Related Experiment Videos
n--> infinity limit of O(n) ferromagnetic models on graphs
1Dipartimento di Fisica, Istituto Nazionale di Fisica della Materia, Universita di Parma, Viale delle Scienze, 43100 Parma, Italy.
Physical Review Letters
|September 6, 2000
Summary
The singular free energies of O(n) spin models coincide with spherical models on graphs, revealing critical exponents depend solely on spectral dimension.
Area of Science:
- Statistical Mechanics
- Condensed Matter Physics
- Graph Theory
Background:
- H. E. Stanley's 30-year-old finding: O(n) spin models on lattices converge to spherical models as n approaches infinity.
- This convergence implies identical free energies at all temperatures.
- The result is not directly applicable to general discrete structures like graphs due to lack of translation invariance.
Purpose of the Study:
- To investigate if the singular parts of free energies, crucial for critical behavior, coincide for O(n) and spherical models on graphs.
- To determine if critical exponents for O(n) models on graphs (as n approaches infinity) align with spherical model exponents.
- To explore the dependence of these critical exponents on graph properties.
Main Methods:
- Analysis of ferromagnetic O(n) spin models on general graphs.
- Focus on the thermodynamic limit to study critical phenomena.
- Examination of the singular components of free energies.
Main Results:
- Demonstrated that the singular parts of free energies for O(n) and spherical models coincide on graphs.
- Established that critical exponents for O(n) models (n-->infinity) on graphs converge to spherical model exponents.
- Showed that these critical exponents are determined by the graph's spectral dimension.
Conclusions:
- The universality of critical behavior between O(n) and spherical models extends to graphs, despite the breakdown of direct free energy coincidence.
- Graph spectral dimension is a key factor governing critical exponents in the large-n limit.
- This finding offers insights into phase transitions on complex network structures.