Large Area Flexible Thin Layer Terahertz Detector

Due to the low photon energy in the relevant frequency band, fewer materials can directly excite carriers, and an extreme lack of high‐performance, large‐area detectors, limit the promotion, and development of 6G technology. In addition, flexible detectors with the integration of 6G communication and Internet of Things technology has good prospects for application in the development of optoelectronic devices, which are also urgently needed. A large‐area flexible thin layer terahertz detector is designed that combines the advantages of excellent optoelectronic performance in the detection of electromagnetic waves with low photon energy and the localized surface plasmon (LSP) effect in a sub‐wavelength structure detector. Due to the large effective area, the spiral electrode device demonstrates the best performance at room temperature. The noise equivalent power (NEP) and photoresponsivity (RV) are achieved with 0.4 pW Hz−1/2 and 154 MV W−1 at 0.28 THz. In addition, the bending experiments show that the device has extreme advantages for flexible 6G detector applications.


Introduction
6G is the sixth generation of mobile communication technology that has the potential to help human society move toward a smart era due to its better spectrum and energy efficiency, as well as its Figure 1. Main performance and detection bandwidth of terahertz detectors based on different materials. [1][2][3][4][5][6][7][8][9][10][11] is on the nanoscale, which can limit its application. The equivalent noise power is another important metric for detectors, and further improvement of the optical responsivity and reduction of the total noise level can reduce the equivalent noise power, which can be seen in Figure 1. The use of metal sub-wavelength spacing electrodes introduces a strongly localized surface plasmon (LSP) effect, which greatly facilitates the interaction between the unfocused terahertz waves and the active layer.
For the active layer material, we have chosen a Weyl semimetal material in which the valence and conduction bands contact at certain points in the Brillouin zone and pair up to form a chiral Dirac cone. The energy level difference can vary continuously to zero as the crystal momentum approaches the Dirac point. The bandgap width is proportional to the displacement of the crystal momentum near the Dirac point, and the unique energy level properties allow the Weyl semimetal to be well excited by low-energy terahertz photons. [12,13] The absorption of terahertz photons by Weyl semimetal films results in electron-hole pairs, which lead to a concentration of carriers. Due to the lower dark current of Weyl semimetals and the Weyl dots enhancement effect, the Weyl dots can enhance the momentum of carriers excited to the vicinity of the Weyl dots, allowing for stronger bright current excitation, [14] giving it great potential for long wavelength detection.
The flexibility of various microelectronic devices as information carriers can facilitate efficient communication between people and information. [15][16][17][18] Flexible electronic devices have received much attention in recent years as one of the directions of future electronic device development. As a result, wearable devices, medical implantable devices, electronic skin, and smart electronic fabrics are emerging as application directions. To meet the requirements of flexible electronic devices such as thin, transparent, flexible, and stretchable, insulation, and corrosion resistance, common flexible materials are: polyvinyl alcohol (PVA), polyester (PET), polyimide (PI), etc., the above materials have the advantages of convenient and easy to obtain, chemically stable, transparent, and good thermal stability. [19][20][21] The flexible 6G detector will meet the integration of 6G communication and Internet of Things technology, better promote human-computer interaction, and further popularize 6G technology, so the design of a flexible large-area high responsiveness terahertz detector has good prospects for application in the development of optoelectronic devices.
For current flexible detectors, a large number of functional materials have been extensively studied, including 0D nanostructured materials, including quantum dots and nanocrystals, 1D nanostructured materials represented by nanowires and nanorods, and 2D thin film materials represented by optoelectronic films and transparent conductive films. These materials all have unique optoelectronic properties and excellent mechanical flexibility. To obtain a high-performance flexible terahertz detector with high optical response sensitivity, low equivalent noise power, and good mechanical flexibility, we designed a terahertz detector based on a metallic subwavelength electrode structure on PI substrate combined with Weyl semimetal film.
We have designed a large area flexible thin layer terahertz detector that combined advantages of the excellent optoelectronic performance in the detection of electromagnetic waves with low photon energy and the LSP effect in a sub-wavelength structure detector. In our experiment, grating and spiral electrode devices based on PI substrate were designed and compared on the detection of 6G communication frequency covered 0.1 to 0.28 THz. Due to the large effective area, the spiral electrode device demonstrates the best performance at room temperature. The detectivity (D * ), NEP, and R V are achieved with 1.1 × 10 10 cmHz −1/2 W −1 , 0.4 pW Hz −1/2 , and 154 MV W −1 at 0.28 THz. In addition, the bending experiments show that the device has extreme advantages for flexible 6G detector applications.

Device Fabrication and Experiment Setup
Practical photos of the devices are shown in Figure 2. Polyimide (PI) with high-temperature resistance and low dielectric loss was employed as the substrate material. The thickness of the PI substrate for the two samples is 13 μm. The dimensions of these devices are 11 mm × 16 mm with a line spacing of 100 μm and a line width of 100 μm. For the grating device, the electrode length is 7.7 mm and the number of fork-finger electrodes is 20. The metal layer structure is Cu/Ni/Au with thicknesses of 12, 1, and 1 μm respectively. The WTe 2 layers are processed on this substrate by magnetron sputtering deposition.
At first, the air pressure was set as 1 × 10 −4 Pa, which injected argon into the cavity. coating the WTe 2 target by RF drive. The argon flow rate of 50 SCCM and power of 100 W were used. To  maintain the continuity of the nanofilm and good carrier transport properties when the device is bent, we have used a secondary coating method to prepare the WTe 2 material layer. First, place the substrate directly flat and coat for 150 s, then bend the sample, i.e., advancing the two short sides of the rectangle inward by 2 mm each, fixing it, and coating it again for 150 s. The final device was obtained as a double-coated device with a total active layer thickness of 320 nm. Figure 3a shows the thickness of the WTe 2 film. The cross-sectional view shows a thickness of 320 nm and good thickness uniformity. Figure 3b shows the complex conductivity of the WTe 2 nanofilm at room temperature. The real part of the complex conductivity for 0.5, 1, and 1.5 THz are 0.654732, 0.474484, and 0.0014077, and the imaginary part of the complex conductivity are −1.8611, −1.47988, and −1.12853, respectively. Ordinary Drude equation was also employed to fit the frequency dependence of conductivity: [23] here, indicate the frequency of the THz wave, the quasiparticle relaxation time, and 0 DC conductivity. An X-ray photoelectron spectrometer (PHI 5000VersaProbe II) was also used to record the X-ray photoelectron spectroscopy (XPS) of WTe 2 nanofilm to show the quality of the coating film, especially the material homogeneity. The results were demonstrated in Figure 3c,d. In Figure 3c, the peaks belong to the Te 3p 3 . In Figure 3d, the peaks are related to W 4d 5 and W 4d 3.

THz Detector
A light-dark current comparison method based on light current excitation was used to characterize the detection capability of the device. The detection setup is shown in Figure 4a. The terahertz detector was placed in the center of the probe platform, and two tungsten probes (needle tip of 1 μm) were used to contact the metal electrodes of the device and were connected in series with the DC power supply and the picoammeter. The terahertz avalanche diodes were used as terahertz emitters. The model type was TeraSense IMPATT diodes at 0.1 THz with a max output power of 7.1 mW 0.14 THz with a max output power of 5.6 mW and 0.28 THz with a max output power of 1.2 mW. Due to the long use time, the diodes exhibited a loss in power. The output power was measured with a THz power meter (ELVA-1 DPM) before the experiment. The I-V curves were recorded by the picoammeter (Keithley 6485). The total area of the device was larger than the diffraction-limited area. Our calculated responsivity was based on photocurrent, and the conversion to voltage (considered device resistance) was to facilitate comparison with other devices. The detection area was the contact area of the device structure itself, which was determined using a terahertz camera that measured the size of the terahertz spot within the same distance from the source to the detector. Figure 4 shows the detection performance that contains the light current and dark current of the grating and spiral electrode devices at 0.28 THz at room temperature. The grating electrode device showed a lower light current and dark current. As the spiral electrodes were always continuous, they had a larger effective contact area, resulting in higher currents, whereas, for the gate electrode devices, the fork-finger electrodes were disconnected at the end and therefore had a smaller effective detection area than spiral electrode devices with the same electrode width and spacing. The photoresponsivity (R V ), noise equivalent power(NEP), and detectivity (D * ) were also calculated by the light-dark current comparison to demonstrate the photo-electrical conversion capability of their detectors. R V , NEP, and D * are defined as: [24,25] R v = V∕P, NEP = v n ∕R, and D * = V is the photovoltage of the device. P is the incident THz power, and S is the effective detection area of the detector. And n is the noise voltage as shown in Equation 2. [8] v n = k B , T, r, q, and I d are Boltzmann's constant, the detector's absolute temperature, resistance value, elementary charge, and dark current, respectively. The noise and the signal-to-noise ratio were shown in Supporting Information 1.
With different voltage at room temperature for the incident THz frequency of 0.28 THz. The R V linearly increases with the applied voltage, for the nonequilibrium electrons proportional to the drift velocity. With the applied voltage of 25 V, the R V of grating and spiral electrode devices were 25 and 154 MV W −1 , which was higher than that of a Golay cell (0.1 MV W −1 ). The Weyl semimetal and LSP effect make our device an excellent room temperature detector. With the applied voltage of 25 V, the NEP were 2.9 and 0.4 pW Hz −1/2 for the grating and spiral electrode devices, which is just 1/50 to 1/350 of the commercial Golay cell (140 pW Hz −1/2 ). [26] The D* of the rating and spiral electrode devices were 2 × 10 9 cm Hz −1/2 W −1 and 1.1 × 10 10 cm Hz −1/2 W −1 at the applied voltage of 25 V (Figure 5).
6G communications preferentially define 0.28 THz as the experimental band, so the device design focused on this frequency. As the electrode width and electrode spacing were designed to be 100 μm, it was a better match for terahertz at 0.28 THz. It could also be seen from the experimental results that the test results were better at 0.28 THz, but as the device design combines both Weyl semimetal and LSP effect, it also had better detection results for frequency bands such as 0.1 THz and 0.14 THz.
To test the reliability of the flexible device, the relationship between the number of bends and the current loss was tested, with the bend state being 2 mm inward on each of the two short sides of the device (as shown in the inset in Figure 6a) and the unfolded state being with the external force removed.
The loss of current was tested by bending the spiral and gate electrode devices 50-500 times, respectively. The experimental results showed that the current of the spiral device gradually decreases after bending, with the spiral electrode current decreasing by 4.5% and the gate electrode current decreasing by 8.3% after 500 bends. From the experimental results, the flexibility test results met the requirements of wearable devices, and although there was a current loss after a simple bend, the loss was not significant. The spiral electrode had a larger contact surface between the electrode and the material, so the degradation in performance, such as film continuity after bending did not have a significant impact on the spiral electrode. Therefore, spiral electrodes were an ideal choice for practical applications.
The response rate of the spiral electrode device was also tested and according to the test results the current response time was 98 ms. The large response time was mainly caused by the low sampling efficiency of the picoammeter.

THz Active Modulation
The active modulation experiments were performed by a THz-TDS based on photoconductive antennas. The femtosecond laser had a pulse width of 98 fs, a repetition rate of 80 MHz, a central wavelength of 780 nm, a spectral width of 2 nm, and a power   input to the PCA of 15 mW. The photoconductive antenna was Batop PCA-40-05-10-800, the terahertz wave was collimated by a TPX lens, and the device was placed in the focal position of the system. The terahertz spot diameter was 2 mm. The power supply used for the electrically modulated experiments was a DC supply with a current of 1 A and a maximum voltage of 30 V. The laser source used for the optically modulated experiments was a continuous laser with a central wavelength of 980 nm and a maximum output power of 1 W.
The active modulation depth of the THz modulator was defined as: [23] where T p and T 0 are the THz transmittances with and without laser incident on the modulators, respectively. This definition applied to both optically and electronically modulated results. The results of the electrically controlled experiments are shown in Figure 7, where it could be seen that the degree of modulation is greater for the spiral electrode sample at 48% and the grating electrode at 16%, which was ≈1/3 of that of the spiral electrode.
The electric modulation experiment was mainly to test the movement of carriers under the action of an electric field. The results of this experiment showed that the spiral electrode sample had a larger contact surface with the material and a continuous spiral of electrodes when the same applied electric field was applied, so the electric field driving effect was more pronounced, increasing the modulation depth by a factor of approximately three. In contrast, in the gate electrode sample, the contact area between the gate and the material was smaller than that of the spiral electrode, and there was no continuity between the gates, resulting in a lower efficiency of the applied electric field driving the carriers, and therefore a lower modulation depth for the gate electrode. This part of the experiment also verified that the spiral electrode sample was more suitable as a detector for flexible devices due to the large and continuous electrode contact area, and therefore a higher efficiency of the applied electric field driving the carriers This was in agreement with the experimental results in the detector section. Figure 7c shows a rather high transmission of these devices. The transmission was in the range of 40% to 60%, which was also sufficient to support the devices used as modulated devices.
The optical modulation experimental setup is shown in Supporting Information 2. As shown in Figure 8, the optical modulation depths of the spiral and grating electrode samples were 73% and 40%, respectively. The optical modulation depth of the device reflected the carrier motion after laser pumping of the device. For the spiral electrode device, the large contact area between the electrode and the material allowed for a wider distribution of carriers between the electrodes, resulting in a greater inter-electrode density and therefore a greater modulation depth, while the grating electrode had a smaller area compared to the spiral electrode and a lower inter-electrode carrier density than the spiral electrode, resulting in a lower modulation depth. The experimental results www.advancedsciencenews.com www.advelectronicmat.de showed that a continuous spiral electrode with a larger contact area makes it easier for carriers to accumulate around the electrode, resulting in a larger carrier density, and therefore a higher sensitivity was obtained by using the spiral electrode as a detector.

Analysis
Electric field strength distribution and the surface current distribution of the spiral electrodes device and grating electrodes device are shown in Figure 9. The extremely obvious LSP effect could be recognized in these samples. [27] With the subwavelength scale electrode layer on the surface, the limitation of the THz electromagnetic field energy into a small scale, the interaction of the WTe 2 and the THz wave was enhanced. The concentrate effects in the grating electrodes device and spiral electrodes device are quite different. For the former, the terahertz electric field as well as the current density distribution is mainly distributed at the junction of each back end of the grating electrodes, as the front end of the metal gate is not continuous, resulting in an ineffective convergence of the electric field and therefore a smaller distribution of carriers generated by the interaction of the terahertz wave and the Weyl semimetal.
For the latter, the terahertz electric field and the resulting carrier density distribution are distributed over the entire spiral area, and the spiral-shaped metal electrode is continuous over the effective detection area, thus greatly increasing the contact area between the terahertz wave and the electrode, where carriers can be generated within the contact surface, resulting in a more significant LSP effect and better detection performance for spiral electrodes device. Furthermore, since we have used a single polarization direction of the terahertz field for the simulation, a polarization distribution appears on the surface of the spiral electrode, however, in reality, the spiral electrode device is not sensitive to the polarization of the terahertz field. In practical 6G communication, the reception of terahertz waves is more efficient due to its larger effective area, which is more favorable for 6G signals, which are weaker in strength compared to microwave signals. The Seebeck effect is also one of the factors that potentially increase carrier energy. [28] As shown in Figure 1, in comparison with related terahertz detector studies in recent years, our device has a lower NEP and higher detection efficiency, especially for spiral electrodes device, with a larger effective detection area and non-polarization sensitive characteristics, which is more practical for receiving 6G signals with relatively weak fields.

Conclusion
In conclusion, we have designed a large area flexible thin layer terahertz detector that combines the advantages of excellent optoelectronic performance in the detection of electromagnetic waves with low photon energy and the LSP effect in a sub-wavelength structure detector. Due to the large effective area, the spiral electrode device demonstrates the best performance at room temperature. The NEP and R V are obtained with 0.4 pW Hz −1/2 and 154 MV W −1 at 0.28 THz. The effective detection area is almost 1 mm 2 . In addition, the bending experiments show that the device has extreme advantages for flexible detector applications.
Our results show that Weyl semimetal nanofilm-based large-area flexible THz detectors have great potential to be extended to 6G technology.

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