Fully Printed HTL-Free MAPbI₃ Perovskite Solar Cells with Carbon Electrodes

Abstract

This study investigates fully printed methylamine vapour-treated methylammonium lead iodide (MAPbI₃) hole transport layer (HTL)-free perovskite solar cells (PSCs) with a carbon electrode. We describe a method that can be used to deposit MAPbI₃ films in an ambient environment with doctor blading that is entirely free of spin coating and has precise morphology control, in which the varying input N₂ pressure affects the film morphology. Consequently, a fully printed perovskite solar cell with an ITO/SnO₂/MAPbI₃/carbon structure was fabricated using a doctor-blading SnO₂ electron transport layer and a screen-printed carbon counter electrode. The low-temperature-derived PSCs exhibited a superior power conversion efficiency (PCE) of 14.17% with an open-circuit voltage (V_oc) of 1.02 V on a small-active-area device and the highest efficiency of >8% for an illumination exposure area of 1.0 cm², with high reproducibility. This work highlights the potential of doctor blading and methylamine vapour treatment as promising methods for fabricating high-performance perovskite solar cells. A doctor-blading approach offers a wide processing window for versatile high-performance perovskite optoelectronics in the context of large-scale production.

1. Introduction

Perovskite photovoltaics (PVs) have gained significant attention in recent years due to their high power conversion efficiency (PCE) [1]. The accelerated development of applications for energy storage [2,3] is largely attributable to the superior photovoltaic properties of perovskite materials [4,5,6] and the solution approach for film fabrication [7,8]. The most prevalent method for preparing perovskite films for efficient devices is the combination of a single-step spin-coating methodology and an antisolvent strategy [9,10]. However, this method has limitations, such as a constrained timeframe for incorporating the antisolvent substance; this can result in a restricted threshold during the device’s development and potentially hinder the productivity and repeatability of perovskite production, particularly mass production [11,12]. Moreover, spin coating applies a potent rotational force to coat the thin films and extract superfluous solutions during film fabrication. Non-uniform gravitational pull at various substrate locations can result in substantial thickness disparities in the developed films, which is undesirable for large-area devices and multi-device units [13,14,15]. In short, it is difficult to scale up the spin-coating/antisolvent method to accommodate large-area surfaces and bulk production. Currently, numerous scalable approaches are being devised to address these issues for green energy [16,17], including spray deposition [18,19], printing deposition [20,21,22], brush-painting deposition [23], electrodeposition [24,25], slot-die deposition [26,27], soft-cover deposition [28], a pressure-processing method [29], inkjet printing [30], a blade-coating method [31] and a gas-assisted method [32], which need to be addressed for successful commercialization, especially the suppression of the coffee-ring [33] effect that can cause film non-uniformity and poor device reproducibility and increase film defects. The demand for effective, scalable techniques has not yet been adequately met. The efficacy of devices fabricated using these procedures is still subpar compared to those manufactured using spin coating. Some of these approaches can only be utilized for perovskite film deposition. Conversely, spin-coating strategies are still needed for the deposition of further layers in devices (such as electron transport layers (ETLs) and hole transport layers (HTLs)) [34,35]. Nevertheless, some of these approaches still employ antisolvent procedures; gas-assisted strategies, for instance, were designed to produce perovskite devices without spin-coating and antisolvent procedures. However, the device’s performance (e.g., efficiency, hysteresis, and stability) requires significant research and advancement. In addition, the scalability of these technologies for the efficient manufacturing of large-area devices has not been established. In short, it is still challenging to construct highly scalable, spin-coating- and antisolvent-free perovskite solar cells with high efficiency.

The doctor blade, with nitrogen gas blown via an air knife with varying input pressure, which can create a uniform sheet of laminar airflow, is a commonly used type of industrial equipment employed to coat thin films, eradicate extraneous particles/liquids, and evaporate liquid coatings. Typically, the doctor-blading approach for thin film coating uses a steady airflow and linearly moving airflow to apply the solution onto the substrates that are being coated [36,37,38,39]. In contrast to spin coating, the continuous airflow with a linear trajectory induces consistent thrust for thin film formation; consequently, the doctor-blading approach does not lead to homogeneity concerns for the thin films. Moreover, doctor blading can result in the swift removal of the reagent from the moist films, which may contribute to the formation of films. In light of these benefits, doctor blading is a viable technology for producing perovskite photovoltaic cells without spin-coating and antisolvent procedures. In particular, the entire mechanism for the doctor-blading method was built in industry for mass production. Because of this, the doctor-blading technique in the perovskite area can offer perovskite technology significant commercialization potential. Doctor blading is also a scalable process, making it a promising solution for the commercial production of perovskite solar cells. In conclusion, doctor blading is a promising alternative to spin coating for perovskite solar cell fabrication, offering an improved performance and scalability.

In this study, we propose a doctor-blading technique for depositing thin films with a tuneable thickness and assembling whole perovskite solar cells without spin coating. All of the films in the device’s construction, including the ETLs, perovskite films, and carbon electrodes, are fabricated using a simple doctor-blading and screen-printing approach. Notably, the doctor-blading technique for perovskite film deposition can be used to coat the precursor solution and vaporise the water-soluble solvent from the perovskite precursor solution, thereby producing a dense, uniform/compact perovskite film. The resultant MAPbI₃ solar cells with low-temperature-printed carbon electrodes demonstrate a PCE of up to 14.17%. In addition, the doctor-blading method’s mechanism for producing high-quality and reproducible perovskite films is thoroughly investigated [40,41]. Regarding the solar cell’s characteristics, we conducted a comprehensive comparison (Table S1) of our findings with those reported in the literature using the same absorbing layer (MAPbI₃). These findings imply that the doctor-blading method is a viable, reproducible technique for thin film deposition and device creation, offering numerous opportunities for real-world applications of perovskite technology.

2. Experimental Section

2.1. Materials

All the chemicals and solvents were employed as obtained. Tin (IV) oxide (SnO₂, 15 wt$\%$ in H₂O colloidal dispersion) and acetonitrile were attained from Alfa Aesar. The Sinopharm Chemical Reagent Corporation, Ltd. (Shanghai, China) supplied a 30% solution of methylamine in ethanol. The methylammonium lead iodide (MAPbI₃, 99.9%) was from Xi’an p-OLED. Shanghai MaterWin New Materials CO., LTD. (Shanghai, China) supplied the conductive carbon material.

2.2. Formation of SnO₂ ETLs

A total of 1 g of the initial SnO₂ dispersion was diluted alongside 5 g of DI water in order to formulate the SnO₂ as a precursor, which was then sonicated for an hour and filtered through a 0.45 m polytetrafluoroethylene (TPFE) filter. The thin SnO₂ layer was formed on the UV-treated ITO substrate by doctor blading 50 µL of SnO₂ dispersion in DI water at 50 mms⁻¹ with an input N₂ flow pressure of 0.4 MPa, after which it underwent calcination at 150 °C for 30 min an ambient environment.

2.3. Preparation of Perovskite Films with Doctor Blading and Fabrication of Devices

Our prior work discussed the perovskite precursor method [42], and Figure 1 demonstrates film formation with doctor blading. A MAPbI₃ precursor solution (30 µL) was deposited on surfaces of 2 cm × 2 cm, and a doctor blade was used to coat the precursor. The dimensions of the length and girth of the air knife (N₂ knife) were 1 mm and 2.5 cm, respectively, and the blade coater length was 10 cm. The air-knife-to-substrate angle was kept constant at 55°, and the doctor blade moved at a steady rate of 50 mms⁻¹. The pressure of the N₂ flow supply fluctuated between 0.2 and 0.6 MPa. The coated films were subsequently annealed at 130 °C for 20 min on a heated plate in order to produce perovskite films. In fact, our investigations demonstrated that 25 µL of precursor solution was sufficient to generate perovskite films on a 2 cm × 2 cm surface with comparable device functionality using the doctor-blading method. Optimizing the supplied N₂ flow pressure, the framework of the doctor blade, and the distance between the air knife and the substrate surface, a better material productivity should be feasible for the doctor-blading mechanism. The perovskite films were wholly manufactured in an ambient environment with a relative humidity of ~50%. Carbon pastes were then screen-printed onto the perovskite layer, and the materials were baked for 20 min at 115 °C.

2.4. Characterizations

The morphological and crystallographic properties of the films were observed using a scanning electron microscope (JSM-6307, JEOL Inc., Tokyo, Japan) and an X-ray diffractometer (SMART LAB, Rigaku, Tokyo, Japan), respectively. The UV–Vis transmission spectra of the films were measured with a JASCO V-570 UV/VIS/NIR Spectrometer. Using a fluorescence spectrometer (FLS980, Edinburgh Instruments, Livingston, UK), the photoluminescence (PL) and time-resolved photoluminescence (TRPL) spectra were acquired. To measure the J–V curves of the manufactured devices, a PV-IV-201V I-V Station (Newport Oriel, Irvine, CA, USA) was used. Previously, the source of illumination was rectified using a Newport 91150 V reference cell system. The J–V contours were scanned at 0.1 Vs⁻¹, either from −0.1 V to 1.2 V or in the opposite direction. A Qtest Station 1000ADX (Growntech, Inc., Brentwood, Tennessee) was used to measure the external quantum efficiency (EQE) spectra of the devices. Electrochemical impedance spectroscopy (EIS) was performed using a Chenhua, Shanghai CHI660E electrochemical workstation.

3. Results and Discussion

Figure 2a demonstrates that the XRD patterns of the MAPbI₃ films of varying thickness are identical, suggesting that N₂ pressure does affect the structural properties of doctor-bladed perovskite films. All the samples show a pure phase of MAPbI₃ with characteristic diffraction peaks at 14.2°, 28.5°, and 31.8°, which can be indexed to the (110), (220), and (222) planes of the MAPbI₃, respectively. The mean crystal size was calculated using the Scherrer equation, which relates the FWHM of the XRD peaks to the crystal size. Using this equation, we calculated the high-intensity peak (110) for various N₂ pressures of 0.2, 0.3, 0.4, 0.5, and 0.6 MPa and crystallite sizes of 694.41, 689.63, 656.69, 648.92, and 648.22 nm, respectively. The diffraction peak intensity decreases with the increase in N₂ pressure, which should be ascribed to the decrease in film thickness, as demonstrated in Figure 2d. The optical absorption spectra in Figure 2c indicate that the MAPbI₃ films have the same absorption onset. Nonetheless, the light absorption beyond the bandgap gradually decreases with reduced thickness. Figure 2d shows the MAPbI₃ film thickness as a function of N₂ flow pressure, indicating that the film thickness is inversely proportional to the N₂ pressure. The corresponding cross-sectional SEM images are depicted in Figure 3.

$$\mathrm{D}=\frac{\mathrm{k}\mathsf{\lambda }}{\mathsf{\beta }\cos \mathsf{\theta }}$$

(1)

The top-view scanning electron microscopy (SEM) images in Figure 3a–e portray the morphology of perovskite films (on glass substrates) prepared via doctor blading, from which it can be observed that the grain size of the doctor-bladed films is larger than that of the conventional-method-produced film (which is consistent with the literature [43]). Increasing the N₂ pressure during blade coating can result in more uniform film formation and enhanced surface coverage, which can facilitate the growth of larger grains and improve the overall crystallinity of the material. On the other hand, decreasing the N₂ pressure can reduce the spread of the solution and result in a more porous and less compact, thicker film with a smaller grain size and reduced crystallinity. For the sake of simplicity, we solely examined the thickness control of 2 cm × 2 cm substrates. During the doctor-blading process, we regulated the pressure of the input N₂ flow to customize the thickness of the MAPbI₃ films. The thickness of the MAPbI₃ film was inversely proportional to the N₂ flow pressure. Figure 3a–e illustrates the SEM images of the respective top views and cross-sections. Since thickness can significantly affect film quality, perovskite films of varying thicknesses were evaluated to determine the ideal thickness for PV devices. The MAPbI₃ films showed similar structural properties after tuning their thickness, while the optical and optoelectronic properties varied with different thicknesses.

Steady-state photoluminescence (PL) spectra were obtained to reveal the quality of the perovskite films deposited on quartz with varying input N₂ pressure, with the subsequent outcomes shown in Figure 4a. The photoluminescence (PL) intensity of perovskite films can vary with the N₂ pressure used during deposition, mainly when blade coating is employed. This is because the N₂ pressure can impact the growth conditions and the film quality, affecting the PL performance. On the one hand, high N₂ pressure can result in denser and more compact films, leading to a better PL performance. However, if the pressure is too high, it can also cause defects or inhomogeneities in the film, which can degrade the PL. On the other hand, low N₂ pressure can lead to less compact and more porous films, resulting in reduced PL intensity. Therefore, identifying the optimal N₂ pressure that balances the trade-off between film quality and PL performance can be challenging, and the ideal pressure can vary depending on the specific blade-coating conditions. In doctor blading, the most vigorous PL emission arises at a sample input N₂ pressure of 0.5 MPa. Figure 4b shows the TRPL measurements, which provide additional evidence for the PL spectrum. The PL decay curves clearly exhibit biexponential decay behaviour, which can be modelled using the following biexponential equation [44]:

$$\mathrm{y}={\mathrm{y}}_0+{\mathrm{A}}_1\exp \left.\left(-\frac{\mathrm{t}}{{\mathsf{\tau }}_1}\right.\right)+{\mathrm{A}}_2\exp \left.\left(-\frac{\mathrm{t}}{{\mathsf{\tau }}_2}\right.\right)$$

(2)

where A₁ and A₂ represent the decay amplitudes, τ₁ and τ₂ represent the carrier lifetime, and y₀ represents an offset constant. The constants for trap-mediated nonradiative recombination and free carrier recombination from the bulk perovskite are the fast decay component τ₁ and slow decay component τ₂ [45,46]. The detailed results of the calibration are listed in

Table 1. A concise overview of the fitted parameters for the TRPL of perovskite films on bare glass substrates with varying N₂ pressure inputs.

Samples (MPa)	τ1 (ns)	Fraction 1 (%)	τ2 (ns)	Fraction 2 (%)	τavg (ns)
0.2	37.85	21.73	116.51	78.57	91.54
0.3	25.86	7.76	125.15	92.24	117.45
0.4	28.11	7.80	128.01	92.20	120.21
0.5	34.55	5.20	157.68	94.80	151.28
0.6	32.31	6.32	136.94	93.68	130.31

. Using the equation below, the average carrier lifetime (τ_avg) can be calculated [47]:

$${\mathsf{\tau }}_{\mathrm{avg}}=\frac{\sum {\mathrm{A}}_{\mathrm{i}}{\mathsf{\tau }}_{\mathrm{i}}^2}{\sum {\mathrm{A}}_{\mathrm{i}}{\mathsf{\tau }}_{\mathrm{i}}}$$

(3)

The average carrier lifetimes for the doctor-bladed films created with varying input N₂ pressures (0.2, 0.3, 0.4, 0.5, and 0.6 MPa) are 91.54, 117.45, 120.21, 151.28, and 130.31 ns, respectively. The 0.5 MPa film’s carrier lifetime is longer than that of the other films, which can be attributed to its slower recombination rate and lower defect density than the other doctor-bladed films.

To fabricate fully printed perovskite solar cells, we tried to deposit SnO₂ layers using the doctor-blading method and compared them with the spin-coated SnO₂ layers. The top-view and cross-sectional SEM images in Figure S1 (see Supplementary Materials) illustrate the morphology of the SnO₂ films (on the glass/ITO substrates) fabricated via spin-coating and doctor-blading procedures, respectively. It is clear that the doctor-bladed SnO₂ film is thicker than the spin-coated SnO₂ film. The transmission spectra in Figure S2 demonstrate that the spin-coated SnO₂ film is comparable with the doctor-bladed SnO₂ film. To investigate the effect of varying N₂ pressure on cell performance, as shown in Figure 5a, we manufactured spin-coating-free PSCs with ITO/SnO₂/perovskite utilizing methylamine-treated MAPbI₃ as the light absorber. Figure 5b depicts a cross-sectional SEM image of a solar cell constructed with a MAPbI₃ film at an N₂ pressure input of 0.5 MPa. The overall device is highly consistent in appearance, and the ITO layer and perovskite film are around 190 and 315 nm thick, respectively. The SnO₂ ETL is too thin to be discernible in the SEM image. Figure 5c demonstrates the J–V curves for the best-performing carbon devices with various perovskite films, whereas

Table 2. Average photovoltaic parameters for the best-performing PSC devices manufactured with different N₂ pressures.

Samples (MPa)	Voc (V)	Jsc (mA cm−2)	FF	PCE (%)
0.2	0.99	18.71	0.52	9.65
0.3	1.01	19.40	0.56	11.02
0.4	1.02	20.50	0.54	11.38
0.5	1.01	22.77	0.61	14.17
0.6	1.00	21.90	0.61	13.47

summarises the full photovoltaic parameters. The device deposited with a 0.5 MPa N₂ pressure yielded the highest PCE, with a PCE of 14.17%, V_oc of 1.01 V, J_sc of 22.77 mA cm⁻², and FF of 0.61.

Figure 5d conveys the external quantum efficiency (EQE) spectra for the devices, and the integral current densities for the fluctuating input N₂ pressure (0.2, 0.3, 0.4, 0.5, and 0.6 MPa) devices are 16.65 mA cm⁻², 17.33 mA cm⁻², 18.44 mA cm⁻², 20.97 mA cm⁻², and 19.20 mA cm⁻², respectively, which are comparable to those obtained from the J–V curves. We also attempted to create bigger devices, and Figure S3 depicts an example of a typical carbon-printed device pattern (over a square-inch substrate). This finding is consistent across multiple samples, indicating that the devices created with doctor blading are uniform. Figure S3 shows the related J–V curves and reliable PCE output; a stabilized output PCE 8.38% was attained. As demonstrated previously, the lower PCE with a greater cell size is primarily due to the reduced fill factor (FF) associated with higher series resistance.

We further performed electrochemical impedance measurements under light illumination to investigate the impact of varying input N₂ pressure on the charge carrier behaviour of the devices under working conditions. Figure 6 shows the Nyquist plots with different N₂ pressures under 1 sun illumination without bias at a minor voltage fluctuation (5 mV). By fitting the Nyquist plots to the equivalent circuit depicted in Figure 6, the charge transfer resistance of the devices at a given voltage could be determined. The curves are separated into two regions, with a high-frequency-area semicircle representing the charge transfer resistance (R_ct) and a low-frequency-zone incomplete semicircle representing the recombination resistance (R_rec), and the resulting values are displayed in

Table 3. Fitting parameters for EIS of PSC devices based on doctor blading with varying input N₂ pressure.

Samples (MPa)	Rs/Ω	Rct/Ω	Rrec/Ω
0.2	67.08	129.8	97.3
0.3	66.91	138.0	189.2
0.4	56.67	148.5	252.5
0.5	44.10	102.5	345.7
0.6	45.62	133.8	306.8

. These results verify an improved charge collection at a 0.5 MPa N₂ pressure, consistent with the TRPL and stabilized output results. However, the 0.5 MPa sample, in the case of the N₂ pressure input device, demonstrates the most substantial recombination resistance among all the samples, indicating efficient carrier extraction and less recombination [47], which enriches the device’s performance (FF) [44,47].

All types of devices were constructed in many batches in order to further test for repeatability, and Figure 7a–d depicts the statistical results for the photovoltaic parameters. The device’s performance characteristics are constrained to an equitably slight range, and the device with an 0.5 MPa N₂ pressure shows some variation among all the parameters.

4. Conclusions

In this study, we made a monumental breakthrough in the realm of doctor-blading techniques for the deposition of perovskite films and ETLs and the fabrication of entirely air-printed perovskite photovoltaic devices. We demonstrated a strategy for depositing MAPbI₃ films in an ambient environment with doctor blading that is entirely free of spin coating and has precise morphological control. Consequently, fully printed perovskite solar cells with the structure of ITO/SnO₂/MAPbI₃/carbon were fabricated using a doctor-bladed SnO₂ electron transport layer and a screen-printed carbon counter electrode. The device with an input N₂ pressure of 0.5 MPa obtained a superior PCE of 14.17%, whereas the large-area device, with an active area of 1 cm², yielded a stabilised PCE of 8.38%. This work addressed ambient-processed, fully spin-coating-free PSCs with the merits of industrialization, scalability, and stability. This research could have significant implications for the development of efficient and cost-effective perovskite solar cells.