Origin of Nonlinear Transport across the Magnetically Induced Superconductor-Metal-Insulator Transition in Two Dimensions

Abstract

We have studied the effect of perpendicular magnetic fields and temperatures on the nonlinear electronic transport in amorphous Ta superconducting thin films. The films exhibit a magnetic field induced metallic behavior intervening the superconductor-insulator transition in the zero temperature limit. We show that the nonlinear transport in the superconducting and metallic phase is of non-thermal origin and accompanies an extraordinarily long voltage response time.In recent years, the suppression of superconductivity in two-dimensions (2D) by means of increasing disorder (usually controlled by film thickness) or applying magnetic fields has been a focus of attention. Conventional treatments [1,2,3,4,5] of electronic transport predict that in 2D the suppression of the superconductivity leads to a direct superconductor-insulator transition (SIT) in the limit of zero temperature (T = 0). This traditional view has been challenged by the observation of magnetic field (B) induced metallic behavior in amorphous MoGe [6,7,8] and Ta thin films [9]. The unexpected metallic behavior, intervening the B-driven SIT, is characterized by a drop in resistance (ρ) followed by a saturation to a finite value as T → 0. The metallic resistance can be orders of magnitude smaller than the normal state resistance (ρ n ) implying that the metallic state exists as a separate phase rather than a point in the phase diagram. Despite many theoretical treatments [8,10,11,12,13,14,15,16,17], a consensus on the mechanism behind the metallic behavior is yet to be reached. Proposed origins of the metallic behavior include bosonic interactions in the nonsuperconducting phase [10,11], contribution of fermionic quasiparticles to the conduction [12,13], and quantum phase fluctuations [14,15].In a recent paper [9] on the magnetically induced metallic behavior in Ta films, we have reported the nonlinear voltage-current (I-V ) characteristics that can be used to identify each phase. The superconducting phase is unique in having both a hysteretic I-V and an "immeasurably" small voltage response to currents below an apparent critical current I c . The metallic phase can be identified by a differential resistance (dV /dI) that increases with increasing I, whereas the insulating phase is identified by a dV /dI that decreases with increasing I. The contrasting nonlinear I-V in the metallic and insulating phase are shown in Fig. 1(a).The main purpose of this Letter is to report that the origin of the nonlinear transport, particularly in the superconducting and metallic phase, is not a simple reflection of T -dependence of ρ via the unavoidable Joule heating. We describe the effect of B and T on the nonlinear TABLE I: List of sample parameters: nominal film thickness, mean field Tc at B = 0, normal state resistivity at 4.2 K, critical magnetic field at which the resistance reaches 90% of the high field saturation value, and correlation length calculated from ξ = Φ0/2πBc where Φ0 is the flux quantum.

Insulator-to-metal transition in vanadium supersaturated silicon: variable-range hopping and Kondo effect signatures

“…The insulator-to-metal transition could be driven by means of control of key parameters such as band filling or energy bandwidth. These parameters are experimentally varied by doping, pressure, chemical composition, and magnetic fields [4][5][6]. Particularly the insulator-to-metal transition controlled by doping depends on a critical doping concentration (n crit ) which for a variety of systems satisfies = a n 0.25 b crit 1 3 / [7], where a b is the effective Bohr radius of the dopant impurity.…”

Superconducting proximity effect in inverted InAs/GaSb quantum well structures with Ta electrodes

“…Since the separation between the two superconducting electrodes in our devices is always much greater than the correlation length of Ta (only a few tens of nm) 24 , no supercurrent is observed across the Ta−InAs−Ta junction and no cross-talk effect is expected between the transport across the two InAs−Ta junctions. Figure 1a shows the dV/dI spectra of two InAs−Ta junctions as a function of the dc current I dc .…”

Theoretical analysis of drag resistance in amorphous thin films exhibiting superconductor-insulator transitions

“…Furthermore, experiments on Tantalum films show distinct signatures in the metallic regime which disappear in the insulating and superconducting regimes, and also distinguish it from the thermally destroyed superconducting phase. 24 Third, even if the metallic behavior of the films is a finite temperature phenomena, within the vortex paradigm, the resistance still arises due to vortex motion. Therefore the drag calculated within this paradigm using a diffusive vortex model should still be adequate, and our results do not depend crucially on the existence of a zero-temperature intervening metallic state.…”

“…[19][20][21][22][23][24][25][26] Near the "SIT critical point," as temperature is lowered below ϳ100 mK, the resistance curve starts to level off, indicating the existence of a novel metallic phase, with a distinct nonlinear I-V characteristics at least in Ta films that are interpreted as a consequence of vortex dynamics. 24 Another interesting experimental finding is the nonmonotonic behavior of the magnetoresistance. 19,23,27,28 As one increases magnetic field further from the SIT point, the resistance climbs up quickly to very large value in InO and TiN films, before it plummeting back to the normal state resistance, as shown in Fig.…”

“…In Ta and MoGe films, as well as some InO films, the resistance peak is not as large but is still apparent. [20][21][22][23][24][25] Two competing paradigms may account for these phenomena. On the one hand, within the quantum vortex pictures, 2,29-31 the insulating phase at the peak of the magnetoresistance implies the condensation of quantum vortices, and the high field negative magnetoresistance indicates the gradual depairing of Cooper pairs and the appearance of a finite electronic density of states at the Fermi level.…”