To further probe the CDW phase transition characteristics ofTaTe2, we explore the opto- electronic properties of the MTM device through spatially resolved scanning photocurrent measurements at various temperatures. The photocurrent is measured as Ipc=Ids,laser− Ids,dark, whereIds,laseris the total current under laser illumination, whileIds,darkillustrates the current in the dark environment. In order to obtain spatial photocurrent mapping of the MTM device, we use a diffraction-limited laser beam (∼1 µm) sourced from a continu- ous wave laser (NKT Photonics SuperK Supercontinuum Laser) to scan the entire device controlled by a piezo-controlled scan mirror. The photocurrent is collected as a function of the laser spot position via a preamplifier, while the reflected light can be detected by a sili- con photodetector to record the laser spot position. Figure 6.2a illustrates the photocurrent images obtained with a 1064 nm laser beam scanning from one contact to the other at dif-
the metallic and CDW phase ofTaTe2. A strong photocurrent is observed at TaTe2-metal junctions even without applying a bias voltage between the source and drain electrodes.
This is because the built-in electric field created by the work function mismatch between TaTe2 (3.11 eV) and gold electrodes (5.1 eV) can separate the generated electron-hole pairs efficiently [177; 178; 179]. Furthermore, Figure 6.2b shows the normalized pho- tocurrent intensity versus the laser beam position extracted fromTaTe2-metal junctions at 130 K. The solid red line is a Gaussian fitting to the photocurrent, which corresponds to a diffraction-limited laser spot. Moreover, through comparing the photocurrent profile with the Gaussian fitting curve, we noticed a photocurrent ‘tail’ near the gold electrode region as indicated by the blue arrow, which suggests that the photo-thermoelectric effect (PTE) also contributes to the photocurrent generation [180; 181]. When the TaTe2-metal junc- tion is under local laser illumination, a PTE voltage can be generated due to the difference between the Seebeck coefficients of the metal electrode andTaTe2, which is described as VPT E=∆S∆T. This PTE voltage across the channel produces a photocurrent [182].
By recording photocurrent images of the MTM device at different temperatures using a 1064 nm laser beam with an intensity of 1.76 mW, we obtain a temperature-dependent photocurrent curve in Figure 6.2c. When the temperature is above TCDW, the photocur- rent increases rapidly as the temperature drops. More remarkably, a significant jump in the photocurrent is observed when the temperature reaches TCDW. However, in the CDW phase upon cooling below 170 K, the photocurrent only increases at a rather slow rate as the temperature further decreases. Compared to the insignificant change in the electrical resistance as the CDW phase transition occurs, the photocurrent measurements disclose a remarkable variation upon cooling across TCDW. The pronounced photocurrent change at theTaTe2junctions is due to both PTE and photovoltaic effect (PVE), which are related to the dislocation of atomic arrangement atTaTe2-metal interface [174]. Indeed, both factors depend on the local electronic structure reconstruction and lattice distortion, leading to a more obvious change in photocurrent than electrical resistance.
Figure 6.2: (a) Spatially resolved photocurrent images of MTM device at 130 K, 160 K, 200 K, 230 K, 270 K using 1064 nm illumination. The laser power is 1.76 mW for all pho- tocurrent images. (b) Spatial distribution of photocurrent generated as a function of laser position fromTaTe2-metal junctions at 130 K. The black dots profile and the solid red curve represents recorded photocurrent data and corresponding Gaussian fitting. The blue arrow highlights the photocurrent “tail” on the electrode region. The yellow backgrounds repre- sent the electrode region. (c) Temperature-dependent photocurrent measurement indicates a step-like transition from normal metallic phase to CDW phase, illustrating an increasing photocurrent with decreasing temperature.
To further explore the optoelectronic characteristics of the CDW phase transition of TaTe2, we measure photocurrent response under 1064 nm laser versus light intensities as a function of temperature (Figure 6.3a). Results show that the photocurrent linearly de- pends on the light intensity at each specific temperature (Ipc≈Pα,α =1), indicating that the number of photo-induced electron-hole pairs (EHPs) is proportional to the amount of incident photons [183]. Moreover, at temperatures (140 K, 160 K) belowTCDW the MTM device yields a higher photocurrent than at temperatures (180 K, 200 K) aboveTCDW at each light intensity. One more trend is that while the photocurrent difference at different tem- peratures belowTCDW is approximately independent of the light intensity, the photocurrent difference becomes larger as the light intensity enhances between different temperatures that are aboveTCDW.
Figure 6.3: (a) Power-dependent photocurrent measurements of MTM device under 1064 nm illumination at various temperatures. The photocurrent illustrates a linear dependence with laser intensity. (b) Photoresponsivity versus wavelength from visible to near-infrared spectrum under various temperatures.
Next we probe the photocurrent response of the MTM structure as the incident light wavelength changes from visible to the near-infrared. Figure 6.3b shows the corresponding wavelength-dependent photoresponsivity (R= Ipc/P, where P is the incident laser power) as a function of incident light wavelength for various temperatures spanning from 140 K to 200 K with a 20 K step. The results suggest a jump in the photoresponsivity across TCDW at each wavelength, indicating that the significant change in photoresponsivity is in-
dependent of incident light wavelength. As such, we can identify CDW phase transition through optoelectronic characterization using a light source spanning a wide range of wave- lengths. Interestingly, the photoresponsivity of the MTM device decreases as the incident light wavelength increases. The maximum responsivity achieved is at 140 K with a value of 1.4 mA/W under 650 nm (1.9 eV) laser illumination, which is about twice of the value 0.7 mA/W under 1100 nm wavelength illumination (1.13 eV). For each wavelength, under the same illumination power the number of photons can be calculated asnphoton=
Pλ hc
(h is the Plank constant and c is the speed of light), indicating that a longer wavelength corre- sponds to a larger number of photons. As we use a diffraction-limited laser spot, when the laser beam is focused on the device, the radius of the Gaussian beam can be estimated by r= λ
πNA. Therefore, the photon density is∼ 1
λ, which is in a reverse relationship with the incident wavelength value. On the other hand, the absorption spectrum of theTaTe2device also needs to be considered. The penetration depths are deeper than 1100 nm for a shorter wavelength excitation for metallic-like materials, which leads to an enhanced optical ab- sorbance at shorter wavelength [161]. Overall, the MTM device shows an ultra-broadband photocurrent response from visible to the near-infrared range, illustrating the potential ap- plications in electronic devices.
For optoelectronic applications, response time is an important characteristic and we fur- ther explore the photoresponse dynamics ofTaTe2-metal junctions at various temperatures through temporally resolved measurements under 1064 nm laser illumination. An optical chopper was introduced to mechanically apply on/off light modulations to the laser beam to extract the rise and decay time by collecting photocurrent signals as a function of time. The laser beam is locally focused at the TaTe2-metal interface. Figure 6.4a shows two cycles of typical temporal on-off response when the 1064 nm illumination is turned on and off at 130 K. To obtain the rise or decay time constants (τriseorτdecay) for the device, a single ex- ponential function was applied to fit the rising or decaying region of the curve, respectively (Figure 6.4b). As shown in Figure 6.4c, the rise and decay time constants at temperatures
all indicate high response speeds, and the response time changes from an average of 40µs to 45µs as theTaTe2changes from a normal metallic to CDW phase.
Figure 6.4: (a) The typical photocurrent temporal response of the MTM device and (b) the normalized photocurrent response with the rise and decay time of 45µs at 130 K under 1064 nm illumination. (c) Temperature-dependent response for the response time constants as shown in black dots. The orange background represents the transition from a metallic state to a CDW phase.