RESEARCHARTICLE

Copyright © 2008 American Scientific PublishersAll rights reservedPrinted in the United States of America

Journal ofNanoscience and Nanotechnology

Vol. 8, 1–6, 2008

Solar Selective Coatings Based onNickel Oxide Obtained via Spray Pyrolysis

Mihaela Voinea∗, Elena Ienei, Cristina Bogatu, and Anca DutaThe Centre: Product Design for Sustainable Development, Transilvania University of Brasov,

Eroilor 29, 500036, Brasov, Romania

The paper presents the optimization process for obtaining NiO thin layers on copper substrate forsolar absorber coatings, using an inexpensive and up-scalable technique: spray pyrolysis deposi-tion (SPD). Efficient selective coatings must present a high absorption coefficient of the incidentsolar irradiation, and low emission of heat. The solar selective coatings design involves tailoringthe surface properties for superior optical properties. The deposition parameters were varied formaximizing the solar absorbance and minimizing the thermal emittance. The film morphology wascontrolled using copolymers of the maleic anhydride as additives into the precursors’ solution. Thestructural and surface properties of the films were investigated by X-ray diffraction and atomic forcemicroscopy, respectively. The Cu/CuOx /NiO solar absorber shows good values for the solar absorp-tance (�s = 0�95) and thermal emittance (�T = 0�05) compared with the ones obtained by othermethods employed in literature and new additives are recommended in tailoring the surface of solarselective coatings.

Keywords: Solar Absorber, Spectrally Selective Coatings, Thermal Emittance, SolarAbsorptance, Maleic Anhydride Copolymers.

1. INTRODUCTION

Selective surfaces are required to improve the solar col-lectors’ efficiency, thus the conversion of solar radiationinto thermal energy. Advanced coatings should fulfill twomain conditions: superior optical coefficients expressed ashigh solar absorptance (�s > 0�9) in the solar radiationwavelength range (0.29 to 2.5 �m.) and low emittance(�T < 0�1) in the I.R. wavelength range (� > 2�5 �m)along with good stability and low fabrication cost.1�2

The solar absorber for flat plate collectors consists of ahigh thermal conductivity metal plate (copper, aluminium)coated with a selective surface that can be produced bydifferent techniques like: magnetron sputtering,3�4 chemi-cal and physical vapour deposition,5–7 electrodeposition,8�9

etc. Although chemical methods have been widely used,the process involves and produces a lot of chemicalwaste materials. Today vacuum deposition techniques areestablished in the production of the available performingabsorbers.10�11 Table I present the performance of the mostimportant spectrally selective absorbers.12�13

The main drawback of the present solar absorbers is thatthey are rather expensive and energetic intensive whichrise sustainability issues for the final products.

∗Author to whom correspondence should be addressed.

The studies made in this paper are focusing on thedevelopment of solar absorbers based on NiO for flatplate collectors. Selective coatings with improved solarabsorption and thermal emittance were deposited on cop-per substrate using a low cost, up-scalable technique: spraypyrolysis deposition (SPD).14�15 The operational tempera-ture ranges of these materials for solar applications can becategorized as low temperature (T < 100 �C), mid tem-perature (100 �C < T < 400 �C), and high temperature(T > 400 �C). The NiO based solar absorber was designedto be used in flat plate collectors, thus for low and midtemperature. During days with maximum radiation inten-sity (1000 W/m2), maximum stagnation temperatures ofabout 200 �C are reached.13 Structures obtained at deposi-tion temperature (Tdep) higher then the operating tempera-ture or annealed at higher temperatures are expected to bethermally stable for T < Tdep.

Multifunctional nickel oxide thin films have beenintensely studied due to their good optical, electrical andmagnetic properties and chemical stability, making themsuitable for a large number of applications including elec-trochromic display devices,16�17 chemical sensors,18 andselective coatings for solar thermal applications.19�20 Themorphology of the absorber coating plays an importantrole in obtaining high performance materials. This study

J. Nanosci. Nanotechnol. 2008, Vol. 8, No. xx 1533-4880/2008/8/001/006 doi:10.1166/jnn.2008.S345 1

RESEARCHARTICLE

Solar Selective Coatings Based on Nickel Oxide Obtained via Spray Pyrolysis Voinea et al.

Table I. Performance comparison between some spectrally selective absorbers available.12�13

Company Absorber Deposition technique �s �T

Alanod Cr2O3-Cr(NO3)3/Cu Sputtering 0�95±0�01 0�05±0�01Helios Pigment/Stainless steel Painting 0.91 0.25MTI Black chrome/Cu Electroplating 0�96±0�02 0�08±0�03Sunstrip Ni-NixOx-Cu-Ni/ Al Sputtering 0�96±0�02 0�07±0�02Interpane (Sunselect) CrN-CrOx/Cu Sputtering 0.93 0.07TINOX TiNxOx/Cu Evaporation 0�95±0�02 0�05±0�02

presents the results focusing on the optical performance ofthe Cu/CuOx/NiO thin films with surface morphology con-trolled by using polymeric complexing agents into the pre-cursors’ solution. The critical functional properties (solarabsorptance and thermal emittance) of the obtained solarabsorber were determined by using the standard methodISO 12592.2. There is no comprehensive European stan-dardized procedure to test the durability of a solar absorbersurface, but some assessment methods have been pro-posed by the IEA Task X Working Group,21 designatedISO/DIN 12952 and recently by IEA SHC Task 27.22

2. EXPERIMENTAL DETAILS

Thin films of nickel oxide were deposited by SPD ontomicroscopic glass—(G, reference sample) and copper sub-strate (C) from aqueous solutions of Ni(CH3COO)2 ·4H2O 0.2 M (NiAc2, 99%, Acros Organics). Aqua-solublecopolymers of maleic anhydride with controlled hydropho-bia (synthesized at Petru Poni Institute, Romania) wereadded (50 ppm in 50 mL solution) in the precursor sys-tem and their influence on the surface morphology of thesprayed films was investigated.

The glass substrates (1�5× 3 cm pieces of microscopicglass, Heinz Herenz) were cleaned before each depositionin an ultrasonic bath with ethanol, while the copper sub-strate (1�5× 3 cm flat pieces of Cu 99.9%, Beofon) wasmechanically polished (sand paper No. 800).

The substrate temperature (T ) was kept constant dur-ing the deposition (330 �C) using a CERAN 500± 1 �Cceramic plate. Air was used as carrier gas (p = 1�4 bar).The deposition was done in open atmosphere, at a spray-ing angle of 45�. The distance from nozzle to the heatedsubstrate was 20 cm and the spraying sequences numberwas varied from 60, 90 to 120, with a break between twopulses of 30 sec.

The deposition temperature was chosen based on thedifferential scanning calorimetry, DSC (Perkin Elmer,DSC-2). The thermal parameters (�H ) for each thermalevent were calculated. The experiments were carried outin inert atmosphere (N2) from −15 �C to 600 �C in analuminium pan with a heating rate of 20 �C/min. Thefilms structure and composition were investigated using anX-ray diffractometer (Bruker-AXS-D8) with CuK� radia-tion (�= 0�154059 nm), in the range 2� = 2–80�. Surfacemorphology, roughness and microstructural properties of

thin films were examined using Atomic Force Microscopy(AFM, NT-MDT model NTGRA PRIMA EC). The imageswere taken in semi-contact mode with “GOLDEN” sil-icon cantilever (NCSG10, force constant 0.15 N/m, tipradius 10 nm). The solar absorbance (�s) was calculatedfrom reflectance spectra recorded in the wavelength range290–2500 nm, using a Perkin Elmer UV-VIS spectropho-tometer (Lambda 25). The reflectance values in the wave-length 1100–2500 nm were estimated by interpolation. Thethermal emittance (�T) of as-deposited films was deter-mined from the reflectance spectra in the range of 2500 to16500 nm with a Perkin Elmer FT-IR spectrophotometer,model Spectrum BX.

3. RESULTS AND DISCUSSION

3.1. Optimization of the Deposition Temperature

The DSC thermograms and the values of the thermophys-ical parameters (�H ) obtained for nickel acetate powderare presented in Figure 1. The DSC measurements indi-cate that nickel acetate decomposes via two processes ofweight loss (thermal events I and II) over the temperaturerange 95–400 �C:• Process I (95–150 �C): corresponds to dehydration, withan endothermic maximum peak at 113.06 �C, correspond-ing to the process indicated in Eq. (1).

Ni(C2H3O2)2 ·4H2O → Ni(C2H3O2)2 +4H2O (1)

• Process II (330 �C to 400 �C) corresponds to decom-position of NiAc2. In this range, DSC showed two dou-ble endothermic peaks: at 366.29 �C and 375.94 �C.These peaks are related to complex solid-gas-solid reac-tions involving most probably the production of nickeloxide and volatile or gaseous products (Eq. (2)).

Ni(C2H3O2)2 → NiO+3H2O+4CO2 (2)

The results show that nickel oxide was formed from tem-peratures starting with 330 �C and they are in agreementwith the literature data.23–25 The deposition temperaturewas chosen based on the thermal analysis (Tdep = 330 �C).Higher deposition temperatures result in the formation ofnon adherent films of copper and nickel oxides; althoughtemperatures Tdep > 330 �C could have lead to higher NiOcrystallinity degree.

2 J. Nanosci. Nanotechnol. 8, 1–6, 2008

RESEARCHARTICLE

Voinea et al. Solar Selective Coatings Based on Nickel Oxide Obtained via Spray Pyrolysis

Fig. 1. DSC curves for a freeze-dried nickel(II) acetate tetrahydrate powder.

3.2. Optimisation of the Precursor Concentration

After selecting the Tdep, the solution concentration in therange of 0.05–0.2 M was chosen in the limits of theroom temperature dissolution of NiAc2 in water, withoutblocking spray-nozzle due to precursor crystallites. Duringthe optimisation steps, reference samples (on glass) weredeveloped, for avoiding the interference of results withthose provided by the metal substrate. The other depositionparameters were kept constant. After spraying the filmswere immediately removed from the heater and allowed tocool in air.

The optimisation criteria were based on the solarabsorptance (�s) and the thermal emittance (�T) whichwere calculated from the reflectance spectra according toDuffie et al.26 The approximate film thickness was calcu-lated based on the UV-VIS reflectance spectra using therelation:

d = N�1�2

2��2 −�1� · �n2 − sin2 ��1/2[nm] (3)

where:N = number of the fringes in the wavelength �1 to �2

(�1 < �2);n= 2�1818 (the refractive index of NiO);27

�= angle of incidence (6�).The results are presented in Table II. As the concen-

tration increases, the amount of Ni2+ that participatesin nickel oxide films formation increases and the filmthickness is higher, confirming that the growth repre-sents the predominant step in the film. Higher concen-trations also favour better values for �T and �s, due tofilm homogeneity and increased heat storage capacity,thus the optimum concentration was chosen c = 0�2 M.The layer thickness doesn’t exhibit a linear variation withthe number of spraying sequences, proving that structural

re-organizations and compacting are parallel process withgrowth.

3.3. Optimisation of the SprayingSequences Number

Thin films of variable thickness were grown on glass andalso copper substrate by varying the spraying sequencesnumber (Table II). By increasing the value of nsp moreuniform and thicker films are obtained.

The deposition parameters for sample G6, with the bestperformance in terms of �s and �T, were applied on cop-per substrate (sample C_test), with very good results (�s =0�93 and �T = 0�07). Higher values for nsp were also stud-ied for copper substrate. However, large spraying periods(over 90 min for nsp > 120) lead to non adherent films.The explanation is related to the concomitant formation ofa tick enough copper oxide film, due to the thermal oxida-tion of the substrate. For nsp = 120 there is no significantdifferences comparing the performance of the two films.

Table II. The �s and �T values function of the deposition parametersvaried for optimising.

c Tdep d

Sample [mol/L] [�C] nsp [nm] �s �T

(a) The precursor concentrationG1 0�05 330 30 254 0�74 0�25G2 0�1 330 30 317 0�81 0�17G3 0�15 330 30 392 0�86 0�14G4 0�2 330 30 425 0�89 0�1

(b) The sprayingsequences number

G5 0�2 330 60 0.481 0�90 0�17G6 0�2 330 90 0.532 0�92 0�13G7 0�2 330 120 0.510 0�91 0�14C_test 0�2 330 90 0.561 0�93 0�07

J. Nanosci. Nanotechnol. 8, 1–6, 2008 3

RESEARCHARTICLE

Solar Selective Coatings Based on Nickel Oxide Obtained via Spray Pyrolysis Voinea et al.

(a) (b)

Fig. 2. Structures of the maleic anhydride copolymers: (a) hydrophilicform (HFL) and (b) hydrophobic form (HFB).

Therefore, the optimum number for the spraying sequenceswas chosen to be 90.

3.4. The Influence of the ComplexingAgents on the Surface Morphology andon the �s and �T Values

Morphology controlling agents were added in the spray-ing solutions: hydrophobic (HFB) and hydrophilic (HFL)copolymers of maleic anhydride, Figure 2.

Depending on their textural organization, maleic anhy-dride copolymers and their derived sodium form are usedin several applications such as surface control agents, soilconditioners, drug carriers, or biomaterials, and the num-ber is continuously growing.28–30 They can have variablehydrophobic character which, in aqueous solvent, resultsin variable conformations (macromolecular stretched chainor macromolecular coil). The interaction of the polymerwith the nickel precursors depends on the conformationand it is increased by the relaxed configurations, i.e., whenusing hydrophilic polymers. Strong polymer–nickel inter-actions have as direct consequence lower nucleation and/orgrowth rates and, morphologically speaking, smoothersurfaces.31

The influence of both hydrophilic and hydrophobic formof the copolymers in controlling the surface morphol-ogy and their effect on the solar absorptance and ther-mal emittance was studied. Thin layers of nickel oxidewere deposited on copper substrate using the previouslyoptimised conditions, with and without additives in theprecursor solutions. The results of the film thickness,solar absorptance and thermal emittance are presented inTable III. The films are slightly thinner in the case of usingcopolymers.

3.5. Structural Characterization

The crystal structures of the samples were investigated byX-ray diffraction, Figures 3–5. The XRD spectra recorded

Table III. The influence of the complexing agents on the �s and �T

values.

c Tdep d

Sample [mol/L] [�C] nsp Additive [nm] �s �T

C1 0�2 330 90 — 0�548 0�92 0�07C2 0�2 330 90 Hfl 0�461 0�95 0�05C3 0�2 330 90 Hfb 0�432 0�94 0�06

10 20 30 40 50 60 70 80

**

****

** - NiO cubic

Inte

nsity

[a.u

.]

2θ [degree]

Fig. 3. XRD patterns for the sample obtained on glass substrate.

for the thin layers deposited on microglass substrate(Fig. 3) show that NiO cubic with low crystallinity wasformed (ICCD, PDF 47-1049). The peaks corresponds tothe {111}, {200}, {220} plans and the grain size calculatedwith Scherrer’s formula was about 4–10 nm. The latticeparameters for NiO were: a= b= c= 0�417 nm, in agree-ment with the literature values (0.416 nm32), showing thatno distortion affects the unit cell.

The XRD pattern of the samples obtained on coppersubstrate indicates that all the films deposited at 330 �Care containing mainly crystalline CuO (ICCD, PDF 80-0076), Cu2O (ICCD, PDF 74-1230) and amorphous nickeloxide (Figs. 4–6). The crystalline phase of NiO couldnot be marked out in the case of copper substrate sam-ples, although the films were obtained in the same condi-tions, only with different substrate (glass and copper). Thediffraction lines for NiO were most probably too smallcompared with the copper and copper oxides signal.

The XRD spectra of the samples obtained on coppersubstrate using complexing agents show more crystallinestructures (crystallite size of 20–27 nm) then in the case

0 10 20 30 40 50 60 70 80

^ #

##

* ^

^ - Cu cubic* - CuO monoclinic# - Cu2O cubic

Inte

nsity

[a.u

.]

2θ [degree]

Fig. 4. XRD patterns for the sample obtained on copper substratewithout any addition.

4 J. Nanosci. Nanotechnol. 8, 1–6, 2008

RESEARCHARTICLE

Voinea et al. Solar Selective Coatings Based on Nickel Oxide Obtained via Spray Pyrolysis

0 10 20 30 40 50 60 70 80

^^

*

#

#

^

*^

^

#

^

#

##

*

^ - Cu cubic* - CuO monoclinic

# - Cu2O cubic

Inte

nsity

[a.u

.]

2θ [degree]

Fig. 5. XRD patterns for the sample obtained on copper substrate andHfb addition.

with no copolymers (crystallite size of 13–16 nm). Thethermal emittance is higher for these samples, Table III.

3.6. Surface Morphology Characterization

Atomic force microscopy was used for visualizing the sur-face nano-texture and for measuring the nanometer scalesurface roughness. The two-dimensional AFM images arepresented in Figures 7–9. The thin layers obtained by SPDon copper substrate using the optimised conditions are rel-atively uniform, with different values of the surface rough-ness, depending on the additive (Hfl, Hfb or none).

The surface texture and roughness can alter or have apositive effect on the optical properties of the thin filmsobtained. The mean roughness of the samples depositedfrom a precursor solution containing copolymers of maleicanhydride (C2 and C3) is lower (28.3 nm and 41.9 nm,respectively) than the roughness of sample C1 (64.7 nm),obtained in the same deposition conditions, but withoutcomplexing agent.

0 10 20 30 40 50 60 70 80

^^

*

#

#

^

*^

^

#

^

#

##

*

^ - Cu cubic* - CuO monoclinic# - Cu2O cubic

Inte

nsity

[a.u

.]

2θ [degree]

Fig. 6. XRD patterns for the sample obtained on copper substrate andHfl addition.

Fig. 7. AFM image of the sample obtained on copper substrate withoutany addition (5×5 �m); mean surface roughness: 94.7 nm.

Fig. 8. AFM image of the sample obtained on copper substrate and Hfbaddition (5×5 �m); mean surface roughness: 41.9 nm.

Fig. 9. AFM image of the sample obtained on copper substrate and Hfladdition (5×5 �m); mean surface roughness: 28.3 nm.

J. Nanosci. Nanotechnol. 8, 1–6, 2008 5

RESEARCHARTICLE

Solar Selective Coatings Based on Nickel Oxide Obtained via Spray Pyrolysis Voinea et al.

4. CONCLUSIONS

Thin films of nickel oxide were obtained by spray pyrolysisto be used as selective coatings for flat plate collectors. Inthe first step the deposition conditions were optimised onglass substrate, and then the results were applied for cop-per substrate. Moreover, the influence of two copolymersof maleic anhydride with controlled hydrophobia on thesurface morphology and on the solar absorptance/thermalemittance was studied.

The structural analysis showed the formation of NiO onglass substrate at 330 �C, with the crystallite size of about4–10 nm. For the films deposited in the same conditions,on copper substrate, only crystalline copper and copperoxides formed due to the thermal oxidation of the substratewere observed. The diffraction lines corresponding to NiOwere most probably too small compared with copper andcopper oxides signal or correspond to a totally amor-phous phase. The XRD spectra of the samples obtainedon copper substrate using complexing agents showedmore crystalline structures (crystallite size of 20–27 nm)then in the case with no copolymers (crystallite size of13–16 nm).

The copolymers are influencing the layer formationprocess, influencing the nucleation rate and leading tosmoother films. The mean roughness of the samplesdeposited from a precursor solution containing copolymersof maleic anhydride is lower (41.9 nm for Hfb and 28.3 nmfor Hfl) than the roughness of the sample obtained in thesame deposition conditions, but without complexing agent(94.7 nm).

The solar absorber obtained Cu/CuOx/NiO shows goodvalues for the solar absorptance and thermal emittance(�s = 0�95, �T = 0�05 for Hfl and �s = 0�94, �T = 0�06for Hfb) when copolymers of the maleic anhydride wereused. The results recommend the use of these copolymersin tailoring the surface of solar selective coatings.

Acknowledgments: The research work was supportedby the CEEX program MATSOL 277/2006 and CNCSISTd 312/2007.

References and Notes

1. P. Oelhafen and A. Schüler, Solar Energy 79, 110 (2005).2. S. Brunold, U. Frei, B. Carlsson, K. Moller, and M. Kohl, Solar

Energy 68, 313 (2000).3. S. Zhao, C.-G. Ribbing, and E. Wäckelgård, Solar Energy 78, 125

(2005).

4. H. C. Barshilia, N. Selvakumar, K. S. Rajam, and A. Biswas, Sol.Energy Mater. Sol. Cells 92, 495 (2008).

5. J. Boudaden, P. Oelhafen, A. Schüler, C. Roecker, and J.-L.Scartezzini, Sol. Energy Mater. Sol. Cells 89, 209 (2005).

6. C. Nunes, V. Teixeira, M. Collares-Pereira, A. Monteiro, E. Roman,and J. M. Gago, Vacuum 67, 623 (2002).

7. Y. P. Sukhorukov, Technical Physics Letter 32, 132 (2006).8. C. N. Tharamani and S. M. Mayanna, Sol. Energy Mater. Sol. Cells

91, 664 (2007).9. T. K. Lee, D. H. Kim, and P. Chungmoo Auh, Sol. Energy Mater.

Sol. Cells 29, 149 (1993).10. M. Farooq and M. G. Hutchins, Sol. Energy Mater. Sol. Cells 71,

523 (2002).11. A. Schüler, C. Roecker, J. L. Scartezzini, J. Boudaden, I. R.

Videnovic, R. S. C. Ho, and P. Oelhafen, Sol. Energy Mater. Sol.Cells 84, 241 (2004).

12. K. Gelin, Preparation and characterisation of sputter depositedspectrally selective solar absorbers, Ph.D. Thesis, Uppsala (2004),pp. 19–22.

13. C. E. Kennedy, Review of mid- to high-temperature solar selectiveabsorber materials, NREL Technical Report NREL/TP-520-31267(2002).

14. A. Duta, S. A. Manolache, M. Nanu, A. Goossens, andJ. Schoonman, Thin Solid Films 511–512, 195 (2006).

15. L. Andronic, S. Manolache, and A. Duta, Journal of Optoelectronicsand Advanced Materials 5, 1403 (2007).

16. T. Maruyama and S. Arai, Sol. Energy Mater. Sol. Cells 30, 257(1993).

17. E. O. Zayim, I. Turhan, F. Z. Tepehan, and N. Ozer, Sol. EnergyMater. Sol. Cells 92, 164 (2008).

18. C. Cantalini, M. Post, D. Buso, M. Guglielmi, and A. Martucci,Sensors and Actuators B: Chemical 108, 184 (2005).

19. M. Voinea and A. Duta, Journal of Optoelectronics and AdvancedMaterials 9, 1451 (2007).

20. S. Zhao, E. Avendaño, K. Gelin, J. Lu, and Ewa Wäckelgård, Sol.Energy Mater. Sol. Cells 90, 308 (2006).

21. M. Köhl, Sol. Energy Mater. Sol. Cells 84, 275 (2004).22. www.iea-shc.org/task27/index.html.23. J. Hong, G. Guo, and K. Zhang, J. Anal. Appl. Pyrolysis 77, 111

(2006).24. J. D. Desai, S.-K. Min, K.-D. Jung, and O.-S. Joo, Appl. Surf. Sci.

253, 1781 (2006).25. J. C. De Jesus, I. González, A. Quevedo, and T. Puerta, J. Mol.

Catal. A: Chem. 228, 283 (2005).26. J. A. Duffie and W. A. Beckman, Solar Engineering of Thermal Pro-

cesses, 2nd edn., Wiley Interscience, New York (1991), pp. 196–197.27. D. R. Lide, Handbook Chemistry and Physics, 57th edn., edited by

R. C. Weast, CRC Press, Cleveland, Ohio (1976), p. E-219.28. G. Aldea, H. Gutiérrez, J.-M. Nunzi, G. C. Chitanu, M. Sylla, and

B. C. Simionescu, Opt. Mater. 29, 1640 (2007).29. M. Treeby, G. C. Chitanu, and K. Kogej, J. Colloid Interface Sci.

288, 280 (2005).30. E. Bacu, G. C. Chitanu, A. Couture, P. Grandclaudon, Gh. Singurel,

and A. Carpov, Eur. Polym. J. 38, 1509 (2002).31. E. Purghel, M. Voinea, L. Isac, and A. Duta, Rev. Chim. (Bucuresti)

59, 469 (2008).32. S. A. Mahmouda, A. A. Akla, H. Kamalb, and K. Abdel-Hady, Phys-

ica B 311, 366 (2002).

Received: 24 January 2008. Accepted: 23 March 2008.

6 J. Nanosci. Nanotechnol. 8, 1–6, 2008