Surface-inspired molecular vanadium oxide catalysts for the oxidative dehydrogenation of alcohols: Evidence for metal cooperation and peroxide intermediates

Article abstract of DOI:10.1002/chem.201101442

On the basis that thiacalix[4]arene (H₄T4A) complex (PPh ₄)₂[H₂T4A(VO₂)]₂ (Ia) was found to be an adequate functional model for surface species occurring on vanadium oxide based catalysts and itself catalyses the oxidative dehydrogenation (ODH) of alcohols, an analogue containing 2,2′-thiobis(2, 4-di-tert-butylphenolate), ^SL^2-, as ligand, namely, (PPh₄)₂[^SLVO₂]₂ (II) was investigated in the same context. Despite the apparent similarity of Ia and II, studies on II revealed several novel insights, which are also valuable in connection with surfaces of vanadia catalysts: 1) For Ia and II similar turnover numbers (TONs) were found for the ODH of activated alcohols, which indicates that the additional OH units inherent to Ia do not contribute particularly to the activity of this complex, for instance, through prebinding of the alcohol. 2) On dissolution II enters into an equilibrium with a monomeric form, which is the predominant species in solution; nevertheless, ODH proceeds exclusively at the dimeric form, and this stresses the need for cooperation of two vanadium centres. 3) By omitting O₂ from the system during the oxidation of 9-fluorenol, the reduced form of the catalyst could be isolated and fully characterised (including single-crystal X-ray analysis). The corresponding intermediate had been elusive in case of thiacalixarene system Ia. 4) Reoxidation was found to proceed via a peroxide intermediate that also oxidises one alcohol equivalent. As the peroxide can also perform mono- and dioxygenation of the thioether group in II, after a number of turnovers the active catalyst contains a sulfone group. The reduced form of this ultimate catalyst was also isolated and structurally characterised. Possible implications of 1)-4) for the function of heterogeneous vanadia catalysts are discussed.

Full text of DOI:10.1002/chem.201101442

FULL PAPER
DOI: 10.1002/chem.201101442
Surface-Inspired Molecular Vanadium Oxide Catalysts for the Oxidative
Dehydrogenation of Alcohols: Evidence for Metal Cooperation and Peroxide
Intermediates
C. Gunnar Werncke, Christian Limberg,* Christina Knispel, and Stefan Mebs[a]
Abstract: On the basis that thiacalix[4]-
arene (H4T4A) complex (PPh4)2-
ACHTUNGTRENNUNG[H2T4AACHTUNGTRENNUNG(VO2)]2 (Ia) was found to be
an adequate functional model for sur-
face species occurring on vanadium
oxide based catalysts and itself cataly-
ses the oxidative dehydrogenation
(ODH) of alcohols, an analogue con-
taining 2,2’-thiobis(2,4-di-tert-butylphe-
nolate), SL2À, as ligand, namely,
tional OH units inherent to Ia do not
contribute particularly to the activity of
this complex, for instance, through pre-
binding of the alcohol. 2) On dissolu-
tion II enters into an equilibrium with
a monomeric form, which is the pre-
dominant species in solution; neverthe-
less, ODH proceeds exclusively at the
dimeric form, and this stresses the
need for cooperation of two vanadium
centres. 3) By omitting O2 from the
system during the oxidation of 9-fluor-
could be isolated and fully character-
ised (including single-crystal X-ray
analysis). The corresponding intermedi-
ate had been elusive in case of thiaca-
lixarene system Ia. 4) Reoxidation was
found to proceed via a peroxide inter-
mediate that also oxidises one alcohol
equivalent. As the peroxide can also
perform mono- and dioxygenation of
the thioether group in II, after
a
(PPh4)
N
number of turnovers the active catalyst
contains a sulfone group. The reduced
form of this ultimate catalyst was also
isolated and structurally characterised.
Possible implications of 1)–4) for the
function of heterogeneous vanadia cat-
alysts are discussed.
the same context. Despite the apparent
similarity of Ia and II, studies on II re-
vealed several novel insights, which are
also valuable in connection with surfa-
ces of vanadia catalysts: 1) For Ia and
II similar turnover numbers (TONs)
were found for the ODH of activated
alcohols, which indicates that the addi-
AHCTUNGTRENNeGUN nol, the reduced form of the catalyst
Keywords: dehydrogenation · ho-
mogeneous catalysis · peroxo li-
gands · reaction mechanisms · vana-
dium
Introduction
compounds with respect to the ODH of alcohols.[2,4] It
turned out that dinuclear compounds are more active than
mononuclear ones, and that thiacalixarene complexes are
more efficient than complexes containing the classical calix-
arene ligands.
Supported vanadium oxide catalysts are among the most ef-
ficient catalysts for the oxidative dehydrogenation (ODH)
of alcohols and alkanes,[1] and for a couple of years we have
been interested in their active sites on the molecular level.
Hence, we have pursued the synthesis of molecular com-
pounds which can simulate the structure, reactivity and/or
the spectroscopic properties of corresponding sites postulat-
ed to occur on the surfaces of such catalysts.[2-8] For simula-
tion of the oxidic environments that oxidovanadium moi-
Oxidovanadium(V) complex Ia (see Scheme 1) reached
the highest turnover numbers (TONs), and the results of
more detailed mechanistic investigations, inter alia isolation
of intermediate Ic, led to the proposal of a catalytic cycle as
exemplified in Scheme 1 for the substrate 9-fluorenol. While
it was not clear whether pre-binding of the alcohol is neces-
sary prior to the oxidation step, in any case starting from Ia
the next step was proposed to be a dehydrogenation (pre-
sumably in form of two consecutive H-atom abstractions)
leading to 9-fluorenone and an oxidovanadium(IV) complex
with two hydroxide bridges (Ib). To transform Ib into inter-
mediate Ic, which could be isolated, an excess of alcohol has
to replace one of the two hydroxide bridges by an alcohol-
ate ligand under elimination of water, and it is reasonable to
assume that this step proceeds sufficiently fast at 808C, es-
pecially in the presence of molecular sieves, which were
added to constantly remove the water formed in this pro-
cess. Starting from Ic, O2 is then capable of regenerating the
starting situation, as shown independently, and for this it has
to use all of its four oxidation equivalents.[4]
ACHTUNGTRENNUNGeties experience on the surfaces of oxidic support materials
we have successfully employed two different types of li-
gands: The deprotonated forms of calixarenes[2-5] and the de-
protonated forms of incompletely condensed silsesquiox-
anes.[6–8] Recently we reported the results of structure–func-
tion analyses of calixarene-based oxidovanadium model
[a] C. G. Werncke, Prof. Dr. C. Limberg, Dr. C. Knispel, Dr. S. Mebs
Institut fꢀr Anorganische Chemie
Humboldt-Universitꢁt zu Berlin
Brook-Taylor-Strasse 2, 12489 Berlin (Germany)
Fax : (+49)030-2093-6966
E-mail: christian.limberg@chemie.hu-berlin.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101442.
Chem. Eur. J. 2011, 17, 12129 – 12135
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12129

Scheme 1. Proposed catalytic cycle for the oxidation of alcohols (exemplified for 9-fluorenol) catalysed by thiacalix[4]arene-based compound Ia.
While this mechanism is plausible and was supported by
several observations, it still posed some questions, for in-
stance, concerning the role of the remaining OH groups in
the ligand system. Are they beneficial in terms of precoordi-
nation of alcohols via H-bonds? We addressed this and
other questions by preparing a derivative without these
groups: by employing the 2,2’-thiobis(2,4-di-tert-butylpheno-
Results and Discussion
Influence of dangling OH groups in Ia: To determine the
contribution of the dangling OH groups in Ia to its high ac-
tivity, for instance, through pre-binding of alcohol substrates,
it was of interest to test II, which lacks such OH groups, as
a catalyst for the ODH of alcohols under the same condi-
tions as Ia (Table 1).
S
late) ligand L2À (Scheme 2a), which can be regarded as the
relevant section of the thiacalixarene framework needed for
construction of Ia, we managed to prepare analogous com-
plex II (Scheme 2b),[9] which, however, was found to be
mainly monomeric in solution.
While the trend is somewhat non-uniform, it is clear that
reduction of the thiacalixarene ligand system to SL2À does
not lead to a loss in activity that would point to the OH
groups being responsible for the superiority of Ia over other
calixarene-based catalysts. This
holds all the more given that
only a fraction of II employed,
namely, the equilibrium concen-
tration of IID, can be effective
during the catalytic procedure
(see below), so that it has to be
concluded, that IID is even su-
perior to Ia. Hence, we can
now be sure that the free OH
groups of the thiacalix[4]arene
Scheme 2.
ligand in Ia do not serve any
particular purpose during catal-
ysis.
Recently, we reported on bioinspired investigations con-
cerning haloperoxidase-like behaviour of II and its catalytic
activity in sulfoxidation with tBuOOH.[9] We now report its
properties as an ODH catalyst and detailed insight into its
function, which at the same time improve and extend the
mechanistic view on systems such as Ia.
Table 1. TONs for the ODH of selected alcohols to the respective alde-
hydes/ketones with O2, catalysed by 1 mol% of Ia or II, after 60 min.
Catalyst
Benzyl
alcohol
9-Fluorenol
Cinnamic
alcohol
Ia
II
34
50
83
83
45
25
12130
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12129 – 12135

Oxidative Dehydrogenation of Alcohols
FULL PAPER
Dinuclearity as a prerequisite: Since II crystallises as a
dimer (IID) but after dissolution mainly exists in form of a
monomeric species (IIM, Keq =20 molLÀ1 for the equilibrium
shown in Scheme 2b), the question arose whether the mech-
anism shown in Scheme 1 even holds for II. However, all
subsequent investigations as outlined below confirmed that
the ODH reactivity is exhibited exclusively by the dimer IID,
which is a further important finding, also with regard to the
model character of these compounds and inferences for sur-
face chemistry.
Scheme 1, the V centres in IID must switch their oxidation
states only between +V and +IV, while function of IIM
would require a change between +V and +III.
The rate-determining step: To obtain information on the
rate-determining step of the catalytic oxidation, 9-deutero-9-
fluorenol was employed to compare its reaction rate with
that of undeuterated 9-fluorenol (Figure 2). This gave a ki-
Provided that a rate law for a pseudo-first-order reaction
applies, a linear dependence between the concentration of
the catalytically active species and the reaction rate should
be observed. Figure 1 shows corresponding plots for various
Figure 2. Reaction rates for the oxidation of 9-fluorenol and 9-deutero-9-
fluorenol with O2 catalysed by 1 mol% II.
netic isotope effect (KIE) of 2.8. Considering, that the maxi-
mum KIE reachable at 808C is about 5, a value of 2.8 is cer-
tainly indicative of a rate-determining step with significant
Figure 1. Observed reaction rate in dependence on the amount of used
catalyst for ODH of 9-fluorenol to 9-fluorenone with O2 catalysed by var-
ious amounts of II.
À
involvement of the C H bond and a quite symmetric transi-
tion state along the H-atom transfer coordinate O···H···C.
This finding is consistent with recent investigations concern-
ing the H-atom abstraction reactivity of ligated [VO2]+
units.[8,10]
weigh-ins of catalyst II. Interestingly, doubling the amount
of added catalyst from 0.125 to 0.25 mol% or from 0.25 to
0.5 mol% leads to an approximately fourfold (4.1 and 3.6,
respectively) increase in reaction rate, which can only be ex-
plained if the ODH proceeds not through the monomeric
but through the dimeric form: On the basis of the equilibri-
Isolation of the reduced catalyst: As in the case of Ia, we
also tried to identify reaction intermediates of the catalytic
system involving II. Interestingly, omitting O2 during conver-
sion of 9-fluorenol led to isolation of (PPh4)2ACTHNUTRGENNUG ACHTUNGTRENNUNG(m-
[SLVO
um equation Keq =[IIM]2/
N
OH)2VOSL] (1), which is an analogue of Ib. Compound Ib
itself had been elusive, supposedly as it quickly reacts with
excess alcohol to give 1c. Its existence was only inferred,[4]
but isolation of 1 now confirms that its proposed intermedia-
cy was reasonable. Furthermore, it provides additional sup-
port that the ODH indeed proceeds via dimer IID.
Scheme 2b, at these low concentrations a doubling of the
overall amount of catalyst added leads to a fourfold increase
of [IID], while [IIM] is only doubled. When the weigh-in is
doubled at higher catalyst concentrations, the multiplying
factor for IID should be somewhat lower than four, as the
equilibrium gradually shifts more to the side of the dimer,
and this is observed here, too. However, the decrease of the
observed factor is more pronounced than should be theoret-
ically (2.45 from 1 to 2 mol%), which we cannot rationalise
at present.
The structure of the dianion of 1 (Figure 3) is nearly iden-
tical with that of IID, from which it can be derived formally
by replacing the m-O bridges by m-OH groups. The protons
could be located in the X-ray analysis, and the appearance
À1
~
of a corresponding OH stretching band at n=3678 cm in
Nevertheless, the behaviour of the reaction rates with var-
iation of the overall concentration at low catalyst weigh-ins
clearly indicates that IID mediates alcohol oxidation, while
IIM itself is inactive, which nicely illustrates the advantage of
the cooperation of two vanadium ions: Referring to
the IR spectrum of 1 further proved their existence. The re-
placement of the bridging divalent oxido ligands by mono-
À
valent OH ligands leads to lengthening of corresponding V
O
distances from approximately 1.84 ꢃ in IID to
À
À
1.9870(11) ꢃ (O1 V1) and 2.0103(11) ꢃ (O1’ V1) in 1. Fur-
Chem. Eur. J. 2011, 17, 12129 – 12135
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12131

C. Limberg et al.
Figure 3. Structure of 1·2MeCN in the crystal. All hydrogen atoms
except H1 and H1’, the Ph4P+ cations and co-crystallised acetonitrile sol-
vent molecules are omitted. Selected bond lengths [ꢃ] and angles [8]:
À
À
À
À
O1 H1 0.8489, O1 V1 1.9870(11), O1’ V1 2.0103(11), O2 V1
À
À
À
À
1.6050(12), O3 V1 1.9669(11), O4 V1 1.9842(11), S1 V1 2.8229(5), V1
À
V1’ 3.1705(4), V1 (O1, O1’, O3, O4) 0.3991(3), O1-V1-O1’ 75.04(5), O1-
V1-O2 104.22(6), O1’-V1-O2 102.96(5), O2-V1-O3 100.56(6), O2-V1-O4
99.19(6), O2-V1-S 170.01(5), V1-O1-V1’ 104.96(5).
thermore, the distance between the two vanadium atoms,
which are now in the oxidation state +IV, is lengthened
from 2.7782(7) ꢃ in IID to 3.1705(4) ꢃ. The widening of the
V2O2 core is in agreement with the observations made for
intermediate Ic containing thiacalixarene ligands. In solution
1 exhibits a magnetic moment of 2.27 mB at room tempera-
ture, which is indicative of weak antiferromagnetic coupling
of the two VIV ions. EPR studies on 1 (in frozen THF,
MeCN or CH2Cl2 solution) at 77 or 293 K (THF or CH2Cl2
solutions) revealed, if at all, only a very weak and broad
signal (see Supporting Information for selected EPR spectra
of 1), which excludes the presence of a monomeric form of
1 and again suggests a magnetic communication between
the two VIV ions. We were not able to isolate a derivative of
1 containing an alkoxide bridge as in Ic even when using a
fivefold excess of the substrate.
Scheme 3. Products resulting from the oxygenation of II. Compounds IV
and VI are involved in monomer/dimer equilibrium (a), whereas perox-
ides III, V and VII exist only in monomeric form (b).[9]
to the sulfoxide group to yield VIM, which was identified
previously as the final product of H2O2 oxidation of II[9] and
here represents the monomer of the ultimate catalyst.
Consistently, the reaction of VI with 9-fluorenol led to
isolation
of
VIV
intermediate
(PPh4)2ACHTUNGTERNNNU[G ACHTUNGTRENNUNG(m-
SO2LVO
OH)2VOSO2L] (2). The structure of the dianion of
2
(Figure 4) can be derived from that of 1 by replacement of
the thioether groups by sulfone units, which interact with
the vanadium centres through one of their oxygen atoms.
The protons of the bridging OH unit could be located in the
The reoxidation step: We also investigated the reoxidation
step starting from 1. Treatment of 1 with O2 at room temper-
ature led to an immediate colour change from green to
violet. Monitoring this reaction by 51V NMR spectroscopy
revealed, in addition to a signal at d=À502 ppm, which cor-
responds to the starting compound IIM, a further signal at
d=À448 ppm. The latter belonged to peroxide III
(Scheme 3b), which was identified previously as a product
of the reaction between II and H2O2.[9] On storing of the
sample at room temperature the peroxide signal decreased
in intensity in favour of a new one at d=À512 ppm, charac-
teristic of the monomer of singly S-oxidised compound IVM
(Scheme 3a), generated from III in the course of an intra-
molecular transfer of a peroxidic oxygen atom to the thio-
ether linker. Like IIM, also IVM enters into an equilibrium
with a dimeric form, IVD,[9] and it is reasonable to assume
that IVD will oxidise a further alcohol equivalent; reoxida-
tion can then be anticipated to lead to IVM and V. Finally,
the peroxide group of V will transfer another oxygen atom
Figure 4. Structure of 2·8CH2Cl2 in the crystal. All hydrogen atoms
except H2 and H2’, the Ph4P+ cations and co-crystallised CH2Cl2 solvent
À
molecules are omitted. Selected bond lengths [ꢃ] and angles [8]: O2 H2
À
À
À
À
0.74(6), O1 V1 1.596(2), O2 V1 2.004(2), O2’ V1 2.000(2), O3 V1
À
À
À
À
1.957(2), O4 V1 1.980(2), O5 V1 2.410(2), V1 V1’ 3.1378(7), V (O2,
O2’, O3, O4) 0.3798(6), O1-V1-O2 102.65(11), O1-V1-O2’ 103.78(11),
O1-V1-O3 100.43(11), O1-V1-O4 97.76(11), O1-V1-O5 176.65(10), V1-
O2-V2 103.20(11).
12132
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12129 – 12135

Oxidative Dehydrogenation of Alcohols
FULL PAPER
X-ray analysis, and the appearance of a corresponding OH
stretching band at 3664 cmÀ1 in the IR spectrum of 2 further
proved their existence. As in the case of 1, treatment of 2
with O2 leads to a colour change from green to violet, and a
51V NMR spectrum recorded directly afterwards indicated
the presence of equimolar amounts of starting compound
VIM and known peroxide VIIM.[9] Independent reactivity
studies with peroxide VIIM itself showed that it does not oxi-
dise intermediate 2, but it is capable of oxidising the alcohol,
although this reaction proceeds more slowly than oxidation
of the alcohol by VI. Hence, this is the route through which
the peroxide, formed in the course of reoxidation of the re-
duced catalyst, re-enters the catalytic cycle after the thio-
with all of the S atoms being doubly oxidised. This leads to
VI as the actual catalytically active species, as proved by
NMR spectroscopy.
Considering these findings the question arises how they
relate to the thiacalixarene system and the proposed catalyt-
ic cycle shown in Scheme 1? First, they indicate, that cataly-
sis with Ia need not necessarily proceed via an alcoholate in-
termediate Ic. In the case of II, the reduced bis(m-hydroxi-
do) intermediate is readily reoxidized by O2, and now the
same seems possible also for I, although there was no evi-
dence for a peroxide intermediate in this system. Perhaps
catalysis with Ia proceeds by two simultaneous cycles, one
as shown in Scheme 1 and one according to Scheme 4. A
further possibility is that I exclusively utilises the Scheme 4
cycle, too, and that 1c was only isolated under non-turnover
conditions as it is thermodynamically more stable than Ib.
ACHTUNGTRENNUNGether group has been fully oxidised: it oxidises a further al-
cohol equivalent. In the course of catalytic alcohol oxida-
tions setting out with VI no new signals belonging to any
other V species could be observed, which supports that VI
(and VII) are the sole species responsible for substrate oxi-
dation in the later stage of the catalytic process.
Implications for surface chemistry: By investigating system
II as an ODH catalyst two important findings could be
made: 1) catalysis proceeds (in the temperature regime that
applies here) exclusively via dinuclear species, and 2) re-oxi-
dation proceeds via the participation of peroxides. One hy-
pothesis which 1) thus poses for heterogeneous catalysts is
that, also on their surfaces, dinuclear species are more reac-
tive than mononuclear ones, and this nicely complements
from the molecular side corresponding proposals based on
experimental and theoretical results concerning heteroge-
Catalytic cycle: Based on all these results, it is reasonable to
propose the mechanism exemplified for the substrate 9-
fluorACHTUNGTRENNUNG
enol in Scheme 4: Starting from IID ODH of one alco-
hol equivalent occurs in the form of two H-atom abstrac-
tions, whereas the rate-determining step consists of an H-
atom transfer from the HOCH unit to the oxidovanadium
core. This produces 1 and one equivalent of the carbonyl
compound. Reoxidation of 1 with O2 leads to one equivalent
of the monomer of the starting material, IIM, and one equiv-
alent of peroxide III. The peroxide then oxidises a further
alcohol equivalent to close the cycle, as shown in Scheme 4.
This cycle, however, is somewhat simplified: As the perox-
ide III can also oxidise the thioether group of the thio-
AHCTUNGTRENNUNG
neous catalysts.[11] Considering 2), it is noteworthy that only
recently after detailed spectroscopic investigations and com-
parison with peroxido model compounds was it concluded
that “an umbrella-type vanadium oxido peroxido structure
does not exist for hydrated or dehydrated V2O5/SiO2 cata-
lysts”,[12] although DFT results suggested that reoxidation
takes place via peroxide species.[13] In this context it is im-
portant to note that the failure to detect peroxides does not
mean that they are not formed
ACHTUNGTRENNUNGbisphenolate ligand, after approximately 6–7 turnovers the
catalytic system works exactly as shown in Scheme 4 but
intermediately; they may just
not be stable under the condi-
tions of analysis or do not accu-
mulate due to their high reac-
tivity. Dissolved in MeCN, per-
oxide VII, for instance, only has
a half-life of 12 h at room tem-
perature and merely 5 min at
808C.[9] With this background
our results support the peroxide
route for reoxidation.
Conclusion
Investigation of the thiobisphe-
nolato oxidovanadium complex
(PPh4)2ACHTUNGTRNEUNG
[(SLVO2)2] (II) as a cata-
lyst for the ODH of alcohols re-
vealed valuable insights into ox-
idation catalysis at vanadium
Scheme 4. Catalytic cycle for the oxidation of alcohols (exemplified for 9-fluorenol) with II as catalyst.
Chem. Eur. J. 2011, 17, 12129 – 12135
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12133

C. Limberg et al.
partially degassed, flushed with 1.2 bar of O2 and heated to 808C. Con-
versions were determined by 1H NMR spectroscopic analysis of samples
taken at the depicted times under a dioxygen stream. Blind studies in the
absence of the catalyst showed no alcohol conversion.
centres. First, it was shown that dangling OH groups present
in a related thiacalix[4]arene-based catalyst do not contrib-
ute particularly to its activity, and the proposal of an “OVIV-
ACHTUNGTRENNUNG
(m-OH)2VIVO” intermediate, which was elusive in the case
Crystal structure determinations: Data for 1·2MeCN and 2·8CH2Cl2
were collected at 100 K on a Stoe IPDS 2T diffractometer with MoKa ra-
diation (l=0.71073 ꢃ) from a sealed-tube generator with graphite mono-
chromator. Numerical absorption correction[15] were applied for
1·2MeCN and 2·8CH2Cl2. The structures were solved by direct methods
and refined by full-matrix least-squares procedures based on F2 with all
measured reflections.[16] All non-hydrogen atoms were refined anisotropi-
cally. H atoms were introduced in their idealised positions and refined by
of that system, was supported: An analogue (1) has now
successfully been isolated by employing the simplified thio-
bisphenolate ligand SL2À. Reoxidation of 1 was found to pro-
ceed through a peroxide intermediate and, importantly, it
was shown that ODH proceeds via a dinuclear species,
while the corresponding monomer is inactive. The rate-de-
termining step is an H-atom abstraction from the CH unit
connected to the OH group. Altogether these results al-
lowed a plausible reaction mechanism to be proposed. They
suggest that also on surfaces dinuclear species are more re-
active than mononuclear ones and that the reoxidation pro-
ceeds via a peroxide route.
using
a riding model. CCDC-817517 (1·2MeCN) and CCDC-817518
(2·8CH2Cl2) contain the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Crystal data for 1·2MeCN: C108H128N2O8P2S2V2, Mr =1810.06, T=
100(2) K, l=0.71073 ꢃ, monoclinic, space group P21/n, a=12.5041(4),
b=31.2121(6), c=13.9180(4) ꢃ, b=114.478(2)8, V=4943.7(2) ꢃ3, Z=2,
1calcd =1.216 mgmÀ3, m=0.320 mmÀ1; 42882 reflections measured, 9715
unique (Rint =0.0380); final R indices [I>2s(I)]: R=0.0379, wR=0.0924.
Crystal data for 2·8CH2Cl2: C112H138Cl12O12P2S2V2, Mr =2471.49, T=
ꢀ
Experimental Section
100(2) K, l=0.71073 ꢃ, triclinic, space group P1, a=13.1043(4), b=
14.8568(5), c=17.0432(6) ꢃ, a=75.750(3)8, b=83.516(3)8, g=70.633(3)8,
V=3032.25(19) ꢃ3, Z=1, 1calcd =1.352 mgmÀ3, m=0.623 mmÀ1; 31167 re-
flections measured, 11841 unique (Rint =0.0423); final R indices [I>
2s(I)]: R=0.0544, wR=0.1390.
General remarks: All manipulations were carried out in a glove box or
by Schlenk-type techniques under a dry argon atmosphere. The 1H and
51V NMR spectra were recorded on a Bruker AV-400 NMR or DPX-300
spectrometer. The 1H NMR spectra were calibrated against the residual
protons of the deuterated solvent, and the 51V NMR spectra against
VOCl3 as external standard. Microanalyses were performed on a Leco
CHNS-932 or HEKAtech Euro EA 3000 elemental analyser. Infrared
spectra were recorded on samples prepared as KBr pellets with a Shi-
madzu FTIR 8400S spectrometer. The high-resolution mass spectra were
recorded with an Agilent Technologies 1200 Series mass spectrometer
with a time-of-flight analyser. EPR spectra were measured on an ERS
300 (ZWG/Magnettech GmbH, Berlin Adlershof, Germany). Solution
magnetic susceptibilities were determined by the Evans method[14] with a
Bruker AV-400 NMR spectrometer. Compounds II, VI and VII were ob-
tained by literature procedure.[9]
Acknowledgements
We are grateful to the Deutsche Forschungsgemeinschaft, the CRC546,
the Fonds der Chemischen Industrie and the BMBF for financial support.
We thank C. Ohde for the EPR measurements.
[1] a) A. Comite, A. Sorrentino, G. Capannelli, M. Di Serio, R. Tesser,
E. Santacesaria, J. Mol. Catal. A: Chem. 2003, 198, 151–165; b) R.
Zhou, Y. Cao, S. Yan, J. Deng, Y. Liao, B. Hong, Catal. Lett. 2001,
75, 107–112; c) T. Blasco, A. Galli, J. Lopez Nieto, F. Trifido, J.
Catal. 1997, 169, 203–211; d) M. A. Chaar, D. Patel, H. H. Kung, J.
Catal. 1988, 109, 463–467; e) A. Corma, J. M. Lopez Nieto, N. Par-
edes, J. Catal. 1993, 144, 425–438; f) F. D. Hardcastle, I. E. Wachs, J.
Mol. Catal. 1988, 46, 173–186; g) G. Deo, I. E. Wachs, J. Haber,
Crit. Rev. Surf. Chem. 1994, 4, 141–187; h) I. E. Wachs, B. M. Weck-
huysen, Appl. Catal. A 1997, 157, 67–90; i) J. G. Eon, R. Olier, J. C.
Volta, J. Catal. 1994, 145, 318–326; j) T. Balsko, J. M. Lopez Nieto,
Appl. Catal. A 1997, 157, 117–142; k) A. Khodakov, B. Olthof, A. T.
Bell, E. Iglesia, J. Catal. 1999, 181, 205–216; l) A. Khodakov, J.
Yang, S. Su, E. Iglesia, A. T. Bell, J. Catal. 1998, 177, 343–351;
m) E. V. Kondratenko, M. Baerns, Appl. Catal. A 2001, 222, 133–
143; n) M. D. Argyle, K. Chen, E. Iglesia, A. T. Bell, J. Catal. 2002,
208, 139–149; o) L. Owens, H. H. Kung, J. Catal. 1993, 144, 202–
213; p) O. R. Evans, A. T. Bell, T. D. Tilley, J. Catal. 2004, 226, 292–
300.
Synthesis of 1: Compound II·2MeCN (90 mg, 50 mmol, 1 equiv), 9-fluor-
ACHTUNGTRENNUNGenol (9.1 mg, 50 mmol, 1 equiv) and 3 ꢃ molecular sieves (500 mg) were
suspended in 10 mL of acetonitrile and stirred at 808C. After 1.5 h the
solvent was removed thoroughly, the residue was extracted with 5 mL of
CH2Cl2 and layered with 10 mL of hexane. Removal of the solvent by fil-
tration after 3 d and drying of the remaining crystalline solid afforded
~
40 mg of 1·2CH2Cl2 (21 mmol, 42%). IR (KBr): n=3678 (w), 3062 (w),
2948 (s), 2900 (m), 2865 (m), 1586 (w), 1467 (s), 1436 (vs), 1381 (m), 1357
(w), 1312 (s), 1299 (w), 1256 (m), 1200 (w), 1108 (vs), 995 (w), 959 (s),
838 (m), 758 (m), 723 (s), 688 (s), 527 cmÀ1 (vs); elemental analysis calcd
(%) for C104H122O12P2S2V2·1CH2Cl2 (1897.9): C 67.08, H 6.69, S 3.37;
found: C 67.54, H 7.05, S 2.99. Crystals that were suitable for X-ray dif-
fraction analysis were obtained by slowly evaporating a saturated solu-
tion of 1 in acetonitrile.
Synthesis of 2: Compound VI (90 mg, 50 mmol, 1 equiv), 9-fluorenol
(9.1 mg, 50 mmol, 1 equiv) and 3 ꢃ molecular sieves (500 mg) were sus-
pended in 5 mL of CH2Cl2. After 1 h of stirring the solution was filtered
and layered with 5 mL of hexane. Removal of the solvent by filtration
after 3 d and drying of the remaining crystalline solid afforded 58 mg of
[2] E. Hoppe, C. Limberg, B. Ziemer, Inorg. Chem. 2006, 45, 8308–
8317.
~
2·CH2Cl2 (31 mmol, 62%). IR (KBr): n=3681 (w), 3061 (w), 2950 (s),
2902 (m), 2864 (m), 1598 (w), 1530 (w), 1480 (vs), 1441 (s), 1436 (s), 1381
(w), 1358 (m), 1336 (s), 1273 (s), 1256 (m), 1193 (w), 1139 (w), 1122(m),
1108 (vs), 1085 (m), 995 (w), 972 (s), 868 (w), 841 (w), 786 (m), 777(m),
750 (w), 723 (s), 689 (s), 619 (vs), 526 (vs), 486 cmÀ1 (w); elemental analy-
sis calcd (%) for C104H122O8P2S2V2·CH2Cl2 (1877.0): C 67.18, H 6.66, S
3.41, Cl 3.77; found: C 66.91, H 6.71, S 3.25, Cl 3.24.
[3] E. Hoppe, C. Limberg, B. Ziemer, C. Mꢀgge, J. Mol. Catal. A:
Chem. 2006, 251, 34–40.
[4] E. Hoppe, C. Limberg, Chem. Eur. J. 2007, 13, 7006–7016.
[5] C. Limberg, Eur. J. Inorg. Chem. 2007, 3303–3314.
[6] C. Ohde, M. Brandt, C. Limberg, J. Dçbler, B. Ziemer, J. Sauer,
Dalton Trans. 2008, 326–331.
General procedure for oxidation studies: Alcohol, catalyst (mol% with
respect to V centres) and 1.5 g of 3 ꢃ molecular sieves were suspended
in 6 mL of acetonitrile under an argon atmosphere. Then the vessel was
[7] C. Ohde, C. Limberg, R. Stçsser, S. Demeshko, Inorg. Chem. 2010,
49, 2479–2485.
[8] C. Ohde, C. Limberg, Chem. Eur. J. 2010, 16, 6892–6899.
12134
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12129 – 12135

Oxidative Dehydrogenation of Alcohols
FULL PAPER
[9] C. G. Werncke, C. Limberg, C. Knispel, R. Metzinger, B. Braun,
Chem. Eur. J. 2011, 17, 2931–2938.
[13] X. Ding, W. Xue, Y. Ma, Y. Zhao, X. Wu, S. He, J. Phys. Chem. C
2010, 114, 3161–3169.
[10] a) C. R. Waidmann, X. Zhou, E. A. Tsai, W. Kaminsky, D. A.
Hrovat, W. T. Borden, J. M. Mayer, J. Am. Chem. Soc. 2009, 131,
4729–4743; b) M. H. V. Huynh, T. J. Meyer, Chem. Rev. 2007, 107,
5004–5064; c) J. M. Mayer, Annu. Rev. Phys. Chem. 2004, 55, 363–
390.
[14] a) D. F. Evans, J. Chem. Soc. 1959, 2003–2005; b) E. M. Schubert, J.
Chem. Ed. 1992, 69, 62; c) T. Ayers, R. Turk, C. Lane, J. Goins, D.
Jameson, S. J. Slattery, Inorg. Chim. Acta 2004, 357, 202–206.
[15] W. R. Busing, H. A. Levy, Acta Crystallogr. Sect. B 1967, 22, 457–
464.
[11] H. Feng, J. W. Elam, J. A. Libera, M. J. Pellin, P. C. Stair, J. Catal.
2010, 269, 421–431.
[16] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112–122.
[12] J. E. Molinari, I. E. Wachs, J. Am. Chem. Soc. 2010, 132, 12559–
12561.
Received: May 11, 2011
Published online: September 13, 2011
Chem. Eur. J. 2011, 17, 12129 – 12135
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12135