Answers are grounded in this paper and cite their sources.
We map out the ground-state phase diagram of a one-dimensional SU(N) Kondo-Heisenberg lattice model at half-filling and in the fully antisymmetric self-adjoint representation as a function of 1/N and Kondo coupling Jₖ/t. On the basis of auxiliary field quantum Monte Carlo (QMC) simulations with even N up to 8, we show that the enlarged SU(N≥4) symmetry realizes a quantum phase transition separating a valence bond solid (VBS) phase occupying a weak-coupling part of the phase diagram and the Kondo insulator (KI) state dominating in the strong Jₖ/t limit. Along the phase boundary, we always observe critical exponents that belong to a two-dimensional classical Ising universality class. We next trace the evolution of the composite fermion and spin spectral functions across the phase boundary and conclude that VBS order triggers a bond order wave state of conduction electrons, both coexisting with Kondo screening. Upon further reducing Jₖ/t, we observe that composite quasiparticles lose their spectral weight indicating that VBS order gradually liberates localized f spins from forming Kondo singlets with conduction electron spins. We contrast the QMC results with a static large-N approximation, and we show that in the limit of infinite degeneracy N→∞ the order-disorder transition becomes of first order and is accompanied by a full decoupling of conduction electrons and localized f spins. We complete our analysis by considering the limit of a vanishing Heisenberg coupling Jₕ=0, and we provide evidence that the enlarged SU(N) symmetry and low dimensionality of the Ruderman-Kittel-Kasuya-Yosida exchange interaction are not sufficient to induce VBS order in the conventional SU(N) Kondo chain, which hosts solely a KI phase. Finally, we quantify the role of increasing N in reducing the separation of charge and spin timescales specific to the KI state, and we show that the latter is adiabatically connected to a trivial band insulator captured by the large-N approximation.
No takes yet. Share an insight, caveat, or question.
Raczkowski et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: