Journal of Magnetism and Magnetic Materials 471 (2019) 501–503

Contents lists available at ScienceDirect

Journal of Magnetism and Magnetic Materials

journal homepage: www.elsevier.com/locate/jmmm

Research articles

Oguchi’s method and random interaction fields’ method: Investigation of T

properties of ferromagnetic materials
V. Belokona, A. Trofimova, O. Dyachenkob,
⁎

a
Department of Theoretical and Nuclear Physics, The School of Natural Sciences, Far Eastern Federal University, 8, Sukhanova St., 690950 Vladivostok, Russian
Federation
b
Department of General and Experimental Physics, The School of Natural Sciences, Far Eastern Federal University, 8, Sukhanova St., 690950 Vladivostok, Russian
Federation

ARTICLE INFO ABSTRACT

Keywords: The Curie point of a ferromagnetic material is determined in the framework of the Ising model by Oguchi’s
Magnetic phase transitions method and the random interaction fields’ method. The heat capacity of the system is also investigated. The
Random field method jump in the heat capacity is calculated at the Curie point. Comparison with known results is made.
Curie point
Ferromagnetism
Heat capacity
Oguchi’s method

1. Introduction For example, we consider the association of Oguchi’s method and

the random interaction fields’ method to study the properties of ferro-
A well-known, effective (molecular) fields’ method has enabled the magnetic materials.
qualitative understanding of the properties of ferromagnetic materials
in the theory of magnetism, although it is far from perfect quantita- 2. Random interaction fields’ method
tively. This method overestimates the Curie point Tc . A noticeable im-
provement in this method, allowing for local interactions and correla- In a random distribution of exchange-interacting ions in the crystal
tions of spins, is associated with the use of cluster methods [1–5] or by lattice, the distribution function of the random fields of the exchange
means the random interaction fields’ method [6–10]. interaction has the form [6]:
In contrast to the molecular field method, in Oguchi’s and Bethe-
1 [H ( ) H0 ]2
Peierls methods, a cluster consisting of two (Oguchi’s method or [z + 1] W (H ) = exp .
Bethe-Peierls method) interacting magnetic moments, where z is the
B B2 (1)
number of nearest neighbors, is placed in an effective field. Thus, The atom creating a molecular field H (ferromagnetic atoms) is as-
considering the correlation effectively decreases the exchange integral. sumed to have a random coordinate for amorphous magnetic material
In particular, in the framework of the Bethe-Peierls method, the ex- or a random lattice site. A magnetic moment also has a random inter-
change integral decreases approximately in (z 1)/z times [8]. action. Moments of the distribution function are defined as
In the Ising model framework, the random interaction fields’
method allows us to determine the distribution function, with para- H0 = p n,
n (2)
meters calculated by the spin interaction law (or the magnetic moments
of particles, clusters, and grains). This approach has advantages over B2 = 2p [1 ( )2p] 2
n,
traditional ones since it allows us to connect possible phase transitions n (3)
with the concentration of exchange-interacting atoms (ions) and cal-
where n describes the law of interaction among particles, is the re-
culate critical concentrations. Simultaneously applying cluster methods
lative probability of the spin being up, is the relative probability of
and the random interaction fields’ method enables a more accurate
the spin being down, and p is the concentration of the magnetic parti-
calculation of the phase transition conditions.
cles. The mean magnetization is defined as

⁎
Corresponding author.
E-mail addresses: belokon.vi@dvfu.ru (V. Belokon), trofimov.an@dvfu.ru (A. Trofimov), dyachenko.oi@dvfu.ru (O. Dyachenko).

https://doi.org/10.1016/j.jmmm.2018.09.065
Received 21 May 2018; Received in revised form 13 September 2018; Accepted 19 September 2018
Available online 19 September 2018
0304-8853/ © 2018 Elsevier B.V. All rights reserved.
V. Belokon et al. Journal of Magnetism and Magnetic Materials 471 (2019) 501–503

mH appears. Using the critical value H0 / B = 1, determining the critical

M = tanh W (H ) dH ,
KT (4) concentrations for other exchange interaction laws is possible, for ex-
ample, the RKKY.[9].
where M = ¯ ¯ . Here … denotes thermodynamic and configura-
tional averaging, …¯ denotes thermodynamic averaging, m is the mag- 3. The curie point: Oguchi’s method and random interaction
netic moment per ion, k is the Boltzmann constant, and T is the tem-
fields’ method
perature. The angle brackets will be omitted.
After integrating the expression, the magnetization can be de-
In Oguchi’s method, a pair of exchange-interacting atoms is placed
termined as
in the effective field H, created by means (z 1) neighbors, the partition
1 B m (H0 M + H ) function has the form:
M= tanh dH .
2B B kT (5) 1 + 2H 1 2H 1 1 2H
S = exp + exp + 2exp = 2exp cosh
Small M magnetization can be represented in the form T T T T T
1

M2 =
3 ( H0
B
tanh
mB
kT
1 ) ,
+ 2exp
T
,
(12)
m2H03
Bk 2T 2 (tanh mB
kT
tanh3
mB
kT ) (6)
and the magnetic moment is determined as
2H
sinh
As follows from Eq. (6) a spontaneous magnetic moment can arise M=
T
,
only under the condition cosh
2H
+ exp
2
T T (13)
H0 mB
tanh > 1. where H = (z 1) M .
B kT (7)
The Curie point can be obtained by requiring the derivative of the
Accordingly, the Curie point is determined from the equation right-hand side of Eq. (13) to M be equal to one. Thus, Tc = 3.78.
Using the random interaction fields’ method, the relative magnetic
H0 mB
tanh = 1. moment can be represented as
B kTc (8)
2H
sinh
Fields and temperatures are usually measured in energy units M=
T
W (H ) dH ,
2H 2
B = mB, H0 = mH0, T = kT . Omitting the primes, we obtain: cosh T
+ exp T (14)
H0 B (H MH0 )2
tanh = 1. where W (H ) = B e B = 2(z 1) .
1
B2 ,
B Tc (9)
Replacing W (H ) by the “rectangular” function [6] and passing to
The theory of the molecular field (TMF) is obtained from the lim- the new coordinates H = H + MH0 , we obtain
iting transition W (H ) (H mH0 ) , where (x ) is the Dirac delta
2(H + MH0)
function. Thus, 1 B sinh T
M= dH .
MpH0 2B B
cosh
2(H + H0)
+ exp
2
M = tanh . T T (15)
T (10)
Considering Eq. (15) near Tc for small M, we have
In the direct exchange n = J , H0 = zJ , J is the exchange integral. At
2B
J = 1, Tc = pz , where z is the number of the nearest neighbors.
(cosh[ )
sinh 3
H0 T 2B 2
1 ] + exp
Function W (H ) can be replaced by “rectangular” function W (x ) to B cosh 2B + exp 2 T T
T T
simplify the computation: M2 = .

0, B > x, B < x
2 H0
3 B ( ) H0 2
T
2B
sinh[ T ] (2 + cosh[ 2B
T
]exp
2
T
exp
4
T ) (16)
W (x ) = 1
2B
, B x B. The Curie point is found from the equation
2B
Obviously, W (x ) also tends to the Dirac delta function at B 0. H0 sinh[ T ]
= 1.
In [6], examples of the numerical solution of the equation for M are B cosh 2B
+ exp
2
T T (17)
given by the exact and approximate functions. Thus, near the phase
transition points, where M and B are small, the error in the calculations At z = 4 , the Curie point is Tc = 2.75. If we do not use the approxima-
is negligible. Additional arguments favoring such a possible substitution tions adopted above and determine the distribution function as the
can be considered as an estimate of the critical density pc corresponding exact binomial function for z = 4 (in our case, z 1 = 3)
to the percolation threshold, which is easy to obtain for the direct ex- 3
change interaction. (1 + M )k (1 M ) (3 k)
W (H ) = Ck3 H 2k 3 ,
We note immediately that for concentration p = 1 and the square k=0
23 (18)
lattice in Eq. (8), the Curie point is Tc = 3.2 , which is closer to the exact
B where C is the binomial coefficient, then the Curie point is determined
solutions than in the conventional TMP. Since tanh Tc
1, a sponta- from Equation:
neous magnetic moment can arise when the condition B0 > 1 is sa-
H
3 2(2k 3)
(1 + M ) k (1 M )(3 k) sinh[ ]
tisfied. Then for direct exchange, it means (see the definition of H0 and M= Ck3 T
,
B2 under the condition M 1): 23 2(2k 3)
cosh[ T ] + exp
2
k=0 T (19)
2
p > pc = . whose solution is Tc = 2.74 . The use of the binomial function, even for
z (11)
such a small value as z = 3, does not give a noticeable improvement in
The values of the critical concentration pc turned out to be close to the result
the percolation thresholds, so in the direct exchange, naturally, ferro- This result provides a more accurate value of the Curie temperature
magnetic ordering is possible only when a so-called flowing cluster than in the Bethe-Peierls method, wherein the Curie point is

502
V. Belokon et al. Journal of Magnetism and Magnetic Materials 471 (2019) 501–503

Tc = 2.89.The exact Curie temperature obtained by Onsager is Tc = 2.28. 2H 2H

1 H sinh T
cosh T z
U = .
4. Heat capacity: Oguchi’s method and random interaction fields’
method
2
( cosh
2H
T
+ exp
2
T ) 2
(26)
The configuration averaging by the function W (H ) allows us to
In the theory of random interaction fields, the mean value of the calculate U by the formula
square of the magnetic moment is defined as
2(H + MH0 ) 2(H + MH0 )
z 1 H sinh cosh
1 BH + MH0 B T T
M2 = tanh2 dH U= dH .
2B
T
B

B
T
MH0 B MH0
4B B 2
(cosh 2(H + MH0 )
T
+ exp
2
T ) (27)
=1 tanh + + tanh .
2B T T T T (20) Determining the heat capacity “from the left” of the Curie point,
differentiating the expression for U with respect to temperature, and
If T > Tc then M = 0 ; if T < Tc then M 0 ; and near the Curie point Tc
using Eq. (17) are necessary. At T = 2.75, the heat capacity “to the left”
the magnitude of M is determined from Eq. (6).
of the Curie point is C1 = 3.98. For M = 0 , the heat capacity “to the
Because of the expansion in a series of a value tanh T ± T 0 , we
B MH
right” of the Curie point is C2 = 0.24 .? Thus, the heat capacity experi-
obtain ences a stronger jump at T = Tc than in the random interaction field
B MH0 B B MH0 approximation applied to an isolated atom.
tanh ± = tanh ± 1 tanh2
T T T T T
5. Conclusion
B B MH0 2
+ tanh + tanh3 .
T T T (21)
Preserving the relative simplicity of the molecular field theory, the
When Eq. (21) is substituted into Eq. (20), the second term drops random interaction fields’ method combined with Oguchi’s correlations
out during summation. Thus the average value of the square of the provides a closer estimate of the Curie point and specific heat dis-
magnetic moment is defined as continuity than the Bethe-Peierls and Bragg-Williams methods.
2
T B B B MH0 Acknowledgement
M2 = 1 2tanh +2 tanh + tanh3 .
2B T T T T
(22) This work was financially supported by the State Program of the
The energy per atom is U = zM 2/2. Ministry of Education and Science of the Russian Federation
The heat capacity “from the left” of the Curie point has the form: 3.7383.2017/8.9.

Appendix A. Supplementary data

3 1 1 B 1 z
C1 = sec2 .
Tc cosh2 B Tc Tc H0 2
Tc (23) Supplementary data associated with this article can be found, in the
online version, athttps://doi.org/10.1016/j.jmmm.2018.09.065.
( ) ( ).
2 2
Given that tanh2 then
B 1 B 1 B
Tc
=1 B
= H0 B
=1 H0
cosh2 cosh2
Tc Tc References
After this the heat capacity can be defined as:
2 2 [1] P. Weiss, The application of the Bethe-Peierls method to ferromagnetism, Phys. Rev.
3 B 1 B 1 z 74 (10) (1948) 1493.
C1 = 1 1 .
Tc H0 Tc H0 H0 2 [2] W. Bragg, E. Williams, The effect of thermal agitation on atomic arrangement in
(24) alloys, Proc. R. Soc. Lond. Ser. A 145 (1934) 699–730.
[3] J.S. Smart, Effective field theory of magnetism.
In the direct exchange = 2pz, = near T = Tc . At
B2 H02 p2 z 2 [4] S. Semkin, V. Smagin, Mean-field approximation for the potts model of a diluted
p = 1, z = 4 ,and Tc = 3.2 , the heat capacity at the Curie point is magnet in the external field, Phys. Solid State 58 (7) (2016) 1350–1354.
C1 = 1.2 . [5] S. Semkin, V. Smagin, Bethe approximation in the Ising model with mobile im-
purities, Phys. Solid State 57 (5) (2015) 943–948.
The heat capacity “to the right” from the Curie point is determined [6] V. Belokon, K. Nefedev, Distribution function for random interaction fields in dis-
by the right-hand term of Eq. (24) ordered magnets: Spin and macrospin glass, J. Exp. Theor. Phys. 93 (1) (2001)
136–142.
2
1 B 1 z [7] V. Belokon, K. Nefedev, M. Savunov, Spin glass state with finite interaction range in
C2 = 1 . the Ising model, Solid State Phys. 48 (2006) 1649–1656.
Tc H0 H0 2 (25) [8] V. Belokon, V. Kapitan, O. Dyachenko, The combination of the random interaction
fields’ method and the bethe-peierls method for studying two-sublattice magnets, J.
In the direct exchange at the Curie point, the heat capacity is C2 = 0.2. Magn. Magn. Mater. 401 (2016) 651–655.
Thus, at a point T = Tc , the heat capacity experiences a jump rela- [9] V. Belokon, O. Dyachenko, Random interaction fields method: magnetic phase
transitions in the thin films, J. Magn. Magn. Mater. 374 (2015) 92–95.
tively smaller than in the Bragg-Williams approximation, but a nonzero [10] L. Afremov, S. Anisimov, I. Iliushin, Size effect on the hysteresis characteristics of a
heat capacity is associated with random fluctuations of the effective system of interacting core/shell nanoparticles, J. Magn. Magn. Mater. 444 (2018)
field of the exchange interaction. 88–95.
In Oguchi’s method, the thermodynamically averaged internal en-
ergy U per atom is determined by the formula

503