Defect passivation in perovskite films by p-methoxy phenylacetonitrile for improved device efficiency and stability

The certification efficiency of halide perovskite solar cell is as high as 25.7%, which is one of the most efficient solar cells at present. However, the defects in the halide perovskite including grain boundary, interface defects, and ionic defects often act as nonradiative composite sites, which lead to rapid degradation of halide perovskite films, deteriorate the performance of perovskite devices, and lead to instability. In this work, a suitable multifunctional molecule additive p-methoxy phenylacetonitrile (pMP) is selected to improve the film and device stability. Specifically, pMP delays the crystallization rate of halide perovskite and promotes the formation of high-quality large grain halide perovskite films, and C≡N in pMP forms a coordination bond with Pb2+ and passivates the uncoordinated Pb2+ defects. Moreover, the π bonds increase electron transport. In addition, the methoxy group in pMP forms an effective barrier on halide perovskite to enhance its water stability. With the influence of the comprehensive effect of these factors of pMP, the PSC with pMP additive achieved the highest efficiency of 21.26% and significantly improved the stability of moisture resistance.


Introduction
In recent years, the energy field has received increasing attention, especially new energy and new materials [1][2][3][4].Organic-inorganic hybrid perovskite solar cells (PSCs) fabricated by the solution process have experienced rapid development, and power conversion efficiency (PCE) has soared from 3.8 to 25.7% in 10 years, which shows the most promising commercial application potential among the 3 rd generation thin film solar cells [5,6].The high photovoltaic efficiency is mainly due to its good photoelectric properties, including excellent carrier mobility [7,8], high light absorption coefficient [9], small exciton binding energy [6], and adjustable band gap [10].Despite these excellent properties and the rapid developments of halide perovskite materials, the low temperature solution treatment and fast crystal growth of perovskite films inevitably introduce various of crystal defects.These defects mainly exist at the grain boundary and the interface of halide perovskite with electrons or holes extraction layer.The massive defects (such as 155 Page 2 of 8 impurity atoms, ion vacancy, interstitial ion, and undercoordinated ions), which causes ion migration and defect assisted non-radiative carrier recombination, lead to abnormal J-V hysteresis and accelerate water/oxygen penetration.Thus, the stability and PCE of devices would be reduced.In this regard, an effective strategy must be found to passivate or eliminate these various defects in order to further improve the photovoltaic properties and stability of cells.
Adding additives to precursors is an effective and common method to optimize the crystallization process of halide perovskite, passivate grain boundaries and/or surface defects, promote interfacial charge diffusion, and improve the operational stability of PSCs.The Lewis base or Lewis acid usually contains some functional groups, amino, amide, isopropyl, carbonyl, phenylethyl, aminocyano, etc., which can effectively passivate defects in halide perovskite by coordinating or eliminating dangling bonds, so as to significantly increase carrier lifetime and enhance the photoelectric performance of perovskite devices [11][12][13][14][15][16][17][18].Thus, the passivation strategy based on Lewis acid/base is certainly worthy of further study [19].However, most of these passivation molecules are insulating materials, which hinder the separation and transport of carriers [20][21][22][23].In addition, some selected passivating molecules usually have only one functional group, which can only passivate a single trap state and lack hydrophobicity.In order to solve these problems, it is very meaningful to develop new functional molecules, which can combine many types of traps, and has certain carrier conductivity and good moisture resistance [24][25][26].Recently, Mai et al. introduced D-π-A-type porphyrin derivatives in PSCs, that can effectively passivate the perovskite surface, increase V OC and FF, reduce the hysteresis effect, and achieve best-performing PCEs of 22.37% [27].
Here, pMP was selected as a functional additive aiming to improve the electrical or physical properties of halide perovskite, such as a large number of traps, water sensitivity, and ion migration.Specifically, C≡N in pMP forms coordination bond with Pb 2+ , passivating the under-coordinated Pb 2+ defects.Therefore, not only crystallization rate of halide perovskite is reduced, but also high-performance halide perovskite films with fewer traps can be prepared.The C≡N group, methomyl group, and benzene ring form D-π-A structure to promote electron transport.In addition, the methoxy group in pMP forms an effective barrier on halide perovskite to enhance its water stability.The theory calculation, FTIR, and PL/TRPL analysis confirmed the protection and passivation effect of pMP additive.With the help of the comprehensive action of multi-functional groups in pMP molecules, the best PSC with pMP additives reached a PCE of 21.26% and significantly enhanced the stability against humidity.And the unpackaged PSC containing pMP additive has better environmental stability and retains 87% of the initial PCE after exposure in the air for 1000 h.

Results and discussion
As shown in Fig. 1a, the pMP consists of a carbonyl group, a benzene ring and a methyl group.To visualize electron cloud distribution, the electrostatic potential of pMP was analyzed by DFT method (Fig. 1a).The benzene ring has delocalized π electrons.While the strong electron withdraw cyano group shows high electron density in red region.The methoxy group is an electron-donating group.Thus, a D-π-A system is formed, which can coordinate or bind with uncoordinated Pb 2+ in halide perovskite.In addition, the methoxy group in pMP leads to strong hydrophobicity, which is very important to enhance the moisture resistance of halide perovskite in air environmental conditions.
In this work, mixed cation halide perovskite thin film was used as a photoactive layer.The additive molecules are directly mixed in the halide perovskite precursor solution, and the film is prepared by the one-step antisolvent method, which ensures the good distribution of pMP molecules on the grain boundary and the surface of polycrystalline film.Through the interaction with halide perovskite, pMP delayed the perovskite crystallization process, and induced the formation of high-quality large-grain halide perovskite films (see Fig. 1b).
To further clarify the effect of pMP concentration on halide perovskite layers, 0.1 ~ 0.6 mM pMP was added to the halide perovskite precursor.Top-view SEM images of pMP added halide perovskite are shown in Fig. 2a, and the corresponding halide perovskite grain size distribution is summarized in Fig. 2b.When the concentration of pMP increased from 0 to 0.4 mM, the grain size increased, while the grain size decreased when the concentration further increased to 0.6 mM.The average halide perovskite grain sizes for control and 0.4 mM pMP were 560 and 760 nm, respectively.The increase in grain size is attributed to interaction between PbI 2 and pMP.As reported earlier, the formation of PbI 2 -pMP adducts helps to increase the activation energy and reduce the nucleation number [28][29][30].Film surface wettability test was carried out to evaluate water resistance of halide perovskite layers (The water contact angle test results are shown in Fig. 2c).Water contact angle increases with the increase of pMP concentration.The increase of the contact angle can be attributed to the exposure of the hydrophobic methoxy group of pMP on halide perovskite surface, which can help to improve moisture resistance of halide perovskite and slow down its degradation in the air [31,32].The photoelectric performance of complete devices with a series of different pMP concentration was tested.Compared with control solar cell, the device efficiency of 0.4 mM pMP increased significantly to 21.26% (Fig. S1).To confirm the effect of pMP on films or devices, the performance of the control and 0.4-pMP samples (labeled as pMP) was analyzed and discussed emphatically in the following part.
FTIR was measured to evaluate interaction between halide perovskite and pMP molecule.In Fig. 3a, the peaks at 820, 2925, and 2248 cm −1 in pure pMP belonged to phenyl ring, methoxy and C≡N.It can be seen from the high-resolution spectrum, pMP-added halide perovskite also shows these characteristic peaks, indicating that pMP molecules were successfully added to halide perovskite.The characteristic peak of C≡N shifts towards about 2241 cm −1 , confirming that pMP interacted with halide perovskite.Because C≡N group locate at the strong negative electrostatic region, the peak shift is likely caused by electrostatic coupling of C≡N and uncoordinated Pb 2+ .[33] In addition, we further tested the mix sample of pure pMP and PbI 2 to further observer the change of pMP characteristic peak caused by PbI 2 .In Fig. S2, the C≡N stretching vibration mode locates at 2248 cm −1 for pure pMP and it shifted to 2242 cm −1 after mixing with PbI 2 , which further confirming the electrostatic coupling interaction between Pb 2+ and pMP molecules.The XPS results show that the introduction of pMP triggers a shift of Pb4f toward lower binding energy, indicating that an interaction between pMP and Pb 2+ is indeed present in the perovskite film (Fig. S3).
To further explore the effects of pMP on lattice structure and halide perovskite crystallinity, XRD spectra were obtained for control halide perovskite and pMP added halide perovskite.Figure 3b shows the XRD spectra of control and pMP halide perovskite.Usually, heteromorphic atoms or molecules entering halide perovskite lattice can cause lattice distortion and lead to peak shift.From the Fig. 3b, it can be seen that all peaks of the two films are in the same position, while the intensity of the pMP perovskite XRD peak was enhanced, indicating an improvement in the crystallization without the observable lattice parameter changed.In addition, trace PbI 2 was observed at 12.7° diffraction peak.It is worth noting that comparing with control sample, the diffraction peak intensity of halide perovskite film added with pMP is significantly increased, while the intensity of PbI 2 diffraction peak is reduced.This shows that the interaction between pMP and uncoordinated Pb 2+ reduces the concentration of Pb 2+ and improves the crystallinity of halide perovskite films.
Effects of pMP molecules on the optical properties for halide perovskite further explored by UV-vis absorption spectra, PL, and TRPL spectra.In Fig. 3c, the enhanced UV-vis absorption of pMP halide perovskite film shows that the pMP additive improves the optical properties of the film, which is consistent with the previously mentioned large grain halide perovskite with good crystallinity.It is also shown that absorption band edge of the sample remains the same due to the introduction of pMP, which indicates that there is change of band gap for the halide perovskite.Further evidence can be found in the Tauc plot in Fig. S4.The results show that pMP molecules do not enter halide perovskite lattice, which is consistent well with XRD results.Analogously, the PL intensity of pMP halide perovskite increases significantly compared with control film (Fig. 3d), which can be attributed to the reduction in the trap density and enhanced crystallinity of halide perovskite with pMP additive.TRPL tests were carried out to further understand pMP molecules effect to the halide perovskite films on carrier recombination.The TRPL results (Fig. 3e) show double exponential decay with fast and slow decay components.The function ( 1) was used to fit the TRPL spectrum [34,35].
where A is attenuation amplitude and τ is decay lifetime (τ 1 fast decay, τ 2 slow decay).TRPL decay can be attributed to defect assisted recombination at grain boundary or interface.The corresponding fitting parameters of the control and pMP samples are listed in Table S1.The fitting results show that halide perovskite films with pMP additive have longer photoexcited carrier lifetime (55.76 ns) than that of control (35.61 ns), which shows the traps of the halide perovskite films decreased after pMP addition.The results indicate that the addition of pMP molecules can effectively passivate lead ion traps, prolong decay lifetime of charges in perovskite layers and enhance perovskite crystallization.This is consistent with grain changes observed in SEM images: halide perovskite with added pMP has big grains and relatively few defects or grain boundaries.
To study the trap density in control and pMP perovskite films, SCLC tests were performed under dark conditions on electronic-only devices (ITO/SnO 2 /Perovskite/PCBM/Ag).The results are shown in Fig. 3f-g.The J-V characteristic curves of electronic-only transmission devices include 3 regions: ohmic region (low), trap filling limit region (TFL, intermediate), trap free SCLC region (high).As voltage increases, traps are gradually filled until V TFL is reached.The V TEL value of control device is 0.56 V, while the V TEL value of the pMP device is 0.48 V. Trap density (n t ) was calculated by Eq. ( 2) [36,37].
where e, ε 0 , ε, and L are the basic charge, vacuum dielectric constant, relative dielectric constant, and film thickness.According to the calculation, the n t value of halide perovskite with added pMP is 4.85 × 10 15 cm −3 while the n t of the control halide perovskite is 4.16 × 10 15 cm −3 .The reduction of defects, on one hand, is due to the electrostatic coupling between the C≡N group in pMP molecule and Pb 2+ in halide ( 1) perovskite to passivate Pb 2+ traps.On the other hand, pMP molecule reduced grain boundary defects by increasing the halide perovskite grains size.
To evaluate the effect of pMP molecules on the photovoltaic property and stability of the solar cells, planar structure PSCs were fabricated (ITO/ETL/Perovskite/HTL/Ag). Figure 4a, b show a structure diagram of solar cell and the cross-sectional SEM image.From the cross-sectional image, it shows that halide perovskite thickness is about 900 nm, and big grain is in the whole film.It should be pointed out that the introduction of pMP did not cause the change of halide perovskite thickness due to the small amount of additives.Figure S5 is the cross-sectional images of control and pMP devices.The best photoelectric performance of the control and the pMP cells were compared in Fig. 4c.The champion devices of pMP PSC show excellent performance with an V OC of 1.17 V, a J SC of 23.10 mA cm −2 , a FF of 78.54%, and overall efficiency of 21.26%, while the control solar cell has an V OC of 1.10 V, a J SC of 21.33 mA cm −2 , and a FF of 77.24%, and the efficiency is 18.18%.In addition, it can be seen from reverse/forward J-V curves that pMP device has smaller hysteresis.The hysteresis index (HI = [(PCE R -PCE F )/PCE R ]*100%) are 7.5% and 5.1% respectively, indicating that the pMP additive can effectively inhibit the hysteresis of the solar cell.The mMP devices exhibited considerable reproducibility with an average efficiency of 20.14% in a count of 20 devices, whereas the control showed an average efficiency of 17.28%, as illustrated in Fig. S6. Figure 4d shows corresponding IPCE of control and pMP champion cells, which also demonstrates the good performance of pMP devices.The integrated J SC of control and pMP cells are 22.43 and 20.73 mA cm −2 , which is in accordance well with J-V results.The enhanced performance and inhibited hysteresis should be attributed to the electrostatic coupling between C≡N and incoordination Pb 2+ traps.On one hand, it promotes the formation of large grains and reduced grain boundaries defects.On the other hand, C≡N groups and benzene rings form D-π-A transport channels to accelerate carrier transport and reduce carrier accumulation at grain boundaries or interfaces.Compared with control cells, pMP devices also show a lower dark current (Fig. 4e).
The interfacial carrier separation, transfer, and recombination dynamics process were further studied by EIS (see Fig. 4f).Nyquist plots show that solar cell with pMP additive has a larger recombination resistance (R rec = 1274.4Ω) than the control device (R rec = 696.5 Ω), indicating the enhanced carrier transfer and reduced carrier recombination in pMP solar cell (Table S2).To reveal the carrier recombination characteristics in solar cells, J-V results under a series of irradiation intensities were calculated [38,39].
(3) V OC = nkT∕q ln(I) where q, T, n, and k are elementary charge, absolute temperature, ideal factor, and Boltzmann constant, respectively.The J sc in well consistent with power law on illumination intensity (Fig. 4g), while the fitting slope of the light intensity-V oc curves decreased from 1.58 to 1.39 with the adding of pMP, indicating the effect of Shockley Read-Hall recombination (see Fig. 4h).
We also explored TPV decay under V OC condition.The carrier recombination lifetime for pMP devices is significantly longer than that of control, shown in Fig. 4i.Upon illumination condition under V OC , holes and electrons are generated and recombine in the halide perovskite layer without going to outward circuit.The separated carriers recombine again when illumination stops, causing photovoltage rapid decay.From TRPL results discussed previously, halide perovskite films with and without pMP additive have a similar PL lifetimes and optoelectronic quality.Thus, the enhanced carrier recombination lifetime from TPV decay for pMP device is attributed to less carrier recombination, which mainly related to faster charge extraction reduces carrier accumulation at grain boundary or interfaces.
As described above, the pMP can improve halide perovskite quality and also can significantly inhibit reactivity between perovskite films and water.So, pMP additive is hoped to enhance long-term stability of perovskite devices, especially in high humidity condition.Therefore, the efficiency changes of unpackaged cells in air were examined (relative humidity ~ 50%, temperature ~ 25 °C).In Fig. 5a, PCE of control cell decreased 40% within 1000 h.In contrast, pMP cell can maintain an initial efficiency of about 87% even after 1000 h of storage.The XRD of pMP film (see Fig. 5b) in the air for 2 weeks has almost no change.In addition, the UV-Vis absorption intensity of perovskite films, as shown in Fig. S7, decreases with time.Absorption intensity of pMP perovskite has little change after 7 days, which means that the degradation process of the film is effectively inhibited.After 15 days, the light absorption intensity of the two films both decreased, but the decrease was more obvious in the control.These results indicate that suitable additives can enhance device stability in worse environments.

Conclusion
In conclusion, we propose a strategy in halide perovskite defect regulate and control by adding small organic molecules which containing cyano units as end groups in precursor to prepare high-quality halide perovskite with big grain.We also found that pMP molecules modify grain boundaries and passivate defects by forming coordinate bond with uncoordinated Pb 2+ in halide perovskite films.Importantly, interaction between C≡N and undercoordinated Pb 2+ helps the carrier separation and transport at grain boundary or interfaces.Therefore, the PCE of the champion solar cell with adding pMP increases significantly up to 21.26%, and the unpackaged device stability at room temperature also improved.This work provides a useful strategy for reasonably design molecules additives, passivate defects, and enhance carrier extraction/transport to get high-quality perovskite films and high-performance perovskite devices.

Fig. 1 a
Fig. 1 a Schematic diagram of pMP molecular structure, surface electrostatic potential, and its interaction with halide perovskite.b Preparation process of halide perovskite films

Fig. 3 a
Fig. 3 a Fourier transform infrared spectra of pure pMP, perovskite (labeled as control) and pMP-additive perovskite (labeled as pMP).b XRD, c UV-vis spectra, d PL, and e TRPL spectra of control and pMP perovskite films.SCLC versus voltage of f control and g pMP-based devices

Fig. 4 a
Fig. 4 a Structure diagram of solar cell.b complete device cross-sectional SEM photo.c J-V curves, d IPCE spectrum, e dark current density, f EIS results, g J SC and h V OC versus light-intensity of control and pMP cells, i TPV for the control and pMP cells

Fig. 5 a
Fig. 5 a Long-term stability of unpacked devices, b XRD patterns of pMP perovskite film