Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Advertisement
Communications Materials volume 7, Article number: 79 (2026)
Achieving both low voltage loss and efficient charge generation remains a major challenge in advancing high-performance organic photovoltaics (OPVs). Here, we show that photovoltaic cells using PTNT1-F—a dithienonaphthobisthiadiazole (TNT)-based polymer recently developed by our group—exhibit a notably low nonradiative voltage loss (∆Vnr) of 0.18 V, suggesting a minimal driving force for charge generation. Remarkably, when combined with a nonfullerene acceptor Y12, the PTNT1-F device achieved high photocurrents and charge generation efficiencies exceeding 80% of the theoretical limit—to the best of our knowledge, the highest reported for OPVs with similarly low ∆Vnr. PTNT1-F features a rigid, ordered backbone that preserves the density of states (DOS) upon blending with the acceptor. In contrast, reference polymers such as D18 and PM6 show significant DOS changes, emphasizing advantage of PTNT1-F in promoting hole delocalization and efficient charge dissociation even with limited driving force. These results offer valuable insights into designing polymer donors for simultaneous low voltage loss and efficient charge generation.
Organic photovoltaics (OPVs) have made remarkable progress in the past two decades, primarily owing to advances in π-conjugated materials used as donors and acceptors. As a result, power conversion efficiencies (PCEs) have exceeded 20% in state-of-the-art devices1,2,3,4. Despite these improvements, OPVs still lag behind inorganic photovoltaics, largely due to the relatively large voltage loss (Vloss), leading to low open-circuit voltage (VOC)5. In particular, the voltage loss due to nonradiative recombination (∆Vnr) is particularly pronounced compared to inorganic photovoltaics, making its reduction critical for further performance enhancement6,7,8,9. Recent studies have shown that the ∆Vnr can be suppressed by minimizing the energy offset between the locally excited (LE) state and the interfacial charge transfer (CT) state. This offset roughly corresponds to the difference in the lowest unoccupied molecular orbital (LUMO) energy levels or the highest occupied molecular orbital (HOMO) energy levels between the donor and acceptor9,10,11. However, since these energy offsets serve as the driving force for charge generation, reducing them inevitably results in lower charge generation efficiency and thereby lower photocurrent12,13. Hence, overcoming this inherent trade-off remains a fundamental challenge for OPVs.
While some polymer:fullerene systems have demonstrated relatively small ∆Vnr values of around 0.25 V9,14,15,16,17, their external quantum efficiencies (EQEPVs) are typically limited to approximately 60%. The emergence of nonfullerene acceptors (NFAs), such as Y6 and its derivatives, has enabled EQEPV values of more than 80% with a similarly small ∆Vnr (0.23–0.25 V)18,19,20, suggesting progress toward resolving the trade-off21. Furthermore, several reports have claimed efficient charge generation even with negligible energy offsets, implying that the trade-off may no longer be universal22,23,24. However, a survey of more recently reported polymer:NFA blend systems reveals that the trade-off still persists—charge generation efficiency tends to decline as ∆Vnr decreases, particularly when ∆Vnr falls below 0.2 V, which will be discussed later in this paper.
One promising strategy to address this challenge is to enhance charge delocalization, which can mitigate Coulombic attraction between holes and electrons at the donor and acceptor (D/A) interface25. Although the importance of charge delocalization as well as exciton delocalization in NFAs in facilitating charge generation has been discussed25,26, charge delocalization in π-conjugated polymers as donors—despite being equally critical—has received far less attention. Recently, delocalization of excitons in polymer donors has been discussed as a pathway to enhance charge generation27. However, a more critical factor would be enabling efficient dissociation of excitons on the acceptor under small ∆Vnr conditions (under small driving force), i.e., enabling efficient delocalization of holes transferred to the polymer donor from the acceptor.
In π-conjugated polymers, the charge delocalization can be improved by increasing the π-conjugation length and the π–π interaction28. Therefore, extending the heteroaromatic fused rings in the polymer backbone is a viable design strategy. Recently, we reported that a dithienonaphthobisthiadiazole (TNT)-based π-conjugated polymer, named PTNT1-F (Fig. 1a) demonstrates reasonably high PCEs of more than 17% when blended with Y6, one of the high-performance NFAs. Notably, the PTNT1-F:Y6 cell exhibited a VOC of 0.86 V, which is relatively high for the Y6-based cells and is due to the relatively deep HOMO energy level of the polymer. Furthermore, PTNT1-F features a highly rigid polymer backbone derived from the π-extended TNT unit, which potentially facilitates π-electron delocalization. The result motivated us to explore the potential of PTNT1-F further.
a Chemical structures of PTNT1-F, D18 and PM6 b chemical structure of Y12 (EH = 2-ethylhexyl, BO = 2-butyloctyl, HD = 2-hexyldecyl). c Energy diagrams of the materials, where HOMO and LUMO energy levels (EHOMO and ELUMO) were determined by the cyclic voltammetry. EHOMO determined by the PYS measurement was also shown in parentheses. d UV–vis–NIR absorption spectra of the materials in thin films.
Here, we demonstrate that PTNT1-F when blended with Y12 (Fig. 1b), a derivative of Y6, achieves PCEs of more than 18% with an increased VOC of 0.87 V. Remarkably, PTNT1-F exhibits a small ∆Vnr of 0.18 V while maintaining higher charge generation efficiency than benchmark polymers, such as PM6 and D18 (Fig. 1a) under comparable similar ∆Vnr conditions. More importantly, we attribute the unconventional photovoltaic performance of PTNT1-F to its low energetic disorder, specifically, a narrow density of states (DOS) distribution in the blend. This originates from its rigid π-conjugated backbone and ensure π-electron delocalization along the polymer backbone, which facilitates efficient charge separation even under limited driving force. These results provide new insights into designing high-performance π-conjugated polymers capable of overcoming the inherent trade-off in OPVs.
PTNT1-F was synthesized by a polycondensation of dibrominated dithienyl-TNT and a distannylated benzodithiophene derivative via the Migita–Kosugi–Stille cross-coupling reaction as reported previously (Supplementary Scheme 1)29. The number-average molecular weights (Mn) and a dispersity (Đ) of PTNT1-F were 41,300 and 2.1, respectively (Supplementary Fig. 1 and Supplementary Table 1). The HOMO and LUMO energy levels (EHOMO and ELUMO, respectively) of PTNT1-F evaluated by cyclic voltammetry (Supplementary Fig. 2) were −5.50 and −3.18 eV, respectively, which were deeper than those of D18 (EHOMO = −5.46 eV, ELUMO = −3.08 eV) and PM6 (EHOMO = −5.43 eV, ELUMO = −3.20 eV) (Fig. 1c). In this study, we used Y12, a derivative of Y6, as the optimal acceptor for PTNT1-F. The EHOMO for Y12 was determined to be −5.69 eV, which resulted in the HOMO energy offset (∆EHOMO) values were 0.19, 0.23, and 0.26 eV for the blends of PTNT1-F, D18, and PM6, respectively. We also carried out the photoemission yield spectroscopy (PYS) (Supplementary Fig. 3), and observed a similar trend in EHOMO: the EHOMO and ∆EHOMO were −5.38 and 0.48 eV for PTNT1-F, −5.33 and 0.54 eV for D18, and −5.28 and 0.58 eV for PM6 (Fig. 1d), respectively. Thus, the PTNT1-F:Y12 blend had the smaller ∆EHOMO than the D18:Y12 and PM6:Y12 blends. Reflecting its smaller ∆EHOMO, the PTNT1-F:Y12 blend showed a slightly lower photoluminescence (PL) quenching efficiency (91%) compared with the D18:Y12 (93%) and PM6:Y12 (94%) blends (Supplementary Fig. 4). Nevertheless, the PL quenching efficiency of more than 90% for the PTNT1-F:Y12 blend indicates the efficient charge transfer from Y12 to PTNT1-F. The UV–vis–NIR absorption spectra of the polymers and Y12 in the thin film are shown in Fig. 1e. PTNT1-F exhibited main absorption at around 500–700 nm with the maximum (λmax) at 629 nm and the onset (λonset) at 704 nm, which were red-shifted compared to those of D18 (λmax = 578 nm, λonset = 631 nm) and PM6 (λmax = 612 nm, λonset = 680 nm). Accordingly, the optical bandgap (Egopt) of PTNT1-F, determined by the λonset, was 1.76 eV, which was smaller than that of D18 (1.97 eV) and PM6 (1.82 eV). The differences in the Egopt and energy levels between the three polymers were consistent with DFT calculations at the B3LYP/6-31 g(d) level (Supplementary Fig. 5).
Photophysical properties of polymers were evaluated by time-resolved PL measurements (Table 1 and Supplementary Fig. 6). The PL decay of PTNT1-F was slower than that of D18 and PM6, resulting in a longer PL lifetime (τf) of PTNT1-F (0.682 ns) than that of D18 (0.309 ns) and PM6 (0.205 ns). This τf value was also comparable to that for Y12 (0.608 ns) (Table S2). Furthermore, PTNT1-F exhibited a higher PL quantum yield (ΦPL) (0.055) than D18 (0.024) and PM6 (0.007), indicating its superior emissive property. Such enhanced PL can be beneficial for reducing the ∆Vnr in corresponding devices. The rate constants of the radiative and nonradiative decays (kr and knr) are also summarized in Table 1. The kr of PTNT1-F (8.0 × 107 s−1) was comparable that of D18 (7.8 × 107 s−1) but higher than that of PM6 (3.6 × 107 s−1). In contrast, the knr of PTNT1-F (1.4 × 109 s−1) was substantially lower than those of D18 (3.2 × 109 s−1) and PM6 (4.8 × 109 s−1), which is likely attributed to its rigid polymer backbone that effectively restricts nonradiative decay pathways. Thus, the suppressed knr mainly accounts for the prolonged τf and elevated ΦPL observed in PTNT1-F.
The OPV cells were fabricated using a conventional architecture (ITO/PEDOT:PSS/polymer:Y12/PNDIT-F3N-Br/Ag). The PTNT1-F:Y12 blend solution was prepared with chlorobenzene (CB) containing 0.5% (v/v) diphenyl sulfide (DPS) and spin-coated to obtain a photoactive layer thickness of approximately 100 nm. Figure 2a, b show the current density–voltage (J–V) curves and EQEPV spectra of the optimized cells. The JSC values were consistent with the JSC values calculated from the EQEPV spectra (JSCEQE). The photovoltaic parameters are listed in Table 2. The data for the OPV cells fabricated under different conditions are summarized in Supplementary Fig. 7 and Supplementary Table 3. The PTNT1-F cell exhibited VOCs of 0.87–0.88 V, which were higher than those for the D18 (0.85–0.86 V) and PM6 cells (0.83–0.84 V). Furthermore, the PTNT1-F cell exhibited a JSC of 28.3 mA cm−2, which was similar to and higher than the D18 (27.9 mA cm−2) and PM6 cells (26.9 mA cm−2) despite having higher VOCs. Although the FF of the PTNT1-F cell was 0.737, which was lower than the D18 (0.773) and PM6 cells (0.767), the PCE of the PTNT1-F cell was 18.3%, which was similar to and higher than the D18 (18.3%) and PM6 cells (17.3%), respectively. We note that a similar trend in the photovoltaic performance among these polymers was also observed when Y6 was used as the acceptor (Supplementary Fig. 8 and Supplementary Table 4). In addition, the PCEs of the D18 and PM6 cells blended with Y6 and Y12 were consistent with the values reported in a number of literatures so far. These results indicate that the JSC and VOC values of the PTNT1-F cell were remarkably high.
a J–V curves, b EQEPV spectra, c temperature dependence of VOC, and d semilogarithmic plots of EQEEL as a function of current for OPVs based on the PTNT1-F:Y12, D18:Y12, and PM6:Y12 cells. e Plots of JSC/JSC, SQ as a function of ΔVnr for reported OPV cells based on binary blend systems, along with the PTNT1-F:Y12 cells. The red dashed line is a guide to the eye showing current limits of JSC/JSC, SQ with respect to ΔVnr.
The higher VOC observed in the PTNT1-F cell compared to the D18 and PM6 cells is most likely due to the deeper EHOMO and the smaller ∆EHOMO of PTNT1-F. However, the ∆EHOMO values were determined from the EHOMO values of the single materials and do not necessarily represent those of the blends. To confirm the energy offset in the blends, we measured the temperature dependence of VOC for the cells by varying the temperature from 298 to 213 K (Fig. 2c)30,31. By fitting the plots by linear relation, the CT state energy (ECT) was estimated to be 1.21, 1.16, and 1.07 eV for the PTNT1-F, D18, and PM6 blends, respectively, from the intersection at 0 K. Note that ECT at room temperature should be slightly higher than these values. Thus, the energy offset between the LE and CT states (∆E), which is qualitatively equal to ∆EHOMO, was determined to be 0.19, 0.24, and 0.33 eV or less for the PTNT1-F, D18, and PM6 blends, respectively, by using the equation ∆E = EgPV– ECT, where EgPV is the bandgap of Y12 in the blend determined from the inflection point of the EQEPV spectra (Supplementary Fig. 9); EgPV was 1.40 eV for all the blends. Therefore, the trends in the energy offset determined from the single materials (∆EHOMO) and the blends (∆E) agreed well.
To investigate the ∆Vnr, we performed electroluminescence (EL) measurements of the photovoltaic cells; ∆Vnr can be determined as ∆Vnr = −kBT/q ln(EQEEL), where kB is the Boltzmann constant, T is the temperature, q is the elementary charge, and EQEEL is the EL external quantum efficiency (Fig. 2d). The PTNT1-F cell showed an EQEEL of 8.0 × 10−4, which was higher than that of the D18 (2.0 × 10−4) and PM6 cells (9.3 × 10−5). Accordingly, the ∆Vnr of the PTNT1-F cell was calculated to be 0.182 V, which was remarkably smaller than that of the D18 (0.214 V) and PM6 cells (0.235 V). The results indicate that the difference in the VOC mainly originates from the difference in ∆Vnr, and that the difference in ∆Vnr originates from the difference in the energy offset (in both ∆EHOMO and ∆E). In addition, it is noted that the difference in ∆Vnr is also consistent with the difference in ΦPL of the polymers. Thus, the backbone rigidity as well as the lower EHOMO of PTNT1-F plays an important role in reducing ∆Vnr.
Figure 2e depicts the plots of the ratio JSC/JSC, SQ for the PTNT1-F cells and the reported OPV cells based on binary blend systems (Supplementary Table 5); JSC, SQ is the theoretical JSC calculated from the Shockley–Queisser (SQ) model. The ratio JSC/JSC, SQ can be regarded as the charge generation efficiency24 and using the ratio provides fairer comparison than using the maximum EQEPV values at a specific wavelength. Note that for the calculation of JSC, SQ, bandgaps were determined from the onset of the EQEPV spectra, as not all the literatures report the bandgap values determined as described above; however, most of the literatures report the EQEPV spectrum, and thus the onset can be approximately determined (Supplementary Fig. 10). Although it was discussed that there is no trade-off between ∆Vnr and charge generation efficiency24, it is clear that, when more recent data was added here, there exists the trade-off and the JSC/JSC, SQ values decrease with the decrease in ∆Vnr (red-dashed line in Fig. 2e). The JSC/JSC, SQ values were estimated to be 80%, 79%, and 76% for the PTNT1-F:Y12 (∆Vnr = 0.182 V), D18:Y12 (∆Vnr = 0.214 V), and PM6:Y12 (∆Vnr = 0.235 V) systems, respectively. Thus, PTNT1-F exhibited charge generation efficiencies similar to D18 and higher than PM6, even though ∆Vnr was significantly smaller.
Importantly, to the best of our knowledge, the JSC/JSC, SQ value for the PTNT1-F:Y12 system is most likely the highest value reported so far for binary blend OPVs with similarly small ∆Vnr values (Fig. 2e). We also note that this JSC/JSC, SQ value was similar to or even higher than that of the blends of D18 or its derivative D18-Cl combined with Y6 or a Y-series acceptor even though the ∆Vnr value for the PTNT1-F:Y12 system was smaller than that for the D18 systems; for example, JSC/JSC, SQ = 80% and ∆Vnr = 0.193 V for D18:L8-BO and JSC/JSC, SQ = 76% and ∆Vnr = 0.189 V for D18-Cl:L8-BO32,33, which are depicted as yellow circles. Furthermore, the JSC/JSC, SQ values of PTNT1-F systems were higher than those of the blends of PM6, another benchmark polymer, and Y-series acceptors, which are depicted as green circles (JSC/JSC, SQ = ~75% when ∆Vnr = ~0.18 V)34,35,36,37,38. This demonstrates that PTNT1-F is indeed a high-performance π-conjugated polymer that can improve the trade-off between small ∆Vnr and high charge generation efficiency.
We investigated the packing order of the polymers in the neat and blend films using grazing incidence wide-angle X-ray diffraction (GIXD) measurements (Fig. 3). The neat polymer film exhibited a diffraction corresponding to π–π stacking in the wide-angle region along the quasi-qz axis (Fig. 3a–c), which is indicative of a predominant face-on orientation. PTNT1-F exhibited more dominant face-on orientation than D18 and PM6, as the lamellar diffraction appeared in the small-angle region along the qz axis was more significant for D18 and PM6 than PTNT1-F. The π–π stacking distance (dπ) of PTNT1-F was 3.74 Å, which was slightly wider than that of D18 (3.71 Å) and PM6 (3.69 Å) (Fig. 3d). The coherence length (LC) for the π–π stacking of PTNT1-F was 14 Å, which was slightly smaller than that of D18 (15 Å) and PM6 (20 Å). The blend films also showed similar π–π stacking diffractions (Fig. 3e–g), with dπ and Lπ values of around 3.5 and 16–19 Å, respectively (Fig. 3h). Note that, however, these π–π stacking diffractions resulted from the superposition of the diffractions corresponding to the polymer and Y12. More specifically, considering that the dπ in the blend films is consistent with that of Y12 (Supplementary Fig. 11), the difference in Lπ may be ascribed to the difference in the Y12 order. Although these results indicated that each polymer remained the packing order even in blend film, it is difficult to fairly compare the packing order of the polymers in the blend films.
2D GIXD patterns of the polymer neat films for a PTNT1-F, b D18, and c PM6. 2D GIXD patterns of the polymer:Y12 blend films for e PTNT1-F:Y12, f D18:Y12, and g PM6:Y12. Cross-sectional diffraction profiles cut from the 2D GIXD patterns along the quasi-qz (solid line) and qxy (dotted line) for the d polymer neat and h polymer:Y12 blend films, respectively.
We also conducted the differential phase contrast (DPC) scanning transmission electron microscopy (STEM)39 which can observe more precise internal morphology of the films than typical transmission electron microscopy, and the atomic force microscopy (AFM) (Fig. 4 and Supplementary Fig. 12). All the polymer:Y12 blend film similarly exhibited well-interpenetrated networks with small domain sizes, which were beneficial for charge separation, and did not show significant differences.
Electrostatic potential images of a PTNT1-F:Y12, b D18:Y12, and c PM6:Y12 observed in DPC STEM. Black and white areas refer to polymer and NFA, respectively.
As discussed above, the packing structure and morphology do not explain why PTNT1-F show efficient charge generation with a small driving force energy. We therefore carried out more detailed structural analysis using the density of states (DOS) evaluated by PYS with the highly sensitized conditions reported previously40. This measurement probe the entire electronic states, including both crystalline and amorphous regions. The DOS width reflects the energetic distribution of HOMO averaged over the whole film, not only the ordered domains captured by 2D GIXD. Figure 5a–c display the DOS of HOMO for PTNT1-F, D18, and PM6 in the neat and Y12 blend films, respectively. Interestingly, the change in DOS width between the polymer neat and blend films differed by the polymer. PTNT1-F showed almost the same DOS width of 164 and 163 meV for the neat and blend films, respectively, suggesting that PTNT1-F maintained its structural order by blending with Y12. By contrast, D18 and PM6 showed significantly larger DOS widths in the blend film than in the neat film: 135 and 158 meV for the D18 neat and blend films, and 143 and 168 meV for the PM6 neat and blend films, respectively. This suggests that the structural order was reduced (structural defect was increased) for D18 and PM6 by blending with Y12.
Highly-sensitive PYS spectra of the polymer neat and Y6 blend films for a PTNT1-F, b D18, and c PM6. The dotted lines represent Gaussian functions fitted to the DOS edge. The differential absorption spectra obtained by subtracting the absorption of Y12 measured using the Y12:polystyrene (PS) blend film from the polymer:Y12 blend absorption: d PTNT1-F, e D18, and f PM6. Temperature-dependent UV–vis absorption spectra of g PTNT1-F, h D18, and i PM6 in TCB solution.
To confirm the difference in DOS width, we further investigated the polymer backbone order by measuring the UV–vis absorption spectra of the polymer:Y12 blend films. Since the absorptions of the polymers and Y12 are somewhat overlapped, we also measured the spectrum of the Y12:polystyrene (PS) blend film and subtracted it from the spectra of polymer:Y12 blend to probe the polymer absorption in the blend film. Figure 5d–f show the differential and polymer neat film spectra for PTNT1-F, D18, and PM6, respectively. The λmax of PTNT1-F red-shifted for 10 nm by blending Y12 (λmax = 626 nm for the neat film and λmax = 636 nm for the Y12 blend film). Interestingly, this sharply contrasts to the fact that D18 and PM6 showed blue-shift in λmax by blending Y12; In D18, λmax was 588 nm for the neat and 584 nm for the blend, and in PM6, λmax was 620 nm for the neat and 617 nm for the blend. Such difference in the shift of polymer absorption was consistently seen in the polymer:PS blend films (Supplementary Fig. 13). In addition, for D18 and PM6, the intensity of the 0–0 band relative to that of the 0–1 band slightly reduced by blending Y12. These results suggest that the local backbone order of PTNT1-F in the blend film was similar to (or even higher than) that in the neat film, whereas the local backbone order of D18 and PM6 was reduced by blending with Y12. These results support the difference in the change in DOS width in these polymers. Here, we note that the absorption band of Y12 slightly differed in the PTNT1-F blend and D18 and PM6 blends. Y12 band in the PTNT1-F blend had a more intense shoulder at the short-wavelength region than the D18 and PM6 blends, suggesting that Y12 was more disordered when blended with PTNT1-F than blended with D18 and PM6 (Supplementary Fig. 14). Nevertheless, the difference in the backbone order would play the predominant role in the superior charge generation in PTNT1-F as the more disordered Y12 would not enhance the photovoltaic property.
The unchanged local backbone order in PTNT1-F by blending with NFA can be ascribed to the more rigid backbone than D18 and PM6. Figure 5g–i show the temperature-dependent UV–vis absorption spectra of PTNT1-F, D18, and PM6, respectively, measured in trichlorobenzene (TCB) at 20–180 °C. Notably, the absorption spectrum of PTNT1-F changed slowly with increasing temperature; the 0–0 band was still more intense than the 0–1 band even at 100 °C. In contrast, the absorption spectra of D18 and PM6 changed more significantly as the 0–0 band became weaker than the 0–1 band at 60 °C and 80 °C, respectively. This indicates that PTNT1-F has a more rigid backbone and/or has stronger aggregation property than D18 and PM6. Nevertheless, it is somewhat surprising that PTNT1-F remained soluble at room temperature after complete dissolution, and that the PTNT1-F blend solution can be spin-coated at below 40 °C to form uniform films (Supplementary Fig. 15), which exhibits well-mixed phase-separated morphology as shown above (Fig. 4). On the other hand, with the even smaller size of the building unit, D18 can be dissolved in solvents, such as chloroform (CF), CB, and TCB at more than 120 °C, and the solution becomes a complete gel at around 30 °C: the D18 blend film must be formed by hot spin-coating at around 60 °C. This suggests that the aggregation property is much less for PTNT1-F than for D18, and therefore, it is highly likely that PTNT1-F has significantly higher backbone rigidity than D18.
Here, we discuss the efficient charge generation with the significantly small ∆Vnr (∆EHOMO) in the PTNT1-F system. It has been reported that in nonfullerene acceptors, such as Y6, their distinctive π–π molecular packing leads to a remarkable delocalization of the electron wavefunctions, which facilitates the charge separation and thereby leads to efficient charge separation with smaller ∆Vnr compared to the fullerene acceptors25. It should be noted that such delocalization in polymer donors is also important for efficient charge separation. When the holes are transferred from the acceptor to the polymer donor, hole delocalization would facilitate charge dissociation by avoiding geminate recombination. As discussed above, there is a clear difference in the polymer order in the blend between PTNT1-F and D18 and PM6. Thus, it is very likely that the hole delocalization is greater in PTNT1-F than in D18 and PM6.
To support our argument on the hole delocalization, we estimated the effective mass for the hole (mh) along the polymer backbone by using first-principles pseudopotential density functional calculations (Supplementary Figs. 16–19 and Supplementary Table 6)41. In π-conjugated polymers, the extent of the π-electron system, i.e., π-electron delocalization, plays a crucial role in determining the mh42. Thus, mh is a useful metric for gauging the degree of hole delocalization along the polymer backbone. For PTNT1-F, the calculated mh was 0.167m0, which was almost identical to that for D18 (0.168m0) and significantly smaller than that for PM6 (0.195m0) where m0 is the mass of electron in vacuum. This indicates that, under an optimized backbone geometry where the π-electron system is maximally extended, PTNT1-F achieves a level of π-electron delocalization comparable to that of D18 and more extensive than PM6. We assume that the difference in the mh among these polymers would correspond to the difference in the hole delocalization in their neat film. However, in the blend films, the polymer backbones of D18 and PM6 are more disordered than in the neat film as discussed above, and as a result, the mh in the blend of D18 and PM6 are expected to be higher than the values calculated for the optimized geometry. This means that, in the blend, PTNT1-F is expected to exhibit greater hole delocalization than either D18 or PM6.
A question arose whether the change or no change in the backbone order by blending with Y12 occurred at the donor and acceptor (D/A) interface, where the charge separation occurs. To investigate this, we measured the EL spectra of the polymer:Y12 cells while increasing the applied current. Since EL emission in OPV cells originates from the CT states at the D/A interface, the measurements provide insight into structural disorder at the D/A interface. Figure 6a–c show normalized EL spectra for the PTNT1-F, D18, and PM6 OPV cells at the current ranging from 5 to 500 mA (Supplementary Fig. 20). For the PTNT1-F cell, the EL peak at 5 mA was 1.321 eV, which shifted to a higher energy of 1.331 eV at 500 mA, resulting in a shift of the EL peak (∆EEL) of 0.010 eV. The high-energy shift in the EL emission may be attributed to deep traps in the CT state being filled by charge injection, raising the CT state energy. Similarly, the D18 and PM6 cells also showed a high-energy shift in their EL emission. However, the ∆EEL values for the D18 and PM6 cells were 0.015 and 0.018 eV, respectively, which were larger than that of the PTNT1-F cell. The greater shift in ∆EEL may suggest that the trap states were more broadly distributed in the D18 and PM6 blends than in the PTNT1-F blend, resulting in a more pronounced elevation of the CT state energy when the traps are filled. If we assume that the ∆EEL shift reflects structural defects at the D/A interface, these results are consistent with the differences in the DOS change in these polymer systems and, importantly, the change upon blending with Y12 occurs primarily at the D/A interface.
Normalized EL spectra of a PTNT1-F:Y12, b D18:Y12, and c PM6:Y12 cells measured with different applied currents.
Overall, due to its high rigidity, the polymer backbone at the D/A interface in PTNT1-F can adopt a more ordered structure than in D18 and PM6 (Fig. 7), resulting in greater hole delocalization that facilitates charge dissociation, even though the ΔEHOMO is smaller. We propose that key factors in the chemical structure enabling high charge generation with a small ∆Vnr are the symmetry and size of the fused ring. Whereas TNT has a centrosymmetric (C2h) structure, DTBT and 4H,8H-benzo[1,2-c:4,5-c′]dithiophene-4,8-dione (BDTD)—electron-deficient building units for D18 and PM6, respectively—possess axisymmetric (C2v) structures (Fig. 1a–c). In addition, TNT has a more π-extended structure than DTBT and BDTD and is incorporated such that its long axis aligns along the polymer backbone, whereas, in particular, the long axis of BDTD is oriented orthogonally to the backbone. These structural characteristics of PTNT1-F likely account for the difference in backbone rigidity and local ordering.
a Polymer neat films for all three polymers and blend films for b PTNT1-F and c D18 and PM6. In the polymer:Y12 mixed phase, it is assumed that the backbone of PTNT1-F is more ordered than that of D18 and PM6, promoting hole delocalization, facilitating charge dissociation and thereby charge generation.
In conclusion, we have demonstrated that OPV cells based on PTNT1-F exhibits high charge generation efficiencies, defined by JSC/JSC, SQ, exceeding 80%—the highest reported among the OPV systems with similarly small ∆Vnr values, thus small energy offsets, to date. These results establish PTNT1-F as a high-performance π-conjugated polymer that effectively mitigates the fundamental trade-off between voltage loss and charge generation in OPV. Notably, as evidenced by the nearly unchanged DOS width before and after blending with Y12, PTNT1-F distinguishes itself from benchmark polymers, such as D18 and PM6 by its highly rigid backbone—the rigid backbone also accounts for the higher PL quantum yield, which likely contribute to the suppression of ∆Vnr. We propose that this structural characteristic promotes efficient hole delocalization along the polymer backbone, enabling effective charge separation at the D/A interfaces despite the small driving force. We believe these findings will pave the way for designing high-performance π-conjugated polymers that overcome the critical trade-offs between voltage loss and charge generation in OPVs.
The synthetic scheme for PTNT1-F is displayed in Supplementary Scheme 1. 5,11-Bis(5-bromo-3-(2-hexyldecyl)thiophen-2-yl)dithieno[3′,2′:3,4;3′′,2′′:7,8]naphtho[1,2-c:5,6-c′]bis[1,2,5]thiadiazole (1) (56.4 mg, 0.05 mmol), (4,8-bis(5-(2-ethylhexyl)4-fluorothiophen-2-yl)benzo[1,2-b:4,5-b′]dithiophene-2,6-diyl)bis(trimethylstannane) (2) (47.0 mg, 0.05 mmol) (Ossila Ltd.), Pd2(dba)3·CHCl3 (0.53 mg, 0.0005 mmol), P(o-tol)3 (1.22 mg, 0.002 mmol), and 2 mL of chlorobenzene were placed in a reaction vessel. The vessel was purged with argon and sealed. The vessel was inserted into a microwave reactor (Biotage Initiator) and heated at 140 °C for 2 h. After cooling to room temperature, the reaction mixture was poured into 50 mL of methanol (HCl 5 vol%), and then vigorously stirred for 3 h at room temperature. The precipitate was collected by filtration and subjected to sequential Soxhlet extraction with methanol, hexane, and DCM to remove low molecular weight fractions. The residue was extracted with CF and then reprecipitated in methanol. The precipitate was filtered and dried in vacuo to obtain PTNT1-F (74.0 mg, 0.047 mmol) in 94% yield as lustrous blue-black solid. Mn was determined to be 41,300. 1H-NMR (500 MHz, DCB-d4): δ (ppm) 8.40–6.50, 3.40–2.70, and 2.18–0.63.
Density functional theory (DFT) calculations of the model compounds for the polymers were carried out based on the B3LYP/6-31G(d) basis set implemented in the Gaussian 16 Revision C.01 suite programs with default thresholds and algorithms, where the model of the polymer with methyl groups instead of long branched alkyl groups were employed for the calculation. Cyclic voltammograms were recorded on an ALS Electrochemical Analyzer Model 612D with the three electrodes system consisting of a platinum disc working electrode (φ = 3 mm), a platinum wire counter electrode, and an Ag/Ag+ reference electrode in acetonitrile containing tetrabutylammonium hexafluorophosphate (0.1 M) at a scan rate of 100 mV s−1. The thin films of the materials were cast from CB solution directly on the working electrode. All the potentials were calibrated with the half-wave potential of the ferrocene/ferrocenium redox couple measured under identical condition (Fc/Fc+ = 0.17 V for PTNT1-F). HOMO and LUMO energy levels (EHOMO and ELUMO) were estimated by the following equations, EHOMO: −4.80−Eoxonset, ELUMO: −4.80−Eredonset, where Eoxonset and Eredonset are onset potentials for the oxidation and reduction peaks. Photoemission yield spectroscopy (PYS) in air was performed by a spectrometer, model AC-2S (Riken Keiki Co., Ltd). The thin film samples for the PYS measurements were prepared by spin-coating from CB solution on ITO/glass substrate. UV–vis absorption spectroscopy was performed with Shimadzu UV-3600 Plus spectrometer. As for the samples for the UV–vis measurements, the polymer solutions (5 × 10−6 M) were prepared using TCB as the solvent and the polymer thin films were prepared by spin-coating from the CB solution for PTNT1-F and the CF solution for D18, PM6, and Y12.
Photoluminescence (PL) decay was measured by a time-correlated single-photon-counting system (Unisoku, SP-200T-VN-EM) equipped with a photon-counting detector module (Micro Photon Devices, PDM-UV). The thin film samples were prepared by spin-coating on quartz substrate from the CB solution for PTNT1-F and the CF solution for D18, PM6 and Y12. The samples were excited at the absorption maxima with picosecond pulses (660 nm, 100 ps duration, 440 pJ cm−2 pulse−1, or 520 nm, 100 ps duration, 85 pJ cm−2 pulse–1). The FWHM of the instrument response function was ca. 120 ps. The PL decays were biexponential. Therefore, the average PL lifetimes (τf) were used in the analysis of radiative and nonradiative transition rates (kr and knr). The absolute PL quantum yields (ΦPL) were evaluated with an absolute PL/EL quantum yield spectrometer equipped with an integrating sphere (Bunkoukeiki, BEL-300)43. The kr and knr can be evaluated from the following relations: ΦPL = kr × (kr + knr)−1 and τf = (kr + knr)−1.
GIXD experiments were conducted at the SPring-8 on the beamline BL46XU. The sample was irradiated with the X-ray energy of 12.39 keV (λ = 1 Å) at a fixed incident angle on the order of 0.12° through a Huber diffractometer. The two-dimensional (2D) GIXD patterns were recorded with a 2D image detector (Pilatus 300K). Samples for the X-ray measurements were prepared in the same manner as that for solar cell fabrication. Tapping mode atomic force microscopy was carried out on an SPM-9700HT scanning probe microscope (a transmission electron microscope (TEM) (JEM-ARM300F2, JEOL) was used for differential phase contrast (DPC) scanning transmission electron microscopy (STEM) imaging. The microscope was equipped with an aberration corrector for probe-forming systems (Delta corrector, JEOL) and a segmented annular all filed (SAAF) detector (SAAF Quad, JEOL). The observations were performed at an accelerating voltage of 300 kV39.
ITO/glass substrates were pre-cleaned sequentially by sonicating in a detergent bath, then with deionized water, acetone, and isopropanol at room temperature and in a boiled isopropanol bath, each for 10 min, and then baked at 120 °C for 3 min in air. They were then subjected to a UV/ozone treatment at room temperature for 20 min. The pre-cleaned ITO/glass substrates were coated with PEDOT:PSS (Clevios P VP Al 4083) by spin-coating at 4000 rpm for 30 s. The photoactive layer was deposited by spin-coating a polymer:acceptor solution in a glove box (KIYON, KK-011AS-EXTRA) as follows. For the PTNT1-F:Y12 (eFlexPV, Ltd.) blends, 8 g L−1 of the polymer:Y12 (1/1 w/w) solution in CB containing 0.5 vol% of diphenyl sulfur as the solvent additive, where the concentration was based on the polymer weight, was stirred at 120 °C for 30 min, during which time the materials were completely dissolved. The solution was then cooled to 40 °C and spun at 1000 rpm for 20 s. Then, the photoactive layer was immediately baked at 120 °C for 5 min. For the D18:Y12 blends, 4 g L−1 of the polymer:Y12 (1/1.2 w/w) solution in CF was stirred at 120 °C for 10 min, during which time the materials were completely dissolved. The solution was then cooled to 60 °C and spun at 3000 rpm for 10 s. Then, the photoactive layer was immediately baked at 90 °C for 5 min. For the PM6:Y12 (eFlexPV, Ltd.) blends, 7 g L−1 of the polymer:Y12 (1/1.2 w/w) solution in CF was stirred at 60 °C for 30 min, during which time the materials were completely dissolved. The solution was then cooled to 30 °C and spun at 3000 rpm for 10 s. Then, the photoactive layer was immediately baked at 90 °C for 5 min. PNDIT-F3N-Br (Luminescence Technology Corp.) was subsequently deposited by spin-coating from 0.5 g L−1 of the methanol solution (0.5 vol% acetic acid) onto the photoactive layer at 1000 rpm for 30 s in the glove box. The thin films were transferred into a vacuum evaporator (ALS Technology, E-100J) connected to the glove box. Ag layer (150 nm) was deposited by thermal evaporation through a shadow mask under ~10−5 Pa, where the photoactive area was 0.04 cm2.
The J–V characteristics of the cells, with a photomask (0.0351 cm2), were measured in the forward direction using a Keithley 2400 source-measure unit under 1 sun (AM1.5 G) conditions using a solar simulator (SAN-EI Electric, XES-40S1) in the glove box. The light intensity for the J–V measurements was calibrated with a reference PV cell (Bunkoukeiki, BS520BK). EQEPV spectra were measured with a spectral response measuring system (SOMA OPTICS, S-9241). More than 10 different substrates (four photoactive areas each) were prepared for the optimized cells, and their photovoltaic properties were measured. High-sensitive EQEPV spectra were measured with a spectral response measuring system (Bunkoukeiki, ECT-250D). Photoactive layer thickness was measured with an ET4000 (Kosaka Laboratory, Ltd.), where the optimal photoactive layer thickness was around 100 nm.
For the electroluminescence (EL) measurements of the OPV cells, the cells were fabricated on the substrate with a patterned ITO layer, where the photoactive area was 0.16 cm2. The EL spectra and EQEEL were measured using the absolute PL/EL quantum yield spectrometer with an integrating sphere (Bunkoukeiki, BEL-300).
The band-structure calculations of isolated polymers have been performed using the Quantum ESPRESSO44. We use the van der Waals density functional (vdW-DF2)45. The cutoff energies for the plane wave and charge density are 50 and 500 Ry, respectively. The Brillouin zone integration is performed with a 4 × 4 × 4 k-point set. The effective mass for hole (mh) and electron (me) for each polymer can be extracted from the band structure of the polymer structure simulated by the periodic boundary condition using the central repeat unit in the optimized trimer model.
The authors declare that the main data supporting the findings of this study are available within the article and its Supplementary Information files. Other datasets generated and/or analyzed during the current study are available from the corresponding authors upon reasonable request.
Guan, S. et al. Self-assembled interlayer enables high-performance organic photovoltaics with power conversion efficiency exceeding 20%. Adv. Mater. 36, e2400342 (2024).
Article PubMed Google Scholar
Fu, J. et al. Two-step crystallization modulated through acenaphthene enabling 21% binary organic solar cells and 83.2% fill factor. Nat. Energy 10, 1251–1261 (2025).
Article CAS Google Scholar
Wang, L. et al. Diluted ternary heterojunctions to suppress charge recombination for organic solar cells with 21% efficiency. Adv. Mater. 37, e2419923 (2025).
Article PubMed Google Scholar
Chen, H. et al. Organic solar cells with 20.82% efficiency and high tolerance of active layer thickness through crystallization sequence manipulation. Nat. Mater. 24, 444–453 (2025).
Article CAS PubMed Google Scholar
Yao, J. et al. Quantifying losses in open-circuit voltage in solution-processable solar cells. Phys. Rev. Appl. 4, 014020 (2015).
Article Google Scholar
Ross, R. T. Some thermodynamics of photochemical systems. J. Chem. Phys. 46, 4590–4593 (1967).
Article CAS Google Scholar
Rau, U. Reciprocity relation between photovoltaic quantum efficiency and electroluminescent emission of solar cells. Phys. Rev. B 76, 085303 (2007).
Article Google Scholar
Liu, J. et al. Fast charge separation in a non-fullerene organic solar cell with a small driving force. Nat. Energy 1, 16089 (2016).
Article CAS Google Scholar
Qian, D. et al. Design rules for minimizing voltage losses in high-efficiency organic solar cells. Nat. Mater. 17, 703–709 (2018).
Article CAS PubMed Google Scholar
Eisner, F. D. et al. Hybridization of local exciton and charge-transfer states reduces nonradiative voltage losses in organic solar cells. J. Am. Chem. Soc. 141, 6362–6374 (2019).
Article CAS PubMed Google Scholar
Saito, T., Natsuda, S., Imakita, K., Tamai, Y. & Ohkita, H. Role of energy offset in nonradiative voltage loss in organic solar cells. Sol. RRL 4, 2000255 (2020).
Article CAS Google Scholar
Li, W., Hendriks, K. H., Furlan, A., Wienk, M. M. & Janssen, R. A. J. High quantum efficiencies in polymer solar cells at energy losses below 0.6 eV. J. Am. Chem. Soc. 137, 2231–2234 (2015).
Article CAS PubMed Google Scholar
Bertrandie, J. et al. The energy level conundrum of organic semiconductors in solar cells. Adv. Mater. 34, 2202575 (2022).
Article CAS Google Scholar
Wang, M. et al. High open circuit voltage in regioregular narrow band gap polymer solar cells. J. Am. Chem. Soc. 136, 12576–12579 (2014).
Article CAS PubMed Google Scholar
Kawashima, K., Tamai, Y., Ohkita, H., Osaka, I. & Takimiya, K. High-efficiency polymer solar cells with small photon energy loss. Nat. Commun. 6, 10085 (2015).
Article CAS PubMed PubMed Central Google Scholar
Ran, N. A. et al. Harvesting the full potential of photons with organic solar cells. Adv. Mater. 28, 1482–1488 (2016).
Article CAS PubMed Google Scholar
Rosenthal, K. D. et al. Quantifying and understanding voltage losses due to nonradiative recombination in bulk heterojunction organic solar cells with low energetic offsets. Adv. Energy Mater. 9, 1901077 (2019).
Article Google Scholar
Yuan, J. et al. Enabling low voltage losses and high photocurrent in fullerene-free organic photovoltaics. Nat. Commun. 10, 570 (2019).
Article CAS PubMed PubMed Central Google Scholar
Yuan, J. et al. Single-junction organic solar cell with over 15% efficiency using fused-ring acceptor with electron-deficient core. Joule 3, 1140–1151 (2019).
Article CAS Google Scholar
Li, C. et al. Non-fullerene acceptors with branched side chains and improved molecular packing to exceed 18% efficiency in organic solar cells. Nat. Energy 6, 605–613 (2021).
Article CAS Google Scholar
Saito, M., Ohkita, H. & Osaka, I. π-conjugated polymers and molecules enabling small photon energy loss simultaneously with high efficiency in organic photovoltaics. J. Mater. Chem. A 8, 20213–20237 (2020).
Article CAS Google Scholar
Li, S. et al. Highly efficient fullerene-free organic solar cells operate at near zero highest occupied molecular orbital offsets. J. Am. Chem. Soc. 141, 3073–3082 (2019).
Article CAS PubMed Google Scholar
Li, X. et al. Roles of acceptor guests in tuning the organic solar cell property based on an efficient binary material system with a nearly zero hole-transfer driving force. Chem. Mater. 32, 5182–5191 (2020).
Article CAS Google Scholar
Chen, X.-K. et al. A unified description of non-radiative voltage losses in organic solar cells. Nat. Energy 6, 799–806 (2021).
Article CAS Google Scholar
Zhang, G. et al. Delocalization of exciton and electron wavefunction in non-fullerene acceptor molecules enables efficient organic solar cells. Nat. Commun. 11, 3943 (2020).
Article CAS PubMed PubMed Central Google Scholar
Cai, G. et al. Deuteration-enhanced neutron contrasts to probe amorphous domain sizes in organic photovoltaic bulk heterojunction films. Nat. Commun. 15, 2784 (2024).
Article CAS PubMed PubMed Central Google Scholar
Jiang, K. et al. Photoluminescent delocalized excitons in donor polymers facilitate efficient charge generation for high-performance organic photovoltaics. Nat. Commun. 16, 3176 (2025).
Article CAS PubMed PubMed Central Google Scholar
Fratini, S., Nikolka, M., Salleo, A., Schweicher, G. & Sirringhaus, H. Charge transport in high-mobility conjugated polymers and molecular semiconductors. Nat. Mater. 19, 491–502 (2020).
Article CAS PubMed Google Scholar
Mikie, T. et al. Dithienonaphthobisthiadiazole synthesized by thienannulation of electron-deficient rings: an acceptor building unit for high-performance π-conjugated polymers. Chem. Sci. 15, 19991–20001 (2024).
Article CAS PubMed PubMed Central Google Scholar
Vandewal, K., Tvingstedt, K., Gadisa, A., Inganäs, O. & Manca, J. V. Relating the open-circuit voltage to interface molecular properties of donor: acceptor bulk heterojunction solar cells. Phys. Rev. B 81, 125204 (2010).
Article Google Scholar
Hörmann, U. et al. Quantification of energy losses in organic solar cells from temperature-dependent device characteristics. Phys. Rev. B 88, 235307 (2013).
Article Google Scholar
Gao, J. et al. Over 19.2% efficiency of organic solar cells enabled by precisely tuning the charge transfer state via donor alloy strategy. Adv. Sci. 9, 2203606 (2022).
Article CAS Google Scholar
Wei, Y. et al. Binary organic solar cells breaking 19% via manipulating the vertical component distribution. Adv. Mater. 34, 2204718 (2022).
Article CAS Google Scholar
Shi, Y. et al. Small reorganization energy acceptors enable low energy losses in non-fullerene organic solar cells. Nat. Commun. 13, 3256 (2022).
Article CAS PubMed PubMed Central Google Scholar
Shi, Y. et al. Small energetic disorder enables ultralow energy losses in non-fullerene organic solar cells. Adv. Energy Mater. 13, 2300458 (2023).
Article CAS Google Scholar
Zhan, L. et al. Multiphase morphology with enhanced carrier lifetime via quaternary strategy enables high-efficiency, thick-film, and large-area organic photovoltaics. Adv. Mater. 34, 2206269 (2022).
Article CAS Google Scholar
He, C. et al. Simultaneous improvements in efficiency and stability of organic solar cells via a symmetric-asymmetric dual-acceptor strategy. Adv. Energy Mater. 13, 2204154 (2023).
Article CAS Google Scholar
Bi, Z. et al. Tuning acceptor composition in ternary organic photovoltaics–impact of domain purity on non-radiative voltage losses. Adv. Energy Mater. 12, 2103735 (2022).
Article CAS Google Scholar
Inamoto, S., Shimomura, S. & Otsuka, Y. Electrostatic potential imaging of phase-separated structures in organic materials via differential phase contrast scanning transmission electron microscopy. Microscopy 69, 304–311 (2020).
Article CAS PubMed Google Scholar
Nakano, K., Kaji, Y. & Tajima, K. Highly sensitive evaluation of density of states in molecular semiconductors by photoelectron yield spectroscopy in air. ACS Appl. Mater. Interfaces 13, 28574–28582 (2021).
Article CAS PubMed Google Scholar
Northrup, J. E. Atomic and electronic structure of polymer organic semiconductors: P3HT, PQT, and PBTTT. Phys. Rev. B 76, 245202 (2007).
Article Google Scholar
Mikie, T. et al. Extended π-electron delocalization in quinoid-based conjugated polymers boosts intrachain charge carrier transport. Chem. Mater. 33, 8183–8193 (2021).
Article CAS Google Scholar
Mello, J. C., de, Wittmann, H. F. & Friend, R. H. An improved experimental determination of external photoluminescence quantum efficiency. Adv. Mater. 9, 230–232 (1997).
Article Google Scholar
Giannozzi, P. et al. QUANTUM ESPRESSO: a modular and open-source software project for quantum simulations of materials. J. Phys. Condens. Matter 21, 395502 (2009).
Article PubMed Google Scholar
Thonhauser, T. et al. Spin signature of nonlocal correlation binding in metal-organic frameworks. Phys. Rev. Lett. 115, 136402 (2015).
Article CAS PubMed Google Scholar
Download references
This work was supported by the MIRAI Program (grant no. JPMJMI20E) and Advanced Technologies for Carbon-Neutral (ALCA-Next) (grant no. JPMJAN25G1) from the Japan Science and Technology Agency, KAKENHI from Japan Society for the Promotion of Science (21H04692 and 22K14745), and the HIRAKU-Global Program from the Ministry of Education, Culture, Sports, Science and Technology (MEXT). 2D GIXD experiments were performed at BL46XU of SPring-8 with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (Proposal Nos. 2023B1606 and 2023B1933). The authors thank Dr. T. Koganezawa (JASRI) for support in the 2D GIXD measurements and Mr. H. Motozawa (Kyoto University) for his support in the EQEEL measurements.
Applied Chemistry Program, Graduate School of Advanced Science and Engineering, Hiroshima University, Higashi-Hiroshima, Hiroshima, Japan
Shota Suruga, Tsubasa Mikie & Itaru Osaka
Department of Polymer Chemistry, Graduate School of Engineering, Kyoto University, Katsura, Nishikyo-ku, Kyoto, Japan
Yuki Sato, Kazuki Kohzuki, Jihun Jeon, Hyung Do Kim & Hideo Ohkita
Morphological Research Laboratory, Toray Research Center, Otsu, Shiga, Japan
Shin Inamoto
Department of Materials Science, Institute of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Ibaraki, Japan
Hiroyuki Ishii
RIKEN Center for Emergent Matter Science, Wako, Saitama, Japan
Kyohei Nakano & Keisuke Tajima
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
S.S. synthesized the polymer sample and conducted the measurements and analyses for the polymer properties, GIXD, and TEM. S.S., T.M. fabricated and measured the OPV cells and measurement. Y.S., K.K., and J.J. conducted the PL and temperature-dependent VOC measurement and analysis supervised by H.D.K. and H.O. S.S., T.M., and Y.S. carried out EL measurements and analysis supervised by H.O. S.I. conducted the STEM measurements. H.I. carried out the band calculations. K.N. and K.T. carried out the high-sensitive PYS measurements and analysis for the DOS. T.M. and I.O. designed and directed the project. S.S. and T.M. wrote the draft of the manuscript. All authors reviewed and revised the manuscript.
Correspondence to Tsubasa Mikie, Keisuke Tajima, Hideo Ohkita or Itaru Osaka.
The authors declare no competing interests.
Communications Materials thanks the anonymous reviewers for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
Suruga, S., Mikie, T., Sato, Y. et al. Backbone rigidity promoting hole delocalization and enabling efficient charge generation with minimal voltage loss in nonfullerene organic photovoltaics. Commun Mater 7, 79 (2026). https://doi.org/10.1038/s43246-026-01115-y
Download citation
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s43246-026-01115-y
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
Advertisement
Communications Materials (Commun Mater)
ISSN 2662-4443 (online)
© 2026 Springer Nature Limited
Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.
