English

• In recent years, organic solar cells (OSCs) have drawn substantial attention from both the academic community and the industrial realm, owing to their remarkable flexibility, light-weight, and printability [1–7]. These characteristics endow them with exceedingly broad application prospects in the field of emerging next-generation energy technologies. At present, the power conversion efficiency (PCE) of single-junction devices based on fused-ring electron acceptors (FREAs) has exceeded 20%, marking a significant stride towards their commercialization [8–17]. Nevertheless, the synthesis of FREAs necessitates cyclization reactions to integrate multiple aromatic units into a planar conjugated framework. This intricate synthetic procedure not only escalates the synthesis complexity but also leads to relatively low yields, consequently driving up the production costs and posing a challenge to large-scale commercial application.

  Non-fused-ring electron acceptors (NFREAs) have emerged as highly promising alternative materials, primarily owing to their cost-effectiveness and straightforward synthetic routes. To date, the highest documented PCE of NFREAs has surpassed 19%, demonstrating a level of performance comparable to that of FREAs [18]. From the standpoint of optimizing the photovoltaic performance of NFREAs, molecular structure design is the pivotal strategy for enhancing their properties. Notably, the incorporation of halogen substituents and the introduction of molecular asymmetry have proven to be effective molecular modification approaches for optimizing the photovoltaic performance. These modifications can induce significant alterations in molecular structure, absorption characteristics, and energy level distributions, thereby playing a pivotal role in advancing the field of organic semiconductors [19]. In the context of this study, the introduction of asymmetric side chains and halogenation into NFREAs leads to marked changes in the spatial distribution of electron clouds and the molecular spatial conformation. Specifically, these modifications can fine-tune the electrostatic potential (ESP) distribution of molecules, strengthen intermolecular interactions, and facilitate the ordered packing of molecules within thin films. Consequently, this results in more efficient charge transport, a reduction in charge recombination losses, and ultimately an enhancement in the performance of photovoltaic devices [20–24].

  Herein, we chose a non-fused ring electron acceptor material featuring diphenylamine (DPA) side groups, which was reported by our research group in 2021, as a foundational platform for structural modification investigations [25]. Theoretically, two strategies hold promise for precisely regulating the aggregation behavior and optoelectronic properties of the materials: One involves introducing fluorine (F) atoms to establish intermolecular halogen bonds with adjacent acceptors, and the other focuses on incorporating asymmetric structures to augment molecular dipoles. Building on these theoretical insights, we synthesized two asymmetric non-fused ring acceptors. This was accomplished by introducing fluorine atoms to modify the DPA side groups and constructing asymmetric molecular architectures, with the objective of achieving precise control over the side-group molecular structure. Among these two asymmetric acceptors, the molecular polarity index (MPI) of the ortho-difluoro configuration (o-DPA) is markedly higher than that of its meta-difluoro isomer (m-DPA). This disparity suggests that relative to m-DPA, the strong electron-withdrawing nature of the ortho-difluoro atoms in o-DPA is more pronounced, thereby more readily inducing a redistribution of the electron cloud within the π-backbone. In line with our predictions, 3TT-o-2F demonstrates more compact molecular packing and a longer exciton diffusion length compared to 3TT-m-2F. Moreover, its blend films exhibit a well-defined nanoscale phase morphology, leading to reduced device energy losses. Owing to these advantageous characteristics, the device based on D18:3TT-o-2F attains an exceptional PCE of up to 17.16%, significantly surpassing that of the device utilizing D18:3TT-m-2F (15.36%). In summary, this study underscores the pivotal role of introducing F atoms and employing asymmetric strategies in the development of high-performance non-fused ring electron acceptor materials, offering valuable insights for future material design and optimization (Scheme 1).

  Scheme 1

  

  Scheme 1.  Molecular structures and synthetic routes of 3TT-m-2F and 3TT-o-2F. Reagents and conditions: (ⅰ) Pd(OAc)₂, NaOt-Bu, Xantphos, o-Xylol, 140 ℃; (ⅱ) N-bromosuccinimide (NBS), N,N-dimethylformamide (DMF), 0 ℃; (ⅲ) trimethyl(6-hexylthieno [3,2-b]thiophen-2-yl)stannane, Pd(PPh₃)₄, toluene, 110 ℃; (ⅳ) POCl₃, 1,2-dichloroethane, DMF, 80 ℃; (ⅴ) 2-(5,6-difluoro-3-oxo-2,3-dihydro-1H-inden-1-ylidene)malononitrile, acetic anhydride, toluene.

  DownLoad: Full-Size Img PowerPoint

  We conducted a systematic investigation into the absorption characteristics and aggregation behavior of two acceptor materials, 3TT-m-2F and 3TT-o-2F, in both dilute chloroform solutions and thin-film states by employing ultraviolet-visible (UV–vis) absorption spectroscopy. As depicted in Figs. 1a and b, these two acceptor materials manifest prominent and well-defined absorption peaks within the wavelength range of 600–800 nm. In dilute chloroform solutions, their maximum absorption peaks (λ_max) are precisely located at 694 and 705 nm, respectively. Upon transitioning to thin-film states, these absorption peaks exhibit marked red-shifts, reaching wavelengths of 753 and 769 nm, respectively. In comparison with their solution states, the magnitudes of red-shift are approximately 59 and 64 nm, respectively. Concurrently, within the thin-film systems, both 3TT-m-2F and 3TT-o-2F clearly display distinct 0–1 vibrational shoulder peaks, with the corresponding shoulder-peak wavelengths positioned at 690 and 700 nm, respectively. Further quantitative analyses reveal that the intensity ratio of the 0–0 to 0–1 electronic vibrational transitions for compound 3TT-o-2F is significantly higher than that for compound 3TT-m-2F. These features unequivocally demonstrate that the molecular system of 3TT-o-2F exhibits more pronounced J-aggregate structural characteristics, suggesting stronger intermolecular interactions in the thin-film state. Such enhanced intermolecular interaction is beneficial for facilitating efficient charge transport and effective exciton diffusion. The optical bandgaps (E_g^opt) of 3TT-m-2F and 3TT-o-2F were calculated using the formula E_g^opt = 1240/λ_onset, where λ_onset represents the absorption edge wavelength of the thin films. The calculated E_g^opt values are 1.47 and 1.45 eV for 3TT-m-2F and 3TT-o-2F, respectively. Simultaneously, the energy levels of these two acceptor materials were determined via cyclic voltammetry (CV), with the relevant results presented in Fig. S1 (Supporting information). The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels were computed using the formula E_HOMO/LUMO = −e(E_ox/red − E_(Fc/Fc+) + 4.8) (eV). As illustrated in Fig. 1c and Table 1, the HOMO/LUMO energy levels of 3TT-m-2F and 3TT-o-2F are −5.66/−3.94 eV and −5.55/−3.95 eV, respectively. These research findings unequivocally reveal that both the introduction position of fluorine atoms on the side-groups of non-fused-ring acceptors and the asymmetric structure of the acceptors exert a significant influence on the absorption spectral characteristics and energy-level alignment of the acceptor molecules.

  Figure 1

  

  Figure 1.  (a) UV–vis absorption spectra of 3TT-m-2F and 3TT-o-2F in chloroform solutions. (b) UV–vis absorption spectra of their corresponding thin films. (c) Energy band diagrams of 3TT-m-2F and 3TT-o-2F. (d) ESP distributions and calculated MPI values of the side groups of 3TT-m-2F and 3TT-o-2F. (e) ESP distributions and calculated MPI values for the entire molecular systems of 3TT-m-2F and 3TT-o-2F. (f) Optimal molecular conformations of 3TT-m-2F and 3TT-o-2F calculated via DFT.

  DownLoad: Full-Size Img PowerPoint

  Table 1

  

  Table 1.  Optical properties and energy levels of 3TT-m-2F and 3TT-o-2F.

  DownLoad: CSV

  Acceptor	λmax (nm) a	λmax (nm) b	λonset (nm) b	Egopt (eV)	HOMO (eV)	LUMO (eV)
  3TT-m-2F	694	753	842	1.47	−5.66	−3.94
  3TT-o-2F	705	769	853	1.45	−5.55	−3.95
  a In dilute chloroform solution. b As thin film.

  As demonstrated by the molecular ESP distribution characteristics depicted in Fig. 1d. Upon the introduction of meta-difluoro substituents into 3TT-m-2F, a cross-distribution configuration manifests. In this configuration, the positive ESP is arrayed around the molecular backbone, whereas the negative ESP is arranged around the side-groups. In marked contrast, following the incorporation of ortho-difluoro substituents to 3TT-o-2F, notable alterations occur in its ESP distribution characteristics, with the non-uniformity becoming further exacerbated. Specifically, the disparities in the distribution of positive and negative ESPs between the molecular backbone region and the side-group region are more pronounced, thereby presenting a well-defined spatial ESP distribution pattern.

  To enable a quantitative assessment of molecular polarity, we introduce the MPI as a pivotal parameter [26]. This index is capable of accurately quantifying the local polarity deviations arising from the non-uniform distribution of ESP, thereby offering an effective quantitative criterion for in-depth investigation into the characteristics of molecular polarity [27–29]. As illustrated in Fig. 1d, for the compounds 3TT-m-2F and 3TT-o-2F, the MPI values between their side-groups and molecular backbones have been experimentally determined to be 7.93 and 8.59 kcal/mol, respectively. These experimental data unequivocally demonstrate that the incorporation of ortho-difluoro substituents can substantially augment the polarity characteristics of the side-groups.

  Meanwhile, we conducted a comprehensive computation of the ESP distribution for the entire molecules (Fig. 1e and Fig. S2 in Supporting information). The results show that the side-group structure exerts a specific regulatory effect on the ESP distribution of the molecular backbone, and this effect, in turn, influences the overall electronic cloud distribution characteristics of the molecule. The MPI values for 3TT-m-2F and 3TT-o-2F are 9.20 and 9.35 kcal/mol, respectively. This result not only validates the enhancing effect of ortho-difluoro substituents on side-group polarity but also reveals that the increase in side-group polarity can be effectively transferred to the entire molecular system, facilitating the formation of a more ordered and efficient stacking for the acceptor molecules.

  As illustrated in Fig. 1f, the optimal molecular conformations of 3TT-m-2F and 3TT-o-2F were determined via density functional theory (DFT) calculations. From a top-view perspective, the dihedral angles between the indanone cyanine (IC) end-groups and the π-bridges are below 1°, signifying a near-perfect planar geometry. Nevertheless, for 3TT-m-2F and 3TT-o-2F, the twist angles between the central dithienothiophene cores and the adjacent dithienothiophene bridges measure approximately 22.2° and 18.2°, respectively. When viewed from the side (Fig. S3 in Supporting information), the side chains of both acceptor molecules exhibit relatively large dihedral angles with respect to the molecular backbones. Specifically, the dihedral angle for 3TT-m-2F is 50.5°, and for 3TT-o-2F, it is 53.8°. This structural characteristic effectively mitigates excessive molecular aggregation. DFT calculations reveal that the HOMO and LUMO energy levels of 3TT-m-2F are −5.91 and −3.65 eV, respectively, while those of 3TT-o-2F are −5.88 and −3.61 eV, respectively. These findings align with the trends observed in CV measurements. As shown in Fig. S3, the electron cloud of the LUMO is delocalized over the entire molecular backbone, whereas the electron cloud of the HOMO is predominantly localized on the electron-donating moieties.

  The photovoltaic performance of 3TT-m-2F and 3TT-o-2F was systematically evaluated in OSCs with an optimized device architecture of ITO/2PACz/active layer/PDINN/Ag. The polymer D18 was selected as the electron donor due to its complementary absorption spectrum and well-aligned energy levels with the acceptors, ensuring efficient photon harvesting and charge transfer [30,31]. The detailed device processing procedures are listed in Supporting information. The J-V curves of the devices are shown in Fig. 2a, while the photovoltaic performance parameters during the optimization process are presented in Table 2, and the optimal parameters are listed in Supporting information. The device based on 3TT-o-2F achieved a PCE of 17.16%, with an open-circuit voltage (V_oc) of 0.89 V, a J_sc of 24.16 mA/cm², and an FF of 78.72%, significantly outperforming the device based on 3TT-m-2F, which achieved a PCE of 15.36%. To elucidate the physical mechanism underlying the enhancement in J_sc, this study systematically evaluated the carrier collection efficiency of the 3TT-m-2F and 3TT-o-2F devices using external quantum efficiency (EQE) spectroscopy. As depicted in Fig. 2b, the device based on the ortho-fluorinated 3TT-o-2F maintained an EQE peak response of 84.6% across a broad spectral range of 300–850 nm, demonstrating a wider spectral coverage and higher response intensity compared to the meta-substituted 3TT-m-2F system. This further explains the high J_sc observed in the 3TT-o-2F device. The current densities calculated by integrating the EQE curves for the two devices reached 22.20 and 23.21 mA/cm², respectively.

  Figure 2

  

  Figure 2.  Performance and photoelectric characteristics of the 3TT-m-2F and 3TT-o-2F based devices, (a) J-V curves, (b) EQE spectra, (c) EQE_EL curves, (d) dependence of J_ph on V_eff and (e) dependence of V_oc on P_light. (f) The detailed E_loss of OSCs.

  DownLoad: Full-Size Img PowerPoint

  Table 2

  

  Table 2.  Photovoltaic parameters of devices based on 3TT-m-2F and 3TT-o-2F.

  DownLoad: CSV

  Sample	Voc (V)	Jsc (mA/cm2)a	FF (%)	PCE (%)b
  D18:3TT-m-2F	0.87	23.19 (22.20)	75.04	15.36 (14.91 ± 0.39)
  D18:3TT-o-2F	0.89	24.16 (23.21)	78.72	17.16 (16.99 ± 0.14)
  a Integrated from EQE curves. b Average of PCE.

  We investigated the charge carrier recombination, exciton dissociation, and charge collection in devices based on the two acceptor molecules. Firstly, to elucidate the mechanisms of exciton dissociation and charge collection in OSCs, we evaluated the photocurrent density (J_ph) and effective voltage (V_eff), and calculated the exciton dissociation efficiency (P_diss) using the formula P_diss = J_ph/J_sat and the charge collection efficiency (P_coll) using the formula P_coll = J_sc/J_sat [32,33]. The results revealed that the P_diss/P_coll values for devices based on 3TT-m-2F and 3TT-o-2F are 96.2%/84.1% and 97.1%/84.7%, respectively (Fig. 2d). This indicates that the exciton dissociation and charge collection efficiencies in the device based on 3TT-o-2F are higher than those in the device based on 3TT-m-2F, providing further explanation for the higher J_sc and FF values.

  Additionally, we studied the relationships between J_sc and V_oc, as well as between J_sc and light intensity (P_light), to assess the charge carrier recombination kinetics within the devices. As shown in Fig. S4 (Supporting information), the relationship between J_sc and P_light is expressed by the formula J_sc ∝ P_light^α, where the proximity of the α value to 1 is closely related to the recombination mechanism; an α value close to 1 indicates that the bimolecular recombination process in the device is negligible [34,35]. The α values for devices based on 3TT-m-2F and 3TT-o-2F are 0.989 and 0.992, respectively. The α value for the device based on 3TT-o-2F is closer to the theoretical ideal value, confirming the correlation between the fluorine atom substitution site and the suppression of bimolecular recombination. The ideality factor (n) of the relationship between V_oc and P_light is explained by the slope kT/q (where k represents the Boltzmann constant, T denotes the Kelvin temperature, and q represents the elementary charge) [5,36,37]. The n values for devices based on 3TT-m-2F and 3TT-o-2F were 1.03 and 1.01 kT/q, respectively. As shown in Fig. 2e, the device based on 3TT-o-2F is closer to 1, indicating that trap-assisted recombination has been effectively suppressed in the device based on 3TT-o-2F compared to that based on 3TT-m-2F. These findings suggest that the device based on 3TT-o-2F experiences less bimolecular recombination loss and exhibits more efficient charge carrier transport.

  The energy loss (E_loss) can be calculated using the formula E_loss = E_g - qV_oc = ΔE₁ + ΔE_rad + ΔE_non-rad, where E_g represents the energy band corresponding to the intersection point of the thin-film UV–vis and photoluminescence spectra [38,39]. To delve deeper into the reasons for these E_loss differences, we conducted highly sensitive photovoltaic quantum efficiency (EQE_PV) and electroluminescence (EL) measurements to determine the energy of the charge transfer state (E_CT). As shown in Fig. 2c and Fig. S5 (Supporting information), the E_CT values for devices based on D18:3TT-m-2F and 3TT-o-2F are 1.43 and 1.41 eV, respectively. Furthermore, we performed EQE_EL measurements to evaluate ΔE_non-rad in the OSCs. According to the formula: ΔE_non-rad = −kTln(EQE_EL)/q, where k, T, and q represent the Boltzmann constant, Kelvin temperature, and elementary charge, respectively [40]. The calculated ΔE_non-rad values for devices based on D18:3TT-m-2F and 3TT-o-2F were 0.34 and 0.29 eV, respectively. As shown in Fig. 2f and Table 3, the E_loss values for devices based on 3TT-m-2F and 3TT-o-2F were 0.66 and 0.61 eV, respectively.

  Table 3

  

  Table 3.  The E_loss of 3TT-m-2F and 3TT-o-2F-based devices.

  DownLoad: CSV

  Sample	Eg (eV)	Ect (eV)	Eloss (eV)	ΔE1 (eV)	ΔErad (eV)	ΔEnon-rad (eV)
  D18:3TT-m-2F	1.53	1.43	0.66	0.22	0.10	0.34
  D18:3TT-o-2F	1.50	1.41	0.61	0.23	0.09	0.29

  We utilized the space-charge-limited current (SCLC) technique to systematically investigate the charge transport characteristics of devices based on D18:3TT-m-2F and D18:3TT-o-2F. Specifically, hole mobility (μ_h) was measured employing an ITO/PEDOT: PSS/active layer/Au device architecture, while electron mobility (μ_e) was determined utilizing an ITO/zinc oxide (ZnO)/active layer/Al configuration. As depicted in Fig. S6 and Table S1 (Supporting information), the μ_e values for devices based on D18:3TT-m-2F and D18:3TT-o-2F are 4.44 × 10⁻⁴ and 6.06 × 10⁻⁴ cm² V⁻¹ s⁻¹, respectively. Additionally, the μ_h values for devices based on D18:3TT-m-2F and D18:3TT-o-2F are 5.33 × 10⁻⁴ and 5.61 × 10⁻⁴ cm² V⁻¹ s⁻¹, respectively. Compared to the device based on D18:3TT-m-2F, the device based on D18:3TT-o-2F exhibits higher μ_h and μ_e values and a more balanced μ_h/μ_e ratio, which further elucidates the superior J_sc and FF values observed in the device based on D18:3TT-o-2F.

  We have carried out an in-depth exploration of hole transport dynamics within blend films, employing femtosecond transient absorption (fs-TA) spectroscopy techniques. During the experimental procedure, excitation was achieved using a 750 nm pump laser, and the dynamic responses across the 520–850 nm spectral range were visualized through two-dimensional color mapping diagrams (Fig. 3). Additionally, transient spectra at characteristic delay time points were deliberately selected for comparative analysis. The results demonstrate that both blend systems, namely D18:3TT-o-2F and D18:3TT-m-2F, exhibit prominent ground-state bleaching (GSB) characteristic signals. As depicted in Fig. 3 and Fig. S7 (Supporting information), two new bleaching peaks emerge at 550 and 600 nm within the donor absorption region. Furthermore, the peak intensity of the D18:3TT-o-2F blend film surpasses that of the D18:3TT-m-2F blend film, indicating a more efficient hole transport process in the D18:3TT-o-2F blend film. To facilitate a quantitative analysis of the hole transport process, we extracted kinetic traces at specific wavelengths. By applying a double-exponential fitting algorithm to the rising phase of the donor GSB signal, we derived the kinetic time (Table S2 in Supporting information). The time constants for the different blend films are as follows: For D18:3TT-m-2F, τ₁ = 3.31 ± 0.72 ps and τ₂ = 33.11 ± 5.71 ps; for D18:3TT-o-2F, τ₁ = 1.03 ± 0.19 ps and τ₂ = 31.52 ± 1.91 ps. Here, the faster τ₁ predominantly corresponds to the ultrafast exciton dissociation process occurring at the donor-acceptor interface, while the slower τ₂ mainly reflects the exciton diffusion process within the acceptor domains [19,41]. A comparative analysis reveals that, in contrast to the D18:3TT-m-2F blend film, the D18:3TT-o-2F blend film exhibits a faster exciton dissociation rate at the interface and more efficient intra-domain exciton diffusion. This unequivocally demonstrates its superior hole transport rate. Such a characteristic is highly favorable for the generation and collection of charges in OSCs, ultimately conferring them with exceptional optoelectronic performance.

  Figure 3

  

  Figure 3.  The TA image and various decay times of D18:3TT-m-2F (a, b), D18:3TT-o-2F (d, e); Singlet exciton decay dynamics with different excitation fluence for neat films of 3TT-m-2F (c) and 3TT-o-2F (f).

  DownLoad: Full-Size Img PowerPoint

  In addition, we further explored the impact of exciton diffusion on charge transport performance. As shown in Figs. 3c and f, by combining fs-TA spectroscopy measurements with the exciton-exciton annihilation (EEA) model, we estimated the exciton diffusion length (L_D) [29,42,43]. According to the EEA model, excitons primarily undergo two quenching mechanisms: Defect-assisted recombination and bimolecular exciton-exciton annihilation recombination. These two processes jointly determine the diffusion coefficient (D) and further influence the exciton diffusion length. We selected 750 nm as the excitation wavelength and investigated the dynamic evolution of excitons by adjusting the excitation intensity. The results revealed that an increase in excitation fluence accelerated the decay rate of excitons, and the overall evolution process could be well-fitted into a rate equation model. By integrating experimental results with theoretical analysis, we obtained L_D values of 13.40 and 17.10 nm for 3TT-m-2F and 3TT-o-2F, respectively (Table S3 in Supporting information). It is evident that 3TT-o-2F exhibits a larger diffusion length, indicating superior performance in charge transport.

  We further employed grazing-incidence wide-angle X-ray scattering (GIWAXS) to investigate the packing and orientation of the neat acceptor films and blend films. After signal normalization, as illustrated in Fig. 4a, the 3TT-o-2F thin film exhibits significantly higher scattering intensity in the out-of-plane (q_z) direction compared to the 3TT-m-2F thin film. As can be observed from both the 1D profiles and 2D images, both acceptor molecules display distinct (010) peaks in the out-of-plane direction, indicating a predominant in-plane orientation. As shown in Fig. 4b, the (010) π-π stacking diffraction peaks for the neat 3TT-m-2F and 3TT-o-2F thin films are located at 1.70 and 1.72 Å⁻¹, respectively. Consequently, 3TT-o-2F demonstrates a π-π stacking distance of 3.65 Å, which is slightly shorter than that of 3TT-m-2F (3.69 Å). Additionally, both the D18:3TT-m-2F and D18:3TT-o-2F blend films exhibit a predominantly in-plane molecular orientation with intense (010) diffraction peaks. Furthermore, the crystal coherence lengths (CCLs) [44] are calculated to be 19.27 and 25.40 Å for D18:3TT-m-2F and D18:3TT-o-2F blend films, respectively (Table S4 in Supporting information). Compared to 3TT-m-2F, 3TT-o-2F exhibits a shorter intermolecular π-π stacking distance, along with a significantly increased CCL in its blend film. This optimized molecular ordering characteristic reflects its strong crystallinity, which in turn leads to higher carrier mobility in the device and provides a structural basis for efficient charge transport.

  Figure 4

  

  Figure 4.  (a) 2D GIWAXS patterns and (b) 1D scattering profiles of the neat 3TT-m-2F and 3TT-o-2F films and the corresponding blend films.

  DownLoad: Full-Size Img PowerPoint

  To gain a more profound understanding of the photovoltaic performance exhibited by different acceptors, we utilized transmission electron microscopy (TEM) and atomic force microscopy (AFM) to conduct in-depth characterizations of the bulk microstructure and surface morphology of the blend films. The corresponding results are presented in Fig. 5. As shown in Figs. 5a and b, in comparison with the coarse fiber structure of the D18:3TT-m-2F blend film, the D18:3TT-o-2F blend film demonstrates more homogeneous morphological features, characterized by a finer and more densely packed nanofiber network. Furthermore, as illustrated in Figs. 5c and d, through high-precision measurement, the surface roughness of the D18:3TT-o-2F blend film is measured to be 1.03 nm, which is significantly lower than the 1.14 nm measured for the D18:3TT-m-2F blend film. Simultaneously, the outcomes of atomic force microscopy phase imaging, as shown in Figs. 5e and f, indicate the existence of fine and closely-arrayed fibrous crystallites within the D18:3TT-o-2F blend film. This finding strongly corroborates the observations made via TEM. The D18:3TT-o-2F blend film showcases a more optimized bicontinuous network structure. Such a structure is highly conducive to the efficient transport of charges, ultimately conferring upon it exceptional photovoltaic performance.

  Figure 5

  

  Figure 5.  TEM and AFM images of D18:3TT-m-2F (a, c, e), D18:3TT-o-2F blend films (b, d, f).

  DownLoad: Full-Size Img PowerPoint

  In conclusion, we have successfully designed and synthesized two asymmetric NFREA materials, designated as 3TT-m-2F and 3TT-o-2F, respectively. These materials are structural isomers resulting from the introduction of two fluorine atoms at different substitution positions on the side chains. The adoption of asymmetric molecular architectures and side-group fluorination strategies exerts a profound influence on the electronic characteristics and aggregation behaviors of these materials. Specifically, the MPI of the ortho-difluoro configuration is markedly higher than that of its meta-difluoro counterpart. Compared with 3TT-m-2F, 3TT-o-2F exhibits distinct features, including a redshift in the absorption spectrum, enhanced molecular packing density, and an extended exciton diffusion length. Moreover, the D18:3TT-o-2F blend film exhibits a well-defined nanofiber network morphology. This unique structure significantly promotes efficient charge migration, thus enabling the achievement of a high charge mobility. As a result, the photovoltaic device based on D18:3TT-o-2F achieves an impressive PCE of 17.16%, along with a V_oc of 0.89 V, a J_sc of 24.16 mA/cm², and an FF of 78.72%. In sharp contrast, the device based on 3TT-m-2F has a relatively lower PCE of 15.36%. In essence, this study has significantly elevated the photovoltaic performance of acceptor materials through the strategic introduction of fluorine atoms into the side chains and the adoption of asymmetric molecular designs. The findings offer highly valuable references and insights for the rational design and optimization of high-performance NFREA materials in the field of OSCs.

  CRediT authorship contribution statement

  Dawei Li: Writing – original draft, Methodology, Investigation, Data curation. Haoming Song: Methodology, Investigation, Formal analysis. Yibo Shi: Software, Data curation. Guangliu Ran: Methodology, Formal analysis, Data curation. Cuifen Zhang: Software, Methodology, Formal analysis, Data curation. Andong Zhang: Validation, Software, Resources. Ya-Nan Chen: Formal analysis, Data curation. Gendi Zhang: Methodology, Formal analysis. Zaifei Ma: Formal analysis, Data curation. Wenkai Zhang: Investigation, Data curation, Conceptualization. Cuihong Li: Writing – review & editing, Supervision, Project administration. Yahui Liu: Project administration, Methodology, Investigation. Hao Lu: Validation, Resources, Investigation. Zhishan Bo: Writing – review & editing, Supervision, Project administration, Investigation.

  Declaration of competing 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.

  Acknowledgments

  Financial support from the National Natural Science Foundation of China (Nos. 22375024, 21975031, 52433007, and 52173174), the Natural Science Foundation of Shandong Province (No. ZR2022YQ45), the Taishan Scholars Program (Nos. tstp20221121 and tsqnz20221134), The Beijing Natural Science Foundation (No. 2244073), Supported by State Key Laboratory of Bio-Fibers and Eco-Textiles (Qing-dao University) (No. RZ2200002821) is acknowledged.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2026.112438.

  
  
• 1. [1]

     Y. Liu, B. Li, C. Ma, et al., Sci. China Chem. 65 (2022) 224–268. doi: 10.1007/s11426-021-1180-6

  2. [2]

     W. Lai, S. Karuthedath, C.Y. Xiao, et al., Chin. Chem. Lett. 35 (2024) 108287. doi: 10.1016/j.cclet.2023.108287

  3. [3]

     Q. Bei, B. Zhang, K. Wang, et al., Chin. Chem. Lett. 35 (2024) 108438. doi: 10.1016/j.cclet.2023.108438

  4. [4]

     H. Wang, H. Lu, Y.N. Chen, et al., Adv. Mater. 34 (2022) 2105483. doi: 10.1002/adma.202105483

  5. [5]

     H. Lu, H. Wang, G. Ran, et al., CCS Chem. 6 (2024) 1757–1766. doi: 10.31635/ccschem.023.202303239

  6. [6]

     Z. Xu, T. Xue, Q. Guo, et al., Inf. Funct. Mater. 2 (2025) 1–39.

  7. [7]

     Y. Guo, X. Ma, H. Song, et al., Adv. Funct. Mater. 32 (2025) 2416356.

  8. [8]

     M. Du, N. Sun, H. Cheng, et al., Angew. Chem. Int. Ed. 64 (2025) e202515114. doi: 10.1002/anie.202515114

  9. [9]

     H. Chen, Y. Huang, R. Zhang, et al., Nat. Mater. 24 (2025) 444–453. doi: 10.1038/s41563-024-02062-0

  10. [10]

      C. Li, J. Song, H. Lai, et al., Nat. Mater. 24 (2025) 433–443. doi: 10.1038/s41563-024-02087-5

  11. [11]

      R. Ma, Z. Luo, Y. Zhang, et al., Sci. China Mater. 68 (2025) 1689–1701. doi: 10.1007/s40843-025-3366-9

  12. [12]

      H. Lu, D. Li, W. Liu, et al., Angew. Chem. Int. Ed. 63 (2024) e202407007. doi: 10.1002/anie.202407007

  13. [13]

      R. Zeng, F. Han, W. Zhong, et al., Adv. Mater. 37 (2025) 202501812.

  14. [14]

      C. Chen, L. Wang, W. Xia, et al., Nat. Commun. 15 (2024) 6865. doi: 10.1038/s41467-024-51359-w

  15. [15]

      J. Zhang, W. Wei, Z. Luo, et al., Angew. Chem. Int. Ed. 64 (2025) 202512237. doi: 10.1002/anie.202512237

  16. [16]

      Z. Zhao, S. Chung, L. Bai, et al., Adv. Funct. Mater. 35 (2025) 202506593.

  17. [17]

      H. Fang, C. Xiao, S. Liang, et al., Angew. Chem. Int. Ed. 64 (2025) e202514735. doi: 10.1002/anie.202514735

  18. [18]

      R. Zeng, M. Zhang, X. Wang, et al., Nat. Energy 9 (2024) 1117–1128.

  19. [19]

      D. Li, H. Zhang, X. Cui, et al., Adv. Mater. 36 (2024) 2310362. doi: 10.1002/adma.202310362

  20. [20]

      X. Wang, X. Shen, J. Wang, et al., Energy Environ. Sci. 18 (2025) 5586–5598. doi: 10.1039/d5ee00455a

  21. [21]

      J. Liu, W. Zheng, G. Han, et al., Sol. RRL 5 (2021) 202000780.

  22. [22]

      W. Hu, X. Du, W. Zhuang, et al., ACS Appl. Energy Mater. 3 (2020) 5734–5744. doi: 10.1021/acsaem.0c00689

  23. [23]

      Y. Sun, C. Chen, L. Wang, et al., ACS Mater. Lett. 5 (2023) 1710–1717. doi: 10.1021/acsmaterialslett.3c00331

  24. [24]

      R. Sewak, K. Singh, A. Mondal, ACS Appl. Energy Mater. 8 (2025) 11456–11467. doi: 10.1021/acsaem.5c01643

  25. [25]

      X. Wang, H. Lu, Y. Liu, et al., Adv. Energy Mater. 11 (2021) 202102591.

  26. [26]

      Y. Cui, P. Zhu, X. Shi, et al., J. Phys. Chem. C 125 (2021) 10250–10259. doi: 10.1021/acs.jpcc.1c01488

  27. [27]

      Z. Chen, S. Zhang, J. Ren, et al., Adv. Mater. 36 (2024) 202310390.

  28. [28]

      X. Liao, H. Pei, H. Zhao, et al., Chem. Eng. J. 435 (2022) 135020. doi: 10.1016/j.cej.2022.135020

  29. [29]

      R. Xu, Y. Jiang, H. Lei, F, et al., Sci. Adv. 11 (2025) eadt6024. doi: 10.1126/sciadv.adt6024

  30. [30]

      X. Meng, K. Jin, Z. Xiao, et al., J. Semicond. 42 (2021) 100501. doi: 10.1088/1674-4926/42/10/100501

  31. [31]

      H. Lu, W. Liu, G. Ran, et al., Angew. Chem. Int. Ed. 62 (2023) e202314420. doi: 10.1002/anie.202314420

  32. [32]

      G. Chai, Y. Chang, J. Zhang, et al., Energy Environ. Sci. 14 (2021) 3469–3479. doi: 10.1039/d0ee03506h

  33. [33]

      X. Kong, C. Zhu, J. Zhang, et al., Energy Environ. Sci. 15 (2022) 2011–2020. doi: 10.1039/d2ee00430e

  34. [34]

      Y.N. Chen, M. Li, Y. Wang, et al., Angew. Chem. Int. Ed. 59 (2020) 22714–22720. doi: 10.1002/anie.202010856

  35. [35]

      W. Ma, X. He, T. Chen, et al., Chin. Chem. Lett. 35 (2024) 109099. doi: 10.1016/j.cclet.2023.109099

  36. [36]

      Y. Li, Y. Zhang, B. Wu, et al., Sol. RRL 6 (2022) 00073.

  37. [37]

      L. Xu, Y. Gao, W. Wang, et al., Energy Environ. Sci. 16 (2023) 3942–3950. doi: 10.1039/d3ee01317k

  38. [38]

      G. Kim, C. Sun, D. Lee, et al., Adv. Funct. Mater. 32 (2022) 202201150.

  39. [39]

      D. Li, H. Lu, Y. Chen, et al., Chem. Mater. 34 (2022) 8840–8848. doi: 10.1021/acs.chemmater.2c02146

  40. [40]

      Z. Zhang, D. Li, H. Zhang, et al., J. Mater. Chem. C 10 (2022) 2792–2799. doi: 10.1039/d1tc04425g

  41. [41]

      K. Liu, Y. Jiang, F. Liu, et al., Adv. Mater. 37 (2024) 202413376.

  42. [42]

      P. Bi, S. Zhang, Z. Chen, et al., Joule 5 (2021) 2408–2419. doi: 10.1016/j.joule.2021.06.020

  43. [43]

      H. Cha, S. Wheeler, S. Holliday, et al., Adv. Funct. Mater. 28 (2018) 01704389. doi: 10.1002/adfm.201704389

  44. [44]

      H. Zhuo, X. Li, J. Zhang, et al., Angew. Chem. Int. Ed. 62 (2023) 202303551. doi: 10.1002/anie.202303551

• 

  Scheme 1  Molecular structures and synthetic routes of 3TT-m-2F and 3TT-o-2F. Reagents and conditions: (ⅰ) Pd(OAc)₂, NaOt-Bu, Xantphos, o-Xylol, 140 ℃; (ⅱ) N-bromosuccinimide (NBS), N,N-dimethylformamide (DMF), 0 ℃; (ⅲ) trimethyl(6-hexylthieno [3,2-b]thiophen-2-yl)stannane, Pd(PPh₃)₄, toluene, 110 ℃; (ⅳ) POCl₃, 1,2-dichloroethane, DMF, 80 ℃; (ⅴ) 2-(5,6-difluoro-3-oxo-2,3-dihydro-1H-inden-1-ylidene)malononitrile, acetic anhydride, toluene.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  (a) UV–vis absorption spectra of 3TT-m-2F and 3TT-o-2F in chloroform solutions. (b) UV–vis absorption spectra of their corresponding thin films. (c) Energy band diagrams of 3TT-m-2F and 3TT-o-2F. (d) ESP distributions and calculated MPI values of the side groups of 3TT-m-2F and 3TT-o-2F. (e) ESP distributions and calculated MPI values for the entire molecular systems of 3TT-m-2F and 3TT-o-2F. (f) Optimal molecular conformations of 3TT-m-2F and 3TT-o-2F calculated via DFT.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Performance and photoelectric characteristics of the 3TT-m-2F and 3TT-o-2F based devices, (a) J-V curves, (b) EQE spectra, (c) EQE_EL curves, (d) dependence of J_ph on V_eff and (e) dependence of V_oc on P_light. (f) The detailed E_loss of OSCs.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The TA image and various decay times of D18:3TT-m-2F (a, b), D18:3TT-o-2F (d, e); Singlet exciton decay dynamics with different excitation fluence for neat films of 3TT-m-2F (c) and 3TT-o-2F (f).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) 2D GIWAXS patterns and (b) 1D scattering profiles of the neat 3TT-m-2F and 3TT-o-2F films and the corresponding blend films.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  TEM and AFM images of D18:3TT-m-2F (a, c, e), D18:3TT-o-2F blend films (b, d, f).

  下载: 全尺寸图片 幻灯片

  Table 1.  Optical properties and energy levels of 3TT-m-2F and 3TT-o-2F.

  Acceptor	λmax (nm) a	λmax (nm) b	λonset (nm) b	Egopt (eV)	HOMO (eV)	LUMO (eV)
  3TT-m-2F	694	753	842	1.47	−5.66	−3.94
  3TT-o-2F	705	769	853	1.45	−5.55	−3.95
  a In dilute chloroform solution. b As thin film.

  下载: 导出CSV

  Table 2.  Photovoltaic parameters of devices based on 3TT-m-2F and 3TT-o-2F.

  Sample	Voc (V)	Jsc (mA/cm2)a	FF (%)	PCE (%)b
  D18:3TT-m-2F	0.87	23.19 (22.20)	75.04	15.36 (14.91 ± 0.39)
  D18:3TT-o-2F	0.89	24.16 (23.21)	78.72	17.16 (16.99 ± 0.14)
  a Integrated from EQE curves. b Average of PCE.

  下载: 导出CSV

  Table 3.  The E_loss of 3TT-m-2F and 3TT-o-2F-based devices.

  Sample	Eg (eV)	Ect (eV)	Eloss (eV)	ΔE1 (eV)	ΔErad (eV)	ΔEnon-rad (eV)
  D18:3TT-m-2F	1.53	1.43	0.66	0.22	0.10	0.34
  D18:3TT-o-2F	1.50	1.41	0.61	0.23	0.09	0.29

  下载: 导出CSV
•