Visible Near-Infrared Photodetection Based on Ta2NiSe5/WSe2 van der Waals Heterostructures

The increasing interest in two-dimensional materials with unique crystal structures and novel band characteristics has provided numerous new strategies and paradigms in the field of photodetection. However, as the demand for wide-spectrum detection increases, the size of integrated systems and the limitations of mission modules pose significant challenges to existing devices. In this paper, we present a van der Waals heterostructure photodetector based on Ta2NiSe5/WSe2, leveraging the inherent characteristics of heterostructures. Our results demonstrate that this detector exhibits excellent broad-spectrum detection ability from the visible to the infrared bands at room temperature, achieving an extremely high on/off ratio, without the need for an external bias voltage. Furthermore, compared to a pure material detector, it exhibits a fast response and low dark currents (~3.6 pA), with rise and fall times of 278 μs and 283 μs for the response rate, respectively. Our findings provide a promising method for wide-spectrum detection and enrich the diversity of room-temperature photoelectric detection.


Introduction
The photoelectric industry has been experiencing continuous development over the past few decades and photoelectric devices have become ubiquitous in our daily lives. As the core of photoelectric devices, the photo-detector is a type of application equipment that converts optical signals into electrical signals. It is a key component of equipment that is frequently used in various fields such as medical diagnosis, optical communication, security monitoring, and military applications [1][2][3][4][5]. Conventional photodetectors, including Si, SiC, GaN, and ZnO [6,7], exhibit excellent single-band optical responses due to their band gaps. Silicon-based materials are generally used in visible light detectors, which have a wide range of applications in color imaging for realistic everyday scenes, while wide-bandgap semiconductor materials such as SiC, GaN, and ZnO are more suitable for UV detectors and have a wide range of application scenarios for anti-missile and tracking in the solar blind range of ultraviolet light [8,9]. However, as the application of photodetectors expands, the need to identify complex working environmental information is increasing. In the face of wide-spectrum detection, multiple detectors of different bands are often required to perform together; however, there has been a gradual failure to meet the current needs of the actual photoelectric application field. Therefore, the research on and development of broadband photodetectors that can cover multi-band response is particularly urgent at present [10][11][12].
Over the past decade, two-dimensional (2D) materials have gradually attracted considerable attention due to their atomic thin thickness, bandgap tunability, and easy integration of van der Waals (vdWs) heterostructures [13][14][15][16][17]. In particular, graphene possesses a zero bandgap and high carrier mobility, making it an attractive material for broadband response research. However, its natural spectrum absorption is only 2.3% and its zero bandgap results in relatively low response performance and limitations in photodetection research [18]. While single-layer MoS 2 has demonstrated superior performance in the visible spectrum range, its application in the infrared spectrum of equal width is significantly restricted by its large bandgap [19][20][21][22]. Although black phosphorus (BP) exhibits a tunable narrow-bandgap (0.3-1.5 eV) and high carrier mobility, its poor stability currently impedes its widespread utilization in photoelectric detection research [23][24][25][26]. Recently, terephthalic thiocompound (Ta 2 NiSe 5 ), a layered material with weak van der Waals force interactions, has shown significant potential in wide-band photodetection [27,28]. Unlike other two-dimensional materials, Ta 2 NiSe 5 is a ternary chalcogenide compound that possesses a direct narrow bandgap (E g~0 .33 eV) [29][30][31][32]. However, the Ta 2 NiSe 5 photodetector suffers from the drawbacks of excessive dark current and high power consumption, which restrict its wide application in the field of photodetection [33]. The fabrication of heterogeneous devices may offer a potential solution to these problems associated with Ta 2 NiSe 5 . Heterojunctions based on two-dimensional materials can achieve self-powered, high photo response and a good signal ratio, which is determined by the mechanism of photocurrent generation. Moreover, in some structures, the appearance of heterogeneous interface Schottky barriers generates a built-in electric field that separates the electron-hole pairs generated by light, thereby inhibiting the rapid recombination of photocarriers [34][35][36].
In this work, we fabricated a photodetector by constructing a Ta 2 NiSe 5 /WSe 2 van der Waals heterostructure on a metal electrode. A series of experimental data showed that Ta 2 NiSe 5 /WSe 2 heterojunctions not only reduce the dark current and power consumption of single-material photodetectors by utilizing the Schottky barrier formed but also address the issue of limited detection bands in single-material photodetectors, thereby enabling wide-spectral detection from visible to infrared light. They also show good application potential in the field of visible light imaging. Our experimental results confirm the feasibility of Ta 2 NiSe 5 /WSe 2 heterojunction photodetectors for wide-spectrum detection, low power consumption, and fast response.

Device Design and Fabrication
We obtained bulk Ta 2 NiSe 5 and WSe 2 single crystals from Shanghai Onway Technology Co., Ltd. (Shanghai, China). Few-layer Ta 2 NiSe 5 and WSe 2 were exfoliated using scotch tape and then assembled using the all-dry transfer method to form van der Waals heterojunctions. As illustrated in Figure 1, the Ta 2 NiSe 5 /WSe 2 heterostructure was constructed using the all-dry transfer method, which shows the process of heterojunction fabrication. Initially, we mechanically peeled off Ta 2 NiSe 5 sheets of appropriate size using tape, followed by placing polydimethylsiloxane (PDMS) as an intermediate carrier onto the tape with the materials during the transfer process. Subsequently, we oriented the other side of the PDMS towards the substrate, with the material side facing down, to transfer the Ta 2 NiSe 5 sheet onto the substrate. To transfer the WSe 2 sheets, a similar technique was employed, but with the added requirement of operating and observing the specific position under a microscope in real-time to ensure the successful stacking of the WSe 2 sheet onto the Ta 2 NiSe 5 sheet to obtain a complete Ta 2 NiSe 5 /WSe 2 heterojunction.
In this study, we investigated the photoelectric performance of the Ta 2 NiSe 5 /WSe 2 heterojunction device under various environmental conditions in real-time, as illustrated in Figure 2a. Initially, Cr/Au (10 nm/50 nm) metal electrodes were deposited onto a lowresistance silicon substrate coated with SiO 2 (300 nm) using UV lithography technology. Next, the Ta 2 NiSe 5 sheet, prepared by mechanical exfoliation, was placed on the substrate via the standard PDMS pressing method, with the same preparation transfer process for the WSe 2 sheet, to create a Ta 2 NiSe 5 /WSe 2 heterostructure. The experimental materials were characterized using optical microscopy (Olympus BX51M), scanning electron microscopy with X-ray energy dispersive spectroscopy (SEM-EDS) (Zeiss/GEMINISEM 360), and microscopic laser confocal Raman spectroscopy (Dugout/LABRAM ODYSSEY). Figure 2b presents the optical image of the Ta 2 NiSe 5 /WSe 2 heterostructure-based photodetector, with A-D denoting the four electrode directions. Different electrodes were connected to apply voltage for the performance testing of various devices. The scanning electron microscopy (SEM) image and corresponding energy dispersive spectra (EDS) mappings are shown in Figure 2c. The Raman spectra of the Ta 2 NiSe 5 , WSe 2 , and Ta 2 NiSe 5 /WSe 2 heterojunction are represented in Figure 2d. For pure Ta 2 NiSe 5 [37], three characteristic peaks located at 97.7, 121.8, and 288.8 cm −1 , attributed to the A g vibration mode, can be clearly observed and are denoted by the orange line. For pure WSe 2 [38], the Raman spectrum exhibited two distinct diffraction peaks at 247.3 and 257.6 cm −1 , corresponding to the E 2g and A 1g lattice vibrational modes in WSe 2 , respectively (red line). It has been reported that monolayers of WSe 2 do not produce the characteristic peak at 308.3 cm −1 under 532 nm laser excitation [39], and the presence of this peak suggests that the exfoliated WSe 2 flakes were less laminated. All the vibration modes of the Ta 2 NiSe 5 /WSe 2 heterojunction, regardless of whether Ta 2 NiSe 5 or WSe 2 was used (blue line), were clearly observed in the overlapping region, indicating the effective preparation of the Ta 2 NiSe 5 /WSe 2 heterojunction. According to previous reports, the conductance band (E c ) and valence band (E v ) of Ta 2 NiSe 5 (WSe 2 ) are located at −4.6 eV (−4.2 eV) and −5.28 eV (−4.93 eV) [40][41][42][43][44], respectively, forming the type I band alignment of the heterostructure. Figure 2e shows the band structure before and after contact. When the two materials came into contact to form the heterojunction, electrons flowed from Ta 2 NiSe 5 to the latter due to a higher Fermi energy level than that for Ta 2 NiSe 5 , resulting in a depletion layer near Ta 2 NiSe 5 and an accumulation layer on the other side, gradually forming an internal electric field [45,46]. The direction of the electric field went from Ta 2 NiSe 5 to WSe 2 . Applying a forward bias reduced the barrier, facilitating the flow of electrons from Ta 2 NiSe 5 to WSe 2 , resulting in an increase in current. Conversely, under reverse bias, the barrier increased, impeding the movement of electrons and causing a decrease in current, as illustrated in Figure 2f.  In this study, we investigated the photoelectric performance of the Ta2NiSe5/WSe2 heterojunction device under various environmental conditions in real-time, as illustrated in Figure 2a. Initially, Cr/Au (10 nm/50 nm) metal electrodes were deposited onto a lowresistance silicon substrate coated with SiO2 (300 nm) using UV lithography technology. Next, the Ta2NiSe5 sheet, prepared by mechanical exfoliation, was placed on the substrate via the standard PDMS pressing method, with the same preparation transfer process for the WSe2 sheet, to create a Ta2NiSe5/WSe2 heterostructure. The experimental materials were characterized using optical microscopy (Olympus BX51M), scanning electron mi-

Results and Discussion
Before discussing the photoelectric performance of the Ta2NiSe5/WSe2 heterojunction detectors, certain photoelectric properties of the individual detector unit are examined to emphasize the advantages of heterojunction detectors. The external circuit was connected to electrodes B and D to conduct the photoelectric performance test of the few-layered WSe2 detector. Figure 3a displays the current-voltage (I-V) curve of the few-layered WSe2 photodetector in the dark, which exhibited Schottky characteristics at room temperature and provided a built-in electric field in the device (Mstarter 200). The optoelectronic properties of the few-layered WSe2 photodetector were evaluated by varying the incident power (P) of 638 nm visible radiation, as shown in Figure 3b. Additionally, Figure 3c depicts the time-resolved photo responses of the various incident powers collected by highspeed oscilloscopes at room temperature without any bias voltage, where the photocurrent Iph and the current density J are obtained using Equations (1) and (2):

Iph Ids Idark
(1) where Ids represents the source-drain current under illumination, Idark represents the source-drain current in the dark, and S represents the effective area of the device. To elucidate the photocurrent generation mechanism of the few-layered WSe2 photodetector, scanning photocurrent microscopy mapping was performed without bias voltage under 638 nm laser irradiation at room temperature. As observed in Figure 3d, the area of the device was mainly distributed between the material and the metal electrode, and there was clear carrier movement within the device, displaying excellent photo response per-

Results and Discussion
Before discussing the photoelectric performance of the Ta 2 NiSe 5 /WSe 2 heterojunction detectors, certain photoelectric properties of the individual detector unit are examined to emphasize the advantages of heterojunction detectors. The external circuit was connected to electrodes B and D to conduct the photoelectric performance test of the few-layered WSe 2 detector. Figure 3a displays the current-voltage (I-V) curve of the few-layered WSe 2 photodetector in the dark, which exhibited Schottky characteristics at room temperature and provided a built-in electric field in the device (Mstarter 200). The optoelectronic properties of the few-layered WSe 2 photodetector were evaluated by varying the incident power (P) of 638 nm visible radiation, as shown in Figure 3b. Additionally, Figure 3c depicts the time-resolved photo responses of the various incident powers collected by high-speed oscilloscopes at room temperature without any bias voltage, where the photocurrent I ph and the current density J are obtained using Equations (1) and (2): where I ds represents the source-drain current under illumination, I dark represents the source-drain current in the dark, and S represents the effective area of the device. To elucidate the photocurrent generation mechanism of the few-layered WSe 2 photodetector, scanning photocurrent microscopy mapping was performed without bias voltage under 638 nm laser irradiation at room temperature. As observed in Figure 3d, the area of the device was mainly distributed between the material and the metal electrode, and there was clear carrier movement within the device, displaying excellent photo response performance. The intrinsic Schottky junction facilitated the rapid separation and collection of electronhole pairs, producing a photovoltaic current at visible radiation. However, due to the large bandgap limitation of WSe 2 , the detector failed to detect the photo response in the infrared band. To unveil the photo response of the few-layered Ta2NiSe5 device in the infrared band, we characterized its optical image by switching on electrodes A and C, as shown in the inset of Figure 4a. In contrast to the WSe2 device, the current-voltage (I-V) curve of the Ta2NiSe5 device exhibited excellent linearity in the dark, as depicted in Figure 4a. Here, the dark current was measured to be 126 uA at a bias voltage of −0.1 V. Figure 4b presents the linear current-voltage output curve of the Ta2NiSe5 device under illumination of 1550 nm. However, due to a large dark current, the photocurrent was not very apparent, even after applying the bias voltage. To further assess the photo response at 1550 nm, we measured the time-resolution photo response under periodic illumination with a fixed bias (Vds = 0.1 V) by switching to different optical powers (see Figure 4c, which shows a significant power dependency). In the near-infrared band, since the energy of the incident photon exceeded the bandgap of Ta2NiSe5 (hv > Eg), the photoinduced electron-hole pairs were separated by an external electric field and collected by the electrodes. The scanning photocurrent microscopy mapping revealed that the photocurrent region was far from the contact region of the electrode and Ta2NiSe5 material. It is noteworthy that, although the Ta2NiSe5 material is capable of realizing near-infrared photodetection, it is uncompetitive for future applications due to the large dark current caused by bias in photoconductivity mode. To unveil the photo response of the few-layered Ta 2 NiSe 5 device in the infrared band, we characterized its optical image by switching on electrodes A and C, as shown in the inset of Figure 4a. In contrast to the WSe 2 device, the current-voltage (I-V) curve of the Ta 2 NiSe 5 device exhibited excellent linearity in the dark, as depicted in Figure 4a. Here, the dark current was measured to be 126 uA at a bias voltage of −0.1 V. Figure 4b presents the linear current-voltage output curve of the Ta 2 NiSe 5 device under illumination of 1550 nm. However, due to a large dark current, the photocurrent was not very apparent, even after applying the bias voltage. To further assess the photo response at 1550 nm, we measured the time-resolution photo response under periodic illumination with a fixed bias (V ds = 0.1 V) by switching to different optical powers (see Figure 4c, which shows a significant power dependency). In the near-infrared band, since the energy of the incident photon exceeded the bandgap of Ta 2 NiSe 5 (hv > E g ), the photoinduced electron-hole pairs were separated by an external electric field and collected by the electrodes. The scanning photocurrent microscopy mapping revealed that the photocurrent region was far from the contact region of the electrode and Ta 2 NiSe 5 material. It is noteworthy that, although the Ta 2 NiSe 5 material is capable of realizing near-infrared photodetection, it is uncompetitive for future applications due to the large dark current caused by bias in photoconductivity mode. The experimental results of the detectors based on the two individual materials mentioned above led to the proposal of a novel Ta2NiSe5/WSe2 vdWs heterogeneous structure that achieves low dark current levels from visible to infrared detection. During the testing process, we varied the wiring configuration by connecting the source and drain electrodes at terminals C and D. The current-voltage (I-V) output curve of the vdWs heterogeneous device at room temperature with and without laser illuminations of 638 nm and 1550 nm, respectively, is presented in Figure 5a. By increasing the light incident power of 638 nm and 1550 nm, the Ta2NiSe5/WSe2 heterojunction device exhibited excellent photo response performance, as shown in Figure 5b,c, where a heterojunction is clearly visible. The phenomenon of negative short-circuit current (~4.3 pA) and positive open-circuit voltage (~0.1 V), as revealed in Figure 5a-c, is indicative of a photovoltaic-type mode. At a bias voltage of −1 V, the dark current of the proposed device was measured to be 3.6 pA, which is eight orders of magnitude lower than that of the Ta2NiSe5 device. The presence of the heterojunction not only reduced the dark current and improved the signal-to-noise ratio but also enhanced the light on/off ratio of the photocurrent. Figure 5d,e demonstrate the time-resolved photo response without any external electric field and with increasing incident power at 638 nm and 1550 nm, respectively, showing a fast photo-switching process. To further confirm the photo response mechanism, the location of the scanning photocurrent mapping focused on the junction area (see Figure 5f), which confirmed the origin of the photocurrent at the heterojunction. In the heterojunction region, electrons in Ta2NiSe5 were excited to the conduction band by IR radiation, generating electron-hole pairs, which were separated by the built-in electric field and flowed towards WSe2, being ultimately collected by electrodes. Conversely, holes flowed in the opposite direction, forming a circuit and producing photocurrent. The experimental results of the detectors based on the two individual materials mentioned above led to the proposal of a novel Ta 2 NiSe 5 /WSe 2 vdWs heterogeneous structure that achieves low dark current levels from visible to infrared detection. During the testing process, we varied the wiring configuration by connecting the source and drain electrodes at terminals C and D. The current-voltage (I-V) output curve of the vdWs heterogeneous device at room temperature with and without laser illuminations of 638 nm and 1550 nm, respectively, is presented in Figure 5a. By increasing the light incident power of 638 nm and 1550 nm, the Ta 2 NiSe 5 /WSe 2 heterojunction device exhibited excellent photo response performance, as shown in Figure 5b,c, where a heterojunction is clearly visible. The phenomenon of negative short-circuit current (~4.3 pA) and positive open-circuit voltage (~0.1 V), as revealed in Figure 5a-c, is indicative of a photovoltaic-type mode. At a bias voltage of −1 V, the dark current of the proposed device was measured to be 3.6 pA, which is eight orders of magnitude lower than that of the Ta 2 NiSe 5 device. The presence of the heterojunction not only reduced the dark current and improved the signal-to-noise ratio but also enhanced the light on/off ratio of the photocurrent. Figure 5d,e demonstrate the timeresolved photo response without any external electric field and with increasing incident power at 638 nm and 1550 nm, respectively, showing a fast photo-switching process. To further confirm the photo response mechanism, the location of the scanning photocurrent mapping focused on the junction area (see Figure 5f), which confirmed the origin of the photocurrent at the heterojunction. In the heterojunction region, electrons in Ta 2 NiSe 5 were excited to the conduction band by IR radiation, generating electron-hole pairs, which were separated by the built-in electric field and flowed towards WSe 2 , being ultimately collected by electrodes. Conversely, holes flowed in the opposite direction, forming a circuit and producing photocurrent. As illustrated in Figure 6a,b, photocurrent was observed in both the visible and infrared bands of the Ta2NiSe5/WSe2 heterojunction device, without applying any bias. Laser irradiation of the Ta2NiSe5/WSe2 heterojunction device generated photogenerated electrons, which then moved directionally to form photocurrent under the influence of the built-in electric field within the device. To evaluate the crucial parameters of the Ta2NiSe5/WSe2 heterojunction photodetector, responsivity R, characterized as the value of the photocurrent generated per unit of optical power, is defined as follows [47]: Specific detectivity D* is another important parameter of the photodetector and represents the sensitivity of the photodetector; it is defined as follows [48]: where Iph is the photocurrent, Pin is incident light power, q is the elementary charge, and S is the illumination active area on the device. For the visible/IR band, the extracted powerdependent responsivity and detectivity are plotted in Figure 6c,d. The responsivity of the heterojunction device was ~0.04 mA/W (P = 11.12 µW) at 638 nm and ~0.82 nA/W (P = 231.8 µW) at 1550 nm. The detectivity of the heterojunction device was approximately 3.6 × 10 5 Jones (P = 11.12 µW) at 638 nm and approximately 90 Jones (P = 231.8 µW) at 1550 nm. Figure 6e presents the time-resolved optical response at 638 nm in the absence of any external electric field, revealing a rapid optical switching process. During this period, the photocurrent response of the device remained relatively stable, indicating good stability of the heterojunction device. Response time, another key parameter, which is routinely defined as the time required by the photo response to rise from 10% to 90% or fall from 90% to 10% in a single impulse, is retrieved directly from the high-speed sampling oscilloscope (Tektronix/Afg 3100 series) [49]. The rise and fall times of the heterojunction device were 278 µs and 283 µs, respectively, as shown in Figure 6f. We also compared the performance test results with other van der Waals heterojunction photodetectors, as shown in Table 1 [50][51][52][53][54]. The As illustrated in Figure 6a,b, photocurrent was observed in both the visible and infrared bands of the Ta 2 NiSe 5 /WSe 2 heterojunction device, without applying any bias. Laser irradiation of the Ta 2 NiSe 5 /WSe 2 heterojunction device generated photogenerated electrons, which then moved directionally to form photocurrent under the influence of the built-in electric field within the device. To evaluate the crucial parameters of the Ta 2 NiSe 5 /WSe 2 heterojunction photodetector, responsivity R, characterized as the value of the photocurrent generated per unit of optical power, is defined as follows [47]: Specific detectivity D * is another important parameter of the photodetector and represents the sensitivity of the photodetector; it is defined as follows [48]: where I ph is the photocurrent, Pin is incident light power, q is the elementary charge, and S is the illumination active area on the device. For the visible/IR band, the extracted power-dependent responsivity and detectivity are plotted in Figure 6c,d. The responsivity of the heterojunction device was~0.04 mA/W (P = 11.12 µW) at 638 nm and~0.82 nA/W (P = 231.8 µW) at 1550 nm. The detectivity of the heterojunction device was approximately 3.6 × 10 5 Jones (P = 11.12 µW) at 638 nm and approximately 90 Jones (P = 231.8 µW) at 1550 nm. Figure 6e presents the time-resolved optical response at 638 nm in the absence of any external electric field, revealing a rapid optical switching process. During this period, the photocurrent response of the device remained relatively stable, indicating good stability of the heterojunction device. Response time, another key parameter, which is routinely defined as the time required by the photo response to rise from 10% to 90% or fall from 90% to 10% in a single impulse, is retrieved directly from the high-speed sampling oscilloscope (Tektronix/Afg 3100 series) [49]. The rise and fall times of the heterojunction device were 278 µs and 283 µs, respectively, as shown in Figure 6f. We also compared the performance test results with other van der Waals heterojunction photodetectors, as shown in Table 1 [50][51][52][53][54]. The comparison revealed that the Ta 2 NiSe 5 /WSe 2 heterojunction detectors prepared in our experiment exhibit superior performance in terms of dark current and response times, surpassing previously reported photodetectors.
x FOR PEER REVIEW 8 of 12 comparison revealed that the Ta2NiSe5/WSe2 heterojunction detectors prepared in our experiment exhibit superior performance in terms of dark current and response times, surpassing previously reported photodetectors.   [54] In addition, in order to explore the potential application of the Ta2NiSe5/WSe2 heterojunction photodetectors in the imaging field, we carried out imaging experiments on the detectors. Imaging was conducted using the visible spectrum (638 nm) on the Ta2NiSe5/WSe2 heterojunction photoelectric device, as illustrated in Figure 7a (equipment type: THORLABS CM401). Figure 7b depicts a schematic diagram of the overall device used in the imaging system. Prior to testing, two pieces of white paper with a certain thickness were prepared, upon which letter graphics (ZJUT, HIAS) were cut out for presentation. During the test, visible light emitted directly at the paper was reflected, while visible light shining on the letter spaces penetrated.   [54] In addition, in order to explore the potential application of the Ta 2 NiSe 5 /WSe 2 heterojunction photodetectors in the imaging field, we carried out imaging experiments on the detectors. Imaging was conducted using the visible spectrum (638 nm) on the Ta 2 NiSe 5 /WSe 2 heterojunction photoelectric device, as illustrated in Figure 7a (equipment type: THORLABS CM401). Figure 7b depicts a schematic diagram of the overall device used in the imaging system. Prior to testing, two pieces of white paper with a certain thickness were prepared, upon which letter graphics (ZJUT, HIAS) were cut out for presentation. During the test, visible light emitted directly at the paper was reflected, while visible light shining on the letter spaces penetrated.

Conclusions
In conclusion, we prepared a van der Waals heterostructure device based on Ta2NiSe5/WSe2 material by stacking thin sheets of Ta2NiSe5 and WSe2 on a SiO2 substrate. The resulting heterojunction photodetector exhibited wide-spectrum detection from visible to infrared light, with a response rate of 0.04 mA/W at 638 nm (0.82 nA/W at 1550 nm), utilizing the photoelectric response characteristics of the Ta2NiSe5/WSe2 heterojunction. Furthermore, the device demonstrated extremely low dark current (~3.6 pA) due to the built-in electric field within the van der Waals heterojunction structure, effectively inhibiting dark current generation and promoting the separation of photogenerated electronhole pairs, resulting in reduced power loss. In addition, the rapid transport and separation of photogenerated carriers under an electric field enabled the Ta2NiSe5/WSe2 heterojunction device to respond quickly, with an up time of approximately 278 µs and a down time of approximately 283 µs. Our results suggest that the van der Waals heterojunction photodetector based on Ta2NiSe5/WSe2 has immense potential for achieving high detection performance, such as broadening the detection band and improving response times.

Conclusions
In conclusion, we prepared a van der Waals heterostructure device based on Ta 2 NiSe 5 / WSe 2 material by stacking thin sheets of Ta 2 NiSe 5 and WSe 2 on a SiO 2 substrate. The resulting heterojunction photodetector exhibited wide-spectrum detection from visible to infrared light, with a response rate of 0.04 mA/W at 638 nm (0.82 nA/W at 1550 nm), utilizing the photoelectric response characteristics of the Ta 2 NiSe 5 /WSe 2 heterojunction. Furthermore, the device demonstrated extremely low dark current (~3.6 pA) due to the built-in electric field within the van der Waals heterojunction structure, effectively inhibiting dark current generation and promoting the separation of photogenerated electron-hole pairs, resulting in reduced power loss. In addition, the rapid transport and separation of photogenerated carriers under an electric field enabled the Ta 2 NiSe 5 /WSe 2 heterojunction device to respond quickly, with an up time of approximately 278 µs and a down time of approximately 283 µs. Our results suggest that the van der Waals heterojunction photodetector based on Ta 2 NiSe 5 /WSe 2 has immense potential for achieving high detection performance, such as broadening the detection band and improving response times.