| Historically, it has been appreciated that glassy effects, which appear due to the competition between electron-electron interactions and disorder, play a key role in the insulator and may be crucial even in the critical regime of the metal-insulator transition (MIT). In this dissertation we present the simplest method for studying the systems of spinless electrons, which describes a MIT and glassy features associated with it. An experimental realization of such model would be a system of 2D electrons, such as a metal oxide semiconductor field effect transistor (MOSFET), where spins are polarized with the external magnetic fields, thus rendering the spin degree of freedom inactive.;We start by studying the electron glass transition in a classical case, which is then extended to include quantum hopping. We find that even though the glassy phase persists into quantum domain and only "melts" when hopping t becomes larger than some finite t G, the system becomes metallic as soon as t > 0. We argue that Anderson localization effects should be considered for the insulating phase to exist at finite t. To account for Anderson localization effects we introduce the typical medium theory, that allows a treatment of localization effects on an equal footing with electron-electron interactions, establishing the insulating phase for t < tcrit. We find that adding localization significantly enhances the size of the glassy phase, making the entire insulating phase glassy. Most interestingly, we show that within our theory, glassy behavior is not suppressed as soon as electrons become mobile, but it persists close to MIT as long as mobility remains small and quantum fluctuations are not sufficiently large to overcome barriers between metastable states. Our prediction of an intermediate metallic glass phase separating the insulator and the normal metal has been experimentally confirmed [1]. |
