On the isotope effect in compressed superconducting H$_\textrm{3}$S and D$_\textrm{3}$S

TL;DR

This study investigates the isotope effect in high-pressure superconducting H3S and D3S, revealing that material quality significantly influences observed transition temperatures and isotope effect exponents, challenging phonon-based superconductivity interpretations.

Contribution

The paper demonstrates that variations in superconducting transition temperatures and isotope effect exponents are strongly affected by material disorder, providing a corrected upper limit that questions phonon-based theories.

Findings

Material quality impacts measured $T_C$ and isotope effect.

Corrected isotope effect exponent is much lower than theoretical predictions.

Disorder levels explain variations in superconducting properties.

Abstract

A maximum superconductive transition temperature $T_\textrm{C}$ = 203.5 K has recently been reported for a sample of the binary compound tri-hydrogen sulfide (H$_\textrm{3}$S) prepared at high pressure and with room temperature annealing. Measurements of $T_\textrm{C}$ for H$_\textrm{3}$S and its deuterium counterpart D$_\textrm{3}$S have suggested a mass isotope effect exponent ${\alpha}$ with anomalous enhancements for reduced applied pressures. While widely cited for evidence of phonon-based superconductivity, the measured $T_\textrm{C}$ is shown to exhibit important dependences on the quality and character of the H$_\textrm{3}$S and D$_\textrm{3}$S materials under study; examination of resistance versus temperature data shows that variations in $T_\textrm{C}$ and apparent ${\alpha}$ are strongly correlated with residual resistance ratio, indicative of sensitivity to metallic order. Correlations also extend to the fractional widths of the superconducting transitions. Using resistance data to quantify and compensate for the evident materials differences between H$_\textrm{3}$S and D$_\textrm{3}$S samples, a value of ${\alpha}$ = 0.043 $\pm$ 0.140 is obtained. Thus, when corrected for the varying levels of disorder, the experimental upper limit ($\leq$0.183) lies well below ${\alpha}$ derived in phonon-based theories.