Significant pressure-induced enhancement of photoelectric properties of WS2 in the near-infrared region

Here, a WS2-based photodetector is fabricated and its detection range is extended to optical communication waveband by pressure. At ambient pressure, WS2 exhibits a fast photoelectric response speed (rising/decaying time: 35/49 μs) under 635 nm laser illumination. Upon compression, the responsiveness and external quantum efficiency increase by two orders of magnitude under NIR illumination with different wavelengths as the pressure increases up to 17.2 GPa. The pressure-induced band gap reduction is mainly ascribed to the enhanced S-S interlayer interaction, leading to significant enhancement in the WS2 photoresponsiveness. This study paves the way for designing high-performance broadband photodetectors through pressure modification. IMPACT STATEMENT Fast photoelectric response speed and extended detection range were achieved in the WS2-based photodetector via pressure modulation, which paves a new way for designing high-performance broadband photodetectors. GRAPHICAL ABSTRACT


Introduction
High-performance infrared photodetectors have attracted widespread attention owing to their wide applications in industrial and scientific fields [1][2][3][4][5]. Layered materials, such as graphene and black phosphorus, are often used as promising infrared photodetectors, presenting novel physical properties and excellent optoelectronic properties. However, with the ever-increasing demands, the disadvantages of the zero bandgap in graphene and the environmental instability in black phosphorus limit their further development in infrared photoelectric detection technology [6][7][8]. Therefore, there is an urgent requirement for novel infrared detection materials and feasible CONTACT  methods for the improvement of the detection performance of photoelectric materials in the infrared waveband.
Recently, transition metal dichalcogenides (TMDs) have attracted extensive attention in the scientific community [9][10][11][12]. TMDs exhibit various outstanding physical properties such as tunable band gap, high optical absorption, and strong spin-orbit coupling that have immense potential in the fields of energy, catalysis, and optoelectronics [13][14][15][16][17][18][19]. TMDs have a typical sandwiched structure. The middle layer consists of transition metal atoms, and the top and bottom layers are made of chalcogen atoms. The intralayer atoms are covalently bonded, while the interlayers are connected by van der Waals forces [20]. Therefore, they can be easily exfoliated into multilayers or even monolayer nanosheets via ultrasonic or mechanical methods [21][22][23][24]. Similar to the most well-known TMDs, 2H-WS 2 is a widely studied TMD with thickness-dependent indirect-direct bandgap transitions and various bandgap values [25][26][27], for its individual use as a substrate material for photodetectors, the performance-tuning methods for WS 2 -based photodetectors often focus on improving the quality of the prepared WS 2 and tuning its number of layers. Substantial efforts have been devoted to improving the WS 2 photoelectronic properties since Perea-Lopez et al. [28] first used it in the preparation of photodetectors, increasing its responsiveness, response time, and response range to 53.3 A W −1 , 560 µs, and 1064 nm, respectively [29][30][31][32][33]. This series of studies provide new prospects for enhancing the photoelectronic properties of WS 2 . However, improving the photoelectric properties of WS 2 devices further with faster response speeds and wider spectral response ranges is still an enormous challenge.
Employment of high pressure is an effective method of modulating the bandgap and electronic states of materials that can further tune photoelectronic-related properties. Unlike chemical doping and molecular modifications [34,35], the application of pressure does not introduce chemical impurities and is more controllable. Recently, pressure-induced enhancement of photoelectric properties has been realized in a variety of materials, such as perovskites, Bi 2 O 2 S, AgIn 5 S 8 , and iodine [36][37][38][39][40]. In TMDs, enhanced photoelectric properties have been observed in ReS 2 and PtS 2 under pressure [41,42]. 2H-WS 2 has excellent mobility and carrier density than PtS 2 and ReS 2 , which shows the advantages in photoelectric applications [43][44][45][46][47][48]. Previous experimental and theoretical calculation results have revealed that the applied pressure not only effectively regulates the charge carrier characteristics of WS 2 but also substantially modulates the light absorption ability of WS 2 [47,49]. All these properties are inextricably linked to the photoelectronic properties of a semiconductor, and therefore, it is anticipated that pressure regulation may be an effective method of modifying the photoelectric performance of WS 2 .
In this study, we systematically investigate the photoelectric properties of WS 2 under pressure. The WS 2based photodetector prepared in a pressure cell simultaneously achieves an ultra-fast response speed and a broad detection range at ambient pressure. The photoelectrical properties of WS 2 are significantly enhanced under high-pressure modulation in the detection range from visible to NIR. The continuous tuning of the photoelectric properties provides a platform for developing pressure-modulated devices for future functional material technologies.

Sample characterization
WS 2 crystal was purchased from 6Carbon Technology (Shenzhen, China). The sample was characterized by X-Ray diffraction (XRD, Rigaku Synergy Custom FR-X) with Mo Kα radiation (λ = 0.7099 Å) and scanning electron microscopy (SEM, FEI Magellan 400). A JEOL (JEM-2200FS) instrument with an accelerating voltage of 200 kV was employed for high-resolution transmission electron microscopy (HRTEM) and selected area electron diffraction (SAED).

In-situ XRD and Raman measurements
High pressure was generated using a diamond anvil cell (DAC) with a pair of diamond anvils (300 μm culet). Pressure was measured via the ruby fluorescence method. Silicone oil was used as the pressuretransmitting medium in the experiments. In-situ highpressure XRD was measured by Rigaku Synergy Custom FR-X. The Raman spectra were recorded using a Renishaw inVia notch filter spectrometer equipped with a 514.5 nm exciting laser.

Photocurrent
In-situ high-pressure photocurrent measurements were conducted in a symmetrical DAC with a pair of diamonds with a culet size of 400 µm. A T301 stainless steel gasket was pre-indented to ∼ 50 µm (thickness), and a sample chamber with a diameter of ∼ 380 µm was drilled at the center of the indentation by a laser. An epoxy-boron nitride layer was inserted between the steel gasket and the diamond culet to achieve electrical insulation between the electrical leads and the metal gasket. Two platinum sheets were arranged to contact the sample in the chamber for photocurrent measurements as electrodes. The sample area between the two electrodes was approximately 8 × 10 −9 m 2 . No pressure-transmitting medium was used for the photocurrent measurements. The photocurrents under pressure were recorded using a Keithley 2461 semiconductor characterization system under the illumination of a xenon lamp. A high-precision photocurrent scanning test microscope system (MStarter 200) with a source meter (Keithley 6482) was used for laser illumination of different wavelengths, including 635, 980 nm (power density: 15.3 mW cm −2 ), 1270 nm (power density: 5.7 mW cm −2 ), 1450 nm (power density: 5.4 mW cm −2 ) and 1650 nm (power density: 8.3 mW cm −2 ), the light power density under different wavelengths were measured by an optical power meter, the dark environment was required to avoid the interference of environmental light.

Computational details
All DFT calculations were performed using the projector augmented wave method (PAW) as implemented in the Vienna ab initio simulation package (VASP). The Perdew−Burke−Ernzerhof (PBE) was treated as the exchange-correlation potential within the generalized gradient approximation (GGA). We calculated the lattice parameters using the optB88b-vdW functional under pressure. The band gap and optical properties of WS 2 under pressure were postprocessed using VASPKIT. The cut-off energy for the plane-wave was determined to be 600 eV, and Monkhorst−Pack k-point meshes spanning not more than 0.03 Å −3 in the Brillouin zone were chosen. The convergence for energy was set to 10 −6 eV on each atom, and the internal atomic positions and lattice constants under each hydrostatic pressure were optimized until the force on each atom was less than 0.01 eV Å −1 .

Results and discussion
The XRD pattern of layered WS 2 at ambient pressure is shown in Figure 1(a) which indicates that our sample has a 2H structure with space group P6 3 /mmc, and its structure is schematically represented in Figure 1(b). Moreover, SEM and HRTEM measurements were obtained to investigate the crystal structure of WS 2 , and the results are displayed in Figure 1(c,d). These images clearly illustrate the layered stacking structure with a lattice spacing of 0.27 nm in WS 2 , and the corresponding SAED pattern (inset of Figure 1(d)) proves its hexagonal structure.
To evaluate the photoresponse performance of WS 2 , a series of optoelectronic measurements was performed within the DAC device using a two-point probe. The inset image of Figure 2(a) shows a schematic of the prepared installation. The relationship between the photocurrents and different biases with the light on/off time interval of 5 × 10 −3 s is shown in Figure 2(a). In addition, the response speed is an important parameter to describe the quality of a photodetector, the rise time τ rise (and decay time τ decay ) is defined as the time intervals for the response rising (falling) from 10% (90%) to 90% (10%) of its peak value. As shown in Figure 2(b), under 635 nm laser illumination, τ rise and τ decay are calculated as ∼ 35 µs and ∼ 49 µs, respectively, at ambient pressure. Compared with the previously reported WS 2 photodetectors [28][29][30][31][32][33][50][51][52][53][54][55], our measured response and recovery times are the most rapid (Table S1). In addition, the effect of the illumination intensity (P in ) on the photoelectronic properties of WS 2 was also studied. Figure 2(c) plots the photocurrents of WS 2 under different P in ranging from 0.1 to 25.7 mW cm −2 under 5 V bias. It is clearly observed that the photocurrents increase with increasing P in , which could be understood that more photogenic electron-hole pairs contribute to the photocurrent due to the higher illumination intensity [56]. Figure S1 displays the sublinear relationship between the photocurrent and P in extracted from Figure 2(c), this phenomenon has been observed in several photoconductive detectors and is considered to be the result of electron-hole generation, trapping, and recombination in semiconductors [4,[57][58][59].
The effects of pressure on the photoresponsiveness of WS 2 were further investigated. Figure 3(a) displays the photocurrents of WS 2 device under xenon lamp irradiation (P in = 1.2 mW cm −2 ) at selected pressures with 5 V bias, and the observed linear I-V curves demonstrate that there is good Ohmic contact between WS 2 and electrodes, as shown in Figure S2a. The responsivity (R) is described as the photocurrent generated per unit power of the incident light on the effective area, that is, I ph = I illumination − I dark , S is the effective illuminated area [60,61]. During the compression process, the increase photoelectric response time of WS 2 is mainly owing to enhanced interatomic interaction, as shown in Figure S2b [62]. In addition, the effect of pressure on the WS 2 optoelectronic properties can be expressed by the dependence of the photocurrent and R on the pressure, as presented in Figure 3(b). The value of R reaches ∼ 692.9 A W −1 at 20.1 GPa that is ∼ 20 times that at 1.4 GPa (35.6 A W −1 ). In addition, according to the given formula, when P in and S remain constant, as the pressure increases, ∼ 20-fold increase in the photocurrent of the WS 2 device is achieved by pressure tuning. However, the absence of a significant increase in specific detectivity is considered mainly due to the inevitable increase in dark current with the applied pressure, as shown in Figure S2c and d. However, compared to the previous literatures, the specific detectivity under high pressure remains basically stable at a relatively high order of 10 11 [29][30][31][32]50,51], as shown in Table S1. The pressure-induced photocurrent increase is closely related to the decrease of bandgap, which is calculated in the following theoretical section. As the band gap decreases, electrons in the conduction band are more likely to be excited into the conduction band by the same illuminated wavelength. This phenomenon has also been   observed in other single-crystal TMDs, such as PtS 2 and ReS 2 , however, the photocurrent enhancement of WS 2 is much higher than the previous reports [41,42]. Enhancing the response of optoelectronic materials is still an immense challenge for developing the functionalities and applications of high-performance NIR photodetectors. Hence, we used several specific NIR wavelength lasers as the light source, including 980, 1270, 1450, and 1650 nm, as shown in Figure 3(c,d) and Figure S3a and b, to observe the variation in the photoelectric properties of WS 2 under applied pressures up to 17.2 GPa. Similar to the variation observed under xenon lamp illumination, when compared with the photocurrents of WS 2 under ambient pressure ( Figure S4), the photocurrent increased significantly upon the application of pressure under illumination of different NIR wavelengths. It is worth mentioning that, as shown in Figure 3(e), in the spectral range of 1450-1650 nm, the pressure-induced photocurrent increasement reaches up to three orders of magnitude compared with the initial pressure. The external quantum efficiency (EQE) is another important parameter for photoelectric behavior characterization and can be calculated using the relationship where c is the light velocity, e is the electron charge, and λ is the wavelength of illumination. As shown in Figure 3(f), under the illumination of the selected NIR wavelengths, the photoelectric characteristic parameters R and EQE, like the photocurrent, exhibit a significant increase upon compression. As the pressure increases to 17.2 GPa, WS 2 exhibits an increase of approximately two orders of magnitude in R and EQE in the NIR region relative to the initial pressure. Intriguingly, although the pressure leads to an enhancement of the photoelectric properties throughout the NIR waveband, the photoelectric properties at shorter wavelengths is consistently better than those at longer wavelengths, as shown in Figure 3(e,f). This is in good agreement with the optical absorption properties of WS 2 obtained from our theoretical calculation results (the insets in Figure 5(b,d)). Furthermore, we compared the photocurrent at 0.5 GPa in the compression process with that under ambient pressure after decompression. We surprisingly found that the photocurrent of WS 2 after compression treatment is seven times as large as that at 0.5 GPa (upon compression), as shown in Figure S3c. This indicates that pressure is of great significance as a means of adjusting high-performance infrared photodetectors.
To further understand the mechanism of the pressuremodified photoelectric property enhancement of WS 2 , in-situ XRD and Raman spectroscopy measurements were conducted to decipher the structural modification upon compression. It can be seen in Figure S5 that when the applied pressure is up to 21.8 GPa, the insitu high-pressure XRD patterns prove that there is no structural phase transition occurring within pressures up to 21.8 GPa, suggesting that the pressure-influenced enhancement of photoelectronic properties is not associated with it. To provide insight into the effects of the applied pressure, we investigated the Raman spectra of WS 2 . As shown in Figure 4(a), at 0.8 GPa, due to the 514.5 nm excitation wavelength is close to the B excitonic resonance for WS 2 , there are several second-order Raman peaks could be observed in the Raman spectrum. As the pressure increases, the B exciton energy shows a blue shift, and the second-order Raman vibration modes disappeared above 4.4 GPa (as shown in Figure S6), which is consistent with the previous literature results [63]. Besides that, there are two obvious first-order Raman peaks at ∼ 355 and 421 cm −1 that belong to E 1 2g (in-plane vibrational mode of W and S atoms) and A 1g (out-of-plane S-atom vibrational mode) during the whole compression process. With an increase in pressure, these two Raman modes undergo rapid hardening (Figure 4(b)). The slope of the A 1g vibration mode is larger than that of the E 1 2g mode, which is not only indicating the A 1g mode is more responsive to the external pressure, but also means the S-S interlayer interaction is enhanced under compression. To further understand the chemical bonding, we performed calculations on the electron localization function (ELF), as shown in Figure 4(c). The enhanced interlayer interaction is also evident from the evolution of the S-S interlayer distances that reduce from 3.604 Å at ambient pressure to 2.965 Å at 20 GPa. The Raman spectra and ELF calculations reveal that the reduction of the interlayer spacing leads to the enhancement of S-S interlayer interaction upon compression.
To explicitly provide information regarding the orbitals involved in the charge transfer of the band structure, we also performed theoretical calculations of the WS 2 electronic structure under different pressures using the VASP as shown in Figure S7. Upon compression, the bandgap of WS 2 shows a successive decrease due to the broadening of valence band maxima and conduction band minima with the contribution of the W 5d and Se 3p orbitals. Similar to MoS 2 and ReS 2 , the enhanced S-S interlayer interaction of WS 2 leads to a reduction in the bandgap during compression [42,47,64]. The narrowed bandgap causes the valence electrons to be more easily excitable by illumination, such that they may cross over the bandgap into the conduction band, resulting in more photo-generated carriers, eventually causing higher photocurrents during compression [42,65].
Evaluating the variations of optical parameters (particularly the absorption coefficient α(ω) and the photoconductivity σ (ω) obtained by theoretical calculations) with pressure facilitates the further understanding of the photoelectric properties of WS 2 under high pressure. As shown in Figure 5, under ambient pressure, the α(ω) and σ (ω) is consistent with the experimental results [66,67], which proves the reliability of the theoretical calculations. With the increase of pressure, α(ω) and σ (ω) both increase to varying degrees, which is in accordance with our experimental observation of pressure-induced photocurrent increases. In addition, as shown in Figure S2a, the slope change of I-V curves for WS 2 under xenon lamp irradiation also provides reliable evidence of the increasing trend of σ (ω) with increasing pressure. However, the α(ω) and σ (ω) are anisotropic, with considerably different values for different incident light polarization directions. Upon compression, the α(ω) and σ (ω) perpendicular to c axis are almost unchanged (Figure 5(a,c)); however, these two parameters parallel to c axis exhibit a significant enhancement and redshift. These results suggest the experimentally enhanced photoelectric properties of WS 2 during compression is principally attributable to changes in the optical parameters along the c axis, and the red-shifted absorption edge is attributable to the pressure-induced bandgap reduction.

Conclusion
In summary, we investigated the photoelectric properties and structural changes of WS 2 under pressure through in-situ XRD, Raman spectroscopy, photocurrent measurements, and theoretical calculations. The WS 2 -based photodetector prepared in the pressure cell exhibits an ultra-fast response speed of ∼ 35 µs and a broad photodetection range up to 1650 nm. Moreover, the photoelectronic properties of WS 2 are significantly enhanced from the visible to the NIR range upon the application of pressure. Extensive high-pressure spectroscopic analyses and theoretical calculations reveal that upon compression, more photogenerated carriers are modulated by the decrease in the WS 2 bandgap owing to the enhanced S-S interlayer interaction, leading to significantly enhanced photoelectric characteristics of WS 2 . The employment of high pressure as a modulation strategy leads to a remarkable enhancement in the photoelectronic properties of WS 2 . Our work provides an experimental and theoretical basis for the practical application of WS 2 as a functional material in future optoelectronic devices.

Disclosure statement
No potential conflict of interest was reported by the author(s).