Phase separation dynamics in a two-dimensional magnetic mixture

Abstract

Based on classical density functional theory (DFT), we investigate the demixing phase transition of a two-dimensional, binary Heisenberg fluid mixture. The particles in the mixture are modeled as Gaussian soft spheres, where one component is characterized by an additional classical spin-spin interaction of Heisenberg type. Within the DFT we treat the particle interactions using a mean-field approximation. For certain magnetic coupling strengths we calculate phase diagrams in the density-concentration plane. For sufficiently large coupling strengths and densities, we find a demixing phase transition driven by the ferromagnetic interactions of the magnetic species. We also provide a microscopic description (i.e., density profiles) of the resulting non-magnetic/magnetic fluid-fluid interface. Finally, we investigate the phase separation using dynamical density functional theory (DDFT), considering both nucleation processes and spinodal demixing.Comment: 15 pages, 10 figure

Novel structure formation of a phase separating colloidal fluid in a ratchet potential

“…Experimentally, such systems involve, e.g., ferro-colloids and polymers [28]. In our model, the asymmetric interaction potentials yield (in two dimensions) a first-order demixing transition for a broad range of parameters, as we have already shown in earlier studies [29,30]. In the present letter we combine this interaction potential with a (one-dimensional) ratchet potential coupling to the spins.…”

“…We make the choice J * = J/(k B T ) > 0 such that ferromagnetic ordering is favored. We also note that we set J(|r − r |) = 0 for distances |r−r | < σ, i.e., we assume that at these separations the interaction between two magnetic particles is negligible as compared to the repulsion from the core potentials [29]. In fact, for J * = 0 the particles are identical and no demixing occurs.…”

“…Equation (2) can be simplified drastically by assuming that the magnetic moments relax instantaneously. To this end we factorize the one-body density profile into a translational number density part, ρ α (r, t), and a (normalized) orientational distribution function, h α (r, ω, t) and then set the functional derivative δF/δh m (r, ω, t) = 0 for all times t [29]. This yields a self-consistency relation for the orientational distribution h m (r, ω, t) = exp(B(r, t) · s(ω, t))/ dω exp(B(r, t) · s(ω, t)) where the (self-consistent) effective field is given by B(r, t) = − dr dω ρ m (r , t)h m (r , ω , t)J(|r − r |)s [29].…”

“…To this end we factorize the one-body density profile into a translational number density part, ρ α (r, t), and a (normalized) orientational distribution function, h α (r, ω, t) and then set the functional derivative δF/δh m (r, ω, t) = 0 for all times t [29]. This yields a self-consistency relation for the orientational distribution h m (r, ω, t) = exp(B(r, t) · s(ω, t))/ dω exp(B(r, t) · s(ω, t)) where the (self-consistent) effective field is given by B(r, t) = − dr dω ρ m (r , t)h m (r , ω , t)J(|r − r |)s [29]. Before we discuss the impact of the full ratchet potential on the dynamics of the system we briefly recall the phase behavior of the "bulk" two-dimensional binary mixture with V ext m = 0 (for details see Ref.…”

“…-Our model system consists of a binary mixture of Brownian particles confined to a two-dimensional substrate, where one species is magnetic (m) and the other one is non-magnetic (n) [29,30] (for an experimental counterpart see, e.g., Ref. [28]).…”