Buried Interface Modification Using Diammonium Ligand Enhances Mechanical Durability of Flexible Perovskite Solar Cells

Abstract

Flexible perovskite solar cells (F-PSCs) hold great potential for lightweight photovoltaic applications due to their flexibility, bending compatibility, and low manufacturing cost. However, tin oxide (SnO₂), as a common electron transport layer (ETL) used in F-PSCs, typically suffers from high-density surface defects that hinder the charge extraction efficiency and deteriorate the crystallization quality of the upper perovskite film. Additionally, the poor buried interface quality intensifies lattice extrusion and strain residue across the perovskite films, further aggravating the mechanical brittleness in devices. To address the issues, we developed a molecular bridging strategy by introducing the 2,2′-oxybis(ethylenediamine) dihydrochloride (DO) at the perovskite/SnO₂ interface. The diammonium groups of spacer ligands can achieve the bidentate anchoring on the SnO₂ and perovskite films, cooperating with the oxygen atom on the alkyl chain to passivate the charged defects at the buried interface. The tailored interface properties also endow the optimized crystallization quality of perovskite films and significantly alleviate tensile strain to strengthen the perovskite’s pliability. As a result, the F-PSCs achieved a champion efficiency of 23.50%, outperforming the value of 21.87% for the control device. Furthermore, the devices exhibited excellent mechanical robustness, maintaining 90% of the initial PCE after 6000 bending cycles at a radius of 4 mm. This work presents a reliable strategy for the synergistic optimization of the buried contact at the electron extraction interface, contributing to the further development of efficient and stable F-PSCs.

1. Introduction

Perovskite solar cells (PSCs) have gained tremendous attention due to their exceptional optoelectronic properties and facile solution processability [1,2]. The inherent soft lattice structure of metal halide perovskite materials facilitates the ongoing development of flexible perovskite solar cells (F-PSCs) [3,4]. This category of devices occupies notable advantages, including light weight, mechanical flexibility, and bending durability, thereby conferring its significant application potential in wearable electronics and aerospace fields [5,6]. The advance of perovskite manufacturing methodologies has promoted the rapid development of F-PSCs, with reported efficiencies exceeding 25% to date [7]. However, the current state-of-the-art F-PSCs still suffer from a remarkable efficiency deficit when compared to their rigid counterparts, primarily attributed to the compromised interfacial contact and deteriorated crystalline quality of upper perovskite films following the substrate variation [8,9]. Additionally, the mechanical bending stability of F-PSCs remains a critical basis that necessitates further enhancement to ensure a reliable operation lifetime in practical applications involving curved settings [10,11,12].

SnO₂ is a commonly employed ETL for high-performance PSCs due to its good optical transmittance, low chemical activities, and compatibility with moderate manufacturing temperatures [13]. Nevertheless, the solution-processed SnO₂ layer generally suffers from severe defect accumulation and colloidal agglomeration within films [14]. The presence of hydroxyl groups on the SnO₂ surface facilitates reactions with lattice oxygen, resulting in the formation of unsaturated Sn dangling bonds that serve as the charge-trapping center [15]. Furthermore, the undercoordinated metal ions and reactive oxygen species can adsorb O₂ and H₂O from the environment, thereby accelerating the degradation of the overlying perovskite layer [16,17]. As such, numerous regulation strategies have been proposed to optimize the physicochemical properties of SnO₂, which, in turn, contribute to suppressing the deep-level defects and improving carrier dynamics at the cathode interface [18,19]. Notably, in the case of F-PSCs, it is also essential to address the critical requirement for mechanical reinforcement to enhance device durability. However, the influence of the buried interface on the mechanical stability of F-PSCs remains inadequately understood, warranting further elucidation. In addition, multifunctional passivators, capable of collectively tailoring the interfacial properties of SnO₂ and strengthening the lattice stability of perovskite, are still in urgent need for the advancement of high-performance flexible photovoltaic cells [20,21,22].

In this study, we proposed a molecular bridge strategy by introducing the functional modifier 2,2′-oxybis(ethylenediamine) dihydrochloride (DO) at the buried interface to regulate the residual strain and strengthen the mechanical robustness of perovskite films. The ammonium groups at both ends of spacer ligands can simultaneously achieve the bidentate anchoring on the SnO₂ and perovskite films, effectively suppressing the trap defects and enhancing the interfacial contact [23]. The electron-donating oxygen atom on the ligand structure can further coordinate with the under-coordinated metal ions at the buried interface and limit the chemical reaction sites. The obtained perovskite films grown on the optimized buried interface exhibited enhanced morphology quality and crystalline orientation, endowing improved charge transport dynamics and non-radiative recombination loss [24]. Furthermore, the spacer ligands accommodated at the buried interface can release the lattice strain of the upper perovskite films to enhance the mechanical strength of F-PSCs. As a result, the DO-treated devices achieved a champion power conversion efficiency (PCE) of 23.50%, outperforming the value of 21.87% for the control device. Simultaneously, the target F-PSCs exhibited enhanced bending stability, retaining 90% of their original efficiency after 6000 bending cycles under a bending radius of 4 mm.

2. Method and Characterization

2.1. Material Preparation

Lead (II) iodide (PbI₂, 99.5%), Formamidine iodide (FAI, 99.5%), methylammonium iodide (MAI, 99.5%), cesium iodide (CsI, 99.5%), and methylammonium chloride (MACl, 98%) were purchased from Advanced Election Technology Co., Ltd., Liaoning, China. 2-Phenylethanamine hydroiodide (PEAI) was purchased from Xi’an Polymer Light Technology Cory, Xi’an, China. N, Ndimethylformamide (DMF, 99.8%), dimethyl sulfoxide (DMSO, 99.9%), anhydrous ether, tin (IV) oxide (SnO₂) colloidal dispersion (15% in H₂O), acetonitrile, lithium salt, Li-bis-(trifluoromethanesulfonyl) imide (Li TFSI, ≥99%), 2,2′,7,7′-tetrakis (N,N-dimethoxyphenylamine)-9,9′-spirobifluorene (Spiro-OMeTAD), and 4-tert-butylpyridine (tBP, ≥96%), were obtained from Alfa Aesar, Ward Hill, MA, USA. Chlorobenzene (CB, 99.8%) and isopropanol (IPA, 99.5%) were purchased from China National Pharmaceutical Group Corporation, Beijing, China.

2.2. Device Fabrication

The flexible polyethylene 2,6-naphthalate (PEN)/ITO substrates were ultrasonically cleaned with detergent, deionized water, and ethanol for 30 min each, and then dried by N₂ blowing and treated with ultraviolet–ozone plasma for 10 min. SnO₂ colloidal dispersion was diluted with deionized H₂O in a volume ratio of 1:3 before use. The SnO₂ precursor was spin-coated onto the PEN/ITO substrates at 5000 rpm for 40 s. Finally, the prepared PEN/ITO/SnO₂ substrate was annealed at 100 °C for 60 min. For DO modification, DO solution (3 mg mL⁻¹ in deionized water) was spin-coated onto SnO₂ ETL at 5000 rpm for 20 s and annealed at 100 °C for 5 min. The perovskite solution was prepared by dissolving 783 mg PbI₂, 269 mg FAI, 18 mg CsI, 11 mg MAI, and 45 mg MACl in a 1 mL mixed solvent of DMF and DMSO (volume ratio, 4:1). For the perovskite deposition process, perovskite solution was spin-coated onto SnO₂ ETL at 6000 rpm for 30 s, followed by annealing at 120 °C for 20 min in ambient air (~40% humidity). PEAI solution (5.0 mg mL⁻¹ in IPA) was spin-coated on the perovskite films, followed by 5 min annealing at 100 °C. Then, the hole transport layer (HTL) was fabricated using spin-coated 45 μL Spiro-OMeTAD solution (90 mg Spiro-OMeTAD, 36 μL TBP, and 22 μL of Li-TFSI in 1 mL CB). Finally, the 80 nm gold electrode was deposited on the HTL.

2.3. Device Characterization

X-ray diffraction (XRD) measurements were executed using a DX-2700BH diffractometer (Haoyuan Instrument, Dandong, China). Grazing incidence X-ray scattering (GIWAXs) pattern at an incidence angle of 0.4° was conducted by Shanghai Synchrotron Radiation Facility at the BL17B1 line station (Shanghai, China). The morphology of perovskite films was assessed by field emission scanning electron microscopy (SEM, Hitachi, SU-8020, Tokyo, Japan) and atomic force microscopy (AFM, Bruker Dimension Icon instrument, AZ, USA). X-ray photoelectron spectroscopy (XPS) was examined by ESCALAB 250, Al Kα, Thermo Fisher Scientific. Young’s modulus was determined through a Dimension ICON SPM system. The grazing incidence X-ray diffraction (GIXRD) tests were achieved by using a high-resolution X-ray Diffractometer (Smartlab (9), Tokyo, Japan). UV-visible absorption was recorded with a UV-visible spectrometer (PerkinElmer UV-Lambda 950, ON, Canada). Steady-state photoluminescence (PL) spectra and time-resolved photoluminescence (TRPL) spectra were achieved by using a PicoQuant FT-300 spectrometer under an excitation wavelength of 510 nm (PicoQuant, Berlin, Germany). The current density–voltage (J–V) characteristics of the devices were gathered by solar simulator equipment (Enlitech, SS-F5, Taiwan, China), and the illumination intensity (100 mW·cm⁻², AM 1.5 G) was calibrated via a reference silicon cell with a KG5 filter. The scan range was from 1.5 to 0 V with a 0.02 V increment and a delay time of 20 ms. The external quantum efficiency (EQE) was quantified by a QE-R system (Enli Technology Co., Ltd., Taiwan, China) using a 300-WXe lamp as the light source. The impedance spectroscopic measurements (EIS) and capacitance–voltage (C–V) measurements were conducted using the electrochemical workstation (IM6ex, Zahner, Karlsruhe, Germany) in dark conditions.

3. Results and Discussion

To resolve the mechanical bending tolerance of F-PSCs, an organic chlorinated diammonium, namely 2,2′-oxybis(ethylenediamine) dihydrochloride (DO), was employed at the perovskite-SnO₂ buried interface to enforce the structural stability of F-PSCs. The DO molecule occupies multiple functional groups in a molecular structure, endowing the bidentate interaction with perovskite and SnO₂ layers. The ammonium groups in DO can eliminate the hydroxyl group accumulated on the SnO₂ surface, thereby inhabiting the deep-level defects and improving the electron transport capability of SnO₂ [25]. The ammonium moiety at the other end of the DO ligand can neutralize the negatively charged defects within the buried interface of the perovskite layer and from the N-H·I hydrogen bond with the Pb-I octahedra framework to modulate the perovskite crystallization [26]. The oxygen atom on the alkyl chain can further support the lone pair electron to form the Lewis acid-based interactions with the undercoordinated metal ions from SnO₂ and perovskite, suppressing the charge-trapping centers. The long alkyl chains of the spacer ligand could act as an insulating barrier, which inhibits the back charge transfer, facilitating efficient charge collection [27]. In addition, the dissociated Cl ions can passivate the oxygen vacancy on the SnO₂ layer and effectively hinder the chemical reaction activity of SnO₂ under moisture conditions (Figure 1a). Therefore, the molecular bridge strategy at the buried interface may enhance the interface adhesion and simultaneously optimize the morphology quality of the upper absorber, contributing to the carrier dynamics and mechanical durability of F-PSCs.

X-ray photoelectron spectroscopy (XPS) measurements were conducted to investigate the chemical coordination between the perovskite and DO ligand. As shown in Figure 1b,c, with the introduction of DO, the Pb 4f signals shifted to the lower binding energies from Pb 4f_5/2 at 143.3 eV and Pb 4f_7/2 at 138.4 eV to 143.2 and 138.3 eV, respectively. This downward shift was associated with the increased electron cloud density around Pb atoms due to the strong coordination between the spacer ligand with Pb-I frameworks [28]. Additionally, the N 1s signal of the DO-SnO₂-based perovskite film shifted to lower binding energy compared to the control film, indicating the formation of hydrogen bonds between the ammonium cation and the perovskite, which enhanced the electronic structure of the perovskite film [29].

Scanning electron microscope (SEM) measurements were then preformed to study the surface morphology and crystalline quality of perovskite films deposited on different SnO₂ substrates (Figure 1d,e). Both films presented compact grain stacking without distinct pinholes on the surface, whereas traces of PbI₂ were retained on the control film. In contrast, the average grain size was increased from approximately 0.7 to 1.1 μm after introducing the DO molecular bridge. The perovskite film using DO-SnO₂ showed a smoother surface topography with the root mean square (RMS) reduced from 29.1 nm to 17.3 nm concerning the control film (Figure S1). The exfoliation of the perovskite films from the substrates revealed that the buried interface of the film on DO-SnO₂ exhibited a more uniform morphology with fewer pinholes compared to the control. The homogeneous grain distribution ensures the strong adhesion between the perovskite and substrate and the mitigated lattice strain across the films. Cross-sectional SEM images illustrated the well-connected grains with a comparable thickness of ca. 600 nm for the different cases (Figure 1f). Notably, the DO-SnO₂-based film presented a decreased density of grain boundaries, which would facilitate the lower defect density and efficient charge transport within the absorber layer [30,31].

The X-ray diffraction (XRD) patterns presented in Figure S2 were analyzed to assess the growth quality of the perovskite layers. The peak observed at 2θ = 14.0° corresponds to the (100) plane of the perovskite. Upon deposition of the perovskite film on the DO-SnO₂ interface, a significant enhancement in the intensity of the (100) diffraction peak was observed, concomitant with a decrease in the full width at half maximum (FWHM). Based on the Scherrer formula, the grain sizes for the perovskite films were calculated to be 0.7 and 1.1 μm for the control and DO-SnO₂ samples, respectively. Additionally, the grazing incidence wide-angle X-ray scattering (GIWAXS) measurements were conducted to further elucidate the crystalline properties of perovskite films (Figure 1g). In the control film, a diffraction signal of PbI₂ was detected at q = 0.9 Å⁻¹, which was suppressed in the DO-SnO₂-based film, suggesting the enhanced film stability under the ambient testing condition. In addition, compared to the control perovskite, the DO-SnO₂ case showed a stronger diffraction signal, illustrating the improved orientation and crystalline structure of perovskite films. Line-cut profiles further indicated the orientation preference of perovskite films [32]. The diffraction intensity of the (100) peak at q = 1.0 Å was enhanced along the out-of-plane direction for the DO-SnO₂-based perovskite, while reducing along the in-plane direction compared to the control, indicative of a preferential growth direction perpendicular to the substrate (Figure 1h and Figure S3). These above results demonstrated the improved crystallinity, optimized crystal orientation, and higher phase purity of the perovskite films on the DO-SnO₂ substrate, which would facilitate the improvement of contact quality at the buried interface and promote efficient charge transport in F-PSCs [33].

To probe the structural characteristics of perovskite films, the peak-force quantitative nanomechanical atomic force microscopy (PFQNM-AFM) imaging technique was employed (Figure 2a,b). The modulus diagrams are intrinsically linked to variations in the grain structures, thereby reflecting the mechanical flexibility of perovskite films [13]. Compared to the perovskite deposited on pristine SnO₂, the DO-treated case exhibited a significant reduction in Young’s modulus from 21.6 GPa to 9.7 Gpa. The enhanced mechanical resilience of perovskite films on the DO-SnO₂ substrate may contribute to alleviating grain cracking and lattice extrusion during bending cycles. Furthermore, the impact of DO-based interface modification on the residual strain within perovskite films was examined using grazing incidence X-ray diffraction (GIXRD). The (012) plane, characterized by a high diffraction angle of 31.7°, was selected to ensure the reliability of the grain information while minimizing the crystal orientation on strain measurements [34]. As shown in Figure 2c,d, the depth-resolved GIXRD patterns illustrated a gradual shift of the (012) peak towards the lower angle following the increased tilt angle Ψ from 10° to 50°. This reflects an enlarged lattice distance of the (012) plane along the in-plane direction as the detection depth increases. According to the linear relationship between sin² Ψ and 2θ, the negative slopes of the fitting lines reflect the gradual tensile strain within perovskite films across the vertical direction of substrates (Figure 2e). In comparison, the perovskite film using DO-SnO₂ exhibited a lower slope of 0.13 relative to the control of 0.17, illustrating the effective mitigation of tensile strain in the perovskite lattices on the DO-SnO₂ interface. According to Bragg’s Law and generalized Hooke’s Law, the residual strain can be further quantified by the slope of the following equation [35]:

$$\sigma =-{\displaystyle \frac{E}{2\left(1+\mathsf{\nu }\right)}}{\displaystyle \frac{\pi }{180°}}cot{\theta }_0{\displaystyle \frac{\partial \left(2\theta \right)}{\partial \left({sin}^2\phi \right)}}$$

(1)

where $\phi$, $E$, and $\mathsf{\nu }$ represent the angle between the diffraction vector respective to the normal direction of sample surface, perovskite modulus, and Poisson’s ratio of perovskite, respectively. Based on the calculation result, the residual stress of control film was determined to be 49.5 MPa, while the value of target film showed a nearly five-fold reduction to 10.9 MPa (Figure 2f). The significantly mitigated residual stain within films indicates that the perovskite crystals occupied weaker lattice distortion and better crystalline quality, benefiting the enhanced mechanical robustness and reduced trap density in bulk films.

Collecting the observations above, the introduction of the DO ligand at the buried interface of F-PSCs can effectively passivate the trap sites at the SnO₂–perovskite interface and serve as the growth template to optimize the crystalline quality of upper perovskite films. The improved grain sizes with reduced grain boundaries within perovskite films using DO-SnO₂ would relieve the microcrack formation under mechanical stress and boost their bending resistance. Additionally, the uniform morphology and dense grain distribution at the buried interface of perovskite film further strengthen the interfacial adhesion, thereby largely escaping from mechanical delamination during bending treatments [36]. More importantly, the significantly mitigated stain residue within perovskite films upon DO treatment would further boost the mechanical robustness of the absorber, which may synergistically enhance the durability of F-PSCs.

We then explored the effects of interfacial molecular modification on the optoelectronic properties of perovskite films. The ultraviolet-visible (UV-Vis) absorption spectra revealed that both perovskite films using SnO₂ with and without DO decoration presented no distinct difference in absorbance within the visible light region with a similar optical bandgap (Figure 3a,b). The quantitative analysis of trap density and electron mobility of perovskite films were evaluated through dark I–V measurements based on the space–charge–limited current (SCLC) model by constructing an electron-only device (Figure 3c). Following DO treatment, the trap density of perovskite films decreased from 2.60 × 10¹⁶ to 2.14 × 10¹⁶ cm⁻³ (Table S1). Furthermore, compared to the control perovskite film, the DO-based perovskite film showed a considerable enhancement in charge conductivity, rising from 3.64 × 10⁻³ to 5.28 × 10⁻³ mS cm⁻¹ (Figure 3d). This improvement is closely correlated with improved morphology quality and reduced trap density within the perovskite films. Additionally, the steady-state photoluminescence (PL) spectra of perovskite films presented a clear intensity quenching upon the introduction of DO at the SnO₂/perovskite interface (Figure 3e). The average carrier lifetimes (τ_ave) calculated from the time-resolved PL (TRPL) spectra illustrated the shortened carrier lifetime of 0.51 μs for the DO-SnO₂-based films compared to 3.67 μs for the control (Figure 3f, Table S2). Moreover, time-resolved confocal PL mappings further indicated the reduction in carrier lifetime observed in the DO-treated perovskite films with a uniform PL emission on the surface (Figure S4). Collectively, these findings affirm that the introduction of molecular bridge at the buried interface significantly passivates charged defects, thereby facilitating accelerated electron extraction, which is advantageous for enhancing device performance [37].

We fabricated n-i-p structured F-PSCs using a device configuration of PEN/ITO/SnO₂/Perovskite/Spiro-OMeTAD/Au to evaluate the impact of DO decoration on device performance (Figure 4a). The influence of the DO incorporation concentration on the photovoltaic performance of F-PSCs using DO-SnO₂ was verified (Figure S5, Table S3). When the DO concentrations increased from 2 mg mL⁻¹ to 4 mg mL⁻¹, the power conversion efficiency (PCE) showed a trend of first rising and then falling, determining the optimal case of 3 mg mL⁻¹. Figure 4b depicts the current density–voltage (J–V) curves of the optimal devices, and the detailed parameters were summarized in Table S4. The control devices delivered a PCE of 21.87%, with a short-circuit current density (J_SC) of 24.52 mA cm⁻², an open-circuit voltage (V_OC) of 1.11 V, and a fill factor (FF) of 80.62%. Upon introducing the DO, the target devices achieved a significantly increased PCE of 23.50%, primarily attributed to the improvements in FF and V_OC. Furthermore, the statistical analysis of the photovoltaic parameters across 20 individual devices further corroborates the efficiency increase and good reproducibility of DO-SnO₂-based F-PSCs (Figure S6). The integrated current densities derived from the external quantum efficiency (EQE) spectra were determined to be 23.78 and 24.01 mA cm⁻² for the control and target devices, respectively, well-matched with the J–V characterization results (Figure 4c).

To gain insight into the effects of charge transfer and recombination behaviors on the device performance, we then performed the electrochemical impedance spectroscopy (EIS) measurements on devices (Figure 4d) [38]. According to the fitted parameters in Table S5, the target case exhibited the smaller series resistance (R_s) and larger composite resistance (R_rec) than those of the control, indicating the suppressed trap-induced non-radiative recombination upon DO decoration. The dark J–V measurement confirmed the lower leakage currents within the target device, which further indicated the mitigated charge accumulation and effective defect suppression at the contact interface (Figure 4e). The Mott–Schottky (M–S) measurements were also employed to evaluate the built-in potential (V_bi) within F-PSCs. As shown in Figure 4f, the target device presented an improved V_bi from 1.07 to 1.15 V concerning the control device. In addition, the residual charge densities in devices were reduced from 12.7 × 10¹⁵ to 8.05 × 10¹⁵ cm⁻³ with the DO addition, associated with the dark J–V results. These results indicated that the DO modification on the buried interface can synergistically enhance the driving force for the charge transport and suppress the interfacial defects, endowing the reduced carrier accumulation and recombination loss at the electron–extraction interface.

Finally, we evaluated the influences of DO decoration on the long-term stability of F-PSCs. Figure 4g illustrates the efficiency evolution of unencapsulated devices when stored under ambient conditions with a relative humidity of 40% at room temperature. The control devices showed a rapid degradation within the initial 400 h with 72% initial efficiency retainment. In contrast, the DO-treated F-PSCs delivered improved humidity resistance, which retained over 80% of the original performance after 1200 h. The enhanced humidity stability can be attributed to the improved crystalline size with the reduced density of grain boundaries, impeding the immersion of moisture and oxygen into the perovskite lattices. More importantly, the mechanical durability of F-PSCs was significantly boosted with DO modification. The target devices maintained 90% of their initial efficiency after 6000 bending cycles at a bending radius of 4 mm, while the value rapidly reduced to below 85% after 6000 cycles for the control case (Figure 4h, Table S6). These findings highlighted the importance of buried interface modification on the synergistic improvements in the photoelectric performance and operation reliability of F-PSCs.

4. Conclusions

In this work, we proposed a molecular bridge strategy by introducing the DO ligand at the SnO₂/perovskite interface. The systematic photoelectric characterizations elucidate the synergistic optimization effects of diammonium ligands on the buried interface, including defect passivation, crystallization optimization, and charge transport reinforcement. Collecting these advantages, the champion F-PSCs achieved an impressive efficiency of 23.50% with enhanced humidity resistance. More importantly, the reduced voids on the bottom surface and released strain residue across the perovskite contributed to the enhanced mechanical durability of F-PSCs, achieving a T₉₀ lifetime of over 6000 cycles under a bending radius of 4 mm. The findings highlight the importance of multifunctional interface modifiers on the structural integrity of the perovskite structure and photoelectric properties of the charge extraction interface, which may promote the further development of F-PSCs towards the commercialization process.