Performance enhancement of perovskite solar cells with a modified TiO₂ electron transport layer using Zn-based additives

Abstract

Perovskite solar cells (PVSCs) have recently emerged as a very attractive option in film photovoltaics. Both solvent-engineering techniques and electron transport layer (ETL) properties are necessary for high-performance PVSCs. Herein, we report a one-step spin-coating approach for the preparation of uniform and dense perovskite layers, using a mixed solvent of dimethylacetamide (DMAC) and N-methyl-2-pyrrolidone (NMP) followed by exposure to toluene to induce crystallization. We also developed a simple and quantitative method to improve the traditional compact TiO₂ ETL properties and device performance based on Zn precursor doping. The effect of Zn-doped TiO₂ was investigated using atomic force microscopy, ultraviolet-visible spectra, photoluminescence, and open-circuit photovoltage decays. The results indicate that Zn-doped TiO₂ provides a better interface between the ETL and perovskite layer than non-doped TiO₂. Light Zn doping (2 vol%) was found to be the most effective additive, and the average power conversion efficiency improved from 13.61% to 15.25%.

---

Introduction

More than 5 years since the first perovskite solar cells (PVSCs) were fabricated using a perovskite light absorber as a sensitizer in 2009,¹ PVSCs continue to attract attention from researchers. After continuous development and attempts to perfect PVSCs, a high power conversion efficiency (PCE) of 20.1% has been obtained,² and this spectacular progress in thin-film solar cells is unsurpassed. In practice, the performance improvement of PVSCs, whether they have planar- or mesoscopic-type device structures, is mainly due to the morphology control of the perovskite films and interface-engineering techniques.³ The growth of crystalline perovskite films is strongly dependent on the deposition method used, precursor composition, solution concentration, solvent choice, and so on.⁴ Solvent-engineering techniques are an effective means of improving the efficiency of the cells by optimizing the film morphology and coverage by manipulating the perovskite nucleation and growth.^5–8 Hence, the enhanced efficiency is attributed to homogeneous high-quality perovskite thin films. Moreover, the electron transport layer (ETL) properties are also crucial to the cell performance, especially for planar-structure devices, because a compact layer (for example TiO₂ or ZnO) is the only ETL, and plays a double role in electron-injection and hole-blocking in planar-structure PVSCs.⁹ The use of doping techniques is a simple and effective way of modifying the electrical and optical properties of films. They are widely used in the ETL of PVSCs, and doping can be achieved by adding Nb, Y, and graphene nanoflakes in TiO₂ layer.^10–12 The introduction of these additives mainly improves electron transport properties and improves the photovoltaic performance. However, this also leads to scarcer adulterated elements and more complex procedures, which can increase the cost.

In this study, we developed a simple and practical method to introduce Zn into the compact TiO₂ ETL of PVSCs, which can result in better electron transport properties of the interface between the TiO₂ and perovskite. The effect of Zn-doped TiO₂ on the device performance was investigated. Our results indicate that the PVSCs with light Zn doping (2 vol%) result in enhanced device performance and excellent stability against ambient air without encapsulation. In addition, we obtained uniform and dense perovskite layers using a one-step, solvent-induced method involving spin-coating of a mixed solvent of dimethylacetamide (DMAC) and N-methyl-2-pyrrolidone (NMP) followed by toluene drop-casting to induce fast crystallization. This solvent-engineering and ETL modification technology enabled planar-heterojunction PVSCs, yielding a maximum PCE of 16.2% under standard AM 1.5 conditions.

Experimental section

Materials

CH₃NH₃I was synthesized by reacting 27.86 ml methylamine (40% in methanol, Junsei Chemical Co.) and 30 ml hydroiodic acid (57 wt% in water, Aldrich) in a 250 ml round-bottomed flask at 0 °C for 2 h with stirring. The precipitate was recovered by evaporation at 50 °C for 1 h. Then, the resulting CH₃NH₃I, was dissolved in ethanol, recrystallized from diethyl ether, and dried at 60 °C in a vacuum oven for 24 h. Further, 2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenyl amine)-9,9′-spirobifluorene (spiro-MeOTAD) was purchased from Aldrich and used for the preparation of hole transporting material (HTM).

Solar cell fabrication

Fluorine-doped, transparent, conducting SnO₂-coated (fluorine-doped tin oxide; FTO) glass substrates (OPV Tech Co., Ltd) were cleaned sequentially by ultrasonication in ethanol, acetone, and ultrapure water, and were subsequently treated in an oxygen plasma cleaning machine for 10 min. The FTO substrate was immersed in 0.04 M aqueous TiCl₄ solution at 70 °C for 1 h as a preprocessing step. A 0.15 mol l⁻¹ titanium isopropoxide with 50 μl acetylacetone was slowly added dropwise under stirring to a 2 mol l⁻¹ HCl solution in 25 ml absolute ethanol, and the TiO₂ precursor solution was synthesized. Different amounts of zinc acetate dehydrated solution (0.1 mol l⁻¹ of ethylene glycol monomethyl ether (EGME)) were added to the TiO₂ precursor solution with volume ratios of 0.01, 0.02, 0.04, and 0.06. The precursor solution changed from colorless to bright yellow to yellow with the increasing volume ratios (Fig. S1† in the ESI). Then, the above-mentioned solutions were coated on the substrates by spin coating at 3000 rpm for 30 s and the resultant film was annealed at 450 °C for 60 min.

The previously synthesized CH₃NH₃I powder and PbI₂ (99.5%, Alfa-Aesar) were stirred in a mixture of DMAC and NMP (5 : 1 v/v) at 60 °C for 12 h. The resulting concentration of the CH₃NH₃PbI₃ solution was 1.2 mol l⁻¹. The resulting solution was coated onto the TiO₂/FTO substrate using a consecutive spin-coating process at 800 rpm for 10 s and 5500 rpm for 40 s. During the high-speed spin-coating, toluene (50 μl) was quickly dropped onto the center of the substrate. This instantly changed the color of the substrate from transparent to light brown. The film was then subjected to annealing on a hot plate at 105 °C for 10 min. Subsequently, the spiro-MeOTAD-based hole-transfer layer (i.e., 170 mg spiro-MeOTAD, 28.5 μl 4-tert-butylpyridine, and 20 mg lithium-bis(trifluoromethanesulfonyl)imide (Li-TFSI), all dissolved in 1 ml chlorobenzene) was deposited by spin coating at 4000 rpm for 30 s. Before evaporating the silver electrodes, spiro-MeOTAD was allowed to oxidize in dry air overnight at room temperature. Finally, a 120 nm-thick silver layer was deposited by vacuum evaporation at a pressure of 1.7 × 10⁻³ Pa.

Characterization

The current–voltage (J–V) curves of the solar cells were determined in ambient air using a Keithley 2400 source meter (Keithley Instruments Inc., Cleveland, OH, USA). The cells were illuminated using a 450 W Class AAA simulator equipped with an AM 1.5G filter (XES-40S1) at a calibrated intensity of 100 mW cm⁻², as determined by a standard silicon reference cell. The J–V scan rate was 100 mV s⁻¹. The effective area of the cell was set to 0.09 cm² by using a non-reflective metal mask. The crystallographic structure of the perovskite was analyzed by X-ray diffraction (XRD) (D/MAX Ultima III, Rigaku Corporation, Tokyo, Japan) using Cu Kα radiation. The morphology was determined by scanning electron microscopy (SEM). Atomic force microscopy (AFM) images were obtained using a Digital Instrument NanoScope NS3A system to study the surface morphologies, including the surface roughness. UV-visible (UV-Vis) spectra were recorded using a Hitachi U-3010 spectrophotometer (Hitachi, Ltd., Chiyoda, Tokyo, Japan). Steady-state photoluminescence (PL) spectra were measured using a fluorescence spectrophotometer (PerkinElmer Instruments, LS-45), where the excited wavelength was 460 nm and was detected in the wavelength range of 500 to 900 nm with 10 nm increments. The open-circuit photovoltage decays (OCVD) were recorded using a digital oscilloscope (GDS3352, 350 MHz). External quantum efficiency (EQE) measurements were carried out using a system consisting of a Xe lamp (300 W) with a monochromator (Oriel 74100). The light intensity was measured with an optical power meter (OphirOptronics 70310) equipped with a calibrated thermopile head (OphirOptronics 71964). To investigate the electrical properties of the interfaces, electrochemical impedance spectroscopy (EIS) was performed using a Zahner IM6ex electrochemical workstation, in which a perturbation of 10 mV was applied and the frequency ranged from 1 Hz to 1 MHz.

Results and discussion

Fig. 1(A) presents the XRD patterns of the CH₃NH₃PbI₃ thin films prepared by solvent-induced one-step deposition and the inset shows an image of a perovskite film. Strong diffraction peaks at 14.08°, 28.40°, and 31.86° can be respectively assigned to (110), (220), and (310) diffractions of the tetragonal CH₃NH₃PbI₃ phase. Based on the solvent-engineering procedure, when the perovskite film was deposited using a mixture of DMAC and NMP as precursor solvents, followed by a toluene drip while spinning, extremely smooth and dense layers were formed. The solvent mixture needs to have a high-low volatility and good compatibility, so that the crystallizing rate of CH₃NH₃PbI₃ can be controlled. The toluene does not dissolve the perovskite materials, and can reduce the solubility of CH₃NH₃PbI₃ in the mixed solvent, thereby promoting fast nucleation and growth of the perovskite crystals. The SEM images (Fig. 1(B) and (C)) clearly show that the perovskite film has a full coverage and a homogeneous surface. Hence, it is proved that the solvent-engineering procedure is effective for the deposition of perovskite films.

Fig. 1 Morphological and structural characterization of CH₃NH₃PbI₃ films. (A) Indexed XRD pattern, and (B) low- and (C) high-magnification SEM micrographs of MAPbI₃ perovskite films deposited on glass/FTO/bl-TiO₂ substrate.

Fig. 2(A) shows the device configuration of the PVSCs used in this study. Electron–hole pairs are generated in perovskite CH₃NH₃PbI₃ following light absorption, and charge separation can then occur.¹³ Electrons and holes are subsequently injected into an n-type ETL (such as TiO₂) and a p-type hole-transporting material (HTM) (such as spiro-OMeTAD), respectively. The TiO₂ compact layer plays an important role in reducing the surface recombination rate at the interface and the transport electrons as well. Fig. 2(B) shows the J–V curves of the PVSCs with the non-doped and Zn-doped TiO₂ compact layers. The photovoltaic characteristics are summarized in Table 1. The ETLs at increasing volume ratios of the Zn–Ti precursor solution of 0, 0.01, 0.02, 0.04, and 0.06 are expressed as non-doped TiO₂, 0.01ZTP, 0.02ZTP, 0.04ZTP, and 0.06ZTP, respectively. The non-doped TiO₂ cell showed a moderate performance, a short-circuit current density (J_sc) of 21.78 mA cm⁻², and a PCE of 13.61%. After adding 1 and 2 vol% Zn-containing solutions to the TiO₂ ETL, the J_sc increased to 22.56 and 23.54 mA cm⁻², respectively. Meanwhile, the filling factor (FF) and PCE also improved as expected. Nevertheless, the overall device performance declined when the doping amount was up to 4 and 6 vol%. This result shows that the 0.02ZTP-based cell exhibited the best performance.

Fig. 2 (A) Schematic of the PVSCs device structure. (B) J–V curves of the perovskite solar cells with the none-doped and Zn-doped TiO₂ compact layers.

Table 1 Photovoltaic parameters (average values) for the perovskite solar cells with different amount of Zn doping

	Voc (V)	Jsc (mA cm^−2)	FF (%)	PCE (%)
None-doped TiO2	1.01 ± 0.01	21.78 ± 0.36	61.9 ± 2.23	13.61 ± 0.87
0.01ZTP	1.01 ± 0.02	22.56 ± 0.43	64.6 ± 1.95	14.68 ± 1.08
0.02ZTP	1.03 ± 0.01	23.54 ± 0.29	63.0 ± 1.86	15.25 ± 0.82
0.04ZTP	0.99 ± 0.02	21.67 ± 0.61	57.1 ± 3.07	12.19 ± 1.35
0.06ZTP	1.01 ± 0.03	20.46 ± 0.72	53.7 ± 3.44	11.10 ± 1.49

---

In order to elucidate the mechanism of the enhanced performance of the ZTP-based PVSCs, a series of characterizations were done. The surface topography and crystal structure of the ZTP-based ETL were determined using SEM and XRD (Fig. S2 and S3† in the ESI). As is well known, the morphology of FTO includes F-doped SnO₂ grains with sizes ranging from tens to hundreds of nanometers, and can be distinguished from that of the thin ETL.¹⁴ The crystallinity and coverage of the compact ETL increase with the increase of additives, and the XRD results exhibit a constantly enhanced anatase and brookite phase of the TiO₂ crystal at 25.281° and 64.601°, respectively. Meanwhile, the Zn oxides is not noticed because of the high solubility of TiO₂ crystal type for Zn doping and the extreme thinness of the compact ETL. The elements of compact films were further investigated by energy dispersive X-ray spectroscopy (EDX) and X-ray photoelectron spectroscopy (XPS) spectra, as shown in Fig. S4 and S5.† The SEM-EDX and XPS spectrum for 0.02ZTP compact layer demonstrated the existence of zinc in the TiO₂ film after annealed at 450 °C, and EDX parameters for normalized atomic percentage of none-doped TiO₂ and 0.02ZTP based ETL films were summarized in Table S1.† Based on XPS spectrum, the atomic concentration of C 1s, O 1s, Ti 2p and Zn 2p3 for 0.02ZTP compact film are 43.02, 46.92, 9.87 and 0.19, respectively. The surface roughness of the ETL films was further examined by high-resolution AFM (Fig. 3 and Fig. S6†). We observed a change in the morphologies and surface area of the ETL films with a varying doping amount. The measured average roughness (R_a) values when the amount of Zn-containing solutions was 0, 1, 2, 4, and 6 vol% were 21.8, 19.1, 20.2, 23.5, and 45.5 nm, respectively; the surface areas of these films were 27.9, 27.2, 27.5, 28.8 and 39.0 μm², respectively. This confirmed that the R_a of over 0.04ZTP-based ETLs increased significantly, and hence, the ∼0.02ZTP-based ETL exhibited proper surface roughness.

Fig. 3 AFM topographical images (10 μm × 10 μm) of ETL films based on various doping amount. (A and D) None-doped TiO₂, (B and E) 0.02ZTP, (C and F) 0.06ZTP.

Fig. S7† shows the UV-Vis spectra of different ZTP-based ETLs. Interestingly, the intensity of the UV absorption peak increased with an increase in the amount of the Zn-containing solutions, which shows that Zn-doping improved the UV absorption spectrum. That is, light contact area increased with the increase of surface areas of these films. The optical band gaps derived from Tauc plot, as shown in the inset of Fig. S7,† demonstrate the smaller band gaps of ZTP-based ETLs than the none-doped TiO₂. The energy level diagram of perovskite solar cells with none-doped TiO₂ and 0.02ZTP compact layer is shown in Fig. S8.† The band gaps of none-doped and 0.02ZTP TiO₂ are 3.43 eV and 3.29 eV, respectively. The modified work function of 0.02ZTP-based ETL could be well matched to that of the CH₃NH₃PbI₃ emissive layer, thereby facilitating efficient electron injection at the interface. Fig. 4 presents the absorption spectra of perovskite films on different compact ETL substrates. All the samples exhibited panchromatic absorption of light with spectra extended over the visible to near-infrared region. However, the 0.04ZTP- and 0.06ZTP-based samples showed lower light harvesting, implying that rough substrates of perovskite films may decrease the optical absorption coefficient. We observed that the 0.02ZTP-based ETL exhibited stronger absorption in the wavelength range of 300–550 nm, owing to the appropriate doping amount and R_a.

Fig. 4 Absorption spectra of perovskite films on different compact ETLs substrate.

The OCVD technique involves turning off the illumination in a steady state and monitoring the subsequent decay of the photovoltage, V_oc.¹⁵ We studied the effect of different ETLs on the electron recombination and lifetime for devices using the OCVD method. Fig. 5 illustrates the voltage decay curves of the PVSCs with the non-doped and ZTP-based ETLs. The figure indicates that the electron lifetimes of the 0.02ZTP-based cells are longer than those of the non-doped ETLs. In addition, the 0.04ZTP- and 0.06ZTP-based devices have reduced electron lifetimes. Accordingly, we found that the right amount of Zn-doping can delay the electron recombination process, which is beneficial for electron transportation as the electron lifetimes increase.

Fig. 5 OCVD curves of the perovskite solar cells with the none-doped TiO₂ and ZTP-based ETLs.

Fig. 6(A) compares the PL intensities of the perovskite films for different ETLs and the emission peaks appearing in the 775–800 nm range. Most of the samples exhibit weak emission, which indicates effective electron injection. However, the 0.06ZTP-based CH₃NH₃PbI₃ films have a strong photoluminescence, indicating poor electron injection, probably resulting from the rough ETL with excess doping. The EQE spectra are shown in Fig. 6(B), and a significant contribution at wavelengths between 300 and 800 nm was revealed. The integrated J_sc values of the spectral responses are 21.73, 22.52, 23.37, 21.36, and 20.34 mA cm⁻², respectively, for the devices prepared with Zn-containing solution amounts from 0 to 6 vol% under irradiation, which agree with the J–V measurement results. To analyze the interfacial recombination at the interface between the ETL and CH₃NH₃PbI₃, impedance spectroscopy measurements were performed in the dark under a forward bias of 600 mV. Fig. 6(C) presents the Nyquist plots of the PVSCs with the none-doped and ZTP-based ETLs, and the inset shows the equivalent circuit consisting of a series resistance (R_s), a recombination resistance (R_rec), and a chemical capacitance (C), which is replaced with a constant phase element (CPE). In the devices with the non-doped and ZTP-based ETLs, no distinct transmission line (TL) behavior is observed, which is likely due to the very thin ETLs employed.¹⁶ In the circuit model, R_s represents the ohmic resistance due to electrodes. The impedance spectra are dominated by a large semicircle and the R_rec is related to the electron and hole recombination within the solar cells.¹⁷ As can be seen, the R_rec of the 0.02ZTP-based cell is obviously larger than that of the 0.06ZTP-based cell, agreeing well with the results discussed above. Fig. 6(D) shows the recombination resistance as a function of the applied bias in the relevant voltage range; a decrease in R_rec is observed with increasing forward bias voltage. The device with the 0.02ZTP-based ETL exhibited a higher R_rec than the device with the 0.06ZTP-based ETL. Therefore, it is concluded that the charge recombination in the 0.02ZTP-based cell has been significantly suppressed. In other words, a lower recombination rate occurs at low Zn doping levels (≤4 vol%). Finally, the long-term stabilities of the PVSCs with the non-doped and 0.02ZTP-based ETLs under 35% humidity were tested. The 0.02ZTP-based devices were more stable, retaining over 80% of their initial PCEs over 28 days (Fig. S6†), which may be due to a good contact interface between the ETL and perovskite layer.

Fig. 6 (A) PL of perovskite films on different compact ETLs substrate. (B) EQE spectra and (C) Nyquist plot of the PVSCs with the none-doped TiO₂ and ZTP-based ETLs; (D) variation of R_rec with the applied bias.

In summary, we reported solvent-engineering techniques for the formation of uniform and dense perovskite layers. We successfully demonstrated a simple and practical approach, which is based on the incorporation of zinc acetate dihydrate solution into a TiO₂ precursor. The Zn doping resulted in negative effects on electron transport and the photovoltaic properties depending on the doping amount. We consequently found that lightly doped Zn precursors can reduce electron recombination and improve the optical and electrical properties of the ETL/perovskite layer, resulting in enhanced device performance, particularly pertaining to J_sc and FF. These results can be used to pursue the design of appropriate n-type ETLs and fabrication of perovskite films.

Acknowledgements

This work was supported by the Privileged Development Program of Jiangsu High Education on New Energy Material Science and Engineering, the National Natural Science Foundation of China (Grant No. 51572037), the Jiangsu Province Industry-University-Research Joint Innovation Fund (BY2013024-01), the Natural Science Foundation of the Jiangsu Higher Education Institutions of China (Grant No. 14KJA430001, EEKJA48000).

References

1. A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 2009, 131, 6050–6051.
2. W. S. Yang, J. H. Noh, N. J. Jeon, Y. C. Kim, S. Ryu, J. Seo and S. I. Seok, Science, 2015, 348, 1234–1237.
3. X. Qifan, S. Chen, H. Zhicheng, H. Fei, Y. Hin-Lap and C. Yong, Acta Chim. Sin., 2015, 73, 179–192.
4. X. Gong, M. Li, X. B. Shi, H. Ma, Z. K. Wang and L. S. Liao, Adv. Funct. Mater., 2015, 25, 6671–6678.
5. S. D. Stranks and H. J. Snaith, Nat. Nanotechnol., 2015, 10, 391–402.
6. N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. Ryu and S. I. Seok, Nat. Mater., 2014, 13, 897–903.
7. K. Wang, C. Liu, P. Du, H. L. Zhang and X. Gong, Small, 2015, 11, 3369.
8. Y. Rong, Z. Tang, Y. Zhao, X. Zhong, S. Venkatesan, H. Graham, M. Patton, Y. Jing, A. M. Guloy and Y. Yao, Nanoscale, 2015, 7, 10595.
9. M. He, D. Zheng, M. Wang, C. Lin and Z. Lin, J. Mater. Chem. A, 2014, 2, 5994–6003.
10. D. H. Kim, G. S. Han, W. M. Seong, J. W. Lee, B. J. Kim, N. G. Park, K. S. Hong, S. Lee and H. S. Jung, ChemSusChem, 2015, 8, 2392–2395.
11. H. Zhou, Q. Chen, G. Li, S. Luo, T. B. Song, H. S. Duan and Y. Yang, Science, 2014, 345, 542.
12. J. T. W. Wang, J. M. Ball, E. M. Barea, A. Abate, J. A. Alexander-Webber, J. Huang, M. Saliba, I. Mora-Sero, J. Bisquert, H. J. Snaith and R. J. Nicholas, Nano Lett., 2013, 14, 724–730.
13. J. M. Ball, S. D. Stranks, M. T. Hörantner, S. Hüttner, W. Zhang, E. J. W. Crossland, I. Ramirez, M. Riede, M. B. Johnston, R. H. Friend and H. J. Snaith, Energy Environ. Sci., 2015, 8, 602–609.
14. Y. Wu, X. Yang, H. Chen, K. Zhang, C. Qin, J. Liu, W. Peng, A. Islam, E. Bi and F. Ye, Appl. Phys. Express, 2014, 7, 052301.
15. J. Bisquert, A. Zaban, M. Greenshtein and I. Mora-Seró, J. Am. Chem. Soc., 2004, 126, 13550–13559.
16. J. C. Yu, D. B. Kim, G. Baek, B. R. Lee, E. D. Jung, S. Lee, J. H. Chu, D. K. Lee, K. J. Choi, S. Cho and M. H. Song, Adv. Mater., 2015, 27, 3492–3500.
17. J. Dong, Y. Zhao, J. Shi, H. Wei, J. Xiao, X. Xu, J. Luo, J. Xu, D. Li, Y. Luo and Q. Meng, Chem. Commun., 2014, 50, 13381–13384.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra01149g

---