Heterolytic activation of C-H bonds on CrIII-O surface sites is a key step in catalytic polymerization of ethylene and dehydrogenation of propane

Article abstract of DOI:10.1021/ic502696n

We describe the reactivity of well-defined chromium silicates toward ethylene and propane. The initial motivation for this study was to obtain a molecular understanding of the Phillips polymerization catalyst. The Phillips catalyst contains reduced chromium sites on silica and catalyzes the polymerization of ethylene without activators or a preformed Cr-C bond. Cr^II sites are commonly proposed active sites in this catalyst. We synthesized and characterized well-defined chromium(II) silicates and found that these materials, slightly contaminated with a minor amount of Cr^III sites, have poor polymerization activity and few active sites. In contrast, chromium(III) silicates have 1 order of magnitude higher activity. The chromium(III) silicates initiate polymerization by the activation of a C-H bond of ethylene. Density functional theory analysis of this process showed that the C-H bond activation step is heterolytic and corresponds to a -bond metathesis type process. The same well-defined chromium(III) silicate catalyzes the dehydrogenation of propane at elevated temperatures with activities similar to those of a related industrial chromium-based catalyst. This reaction also involves a key heterolytic C-H bond activation step similar to that described for ethylene but with a significantly higher energy barrier. The higher energy barrier is consistent with the higher pK_a of the C-H bond in propane compared to the C-H bond in ethylene. In both cases, the rate-determining step is the heterolytic C-H bond activation.

Full text of DOI:10.1021/ic502696n

Forum Article
pubs.acs.org/IC
III
Heterolytic Activation of C−H Bonds on Cr −O Surface Sites Is a Key
Step in Catalytic Polymerization of Ethylene and Dehydrogenation of
Propane
†
†
†
Matthew P. Conley, Murielle F. Delley, Francisco Nun
́
̃
ez-Zarur, Aleix Comas-Vives,
and Christophe Coperet*
́
̈ ̈
Department of Chemistry and Applied Biosciences, ETH Zurich, Vladimir-Prelog-Weg 1−5, CH-8093 Zurich, Switzerland
*
S Supporting Information
ABSTRACT: We describe the reactivity of well-defined
chromium silicates toward ethylene and propane. The initial
motivation for this study was to obtain a molecular understanding
of the Phillips polymerization catalyst. The Phillips catalyst
contains reduced chromium sites on silica and catalyzes the
polymerization of ethylene without activators or a preformed Cr−
II
C bond. Cr sites are commonly proposed active sites in this
catalyst. We synthesized and characterized well-defined
chromium(II) silicates and found that these materials, slightly
III
contaminated with a minor amount of Cr sites, have poor polymerization activity and few active sites. In contrast,
chromium(III) silicates have 1 order of magnitude higher activity. The chromium(III) silicates initiate polymerization by the
activation of a C−H bond of ethylene. Density functional theory analysis of this process showed that the C−H bond activation
step is heterolytic and corresponds to a σ-bond metathesis type process. The same well-defined chromium(III) silicate catalyzes
the dehydrogenation of propane at elevated temperatures with activities similar to those of a related industrial chromium-based
catalyst. This reaction also involves a key heterolytic C−H bond activation step similar to that described for ethylene but with a
significantly higher energy barrier. The higher energy barrier is consistent with the higher pK of the C−H bond in propane
a
compared to the C−H bond in ethylene. In both cases, the rate-determining step is the heterolytic C−H bond activation.
INTRODUCTION
chromium on oxide supports are efficient dehydrogenation
■
3
catalysts that can convert propane to propene and H , an
2
The activation of small molecules is a common theme in the
petrochemical and commodity chemical industries. For
example, the synthesis and upgrading of small molecules is
central to the Haber−Bosch (N /H ), Fischer−Tropsch (H /
4
increasingly important process run on industrial scales. In this
Forum Article, we will describe how these two catalytic
processes are related by similar heterolytic C−H bond
activation steps. We will first summarize the classical and
well-defined chromium silicates for ethylene polymerization
and then show through combined experimental and computa-
tional studies that the well-defined chromium species are also
active in the dehydrogenation of propane.
2
2
2
CO), methane reforming (CH /H O/CO ), and methanol-to-
4
2
2
1
olefin processes, to name but a few. In addition to these
examples, metathesis, oligomerization, and polymerization of
olefins are also crucial industrial examples that in most cases
rely on the conversion of small molecules (ethylene, propylene,
etc.) to products (olefins, polymers, etc.). Traditionally, the
low-molecular-weight olefins for these processes are synthe-
sized by cracking techniques. However, the current abundance
of low-molecular-weight alkanes in shale gas has caused a
renewed interest in dehydrogenation processes for the
production of olefins. The catalysts used industrially for the
reactions mentioned above will vary from process to process,
although they are exclusively heterogeneous because of the ease
of separating the desired product from the reaction mixture and
the ability of these catalysts to be easily regenerated.
The Phillips Catalyst: A Historical Perspective. In the
early 1950s, Banks and Hogan, researchers at Phillips
Petroleum, found that chromium oxides supported on silica
(CrO /SiO ) catalyzed the polymerization of ethylene in the
x
2
absence of activators or cocatalysts to form high-density
2
,5
polyethylene (HDPE). The CrO /SiO material, typically
x
2
referred to as the Phillips catalyst, was the first commercialized
HDPE catalyst and currently accounts for roughly half of global
HDPE production per year.
Supported chromium species have a rich history in the
activation of small molecules. Chromium-supported on silica2
was the first commercialized ethylene polymerization catalyst,
a discovery that predates the discovery of Ziegler−Natta
polymerization by roughly 2 years. Related catalysts containing
Special Issue: Small Molecule Activation: From Biological Principles
to Energy Applications
Received: November 8, 2014
©
XXXX American Chemical Society
A
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
The Phillips catalyst was discovered 2 years prior to Ziegler’s
The Phillips catalyst is prepared by impregnating a silica
support with aqueous solutions containing a chromium source,
typically CrO or Cr(OAc) , followed by a calcination step at
discovery of the TiCl /Et AlCl polymerization catalyst and well
4
2
before the generally accepted mechanism for the polymer-
3
3
ization of olefins catalyzed by homogeneous transition-metal
complexes was established. Extensive studies on homogeneous
olefin polymerization catalysts led to the classical insertion
mechanism shown in Scheme 1, a mechanism that probably
high temperatures (>400 °C). At this stage, the surface species
are chromate(VI) esters supported on the silica surface
(Scheme 2). At low chromium loadings (<1% by weight),
6
VI
monomeric Cr sites E are usually formed on the silica
15
surface, while dimeric and polymeric species can also be
obtained at higher loadings. Polymerization of ethylene occurs
when E is contacted with ethylene to form “reduced”
chromium species that are the proposed active sites. Because
Scheme 1. Migratory Insertion Mechanism for Transition-
Metal-Catalyzed Olefin Polymerization
VI
the Cr species must be reduced prior to polymerization
initiation, there is a pronounced induction period for polymer
formation when using CrO /SiO . Alternatively, E can be
x
2
prereduced by carbon monoxide (CO) at 300 °C to produce a
catalyst that polymerizes ethylene with a less pronounced
16
induction period.
Several leading reviews have shown the difficulty in
17
determining the active site in the Phillips catalyst. Despite
these extensive efforts, the active site of the Phillips catalyst is
unknown. A major challenge in determining the active site in
18
this catalyst is the presence of only ca. 10% active chromium.
7
Early work by Baker and Carrick showed that ethylene reacts
with chromium(VI) silicate to form aldehydes, ketones, and
also applies to the heterogeneous Ziegler−Natta catalyst. An
unsaturated metal alkyl A is formed by the treatment of a
precatalyst with an alkylaluminum cocatalyst or by activation
with a Lewis acid. A coordinates ethylene to form B and
undergoes migratory insertion to form C. Repeating the
coordination and insertion of ethylene propagates the polymer
chain. Chain transfer or termination releases the polymer and
regenerates A. This general mechanism applies to a large
family of catalysts based on early- and late-transition-metal
II
19
8
predominately Cr on the silica surface. In agreement with
II
9
this report, several spectroscopic studies show that Cr is the
20
major oxidation state on the surface after treatment with CO.
In particular, Cr K-edge X-ray absorption spectroscopy (XAS)
has been used extensively to study the Phillips catalyst, which
showed that isolated species assigned to F are the major surface
1
0
21
chromium site.
6
b,11
metals,
which can be either homogeneous and heteroge-
neous. This mechanism also applies to catalysts that
selectively incorporate long-chain branches, stereoselectively
Several initiation species were proposed based on isolated
1
2
II
Cr surface sites (F), which are summarized in Scheme 2b. The
13
common theme in each of these proposals is two-electron
14
II
IV
polymerize α-olefins, and control the molecular weight
oxidation of Cr sites to form Cr organometallic inter-
mediates. For example, oxidative coupling of two ethylene
molecules would form the chromacyclopentane (G), which
11a
distribution of the polymer. The Phillips catalyst polymerizes
ethylene without an activator or a preformed Cr−C bond, a
unique feature among all known olefin polymerization catalysts.
22
could insert ethylene by ring expansion. Alternatively, G can
a
Scheme 2. (a) Reactivity of the Phillips Catalyst
a
VI
II
(
a) The Cr -containing silica material (E) is reduced by ethylene or by CO to form predominately Cr on silica (F). F is proposed to polymerize
ethylene. (b) Proposed structures that initiate polymerization: chromacyclopentane species (G), chromium allyl hydride (H), chromium(IV) vinyl
hydride (I), or chromium(IV) alkyidene (J).
B
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
t
undergo β-hydride elimination to form chromium allyl hydride
Scheme 3. Reaction of [Cr[OSi(O Bu) ] ] with Silica
3 2 2
(
H), which can insert ethylene into the Cr−H bond to
Dehydroxylated at 700 °C (SiO
) To Form [(
SiO)Cr (OSi(O Bu) ) ] and Subsequent Thermal Treatment
2 3 3
2
‑(700)
5,23
II
t
propagate the polymer.
The Cr sites could also activate a
C−H bond of ethylene by oxidative addition to form
To Yield [(SiO) Cr ]
4
2
24
chromium(IV) vinyl hydride (I) or coordinate ethylene and
25
rearrange to chromium(IV) alkyidene (J). Each of these
intermediates could, in principle, polymerize ethylene either by
26
a Cossee−Arlman insertion (for H and I) or by a Rooney−
27
Green mechanism (for J).
II
Although monomeric Cr surface species are the most
commonly proposed active site in the Phillips catalyst, the low
quantity of active sites indicates that most of the spectroscopic
data are probably related to inactive sites. Indeed, the debate
concerning the active sites in the Phillips catalysts is still
vigorous, and even the oxidation state of the active sites is
controversial. For example, Lunsford et al. showed that
chromium(III) salts supported on silica were competent
28
polymerization catalysts. Organometallic chromium(III) and
chromium(IV) complexes supported on silica are also active in
25a,29
II
the polymerization of ethylene.
Although Cr is the
predominant oxidation state in a CO-reduced Phillips catalyst,
these latter studies indicate that higher oxidation state
chromium sites are also competent ethylene polymerization
catalyst precursors. The major challenge in determining the
active site in the Phillips catalyst, and the formation of the first
Cr−C bond, is obtaining a chromium silicate material with a
high percentage of active sites, controlled chromium oxidation
state, and a coordination environment.
Preparation of Well-Defined Cr/SiO Catalysts. Con-
2
trolled functionalization of supports with molecular precursors
can give well-defined chromium surface species, a general
12b,30
method referred to as surface organometallic chemistry.
Several well-defined chromium surface species have been
2
5a,29a,31
prepared using this method,
in some cases with the
31
goal to understand the active sites of the Phillips catalyst.
These studies produced catalysts that either already contained a
M−C bond or led to chromium(VI) species after calcination,
which must undergo the complicated reduction to low-valent
chromium discussed above.
Using the thermolytic molecular precursor (TMP) approach
and [(SiO) Cr ]. However, the electron paramagnetic
4 2
4
2
31a,b,32
developed by Tilley and co-workers.
we synthesized
resonance (EPR) spectrum of [(SiO) Cr ] indicated the
III
silica-supported materials containing “ligandless” chromium
silicates with defined oxidation states and nuclearity. This
presence of traces of Cr sites (several percent), presumably
II
resulting from the adventitious oxidation Cr during
approach involves grafting of a molecular precursor having
thermolysis.
t
II
−
OSi(O Bu)3 ligands onto silica containing a known
[(SiO) Cr ], containing mainly Cr sites, polymerizes
4
2
concentration of surface silanol groups, followed by thermal
treatment of the grafted complex to eliminate the organic
ethylene with low activity. Poisoning experiments revealed that
only 0.2 equiv of 4-methylpyridine shuts down all polymer-
ization activity, indicating a low amount of active sites in this
t
residues (isobutene and BuOH). We synthesized a well-
t
defined chromium(II) siloxide [Cr[OSi(O Bu) ] ] and grafted
material. N O is a known activator of a CO-reduced Phillips
3
2 2
2
t
31d
this complex on silica to form [(SiO)Cr (OSi(O Bu) ) ], as
catalyst, which prompted us to investigate the activation of
2
3 3
shown in Scheme 3. In TMP, removal of the organic groups is
generally accomplished by heating to high temperatures under
air, an approach that led to numerous well-defined oxidation
[(SiO) Cr ] with N O. This treatment resulted in a 1 order
4
2
2
of magnitude increase in the initial polymerization rate
compared to [(SiO) Cr ], and a dramatic increase in the
4
2
31a,b,32a,33
catalysts.
We desired materials that contained low-
quantity of active sites (65% active sites according to
quantitative poisoning experiments with 4-methylpyridine).
The resulting polyethylene had a large dispersity Đ = M /M =
oxidation-state chromium. Conducting thermolysis of [(
t
SiO)Cr (OSi(O Bu) ) ] under vacuum resulted in the release
2
3 3
w
n
−1
−1
of C organics, which forms [(SiO) Cr ] and conserves the
9.4 (with M = 5500 g mol and M = 52000 g mol ), which
4
4
2
n w
3
4
oxidation state and nuclearity of the molecular precursor. In
fact, the X-ray absorption near-edge structure (XANES) of
is typical for polymers produced with the Phillips catalyst.
In order to determine the structure of this active catalyst we
t
t
t
[
Cr[OSi(O Bu) ] ] , [(SiO)Cr (OSi(O Bu) ) ] and [(
treated the molecular [Cr[OSi(O Bu) ] ] complex with N O,
3
2 2
2
3 3
3 2 2
2
SiO) Cr ] contain very similar features, indicating that each
which resulted in isolation of the chromium(III) dimer K
shown in Scheme 4. We determined the XAS signature of K
[XANES and extended X-ray absorption fine structure
4
2
II
contains Cr , and the coordination environment is similar to
that of the molecular complex in [(SiO)Cr (OSi(O Bu) ) ]
t
2
3 3
C
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
Scheme 4. (a) Reaction of [(SiO) Cr ] with N O Giving the Chromium(III) Dinuclear Species [(SiO) Cr ] and (b)
4
2
2
6
2
t
Reaction of [Cr[OSi(O Bu) ] ] with N O Giving K
3
2 2
2
(
EXAFS)] and [(SiO) Cr ] treated with N O and found
Scheme 5. Preparation of [(SiO) Cr] by Grafting
4
2
2
3
t
remarkably similar spectral signatures. This result indicates that
(SiO) Cr ] reacts with N O to form the chromium(III)
Cr[OSi(O Bu) ] ·2THF onto SiO
and Subsequent
3
3
2‑(700)
[
Thermal Treatment
4
2
2
material [(SiO) Cr ] shown in Scheme 4 and established
6
2
III
II
that Cr sites are active, while Cr sites display low, if any,
polymerization activity.
The higher activity of dinuclear chromium(III) silicates than
chromium(II) materials is consistent with numerous examples
35
in homogeneous chemistry. However, most studies of the
Phillips catalyst suggest that monomeric chromium sites are
1
7b,21
active.
Using the TMP strategy, we synthesized the
mononuclear chromium(III) surface species [(SiO) Cr]
3
t
from thermolysis of [(SiO)Cr(OSi(O Bu) ) ·THF], as
3
2
shown in Scheme 5. The similar XANES of Cr[OSi-
t
t
(
O Bu) ] ·2THF, [(SiO)Cr(OSi(O Bu) ) ·THF], and [(
3
3
3 2
III
SiO) Cr] indicates that the Cr oxidation state is conserved
3
36
during the grafting and thermolysis steps.
[
(SiO) Cr] polymerized ethylene to form linear poly-
3
ethylene with rates and dispersities that are similar to those
obtained for the [(SiO) Cr ]-catalyzed polymerization of
6
2
ethylene. Adsorption of CO on [(SiO) Cr] gives two ν
3
CO
−1
stretches in the IR spectrum at 2202 and 2188 cm , which are
blue-shifted with respect to free CO. According to labeling
studies, these two νCO bands correspond to two different
chromium surface sites and not to symmetric and antisym-
metric vibrations. Consistent with the presence of different
III
chromium sites in [(SiO) Cr], roughly 60% of the Cr
3
surface sites in [(SiO) Cr] are active in polymerization.
3
We turned to ab initio calculations in order to understand the
structure of these different sites. We constructed the two cluster
models shown in Figure 1. The first model 1a contains three
Si−O groups coordinated to a Cr ion, while the second model
b incorporates an additional siloxane bridge coordinated to
III
1
1
a and 1b. The binding of up to three molecules of CO on 1a is
−1
symm
CO
the metal center. The clusters contain six- and eight-membered
exoergic (ΔG = −10.1 kcal mol ). The corresponding ν
and ν
cm with respect to free CO. In contrast, 1b only binds two
molecules of CO (ΔG = −8.7 kcal mol ), and the calculated
CO vibrations are shifted by +58 and +55 cm for the
symmetric and antisymmetric modes. Overall, combined
experimental [+45 cm (major) and +59 cm (minor)] and
computed frequencies are consistent with the presence of a
tricoordinate chromium(III) species as a major surface species,
37
asymm
siloxane rings that are typically observed in bulk silica. We
completed the silicon environments of both clusters with
fluorine, a common approach when using cluster models for
in 1a-3CO are blue-shifted by +53 and +46/+50
CO
−1
−1
2
2b,38
−1
silica.
For all of the calculated stationary points discussed below, we
3
−1
−1
found that the high-spin state (S = / ) is always more stable
2
1
than the low-spin one (S = / ). We calibrated these models by
2
studying the IR frequency of CO coordinated to chromium in
D
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
of [(SiO) Cr ] and [(SiO) Cr] contacted with ethylene
6
2
3
contained strong C−H bands associated with polyethylene, a
3
9
red-shifted SiOH band interacting with the polymer chain,
−1
and signals near 3600 cm previously assigned to the Si−(μ-
31c,36
OH)−Cr species.
The assignment of Si−(μ-OH)−Cr was
based on similar IR signatures of related species in zeolitic
40
materials. However, the IR spectrum of polyethylene contains
combination bands from C−H vibrations at the exact same
41
frequency as that of the Si−(μ-OH)−Cr species. This result
indicates that the IR spectra of [(SiO) Cr ] and [(
6
2
SiO) Cr] after contact with ethylene are dominated by these
3
combination bands, making the unequivocal observation of Si−
(
μ-OH)−Cr species impossible even with deuterium-labeling
Figure 1. Cluster models constructed to represent CrIII sites in silica
and the corresponding CO adducts.
experiments.
Using the cluster model of the isolated CrIII site 1a, we also
investigated their reactivity toward ethylene. Calculations of the
Gibbs free energies were carried out using standard conditions
along with a minor species having the CrIII ion coordinated to
an additional siloxane bridge. Consistent with these blue-shifted
CO signatures, the chromium centers in both 1a and 1b carry a
partial charge of 1.5+ from natural bond order analysis.
We also modeled the chromium sites using cluster models
terminated with −OH groups, resulting in the cluster models
(
25 °C and 1 atm). One or two molecules of ethylene can
coordinate to 1a with favorable energies (see Figure 2). When
1
a′ and 1b′ (Figure S1 in the Supporting Information, SI). The
CO binding free energies for the carbonyl adducts in the −OH-
terminated clusters are −5.4 and −4.6 kcal mol for 1a′ and
−1
1
b′, respectively. The CO stretching bands are blue-shifted
−1
with respect to free CO by +42 and +37 cm for the
symmetric stretching mode and by +39/+36 and +34 cm− for
the antisymmetric stretching mode for 1a′ and 1b′,
respectively. These values are consistent with the experimental
data.
1
Activation and Polymerization Mechanisms. Both [(
Figure 2. Activation of a C−H bond in ethylene by 1a. The numbers
in parentheses are the Gibbs free energies of the corresponding
intermediates, and those on the arrows are the Gibbs free energies of
the transition states, normalized with respect to 1a + ethylene. All
SiO) Cr ] and [(SiO) Cr] still lack a Cr−C bond necessary
6
2
3
to propagate polymerization. We proposed that ethylene reacts
III
with a Cr −O bond by a heterolytic C−H bond activation
mechanism to form chromium vinyl and Si−(μ-OH)−Cr
species based on combined experimental and computational
studies (Scheme 6).
−1
energies are given in kcal mol .
one molecule of ethylene is coordinated to 1a to form 2, C−H
Scheme 6. Proposed Mechanism of Ethylene Polymerization
Involving a Heterolytic Splitting of a C−H Bond across a
Cr−O Bond as the Initiation Step
bond activation occurs with an intrinsic barrier of 41.9 kcal
−1
mol . The transition state has the typical structural features
42
associated with σ-bond metathesis, with a wide O−H−
Cethylene angle of about 144° and a C−H distance equal to 1.53
Å, consistent with the heterolytic splitting of a C−H bond in
the reaction of 2 to 3. Coordinating two molecules of ethylene
to 1a forms 4. In this case, the activation of a C−H bond of
−1
ethylene has a lower barrier of 36.6 kcal mol . The transition
state for the C−H bond activation from 4 is qualitatively similar
We showed that both [(SiO) Cr ] and [(SiO) Cr]
to that found for 2 with a wide O−H−C
angle of 148°
6
2
3
ethylene
polymerize ethylene without an induction period. In addition,
poisoning studies with 4-methylpyridine revealed that most
surface sites (ca. 65%) can initiate polymerization. The quantity
of the active sites in these materials is significantly higher than
that in the Phillips catalysts (<10%) and indicates that the
active site can be detected by spectroscopic methods. The
XANES of [(SiO) Cr ] or [(SiO) Cr] after exposure to
and a C−H distance equal to 1.48 Å, although in this case a
second ethylene ligand is coordinated to the chromium center
at 2.53 Å.
Exchanging the F-termination in 1a by −OH groups does
not noticeably change the energetics of ethylene C−H bond
activation (Figure S3 in the SI).
We also investigated the reactivity of 1b toward ethylene.
The presence of an additional siloxane bridge prevents the
coordination of two ethylene molecules and yields a much
6
2
3
ethylene contains a near-identical edge position relative to
pristine [(SiO) Cr ] or [(SiO) Cr], indicating that the
6
2
3
III
−1
Cr oxidation state is maintained. The polymer produced by
higher C−H bond activation energy barrier (48.4 kcal mol )
either [(SiO) Cr ] or [(SiO) Cr] contains both olefinic
than that found for 1a. This result suggests that sites
coordinating additional siloxane bridges are inactive in
polymerization, which is consistent with the presence of 60%
6
2
3
and alkyl end groups according to NMR analysis.
We previously proposed that IR spectroscopy could be used
to observe the formation of Si−(μ-OH)−Cr. The IR spectrum
of active sites in [(SiO) Cr].
3
E
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
Figure 3. Ethylene polymerization mechanism by 1a. The numbers in parentheses are the Gibbs free energies of the corresponding intermediate, and
−1
those on the arrows are the Gibbs free energies of the transition states, normalized with respect to 1a + ethylene. All energies are given in kcal mol .
−1
Figure 4. Free-energy profile for the ethylene polymerization reaction by 1a. All energies are given in kcal mol .
We also investigated the species classically proposed in the
Phillips catalyst shown in Scheme 2 using the tricoordinate
chromium(III) model 1a. Each surface species in Scheme 2
requires two-electron oxidation at the chromium center, either
oxidative coupling or oxidative C−H activation, which would
lead to chromium(V) intermediates using 1a. Neither
chromium vinyl hydride nor chromacycle could be located as
stable minima. Optimization of the chromium vinyl hydride
species from oxidative C−H activation gives the starting
chromium ethylene adduct (2 in Figure 2). Attempts to
optimize the chromacyclic structure of chromium(V) by the
coupling of two ethylene molecules led to a chromium alkyl
F
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
6,47
4
species, where the terminal CH group is about 2.7 Å away
from the chromium center, indicating that there is no bond
between chromium and that terminal carbon atom (Figure S4
in the SI). This intermediate is 29 kcal mol less stable than
separated reactants and therefore unlikely. The difficulty of
finding stable chromium(V) species is not too surprising
added to the reaction mixture or generated in situ.
The
2
III
heterogeneous Cr /Al O dehydrogenation catalyst operates at
high temperatures to overcome the thermodynamic limitations
2
3
−1
III
of the reaction. Cr /Al O probably activates the C−H bond
2
3
in propane by a heterolytic mechanism on a Cr−O site,
forming chromium alkyl species, as shown in Scheme 7, which
5+
considering the positive redox potential for the reaction Cr →
3+
43
Cr (0.84 V vs NHE). Therefore, neither chromium vinyl
hydride nor chromacyclic species are likely intermediates in the
polymerization of ethylene when chromium(III) species initiate
polymerization. These results are also in agreement with the
conservation of the 3+ oxidation state during polymerization, as
observed by XANES.
Scheme 7. Proposed Mechanism for Propane
III
Dehydrogenation on Cr /Al O
2
3
Figure 3 shows the proposed catalytic cycle for the formation
of an ethylene oligomer catalyzed by 1a. Two molecules of
ethylene coordinate to 1a to form 4, one of which undergoes
C−H bond activation to form 5. Species 5 contains a Cr−C
bond and a coordinated ethylene that can undergo migratory
insertion. After this first insertion, the formation of chromium
−1
butenyl (6) is only 1.8 kcal mol above the initial reactants.
Coordination of an ethylene molecule to 6 forms the π-
coordinated species 7, which is almost thermoneutral with
48
is supported by labeling and operando studies of the catalyst.
−1
respect to 1a (+0.7 kcal mol ). All successive ethylene
Chromium alkyl then undergoes β-hydride elimination to
release the olefin. The high operating temperatures (>400 °C)
allow for the direct liberation of H , either by recombination of
−1
insertions are exoergic by about 10 kcal mol per inserted
ethylene unit, and the intrinsic energy barriers for insertion of
ethylene into chromium alkyl are low, ranging from 9.4 to 14.0
2
III
a Cr−H and a surface O−H to reform isolated Cr sites or by
−1
kcal mol . These energies are typical of M−R-catalyzed olefin
4,48a
σ-bond metathesis to regenerate chromium alkyl.
The C−
4
4
insertion. The most favorable pathway for chain termination
corresponds to the proton transfer of Cr−(μ-OH)−Si to
chromium alkyl, which is the microreverse of the initial C−H
bond activation step on Cr−O sites. Other termination
pathways such as direct chain transfer to monomer and β-
hydride transfer have less unstable intermediates and/or higher
energy barriers than proton transfer.
H bond activation step proposed in dehydrogenation parallels
the initiation step in ethylene polymerization on well-defined
chromium silicates. This led us to investigate the propane
dehydrogenation chemistry of [(SiO) Cr].
3
The reaction of [(SiO) Cr] with propane in a continuous-
3
−1
flow reactor (10 mL min , 20% in argon, 1.5 bar) forms
propene with 72% selectivity. Methane (14%), ethane (2%),
and ethylene (11%) are the other products detected (Figures
S5−S7 in the SI). Propene is produced with an initial turnover
Figure 4 shows the free-energy profile associated with
initiation, polymer growth, and termination involving three
ethylene insertions. The C−H bond activation of ethylene is
the rate-determining step. The driving force for this reaction is
the highly exoergic ethylene insertion reaction (ΔG ≈ −10 kcal
−1
frequency (TOF) of 10.3 h at 550 °C and steadily decreases
−1
over 2 h to a TOF of 2.8 h . The initial rate and steady decay
III
of activity over time are also typical of classical Cr /Al O
2
3
−
1
⧧
−1
mol ) coupled with low barriers (ΔG ≈ 9−14 kcal mol for
each insertion). Proton transfer to terminate the polymer
chains at each chromium alkyl intermediate (red profile in
Figure 4) is associated with a higher transition state than
4
catalysts. The activity of [(SiO) Cr] in propane dehydro-
3
genation is stable for 20 h after this initial rate decrease,
corresponding to a turnover number of 46. All activities and
output ratios can be viewed as an upper limit for the
equilibrium conversion, where coking is not accounted for (a
thermodynamic maximum of 65% of propene is formed at 550
⧧
−1
insertion (ΔΔG = 12 kcal mol ), consistent with polymer
growth over termination.
°
C in propane dehydrogenation).
Propane Dehydrogenation Pathways. Propane dehy-
RESULTS AND DISCUSSION
Propane Dehydrogenation Catalyzed by Well-De-
fined Chromium(III) Silicates. The dehydrogenation of
■
drogenation and associated byproduct formation were also
investigated by density functional theory (DFT) calculations
using 1a as a model for the active sites. The free energies were
calculated using the experimental conditions 550 °C and 1 atm.
The most favorable dehydrogenation pathway catalyzed by 1a
alkanes to alkenes and H (eq 1) is an increasingly important
2
industrial process because of the current abundance of low-
45
molecular-weight alkanes in shale gas. Not surprisingly, this
reaction is highly endoergic and requires high reaction
temperatures. In industry, the catalysts are based on either
3
is shown in Figure 5. The formation of a chromium propane η -
49
H,H,H adduct 12 is highly endoergic, with Cr−H distances
between 2.44 and 2.55 Å and a Cr−C distance of 2.63 Å. The
heterolytic activation of a C−H bond yields chromium(III)
propyl intermediate 13, an endoergic step of about 17 kcal
mol , associated with a rather high activation barrier of 57.0
kcal mol− with respect to the initial reactants (1a + propane).
The transition state associated with the C−H bond activation
step is again typical for a σ-bond metathesis step with a wide
4
supported PtSn or Cr/Al O .
2
3
−1
Two approaches are generally used to overcome the
thermodynamic demands of this reaction, which depend on
whether the catalyst is homogeneous or heterogeneous.
Brookhart and Goldman have shown that homogeneous
1
(
PCP)Ir complexes dehydrogenate alkanes at moderate
O−H−C
angle of 150° and a C−H distance equal to 1.58
propane
4
2
temperatures in the presence of olefin scavengers, either
Å.
G
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
−1
state from 13 to 12 (+57.0 kcal mol ), indicating that the
reaction can proceed via 14. The subsequent release of propene
forms the chromium hydride species 15. This step is exergonic
−1
by 6.8 kcal mol . We could not locate the transition state
associated with propene loss. The formation of H2 and
completion of the catalytic cycle take place by proton transfer
from 15 to form a dihydrogen adduct, Cr(H ) (16), a highly
2
−1
exergonic process by almost 20 kcal mol . The transition state
is 53.1 kcal mol− above that of the initial reactants and displays
the typical geometry for proton transfer by a σ-bond metathesis
transition state with a wide O−H−H angle (137°) and a H−H
distance equal to 1.15 Å.
1
Alternative pathways were also calculated (Scheme 8).
Propane could undergo C−H bond activation at the methylene
fragment, leading to 13′. This C−H bond activation is
associated with a slightly lower energy transition state (+55.8
−1
kcal mol ) than the activation of the primary carbon, leading
−1
to 13 (57.0 kcal mol ). The subsequent β-H transfer from 13′
has an energy profile similar to that of 13, showing that the rate
of formation of propene by either pathway will be only
marginally different (Scheme 8a). Other possible pathways that
do not involve stepwise β-H transfer, olefin decoordination, and
H release are energetically much less favored (Figure S8 in the
2
SI). For example, coupling of the hydrogen atoms in 14 prior to
Figure 5. Dehydrogenation of propane catalyzed by 1a. The numbers
in parentheses are the Gibbs free energies of the corresponding
intermediates, and those on the arrows are the Gibbs free energies of
the transition states, normalized with respect to 1a + propane. All
propene release forms the Cr(H )(olefin) species 17 and is
2
associated with a high energy barrier (Scheme 8b). Another
possibility would be the reaction of a propane molecule with 15
to re-form 13 by a σ-bond metathesis transition state (Scheme
−1
energies are given in kcal mol .
8
c). However, each process is associated with very high
transition-state energies.
From the chromium(III) propyl intermediate, three path-
ways for dehydrogenation were investigated. The chromium
propyl species 13 can undergo β-H elimination to form the
chromium propene hydride surface species 14 (Figure 5). This
These results show that dehydrogenation occurs by β-H
transfer, followed by decoordination of propene, recombination
of Si−(μ-OH)−Cr and Cr−H, and release of H . The overall
2
thermodynamics for dehydrogenation of propane is calculated
50
−1
step is characterized by a late transition state containing a C−
C distance of 1.38 Å, which is very close to that calculated for
free propene (1.33 Å). Although 14 is 14.7 kcal mol− less
stable than 13, the energy of the transition state from 13 to 14
to be +4.5 kcal mol at 550 °C, and the rate-determining step
is the first C−H bond activation of propane.
1
Byproduct Formation Pathways. The formation of
methane, ethane, and ethylene indicates that cracking processes
also occur under experimental conditions. Cracking is
−1
(
+50.7 kcal mol ) is lower than the energy of the transition
Scheme 8. Alternative Pathways for Dehydrogenation of Propane by (a) Dehydrogenation Pathway through the C−H Bond
Activation of the Methylene of Propane, (b) Direct Coupling of the Hydrogen Atoms in 14, and (c) Regeneration of 13 by σ-
Bond Metathesis
H
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
associated with similar elementary steps described for
dehydrogenation (Figures 6 and 7 and Scheme 9), with the
Scheme 9. Alternative Pathways for the Cracking of Propane
by (a) β-Methyl Transfer Followed by Proton Transfer (b)
Regeneration of 13 via σ-Bond Metathesis
H bond of propane, yielding 13. From 13, β-alkyl transfer
forms the Cr(ethylene)(CH ) species 18, an endergonic
3
−1
process by 11.7 kcal mol associated with a transition state
of 60.3 kcal mol− above separated reactants. 18 releases
ethylene to form 19 through an exergonic process of 18.8 kcal
1
−1
mol . Subsequent proton transfer from Si−(μ-OH)−Cr to
Cr−CH in 19 forms the methane adduct 20 by a transition
3
−1
state that is 49.2 kcal mol above separated reactants (1a and
propane). Low barrier release of methane regenerates the initial
catalyst (1a).
Figure 6. Byproduct formation for the reaction of propane with 1a.
The numbers in parentheses are the Gibbs free energies of the
corresponding intermediate, and those on the arrows are the Gibbs
free energies of the transition states, normalized with respect to 1a +
We investigated alternative pathways to crack propane
Scheme 9), all of which are similar to the alternative pathways
(
−1
discussed for dehydrogenation of propane (Figure S9 in the
propane. All energies are given in kcal mol .
SI). The transfer of a proton from Si−(μ-OH)−Cr to Cr−CH
3
prior to ethylene decoordination (Scheme 9a) or σ-bond
metathesis between propane and the Cr−CH species 19
3
(
Scheme 9b) is associated with much higher energies than β-
Me transfer.
The catalytic cycle for producing methane and ethylene is
−1
thermodynamically favored by −6.6 kcal mol . However, β-
alkyl transfer presents a slightly higher overall energy barrier
−1
than the C−H bond activation step (60.3 vs 57.0 kcal mol ),
suggesting that this step is rate-limiting in cracking. This is in
contrast to the dehydrogenation reaction, where the C−H
bond activation is rate-limiting.
The formation of ethane can proceed by two pathways, with
the most favorable one shown in Figure 7; a more complete set
of possibilities can be found in the SI (Figure S10). The
coordination of ethylene generated from the cracking of
propane to 15 forms the π-coordinated species 22, which
inserts ethylene to form the chromium ethyl complex 23. As
expected, the insertion of ethylene into Cr−H is highly
−1
exergonic by 14.2 kcal mol . Proton transfer from 23 forms
ethane and regenerates 1a. The associated transition states for
ethylene insertion and proton transfer are 48.2 and 53.4 kcal
−1
mol above separated reactants. Alternatively, ethane can be
formed by the direct reaction of ethylene and 13 (Figure S10 in
the SI). However, this pathway is associated with higher energy
intermediates and higher energy barriers than those of the
hydrogenation sequence.
Figure 7. Hydrogenation of ethylene by 1a to form ethane. The
numbers in parentheses are the Gibbs free energies of the
corresponding intermediate, and those on the arrows are the Gibbs
free energies of the transition states, normalized with respect to 1a +
−1
propane. All energies are given in kcal mol .
Activation of CH and H . Contacting [(SiO) Cr] with
4
2
3
either H or CH at 150 °C does not result in the detection of
2
4
only difference being cleavage of a C−C bond (β-alkyl transfer)
in place of a C−H bond (β-H transfer). The catalytic cycle
starts with 12 and the successive heterolytic activation of a C−
reaction intermediates (chromium alkyl or hydride) by IR
spectroscopy under the explored conditions. Consistent with
III
this observation, the reaction of H and CH on Cr sites was
2
4
I
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
Figure 8. Most favorable free-energy pathways for dehydrogenation (in green) and cracking (in blue) of propane by 1a. All energies are given in kcal
−
1
mol .
−1
III
calculated to be endergonic by 19.9 and 17.5 kcal mol ,
which tricoordinate Al −O sites react with methane through a
respectively. The intrinsic barrier to activate H is 32.9 kcal
heterolytic C−H bond activation step. One major difference is
2
−
1
III
mol , while the analogous barrier to activate CH is 40.6 kcal
mol . However, if a sufficient thermodynamic driving force is
that the C−H bond activation step is exoergic on Al −O sites
4
−1
III
and endothermic on Cr −O sites, which leads to observable
present, the activation of H is possible. The hydrogenation of
M−CH surface species on γ-alumina and not on the silica-
2
3
ethylene [(SiO) Cr] occurs readily at 70 °C with a 1:1
supported Cr species.
3
mixture of ethylene and H , consistent with the formation of
[(SiO) Cr] also catalyzes the dehydrogenation of propane
2
3
hydride intermediates.
with 72% selectivity for propene at 550 °C; the byproducts are
methane, ethane, and ethylene. DFT analysis shows that the
rate-determining step for dehydrogenation is also the C−H
bond activation of propane on a Cr−O site (Figure 8). The
transitions states for β-H- and proton-transfer steps generating
DISCUSSION
■
The Phillips catalyst is an important industrial catalyst for the
production of HDPE. Since the discovery of the Phillips
catalyst in 1951, the catalytically active site has been
controversial. Counter to several proposals inferring Cr active
sites, we found that well-defined Cr sites are active while Cr
propene and H and regenerating 1a are lower in energy than
2
II
the microreverse of C−H bond activation. The overall reaction
is slightly endergonic at 550 °C according to calculation by +4.5
III
II
−
1
sites have poor to no activity. This finding is in agreement with
several reports in homogeneous catalysis showing that
chromium(III) alkyls are active in the polymerization of
kcal mol .
In contrast, the thermodynamics for the formation of
methane and ethylene byproducts from propane are actually
3
5
−1
ethylene.
much more favorable (−6.6 kcal mol ). However, the key step
The well-defined chromium(III) silicates do not contain a
preformed Cr−C bond that is required for olefin polymer-
ization. A mechanism consistent with the NMR of the polymer
and XAS data is the heterolytic activation of a C−H bond in
ethylene to form a chromium vinyl species that can coordinate
and insert ethylene. DFT calculations are in agreement with
this proposal and indicate that tricoordinate chromium(III)
species are the active sites. Including additional interaction
determining selectivity is the β-X transfer step (X = H, CH ),
3
which is more favored for dehydrogenation (X = H) than for
−1
cracking (X = CH ) by 9.6 kcal mol . These results indicate
3
that C−H bond activation is rate-determining in propane
dehydrogenation, while β-methyl transfer is likely rate-limiting
for the cracking of propane. However, β-methyl transfer can
compete with the initial C−H bond activation of propane
because the overall energetic difference between the two
III
−1
between the Cr site and an adjacent siloxane bridge in our
transition states is ca. 3 kcal mol . This difference is small and
model leads to significantly higher C−H bond activation
barriers, suggesting that these sites are inactive in polymer-
ization. The activation of a C−H bond in ethylene is endoergic
and rate-limiting, but coupling with the highly exergonic
ethylene insertion steps makes the chromium system an active
polymerization catalyst. This proposal parallels what was found
indicates that the two processes are competitive, in agreement
with the detection of both dehydrogenation and cracking
products in the gas phase under experimental conditions.
The activation of an H−X bond (X = H, CH , C H , C H )
3
2
3
3
7
is endoergic in all cases explored here. With the calculated
transition states, we can now compare the activation step for
each substrate. Figure 9 shows the transition states for the
51
for the activation of C−H bonds of methane on γ-alumina, in
J
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
Figure 9. Transition-state structures for the X−H bond activation in C H , CH , C H , and H . The enthalpic transition-state energy [H(TS)] at
3
8
4
2
4
2
−1
2
98 K and 1 atm (in kcal mol , calculated with respect to separated reactants with the BS1 basis set and no dispersion correction) for the activation
of the X−H bond, the pK value, and the gas-phase acidities (ΔH °, in kcal·mol ) for all substrates are given below the structures.
−1
a
r
activation of C H , CH , C H , and H and gives their energy
compared to separated reactants under standard conditions (25
The chromium(III) silicates also catalyze the dehydrogen-
3
8
4
2
4
2
ation of propane to form propene and H at 550 °C with rates
2
°
C and 1 atm). The barriers of the H−X bond activation step
that are similar to the industrial catalysts. Dehydrogenation also
occurs by the C−H bond activation of propane on a Cr−O
bond. The transition state also has σ-bond metathesis character
and structural similarity to that obtained for the C−H bond
activation of ethylene. The calculated barrier to heterolytic
activation of the C−H bond of propane is much higher than
decrease in the order propane ≈ methane > ethylene ≈ H . The
2
transition-state structures for the activation processes of all of
these substrates are remarkably similar. All of the transition-
state structures contain wide O−H−X angles ranging from
1
1
37° to 150° and a H−X distance ranging between 1.15 and
that found for ethylene, consistent with the higher pK value for
.55 Å. The lowest values are for H and ethylene, while the
a
2
the C−H bond of propane, as expected for a heterolytic C−H
values for saturated hydrocarbons are very similar. In all cases, a
formal four-center σ-bond metathesis transition state, con-
sistent with the heterolytic cleavage of the H−X bond, was
bond activation process.
These studies establish that silica-supported Cr sites can
III
4
2,51c,52
activate C−H bonds of small hydrocarbons by a σ-bond
metathesis mechanism. The highly reactive intermediates
generated by C−H bond activation lead to subsequent catalytic
reactivity in polymerization or dehydrogenation. We expect that
the heterolytic C−H bond activation encountered in these
studies is general and will lead to new materials capable of
activating small molecules by similar or related mechanisms.
We are currently exploring this possibility.
obtained.
Analysis of the partial charges in these
transition-state structures indicates that the chromium ion
and the activated C−H bond carry partial positive charges
(
the “X” group carry a partial negative charge (atoms in red).
The easier activation of C −H over C −H bonds can be
related to the difference of pK values and the gas-phase
acidities of these substrates (ΔH °), which is also consistent
with a heterolytic activation process.
disfavored H activation on Cr −O sites in comparison to
ethylene is likely due to the shorter H−H versus C−H bond.
atoms in blue), while the oxygen acceptor of the proton and
sp
2
sp
3
a
r
5
1c,52c
The slightly
EXPERIMENTAL SECTION
General Considerations. All manipulations were performed
■
III
2
under an inert gas atmosphere using Schlenk (argon), glovebox
−5
(
argon), or high-vacuum (10 mbar) techniques. O and water were
2
removed from all gases (propane, ethylene, methane, and H ) by
2
CONCLUSION
■
passing them through a copper catalyst (R3-11, BASF) and activated 4
Å molecular sieves before use. IR spectra were recorded in
transmission mode on a Nicolet 6700 FT-IR spectrophotometer.
Catalytic experiments under gas-flow conditions were performed in a
flow reactor made of stainless steel SS316 (diameter 1.4 cm)
surrounded by an oven connected via a thermocouple reaching into
the catalyst mixed with SiC (2.5 g). The synthesis of [(SiO) Cr ]
The initial motivation of this endeavor was to obtain a
molecular understanding of the active sites in the PhillipIsI
catalyst. Although numerous studies conclude that isolated Cr
sites are active in olefin polymerization, we found that well-
defined dinuclear chromium(II) silicates have, at best, poor
activity. In contrast, well-defined CrIII sites are active. The
heterolytic C−H bond activation of ethylene forms the first
6
2
and [(SiO) Cr] and experimental details for the polymerization of
3
31c,36
ethylene using these materials were described previously.
III
Cr−C bond in these well-defined Cr -containing materials, as
Propane Dehydrogenation Using [(SiO)
60 mg, 0.011 mmol of chromium) was mixed with SiC (2.5 g) and
placed in a flow reactor. Argon was flowed through the reactor (10 mL
3
Cr]. [(SiO)
Cr]
3
(
well as for the CO-reduced Phillips catalyst. The C−H bond
activation step is rate-determining and follows a σ-bond-
metathesis-type transition state. The microreverse of the C−H
bond activation (chain transfer) has a higher barrier than the
insertion of ethylene into a chromium alkyl, which leads to
polymer growth.
−1
min ), while the reactor was heated to 550 °C. At this temperature, a
−1
mixture of propane and argon (10 mL min ; 1:4 propane/argon; gas
−1
hourly space velocity = 353 h ) was passed through the reactor, and
the output gases were monitored by gas chromatography with a flame
ionization detector (FID-GC) for 20 h.
K
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
Reactivity of [(SiO) Cr] toward Propane. A disk of [(
3
ASSOCIATED CONTENT
Supporting Information
Listings of selectivity, activity, and conversion data for propane
dehydrogenation and further DFT calculations. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
SiO) Cr] (30 mg, 0.006 mmol of chromium) was contacted with
3
*
S
propane (150 mbar, 310 equiv per chromium) at room temperature
for 30 min, then at 100 °C for 30 min, then at 150 °C for 30 min, then
at 300 °C for 30 min, then at 400 °C for 5 h, and then at 500 °C for 14
h. The reaction was monitored by IR transmission; after heating at
each temperature, two IR spectra were collected with and without
condensation of the gas phase. After the experiment, FID-GC of the
gas phase was performed, which apart from propane showed methane
AUTHOR INFORMATION
Corresponding Author
E-mail: ccoperet@ethz.ch. Phone: +41446339394.
■
*
(
0.02 mmol, 3.2 equiv per chromium), ethane (0.01 mmol, 1.5 equiv
per chromium), ethylene (0.01 mmol, 1.2 equiv per chromium), and
propylene (0.10 mmol, 15.4 equiv per chromium). The gas phase was
Author Contributions
These authors contributed equally to the work.
†
−5
then evacuated under high vacuum (10 mbar) for 1 h, and another
IR spectrum was recorded.
Notes
Reactivity of [(SiO) Cr] toward CH Monitored by IR
The authors declare no competing financial interest.
3
4
Spectroscopy. A disk of [(SiO) Cr] (30 mg, 0.006 mmol of
3
chromium) was contacted with methane (150 mbar, 320 equiv per
chromium) at room temperature for 1 h, then at 100 °C for 1 h, and
then at 150 °C for 1 h and monitored by IR transmission spectrometry
after heating at each temperature. The gas phase was evacuated under
ACKNOWLEDGMENTS
■
M.F.D. thanks ETH Zu
̈
rich and the Scholarship Fund of the
Swiss Chemical Industry for financial support. A.C.-V. acknowl-
edges support from the Swiss National Foundation (Grant
PZ00P2_148059).
−5
high vacuum (10 mbar) for 1 h, and another IR spectrum was
recorded.
Reactivity of [(SiO) Cr] toward H . A disk of [(SiO) Cr]
3
2
3
DEDICATION
(
3
30 mg, 0.006 mmol of chromium) was contacted with H (150 mbar,
■
2
20 equiv per chromium) at room temperature for 1 h, then at 100 °C
for 1 h, and then at 150 °C for 1 h and monitored by IR transmission
Dedicated to Prof. E. Negishi on the occasion of his 80th
birthday.
spectrometry after heating at each temperature. The gas phase was
−5
evacuated under high vacuum (10 mbar) for 1 h, and another IR
REFERENCES
■
spectrum was recorded.
(
1) (a) Lloyd, L. Handbook of Industrial Catalysts; Springer: Berlin,
011. (b) Bartholomew, C. H. F.; Robert, J. Fundamentals of Industrial
Catalytic Processes, 2nd ed.; John Wiley & Sons, Inc.: New York, 2006.
2) McDaniel, M. P. A Review of the Phillips Supported Chromium
Catalyst and Its Commercial Use for Ethylene Polymerization. In
Advances in Catalysis; Bruce, C. G., Helmut, K., Eds.; Academic Press:
New York, 2010; Chapter 3.
(3) Resasco, D. E. Dehydrogenation−Heterogeneous. In Encyclopedia
of Catalysis; John Wiley & Sons, Inc.: New York, 2002.
Hydrogenation of Ethylene with [(SiO) Cr] Using H . [(
3
2
2
SiO) Cr] on a Grace Sylopol 948 (40 mg, 0.012 mmol of chromium)
3
was preheated at 70 °C and then contacted with a 1:1 mixture of H2/
(
C H (650 mbar, 680 equiv of H /C H mixture per chromium) for 5
2
4
2
2
4
h. The gas phase was monitored and analyzed by IR and FID-GC and
GC−mass spectrometry. After 4 h, the gas phase contained, apart from
ethylene and H , ethane (11 equiv per chromium) and traces of butane
2
(
0.085 equiv per chromium) and hexane (0.022 equiv per chromium).
1
After the experiment, the surface was characterized via IR and H and
(
4) Sattler, J. J. H. B.; Ruiz-Martinez, J.; Santillan-Jimenez, E.;
1
3
C NMR, which evidenced the formation of polyethylene on the
surface.
Weckhuysen, B. M. Chem. Rev. 2014, 114, 10613−10653.
(5) McDaniel, M. P. Supported Chromium Catalysts for Ethylene
Polymerization. In Advances in Catalysis; Eley, H. P. D. D., Paul, B. W.,
Eds.; Academic Press: New York, 1985.
COMPUTATIONAL DETAILS
■
́
(6) (a) Rappe, A. K.; Skiff, W. M.; Casewit, C. J. Chem. Rev. 2000,
The theoretical calculations were carried out at the DFT level with the
100, 1435−1456. (b) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem.
Rev. 2000, 100, 1169−1204.
53
B3LYP hybrid density functional, as implemented in the Gaussian 09
54
suite of programs. All considered reaction pathways were calculated
(7) Corradini, P.; Guerra, G.; Cavallo, L. Acc. Chem. Res. 2004, 37,
231−241.
III
3
3
with the high-spin configuration of the Cr d ion (S = / ) because
2
(
(
8) Kaminsky, W. J. Chem. Soc., Dalton Trans. 1998, 1413−1418.
9) Chen, E. Y.-X.; Marks, T. J. Chem. Rev. 2000, 100, 1391−1434.
the doublet is much higher in energy and no spin crossing was
detected in the potential energy surface exploration. Geometries were
5
5
(10) Amin, S. B.; Marks, T. J. Angew. Chem., Int. Ed. 2008, 47, 2006−
obtained through a combination of basis sets (BS1): the LanL2DZ
effective core potential of Hay and Wadt for chromium and the 6-
2
025.
56
(11) (a) Coates, G. W.; Hustad, P. D.; Reinartz, S. Angew. Chem., Int.
3
11G(d,p) basis set for all other atoms. The nature of all stationary
Ed. 2002, 41, 2236−2257. (b) Nakamura, A.; Anselment, T. M. J.;
Claverie, J.; Goodall, B.; Jordan, R. F.; Mecking, S.; Rieger, B.; Sen, A.;
van Leeuwen, P. W. N. M.; Nozaki, K. Acc. Chem. Res. 2013, 46,
438−1449.
12) (a) Marks, T. J. Acc. Chem. Res. 1992, 25, 57−65. (b) Wegener,
S. L.; Marks, T. J.; Stair, P. C. Acc. Chem. Res. 2011, 45, 206−214.
c) Nicholas, C. P.; Ahn, H.; Marks, T. J. J. Am. Chem. Soc. 2003, 125,
325−4331. (d) Jezequel, M.; Dufaud, V.; Ruiz-Garcia, M. J.; Carrillo-
Hermosilla, F.; Neugebauer, U.; Niccolai, G. P.; Lefebvre, F.; Bayard,
F.; Corker, J.; Fiddy, S.; Evans, J.; Broyer, J.-P.; Malinge, J.; Basset, J.-
M. J. Am. Chem. Soc. 2001, 123, 3520−3540. (e) Joubert, J.; Delbecq,
F.; Sautet, P.; Roux, E. L.; Taoufik, M.; Thieuleux, C.; Blanc, F.;
points along the potential energy surfaces was confirmed by
computation of the Hessian matrix, obtaining no imaginary
frequencies for minima and only one for transition-state structures.
Intrinsic reaction coordinate calculations were carried out in the C−H
bond activation steps to ensure that the transition states connect
properly with the reactants and products. Gas-phase thermal
corrections were calculated at standard conditions (25 °C and 1
atm) for the ethylene polymerization reaction and at 550 °C and 1 atm
for the propane dehydrogenation reaction, including the formation of
byproducts. These calculations allow us to obtain entropic
contributions in the reaction pathways. Electronic energies were
1
(
(
4
5
5a,57
refined with extended basis sets (BS2): LanL2TZ(f)
for
́
Coperet, C.; Thivolle-Cazat, J.; Basset, J.-M. J. Am. Chem. Soc. 2006,
5
6,58
chromium and 6-311++G(d,p) for all other atoms. Additionally,
Grimme’s D3 dispersion corrections with Becke and Johnson
128, 9157−9169.
59
(13) Braunschweig, H.; Breitling, F. M. Coord. Chem. Rev. 2006, 250,
2691−2720.
60
damping were added to the final energies.
L
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
(
(
14) Coates, G. W. Chem. Rev. 2000, 100, 1223−1252.
L.; Theopold, K. H. J. Am. Chem. Soc. 2005, 127, 1082−1083.
(c) Thomas, B. J.; Noh, S. K.; Schulte, G. K.; Sendlinger, S. C.;
Theopold, K. H. J. Am. Chem. Soc. 1991, 113, 893−902.
(36) Delley, M. F.; Nunez-Zarur, F.; Conley, M. P.; Comas-Vives, A.;
Siddiqi, G.; Norsic, S.; Monteil, V.; Safonova, O. V.; Coperet, C. Proc.
Natl. Acad. Sci. U. S. A. 2014, 111, 11624−11629.
15) McDaniel, M. P.; Collins, K. S.; Benham, E. A. J. Catal. 2007,
2
(
52, 281−295.
16) Zecchina, A.; Garrone, E.; Ghiotti, G.; Morterra, C.; Borello, E.
J. Phys. Chem. 1975, 79, 966−972.
17) (a) Weckhuysen, B. M.; Schoonheydt, R. A. Catal. Today 1999,
1, 215−221. (b) Groppo, E.; Lamberti, C.; Bordiga, S.; Spoto, G.;
Zecchina, A. Chem. Rev. 2005, 105, 115−184.
18) McDaniel, M. P.; Martin, S. J. J. Phys. Chem. 1991, 95, 3289−
293.
(
5
(37) (a) Stallons, J. M.; Iglesia, E. Chem. Eng. Sci. 2001, 56, 4205−
4216. (b) Zhuravlev, L. T. Colloid Surf. A 2000, 173, 1−38.
(38) (a) Espelid, O.; Borve, K. J. J. Catal. 2000, 195, 125−139.
(b) Espelid, O.; Borve, K. J. Catal. Lett. 2001, 75, 49−54. (c) Espelid,
O.; Borve, K. J. J. Catal. 2002, 205, 177−190. (d) Espelid, O.; Borve,
K. J. J. Catal. 2002, 205, 366−374.
(39) van der Meer, J.; Bardez-Giboire, I.; Mercier, C.; Revel, B.;
Davidson, A.; Denoyel, R. J. Phys. Chem. C 2010, 114, 3507−3515.
(40) (a) Dzwigaj, S.; Shishido, T. J. Phys. Chem. C 2008, 112, 5803−
5809. (b) Hadjiivanov, K.; Penkova, A.; Kefirov, R.; Dzwigaj, S.; Che,
M. Microporous Mesoporous Mater. 2009, 124, 59−69. (c) Tielens, F.;
Islam, M. M.; Skara, G.; De Proft, F.; Shishido, T.; Dzwigaj, S.
Microporous Mesoporous Mater. 2012, 159, 66−73. (d) Cappus, D.; Xu,
C.; Ehrlich, D.; Dillmann, B.; Ventrice, C. A., Jr.; Al Shamery, K.;
Kuhlenbeck, H.; Freund, H. J. Chem. Phys. 1993, 177, 533−546.
(41) We thank a reviewer of this manuscript for directing us to
literature data containing the FTIR spectrum of HDPE. The FTIR
spectrum of HDPE synthesized in these studies is given in the SI
(Figure S2).
(42) (a) Balcells, D.; Clot, E.; Eisenstein, O. Chem. Rev. 2010, 110,
749−823. (b) Waterman, R. Organometallics 2013, 32, 7249−7263.
(43) Bose, R. N.; Fonkeng, B.; Barr-David, G.; Farrell, R. P.; Judd, R.
J.; Lay, P. A.; Sangster, D. F. J. Am. Chem. Soc. 1996, 118, 7139−7144.
(44) Niu, S. Q.; Hall, M. B. Chem. Rev. 2000, 100, 353−405.
(45) Malakoff, D. Science 2014, 344, 1464−1467.
(46) (a) Haibach, M. C.; Kundu, S.; Brookhart, M.; Goldman, A. S.
Acc. Chem. Res. 2012, 45, 947−958. (b) Choi, J.; MacArthur, A. H. R.;
Brookhart, M.; Goldman, A. S. Chem. Rev. 2011, 111, 1761−1779.
(47) (a) Eisenstein, O.; Crabtree, R. H. New J. Chem. 2001, 25, 665−
666. (b) Baudry, D.; Ephritikhine, M.; Felkin, H. J. Chem. Soc., Chem.
Commun. 1982, 606−607. (c) Baudry, D.; Ephritikhine, M.; Felkin, H.
J. Chem. Soc., Chem. Commun. 1980, 1243−1244. (d) Gupta, M.;
Hagen, C.; Kaska, W. C.; Cramer, R. E.; Jensen, C. M. J. Am. Chem.
Soc. 1997, 119, 840−841. (e) Belli, J.; Jensen, C. M. Organometallics
1996, 15, 1532−1534.
(
3
(
(
19) Baker, L. M.; Carrick, W. L. J. Org. Chem. 1968, 33, 616−618.
20) (a) Weckhuysen, B. M.; Wachs, I. E.; Schoonheydt, R. A. Chem.
Rev. 1996, 96, 3327−3350. (b) Chakrabarti, A.; Wachs, I. E. Catal.
Lett. 2014, 144, 1−10.
(
21) Bordiga, S.; Groppo, E.; Agostini, G.; van Bokhoven, J. A.;
Lamberti, C. Chem. Rev. 2013, 113, 1736−1850.
22) (a) Rebenstorf, B.; Larsson, R. J. Mol. Catal. 1981, 11, 247−256.
b) Espelid, O.; Borve, K. J. J. Catal. 2002, 206, 331−338. (c) Groppo,
E.; Lamberti, C.; Bordiga, S.; Spoto, G.; Zecchina, A. J. Catal. 2006,
40, 172−181.
23) Ruddick, V. J.; Badyal, J. P. S. J. Phys. Chem. B 1998, 102, 2991−
994.
24) Cheng, R.; Liu, Z.; Zhong, L.; He, X.; Qiu, P.; Terano, M.;
(
(
2
(
2
(
Eisen, M.; Scott, S.; Liu, B. Phillips Cr/Silica Catalyst for Ethylene
Polymerization. In Polyolefins: 50 years after Ziegler and Natta I;
Kaminsky, W., Ed.; Springer: Berlin, 2013.
(
25) (a) Amor Nait Ajjou, J.; Scott, S. L.; Paquet, V. J. Am. Chem. Soc.
1
998, 120, 415−416. (b) Ghiotti, G.; Garrone, E.; Zecchina, A. J. Mol.
Catal. 1988, 46, 61−77. (c) Rebenstorf, B. J. Mol. Catal. 1988, 45,
2
2
63−274. (d) Liu, B.; Nakatani, H.; Terano, M. J. Mol. Catal. A: Chem.
003, 201, 189−197. (e) Amor Nait Ajjou, J.; Scott, S. L.
Organometallics 1997, 16, 86−92. (f) Ghiotti, G.; Garrone, E.;
Coluccia, S.; Morterra, C.; Zecchina, A. J. Chem. Soc., Chem. Commun.
1
979, 1032−1033.
26) (a) Cossee, P. J. Catal. 1964, 3, 80−88. (b) Arlman, E. J. J.
Catal. 1964, 3, 89−98. (c) Arlman, E. J.; Cossee, P. J. Catal. 1964, 3,
9−104.
27) Ivin, K. J.; Rooney, J. J.; Stewart, C. D.; Green, M. L. H.;
Mahtab, R. J. Chem. Soc., Chem. Commun. 1978, 604−606.
28) (a) Przhevalskaya, L. K.; Shvets, V. A.; Kazansky, V. B. J. Catal.
(
9
(
(
1
975, 39, 363−368. (b) Myers, D. L.; Lunsford, J. H. J. Catal. 1986,
99, 140−148.
(48) (a) Weckhuysen, B. M.; Schoonheydt, R. A. Catal. Today 1999,
51, 223−232. (b) Coperet, C. Chem. Rev. 2010, 110, 656−680.
(c) Olsbye, U.; Virnovskaia, A.; Prytz, Ø.; Tinnemans, S.;
Weckhuysen, B. Catal. Lett. 2005, 103, 143−148. (d) Weckhuysen,
B. M.; Bensalem, A.; Schoonheydt, R. A. J. Chem. Soc., Faraday Trans.
1998, 94, 2011−2014. (e) Weckhuysen, B. M. Phys. Chem. Chem. Phys.
2003, 5, 4351−4360. (f) For related C−H bond activation on M−O
sites, see: Hu, B.; Getsoian, A.; Schweitzer, N. M.; Das, U.; Kim, H. S.;
Niklas, J.; Poluektov, O.; Curtiss, L. A.; Stair, P. C.; Miller, J. T.; Hock,
A. S. J. Catal. 2014, 322, 24−37.
(
29) (a) Amor Nait Ajjou, J.; Scott, S. L. J. Am. Chem. Soc. 2000, 122,
8968−8976. (b) Ikeda, H.; Monoi, T.; Sasaki, Y. J. Polym. Sci., Part A:
Polym. Chem. 2003, 41, 413−419. (c) Monoi, T.; Haruhiko, H. I.;
Hiroyuki, O.; Yasuaki, S. Polym. J. 2002, 34, 461−465. (d) Takashi, M.;
Haruhiko, I.; Yasuaki, S.; Yasumichi, M. Polym. J. 2003, 35, 608−611.
(
́
30) (a) Coperet, C.; Chabanas, M.; Petroff Saint-Arroman, R.;
Basset, J.-M. Angew. Chem., Int. Ed. 2003, 42, 156−181. (b) Tada, M.;
Iwasawa, Y. Coord. Chem. Rev. 2007, 251, 2702−2716. (c) Thomas, J.
M.; Raja, R.; Lewis, D. W. Angew. Chem., Int. Ed. 2005, 44, 6456−
6
482.
31) (a) Fujdala, K. L.; Tilley, T. D. J. Catal. 2003, 216, 265−275.
b) Fujdala, K.; Brutchey, R.; Tilley, T. D. Tailored Oxide Materials via
Thermolytic Molecular Precursor (TMP) Methods. In Surface and
Interfacial Organometallic Chemistry and Catalysis; Coperet, C.,
(49) Siegbahn, P. E. M.; Svensson, M. J. Am. Chem. Soc. 1994, 116,
10124−10128.
(50) Siegbahn, P. E. M. J. Am. Chem. Soc. 1993, 115, 5803−5812.
(51) (a) Joubert, J.; Salameh, A.; Krakoviack, V.; Delbecq, F.; Sautet,
(
(
́
P.; Coper
23950. (b) Wischert, R.; Coper
Chem., Int. Ed. 2011, 50, 3202−3205. (c) Wischert, R.; Laurent, P.;
Coperet, C.; Delbecq, F.; Sautet, P. J. Am. Chem. Soc. 2012, 134,
14430−14449.
́
et, C.; Basset, J. M. J. Phys. Chem. B 2006, 110, 23944−
Chaudret, B., Eds.; Springer: Berlin, 2005. (c) Conley, M. P.;
Delley, M. F.; Siddiqi, G.; Lapadula, G.; Norsic, S.; Monteil, V.;
Safonova, O. V.; Coperet, C. Angew. Chem., Int. Ed. 2014, 53, 1872−
́
et, C.; Delbecq, F.; Sautet, P. Angew.
1
876. (d) Groppo, E.; Damin, A.; Otero Arean, C.; Zecchina, A.
Chem.Eur. J. 2011, 17, 11110−11114.
32) (a) Coles, M. P.; Lugmair, C. G.; Terry, K. W.; Tilley, T. D.
(52) (a) Maron, L.; Perrin, L.; Eisenstein, O. J. Chem. Soc., Dalton
Trans. 2002, 534−539. (b) Coperet, C.; Grouiller, A.; Basset, J. M.;
Chermette, H. ChemPhysChem 2003, 4, 608−611. (c) Werkema, E. L.;
Maron, L.; Eisenstein, O.; Andersen, R. A. J. Am. Chem. Soc. 2007, 129,
6662−6662.
(53) (a) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J.
J. Phys. Chem. 1994, 98, 11623−11627. (b) Miehlich, B.; Savin, A.;
Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989, 157, 200−206. (c) Lee, C.
T.; Yang, W. T.; Parr, R. G. Phys. Rev. B: Condens. Matter 1988, 37,
(
Chem. Mater. 2000, 12, 122−131. (b) Terry, K. W.; Gantzel, P. K.;
Tilley, T. D. Inorg. Chem. 1993, 32, 5402−5404. (c) Terry, K. W.;
Lugmair, C. G.; Tilley, T. D. J. Am. Chem. Soc. 1997, 119, 9745−9756.
(
(
(
33) Fujdala, K. L.; Tilley, T. D. Chem. Mater. 2001, 13, 1817−1827.
34) Conley, M.; Coperet, C. Top. Catal. 2014, 57, 843−851.
35) (a) Theopold, K. H. Eur. J. Inorg. Chem. 1998, 1998, 15−24.
́
(
b) MacAdams, L. A.; Buffone, G. P.; Incarvito, C. D.; Rheingold, A.
M
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Forum Article
7
(
1
(
85−789. (d) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
e) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200−
211.
54) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
̈
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
revision D.01; Gaussian, Inc.: Wallingford, CT, 2009.
(
55) (a) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299−310.
(
(
b) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270−283.
c) Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284−298.
(
56) (a) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem.
Phys. 1980, 72, 650−654. (b) Mclean, A. D.; Chandler, G. S. J. Chem.
Phys. 1980, 72, 5639−5648.
(
57) (a) Roy, L. E.; Hay, P. J.; Martin, R. L. J. Chem. Theory Comput.
2
008, 4, 1029−1031. (b) Ehlers, A. W.; Bohme, M.; Dapprich, S.;
Gobbi, A.; Hollwarth, A.; Jonas, V.; Kohler, K. F.; Stegmann, R.;
Veldkamp, A.; Frenking, G. Chem. Phys. Lett. 1993, 208, 111−114.
(
58) (a) Frisch, M. J.; Pople, J. A.; Binkley, J. S. J. Chem. Phys. 1984,
8
0, 3265−3269. (b) Clark, T.; Chandrasekhar, J.; Spitznagel, G. W.;
Schleyer, P. V. J. Comput. Chem. 1983, 4, 294−301.
59) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys.
010, 132, 154104−154119.
60) Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comput. Chem. 2011, 32,
456−1465.
(
2
(
1
N
DOI: 10.1021/ic502696n
Inorg. Chem. XXXX, XXX, XXX−XXX