A 4.92% efficiency Cu₂ZnSnS₄ solar cell from nanoparticle ink and molecular solution

Abstract

Quaternary Cu₂ZnSnS₄ (CZTS) thin films were prepared by a low-cost, simple and environmentally-friendly ink method. By depositing molecular solution on the nanoparticle thin film, the quality of as-prepared CZTS thin films was greatly improved (e.g. reduction of fine grain layer formation and improved crystallinity). The effect of the number of spin-coated layers from molecular solution on solar cell performance was investigated. The results indicated that the CZTS thin film had the best performance when 5 layers were spin-coated from molecular solution on the nanoparticle thin film. The crystallinity of the as-prepared CZTS thin film and the interface at Mo/CZTS was found to be obviously enhanced by addition of a molecular solution layer. Finally, a CZTS thin film solar cell with an efficiency of 4.92% has been fabricated.

---

1. Introduction

The thin film solar cell materials, such as Cu(In,Ga)(S,Se)₂ (CIGS), Cu₂ZnSnS₄ (CZTS), CdTe, etc., are attracting widespread attention. Particularly, CZTS which contains earth-abundant, low-cost, and nontoxic elements and retains a similar structure to CIGS, has garnered considerable interest. Moreover, CZTS has an optical band gap (∼1.5 eV) and a large absorption coefficient (>10⁴ cm⁻¹).¹ The theoretical limit for Cu₂ZnSnS₄ thin film solar cells is nearly 32.2% by the Shockley–Queisser photon balance calculations.²

Various methods including vacuum and non-vacuum based technologies have been used to prepare CZTS thin films for solar cells. CZTS solar cell with an active area efficiency of 6.77% was fabricated by sputtering method.³ Byungha Shin et al. reported 8.4% efficiency CZTS solar cell fabricated by co-evaporation deposition method.⁴ However, the deposition technique generally requires high vacuum system increasing preparation cost. In contrast, non-vacuum based approaches can provide low-cost routes to CZTS solar cells. These non-vacuum based approaches include spray pyrolysis, electrodeposition, and ink approach, etc.^5–8 Among these non-vacuum methods, ink approach containing of nanoparticles-based ink and molecular solution ink, has been extensively reviewed due to suitability for large-scale production.⁹ Su et al. used molecular solution ink method to prepare CZTS thin film, and the efficiency of CZTS solar cell reached to 5.1%.¹⁰ Recently, Su et al. reported that Cd replaced Zn partially to form Cu₂Zn_1−xCd_xSnS₄ thin film and the efficiency of Cu₂Zn_1−xCd_xSnS₄ solar cell reached to 9.24%.¹¹ Meanwhile, nanocrystals-based approach has obtained more and more attention. This method can control phase formation, since phase formation occurs prior to film deposition. Gu et al. fabricated CZTS thin film solar cell with 2.29% by nanocrystals-ink method.¹² Caleb K. Miskin et al. selenized nanoparticle inks to prepare Cu₂ZnSn(S,Se)₄ thin film and obtained 9% efficient solar cell.¹³

At present, the CZTS nanoparticle ink is prepared by hydrophobic CZTS nanoparticles, which are synthesized using organic solvent such as oleylamine (OM), trioctylphosphine oxide (TOPO), oleic acid (OA), and octadecene (ODE).¹⁴ These solvents are high-cost and not environmental-friendly. The hydrophobic nanoparticles are coated with a chemisorbed layer of organic ligands to prevent aggregation, which provide good dispersibility in solvents. But, the organic ligands will cause carbon residue, which can weaken the devices performance.¹⁵ In this paper, we used hydrophilic CZTS nanoparticles to prepare ink, which were synthesized using low-cost and environmental-friendly ethylene glycol as solvent.

In order to improve quality of as-prepared thin films, we combined molecular solution with nanoparticle ink to fabricate CZTS thin films. Compared with nanoparticles ink-derived CZTS solar cell, the CZTS solar cell prepared by combination of molecular solution with nanoparticle ink had higher efficiency and its efficiency reached 4.92%.

2. Experimental

2.1 Starting materials

Copper(II) acetate monohydrate (Cu(CH₃COO)₂·H₂O), tin(II) chloride dihydrate (SnCl₂·2H₂O), zinc(II) acetate dihydrate (Zn(CH₃COO)₂·2H₂O), thiourea (CN₂H₄S, Tu), n-propanol (C₃H₇OH), 2-methoxyethanol (C₃H₈O₂), monoethanolamine (HO(CH₂)₂NH₂, MEA) and triethanolamine ((HOCH₂CH₂)₃N, TEA) were used as starting materials of analytical grade.

2.2 Preparation of CZTS nanoparticle ink and molecular solution

Preparation of CZTS nanoparticle ink: CZTS nanoparticles were synthesized by microwave irradiation method using thioacetamide as sulfur source. Detail preparation of CZTS nanoparticles could be found in our previous literature report.¹⁶ 600 mg CZTS nanoparticles were added to 6 mL n-propanol. Then, 0.5 mL TEA was added to this ink and the ink was ultra-sonicated for 1 h.

Preparation of molecular solution: 7.74 mmol Cu(CH₃COO)₂·H₂O, 5 mmol Zn(CH₃COO)₂·2H₂O, 4 mmol SnCl₂·2H₂O and 32 mmol Tu were dissolved in 15 mL 2-methoxyethanol. This preparation method of molecular solution was similar with the previous literature report.¹¹

2.3 Preparation of CZTS thin film

Fig. 1 showed diagram of the preparation of CZTS thin film. The CZTS nanoparticle ink was spin-coated on molybdenum coated glass substrate at 2000 rpm for 20 s followed by heating at 300 °C for 2 min on a hotplate in air. This process was repeated for 20 times and we marked this sample as S0. The CZTS nanoparticle ink was spin-coated for 12 times and then molecular solution was spin-coated for 3 times. We marked the sample as S3. The CZTS nanoparticle ink was spin-coated for 10 times and then molecular solution was spin-coated for 5 times. We marked the sample as S5. In order to compare their properties, we also used all molecular solution to spin-coat for 15 times to prepare thin film, and we marked this sample as PS15. After that, all samples were sulfurized at 600 °C for 40 min. The sulfurization pressure was 300 mbar.

Fig. 1 Diagram of the preparation process of CZTS thin film by nanoparticle ink and molecular solution method.

2.4 Preparation of CZTS solar cell

CdS buffer layers were deposited on the as-prepared CZTS thin films at 80 °C for 10 min by chemical bath deposition (CBD) method. The CBD solution consisted of 140 mL deionized H₂O, 20 mL NH₃·H₂O (25 wt%, AR), 20 mL CdSO₄ (15 mmol L⁻¹, AR), 20 mL CS(NH₂)₂ (75 mmol L⁻¹, AR). And then, 50 nm i-ZnO was deposited by RF sputtering. The growth parameters such as sputtering power, work pressure and deposition time were fixed at 150 W, 10⁻² mbar and 15 min, respectively. Then, 450 nm ZnO : Al (AZO) layer was deposited by DC sputtering. The sputtering power was 200 W, and work pressure was 4 × 10⁻³ mbar. The sheet resistivity of as-prepared AZO thin film was 20 Ω □⁻¹. Finally, Ag grid was printed as the top electrode, and dried at 60 °C.

2.5 Characterization

The morphologies of the as-synthesized samples were characterized using scanning electron microscope (SEM, JEOL JSM 7600F). The phase and crystallographic information were obtained from the X-ray diffraction patterns (XRD, Bruker D8 Advance equipped with Cu-K_α radiation source) and Raman spectroscopy (Thermo Fisher DXR). Light I–V curves were plotted under 1.5 AM illumination (VS-0852) at room temperature with cell area of 0.16 cm². External quantum efficiency was obtained with light wavelength range from 300 nm to 1000 nm (PVE300 Photovoltaic Devices Characterization System).

3. Results and discussions

The detail characterization of CZTS nanoparticles could be found in our previous report.¹⁶ Fig. 2 showed XRD patterns of the as-prepared CZTS thin films. It revealed that all films had good crystallinity. The diffractions peaks were attributed to (002), (101), (110), (112), (103), (200), (220), (312), (008) and (332) planes of the kesterite structure of CZTS (JCPDS no. 26-0575).¹⁷ The XRD patterns of all samples showed a preferred orientation in the (112) direction. But we could find that (112) peak of PS15 was weaker than the others, which indicated the crystallinity of CZTS from solution was weaker compared to the other thin films. As shown in the XRD patterns, no other phase was observed except for Mo.

Fig. 2 The XRD patterns of as-prepared CZTS thin films.

Because the XRD diffraction patterns of CZTS are similar with those of ZnS, CuS and Cu₂SnS₃ secondary phases, Raman spectroscopy was used to further confirm the phase structure information. As shown in Fig. 3, the Raman peaks located at 286, 335 and 372 cm⁻¹ could be assigned to the characteristic modes of kesterite structure Cu₂ZnSnS₄, which matched well with previous literature.¹⁸ The Raman main peak of CZTS thin film prepared by molecular solution was weaker and broader than others, which might be due to smaller crystalline size. The results were agreed well with those of XRD. Raman peaks from ZnS, Cu₂SnS₃, SnS₂ and Cu_2−xS (351 cm⁻¹, 305 cm⁻¹, 315 cm⁻¹ and 475 cm⁻¹) were not observed, further indicating that pure CZTS thin films were obtained.¹⁹

Fig. 3 Raman spectroscopy of the as-prepared CZTS thin films.

Shown in Fig. 4 were the surface morphologies of the as-prepared CZTS thin films. From low magnification SEM images, there were several cracks in the surface of S0 (Fig. 4a). S0 was prepared using nanoparticle ink without molecular solution. During the thermal treatment process, the evaporation of organic compounds leads to bring large shrinkage stress which produced cracks. The crack would lead to generate larger leakage current to decrease the efficiency of solar cells. From Fig. 4c and e, we could find that using the molecular solution to cover the surface can effectively remove cracks. However, spin-coated 3 layers from the molecular solution on the nanoparticles film were too thin to cover all area. Thus, S3 had some holes on the surface (Fig. 4c and d). When the molecular solution was spin-coated on the nanoparticles film for 5 times, the cracks and holes disappeared as shown in Fig. 4e and f. Moreover, the as-prepared thin film was compact, uniform and even. But some holes could be also observed on the surface of PS15 (Fig. 4g and h). EDS measurement showed that the chemical compositions of S0 and PS15 were Cu : Zn : Sn : S = 20.37 : 10.91 : 12.17 : 56.56 and Cu : Zn : Sn : S = 21.22 : 13.20 : 12.41 : 53.17. The composition of S0 was close to the predicted composition. The result indicated that Sn-loss was no observed in S0 prepared by nanoparticle ink method.

Fig. 4 Low and high magnification SEM images of the as-prepared CZTS thin films: (a and b) S0, (c and d) S3, (e and f) S5, (g and h) PS15.

Fig. 5 presented the cross-section SEM images of the CZTS thin film solar cells with structure of Mo/CZTS/CdS/i-ZnO/AZO. The thickness of as-prepared CZTS thin films was 0.81, 0.73, 0.70 and 0.9 μm, respectively. As shown in Fig. 5a, S0 exhibited a bi-layers structure consisting of a large-grain top layer (∼0.66 μm) and a fine-grain bottom layer (∼0.15 μm). This structure was similar with the literature.²⁰ But we could also find that the fine-grain layer had a good contact with Mo. Compared with S0, S3 and S5 exhibited that the thickness of the fine-grain layer reduced and even disappeared. The fine-grain layer disappeared due to reduction nanoparticles precursor layers. Fig. 5c presented the grain size was larger than 500 nm. According to previous literature reports, large grain size was necessary to prepare solar cells with high efficiency. The fine-grain layer and MoS₂ at the interface of CZTS/Mo were no observed. However, the grain size of pure molecular solution-based thin film was smaller than the other samples (Fig. 5d). There were several holes at the interface of CZTS/Mo, which would decrease the short circuit current density (J_SC) and fill factor (FF) of solar cells. Thus, we speculated that the several layers from molecular solution not only improved the interface of CZTS/Mo, but also enhanced the crystallinity of the as-prepared CZTS thin films.

Fig. 5 The cross-section SEM images of as-prepared CZTS thin film solar cells with a structure of Mo/CZTS/CdS/i-ZnO/AZO: (a) S0, (b) S3, (c) S5, (d) PS15.

After sulfurization and without further treatments, we used standard method to fabricate the CZTS thin film solar cells with the structure of Glass/Mo/CZTS/CdS/i-ZnO/AZO/Ag. The current density–voltage characterization of the as-prepared devices was obtained under AM 1.5 simulated illumination as shown in Fig. 6. The devices parameters were calculated based on the active device area (0.15 cm²) which did not include shadowing by Ag electrode. The total device area was 0.16 cm². Table 1 presented the detail parameters of the as-prepared devices. The power conversion efficiency (PCE) of sample S0 with pure nanoparticles-based fabrication was 2.1%, and the open circuit voltage (V_OC), J_SC and FF were 415 mV, 11.02 mA cm⁻² and 45.9%, respectively. The absorber layer of S0 device had some cracks, which could lead to low V_OC and J_SC. The cracks were removed by several molecular solution layers to cover the surface (S3 and S5). So, the PCE of sample S3 and S5 was improved to 4.18% and 4.92%. Moreover, we can also find that the V_OC, J_SC and FF of the sample S3 and S5 all improved greatly. However, when we used pure molecular solution to prepare thin film, the corresponding device performance was worse than S3 and S5. Particularly, the FF and J_SC of sample PS15 device decreased to 36.74% and 14.77 mA cm⁻², respectively. According to the result of the cross-section SEM image, there were some holes at the interface of CZTS/Mo, which were the main reason of decreasing the device performance. The V_OC was lower than the previous report, which maybe due to general interface recombination at CdS/CZTS or Mo/CZTS.⁴ Nevertheless, combining nanoparticle ink with molecular solution to prepare CZTS thin films could greatly improve the devices performance.

Fig. 6 I–V characterization of the as-prepared CZTS thin films solar cells under AM 1.5 illumination.

Table 1 The detail performance parameters of the as-prepared CZTS solar cells

Sample	VOC (mV)	JSC (mA cm^−2)	FF (%)	η (%)
S0	415	11.02	45.9	2.10
S3	502.5	17.89	46.51	4.18
S5	512.8	18.55	51.74	4.92
PS15	515.5	14.77	36.74	2.80

---

In order to further analyze the devices, external quantum efficiency (EQE) measurements were performed and the results were shown in Fig. 7. The devices presented good photo-responses in the range of 400–800 nm. At the near blue spectrum region, the EQE results exhibited a little decay, indicating the light losses in the CdS buffer layer, which might be attributed to the relative thickness of the CdS layer. On the other hand, nanoparticle ink-derived CZTS thin film had some cracks to cause quality of CdS. So, S0 had lower EQE than the others at the near blue spectrum region. However, at the near red spectrum region, the EQE results exhibited a steep decay, which may be due to high recombination losses in the bulk and depletion regions.¹⁰ As can be seen from Fig. 7, S5 showed the best EQE compared with the other devices, which was consistent with its J_SC. Fig. 8 demonstrated a plot of [hv ln(1 − EQE)]² vs. photon energy (hv) to estimate the band gap of the absorber layer. The band gaps were 1.56 eV for S0, 1.5 eV for S3 and S5, 1.49 eV for PS15, respectively. These results were also in agreement with the band gap of CZTS in previous reports.²¹ The different band gaps might be due to different composition.

Fig. 7 External quantum efficiency (EQE) curves from the CZTS solar cells.

Fig. 8 The band gap determination by the [hv ln(1 − EQE)]² vs. hv curves.

4. Conclusions

A low-cost, versatile, and easily up-scalable method to fabricate CZTS thin film using hydrophilic CZTS nanoparticle and molecular solution was developed. By combining nanoparticle ink with molecular solution, the as-prepared CZTS thin films showed compactness and good crystallinity. Moreover, it also could remove carbon-rich fine-grain layer and improve contact at interface of CZTS/Mo. After optimization, the as-prepared CZTS thin film consisting of 10 layers nanoparticle ink and 5 layers from molecular solution had the best photoelectric performance. CZTS thin film solar cells with a structure of Mo/CZTS/CdS/i-ZnO/AZO/Ag were fabricated and the best solar cell exhibited an active area efficiency of 4.92% without any anti-reflection coating.

Acknowledgements

This work has been financially supported by National Natural Science Foundation of China (61176062), a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions, the research fund of Jiangsu Province Cultivation base for State Key Laboratory of Photovoltaic Science and Technology (SKLPSTKF201506).

Notes and references

1. J. He, L. Sun, Y. Chen, J. C. Jiang, P. X. Yang and J. H. Chu, J. Power Sources, 2015, 273, 600–607.
2. H. Yoo and J. H. Kim, Sol. Energy Mater. Sol. Cells, 2011, 95, 239–244.
3. H. Katagiri, K. Jimbo, S. Yamada, T. Kamimura, W. S. Maw, T. Fukanol, T. Ito and T. Motohiro, Appl. Phys. Express, 2008, 1, 041201.
4. B. Shin, O. Gunawan, Y. Zhu, N. A. Bojarczuk, S. J. Chey and S. Guha, Prog. Photovoltaics Res. Appl., 2013, 21, 72–76.
5. X. Zeng, K. F. Tai, T. L. Zhang, C. W. John Ho, X. D. Chen, A. Huan, T. C. Sum and L. H. Wong, Sol. Energy Mater. Sol. Cells, 2014, 124, 55–60.
6. S. Ahmed, K. B. Reuter, O. Gunawan, L. Guo, L. T. Romankiw and H. Deligianni, Adv. Energy Mater., 2012, 2, 253–259.
7. Z. G. Li, J. C. W. Ho, K. K. Lee, X. Zeng, T. L. Zhang, L. H. Wong and Y. M. Lam, RSC Adv., 2014, 4, 26888–26894.
8. H. Xin, J. K. Katahara, I. L. Braly and H. W. Hillhouse, Adv. Energy Mater., 2014, 4, 1301823.
9. H. P. Zhou, W. C. Hsu, H. S. Duan, B. Bob, W. B. Yang, T. B. Song, C. J. Hsu and Y. Yang, Energy Environ. Sci., 2013, 6, 2822–2838.
10. Z. H. Su, K. W. Sun, Z. L. Han, H. T. Cui, F. Y. Liu, Y. Q. Lai, J. Li, X. J. Hao, Y. X. Liu and M. A. Green, J. Mater. Chem. A, 2014, 2, 500–509.
11. Z. H. Su, J. M. R. Tan, X. L. Li, X. Zeng, S. K. Batabyal and L. H. Wong, Adv. Energy Mater., 2015, 5, 19.
12. E. N. Gu, C. Yan, F. Y. Liu, Y. K. Liu, Z. H. Su, K. Zhang, Z. W. Chen, J. Li and Y. X. Liu, J. Mater. Sci.: Mater. Electron., 2015, 26, 1932–1939.
13. C. K. Miskin, W. C. Yang, C. J. Hages, N. J. Carter, C. S. Joglekar, E. A. Stach and R. Agrawal, Prog. Photovoltaics Res. Appl., 2015, 5, 654–659.
14. J. M. R. Tan, Y. H. Lee, S. Pedireddy, T. Baikie, X. Y. Ling and L. H. Wong, J. Am. Chem. Soc., 2014, 136, 6684–6692.
15. V. T. Tiong, Y. Zhang, J. Bell and H. X. Wang, RSC Adv., 2015, 5, 20178–20185.
16. W. Wang, H. L. Shen, X. C. He and J. Z. Li, J. Nanopart. Res., 2014, 16, 2437.
17. K. Woo, Y. Kim and J. Moon, Energy Environ. Sci., 2012, 5, 5340–5345.
18. S. Suehiro, K. Horita, K. Kumamoto, M. Yuasa, T. Tanaka, K. Fujita, K. Shimanoe and T. Kida, J. Phys. Chem. C, 2014, 118, 804–810.
19. P. A. Fernandes, P. M. P. Salomé and A. F. da Cunha, Thin Solid Films, 2009, 517, 2519–2523.
20. Q. J. Guo, G. M. Ford, W. C. Yang, C. J. Hages, H. W. Hillhouse and R. Agrawal, Sol. Energy Mater. Sol. Cells, 2012, 105, 132–136.
21. Y. L. Zhou, W. H. Zhou, M. Li, Y. F. Du and S. X. Wu, J. Phys. Chem. C, 2011, 115, 19632–19639.

---