Extended Data Fig. 10: Magnetoresistance oscillations in non-graphite-gated devices.
From: Landau quantization and highly mobile fermions in an insulator
a, Two-probe resistance (R) measured for device NG1 as a function of the bottom-gate voltage (Vbg) at a fixed top-gate voltage (Vtg). Inset, the device image and layer schematic. b, The magnetoresistance at a fixed Vbg, corresponding to a location indicated by the green arrow in a. The oscillations are visible at 200 mK but disappear at higher temperatures. c, The normalized oscillating component of the magnetoresistance (Rosc/R0) shown in b after background extraction. d, FFT of the 200 mK curve in c. e, Gate-dependent R and device information for another non-graphite-gated device, NG2. Inset, the device image and layer schematic. f, A magnetoresistance trace observed in the insulating regime (location indicated by the green curve in e). g, Its oscillation component together with the FFT (inset). h–j, The same summary as e–g but for device NG3. The top gate is replaced by a thin ZrTe2 flake, which is air-sensitive and its degradation could introduce inhomogeneities to the WTe2 channel. We note that in all three devices, the observed sample resistances are lower than our typical graphite devices. Higher top-gate voltages are also needed for the metal top-gated devices in order to perform measurements in the insulating regime (that is, we expect substantial contact resistance in the measured curves). Magnetoresistance oscillations only develop in the sub-kelvin regime, indicating that freezing out the excess charge carrier is perhaps important to their appearance. All scale bars in the device images are 3 μm.
