- Email: [email protected]
ZnO solar cells, with chemically deposited ZnS buffer layers from acidic solutions
ARTICLE IN PRESS
Solar Energy Materials & Solar Cells 92 (2008) 302–306 www.elsevier.com/locate/solmat
Study of CuInS2/ZnS/ZnO solar cells, with chemically deposited ZnS buffer layers from acidic solutions B. Asenjoa,, A.M. Chaparroa, M.T. Gutie´rreza, J. Herreroa, J. Klaerb a
Department of Energy, CIEMAT, Avda. Complutense 22, 28040 Madrid, Spain b Hahn-Meitner-Institut, SE3, Glienicker Str. 100, D-14109 Berlin, Germany Received 28 June 2007; accepted 11 September 2007 Available online 22 October 2007
Abstract Thin film solar cells based on CuInS2/ZnS/ZnO have been prepared with ZnS buffer film of different thickness. ZnS films are grown by chemical bath deposition (CBD) from acidic solutions of ZnSO4 and thioacetamide (TA). The change of the growth rate with time is studied by means of the quartz crystal microbalance. Films with different thickness show variable physical, chemical and morphological properties. The structure is studied with X-ray diffraction, showing different crystallinity with deposition time. The absorption coefficient depends also on the CBD deposition time, and shows absorption edges between 2.70 and 3.65 eV. The compositional analysis carried out with XPS (surface) and EDAX (bulk). Bulk composition reflects highly stoichiometric films, with Zn/S ratios close to unit. Preliminary results with CuInS2-based solar cell show efficiencies around 5%, lower than usually found with standard CdS buffer films (around 9%). r 2007 Elsevier B.V. All rights reserved. Keywords: Zinc sulphide; Buffer; Chemical deposition; Thin film
1. Introduction Zinc suphide (ZnS) is an important semiconducting material with a wide direct band gap of 3.65 eV [1]. It is of interest for replacement of CdS as buffer layer of thin filmbased solar cells due to higher energy gap, good transparency, and general good film properties (compact, adherent, conforming). Several techniques can be used for thin film growth of ZnS, such as chemical vapour deposition [2], spray pyrolysis [3] and chemical bath deposition (CBD) [4–7]. Some authors have obtained good solar cell results with this buffer type; Neve et al. [8] prepared ZnS by the CBD method in basic solution and obtained 10.7% efficiency on CuInS2-based cells. However, few authors prepare ZnS in acidic solutions; O’Hare et al. [9] obtained ZnS from ZnCl2 between pH 2.0 and 5.0 and Makhova et al. [10] compared ZnS films obtained by acidic and basic solutions, arriving at the conclusion that thin films prepared in acidic solutions present more stoichiometric composition. Corresponding author.
E-mail address: [email protected] (B. Asenjo). 0927-0248/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2007.09.005
The aim of this work is to prepare by CBD-ZnS thin films of different thinckness by CBD, using thioacetamide (TA) solution and ZnSO4. The growth rate of the film is monitored with a quartz crystal microbalance (QCM). Composition, optical and structural properties are studied for the films. Their properties as buffer layers of CuInS2/ ZnS/ZnO solar cells are also analysed. 2. Experimental The films with ZnS composition were deposited on the different substrates (glass, Au, and CuInS2) from an acidic solution (pH ¼ 2) using 37% hydrocloric acid 0.01 M (HCl, Merck), thioacetamide 0.5 M (TA, CH3CSNH3, Fluka), zinc sulphate 0.01 M (ZnSO4, Fluka) and acetic acid 0.3 M (CH3COOH, Merck). Growth rate was measured in a thermostatised bath at 70 1C equipped with a QCM (Maxtek Inc.), as described elsewhere [11]. Au covered quartz substrates (unpolished, AT cut, Maxtek) were used for the QCM study. Surface and bulk composition of the films was determined with XPS technique (Perkin-Elmer PHI 5400 spectrometer, hn ¼ 1253.6 eV Mg Ka radiation), using
ARTICLE IN PRESS B. Asenjo et al. / Solar Energy Materials & Solar Cells 92 (2008) 302–306
303
angle-resolved X-ray photoelectron spectroscopy (ARXPS) detection. The energy position of the signals was measured with respect to the energy of adventitious carbon (EB ¼ 284.8 eV). Parameters of the signals were obtained by fitting to a symmetric Gaussian–Lorentzian (0.8–0.2) sum function after subtracting the background by the Shirley method. Absorption coefficient (a) and band gap (Eg) of the film was determined from transmittance (T) and reflectance (R) spectra (Perkin-Elmer Lambda 9 spectrometer) using the equation [12]:
(1) An induction time, defined as the time interval from the addition of the last reactant (TA) to the start of film growth, is observed of about 10 min. During this time, the beginning of the decomposition of the thioacetamide (reaction (2)) and formation of the first crystalline centre takes place:
ð1 RÞ2 ead , (1) 1 R2 e2ad where d is the film thickness. For the preparation of CuInS2/ZnS/ZnO solar cells, CuInS2 thin films were grown on Mo substrate by a sequential sputtering-evaporation process [13]. Previous to ZnS buffer layer deposition, the surface of CuInS2 is treated in 0.1 M KCN solution at 40 1C during 2 min to remove the CuS overlayer that results from the growth process. Then, the buffer film of ZnS composition is grown by CBD in a thermostatised reactor. The ZnO window layer is deposited by sputtering onto the CuInS2/buffer bilayer and, finally, aluminium grid contacts are evaporated on top of the ZnO layer. The solar cell results presented here correspond to a set fabricated with CuInS2 films generated in the same batch process, with identical termination of the cell (i.e., deposition of ZnO and Al contacts), in order to differentiate solely buffer layer deposition effects. The current vs voltage (I–V) curves were obtained under 100 mW cm2 AM1.5 illumination at 25 1C.
(2) A linear increase of the reaction rate, which corresponds with the growth of the most compact film. The growth of nucleation centres occurs during this period:
T¼
3. Results and discussion Fig. 1 shows general features of the growth rate of ZnS composition films, measured with QCM technique on Au covered substrates. Different stages for growth of ZnS can be differentiated in the acid dissolution. 20 1250
nm
750
10
500
V / nm min-1
15
1000
5 250 0 0
50
100
150
200
0 250
min Fig. 1. Growth rate and thickness plots of ZnS thin films deposited on Au substrate. Solution composition: [TA] ¼ 0.5 M, [ZnSO4] ¼ 0.01 M, [HCl] ¼ 0.01 M and [CH3COOH] ¼ 0.3 M at 70 1C.
CH3 CSNH2 þ 2H2 O ! H2 S þ NH4 þ þ CH3 COO (2)
H2 S þ Zn2þ ! ZnSk þ 2Hþ
(3)
(3) A period of constant rate of growth, when a three dimensional film is growing after coalescence of the centres formed in the previous stage. There is a constant growth rate of 10 nm min1. (4) Finally, a decrease of the growth rate. In this period, ZnS colloids are formed in the bulk of the dissolution, which does not adhere to the film so they do not contribute to its growth. Similar succession of stages has been proposed by Eshuis et al. [14]. Other authors [15] propose reaction with two mechanisms, in the basic solution: one heterogeneous due to the reaction of the cations of Zn2+ adsorbed on the surface of the substrate, that gives rise to the formation of hydroxides and oxides; and a homogeneous mechanism due to deposition of aggregates of ZnS giving rise to films more stoichiometric and porous. Both mechanisms coexist in the films preparation of ZnS in the basic medium. In the acid medium, however, the formation of oxides and hydroxides is not favored by pH conditions, and the resulting films are more stoichiometric. Crystallographic analysis of the films by X-ray diffraction (XRD) is shown in Fig. 2. The results correspond to films grown during 30 and 40 min on lime glass substrate previously activated by immersion in SnCl2 (0.01 M). It is also shown the diffractogram of the second film after thermal treatment at 300 1C during 30 min in N2 atmosphere. The results show a mixture of cubic (JCPDS 050566) and hexagonal (JCPDS 89-1363) phases, as observed by others [16,17]. Crystallinity of the films increases with deposition time. Fig. 3 shows the optical spectra of the films grown after different deposition time (25, 30 and 40 min). Because the reference for these optical measurements was air, the transmittance and reflectance of the bare glass substrate have been included for comparison. Fig. 4 shows the absorption spectra of the films calculated from the data in Fig. 3, using Eq. (1). The absorption edge increases with the time of deposition, towards values close to the gap energy of ZnS (3.6 eV) [1]. For the film prepared
ARTICLE IN PRESS B. Asenjo et al. / Solar Energy Materials & Solar Cells 92 (2008) 302–306
304
25min 30min 40min
105 α (cm-1)
-(311)cub/(118)hex
Intensity / a.u.
(220)cub (110)hex
-(111)cub -(008)hex
106
40 min+ 300 °C
104
40 min
103
ZnO
30 min
2.0 0
10
20
30
40
50
60
70
80
2.5
90
2θ Fig. 2. XRD of ZnS films. Solution composition: [TA] ¼ 0.5 M, [ZnSO4] ¼ 0.01 M, [HCl] ¼ 0.01 M and [CH3COOH] ¼ 0.3 M at 70 1C, pH ¼ 2.0. Films deposited during 20, 30 and 40 min, and after heat treatment (300 1C, 30 min N2).
ZnS
3.0 3.5 Energía (eV)
4.0
4.5
Fig. 4. Absorption coefficient spectra of the films, calculated according to Eq. (1) from results in Fig. 3.
ZnO
ZnS
Zn0
80 25min
T T, R (%)
30min 60
Intensity / u.a.
100
40min
70° 45° 20°
SLG 40
20
980
R
0 0
500
1000
1500
2000
2500
982
984
990 986 988 Kinetic energy / eV)
992
994
996
Fig. 5. XPS signal corresponding to Zn KMM Auger transition of a sample deposited during 30 min. Same solution composition as in Fig. 1.
λ (nm) Fig. 3. Optical transmittance and reflectance spectra for ZnS films. Solution composition: [TA] ¼ 0.5 M, [ZnSO4] ¼ 0.01 M, [HCl] ¼ 0.01 M and [CH3COOH] ¼ 0.3 M at 70 1C, pH ¼ 2.0. Films deposited during 25, 30 and 40 min.
Intensity / a.u.
at 25 min the absorption edge value is the lowest, possibly reflecting the formation of secondary compounds, such as oxides and hydroxides, which may be formed at the beginning of the process by heterogeneous growth mechanism. Composition of the films was studied with XPS. The analysis of Zn LMM signal is shown for the film prepared at 30 and 40 min in Figs. 5 and 6, respectively, under 201, 451 and 701 detection angle. For the film deposited at 30 min the signal shows little dependence on exit angle, positioned at the value of the ZnS (988.1 eV). However, the film prepared after 40 min deposition, the energy shifts upon increasing the exit angle, probably reflecting the presence of the oxide in the surface.
ZnS
Zn0
ZnO
70°
45°
20°
980
982
984
986 988 990 Kinetic energy / eV)
992
994
996
Fig. 6. XPS signal corresponding to Zn KMM Auger transition of a sample deposited during 40 min. Same solution composition as in Fig. 1.
ARTICLE IN PRESS B. Asenjo et al. / Solar Energy Materials & Solar Cells 92 (2008) 302–306
The elemental composition obtained by XPS is shown in Table 1. It is observed that the percentage of S is highest in all the samples while the relationship Zn/S oscillates between 0.5 and 0.7, indicating a different composition in the surface that in the bulk of the film. Results of EDAX (Table 2), more sensitive to the bulk of the film, show in fact a relationship Zn/S close to unit. Table 2 also shows the analysis of sulphur compounds present in the films, obtained from the S 2p signal at the three exit angles for electrons. Sulphide shows major proportion, above 80% for the film without heating. The increase in the sulphates
Table 1 XPS results of the elemental composition ZnS films at 451 detection angle prepared at two different deposition times: 30 min (a), 40 min (b), and 40 min with thermal treatment at 300 1C atmosphere N2 during 30 min (c) Time of deposition
(a) 30 min
(b) 40 min
(c) 40 min at 300 1C
Detection angle (1)
20
45
70
20
45
70
20
45
70
O Zn S C Zn/S Sulphates Sulphides Ek (eV; Zn LMM)
20.1 8.8 17.5 53.7 0.50 18.0 82
21.6 10.4 21.6 46.5 0.48 17.0 83 988.8
17.9 10.8 22.1 49.1 0.49 19.0 81
18.3 12.2 20.1 49.4 0.60 16.0 84
19.4 14.7 23.4 42.5 0.63 18.0 82 989.2
19.0 15.3 24.7 41.0 0.62 19.0 81
34.4 9.5 12.4 43.7 0.77 23.0 77
41.3 9.7 15.5 35.5 0.63 30.0 70 989.9
40.0 9.8 15.0 31.9 0.66 24.0 76
Other conditions as in Fig. 1. Table 2 Composition analysis of three films prepared at different time deposition determined by EDAX Films
Zn (at%)
S (at%)
Zn/S
25 min 30 min 40 min 40 min at 300 1C
44.86 55.05 50.02 53.62
55.14 44.95 49.98 46.38
0.8 1.2 1.0 1.2
305
proportion for the heated films reflects some oxidation of the sulphur. 3.1. CuInS2/ZnS/ZnO solar cell results solar cell results The solar cells results, corresponding to different buffer layer conditions, are shown in Table 3. The buffer layers were deposited on CuInS2 at 70 1C, at different deposition times, maintaining other conditions constant. In principle, they show lower performance than results obtained with other buffer films type, as In2S3 and In2S3–ZnS prepared under similar conditions [18]. Maximum efficiency of 5.3% is obtained with ZnS grown up to 25 min (Table 3). This buffer film is thinner and has higher oxygen proportion than those grown to 30 and 40 min (Table 2), which could result in a beneficial effect on absorber-buffer band alignment [19,20]. Below 25 min deposition time, the buffer film performance decreases, probably because it is not fully covering the substrate surface. In general, cell performances are below those obtained with CdS buffer layers, in the 9–10% range under the same preparation conditions. The most significant differences are encountered in the values of fill factor (FF) and short circuit current (Isc) (Table 3). Such low values are also reflected by low quantum efficiencies of the cells compared with CdS buffer cell (Fig. 7). Given the higher transparency of ZnS buffer layers, the decrease of the quantum efficiency on CuInS2/ZnS/ZnO must be attributed to the low quality of the absorbent/buffer interface. 4. Conclusions Buffer layers of ZnS composition have been prepared by the CBD method. The kinetics of growth of the layers is analysed with a QCM, showing four growth stages. Crystallographic analysis shows a mixture of phases and compositional analysis shows more stoichiometric composition in the bulk of the film that in the surface, where there is probably a higher concentration of zinc oxides and
Table 3 Deposition conditions for the growth of ZnS buffer films on CuInS2 from aqueous solution at 70 1C CuInS2/buffer
Time (min)a Thickness (nm) FF (%) Voc (mV) Isc (mA cm2) Rs (O cm2) Rp (O cm2) Z (%)
934-1/ZnS
934-2/ZnS
934-3/ZnS
934-5/ZnS
CdS
20 38 36.572.1 685710 13.471.1 1.5370.35 10479 3.470.4
25 64 44.372.0 613719 19.671.2 5.6772.67 343771 5.370.5
30 108 35.774.2 644716 14.173.5 2.5171.07 100714 3.371.1
40 195 46.177.2 632732 14.674.1 1.8670.92 160737 4.471.8
– – 66.071.8 69974 11.0470.28
Also results for a CdS buffer cell prepared under similar conditions. a Other solution components are 0.5 M TA, 0.025 M InCl3, 0.01 M HCl and 0.3 M CH3COOH.
10.270.2
ARTICLE IN PRESS B. Asenjo et al. / Solar Energy Materials & Solar Cells 92 (2008) 302–306
306
1.0 934-1/ZnS 934-2/ZnS 934-3/ZnS 934-5/ZnS CdS
0.8
QE
0.6
0.4
0.2
0.0 400
500
600 700 800 Wavelength (nm)
900
1000
Fig. 7. Quantum efficiency (QE) of CuInS2/ZnS/ZnO solar cells, with ZnS buffer deposited at 70 1C, other conditions as in Table 3. Also shown the QE results of cells from the same batch with CdS buffer layer.
hydroxides. Solar cells, CuInS2/ZnS/ZnO, have given efficiencies above 5%. Acknowledgements This work has been supported by FOTOFLEX Project (IV-PRICIT-CAM) and ARECES Project (Ramo´n Areces Fundation). References [1] R. Ortega Borges, D. Lincot, J. Vedel, in: Proceedings of the 11th E.C. Photovoltaic Solar Energy Conference, 1992, p. 862.
[2] E.Y.M. Lee, N.H. Tran, R.N. Lamb, Appl. Surf. Sci. 241 (3–4) (2005) 493. [3] O¨. Mustafa, M. Bedir, S. Ocak, R.G. Yildirim, J. Mater. Sci.: Mater. Electron. 15 (5) (2007) 505. [4] F. Go¨de, C. Gu¨mu¨s, M. Zor, J. Crystal Growth 299 (1) (2007) 136. [5] C. Kaufmann, P.J. Dobson, in: Proceedings of 28th IEEE Photovoltaic Specialists Conference, 2000, p. 688. [6] R.N. Bhattacharya, K. Ramanathan, L. Gedvilas, B. Keyes, J. Phys. Chem. Sol. 66 (2006) 1862. [7] B. Mokili, Y. Charreire, R. Cortes, D. Lincot, Thin Solid Films 288 (1996) 21. [8] S. Neve, W. Bohne, J. Klaer, R. Klenk, R. Scheer, in: Proceedings of 17th European Photovoltaic Solar Energy Conference, Munich, 2001, p. 1102. [9] J.P. O’Hare, K. Govender, P. O’Brien, D. Smyth-Boyle, Mater. Res. Soc. Symp. Proc. 668 (2001) H8.15.1. [10] L.V. Makhova, I. Konovalov, R. Szargan, N. Aschkenov, M. Schubert, T. Chasse´, Phys. Status Solidi C 2 (3) (2005) 1206. [11] A.M. Chaparro, M.T. Gutie´rrez, J. Herrero, Electrochim. Acta 47 (2001) 977. [12] G. Connel, A. Lewis, Phys. Status Solidi B 60 (1973) 291. [13] J. Klaer, J. Bruns, R. Henninger, K. Siemer, R. Klenk, K. Ellmer, D. Bra¨unig, Semicond. Sci. Technol. 13 (1998) 1456. [14] A. Eshuis, G.R.A. van Elderen, C.A.J. Koning, Colloids Surf., A: Physicochem. Eng. Aspects 151 (1999) 505. [15] C.-Y. Yeh, Z.W. Lu, S. Froyen, A. Zunger, Phys. Rev. B46 (1992) 10086. [16] J. Cheng, D.B. Fan, H. Wang, B.W. Liu, Y.C. Zhang, H. Yan, Semicond. Sci. Technol. 18 (2003) 676. [17] T. Ben Nasr, N. Kamoun, M. Kanzari, R. Bennaceu, Thin Solid Films 500 (2006) 4. [18] B. Asenjo, A.M. Chaparro, M.T. Gutie´rrez, J. Herrero, J. Klaer, Thin Solid Films 515 (2007) 6036. [19] S.B. Zhang, S.-H. Wei, A. Zunger, J. Appl. Phys. 83 (1998) 3192. [20] C. Persson, C. Platzer-Bjo¨rkman, J. Malmstro¨m, T. To¨rndahl, M. Edoff, Phys. Rev. Lett 97 (2006) 146403.
Solar Energy Materials & Solar Cells 92 (2008) 302–306 www.elsevier.com/locate/solmat
Study of CuInS2/ZnS/ZnO solar cells, with chemically deposited ZnS buffer layers from acidic solutions B. Asenjoa,, A.M. Chaparroa, M.T. Gutie´rreza, J. Herreroa, J. Klaerb a
Department of Energy, CIEMAT, Avda. Complutense 22, 28040 Madrid, Spain b Hahn-Meitner-Institut, SE3, Glienicker Str. 100, D-14109 Berlin, Germany Received 28 June 2007; accepted 11 September 2007 Available online 22 October 2007
Abstract Thin film solar cells based on CuInS2/ZnS/ZnO have been prepared with ZnS buffer film of different thickness. ZnS films are grown by chemical bath deposition (CBD) from acidic solutions of ZnSO4 and thioacetamide (TA). The change of the growth rate with time is studied by means of the quartz crystal microbalance. Films with different thickness show variable physical, chemical and morphological properties. The structure is studied with X-ray diffraction, showing different crystallinity with deposition time. The absorption coefficient depends also on the CBD deposition time, and shows absorption edges between 2.70 and 3.65 eV. The compositional analysis carried out with XPS (surface) and EDAX (bulk). Bulk composition reflects highly stoichiometric films, with Zn/S ratios close to unit. Preliminary results with CuInS2-based solar cell show efficiencies around 5%, lower than usually found with standard CdS buffer films (around 9%). r 2007 Elsevier B.V. All rights reserved. Keywords: Zinc sulphide; Buffer; Chemical deposition; Thin film
1. Introduction Zinc suphide (ZnS) is an important semiconducting material with a wide direct band gap of 3.65 eV [1]. It is of interest for replacement of CdS as buffer layer of thin filmbased solar cells due to higher energy gap, good transparency, and general good film properties (compact, adherent, conforming). Several techniques can be used for thin film growth of ZnS, such as chemical vapour deposition [2], spray pyrolysis [3] and chemical bath deposition (CBD) [4–7]. Some authors have obtained good solar cell results with this buffer type; Neve et al. [8] prepared ZnS by the CBD method in basic solution and obtained 10.7% efficiency on CuInS2-based cells. However, few authors prepare ZnS in acidic solutions; O’Hare et al. [9] obtained ZnS from ZnCl2 between pH 2.0 and 5.0 and Makhova et al. [10] compared ZnS films obtained by acidic and basic solutions, arriving at the conclusion that thin films prepared in acidic solutions present more stoichiometric composition. Corresponding author.
E-mail address: [email protected] (B. Asenjo). 0927-0248/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2007.09.005
The aim of this work is to prepare by CBD-ZnS thin films of different thinckness by CBD, using thioacetamide (TA) solution and ZnSO4. The growth rate of the film is monitored with a quartz crystal microbalance (QCM). Composition, optical and structural properties are studied for the films. Their properties as buffer layers of CuInS2/ ZnS/ZnO solar cells are also analysed. 2. Experimental The films with ZnS composition were deposited on the different substrates (glass, Au, and CuInS2) from an acidic solution (pH ¼ 2) using 37% hydrocloric acid 0.01 M (HCl, Merck), thioacetamide 0.5 M (TA, CH3CSNH3, Fluka), zinc sulphate 0.01 M (ZnSO4, Fluka) and acetic acid 0.3 M (CH3COOH, Merck). Growth rate was measured in a thermostatised bath at 70 1C equipped with a QCM (Maxtek Inc.), as described elsewhere [11]. Au covered quartz substrates (unpolished, AT cut, Maxtek) were used for the QCM study. Surface and bulk composition of the films was determined with XPS technique (Perkin-Elmer PHI 5400 spectrometer, hn ¼ 1253.6 eV Mg Ka radiation), using
ARTICLE IN PRESS B. Asenjo et al. / Solar Energy Materials & Solar Cells 92 (2008) 302–306
303
angle-resolved X-ray photoelectron spectroscopy (ARXPS) detection. The energy position of the signals was measured with respect to the energy of adventitious carbon (EB ¼ 284.8 eV). Parameters of the signals were obtained by fitting to a symmetric Gaussian–Lorentzian (0.8–0.2) sum function after subtracting the background by the Shirley method. Absorption coefficient (a) and band gap (Eg) of the film was determined from transmittance (T) and reflectance (R) spectra (Perkin-Elmer Lambda 9 spectrometer) using the equation [12]:
(1) An induction time, defined as the time interval from the addition of the last reactant (TA) to the start of film growth, is observed of about 10 min. During this time, the beginning of the decomposition of the thioacetamide (reaction (2)) and formation of the first crystalline centre takes place:
ð1 RÞ2 ead , (1) 1 R2 e2ad where d is the film thickness. For the preparation of CuInS2/ZnS/ZnO solar cells, CuInS2 thin films were grown on Mo substrate by a sequential sputtering-evaporation process [13]. Previous to ZnS buffer layer deposition, the surface of CuInS2 is treated in 0.1 M KCN solution at 40 1C during 2 min to remove the CuS overlayer that results from the growth process. Then, the buffer film of ZnS composition is grown by CBD in a thermostatised reactor. The ZnO window layer is deposited by sputtering onto the CuInS2/buffer bilayer and, finally, aluminium grid contacts are evaporated on top of the ZnO layer. The solar cell results presented here correspond to a set fabricated with CuInS2 films generated in the same batch process, with identical termination of the cell (i.e., deposition of ZnO and Al contacts), in order to differentiate solely buffer layer deposition effects. The current vs voltage (I–V) curves were obtained under 100 mW cm2 AM1.5 illumination at 25 1C.
(2) A linear increase of the reaction rate, which corresponds with the growth of the most compact film. The growth of nucleation centres occurs during this period:
T¼
3. Results and discussion Fig. 1 shows general features of the growth rate of ZnS composition films, measured with QCM technique on Au covered substrates. Different stages for growth of ZnS can be differentiated in the acid dissolution. 20 1250
nm
750
10
500
V / nm min-1
15
1000
5 250 0 0
50
100
150
200
0 250
min Fig. 1. Growth rate and thickness plots of ZnS thin films deposited on Au substrate. Solution composition: [TA] ¼ 0.5 M, [ZnSO4] ¼ 0.01 M, [HCl] ¼ 0.01 M and [CH3COOH] ¼ 0.3 M at 70 1C.
CH3 CSNH2 þ 2H2 O ! H2 S þ NH4 þ þ CH3 COO (2)
H2 S þ Zn2þ ! ZnSk þ 2Hþ
(3)
(3) A period of constant rate of growth, when a three dimensional film is growing after coalescence of the centres formed in the previous stage. There is a constant growth rate of 10 nm min1. (4) Finally, a decrease of the growth rate. In this period, ZnS colloids are formed in the bulk of the dissolution, which does not adhere to the film so they do not contribute to its growth. Similar succession of stages has been proposed by Eshuis et al. [14]. Other authors [15] propose reaction with two mechanisms, in the basic solution: one heterogeneous due to the reaction of the cations of Zn2+ adsorbed on the surface of the substrate, that gives rise to the formation of hydroxides and oxides; and a homogeneous mechanism due to deposition of aggregates of ZnS giving rise to films more stoichiometric and porous. Both mechanisms coexist in the films preparation of ZnS in the basic medium. In the acid medium, however, the formation of oxides and hydroxides is not favored by pH conditions, and the resulting films are more stoichiometric. Crystallographic analysis of the films by X-ray diffraction (XRD) is shown in Fig. 2. The results correspond to films grown during 30 and 40 min on lime glass substrate previously activated by immersion in SnCl2 (0.01 M). It is also shown the diffractogram of the second film after thermal treatment at 300 1C during 30 min in N2 atmosphere. The results show a mixture of cubic (JCPDS 050566) and hexagonal (JCPDS 89-1363) phases, as observed by others [16,17]. Crystallinity of the films increases with deposition time. Fig. 3 shows the optical spectra of the films grown after different deposition time (25, 30 and 40 min). Because the reference for these optical measurements was air, the transmittance and reflectance of the bare glass substrate have been included for comparison. Fig. 4 shows the absorption spectra of the films calculated from the data in Fig. 3, using Eq. (1). The absorption edge increases with the time of deposition, towards values close to the gap energy of ZnS (3.6 eV) [1]. For the film prepared
ARTICLE IN PRESS B. Asenjo et al. / Solar Energy Materials & Solar Cells 92 (2008) 302–306
304
25min 30min 40min
105 α (cm-1)
-(311)cub/(118)hex
Intensity / a.u.
(220)cub (110)hex
-(111)cub -(008)hex
106
40 min+ 300 °C
104
40 min
103
ZnO
30 min
2.0 0
10
20
30
40
50
60
70
80
2.5
90
2θ Fig. 2. XRD of ZnS films. Solution composition: [TA] ¼ 0.5 M, [ZnSO4] ¼ 0.01 M, [HCl] ¼ 0.01 M and [CH3COOH] ¼ 0.3 M at 70 1C, pH ¼ 2.0. Films deposited during 20, 30 and 40 min, and after heat treatment (300 1C, 30 min N2).
ZnS
3.0 3.5 Energía (eV)
4.0
4.5
Fig. 4. Absorption coefficient spectra of the films, calculated according to Eq. (1) from results in Fig. 3.
ZnO
ZnS
Zn0
80 25min
T T, R (%)
30min 60
Intensity / u.a.
100
40min
70° 45° 20°
SLG 40
20
980
R
0 0
500
1000
1500
2000
2500
982
984
990 986 988 Kinetic energy / eV)
992
994
996
Fig. 5. XPS signal corresponding to Zn KMM Auger transition of a sample deposited during 30 min. Same solution composition as in Fig. 1.
λ (nm) Fig. 3. Optical transmittance and reflectance spectra for ZnS films. Solution composition: [TA] ¼ 0.5 M, [ZnSO4] ¼ 0.01 M, [HCl] ¼ 0.01 M and [CH3COOH] ¼ 0.3 M at 70 1C, pH ¼ 2.0. Films deposited during 25, 30 and 40 min.
Intensity / a.u.
at 25 min the absorption edge value is the lowest, possibly reflecting the formation of secondary compounds, such as oxides and hydroxides, which may be formed at the beginning of the process by heterogeneous growth mechanism. Composition of the films was studied with XPS. The analysis of Zn LMM signal is shown for the film prepared at 30 and 40 min in Figs. 5 and 6, respectively, under 201, 451 and 701 detection angle. For the film deposited at 30 min the signal shows little dependence on exit angle, positioned at the value of the ZnS (988.1 eV). However, the film prepared after 40 min deposition, the energy shifts upon increasing the exit angle, probably reflecting the presence of the oxide in the surface.
ZnS
Zn0
ZnO
70°
45°
20°
980
982
984
986 988 990 Kinetic energy / eV)
992
994
996
Fig. 6. XPS signal corresponding to Zn KMM Auger transition of a sample deposited during 40 min. Same solution composition as in Fig. 1.
ARTICLE IN PRESS B. Asenjo et al. / Solar Energy Materials & Solar Cells 92 (2008) 302–306
The elemental composition obtained by XPS is shown in Table 1. It is observed that the percentage of S is highest in all the samples while the relationship Zn/S oscillates between 0.5 and 0.7, indicating a different composition in the surface that in the bulk of the film. Results of EDAX (Table 2), more sensitive to the bulk of the film, show in fact a relationship Zn/S close to unit. Table 2 also shows the analysis of sulphur compounds present in the films, obtained from the S 2p signal at the three exit angles for electrons. Sulphide shows major proportion, above 80% for the film without heating. The increase in the sulphates
Table 1 XPS results of the elemental composition ZnS films at 451 detection angle prepared at two different deposition times: 30 min (a), 40 min (b), and 40 min with thermal treatment at 300 1C atmosphere N2 during 30 min (c) Time of deposition
(a) 30 min
(b) 40 min
(c) 40 min at 300 1C
Detection angle (1)
20
45
70
20
45
70
20
45
70
O Zn S C Zn/S Sulphates Sulphides Ek (eV; Zn LMM)
20.1 8.8 17.5 53.7 0.50 18.0 82
21.6 10.4 21.6 46.5 0.48 17.0 83 988.8
17.9 10.8 22.1 49.1 0.49 19.0 81
18.3 12.2 20.1 49.4 0.60 16.0 84
19.4 14.7 23.4 42.5 0.63 18.0 82 989.2
19.0 15.3 24.7 41.0 0.62 19.0 81
34.4 9.5 12.4 43.7 0.77 23.0 77
41.3 9.7 15.5 35.5 0.63 30.0 70 989.9
40.0 9.8 15.0 31.9 0.66 24.0 76
Other conditions as in Fig. 1. Table 2 Composition analysis of three films prepared at different time deposition determined by EDAX Films
Zn (at%)
S (at%)
Zn/S
25 min 30 min 40 min 40 min at 300 1C
44.86 55.05 50.02 53.62
55.14 44.95 49.98 46.38
0.8 1.2 1.0 1.2
305
proportion for the heated films reflects some oxidation of the sulphur. 3.1. CuInS2/ZnS/ZnO solar cell results solar cell results The solar cells results, corresponding to different buffer layer conditions, are shown in Table 3. The buffer layers were deposited on CuInS2 at 70 1C, at different deposition times, maintaining other conditions constant. In principle, they show lower performance than results obtained with other buffer films type, as In2S3 and In2S3–ZnS prepared under similar conditions [18]. Maximum efficiency of 5.3% is obtained with ZnS grown up to 25 min (Table 3). This buffer film is thinner and has higher oxygen proportion than those grown to 30 and 40 min (Table 2), which could result in a beneficial effect on absorber-buffer band alignment [19,20]. Below 25 min deposition time, the buffer film performance decreases, probably because it is not fully covering the substrate surface. In general, cell performances are below those obtained with CdS buffer layers, in the 9–10% range under the same preparation conditions. The most significant differences are encountered in the values of fill factor (FF) and short circuit current (Isc) (Table 3). Such low values are also reflected by low quantum efficiencies of the cells compared with CdS buffer cell (Fig. 7). Given the higher transparency of ZnS buffer layers, the decrease of the quantum efficiency on CuInS2/ZnS/ZnO must be attributed to the low quality of the absorbent/buffer interface. 4. Conclusions Buffer layers of ZnS composition have been prepared by the CBD method. The kinetics of growth of the layers is analysed with a QCM, showing four growth stages. Crystallographic analysis shows a mixture of phases and compositional analysis shows more stoichiometric composition in the bulk of the film that in the surface, where there is probably a higher concentration of zinc oxides and
Table 3 Deposition conditions for the growth of ZnS buffer films on CuInS2 from aqueous solution at 70 1C CuInS2/buffer
Time (min)a Thickness (nm) FF (%) Voc (mV) Isc (mA cm2) Rs (O cm2) Rp (O cm2) Z (%)
934-1/ZnS
934-2/ZnS
934-3/ZnS
934-5/ZnS
CdS
20 38 36.572.1 685710 13.471.1 1.5370.35 10479 3.470.4
25 64 44.372.0 613719 19.671.2 5.6772.67 343771 5.370.5
30 108 35.774.2 644716 14.173.5 2.5171.07 100714 3.371.1
40 195 46.177.2 632732 14.674.1 1.8670.92 160737 4.471.8
– – 66.071.8 69974 11.0470.28
Also results for a CdS buffer cell prepared under similar conditions. a Other solution components are 0.5 M TA, 0.025 M InCl3, 0.01 M HCl and 0.3 M CH3COOH.
10.270.2
ARTICLE IN PRESS B. Asenjo et al. / Solar Energy Materials & Solar Cells 92 (2008) 302–306
306
1.0 934-1/ZnS 934-2/ZnS 934-3/ZnS 934-5/ZnS CdS
0.8
QE
0.6
0.4
0.2
0.0 400
500
600 700 800 Wavelength (nm)
900
1000
Fig. 7. Quantum efficiency (QE) of CuInS2/ZnS/ZnO solar cells, with ZnS buffer deposited at 70 1C, other conditions as in Table 3. Also shown the QE results of cells from the same batch with CdS buffer layer.
hydroxides. Solar cells, CuInS2/ZnS/ZnO, have given efficiencies above 5%. Acknowledgements This work has been supported by FOTOFLEX Project (IV-PRICIT-CAM) and ARECES Project (Ramo´n Areces Fundation). References [1] R. Ortega Borges, D. Lincot, J. Vedel, in: Proceedings of the 11th E.C. Photovoltaic Solar Energy Conference, 1992, p. 862.
[2] E.Y.M. Lee, N.H. Tran, R.N. Lamb, Appl. Surf. Sci. 241 (3–4) (2005) 493. [3] O¨. Mustafa, M. Bedir, S. Ocak, R.G. Yildirim, J. Mater. Sci.: Mater. Electron. 15 (5) (2007) 505. [4] F. Go¨de, C. Gu¨mu¨s, M. Zor, J. Crystal Growth 299 (1) (2007) 136. [5] C. Kaufmann, P.J. Dobson, in: Proceedings of 28th IEEE Photovoltaic Specialists Conference, 2000, p. 688. [6] R.N. Bhattacharya, K. Ramanathan, L. Gedvilas, B. Keyes, J. Phys. Chem. Sol. 66 (2006) 1862. [7] B. Mokili, Y. Charreire, R. Cortes, D. Lincot, Thin Solid Films 288 (1996) 21. [8] S. Neve, W. Bohne, J. Klaer, R. Klenk, R. Scheer, in: Proceedings of 17th European Photovoltaic Solar Energy Conference, Munich, 2001, p. 1102. [9] J.P. O’Hare, K. Govender, P. O’Brien, D. Smyth-Boyle, Mater. Res. Soc. Symp. Proc. 668 (2001) H8.15.1. [10] L.V. Makhova, I. Konovalov, R. Szargan, N. Aschkenov, M. Schubert, T. Chasse´, Phys. Status Solidi C 2 (3) (2005) 1206. [11] A.M. Chaparro, M.T. Gutie´rrez, J. Herrero, Electrochim. Acta 47 (2001) 977. [12] G. Connel, A. Lewis, Phys. Status Solidi B 60 (1973) 291. [13] J. Klaer, J. Bruns, R. Henninger, K. Siemer, R. Klenk, K. Ellmer, D. Bra¨unig, Semicond. Sci. Technol. 13 (1998) 1456. [14] A. Eshuis, G.R.A. van Elderen, C.A.J. Koning, Colloids Surf., A: Physicochem. Eng. Aspects 151 (1999) 505. [15] C.-Y. Yeh, Z.W. Lu, S. Froyen, A. Zunger, Phys. Rev. B46 (1992) 10086. [16] J. Cheng, D.B. Fan, H. Wang, B.W. Liu, Y.C. Zhang, H. Yan, Semicond. Sci. Technol. 18 (2003) 676. [17] T. Ben Nasr, N. Kamoun, M. Kanzari, R. Bennaceu, Thin Solid Films 500 (2006) 4. [18] B. Asenjo, A.M. Chaparro, M.T. Gutie´rrez, J. Herrero, J. Klaer, Thin Solid Films 515 (2007) 6036. [19] S.B. Zhang, S.-H. Wei, A. Zunger, J. Appl. Phys. 83 (1998) 3192. [20] C. Persson, C. Platzer-Bjo¨rkman, J. Malmstro¨m, T. To¨rndahl, M. Edoff, Phys. Rev. Lett 97 (2006) 146403.