Reactivity and properties of [-O-BiIII...O=Mo-]n chains

Article abstract of DOI:10.1021/ic061198c

A coordination polymer [Cp*(O)₂Mo-O-Bi(o-tolyl) ₂]_n, II, containing Mo-O-Bi and Mo=O...Bi moieties was investigated with respect to its behavior in contact with OH^- and Cp₂MoH₂ and as potential single source precursor in the polyol method. It turned out that hydroxide as a base breaks up the polymer to yield Cp*MoO₃^- and (o-tolyl)₂BiOH. The latter polymerizes to give the coordination polymer [(o-tolyl) ₂BiOH]_n, 1. Alternatively, 1 can be prepared by reacting [(o-tolyl)₂Bi(hmpa)₂]SO₃CF₃ with NBu₄OH/H₂0 in thf/water. If, however, NBu ₄OH/MeOH is used in dichloromethane as the solvent, the (o-tolyl)₂BiOH formed intermediately undergoes methanolysis, and finally, [(o-tolyl)₂BiOMe]_n, 3, is isolated. Although 1 and 3 are very similar compounds, their crystal structures differ significantly: while the structure of 1 is dominated by secondary bonding leading to seesaw-type coordination geometries around the Bi centers, the Bi atoms in 3 are coordinated in a distorted tetrahedral fashion, and secondary bonding plays only a minor role. If 1 is dissolved in a nonpolar, nonprotic solvent, condensation reactions occur immediately leading to [(otolyl) ₂BiOBi(o-tolyl)₂], 2, which can be obtained on a preparative scale this way. Compound 3 which can be prepared in good yields may prove to be a useful starting material in bismuth chemistry. Here, it was shown to react with molybdocene dihydrides to provide stable Bi-substituted molybdocene monohydrides [^RCp₂-Mo(H)(Bi(o-tolyl) ₂)] (R = Me 4, R = H 5); compounds of that type were identified in solution before but had so far eluded isolation. Compound 4, whose crystal structure is discussed, also forms when II is treated with methylated molybdocene dihydride. This obviously leads to the formation of Mo-Bi bonds (→ 4), as well as Mo-OH units, which undergo condensation reactions leading to Mo-O-Mo moieties (i.e., [Cp*₂Mo₂O₅] is formed as a byproduct). The use of II as precursor in the polyol method successfully led to bismuthmolybdate nanoparticles (accompanied by crystallites); however, no single phase is obtained, but biphasic materials consisting of Bi₂Mo₂O₉ and Bi ₂MoO₆, whose ratio can be determined by the choice of the hydrolyzing reagent, are formed instead. One of these materials proved to be capable of sensing EtOH selectively at elevated temperatures.

Full text of DOI:10.1021/ic061198c

Inorg. Chem. 2006, 45, 9020−9031
Reactivity and Properties of [−O− OdMo−]n Chains
BiIII...
Stefan Roggan,† Christian Limberg,*,† Burkhard Ziemer,† Maike Siemons,‡ and Ulrich Simon*,‡
Humboldt-UniVersita¨t zu Berlin, Institut fu¨r Chemie, Brook-Taylor-Strasse 2,
12489 Berlin, Germany, and RWTH Aachen, Institut fu¨r Anorganische Chemie, Landoltweg 1,
52056 Aachen, Germany
Received June 30, 2006
A coordination polymer [Cp*(O)2Mo−O−Bi(o-tolyl)2]n, II, containing Mo−O−Bi and ModO‚‚‚Bi moieties was
investigated with respect to its behavior in contact with OH- and Cp2MoH2 and as potential single source precursor
-
in the polyol method. It turned out that hydroxide as a base breaks up the polymer to yield Cp*MoO3 and (o-
tolyl)2BiOH. The latter polymerizes to give the coordination polymer [(o-tolyl)2BiOH]n, 1. Alternatively, 1 can be
prepared by reacting [(o-tolyl)2Bi(hmpa)2]SO3CF3 with NBu4OH/H2O in thf/water. If, however, NBu4OH/MeOH is
used in dichloromethane as the solvent, the (o-tolyl)2BiOH formed intermediately undergoes methanolysis, and
finally, [(o-tolyl)2BiOMe]n, 3, is isolated. Although 1 and 3 are very similar compounds, their crystal structures differ
significantly: while the structure of 1 is dominated by secondary bonding leading to seesaw-type coordination
geometries around the Bi centers, the Bi atoms in 3 are coordinated in a distorted tetrahedral fashion, and secondary
bonding plays only a minor role. If 1 is dissolved in a nonpolar, nonprotic solvent, condensation reactions occur
immediately leading to [(o-tolyl)2BiOBi(o-tolyl)2], 2, which can be obtained on a preparative scale this way. Compound
3 which can be prepared in good yields may prove to be a useful starting material in bismuth chemistry. Here, it
was shown to react with molybdocene dihydrides to provide stable Bi-substituted molybdocene monohydrides [RCp2-
Mo(H)(Bi(o-tolyl)2)] (R
eluded isolation. Compound 4, whose crystal structure is discussed, also forms when II is treated with methylated
molybdocene dihydride. This obviously leads to the formation of Mo Bi bonds ( 4), as well as Mo OH units,
which undergo condensation reactions leading to Mo Mo moieties (i.e., [Cp*2Mo2O5] is formed as a byproduct).
) Me 4, R ) H 5); compounds of that type were identified in solution before but had so far
−
f
−
−O−
The use of II as precursor in the polyol method successfully led to bismuthmolybdate nanoparticles (accompanied
by crystallites); however, no single phase is obtained, but biphasic materials consisting of Bi2Mo2O9 and Bi2MoO6,
whose ratio can be determined by the choice of the hydrolyzing reagent, are formed instead. One of these materials
proved to be capable of sensing EtOH selectively at elevated temperatures.
Introduction
ular Mo-O-Bi complexes existed in the literature until
recently, when we achieved the synthesis and isolation of
such complexes containing Bi atoms in oxidation states +III
and +V (see Chart 1).2 Here, we describe the results of an
investigation concerning the reactivity and properties of one
of these compounds, the coordination polymer II.
nMoO3/Bi2O3 phases represent heterogeneous catalysts for
the allylic oxidation of propene. The reasons why exactly
this combination of metals is most efficient are still discussed
controversially.1 The favorable properties of bismuthmolyb-
dates might be related inter alia to the presence of Mo-O-
Bi linkages on the surface of these solids, assuming that such
assemblies possess favorable H abstraction or radical trapping
properties.1g,h This hypothesis naturally stimulates attempts
aimed at establishing Mo-O-Bi units in molecular com-
pounds also. However, no structurally characterized molec-
(1) (a) Keulks, G. W.; Krenzke, L. D.; Notermann, T. M. AdV. Catal.
1978, 27, 183. (b) Grasselli, R. K.; Burrington, J. D. AdV. Catal. 1981,
30, 133. (c) Grasselli, R. K.; Burrington, J. D. Ind. Eng. Chem. Prod.
Res. DeV. 1984, 23, 393. (d) Belagem, J.; Osborn, J. A.; Kress, J. J.
Mol. Catal. 1994, 86, 267. (e) Grasselli, R. K. In Handbook of
Heterogeneous Catalysis; Ertl, G., Kno¨zinger, H., Weitkamp, J., Eds.;
VCH Verlagsgesellschaft mbH: Weinheim, Germany, 1997; p 2302.
(f) Jang, Y. H.; Goddard, W. A., III. Top. Catal. 2001, 15, 273. (g)
Limberg, C. Angew. Chem. 2003, 115, 6112; Angew. Chem., Int. Ed.
2003, 42, 5932. (h) Hanna, T. A. Coord. Chem. ReV. 2004, 248, 429
and references therein.
* To whom correspondence should be addressed. E-mail
Christian.limberg@chemie.hu-berlin.de
ac.rwth-aachen.de (U.S.).
(C.L.);
ulrich.simon@
† Humboldt-Universita¨t zu Berlin.
‡ RWTH Aachen.
9020 Inorganic Chemistry, Vol. 45, No. 22, 2006
10.1021/ic061198c CCC: $33.50
© 2006 American Chemical Society
Published on Web 10/03/2006

[-O-BiIII‚‚‚OdMo-]n Chains
Chart 1
A white precipitate formed, and the suspension was stirred for 2 h
at room temperature. Concentration and filtration yielded a white
solid consisting of 1 and 2 (according to IR analysis), which was
washed with acetone and dried in vacuo. The solid was redissolved
in CH2Cl2, and the resulting solution was stirred for 10 min at room
temperature. Subsequently, all volatiles were removed to yield a
bright yellow powder (320 mg, 73%), which was shown to consist
solely of 2 by means of IR spectroscopy. Analytically pure samples
can be obtained by recrystallization from THF or toluene. Anal.
Calcd for C28H28Bi2O: C, 42.12; H, 3.53. Found: C, 42.52; H,
3.60. 1H NMR (CDCl3): δ 8.20 (d, 4H CH), 7.40 (t, 4H CH), 7.31
(d, 4H CH), 7.24 (t, 4H CH), 2.15 (s, 12H o-CH3) (in agreement
with ref 6a). 13C{1H} NMR (CDCl3): δ 181.4/143.2/135.6/131.3/
128.9/128.3 (CAr), 23.9 (o-CH3). IR (KBr, cm-1): ν˜ 3052 m, 3039
m, 3026 m, 2991 w, 2966 m, 2933 w, 2912 w, 2854 w, 1578 m,
1562 w, 1444 s, 1374 m, 1277 w, 1266 m, 1201 s, 1153 m, 1112
m, 1043 m, 1030 m, 1017 m, 984 w, 951 w, 943 w, 870 w, 792 m,
744 vs, 618 vs, 535 w, 484 w 426 s.
Experimental Section
General Procedures. All manipulations were carried out in a
glovebox or with Schlenk-type techniques with a dry argon
1
atmosphere. The H and 13C{1H} NMR spectra were recorded on
a Bruker AV 400 NMR spectrometer (1H, 400.13 MHz; 13C, 100.61
MHz) with C6D6 and CD2Cl2 as the solvents at 22 °C. The 1H and
13C{1H} NMR spectra were calibrated against the residual proton
and natural abundance 13C resonances of the deuterated solvent
(benzene-d6, δH 7.15 and δC 128.02; dichloromethane-d2, δH 5.32
and δC 53.5). Microanalyses were performed on a Leco CHNS-
932 elemental analyzer. Infrared (IR) spectra were recorded using
samples prepared as KBr pellets with a Digilab Excalibur FTS 4000
FTIR spectrometer. The powder morphology was examined by SEM
analysis using a LEO Supra 35VP equipped with an EDX system
(Inca Energy 200, Oxford, U.K.). For that purpose, all samples were
coated with carbon. Conductivity measurements were performed
with an Agilent 4192 A impedance analyzer. Electrical character-
ization was performed using a Keithley 2400 source meter and an
Agilent 4192 A (for response measurements).
[(o-Tolyl)2BiOCH3]∞, 3. A 1 M solution (0.9 mL) of (n-C4H9)4-
NOH in CH3OH (0.9 mmol) was added to a solution of 809 mg
(0.9 mmol) of [(o-tolyl)2Bi(OTf)(hmpa)2] in 10 mL of CH2Cl2. A
turbid solution formed, which was stirred for 2 h at room
temperature and filtered. The filtrate was concentrated until a white
precipitate began to form. After the addition of a minimum amount
of CH2Cl2 to redissolve the precipitate, the clear solution was cooled
to -30 °C. After a few days, a white precipitate had formed, which
was filtered off and dried in vacuo to yield 0.262 mg (0.62 mmol,
Materials. Solvents were purified, dried, and degassed prior to
use. [(o-Tolyl)2Bi(OTf)(hmpa)2] (hmpa ) hexamethylphosphor-
amide),3 [Cp2MoH2],4 and [MeCp2MoH2]5 were prepared according
to the literature procedures. (n-C4H9)4NOH (a 1 M solution in
methanol or a 40% solution in water) was obtained commercially
and used as received.
[(o-Tolyl)2BiOH]∞, 1. A 7.25 g portion of a 40% solution of
(n-C4H9)4NOH in H2O (11.1 mmol) was diluted with an additional
60 mL of H2O and then added to a solution of 2 g (2.2 mmol) of
[(o-tolyl)2Bi(OTf)(hmpa)2] in 20 mL of THF. A colorless solid
precipitated, which was filtered off after the suspension had been
stirred for 1 h at room temperature. The residue was washed with
water and diethyl ether. Finally, it was dried in vacuo to yield 591
mg (1.45 mmol, 65%) of a crude product, which was shown by IR
spectroscopy and elemental analysis to consist mainly of 1.
Analytically pure samples can be obtained by recrystallization from
ethanol (96%, used as received). Anal. Calcd for C14H15BiO: C,
41.19; H, 3.70. Found: C, 41.04; H, 3.63. IR (KBr, cm-1): ν˜ 3617
m, 3065 w, 3045 m, 2992 w, 2961 w, 2950 w, 2917 w, 2856 w,
1579 w, 1560 w, 1445 br, 1376 w, 1265 m, 1197 m, 1157 w, 1116
w, 1043 w, 1029 w, 938 w, 858 w, 834 s, 790 m, 737 s, 707 w,
644 w, 540 m, 486 m, 425 m.
1
69%) of 3. H NMR spectroscopic data revealed a contamination
by small amounts of the condensation product [(o-tolyl)2BiOBi(o-
tolyl)2], 2, which can be separated by recrystallization from CH2-
Cl2 at -30 °C. The first fraction of the crystallization is purer when
5 mL of MeOH are added to the initial solution of [(o-tolyl)2Bi-
(OTf)(hmpa)2] in CH2Cl2. Anal. Calcd for C15H17BiO: C, 42.66;
1
H, 4.06. Found: C, 42.41; H, 3.93. H NMR (C6D6): δ 8.38 (d,
2H CH), 7.30 (t, 2H CH), 7.19 (d, 2H CH), 7.09 (t, 2H CH), 4.23
(s, 3H OCH3), 2.11 (s, 6H o-CH3). 13C{1H} NMR (C6D6): δ 143.5
(o-CCH3), 135.9/132.0/129.4/128.8 (CH), 57.3 (OCH3), 23.8 (o-
CCH3). The ipso C atom was not detected. IR (KBr, cm-1): ν˜ 3119
w, 3048 s, 2993 m, 2969 m, 2899 m, 2856 m, 2785 s, 1579 m,
1561 w, 1459 m, 1446 s, 1376 m, 1275 m, 1267 m, 1201 s, 1154
m, 1114 m, 1020 vs, 947 w, 872 w, 792 m, 751 vs, 738 vs, 644 w,
616 s, 537 m, 490 w, 444 m, 430 s. MS (EI): 482 [(o-tolyl)3Bi],
422 [M•+], 391 [M•+ - OCH3], 300 [(o-tolyl)Bi], 182 [{(CH3)-
C6H4}2], 91 [(CH3)C6H4].
[(o-Tolyl)2BiOBi(o-tolyl)2], 2. A 40% solution of (n-C4H9)4NOH
in H2O (0.72 mL, 1.1 mmol) was added dropwise to a solution of
1.0 g (1.1 mmol) of [(o-tolyl)2Bi(OTf)(hmpa)2] in 10 mL of acetone.
[
MeCp2Mo(H)(Bi(o-tolyl)2)], 4, and [Cp2Mo(H)(Bi(o-tolyl)2)],
5. Compound 3 (83.2 mg, 0.20 mmol) was dissolved in 5 mL of
CH2Cl2, and a solution of 45.9 mg (0.18 mmol) of [MeCp2MoH2]
in 2 mL of CH2Cl2 was added at room temperature.The resulting
red solution was left to stir for 2 h, and subsequently, the solvent
(2) (a) Roggan, S.; Limberg, C.; Ziemer, B. Angew. Chem. 2005, 117,
5393; Angew. Chem., Int. Ed. 2005, 44, 5259. (b) Roggan, S.; Limberg,
C.; Brandt, M.; Ziemer, B. J. Organomet. Chem. 2005, 690, 5282.
(3) Matano, Y.; Miyamatsu, T.; Suzuki, H. Organometallics 1996, 15,
1951.
(4) Silavwe, N. D.; Castellani, M. P.; Tyler, D. R.; Beck, M. A.;
Lichtenhan, J. D.; Doherty, N. M. Inorg. Synth. 1992, 29, 204.
(5) Luo, L.; Lanza, G.; Fragala, I. L.; Stern, C. L.; Marks, T. J. J. Am.
Chem. Soc. 1998, 120, 3111.
(6) (a) Matano, Y.; Nomura, H.; Suzuki, H.; Shiro, M.; Nakano, H. J.
Am. Chem. Soc. 2001, 123, 10954. (b) Breunig, H. J.; Ebert, K. H.,
Schulz, R. E. Z. Naturforsch. 1995, 50b, 735.
Inorganic Chemistry, Vol. 45, No. 22, 2006 9021

Roggan et al.
was evaporated in vacuo to yield a red oil. Multiple extractions of
the residue with small portions of hexane resulted in red colored
hexane fractions and a white residue consisting of unreacted 3. The
unified hexane extracts were concentrated to approximately 2 mL
and cooled to 4 °C. After a few days, red crystals had formed,
which were collected and washed with hexane to yield 98 mg (0.15
mmol, 83%) of analytically pure 4. Anal. Calcd for C26H29BiMo:
the diffractometer (Stoe IPDS I for 1 and 3, Stoe IPDS2T for 4)
using Mo KR radiation, λ ) 0.71073 Å, and the structures were
solved by direct methods (SHELXS-97),8 refined versus F2
(SHELXL-97)9 with anisotropic temperature factors for all non-
hydrogen atoms. The hydrogen atoms of the BiOH and MoH units
in 1 and 4 were found from a difference Fourier map and freely
refined in the following least-squares cycles. All remaining
hydrogen atoms were added geometrically and refined using a riding
model. The crystal of 3 was twinned with disordered tolyl groups
at Bi.
1
C, 48.31; H, 4.52. Found: C, 48.51; H, 4.82. H NMR (CD2Cl2):
δ 7.59 (d, 2H CtolH), 7.16 (m, 4H, CtolH), 6.88 (m, 2H, CtolH),
4.51 (bs, 2H, CCpH), 4.45 (bs, 2H, CCpH), 4.27 (bs, 2H, CCpH),
4.23 (bs, 2H, CCpH), 2.43 (s, 6H, CtolCH3), 2.01 (s, 6H, CcpCH3),
-9.31 (s, 1H, MoH). 13C{1H} NMR (CD2Cl2): δ 144.3 (CtolCH3),
140.7/127.7/127.2/126.6 (CtolH), 97.3 (CCpCH3), 84.0/83.1/80.8/79.6
(CcpH), 27.8 (CtolCH3)/15.6 (CcpCH3). IR (KBr, cm-1): ν˜ 3101 w,
3084 w, 3047 m, 3024 sh, 2978 m, 2953 w, 2918 w, 2864 w, 1813
m, 1574 m, 1558 w, 1481 w, 1456 s, 1445 s, 1402 w, 1375 m,
1273 w, 1261 m, 1226 w, 1200 m, 1157 m, 1113 m, 1061 w, 1030
s, 1015 m, 978 w, 937 w, 926 w, 878 m, 841 m, 826 w, 787 s, 745
vs, 706 w, 600 w, 575 m, 432 s, 417 s.
For the structural characterization of the materials obtained from
the polyol method, we carried out powder XRD measurements (Cu
KR, λ ) 1.5418 Å) on thin films with a Huber Image Plate in
transmission. The powder diffraction patterns were analysed using
Stoe WinXPow 1.06 Software (Stoe & CIE GmbH).
Results and Discussion
Breaking Up II with Hydroxide. Previous studies probing
the scope of breaking up polymer II by external donors to
obtain more soluble molecular compounds had shown that
The synthesis of 5 was performed analogously with 100 mg (0.24
mmol) of 3 and 54 mg (0.24 mmol) of [Cp2MoH2]. Yield: 120
mg (0.19 mmol, 79%). Anal. Calcd for C24H25BiMo: C, 46.62; H,
-
treatment with Cp*MoO3 yields the complex anion III
1
4.08. Found: C, 46.51; H, 4.40. H NMR (CD2Cl2): δ 7.53 (d,
(Chart 1).2b Envisaging the formation of novel compounds,
possibly containing Mo-O-Bi-O-Bi units, via condensa-
tion reactions, we offered the most simple Lewis base, HO-,
as a reagent in the form of NBu4OH dissolved in MeOH.
After the addition of NBu4OH/MeOH, the stirring of the
mixture, and the subsequent workup, colorless crystals were
obtained that were investigated by means of X-ray analysis
(Figure 1). It turned out that these crystals belonged to a
[(o-tolyl)2BiOH]n coordination polymer, 1, where the hy-
droxide ligands bridge two Bi centers each.
2H, CtolH), 7.15 (m, 4H, CtolH), 6.90 (m, 2H, CtolH), 4.57 (s, 10H,
CpH), 2.42 (s, 6H, CCH3), -9.40 (s, 1H, MoH). 13C{1H} NMR
(CD2Cl2): δ 144.2 (CArCH3), 140.3/127.8/127.4/126.7 (CArH), 79.9
(CcpH), 27.7 (CArCH3). IR (KBr, cm-1): ν˜ 3108 w, 3044 m, 3023
w, 2983 w, 2959 w, 2930 w, 2852 w, 1836 m, 1575 m, 1559 w,
1444 s, 1426 m, 1409 m, 1374 m, 1263 m, 1199 w, 1153 w, 1106
m, 1061 w, 1043 w, 1030 w, 1013 m, 993 m, 896 w, 864 w, 831
m, 820 m, 785 s, 751 vs, 741 vs, 587 m, 432 m, 417 m.
Deposition of Thick-Film Sensing Layer. According to a well-
established protocol,7 the powder material was dispersed by mixing
it in a mortar with an aqueous polyethylene imine (0.5 wt %)
solution, and the resulting dispersion with a solid amount of 0.08
mol/L was then applied to an interdigital capacitor electrode
structure. The substrate was first dried for 48 h at room temperature
in air. After that it was held for 6 h at 700 °C in air to remove the
organic residues.
There is some very early work describing reactions (like
treatment of Ph2BiI with NaOH in EtOH and [(p-CH3C6H4)2-
BiC]2 with an alkaline Cu+ or Ag+ solution) whose products
were assigned the formula Ar2BiOH (Ar ) Ph, p-tolyl) even
though analytic/spectroscopic support is missing. In general,
structurally characterized complexes containing an OH-
ligand covalently bonded to a BiIII center are rather rare and
include the hexanuclear cationic bismuth complex [Bi6O4-
(OH)4](ClO4)6‚7H2O, IV, whose crystal structure was de-
termined by X-ray diffraction, as well as by neutron
diffraction studies,11a and the related citrate cluster anion
[Bi6O4(OH)(cit)3(H2O)3]3-,V, where three of the hydroxide
ligands in IV are substituted by the hydroxylic groups of
three citrate ligands.11b In clusters IV and V, the hydroxide
anions act as µ3 bridging ligands, while µ2-OH bridges are
found in the dinuclear compounds [(η2-NO3)2(tpy)Bi(µ-
OH)2Bi(tpy)(η2-NO3)2] (tpy ) terpyridine), VI,12 and [(η2-
NO3)2(phen)Bi(µ-OH)2Bi(phen)(η2-NO3)2] (phen ) 1,10
phenanthroline), VII.13 A nonbridging terminal Bi-OH unit
Electrical Characterization. Electrical measurements were
performed using a source meter at a measuring voltage of 1 V.
The measurements were carried out in a temperature range between
500 and 400 °C in 25° steps (at lower temperature the resistance
of the material exceeds the measurement limits). Reference
measurements were performed in synthetic air, which was also used
as the carrier gas for the test gases H2, CO, NO2 EtOH, and
propylene, whereas for NO nitrogen was chosen as the carrier gas.
For the gas-sensing experiments, the test gases were mixed with
synthetic air to reach a continuous gas flow of 100 sccm each.
Conditioning of the materials was carried out at 500 °C for 240
min in air and at the next temperatures for 90 min.
X-ray Data Collection. Suitable single crystals of 1 were
obtained by slow evaporation of the solvent from a CH2Cl2 solution
of the precipitate obtained by cooling the solution produced in the
reaction between II and an equimolar amount of (n-C4H9)4NOH
(1 M solution in MeOH) in MeOH. Single crystals of 3 were
obtained by cooling the filtrate of the reaction mixture leading to
3 to -30 °C. Suitable single crystals of 4 were obtained by cooling
a hexane solution of 4 to 4 °C. The crystals were mounted on a
glass fiber and then transferred to the cold nitrogen gas stream of
(8) Sheldrick, G. M. SHELXS-97, Program for Crystal Structure Solution;
University of Go¨ttingen: Go¨ttingen, Germany, 1997.
(9) Sheldrick, G. M. SHELXL-97, Program for Crystal Structure Refine-
ment; University of Go¨ttingen: Go¨ttingen,Germany, 1997.
(10) (a)Rozenblumowna S.; Weil S. Bull. TraV. Inst. Pharm. Etat 1927, 1,
3. (b) Hartmann, H.; Habenicht, G.; Reiss, W. Z. Anorg. Allg. Chem.
1962, 317, 54.
(11) (a) Sundvall, B. Inorg. Chem. 1983, 22, 1906. (b) Asato, E.; Katsura,
K.; Mikuriya, M.; Fujii, T.; Reedijk, J. Chem. Lett. 1992, 1967.
(12) Junk, P. C.; Louis, L. M. Z. Anorg. Allg. Chem. 2000, 626, 556.
(7) Siemons, M.; Simon, U. Sens. Actuators, B 2006, published online
March 9, http://dx.doi.org/10.1016/j.snb.2006.01.049.
9022 Inorganic Chemistry, Vol. 45, No. 22, 2006

[-O-BiIII‚‚‚OdMo-]n Chains
Table 1. Crystallographic Data of 1, 3, and 4
1
3
4
empirical formula
fw
C14H15BiO
408.24
C15H17BiO
422.27
C26H29BiMo
646.41
cryst size (mm)
cryst syst
space group
a (Å)
b (Å)
c (Å)
0.40 × 0.36 × 0.28
orthorhombic
Pbca
13.8069(7)
8.4987(6)
21.167(2)
90
90
90
2483.7(2)
2.183
180(2)
0.30 × 0.22 × 0.18
orthorhombic
Cm2a
21.013(4)
16.391(3)
4.1400(5)
90
90
90
1425.9(4)
1.967
180(2)
0.20 × 0.17 × 0.16
triclinic
P1h
7.9109(5)
9.1244(7)
15.7401(9)
79.837(5)
84.125(5)
81.073(5)
1101.5(2)
1.949
R (deg)
â (deg)
γ (deg)
V (Å3)
F
calcd (g cm-3
)
T (K)
100(2)
abs corrn
numerical
14.173
1520
numerical
8.555
620
abs coeff (µ, mm -1
F(000)
)
12.347
792
index ranges
-18 e h e 18
-11 e k e 11
-28 e l e 28
3.41-29.37
3344
3344 (Rint ) 0.0000)
3344/1/151
0.997
-25 e h e 25
-20 e k e 20
-5 e l e 5
3.15-25.96
7439
1418 (Rint ) 0.0475)
1418/16/50
1.080
-10 e h e 11
-13 e k e 13
-23 e l e 23
3.57-31.84
24 860
7476 (Rint ) 0.0395)
7476/0/262
1.194
θ range (deg)
reflns collected
independent reflns
data/restraints/params
GOF on F2 a
Z
8
4
2
R1b [I > 2σI]
0.0278
0.0494
1.706 and -1.962
0.0215
0.0532
1.188 and -0692
0.0274
0.0459
1.560 and -1.782
wR2c (all data)
largest diff. peak and hole (e Å-3
)
2
2
2
2
2
a GOF ) {∑[w(Fo - Fc )2]/(Nobsd - Nparams)}1/2
.
b R1 ) ∑(||Fo| - |Fc||)/∑|Fo|. c wR2 ) [∑(|Fo| - |Fc| )2/∑(Fo )]1/2
.
a pseudo-trigonal bipyramidal fashion (also called “seesaw”)
where the oxygen atoms occupy the apical positions. The
angle between them amounts to 161.09(3)° indicating a
distorted coordination. Aggregation of R2BiX molecules
(where X represents an electronegative substituent) under
the formation of polymeric chains is not uncommon and
reflects the lewis acidic character of the BiIII center.15 The
linkages between the monomeric units are accomplished via
secondary interactions between lone pairs of the X ligand
and σ* Bi-X acceptor orbitals of neighboring molecules,
which results in the observed trans arrangement of the
electronegative ligands in the polymers formed. Similar
structural features are observed, for example, in II,2a where
the O-Bi-O angle is 170.46(9)°, or in a series of bismuth
halides ligated by organometallic complex fragments such
as, for example, [BiX{Mo(CO)3Cp}2] (X ) Cl or Br) with
a Cl-Bi-Cl angle of 151.6(1)° and a Br-Bi-Br angle of
148.2(1)°.16 Deviation of the X-Bi-X units from linearity
indicates a less-efficient overlap between X and the σ* Bi-X
orbitals and thus a weakening of the secondary bond; in fact,
the BiX2 angle correlates with the strength of this interaction.
Moreover, with the increasing strength of the secondary
bond, the corresponding primary bond is weakened so that
Figure 1. Molecular structure of 1; all hydrogen atoms, except for the
hydroxylic H atom, were omitted for clarity. Selected bond lengths (Å)
and angles (deg): Bi-C1 ) 2.255(4), Bi-C8 ) 2.256(4), Bi-O ) 2.319(3),
Bi-O′′ ) 2.343(3), O-H1 ) 0.83(4), Bi-O-Bi′ ) 135.6(2), O′′-Bi-O
) 161.09(3), C1-Bi-O ) 87.0(2), C8-Bi-O ) 82.0(2), C8-Bi-O′′ )
81.9(2), C1-Bi-O′′ ) 85.4(2), C8-Bi-C1 ) 97.7(2), Bi-O-H1 ) 108-
(4).
(13) Barbour, L. J.; Belfield, S. J.; Junk, P. C.; Smith, M. K. Aust. J. Chem.
1998, 51, 337.
(14) Battaglia, L. P.; Bonamartini Corrradi, A.; Pelizzi, C.; Pelosi, G.;
Tarasconi, P. J. Chem. Soc., Dalton Trans. 1990, 3857.
(15) Roggan, S.; Limberg. C. Inorg. Chim. Acta in press.
(16) (a) Clegg, W.; Compton, N. A.; Errington, R. J.; Norman, N. C.;
Tucker, A. J.; Winter, M. J. J. Chem. Soc., Dalton Trans. 1988, 2941.
(b) Clegg, W.; Compton, N. A.; Errington, R. J.; Fisher, G. A.;
Hockless, D. C. R.; Norman, N. C.; Williams, N. A. L.; Stratford, S.
E.; Nichols, S. J.; Jarrett, P. S.; Orpen, A. G. J. Chem. Soc., Dalton
Trans. 1992, 193.
is only found in [Bi(dapt)(OH)] (H2dapt ) 2,6-diacetylpyr-
idine bis (2-theonylhydrazone)), VIII.14
Compound 1 crystallizes as a coordination polymer where
(o-tolyl)2BiOH moieties are linked via intermolecular inter-
actions between the hydroxide ligands and the bismuth atoms
of adjacent molecules. The bismuth center is coordinated in
Inorganic Chemistry, Vol. 45, No. 22, 2006 9023

Roggan et al.
Scheme 1
Scheme 2
in extreme cases the two Bi-X bond lengths align. In 1, the
first Bi-O bond length is 2.319(3) Å being only a little
shorter than the second Bi-O′ bond (2.343(3) Å), so that a
distinction between the covalently bonded O atom and the
coordinatively bonded O atom (i.e., primary vs secondary
interaction) is hardly possible. These similar Bi-O bond
lengths, as well as the nearly linear O-Bi-O unit, suggest
a strong secondary bond. In the case of II, where the
O-Bi-O angle is even closer to 180° (170.46(9)°), the
Bi-O bond distances are nearly identical, too (2.310(2) vs
2.385(2) Å). The Bi-OH bond lengths of dinuclear com-
plexes VI and VII are similar (2.20(1) and 2.32(1) Å in the
terpyridine derivative, VI, and 2.214(3) and 2.232(3) Å in
the more symmetrically bridged VII). As a result of the
planar Bi2O2 diamond core structure, the O-Bi-O angles
are very acute and amount to 66.5(4)°. Consequently, the
acceptor orbital for the linking of one (formally separated)
Bi-OH unit to the second Bi center cannot be the Bi-OH
σ* orbital of the latter, and hence, no bond elongation as
observed in 1 occurs resulting in shorter Bi-OH bonds in
VI and VII, as compared to those in 1. For the abovemen-
tioned complexes, the shortest Bi-OH bond length is found
in VIII (2.08(1) Å) which contains an isolated Bi-OH bond
without any bridging interaction.
Accordingly, the addition of hydroxide to II breaks up all
of the Mo-O-Bi contacts to give a homometallic compound
1; the condensation product [(o-tolyl)2Bi-O-Bi(o-tolyl)2],
2, (that has been mentioned previously6a as a reaction
byproduct according to 1H NMR spectroscopic assignments)
and [NBu4][Cp*MoO3]17 were identified as additional prod-
ucts by means of NMR spectroscopy (Scheme 1), and
unfortunately, these crystallize together with 1 which con-
sequently can only be separated from them in small amounts
mechanically.
replacement of hydroxide by methoxide ligands occurred so
that [(o-tolyl)2BiOMe]n, 3, was isolated instead of 1 (Scheme
2). The result of an X-ray crystal structure determination is
shown in Figure 2.
In recent years, the increasing interest in oxidic materials
based on bismuth has led to the synthesis of numerous
soluble bismuth alkoxides which were employed as starting
materials for sol gel or chemical vapor-deposition processes.
However, most of the structurally characterized bismuth-
(III) alkoxides are homoleptic: for example, [Bi(OC6H3Me2-
2,6)3],18 [{Bi(OCH(CF3)2)3(thf)}2],19 IX, [Bi(OCH2CH2-
OCH3)3]∞,20 X, and [Bi(OEt)3]8,20c,21 XI. M. Wieber and U.
Baudis in 1978 reported the first nonhomoleptic BiIII-
alkoxide complexes with the general formula Ph2BiOR (R
) Me, Et, iPr, Ph), as well as CH3Bi(OR)2 (R )Me, Et,
iPr), which were prepared by metathetical reactions between
the organobismuthbromides and the corresponding sodium
(18) Evans, W. J.; Hain, J. H., Jr.; Ziller, J. W. J. Chem. Soc., Chem.
Commun. 1989, 1628.
(19) (a) Jones, C. M.; Burkart, M. D.; Whitmire, K. H. Angew. Chem. 1992,
104, 466; Angew. Chem., Int. Ed. Engl. 1992, 31, 451. (b) Jones, C.
M.; Burkart, M. D.; Bachman, R. E.; Serra, D. L.; Hwu, S.-J.;
Whitmire, K. H. Inorg. Chem. 1993, 32, 5136.
(20) (a) Matchett, M. A.; Chiang, M. Y.; Buhro, W. E. Inorg. Chem. 1990,
29, 358. (b) Massiani, M.-C.; Papiernik, R.; Hubert-Pfalzgraf, L. G.;
Daran, J.-C. J. Chem. Soc., Chem. Commun. 1990, 301. (c) Massiani,
M.-C.; Papiernik, R.; Hubert-Pfalzgraf, L. G.; Daran, J.-C. Polyhedron
1991, 10, 437.
Compound 1 is by far the most simple well-characterized
Bi-OH complex known so far, so that it seemed to be an
interesting compound in its own right (for instance as a
starting material for further syntheses). Hence, attempts were
made to prepare it in a higher yield via the reaction of [(o-
3
tolyl)2Bi(hmpa)2]SO3CF3 with NBu4OH in MeOH. These
might have been successful in the first place, but then
(21) (a) Kessler, V. G.; Turova, N. Ya.; Turevskaya, E. P. Inorg. Chem.
Commun. 2002, 5, 549.
(17) Sundermeyer, J.; Radius, U.; Burschka, C. Chem. Ber. 1992, 125, 2379.
9024 Inorganic Chemistry, Vol. 45, No. 22, 2006

[-O-BiIII‚‚‚OdMo-]n Chains
its peculiar structural behavior remains unclear, particularly
considering the structures of the aryloxides XII and XIII,
which consist of polymeric chains similar to the ones of 1
and 3 but contain Bi centers in a seesaw-like coordination
sphere, as found in hydroxide 1. The Bi-O bonds in 3
(2.295(5) and 2.399(5) Å) are more distinct than those in 1
or in II, which again points to a weak secondary interaction.
They are shorter than those in aryloxides XII (2.4105(7) and
XIII (2.382(7) Å), where the secondary bonds are as strong
as the primary ones leading to equal distances. Because a
comparison with other diorganobismuthalkoxides (where the
alkoxide ligand is aliphatic) is not possible (vide supra), the
structural parameters of 3 involving the alkoxidic units are
compared with the ones of homoleptic bismuth alkoxides
bearing aliphatic alkoxide ligands. Because of the strong
Lewis acidity of the BiIII center, there is only one monomeric
Bi(OR)3 complex known with R representing a saturated
hydrocarbon residue, Bi(OCPh3)3, whose Bi-O bonds aver-
age 2.074(4) Å which is typical for terminal Bi-alkoxide
ligands.26a In other cases with less bulky organic residues,
dimers or oligomers are formed that exhibit terminal OR
ligands with or without a trans-coordinated OR counterligand.
Not surprisingly, the Bi-O bonds displayed by OR ligands
positioned trans to a vacant coordination site compare to
those in Bi(OCPh3)3 and are shorter than those in 3 (e.g.,
2.064(8), 2.07(2), and 2.06 (averaged) Å in IX, X, and XI,
respectively). On the other hand, as expected, the Bi-O bond
lengths found in trans RO-Bi-OR arrangements compare
to those in 3: the Bi-O distances are 2.188(7) Å in IX,
2.20(1) Å in X or 2.265(8) and 2.285(9) Å in the potassium-
(tertbutoxy) bismuthate K[(OtBu)4Bi], XIV.26b
The question of why the synthesis shown in Scheme 1
proceeds to give 1 and not 3 cannot be answered with
certainty. Compound 1 is poorly soluble in methanol, which
was used for the reaction shown in Scheme 1, so that rapid
precipitation may have prevented it from undergoing alco-
holysis with MeOH, while some condensation to 2 could
still proceed. This hypothesis is supported by the slow
conversion of 1 to 3 and 2 (accompanied by some decom-
position to (o-tol)3Bi) that can be observed when pure 1 (vide
infra) is stirred in dry MeOH for a prolonged time. The
addition of CH2Cl2 to a suspension of 1 in MeOH accelerates
the formation of 3 and, at the same time, reduces that of 2.
If this reactivity of pure 1 is considered, the formation of 3,
as shown in Scheme 2, can be well understood in terms of
the solubilities/aggregation and precipitation rates in different
solvents. In solution, the initially formed hydroxide reacts
readily with any MeOH present, while the condensation to
2 still occurs. The latter can be suppressed by the addition
of excess MeOH (Experimental Section). This alcoholysis
generates water, and it seems peculiar that 3 should be
resistant to that. However, during the reaction, small amounts
Figure 2. Molecular structure of 3; all hydrogen atoms were omitted for
clarity. Selected bond lengths (Å) and angles (deg): Bi-C1 ) 2.270(6),
Bi-C8 ) 2.308(6), Bi-O ) 2.295(5), Bi-O′ ) 2.399(5), O-C15 )
1.42(2), Bi-O-Bi′′ ) 123.7(3), O′-Bi-O ) 123.7(3), C1-Bi-O )
129.4(2), C8-Bi-O ) 117.5(2), C1-Bi-O′ ) 90.9(2), C8-Bi-O′ )
89.2(2), C1-Bi-C8 ) 96.2(3), Bi-O-C15 ) 124.5(5), C15-O-Bi′′ )
111.7(5).
alkoxides. However, none of them could be characterized
via X-ray diffraction, and their preparations involve tedious
filtrations of the alkali salt byproducts.22 In this context it
should also be noted that [(o-tolyl)2BiOMe] has been
mentioned before as a byproduct of the reaction of [(o-
tolyl)3BidN-C(O)-Ar] with methanol in an NMR tube
experiment: the signals in the 1H NMR spectrum of the crude
product were “tentatively assigned to [(o-tolyl)2BiOMe] but
attempts to prepare the sample independently failed”.6 In fact
the only structurally characterized diorganobismuthalkoxide
complexes we are aware of are the aryloxides 10-arylphe-
nothiabismine 5,5-dioxide (aryl ) 2-methoxyphenoxy and
4-methoxyphenoxy)23 and [Et2Bi(OAr)]∞,24 (Ar ) C6F5 (XII),
Ph (XIII)). Compound 3 represents a diorgano bismuth
alkoxide, and its bismuth center is covalently bound to two
carbon atoms of the tolyl ligands and an oxygen atom of
the methoxy ligand. As in 1, an additional secondary
interaction with the methoxy oxygen of a neighboring
molecule is observed leading to a one-dimensional polymeric
structure. However, in contrast to 1 the O′-Bi-O angle is
only 123.7(3)°, indicating that this secondary interaction is
comparatively weak. The structure can thus not be described
as ψ-trigonal bipyramidal, and the assignment to a ψ-square
pyramidal geometry which could have been expected if the
lone pair was stereoactive does not seem to be appropriate,
either. The structure should thus be discussed most ad-
equately as being distorted tetrahedral. Variations between
these two different structural types have been observed
previously for 10-electron AB4E compounds, and electronic
reasons (e.g., π-donor properties of ligands) were discussed.25
However, 3 differs from 1 only in that a proton has been
exchanged by a methyl group, and therefore, the reason for
(25) (a) Gimarc, B. M.; Khan, S. A. J. Am. Chem. Soc. 1978, 100, 2340.
(b) Clegg, W.; Compton, N. A.; Errington, R. J.; Fisher, G. A.;
Hockless, D. C. R.; Norman, N. C.; Orpen, A. G.; Stratford, S. E. J.
Chem. Soc., Dalton Trans. 1992, 3515.
(26) (a) Hanna, T. A.; Keitany, G.; Ibarra, C.; Sommer, R. D.; Rheingold,
A. L. Polyhedron, 2001, 20, 2451. (b) Veith, M.; Yu, E.-C.; Huch, V.
Chem.sEur. J. 1995, 1, 26.
(22) (a) Wieber, M.; Baudis, U. Z. Anorg. Allg. Chem. 1978, 439, 134. (b)
Wieber, M.; Baudis, U. Z. Anorg. Allg. Chem. 1978, 439, 139.
(23) Murafuji, T.; Nagasue, M.; Tashiro, Y.; Sugihara, Y. Organometallics
2000, 19, 1003.
(24) Whitmire, K. H.; Hutchison, J. C.; McKnight, A. L.; Jones, C. M. J.
Chem. Soc., Chem. Commun. 1992, 1021.
Inorganic Chemistry, Vol. 45, No. 22, 2006 9025

Roggan et al.
Scheme 3
of a white precipitate form, which is purely inorganic in
nature according to spectroscopic analysis. A possible
explanation is, therefore, that all water generated is scavenged
in the course of successive reactions that lead to the complete
decomposition of a small part of 3.
As outlined above diorganobismuthalkoxides are more
prominent than bismuthhydroxides; however, well-character-
ized examples that are available in high yields are rare.
Therefore, 3 might be an interesting starting material for
future research, too, and the experiments described below
support this potential.
Turning the attention back onto 1 first and a suitable
synthesis leading to it in a pure form, we decided to
reinvestigate the reaction between [(o-tolyl)2Bi(hmpa)2]SO3-
CF3 3 and NBu4OH in a mixture of thf/water as the solvent.
This indeed led to a white precipitate consisting of 1 as
shown by elemental analysis and IR spectroscopy. Dissolu-
tion of this solid in CDCl3 or C6D6 for NMR spectroscopic
investigations, however, only showed signals for (o-tolyl)2Bi-
O-Bi(o-tolyl)2, 2, and H2O, in addition to signals belonging
to small quantities of a second unidentified component (i.e.,
dissolution in nonprotic solvents immediately triggers the
condensation process (also compare, for instance, ref 6b)).
The latter can also be performed on a preparative scale, thus
offering a synthetic route to 2 which so far had only been
characterized via a 1H NMR spectrum:6a recrystallization of
the abovementioned white precipitate 1 from hot toluene
leads to a different white precipitate which was shown by
analysis to be pure 2. Accordingly, all results can be
summarized as follows: Reactions of II or [(o-tolyl)2Bi-
(hmpa)2]SO3CF3 with NBu4OH lead to 1 initially. However,
appreciable amounts of 1 can only be isolated when
precipitation occurs rapidly, as in the presence of alcohols
and water (in many instances, we noted that even small
amounts of water are very beneficial). Otherwise, it is labile
in solution with respect to condensation reactions leading to
2; it can only escape them if precipitation proceeds faster
than condensation. In the presence of MeOH, the alcoholysis
of 1 to 3 is apparently faster than its condensation to 2, so
that 3 is isolated predominantly.
Breaking Up II with Molybdocenedihydrides. It further
seemed interesting to explore the behavior of II in contact
with molybdocenedihydrides: Previously, it had been shown
that [MeCp2MoH2] and [Bi(OtBu)3] react to give, in course
of the reactions shown in Scheme 3, three novel complexes
(XV, XVI, and XVII) which were isolated and character-
ized.27
C-H-activation reactions initiated by complex-induced
proximity effects in XVI allow for the formation of XVII
featuring a, until then, unprecedented bonding situation,
where a planar µ3-η5: η1: η1-MeCp ligand bridges three metal
centers that are additionally linked via metal bonds. As a
reaction intermediate, the novel molybdocene monohydride
species XVIII was identified by NMR spectroscopy, but it
eluded isolation as its half-life in solution is only ∼15 min.
If the low Brønstedt acidity of molybdocene dihydride is
taken into account, its reaction with the bismuthalkoxide to
form XVIII under elimination of alcohol should not proceed
by a simple acid-base reaction but via a precursor complex
involving the lone pair of either the molybdocene unit or
the bismuth center.28a Hence, the question of whether
(27) Roggan, S.; Limberg, C.; Ziemer, B.; Brandt, M. Angew. Chem. 2004,
116, 2906; Angew. Chem., Int. Ed. 2004, 43, 2846.
(28) (a) Hunger, M.; Limberg, C.; Kaifer, E.; Rutsch, P. J. Organomet.
Chem. 2002, 641, 9. (b) Roggan, S.; Schnakenburg, G.; Limberg, C.;
Sandho¨fner, S.; Pritzkow, H.; Ziemer, B. Chem.sEur. J. 2005, 11,
225.
9026 Inorganic Chemistry, Vol. 45, No. 22, 2006

[-O-BiIII‚‚‚OdMo-]n Chains
Scheme 4
molybdocene dihydride is also capable of breaking up the
polymer II, either by formation Mo-OH/Bi-Mo units or
by generation Bi-OH/Mo-Mo moieties, arose. The reaction
of [MeCp2MoH2] (or [Cp2MoH2]) with II in benzene led to a
red solution that according to 1H NMR spectroscopic results
contained the known17,29 [Cp*2Mo2O5] as well as a new
product whose data in combination with those obtained for
XVIII hinted to the formation of a monohydride compound
[RCp2Mo(H)(Bi(o-tolyl)2)], (R ) Me (4), H (5)); remarkably,
in contrast to XVIII, compounds 4 and 5 appeared to be
thermally stable. Because they could not be isolated from
this mixture in a pure form, we attempted an alternative
synthesis, which we hoped would provide a purer product
on a larger scale, starting from the molybdocene dihydrides
and 3, by analogy to the formation of XVIII according to
Scheme 3 (see Scheme 4).
compares well with the value of 145.8° in [Cp2MoH2]30a or
143.3° in XV,27 but is slightly smaller than the one in [MeCp2-
MoH2] (149.5(1)°).30b The Mo-Bi bond length amounts to
2.8976(3) Å and is thus similar to those found in the other
compounds with bismuth atoms bonded to the Cp2Mo or
MeCp2Mo moieties. These range from 2.788(2) Å in [Cp2-
Mo{Bi(OCH(CF3)2)2}2]28a to 2.966(2) Å in the tetranuclear
compound [{Cp2Mo}2{µ-Bi(OtBu)}2],28b which has a struc-
ture similar to that of XVI. The Mo-H bond length of
1.56(5) Å is significantly shorter than that in [Cp2MoH2]
(neutron diffraction, 1.685(3) Å)30a or [MeCp2MoH2] (1.63(3)
Å).30b The only known homologous molybdocenemonohy-
drides that have so far been structurally characterized appear
to be [Cp2Mo(H)(µ-PPh2)Mn(CO)2Cp]31 and the ionic com-
plex [Cp2Mo(H)(PPh3)]I,32 while no compounds containing
one As or Sb atom bonded to the Mo center are known. In
both cases, the Mo-H bond is longer than in 4 (1.71(9) and
1.74(8) Å, respectively). Mo-H bond distances more similar
to the one in 4 are found in the complexes [Cp2Mo(H)-
(SiH2(C5Me4H)],33 XIX (1.58(6) Å), and [Cp2Mo(H)-
(SnMe3)],34 XX (1.60(5) and 1.51(6) Å in two independent
molecules in the unit cell).
The H-Mo-Bi angle in 4 is 72.0(3)° and, as expected,
is somewhat smaller than the Bi-Mo-Bi angles (74.76(3)-
77.41(3)°) which were found in the above-mentioned
[MoBi2] complexes (replacement of BiR2 by H). The angle
is similar to the H-Mo-E angles measured for the related
group 14 compounds mentioned (e.g., 76(2)° in XIX (E )
Si) or 67.5 and 69.5° in XX (E ) Sn)). The aryl ligands
surround the Bi center in a manner similar to that of the
OtBu ligands coordinated to the Bi atom in XV (i.e., they
show an acute C-Bi-C angle of 94.8(2)°, indicating the
predominant p character of the orbitals involved).
Indeed, 4 and 5 formed in good yields which also allowed
crystallization. The result of the single-crystal X-ray analysis
for 4 is shown in Figure 3.
As mentioned above, in our studies concerning the [MeCp2-
MoH2]/[Bi(OtBu)3] system, there was 1H NMR spectroscopic
evidence that the reaction mixture leading to XVI contained
[MeCp2Mo(H)(Bi(OtBu)2)], XVIII, intermediately, and 4
clearly represents a stable derivative of it. In 4, the common
structure of M(IV) Cp2ML2 complexes is retrieved with a
bent-sandwich configuration of the cyclopentadienyl ligands
about the metal. The bending angle in 4 is 145.3°; thus it
Accordingly, it can be confirmed that 4 and 5 are products
of the reaction between molybdocenedihydrides and II, and
(29) (a) Herberhold, M.; Kremnitz, W.; Razavi, A.; Thewalt, U. Angew.
Chem. 1985, 97, 603; Angew. Chem., Int. Ed. Engl. 1985, 24, 601.
(b) Saurenz, D.; Demirhan, F.; Richard, P.; Poli, R.; Sitzmann, H.
Eur. J. Inorg. Chem. 2002, 1415.
(30) (a) Schultz, A. J.; Stearley, K. L.; Williams, J. M.; Mink, R.; Stucky,
G. D. Inorg. Chem. 1977, 16, 3303. (b) Stender, M.; Oesen, H.;
Blaurock, S.; Hey-Hawkins, E. Z. Anorg. Allg. Chem. 2001, 627, 980.
(31) Barre, C.; Kubicki, M. M.; Leblanc, J.-C.; Moise, C. Inorg. Chem.
1990, 29, 5244.
(32) Azevedo, C. G.; Calhorda, M. J.; Carrondo, M. A. A. F. de C. T.;
Dias, A. R.; Fe´lix, V.; Romao, C. C. J. Organomet. Chem. 1990, 391,
345.
(33) Petri, S. H. A.; Neumann, B.; Stammler, H.-G.; Jutzi, P. J. Organomet.
Chem. 1998, 553, 317.
(34) Protsky, A. N.; Bulychev, B. M.; Soloveichik, G. L.; Belsky, V. K.
Inorg. Chim. Acta 1986, 115, 121.
Figure 3. Molecular structure of 4; all hydrogen atoms, except for those
of the hydride ligand, were omitted for clarity. Selected bond lengths (Å)
and angles (deg): Mo-Bi ) 2.8976(3), Mo-H1 ) 1.56(5), Bi-C1 )
2.279(3), Bi-C8 ) 2.292(3), H1-Mo-Bi ) 72(2), C1-Bi-C8 ) 94.8(2),
C1-Bi-Mo ) 103.14(6), C8-Bi-Mo ) 102.03(7).
Inorganic Chemistry, Vol. 45, No. 22, 2006 9027

Roggan et al.
Figure 4. X-ray diffraction patterns of material A and B from complex II.
Figure 5. SEM pictures of material A.
this, as well as the identification of [Cp2*Mo2O5] as the
byproduct, suggests an attack of a Mo-H function at the
Bi-O-Mo bonds under the formation of Mo-Bi and Mo-
OH linkages; the Mo-OH units then undergo intermolecular
condensation processes to yield the Mo-O-Mo moiety in
[Cp2*Mo2O5]. The stabilities of 4 and 5 are remarkable
considering that they differ from XVIII only in the residues
at the bismuth centers.
Hence, the first representative of the hitherto elusive
substances [Cp2Mo(H)(BiR2)] has been obtained simply by
variation of R from the tert-butoxy ligands to the o-tolyl
ligands, and 3 has been very helpful as a starting material
for a high-yield synthesis.
Precursor Chemistry with II and Other Mo/Bi Com-
pounds. Finally, we were interested in whether II could serve
as a starting material in the polyol method35-37 to produce
bismuthmolybdate (Bi2O3‚nMoO3, n ) 1, 2, or 3) nanopar-
ticles that might find application in gas sensing, catalysis,
etc. Hence, 160 mg of complex II was dispersed in 10 mL
of diethylene glycol (DEG) and heated to 120 °C under
vigorous stirring until a clear solution was obtained. Sub-
sequently, 0.4 mL of deionized water in route A and 0.4
mL 65% nitric acid in route B were added as hydrolyzing
agents, respectively. The emerging suspensions were heated
for 5 h to 170 °C, and then they were allowed to cool to
room temperature. To obtain crystalline materials, the
suspensions were dried at 400 °C (1 h), and the solid obtained
was then annealed at 700 °C for 12 h. A detailed description
of this synthesis method can be found in ref 37. The initial
characterization of the products was carried out by powder
XRD measurements on thin films, and Figure 4 shows the
X-ray diffraction patterns for the products obtained from
routes A and B.
It becomes evident that in both cases a mixture of Bi2-
Mo2O9 and Bi2MoO6 was formed. For route A, about 70%
Bi2MoO6 was achieved, and for route B, we obtained about
80% Bi2Mo2O9. No further byproduct was identified. SEM
characterization of the materials showed that not only
nanoparticles were formed. Figures 5 and 6 show the SEM
pictures of the products obtained from routes A and B,
respectively.
In both cases, mixtures of needles (crystallites) and
nanoparticles are found. The sizes of the primary particles
are around 30-60 nm (route A) and 80-200 nm (route B),
whereas the needles span from the micrometer to millimeter
range. From the energy-dispersive X-ray microanalysis
(EDX) measurements, an accurate assignment to the different
products could not be made. Considering the potential of II
to act as a single-source precursor for the bismuthmolybdate
(35) Figlarz, M.; Fievet, F.; Lagier, J. P. EU 0113281, US A 4539041, JP
04024402, 1982.
(36) Feldmann, C.; Jungk, H.-O. Angew. Chem. 2001, 113, 372; Angew.
Chem., Int. Ed. 2001, 40, 359;.
(37) Siemons, M.; Weirich, Th.; Mayer, J.; Simon, U. Z. Anorg. Allg. Chem.
2004, 630, 2083.
9028 Inorganic Chemistry, Vol. 45, No. 22, 2006

[-O-BiIII‚‚‚OdMo-]n Chains
Figure 6. SEM pictures of material B.
of the gases was 45% at room temperature so that the
measurements were performed in a typical humidity range
representative for ambient conditions. For sample condition-
ing, a preliminary gas flow was applied over 30 min to reach
equilibrium. A detailed description of the experimental setup
can be found in ref 40.
A plot of conductance versus temperature for the com-
pounds prepared is shown in Figure 8, which indicates that
the conductivity is thermally activated, as expected. Accord-
ing to the Arrhenius relation, an activation energy, EA, of
the process could be estimated to be 0.98 eV. Figure 9 shows
the sensitivity dependence versus the temperature. The
absolute sensitivity, S, of a material to a test gas is defined
by
Figure 7. Homogeneous sensor layer formed from material A (Bi2MoO6
(70%)/Bi2Mo2O9 (30%)).
materials, we thus concluded that it facilitates mixing on the
molecular level and the avoidance of sample inhomogene-
ities; however, the preparation of pure phases seems to not
be possible because the crystallization of more than one phase
occurs at the calcination temperature. In addition, the
chelating effect of the diethylene glycol used as the solvent
seems to not be strong enough for the exclusive preparation
of nanoparticles. On the other hand, the product distribution
can be influenced by the use of different hydrolysis agents.
However, for various applications (e.g., as catalyst for the
oxidation of light olefins38 and gas-sensing materials for
hydrocabons),39 bismuthmolybdates are employed not only
in form of single-phases but also as mixtures of different
bismuth molybdate materials.
Therefore the product obtained via preparation A was used
for further investigations with respect to gas-sensing proper-
ties. A photograph of the sensing layer on the interdigital
capacitor structure is shown in Figure 7. The sensing layer
was around 4 mm in diameter. The distance between
electrodes was around 110 µm.
R
S ) g - 1
Rg > Rr
Rg < Rr
Rr
Rr
S ) - 1
Rg
where Rg is the resistance obtained under test gas and Rr is
the reference resistance value for a measurement under
synthetic air.
Only NO, propylene, and ethanol are shown because no
sensitivity was observed toward the other gases. Over the
temperature range investigated, a pronounced sensitivity
toward ethanol could be found, while the sensitivities toward
nitrogen oxide and propylene were only weak.
To measure the response time of the sample, conductivity
measurements at a single frequency (100 Hz) with an
impedance analyzer were performed. The 100 Hz frequency
was chosen to remain in the quasi-dc plateau (i.e., the
frequency-independent part at low frequencies of the imped-
ance spectra). While each measurement took 1.5 s, the time
interval between the measurements was 4 s.
Figure 10 shows the response and recovery time, τ50 (i.e.,
time needed to reach 50% of the maximum response upon
exposure to ethanol), at 500 °C. The material responds fast
upon exposure to 40 and 60 ppm of ethanol (τ50 < 15 s),
and the recovery time is slightly higher (τ50 ≈ 18 s). An
The test gas sequence was H2 (25 ppm), air, CO (50 ppm),
air, NO (5 ppm), air, NO2 (5 ppm), air, EtOH (40 ppm), air,
propylene (25 ppm), air. The intermediate measurements in
air were intended to verify the recovery of the resistance,
before the next test gas was applied. The relative humidity
(38) Le, M. T.; Van Well, W. J. M.; Stoltze, P.; Van Driessche, I.; Hoste,
S. Appl. Catal., A 2005, 282, 189.
(39) Morrison, S. R.; Hykaway, N. J.; Sears, W. M.; Frindt, R. F. US
5082789, 1992.
(40) Simon, U.; Sanders, D.; Jockel, J.; Brinz, T. J. Comb. Chem. 2005, 7,
682.
Inorganic Chemistry, Vol. 45, No. 22, 2006 9029

Roggan et al.
Figure 10. Response and recovery time of material A at T ) 500 °C.
The bars indicate the concentration steps of the admixed gases and the period
of the test gas pulse, while the curve above shows the response of the
resistance during the application of propylene and ethanol.
Figure 8. Arrhenius plot of the conductivity of material A.
Chart 2
of a single-phase compound. Additionally, XXII, a molecular
oxo complex2b with a Mo/Bi ratio of 2:1 and with Bi in the
oxidation state +V, was employed in the polyol method.
Different preparations were performed with variation of
hydrolyzing agent (H2O, H2O2, HNO3), reaction time (1-5
h), and temperature (150-170 °C). All samples were calcined
as described before. In all cases, monoclinic Bi2Mo3O12
accompanied by MoO3 was obtained. SEM pictures of
materials obtained from both complexes revealed the forma-
tion of needles with a length in the micrometer to millimeter
range. The outcome of this additional experiment thus shows
that the relative molar ratio of Bi/Mo also determines the
formation of different bismuth molybdate phases.
In the sensor-test measurements, a pronounced sensitivity
toward the applied gases was not found in either case
(materials derived from XXI or XXII); only a small
sensitivity toward ethanol was noted.
Hence, promising results could be obtained only starting
from II, but the polyol method so far seems to not be the
ideal synthetic strategy to form single-phase nanoparticles
from the complexes.
Figure 9. Temperature dependence of the sensitivity of material A.
increase of the resistance is observed; this is the typical
behavior of p-type semiconductors. Additionally, as men-
tioned before, almost no cross sensitivity toward propylene
could be found.
After these promising results, other Mo/Bi compounds
prepared previously were tested for comparison and to see
if they could serve as suitable precursors for the preparation
of bismuthmolybdate single phases. A necessary condition
for the application of the polyol method is a short-time (at
least) stability of the Mo/Bi heterodinuclear arrangements
against OH functions, and in this context the molybdenum/
bismuth alkoxide complex XXI (Chart 2) recommended itself
because it is stable, at least for short periods, in air. It,
therefore, had been investigated in the past with respect to
its behavior on silica surfaces.41
Conclusion
The use of XXI in the polyol method led to Bi2Mo2O9
being found as the core product after calcinations, and Bi2O3
was found as a byproduct. Variation of the reaction param-
eters did not alter the process of product formation in favor
The reaction of [Cp*(O)2Mo-O-Bi(o-tolyl)2]n, II, with
NBu4OH/MeOH intermediately leads to the species (o-
tolyl)2BiOH which polymerizes to give [(o-tolyl)2BiOH]n, 1
(NBu4[Cp*MoO3] is formed as the byproduct). Compound
1 represents a rare example of a bismuthhydroxide com-
pound, and it can be obtained in pure state from the reaction
(41) Limberg, C.; Hunger, M.; Habicht, W.; Kaifer, E. Inorg. Chem. 2002,
41, 3559.
9030 Inorganic Chemistry, Vol. 45, No. 22, 2006

[-O-BiIII‚‚‚OdMo-]n Chains
of [(o-tolyl)2Bi(hmpa)2]SO3CF3 with NBu4OH/H2O in thf/
water. Dissolved in dichloromethane, 1 immediately elimi-
nates water to give [(o-tolyl)2BiOBi(o-tolyl)2], 2, which can
be obtained on a preparative scale in this manner. If [(o-
tolyl)2Bi(hmpa)2]SO3CF3 is reacted with NBu4OH/MeOH in
dichloromethane, (o-tolyl)2BiOH undergoes methanolysis
before condensation leading to the isolation of [(o-tolyl)2-
BiOMe]n, 3. With molybdocene dihydrides, II reacts under
the formation of Mo-Bi bonds and Mo-OH units, so that
finally [RCp2Mo(H)(Bi(o-tolyl)2)], (R ) Me 4, R ) H 5)
and [Cp*2Mo2O5] are isolated as products. Compounds 4 and
5 can be obtained in pure state if 3 is chosen as the starting
material to be reacted with the corresponding molybdocene
dihydride, and it is noteworthy that 4 and 5 are the first
representatives of this class of compounds that are isolable
and thermodynamically stable. Unlike other molecular Mo/
Bi compounds, II can be employed as precursor in the polyol
method for the preparation of materials that exclusively
contain bismuthmolybdates in form of nanoparticles and
crystallites. These materials are biphasic and consist of Bi2-
Mo2O9 and Bi2MoO6, whose ratio can be determined by the
choice of the hydrolyzing reagent. One of these materials
proved to be capable of sensing EtOH selectively at elevated
temperatures.
neous phase that give access to novel compounds. Further-
more, they show promising properties as single-source
precursors for bismuthmolybdates. Future research will have
to reveal the most suitable method for their conversion to
functional materials.
Acknowledgment. We are grateful to the Deutsche
Forschungsgemeinschaft, the Fonds der Chemischen Indus-
trie, the BMBF, and the Dr. Otto Ro¨hm Geda¨chtnisstiftung
for financial support. We also would like to thank P.
Neubauer for crystal structure analyses and C. Jankowski
for the preparation of starting materials. We further appreciate
the contributions of D. Zornik to some of the experimental
procedures reported here. M.S. gratefully acknowledges
support by the Studienstiftung des Deutschen Volkes.
Supporting Information Available: X-ray crystallographic data
in CIF format. This material is available free of charge via the
Internet at http://pubs.acs.org. The crystallographic data (apart from
structure factors) of 1, 3, and 4 were deposited at the Cambridge
Crystallographic Data Centre as supplementary publications 607643,
607644, and 607645. Copies of the data can be ordered free of
charge upon application to CCDC, 12 Union Road, Cambridge CB2
1EZ (Fax (+44)1223-336-033; e-mail deposit@ccdc.cam.ac.uk).
In conclusion, molecular compounds containing Mo-O-
Bi moieties display interesting reactivitites in the homoge-
IC061198C
Inorganic Chemistry, Vol. 45, No. 22, 2006 9031