- Email: [email protected]
Earth abundant thin film solar cells from co-evaporated Cu2SnS3 absorber layers
Accepted Manuscript Earth abundant thin film solar cells from co-evaporated Cu2SnS3 absorber layers Jose A. Marquez Prieto, Sergiu Levcenko, Justus Just, Hannes Hampel, Ian Forbes, Nicola M. Pearsall, Thomas Unold PII:
S0925-8388(16)32329-5
DOI:
10.1016/j.jallcom.2016.07.293
Reference:
JALCOM 38462
To appear in:
Journal of Alloys and Compounds
Received Date: 14 March 2016 Revised Date:
4 July 2016
Accepted Date: 26 July 2016
Please cite this article as: J.A. Marquez Prieto, S. Levcenko, J. Just, H. Hampel, I. Forbes, N.M. Pearsall, T. Unold, Earth abundant thin film solar cells from co-evaporated Cu2SnS3 absorber layers, Journal of Alloys and Compounds (2016), doi: 10.1016/j.jallcom.2016.07.293. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Earth abundant thin film solar cells from co-evaporated Cu2SnS3 absorber layers Jose A. Marquez Prieto1*, Sergiu Levcenko2, Justus Just2, Hannes Hampel2, Ian Forbes1, Nicola M. Pearsall1 and Thomas Unold2
RI PT
Corresponding author: [email protected];
1. Northumbria Photovoltaic Applications Group, Department of Physics and Electrical Engineering, Northumbria University, Newcastle upon Tyne, UK
SC
2. Helmholtz-Zentrum Berlin for Materials and Energy, Hahn-Meitner-Platz 1, 14109 Berlin, Germany Abstract:
Introduction
AC C
1.
EP
TE D
M AN U
Cu2SnS3 (CTS) is starting to gain interest in the PV research community as an alternative earth abundant absorber for thin film photovoltaics. In this work, the structure, morphology and the composition of the CTS absorbers as well as their influence on the optoelectronic properties of the solar cells are analysed. The synthesis of Cu-Sn-S thin films by coevaporation at a nominal temperature of 400 ºC is presented. A combination of X-ray diffraction, Raman and UV-Vis spectroscopy suggests that the Cu2SnS3 is crystallising in a cubic structure with disorder in the Cu and Sn sites, leading to substantial band tailing. The best device was fabricated from absorbers exhibiting a Cu/Sn ratio of approximately 1.7 and had an efficiency of 1.8 %, a short circuit current of 28 mA cm-2, and an open circuit voltage of 147 mV with a fill factor of 42.9 %. From the quantum efficiency measurement, we estimate a band gap of 1.06 eV for the CTS absorber material. Capacitance-voltage measurements show charge carrier concentrations between 4 and 6 x 1016 cm-3.
Cu-Sn-S derived materials are increasing interest as absorber layers produced from earth abundant elements for application in thin film solar cells. Efficiencies up to 4.6% have been already achieved using the same device architecture used for Cu(In,Ga)Se2(CIGS) and for Cu2ZnSn(S,Se)4 (CZTSSe) based solar cells[1]. In the Cu-Sn-S system, several compounds with different stoichiometry have been reported in literature, including Cu2SnS3[2-8], Cu3SnS4[3, 6, 9-11], Cu4SnS4[2], Cu2Sn3S7[2] and Cu4Sn7S16[12]. Particularly, Cu2SnS3 is so far the ternary compound of the Cu-Sn-S system that has been studied most for solar cells applications. Several synthesis approaches have been demonstrated so far to be suitable for producing Cu2SnS3 films including sputtering [6], reactive sputtering [13], electrodeposition [5, 14], pulsed layer deposition [15] and several non-vacuum routes [16, 17].
ACCEPTED MANUSCRIPT
SC
RI PT
Cu2SnS3, has been reported to crystallize in several structural polytypes including a cubic (F43m)[3, 6], a tetragonal (I-42m)[3, 6, 12] and a monoclinic (C1c1) [4, 5, 7, 8, 18, 19] structure depending on the temperature during synthesis. With increasing temperature, a change in the crystal structure to unit cells with higher symmetry has been predicted, changing from monoclinic to orthorhombic, then tetragonal and cubic [20]. However, experimental results in literature suggest that for synthesis process at temperatures below 550ºC, Cu2SnS3 adopts a cubic or a tetragonal structure [3, 6] and when the synthesis temperature is above 550ºC it adopts a monoclinic structure [8, 19, 21]. A large discrepancy in the value of the bandgap of Cu2SnS3 in its different crystal structures is also found in literature. For the monoclinic form, the presence of two bandgaps was reported, one at 0.920.93 eV and a second one around 0.99 eV [5, 8, 18, 19]. Fernandes et al. reported a bandgap of 1.35 eV for the tetragonal Cu2SnS3 and 0.96 eV for the cubic structure [3].
M AN U
Most of the Cu2SnS3 absorber layers that have been used in solar cell devices have been synthesised in a two stage process with a reactive annealing step at temperatures higher than 550ºC under a sulphur or sulphur-tin atmosphere. This high temperature step resulted in monoclinic Cu2SnS3 [1, 5, 21]. However, it is difficult to find reports of solar cells prepared at low or moderate temperature processes, for which the Cu2SnS3 is not monoclinic. In addition to the fact that in low temperature synthesis, the grain growth is more limited, the cubic CTS structure might present mid gap states and band tailing, making it difficult to make efficient solar cells out of this structure [22].
2.1.
Experimental
AC C
2.
EP
TE D
In this work, we report the synthesis of Cu-Sn-S absorber layers by a single step coevaporation method with a nominal substrate temperature of 400 ºC. Microstructural analysis of the absorber layers based on SEM, Raman spectroscopy and X-Ray diffraction is presented. It was found that Cu2SnS3 crystallises mostly in the cubic structure, however the presence of other Cu-Sn-S phases in the absorber layers have been identified. Solar cells fabricated from these absorber layers show a power conversion efficiency of 1.8 %, mainly limited by a low value of the open circuit voltage.
CTS absorber synthesis
A series of CTS absorber layers was deposited using a PVD system equipped with elemental Sn and Cu thermal evaporation sources and a sulphur evaporator with a cracking unit. The temperature of the Cu and Sn source and S cracking unit were 1285, 1295 and 500 oC respectively. 50x12.5 mm2 molybdenum coated soda-lime glass substrates were used in this study, which were rotated during the deposition. The total pressure of the chamber during the deposition was around 1.5x10-3Pa. For all the depositions the substrate temperature was set to a nominal temperature of 400 ºC. Infrared reflectometry was used for in situ process control. The deposition was stopped after achieving a layer thickness of approximately 1 µm.
ACCEPTED MANUSCRIPT 2.2.
CTS absorber characterisation
Device fabrication and characterisation
SC
2.3.
RI PT
The CTS thin films were measured with a PANalytical X’Pert MPD Pro X-ray diffractometer in Bragg-Brentano (BB) configuration using a Cu Kα radiation source (λ=0.15406 nm).The composition of the samples was determined using a FEI Quanta 200 scanning electron microscope (SEM) equipped with an Oxford Instruments energy dispersive X-ray analyser (EDS) with an acceleration voltage of 20kV from the top view. The cross section images of the solar cells were acquired with a Zeiss Gemini SEM. Raman scattering measurements were performed with He-Ne laser (633nm) excitation using a CCD detector coupled to a 0.5m spectrometer. The transmittance/reflectance of the absorber layers were measured with Perkin-Ellmer UV-Vis setup after a lift off the films from the Mo substrate.
TE D
M AN U
A CdS buffer layer of approximately 50 nm in thickness was deposited by chemical bath deposition (CBD) on the absorber layers followed by a i-ZnO/ZnO:Al window layer deposited by RF sputtering. Ni:Al metallic front contact grids were evaporated on top of the window. No etching treatment was applied before the CdS deposition. Cells of area 0.5 cm2 were mechanically scribed. Illuminated J-V curves were measured under 100mW/cm2 simulated AM1.5 solar illumination calibrated with a Si reference cell and the external quantum efficiency (EQE) was measured using a lock-in amplifier combined with a 1/4m monochromator. Capacitance-voltage (CV) measurements were performed with an HP4284 LCR meter and four-point probes.
Results and discussion
3.1.
Microstructure and morphology of CTS absorber layers
EP
3.
AC C
The absorber layers showed a Cu/Sn ratio of around 1.7. Figure 1 shows a cross sectional view of the solar cells processed for this study. The CTS absorber layer is about 1 µm thick and shows columnar like morphology suggesting that the films are highly textured. A similar morphology has been also reported for CTS samples grown by RF sputtering from Cu2S and SnS targets heating the substrates to around 300º C [6], suggesting that this columnar growth type is typical for low temperature synthesis.
RI PT
ACCEPTED MANUSCRIPT
Figure 1. SEM cross section of a complete CTS device. The different layers are identified in the right hand side of the image.
AC C
EP
TE D
M AN U
SC
Figure 2 shows the X-Ray Diffraction (XRD) pattern of CTS absorber layer acquired in Bragg-Brentano configuration. An intense peak appears at 28.3º that can be attributed to Cu2SnS3 in the cubic (F-43m, ICSD 43532), the tetragonal (I-42m, ICSD 50965) or the monoclinic (C1c1, ICSD 91762) structure. No other peaks can be observed related to this phase. Therefore, it is difficult to determine the crystal structure only based on the information provided by the XRD pattern. The single peak suggests that the films are preferentially orientated, which is consistent with the columnar growth morphology observed for the cross sectional images in figure 1. The inset of figure 2 shows a magnification of the XRD pattern between 20 and 50º. The diffractogram shows the presence of low intensity peaks at 27.4 and 30.7º that can be attributed to the presence of orthorhombic Cu3SnS4[11]. The presence of Cu3SnS4 in films has been already reported to coexist with Cu2SnS3 [6, 22], however the relatively low intensity of these reflexes in comparison with the strong peak attributed to Cu2SnS3 suggests that the amount of the Cu3SnS4 orthorhombic phase is rather low. No other phases have been identified with this technique.
Cu2SnS3 Cu3SnS4
20
Mo Cu3SnS4
30
40
2 theta (º)
Mo
RI PT
Intensity (arb. units)
Cu2SnS3
50
SC
Intensity (arb. units)
ACCEPTED MANUSCRIPT
Mo
Mo
20
30
40
50
60
M AN U
10
70
80
2 theta (º)
Figure 2. X-Ray diffraction pattern of the CTS absorber layer. The inset shows a magnification of low intensity range between 20 and 50º. The phases assigned to each peak are labelled in the plots.
AC C
EP
TE D
In order to confirm the crystal structure of the CTS absorber layers Raman spectroscopy was performed on the sample and it is shown in figure 3. The peaks observed in the spectrum have been qualitatively fitted with Lorentzian functions to determine the position of the observed Raman modes. Two high intensity peaks at 303 and 359 cm-1 suggest that Cu2SnS3 is mainly present in the cubic structure, in accordance with the studies reported by Fernandes et al. [3, 23]. Unfortunately, no calculations of the phonon spectra of the different crystal structures of Cu2SnS3 have been reported up to now and the assignments for the cubic Cu2SnS3 are made based on reported experimental data. In addition, lower intensity peaks at 319 cm-1 and 337-346 cm-1can be observed. In the literature, peaks at these positions have been correlated with the presence of orthorhombic Cu3SnS4 [11, 23], which agrees with its identification by XRD presented in figure 2. The position of the modes for the orthorhombic Cu3SnS4 to perform the fitting for this phase was adjusted according to the phonon spectrum calculated by Dzhagan et al [11]. Raman measurements with λexc =633nm would potentially be very sensitive to the modes of Cu3SnS4, due to near resonant conditions [24], since a bandgap of 1.6 eV has been reported for this phase [3]. This would explain the fact that we clearly observe modes related to this phase even when its concentration in the film is low, as suggested by the low intensity peaks in the XRD patterns corresponding to Cu3SnS4. Also a Raman mode at 265 cm-1 is observed, which we are not able to attribute to a specific phase. Although calculations have predicted Raman active modes close to this frequency [11], the experimental data in literature for orthorhombic Cu3SnS4 does not agree with this mode being the most intense for this phase [11].
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
A bandgap value of 1.02 eV was calculated form the Tauc’s plot generated with the optical transmission spectrum of a lifted-off CTS absorber as shown in figure 3b. Strong subgap absorption can be observed up to 0.85 eV, in good agreement with theoretical calculations of the density of states for the disorder cubic CTS that have shown strong band tails for this material [22].
EP
Photovoltaic devices.
AC C
3.2.
TE D
Figure 3. (a) Raman spectrum of the CTS absorber (dark grey). Lorentzian fitting of the Cu2SnS3 cubic modes (red), Cu3SnS4 (blue) and an unidentified mode (green). The black line shows the sum of the fits. (b) Tauc’s plot generated from the optical transmission and reflection spectrum of a lifted-off CTS absorber.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 4. (a) Dark and illuminated JV curves of the best CTS cell. (b) External (black line) and internal (red line) quantum efficiency of the best solar cell. (c) CV profiles measured at 10 kHz, 100 kHz and 1000 kHz.
The current-voltage (J-V) characteristic of the best performing device is shown in figure 4a. The efficiency of the device was 1.8%, with a short circuit current density (JSC) of 28 mA/cm2, an open circuit voltage (VOC) of 147 mV and a fill factor (FF) of 44% for a total cell area of 0.5 cm2. The reason for the low value of VOC is not well understood but several possibilities can be considered. The cubic structure of Cu2SnS3 has been reported to be highly disordered implying band tailing, compositional inhomogeneities at the nanometre scale and
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
potential fluctuations [22]. Also, a cliff-like band offset has been reported for the CdS-CTS junction [25]. Assuming that this is the case for this device, a reduction of the interface bandgap would reduce VOC and increase the recombination at the interface of the device [26]. In addition, the presence of other phases found in the absorber layer, such as Cu3SnS4, could contribute detrimentally, increasing the metallic character of the absorber layers and decreasing the performance of the solar cells due to an increase of shunting paths [6]. The External Quantum Efficiency (EQE) and the Internal Quantum Efficiency (IQE) are shown in figure 4b. The IQE (λ) was calculated from the ratio of the EQE (λ) and the spectral refelectance of the device as IQE (λ) = EQE (λ) / [1−R (λ)] . A step between 650 and 800 nm observed in the EQE, which is not evident in the IQE, suggesting that it is due to interferences of the device layers. At wavelengths above the bandgap of CdS (510-520nm) both, the EQE and IQE start to decay significantly, suggesting that a measurable contribution to the JSC occurs from hole generation in the CdS. This contribution ceases for wavelengths with energies below the bandgap of CdS in agreement with a small collection function in the bulk of the absorber [27]. Nakashima et al. reported the appearance of the same decay when the EQE was measured without white light bias [1], suggesting that the light intensity dependent properties of CTS solar cells should be further studied to contribute to the progress of this material. The bandgap was calculated from the inflection point of the EQE and IQE giving a value of approximately 1.06 eV. Strong band tailing is also observed in the QE, in agreement with the Tauc’s plot presented in figure 3b and in accordance with previously reported theoretical calculations for the disorder cubic structure [22]. The presence of this strong band tailing could be a potential issue for the development of CTS solar cells when the absorber layer adopts the cubic structure. This band tail is not observed in the EQE of devices made out of the monoclinic Cu2SnS3 where two transitions are observed at around 0.93 and 0.99 eV [1], suggesting that this crystal structure could be a better candidate for CTS solar cells. JSC has been calculated by integrating the EQE with the AM 1.5 solar spectrum yielding a value of JSC (EQE) =16.4 mA/cm2 which is significantly lower than the value measured in the JV curve (28.0 mA/cm2). This behaviour has been previously observed in other CTS solar cells found in literature when the EQE has been measured without white light bias [1], suggesting again that the optoelectronic properties of these CTS solar cells are affected by the light intensity. CV measurements were performed on the CTS device and are shown in figure 4c. Charge carrier concentrations of 3 - 5 x 1016 cm-3and a depletion region width of around 120 nm were estimated for this solar cell by assuming a static dielectric constant ε0=10. The charge carrier concentration values measured for this solar cell are significantly lower than what was previously reported for cubic CTS, where the lowest values were around 7 x 1017 cm-3 and 1021 cm-3 when Cu3SnS4 was present. These values are considered to be too high for a p-type absorber layer in a photovoltaic device [6].
4. Conclusion
ACCEPTED MANUSCRIPT
SC
RI PT
In this work, we have reported the synthetisation of CTS absorber layers by single step coevaporation with a nominal substrate temperature of 400ºC. A combination of Raman, SEM and XRD analysis leads us to conclude that the Cu2SnS3 is highly textured and adopts a disordered cubic structure. The presence of Cu3SnS4 was also detected in the samples. A 1.8% efficient CTS/CdS/i-ZnO/Al:ZnO solar cell was obtained with JSC of 28 mA cm-2, VOC of 147 mV and a FF of 42.9 %. A bandgap value of 1.06 eV was extracted from the inflection point of the EQE. The Jsc calculated by integrating the EQE with the AM 1.5 solar spectrum shows disagreement with the measured JSC indicating that the devices suffer from changes in its electronic properties depending on the intensity of the illumination. A charge carrier concentration of 4 x 1016 cm-3 is extracted from C-V measurements with a depletion region of approximately 120 nm. The limitations of the CTS solar cells presented in this work provide insights for the future development of higher efficiency thin film solar cells based on the Cu2SnS3 ternary compound.
M AN U
5. Acknowledgements
References
TE D
This work has been undertaken within a Marie Curie, Initial Training Network (ITN) project, “KESTCELLS” co-financed by the Seventh Framework Programme of the European Commission (FP7/2007-2013) under grant agreement nº316488. The authors gratefully acknowledge support by the Helmholtz Association Initiative and Network Fund (HNSEIProject SO-075). A. Irkhina, C. Ferber and L. Steinkopf from HZB are also acknowledged for helping with processing and characterisation of the solar cells
AC C
EP
[1] M. Nakashima, J. Fujimoto, T. Yamaguchi, M. Izaki, Cu2SnS3thin-film solar cells fabricated by sulfurization from NaF/Cu/Sn stacked precursor, Applied Physics Express, 8 (2015) 042303. [2] S. Fiechter, M. Martinez, G. Schmidt, W. Henrion, Y. Tomm, Phase relations and optical properties of semiconducting ternary sulfides in the system Cu–Sn–S, Journal of Physics and Chemistry of Solids, 64 (2003) 1859-1862. [3] P.A. Fernandes, P.M.P. Salomé, A.F.d. Cunha, A study of ternary Cu2SnS3and Cu3SnS4thin films prepared by sulfurizing stacked metal precursors, Journal of Physics D: Applied Physics, 43 (2010) 215403. [4] D.M. Berg, R. Djemour, L. Gütay, S. Siebentritt, P.J. Dale, X. Fontane, V. Izquierdo-Roca, A. PérezRodriguez, Raman analysis of monoclinic Cu2SnS3 thin films, Applied Physics Letters, 100 (2012) 192103. [5] D.M. Berg, R. Djemour, L. Gütay, G. Zoppi, S. Siebentritt, P.J. Dale, Thin film solar cells based on the ternary compound Cu2SnS3, Thin Solid Films, 520 (2012) 6291-6294. [6] L.L. Baranowski, P. Zawadzki, S. Christensen, D. Nordlund, S. Lany, A.C. Tamboli, L. Gedvilas, D.S. Ginley, W. Tumas, E.S. Toberer, A. Zakutayev, Control of Doping in Cu2SnS3through Defects and Alloying, Chemistry of Materials, 26 (2014) 4951-4959. [7] T. Nomura, T. Maeda, T. Wada, Fabrication of Cu2SnS3solar cells by screen-printing and highpressure sintering process, Japanese Journal of Applied Physics, 53 (2014) 05FW01.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
[8] A. Kanai, H. Araki, A. Takeuchi, H. Katagiri, Annealing temperature dependence of photovoltaic properties of solar cells containing Cu2SnS3 thin films produced by co-evaporation, physica status solidi (b), 252 (2015) 1239-1243. [9] M. Bouaziz, J. Ouerfelli, M. Amlouk, S. Belgacem, Structural and optical properties of Cu3SnS4sprayed thin films, physica status solidi (a), 204 (2007) 3354-3360. [10] V.A. Kovalenker, Kuramite, Cu3SnS4, a new mineral of the stannite group, International Geology Review, 23 (2010) 365-370. [11] V.M. Dzhagan, A.P. Litvinchuk, M. Kruszynska, J. Kolny-Olesiak, M.Y. Valakh, D.R.T. Zahn, Raman Scattering Study of Cu3SnS4Colloidal Nanocrystals, The Journal of Physical Chemistry C, 118 (2014) 27554-27558. [12] X.-a. Chen, H. Wada, A. Sato, M. Mieno, Synthesis, Electrical Conductivity, and Crystal Structure of Cu4Sn7S16and Structure Refinement of Cu2SnS3, Journal of Solid State Chemistry, 139 (1998) 144-151. [13] Y. Ren, J.J. Scragg, T. Ericson, T. Kubart, C. Platzer-Björkman, Reactively sputtered films in the CuxS–ZnS–SnSy system: From metastability to equilibrium, Thin Solid Films, 582 (2015) 208-214. [14] J. Koike, K. Chino, N. Aihara, H. Araki, R. Nakamura, K. Jimbo, H. Katagiri, Cu$_{2}$SnS$_{3}$ Thin-Film Solar Cells from Electroplated Precursors, Japanese Journal of Applied Physics, 51 (2012) 10NC34. [15] S.A. Vanalakar, G.L. Agawane, A.S. Kamble, C.W. Hong, P.S. Patil, J.H. Kim, Fabrication of Cu2SnS3 thin film solar cells using pulsed laser deposition technique, Solar Energy Materials and Solar Cells, 138 (2015) 1-8. [16] S.A. Vanalakar, G.L. Agawane, S.W. Shin, H.S. Yang, P.S. Patil, J.Y. Kim, J.H. Kim, Non-vacuum mechanochemical route to the synthesis of Cu2SnS3 nano-ink for solar cell applications, Acta Materialia, 85 (2015) 314-321. [17] Q. Chen, X. Dou, Y. Ni, S. Cheng, S. Zhuang, Study and enhance the photovoltaic properties of narrow-bandgap Cu2SnS3 solar cell by p-n junction interface modification, Journal of colloid and interface science, 376 (2012) 327-330. [18] N. Aihara, H. Araki, A. Takeuchi, K. Jimbo, H. Katagiri, Fabrication of Cu2SnS3thin films by sulfurization of evaporated Cu-Sn precursors for solar cells, physica status solidi (c), 10 (2013) 10861092. [19] Y. Dong, J. He, L. Sun, Y. Chen, P. Yang, J. Chu, Effect of sulfurization temperature on properties of Cu2SnS3 thin films and solar cells prepared by sulfurization of stacked metallic precursors, Materials Science in Semiconductor Processing, 38 (2015) 171-176. [20] A. Zakutayev, L.L. Baranowski, A.W. Welch, C.A. Wolden, E.S. Toberer, Comparison of Cu2SnS3 and CuSbS2 as potential solar cell absorbers, in: Photovoltaic Specialist Conference (PVSC), 2014 IEEE 40th, 2014, pp. 2436-2438. [21] N. Aihara, A. Kanai, K. Kimura, M. Yamada, K. Toyonaga, H. Araki, A. Takeuchi, H. Katagiri, Sulfurization temperature dependences of photovoltaic properties in Cu2SnS3-based thin-film solar cells, Japanese Journal of Applied Physics, 53 (2014) 05FW13. [22] L.L. Baranowski, K. McLaughlin, P. Zawadzki, S. Lany, A. Norman, H. Hempel, R. Eichberger, T. Unold, E.S. Toberer, A. Zakutayev, Effects of disorder on carrier transport in Cu2SnS3, Physical Review Applied, 4 (2015) 044017. [23] P.A. Fernandes, P.M.P. Salomé, A.F. da Cunha, CuxSnSx+1(x = 2, 3) thin films grown by sulfurization of metallic precursors deposited by dc magnetron sputtering, physica status solidi (c), (2010). [24] A. Fairbrother, V. Izquierdo-Roca, X. Fontané, M. Ibáñez, A. Cabot, E. Saucedo, A. PérezRodríguez, ZnS grain size effects on near-resonant Raman scattering: optical non-destructive grain size estimation, CrystEngComm, 16 (2014) 4120-4125. [25] S. Sato, H. Sumi, G. Shi, M. Sugiyama, Investigation of the sulfurization process of Cu2SnS3 thin films and estimation of band offsets of Cu2SnS3-related solar cell structure, physica status solidi (c), 12 (2015) 757-760.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
[26] S. Siebentritt, Why are kesterite solar cells not 20% efficient?, Thin Solid Films, 535 (2013) 1-4. [27] R.S.a.H.-W. Schock, Chalcogenide Photovoltaics: Physics, Technologies, and Thin Film Devices, 2011.
ACCEPTED MANUSCRIPT
Highlights
AC C
EP
TE D
M AN U
SC
RI PT
1. Cu2SnS3 thin films deposited by co-evaporation 2. We fabricated a Cu2SnS3 thin film solar cell 3. The Cu2SnS3 absorber layer crystallises in a disorder cubic structure
S0925-8388(16)32329-5
DOI:
10.1016/j.jallcom.2016.07.293
Reference:
JALCOM 38462
To appear in:
Journal of Alloys and Compounds
Received Date: 14 March 2016 Revised Date:
4 July 2016
Accepted Date: 26 July 2016
Please cite this article as: J.A. Marquez Prieto, S. Levcenko, J. Just, H. Hampel, I. Forbes, N.M. Pearsall, T. Unold, Earth abundant thin film solar cells from co-evaporated Cu2SnS3 absorber layers, Journal of Alloys and Compounds (2016), doi: 10.1016/j.jallcom.2016.07.293. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Earth abundant thin film solar cells from co-evaporated Cu2SnS3 absorber layers Jose A. Marquez Prieto1*, Sergiu Levcenko2, Justus Just2, Hannes Hampel2, Ian Forbes1, Nicola M. Pearsall1 and Thomas Unold2
RI PT
Corresponding author: [email protected];
1. Northumbria Photovoltaic Applications Group, Department of Physics and Electrical Engineering, Northumbria University, Newcastle upon Tyne, UK
SC
2. Helmholtz-Zentrum Berlin for Materials and Energy, Hahn-Meitner-Platz 1, 14109 Berlin, Germany Abstract:
Introduction
AC C
1.
EP
TE D
M AN U
Cu2SnS3 (CTS) is starting to gain interest in the PV research community as an alternative earth abundant absorber for thin film photovoltaics. In this work, the structure, morphology and the composition of the CTS absorbers as well as their influence on the optoelectronic properties of the solar cells are analysed. The synthesis of Cu-Sn-S thin films by coevaporation at a nominal temperature of 400 ºC is presented. A combination of X-ray diffraction, Raman and UV-Vis spectroscopy suggests that the Cu2SnS3 is crystallising in a cubic structure with disorder in the Cu and Sn sites, leading to substantial band tailing. The best device was fabricated from absorbers exhibiting a Cu/Sn ratio of approximately 1.7 and had an efficiency of 1.8 %, a short circuit current of 28 mA cm-2, and an open circuit voltage of 147 mV with a fill factor of 42.9 %. From the quantum efficiency measurement, we estimate a band gap of 1.06 eV for the CTS absorber material. Capacitance-voltage measurements show charge carrier concentrations between 4 and 6 x 1016 cm-3.
Cu-Sn-S derived materials are increasing interest as absorber layers produced from earth abundant elements for application in thin film solar cells. Efficiencies up to 4.6% have been already achieved using the same device architecture used for Cu(In,Ga)Se2(CIGS) and for Cu2ZnSn(S,Se)4 (CZTSSe) based solar cells[1]. In the Cu-Sn-S system, several compounds with different stoichiometry have been reported in literature, including Cu2SnS3[2-8], Cu3SnS4[3, 6, 9-11], Cu4SnS4[2], Cu2Sn3S7[2] and Cu4Sn7S16[12]. Particularly, Cu2SnS3 is so far the ternary compound of the Cu-Sn-S system that has been studied most for solar cells applications. Several synthesis approaches have been demonstrated so far to be suitable for producing Cu2SnS3 films including sputtering [6], reactive sputtering [13], electrodeposition [5, 14], pulsed layer deposition [15] and several non-vacuum routes [16, 17].
ACCEPTED MANUSCRIPT
SC
RI PT
Cu2SnS3, has been reported to crystallize in several structural polytypes including a cubic (F43m)[3, 6], a tetragonal (I-42m)[3, 6, 12] and a monoclinic (C1c1) [4, 5, 7, 8, 18, 19] structure depending on the temperature during synthesis. With increasing temperature, a change in the crystal structure to unit cells with higher symmetry has been predicted, changing from monoclinic to orthorhombic, then tetragonal and cubic [20]. However, experimental results in literature suggest that for synthesis process at temperatures below 550ºC, Cu2SnS3 adopts a cubic or a tetragonal structure [3, 6] and when the synthesis temperature is above 550ºC it adopts a monoclinic structure [8, 19, 21]. A large discrepancy in the value of the bandgap of Cu2SnS3 in its different crystal structures is also found in literature. For the monoclinic form, the presence of two bandgaps was reported, one at 0.920.93 eV and a second one around 0.99 eV [5, 8, 18, 19]. Fernandes et al. reported a bandgap of 1.35 eV for the tetragonal Cu2SnS3 and 0.96 eV for the cubic structure [3].
M AN U
Most of the Cu2SnS3 absorber layers that have been used in solar cell devices have been synthesised in a two stage process with a reactive annealing step at temperatures higher than 550ºC under a sulphur or sulphur-tin atmosphere. This high temperature step resulted in monoclinic Cu2SnS3 [1, 5, 21]. However, it is difficult to find reports of solar cells prepared at low or moderate temperature processes, for which the Cu2SnS3 is not monoclinic. In addition to the fact that in low temperature synthesis, the grain growth is more limited, the cubic CTS structure might present mid gap states and band tailing, making it difficult to make efficient solar cells out of this structure [22].
2.1.
Experimental
AC C
2.
EP
TE D
In this work, we report the synthesis of Cu-Sn-S absorber layers by a single step coevaporation method with a nominal substrate temperature of 400 ºC. Microstructural analysis of the absorber layers based on SEM, Raman spectroscopy and X-Ray diffraction is presented. It was found that Cu2SnS3 crystallises mostly in the cubic structure, however the presence of other Cu-Sn-S phases in the absorber layers have been identified. Solar cells fabricated from these absorber layers show a power conversion efficiency of 1.8 %, mainly limited by a low value of the open circuit voltage.
CTS absorber synthesis
A series of CTS absorber layers was deposited using a PVD system equipped with elemental Sn and Cu thermal evaporation sources and a sulphur evaporator with a cracking unit. The temperature of the Cu and Sn source and S cracking unit were 1285, 1295 and 500 oC respectively. 50x12.5 mm2 molybdenum coated soda-lime glass substrates were used in this study, which were rotated during the deposition. The total pressure of the chamber during the deposition was around 1.5x10-3Pa. For all the depositions the substrate temperature was set to a nominal temperature of 400 ºC. Infrared reflectometry was used for in situ process control. The deposition was stopped after achieving a layer thickness of approximately 1 µm.
ACCEPTED MANUSCRIPT 2.2.
CTS absorber characterisation
Device fabrication and characterisation
SC
2.3.
RI PT
The CTS thin films were measured with a PANalytical X’Pert MPD Pro X-ray diffractometer in Bragg-Brentano (BB) configuration using a Cu Kα radiation source (λ=0.15406 nm).The composition of the samples was determined using a FEI Quanta 200 scanning electron microscope (SEM) equipped with an Oxford Instruments energy dispersive X-ray analyser (EDS) with an acceleration voltage of 20kV from the top view. The cross section images of the solar cells were acquired with a Zeiss Gemini SEM. Raman scattering measurements were performed with He-Ne laser (633nm) excitation using a CCD detector coupled to a 0.5m spectrometer. The transmittance/reflectance of the absorber layers were measured with Perkin-Ellmer UV-Vis setup after a lift off the films from the Mo substrate.
TE D
M AN U
A CdS buffer layer of approximately 50 nm in thickness was deposited by chemical bath deposition (CBD) on the absorber layers followed by a i-ZnO/ZnO:Al window layer deposited by RF sputtering. Ni:Al metallic front contact grids were evaporated on top of the window. No etching treatment was applied before the CdS deposition. Cells of area 0.5 cm2 were mechanically scribed. Illuminated J-V curves were measured under 100mW/cm2 simulated AM1.5 solar illumination calibrated with a Si reference cell and the external quantum efficiency (EQE) was measured using a lock-in amplifier combined with a 1/4m monochromator. Capacitance-voltage (CV) measurements were performed with an HP4284 LCR meter and four-point probes.
Results and discussion
3.1.
Microstructure and morphology of CTS absorber layers
EP
3.
AC C
The absorber layers showed a Cu/Sn ratio of around 1.7. Figure 1 shows a cross sectional view of the solar cells processed for this study. The CTS absorber layer is about 1 µm thick and shows columnar like morphology suggesting that the films are highly textured. A similar morphology has been also reported for CTS samples grown by RF sputtering from Cu2S and SnS targets heating the substrates to around 300º C [6], suggesting that this columnar growth type is typical for low temperature synthesis.
RI PT
ACCEPTED MANUSCRIPT
Figure 1. SEM cross section of a complete CTS device. The different layers are identified in the right hand side of the image.
AC C
EP
TE D
M AN U
SC
Figure 2 shows the X-Ray Diffraction (XRD) pattern of CTS absorber layer acquired in Bragg-Brentano configuration. An intense peak appears at 28.3º that can be attributed to Cu2SnS3 in the cubic (F-43m, ICSD 43532), the tetragonal (I-42m, ICSD 50965) or the monoclinic (C1c1, ICSD 91762) structure. No other peaks can be observed related to this phase. Therefore, it is difficult to determine the crystal structure only based on the information provided by the XRD pattern. The single peak suggests that the films are preferentially orientated, which is consistent with the columnar growth morphology observed for the cross sectional images in figure 1. The inset of figure 2 shows a magnification of the XRD pattern between 20 and 50º. The diffractogram shows the presence of low intensity peaks at 27.4 and 30.7º that can be attributed to the presence of orthorhombic Cu3SnS4[11]. The presence of Cu3SnS4 in films has been already reported to coexist with Cu2SnS3 [6, 22], however the relatively low intensity of these reflexes in comparison with the strong peak attributed to Cu2SnS3 suggests that the amount of the Cu3SnS4 orthorhombic phase is rather low. No other phases have been identified with this technique.
Cu2SnS3 Cu3SnS4
20
Mo Cu3SnS4
30
40
2 theta (º)
Mo
RI PT
Intensity (arb. units)
Cu2SnS3
50
SC
Intensity (arb. units)
ACCEPTED MANUSCRIPT
Mo
Mo
20
30
40
50
60
M AN U
10
70
80
2 theta (º)
Figure 2. X-Ray diffraction pattern of the CTS absorber layer. The inset shows a magnification of low intensity range between 20 and 50º. The phases assigned to each peak are labelled in the plots.
AC C
EP
TE D
In order to confirm the crystal structure of the CTS absorber layers Raman spectroscopy was performed on the sample and it is shown in figure 3. The peaks observed in the spectrum have been qualitatively fitted with Lorentzian functions to determine the position of the observed Raman modes. Two high intensity peaks at 303 and 359 cm-1 suggest that Cu2SnS3 is mainly present in the cubic structure, in accordance with the studies reported by Fernandes et al. [3, 23]. Unfortunately, no calculations of the phonon spectra of the different crystal structures of Cu2SnS3 have been reported up to now and the assignments for the cubic Cu2SnS3 are made based on reported experimental data. In addition, lower intensity peaks at 319 cm-1 and 337-346 cm-1can be observed. In the literature, peaks at these positions have been correlated with the presence of orthorhombic Cu3SnS4 [11, 23], which agrees with its identification by XRD presented in figure 2. The position of the modes for the orthorhombic Cu3SnS4 to perform the fitting for this phase was adjusted according to the phonon spectrum calculated by Dzhagan et al [11]. Raman measurements with λexc =633nm would potentially be very sensitive to the modes of Cu3SnS4, due to near resonant conditions [24], since a bandgap of 1.6 eV has been reported for this phase [3]. This would explain the fact that we clearly observe modes related to this phase even when its concentration in the film is low, as suggested by the low intensity peaks in the XRD patterns corresponding to Cu3SnS4. Also a Raman mode at 265 cm-1 is observed, which we are not able to attribute to a specific phase. Although calculations have predicted Raman active modes close to this frequency [11], the experimental data in literature for orthorhombic Cu3SnS4 does not agree with this mode being the most intense for this phase [11].
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
A bandgap value of 1.02 eV was calculated form the Tauc’s plot generated with the optical transmission spectrum of a lifted-off CTS absorber as shown in figure 3b. Strong subgap absorption can be observed up to 0.85 eV, in good agreement with theoretical calculations of the density of states for the disorder cubic CTS that have shown strong band tails for this material [22].
EP
Photovoltaic devices.
AC C
3.2.
TE D
Figure 3. (a) Raman spectrum of the CTS absorber (dark grey). Lorentzian fitting of the Cu2SnS3 cubic modes (red), Cu3SnS4 (blue) and an unidentified mode (green). The black line shows the sum of the fits. (b) Tauc’s plot generated from the optical transmission and reflection spectrum of a lifted-off CTS absorber.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 4. (a) Dark and illuminated JV curves of the best CTS cell. (b) External (black line) and internal (red line) quantum efficiency of the best solar cell. (c) CV profiles measured at 10 kHz, 100 kHz and 1000 kHz.
The current-voltage (J-V) characteristic of the best performing device is shown in figure 4a. The efficiency of the device was 1.8%, with a short circuit current density (JSC) of 28 mA/cm2, an open circuit voltage (VOC) of 147 mV and a fill factor (FF) of 44% for a total cell area of 0.5 cm2. The reason for the low value of VOC is not well understood but several possibilities can be considered. The cubic structure of Cu2SnS3 has been reported to be highly disordered implying band tailing, compositional inhomogeneities at the nanometre scale and
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
potential fluctuations [22]. Also, a cliff-like band offset has been reported for the CdS-CTS junction [25]. Assuming that this is the case for this device, a reduction of the interface bandgap would reduce VOC and increase the recombination at the interface of the device [26]. In addition, the presence of other phases found in the absorber layer, such as Cu3SnS4, could contribute detrimentally, increasing the metallic character of the absorber layers and decreasing the performance of the solar cells due to an increase of shunting paths [6]. The External Quantum Efficiency (EQE) and the Internal Quantum Efficiency (IQE) are shown in figure 4b. The IQE (λ) was calculated from the ratio of the EQE (λ) and the spectral refelectance of the device as IQE (λ) = EQE (λ) / [1−R (λ)] . A step between 650 and 800 nm observed in the EQE, which is not evident in the IQE, suggesting that it is due to interferences of the device layers. At wavelengths above the bandgap of CdS (510-520nm) both, the EQE and IQE start to decay significantly, suggesting that a measurable contribution to the JSC occurs from hole generation in the CdS. This contribution ceases for wavelengths with energies below the bandgap of CdS in agreement with a small collection function in the bulk of the absorber [27]. Nakashima et al. reported the appearance of the same decay when the EQE was measured without white light bias [1], suggesting that the light intensity dependent properties of CTS solar cells should be further studied to contribute to the progress of this material. The bandgap was calculated from the inflection point of the EQE and IQE giving a value of approximately 1.06 eV. Strong band tailing is also observed in the QE, in agreement with the Tauc’s plot presented in figure 3b and in accordance with previously reported theoretical calculations for the disorder cubic structure [22]. The presence of this strong band tailing could be a potential issue for the development of CTS solar cells when the absorber layer adopts the cubic structure. This band tail is not observed in the EQE of devices made out of the monoclinic Cu2SnS3 where two transitions are observed at around 0.93 and 0.99 eV [1], suggesting that this crystal structure could be a better candidate for CTS solar cells. JSC has been calculated by integrating the EQE with the AM 1.5 solar spectrum yielding a value of JSC (EQE) =16.4 mA/cm2 which is significantly lower than the value measured in the JV curve (28.0 mA/cm2). This behaviour has been previously observed in other CTS solar cells found in literature when the EQE has been measured without white light bias [1], suggesting again that the optoelectronic properties of these CTS solar cells are affected by the light intensity. CV measurements were performed on the CTS device and are shown in figure 4c. Charge carrier concentrations of 3 - 5 x 1016 cm-3and a depletion region width of around 120 nm were estimated for this solar cell by assuming a static dielectric constant ε0=10. The charge carrier concentration values measured for this solar cell are significantly lower than what was previously reported for cubic CTS, where the lowest values were around 7 x 1017 cm-3 and 1021 cm-3 when Cu3SnS4 was present. These values are considered to be too high for a p-type absorber layer in a photovoltaic device [6].
4. Conclusion
ACCEPTED MANUSCRIPT
SC
RI PT
In this work, we have reported the synthetisation of CTS absorber layers by single step coevaporation with a nominal substrate temperature of 400ºC. A combination of Raman, SEM and XRD analysis leads us to conclude that the Cu2SnS3 is highly textured and adopts a disordered cubic structure. The presence of Cu3SnS4 was also detected in the samples. A 1.8% efficient CTS/CdS/i-ZnO/Al:ZnO solar cell was obtained with JSC of 28 mA cm-2, VOC of 147 mV and a FF of 42.9 %. A bandgap value of 1.06 eV was extracted from the inflection point of the EQE. The Jsc calculated by integrating the EQE with the AM 1.5 solar spectrum shows disagreement with the measured JSC indicating that the devices suffer from changes in its electronic properties depending on the intensity of the illumination. A charge carrier concentration of 4 x 1016 cm-3 is extracted from C-V measurements with a depletion region of approximately 120 nm. The limitations of the CTS solar cells presented in this work provide insights for the future development of higher efficiency thin film solar cells based on the Cu2SnS3 ternary compound.
M AN U
5. Acknowledgements
References
TE D
This work has been undertaken within a Marie Curie, Initial Training Network (ITN) project, “KESTCELLS” co-financed by the Seventh Framework Programme of the European Commission (FP7/2007-2013) under grant agreement nº316488. The authors gratefully acknowledge support by the Helmholtz Association Initiative and Network Fund (HNSEIProject SO-075). A. Irkhina, C. Ferber and L. Steinkopf from HZB are also acknowledged for helping with processing and characterisation of the solar cells
AC C
EP
[1] M. Nakashima, J. Fujimoto, T. Yamaguchi, M. Izaki, Cu2SnS3thin-film solar cells fabricated by sulfurization from NaF/Cu/Sn stacked precursor, Applied Physics Express, 8 (2015) 042303. [2] S. Fiechter, M. Martinez, G. Schmidt, W. Henrion, Y. Tomm, Phase relations and optical properties of semiconducting ternary sulfides in the system Cu–Sn–S, Journal of Physics and Chemistry of Solids, 64 (2003) 1859-1862. [3] P.A. Fernandes, P.M.P. Salomé, A.F.d. Cunha, A study of ternary Cu2SnS3and Cu3SnS4thin films prepared by sulfurizing stacked metal precursors, Journal of Physics D: Applied Physics, 43 (2010) 215403. [4] D.M. Berg, R. Djemour, L. Gütay, S. Siebentritt, P.J. Dale, X. Fontane, V. Izquierdo-Roca, A. PérezRodriguez, Raman analysis of monoclinic Cu2SnS3 thin films, Applied Physics Letters, 100 (2012) 192103. [5] D.M. Berg, R. Djemour, L. Gütay, G. Zoppi, S. Siebentritt, P.J. Dale, Thin film solar cells based on the ternary compound Cu2SnS3, Thin Solid Films, 520 (2012) 6291-6294. [6] L.L. Baranowski, P. Zawadzki, S. Christensen, D. Nordlund, S. Lany, A.C. Tamboli, L. Gedvilas, D.S. Ginley, W. Tumas, E.S. Toberer, A. Zakutayev, Control of Doping in Cu2SnS3through Defects and Alloying, Chemistry of Materials, 26 (2014) 4951-4959. [7] T. Nomura, T. Maeda, T. Wada, Fabrication of Cu2SnS3solar cells by screen-printing and highpressure sintering process, Japanese Journal of Applied Physics, 53 (2014) 05FW01.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
[8] A. Kanai, H. Araki, A. Takeuchi, H. Katagiri, Annealing temperature dependence of photovoltaic properties of solar cells containing Cu2SnS3 thin films produced by co-evaporation, physica status solidi (b), 252 (2015) 1239-1243. [9] M. Bouaziz, J. Ouerfelli, M. Amlouk, S. Belgacem, Structural and optical properties of Cu3SnS4sprayed thin films, physica status solidi (a), 204 (2007) 3354-3360. [10] V.A. Kovalenker, Kuramite, Cu3SnS4, a new mineral of the stannite group, International Geology Review, 23 (2010) 365-370. [11] V.M. Dzhagan, A.P. Litvinchuk, M. Kruszynska, J. Kolny-Olesiak, M.Y. Valakh, D.R.T. Zahn, Raman Scattering Study of Cu3SnS4Colloidal Nanocrystals, The Journal of Physical Chemistry C, 118 (2014) 27554-27558. [12] X.-a. Chen, H. Wada, A. Sato, M. Mieno, Synthesis, Electrical Conductivity, and Crystal Structure of Cu4Sn7S16and Structure Refinement of Cu2SnS3, Journal of Solid State Chemistry, 139 (1998) 144-151. [13] Y. Ren, J.J. Scragg, T. Ericson, T. Kubart, C. Platzer-Björkman, Reactively sputtered films in the CuxS–ZnS–SnSy system: From metastability to equilibrium, Thin Solid Films, 582 (2015) 208-214. [14] J. Koike, K. Chino, N. Aihara, H. Araki, R. Nakamura, K. Jimbo, H. Katagiri, Cu$_{2}$SnS$_{3}$ Thin-Film Solar Cells from Electroplated Precursors, Japanese Journal of Applied Physics, 51 (2012) 10NC34. [15] S.A. Vanalakar, G.L. Agawane, A.S. Kamble, C.W. Hong, P.S. Patil, J.H. Kim, Fabrication of Cu2SnS3 thin film solar cells using pulsed laser deposition technique, Solar Energy Materials and Solar Cells, 138 (2015) 1-8. [16] S.A. Vanalakar, G.L. Agawane, S.W. Shin, H.S. Yang, P.S. Patil, J.Y. Kim, J.H. Kim, Non-vacuum mechanochemical route to the synthesis of Cu2SnS3 nano-ink for solar cell applications, Acta Materialia, 85 (2015) 314-321. [17] Q. Chen, X. Dou, Y. Ni, S. Cheng, S. Zhuang, Study and enhance the photovoltaic properties of narrow-bandgap Cu2SnS3 solar cell by p-n junction interface modification, Journal of colloid and interface science, 376 (2012) 327-330. [18] N. Aihara, H. Araki, A. Takeuchi, K. Jimbo, H. Katagiri, Fabrication of Cu2SnS3thin films by sulfurization of evaporated Cu-Sn precursors for solar cells, physica status solidi (c), 10 (2013) 10861092. [19] Y. Dong, J. He, L. Sun, Y. Chen, P. Yang, J. Chu, Effect of sulfurization temperature on properties of Cu2SnS3 thin films and solar cells prepared by sulfurization of stacked metallic precursors, Materials Science in Semiconductor Processing, 38 (2015) 171-176. [20] A. Zakutayev, L.L. Baranowski, A.W. Welch, C.A. Wolden, E.S. Toberer, Comparison of Cu2SnS3 and CuSbS2 as potential solar cell absorbers, in: Photovoltaic Specialist Conference (PVSC), 2014 IEEE 40th, 2014, pp. 2436-2438. [21] N. Aihara, A. Kanai, K. Kimura, M. Yamada, K. Toyonaga, H. Araki, A. Takeuchi, H. Katagiri, Sulfurization temperature dependences of photovoltaic properties in Cu2SnS3-based thin-film solar cells, Japanese Journal of Applied Physics, 53 (2014) 05FW13. [22] L.L. Baranowski, K. McLaughlin, P. Zawadzki, S. Lany, A. Norman, H. Hempel, R. Eichberger, T. Unold, E.S. Toberer, A. Zakutayev, Effects of disorder on carrier transport in Cu2SnS3, Physical Review Applied, 4 (2015) 044017. [23] P.A. Fernandes, P.M.P. Salomé, A.F. da Cunha, CuxSnSx+1(x = 2, 3) thin films grown by sulfurization of metallic precursors deposited by dc magnetron sputtering, physica status solidi (c), (2010). [24] A. Fairbrother, V. Izquierdo-Roca, X. Fontané, M. Ibáñez, A. Cabot, E. Saucedo, A. PérezRodríguez, ZnS grain size effects on near-resonant Raman scattering: optical non-destructive grain size estimation, CrystEngComm, 16 (2014) 4120-4125. [25] S. Sato, H. Sumi, G. Shi, M. Sugiyama, Investigation of the sulfurization process of Cu2SnS3 thin films and estimation of band offsets of Cu2SnS3-related solar cell structure, physica status solidi (c), 12 (2015) 757-760.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
[26] S. Siebentritt, Why are kesterite solar cells not 20% efficient?, Thin Solid Films, 535 (2013) 1-4. [27] R.S.a.H.-W. Schock, Chalcogenide Photovoltaics: Physics, Technologies, and Thin Film Devices, 2011.
ACCEPTED MANUSCRIPT
Highlights
AC C
EP
TE D
M AN U
SC
RI PT
1. Cu2SnS3 thin films deposited by co-evaporation 2. We fabricated a Cu2SnS3 thin film solar cell 3. The Cu2SnS3 absorber layer crystallises in a disorder cubic structure