A high-performance broadband phototransistor array of a PdSe₂/SOI Schottky junction†

Yexin Chen‡ ^a, Qinghai Zhu‡ ^a, Jiabao Sun ^b, Yijun Sun ^b, Nobutaka Hanagata ^c and Mingsheng Xu *^a
aCollege of Integrated Circuits, State Key Laboratory of Silicon and Advanced Semiconductor Materials, Zhejiang University, Hangzhou 310027, China. E-mail: msxu@zju.edu.cn
^bCollege of Information Science & Electronic Engineering, Zhejiang University, 38 Zheda Road, Hangzhou 310027, China
^cResearch Center for Functional Materials and Nanotechnology Innovation Station, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan

First published on 20th February 2024

---

---

Introduction

Photodetectors (PDs), as vital optoelectronic devices that can convert light signals into electrical signals, are one of the key components of modern optoelectronic systems.¹ They play a role in important fields such as optical communications, intelligent imaging, environmental monitoring, and industrial automation.² So far, high-performance PDs have been achieved by using traditional semiconductor materials including Si, GaN, and InGaAs.^3,4 Among them, Si is most widely used and its preparative technology is mature, especially in imaging sensor CCD and complementary metal oxide semiconductor (CMOS) devices. However, due to the limitation of the band gap, the detection wavelength of traditional Si-based photodetectors cannot exceed 1100 nm, making it impossible to achieve wide spectrum detection.⁵ Although other narrow bandgap semiconductor materials (such as Ge and HgCdTe) have been used to fabricate broadband photodetectors, the complex preparation process, low operating temperature, and high price prevent these photodetectors from being widely used.⁶ Therefore, it is essential to introduce new materials to produce PDs with a low cost, wide detection range, high sensitivity and low power consumption.

Until now, two-dimensional (2D) transition metal dichalcogenides (TMDs) have become a hot research topic due to their exotic optoelectronic properties.^7,8 PDs based on 2D TMDs can overcome the shortcomings of traditional photodetectors. As an emerging group-10 TMD, palladium diselenide (PdSe₂) has high carrier mobility, long-term material stability, and strong interlayer coupling.⁹ Meanwhile, PdSe₂ shows a thickness-dependent bandgap ranging from about 0.0 eV to 1.3 eV.¹⁰ This special property of tunable bandgap makes it possible for PDs based on PdSe₂ to extend the detection range from visible light to near-infrared and mid-infrared, achieving a broadband of photon detection.^11,12 Based on these characteristics, PdSe₂ is expected to become a candidate material for high-sensitivity broadband PDs. Compared with traditional single-material photodetectors, PDs based on van der Waals heterostructures (vdWHs) have been extensively explored due to their strong interlayer coupling, designable energy band structure, and excellent light absorption capabilities.¹³ For example, Zeng et al. used a simple selenization method to controllably synthesize a uniform and wafer-scale 2D PdSe₂ film on a Si substrate, forming a PdSe₂/n-Si vertical structure heterojunction photodetector.¹⁴ Under 780 nm illumination, the device has a high responsivity (300 mA W⁻¹) and high specific detectivity (≈ 10¹³ Jones) at zero bias. Liang et al. investigated a highly sensitive and self-driven photodetector based on PdSe₂/pyramid Si heterojunction arrays, which are fabricated by pre-depositing Pd film on a Si substrate.¹⁵ The responsivity and specific detectivity of the device can reach 456 mA W⁻¹ and 9.97 × 10¹³ Jones under 980 nm illumination at zero bias, respectively. The detection wavelength of the photodetector arrays can reach the near-infrared light region and the arrays exhibit excellent uniformity and repeatability in imaging applications. These devices only considered the influence of direct current dark noise in the calculation of specific detectivity and underestimated the noise current, resulting in an overestimated value of the specific detectivity.¹⁶ In addition, these devices suffered from weak photoresponsivity, which is still difficult to be effectively applied in high sensitivity detection.

Recently, we have reported the PD's performance of a semiconductive PtSe₂ film on a silicon-on-insulator (SOI).¹⁷ PtSe₂ is a 2D TMD with a highly symmetrical hexagonal crystal structure, where each Pt atom is closely surrounded by six Se atoms to form an octahedral coordination structure and atoms in the same layer are connected by covalent bonds. Different from the structure of PtSe₂, 2D PdSe₂ has a unique pentagonal crystal structure and the Pd atoms exhibit unusual planar four-coordination, which can still have a relatively low symmetry in the regular corrugation pattern between layers. Furthermore, it is known that the barrier region of a Schottky junction can directly absorb photons, avoiding the need for photogenerated carriers to diffuse to the space charge region and then be separated, thereby effectively improving the response speed of the device. Compared to other TMD materials, the larger work function of semi-metallic PdSe₂ forms a higher Schottky barrier by coming into contact with Si, which can effectively suppress the formation of dark current. In addition, the highly inclined Dirac cone band structure of semi-metallic PdSe₂ could promote the rapid separation of photogenerated carriers in the PdSe₂/Si Schottky junction.¹⁸ Previously, we have found that the thickness of the PdSe₂ film in PdSe₂/Si heterostructure PDs has an influence on the performance of PDs.¹⁹ In order to improve the photoelectric performance of the device, we chose to use a SOI substrate and semi-metallic PdSe₂ to form a Schottky junction phototransistor. By regulating the gate voltage, the responsivity of the device can be effectively improved.

Here, we report a self-driven broadband PdSe₂/Si heterojunction phototransistor and its array fabricated by patterning the growth of a semi-metallic PdSe₂ film directly on a SOI substrate, and the photoresponse performance of the PdSe₂/SOI phototransistor can be modulated by a gate bias. The PdSe₂/SOI phototransistor exhibits good photoresponse performance tested in the wavelength range from ultraviolet (375 nm) to near-infrared (1550 nm), which can also be modulated by a gate bias. Under 808 nm illumination, the responsivity and rising time/falling time of this device can reach 1.15 A W⁻¹ and 27.1/40.3 μs, respectively, exceeding most reported Si-based infrared detectors. We verify that the 4 × 4 PdSe₂/Si phototransistor array has outstanding uniformity, stability, and image sensing capabilities under incident light of different wavelengths. Our work demonstrates that 2D PdSe₂ is promising as a potential candidate material for broadband high-performance photodetection applications.

Experimental

Materials synthesis and device fabrication

PdSe₂/Si heterojunction phototransistors were prepared using a silicon-on-insulator (SOI) wafer with a 1500 nm top silicon (resistivity: 1–10 Ω cm) and 1000 nm silica insulating layer. Firstly, a series of silicon window areas with a size of 300 × 300 μm² were reserved on the top silicon by ultraviolet photolithography technology and a SiO₂ material with a thickness of 150 nm was sputtered outside the silicon window to form an insulating layer. Then, the growth region of the PdSe₂ material on the top layer was defined by the photolithography process, thus obtaining the PdSe₂/Si heterojunction with an area of 300 × 50 μm². The Pd layer was deposited in the pre-defined growth area by magnetron sputtering. Next, 600 mg selenium powder (99.99%) and the pre-treated SOI wafer were placed in the upstream and downstream areas of a tube furnace. A mechanical pump was used to pump air and the temperatures of the two heating zones were set to 230 °C and 300 °C, respectively. In the process of thermal-assisted selenization, a 50/10 sccm Ar/H₂ mixed gas was introduced into the quartz tube as a carrier gas. After reacting for 1 hour, the deposited Pd metal has been completely selenized into the PdSe₂ material. The temperature of the tube furnace dropped to room temperature under the protection of Ar gas. Then, the contact area of the electrode was defined on the PdSe₂ layer by photolithography technology and the Au metal was deposited as the drain electrode using magnetron sputtering equipment. Similarly, the Al metal was deposited on part of the Si window to form a good ohmic contact. Finally, we successfully obtained a 4 × 4 PdSe₂/Si phototransistor array by using the heavily doped n-Si material at the bottom of the SOI wafer as the back gate electrode.

Characterization

The molecular structure of the PdSe₂ material was characterized using Raman equipment (Renishaw) with a 532 nm laser and its lattice structure was determined by X-ray diffraction (XRD, Bruker) with Cu Kα X-rays at a working voltage of 40 kV. An atomic force microscope (AFM, Bruker Dimension Edge) was used to measure the thickness of PdSe₂ films and ultraviolet electron spectroscopy (UPS) was performed to determine the energy band structure of PdSe₂ materials. The absorption spectra of PdSe₂, Si and PdSe₂/Si materials within the range of ultraviolet to near-infrared light were recorded using a UV-vis-NIR spectrometer (Hitachi U-4100). As for the measurement of the photodetection performance of our PdSe₂/Si phototransistor, lasers with wavelengths of 375, 532, 808, 940 and 1550 nm were used as light sources, a signal generator (RIGOL DG5100) was used to control the generation of laser signals, and a semiconductor analyzer (PDA Co., Ltd) was utilized to collect the output signals of our device.

Results and discussion

Fig. 1 shows the manufacturing process of the PdSe₂/Si phototransistor. In brief, a 300 × 300 μm² Si window was patterned on a SOI substrate by UV lithography. 150 nm SiO₂ was then deposited on the top silicon of the SOI as an insulating layer to isolate devices in the array. The thicknesses of top silicon and SiO₂ in the SOI are 1.5 μm and 1.0 μm, respectively. The Pd film was positioned and grown on part of the Si window by magnetron sputtering. The Pd film and Se powder were fully reacted by chemical vapor deposition (CVD) to form a PdSe₂ film, and finally the 300 × 50 μm² PdSe₂/Si heterojunction was formed. Since the thickness of PdSe₂ was determined by the Pd layer, the thickness of the PdSe₂ film can be changed by appropriately adjusting the thickness of the Pd film in this process.¹⁹ Finally, Au and Al were used as source and drain electrodes to achieve ohmic contact in the PdSe₂ and Si regions, respectively. The heavily doped n-type Si on the SOI bottom layer serves as the back gate electrode of the PdSe₂/Si phototransistor. Fig. S1† displays the optical microscopy image of the PdSe₂/Si phototransistor.

As shown in Fig. 2a, the Raman spectrum of the fabricated film has four obvious peaks located at about 145, 207, 223, and 257 cm⁻¹, corresponding to the A¹_g, A²_g, B_1g, and A³_g movement modes of the PdSe₂ lattice, respectively.²⁰ The X-ray diffraction (XRD) pattern shown in Fig. 2b confirms the physical phase of PdSe₂ crystals. These results are consistent with previous work,²¹ indicating that we have successfully synthesized the PdSe₂ material. Fig. 2c shows the UV-Vis-NIR absorption spectrum of the PdSe₂ film, Si and the PdSe₂/Si heterojunction. Obviously, the intrinsic absorption range of Si is mainly in the 350 nm–1100 nm band. When the wavelength continues extending to the infrared region, the absorption effect of Si is extremely weak. However, the light absorption range of the PdSe₂/Si heterojunction can cover from 350 nm to 2500 nm and the light absorption capacity is significantly enhanced as compared to those of PdSe₂ and Si alone materials. This means that the PdSe₂/Si heterojunction is suitable for broadband photodetectors. The related Tauc diagram in Fig. 2d indicates that the band gap of the PdSe₂ film is about 0.0 eV. In addition, the synthesized PdSe₂ film was further characterized by using atomic force microscopy (AFM) and the corresponding height profile (Fig. 2e and f), revealing that the PdSe₂ film has a multilayered structure with a thickness of approximately 37 nm.

In Fig. 3a, the dark-state output characteristics of the PdSe₂/Si phototransistor were measured by changing the gate voltage (V_G) from −4 V to 4 V. In a larger bias range of V_DS = ±5 V, we observed the typical nonlinearity and asymmetry of output curves with a strong rectification ratio. Fig. 3b displays the transfer characteristics of the phototransistor under different biases (V_DS) in the dark. As V_DS sweeps from −3.0 V to 3.0 V, the threshold voltage of the PdSe₂/SOI phototransistor changes with the V_DS from −3.4 V to 1.7 V. Since the photodetection performance was measured at V_DS = 0 V, we emphasize in the inset of Fig. 3b that the device has the V_th value of −1.4 V. It can be found that the output current of the PdSe₂/Si phototransistor will increase sharply when V_G is less than −1.4 V. To further evaluate the effect of V_G on the phototransistor's photoresponse, as shown in Fig. 3c, the photoresponses of the device at different V_G values were investigated under 808 nm illumination (V_DS = 0 V). Apparently, the device can be easily switched between on and off states. When a positive gate voltage is applied, a high photocurrent is recorded, while the dark current does not change significantly as compared to V_G = 0 V. When a negative gate voltage is applied, the photocurrent also increases, but the enhancement of dark current is non-negligible. Normally, PDs with a large dark state current are not conducive to the detection of weak light signals. Thus, in this work, we applied a V_G of 5 V to the phototransistor. These phenomena indicate that the gate voltage can effectively adjust the photocurrent of the device, thereby improving the device performance.^17,22

Elucidating the generation mechanism of photocurrent is essential for our subsequent study of the photoelectric properties of the PdSe₂/Si phototransistor. According to previous reports, we know that the valence band position, Fermi level, and conduction band position of n-type-doped silicon (E_g = 1.12 eV) with a resistivity of 1–10 Ω·cm are about 5.17 eV, 4.25 eV, and 4.05 eV, respectively.²³ According to the ultraviolet photoelectron spectroscopy (UPS) measurement in Fig. S2,† it can be obtained that the Fermi level position of the PdSe₂ film (E_g = 0.0 eV) is approximately 5.24 eV. The working mechanism of the PdSe₂/Si phototransistor can be explained by the energy band diagram shown in Fig. 3d and e. Due to the different work functions of two materials and the semi-metallic properties of the PdSe₂ film, when PdSe₂ and n-Si form a Schottky junction under dark conditions, electrons diffuse from the n-Si to PdSe₂, leaving immovable positive charges in the depletion region of n-Si, and the energy band bends at the interface to reach an equilibrium state. When the thermal equilibrium state is reached, the Fermi level positions of PdSe₂ and n-Si are consistent and a strong built-in electric field is formed from the n-Si to PdSe₂. As shown in Fig. 3d, when light irradiates the PdSe₂/Si Schottky junction, the built-in electric field can quickly and effectively separate the electron–hole pairs and flow to the opposite sides. The photogenerated electrons are directionally transported to the n-Si and the photogenerated holes are directionally transported to the PdSe₂. Eventually, the photogenerated carriers are collected by the electrodes and form the photocurrent in the device. This phenomenon of photogenerated carrier transport does not require the application of any voltage to the PdSe₂/Si phototransistor, which means that the device has self-driven feature. In addition, the regulation of the PdSe₂/Si phototransistor by the gate voltage can be explained by the energy band model in Fig. 3e. When the gate voltage is applied to the phototransistor, the Fermi level positions on both sides of the phototransistor, PdSe₂ and Si, move due to changes in charge carriers under the electric field effect.^24,25 However, due to the charge shielding effect of Si on the PdSe₂ film, the back gate mainly affects the Fermi level of Si and the impact on the Fermi level of PdSe₂ is almost negligible.²⁶ Since the bias V_DS is not applied, the Fermi levels of PdSe₂ and Si are still arranged at the same energy level. Fig. 3e shows the energy band arrangement of the PdSe₂/Si phototransistor when a positive gate voltage is applied (V_G > 0 V and V_DS = 0 V). The Fermi level of Si shifts to the conduction band under the gate voltage, thereby increasing the built-in electric field in the heterojunction. The built-in electric field enhancement promotes the separation of photogenerated carriers and increases the photocurrent of the device.²⁷ In evidence, by changing the gate voltage, the photoresponse of the phototransistor can be effectively improved and better photodetection performance can be obtained.

We conducted a systematic study on the photoelectric properties of the PdSe₂/Si phototransistor. First, we investigated the effects of different optical power densities on device performance. Fig. 4a displays the output characteristic curves under 808 nm illumination with the optical power density transforming from 0.1 mW cm⁻² to 16 mW cm⁻². The phototransistor has excellent photovoltaic effect under illumination. Fig. 4b exhibits the effect of different optical power densities on the time-dependent photocurrent at a fixed wavelength (λ = 808 nm) and zero bias (V_G = 0 V and V_DS = 0 V). Due to higher power density laser irradiation, more photogenerated electron–hole pairs are excited at the heterojunction interface of the PdSe₂/Si phototransistor and separated by the built-in electric field. Therefore, as the optical power density increases, the photocurrent of the device increases, which is consistent with the experimental results. In order to verify the application of the PdSe₂/Si phototransistor in the field of near-infrared light detection, we measured the time-dependent photocurrents of the device under incident light of different wavelengths, as shown in Fig. 4c. The PdSe₂/Si phototransistor shows obvious photoresponsive characteristics in the wavelength range from the ultraviolet (λ = 375 nm) to near-infrared (λ = 940 nm) spectrum. Meanwhile, the device can effectively switch between “on” and “off” states at different wavelengths of light. To further quantitatively evaluate the photoresponse of the PdSe₂/Si phototransistor to light of different wavelengths, we used the following two formulae to calculate the responsivity (R) and specific detectivity (D^*) of the device.^16,28

	(1)

	(2)

where I_light is the photocurrent, I_dark is the dark current, P is the incident light power density, S is the effective working area, Δf is the bandwidth, and i_N is the noise current. Fig. 4d depicts the relationship between R and D^* as a function of the incident light wavelength. The PdSe₂/Si phototransistor has the highest R and D^* under 808 nm illumination, which can be attributed to the inherent light absorption frequency of Si. R is the sensitivity of the photodetector to incident light and D^* is used to evaluate the detection ability of the photodetector to weak incident light from noise.²⁹ Under 808 nm illumination, the highest R value of the PdSe₂/Si phototransistor can reach 1.15 A W⁻¹. The existing surface states and charge traps of the heterostructure could capture photogenerated carriers, which has an effect similar to the local gate voltage. The phenomenon can effectively adjust the conductance of the material, thereby improving the responsivity of the device.³⁰ It can be seen from the noise density spectrum in Fig. S3† that the noise current density is 1.50 × 10⁻¹³ A/Hz^1/2 when Δf = 1 Hz and V_G = 0 V. It can be calculated that the maximum D^* value of the device is 9.39 × 10¹⁰ Jones under 808 nm illumination and both R and D^* values are higher than those of previously reported self-driven photodetectors.^14,15 We consider the effect of noise current when calculating D^*, but this effect was not considered in some previous reports, which may overestimate the calculation of D^*.^16,31 In addition, benefiting from the narrow band gap of the PdSe₂ material, the PdSe₂/Si phototransistor also has a stable and repeatable photoresponse phenomenon under 1550 nm illumination (Fig. S4†). It is further verified that the use of the PdSe₂ material can make the detection range of the device exceed the cut-off wavelength of Si and surpass those of traditional Si-based photodetectors. The fabricated PdSe₂/Si phototransistor has the advantage of a broadband light response.

According to the energy band model in Fig. 3e, applying a gate voltage of 5 V can move the Fermi level on the Si side of the PdSe₂/Si heterojunction to the conduction band. Increasing the built-in electric field at the heterojunction interface allows more photogenerated carriers to separate, thereby improving the performance of the photodetector. In Fig. 4e, we measured the effect of different optical power densities on the time-dependent photocurrent under 808 nm illumination at V_G = 5 V. Obviously, under the same optical power density, the PdSe₂/Si phototransistor produces a larger photocurrent when V_G (= 5 V) is applied. As shown in Fig. 4f, the responsivities of the PdSe₂/Si phototransistor were compared at V_G = 5 V and V_G = 0 V under 808 nm illumination. At a gate voltage of 5 V, the R of the device can reach 1.61 A W⁻¹, which further verifies that appropriate application of the gate voltage can improve the light detection capability of the photodetector. Upon applying a gate voltage, the PdSe₂/Si heterojunction generates a larger built-in electric field to promote photogenerated carrier separation and transport. On the other hand, due to the waved pentagonal mesh structure of the PdSe₂ material, defects may exist, and the photogenerated carriers separated under the action of the gate voltage could be captured by traps. As a result, the gate voltage does not significantly enhance the photoelectric performance of the PdSe₂/Si phototransistor. The R of the device increases as the incident optical power density decreases, which is attributed to the increased carrier recombination activity at higher optical power densities.

Apart from the high sensitivity to the incident light of different wavelengths and different light intensities, the PdSe₂/Si phototransistor exhibits the ability to rapidly and accurately track the optical signals of different frequencies. We used a function signal generator to control the 808 nm laser source to generate pulsed light of different frequencies and used a digital oscilloscope to record the transient photoresponse. Fig. S5a–d† show the time-dependent photoresponse of the PdSe₂/Si phototransistor under 808 nm illumination in the frequency range from 1 kHz to 15 kHz. The device shows a fast switching response and repeatability to the incident light of various frequencies. Under high frequency pulsed light (e.g., f = 15 kHz), the photoresponse has obvious composite degradation, which is caused by trap densities in the PdSe₂ material.³² Response time (rising time/falling time) is another critical parameter in measuring the photodetector performance. The rising time τ_r (or falling time τ_f) is defined as the time required for the photocurrent to rise from 10% to 90% of the peak value (or to fall from 90% to 10% of the peak value).³³ Fig. S5e† shows that the device response times τ_r and τ_f measured at f = 5 kHz and V_G = 0 V are 27.1 μs and 40.3 μs, respectively. The quick response time of the PdSe₂/Si phototransistor is attributed to the built-in electric field formed by the vertical structure heterojunction, which allows the photogenerated carriers to be quickly separated and transported on the vertical structure and collected by the electrodes.³⁴ Under illumination, a large number of photogenerated carries are generated at the PdSe₂/Si heterojunction interface and the trapping states existing in the PdSe₂ material could capture photogenerated carriers. When the laser illumination is turned off, the photogenerated carriers recombine rapidly and the photocurrent decreases rapidly. On the other hand, the release process of carriers from trapping states is relatively slow, causing the falling time of the device to be significantly higher than the rising time.³⁵ In addition, the relationship between the relative balance (I_max – I_min)/I_max of the PdSe₂/Si phototransistor and the modulation frequency is shown in Fig. S5f.† It can be seen that the cut-off frequency of the device is approximately 15.0 kHz. Obviously, the relative balance value of the device decreases slowly and still maintains the photoresponse value higher than 20% under high-frequency pulsed light of 50 kHz, which shows that the PdSe₂/Si phototransistor has application prospects in detecting high-frequency optical signals. Table 1 compares the main parameters of the PdSe₂/Si phototransistor in this work with other PdSe₂-based photodetectors. Due to the unique properties of the PdSe₂ material and the vertical structure of the PdSe₂/Si heterojunction, the responsivity, response speed and detection wavelength range of our PdSe₂/Si phototransistor are comparable to, or even better than, most PdSe₂-based photodetectors.

---

Fig. 5a displays the digital photograph of a 4 × 4 PdSe₂/Si phototransistor array. In order to avoid the crosstalk of device units in the array, we designed the spacing between each unit to be large enough to avoid mutual influence between devices. On the other hand, the PdSe₂ films of each device unit in the array are separated from each other rather than connected together. When designing the electrodes pattern, we avoid contact between different electrodes in the array and the electrodes do not contact the PdSe₂ films of other device units. The time-dependent photoresponses of the PdSe₂/Si phototransistor after 100 switching cycles and 1 month are depicted in Fig. 5b and c, respectively, which demonstrate the excellent reproducibility and environmental stability of the device. To explore the practical application potential of PdSe₂/Si phototransistor, we further investigated its imaging capabilities under incident light irradiation of different wavelengths. Fig. 5d–i present the imaging principal diagrams and the corresponding response current comparison diagrams of the device array under ultraviolet (375 nm, 0.1 mW cm⁻²), visible (532 nm, 0.1 mW cm⁻²), and near-infrared (808 nm, 0.1 mW cm⁻²) laser irradiation. The laser source was driven to generate lasers of different wavelengths and the laser was irradiated on the surface of the device array through a mask with a hollow letter pattern. Finally, by changing the position of the probes in the semiconductor analyzer multiple times, a digital acquisition computer records the response current of each pixel unit in the array point by point. In this process, the device unit under the hollow area of the mask was illuminated with a laser and generates a photocurrent in the device, while other devices that are not illuminated can only detect minimal dark current. We tried to minimize the distance between the mask and the device array to ensure that pixels in the shadow area of the mask are not illuminated by the laser. Therefore, as shown in Fig. 5g–i, the letter patterns “Z”, “J”, and “U” on the mask can be mapped into images, respectively, through the current values recorded by the array pixels. In addition, the 16 fabricated device units all exhibit slight photocurrent fluctuations under the same incident light power density, indicating outstanding device uniformity. These results demonstrate that the PdSe₂/Si phototransistor array has excellent imaging capabilities and high uniformity in the ultraviolet to near-infrared wavelength range and has great potential in the future development of integrated optoelectronic systems.

Conclusions

In summary, we report a high-performance broadband phototransistor based on a PdSe₂/SOI heterojunction. Benefiting from the powerful built-in electric field of the PdSe₂/Si vertical heterostructure, the device has excellent photoresponse with the responsivity, specific detectivity, and response time being 1.15 A W⁻¹, 9.39 × 10¹⁰ Jones, and 27.1/40.3 μs respectively under 808 nm illumination at zero bias, which are better than other PdSe₂-based photodetectors. Appropriate application of the gate voltage can effectively change the optoelectronic properties of the device. Under a gate voltage of 5 V, the Fermi level position of n-Si could move to the conductor band due to changes in charge carriers caused by the electric field effect, resulting in an enhancement of the built-in electric field in the Schottky junction, thereby promoting the separation of photogenerated carriers. The responsivity of the device can be further improved to 1.61 A W⁻¹. Due to the narrow band gap of the PdSe₂ material, the prepared device also displays a pronounced photoresponse to other near-infrared light sources with wavelengths of 940 nm and 1550 nm. It was also found that the PdSe₂/Si phototransistor can function as an imaging sensor array, which can easily record images generated by 375 nm, 532 nm, and 808 nm laser sources. The above results prove that the PdSe₂/Si phototransistor prepared on the SOI substrate provide important guidance for the development of highly sensitive energy-efficient broadband photodetectors and integrated imaging applications.

Author contributions

Y. Chen, Q. Zhu and M. Xu conceived the idea and designed the experiments. Y. Chen and Q. Zhu performed the experiments with the assistance of J. Sun, Y. Sun and N. Hanagata. Y. Chen, Q. Zhu and M. Xu analyzed the data. Y. Chen, Q. Zhu and M. Xu co-wrote the manuscript. All authors discussed the results.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 62090030/62090031, 62274145), the National Key R & D Program of China under Grant No. 2021YFA1200502, the Natural Science Foundation of Zhejiang Province, China under Grant No. LZ20F040001, and the Fund of China Scholarship Council (CSC). The authors would like to acknowledge the device fabrication support from the ZJU Micro-Nano Fabrication Center in Zhejiang University.

References

1. F. Cao, L. Liu and L. Li, Mater. Today, 2023, 62, 327–349 CrossRef CAS .
2. Q. Zhang, N. Li, T. Zhang, D. Dong, Y. Yang, Y. Wang, Z. Dong, J. Shen, T. Zhou and Y. Liang, Nat. Commun., 2023, 14, 418 CrossRef CAS PubMed .
3. J. Michel, J. Liu and L. C. Kimerling, Nat. Photonics, 2010, 4, 527–534 CrossRef CAS .
4. J. Klem, J. Kim, M. Cich, G. Keeler, S. Hawkins and T. Fortune, Appl. Phys. Lett., 2009, 95, 031112 CrossRef .
5. G. Wu, X. Wang, Y. Chen, Z. Wang, H. Shen, T. Lin, W. Hu, J. Wang, S. Zhang and X. Meng, Nanotechnology, 2018, 29, 485204 CrossRef PubMed .
6. A. M. Itsuno, J. D. Phillips and S. Velicu, Appl. Phys. Lett., 2012, 100, 161102 CrossRef .
7. M. Long, P. Wang, H. Fang and W. Hu, Adv. Funct. Mater., 2019, 29, 1803807 CrossRef .
8. M. Xu, T. Liang, M. Shi and H. Chen, Chem. Rev., 2013, 113, 3766–3798 CrossRef CAS PubMed .
9. Q. Liang, Q. Wang, Q. Zhang, J. Wei, S. X. Lim, R. Zhu, J. Hu, W. Wei, C. Lee and C. Sow, Adv. Mater., 2019, 31, 1807609 CrossRef PubMed .
10. J. Sun, H. Shi, T. Siegrist and D. J. Singh, Appl. Phys. Lett., 2015, 107, 153902 CrossRef .
11. C. H. Mak, S. Lin, L. Rogee and S. P. Lau, J. Phys. D: Appl. Phys., 2019, 53, 065102 CrossRef .
12. R. Zhang, Z. Yang, L. Liu, J. Lin, S. Wen, Y. Meng, Y. Yin, C. Lan, C. Li, Y. Liu and J. C. Ho, Adv. Opt. Mater., 2023, 11, 2301055 CrossRef CAS .
13. P. Wu, L. Ye, L. Tong, P. Wang, Y. Wang, H. Wang, H. Ge, Z. Wang, Y. Gu and K. Zhang, Light: Sci. Appl., 2022, 11, 6 CrossRef CAS PubMed .
14. L. H. Zeng, D. Wu, S. H. Lin, C. Xie, H. Y. Yuan, W. Lu, S. P. Lau, Y. Chai, L. B. Luo and Z. J. Li, Adv. Funct. Mater., 2019, 29, 1806878 CrossRef .
15. F. X. Liang, X. Y. Zhao, J. J. Jiang, J. G. Hu, W. Q. Xie, J. Lv, Z. X. Zhang, D. Wu and L. B. Luo, Small, 2019, 15, 1903831 CrossRef CAS PubMed .
16. P. Ye, H. Xiao, Q. Zhu, Y. Kong, Y. Tang and M. Xu, Sci. China Mater., 2023, 66, 193–201 CrossRef CAS .
17. Y. Chen, Q. Zhu, X. Zhu, Y. Sun, Z. Cheng, J. Xu and M. Xu, Nano Res., 2023, 16, 7559–7567 CrossRef CAS .
18. L. Zeng, W. Han, X. Ren, X. Li, D. Wu, S. Liu, H. Wang, S. P. Lau, Y. H. Tsang and C.-X. Shan, Nano Lett., 2023, 23, 8241–8248 CrossRef CAS PubMed .
19. Y. Hu, Q. Zhu, J. Sun, Y. Sun, N. Hanagata and M. Xu, Nanotechnology, 2023, 34, 495202 CrossRef PubMed .
20. M. Long, Y. Wang, P. Wang, X. Zhou, H. Xia, C. Luo, S. Huang, G. Zhang, H. Yan and Z. Fan, ACS Nano, 2019, 13, 2511–2519 CAS .
21. L. H. Zeng, D. Wu, S. H. Lin, C. Xie, H. Y. Yuan, W. Lu, S. P. Lau, Y. Chai, L. B. Luo and Z. J. Li, Adv. Funct. Mater., 2019, 29, 1970005 CrossRef .
22. H. Xu, J. Wu, Q. Feng, N. Mao, C. Wang and J. Zhang, Small, 2014, 10, 2300–2306 CrossRef CAS PubMed .
23. C. Xie, L. Zeng, Z. Zhang, Y.-H. Tsang, L. Luo and J.-H. Lee, Nanoscale, 2018, 10, 15285–15293 RSC .
24. O. Lopez-Sanchez, D. Lembke, M. Kayci, A. Radenovic and A. Kis, Nat. Nanotechnol., 2013, 8, 497–501 CrossRef CAS PubMed .
25. M. Huang, S. Li, Z. Zhang, X. Xiong, X. Li and Y. Wu, Nat. Nanotechnol., 2017, 12, 1148–1154 CrossRef CAS PubMed .
26. N. Huo and G. Konstantatos, Nat. Commun., 2017, 8, 572 CrossRef PubMed .
27. L. Su, W. Yang, J. Cai, H. Chen and X. Fang, Small, 2017, 13, 1701687 CrossRef PubMed .
28. H. Xiao, T. Liang, J. Xu and M. Xu, Adv. Opt. Mater., 2021, 9, 2100664 CrossRef CAS .
29. X. Li, M. Zhu, M. Du, Z. Lv, L. Zhang, Y. Li, Y. Yang, T. Yang, X. Li and K. Wang, Small, 2016, 12, 595–601 CrossRef CAS PubMed .
30. Y. Yang, J. Li, S. Choi, S. Jeon, J. H. Cho, B. H. Lee and S. Lee, Appl. Phys. Lett., 2021, 118, 013103 CrossRef CAS .
31. Y. Fang, A. Armin, P. Meredith and J. Huang, Nat. Photonics, 2019, 13, 1–4 CrossRef CAS .
32. D. Wu, J. Guo, C. Wang, X. Ren, Y. Chen, P. Lin, L. Zeng, Z. Shi, X. J. Li and C.-X. Shan, ACS Nano, 2021, 15, 10119–10129 CrossRef CAS PubMed .
33. Q. Zhu, Y. Chen, T. Chen, L. Zuo, Y. Sun, R. Wang and M. Xu, Nano Res., 2022, 15, 8595–8602 CrossRef CAS .
34. Z. Lu, Y. Xu, Y. Yu, K. Xu, J. Mao, G. Xu, Y. Ma, D. Wu and J. Jie, Adv. Funct. Mater., 2020, 30, 1907951 CrossRef CAS .
35. D. Wu, J. Guo, C. Wang, X. Ren, Y. Chen, P. Lin, L. Zeng, Z. Shi, X. J. Li and C. X. Shan, ACS Nano, 2021, 15, 10119–10129 CrossRef CAS PubMed .
36. R. Zhang, Z. Yang, L. Liu, J. Lin, S. Wen, Y. Meng, Y. Yin, C. Lan, C. Li and Y. Liu, Adv. Opt. Mater., 2023, 2301055 CrossRef CAS .
37. H. Gao, C. Du, P. Dai, J. Leng, W. Wang and K. Li, Vacuum, 2023, 216, 112475 CrossRef CAS .
38. 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 RSC .
39. L. H. Zeng, Q. M. Chen, Z. X. Zhang, D. Wu, H. Yuan, Y. Y. Li, W. Qarony, S. P. Lau, L. B. Luo and Y. H. Tsang, Adv. Sci., 2019, 6, 1901134 CrossRef CAS PubMed .
40. J. Zhong, B. Wu, Y. Madoune, Y. Wang, Z. Liu and Y. Liu, Nano Res., 2022, 15, 2489–2496 CrossRef CAS .

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3nr06643f

‡ These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2024