Electron Dephasing in Mesoscopic Metal Wires

The low-temperature behavior of the electron phase coherence time, ${\tau }_{\varphi }$, in mesoscopic metal wires has been a subject of controversy recently. Whereas theory predicts that ${\tau }_{\varphi }(T)$ in narrow wires should increase as $T^{-2\mathrm{/}3}$ as the temperature $T$ is lowered, many samples exhibit a saturation of ${\tau }_{\varphi }$ below about 1 K. We review here the experiments we have performed recently to address this issue. In particular we emphasize that in sufficiently pure Ag and Au samples we observe no saturation of ${\tau }_{\varphi }$ down to our base temperature of 40 mK. In addition, the measured magnitude of ${\tau }_{\varphi }$ is in excellent quantitative agreement with the prediction of the perturbative theory of Altshuler, Aronov and Khmelnitskii. We discuss possible explanations why saturation of ${\tau }_{\varphi }$ is observed in many other samples measured in our laboratory and elsewhere, and answer the criticisms raised recently by Mohanty and Webb regarding our work.