SciPost Submission Page
Reentrant supersolidity
by Leo Radzihovsky
Submission summary
| Authors (as registered SciPost users): | Leo Radzihovsky |
| Submission information | |
|---|---|
| Preprint Link: | https://arxiv.org/abs/2311.04266v1 (pdf) |
| Date submitted: | 2023-12-14 10:17 |
| Submitted by: | Radzihovsky, Leo |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
| Approach: | Theoretical |
Abstract
A "supersolid" -- a crystal that exhibits an off-diagonal long-range order and a superflow -- has been a subject of much research since its first proposal [Andreev and Lifshitz 1969], but has not been realized as a ground state of short-range interacting bosons in a continuum. In this note I point out a simple and generic mechanism for a thermally-driven reentrant supersolidity, and discuss challenges of experimental realization of this idea. In the limit of bosons in a periodic potential, this mechanism reduces to a {\em reentrant} low-temperature normal-superfluid transition, that should be accessible to simulations and in current experiments on bosonic atoms in an optical periodic potential.
Current status:
Reports on this Submission
Report
This paper considers defects in solids forming a Bose condensate. The defects in solids interact with long range strain induced interaction, an example of which is that between dislocations in the Kosterlitz Thouless work. These interactions can lead to many new interesting phases. My recollection is that Franz Wegner is one of the authors who has considered this.
This paper has not included any interaction between defects and I cannot recommend publication.
Report
The author argues that there can exist a phenomenon of reentrant superfluidity and supersolidity emerging on top of the insulating ground state—either a crystal or a Mott insulator—in a certain interval of temperatures. The proposed scenario is based on a favorable competition between the increasing density of thermally activated bosonic excitation and decreasing de Broglie wavelength.
If the density of excitation wins, a superfluid transition takes place. The scenario appears to be free from logical contradictions. I believe that this work might stimulate an interest of the community to the problem and recommend the manuscript for publication.
However, I have a remark concerning the putative phase boundary shown in Fig. 2. In a standard case, the quantum phase transition from superfluid (SF) to Mott insulator (MI) is in the (d+1)-dimensional U(1) universality class. Corresponding emergent universal field theory implies that, in the vicinity of the quantum critical point, the reentrant behavior is guaranteed to be absent. The author has to address this fact and either correct the shape of the curve in Fig. 2 or explain why there is a possibility of a non-standard SF-MI criticality implied by the shape of the curve Fig. 2.
Author: Leo Radzihovsky on 2024-01-26 [id 4287]
(in reply to Report 1 on 2024-01-13)Dear Editor,
Thank you for sharing the referee reports.
Referee 1's review is thorough, positive, supportive, and constructive. There is only one constructively critical comment by the referee on the nature of Fig. 2 given that the SF-MI transition is of well-understood d+1 universality class with a well studied nonzero T boundary. It is pointed out by the referee that the conventional shape of the Tc(g) boundary does not display the scenario that I illustrated (as a possibility) in Fig. 2. I did not appreciate that the orientation of this critical boundary is fully universal in contrast to an insightful suggestion by the referee. This sounds quite plausible, but I am not fully convinced of it and need to think more about it. Thus, if indeed the shape of the critical boundary is fully universal, then the scenario (while still viable in other universality classes) is not realized in SF-MI transition. I thank the referee for his/her insightful comment on this point and will need to think carefully about this comments before resubmitting.
Referee 2's report is brief and quite shallow and sloppy (felt like the referee spent on the order of 5 mins reading the draft). He/she also brings up irrelevant nature of dislocation defects that (although indeed strongly interacting) are irrelevant to the key arguments of the manuscript. The key defects are atomic vacancies /interstitials for the supersolid and gapped quasi-particles (atoms) in the MI case. These defects are short-range interacting and one can consider a limit (as a zeroth order approach to my qualitative argument) in which they are dilute and therefore weakly interacting. Per above comments, this report is not thoughtful enough to warrant any further response.