Nitrogen doping.

1.0 Introduction

Meeting human energy requirements has been an elusive task even after the discovery of the immense energy from the atom when the phrase “too cheap to meter” was coined [1]. The energy cost has been soaring each passing day. The energy crisis in the 1970’s which saw the cost of petroleum almost quadruple necessitated intensive research and development on renewable energy sources [1] to attain higher conversion efficiency and cost reduction.

Of the renewable energy sources, photovoltaic conversion of solar energy is environmentally friendly since it has no greenhouse gas emission during the harnessing process and

even at the utilization stage. The other advantage solar energy has over other renewable energy sources is its universality and ability to be made available even to the remotest parts of the world. Solar energy systems do not involve moving parts and hence present minimal wear and tear and therefore require very low maintenance once installed. The initial cost of installation for solar energy systems is high just as for other energy sources but due to low maintenance costs, it guarantees free energy after a short payback period. This makes solar energy systems cost effective and economical [2].

Improving the efficiency of solar cells which are under laboratory test has been a major concern for photovoltaic applications. A number of factors have been considered that may lead to improved conversion efficiency of a solar cell device. These include: (i) use of suitable front and back contacts [3], (ii) use of tunneling insulation layer for improved charge collection [4], (iii) use of extremely thin layer as absorber material of the solar cell device and suppression of recombination of the charge carriers [5] and (iv) use of blends especially in organic solar cell [6].

The concept of extremely thin absorber (ETA) realized through reduced absorber thickness reduces transport path for

the carrier charges resulting in higher probability of charge carrier collection. The highly porous window layer used in fabrication of solar cells based on ETA concept enhances scattering at the interpenetrating surfaces resulting in increased optical path for radiation thereby increasing absorption [5].

Titanium dioxide has been used as window layer for both all-solid state and dye sensitized solar cells (DSSC). A porous TiO2 deposited by spray pyrolysis was used as a window layer in the fabrication of TiO2/In(OH)iSj/Pb(OH)xSy/PEDOT:PSS ETA solar cell and observation that the highly porous titanium dioxide ensured

2.2. Deposition of lead hydroxy sulphide

(Pb(OH)xSy) coating

Lead hydroxy sulphide (Pb(OH)xSy) coating was deposited by successive ion layer adsorption and reaction (SILAR). In this process, saturated lead acetate (Pb(CH3OOH)2) solution (97% pure) and sodium sulfide (Na2S) solution (97% pure) both commercially obtained from Sigma Aldrich Ltd. were used. To deposit lead hydroxy sulphide (Pb(OH)xSy) coating, the following four-stage process were followed; (a) the samples were submerged in lead acetate solution for 10 seconds for the metal ions to be adsorbed, (b) the samples were removed and then rinsed in distilled water, (c) the samples were submerged in sodium sulfide solution for 10 seconds to allow sulphur ions to react with metal ions, (d) finally the samples were rinsed in distilled water. These four steps constituted one cycle giving a particular Pb(OH)xSy deposit thickness. After all the layers (the TiO2 thin film window layer and doped with nitrogen, indium hydroxy sulphide buffer layer, lead hydroxy sulphide absorber layer and the silver paste back contact) were deposited, the schematic cross-section of the TiO2/In(OH)iSj/Pb(OH)xSy/Ag ETA solar cell appears as shown in Figure 1.

Figure 1. A schematic cross - section of the N:TiO2/In(OH)iSj/Pb(OH)xSycomposite (ETA) solar cell.

2.3. I-V Characterization

Current voltage (I-V) characterization was done using a 0.5kW Xenon arc lamp model 66902 Oriel instruments Ltd. solar simulator. The source of illumination simulated one sun obtained by use of AM 1.5 Global air mass filter. The cell was powered from the DC source meter (Keithly meter-2400).Values useful for I-V curve plot were keyed in a computerized system on a Visual Basic program platform. These values include start voltage (0 V), end voltage (5 V), cell area (0.1256 cm2), the number of data points (100) and the value of radiation (1000 W/m2).Characteristic I-V curves were plotted using Visual Basic software and relevant performance parameters such as open circuit voltage, Voc, and short circuit current density, Jsc, deduced from the curves.

3.0 Results and Discussion

3.1. Structural Analysis

Figure 2 shows a SEM micrograph of Nitrogen-doped TiO2window layer deposited by spray pyrolysis with indium hydroxy sulphide (Pb(OH)xSy) deposited on top of porous nitrogen-doped TiO2 window layer. Titanium dioxide deposited by spray pyrolysis method has a characteristic porous film towards the surface from the base of the substrate (FTO). The film is characterized by nano-scale particles and nano-scale pores of approximately 500 nm. This morphology is favourable for light scattering effect and increases the light optical path, thus enhancing light absorption which makes the films appropriate for solar cell applications.

Figure 2. SEM micrograph of nitrogen-doped TiO2 (N:TiO2) coating on FTO glass substrates and annealed at 400 oC with indium hydroxy sulphide (In(OH)iSj) deposited on top of porous nitrogen-doped TiO2.


3.2.1 Reflectance and Transmittance

Figure 3 shows absorption coefficient, α, of the various samples of nitrogen-doped and undoped TiO2 thin films, both annealed and unannealed, and it is observed that at all wavelengths the absorption coefficient, α, is higher for the nitrogen-doped films compared to the undoped films. This means that within the wavelength range of 300 – 1200 nm considered here there is enhanced absorption. Figure 4 shows transmittance and reflectance characteristics for undoped and nitrogen-doped samples both annealed and unannealed. Absorption, reflectance and transmittance should all add up to unity. Notable from Figure 4 is that the transmittance of the undoped films is higher than for the doped for all wavelengths, but reflectance is fairly constant.

This implies that only transmittance and absorptance of the film vary. Therefore, if transmittance is high then absorption is reduced and vice versa and we can infer from both figures that doping increases absorptance.

The graphs in Figure 3 and Figure 4 shift towards the right, that is, from the ultra violet region to the visible region on

doping the films whether annealed or unannealed. This is indicative of an improvement in photo-absorption. The improvement is attributed to the reduction of band gap upon nitrogen doping.

Considering the graphs for the doped samples in Figure 3 it isnoted that absorption coefficients are higher for the doped TiO2 thin films within the visible part of the spectrum. The

Silver paste back contact

FTO glass substrate

In(OH)iSj layerPb(OH)xSy layer

N:TiO2 coating

TiO2 matrix

In(OH)iSj

Nano-pores

93 90

Improved Performance of TiO2/In(OH)iSj/Pb(OH)xSy Composite ETA Solar Cell through Nitrogen doped TiO2 Thin Film

window layer.C. O. Ayieko*1, R. J. Musembi1, S. M. Waita1, B. O. Aduda1 and P. K. Jain2

1 Department of Physics, University of Nairobi, P.O. Box 00100-30197, Nairobi, Kenya. 2 Department of Physics, University of Botswana, P/ Bag 0022, Gaborone, Botswana.

Abstract

TiO2 thin films were doped with nitrogen (N2) gas and the presence of nitrogen atoms was confirmed in the films by Energy Dispersive X-Ray (EDX) spectrum analysis. Undoped TiO2 thin films had an energy band gap of 3.25 eV while the doped films had 2.90 eV. The reduction in energy band gap was attributed to the introduction of N2 impurity states on the bands (conduction band and or valence). Effect of doping titanium dioxide window layer on the efficiency of the extremely thin absorber (ETA) TiO2/In(OH)iSj/Pb(OH)xSy solar cell was investigated using conventional current-voltage (I-V) technique. The photovoltaic conversion efficiency (η) was 1.06% for the solar cell with undoped films and 1.32% for the solar cell with doped films. The increase in photovoltaic conversion efficiency on doping was attributed to decrease in band gap due to nitrogen doping.

* Corresponding author: [email protected] (C. O. Ayieko)

Keywords Titanium dioxide, Conversion efficiency, Nitrogen doping.

1.0 Introduction

Meeting human energy requirements has been an elusive task even after the discovery of the immense energy from the atom when the phrase “too cheap to meter” was coined [1]. The energy cost has been soaring each passing day. The energy crisis in the 1970’s which saw the cost of petroleum almost quadruple necessitated intensive research and development on renewable energy sources [1] to attain higher conversion efficiency and cost reduction.

Of the renewable energy sources, photovoltaic conversion of solar energy is environmentally friendly since it has no greenhouse gas emission during the harnessing process and

even at the utilization stage. The other advantage solar energy has over other renewable energy sources is its universality and ability to be made available even to the remotest parts of the world. Solar energy systems do not involve moving parts and hence present minimal wear and tear and therefore require very low maintenance once installed. The initial cost of installation for solar energy systems is high just as for other energy sources but due to low maintenance costs, it guarantees free energy after a short payback period. This makes solar energy systems cost effective and economical [2].

Improving the efficiency of solar cells which are under laboratory test has been a major concern for photovoltaic applications. A number of factors have been considered that may lead to improved conversion efficiency of a solar cell device. These include: (i) use of suitable front and back contacts [3], (ii) use of tunneling insulation layer for improved charge collection [4], (iii) use of extremely thin layer as absorber material of the solar cell device and suppression of recombination of the charge carriers [5] and (iv) use of blends especially in organic solar cell [6].

The concept of extremely thin absorber (ETA) realized through reduced absorber thickness reduces transport path for

the carrier charges resulting in higher probability of charge carrier collection. The highly porous window layer used in fabrication of solar cells based on ETA concept enhances scattering at the interpenetrating surfaces resulting in increased optical path for radiation thereby increasing absorption [5].

Titanium dioxide has been used as window layer for both all-solid state and dye sensitized solar cells (DSSC). A porous TiO2 deposited by spray pyrolysis was used as a window layer in the fabrication of TiO2/In(OH)iSj/Pb(OH)xSy/PEDOT:PSS ETA solar cell and observation that the highly porous titanium dioxide ensured

increased absorption of incident radiation through scattering [7]. Indium hydroxyl sulphide (In(OH)iSj) was used as a layer to suppress charge recombination.

The effect of nitration of pressed TiO2 was investigated and the nitrated TiO2 used as window layer in DSSC, however the challenge encountered with DSCC was leakages. This was because not all the components of the cell were solid [8]. The technique of nitrogen doping was applied on titanium dioxide and the band gap was reported to have been tuned to respond to visible spectrum [9]. The change in titanium dioxide band gap implied that its optical properties had changed.

In this work we shift from doping TiO2 by pressing method used by [8] and adopt a heat treatment method suitable for doping TiO2 coated by an upscalable spray pyrolysis technique. We incline to heat treatment method so as to advance and exploit the observation by [9] that doping TiO2shifts absorption to the visible range to the solar cell as he also used the heat treatment method. We choose to use nitrogen doped TiO2 as a window layer on all-solid state N:TiO2/In(OH)iSj/Pb(OH)xSy ETA solar cell to improve photo response as opposed to the use of nitrogen-doped TiO2as window layer in DSSC which posed the challenge of sealing. With this concept of doping, it is anticipated that there would be improved photovoltaic conversion efficiency of all-solid state ETA solar cell.

2.0 Experimental

Fluorine-doped tin oxide coated glass substrates (FTO) were used and were cleaned in an ultrasonic bath filled with distilled water for 2 minutes. The substrates were then kept in acetone for 20 minutes and finally rinsed in distilled water.

For the deposition of TiO2 by spray pyrolysis, 120 µl ofcommercially available titanium (iv) isopropoxide (TiC12H28O4) (99.7% pure) and 200 ml isopropanol (C3H8O)(99.7% pure) from Fluka Ltd were mixed in a 500 ml beaker. The mixture was maintained at a temperature of 50 oC using an electric hot plate while stirring with a magnetic stirrer for about 15 minutes. This mixture formed the precursor which was atomized by the spray system using compressed nitrogen as the carrier gas.

The spray nozzle-to-substrate distance was set to about 15.0cm. The substrate temperature was set to 200 oC, 150 oC and 100 oC for different samples respectively. Spraying was done in pulses. A pulse consisted of 5 seconds of spraying and 30 seconds of pause. Ten pulses were done for every sample at a precursor flow rate of 2.6 ml/min. The chemical reaction resulting in the formation of amorphous TiO2 thin film was as follows;

TiC12H28O4 (aq) + 17O2 (g) + 2C3H8O (aq) Ts (100 oC- 200 oC)

TiO2(s)(amorph) + 8CO2(g) + 22H2O(g).................................(1)

TiO2 (s) (amorph) annealing (30 min) at 400 oC TiO2 (s)

(polycrystalline)................................................................... (2)

The film thicknesses were determined using a computerized KLA-Tencor Alpha –Step IQ surface profiler.

For nitrogen doping, a pyrex glass tube placed in a programmable horizontal tube furnace (ThermoScientific Lindberg/Blue Mini-Mite) was used for heating the samples in nitrogen gas atmosphere. The tube furnace was fitted with a digital themometer which was used to measure the temperature of the samples’ chamber. Samples were loaded into the chamber using a metallic sample holder and heated gradually from room temperature upto a fixed temperature of 450 oC. Nitrogen gas flow into the chamber was maintained at the rate of 20 cm3/sec via Labview Inc. Computer software that interfaced with digital flowmeter connected to a nitrogen gas cylinder. After 30 minutes the samples were allowed to cool to room temperature without cutting off nitrogen flow.The structure of the films was characterized by X-ray diffraction (XRD) using a Phillips, PW 3710 XRD interfaced with computer installed with Analytical X’Pert 2000 data collection software. EDX studies were carried out using scanning electron microscope (SEM) model Phillips, XL 30 ESEM, integrated with energy dispersive X-ray (EDX) unit.The SEM micrographs for the study of surface morphology were obtained by examination using Carl ZeissTM LEO 1530 SEM model.

Total transmittance and reflectance measurements were taken at near normal incidence in the wavelength range of 300-1200 nm using a computerized double beam solid-spec 3700 DUV Shimadzu Spectrophotometer equipped with 198851 Barium Sulphate (BaSO4) integrating sphere. The reflectance and transmittance data were used for calculation of optical properties such as band gap, Urbach energy and refractive index.

2.1. Deposition of indium hydroxy sulphide

(In(OH)iSj) buffer layer

The nitrogen-doped and undoped TiO2 thin films were coated with indium hydroxy sulphide buffer layer by chemical bath deposition (CBD) in a solution prepared using commercially obtained reagents from Sigma Aldrich Ltd. The following procedure was used to make the solution: 20000 ± 1µl of distilled water was mixed with 1250 ± 1µl of 0.005M of hydrochloric acid (HCL) which was then mixed with 1250 ± 1µl of 0.005M indium (III) chloride (In2Cl3) and 2500 ± 1µl of 0.1M thioacetamide (CH3CSNH2).The resulting solution which was in a beaker was immersed into a water bath maintained at a temperature of 70 ± 1 oC.

The thin film samples were dipped three times for 30 minutes each into the same solution. Each dipping was followed by rinsing with distilled water to clean off unwanted chemical remnants that could dry on the coating after removal. The dipping, withdrawal and rinsing procedure was repeated three times for each sample. The samples were finally annealed at a temperature of 300 oC in air.

REFERENCES

[1] Quashing, V. (2005). Understanding renewable systems, Earthscan, London, 1-130.

[2] Hankins, M. (1995), Solar electric systems for Africa,Motif Creative Arts, Ltd. Nairobi, Kenya.

[3] Zeman, M., R. VanSwaaij, M. Zuiddam, J. W. Metselaar , R.E.I Schropp, Solar Energy Materials and Solar Cells66, (1-4) 353-359 (2001).

[4] Konig, D. and G.Ebest, Solar Energy Materials and Solar Cells 75 (2003), (3-4) 335-343.

[5] Nanu, M., J. Schooman, A. Goosens, Advanced Materials16(2004), (5) 453-455

[6] Bayon, R., R. Musembi, A. Belaidi, M. Bar, T. Guminskaya, M. Lux-Steiner, Th. Dittrich, Solar Energy Materials and Solar Cells, 89(2005), 18-19.

[7] Musembi, R. (2009), Fabrication and characterization of In(OH)iSj modified highly structured TiO2/ Pb(OH)xSy/PEDOT :PSS eta Solar cell, and study of its transport mechanisms, PhD Thesis, Department of Physics, University of Nairobi, 88-91.

[8] Wafula, B., J. Simiyu, S. Waita, B. Aduda, J. Mwabora Effect of nitration on pressed TiO2 Photoelectrodes for Dye-Sensitized solar cells, African Journal of Science and Technology (AJST), Science and Engineering Series8(2007), (2) 63-71

[9] Gartner, M., P. Osiceanu, M. Anastasescu, T. Stoica, T.F. Stoica, C. Trapalis, T. Giannakopoulou, N. Todoroa, A., Thin solid Films, 516(2008), 8184-8189

[10] Anderson, D. (2001).Clean Electricity from Photovoltaics,M.D Archer and R.D. Hill Edition, Imperial College Press: London.

[11] Baoshun, L., W. Liping, Z. Xiujian, Solar Energy Materials and Solar Cells, 92(2007), 1-10.

91 92

increased absorption of incident radiation through scattering [7]. Indium hydroxyl sulphide (In(OH)iSj) was used as a layer to suppress charge recombination.

The effect of nitration of pressed TiO2 was investigated and the nitrated TiO2 used as window layer in DSSC, however the challenge encountered with DSCC was leakages. This was because not all the components of the cell were solid [8]. The technique of nitrogen doping was applied on titanium dioxide and the band gap was reported to have been tuned to respond to visible spectrum [9]. The change in titanium dioxide band gap implied that its optical properties had changed.

In this work we shift from doping TiO2 by pressing method used by [8] and adopt a heat treatment method suitable for doping TiO2 coated by an upscalable spray pyrolysis technique. We incline to heat treatment method so as to advance and exploit the observation by [9] that doping TiO2shifts absorption to the visible range to the solar cell as he also used the heat treatment method. We choose to use nitrogen doped TiO2 as a window layer on all-solid state N:TiO2/In(OH)iSj/Pb(OH)xSy ETA solar cell to improve photo response as opposed to the use of nitrogen-doped TiO2as window layer in DSSC which posed the challenge of sealing. With this concept of doping, it is anticipated that there would be improved photovoltaic conversion efficiency of all-solid state ETA solar cell.

2.0 Experimental

Fluorine-doped tin oxide coated glass substrates (FTO) were used and were cleaned in an ultrasonic bath filled with distilled water for 2 minutes. The substrates were then kept in acetone for 20 minutes and finally rinsed in distilled water.

For the deposition of TiO2 by spray pyrolysis, 120 µl ofcommercially available titanium (iv) isopropoxide (TiC12H28O4) (99.7% pure) and 200 ml isopropanol (C3H8O)(99.7% pure) from Fluka Ltd were mixed in a 500 ml beaker. The mixture was maintained at a temperature of 50 oC using an electric hot plate while stirring with a magnetic stirrer for about 15 minutes. This mixture formed the precursor which was atomized by the spray system using compressed nitrogen as the carrier gas.

The spray nozzle-to-substrate distance was set to about 15.0cm. The substrate temperature was set to 200 oC, 150 oC and 100 oC for different samples respectively. Spraying was done in pulses. A pulse consisted of 5 seconds of spraying and 30 seconds of pause. Ten pulses were done for every sample at a precursor flow rate of 2.6 ml/min. The chemical reaction resulting in the formation of amorphous TiO2 thin film was as follows;

TiC12H28O4 (aq) + 17O2 (g) + 2C3H8O (aq) Ts (100 oC- 200 oC)

TiO2(s)(amorph) + 8CO2(g) + 22H2O(g).................................(1)

TiO2 (s) (amorph) annealing (30 min) at 400 oC TiO2 (s)

(polycrystalline)................................................................... (2)

The film thicknesses were determined using a computerized KLA-Tencor Alpha –Step IQ surface profiler.

For nitrogen doping, a pyrex glass tube placed in a programmable horizontal tube furnace (ThermoScientific Lindberg/Blue Mini-Mite) was used for heating the samples in nitrogen gas atmosphere. The tube furnace was fitted with a digital themometer which was used to measure the temperature of the samples’ chamber. Samples were loaded into the chamber using a metallic sample holder and heated gradually from room temperature upto a fixed temperature of 450 oC. Nitrogen gas flow into the chamber was maintained at the rate of 20 cm3/sec via Labview Inc. Computer software that interfaced with digital flowmeter connected to a nitrogen gas cylinder. After 30 minutes the samples were allowed to cool to room temperature without cutting off nitrogen flow.The structure of the films was characterized by X-ray diffraction (XRD) using a Phillips, PW 3710 XRD interfaced with computer installed with Analytical X’Pert 2000 data collection software. EDX studies were carried out using scanning electron microscope (SEM) model Phillips, XL 30 ESEM, integrated with energy dispersive X-ray (EDX) unit.The SEM micrographs for the study of surface morphology were obtained by examination using Carl ZeissTM LEO 1530 SEM model.

Total transmittance and reflectance measurements were taken at near normal incidence in the wavelength range of 300-1200 nm using a computerized double beam solid-spec 3700 DUV Shimadzu Spectrophotometer equipped with 198851 Barium Sulphate (BaSO4) integrating sphere. The reflectance and transmittance data were used for calculation of optical properties such as band gap, Urbach energy and refractive index.

2.1. Deposition of indium hydroxy sulphide

(In(OH)iSj) buffer layer

The nitrogen-doped and undoped TiO2 thin films were coated with indium hydroxy sulphide buffer layer by chemical bath deposition (CBD) in a solution prepared using commercially obtained reagents from Sigma Aldrich Ltd. The following procedure was used to make the solution: 20000 ± 1µl of distilled water was mixed with 1250 ± 1µl of 0.005M of hydrochloric acid (HCL) which was then mixed with 1250 ± 1µl of 0.005M indium (III) chloride (In2Cl3) and 2500 ± 1µl of 0.1M thioacetamide (CH3CSNH2).The resulting solution which was in a beaker was immersed into a water bath maintained at a temperature of 70 ± 1 oC.

The thin film samples were dipped three times for 30 minutes each into the same solution. Each dipping was followed by rinsing with distilled water to clean off unwanted chemical remnants that could dry on the coating after removal. The dipping, withdrawal and rinsing procedure was repeated three times for each sample. The samples were finally annealed at a temperature of 300 oC in air.

enhanced absorption is due to nitrogen doping which increases the absorption coefficient of the TiO2 thin film.

300 400 500 600 700 800 900 1000 1100 1200 13000.0

5.0x106

1.0x107

1.5x107

2.0x107

Abso

rptio

n C

oeffi

cien

t (

)(m

-1)

Wavelength (nm)

undoped and annealed, 400 oC doped and unannealed undoped and unannealed doped and annealed,400 oC

Figure 3. Dependence of absorption co-efficient on wavelength of nitrogendoped and undoped TiO2 thin films of approximate thickness of 400 nm deposited on fluorine doped tin oxide (FTO) glass slides.

From Figure 4 reflectance curves cluster in a narrow range for the wavelength range considered here as opposed to transmittance that varies greatly with the sample treatment. The nitrogen-doped sample records the lowest transmittance across the considered spectrum. This shows that the nitrogen-doped sample has high absorption for the wavelength range presented.

300 400 500 600 700 800 900 1000 1100 1200 13000

10

20

30

40

50

60

70

80

Tran

smitt

ance

(%)

Wavelength (nm)

undoped and annealed, 400 oC doped and unannealed undoped and unannealed doped and annealed 400 oC

Ref

lect

ance

(%)

10% transmittance line

Figure 4. Dependence reflectance and transmittance on wavelength of nitrogen doped and undoped TiO2 thin films of approximate thickness of 400 nm deposited on fluorine doped tin oxide (FTO) glass slides.

To establish the part of the spectrum where enhanced photoresponse is evident due to nitrogen doping, equation (3)is used [10]:

mE

24.1

................................................................. (3)

where the Eλ is energy in eV and λ is wavelength in micrometers (µm). The equation gives wavelengths which correspond to a minimum of 428 nm for 2.90 eV and a maximum of 384 nm for 3.25 eV that is also confirmed by the observation of band gap reduction due to doping in Section 3.2.2. This is very remarkable since nitrogen doping enhances light absorption beyond the visible to near infra-red.

3.2.2. Effect of Nitrogen-doping on the band gap of TiO2

coated on fluorine doped tin oxide (FTO) glass slides

Figure 5 presents the band gaps of films coated on fluorine doped tin oxide (FTO) substrate and annealed at 400 oC. A significant decrease in the band gap is observed, the undoped film had a band gap of 3.25 eV compared to the nitrogen-doped which had 2.90 eV. This observation concurs with

results earlier obtained by Baoshun [11] who also observed a significant decrease in the band gap TiO2 prepared by sputtering from 3.2 eV to 2.7 eV upon nitrogen doping. This reduction is as a result of nitrogen impurity states which introduce tail energy levels either in the conduction band or

valence or both of the undoped titanium dioxide.

0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.40

1x103

2x103

3x103

4x103

5x103

6x103

7x103

8x103

9x103

1x104

h

)2 (eV/

m)2

Energy, h eV

FTO Ts= 200oC (Undoped Eg= 3.25 eV)

FTO Ts= 200oC (Doped Eg= 2.90 eV)

Figure 5. Graph showing the effect of doping TiO2 thin film coated on conducting flourine doped tin oxide (FTO) glass slides with nitrogen (N2)and annealed at 400 oC.

3.3. TiO2/In(OH)iSj/Pb(OH)xSy ETA solar cell

performance 3.3.1 I-V characterization

Figure 6 and 7 shows I-V curves for ETA solar cells obtained in dark conditions (no illumination) when fabricated from nitrogen-doped and undoped TiO2 films cell with the cell’s active area of 0.1256 cm2. In dark, solar cells from doped and undoped TiO2 films show no significant difference since there are no photogenerated electrons. The low short circuit current densities (Jsc ≈ 0.085 mA /cm2 for nitrogen-doped cell and Jsc ≈ 0.070 mA /cm2 for undoped cell) are due to very few thermally excited electrons and holes drifting across to p-side and n-side respectively. The thermal excitation is with reference to the electrons occupying the acceptor levels in TiO2 which are very close to the conduction band and therefore requires just a little energy to excite them to conduction band as in the range of ordinary room temperatures of approximately 23o C translating to about 0.026 eV. However, the magnitude of the drift current is higher for nitrogen-doped cell than for the undoped cell suggesting that doping introduced impurity levels even closer to the conduction band compared to the undoped films.

0.0 0.5 1.0 1.5 2.0-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Cur

rent

den

sity

J (m

A/c

m2 )

Voltage (V)

Doped cell Undoped cell

Figure 6. I-V curves (in dark) for doped and undoped TiO2/In(OH)iSj/Pb(OH)xSy solar cell at 23 oC with a cell area = 0.1256 cm2.

0.00 0.03 0.06-0.10

-0.08

-0.06

-0.04

-0.02

0.00

Cur

rent

den

sity

J (m

A/c

m2 )

Voltage (V)


Figure 7. Magnification of figure 6 to show values of Jsc ≈ 0.085 mA /cm2 for nitrogen-doped cell and Jsc ≈ 0.070 mA /cm2 for undoped cell.

Figure 8 shows the I-V characteristics of a nitrogen-doped

and undoped-TiO2/In(OH)iSj/Pb(OH)xSy solar cell under

1000 W/ m2 illumination with a cell’s active area of 0.1256

cm2.

REFERENCES

[1] Quashing, V. (2005). Understanding renewable systems, Earthscan, London, 1-130.

[2] Hankins, M. (1995), Solar electric systems for Africa,Motif Creative Arts, Ltd. Nairobi, Kenya.

[3] Zeman, M., R. VanSwaaij, M. Zuiddam, J. W. Metselaar , R.E.I Schropp, Solar Energy Materials and Solar Cells66, (1-4) 353-359 (2001).

[4] Konig, D. and G.Ebest, Solar Energy Materials and Solar Cells 75 (2003), (3-4) 335-343.

[5] Nanu, M., J. Schooman, A. Goosens, Advanced Materials16(2004), (5) 453-455

[6] Bayon, R., R. Musembi, A. Belaidi, M. Bar, T. Guminskaya, M. Lux-Steiner, Th. Dittrich, Solar Energy Materials and Solar Cells, 89(2005), 18-19.

[7] Musembi, R. (2009), Fabrication and characterization of In(OH)iSj modified highly structured TiO2/ Pb(OH)xSy/PEDOT :PSS eta Solar cell, and study of its transport mechanisms, PhD Thesis, Department of Physics, University of Nairobi, 88-91.

[8] Wafula, B., J. Simiyu, S. Waita, B. Aduda, J. Mwabora Effect of nitration on pressed TiO2 Photoelectrodes for Dye-Sensitized solar cells, African Journal of Science and Technology (AJST), Science and Engineering Series8(2007), (2) 63-71

[9] Gartner, M., P. Osiceanu, M. Anastasescu, T. Stoica, T.F. Stoica, C. Trapalis, T. Giannakopoulou, N. Todoroa, A., Thin solid Films, 516(2008), 8184-8189

[10] Anderson, D. (2001).Clean Electricity from Photovoltaics,M.D Archer and R.D. Hill Edition, Imperial College Press: London.

[11] Baoshun, L., W. Liping, Z. Xiujian, Solar Energy Materials and Solar Cells, 92(2007), 1-10.

97 94

enhanced absorption is due to nitrogen doping which increases the absorption coefficient of the TiO2 thin film.

300 400 500 600 700 800 900 1000 1100 1200 13000.0

5.0x106

1.0x107

1.5x107

2.0x107

Abso

rptio

n C

oeffi

cien

t (

)(m

-1)

Wavelength (nm)

undoped and annealed, 400 oC doped and unannealed undoped and unannealed doped and annealed,400 oC

Figure 3. Dependence of absorption co-efficient on wavelength of nitrogendoped and undoped TiO2 thin films of approximate thickness of 400 nm deposited on fluorine doped tin oxide (FTO) glass slides.

From Figure 4 reflectance curves cluster in a narrow range for the wavelength range considered here as opposed to transmittance that varies greatly with the sample treatment. The nitrogen-doped sample records the lowest transmittance across the considered spectrum. This shows that the nitrogen-doped sample has high absorption for the wavelength range presented.

300 400 500 600 700 800 900 1000 1100 1200 13000

10

20

30

40

50

60

70

80

Tran

smitt

ance

(%)

Wavelength (nm)

undoped and annealed, 400 oC doped and unannealed undoped and unannealed doped and annealed 400 oC

Ref

lect

ance

(%)

10% transmittance line

Figure 4. Dependence reflectance and transmittance on wavelength of nitrogen doped and undoped TiO2 thin films of approximate thickness of 400 nm deposited on fluorine doped tin oxide (FTO) glass slides.

To establish the part of the spectrum where enhanced photoresponse is evident due to nitrogen doping, equation (3)is used [10]:

mE

24.1

................................................................. (3)

where the Eλ is energy in eV and λ is wavelength in micrometers (µm). The equation gives wavelengths which correspond to a minimum of 428 nm for 2.90 eV and a maximum of 384 nm for 3.25 eV that is also confirmed by the observation of band gap reduction due to doping in Section 3.2.2. This is very remarkable since nitrogen doping enhances light absorption beyond the visible to near infra-red.

3.2.2. Effect of Nitrogen-doping on the band gap of TiO2

coated on fluorine doped tin oxide (FTO) glass slides

Figure 5 presents the band gaps of films coated on fluorine doped tin oxide (FTO) substrate and annealed at 400 oC. A significant decrease in the band gap is observed, the undoped film had a band gap of 3.25 eV compared to the nitrogen-doped which had 2.90 eV. This observation concurs with

results earlier obtained by Baoshun [11] who also observed a significant decrease in the band gap TiO2 prepared by sputtering from 3.2 eV to 2.7 eV upon nitrogen doping. This reduction is as a result of nitrogen impurity states which introduce tail energy levels either in the conduction band or

valence or both of the undoped titanium dioxide.

0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.40

1x103

2x103

3x103

4x103

5x103

6x103

7x103

8x103

9x103

1x104

h

)2 (eV/

m)2

Energy, h eV

FTO Ts= 200oC (Undoped Eg= 3.25 eV)

FTO Ts= 200oC (Doped Eg= 2.90 eV)

Figure 5. Graph showing the effect of doping TiO2 thin film coated on conducting flourine doped tin oxide (FTO) glass slides with nitrogen (N2)and annealed at 400 oC.

3.3. TiO2/In(OH)iSj/Pb(OH)xSy ETA solar cell

performance 3.3.1 I-V characterization

Figure 6 and 7 shows I-V curves for ETA solar cells obtained in dark conditions (no illumination) when fabricated from nitrogen-doped and undoped TiO2 films cell with the cell’s active area of 0.1256 cm2. In dark, solar cells from doped and undoped TiO2 films show no significant difference since there are no photogenerated electrons. The low short circuit current densities (Jsc ≈ 0.085 mA /cm2 for nitrogen-doped cell and Jsc ≈ 0.070 mA /cm2 for undoped cell) are due to very few thermally excited electrons and holes drifting across to p-side and n-side respectively. The thermal excitation is with reference to the electrons occupying the acceptor levels in TiO2 which are very close to the conduction band and therefore requires just a little energy to excite them to conduction band as in the range of ordinary room temperatures of approximately 23o C translating to about 0.026 eV. However, the magnitude of the drift current is higher for nitrogen-doped cell than for the undoped cell suggesting that doping introduced impurity levels even closer to the conduction band compared to the undoped films.

0.0 0.5 1.0 1.5 2.0-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Cur

rent

den

sity

J (m

A/c

m2 )

Voltage (V)


Figure 6. I-V curves (in dark) for doped and undoped TiO2/In(OH)iSj/Pb(OH)xSy solar cell at 23 oC with a cell area = 0.1256 cm2.

0.00 0.03 0.06-0.10

-0.08

-0.06

-0.04

-0.02

0.00

Cur

rent

den

sity

J (m

A/c

m2 )

Voltage (V)


Figure 7. Magnification of figure 6 to show values of Jsc ≈ 0.085 mA /cm2 for nitrogen-doped cell and Jsc ≈ 0.070 mA /cm2 for undoped cell.

Figure 8 shows the I-V characteristics of a nitrogen-doped

and undoped-TiO2/In(OH)iSj/Pb(OH)xSy solar cell under

1000 W/ m2 illumination with a cell’s active area of 0.1256

cm2.

0.00 0.05 0.10 0.15 0.20 0.25 0.300

2

4

6

8

10

Undoped cellVoc=0.26 VFF=0.71Jsc=5.75 mA/cm2

=1.06%

Doped cellVoc=0.275 VFF=0.51Jsc=9.5 mA/cm2

=1.32%

Cur

rent

den

sity

J (

mA

/cm

2 )

Voltage (V)

Figure 8. I-V curves for nitrogen-doped and undoped TiO2 window layer in TiO2/In(OH)iSj/Pb(OH)xSy ETA solar cell taken under illumination of 1000 W/m2 at 23 oC with a cell area= 0.1256 cm2.

From this important performance parameters of the solar cell such as fill factor (FF), open circuit voltage (Voc), short circuit current density (Jsc) and photovoltaic conversion efficiency (η) are obtained which are given in Table 1. Short circuit current density for undoped-TiO2/In(OH)iSj/Pb(OH)xSy solar cell is smaller (Jsc = 5.75 mA /cm2) compared to the value obtained for the doped-TiO2/In(OH)iSj/Pb(OH)xSy solar cell (Jsc = 9.5 mA /cm2). This means that on doping, the short circuit current density increased by 65%. This increase in short circuit current density in the doped-TiO2/In(OH)iSj/Pb(OH)xSy solar cell is due the reduction in the band gap caused by doping which improves the photoresponse as discussed in section 3.2.2 and shown in Fig. 5. Upper wavelength limit of absorption is increased from 384 nm to 428 nm on doping because reduced band gap extends absorption to longer wavelengths. This increased absorption plays a role in increasing the energy captured by the window layer to be used in electron-hole generation in the absorber.

The Photovoltaic conversion efficiency (η) is directly related to fill factor, open circuit voltage and short circuit current density. The fill factor is higher in the undoped cell, an observation attributed to the increase in series resistance of the doped cell brought about by the defects introduced in the TiO2 lattice by nitrogen atoms. These defects produce mid-gap levels within the forbidden band gap, which act as indirect recombination centers. Increase in series resistance lowers the fill factor of doped cell as the fill factor is inversely dependent on series resistance.

Table 1: Photovoltaic parameters for the TiO2/In(OH)iSj/Pb(OH)xSy ETAsolar cell

The series resistance Rs,Dark and Rs,Light in the dark and under illumination, respectively and Jmax, the maximum current achievable under illumination for the doped and undoped cells are also given in Table 1.4.0 Conclusions Optical studies were carried out on the nitrogen-doped and undoped TiO2 films. From the transmittance and reflectance spectra, nitrogen-doped films had higher absorption compared to the undoped films. Doping the TiO2 thin film with nitrogen reduces the band gap which is attributed to the introduction of impurity states either on the conduction band or valence or both conduction and valence bands. There was a significant decrease in the band gap upon doping, the undoped film had a band gap of 3.25 eV compared to the nitrogen-doped which had 2.90 eV.

The ETA solar cells fabricated from the TiO2 films varied in performance depending on whether the films were doped or undoped. The doped solar cell had higher photovoltaic conversion efficiency of 1.32% compared to the undoped which had 1.06%. The short circuit current density increases when TiO2/In(OH)iSj/Pb(OH)xSy solar cell is doped with nitrogen. Short circuit current density for undoped-TiO2/In(OH)iSj/Pb(OH)xSy solar cell was smaller (Jsc = 5.75 mA /cm2) compared to the value obtained for the doped-TiO2/In(OH)iSj/Pb(OH)xSy solar cell (Jsc = 9.5 mA /cm2). Series resistance also increased when the cell was doped due to defects introduced by nitrogen atoms in the TiO2 matrix.

ACKNOWLEDGEMENTS This work is supported by African Materials Science and Engineering Network (AMSEN) and International Programme in the Physical Sciences (IPPS), Uppsala University (Sweden).

Rs,Dark

(Ωcm2)

Rs,Light

(Ωcm2)

FF Voc(V)

Jsc(mA/cm2)

Jmax(mA/cm2)

η (%)

Undoped 0.91 0.83 0.71 0.260 5.75 5.16 1.06

Doped 2.47 1.45 0.51 0.275 9.50 7.12 1.32

95 96

Construction and Evaluation of a PV/T Collector Using the Batch Method

G. Samukonga1*, S. Hatwaambo1, G. Chinyama2.1 Department of Physics, University of Zambia, P.O. Box 32379, Lusaka, Zambia.2 Department of Physics, Copperbelt University, P.O. Box 22379,Kitwe, Zambia.

*Corresponding author: [email protected]

AbstractPhotovoltaics are generally made from doped semiconductors whose conductivity decrease with increase in temperature. Suppressing the rise in temperature for cells under low concentration by using water could improve the overall efficiency of a Photovoltaic-Thermal hybrid collector (PV/T). In this study a batch method was employed to analyze the performance of a PV/T under low concentration. The batch methods employed a system of collecting thermal energy in a predetermined period of time and analyzing the consequences on the electrical efficiency. The study experimentally revealed that the thermal efficiency reached a maximum of 70% while electrical efficiency was increased to a maximum of 23% from a standard reference value of 14% for monocrystalline solar cells. The method produced two-third of the tank capacity as hot water per batch .The maximum temperatures were in the range of 650C to 700C. It was concluded that thermal and electrical efficiency decreased exponentially with increase in outlet temperature beyond 46oC regardless of any further increase in insolation. Therefore, the stagnant temperature of the system was approximated to be 460C. The wind profile is important when designing a PV/T system because high wind profiles favor electrical efficiency than thermal output due to high thermal losses which occur at high wind speeds. The batch method provided a large surface area for thermal absorption as a consequence there was better cooling system compared to risers which are usually used in many solar heaters and PV/T systems. Hence, the batch method has been observed to help optimize the efficiency of a PV/T under concentration and it is a tool which can be used to choose the best cells and designing a PV/T . The temperature of usable water was attained by losingslightly part of electrical efficiency (temperature beyond 460 C) which is almost negligible. The electrical and thermal efficiencies were high at insolation close to one sun (1000 -1340W/m2 and temperature of 450 C). The other conclusion is that the PV/T overall system efficiency is adversely affected by type of solar cells used than other factors.

Keywords: PV/T, Batch, stagnant temperature, wind profile

___________________________________________________________________________

1.0 IntroductionThe conversion of energy carried by optical electromagnetic radiation into electrical energy is a physical phenomenon. Photovoltaics are one of the most important types of transducer which carry out such conversions [1]. However, not all radiant energy is transformed by a photovoltaic cell known as the photovoltaic effect [2]. Photovoltaic cells are made from doped semiconductors whose conductivity decrease with increase in temperature. For silicon based solar cells, the open circuit voltage drop at 2mV per degree Celsius for any rise in temperature above room temperature [3]. Suppressing the rise in temperature for cells under concentration and using hot water for other purposes, the overall efficiency of the photovoltaic-Thermal hybrid system (PV/T) could improve. Concentrators in photovoltaic/thermal collectors reduce the cost of solar cells by

minimizing the solar cells collecting area with cheap CPC concentrator materials. The second advantage is that the temperature of the transfer fluid (water) can reach acceptable practical values of 55oc and above [2, 4]. It is for this reason that building PV/T systems at an acceptable cost can take care of limited accessible land and costs needed for huge power generation stations. Stand alone PV/T systems can become more practical in grid and none grid areas .This would reduce stress on the already over stressed grids especially in developing countries due to dual source of energy from a PV/T.

The energy conversion efficiency of solar cells depends on the band gap which is affected by the temperature, electron density and type of material. A material with small band gap responds to long wavelength photons, while the other photons of

longer wavelengths are wasted as heat. Large band gap responds to short wavelength photons of the solar spectrum which means that no single material can respond to the whole solar spectrum. Therefore, there is always some solar energy wasted from any kind of a solar cell.

The aim of the study was to construct a hybrid solar photovoltaic-thermal system to generate both electricity and hot water in batches by using a low parabolic compound concentrator system in an attempt to increase PVT system efficiency. The study is aimed at maximizing the overall efficiency of a PV/T system and demand for the fabrication of solar cells on substrates as heat sinks for effective cooling of solar cells and thermal heat transfer.

2.0: Experimental DetailsUpon choosing the acceptance angle from the dimensions of the PV/T collector and the solar window, a computer program was written to convert polar coordinates into Euclidean coordinates so that the locus of the CPC profile was drawn on two A0 papers. The profile was then transferred onto a plywood to form the CPC of the PV/Tmould. A slot at the base of the wood profile was then made for the thermal collector. The complete assembly of the photovoltaic-thermal collector (PV/T) hybrid system is shown in Figure 1. It consists of a bronze tank retrofitted to a commercial solar panel with diffuse CPC reflectors.

Figure 1: The PV/T system with diffuse booster reflectors.

The PV/T was put on an inclined frame tilted at the latitude of Lusaka and facing the North .

Figure 2: The thermal collector tank

A bronze tank 4mm thickness and heat capacity of 360j/kg was constructed as a thermal collector as shown in Figure

2. The fluid capacity of the tank was approximately 6litres.. Glass wool was selected as the insulator. Glass wool has thermal conductivity of 0.044W/m2 and thickness of 30mm used at the base and sides. Two gate valves were incorporated to assist in controlling the collection of fluids in batches. A simple pressure difference open loop forced circulation per batch was used as a pumping system. A measuring cylinder was used to measure how much hot water was extracted per batch.

Figure.3: The two data loggers and the sensor cables

The electrical performance of the PV/T was monitored by V-I tracker data logger (Figure 3) before and after every batch on the computer moniter. The model cell electrical efficiency of 14.2% was compared to the experimental I-Vdata obtained using the data logger. The experimental data was then inserted in the expression below (54) to calculate

the electrical efficiency of the PV/T.

(1)

where was the current–voltage product, was theeffective solar cells area, and ) was the irradiance at a given temperature.

The CR10X data logger was programmed to sense inlet, outlet, ambient temperatures and solar insolation every 10 seconds. The computer interfaced with the data loggers, was used to view and store the data from the sensors. The performance of the thermal collection from the PV/T collector was done by considering the total incoming energy (the incident irradiation ( , effective collector area (Ac) and the overall transmittance-absorptance of the grazing and substrate materials using equation (2).

(2)

However, there are thermal losses after thermal convertionfrom the insolation. Total thermal energy lost was taken into account before obtaining the amount of useful thermal energy produced. This was done because the thermal energy generated cannot easily be quantified. But ,evaluating thermal energy losses then compare them to the incoming insolation using the double grazing model system. Thermal losses occur through various methods such as convection. This is the process by which heat is transferred between a solid surface and the fluid surrounding it due to macroscopic and molecular motion.Equation (3) shows total heat losses through convection;

(3)

Diffuse reflectors

CR10X Logger

I-V Tracker Logger

software to obtain the graphical representation of the results.

3.0: Data collection, analysis and discussionsThe water in the tank was replaced after a predetermined period of time (batch time). At the time of introducing cold water in the PVT collector, the instantaneous efficiency was observed to increase to its maximum value, and then decreased as the temperature of the water in the tank increased. The higher the output temperature of hot water, the less the electrical efficiency obtained as shown in Figure 4 below.

Water output Temperature 0C

30 35 40 45 50 55 60 65

Eff

icie

ncy

%

0

10

20

30

40

50

60

70

80

Electrical

Thermal

Figure 4: The effect of output temperature on PV/Tefficiency.

The stagnation temperature of a PV/T system is the temperature at which maximum electrical and thermal efficiencies are obtained. Therefore, this should be the operational temperature of a PV/T system to enable it effectively yield optimum efficiency. The efficiency increased exponentially being highest at low water output temperatures but steadily decreased as the output temperatures increased above 450C (stagnant temperature).This phenomenal can be attributed to increase in shunt resistance of the cells with temperature Hence. the shunt resistance losses become a source of power loss after stagnant temperature.

3.1 PV/T thermal efficiencyIt is clear from literature that the amount of energy generated by a solar cell or thermal collector systems increases with increase in insolation. However, there is a limit to which the electrical or thermal energy can be increased as suggested by the experimental results in Figure 5

Figure 5: Thermal efficiency of PV/T system as a function of insolation

The graph in Figure 5 shows that the thermal efficiency increased exponentially with insolation until about 1100W/m2, and started to decrease to 1600W/m2 .The best and maximum efficiencies were attained when theinsolation was between 1000W/m2 and 1200W/m2.Therefore, the optimal insolation for the PV/T prototype was approximated to be 1200W/m2. The decrease in thermal efficiency with temperature might be as a result of the high thermal losses at high temperatures in accordance to Stefan’s law. The law states that radiant losses increase to the power of four when the temperature increases. The high overall PV/T temperatures could be caused by unconverted insolation by the solar cells as well as unused phonons by the thermal collector after the equilibrium temperature (stagnant temperature) of the system. Hence, any further increase in irradiance does not mean more phonons would be available for the heating of the water in the PV/T but instead it is wasted as radiant heat to the surroundings. It was observed that whenever the temperature was above a critical value called stagnant temperature, no meaningful thermal energy efficiency was obtained. This limited the overall thermal energy efficiency of a PV/T.

101 98

Construction and Evaluation of a PV/T Collector Using the Batch Method

G. Samukonga1*, S. Hatwaambo1, G. Chinyama2.1 Department of Physics, University of Zambia, P.O. Box 32379, Lusaka, Zambia.2 Department of Physics, Copperbelt University, P.O. Box 22379,Kitwe, Zambia.

*Corresponding author: [email protected]

AbstractPhotovoltaics are generally made from doped semiconductors whose conductivity decrease with increase in temperature. Suppressing the rise in temperature for cells under low concentration by using water could improve the overall efficiency of a Photovoltaic-Thermal hybrid collector (PV/T). In this study a batch method was employed to analyze the performance of a PV/T under low concentration. The batch methods employed a system of collecting thermal energy in a predetermined period of time and analyzing the consequences on the electrical efficiency. The study experimentally revealed that the thermal efficiency reached a maximum of 70% while electrical efficiency was increased to a maximum of 23% from a standard reference value of 14% for monocrystalline solar cells. The method produced two-third of the tank capacity as hot water per batch .The maximum temperatures were in the range of 650C to 700C. It was concluded that thermal and electrical efficiency decreased exponentially with increase in outlet temperature beyond 46oC regardless of any further increase in insolation. Therefore, the stagnant temperature of the system was approximated to be 460C. The wind profile is important when designing a PV/T system because high wind profiles favor electrical efficiency than thermal output due to high thermal losses which occur at high wind speeds. The batch method provided a large surface area for thermal absorption as a consequence there was better cooling system compared to risers which are usually used in many solar heaters and PV/T systems. Hence, the batch method has been observed to help optimize the efficiency of a PV/T under concentration and it is a tool which can be used to choose the best cells and designing a PV/T . The temperature of usable water was attained by losingslightly part of electrical efficiency (temperature beyond 460 C) which is almost negligible. The electrical and thermal efficiencies were high at insolation close to one sun (1000 -1340W/m2 and temperature of 450 C). The other conclusion is that the PV/T overall system efficiency is adversely affected by type of solar cells used than other factors.

Keywords: PV/T, Batch, stagnant temperature, wind profile

___________________________________________________________________________

1.0 IntroductionThe conversion of energy carried by optical electromagnetic radiation into electrical energy is a physical phenomenon. Photovoltaics are one of the most important types of transducer which carry out such conversions [1]. However, not all radiant energy is transformed by a photovoltaic cell known as the photovoltaic effect [2]. Photovoltaic cells are made from doped semiconductors whose conductivity decrease with increase in temperature. For silicon based solar cells, the open circuit voltage drop at 2mV per degree Celsius for any rise in temperature above room temperature [3]. Suppressing the rise in temperature for cells under concentration and using hot water for other purposes, the overall efficiency of the photovoltaic-Thermal hybrid system (PV/T) could improve. Concentrators in photovoltaic/thermal collectors reduce the cost of solar cells by

minimizing the solar cells collecting area with cheap CPC concentrator materials. The second advantage is that the temperature of the transfer fluid (water) can reach acceptable practical values of 55oc and above [2, 4]. It is for this reason that building PV/T systems at an acceptable cost can take care of limited accessible land and costs needed for huge power generation stations. Stand alone PV/T systems can become more practical in grid and none grid areas .This would reduce stress on the already over stressed grids especially in developing countries due to dual source of energy from a PV/T.

The energy conversion efficiency of solar cells depends on the band gap which is affected by the temperature, electron density and type of material. A material with small band gap responds to long wavelength photons, while the other photons of

longer wavelengths are wasted as heat. Large band gap responds to short wavelength photons of the solar spectrum which means that no single material can respond to the whole solar spectrum. Therefore, there is always some solar energy wasted from any kind of a solar cell.

The aim of the study was to construct a hybrid solar photovoltaic-thermal system to generate both electricity and hot water in batches by using a low parabolic compound concentrator system in an attempt to increase PVT system efficiency. The study is aimed at maximizing the overall efficiency of a PV/T system and demand for the fabrication of solar cells on substrates as heat sinks for effective cooling of solar cells and thermal heat transfer.

2.0: Experimental DetailsUpon choosing the acceptance angle from the dimensions of the PV/T collector and the solar window, a computer program was written to convert polar coordinates into Euclidean coordinates so that the locus of the CPC profile was drawn on two A0 papers. The profile was then transferred onto a plywood to form the CPC of the PV/Tmould. A slot at the base of the wood profile was then made for the thermal collector. The complete assembly of the photovoltaic-thermal collector (PV/T) hybrid system is shown in Figure 1. It consists of a bronze tank retrofitted to a commercial solar panel with diffuse CPC reflectors.

Figure 1: The PV/T system with diffuse booster reflectors.

The PV/T was put on an inclined frame tilted at the latitude of Lusaka and facing the North .

Figure 2: The thermal collector tank

A bronze tank 4mm thickness and heat capacity of 360j/kg was constructed as a thermal collector as shown in Figure

2. The fluid capacity of the tank was approximately 6litres.. Glass wool was selected as the insulator. Glass wool has thermal conductivity of 0.044W/m2 and thickness of 30mm used at the base and sides. Two gate valves were incorporated to assist in controlling the collection of fluids in batches. A simple pressure difference open loop forced circulation per batch was used as a pumping system. A measuring cylinder was used to measure how much hot water was extracted per batch.

Figure.3: The two data loggers and the sensor cables

The electrical performance of the PV/T was monitored by V-I tracker data logger (Figure 3) before and after every batch on the computer moniter. The model cell electrical efficiency of 14.2% was compared to the experimental I-Vdata obtained using the data logger. The experimental data was then inserted in the expression below (54) to calculate

the electrical efficiency of the PV/T.

(1)

where was the current–voltage product, was theeffective solar cells area, and ) was the irradiance at a given temperature.

The CR10X data logger was programmed to sense inlet, outlet, ambient temperatures and solar insolation every 10 seconds. The computer interfaced with the data loggers, was used to view and store the data from the sensors. The performance of the thermal collection from the PV/T collector was done by considering the total incoming energy (the incident irradiation ( , effective collector area (Ac) and the overall transmittance-absorptance of the grazing and substrate materials using equation (2).

(2)

However, there are thermal losses after thermal convertionfrom the insolation. Total thermal energy lost was taken into account before obtaining the amount of useful thermal energy produced. This was done because the thermal energy generated cannot easily be quantified. But ,evaluating thermal energy losses then compare them to the incoming insolation using the double grazing model system. Thermal losses occur through various methods such as convection. This is the process by which heat is transferred between a solid surface and the fluid surrounding it due to macroscopic and molecular motion.Equation (3) shows total heat losses through convection;

(3)

Diffuse reflectors

CR10X Logger

I-V Tracker Logger

where is the total amount of thermal energy lost through convection, is overall convective coefficient, is the total effective area of the PV/T system, To is the fluid output temperature and is the ambient or input temperature. The heat losses through radiation occur every time there is a temperature gradient as in equation (4)

(4)The total insolation absorbed for photo-thermal effect is given by equation (5).

(5)

Where is the absorbed and is the lost thermal energy. Hence, the thermal efficiency was calculated by assuming that unused insolation (phonons) will almost be effectively be transmitted through the substrates ( =0.9)using expression (6).

(6)

Where is the useful thermal energy, incoming irradiation G and collecting area Ac. .In steady state condition, the rate of energy absorbed by the fluid per unit area should be equal to the sum of the rate of useful energy (used to actually heat the water) and the rate of heat lost per unit area by the system to the surroundings. A model modified from Hottel-Whillier-Bliss equations presented by Duffie and Beckman (2006) was adopted [6]. Under these conditions the useful heat gain can be calculated usingequation (6). This expression of useful heat gain (Qu) is represented as a function of the collector area (Ac), the packing factor (F), the transmittance-absorptance product of the photovoltaic cells (ατ) glass cover and substrate, the solar irradiation available (G), the collector heat loss coefficient (U

L) and the temperature difference between the

cooling medium (inlet temperature(Ta) and the output temperature (T

p).

(7)

The heat loss from the PV/T collector is the sum of the heat loses from top grazing, sides (edges) and the bottom of the collector due convection and radiation, as shown below.

(8)

UL is the total amount of heat lost through the three modes;Us represent losses on the sides of the thermal collectorequation (9).

(9)The side loses ( ) depend on conductivity of the insulator k, thickness of insulation (Lin) and the ratio of the top (Ac)

to the sides area (As) as shown by the expression (4). While, Ub are losses at the bottom equation (10)

(10)The losses on top of the thermal collector UT . Most thermal loses in a PV/T are mainly due to convection and radiation because conduction is the initial form of the heat transfer from the PV/T body. Therefore, any good lagging may reduce the losses by conduction. The convective heat loss to the surroundings can be accelerated by wind velocity which help blow out the heat from the PV/T tank surface to the surrounding as shown by equation (11)..

(11)Where V is the average wind velocity and is the subsequent convectional heat loss coefficient due to wind velocity.The other major method of thermal loss in a PV/T is through radiation. The losses increase as a power of 4 with temperature rise according to Stefan-Boltzmann’s lawin expression 12.

(12)

Where s the Stefan’sconstant. is effective the emissivity of the cover and substrate material. Radiant energy falling on the grazing surface heats up the cover andthen the substrate surface. Since thermal losses increases with high temperatures.The expression (13) below wasused to find out how much thermal energy is lost through radiation at the surface cover of the PV/T at any given temperature.

(13)

The overall heat loss due to convection and radiation wasquantified by looking at sum of convective and irradiative heat transfer coefficients h1and h2.

(14)

(15)The subscript c represents coefficient of thermal losses through convection. While r stands for irradiative heat coefficient losses. The overall heat loss coefficient is obtained by finding the reciprocal of the sum of overall convective and radiated heat loss coefficients of expressions 16.

= (16) The overall PV/T performance was found by combining of thermal and electrical efficiencies by using the expression below.

(18)

The thermal and electrical efficiencies were analyzed by comparing them to the input, output water temperatures and the insolation data obtained was then fed into a sigma plot

99 100

where is the total amount of thermal energy lost through convection, is overall convective coefficient, is the total effective area of the PV/T system, To is the fluid output temperature and is the ambient or input temperature. The heat losses through radiation occur every time there is a temperature gradient as in equation (4)

(4)The total insolation absorbed for photo-thermal effect is given by equation (5).

(5)

Where is the absorbed and is the lost thermal energy. Hence, the thermal efficiency was calculated by assuming that unused insolation (phonons) will almost be effectively be transmitted through the substrates ( =0.9)using expression (6).

(6)

Where is the useful thermal energy, incoming irradiation G and collecting area Ac. .In steady state condition, the rate of energy absorbed by the fluid per unit area should be equal to the sum of the rate of useful energy (used to actually heat the water) and the rate of heat lost per unit area by the system to the surroundings. A model modified from Hottel-Whillier-Bliss equations presented by Duffie and Beckman (2006) was adopted [6]. Under these conditions the useful heat gain can be calculated usingequation (6). This expression of useful heat gain (Qu) is represented as a function of the collector area (Ac), the packing factor (F), the transmittance-absorptance product of the photovoltaic cells (ατ) glass cover and substrate, the solar irradiation available (G), the collector heat loss coefficient (U

L) and the temperature difference between the

cooling medium (inlet temperature(Ta) and the output temperature (T

p).

(7)

The heat loss from the PV/T collector is the sum of the heat loses from top grazing, sides (edges) and the bottom of the collector due convection and radiation, as shown below.

(8)

UL is the total amount of heat lost through the three modes;Us represent losses on the sides of the thermal collectorequation (9).

(9)The side loses ( ) depend on conductivity of the insulator k, thickness of insulation (Lin) and the ratio of the top (Ac)

to the sides area (As) as shown by the expression (4). While, Ub are losses at the bottom equation (10)

(10)The losses on top of the thermal collector UT . Most thermal loses in a PV/T are mainly due to convection and radiation because conduction is the initial form of the heat transfer from the PV/T body. Therefore, any good lagging may reduce the losses by conduction. The convective heat loss to the surroundings can be accelerated by wind velocity which help blow out the heat from the PV/T tank surface to the surrounding as shown by equation (11)..

(11)Where V is the average wind velocity and is the subsequent convectional heat loss coefficient due to wind velocity.The other major method of thermal loss in a PV/T is through radiation. The losses increase as a power of 4 with temperature rise according to Stefan-Boltzmann’s lawin expression 12.

(12)

Where s the Stefan’sconstant. is effective the emissivity of the cover and substrate material. Radiant energy falling on the grazing surface heats up the cover andthen the substrate surface. Since thermal losses increases with high temperatures.The expression (13) below wasused to find out how much thermal energy is lost through radiation at the surface cover of the PV/T at any given temperature.

(13)

The overall heat loss due to convection and radiation wasquantified by looking at sum of convective and irradiative heat transfer coefficients h1and h2.

(14)

(15)The subscript c represents coefficient of thermal losses through convection. While r stands for irradiative heat coefficient losses. The overall heat loss coefficient is obtained by finding the reciprocal of the sum of overall convective and radiated heat loss coefficients of expressions 16.

= (16) The overall PV/T performance was found by combining of thermal and electrical efficiencies by using the expression below.

(18)

The thermal and electrical efficiencies were analyzed by comparing them to the input, output water temperatures and the insolation data obtained was then fed into a sigma plot

PV/T output temperature

Water output temperature 0C

30 35 40 45 50 55 60 65

The

rmal

Eff

icie

ncy

%

50

55

60

65

70

75

Figure 6: The PV/T thermal efficiency as a function of the output temperature.

It was deduced that the rise in temperature of the system causes an exponential decrease in the thermal efficiency as illustrated in Figure 6. Similarly, at lower temperatures, the thermal efficiency was low due to the fact that less insolation with low intensity was available and most of it was being used for photovolatic effect than thermal energy production.Therefore, less radiation was being rejected through the substrate material for creation of phonons in the thermal collector tank.

Insolation W/m 2

600 800 1000 1200 1400 1600 1800

Effic

ienc

y %

0

10

20

30

40

50

60

70

80

Figure 7: The PV/T Electrical and thermal efficiency under Concentration

As the temperature increased above 55 oC, the efficiency started decreasing. The highest thermal efficiency wereobtained between the temperatures 45oC and 55 oC. At the time of introducing cold water in the PV/T collector, the thermal efficiency was observed to steadily increase and later decreased as the temperature of the water in the tank increased beyond 46oC. The electrical and thermal efficiency of a monocrystalline cells in a PVT increased as

the insolation was increasing towards 1000 W/m2 then started to decrease exponentially as the insolation increased as shown in Figure 7.This is due to the fact that the efficiency of monocrystalline cells respond effectively to insolation close to one sun. Hence, a PVT system will need special cells and substrates to yield high thermal efficiency and good electrical.

Figure 8: The electrical conversion efficiency and the temperature difference from ambient of a PVT.

For low insolation the electrical efficiency was low. However, fpr insolation close to 1000W/m the electrical efficiency went to their maximum. Thereafter, the efficiencies decreased but the cell PV/T temperature kept on increasing as shown in figure 7.

The electrical efficiency solar cell increase with increase in insolation. (Figure 8). However, there is a limit to which this power can be increased. The experimental results show that electrical output adversely affect the thermal output of a PVT system even though other factors such as the optical (reflector ,concentrator) and thermal transfer characteristics of the materials can limit the PVT system efficiency . In PVT systems the thermal efficiency does not linearly increase as the concentration increases due to thermal losses and material mismatch between the substrate and the PV/T thermal collector. These losses are mainly due to radiation and convection. The radiant losses increase as a power of four in accordance to Stefan law. Moreover, the conductivity of air and the insulator increases with temperature rise which make the thermal efficiency to decrease due to such added sources of energy losses.

It has been observed that a PV/T system without a concentrator may not be viable because the thermal energy output would be low at most times and require longer batch time. Therefore, a PV/T without a concentrator cannot be an energy solution because only electrical energy efficiency would be biased than the thermal energy. The thermal output from a PV/T with a concentrator was six times more

There are various categories of deposition techniques for thin

films; Physical vapour deposition, Chemical vapour

deposition, Oxidation, Spin coating, Plating and Solution

growth technique. Solution growth techniques include;

Chemical bath deposition (CBD) and Spray Pyrolysis. CBD

method uses a controlled chemical reaction to effect the

deposition of a thin film by precipitation. The substrates are

immersed in an alkaline solution containing the chalcogenide

source, the metal ion and added base. A chelating agent is

added to control the release of the metal ions. At a given

temperature when ionic product (IP) of reactants exceeds the

solubility product (SP), precipitation occurs. If the IP<SP,

then the solid phase produced will dissolve back to the

solution resulting in no net precipitation. In this study, CBD

has been chosen because this technique is simple and low

cost technique, it promotes large area deposition allowing

easy doping of the thin films hence the process is scalable

and can be used for large scale production of high quality thin

films [23]. No elaborate arrangement is required for CBD

(Figure 2) and the deposition is usually carried out at normal

atmospheric pressure and at low temperature. High purity

chemicals are also not an essential requirement hence, CBD

is a low cost technique. There is little chance of incorporation

of impurities from the reaction bath onto the deposited thin

films unless the IP of the impurity species is greater than its

SP [5] For CBD, it is not easy to dope intrinsic

semiconductor films with external dopants at the time of film

formation because the concentrations of dopants included in

the reaction bath will not be incorporated to the film unless

IP > SP for that particular dopant [19].

2. Experimental Procedure

2.1. Preparation of substrates

The substrates used for film growth were ordinary

microscope glass slides measuring 762cm by 254cm by 1 cm.

Before use; they were soaked in a bath of acetone and then

rinsed in deionized water. They were then degreased in

hydrochloric acid, rinsed in deionized water, left to dry in air

and then stored in a desiccator ready to be used.

2.2. Deposition of CdxNi1-xS thin films

CBD technique was used in the deposition of the thin films.

A simple experimental set up was used as shown by the

sketch in figure 2. The set up consisted of a hot plate,

magnetic stirrer, thermocouple, DR359TX pH meter and the

reaction bath (RB) in the beakers. The RB was composed of

0.01M Cadmium Chloride (Cd ) solution, 1.0 M Thiourea

( ), some drops of Concentrated Ammonia ( )

solution and various volumes of 0.01M Nickel Chloride

( ) solution. (As shown in Table 1). The deposited thin

films on the glass substrates were labelled to (table 1).

To investigate the effect of the concentration of the RB on

the optical properties and the band gap of the deposited thin

films, a repeat of the deposition was done at a concentration

of 0.8M and 0.8 M Cd .The deposited thin films on

the glass substrates were labelled to ( As shown in

table 2).

In the preparation of the reaction baths, 10 ml Cd solution

was put in 100ml beaker, some drops of conc solution

added, the mixture well stirred and the specified volumes of

solution added. 1M was then added to the

mixture followed by addition of some distilled water at room

temperature and put in a warm water bath at 318 K. Six

depositions ( to ) and ( to ) were made at different

105 102

effective and the electrical output was two times better than the one without a concentrator under the same incident radiation. Morever, most of the water was not hot enough to be used at the end of each batch. But a prolonged batch time would perform better and generate the expected amount of hot water.

Temperature difference from ambient

8 9 10 11 12 13 14

Hot

wat

er in

litr

es

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Figure 9: The amount of hot water with temperature rise.

The amount of hot water increased with rise in temperature. This implies that water with temperature difference below ten degrees was not significant because the intended purpose was not achieved. (Figure 9). The graph shows that the amount of hot water was reducing as the temperature was rising as the system was becoming inefficient at high temperatures.

4.0 ConclusionIn this research, a batch method of collecting heated water and its effect on the power output of a solar panel has been investigated. It has been experimentally observed that the thermal efficiency reached a maximum of 70% while electrical efficiency was increased by a maximum of 23% from a standard reference value of 14% for monocrystaline solar cells. We concluded that thermal and electrical efficiency decreased exponentially with increase in outlet temperature beyond 46oC regardless of any further increase in insolation. Therefore, the overall efficiency of the PVT increases logarithmically with increase in insolation until stagnant temperature of the system is reached. We also observed that the wind profile is important when designing a PVT system because high wind profiles favor electrical efficiency than thermal output. High wind velocities help the PVT systems lose a lot of thermal energy which keep the cells at low operating temperatures. The batch method provided a large surface area for thermal absorption as a consequence, the cells operated at a lower. A combination of solar cells and thermal collectors so that both electrical and thermal energy are harvested can be complimented by a Thermol- electric feedback so that the electrical output so that the electrical energy collected through the usual

photovoltaic-effect would be boosted. At present, cooling the cells highly help increase the efficiency of solar cell in turn the overall efficiency of PV/T increases because the

total efficiency (η) of the PV/T is the sum of electrical and thermal efficiencies as expressed below. temperature delivering more power because it makes the minority carriers lifetime to increase by effectively cooling both the substrate and solar cells. The electrical and thermal efficiencies were high to insolation close to one sun (1000-1342 W/m2) and ambient temperatures close to 45oC than other wise. Therefore, the batch method is a key tool in selecting the best cells for a PV/T and hence, its desgn.

AcknowledgmentsI would like to give my gratitude to the international Programme in Physical Sciences of Uppsala University in Sweden for financial support .I’m also indebted to Professor Manyala, P. Kaloyeru, Dr. H. Mweene and Dr. Kalezhi for their academic assistance; Dr. Chuba, the assistant Dean of postgraduate for his valuable advice and the technical staff at the Department of Physics for their tireless assistance and MSSEESA for support.

References[1] Yiannis Tripanagnostopoulos, (1st July-10th July

2004, Patra),”low concentration hybrid

photovoltaic/thermal solar energy systems”, PV

systems Teaching and Learning.

[2] Brogren M. P. Nostell , and B. Karlsson solar

energy, 69, (2000), pp 173-183.

[3] Ommen R, S. Jayaraman. energy conversion and

management 42, (2001), pp1379-1399

[4] Goetz M, P. Torres, P. Pernet, J.Meier, D. Fischer,

H. Keppner, A.Shah. (2000), ”N-I-P

michromorph solar cells on aluminium

substrates”, Institut de Microtechnique (IMT) de

l'Université de Neuchâtel, Rue Breguet 2, CH-

2000, Neuchâtel, Switzerland.

[5] Brogren M., M. Rönnelid, B. Karlsson,(2000).

“PV-thermal hybrid low concentrating CPC

module,” in Proceedings of the 16th European

Photovoltaic Solar Energy Conference and

Exhibition, 1–5, Glasgow, UK.

[6] Winston R. solar energy, 17, (1974), 255 - 8.

[7] Zondag. H. A., D.W. De Vries, W.G.J Van

Helden, R. J. C VanZolingen. A. A. Van

Steenhoven, Solar Energy, 72, (2002), 113-128.

The effect of surface passivation on optical properties of as-grown and annealed chemically deposited CdxNi1-xS

thin films as a material for photovoltaic application R. T. Shikambe1, 2,*, M. K. Munji1, R. J. Musembi2, P. M. Mwathe1

1Department of Physics, Kenyatta University, P.O. Box 43844-00100 GPO, Nairobi, Kenya

2Department of Physics, University of Nairobi, P.O. Box 30197- 00100 GPO, Nairobi, Kenya

Email address:

[email protected] (R.Shikambe), [email protected] (R.Musembi), [email protected] (M. Munji)

Abstract: Thin films of CdxNi1-xS were grown on plain glass substrates by the chemical bath deposition (CBD) technique. Cadmium chloride, nickel chloride and thiourea were used as sources for cadmium, nickel and sulphur ions respectively. Optimization on the deposition parameters was done in order to obtain high quality thin films. The effect of varying the nickel volume on the optical properties of CdxNi1-xS thin films was studied. Transmittance and reflectance data in the range 300 nm-1100 nm were measured using UV-VIS NIR Spectrophotometer Solid State 3700 DUV for all the as-grown, annealed and passivated thin film samples. Transmittance values above 78% for both annealed and as-grown samples were observed. Passivated samples had transmittance values within the same range. From the optical analysis, the films showed low reflectance values below 20% in the UV, VIS–NIR regions. Passivating as-grown and annealed samples had very little effect on the transmittance, reflectance, absorbance and the band gap of the thin film samples. Increase in the concentration of Cd2+

and Ni2+ ions in the reaction bath was found to increase the band gap. The optical measurements were simulated using SCOUT 2.4 software to determine optical constants and then plotted on Origin 9.1 to obtain energy band gap of the thin films. Band gap values of 2.55 eV - 3.49 eV for as-grown samples and 2.82 eV – 3.50 eV for annealed samples were obtained. For passivated as-grown samples, the band gap values were within the range 2.85 eV – 3.12 eV and 2.81eV -3.09 eV for passivated annealed samples. EDX-800HS Energy Dispersive X-rays Spectrometer was used to determine the elemental composition of CdxNi1-xS samples. S, Ni and Cd ions were the major elements of interest. The qualitative analysis showed the presence of all the mentioned ions.

Keywords: Cadmium Nickel Sulphide, Thin films, CBD, Optical properties, EDX, Annealing, Passivation, solar cell

1. Introduction

All fuels derive their source by utilizing energy from the sun [4]. A thin film photovoltaic cell (TFPV) is made by depositing one or more thin layers of

Photovoltaic material on a substrate [3, 7]. Photovoltaic cells absorb light (photons) from the sun and convert them into electricity (Figure 1).

Figure1. P-type and n-type thin film semiconductors

103 104

volumes of nickel as a dopant at the same deposition time of

1 hour and same temperature of 318 K but different

concentration of Cd and solution (as shown in

table 1 and table 2). During the deposition time, the RB was

stirred continuously using a magnetic stirrer at each stage to

obtain a homogeneous temperature. Solution was used

as a complexing agent as well as maintaining the pH of the

RB between 9 and 10. The microscope glass substrates were

inserted in the RB inclined to the wall of the glass beaker.

After each deposition, they were removed; rinsed in distilled

water, drip dried in open air and then kept in a desiccator for

Characterization. The value of x varied between 1.0 and 0.5

according to the equation; . The volume of

solution was varied according to the formula;

[1]

where, is the value of varying volume of

solution, is the value of the volume of Cd solution.

2.3. Annealing of the Thin Films

The thin films were annealed in a tube furnace (schematic

diagram shown in Fig 3), in argon atmosphere at a

temperature of 573 K for 1hour. Annealing was done in order

to improve the microstructure and crystallinity of the

coatings. It also helped in hardening the coatings and also

improves the electrical conductivity of CdS: Ni thin films.

2.4. Thin Film Passivation

The thin films were passivated by annealing them in a

nitrogen atmosphere in a tube furnace (fig 3) for 1 hour at a

temperature of 573 K. Passivation is known to improve on

the stability of doped thin films [21], in this study it was one

in order to study its effect on optical properties and band gap

of CdNiS thin films.

Table 1.Preparation of CdxNi1-xS at 0.01M and 0.01M Cd fort optimized time and temperature

Table 2.Preparation of CdxNi1-xS at 0.8M and 0.8M Cd for optimized time and temperature

RB CdCl2 NiCl2 value of x Value of 1-x

Con(M)

Vol(ml)

Con(M)

Vol(ml)

N1 0.01 10 0.01 0 1 0

N2 0.01 10 0.01 1.1 0.9 0.1N3 0.01 10 0.01 2.5 0.8 0.2N4 0.01 10 0.01 4.3 0.7 0.3N5 0.01 10 0.01 6.7 0.6 0.4N6 0.01 10 0.01 10 0.5 0.5

RB CdCl2 NiCl2 value of x Value of 1-x

Con(M)

Vol(ml)

Con(M)

Vol(ml)

N7 0.8 10 0.8 0 1 0

N8 0.8 10 0.8 1.1 0.9 0.1N9 0.8 10 0.8 2.5 0.8 0.2

N10 0.8 10 0.8 4.3 0.7 0.3N11 0.8 10 0.8 6.7 0.6 0.4N12 0.8 10 0.8 10 0.5 0.5

Figure 5. XRF spectrum of as-deposited optimized CdS thin films from EDX 800HS

Figure 6. XRF spectrum of as-deposited optimized CdNiS thin films from EDX 800HS

109 106

Figure 2: Set up for chemical bath deposition technique of thin

Figure3.Horizontal tube furnace for annealing and passivation


Optical study was aimed at determining optical properties

like; reflectance, transmittance, absorption coefficient,

refractive index and dielectric function. Reflectance and

transmittance measurements were obtained using a

computerized double beam Solid-Spec3700 DUV Shimadzu

Spectrophotometer. The spectra measurements were taken in

the spectral range from 300-1100 nm of wavelength.

Analysis of the collected data was done using pre-developed

models in Scout 2.4 software [26, 27].

Transmittance data were fitted to the simulated spectra from

which refractive indices, absorption coefficients, dielectric

constants, real and imaginary parts and energy loss

parameters were obtained. The graphs were drawn using the

Origin Pro 2.4 software. Drude model was used since the

films had free carriers like metals and conductors. With this

model, the parameters adjusted were; plasma frequency,

damping constant and charge carrier density. OJL model was

used for the sake of interband transitions. The parameters

varied were; decay, gap energy and gamma. Tauc Lorentz

model was used due to the crystallinity of the films and

interband transition. Parameters varied were; resonance

frequency and damping and Dielectric Background was used

in which parameters varied were real part and imaginary part.

2.6. Elemental Composition Analysis of

Energy Dispersive X-rays Spectrometer model EDX-800HS

was used to perform the elemental composition analysis of

thin films to ascertain presence of S2-, Ni2+ and

Cd2+ ions. Qualitative analysis involved the identification of

the elements present in the samples. Quantitative analysis

involved determination of the concentrations of the elements

present.


3.1. Elemental Composition Analysis

In table 3: (a), S and Cd were the major elements of interest

in sample N1. Other elements like; Si is a trace impurity

which might have come as a result of silicon internal

fluorescence peak due to the photoelectric absorption of X-

rays by the silicon dead layer in the detector resulting in the

emission of Si K- X-rays from this layer into the active

volume of the detector. Presence of Rh element was as a

result of the X-ray tube which is made of this material.

Presence of Sb, Ca and Fe might have been caused by

spectral interference problems in EDX machine which occur

at low energy (< 3 keV) [24].Figure 5 shows the peak based

analysis of the elemental composition of CdS thin films.

From the spectrum analysis, the thin films had composition

of the elements at Kα, Kβ, Lα and Lβ lines. Qualitative

results in table 3 (b) shows presence of S, Rh, Ni, Cd and K

in sample N5. The sample comprised of Cd, Ni and S ions

with Ni appearing in trace quantity. This implies that Ni acted

as an impurity in the as-deposited CdNiS. Figure 6 shows the

peak based analysis of the elemental composition of CdNiS

thin film samples.

3.2. Optical Properties

Effect of surface Passivation on optical properties

of as-grown and annealed thin film samples was investigated

basing on the transmittance, reflectance and absorbance of

various samples under different deposition parameters.

Table 3. The elemental composition analysis results

Sample Analyzed Qualitative Results Quantitative Results

(a) N1 (CdxNi1-xS) , x=1 S 55.223%

Rh --------Cd 44.642%Sb 0.029%Ca 0.101%Fe 0.005 %

(b)N5 (CdxNi1-xS), x=0.6S 21.643%

Rh --------Ni 4.006%Cd 74.345%K 0.006 %

107 108

Figure 7(a). A Plot of Spectral Transmittance of CdNiS Thin Film (Sample N1 – N6, As-deposited thin films).

300 400 500 600 700 800 900 1000 110010

20

30

40

50

60

70

80

90

100

Tra

nsm

ittan

ce,T

(%)

Wavelength, (nm)

AN1 AN2 AN3 AN4 AN5 AN6

Figure 7(b).A plot of Spectral Transmittance of CdNiS Thin Film (Sample AN1 – AN6, Annealed samples).

Figures 7 (a) and (b) show the spectral transmittance

of CdNiS thin film deposited at 318 K for 0.01M

concentaration of CdCl2 and NiCl2 solutions.It is observed

that the transmittance value of CdXNiX-1S thin films

increased as the volume of NiCl2 increased i.e., it was high

at 10 ml and 6.7 ml for as-grown and annealed thin films

respectively. It is also noted that the transmittance values of

the film is a bout 20% at the UV region. The average

transmittance of the as-deposited thin films was 78% at 10

ml of Ni ions within the visible part of the spectrum. It was

observed to increase to 80 % for annealed samples at 6.7 ml

doping. The plot indicates that film of high transmittance

value above 84% in NIR was obtained for samples deposited

at 6.7ml-10ml volumes of NiCl2 solution. Annealing slightly

improved the transmittance spectra of the films due the fact

that annealing makes a material more crystalline and packing

order of molecules is enhanced causing transmittance to be

enhanced. An increase in transmitance at 6.7 ml -10 ml of

nickel chloride is attributed to decrease in diffuse and

multiple reflections caued by increase in grain size and in

light- scattering effect[21,25,2].

Figure 7(c).A plot of Spectral Transmittance of CdNiS Thin Film (Sample N7 – N12, As-deposited thin films).

Figure 7(d). A Plot of Spectral Transmittance of CdNiS Thin Film (Sample AN7 – AN12, annealed samples).

Figure 7 (c) and (d) show the plot of spectral

transmittance for samples N7 – N12 deposited at a

temperature 318 K for 0.8 M concentaration of CdCl2 and

NiCl2 solutions. There is an average decrease in

transmittance from a value of 80 % to 76 % within the visible

part of the spectrum, as the concentration of NiCl2 and CdCl2

300 400 500 600 700 800 900 1000 110010

20

30

40

50

60

70

80

90

Tran

smitt

ance

,T(%

)

Wavelength,(nm)

N1 N2 N3 N4 N5 N6

300 400 500 600 700 800 900 1000 11000

10

20

30

40

50

60

70

80

90

Tran

smitt

ance

,T(%

)

Wavelength,(nm)

N7 N8 N9 N10 N11 N12

300 400 500 600 700 800 900 1000 11000

10

20

30

40

50

60

70

80

90

Tran

smitt

ance

,T(%

)

Wavelength,(nm)


Figure 10(a). A plot of Absorbance of CdNiS Thin Films (Sample AN7 – AN12, As-grown samples) The optical absorption coefficients (α) of were determined

from the T% and R% values [18];

α = [2]

where, d is the film thickness. Scout 2.4 software was used

to validate the values of α obtained and the two agreed within

a discrepancy of not more than 5%. There was generally low

absorption at photon energy less than 2.5 eV at 0.01M and at

photon energy of 3.75 eV for 0.8M. The extinction

coefficients (k) of the thin films were calculated using the

expression [12];

k = [3]

These values were validated using Scout 2.4 software. The

value of k increased with increase in photon energy for all

the samples. Low values of k, 0 to 1 were noted for films

deposited at concentration of 0.01M and 0 to 11 for thin films

at 0.8M. The refractive index, n for the thin films were

obtained using Scout 2.4 software. We observed that, at

0.01M concentration of the RB, the range for n for both as-

grown and annealed film was 1 to 2. The range remained

same for 0.8M concentration. The average maximum value

of n was 2.1 at 4.25 eV. The values were low at low photon

energy. The maximum values of k and n for all the samples

occurred at same photon energy of 4.25 eV, which is agrees

with the findings of Greenway and Harbeke [7, 11], which

states that for semiconductors, the maximum value of n

occur at the energy near that at which the maximum change

in k occurs. The optical conductivity (o) was obtained using

the equation [11];

o = [4]

where, c is the speed of light in vacuum.

The high magnitude of Optical conductivity (σo) of 1012 s-1

was obtained for all the films. This shows that they have good

photo-response as a material for solar cell application. On

average, annealing did not have much effect on the optical

conductivity of the thin films but increased concentration of

the reaction baths increased the conductivity. Higher values

of σo for all the samples (as-grown and annealed) occurred at

high photon energy. The dielectric function is a complex

quantity and a fundamental intrinsic property of the material

which consists of both the real and imaginary parts [12].

ɛ = ɛr + ɛi [5] where,

ɛ is the dielectric function, ɛr the real part and ɛi the

imaginary part. The real part indicates how the speed of light

in the material can be slowed down while the imaginary part

deals with the absorption of energy by a dielectric from the

electric field due to dipole motion [12]. The real and

imaginary parts were obtained from Scout 2.4 software by

fitting the experimental transmittance data within the

wavelength range 300nm-1100nm. They were validate using

the relations [12],

Ɛr = n2-k2 [6]

Ɛi = 2nk [7]

Real part for all the samples showed almost constant values

averagely above 2.0 from the lowest to the highest photon

energy values. The imaginary part values were below 0.5 for

lower energy values and showed an upward trend for energy

values above 3.5eV. The constant values for the real part are

attributed to the fact that for semiconductors [6], k2 <<

n2.This showed that indeed CdNiS thin films deposited is a

semiconductor material.

113 110

solution increased from 0.01 M to 0.8 M. The value is close

to 0% at UV region. A close observation on graphs 7 (c)-(d)

reveals that, the transmittance value of CdNiS thin films

increases as the volume of the dopant (Ni2+) increased. It is a

maxmum at about 81% at 10 ml of doping. The decrease in

transmittance as the concentration increases from 0.01M to

0.8M is due to increase in photon absorption as photon

striking increases with increase in carrier

concentration[21,25]. A close look at the transmittance

spectral graphs 7 (a) – 7(d) reveals highest values a bove 85%

in NIR region since most materials are transparent to infra-

red radiation[27]. Annealing N7-N12 films slightly increased

the transmittance of the films within the VIS region.High

transmittance in VIS-NIR region exhibited by these films

makes them good materials for construction of window

layers in solar cells.

Figure 8 (a).A Plot of Reflectance of CdNiS Thin Film(Sample N1 – N6, As-deposited samples).

Figure 8(c). A plot of Reflectance of CdNiS Thin Film (Sample N7 – N12, As-deposited samples).

Figure 8 (d).A plot of Reflectance of CdNiS Thin Film (Sample AN7 – AN12, Annealed samples).

Figures 8(a) and (b) show the spectral reflectance of the as-

grown and annealed CdNiS thin films deposited at 0.01 M

concentration of Ni and Cd ions. The reflectance values of

the films ranged between 12.5% and 32.5% in the VIS region

for both as-grown and annealed samples with the highest

values at wavelength of 500nm. The un-doped films (at 0.0

ml of nickel ions) had lowest values of reflectance. Those at

2.5 ml doping gave the highest values of reflectance.

Annealing had very small effect on the reflectance values

obtained. Figures 8(c-d) show the spectral reflectance of the

as-grown and annealed CdNiS thin films grown at 0.8 M

concentration of Ni and Cd ions. Increase in RB

concentration from 0.01M to 0.8M reduced the reflectance

range from between 12.5% to 32.5% to between 8% and

21% within the visible part of the spectrum.

Figure 8(b). A Plot of Reflectance of CdNiS Thin Film (Sample AN1 – AN6, Annealed samples).

The highest value was recorded at 500 nm at 2.5ml doping

level and lowest below 9% across VIS-NIR was noted at 10

300 400 500 600 700 800 900 1000 11005

10

15

20

25

30

35

Rele

ctan

ce,R

(%)

Wavelength, (nm)

N1 N2 N3 N4 N5 N6

300 400 500 600 700 800 900 1000 11008

10

12

14

16

18

20

22

24

26

28

Relec

tance

,R(%

)

Wavelength,(nm)


300 400 500 600 700 800 900 1000 11002

4

6

8

10

12

14

16

18

20

22

Relec

tance

,R(%

)

Wavelength,(nm)


300 400 500 600 700 800 900 1000 1100

6

8

10

12

14

16

18

20

22

Refle

ctanc

e,R (%

)

Wavelength,(nm)

N7 N8 N9 N10 N11 N12

ml doping. With low reflectance value in the VIS-NIR region

means the material is best for photovoltaic devices like solar

cells as a window layer. This will decrease reflective loss on

the cell surface [8].

Figure 9(a). A plot of Absorbance of CdNiS Thin Films (Sample N1 – N6, As-grown samples)

Figure 9(b). A plot of Absorbance of CdNiS Thin Films (Sample AN1 – AN6, Annealed samples)

Figures 9(a-b) show the absorbance spectra for films

deposited at 0.01M concentration of Ni2+ and Cd2+. It is high

in the UV region for all the samples and between 0% and

32.5% for the VIS-NIR region. The poor absorption is

attributed to high transmittance within the region. The films

with high absorbance in the UV region and low absorbance

in the VIS-NIR region can be applied in the construction of

window layers of p-n junction solar cells [13, 23].

Figures 9(c-d) show the absorbance spectra for

films deposited at 0.8 M concentration for Ni2+ and Cd2+. The

thin films absorb heavily at UV region with value over 93%.

The absorption increased with the increase in the

concentration from 0.01M to 0.8M. The increase is

attributed to decrease in spectral transmittance-reflectance.

A close look at the spectral graphs 9(c-d) reveals low

absorbance value close to 10% within the VIS-NIR region of

the electromagnetic spectrum. Annealing the thin films did

not have much effect on the absorbance. Thin films at 2.5ml

doping showed highest value of absorption with the film at

10ml doping giving the least values.

Figure 10(a). A plot of Absorbance of CdNiS Thin Films (Sample N7 – N12, As-grown samples)

300 400 500 600 700 800 900 1000 11000

10

20

30

40

50

60

Abs

orba

nce,

A(%

)

Wavelength, (nm)

N1 N2 N3 N4 N5 N6

300 400 500 600 700 800 900 1000 11000

10

20

30

40

50

60

AN1 AN2 AN3 AN4 AN5 AN6A

bsor

banc

e,A

(%)

Wavelength,(nm)300 400 500 600 700 800 900 1000 1100

0

10

20

30

40

50

60

70

80

90

Abso

rban

ce, R

(%)

Wavelength, (nm)

N7 N8 N9 N10 N11 N12

300 400 500 600 700 800 900 1000 11000

10

20

30

40

50

60

70

80

90

AN7

A


111 112

3.3. Band gap variation

The relation between the absorption coefficient (α)

and the incident photon energy (hν) can be written as [9];

(Αhν) 1/n = A (hν - Eg) (6)

where, A is a constant, Eg is the band gap of the material and

the exponent n depends on the type of transition. The values

of n for direct allowed and indirect allowed transition are n =

½ and n= 2 respectively. We obtained the band gaps of the

films from the linear portion of (αhν) 2 versus hν plot. The

band gap of the as-prepared thin films deposited at 0.01M

concentration (N1-N6) was found decrease from 3.17 eV to

2.55 eV as the volume of NiCl2 in the RB increased (fig 13).

By increasing the concentration to 0.8 M (N7-N12), the band

gap increased from between 2.55 eV-3.17 eV (for N1-N6) to

between 3.16 eV-3.49 eV (for N7-N12, fig 14). Annealing

the as-prepared thin films had a narrowing effect in the band

gap. The band gap range for as-deposited samples at 0.01 M

(N1-N6) was 2.55 eV-3.17 eV and 2.83 eV-3.11 eV for the

annealed samples (AN1-AN6) at same concentration of Ni2+

and Cd2+. Same narrowing effect was observed for the

sample prepared at 0.8M concentration whose band gap

range was 3.16 eV - 3.49 eV (for N7-N12) and 3.24 eV - 3.50

eV (for AN7-AN12). There was an increase in band gap with

the increase in the concentration of the RB.The decrease in

the band gap as the increase in the CdS dopant volume (i.e.,

Nickel Chloride solution) from 0 ml to 10 ml can be

attributed to the incorporation of Ni2+ ions in the crystal

lattice, which gives rise to donor levels in the CdNiS band

gap. This causes the conduction band to lengthen leading to

a reduction in the band gap [21]. An increase in the band gap

as the concentration of Ni2+ and Cd2+ ions increased from

0.01 M to 0.8 M can be attributed to the Burstein-Moss effect

which states that: Increase in free carrier concentration, due

to the high doping levels, fills empty states belonging to the

conduction band of the thin films thereby increasing the

energy magnitude required for the valence band to

conduction band transitions as reported by Muathe et al.,

2014 [21]. The decrease in the band gap due to annealing of

the thin films is due to the fact that annealing process allows

the packing of crystallites of molecules in the material to be

ordered hence reducing the intermolecular defects within the

material causing the reduction in the band gap value [21].The

high band gap values ranging between 2.56 eV and 3.6 eV

are in good agreement with results obtained by Ottih I.E. and

Ekpunobi A.J., (2011) [20] whose value was between 2.49

eV- 3.2 eV and Ezema et al., (2009) [9] whose value was 2.6

eV- 2.82 eV. CdNiS is a good candidates for window layer

material for photovoltaic devices because of its large band

gap which vary from 2.49 eV to 3.50 eV [8].In order to have

a reasonable chance of capturing a photon, the n-type layer

has to be fairly thick. This also increases the chance that an

ejected electron will meet up with a previously created hole

in the material before reaching the p-n junction [10, 17].

3.4. Effect of Passivation

The effect of varying growth parameters such as deposition

time, bath composition and bath temperature on the various

properties of thin films have been reported by several

researchers [20,7,14,9,16,28]. In this paper, we were

concerned with investigating the effect of surface passivation

on optical properties of as-grown and annealed CdNiS thin

films. Figures 11(a-b) and 12 (a-b) show the effect of surface

passivation on the transmittance and reflectance spectra of

as-grown and annealed thin films. A close look at these

graphs and graphs in figures 7(a-b) and 8(a-b) reveals no

much difference in transmittance and reflectance spectra

[6] Dieter W. and Dieter M. (1991). Organic Solar Cells.Institute of Advanced Materials, Global journal of science 3:132-140.

[7] Ezema F.I (2009). “Effect of Deposition Time on the Band Gap and Optical Properties of Chemical Bath Deposited CdNiS Thin Films”. Optoelectronics and Advanced Material Rapid Communications. 3:141-144.

[8] Ezenwa I. and Ekpunobi A. (2009). Optical properties and band offsets of CdS/PbS Super lattice. The Pacific Journal of Science and Technology, 11: 404-407.

[9] Ezugwu S. C., Ezema F. I. *, Osuji R. U., Asogwa P. U., Ekwealor A. B. C. and Ezekoye B. A. (2009). Effect of deposition time on the band-gap and optical properties of chemical bath deposited cdnis thin films. Optoelectronics and advanced materials – rapid communications, 3: 141 –144.

[10] Green M. A. (April 2002).Third generation photovoltaic solar cells for 2020 and beyond. Journal for Low –dimensional Systems and Nanostructure, 14: 65-67.

[11] Ilenikhena P. A. (2008). Comparative Studies of Improved

Chemical Bath Deposited Copper Sulphide (CuS) and Zinc Sulphide (ZnS) Thin Films at 320K and Possible Applications. African Physical Review, 2:0007

[12] Jeroh M.D. and Okoli D.N. (2012). Optical, Structural and Morphological Studies of Chemical Bath Deposited Antimony Sulphide Thin Films. Global Journal of Science Frontier Research (Physics & Space Science) 12:2.

[13] Kassim U., Narayanan H. and Anthony O. (2008). Optimization of process parameters of chemical bath deposition of Cd1-xZnxS thin films. Leonardo Journal of Sciences, 12: 111-120.

[14] Khallif H., Chai G., Lupan O., Chow L., Heinrich H. and Park L. (2009). In-situ boron doping of chemical-bath deposited CdS thin films. Physica Status Solidi. A206:256–262.

[15] Kumar S. (2006). "Mathematical modeling and thermal performance analysis of unglazed transpired solar collectors”. Solar Energy 81: 62–75

[16] Manolache S.A., Andronic L., Duta A., Enesca A. (2007). The Influence of the Deposition Condition on Crystal Growth and on the Band Gap of CuSb2S3 Thin Film Absorber Used For Solid State Solar Cells (SSSC).

[17] Mosiori C. (2012). Electrical and optical properties of CdZnS and PbS thin films for photovoltaic applications. Msc. Thesis, Kenyatta University, Kenya

[18] Nair P.K., Ocampo M., Fernandez A. and Nair M.T.S. (1989). “Solar Control Characteristics of Chemically Deposited PbS Films for Solar Control Applications”. Journal of Solar Energy Materials 20:235.

[19] Nair P. K. and Nair M. T. S. (1990). Solar Cell Technology. Journal of Physics D, 23: 150.

[20] Ottih I.E. and Ekpunobi A.J. (2011). Fabrication and Characterization of High Efficiency Solar Cell Thin Films (CdNiS). Pacific Journal of Science and Technology. 12:351-355.

[21] Patrick M.M., Musembi R., Munji K.M., Odari V., Munguti L., Ntilakigwa A.A., Nguu J., Aluda B. and

Muthoka B. (2014). Influence of Surface Passivation on Optical Properties of Spray Pyrolysis Deposited Pd-F: SnO2. International Journal of Materials Science and Applications

[22] Reed S.J.B. (1993). Electron Microprobe Analysis, 2nd ed. Cambridge University Press, Cambridge

[23] Reynolds J. E. (1884).’’An overview of e- Beam Mask- Making in Solid State Technology’’. Journal for Chemical Society 45: 162.

[24] Sankara N.S, Santhi B, Sundareswaran S, Venkatakrishnan KS. Studies on Spray Deposited SnO2, Pd: SnO2 and F: SnO2 Thin Films for Gas Sensor Applications, Synthesis and Reactivity in Inorganic. Metal-Organic and Nano-Metal Chemistry. 2006; 36(1):131-135.

[25] Sankara N.S, Santhi B, Sundareswaran S, Venkatakrishnan KS. Studies on Spray Deposited SnO2, Pd: SnO2 and F: SnO2 Thin Films for Gas Sensor Applications, Synthesis and Reactivity in Inorganic. Metal-Organic and Nano-Metal Chemistry. 2006; 36(1):131-135

[26] Theiss W. in: W. Theiss (Ed.), Scout Thin Film Analysis Software Handbook, Hard and Software, Aachen, Germany, 2001 www.mtheiss.com, Pg. 54-57

[27] Theiss W. (2000). Scout thin films analysis software handbook, edited by Theiss M (Hand and Software Aachen German) www.mtheiss.com

[28] Ugwu I. and Onah U. (2007). Optical characteristics of chemical bath deposited CdS thin film characteristics within UV, Visible, and NIR radiation. The Pacific Journal of Science and Technology, 8: 155-161

117 114

across the UV-VIS-NIR region of the EM spectrum. Hence

surface passivation had very little influence on transmittance

and reflectance spectra.

Figure 11(a). A plot of Spectral Transmittance of passivated CdNiS Thin Films (Sample PN1 – PN6).

300 400 500 600 700 800 900 1000 110010

20

30

40

50

60

70

80

90

Tran

smitt

ance

, T (%

)

Wavelength, (nm)

PAN1 PAN2 PAN3 PAN4 PAN5 PAN6

Figure 11(b). A plot of Spectral Transmittance of passivated annealed CdNiS Thin Films (Sample PAN1 – PAN6).

Figure 12(a). A plot of Spectral Reflectance of passivated CdNiS Thin Films (Sample PN1 – PN6).

Figure 12(b). A plot of Spectral Reflectance of passivated annealed CdNiS Thin Films (Sample PAN1 – PAN6).

We observed that the absorbance was high in the

UV region for all the samples and between 0% and 35 % in

the VIS-NIR region for as-grown passivated samples (PN1-

PN6) and annealed passivated samples (PAN1-PAN6) as

compared to 0 % to 32.5 % for as-grown (N1-N6) and

annealed (AN1-AN6) thin films deposited. Hence,

passivation had also minimal influence on the absorbance

spectra of CdNiS thin films. Low absorption at photon

energy less than 2.5 eV for both as-grown passivated (PN1-

PN6) and annealed passivated thin films (PAN1-PAN6) was

also observed. There was more absorption rate at higher

photon energy for the films in UV region. Optical

conductivity, Real part of the dielectric function, Imaginary

part of dielectric function and the Refractive indices were not

much influenced by the effect of passivation. Values

obtained for these constants for passivated and un-passivated

thin films were within the same range. The band gap range

for as-grown passivated thin films (PN1-PN6, figure 13) was

found to range between 2.85 eV and 3.12 eV compared to the

range for as-grown thin films (N1-N6) which was between

2.56 eV and 3.42 eV. The band gap range for annealed

passivated thin films (PAN1-PAN6) ranged between 2.81 eV

and 3.09 eV compared to 3.12 eV -3.482 eV for annealed thin

films (AN1-AN6).

300 400 500 600 700 800 900 1000 110010

20

30

40

50

60

70

80

90

Tran

smitta

nce,

T(%)

Wavelength,(nm)

PN1 PN2 PN3 PN4 PN5 PN6

300 400 500 600 700 800 900 1000 11005

10

15

20

25

30

35

Refle

ctanc

e,R(%

)

Wavelength, (nm)

PN1 PN2 PN3 PN4 PN5 PN6

300 400 500 600 700 800 900 1000 11005

10

15

20

25

30

35

Refle

ctanc

e,R (%

)

Wavelength,(nm)

PAN1 PAN2 PAN3 PAN4 PAN5 PAN6

Figure 13. Variation of the Band gaps of the films (N1-N6, AN1-AN6, PN1-PN6 and PAN1-PAN6) with the deposition parameters

Figure 14. Variation of the Band gaps of the films (N7-N12 and AN7-AN12) with the deposition parameters

4. Conclusion

CdxNi1-xS thin films deposited onto glass substrates chemical

bath deposition technique have low reflectance values in the

UV- VIS-NIR regions, low transmittance values at UV

region and high at VIS-NIR regions. The average band gap

has been found to be above 2.80 eV. This high band gap

value of the film and the optical properties stated make it

suitable for use as a window layer in solar cells application.

Surface passivation has been found to have minimal effect

on the optical properties and band gap of CdxNi1-xS thin

films. This implies that, when used as a window layer in solar

cells, changing environmental conditions will have

insignificant effect on its performance.

Acknowledgements

The authors would wish to thank University of Nairobi and

Kenyatta University for access of equipment and the

technical staffs of both universities, particularly Mr.

Muthoka and Mr. Mudimba, for their advices while carrying

out the experiments in the laboratory.

REFERENCES

[1] Amanullah F., Al-Shammari S. and Al-Dhafiri A.(2005).Co-activation effect of chlorine on the physical properties of CdS thin films prepared by CBD technique for solar applications Journal of physica status Solidi (Applied research), 201:2470 – 2480.

[2] Benjamin V. Odari, Musembi R.J., Mageto M.J., Othieno H., Gaitho F., Mwamburi M.and Muramba V.(2013). Optoelectronic Properties of F-co-doped PTO Thin Films Deposited by Spray Pyrolysis. American Journal of Materials Science, 3(4): 91-99.

[3] Butti K. and Perlin J. (1981). A Golden Thread (2500 Years of Solar Architecture and Technology. Van Nostrand Reinhold.ISBN 0-442-24005-8.

[4] Chapin A. M. and Pearson L. (1954). A new Silicon p-njunction photocell for converting solar radiation into electrical power. Journal of Applied Physics, 25: 676-681.

[5] Ezugwu S. C., Ezema F. I. *, Osuji R. U., Asogwa P. U., Ekwealor A. B. C. and Ezekoye B. A. (2009). Effect of deposition time on the band-gap and optical properties of chemical bath deposited cdnis thin films. Optoelectronics and advanced materials – rapid communications, 3: 141 –144.

As-deposited

Annealed

Passivated

Annealed then Passivated

N1, 3.1

N1, 2.98

N1, 3.04

N1, 2.97

N2, 3.17

N2, 2.92

N2, 2.97

N2, 3.09

N3, 2.99

N3, 2.83

N3, 3.06

N3, 2.87

N4, 3.02

N4, 3.08

N4, 3.12

N4, 2.81

N5, 2.55

N5, 3.11

N5, 2.97

N5, 3.06

N6, 2.94

N6, 3.02

N6, 2.86

N6, 2.81

Band Gap variation with deposition parameters

N1N2N3N4N5N6

As-deposited samples

Annealed samples

N7, 3.47

N7, 3.32

N8, 3.24

N8, 3.5

N9, 3.16

N9, 3.25

N10, 3.2

N10, 3.47

N11, 3.35

N11, 3.29

N12, 3.49

N12, 3.42

Band Gap variation with deposition parameters

N7N8N9N10N11

Table 2: Parameters of P3HT:SQ3:PCBM solar cells

Solar Cell

Voc

(V)Isc

(mA)Vmax

(V)Jmax

(mA)Pmax FF

(%)A - - - - - - -B 0.53 5.78 0.32 3.39 1.08 0.35 1.3C 0.61 6.88 0.37 4.29 1.59 0.38 2.5D 0.62 7.88 0.36 5.04 1.81 0.37 2.8E 0.64 9.68 0.36 6.86 2.41 0.40 3.9

Table 2 is extracted from figure 5 and shows the meas-ured Isc and Voc. Solar cell A was characterized in the dark and as such had no open circuit voltage and no short circuit current as shown in figure 5. The measured Voc and Isc for solar cells B, C, D and E was used to calculate the FF, maximum current density (Imax), maximum voltage (Vmax), Pmax and power conversion efficiency () of P3HT:SQ3:PCBM organic solar cells.

Conclusion

We have enhancement the device performance of P3HT:PCBM polymer solar cells by heat treatment and dye sensitization using SQ3 dye III. The effect of SQ3 dye on light harvesting of the P3HT:PCBM solar cell occurs by increasing the dye concentration. In the P3HT:SQ3:PCBM combination, SQ3 dye appears as a donor material and in-creases the site for exciton dissociation in the blend. The combined contribution of SQ3 dye and thermal annealing at 140 C resulted in increased power conversion efficiency (η) of pristine P3HT:PCBM solar cell from 2.8% to 3.9%. This technique will come to play an important role in the practi-cal application of the polymer solar cells.

ACKNOWLEDGMENTS

The authors would like to express gratitude to the finan-cial support from the International Science Project (ISP) of Sweden, Nanoscience Africa Network (NANOAFNAC) and Material Science for East Central and Southern Africa (MSSEESA). Special mention should also be made to iThemba Labs, Physics and Chemistry Departments of University of Zambia, Physics Department of University of Western Cape in South Africa and Physics Department of the University of Dar-es-salaam in Tanzania.

REFERENCES[1] [14] W. J. Belcher, K. I. Wagner, P. C. Dastoor, The effect of

porphyrin inclusion on the spectral response of ternary P3HT:porphyrin:PCBM bulk heterojunction solar cells, Sol.Energy Mater. Sol. Cells 91 (2007) 447–452.

[2] N. S. Sariciftci, L. Smilowitz, A. J. Heegerand F. Wudl, Science, (1992), 258: 1474

[3] A. Haugeneder, M. Neges, C. Kallinger, W. Spirkl, U. Lem-mer, J. Feldmann, U. Scherf, E. Harth, A. Gügeland K. Mül-len, (1999), Physical Review B, 59: 15346

[4] J. J. M. Halls, C.A. Walsh, N. C. Greenham, E. A. Marseglia, R. H. Friend, S.C.Morattiand A. B. Holmes, Nature,(1995), 376: 498

[5] S. E. Shaheen, C. J. Brabec, N. S. Sariciftci, F. Padinger, T. Fromherz and J. C. Hummelen, (2001) Applied Physics Let-ters, 78: 841

[6] P. Schilinsky, C. Waldaufand C. J. Brabec, (2002), Applied Physics Letters, 81: 3885

[7] G. Yu, J. Gao, J. C. Hummelen, F. Wudl, A. J. Heeger, Poly-mer photovoltaic cells: enhanced efficiencies via a network of internal donor–acceptor heterojunc- tions, Science 270(1995) 1789–1791.

[8] F. Padinger, R.S. Ritterberger, N.S. Sariciftci, Effects of post-production treatment on plastic solar cells, Adv. Funct.Mater.13 (2003) 85–88.

[9] National Renewable Energy Laboratory. Best Research-Cell Efficiencies. [accessed on 11 December 2012]. Available online: http://www.nrel.gov/ncpv/images/efficiencychart.jpg.

[10] Y. Li. Molecular design of photovoltaic materials for polymer solar cells: Toward suitable electronic energy levels and broad absorption. Acc. Chem. Res. 2012;45:723–733.[PubMed]

[11] Y. Kim, S. Cook, S. M. Tuladhar, S. A. Choulis, J. Nelson, J. R. Durrant, D. D. C. Bradley, M. Giles, I. McCulloch, C. S. Ha, M. Ree, A strong regioregularity effect in self-organizing conjugated polymer films and high-efficiency polythio-phene:fullerene solar cells, Nat. Mater. 5 (2006) 197–203.

[12] J. Y. Kim, S. H. Kim, H. H. Lee, K. Lee, W. Ma, X. Gong, A. J. Heeger, New architecture for high-efficiency polymer photovoltaic cells using solution- based titanium oxide as an optical spacer, Adv. Mater. 18 (2006) 572–576.

[13] G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, High- efficiency solution processable polymer pho-tovoltaic cells by self-organization of polymer blends, Nat.Mater. 4 (2005) 864–868.

[14] J. A. Hauch, P. Schilinsky, S. A. Choulis, R. Childers, M. Biele, C. J. Brabec, Flexible organic P3HT:PCBM bulk-heterojunction modules with more than 1 year outdoor lifetime, Sol. Energy Mater. Sol. Cells 92 (2008) 727–731.

[15] E. Klimov, W. Li, X. Yang, G. G. Hoffmann and J. Loos,

“Scanning Near-Field and Confocal Raman Microscopic In-

vestigation of P3HT-PCBM Systems for Solar Cell Applica-

tions”, Macromolecules 2006, 39, 4493-4496

[16] H. Kim, W. So and S. Moon, “Effect of Thermal Annealing

on the Performance of P3HT/PCBM Polymer Photovoltaic

Cells”, Journal of the Korean Physical Society, Vol. 48, No.

3, March 2006, pp. 441-445

115 116

Enhancement of Photo-Absorption in P3HT:PCBM Blends using Squarylium III Dye

M. Tembo1, O. Munyati2, S. Hatwaambo1, M. Maaza3

1Department of Physics, School of Natural Sciences, University of Zambia, P.O. Box 32379, Lusaka, Zambia2Department of Chemistry, School of Natural Sciences, University of Zambia, P.O. Box 32379, Lusaka, Zambia

3Nanosciences Laboratories, Themba LABS, P.O. Box 722, Somerset West 7129, Western Cape, South Africa

AbstractNano-size thin films comprising poly (3-hexylthiophene) (P3HT) and a fullerene derivative [6, 6] Phenyl-C61-butyric acid methyl ester (PCBM) incorporating squarylium dye III (sq3) are reported. The materials prepared were evaluated for their optical, electrical and photo-conversion efficiency. Materials comprising a blend ratio 1:0.6:1 of P3HT:SQ3:PCBM were deposited by spin-coating or screen printing to produce thin films measuring 100 nm and subsequently annealed at 140°C for 10 minutes. The films were characterized by UV-Vis-NIR spectroscopy for their optical properties, atomic force microscopy for surface morphology and film thickness, and electrical properties. Optical measurements for blends incorporating different amounts of dye showed increased photo-absorbance with increasing dye concentration. The combined contribution of squarylium III dye and thermal annealing resulted in increased power conversion efficiency (η) of pristine P3HT:PCBM solar cells from 2.8 % to 3.9 %. The dye in the active layer improved photo-absorption by enhanced light harvesting while thermal treatment improved the nanoscale morphology leading to better metal-film interface contact and broadening of the absorption wavelength range.

Keywords: polymer solar cell, squarylium dye III, P3HT (Poly-3-hexylthiophene),PCBM ([6,6]-phenyl-C61-butyric acid methyl ester),bulk heterojunction, spin coating

1. Introduction

The ever increasing demand for energy has lead to re-searchers exploring various options in attempt to mitigate the energy deficit which include tapping of solar energy. Polymer solar cells (PSC) have attracted a lot attention in recent years due to lower fabrication cost, possibility of flexible substrate and high volume production. Additionally, polymer solar cells properties can be fine-tuned by modify-ing the architecture of the polymer or blend. One of the most promising PSC in terms of efficiency and long term stability, is the system based on region-regular poly(3-hexylthiophene) (rrP3HT) and the fullerene deriva-tive [6,6]-phenyl-C61-butyric acid methyl ester (PCBM). P3HT functions as an electron donor while PCBM the elec-tron acceptor material respectively. Bulk heterojunction (BHJ) solar cells based on P3HT and PCBM is currently the most promising system for use in polymer based solar cells. This is due, in-part, to their flexible material proper-ties and potential for low-cost manufacturing [1].

In a typical polymer solar cell, light absorption creates strongly bound excitons, which can dissociate into free charges only at donor/acceptor (D/A) interfaces by rapid electron transfer from the donor to acceptor [2, 3]. The chemical architecture of the components used and the local nano-scale organization of the photoactive layer are largely responsible for the efficiency of the PSC. The challenge is to generate a maximized interface between the donor and

acceptor materials within the whole volume of the photoac-tive layer for maximum exciton dissociation. Additionally,to ensure the creation of continuous pathways for charge transport to both of the electrodes [4-6].

The reported efficiencies for organic solar cells have seen an increase from the 1% recorded in the first BHJ device in 1995 [7,8]. This increase in efficiency is a resulted of the progress made in better understanding the photo-conversion mechanism, which instructs device design and material synthesis [9,10]. By 2003, P3HT:PCBM cells with 3.5% [11] were reported, and in 2007 multiple groups reported efficiencies of 4–4.5% with P3HT/PCBM active layer ma-terial combination [12-14].

One of the reasons for organic solar cells having lower efficiencies in comparison to their inorganic counterparts is that the spectral range of the optical absorption of the con-ducting polymers is relatively narrow compared to the solar spectrum. The limitation in light absorption across the solar spectrum limits the photocurrent of the solar cells. To achieve high efficiencies in PSCs, more research is required to understand the fundamental electronic interactions be-tween the polymeric donors and the fullerene acceptors. In addition, the complex interplay of device architecture, morphology, processing and the fundamental electronic processes involved needs to be better understood. Here we describe a strategy to use of squarylium III (SQ3) dye and thermal annealing to expand the absorption range.

Table 2: Parameters of P3HT:SQ3:PCBM solar cells

Solar Cell

Voc

(V)Isc

(mA)Vmax

(V)Jmax

(mA)Pmax FF

(%)A - - - - - - -B 0.53 5.78 0.32 3.39 1.08 0.35 1.3C 0.61 6.88 0.37 4.29 1.59 0.38 2.5D 0.62 7.88 0.36 5.04 1.81 0.37 2.8E 0.64 9.68 0.36 6.86 2.41 0.40 3.9

Table 2 is extracted from figure 5 and shows the meas-ured Isc and Voc. Solar cell A was characterized in the dark and as such had no open circuit voltage and no short circuit current as shown in figure 5. The measured Voc and Isc for solar cells B, C, D and E was used to calculate the FF, maximum current density (Imax), maximum voltage (Vmax), Pmax and power conversion efficiency () of P3HT:SQ3:PCBM organic solar cells.

Conclusion

We have enhancement the device performance of P3HT:PCBM polymer solar cells by heat treatment and dye sensitization using SQ3 dye III. The effect of SQ3 dye on light harvesting of the P3HT:PCBM solar cell occurs by increasing the dye concentration. In the P3HT:SQ3:PCBM combination, SQ3 dye appears as a donor material and in-creases the site for exciton dissociation in the blend. The combined contribution of SQ3 dye and thermal annealing at 140 C resulted in increased power conversion efficiency (η) of pristine P3HT:PCBM solar cell from 2.8% to 3.9%. This technique will come to play an important role in the practi-cal application of the polymer solar cells.

ACKNOWLEDGMENTS

The authors would like to express gratitude to the finan-cial support from the International Science Project (ISP) of Sweden, Nanoscience Africa Network (NANOAFNAC) and Material Science for East Central and Southern Africa (MSSEESA). Special mention should also be made to iThemba Labs, Physics and Chemistry Departments of University of Zambia, Physics Department of University of Western Cape in South Africa and Physics Department of the University of Dar-es-salaam in Tanzania.

REFERENCES[1] [14] W. J. Belcher, K. I. Wagner, P. C. Dastoor, The effect of

porphyrin inclusion on the spectral response of ternary P3HT:porphyrin:PCBM bulk heterojunction solar cells, Sol.Energy Mater. Sol. Cells 91 (2007) 447–452.

[2] N. S. Sariciftci, L. Smilowitz, A. J. Heegerand F. Wudl, Science, (1992), 258: 1474

[3] A. Haugeneder, M. Neges, C. Kallinger, W. Spirkl, U. Lem-mer, J. Feldmann, U. Scherf, E. Harth, A. Gügeland K. Mül-len, (1999), Physical Review B, 59: 15346

[4] J. J. M. Halls, C.A. Walsh, N. C. Greenham, E. A. Marseglia, R. H. Friend, S.C.Morattiand A. B. Holmes, Nature,(1995), 376: 498

[5] S. E. Shaheen, C. J. Brabec, N. S. Sariciftci, F. Padinger, T. Fromherz and J. C. Hummelen, (2001) Applied Physics Let-ters, 78: 841

[6] P. Schilinsky, C. Waldaufand C. J. Brabec, (2002), Applied Physics Letters, 81: 3885

[7] G. Yu, J. Gao, J. C. Hummelen, F. Wudl, A. J. Heeger, Poly-mer photovoltaic cells: enhanced efficiencies via a network of internal donor–acceptor heterojunc- tions, Science 270(1995) 1789–1791.

[8] F. Padinger, R.S. Ritterberger, N.S. Sariciftci, Effects of post-production treatment on plastic solar cells, Adv. Funct.Mater.13 (2003) 85–88.

[9] National Renewable Energy Laboratory. Best Research-Cell Efficiencies. [accessed on 11 December 2012]. Available online: http://www.nrel.gov/ncpv/images/efficiencychart.jpg.

[10] Y. Li. Molecular design of photovoltaic materials for polymer solar cells: Toward suitable electronic energy levels and broad absorption. Acc. Chem. Res. 2012;45:723–733.[PubMed]

[11] Y. Kim, S. Cook, S. M. Tuladhar, S. A. Choulis, J. Nelson, J. R. Durrant, D. D. C. Bradley, M. Giles, I. McCulloch, C. S. Ha, M. Ree, A strong regioregularity effect in self-organizing conjugated polymer films and high-efficiency polythio-phene:fullerene solar cells, Nat. Mater. 5 (2006) 197–203.

[12] J. Y. Kim, S. H. Kim, H. H. Lee, K. Lee, W. Ma, X. Gong, A. J. Heeger, New architecture for high-efficiency polymer photovoltaic cells using solution- based titanium oxide as an optical spacer, Adv. Mater. 18 (2006) 572–576.

[13] G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, High- efficiency solution processable polymer pho-tovoltaic cells by self-organization of polymer blends, Nat.Mater. 4 (2005) 864–868.

[14] J. A. Hauch, P. Schilinsky, S. A. Choulis, R. Childers, M. Biele, C. J. Brabec, Flexible organic P3HT:PCBM bulk-heterojunction modules with more than 1 year outdoor lifetime, Sol. Energy Mater. Sol. Cells 92 (2008) 727–731.

[15] E. Klimov, W. Li, X. Yang, G. G. Hoffmann and J. Loos,

“Scanning Near-Field and Confocal Raman Microscopic In-

vestigation of P3HT-PCBM Systems for Solar Cell Applica-

tions”, Macromolecules 2006, 39, 4493-4496

[16] H. Kim, W. So and S. Moon, “Effect of Thermal Annealing

on the Performance of P3HT/PCBM Polymer Photovoltaic

Cells”, Journal of the Korean Physical Society, Vol. 48, No.

3, March 2006, pp. 441-445

121 118

Enhancement of Photo-Absorption in P3HT:PCBM Blends using Squarylium III Dye

M. Tembo1, O. Munyati2, S. Hatwaambo1, M. Maaza3

1Department of Physics, School of Natural Sciences, University of Zambia, P.O. Box 32379, Lusaka, Zambia2Department of Chemistry, School of Natural Sciences, University of Zambia, P.O. Box 32379, Lusaka, Zambia

3Nanosciences Laboratories, Themba LABS, P.O. Box 722, Somerset West 7129, Western Cape, South Africa

AbstractNano-size thin films comprising poly (3-hexylthiophene) (P3HT) and a fullerene derivative [6, 6] Phenyl-C61-butyric acid methyl ester (PCBM) incorporating squarylium dye III (sq3) are reported. The materials prepared were evaluated for their optical, electrical and photo-conversion efficiency. Materials comprising a blend ratio 1:0.6:1 of P3HT:SQ3:PCBM were deposited by spin-coating or screen printing to produce thin films measuring 100 nm and subsequently annealed at 140°C for 10 minutes. The films were characterized by UV-Vis-NIR spectroscopy for their optical properties, atomic force microscopy for surface morphology and film thickness, and electrical properties. Optical measurements for blends incorporating different amounts of dye showed increased photo-absorbance with increasing dye concentration. The combined contribution of squarylium III dye and thermal annealing resulted in increased power conversion efficiency (η) of pristine P3HT:PCBM solar cells from 2.8 % to 3.9 %. The dye in the active layer improved photo-absorption by enhanced light harvesting while thermal treatment improved the nanoscale morphology leading to better metal-film interface contact and broadening of the absorption wavelength range.

Keywords: polymer solar cell, squarylium dye III, P3HT (Poly-3-hexylthiophene),PCBM ([6,6]-phenyl-C61-butyric acid methyl ester),bulk heterojunction, spin coating

1. Introduction

The ever increasing demand for energy has lead to re-searchers exploring various options in attempt to mitigate the energy deficit which include tapping of solar energy. Polymer solar cells (PSC) have attracted a lot attention in recent years due to lower fabrication cost, possibility of flexible substrate and high volume production. Additionally, polymer solar cells properties can be fine-tuned by modify-ing the architecture of the polymer or blend. One of the most promising PSC in terms of efficiency and long term stability, is the system based on region-regular poly(3-hexylthiophene) (rrP3HT) and the fullerene deriva-tive [6,6]-phenyl-C61-butyric acid methyl ester (PCBM). P3HT functions as an electron donor while PCBM the elec-tron acceptor material respectively. Bulk heterojunction (BHJ) solar cells based on P3HT and PCBM is currently the most promising system for use in polymer based solar cells. This is due, in-part, to their flexible material proper-ties and potential for low-cost manufacturing [1].

In a typical polymer solar cell, light absorption creates strongly bound excitons, which can dissociate into free charges only at donor/acceptor (D/A) interfaces by rapid electron transfer from the donor to acceptor [2, 3]. The chemical architecture of the components used and the local nano-scale organization of the photoactive layer are largely responsible for the efficiency of the PSC. The challenge is to generate a maximized interface between the donor and

acceptor materials within the whole volume of the photoac-tive layer for maximum exciton dissociation. Additionally,to ensure the creation of continuous pathways for charge transport to both of the electrodes [4-6].

The reported efficiencies for organic solar cells have seen an increase from the 1% recorded in the first BHJ device in 1995 [7,8]. This increase in efficiency is a resulted of the progress made in better understanding the photo-conversion mechanism, which instructs device design and material synthesis [9,10]. By 2003, P3HT:PCBM cells with 3.5% [11] were reported, and in 2007 multiple groups reported efficiencies of 4–4.5% with P3HT/PCBM active layer ma-terial combination [12-14].

One of the reasons for organic solar cells having lower efficiencies in comparison to their inorganic counterparts is that the spectral range of the optical absorption of the con-ducting polymers is relatively narrow compared to the solar spectrum. The limitation in light absorption across the solar spectrum limits the photocurrent of the solar cells. To achieve high efficiencies in PSCs, more research is required to understand the fundamental electronic interactions be-tween the polymeric donors and the fullerene acceptors. In addition, the complex interplay of device architecture, morphology, processing and the fundamental electronic processes involved needs to be better understood. Here we describe a strategy to use of squarylium III (SQ3) dye and thermal annealing to expand the absorption range.

2. Experimental Procedures

2.1. Preparation of Polymer Blends

P3HT (Kintec Company of China, 99.9%), PCBM (Kintec Company of China, 99.9 %) and Squarylium dye III (SQ3, Hallo Chem Pharma Import and Export Limited, 99.8%)were used as received. The blends comprising P3HT:SQ3:PCBM (1:0.6:1) in chloroform solvent were vigorously stirred for more than 24 hours at room tempera-ture under nitrogen atmosphere in a glove box to maximize mixing.

Figure 1. Molcular structures of P3HT and PCBM

2.2. Film and Device Fabrication

P3HT:PCBM (neat) and P3HT:SQ3:PCBM blend films were prepared by spin-coating onto cleaned micro-glass substrates. The optical properties of P3HT:SQ3:PCBM thin films were subsequently measured without annealing to evaluate the effect of dye concentration on photo-absorption.Neat P3HT:PCBM films were heated at different tempera-tures to investigated the effects of thermal annealing.

Solar cell devices were fabricated by using the procedure as follows. ITO-glass (Sheet resistance 15 ohm/square)substrates were sequentially cleaned in an ultrasonic bath using acetone and methanol, rinsed with de-ionized water and finally dried. A layer of PEDOT:PSS then was spin-coated (1500 rpm) onto the ITO glass substrate in air and dried using a digitally controlled hot plate at 100˚C for 10minutes. The thin films were heated at 100˚C for 10 minutes to removing any residual solvent. A 100 nm thick aluminum (Al) electrode was thermally deposited onto the active layer using a vacuum deposition system at a pressure of about 3 x 10-4 Pa through a shadow mask to obtain cells with an active area of 1 cm2. The devices were then ther-mally annealed at 140˚C for 4 minutes. The organic solar cells, with a configuration as shown in Figure 2, were stored in the dark.

Figure 2. Device configuration of a typical of polymer solar cell

2.3. Characterization of Thin Films and Solar Cell Devices

Film thickness was measured using Tencor Alpha-step Profiler while optical characteristics were determined using a Lambda 90 UV-Vis-NIR Spectrophotometer. Electrical properties were evaluated using a KEITHLEY 4200 at room temperature under a nitrogen atmosphere. The power con-version efficiency (PCE) was calculated from the currentdensity-voltage (I-V) characteristics under air mass 1.5 global solar simulated light irradiation of 100 mW cm-2.


3.1. Optical Properties of P3HT:SQ3:PCBM Blends

The contribution of SQ3 in the optical absorption of the ternary P3HT:SQ3:PCBM solar cell active layer is shown Figure 3. The Figure shows absorption spectra at SQ3 loading levels of 2, 4 and 6 mg. Varying the amount of dye content in the blend as 2 mg, 4 mg and 6 mg. Optical ab-sorption is seen to increase with increasing loading of the dye.

c c Figure2. Cross-sectional view of organic photovoltaic device

P3HT

350 400 450 500 550 600 650 7000.0

0.2

0.4

0.6

0.8

1.0

Abso

rban

ce (a

.u)

Wavelength (nm)

P3HT:PCBM film P3HT:PCBM with 2mg dye P3HT:PCBM with 4mg dye P3HT:PCBM with 6mg dye

Al

LIGHT

Glass

ITO

PEDOT:PSS

P3HT:SQ3:PCBM

Al

Figure 3. Absorption spectra of pristine P3HT:PCBM (black) ans P3HT:SQ3:PCBM blend films with 2 mg dye (red), 4 mg (blue) and 6 mg (green) of SQ3 dye.

The addition of 2, 4 and 6 mg of SQ3 to the P3HT:PCBM blend increased the peak absorbance from 0.63 to 0.67, 0.74 and 0.82 respectively. Therefore, in-creasing the SQ3 dye concentration in the P3HT:PCBM blend films increases light harvesting through the increase in the peak absorption value with a little broadening in the absorption range. This means that the total number of ab-sorbed photons by the solar cell active layer increases when SQ3 is included in the P3HT:PCBM blend.

3.2. Effect of Thermal Annealing on Optical Properties of P3HT:PCBM Blends

UV-Vis absorption spectroscopy was used to study theeffect of thermal annealing on thin films of P3HT:PCBM spin coated on glass substrates. The profile of non-annealed blend shows an absorption maximum of 0.30 at 501 nm with shoulders at 545 nm and at 597 nm. Thermal annealing at temperatures of 100 °C, 120 °C and 140 °C increased the peak absorptions to 0.31, 0.32 and 0.36 respectively. Thin films treated at 140 °C show a red shift of peak absorption wavelength from 501 nm to 510 nm while the shoulders shift to 549 nm and 604 nm respectively.

Figure 4. UV-Vis absorbance spectra of 100 nm P3HT:PCBM blend films spin coated at 1400 rpm treated in the following ways: as cast (green), thermally annealed at 100 ˚C (blue), 120 ˚C (red) and 140 ˚C (black).

Heating the P3HT:PCBM active layer above 100C ena-bles re-ordering of the polymer chains and free diffusion of the fullerene molecules in the composite. Thermal anneal-ing optimized the morphology of P3HT:PCBM to a ther-modynamically favorable state [15]. The thermal annealing contributes to the enhanced PSC performance by optimiz-ing both the donor/acceptor morphology in the BHJ active layer and the interfacial contact between the metal electrode and the active layer [16].

3.4. Electrical Properties

The solar cell electrical characteristics are determined by measuring the current density to voltage (I-V) characteris-tics, both in dark and under illumination. The open circuit voltage (Voc), short circuit current (Jsc), fill factor (FF) and maximum power point (Pmax) were deduced from the I-Vcurve.

3.5Current density-Voltage (J-V) Char-acterization

The polymer solar cell device configuration is shown in table 1 and the corresponding I-V curve is shown in figure 6.Thermal annealing was done at a temperature of 140˚C.

Table 1. Polymer solar cells device structure

Solar Cell

DEVICE STRUCTURE

A ITO/PEDOT:PSS/P3HT:SQ3:PCBM/Al(NOT ANNEALED, NOT ILLUMINATED)

B ITO/PEDOT:PSS/P3HT:PCBM/Al(NOT ANNEALED, ILLUMINATED)

C ITO/PEDOT:PSS/P3HT:SQ3:PCBM/Al(ANNEALED, ILLUMINATED)

D ITO/PEDOT:PSS/P3HT:PCBM/Al(ANNEALED, ILLUMINATED)

E ITO/PEDOT:PSS/P3HT:SQ3:PCBM/Al(ANNEALED, ILLUMINATED)

Figure 5. Current-density versus Voltage (I-V) curves obtained for solar cells under AM 1.5 solar spectrum simulations (light) at irradiation inten-sity of 100 mW cm-2.

The performance of the polymer solar cells is summa-rized in table 2. The devices with the active layers of P3HT:PCBM blend thermally annealed at 140° performed better than the device with as-spun active layer. The open circuit voltage (Voc) also shows significant enhancementafter thermal annealing.

119 120

Department of Physics MSSEESA Conference on P.O. Box 32379 … · 2019-09-05 · Thin Films for...

Documents

Transcript of Department of Physics MSSEESA Conference on P.O. Box 32379 … · 2019-09-05 · Thin Films for...