Observation of Spontaneous Ferromagnetism in a Two-Dimensional Electron System

TL;DR

This study observes spontaneous ferromagnetism in a high-quality two-dimensional electron system at low densities, revealing a sequence of phases including paramagnetic, ferromagnetic, and Wigner solid states, with disorder influencing phase transitions.

Contribution

First experimental observation of spontaneous ferromagnetism in a clean 2D electron system at high interaction strength, advancing understanding of correlated electron phases.

Findings

Ferromagnetic transition occurs at $r_s > 35$.

Wigner solid phase appears at $r_s > 38$.

Insulating state emerges at $r_s \,\approx 27$, before ferromagnetism.

Abstract

What are the ground states of an interacting, low-density electron system? In the absence of disorder, it has long been expected that as the electron density is lowered, the exchange energy gained by aligning the electron spins should exceed the enhancement in the kinetic (Fermi) energy, leading to a (Bloch) ferromagnetic transition. At even lower densities, another transition to a (Wigner) solid, an ordered array of electrons, should occur. Experimental access to these regimes, however, has been limited because of the absence of a material platform that supports an electron system with very high-quality (low disorder) and low density simultaneously. Here we explore the ground states of interacting electrons in an exceptionally-clean, two-dimensional electron system confined to a modulation-doped AlAs quantum well. The large electron effective mass in this system allows us to reach very large values of the interaction parameter $r_s$, defined as the ratio of the Coulomb to Fermi energies. As we lower the electron density via gate bias, we find a sequence of phases, qualitatively consistent with the above scenario: a paramagnetic phase at large densities, a spontaneous transition to a ferromagnetic state when $r_s$ surpasses 35, and then a phase with strongly non-linear current-voltage characteristics, suggestive of a pinned Wigner solid, when $r_s$ exceeds $\simeq 38$. However, our sample makes a transition to an insulating state at $r_s\simeq 27$, preceding the onset of the spontaneous ferromagnetism, implying that, besides interaction, the role of disorder must also be taken into account in understanding the different phases of a realistic dilute electron system.