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
Nature Communications volume 17, Article number: 2330 (2026)
The realization of high-efficiency thick-film organic solar cells (OSCs) is crucial for scalable manufacturing yet remains challenging due to limited exciton diffusion. Here, we introduce a magnetic strategy by incorporating two-dimensional (2D) ferromagnetic MoPS3 nanocrystal into the active layer to manipulate exciton spin dynamics. We demonstrate that the interaction between the ferromagnetic MoPS3 nanocrystal and excitons promotes the formation of weak intrinsic magnetic fields within the active layer. These fields effectively promote the intersystem crossing (ISC) from short-lived singlet excitons to long-lived triplet excitons, thereby extending exciton diffusion length and reducing non-radiative recombination losses. Consequently, MoPS3 nanocrystal doped D18-Cl:L8-BO devices achieve power conversion efficiencies of 20.37% at an active layer thickness of 100 nm and 19.36% (19.13% certified value) at an active layer thickness of 300 nm, representing one of the highest reported values for thick-film ( > 300 nm) OSCs. Universal applicability is demonstrated with power conversion efficiencies of 20.91%/19.63% (D18:L8-BO) and 19.13%/17.92% (PM6:Y6) at 100/300 nm. This work establishes 2D ferromagnetic MoPS3 nanocrystal as effective spin manipulators in organic semiconductors and provides a universal strategy to overcome the critical thickness-performance trade-off in OSCs.
The pursuit of organic solar cells (OSCs) has been driven by their potential for lightweight, flexible, and large-area photovoltaic applications via cost-effective solution-processing techniques1,2,3,4,5,6,7. Significant advancements have been achieved with power conversion efficiencies (PCEs) now beyond 20% in state-of-the-art laboratory devices8,9,10,11,12. However, these record efficiencies are invariably obtained from devices with thin active layers (typically ≤ 100 nm), which are amenable to efficient exciton diffusion and charge collection13,14. For industrial-scale manufacturing using printing and coating techniques, a thicker active layer ( ≥ 300 nm) is highly desirable to enhance process tolerance, improve film uniformity over large areas, and ensure robust, defect-free fabrication15,16. Unfortunately, the inherent trade-off between active layer thickness and device performance remains a formidable challenge17,18,19. In thick-film OSCs, limited charge-carrier mobility and pronounced bimolecular recombination lead to severe photocurrent and fill factor losses, ultimately constraining the PCE20,21.
The core of this limitation lies in the photophysics of the primary photogenerated species: the singlet exciton (S1)22,23. In a thick-film active layer ( ≥ 300 nm), the probability of an S1 exciton diffusing from its generation site to the donor/acceptor (D/A) interface for charge separation becomes critically low. This is governed by two intrinsic constraints: a short exciton diffusion length (LD, typically 5–20 nm in organic blends) and a fleeting lifetime on the nanosecond scale. A promising route to overcome this barrier is to populate the triplet exciton (T1) state. T1 excitons exhibit microsecond to millisecond lifetimes due to their spin-forbidden decay, which translates to a substantially longer effective diffusion length, dramatically increasing the probability of successful charge separation in a thick-film active layer24. However, harnessing this potential is non-trivial. The intersystem crossing (ISC) process from S1 to T1 is spin-forbidden in most organic semiconductors due to weak spin-orbit coupling (SOC) inherent in light elements (e.g., C, H, O, N) that constitute these materials, creating a kinetic bottleneck.
In this work, we introduce a materials-based strategy to overcome this spin-dynamic bottleneck. We synthesized a novel two-dimensional ferromagnetic nanocrystal, MoPS3, via chemical vapor transport (CVT) and derived nanocrystals through solvent-assisted ultrasonic exfoliation25,26,27,28,29, and incorporated it as a spin-functional additive. The interaction between the ferromagnetic MoPS3 nanocrystals and photoexcited species promotes the formation of weak intrinsic magnetic fields within the active layer. These fields, synergizing with the enhanced SOC provided by the heavy molybdenum (Mo) atoms, effectively promote the ISC process. This catalyzes the conversion of short-lived singlets into long-lived triplets, thereby extending the LD and suppressing non-radiative energy loss. Remarkably, MoPS3 nanocrystal doped D18-Cl:L8-BO devices achieve a PCE of 20.37% at a 100 nm thickness and retain a high efficiency of 19.36% at 300 nm, ranking among the highest reported for thick-film OSCs. More importantly, the approach demonstrates remarkable generality and thickness tolerance across high-performance systems, for which D18:L8-BO based devices attain PCEs of 20.91% (100 nm) and 19.63% (300 nm), while PM6:Y6-based devices reach 19.13% (100 nm) and 17.92% (300 nm). Our work establishes ferromagnetic MoPS3 nanocrystal integration as a paradigm for designing thickness-resilient, high-efficiency OSCs, directly addressing a key manufacturability obstacle.
MoPS3 bulk crystal was synthesized via CVT using iodine as a transport agent, with the detailed synthetic route summarized in supplementary information (SI). MoPS3 nanocrystals were subsequently obtained in high yield through solvent-assisted ultrasonic exfoliation (Supplementary Fig. 1). X-ray photoelectron spectroscopy (XPS) depth profiling verifies chemical structure of MoPS3 nanocrystal (Supplementary Fig. 2). X-ray diffraction (XRD) patterns exhibited characteristic (001), (002), and (003) reflections at 15.3°, 9.5°, and 30°, respectively, verifying preserved long-range layered ordering (Supplementary Fig. 3). Atomic force microscopy (AFM) revealed a monolayer-dominated morphology with uniform surface topography and a thickness distribution centered around 0.41 nm (Supplementary Fig. 4). High-resolution transmission electron microscopy (HR-TEM) further resolved crystalline domains with lattice spacings of 0.65 nm and 0.31 nm, corresponding to the (001) and (110) planes, respectively (Supplementary Fig. 5). Dielectric and magnetic profiling demonstrated notably high permittivity and permeability, confirming the capability of MoPS3 nanocrystals to modulate exciton-spin interactions (Supplementary Figs. 6–7). The pronounced SOC in MoPS3 nanocrystals, a direct consequence of the heavy Mo element, manifests as significant band splitting and a complete lifting of spin degeneracy (Fig. 1a-c). This strong SOC, primarily driven by the heavy d-orbital of Mo, leads to a substantial reconstruction of the low-energy electronic structure near the Fermi level. These features provide clear evidence for the decisive role of heavy-metal-derived SOC in defining the electronic and magnetic properties of MoPS3 nanocrystals.
a Single-layer structure diagram of MoPS3 and bipyramid structure of [P2S6]. b Band structure diagram of MoPS3 with spin-orbit coupling. c Band structure diagram of MoPS3 without spin-orbit coupling. d Magnetic hysteresis loop of D18-Cl:L8-BO and D18-Cl:L8-BO:MoPS3 films. SD, saturation magnetization. e–h Absorption, fluorescence and phosphorescence spectrum of D18-Cl, D18-Cl:MoPS3, L8-BO and L8-BO:MoPS3 films. S1, singlet exciton; T1, triplet exciton; ΔEST, singlet-triplet energy gap. i EQEEL spectrum of D18-Cl:L8-BO and D18-Cl:L8-BO:MoPS3 devices. EQEEL, electroluminescence external quantum efficiency.
The influence of MoPS3 nanocrystals on the intrinsic magnetic field within the D18-Cl:L8-BO system (Supplementary Fig. 8) was examined using a vibrating-sample magnetometer (VSM). Incorporation of MoPS3 into the D18-Cl:L8-BO composite significantly enhanced the saturation magnetization (SA) from 0.66 emu/g to 1.04 emu/g (Fig. 1d), confirming the reinforcement of intrinsic magnetic fields30. To probe the impact on excited states, we investigated the singlet (ES) and triplet (ET) energy levels. The ES values of D18-Cl and L8-BO, determined from absorption and photoluminescence (PL) spectra intersection, were 2.06 eV and 1.47 eV (Supplementary Table 1), respectively. Due to the large singlet-triplet energy gap (ΔEST > 0.82 eV for D18-Cl and > 0.65 eV for L8-BO), phosphorescent emission at 77 K was nearly undetectable for both neat materials (Fig. 1e, f), indicating energetically blocked ISC. In contrast, dilute solutions of D18-Cl:MoPS3 and L8-BO:MoPS3 at 77 K exhibited clear phosphorescence peaks around 690 nm and 945 nm (Fig. 1g-h), corresponding to ET values of 1.80 eV and 1.31 eV, respectively. The resulting ΔEST values were reduced to 0.25 eV and 0.16 eV, respectively, markedly smaller than those of the pristine materials and favorable for efficient ISC from S1 to T1. Notably, the ET of D18-Cl:MoPS3 (1.80 eV) exceeds the ES of L8-BO (1.47 eV), suggesting potential for triplet-to-singlet energy transfer.
The charge-transfer (CT) state energy (ECT) was determined from the low-energy tail of the Fourier-transform photocurrent spectroscopy (FTPS) and electroluminescence external quantum efficiency (EQEEL) spectra (Supplementary Fig. 9; Fig. 1i). The extracted ECT values are 1.34 eV for the pristine D18-Cl:L8-BO blend and 1.35 eV for the MoPS3-incorporated device. In the control system, ECT lies below the triplet energies (ET) of both donor and acceptor, permitting non-radiative recombination loss via triplet-state back-transfer. In contrast, the incorporation of MoPS3 elevates the triplet energy of D18-Cl above ECT, effectively blocking this loss channel. Furthermore, the reduced energy offset between ET and ECT also implies a more competitive forward CT rate from L8-BO triplet excitons to the CT state (Supplementary Table 2). Additional support comes from Urbach energy analysis of the sub-bandgap EQE spectra, which shows a decrease from 23.66 meV to 21.72 meV upon MoPS3 addition, indicating suppressed energetic disorder31. Collectively, these findings demonstrate that MoPS3 nanocrystals serve a dual function: they reduce ΔEST to promote ISC (suppressing geminate recombination) and strategically realign the triplet and CT energy landscape to minimize non-radiative losses.
Electron paramagnetic resonance (EPR) spectroscopy provides direct evidence for the MoPS3-mediated modulation of triplet exciton generation. In the absence of MoPS3, the superoxide radical signal that resulting from the reaction of T1 excitons with oxygen is weak and broadened (Fig. 2a), indicating low ISC efficiency and limited T1 yield. Meanwhile, the singlet oxygen signal originating from direct S1 – oxygen interaction is relatively intense and broad, reflecting that most S1 excitons are lost non-radiatively. Upon incorporation of MoPS3, the superoxide radical signal increases markedly and sharpens (Fig. 2b), which may be attributed to the Mo heavy-atom effect enhancing SOC and thus accelerating ISC. Correspondingly, the singlet oxygen signal is substantially suppressed, demonstrating efficient S1 to T1 conversion. Transient absorption (TA) spectroscopy (Fig. 2c, d) conclusively establishes MoPS3 nanocrystal-mediated triplet exciton generation and accelerated ISC within the D18-Cl:L8-BO system. Near-IR kinetics (1200 nm) assigned to T1 → Tn transitions show extended triplet lifetime from 802 ps (pristine D18-Cl:L8-BO) to 1198 ps (MoPS3 nanocrystal doped D18-Cl:L8-BO), confirming interfacial stabilization of triplets (Fig. 2e).
a Superoxide radical and (b) singlet oxygen for EPR spectroscopy of D18-Cl:L8-BO and D18-Cl:L8-BO:MoPS3 films. (c, d) TA spectrum and (e) the trace kinetic of GSB signals of D18-Cl:L8-BO and D18-Cl:L8-BO:MoPS3 films.
The generation of triplet excitons extends the LD. Excitation-fluence-dependent TA measurements were performed (Fig. 3a, b). The calculated LD values increased from 16 nm (pristine D18-Cl) to 25 nm (D18-Cl:MoPS3) and from 21 nm (neat L8-BO) to 37 nm (L8-BO:MoPS3), confirming enhanced exciton diffusion in MoPS3 nanocrystal-doped phases. MoPS3 nanocrystal incorporation elevated hole (μh) and electron (μe) mobility in both individual and blended films (Supplementary Figs. 10–13; Supplementary Tables 3 and 4), indicating effective promotion of charge transport pathways32. In the D18-Cl:L8-BO blend ( ~ 100 nm), both μh and μe are enhanced from 1.16 × 10−3 and 1.31 × 10−3 to 2.21 × 10−3 and 2.09 × 10−3 cm2 V−1 s−1 upon MoPS3 incorporation. Notably, when the thickness increases to ~300 nm, the pristine blend suffers a severe mobility reduction (μh = 0.38 × 10−3 and μe = 0.42 × 10−3 cm2 V−1 s−1), whereas the MoPS3-modified film maintains high and well-balanced mobilities (μh = 1.86 × 10−3 and μe = 1.77 × 10−3 cm2 V−1 s−1). This thickness-robust and balanced charge transport effectively suppresses space-charge accumulation and recombination losses, enabling efficient charge extraction in thick-film active layers. Complementary transient photovoltage (TPV) measurements (Supplementary Fig. 14) under open-circuit conditions independently assessed LD via the Einstein-Smoluchowski relation: LD = (μτkT/q)1/2, where μ indicates charge-carrier mobility, k denotes the Boltzmann constant, and q stands for the elementary charge. Fitting TPV decays [y = Aexp( − x/τ)] yielded longer lifetimes for MoPS3 nanocrystal doped D18-Cl:L8-BO devices (1.61 μs vs. 1.00 μs control)33, corresponding to LD enhancements from 55 nm to 96 nm (D18-Cl) and 58 nm to 93 nm (L8-BO). Although the absolute LD values derived from the ensemble-averaged TPV method and the local-probe EEA analysis are not directly comparable, both techniques consistently demonstrate a substantial increase in the LD upon MoPS3 incorporation (Supplementary Fig. 15). These findings are corroborated by time-resolved photoluminescence (TRPL), which showed an extended average fluorescence lifetime from 0.26 ns to 0.39 ns upon MoPS3 incorporation (Supplementary Fig. 16).
a, b Singlet excitons dynamics of control and MoPS3 nanocrystal doped D18-Cl (or L8-BO) films measured at different densities. LD, exciton diffusion length. c, d TA spectrum. e, f The trace kinetic of GSB signals for control and MoPS3 nanocrystal doped D18-Cl:L8-BO films.
Complementary TA spectroscopy under 800 nm excitation (L8-BO selective) revealed exciton dynamics. Collective analysis of 2D color plots (Supplementary Fig. 17), TA spectra (Fig. 3c, d), and ground-state bleaching (GSB) kinetics (Fig. 3e, f) demonstrates synchronized decay of L8-BO GSB (798 nm) with D18-Cl GSB rise (587 nm), providing direct evidence of hole transfer from L8-BO to D18-Cl. Biexponential fitting resolved two lifetime components: τ1 (exciton dissociation lifetime at D/A interfaces) and τ2 (exciton diffusion lifetime to D/A interfaces). Notably, MoPS3 nanocrystal incorporation significantly boosted these exciton lifetime constants at both D18-Cl (τ1 increased from 1.78 ps to 2.15 ps and τ2 from 25.39 ps to 29.95 ps) and L8-BO (τ1 increased from 1.23 ps to 1.35 ps and τ2 from 15.16 ps to 19.70 ps). This extension aligns with the enhanced exciton diffusion lengths. The control device showed a maximum exciton generation rate (Gmax) of 1.74×1028 m−3 s−1 and an exciton dissociation efficiency [P(E,T)] of 97.2%. MoPS3 incorporation elevated these to 1.75×1028 m−3 s−1 and 97.6%, respectively (Supplementary Fig. 18; Supplementary Table 5)34,35.
Time-dependent density functional theory (TD-DFT) simulations examined the photoexcitation characteristics (Supplementary Fig. 19). The molar absorptivity increased from 9.1×104 L mol⁻1 cm⁻1 (pristine) to 1.30×105 L mol⁻1 cm⁻1 (MoPS3-doped), with a red-shifted excitation energy and increased oscillator strength. Hole-electron distribution mapping showed that MoPS3 doping markedly increased and continuouslized the charge density networks (Supplementary Figs. 20–23). Calculations of the highest occupied molecular orbital (HOMO)/ lowest unoccupied molecular orbital (LUMO) levels using a polarization functions (DZVP-MOLOPT-SR-GTH) in a hybrid Gaussian/plane method (Supplementary Figs. 24–25) quantified the electronic structure modification. The HOMO-LUMO gap narrowed upon MoPS3 nanocrystal incorporation, which may strengthen donor-acceptor orbital overlap, reduce the exciton dissociation barrier, and extend the excited-state lifetime, thereby promoting charge separation.
To systematically assess photovoltaic implications, we fabricated control and MoPS3 nanocrystal doped D18-Cl:L8-BO devices with architecture of ITO/2PACz/active layer/PNDIT-F3N/Ag (Supplementary Fig. 26). Current density-voltage (J-V) characterization (Fig. 4a) show the reference D18-Cl:L8-BO device (100 nm active layer) delivers a PCE of 18.97%, with open-circuit voltage (VOC) of 0.92 ± 0.01 V, short-circuit current density (JSC) of 26.19 ± 0.64 mA cm−2, and fill factor (FF) of 76.34 ± 1.23%. Increasing thickness to 300 nm degrades PCE to 16.79%, primarily from FF reduction (69.12 ± 0.88%). Conversely, optimized MoPS3 nanocrystal (0.6 wt%) doped D18-Cl:L8-BO devices (100 nm active layer) achieve peak PCE of 20.37% (VOC = 0.92 ± 0.01 V, JSC = 27.45 ± 0.73 mA cm−2, FF = 79.63 ± 0.68%), as detailed in Table 1 and Supplementary Table 6. This optimum is consistent with the concentration-dependent SCLC results, in which 0.6 wt% MoPS3 nanocrystal incorporation yields enhanced and balanced charge mobilities (Supplementary Fig. 27). Critically, 300 nm MoPS3 nanocrystal doped D18-Cl:L8-BO devices retain a PCE of 19.36% (VOC = 0.92 ± 0.01 V, JSC = 27.24 ± 0.86 mA cm−2, FF = 76.41 ± 0.82%), exhibiting exceptional thickness tolerance (Fig. 4b). This 19.36% efficiency was certified at 19.13% (Supplementary Fig. 28), standing as the record-high for 300 nm-thick OSCs (Fig. 4c; Supplementary Table 7). External quantum efficiency (EQE) spectra (Fig. 4d) confirm enhanced photo response across wavelengths for MoPS3 nanocrystal doped D18-Cl:L8-BO devices at both thicknesses versus undoped counterparts. To assess operational resilience, we tracked parameter evolution under thermal stress (50 °C), LED illumination, and extended N2 storage (Supplementary Fig. 29; Supplementary Tables 8−13). After 1008 h, MoPS3-doped devices consistently showed higher PCE retention than controls across all conditions.
a J–V characteristic curves and (b) the PCE versus film thickness plots of control and MoPS3 nanocrystal doped D18-Cl:L8-BO devices. c PCE versus JSC plots for 300 nm-thick OSCs reported in the literature. d EQE spectrum of control and MoPS3 nanocrystal doped D18-Cl:L8-BO devices. e–h Temperature-dependent PL spectrum and (i) fitted low frequency reorganization energy for control and MoPS3 nanocrystal doped D18-Cl (or L8-BO) films.
Charge recombination and extraction dynamics were further examined. The relationship VOC∝nkT/q ln Plight (k: Boltzmann constant, T: absolute temperature, q: elementary charge) reflects diminishing trap-assisted recombination as the ideality factor n approaches unity36. Linear regression yielded slopes of 1.21 kT/q (control) versus 1.16 kT/q (MoPS3 nanocrystal doped) devices (Supplementary Fig. 30a), signifying mitigated trap-assisted recombination with MoPS3 nanocrystal integration. Complementarily, the power exponent S in JSC ∝ (Plight)S indicates recombination kinetics, where S ≈ 1 corresponds to minimal bimolecular losses37. Fitting yielded S = 0.89 for control D18-Cl:L8-BO devices and S = 0.91 for MoPS3 nanocrystal doped D18-Cl:L8-BO devices (Supplementary Fig. 30b), indicating restrained bimolecular recombination in MoPS3 nanocrystal doped devices. Charge recombination dynamics were examined via transient photocurrent (TPC) measurements38,39. The extracted charge transport times decreased from 0.39 μs (control D18-Cl:L8-BO device) to 0.37 μs (MoPS3 nanocrystal doped D18-Cl:L8-BO device), confirming accelerated charge extraction enabled by MoPS3 nanocrystal (Supplementary Fig. 31)40. Temperature-dependent PL spectroscopy quantified energetic disorder via a Marcus-Levich-Jortner framework (Fig. 4e-h; Supplementary Figs. 32–33; Supplementary Tables 14−15)41. The D18-Cl:MoPS3 and L8-BO:MoPS3 blends exhibited smaller low-frequency reorganization energy (λ) compared to their neat counterparts (Fig. 4i), demonstrating that MoPS3 suppresses non-radiative recombination by reducing dynamic disorder.
Ab initio molecular dynamics (AIMD) simulations quantified interaction energies (E) to evaluate interfacial affinity (Supplementary Fig. 34). The computed E for MoPS3/D18-Cl (−23.83 eV) and MoPS3/L8-BO (−20.12 eV) were comparably strong. The interaction energy for the pristine D18-Cl:L8-BO blend was −13.44 eV, which shifted to −4.01 eV upon MoPS3 incorporation. This decrease implies that MoPS3 promotes spatial decoupling between the donor and acceptor phases. Fourier transform infrared (FTIR) and XPS spectroscopy probed interaction sites. FTIR analysis of D18-Cl films reveals MoPS3 nanocrystal selectively coordinates chlorine atoms via S···Cl halogen bonding, marked by attenuation of the C-Cl vibrational mode at 824 cm−1 (Fig. 5a; Supplementary Fig. 35). This aligns with the decreased Cl 2p binding energy (from 200.7 eV to 200.4 eV) and S 2 P binding energy (from 164.2 eV to 163.9 eV) (Supplementary Fig. 36). Such position-selective coordination promotes D18-Cl crystallinity through nanocrystal-directed molecular alignment. Conversely, L8-BO films exhibit complete C = O peak suppression at 1159 cm−1 alongside a broad emergent feature at 1103 cm−1 (Fig. 5a; Supplementary Fig. 37), coupled with a 1.2 eV O 1 s binding energy decrease, directly evidencing Mo-O coordination between Mo6+ sites in MoPS3 nanocrystal and oxygen atoms in L8-BO. Significantly, a synchronous >1 eV decreases in S 2p, C 1 s, N 1 s, and F 1 s binding energies (Supplementary Fig. 38) upon the incorporation of MoPS3 nanocrystal suggests bidirectional charge transfer: initial electron donation from L8-BO oxygen to Mo6+ is partially back-transferred to L8-BO’s π-system via Mo-S bonds, elevating its electron density.
a FTIR for control and MoPS3 nanocrystal doped D18-Cl (or L8-BO) films. b Two-dimensional GIWAXS patterns. c One-dimensional integrated scattering profiles, π-π stacking distance and CCL for the corresponding films. d The corresponding 2D in situ absorption spectra during spin-coating. e PiFM images and line profiles acquired at different wavelengths. f Statistical graph of fibril widths for D18-Cl:L8-BO and D18-Cl:L8-BO:MoPS3 films at different wavelengths.
The impact on molecular packing was investigated using two-dimensional grazing incidence wide-angle X-ray scattering (GIWAXS). MoPS3 incorporation reduced the out-of-plane (OOP) π-π stacking distance for both D18-Cl (from 4.142 Å to 4.009 Å) and L8-BO (from 3.959 Å to 3.944 Å) and increased their crystal coherence lengths (CCL), indicating denser, more ordered packing (Supplementary Figs. 39–41; Supplementary Tables 16−17)42,43. This ordering persisted in the blend film (Fig. 5b). Both control and MoPS3-doped blend films exhibited a preferred face-on orientation. MoPS3 sharpened and intensified the OOP (010) π-π stacking peak (qz from 1.546 Å−1 to 1.577 Å−1), corresponding to distances of 4.064 Å and 3.985 Å, respectively, and increased the CCL from 2.719 nm to 3.821 nm (Supplementary Table 18). The reduced π-π distance and increased CCL confirm enhanced π-orbital coherence and crystallinity (Fig. 5c). Spectroscopic signatures (vibronic coupling ratio44, PL intensity45 further corroborated improved molecular ordering in both components (Supplementary Fig. 42), consistent with AFM and TEM results (Supplementary Fig. 43).
In situ absorption spectroscopy reveals distinct film formation dynamics during spin coating of D18-Cl:L8-BO and D18-Cl:L8-BO:MoPS3 blends (Fig. 5d). The pristine film shows a rapid rise and red shift of the long-wavelength absorption within ~0.5 s, indicating accelerated acceptor aggregation and abrupt solidification, while donor-related absorption evolves more gradually. In contrast, the MoPS3-containing film exhibits a smoother and delayed absorption evolution across the entire spectral range, with suppressed abrupt transitions during the critical solidification stage (Supplementary Fig. 44). This moderated kinetics suggests that MoPS3 retards excessive acceptor aggregation and synchronizes donor–acceptor ordering, thereby extending the molecular rearrangement window and promoting more homogeneous film formation. Photo-induced force microscopy (PiFM) was used to probe the nanoscale chemical morphology of the active layers (Fig. 5e). The pristine D18-Cl:L8-BO film exhibits relatively coarse fibrillar networks at both 1455 cm−1 (L8-BO) and 1532 cm−1 (D18-Cl), with average fibril diameters of ~25 nm, indicative of less confined phase separation. Upon incorporation of MoPS3, both donor- and acceptor-rich fibrils become significantly thinner ( ~ 21 nm) with narrower diameter distributions (Fig. 5f; Supplementary Fig. 45), reflecting a more uniform and finely interpenetrating morphology. Such refined domain sizes are expected to reduce carrier trapping and non-radiative recombination at donor–acceptor interfaces, consistent with the suppressed non-radiative losses.
To validate universal thickness tolerance, we incorporated MoPS3 into D18:L8-BO and PM6:Y6 blends. MoPS3 incorporation significantly enhanced saturation magnetization in both systems (Fig. 6a,b; Supplementary Fig. 46). Phosphorescence measurements confirmed that MoPS3 markedly reduced ΔEST for all components (D18, L8-BO, PM6, Y6), promoting efficient ISC (Supplementary Figs. 47–49; Tables 19–20). ECT analysis and Urbach energy decreases confirmed suppressed non-radiative losses and energetic disorder in both MoPS3-modified systems (Fig. 6c; Supplementary Figs. 50–51; Supplementary Tables 21–22). In both control systems, ECT lies below the ET of donor and acceptor, permitting loss via triplet back-transfer. MoPS3 elevates the donor triplet energy above ECT, effectively blocking this channel (Fig. 6d). SCLC analysis confirmed substantial charge mobility enhancements in both systems following MoPS3 integration (Supplementary Figs. 52–55; Supplementary Tables 23–24). TPV measurements provided longer decay times and correspondingly longer calculated LD values for all components (D18, PM6, L8-BO, Y6) in MoPS3-doped devices (Fig. 6e; Supplementary Fig. 56–57). These results demonstrate the universal applicability of MoPS3 for promoting ISC and extending exciton diffusion length.
a, b Magnetic hysteresis loop. SD, saturation magnetization. c EQEEL spectrum. d The corresponding S1, T1, CT and probable charge transfer between D18 (or PM6) and L8-BO (or Y6) with/ without MoPS3 nanocrystal. e Calculated exciton diffusion length. f, g J–V curves for control and MoPS3 nanocrystal doped D18:L8-BO (or PM6:Y6) devices.
Photovoltaic metrics for devices across 100–300 nm active layers are shown in J–V curves (Fig. 6f-g; Supplementary Figs. 58–59). The reference D18:L8-BO device (100 nm active layer) delivered a PCE of 19.92%, with a VOC of 0.92 ± 0.01 V, JSC of 27.05 ± 0.71 mA cm−2, and FF of 79.27 ± 0.74%. Performance dropped to 17.59% at an active layer thickness of 300 nm due to FF attenuation (70.21 ± 0.82%). The 0.6 wt% MoPS3 nanocrystal incorporated D18:L8-BO device demonstrated optimal performance with an active layer of about 100 nm (PCE = 20.91%, VOC = 0.92 ± 0.01 V, JSC = 27.69 ± 0.57 mA cm−2, FF = 81.02 ± 0.78%; Supplementary Table 25), as well as maintained a PCE of 19.63% at an active layer thickness of 300 nm (with VOC of 0.92 ± 0.01 V, JSC of 27.82 ± 0.54 mA cm−2, and FF of 76.93 ± 0.67%; Supplementary Table 26), exhibiting excellent film thickness insensitivity. Similarly, PM6:Y6 devices retained 17.92% PCE at 300 nm active-layer thickness (vs. 19.13% at 100 nm active-layer thickness) with 0.6 wt% MoPS3 nanocrystal integration (Supplementary Tables 27 and 28).
In summary, we synthesized a novel two-dimensional magnetic MoPS3 nanocrystal and demonstrated its efficacy as a multifunctional additive in overcoming the thickness-performance trade-off in OSCs. Through site-specific coordination, MoPS3 optimizes molecular packing and simultaneously creates a weak intrinsic magnetic field within the active layer, thus significantly accelerates the ISC from short-lived singlets to long-lived triplets, via enhancing SOC. This extends the LD and, together with an elevated triplet energy landscape, suppresses non-radiative recombination losses. Consequently, MoPS3-doped D18-Cl:L8-BO devices achieve a high PCE of 20.37% at 100 nm and, critically, retain a PCE of 19.36% at 300 nm, representing the highest reported values for 300 nm thick-film OSCs. Crucially, this strategy demonstrates universal thickness tolerance across diverse OSC architectures, for which the PCE of MoPS3 nanocrystal incorporated D18:L8-BO and PM6:Y6 device only decreased from 20.91% and 19.13% to 19.63% and 17.92%, respectively, as the active-layer thickness boosted from 100 to 300 nm. This MoPS3 nanocrystal integration pioneers a rational materials engineering pathway to decouple efficiency from thickness constraints, advancing scalable OSC manufacturing.
Organic semiconductors, including PM6, Y6, L8-BO, and PNDIT-F3N, were purchased from Nanjing Zhiyan Technology Co., Ltd. D18 was obtained from Shenzhen Ruixun Optoelectronic Material Technology Co., Ltd. D18-Cl was sourced from eFlexPV Limited Tech. Co., Ltd. High-purity solvents (e.g., chloroform, CF) and material 2PACz were procured from Sigma-Aldrich.
Patterned indium tin oxide (ITO)-coated glass substrates were cleaned sequentially with detergent, deionized water, acetone, and isopropyl alcohol under sonication (20 min each), followed by overnight drying in an oven. After 5 min of oxygen-plasma treatment, a 2PACz solution in ethanol was spin-coated onto the ITO substrates at 3000 rpm for 30 s and then annealed at 100 °C for 8 min. For the D18-Cl:L8-BO system, the donor:acceptor (D:A) mass ratio was optimized to 1:1.2 (w/w). The concentration of D18-Cl was fixed at 6 mg/mL in chloroform. Active layers with nominal thicknesses of ~100 nm and ~300 nm were deposited via spin-coating at 3500 rpm and 2000 rpm for 30 s, respectively. For the PM6:Y6 system, the optimized D:A mass ratio was 1:1.5 (w/w), with a PM6 concentration of 7 mg/mL in chloroform. Spin-coating at 4000 rpm and 2500 rpm for 30 s yielded films of ~100 nm and ~300 nm, respectively. For the D18:L8-BO system, the optimized D:A mass ratio was 1:1.4 (w/w), with a D18 concentration of 3.5 mg/mL in chloroform. Films with thicknesses of ~100 nm and ~300 nm were obtained by spin-coating at 2000 rpm and 1300 rpm for 30 s, respectively. A 30-second spin-coating process at 2000 rpm forms a uniform thin film around 100 nm thick, whereas spin-coating at 1300 rpm for 30 seconds produces a homogeneous film roughly 300 nm in thickness. Subsequently, a methanol solution of PNDIT-F3N (0.5 mg/mL) was spin-coated at 2000 rpm for 30 s to form the cathode interlayer. Finally, a 100 nm-thick Ag electrode was thermally evaporated under high vacuum ( < 1×10−4 Pa). The effective device area was 0.0516 cm2, which was further defined as 0.04 cm2 using a square metal mask.
Hole-only and electron-only devices with structures of ITO/2PACz/active layer/Ag and ITO/ZnO/active layer/PNDIT-F3N/Ag, respectively, were fabricated to evaluate charge-carrier mobility. Dark current J–V characteristics were measured using a Keithley 2450 source meter. Mobilities were extracted by fitting the SCLC region using the equation: (J=(9/8){varepsilon }_{0}{varepsilon }_{r}mu {{V}_{{eff}}}^{2}{/d}^{3}), where the parameters are defined as follows: J is the current density, μ is the zero-field charge mobility, ε0 is the permittivity of free space, εᵣ is the relative permittivity of the material, d is the thickness of the active layer, and Veff is the effective voltage.
UV–vis–NIR absorption and PL spectra were recorded on an APL-AS II spectral analysis system. J–V characteristics of photovoltaic devices were measured under simulated AM 1.5 G illumination using a Zolix Sirius-SS150A solar simulator and a Keithley 2450 source meter, calibrated with a certified silicon reference cell. EQE spectra were obtained using a Zolix-built EQE measurement system calibrated with a certified silicon reference cell. TA spectra were acquired by probing differential transmission before and after excitation with an 805 nm pump pulse. TPC and TPV were measured using a CEL-TPV2000 system. AFM images were taken on a Bruker Multimode 8 microscope in tapping mode. TEM was performed on a JEOL JEM-F200 microscope. XRD patterns of MoPS3 nanocrystals were collected on a Bruker D8 Advance diffractometer using Cu Kα radiation (λ = 0.15418 nm). GIWAXS measurements were conducted on a Xeuss 2.0 system (λ = 1.54189 Å). XPS was performed on a Thermo Fisher Nexsa spectrometer. In-situ absorption spectrum were collected using a Maya2000 Pro spectrometer (Ocean Insight) with illumination from a xenon lamp source. PiFM results were acquired using a VistaScope microscope from Molecular Vista inc. from Songshan Lake Materials Laboratory. PiFM experiments were excited by a pulsed quantum cascade laser (Block Engineering) with a gap-free narrowband tunable wavenumber of 760-1950 cm−1. The spectral linewidth is ~2 cm−1 with a wavenumber resolution of 0.5 cm−1.
SOC calculations are performed using the Vienna ab initio simulation package (VASP) by DFT theory calculations. The exchange correlation functional is parametrized by the Perdew, Burke, and Ernzerhof (PBE) model. For the structural relaxation, a cutoff energy of 450 eV is used. A 3×3×1 Monkhorst-Pack K-mesh is used to sample the two-dimensional Brillouin zone. The energy convergence threshold is 1.0 × 10−5 eV and the criteria for the convergence of the Hellmann-Feynman forces on the atoms is 0.02 eV/Å.
AIMD simulations were performed using the CP2K software package to investigate the structural evolution of D18-Cl and L8-BO molecules in the presence and absence of MoPS3. The E between D18-Cl and L8-BO were computed for both scenarios. A sufficiently large vacuum thickness was applied along the non-periodic direction to eliminate spurious interactions between periodic images, ensuring that the calculated results were solely governed by the sampled k-points in reciprocal space. Electronic structure calculations were conducted within the framework of DFT using CP2K. The Perdew–Burke–Ernzerhof (PBE) exchange–correlation functional was employed to describe electron–electron interactions. All simulations utilized a double-ζ valence polarized basis set (DZVP-MOLOPT-SR-GTH) for geometry optimizations and energy evaluations of the D18-Cl and L8-BO systems, with and without MoPS3. The optical absorption spectra and electronic excited-state properties were investigated using TDDFT calculations in CP2K with the same DZVP-MOLOPT-SR-GTH basis set. A total of 50 excited states were computed under periodic boundary conditions. Subsequent wave function analysis for the excited states was carried out using the Multiwfn software package.
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
The data that support the findings of this study are available within the article and its Supplementary Information/Source Data file. Source data are provided with this paper.
Wang, Y. et al. Fully non-fused electron acceptor solar cells with 18% efficiency via a synergistic peripheral substituent strategy. Nat. Commun. 16, 5449 (2025).
Article ADS CAS PubMed PubMed Central Google Scholar
Chen, X. et al. Siloxane-tethered halogenated thiophene additives enable excellent fibrillar morphology and air processing for high-performing polymer solar cells. Adv. Mater. 37, e12197 (2025).
Article CAS PubMed Google Scholar
Hung, C.-M. et al. Interspersed assembled monolayers enhance hole transport in high-efficiency organic and perovskite solar cells. J. Am. Chem. Soc. 147, 23683–23695 (2025).
Article ADS CAS PubMed PubMed Central Google Scholar
Guo, J. et al. Asymmetric small-molecule acceptor enables suppressed electron-vibration coupling and minimized driving force for organic solar cells. Nat. Comm. 16, 1503 (2025).
Article ADS CAS Google Scholar
Dai, T. et al. Modulation of molecular quadrupole moments by phenyl side-chain fluorination for high-voltage and high-performance organic solar cells. J. Am. Chem. Soc. 147, 4631–4642 (2025).
Article ADS CAS PubMed Google Scholar
Sun, F. et al. Constructing continuous acceptor fibrillar networks and achieving uniform phase separation via polymer-assisted morphology control for 20.3% efficiency additive-free organic solar cells. Energy Environ. Sci. 18, 7071–7081 (2025).
Article CAS Google Scholar
He, H. et al. Dynamic hydrogen-bonding enables high-performance and mechanically robust organic solar cells processed with non-halogenated solvent. Nat. Comm. 16, 787 (2025).
Article ADS CAS Google Scholar
Xu, C. et al. A bithiophene imide-based polymer donor for alloy-like ternary organic solar cells with over 20.5% efficiency and enhanced stability. Energy Environ. Sci. 18, 5913–5925 (2025).
Article CAS 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 ADS CAS Google Scholar
Li, C. et al. Non-fullerene acceptors with high crystallinity and photoluminescence quantum yield enable >20% efficiency organic solar cells. Nat. Mater 24, 433–443 (2025).
Article ADS CAS PubMed Google Scholar
Li, C. et al. Organic solar cells with 21% efficiency enabled by a hybrid interfacial layer with dual-component synergy. Nat. Mater. 24, 1626–1634 (2025).
Article ADS CAS PubMed Google Scholar
Zhou, W. et al. Solvent vapor diffusion–driven multiscale pre-aggregation of non-fullerene acceptors enables high-performance organic solar cells. Nat. Comm. 16, 11188 (2025).
Article ADS CAS 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 ADS CAS PubMed Google Scholar
Sun, F. et al. Precise control over crystallization kinetics by combining nucleating agents and plasticizers for 20.1% efficiency organic solar cells. Adv. Mater. 37, 2500357 (2025).
Article Google Scholar
Wang, X. et al. Elevating dielectric constant via additive engineering: Achieving 19.23% certified efficiency in thick-film binary organic solar cells. Joule 9, 102135 (2025).
Article CAS Google Scholar
Ge, Z. et al. Balanced electron and hole transfer behavior enabled approaching 19% efficiency in thick-film organic solar cells with improved fill factor. Energy Environ. Sci. 18, 6667–6675 (2025).
Article CAS Google Scholar
Chen, L. et al. High efficiency (∼18%) organic solar cells with 500 nm-thick toluene cast active layer by aggregation manipulation and additive engineering. Adv. Mater. 38, e08209 (2026).
Article CAS PubMed Google Scholar
Meng, Y. et al. Critical length screening enables 19% efficiency in thick-film organic solar cells. Nat. Comm. 16, 9840 (2025).
Article Google Scholar
Shi, B. et al. High-Performance thick-film all-polymer solar cells enabled by a blade-coating process assisted by a direct nonuniform electric field. Adv. Mater. 37, e05313 (2025).
Article CAS PubMed Google Scholar
Chen, X. et al. Efficient perovskite/organic tandem photovoltaic devices and large-area modules featuring thick-film organic solar cells. Adv. Mater. 37, 2500190 (2025).
Article CAS Google Scholar
Cao, C. et al. Pressure-modulated molecular stacking strategy extends exciton diffusion length for thick-film (300 nm) organic photovoltaics exceeding 19% efficiency. Adv. Mater. 38, e14047 (2026).
Article CAS PubMed Google Scholar
González, D. M. et al. Improved power conversion efficiency of P3HT:PCBM organic solar cells by strong spin–orbit coupling-induced delayed fluorescence. Adv. Energy Mater. 5, 1401770 (2015).
Article Google Scholar
Xu, B. et al. Phosphorescence and delayed fluorescence of poly(3-hexylthiophene) films. Thin Solid Films 242, 174–177 (1994).
Article ADS CAS Google Scholar
Luppi, B. T. et al. Triplet excitons: improving exciton diffusion length for enhanced organic photovoltaics. J. Mater. Chem. A 7, 2445–2463 (2019).
Article CAS Google Scholar
Ohno, H. A window on the future of spintronics. Nat. Mater. 9, 952–954 (2010).
Article ADS CAS PubMed Google Scholar
Blom, P. W. et al. Device physics of polymer:fullerene bulk heterojunction solar cells. Adv. Mater. 19, 1551–1566 (2007).
Article CAS Google Scholar
Cowan, S. R. et al. Charge formation, recombination, and sweep-out dynamics in organic solar cells. Adv. Funct. Mater. 22, 1116–1128 (2012).
Article CAS Google Scholar
Shuttle, C. G. et al. Bimolecular recombination losses in polythiophene: Fullerene solar cells. Phys. Rev. B 78, 113201 (2008).
Article ADS Google Scholar
Hwang, J. G. et al. Effects of nanoparticle charging on streamer development in transformer oil-based nanofluids. J. Appl. Phys. 107, 014310 (2010).
Article ADS Google Scholar
Desai, P. et al. Magnetoresistance in organic light-emitting diode structures under illumination. Phys. Rev. B 76, 235202 (2007).
Article ADS Google Scholar
Liu, S. et al. High-efficiency organic solar cells with low non-radiative recombination loss and low energetic disorder. Nat. Photon. 14, 300–305 (2020).
Article ADS CAS Google Scholar
Li, Y. et al. Solution-processable pedot for organic solar cells: From one-pot synthesis to kinetically-controlled polymerization. Adv. Mater. 37, 2502394 (2025).
Article CAS Google Scholar
Credgington, D. et al. Insights from transient optoelectronic analyses on the open-circuit voltage of organic solar cells. J. Phys. Chem. Lett. 3, 1465 (2012).
Article CAS PubMed Google Scholar
Li, Z. et al. Regulating polymeric adsorption dynamics via monolayer-architected hydrotalcite integration for high-performance organic solar cells. Adv. Mater. 37, 2508049 (2025).
Article CAS Google Scholar
Wu, J. et al. Balance processing and molecular packing via structural disordering in a random terpolymer for over 19% efficiency non-halogenated solvent organic solar cells. Adv. Energy Mater. 15, 2500024 (2025).
Article CAS Google Scholar
Hu, H. et al. Revealing electron transport connectivity as an important factor influencing stability of organic solar cells. Nat. Commun. 16, 6546 (2025).
Article ADS CAS PubMed PubMed Central Google Scholar
Dong, J. et al. Dielectric constant engineering of nonfullerene acceptors enables a record fill factor of 83.58% and a high efficiency of 20.80% in organic solar cells. Energy Environ. Sci. 18, 4982–4995 (2025).
Article CAS Google Scholar
Deng, J. et al. Acceptor crystallinity engineering enables >20% efficiency binary organic solar cells with 83.0% fill factor. Adv. Mater 37, 2501243 (2025).
Article CAS Google Scholar
Xie, X. et al. Thickness insensitive uv blocking layer meliorating carrier extraction and deep trap towards stable organic solar cells. Adv. Funct. Mater. 35, 2420940 (2025).
Article CAS Google Scholar
Li, Z. et al. Monolayer MnPS3 nanosheet integration for extended exciton diffusion and charge transport in high-performance thickness-insensitive organic photovoltaics. Adv. Mater. 37, e13125 (2025).
Article CAS PubMed Google Scholar
Yan, J. et al. Influence of static disorder of charge transfer state on voltage loss in organic photovoltaics. Nat. Commun. 12, 3642 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Gao, Y. et al. Highly efficient all-small-molecule organic solar cells with excellent operational stability and blend-thickness tolerance. Energy Environ. Sci. 18, 7302–7312 (2025).
Article CAS Google Scholar
Gao, C. et al. Prolonged-nucleation strategy via an asymmetric brominated acceptor enables > 20% efficiency in five different organic solar cells. Adv. Mater 38, e17576 (2026).
Article CAS PubMed Google Scholar
Zhao, Q. et al. H- and J-aggregation of conjugated small molecules in organic solar cells. J. Energy Chem. 93, 174–192 (2024).
Article CAS Google Scholar
Wan, J. et al. All annealing-free solution-processed highly flexible organic solar cells. J. Mater. Chem. A. 9, 5425–5433 (2021).
Article CAS Google Scholar
Download references
This work was financially supported by the National Natural Science Foundation of China: 62404094 (Z. L.), the Natural Science Foundation of Hunan Province: 2025RC3191 (H. L.), 2023JJ40532 (Z. L.), and the Fund of University of South China: 210XQD018 (Z. L.), 5524GC017 (H. L.).
College of Mechanical Engineering & School of Electrical Engineering, University of South China, Hengyang, China
Zhenye Li, Xinyu Pu, Zhaoxiong Su, Maoting Chen, Haiyang Wang, Yuhui Qin, Xingjie Mi & Hanjian Lai
School of Mathematics and Physics, University of South China, Hengyang, China
Yu-Feng Ding
Dongguan Key Laboratory of Interdisciplinary Science for Advanced Materials and Large-Scale Scientific Facilities, School of Physical Sciences, Great Bay University, Dongguan, Guangdong, China
Guoli Wang & Sha Liu
Guangdong Basic Research Center of Excellence for Aggregate Science, School of Science and Engineering, The Chinese University of Hong Kong (Shenzhen), Shenzhen, Guangdong, China
Liming Liu & Jun Yan
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
PubMed Google Scholar
Z. L. conceived the idea, designed the experiment, analyzed the experimental data, as well as prepared and revised the manuscript. X. P. prepared OSC devices (or samples) and carried out measurements. Z. S. carried out AIMD and TDDFT calculations. Y.-F. D. carried out SOC calculations. M. C., H. W., Y. Q., and X. M. assisted in OSC sample preparation. G. W. and S. L. carried out the energy loss and temperature-dependent PL spectrum measurement. L. L. and J. Y. fitted the temperature-dependent PL spectrum data. Z. L. and H. L. supervised the project. All authors commented on the manuscript.
Correspondence to Zhenye Li, Yu-Feng Ding, Sha Liu or Hanjian Lai.
The authors declare no competing interests.
Nature Communications thanks Guangye Zhang, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
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
Li, Z., Pu, X., Su, Z. et al. Spin-manipulation via novel MoPS3 nanocrystal for high-performance thick-film organic solar cells. Nat Commun 17, 2330 (2026). https://doi.org/10.1038/s41467-026-70320-7
Download citation
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-026-70320-7
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
Nature Communications (Nat Commun)
ISSN 2041-1723 (online)
© 2026 Springer Nature Limited
Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.
