G.N. Papatheodorou / Journal of Molecular Structure 828 (2007) 102–110
109
agreement between experimental and calculated values is
found for the spin allowed ligand field transitions arising
structural models a, b, and c in Fig. 1 and the ‘‘tetrahedral’’
coordination (versus, e.g., a square planar) is supported, at
least for Cr(III), from the above analysis of the absorption
spectra. On the other hand the most probable structural
models for the dimer A M X , trimer A M X , etc., more
likely involve mixed ‘‘tetrahedral’’ and ‘‘octahedral’’ coor-
dinations of M(III). This is shown in Fig. 1 with examples
for the dimer (d) and a trimer (e).
In the case of Cr(III) both the ‘‘octahedra’’ and ‘‘tetra-
hedra’’ present in the dimers (Fig. 1) should contribute to
the absorption spectra of the vapors. On the other hand
since the molar absorptivity of the ‘‘octahedra’’ is lower
than that of the ‘‘tetrahedra’’ and since the dimers are more
likely the minor components, then the absorption spectra
should be dominated from the ‘‘tetrahedral’’ absorption
bands. In other words the ‘‘tetrahedral’’ spectra in Fig. 2
cannot exclude the presence of more than one ‘‘tetrahe-
4
4
from the F and P free ion states. These transitions depend
only on the DqTd and BTd parameters which confirms the
correctness of the estimation of the parameters as listed
in Table 4. Some disagreement, however, exists for the
intercombination bands (spin forbidden states) arising
2
2
8
3
3
12
2
from the splitting of the G free ion states. The energies
2
2
2
2
of the resulting E(G), T (G), T (G) and A (G) terms
1
2
1
depend on the DqTd and BTd parameters and in addition
on the CTd Racah parameter which determines the overall
2
4
4
position of the G relative to P and F free ion states. Thus
it seems that the origin of disagreement is due to inaccura-
cies in the estimation of the CTd parameter.
Finally, the spectra of vapors over both the LiCl–CrCl3
and the CsCl–CrCl3 systems show bands near the UV
region marked as VI and VII (Fig. 2 and Table 2). These
bands may arise from either the splitting of the higher
dral’’ species in the vapor phase; i.e., both ACrX and
4
2
2
2
energy H, P or D free ion states or are related to the elec-
tron donating/accepting properties of the ligand (charge
transfer bands, CT). In the first case the transitions are spin
forbidden and weak bands are expected; thus with molar
A Cr X or even higher polymers may be present. This
2
2
8
would imply that the volatility enhancements in Table 1
are underestimated.
The presence of more than one vapor species over the
ꢀ
1
ꢀ1
absorptivities higher than 100 L mol cm
it is more
ACl–CrCl systems is also supported from a thermodynamic
3
likely that these high energy transitions have a charge
transfer origin.
and a LFSE (Section 3.4) point of view. The Raman spectro-
scopic studies in molten CsCl–ScCl mixtures have shown
3
In the vapor molecule LiCrCl the T symmetry of the tet-
that near the 1:1 composition an equilibrium takes place:
4
d
rahedral is expected to be distorted by the field of the neigh-
boring Li cation. Depending on the position of Li along the
edge, the face or the corner of the tetrahedra (Fig. 1a–c) the
3ꢀ
¼ 3ScCl4ꢀ þ ScCl63ꢀ
DHSc ¼ 38 kJ molꢀ1;
+
2Sc Cl
;
2
9
ð6Þ
symmetry will be lowered to C3v or C . These lower symme-
2v
where the dimer structure consists of two ‘‘octahedra’’
bound by a face [16]. The vaporization of the 1:1 melt mix-
ture to form CsScCl would imply that ScCl leaves the
melt and thus the above equilibrium should shift to the
right. On the average the equilibrium transforms three
tries should further split the degenerate tetrahedral ligand
field terms to sets of component states. At elevated tempera-
tures these components would be rather difficult to resolve in
the spectra resulting broader bands centered around the
energy of the degenerate tetrahedral term. The spectra in
ꢀ
4
4
‘
‘octahedra’’ to three ‘‘tetrahedra’’ and in the absence of
Fig. 2 support this view and show that the degenerate
4
LFSE the enthalpy of the above reaction is mainly associ-
ated to breaking the bridging bonds of the dimer and low-
ering the coordination. If, as discussed in Section 3.4, the
structure of molten ACl–CrCl (1:1) is similar to that of
the corresponding ScCl melts, then an analogous equilib-
T (P) band possesses a broader maximum and a higher
1
4
half-width relative to the non-degenerate A (F) band.
1
In conclusion the above analysis establishes for the first
time the near IR/Vis ligand field states of an all chloride
coordinated chromium (III) in nearly tetrahedral
symmetry.
3
3
rium (6) for the ACl–CrCl system should be influenced by
3
the ‘‘octahedra’’/‘‘tetrahedra’’ LFSE of Cr(III). From the
Dq values in Tables 3 and 4 the LFSE of the two different
geometries can be estimated yielding the enthalpy for the
coordination change:
3
.7. The multi-structure of the 1:1 vapor species
The above data and discussion infer that the vaporiza-
CrCl6 ¼ CrCl4ꢀ þ 2Clꢀ; DHLF ¼ 130 kJ molꢀ1:
3
ꢀ
ð7Þ
tion occurs easier from the 1:1 melt and that the same stoi-
chiometry is kept in the vapor. The absorption spectra
indicate that a ‘‘tetrahedrally’’ coordinated Cr(III) is pres-
ent in the gas phase thus the most probable vapor species is
And for corresponding equilibrium in the melt we may
write:
3
ꢀ
2Cr Cl
9
¼ 3CrCl4ꢀ þ CrCl6
3ꢀ
;
the ACrX monomer. Such molecules have been found to
4
2
be the predominant vapor species over a large number of
DHCr ¼ DHSc þ 3DHLF ¼ 430 kJ molꢀ1:
ð8Þ
AX–MX molten systems [2–5]. Mass spectrometric studies
3
have also shown that apart from AMX other species like
Thus the Cr(III) evaporation from the melt to form the 1:1
monomer ACrCl requires that equilibrium (8) shift to the
4
the dimers A M X and/or the five coordinated A MX
5
4
2
2
8
2
species exist as minor components [2,5]. The structure of
right, a process which is inhibited by the high endothermic
enthalpy. On the other hand if the vaporization leads in
the monomer AMX can be easily visualized from the three
4
Papatheodorou
