Fig. 1. (a) Schematic of the interactions among the pump, probe laser pulses, and Ta2NiSe5 specimen, where the green, gray, and orange spheres represent the Ta, Ni, and Se atoms, respectively. (b) Schematic illustration of the DMD-based pump-probe microscopy. A spatial filter selects the −1st order of the probe beam to perform scanning. A half-wave plate (HWP) adjusts the probe beam’s polarization. DM, dichroic mirror; OBJ, objective lens. (c) Wide-field optical image of the Ta2NiSe5 flake, where the armchair (a) and zigzag (c) directions are labeled. θ indicates the angle between the probe beam’s polarization and the long edge of the Ta2NiSe5 specimen. Scale bar: 2 μm. (d) High-resolution transmission electron microscopy and selected area electron diffraction pattern measurement of Ta2NiSe5. (e) Enlarged view of the objective lens and the Ta2NiSe5 specimen on the quartz substrate from (b), where the probe beam scans along the armchair and zigzag directions to acquire the anisotropic carrier diffusion coefficients.

Fig. 2. (a) Absolute value of signals (|ΔR/R|) as a function of the delay time of the probe beam in Ta2NiSe5 excited by the pump beam from 50 to 300 μW (probe power, 100 μW). The solid lines are fitted curves based on the exponential decay model. (b) Exponential decay time constants (cyan dots, right axis) and the peak |ΔR/R| signals (red dots, left axis) as a function of the pump power. The red solid line indicates the linear relation between the peak signal and pump power.

Fig. 3. (a) Polarization-resolved |ΔR/R| transient mapping of Ta2NiSe5 from −3 to 30 ps. (b) |ΔR/R| at four selected (probe) polarization angles: 0°, 30°, 60°, and 90°. The solid lines show the fitted decay curves. (c) |ΔR/R| plotted as a function of the polarization angle with delays of 0 and 10 ps. (d) Exponential decay time constants as a function of polarization angles.

Fig. 4. Photocarrier diffusion along the armchair direction of Ta2NiSe5. (a) Contour color map of |ΔR/R| as a function of the probe delay time (horizontal axis) and probe position (vertical axis). (b) |ΔR/R| as a function of probe position at four selected time delays from (a). (c) |ΔR/R| as a function of the probe position at 0, 2, 3, and 5 μm from (a), where the inset shows corresponding decay time constants. The solid lines show the fitted decay curves. (d) FWHM2 as a function of the probe time delay in the armchair (red dots) and zigzag (blue dots) directions. The FWHMs are obtained from fitting the data in (a) with a Gaussian function. The red solid and dashed lines are fitted by a linear function with diffusion coefficients of 500±30  cm2 s−1 and 56±8  cm2 s−1, respectively. The blue solid and dashed lines are fitted by a linear function with diffusion coefficients of 100±10  cm2 s−1 and 35±5  cm2 s−1, respectively. The inset plots the carrier diffusion coefficient at pump powers from 100 to 300 μW along the armchair direction in the first 5 ps.

Fig. 5. 2D transient differential reflection images of Ta2NiSe5 at time delays of (a) 0 ps, (b) 5 ps, and (c) 10 ps with a probe beam polarization angle θ=0°. The color bar shows the logarithmic intensity distribution; the dotted lines show the ovality (O) curves with an intensity value of 1.8×10−5. (d) Gaussian fitted profile of (a) at 0 ps time delay. Scale bar: 2 μm.

Fig. 6. (a) Optical configuration of the DMD-based pump-probe microscope system. L1−L9, lenses; M1−M4, high-reflectivity mirrors; ISO, isolator (central wavelength=800  nm); BE, beam expander; HWP1 and HWP2, half-wave plates; PBS, polarizing beam splitter; ND1 and ND2, neutral density filters; TG, transmission grating; DM, long-pass dichroic mirror; BS, beam splitter (50:50); CP, chopper; PD, photodiode; SP, short-pass filter; LP, long-pass filter; OBJ, objective lens. (b) Front view of the spatial filter that illustrates how redundant diffraction orders are blocked and the −1st-order beam is collected.

Fig. 7. Binary holograms and the corresponding focus positions. (a)–(d) Holograms with (a) blank pattern, (b) vertical stripe pattern, (c) 45° sloped stripe pattern, and (d) concentric circle pattern. (e)–(h) Location of the −1st, 0th, and 1st order beams. (e) Only 0th order, (f) −1st order beam moved laterally, (g) −1st order beam moved along a 45° line, and (h) −1st order beam moved axially (in the z direction, out of the focal plane).

Fig. 8. Measured dynamics of |ΔR/R| over 120 ps with the pump power of 200 μW. (a) Raw data (blue line) and the corresponding biexponential decay fit (red line) for showing the two-step photocarrier relaxation dynamics. (b) Raw data plotted in the logarithmic scale. The dashed lines show the biexponential decay trends extracted from (a). The inset shows the corresponding frequency analysis.

Fig. 9. (a) Wide-field optical microscopy image of the Ta2NiSe5 flake. (b) Transient absorption mapping of the Ta2NiSe5 flake on the quartz substrate at the time delay of 0 ps. The intensity scale selects a logarithmic distribution. (c) |ΔR/R| as a function of the probe delay at positions 1, 2, and 3 in (a). (d) Corresponding time constants extracted by the exponential decays. The error bars show the 80% confidence intervals of the exponential decay fitting.

Fig. 10. Photocarrier diffusion along the zigzag direction of Ta2NiSe5. (a) Contour color map of the absolute differential reflection as a function of the probe delay time (horizontal axis) and probe position (vertical axis). (b) Intensity profiles of the four selected delay times (i.e., Δt=0, 1, 10, and 20 ps) extracted from (a).