Thickness of the hole transport layer in perovskite solar cells: performance versus reproducibility

Abstract

We studied the effect of the hole transport layer (HTL) thickness on photovoltaic properties of meso-superstructured perovskite solar cells based on CH₃NH₃PbI_3−xCl_x. We found that there is an interplay between photovoltaic performance and reproducibility: thinning the HTL increased performances of the devices but reduced their reproducibility.

---

Recent findings of organic-inorganic hybrid perovskites of formula MAPbX₃ (where, MA = methylammonium cation, X = Cl, Br or I) have attracted the attention of researchers in various fields, to develop thin layer photocells with a high power conversion efficiency. This is mainly due to their excellent visible light absorption, high charge carrier mobilities, high carrier diffusion lengths and ambipolar transport characteristics.^1–9 Although perovskite solar cells are fabricated in various device configurations such as meso-superstructured,¹⁰ pillared structure,¹¹ planar^12,13 and hole-conductor-free structure,^14,15 they generally consist of fluorine-doped tin oxide (FTO), compact hole blocking layer, perovskite absorber, hole transport layer (HTL), and counter electrode. In principle, photogenerated electrons should migrate through the perovskite layer to the FTO, and holes should migrate through the HTL to the counter electrode for the continuous operation with effective transformation of light to electricity.^16–18

Considering the operation of perovskite photovoltaic device, it is obvious that the most important layer is perovskite and thus, optimization of perovskite layer morphology, thickness and surface passivation has been intensively investigated by various groups. Sequential deposition by M. Grätzel firstly lead to the efficiency of 15%.¹¹ Similarly, interdiffusion of solution-processed precursor stacking layer also induced uniform perovskite film formation, leading to the efficiency of 17%.¹⁹ Recently, Seok et al. reported intramolecular exchange method to prepare smooth perovskite layers which exhibited power conversion efficiency in excess of 20%.⁵

In contrast, studies regarding the optimization of the most widely-utilized HTL in perovskite solar cells, 2,2′7,7′-tetrakis-(N,N-di-p-methoxyphenyl amine)-9,9′-spirobifluorene (spiro-MeOTAD), are rare and therefore warrant immediate investigation.²⁰ The HTL primarily serves two main functions: transporting photogenerated holes to the counter electrode; and preventing direct contact between perovskite layer and the counter electrode.³ Both of these functions are vitally important for achieving high performance and stability. A hole transport material (HTM) with high hole mobility is preferred, because it would result in fast transport of holes to the counter electrode.^21,22 The thickness of the HTL affects both performance and stability of devices. Thinning the HTL should reduce the distance traveled by holes to reach the counter electrode, and reduce the probability that they will undergo a recombination event; however, at the same time it might reduce the uniformity of coverage, especially when the perovskite surface is composed of large crystallites. In addition, it is well known fact that the counter electrode reacts with iodide ion in perovskite and quickly degrades the devices. Therefore, it is important to avoid direct contact between perovskite and counter electrode by employing sufficiently thick HTL.²³ Thicker HTL would cover the rough perovskite surface uniformly, but would cause high series resistance due to much lower hole mobility of HTL as compared to perovskite. Herz et al. reported the charge carrier mobility in a mixed halide perovskite (MAPbI_3−xCl_x), is 11.6 cm² V⁻¹ s⁻¹, which is much higher than that (10⁻⁴ to 10⁻³ cm² V⁻¹ s⁻¹) of spiro-MeOTAD doped with tert-butylpyridine (tBP) and lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) salt.^21,24 Therefore, it is vitally important to carefully control the thickness of HTL to obtain full coverage of the rougher perovskite layer, without increasing the series resistance of the devices. This would lead to devices with high performance, stability and reproducibility.

Here, we report how the thickness of a HTL influences the performances and reproducibility in the meso-superstructured solar cells comprising FTO, a compact TiO₂ (c-TiO₂) hole blocking layer, a Al₂O₃ scaffold layer, a MAPbI_3−xCl_x perovskite absorber layer, a spiro-MeOTAD HTL, and a Ag counter electrode. To the best of our knowledge, this is the first report describing the effect of HTL on photovoltaic performance of meso-superstructured photovoltaic devices that use mixed-halide perovskite.

Results and discussion

Fig. 1a is a secondary ion mass spectroscopy (SIMS) depth profile showing vertical chemical composition of a device-FTO/Al₂O₃/perovskite/HTL. Illustration of the as prepared device configuration is shown in Fig. 1b, which shows 3 distinct layers of FTO, Al₂O₃-perovskite and HTL. However, there existed three layers of HTL, HTM/perovskite and HTM/perovskite/Al₂O₃ (Fig. 1c), due to the gradient composition of perovskites in the range of 160–330 nm depth, judging from the intensities of lithium (Li), lead (Pb), and Tin (Sn), representing spiro-MeOTAD HTL with additives (LiTFSI and tBP), perovskite, and FTO, respectively. This suggests the formation of the partial protruding perovskite crystals with ca. 140 nm-thickness, as illustrated in Fig. 1d. It can be seen from the SIMS profile that Li atoms are present until ca. 800 nm depth, indicating the penetration of spiro-MeOTAD inside of the porous perovskite/Al₂O₃ scaffold layer.

Fig. 1 (a) Depth profile in perovskite hybrid solar cell using secondary ion mass spectroscopy (SIMS). Schematic device structures for 3 types: (b) prepared, (c) measured and (d) possible structures.

The extent to which the HTL covers the protruding perovskite crystals might greatly influence both photovoltaic performance of the device, and the reproducibility of device fabrication. Therefore, we further investigated how the formation of the partially-protruding perovskite crystals on the perovskite layer influenced the thickness and coverage of HTL. We prepared spiro-MeOTAD HTLs of different thicknesses on the perovskite layer infiltrated into Al₂O₃ scaffold layer. Scanning electron microscopy (SEM) images of perovskite layer on Al₂O₃ and covered with thick (ca. 400 nm) and thin (ca. 100 nm) spiro-MeOTAD are shown in Fig. 2a, b and c, respectively.

Fig. 2 Top scanning electron microscope (SEM) images and atomic force microscope (AFM) images (scan area: 5 μm × 5 μm) with roughness profile of (a) perovskite film on Al₂O₃ scaffold layer, (b) thick (∼400 nm) spiro-MeOTAD film on (a) layer and (c) thin (∼100 nm) spiro-MeOTAD film on (a) layer. Cross section images of (d) thick spiro-MeOTAD layer which can block perovskite from electrode and (e) thin spiro-MeOTAD layer which cannot block perovskite from electrode.

The perovskite layer on the Al₂O₃ scaffold is intrinsically very rough as can be seen from SEM and atomic force microscope (AFM) image (Fig. 2a), giving ca. 82.4 nm of RMS value. The rough surface becomes much smoother (RMS = ca. 27.5 nm) when the thicker HTL is coated, indicating that spiro-MeOTAD layer fully covers the rough perovskite surface (Fig. 2a). This thick layer could increase series resistance of the device, but it also prevents direct contact of perovskite layer with the counter electrode as shown in cross-sectional SEM image in Fig. 2d. This may probably give less efficient device with a reasonable reproducibility. Meanwhile, rough morphology (RMS = ca. 50.4 nm) was still observed at the surface employing the thinner HTL layer, suggesting that bare perovskites could be partially exposed as exemplified in Fig. 2e. These partially exposed, bare perovskite crystals could form direct contact with a counter electrode, which could cause degradation of counter electrode as well as perovskite. In addition, there is a possibility of counter electrode penetrating the perovskite layer, thereby short-circuiting the device. Therefore, it is important to carefully optimize the thickness of HTL, in order to achieve full coverage of perovskite as well as low series resistance which will lead to high power conversion efficiencies as well as reproducibility.

We further measured the charge carrier mobility of spiro-MeOTAD doped with tBP and LiTFSI, by measuring the space-charge-limited current (SCLC) of the hole-only devices of configuration ITO/PEDOT-PSS/spiro-MeOTAD/Au (Fig. 3a). At higher voltages, assuming ohmic contact and trap free transport, mobility from SCLC can be calculated by Mott–Gurney's equation,

where J is the current density, ε_r is the dielectric constant of the organic semiconductor (generally 3), ε₀ is the permittivity of free space, μ is the mobility, V is the voltage, and L is the thickness of the active layer. The mobility of spiro-MeOTAD calculated from above equation is of the order of ∼10⁻³. All the devices showed nearly similar mobility value regardless of the thickness of HTL, which is well-matched with previous reports.^21,25 To determine the charge carrier mobility during the operation of photovoltaic devices, photo CELIV (charge extraction by a linearly-increasing voltage) was measured by fabricating complete devices and the mobility can be calculated using
where d is the thickness of the active layer, A is the ramp rate, t_max is the time at the maximum Δj of the extraction peak, and j(0) is the capacitive displacement current.^26–28 The mobility can be inferred from t_max at which current density is maximal; t_max is inversely proportional to mobility. As shown in Fig. 3b, our devices with various thickness of HTL show similar values of t_max, indicating that all the devices have similar charge mobility in operation. Because the mobility in device is not dependent on the thickness of HTL, it can be concluded that the photovoltaic performance would be higher as the thickness of HTL becomes thinner, because time for charges to move towards counter electrodes is reduced, and charge carriers undergo less recombination events before reaching each electrode.

Fig. 3 (a) Space-charge limited current (SCLC) of hole-only device in various thickness of spiro-MeOTAD (b) photo-CELIV (charge extraction by a linearly increasing voltage) measurement in meso-superstructured perovskite solar cell in various thickness of HTL.

We fabricated photovoltaic devices with various thickness of HTL. Fig. S1† shows cross-sectional SEM images of the meso-superstructured perovskite hybrid solar cells with various thickness of HTL, occasionally showing protruding perovskite crystals in thin HTL. The thickness of HTL was controlled by changing the spin-coating speed of spiro-MeOTAD, because controlling the concentration of spiro-MeOTAD also needs additional optimization of additives (tBP and LiTFSI). We fabricated more than 40 devices to obtain statistical data and to see the reproducibility of photovoltaic parameters. J–V curves of the best performing devices of each HTL thickness are shown in Fig. 4. The photovoltaic parameters extracted from J–V curves are summarized in Table 1. As expected, the power conversion efficiencies are much higher when employing thinner HTL, than when employing thicker HTL. The device that used the 180 nm spiro-MeOTAD layer showed best power conversion efficiency (PCE) of 15.5% with 23.5 mA cm⁻² of short-circuit current (J_SC), 1.02 V of open-circuit voltage (V_OC) and 64.7% of fill factor (FF). Especially, as HTL becomes thinner, FF increases gradually due to low series resistance. It agrees well with our hypothesis that a thin HTL would have a good effect on the performance, due to its low resistance and reduced time for charges to move toward each electrode. However, the deviation in PCE (Fig. 5a), J_SC (Fig. 5b) and FF (Fig. 5c) increased as thickness of HTL decreased, as predicted from SEM and AFM images that showed protruding perovskite crystals (Fig. 2). In contrast, the variation in V_OC is rather small (Fig. 5d), because the highest occupied molecular orbital (HOMO) energy level of spiro-MeOTAD is nearly constant regardless of its thickness, as measured from ultraviolet photoelectron spectroscopy (UPS) measurements (Fig. S2†).

Fig. 4 J–V curves for meso-superstructured devices, measured under simulated 1 sun illumination, depending on the thickness of spiro-MeOTAD layer.

Table 1 Summary of the best-performing device parameters about each thickness; short-circuit current (J_SC), open-circuit voltage (V_OC), fill factor (FF) and power conversion efficiency (PCE)^a

HTL thickness (nm)	JSC (mA cm^−2)	VOC (V)	FF (%)	PCE (%)	Rseries^b (Ω cm^2)
a Cell size: 0.09 cm^2.b Series resistance.
700	22.2	1.05	46.3	10.8	11.73
550	21.3	1.04	51.8	11.4	8.00
450	20.2	1.00	52.5	10.6	9.09
400	20.7	0.97	53.2	10.7	9.93
250	21.5	1.00	64.6	14.0	5.35
180	23.5	1.02	64.7	15.5	5.00
100	12.5	0.43	43.2	2.3	—

---

Fig. 5 (a). Statistical distribution of (a) short-circuit current (J_SC), (b) open-circuit voltage (V_OC), (c) fill factor (FF) and (d) power conversion efficiency (η) depending on the thickness of spiro-MeOTAD.

Most perovskite solar cells employing about 100 nm HTL did not work effectively because it cannot fully cover the protruding perovskite crystals, resulting in large leakage current. Therefore, it is necessary that the perovskite layer should be thicker than 150 nm, perfectly covering protruding perovskite crystals preventing direct contact between perovskite and electrode. In addition, it is observed that, thicker HTL gives lower performing photovoltaic devices, but with higher reproducibility as seen from box plots shown in Fig. 5. The main difference between high and low performing devices is J_SC and FF, which are related to series resistance. Employing thicker HTL layer having high mobility and conductivity in the present device architecture may provide the high performance and reproducible devices. Also, fabricating smoother perovskite capping layer with less thick HTL layer would result in better performing devices.

Conclusions

In conclusion, we successfully demonstrated the dependence of device performance on hole transport layer (HTL) thickness. We showed that there is interplay between performance and reproducibility. Thinner HTL (∼180 nm) gives rise to high performing devices, while thicker HTL layer provides more reproducible performance. Achieving conformal coating of HTL over rough perovskite layer is critically important to achieve high performance and reproducibility. However, deposition of such thin layers on rough perovskite layer is difficult by using conventional spin-coating method, showing low reproducibility. Our study put emphasis on the importance of fabricating uniform HTL with carefully optimized thickness for obtaining maximum photovoltaic performance.

Experimental section

Synthesis of a CH₃NH₃PbI_3−xCl_x perovskite

The CH₃NH₃PbI_3−xCl_x perovskite was synthesized as already reported. Firstly, CH₃NH₃I was synthesized by reacting 27.86 mL methylamine (40 wt% in methanol, Junsei Chemical Co.) and 30 mL hydroiodic acid (57 wt% in water, Aldrich) in a round-bottomed flask at 0 °C for 2 hours with stirring. The resulting materials were obtained in the process of evaporation and recrystallization and dried at 60 °C in a vacuum oven overnight. Then, synthesized CH₃NH₃I powder was mixed with PbCl₂ (Aldrich) with a ratio of 40 wt% in dimethylformamide (DMF).

Device fabrication

Solar cell devices were fabricated on 40 × 40 mm fluorine-doped tin oxide (FTO) coated glass substrates. The FTO coated glass substrates were etched by brushing with zinc powder and 2 M hydrochloric acid (HCl). Then, they were washed using detergent, deionized water (DI-water), acetone and isopropyl alcohol (IPA). The FTO coated glass substrates were under UV-ozone treatment for 15 minutes, and then coated by TiO₂ compact layer at 2000 rpm for 60 seconds as reported. The coated films were sintered at 500 °C. Al₂O₃ (1 : 2 vol%, Aldrich) layer was coated on them at 2500 rpm for 60 seconds and annealed at 150 °C for 1 hour. The perovskite solution was spin-coated at 2000 rpm for 30 seconds, and then slowly annealed at 100 °C for 2 hours. Spiro-MeOTAD solution (83.3292 mg mL⁻¹ in CB) with addition of Li-TFSI in acetonitrile (0.18 M) and tert-butylpyridine (1 mL in 9 mL CB) were spin-coated on the films at different speed, and oxidized for 24 hours. Finally, silver electrodes (120 nm) were vacuum-deposited at the pressure of ∼10⁻⁷ torr. Devices for space-charge-limited-current (SCLC) measurement were fabricated on the 25 × 25 mm indium tin oxide (ITO) coated glass substrates. Firstly, they were washed using detergent, DI-water, acetone and IPA. After washing, the substrates were dried in the 150 °C oven overnight and UV-ozone-treated for 15 minutes. PEDOT-PSS (Clevios P, VP AI 4083) was spin-coated on the substrates in the condition of 5000 rpm/30 seconds and annealed at 150 °C for 10 minutes. The spiro-MeOTAD solution in above condition was spin-coated at different speed on the PEDOT-PSS layer. After then, gold electrodes (100 nm) were deposited in a vacuum condition.

Device characterization

Using a Keithley 2400 SMU and an Oriel xenon lamp (450 W) with an AM1.5 filter, the solar cells were characterized in air under AM 1.5G illumination of 100 mW cm⁻² (Oriel 1 kW solar simulator), which was calibrated with a KG5 filter certified by NREL. The J–V curves of all devices were measured with 0.09 cm² of active area. SCLC was also measured in the same condition but in dark condition.

Atomic force microscopy (AFM) measurement

The atomic force microscope (using a VEECO Dimension 3100 + Nanoscope V) was operated in tapping mode to acquire the images of the surfaces of perovskite and HTL coated on the perovskite film.

Secondary ion mass spectroscopy (SIMS) measurement

The secondary ion mass spectroscopy (SIMS) (IMS 6F with Cs⁺ gun, CAMECA) was operated to acquire the atomic depth profile of perovskite hybrid solar cells.

Photo-CELIV measurement

Photo-CELIV was conducted on FTO/bl-TiO₂/Al₂O₃/CH₃NH₃PbI_3−xCl_x/spiro-MeOTAD/Ag devices in ambient air. The active area was defined by the size of the counter-electrode (0.09 cm²).

Acknowledgements

This work was supported by grants from the Nano Material Technology Development Program (2012M3A7B4049989) and the Center for Advanced Soft Electronics under the Global Frontier Research Program (Code No. NRF-2012M3A6A5055225) through the National Research Foundation of Korea (NRF) by the MSIP, Korea.

Notes and references

1. H.-S. Kim, C.-R. Lee, J.-H. Im, K.-B. Lee, T. Moehl, A. Marchioro, S.-J. Moon, R. H. Baker, J. H. Yum, J. E. Moser, M. Grätzel and N.-G. Park, Sci. Rep., 2012, 2, 1.
2. L. Etgar, P. Gao, Z. Xue, Q. Peng, A. K. Chandiran, B. Liu, M. K. Nazeeruddin and M. Grätzel, J. Am. Chem. Soc., 2012, 134, 17396.
3. M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami and H. J. Snaith, Science, 2012, 338, 643.
4. S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza and H. J. Snaith, Science, 2013, 342, 341.
5. N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. Ryu and S. I. Seok, Nat. Mater., 2014, 13, 897.
6. G. Xing, N. Mathews, S. Sun, S. S. Lim, Y. M. Lam, M. Grätzel, S. Mhaisalkar and T. C. Sum, Science, 2013, 342, 344.
7. J.-H. Im, C.-R. Lee, J.-W. Lee, S.-W. Park and N.-G. Park, Nanoscale, 2011, 3, 4088.
8. H. Zhou, Q. Chen, G. Li, S. Luo, T.-B. Song, H.-S. Duan, Z. Hong, J. You, Y. Liu and Y. Yang, Science, 2014, 345, 542.
9. D. Shi, V. Adinolfi, R. Comin, M. Yuan, E. Alarousu, A. Buin, Y. Chen, S. Hoogland, A. Rothenberger, K. Katsiev, Y. Losovyj, X. Zhang, P. A. Dowben, O. F. Mohammed, E. H. Sargent and O. M. Bakr, Science, 2015, 347, 519.
10. J. M. Ball, M. M. Lee, A. Hey and H. J. Snaith, Energy Environ. Sci., 2013, 6, 1739.
11. J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin and M. Grätzel, Nature, 2013, 499, 316.
12. J. H. Heo, H. J. Han, D. Kim, T. K. Ahn and S. H. Im, Energy Environ. Sci., 2015, 8, 1602.
13. J. H. Heo, D. H. Song, H. J. Han, S. Y. Kim, J. H. Kim, D. Kim, H. W. Shin, T. K. Ahn, C. Wolf, T.-W. Lee and S. H. Im, Adv. Mater., 2015, 27, 3424.
14. S. Gamliel, A. Dymshits, S. Aharon, E. Terkieltaub and L. Etgar, J. Phys. Chem. C, 2015, 119, 19722.
15. A. Mei, X. Li, L. Lin, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu, J. Chen, Y. Yang, M. Grätzel and H. Han, Science, 2014, 345, 295.
16. V. Gonzalez-Pedro, E. J. Juarez-Perez, W.-S. Arsyad, E. M. Barea, F. Fabregat-Santiago, I. Mora-Sero and J. Bisquert, Nano Lett., 2014, 14, 888.
17. Y. Zhao, A. M. Nardes and K. Zhu, J. Phys. Chem. Lett., 2014, 5, 490.
18. R. Lindblad, D. Bi, B.-W. Park, J. Oscarsson, M. Gorgoi, H. Siegbahn, M. Odelius, E. M. J. Johansson and H. Rensmo, J. Phys. Chem. Lett., 2014, 5, 648.
19. Z. Xiao, C. Bi, Y. Shao, Q. Dong, Q. Wang, Y. Yuan, C. Wang, Y. Gao and J. Huang, Energy Environ. Sci., 2014, 7, 2619.
20. N. Marinova, W. Tress, R. Humphry-Baker, M. I. Dar, V. Bojinov, S. M. Zakeeruddin, M. K. Nazeeruddin and M. Grätzel, ACS Nano, 2015, 9, 4200.
21. Y. S. Kwon, J. Lim, H.-J. Yun, Y.-H. Kim and T. Park, Energy Environ. Sci., 2014, 7, 1454.
22. G.-W. Kim, J. Kim, G.-Y. Lee, G. Kang, J. Lee and T. Park, Adv. Energy Mater., 2015, 5, 1500471.
23. Y. Han, S. Meyer, Y. Dkhissi, K. Weber, J. M. Pringle, U. Bach, L. Spiccia and Y.-B. Cheng, J. Mater. Chem. A., 2015, 3, 8139.
24. C. Wehrenfennig, G. E. Eperon, M. B. Johnston, H. J. Snaith and L. M. Herz, Adv. Mater., 2014, 26, 1584.
25. H. J. Snaith and M. Grätzel, Appl. Phys. Lett., 2006, 89, 262114.
26. J.-H. Im, I.-H. Jang, N. Pellet, M. Grätzel and N.-G. Park, Nat. Nanotechnol., 2014, 9, 927.
27. Y. Chen, J. Peng, D. Su, X. Chen and Z. Liang, ACS Appl. Mater. Interfaces, 2015, 7, 4471.
28. N. Anh, D.-Y. Son, I.-H. Jang, S. M. Kang, M. Choi and N.-G. Park, J. Am. Chem. Soc., 2015, 137, 8696.

---

Footnote

† Electronic supplementary information (ESI) available: Cross-sectional SEM images, and UPS spectra. See DOI: 10.1039/c5ra18648j

---