Full text of DOI:10.1557/JMR.2000.0410

Formation of silicate structures in Cu-containing
silica xerogels
F. Ruiz
Instituto de F´ısica “Manuel Sandoval Vallarta,” Universidad Auto´noma de San Luis Potos´ı,
78000 San Luis Potos´ı, S.L.P., Mexico
J.R. Mart´ınez
Facultad de Ciencias, Universidad Auto´noma de San Luis Potosí, 78000 San Luis Potos´ı,
S.L.P., Mexico
J. Gonza´lez-Herna´ndez
Centro de Investigacio´n y de Estudios Avanzados del IPN, Unidad Quere´taro, Apdo. Postal 1-1010,
C.P. 76010, Quere´taro, Qro., Mexico
(Received 21 March 2000; accepted 22 September 2000)
Using infrared absorption and Raman spectroscopy, the structure of SiO2 sol-gel
samples containing copper were analyzed as a function of air heat treatments. The
weight composition of the samples was 70(SiO2)–30(CuO) and the thermal treatments
were carried out in air at various temperatures in the range of 100–600 °C. It was
found that the Cu might be incorporated into the SiO2 matrix at least in two ways:
forming oxide particles and forming metasilicatelike structures. The copper metasilicate
found had the chemical formula, Cu6[Si6O18] и 6H2O, which is known as dioptase with
a thermal stability at temperatures up to 600 °C.
I. INTRODUCTION
smaller than 50 nm, and that a small amount of the cop-
per atoms form copper silicate structures. According to
our results, the copper silicate formed is probably the
dioptase, which is a metasilicate with the chemical for-
mula Cu6[Si6O18] и 6H2O.12 Its chemical structure is
formed with sixfold beryl-like rings of the form
[Si6O18]−12. According to Belov et al.,13 the Cu cations
with the O atoms and the OH groups form octahedral
columns oriented along a threefold screw axis with six-
fold silicon–oxygen rings at three levels, while sixfold
rings consisting of H2O molecules are distributed be-
tween the columns. The Cu atoms, in dioptase, have
(4 + 2)–fold coordination; their four nearest atoms are 4
oxygen of three different Si6O18 ring surrounding the
copper atom in an almost perfect square. The coordina-
tion of the Cu atom is raised to 6 by the two O atoms
from the water molecules. Our results indicate that,
among the two known stable copper silicates, the heat
treatments promote the formation of small dioptase struc-
tural units, because one of the distinct features of the
silica xerogel structures is the presence of the sixfold
rings (six oxygen and six silicon atoms).14
The sol-gel method is a versatile technique for the
preparation of glass coatings and powdered samples that
allows the incorporation of metals and other materials in
their matrixes. The modifications to the glass structure
that results from the incorporations of different elements
depend on the type of the incorporated elements, post-
deposition treatments, and the interactions of the ele-
ments with the host matrix.1 The introduction of
transition metals in a SiO2 glass matrix has a strong
influence on the optical visible absorption spectrum.2
Nano- and microsized particles of various materials
embedded in the glass matrix produce quantum and non-
linear optical effects when the particles have some criti-
cal size.3
Several authors have reported studies of sol-gel
prepared SiO2 samples containing copper.4–8 To our
knowledge none of the previous studies observed the
formation of metastable copper silicates; instead, they
report the formation of oxides or metallic particles,
depending on whether the samples are subjected to
heat treatments in oxidizing or reducing atmospheres,
respectively.9–11
II. EXPERIMENTAL METHODS
In the present work, we prepared sol-gel SiO2 glasses
containing copper, and the structural evolutions, due to
heat treatments in air, have been analyzed using Raman
spectroscopy, infrared (IR) absorption, and x-ray diffrac-
tion. From these measurements we can conclude that
most of the copper precipitates into oxide particles
The starting solutions were prepared by mixing tetra-
ethyl orthosilicate (TEOS), water, ethanol, and copper
nitrate. The molar ratios of ethanol to TEOS and water to
TEOS were 4:1 and 11.67:1, respectively. The concen-
tration of Cu nitrates were calculated to obtain 30 wt% of
J. Mater. Res., Vol. 15, No. 12, Dec 2000
© 2000 Materials Research Society
2875
IP address: 128.218.248.209
http://journals.cambridge.org
Downloaded: 23 Mar 2015

F. Ruiz et al.: Formation of silicate structures in Cu-containing silica xerogels
cupric oxide. The nitrate used for the incorporation of
copper was the Cu(NO3)2 и 2.5H2O and the amount
added was such that the sample weight composition was
70(SiO2)–30(CuO), assuming that all the copper forms
the cupric oxide. To catalyze the hydrolysis/condensation
reactions small amounts of nitric acid were added. The
metal nitrate was dissolved in the water, and the TEOS
and the ethanol were also separately mixed. A homog-
enous solution of all components was obtained mixing
them for about 15 min using a magnetic stirrer to form
the starting solution. This solution was placed in a con-
stant bath temperature of 35 °C. Soft pieces of the glass
were obtained after about 48 h. Previously, the bath tem-
perature was fixed at 50 °C during 2 h. For subsequent
annealing the samples were introduced in the oven at the
desired temperature for 15 min at atmospheric air con-
ditions. Those pieces were ground to form a fine powder.
The Raman spectra were obtained at room temperature
using a Labram-Dilor micro-Raman system in which the
sample was excited with a He–Ne laser and the signal
analyzed with a low-noise charge-coupled device detec-
tor. To check for possible changes in the sample induced
by laser heating, the laser spot diameter was varied in the
1–10 ␮m range and the power density changed using
neutral density filters.
Fig. 1. IR absorption spectra in the range of 400 to 2200 cm−1 for the
70(SiO2)–30(CuO) powered samples.
shoulder at its high-energy side. The absorption band at
about 950 cm−1, which intensity increases with the an-
nealing temperature, is due to vibrations of the silanol
(Si–OH) groups.18 The absorption band at about 1650
cm−1 has been previously assigned to vibrations of free
molecular water.19
Several reports have been published concerning the
high-energy shoulder of the stretching band, in the fre-
quency range of 1150–1300 cm−1; however, its origin is
not yet clear. For instance, in previous work IR spectra
have been reported in which these subbands can have an
amplitude comparable or larger than the main stretching
band at 1078 cm−1.20,21 This has been achieved in vitre-
ous SiO2 samples prepared by the sol-gel method from
solutions with a large water/TEOS ratio. The enhanced
intensity of these subbands has been interpreted as due to
the formation of a chain- or ringlike structure. Lange22
has also observed that the intensity of these subbands
increases when vitreous SiO2 thin films are subjected to
ion bombardment. He has proposed that the relative in-
crease in the intensity of the subband is due to an increase
in the structural disorder.
From Fig. 1 it can be seen that the relative intensity of
the bending (B) and rocking (R) bands basically does not
change for the former and increases for the latter with the
increase in the annealing temperature. This could be due
to some absorption from cupric oxide, whose vibrational
modes has been reported at 420, 425, and 528 cm−1.23
The Si–O stretching (S) band at approximately 1078
cm−1 suffers significant modification with the increase in
the annealing temperature. An explanation of this will be
proposed in next parts of this section.
The IR spectra were measured in a Fourier Transform
infrared spectrometer Perkin Elmer System Model 2000
using the diffuse reflectance mode: For that, 0.05 g of
powder sample was mixed with 0.3 g of KBr.
III. RESULTS
Figure 1 shows the IR absorption spectra in the range
of 400 and 2200 cm−1 for a sample with copper after heat
treatments in air at the indicated temperatures. The strong
absorption band, at about 1400 cm−1 in the sample heat
treated at 100 °C, is associated with vibrations of the
−
nitrate radical NO3 . This band has less intensity in
the sample treated at 200 °C and it does not appear
in samples treated at 400 and 600 °C. The latter is be-
cause the copper nitrate thermally decomposes, and the
nitrogen evolves to the atmosphere in the form of nitro-
gen oxides.15 The three features identified with the letters
R, B, and S are absorption bands related with particular
vibrational modes of the oxygen (O) atom with respect to
the silicon (Si) atom that they bridge. The rocking (R)
of the O atom along an axis through the two Si atoms is
associated with the lowest-frequency band centered at
about 457 cm−1. The bending (B) of the O atom along a
line bisecting the axis formed by the two Si atoms cor-
responds to the band centered at about 800 cm−1. The
band at about 1078 cm−1 is associated with the asym-
metrical stretch (S) motion, in which the O atom moves
back and forth along a line parallel to the axis along the
two Si atoms.16,17 This mode has a strong absorption
Several authors have deconvoluted the high-energy
shoulder of the stretching band into several bands, which
according to them are associated with specific vibrations
related to some detail in the SiO2 structure. Lange and
2876
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 23 Mar 2015
http://journals.cambridge.org
IP address: 128.218.248.209

F. Ruiz et al.: Formation of silicate structures in Cu-containing silica xerogels
Kirk,22,24 for example, have reported vibrational modes
per, in which the relative intensity of these bands clearly
decreases with the annealing temperature, as a conse-
quence of the glass densification.1,15,20,27
at 1254, 1170, and 1200 cm−1, and related them with
disorder-induced modes in amorphous SiO2 films. Seifert
et al.25 suggested that the presence of these subbands
could be related to two or more structure types present in
the fully polymerized glass. Vuckcetich has proposed a
two-state model26 for the SiO2 glass structure, in which
there are at least two distinct populations of Si–O–Si
angles within the tetrahedral network with value near
138° and 145°. We have assigned these bands to the
presence of ringlike structures.20,21
From the results in Fig. 2, one can see that the relative
intensity and broadness of subbands 3, 4, and 5 change
with the annealing temperature. For instance, subband 3
becomes broader and its relative intensity, with respect
to the higher energy bands, increases with the increase in
the annealing temperature. The linewidth of this band has
been directly related with the Si–O–Si bond angle distri-
bution, indicating the degree of structural relaxation of
the SiO2 network.28 The relative intensity and linewidth
of subbands 4 and 5 also change; however, to our knowl-
edge, the reason for these changes has not yet been fully
understood. In previous publications we proposed that
these subbands can be associated with vibrations of rings
with a different number of members (Si–O–Si). Thus, the
change in the intensity of any of these subbands reflects
the change in the statistical population of the particular
ring involved.20 According to our observations20 and
other investigations,14,29–32 the structure of samples an-
nealed at low temperatures are basically formed by four-,
five-, and mainly sixfold rings. For higher annealing tem-
peratures the amount of larger rings is decreased and so
are the intensities of the corresponding subbands.
Figure 2 shows the deconvoluted IR spectra in the
range of 700–1300 cm−1 for the SiO2 with copper. The
spectra were decomposed in six absorption bands, which
gave the best fitting to the experimental data. The dotted
curve denotes the measured spectra and the solid curve,
traced along the data points, is the sum of the six bands
(thin lines) obtained from the decomposition. As can be
seen from these figures, the positions of the subbands are
about 800, 930, 964, 1078, 1168, and 1220 cm−1, labeled
with B, 1, 2, 3, 4, and 5, respectively. It is observed that
the band assigned to the Si–OH vibration is composed of
two bands. This has been explained as due to the exis-
tence of at least two different configurations for the Si–
OH groups.18 For annealing temperatures above 200 °C
the relative intensity and width of these last two bands
increase. This result differs from the observations in
other SiO2 sol-gel samples, including those without cop-
Figure 3 shows the IR spectra for a SiO2 film grown
on a silicon wafer by thermal oxidation33 and for a
silica xerogel sample with Cu and annealed at 600 °C for
Fig. 2. Deconvoluted IR spectra of 70(SiO2)–30(CuO) in the 700–1300 cm−1 range. The spectra were decomposed in six absorption bands at:
approximately 800 cm−1 (B), approximately 930 cm−1 (1), approximately 964 cm−1 (2), approximately 1078 cm−1 (3), approximately 1068 cm−1
(4), and approximately 1220 cm−1 (5).
J. Mater. Res., Vol. 15, No. 12, Dec 2000
2877
IP address: 128.218.248.209
http://journals.cambridge.org
Downloaded: 23 Mar 2015

F. Ruiz et al.: Formation of silicate structures in Cu-containing silica xerogels
15 min. There are several differences between the two
spectra: the relative intensity of the stretching band, in
the region of 1100–1250 cm−1, as well as that of the
Si–OH vibration band is more pronounced in the xerogel
sample, and the rocking mode appears broader, which,
at least, in part is due to the absorption of the copper
oxides, in the region of 500–650 cm−1. Obviously, the
two preparation processes are different. In the sol-gel
process, a wet process, the presence of water at high
concentrations promotes a high degree of hydrolization,
which in turn produces samples with longer Si–O–Si
chains or rings with a larger number of members.1,18,20 It
is known that thermally grown SiO2 on silicon wafers
has a dense structure, formed by a statistical distribu-
tion of n-fold rings predicted by previous theoretical
calculations.14,29–32
Figure 4 shows the Raman spectra in the range of 200
to 1100 cm−1 for the same set of samples shown in Fig. 1.
For comparison a spectrum is included (top) for a pure,
that is, without copper, sol-gel-made SiO2 sample an-
nealed at 200 °C. For this sample, four major features are
observed, characterizing the amorphous SiO2 Raman
spectrum.16 The position of these bands is at about 430,
495, 850, and 980 cm−1. The sharp band at approxi-
mately 495 cm−1, known as D1 band, along with a band
at about 606 cm−1, known as the D2 band, have received
considerable attention and several interpretations have
been proposed29,34–36; the most accepted was made by
Galeener.29 He has suggested that the D1 and D2 bands
correspond to decoupled modes of in-phase breathing
motions of oxygen atoms in planar three- and four-
member rings, respectively. In agreement with other Ra-
man results in sol-gel-made samples, our data do not
Fig. 4. Raman spectra in the range of 200 to 1100 cm−1 for the same
set of samples of Fig. 1. For comparison a spectrum of a nondoped
xerogel sample annealed at 200 °C for 15 min is included.
show the D2 mode.37,38 According to Riegel et al.,37 in
SiO2 samples prepared using a wet method, the chemi-
sorbed water beaks up the siloxane bridges, especially
those under mechanical stress, as in threefold rings,
transforming those in rings with a larger number of mem-
bers with a free hydroxyl group. The Raman band at 980
cm−1 supports this interpretation, because it has been
previously related with vibrations of SiO4 units with two
nonbridging oxygen, in which these oxygen are bound to
OH groups39; this band, 980 cm−1, is absent in the Ra-
man spectrum of fused quartz, because in this material,
the four oxygen are bridging atoms. For sol-gel-made
samples annealed at higher temperatures the D1 band
appears due to the dehydroxylation process. This process
transforms some sixfold rings onto three- and fourfold
rings.37,38
In Fig. 4 it can be observed that the Raman spectra of
the sol-gel sample with copper is quite different from the
corresponding fused quartz sample. The spectrum of the
100-°C-treated sample has as the main Raman feature an
intense and sharp peak at about 1050 cm−1, which cor-
−
responds to vibrations of the NO3 ions or forming ni-
trates. The other peaks in the range of 250 and 949 cm−1
have the same origin. Notice that, in agreement with the
IR results, the intensity of this band is strongly reduced
after the annealing at 200 °C and it disappears for an-
nealings above 400 °C. Also, in the Raman spectrum of
the 200-°C-treated sample appears a band at approxi-
mately 288 cm−1, and weak peaks at approximately 350
and at approximately 430 cm−1. These features are as-
cribed to vibrations of Cu–O bonds in cupric oxide.40 In
the same spectrum, two broad bands appear, one in the
region of 400 to 700 cm−1 and the other above 850 cm−1.
On top of this latter band, weak peaks at 918, 949, and
973 can be resolved. The Raman spectra of the samples
Fig. 3. IR spectra for a thermal-growth SiO2 sample and for silica
xerogel sample with composition 70(SiO2)–30(CuO), annealed at
600 °C for 15 min.
2878
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 23 Mar 2015
http://journals.cambridge.org
IP address: 128.218.248.209

F. Ruiz et al.: Formation of silicate structures in Cu-containing silica xerogels
annealed at 400 and 600 °C show basically the same
features, but the broad bands are better defined, particu-
larly in the 600-°C-annealed sample.
Figure 5 shows the x-ray pattern of the samples an-
nealed at 200, 400, and 600 °C. In the sample annealed at
200 °C, the sharp lines have been identified with the
partially hydrolyzed Cu2(OH) и 3(NO3) copper nitrate
phase. For the 400- and 600 °C-annealed samples, the
nitrate phase disappears and the CuO phase appears.41
presence of the Cu incorporated into the glass matrix.
The latter band is centered at about 950–980 cm−1. Ra-
man vibrations in this frequency range have been asso-
ciated with SiO4 units with two nonbridged oxygens
(NBOs).39 SiO4 units with two NBOs may form different
kinds of cyclosiloxanes, with structures formed basically
by rings with three and six members.12 Because threefold
rings do not exist in the presence of large amounts of
OH groups, the main contribution for the 850–1050
cm−1 band could come from SiO4 units with two NBOs
forming rings of six members. The SiO4 units with two
NBOs are the only constant factor in the structure of
most silicates. The high-frequency Raman line in crys-
talline and glass silicates has been related to symmetric
silicate stretching motions involving the nonbridged oxy-
gens, and the low frequency to vibrations of the bridged
oxygens.42,43
To our knowledge, there are only two stable copper
silicates, one the dioptase and the other the shattuckite
with the composition Cu5(SiO3) и 4(OH), which is a
single-width unbranched inosilicate with the structural
formula (SiOx)3+. According to previous work,42 silicate
molecular groups of the form (SiOx)3+ should give an
additional Raman line at about 700 cm−1, which we do
not observe in our Raman results. Therefore, compar-
ing our Raman and IR results with previous reports, it
is quite possible that in the samples studied in this
work there exists the formation of small units that
have the structure of dioptase, because one of the es-
sential features in the structure of this silicate is that
the sixfold rings are present in the structure of silica
xerogel.14,37
IV. DISCUSSION
The main features in the IR and Raman spectra of
Cu-doped samples annealed at 100 and 200 °C are the
nitrate band. For samples annealed to 400 and 600 °C the
nitrates bands disappear and most of the copper segre-
gates to form CuO particles embedded in the glass ma-
trix. In the IR spectra of samples containing copper (Fig.
1), it is clear that the relative intensities of the Si–(OH)
bands at 940–970 cm−1 and of the Cu–O band at around
530 cm−1 increase with the increase in the annealing
temperature. The presence of the Si–(OH) groups even
after the treatment at 600 °C is contrary to the annealing
behavior of similar samples without copper. In these lat-
ter samples, the Si–(OH) groups gradually disappear with
the increase in the annealing temperature, being basically
removed for temperatures around 600 °C.
Several previous reports have pointed out that most
Raman spectra of silicates have major bands near 600
and 1000 cm−1.42,43 The spectra of crystalline silicates
are similar to that of the corresponding glasses, but with
narrower lines.42,44,45 The Raman results, in Fig. 4 for
samples treated at 200, 400, and 600 °C, show two weak
but well-defined bands in the ranges of 400–700 cm−1
and 850–1050 cm−1, which may be associated with the
As has been mentioned above, six member rings are
the basic structural units of the metasilicate known as
dioptase. These rings can trap the water molecules that
are observed in the IR spectra, being stable to heat treat-
ments at elevated temperatures.
The fact that the two Raman lines related to the for-
mation of structural units of silicates are weak and broad
means that in our samples the silicate formation, perhaps
around the oxide particles, is still incipient, and therefore
without any long-range structural order, the reason why
they do not show the characteristic x-ray and/or IR sig-
nals of micro- or nanocrystalline dioptase. The broadness
of these two Raman lines also agrees with the calcula-
tions by Rino et al.14 They have shown that the Si–O–Si
bond angle distribution for nonplanar six-member rings
is wide, and therefore, wide bands should be expected
in the Raman spectra. The situation is different for the
planar threefold and quasi-planar fourfold rings. In
these cases, their vibrations are decoupled and the
bond angle distributions are narrower, producing sharp
peaks in the Raman spectrum at frequencies of about
606 and 490 cm−1, for the three- and fourfold rings,
respectively.29,34–36
Fig. 5. X-ray pattern for the 70(SiO2)–30(CuO) sample annealed at
200 °C, 400 °C, and 600 °C.
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 23 Mar 2015
2879
IP address: 128.218.248.209
http://journals.cambridge.org

F. Ruiz et al.: Formation of silicate structures in Cu-containing silica xerogels
16. P. Sen and M.F. Thorpe, Phys. Rev. B 15, 4030 (1979).
V. CONCLUSIONS
17. F.L. Galeener, Phys. Rev. B 15, 4292 (1979).
18. A. Duran, C. Serna, V. Fornes, and J.M. Ferna´ndez-Navarro,
J. Non-Cryst. Solids 63, 45 (1984).
19. N.B. Colthup, L.H. Daly, and S.E. Wiberly, Introduction to Infra-
red and Raman Spectroscopy, 2nd ed. (Academic Press, New
York, 1975).
20. J.R. Mart´ınez, F. Ruiz, Y.V. Vorobiev, F. Pe´rez-Robles, and
J. Gonza´lez-Herna´ndez, J. Chem. Phys. 109, 7511 (1998).
21. J.R. Mart´ınez, F. Ruiz, J.A. de la Cruz-Mendoza, P. Villasen˜or-
Gonza´lez, J. Gonza´lez-Herna´ndez, M.M. Gonza´lez-Cha´vez, and
L. Valle-Aguilera, Rev. Mex. F´ıs. 44, 575 (1998).
22. P. Lange, J. Appl. Phys. 66, 201 (1989).
It is well known that silica xerogels have an open
structure formed basically by regular n-fold rings with
n ס 4, 5, and 6. When these glasses are heated for
densification, the ring statistics change, forming rings
with a low number of members including threefold
rings. In samples containing copper, the metal avoids the
dehydroxylation process preserving the open (polymeric)
structure and allowing the formation of metal silicates.
From our analysis, we propose that upon thermal anneal-
ing at temperatures above 200 °C, there is the formation
of small silicate structures corresponding to the dioptase.
23. S.B. Ogale, P.G. Bilukar, N. Mate, S.M. Kanetkar, N. Parikh, and
B. Patnaik, J. Appl. Phys. 72, 3765 (1992).
24. C.T. Kirk, Phys. Rev. B 38, 1255 (1988).
25. F.A. Seifert, B.O. Mysen, and D. Virgo, Am. Mineral. 67, 696
(1982).
REFERENCES
1. C.J. Brinker and G.W. Scherer, Sol-Gel: The Physics and Chem-
istry of Sol Gel Processing (Academic Press, New York, 1990).
2. F. Orgaz and H. Rawson, J. Non-Cryst. Solids 82, 390 (1986).
3. M. Nogami, in Sol-Gel Optics, Processing and Applications, ed-
ited by L.C. Klein (Kluwer Academic, Hingham, MA, 1994).
p. 329.
4. J-H. Gwak, S-J. Kim, and M. Lee, J. Phys. Chem. B 102, 7699
(1998).
5. M. Iwamoto, Stud. Surf. Sci. Catal. 54, 121 (1990).
6. M. Shimokawage, N. Takezawa, and H. Kobayashi, Appl. Catal.
2, 379 (1982).
26. M.R. Vukcetich, J. Non-Cryst. Solids 48, 31 (1982).
27. R.K. Iler, The Chemistry of Silica (Wiley, New York, 1979).
28. G. Lucovsky and M.J. Manitini, J. Vac. Sci. Technol. B 5, 530
(1987).
29. F.L. Galeener, Solid State Commun. 44, 1037 (1982).
30. Pasquarello and R. Car, Phys. Rev. Lett. 80, 5145 (1998).
31. P. Dean, Nature, 210, 257 (1966).
32. F. Galeener, R.A. Barrio, E. Mart´ınez, and R.J. Elliot, Phys. Rev.
Lett. 53, 2429 (1984).
33. A. Zun˜iga, F. Ruiz, C. Va´zquez, J. Gonza´lez-Herna´ndez,
G. Torres, and D.V. Tsu, J. Vac. Sci. Technol. A 12, 1445 (1994).
34. D.M. Krol and J.G. van Lierop, J. Non-Cryst. Solids 63, 131
(1984).
7. J. Yan, A.M. Buckley, and M. Greenblatt, J. Non-Cryst. Solids
180, 180 (1995).
8. M.G.F. da Silva and J.M. Fernandez Navarro, J. Non-Cryst.
Solids 100, 447 (1988).
35. D.A. Armitage, Inorganic Rings and Cages (Edward Arnold, Lon-
don, 1972).
9. J.F. Pe´rez-Robles, F.J. Garc´ıa-Rodr´ıguez, J.M. Yan˜ez-Limo´n,
F.J. Espinoza-Beltra´n, Y.V. Vorobiev, and J. Gonza´lez-
Herna´ndez, J. Phys. Chem. Solids 60, 1729 (1999).
10. S.B. Ogale, P.G. Bilukar, N. Mate, S.M. Kanetkar, N. Parikh, and
B. Patnaik, J. Appl. Phys. 72, 3765 (1992).
11. G. Cordoba, R. Arroyo, J.L.G. Fierro, and M. Viniegra, J. Solid
State Chem. 123, 93 (1994).
12. A.F. Wells, Structural Inorganic Chemistry (Clarendon Press,
Oxford, 1978).
13. N.V. Belov, B.A. Maksirnov, Yu. Z. Noskik, and L.A. Muradyan,
Sov. Phys. Dokl. 23, 215 (1978).
14. J.P. Rino, I. Ebbsjo¨, R.K. Kalia, A. Nakana, and P. Vashishta,
Phys. Rev. B 47, 3053 (1993).
15. J.F. Pe´rez-Robles, L.A. Garc´ıa-Cerda, F.J. Espinoza-Beltra´n,
M. Ya´nez-Limo´n, J. Gonza´lez-Herna´ndez, Y.V. Vorobiev, and
J.R. Parga-Torres, F. Ruiz, and J. Me´ndez-Nonell, Phys. Status
Solidi 172, 49 (1999).
36. R.A. Barrio, F.L. Galeener, E. Martinez, and R.J. Elliott, Phys.
Rev. B 48, 15672 (1993).
37. B. Riegel, I. Hartmann, W. Kiefer, J. Gro␤, J. Fricke, J. Non-
Cryst. Solids 211, 294 (1997).
38. F. Pe´rez-Robles, F.J. Garc´ıa-Rodr´ıguez, S. Jimenez-Sandoval, and
J. Gonza´lez-Herna´ndez, J. Raman Spectrosc. 30, 1099 (1999).
39. K. Nakamoto, Infrared and Raman Spectra of Inorganic and Co-
ordination Compounds, 5th ed. (John Wiley, New York, 1997).
40. J. Chrzanowski and J.C. Irwin, Solid State Phys. 70, 11 (1989).
41. International Center for Diffraction Data, Inorganic Phases (Joint
Committee on Powder Diffraction Standards, Swarthmore, PA,
1995), Powder Diffraction File, Card No. 41-0254.
42. P. McMillan and B. Piriou, Bull. Mineral 106, 57 (1983).
43. T. Furukawa, T. Fox, and W.B. White, J. Chem. Phys. 75, 3226
(1981).
44. P. McMillan and B. Piriou, J. Non-Cryst. Solids, 53, 279 (1982).
45. D. Virgo, O.B. Mysen, and I. Kushiro, Science 208, 1371 (1980).
2880
J. Mater. Res., Vol. 15, No. 12, Dec 2000
Downloaded: 23 Mar 2015
http://journals.cambridge.org
IP address: 128.218.248.209