Abstract
In superconductors the zero-resistance current-flow is protected from dissipation at finite temperatures (T) by virtue of the short-circuit condition maintained by the electrons that remain in the condensed state. The recently suggested finite-T insulator and the “superinsulating” phase are different because any residual mechanism of conduction will eventually become dominant as the finite-T insulator sets-in. If the residual conduction is small it may be possible to observe the transition to these intriguing states. We show that the conductivity of the high magnetic-field insulator terminating superconductivity in amorphous indium-oxide exhibits an abrupt drop and seem to approach a zero conductance at T < 0.04 K. We discuss our results in the light of theories that lead to a finite-T insulator.
Similar content being viewed by others
Introduction
In 2005, two theoretical groups1,2 considered a disordered, strongly interacting, many-body system of electrons that is not coupled to an external environment (phonons). They posed the fundamental question of whether thermal excitations, which are essential to the mechanism of charge transport, can equilibrate via the interaction with the electron bath or stay frozen as a consequence of, what they termed, the many-body localization (MBL). Their analyses indicated that in such a system an insulating, zero conductance (σ), state is identified at finite-T up to a well-defined critical T, T*. Numerical calculations3,4 based on the analytical approach of ref.1provide ambiguous results regarding the existence of such a phase at nonzero T’s.
In order to experimentally search for this finite-T insulator, it was later suggested5, one should look in disordered systems in which the electrons decouple, at low T, from the phonons. A clear signature of this decoupling is the appearance of discontinuities in the current-voltage (I-V) characteristics6 that result from bi-stability of the electrons T (Te) under V-bias conditions.
We focus on highly disordered superconductors that, at high magnetic-field (B), undergo a superconductor-insulator transition (SIT)7,8. The SIT is a quantum phase transition9 that can be driven by B10,11,12, disorder13, thickness14, gate voltage15 or other parameters in the Hamiltonian. It is observed in variety of systems10,11,12,14,16 and by various experimental techniques10,17,18.
In the B-driven SIT the superconductor goes into an insulating phase at a critical B, BC. In many cases11,17,19,20,21,22 strong insulating behavior is seen only over a narrow range of B to form an “insulating peak” (see Fig. 1). Both theoretical23,24 and experimental20,22,25 studies associate the insulating peak with Cooper-pair localization.
To characterize this B-induced insulating peak, we25 studied its I-V characteristics and found that they exhibit a discontinuous jump in I of more than 4 orders of magnitude as a threshold V, Vth, is exceeded (see top right inset of Fig. 1). This finding26,27 was theoretically linked28 to the formation of a ‘superinsulating’ state that in a manner akin, but opposite, to superconductivity is characterized by an abrupt vanishing of σ at low V-bias.
An alternative view of the discontinuous I-V characteristics was offered by Altshuler et al.29 who analyzed the steady state heat balance in the insulating-peak region under V-bias. They suggested that the I jumps resulted from bi-stability of Te that, at low T, can be very different from the T of the host phonons (Tph). We followed this theoretical work with a systematic study and obtained a good agreement30. We were also able to estimate the T dependence of the e-ph scattering rate, τe−ph, on the high B side of the insulating peak and found a rather strong dependence of τe−ph ∼ T−4, which is in agreement with the modified dirty metal model29,31. The success of this theoretical description provides an essential indication that, in our regime of measurements, the electrons are decoupled from the phonons.
The realization that our samples exhibit a strongly T-dependent insulating behavior with diminishing e-ph coupling motivated us to conduct a systematic study of their Ohmic transport at very low T (T < 0.3 K). In order to achieve that, we had to greatly improve our ability to measure very high sheet resistance (R). While our earlier studies21 were limited to R up to 109 Ω, several improvements (described in the supplementary materials) extended the range of our measurements to 1012 Ω. These improvements enabled the results that follow.
The data presented here are obtained from the sample S1aHiR, a thin film of a:InO, patterned in Hall bar geometry, 0.5 × 0.25 mm2 in size. The sample is superconducting at B = 0 with a Tc ≈ 1.1 K (see left inset of Fig. 1) and undergoes a B-driven SIT. In Fig. 1 we show two isotherms of R in the insulating region, as a function of B from 0.5 to 12 T, at T = 0.08 and 0.1 K. Both show the insulating peak at 5 T. Due to technical reasons we were unable to pinpoint the Bc of our sample but located it to be between 0.16 and 0.4 T. The sample exhibited the thermal bi-stability in the insulating phase as evident by a typical I-V characteristic30, at B = 0.55 T and T = 13 mK, shown in top right inset of Fig. 1.
Our main results are presented in Fig. 2 where we plot the T-dependence of R at various B’s, from 0.5–12 T, spanning the insulating peak. Depending on the R-range, measurements were done using two different techniques. For the moderate-R range (R < 108 Ω) data were obtained by continuous two-terminal measurements (solid lines), whereas for R > 108 Ω each datum (marker) was obtained from a full I-V scan (see methods). The dashed lines joining the markers are guides to the eye.
Based on earlier studies which were limited to a much lower R-range, we were anticipating activated behavior21,26 and adopted an Arrhenius form to present our data. However, the broad range of R in this study brings about the observation of clear deviations from activated transport. While the low R (R < 106 Ω) data are still consistent with activated behavior (for reference we added a dashed black straight line, indicating activated behavior in Fig. 2) the high R data, offering several orders of magnitude broader range, clearly are not.
The deviations, seen in all B values of Fig. 2, crucially differ depending on the value of B. At the high B’s, the convex shape of the curve indicates sub-activation behavior. This behavior is illustrated in Fig. 3(a) where R(B = 12 T) is plotted (in red), using a logarithmic scale, vs. T−1/2. The data convincingly follow a straight line over our full T-range indicating,
This is consistent with the Efros-Shklovskii (ES) variable range hopping (VRH) mechanism of transport32. TES and RES are the ES temperature (TES = 14.8 K) and pre-factor respectively. This dependence holds, with increasing TES, for B’s down to the peak position (at B = 5 T, TES = 23.6 K).
The picture changes dramatically at lower B’s, approaching the SIT (0.5 < B < 2 T). An attempt, shown in blue in Fig. 3(a), to plot data taken in this B range using the ES form clearly fails. A simple activated form is also inadequate as the data clearly appear concave (see Fig. 2).
The concave curvature evident in the B < 2 T data of Fig. 2 signals an unusual, faster than exponential33, R(T) dependence. The anomaly is clearly seen when we plot, in Fig. 3(b), σ as a function of T at B = 0.75 T. Focusing on the T < 0.3 K range we see that σ decreases moderately upon cooling until T = 0.1 K and then undergoes a precipitous drop of 6 orders of magnitude to the noise level in our measurement (σ = 10−12 Ω−1). As we stated earlier, our attempts, indicated by the black curve in Fig. 3(b), to fit these data with an Arrhenius form, failed. For reference we add σ(T) taken at B = 12 T where ES dependence holds (shown in red in that figure).
Our inability to fit the data using an exponential or stretched exponential dependence along with the e-ph decoupling we observe in our samples point in the direction of a finite-T insulator5. To test this possibility we fit our data with the following phenomenological form:
which describes the vanishing of the conductivity at finite T = T*. The result of our fit is plotted using the black dashed line in Fig. 3(b), from which we obtain T0 = 0.138 K and T* = 0.031 K. The data follow this functional form down to T = 0.042 K and σ = 1.3 × 10−10 Ω−1, where deviation larger than our measurement accuracy develop.
In any real system σ = 0 is not a realistic expectation. This is because when σ becomes very small other, parallel, channels will carry the electronic current and contribute to σ. Each such channel will lead to the measured σ being higher and can account for the deviations we observe at σ < 1.3 × 10−10 Ω−1. These can be due to physical processes within the sample or, possibly, due to leakage currents elsewhere in the measurement circuit. More recently, a theoretical paper utilizing a mean field description to a system near the MBL transition34 suggested such deviations should be expected.
By using Eq. (2) we do not intend to adhere to a specific theoretical model2. It is merely a phenomenological description intended to highlight the unusual aspect of our data: σ(T) exhibits a dramatic drop at T < 0.1 K and appear to approach σ = 0 at a finite T = T*. The B-dependence of T* and T0 obtained by fitting our data using Eq. (2) are plotted as the inset in Fig. 3(b). The shaded region indicates the approximate location of the SIT in this sample. It is worth noting that both T* and T0 seem to approach zero in this region.
Another way to illustrate the abrupt nature of the conductivity drop near T* is to compare it to the superconductivity transition in one of our disordered a:InO films. In Fig. 4 we plot σ vs. T at B = 0.75 T for this sample, whereas in the inset we plot R vs. T for sample MInOLa4 at B = 0 T. Despite the different T-range their appearance is remarkably similar: both quantities exhibit a sharp drop over a rather narrow T-range.
It is important to discuss one alternative to Eq. (2) that, on first sight, appears to agree with our results. At least some of the lower B data of Fig. 2 can be described, at T < 0.05 K, by an Arrhenius form indicating activated transport, which results from a mobility gap in the spectrum. A quantitative analysis clearly renders this view inadequate for the following reason. Fitting the B = 0.75 T data using an Arrhenius form leads to an activation T of 0.91 K. If a mobility gap of such magnitude existed in our system we would expect a much sharper increase in R at 0.91 > T > 0.05 K, as seen in the fit presented in the supplementary material. This drop is clearly missing in our data rendering an activated interpretation highly unlikely unless the 0.91 K gap only opens at T < 0.1 K. We are not aware of a theoretical work predicting such a possibility.
While the new results presented here appear to be in contradiction with earlier findings21,26 of activated transport in the peak region, this is not the case: the activation behavior is seen at T’s higher than 0.2 K, below which deviations from activation are seen (see Fig. 2). For these higher T’s, where activation is seen, the maximum value of the activation energy is close to TC(B = 0), confirming earlier observations.
The data we are showing here is consistent with transition into a finite-T insulating state. It is tempting to associate this state with the MBL state suggested theoretically1,2,3,4. Some of the ingredients are certainly present: our system is highly disordered, strongly interacting and, at the relevant T, the electrons decouple from the phonons.
There are other tests that are needed to fully establish the link between our observations and the MBL state chief among which is showing that our electrons are ineffective in reaching equilibrium1,2. This is usually indicated by the presence of long relaxation times in transport. So far, in our experiments, we have not seen such effects but Ovadyahu’s group, who study similar materials in a different regime, reported such slow relaxation phenomena35,36.
On the other hand, we recall that the systems in which we observe the transition to the finite-T insulating state are superconductors at low B and only becomes insulating as B is increased beyond the SIT. Furthermore Cooper-pairing is still dominant in transport even within the insulating regime. While the possible role of Cooper-pairs in forming the finite-T insulator was not considered within the framework of the MBL theories, it was explicitly considered by Vinokur28 et al., in accordance with the suggested duality37 nature of the ‘superinsulating’ state and, more recently, by Feigel’man et al.38 who considered the fractal nature of the electronic wave function near a mobility edge and suggested that, if an attractive interaction near the SIT is considered, a finite-T insulator become feasible. More detailed experiments are needed to test the relevance of these theories.
In summary, we have been able to observe an abrupt drop in σ by several orders of magnitude occurring at T < 0.1 K in a:InO thin film near B induced SIT. This has been found to occur at T and B where the electrons decouple from the host lattice phonons. The measured data cannot be explained using ES model but fit well with the finite-T electron localization down to a certain conductivity.
Additional Information
How to cite this article: Ovadia, M. et al. Evidence for a Finite-Temperature Insulator. Sci. Rep. 5, 13503; doi: 10.1038/srep13503 (2015).
References
Basko, D. M., Aleiner, I. L. & Altshuler, B. L. Metal - insulator transition in a weakly interacting many-electron system with localized single-particle states. Annals of Phys. 321, 1126 (2006).
Gornyi, I. V., Mirlin, A. D. & Polyakov, D. G. Interacting electrons in disordered wires: Anderson localization and low-t transport. Phys. Rev. Lett. 95, 206603 (2005).
Oganesyan, V. & Huse, D. A. Localization of interacting fermions at high temperature. Phys. Rev. B 75, 155111 (2007).
Iyer, S., Oganesyan, V., Refael, G. & Huse, D. A. Many-body localization in a quasiperiodic system. Phys. Rev. B 87, 134202 (2013).
Basko, D. M., Aleiner, I. L. & Altshuler, B. L. Possible experimental manifestations of the many-body localization. Phys. Rev. B 76, 052203 (2007).
Ladieu, F., Sanquer, M. & Bouchaud, J. P. Depinning transition in mott-anderson insulators. Phys. Rev. B 53, 973 (1996).
Goldman, A. M. & Markovic, N. Superconductor-insulator transitions in the two-dimensional limit. Phys. Today 51, 39 (1998).
Gantmakher, V. F. & Dolgopolov, V. T. Superconductor-insulator quantum phase transition. Phys.-Usp. 53, 1 (2010).
Sondhi, S. L., Girvin, S. M., Carini, J. P. & Shahar, D. Continuous quantum phase transitions. Rev. Mod. Phys. 69, 315 (1997).
Hebard, A. F. & Paalanen, M. A. Magnetic-field-tuned superconductor- insulator transition in two-dimensional films. Phys. Rev. Lett. 65, 927 (1990).
Yazdani, A. & Kapitulnik, A. Superconducting-insulating transition in two-dimensional amoge thin films. Phys. Rev. Lett. 74, 3037 (1995).
Baturina, T. I. et al. Superconductivity on the localization threshold and magnetic-field-tuned superconductor-insulator transition in tin films. JETP Lett. 79, 416 (2004).
Shahar, D. & Ovadyahu, Z. Superconductivity near the mobility edge. Phys. Rev. B 46, 10917 (1992).
Haviland, D. B., Liu, Y. & Goldman, A. M. Onset of superconductivity in the two-dimensional limit. Phys. Rev. Lett. 62, 2180 (1989).
Parendo, K. A. et al. Electrostatic tuning of the superconductor-insulator transition in two dimensions. Phys. Rev. Lett. 94, 197004 (2005).
Allain, A., Han, Z. & Bouchiat, V. Electrical control of the superconducting-to-insulating transition in graphenemetal hybrids. Nat. Mater. 11, 590 (2012).
Sacépé, B. et al. Disorder-induced inhomogeneities of the superconducting state close to the superconductor-insulator transition. Phys. Rev. Lett. 101, 157006 (2008).
Crane, R. W. et al. Fluctuations, dissipation and nonuniversal superfluid jumps in twodimensional superconductors. Phys. Rev. B 75, 094506 (2007).
Paalanen, M. A., Hebard, A. F. & Ruel, R. R. Low-temperature insulating phases of uniformly disordered two-dimensional superconductors. Phys. Rev. Lett. 69, 1604 (1992).
Gantmakher, V. F., Golubkov, M. V., Lok, J. G. S. & Geim, A. K. Giant negative magnetoresistance of semi - insulating amorphous indium oxide films in strong magnetic fields. JETP 82, 951 (1996).
Sambandamurthy, G., Engel, L. W., Johansson, A. & Shahar, D. Superconductivity-related insulating behavior. Phys. Rev. Lett. 92, 107005 (2004).
Nguyen, H. Q. et al. Observation of giant positive magnetoresistance in a cooper pair insulator. Phys. Rev. Lett. 103, 157001 (2009).
Feigel’man, M., Ioffe, L., Kravtsov, V. & Cuevas, E. Fractal superconductivity near localization threshold. Annals of Phys. 325, 1390 (2010).
Dubi, Y., Meir, Y. & Avishai, Y. Nature of the superconductorinsulator transition in disordered superconductors. Nature 449, 876 (2007).
Sambandamurthy, G., Engel, L. W., Johansson, A., Peled, E. & Shahar, D. Experimental evidence for a collective insulating state in two-dimensional superconductors. Phys. Rev. Lett. 94, 017003 (2005).
Baturina, T. I., Mironov, A. Y., Vinokur, V. M., Baklanov, M. R. & Strunk, C. Localized superconductivity in the quantum-critical region of the disorder-driven superconductor-insulator transition in tin thin films. Phys. Rev. Lett. 99, 257003 (2007).
Kalok, D. et al. Non-linear conduction in the critical region of the superconductor-insulator transition in tin thin films. J. Phys.:Conference Series 400, 022042 (2012).
Vinokur, V. M. et al. Superinsulator and quantum synchronization. Nature 452, 613 (2008).
Altshuler, B. L., Kravtsov, V. E., Lerner, I. V. & Aleiner, I. L. Jumps in current-voltage characteristics in disordered films. Phys. Rev. Lett. 102, 176803 (2009).
Ovadia, M., Sacepe, B. & Shahar, D. Electron-phonon decoupling in disordered insulators. Phys. Rev. Lett. 102, 176802 (2009).
Shtyk, A. V., Feigel’man, M. V. & Kravtsov, V. E. Magnetic field-induced giant enhancement of electron-phonon energy transfer in strongly disordered conductors. Phys. Rev. Lett. 111, 166603 (2013).
Efros, A. L. & Shklovskii, B. I. Coulomb gap and low temperature conductivity of disordered systems. J. Phys. C 8, L49 (1975).
Baturina, T. I., Mironov, A. Y., Vinokur, V. M., Baklanov, M. R. & Strunk, C. Hyperactivated resistance in tin films on the insulating side of the disorder-driven superconductor-insulator transition. JETP Lett. 88, 752 (2008).
Gopalakrishnan, S. & Nandkishore, R. Mean-field theory of nearly many-body localized metals. Phys. Rev. B 90, 224203 (2014).
Ben-Chorin, M., Ovadyahu, Z. & Pollak, M. Nonequilibrium transport and slow relaxation in hopping conductivity. Phys. Rev. B 48, 15025 (1993).
Ovadyahu, Z. Suppression of inelastic electron-electron scattering in anderson insulators. Phys. Rev. Lett. 108, 156602 (2012).
Ovadia, M., Kalok, D., Sacepe, B. & Shahar, D. Duality symmetry and its breakdown in the vicinity of the superconductor-insulator transition. Nat. Phys 9, 415 (2013).
Feigel’man, M. V., Ioffe, L. B. & Mézard, M. Superconductor-insulator transition and energy localization. Phys. Rev. B 82, 184534 (2010).
Acknowledgements
We are grateful to B. Altshuler, I. Aleiner, E. Altman, D. M. Basko, M. Feigelman, V. Kravtsov, M. Müller, Y. Oreg, Z. Ovadyahu, S. Sondhi and V. M. Vinokur for fruitful discussions. This work was supported by the Israeli Science Foundation and the Minerva Foundation with funding from the Federal German Ministry for Education and Research.
Author information
Authors and Affiliations
Contributions
M.O., D.K. and B.S. prepared the samples and carried out the experiment. M.O., D.K, B.S. and D.S. analyzed the data. B.S. and D.S. initiated this work. I.T., B.S., S.M. and D.S. wrote the paper. All the authors discussed the results and commented on the manuscript.
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Electronic supplementary material
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Cite this article
Ovadia, M., Kalok, D., Tamir, I. et al. Evidence for a Finite-Temperature Insulator. Sci Rep 5, 13503 (2015). https://doi.org/10.1038/srep13503
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/srep13503
This article is cited by
-
Observation of an exotic insulator to insulator transition upon electron doping the Mott insulator CeMnAsO
Nature Communications (2023)
-
Screening the Coulomb interaction leads to a prethermal regime in two-dimensional bad conductors
Nature Communications (2023)
-
Electron-phonon decoupling in two dimensions
Scientific Reports (2021)
-
Quantum magnetic monopole condensate
Communications Physics (2021)
-
Overactivated transport in the localized phase of the superconductor-insulator transition
Nature Communications (2021)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.