2248 Inorg. Chem.

1997, 36, 2248-2251

Articles
Dicesium trans-Tetraaquadichlorochromium(III) Chloride: Redetermination of the Crystal
Structure and Infrared Study of the Water Spectrum

Stanley C. Nyburg,*,1a Bojan Šoptrajanov,*,1b Viktor Stefov,1b and Vladimir M. Petruševski1b

Department of Chemistry, King’s College, University of London, Strand, London WC2R 2LS, England,
and Institut za hemija, PMF, Univerzitet “Sv. Kiril i Metodij”, PO Box 162, 91001 Skopje, Macedonia
ReceiVed July 3, 1996X

Infrared spectral studies of the solid led us to believe that the published crystal structure of dicesium trans-
tetraaquadichlorochromium(III) chloride, trans-Cs2[CrCl2(H2O)4]Cl3 might be incorrect. Crystal data: Cs2CrCl5‚
4H2O, a ) 17.484(9) Å, b ) 6.099(3) Å, c ) 6.928(3) Å, β ) 106.06(5)°, monoclinic, C2/m, Z ) 2 molecules
per cell. The redetermination has revealed disorder in the positions of the water molecules. Instead of one type
of H2O molecule being present as found in the original study, two sets of such molecules with four nonequivalent
O‚‚‚Cl contacts were found. The presence, in the O-D stretching region of the spectra of samples with low
deuterium content, of three bands with intensities close to 2:1:1 (rather than the expected four) is believed to be
a consequence of different degrees of nonlinearity of the two hydrogen bonds formed by the water molecules of
one of the two existing types.

Introduction Diffraction data were collected at 24 °C on an automated four-circle

Picker diffractometer using Zr-filtered Mo KR radiation (λ ) 0.7107
Data on the infrared spectra of the title compound (of both Å) with pulse height analysis to a limiting 2θ of 50° covering the ranges
protiated and partially deuterated samples) have already been h ) (18, k ) 0-7, and l ) 0-8 (see Table 1 for details). Of 686
reported by us.2 While it was found that our spectrum of the reflections measured, 660 were significant on the criterion Inet >
protiated trans-Cs2[CrCl2(H2O)4]Cl3 agrees well with the spectra 2.5σ(Inet). The systematic absences hkl for h + k ) 2n + 1 were
reported in 1983 by Michalska-Fong et al.,3 the spectral picture confirmed, thus allocating the space group to C2, Cm, or C2/m.
in the O-D stretching region of the spectra of isotopically McCarthy et al.4 assumed C2/m, and we used this assignment also.
isolated HOD molecules was incompatible with the reported Using NRCVAX computer software, after applying Gaussian absorption
X-ray structure analysis of the crystal.4 This analysis has corrections to the intensity data5 (µ ) 67.4 cm-1; minimum and
maximum transmission factors 0.18 and 0.36), we used the fractional
Cr(III) hexacoordinated to four equivalent water molecules
atomic coordinates and anisotropic temperature factors listed by
located in the equatorial plane and to two axial chlorine atoms. McCarthy et al.4 to carry out refinement, varying only the scale factor
Thus only two ν(O-D) bands should be expected in the and extinction coefficient. The resulting residual was somewhat better,
spectrum of the samples with low deuterium content. The at 0.070.6
number of observed bands led us to conclude that either a wrong A ∆F Fourier map showed immediately that, instead of four
space group had been assigned to the crystal or else the equatorial oxygen atoms associated with Cr, there were, in fact, eight
coordinated water molecules were disordered in some way not partially occupied sites. The eight oxygen atoms had their occupancies
revealed by the crystal structure analysis. Our view that the set to 0.5. Full-matrix anisotropic least-squares refinement was
published crystal structure might be wrong was strengthened uneventful, the final residual being 0.041. Attempts were made to refine
by the fact that the residual R was somewhat high (at 0.083) the structure (with appropriately modified occupancies, etc.) in space
groups C2 and Cm. In the former, the oxygen atoms could not be
but even more because the anisotropic temperature factors for
refined anisotropically, and the latter led to a false minimum showing
the oxygen atoms were an order of magnitude greater than would high, but chemically unfeasible, disorder.
be expected for an ordered structure. We report here a The positions for hydrogen atoms could not be determined with any
redetermination of the crystal structure showing that the certainty, but the most likely positions for them appear to be those
coordinated water molecules are, in fact, disordered. which allow both O1 and O2 to be hydrogen bonded to Cl1 and Cl3.
(Because the oxygen atoms are disordered, they cannot be simulta-
Experimental Section neously bonded this way in any one Cr coordination polyhedron). The
Green crystals of the title compound were from the same sample as geometries then are as follows: O1-Cl1, 3.009 Å; O1-Cl3, 3.045 Å;
those used in the spectral analysis.2 The crystal chosen for X-ray Cl1‚‚‚O1‚‚‚Cl3 angle, 88.7°; O2-Cl1, 3.086 Å; O2-Cl3, 3.178 Å;
analysis was columnar, 0.08 × 0.1 × 0.4 mm, and preliminary X-ray Cl1‚‚‚O2‚‚‚Cl3 angle, 85.0°.
photographs showed it to be of good quality. Final fractional atomic coordinates are given in Table 2. Figure 1
shows an ORTEP diagram7 of the coordination sphere, and Figure 2
X Abstract published in AdVance ACS Abstracts, April 15, 1997. shows a packing diagram with putative hydrogen bonds marked in.
(1) (a) King’s College, London. (b) Institut za hemija, Skopje.
(2) Šoptrajanov, B.; Stefov, V.; Petruševski, V. M. Spectrosc. Lett. 1993, (5) Gabe, E. J.; Le Page, Y.; Charland, J.-P.; Lee, F. L.; White, P. S. J.
26, 1839. Appl. Crystallogr. 1989, 22, 384.
(3) Michalska-Fong, D.; McCarthy, P. J.; Nakamoto, K. Spectrochim. Acta (6) In order to obtain the correct Uij factors from the βij values as published,
1983, 39A, 835. those with i * j had to be divided by two.
(4) McCarthy, P. J.; Lauffenburger, J. C.; Skonezny, P. M.; Rohrer, D. (7) Johnson, C. K. ORTEP; Report ORNL-3794, Oak Ridge National
C. Inorg. Chem. 1981, 20, 1566. Laboratory: Oak Ridge, TN, 1965.

S0020-1669(96)00790-2 CCC: $14.00 © 1997 American Chemical Society

Cs2CrCl5‚4H2O Inorganic Chemistry, Vol. 36, No. 11, 1997 2249

Table 1. Crystal Data

empirical formula Cs2CrCl5‚4H2O fw 567.13
a ) 17.484(9) Å (17.604(1))† space group C2/m No. 12
b ) 6.099(3) (6.140(1))a T ) 24 °C
c ) 6. 928(3) (6.979(1)) λ ) 0.7107 Å
β ) 106.06(5)° (106.040(7)) Fcalcd ) 2.653 g cm-3
V ) 709.9(6) Å3 µ ) 67.4 cm-1
Z)2 R(Fo) ) 0.0403b
Rw(Fo) 0.0530b
a
Parameters in parentheses are from McCarthy et al.4 Differences
of the order 0.7% could be due to a higher ambient temperature. b R(Fo)
) ∑(Fo - |Fc|)/∑Fo; Rw(Fo) ) (∑(w(Fo - |Fc|)2)/∑(wFo2))1/2 with w
based on counting statistics.

Table 2. Atomic Parameters and Biso Values, Where Esd’s Refer to

the Last Digit(s) Printed
x y z Bisoa (Å2)
Cs 0.29779(5) 0 0.25146(12) 2.84(6)
Cr 0 0 0 1.72(11)
Cl1 0.19339(19) 1/ 0.2433(5) 2.71(14)
2
Cl2 -0.10693(19) 0 0.1296(5) 2.55(14)
Cl3 0 1/ 1
/2 2.72(20) Figure 3. Left: (a) View normal to plane containing oxygen atoms
2
O1 0.0329(8) 0.2810(24) 0.1362(21) 2.5(6) (as reported by McCarthy, et al.4). (b) Right: Disordered, structure,
O2 0.0647(7) 0.153(3) 0.2410(19) 2.5(6) showing that the rotational sense of disordering must be preserved in
any one column.
a Biso is the mean of the principal axes of the thermal ellipsoid.
Table 3. Selected Interatomic Distances (Å) and Bond Angles
(deg)a
Cs-Cl1 3.546(2) Cs-Cl3 3.478(2)
Cs-Cl1a 3.461(4) Cr-Cl2 2.287(3)
Cs-Cl2c 3.683(2) Cr-O1 1.964(13)
Cs-Cl2b 3.637(4) Cr-O2 1.972(13)
O1-Cl1 3.005(13) O2-Cl1 3.086(13)
O1-Cl3 3.045(13) O2-Cl3 3.178(13)
O1-Cr-O2d 91.0(7) O2-Cr-Cl2 90.4(3)
O1-Cr-Cl2 88.5(3)
Cl1-O1-Cl3 88.7(10) Cl1-O2-Cl3 85.0(9)
a 1/ 1/
Symmetry operations: (a) - x,
2 2 - y, 1 - z; (b) -x, -y, -z;
(c) 1/2 + x, -1/2 + y, z; (d) -x, y, -z.

nitrogen was employed as a cooling agent. The temperature was

maintained at 100 K, which is close to the lowest temperature attainable
with this system.
The software package GRAMS 20008 was used in the spectra
acquisition, and GRAMS/3869 was used to analyze the spectra.

Figure 1. ORTEP plot7 of the coordination sphere showing 50% Discussion

probability ellipsoids. The Cr atoms lie along 0, y, 0 and 1/2, y, 0 equally spaced at
one cell dimension (i.e. b) apart with their Cl2-Cr-Cl2 polar
axes all parallel in the xz-plane (see Figures 2 and 3). The O
atoms lie in planes containing the Cr atoms and normal to the
polar axes. Had the oxygen atoms been ordered (as in the
published structure4), they would be disposed as in Figure 3a.
The disordered structure presents the O atoms with two choices,
either a clockwise or a counterclockwise rotation (of 28.2°)
about the polar axes through Cr (Figure 3b). Whichever sense
of rotation is adopted for a particular row along y, it must be
the same throughout the entire row. This is because any change
in the sense of rotation at a point along the row would result in
an unacceptably short O‚‚‚O distance. Figure 3 provides the
Figure 2. Packing diagram showing putative hydrogen bonds. (Only geometric details, and Table 3 gives selected interatomic
two of the four shown can be present together.) distances and angles.
The existence of two crystallographically different types of
The Fourier-transform infrared (FTIR) spectra were recorded on a water molecule is consistent with the appearance (both at room
Perkin-Elmer System 2000 FT-IR (resolution 2 cm-1, OPD rate 0.2 and at liquid-nitrogen temperature) of two bands due to
cm/s, 32 background and 64 sample scans) from pellets in CsCl. Mulls
in Nujol between CsBr plates were also used. The spectra were
(8) GRAMS ANALYST for PE 2000 FT-IR, Version 3.01B, Galactic
identical, thus guaranteeing that no reactions occurred during pellet Industries Corp., 1991-1994.
preparation. For low-temperature work, a P/N 21525 low-temperature (9) GRAMS/386 for Microsoft Windows, Version 2.02, Galactic Industries
cell (Graseby Specac) with a temperature controller was used. Liquid Corp., 1991-1993.
2250 Inorganic Chemistry, Vol. 36, No. 11, 1997 Nyburg et al.

Table 4. Spectral Parameters (Obtained by Curve-Fitting) for the

OD Stretching and HOD Bending Bands of Isolated HDO
Molecules in the Difference Liquid-Nitrogen-Temperature IR
Spectra of trans-Cs2[CrCl2(H2O)4]Cl3
ν/cm-1 height/arbitrary units width/cm-1 area/arbitrary units
2427.6 0.1838 16.7 4.0751
2322.1 0.1428 13.2 2.8232
2300.6 0.1228 14.1 2.3532
1462.4 0.0204 13.0 0.3758
1445.3 0.0191 13.7 0.2924
1439.1 0.0169 11.9 0.2495
1428.6 0.0316 10.9 0.4340

Figure 4. HOH bending region in the spectrum of isotopically isolated

H2O molecules in practically perdeuterated Cs2[CrCl2(H2O)4]Cl3: (Top)
room temperature; (bottom) liquid-nitrogen temperature.

Figure 6. Difference spectrum of slightly deuterated (≈2% D) and

protiated Cs2[CrCl2(H2O)4]Cl3 in the O-D stretching region: (top)
actual difference spectrum; (middle) spectrum obtained by curve-fitting;
(bottom) component profiles of the fitted spectrum.

the bending vibrations of water isotopomers in crystalline

hydrates.11 The reasons for such behavior remain unclear.
Since the true positions of the hydrogen atoms are unknown,
Figure 5. FT-IR liquid-nitrogen temperature difference spectrum of the geometry and the details for the bonding of the water
slightly deuterated (≈2% D) and protiated Cs2[CrCl2(H2O)4]Cl3 in the molecules could not be determined with any degree of confi-
region of the HOD bendings: (Top) actual difference spectrum;
(middle) spectrum obtained by curve-fitting; (bottom) component dence by crystallographic means. It is here that the infrared
profiles of the fitted spectrum. spectra, particularly those of samples containing isotopically
isolated HOD molecules, may shed some light.
H-O-H bending (Figure 4) in the spectra of samples with a The appearance (Figure 6) of three distinct bands in the O-D
high deuterium content. As has been pointed out,2 under such stretching region (the one at highest frequency being more
circumstances the H2O molecules are isotopically isolated by intense) suggests that the four crystallographically distinct
HOD and D2O molecules and all interactions between identical hydrogen bonds are grouped in pairsstwo of them must be
oscillators (correlation-field effects) are excluded. The two stronger than the remaining ones. That this is indeed so is
bands then must originate from vibrations of two distinct types shown by the ratio of the measured intensities of these three
of water molecule. This is confirmed by the analysis of the bands (1.73:1.20:1) which is close to the ideal 2:1:1 ratio
expected for a quartet with two overlapping bands. The half-
HOD bending region. For two types of water molecule, four
width of the 2427.6 cm-1 band is larger (at approximately
types of half-deuterated molecules should be present (with each
17 cm-1) than that of the two bands at lower frequencies (at
of the protons being alternatively substituted by deuterons). In
13 and 14 cm-1, respectively). Qualitatively, this is consistent
the difference spectrum10 of the slightly deuterated sample, a
with the hydrogen-bonding scheme outlined above since the
complex feature with three maxima is observed (Figure 5), but O1-Cl distances were found to be shorter than the O2-Cl ones
the envelope could be better reproduced by assuming that it and, furthermore, the two O1-Cl contacts differ slightly from
consists of four components (Table 4). each other. It is then safe to conclude that the bands at 2322
As seen from Figures 4 and 5, as well as from Table 4, the and 2301 cm-1 originate from O-D stretching modes of
intensities of the individual components of the HOH/HOD isotopically isolated H-O1-D molecules. What remains
bending modes vary considerably. This is, however (perhaps unclear is why there is only a single (albeit stronger and broader)
contrary to the intuitive expectations), a common behavior of band at 2427.6 cm-1, contrary to the fact that the difference
between the two O2-Cl distances (≈0.09 Å) is much larger
(10) The spectrum of the protiated compound, appropriately scaled, was
subtracted from the spectrum of the deuterate. The same procedure (11) Cvetković, J.; Petruševski, V. M.; Šoprajanov, B. J. Mol. Struct., in
was performed in the region of the OD stretching bands. press.
Cs2CrCl5‚4H2O Inorganic Chemistry, Vol. 36, No. 11, 1997 2251

than the difference between the two O1-Cl distances. Our Full crystal data and anisotropic temperature factors are
attempts to resolve (either experimentally or by curve-fitting provided as Supporting Information.
procedures) this band into two components were unsuccessful Acknowledgment. One of us (V.S.) gratefully acknowledges
so that the frequency difference between the components must the financial support by the British Council.
be less than some 7 cm-1. One of the possible explanations
for the discrepancy between our expectations and the experi- Supporting Information Available: Tables of crystal data and
mental results is the conjecture that the two O2‚‚‚Cl2 hydrogen anisotropic temperature factors (2 pages). Ordering information is given
on any current masthead page.
bonds deviate to a different degree from linearity and that thus
they are, accidentally, almost identical in strength. IC9607907