• Aucun résultat trouvé

DYNAMICS OF THE CLASSICAL PLANAR SPIN CHAIN

N/A
N/A
Protected

Academic year: 2021

Partager "DYNAMICS OF THE CLASSICAL PLANAR SPIN CHAIN"

Copied!
3
0
0

Texte intégral

(1)

HAL Id: jpa-00217761

https://hal.archives-ouvertes.fr/jpa-00217761

Submitted on 1 Jan 1978

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

DYNAMICS OF THE CLASSICAL PLANAR SPIN

CHAIN

H. de Raedt, B. de Raedt

To cite this version:

(2)

JOURNAL DE PHYSIQUE Colloque C6, supplhment au no 8, Tome 39, aolit 1978, page C6-708

DYNAMICS OF THE CLASSICAL PLANAR SPIN CHAIN

H. De Raedt and B. De ~aedt+

+ Department Natuurkunde

,

Universitaire Inste ZZing Antuerpen B-2610 W i Z r i jk, Be lgiwn

I n s t i t u t ffir Theoretische Physik, Universitiit desSaarlandes 0-6600 Saarbrilcken,Gemany

R8sum6.- La technique Monte Carlo a 8tE utilis6e pour calculer les moments de la fonction de rela- xation pour la charne de spins classiques avec anisotropie dans le plan x-y. Ainsi on obtient des rgsultats d6taill6s pour les propri6tCs dynamiques. Comparaison est faite avec 11exp8rience et au-

tres th8ories.

Abstract.- The Monte Carlo method has been used to calculate the moments of the relaxation function for the classical planar spin chain. In this way extended results for the dynamic properties of this system are obtained. Comparison is made with experiment and other theories.

In a previous paper the authors presented a method, based on the projection operator formalism.

to obtain expressions for the relaxation function in a systematic way /l/. Ti was shown, that taking a spin component and its first two time derivatives as relevant variables, the normalized relaxation function can be written as

z2+zrj(z)-(<ak> /<a2> -<a2>

@ (2)' k k k ) (1)

k z(z2 -<W"> /<w~>~)+c~(z) k

In the simplest approximation the transport coeffi- cient Ck(z) was found to be given by / l /

<u6> -<a4>

k ;/

C,(z) =

-

(2)

< w ~ > ~ - ~ < w ~ > <a4> + < u ~ > ~

' / *

z+i

[

k k

5

The relaxation function is now completely determi- ned by its moments and < w ~ > ~ , and it satisfies the sum rules

identically.

The classical planar spin chain is described by the hamiltonian / 2 , 3 , /

where the spins are classical unit vectors, and we will take A = 0.21, corresponding to the anisotro- py in CsNi-F3, for which measurements were made by Steiner and a1./4/. Loveluck and al. calculated some moments using the transfer operator method /5/. But because the higher moments are very complicated ex- pressions, that give not much physical insight, we have used a Monte Carlo simulation to calculate the-

se quantities numerically. A more complete descrip- tion of the method, and extensed results will be presented in a forthcoming paper 161. Some results will now be summarized.

Fig.1 : The second, fourth and sixth moments for both the X- (left) and z-(right)components, for T = 0.5, and as a function of the wave vector k.

In figure I bare results for the moments are shown for the temperature T = 0.5. The moments of the X-component do not disappear at zero wave vec- tor, because the total spin in the X-direction is not a conserved quantity, in contrast with the to- tal spin in the z-direction. In figure 2 we give some results for the imaginary part of the relaxa- tion function for the wave vector k = 0.25 n. At lower temperatures the width of the spin wave peak of the z-component is extremely small, which means. that the excitations are almost undamped. At higher

(3)

Fig. 2 : The imaginary part of the normalized rela- xation function for the wave vector k P 0.2511 for

the X- (left) and z- (right) components.

temperatures the peak of the z-component seems to disappear more rapidly than that of the X-component. The difference between the X-and z-components at low-temperatures can be understood in terms of the correlation lengths. At zero temperature the corre- lation length

E x

diverges, because the spins order ferromagnetically in the X-y plane, while

CZ

remains finite /6/.

Our results agree with the experimental observations on CsNiF3 /4/, and with the theoreti-

References

/ l / De Raedt, H. and De Raedt, E. Phys. Rev. B

2

(1977) 5379.

/2/ Villain, J., J.Physique

35

(1974) 27. /3/ Loveluck, J.M., Lovesey, S.W. and Aubry, S.,

J. Phys. C

8

(1975) 3841.

/ 4 / Steiner, M., Villain, J. and Windsor C.G., Adv. Phys.

5

(1976) 87.

/5/ Loveluck, J.M. and Lovesey, S.W., J.Phys. C

8

(1975) 3857.

/6/ De Raedt,B. and De Raedt, H., to be published in Phys. Rev. B.

Références

Documents relatifs

– MCMC yields Up component velocity uncertainties which are around [1.7,4.6] times larger than from CATS, and within the millimetre per year range

Fujuki and Betts [14], and also Nishimori and Nakan- ishi [15] discussed the disappearance of long-range or- der in the study of quantum spin systems on the trian-

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des

Probabilities of errors for the single decoder: (solid) Probability of false negative, (dashed) Probability of false positive, [Down] Average number of caught colluders for the

We can explain this by the following argument: below the transition temperature, there exists a single large cluster of lattice spins with some isolated ”defects” (i. clusters

On the other hand, if one thinks that solving an integrable model means that one is able to reduce the complexity of the system using unified mathematical structures (quantum

Using results on the mass gap in the sine-Gordon model combined with the exact amplitudes in the bosonized representation of the Heisenberg spin-1/2 chain and one-loop

It yields upper bounds on the probability of outcomes of the program, according to the supplied random distributions, with worse- case behavior according to the nondeterminism;