Van der Waals Template‐Assisted Growth of Two‐dimensional Sb₂S₃

Abstract

Antimony sulfide (Sb₂S₃), a representative of the pnictogen chalcogenide family, possesses a tunable bandgap, strong optical absorption, and phase‐change functionality, making it a promising candidate for next‐generation optoelectronic and memory devices. However, its intrinsic quasi‐one‐dimensional (1D) crystal structure favors nanowire or nanorod growth, hindering synthesis in two‐dimensional (2D) form and limiting integration into ultrathin planar device architectures. Here, van der Waals (vdW) template‐assisted growth of atomically thin 2D Sb₂S₃ nanosheets on monolayer molybdenum disulfide (MoS₂) single crystals is demonstrated, using a low‐temperature chemical vapor deposition process. The density functional theory calculations reveal that MoS₂ lowers diffusion barriers and weakens precursor molecules adsorption, promoting lateral diffusion and 2D growth while suppressing thermodynamically favored 1D morphologies. The resulting 2D Sb₂S₃ exhibits sub‐8 nm thickness and with lateral dimensions dictated by the underlying MoS₂ single‐crystal template. Remarkably, devices fabricated on the resulted Sb₂S₃ integrated MoS₂ heterostructure demonstrate broadband photodetection from ultraviolet to near‐infrared, with photoresponsivity enhanced by two orders of magnitude and improved field‐effect mobility compared to bare monolayer MoS₂. These results establish a scalable route to access 2D forms of quasi‐1D chalcogenides, bridging the critical gap between theoretical predictions and practical applications while enabling their integration into ultrathin, interface‐engineered optoelectronic and memory devices.

Van der Waals (vdW) template‐assisted growth of atomically thin two‐dimensional (2D) Sb₂S₃ on monolayer MoS₂ single crystals is demonstrated, overcoming its thermodynamically preferred one‐dimensional (1D) growth. The resulting sub‐8 nm nanosheets show uniform coverage, atomically sharp interfaces, and enhanced broadband optoelectronic performance as Sb₂S₃/MoS₂ heterostructures, offering a scalable route to 2D pnictogen chalcogenides.

1. Introduction

Atomically thin two‐dimensional (2D) materials have gained immense attention due to their unique physical and chemical properties, which emerge from their reduced dimensionality and strong quantum confinement effects.^{[ 1} , ² , ³ , ⁴ , ^{5 ]} These materials exhibit layer‐dependent electronic band structures, high carrier mobility, and strong light‐matter interactions, making them highly promising for applications in nanoelectronics, optoelectronics, energy storage, quantum, and photonic devices.^{[ 6} , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ^{12 ]} While transition metal dichalcogenides (TMDs) are widely studied for their tunable electronic and optical properties,^{[ 13} , ¹⁴ , ¹⁵ , ^{16 ]} pnictogen chalcogenides (M₂X₃, where M = group V metal and X = group VI chalcogen), are an emerging yet relatively underexplored class of 2D materials. Owing to their intrinsically anisotropic crystal structures and broad tunability in optoelectronic behavior, pnictogen chalcogenides present unique opportunities for functional diversification and scalable integration in next‐generation electronic and photonic devices.^{[ 17} , ¹⁸ , ¹⁹ , ²⁰ , ^{21 ]} Among them, antimony sulfide (Sb₂S₃) has gained increasing interest due to its tunable bandgap (≈1.7–2.8 eV), high absorption coefficient (1.8 × 10⁵ cm⁻¹ at 450 nm), stability in ambient conditions, and its potential for use in photodetectors, photovoltaics, and phase‐change memory applications.^{[ 22} , ²³ , ²⁴ , ²⁵ , ^{26 ]} However, despite these incredible properties, its intrinsic quasi‐one‐dimensional (1D)m crystal structure makes it extremely challenging to grow as a layered 2D material. Sb₂S₃ consists of ribbon‐like chains of [Sb₄S₆]_n aligned along the [010] crystallographic direction, which are covalently bonded along one axis but weakly held together by van der Waals (vdW) interactions, leading to its preferential growth along the z‐axis into 1D nanostructures such as nanowires, ribbons, or thin films from three‐dimensional (3D) particulates, instead of forming continuous 2D nanosheets in lateral configuration.^{[ 27} , ²⁸ , ²⁹ , ³⁰ , ^{31 ]} Beyond overcoming this synthesis challenge, realizing Sb₂S₃ in a 2D form offers superior compatibility with planar device architectures, wafer‐scale fabrication, and seamless integration with other 2D materials for complementary metal–oxide–semiconductor (CMOS) technologies. Confining the quasi‐1D chains within an ultrathin geometry enhances in‐plane anisotropic responses, tunes interchain coupling, and enables atomically sharp vdW heterostructures with tailored charge‐transfer properties critical for next‐generation atomically thin, interface‐driven devices.

Although liquid‐phase exfoliation, a top‐down approach, has been explored for obtaining Sb₂S₃ 2D nanosheets or nanoplatelets, it suffers from limited thickness control, solvent complexity, and restricted scalability, particularly in terms of lateral flake size.^{[ 32} , ³³ , ³⁴ , ^{35 ]} Similarly, bottom‐up techniques such as chemical vapor deposition (CVD) have been employed for Sb₂S₃ synthesis, however, these methods predominantly favor 1D growth, yielding nanowires and nanotubes, rather than lateral 2D sheets.^{[ 30} , ^{36 ]} As a result, achieving controlled 2D Sb₂S₃ growth remains an ongoing challenge. An alternative approach to address these challenges is epitaxial growth, which has been widely used for fabricating high‐quality 2D semiconductor thin films, allowing precise thickness control, crystallinity, and lattice registry with the single‐crystalline substrate.^{[ 37} , ³⁸ , ^{39 ]} Conventionally, epitaxial growth, including molecular beam epitaxy (MBE) and CVD‐based epitaxy, relies on lattice matching between the growing material and the substrate.^{[ 40} , ^{41 ]} However, lattice mismatch and strain accumulation often introduce defects, dislocations, and grain boundaries, which severely limit the quality and scalability of epitaxially grown materials.^{[ 42 ]} In contrast, template‐assisted growth on vdW surfaces has emerged as a transformative strategy for integrating diverse 2D materials, overcoming the fundamental constraints of lattice mismatch.^{[ 43 ]} Here, weak vdW interactions rather than strong covalent bonding govern interfacial energetics, enabling strain‐free growth even with lattice‐mismatched systems. When deposited on atomically flat 2D templates such as graphene and TMDs, the reduced diffusion barrier promotes lateral adatom migration and suppresses vertical aggregation, yielding high‐crystallinity 2D films without the constraints of strict lattice matching.^{[ 44} , ^{45 ]} This vdW mediated, templated‐assisted approach has enabled the synthesis of high‐quality 2D heterostructures, quantum dots, and layered chalcogenides, demonstrating its versatility for integrating diverse material systems.^{[ 44} , ⁴⁵ , ^{46 ]}

In this work, we demonstrate the first synthesis of atomically thin 2D Sb₂S₃ nanosheets enabled by vdW template‐assisted growth on monolayer MoS₂ single crystals at a low growth temperature of 200 °C. Unlike conventional oxide substrates such as SiO₂, where strong surface interactions impede lateral diffusion and favor the formation of 3D particulates, MoS₂ template provides an ideal vdW surface that lowers the formation energy barrier and promotes layer‐by‐layer lateral growth. Our theoretical calculations validate this growth mechanism, demonstrating that MoS₂ surface enhances adatom mobility via reduced diffusion barrier and thus stabilizes the 2D morphology. The resulting Sb₂S₃ nanosheets exhibit uniform coverage, atomically sharp interfaces, and sub‐8 nm thickness, with lateral dimensions dictated by the MoS₂ single‐crystal domain size. Our approach bypasses the thermodynamically favored 1D growth by using low‐temperature, kinetically controlled conditions, where the MoS₂ vdW template reduces the diffusion barrier and directs lateral 2D growth. Importantly, the planar Sb₂S₃/MoS₂ heterostructures enable enhanced device performance, while achieving an order‐of‐magnitude higher photoresponsivity, broadband detection from ultraviolet to near‐infrared (UV–NIR), and improved field‐effect transistor (FET) characteristics compared to bare MoS₂. These enhancements are attributed to strong optical absorption and efficient interfacial charge transfer through the high‐quality vdW interface. Beyond Sb₂S_3, this vdW template‐assisted growth strategy holds strong potential for other pnictogen chalcogenides and related 2D material systems, enabling a generalizable pathway to integrate them into atomically thin electronics, optoelectronics, energy storage, and phase‐change memory devices.

2. Results and Discussion

To achieve the template‐assisted growth of ultrathin 2D Sb₂S₃ nanosheets on a monolayer MoS₂ template, we employed a two‐step CVD process. In the first step, monolayer MoS₂ single crystals were synthesized using a two‐zone CVD system, as reported in our previous study.^{[ 47 ]} Briefly, sulfur (S) powder was heated to 180 °C in zone‐1, generating vapors that was carried downstream by argon (Ar) as carrier gas to zone‐2, where 300 nm SiO₂/Si substrates pre‐coated with Mo‐precursor were placed at 750 °C under low‐pressure conditions, as illustrated in Figure  1a. In Figure 1b, the cross‐sectional atomic model of a typical as‐grown MoS₂ single‐crystal is depicted. Subsequently, using this single crystal MoS₂ as vdW seed layer, Sb₂S₃ deposition was carried out in the 2nd step of CVD growth, as schematically presented in Figure 1c. In a typical CVD process, Sb₂S₃ precursor powder was sublimated at 550 °C in zone‐1, and the resulting vapors were transported by Ar as a carrier gas into zone‐2, where 2D Sb₂S₃ nanosheets were deposited on a single‐crystal MoS₂ template at a much lower temperature of 200 °C. Figure 1d shows the atomic model of the resulting 2D Sb₂S₃ on MoS₂ vdW template, representing the intended vdW‐stacked configuration formed through this sequential growth process. To evaluate the suitability of monolayer MoS₂ as a vdW template for the lateral growth of 2D Sb₂S₃, we conducted preliminary growth experiments by depositing Sb₂S₃ on both MoS₂ single‐crystals and bare SiO₂/Si substrates. However, the optical images (Figure S1, Supporting Information) revealed that Sb₂S₃ adhered preferentially and more uniformly to the MoS₂ surface compared to bare SiO₂/Si. This observation suggests that MoS₂ offers a more favorable surface for Sb₂S₃ nucleation and growth, attributed to the vdW interaction between the two materials. Notably, in control experiments at a higher growth temperature of 350 °C, the Sb₂S₃ morphology transitioned from lateral 2D sheets to 1D nanorod‐like domains (Figure S2, Supporting Information). This transition reflects the intrinsic thermodynamic preference for anisotropic growth along the [010] direction at higher temperatures, consistent with the quasi‐1D crystal structure of Sb₂S₃ and prior observations of 1D Sb₂S₃ growth on tungsten disulfide (WS₂) at 650 °C.^{[ 48 ]} In contrast, the lower growth temperature (≈200 °C) used in this work establishes non‐thermodynamically conditions that facilitate lateral adatom diffusion on MoS₂ vdW surfaces, stabilizing the 2D morphology. While higher temperatures can improve crystallinity, they hinder lateral growth, which is essential for forming layered vdW heterostructures with MoS₂. Based on these insights, we optimized the CVD parameters to achieve the controlled growth of 2D Sb₂S₃ on MoS₂ single‐crystal template. Figure 1e shows the optical image of a representative as‐grown single‐crystal MoS₂ obtained from the 1st CVD growth, which served as the vdW template for subsequent Sb₂S₃ deposition. Following the 2nd CVD step, the optical image of the resulting Sb₂S₃ grown on MoS₂ template is shown in Figure 1f. The clear optical contrast between the monolayer MoS₂ and the Sb₂S₃/MoS₂ stack on 300 nm SiO₂/Si substrate highlights the difference in flake thickness, confirming the successful synthesis of Sb₂S₃ on MoS₂ single‐crystal template.

Figure 1.

Synthesis and characterization of MoS₂ and 2D Sb₂S₃ on MoS₂ template. a) Schematic of CVD growth setup of monolayer MoS₂. b) Atomic model of monolayer MoS₂ atomic structure c) A schematic of CVD growth setup to grow Sb₂S₃ on a pre‐grown monolayer MoS₂ template. d) Atomic model of Sb₂S₃/MoS₂ stack. e,f) Optical images of a typical as‐grown single crystal MoS₂ template and the resulting Sb₂S₃/MoS₂ heterostructure samples, respectively. g,h) AFM images of MoS₂ and Sb₂S₃/MoS₂ heterostructure flakes, respectively. Inset height profiles revealing change in thickness of the monolayer MoS₂ flake upon 2nd growth of Sb₂S₃ i) Raman spectra of bare monolayer MoS₂ and Sb₂S₃/MoS₂ heterostructure. The A_1g and E_2g modes correspond to typical monolayer MoS₂ (labelled in blue). The additional peaks belong to B_3g and A_g groups corresponding to Sb₂S₃(labelled in green).

To gain quantitative insight into the thickness variation across the heterostructure, we conducted atomic force microscopy (AFM) analysis following the second CVD step. The AFM image shown in Figure 1g reveals a uniform thickness of ≈0.8 nm of as‐grown MoS₂ single crystal domain, consistent with a CVD grown monolayer MoS₂ films.^{[ 49 ]}However, following the deposition of Sb₂S₃, the AFM image shown in Figure 1h reveals an increased thickness of ≈8.2 nm across the Sb₂S₃/MoS₂ stack, confirming the formation of a Sb₂S₃ film of ≈7.4 nm thickness. Notably, the AFM line‐scan reflects that the Sb₂S₃ grows uniformly on the MoS₂ surface, whereas on adjacent SiO₂ regions, the deposition is rough and non‐uniform, with randomly distributed 3D particulates indicating uncontrolled nucleation. This preferential growth behavior is further corroborated by scanning electron microscopy (SEM) imaging (Figure S3, Supporting Information), which reveals quasi‐continuous Sb₂S₃ film, formed by laterally coalesced domains on the MoS₂ template and random 3D particulate deposition on bare SiO₂. These observations highlight the role of MoS₂ as a vdW template, where its atomically flat surface facilitates lateral adatom diffusion and ordered nucleation of Sb₂S₃ via interfacial vdW interactions (discussed later). Furthermore, we found that the thickness of 2D Sb₂S₃ can be systematically tuned by adjusting the duration of the second CVD growth, achieving isolated flakes as thin as ≈5.8 nm (Figure S4, Supporting Information). While prolonged growth for 30 min yielded quasi‐continuous Sb₂S₃ films with a thickness of ≈28.6 nm (Figure S5, Supporting Information). For all subsequent characterization and device fabrication in this study, we focused on ultrathin films (≈8 nm) to ensure consistency and maximize performance in atomically thin heterostructure devices.

To assess the structural evolution and interfacial coupling in the vdW heterostructure, we carried out Raman spectroscopy on both monolayer MoS₂ and the resulting Sb₂S₃ films grown on MoS₂. As shown in Figure 1i, the Raman spectrum of monolayer MoS₂ displays two prominent peaks: the E¹ _2g mode (≈382 cm⁻¹) corresponding to in‐plane vibrations, and the A_1g mode (≈404 cm⁻¹) associated with out‐of‐plane vibrations of Mo–S atoms within the lattice structure.^{[ 49 ]} These peaks exhibit a typical frequency separation (Δ) of ≈21 cm⁻¹, consistent with our previous LPCVD grown monolayer MoS₂ with S vacancy defects.^{[ 49 ]} Following Sb₂S₃ deposition, the Raman spectrum of MoS₂ reveals a slight red shift in the A_1g mode from 404 to 403 cm⁻¹, while the E_2g mode remains unchanged at ≈382 cm⁻¹. This red shift in A_1g is attributed to n‐type doping in MoS₂ induced by the Sb₃S₃ overlayer, consistent with previous studies and corroborated by our FET measurements (discussed later), which show increased electron concentration. In addition, to further probe the spectral features, peak deconvolution was performed(Figure S6, Supporting Information), which revealed broadening of the peaks, indicating a modification in the phonon environment and bonding dynamics at the interface.^{[ 50 ]} The shift and broadening are likely arising from interlayer interactions, strain, or charge transfer effects induced during the Sb₂S₃/MoS₂ heterostructure formation.^{[ 51 ]} The broadening of the E_2g mode, which is attributed to inhomogeneous tensile strain resulting from the lattice mismatch between Sb₂S₃ and MoS_2, which is consistent with strain modulation commonly observed in vdW heterostructures. Typically, the high‐temperature CVD growth promotes strong interfacial contact that helps retain this strain, thereby contributing to the observed phonon broadening.^{[ 52 ]} Nevertheless, the more pronounced shift in the A_1g mode suggests that out‐of‐plane vibrations are more strongly affected by the overlayer, consistent with vdW coupling at the heterointerface.^{[ 53 ]} In addition, weak shoulder features near 375 and 408 cm⁻¹, corresponding to LO(M) and ZO(M) modes, respectively, are evident. These features suggest the presence of S vacancy‐induced or interfacial defects, in line with our previous work on defect‐engineered CVD‐grown MoS₂.^{[ 47 , 54} Importantly, our high‐resolution transmission electron microscopy (HRTEM) and selected area electron diffraction (SAED) analyses (discussed later) confirm that the underlying MoS₂ retains its single‐crystalline nature after Sb₂S₃ growth, demonstrating that the template‐assisted growth process preserves the structural integrity of the MoS₂ template. Beyond the MoS₂ modes, the Raman spectrum of the Sb₂S₃/MoS₂ heterostructure exhibits multiple peaks(Figure 1i) at ≈50, ≈60, ≈71, ≈129, ≈148, ≈238, ≈282, and ≈302 cm⁻¹, which can be assigned to A_g and B_3g vibrational modes of Sb–S bonds, consistent with the orthorhombic Pnma phase of Sb₂S₃.^{[ 54 ]} Furthermore, our preliminary polarized Raman of the ≈282 cm⁻¹ mode shows angular variation (Figure S7, Supporting Information), suggesting anisotropic vibrational behavior consistent with orthorhombic symmetry. In the as‐grown samples, Raman peaks appear broadened (Figure S8, Supporting Information), which we attribute to lattice disorder and interfacial strain introduced during the initial CVD growth. However, following mild annealing at 300 °C in an N₂ environment, these peaks become sharper and more clearly resolved. A weaker peak at ≈188 cm⁻¹ is assigned to α‐Sb₂O₃, likely originating from laser‐induced photo‐oxidation of Sb₂S₃ in ambient conditions, which is an effect commonly observed due to the chemical instability of the stibnite phase.^{[ 50} , ^{55 ]}

To further evaluate the chemical composition and vertical stacking of the as‐grown MoS₂ and Sb₂S₃/MoS₂ heterostructure samples, we employed a combination of Time‐of‐Flight Secondary Ion Mass Spectrometry (ToF‐SIMS) and X‐ray Photoelectron Spectroscopy (XPS). ToF‐SIMS 3D analysis (depth profiling and imaging) is particularly well‐suited for analyzing 2D heterostructures, as it provides spatially resolved chemical information both at the surface and along the depth of the material.^{[ 56 ]} Figure 2a,b displays the ToF‐SIMS surface maps of S⁻ and SbS⁻ ions, corresponding to monolayer MoS₂ and Sb₂S₃ grown on MoS₂ template, respectively. The observed triangular features in both maps closely match the single‐crystal shapes seen in the optical images (Figure 1e,f), confirming the spatial correlation between the optical and chemical signatures. To further resolve the vertical composition, we performed ToF‐SIMS depth profiling using Cs⁺ sputtering on the Sb₂S₃/MoS₂ stacked sample. As shown in Figure 2c, the initial increase in Sb₂S₃ ⁻ signal indicates the presence of the Sb₂S₃ top layer. After ≈80 s of sputtering, the Sb₂S₃‐ intensity begins to decrease, coinciding with a sharp rise in MoS₂ ⁻ signal, confirming the underlying MoS₂ layer. Continued sputtering results in a decline of MoS₂ ⁻ signal and a subsequent rise in Si⁻ ion intensity, corresponding to the Si/SiO₂ substrate beneath the heterostructure. These depth‐resolved chemical profiles unambiguously validate the vertical stacking sequence of the heterostructure, with Sb₂S₃ on top of MoS_2, and demonstrate the layer‐specific distribution of chemical species across the interface. Moreover, a mixed ion MoSbS₂ ⁻ associating with Sb₂S₃ and MoS₂ was found at the interface of the heterostructure, which evidences the interaction between Sb₂S₃ and MoS₂. Figure 2d–f present the cumulative ion maps of Sb₂S₃ ⁻, MoS₂ ⁻, and Si⁻, integrated over the full sputtering duration (0–700 s), respectively, depicting the elemental distribution across the vertical depth of the Sb₂S₃/MoS₂ stack. To visualize the spatial arrangement of these species along the z‐axis (depth direction), we extracted cross‐sectional slices from both the horizontal (xz) and vertical (yz) planes across a representative single‐crystal domain. As shown in Figure 2g, the yz cross‐sectional map reveals a well‐defined vertical sequence: Sb₂S₃‐ ions are concentrated near the surface, followed by a distinct layer of MoS₂‐ ions, and finally, Si‐ signals corresponding to the SiO₂/Si substrate. This sequence confirms the vertical stacking order of the structure, with Sb₂S₃ grown atop MoS₂, as synthesized by vdW template‐assisted growth. Furthermore, the 3D ions distribution rendering (Figure 2h), provides a comprehensive visualization of the chemical architecture in vertical direction, clearly outlining the layered arrangement of Sb₂S₃, MoS₂, and the substrate, and consistent with the ToF‐SIMS depth profile (Figure 2c), reinforcing the chemical and spatial fidelity of the obtained heterostructure.

Figure 2.

Chemical compositional analysis of 2D Sb₂S₃ grown on MoS₂ vdW template a,b) ToF‐SIMS high lateral resolution surface map of S‐ and SbS‐ secondary ions representing as‐grown MoS₂ crystal and Sb₂S₃/MoS₂ stacked sample. The distribution of chemical maps is consistent with optical images of triangular single crystals. c) Depth profile of Sb₂S₃‐, SbMoS₂‐, MoS₂‐ and Si‐ ions against sputtering time of Sb₂S₃/MoS₂ stack grown on SiO₂/Si substrate.d–f) Sb₂S₃ ^–, MoS₂‐ and Si‐ ion map accumulated from depth profiling after a certain time of Cs⁺ sputtering, unveiling the Sb₂S₃, MoS_2, and SiO₂ surfaces underneath, respectively. g) 2D and h) 3D cross‐section view of Sb₂S₃ ^–, MoS₂‐ and Si‐ ion distribution in the z‐axis obtained from mapping xz and yz slices. The yz cross‐section reveals the stacking of the Sb₂S₃/MoS₂ heterostructure on SiO₂/Si substrate. i,j) XPS spectra of the Mo 3d and Sb 3d obtained for as‐grown MoS₂ and Sb₂S₃ grown on MoS₂ template, showing characteristic peaks of Mo⁴⁺ and Sb³⁺ 3d doublets, respectively. Also shows S 2s associated with MoS₂ and Sb₂S₃.

To corroborate these insights from ToF‐SIMS analysis, we further investigated the chemical states and bonding configurations in the MoS₂ and Sb₂S₃ nanosheets on MoS₂ template using XPS measurements. The high‐resolution XPS spectrum of the bare MoS₂ sample is shown in Figure 2i, where clear peaks from the Mo 3d and S 2s orbitals are observed. The dominant peaks at 232.51 and 229.31 eV are assigned to Mo 3d_3/2 and Mo 3d_5/2 orbitals of Mo⁴⁺, respectively, while the peak at 226.41 eV corresponds to S 2s core level, confirming the presence of chemically bonded Mo and S species consistent with MoS₂.^{[ 49 ]} In addition to these main peaks, weaker doublets centered at approximately 232 and 235 eV suggest the presence of oxidized Mo species (MoO_x), which likely arise from minor Mo–O bonding due to ambient air exposure.^{[ 47 ]} In contrast, the Mo 3d signal is largely suppressed in the XPS spectrum of the Sb₂S₃/MoS₂ stacked sample (Figure S9, Supporting Information), which we attribute to photoelectron attenuation by the overlying Sb₂S₃ layer. Nevertheless, the presence and stacking order of MoS₂ beneath the Sb₂S₃ layer are independently validated by Raman spectroscopy and ToF‐SIMS depth profiling (Figure 2c,g,h), which confirm the layered stacking order and spatial continuity of the heterostructure. The corresponding Sb 3d spectrum for the Sb₂S₃/MoS₂ heterostructure (Figure 2j) shows distinct peaks at 529.3 eV (Sb 3d_5/2) and 538.6 eV (Sb 3d_3/2), attributed to Sb³⁺ in Sb₂S₃, confirming the formation of the intended orthorhombic phase. Quantitative analysis yields an Sb 3d (Sb₂S₃) to S 2s signal ratio of ≈0.4, in excellent agreement with the ideal 2:3 stoichiometry.^{[ 57 ]} A weaker secondary Sb 3d doublet at 529.9 eV (Sb 3d_5/2) and 539.3 eV (Sb 3d_3/2) corresponds to Sb₂O_3. Based on the inelastic mean free path of Sb 3d electrons (1.9 nm for Sb₂S₃ and 1.8 nm for Sb₂O₃), the surface oxide overlayer thickness is estimated to be ≈0.7 nm (accounting for ≈33% of the total Sb signal).^{[ 58 ]} This surface oxide layer is consistent with the minor oxide peak in our Raman spectra and is attributed to the known chemical instability of Sb₂S₃ in ambient oxygen, reported previously.^{[ 55 ]} Collectively, the complementary results from ToF‐SIMS, and quantitative XPS confirm that the overlayer is predominantly stoichiometric Sb₂S₃ with slight surface oxidation arising from ambient exposure.

To gain deeper insight into the mechanism driving the preferential lateral growth of 2D Sb₂S₃ on MoS₂ template, we examined the morphological evolution at the growth interface using high‐resolution AFM and TEM analyses. As illustrated in Figure 3a, the vdW surface of monolayer MoS₂ plays a decisive role in directing the lateral 2D growth of 2D Sb₂S₃, in contrast to the randomly oriented 3D particulates observed on bare SiO₂. High‐resolution AFM images (Figure 3b,c) of Sb₂S₃ grown on MoS₂ single‐crystal template supported on SiO₂/Si substrates enable a direct comparison of growth behavior across the two surfaces. The AFM scans, along with the corresponding line profiles (insets), clearly show that Sb₂S₃ grown on MoS₂ is atomically smooth and has uniform coverage, indicative of layered growth. In contrast, Sb₂S₃ nucleated on bare SiO₂ exhibits discrete, randomly distributed 3D clusters, as evidenced by the rough and irregular height profile, and Raman spectra acquired in this region shows no peaks corresponding to Sb₂S₃ (Figure S10, Supporting Information). Interestingly, at longer growth durations, the Sb₂S₃ films formed on MoS₂ become thicker yet remain flat, forming well‐defined 2D nanosheets with lateral dimensions of several tens of nanometers, as shown in Figure 3c. This evolution underscores the stability of the Sb₂S₃ layered morphology and further validates a template‐driven 2D growth mechanism, enabled by the vdW surface of MoS₂. To directly probe the atomic structure and crystallinity of 2D Sb₂S₃ grown on MoS₂, we performed HRTEM analysis on Sb₂S₃/MoS₂ stacks transferred directly from the growth substrate to the holey carbon‐coated Cu TEM grids.^{[ 59 ]} The low‐magnification TEM image (Figure 3d) reveals a uniform distribution of well‐faceted triangular Sb₂S₃ flakes across the underlying MoS₂ domains. Alongside these flakes, darker diffuse regions are visible, which we attribute to the initial nucleation of amorphous Sb₂S₃ seeds formed during the early growth stage, indicative of a nucleation–growth transition from amorphous clusters to well‐defined crystalline domains. The corresponding annular dark‐field STEM image (Figure 3e), together with energy‐dispersive X‐ray spectroscopy (EDS) mapping (Figure S11, Supporting Information), further confirms the presence of discrete triangular Sb₂S₃ flakes, randomly oriented on the MoS₂ surface. High‐resolution TEM imaging (Figure 3f) shows clear lattice fringes with a measured interplanar spacing of 3.05 Å, corresponding to the (004) planes of orthorhombic Sb₂S₃. The selected‐area electron diffraction (SAED) pattern from the same region exhibits diffraction spots indexed to orthorhombic Sb₂S₃ ((110), (1–10), (–110), (–1–10)) alongside the six‐fold symmetric spots of hexagonal MoS₂, confirming the stacking of Sb₂S₃ crystals on MoS₂ template. Importantly, the sharp diffraction spots inherent to single‐crystal MoS₂ (Figure S12, Supporting Information) indicate that the underlying single‐crystal remains structurally intact after Sb₂S₃ growth, demonstrating preservation of the MoS₂ lattice during the second growth step. Notably, SAED analysis of multiple flakes indicates that Sb₂S₃ adopts arbitrary in‐plane orientations relative to MoS₂. The absence of any preferred alignment confirms the lack of a strictly epitaxial relationship, given the crystallographic incompatibility between hexagonal MoS₂ and orthorhombic Sb₂S₃ lattices. Instead, the appearance of laterally oriented triangular Sb₂S₃ flakes points to a, template‐assisted growth process, in which vdW interactions on the atomically flat MoS₂ surface suppress the intrinsic chain‐growth tendency and promote lateral growth governed by surface energy minimization rather than lattice matching.

Figure 3.

Mechanism of vdW template‐assisted growth of Sb₂S₃ on MoS₂. a) Schematic illustration of lateral 2D Sb₂S₃ growth on MoS₂ surfaces contrasted with uncontrolled Sb₂S₃ cluster nucleation on SiO₂ surface. b,c) High‐resolution AFM scan of Sb₂S₃ film grown on MoS₂ template with in‐set height profiles unveiling the quasi‐continuous 2D Sb₂S₃ film formation on MoS₂ and Sb₂S₃ particulates on SiO₂ surface. d) Low‐magnification TEM image revealing triangular Sb₂S₃ flakes and amorphous seed‐like features on MoS₂ (inset: high magnification view). e) Annular dark‐field STEM image showing randomly oriented triangular Sb₂S₃ flakes on MoS₂. f) Atomic resolution TEM image of the Sb₂S₃/MoS₂ showing (004) planes of Sb₂S₃ (d = 3.05 Å). SAED pattern (inset) displays diffraction spots from orthorhombic Sb₂S₃ and hexagonal MoS₂. g,h) Atomic structural model obtained from DFT showing adsorption of Sb₂S₃ on MoS₂ and SiO_2, respectively. i) DFT energy profiles for diffusion of S and Sb atoms and adsorption of Sb₂S₃ molecules on amorphous SiO₂ and MoS₂ vdW surface, respectively.

To validate our experimental observations and gain deeper insight into the growth kinetics and interfacial energetics governing the selective lateral growth of 2D Sb₂S₃, we performed a theoretical investigation including DFT, comparing its behavior on MoS₂ and amorphous SiO₂ surfaces. As shown in Figure 3g–i, we calculated the diffusion energy barriers for S atoms and Sb atoms, which are critical constituents in Sb₂S₃ formation on both surfaces. Our DFT calculations reveal a stark contrast in adatom mobility on MoS₂ versus the bare SiO₂ surface, such as S and Sb atoms exhibit low diffusion energy barriers of 0.3 and 0.275 eV, respectively, whereas on SiO₂, the corresponding values rise sharply to 2.5 and 2.41 eV (Figure 3i). These results confirm that the MoS₂ surface provides a favorable pathway for surface diffusion, promoting efficient lateral adatom migration, promoting layered growth of Sb₂S₃, which is consistent with the smooth, uniform 2D films observed in AFM and TEM analysis for 2D Sb₂S₃/MoS₂ heterostructure. In addition to diffusion kinetics, adsorption energy (E_ads) calculations further elucidate the difference in interfacial interaction. The E_ads of an Sb₂S₃ molecule on MoS₂ is –0.011 eV, indicating weak physisorption governed by vdW interactions, allowing the precursors to diffuse freely. In contrast, the adsorption of Sb₂S₃ on amorphous SiO₂ is much stronger (E_ads = –4.027 eV), indicating strong chemisorption, which traps adatoms and impedes their mobility. This explains the formation of 3D clusters and amorphous domains observed on SiO₂ in the absence of a vdW template. Collectively, these DFT results provide a thermodynamic and kinetic rationale for the experimentally observed selective lateral growth of 2D Sb₂S₃ on MoS₂. The MoS₂ vdW surfaces lowers the energy barriers, enhances adatom mobility, and promotes ordered 2D growth, in contrast to oxide substrate (non‐vdW surfaces) that that trap adatoms due to stronger, immobilizing interactions leading to random 3D nucleation. Furthermore, the versatility of vdW template‐assisted approach is demonstrated by the successful growth of 2D Sb₂S₃ on monolayer WS₂ single crystals, yielding uniform coverage and distinct Raman modes from both constituents (Figure S13, Supporting Information), confirming its applicability across a broader range of TMD templates beyond MoS₂.

With the successful synthesis and structural verification of template‐assisted 2D Sb₂S₃ growth on MoS₂, we next evaluated their intrinsic optical and electronic properties and explored functional potential for optoelectronic and electronic device applications. We acquired room temperature photoluminescence (PL) spectra from monolayer MoS₂ and Sb₂S₃/MoS₂ heterostructure samples, shown in Figure 4a. The PL spectrum of monolayer MoS₂, exhibits a strong emission peak at ≈1.86 eV, which corresponds to A exciton, and a weaker excitonic peak at 2.01 eV associated with B exciton, arising from direct bandgap transition and spin–orbit‐induced valence band splitting, respectively.^{[ 49 ]} Interestingly, following Sb₂S₃ deposition, the PL spectrum of the Sb₂S₃/MoS₂ flakes exhibits a red shift in both peaks to ≈1.79 eV (exciton‐A) and ≈1.96 eV (exciton‐B), accompanied by an enhancement in the PL intensity, suggesting an interlayer electronic coupling.^{[ 60 ]} This behavior suggests interfacial charge transfer between Sb₂S₃ and MoS₂ layers. To investigate this further, time‐resolved PL (TRPL) measurements were performed to observe the decay lifetime of excitons, as presented in Figure 4b. A bi‐exponential decay model was used to extract the fast (τ ₁) and slow (τ ₂) components of exciton recombination in both monolayer MoS₂ and Sb₂S₃/MoS₂ heterostructure samples. The fast decay (τ ₁ ≈0.3 ns) remained unchanged across both samples, while the slow component (τ ₂) increased slightly from 1.31 ns in MoS₂ to 1.40 ns in the heterostructure, suggesting modest suppression of non‐radiative recombination and evidence of interfacial charge transfer dynamics. Figure 4c presents in the absorbance spectra acquired from monolayer MoS₂ and Sb₂S₃/MoS₂ stack samples, further supporting these findings of modifications in optical behavior following Sb₂S₃ deposition. Monolayer MoS₂ exhibits typical excitonic absorption peaks at ≈610 nm (≈2.02 eV) and ≈658 nm (≈1.88 eV), consistent with the A and B excitons observed in the respective PL spectrum. In the Sb₂S₃/ MoS₂ stack samples, these features are not only retained but also significantly enhanced, confirming improved visible light absorption. Importantly, while the absorbance of bare MoS₂ declines sharply beyond 750 nm, the heterostructure retains substantial absorption well into the near‐infrared (NIR) region, indicative of an extended spectral photoresponse enabled by Sb₂S₃, which we observed (as discussed in the next section). Additional absorbance enhancement is also observed in the UV region (<500 nm), where Sb₂S₃ dominates due to its high absorption coefficient in the UV range.^{[ 31 ]} These results demonstrate that the integration of Sb₂S₃ onto MoS₂ not only preserves the inherent excitonic features of MoS₂ but also broadens and amplifies the optical response, effectively extending the operational window from UV–visible in pristine MoS₂ to UV–visible–NIR in the heterostructure. Moreover, the Sb₂S₃/ MoS₂ heterostructure is expected to form a type‐II band alignment (Figure S14, Supporting Information),^{[ 57 ]} which facilitates efficient spatial separation of charge carriers, enabling the extraction of lower‐energy photons and supporting enhanced device performance based on this template‐assisted grown Sb₂S₃/ MoS₂ heterostructure.

Figure 4.

Optical, electronic, and optoelectronic characterization a) PL spectra reveal a decrease in bandgap in Sb₂S₃/MoS₂ heterostructure compared to only monolayer MoS₂ b) Time‐resolved PL measurements showing longer decay lifetimes for Sb₂S₃/MoS₂ sample than MoS₂ only sample. c) UV–visible‐IR light absorbance spectrum showing higher and broader range absorption in the heterostructure than bare MoS₂. d) 3D schematic of Sb₂S₃/MoS₂ heterostructure on 300 nm SiO₂/Si substrate. The in‐set shows the optical image of a representative fabricated device. The channel length is ≈ 3 µm. e) Transfer characteristics of MoS₂ and Sb₂S₃/MoS₂ based FET, with current on a log and linear scale f) comparison of field‐effect electron mobility, carrier density, and I_on/I_off for MoS₂ and Sb₂S₃/MoS₂ FETs g,h) Time‐based photo‐response measurements of MoS₂ and Sb₂S₃/MoS₂ photodetectors under 660 and 1050 nm illumination, 4 mW cm⁻² power density and at constant V_ds of 2 V i) Photoresponsivity and detectivity of MoS₂ and vdW surface enabled grown Sb₂S₃/MoS₂ stack, respectively.

To demonstrate the functional potential of the vdW template‐assisted grown 2D Sb₂S₃/MoS₂ heterostructure for electronic and optoelectronic device applications, we fabricated devices directly on the as‐grown samples. Maskless photolithography was employed to define metal contacts and evaluate both the electronic and optoelectronic performance of the heterostructure, with monolayer MoS₂ devices used as reference. Figure 4d shows the 3D schematic and corresponding optical micrograph (inset) of a representative Sb₂S₃/MoS₂ device fabricated on a 300 nm SiO₂/Si substrate. The devices utilize lateral Ni/Au top contacts as source and drain, enabling both configurations for operating as two‐terminal photodetectors and three‐terminal FETs, using the highly doped Si (p⁺⁺) as a global back gate. We first evaluated the electrical performance of Sb₂S₃/MoS₂ devices as compared to MoS₂‐based FETs, as shown in Figure 4e,f. The transfer characteristics (Figure 4e) reveal a marked improvement in the on‐state current (I_o) for epitaxial grown Sb₂S₃/MoS₂ devices compared to pristine MoS₂ based FETs, indicating enhanced channel conductivity. Furthermore, both devices exhibit n‐type behavior with negative threshold voltages (V_th); however, the Sb₂S₃/MoS₂ device shows a larger shift (V_th ≈–57 V) compared to the monolayer MoS₂ device (V ≈–32 V), suggesting a higher degree of n‐doping induced in the heterostructure. This threshold shift is attributed to interfacial charge transfer in the vdW‐stacked 2D Sb₂S₃/MoS₂ heterostructure, which increases the intrinsic carrier concentration. As supported by our calculations (see Supporting Information Section on page 10 for further details), the field‐effect carrier density (n _e) in the Sb₂S₃/MoS₂ device reaches 1.94 × 10¹² cm⁻², which is an order of magnitude higher than that of the MoS₂ device (n _e  =  1.44 × 10¹¹ cm⁻²), as shown in Figure 4f. Furthermore, the Sb₂S₃/MoS₂ devices exhibit a significantly improved field‐effect mobility of≈ 5 cm² V⁻¹ s⁻¹, compared to ≈1.2 cm² V⁻¹ s⁻¹ for pristine monolayer MoS₂ FETs (Figure 4f). This enhancement arises primarily from efficient interfacial charge transfer. The type‐II band alignment between Sb₂S₃ and MoS₂ facilitates rapid electron–hole separation and generates a built‐in electric field that screens Coulomb scattering from charged impurities and mitigates surface‐related scattering in MoS₂.^{[ 48 ]} Such interlayer charge‐transfer‐induced screening, as reported in 2D heterostructures,^{[ 61 ]} enables higher mobility even in the presence of increased carrier concentration. Furthermore, the output current‐voltage (I _DS‐V _DS) characteristics measured at room temperature (Figure S15, Supporting Information), exhibit linear ohmic behavior for both device types, consistent with our previous studies on monolayer MoS₂ FETs. This ohmic response indicates low contact resistance and a small Schottky barrier height (SBH) at the Ni/Au–semiconductor interface.^{[ 62 ]} Therefore, in Sb₂S₃/MoS₂ devices, the combined effects of interlayer charge‐transfer‐induced screening, reduced SBH, and improved interfacial charge injection lead to enhanced on‐state current (I_on) and field‐effect mobility, confirming that vdW surfaces integration of Sb₂S₃ with monolayer MoS₂ is an effective strategy for modulating and boosting the electronic transport properties of potential atomically thin FET platforms.

The combination of broadband optical absorption and favorable interfacial charge dynamics makes the Sb₂S₃/MoS₂ hybrid structure a highly promising platform for advanced optoelectronic applications, particularly broadband photodetectors. To evaluate its potential as a photodetector, we performed time‐resolved photocurrent measurements under illumination at multiple wavelengths spanning the UV–visible–NIR spectrum (365, 455, 565, 660, 850, and 1050 nm), using a constant LED power density of 4 mW cm⁻² and a fixed readout voltage of 2 V (See Figure S16, Supporting Information for all wavelength). The Sb₂S₃/MoS₂ devices exhibit a clear and consistent increase in photocurrent (I_DS) across all tested wavelengths, demonstrating broadband photodetection capability. Notably, under 660 nm illumination, the heterostructure device shows an enhancement of approximately two orders of magnitude in photocurrent compared to the MoS₂‐only device, as shown in Figure 4g. This substantial increase in photoresponse is characteristic of type‐II band alignment,^{[ 57 ]} where efficient spatial separation of photogenerated electron–hole pairs enhances carrier extraction. The formation of a type‐II heterojunction between Sb₂S₃ and MoS₂ is further supported by the energy band alignment diagram derived from literature‐reported energy levels (see Figure S14, Supporting Information),^{[ 57 ]} confirming that the conduction and valence band offsets facilitate interlayer charge separation, also consistent with our optical characterization results discussed above. In addition to the enhanced photocurrent, the dark current (I_dark) of the Sb₂S₃/MoS₂ device is also approximately an order of magnitude higher than that of bare MoS₂, consistent with the previously observed n‐type doping effect introduced by Sb₂S₃. This increased conductivity is attributed to a higher free carrier concentration, corroborating our FET measurements and thus confirming that the Sb₂S₃ overlayer introduces interfacial charge transfer that boosts both dark and photo‐induced currents.

To quantitatively compare the performance of these devices as photodetectors, we calculated key figures of merit, including photoresponsivity (R) and specific detectivity (D^*), across a broad spectral range from UV to NIR, using the standard expressions (Supporting Information for details). Under a constant drain‐source bias of 2 V, monolayer MoS₂ photodetectors fail to respond to illumination in the NIR region, as shown in Figure 4h. In contrast, the Sb₂S₃/MoS₂ heterostructure devices exhibit strong and consistent photoresponse across the entire tested range, from 365 nm (UV) to 1050 nm (NIR) (Figure S16, Supporting Information), highlighting their broadband detection capability. As shown in Figure 4i, the heterostructure photodetectors exhibit an order of magnitude higher photoresponsivity and a broadened spectral detectivity range compared to devices based on pristine monolayer MoS₂. This superior performance is attributed to the efficient interfacial charge separation facilitated by the type‐II band alignment at the Sb₂S₃/MoS₂ interface, coupled with the enhanced broadband absorption of the heterostructure, as previously demonstrated in the absorbance spectra (Figure 4c). The absence of NIR response in monolayer MoS₂ devices is consistent with its limited absorption beyond the visible range, whereas the Sb₂S₃/MoS₂ heterostructure retains significant absorption well into the NIR, which we attribute to mid‐gap states introduced at the interface. These defect‐mediated states enable sub‐bandgap photon absorption, supported by our TRPL results (Figure 4b) showing an increased slow decay component (τ₂) from 1.31 ns in MoS₂ to 1.40 ns in the heterostructure. The type‐II band alignment further promotes carrier extraction from these mid‐gap states via a built‐in electric field,^{[ 48 ]} sustaining the NIR photoresponse up to ≈1050 nm. Together, these results establish 2D Sb₂S₃/MoS₂ heterostructures grown via vdW template‐assisted growth as a versatile and high‐performance material platform that simultaneously enhancing both electronic transport and broadband photodetection, offering strong prospects for integration into next‐generation 2D nanoelectronics and optoelectronic devices.

3. Conclusion

In this work, we have demonstrated a vdW template‐assisted growth approach that realizes controlled synthesis of layered 2D Sb₂S₃ nanosheets on monolayer MoS₂ single‐crystals. Unlike conventional oxide surfaces that promote 3D particulate growth due to strong substrate interactions, the vdW surface of MoS₂ facilitates low‐energy diffusion and layered lateral growth, as supported by experimental observations and DFT calculations. Furthermore, the resulting Sb₂S₃/MoS₂ heterostructure exhibits well‐defined interfaces, large lateral domain sizes, and sub‐8 nm thickness, forming structurally coherent vertical stacks. Since Sb₂S₃ coverage directly follows the lateral dimensions of the underlying MoS₂ single‐crystal domains, our approach is inherently scalable to wafer scale with continuous large‐area MoS₂ templates. Comprehensive optical and electrical characterizations reveal that vdW integration of Sb₂S₃ significantly modulates the properties of MoS₂, leading to enhanced light absorption from UV to NIR, interfacial charge transfer, and improved carrier transport. FETs based on the heterostructure show higher on‐state current, increased field‐effect mobility, and greater carrier densities compared to pristine MoS₂. In addition, broadband photodetectors fabricated from the Sb₂S₃/MoS₂ heterostructure exhibit strong responsivity and detectivity across a wide spectral range, enabled by type‐II band alignment and efficient photocarrier separation. In conclusion, this work overcomes the intrinsic 1D/3D growth tendency of Sb₂S₃ and establishes a scalable vdW template‐assisted growth strategy for realizing layered 2D pnictogen chalcogenide heterostructures with TMDs. The demonstrated ultrathin Sb₂S₃/MoS₂ platform, with its well‐defined interfaces and enhanced optoelectronic performance, opens new opportunities for future integration into the next generation of nanoelectronics and optoelectronic devices.

4. Experimental Section

CVD Growth of Monolayer MoS₂ and Sb₂S₃/MoS₂ Heterostructure

Monolayer MoS₂ single‐crystals were grown using a two‐temperature zone tube furnace in a low‐pressure (≈1 torr) CVD condition. The growth substrates, 300 nm SiO₂/Si were ultrasonically cleaned by acetone, IPA, and water sequentially, followed by N₂ blow drying. An aqueous solution of ammonium molybdate tetrahydrate was used as a Mo precursor, and was drop‐casted (5 µl droplet) onto the cleaned SiO₂/Si substrates, followed by baking at 110 °C for 5 min. The samples were then loaded into zone‐2 of the tube furnace, under the constant flow of 500 sccm Ar for 10 min. 200 mg of sulfur (S) powder was placed in zone‐1 and heated to 180 °C to create S vapors, which were carried downstream into zone‐2 by 70 sccm flow of carrier gas (Ar). The MoS₂ single crystals were grown at 750 °C for 20 min. For the 2nd growth of Sb₂S₃/MoS₂ heterostructure, the as‐grown MoS₂ single‐crystal was loaded into zone‐2. Sb₂S₃ precursor powder was in zone‐1 and sublimated at 550 °C, and the resulting vapors were transported by Ar as a carrier gas into zone‐2, where 2D Sb₂S₃ nanosheets were deposited on a single‐crystal MoS₂ template at 200 °C with a growth time of 10 min, while maintaining the growth pressure at 1 torr. After growth, the furnace was cooled to room temperature under the 500 sccm flow of Ar.

Material Characterization

To obtain optical images of the grown samples, a Leica microscope was utilized. For thickness and morphology analysis, Bruker Dimension Icon AFM with ScanAsyst‐air tip was used for measuring the thickness of the as grown samples. To obtain details about the structural properties of the samples, Raman and PL measurements were undertaken using a HORIBA LabRAM Raman spectrometer with a 532 nm laser source. In addition, the UV–vis‐NIR absorbance spectra were obtained using the CRAIC 20/30 UV−vis microspectrophotometer (CRAIC). JEOL JEM‐F200 CFEG TEM with an acceleration voltage of 200 kV was utilised for physical characterisations, i.e., TEM, HRTEM images, and SAED patterns. XPS data was acquired for the analysis of elemental composition of the defect‐enriched and passivated MoS₂ using a Kratos AXIS Supra XPS spectrometer with 1486.7 eV (Al Kα) X‐ray source.

ToF‐SIMS Measurements

ToF‐SIMS measurements were carried out with an IONTOF M6 Hybrid SIMS instrument (ION‐TOF GmbH, Münster, Germany). In this work, two types of ToF‐SIMS analyses were performed: 1) high lateral resolution surface chemical mapping and 2) ToF‐SIMS 3D analysis. For surface chemical mapping, a 30 keV Bi₃ ⁺ cluster beam was scanned over a specified raster area of typically 150×150 mm² with a pixel resolution of 256×256. The primary ion beam was optimized for the best spatial resolution (≈ 100 nm). For ToF‐SIMS 3D analysis, a 500 V‐Cs+ beam scanning over typically an area of 400×400 mm² with a current of 7 nA was used to sputter through the 2D materials. High spatial resolution images were taken at the center of Cs sputtering with 256×256 raster pixel sizes. The imaged area was typically 120 × 120 mm².

Device Fabrication

A standard photolithography process was used to fabricate two‐terminal lateral devices, which include source and drain channels with an active area of ≈25 µm². AZ5214E photoresist was spin‐coated onto the sample, and the source/drain electrodes were patterned with a maskless aligner (MLA150—Heidelberg Instruments). AZ400K and H₂O mixture (1:3) and AZ726MIF developers were used to develop MoS₂ and Sb₂S₃/MoS₂‐samples, respectively. After the development of patterned devices, 10 nm Ni followed by 100 nm Au metal contacts were evaporated onto the sample via electron beam deposition (PVD75—Kurt J. Lesker) with a base pressure of <5 × 10⁻⁷ Torr. The device fabrication process was ended by the lift‐off process in acetone to obtain MoS₂ and Sb₂S₃/MoS₂ devices.

Electrical and Optoelectronic Measurements

The electrical measurements were performed using a Keithley 4200SCS semiconductor parameter analyzer and an Agilent 2912A source meter.  Optoelectronic measurements were performed using a Keysight B2912A source measurement unit. All measurements for the devices were taken under dark ambient conditions except during optical excitation with uncollimated monochromatic LEDs (from Thorlabs Inc.) with wavelengths of 285, 365, 455, 565, 660, 850, and 1050 nm. A programmable Arduino‐uno microcontroller was used for pulse width modulations of these excitation sources.

Computational Methods

The optimized adsorption configuration was obtained via density functional theory using the projector augmented wave as implemented in the Vienna Ab initio Simulation Package.^{[ 63} , ^{64 ]} The Sb₂S₃ adsorption energy was calculated based on the equation: E_ads = E_tot – E_surface – E_molecule, where E_tot, E_surface, and E_molecule correspond to the total energy of the combined system, the total energy of the clean surface, and the total energy of the Sb₂S₃ isolated molecule. The Perdew–Burke–Ernzerhof approximation was used for the exchange‐correlation functional.^{[ 65 ]} The cut‐off energy of the plane‐wave basis was set to 400 eV, and self‐consistent calculations did not stop until energy changes were <1 × 10⁻⁴ eV. The initial geometries of amorphous SiO₂ were obtained by performing molecular dynamics calculations on the SiO₂(001) surface with O bond termination, as previously reported to be the lowest energy facet on SiO₂.^{[ 66 ]} The calculations were carried out using the NVT ensemble with a Nose–Hoover thermostat set to a target temperature of 700 K.^{[ 67 ]} All NVT‐MD calculations were performed at the Γ‐point. Simulation times of 20 ps with a timestep of 1 fs were found to be sufficient for surface reconstructions to take place. The formed new siloxane bridges under coordinated Si cations were chosen to be the diffusion path of S and Sb atoms. The minimum‐energy pathway and diffusion barriers for S and Sb atoms on MoS₂ and the SiO₂ surface were obtained from the climbing‐image nudged elastic band method as implemented within the Atomic Simulation Environment package with the FIRE algorithm, and CHGnet was employed as the calculator.^{[ 68 ]} All structures were fully relaxed until the residual force per atom reached <0.1 eV Å⁻¹.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Giridhar S. P., Abidi I. H., Qiu J., et al. “Van der Waals Template‐Assisted Growth of Two‐dimensional Sb₂S₃ .” Adv. Sci. 12, no. 46 (2025): e09903. 10.1002/advs.202509903

Data Availability Statement

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