Zubov1997 PDF

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

V. I. Zubov et al.

: Thermal and Elastic Properties of Solid Sodium 27

phys. stat. sol. (b) 200, 27 (1997)


Subject classification: 63.10.+a; 62.20.Dc; 65.50.+m; S2

Thermal and Elastic Properties


of Solid Sodium Based
on an Effective Interionic Potential
V. I. Zubov 1 †, J. F. Sanchez 2 †, N. P. Tretiakov 1 †, and A. A. Caparica

Departamento de F{sica, Universidade Federal de Goi


as, C.P. 131,
74001-970 Goi^
ania-Goi as, Brazil

(Received August 19, 1996)

Basing on the Schiff effective interionic potential which has oscillatory character we have calculated
a complete set of equilibrium thermodynamic properties of solid sodium at normal pressure above
the martensitic transition temperature. To take into account the anharmonicity which is strong at
high temperatures we use the correlative method of unsymmetrized self-consistent field elaborated
previously. Our results are in agreement with measured values and with available data of computer
simulations. We discuss the thermodynamic stability of the b.c.c. lattice of sodium and the mechan-
ism of its loss. The metastability of a hypothetical f.c.c. lattice with Schiff potential is discussed.

1. Introduction
Difficulties encountered in theoretical investigations of the atomic properties of metals
are associated with the presence of almost free electrons. The majority of calculations of
such properties were based on the pseudopotential theory [1 to 6] which usually consid-
ered the lattice vibrations to be harmonic or quasi-harmonic and, at most, included
anharmonic effects using perturbation methods. However, at temperatures above half
the melting point of a crystal, its lattice anharmonicity is strong. In this case, the low-
est-order anharmonic terms cannot be treated as a perturbation (see, e.g. [7 to 9]). The
application of the pseudopotential theory to strongly anharmonic crystals is difficult.
Swanson et al. [10] used for solid sodium the quantum quasi-harmonic approximation at
low temperatures and classical molecular dynamics (without using the perturbation the-
ory for anharmonic terms) at high temperatures. They calculated its scalar thermody-
namic functions and obtained a good fit to experimental results.
Another approach is to use effective interionic potentials taking into account the influ-
ence of the electron gas on the ionic cores. The possibility of such potentials, at least up
to the fourth-order anharmonic terms, readily follows from the Born-Oppenheimer adia-
batic theorem [11], but their constructions for specific materials is a nontrivial problem.
Because of this, the attempts in this approach in general have not been successful. It is
common for effective interionic potentials in metals to be of an oscillatory form resulting
from the screening of Coulomb interionic forces by conduction electrons. Such Friedel

1
† On leave from Department of Theoretical Physics, Peoples' Friendship University, Moscow,
Russia.
2
† Present address: Departamento de Fsica e Matematika, Universidad Cayetano Heredia,
Lima, Peru.
28 V. I. Zubov, J. F. Sanchez, N. P. Tretiakov, and A. A. Caparica

oscillations are of particular importance for not close-packed lattices providing their ab-
solute stability. For some metals, oscillations may be smoothed up by electrons of incom-
plete inner shells. On occasion even model potentials of the Morse or Lennard-Jones
type are used.
In our calculations of the thermodynamic properties of solid sodium we utilize the
Schiff interionic potential
   
F…r† B C cos 2kF R E sin 2kF R
ˆ A‡ 2‡ 4 ‡ D ‡ ; …1†
e R R R3 R2 R4

where e is the depth of the potential well, R ˆ r=s, and s is the effective diameter of an
ion screened by conduction electrons. The parameters are: e=k ˆ 599 K, s ˆ 3:24  A,
A ˆ 0:19, B ˆ ÿ1:02, C ˆ ÿ0:08, D ˆ ÿ0:43, E ˆ ÿ2:54, 2kF ˆ 5:987, k is the Boltz-
mann constant. Originally this potential was proposed for the liquid phase [12]. More
recently Galashev [13] applied it to the solid state on the basis of the molecular dy-
namics technique. To prevent extrapolation of this function to large interionic distances
he truncated it at r  2:5s that coincides with one-half edge length of the basic cell in
his machine experiment.
We follow the correlative method of unsymmetrized self-consistent field (CUSF) which
enables one to calculate a complete set of thermal and elastic properties of strongly
anharmonic crystals using a personal computer, with little computing time being re-
quired. According to our estimates, in the body-centred cubic (b.c.c.) lattice the inclu-
sion of seven coordination spheres for which r  3s provides best results.

2. Method and Outline of Computations


The basis of the CUSF for scalar thermodynamic functions of strongly anharmonic crys-
tals are explained in [14 to 17]. This method appears to be efficient for the investigation
of such functions for simple van der Waals crystals [17 to 19], alkali halides [19, 20] and
also sodium [21]. Recently, we have extended it to calculations of the elastic constant
tensors [22] and applied it to a molecular crystal ±± fullerite C60 [23].
One can find formulae of the scalar thermodynamic functions of cubic strongly anhar-
monic crystals in [15 to 17]. They contain the zeroth approximation, the first quantum
corrections (a quasi-classical approach) and the contributions of the statistical perturba-
tion theory. Here, we write the zeroth approximation for the components of the isother-
mal elastic constant tensor and the first quantum corrections to them taking into con-
sideration anharmonic terms up to the fourth order,
T …0; Q† P …400†
C11 ˆ h0; Q …q; a† ‡ Zk x0; Q …q; a; k† ÿ P0; Q ; …2†
k1

T …0; Q† P …220†
C12 ˆ h0; Q …q; a† ‡ Zk x0; Q …q; a; k† ÿ P0; Q ; …3†
k1
  
T …0† P a2
…220† 1 6 1 2 4
C44 ˆ Zk j2 q4 ‡ j…2w ‡ c†
x0 ÿ q ‡ q q
k1 15qn 7 3
  #
q2 2 6 1 2 4 2 q8 2 2
‡ q ‡ q q …2w ‡ c† ‡ …16w ‡ 12wc ‡ 9c † ÿ P0 ; …4†
12 7 3 756
Thermal and Elastic Properties of Solid Sodium 29

T …Q† P …220† 2
h
C44 ˆ Zk xQ …q; a; k† ÿ
k1 1080mqn
   
1 6 1 2 4
 j…2w ‡ 3c† q4 ‡ q ‡ q q …3w2 ‡ 4wc ‡ 3c2 † ÿ PQ ; …5†
7 3

where P0 and PQ are the zeroth approximation and the first quantum correction for
pressure determined from the equation of state at given temperature q ˆ kB T and
interionic distance a (or volume of the unit cell n…a† ˆ V =N†, and m is the atomic
mass.
We also use the following notations:
"  
a2 q 1 dK2 2 Xb0 dK2 dK4
h0 ˆ …Xb0 ÿ b† ÿ
18n K2 da K2 K4 da da
 2 #
1 dK4
‡ …2b ‡ Xb0 ÿ 6† ; …6†
2K4 da
 
R4k ^ 2 bq ^ 2 …3 ÿ b† q ^ 2
x0 ˆ D F…Rk † ‡ D F2 …Rk † ‡ D F4 …Rk † ; …7†
2n K2 2K4
 
P …220† ^ 3 Zk R2k ^ 2
j…a† ˆ R4k Zk D ‡ D F…Rk † ; …8†
k1 3
 
P …420† ^ 5 R4 …400† …220† ^ 4 ^ 3 F…Rk † ;
w…a† ˆ R6k Zk D ‡ k …Zk ‡ 9Zk † D ‡ Zk R2k D …9†
k1 2
 
P …222† ^ 5 …220† ^ 4 Zk R2k ^ 3
c…a† ˆ R6k Zk D ‡ R4k Zk D ‡ D F…Rk † ; …10†
k1 3
"     #
h2 a 2 K 2
5 00 1 dK2 1 dK4 2  0 g 1 dK2 2
hQ ˆ Xg ÿ ‡ g ÿ ;
108mqn K2 da 2K4 da X 2K2 da
…11†
  
 2 R4k
h 5g0 ^ 2 5K2  g 0

^ 2
xQ ˆ 1‡ D F2 …Rk † ‡ ÿ g D F4 …Rk † : …12†
4mqn 3 6K4 X

Here
1 P ^ P
F2l …r† ˆ r2l F…r† ; K2l ˆ F2l …jAnj† ˆ Zk F2l …Rk † ; …13†
2l ‡ 1 n 6ˆ 0 k1

^ˆ 1 d ;
D …14†
r dr
^ is the lattice matrix, n are integer component vectors, Zk and Rk the coordination
A
numbers and the radii of the coordination spheres, respectively,

…ijl† 1 P ^ i …An†
^ j …An†
^ l
Zk ˆ …An†x y z …15†
Rik‡ j ‡ l jAnj
^ ˆ Rk
30 V. I. Zubov, J. F. Sanchez, N. P. Tretiakov, and A. A. Caparica

are the partial coordination numbers [22, 23]. For cubic crystals they are invariant un-
der permutations of upper indices and possess the properties
…000† …200† …400† …220†
Zk ˆ 3Zk ˆ 3Zk ‡ 6Zk
…600† …420† …222†
ˆ 3Zk ‡ 18Zk ‡ 6Zk ˆ . . . ˆ Zk : …16†
p
The function b…K2 3=qK4 † is an implicit one determined by the transcendental equa-
tion
Dÿ2:5 …X ‡ 5b=6X†
b…X† ˆ 3X ; …17†
Dÿ1:5 …X ‡ 5b=6X†
Dm …Y † are the parabolic cylinder functions, and
b
gˆ : …18†
X
The behaviour of the functions b…X† and g…X† on the real axis has been thoroughly
studied (see, e.g. [15, 20]). Their derivatives can be expressed in terms of these functions
themselves and their argument. In particular,
54 ÿ 18Xg ÿ 24g2
g0 ˆ ; …19†
63 ÿ 15Xg ÿ 20g2
18g ÿ 18Xg0 ÿ 63gg0 ‡ 5…g0 †2 …3X ‡ 8g†
g00 ˆ : …20†
63 ÿ 5g…3X ‡ 4g†
The variance of the atomic position is represented by
bq
q2 ˆ : …21†
K2
The higher-order moments can be calculated using both the definition
…
1 ( " ! #)
2l ‡ 2 1 1 5K4 q2 2 K4 4
q exp ÿ K2 ‡ q ‡ q dq
q 2 18 24
0
q2l ˆ 1 … ( " ! #) …22†
2 1 1 5K4 q2 2 K4 4
q exp ÿ K2 ‡ q ‡ q dq
q 2 18 24
0

and the recurrence relation


 
6…2l ÿ 1† 2l ÿ 4 6K2 5 2 2l ÿ 2
q2l ˆ q ÿ ‡ q q ; l ˆ 2; 3; . . . …23†
K4 K4 3
with q0 ˆ 1 and q2 from (21).
It is known that the quasi-classical approach including the first quantum corrections
is valid when T  TD =3, where TD is the Debye temperature. Sodium has a body-centred
cubic (b.c.c.) lattice above the martensitic phase transition region 60 to 78 K [24, 25]
and TD  150 K [26] (at normal pressure). Hence, such an approach is appropriate in
this case.
The thermodynamic perturbation theory improves the results taking into account, in
particular, the higher (fifth and sixth) anharmonic terms. Corrections to the isothermal
Thermal and Elastic Properties of Solid Sodium 31

elastic constants (2) to (5) are expressed in terms of the second derivatives of the corre-
sponding corrections to the Helmholtz free energy [16, 17] with respect to the compo-
nents of the strain tensor uab . They are rather cumbersome and we do not write them
here. Notice only that they, unlike the zeroth approximation and quantum corrections,
are quadratic forms of the derivatives of the interionic potential and along with the
moments (21) to (23), contain their first and second derivatives with respect to uab .
Since such quadratic forms decrease with increasing interionic distance much more ra-
pidly than linear forms contained in (2) to (5), the perturbation theory requires the
consideration of a smaller number of coordination spheres than the zeroth approxima-
tion and quantum corrections. In case of the Schiff potential (1) in the b.c.c. lattice we
have summed (2) to (5) over seven spheres. For the perturbation theory it suffices to
include two spheres. (Note that in the case of face-centred cubic (f.c.c.) van der Waals
crystals, the perturbation theory needs only the first coordination sphere [22].)
For the derivatives of moments with respect to the strain tensor the following formu-
lae have been obtained [22]:
@q2l a @q2l
ˆ dab ; a; b ˆ 1; 2; 3 ;
@uab 3 @a
" #
…400†
@ 2 q2l a2 @ 2 q2l @q2l
Z1 C
2
 2
ÿa …C ‡ 1† ÿ ;
@uaa 9 @a @a Z1 9
" #
…220†
@ 2 q2l a @ 2 q2l @q2l C Z1
 ‡a ÿ …C ‡ 1† ;
@uaa ubb 9 @a2 @a 9 Z1
…220†
@ 2 q2l Z @q2l
 ÿa…C ‡ 1† 1 ; a 6ˆ b ; …24†
@u2ab Z1 @a

where
 00 
K2 K400 0 dK2l 00 d2 K2l
C  ÿa ‡ ; K2l  ; K2l  : …25†
K20 K40 da da2
From (21) it immediately follows that
 
@q2 q 0 @X bK20
ˆ b ÿ ;
@a K2 @a K2
"    0 2 #
2
@ 2 q2 q 00 @X 0 @ X
2
K20 b0 @X bK200 K2
2
ˆ b ‡b 2
ÿ2 ÿ ‡ 2b : …26†
@a K2 @a @a K2 @a K2 K2

The derivatives of the higher-order moments with respect to the interionic distance can
be deduced by differentiation of (22) or (23) and substitution of expressions (26) there.
As a result, for these derivatives we have integral representations and recurrence rela-
tions, respectively.
The adiabatic elastic constants are related to the isothermal ones, the thermal expansion
coefficient and the isochoric specific heat by the familiar thermodynamic formulae [27]

S T S T TV a2 B2T S T
C11 ÿ C11 ˆ C12 ÿ C12 ˆ ; C44 ÿ C44 ˆ 0; …27†
CV
32 V. I. Zubov, J. F. Sanchez, N. P. Tretiakov, and A. A. Caparica

where all functions consist of the zeroth approximation, the quantum corrections and
the contributions of the perturbation theory. The last-named ones for a and BT contain
the first and second derivatives of moments with respect to temperature and cross deri-
vatives. They can be obtained using integral representations or recurrence relations as
well.
In principle, the latter are more convenient in calculations because they are less com-
puter time consuming. However, they consists of several terms of opposite signs, so that
their difference is far less than the absolute value of each term, especially at low tem-
peratures. As this takes place, they do not provide a reasonable accuracy in the case of a
b.c.c. crystal with Schiff potential. On the other hand, the integrands for second deriva-
tives of the moments peak very sharply. This makes numerical integration with neces-
sary accuracy difficult at low temperatures. That is why we used the integral representa-
tions for the higher-order moments and their first derivatives. Then, differentiating
numerically the latter, we calculated the second derivatives.

3. Results and Discussion


We have solved the equations of state for solid sodium based on the CUSF [21] at
normal pressure and T  80 K. At temperatures less than a limit Tl  662:1 K, it has
two roots a1 …T † < a2 …T † which coalesce at the limiting temperature a1 …Tl † ˆ a2 …Tl †,
and when T > Tl there is no real root. At the lower branch of the isobar, the isother-
mal bulk modulus is positive while at the upper one is negative. The latter stands for
the absolute instability of the states represented by a2 …T †. It is known [28, 29] that in
the stable and metastable regions of a thermodynamic phase, the stability determinant
is positive together with its principal minors. For cubic crystals, these conditions are

Fig. 1. The normal isobar of solid sodium obtained using (1) CUSF and (2) quasi-harmonic approx-
imation, Tm is the experimental melting temperature, Ts the spinodal point
Thermal and Elastic Properties of Solid Sodium 33

T
Fig. 2. The stability coefficients of solid sodium: (1) C11 , (2) BT , (3) C44 , (4) C11 ÿ C12 (all in
kbar), and (5) T =CV (in mol K2 /J)

of the form
T
BT > 0 ; C11 > 0; C11 ÿ C12 > 0 ; C44 > 0 ; T =CV > 0 : …28†
The violation of any of these inequalities signifies the loss of stability (spinodal point) of
a given phase.
In Fig. 1 the normal isobar of solid sodium is shown obtained using CUSF together
with that in the quasi-harmonic approximation. At not high temperatures, the lower
branches of both curves are nearly the same and very close to the experimental data
[30]. However, in the vicinity of the experimental melting point Tm ˆ 370:95 K and
above it (in the metastable region) the quasi-harmonic curve deviates. The anharmonic
effects will be discussed in more detail below. Note also that the lower branch of curve 1
is very close to the results calculated by Swanson et al. [10] using a combination of the
quasi-harmonic approximation at low temperatures and computer simulations in the
classical regime, therefore these latter are omitted in Fig. 1.
Along the lower branch of the isobar of solid sodium we have calculated its thermody-
namic properties. Fig. 2 displays its stability coefficients (28). One can see that all but
T =CV decrease with increasing temperature. Up to 498 K where the shear coefficient
C11 ÿ C12 turns to zero, all the values (28) are positive. The stability coefficient C44
T
passes through zero at about 570 K and C11 does this in the vicinity of the temperature
of merging branches of the isobar Tl , where BT ˆ 0. Hence, the normal pressure spinodal
point of solid sodium is Ts  498 K, in which the shear instability takes place with
…C11 ÿ C12 † / …T ÿ Ts †. The region of the lower branch between Tm and Ts represents
metastable states. Its domain T > Ts and the whole upper branch correspond to abso-
lutely unstable states. It is interesting that the relation Ts =Tm  1:34 is very close to
that for van der Waals crystals [17], however, for them Ts ˆ Tl , i.e. their instability has
bulk nature, and CUSF gives BT / …T ÿ Ts †0:5 [31].

3 physica (b) 200/1


34 V. I. Zubov, J. F. Sanchez, N. P. Tretiakov, and A. A. Caparica

Fig. 3. The theoretical (1) isothermal and (2) adiabatic bulk moduli, (3) results of Swanson et al.
[10] for BT , and the experimental data for BT (circles) [32] and for BS (asterisks) [25]

In Fig. 3 we compare the isothermal and adiabatic bulk moduli with experimental
data [25, 32] and also with values of BT computed by Swanson et al. [10]. The discre-
pancy with experiment comprising 20 to 25% at 80 K, rapidly decreases as the tempera-
ture increases. Note that the experimental data are between our results and those of

S S
Fig. 4. The adiabatic elastic constants: (1) and asterisks ±± C11 , (2) and solid circles ±± C12 , (3)
and diamonds ±± C44 . The experimental data were taken from [25]
Thermal and Elastic Properties of Solid Sodium 35

Fig. 5. The (1) isobaric and (2) isochoric specific heats (in cal Kÿ1 molÿ1 ), (3) the thermal expan-
sion coefficient (in 10ÿ4 Kÿ1 †, …30 † a in the quasi-harmonic approximation, and (4) the Gr uneisen
parameter. The experimental data were taken for CP (asterisks) from [33, 34], for a (solid circles)
from [30], and for g (diamonds) from [25]

Swanson et al. Fig. 4 presents the adiabatic elastic constants of solid sodium. Here, one
can see that the agreement with experimental data improves with increasing tempera-
ture as well.
Finally, in Fig. 5 we give the specific heats, the thermal expansion coefficient and the
Gruneisen parameter g ˆ V aBT =CV . For the isobaric specific heat, the discrepancy be-
tween our results and experimental data [33, 34] is mainly in the range from 1 to 3%.
Exceptions occur for low temperatures (80 to 100 K) where the first quantum correc-
tions do not provide a high precision for specific heats and in the vicinity of the melting
point in which a major portion of the discrepancy can be attributed to the contribution
from vacancies. For the thermal expansion coefficient and the Gr uneisen parameter, the
fit is somewhat worse than for the above-listed properties. In the case of a, the discre-
pancy with experimental data increases with temperature and comes to about 23% at
the melting point. For g, the discrepancy is about 30% throughout the entire tempera-
ture interval.
As to the quasi-harmonic approximation, it gives worse results at all. For instance,
the ratio between values of a calculated in this approximation (curve 30 in Fig. 5) and
experimental ones ranges up to 1.57 at the melting point. This testifies that at high
temperatures, the anharmonicity in sodium is strong (but it is somewhat weaker than in
van der Waals crystals [17, 21]).
We have also computed the properties of a hypothetical f.c.c. lattice with Schiff poten-
tial at normal pressure. It has been found that up to Ts ˆ Tl  701 K where BT goes to
zero, all its stability coefficients (28) remain positive; that is the specific phase is at least
metastable. The cohesive energies of both lattices are very close to each other, with their
difference depending on the number of coordination spheres included. Taking into ac-

3*
36 V. I. Zubov, J. F. Sanchez, N. P. Tretiakov, and A. A. Caparica

count seven spheres, the cohesive energy of the b.c.c. lattice is about 1% bigger than
that of the f.c.c. lattice. It should be considered as an evidence for the adequacy of the
truncation of the Schiff potential for solid sodium after the seventh coordination sphere
rather than a proof of the absolute stability of the first lattice and the metastability of
the latter.

4. Conclusions
So, basing on the correlative method of unsymmetrized self-consistent field for strongly
anharmonic crystals and using the Schiff effective interionic potential (1) in the Born-
Oppenheimer adiabatic approximation which takes into account the influence of the con-
duction electrons, we have studied a complete set of thermal and elastic properties of
solid sodium including the thermodynamic stability of the real b.c.c. and hypothetical
f.c.c. lattices at normal pressure. On the strength of the obtained results we conclude
that:
1. The Schiff potential depicts adequately the effective interionic forces in solid so-
dium up to the seventh coordination spheres.
2. This potential truncated after the seventh sphere provides a satisfactory agreement
with experimental data, absolute stability of the first lattice, metastability of the latter
and the realistic (shear) mechanism for the loss of stability of the b.c.c. lattice.
3. At high temperatures (near the melting point and especially in the metastable re-
gion), the anharmonicity of the lattice vibrations in sodium is strong but somewhat
weaker than in van der Waals crystals.

Acknowledgements Two of the authors (V.I.Z. and N.P.T.) are grateful to Conselho
Nacional de Desenvolvimento Cientfico e Tecnol
ogico (CNPq), Brazil, for financial sup-
port.

References
[1] W. Harrison, Pseudopotentials in the Theory of Metals, W. A. Benjamin, Inc., New York
1966.
[2] V. Heine and I. V. Abarenkov, Phil. Mag. 9, 451 (1964).
[3] S. Cohen and M. L. Klein, Phys. Rev. B 12, 2984 (1975).
[4] D. Sen, A. Sarkar, and S. Sengupta, Acta phys. Polon. A 67, 773 (1985).
[5] S. S. Cohen, M. L. Klein, M. S. Duesbery, and R. Taylor, J. Phys. F: Metal Phys. 6,
L271 (1976).
[6] L. F. Magan ~a and G. J. Va zquez, J. Phys.: Condensed Matter 2, 4807 (1990).
[7] Ph. F. Choquard, The Anharmonic Crystal, W. A. Benjamin, Inc., New York 1967.
[8] N. M. Plakida and T. Siklos, phys. stat. sol. (b) 39, 171 (1970).
[9] M. L. Klein and G. K. Horton, J. Low-Temp. Phys. 9, 151 (1972).
[10] R. E. Swanson, G. K. Straub, B. L. Holian, and D. C. Wallace, Phys. Rev. B 23, 7807
(1981).
[11] M. Born and K. Huang, Dynamical Theory of Crystal Lattices, Clarendon Press, Oxford
1954.
[12] D. Schiff, Phys. Rev. 186, 151 (1969).
[13] A. E. Galashev, in: Thermophysical Properties of Metastable Systems, Izd. Nauka, Sverd-
lovsk 1984 (p. 35) (in Russian).
[14] V. I. Zubov and Ya. P. Terletsky, Ann. Phys. (Leipzig) 24, 97 (1970).
[15] V. I. Zuvbov, Ann. Phys. (Leipzig) 31, 33 (1974).
[16] V. I. Zubov, phys. stat. sol. (b) 72, 71, 483 (1975).
Thermal and Elastic Properties of Solid Sodium 37

[17] V. I. Zubov, phys. stat. sol. (b) 87, 385, 88, 43 (1978).
[18] V. I. Zubov, phys. stat. sol. (b) 101, 95 (1980).
[19] V. I. Zubov and S. Sh. Soulayman, Kristallografiya 27, 588 (1982).
[20] V. I. Yukalov and V. I. Zubov, Fortschr. Phys. 31, 627 (1983).
[21] J. F. Sanchez Ortiz, N. P. Tretyakov, and V. I. Zubov, phys. stat. sol. (b) 181, K7
(1994).
[22] V. I. Zubov, J. F. Sanchez, N. P. Tretiakov, and A. E. Yusef, Internat. J. mod. Phys.
B 9, 803 (1995).
[23] V. I. Zubov, J. F. Sanchez, N. P. Tretiakov, and A. A. Caparica, Phys. Rev. B 53,
12080 (1996).
[24] R. Stern, G. G. Natale, and J. Rudnick, J. Phys. Chem. Solids 27, 9 (1966).
[25] G. Fritsch, M. Nehmann, P. Korpiun, and E. Lu  scher, phys. stat. sol. (a) 19, 555 (1973).
[26] N. W. Ashcroft and N. D. Mermin, Solid State Physics, Holt. Rinehart and Winston, New
York 1976.
[27] G. Leibfried, Gittertheorie der mechanischen und thermischen Eigenschaften der Kristalle,
Springer-Verlag, Berlin 1955.
[28] V. K. Semenchenko, Izbrannye glavy teoreticheskoi fiziki, Vysshaya Shkola, Moscow 1966.
[29] V. K. Semenchenko, In: Peregretye zhidkosti i fazovye perekhody, Akad. Nauk SSSR, Sverd-
lovsk 1979 (p. 3).
[30] W. Adlhart, G. Fritsch, A. Heidemann, and E. Lu  scher, Phys. Letters A 47, 91 (1974).
[31] V. I. Zubov, A. A. Caparica, N. P. Tretiakov, and J. F. Sanchez Ortiz, Solid State
Commun. 91, 941 (1994).
[32] G. Fritsch, F. Geipel, and P. Prasetyo, J. Phys. Chem. Solids 34, 1961 (1973).
[33] D. L. Martin, Phys. Rev. 154, 571 (1967).
[34] C. B. Alcock, M. W. Chase, and V. P. Itkin, J. Phys. Chem. Ref. Data 27, 385 (1995).

You might also like