arXiv subjects
Publications and source records attributed to V. Gurarie.
This paper examines the problem of molecule production in an atomic fermionic gas close to an s-wave Feshbach resonance by means of a magnetic field sweep through the resonance. The density of molecules at the end of the process is derived for narrow resonance and slow sweep.
We consider the behavior of an interacting many particle system under slow external driving -- a many body generalization of the Landau-Zener paradigm. We find that a conspiracy of interactions and driving leads to physics profoundly different from that of the single particle limit: for practically all values of the driving rate the particle distributions in Hilbert space are very broad, a phenomenon caused by a strong amplification of quantum fluctuations in the driving process. These fluctuations are 'non-adiabatic' in that even at very slow driving it is exceedingly difficult to push the center of the distribution towards the limit of full ground state occupancy. We obtain these results by a number of complementary theoretical approaches, including diagrammatic perturbation theory, semiclassical analysis, and exact diagonalization.
We study the stability of the paired fermionic p-wave superfluid made out of identical atoms all in the same hyperfine state close to a p-wave Feshbach resonance. First we reproduce known results concerning the lifetime of a 3D superfluid, in particular, we show that it decays at the same rate as its interaction energy, thus precluding its equilibration before it decays. Then we proceed to study its stability in case when the superfluid is confined to 2D by means of an optical harmonic potential. We find that the relative stability is somewhat improved in 2D in the BCS regime, such that the decay rate is now slower than the appropriate interaction energy scale. The improvement in stability, however, is not dramatic and one probably needs to look for other mechanisms to suppress decay to create a long lived 2D p-wave fermionic superfluid.
We consider the evolution of a two-level system coupled to a photon field initially in a coherent state, as the energy of the two-level system is linearly varied through resonance with the photon field. At a fixed time after the resonance, the amplitude of the photon field is found to show a collapse and subsequent revivals as a function of rate of energy variation. Including decay of the photon field, we find that the observation of such collapse and revivals is near the technological limit of current cavity QED experiments but should be achievable.
We examine bosons hopping on a one-dimensional lattice in the presence of a random potential at zero temperature. Bogoliubov excitations of the Bose-Einstein condensate formed under such conditions are localized, with the localization length diverging at low frequency as $\ell(ω)\sim 1/ω^α$. We show that the well known result $α=2$ applies only for sufficiently weak random potential. As the random potential is increased beyond a certain strength, $α$ starts decreasing. At a critical strength of the potential, when the system of bosons is at the transition from a superfluid to an insulator, $α=1$. This result is relevant for understanding the behavior of the atomic Bose-Einstein condensates in the presence of random potential, and of the disordered Josephson junction arrays.
We formulate and approximately solve a specific many-body generalization of the Landau-Zener problem. Unlike with the single particle Landau-Zener problem, our system does not abide in the adiabatic ground state, even at very slow driving rates. The structure of the theory suggests that this finding reflects a more general phenomenon in the physics of adiabatically driven many particle systems. Our solution can be used to understand, for example, the behavior of two-level systems coupled to an electromagnetic field, as realized in cavity QED experiments.
We study theoretically a dilute gas of identical fermions interacting via a p-wave resonance. We show that, depending on the microscopic physics, there are two distinct regimes of p-wave resonant superfluids, which we term "weak" and "strong". Although expected naively to form a BCS-BEC superfluid, a strongly-resonant p-wave superfluid is in fact unstable towards the formation of a gas of fermionic triplets. We examine this instability and estimate the lifetime of the p-wave molecules due to the collisional relaxation into triplets. We discuss consequences for the experimental achievement of p-wave superfluids in both weakly- and strongly-resonant regimes.
We study a gas of fermions undergoing a wide resonance s-wave BCS-BEC crossover, in the BEC regime at zero temperature. We calculate the chemical potential and the speed of sound of this Bose-condensed gas, as well as the condensate depletion, in the low density approximation. We discuss how higher order terms in the low density expansion can be constructed. We demonstrate that the standard BCS-BEC gap equation is invalid in the BEC regime and is inconsistent with the results obtained here. We indicate how our theory can in principle be extended to nonzero temperature. The low density approximation we employ breaks down in the intermediate BCS-BEC crossover region. Hence our theory is unable to predict how the chemical potential and the speed of sound evolve once the interactions are tuned towards the BCS regime. As a part of our theory, we derive the well known result for the bosonic scattering length diagrammatically and check that there are no bound states of two bosons.
We present a theory of a degenerate atomic Fermi gas, interacting through a narrow Feshbach resonance, whose position and therefore strength can be tuned experimentally, as demonstrated recently in ultracold trapped atomic gases. The distinguishing feature of the theory is that its accuracy is controlled by a dimensionless parameter proportional to the ratio of the width of the resonance to Fermi energy. The theory is therefore quantitatively accurate for a narrow Feshbach resonance. In the case of a narrow s-wave resonance, our analysis leads to a quantitative description of the crossover between a weakly-paired BCS superconductor of overlapping Cooper pairs and a strongly-paired molecular Bose-Einstein condensate of diatomic molecules. In the case of pairing via a p-wave resonance, that we show is always narrow for a sufficiently low density, we predict a detuning-temperature phase diagram, that in the course of a BCS-BEC crossover can exhibit a host of thermodynamically-distinct phases separated by quantum and classical phase transitions. For an intermediate strength of the dipolar anisotropy, the system exhibits a p_x + i p_y paired superfluidity that undergoes a topological phase transition between a weakly-coupled gapless ground state at large positive detuning and a strongly-paired fully-gapped molecular superfluid for a negative detuning. In two dimensions the former state is characterized by a Pfaffian ground state exhibiting topological order and non-Abelian vortex excitations familiar from fractional quantum Hall systems.
We discuss fermionic zero modes in the two-dimensional chiral p-wave superconductors. We show quite generally, that without fine-tuning, in a macroscopic sample there is only one or zero of such Majorana-fermion modes depending only on whether the total vorticity of the order parameter is odd or even, respectively. As a special case of this, we find explicitly the one zero mode localized on a single odd-vorticity vortex, and show that, in contrast, zero modes are absent for an even-vorticity vortex. One zero mode per odd vortex persists, within an exponential accuracy, for a collection of well-separated vortices, shifting to finite E or -E energies as two odd vortices approach. These results should be useful for the demonstration of the non-Abelian statistics that such zero-mode vortices are expected to exhibit, and for their possible application in quantum computation.
We present a study of a gas of bosons confined in one dimension with Feshbach resonant interactions, at zero temperature. Unlike the gas of one dimensional bosons with non-resonant interactions, which is known to be equivalent to a system of interacting spinless fermions and can be described using the Luttinger liquid formalism, the resonant gas possesses novel features. Depending on its parameters, the gas can be in one of three possible regimes. In the simplest of those, it can still be described by the Luttinger liquid theory, but its Fermi momentum cannot be larger than a certain cutoff momentum dependent on the details of the interactions. In the other two regimes, it is equivalent to a Luttinger liquid at low density only. At higher densities its excitation spectrum develops a minimum, similar to the roton minimum in helium, at momenta where the excitations are in resonance with the Fermi sea. As the density of the gas is increased further, the minimum dips below the Fermi energy, thus making the ground state unstable. At this point the standard ground state gets replaced by a more complicated one, where not only the states with momentum below the Fermi points, but also the ones with momentum close to that minimum, get filled, and the excitation spectrum develops several branches. We are unable so far to study this new regime in detail due to the lack of the appropriate formalism.
We study a single-species polarized Fermi gas tuned across a narrow p-wave Feshbach resonance. We show that in the course of a BEC-BCS crossover the system can undergo a magnetic field-tuned quantum phase transition from a p_x-wave to a p_x+i p_y-wave superfluid. The latter state, that spontaneously breaks time-reversal symmetry, furthermore undergoes a topological p_x+ i p_y to p_x+ i p_y transition at zero chemical potential μ. In two-dimensions, for μ>0 it is characterized by a Pfaffian ground state exhibiting topological order and non-Abelian excitations familiar from fractional quantum Hall systems.
We show that the density of states of random wave equations, normalized by the square of the frequency, has a peak - sometimes narrow and sometimes broad - in the range of wave vectors between the disorder correlation length and the interatomic spacing. The results of this letter may be relevant for understanding vibrational spectra and light propagation in disordered solids.
We develop techniques which allow us to calculate the spectra of pinned charge density waves with background current.
We examine two-dimensional conformal field theories (CFTs) at central charge c=0. These arise typically in the description of critical systems with quenched disorder, but also in other contexts including dilute self-avoiding polymers and percolation. We show that such CFTs must in general possess, in addition to their stress energy tensor T(z), an extra field whose holomorphic part, t(z), has conformal weight two. The singular part of the Operator Product Expansion (OPE) between T(z) and t(z) is uniquely fixed up to a single number b, defining a new `anomaly' which is a characteristic of any c=0 CFT, and which may be used to distinguish between different such CFTs. The extra field t(z) is not primary (unless b=0), and is a so-called `logarithmic operator' except in special cases which include affine (Kac-Moody) Lie-super current algebras. The number b controls the question of whether Virasoro null-vectors arising at certain conformal weights contained in the c=0 Kac table may be set to zero or not, in these nonunitary theories. This has, in the familiar manner, implications on the existence of differential equations satisfied by conformal blocks involving primary operators with Kac-table dimensions. It is shown that c=0 theories where t(z) is logarithmic, contain, besides T and t, additional fields with conformal weight two. If the latter are a fermionic pair, the OPEs between the holomorphic parts of all these conformal weight-two operators are automatically covariant under a global U(1|1) supersymmetry. A full extension of the Virasoro algebra by the Laurent modes of these extra conformal weight-two fields, including t(z), remains an interesting question for future work.
We consider a two-species degenerate Fermi gas coupled by a diatomic Feshbach resonance. We show that the resulting superfluid can exhibit a form of coherent BEC-to-BCS oscillations in response to a nonadiabatic change in the system's parameters, such as for example a sudden shift in the position of the Feshbach resonance. In the narrow resonance limit, the resulting soliton-like collisionless dynamics can be calculated analytically. In equilibrium the thermodynamics can be accurately computed across the full range of BCS-BEC crossover, with corrections controlled by the ratio of the resonance width to the Fermi energy.
We study the spectral and transport properties of magnons in a model of a disordered magnet called Mattis glass, at vanishing average magnetization. We find that in two dimensional space, the magnons are localized with the localization length which diverges as a power of frequency at small frequencies. In three dimensional space, the long wavelength magnons are delocalized. In the delocalized regime in 3d (and also in 2d in a box whose size is smaller than the relevant localization length scale) the magnons move diffusively. The diffusion constant diverges at small frequencies. However, the divergence is slow enough so that the thermal conductivity of a Mattis glass is finite, and we evaluate it in this paper. This situation can be contrasted with that of phonons in structural glasses whose contribution to thermal conductivity is known to diverge (when inelastic scattering is neglected).
This paper studies the 1D pressureless turbulence (the Burgers equation). It shows that reliable numerics in this problem is very easy to produce if one properly discretizes the Burgers equation. The numerics it presents confirms the 7/2 power law proposed for probability of observing large negative velocity gradients in this problem. It also suggests that the entire probability function for the velocity gradients could be universal, perhaps in some approximate sense. In particular, the probability that the velocity gradient is negative appears to be $p \approx 0.21 \pm 0.01$ irrespective of the details of the random force. Finally, I speculate that the theory initially proposed by Polyakov, with a particular value of the "anomaly" parameter, may indeed be exact, at least as far as velocity gradients are concerned.