three IR active modes, and presents a broad multiple maximum near 500 cm
− 1
,
with weaker components near 575 and 160 cm
− 1
. Monoclinic zirconia (baddeleyite)
has a much more complex spectrum, with 15 IR active modes. In particular it is
well characterized by a strong band evident at 770 cm
− 1
, associated with the move-
ment of oxygen in threefold coordination.
IR spectroscopy allows the analysis of the many polymorphs of manganese
dioxide [54] such as α - MnO
2
(hollandite, tetragonal, SG I 4/ m ), R - MnO
2
(ramsdel-
lite, orthorhombic, Pbnm ), β - MnO
2
(pyrolusite, tetragonal rutile structure, P 4
2
/
mnm ), γ - MnO
2
(nsutite, hexagonal), δ - MnO
2
(vernadite, hexagonal) and λ - MnO
2
(spinel - type, cubic, Fd 3 m ).
α - and β - quartz, α - and β - cristobalite, tridymites, coesite, morganite and sili-
calite are different polymorphs of silica with tetrahedrally coordinated silicon. So,
their vibrational structure, as well as that of amorphous or vitreous silica, in which
Si is also tetrahedral, can be discussed in parallel. The IR spectra of all tetrahedral
silica polymorphs have bands in three regions: in the region 1300 – 950 cm
− 1
(very
strong, denoted ν
1
), 850 – 600 cm
− 1
(medium strength, ν
2
) and near 450 cm
− 1
(very
strong, ν
3
). Raman spectra show very weak peaks in the 1300 – 950 cm
− 1
and 850 –
700 cm
− 1
regions ( ν
1
and ν
2
modes), and one or more very strong peaks below
550 cm
− 1
, in a position very sensitive to the overall crystal structure of the solid
( ν
4
). These features are also present in the spectra of amorphous or vitreous silica;
so they represent vibrations of the basic structural units of both crystalline and
amorphous silica, that is, SiO
4
tetrahedra and bridging oxygens.
Four main vibrational modes are expected for a tetrahedral silica network. In
the C
2v
symmetry of the Si
–
O
–
Si bent units, a symmetric stretching (A
1
), an asym-
metric stretching (B
1
), a bending (A
1
) and a rocking mode (B
2
) are expected, the
modes with A
1
, B
1
and B
2
symmetry being both IR and Raman active. The stron-
gest Raman mode, ν
4
, is assumed to be a symmetric stretching mode, mainly
because Raman peak intensity is generally greater for stretching than for bending
modes. However, the coupling of this mode with the Si
–
O
–
Si scissoring mode
provides an explanation of the low frequency of the strongest Raman peak (550 –
350 cm
− 1
) in all silica polymorphs. On the other hand, it is clear that in the solid
state in order for all Si
–
O bonds to expand in - phase, most Si
–
O
–
Si angles must
contract simultaneously. This mode is consequently highly sensitive to the struc-
ture of the polymorph, as is indeed observed. The position of the ν
4
peak, in fact,
strongly depends on the type of ring present in silicas and silicates. As for the IR
spectrum, the strongest complex band is ν
1
, observed in the 1300 – 950 cm
− 1
region
and is associated to the asymmetric stretching of the Si
–
O
–
Si bridges. However,
this mode is split into two components even in the case of vitreous silica, due to
either the in - phase
′
()
ν
1
or the out - of - phase coupling
′′
()
ν
1
of the asymmetric
stretching modes of nearest Si
–
O
–
Si groups. In other words, this mode couples
with the symmetric and asymmetric stretching of the four Si
–
O bonds of the SiO
4
octahedra. The position of both
′
ν
1
and
′′
ν
1
components is also somewhat sensitive
to the size of the rings and/or to the density of the different silica polymorphs.
This is particularly true for
′
ν
1
, whose shift is stronger, so that it is actually resolved
from
′′
ν
1
only in some cases. As discussed above, the ν
2
mode is essentially a
3.4 The Skeletal IR Spectra of Metal Oxides 119