Efficient room temperature aqueous Sb₂S₃ synthesis for inorganic–organic sensitized solar cells with 5.1% efficiencies

Abstract

Sb₂S₃ sensitized solar cells are a promising alternative to devices employing organic dyes. The manufacture of Sb₂S₃ absorber layers is however slow and cumbersome. Here, we report the modified aqueous chemical bath synthesis of Sb₂S₃ absorber layers for sensitized solar cells. Our method is based on the hydrolysis of SbCl₃ to complex antimony ions decelerating the reaction at ambient conditions, in contrast to the usual low temperature deposition protocol. This simplified deposition route allows the manufacture of sensitized mesoporous-TiO₂ solar cells with power conversion efficiencies up to η = 5.1%. Photothermal deflection spectroscopy shows that the sub-bandgap trap-state density is lower in Sb₂S₃ films deposited with this method, compared to standard deposition protocols.

---

Antimony sulfide (Sb₂S₃) is a promising material for several optoelectronic applications. Due to its high absorption coefficient (α ≈ 1.8 × 10⁵ cm⁻¹ at λ = 450 nm) and a suitable direct band-gap of E_g ≈ 1.7 eV, crystalline Sb₂S₃ (stibnite) is interesting as light absorber for solid-state sensitized solar cells (Fig. 1).^1,2 In particular, Sb₂S₃-based solar cells excel in their stability of operation when compared to other organic–inorganic hybrid devices. Recently, power-conversion efficiencies of η = 6.2% (ref. 3) and η = 7.5% (ref. 4) were achieved using Sb₂S₃ as the absorber material obtained from chemical bath deposition. Further, the material has been used to improve the stability of methyl-ammonium lead iodide perovskite solar cells.⁵

Fig. 1 (a) Schematic of the solar cell cross section. (b) Simplified band-diagram of the Sb₂S₃ sensitised photovoltaic cells. The band edge values were taken from ref. 10.

Antimony sulfide synthesis typically involves deposition in aqueous and non-aqueous chemical baths at low temperatures (low-T deposition).^6–9 The standard method is the aqueous chemical bath deposition (CBD) using antimony chloride and sodium thiosulfate. This technique is however problematic since it requires a precise temperature control of the solution when cooling below 10 °C and maintaining the sample at low temperatures. For large-scale applications such a cooling protocol is cumbersome, costly and energy-intensive.

Here, we present an aqueous room temperature (RT) deposition route of Sb₂S₃ using the same precursor materials as the standard CBD method. We have fabricated Sb₂S₃-sensitized solar cells using this RT deposition method and demonstrate excellent device performance with efficiencies of up to η = 5.1%.

The low-temperature synthesis of Sb₂S₃ for photovoltaic applications involves the chemical reaction equations⁷

2SbCl₃ + 3Na₂S₂O₃ → Sb₂(S₂O₃)₃ + 6NaCl

Sb₂(S₂O₃)₃ + 6H₂O → Sb₂S₃ + 3HSO₄⁻ + 3H₃O⁺.

These reactions have to be slowed down by cooling below 10 °C to avoid immediate precipitation⁷ and to enable strong adhesion of Sb₂S₃ to the substrate. Thus, the standard CBD method, termed low-T deposition, requires cooling of the reaction solution, whereas our modified method, RT deposition, can be performed at room temperature. By changing the order of reactant addition at RT, Sb₂S₃ formation is slowed down and well-adhering films are obtained.

For the RT deposition, a 1.4 M SbCl₃ solution in acetone was prepared. Note that the optimal concentration of SbCl₃ for the RT method is slightly higher compared to low-T deposition (Fig. S7, ESI†). SbCl₃ can be used without dissolution in acetone, changing the reaction behaviour very little (Fig. S8, ESI†). The addition of acetone facilitates however the handling of the highly hygroscopic SbCl₃. Deionised water is added under vigorous stirring to reduce the total concentration of SbCl₃ to 46 mM. The addition of water hydrolyses SbCl₃, which leads to a solid white precipitate. The product of the hydrolysis reaction of SbCl₃ is not very well defined. It depends on many parameters such as the dilution of the reaction medium, the pH value of the solution and solvent composition.^11,12 The aqueous solution containing the hydrolysed SbCl₃ has a pH of 1.4.

The precipitate was filtered and dried and X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR) and energy-dispersive X-ray (EDX) spectroscopy were performed on the white powder (Fig. S10–S12, ESI†). The XRD pattern shows crystalline phases of Sb₄O₅Cl₂, Sb₈(OH)₆O₈Cl₂(H₂O), Sb₈O₁₁Cl₂(H₂O)₆ and Sb₃O₆(OH). This is in contrast to the reports by Li et al. and Yu et al. describing similar reactions.^13,14 They also reporting the formation of a white precipitate, which they identify as antimony oxychloride SbOCl. The XRD pattern of Fig. S10 (ESI†) however shows no evidence of crystalline SbOCl formation. According to Chen et al. hydrolysis at pH 1–2 leads to the formation of Sb₄O₅Cl₂ for mixed solvents such as water and ethanol or water and ethylene glycol.¹² We also do not observe an immediate colour change of the precipitate to orange upon the addition of the sulphur source, as reported by Li et al.¹³ and Yu et al.¹⁴

Subsequently, a 1 M aqueous Na₂S₂O₃ solution was added at a final concentration of 0.25 M in the chemical bath. This causes the solution to turn clear as most of the precipitate dissolves, suggesting the formation of a water-soluble complex. After 5–10 min at 20 °C, the solution starts to turn orange, indicating the formation of amorphous Sb₂S₃. This is accompanied by a pH change of the solution from pH = 3.3 upon sodium thiosulfate (pH = 7.3) addition to pH = 4.3 when antimony sulfide deposition is complete after two hours.

Fig. S2 (ESI†) shows the UV-vis spectra of RT-deposited Sb₂S₃ films on mesoporous-TiO₂, annealed at 300 °C for 5 min.

The annealed antimony sulfide was characterised by powder X-ray diffraction (XRD). Fig. 3a compares the XRD pattern of the low-T and RT deposition methods. Both pattern are very similar and match the stibnite reference pattern.¹⁵ Sb₂S₃ films deposited by the RT method onto mesoporous TiO₂ substrates also show the characteristic XRD peaks of crystalline Sb₂S₃ (Fig. S3, ESI†). A Rietveld analysis of the two patterns using the powder diffraction software ReX¹⁶ yields an average Sb₂S₃ crystallite size of 40 nm and 35 nm for RT and low-T deposition, respectively. The structure of the solar cell studied in this work and a schematic band diagram are shown in Fig. 1.

Fig. 2 shows (a) the current–voltage-characteristic and (b) the external quantum efficiency EQE of the best performing solar cell employing RT deposited Sb₂S₃. Its photovoltaic parameters are summarised in Fig. 2. The shunt and series resistances were obtained by a least square fit of the diode function. The RT method enables the fabrication of solar cells with high reproducibility. The deviations from batch to batch and from device to device were small (Fig. S13, ESI†). The hole transport layer consists of a PCPDTBT–PCBM blend. Charge carriers which are generated by photon absorption in the hole transport layer can also be harvested due to the formation of PCBM electron conducting channels. The PCBM–PCPDTBT blend thus contributes to the current and causes the near-infrared shoulder in the EQE spectrum.¹⁷ Earlier cells using P3HT as hole conductor showed lower efficiencies compared to the PCPDTBT:PCBM blend. We therefore concentrated on the donor–acceptor PCPDTBT–PCBM blend. A comparison of a RT solar cell and a solar cell with the same device architecture using the low-temperature deposition method is shown in the ESI† (Fig. S4). The reference cells were prepared in the same laboratory using the identical procedures as the better performing earlier published devices.⁴ The lower performance may reflect the variation in this system that possibly arises from minor variations in device fabrication. It is however important to point out that the reference devices and devices made with our new methodology were created in parallel and therefore are much less likely to be subject to this type of variation.

Fig. 2 (a) Current–voltage curve of the best performing solar cell device employing Sb₂S₃ synthesised using the RT method. The inset table shows the photovoltaic parameters of the device. (b) External quantum efficiency of the device.

X-ray photoelectron spectroscopy (XPS) measurements were carried out on Sb₂S₃ films formed by both deposition techniques, shown in Fig. 3b.

Fig. 3 (a) XRD pattern of crystalline Sb₂S₃, synthesised via the low temperature deposition method (upper trace) and the RT technique (center trace). The lower trace shows the difference of the two patterns. The stibnite reference pattern was taken from ref. 15. (b) XPS Sb3d_3/2 peak of films from both deposition methods. The solid lines show a least squares fit of a superposition of two Gaussian curves. (c) PDS measurements showing the sub band-gap energy levels for both deposition methods. The inset shows the corresponding Urbach energies.

To compare the oxide content of the samples, the antimony Sb3d_3/2 peak was examined, because the oxygen O1s peak directly overlaps with the antimony Sb3d_5/2 peak. The Sb3d_3/2 peak can be modelled using a superposition of two Gaussians, one at ≈538.5 eV representing Sb₂S₃ and one at ≈539.5 eV for SbO_x, most likely Sb₂O₃.¹⁸ The RT sample has a marginally higher oxide content compared to the low-T material. This probably causes the lower conductivity seen in these films (Fig. S6, ESI†).

One of the biggest challenges of using antimony sulfide as absorber in sensitized solar cells is the high density of electronic traps in this material, i.e. the number of energy states which lie in the band-gap of Sb₂S₃.^4,19 These trap-states lead to a significant loss in potential and to charge carrier recombination in the solar cell. To explore this, we employed photo-thermal deflection spectroscopy (PDS) to determine the trap-state density and the energetic disorder of Sb₂S₃. PDS is a highly sensitive absorption measurement technique, which can detect absorbance values down to 10⁻⁵ AU. Thus, PDS is able to accurately measure weak absorption in the bandgap. Fig. 3c shows the PDS spectra of Sb₂S₃ samples on mesoporous TiO₂ for both deposition methods. The corresponding Urbach energies are given in the inset.

The absorption in the RT-deposited sample was significantly lower at energies below the band-gap of Sb₂S₃ compared to the low-T sample, by nearly one order of magnitude at energies below 1.5 eV. This indicates a clear reduction in the density of deep-trap states for the RT deposited Sb₂S₃. The difference in the open-circuit voltage for low-T and RT deposited Sb₂S₃ in optimized solar cells are however similar (Fig. S4, ESI†). As the band-gap of antimony oxide is higher than that of Sb₂S₃,²⁰ the PDS spectrum cannot show a potential increase of deep traps caused by the higher content of antimony oxide in the RT-deposited sample.

We have demonstrated an aqueous deposition technique of antimony sulfide for sensitized solar cells, which can be carried out at room temperature. The chemical bath deposition method is based on the same precursor materials but uses the hydrolysis of SbCl₃ to complex antimony ions. The resulting Sb₂S₃ films were investigated using UV-vis spectroscopy, XRD, PDS and XPS. PDS shows a reduction in sub-band gap trap states in RT-deposited Sb₂S₃. Manufactured devices achieved a maximum power conversion efficiency η = 5.1% for Sb₂S₃ sensitized solar cells using the RT deposition method. A more detailed optimization of the deposition step, interfacial surface treatments^21,22 or doping of the Sb₂S₃²³ could lead to a further improvement in solar cell performance. This work is therefore an important step in the development of low-cost, stable and highly efficient solar cells.

We acknowledge Adam Brown's help with the XPS measurements. K.C.G. would like to thank the Cambridge Trust, the Mott Fund for Physics of the Environment and Corpus Christi College Cambridge for funding. A.S. acknowledges funding from the Engineering and Physical Sciences Research Council (EPSRC).

References

1. M. Y. Versavel and J. A. Haber, Thin Solid Films, 2007, 515, 7171–7176.
2. S.-J. Moon, Y. Itzhaik, J.-H. Yum, S. M. Zakeeruddin, G. Hodes and M. Grätzel, J. Phys. Chem. Lett., 2010, 1, 1524–1527.
3. S. H. Im, C.-S. Lim, J. A. Chang, Y. H. Lee, N. Maiti, H.-J. Kim, M. K. Nazeeruddin, M. Grätzel and S. I. Seok, Nano Lett., 2011, 11, 4789–4793.
4. Y. C. Choi, D. U. Lee, J. H. Noh, E. K. Kim and S. I. Seok, Adv. Funct. Mater., 2014, 24, 3587–3592.
5. S. Ito, S. Tanaka, K. Manabe and H. Nishino, J. Phys. Chem. C, 2014, 118, 16995–17000.
6. I. Grozdanov, Semicond. Sci. Technol., 1994, 9, 1234.
7. M. T. S. Nair, Y. Pena, J. Campos, V. M. Garcia and P. K. Nair, J. Electrochem. Soc., 1998, 145, 2113–2120.
8. S. Messina, M. Nair and P. Nair, Thin Solid Films, 2007, 515, 5777–5782.
9. N. Maiti, S. H. Im, C.-S. Lim and S. I. Seok, Dalton Trans., 2012, 41, 11569–11572.
10. J. A. Chang, J. H. Rhee, S. H. Im, Y. H. Lee, H.-j. Kim, S. I. Seok, M. K. Nazeeruddin and M. Gratzel, Nano Lett., 2010, 10, 2609–2612.
11. A. Koliadima, A. Henglein and E. Matijević, Colloid Polym. Sci., 1997, 275, 972–978.
12. X. Y. Chen, H. S. Huh and S. W. Lee, J. Solid State Chem., 2008, 181, 2127–2132.
13. C. Li, X. Yang, Y. Liu, Z. Zhao and Y. Qian, J. Cryst. Growth, 2003, 255, 342–347.
14. Y. Yu, R. H. Wang, Q. Chen and L.-M. Peng, J. Phys. Chem. B, 2005, 109, 23312–23315.
15. A. Kyono and M. Kimata, Am. Mineral., 2004, 89, 932–940.
16. M. Bortolotti, L. Lutterotti and I. Lonardelli, J. Appl. Crystallogr., 2009, 42, 538–539.
17. J. A. Chang, S. H. Im, Y. H. Lee, H.-J. Kim, C.-S. Lim, J. H. Heo and S. I. Seok, Nano Lett., 2012, 12, 1863–1867.
18. V. P. Zakaznova-Herzog, S. L. Harmer, H. W. Nesbitt, G. M. Bancroft, R. Flemming and A. R. Pratt, Surf. Sci., 2006, 600, 348–356.
19. A. Darga, D. Mencaraglia, C. Longeaud, T. J. Savenije, B. O'Regan, S. Bourdais, T. Muto, B. Delatouche and G. Dennler, J. Phys. Chem. C, 2013, 117, 20525–20530.
20. Z. Deng, F. Tang, D. Chen, X. Meng, L. Cao and B. Zou, J. Phys. Chem. B, 2006, 110, 18225–18230.
21. K. Tsujimoto, D.-C. Nguyen, S. Ito, H. Nishino, H. Matsuyoshi, A. Konno, G. R. A. Kumara and K. Tennakone, J. Phys. Chem. C, 2012, 116, 13465–13471.
22. T. Fukumoto, T. Moehl, Y. Niwa, M. K. Nazeeruddin, M. Grätzel and L. Etgar, Adv. Energy Mater., 2013, 3, 29–33.
23. S. Ito, K. Tsujimoto, D.-C. Nguyen, K. Manabe and H. Nishino, Int. J. Hydrogen Energy, 2013, 38, 16749–16754.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5cc01966d

---