Phase transitions in one matrix models

doi:10.1016/0370-2693(90)90723-J

Physics Letters B

Volume 251, Issue 3, 22 November 1990, Pages 388-392

https://doi.org/10.1016/0370-2693(90)90723-J Get rights and content

Abstract

We discuss in detail the phase diagram of a matrix model with a sixth degree potential. We compare with some of the recent results on k=2 and k=3 multicritical models. The k=2 universality class seems to be in the middle of a “wrong phase”.

References (10)

E. Brézin et al.
Commun. Math. Phys.
(1978)
O. Alvarez and P. Windey, LPTHE (Paris) preprint (February...
M. Douglas, N. Sieberg and S. Shenker, Rutgers University preprint (April...
L. Mollinari and E. Montaldi, INFN (Milano) preprint...
D.J. Gross et al.
PUPT-1159
(December 1989)

There are more references available in the full text version of this article.

Cited by (20)

Stokes phenomena and non-perturbative completion in the multi-cut two-matrix models
2012, Nuclear Physics B
The Stokes multipliers in the matrix models are invariants in the string-theory moduli space and related to the D-instanton chemical potentials. They not only represent non-perturbative information but also play an important role in connecting various perturbative string theories in the moduli space. They are a key concept to the non-perturbative completion of string theory and also expected to imply some remnant of strong coupling dynamics in M theory. In this paper, we investigate the non-perturbative completion problem consisting of two constraints on the Stokes multipliers. As the first constraint, Stokes phenomena which realize the multi-cut geometry are studied in the $Z_{k}$ symmetric critical points of the multi-cut two-matrix models. Sequence of solutions to the constraints are obtained in general k-cut critical points. A discrete set of solutions and a continuum set of solutions are explicitly shown, and they can be classified by several constrained configurations of the Young diagram. As the second constraint, we discuss non-perturbative stability of backgrounds in terms of the Riemann–Hilbert problem. In particular, our procedure in the 2-cut $(1, 2)$ case (pure-supergravity case) completely fixes the D-instanton chemical potentials and results in the Hastings–McLeod solution to the Painlevé II equation. It is also stressed that the Riemann–Hilbert approach realizes an off-shell background independent formulation of non-critical string theory.
Symmetry breaking in the double-well hermitian matrix models
1993, Nuclear Physics, Section B
We study symmetry breaking in Z₂ symmetric large N matrix models. In the planar approximation for both the symmetric double-well φ⁴ model and the symmetric Penner model, we find there is an infinite family of broken symmetry solutions characterized by different sets of recursion coefficients R_n and S_n that all lead to identical free energies and eigenvalue densities. These solutions can be parameterized by an arbitrary angle θ(x), for each value of x = n/N<1. In the duoble scaling limit, this class reduces to a smaller family of solutions with distinct free energies already at the torus level. For the double-well φ⁴ theory the double scaling string equations are parameterized by a conserved angular momentum parameter in the range 0 ⩽ l < ∞ and a single arbitrary U(1) phase angle.
D = 0 matrix model as conjugate field theory
1993, Nuclear Physics, Section B
The D = 0 matrix model is reformulated as a nonlocal quantum field theory in two dimensions, in which the interactions occur on the one-dimensional line of hermitian matrix eigenvalues. The field can be thought of as a fluctuation in the potential V, and is conjugate to the density of matrix eigenvalues which appears in the Jevicki collective field theory. The classical solution of the field equation is either unique or labeled by a discrete index. Such a solution corresponds to the Dyson sea modified by a entropy term. The modification smoothes the sea edges, and interpolates between different eigenvalue bands for multiple-well potentials. Our classical eigenvalue density contains nonplanar effects, and satisfies a local nonlinear Schrödinger equation with similarities to the fifth-time D = 1 reformulation. The quantum fluctuations about a classical solution are computable, and the IR and UV divergences are manifestly removed to all orders. The quantum corrections greatly simplify in the double-scaling limit, and include both string-perturbative and nonperturbative effects. The latter are unambiguous for V bounded from below, and even for k = 2 can be compared with the various nonperturbative definitions of these theories proposed in the literature. The entire double-scaled, string-perturbative series for the free energy resides in two terms: the classical term, plus the one-loop (semiclassical) functional determinant about the classical solution.
A quantum mechanical framework for pure two-dimensional gravity
1991, Physics Letters B
A natural quantum mechanical framework for the study of pure 2D quantum gravity arises from the observation that a Fokker-Planck potential fixes the critical behavior of the system. This clarifies the Marinari-Parisi suggestion of defining the zero-dimensional bosonic theory as the dimensional reduction of a supersymmetric one-dimensional counterpart, and can be used in all its strenght to approach the D = 0 problem.
Multiloop correlators for two-dimensional quantum gravity
1990, Physics Letters B
We find explicitly all multi-loop correlators in the complex matrix model to the leading order in $1 N$ and show that they are identical to the even part of the multi-loop correlators in the hermitean matrix model. The scaling limit for the corresponding macroscopic loop correlators is constructed and agrees with the one of the hermitean model to all orders in $1 N^{2}$ . In particular the double scaling limits of the two models will lead to identical “string equations”.
Bigeneric nonperturbative strings
1990, Physics Letters B
The zero dimensional hermitian matrix models with asymmetric potentials include a family of bigeneric (p,q) multicritical models which have nontrivial double-scaling continuum limits. These exhibit more general critical behaviors than the k multi-critical models that occur for even potentials. They are related to a simpler construction which we call the k bigeneric models. The specific heat of the kth bigeneric model coincides with that of the kth multicritical model for odd genus, but vanishes for all even genus. We provide a number of explicit examples of bigeneric string equations.

View all citing articles on Scopus

¹: Permanent address: Thinking Machines Corporation, 245 First Street, Cambridge, MA 02142, USA.

²: On leave from: Tata Institute of Fundamental Research, Homi Bhabha Road, Bombay 400 005, India.

³: Address after August 1990: Physics Department, Harvard University, Cambridge, MA 02138, USA.

View full text

Phase transitions in one matrix models

Abstract

Commun. Math. Phys.

PUPT-1159