Hydrothermal synthesis of two cationic bismuthate clusters: An alkylenedisulfonate bridged hexamer, [Bi6O4(OH) 4(H2O)2][(CH2)2(SO 3)2]3 and a rare nonamer templated by triflate, [Bi9O8(OH)6][CF3SO 3]5

Article abstract of DOI:10.1021/ic1004402

This paper reports the synthesis, characterization, and application of two cationic bismuthate clusters by anion templating. The compounds were synthesized under mild hydrothermal treatment and characterized by powder and single-crystal X-ray diffraction, infrared spectroscopy, and thermogravimetric analysis. The first material consists of a cationic hexanuclear bismuthate cluster octahedral in geometry and linked by 1,2-ethanedisulfonate molecules. This structure is thermally stable to about 235 °C. In the second compound, discrete cationic nonanuclear bismuthate clusters interact electrostatically with trifluoromethanesulfonate anions to pack into a nearly layered assembly. The material undergoes a transformation to Bi₂O₃ upon loss of the triflate groups at about 385 °C. Both materials demonstrate the use of sulfonate groups for the anion-directed assembly of these rare cationic inorganic structures. The application of the 3D octahedral bismuth cluster material toward acidic heterogeneous catalysis is also reported.

Full text of DOI:10.1021/ic1004402

Inorg. Chem. 2010, 49, 5619–5624 5619
DOI: 10.1021/ic1004402
Hydrothermal Synthesis of Two Cationic Bismuthate Clusters: An
Alkylenedisulfonate Bridged Hexamer, [Bi6O4(OH)4(H2O)2][(CH2)2(SO3)2]3
and a Rare Nonamer Templated by Triflate, [Bi9O8(OH)6][CF3SO3]5
David L. Rogow,† Honghan Fei,† Daniel P. Brennan,† Mariko Ikehata,† Peter Y. Zavalij,‡ Allen G. Oliver,† and
Scott R. J. Oliver*,†
†Department of Chemistry and Biochemistry, University of California, Santa Cruz, 1156 High Street,
Santa Cruz, California 95064, and ‡Department of Chemistry and Biochemistry, University of Maryland,
College Park, Maryland 20742
Received March 5, 2010
This paper reports the synthesis, characterization, and application of two cationic bismuthate clusters by anion
templating. The compounds were synthesized under mild hydrothermal treatment and characterized by powder and
single-crystal X-ray diffraction, infrared spectroscopy, and thermogravimetric analysis. The first material consists of a
cationic hexanuclear bismuthate cluster octahedral in geometry and linked by 1,2-ethanedisulfonate molecules. This
structure is thermally stable to about 235 °C. In the second compound, discrete cationic nonanuclear bismuthate
clusters interact electrostatically with trifluoromethanesulfonate anions to pack into a nearly layered assembly. The
material undergoes a transformation to Bi2O3 upon loss of the triflate groups at about 385 °C. Both materials
demonstrate the use of sulfonate groups for the anion-directed assembly of these rare cationic inorganic structures.
The application of the 3D octahedral bismuth cluster material toward acidic heterogeneous catalysis is also reported.
Introduction
The extensively studied hydrolysis behavior of Bi3þ makes
it a particularly desirable system for hydrothermal studies.6
The existence of polyoxo Bin clusters with n = 2 to 9 has been
supported by solution NMR evidence,7-9 but only octahe-
dral Bi6 oxocations and a single case of a polyoxo Bi9 cluster
have been characterized crystallographically.10 In the case of
the only nonanuclear report, the cluster is synthesized by base
hydrolysis of BiO(ClO4), with ClO4- anions directing cluster
formation and providing charge balance.10 By understand-
ing the conditions favorable for the formation of larger
clusters, it is possible that infinitely extended cationic bis-
muthate materials can be synthesized. While not nearly as
well explored as the cluster cations, the presence of higher
dimensionality cationic bismuthate features such as a one-
dimensional (1D) ribbon11 and a 2D layer12-14 support the
A major goal of our research has been the synthesis of
cationic three-dimensional (3D) inorganic materials for anion-
based applications such as anion exchange and heteroge-
neous catalysis.1 The two-dimensional (2D) cationic materials
Pb3F5 NO3, [Sb4O4(OH)2][C2H4(SO3)2] H2O, and Pb4.5F8
3
3
3
ClO4 reported previously by this group represent our first
major steps toward achieving this goal.2-4 Because of the
success with antimony and lead-based materials, a logical
extension for creating cationic inorganic materials was to
utilize other metal centers with the inert pair effect. The
naturally occurring francisite5 is a 2D mixed metal (Cu/Bi
and SeO3) cationic oxychalcogenide with the inert pair of the
selenium pointing to the chloride ions that are locked inside the
channels. We therefore extended our studies to bismuth for
the formation of higher dimensionality cationic frameworks.
(7) Pokric, B.; Pucar, Z. J. Inorg. Nucl. Chem. 1973, 35(9), 3287–3289.
(8) Olin, A. Acta Chem. Scand. 1959, 13(9), 1791–1808.
(9) Sundvall, B. Acta Chem. Scand. 1980, 34(2), 93–98.
(10) Thurston, J. H.; Swenson, D. C.; Messerle, L. Chem. Commun. 2005,
33, 4228–4230.
(11) Colmont, M.; Huve, M.; Ketatni, E.; Abraham, F.; Mentre, O. J. Sol.
State Chem. 2003, 176(1), 221–233.
(12) Tsuji, M.; Ikeda, Y.; Sazarashi, M.; Yamaguchi, M.; Matsunami, J.;
Tamaura, Y. Mater. Res. Bull. 2000, 35(13), 2109–2121.
(13) Henry, N.; Evain, M.; Deniard, P.; Jobic, S.; Abraham, F.; Mentre,
O. Z. Naturforsch. 2005, 60b(3), 322–327.
(14) Tsuji, M.; Yamaguchi, M.; Murao, S. Anion Intercalation and Anion
Exchange in Bismuth Compounds. In Solid-State Chemistry of Inorganic
Materials IV; Proceedings of the Symposium of the Materials Research Society,
Dec 2-6, 2002, Boston, MA; Materials Research Society: Warrendale, PA, 2003;
pp 197-202.
*To whom correspondence should be addressed. E-mail: soliver@
chemistry.ucsc.edu. Fax: (831) 459-2935. Phone: (831) 459-5448.
(1) Oliver, S. R. J. Chem. Soc. Rev. 2009, 38(7), 1868–1881.
(2) Tran, D. T.; Zavalij, P. Y.; Oliver, S. R. J. J. Am. Chem. Soc. 2002, 124
(15), 3966–3969.
(3) Swanson, C. H.; Shaikh, H. A.; Rogow, D. L.; Oliver, A. G.; Campana,
C. F.; Oliver, S. R. J. J. Am. Chem. Soc. 2008, 130(35), 11737–11741.
(4) Rogow, D. L.; Russell, M. P.; Wayman, L. M.; Swanson, C. H.;
Oliver, A. G.; Oliver, S. R. J. Cryst. Growth Des. 2010, 10(2), 823–829.
(5) Pring, A.; Gatehouse, B. M.; Birch, W. D. Am. Mineral. 1990, 75
(11-12), 1421–1425.
(6) Baes, C. F., Mesmer, R. E. The Hydrolysis of Cations; John Wiley &
Sons: New York, 1976.
r
2010 American Chemical Society
Published on Web 04/29/2010
pubs.acs.org/IC

5620 Inorganic Chemistry, Vol. 49, No. 12, 2010
Rogow et al.
Table 1. Crystallographic Data Collection and Refinement Statistics for SLUG-9
and SLUG-16
octahedral bismuth cluster material is stable and displays
activity toward acid catalyzed ketone protection. Cationic
organotin clusters have been used to increase the catalytic
activity of neutral organotin clusters toward Lewis acidic
alcohol acetylation reactions.19 The tin-oxo clusters were
discrete clusters and used as homogeneous catalysts, in con-
trast to the organosulfonate bismuth-oxo cluster described
here which behaves as a reusable, easily recoverable hetero-
geneous catalyst.
SLUG-9
SLUG-16
empirical formula
formula weight (g mol-1
C6H12Bi6O28S6
1978.40
C5Bi9F15O29S5
2850.17
)
3
crystal size (mm3)
0.340 ꢀ 0.230 ꢀ
0.310 ꢀ 0.145 ꢀ
0.170
monoclinic
P21
0.015
monoclinic
C2/c
crystal system
space group
crystal color, habit
colorless, block
12.561(1)
9.462(1)
13.727(2)
114.062(1)
1489.6(3)
2
4.411
35.847
1732
colorless, plate
17.463(6)
10.114(4)
23.924(9)
99.665(7)
4166(3)
4
4.545
38.279
4896
˚
a (A)
˚
b (A)
Experimental Section
˚
c (A)
Synthesis. [Bi6O4(OH)4(H2O)2][(CH2)2(SO3)2]3. [Bi6O4-
(OH)4(H2O)2][(CH2)2(SO3)2]3 (which we denote SLUG-9, for
University of California, Santa Cruz, Structure No. 9) was syn-
thesized hydrothermally. Bismuth(III) oxide (Acros, 99.9%),
1,2-ethanedisulfonic acid (TCI America, 95%), and deionized
water in a molar ratio of 1:4:400, respectively, were scaled to
15.3 mL and added to a 23 mL Teflon lined stainless steel
autoclave (Parr Instrument Company). The autoclave was
heated at 175 °C for 2 days, then allowed to cool to room
temperature before filtering and rinsing the crystals with
deionized water. The pH increased slightly from 1.54 before
the reaction to 1.71 after opening the autoclave. The average
yield was 1.33 g (90% based on Bi2O3).
β (deg)
volume (A )
3
˚
Z
Dc (Mg m3)
3
μ (mm-1
)
F(000) (e-)
θ range (deg)
index ranges
1.78 to 28.64
-16 e h e 6
-12 e k e 12
-18 e l e 18
16758
2.37 to 28.28
-23 e h e 23
-13 e k e 13
-31 e l e 31
23984
reflections collected
independent reflections
completeness to θ = 28.64 (%) 98.3
7425
5161
99.3
absorption correction type
min. and max. transmission
data/restraints/parameters
goodness of fit on F2
R1 [I >2σ(I)]
empirical
empirical
0.067 and 0.563
5161/105/325
1.028
0.0382
0.0596
0.51 and 1.00
7425/205/416
1.074
0.0286
0.0716
[Bi9O8(OH)6][CF3SO3]5. [Bi9O8(OH)6][CF3SO3]5 (denoted
SLUG-16) was synthesized as above. Bismuth(III) oxide, tri-
fluoromethanesulfonic acid (Acros, 99%), and deionized water
in a molar ratio of 1:2:400 were scaled to 15.3 mL. The yellow
suspension was allowed to stir for ∼10 min, after which time the
mixture appeared slightly more opaque, with a milky consis-
tency. The pH of the reaction mixture was 0.9. The pH was
adjusted to ∼3.1 by the dropwise addition of concentrated
aqueous ammonia (NH4OH, Fisher, 29.7%, ACS reagent
grade) before it was transferred to a 23 mL Parr autoclave.
The autoclave was sealed and heated at 175 °C for 2 days, after
which time it was removed from the oven and allowed to cool to
room temperature. When the autoclave was opened, a crop of
crystals was visible at the bottom of the Teflon liner. The pH of
the mother liquor was 2.8. The crystals were recovered by
vacuum filtration, rinsed with deionized water, and allowed to
air-dry overnight (average yield: 1.17 g, 89% based on Bi2O3).
Characterization Methods. Instrumental Details. Powder
X-ray diffraction (PXRD) data were recorded using a Rigaku
MiniFlex II Plus powder diffractometer with Cu-KR radiation
wR2 (all data)
largest diff. peak and
2.348 and -2.420 2.335 and -1.248
-
3
hole (e /A )
˚
potential of Bi3þ as the metal center for 3D cationic inorganic
materials.
The major drawback in applying our nitrate templating
strategy that was successful for Pb3F5(NO3) to Bi-based
systems concerns their tendency to form oxynitrates in the
presence of the nitrate ion. Other non-coordinating anions
such as BF4- and PF6- do not survive under hydrothermal
conditions. For these reasons, it was necessary to explore
other possible anion templating species for the growth of a
cationic material. The weakly coordinating sulfonate anions
provided a logical choice for a non-coordinating species that
would direct the structure without ligating the metal centers,
as observed for Sb.3 Sulfonate groups can, however, coordi-
nate directly to certain metals such as silver, barium, and
lead.15,16 The use of trifluoromethanesulfonate as a counter-
anion for inorganic complexes is well-known in the litera-
ture.17,18
We describe here the synthesis and solid state character-
ization of the first organosulfonate-bridged hexabismuth
cluster [Bi6O4(OH)4(H2O)2][(CH2)2(SO3)2]3 and a sulfonate-
templated nonabismuth polyoxocation cluster, [Bi9O8-
(OH)6][CF3SO3]5, only the second crystal structure of a Bi
cluster of this nuclearity reported (Table 1). In our prepara-
tion, hydrothermal treatment of Bi2O3 in the presence of
trifluoromethanesulfonic (triflic) acid produces the nonabis-
muth cluster arranged in layers with interlamellar triflate
groups. While this structure decomposes under reflux, the
˚
(λ = 1.5418 A). The irradiated area of the sample was a constant
4 mm2 by a variable slit mechanism. Thermogravimetric Ana-
lysis (TGA) was performed using a TGA 2050 (TA Instru-
ments). The samples were heated on a platinum pan from 25
to 1000 °C at a rate of 15 °C per minute under a nitrogen gas
purge. Fourier Transform Infrared Spectroscopy (FTIR) was
carried out with a Perkin-Elmer Spectrum One spectrophoto-
meter. Samples were prepared as KBr pellets. 1H NMR spectra
were measured on a Varian Unity Inova spectrometer equipped
with a 600 MHz HCN 5 mm cold probe system. Spectra were
taken in deuterated chloroform.
X-ray Crystal Structure Determination. Single-crystal data
for SLUG-9 were recorded at 150(2) K using a Bruker SMART
APEX II CCD area detector X-ray diffractometer using gra-
˚
phite monochromated Mo-KR radiation (λ = 0.71073 A) from
a fine-focus sealed tube operated at 50 kV and 30 mA. Single-
crystal data for SLUG-16 were measured at 299(2) K on a three-
circle diffractometer system equipped with a Bruker SMART
APEX CCD area detector using a graphite monochromator and
(15) Rogow, D. L.; Zapeda, G.; Swanson, C. H.; Fan, X.; Campana,
C. F.; Oliver, A. G.; Oliver, S. R. J. Chem. Mater. 2007, 19(19), 4658–4662.
(16) Cote, A. P.; Shimizu, G. K. H. Coord. Chem. Rev. 2003, 245(1-2),
49–64.
(17) Hubner, G. M.; Glaser, J.; Seel, C.; Vogtle, F. Angew. Chem., Int. Ed.
1999, 38(3), 383–386.
(18) Vilar, R. Angew. Chem., Int. Ed. 2003, 42(13), 1460–1477.
˚
a Mo-KR fine-focus sealed tube (λ= 0.71073 A) operated at
(19) Durand, S.; Sakamoto, K.; Fukuyama, T.; Orita, A.; Otera, J.;
Duthie, A.; Dakternieks, D.; Schulte, M.; Jurkschat, K. Organometallics
2000, 19(16), 3220–3223.

Article
Inorganic Chemistry, Vol. 49, No. 12, 2010 5621
50 kV and 30 mA. The structures were solved by direct methods
and expanded routinely. The models were refined by full-matrix
least-squares analysis of F2 against all reflections. All non-
hydrogen atoms were refined with anisotropic thermal dis-
placement parameters. Thermal parameters for the hydrogen
atoms were tied to the isotropic thermal parameter of the
atom to which they are bonded. Programs used were APEX-II
v2.1.4,20 SHELXTL v6.14,21 and Diamond v3.1e.22 Crystal-
lographic data collection and refinement parameters are listed in
Table 1.
Catalytic Studies. In a typical reaction, 0.05 g (25 μmol) of
SLUG-9 was placed in a 100 mL round-bottom flask along with
15.0 mL (168 mmol) of 2-butanone (methylethyl ketone, MEK,
Acros 99þ %, extra pure) and 9.35 mL (168 mmol) of ethylene
glycol (Acros, 99þ %). Other carbonyl substrates used were
acetone (Fisher, ACS reagent grade, 99.7%), benzaldehyde
(Aldrich,g99%), 2-pentanone (TCI America, min. 97%), and
acetaldehyde (Acros, 99%). This mixture was refluxed under
Dean-Stark conditions to remove water product. After a speci-
fied amount of reflux time, the reaction mixture was allowed to
cool to room temperature, and the catalyst was removed by
suction filtration. PXRD patterns of the solid were collected as
well as NMR spectra of the filtrate to determine the robustness
of the catalyst and percent yield of the reaction, respectively. 1H
NMR peaks used to calculate the percent yields were those of
ethylene glycol (δ 3.726, d or t, 4H) and the corresponding peaks
in the product: 2-ethyl-2-methyl-1,3-dioxolane (δ 3.88-3.97,
m, 4H), 2,2-dimethyl-1,3-dioxolane (δ 3.912, s, 4H), 2-methyl-
1,3-dioxolane (δ 3.83-3.92, m, 4H), 2-phenyl-1,3-dioxolane
(δ 4.00-4.20, m, 4H) and 2-methyl-2-propyl-1,3-dioxolane
(δ 3.89-3.98, m, 4H).
Figure 1. ORTEP diagrams of SLUG-9 (a) and SLUG-16 (b; triflate
anions omitted for clarity). Thermal ellipsoids are calculated at 50%
probability.
Four of the cluster capping oxygen atoms are likely
protonated. The distances of these oxygen atoms from
the plane of the three Bi atoms which they bridge are
˚
˚
as follows: O(1) = 1.101(7) A, O(3) = 1.106(7) A, O(4) =
Results and Discussion
˚
˚
1.108(7) A, and O(6) = 1.127(6) A. The rest of the cluster
oxygens are likely unprotonated as their distances from
the planes of the metals that they bridge are much smaller
SLUG-9. The synthesis of [Bi6O4(OH)4(H2O)2][(CH2)2-
(SO3)2]3 is straightforward and results in a pure phase of
the structure with high yield (90%). The Bi2O3 reagent
does not dissolve in the water before the autoclave is
heated, but the reaction conditions are sufficient to dis-
solve the reagent. The resulting colorless crystals are
large in size (average ca. 400 ꢀ 400 ꢀ 200 μm3) and block
shaped. The complex consists of an octahedral core of Bi
atoms. Each face of the octahedron is bridged by one
oxygen atom (μ3-oxo). These bridging oxygens have short-
er than usual bond distances to the Bi atoms [ranging from
˚
than the four hydroxyl groups: O(2)=0.319(6) A, O(5)=
˚
˚
˚
0.386(7) A, O(7) = 0.356(7) A, and O(8) = 0.465(6) A.
For charge balance, four cluster hydroxyls would give the
entire cluster a net charge of þ6. The three disulfonate
groups balance this charge and link the clusters with
Bi-O-S covalent bonds as well as two long electrostatic
contacts [Bi(3)-O(12) = 2.712(7) and Bi(1)-O(26)A =
˚
2.751(7) A] (Figure 1a). The bismuth cluster is coordi-
nated by two oxygen atoms on each end of the ethanedi-
sulfonate molecules. The third ethanedisulfonate coordi-
nates by only one oxygen atom, Bi(2) to O(11). The
organosulfonate groups cross-link to adjacent Bi octahe-
dra resulting in a 3D metal-organic network. The clus-
ters are discrete from each other and arranged in an
alternating manner, nearly resembling cubic close packed
(Figure 2a,b). The experimental powder pattern of
SLUG-9 matches closely to the theoretical pattern gen-
erated by the structure solution (Figure 3).
SLUG-16. The entire product after recovery was a crop
of colorless plates (average size ca. 450 ꢀ 400 ꢀ 20 μm3).
The product was synthesized in nearly 90% yield, a
significant improvement over the 70% yield reported
for the first nonabismuth cluster compound.10 After the
product had dried, a small amount of sample was used for
PXRD analysis. The resultant PXRD pattern (Figure 3)
could not be matched to any known phase. A suitable
single crystal was selected for SCXRD, and the structure
solution confirmed that the product is a new crystal
structure. The close agreement between the experimental
PXRD pattern and the pattern projected from single
˚
2.119(6) to 2.505(6) A, (Supporting Information, Table
S1); mean Bi-O distance = 2.424 ( 0.245 A: CSD].23 Two
˚
of the bismuth atoms have a terminal oxygen [O(9) and
O(10)], which are most likely water molecules. Hydrogen
atoms could not be observed, however, in the Fourier
difference electron density maps because of the high elec-
tron density of the six bismuth atoms. The bond distances
˚
are 2.618(7) and 2.671(7) A for Bi(2)-O(9) and Bi(5)-
O(10), respectively (Supporting Information, Table S1).
These distances are longer than the average Bi-O bond
˚
(∼ 2.5 A: CSD), indicating that they may be coordinated
water molecules. Further evidence for water molecules is
,
corroborated by the FTIR spectrum (3300 to 3500 cm-1
Supporting Information, Figure S1). Bound hydroxyl is
also evident in the infrared spectrum.
(20) APEX-II, 2.1.4; Bruker-AXS: Madison, WI, 2007.
(21) SHELXTL, Crystal Structure Determination Package; Bruker Analy-
tical X-ray Systems Inc.: Madison, WI, 1995-99.
(22) Brandenburg, K.; Putz, H. Diamond, 3.1; Crystal Iimpact: Bonn,
Germany, 2007.
(23) Allen, F. H. Acta Crystallogr., Sect. B 2002, No. 58, 380–388.

5622 Inorganic Chemistry, Vol. 49, No. 12, 2010
Rogow et al.
The rest of the Bi atoms are each coordinated to 5 O
atoms within the cluster to define a D3h trigonal bipyri-
midal geometry around each metal center. The overall
formula of the cationic cluster is [Bi9O8(OH)6]5þ. The
clusters are arranged into sheets with long electrostatic
contacts between three cluster bismuth atoms on the cor-
ners of the cluster and all three oxygen atoms of the tri-
fluoromethanesulfonate anions. The through space dis-
˚
tances range from 2.987(7) A between Bi(1) and O(13), to
˚
4.317(7) A between Bi(1) and O(11). These distances are
beyond the accepted range of covalent Bi-O bonds.
There are also triflate anions between the clusters that
are not close enough to interact with the clusters and
remain in the periphery (Figure 2c,d).
All O atoms are triply bridging (μ3-oxo), with each Bi
center linking to the eight other Bi atoms through
Bi-O-Bi bonding. The H atoms could not be located
from SCXRD data, but their presence, necessary for the
charge balance of the compound, is confirmed by the
occurrence of two sharp peaks in the IR spectrum (3600
and 3669 cm-1, Supporting Information, Figure S1),
indicating the presence of bound hydroxyl groups.
Furthermore, the nonabismuth cluster characterized pre-
viously by Thurston et al. contained 6 μ3 cluster hydro-
xyls. These were assigned by necessity of charge balance
as well as the distance of the oxygen atoms above the
plane of the three Bi atoms to which they are bonded.10
The average above-plane distance of the μ3 cluster
hydroxyls [O(3), O(5), O(6), O(3)A, O(5)A and O(6)A], is
Figure 2. Crystallographic images of SLUG-9 in the a-c (a) and b-c
(b) planes, and SLUG-16 in the a-c (c) and b-c (d) planes (bismuth, violet;
oxygen, red; sulfur, yellow; fluorine, green; carbon, dark gray; and hydro-
gen, light gray).
˚
1.207(7) A. The unprotonated μ3-oxos [O(2), O(4), O(7),
˚
O(2)A, O(4)A, and O(7)A] lie an average of 0.368(6) A
above the plane of the three Bi atoms to which they
bridge. The other two unprotonated μ3 oxygens [O(1)
˚
and O(1)A] actually lie 0.234(7) A concave with respect to
the cluster.
FTIR spectra for both structures are shown in Sup-
porting Information, Figure S1. Strong peaks for sulfo-
nate stretching between about 1100 and 1300 cm-1 are
observed for SLUG-9. Peaks corresponding to the tri-
fluoromethane (CF3) group in SLUG-16 are observed
between 1100 and 1350 cm-1, in the stretching region
known for perfluorinated sulfonate.24 The sharp absor-
bance peaks in both spectra between 1000 and 1050 cm-1
are also assigned to the sulfonate groups. A broad
adsorbed water peak is observed centered at 3436 cm-1
for SLUG-16. Sharp bound hydroxyl stretching peaks
appear at 3600 and 3669 cm-1 for SLUG-16, and at 3445
and 3499 cm-1 for SLUG-9. A somewhat broad and weak
bound water stretch appears at 3233 cm-1 in the SLUG-9
spectrum. Sharp peaks are observed as a doublet in
SLUG-9 at 1603 and 1659 cm-1 and are due to lattice
H2O scissoring.25,26 The presence of hydroxyl and water
in SLUG-9 confirms the structural assignment based on
charge balance and oxygen distances above the plane
of the three Bi atoms they bridge. The sharp peak at
Figure 3. PXRD patterns of SLUG-9 (bottom, blue) and SLUG-16
(top, red). Theoretical patterns are shown below as bars (asterisk: peak
from the aluminum sample holder).
crystal data indicates that the as-synthesized product is a
pure phase (Figure 3). The SLUG-16 structure is not
stable under reflux conditions and was therefore not
pursued as a possible heterogeneous catalyst.
The structure is made up of discrete nonabismuth
cluster cations charge-balanced by trifluoromethanesul-
fonate anions. Figure 1b shows the Oak Ridge Thermal
Ellipsoid Plot (ORTEP) diagram and atom labeling
scheme for the nonabismuth structure (SLUG-16). The
cluster is composed of nine Bi atoms. Bi(4), Bi(4)A, Bi(5),
and Bi(5)A are four coordinate in a slightly distorted
square pyramidal geometry where the Bi atom is at the
vertice (Figure 1b, Supporting Information, Table S2).
(24) Stoyanov, E. S.; Kim, K.-C.; Reed, C. A. J. Phys. Chem. A 2004, 108
(42), 9310–9315.
(25) Nakamoto, K. Infrared and Raman Spectra of Inorganic and
Coordination Compounds, 5th ed.; John Wiley & Sons, Inc.: New York, 1997;
pp 227-231.
(26) Mayo, D. W.; Miller, F. A.; Hannah, R. W. Course Notes on the
Interpretation of Infrared and Raman Spectra; John Wiley & Sons, Inc.:
Hoboken, NJ, 2004; pp 300-302.

Article
Inorganic Chemistry, Vol. 49, No. 12, 2010 5623
1630 cm-1 in SLUG-16 is most likely due to physisorbed
water on the surface of the sample, as observed in the
TGA trace below 50 °C (Supporting Information, Figure
S2). This peak appears as a similarly shaped singlet in the
spectrum of liquid H2O.26 The CF3 stretching peak is very
strong and overshadows the alkylene peaks of the sul-
fonate. In SLUG-9, sharp peaks at 741, 761 are from the
-CH2- group, and the absorbance at 1416 cm-1 is
assigned to -CH2-S.
Scheme 1. Ketal Formation Reaction between Carbonyl Substrate
and Ethylene Glycol
Thermal Characterization. Thermogravimetric analysis
was performed to observe the thermal behavior of the
clusters, including possible linkage into higher dimen-
sionality. The TGA trace of SLUG-9 indicates that the
structure is thermally stable until about 235 °C, where a
small loss of weight is followed by a plateau until about
350 °C (blue, Supporting Information, Figure S2). After
this first weight loss step, about 96.5 wt % of the structure
remains intact. This corresponds closely to 2 hydroxyl
groups and two water molecules (3.5 wt % theoretical).
Ex-situ heating to 250 °C and subsequent PXRD revealed
that the structure remains intact but undergoes a slight
contraction in the spaces between the clusters. The water
molecules are lost, and hydroxides are most likely being
evolved by intermolecular condensation. The second,
major weight loss step with onset at about 350 °C leaves
68 wt % of the original structure by 450 °C. This likely
corresponds to loss of the three ethanedisulfonate mole-
cules in the structure (28.4 wt % theoretical). A slight
weight gain is observed in the temperature region of 400
to 1000 °C, possibly a result of nitrogen uptake by the
unsaturated bismuth atoms. Ex-situ heating to 800 °C
under a nitrogen atmosphere gave bismuth metal (ICDD
PDF # 44-1246). The final product obtained after ex-situ
heating to 1000 °C under ambient atmosphere was also
bismuth metal (ICDD PDF # 44-1246).
The TGA trace of SLUG-16 (red, Supporting Informa-
tion, Figure S2) displays a small weight loss of ∼0.5 wt %
below 50 °C because of removal of adsorbed water on the
surface of the solid. The sample plateaus in the region
from 50 to 350 °C. There is a small weight loss of 1.8%
between 325 and 375 °C likely because of the loss of 3
hydroxyls to intermolecular condensation (1.8 wt %
theoretical). The major weight loss step occurs over the
range of about 385 to 425 °C and then slows considerably
until the end of the measurement range. The rate of
weight loss decreases and almost plateaus at about
600 °C where about 77% of the original weight remains.
The loss is likely due to the removal of 79% of the
trifluoromethanesulfonate anions (theoretical 26.1% for
all triflates). The product obtained after heating to 600 °C
is Bi2O3 (ICDD PDF # 41-1449). After ex-situ heating to
1000 °C in ambient atmosphere, the final product is
bismuth metal (ICDD PDF # 44-1246).
Table 2. Percent Yields for the Different Ketal Formation Substrates
reaction time
(hours)
catalyst
carbonyl substrate
yield (%)
SLUG-9
SLUG-9
SLUG-9
SLUG-9
SLUG-9
Bi2O3
Bi2O3
Bi2O3
Bi2O3
acetone
6
6
3
3
4
6
6
3
3
30
54
86
96
2-butanone
2-pentanone
benzaldehyde
acetaldehyde
acetone
2-butanone
2-pentanone
benzaldehyde
10
<0.1
<0.1
<0.1
81
protecting carbonyl groups, especially the cyclic ketal
formation reaction demonstrated here.27,28 In our case,
no harmful solvents are needed, only the reactants them-
selves, and thus the reaction can be considered “green”.
Also, the catalyst can be used as prepared with no activa-
tion or pretreatment required. The reaction using benzal-
dehyde as the carbonyl substrate goes to 96% completion
in 3 h, as confirmed by 1H NMR (Table 2 and Supporting
Information, Figure S3).
A typical homogeneous iodine catalyst reported by
Banik et al. shows ketal formation reactions going to
90% in 16 h.28 In addition, the catalyst reported here is a
heterogeneous catalyst. Solid metal phosphonates of Sn,
Ti, and Zr show reaction yields as high as 90% in 3 h.27
Although SLUG-9 remains intact after the reaction, the
crystallinity is reduced slightly after the catalytic reac-
tion as observed by decreased intensity of the majority
peaks in the PXRD pattern (Supporting Information,
Figure S8). Control reactions were conducted using only
Bi2O3 added to the reactants. Very little products were
formed (<0.1% with MEK substrate, Table 2) using the
starting material Bi2O3. A reaction was also run for 3 h,
after which the solid catalyst was filtered off and then the
reaction was run for 3 more hours. No further product
formation occurred after the catalyst was removed, prov-
ing that (i) turnover requires the presence of the solid
catalyst and (ii) no species were being leached from the
solid that could be active homogenously. Furthermore,
1
no evidence of ethanedisulfonate was found in the H
NMR spectra of any of the products. The catalyst is
reusable for at least two more cycles with little diminished
yield (average ∼5%).
It is possible that the hydroxyl groups are responsible
for the catalytic acivity. Bi(III) is Lewis acidic; however,
it is bonded to the sulfonate oxygen atoms and may be
coordinatively inaccessible. Organosulfonate ligands,
however, are known to be “flexible” in that they may
adjust their coordination mode to accommodate struc-
tural changes.29 In either case, the reaction is occurring at
Catalytic Reaction of SLUG-9. The SLUG-9 structure
showed significant activity as an acidic catalyst toward
formation of a cyclic ketal protecting group in the reac-
tion between ketones or aldehydes and ethylene glycol
(Scheme 1). This reaction is useful in organic synthesis for
(27) Patel, S. M.; Chudasama, U. V.; Ganeshpure, P. A. J. Mol. Catal. A:
Chem. 2003, 194(1-2), 267–271.
(28) Banik, B. K.; Chapa, M.; Marquez, J.; Cardona, M. Tetrahedron
Lett. 2005, 46(13), 2341–2343.
(29) Melcer, N. J.; Enright, G. D.; Ripmeester, J. A.; Shimizu, G. K. H.
Inorg. Chem. 2001, 40(18), 4641–4648.

5624 Inorganic Chemistry, Vol. 49, No. 12, 2010
Rogow et al.
the surface of the material, inferred from the observation
that the bulkier substrates such as 2-pentanone and
benzaldehyde showed the greatest percent conversion.
Indeed, there is no real void space inside the struc-
ture (no solvent accessible void space, calculated by
PLATON30) so reactant diffusion into the material can
be ruled out. The substrates that have the most electron
delocalization ability gave the highest percent yields,
possibly because of the stabilization of the oxocation
intermediate. Ketal formation in general requires protic
acids; therefore, we would conclude that the cluster
hydroxyl groups are responsible for the catalytic activity.
does not survive under reflux conditions. The hexabismuth
octahedral cluster bridged by alkylenedisulfonate shows
significant activity asan acidic heterogeneous catalyst toward
cyclic ketal formation without the need for activation.
We believe that higher dimensionality cationic bismuthate
materials are possible using suitably shaped anionic sulfonate
templates under favorable conditions of reaction time, tem-
perature, and pH.
Acknowledgment. This research was supported by an
NSF Career Award (DMR-0506279). The single crystal
X-ray diffraction data in this work were recorded on an
instrument supported by the NSF Major Research Instru-
mentation (MRI) Program under Grant CHE-0521569.
Conclusions
In summary, two cationic bismuthate clusters have been
synthesized by anion templating. Infrared spectroscopy and
thermogravimetric analysis confirms the presence of hydro-
xyl groups on both clusters and coordinated water on the
octahedral cluster. The trifluoromethansulfonate-templated
nonabismuth structure displays greater thermal stability, but
Supporting Information Available: Crystallographic informa-
tion files (.cif) containing fractional atomic coordinates, bond
lengths and angles and refinement parameters, Infrared spectra,
TGA traces, proton NMR spectra of the catalytic reaction
products and PXRD of the catalyst after reaction and tables
of bond lengths and angles. This material is available free of
charge via the Internet at http://pubs.acs.org.
(30) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7–13.