Fig. 4: Physical mechanism of noise inhibition in (TaSe4)2I nanowires.
a The relative conductance contribution of normal, \([{G}_{n}/({G}_{n}+{G}_{c})]^{2}\), and CDW carriers, \([{G}_{c}/({G}_{n}+{G}_{c})]^{2}\), as a function of bias voltage. b CDW conductance, \({G}_{c}\) dependence on CDW current \({I}_{c}\). The blue dashed line shows a power-law fit \({G}_{c}\propto {I}_{c}^{\beta }\) to the data on the log-log scale. c Noise of electrons (gray symbols), weighted noise of electrons (blue), and weighted noise of the sliding CDW condensate (red) as a function of bias voltage. d \({S}_{I}/{I}^{2}\), as a function of current. Note that the normalized noise scales inversely with the current, unlike passive resistors, which maintain a constant noise level. The blue dashed line shows a power-law fit \({S}_{I}/{I}^{2}\, {{\rm{\propto }}}\, 1/I\) to the data on the log–log scale. e Measured I–Vs fitted with our model based on Eq. (2). f Noise calculated from Eq. (3) with the experimental data superimposed. No saturation is observed for the total normalized noise level in (TaSe4)2I nanowires, indicating that we are not reaching the intrinsic CDW noise floor. Details for the theory in (e, f) are in the “Methods”. The data are shown for device 1.
