Enhanced photoresponse in 2D seamless metal–semiconductor contact photodiodes via one-step sulfurization synthesis

Hyun Young Seo ab, Ojun Kwon ab, Minhee Kim ab, Seyoung Oh ab, Woojin Park b and Byungjin Cho *abc
aDepartment of Advanced Materials Engineering, Chungbuk National University, Chungdae-ro 1, Seowon-gu, Cheongju, Chungbuk 28644, Republic of Korea. E-mail: bjcho@chungbuk.ac.kr
bDepartment of Urban, Energy, Environmental Engineering, Chungbuk National University, Chungdae-ro 1, Seowon-gu, Cheongju, Chungbuk 28644, Republic of Korea
cBiomedical Research Institute, Chungbuk National University Hospital, 776, 1Sunhwan-ro, Seowon-gu, Cheongju, Chungbuk 28644, Republic of Korea

Received 28th December 2024 , Accepted 11th March 2025

First published on 21st March 2025


Abstract

Two-dimensional (2D) seamless metal–semiconductor contacts can significantly reduce the charge transport barrier in electronic switching devices by minimizing metal-induced defect states and eliminating dangling bonds. This approach outperforms conventional top contact and van der Waals (vdW) contact devices. However, the fabrication of a laterally connected 2D metal–semiconductor for photonic devices is highly complex and has been rarely studied. In this study, we demonstrate a high photoresponse and fast switching characteristics in a MoxNb1−xS2/n-MoS2/p-Si/Au stacked photodiode with a 2D seamless metal (MoxNb1−xS2)–semiconductor (MoS2) junction. This junction, formed by a one-step H2S sulfurization process of pre-deposited Nb2O5/MoO3 precursor films, allows for efficient synthesis. The 2D seamless contact-based diode exhibits excellent rectifying electrical properties, including a high forward current and a relatively low ideality factor of 2.61, outperforming those of photodiodes with conventional Au metal electrodes. It exhibited a high photo-responsivity of 2.1 A W−1 and a specific detectivity of 1.55 × 1012 Jones at −5 V under 530 nm illumination. Moreover, the device demonstrated reliable and fast switching dynamics, with a rise and fall time of approximately 17 ms at 850 nm. The high power-law exponent (∼0.89) suggests that the lateral covalent bonding between the alloyed MoxNb1−xS2 metallic layer and the MoS2 activation layer ensures efficient transport of photo-generated charge carriers, leading to a relatively low Schottky barrier height without a vdW tunneling gap. This simple and efficient creation of a smooth 2D metal–semiconductor junction paves the way for reproducible and fast photoelectronics.


1. Introduction

Two-dimensional (2D) transition metal dichalcogenide (TMDC) atomic layered nanomaterials have attracted significant attention for application in optoelectronics due to their exceptional electrical and optical properties. Among them, molybdenum disulfide (MoS2) has been extensively studied as a promising candidate for next-generation electronics because of its high mobility atomic thickness, tunable energy bandgap, and ease of synthesis. However, the conventional vacuum-based deposition process for top metal contact on a semiconducting 2D layer often causes local heating and cluster bombardment, leading to physical damage of the 2D surface and the formation of metal induced gap states (MIGSs).1,2 These MIGS-induced pinning effects result in a high Schottky barrier and increasing contact resistance degrading the device performance. Alternatively, the concept of using a side metal contact to the exposed sidewalls of thin 2D materials forms lateral covalent bonds, reducing the contact resistance.3–5 However, this device configuration involves a complex fabrication process, and suffers from low reproducibility and scalability. A clean, defect-free interface can also be achieved by stacking 2D–2D van der Waals (vdW) heterostructures,3 naturally eliminating the MIGSs and preventing Fermi level pinning.6–8 Nevertheless, the presence of a vdW tunneling gap at the interface still limits charge injection efficiency.

To overcome this limitation, seamless contacts that laterally connect metallic 2D materials to semiconducting 2D materials have been suggested in a few recent studies.3,9–11 Ideal 2D metal–semiconductor seamless contacts, featuring lateral covalent bonds, provide a relatively clean interface without local impurity states, offering a path toward achieving near-Ohmic contact.9,10 Typical 2D TMDC layered nanomaterials exhibit both the metallic and semiconducting phases.12 The phase transition, without breaking the chemical bonds, can be induced by a distortion in the movement of chalcogenide atoms.12 Kappera et al. demonstrated a method to partially open the semiconducting 2H-MoS2 phase and selectively transform it into a metallic 1T phase through chemical treatment, achieving extremely low contact resistance in the transistor device.13 Controlled laser beam irradiation has also been shown to induce a phase transition from 2H-MoTe2 to metallic 1T′-MoTe2, creating 2D–2D seamless contacts. The resulting field effect transistor exhibited a very low Schottky barrier height of 10 meV, along with high mobility and contact stability.14 Seamless contacts can also be achieved through heterogeneous lateral epitaxy growth of 2D TMDC materials adjacent to pre-patterned monolayer graphene.15 Additionally, spatially controlled intercalation of Co atoms into the van der Waals gap of bilayer SnS2 nanomaterials has been used to tune the semiconducting 2D material into a degenerately doped 2D state, offering a promising method for fabricating 2D–2D seamless contacts. Optoelectronic devices, in particular, require strategies to facilitate charge transport of photo-generated carriers. However, creating 2D lateral heterostructures with dangling bond-free interfaces remains challenging. Therefore, easily applicable synthesis methods for forming 2D seamless metal–semiconductor contacts are essential to achieve high photoresponsivity and reliability in 2D TMD-based optoelectronic devices. Despite the potential benefits, 2D seamless contacts for photoelectronic applications have been rarely studied.

In this study, we present the high-performance photocurrent behavior of a MoxNb1−xS2/n-MoS2/p-Si/Au photodiode, featuring a 2D seamless metal–semiconductor contact between MoxNb1−xS2 and MoS2. MoxNb1−xS2 serves as the top electrode, while MoS2 acts as the photo-responsive active layer. Both materials were synthesized using a simple one-step H2S sulfurization process of the Nb2O5/MoO3 precursor films. Although the device structure inherently includes an n-MoS2/p-Si p–n junction, the dominant improvement in charge injection and transport efficiency originates from the MoxNb1−xS2/n-MoS2 seamless contact, as it effectively lowers the Schottky barrier and minimizes interfacial recombination losses. The resulting MoxNb1−xS2/n-MoS2/p-Si photodiode exhibits excellent rectifying behavior, characterized by a low ideality factor and high forward current. Notably, the photocurrent generated by the seamless contact photodiode is nearly three orders of magnitude higher than that of a conventional Au electrode-based photodiode. The lateral seamless contact formation between the 2D metallic MoxNb1−xS2 and the MoS2 semiconductor layer effectively lowers the Schottky barrier height, enhancing the transport of the photo-generated charge carriers across the interface and leading to an improved photoresponse. During the dynamic light switching test, the seamless contact photodiode demonstrated consistent and rapid photoresponse performance, with a rise and fall time of approximately 17 ms. This simple approach to creating an ideal 2D metal–semiconductor junction paves the way for reproducible, fast photoelectronic applications, as well as conventional electronic switching devices.

2. Experimental

2.1. Fabrication of the MoxNb1−xS2/n-MoS2/p-Si structure photodiode

Boron-doped p-type silicon (Si) was cut into a size of 1.5 × 1.5 cm2 and cleaned ultrasonically for 3 minutes each in acetone, ethanol, and deionized water in sequence. To remove residual impurities and render the surface hydrophilic, an ultraviolet (UV)-ozone treatment was applied for 20 minutes. A 9 nm thick MoO3 layer was then deposited as a precursor for MoS2 using radio frequency (RF) sputtering. The chamber pressure was maintained at 3 mTorr by continuously flowing 10 sccm of high-purity argon (Ar, 99.999%) gas, while sputtering from a 2-inch MoO3 (99.95%) target at 50 W RF power.

Subsequently, the MoO3-predeposited p-Si substrate was heat-treated on a hot plate at 200 °C for 10 minutes to remove surface impurities and improve adhesion to the substrate. To pattern the top electrode on the MoO3 layer, the Nb2O5 precursor film was sputtered using a shadow mask on the MoO3/p-Si substrate. The chamber pressure was kept at 3 mTorr, with 10 sccm of Ar gas flow, and 5 nm thick Nb2O5 was sputtered from a 2-inch Nb2O5 (99.95%) target at 50 W RF power. The Nb2O5/MoO3/p-Si substrate was then placed at the center of a quartz chamber, where Ar/H2S (99:1) mixed gas was introduced to sulfurize the pre-deposited oxide films, maintaining a synthesis pressure of 800 torr. The furnace temperature was increased to 900 °C and held for 1 hour to synthesize a MoxNb1−xS2/MoS2 heterostructure. The edge region of the as-synthesized MoS2 was etched away using conventional photolithography and the CF4 dry etching process. Finally, a 50 nm thick Au layer was thermally evaporated on the backside of the substrate to form the bottom contact of the photodiode.

3. Results and discussion

Utilizing the one-step sulfurization process of pre-deposited Nb2O5/MoO3 precursor oxide films, a 2D MoxNb1−xS2/MoS2 was successfully formed on p-Si substrate. Specifically, Fig. 1a shows the sequential fabrication steps of the photodiode device with MoxNb1−xS2/n-MoS2/p-Si/Au structures. As depicted in the enlarged images of the CVD-based sulfurization process, the alloyed metallic MoxNb1−xS2 (serving as the top electrode) was not only vertically bonded to MoS2 (the light-absorbing active layer) via van der Waals forces but also laterally covalently bonded to MoS2. This 2D–2D metal–semiconductor seamless contact naturally forms without the need for a complex synthesis technique, enabling the creation of a smooth and clean heterojunction interface with ease. The 2D MoxNb1−xS2 metallic layers, being semi-transparent to the visible lights in comparison to metals with stronger reflective properties, allow for superior light transmission, making them ideal for photonic applications. Fig. 1b (left) shows high-resolution transmission electron microscopy (HR-TEM) images of the cross-sectional 2D heterostructure, with corresponding elemental distribution mapping profiles as shown in Fig. 1b (right). The image reveals a layered MoxNb1−xS2/MoS2 heterojunction structure, with concentration of Nb atoms in the upper region and Mo atoms distributed throughout the layers, confirming the formation of an Nb-gradient 2D heterostructure across the vertical profile. Fig. 1c (left) displays a scanning transmission electron microscopy (STEM) image of the layered MoxNb1−xS2/MoS2/p-Si structures. The yellow line marks the boundary between the distinct 2D layers, while the green arrow indicates the direction of the line scan for elemental analysis. Fig. 1c (right) presents the depth profile of atomic elements (Nb, Mo, S, O, and Si) obtained from energy-dispersive x-ray spectroscopy (EDS). The elemental line scan across the heterostructure layers highlights the variation in the interface, emphasizing the Nb-gradient region within the MoxNb1−xS2 layer. A high percentage of oxygen atoms between the MoS2 and p-Si layers suggests the presence of native oxides grown from the Si substrate, consistent with the HR-TEM mapping image. The co-sulfurization and inter-diffusion of Nb2O5 and MoO3 precursor layers during the high-temperature synthesis process results in the distinct alloyed MoxNb1−xS2 layer, differentiated from the bottom MoS2. Previous studies have reported that transition metals such as Nb and Mo can diffuse into one another, leading to the formation of intermetallic 2D compounds.7,8
image file: d4tc05484a-f1.tif
Fig. 1 Sequential fabrication process and material characterization of the MoxNb1−xS2/n-MoS2/p-Si/Au photodiode. (a) Schematic illustration of the seamless 2D metal–semiconductor contact photodiode fabrication process, with an enlarged image highlighting the 2D sulfurization process from a pre-deposited oxide precursor film. (b) Cross-sectional HR-TEM and the corresponding EDS elemental mapping image of the fabricated device. (c) STEM image and elemental depth profile analysis from the EDS line scan (the green arrow on the left figure indicating the direction of the line scan) of the MoxNb1−xS2/n-MoS2/p-Si layers.

Fig. 2a provides a schematic planar view of the 2D heterostructure, comprising MoxNb1−xS2 and MoS2 layers. Unlike a conventional top contact to a 2D semiconductor,16 this structure features an unprecedented smooth lateral covalent bonding between two distinct 2D layers. This lateral bonding significantly reduces the contact resistance in the electronic device.3,10 The atomic-scale seamless contact between the 2D metal and semiconductor forms a clean, covalent bond-based interface devoid of local interfacial states or electronic boundaries, facilitating Ohmic contact.9,10,17 To characterize the optical properties of the 2D layered materials in the photodiode device, Raman spectra, photoluminescence (PL), and UV-visible absorption spectra were analyzed. Fig. 2b shows the Raman spectrum of the as-synthesized MoS2, with prominent peaks corresponding to in-plane (E2g) and out-of-plane (A1g) vibration modes at 383.7 cm−1 and 409.4 cm−1, respectively. The 25.7 cm−1 separation between these peaks indicates the presence of a multi-layered MoS2 film. For comparison, the Raman spectrum of as-grown NbS2, used as a reference for the alloyed MoxNb1−xS2, exhibited E1 and E2 vibration modes. These two E modes indicate a doubly degenerate mode, where Nb and S atoms vibrate in opposite directions,18–20 with peak positions at 295 cm−1 and 327 cm−1. The A1 vibration mode, representing symmetrical vibrations along the axial direction with S atoms vibrating perpendicular to the Nb–S plane,21–23 was observed at 381 cm−1, indicating the 3R-NbS2 phase. Meanwhile, the MoxNb1−xS2 layer, synthesized through a one-step CVD process, showed the 2H-MoS2 phase in most regions with 2H-phase NbS2. This resulted in the appearance of the E1g vibrational mode peaks at 251 and 304 cm−1.


image file: d4tc05484a-f2.tif
Fig. 2 Optical characterization of the MoxNb1−xS2/MoS2 heterostructure with 2D metal–semiconductor seamless contact. (a) Schematic top-view illustration of the MoxNb1−xS2/MoS2 heterostructure on a p-Si substrate. (b) Raman spectra of the MoS2, NbS2, and MoxNb1−xS2 films and (c) PL spectra of the MoS2 and MoxNb1−xS2. (d) UV-visible absorption spectra of MoS2 and MoxNb1−xS2.

Fig. 2c shows the photoluminescence (PL) spectra of multilayer MoS2 and MoxNb1−xS2. Compared to pristine MoS2, the PL intensity of MoxNb1−xS2 is quenched, indicating that radiative recombination of photo-generated carriers was suppressed. This is attributed to more efficient charge transfer from MoS2 to MoxNb1−xS2, resulting in a longer carrier lifetime.24Fig. 2d shows the UV-visible absorbance spectra of the pristine MoS2 and the MoxNb1−xS2 films. The pristine MoS2 exhibits a broad absorption spectrum, primarily due to strong absorption in the visible light range. The absorption peaks corresponding to A and B excitons in pristine MoS2 result from energy splitting caused by valence band spin-orbital coupling.25 Even though the direct exciton peaks of the MoxNb1−xS2 alloy layer, as observed in MoS2, disappeared with the emergence of Nb-related energy states, its strong absorption within the measured wavelength range suggests its suitability for photonic devices. To determine the band gap energy (Eg) of MoS2, the Tauc plot was derived from the UV-visible absorption spectrum using Tauc's equation as follows:

 
(αhv)2 = hvEg(1)
where α is the absorption coefficient, h is Planck's coefficient, and v is the frequency. From the Tauc plot, the energy band gap (Eg) of MoS2 was determined to be ∼1.68 eV. Fig. S1 (ESI) shows additional ultraviolet photoelectron spectra (UPS) of the as-synthesized pristine MoS2, NbS2, and MoxNb1−xS2 for determining work function (Φ). The work functions of MoS2 and NbS2 were calculated using the equation Φ = hvEoffset, where hv is the incident photon energy (21.22 eV) and Eoffset is the cutoff edge of the secondary electrons. Using the simple formula, the work functions of MoS2, NbS2, and MoxNb1−xS2 were determined to be ∼3.51, 3.91, and 3.39 eV, respectively. Additionally, by performing a linear extrapolation of the leading edge to the baseline, the valence band maxima (VBM, EV) of MoS2 and NbS2 were determined. Fig. S2 (ESI) shows the position of the Fermi energy (EF) relative to EV (EFEV), with EF located 0.97 eV above EV for MoS2 and 0.81 eV below EV for NbS2. It was found that the synthesized MoS2 exhibits n-type semiconducting properties, while NbS2 shows degenerate p-type characteristics with major hole carriers.

Fig. 3a shows a schematic illustration of the MoxNb1−xS2/n-MoS2/p-Si/Au photodiode featuring a seamless 2D metal–semiconductor lateral contact under light illumination. Fig. 3b compares the typical rectifying IV curves of the vertical diode devices using conventional Au metal electrodes or atomically thin 2D MoxNb1−xS2 to the active MoS2. The 2D metal–semiconductor seamless contact diode exhibited significantly higher forward current and an improved ideality factor (lower value of 2.61) compared to those of the Au reference devices. While the n-MoS2/p-Si junction contributes to rectification behavior, the dominant enhancement in charge injection and transport efficiency originates from the 2D seamless contact at the MoxNb1−xS2/MoS2 interface. This improvement is attributed to the lateral covalent bonding between metallic MoxNb1−xS2 and semiconducting MoS2 preventing metal-induced gap states and dangling bond-related defects, which avoid charge trapping and enhance charge transport and injection efficiency across the 2D heterostructure interface.9 Replacing Au with MoxNb1−xS2 significantly improved the diode's efficiency and reduced charge carrier recombination. Fig. 3c shows the current characteristics of the MoxNb1−xS2/n-MoS2/p-Si/Au seamless device in the reverse bias region (VDS = −5 V to −1 V) under both dark and light illuminated conditions, with varying wavelengths (455, 530, 656, and 850 nm) and intensities (1200, 473, 1190, and 973 μW cm−2). A significant increase in photocurrent was observed, from dark current (Idark) of 10−7A to 10−5 A of photocurrent (Iph) at −5 V, suggesting that the seamless photodiode can respond to a broad range of light wavelengths. As shown in the inset of Fig. 3c, the device demonstrates clear asymmetrical rectifying IDSVDS characteristics under dark conditions, with a considerable increase in reverse current under illumination.


image file: d4tc05484a-f3.tif
Fig. 3 Photocurrent characteristics of the seamless contact photodiode (MoxNb1−xS2/n-MoS2/p-Si/Au). (a) Schematic diagram of the seamless contact photodiode under light illumination. (b) Comparison of the IV rectifying properties and ideality factors between the seamless contact photodiode and the Au reference photodiode (Au/n-MoS2/p-Si/Au). (c) IV curves of the seamless contact photodiode in the reverse bias region under both dark and light conditions at varying wavelengths (455, 530, 656, and 850 nm). The inset shows its photocurrent IV characteristics over the entire voltage sweep. (d) Comparison of light intensity-dependent photocurrents between the seamless contact photodiode and the Au reference device under 656 nm. Energy band diagrams illustrating the difference of the Schottky barrier height affecting the transport of photo-induced carriers for (e) the Au reference device and (f) seamless contact device.

Fig. 3d compares light intensity-dependent photocurrents for both 2D seamless and Au reference devices at 656 nm light intensities ranging from 119 to 1190 μW cm−2. Iph increased gradually for both devices as intensity increased. However, the 2D seamless contact photodiode showed over 600 times greater photocurrent at VDS = −3 V than the Au reference device. Specifically, under 656 nm light at 1190 μW cm−2, ΔIph (IphIdark) for the Au reference was only ∼30.3 nA, whereas for the seamless 2D contact photodiode, it was significantly higher at ∼23.6 μA. This indicates that the smooth and clean interface between the 2D metal MoxNb1−xS2 and MoS2 effectively enhances the transport and separation of photo-generated carriers. Fig. S3 (ESI) shows the Iph/Idark ratio as a function of VDS, demonstrating that the increased dark current in the MoxNb1−xS2 device is well compensated by its enhanced photocurrent response, which ensures stable photodetector performance. Fig. 3e and f show the transport of photo-generated carriers from MoS2 through the Au and MoxNb1−xS2 electrodes, where a Schottky barrier is formed due to imperfect Fermi level alignment and interface defects.26,27 Despite the strong chemical bonds between Au and S atoms, metal-induced impurity and dangling bond-related defects contribute to a higher Schottky barrier.28,29 A relatively high Schottky barrier was formed at the contact between Au and n-MoS2 (Fig. 3e). In addition, Au atoms or clusters from direct evaporation often damage the 2D surface.30 The seamless MoxNb1−xS2/MoS2 interface, formed through lateral covalent bonding during high-temperature CVD synthesis, exhibits an almost negligible Schottky barrier (Fig. 3f). When photons from the incident light strike the MoS2 surface, electrons near the conduction band edge are excited to the valence band edge and then move toward the top electrode. The MoxNb1−xS2 layer facilitates easier charge transport through the smooth interface, leading to increased photocurrent. This feature underscores the advantage of 2D alloyed MoxNb1−xS2 electrodes over the Au electrode in the charge transport of photogenerated carriers in photodiode devices. Additionally, since the n-MoS2/p-Si interface remains the same across different photodiode structures, its effect on the photocurrent is negligible.

The photoresponse dynamics of the 2D seamless contact-based MoxNb1−xS2/n-MoS2/p-Si/Au photodiode, under varying wavelengths (450 to 850 nm) and light intensities (47 to 1220 μW cm−2), are shown in Fig. 4a–d. The time-dependent photoresponse measurements reveal distinct characteristics at different light intensities for each wavelength. As light intensity increases, more electron–hole pairs are generated, leading to a corresponding increase in photocurrent. These results demonstrate the photodiode's ability to respond to a broad range of light intensities, highlighting its potential for applications requiring sensitivity to diverse optical conditions. To estimate the response to a single light pulse, the rise and fall times of the photodiodes were measured to be ∼16.9 and 17.4 ms, respectively (Fig. 4e). The rise and fall times are defined as the time required for the photoresponse to increase from 10% to 90% of its maximum when the light is turned on, while the fall time refers to the time needed for the photoresponse to decrease from 90% to 10% when the light is turned off. These short switching times indicate that the 2D seamless contact-based photodiode can rapidly respond to incident light signals and return to its baseline state without any current fluctuation or drift. In contrast, the Au/n-MoS2/p-Si/Au photodiode exhibited a prolonged temporal response (Fig. S4, ESI), suggesting transport limitations due to interfacial recombination at the metal–semiconductor junction. To evaluate the reproducibility of the device's photo-switching behavior, the light on/off switching was modulated every 0.5 seconds under 850 nm wavelength illumination with a reading voltage of −0.3 V (Fig. 4f). A consistent photocurrent (∼7.7 μA) was generated at a frequency of 17.9 Hz during repeated switching, demonstrating highly reliable light-switching performance. Fig. S5 (ESI) also shows the fast and stable photo-switching characteristics at different wavelengths (656 and 850 nm). However, the reference device exhibited relatively poor photoresponse dynamics due to inefficient charge injection and partial trapping at defect states in the interface between the Au metal and the MoS2 semiconductor (Fig. S6, ESI).


image file: d4tc05484a-f4.tif
Fig. 4 Photoresponse characteristics of the seamless contact photodiode. Time-dependent switching dynamics of the photodiode under varying light intensities (47 to 1220 μW cm−2) with wavelengths of (a) 455 nm, (b) 530 nm, (c) 656 nm, and (d) 850 nm, with a reading voltage of −5 V. (e) Temporal photoresponse showing the rise and fall times under a single 850 nm light pulse. (f) Stable photocurrent switching dynamics of the photodiode under repeated light on/off switching.

To understand the correlation between incident light intensity and the photo-generated current, the photocurrents as a function of light intensities (119 to 1190 μW cm−2) at 656 nm light were plotted on a double logarithmic scale (Fig. 5a). The photocurrent increases with the light intensity, following a power-law relationship:31,32

 
IphP0.89(2)
where Iph is the photocurrent, and P is the incident light power. The linear behavior of the double-log[thin space (1/6-em)]Iphversus P can be explained by the photoelectric effect where the photocurrent depends on the light intensity. The efficiency of the energy transfer from a photon to a photo-generated charge carriers is estimated by the slope of the fitted line. The exponent values are often smaller than the ideal value of 1, mainly due to the trapping of photo carriers within the active materials and interfaces. Nevertheless, the power-law exponent of the 2D seamless contact-based photodiode was calculated to be ∼0.89, indicating that excellent charge transport of photo-generated carriers is facilitated by the ideal seamless edge contact between 2D metal–semiconductor (MoxNb1−xS2/n-MoS2) van der Waals layered nanomaterials. In contrast, the Au/n-MoS2/p-Si/Au photodiode exhibits negligible dependence on light intensity, as reflected by its near-zero power-law exponent (Fig. S7, ESI). This difference highlights the advantage of the 2D edge contact structure, which results in significantly lower contact resistance, approaching ohmic behavior, and leads to much higher performance compared to that of conventional 3D metal-based devices.9,10 Unlike conventional Au electrodes, which reflect the visible light, the high absorbance property of 2D metallic atomic layered nanomaterials is also advantageous for photoelectron generation, as illustrated in Fig. 2d.


image file: d4tc05484a-f5.tif
Fig. 5 (a) Photocurrent of the seamless contact photodiode as a function of light intensities under 656 nm light illumination. (b) Responsivity and detectivity of the photodiode as a function of wavelength (455 to 850 nm). (c) The EQE values as a function of wavelength. (d) Comparison of the performance parameters (responsivity and detectivity) of the seamless contact photodiode with those of previously reported 2D photodiodes.

The photo-responsivity (R) and specific detectivity (D*) were calculated as key parameters for comparing the photodiode performance, using the following equations:31–33

 
image file: d4tc05484a-t1.tif(3)
 
image file: d4tc05484a-t2.tif(4)
Here, Pin is the power of the incident light, q is the charge of an electron, and Jdark is the current density under dark conditions. The photo-responsivity and specific detectivity at varying wavelengths are summarized in Fig. 5b. The highest values of R and D* were calculated to be 2.1 A W−1 and 1.6 × 1012 Jones, respectively, at 530 nm. It is reasonable that the highest responsivity was observed at 530 nm wavelength because the PL peaks and absorption characteristics of Mo1−xNbxS2 and MoS2 become prominent at the wavelength (Fig. S8, ESI). In comparison, the Au/n-MoS2/p-Si/Au photodiode exhibited a peak responsivity of 0.29 A W−1 and a detectivity of 2.06 × 1011 Jones at 850 nm (Fig. S9a, ESI). The external quantum efficiency (EQE) of the 2D seamless contact-based photodiodes was also estimated at an applied bias of −5 V (Fig. 5c) using the equation:34
 
image file: d4tc05484a-t3.tif(5)
where h is the Planck's constant, c is the speed of light, and λ is the wavelength of the incident light. EQE highlights the ratio of collected charge carriers to the incident photons, expressed as a percentage. The EQE reached a maximum of 490% at 530 nm, which is consistent with the highest values of photo-responsivity and detectivity shown in Fig. 5b. The Au/n-MoS2/p-Si/Au device exhibited a maximum EQE of 50%, further demonstrating the improved light absorption and carrier transport properties enabled by the seamless 2D contact (Fig. S9b, ESI). Fig. 5d compares the performance of this work (indicated by the red star) with that of other studies32,35–41 in terms of responsivity and specific detectivity. The results highlight that our device demonstrates relatively high specific detectivity and responsivity compared to previous reports, showcasing the superior performance of the seamless contact photodetector.

4. Conclusions

In summary, we have demonstrated enhanced photoresponse characteristics in a MoxNb1−xS2/n-MoS2/p-Si/Au stacked photodiode with a uniquely seamless metal–semiconductor junction (MoxNb1−xS2/n-MoS2), easily formed by a one-step sulfurization process of the pre-deposited Nb2O5/MoO3 films. The MoxNb1−xS2/n-MoS2 heterostructure revealed the gradient variation of the Nb atoms across the vertical layers, showing the formation of the Nb diffusion-driven 2D mixed alloy layers. Unlike the reference device, which used a conventional Au metal electrode on MoS2, the lateral covalent bonding between the alloyed MoxNb1−xS2 top electrode and the MoS2 light absorption layer prevents the metal-induced defect states and Fermi level pinning near the 2D semiconductor, leading to the improved charge transport across the interface. As a result, the 2D seamless contact diode exhibited excellent rectifying electrical properties with a low ideality factor and significantly outperformed the reference photodiode. The 2D contact photodiode exhibited a higher photo-responsivity of 2.1 A W−1 and a specific detectivity of 1.55 × 1012 Jones at −5 V under 530 nm light illumination. The reproducible and fast dynamic behaviors, with rising and falling times of approximately 17 ms at 850 nm, were also highlighted during the light-switching tests. The high power-law exponent (∼0.89) of the seamless contact photodiode indicates that the lateral covalent bonding between the alloyed MoxNb1−xS2 metallic layer and the MoS2 light absorption layer ensures efficient transport of photo-generated charge carriers across the interface. This was attributed to the relatively low Schottky barrier height without a vdW tunneling gap. Consequently, the seamless 2D metal–semiconductor junction enables fast and reliable photoelectronic switching along with high photoresponse characteristics, paving the way for next-generation photoelectronic applications.

Author contributions

Hyun Young Seo: conceptualization, data curation, and writing – original draft. Ojun Kwon: data curation and writing – original draft. Minhee Kim: formal analysis. Seyoung Oh: formal analysis. Woojin Park: investigation. Byungjin Cho: supervision, conceptualization, and writing – review and editing.

Data availability

The data that support the findings of this study are available in the ESI.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This work was supported by the Basic Science Research Program of the National Research Foundation of Korea (NRF) and funded by the Ministry of Education (RS-2023-002419690). This work was also supported by the Technology Innovation Program Development Program-Development of alternative gases and process technologies with a GWP of 150 or less for display deposition and cleaning processes (RS-2023-00266568, development of alternative gas and process technology below GWP 150 for vapor deposition process for film deposition of display TFT protective film) funded by the Ministry of Trade, Industry & Energy (MOTIE, Korea).

References

  1. Y. Liu, J. Guo, E. Zhu, L. Liao, S.-J. Lee, M. Ding, I. Shakir, V. Gambin, Y. Huang and X. Duan, Nature, 2018, 557, 696–700 CAS .
  2. R. Zan, Q. M. Ramasse, R. Jalil, T. Georgiou, U. Bangert and K. S. Novoselov, ACS Nano, 2013, 7, 10167–10174 CAS .
  3. S. Chen, S. Wang, C. Wang, Z. Wang and Q. Liu, Nano Today, 2022, 42, 101372 CAS .
  4. L. Wang, I. Meric, P. Y. Huang, Q. Gao, Y. Gao, H. Tran, T. Taniguchi, K. Watanabe, L. M. Campos, D. A. Muller, J. Guo, P. Kim, J. Hone, K. L. Shepard and C. R. Dean, Science, 2013, 342, 614–617 CAS .
  5. H. Choi, B. H. Moon, J. H. Kim, S. J. Yun, G. H. Han, S. G. Lee, H. Z. Gul and Y. H. Lee, ACS Nano, 2019, 13, 13169–13175 CAS .
  6. Y. Liu, P. Stradins and S.-H. Wei, Sci. Adv., 2016, 2, e1600069 Search PubMed .
  7. Y. Kim, A. R. Kim, J. H. Yang, K. E. Chang, J. D. Kwon, S. Y. Choi, J. Park, K. E. Lee, D. H. Kim, S. M. Choi, K. H. Lee, B. H. Lee, M. G. Hahm and B. Cho, Nano Lett., 2016, 16, 5928–5933 CAS .
  8. A. R. Kim, Y. Kim, J. Nam, H. S. Chung, D. J. Kim, J. D. Kwon, S. W. Park, J. Park, S. Y. Choi, B. H. Lee, J. H. Park, K. H. Lee, D. H. Kim, S. M. Choi, P. M. Ajayan, M. G. Hahm and B. Cho, Nano Lett., 2016, 16, 1890–1895 CAS .
  9. A. Allain, J. Kang, K. Banerjee and A. Kis, Nat. Mater., 2015, 14, 1195–1205 CAS .
  10. J. H. Sung, H. Heo, S. Si, Y. H. Kim, H. R. Noh, K. Song, J. Kim, C. S. Lee, S. Y. Seo, D. H. Kim, H. K. Kim, H. W. Yeom, T. H. Kim, S. Y. Choi, J. S. Kim and M. H. Jo, Nat. Nanotechnol., 2017, 12, 1064–1070 CAS .
  11. S. Aftab, M. Z. Iqbal, M. W. Iqbal, M. Asghar and H. Ullah, J. Mater. Chem. C, 2022, 10, 14795–14811 CAS .
  12. K. Liu, P. Luo, W. Han, S. Yang, S. Zhou, H. Li and T. Zhai, Sci. Bull., 2019, 64, 1426–1435 Search PubMed .
  13. R. Kappera, D. Voiry, S. E. Yalcin, B. Branch, G. Gupta, A. D. Mohite and M. Chhowalla, Nat. Mater., 2014, 13, 1128–1134 CAS .
  14. S. Cho, S. Kim, J. H. Kim, J. Zhao, J. Seok, D. H. Keum, J. Baik, D. H. Choe, K. J. Chang, K. Suenaga, S. W. Kim, Y. H. Lee and H. Yang, Science, 2015, 349, 625–628 CAS .
  15. M. H. D. Guimarães, H. Gao, Y. Han, K. Kang, S. Xie, C. J. Kim, D. A. Muller, D. C. Ralph and J. Park, ACS Nano, 2016, 10, 6392–6399 Search PubMed .
  16. S. M. Song, J. K. Park, O. J. Sul and B. J. Cho, Nano Lett., 2012, 12, 3887–3892 CAS .
  17. J. Kang, D. Sarkar, Y. Khatami and K. Banerjee, Appl. Phys. Lett., 2013, 103, 083113 CrossRef .
  18. W. Ge, K. Kawahara, M. Tsuji and H. Ago, Nanoscale, 2013, 5, 5773–5778 RSC .
  19. Z. Li, W. Yang, Y. Losovyj, J. Chen, E. Xu, H. Liu, M. Werbianskyj, H. A. Fertig, X. Ye and S. Zhang, Nano Res., 2018, 11, 5978–5988 CrossRef CAS .
  20. S. Nakashima, Y. Tokuda, A. Mitsuishi, R. Aoki and Y. Hamaue, Solid State Commun., 1982, 42, 601–604 CrossRef CAS .
  21. S. Zhao, T. Hotta, T. Koretsune, K. Watanabe, T. Taniguchi, K. Sugawara, T. Takahashi, H. Shinohara and R. Kitaura, 2D Mater., 2016, 3, 025027 CrossRef .
  22. X. Wang, J. Lin, Y. Zhu, C. Luo, K. Suenaga, C. Cai and L. Xie, Nanoscale, 2017, 9, 16607–16611 Search PubMed .
  23. A. R. Mohmad, A. A. Hamzah, J. Yang, Y. Wang, I. Bozkurt, H. S. Shin, H. Y. Jeong and M. Chhowalla, Faraday Discuss., 2021, 227, 332–340 RSC .
  24. J. H. Nam, S. Oh, H. Y. Jang, O. Kwon, H. Park, W. Park, J. Kwon, Y. Kim and B. Cho, Adv. Funct. Mater., 2021, 31, 2104174 Search PubMed .
  25. N. Dong, Y. Li, Y. Feng, S. Zhang, X. Zhang, C. Chang, J. Fan, L. Zhang and J. Wang, Sci. Rep., 2015, 5, 14646 Search PubMed .
  26. Y. Guo, D. Liu and J. Robertson, Appl. Phys. Lett., 2015, 106, 173106 Search PubMed .
  27. J. Kang, W. Liu, D. Sarkar, D. Jena and K. Banerjee, Phys. Rev. X, 2014, 4, 031005 Search PubMed .
  28. C. Kim, I. Moon, D. Lee, M. S. Choi, F. Ahmed, S. Nam, Y. Cho, H. J. Shin, S. Park and W. J. Yoo, ACS Nano, 2017, 11, 1588–1596 CAS .
  29. R. T. Tung, Appl. Phys. Rev., 2014, 1, 011304 Search PubMed .
  30. C. Gong, L. Colombo, R. M. Wallace and K. Cho, Nano Lett., 2014, 14, 1714–1720 CrossRef CAS PubMed .
  31. H. Y. Jang, J. H. Nam, J. Yoon, Y. Kim, W. Park and B. Cho, Nanotechnology, 2020, 31, 225205 CAS .
  32. W. Seo, W. Park, H. Y. Seo, S. Oh, O. Kwon, S. H. Jeong, D. H. Kim, M. J. Kim, S. K. Lee, B. H. Lee and B. Cho, Appl. Surf. Sci., 2022, 604, 154485 CAS .
  33. C. H. Ji, K. T. Kim and S. Y. Oh, RSC Adv., 2018, 8, 8302–8309 Search PubMed .
  34. M. S. Alam, M. A. Boby, F. A. Chowdhury, H. Albrithen and M. A. Hossain, RSC Adv., 2019, 9, 18996–19005 CAS .
  35. X. Liu, F. Li, M. Xu, T. Shen, Z. Yang, W. Fan and J. Qi, Langmuir, 2018, 34, 14151–14157 CAS .
  36. J. M. Choi, H. Y. Jang, A. R. Kim, J. D. Kwon, B. Cho, M. H. Park and Y. Kim, Nanoscale, 2021, 13, 672–680 CAS .
  37. L. Wang, J. Jie, Z. Shao, Q. Zhang, X. Zhang, Y. Wang, Z. Sun and S. T. Lee, Adv. Funct. Mater., 2015, 25, 2910–2919 CAS .
  38. S. Qiao, R. Cong, J. Liu, B. Liang, G. Fu, W. Yu, K. Ren, S. Wang and C. Pan, J. Mater. Chem. C, 2018, 6, 3233–3239 CAS .
  39. D. Wu, C. Jia, F. Shi, L. Zeng, P. Lin, L. Dong, Z. Shi, Y. Tian, X. Li and J. Jie, J. Mater. Chem. A, 2020, 8, 3632–3642 Search PubMed .
  40. P. Vabbina, N. Choudhary, A. A. Chowdhury, R. Sinha, M. Karabiyik, S. Das, W. Choi and N. Pala, ACS Appl. Mater. Interfaces, 2015, 7, 15206–15213 Search PubMed .
  41. W. Zheng, R. Lin, Y. Zhu, Z. Zhang, X. Ji and F. Huang, ACS Appl. Mater. Interfaces, 2018, 10, 20696–20702 CrossRef CAS PubMed .

Footnotes

Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tc05484a
These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2025
Click here to see how this site uses Cookies. View our privacy policy here.