arXiv subjects

Andrey V. Chubukov

Publications and source records attributed to Andrey V. Chubukov.

At least 145 records · Page 8Linked to original sources

arXiv2010

Electron self-energy near a nematic quantum critical point

We consider an isotropic Fermi liquid in two dimensions near the n=2 Pomeranchuk instability in the charge channel. The order parameter is a quadrupolar stress tensor with two polarizations, longitudinal and transverse to the quadrupolar momentum tensor. Longitudinal and transverse bosonic modes are characterized by dynamical exponents z_parallel=3 and z_perp=2, respectively. Previous studies have found that such a system exhibits multiscale quantum criticality with two different energy scales omega ~ xi^{-z_{parallel,perp}}, where xi is the correlation length. We study the impact of the multiple energy scales on the electron Green function. The interaction with the critical z_parallel =3 mode is known to give rise to a local self-energy that develops a non-Fermi liquid form, Sigma(omega) ~ omega^{2/3} for frequencies larger than the energy scale omega ~ xi^{-3}. We find that the exchange of transverse z_perp=2 fluctuations leads to a logarithmically singular renormalizations of the quasiparticle residue Z and the vertex Gamma. We derive and solve renormalization group equations for the flow of Z and Gamma and show that the system develops an anomalous dimension at the nematic quantum-critical point (QCP). As a result, the spectral function at a fixed omega and varying k has a non-Lorentzian form. Away from the QCP, we find that the flow of Z is cut at the energy scale omega_{FL} ~ xi^{-1}, associated with the z=1 dynamics of electrons. The z_perp=2 energy scale, omega ~ xi^{-2}, affects the flow of Z only if one includes into the theory self-interaction of transverse fluctuations.

cond-mat.str-el↗

arXiv2010

Quantum-critical pairing with varying exponents

We analyse the onset temperature T_p for the pairing in cuprate superconductors at small doping, when tendency towards antiferromagnetism is strong. We consider the model of Moon and Sachdev (MS), which assumes that electron and hole pockets survive in a paramagnetic phase. Within this model, the pairing between fermions is mediated by a gauge boson, whose propagator remains massless in a paramagnet. We relate the MS model to a generic γ-model of quantum-critical pairing with the pairing kernel λ(Ω) \propto 1/Ω^γ. We show that, over some range of parameters, the MS model is equivalent to the γ-model with γ=1/3 (λ(Ω) \propto Ω^{-1/3}). We find, however, that the parameter range where this analogy works is bounded on both ends. At larger deviations from a magnetic phase, the MS model becomes equivalent to the γ-model with varying γ>1/3, whose value depends on the distance to a magnetic transition and approaches γ=1 deep in a paramagnetic phase. Very near the transition, the MS model becomes equivalent to the γ-model with varying γ<1/3. Right at the magnetic QCP, the MS model is equivalent to the γ-model with γ=0+ (λ(Ω) \propto \log Ω), which is the model for color superconductivity. Using this analogy, we verified the formula for T_c derived for color superconductivity.

cond-mat.supr-con↗

arXiv2010

Pseudogap in underdoped cuprates and spin-density-wave fluctuations

We analyze fermionic spectral function in the spin-density-wave (SDW) phase of quasi-2D cuprates at small but finite T. We use a non-perturbative approach and sum up infinite series of thermal self-energy terms, keeping at each order nearly-divergent (T/J) |log epsilon| terms, where epsilon is a deviation from a pure two-dimensionality, and neglecting regular T/J corrections. We show that, as SDW order decreases, the spectral function in the antinodal region acquires peak/hump structure: the coherent peak position scales with SDW order parameter, while the incoherent hump remains roughly at the same scale as at T=0 when SDW order is the strongest. We identify the hump with the pseudogap observed in ARPES and argue that the presence of coherent excitations at low energies gives rise to magneto-oscillations in an applied field. We show that the same peak/hump structure appears in the density of states and in optical conductivity.

cond-mat.str-el↗

arXiv2010

Universal and non-universal renormalizations in Fermi liquids

We discuss an interplay between the Fermi-liquid (FL) theory and diagrammatic perturbative approach to interacting Fermi systems. In the FL theory for Galilean-invariant systems, mass renormalization $m^*/m$ comes exclusively from fermions at the Fermi surface. We show that in a diagrammatic perturbation theory the same result for $m^*/m$ comes from fermions both at and away from the Fermi surface. The equivalence of the FL and pertubative approaches is based on a particular relation between self-energy contributions from high- and low-energy fermions. We argue that care has to be exercised in the renormalization group approach to a FL in order not to miss the high-energy contribution to $m^*/m$. As particular examples, we discuss $m^*/m$ and the quasiparticle residue $Z$ for 2D and 3D systems with both SU(2) and SU(N) symmetries, and with a short-range interaction. We derive an expression for the anisotropic part of the Fermi-liquid vertex in the large-$N$ limit of the SU(N) case.

cond-mat.str-el↗

arXiv2010

Optical Integral and Sum Rule Violation

The purpose of this work is to investigate the role of the lattice in the optical Kubo sum rule in the cuprates. We compute conductivities, optical integrals W, and ΔW between superconducting and normal states for 2-D systems with lattice dispersion typical of the cuprates for four different models -- a dirty BCS model, a single Einstein boson model, a marginal Fermi liquid model, and a collective boson model with a feedback from super-conductivity on a collective boson. The goal of the paper is two-fold. First, we analyze the dependence of W on the upper cut-off w_c placed on the optical integral because in experiments W is measured up to frequencies of order bandwidth. For a BCS model, the Kubo sum rule is almost fully reproduced at w_c equal to the bandwidth. But for other models only 70%-80% of Kubo sum rule is obtained up to this scale and even less so for ΔW, implying that the Kubo sum rule has to be applied with caution. Second, we analyze the sign of ΔW. In all models we studied ΔW is positive at small w_c, then crosses zero and approaches a negative value at large w_c, i.e. the optical integral in a superconductor is smaller than in a normal state. The point of zero crossing, however, increases with the interaction strength and in a collective boson model becomes comparable to the bandwidth at strong coupling. We argue that this model exhibits the behavior consistent with that in the cuprates.

cond-mat.str-el↗

arXiv2009

Magnetic degeneracy and hidden metallicity of the spin density wave state in ferropnictides

We analyze spin density wave (SDW) order in iron-based superconductors and electronic structure in the SDW phase. We consider an itinerant model for Fe-pnictides with two hole bands centered at $(0,0)$ and two electron bands centered at $(0,π)$ and $(π,0)$ in the unfolded BZ. A SDW order in such a model is generally a combination of two components with momenta $(0,π)$ and $(π,0)$, both yield $(π,π)$ order in the folded zone. Neutron experiments, however, indicate that only one component is present. We show that $(0,π)$ or $(π,0)$ order is selected if we assume that only one hole band is involved in the SDW mixing with electron bands. A SDW order in such 3-band model is highly degenerate for a perfect nesting and hole-electron interaction only, but we show that ellipticity of electron pockets and interactions between electron bands break the degeneracy and favor the desired $(0,π)$ or $(π,0)$ order. We further show that stripe-ordered system remains a metal for arbitrary coupling. We analyze electronic structure for parameters relevant to the pnictides and argue that the resulting electronic structure is in good agreement with ARPES experiments. We discuss the differences between our model and $J_1-J_2$ model of localized spins.

cond-mat.supr-con↗

arXiv2009

Spin conservation and Fermi liquid near a ferromagnetic quantum critical point

We propose a new low-energy theory for itinerant fermions near a ferromagnetic quantum critical point. We show that the full low-energy model includes, in addition to conventional interaction via spin fluctuations, another type of interaction, whose presence is crucial for the theory to satisfy SU(2) spin conservation. We demonstrate the consistency between a loop-wise expansion and a Fermi liquid description for the full model. We further show that, prior to the ferromagnetic instability, the system develops a Pomeranchuk-type instability into a state with zero magnetization but with p-wave deformations of the Fermi surfaces of spin-up and -down electrons (a spin nematic).

cond-mat.str-el↗

arXiv2009

Fermi liquid near Pomeranchuk quantum criticality

We analyze the behavior of an itinerant Fermi system near a charge nematic(n=2) Pomeranchuk instability in terms of the Landau Fermi liquid (FL) theory. The main object of our study is the fully renormalized vertex function $ΓΩ$, related to the Landau interaction function. We derive $Γ^Ω$ for a model case of the long-range interaction in the nematic channel. Already within the Random Phase Approximation (RPA), the vertex is singular near the instability. The full vertex, obtained by resumming the ladder series composed of the RPA vertices, differs from the RPA result by a multiplicative renormalization factor $Z_Γ$, related to the single-particle residue $Z$ and effective mass renormalization $m^*/m$. We employ the Pitaevski-Landau identities, which express the derivatives of the self-energy in terms of $Γ^Ω$, to obtain and solve a set of coupled non-linear equations for $Z_Γ$, $Z$, and $m^*/m$. We show that near the transition the system enters a critical FL regime, where $Z_Γ\sim Z \propto (1 + g_{c,2})^{1/2}$ and $m^*/m \approx 1/Z$, where $g_{c,2}$ is the $n=2$ charge Landau component which approaches -1 at the instability. We construct the Landau function of the critical FL and show that all but $g_{c,2}$ Landau components diverge at the critical point. We also show that in the critical regime the one-loop result for the self-energy $Σ(K) \propto \int dP G(P) D (K-P)$ is asymptotically exact if one identifies the effective interaction $D$ with the RPA form of $Γ^Ω$.

cond-mat.str-el↗

arXiv2009

Theory of Raman response in a superconductor with extended s-wave symmetry: Application to Fe-pnictides

We argue that Raman study of Fe-pnictides is a way to unambiguously distinguish between various superconducting gaps proposed for these materials. We show that $A_{1g}$ Raman intensity has a true resonance peak below $2Δ$ for extended s-wave superconducting gap, $Δ(\mathbf{k}) = Δ(\cos k_x + \cos k_y)/2$ in the folded Brillouin zone. No such peak emerges for a pure s-wave gap, a d-wave gap, and another extended s-wave gap with $Δ(\mathbf{k}) = Δ\cos{\frac{k_x}{2}} \cos{\frac{k_y}{2}}$ proposed by several groups.

cond-mat.supr-con↗

arXiv2009

Fermionic propagators for 2D systems with singular interactions

We analyze the form of the fermionic propagator for 2D fermions interacting with massless overdamped bosons. Examples include a nematic and Ising ferromagnetic quantum-critical points, and fermions at a half-filled Landau level. Fermi liquid behavior in these systems is broken at criticality by a singular self-energy, but the Fermi surface remains well defined. These are strong-coupling problems with no expansion parameter other than the number of fermionic species, N. The two known limits, N >>1 and N=0 show qualitatively different behavior of the fermionic propagator G(ε_k, ω). In the first limit, G(ε_k, ω) has a pole at some ε_k, in the other it is analytic. We analyze the crossover between the two limits. We show that the pole survives for all N, but at small N it only exists in a range O(N^2) near the mass shell. At larger distances from the mass shell, the system evolves and G(ε_k, ω) becomes regular. At N=0, the range where the pole exists collapses and G(ε_k, ω) becomes regular everywhere.

cond-mat.str-el↗

arXiv2008

Magnetic resonance in the cuprates -- exciton, plasmon, or π-mode

We re-analyzed the issue whether the resonance peak observed in neutron scattering experiments on the cuprates is an exciton, a π-resonance, or a magnetic plasmon. We considered a toy model with on-cite Hubbard U and nearest-neighbor interactions in both charge and spin channels. We found that the resonance is predominantly an exciton, even if magnetic interaction is absent and d-wave pairing originates from attractive density-density interaction. Our results indicate that one cannot distinguish between spin and charge-mediated pairing interactions by just looking at the resonance peak in the dynamic spin susceptibility.

cond-mat.str-el↗

arXiv2008

Nonanalytic paramagnetic response of itinerant fermions away and near a ferromagnetic quantum phase transition

We study nonanalytic paramagnetic response of an interacting Fermi system both away and in the vicinity of a ferromagnetic quantum phase transition (QCP). Previous studies found that (i) the spin susceptibility scales linearly with either the temperature $T$ or magnetic field H in the weak-coupling regime; (ii) the interaction in the Cooper channel affects this scaling via logarithmic renormalization of prefactors of the $T$, |H| terms, and may even reverse the signs of these terms at low enough energies. We show that Cooper renormalization becomes effective only at very low energies, which get even smaller near a QCP. However,even in the absence of such renormalization, generic (non-Cooper) higher-order processes may also inverse the sign of T,|H| scaling. We derive the thermodynamic potential as a function of magnetization and show that it contains, in addition to regular terms, a non-analytic |M|^3 term, which becomes M^4/T at finite $T$. We show that regular (M^2, M^4, ...) terms originate from fermions with energies of order of the bandwidth, while the non-analytic term comes from low-energy fermions. We consider the vicinity of a ferromagnetic QCP by generalizing the Eliashberg treatment of the spin-fermion model to finite magnetic field, and show that the |M|^3 term crosses over to a non-Fermi-liquid form |M|^{7/2} near a QCP. The prefactor of the |M|^{7/2} term is negative, which indicates that the system undergoes a first-order rather than a continuous transition to ferromagnetism. We compare two scenarios of the breakdown of a continuous QCP: a first-order instability and a spiral phase. In a model with a long-range interaction in the spin channel, we show that the first-order transition occurs before the spiral instability.

cond-mat.str-el↗

arXiv2008

Quantum stabilization of 1/3-magnetization plateau in Cs_2CuBr_4

We consider the phase diagram of a spatially anisotropic 2D triangular antiferromagnet in a magnetic field. Classically, the ground state is umbrella-like for all fields, but we show that the quantum phase diagram is much richer and contains a 1/3 magnetization plateau, two commensurate planar states, two incommensurate chiral umbrella phases, and, possibly, a planar state separating the two chiral phases. Our analysis sheds light on several recent experimental findings for the spin-1/2 system Cs_2CuBr_4.

cond-mat.str-el↗

arXiv2008

Non-analytic behavior of 2D itinerant ferromagnets

We consider an ordered ferromagnet in the vicinity of a T=0 transition into a paramagnet. We show that the free energy and the transverse and longitudinal static susceptibilities contain non-analyticities which destroy a continuous second-order transition. Depending on the parameters, the transition either becomes first-order, or occurs via an intermediate spiral phase.

cond-mat.str-el↗

arXiv2008

A test of the g-ology model for one-dimensional interacting Fermi systems

Bosonization predicts that the specific heat, C(T), of a one-dimensional interacting Fermi system is a sum of the specific heats of free collective charge and spin excitations, plus the term with the running backscattering amplitude which flows to zero logarithmically with decreasing T. We verify whether this result is reproduced in the g-ology model. Of specific interest are the anomalous terms in C(T) that depend on the bare backscattering amplitude. We show that these terms can be incorporated into a renormalized spin velocity. We do this by proving the equivalence of the results for C(T) obtained within the g-ology model and by bosonization with velocities obtained by the numerical solution of the Bethe-ansatz equations for the Hubbard model.

cond-mat.str-el↗

arXiv2007

Specific heat of a one-dimensional interacting Fermi system: the role of anomalies

We re-visit the issue of the temperature dependence of the specific heat C(T) for interacting fermions in 1D. The charge component C_c(T) scales linearly with T, but the spin component C_s (T) displays a more complex behavior with T as it depends on the backscattering amplitude, g_1, which scales down under RG transformation and eventually behaves as g_1 (T) \sim 1/\log T. We show, however, by direct perturbative calculations that C_s(T) is strictly linear in T to order g^2_1 as it contains the renormalized backscattering amplitude not on the scale of T, but at the cutoff scale set by the momentum dependence of the interaction around 2k_F. The running amplitude g_1 (T) appears only at third order and gives rise to an extra T/\log^3 T term in C_s (T). This agrees with the results obtained by a variety of bosonization techniques. We also show how to obtain the same expansion in g_1 within the sine-Gordon model.

cond-mat.str-el↗

arXiv2007

Signatures of non-monotonic d-wave gap in electron-doped cuprates

We address the issue whether the data on optical conductivity and Raman scattering in electron-doped cuprates below $T_c$ support the idea that the $d-$wave gap in these materials is non-monotonic along the Fermi surface. We calculate the conductivity and Raman intensity for elastic scattering, and find that a non-monotonic gap gives rise to several specific features in optical and Raman response functions. We argue that all these features are present in the experimental data on Nd$_{2-x}$Ce$_{x}$CuO$_4$ and Pr$_{2-x}$Ce$_{x}$CuO$_4$ compounds.

cond-mat.supr-con↗

arXiv2007

Cooper channel and the singularities in the thermodynamics of a Fermi liquid

We analyze how the logarithmic renormalizations in the Cooper channel affect the non-analytic temperature dependence of the specific heat coefficient γ(T) - γ(0)=A(T)T in a 2D Fermi liquid. We show that A(T) is expressed exactly in terms of the fully renormalized backscattering amplitude which includes the renormalization in the Cooper channel. In contrast to the 1D case, both charge and spin components of the backscattering amplitudes are subject to this renormalization. We show that the logarithmic renormalization of the charge amplitude vanishes for a flat Fermi surface, when the system becomes effectively one-dimensional.

cond-mat.str-el↗