View source for Spin glass

{{Multiple issues|
{{technical|date=September 2011}}
{{more citations needed|date=September 2011}}
{{context|date=September 2011}}
}}

[[File:Spin glass by Zureks.svg|thumb|Schematic representation of the '''random''' spin structure of a '''spin glass''' (top) and the '''ordered''' one of a '''ferromagnet''' (bottom)]]
{{multiple image
| direction = horizontal
| width = 150
| footer = The magnetic disorder of spin glass compared to a ferromagnet is analogous to the positional disorder of glass (left) compared to quartz (right).
| image1 = SiO2_-_Glas_-_2D.png
| alt1 = Amorphous SiO<sub>2</sub>
| caption1 = Glass (amorphous SiO<sub>2</sub>)
| image2 = SiO2_-_Quarz_-_2D.png
| alt2 = Crystalline SiO<sub>2</sub>)
| caption2 = Quartz (crystalline SiO<sub>2</sub>)
| align = 
| total_width = 
}}

In [[condensed matter physics]], a '''spin glass''' is a model of a certain type of [[magnet]].  Magnetic spins are, roughly speaking, the orientation of the north and south magnetic poles in three-dimensional space.  In [[Ferromagnetism|ferromagnetic]] solids, component atoms' magnetic [[Spin (physics)|spins]] all align in the same direction.  Spin glasses are contrasted with ferromagnets as "[[Entropy|disordered]]" magnets in which their atoms' spins are not aligned in a regular pattern.

The term "glass" comes from an analogy between the ''magnetic'' disorder in a spin glass and the ''positional'' disorder of a conventional, chemical [[glass]], e.g., a window glass.  In window glass or any [[Amorphous solid|amorphous]] solid the atomic bond structure is highly irregular; in contrast, a [[crystal]] has a uniform pattern of atomic bonds.  In [[ferromagnetic]] solids, magnetic spins all align in the same direction; this is analogous to a crystal's [[Crystal lattice structure|lattice-based structure]].

The individual atomic bonds in a spin glass are a mixture of roughly equal numbers of ferromagnetic bonds (where neighbors have the same orientation) and [[antiferromagnet]]ic bonds (where neighbors have exactly the opposite orientation: north and south poles are flipped 180 degrees).  These patterns of aligned and misaligned atomic magnets create what are known as [[Geometrical frustration|frustrated interactions]] – distortions in the geometry of atomic bonds compared to what would be seen in a regular, fully aligned solid.  They may also create situations where more than one geometric arrangement of atoms is stable.

Spin glasses and the complex internal structures that arise within them are termed "[[Metastability|metastable]]" because they are "stuck" in stable configurations other than the [[Ground state|lowest-energy configuration]] (which would be aligned and ferromagnetic).  The mathematical complexity of these structures is difficult but fruitful to study experimentally or in [[simulation]]s; with applications to physics, chemistry, materials science and [[Artificial neural network|artificial neural networks]] in computer science.

==Magnetic behavior==
It is the time dependence which distinguishes spin glasses from other magnetic systems.

Above the spin glass [[phase transition|transition temperature]], ''T''<sub>''c''</sub>,<ref group="note"><math>T_c</math> is identical with the so-called "freezing temperature" <math>T_f</math></ref> the spin glass exhibits typical magnetic behaviour (such as [[paramagnetism]]).

If a [[Applied magnetic field|magnetic field]] is applied as the sample is cooled to the transition temperature, magnetization of the sample increases as described by the [[Curie's law|Curie law]]. Upon reaching ''T''<sub>''c''</sub>, the sample becomes a spin glass and further cooling results in little change in magnetization. This is referred to as the ''field-cooled'' magnetization.

When the external magnetic field is removed, the magnetization of the spin glass falls rapidly to a lower value known as the ''remanent'' magnetization.

Magnetization then decays slowly as it approaches zero (or some small fraction of the original value—this [[List of unsolved problems in physics|remains unknown]]). This [[exponential decay|decay is non-exponential]] and no simple function can fit the curve of magnetization versus time adequately.<ref name="JPhys">{{cite journal|last1=Joy|first1=P A|last2=Kumar|first2=P S Anil|last3=Date|first3=S K|title=The relationship between field-cooled and zero-field-cooled susceptibilities of some ordered magnetic systems|journal=J. Phys.: Condens. Matter|date=7 October 1998|volume=10|issue=48|pages=11049–11054|doi=10.1088/0953-8984/10/48/024|bibcode = 1998JPCM...1011049J }}</ref> This slow decay is particular to spin glasses. Experimental measurements on the order of days have shown continual changes above the noise level of instrumentation.<ref name="JPhys" />

Spin glasses differ from ferromagnetic materials by the fact that after the external magnetic field is removed from a ferromagnetic substance, the magnetization remains indefinitely at the remanent value. Paramagnetic materials differ from spin glasses by the fact that, after the external magnetic field is removed, the magnetization rapidly falls to zero, with no remanent magnetization. The decay is rapid and exponential.{{Citation needed|date=September 2011}}

If the sample is cooled below ''T''<sub>''c''</sub> in the absence of an external magnetic field and a magnetic field is applied after the transition to the spin glass phase, there is a rapid initial increase to a value called the ''zero-field-cooled'' magnetization. A slow upward drift then occurs toward the field-cooled magnetization.

Surprisingly, the sum of the two complicated functions of time (the zero-field-cooled and remanent magnetizations) is a constant, namely the field-cooled value, and thus both share identical functional forms with time,<ref name="Nordblad">{{cite journal|last1=Nordblad|first1=P|last2=Lundgren|first2=L|last3=Sandlund|first3=L|title=A link between the relaxation of the zero field cooled and the thermoremanent magnetizations in spin glasses|journal=Journal of Magnetism and Magnetic Materials|date=February 1986|volume=54-57|issue=1|pages=185–186|doi=10.1016/0304-8853(86)90543-3|bibcode = 1986JMMM...54..185N }}</ref> at least in the limit of very small external fields.

==Edwards–Anderson model==
In this model, we have spins arranged on a <math>d</math>-dimensional lattice with only nearest neighbor interactions similar to the [[Ising model]]. This model can be solved exactly for the critical temperatures and a glassy phase is observed to exist at low temperatures.<ref name=nishimori>{{cite book|last=Nishimori|first=Hidetoshi|title=Statistical Physics of Spin Glasses and Information Processing: An Introduction|year=2001|publisher=Oxford University Press|location=Oxford|isbn=9780198509400|pages=243}}</ref> The [[Hamiltonian mechanics|Hamiltonian]] for this spin system is given by:

: <math>
H = -\sum_{\langle ij\rangle} J_{ij} S_i S_j,
</math>

where <math>S_{i}</math> refers to the [[Pauli spin matrix]] for the spin-half particle at lattice point <math>i</math>. A negative value of <math>J_{ij}</math> denotes an antiferromagnetic type interaction between spins at points <math>i</math> and <math>j</math>. The sum runs over all nearest neighbor positions on a lattice, of any dimension. The variables <math>J_{ij}</math> representing the magnetic nature of the spin-spin interactions are called bond or link variables.

In order to determine the partition function for this system, one needs to average the free energy <math>f[J_{ij}] = -\dfrac 1 \beta \ln \mathcal{Z}\left[J_{ij}\right]</math> where <math>\mathcal{Z}\left[J_{ij}\right] = \operatorname{Tr}_S e^{(-\beta H)}</math>, over all possible values of <math>J_{ij}</math>. The distribution of values of <math>J_{ij}</math> is taken to be a gaussian with a mean <math>J_0</math> and a variance <math>J^2</math>:

: <math>
P(J_{ij}) = \sqrt{\frac{N}{2\pi J^2}} \exp\left\{-\frac N {2J^2} \left(J_{ij} - \frac{J_0} N \right)^2\right\}.
</math>

Solving for the free energy using the [[replica trick|replica method]], below a certain temperature, a new magnetic phase called the spin glass phase (or glassy phase) of the system is found to exist which is characterized by a vanishing magnetization <math>m = 0</math> along with a non-vanishing value of the two point correlation function between spins at the same lattice point but at two different replicas: 

<math>q = \sum_{i=1}^N S^\alpha_i S^\beta_i \neq 0, </math>

where <math>\alpha, \beta</math> are replica indices. The [[order parameter]] for the ferromagnetic to spin glass phase transition is therefore <math>q</math>, and that for paramagnetic to spin glass is again <math>q</math>. Hence the new set of order parameters describing the three magnetic phases consists of both <math>m</math> and <math>q</math>.

Under the assumption of replica symmetry, the mean-field free energy is given by the expression<ref name=nishimori />

: <math>
\begin{align}
\beta f = &- \frac{\beta^2 J^2}{4}(1-q)^2 + \dfrac{\beta J_0 m^2}{2} \\
&- \int \exp\left( -\frac{z^2} 2 \right) \log \left(2\cosh\left(\beta J z + \beta J_0 m\right)\right) \, \mathrm{d}z.
\end{align}
</math>

==The model of Sherrington and Kirkpatrick==
In addition to unusual experimental properties, spin glasses are the subject of extensive theoretical and computational investigations. A substantial part of early theoretical work on spin glasses dealt with a form of [[mean field theory]] based on a set of [[replica trick|replicas]] of the [[partition function (statistical mechanics)|partition function]] of the system.

An important, exactly solvable model of a spin glass was introduced by [[David Sherrington (physicist)|David Sherrington]] and Scott Kirkpatrick in 1975. It is an [[Ising model]] with long range frustrated ferro- as well as antiferromagnetic couplings. It corresponds to a [[mean field theory|mean field approximation]] of spin glasses describing the slow dynamics of the magnetization and the complex non-ergodic equilibrium state.

Unlike the Edwards–Anderson (EA) model, in the system though only two-spin interactions are considered, the range of each interaction can be potentially infinite (of the order of the size of the lattice). Therefore, we see that any two spins can be linked with a ferromagnetic or an antiferromagnetic bond and the distribution of these is given exactly as in the case of Edwards–Anderson model. The Hamiltonian for SK model is very similar to the EA model:

: <math>
H = -\sum_{i<j} J_{ij} S_i S_j
</math>

where <math>J_{ij}, S_i, S_j</math> have same meanings as in the EA model. The equilibrium solution of the model, after some initial attempts by Sherrington, Kirkpatrick and others, was found by [[Giorgio Parisi]] in 1979 with the replica method. The subsequent work of interpretation of the Parisi solution—by M. Mezard, G. Parisi, M.A. Virasoro and many others—revealed the complex nature of a glassy low temperature phase characterized by ergodicity breaking, ultrametricity and non-selfaverageness. Further developments led to the creation of the [[cavity method]], which allowed study of the low temperature phase without replicas. A rigorous proof of the Parisi solution has been provided in the work of [[Francesco Guerra]] and [[Michel Talagrand]].<ref>Michel Talagrand, ''[http://michel.talagrand.net/challenge/volume1.pdf Mean Field Models for Spin Glasses Volume I: Basic Examples]'' (2010)</ref>

The formalism of replica mean field theory has also been applied in the study of [[neural networks]], where it has enabled calculations of properties such as the storage capacity of simple neural network architectures without requiring a training algorithm (such as [[backpropagation]]) to be designed or implemented.<ref name="Gardner">{{cite journal|last1=Gardner|first1=E|last2=Deridda|first2=B|title=Optimal storage properties of neural network models|journal=J. Phys. A|date=7 January 1988|volume=21|number=1|pages=271|doi=10.1088/0305-4470/21/1/031}}</ref>

More realistic spin glass models with short range frustrated interactions and disorder, like the [[Carl Gauss|Gaussian]] model where the couplings between neighboring spins follow a [[Gaussian distribution]], have been studied extensively as well, especially using [[Monte Carlo simulation]]s. These models display spin glass phases bordered by sharp [[phase transition]]s.

Besides its relevance in condensed matter physics, spin glass theory has acquired a strongly interdisciplinary character, with applications to [[neural network]] theory, computer science, theoretical biology, [[econophysics]] etc.

==Infinite-range model==
The infinite-range model is a generalization of the [[Spin glass#The model of Sherrington and Kirkpatrick|Sherrington–Kirkpatrick model]] where we not only consider two spin interactions but <math>r</math>-spin interactions, where <math>r \leq N</math> and <math>N</math> is the total number of spins. Unlike the Edwards–Anderson model, similar to the SK model, the interaction range is still infinite. The Hamiltonian for this model is described by:

: <math>
H = -\sum_{i_1 < i_2 < \cdots < i_r} J_{i_1 \dots i_r} S_{i_1}\cdots S_{i_r}
</math>

where <math>J_{i_1\dots i_r}, S_{i_1},\dots, S_{i_r}</math> have similar meanings as in the EA model. The <math>r\to \infty</math> limit of this model is known as the [[random energy model]]. In this limit, it can be seen that the probability of the spin glass existing in a particular state, depends only on the energy of that state and not on the individual spin configurations in it.
A gaussian distribution of magnetic bonds across the lattice is assumed usually to solve this model. Any other distribution is expected to give the same result, as a consequence of the [[central limit theorem]]. The gaussian distribution function, with mean <math>\dfrac{J_0} N </math> and variance <math>\dfrac{J^2} N</math>, is given as:

: <math>
P(J_{i_1\cdots i_r}) = \sqrt{\dfrac{N^{r-1}}{J^2 \pi r!}} \exp\left\{-\frac{N^{r-1}}{J^2 r!} \left(J_{i_1\cdots i_r} - \frac{J_0 r!}{2N^{r-1}}\right)\right\}
</math>

The order parameters for this system are given by the magnetization <math>m</math> and the two point spin correlation between spins at the same site <math>q</math>, in two different replicas, which are the same as for the SK model. This infinite range model can be solved explicitly for the free energy<ref name=nishimori/>  in terms of <math>m</math> and <math>q</math>, under the assumption of replica symmetry as well as 1-Replica Symmetry Breaking.<ref name=nishimori/>

: <math>
\begin{align}
\beta f = {} & \frac{\beta^2 J^2 q^r} 4 - \frac{r\beta^2 J^2 q^r} 2 - \frac{\beta^2 J^2} 4 + \frac{\beta J_0 r m^r} 2 + \frac{r\beta^2 J^2 q^{r-1}}{4\sqrt{2\pi}} \\
& {} + \int \exp\left( -\frac{z^2}2 \right)\log \left(2\cosh\left(\beta Jz \sqrt{\frac{rq^{r-1}}{2}} + \frac{\beta J_0 r m^{r-1}}{2}\right)\right) \, \mathrm{d}z
\end{align}
</math>

==Non-ergodic behavior and applications==
A thermodynamic system is [[ergodic]] when, given any (equilibrium) instance of the system, it eventually visits every other possible (equilibrium) state (of the same energy). One characteristic of spin glass systems is that, below the freezing temperature <math>T_\text{f}</math>, instances are trapped in a "non-ergodic" set of states: the system may fluctuate between several states, but cannot transition to other states of equivalent energy. Intuitively, one can say that the system cannot escape from deep minima of the hierarchically disordered energy landscape; the distances between minima are given by an [[ultrametric]], with tall energy barriers between minima.<ref group="note">The hierarchical disorder of the energy landscape may be verbally characterized by a single sentence: in this landscape there are "(random) valleys within still deeper (random) valleys within still deeper (random) valleys, ..., etc."</ref> The [[participation ratio]] counts the number of states that are accessible from a given instance, that is, the number of states that participate in the [[ground state]].

For physical systems, such as dilute magnesium in copper, the freezing temperature is typically as low as 30 [[kelvin]]s (−240 °C), and so the spin-glass magnetism appears to be practically without applications in daily life. The non-ergodic states and rugged energy landscapes are, however, quite useful in understanding the behavior of certain [[neural networks]], including [[Hopfield network]]s, as well as many problems in [[computer science]] [[optimization (mathematics)|optimization]] and [[genetics]].

==History of the field==

A detailed account of the history of spin glasses from the early 1960s to the late 1980s can be found in a series of popular articles by [[Philip Warren Anderson|Philip W. Anderson]] in ''[[Physics Today]]''.<ref>
{{cite journal
 |title=Spin Glass I: A Scaling Law Rescued
 |journal = Physics Today|volume = 41|pages = 9–11|author=Philip W. Anderson
 |year=1988
 |doi=10.1063/1.2811268
 |url=http://www.physics.rutgers.edu/~pchandra/physics681/sglass1.pdf
 }}</ref><ref>
{{cite journal
 |title=Spin Glass II: Is There a Phase Transition?
 |journal = Physics Today|volume = 41|issue = 3|pages = 9|author=Philip W. Anderson
 |year=1988
 |doi=10.1063/1.2811336
 |url=http://www.physics.rutgers.edu/~pchandra/physics681/sglass2.pdf
 }}</ref><ref>
{{cite journal
 |title=Spin Glass III: Theory Raises its Head
 |journal = Physics Today|volume = 41|issue = 6|pages = 9–11|author=Philip W. Anderson
 |year=1988
 |doi=10.1063/1.2811440
 |url=http://www.physics.rutgers.edu/~pchandra/physics681/sglass3.pdf
 }}</ref><ref>
{{cite journal
 |title=Spin Glass IV: Glimmerings of Trouble
 |journal = Physics Today|volume = 41|issue = 9|pages = 9–11|author=Philip W. Anderson
 |year=1988
 |doi=10.1063/1.881135
 |url=http://www.physics.rutgers.edu/~pchandra/physics681/sglass4.pdf
 }}</ref><ref>
{{cite journal
 |title=Spin Glass V: Real Power Brought to Bear
 |journal = Physics Today|volume = 42|issue = 7|pages = 9–11|author=Philip W. Anderson
 |year=1989
 |doi=10.1063/1.2811073
 |url=http://www.physics.rutgers.edu/~pchandra/physics681/sglass5.pdf
 }}</ref><ref>
{{cite journal
 |title=Spin Glass VI: Spin Glass As Cornucopia
 |journal = Physics Today|volume = 42|issue = 9|pages = 9–11|author=Philip W. Anderson
 |year=1989
 |doi=10.1063/1.2811137
 |url=http://www.physics.rutgers.edu/~pchandra/physics681/sglass6.pdf
 }}</ref><ref>
{{cite journal
 |title=Spin Glass VII: Spin Glass as Paradigm
 |journal = Physics Today|volume = 43|issue = 3|pages = 9–11|author=Philip W. Anderson
 |year=1990
 |doi=10.1063/1.2810479
 |url=http://www.physics.rutgers.edu/~pchandra/physics681/sglass7.pdf
 }}</ref>

==See also==
*[[Antiferromagnetic interaction]]
*[[Cavity method]]
*[[Crystal structure]]
*[[Geometrical frustration]]
*[[Orientational glass]]
*[[Phase transition]]
*[[Quenched disorder]]
*[[Random energy model]]
*[[Replica trick]]
*[[Spin ice]]

==Notes==
{{Reflist|group="note"}}

==References==
{{Reflist}}

==Literature==
*{{citation|first1=S.F.|last1=Edwards|first2=P.W.|last2=Anderson|journal=Journal of Physics F: Metal Physics|title=Theory of spin glasses|volume=5|issue=5|pages=965–974|year=1975|doi=10.1088/0305-4608/5/5/017|bibcode=1975JPhF....5..965E}}. [http://iopscience.iop.org/article/10.1088/0305-4608/5/5/017/meta;jsessionid=4B8D9A38523A828CD28C8CE67DD973E8.c5.iopscience.cld.iop.org]
*{{citation|first1=David|last1=Sherrington|first2=Scott|last2=Kirkpatrick|journal=Physical Review Letters|title=Solvable model of a spin-glass|volume=35|pages=1792–1796|doi=10.1103/PhysRevLett.35.1792|issue=26|year=1975|bibcode=1975PhRvL..35.1792S}}. [http://papercore.org/Sherrington1975 Papercore Summary http://papercore.org/Sherrington1975]
*{{citation|first1=P.|last1=Nordblad|first2=L.|last2=Lundgren|first3=L.|last3=Sandlund|title=A link between the relaxation of the zero field cooled and the thermoremanent magnetizations in spin glasses|journal=Journal of Magnetism and Magnetic Materials|volume=54|pages=185–186|year=1986|doi=10.1016/0304-8853(86)90543-3|bibcode = 1986JMMM...54..185N }}.
*{{citation|author1-link=Kurt Binder|first1=K.|last1=Binder|first2=A. P.|last2=Young|title=Spin glasses: Experimental facts, theoretical concepts, and open questions|journal=Reviews of Modern Physics|volume=58|issue=4|pages=801–976|year=1986|doi=10.1103/RevModPhys.58.801|bibcode=1986RvMP...58..801B}}.
*{{citation|last1=Bryngelson|first1=Joseph D.|first2=Peter G.|last2=Wolynes|title=Spin glasses and the statistical mechanics of protein folding|journal=[[Proceedings of the National Academy of Sciences]]|volume=84|issue=21|pages=7524–7528|year=1987|doi=10.1073/pnas.84.21.7524|pmid=3478708|bibcode = 1987PNAS...84.7524B |pmc=299331}}.
*{{citation|first1=K. H.|last1=Fischer|first2=J. A.|last2=Hertz|title=Spin Glasses|publisher=Cambridge University Press|year=1991}}.
* {{citation
 | last1 = Mezard
 | first1= Marc
 | last2=Parisi|first2=Giorgio|author2-link=Giorgio Parisi
 | last3=Virasoro|first3=Miguel Angel|author3-link=Miguel Ángel Virasoro (physicist)
 | year = 1987
 | title = Spin glass theory and beyond
 | publisher = World Scientific
 | location = Singapore
 | isbn =  978-9971-5-0115-0
}}.
*{{citation|first=J. A.|last=Mydosh|title=Spin Glasses|publisher=Taylor & Francis|year=1995}}.
*{{citation| first=G.| last=Parisi|title=The order parameter for spin glasses: a function on the interval 0-1|journal=J. Phys. A: Math. Gen.| volume= 13| issue=3| pages=1101–1112| year=1980| doi=10.1088/0305-4470/13/3/042|bibcode = 1980JPhA...13.1101P |url=https://www.openaccessrepository.it/record/19057/files/LNF_79_038%28P%29.pdf}} [http://papercore.org/Parisi1980 Papercore Summary http://papercore.org/Parisi1980].
*{{citation|authorlink=Michel Talagrand|first=Michel|last=Talagrand|journal=Annals of Probability|volume=28|pages=1018–1062|year=2000|jstor=2652978|title=Replica symmetry breaking and exponential inequalities for the Sherrington–Kirkpatrick model|issue=3|doi=10.1214/aop/1019160325}}.
*.{{citation|first1=F.|last1=Guerra|first2=F. L.|last2=Toninelli|title=The thermodynamic limit in mean field spin glass models|journal=Communications in Mathematical Physics|volume=230|issue=1|pages=71–79|year=2002|doi=10.1007/s00220-002-0699-y|arxiv = cond-mat/0204280 |bibcode = 2002CMaPh.230...71G }}
*{{citation|first1=T. G.|last1=Aminov|first2=V. N.|last2=Novotortsev|title=Spin Glasses in Cu<sub>0.5</sub>Fe<sub>0.5</sub>Cr<sub>2</sub>S<sub>4</sub> - Based Solid Solutions|journal=Inorganic Materials|volume=50|issue=13|pages=1343–00|year=2014|doi=10.1134/s0020168514130020|issn=0020-1685 }}

==External links==
*[http://papercore.org/summaries/solvable-model-of-a-spin-glass Papercore summary of seminal Sherrington/Kirkpatrick paper]
*[http://xstructure.inr.ac.ru/x-bin/theme3.py?level=2&index1=125728 Statistics of frequency of the term "Spin glass" in arxiv.org]
{{magnetic states}}

{{Authority control}}

{{DEFAULTSORT:Spin Glass}}
<!--Categories-->
[[Category:Magnetic ordering]]
[[Category:Theoretical physics]]
[[Category:Mathematical physics]]