Modulating the Photoresponse Performance of Two‐Dimensional GeSe Photodetectors in Visible Region by Ion Irradiation

The 2D GeSe‐based photodetectors have exhibited ultrahigh photoresponsivity (Rλ), sensitive‐specific detectivity (D*), and large external quantum efficiency (EQE) in previous researches. Ion beam techniques have been utilized to effectively modify the surface of nanomaterials for recent years. Herein, the authors propose to engineer the 2D GeSe nanosheets via low‐energy ion irradiation for improving the photoresponse in visible region. The nonmetallic nitrogen and metallic silver elements are selected to modulate the performance of 2D GeSe FETs, respectively. The results show that N‐irradiated GeSe nanosheets have exhibited twice faster photoresponse for 532 nm laser for making up the trailing phenomenon in the decay process during the dynamic response of pristine 2D GeSe. More importantly, via Ag ion irradiation, a self‐driven and higher photoresponsivity GeSe‐based photodetector is realized. The Ag‐irradiated GeSe nanosheets have shown the considerably high Rλ of 9.6 × 102 A W−1 with no bias and no gate voltage applied. The work provides a new direction for modification of other 2D materials by ion beam technique for optoelectronic devices.


Introduction
Photodetector is a device that can convert an optical signal into an electrical signal. Photodetectors based on 2D materials DOI: 10.1002/aelm.202300177 have immerged in recent years for their unique band gaps and excellent optoelectronic properties. [1][2][3][4][5][6] The mechanisms of photocurrent generation mainly include the excitation of free carriers during optical transition and the thermal effect. The former includes the photoconductive effect, photovoltaic effect, and photogating effect; the latter includes the photo-thermoelectric and bolometric effects. [7,8] Up to now, many 2D material-based photodetectors have been reported, including zero band gap graphene, [9] narrow band gap black phosphorus, moderate band gap transition metal dichalcogenides, and wide band gap hexagonal boron nitride. [10] Among the 2D semiconducting materials, IVA-VIA group compounds (MX: M = Ge, Sn; X = S, Se) have a great potential in the application of new-generation optoelectronics and electronics because of their earth-abundant and environmentfriendly. [11,12] More basically, their intrinsic anisotropy contributing from anisotropic crystal structures provide many opportunities for polarization-sensitive photodetectors. [13,14] In addition, their intriguing properties like high carrier mobility and large coefficient of light absorption are favorable for realizing multifunctional devices. [15,16] Recently, 2D-layered GeSe nanosheets have attracted tremendous attention from researchers as the high-performance photodetectors including good air stability, ultrahigh photoresponsivity, excellent external quantum efficiency (EQE), and highly inplane anisotropy. [17][18][19][20][21][22][23] GeSe, a p-type semiconductor, possesses a moderate band gap within 1.13-1.2 eV, which can response for ultraviolet to visible even infrared light. [16,24] Furthermore, monolayer GeSe and double-layer GeSe have a direct band gap and strong absorbance in the visible region by first-principle calculation. [15] However, there is a severe trailing phenomenon in the falling edge during the dynamic response of GeSe to light. It may result from the slow recombination of photon-generated carrier. [25] Unfortunately, researchers have not proposed a reliable way to modify the weakness of GeSe. Ion beam techniques have been used as an effective method of surface modification for modulating the performance of nanomaterials. [26][27][28][29] Among ion beam The Raman vibrations that contains in-and out-of-plane modes in GeSe crystals. c) Raman Spectra of the pristine and N-GeSe nanosheets before and after annealing; d) Raman spectra of the pristine and Ag-GeSe nanosheets.
techniques, low-energy ion irradiation is hoped to avoid excessive damage while introducing foreign impurities in the surface of 2D materials. [26] Therefore, it is an ideal way to introduce foreign impurities in 2D GeSe nanosheets via ion irradiation for accelerating the recombination of photon-generated carrier. [30] Recently, defect engineering is also utilized to increase the internal electronic field of MoSe 2-x /graphene heterostructure by Liu et al. for improving the response speed of the photodetector. [31] In this work, we have successfully realized the optimal modulation in response time and photoresponsivity of 2D GeSe photodetector via different elements ion irradiation. A faster response photodetector based on 2D GeSe nanosheet irradiated by nitrogen ions (N-GeSe) has been demonstrated. The falling response of the N-GeSe-based photodetector is more than twice as fast as pristine GeSe under the illumination of 532 nm laser. Furthermore, a low-voltage driving even self-powered and more sensitive GeSe-based photodetector was proposed by low-energy Ag ions irradiation (Ag-GeSe). At bias of 100 mV, the photocurrent of Ag-GeSe-based photodetector is about ten times larger than that of pristine GeSe. Importantly, the Ag-irradiated GeSe nanosheets have shown the considerably high photoresponsivity of 9.6 × 10 2 A W −1 with no bias and no gate voltage. The value of photoresponsivity is higher than most 2D self-driven photodetectors currently. The results represent that ion irradiation is feasible for modulating the photoresponse performance of 2D materials.

Results and Discussion
Owing to the weak interlayer van der Waal forces, the pristine 2D GeSe nanosheet was mechanically exfoliated onto SiO 2 /Si sub-strate by scotch tapes and polydimethylsiloxane (PDMS). Then, they were irradiated by N ions and Ag ions. Details are shown in the Experimental Section and Figure S1 (Supporting Information).
GeSe, which has a distorted orthorhombic crystal similar to BP, belongs to the space group of Pnma (#62). [32] As shown in Figure 1a, the two typical in-plane directions of zigzag and armchair correspond to the crystallographic axes of b and c, respectively, and the layers of GeSe are stacked along the a axis. [33] Due to the weak van der Waal forces interlayer, GeSe can be cleaved easily in the plane perpendicular to the a axis. We conducted energy-dispersive spectrum mapping to characterize the pristineexfoliated GeSe. The mapping images ( Figure S2a, Supporting Information) exhibit the uniform distributions of Ge and Se elements in the GeSe nanosheet. And the atomic ratio of Ge and Se is about 1:1 (see Figure S2b, Supporting Information), which is in accordance with the expectation. Raman spectroscopy was then performed to characterize the 2D GeSe nanosheets. [5] Generally, the pristine alpha-phase GeSe exfoliated from bulk GeSe has four vibrational modes of A 3 g , B 1 3g , A 2 g (or B 2 2g ), and A 1 g . It is reported that the A 2 g and B 2 2g modes should be 174 and 178 cm −1 , respectively. The A g and B 3g phonons are all shear modes that phonons move parallel along the zigzag and armchair directions in the plane of y and z axes (also b and c axes), whereas the B 2g phonon is a compressive mode that vibrates against out of plane in the x direction (also a direction) as depicted in Figure 1b. [33][34][35] Respectively, the peak intensities were observed in 82, 151.3, 177.1, and 189.5 cm −1 (Figure 1c), which are constant with previous reports. [21,34,35] After ion irradiation, there is no obvious deviation among the three peak positions of A 3 g , B 1 3g , and A 1 g . However, the A 2 g mode shows a slight red shift about 3 cm −1 . It demonstrates that the ion irradiation disturbs the lattice structure of the GeSe to a certain extent. After annealing, the Raman peaks both pristine and irradiated GeSe have a slight increase with no change in the peak positions. This indicates that annealing is beneficial for GeSe to reform better crystal quality. Furthermore, we also investigated the morphologies and thicknesses of the GeSe nanosheets by atomic force microscopy (AFM) before and after ion irradiation (see Figure S3a,b, Supporting Information, respectively). We found that there is little difference in morphology and thickness between the pristine and irradiated GeSe nanosheets. This clearly indicates that rarely damage occurs in the GeSe nanosheets during low-energy N ions irradiation. And the X-ray photoelectron spectroscopy (XPS) characterization (see Figure S4a, Supporting Information) confirmed the presence of nitrogen element in the N-GeSe nanosheet. [36,37] Similarly, as shown in Figure 1d, the Raman spectra also indicate that there is little damage in GeSe crystal structure even irradiated by Ag ions because of the slight peak attenuation in the A 3 g , B 1 3g , and A 1 g vibrational modes. Interestingly, the peak at about 177.1 cm −1 exhibited a certain peak enhancement. This means that from the Raman modes analyzed above, there was a larger interrigid compression vibration. [34,35] Combining with the fitted envelope of XPS spectra (see Figure S4b, Supporting Information), the Ag 3d 3/2 and Ag 3d 5/2 located in 374.5 and 368.5 eV, respectively. Therefore, the existence of Ag element could be verified. [38] Then, the silver atom with larger radius than Ge and Se atoms could increase layer spacing and subsequently weaken the interlayer van der Waals force to enhance the inter-rigid compression vibration. [35,39] The electrical properties of N-GeSe nanosheets were investigated and shown in the inset of Figure 2a. After irradiation, the I-V curve (Figure 2a) indicates Schottky contact between the electrodes and N-GeSe nanosheet different from that of the pristine GeSe showing good Ohmic contact ( Figure S5a, Supporting Information). During the ion irradiation, both the bombardment and chemical bonding may occur. [40] The crystal structure damage induced by irradiation process would bring a bounty of lattice defects on the surface of GeSe. [41] The increase of surface resistance could originate from the capture to carrier of defectstates. In addition, the electrical conductivity of the device also significantly improved after annealing. It is shown that annealing contributes to the increase of free carriers driven by electric field corresponding to the repair of lattice defects including vacancies healing and interstitials migration. [41] Therefore, we conducted all measurements based on the N-GeSe device after annealing. Figure 2b illustrates the output characteristics adjusted by the voltage of back gate representing that the N-GeSe-based transistor still keeps the typical p-type semiconductor behavior, which are the same as that of the pristine GeSe. According to Figure 2c, we can conclude that the on/off ratio of GeSe decreased to about 5 after irradiation under the bias of 3 V, the immediate cause of which is the decrease of carrier mobility after irradiation. The defects induced by irradiation could slower the drift velocity of carriers via blocking and trapping most holes. [42,43] It also makes it hard to tune the migration of holes vertically via backgate voltages. The transfer characteristics of the pristine GeSe ( Figure 2c) and N-GeSe ( Figure S5b, Supporting Information) further confirmed that both pristine and irradiated GeSe show hole-dominated conductivities.
To further explore the hole transport mechanism of N-GeSebased field effect transistors (FETs), the carrier injection model should be discussed. Generally, at a metal-semiconductor contact, thermionic emission (Schottky emission), direct tunneling (DT), and Fowler-Nordheim (F-N) tunneling mainly are the leading of charge transport. [44,45] The fitted curve of Ln( I V 2 ) versus Ln( 1 V ) was plotted as Figure 2d, the linear characteristics of which suggested that both F-N tunneling and direct tunneling were observed. At the same time, the transition from direct tunneling at low voltage to F-N tunneling at high voltage confirms a charge injection barrier at the N-GeSe/electrode interface. [46] And a series of fits for output curves collected from the pristine GeSebased device and N-GeSe-based device were displayed in Figure  S6 (Supporting Information). As shown in Figure 2e, the shape of the hole injection barrier converts from trapezoidal to triangular. At low bias, the charge transport barrier is trapezoidal and the I-V characteristic is consistent with the direct tunneling mechanism. The I-V relationship can be described as follow: where d is the width of the interface barrier; m is the effective mass of the charge carrier; Φ B is the barrier height; ℏ is the Plank's constant divided by 2 . However, with the increase of voltage, the tunneling barrier becomes triangular and the F-N tunneling comes up. The I-V relation can be described as follow: where e is the charge of an electron.
To linearized Equations (1) and (2) in a logarithm scale, respectively, as follow: It is clearly shown that the two regions separated by the threshold voltage (V t ) in Figure 2d are in accordance with the Equations (3) and (4) that are consequently consistent with DT and F-N tunneling models, respectively. And the calculated value of V t is about 1.6 V. The inflection point of the voltage also illustrates the transition of the barrier at the N-GeSe/electrode interface. Furthermore, it is notable that the deepest implantation depth via the ion irradiation is about 5 nm, which was simulated by SRIM as shown in Figure S7a (Supporting Information). It was inferred that there is an ultrahigh contact resistance between the N-GeSe nanosheet and electrodes and a large barrier for holes at the surface. Only when the bias voltage was high, the barrier height was exceed. To investigate the photoresponse of the N-GeSe-based phototransistor (Figure 3a), we measured the I-V curves of the N-GeSe-based photodetector under 532-nm laser illumination with different power densities as shown in Figure 3b (see Figure S8, Supporting Information, under 638-nm laser illumination). The Schottky contact behavior has not changed with the illumination, which also differed from the I-V curves of pristine GeSe-based photodetector ( Figure S9a,c, Supporting Information). With the increase of laser power density, the current gradually increases and approaches saturation finally. To contrast response time, we tested the time-resolved photoresponse of the pristine and irradiated GeSe under 532-nm laser illumination with the power intensity of 0.5 mW cm −2 at 5 V bias voltage and zero gate voltage ( Figure 3c). As shown in Figure S7b (Supporting Information), the time-resolved response of the irradiated GeSe was also demonstrated at the 638-nm laser illumination of 0.45 mW cm −2 under the gate voltage of 40 V. They both exhibited good circulation and represented sensitive photoresponse. As illustrated in Figure 3d, the N-GeSe-based photodetector shows excellent stability during 532-nm laser illuminations with different power densities at the bias voltage of 8 V and the back-gate voltage of 60 V. To evaluate the performance of the N-GeSe-based photodetector, we calculated the critical parameters including photore-sponsivity (R ), specific detectivity (D*), and external quantum efficiency (EQE). The calculation equations are as follows: where I ph is photocurrent and I ph = I light − I dark ; P is the power density of incident light; A is the effective area under illumination. where e is the charge of an electron.
where hv is the energy of an incident photon. By those equations, the responsivity R of the N-GeSe-based photodetector was calculated to be about 1.568 A W −1 at the bias of 5 V and without back-gate voltage. Under the same circumstances, the specific detectivity (D*) and external quantum efficiency (EQE) were estimated to be ≈1.05 × 10 10 Jones and 3.67 × 10 2 %, respectively. Furthermore, response time is a vital figure-of-merits for photodetectors. Response time is the time that current increases (decreases) from 10% (90%) to 90% (10%) of the net photocurrent. Response time ( ) includes the rise time ( rise ) and the fall time ( fall ) that indicate the generation and recombination of the photon-generated carriers, respectively. As illustrated in Figure  S10 (Supporting Information), the response time of the irradiated GeSe-based photodetector were described in detail. It is notable that the decay time of the N-GeSe-based photodetector is faster than that of the pristine GeSe (see Table S1, Supporting Information). Under the condition of no gate voltage, the fall time of the N-GeSe-based photodetector can reach twice faster than www.advancedsciencenews.com www.advelectronicmat.de that of the pristine GeSe under the 532-nm illumination. The rise time of the N-GeSe-based photodetector does not change much compared with that of the pristine GeSe-based photodetector. As shown in Table S1 (Supporting Information), the response time of pristine GeSe and N-GeSe under different measurement conditions is given. The fast fall time of N-GeSe-based photodetector means the shorter life of the photon-generated carriers in the irradiated GeSe than that of the pristine GeSe. From the output characteristics that we have discussed above, it is obvious that the N-GeSe-based device has an increased resistance at the surface. It may contribute from an ultrathin passivation layer at the surface introduced by ion irradiation that effectively reduce the carrier traps at the photosensitive area of the channel, and further improve the response speed. [47] Generally, it is difficult to balance the performance of photodetectors based on 2D materials, which means that faster response time always coexists with relatively lower photoresponsivity. Hence, we further proposed a low-power consumption and highly sensitive photodetector based on GeSe through Ag ion irradiation (Ag-GeSe). The results declare that a modulated GeSe-based photodetector was fabricated by ion irradiations. As shown in Figure 4a, Ag-GeSe-based photodetector could provide more majority carriers in both darkness and illumination. The increased carriers derive from the metal (Ag) ions via irradiation. After irradiation, at an extremely low bias voltage of 100 mV, the Ag-GeSe device maintained good Ohmic contact and showed about ten times current as large as that of pristine GeSe device (Figure 4b). The transfer curve of Ag-GeSe device still exhibited obviously p-type semiconductor feature that is the same as that of pristine GeSe device, which was demonstrated in Figure 4c. In addition, Ag-GeSe device also have sensitive photoresponse for visible light shown in Figure 4d,e. Furthermore, it is worth noting that the photocurrent of the Ag-GeSe-based photodetector is also about ten times larger than that of the pristine GeSe-based photodetector, as shown in Figure 4f. The driving voltage and photoresponsivity of Ag-GeSe-based photodetector is very comparable with other 2D material photodetectors and self-powered photodetectors (see Table 1). [4,7,48,49] By the equations mentioned above, at no bias voltage and gate voltage, the R, D*, and EQE of Ag-GeSe-based photodetector were calculated to be about 9.6 × 10 2 A W −1 , 1.632 × 10 10 Jones, and 2.25 × 10 3 %, respectively.
As shown in Figure S6b (Supporting Information), the largest ion range of the Ag ion irradiation is about 5 nm in the surface of the GeSe crystal. The presence of Ag element was demonstrated by XPS spectroscopy. It is speculated that the concentration of majority carriers (holes) in the multilayer GeSe crystal increased after Ag ion implantation, which may due to the substituting of Ag ions with Ge ions. [50,51] The photoluminescence (PL) spectra could indicate the defects and dopants in the emission band to further explain the optical characterization. [52] According to the PL results (see Figure S11, Supporting Information), there is an obvious quenching after Ag ion irradiation. This means the less recombination of photo-generated electron-hole pair, which would contributed to larger photocurrent and higher photoresponsivity. [23,[53][54]

Conclusion
In summary, we have fabricated a N-GeSe-based photodetector with a fast response in visible spectrum. Importantly, the decay time of the device is able to shorten half more than that of the pristine GeSe-based photodetector, as well as remain good on/off stability. Furthermore, a Ag-GeSe-based transistor was put forward showing ten times photocurrent than that of pristine GeSe. Owing to the irradiation of nitrogen and silver ions, the irradiated GeSe-based photodetector can realize the modification of performance without damage in crystal. The results demonstrate that low-energy irradiation will be a good way to improve even modulate the photoresponse of 2D materials-based photodetectors.

Experimental Section
Preparation of Irradiated GeSe Nanosheets: The GeSe nanosheet was exfoliated from the commercial bulk crystals (HQ graphene co.). At first, repeatedly sticking a little patch of bulk GeSe (0.1-0.4 mm 2 ) via scotch tape (3M co.) until crystal distributed uniformly. Then, the GeSe crystal was entirely pasted on tape to prepared same-sized PDMS on a glass slide, thereby, the GeSe nanosheets on PDMS. Subsequently, GeSe nanosheet was transferred to a clean silicon with 300-nm SiO 2 (ultrasonic cleaning for 15 min with ethanol, acetone, and isopropanol, respectively, at room temperature). Finally, the GeSe nanosheets on SiO 2 /Si substrate were irradiated by nitrogen ions (irradiation energy is 300 eV, dosage is 10 15 ions cm −2 ) and silver ions (irradiation energy is 450 eV, dosage is 10 13 ions cm −2 ).
Characterizations: The morphology and thickness were determined by an optical microscope (BX51, Olympus) and atomic force microscope (Multimode 8, Bruker). Raman and PL spectra were conducted by a Raman system (LabRam HR Evolution, Horiba) with the source of 532 nm laser, the grating of which were 1800 and 600 lines per millimeter. Energydispersive spectrum was performed by the MIRA3 TESCAN system to characterize the element distributions of pristine GeSe nanosheets under a 20 kV voltage.
Device Fabrication: The source and drain electrodes were patterned via electron-beam lithography system (JEOL 6510 with NPGS system). Then, the Cr/Au (15/50 nm) metal film was deposited by the thermal evaporation, followed immersing in 40°C acetone for 40 min for lift-off. The N-GeSe device was annealed for an hour in 300°C with the N 2 gas flow of 115.4 sccm (scrubbing the tube via N 2 for an hour with the same gas flow before heating).
Device Measurements: The electrical characteristics of the device using a semiconductor system (TOSTAR5514B) with a probe station (Lake Shore) and vacuum system (Edwards) were tested. All tests were performed in a vacuum of 10 −4 mbar. For the photoresponse test, the fixed wavelength of lasers (532 and 638 nm) was used with tunable power densities and adjustable optical attenuator. The response time test was analyzed by a preamplifier (SR570, Stanford Research System) and an oscilloscope (UPO2102CS, UNIT) chopping with an electronic timer (GCI-73).

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.