Mechanistic insights into the rhenium-catalyzed alcohol-to-olefin dehydration reaction

Article abstract of DOI:10.1002/chem.201300209

Rhenium-based complexes are powerful catalysts for the dehydration of various alcohols to the corresponding olefins. Here, we report on both experimental and theoretical (DFT) studies into the mechanism of the rhenium-catalyzed dehydration of alcohols to olefins in general, and the methyltrioxorhenium-catalyzed dehydration of 1-phenylethanol to styrene in particular. The experimental and theoretical studies are in good agreement, both showing the involvement of several proton transfers, and of a carbenium ion intermediate in the catalytic cycle. Ionic or concerted? Rhenium-based complexes are powerful catalysts for the dehydration of various alcohols to the corresponding olefins. Experimental and DFT studies into the mechanism of the rhenium-catalyzed dehydration of alcohols to olefins are reported. The experimental and theoretical studies are in good agreement, both showing the involvement of a carbenium ion intermediate in the catalytic cycle (see figure). Copyright

Full text of DOI:10.1002/chem.201300209

DOI: 10.1002/chem.201300209
Mechanistic Insights into the Rhenium-Catalyzed Alcohol-To-Olefin
Dehydration Reaction
[a]
Ties J. Korstanje, Johann T. B. H. Jastrzebski, and Robertus J. M. Klein Gebbink*
Abstract: Rhenium-based complexes are powerful catalysts for the dehydration of
various alcohols to the corresponding olefins. Here, we report on both experimen-
Keywords: alcohols · dehydration
reaction · density functional calcu-
tal and theoretical (DFT) studies into the mechanism of the rhenium-catalyzed de-
hydration of alcohols to olefins in general, and the methyltrioxorhenium-catalyzed
dehydration of 1-phenylethanol to styrene in particular. The experimental and the-
oretical studies are in good agreement, both showing the involvement of several
proton transfers, and of a carbenium ion intermediate in the catalytic cycle.
lations
· reaction mechanisms ·
rhenium
Introduction
involves protonation of the hydroxyl group of the alcohol,
followed by removal of water and deprotonation at the b-
[18]
Rhenium-catalyzed oxygen atom transfer (OAT) reactions,
especially deoxygenation reactions, have received growing
attention in recent years due to their potential application in
the processing of biomass-derived substrates to chemical
building blocks. The potential of rhenium complexes as
catalysts for deoxygenation reactions was first recognized in
carbon.
Although for simple Brønsted acids, such as sulfuric or
phosphoric acid, this E1 mechanism seems generally accept-
ed and taught even in advanced organic chemistry courses,
there are indications that the mechanism of operation is
more complicated. One of the issues involved with a pure
E1 mechanism is the deprotonation of the carbenium ion,
which in acidic media is very slow. Several kinetic investiga-
tions have indicated that the formation of ethers through an
[1]
the 1990s, by the discovery that rheniumACTHUNGTRENNUNG
[2–7]
expel alkenes upon heating.
were the first to convert this stoichiometric reaction to a cat-
alytic deoxydehydration reaction by using Cp*ReO (Cp*=
Ad 2 mechanism and subsequent ether-splitting could pro-
3
E
[19,20]
C
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(CH ) anion) and PPh as a reductant to convert phenyl-
vide an alternative pathway.
For the hydrochloric acid-
5
3
5
3
[8]
1,2-ethanediol to styrene in good yield. In the same year
catalyzed gas phase dehydration, a concerted six-membered
transition-state has been proposed on the basis of DFT cal-
culations, with the HCl proton linked to the OH group and
the chloride linked to the hydrogen atom at the b-posi-
the use of methyltrioxorhenium (CH ReO , MTO) as a cata-
3
3
lyst for the dehydration reaction of various alcohols to ole-
[9]
fins was reported by Zhu and Espenson. Very recently
[
10–15]
[21]
many extensions to both the deoxydehydration reaction
and the alcohol-to-olefin dehydration reaction
tion. In the case of alcohol dehydration over g-alumina,
[16,17]
catalyzed
the mechanism seems different from that of simple Brønsted
acids, as the involvement of free carbenium ions, as in the
E1 mechanism, can be ruled out on the basis of kinetic iso-
by various rhenium complexes have been reported. Quite a
number of mechanistic investigations have been performed
on the rhenium-catalyzed deoxydehydration reaction, yet
for the rhenium-catalyzed dehydration, only a few mecha-
nistic studies have been reported.
tope effect (KIE) studies. Both an S 2-type mechanism is
N
proposed for ether formation and an E2-type mechanism for
[22–24]
dehydration towards the olefin.
Interestingly, for
For the various Brønsted acid-catalyzed alcohol dehydra-
tion reactions, many more mechanistic studies have been re-
ported. The classical E1 mechanism taught in organic chem-
istry for the Brønsted acid-catalyzed dehydration reaction
Keggin-type polyoxometallates based on tungsten, an E1
mechanism was found, whereas no indications for an E2-
type elimination were observed either by experiment or in
[25]
theoretical calculations.
For the various rhenium-catalyzed reactions involving al-
[
9,26]
[27]
cohols, including dehydration,
etherification,
dehydrative coupling,
[
a] Dr. T. J. Korstanje, Dr. J. T. B. H. Jastrzebski,
Prof. Dr. R. J. M. Klein Gebbink
[9,28]
[29]
and 1,3-transposition, different reaction
Organic Chemistry & Catalysis
mechanisms have been proposed in the literature. In most
proposed mechanisms, the first step involves the reaction of
Debye Institute for Nanomaterials Science
Faculty of Science, Utrecht University
Universiteitsweg 99, 3584 CG Utrecht (The Netherlands)
Fax : (+31)30-252-3615
MTO with the alcohol to form a [ReO (Me)(OH)(OR)] in-
2
termediate. Rhenium ACHTUNGTRENNGNU( VII) alkoxide complexes are known
and some have been isolated, such as [ReO
A
H
U
G
R
N
N
(OMe)] and
E-mail: r.j.m.kleingebbink@uu.nl
3
[30,31]
[
ReO
ACHTUNGERTUNNNG( OtBu)].
3
The latter was found to be unstable
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201300209.
above À208C, whereas the former is stable at room temper-
13224
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 13224 – 13234

FULL PAPER
ature, but decomposes at elevated temperatures. The next
step in the proposed mechanisms can be divided into two
different groups: a concerted pathway, involving all transfor-
mations on the rhenium center, resembling an E2 mecha-
Results and Discussion
Experimental results: Previously, we reported on the Re O -
2
7
catalyzed dehydration of benzylic alcohols to the corre-
[9,29]
[16]
nism,
and an ionic pathway, involving a heterolytic cleav-
sponding styrene moieties. There we found a strong influ-
age of the alcohol CÀO bond to form a carbenium ion and a
ence of the electronic parameters of para substitution on the
À
[
ReO (Me)(OH)] anion, resembling an E1 mechanism
phenyl ring of 1-phenylethanol with both Re O or sulfuric
3
2
7
[9,28]
(
Scheme 1).
acid (H SO ) as catalyst: electron-donating substituents
2 4
result in an increase in reaction rate, whereas electron-with-
drawing substituents halt the reaction. Interestingly, p-nitro-
phenylethanol, the most electron-poor alcohol tested, does
react when using H SO as catalyst, whereas with Re O as
2
4
2
7
catalyst no reaction is observed. p-Cyanophenylethanol,
being less electron-poor than p-nitrophenylethanol, does not
react with either catalyst. The very strong electronic effect
observed indicates the build-up of positive charge during
the rate-determining step, possibly through the formation of
a carbenium intermediate.
When monitoring the alcohol-to-olefin dehydration reac-
tion in time, a typical reaction profile as depicted in Fig-
ure 1a is observed. At the start of the reaction 1-phenyle-
thanol is being consumed, whereas styrene forms simultane-
ously with diphenethyl ether. When the starting material is
nearly exhausted, the ether concentration reaches a maxi-
mum, after which its concentration gradually decreases,
whereas the styrene concentration further increases to near-
quantitative yield after 24 h. When starting from dipheneth-
yl ether, formation of styrene and minor amounts of 1-phe-
nylethanol occur, although the reaction proceeds slowly
compared with the alcohol-to-olefin reaction (initial rate of
Scheme 1. Proposed concerted (top half) and ionic (bottom half) alcohol-
to-olefin dehydration reaction mechanisms catalyzed by MTO.
À1
As various pathways can be active (primarily dependent
on the catalyst applied), we considered both types of mecha-
nisms in our investigations. Here, we report on both our ex-
perimental and theoretical re-
0.5 mmolh for both Re O and H SO as the catalyst, com-
2
7
2
4
À1
pared with 4.3 and 2.6 mmolh for Re O and H SO , re-
2
7
2
4
spectively, for the dehydration of 1-phenylethanol). This in-
search efforts to elucidate the
mechanism of the rhenium-cat-
alyzed dehydration reaction.
We have investigated various
aspects of the mechanism ex-
perimentally by determining
the influence of electronic pa-
rameters of substituents and
the influence of the presence
of nitrogen bases, water, and
oxygen on the rhenium-cata-
lyzed alcohol-to-olefin dehy-
dration reaction. Furthermore,
indications for the rate-deter-
mining step were obtained by
kinetic isotope effect studies,
reaction order determination,
and substrate reactivity trends.
Next, both the concerted and
ionic pathways were investigat-
ed by using DFT calculations,
and compared with the experi-
mental results.
Figure 1. a) Reaction profile of MTO-catalyzed dehydration reaction of 1-phenylethanol (*,c) to styrene
, b) and the corresponding ether (^, d); b) Conversion–yield plot for the dehydration reaction of 1-
(~
phenylethanol to styrene catalyzed by Re
c) Conversion–yield plot for the side reaction of 1-phenylethanol to diphenethyl ether catalyzed by Re
c), MTO (*, b), H
SO
of styrene, resulting from the dehydration of 1-phenylethanol, catalyzed by Re
SO
2
O
7
(
&
, c), MTO (*, b), H
2 4
SO
(~, d), and pTSA (^, l);
O
7
( ,
&
2
(~, d), and pTSA (^, l); d) Conversion/yield plot for the oligomerization
2
4
(&, c), MTO (*, b),
2
O
7
H
2
(~, d), and pTSA (^, l). Negative values are due to measurement errors in the GC.
4
Chem. Eur. J. 2013, 19, 13224 – 13234
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13225

R. J. M. Klein Gebbink et al.
[
33]
dicates that the ethers can be both reversibly formed from
the alcohol and directly transformed to styrene and 1-phe-
nylethanol.
Wagner–Meerwein rearrangement. The fact that the rear-
rangement product camphene is quantitatively formed, indi-
cates that a carbenium ion is a part of the reaction mecha-
nism. Another indication for a carbenium intermediate is
found in the Re O -catalyzed dehydration of linalool, which
The monitoring in time of the dehydration reaction also
shows an interesting distinction between the different cata-
lysts used. Plotting of the styrene yield against the conver-
sion of 1-phenylethanol using Re O , MTO, H SO , or p-tol-
2
7
yields limonene and terpinolene in a combined selectivity of
65% (Scheme 2, bottom). This ring-closuring reaction likely
also proceeds through a carbenium ion, although the lower
selectivity observed together with the formation of 2,6-dime-
thyloctatriene (32% selectivity, cyclic/linear ratio is 2:1) in-
dicates that the cyclization kinetically competes with depro-
2
7
2
4
uenesulfonic acid (pTSA) as the catalyst reveals that both
rhenium-based catalysts show significantly higher styrene
yields during the reaction compared with the Brønsted acid
catalysts (Figure 1b). The final yield of styrene is a lot lower
with H SO as the catalyst than for the rhenium catalysts,
tonation to form the linear olefin. Interestingly, with H SO4
2
4
2
but in the case of pTSA the yield surprisingly increases rap-
idly near the end of the reaction, to give a final yield of sty-
ACHTUNGTRENNUNGr ene comparable to the rhenium catalysts. Likewise, the con-
as the catalyst, the amount of cyclic olefins relative to the
linear olefins is much higher (7.85:1), showing that the cycli-
zation process is much faster than deprotonation of the
linear cationic intermediate when H SO is used as the cata-
version/yield relationship for ether formation shows the in-
verse trend, with the rhenium-based catalysts giving a low
amount of ethers during the reaction, whereas the Brønsted
acid catalysts show a significantly higher amount of ether
formation (Figure 1c). Both show a sharp decrease of the
amount of ether upon reaching full conversion, as was also
deduced from the reaction profile. A consideration of the
amount of missing components, likely consistent of oligom-
ers or polymers of styrene that do not elute from the GC
column, against the conversion shows a similar trend as for
ether formation (Figure 1d): Re O and MTO show virtually
2
4
lyst.
The reactivity of the various alcohols tested in the dehy-
dration reaction could also give relevant information on the
reaction mechanism. Previously, we tested a broad range of
alcohols, benzylic, allylic, aliphatic, and homoallylic, and ter-
[16,17]
tiary, secondary, and primary alcohols.
Within one class
of alcohols (tertiary, secondary), the following trend in reac-
tivity was observed, based on the initial rate of disappear-
ance of the alcohol:
2
7
no oligomerization during the reaction, whereas H SO
benzylic>allylic>homoallylicꢀaliphatic alcohols
2
4
shows the formation of considerable amounts of oligomers
over 60% at full conversion of 1-phenylethanol). In the
(
When considering the reactivity of substrates within the
classes of either benzylic, aliphatic, or homoallylic alcohols,
the following reactivity trend was observed:
case of pTSA very little oligomers are formed during the re-
action, although at conversions above 90% oligomers are
also formed in this case. These data clearly show that the se-
lectivity profile for the rhenium-based catalysts is distinc-
tively different from those of the Brønsted acid catalysts, in-
dicating that the kinetic parameters differ substantially.
An interesting indication for the mechanism at operation
can be derived from the products formed in the dehydration
reaction. Earlier we have also reported on the Re O -cata-
tertiary>secondary@primary alcohols
Both these trends clearly follow the trend in stability of the
corresponding carbenium ion, although the homoallylic car-
benium ion is expected to be somewhat more stable than
the comparable alkyl carbenium ion due to the involvement
2
7
[34–36]
lyzed dehydration reaction of (iso)borneol, yielding cam-
of the bicyclobutonium and cyclopropylcarbinyl ions.
[17]
phene as the only product (Scheme 2, top). This reaction
proceeds through a 1,2-sigmatropic shift, well-known as the
Nevertheless, the observed trend is a strong indication that
a carbenium ion is involved in the dehydration mechanism.
Next, the influence of various additives on the alcohol-to-
olefin dehydration reaction was tested. For rhenium-cata-
lyzed epoxidation reactions it is known that Lewis bases can
coordinate to the highly Lewis acidic MTO, influencing the
[37]
activity and/or selectivity.
On the other hand, Brønsted
bases could also have an influence by decreasing the proton
activity in the reaction medium. Therefore, we performed
the reaction of 1-phenylethanol to styrene, catalyzed by
MTO, in the presence of 20 equivalents of nitrogen base
(
relative to MTO, Table 1). A wide variety of nitrogen bases
were tested, ranging from low to high-base strength and
from unhindered to strongly hindered bases. Coordination
of the base to MTO was probed by using solution-phase IR
spectroscopy, monitoring the very strong anti-symmetrical
[
16]
Re=O vibration. As reported previously, MTO catalyzes
Scheme 2. Dehydration of isoborneol to camphene (top) and of linalool
to limonene, terpinolene, and 2,6-dimethyloctatrienes (bottom).
the dehydration of 1-phenylethanol very efficiently, with full
13226
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 13224 – 13234

Rhenium-Catalyzed Alcohol-To-Olefin Dehydration
FULL PAPER
Table 1. Influence of nitrogen bases on the MTO-catalyzed dehydration
reaction of 1-phenylethanol to styrene and the corresponding ether, or-
dered by increasing pK .
a
the proton activity of the reaction medium. This is an indica-
tion that a proton transfer is involved in the reaction mecha-
nism.
[
a]
Next, both the MTO- and H SO -catalyzed dehydration
2
4
reaction of 1-phenylethanol to styrene were used to deter-
mine the kinetic order of the different components in the re-
action (Figure 2). For 1-phenylethanol the reaction order
was found to be unity in the 20 mm–1.2m initial concentra-
tion regime using Re O as the catalyst. For H SO also a
first-order relationship was found for the 2–20 mm concen-
tration regime. The order in MTO was found to be unity in
the 0.1–3.2 mm concentration range, but in the higher con-
centration regime (3.2–16 mm), the reaction order seems to
decrease, approaching zero at a concentration of 16 mm.
This indicates a change of rate-determining step from one in
which MTO is involved to one in which MTO is not in-
volved upon increasing the MTO concentration. The experi-
ments described here are all performed at a MTO concen-
tration of 2 mm, thus in the regime in which the dependence
on MTO is first order.
[
b]
Additive
pK
a
Conversion
Yield
[%]
Styrene Ether
n (antisym
Re=O) [cm ]
[c]
[c]
À1 [d]
[%]
2
7
2
4
–
–
>99
13
89
4
3
11
5
965
938, 965
965
Pyrazole
2.5
3.6
2
,6-Di-tert-
butylpyridine
,2’-Bipyridyl
Pyridine
,6-Lutidine
Piperidine
9
4
[
e]
2
4.3
5.2
6.7
9.7
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
915
934
965
930
965, 934
930
965
2
Triethylamine 10.8
[
f]
Quinuclidine
11.0
11.4
[
g]
[h]
DIPEA
[
a] Reaction conditions: 1-phenylethanol (2 mmol), MTO (0.02 mmol),
additive (0.40 mmol), pentadecane (internal standard) (250 mL), toluene
10 mL), 1008C, 2 h. [b] pK of the conjugate acid, values from ref. [38],
unless otherwise mentioned. [c] Based on GC. [d] Solution infrared stud-
ies of MTO and additive (2 equiv) in toluene. [e] Value from ref. [39].
(
a
In contrast with the reaction orders found for the other
catalysts, the reaction order in Re O was determined at
2
7
0
.15. Re O , however, does not completely dissolve in the
2 7
[
f] Value from ref. [40]. [g] N,N-diisopropylethylamine. [h] Value from
ref. [41].
reaction medium, making mass-transfer limitations an im-
portant issue, which could explain the fractional reaction
order. To probe mass-transfer limitations, we prepared vari-
ous sieve fractions of Re O . Three sieve fractions were
conversion and in a good yield (89%) of styrene after 24 h.
In the presence of all but two of the nitrogen bases, pyrazole
and 2,6-di-tert-butylpyridine, the reaction does not proceed
at all. From the infrared spectrum it is observed that pyra-
zole weakly coordinates to the MTO center, showing both
the Re=O vibration of free MTO (965 cm ) and of the
MTO-pyrazole adduct (938 cm ), whereas the sterically hin-
dered 2,6-di-tert-butylpyridine does not coordinate, as shown
by the presence of the 965 cm absorption and no absorp-
2
7
tested as catalyst in the dehydration reaction of a-terpineol
[17]
to give limonene and terpinolene,
using Re O7 particle
2
sizes of 850–500 mm, 500–150 mm, and <150 mm (Table 2).
The measured initial rates of consumption of the starting
À1
À1
material were 6.5, 8.2, and 10.2 mmolh , respectively,
À1
indeed showing a correlation with particle size. This con-
firms that mass-transfer limitations play an important role in
the Re O -catalyzed dehydration reaction.
À1
2
7
tion in the lower energy
region. These two additives are
also the weakest bases tested,
with a pKa of the conjugate
acid of 2.5 and 3.6, respective-
ly. When considering the more
basic additives tested, it is ob-
served that coordination to the
rhenium center does not play a
significant role, as both coordi-
native (e.g., pyridine, piperi-
dine) and non-coordinative
bases (e.g., 2,6-lutidine, N,N-
diisopropylethylamine
(
DIPEA)) completely prevent
any reaction from occurring.
These results show that bases
with a pKa of the conjugate
acid higher than 4 prevent the
reaction from occurring; not
through coordination to the
metal, but likely by decreasing
Figure 2. Order determination for a) 1-phenylethanol (starting concentration varied between 20 mm–1.2m; y=
2
2
0
.9482x+0.1643 with R =0.9935), b) MTO (0.1–16 mm; y=0.8859x+0.4957 with R =0.9908), c) Re
2
O
7
(1–
2
2
20 mm; y=0.1487x+0.2745 with R =0.9074), and d) H SO (2–20 mm; y=1.0140x+1.0877 with R =0.9486).
2 4
Chem. Eur. J. 2013, 19, 13224 – 13234
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13227

R. J. M. Klein Gebbink et al.
Table 2. Influence of Re
2
O
7
particle size on the Re
2
O
7
-catalyzed dehy-
Table 3. Kinetic isotope effects of the Re
2
O
7
- and H
2 4
SO -catalyzed dehy-
dration of a-terpineol.
dration of [D
]1-phenylethanol and [D ]1-phenylethanol.
1 3
Re
size [mm]
2
O
7
particle
Initial rate
t
Olefin
yield [%]
Substrate
Kinetic isotope effect
(k /k
À1
A[ mmolh
H
U
G
R
N
U
G
]
A[ min]
H
U
G
R
N
U
G
A
H
U
G
R
N
U
G
H
D
)
Re
2
O
7
2 4
H SO
8
5
<
50–500
00–150
150
6.5
8.2
10.2
60
45
30
84
85
89
1
.28
1.18
1.65
During our catalytic experiments we found an interesting
1.48
dependence of the dehydration reaction on the presence of
water and oxygen. When performing the reaction in water-
and oxygen-free conditions, no reaction is observed. When
performing the reaction in oxygen-free conditions, but in the
presence of water, the reaction proceeds very sluggishly,
giving only 6% conversions, corresponding to 6 turnovers
after 24 h. This indicates that both water and oxygen play a
vital role in the reaction mechanism. The dependence on
water can be rationalized if a proton transfer is involved in
the reaction mechanism, since this can be facilitated by
small amounts of water in the reaction medium. The de-
pendence on oxygen could indicate that during the reaction
the catalyst is reduced, rendering it inactive, and an oxidant
is required for enduring catalytic performance. It is known
that one of the oxygen atoms of MTO can be removed by
the dehydration reaction could influence the KIE, although
further studies are required to determine the influence of
temperature on the KIE in this reaction. On the other hand,
the KIE values observed here are much lower than those to
be expected when proton loss at the b-carbon is rate-deter-
[44]
mining. Overall, it can be concluded from the KIE mea-
surements that with both [D ]- and [D ]1-phenylethanol a
AHCTUNGTRENNUNG
1
3
secondary KIE is at play, thus that both the a- and b-CÀH/
D bonds are not broken in the rate-determining step, ruling
out proton loss as the rate-determining step.
Finally, we determined the activation barriers of the dehy-
dration reaction by varying the temperature between 50 and
1008C and measuring the initial rate of consumption of the
1-phenylethanol starting material (Figure 3). A linear fit on
[10]
[42]
[11]
mild reductants such as H2, PPh3, or sulfite to form
methyldioxorhenium(V) (MDO). At higher temperatures it
has also been demonstrated that alcohols can be oxidized to
the corresponding ketone and as such act as a reducing
[13]
agent toward MTO. Under our reaction conditions no ke-
tones have been observed, and also when performing a stoi-
chiometric reaction between MTO and 1-phenylethanol
under oxygen-free conditions only trace amounts of ketone
1
and no reduction of MTO to MDO is observed by H NMR
spectroscopy after 24 h. Therefore the reduction of MTO by
the alcohol can be considered unlikely.
An important tool to probe the rate-determining step is
the use of isotopic labeling at various positions in the sub-
strate to probe the involvement of a kinetic isotope effects
Figure 3. Eyring plot for the dehydration reaction of 1-phenylethanol to
2
ꢀ
styrene using MTO (left; y=À6746xÀ2.977 with R =0.9514; DG
=
98
2
À1
ꢀ
À1
ꢀ
À1 À1
5
8.57 kJmol ; DH =56.10 kJmol ; DS =À8.30 Jmol
K
) or H
2
SO
4
(
KIE). Therefore, we synthesized 1-phenylethanol with
2
ꢀ
À1
ꢀ
(
right; y=À6197xÀ2.773 with R =0.6757; DG =54.00 kJmol ; DH =
2
98
À1
ꢀ
À1 À1
either one deuterium at the a-carbon of the alcohol or three
deuterium atoms at the b-position. Both compounds were
used as substrate in the dehydration reaction, using either
Re O or H SO as catalyst, and the initial rate of disappear-
51.53 kJmol ; DS =À8.29 Jmol
-phenylethanol (2 mmol), MTO (0.02 mmol) or H
pentadecane (250 mL, internal standard), toluene (10 mL), 50–1008C.
K
) as catalyst. Reaction conditions:
1
2
SO (0.05 mmol),
4
ꢀ
ꢀ
ꢀ
DH derived from Eyring equation, DG from Arrhenius equation, DS
2
7
2
4
follows from the Gibbs free energy equation.
ance of the starting material was compared with that of
non-deuterated 1-phenylethanol (Table 3). The KIE of
[
D ]1-phenylethanol was determined at 1.3 and 1.2 for
the ln (initial rate/T) versus 1/T was performed and from
ꢀ ꢀ
1
Re O and H SO , respectively, whereas for [D ]1-phenyle-
thanol the KIE was 1.5 and 1.6 for Re O and H SO , re-
this the activation enthalpy (DH ), entropy (DS ) and
2
7
2
4
3
ꢀ
Gibbs free energy (DG ) were obtained by using the Eyring
2
7
2
4
spectively. For the solvolysis of 1-phenylethyl chloride in
and Arrhenius equations. For the MTO-catalyzed dehydra-
tion reaction the free energy of activation DG =
ꢀ
polar solvents, known to proceed through an S 1 mecha-
N
À1
ꢀ
nism, an a-KIE of 1.15 and a b-KIE of 1.23 at 258C was re-
58.6 kJmol , for the H SO -catalyzed reaction DG =
2 4
[43]
À1
ported. The somewhat higher KIE values observed in the
-phenylethanol dehydration compared with the solvolysis
54.0 kJmol . In both cases a small and negative entropy of
ꢀ
À1 À1
1
activation DS =À8.3 Jmol
K
was found, indicating that
of phenylethyl chloride could be explained by the minor in-
volvement of nucleophilic attack in the solvolysis, resulting
in a lower KIE. Also the elevated temperature required for
the transition state is in a more ordered state than the start-
ing state, likely caused by the encounter of two reactant
molecules.
13228
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 13224 – 13234

Rhenium-Catalyzed Alcohol-To-Olefin Dehydration
FULL PAPER
Summarizing the experimental mechanistic work reported
here, the addition of both coordinating and non-coordinat-
ing nitrogen bases of sufficient base strength inhibit the re-
action, indicating that proton transfer plays a role in the
mechanism. In addition, the strong dependence of the dehy-
dration reaction on water also indicates that proton transfer
is involved in the mechanism. The found carbon skeletal re-
arrangements and the trends in reactivity of different alco-
hols point to the involvement of a carbenium species, which
is in accordance with the role of proton transfer in the
mechanism. Proton loss is, however, not involved in the
rate-determining step, as proven by the experimental KIEs.
Finally, kinetic-order determinations have shown that both
the reaction order in substrate and catalyst is unity and ther-
modynamic analysis has shown a small and negative entropy
of activation, indicating that both substrate and catalyst are
involved in the rate-determining step.
tional and 6-31G** basis set for H, C, and O, and
LANL2DZ plus ECP for Re in these calculations.
In these calculations, the first step in both pathways is the
coordination of the 1-phenylethanol to MTO. We have in-
vestigated four different reactions: a [2+2] addition of the
alcohol OÀH bond over the ReÀCH bond (TS
,
3
1ÀA
Figure 4), a [2+2] addition of the alcohol OÀH bond over a
Re=O bond (TS ), and two additions via a six-membered
1
ÀB
transition state (TS1 and TS ). TS involves a proton
ÀC
1À2
1À2
transfer from the alcohol OH to a second alcohol molecule,
concurrent proton transfer from the second alcohol mole-
cule to a Re=O group and alkoxide coordination to the rhe-
nium center. TS1 also involves two 1-phenylethanol mole-
ÀC
cules, yet three extra water molecules are added to provide
stabilization of the proton-transfer step. In the case of TS1
,
ÀB
TS1 , and TS the result is the formation of a coordinated
ÀC
1À2
alkoxide and protonation of a rhenium oxo group, whereas
with TS1 the formation of a coordinated alkoxide is ac-
ÀA
DFT calculations: Next to the experimental investigations,
we decided to investigate the mechanism of the rhenium-
catalyzed alcohol-to-olefin dehydration reaction by means
of density functional theory (DFT). In doing so, we have fo-
cused on both concerted as well as ionic pathways
companied by the formation of methane.
The [2 +2] addition of the OÀH bond over a Re=O bond
ꢀ
(TS ) has a Gibbs free energy of activation DG =
1
ÀB
À1
121.8 kJmol (relative to the van der Waals complex) and
is endergonic (D G=41.4 kJmol ), whereas the [2+2] addi-
À1
r
(
Scheme 2, see above). For the DFT calculations we have
tion over the ReÀCH bond (TS ) is inaccessibly high in
3
1ÀA
ꢀ
À1
used MTO as the catalyst; MTO is completely soluble in tol-
uene, which is the solvent applied for all experimental inves-
tigations, and is monomeric in solution, in contrast to
Re O , which has a polymeric form, thereby complicating
the DFT calculations. In addition, we have used 1-phenyle-
thanol as the substrate. We have applied the B3LYP func-
activation energy (DG =187.3 kJmol ), but is quite exer-
À1
gonic (D G=À68.2 kJmol ). The six-membered transition-
r
À1
state TS has a free energy of DG=68.0 kJmol , which is
1À2
energetically the most favorable pathway for the formation
of an alkoxide species. This pathway is likely the most fa-
vored due to the reduced strain in the six-membered transi-
2
7
Figure 4. Calculated Gibbs free energies of the reaction of MTO with 1-phenylethanol. All complexes except 1 are van der Waals complexes.
Chem. Eur. J. 2013, 19, 13224 – 13234
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13229

R. J. M. Klein Gebbink et al.
tion-state, compared with the much more strained four-
membered transition-states of both TS1 and TS . The re-
After coordination of 1-phenylethanol to the rhenium cat-
alyst (2), we first investigated the concerted mechanism, in
which the b-proton transfers intramolecularly to either a
Re=O or the ReÀOH group via a six-membered transition-
ÀA
1ÀB
lated six-membered transition-state TS1 , involving three
ÀC
additional water molecules, has a free energy of DG=
À1
1
17.3 kJmol . The much higher free energy of activation of
state (TS
or TS , Figure 6). The first reaction yields-
BÀI
BÀ3
the latter pathway is likely caused by the reduced nucleophi-
licity of the oxo groups of MTO, as the oxo groups donate
electron density to the additional water molecules, thereby
stabilizing the MTO molecule and raising the activation
energy. Intermediate C, surprisingly, is very high in energy
A
T
N
T
E
N
N
[ReO (Me)(OH) ] together with styrene, whereas the latter
2
2
directly recovers the MTO catalyst while expelling water
and styrene. Both TS
activation energy DG =138 and 166 kJmol , respectively,
and are therefore implausible.
and TS , however, require a high
BÀ3
BÀI
ꢀ
À1
and even higher than TS1 . Likely this is caused by the dif-
As the concerted pathway, either through a hydrated
(Figure 5) or non-hydrated (Figure 6) rhenium species, lies
very high in free energy, we further investigated the ionic
pathway. In this pathway, the CÀO bond of the alcohol is
ÀC
ficulty of doing a full conformational space analysis due to
the high number of fragments in this state, combined with a
poor treatment of the entropic term (see below).
Another possible reaction pathway for coordination of 1-
phenylethanol to the catalyst proceeds through a hydration
reaction of MTO with water to form a [ReO (Me)(OH) ] in-
broken after coordination to the rhenium catalyst, yielding
an anionic rhenium species and a carbenium intermediate.
As DFT disfavors charge separation, even when using an
implicit solvent model such as the polarizable continuum
model (PCM) applied here, the direct breakage of the CÀO
2
2
termediate (D), again either in a [2+2] fashion (TS1
,
ÀD
Figure 5) or via either the six-membered transition-state
TS1 or TS . Subsequent reaction with 1-phenylethanol
bond, forming a styrene carbenium ion and a rhenium
ÀE
1ÀF
À1
can proceed through a [2+2] addition TS
or either of the
anion, is very high in energy (D G>450 kJmol ). Upon ad-
FÀH
r
six-membered transition states TS
or TS
involving ad-
dition of explicit solvent molecules though the charges are
more stabilized and more reasonable energies can be ob-
tained. We, therefore, have calculated the CÀO bond-break-
EÀG
EÀI
ditional water molecules. TS
has a free energy of DG=
is somewhat lower in free
1ÀE
À1
8
4.3 kJmol , whereas TS
1ÀF
À1
energy (DG=70 kJmol ), making both pathways compara-
ble in energy to the most favorable alkoxide formation men-
tioned above. The following reaction to coordinate 1-phe-
nylethanol, however, leads to inaccessibly high overall acti-
vation energies (DG =170–180 kJmol ), indifferent to the
followed pathway (TS , TS , or TS ), rendering the
age reaction using two additional water molecules and an
additional 1-phenylethanol molecule, which are hydrogen-
bonded to the oxo groups of the rhenium complex
(Figure 7). The van der Waals starting complex, intermedi-
ate C (Figure 4), has a very high free energy in comparison
ꢀ
À1
with the separated fragments (intermediate 2, DG =
EÀG
FÀH
EÀI
C
À1
À1
hydrated pathway implausible. Subsequent reactions from
intermediates G, H, or I towards dehydration of the 1-phe-
nylethoxide and closure of the catalytic cycle were therefore
not considered.
254.1 kJmol vs. DG =48.3 kJmol ), indicating that the
2
entropic term is not well-treated. We have compensated for
this error by treating the consecutive steps as a separate re-
action and setting intermediate 2 at zero, thereby relating
Figure 5. Calculated Gibbs free energies of the reaction of MTO with water and subsequently with 1-phenylethanol. All complexes except 1 are van der
Waals complexes.
13230
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 13224 – 13234

Rhenium-Catalyzed Alcohol-To-Olefin Dehydration
FULL PAPER
Figure 6. Calculated Gibbs free energies for the intramolecular proton transfer from the b-position of the 1-phenylethoxide to either a rhenium oxo or
rhenium hydroxyl group, starting from [ReO
complexes.
2
(Me)(OH) ACHUTTGNENRNUG( 1-PE)] and finally yielding styrene, MTO, and water. All complexes except 3 are van der Waals
Figure 7. Calculated free energies for the ionic pathway, starting from [ReO
2
(Me)(OH) AHCTUTGNENRNUG( 1-PE)] and finally yielding MTO and styrene. Additional hydro-
gen-bonded water is omitted from the transition states for clarity, all complexes are van der Waals complexes.
the consecutive complexes to that point, yet the relative free
energies of activation are unchanged.
mates the energy of charge-separated fragments. Likely, this
is caused by errors in the treatment of coulombic interac-
tions and charge dispersion forces, two factors that are
known to be poorly represented by standard DFT meth-
This pathway consists of a number of consecutive reac-
tions; the first is the CÀO breakage of the alkoxide frag-
[44]
ment 2, which is stabilized by an alcohol molecule (TS
,
À4
ods.
Although this hampers the reliability of the Gibbs
2
ꢀ
À1
DG =49.4 kJmol ), resulting in the anionic rhenium spe-
free energies of intermediates 5 and 6, both are sufficiently
low in energy that neither is likely to actually lie above the
energy of any of the preceding intermediates (2 or 4), there-
fore not playing a limiting role in the catalytic cycle.
À
cies [ReO (Me)(OH)] and a carbenium ion (4). Next, the
3
b-proton of the carbenium ion is transferred to a water mol-
ecule (TS4 ), yielding a molecule of styrene (5). The fact
À5
that TS4 is slightly lower in energy than 4 indicates that
The formed [ReO (Me)(OH) ] can finally expel an equiv-
2 2
À5
the energy surface is very shallow at this point. Subsequent-
ly another proton transfer occurs from the hydronium ion to
the rhenium anion, yielding the neutral [ReO (Me)(OH) ]
alent of water, to reproduce the original catalyst species
MeReO . This reaction is the reverse reaction of the hydra-
3
tion reaction, 1ÀE (Figure 5), having a relatively low activa-
3
2
ꢀ
À1
ꢀ
À1
species and water (TS5 , DG =11.9 kJmol ). The overall
tion energy barrier (DG =21.0 kJmol ), thereby closing
À6
reaction of this pathway is slightly exergonic (D G=
the catalytic cycle. The overall reaction leading from 1 to 3
r
À1
À1
À37.4 kJmol ). The very low relative energy of the charge-
is slightly endergonic (D G=4.9 kJmol ).
r
separated fragments is noteworthy, as usually DFT overesti-
Chem. Eur. J. 2013, 19, 13224 – 13234
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13231

R. J. M. Klein Gebbink et al.
Scheme 3. Calculated lowest-energy pathway for the MTO-catalyzed dehydration of 1-phenylethanol, yielding styrene and water.
The DFT results, in summary, show that the lowest energy
pathway found proceeds through the coordination of an al-
cohol molecule to the MTO catalyst via a proton transfer,
The mechanisms proposed in literature for the g-alumina-
catalyzed dehydration reaction, proceeding through an E2-
[22–24]
type mechanism,
can be ruled out for the rhenium-cata-
involving six-membered transition state TS1 , assisted by an
lyzed reaction on the basis of the high activation energies
found for the concerted mechanism, depicted in Figure 6.
The mechanism presented here does resembles the one
found for tungsten-based polyoxometallates, for which also
À2
additional alcohol molecule, forming the alkoxyhydroxyrhe-
nium species 2 (Scheme 3). Next, the alkoxide in intermedi-
ate 2 undergoes CÀO cleavage, stabilized by an additional
[25]
alcohol molecule and two additional water molecules
an E1-type mechanism has been proposed, as well as the
mechanism proposed for the dehydrative coupling of allylic
alcohols using Re O , for which an S 1 mechanism is pro-
(
TS2 ), to form the charge-separated intermediate 4. Subse-
À4
quently, proton transfer from the b-position of the carbeni-
um ion to a water molecule (5), and proton transfer from
the hydronium ion to the rhenium anionic species occurs,
forming styrene and [ReO (Me)(OH) ] (6). Finally, water is
2
7
N
[27]
posed.
As MTO and tungsten-based polyoxometallates
are based on neighboring third-row transition metals, and
both are in high oxidation states, primarily ligated by
oxygen atoms, the similarity in mechanism is not unexpect-
ed.
2
2
expelled from intermediate 6, assisted by an additional
water molecule via six-membered transition state TS1
,
ÀE
thereby recovering the catalyst MTO and closing the cata-
lytic cycle.
The activation energies that we have found for the coordi-
nation of either a water molecule or the alcohol fragment
through the [2+2] addition are in good agreement with
those reported in literature for the very similar coordination
of a diol to MTO, forming a diolate species under the expul-
This catalytic cycle can be described best as an E1-type
mechanism and is very similar to that generally proposed
for the Brønsted acid-catalyzed alcohol-to-olefin dehydra-
[18]
[46]
tion reaction involving a carbenium ion. The main differ-
sion of water.
ence, however, lies in the pK of the hydroxyrhenium spe-
When correlating the experimental and theoretical calcu-
lations, many observations found in the experimental data
are complemented by the DFT calculations. The involve-
ment of a carbenium intermediate was found by the ob-
served substrate reactivity trends and rearrangements and
was also found in the DFT calculations. The role of proton
transfers, as indicated by the influence of water and nitrogen
bases, is also involved in the mechanistic cycle determined
by DFT. A good correlation between the kinetic order
a
cies, which is expected to be lower than that of the strong
Brønsted acids usually applied in dehydration reactions,
thereby making the deprotonation step of the carbenium
ion more plausible. We, however, did not calculate the ether
formation step here, which is proposed to be a relevant
[19,20]
pathway in the Brønsted acid-catalyzed reaction,
and
therefore we do not rule out the involvement of such a path-
way in the rhenium-catalyzed reaction.
13232
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 13224 – 13234

Rhenium-Catalyzed Alcohol-To-Olefin Dehydration
FULL PAPER
found in catalyst and substrate and the DFT calculations
can be found. The additional alcohol molecule considered in
the DFT calculations is part of an explicit hydrogen-bonded
solvent shell, also involving some water molecules in the cal-
culations depicted in Figure 7, required to stabilize the
proton-transfer steps involved. Although we depicted an al-
cohol molecule in the calculations, this molecule is inter-
changeable by a water molecule present in solution. There-
fore this alcohol molecule does not contribute to the kinetic
equation, thus confirming the first order dependence found
experimentally for the alcohol. The role of oxygen, experi-
mentally found to be required for the rhenium-catalyzed de-
hydration reaction, cannot be correlated with the DFT cal-
culations, and it remains unclear what its influence is caused
by. The exact activation energy as found by experiment
could not be confirmed by DFT, due to uncertainty in the
Gibbs free energy in the second part of the cycle, which in-
volves several fragments, and causes poor treatment of the
entropic term in DFT. Overall, a good agreement between
experiment and theory was found, indicating that the mech-
anism found by DFT calculation is plausible.
these differences, although further studies are required to
elucidate this in more detail.
Experimental Section
[
47]
1 3
General: [D ]- and [D ]1-Phenylethanol were synthesized from aceto-
phenone according to literature procedures. Sulfuric acid (95–97%) was
obtained from Merck and used without further purification. All other
compounds were obtained from Sigma–Aldrich or ABCR and used with-
out further purification. GC analysis was performed on a Perkin–Elmer
Autosystem XL Gas Chromatograph equipped with an Elite-17 column
(
30 mꢁ0.32 mmꢁ0.50 mm) and a flame ionization detector. Solution in-
frared analyses were performed by using an ASI Applied Systems Reac-
tIR 1000.
Typical procedure for the rhenium-catalyzed dehydration of alcohols to
olefins: The catalyst (0.01 or 0.02 mmol) was dissolved or suspended in
technical grade toluene (10 mL). The substrate (2 mmol), pentadecane
(
250 mL), and, where applicable, nitrogen base (0.4 mmol) were added by
syringe. The reaction flask was sealed with a septum and the mixture was
heated with magnetic stirring to 1008C for 24 h. Samples for GC analysis
were taken by syringe and filtered over Florisil, eluting with ethyl ace-
tate.
Quantum-chemical calculations: Quantum-chemical calculations were
performed by using density functional theory (DFT) methods using the
[
49]
Gaussian 09 program. The hybrid B3LYP functional was used, which
includes a mixture of Hartree–Fock exchange with DFT exchange-corre-
Conclusion
[50]
lation given by Beckeꢂs three-parameter hybrid functional
with Lee,
[
51]
Yang, and Parrꢂs gradient-corrected correlation functional. For hydro-
[
52]
We have performed both experimental and theoretical stud-
ies on the rhenium-catalyzed alcohol-to-olefin dehydration
reaction in general, and the MTO-catalyzed dehydration re-
action of 1-phenylethanol to styrene in particular. The ex-
perimental studies showed the dependence on water and the
intolerance to base, and the involvement of proton transfer
in the catalytic cycle. In combination with the observed sub-
strate reactivity trends and rearrangements, these strongly
suggest the involvement of a carbenium ion intermediate.
Kinetic isotope effect studies, furthermore, ruled out proton
loss from the b-position of the alcohol as the rate-determin-
ing step.
gen, carbon and oxygen the 6–31G** basis set was used, for rhenium
[
53]
the LANL2DZ basis set and effective core potential (ECP) was used,
a commonly used functional/basis set combination for DFT calculations
[
54–56]
yielding good results for rhenium-based complexes.
Addition of an
extra f-polarization function to the LANL2DZ basis set did not yield sig-
nificantly different results. Furthermore, various combinations of func-
tionals and basis sets were tested on the hydration reaction of MTO, and
À1
the found activation energies for TS1ÀD varied with Æ6.7 kJmol , indi-
cating a good reproducibility of the results with various functionals and
basis sets.
The polarizable continuum model (PCM) was used as solvation model,
using toluene as the solvent. Frequency calculations were performed on
the same level of theory to obtain the Gibbs free energies and to confirm
the nature of the stationary points, yielding a single imaginary frequency
for transition states and no imaginary frequency for minima, except for
The DFT calculations indicated that the lowest energy
pathway most likely proceeds through coordination of the
alcohol to the rhenium center, subsequent CÀO breakage,
À1
vdW1’ and TS1ÀC, which showed one (À11.3 cm ) and two imaginary fre-
À1
quencies (À35.2 cm ), respectively. Based on the very low values of the
imaginary frequencies, these were considered as an artifact in the calcula-
tions and as such of no relevance for the obtained results.
yielding a carbenium ion, proton transfer from the carbeni-
um ion to water, forming styrene, and subsequent proton
transfer from the hydronium ion to the rhenium anion, and
finally water expulsion to close the catalytic cycle. The ex-
perimental and theoretical results are in good agreement
with each other, as all observations found in experiment can
be correlated with the results obtained in theory.
Acknowledgements
This research has been performed within the framework of the CatchBio
program. The authors gratefully acknowledge the support of the Smart
Mix Program of the Netherlands Ministry of Economic Affairs and the
Netherlands Ministry of Education, Culture and Science. Dr. E.A. Pidko
The catalytic cycle proposed here for the rhenium-cata-
lyzed alcohol-to-olefin dehydration reaction resembles the
E1 mechanism commonly proposed for the Brønsted acid-
catalyzed reaction. Interestingly, we do find differences in
reactivity and in the reaction profiles between the two types
of catalyst, which are not reflected in the currently proposed
mechanisms for both types of catalyst. Possibly the involve-
ment of an ether formation pathway in both the Brønsted
acid- and rhenium-catalyzed reaction is crucial in explaining
(
Eindhoven University of Technology) and Dr. M. Moret (Utrecht Uni-
versity) are acknowledged for input and discussions concerning the DFT
studies.
[
[
1] T. J. Korstanje, R. J. M. Klein Gebbink, Top. Organomet. Chem.
012, 39, 129–174.
2] W. A. Herrmann, D. Marz, E. Herdtweck, A. Schafer, W. Wagner,
H. J. Kneuper, Angew. Chem. 1987, 99, 462–464; Angew. Chem. Int.
Ed. Engl. 1987, 26, 462–464.
2
Chem. Eur. J. 2013, 19, 13224 – 13234
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13233

R. J. M. Klein Gebbink et al.
[
3] W. A. Herrmann, D. Marz, E. Herdtweck, J. Organomet. Chem.
990, 394, 285–303.
[36] G. A. Olah, G. K. S. Prakasg, G. Rasul, J. Am. Chem. Soc. 2008, 130,
9168–9172.
1
[
[
[
[
[
4] K. P. Gable, T. N. Phan, J. Am. Chem. Soc. 1994, 116, 833–839.
5] K. P. Gable, J. J. J. Juliette, J. Am. Chem. Soc. 1995, 117, 955–962.
6] K. P. Gable, J. J. J. Juliette, J. Am. Chem. Soc. 1996, 118, 2625–2633.
7] K. P. Gable, Organometallics 1994, 13, 2486–2488.
[37] W. D. Wang, J. H. Espenson, J. Am. Chem. Soc. 1998, 120, 11335–
11341.
[38] CRC Handbook of Chemistry and Physics; 92nd ed. (Ed.: W. M.
Heynes), CRC Press/Taylor and Francis, Boca Raton, FL, 2012, Sec-
tion 5.
8] G. K. Cook, M. A. Andrews, J. Am. Chem. Soc. 1996, 118, 9448–
9
449.
[39] R. A. Jones, B. D. Roney, W. H. F. Sasse, K. O. Wade, J. Chem. Soc.
B 1967, 106–111.
[
9] Z. Zhu, J. H. Espenson, J. Org. Chem. 1996, 61, 324–328.
[
[
[
[
[
[
[
[
[
10] J. E. Ziegler, M. J. Zdilla, A. J. Evans, M. M. Abu-Omar, Inorg.
Chem. 2009, 48, 9998–10000.
11] S. Vkuturi, G. Chapman, I. Ahmad, K. M. Nicholas, Inorg. Chem.
[40] N. M. Hext, J. Hansen, A. J. Blake, D. E. Hibbs, M. B. Hursthouse,
O. V. Shishkin, M. Mascal, J. Org. Chem. 1998, 63, 6016–6020.
[41] R. Hrdina, C. E. Muller, R. C. Wende, K. M. Lippert, M. Benassi, B.
Spengler, P. R. Schreiner, J. Am. Chem. Soc. 2011, 133, 7624–7627.
[42] W. A. Herrmann, J. Kuchler, J. Felixberger, E. Herdtweck, W.
Wagner, Angew. Chem. 1988, 100, 420–422; Angew. Chem. Int. Ed.
Engl. 1988, 27, 394–396.
[43] V. J. Shiner, W. E. Buddenba, B. L. Murr, G. Lamaty, J. Am. Chem.
Soc. 1968, 90, 418–426.
[44] D. S. Noyce, D. R. Hartter, R. M. Pollack, J. Am. Chem. Soc. 1968,
90, 3791–3794.
[45] M. Rapacioli, F. Spiegelman, D. Talbi, T. Mineva, A. Goursot, T.
Heine, G. Seifert, J. Chem. Phys. 2009, 130, 244304.
[46] S. Bi, K. Wang, L. Liu, P. Li, Z. Lin, Organometallics 2012, 31,
6139–6147.
[47] P. L. Polavarapu, L. P. Fontana, H. F. Smith, J. Am. Chem. Soc. 1986,
108, 94–99.
2
010, 49, 4744–4746.
12] I. Ahmad, G. Chapman, K. M. Nicholas, Organometallics 2011, 30,
810–2818.
13] E. Arceo, J. A. Ellman, R. G. Bergman, J. Am. Chem. Soc. 2010,
32, 11408–11409.
14] M. Shiramizu, F. D. Toste, Angew. Chem. Int. Ed. 2012, 51, 8082–
086.
15] S. C. A. Sousa, A. C. Fernandes, Tetrahedron Lett. 2011, 52, 6960–
962.
2
1
8
6
16] T. J. Korstanje, J. T. B. H. Jastrzebski, R. J. M. Klein Gebbink, Chem-
SusChem 2010, 3, 695–697.
17] T. J. Korstanje, E. F. de Waard, J. T. B. H. Jastrzebski, R. J. M. Klein
Gebbink, ACS Catal. 2012, 2, 2173–2181.
18] F. A. Carey, R. J. Sundberg, Advanced Organic Chemistry: Part A:
Structure and Mechanisms, 5th ed., Springer, New York, 2007,
pp. 563–564.
[48] T. Hosokawa, T. Ohta, S. Kanayama, S. Murahashi, J. Org. Chem.
1987, 52, 1758–1764.
[
[
[
19] D. Bother, A. Steinkemper, H. Warnecke, Ind. Eng. Chem. Res.
[49] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X.
Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.
Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, ꢄ.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox,
Gaussian, Inc.: Wallingford, CT, 2009.
2
006, 45, 2896–2993.
20] X. Xu, M. J. Antal Jr., D. G. M. Anderson, Ind. Eng. Chem. Res.
997, 36, 23–41.
1
21] J. R. Mora, D. J. Marquez, E. Marquez, M. LoroÇo, T. Cordova, G.
Chuchani, Int. J. Quantum Chem. 2012, 112, 78–88.
[
22] B. Shi, B. H. Davis, J. Catal. 1995, 157, 359–367.
[
23] B. Shi, H. A. Dabbagh, B. H. Davis, J. Mol. Catal. A Chem 1999,
1
41, 257–262.
[
[
24] B. Shi, H. A. Dabbagh, B. H. Davis, Top. Catal. 2002, 18, 259–264.
25] M. J. Janik, J. Macht, E. Iglesia, M. Neurock, J. Phys Chem. C 2009,
1
13, 1872–1885.
[
[
26] J. Yi, S. Liu, M. M. Abu-Omar, ChemSusChem 2012, 5, 1401–1404.
27] B. G. Das, R. Nallagona, P. Ghorai, J. Org. Chem. 2012, 77, 5577–
5
583.
28] Y. Liu, R. Hua, H. B. Sun, X. Qiu, Organometallics 2005, 24, 2819–
821.
29] J. Jacob, J. H. Espenson, J. H. Jensen, M. S. Gordon, Organometallics
998, 17, 1835–1840.
[
[
2
[50] A. D. Becker, J. Chem. Phys. 1993, 98, 1372–1377.
[51] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[52] P. C. Hariharan, J. A. Pople, Theor. Chim. Acta 1973, 28, 213–222.
[53] P. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 270–283.
[54] C. Guo, J. Song, J. Jia, X. Zhang, H. Wu, Organometallics 2010, 29,
2069–2079.
1
[
[
30] P. Edwards, G. Wilkinson, J. Chem. Soc. Dalton 1984, 2695–2702.
31] W. A. Herrmann, W. A. Wojtczak, G. R. J. Artus, F. E. Kꢃhn, M. R.
Mattner, Inorg. Chem. 1997, 36, 465–471.
[
[
[
32] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.
33] G. A. Olah, Acc. Chem. Res. 1976, 9, 41–52.
34] G. A. Olah, D. P. Kelly, C. L. Jeuell, R. D. Porter, J. Am. Chem. Soc.
[55] A. B. Bengali, M. B. Hall, H. Wu, Organometallics 2008, 27, 5826–
5829.
[56] B. Safi, J. Mertens, F. de Proft, R. Alberto, P. Geerlings, J. Phys.
Chem. A 2005, 109, 1944–1951.
1
970, 92, 2544–2546.
[
35] J. S. Staral, I. Yavari, J. D. Roberts, G. K. S. Prakash, D. J. Donovan,
Received: January 20, 2013
G. A. Olah, J. Am. Chem. Soc. 1978, 100, 8016–8018.
Published online: August 14, 2013
13234
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 13224 – 13234