Dynamic Article LinksC<CrystEngComm

Cite this: CrystEngComm, 2012, 14, 1825

www.rsc.org/crystengcomm PAPER

Dow

nloa

ded

by U

nive

rsity

of

Ten

ness

ee a

t Kno

xvill

e on

02/

04/2

013

09:0

7:06

. Pu

blis

hed

on 0

9 Ja

nuar

y 20

12 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

E05

756A

View Article Online / Journal Homepage / Table of Contents for this issue

In situ preparation of CuInS2 films on a flexible copper foil and theirapplication in thin film solar cells†

Minghua Tang,a Qiwei Tian,a Xianghua Hu,a Yanling Peng,a Yafang Xue,a Zhigang Chen,*a Jianmao Yang,b

Xiaofeng Xu*c and Junqing Hu*a

Received 21st June 2011, Accepted 11th November 2011

DOI: 10.1039/c1ce05756a

The in situ preparation of semiconductor films on a flexible metal foil has attracted increasing attention

for constructing flexible solar cells. In this work, we have developed an in situ growth strategy for

preparing CuInS2 (CIS) films by solvothermally treating flexible Cu foil in an ethylene glycol solution

containing InCl3$4H2O and thioacetamide with a concentration ratio of 1 : 2. The effects of

solvothermal temperature, time and concentration on the morphology and phase of the CIS films are

investigated. Solvothermal temperature has no obvious effect on the morphology of the final films, but

higher temperature is favorable for the growth of CIS films with higher crystallinity. Reactant

concentration plays a significant role in controlling the morphology of CIS films; if InCl3$4H2O

concentration is relatively low (#0.042 M), single-layered CIS films can be produced, which are

composed of high ordered potato chips shaped nanosheets, otherwise, it prefers to form a double-

layered film, for which the lower layer is similar CIS ordered nanosheets while the upper layer is

composed of flower shaped superstructures. A possible mechanism of the CIS films is also investigated.

UV-vis measurements show that all these CIS films possess a direct bandgap energy of 1.48 eV,

appropriate for the absorption of the solar spectrum. Finally, single-layered CIS films on Cu foil were

employed for fabricating flexible solar cells with a structure of Cu foil/CuInS2/CdS/i–ZnO/ITO/Ni–Al,

and the resulting cells yield a power conversion efficiency of 0.75%. Further improvement of the

efficiencies of the solar cells can be expected by optimizing the morphology, structure and composition

of the CIS films, as well as the fabrication technique.

1. Introduction

The quest and demand for clean and economical energy sources

have increased widespread interest in the development of solar

applications. In particular, direct conversion of solar energy to

electrical energy by using semiconductor photoelectrodes has

attracted much attention for many decades. Among various

semiconductor materials used in solar cells, CuInS2 (CIS) and

related I–III–VI2 chalcopyrite compounds have garnered consid-

erable interest due to their myriad benefits,1–5 including large

absorption coefficients, adjustable band gap, low toxicity and

long-term stability. CIS-based film solar cells have exhibited

a confirmed conversion efficiency of about 12.5%,6 but they are

aState Key Laboratory for Modification of Chemical Fibers and PolymerMaterials, College of Materials Science and Engineering, DonghuaUniversity, Shanghai, 201620, China. E-mail: hu.junqing@dhu.edu.cn;zgchen@dhu.edu.cnbResearch Center for Analysis and Measurement, Donghua University,Shanghai, 201620, ChinacDepartment of Applied Physics, Donghua University, Shanghai, 201620,China. E-mail: xxf@dhu.edu.cn

† Electronic supplementary information (ESI) available. See DOI:10.1039/c1ce05756a

This journal is ª The Royal Society of Chemistry 2012

mostly fabricated via a high-cost vacuum-based process such as

evaporation7,8 and sputtering9–11 on Mo coated glass substrates.

Sophisticated vacuum-based set-ups and control systems require

large capital investment, thereby hindering the emerging CIS-

type solar cells from commercial utilization. In addition, the use

of glass substrates in solar cells can raise some problems such as

inflexibility and fragility, leading to much inconvenience in

transport and installation.12

To settle these problems, many efforts have been devoted to

developing alternative deposition techniques for thin film CIS

solar cells via non- or low-vacuum processes. They can be

roughly divided into two categories depending on precursor

types, in which the first approach uses nanoparticle-based

precursors13,14 and the other uses the solution type precur-

sors.15–19 On the one hand, the nanoparticle-based precursors are

based on the preparation of CIS and related I–III–VI2 chalcopyrite

nanocrystal inks.20,21 Nanocrystal inks offer the advantages of

dispersibility in organic solvents for printing CIS and related

I–III–VI2 chalcopyrite films on inflexible and/or flexible substrates,

but a complicated synthesis of nanocrystals is necessary and the

presence of organic ligands can leave residues that hurt device

performance.13 On the other hand, the solution type precursors

can be used to prepare CIS and related I–III–VI2 chalcopyrite film

CrystEngComm, 2012, 14, 1825–1832 | 1825

Dow

nloa

ded

by U

nive

rsity

of

Ten

ness

ee a

t Kno

xvill

e on

02/

04/2

013

09:0

7:06

. Pu

blis

hed

on 0

9 Ja

nuar

y 20

12 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

E05

756A

View Article Online

by spray pyrolysis,22,23 electrochemical deposition24–26 and direct

coating-sintering.27–29 Among these deposition techniques, direct

coating-sintering method is the most promising for the produc-

tion of low-cost and high efficiency solar cells. Mitzi et al.17,18

utilized hydrazine as the solvent for dissolving metal chalco-

genides as ‘‘precursor ink’’, and then spin-casted onto the Mo

coated glass substrate to fabricate Cu(In,Ga)Se2 and CuIn(Se,S)2based solar cells, yielding a conversion efficiency of 10.3% and

12.2%, respectively. In addition, Moses16 and Cui19 developed,

respectively, a molecular-based precursor ink with 1-butylamine

as a solvent and an easily decomposable vulcanized polymeric

ink with pyridine as a solvent to prepare CIS films on a Mo

coated glass substrate, and the resulting CIS solar cells exhibited

conversion efficiencies of about 4% and 2.15%, respectively.

However, although direct coating-sintering method is a very

exciting approach, it usually involves environmentally unfriendly

solvent (such as hydrazine and pyridine) and a high temperature

(250–550 �C) sintering process, which are considered to be an

obstacle for the large-scale profitable operation and/or the

fabrication of solar cells. Therefore, it is still desirable to develop

novel methods to prepare CIS and related I–III–VI2 chalcopyrite

films for solar cells.

It is well known that the rigid substrate materials for the solar

cells have been limited due to mechanical brittleness and diffi-

cult processing, such as silicon, indium tin oxide (ITO) and

fluorine-doped tin oxide (FTO).30–32 The flexible solar cells have

the advantages of light weight and foldability, and could be

used on the curved surfaces of many buildings and instruments.

Therefore, the development of solar cells based on the flexible

substrates has become another hot focus following traditional

cells. Very early, the flexibilization of silicon solar cells was

proposed to reduce their thicknesses for the purpose of the

flexible silicon solar arrays.33 Recently, in spite of a good flexible

performance of the polymer substrates for the organic/inorganic

hybrid solar cells, there are also worse problems in the practical

applications, such as poor thermal stability, low melting point,

and weak bonding between the substrates and as-grown mate-

rials.34,35 Hence, it is still meaningful to explore the solar cells

based on flexible substrate materials for their wide applications.

In particular, in situ preparation of semiconductor films on

flexible metal foils has attracted much attention.36 For example,

different semiconductor films, including TiO2 nanotube or

nanowire films on a Ti foil37,38 and ZnO nanowire films on a Zn

foil,39 have been successfully obtained for different kinds of

solar cells. In addition, some groups also prepared Cu2O

nanostructures,40 Cu(OH)2 and CuO nanoribbon arrays41 on Cu

foils, and our group recently reported the in situ preparation of

well-aligned Cu2-xSe nanostructures on Cu foils.42 Although

some chalcogenide semiconductor films have been grown

directly on Cu foil or other metal surfaces by a solvothermal

route,43–47 there are few reports on the preparation of CIS and

related I–III–VI2 chalcopyrite films on Cu foils for flexible solar

cells.

Herein, we present a simple route for the in situ growth of

CIS thin films on the flexible Cu foil, thereby circumventing

a complicated film fabrication process as above mentioned and

employing toxic materials, in which the effects of reaction

temperature, time and concentration on the morphology and

phase of CIS films were carefully investigated. Single-layered

1826 | CrystEngComm, 2012, 14, 1825–1832

and double-layered CIS films are obtained, and the solar cells

with the structure of the flexible Cu foil/CuInS2/CdS/i–ZnO/

ITO/Ni–Al are fabricated, and the solar cells based on the

single-layered structures of CIS films exhibit a conversion effi-

ciency of 0.75%.

2. Experimental section

2.1. CIS thin film growth

All reagents were analytical grade and used without further

purification. In a typical synthetic process, 0.6 mmol of thio-

acetamide was added into 12 mL of ethylene glycol solution

containing InCl3$4H2O (0.025 M) under magnetic stirring,

forming a clear and colorless solution. The resulting solution was

transferred into a Teflon-lined stainless steel autoclave with

30 mL capacity. Subsequently, a piece of Cu foil, which had been

ultrasonically cleaned in a 3.0 M HCl aqueous solution, ethanol

and distilled water for 10 min, respectively, was vertically and

partly (� 3/4 of the foil length) immersed into the solution. The

unsubmerged part of the Cu foil is kept to have a clean surface

and act as the anode of the later fabricated solar cell device.

Lastly, the autoclave was kept in a fan-forced oven at 180 �C for

16 h. After being air-cooled to room temperature, the CIS film

deposited foil was washed with deionized water and absolute

ethanol successively, and then dried in air. For comparison, the

effects of solvothermal temperature (120–180 �C), reaction time

(40 min–16 h), and InCl3$4H2O concentration (0.025–0.083 M)

on the CIS films’ growth were investigated. The concentration

ratio of InCl3$4H2O and thioacetamide was maintained as

a constant (1 : 2) for all the cases.

2.2. Fabrication of solar cells

CIS film solar cells were fabricated as follows. CIS films (0.5 �0.8 cm2) on the flexible Cu foils (� 0.1 cm thickness) were used as

the absorber layer. The n-type junction partner CdS (� 50 nm

thickness) was deposited on the CIS films by using a chemical

bath approach, described as follows: the bath solution used in

this process was composed of 30 mL of deionized H2O, 6.5 mL of

NH4OH solution (25–28 wt%), 5 mL of CdCl2 (0.015 M), and 10

mL of NH2CSNH2 (0.375 M), which are mixed at room

temperature. The as-formed bath solution together with the CIS

film sample was transferred to a water-heated vessel, which was

kept at 80 �C and constantly stirred by a magnetic chuck during

the deposition process. After 5 min, the CIS film sample was

taken out of the vessel, and washed with deionized water and

absolute ethanol, successively. After that, two magnetron sput-

tering processes were utilized to deposit layers of intrinsic ZnO

(� 70 nm in thickness) and transparent conductive indium tin

oxide (ITO) (� 180 nm in thickness), respectively, on the CdS

layer. Finally, a Ni (50 nm)/Al (3 mm) dual layer top contact grid

was evaporated through a metal mask to facilitate collecting the

photogenerated carriers as the cathode in the as-fabricated solar

cell device. The resulting solar cell device (total area: � 0.4 cm2)

with a structure of Cu foil/CuInS2/CdS/i–ZnO/ITO/Ni–Al was

defined by isolating the multilayer film areas via mechanical

scribing.

This journal is ª The Royal Society of Chemistry 2012

Dow

nloa

ded

by U

nive

rsity

of

Ten

ness

ee a

t Kno

xvill

e on

02/

04/2

013

09:0

7:06

. Pu

blis

hed

on 0

9 Ja

nuar

y 20

12 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

E05

756A

View Article Online

2.3. Characterization and photoelectrical measurement

The morphology and phase of CIS films were investigated by

using scanning electron microscopy (SEM; Hitachi S-4800) and

powder X-ray diffraction (XRD; Rigaku D/max-g B) with

a Cu-Ka radiation source (l ¼ 1.5418 �A). Structural analysis of

the CIS films was carried out using transmission electron

microscopy (TEM; JEM-2010F, at 200 kV). The samples for

TEM characterization were prepared as follows: a drop of

dispersing solution of the CIS material scraped from the Cu foils

was placed on a Cu mesh coated with an amorphous carbon film,

followed by evaporation of the solvent. An UV-vis spectropho-

tometer (Lambda 950 Spectrophotometer) was directly used to

carry out the optical measurements of CIS films on the flexible

Cu foils.

The photocurrent density–voltage curves of the cells were

recorded under standard AM 1.5 solar illumination with an

intensity of 80 mW cm�2 using a computerized Keithley Model

2400 SourceMeter unit. A 300 W xenon lamp (Newport Oriel)

served as the light source.

3. Results and discussion

3.1. Structure and morphology of CIS films

Using a one-step solvothermal route, we have fabricated a series

of in situ grown CIS films on the flexible Cu foil. The effects of the

solvothermal temperature, time and reactant concentration on

the morphology and phase of the CIS films have been system-

atically studied as follows.

3.1.1. Effect of solvothermal temperature. The CIS films were

prepared by treating Cu foils in ethylene glycol solution con-

taining InCl3$4H2O (0.025 M) and thioacetamide (0.050 M) at

different temperatures (120, 140, 160, 180 �C) for 16 h. Fig. 1

Fig. 1 The XRD patterns of the CIS films obtained by treating Cu foils

in an ethylene glycol solution containing InCl3$4H2O (0.025 M) and

thioacetamide (0.05 M) at 120, 140, 160, and 180 �C for 16 h in

an autoclave, and the standard powders of CIS from JCPDS card

(no. 85-1575).

This journal is ª The Royal Society of Chemistry 2012

shows the XRD patterns of the resulting CIS films. Besides those

existing peaks from the Cu foil (marked by triangles), the

diffraction peaks at 27.8, 46.5 and 55.1� are assigned to (112),

(204)/(220) and (312)/(116) planes of the CIS material, respec-

tively, which are consistent with those in the literature13 and from

the JCPDS card (no. 85-1575). This fact confirms that all films

prepared at the above temperatures are chalcopyrite CuInS2.

With the temperature increasing to 180 �C, the intensities of threemain peaks of (112), (204) and (312) planes are significantly

strong, suggesting that the CIS films grown at this temperature

have higher crystallinity, compared to those grown at 120, 140,

and 160 �C, as later revealed by the HRTEM imaging (Fig. 2d).

A peak at � 26.2� in the XRD patterns could be attributed to

that from Cu9S5 (JCPDS card no 47-1748). It is suggested that

with the increase of reaction temperature, a small quantity of

Cu9S5 forms, which is also a p-type semiconductor for photo-

voltaic applications.

The morphologies of the samples prepared at different

temperatures were characterized by SEM. Fig. 2 gives the typical

morphologies of the CIS films prepared at 120 and 180 �C. When

the reaction temperature is 120 �C, a large amount of highly

ordered potato chips shaped CIS nanosheet arrays are densely

packed and uniformly covered over the entire surface of the Cu

foil (Fig. 2a), which should be attributed to Cu foil directly

serving as the copper source, thus leading to in situ growth of CIS

nanosheet alignments. A high-magnification SEM image (inset in

Fig. 2a) shows that individual nanosheets within this CIS film

display a crooked shape with a thickness of�50 nm and length of

�4 mm. These CIS potato chips shaped nanosheets with rough

surfaces are assembled and inter-meshed with each other,

Fig. 2 SEM images show the CIS films obtained at (a) 120 �C and (b)

180 �C after 16 h by treating Cu foils in ethylene glycol solution con-

taining InCl3$4H2O (0.025 M) and thioacetamide (0.05 M) in the auto-

clave, (c) cross sectional SEM image and (d) HRTEM image of the

sample shown in (b). (All upper-right insets are the corresponding

magnification SEM images).

CrystEngComm, 2012, 14, 1825–1832 | 1827

Dow

nloa

ded

by U

nive

rsity

of

Ten

ness

ee a

t Kno

xvill

e on

02/

04/2

013

09:0

7:06

. Pu

blis

hed

on 0

9 Ja

nuar

y 20

12 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

E05

756A

View Article Online

forming a continuous net-like flat film. It is found that the sol-

vothermal reaction temperature has no significant effect on the

morphology of the final films. As the reaction temperature is

elevated to 180 �C, keeping the other reaction conditions

constant, the net-like film is composed of a great quantity of

carpenterworm shaped CIS nanosheet arrays with a little higher

density than that grown at 120 �C (Fig. 2b). Each nanosheet is

further grown and well aligned in a denser array that is nearly

perpendicular to the Cu substrate surface, resulting in uniform

and densely aligned net-like film. Simultaneously, some peeled

fragments (a substructure of the net-like film) can also be found,

as shown in Fig. S1 (see ESI†), which further indicates the good

quality of these CIS nanosheets with a clear square laminar

shape. A cross-sectional SEM image, Fig. 2c, reveals that the CIS

film (grown at 180 �C) composed of carpenterworm shaped

nanosheets is single-layered with a large grain size and a thick-

ness of �1.5 mm. From the HRTEM image (Fig. 2d), the lattice

spacing was measured to be �3.2 �A, corresponding to the (112)

planes of the chalcopyrite CuInS2 crystal, further indicating the

well-crystallized single crystals of these nanosheets within the

as-grown CIS film.

To further investigate the composition of the CIS film, the

energy dispersive X-ray spectroscope (EDX) spectrum (Fig. 3)

was obtained from nanosheets scraped from the Cu foil. Only

Cu, In and S signal peaks are observed, suggesting that the film

consists of the elements of Cu, In and S. Quantitative EDX

analysis indicates the atomic ratio of Cu, In and S is 1.6 : 1 : 2.2,

indicating that the composition of the as-synthesized product is

stoichiometric Cu1.6InS2.2. A copper/sulfur-rich CIS film should

be attributed to the rapid dissolution, diffusion and reaction

(with S) of Cu originating from the Cu foils in ethylene glycol

solution at high temperature (180 �C).

3.1.2. Effect of reactant concentration. The effect of reactant

concentration on the morphology and structure of the as-

prepared CIS films has been studied by varying the concentration

of InCl3$4H2O from 0.025M to 0.083M at 180 �C for 16 h, while

the concentration ratio of InCl3$4H2O and thioacetamide was

Fig. 3 An EDX spectrum of the CIS film obtained by treating Cu foil in

an ethylene glycol solution containing InCl3$4H2O (0.025 M) and thio-

acetamide (0.05 M) at 180 �C for 16 h in the autoclave, the quantitative

EDX analysis is shown in the inset.

1828 | CrystEngComm, 2012, 14, 1825–1832

kept as a constant of 1 : 2. It is found that the reactant concen-

tration plays a significant role in the growth of the CIS films.

When the concentration of InCl3$4H2O is low (0.025M), the CIS

film is uniform and composed of highly ordered carpenterworm

shaped nanosheets (Fig. 2b). As the concentration increases to

0.042 M (Fig. 4a), the entire view of the film is not even and is

embedded with some bumps, appearing as micro-scaled hierar-

chical superstructures with an average diameter of �2.6 mm. In

fact, each hierarchical superstructure is completely assembled by

branched nanoplates, which are relatively smooth and crooked

and have an average thickness of �120 nm. With the further

increase of the concentration to 0.058 M (Fig. 4b), there are

numerous ball shaped assemblies on the surface of the flat CIS

film. Each ball-like assembly is also a hierarchical micro-scaled

superstructure, i.e., it is composed of many two-dimensional

vertically standing and closely aligned nanoplates with an

average thickness of �75 nm. When the concentration is as high

as 0.083 M (Fig. 4c), flower shaped architectures with a diameter

of 1.5–3 mm were spread over the whole film, whose ‘‘petals’’ are

ultrathin with an average thickness of �20 nm and aligned

perpendicularly to the spherical surface with clearly oriented

layers, pointing toward a common center. In addition, many

pores with different sizes can be found among spherical super-

structures and their ‘‘petals’’. To further confirm the structure of

this sample, a cross-sectional SEM image (Fig. 4d), taken from

a fracture of a broken Cu foil, clearly reveals the CIS film is

double-layered. The upper layer is made up of many super-

structures with flower shaped morphology, aggregating with

each other intensely; the lower layer is a similar flat net-like film,

Fig. 4 SEM images show the CIS films obtained by treating Cu foils in

ethylene glycol solution containing different concentrations of

InCl3$4H2O (a) 0.042 M, (b) 0.058 M, and (c) 0.083 M at 180 �C for 16 h

in the autoclave, (d) cross sectional SEM image of the sample shown in

(c). (All upper-right insets are corresponding magnification SEM

images).

This journal is ª The Royal Society of Chemistry 2012

Dow

nloa

ded

by U

nive

rsity

of

Ten

ness

ee a

t Kno

xvill

e on

02/

04/2

013

09:0

7:06

. Pu

blis

hed

on 0

9 Ja

nuar

y 20

12 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

E05

756A

View Article Online

where the composed nanosheets partly dissolve into bulks.

The whole thickness of the double layers is as high as �6.5 mm.

3.1.3. Effect of reaction time and growth mechanism.We have

also carried out a series of time-dependent experiments to

examine a possible growth mechanism of the CIS films, which

were prepared by treating Cu foils in ethylene glycol solution

containing InCl3$4H2O (0.025 M) and thioacetamide (0.05 M) at

180 �C for different reaction times (2 h, 8 h, and 16 h). It is found

that the reaction time shows a significant effect on the thickness,

in particular, the density of the ordered nanosheets of the single-

layered CIS film on the Cu foil. In a short time (e.g., 2 h), only

a small amount of the CIS nanosheets grow loosely on the Cu

foil, and the thickness of these nanosheets is estimated to be �30

nm (Fig. S2a, see ESI†). When the reaction time is increased to 8

h, a great quantity of the CIS nanosheets appear on the Cu foil,

and the density of the nanosheets obviously increases (Fig. S2b,

see ESI†). With the reaction time further increased to 16 h, the

thickness and the density of the CIS nanosheets both further

increase to a higher degree (Fig. S2c, the nanosheets with

a thickness of �130 nm†).

To further understand the growth mechanism of the double-

layered CIS film, a medium concentration of InCl3$4H2O (e.g.

0.05 M) was adopted, since high concentration (such as 0.083 M)

will result in a rapid growth and thus it is difficult to observe the

intermediate state. Fig. 5 gives the surface morphologies of the

CIS films prepared by treating Cu foils in ethylene glycol solution

containing InCl3$4H2O (0.05 M) and thioacetamide (0.10 M) at

180 �C for different reaction times. At a short time (40 min,

Fig. 5a), the net-like film was produced, but it is composed of

Fig. 5 SEM images show the CIS films obtained after (a) 40 min, (b) 2 h,

(c) 8 h and (d) 12 h by treating Cu foils in ethylene glycol solution con-

taining InCl3$4H2O (0.05M) and thioacetamide (0.10M) at 180 �C in the

autoclave. (All upper-right insets are corresponding magnification SEM

images).

This journal is ª The Royal Society of Chemistry 2012

many thin nanoplates with an average thickness of �15 nm,

while each nanoplate intermeshes loosely, leaving many pores on

the Cu foil. With further increasing the reaction time to 2 h

(Fig. 5b), all nanoplates become thicker (a thickness of �85 nm)

and denser, coexisting with a lot of nanoparticles on the surface

of the nanosheets within the CIS film. 8 h of the solvothermal

treatment (Fig. 5c) results in the formation of the flower shaped

CIS architectures. When the reaction is further prolonged to 12 h

(Fig. 5d), a mass of flower shaped architectures appear on the

film, upwardly growing and self-aggregating with each other to

form CIS double-layered films; the ‘‘petals’’ of the flower on the

upper layer are as thin as �20 nm.

On the basis of the above results, a possible mechanism can

be proposed for the formation of the CIS films including single-

layered and double-layered structures, as shown in Fig. 6.

Firstly, since the Cu element on the surface of the Cu foil can be

easily oxidized to Cu2+ ions at high temperature while oxygen

was simultaneously reduced, Cu2+ ions are released from the Cu

foil into the solution.48–50 Then, these Cu2+ ions are reduced to

Cu+ ions by ethylene glycol, which acts both as the reaction

solvent and reductant (4Cu2+ + HO–CH2–CH2–OH / 4Cu+ +

4H+ + CHO–CHO).51 It has been revealed that thioacetamide

can react with water to release H2S (CH3CSNH2 + H2O /

CH3CONH2 + H2S), and then H2S decomposes to give S2� ions

at the proceeding temperature (180 �C) (H2S / 2H+ + S2�).52

Herein, the existence of trace water contained in ethylene glycol

and the crystal water (originating from InCl3$4H2O) is believed

to be necessary and plays an important role in the formation of

S2� ions.51 Subsequently, with a continuous supply of Cu+ from

the substrate, S2� from thioacetamide, and In3+ from

InCl3$4H2O, CIS nucleation occurs via a heterogeneous process

and then in situ growth carries out on the Cu foil (Cu+ + In3+ +

S2� / CuInS2). In this growth process, a high density of CIS

nuclei would lead to the concurrent growth of a large number of

CIS nanosheets, resulting in the congested growth of the CIS

nanosheets. The overcrowding effect would confine the propa-

gation of these nanosheets predominantly in the vertical direc-

tion, forming a single-layered CIS film on the Cu foil. If the

concentrations of InCl3$4H2O ($ 0.05 M) and thioacetamide

($ 0.1 M) are both high enough, with a large and continuous

supply of Cu+, S2�, and In3+ in the solution, meanwhile,

a homogeneous nucleation process goes on in the solution, and

then the homogeneous nuclei grow into individual flower sha-

ped architectures by Ostwald ripening process and self-assem-

bling process. Some of these flower shaped architectures deposit

on the earlier formed layer of ordered potato chips shaped

nanosheets, forming a double-layered structure on the Cu foil.

The fact that the above heterogeneous process happening on the

Cu foil and the homogeneous process happening in the solution

are due to the growth of the CIS films on Cu foil has been also

confirmed in Fig. S3 (see ESI†), in which both copper and non

copper substrates (e.g., silicon wafer) are used as substrates for

the CIS material growth upon the same experimental condi-

tions. Of course, a double-layered film consisting of a lower

layer of uniform ordered potato chips shaped nanosheets and

an upper layer of the flower shaped superstructures form on the

Cu foil, while only a few of the flower shaped superstructures

instead of the double-layered films scatters and deposits on the

Si wafer.

CrystEngComm, 2012, 14, 1825–1832 | 1829

Fig. 6 A possible formation mechanism of the CIS films.

Dow

nloa

ded

by U

nive

rsity

of

Ten

ness

ee a

t Kno

xvill

e on

02/

04/2

013

09:0

7:06

. Pu

blis

hed

on 0

9 Ja

nuar

y 20

12 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

E05

756A

View Article Online

3.2. Optical properties of as-grownCIS films and optoelectronic

properties of solar cells

The optical absorption of all CIS films was measured by an UV-

vis spectrometer, and their absorptions show similar spectra.

Fig. 7 shows a typical diffuse reflection spectrum of single-

layered CIS films obtained by treating Cu foils in ethylene glycol

solution containing InCl3$4H2O (0.025 M) and thioacetamide

(0.050 M) at 180 �C for 16 h in the autoclave. The sample pres-

ents a strong adsorption within a broad range between 400 and

800 nm, which is the characteristic absorption of the CIS mate-

rial and consistent with the black color of the grown CIS film.

Since CIS is a direct transition semiconductor, its absorption

coefficient (a) and bang-gap (Eg) satisfy the equation:

(ahv)2 ¼ A(hv � Eg) (1)

The band gap (Eg) is obtained by extrapolation of the plot of

(ahn)2 vs. (hv) and is determined to be 1.48 eV, as shown in the

inset of Fig.7. This band gap is very similar to that of the bulk

CIS material and also appropriate for the absorption of the solar

spectrum.

Fig. 7 The typical diffuse reflection spectrum of the as-grown single-

layered CIS films obtained by treating Cu foils in ethylene glycol solution

containing InCl3$4H2O (0.025 M) and thioacetamide (0.05 M) at 180 �Cfor 16 h in the autoclave. The inset shows a plot of (ahv)2 as a function of

hv for the absorption spectrum of the sample.

1830 | CrystEngComm, 2012, 14, 1825–1832

To clarify the potential of such an in situ growth of the CIS

films as a real low-cost and facile technology for a large scale

operation, a larger area of Cu foil (1.5 � 6.5 cm2) was used to be

directly coated by the CIS films. Fig. 8a and 8b give the

comparable photographs before and after in situ growth of the

CIS film on the flexible Cu foil. One can find that the Cu foil and

the resulting film both have an excellent flexibility. We have

examined the stability (or adhered force) of the CIS films on the

Cu foil during/after the bending process. No CIS material was

observed to fall off from the films during the bending process,

and no obvious cracks were found to form inside the CIS film

after the bending, as shown in Fig. S4b (Fig. S4a showing the film

before the bending).† These results suggest that after the bending

the grown CIS films still have a good stability or a considerably

strong adhered force on the Cu foil, which should be attributed

to the fact that the Cu foil not only acts as one of the source

materials but also directly serves as the substrate, resulting in in

situ growth of the CIS nanosheet alignments. Subsequently,

single-layered (thickness: �1.5 mm) and double-layered (thick-

ness: �6.5 mm) CIS films were used to fabricate the flexible solar

cells with a structure of Cu foil/CuInS2/CdS/i–ZnO/ITO/Ni–Al,

as shown in Fig. 8c. During the fabrication process, CIS films

(0.5 � 0.8 cm2) on the flexible Cu foils (�0.1 cm thickness) were

used as the absorber layer. Firstly, the n-type junction partner

CdS (�50 nm thickness) was deposited on CIS film by using

a chemical bath approach. Secondly, two magnetron sputtering

processes were utilized to deposit layers of intrinsic ZnO (�70 nm

in thickness) and transparent conductive indium tin oxide (ITO)

(�180 nm in thickness), respectively, on the CdS layer. Finally,

a Ni (50 nm)/Al (3 mm) dual layer top contact grid was

Fig. 8 Photographs of (a) the original Cu foil and (b) the -grown CIS

film on the flexible Cu foil, (c) schematic of the as-fabricated device

structure.

This journal is ª The Royal Society of Chemistry 2012

Dow

nloa

ded

by U

nive

rsity

of

Ten

ness

ee a

t Kno

xvill

e on

02/

04/2

013

09:0

7:06

. Pu

blis

hed

on 0

9 Ja

nuar

y 20

12 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

E05

756A

View Article Online

evaporated through a metal mask to facilitate collecting the

photogenerated carriers as the cathode in an as-fabricated solar

cell device. The resulting solar cell device (total area: �0.4 cm2)

with a structure of Cu foil/CuInS2/CdS/i–ZnO/ITO/Ni–Al was

defined by isolating the multilayer film areas via mechanical

scribing.

Photocurrent density–voltage characteristics of the resulting

solar cells were measured under standard AM 1.5 solar illu-

mination with an intensity of 80 mW cm�2, as shown in Fig. 9.

The solar cell based on single-layered CIS film exhibits open-

circuit voltage (Voc) of 268 mV, short-circuit current density

(Jsc) of 4.00 mA cm�2, fill factor (FF) of 0.56, yielding an

overall solar-to-electric energy conversion efficiency (h) of

0.75%. But, the conversion efficiency from the solar cell based

on double-layered CIS film is 0.33%, which is relatively low

compared with that of the single-layered one. It should be

noted that both conversation efficiencies of the as-fabricated

solar cells are relatively not high, which should be mainly

attributed to the following reasons. Firstly, it is well known

that an important morphological feature for the high-efficiency

film solar cells is that the film should be composed of large

densely packed grains.18 In our cases, there are a plenty of

grain boundaries and pores in both the single-layered and

especially the double-layered CIS films, which are composed of

small aligned nanoplates, leading to serious recombination and

a reduction in the conversion efficiency. Secondly, the CIS films

with copper-poor composition are desired for the high effi-

ciency film solar cells, because an excess of copper has been

revealed to lead to the formation of binary and/or ternary

phases of copper chalcogenide, resulting in a poorer photo-

electrical conversion performance.53 Thus, our CIS films with

copper-rich composition result in relatively low conversion

efficiencies. What’s more, the fabrication technique (e.g.,

depositing n-type junction partner CdS layer using a chemical

bath approach, depositing ZnO and ITO layers by magnetron

sputtering processes, and constructing the cathode of a Ni/Al

dual layer via an evaporation method) of the solar cell devices

as a key factor determining the conversion efficiency needs to

be improved, especially for such an in situ growth of CIS thin

films on the flexible Cu foil. Therefore, further improvement of

the conversion efficiency of the flexible solar cells based on the

as-grown CIS films can be expected by optimizing the

Fig. 9 J–V characteristics of the as-fabricated solar cells based on the

single-layered and double-layered CIS films.

This journal is ª The Royal Society of Chemistry 2012

morphology, structure and composition of the CIS films, as

well as the device fabrication technique.

4. Conclusions

In summary, an in situ growth of CIS films has been demon-

strated by solvothermally treating flexible Cu foils in ethylene

glycol solution containing InCl3$4H2O and thioacetamide. Sol-

vothermal temperature has an obvious effect on the crystallinity

of the final films, while reactant concentrations play a significant

role in the morphology of the CIS films. By tuning the concen-

tration of InCl3$4H2O and thioacetamide, single-layered

(thickness: � 1.5 mm) and double-layered (thickness: � 6.5 mm)

CIS films on flexible Cu foil are both obtained, respectively. All

the CIS films exhibit a direct bandgap energy of 1.48 eV, which is

appropriate for the absorption of solar spectrum. Flexible solar

cells based on single-layered and double-layered CIS films have

been constructed, and they exhibit a power conversion efficiency

of 0.75% and 0.33%, respectively. Further improvement of the

conversion efficiency of the flexible solar cells based on the as-

grown CIS films can be expected by optimizing the morphology,

structure and composition of the CIS films, as well as the device

fabrication technique.

Acknowledgements

This work was supported from the National Natural Science

Foundation of China (Grant no. 21171035, 50872020 and

50902021), the Program forNewCentury Excellent Talents of the

University in China, the ‘‘Pujiang’’ Program of Shanghai

Education Commission (Grant no. 09PJ1400500), the ‘‘Dawn’’

Program of the Shanghai Education Commission (Grant no.

08SG32), the Science and Technology Commission of Shanghai-

based ‘‘Innovation Action Plan’’ Project (Grant no.

10JC1400100), the ‘‘Chen Guang’’ project (Grant no. 09CG27)

supported by the Shanghai Municipal Education Commission

and Shanghai Education Development Foundation, the Program

for the Specially Appointed Professor by Donghua University

(Shanghai, P.R. China), the Shanghai Leading Academic Disci-

pline Project (Grant no. B603), and the Program of Introducing

Talents of Discipline to Universities (no. 111-2-04).

References

1 H. J. Lewerenz, H. Goslowsky, K. D. Husemann and S. Fiechter,Nature, 1986, 321, 687.

2 C. C. Landry and A. R. Barron, Science, 1993, 260, 1653.3 M. Nanu, J. Schoonman and A. Goossens,Adv.Mater., 2004, 16, 453.4 D. C. Pan, L. J. An, Z. M. Sun, W. Hou, Y. Yang, Z. Z. Yang andY. F. Lu, J. Am. Chem. Soc., 2008, 130, 5620.

5 M. Kruszynska, H. Borchert, J. Parisi and J. Kolny-Olesiak, J. Am.Chem. Soc., 2010, 132, 15976.

6 J. Klaer, J. Bruns, R. Henninger, K. Seimer, R. Klenk, K. Ellmer andD. Braunig, Semicond. Sci. Technol., 1998, 13, 1456.

7 L. Stolt, J. Hedstrom, J. Kessler, M. Ruckh, K. O. Velthaus andH. W. Schock, Appl. Phys. Lett., 1993, 62, 597.

8 R. Scheer, M. Alt, I. Luck and H. J. Lewerenz, Sol. Energy Mater.Sol. Cells, 1997, 49, 423.

9 T. Watanabe, M. Matsui and K. Mori, Sol. Energy Mater. Sol. Cells,1994, 35, 239.

10 T. Unold, I. Sieber and K. Ellmer,Appl. Phys. Lett., 2006, 88, 213502.11 O. Volobujeva, M. Altosaar, J. Raudoja, E. Mellikov, M. Grossberg,

L. Kaupmees and P. Barvinschi, Sol. Energy Mater. Sol. Cells, 2009,93, 11.

CrystEngComm, 2012, 14, 1825–1832 | 1831

Dow

nloa

ded

by U

nive

rsity

of

Ten

ness

ee a

t Kno

xvill

e on

02/

04/2

013

09:0

7:06

. Pu

blis

hed

on 0

9 Ja

nuar

y 20

12 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

E05

756A

View Article Online

12 H. Sun, Y. Luo, Y. Zhang, D. Li, Z. Yu, K. Li and Q. Meng, J. Phys.Chem. C, 2010, 114, 11673.

13 M. G. Panthani, V. Akhavan, B. Goodfellow, J. P. Schmidtke,L. Dunn, A. Dodabalapur, P. F. Barbara and B. A. Korgel, J. Am.Chem. Soc., 2008, 130, 16770.

14 R. G. Xie, M. Rutherford and X. G. Peng, J. Am. Chem. Soc., 2009,131, 5691.

15 T. T. John, M.Mathew, C. S. Kartha, K. P. Vijayakumar, T. Abe andY. Kashiwaba, Sol. Energy Mater. Sol. Cells, 2005, 89, 27.

16 L. Li, N. Coates and D. Moses, J. Am. Chem. Soc., 2010, 132, 22.17 D. B.Mitzi, M. Yuan,W. Liu, A. J. Kellock, S. J. Chey, V. Deline and

A. G. Schrott, Adv. Mater., 2008, 20, 3657.18 W. Liu, D. B. Mitzi, M. Yuan, A. J. Kellock, S. J. Chey and

O. Gunawan, Chem. Mater., 2010, 22, 1010.19 B.D.Weil, S. T.Connor andY.Cui, J. Am.Chem. Soc., 2010, 132, 6642.20 Q. Guo, S. J. Kim, M. Kar, W. N. Shafarman, R. W. Birkmire,

E. A. Stach, R. Agrawal and H. W. Hillhouse, Nano Lett., 2008, 8,2982.

21 Q. Guo, G. M. Ford, H. W. Hillhouse and R. Agrawal, Nano Lett.,2009, 9, 3060.

22 T. Terasako, Y. Uno, T. Kariya and S. Shirakata, Sol. Energy Mater.Sol. Cells, 2006, 90, 262.

23 T. Terasako, S. Inoue, T. Kariya and S. Shirakata, Sol. EnergyMater.Sol. Cells, 2007, 91, 1152.

24 S. Jost, F. Hergert, R. Hock, J. Schulze, A. Kirbs, T. Voss,M. Purwins and M. Schmid, Sol. Energy Mater. Sol. Cells, 2007,91, 636.

25 R. N. Bhattacharya and A. M. Fernandez, Sol. Energy Mater. Sol.Cells, 2003, 76, 331.

26 S. Nakamura and A. Yamamoto, Sol. Energy Mater. Sol. Cells, 2003,75, 81.

27 T. Todorov, E. Cordoncillo, J. F. S. S�anchez-Royo, J. Carda andP. Escribano, Chem. Mater., 2006, 18, 3145.

28 S. Ahn, C. Kim, J. H. Yun, J. Gwak, S. Jeong, B. H. Ryu andK. Yoon, J. Phys. Chem. C, 2010, 114, 8108.

29 S. E. Habas, H. A. S. Platt, M. F. A. M. van Hest and D. S. Ginley,Chem. Rev., 2010, 110, 6571.

30 W. U. Huynh, J. J. Dittmer and A. P. Alivisatos, Science, 2002, 295,2425.

31 I. Gur, N. A. Fromer,M. L. Geier and A. P. Alivisatos, Science, 2005,310, 462.

1832 | CrystEngComm, 2012, 14, 1825–1832

32 T. L. Li and H. S. Teng, J. Mater. Chem., 2010, 20, 3656.33 R. L. Crabb and F. C. Treble, Nature, 1967, 213, 1223.34 F. C. Krebs, Sol. Energy Mater. Sol. Cells, 2009, 93, 465.35 F. C. Krebs, S. A. Gevorgyan and J. Alstrup, J. Mater. Chem., 2009,

19, 5442.36 W. Zhang and S. Yang, Acc. Chem. Res., 2009, 42, 1617.37 K. Zhu, N. R. Neale, A. Miedaner and A. J. Frank, Nano Lett., 2007,

7, 69.38 P. Roy, D. Kim, K. Lee, E. Spiecker and P. Schmuki, Nanoscale,

2010, 2, 45.39 Z. Yang, T. Xu, Y. Ito, U. Welp and W. K. Kwok, J. Phys. Chem. C,

2009, 113, 20521.40 L. Ma, Y. Lin, Y. Wang, J. Li, E. Wang, M. Qiu and Y. Yu, J. Phys.

Chem. C, 2008, 112, 18916.41 X. Wen, W. Zhang and S. Yang, Langmuir, 2003, 19, 5898.42 H. H. Chen, R. J. Zou, N. Wang, H. H. Chen, Z. Y. Zhang,

Y. G. Sun, L. Yu, Q. W. Tian, Z. G. Chen and J. Q. Hu, J. Mater.Chem., 2011, 21, 3053.

43 H. M. Jia, W. W. He, X. W. Chen, Y. Lei and Z. Zheng, J. Mater.Chem., 2011, 21, 12824.

44 D. P. Li, Z. Zheng, Y. Lei, F. L. Yang, S. X. Ge, Y. D. Zhang,B. J. Huang, Y. H. Gao, K. W. Wong and W. M. Lau, Chem.–Eur.J., 2011, 17, 7694.

45 D. Li, Z. Zheng, Z. Shui, M. Long, J. Yu, K. W. Wong, L. Yang,L. Zhang and W. M. Lau, J. Phys. Chem. C, 2008, 112,2845.

46 Z. Zheng, A. Liu, S. Wang, B. Huang, K. W. Wong, X. Zhang,S. K. Hark and W. M. Lau, J. Mater. Chem., 2008, 18, 852.

47 D. Li, Z. Zheng, Y. Lei, S. Ge, Y. Zhang, Y. Zhang, K. W. Wong,F. Yang and W. M. Lau, CrystEngComm, 2010, 12, 1856.

48 Z. P. Zhang, H. P. Sun, X. Q. Shao, D. F. Li, H. D. Yu andM. Y. Han, Adv. Mater., 2005, 17, 42.

49 Z. Zhang, X. Shao, H. Yu, Y. Wang and M. Han, Chem. Mater.,2004, 17, 332.

50 J. P. Liu, X. T. Huang, Y. Y. Li, K. M. Sulieman, X. He andF. L. Sun, J. Mater. Chem., 2006, 16, 4427.

51 A. Zhang, Q. Ma, M. K. Lu, G. W. Yu, Y. Y. Zhou and Z. F. Qiu,Cryst. Growth Des., 2008, 8, 2402.

52 S. F. Wang, F. Gu and M. K. Lu, Langmuir, 2006, 22, 398.53 H. Katagiri, K. Jimbo, W. S. Maw, K. Oishi, M. Yamazaki, H. Araki

and A. Takeuchi, Thin Solid Films, 2009, 517, 2455.

This journal is ª The Royal Society of Chemistry 2012