The Ising model: Highlights and perspectives
Christof Külske · 2025
Read The Ising model: Highlights and perspectives by Christof Külske (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-05394.txt.
Introduction
Christof Kuelske gives a short, non-technical tour of the Ising model, a classic model of magnetism. His 2025 paper is based on a lecture at a world congress in Bochum in August 2024. The model is simple to state, yet it has driven a century of work in probability and mathematical physics.
Kuelske reviews the theory of phase transitions in the infinite-volume limit, where the model's sharp changes become exact. He then surveys big ideas that grew from its study: scaling, the renormalization group, and universality. Later topics include curve limits known as SLE and random symmetry breaking in disordered systems and networks. The piece serves as a compact map of triumphs and open problems.
Catalog record
| CREATOR | Christof Külske |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 49 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-05394
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Ising model: Highlights and perspectives Christof Külske Abstract We give a short non-technical introduction to the Ising model, and review some successes as well as challenges which have emerged from its study in probability and mathematical physics. This includes the infinite-volume theory of phase transitions, and ideas like scaling, renormalization group, universality, SLE, and random symmetry breaking in disordered systems and networks. This note is based on a talk given on 15 August 2024, as part of the Ising lecture during the 11th Bernoulli-IMS world congress, Bochum. AMS 2000 subject classification: 60K35, 82B20, 82B26 Key words: Ising model, spin models, Gibbs measures, phase transitions, renormalization group, SLE, disordered systems, extremal decomposition. 0 0 footnotetext: E-Mail: [email protected] ; ORCID iD: 0000-0001-9975-8329 0 0 footnotetext: https://math.ruhr-uni-bochum.de/fakultaet/arbeitsbereiche/stochastik/gruppe-kuelske/mitarbeiter/christof-kuelske/ 1 A theory for ferromagnetic ordering? Ernst Ising (1900 Cologne - 1998 Peoria, Illinois) was born into a Jewish family from Cologne who moved to Bochum, an industrial city in Ruhr area in the western part of Germany. Ernst Ising grew up in Bochum, obtained his Abitur (high-school diploma) in Bochum and went on to study physics in Göttingen, Bonn and Hamburg. In Hamburg he followed Ph.D. studies with Wilhelm Lenz (1888 Frankfurt - 1957 Hamburg) as a Ph.D. advisor. Ising later escaped the prosecution during the Nazi period via Luxemburg and settled in the US. The problem of Ising’s Ph.D. thesis was the following: Can the formalism of statistical mechanics, following Boltzmann and Gibbs, explain a phenomenon like ferromagnetism? Ferromagnetism is the physical property of a magnetic material (e.g. iron) that it can be magnetized by putting on an external magnetic field at low temperatures. If the magnetic field is then decreased to zero, the magnetization does not vanish, and one sees the so-called spontaneous magnetization . If then temperature is slowly increased, the spontaneous magnetization decreases and vanishes sharply at a particular finite critical temperature whose value depends on the material. Statistical mechanics describes equilibrium states of large systems by means of weighted averages (instead of trying to solve time-dependent evolution equations for many-particle systems in detail). The weights are depending on the individual states of all particles. They are given by exponentials of the negative total energy of the system, with a prefactor \beta>0 whose meaning is the inverse of the temperature. An excellent introductory book to the subject is [ 29 ] . 2 Definition of the Ising model It was known at the time of a young Ising that metals have a spatially regular structure, with atoms sitting on regular lattices. Based on this, the following statistical mechanics model which was later called Ising model for ferromagnetism was invented in 1920. The idea was to simplify the description of the local degrees of freedom ( local magnetic moments ) in an extreme way, and further assume only an interaction between nearest neighbors on the lattice of such magnetic moments, sitting on the sites of a lattice \mathbb{Z}^{d} . The spatial dimension d should in physical applications mostly take the values 3 or 2 , but mathematically d may also be an arbitrary integer. In the Ising model, the degrees of freedom of the (finite-volume) model are given by magnetic moments (spins) \sigma_{i}\in\{1,-1\} which are sitting at sites i\in\Lambda\subset\mathbb{Z}^{d} , where \Lambda is a very large but finite subset of the lattice. In today’s probabilistic language they are random variables on a common probability space, which are distributed according to the following Gibbs distribution . This is the probability measure depending on inverse temperature \beta>0 , and external field h\in\mathbb{R} in which the probability to see a configuration \sigma_{\Lambda}=(\sigma_{i})_{i\in\Lambda} is given by \frac{1}{Z_{\Lambda}}\exp\big(\beta\sum_{i\sim j}\sigma_{i}\sigma_{j}+h\sum_{i}\sigma_{i}\big), where the first sum is over nearest neighbors i\sim j in the finite volume \Lambda . The negative exponent is (up to constants) the energy of a configuration. Spin configurations which for most edges i\sim j align are preferred under this measure, as the product of two spins \sigma_{i}\sigma_{j} is maximal iff they are equal. Next to the lattice \mathbb{Z}^{d} , other choices of base spaces which are very interesting today are networks . Then the model is used also with other interpretations of the ”local magnetic moment” \sigma_{i} in mind, e.g. as opinion of a person i in a network. The above exponential formula for the Gibbs measure in finite volume has free boundary conditions. In principle there may (and often should) be terms connecting the sites which sit on an inside layer of the finite volume to a fixed boundary condition. These couple to the spins inside only via its values on the outside boundary layer of the volume. Such a boundary condition could for instance be chosen to be the maximal boundary condition, in which all spins on the sites on the outer boundary are equal to one. Now, does this model show a phase transition and long-range order (in which spins tend to align over long distances)? \bullet Is there a spontaneous magnetization at large \beta ? \bullet Does the model have a magnetic phase transition? \bullet What is actually a phase transition in a precise mathematical sense? \bullet What can be said about the behavior of other physical quantities? Other physical quantities of interest include the specific heat (the amount of energy needed to put into the system in order to increase temperature) and the susceptibility (the increase of the total magnetic moment of the material under increase of the external magnetic field). Ernst Ising was the first person on earth to begin to tackle such questions. He defended his Ph.D. thesis in the year of 1924 in Hamburg, in which he showed by explicit combinatorial computations that the model has no phase transition in spatial dimension d=1 . The result was published in Zeitschrift für Physik (since 1997 European Physical Journal) in the year of 1925 [ 44 ] . He also conjectured in this paper that the absence of a phase transition in the short range model would probably be true also in higher spatial dimensions, and long-range interactions would probably be necessary to create a spontaneous magnetization in which spins over long distances stay aligned. Luckily he was too pessimistic about the short-range Ising model. Independently we note that the study of long-range models (in the form of Dyson models with power-law decay of the strength of the interactions between spins at sites i and j , as a function of the distance d(i,j) ) turned out to be a beautiful story in itself which developed much later, where ferromagnetic ordering may even appear in dimension d=1 , see [ 24 , 30 , 1 ] . 3 Phase transitions and long-range order in lattice dimensions d\geq 2 Boundary condition persistence via Peierls contours and Peierls argument. Rudolph Peierls (1907 Berlin - 1995 Oxford) was a German-Jewish physicist who grew up in Berlin, and then moved to England. Already in 1936 he came up with his famous Peierls argument [ 35 , 29 ] which gave a proof that, contrary to the conjecture of Ising, there actually was long-range order in the two-dimensional short-range Ising model, in zero external field, at sufficiently low temperatures. His statement was about the finite-volume Gibbs measure in large boxes, with an all-plus boundary