Bonding‐Enabled Interfacial Reconstruction at Buried Interface for High‐Efficiency Sb₂Se₃ Solar Cells

ABSTRACT

Antimony selenide (Sb₂Se₃) has attracted intense attention as one of the most promising photovoltaic materials, owing to its outstanding optoelectronic properties and thermal and chemical stability. However, the solar cells based on Sb₂Se₃ suffer from inferior interfacial contact and severe recombination loss at the buried interface, which limits the efficiency improvement of the device. To overcome these limitations, here we develop an acid‐activated interfacial reconstruction strategy to improve the quality of the buried CdS/Sb₂Se₃ interface. We found that HCl post‐treatment on the CdS film can effectively enhance interfacial uniformity and reduce residual SO₄ ²⁻ species on the (100) facet of CdS. These surface characteristics facilitate the formation of Sb‐S bonding at the CdS/Sb₂Se₃ interface, thereby promoting bonding‐mediated oriented growth of Sb₂Se₃. Consequently, the deposited Sb₂Se₃ exhibits a preferred [hk1] orientation and mitigated Se vacancy defects, leading to enhanced carrier transport efficiency and suppressed non‐radiative recombination loss in the device. Ultimately, we achieved a champion power conversion efficiency (PCE) of 9.74% in Sb₂Se₃ superstrate solar cells fabricated via thermal evaporation (TE). This approach establishes a novel paradigm for interfacial engineering, where the in situ activation of CdS surfaces enhances interfacial properties and facilitates chemical bonding during subsequent deposition.

This article reports an HCl‐activated interfacial reconstruction strategy to enhance the uniformity and chemical bonding within the buried CdS/Sb₂Se₃ interface. The enhanced interface characteristics facilitate the stable and continuous formation of interfacial bonding, resulting in the vertical orientation growth of the Sb₂Se₃ film and improved charge carrier transport, leading to an efficiency breakthrough in Sb₂Se₃ solar cells.

1. Introduction

Antimony selenide (Sb₂Se₃) has emerged as a promising light‐absorbing material for high‐efficiency thin‐film photovoltaics owing to its superior properties, including a suitable band gap, a high absorption coefficient (>10⁵ cm⁻¹), and excellent chemical stability [1]. Notably, Sb₂Se₃ exhibits a quasi‐one‐dimensional (Q1D) structure composed of (Sb₄Se₆)_n ribbons [2, 3], which differs from that of the conventional 3D materials such as silicon, CdTe, perovskite, and so on. This unique structural characteristic endows Sb₂Se₃ with highly efficient carrier transport and benign grain boundaries along the ribbons [4, 5, 6]. To take advantage of this property, the (Sb₄Se₆)_n ribbons should be controlled to align perpendicular to certain facets of the underlying charge‐transporting layer [7, 8]. However, controlling the preferred crystallographic orientation has long been a formidable challenge in thin film fabrication.

Fundamentally, the electron transport layer (ETL) plays a crucial role in the fabrication of the absorber layer, as its physicochemical properties and interfacial compatibility directly govern the bonding formation, nucleation behavior, and crystallographic orientation of the absorber layer during deposition. In Sb₂Se₃ solar cell fabrication, cadmium sulfide (CdS) exhibits favorable energy band alignment with the absorber layer and is widely employed as the electron transport layer (ETL), delivering the achievement of the highest power conversion efficiency compared to other ETLs [9, 10]. Chemical bath deposition (CBD) is the dominant method for the fabrication of CdS film. Nevertheless, CBD‐grown CdS films exhibit high surface roughness and residual ionic impurities, which hinder the formation of intimate interfacial contact during the subsequent deposition of Sb₂Se₃ films [9, 11, 12]. Additionally, although CdS surfaces are intrinsically reactive, coupled with interfacial reactions and elemental interdiffusion at the CdS/Sb₂Se₃ heterojunction, often results in non‐ideal and spatially heterogeneous bonding configurations [13, 14, 15], making the formation of well‐defined interfacial chemical bonds challenging. These interfacial limitations hinder carrier transport and lead to non‐radiative recombination loss [11, 16, 17]. Consequently, the development of suitable interface modification and activation strategies is of paramount importance.

Recently, various approaches have been adopted to improve the surface morphology of CdS, including post‐treatment [18, 19, 20, 21], elemental doping [16, 22, 23, 24, 25, 26], and interface modification [27, 28]. These strategies have yielded a moderate improvement in device performance through exogenous treatment strategies. Nevertheless, most studies focus on surface passivation, compositional tuning, or defect engineering; the impact of intrinsic interfacial bonding evolution on charge carrier transport at the CdS/Sb₂Se₃ heterojunction remains insufficiently explored. To address this gap, we herein investigate a CdS/Sb₂Se₃ interfacial engineering strategy based on FTO (fluorine‐doped tin oxide)/SnO₂ substrates, where FTO functions as the transparent conductive electrode and the conformal SnO₂ layer serves as an efficient interfacial layer. Featuring favorable energy band alignment, SnO₂ facilitates efficient electron extraction and suppresses interfacial recombination, benefiting the subsequent deposition of high‐quality thin films and the complete device performance [9]. Building on this, we develop an acid‐activating interfacial reconstruction strategy capable of removing the residual ionic contamination arising from the chemical deposition of CdS films. Through morphological and structural characterizations, we reveal that the optimized interface exhibits reduced surface roughness, enhanced surface potential uniformity, and activated interfacial chemical bonding. Furthermore, the effective removal of sulfate impurities promotes preferentially oriented growth of Sb₂Se₃ along the [hk1] direction while suppressing defects. These effects collectively suppress interfacial and bulk non‐radiative recombination, enabling efficient carrier extraction and transport. The resulting solar cell delivers a high power conversion efficiency (PCE) of 9.74%, representing a 22% relative enhancement over the control device, among the highest efficiencies reported for this structure. The mechanistic insights provide a fundamental paradigm for interfacial bonding reconstruction during subsequent deposition and for improving carrier transport efficiency in high‐performance Sb₂Se₃ photovoltaics.

2. Results and Discussion

2.1. Surface Treatment of CdS Film

The CdS film is prepared on glass/FTO/SnO₂ substrates by chemical bath deposition (CBD), where cadmium sulfate and thiourea are employed as Cd and S sources in aqueous solution. The film preparation is carried out at 66°C for approximately 16 min. The as‐deposited CdS substrate is treated with a diluted HCl solution (Figure 1a). Subsequently, ethanol rinsing and N₂ drying are performed to mitigate damage to CdS caused by excess acid. Afterward, the SbCl₃ solution is spin‐coated onto the freshly deposited CdS films, followed by annealing at 400°C in ambient air to improve crystallinity. For clarity in the subsequent discussion, the films with and without HCl‐treatment are referred to as Treated‐CdS and Control‐CdS, respectively.

FIGURE 1.

Preparation, micro‐structure, and carrier kinetics analysis of CdS film. (a) Schematic illustration of the CdS surface activation process. Surface morphology of (b–d) the Control‐CdS and (e–g) Treated‐CdS. AFM images of (b) unannealed Control‐CdS, (c) annealed Control‐CdS, (e) unannealed Treated‐CdS, and (f) annealed Treated‐CdS. SEM image of annealed CdS films, (d) without and (g) with treatment, showing a smoother and denser surface morphology. (h) GIXRD diffraction patterns for the Control‐CdS (C1 and C2) and Treated‐CdS samples (T1 and T2). All samples are characterized by GIXRD using a glass/CdS configuration. (i) PL spectra for the Control‐CdS and Treated‐CdS samples. (j) Transient decay kinetics (scatter) and curve fitting (solid lines) monitored at 510 nm wavelength for Control‐CdS and Treated‐CdS films, demonstrating prolonged carrier lifetime after treatment.

Scanning electron microscopy (SEM) and atomic force microscopy (AFM) characterization are conducted to reveal the evolution of morphology (Figure 1b–g). After the HCl treatment, the CdS film exhibits a reduced root mean square (RMS) roughness together with a more uniform grain distribution, as shown in Figures S1 and S2. Statistical analysis based on multiple SEM images further indicates a modest increase in the average grain size and a narrower size distribution after treatment (Figures S3 and S4). These results indicate a tendency toward a more homogeneous surface morphology following the HCl treatment. Additionally, we investigate the crystal structure of the CdS thin films by the grazing incidence X‐ray diffraction (GIXRD) technique. GIXRD was performed with a small incident angle to suppress substrate diffraction and characterize the near‐surface region of the CdS layer, resulting in relatively weak diffraction intensity. As shown in Figure 1h, diffraction signals associated with Sb₂O₃ are detected after SbCl₃ treatment, while no other secondary phases are observed within the detection limit. The phase composition of CdS is further confirmed by Raman spectroscopic characterization (Figure S5) [29]. Relative to the control sample, the treated CdS thin film exhibits a shift of the (100) diffraction peak toward lower angles, corresponding to a modest increase in the interplanar spacing according to Bragg's law. This variation may be associated with subtle defect‐related lattice changes in the near‐surface region of CdS film. To further examine defect characteristics, photoluminescence (PL) spectroscopy was employed. As shown in Figure 1i, the characteristic emission band at ∼690 nm, associated with sulfur vacancies, is significantly suppressed in the Treated‐CdS films, suggesting a reduction in sulfur‐vacancy‐related defects [12, 16]. This result provides complementary support for the changes in crystalline ordering suggested by GIXRD. Reduction of sulfur vacancies suppresses carrier recombination, thus extending the carrier lifetime from 2.28 to 3.21 ns (Figure 1j and Table S1), as measured via transient absorption spectroscopy (TAS in Figure S6). We then employed Kelvin probe force microscopy (KPFM) to characterize surface potential (Figure 2a–c). The treated CdS film exhibits enhanced surface potential uniformity, as evidenced by a decrease in root‐mean‐square potential variation (R _q) from 15.7 to 12.6 mV. Additional KPFM measurements were performed on more independent regions of the samples, and the results consistently show a narrower surface potential distribution for the treated CdS films, indicating improved surface potential homogeneity following the HCl treatment (Figure S7). Notably, the treated CdS sample exhibits an increased average contact potential difference (CPD), indicating a clear shift in the surface electrostatic potential. The CPD is related to the work function via V _CPD = (Φ _tip‐Φ _sample)/e. Combined with the band structure analysis derived from ultraviolet photoelectron spectroscopy (UPS, Figure S8) and UV–vis absorption measurements (Figure S9), these results consistently suggest a modification of the surface electronic structure, accompanied by an upward shift of the Fermi level and altered band‐bending behavior. Specifically, the CPD mapping reveals a more pronounced variation in the surface potential, while UPS‐derived results indicate a comparatively smaller but consistent shift in the near‐surface Fermi level position. Such modifications are expected to facilitate more efficient charge transport across the heterointerface [30]. Overall, the CdS treatment improves film quality and electronic properties, yielding enhanced uniformity, reduced defect density, and more efficient charge carrier transport.

FIGURE 2.

Surface potential and chemical structure analysis. KPFM measurements showing the contact potential difference (CPD) images of (a) Control‐CdS, (b) Treated‐CdS, and (c) corresponding statistical CPD distribution. XPS spectra of (d–f) S 2p, Sb 3d, and Cd 3d for as‐deposited and annealed Control‐CdS and Treated‐CdS. The sulfate‐related peak at 167.74 eV and the Sb‐S bonding feature at 530.18 eV are indicated, with the sulfate signal reduced after treatment. XPS spectra of (g–j) Sb₂Se₃/CdS structure, revealing more pronounced Sb‐S and Cd‐Cl bonding features at the buried interface in Sb 3d and Cd 3d regions.

To elucidate the chemical properties of this interfacial treatment, X‐ray photoelectron spectroscopy (XPS) measurements were conducted (Figure 2d–f). In the S 2p spectrum of the as‐deposited CdS film, a weak peak attributed to sulfate species (SO₄ ²⁻) is observed at 167.74 eV. The reduced intensity of this feature after treatment suggests the partial removal of residual sulfur‐oxygen species from the CdS surface, with the atomic percentage of sulfate species decreasing from 4.40% to 1.96% [31, 32]. A similar trend is also observed after annealing (Figure S10 and Table S2), indicating a consistent reduction of surface residual ionic species under the applied treatment conditions. Meanwhile, the annealed film exhibits enhanced characteristic peak intensities at 406.05 and 530.18 eV, corresponding to Cd‐Cl and Sb‐S bonding, respectively, as depicted in Figure 2e,f [17, 33]. The Cd 3d spectra show a slight shift toward lower binding energy (Table S3), suggesting an increase in electron density around Cd atoms and a modified local bonding environment. It should be noted that the Sb 3d _5/2 region partially overlaps with the O 1s signal. Therefore, the Sb 3d doublet was constrained using the well‐established spin–orbit splitting (∼9.4 eV) and the fixed area ratio (3:2), with the Sb 3d _3/2 component serving as the reference to ensure reliable peak separation. Quantitative analysis reveals that the relative fraction of Sb─O species decreases from ∼70% to ∼54% after treatment (Table S4). This reduction is attributed to the partial removal of residual surface moisture induced by HCl treatment, which suppresses SbCl₃ hydrolysis during subsequent deposition and thereby inhibits the formation of interfacial Sb─O bonds. The presence of oxygen at the interface indicates the involvement of oxygen‐containing species, which may originate from surface oxidation or the underlying oxide substrate. Such oxygen‐related species are commonly observed in heterostructures processed under ambient conditions and can influence the local interfacial chemical environment. To further elucidate the interfacial chemical evolution, stepwise XPS depth profiling was conducted on Sb₂Se₃/CdS heterojunctions using Ar cluster ion sputtering (Figure 2g–j). The Se and S peaks of the treated film (Figures S11 and S12) exhibit no discernible changes compared to the control, with detailed peak information provided in Tables S5–S8. The interface region was identified at a sputtering time of approximately 6 to 8 min. Although multiple chemically similar species, including Sb‐Se, Sb‐S, and oxygen‐related components, coexist and partially overlap in binding energy, their relative evolution still provides insight into interfacial chemical changes. Within the interfacial region, the Sb 3d spectra (Figure 2g,h and Tables S9 and S10) reveal a reduction in the Sb⁰‐related component (3d _5/2 at 529.11 eV) and a concurrent increase in the Sb‐S signal in the treated sample, indicating strengthened interfacial Sb‐S bonding. Furthermore, variations in the Cd‐Cl and Cd‐O species are observed at the interface (Figure 2i,j and Tables S11 and S12), suggesting changes in the local chemical environment of Cd. Collectively, these observations suggest that HCl treatment strengthens interfacial chemical interactions, which may contribute to improved heterojunction quality.

We further employed Secondary Ion Mass Spectrometry (SIMS) to probe the depth‐dependent elemental distribution across the CdS/Sb₂Se₃ heterojunction. As highlighted by the region marked in Figure 3a–c and Figure S13, a distinct distribution pattern of sulfur species is observed within the structure in Figure 3g. Both chromatic gradients in 3D spatial distributions and comparative analysis of the depth profiles against the CdS reference signal indicate a reduction in SO₄ ²⁻ content, accompanied by an increased signal associated with Sb‐S species. These observations are consistent with the XPS results, confirming the modification of interfacial bonding configurations induced by the treatment. In addition, 3D SIMS depth profiling of Se indicates a slight suppression of interfacial Se diffusion after treatment (Figure S14). This behavior is likely associated with the modified interfacial chemical environment, particularly the variation in oxygen‐related species revealed by XPS, which may influence local atomic exchange dynamics near the heterojunction. Such interfacial modifications are expected to alleviate parasitic absorption associated with the CdS layer and benefit the spectral utilization of the Sb₂Se₃ absorber.

FIGURE 3.

Elemental distribution, DFT calculation, and bonding‐activated mechanism. (a–c) SIMS depth profiling and 3D tomography results for the Control‐CdS and Treated‐CdS films. (d) Absorption energy between the SO₄ ²⁻ ion and CdS, with stronger binding on the (100) plane. (e,f) Electron cloud distribution (the yellow and blue areas correspond to electron accumulation and loss phenomenon, respectively) between SO₄ ²⁻ ion and two typical crystal planes of CdS is based on DFT calculations. (e) SO₄ ²⁻ ion and (100) planes of CdS, (f) SO₄ ²⁻ ion and (101) planes of CdS. (g–i) Schematic illustration of the bonding‐activated interfacial formation mechanism induced by treatment.

2.2. Sb₂Se₃ Deposition Induced by Activated Interfacial Bond

To elucidate the adsorption mechanism of SO₄ ²⁻ toward specific crystallographic orientations of CdS, we conducted density functional theory (DFT) calculations (see Note S1). The adsorption energies between the SO₄ ²⁻ ion and the (100) and (101) planes of CdS were calculated (Figure 3d). This result reveals that there is a greater adsorption energy between SO₄ ²⁻ and the (100) plane of CdS (2.8722 eV) compared to that with the (101) plane (1.1775 eV), indicating the preferential adsorption of the SO₄ ²⁻ ion on the (100) plane of CdS. Based on the charge‐difference simulation between SO₄ ²⁻ and CdS, the charge density distribution is analyzed in Figure 3e,f. The electron cloud is predominantly localized around the O and Cd atoms, as visualized by the 2D contour plots in Figure S15, suggesting the formation of O‐Cd interactions that align well with our XPS results. Previous studies have shown that exposed CdS (100) crystallographic facets favor the nucleation and interface‐guided oriented growth of Sb₂Se₃ in the [hk1] orientation [14]. As depicted in Figure 3h,i, the adsorption of residual SO₄ ²⁻ ions on CdS (100) crystal planes blocks the interfacial bonding, thereby weakening Sb‐S interfacial interactions and resulting in the formation of an undesirable (hk0) Sb₂Se₃ orientation. In contrast, the effective removal of SO₄ ²⁻ ions exposes surface‐active sites in the CdS film, facilitating the formation of robust Sb‐S interfacial bonds and inducing the growth of Sb₂Se₃ with a preferred (hk1) orientation.

To further investigate the film growth characteristics, we also conducted high‐resolution transmission electron microscopy (HRTEM) to analyze the interface and adjacent bulk phases of the film. Lattice fringe analysis is conducted at several randomly selected locations in Figure 4a,b. For the control sample, the measured interplanar spacings of CdS and Sb₂Se₃ are 0.301 and 0.325 nm, corresponding to the CdS (101) and Sb₂Se₃ (230) crystal planes. For the treated sample, the measured spacings are 0.359 and 0.249 nm, corresponding to the CdS (100) and Sb₂Se₃ (321) crystal planes, respectively. In the control sample, locally disordered lattice features can be observed in the CdS region (highlighted by the light blue circle in Figure 4a), whereas a relatively clearer lattice arrangement near the interface is observed after treatment, suggesting improved local structural ordering. Considering that lattice fringe visibility in HRTEM images can depend on the relative alignment between the crystal lattice and the electron beam, the TEM observations mainly provide local structural information. Therefore, statistical analysis of lattice fringe orientations was conducted on multiple images collected from different regions of both the control and treated CdS/Sb₂Se₃ interfaces. The results indicate that Sb₂Se₃ deposited on the treated CdS exhibits a relatively stronger tendency toward (hk1) orientation, as summarized in Figure 4c.

FIGURE 4.

HRTEM characterization, morphology, and structures of the induced absorber layer. (a–c) HRTEM lattice fringes analysis of the Sb₂Se₃ deposited on control and treated CdS, and interface Sb₂Se₃ lattice fringes statistics determined by random 15 positions, respectively. Surface morphology of (d) Sb₂Se₃/Control‐CdS and (e) Sb₂Se₃/Treated‐CdS. (f) XRD diffraction patterns for Sb₂Se₃ deposited on the Control‐CdS and Treated‐CdS films, (g) FWHM, and (h) TC of the typical diffraction peaks, suggesting enhanced crystallinity and preferred orientation after treatment. All the samples used for XRD characterization have a glass/FTO/SnO₂/CdS/Sb₂Se₃ configuration.

The surface morphology of the absorber layer was further examined by SEM. As shown in Figure 4d,e, the ∼100 nm thick Sb₂Se₃ film grown on the treated CdS substrate shows subtle morphological differences relative to the control sample, manifesting as a slightly more uniform structure with fewer clusters. In addition, a reduction in interfacial pinholes was observed in the cross‐sectional SEM image (Figure S16). To gain further insight into the microstructural evolution of the absorber layer, the crystallographic properties of the Sb₂Se₃ films were analyzed by evaluating the texture coefficient (TC) and full width at half maximum (FWHM) of the corresponding diffraction peaks (Figure 4f–h and Note S2). The film grown on the treated substrate exhibits a relatively enhanced [hk1] preferred orientation along with a marginally reduced FWHM, indicative of improved crystallinity. Together with the interfacial chemical modifications discussed above, these observations demonstrate that HCl‐mediated interfacial engineering promotes more favorable nucleation and growth behavior of Sb₂Se₃ on the CdS substrate, consistent with the strengthened interfacial bonding revealed by XPS and SIMS analyses.

2.3. Defect Analysis of Sb₂Se₃ Film and Device Performance

Furthermore, we employed optical deep‐level transient spectroscopy (O‐DLTS) to analyze defect states of the as‐prepared Sb₂Se₃ by constructing complete photovoltaic devices (Figure 5a–f). In the O‐DLTS spectra, positive peaks correspond to majority carrier traps, while negative peaks indicate minority carrier traps. Ultraviolet photoelectron spectroscopy (UPS) analysis (Figure S17) confirms the n‐type semiconducting behavior of Sb₂Se₃ films, with the Fermi level located closer to the conduction band edge. Consequently, the positive and negative spectral peaks represent electron traps and hole traps, respectively. For the O‐DLTS signal, the b₁ coefficient associated with the measured transient capacitance is defined in Note S3, and the Arrhenius plot is fitted according to it in Figure S18. Using the equipment's simulation software, key defect parameters, including trap energy level (E _T), capture cross‐section (σ), and trap density (N _T), are calculated directly (Table 1). In the control device, one electron trap (E1) and two hole traps (H1 and H2) are detected with trap energy levels of 0.621, 0.546, and 0.779 eV, respectively (Figure 5a–c). However, only two hole traps (H1 and H2) are identified in the treated device, with trap energy levels of 0.517 and 0.764 eV (Figure 5d–f). Based on the DFT calculations, the E1, H1, and H2 traps are attributed to Se vacancy defect (V_Se), Sb_Se1, and Sb_Se2 antisite defects [6, 34, 35, 36]. The effective suppression of V_Se defect can be directly attributed to the optimized crystalline structure of the Sb₂Se₃ absorber layer. Notably, the trap density associated with Sb antisite defects (H1 and H2) is reduced by approximately one order of magnitude upon CdS treatment. These deep‐level antisite defects typically act as potent non‐radiative recombination centers that strongly trap charge carriers. Since the charge heating rate v_th is constant at room temperature, the reduced concentration of hole trap defects leads to a decreased N _T σ product (as depicted in Figure S19) and an extended carrier lifetime based on Equation:

$$\tau =\frac{1}{v_{th}{N}_{\mathrm{T}}\sigma }$$ (1)

FIGURE 5.

Deep‐level defects and device structure analysis. (a) O‐DLTS spectrum of the control device, and (b) an enlarged view of the selective O‐DLTS signals. (d) O‐DLTS spectrum of the treated device, and (e) an enlarged view of the selective O‐DLTS signals. (c, f) Schematic diagram of the energy band and defect level of the control and treated samples, illustrating the suppression of deep‐level defects after treatment. E _C, E _V, and E _F stand for CBM, VBM, and Fermi level, respectively. Heterojunction diagram for (g) Sb₂Se₃/Control‐CdS and (h) Sb₂Se₃/Treated‐CdS. (i) Complete device configuration.

TABLE 1.

Summarized defect parameters obtained from the O‐DLTS signals of the Sb₂Se₃/Control‐CdS‐based and Sb₂Se₃/Treated‐CdS‐based devices.

Samples	defects	E T [eV]	σ [cm2]	N T [cm−3]	Correl
Control‐device	H1 (SbSe1)	E V+0.546	1.63 × 10−17	3.64 × 1014	0.99901
	H2 (SbSe2)	E V+0.779	5.67 × 10−15	4.60 × 1014	0.98810
	E1 (VSe1)	E C−0.621	1.74 × 10−16	4.47 × 1014	0.99898
Treated‐device	H1 (SbSe1)	E V+0.517	7.09 × 10−17	3.14 × 1013	0.99847
	H2 (SbSe2)	E V+0.764	8.34 × 10−15	4.05 × 1013	0.99147

The effective passivation of such deep‐level defects in Sb₂Se₃ thus suppresses Shockley‐Read‐Hall (SRH) recombination and facilitates more efficient charge carrier transport within the device [37, 38].

In addition, to analyze the charge carrier transport behavior of the CdS/Sb₂Se₃ heterojunction, the energy band alignment was constructed based on the UPS‐derived electronic structure parameters (Figures S6 and S20) as illustrated in Figure 5g,h. The reconstructed band diagrams indicate an increased Fermi‐level offset and enhanced band bending at the interface, which are conducive to the separation and extraction of charge carriers across the heterojunction. Such optimized band alignment is expected to mitigate interfacial recombination loss and thereby improve device performance.

The complete device architecture features a glass/FTO/SnO₂/CdS/Sb₂Se₃/Spiro‐OMeTAD/Au configuration (Figure 5i and Figure S21), in which Spiro‐OMeTAD (2,2′,7,7′‐tetrakis[N,N‐di(4‐methoxyphenyl) amino]‐9,9′‐spirobifluorene) functions as the hole‐transport layer. The SnO₂/CdS, Sb₂Se₃, Spiro‐OMeTAD, and Au layers have thicknesses of approximately 70, 350, 70, and 65 nm, respectively. Photovoltaic performance of Sb₂Se₃‐based solar cells with and without CdS interfacial treatment was evaluated under AM 1.5G illumination (100 mW cm⁻²).

Figure 6a shows the statistics of open‐circuit voltage (V _OC), short‐circuit current density (J _SC), fill factor (FF), and power conversion efficiency (PCE) of the devices based on Sb₂Se₃/Control‐CdS and Sb₂Se₃/Treated‐CdS. In the case of the Treated‐CdS electron transport layer, the device delivers an average PCE of 9.48% ± 0.19% (Table 2). This value is higher than the control device, which shows an average PCE of 7.83% ± 0.13%. The champion device based on Treated‐CdS achieves a PCE of 9.74%, representing a 22% improvement compared with the best control device (7.99%). Furthermore, the stability of the champion Treated‐Sb₂Se₃/CdS device was evaluated under ambient conditions (25°C, humidity >15%) without encapsulation. After 8 months of storage, the device retains 94% of its initial PCE, demonstrating excellent long‐term stability (Figure S22 and Table S13). The corresponding photocurrent density–voltage (J‐V) characteristics of the Sb₂Se₃/Control‐CdS‐ and Sb₂Se₃/Treated‐CdS‐based devices are presented in Figure 6b. The optimal control device delivers a V _OC of 0.429 V, J _SC of 29.61 mA cm⁻², FF of 62.91%, and PCE of 7.99%. In comparison, the treated device displays a higher V _OC of 0.474 V, J _SC of 30.57 mA cm⁻², FF of 67.20%, and PCE of 9.74%. The improved efficiency arises from synergistic improvements in both the buried interface and Sb₂Se₃ absorber, owing to (i) superior CdS layer quality and activated heterojunction interface characteristics, coupled with (ii) enhanced vertical orientation of Sb₂Se₃ 1D structures and point defect passivation in the absorber. These factors are anticipated to decrease SRH recombination losses and enhance carrier extraction and transport efficiency.

FIGURE 6.

Photovoltaic performance and electrochemical characterization of devices. (a) Boxplots and corresponding normal distribution curves of V _OC, J _SC, FF, and PCE for Sb₂Se₃/Control‐CdS and Sb₂Se₃/Treated‐CdS devices, based on 13 independently fabricated devices. (b) J– V curves of the devices based on Control‐CdS and Treated‐CdS. (c) EQE spectra and the corresponding integrated current densities of the devices. (d) Dependence of V _OC on the light intensity. (e) Dark J–V curves of the devices. (f) Nyquist plots were measured at 0.25 V in the dark. (g,h) TPV and TPC measurements, indicating suppressed recombination and improved carrier extraction dynamics.

TABLE 2.

Device parameters (average and optimal) of solar cells based on Sb₂Se₃/Control‐CdS‐and Sb₂Se₃/Treated‐CdS‐based devices.

Devices	V OC [V]	J SC [mA cm−2]	FF [%]	PCE [%]
Control	0.429 (0.439 ± 0.008)	29.61 (29.49 ± 1.00)	62.91 (60.60 ± 1.32)	7.99 (7.83 ± 0.13)
Treated	0.474 (0.471 ± 0.007)	30.57 (30.24 ± 0.62)	67.20 (66.55 ± 0.94)	9.74 (9.48 ± 0.19)

To further elucidate the origin of the enhanced J _SC and V _OC, detailed optoelectronic characterizations were performed. The external quantum efficiency (EQE) spectra were measured to evaluate the wavelength‐dependent photoresponse and verify the origin of the photocurrent. Both devices show photocurrent generation ranging from 330 to 1100 nm (Figure 6c). The HCl treatment does not significantly influence the photogeneration process but primarily improves the interfacial quality and carrier collection near the CdS/Sb₂Se₃ interface. The Treated‐CdS device shows slightly enhanced photoresponse in the short and visible wavelength region, resulting in a marginally higher integrated J _SC (29.52 mA cm⁻²) compared with that of the Control‐CdS device (28.69 mA cm⁻²).

To investigate the origin of the V _OC enhancement, dark‐state capacitance–voltage (C–V) measurements were performed (Figure S23). The built‐in potential (V _bi) extracted from the Mott‐Schottky analysis increases from 0.761 V for the Control‐CdS device to 0.874 V for the Treated‐CdS device, suggesting enhanced band bending at the heterojunction and a stronger built‐in electric field that facilitates carrier separation and extraction. Furthermore, the dark J–V characteristics (Figure 6e) reveal a significantly reduced reverse saturation current (J₀ ) for the Treated‐CdS device (5.61 × 10⁻⁵ mA cm⁻²) compared with that of the Control‐CdS device (4.35 × 10⁻⁴ mA cm⁻²). This order‐of‐magnitude suppression of leakage current suggests reduced recombination losses at the heterojunction, which improves carrier separation and extraction and contributes to the increase in V _OC [39, 40].

Complementing the above analysis, light‐intensity‐dependent V _OC measurements probe charge transport and recombination dynamics. As shown in Figure 6d, V _OC follows the relationship:

$$V_{OC}=\frac{\epsilon kTln\left(I\right)}{e}+constant$$ (2)

where ε is the diode ideal factor, k is the Boltzmann constant, T is the absolute temperature, and e is the elementary charge [33, 41, 42]. The Treated‐CdS device yields a lower ideality factor (ε = 1.24) vs. the control (ε = 1.39). This decrease in ε suggests reduced trap‐assisted nonradiative recombination and fewer defect‐assisted carrier loss pathways. Additionally, electrochemical impedance spectroscopy (EIS) shows that the Sb₂Se₃/Treated‐CdS‐based device exhibits a considerably larger recombination resistance R _rec (4329 ohm) and a smaller series resistance R _s (14.15 ohm) compared to the control device (R _rec = 3149 ohm, R _s = 15.59 ohm) in Figure 6f. The enhanced R _rec and reduced R _s indicate improved charge extraction and transport efficiency in the treated device, which are consistent with the light‐intensity‐dependent V _OC measurements. Transient photovoltage/photocurrent (TPV/TPC) measurements were further conducted to investigate the impact of interfacial treatment on the carrier recombination lifetime and charge extraction dynamics in the complete devices. As shown in Figure 6g, the treated device exhibits slower decay profiles compared to the control device. The photovoltage decay lifetime (τ_TPV) increases from 2716.9 µs in the control device to 3699.7 µs in the treated device (Table S14), confirming substantial suppression of non‐radiative recombination. In addition, TPC measurements exhibit that the photocurrent decay lifetime (τ_TPC) is reduced from 36.783 to 31.576 µs after treatment (Figure 6h and Table S15) [43], indicating accelerated charge transport kinetics within the device. These results are fully consistent with the EIS analysis, contributing to its superior optoelectronic performance.

3. Conclusion

This study demonstrates a novel interfacial activation strategy to regulate the surface properties of CdS films, thereby improving heterojunction contact and enhancing photovoltaic device performance. The HCl treatment effectively reduces surface roughness and residual sulfate species, while also improving the uniformity of surface potential distribution. Such interfacial reconstruction promotes nucleation‐guided growth of Sb₂Se₃ with a preferred crystallographic orientation, leading to improved film crystallinity and reduced defect density. Consequently, an optimized heterojunction with enhanced charge carrier transport and suppressed recombination is achieved, enabling high‐performance Sb₂Se₃ solar cells. This study establishes a methodology for fabricating high‐quality heterojunctions and optimizing carrier transport properties, driving the development of high‐performance optoelectronic devices. The treatment strategy, together with new fundamental insights gained in this work, provides a basis for further advancement of Sb₂Se₃ and related metal chalcogenide‐based photovoltaic materials and devices.

4. Experimental Section

CdS electron transport layers (ETLs) were deposited on glass/SnO₂ substrates via chemical bath deposition (CBD), followed by surface treatment with a dilute HCl/ethanol solution. Sb₂Se₃ absorber layers were subsequently deposited by thermal evaporation and annealed to form the heterojunction. Solar cell devices with the structure FTO/SnO₂/CdS/Sb₂Se₃/Spiro‐OMeTAD/Au were fabricated, where the Spiro‐OMeTAD layer was prepared by spin‐coating, and the back electrode was deposited by thermal evaporation.

Structural and morphological characterizations were performed using X‐ray diffraction (XRD), scanning electron microscopy (SEM), atomic force microscopy (AFM), and Raman spectroscopy. Chemical states and energy levels were probed using X‐ray photoelectron spectroscopy (XPS) and ultraviolet photoelectron spectroscopy (UPS). Photovoltaic performance, including photocurrent density–voltage (J–V), external quantum efficiency (EQE), and light‐intensity‐dependent measurements, was evaluated, and carrier dynamics were further investigated using electrochemical impedance spectroscopy (EIS) and transient photocurrent/photovoltage (TPC/TPV) techniques. Detailed experimental procedures and characterization conditions are provided in the Supporting Information.

Conflicts of Interest

The authors declare no conflict of interest.

Supporting information

Supporting File: smll73907‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.