Introduction

Perovskite solar cells (PSCs) have received much attention in recent years due to their outstanding power conversion efficiency (PCE) of up to 26% [1,2,3]. Inverted planar PSCs, denoted as positive–intrinsic–negative (p − i − n), hold greater promise and appeal when contrasted with the conventional n − i − p-structured PSCs, owing to their promising stability under ambient air, ease of fabrication, and use of flexible substrates [4,5,6,7]. Furthermore, the p − i − n architecture advances the technology of silicon/perovskite tandem cells, potentially expediting the widespread commercial adoption of PSCs.

The function of hole-transport materials (HTMs) is pivotal in p − i − n PSCs, as HTMs facilitate the efficient extraction and transport of photoinduced holes while preventing undesired interfacial charge recombination [8, 9]. HTMs can be categorized into three primary classes within the framework of p − i − n planar PSCs: inorganic HTMs, organic small-molecule HTMs, and polymeric HTMs [10,11,12]. Initially, the polymeric HTM poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) gained widespread recognition for its exceptional attributes, which include high optical transparency, adequate electrical conductivity, and superior wettability. However, growing evidence suggests that the acidic and hygroscopic nature of the PSS components, coupled with energy-level mismatches between HTMs and perovskite materials, may impose limitations on the maximum efficiency of these devices and their stability [13]. Although alternative HTMs like poly[bis(4-phenyl)(2,4,6-trimethylphenyl)amine (PTAA) have been used in high-efficiency PSCs, the reliance on deliquescent dopants inevitably compromises the long-term durability of PSCs [14]. 2,2′,7,7′-tetrakis[N, N-di(4-methoxyphenyl) amino]-9,9′-spirobifluorene (Spiro-OMeTAD) and PTAA suffer from low conductivity when used as HTMs in PSCs, which leads to inferior performance. Dopants like lithium bis(trifluoro-methane-sulfonyl)-imide (Li-TFSI) are added to enhance conductivity. However, these dopants can create pinholes, exposing the perovskite layer and reducing the PSC efficiency [15,16,17]. Hence, there exists a pressing need for cost-effective, readily available, and dopant-free HTMs that can deliver exceptional performance and enhanced stability in p − i − n PSCs.

Dopant-free small molecules are still considered the best alternative HTMs due to their facile synthesis, excellent solubility, and modifiable molecular configurations and energy states. To date, p − i − n PSCs employing dopant-free small-molecule HTMs have demonstrated impressive PCE exceeding 20%, and recent advancements with self-assembled monolayers have pushed the efficiency beyond 26%, as illustrated in Fig. S1 and Table S1 in Supplementary Materials [18,19,20,21,22,23,24,25,26,27,28,29]. Most high-performance dopant-free HTMs adopt symmetrical donor–acceptor–donor (D–A–D) structures. This strategy allows for precise control over the energy levels of the highest occupied molecular orbital (HOMO) and the establishment of a harmonious energy alignment with perovskite materials. Molecules with D–A–D structures exhibit favorable intramolecular charge transfer (ICT) characteristics. These attributes enable self-doping phenomena and built-in potentials, thereby augmenting their charge-extraction capabilities [30, 31].

Moreover, halogenation as a molecular design strategy has proven effective in fine-tuning optoelectronic properties and enhancing the film morphology of small molecules [32,33,34,35,36]. Specifically, the C–Cl bond, among the four carbon–halogen bonds, exhibits the highest dipole moment and best intensifies the ICT effect. Furthermore, chlorinated compound synthesis is more straightforward compared to that of their fluorinated counterparts, making chlorination a particularly appealing and promising avenue for designing highly efficient and cost-effective HTMs for PSCs [37,38,39,40]. Despite its potential, the chlorine substitution of HTMs in PSCs had not received adequate attention until recently. Notably, Guo and colleagues [36] have reported the successful development of a chlorinated HTM, mCl-SFXDA, based on the spiro[fluorene-9,9′-xanthene] architecture. This HTM achieved an impressive PCE of 22.14% when used with dopants in n − i − p PSCs.

In this research, we have systematically explored the synthesis and application of chlorinated small-molecule HTMs in dopant-free p − i − n PSCs. The designed chlorinated HTMs, namely fluorenone-triphenylamine (FO-TPA)-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl, are illustrated in Fig. 1a. These HTMs feature fluorenone as the A unit and triphenylamine (TPA) with various chlorine atoms as the D units, forming a D − A − D molecular framework. The specific chlorine substitution positions (para, meta, and ortho) on the terminal benzene units significantly influence the optoelectronic, thermal, and morphological properties of these conjugated organic small molecules. Compared to the chlorine-free FO-TPA-CH3, the FO-TPA-x-Cl (x = p, m, and o) HTMs exhibit markedly improved film coverage, enhanced hole extraction efficiency, and appropriately aligned energy levels with perovskite, resulting in an enhancement of PCE by 0.68–3.15%. Remarkably, the FO-TPA-o-Cl-based device achieved an exceptional PCE of 20.82%, featuring an open-circuit voltage (Voc) of 1.09 V, a short-circuit current density (Jsc) of 24.13 mA/cm2, and a commendable fill factor (FF) of 0.80 under standard AM 1.5G illumination conditions (100 mW/cm2). These performance metrics surpass those of the other two chlorinated HTMs and the chlorine-free FO-TPA-CH3-based devices. Significantly, the unencapsulated devices using FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl HTMs demonstrated robust operational stability, retaining 88.3%, 89.7%, and 93.8% of their initial PCEs, respectively, after an extended 720-h aging test under ambient air conditions. In contrast, the device using chlorine-free FO-TPA-CH3 retained a mere 86.9% of its initial PCE throughout the identical testing duration.

Fig. 1
figure 1

a Molecular structures of hole-transport materials (HTMs); b normalized absorption and emission spectra of HTMs in CH2Cl2 solutions; c energy level schematic of the constructed inverted PSCs

Full size image

Results and Discussion

The synthesis of these dopant-free small-molecule HTMs is elucidated in Fig. S2, and comprehensive experimental procedures can be found in Supplementary Materials. Initially, Intermediate 2 was obtained through a Suzuki–Miyaura coupling reaction, wherein 2, 7-dibromo-9-fluorenone was coupled with 2-methoxyphenyl boronic acid in the presence of a Pd(PPh3)4 catalyst. The resulting intermediate was further modified to yield the dibromo-substituted compound FO-Br, employing N-bromo succinimide as the brominating agent, with a commendable yield of approximately 85%. Following this, C − N bond formations were executed through Buchwald–Hartwig cross-coupling reactions between FO-Br and diphenylamine derivatives, namely 4, 4'-dimethyldiphenylamine, 4-chlorodiphenylamine, 3-chlorodiphenylamine, and 2, 6-dichlorodiphenylamine. These reactions yielded the desired small-molecule HTMs, namely FO-TPA-CH3, FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl, respectively, with respectable yields ranging from 60 to 78%. Purification of the synthesized small-molecule HTMs was achieved through column chromatography, using silica gel and petroleum ether: CH2Cl2 eluent mixture in a 3:1 ratio, followed by recrystallization from methanol. The structural confirmation of the intermediates and the final small-molecule HTMs was accomplished through 1H nuclear magnetic resonance spectra, as depicted in Figs. S3 − S7.

Figure 1b and S8a display the UV − Vis absorption and fluorescence emission spectra of these HTMs in CH2Cl2 solutions and thin film, with pertinent data summarized in Table S2. The newly developed HTMs display distinct absorption characteristics, featuring a prominent absorption peak at approximately 285 nm that can be attributed to the π–π* local electron transition. Simultaneously, the comparatively weaker absorption at 437 nm is attributed to intramolecular charge transitions between the fluorenone core and the TPA moiety [41, 42]. The optical bandgap values (Egopt) for FO-TPA-CH3, FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl were established as 2.68, 2.65, 2.65, and 2.66 eV, respectively, by analyzing the overlapping regions of their normalized absorption and emission spectra (Fig. 1b). As illustrated in Fig. S8b, all HTMs exhibit efficient optical transmittance within the spectral range spanning from 500 to 800 nm, effectively reducing parasitic absorption. Furthermore, the first oxidation potentials (Eox), indicative of the highest occupied molecular orbital (HOMO) energy levels of the HTMs, were determined via cyclic voltammetry, as shown in Fig. S9. To determine the lowest unoccupied molecular orbital (LUMO) energy levels, Egopts were combined with the corresponding HOMO energy levels (LUMO = HOMO + Egopt). Energy-band diagrams of the perovskite and these HTMs are presented in Fig. 1c. Compared with FO-TPA-CH3 (− 5.19 eV), the HOMO energy levels of the three chlorinated HTMs are slightly deeper due to the strong electronegativity of chlorine atoms, measuring approximately − 5.26, − 5.29, and − 5.39 eV for FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl, respectively [43, 44]. The introduction of chlorine substituents into these HTMs enhances the alignment between their HOMO and the valence band of the perovskite, effectively reducing the interfacial energy barrier and promoting hole transfer within p − i − n PSCs. Furthermore, their calculated LUMO energy levels range from − 2.61 to − 2.73 eV, representing a significant increase compared to the perovskite conduction band (− 3.93 eV) and providing efficient electron-blocking properties at the perovskite/HTM interface.

We performed thermogravimetric analysis (TGA) and differential scanning calorimetry to assess the thermal characteristics of these small-molecule HTMs, as shown in Fig. S10. Figure S10a reveals exceptional thermal stability in the chlorinated derivatives, with each exhibiting a thermal decomposition temperature (Td) for a 5% weight loss surpassing that of chlorine-free FO-TPA-CH3. Figure S10b illustrates a distinctive glass transition occurring at 136 °C for FO-TPA-CH3. Conversely, no phase transition was observed in the three chlorinated HTMs up to 300 °C, indicating limited molecular mobility in response to temperature changes. This finding implies that the chlorinated HTMs possess a more rigid molecular structure via the internal heavy-atom effect when compared to their chlorine-free counterpart [45].

Electronic and geometrical properties of the molecules were investigated through density functional theory (DFT) calculations using the B3LYP/6-31G level of theory in the Gaussian 09 software [46, 47]. The optimized geometries of the four small-molecule HTMs are shown in Fig. 2a. The results, illustrated in Fig. 2b, reveal that the electron densities of the HOMOs for the four small-molecule HTMs are predominantly situated on the peripheral TPA moieties, while their LUMOs are concentrated on the fluorenone cores. However, clear differences in HOMOs can be observed as the positions of the Cl atoms change. The terminal Cl atoms exhibit noticeable HOMO distribution for FO-TPA-p-Cl, whereas no such distribution is evident at the Cl atoms for both FO-TPA-m-Cl and FO-TPA-o-Cl. The theoretical calculations also indicate that the introduction of a chlorine atom into the molecule results in reduced HOMO and LUMO energy levels, a trend consistent with the experimental results. Furthermore, the DFT calculation was also performed to obtain the electrostatic potential (ESP) surface (Fig. 2c). The electronegative charges are predominantly situated on the fluorenone moieties, particularly in the vicinities of oxygen atoms, signifying their possession of Lewis base properties conducive to secondary interactions with under-coordinated and electropositive Pb2+ on the perovskite surface [48]. Moreover, in comparison with chlorine-free FO-TPA-CH3, the ESP surface images of all the chlorinated HTMs exhibited negative potential distribution around the Cl atoms due to their pronounced electron-withdrawing capacity. The electron-rich property of the chlorinated area may also function as a soft Lewis base to stabilize the uncoordinated Pb sites in perovskite, passivating the defects at the HTM and perovskite interface [49].

Fig. 2
figure 2

a Optimized molecule geometries; b spatial distribution of the frontier molecular orbitals; c ESPs of the investigated hole-transport materials (HTMs) (the negative and positive potentials are depicted using red and blue, respectively)

Full size image

The film morphologies and quality of these HTMs on indium tin oxide (ITO) substrates were assessed using atomic force microscopy (AFM), as illustrated in Fig. 3a. The root-mean-square roughness of the chlorinated HTM layers of FO-TPA-p-Cl (5.80 nm), FO-TPA-m-Cl (6.44 nm), and FO-TPA-o-Cl (2.44 nm) is notably smaller compared to the chlorine-free FO-TPA-CH3 (8.39 nm), indicating that modifying the position of chlorine atoms in HTMs can optimize the film morphology. The improved coverage of chlorinated films efficiently mitigates shunt paths between perovskite and ITO, thereby suppressing nonradiative recombination at the junction. We further investigated the surface wettability of these four HTMs through water contact-angle testing (Fig. 3b). The water contact angles for FO-TPA-CH3, FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl were measured at 82.2°, 92.0°, 93.8°, and 84.2°, respectively, which are significantly larger than those of benchmark HTMs like small-molecule Spiro-OMeTAD (38.7°) and polymeric PEDOT:PSS (16.9°). All the newly developed HTMs, along with the conventional PTAA (with a contact angle of 90.6°), display hydrophobic properties. These properties can potentially serve as a barrier, preventing moisture ingress that might lead to perovskite absorber degradation. This, in turn, promotes the formation of high-caliber perovskite crystals and facilitates the overall stability of environmental devices [50,51,52].

Fig. 3
figure 3

a Atomic force microscopy (AFM) height images depicting the hole-transport materials (HTMs) on indium tin oxide (ITO) substrates; b visual representation of the water contact angles exhibited by the HTMs on ITO substrates; c scanning electron microscopy (SEM) images, captured from a top-view perspective, showcasing the prepared perovskite films atop the HTMs; d X-ray diffraction (XRD) analysis of the perovskite films grown on HTMs; e photoluminescence (PL) spectra of the perovskite films deposited on glass substrates: glass/fluorenone-triphenylamine (FO-TPA)-CH3, glass/FO-TPA-p-Cl, glass/FO-TPA-m-Cl, and glass/FO-TPA-o-Cl

Full size image

Figure. 3c shows scanning electron microscopy (SEM) images of perovskite films that grew on layers of FO-TPA-CH3, FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl. The FO-TPA-CH3/perovskite film has a random perovskite crystal-size distribution, and pinholes appear at the grain boundaries, while perovskite growth on FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl films is more homogeneous and compact. The better coverage and crystallization of perovskite film may profit from the improved film morphology of chlorinated HTMs [53]. We performed X-ray diffraction (XRD) on perovskite thin films for further assessment of perovskite crystal quality (Fig. 3d). When contrasted with FO-TPA-CH3, perovskite films grown on FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl exhibit more pronounced X-ray diffraction peaks. This observation suggests the superior crystallization of the perovskite film, which in turn contributes to the improved efficiency and stability of the devices [54]. These findings align with the SEM results, which illustrate a denser arrangement of perovskite grains. Thus, manipulating the chlorine atom positions within the HTMs can effectively optimize perovskite film morphology and enhance perovskite crystallinity.

We conducted steady-state photoluminescence (PL) measurements on perovskite films deposited on both glass and various HTMs to explore the dynamics of hole extraction and charge transfer at the interfaces between the HTMs and perovskite (Fig. 3e). Compared to the original perovskite films on glass, the PL spectra of the four HTM/perovskite films exhibit significant quenching, indicating the rapid extraction of holes at the HTM/perovskite interface. The three chlorinated HTMs demonstrate superior quenching of PL emitted by the perovskite films compared to FO-TPA-CH3 HTM, facilitating enhanced charge transfer from the perovskite into the HTM layer. Intriguingly, the FO-TPA-o-Cl/perovskite exhibits the weakest PL intensity among these chlorinated HTMs, signifying a more efficient hole transfer from perovskite to the HTM. This could be ascribed to the inherent advantage of producing high-quality chlorinated HTM films and forging an enhanced interfacial chemical interaction at the HTM/perovskite junction [55].

Building upon these investigations, we assembled dopant-free planar p − i − n PSCs (ITO/HTM/LiF/perovskite/LiF/C60/bathocuproine (BCP)/Ag) using the developed HTMs. The complete device configuration, along with an SEM cross-sectional image depicting the p − i − n PSC, is presented in Fig. 4a and b, respectively. The comprehensive photovoltaic parameters are consolidated in Table 1. Analysis of the current–voltage (J–V) plots reveals that devices incorporating FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl exhibited excellent FF values (ranging from 0.77 to 0.80), surpassing the FF of the device based on FO-TPA-CH3 (0.72) (Fig. 4c). Particularly noteworthy are the Voc values of the PSCs containing FO-TPA-m-Cl and FO-TPA-o-Cl (ranging from 1.04 V to 1.09 V), which exceed those of the device using FO-TPA-CH3 (1.02 V). This enhancement in Voc can be ascribed to the improved alignment of HOMO energy levels between chlorinated HTMs and perovskite, along with the Lewis base passivation of under-coordinated Pb2+ ion defects at the HTM/perovskite interface [45, 56]. Comparing the performance of PSCs based on FO-TPA-CH3, which achieved a PCE of 17.67%, with those based on the three chlorinated HTMs (PCEs of 18.35%–20.82%), reveals significant efficiency gains. Notably, the FO-TPA-o-Cl-derived PSC exhibited the highest Voc of 1.09 V during reverse scanning, Jsc of 24.13 mA/cm2, and FF of 0.80, resulting in the best PCE of 20.82%. This performance surpasses that of PTAA (without doping) and PEDOT:PSS-based devices (Fig. S11), establishing it as a frontrunner compared to other recently documented small-molecule HTMs in dopant-free p–i–n PSCs (Fig. S1 and Table S1). Despite these small-molecule HTMs sharing identical core backbones, the variations in substituent groups exert a considerable influence on the photovoltaic efficacy of PSCs [57].

Fig. 4
figure 4

a Device architecture of the inverted perovskite solar cells (PSCs); b scanning electron microscopy (SEM) cross-sectional image of the device based on fluorenone-triphenylamine (FO-TPA)-o-Cl; c current–voltage (J–V) characteristics of the best-performing devices employing various hole-transport materials (HTMs); d J–V curves for the most efficient device under reverse and forward scans with FO-TPA-o-Cl; e external quantum efficiency (EQE) spectra and integrated Jsc curves; f Histograms depicting the statistical distribution of power conversion efficiencies (PCEs) with FO-TPA-o-Cl

Full size image
Table 1 Comparison of device characteristics of the inverted PSCs with the indicated HTMs
Full size table

During forward scanning, the FO-TPA-o-Cl device achieves a PCE of 20.14% (Fig. 4d). To evaluate device hysteresis, we employ the hysteresis index (HI), calculated as HI = (PCEreverse − PCEforward)/PCEreverse [58]. For FO-TPA-o-Cl-based devices, the HI measures 3.27%. Figure 4e presents the external quantum efficiency (EQE) spectra of FO-TPA-o-Cl-based devices, revealing impressive EQE values exceeding 80% throughout the spectral span from 400 to 800 nm. The integrated Jsc value of 24.10 mA/cm2 aligns with the measured J–V data, affirming the reliability of our measurements. To assess the reproducibility of FO-TPA-o-Cl-based PSCs, we fabricated and characterized 20 devices, and the statistical distribution is presented in Fig. 4f. The mean PCE of FO-TPA-o-Cl PSCs, determined from the results of these 20 devices, is 20.41%.

Subsequent stability assessments were performed under ambient air conditions with relative humidity ranging from approximately 20% to 30%, all conducted without the use of encapsulation. As depicted in Figs. 5 and S12, the attenuation curve illustrates the gradual declines in PCEs for these devices over time. In contrast to devices using the chlorine-free HTM FO-TPA-CH3, which retained 86.9% of its initial PCE, PSCs employing FO-TPA-p-Cl, FO-TPA-m-Cl, and FO-TPA-o-Cl as dopant-free HTMs demonstrate significantly improved stability. Specifically, they maintained 88.3%, 89.7%, and 93.8% of their initial PCEs after 720 h, respectively. In the aging study (Fig. S13), XRD analysis revealed that the chlorine-free FO-TPA-CH3 showed signs of perovskite decomposition with the emergence of PbI2 peaks indicating material degradation, while the chlorinated HTM FO-TPA-o-Cl maintained its perovskite structure but exhibited changes in peak intensities, suggesting alterations at the material interfaces that might contribute to efficiency losses. This enhanced stability can be attributed to a combination of factors that stem from the synergistic interactions between the chlorine atoms and C=O groups present in the chlorinated HTMs, including optimized film morphology, enhanced crystallinity, and the passivation of interfacial and grain boundary defects within the perovskite layer. Consequently, PSCs featuring these chlorinated HTMs exhibit exceptional initial performance and long-term stability when exposed to ambient conditions.

Fig. 5
figure 5

Unencapsulated devices were exposed to ambient conditions at 25 °C with a relative humidity of 20 − 30% and kept in the dark to assess their stability

Full size image

Conclusion

Our study has developed high-performance dopant-free HTMs for p − i − n perovskite solar cells by synthesizing a series of chlorinated small molecules, specifically FO-TPA-x-Cl (x = p, m, and o). The strategic incorporation of chlorine substituents into these small-molecule HTMs optimizes the energy level alignment for efficient hole extraction, improves HTM film morphology for enhanced coverage, and enlarges perovskite crystal grains, collectively boosting device performance. We observed remarkable improvements in key photovoltaic parameters in p − i − n PSCs from these enhancements, including Voc, Jsc, and FF. The FO-TPA-o-Cl variant exhibits exceptional performance—achieving a PCE of 20.82% and maintaining more than 93% of its initial efficiency after 720 h in ambient air. These outcomes not only demonstrate the profound impact of chlorine-substituent control but also highlight the potential of these chlorinated molecules as economical and effective organic HTMs, promising the future development of high-performance PSCs.