SciPost Submission Page
Quantum quenches to the attractive onedimensional Bose gas: exact results
by Lorenzo Piroli, Pasquale Calabrese, Fabian H. L. Essler
This is not the latest submitted version.
This Submission thread is now published as
Submission summary
As Contributors:  Pasquale Calabrese · Fabian Essler · Lorenzo Piroli 
Arxiv Link:  http://arxiv.org/abs/1604.08141v2 (pdf) 
Date submitted:  20160524 02:00 
Submitted by:  Piroli, Lorenzo 
Submitted to:  SciPost Physics 
Academic field:  Physics 
Specialties: 

Approach:  Theoretical 
Abstract
We study quantum quenches to the onedimensional Bose gas with attractive interactions in the case when the initial state is an ideal onedimensional Bose condensate. We focus on properties of the stationary state reached at late times after the quench. This displays a finite density of multiparticle bound states, whose rapidity distribution is determined exactly by means of the quench action method. We discuss the relevance of the multiparticle bound states for the physical properties of the system, computing in particular the stationary value of the local pair correlation function $g_2$.
Ontology / Topics
See full Ontology or Topics database.Current status:
Submission & Refereeing History
You are currently on this page
Reports on this Submission
Anonymous Report 2 on 2016627 (Invited Report)
 Cite as: Anonymous, Report on arXiv:1604.08141v2, delivered 20160626, doi: 10.21468/SciPost.Report.3
Strengths
1. Extremely well written and free of typos.
2. Generously referenced.
3. Technical details of the computations presented in the manuscript are given
in full.
Weaknesses
1. While the computations in the manuscript are technically impressive, the discussion of the physical consequences of these computations is not as expansive as it might be.
Report
This manuscript studies the long time behaviour after a quantum quench in the LiebLiniger model, a model of a onedimensional Bose gas. The particular quench studied takes the gas from c, the strength of its twobody interaction, zero to c finite and negative (i.e. the gas is quenched into its attractive regime).
The authors use the quench action to study this quench. The needed overlaps between the eigenstates of the postquench Hamiltonian and the initial state of the gas had been determined by other authors. However some additional analysis of the overlaps in the case of states that are zero momentum nstrings by the authors was required. With the overlaps in hand, the authors are able to analyze the corresponding generalized TBA equations describing the steady state postquench. Impressively, the authors are able to arrive at an analytical solution to these equations. Finally the authors use this solution to investigate g2, the pair correlation function (which can be obtained as a derivative of the generalized free energy by c). Interestingly the authors show that the density of higher bound states is reduced by increasing the strength of the attractive interaction.
I appreciate the careful presentation of results in this paper. It reads nicely and it is good to see
that all the details of the computations are presented in a comprehensible fashion. The one critique
that I do have is the sole focus on the steady state. It would be interested to understand more about
the approach to steady state. There have been suggestions in the literature (indeed, involving
one of the authors) of unusual entanglement entropy growth after a quench in systems with bound states.
It would be interesting to know if the authors felt something like this would occur here as well. I am not necessarily asking for a detailed computation. But some remarks to this end would be appreciated.
Requested changes
1. As I indicated, I would like to see some comments and/or analysis on the approach of the gas to the steady state postquench.
Anonymous Report 1 on 201668 (Invited Report)
 Cite as: Anonymous, Report on arXiv:1604.08141v2, delivered 20160608, doi: 10.21468/SciPost.Report.2
Strengths
1 Exact results.
2 Experimentally relevant.
3 Well written, clear exposition.
4 Exceptionally good (wide) list of references.
Weaknesses
I don't see any weakness. It is true that most of the theoretical methods are already existing, but they are applied to a new situation, and there are certain steps that required new ideas.
Report
This paper derives exact solutions for a specific quench problem in the attractive LiebLiniger model. The methods are simple generalizations of already available techniques, however, there are certain innovations and all results are new. The g2 local pair correlator is calculated numerically, and it is studied in the small coupling and large coupling limits. The physical consequences of the bound states are discussed. Maybe the most interesting is that g2 does not vanish in the infinite coupling limit. Whereas this is not necessarily surprising, it is useful that exact results are derived. I believe that later the results could be compared to experiments, thus strengthening the link between theory and experiment.
Requested changes
I don't request any changes, but I have a comment to the authors. It would be certainly interesting to consider the higher local correlators as well. For example g3. Maybe the threestrings would show up there in the large coupling limit?
Author: Lorenzo Piroli on 20160706 [id 48]
(in reply to Report 1 on 20160608)
The referee states that
“It would be certainly interesting to consider the higher local correlators as well. For example g3. Maybe the threestrings would show up there in the large coupling limit?”
We agree that the calculation of g3 would be very interesting, and threestrings may well give nonvanishing contributions in the large coupling limit.
While g3 is known for general states in the repulsive LiebLiniger model, the attractive case is significantly harder to treat due to the presence of bound states and requires further development of existing methods. We think that such a calculation is possible, but definitely beyond the scope of our paper.
Author: Lorenzo Piroli on 20160706 [id 49]
(in reply to Report 2 on 20160627)To understand the approach to the stationary state after quantum quenches is without doubt a very interesting and timely problem.
However, it is well known that this is an extraordinarily difficult task to study with integrability methods in truly interacting models. Even the case of the Transverse Field Ising chain, which can be mapped to a free fermion theory, is extremely nontrivial as is clear from the works of Calabrese, Essler and Fagotti.
In the interacting case there are basically only two results available in the literature: one is for particular quenches in the sineGordon model (http://arxiv.org/pdf/1405.4813.pdf, one of us is among the authors), and the other one is a seminumerical paper ( http://arxiv.org/abs/1505.03080, one of us is among the authors ) studying the approach to the stationary state for particular observables in the LiebLinger model in the regime where no string solutions exist. The case of interest to us here is more difficult than either of these two, and studying the time evolution is manifestly beyond the scope of our work.
The referee also mention the anomalous behavior of entanglement entropy and correlation functions found in http://arxiv.org/abs/1604.03571 (of which another one of us is an author). This novel effect is due to the phenomenon of confinement, i.e. not to the presence of the bound states, but to the absence of elementary excitations. In our case, unbound particles are always present as $\rho_1$ is always different from zero. This means that in the quench studied here this phenomenon does not take place.
For the reasons given above we decided not to change the manuscript by adding comments on the finite time dynamics.