Phase transition between two-component and three-component ground states of spin-1 Bose-Einstein condensates

For an antiferromagnetic spin-1 Bose-Einstein condensate under an applied uniform magnetic field, its ground state $(ψ_1,ψ_0,ψ_{-1})$ undergoes a phase transition from a two-component state ($ψ_0 \equiv 0$) to a three-component state ($ψ_j\ne 0$ for all $j$) at a critical value of the magnetic field. This phenomenon has been observed in numerical simulations as well as in experiments. In this paper, we provide a mathematical proof based on a simple principle found by the authors: a redistribution of the mass densities between different components will decrease the kinetic energy.

💡 Research Summary

This paper provides a rigorous mathematical proof for a well-known phase transition phenomenon in antiferromagnetic spin-1 Bose-Einstein condensates (BECs). The system, described by a three-component wavefunction (ψ₁, ψ₀, ψ₋₁), is subjected to a uniform external magnetic field. The central result is that for a fixed magnetization M (where 0 < M < 1), there exists a critical value of the quadratic Zeeman energy, q_c(M), which dictates the structure of the ground state. When the magnetic field is weak (0 ≤ q < q_c(M)), the ground state is a two-component (2C) state where the middle component is identically zero (ψ₀ ≡ 0). As the magnetic field strength increases beyond the critical point (q > q_c(M)), the ground state undergoes a phase transition and becomes a three-component (3C) state where all components are non-zero.

The authors work within a mean-field model. The energy functional consists of kinetic energy, potential energy from a trapping potential V(x), spin-independent interaction (repulsive, c_n > 0), spin-dependent interaction (antiferromagnetic, c_s > 0), and the Zeeman energy. Conservation of total particle number (N=1, after normalization) and magnetization M are enforced as constraints. A key simplification, justified by energy minimization, allows the analysis to focus on the non-negative amplitude functions u_j = |ψ_j|, with constant phases.

The proof is built upon several foundational steps. First, the existence of ground states within the admissible class is established under reasonable assumptions on V(x), c_n, and c_s. The ground states satisfy a system of Euler-Lagrange equations and possess regularity properties. A strong maximum principle implies that any component that is zero at a point must be zero everywhere.

A crucial standalone result is the uniqueness of the minimizer within the restricted two-component subspace (where u₀ ≡ 0). This minimizer, denoted z = (z₁, 0, z₋₁), is independent of the parameter q due to the structure of the energy functional in that subspace.

The central innovation driving the main proof is the “mass redistribution” technique, previously introduced by the authors. This principle states that redistributing the squared amplitudes (mass densities) among the components in a linear, conservative manner can never increase the kinetic energy. This provides a powerful tool for comparing the energies of different configurations.

The proof of the phase transition then proceeds by analyzing the competition between the 2C state z and potential 3C states. For small q, it is shown that any 3C state can be mass-redistributed into a 2C state with lower or equal kinetic energy, while simultaneously increasing the combined spin and Zeeman energy contribution. This proves that the 2C state z is energetically superior for q < q_c(M). Conversely, for large q, the Zeeman energy term (which penalizes the middle component) becomes dominant. In this regime, the authors demonstrate that a 3C state must have a lower total energy than the 2C state z. The boundary between these two behaviors defines the critical point q_c(M). The paper also discusses related issues, such as bounding the critical value and the nature of the ground state exactly at the critical point.