Synthesis and characterization of molybdenum(VI) oxide sulfates and crystal structures of two polymorphs of MoO2(SO4)

Article abstract of DOI:10.1039/b007675i

Reactions between α-MoO₃ or MoO₃·0.43H₂O and 65-95 wt% H₂SO₄ were studied from room temperature up to the boiling point of the acid. Four new compounds, MoO₂(SO₄), (modifications I, II and III) and MoO₂(SO₄)·H₂SO₄·H ₂O, have been isolated depending on the H₂SO₄ concentration, reaction temperature, reaction time and molybdenum source. The characterization of these phases was performed with powder X-ray diffraction, thermogravimetric analysis, differential thermal analysis, chemical analysis and density measurements. Both starting materials react with sulfuric acid and form MoO₂(SO₄) II at temperatures above ca. 75 °C. This compound is unstable in the reaction mixture and converts into MoO₂(SO₄) I or III depending on time and temperature. The three modifications of MoO₂(SO₄) are hygroscopic and decomposition in moist air is studied for the I and II modifications. On heating, all modifications of MoO₂(SO₄) decompose into α-MoO₃ or a mixture of α-MoO₃ and β-MoO₃ depending on heating rate and sample size. The three-dimensional, open-framework structures of I and II, have been determined from single-crystal X-ray diffraction data. I crystallizes in the monoclinic space group C2/c and II in the orthorhombic space group Pna2₁. Both structures are made up of MoO₆ octahedra and SO₄ tetrahedra and contain layers of eight- and four-membered rings of alternating, corner-sharing octahedra and tetrahedra. These layers are linked (also via Mo-S bridges) to form a three-dimensional framework. The MoO₆ octahedra are rather distorted, as demonstrated by large variations in the bonding Mo-O interatomic distances, which reflect the double-bond character of the binding between molybdenum and terminal oxygen atoms pointing into the eight-membered rings. The SO₄ tetrahedra are quite regular. The structural relation to MoO₂(SO₄) III is briefly considered.

Full text of DOI:10.1039/b007675i

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis and characterization of molybdenum(VI) oxide sulfates
and crystal structures of two polymorphs of MoO (SO )
2
4
Alexander F. Christiansen, Helmer Fjellvåg, Arne Kjekshus* and Bernt Klewe
Department of Chemistry, University of Oslo, PO Box 1033 Blindern, N-0315 Oslo, Norway
Received 21st September 2000, Accepted 16th January 2001
First published as an Advance Article on the web 1st March 2001
Reactions between α-MoO or MoO ؒ0.43H O and 65–95 wt% H SO were studied from room temperature up
3
3
2
2
4
to the boiling point of the acid. Four new compounds, MoO (SO ), (modifications I, II and III) and MoO (SO )ؒ
2
4
2
4
H SO ؒH O, have been isolated depending on the H SO concentration, reaction temperature, reaction time
2
4
2
2
4
and molybdenum source. The characterization of these phases was performed with powder X-ray diffraction,
thermogravimetric analysis, differential thermal analysis, chemical analysis and density measurements. Both
starting materials react with sulfuric acid and form MoO (SO ) II at temperatures above ca. 75 ЊC. This compound
2
4
is unstable in the reaction mixture and converts into MoO (SO ) I or III depending on time and temperature. The
2
4
three modifications of MoO (SO ) are hygroscopic and decomposition in moist air is studied for the I and II
2
4
modifications. On heating, all modifications of MoO (SO ) decompose into α-MoO or a mixture of α-MoO and
2
4
3
3
β-MoO depending on heating rate and sample size. The three-dimensional, open-framework structures of I and II,
3
have been determined from single-crystal X-ray diffraction data. I crystallizes in the monoclinic space group C2/c
and II in the orthorhombic space group Pna2 . Both structures are made up of MoO octahedra and SO tetrahedra
1
6
4
and contain layers of eight- and four-membered rings of alternating, corner-sharing octahedra and tetrahedra. These
layers are linked (also via Mo–S bridges) to form a three-dimensional framework. The MoO octahedra are rather
6
distorted, as demonstrated by large variations in the bonding Mo–O interatomic distances, which reflect the double-
bond character of the binding between molybdenum and terminal oxygen atoms pointing into the eight-membered
rings. The SO tetrahedra are quite regular. The structural relation to MoO (SO ) III is briefly considered.
4
2
4
Introduction
with, α-MoO
or MoO ؒ0.43H O. We describe the crystal struc-
3 3 2
tures of MoO (SO ) I and MoO (SO ) II, and their relations to
2
4
2
4
The oxides of molybdenum have been the subject of numerous
investigations back to the early history of chemistry. However,
the reactions in and with concentrated sulfuric acid are little
systematically explored and understood. This is attributed to
the ease of hydrolysis of the molybdenum sulfates and oxide
sulfates, and the fact that work with conc. H SO has certain
MoO (SO ) III. Tentative information concerning the struc-
2
4
tural arrangement of MoO (SO ) (unspecified modification)
has been published on the basis of infrared (IR) spectra. This
work represents a continuation of earlier studies on corre-
2
4
8
11–16
1
17
18
sponding phases of iodine,
titanium, tin, zirconium,
2
4
18
19
hafnium and tellurium at the University of Oslo.
disadvantages. Especially the high viscosity introduces experi-
mental difficulties. Solutes dissolve and crystallize slowly, and
adhered H SO may be difficult to remove from the reaction
Experimental
Synthesis
2
4
product. However, for many purposes the advantages of conc.
H SO may be more important (see Ref. 1).
2
4
The reaction between molybdenum() oxide and hot con-
centrated sulfuric acid was first described by Shultz-Sellack in
α-MoO (Fluka, 99.5%), MoO ؒ0.43H O (Fluka; water content
3 3 2
2
determined by thermogravimetric analysis) and conc. H SO4
2
1
871. The final product was designated MoO (SO ). This and
(Merck; 95–97 wt%; the former value being used throughout
this paper) were used as starting chemicals for the syntheses.
Concentrated H SO in the range 65–95 wt% H SO was made
2
4
products of partially modified reactions have been reported in
Refs. 3–8. Various sulfates and oxide sulfates with Mo , Mo
and Mo have been reported (see Refs. 9 and 10 and references
therein). The existence or composition of many of these com-
pounds is an open question. If we restrict our considerations
to molybdenum() oxide sulfates the selection is limited to
III
V
2
4
2
4
VI
by diluting the as-purchased acid with distilled water. The con-
centration of these acids was checked using a densitometric
flask and confirmed by comparing the measured densities with
20
published values. α-MoO or MoO ؒ0.43H O was added to 20
3
3
2
2
,4,8
4
6
MoO (SO ),
H MoO (SO ) ؒnH O, MoO(SO ) , Mo O -
ml conc. H SO (65–95 wt%) in a round-bottomed flask with
2
4
2
2
4
2
2
4
2
2
3
2 4
7
4
(
SO ) ؒ4H O and Mo O (SO )ؒnH O (see the brief survey in
either a glass stopper for room temperature (r.t.) reactions, or a
reflux cooler for reactions at higher temperature. The molar
ratio [Mo] : [H SO ] of the starting mixtures was for α-MoO
4
3
2
2
5
4
2
Ref. 10). The documented characteristics of these phases are
scarce. Many of these uncertainties seem to be connected with
the fact that too little emphasis has been put on the use of
sensitive and accurate methods for physical characterization of
the samples and verification of their homogeneity. In particular,
insufficient attention has been paid to the stability of these sub-
stances in relation to moisture and heat treatment. There exists
no structure determination or even powder X-ray diffraction
data for characterization.
2
4
3
1 : 9, 1 : 3 and 1 : 2 for 95, 85–90 and 65–80 wt%, respectively;
and for MoO ؒ0.43H O 1 : 6, 1 : 3 and 1 : 2 for 95, 80–90 and 65
3
2
wt%, respectively. The mixtures were stirred with a magnetic
stirrer and treated at different temperatures from r.t. to the boil-
ing point (bp) for periods from 15 min to 6 months. Then the
heating and stirring were turned off. The reaction vessel was
kept untouched until the product was well separated from the
saturated H SO mother liquor. The liquid phase was then
The present report concerns the use of concentrated H SO4
2
2
4
(
in this context 65–95 wt% H SO ) as a solvent for, and reactant
removed by decantation. The product was subsequently trans-
2
4
8
06
J. Chem. Soc., Dalton Trans., 2001, 806–815
DOI: 10.1039/b007675i
This journal is © The Royal Society of Chemistry 2001

View Article Online
Table 1 Crystal data and relevant parameters for the structure refinements of MoO (SO ) I and II. Corresponding unit-cell dimensions for
2
4
MoO (SO ) (modifications I, II, III) and MoO (SO )ؒH SO ؒH O from PXD are included for comparison. Calculated standard deviations are given
2
4
2
4
2
4
2
in parentheses. The 298 K PXD unit-cell dimensions are calibrated relative to a silicon standard whereas the 150 K SXD data are burdened with up to
0
.2% systematic uncertainties.
MoO (SO ) I
MoO (SO ) II
MoO (SO ) III
MoO (SO )ؒH SO ؒH O
2
4
2
4
2
4
2
4
2
4
2
M
224.00
Monoclinic
C2/c
224.00
Orthorhombic
Pna21
224.00
Monoclinic
268.13
Monoclinic
Crystal system
Space group
a/Å (SXD; 150(2) K)
b/Å (SXD; 150(2) K)
c/Å (SXD; 150(2) K)
β/Њ (SXD; 150(2) K)
26.3359(3)
12.2901(1)
8.5892(1)
104.480(1)
2691.76(5)
26.359(4)
12.267(2)
8.555(1)
104.65(2)
2677(1)
24
24.827(5)
8.592(2)
12.294(3)
3
V/Å (SXD; 150(2) K)
2622.5(9)
24.886(8)
8.575(4)
a/Å (PXD; 298(3) K)
b/Å (PXD; 298(3) K)
c/Å (PXD; 298(3) K)
β/Њ (PXD; 298(3) K)
12.249(2)
8.597(1)
8.288(1)
87.10(2)
871.6(4)
8
10.386(8)
8.894(6)
9.872(6)
113.6(1)
835(2)
12.268(4)
3
V/Å (PXD; 298(3) K)
2618(3)
24
3.411
Z
µ/mm
Ϫ1
3.323
Reflections collected
Independent reflections
27317
7699
31314
7894
[
R(int) = 0.0486]
[R(int) = 0.0496]
0.0528, 0.1139
0.0579, 0.1166
R1, wR2 indices [I > 2σ(I)]
R1, wR2 indices (all data)
0.1033, 0.2720
0.1213, 0.2800
ferred to a beaker containing 100 ml glacial acetic acid (Merck)
and stirred for 10 min. The product was filtered off and succes-
sively washed with 50 ml glacial acetic acid, 50 ml acetone
Single crystal X-ray diffraction analysis
The PXD data showed that the unit cell of MoO (SO ) I and
2
4
MoO (SO ) II is monoclinic and orthorhombic, respec-
(
Prolabo) and 50 ml diethyl ether (Den norske Eterfabrikk),
and then moved to a glove box with an argon atmosphere
where all specimens for characterization were prepared). The
2
4
tively (Table 1) and the pycnometric density of MoO (SO )
2
4
I gave a unit-cell content of 24 MoO (SO ) formula units.
(
2
4
The crystals were mounted on thin glass fibers on brass
pins in argon atmosphere and directly (within 5–8 s) placed
whole washing procedure was performed under a carbon
dioxide atmosphere in a box containing solid CO2.
in a cold N gas stream on a Siemens SMART CCD diffrac-
Single crystals of MoO (SO ) I were obtained from a 1 : 9
2
2
4
tometer.
[
Mo] : [H SO ] mixture heated at 240 ЊC for 18 h (without stir-
2 4
Diffraction data covering one hemisphere of the reciprocal
space were collected at 150(2) K by ω scans. Several crystals [of
ring for the last 10 h). MoO (SO ) II crystals were obtained from
2
4
a 1 : 20 [Mo] : [H SO ] mixture heated to bp (ca. 310 ЊC) in an
2
4
MoO (SO ) I in particular] were found unsuitable for structure
open round-bottomed flask. This (evaporation) treatment gave
crystals after 1 h. In both cases the liquid phase was removed by
decantation, and the crystals were washed as described above.
2
4
determination and even for the crystals of MoO (SO ) I used to
2
4
collect data indexing was not straightforward. Data reduction
24
25
(
SAINT ), absorption correction (SADABS ) and space
group establishment, structure determination and refinements
Powder X-ray diffraction (PXD)
26
(
SHELXTL ) were performed with standard programs. The
All samples were characterized by PXD at 25 ЊC with a Siemens
D5000 diffractometer with transmission geometry using mono-
positions of Mo and S were established by direct methods and
O by Fourier methods. Anisotropic displacement factors were
applied for Mo and S and isotropic factors for the O atoms,
since making the latter anisotropic led to non-physical values
for several of them. Table 1 lists crystal data and R factors
for the particular structure models for MoO (SO ) I and
chromatic Cu-Kα radiation (λ = 1.540598 Å) from an incident-
1
beam germanium monochromator. The detector was a Brown
PSD. The powder samples were sealed into 0.5 mm diameter
borosilicate capillaries. The diffraction patterns were collected
over the 2θ range 8–90Њ and autoindexed with help of the
2
4
MoO (SO ) II described below.
21
22
2
4
DICVOL and TREOR programs. Unit-cell dimensions
calibrated to silicon standard) were obtained by least-squares
The final R factors for MoO (SO ) I reflect a somewhat poor
2
4
(
crystal quality owing to the presence of several satellites. For
the 50 most disagreeing reflections in the refinement the
observed intensities were larger than the calculated ones. Forty-
four of these reflections had l = 4, three had l = 5 and three had
l = 9. When all reflections with l = 4 were removed the data set
for MoO (SO ) I refined to R1 = 0.079 and wR2 = 0.1934. The
23
refinements using the ITO program. The hygroscopic decom-
position of the products was studied at 25 ЊC with a Siemens
D5000 diffractometer with flat-plate geometry in reflection
mode. This study was performed with and without “protection”
against moisture of the atmosphere. The samples were mounted
on a silica-glass sample holder and Scotch tape was used as
2
4
differences in positional parameters were small and the crystal
chemical impact of the computational “manicure” was accord-
ingly considered unimportant.
“
protection”. Data were recorded in steps of 0.1 (0.3Њ) in the 2θ
range 19–22 (18–38Њ) at different time intervals over a period of
6 (200 min). Under these conditions an angular range of 3
20Њ) was scanned within 30 s (3 min).
1
(
The structure of MoO (SO ) II could not be determined in
2
4
space group Pnma, but was easily solved in space group Pna21.
The refined structure was afterwards inspected by the find-
Thermal analysis
27
symmetry algorithm in the Insight II program package with
negative results. Moreover, there is no mirror plane symmetry in
the structure, as would have been required by space group
Pnma.
Thermogravimetric (TG) and differential thermal analysis
(
DTA) were, respectively, performed with a Perkin-Elmer
TGA7 and DTA7 system in a nitrogen atmosphere. Silica-glass
containers were used as sample holders. The heating rate was
CCDC reference numbers 158631 and 158632.
See http://www.rsc.org/suppdata/dt/b0/b007675i/ for crystal-
lographic data in CIF or other electronic format.
Ϫ1
2
or 10 ЊC min in the temperature interval 40–550 ЊC and the
sample size ca. 40 mg.
J. Chem. Soc., Dalton Trans., 2001, 806–815
807

View Article Online
Fig. 3 Products resulting from the reaction between α-MoO3 or
MoO ؒ0.43H O and 95 wt% H SO at different temperatures. Fields for
3
2
2
4
MoO (SO ) I, II and III are indicated. The molybdenum source is
2
4
referred to as M and S denotes solution.
Fig. 1 Products resulting from the reaction between α-MoO3 and
PXD showed that different modifications of MoO (SO ) were
2
4
H SO4 as function of H SO concentration, reaction temperature
2
2
4
obtained in the reaction mixtures, depending on reaction time
and temperature. To gain insight in the progressing reaction,
experimental series were performed as a function of time. The
and time. MoO (SO ) I (open square; long reaction period), II (cross;
2
4
shorter reaction period), III (open circle; long reaction period) and
α-MoO3 (filled square). Inset: summary of phase fields for products
isolated after long reaction periods.
outcome for 95 wt% H
SO at 240 ЊC is summarized in Fig. 3.
2
4
First MoO (SO ) II starts to crystallize before all α-MoO is
2
4
3
dissolved, this is followed by a period where the reaction
product is a mixture of MoO (SO ) I and II, before finally a
2
4
single-phase product of I is obtained. Evidently I is the stable
modification at 240 ЊC under these conditions. As seen from
Fig. 3 the reaction progresses almost similarly at 100 ЊC, but the
rate is lower and MoO (SO ) III proves to be the stable reaction
2
4
product.
For 95 wt% H SO the liquid phase was intense yellow at
2
4
2
40 ЊC, but after cooling it turned a weaker yellow. For 90
wt% H SO and mixtures heated at bp the hot mother liquor
2
4
is weakly yellow but turned blue after cooling to r.t.
In the field labelled by α-MoO in the inset of Fig. 1 appar-
3
ently no reaction takes place, but some dissolution of α-MoO3
occurs as evidenced by a blue colouring of the liquid phase.
MoO ؒ0.43H O as reactant. The syntheses with (yellow)
3
2
Fig. 2 Products resulting from the reaction between MoO ؒ0.43H O
MoO ؒ0.43H O generally follow the same evolution patterns as
3
2
3 2
and H SO as function of H SO concentration, reaction temperature
2
4
2
4
those with α-MoO , but as seen from Fig. 2 some products
3
and time. MoO (SO ) I (open square; long reaction period), II (cross;
2
4
differ. The amount of water introduced from MoO ؒ0.43H O
3
2
shorter reaction period), III (open circle; long reaction period),
MoO (SO )ؒH SO ؒH O (H; open upward triangle), α-MoO (filled
changes the H SO4 concentration in the reaction mixture, but
2
2
4
2
4
2
3
this has not been taken into account in the construction of
Fig. 2.
square), solution (open downward triangle) and MoO ؒ0.43H O (filled
3
2
upward triangle). Inset: summary of phase fields for products isolated
after long reaction periods.
When MoO ؒ0.43H O dissolves in 95 wt% H SO a clear
3
2
2
4
yellow solution is first obtained, followed by gradual crystal-
lization of a white product. The time required to go through the
entire reaction sequence depends on the temperature. For
products synthesized above ca. 75 ЊC PXD again shows that
different modifications of MoO (SO ) are obtained under
Results and discussion
Syntheses
During synthesis and characterization several parameters may
influence the reaction product. Altogether 10 parameters which
may be important were identified; molybdenum source, H SO
2
4
different conditions. II is the first solid product obtained in
the entire temperature range. The final product above ca. 175 ЊC
is MoO (SO ) I. At 200 ЊC the average time for dissolution is
2
4
concentration, reaction temperature, reaction time, [Mo]:
H SO ] ratio, stirring rate, oxide particle size, washing pro-
2
4
[
2 min and for onset of crystallization 4 min.
2
4
cedure, characterization atmosphere and storing time before
characterization. Of these the first five were open for more
systematic variation than the others and were varied, primarily
one at the time. The outcome of the syntheses is summarized in
Figs. 1 and 2. In the following consideration of the results
a distinction is conveniently made between α-MoO3 and
MoO ؒ0.43H O as reactant. The solid products obtained were
In the temperature range ca. 75–175 ЊC MoO (SO ) III is
2
4
obtained as the final product. The findings for the reaction
series at 100 ЊC are summarized in Fig. 3. (When the reaction
temperature is below ca. 75 ЊC the final product is MoO (SO )ؒ
2
4
H SO ؒH O. As an average at 50 ЊC, the time for dissolution is 2
2
4
2
days and the time before onset of crystallization is 4 days.) For
the concentration range 65–90 wt% H SO and heating at bp
3
2
2
4
MoO (SO ) I, MoO (SO ) II, MoO (SO ) III, MoO (SO )ؒ
the original MoO ؒ0.43H O dehydrates to α-MoO . In these
2
4
2
4
2
4
2
4
3
2
3
H SO ؒH O and unchanged α-MoO or MoO ؒ0.43H O. (The
syntheses the entire reaction mixture had turned hard after
cooling, and it proved impossible to remove the product from
the round-bottomed flask in the usual way. Therefore, in order
to isolate these products the hot reaction mixture was poured
into a beaker, cooled to room temperature, whereafter 150 ml
glacial acetic acid were added. After stirring for ca. 1 day α-
2
4
2
3
3
2
labels I–III are used to distinguish new modifications syn-
thesized in this study.)
ꢀ
-MoO as reactant The progressing reaction between (pale
3
grey) α-MoO and 85–95 wt% (85 wt% only at 200 ЊC) H SO is
3
2
4
evidenced by gradual dissolution of α-MoO as the reaction
MoO had separated, and this product settled when the stirring
3
3
mixture turns yellow and the crystallization of a white product.
was discontinued.
8
08
J. Chem. Soc., Dalton Trans., 2001, 806–815

View Article Online
In the field labelled by sol in Fig. 2 only a yellow solution was
obtained (viz. no solid precipitate). On doubling the amount of
MoO ؒ0.43H O in the reaction mixture a more viscous solution
3
2
was obtained. The yellow solution was poured into a beaker
containing 150 ml glacial acetic acid, and after stirring for
ca. 30 days a white precipitate separated and settled when the
stirring was discontinued. After repeating the entire washing
procedure PXD showed that somewhat different products had
been obtained in the different syntheses, but attempts to
identify them have so far been unsuccessful.
Reaction scheme
The solid products of the reactions between α-MoO3 or
MoO ؒ0.43H O and concentrated H SO are summarized in
3
2
2
4
Figs. 1–3. These observations, together with knowledge on
the crystal structures of MoO (SO ) (I and II; see below) and
2
4
insight from Ref. 1 make us believe that the reaction between
α-MoO (MoO ؒ0.43H O) and 95 wt% H SO at temperatures
3
3
2
2
4
above ca. 75 ЊC tentatively follows a step-by-step pattern:
2ϩ
MoO (s) → MoO2 (solv) → MoO (SO )
II(s) →
3
2
4
(
MoO SO ) (solv) → MoO (SO ) I(s) for T > 175 ЊC or
2
4
n
2
4
MoO (SO ) III(s) for T < 175 ЊC where n is a small number.
2
4
2ϩ
Between MoO2 (solv) and MoO (SO ) II(s) there may occur a
2
4
step which involves species similar to (MoO SO ) (solv), but if
2
4 n
this is the case n must have a different value. The very fact that
MoO (SO ) II is metastable with respect to I and III shows that
2
4
the species which precedes formation of II must be different
from those which occur when it is redissolved.
The reaction sequences with α-MoO3 and MoO ؒ0.43H O
3
2
are indeed very similar and the differences which do occur are
not discussed here. It may be recalled that a period with only
solution is not observed in the α-MoO3 case, but the yellow
2ϩ
solution, which we take as an indicator of MoO2 (solv), is the
same in both cases.
Perhaps it may be worthwhile to emphasize that concentrated
H SO is a very different solvent from water with which it is
2
4
often compared and sometimes even imputed analogy with.
The one difference which in particular springs to mind in this
context is that the concept of solubility product has limited (or
perhaps no) significance in reactions with concentrated H SO .
2
4
Characterization by PXD and density measurements
All PXD reflections from phase-pure reaction products have
been indexed, and unit-cell data are listed in Table 1. The unit
cells of the three modifications of MoO (SO ) are related, but
Fig. 4 PXD patterns for (a) MoO
(SO ) I, II and III and (b)
2 4
MoO (SO )ؒH SO ؒH O. Miller indices are marked for “stronger”,
2
4
2
4
2
unambiguously indexed reflections.
2
4
their powder patterns (Fig. 4a; corresponding data for MoO2-
SO )ؒH SO ؒH O being depicted in Fig. 4b) show distinctly
solutions). To ensure that the washing procedure had not
changed the products the following simple test was used: a
small amount of each product together with the protecting
mother liquor was sealed into a 0.7 mm borosilicate capillary,
and examined by PXD. The diffraction data thus collected
proved to be virtually identical with those obtained for the
corresponding washed samples. Together with the TG data
these findings show that the washing procedure is satisfactory.
In particular it should be noted that glacial acetic acid is
efficient to remove adhered H SO on these and corresponding
(
4
2
4
2
different profiles, which in turn reflect the structural differences
see below) between the modifications. The PXD profiles for
different samples of MoO (SO ) (I, II and III) showed a certain
(
2
4
intensity variation. In accordance with the fact that the powder
samples were crushed prismatic crystals, introduction of
preferred orientation seems to play a significant role in the PXD
intensities.
The density of MoO (SO ) I was determined pycnometrically
2
4
2
4
at 25.00 ± 0.02 ЊC with the mother liquor (note: a saturated
products obtained with other oxide reactants.
solution) as displacement liquid. The density of the latter was
Ϫ3
established as 1.9271 ± 0.0009 g cm , and the density of I as
Heat treatment and chemical analysis
Ϫ3
3
0
.32 ± 0.03 g cm , which compares quite well with 3.35 ±
Ϫ3
.01 g cm (for unspecified modification) according to Ref. 8.
Results of TG and DTA experiments on reaction products are
presented in Fig. 5. Temperature regions and relative mass
losses for the decomposition reactions are summarized in Table
2. MoO (SO ) (modifications I–III) represents the simplest case,
Hence the unit-cell content is twentyfour (Z = 24.07) MoO -
m
2
(
SO ) formula units.
4
Together with the TG, DTA and chemical analysis data (see
2
4
below), the structural data in Table 1 serve as characterization
documentation for the compounds prepared in this study. The
crystal structure determinations of MoO (SO ) I and II further
verify the composition of these phases.
All these molybdenum compounds dissolve easily in
anhydrous acetone, acetic acid and diethyl ether (giving blue
where the decomposition takes place in one step at almost the
same temperature [eqn. (1), Table 2], and with observed, relative
weight losses in excellent agreement with the calculated values.
Two factors determine the final solid product from the TG or
DTA treatments: the sample size and the heating procedure.
For example with a sample size of 40 mg the final solid product
2
4
J. Chem. Soc., Dalton Trans., 2001, 806–815
809

View Article Online
Ϫ1
Table 2 Summary of TG results for phase-pure reaction products of molybdenum oxide sulfates; heating rate 2 ЊC min
a
∆m/m0
Illustration
no.
Compound
Decomposition reaction
Eqn.
Tstart/ЊC
Tend/ЊC
obs.
calc.
MoO (SO ) I
Fig. 5a
MoO (SO )(s) → MoO (s) ϩ SO (g)
(1)
(1)
(1)
(2)
375
375
375
160
445
445
445
230
0.3576
0.3570
0.3517
0.0509
0.3574
0.3574
0.3574
0.0530
2
4
2
4
3
3
MoO (SO ) II
2
4
MoO (SO ) III
2
4
MoO (SO )ؒH SO ؒH O
Fig. 5b
MoO (SO )ؒH SO ؒH O(s) →
2 4 2 4 2
2
4
2
4
2
MoO (SO )ؒH SO (s) ϩ H O(g)
2
4
2
4
2
MoO (SO )ؒH SO (s) → MoO (SO )(s) ϩ
(3)
230
415
160
415
450
450
0.3018
0.3528
0.5711
0.3043
0.3574
0.5765
2
4
2
4
2
4
H SO (g)
2
4
MoO (SO )(s) → MoO (s) ϩ
(4)
2
4
3
SO (g)
3
MoO (SO )ؒH SO ؒH O(s) → MoO (s) ϩ
(2 ϩ 3 ϩ 4)
2
4
2
4
2
3
2
SO (g) ϩ 2H O(g)
3 2
a
m refers to the mass at the start of the appropriate decomposition reaction.
0
Fig. 6 PXD records of (a) MoO (SO ) I and (b) II at different time
2
4
intervals during decomposition in moist air. Left insets: decomposition
under Scotch tape “protection”. Right insets: intensity of the reflections
concerned as a function of time, directly exposed to moist air (filled
square) and under Scotch tape “protection” (open square).
Fig. 5 (a) TG, DTG and DTA data for MoO (SO ) I. Virtually identi-
2
4
cal curves are obtained for II and III. The DTA scan is adjusted
to constant background signal. (b) TG and DTG data for MoO (SO )ؒ
2
4
H SO ؒH O. For numerical data see Table 2.
2
4
2
decomposition process (see above) the different result is not
surprising.
Ϫ1
is α-MoO at a heating rate of 2 ЊC min and a mixture of
3
Ϫ1
α-MoO and β-MoO at a heating rate of 10 ЊC min . Despite
numerous attempts with a variety of heat treatments, β-MoO3
has not been obtained phase pure from these sources. This is
attributed to the fact that β-MoO is metastable. Ref. 28 sug-
gests that β-MoO should have a fair degree of kinetic stability
near room temperature and that the spontaneous phase-
3
3
The TG curve for MoO (SO )ؒH SO ؒH O shows a more
2
4
2
4
2
complicated decomposition course (Fig. 5b and Table 2).
The sulfur contents of the present molybdenum oxide
sulfates have been verified by chemical analyses (S determined
as BaSO ): for MoO (SO ) I, II and III, obs. 14.7, 15.2, 15.2
28
3
3
4
2
4
wt%, respectively, calc. 14.3 wt%; for MoO (SO )ؒH SO ؒH O,
2
4
2
4
2
conversion temperature for β- → α-MoO is around 450 ЊC.
3
obs. 19.4 wt%, calc. 18.8 wt%).
Despite the close structural relationship between the three
modifications of MoO (SO ) it seems impossible to convert any
2
4
Hydrolysis of molybdenum oxide sulfates in moist air
one of them into one of the others by thermal means [see also
the DTA curve for MoO (SO ) I in Fig. 5a]. Hence the only
The stability of molybdenum oxide sulfates at r.t. depends on
the particle size of the sample and the surrounding atmosphere.
Molybdenum oxide sulfates turn blue, even in a glove box with
an argon atmosphere (fine powdered samples after a week
and sizable single crystals after a few months). Hence, trace
amounts of moist air are sufficient to start hydrolysis of the
2
4
path from, say, MoO (SO ) II to I or III appears to be the
2
4
synthetic route.
The thermal decomposition of MoO (SO ) (unspecified
2
4
modification) has previously been reported to take place
8
about 500 ЊC. On the basis of the factors which influence the
8
10
J. Chem. Soc., Dalton Trans., 2001, 806–815

View Article Online
Table 3 Bond lengths (Å) and bond valences for Mo and S in the crystal structure of MoO (SO ) I. Bond length averages and bond valence sums in
2
4
italics
Mo(1)–O(1)
Mo(1)–O(2)
Mo(1)–O(3)
Mo(1)–O(4)
Mo(1)–O(5)
Mo(1)–O(6)
1.648(11)
1.675(10)
2.028(9)
2.048(7)
2.222(9)
2.279(11)
2.01
1.87
0.72
0.68
0.43
0.37
S(1)–O(6)
1.440(11)
1.459(10)
1.489(9)
1.495(8)
1.64
1.56
1.44
1.42
S(1)–O(17)
S(1)–O(3)
S(1)–O(4)
1.47
6.06
1
.98
6.08
Mo(2)–O(10)
Mo(2)–O(8)
Mo(2)–O(9)
Mo(2)–O(7)
Mo(2)–O(11)
MO(2)–O(12)
1.677(10)
1.680(9)
2.026(8)
2.048(7)
2.212(8)
2.261(8)
1.86
1.85
0.73
0.68
0.44
0.38
S(2)–O(11)
S(2)–O(12)
S(2)–O(7)
S(2)–O(16)
1.445(8)
1.449(8)
1.492(7)
1.494(8)
1.62
1.61
1.43
1.42
1.47
6.08
1
.98
5.94
Mo(3)–O(13)
Mo(3)–O(14)
Mo(3)–O(15)
Mo(3)–O(16)
Mo(3)–O(17)
Mo(3)–O(18)
1.677(10)
1.693(10)
2.024(9)
2.034(7)
2.214(10)
2.292(10)
1.86
1.78
0.73
0.71
0.44
0.35
S(3)–O(18)
S(3)–O(5)
S(3)–O(15)
S(3)–O(9)
1.430(11)
1.455(10)
1.488(9)
1.498(8)
1.69
1.58
1.44
1.41
1.47
6.12
1
.99
5.87
molybdenum oxide sulfates. Under normal laboratory condi-
tions the advancing hydrolysis can be followed by the eye.
The hydrolysis reactions of MoO (SO ) I and II have been
2
4
studied in some detail by PXD. Quantitatively to map the
hydrolysis reactions, diffraction records were taken of samples
either directly exposed to moisture or “protected” by Scotch
tape during exposures. The results are shown in Fig. 6 where it
should be noted that the samples are not identical at the begin-
ning and end of the depicted recordings. The shape of the
11
hydrolysis curves in the insets are similar to those reported for
IO) SO and (IO) SeO . These shapes are typical for reactions,
(
2
4
2
4
whose rate is determined by the size of the surface area. After
complete hydrolysis blue solutions are obtained. The hydrolysis
7
reaction has previously been described according to eqn. (5). It
MoO (SO )(s) ϩ 2H O(g) → H MoO (s) ϩ H SO (1) (5)
2
4
2
2
4
2
4
was accordingly expected that H MoO could be isolated from
2
4
this solution. On adding glacial acetic acid a white precipitate
separated and settled during stirring. Different products were
obtained in different hydrolysis experiments and attempts to
identify them have so far been unsuccessful. One may have been
MoO (SO )ؒ2H O. However, this product was only obtained in
2
4
2
one synthesis. The hydrolysis time and [Mo] : [CH CO H] ratio
3
2
may influence the products after the glacial acetic acid treat-
ment. It is easy to imagine that H O can be incorporated into
2
the MoO (SO ) structure framework (see below) to form an
2
4
intermediate MoO (SO )ؒ2H O. By interaction between H O
2
4
2
2
and MoO (SO ) it is also easy to imagine the formation of
2
4
H SO and H MoO .
2
4
2
4
The crystal structure of MoO (SO ) I
2
4
Fig. 7 (a) The crystal structure of MoO (SO ) I projected along [001].
2
4
Unit-cell data for MoO (SO ) I are given in Table 1, selected
2
4
The three crystallographically different layers are labelled A, B and C
and primes mark mirror-imaged layers.(b) The MoO (SO ) II structure
bond distances are listed in Table 3 (angles are deposited). A
projection of the structure along [001] is shown in Fig. 7a. The
asymmetric unit contains three molybdenum, three sulfur and
eighteen oxygen atoms. Fig. 8 shows the typical coordination
environment for molybdenum and sulfur. All the three crystal-
2
4
projected along [100]. The two crystallographically different layers are
labelled D and E. (c) The MoO (SO ) I peeled-off network projected
2
4
along [001] with its eight- and four-membered rings. (d) The
MoO (SO ) II peeled-off network projected along [100] (note that this
2
4
network becomes broadened because the superimposed nets are not
fully parallel). MoO octahedra are shaded, SO tetrahedra are open
lographicaly different MoO octahedra deviate strongly from
6
6
4
ideal octahedral symmetry, and they contain two short, two
intermediate and two long Mo–O bonds (Table 3 and Fig. 7b).
All these octahedra possess four bridging oxygens (two long
and oxygen atoms are marked by open circles in (a) and (b). For clarity,
all oxygens are omitted in (c) and (d). Unit cells are indicated by thin
lines.
J. Chem. Soc., Dalton Trans., 2001, 806–815
811

View Article Online
Fig. 8 Representative coordination environment for Mo and S atoms in the MoO (SO ) I structure, showing (a) the eight- and four-membered rings
2
4
and the infinite ؒ ؒ ؒ –Mo–O–S–O–Mo– ؒ ؒ ؒ chain along [010]; O (open circles), Mo (small filled circles) and S (large filled circles), and (b) the
octahedral coordination around molybdenum, illustrated for Mo(2) where C, I and T denote chain connection, (layer and chain) interconnection and
terminal, respectively.
2
ment of 4.8 nets is also seen in the [100] projection (see com-
parison of single layers of nets in Fig. 10). The long a axis then
I
2
appears as a result of different stackings of the 4.8 nets.
A microporous-like nature of MoO (SO ) I is indicated by its
2
4
low framework density (17.8). The eight-membered rings are
stacked almost parallel along [001] which gives the structure its
open porosity. The “open” space within the rings (measured
between atomic centres) is approximately 8.5 × 5.5 and 4 ×
2
3
.5 Å for the eight- and four-membered rings, respectively.
The bond valences for Mo and S were calculated according
29
to Brese and O’Keeffe, using eqn. (6) where D is the bond
ij
V = Σ exp[(D Ϫ d )/b]
(6)
Fig. 9 Polyhedral representation of an A layer of the MoO (SO ) I
i
ij
ij
2
4
j
structure projected along [100]. MoO6 octahedra are depicted with
darker shading and the SO tetrahedra with lighter shading.
4
valence parameter for the particular type of bond concerned, dij
the experimentally determined interatomic distance (Table 3)
VI
VI
and b = 0.37; D is 1.907 and 1.624 for Mo –O and S –O,
ij
respectively. Individual bond valences and averages for the
crystallographically different Mo and S atoms for MoO (SO ) I
2
4
are included in Table 3, the overall averages for the structure
being 6.0(1) and 6.09(3) for Mo and S, respectively. Hence the
bond valences are in conformity with the formal oxidation
states according to the formula MoO (SO ).
2
4
The crystal structure of MoO (SO ) II
2
4
Unit-cell data for MoO (SO ) II are included in Table 1,
2
4
selected bond distances in Table 4 (angles are deposited), and a
projection of the structure along [100] is shown in Fig. 7b. As
Fig. 10 The arrangement of nets in a single layer of the MoO (SO ) I
2
4
seen from Fig. 7 and Tables 3 and 4, MoO (SO ) II carries all
structure projected along (a) [001] and (b) [100]. For clarity all oxygens
2
4
are omitted.
the characteristic features of MoO
framework density 18.3), the layers of corner-sharing, alternat-
ing MoO octahedra and SO tetrahedra, the eight- and four-
(SO ) I: the open framework
2 4
(
and two intermediate distances) and two terminal oxygens
6
4
(
two short distances). The shortest Mo–O distances signal
membered ring–channel system and the pattern of short,
intermediate and long Mo–O bonds. The asymmetric unit of II
contains twice as many atoms as the I modification.
double bonds (see below) which comply with the findings of
Eliseev et al. who interpreted the IR spectra of MoO (SO )
8
2
4
(
unspecified modification) in terms of sulfate groups and
A comparison of Figs. 7c and 7d shows that the stacked
2
double-bonded oxygen to molybdenum.
4.8 net layers along [001] in the MoO (SO ) I structure are
2
4
The three-dimensional framework is formed of corner-
sharing, alternating MoO octahedra and SO tetrahedra. The
retained along [100] in the II structure. Moreover, II also con-
tains the characteristic feature of 4.8 nets arranged in layers
2
6
4
thus formed eight- and four-membered rings (counting only the
Mo and S atoms of the ring-forming polyhedra; Figs. 8a and 9)
build infinite layers in the b c plane (designated A, B, C, CЈ, BЈ,
along two mutually perpendicular directions (along [010] as
well as [100]). As a result of the close relationships in the atomic
frameworks of modifications I and II, their unit-cell dimensions
(Table 1) are intimately related in various directions. The space
I
I
AЈ in Fig. 7a). These layers are in turn linked together by long
Mo–O bonds to form a three-dimensional framework, which
exhibits a staggered eight- and four-membered ring–channel
system running along [001] (Fig. 7a). The layers may alter-
natively be considered as built up of infinite ؒ ؒ ؒ –Mo–O–S–O–
Mo– ؒ ؒ ؒ chains along [010] (Fig. 8a) that are interconnected in
the b c plane by the long Mo–O bonds.
2
2
demand of around 4.2 × 12.3 Å for a 4.8 net implies six repeti-
tions along a and a , and two repetitions along c and b . The
I
II
I
II
superstructures along a and aII exhibit different stacking
I
2
sequences of the 4.8 nets. Fig. 11 shows an AЈ layer of
MoO (SO ) I superimposed on a D layer of MoO (SO ) II
2
4
2
4
(layer notations according to Fig. 7a and b). This emphasizes
that although the atomic arrangement within the eight- and
four-membered rings is virtually identical for descriptive
purposes there are significant, individual variations locally.
I
I
It is also instructive to consider the peeled-off network
arrangement of the MoO (SO ) I structure, and Fig. 7c shows
2
4
2
the characteristic 4.8 nets seen along [001]. A similar arrange-
8
12 J. Chem. Soc., Dalton Trans., 2001, 806–815

View Article Online
Table 4 Bond lengths (Å) and bond valences for Mo and S in the crystal structure of MoO (SO ) II. Bond length averages and bond valence sums in
2
4
italics
Mo(1)–O(1)
Mo(1)–O(2)
Mo(1)–O(3)
Mo(1)–O(4)
Mo(1)–O(5)
Mo(1)–O(6)
1.665(7)
1.689(6)
2.022(6)
2.055(7)
2.237(8)
2.286(6)
1.92
1.80
0.73
0.67
0.41
0.36
Mo(4)–O(19)
Mo(4)–O(20)
Mo(4)–O(22)
Mo(4)–O(21)
Mo(4)–O(23)
Mo(4)–O(24)
1.679(8)
1.675(8)
2.019(10)
2.026(10)
2.211(8)
2.240(9)
1.85
1.87
0.74
0.73
0.44
0.41
1
.99
5.89
1.98
6.04
Mo(2)–O(8)
Mo(2)–O(7)
Mo(2)–O(9)
Mo(2)–O(10)
Mo(2)–O(11)
Mo(2)–O(12)
1.673(10)
1.685(8)
2.047(9)
2.053(7)
2.248(8)
2.267(7)
1.88
1.82
0.69
0.67
0.40
0.38
Mo(5)–O(26)
Mo(5)–O(25)
Mo(5)–O(27)
Mo(5)–O(28)
Mo(5)–O(29)
Mo(5)–O(30)
1.675(8)
1.685(6)
2.029(9)
2.055(7)
2.233(6)
2.269(7)
1.87
1.82
0.72
0.67
0.41
0.38
2
.00
5.84
1.99
5.87
Mo(3)–O(13)
Mo(3)–O(14)
Mo(3)–O(15)
Mo(3)–O(16)
Mo(3)–O(17)
Mo(3)–O(18)
1.650(8)
1.686(8)
2.012(9)
2.050(6)
2.238(7)
2.247(7)
2.00
1.82
0.75
0.68
0.41
0.40
Mo(6)–O(31)
Mo(6)–O(32)
Mo(6)–O(33)
Mo(6)–O(34)
Mo(6)–O(35)
Mo(6)–O(36)
1.673(7)
1.669(7)
2.029(8)
2.031(8)
2.226(7)
2.299(7)
1.88
1.90
0.72
0.72
0.42
0.35
1
.98
6.06
1.99
5.99
S(1)–O(12)
S(1)–O(23)
S(1)–O(3)
S(1)–O(16)
1.462(7)
1.477(8)
1.489(7)
1.493(7)
1.55
1.49
1.44
1.43
S(4)–O(24)
S(4)–O(35)
S(4)–O(28)
S(4)–O(27)
1.445(9)
1.465(6)
1.486(7)
1.494(8)
1.62
1.54
1.45
1.42
1
.48
5.90
1.47
6.03
S(2)–O(5)
S(2)–O(17)
S(2)–O(21)
S(2)–O(10)
1.446(8)
1.470(7)
1.482(10)
1.487(7)
1.62
1.52
1.47
1.45
S(5)–O(11)
S(5)–O(15)
S(5)–O(36)
S(5)–O(4)
1.446(7)
1.477(9)
1.472(6)
1.495(7)
1.62
1.49
1.51
1.42
1
.47
6.06
1.47
6.04
S(3)–O(29)
S(3)–O(18)
S(3)–O(22)
S(3)–O(9)
1.438(6)
1.437(7)
1.460(10)
1.483(9)
1.65
1.66
1.56
1.46
S(6)–O(6)
S(6)–O(30)
S(6)–O(34)
S(6)–O(33)
1.439(6)
1.461(7)
1.463(8)
1.496(8)
1.65
1.55
1.55
1.41
1
.46
6.33
1.47
6.16
tetrahedral SO building blocks also carry characteristics of
4
30
molecular complexes. The trans π influence is, e.g., evident
when one considers the deviations of the MoO octahedra and
6
SO tetrahedra from O and T symmetry, respectively. The two
4
h
d
strong, uncoordinated (terminal) Mo–O bonds [average dis-
tance: 1.675(6) and 1.675(3) Å for I and II, respectively] are
appreciably shorter than the four (2 ϩ 2) weakest Mo–O bonds
[
bridging to S; average 2.035(4) and 2.036(4) Å for I and II,
respectively, for the intermediate-sized (cis-positioned to the
terminal oxygens) Mo–O bonds and 2.247(14) and 2.250(7) Å
for the longest (trans-positioned) Mo–O bonds; see Fig. 9]. In
accordance with this, the O–Mo–O bond angles between the
two strongest bonds exceed 90Њ [average: 102 and 103Њ for I
and II, respectively]. Conversely, the angles between the trans-
positioned, weakest Mo–O bonds fall below 90Њ [average: 77
and 76Њ for I and II, respectively]. These features are also
Fig. 11 Superposition of an AЈ layer of the MoO (SO ) I structure
2
4
on a D layer of the MoO (SO ) II structure; O (open circles), Mo
2
4
reflected in the SO tetrahedra although the effects on the S–O
(
small filled circles) and S (larger filled circles). The layers AЈ and D
4
distances are much smaller. The two shortest S–O bonds [aver-
ages 1.446(4) and 1.455(4) Å for I and II, respectively] are bridg-
ing to the weakest Mo–O bonds. Conversely, the two longest
S–O bonds [averages 1.493(2) and 1.483(3) Å for I and II,
respectively] are bridging to the intermediate Mo–O bonds.
correspond to those in Fig. 7.
Individual bond valences and averages for the crystal-
lographically different Mo and S atoms for MoO (SO ) II are
included in Table 4, the averages for the structure as a whole
being 5.96(9) and 6.1(1) for Mo and S, respectively.
2
4
2
The 4.8 nets of interconnected polyhedra (tetrahedra or
Despite the unquestionable three-dimensional character of
the MoO (SO ) I and II structures, the octahedral MoO and
tetrahedra and octahedra) with terminated oxygen atoms are
well known from microporous aluminium phosphates. AlPO4ؒ
2
4
6
J. Chem. Soc., Dalton Trans., 2001, 806–815
813

View Article Online
Table 5 Bond valences and angles around the oxygen atoms in the
crystal structure of MoO (SO ) I
Table 6 Bond valences and angles around the oxygen atoms in the
crystal structure of MoO (SO ) II
2
4
2
4
Bond valence
Angle/Њ
Mo–O bond
Bond valence
Angle/Њ
Mo–O bond
Mo(1)–O(1)
Mo(1)–O(2)
2.01
1.87
Short
Short
Mo(1)–O(1)
Mo(1)–O(2)
1.92
1.80
Short
Short
Mo(1)–O(3)–S(1)
Mo(1)–O(4)–S(1)
Mo(1)–O(5)–S(3)
Mo(1)–O(6)–S(1)
2.16
2.10
2.01
2.01
145.6(6)
135.4(5)
153.3(6)
141.2(7)
Intermediate
Intermediate
Long
Mo(1)–O(3)–S(1)
Mo(1)–O(4)–S(5)
Mo(1)–O(5)–S(2)
Mo(1)–O(6)–S(6)
2.17
2.09
2.03
2.00
142.5(4)
130.4(5)
149.4(5)
137.3(4)
Intermediate
Intermediate
Long
Long
Long
Mo(2)–O(10)
Mo(2)–O(8)
1.86
1.85
Short
Short
Mo(2)–O(7)
Mo(2)–O(8)
1.82
1.88
Short
Short
Mo(2)–O(9)–S(3)
Mo(2)–O(7)–S(2)
Mo(2)–O(11)–S(2)
Mo(2)–O(12)–S(4)
2.13
2.11
2.06
1.99
141.8(5)
136.2(5)
145.7(5)
148.2(5)
Intermediate
Intermediate
Long
Mo(2)–O(9)–S(3)
Mo(2)–O(10)–S(2)
Mo(2)–O(11)–S(5)
Mo(2)–O(12)–S(1)
2.15
2.12
2.01
1.93
147.1(5)
137.2(4)
140.3(5)
145.9(5)
Intermediate
Intermediate
Long
Long
Long
Mo(3)–O(13)
Mo(3)–O(14)
1.86
1.78
Short
Short
Mo(3)–O(13)
Mo(3)–O(14)
2.00
1.82
Short
Short
Mo(3)–O(15)–S(3)
Mo(3)–O(16)–S(2)
Mo(3)–O(17)–S(1)
Mo(3)–O(18)–S(3)
2.22
2.13
2.00
2.04
146.1(6)
140.0(5)
151.5(6)
151.4(7)
Intermediate
Intermediate
Long
Mo(3)–O(15)–S(5)
Mo(3)–O(16)–S(1)
Mo(3)–O(17)–S(2)
Mo(3)–O(18)–S(3)
2.24
2.10
1.93
2.06
160.6(5)
132.0(4)
139.3(5)
148.0(4)
Intermediate
Intermediate
Long
Long
Long
Mo(4)–O(19)
Mo(4)–O(20)
1.85
1.87
Short
Short
31
H O may provide a recent example from this institute in
2
which the structural elements are AlO tetrahedra, AlO (OH )
Mo(4)–O(21)–S(2)
Mo(4)–O(22)–S(3)
Mo(4)–O(23)–S(1)
Mo(4)–O(24)–S(4)
2.19
2.30
1.93
2.03
151.5(6)
147.3(6)
141.1(5)
165.1(6)
Intermediate
Intermediate
Long
4
4
2 2
octahedra and PO tetrahedra. The OH groups point into the
4
2
eight-membered rings. The space demand of the 4.8 net is here
around 4.1 × 11.4 Å . A similar situation occurs in AlPO ؒ
.5H O (AlPO -H3),
Long
2
4
32,33
1
were the half water molecule enters the
2
4
Mo(5)–O(25)
Mo(5)–O(26)
1.82
1.87
Short
Short
eight-membered ring channels (the one water molecule is
located inside six-membered rings) and the framework density
is 17.4. The mentioned values correspond remarkably well to
those found for the MoO (SO ) polymorphs when the differ-
Mo(5)–O(27)–S(4)
Mo(5)–O(28)–S(4)
Mo(5)–O(29)–S(3)
Mo(5)–O(30)–S(6)
2.14
2.12
2.07
1.93
147.0(5)
141.9(4)
160.6(5)
132.7(4)
Intermediate
Intermediate
Long
2
4
ent element combinations are taken into account. It may also
Long
be mentioned that MoO (SO ) (I and II) shows marked
2
4
structural similarities with the molybdenum() phosphates
Mo(6)–O(31)
Mo(6)–O(32)
1.88
1.90
Short
Short
34
KMo(H O)O (PO ) and NH Mo(H O)O (PO ). Again the
2
2
4
4
2
2
4
structural building blocks are MoO octahedra and PO tetra-
hedra, corner shared into layers of eight- and four-membered
rings, the charge-compensating cations being located between
6
4
Mo(6)–O(33)–S(6)
Mo(6)–O(34)–S(6)
Mo(6)–O(35)–S(4)
Mo(6)–O(36)–S(5)
2.14
2.26
1.96
1.85
131.4(5)
148.6(5)
145.3(4)
127.2(4)
Intermediate
Intermediate
Long
the layers. In these structures the MoO octahedra exhibit two
6
Long
short, three intermediate and one longer Mo–O bond. A brief
overview of structural properties of molybdenum oxide phos-
phates and derivatives is found in a recent review by Cheetham
fragments may be imagined to be retained in solution during
the reconstruction process (see above).
35
et al.
The insight into the structural arrangements of MoO (SO ) I
MoO (SO ) (I and II) has an extremely hygroscopic nature in
2
4
2
4
and II allows us to address three more or less related, interesting
questions concerning these substances. (1) How come that
modification II is metastable with respect to I, and yet II is
formed before I during the course of the synthesis (T > 175 ЊC )?
common with several other oxide sulfates. For the controlled
15
hydrolysis of (IO) SO it has been advocated that the reaction
2
4
is initiated by nucleophilic attacks by water oxygens at iodine
and water hydrogens at the nearby oxygens. This increases the
coordination of the originally unsaturated iodine and favours
formation of sulfuric acid. For reasons not fully understood,
(
2) Why has modification II to go through a complete
dissolution process in the mother liquor before modification I
can be formed? (3) Why is MoO (SO ) (I, II as well as III)
19
some oxide sulfates are non-hygroscopic, e.g. Te O (SO ).
2
4
2
3
4
hygroscopic?
However, the Te O (SO ) structure has a saturated TeO
2 3 4
6
The metastability of modification II relative to I is
unexpected in the sense that the latter has a 2.7% larger unit-cell
volume than the former (Table 1). The areas of the b c and b c
II II
coordination (counting short and long Te–O bonds) and no
terminal oxygen atoms on tellurium. Hence, one may speculate
whether the underbonded, terminal oxygens on Mo in
MoO (SO ) (I and II; Tables 5 and 6) are susceptible to electro-
I
I
planes are virtually identical and accordingly it must be the
2
4
distinctions in the atomic arrangements along the (long) a and
philic attacks by water molecules with the formation of O–H
(or O ؒ ؒ ؒ H) bonds. This could accordingly resemble the
situation in many, microporous materials where originally
underbonded, terminal oxygen atoms supplement their bond
valence by means of hydrogen bond interactions with, say,
ammine groups of organic templates or water molecules. In,
I
aII axes which are responsible for the disparity. However, we
have not so far been able to locate the real crystal-chemical core
of this distinction.
2
MoO (SO ) I and II have different stackings of the 4.8 nets.
2
4
Hence, a dissolution is required for their interconversion.
However, since the one- and two-dimensional building blocks
of the modifications are largely identical, sizeable structural
32,33
e.g., AlPO ؒ1.5H O (AlPO -H3)
the half water molecule per
4
2
4
formula unit enters the eight-membered rings at terminal
8
14
J. Chem. Soc., Dalton Trans., 2001, 806–815

View Article Online
oxygens on Al in a configuration resembling that imagined for
the hygroscopically attacking H O on MoO (SO ) (I and II).
13 S. Furuseth, K. Selte, H. Hope, A. Kjekshus and B. Klewe,
Acta Chem. Scand., Ser. A, 1974, 28, 71.
2
2
4
1
4 M. A. K. Ahmed, H. Fjellvåg and A. Kjekshus, Acta Chem. Scand.,
1
994, 48, 537.
Relations to the crystal structure of MoO (SO ) III
15 H. Fjellvåg and A. Kjekshus, Acta Chem. Scand., 1994, 48, 815.
6 M. A. K. Ahmed, H. Fjellvåg and A. Kjekshus, Acta Chem. Scand.,
1995, 49, 457.
17 M. A. K. Ahmed, H. Fjellvåg and A. Kjekshus, Acta Chem. Scand.,
1998, 52, 305.
2
4
1
For MoO (SO ) III no single crystals have been obtained so far.
2
4
The metrics of its unit cell are closely related to those of the
polymorphs I and II (cf. Table 1). Hence, almost certainly III
2
1
8 M. A. K. Ahmed, H. Fjellvåg and A. Kjekshus, Acta Chem. Scand.,
999, 53, 24.
also contains the characteristic 4.8 nets with infinite ؒ ؒ ؒ –Mo–
1
O–S–O–Mo– ؒ ؒ ؒ chains along the aIII axis. The dimensions of
bIII and cIII suggest two repetitions of the basic 4.8 net slab
along both directions. In line with this it is indeed possible to
position a cell with the appropriate dimensions of MoO (SO )
2
19 M. A. K. Ahmed, H. Fjellvåg and A. Kjekshus, J. Chem. Soc.,
Dalton Trans., 2000, 4542.
0 O. Söhnel and P. Novotný, Densities of Aqueous Solutions of
Inorganic Substances, Elsevier, New York, 1985.
2
2
4
III on a projected layer of I or II.
21 A. Boultif and D. Louer, J. Appl. Crystallogr., 1991, 24, 987.
2
2 P. E. Werner, L. Eriksson and M. Westdahl, Program TREOR,
Version 5, J. Appl. Crystallogr., 1985, 18, 367.
2
3 WIN-INDEX Professional Powder Indexing, Version 3.0, Siemens
AG Analytical X-ray System AUT, Karlsruhe, 1996.
References
1
M. A. K. Ahmed, H. Fjellvåg and A. Kjekshus, Acta Chem. Scand.,
24 SAINT Integration Software, Version 4.05, Bruker AXS, Inc.,
Madison, WI, 1995.
25 G. M. Sheldrick, SADABS, Empirical Absorption Correction
Program, University of Göttingen, 1997.
26 G. M. Sheldrick, SHELXTL, Version 5.0, Bruker AXS Inc.,
Madison, WI, 1994.
27 Insight II, Biosym Technologies, San Diego, CA, 1995.
28 E. M. McCarron, J. Chem. Soc., Chem. Commun., 1986, 336.
29 N. E. Brese and M. O’Keefe, Acta Crystallogr., Sect. B, 1991, 47,
192.
30 A. Pidcock, R.E. Richards and L. M. Venanzi, J. Chem. Soc. A,
1966, 1707.
31 S. T. Solevåg, Thesis, University of Oslo, 2000.
32 J. J. Pluth and J. V. Smith, Nature (London), 1985, 318, 165.
33 J. J. Pluth and J. V. Smith, Acta Crystallogr., Sect. C, 1986, 42,
1118.
34 R. Millini and A. Carati, J. Solid State Chem., 1995, 118, 153.
35 A. K. Cheetham, G. Férey and T. Loiseau, Angew. Chem., Int. Ed.,
1999, 38, 3268.
1
996, 50, 275.
2
3
4
C. Shultz-Sellack, Chem. Ber., 1871, 4, 14.
J. Meyer and V. Stateczny, Z. Anorg. Chem., 1922, 122, 19.
K. Ya. Shapiro and I. V. Volk-Karachevskaya, Russ. J. Inorg. Chem.,
1
969, 14, 571.
5
A. N. Zelikman and E. A. Odarenko, Izv. Vyssh. Uchebn. Zaved.
Tsvetn. Metall., 1988, No. 3, 113.
6
7
A. A. Woolf, Chem. Ind. (London), 1954, 1320.
A. A. Palant, V. A. Reznichenko and G. A. Menyailova,
Dokl. Chem., 1973, 213, 804.
S. S. Eliseev, E. E. Vozhdaeva and L. E. Malysheva, Dokl. Akad.
Nauk Tadzh. SSR, 1981, 24, 110.
Gmelins Handbuch der Anorganischen Chemie, System-Nummer 53:
Molybdän, Verlag Chemie, Berlin, 1935.
8
9
1
0 Gmelin Handbook of Inorganic and Organometallic Chemistry,
Mo Suppl. Vol. B8, Springer-Verlag, Berlin, 1995.
1 G. Dæhlie and A. Kjekshus, Acta Chem. Scand., 1964, 18, 144.
2 K. Selte and A. Kjekshus, Acta Chem. Scand., 1970, 24, 1912.
1
1
J. Chem. Soc., Dalton Trans., 2001, 806–815
815