spin glasses

spin glasses
Name	Spin glasses
Field	Condensed matter physics
Notable	David Sherrington; Scott Kirkpatrick; Giorgio Parisi; Daniel Fisher; Marc Mézard

spin glasses Spin glasses are disordered magnetic systems characterized by competing interactions, quenched randomness, and slow dynamics. Originating from studies of dilute magnetic alloys, they exhibit complex energy landscapes, nonergodicity, and anomalous thermodynamic and dynamical behavior. Research on spin glasses connects to statistical mechanics, computational complexity, and information theory.

Introduction

Spin glasses emerged from experiments on noble-metal alloys such as KONSTANTINOPOULOS, George Uhlenbeck, Pierre Curie and theoretical responses by groups including P. W. Anderson's collaborators and the Bell Labs community, leading to paradigms formulated by Samuel Edwards and P. W. Anderson. The term became central after theoretical frameworks by D. Sherrington and S. Kirkpatrick and the exact solution by G. Parisi for mean-field models. Intersections with concepts from John von Neumann-inspired computation, Alan Turing-style optimization, and Claude Shannon-style information theory expanded interest across Bell Labs, IBM Research, and university groups.

Physical Models and Definitions

Canonical models include the Edwards–Anderson model (EA) and the Sherrington–Kirkpatrick model (SK). The Edwards–Anderson model represents short-range interactions on lattices studied by groups at Cavendish Laboratory and Institute for Advanced Study; the Sherrington–Kirkpatrick model is an infinite-range mean-field model solved by Parisi. Key elements are spins on sites, random exchange couplings with ferromagnetic and antiferromagnetic signs, and quenched disorder as formalized by methods used at Princeton University and Harvard University. Mathematically, models use probability distributions introduced in work associated with Norbert Wiener's stochastic ideas and techniques developed at Los Alamos National Laboratory and École Normale Supérieure.

Theoretical Approaches and Methods

Analytical methods encompass replica theory, cavity method, and functional renormalization group. Replica symmetry breaking (RSB) pioneered by Parisi built on earlier statistical mechanics from Lev Landau and Isaak Khalatnikov and was elaborated in seminars at Scuola Normale Superiore and SISSA. The cavity method, developed by teams including M. Mézard and G. Parisi, connects to combinatorial optimization problems studied at Bell Labs and Bellcore. Rigorous probabilistic techniques arose from collaborations between mathematicians at Courant Institute and University of California, Berkeley culminating in proofs attributed to researchers linked to Fields Institute and Clay Mathematics Institute-associated programs. Numerical approaches—Monte Carlo simulations, exchange Monte Carlo, and population dynamics—are implemented in computational groups at Los Alamos National Laboratory and Lawrence Livermore National Laboratory.

Experimental Realizations and Measurements

Experimental systems include dilute alloys like copper-manganese (CuMn) investigated at Bell Labs, insulating compounds studied at Laboratoire Léon Brillouin, and artificially engineered nanomagnets produced in cleanrooms at IBM Research. Techniques to probe spin-glass behavior involve susceptibility and specific-heat measurements developed at NIST, neutron-scattering experiments at facilities such as ISIS Neutron and Muon Source and Institut Laue–Langevin, and muon spin rotation studies from groups at TRIUMF. Aging, memory, and rejuvenation phenomena were characterized in experiments led by teams at École Polytechnique and University of Cambridge, using protocols refined in collaborations with researchers from Max Planck Institute for Solid State Research.

Applications and Related Systems

Conceptual and practical applications span optimization, neural networks, and error-correcting codes. The mapping between spin-glass models and combinatorial optimization informed algorithms researched at MIT and Stanford University and inspired simulated annealing techniques associated with Kirkpatrick, Gelatt, and Vecchi. Connections to associative memory and Hopfield networks draw on work at Princeton University and Hebrew University of Jerusalem. Relations to structural glasses, protein folding experiments at Cold Spring Harbor Laboratory, and satisfiability problems studied at Bell Labs and Microsoft Research highlight interdisciplinary links.

Open Problems and Research Directions

Key open questions include the nature of the low-temperature phase in finite dimensions, the validity of replica symmetry breaking versus droplet/scaling pictures proposed in schools around University of Chicago and University of Toronto, and rigorous characterizations of dynamic heterogeneity pursued at Rutgers University and University of Cambridge. Advances in quantum annealing at D-Wave Systems and quantum Monte Carlo studies from Perimeter Institute raise questions about quantum spin glasses and their computational complexity. Experimental capabilities at large facilities such as CERN-associated centers and national neutron sources motivate new probes of nonequilibrium phenomena, while cross-disciplinary links to machine learning at DeepMind and Google Research continue to evolve.

Category:Condensed matter physics