Significantly Enhanced Photocurrent in High‑Conductance Topological Insulator Nanosheets

Introduction

Elevating photocurrent response is a central goal in photodetector research. Because the optical penetration depth in bulk semiconductors is typically only a few nanometers, surface states dominate the photoresponse. Materials with large surface‑to‑volume ratios—such as nanowires, graphene, transition‑metal dichalcogenides, and topological insulators—have therefore attracted intense study [1–6]. Recent reports show photocurrent values spanning more than two orders of magnitude across similar material systems [17–22]. While variations in growth protocols and experimental conditions explain some of this spread, intrinsic electronic properties such as conductance have received less attention. Here we investigate how the intrinsic sheet conductance of Sb₂Se₂Te nanosheets governs their photocurrent behavior, providing a straightforward electrical metric for predicting photoresponse.

Experimental Method

High‑purity Sb, Se, and Te (99.995 %) were mixed in the stoichiometric ratio for Sb₂Se₂Te, melted at 700–800 °C for 20 h, and slowly cooled in an evacuated quartz tube. The resulting bulk crystals were used as feed rods for a resistance‑heated floating‑zone furnace (RHFZ). After growth, the crystals were cleaved to expose a mirror‑like basal plane. Raman, EDS, XPS, and XRD confirmed the crystal phase and high crystallinity, consistent with ARPES and Shubnikov–de Haas data from our earlier work [27]. Nanosheets were obtained by mechanical exfoliation onto 300 nm SiO₂/n‑Si substrates with pre‑patterned Ti/Au electrodes. Platinum contacts were deposited by focused‑ion beam (FIB) lithography, yielding ohmic interfaces (Fig. 1). Atomic‑force microscopy measured sheet thicknesses of 58 nm, 178 nm, and 202 nm for samples A, B, and C, respectively. Two‑probe I–V measurements with a Keithley 4200-SCS quantified conductance (G) for each sheet, providing the key parameter for subsequent photocurrent studies.

a, b, and c show SEM images of the three Sb₂Se₂Te nanosheets. The thicknesses were measured by AFM. Two Pt contacts were deposited to record the photocurrent. d, e, and f reveal the linear voltage–current relation, confirming ohmic behavior.

Results and Discussion

Figure 1d–f confirms ohmic I–V behavior. The measured conductances were 4 × 10^–5 S (202 nm), 0.006 S (178 nm), and 7 × 10^–5 S (58 nm), all exceeding 1000 S m^–1 and indicating excellent crystal quality. Photocurrent versus light power (532 nm) is linear for all samples (Fig. 2). The relation can be expressed as I_on = β P^α + I_off, where α≈1 for every thickness, confirming consistent optical absorption across the series. Importantly, the ratio β/G ≈ 1.1 × 10⁵ A W^–1 S^–1 for all sheets, demonstrating that the photocurrent scales directly with the effective conductance.

a, b, and c display measured currents as a function of light power for the three thicknesses. d, e, and f show the linear dependence of I_on on P, with larger I_on for sheets that have higher I_off.

The photocurrent I_ph is defined as I_on – I_off. Because the optical penetration depth in Sb₂Se₂Te (~20 nm) is shorter than all sheet thicknesses, the generated electron–hole pairs reside near the surface and traverse the metallic channel to the contacts. Under an applied bias V, I_ph = V G, so the photocurrent is directly proportional to conductance. Accordingly, the responsivity R = I_ph/(P S) (Eq. 1) scales linearly with G, as shown in Fig. 3. Unlike many topological insulator photodetectors where R decreases with increasing power, our vacuum measurements reveal R independent of P, confirming that the light penetration depth remains limited to the surface region across the studied power range.

Figure 4 presents a log–log plot of R versus G. Data from both Sb₂Se₂Te and the related Sb₂SeTe₂ (≈180 nm, 532 nm) agree with a straight‑line trend, reinforcing the proportionality R ∝ G. The highest measured responsivity, 731 A W^–1, occurs at V = 0.1 V for the most conductive sheet, accompanied by a detectivity of 2.6 × 10¹⁰ Jones—exceeding all reported values for (Sb, Bi)₂(Te, Se)₃ topological insulators and many low‑dimensional materials [27, 28].

Responsivity as a function of sheet conductance. The Sb₂SeTe₂ data (gray dots) corroborate the linear trend.

In ambient air the responsivity shows a pronounced drop below 500 W m^–2 (Fig. 5), indicating that adsorbed molecules partially block the active surface. We modeled this by assuming that of the total incident photons Y, a fraction n is blocked by surface species, leaving m = Y – n photons to generate carriers. The resulting quantum efficiency in air is η(air) = (1 – n/2Y) Z, leading to R(air)/R(vac) ∝ (1 – n/2Y). This relationship matches the experimental data (Fig. 6), suggesting that ≈40 % of the surface is covered by adsorbates at atmospheric pressure.

Responsivity versus light power density in vacuum (blue) and air (red). Responsivity is largely constant in vacuum but falls sharply at low power in air.

Time‑resolved photocurrent measurements (Fig. 7) show an exponential charging process with a characteristic time constant k that decreases with increasing thickness, reflecting faster carrier transport in thinner sheets. Charge and discharge constants are similar, and both increase with decreasing ambient pressure, consistent with reduced surface scattering under partial vacuum.

Bottom left: Photocurrent decay versus time. Top right: Charge time constant versus thickness. Bottom right: Charge/discharge time constants versus ambient pressure.

Conclusion

We have demonstrated that the photocurrent and responsivity of Sb₂Se₂Te nanosheets are governed by the intrinsic sheet conductance. The linear dependence of I_ph on dark current and the proportionality of R to G enable a simple electrical metric for predicting photoresponse. In vacuum, responsivity remains flat across the studied power range, whereas in air the response is limited by adsorbate blocking, with a 40 % surface coverage inferred from our model. The highest reported responsivity of 731 A W^–1 and a detectivity of 2.6 × 10¹⁰ Jones at V = 0.1 V set new benchmarks for Sb–Se–Te topological insulator photodetectors. These findings underscore the importance of optimizing conductance and surface chemistry to achieve superior photoresponsivity.

Abbreviations

ARPES:

Angle‑resolved photoemission spectroscopy

EDS:

Energy‑dispersive X‑ray spectroscopy

SdH:

Shubnikov–de Haas oscillations

XPS:

X‑ray photoelectron spectroscopy