Strategy for the isomerization of dibenzo[b,d]furan-based carbazole derivatives as hole transporting materials for perovskite solar cells: theoretical design and experimental study

Abstract

The structural design of hole transport materials (HTMs) is a crucial approach to improving the efficiency and stability of perovskite solar cells (PSCs). In this study, a series of isomeric dibenzo[b,d]furan-based carbazole derivatives (CX11–CX14) were designed to provide a design strategy for the development of HTMs in PSC applications. Side chain isomerism has a significant impact on molecular conjugation, exhibiting distinct isomer-dependent effects in terms of energy levels, planarity, dipole moment, and hole mobility. Furthermore, theoretical calculations and experimental results indicate that the molecule CX11 with superior hole mobility and stronger adsorption on the perovskite surface can act as a potential HTM for PSC applications. According to the results of the optimized PSC devices, the power conversion efficiency (PCE) of the CX11-based PSC exceeded 23%, which is higher than that of devices based on other molecules. The close agreement between computational predictions and experimental validation not only validates the theoretical framework for designing molecular isomers of HTMs but also provides crucial molecular-level insights. The demonstrated methodology is expected to motivate researchers to develop even more efficient HTM isomers for PSCs with higher PCEs.

---

Design, System, Application

Hole transport materials (HTMs) are critical components of perovskite solar cells (PSCs), which can effectively enhance device performance and long-term stability. In this work, based on a carbazole–diphenylamine derivative as the core structure, isomeric dibenzo[b,d]furan as a side-chain was employed to design four molecules (CX11, CX12, CX13, and CX14). The geometric structures, electronic properties, molecular packing morphologies, interfacial characteristics, and hole mobility of these molecules were simulated using density functional theory, molecular dynamics simulations, and the Marcus theory. To validate the potential of these designed molecules as HTMs, the designed molecules were synthesized and utilized to fabricate PSC devices. Experimental results demonstrated that CX11, when used as the HTM, achieved a power conversion efficiency of 23.09% in the PSCs devices, outperforming other isomers of CX12, CX13 and CX14. These findings not only validated the reliability of the computational models but also provided new insights into designing and identifying high-performance HTMs through isomerization strategies.

1 Introduction

As an emerging photovoltaic technology, perovskite solar cells (PSCs) have witnessed continuous improvements in device fabrication and material optimization, showcasing their potential for widespread application. Remarkable progress in power conversion efficiency (PCE) has been achieved, with certified PCE values for single-junction architectures reaching an impressive 27%.¹ Under the premise of proper energy level alignment among different layers, enhancing hole extraction, suppressing interfacial charge recombination, and effectively controlling defect levels remain key challenges for improving device efficiency. In PSCs, hole transport materials (HTMs) in the hole transport layer (HTL) can efficiently extract photogenerated holes from the perovskite layer and passivate defects in the perovskite film. The characteristics of these materials have a significant impact on device performance.^2–4 In recent years, the structural design of organic small molecules has emerged as an effective strategy for functionalizing HTMs.

The furan group possesses an electron-donating ability, and the lone pair electrons of the oxygen atom enable the molecule to interact with Pb ions in the perovskite film. The π-conjugated structure facilitates molecular stacking and hole migration between molecules. Its derivatives are thus widely incorporated into the design of PSC materials.^5–7 Yao et al. achieved efficient PSCs by modulating the crystallization of FAPbI₃ using the furan-based additives (FFEACl). The aromaticity and π-conjugated structure of the furan ring confer strong binding affinity toward Pb²⁺ ions, thereby enhancing its interaction with the perovskite. The addition of FFEACl molecules can result in a more uniform surface potential of the perovskite and reduce its defect density.⁸ Dai et al.⁹ introduced diphenylfuran into the spiro-OMeTAD side chain to replace the anisole units. By altering the substitution positions of dibenzofuran, a series of spiro HTMs were designed, with positional isomerism having a significant impact on the molecular properties and device performance. Moreover, after the introduction of dibenzofuran, the molecules exhibited extended conjugation, lower HOMO levels, and better thermal stability, film-forming ability, and hydrophobicity.

In an increasing number of studies, quantum chemical technology and theoretical calculations have been employed to analyze and predict the geometric configuration, electronic properties and material characteristic of HTMs. These approaches establish a comprehensive theoretical framework for the molecular engineering of HTMs, offering systematic guidance for their rational design. For example, Ding et al.¹⁰ introduced chlorine into the spiro-OMeTAD structure to design spiro-mCl. The combination of molecular dynamics (MD) simulations and first-principles calculations indicated that the oxygen atom of the spiro-mCl side chain interacts more strongly with Pb ions on the perovskite compared to spiro-OMeTAD, with the Cl atom also contributing to the interaction. These studies visually demonstrated the interfacial adsorption behavior and the magnitude of the adsorption energy. Due to the electron-withdrawing ability of the Cl atom, spiro-mCl exhibited better film-forming ability and a superior hole extraction advantage at the interface. After the application of spiro-mCl, the PCE was 25.26%.

In this work, a series of isomeric dibenzo[b,d]furan-based carbazole derivatives (CX11–CX14) were designed to provide a design strategy for the development of HTMs in PSC applications (Fig. 1a). Through a synergistic computational approach integrating density functional theory (DFT), time-dependent DFT (TD-DFT), first-principles calculations, Marcus charge transfer analysis, and molecular dynamics (MD) simulations, we systematically investigated the hole transport properties of DBF-substituted HTMs (CX11–CX14) and their interfacial charge behavior with perovskite films. Theoretical modeling revealed position-dependent molecular mechanisms: variations in the DBF substitution sites induced distinct intermolecular interaction patterns, which governed both the spatial hole mobility across perovskite surfaces and the thermodynamic stability of adsorption configurations. To establish computational-experimental consistency, the synthesized HTMs (CX11–CX14) were successfully implemented in photovoltaic devices, demonstrating remarkable agreement with the theoretical predictions. The study found that the performance differences of the PSCs were related to the isomeric variation of the DBF position. Among them, CX11 exhibited the best hole mobility in PSCs and demonstrated the optimal film morphology and ability to suppress interfacial recombination. Consequently, under the same conditions, the PCE of the CX11-based PSC increased to 23.09%, while the PCE values of the other isomers, CX12, CX13, and CX14, were 22.07%, 18.58%, and 20.61%, respectively. The experimental results were in high agreement with the theoretical simulations, fully validating the reliability and rationality of the calculated model. This work not only reveals the crucial impact of the HTM side chain substitution position on the photovoltaic performance of PSCs but also provides practical significance for the molecular design of HTMs.

Fig. 1 (a) Design of the CX11–CX14 molecules. (b) Electron density distribution of the HOMO and LUMO orbitals for CX11–CX14.

2 Results and discussion

2.1. Theoretical simulations

Fig. S1 (SI) presents the optimized configurations of the DBF-functionalized HTMs (CX11–CX14), obtained through DFT calculations employing the B3P86 hybrid functional with the 6-311G(d,p) basis set.¹¹ The isomerism of the side chains altered the dihedral angles between adjacent functional groups, with smaller dihedral angles between the substituents and the biphenyl backbone generally corresponding to enhanced electronic conjugation.¹² CX11 exhibits better planarity. In Fig. 1(b), the HOMO orbitals of CX11–CX14 are similarly delocalized, while the LUMO of CX13 is mainly distributed on the side chain dibenzofuran, whereas the other molecules exhibit extended LUMO distributions. The observed discrepancy can be attributed to DBF-induced steric/electronic modulation, where ortho- versus para-substitution patterns differentially perturb the continuity of π-electron delocalization. In the ESP visualization, the DBF groups on the side chains of CX11–CX14 show significant negative charge regions, exhibiting Lewis base characteristics that help passivate the interface defects of uncoordinated Pb ions.¹³

The evaluated values of the HOMO and LUMO for CX11–CX14 are presented in Table 1. The well-aligned energy levels between the perovskite and the HTMs facilitate hole transfer to the electrode, suppress interfacial charge recombination, and promote the generation of a photocurrent.^14–16 Fig. S2 (SI) presents the simulated UV-visible absorption spectra of CX11–CX14 in dichloromethane (CH₂Cl₂). In the visible light region, there is no significant absorption, ensuring no interference with the sunlight harvesting of the perovskite.¹¹ As shown in Table 1, the dipole moment (μ) of the designed molecules follows the order: CX11 (2.87 D) > CX12 (2.51 D) > CX13 (2.27 D) > CX14 (1.98 D). In addtion, CX11 exhibits a larger emission wavelength and Stokes shift, which may be due to a greater structural change in the excited state compared to the ground state. CX11 also shows the largest dipole moment. An increased dipole moment can enhance intermolecular interactions, improve molecular packing and interfacial compatibility, and effectively enhance the film quality and hole mobility.¹⁷

Table 1 Calculated and simulated data of electronic and optical properties

	HOMO (eV)	LUMO (eV)	E g (eV)	ΔG (eV)	Absorption					Emission
					λ abs (nm)	μ (D)	η (eV)	f	Assignments	λ em (nm)	Shift (nm)
CX11	−5.08	−2.16	2.93	−0.52	406.32	2.87	2.38	0.23	H → L: 82.9%, H → L + 1: 13.3%	526	120
CX12	−5.10	−2.23	2.87	−0.51	415.04	2.51	2.33	0.29	H → L 94.8%	508	93
CX13	−5.10	−1.99	3.10	−0.51	401.57	2.27	2.46	0.03	H → L + 1: 91.0%, H → L + 2: 5.7%	446	45
CX14	−5.12	−2.14	2.98	−0.50	405.40	1.98	2.55	0.18	H → L: 76.2%, H → L + 1: 20.6%	518	112

---

To obtain the structures of molecular stacking and interface models, using 36 molecules, the interfacial adsorption behavior for the HTM on the perovskite film over 10 ns was studied using MD simulations (Fig. S3) (SI).¹⁸ CX11 exhibits a more compact adsorption morphology on the perovskite, facilitating more efficient interactions with the perovskite. Based on the final adsorption configuration, first-principles calculations were employed to investigate the interfacial adsorption configuration and charge transfer after applying the HTMs. As shown in Fig. 2(a) and Table S1 (SI), CX11 demonstrates the flattest adsorption configuration and the widest interfacial contact, which favor the promotion of hole extraction at the perovskite/HTM interface. CX11 also exhibits the highest adsorption energy (−3.89 eV). The interaction between the PVK surface and the dibenzofuran group is consistent with the computed electron localization function (ELF) diagram (Fig. 2b), which shows that CX11 has a higher degree of electronic delocalization at the interface, promoting charge separation and transfer.¹⁹

Fig. 2 (a) Interface structures of CX11–CX14 on the perovskite surface and the corresponding electron charge density differences. (b) Side view of the electron localization function (ELF) at the perovskite/HTM (CX11–CX14) interface. The oxygen atoms in dibenzofuran are marked with red crosses. (c) Main hole transporting pathways for CX11–CX14. (d) Isodensity surface plots of CX11–CX14 in the transporting pathways (isodensity = 1 × 10⁶). (e) Visualization of weak interactions of CX11–CX14 in the transporting pathways.

The hole mobility (μ_h) of HTMs significantly influences the PCE of PSCs. The hole transport pathways of CX11-CX14 were determined from the final structures and are shown in Fig. 2(c). The hole recombination energies (λ_h) of CX11–CX14, as shown in Table 2, are 0.194, 0.191, 0.192, and 0.192 eV, respectively. As shown in Table 2, the simulated hole mobility (μ_h) values for CX11–CX14 are in the following order: CX11 > CX12 > CX14 > CX13. The charge transfer rate (k_ij) is influenced by the charge transfer integral (v_h). However, among all the transport pathways, the pathway exhibiting the highest v_h value significantly contributes to enhancing the values of μ_h.

Table 2 Calculated and simulated data of hole transporting properties

Molecules	Pathways	λ [eV]	r i [Å]	v [eV]	k ij [s^−1]	μ h [cm^2 V^−1 s^−1]
CX11	1	0.194	11.03	1.77 × 10^−3	1.82 × 10^10	8.88 × 10^−3
	2		10.3	−1.50 × 10^−3	1.30 × 10^10
	3		9.139	1.52 × 10^−3	1.35 × 10^10
	4		9.493	−6.07 × 10^−3	2.41 × 10^11
	5		8.99	3.74 × 10^−3	8.15 × 10^10
	6		8.146	−8.02 × 10^−4	3.73 × 10^9
CX12	1	0.191	15.39	−1.62 × 10^−4	1.59 × 10^8	2.92 × 10^−4
	2		12.76	−1.02 × 10^−4	6.26 × 10^7
	3		15	1.00 × 10^−5	6.06 × 10^5
	4		10.2	3.95 × 10^−5	9.45 × 10^6
	5		12.4	−7.47 × 10^−4	3.38 × 10^9
	6		11.85	−2.24 × 10^−4	3.04 × 10^8
CX13	1	0.192	15.55	2.07 × 10^−4	2.57 × 10^8	3.50 × 10^−5
	2		12.76	−1.00 × 10^−5	5.97 × 10^5
	3		18.3	4.35 × 10^−5	1.13 × 10^7
	4		17.64	1.31 × 10^−5	1.03 × 10^6
	5		15.17	1.87 × 10^−4	2.08 × 10^8
	6		17.04	1.05 × 10^−5	6.62 × 10^5
CX14	1	0.192	16.7	3.68 × 10^−5	8.08 × 10^6	2.90 × 10^−4
	2		13.06	3.79 × 10^−4	8.56 × 10^8
	3		11.3	−2.58 × 10^−5	3.96 × 10^6
	4		11.83	2.14 × 10^−5	2.73 × 10^6
	5		14.02	−1.25 × 10^−4	9.32 × 10^7
	6		11.75	−8.01 × 10^−4	3.83 × 10^9

---

Hole transfer is governed by frontier orbital spatial overlap conditions, and the value of v_h is related to the magnitude of the wave function overlap integral in π-stacked architectures. Fig. 2(d) visualizes the overlapping regions of the HOMO orbitals in the dimer fragments with the highest coupling strength for each HTM molecule. The extent of overlap follows the order of CX11 > CX12 > CX14 > CX13. Fig. 2(e) employs the intermolecular region of interaction (IRI) method to visualize the intermolecular interactions of the dimer fragments.²⁰ The interaction range between CX11 dimers is the largest, dominated by noncovalent interactions, spatially localized near diphenylamine–carbazole moieties, where their HOMO delocalization regions overlap. Compared to other molecules, CX13 shows a significantly smaller interaction range and exhibits weaker steric hindrance, resulting in the lowest v_h. CX14 dimers have a smaller intermolecular distance and show a larger interaction range than CX12. However, the interactions do not primarily localize near diphenylamine–carbazole moieties but are instead dominated by major van der Waals interactions between the dibenzofuran and benzene rings of the dimers, along with stronger steric hindrance, leading to a smaller v_h for CX14 than that for CX12.

The intermolecular interactions of the dimer fragments with the strongest coupling for each HTM molecule were evaluated by the energy decomposition analysis (EDA) using the Gaussian 09 and Multiwfn programs with the B3LYP-D3(BJ) functional and 6-31+G(d,p) basis set.^21–23 Table S2 (SI) shows that the order of the total interaction energy (ΔE_int) is as follows: CX11 (−24.89 kcal mol⁻¹) > CX14 (−13.41 kcal mol⁻¹) > CX12 (−5.17 kcal mol⁻¹) > CX13 (−2.12 kcal mol⁻¹). The size of the Coulomb interaction (ΔE_c) reveals the strength of the intermolecular dispersion forces. The smaller centroid distance (r_i) of CX11 leads to a greater deformation of this dimer, thus enhancing the dispersion force. The attractive interactions between the four dimers are dominated by dispersion forces, and thus, the key to the HOMO–HOMO coupling strength lies in the location of the interaction. Combining this with Fig. 2(d), it is clear that CX11 has a higher v_h than CX12–CX14 because the interaction region overlaps more with the delocalized HOMO.

2.2. Synthesis, characterization and device performance

The isomers CX11–CX14 with dibenzofuran substituents exhibit good energy level alignment with the perovskite. Moreover, in the perovskite/HTM interface simulation, CX11 demonstrates compact molecular stacking and superior interfacial compatibility. To confirm the theoretical results, we synthesized CX11–CX14 (Fig. S4) (SI). The experimental procedures are detailed in the SI, and the ¹³C-NMR, ¹H-NMR, and HRMS spectra of the HTMs are shown in Fig. S5–S16 (SI).

The cyclic voltammetry (CV) and UV-visible absorption spectra (Fig. S17 and S18) (SI) of the materials were measured. CX11–CX14 exhibit strong absorption below 350 nm, mainly due to the π–π* transitions within the dibenzofuran, diphenylamine, and carbazole units. A weaker broad absorption peak around 400 nm may correspond to the π–π stacking of the dibenzofuran unit. The relevant parameters are shown in Table 3. The CV measurements indicate that the HOMO energy levels of CX11–CX14 are −5.14, −5.10, −5.10, and −5.12 eV, respectively. Based on the E_g and HOMO values, the LUMO energy levels are estimated to be −2.33, −2.29, −2.31, and −2.33 eV, respectively. The experimentally obtained HOMO and LUMO energy levels are well matched with the VB and CB of the perovskite. The energy level alignment between the HTM and the perovskite helps reduce charge carrier recombination at the perovskite/HTM interface, which is crucial for the realization of efficient photocurrent generation in PSCs.

Table 3 Experimental data of the optical properties and electrochemistry

HTM	λ max [nm]	λ onset [nm]	E g [eV]	E ox [eV]	HOMO	LUMO
					[eV]	[eV]
a Absorption spectra were measured in a dichloromethane solution. b Optical band gap (Eg) obtained from the onset values of absorption (λonset). c Onset of oxidation potentials measured by cyclic voltammetry.
CX11	410	442	2.81	0.46	−5.14	−2.33
CX12	418	441	2.81	0.42	−5.10	−2.29
CX13	406	445	2.79	0.42	−5.10	−2.31
CX14	408	444	2.79	0.44	−5.12	−2.33

---

The experimental values of hole mobility for the designed HTMs and spiro-OMeTAD in pure hole devices were measured using the space-charge-limited current (SCLC) technique, with the data shown in Fig. 3a and S19(a) (SI). The hole mobility is ranked as follows: CX11 (9.15 × 10⁻⁴ cm² V⁻¹ s⁻¹) > spiro-OMeTAD (6.33 × 10⁻⁴ cm² V⁻¹ s⁻¹) > CX12 (3.91 × 10⁻⁴ cm² V⁻¹ s⁻¹) > CX14 (3.34 × 10⁻⁴ cm² V⁻¹ s⁻¹) > CX13 (1.70 × 10⁻⁴ cm² V⁻¹ s⁻¹), which is consistent with the trends predicted by simulations.

Fig. 3 (a) J–V curves of CX11–CX14 for the characterization of hole mobilities. (b and c) Steady-state PL spectra and time-resolved PL decay of films (e.g. perovskite and CX11–CX14 films). (d–g) AFM images. (h) J–V curves of CX11–CX14 for the characterization of the defect states. (i–m) Top view SEM images (e.g. perovskite and CX11–CX14 films).

As shown in Fig. 3b, the lifetime of the film decreases after coating with HTMs, with the most significant decrease observed for CX11. This suggests more efficient hole extraction at the interface, which can be attributed to CX11's higher hole mobility and stronger interface interactions.²⁴ This indicates that CX11 may more effectively extract free charges from the perovskite, promoting the separation of electrons and holes while suppressing their recombination, thereby enhancing the photovoltaic performance.

Steady-state photoluminescence (PL) was used to measure the emission spectra of the HTM-coated and bare perovskite films (Fig. 3c). The fluorescence emission intensity significantly decreased when the HTMs were deposited on the perovskite. CX11 showed the most pronounced PL quenching among all molecules, indicating its stronger hole extraction ability.²⁵ Furthermore, a blue shift of the PL peak was observed after deposition, which may facilitate more efficient electron transfer to the HTM layer, thereby reducing carrier recombination within the perovskite.²⁶ Atomic force microscopy (AFM) revealed that CX11–CX14 could fully cover the perovskite surface, with distinct differences in the root mean square (RMS) roughness values (Fig. 3d–g). Among them, CX11 had the lowest roughness, with an RMS roughness value of 8.99 nm. Its surface was smoother and more uniform, and it had an excellent film-forming performance. In addition, we provided AFM images of the perovskite surface without HTM in Fig. S20 (SI), showing an RMS roughness value of 40.0 nm.

In order to better understand the photovoltaic performance differences among HTMs in PSCs, the defect state densities of FTO/PEDOT:PSS/perovskite/HTM/Ag devices were determined using the SCLC method.^27–29 As shown in Fig. 3h and S19(b) (SI), the fitted trap-filled limit voltages (V_TFL) for the PSC devices based on CX11–CX14 and spiro-OMeTAD are 0.24, 0.42, 0.64, 0.48, and 0.34 eV, respectively. Correspondingly, the calculated defect densities are 8.40 × 10¹⁴, 1.47 × 10¹⁵, 2.24 × 10¹⁵, 1.68 × 10¹⁵, and 1.19 × 10¹⁵ cm⁻³, respectively. The device based on CX11 has a lower defect state density, leading to reduced non-radiative recombination. This allows the electrons and holes in the perovskite layer to migrate more effectively to the electrodes, resulting in a higher V_oc.

To gain deeper insights into the surface morphological changes of the perovskite films after coating with CX11–CX14, scanning electron microscopy (SEM) measurements were carried out. As shown in Fig. 3i–m, the film surfaces became visibly smoother after the application of HTMs. Among them, CX11 exhibited a more uniform coverage and favorable film morphology, which is beneficial for suppressing interfacial recombination and minimizing charge carrier losses. PSCs were fabricated using CX11–CX14 as the hole transport layers, with the device structure shown in Fig. 4a. Based on the experimental data, the energy-level diagrams of CX11–CX14 as HTMs applied to PSC devices were obtained (Fig. 4b). The fabricated PSC devices were characterized, and the results are summarized in Fig. 4c, Table 4 and Fig. S21 (SI). Furthermore, under the same conditions, the device based on spiro-OMeTAD achieved a PCE of 22.93%. The results indicated that CX11 exhibited the highest PCE in all of the designed HTMs. As shown in Table 4 and Fig. S21 (SI), the statistical data from 20 devices reveal the following trends for the average and maximum PCE: CX11 (22.62% ± 0.27%, 23.09%) > spiro-OMeTAD (22.19% ± 0.41%, 22.93%) > CX12 (21.59% ± 0.35%, 22.07%) > CX14 (20.04% ± 0.48%, 20.61%) > CX13 (17.77% ± 0.39%, 18.58%). Among the 20 tested devices, those based on CX11–CX13 showed a narrower PCE distribution in the box plot, indicating higher reproducibility. The performance enhancement of CX11 is mainly attributed to the improvements in J_sc, V_oc, and FF. The enhanced PCE of CX11 is mainly attributed to the higher hole mobility of CX11 and stronger interface interactions at the perovskite interface. The stability evaluation was conducted on unencapsulated PSC devices under an N₂ atmosphere in a glove box (Fig. 4d). After over 1000 hours, the CX11-based PSC devices exhibited good stability by retaining approximately 93% of their initial PCE, which was higher than that of the CX12–CX14-based devices. Fig. S22 (SI) and Table S3 (SI) show the scanning J–V curves with forward and reverse scans for the CX11–CX14-based PSC devices, along with the corresponding photovoltaic parameters. The hysteresis index (HI) for the devices is presented in Table S3 (SI) in the order of CX11 (2.08%) < CX12 (2.90%) < CX14 (3.59%) < spiro-OMeTAD (4.01%) < CX13 (5.97%). The HI for the CX11-based PSC device was the lowest, indicating that by adjusting the isomerization of the dibenzofuran group, charge accumulation at the perovskite/HTM interface can be minimized.³⁰

Fig. 4 (a) Structure of the fabricated PSC devices. (b) Energy level diagram of the PSC devices. (c) Forward scan J–V curves of CX11–CX14-based PSC devices. (d) Stability of unencapsulated PSCs based on CX11–CX14.

Table 4 Experimental data of the CX11–CX14-based PSC devices in photovoltaic performance

HTMs	J sc [mA cm^−2]	V oc [V]	FF [%]	PCE [%]
a The maximum value. b The average values were obtained from 20 devices.
CX11	24.93 (24.79 ± 0.18)	1.132 (1.122 ± 0.008)	81.82 (81.32 ± 0.78)	23.09^a (22.62 ± 0.27)^b
CX12	24.81 (24.59 ± 0.23)	1.111 (1.107 ± 0.009)	80.09 (79.24 ± 0.91)	22.07 (21.59 ± 0.35)
CX13	24.60 (24.06 ± 0.46)	1.028 (1.026 ± 0.007)	73.51 (71.99 ± 1.22)	18.58 (17.77 ± 0.39)
CX14	24.73 (24.28 ± 0.42)	1.102 (1.091 ± 0.016)	75.58 (75.70 ± 1.20)	20.61 (20.04 ± 0.48)
Spiro-OMeTAD	25.01 (24.71 ± 0.30)	1.121 (1.120 ± 0.010)	81.78 (80.24 ± 1.31)	22.93 (22.19 ± 0.41)

---

Conclusions

This work investigates the influence of isomerization in HTMs on the performance of PSCs. Using computational methods such as DFT, TD-DFT, and MD calculations, the interaction modes of HTMs with the perovskite were simulated. Although the isomers have similar structures and optoelectronic properties, the introduction of the isomeric DBF group alters the conjugation delocalization and intermolecular orbital interactions, leading to changes in charge-transfer capabilities and exhibiting distinct macroscopic adsorption on perovskites. When CX11–CX14 were applied in PSC devices, the PCE based on CX11 was 23.09%, which was higher than that of the other isomers (CX12–CX14). For CX11 as an HTM in PSC application, the improvement can be attributed to its enhanced hole mobility, intermolecular interaction range, interface stacking, and ability to passivate interface defects. The experimental observations validate the reliability of the computational models and provide new insights for designing high-performance HTM candidates.

Author contributions

Xin Chen: investigation, data curation, formal analysis, validation, visualization, and writing – original draft. Jiayi Qi: validation. Xin Jiang: data curation. Fei Wu: supervision and validation. Xiaorui Liu: conceptualization, funding acquisition, project administration, resources, supervision, and writing – review and editing.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). The supplementary data related to this article contains computational details, device fabrication process, characterization methods, details of molecular synthesis, optimized geometries (Fig. S1), simulated absorption spectra (Fig. S2), perovskites/HTMs model for MD simulation (Fig. S3), synthesis route (Fig. S4), molecular structure characterization (Fig. S5–S16), cyclic voltammetry curve (Fig. S17), experimental absorption spectra (Fig. S18), hole mobility and defect state of Spiro-OMeTAD (Fig. S19), AFM image of perovskite film (Fig. S20), photovoltaic parameters (Fig. S21). J–V curves under reverse and forward scan (Fig. S22), and Table S1–S3.

Supplementary information is available. See DOI: https://doi.org/10.1039/d5me00155b.

Acknowledgements

This work was supported by the Chengdu Key Research and Development Program (Grant No. 2023-YF11-00027-HZ) and the Fundamental Research Funds for the Central Universities (Grant No. SWU-KT23009, SWU-XDJH202314).

Notes and references

1. Best Research-Cell Efficiency Chart, https://www.nrel.gov/pv/cell-efficiency.html, (accessed February 25, 2025).
2. J. Zhou, L. Tan, Y. Liu, H. Li, X. Liu, M. Li, S. Wang, Y. Zhang, C. Jiang, R. Hua, W. Tress, S. Meloni and C. Yi, Joule, 2024, 8, 1691–1706.
3. J. Zeng, Z. Liu, D. Wang, J. Wu, P. Zhu, Y. Bao, X. Guo, G. Qu, B. Hu, X. Wang, Y. Zhang, L. Yan, A. K.-Y. Jen and B. Xu, J. Am. Chem. Soc., 2025, 147, 725–733.
4. M. Zhai, K. Du, C. Liu, C. Chen, G. Li, H. Wang, Z. Xia, J. Yang, H. Xu, A. Wang, T. Matsushima, Z. Guo, M. Li, A. Abate, P. J. Dyson, M. K. Nazeeruddin and M. Cheng, ACS Energy Lett., 2025, 10, 915–924.
5. Y. Lin, X. Zhang, J. Lu, X. Lin, Y. Lu, X. Li and S. Tu, Molecules, 2024, 29, 1208.
6. J. Lee, D. H. Lee, Y. Kim, J. Y. Park, H. R. You, T. Park, M. Kim and J. Choi, Energy Environ. Mater., 2023, 6, e12408.
7. Q. Wang, Y. Cao, H. Zhao, B. Li, X. Zhang, X. Ding, G. Rahim, H. Cao, X. Liu, Y. Ding and S. Dai, Sol. RRL, 2024, 8, 2400700.
8. C. Zhao, Q. Zhang, Y. Lyu, J. Liu, F. Shen, H. Liu, H. Kong, H. Han, A. Krishna, J. Xu, H. Zhang and J. Yao, Adv. Funct. Mater., 2024, 34, 2404099.
9. X. Liu, X. Zhang, M. Han, J. Chen, G. Rahim, Y. Liang, B. Li and S. Dai, J. Mater. Chem. C, 2024, 12, 14613–14619.
10. X. Liu, X. Zhang, Y. Zhou, W. Dai, J. Chen, O. A. Syzgantseva, M. A. Syzgantseva, B. Li, R. Ghadari, M. Han, W. Du, Z. Shao, Q. Wang, S. Dai, M. K. Nazeeruddin and Y. Ding, Adv. Funct. Mater., 2025, 35, 2415179.
11. J. Qi, R. Wang, Y. Zeng, X. Gao, X. Chen, W. Shen, F. Wu, M. Li, R. He and X. Liu, Small, 2024, 2312122.
12. Y. Zhang, S. Yang, Y. Yang, J.-C. Ren, C. J. Butch, L. Wang and W. Liu, Adv. Funct. Mater., 2024, 34, 2403836.
13. P. Wang, J. Liu, W. Shang, T. Xu, M. Wang, Y. Shi, R. Cai and J. Bian, J. Phys. Chem. Lett., 2023, 14, 653–662.
14. S. You, F. T. Eickemeyer, J. Gao, J.-H. Yum, X. Zheng, D. Ren, M. Xia, R. Guo, Y. Rong, S. M. Zakeeruddin, K. Sivula, J. Tang, Z. Shen, X. Li and M. Grätzel, Nat. Energy, 2023, 8, 515–525.
15. H. Liu, Q. Chen, F. Wu and X. Liu, J. Phys. Chem. C, 2022, 126, 160–168.
16. Q. Chen, P. Cheng, H. Liu and X. Liu, Sustainable Energy Fuels, 2021, 5, 3403–3413.
17. X. Ji, T. Wang, Q. Fu, D. Liu, Z. Wu, M. Zhang, H. Y. Woo and Y. Liu, Macromol. Rapid Commun., 2023, 44, 2300213.
18. J. Qi, R. Wang, X. Chen, F. Wu, W. Shen, M. Li, R. He and X. Liu, J. Mater. Chem. A, 2024, 12, 4067–4076.
19. A. D. Becke and K. E. Edgecombe, J. Chem. Phys., 1990, 92, 5397–5403.
20. T. Lu and Q. Chen, Chem.: Methods, 2021, 1, 231–239.
21. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian∼09 revision A.01, Gaussian, Inc., Wallingford, CT, 2009.
22. T. Lu and F. Chen, J. Comput. Chem., 2012, 33, 580–592.
23. T. Lu and Q. Chen, J. Phys. Chem. A, 2023, 127, 7023–7035.
24. R. Wang, X. Chen, C. Wu, J. Qi, X. Jiang, F. Wu and X. Liu, Appl. Surf. Sci., 2023, 639, 158192.
25. R. Wang, C. Wu, J. Qi, W. Shen, F. Wu, M. Li, R. He and X. Liu, Adv. Funct. Mater., 2023, 33, 2213843.
26. X. Wang, J. Li, R. Guo, X. Yin, R. Luo, D. Guo, K. Ji, L. Dai, H. Liang, X. Jia, J. Chen, Z. Jia, Z. Shi, S. Liu, Y. Wang, Q. Zhou, T. Wang, G. Pan, P. Müller-Buschbaum, S. D. Stranks and Y. Hou, Nat. Photonics, 2024, 18, 1269–1275.
27. X. Gao, F. Wu, Y. Zeng, K. Chen, X. Liu and L. Zhu, J. Mater. Chem. C, 2023, 11, 11218–11224.
28. J. Qi, R. Wang, X. Chen, F. Wu, W. Shen, M. Li, R. He and X. Liu, J. Mater. Chem. A, 2024, 12, 4067–4076.
29. K. Galkowski, A. Mitioglu, A. Miyata, P. Plochocka, O. Portugall, G. E. Eperon, J. T.-W. Wang, T. Stergiopoulos, S. D. Stranks, H. J. Snaith and R. J. Nicholas, Energy Environ. Sci., 2016, 9, 962–970.
30. R. Wang, J. Qi, X. Chen, F. Jiang, F. Wu and X. Liu, Sol. RRL, 2024, 8, 2400453.

---