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 16, Article number: 8986 (2025)
5182
6
Metrics details
Perovskite/organic tandem solar cells are a promising strategy to surpass the efficiency limits of single-junction devices, yet their performance is restricted by recombination losses in the organic subcells. Here, we investigate these losses by tracking film evolution from the very initial stage of organic film formation. We strategically manipulated donor and acceptor ratios to modulate film growth characteristics, while employing in situ techniques to monitor the real-time crystallization dynamics. Our research findings underscore that the variance in donor content within the organic blend exerts a fine-tuning effect on the solution-to-solid transformation process. When the donor content is inadequate, the acceptor molecules tend to aggregate, disrupting molecular packing and lowering crystallinity. These morphological changes hinder exciton dissociation, thereby leading to charge recombination and deteriorating overall device performance. Optimizing film morphology and crystallization reduces recombination losses, enabling perovskite/organic tandem solar cells with a record 26.42% power conversion efficiency.
In recent years, as the efficiency of single-junction devices has progressively approached their theoretical limits, research on tandem solar cells (TSCs) has garnered increasing attention, emerging as a focal area of investigation. The bandgap of perovskite semiconductors can be broadly tuned through compositional engineering, making them ideal for tandem applications. Wide-bandgap perovskites are particularly suited to pair with narrow-bandgap subcells such as silicon, CIGS, low-bandgap perovskites and organic solar cells (OSCs)1,2. Among the diverse array of tandem configurations, perovskite/organic TSCs have rapidly risen to prominence3,4,5. A notable advantage of perovskite/organic TSCs lies in the excellent process compatibility between perovskite and organic materials, coupled with the orthogonal nature of their processing solvents. This characteristic effectively precludes mutual damage between the two material types during the device fabrication process, thereby facilitating the construction of high-efficiency tandem devices6,7. By synergistically integrating the strengths of perovskite solar cells (PSCs) and OSCs, the tandem cells employ wide-bandgap PSCs as the front subcells, tasked with capturing short-wavelength light. Conversely, narrow-bandgap OSCs serve as the rear subcells, specializing in harvesting long-wavelength light8. This complementary design paradigm enables a more comprehensive exploitation of the solar spectrum, effectively surmounting the performance limitations inherent in single-junction solar cells.
A multitude of strategies have been devised to enhance the efficiency of perovskite/organic TSCs, with a particular focus on optimizing film composition and device structures9,10,11. Up to now, the power conversion efficiency (PCE) of perovskite/organic TSCs has surpassed 26%, while their progress still lags behind that of other tandem architectures, such as perovskite-silicon and all-perovskite ones12. The primary impediment stems from the relatively pronounced recombination losses incurred within OSC subcells, which contribute to the observed efficiency gap13,14. In contrast to the free charge carriers prevalent in inorganic semiconductors, photogenerated electrons and holes in organic domains are typically confined as localized excitons. Given the substantial exciton binding energy, excitons have to traverse the interface between the donor and acceptor for dissociation and collection in OSCs. At this point, recombination losses would be inevitable.
In the OSC field, the amelioration of recombination losses has been extensively investigated, ranging from donor and acceptor synthesis to device fabrication15,16,17,18. It is widely recognized that film morphology and crystallization are two key determinants. Specifically, an interpenetrating fiber network ensures more efficient exciton dissociation, while ordered molecular crystallization facilitates charge transport19,20. On the other hand, during the formation process of active layers, the polymer donor can influence the growth dynamics of the interpenetrating network. Typically, the polymer donor precipitates first from the solution, followed by the subsequent crystallization and growth of the small-molecular acceptor around the polymer donor fibers21,22. Consequently, the growth behavior of donor and acceptor materials constitutes the most direct determinant of the final film characteristics. A comprehensive understanding of the film growth evolution of donor and acceptor materials and its impact on the corresponding crystallization features is paramount for the amelioration of recombination losses, thereby enabling the realization of efficient organic subcells and corresponding perovskite/organic TSCs.
In this study, we conducted a systematic investigation into the influence of film evolution on the corresponding crystallization quality, recombination losses, and device efficiencies. Employing in situ techniques such as grazing-incidence wide-angle X-ray scattering (GIWAXS), ultraviolet-visible (UV-vis) absorption spectroscopy, and photoluminescence (PL) spectroscopy, we scrutinized the morphological evolution of the films. Concurrently, we examined the exciton/charge recombination losses under varying film conditions. Our findings revealed that the donor:acceptor (D:A) content plays a pivotal role in modulating the growth dynamics of the films and their crystallization quality, thereby altering the exciton dissociation and recombination behavior. By judiciously regulating the film characteristics, we effectively mitigated exciton recombination losses and enhanced the corresponding device efficiency. When different OSCs were employed as rear subcells in conjunction with wide-bandgap perovskite front subcells, a champion PCE of 26.42% was achieved in perovskite/organic TSCs.
The molecular structures of the polymer donor material D18 and the small-molecule acceptor BTP-eC9-4F are illustrated in Supplementary Fig. 1a. Supplementary Fig. 1b depicts the UV-vis absorption spectra of D18:BTP-eC9-4F blend films as the D:A weight ratio is varied from 1:1.2 to 0.2:1.2. With the decreasing content of D18 in the blend films, the absorbance of D18 within the 420-630 nm wavelength range gradually diminishes. The absorption peak of BTP-eC9-4F undergoes a progressive redshift, shifting from 810 nm at a 1:1.2 D:A ratio to 825 nm at a 0.2:1.2 ratio. This phenomenon is attributed to the aggregation of BTP-eC9-4F molecules23. To elucidate the aggregation behavior of blend films with varying D18 contents, two-dimensional (2D) GIWAXS measurements were conducted24. The measurement outcomes are presented in Fig. 1a. The out-of-plane (OOP) and in-plane (IP) line-cut profiles extracted from the corresponding 2D GIWAXS patterns are displayed in Fig. 1b, while the lattice parameters are summarized in Supplementary Table 1. All films exhibit a prominent out-of-plane (010) diffraction peak, indicating a preferred face-on orientation. As shown in the GIWAXS profiles, the intensity of the π-π stacking peak in the OOP direction gradually decreases with decreasing donor content. This phenomenon is primarily attributed to the reduction in film thickness (Supplementary Fig. 2). In the out-of-plane direction, as the donor content decreases, the crystalline coherence length (CCL) gradually shortens (with a CCL of 22.0 Å for the 1:1.2 film and 16.0 Å for the 0.2:1.2 film). This is likely due to inadequate interactions between the donor and acceptor, which hinder the effective assembly of molecules into long-range ordered crystalline structures. Simultaneously, in the IP direction, as the donor content decreases, the lamellar stacking distance reduces from 20.7 Å at a 1:1.2 ratio to 20.0 Å at a 0.2:1.2 ratio, and the CCL decreases from 59.9 Å at a 1:1.2 ratio to 28.8 Å at a 0.2:1.2 ratio. These findings imply that an excessive reduction in donor content compromises the molecular packing and the crystallinity quality of the blend films. To further investigate the evolution of molecular orientation, we examined the azimuthal intensity distribution of the (100) peak, where the edge-on and face-on regions were defined by the azimuthal angle (χ): 0° < χ < 30° corresponds to edge-on orientation, 30° < χ < 60° corresponds to random orientation, and 60° <χ < 90° corresponds to face-on orientation25,26,27,28. Due to reshaping effects caused by the missing region in the reciprocal space of the 2D GIWAXS data, the intensity distribution curve was incomplete near 0°29,30. By integrating the signal over 0-30°, 30-60°, and 60-90°, the relative proportions of edge-on, random, and face-on orientations were obtained (as shown in Supplementary Fig. 3 and Supplementary Table 2). With decreasing donor content, the edge-on fraction increases from 26.7% (1:1.2) to 35.2% (0.2:1.2), random orientation increases from 27.4% to 31.3%, while face-on orientation decreases from 45.9% to 33.5%. These results suggest that a reduction in donor content not only weakens crystallization driving forces but also disrupts favorable molecular orientations, particularly the face-on stacking that is often correlated with efficient charge transport and ordered domains. This dual effect ultimately hinders the formation of long-range crystalline structures within the active layer.
a GIWAXS patterns of D18:BTP-eC9-4F films with different D:A content. b OOP (solid line) and IP (dotted line) line cuts extracted from the GIWAXS patterns.
To better understand the microstructural morphology of the active layer, we conducted further characterization using atomic force microscopy (AFM) (Supplementary Fig. 4) and transmission electron microscopy (TEM) (Supplementary Fig. 5). The AFM results revealed that as the donor content decreases, the root mean square (RMS) roughness of the films progressively decreases, from 0.721 nm at the 1:1.2 ratio to 0.561 nm at the 0.2:1.2 ratio. TEM images reveal that the active layer at a 1:1.2 ratio exhibits pronounced fibrous structures, which gradually diminish as the donor content decreases. At a D:A ratio of 0.2:1.2, the fibrous aggregated domains become barely observable.
To further explore crystallization kinetic behavior during solution-solid film formation, in situ 2D GIWAXS measurements were employed25,31. The OOP (010) peak of the blend film was analyzed under various D18 contents. According to Supplementary Fig. 6, this analysis aimed to track the evolution of π-π stacking distance and CCL throughout the film formation process. As illustrated in Fig. 2, the drying process is divided into four stages32, color-coded as follows: stage (I) solution state, stage (II) crystal nucleation, stage (III) crystal growth, and stage (IV) solid state. In stage I, no discernible (010) scattering peaks were observed. In stage II, as the solvent evaporates and the solution concentration increases, the donor and acceptor reach their solubility limit, initiating nucleation. This leads to the appearance of the (010) scattering peak. As solvent evaporation continues, molecular packing becomes more compact, leading to a gradual decrease in the π-π stacking distance. These results suggest that aggregates formed in an orderly manner during this stage. In stage III, at this point, most of the solvent has evaporated, crystallization is complete, and the π-π stacking distance stabilizes. Simultaneously, the movement and rearrangement of molecules lead to an increase in CCL. In stage IV, all solvents have evaporated, and the film morphology is stabilized, with all morphological parameters remaining constant. As the donor content decreases, the duration of each of the four stages shortens to varying degrees (Fig. 2f). For the D:A ratios of 1:1.2, 0.8:1.2, 0.6:1.2, 0.4:1.2, and 0.2:1.2, the duration of stage II remained relatively constant, with values of 1.25, 1, 1, 1, and 0.75 s, respectively. In contrast, the duration of stage III exhibited significant variation (5.5, 5, 2.75, 2, and 1.5 s for the corresponding ratios), indicating that the variation in donor content primarily influenced the crystal growth process. The slope of the CCL and the time taken for the solution to form a film can reflect the relative crystallization rate in the drying process33. The crystallization rates of the CCL corresponding to the (010) peak were 1.7, 1.8, 3.3, 3.9, and 6.1 s−1 for the corresponding D:A ratios. A steeper slope was observed with decreasing donor content, indicating faster CCL evolution in donor-deficient systems.
The d-spacing and CCL evolution of (010) peak for blend films with D:A ratio of a 1:1.2, b 0.8:1.2, c 0.6:1.2, d 0.4:1.2 and e 0.2:1.2 obtained from in situ GIWAXS measurements. f The abridged graph of the film-forming process.
To understand more comprehensively the dynamic evolution of molecular assembly, phase separation, and crystal growth during the transition of thin films from the solution state to the solid-state thin films, we further employed in situ UV-vis absorption and PL spectra characterization techniques. In situ UV-vis absorption spectroscopy was employed to investigate the growth evolution of blend films with different D18 content. The 2D contour maps of D18:BTP-eC9-4F (ratio change from 1:1.2 to 0.2:1.2) from solution to film status are presented in Supplementary Fig. 7 (The time point t = 0 denotes the moment when the signal first appears). The evolution of the absorption peak location represents the aggregation evolution of the donor and the acceptor34,35. The first stage is the solvent evaporating stage, followed by the nucleation and crystal growth stage, and the final one is the dried film stage. The variation in the peak location of D18 was negligible with time changes (Supplementary Fig. 8a), indicating a fast domain formation feature of D18 molecules. On the contrary, as shown in Fig. 3a, the peak location of individual BTP-eC9-4F has a redshift from ~730 nm (solution state) to over 800 nm (film state) in the evolution process, indicating the domain formation speed of BTP-eC9-4F molecules is slower than that of D18 ones. From Supplementary Fig. 8b and Fig. 3b, the absorbance of both D18 and BTP-eC9-4F decreases sharply at the beginning period (solvent evaporates), and continues to decrease as time goes on, and finally reaches its solubility limit and then nucleates36. Under extreme supersaturation, the BTP-eC9-4F absorption peak starts to redshift, indicating the onset of the crystallization process. To quantify this process, the peak locations of BTP-eC9-4F at the onset and completion of the nucleation and crystal growth stage were extracted with varying donor content, as shown in Fig. 3c. With decreasing D18 content, the crystallization process is initiated earlier and completed within a shorter period. To gain deeper insight into the crystallization dynamics, the nucleation and crystal growth stages in the in situ UV-vis spectra could be further separated into the aggregation/nucleation stage and the crystal growth stage37. This allowed for quantitative comparison of the duration of molecular aggregation and crystallization across different D:A ratios. As shown in Supplementary Fig. 9, for 1:1.2, 0.8:1.2, 0.6:1.2, 0.4:1.2, and 0.2:1.2 blend films, the aggregation and crystallization duration were 0.20 s (0.15 s), 0.15 s (0.15 s), 0.10 s (0.15 s), 0.10 s (0.15 s), and 0.10 s (0.10 s), respectively. These results indicate that at higher donor content, aggregation occurs before crystallization, consistent with a phase separation-induced crystallization mechanism. At lower donor ratios (e.g., 0.2:1.2), the two processes become nearly concurrent, suggesting a shift toward a more simultaneous or competing regime. As for the implications on device performance, extended aggregation and crystallization durations (observed at higher donor content) are beneficial for forming favorable phase-separated morphologies with enhanced molecular order38,39. After crystallization, as shown in Fig. 3d, the peak location of BTP-eC9-4F red-shifts progressively with decreasing D18 content, indicating a more tight molecular aggregation23. These results indicate that the reduction of D18 content accelerated the aggregation of BTP-eC9-4F domains. The rapid nucleation and growth of BTP-eC9-4F acceptor influence the ordered aggregation of the blended films.
Time evolution of a peak location and b absorbance of BTP-eC9-4F in blend films from in situ UV-vis absorption. c Time evolution of nucleation and crystal growth extracted from the peak location of BTP-eC9-4F in films. (Symbol * and ** indicate the peak positions of BTP-eC9-4F at the onset and completion of the nucleation and crystal growth stage, respectively.) d Peak variation of BTP-eC9-4F in films with different content. Time evolution of e peak locations and f intensities of BTP-eC9-4F in blend films from in situ PL spectra.
In situ PL spectra were employed to investigate the evolution of phase separation during film formation. As the materials transition from a molecular state in solution to an aggregated state in the solid film, changes in PL intensity provide insight into aggregation-induced emission quenching40. Supplementary Fig. 10 shows the time evolution of 2D PL contour maps of the acceptor peaks of blend films. According to the location variations of the acceptor in Fig. 3e, a more pronounced redshift of BTP-eC9-4F occurs at an earlier stage with decreasing D18 content, suggesting that acceptor aggregation initiates earlier. Similarly, based on the intensity variation of the acceptor in Fig. 3f, as the D18 content decreases, the BTP-eC9-4F intensity declines more rapidly, indicating an accelerated phase separation process between the donor and acceptor.
On the other hand, photoinduced exciton generation and diffusion, as well as carrier dissociation at the donor-acceptor interface, are crucial for the performance of OSCs. Given that all samples contain an equal amount of BTP-eC9-4F, the hole-transfer process, which is dictated by the properties of both the donor and acceptor and the phase-morphology characteristics, can be utilized to analyze how different donor contents affect carrier diffusion and separation. To this end, femtosecond transient absorption (fs-TA) spectroscopy was employed to study the hole-transfer dynamics from BTP-eC9-4F to D18 in devices with varying donor contents (Fig. 4 and Supplementary Fig. 11). For the blend films, a low-power 800 nm pump beam (<2 μJ cm−2) is used to selectively excite the acceptor, BTP-eC9-4F, enabling the identification of hole-transfer pathways. In the blend film with a D:A ratio of 1:1.2, as illustrated in Fig. 4a, b, strong ground-state bleach (GSB) peaks emerge in the long-wavelength region (570–880 nm) after excitation, while a slowly increasing negative signal appears in the short-wavelength region (500–630 nm). These correspond to the generation of acceptor excitons and the gradual hole-transfer process, respectively. A similar phenomenon is observed in blend films with different donor contents (Supplementary Fig. 11). Furthermore, kinetic traces at selected wavelengths (830 nm and 600 nm) are acquired to quantitatively assess the hole-transfer dynamics in the blend film with a D:A ratio of 1:1.2. As shown in Fig. 4c, the attenuation of the GSB signal at 830 nm (corresponding to BTP-eC9-4F) is accompanied by an increase in the GSB signal at 600 nm (corresponding to D18, as shown in Supplementary Fig. 12), indicating efficient hole transfer from BTP-eC9-4F to D18.
a Contour plots of the TA spectra of D18:BTP-eC9-4F (1:1.2) blend film pumped at 800 nm and b TA spectra at selected delay times. c TA traces of D18:BTP-eC9-4F (1:1.2) blend film probed at 830 and 600 nm wavelengths. d TA traces at 600 nm of D18:BTP-eC9-4F blend film with different D18 content. e The time constants and f fitting rate constant correspond to the hole transfer process of D18:BTP-eC9-4F blend films with different D18 content.
The hole-transfer kinetics extracted from the GSB signal at 600 nm in blends with different D18 content are shown in Fig. 4d. By fitting the GSB rise signal of the donor using a double-exponential function41,42, the time constants (τ1 and τ2) and the relevant fitting parameters of the blend films are shown in Fig. 4e, f, respectively. The corresponding data are summarized in Supplementary Table 3, in which τ1 and τ2 are assigned to represent the ultrafast exciton dissociation at the D:A interface and the diffusion of the exciton in the domain, respectively. The values of τ1 and τ2 increase with the decrease of D18 content. In addition, the values of A1 (the interfacial process) and A2 (the corresponding diffusion-mediated process) change significantly. With the decrease of D18 content, the magnitude of A1 decreases, and the magnitude of A2 increases. It is because the D:A interface decreases with the decrease of D18 donor content in the blend film, which makes it difficult for exciton dissociation. At the same time, the decrease of D18 content means the larger acceptor domain, and thus the longer diffusion within the domain. Therefore, the proportion of donors in the blend film is essential for exciton diffusion and dissociation.
A series of OSCs with a typical structure of ITO/PEDOT:PSS/D18:BTP-eC9-4F/PDINN/Ag were fabricated. The ratio of D:A was adjusted from 1:1.2 to 0.2:1.2. The current-density versus voltage (J–V) curves of OSCs are shown in Supplementary Fig. 13a. Meanwhile, the detailed parameters of the devices are summarized in Supplementary Table 4. Clearly, by lowering the donor content from 1:1.2 to 0.2:1.2, the PCEs dropped from 19.01% to 7.86% due to the significantly reduced short-circuit current density ( JSC)43. The external quantum efficiency (EQE) spectra of OSCs are shown in Supplementary Fig. 13b. EQE values were distinctly reduced as the D18 content decreased, and there were larger drops in the D18 absorption range as compared with the drops in the BTP-eC9-4F absorption range. It resulted from the loss of photon harvesting.
To gain more insights into the differences in the photovoltaic performance for the devices with different donor content, we studied their charge dynamic process including the dissociation, transport, and recombination. The photocurrent versus effective voltage ( Jph–Veff) was plotted in Supplementary Fig. 13c, while charge dissociation efficiency (Pdiss) and collection efficiency (Pcoll) derived from the Jph–Veff curve were summarized in Supplementary Table 544,45. The Pdiss/Pcoll values are estimated to be 98.26%/91.58%, 97.67%/90.99%, 97.18%/89.49%, 95.57%/86.24%, and 91.40%/81.79% for the 1:1.2, 0.8:1.2, 0.6:1.2, 0.4:1.2, and 0.2:1.2 devices, respectively. Both Pdiss and Pcoll decreased gradually as the donor content decreased, which agrees with the JSC and fill factor (FF) for the corresponding devices. It may be because the reduced donor content led to a decrease in the D:A heterojunction area and disrupted the donor network, ultimately hindering the dissociation, as well as charge transport and collection, which is consistent with the reported literature46.
The dependence of JSC and open-circuit voltage (VOC) on light intensity (Plight) was tested to examine the charge recombination mechanisms in the OSCs47,48. The slope of n in the function of VOC∝nkBT/qlnPlight (kB is the Boltzmann constant, q is the elementary charge, and T is the temperature) is related to the nonradiative recombination loss of devices. The value of n close to 1 indicates that trap-assisted recombination is negligible. Extracted from Supplementary Fig. 13d, the fitted n values of 1:1.2, 0.8:1.2, 0.6:1.2, 0.4:1.2, and 0.2:1.2 devices are 1.02, 1.04, 1.07, 1.11, and 1.17, respectively, indicating a gradual increase in trap-assisted recombination with decreasing donor content in the OSCs. Moreover, the exponential factor α in the power law relationship of JSC∝(Plight)α can be used to estimate the bimolecular recombination loss. When the value of α approaches 1, it means charge carriers are efficiently extracted without bimolecular recombination. As shown in Supplementary Fig. 13e, the calculated α values for the 1:1.2, 0.8:1.2, 0.6:1.2, 0.4:1.2, and 0.2:1.2 devices are 0.99, 0.99, 0.97, 0.96, and 0.94, respectively, indicating an increasing degree of bimolecular recombination as the donor content decreases in the OSCs. According to the above results, it can be found that the D:A content in the blend film has a significant impact on the dissociation, transport, and recombination of the carriers.
The aforementioned analyses exhibit the relationship of film growth dynamics, crystallization features, recombination losses, and device efficiencies in the OSCs, guiding the optimization of perovskite/organic TSCs. Figure 5a illustrates the structure of tandem devices. The cross-sectional scanning electron microscopy (SEM) image (Fig. 5b) of the TSCs clearly illustrates that the perovskite layer and the organic active layer are distinctly separated by an interconnection layer (ICL) composed of PCBM/C60/SnOX/Au/PEDOT:PSS. In this research, a wide-bandgap perovskite was adopted to construct the front cell, and a PCE of 17.84% was attained in single-junction PSCs. The detailed data are presented in Supplementary Figs. 14 and 15 and Supplementary Table 6. Given the critical role of solar spectrum matching in determining the performance of TSCs, D18:L8-BO system with different bandgap was also employed as rear sub-cells (Supplementary Fig. 16 and Supplementary Table 7). TSCs with rear cells featuring different D:A ratios were fabricated. The corresponding results are displayed in Fig. 5c and Supplementary Table 6. The TSC with a D:A (1:1.2) rear cell exhibited the optimal performance, yielding a PCE of 26.42%, a VOC of 2.242 V, a JSC of 14.85 mA cm-2, and an FF of 79.55%. In contrast, the TSCs with D:A (0.8:1.2) and D:A (0.6:1.2) rear cells exhibited lower PCEs of 24.37% and 22.90%, respectively. This reduction is primarily attributed to a notable decline in the FF, which aligns with the trend observed in the corresponding single-junction OSCs. It is worth highlighting that in single-junction OSCs, a decrease in the donor content results in a substantial reduction in the JSC. However, in TSCs, the decrease in the JSC is relatively less pronounced. The EQE spectra of the TSCs suggest that the photocurrents of the front and rear sub-cells are well-correlated with the measured JSC. For further details, refer to Fig. 5d and Supplementary Fig. 17. Additionally, the stabilized power output curves reveal a stable power output of 25.91%, as shown in Fig. 5e. Finally, the operational stability of the as-fabricated perovskite/organic TSCs was also investigated. As shown in the Supplementary Fig. 18, the PCE evolution of encapsulated devices was monitored via maximum power point (MPP) tracking under continuous one-sun simulated illumination in ambient air (relative humidity of 50–80%) at 25 °C. The results demonstrate that the perovskite/organic TSCs retained 80% of their initial PCE over 250 h of operation, providing a solid foundation for further enhancement of device stability. In addition, the photovoltaic performance comparison with previously reported perovskite/organic TSCs is provided in Supplementary Table 8. The champion PCE of our device reaches 26.42%, which is among the highest reported for solution-processed perovskite/organic TSCs.
a Perovskite/organic TSC structure. b Cross-section SEM image. c J-V curves of perovskite/organic TSCs with different rear subcells. d EQE spectra and e stabilized power output curves of perovskite/organic TSCs.
In conclusion, we have unequivocally elucidated the pivotal role of precise modulation of polymer donor content within organic subcells in achieving high-efficiency perovskite/organic TSCs. Through the establishment of quantitative correlations between the key morphological and electrical parameters, including crystallinity, phase separation, and exciton recombination, it has been demonstrated that an optimized D:A stoichiometry substantially elevates the exciton dissociation yield and charge transport efficiency. Conversely, an inadequate donor loading disrupts the molecular ordering of the active layer, thereby degrading device performance. As a result, following the regulation of the film morphology and mitigation of energy recombination losses in OSC subcells, the PCE has been significantly enhanced. Ultimately, the perovskite/organic TSCs have achieved a remarkable PCE of 26.42%, underscoring the necessity of synergistically optimizing optical absorption and exciton utilization to maximize the performance of tandem devices.
D18, BTP-eC9-4F and L8-BO were synthesized in lab. PEDOT:PSS (Clevios PVP Al4083) was obtained from Heraeus. PDINN was obtained from Derthon. 4PADCB was obtained from TCI. Aluminum oxide nanoparticle (Al2O3 20 wt% in IPA, diameter 15–50 nm) was purchased from Aladdin.
UV-visible absorption spectra were measured on a PerkinElmer UV-vis spectrometer model Lambda 750. TA spectra were recorded using an ultrafast spectrometer (Harpia-TA, Light Conversion). A Yb:KGW laser (1030 nm, 54 kHz, Light Conversion) was split into fundamental light beams: one was converted via an optical parametric amplifier to generate the pump beam, while the other, focused on a 5 mm sapphire, produced a probe beam (575–995 nm). The GIWAXS data were obtained at 1W1A Diffuse X-ray Scattering Station, Beijing Synchrotron Radiation Facility (BSRF-1W1A). The monochromatic of the light source was 1.54 Å. The data were recorded by using the two-dimensional image plate detector of EIGER 1 M from Dectris, Switzerland. The in situ light absorption spectra, fluorescence spectra and light scattering spectra were performed on a multi-spectrometer (DU-200, Shaanxi Puguang Weishi Co. Ltd.). In situ spin-coating GIWAXS measurements were carried out using a Synchrotron-based temperature-controlled spin-coating platform, co-developed by Sichuan University, Beijing Synchrotron Radiation Facility, and Beijing Zhongke Wanyuan Technology Co., Ltd. This system supports remote control of solution dripping, spin speed, and substrate temperature, and is compatible with various atmospheric environments through adjustable gas flow, thereby enabling in situ GIWAXS characterization during the film formation process. All the active layers are monitored by this system during the film-forming process while spin-coating. The in situ GIWAXS data were obtained at beamline BL02U2 and BL6B1 of Shanghai Synchrotron Radiation Facility (SSRF). The monochromatic of the light source was 1.24 Å. The data were recorded by using the two-dimensional image plate detector of Pilatus 2 M from Dectris, Switzerland. The sample-to-detector distance was set to 150 mm for in situ GIWAXS measurement. The transformation to q-space, radial cuts for the in-plane and out-of-plane analysis, and azimuthal cuts for the orientation analysis were processed by GIWAXS-tools.
Femtosecond transient absorption measurements, a Yb:KGW laser (1030 nm, 54 kHz, Light Conversion) is split into two fundamental light beams. One of the light beams is transferred to the optical parametric amplifier (Orpheus, Light Conversion) to generate a high-intensity pulse of a specific wavelength as the pump beam. At the same time, the other is focused on a 5 mm sapphire to generate low-intensity continuum light, employed as the probe beam. The pump and probe beams were spatially set at the magic angle (54.7°) and overlapped at the sample. The time delays between pump and probe beams were achieved using a delay stage monitor, and the transmitted probe light was collected using a charge-coupled device. The pump fluence was kept at <2 μJ/cm2, unless indicated otherwise. For the film samples, the transient absorption experiment was done several times on several spots of the film for each sample, and the average was taken. Upon completion, no sample degradation was detected.
During the measurements, Jph was calculated as the difference between JL and JD, with JL and JD representing current densities in light and darkness, respectively. Veff was defined as V0 minus Vbias, where V0 corresponds to the voltage at which JL equals JD, and Vbias is the applied voltage. The charge separation and collection efficiencies can be individually evaluated by Pdiss and Pcoll, where Pdiss = Jph/Jsat, Pcoll = Jmax/Jsat, Jph is the photocurrent density at which the applied bias voltage is zero, Jmax is the photocurrent density at the maximal output point, respectively45,49.
Organic solar cells with the structure of ITO/PEDOT:PSS/Active layer/PDINN/Ag were fabricated. The PEDOT:PSS solution was deposited onto the pre-cleaned ITO substrates, followed by annealing at 150 °C for 20 min. D18:BTP-eC9-4F (or D18:L8-BO) was dissolved in chloroform, and the concentration of BTP-eC9-4F is 5.4 mg/mL. Then, PDINN was dissolved in methanol (1.5 mg/mL) and spin-coated on the active layer. Finally, 100 nm was thermally evaporated through a shadow mask at a vacuum pressure below 3 × 10−6 Torr. The effective area for each cell was 0.04 cm2.
The structure of the perovskite devices is ITO/4PADCB/Al2O3/FAMACsPb(I0.5Br0.5)3/ PCBM/C60/ALD SnOX/Ag. 4PADCB was dissolved in ethanol with a concentration of 0.4 mg/mL and spin-coated onto the pre-cleaned ITO substrates followed by annealing at 100 °C for 10 min. The diluted Al2O3 dispersion solution was spin-coated on the 4PADCB film at 5000 rpm for 30 s and heated at 100 °C for 10 min. The precursor solution was spin-coated onto Al2O3 film to form the FAMACsPb(I0.5Br0.5)3 perovskite films, and CB was slowly dropped. The films were then immediately annealed at 105 °C for 15 min. After cooling to room temperature, 60 μL of PEABr solution (1.5 mg mL-1 in isopropanol) was swiftly dropped onto the perovskite film and spin-coated at 5000 rpm for 30 s, followed by annealing at 105 °C for 5 min. PCBM was dissolved in CB (10 mg/mL) as electron transport layers and spin-coated. Then the 10 nm of C60 was evaporated in the high-vacuum thermal evaporator. After that, 15–20 nm of SnOX was fabricated by atomic layer deposition. Finally, 90 nm Ag was thermally evaporated through a shadow mark at a vacuum pressure below 3×10-6 Torr. The effective area for each cell was 0.04 cm2.
After deposition of SnOX, 1 nm Au was thermally evaporated, and then PEDOT:PSS was spin-coated as a hole transport layer for the organic subcell. Then, the organic films were spin-coated and PDINN solution was spin-coated on top as an electron transport layer. Finally, 100 nm Ag was thermally evaporated as a top electrode.
The J–V curves of the devices were measured in a glove box with an instrument from Enli Technology Ltd., Taiwan (SS-F53A) under AM 1.5 G illumination (AAA class solar simulator, 100 mW cm−2 calibrated with a standard single crystal Si photovoltaic cell). EQE measurements were performed by a solar cell spectral response measurement system (QER3011, Enli Technology Co. Ltd), and the intensity was calibrated with a standard single-crystal Si photovoltaic cell before the test.
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
All data supporting the results of this study are provided in this article and its supplementary information. Any other information can be requested from the corresponding author. Source data are provided with this paper.
Brinkmann, K. O. et al. Perovskite–organic tandem solar cells. Nat. Rev. Mater. 9, 202–217 (2024).
Article CAS ADS Google Scholar
Xie, G. et al. Recent advances on monolithic perovskite-organic tandem solar cells. Interdiscip. Mater. 3, 113–132 (2024).
CAS Google Scholar
Jiang, X. et al. Isomeric diammonium passivation for perovskite-organic tandem solar cells. Nature 635, 860–866 (2024).
Article PubMed ADS Google Scholar
Zhang, Z. et al. Suppression of phase segregation in wide-bandgap perovskites with thiocyanate ions for perovskite/organic tandems with 25.06% efficiency. Nat. Energy 9, 592–601 (2024).
Article CAS ADS Google Scholar
Chen, W. et al. Monolithic perovskite/organic tandem solar cells with 23.6% efficiency enabled by reduced voltage losses and optimized interconnecting layer. Nat. Energy 7, 229–237 (2022).
Article CAS ADS Google Scholar
Wu, S. et al. Prospects and challenges for perovskite-organic tandem solar cells. Joule 7, 484–502 (2023).
Article CAS Google Scholar
Han, M. et al. Unveiling the potential of two-terminal perovskite/organic tandem solar cells: mechanisms, status, and challenges. Adv. Mater. 36, 2402143 (2024).
Article CAS Google Scholar
Cui, X. et al. Boosting the efficiency of perovskite/organic tandem solar cells via enhanced near-infrared absorption and minimized energy losses. Adv. Mater. 36, 2408646 (2024).
Article CAS Google Scholar
Brinkmann, K. O. et al. Perovskite-organic tandem solar cells with indium oxide interconnect. Nature 604, 280–286 (2022).
Article CAS PubMed ADS Google Scholar
Sun, X. et al. From 20% single-junction organic photovoltaics to 26% perovskite/organic tandem solar cells: self-assembled hole transport molecules matter. Energy Environ. Sci. 18, 2536–2545 (2025).
Article CAS Google Scholar
Xie, Y. et al. Homogeneous grain boundary passivation in wide-bandgap perovskite films enables fabrication of monolithic perovskite/organic tandem solar cells with over 21% efficiency. Adv. Funct. Mater. 32, 2112126 (2022).
Article CAS Google Scholar
Lim, J. et al. All-perovskite tandem solar cells: from fundamentals to technological progress. Energy Environ. Sci. 17, 4390–4425 (2024).
Article CAS PubMed PubMed Central Google Scholar
Menke, S. M. et al. Understanding energy loss in organic solar cells: toward a new efficiency regime. Joule 2, 25–35 (2018).
Article CAS Google Scholar
Liu, Q. et al. Understanding and suppressing non-radiative recombination losses in non-fullerene organic solar cells. Adv. Mater. 35, 2302452 (2023).
Article CAS Google Scholar
Zhang, J. et al. Alkyl-chain branching of non-fullerene acceptors flanking conjugated side groups toward highly efficient organic solar cells. Adv. Energy Mater. 11, 2102596 (2021).
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 ADS Google Scholar
Bi, P. et al. Reduced non-radiative charge recombination enables organic photovoltaic cell approaching 19% efficiency. Joule 5, 2408–2419 (2021).
Article CAS Google Scholar
Ma, R. et al. Organic solar cells: beyond 20%. Sci. China Mater. 68, 1689–1701 (2025).
Article Google Scholar
Zhu, L. et al. Single-junction organic solar cells with over 19% efficiency enabled by a refined double-fibril network morphology. Nat. Mater. 21, 656–663 (2022).
Article CAS PubMed ADS Google Scholar
Zhang, H. et al. Sequentially processed bulk-heterojunction-buried structure for efficient organic solar cells with 500nm thickness. Adv. Mater. 36, 2400521 (2024).
Article CAS Google Scholar
Ma, X. et al. Unraveling the impacts of intermolecular interaction on morphology evolution for highly efficient organic solar cells and modules. Sci. China Chem. 68, 264–272 (2025).
Article CAS Google Scholar
van Franeker, J. J. et al. Polymer solar cells: Solubility controls fiber network formation. J. Am. Chem. Soc. 137, 11783–11794 (2015).
Article PubMed ADS Google Scholar
Wang, X. et al. High-efficiency (16.93%) pseudo-planar heterojunction organic solar cells enabled by binary additives strategy. Adv. Funct. Mater. 31, 2102291 (2021).
Article CAS Google Scholar
Gao, W. et al. Efficient all-small-molecule organic solar cells processed with non-halogen solvent. Nat. Commun. 15, 1946 (2024).
Article CAS PubMed PubMed Central ADS Google Scholar
Li, H. et al. Optimizing the crystallization behavior and film morphology of donor-acceptor conjugated semiconducting polymers by side-chain-solvent interaction in nonpolar solvents. Macromolecules 54, 10557–10573 (2021).
Article CAS ADS Google Scholar
Perez, L. A. et al. Effect of backbone regioregularity on the structure and orientation of a donor-acceptor semiconducting copolymer. Macromolecules 47, 1403–1410 (2014).
Article CAS ADS Google Scholar
Peng, Z. et al. Unveiling the strain-induced microstructural evolution and morphology-stretchability correlations of intrinsically stretchable organic photovoltaic films. Adv. Energy Mater. 14, 2304286 (2024).
Article CAS Google Scholar
Peng, Z. et al. Modulation of morphological, mechanical, and photovoltaic properties of ternary organic photovoltaic blends for optimum operation. Adv. Energy Mater. 11, 2003506 (2021).
Article CAS Google Scholar
Baker, J. L. et al. Quantification of thin film crystallographic orientation using X-ray diffraction with an area detector. Langmuir 26, 9146–9151 (2010).
Article CAS PubMed Google Scholar
Jiang, Z. GIXSGUI: a MATLAB toolbox for grazing-incidence X-ray scattering data visualization and reduction, and indexing of buried three-dimensional periodic nanostructured films. J. Appl. Cryst. 48, 917–926 (2015).
Article CAS ADS Google Scholar
Ma, R. et al. In-situ understanding on the formation of fibrillar morphology in green solvent processed all-polymer solar cells. Natl. Sci. Rev. 11, nwae384 (2024).
Article CAS PubMed PubMed Central Google Scholar
Peng, Z. et al. Real-time probing and unraveling the morphology formation of blade-coated ternary nonfullerene organic photovoltaics with in situ X-ray scattering. Adv. Funct. Mater. 33, 2213248 (2023).
Article CAS Google Scholar
Wang, S. et al. The continuous fiber networks with a balanced bimodal orientation of P(NDI2OD-T2) by controlling solution nucleation and face-on and edge-on crystallization rates. Polymer 262, 125435 (2022).
Article CAS Google Scholar
Zhao, H. et al. High-performance green thick-film ternary organic solar cells enabled by crystallinity regulation. Adv. Funct. Mater. 33, 2210534 (2023).
Article CAS Google Scholar
Xue, J. et al. Nonhalogenated dual-slot-die processing enables high-efficiency organic solar cells. Adv. Mater. 34, 2202659 (2022).
Article CAS Google Scholar
Bi, P. et al. Enhancing photon utilization efficiency for high-performance organic photovoltaic cells via regulating phase transition kinetics. Adv. Mater. 35, 2210865 (2023).
Article CAS Google Scholar
Jiang, Y. et al. 20.6% efficiency organic solar cells enabled by incorporating a lower bandgap guest nonfullerene acceptor without open-circuit voltage loss. Adv. Mater. 37, 2500282 (2025).
Article CAS Google Scholar
Wang, J. et al. Binary organic solar cells with 19.2% efficiency enabled by solid additive. Adv. Mater. 35, 2301583 (2023).
Article CAS Google Scholar
Song, J. et al. Optimizing double-fibril network morphology via solid additive strategy enables binary all-polymer solar cells with 19.50% efficiency. Adv. Mater. 36, 2406922 (2024).
Article CAS Google Scholar
Chen, Z. et al. Simplified fabrication of high-performance organic solar cells through the design of self-assembling hole-transport molecules. Joule 8, 1723–1734 (2024).
Article CAS Google Scholar
Ran, G. et al. High-performance ternary organic solar cells with spectral uniform photocurrent generation by enhanced förster resonance energy transfer induced reverse hole transfer. Adv. Funct. Mater. 35, 2417478 (2025).
Article CAS Google Scholar
Ran, G. et al. Electron, hole, and energy transfer dynamics in non-fullerene small-molecule acceptors. Chem. Sci. 15, 16079–16085 (2024).
Article CAS PubMed PubMed Central Google Scholar
Schopp, N. et al. Unraveling device physics of dilute-donor narrow-bandgap organic solar cells with highly transparent active layers. Adv. Mater. 34, 2203796 (2022).
Article CAS Google Scholar
Ge, Z. et al. Over 18% efficiency of all-polymer solar cells with long-term stability enabled by Y6 as a solid additive. Adv. Mater. 35, 2301906 (2023).
Article CAS Google Scholar
Gao, J. et al. Over 17.7% efficiency ternary-blend organic solar cells with low energy-loss and good thickness-tolerance. Chem. Eng. J. 428, 129276 (2022).
Article CAS Google Scholar
Xue, F. et al. Boosting fill factor of semitransparent donor-poor organic solar cells for the best light utilization efficiency. Adv. Funct. Mater. 35, 2415617 (2025).
Article CAS Google Scholar
Gupta, V. et al. Barium: An efficient cathode layer for bulk-heterojunction solar cells. Sci. Rep. 3, 1965 (2013).
Article PubMed PubMed Central ADS Google Scholar
Xian, K. et al. Refining acceptor aggregation in nonfullerene organic solar cells to achieve high efficiency and superior thermal stability. Sci. China Chem. 66, 202–215 (2023).
Article CAS Google Scholar
Cui, X. et al. Highly efficient solution-processed organic photovoltaics enabled by improving packing behavior of organic semiconductors. Sci. China Chem. 67, 890–897 (2024).
Article CAS Google Scholar
Download references
This work was supported by the National Natural Science Foundation of China (52303240 (D.O.), 52303239 (Y.L.), 51933001 (Z.B.) and 22475114 (Y.L.)), the Natural Science Foundation of Shandong Province (ZR2021QE138 (D.O.), ZR2022QB141 (Y.L.) and 2023HWYQ-087 (Y.L.)), the Postdoctoral Fellowship Program of CPSF (GZB20240067) (G.R.), the Postdoctoral Science Foundation of China (2022M711737) (D.O.), and High Level of Special Funds (G03034K001) (L.Q.). A portion of this work is based on the data obtained at the Beijing Synchrotron Radiation Facility (BSRF) and Shanghai Synchrotron Radiation Facility (SSRF). The authors gratefully acknowledge the cooperation of the beamline scientists at BSRF-1W1A, SRRF-BL16B1, and SRRF-BL02U2.
These authors contributed equally: Xinyue Cui, Guanshui Xie, Guangliu Ran.
College of Textiles and Clothing, State Key Laboratory of Bio-fibers and Eco-textiles, Qingdao University, Qingdao, China
Xinyue Cui, Yuqiang Liu, Qiumin Kong, Dan Ouyang & Zhishan Bo
School of Chemistry and Materials Science, Anhui Normal University, Wuhu, China
Xinyue Cui
Department of Mechanical and Energy Engineering, SUSTech Energy Institute for Carbon Neutrality, Southern University of Science and Technology, Shenzhen, China
Guanshui Xie & Longbin Qiu
School of Physics and Astronomy, Applied Optics Beijing Area Major Laboratory, Center for Advanced Quantum Studies, Beijing Normal University, Beijing, China
Guangliu Ran & Wenkai Zhang
Beijing Key Laboratory of Energy Conversion and Storage Materials, College of Chemistry, Beijing Normal University, Beijing, China
Xueqing Ma, Gendi Zhang & Zhishan Bo
College of Materials and Energy, Guang’an Institute of Technology, Guang’an, China
Hongxiang Li
State Key Laboratory of Polymer Materials Engineering, College of Polymer Science and Engineering, Sichuan University, Chengdu, China
Hongxiang Li & Pei Cheng
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
X.C., G.X. and G.R. contributed equally to this work. Y.L. and X.C. conceived the idea for the study and designed the experiments. X.C. performed the fabrication, testing, and data analysis of OSCs and perovskite/organic TSCs. G.X. performed the fabrication and testing of PSCs and the front subcells for the perovskite/organic TSCs, as well as the stability characterization and analysis of the tandem devices. G.R. and W.Z. carried out the TA characterizations and data analysis. X.M. conducted the fabrication and testing of OSCs. X.M. and G.Z. performed the TEM measurements. H.L. and P.C. provided the GIWAXS measurements. Q.K. contributed to the preparation of UV-vis absorption measurement samples. X.C. and Y.L. wrote the manuscript. Y.L., G.R., H.L., D.O., L.Q., and Z.B. supervised the project. All authors participated in discussions and commented on the manuscript.
Correspondence to Yuqiang Liu, Wenkai Zhang, Hongxiang Li, Dan Ouyang, Longbin Qiu or Zhishan Bo.
The authors declare no competing interests.
Nature Communications thanks Ruijie Ma, Jiangang Liu, 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
Cui, X., Xie, G., Ran, G. et al. Organic film evolution and recombination losses in highly efficient perovskite/organic tandem solar cells. Nat Commun 16, 8986 (2025). https://doi.org/10.1038/s41467-025-64032-7
Download citation
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-025-64032-7
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
Nano-Micro Letters (2026)
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.
