1College of Electronic Information and Optical Engineering, Key Laboratory of Interface Science and Engineering in Advanced Materials of Ministry of Education, Key Laboratory of Advanced Transducers and Intelligent Control System of Ministry of Education, Taiyuan University of Technology, Taiyuan 030024, China
2Shanxi-Zheda Institute of Advanced Materials and Chemical Engineering, Taiyuan 030032, China
The layered two-dimensional material tungsten diselenide () has triggered tremendous interests in the field of optoelectronic devices due to its exceptional carrier transport property. Nevertheless, the limited absorption of in the near infrared (NIR) band poses a challenge for the application of photodetectors in night vision, telecommunication, etc. Herein, the enhanced performance of the photodetector is demonstrated through the incorporation of titanium nitride nanoparticles (TiN NPs), complemented by an atomically-thick layer that aids in suppressing the dark current. It is demonstrated that TiN NPs can dramatically enhance the absorption of light in the proposed photodetector in the NIR regime. This enhancement boosts photocurrent responses through the generation of plasmonic hot electrons, leading to external quantum efficiency (EQE) enhancement factors of 379.66% at 850 nm and 178.47% at 1550 nm. This work presents, for the first time, to our knowledge, that the photodetector is capable of detecting broadband light spanning from ultraviolet to the telecommunication range, all achieved without the reliance on additional semiconductor materials. This achievement opens avenues for the advancement of cost-effective NIR photodetectors.
【AIGC One Sentence Reading】:Incorporating titanium nitride nanoparticles, a WSe2 photodetector's performance in the NIR spectrum is significantly enhanced, boosting external quantum efficiency and enabling broadband light detection from ultraviolet to telecommunication range.
【AIGC Short Abstract】:Incorporating titanium nitride nanoparticles and an atomically-thin alumina layer, we enhanced the performance of a WSe2 photodetector, boosting its external quantum efficiency in the NIR regime without relying on additional semiconductors. This breakthrough enables broadband light detection from ultraviolet to telecommunication wavelengths, paving the way for cost-effective NIR photodetectors.
Note: This section is automatically generated by AI . The website and platform operators shall not be liable for any commercial or legal consequences arising from your use of AI generated content on this website. Please be aware of this.
1. INTRODUCTION
Photodetectors, converting light into electrical signals, have attracted enormous attention for applications in remote sensing, imaging, night vision, telecommunication, and environmental monitoring [1–3]. The emerging of two-dimensional (2D) materials such as graphene and transition metal dichalcogenides (TMDCs) offers a unique opportunity for high-performance photodetectors, with the advantages of superior carrier mobility [4–7]. Various graphene-based ultrafast response photodetectors have been reported; however, the gapless band structure of graphene also generated unsatisfactory dark currents [8]. In addition, the limited absorption of graphene makes graphene-based photodetectors difficult to realize the high photocurrent-to-dark-current-ratio () [9]. Although many techniques have been tried to regulate the bandgap of graphene, the by-product of loss of mobility is undesired [10–12]. By contrast, TMDCs with large bandgap ranges of 1–2 eV offer a competitive platform for efficient electron-hole pair generation under photoexcitation. Besides, TMDCs can form Schottky contact with the electrode, so that the dark current of TMDC-based photodetectors can be significantly suppressed. Compared to the widely used TMDC material, molybdenum disulfide (), tungsten diselenide () stands out due to its superior carrier mobility. Because of the lower electron and hole effective masses, has been widely acknowledged as one of the most promising candidates for the development of highly sensitive photodetectors [13–15]. Recently, Zhou et al. [16] demonstrated a photodetector employing an asymmetric Schottky contact design, which yielded a high responsivity of 2.31 A/W at 650 nm wavelength and an exceptionally low dark current of 1 fA. Bu et al. [17] also fine-tuned the Schottky barrier height of the photodetector, achieving notable results with a high of and an impressive responsivity of 5.16 A/W under 532 nm illumination.
Nevertheless, the limited absorption coefficients of in the near infrared (NIR) wavelength range hinder its potential applications in infrared imaging, night vision, telecommunication [15], etc. In order to boost the photodetection performances in the NIR regime, the heterostructures composed of and other narrow bandgap semiconductor materials have been proposed to serve as the photosensitive layer [18,19]. For example, Sun et al. [20] demonstrated the photocurrent of the photodetector could be remarkably enhanced at 940 nm with the assistance of graphene quantum dots. Xue et al. [21] also reported a broadband photodetector based on the heterostructure of , which displayed an obvious photo-response under the telecommunication wavelength of 1550 nm. Despite this, heterostructured photodetectors that combine with other semiconductor materials inevitably face interfacial challenges, leading to a degradation in performance, especially in dark current [18,19]. Alternatively, a metal nanostructure can excite plasmonic to enhance the absorption and photoelectric properties of optoelectronics [22–24]. Plasmonic nanoparticles (NPs) can also be used to boost the photo-response of photodetectors in the NIR regime, via exciting plasmonic hot carriers [25–28]. For example, Guo et al. [29] integrated the noble metal Au NPs into the photodetector made of , which could remarkably boost the light absorption over a broadband spectral range from 400 nm to 1100 nm. It suggests that the excited plasmonic hot carriers by Au NPs have the potential to overcome the bandgap limitation in semiconductors, ultimately enabling an enhanced photo-response in the NIR regime. In addition to noble metals, transition metal nitrides can serve as functionalized plasmonic materials capable of generating hot carriers. Furthermore, in contrast to Au NPs, transition metal nitride NPs boast a wider plasmonic resonant spectrum and a lower material cost, making them increasingly competitive as plasmonic components introduced into optoelectronic devices [30–32]. Until now, there have been no reports on the use of plasmonic transition metal nitride nanoparticles to enhance the performance of 2D-material-based photodetectors.
Herein, we demonstrate a plasmonic enhanced broadband photodetector with the titanium nitride nanoparticles (TiN NPs) to generate plasmonic hot carriers. The results indicate a significant improvement in the absorption and the photoelectric performances of the photodetector across a broad spectral range from UV to NIR with the incorporation of TiN NPs. Though the increased photocurrent is accompanied by the degradation of dark current, by introducing an atomically thick layer between and TiN NPs, we can maintain the amplitude of the dark current close to its initial level. With the synergistic effects of TiN NPs and layer, the device exhibits a high of at 660 nm wavelength. In addition, a high external quantum efficiency (EQE) enhancement factor of 379.66% and a broad linear dynamic range (LDR) of 99 dB are demonstrated under 850 nm illumination. In literatures, only those heterostructure-based photodetectors comprising and other narrow bandgap semiconductors exhibited a noticeable photoelectric response at 1550 nm wavelength. Surprisingly, the photodetector developed in this work demonstrates responsiveness to the telecommunication illumination at 1550 nm, attributed to the well-regulated energy band diagrams. With the incorporation of TiN NPs and layer, a remarkable enhancement factor in EQE of 178.47% is achieved at 1550 nm. This work provides an alternative approach for developing cost-effective NIR photodetectors with potential applications in areas such as night vision, telecommunication, and more in the near future.
Sign up for Photonics Research TOC Get the latest issue of Advanced Photonics delivered right to you!Sign up now
2. RESULTS AND DISCUSSION
Figure 1 presents the fabrication processes of the plasmonic device, encompassing the functional layers of TiN NPs, , , and TiN electrodes based on a PS nanosphere template (for details see Appendix A). The absorption spectra of TiN NPs with varying deposition thicknesses are depicted in Fig. 6(a) of Appendix D. It shows that the thickness will influence the surface plasmon resonance frequency and change the propagation characteristics of surface plasmon waves, causing shifts in absorption peaks and variations in the absorption intensity of TiN NPs. Thus, TiN NPs exhibit broadband absorption, ranging from 380 nm to 1100 nm, and achieve an average absorption of 45% when the nominal thickness of the TiN NPs is 40 nm. The optimal 40 nm thick TiN NPs exhibit a broad plasmon resonance absorption from 380 nm to 675 nm centered at 520 nm. Besides, the absorption spectra of 40 nm thick TiN NPs under varying incident angles are depicted in Fig. 6(b) of Appendix D. It shows that different incident angles lead to variation in the efficiency of surface plasmon wave excitation, reflection, and interference effects of incident light within the TiN NPs. Larger incident angles may weaken the distribution of the electromagnetic field on the TiN NP surface, decreasing the absorption properties of TiN NPs. In Fig. 2(a), the atomic force microscopy (AFM) image shows the well-ordered and closely-packed monolayer PS nanosphere template with the diameter around 100 nm. Figure 7 in Appendix D displays the AFM image of oxygen-etched monolayer PS nanospheres with a reduced diameter of approximately 70 nm. Figure 2(b) reveals the AFM image of a hexagonal arrangement with 40 nm thick TiN NPs after sputtering TiN and removal of the bottom PS nanosphere template using acetone.
Figure 1.Fabrication processes for integrating TiN NPs array and with photodetector. (a) Cleaning the glass substrate. (b) Preparing PS nanospheres template on the glass substrate. (c) Etching the PS nanospheres template with . (d) Sputtering of TiN on the etched PS nanospheres template. (e) Removal of PS nanospheres template by acetone and obtaining an ordered TiN NPs array. (f) Deposition of on the surface of TiN NPs array. (g) Transfer of onto the substrate. (h) Schematic diagram of the photodetector after sputtering the TiN electrodes.
Figure 2.AFM images of (a) the well-ordered and closely-packed monolayer PS nanosphere template and (b) the hexagonal arrangement TiN NPs after sputtering of TiN film and removal of the bottom PS nanosphere template by acetone. (c) Experimental absorption spectra of the single and the heterostructure of TiN with a 40 nm thick TiN NP layer. (d) Comparative electric field distributions of the single and TiN films at the wavelength of 850 nm.
Figure 3. curves of the , , and devices (a) in the dark and (b) under 660 nm illumination with an intensity of . (c), (d) Transient responses for and single PDs under different wavelengths, ranging from 375 nm to 1550 nm, under the bias of 2 V. The thickness of the layer is 1 nm.
Figure 4.(a) EQE and , and (b) EQE enhancement factor of the single PD and the PD with a 1 nm thick layer. LDR characteristics of the PD at (c) 505 nm and (d) 850 nm.
Figure 5.(a) Raman spectra measured for the , , and TiN films. Working mechanism of the (b) single and (c) device in the UV and visible wavelength range. (d) Energy band diagram of the interfaces illustrating hot electrons excitation and transfer in the NIR and telecommunication range. Working mechanism of (e) the single device and (f) device in the near infrared and telecommunication range.
Benefiting from the broadband plasmonic resonance excited by TiN NPs, the absorption of transition metal sulfides () is expected to be enhanced when combined with TiN NPs. Figure 2(c) shows the experiment absorption spectra of single and TiN films. Besides, the absorption spectra of single and TiN films are also investigated theoretically [refer to Appendix E, Fig. 8(a)]. The broad plasmon resonance absorption of the TiN NPs is overlapped with the absorption peak of and the TiN NPs are covered by , so the resonance absorption peaks of TiN NPs in the absorption spectrum of TiN are not obvious. By integrating simulated and experimental results, we can infer a substantial enhancement in the absorption intensity of , spanning from the wavelength of 390 nm to 900 nm, and the absorption peak in the visible region is more obvious and a bit shifts upon incorporation with TiN NPs. Besides, the absorption curves of become more gentle near 520 nm, with the strong plasmon resonance absorption of TiN NPs. Limited by the imprecision of the experiment, the size and the thickness of the TiN NPs and cannot be precisely controlled compared with simulation. Thus, more typical absorption peaks have been observed in the simulated absorption spectra. In order to gain a more detailed understanding of the absorption enhancement mechanism, we investigate the electric field distributions of the single together with TiN films at different positions for comparison, as shown in Fig. 2(d) and Fig. 9 of Appendix E (for simulation details see Appendix B). It is observed that a dipole-like plasmonic resonance is excited by the TiN NP with a strong localized electric field surrounding the TiN NP, which is responsible for the enhanced absorption in the neighborhood of TiN film compared with single film in a broad wavelength range. Additionally, the intensity of plasmonic resonance gets increased in the wavelength range from 385 nm to 850 nm. Consequently, the absorption enhancement factor [ (%)] between TiN and single films increases from the UV to NIR regime, as shown in Fig. 8(b) of Appendix E, with the maximum (%) reaching 1685% at 850 nm. And the atomically-thick layer has less effect on the absorption of the TiN film. The detailed calculation of (%) can be found in Appendix C.
Figure 8.(a) Simulated absorption spectra and (b) absorption enhancement factor of single film and film.
The characteristics of photodetectors (PDs) with and without TiN-NPs are illustrated in Figs. 3(a) and 3(b), both in dark condition and under illumination of 660 nm with the intensity of . It is noteworthy that the photocurrent of the PD significantly increases from 12 nA to 40 nA under an applied bias of 2 V. This improvement is attributed to the enhanced absorption of light facilitated by the excited plasmonic hot electrons. However, this enhancement introduces a new issue wherein the dark current of the PD experiences a significant degradation to 1.5 pA, compared to the single PD’s dark current of 522 fA. Previous studies have shown that the insertion of an atomic-level thick layer can effectively suppress the dark current without compromising the photocurrent of PDs [33,34]. To maintain the amplitude of the dark current at a level close to its initial state, a 1 nm thick layer is introduced between and TiN NPs. Figure 3(a) demonstrates that the dark current of the PD can be effectively reduced and brought close to its initial level, with the value of 830 fA under the bias of 2 V. Additionally, Fig. 3(b) indicates that the insertion of an atomically thick layer has minimal effect on photocurrent of the PD. For a deeper understanding of the influence of the layer thicknesses on the performances of TiN-NPs based PDs, Fig. 10 in Appendix E presents the compared curves of PD with different layer thicknesses both in dark and under illumination of a 980 nm laser with the intensity of . It can be concluded that all PDs with the layer exhibit significantly higher photocurrent-to-dark-current ratios (), and the largest ratio of is achieved as 37.19 with a 1 nm thick layer. Consequently, we selected PD with a 1 nm thick layer as the optimal device for comparison of other performances against the single PD. Figure 11 in Appendix E displays the curves of PDs in the dark and under different wavelength illuminations with the intensity of . The PD with the layer demonstrates an enhanced from UV to telecommunication range of 1550 nm compared with single PD. The transient responses for and PDs under different wavelength illuminations and applied with the bias of 2 V are also depicted in Figs. 3(c) and 3(d). It is evident that the signal contrast of the PD with is increased over a wide spectral region, ranging from 375 nm to 1550 nm, compared with single PD, consistent with the results of the photocurrent shown in Fig. 11 of Appendix E. Besides, we calculated the response time of the and PDs under visible and NIR wavelengths, shown in Fig. 12 of Appendix E. It can be concluded that both and PDs have shorter response times to visible light (nearly 1 ms) than to NIR light (nearly 100 ms), and the PD with TiN NPs has faster response speed compared with a single device. However, the fastest relaxation time of the device is nearly 100 μs. Controlling the relaxation time of device in ps, we need to make a prodigious effort. There are some strategies that can be considered to improve the relaxation time of devices, such as band structure engineering, surface engineering, external optical excitation, and device structure engineering with transistor structure.
Figure 10.(a)–(e) The curves of PD with different thicknesses of layer in the dark and under illumination of 980 nm laser with the intensity of . (f) The photocurrent to dark current ratio () for PD with different thicknesses of layer in (a)–(e).
Figure 11.The curves of and single PDs in the dark and under different illuminations with the intensity of : (a) 375 nm, (b) 505 nm, (c) 660 nm, (d) 850 nm, (e) 980 nm, (f) 1120 nm, (g) 1208 nm, (h) 1310 nm, and (i) 1550 nm.
The external quantum efficiency (EQE) and responsivity () under bias of 2 V for and PDs are shown in Fig. 4(a). Detailed calculations can be found in Appendix C. The results indicate that both EQE and exhibit an increase over a broad wavelength range from 375 nm to 1550 nm for the PD. The EQE increases from 35% to 90%, and the highest reaches 0.36 A/W at 505 nm for PD, which is 2.5 times higher than that of the single PD. Besides, high EQE enhancement factors () of 379.66% and 178.47% at 850 nm and 1550 nm are obtained, respectively, as shown in Fig. 4(b). Linear dynamic range (LDR) is derived from the ratio of the highest and the lowest detectable illumination power densities () within the linear response range. The LDRs of PD under 505 nm and 850 nm with varied power density illuminations are presented in Figs. 4(c) and 4(d). The device demonstrates a broad LDR of 120 dB at 505 nm, and 99 dB at 850 nm. The limit of detection (LOD) is also a crucial parameter for evaluating the performance of the photodetectors. We also obtained the LOD of the PD from the LDR curves; the results show that the weak-light LODs of the PD are at 505 nm, and at 850 nm.
To further understand the effect of the layer on the performances of films and devices, Raman spectroscopy measurements of , , and films are conducted. As can be seen in Fig. 5(a), the Raman characteristic peak of film is located at , consistent with that in the reported results [29], and the full-width at half-maximum (FWHM) of Raman spectrum for film is . When transferring the on the surface of the layer, the Raman characteristic peak of shifts to , indicating a shift that suggests a weak reaction happened between and [33,34,37]. This reaction reduces the interface defects between and the substrate, improving the crystalline quality of , with the FWHM of the film decreasing to [38–41]. Furthermore, with the addition of TiN-NPs under the film, the Raman intensity of is enhanced due to the plasmonic resonance induced by TiN-NPs, in accordance with the photocurrent changes in Fig. 3(b). Besides, a shift in the Raman characteristic peak occurs because of the reaction between the bottom surface of and TiN-NPs, and the alloying process between TiN-NPs and is suppressed by inserting the layer [42]. These factors are also beneficial for enhancing the crystalline quality of with the FWHM of (details seen in Table 2 in Appendix E). As a result, the performances of the device improved compared with single device under both dark and illumination conditions.
Raman Characteristics of WSe2, Al2O3/WSe2, and TiN-NPs/Al2O3/WSe2 Films
Sample
Peak Position ()
FWHM ()
253
32.70
252
32.49
TiN-
250
31.99
The schematics in Figs. 5(b) and 5(c) show the flow of photogenerated carriers in at the UV and visible wavelengths for the single and devices under an applied forward bias. In the case of the single device, photogenerated holes and electrons yielded by traverse the semiconductor driven by the electric potential, and are collected by the TiN cathode and anode, respectively, as depicted in Fig. 5(b). For the device, the excitation of localized plasmonic resonance by TiN-NPs leads to the generation of more electrons and holes in the ; facilitated by the enhanced electric field in the around TiN-NPs, more photogenerated holes and electrons are transported to the corresponding electrode and collected, as illustrated in Fig. 5(c). Consequently, the absorption and photocurrent of the films and the corresponding device exhibit enhancement in the UV and visible wavelength ranges. This enhancement results in higher EQE values for the TiN- device in the wavelength range from 375 nm to 660 nm.
In the NIR or even the telecommunication range, the limited absorption nature of results in an extremely weak photocurrent. Consequently, the major component in the photocurrent signals for the device becomes the current induced by hot carriers generated by TiN-NPs. To better understand the type of hot carriers, Fig. 5(d) shows the energy band diagram of functional layers at the interfaces without bias, depicting the transfer process of hot electrons generated from TiN-NPs to . The work function of TiN-NPs and is characterized to be 5.08 eV and 4.31 eV, respectively, by the Kelvin probe force microscopy (KPFM) method. The KPFM images of TiN-NPs and are shown in Fig. 14 of Appendix E, and the calculation details can be found in Appendix C. The schematics in Figs. 5(e) and 5(f) illustrate the flow of hot electrons within for the single and devices under an applied forward bias in the NIR and telecommunication ranges. In the case of the single device shown in Fig. 5(e), there is limited hot electron emission from the TiN cathode and only scarce hot electrons can be transported and collected by the anode, resulting in very low EQEs observed from 850 nm to 1550 nm in Fig. 4(a). Conversely, for the device, abundant hot electrons are generated by TiN-NPs. Combined with the hot electrons produced from the TiN electrode, a substantial number of hot electrons are transported into and collected by the anode as depicted in Fig. 5(f). This ultimately contributes to the significantly enhanced EQEs, as shown in Fig. 4(a).
Figure 14.Surface potentials and work functions measured for (a) TiN-NPs and (b) .
In summary, we have demonstrated a novel method to prepare TiN NPs using PS nanospheres as a template and applied them in photodetectors. The results demonstrated a significant enhancement in the absorption and the photoelectric performances of the photodetector across a wide wavelength range from UV to NIR. The incorporation of an atomically thick layer helps maintain the dark current of the PD at the same level as the single PD, resulting in a high of at 660 nm wavelength. The improved performance in the UV and visible wavelength ranges originates from the increased absorption of light due to the localized plasmonic resonance of TiN NPs. Additionally, a substantial number of hot electrons generated by TiN NPs boost photocurrent responses, leading to the improved EQE at NIR wavelengths, with high enhancement factors of 379.66% and 178.47% at 850 nm and 1550 nm, respectively. Moreover, the device exhibits a broad LDR of 120 dB and 99 dB under 505 nm and 850 nm illuminations. Importantly, the device shows much more significant responsivity in the telecommunication range compared with the singe PD. This work introduces an alternative strategy for achieving cost-effective, broadband, and high-performance NIR photodetectors with potential applications in the fields of night vision, telecommunication, and infrared thermal imaging in the near future.
APPENDIX A: EXPERIMENTAL METHODS
1. Preparation of TiN-NPs
The self-assembly polystyrene (PS) nanosphere templates are used to fabricate cost-effective TiN NP arrays, followed by oxygen plasma etching treatment. First, the solution containing 100 nm diameter PS nanospheres is injected on the hydrophilic glass substrate (Shenzhen Huayi Electronics Co., Ltd.) at a temperature of 60°C, with the injection rate of 0.5 mL/min and 0.25 mL/min for the front and back half processes, respectively. Subsequently, the highly ordered and closely packed single-layer PS nanospheres are etched using oxygen, with a flow rate of 60 sccm, etching pressure of 5 Pa, and etching time of 12 min. Next, the TiN film is sputtered on the PS nanospheres substrate with a rate of 0.03 nm/s. Finally, the PS nanospheres are ultimately removed by acetone, leaving an ordered TiN NPs array on the substrate.
2. Preparation of
The mechanical exfoliation method is employed for the dry transfer of onto the substrate with the assistance of PDMS. First, a thin layer is obtained on the blue film tape through repeated mechanical exfoliation from a small piece of crystal. Subsequently, the is transferred from a tap to flexible PDMS using a two-dimensional transfer platform integrated with a microscope. The PDMS, now attached with , serves as the template for preparing subsequent devices by dry transferring onto substrates.
3. Preparation of PDs
The layer is deposited on the surface of the prepared TiN NPs by atomic layer deposition (ALD) method. Then, the is transferred onto the substrate using a two-dimensional transfer platform. Finally, the contact TiN electrodes (the TiN is purchased from Zhongnuo Advanced Material Techbology Co., Ltd.) are deposited using the magnetron sputtering method (Shenyang Kecheng Vacuum Technology Co., Ltd.), employing a copper mesh mask. During the sputtering process, the tray carrying the substrate does not rotate. The thickness of the TiN electrodes is 80 nm and the distance between two TiN electrodes is 20 μm.
4. Characterizations
The absorption spectra of TiN NPs with different thicknesses and are obtained by a micro-region steady-state absorption spectroscopy system (MUV200, Time-Tech Spectra LLC). The morphologies, Raman spectra, and contact potentials of various films are characterized by FM-Nanoview Ra-AFM (Suzhou Flying Man Precision Instrument Co., Ltd.). The current-voltage () curves of the devices are measured in the dark and under the illumination of specific wavelength light (the LEDs: 375 nm, 505 nm, 660 nm, 850 nm, from Thorlabs; the lasers: MGL-F-980 nm, MGL-F-1120 nm, MGL-F-1310 nm, and MGL-F-1550 nm, from Cnilaser). The transient response of the device is tested using a semiconductor analyzer (B1500A, Agilent) with different light sources controlled by a signal generator to produce pulsed light signals. Illumination power intensities are calibrated by an optical power meter (NOVA II, Ophir). The linear dynamic ranges are measured using a semiconductor analyzer (B1500A, Agilent) under 850 nm and 505 nm LEDs at different illumination power intensities. The absorption spectra of TiN NPs under different incident angles were measured by the angle-resolved spectroscopy system (, Shanghai IdeaOptics Instrument Equipment Co., Ltd.).
APPENDIX B: SIMULATION
The optical properties of the films and device are investigated theoretically. In simulation, we simplify the multi-period of the TiN-NPs array into the structure with only one period in COMSOL. All simulations were carried out assuming a periodic boundary along the axis, which is normal to the film plane. The thicknesses of and the glass substrate are 10 nm and 500 nm, respectively. Perfectly matched layer (PML) boundaries are applied at two planes perpendicular to the axis; one lies in the glass and the other is in the air region. Light is illuminated from the glass side. Here, the refractive index of the and TiN is measured by the ellipsometer (Wuhan Eoptics Technology Co., Ltd., SE-VM). The refractive indices and extinction coefficients of other materials used in this work are extracted from Refs. [33,34]. The absorption in the directly affects the performance of the device, so we calculated the absorption spectrum in the single and together with 40 nm diameter TiN NP films over the wavelength range of 420 nm to 900 nm. The electrical field distributions of the device at different absorption positions are also calculated.
APPENDIX C: CALCULATIONS
The (%) of and single films can be calculated by the following equation:
The external quantum efficiency and responsivity values of -based photodetectors were derived according to the number of collected charges and the number of incident photons by Eqs. (C2) and (C3) [36]:
is the photogenerated current density [the measured current density under light () minus the dark current density ()], is the Planck constant, is the frequency of incident light, is photon energy, and is the illumination power density.
The work functions of films were tested by an atomic force microscopy (AFM) instrument equipped with scanning Kelvin probe force microscopy (KPFM). Here, highly oriented pyrolytic graphite (HOPG) was used as a calibration to deduce the work functions of the films by Eq. (C4) [27]: where is the absolute value of the charge of the electron, is the work function of the tested sample, is the work function of HOPG (with a fixed value of 4.65 eV), and and are the average surface potentials of the sample and HOPG obtained from the KPFM images.
Linear dynamic range (LDR) is referred to as the predictable photoresponse generated from the minimum to the maximum optical input, and can be calculated by the following equation [43]: where and are the highest and lowest detectable illumination power densities for the photocurrent density versus light intensity curve, which lie within the LDR range.
APPENDIX D: CHARACTERIZATION OF PS NANOSPHERE AND TiN NPs
Figure 6 shows the absorption spectra of TiN NPs with different deposition thicknesses and incident angles. Figure 7 displays the AFM image of oxygen etched PS nanospheres with the thickness of 40 nm.
APPENDIX E: CHARACTERIZATION OF WSe2 AND TiN-NPs/Al2O3/WSe2 FILMS AND PHOTODETECTORS
Figure 8 exhibits the simulated absorption spectra and absorption enhancement factor of single and films. Figure 9 shows the electric field distributions of the TiN NPs/WSe2 films at different wavelengths. Figure 10 represents the I–V curves and photocurrent to dark current ratios (IP/ID) of PD with different thicknesses of Al2O3 layer in the dark and under illumination of 980 nm laser with the intensity of 10.2 mW/cm2. Figure 11 displays the I–V curves of and single PDs in the dark and under different wavelength illuminations with the intensity of 10.2 mW/cm2. Figure 12 exhibits the response time of and devices in the visible and near infrared wavelength ranges. Figure 13 shows the log I versus log V plots in the dark of , TiN-NPs/WSe2, and devices. Figure 14 demonstrates the surface potentials and work functions measured for TiN-NPs and films. Besides, Table 1 lists the fitted slopes of , TiN-NPs/WSe2, and devices at different regions. Table 2 summarizes Raman characteristics of , Al2O3/WSe2, and films.