Scope and mechanism of homogeneous tantalum/iridium tandem catalytic alkane/alkene upgrading using sacrificial hydrogen acceptors

Article abstract of DOI:10.1021/om500231t

An in-depth investigation of a dual homogeneous catalyst system for the coupling of alkanes and alkenes based on an early-/late-transition-metal pairing is reported. The system is composed of Cp*TaCl₂(alkene) for alkene dimerization and pincer-iridium hydrides for alkane/alkene transfer hydrogenation. Because there is no kinetically relevant interaction between the two catalysts, the tandem mechanism can be entirely described using the two independent catalytic cycles. The alkene dimerization mechanism is characterized by an entropically disfavored pre-equilibrium between Cp*TaCl ₂(1-hexene) + 1-hexene and Cp*TaCl ₂(metallacyclopentane) (ΔH° = -22(2) kcal/mol; ΔS° = -16(2) eu); thus, the overall rate of alkene dimerization is positive order in 1-hexene (exhibiting saturation kinetics), and increases only modestly with temperature. In contrast, the rate of 1-hexene/n-heptane transfer hydrogenation catalyzed by t-Bu[PCP]IrH₄ is inverse order in 1-hexene and increases substantially with temperature. Styrene has been investigated as an alternate sacrificial hydrogen acceptor. Styrene dimerization catalyzed by Cp*TaCl₂(alkene) is considerably slower than 1-hexene dimerization. The conversion of styrene/heptane mixtures by the Ta/Ir tandem system leads to three product types: styrene dimers, coupling of styrene and heptane, and heptene dimers (from heptane). Through careful control of reaction conditions, the production of heptene dimers can be favored, with up to 58% overall yield of heptane-derived products and cooperative TONs of up to 12 and 10 for Ta and Ir catalysts, respectively. There is only slight inhibition of Ir-catalyzed styrene/n-heptane transfer hydrogenation under the tandem catalysis conditions.

Full text of DOI:10.1021/om500231t

Article
pubs.acs.org/Organometallics
Scope and Mechanism of Homogeneous Tantalum/Iridium Tandem
Catalytic Alkane/Alkene Upgrading using Sacrificial Hydrogen
Acceptors
David C. Leitch, Jay A. Labinger,* and John E. Bercaw*
Arnold and Mabel Beckman Laboratories of Chemical Synthesis, California Institute of Technology, Pasadena, California 91125,
United States
S
* Supporting Information
ABSTRACT: An in-depth investigation of a dual homogeneous catalyst
system for the coupling of alkanes and alkenes based on an early-/late-
transition-metal pairing is reported. The system is composed of
Cp*TaCl2(alkene) for alkene dimerization and pincer-iridium hydrides
for alkane/alkene transfer hydrogenation. Because there is no kinetically
relevant interaction between the two catalysts, the tandem mechanism can
be entirely described using the two independent catalytic cycles. The alkene
dimerization mechanism is characterized by an entropically disfavored pre-
equilibrium between Cp*TaCl2(1-hexene) + 1-hexene and
Cp*TaCl2(metallacyclopentane) (ΔH° = −22(2) kcal/mol; ΔS° = −16(2) eu); thus, the overall rate of alkene dimerization
is positive order in 1-hexene (exhibiting saturation kinetics), and increases only modestly with temperature. In contrast, the rate
of 1-hexene/n-heptane transfer hydrogenation catalyzed by t-Bu[PCP]IrH4 is inverse order in 1-hexene and increases
substantially with temperature. Styrene has been investigated as an alternate sacrificial hydrogen acceptor. Styrene dimerization
catalyzed by Cp*TaCl2(alkene) is considerably slower than 1-hexene dimerization. The conversion of styrene/heptane mixtures
by the Ta/Ir tandem system leads to three product types: styrene dimers, coupling of styrene and heptane, and heptene dimers
(from heptane). Through careful control of reaction conditions, the production of heptene dimers can be favored, with up to
58% overall yield of heptane-derived products and cooperative TONs of up to 12 and 10 for Ta and Ir catalysts, respectively.
There is only slight inhibition of Ir-catalyzed styrene/n-heptane transfer hydrogenation under the tandem catalysis conditions.
Such a process must necessarily perform chemistry on alkanes
under relatively mild conditions: the entropic cost of coupling
smaller carbon chains into larger ones means that high
temperatures are thermodynamically incompatible. One
possible technology for achieving this goal, alkane metathesis,
is currently under investigation by a number of research
groups.7 Alkane metathesis operates via combined alkane
dehydrogenation and alkene metathesis using either two
separate catalysts or one catalyst capable of both trans-
formations. In the ideal case, two Cn alkanes afford one
C2n−‑2 alkane and one equivalent of ethane in an approximately
thermoneutral reaction; however, alkane metathesis tends
toward a statistical carbon number distribution of alkanes,
with few examples that can achieve any selectivity for the
desired C2n−2 product.8 While further catalyst development
efforts may result in a viable process, alternative methods for
upgrading light hydrocarbons are also needed.
INTRODUCTION
■
As exploitation of traditional crude oil reserves becomes
economically and geopolitically more volatile, and concern
regarding CO2 emissions from the current inefficient use of
carbon-based energy carriers increases, many countries and
industries are actively pursuing alternate fuel sources for the
global transportation fleet.1 These factors are driving the
development of new technologies for fuel production from
carbon sources such as bitumen and kerogen (so-called oilsands
and shale oil),2 natural gas,3 and biomass.4 In contrast to fuels
derived from crude oil that are predominantly collected by
fractional distillation, obtaining hydrocarbons in the desired
weight range from these alternate carbon sources often requires
extensive refining and synthetic manipulation, such as catalytic
cracking5 and Fischer−Tropsch synthesis.6 In addition to the
optimal fuel-range hydrocarbons (C8−C22), lighter alkanes and
alkenes are abundant products of these processes. Currently,
light hydrocarbons (C<6) have little value as fuels due to their
volatility (liquids) and low volumetric energy density (gases).
This inherent inefficiency in fuel production will increase both
the economic and environmental cost of our continued
exploitation of these emerging carbon sources.
Recently, we outlined a complementary approach toward
light hydrocarbon upgrading based on a tandem9 alkane
dehydrogenation−alkene dimerization sequence.10 Rather than
using alkanes as the sole reactant, this process takes advantage
of the mixed nature of many light byproduct streams by
In order to better utilize low-carbon-number energy carriers,
these feedstocks need to be upgraded to higher molecular
weight compounds, ideally in the diesel fuel range (C10−C22).
Received: March 5, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 1. Tandem Catalytic Alkene Dimerization/Transfer Hydrogenation Approaches toward Hydrocarbon Upgrading
incorporating both alkanes and alkenes as substrates. In an as-
yet unrealized ideal system (Scheme 1, left), one catalyst
dimerizes the 1-alkene component of the mixed gas or liquid
feedstock. Transfer hydrogenation by a second catalyst converts
the alkane component to a 1-alkene, with concomitant
hydrogenation of the C2n alkene to the desired C2n alkane
product. The 1-alkene thus formed is then coupled with
another one equivalent of 1-alkene by the dimerization catalyst,
continuing the synthetic cycle. Overall, this process couples an
alkane and alkene to a higher alkane, with no byproducts
generated. This net alkene hydroalkylation parallels alkane/
alkene coupling reactions using strong acids,11 though using an
entirely different mechanistic approach. While acid-catalyzed
alkylations rely on carbocation intermediates, which give highly
branched products, the tandem catalytic process in Scheme 1
could be tailored to give mostly linear products by a careful
choice of dimerization catalyst.12
In order to elucidate the features of this tandem catalytic
process, we have studied the mechanism of 1-hexene/n-heptane
coupling in considerable detail. These results indicate that the
two catalysts operate with opposing kinetic features, meaning
the ideal reaction conditions (i.e., substrate concentration and
temperature) for each process are very different. Thus, the
optimal set of conditions for tandem catalysis is a compromise
between the two extremes. Furthermore, in an effort to identify
other alkene/alkane combinations that are amenable to tandem
catalysis, we have used styrene as a sacrificial hydrogen
acceptor. By adjusting the ratio of the two catalysts, a high
degree of selectivity (>26:1) for tandem catalysis in this
styrene/alkane system can be realized.
RESULTS AND DISCUSSION
■
Mechanistic Studies of 1-Hexene Dimerization. One of
the most significant findings during our initial tandem catalysis
investigations10 is that the tantalum and iridium catalysts
operate independently in solution. This has been established by
demonstrating that the initial rates of alkene dimerization and
alkene/alkane transfer hydrogenation are the same in tandem
and individual reactions (Figure 2). Thus, a complete picture of
the tandem system can be constructed by overlaying the
mechanisms of the two individual processes.
In our initial communication,10 we reported a dual
homogeneous catalytic system for alkane/alkene coupling
based on an early-/late-transition-metal combination: alkene
dimerization is effected by Cp*TaCl2(alkene) catalyst 1 (Cp* =
C5Me5),13 while transfer hydrogenation is performed by pincer-
ligated iridium catalyst 2 (Figure 1).14−17 This system is able to
Both the tantalum system 113 and iridium system 214−17 have
been well studied by other research groups; however, we
required a detailed understanding of the specific mechanistic
features of each catalyst under the conditions used for the
tandem reactions. Therefore, we have conducted an in-depth
kinetic study of the dimerization of 1-hexene catalyzed by 1 at
elevated temperatures and over a wide range of initial substrate
concentrations. This study complements previous kinetic work
by the Schrock group carried out at lower temperatures (0−90
°C).13d They found that increasing [1-octene] (1.4−6.4 M)
leads to a very small increase (1.3×) in the initial rate of
dimerization at 50 °C, consistent with Cp*TaCl2-
(metallacyclopentane) as the major resting state.
One detail in the original synthetic work by Schrock et al.
that caught our attention concerns the temperature stability of
putative catalytic intermediates: specifically, that Cp*Ta-
Cl2(alkene) complexes undergo rapid decomposition at 100
°C in the absence of added alkene.13c This fact led the authors to
speculate that catalyst decomposition at these elevated
temperatures should be especially problematic at low
concentrations of 1-alkene, since these conditions would
favor Cp*TaCl2(alkene) as the catalyst resting state (vide
Figure 1. Ta and Ir precatalysts investigated for tandem catalytic
alkane/alkene coupling.
couple 1-hexene and n-heptane with a high degree of catalyst
cooperativity; however, the full synthetic cycle shown on the
left of Scheme 1 has not been achieved. Instead, this system
operates by a sacrificial H2 acceptor pathway, in which the
alkene substrate, and not the C2n product, acts as the hydrogen
acceptor (Scheme 1, right). Thus, for every one equivalent of
alkane upgraded, one equivalent of a different alkane is
generated. This sacrificial acceptor reactivity leads to both
dehydrogenative alkane/alkene coupling (Scheme 1, right top)
and double-dehydrogenative alkane coupling (hereafter referred
to as “alkane coupling”, Scheme 1, right bottom) when R and
R′ are different.
B
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 2. Time evolution plot of the tandem catalytic coupling of 1-hexene/n-heptane: (top) reaction conditions and products observed; (bottom
left) time course plot for concentrations of all observed species; (bottom right) expansion and simplification showing formation of C12 (from 1-
hexene dimerization) and n-hexane (from 1-hexene/n-heptane transfer hydrogenation). Colored points are from tandem reactions (same data as
chart on left), while black hollow points represent several individual catalytic runs with either 1 (diamonds) or 2 (squares). These data indicate that
the rates of the individual catalytic reactions are identical with the rates observed during tandem catalysis, and therefore the two catalysts operate
independently.
infra). It is surprising, then, that our tandem catalytic alkane/
alkene coupling operates under exactly this set of conditions
with no apparent catalyst decomposition, even over >30 h with
1-alkene added slowly via syringe pump. Thus, we sought to
identify the catalyst resting state under conditions relevant to
tandem catalysis.
The dimerization of ∼240 mM 1-hexene catalyzed by 1 at
100 °C results in the formation of several products (eq 1). Two
Figure 3. Time evolution plot of the dimerization of 1-hexene (240
mM) catalyzed by 1 (8 mM) at 100 °C (three runs).
assumptions allowed Schrock et al. to derive rate laws for the
C12 regioisomers are observed (∼3:1 ratio), as well as a small
amount of cis- and trans-2-hexene (∼6% combined yield). No
3-hexenes are observed. Monitoring this dimerization reaction
over time by GC/FID gives the profile shown in Figure 3.
There is clearly a rate dependence on the concentration of 1-
hexene (i.e., the reaction is not zero order in [1-hexene]).
Additionally, the regioisomer ratio remains relatively constant,
though there is a small change over time (3.5:1 at 20%
conversion, 3.2:1 at 99% conversion).
Attempts to fit the consumption of 1-hexene or the
production of either regioisomer to simple exponential
functions result in poor correlation. This indicates that a
more complex rate law is operative during the course of the
reaction. An examination of the proposed catalytic cycles
(Scheme 2), and the application of several reasonable
formation of both regioisomers (eqs 2 and 6).13d
These dimerization reactions are proposed to proceed
through reversible oxidative coupling of two 1-alkenes at a
Ta(III) center (A to B or B′) followed by irreversible
decomposition of the tantalacyclopentanes B or B′ through a
remarkable ring-contraction mechanism elucidated previous-
ly13b,d (not shown). Finally, displacement of the product C2n
alkene in C/C′ by the 1-alkene substrate regenerates the
Ta(III) complex A. Recall that it is species A that is thermally
unstable and thus apparently unsuitable as a catalyst resting
state at elevated temperatures. It is therefore important to
determine the values of Keq and Keq′ under tandem catalysis
reaction conditions, which will reveal which side of the A + 1-
alkene ⇆ B equilibrium is favored and thus the predominant
tantalum resting state.
C
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(eq 8). Importantly, Keq can be determined independently from
both sets of data (eqs 5 and 9).
Scheme 2. Proposed Catalytic Cycles for the Dimerization of
1-Alkenes into Two Regioisomers Catalyzed by 1
a
k2Keq′[1‐hexene][Ta]0
rate =
1 + Keq[1‐hexene]
(6)
Keq
1
rate−1
=
[1‐hexene]−1
+
k2Keq′[Ta]0
k2Keq′[Ta]0
(7)
1
k2Keq′ =
m′is slope of line from eq 7
m′[Ta]0
(8)
(9)
b′
m′
Keq
=
b′is intercept of line from eq 7
In order to build a kinetic profile along the lines of eqs 2−9
that is relevant in the context of tandem catalysis, we have
examined the initial rates of formation of both C12 regioisomers
over a wide range of initial 1-hexene concentrations (248−4000
mM) and temperatures (80−125 °C). Figure 4 is a
a
Adapted from ref 13d.
In order to simplify the complex rate equations that govern
this kinetic model into eqs 2 and 6, two key assumptions can be
made. First, Keq ≫ Keq′; Schrock et al. report that they have
never observed α/β metallacycles B′, even for alkenes that give
exclusively the “head-to-tail” dimer (here denoted as “minor”
isomer) that would result from decomposition of B′.13c,d
Second, product displacement by substrate is rapid and
quantitative. Alkene ligand exchange at Cp*TaCl2(alkene)
complexes is known to be rapid;13c furthermore, under initial
rate conditions, [substrate] ≫ [product]. Applying these
assumptions gives the rate law represented in eq 2 for the
major isomer, which can be rearranged by taking the reciprocal
of both sides and constructing a linear equation (eq 3). Thus, a
double-reciprocal plot of initial rate versus initial 1-hexene
concentration allows the calculation of both Keq and k1 (eqs 4
and 5).
Figure 4. Initial rate of formation of the major regioisomer from
dimerization of 1-hexene (248−4000 mM) catalyzed by 1 (8 mM) at
100 °C (each line generated by linear regression of an overlay of at
least three runs).
representative example of the data generated, showing the
initial rates of formation of the major C12 regioisomer at 100 °C
over the entire concentration range examined. Each linear
correlation is an overlay of 12−15 data points collected over 3
runs, and the errors on these rates have been estimated by
linear regression statistics. Table 1 contains the initial rate data
collected at 100 °C.
k1Keq[1‐hexene][Ta]0
rate =
Table 1. Representative Initial Rate Data for the
Dimerization of 1-Hexene Catalyzed by 1 at 100 °C
1 + Keq[1‐hexene]
(2)
1
1
a
rate−1
=
[1‐hexene]−1
+
kobs (10−5 M s−1
)
k1Keq[Ta]0
k1[Ta]0
(3)
entry
[1-hexene]0 (mM)
major
minor
1
2
3
4
5
6
7
248
252
5.87(10)
6.03(12)
9.53(27)
16.73(41)
21.37(52)
37.2(11)
42.9(14)
1.85(4)
2.00(5)
1
k1 =
b is intercept of line from eq 3
b[Ta]0
(4)
(5)
412
2.98(10)
5.56(15)
6.77(19)
12.24(38)
13.76(48)
699
b
m
Keq
=
m is slope of line from eq 3
1010
2000
4000
A similar treatment can be applied to the formation of the
minor isomer (eqs 6−9). This method cannot separate Keq′ and
k2, and only the product of these two values can be calculated
a
Numbers in parentheses are standard errors from regression analysis.
D
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
The data from Table 1 reveal that the dimerization rate is
positive order in [1-hexene] and that the kinetics approach
saturation at high [1-hexene]. A plot of kobs versus [1-hexene]
clearly shows this approach to saturation (Figure 5, top),
Table 2. Rate and Equilibrium Constants for 1-Hexene
Dimerization
a
a
a
entry
temp (°C)
Keq (M−1
)
k1 (s−1
)
Keq′k2 (M−1 s−1
)
b
1
2
72
80
2.4(3)
c
0.98(14)
1.01(16)
0.020(2)
0.0047(3)
d
b
3
4
82
90
0.93(8)
c
0.48(12)
0.50(14)
0.053(11)
0.0072(4)
d
b
5
3
93
0.50(8)
c
100
0.205(60)
0.204(93)
0.111(52)
0.129(56)
0.029(26)
0.031(28)
0.153(44)
0.31(18)
1.47(78)
0.0101(4)
0.0126(3)
0.0186(4)
d
c
4
5
110
125
d
c
d
a
Numbers in parentheses are standard errors from repeat measure-
b
c
ments or from regression analysis. NMR spectroscopy data. Major
isomer data. Minor isomer data.
d
catalyst initially exists as the alkene complex A; at 125 °C, this
value is 99%. This analysis is supported by a simple colorimetric
observation: an 8 mM solution of precatalyst 1, which is a
Ta(III) alkene complex (A), is dark purple; upon addition of 1-
hexene at 25 °C, the solution becomes orange, which indicates
metallacycle (B) formation. When it is heated to reaction
temperature, it becomes dark purple (A) and, upon cooling,
returns to orange (B).
We have previously shown that slow addition of 1-hexene by
syringe pump over 24 h to a refluxing n-heptane solution (∼98
°C) of catalysts 1 and 2 results in >95% conversion of 1-hexene
and a catalyst cooperation factor of 91%. If decomposition of
the tantalum catalyst was occurring during the course of these
experiments, such a high degree of catalyst cooperativity would
not be observed. Furthermore, we have tested the thermal
stability of the tantalum catalyst over ∼4 days at 125 °C.
Precatalyst 1 (0.008 mmol) and an aliquot of 1-hexene (0.2
mmol) was dissolved in d8-toluene in a J. Young NMR tube.
Figure 5. (top) Plot of initial rate (kobs) for 1-hexene dimerization
(blue points, major isomer; red points, minor isomer) catalyzed by 1
(8 mM) at 100 °C versus [1-hexene]0 (248−4000 mM; data from
Table 1), indicating an approach to saturation. (bottom) Correspond-
ing double-reciprocal plot to calculate values for Keq, k1, and k2Keq′
according to eqs 2−9.
1
The tube was heated to 125 °C for 12 h and analyzed by H
NMR spectroscopy, which revealed that all of the 1-hexene had
been consumed. A second 0.2 mmol aliquot of 1-hexene was
added, and the tube was again heated to 125 °C. This process
was repeated for seven cycles, and after each 12 h time period
all of the 1-hexene had been consumed. Finally, for the eighth
addition, ∼2 mmol of 1-hexene was added; after 12 h at 125 °C,
all of the 1-hexene was consumed. Final quantification of the
reaction mixture by GC revealed a total TON of 215.
Importantly, only the expected products were observed (C12
alkenes and ∼5% 2-hexenes), confirming that 1-hexene
consumption was not due to side reactions catalyzed by
decomposition products. Clearly, the temperature sensitivity of
Cp*TaCl2(alkene) complexes is not an important factor under
catalytic conditions, even at low substrate concentration.
While this experiment does not confirm that all of the
tantalum remains catalytically active during this time period,
certainly enough remains after 3.5 days to convert 250 equiv of
1-hexene during the final 12 h. In addition, since 0.2 mmol of 1-
hexene should be consumed fairly quickly at 125 °C (<4 h), for
the majority of the 12 h heating periods only the product
alkenes are present. Thus, it would appear that the product 1,1-
disubstituted alkenes can also stabilize Ta(III) intermediates,
even during prolonged heating. The robust nature of the
meaning that the equilibrium governed by Keq is kinetically
relevant up to high [1-hexene]. In fact, at low [1-hexene]
(248−699 mM), the rate dependence on [1-hexene] is nearly
linear, suggesting that A is the major Ta species.
From the rate data obtained, double-reciprocal plots were
constructed for the major and minor regioisomers at each
temperature; a representative plot for data at 100 °C is shown
in Figure 5 (bottom). All of these plots showed excellent linear
correlation (linear regression R2 ≥ 0.99). From the regression
analyses, values for Keq, k1, and Keq′k2 have been calculated;
these data are summarized in Table 2. Independent estimates of
Keq by 1H NMR spectroscopy under catalytically relevant
conditions are in excellent agreement with the values obtained
from the kinetic analysis: Keq (by NMR) = 2.4(3) M−1 (72 °C);
0.93(8) M−1 (82 °C), and 0.50(8) M−1 (93 °C).18
The most striking feature of these results is that above 80 °C
eq < 1, meaning that the A + 1-alkene ⇆ B equilibrium favors
K
the reactants at these higher temperatures. For example, using 8
mM catalyst at 100 °C and an initial [1-hexene] of 250 mM
(tandem catalysis conditions), nearly 90% of the tantalum
E
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 6. (left) van’t Hoff plot (72−125 °C) for the equilibrium Cp*TaCl2(1-hexene) (A) + 1-hexene ⇆ Cp*TaCl2(metallacycle) (B), with
thermodynamic parameters (blue filled points, from kinetics; red hollow points, from 1H NMR spectroscopy). (right) Eyring plot (80−125 °C) for
the unimolecular decomposition of Cp*TaCl2(metallacyclopentane) (B into C), with activation parameters.
Cp*TaCl2(alkene) system is remarkable in this context: truly,
this catalyst is effectively “indefinitely active” (in a dry and
oxygen-free environment!) as observed previously by Schrock
et al.,13d under conditions even more forcing than those
originally anticipated.
slightly larger than that for the product Keqk1, it would appear
that either Keq′ is reduced less by increasing temperature or k2
is increased more (or both). Again, at higher initial [1-hexene]
(4000 mM), the rate increase from 100 to 125 °C is greater
(3.1-fold). All of these factors lead to a decrease in selectivity
for the major isomer as the temperature increases: upon
completion, the regioisomer ratio is 3.25:1 at 100 °C and 2.35:1
at 125 °C.
A second important consequence of the A + 1-alkene ⇆ B
equilibrium is that the rate of dimerization is relatively
insensitive to temperature. For example, the rate of 1-hexene
consumption at an initial concentration of 248 mM increases
only ∼1.6-fold from 100 to 125 °C. Analysis of the equilibrium
and rate constant data in Table 2 reveals the reason for this
behavior. While k1 increases ∼8-fold over this temperature
range, Keq decreases by a factor of ∼5.5. This decrease in Keq can
be understood in simple thermodynamic terms: the equilibrium
is a 2-into-1 reaction and thus should be exothermic (bond
forming) and entropically disfavored. A van’t Hoff analysis
confirms this intuitive rationale, with ΔH° = −22(2) kcal mol−1
and ΔS° = −16(2) eu for the forward reaction (Figure 6, left).
Likewise, an Eyring analysis of k1 values for the unimolecular
decomposition of B into C gives activation parameters of ΔH⧧
= 26(2) kcal mol−1 and ΔS⧧ = 6(5) eu (Figure 6, right). Thus,
as the reaction temperature is raised, the equilibrium shifts back
toward A + 1-alkene, reducing the steady-state concentration of
B. This effect is offset by the increase in k1, meaning that the
overall rate of dimerization to the major isomer increases by
only 1.4-fold from 100 to 125 °C with an initial [1-hexene] of
250 mM. Notably, at an initial [1-hexene] of 4000 mM, the
increase in initial dimerization rate to the major isomer is 2.4-
fold over the same temperature range, consistent with an
equilibrium shift toward B at the higher alkene concentration.
Formation of the minor isomer follows a similar trend, with
an initial rate increase of ∼2-fold observed from 100 to 125 °C
([1-hexene]0 = 248 mM). Here, the situation is slightly more
complex. Since intermediate A feeds into both major and minor
catalytic cycles, a shift of the A + 1-alkene ⇆ B equilibrium
toward A would increase the amount of tantalum available to
form the minor metallacycle B′. This effect is manifested in the
rate law of eq 6: since Keq is only in the denominator, as Keq
decreases, the rate of minor isomer formation increases. Of
course, Keq′ and k2 are subject to the same offsetting
temperature effects as Keq and k1. Even though individual
determination of these values is not currently possible, the
product Keq′k2 has been calculated and shows a 1.8-fold
increase from 100 to 125 °C (Table 2). Since this increase is
The final step of the catalytic cycles in Scheme 2 is the
displacement of the product alkenes by substrate, in equilibria
that are favored in the forward direction (i.e., Koff and Koff′ >1).
While these equilibria are not significant under initial rate
conditions, they become important at high conversion; in other
words, Koff and Koff′ are measures of product inhibition. While a
rigorous analysis of product inhibition kinetics has not been
carried out, the full mechanism including Koff and Koff′ terms
and isomerization rates has been modeled mathematically using
the calculated rate and equilibrium constants from Table 2,
entry 3. The time course data from Figure 3 were fitted
according to the overall mechanism from Scheme 2.19
Refinement of Koff and Koff′ gives values of 4.4(6) and 2.8(9),
respectively. One would expect that Koff >Koff′ due to the
different sterics of the two products: the minor isomer is less
bulky near the CC bond and would thus exhibit stronger
binding to the Ta(III) center. However, the difference in the
obtained values is not statistically significant. Since Koff and Koff′
are ∼3−4, product inhibition is a factor at high conversion;
modeling the data with large (>10) values for Koff and Koff′
gives a poorer correlation. The fact that the product 1,1-
disubstituted alkenes bind to Ta under conditions of high
[product] and low [substrate] is consistent with the thermal
stability experiment described previously; perhaps this product
inhibition behavior contributes to catalyst longevity at elevated
temperatures by stabilizing the reactive Ta(III) fragment.
In summary, there are two key features of alkene
dimerization catalyzed by 1: the reaction is positive order in
alkene substrate with an approach to saturation at high [1-
alkene], and the rate is relatively insensitive to temperature,
especially under conditions of low initial [1-alkene]. Both of
these factors are a direct consequence of the A + 1-alkene ⇆ B
equilibrium, which has Keq < 1 above 80 °C and ΔS° = −16(2)
eu. From these insights, it is clear that the tantalum catalyst
operates most efficiently at lower temperature and high
[substrate], the exact reverse of tandem catalysis conditions;
F
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
thus, the kinetic features of transfer hydrogenation appear to be
opposite to those of alkene dimerization.
(including 1-decene), but accurate quantification of these
species is not reliable due to overlap with the large signal for n-
decane solvent. Therefore, the formation of n-hexane and
internal hexenes has been monitored to generate initial rate
data for transfer hydrogenation and alkene isomerization
respectively (Table 3).
Mechanistic Studies of Ir-Catalyzed Transfer Hydro-
genation. In order to rationalize the striking dichotomy in
optimal conditions between catalytic alkene dimerization and
tandem catalytic alkane/alkene coupling, the kinetics of 1-
hexene/n-decane transfer hydrogenation catalyzed by 2 have
been examined under a similar set of conditions. Many research
groups, most prominently those of Jensen, Kaska, Goldman,
Krogh-Jespersen, and Brookhart, have extensively studied
transfer hydrogenation14−16 and alkene isomerization17 by
pincer-iridium catalysts. This previous work establishing the
mechanisms of these reactions provides an excellent framework
for understanding catalytic transfer hydrogenation in our
tandem system (Scheme 3); however, for alkene dimerization
by 1, we required a detailed picture of the kinetics under our
specific reaction conditions.
Table 3. Initial Rate Data for 1-Hexene/n-Decane Transfer
Hydrogenation and 1-Hexene Isomerization Catalyzed by 2
a
kobs (10−5 M s−1
)
entry
[1-hexene]0 (mM)
temp (°C)
hydrog
isom
1
2
3
4
5
6
7
248
412
100
100
100
100
125
125
125
1.26(5)
0.753(38)
0.572(20)
0.387(10)
12.4(4)
1.00(4)
1.53(5)
2.31(7)
2.79(6)
4.73(20)
6.30(16)
11.8(5)
699
1010
248
412
8.85(23)
2.88(14)
1010
Scheme 3. Postulated Catalytic Cycles for Alkane/Alkene
Transfer Hydrogenation and Alkene Isomerization
a
Numbers in parentheses are standard errors from regression analysis.
a
Catalyzed by 2
Analysis of the initial rates of transfer hydrogenation and
isomerization versus [1-hexene]0 reveals that transfer hydro-
genation is inverse order in [1-hexene] from 248 to 1010 mM,
whereas isomerization is positive order in [1-hexene],
exhibiting saturation behavior (Figure 7). The inverse depend-
Figure 7. Rates of 1-hexene/n-decane transfer hydrogenation and 1-
hexene isomerization catalyzed by 2 at 100 °C versus initial [1-hexene]
(248−1010 mM). The data clearly show inverse order in 1-hexene for
transfer hydrogenation and positive order in 1-hexene for isomer-
ization (approaching saturation).
a
Adapted from refs 14 and 17.
ence on [1-hexene] is consistent with previous work on
cyclooctane/tert-butylethylene (TBE) transfer hydrogenation
catalyzed by 2: the rate is first order in [TBE] at low
concentrations (where the Ir dihydride F is the major resting
state) and inverse order at higher concentrations.16 While
inhibition by TBE is due to reversible oxidative addition/
reductive elimination of a vinyl C−H bond, linear α-olefins
(LAOs, such as 1-hexene) inhibit catalysis by competitive
formation of the 1-alkene complex G. Indeed, G (where the
alkene is 1-octene) is reportedly a major resting state in n-
octane transfer dehydrogenation.14 Thus, at higher [1-hexene],
the equilibrium in eq 11 would be shifted toward intermediate
G, which is not on the transfer hydrogenation cycle.
Heating a solution of 1-hexene (248 mM) and 2 (5 mM) in
n-decane leads to the generation of several products that can be
analyzed by GC (eq 10). In the C6 fraction, n-hexane (from
transfer hydrogenation) and internal hexenes (from alkene
isomerization) are observed. C10 alkenes are also observed
This shift in equilibrium between resting states is also
consistent with the observation that the transfer hydrogenation
G
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
dimerization catalyzed by 1 and transfer hydrogenation
catalyzed by 2 are major factors in the homogeneous tandem
catalytic coupling of 1-hexene and n-heptane. As for many other
one-pot dual-catalyst processes, the optimal conditions for the
individual catalysts do not necessarily translate into the optimal
conditions for the pair.9 While catalyst 1 operates most
efficiently at lower temperatures and high alkene loading,
catalyst 2 works best at high temperatures and low alkene
loading. Thus, in order to best match the rates of the two
individual reactions, tandem catalysis is effective at a moderate
temperature (100 °C) and an alkene concentration that is as
low as possible. In terms of simple batch reactions, our initial
optimization10 led to a 1-hexene loading of ∼250 mM (Table 4,
entry 1). Raising the substrate loading to 500 or 1000 mM
merely results in the production of more 1-hexene dimer, with
little additional tandem catalysis observed (entries 2 and 3);
this is due to the positive order dependence on [1-hexene] for
1 and the inverse order dependence on [1-hexene] for 2.
Conversely, raising the temperature to 125 or 150 °C
dramatically reduces catalyst cooperativity20 (entries 4 and 7),
due to the much stronger temperature dependence on rate
exhibited by 2 relative to 1.
rate is substantially affected by temperature, increasing nearly
10-fold from 100 to 125 °C (Table 3, entries 1 and 5). The
equilibrium in eq 11 is a 3-into-2 reaction, which means that
the reverse direction is entropically favored; furthermore, this
equilibrium is likely close to thermoneutral, since it is known
that subtle changes to ligand structure can favor F or G (or a
mixture) as the resting state of alkane metathesis.15o Thus, an
increase in temperature might be expected to shift the
equilibrium toward F, which is an on-cycle intermediate for
transfer hydrogenation. Furthermore, since G is also an active
catalyst for alkene isomerization,17 this shift should decrease the
rate of isomerization relative to transfer hydrogenation. This is
exactly the case, where with an initial [1-hexene] of 248 mM,
k
obs(hydrog)/kobs(isom) is 1.26 at 100 °C and 2.62 at 125 °C.
Of course, temperature effects on the other steps of both
catalytic cycles will influence these rates as well.
Notably, our observation that isomerization is positive order
in [1-hexene], determined from initial rate data (Table 3), is in
contrast to the zero-order dependence previously reported for
1-alkene isomerization catalyzed by 2, which was observed over
∼2 half-lives (single run, [1-alkene]0 = 100 mM; kinetics with
higher [1-alkene]0 were not reported).17 This reaction proceeds
by a π-allylic mechanism with a unimolecular turnover-limiting
step and thus should be zero order in [1-alkene]; however, the
positive order dependence observed here can be explained by
the equilibrium between resting states in eq 11. As the
concentration of 1-hexene is increased, the steady-state
concentration of G will increase, leading to a higher rate of
alkene isomerization. Thus, increasing [1-hexene] not only
inhibits transfer hydrogenation but also favors the undesirable
alkene isomerization pathway. The approach to saturation for
the rate of 1-hexene isomerization shown in Figure 7 is entirely
consistent with the proposed17 unimolecular turnover-limiting
step in the π-allylic isomerization catalytic cycle.
While increasing [1-hexene] or temperature results in a rate
imbalance between the two catalysts, it is possible that
increasing both of these variables would have complementary
effects, allowing a higher degree of tandem catalysis at increased
substrate loading. This is indeed the case, though the increase
in tandem catalysis is relatively modest. For example, the
amount of C13 + C14 products at an initial [1-hexene] of 1000
mM is 28 mM at 100 °C, whereas at 125 °C it increases to 46
mM, and at 150 °C it is 57 mM. Exploiting this complementary
temperature and concentration effect allows for higher tandem
TONs for 1 and 2 (7 and 14, respectively). Unfortunately, the
cooperativity factor decreases substantially with increasing
temperature, dropping to ∼25% at 150 °C. This is consistent
with an increase in the rate of alkene isomerization that still
outweighs the increase in dimerization and transfer hydro-
genation rates. Evidently, maintaining low [1-alkene] by slow
addition, as previously reported,10 is a more effective strategy to
enable a high degree of tandem catalysis.
Implications for Tandem Catalytic Alkane/Alkene
Styrene Dimerization and Styrene/1-Heptene Cou-
Coupling. The diametrically opposed kinetic trends in alkene
pling Catalyzed by 1. In the catalytic coupling of 1-hexene/
a
Table 4. Temperature and Concentration Effects on 1-Hexene/n-Heptane Coupling by 1 and 2
b
b
b
c
c
d
entry
[1-hexene]0 (mM)
temp (°C)
n-hexane (mM)
C12 (mM)
C13/C14 (mM)
TON for 1
TON for 2
coop. (%)
1
2
3
4
5
6
7
8
9
250
500
100
100
100
125
125
125
150
150
150
52
88
81
193
432
58
13/10
15/10
18/10
15/13
24/14
32/14
12/12
23/14
42/15
13 (3)
27 (3)
58 (4)
11 (4)
25 (5)
54 (6)
8 (3)
10 (6)
18 (7)
63
40
45
39
37
38
24
26
24
1000
250
87
17 (8)
106
140
161
147
196
296
21 (8)
500
162
386
39
28 (10)
32 (12)
29 (7)
1000
250
500
120
276
20 (5)
42 (7)
39 (10)
59 (14)
1000
a
b
See the Supporting Information for an expanded table. Determined by GC/FID using adamantane as an internal standard; average of at least two
c
d
runs. TONs in parentheses are for production of C13 + C14. Reference 20.
H
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
n-heptane discussed up to this point, competitive dimerization
of 1-hexene is prevalent; even with slow addition of 1-hexene
by syringe pump, the C12 fraction represents ∼50% of the
higher molecular weight products. In an effort to minimize this
side reaction, we initially examined TBE as a possible sacrificial
hydrogen acceptor that would not be incorporated into the
dimerization catalytic cycle, enabling the catalytic coupling of
alkanes.10 While we were able to achieve this with modest yield,
poor conversion and suspected catalyst decomposition led us to
consider other hydrogen acceptors. Styrene seemed to be an
attractive alternative, since competitive alkene isomerization
would not be a factor. Furthermore, the ethylbenzene
generated during transfer hydrogenation could, in principle,
be converted back to styrene by dehydrogenation (steam
cracking), which is practiced industrially on a large scale.21
In their initial studies of stoichiometric and catalytic alkene
dimerization by 1 and related complexes, Schrock et al. did not
report the dimerization of styrenes.13 While the
Cp*TaCl2(styrene) complex was prepared and characterized,
attempts to observe metallacycles derived from two styrene
units, or styrene and an aliphatic alkene, were unsuccessful.13c
These observations (or rather lack thereof) were encouraging
in the context of tandem catalysis, as competitive dimerization
of the hydrogen acceptor (i.e., styrene) should be minimal.
In order to test the ability of 1 to dimerize styrene, we carried
out control experiments under conditions similar to those used
in tandem catalytic 1-hexene/n-heptane coupling. Heating a
500 mM solution of styrene in C6D6 with 25 mM 1 to 110 °C
does in fact lead to dimerization, albeit slowly (t1/2 ≈ 14 h, eq
Figure 8. Plot of ln[styrene] versus time for the dimerization of
styrene (250 mM) catalyzed by 1 (8 mM) at 100 °C through 4 half-
lives (overlay of two runs, final 160 h data point not included in linear
correlation).
hydrogen acceptor in tandem catalysis; however, dehydrocou-
pling of styrene with the alkane is also possible. In order to
determine the relative rates of LAO dimerization and styrene/
LAO coupling, a mixture of styrene (313 mM) and 1-heptene
(258 mM) was converted with 1. Under initial rate conditions,
1-heptene dimerization is ∼4× faster than 1-heptene/styrene
coupling, while 1-heptene dimerization is ∼3× slower than 1-
hexene dimerization in the absence of styrene. Clearly, the
presence of styrene inhibits LAO dimerization, likely because
Cp*TaCl2(styrene) is a significant resting state.
Analysis of the GC traces of these cross-coupling experi-
ments indicated that there are four dominant cross-products,
with one major product.22 If both “head-to-tail” and “tail-to-
tail” couplings are considered, there are four possible
regioisomers (eq 13); assignment of these isomers has been
1
12). By H NMR spectroscopy and GC analysis, there is one
major product of this reaction, the “head-to-tail” dimer formed
in ∼80−90% yield (the balance is unreacted styrene and alkene
isomers of the dimers). Notably, there is a complete switch in
regioselectivity from reactions with LAOs: the “tail-to-tail”
product in styrene dimerization is formed in only ∼2−3% yield.
Monitoring the dimerization of styrene under conditions
analogous to tandem catalysis (250 mM styrene, 8 mM 1, 100
°C) reveals first-order dependence on [styrene] up to ∼3.5
half-lives, with t1/2 ≈ 40 h (Figure 8); deviation from linearity at
high conversion is likely due to product inhibition, as observed
for 1-hexene dimerization. This first-order dependence is
consistent with a mechanism analogous to the top catalytic
cycle in Scheme 2, where the equilibrium Cp*TaCl2(styrene) +
styrene ⇆ Cp*TaCl2(metallacycle) (with equilibrium constant
Keq″) heavily favors reactants; this is in accord with the inability
to observe metallacycles derived from styrenes.13c The observed
first-order rate constant kobs is equal to Keq″k3[Ta]0, and
therefore Keq″k3 = [6.00(6)] × 10−4 M−1 s−1 (where k3 is the
rate constant for the unimolecular decomposition of the styrene
metallacycle, analogous to k2 in Scheme 2). Comparing Keq″k3
to the corresponding values for 1-hexene dimerization to the
major and minor isomers gives a ratio of 53:17:1
(Keqk1:Keq′k2:Keq″k3), while comparing initial rates gives a
ratio of 40:13:1.
carried out by GC/MS fragmentation analysis and independent
synthesis.23 Heating a mixture of 1-heptene and styrene in n-
heptane (250 mM each) with 1 (8 mM) to 100 °C for 18 h
results in complete consumption of 1-heptene and ∼47%
conversion of styrene. The cross-products are formed in 23%
yield (based on 1-heptene): the two “tail-to-tail” isomers
comprise 2% each, and the “head-to-tail” isomers combined
account for 19%.
Styrene as a Sacrificial Hydrogen Acceptor in Tandem
Catalysis. Given the slow rate of styrene dimerization
observed, its use as a sacrificial hydrogen acceptor has been
examined in the tandem catalytic coupling of alkanes (Table 5).
Heating a 250 mM solution of styrene and precatalysts 1 and 2
(8 and 5 mM, respectively) in n-heptane to 100 °C for 18 h
results in the formation of three sets of higher molecular weight
products: styrene dimer, styrene/C7 cross-products, and C14
olefins (entry 1). Ethylbenzene is produced by transfer
hydrogenation between n-heptane and styrene, leading to a
The markedly slower dimerization of styrene relative to 1-
hexene means that styrene should be useful as a sacrificial
I
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
a
Table 5. Evaluation of Catalyst Cooperativity in Styrene/n-Heptane Coupling by 1 and 2
[styrene]0
(mM)
[1]/[2]
(mM)
time
(h)
temp
(°C)
conversn
(%)
EB
(mM)
styrene dimer
cross products
TON for TON for
coop.
d
b
b
b
b
c
c
entry
(mM)
(mM)
C14(mM)
1
2
(%)
1
2
3
4
5
6
7
8
9
10
250
250
250
250
250
250
250
1000
250
250
8/5
18
48
18
18
18
18
18
72
18
18
100
100
100
100
100
100
100
100
125
150
62
75
78
89
99
99
99
67
99
70
71
23
35
11
8
23
28
27
26
31
24
20
42
15
14
11
12
26
35
42
46
44
4
7 (4)
9 (5)
14 (9)
14 (10)
17 (10)
16 (10)
16 (11)
13 (8)
13 (7)
7 (3)
63
72
60
60
70
60
53
50
42
41
8/5
5/8
132
162
163
192
202
99
13 (11)
14 (12)
9 (7)
5/10
10/10
10/15
8/15
10/15
10/15
10/15
17
6
8 (7)
4
8 (8)
229
2
27 (5)
5 (5)
215
209
38
36
14 (6)
14 (6)
>99
2
5 (5)
a
b
See the Supporting Information for an expanded table. Determined by GC/FID using adamantane as an internal standard; average of at least two
c
d
runs. TONs in parentheses are for production of cross-products + C14. Reference 20.
⧫
■
Figure 9. Reaction progress for entry 7 (Table 5): (left) full plot; (right) expansion of product region. Legend: blue , styrene; red , ethylbenzene;
green , C14; purple , cross products; orange , styrene dimer. Lines are drawn as visual guides only.
◊
▲
○
catalyst cooperation factor of 63%. Under these conditions,
only 62% styrene conversion was observed, with the desired C14
alkenes as minor products. Heating this mixture for 48 h results
mainly in the production of more styrene dimer and cross
products, with almost no additional transfer hydrogenation
observed (entry 2), indicating decomposition of the iridium
catalyst.
In an effort to improve conversion and selectivity for the C14
products, the catalyst ratio was inverted to 5 mM 1 and 8 mM 2
(entry 3). After 18 h, 78% styrene conversion is reached, with
higher selectivity for tandem products. Increasing the loading of
2 gives the C14 alkenes as the major products, with 89% styrene
conversion and the highest TONs observed (entry 4). Further
increasing the catalyst loading gives full conversion and a good
degree of selectivity for C14 (entries 5 and 6). Catalyst loadings
of 10 and 15 mM for 1 and 2, respectively, result in the highest
overall yield of tandem products (70 mM, 56%), with C14
formed in 37% yield (46 mM) and the styrene dimer formed in
only 5% yield (6 mM) (entry 6). The ratio of tandem products
to styrene dimer can be increased to 16:1 with a minor
attenuation of catalyst loadings (entry 7); however, coopera-
tivity is slightly diminished (53%). Increasing the styrene
loading to 1000 mM results in mostly styrene dimerization,
analogous to the behavior observed for the 1-hexene/n-heptane
system. Unfortunately, attempts to use slow addition of styrene
to allow higher TONs have been unsuccessful, likely due to the
slower dimerization rates and probable catalyst decomposition.
Increasing the reaction temperature further increases the
selectivity for tandem catalysis over styrene dimerization to
∼26:1 (entries 9 and 10), albeit with lower yield (40−42%) and
cooperativity (41−42%).
In order to compare the features of this tandem reaction to
the 1-hexene/n-heptane system, the runs from entry 7 have
been monitored over time (Figure 9). The time course profiles
reveal steady consumption of styrene, with concomitant
production of ethylbenzene. In contrast to the 1-hexene/n-
heptane system, all three of the higher molecular weight
products are formed from the very beginning of the reaction,
J
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
with initial rates that are very similar despite the much lower
concentration of 1-heptene generated by transfer hydro-
genation. As the reaction progresses, styrene dimerization
ceases quickly while the cross products and C14 alkenes
continue to form until styrene is completely consumed.
As previously described, the tantalum and iridium
precatalysts operate independently in 1-hexene/n-heptane
coupling. In order to determine if this fact holds true with
styrene as the alkene partner, styrene/n-heptane transfer
hydrogenation catalyzed by 2 has been monitored and the
time profile compared to the tandem reaction (Figure 10).
moderate reaction temperature and slow addition of substrate is
crucial to achieving high levels of catalyst cooperation.10
Because the Ta/Ir system operates using sacrificial hydrogen
acceptors in transfer hydrogenation, styrene has been employed
as a more effective acceptor for alkane/alkene dehydrocoupling
and alkane double dehydrocoupling. Tantalum-catalyzed
dimerization of styrene is slow relative to styrene/1-alkene
coupling and 1-alkene dimerization, which enables high
selectivity for tandem catalytic products (up to 26:1 for tandem
catalysis versus styrene dimerization). Unfortunately, this
system appears to undergo catalyst decomposition during the
long reaction times required to reach full conversion. In
contrast to the 1-hexene/n-heptane reactions, there is a slight
inhibition of transfer hydrogenation by the addition of the
tantalum catalyst, which suggests the possibility of catalyst−
catalyst interactions under these conditions.
Finally, while the current dual homogeneous system can
operate in tandem with a high degree of cooperativity, it cannot
achieve the ideal, byproduct-free upgrading process from
Scheme 1. The central challenge to realizing the envisioned
scheme is effecting transfer hydrogenation between the
products of alkene dimerization and the alkane feed. Because
the products formed by tantalum catalyst 1 are highly branched,
transfer hydrogenation is both kinetically and thermodynami-
cally unfavorable. Thus, alternate systems for alkene dimeriza-
tion that give linear products are required.12 Efforts are
currently underway to incorporate such catalysts into a viable
tandem process, as well as to develop heterogeneous dual
catalysts for converting gaseous feedstocks.
Figure 10. Comparison of transfer hydrogenation progress in the
conversion of styrene (blue diamonds) to ethylbenzene (red squares)
for catalysis with 2 only (hollow points, dashed lines) and with 1 and 2
in tandem (filled points, solid lines; from reaction in entry 7, Table 5;
values corrected for styrene converted to cross products and styrene
dimer, see the Supporting Information for more details).
EXPERIMENTAL SECTION
■
General Considerations. Unless otherwise noted, all experiments
were performed under an argon inert atmosphere using standard
Schlenk-line, high-vacuum-line, or glovebox techniques. Transfer
hydrogenation catalysis by many pincer-Ir complexes is known to be
inhibited by N2.15c
Unlike the case for the 1-hexene/n-heptane system, there
appears to be a slight decrease (factor of ∼1.5) in the initial rate
of transfer hydrogenation during tandem catalysis; however,
both reactions reach completion in virtually the same length of
time. Unfortunately, due to the slow rate of styrene
dimerization and competing cross-product and C14 formation,
a direct comparison between individual and tandem reactions
with precatalyst 1 cannot be made. Nevertheless, it is possible
that there is a minor, reversible pathway through which the
tantalum and iridium catalysts could undergo mutual inhibition
in the styrene/n-heptane system.
Materials. Solvents for routine syntheses (pentane, toluene, diethyl
ether, THF) were dried by passage through activated alumina,
degassed under vacuum by several freeze−pump−thaw cycles, and
stored over activated 4 Å molecular sieves under an inert atmosphere.
n-Heptane for use in catalytic reactions (HPLC grade, >99%, Sigma-
Aldrich) was predried by stirring ∼400 mL over CaH2 (∼10 g) for at
least 48 h. The solvent was then vacuum-transferred onto
“titanocene”24 (∼1 g) and stirred overnight; the solution remained
black-green throughout. The n-heptane was collected from this
titanocene solution by a final vacuum transfer and stored under an
argon atmosphere. n-Decane for use in kinetics experiments
(anhydrous, >99%, Sigma-Aldrich) was stored over activated 4 Å
molecular sieves under an argon atmosphere, and aliquots were filtered
through a small column of activated alumina immediately prior to use
in making standard solutions. 1-Hexene and 1-heptene were distilled
under argon from CaH2 after stirring for several days or vacuum-
transferred from LiAlH4 after sitting for several months. Styrene
(>99.9%, 10−15 ppm 4-tert-butylcatechol as inhibitor, Sigma-Aldrich)
was stored under an argon atmosphere at −35 °C and filtered through
a small column of alumina immediately prior to use in catalytic
reactions in order to remove the inhibitor and trace water. Precatalysts
113c and 225 are known compounds and were prepared according to
published procedures.
CONCLUSIONS AND OUTLOOK
■
In summary, we have determined the mechanistic features of
the homogeneous tandem catalytic coupling of alkanes and
alkenes using sacrificial hydrogen acceptors for the combined
Cp*TaCl2(alkene)/[PCP]IrH2 system. These two catalysts
operate independently under tandem catalysis conditions in the
1-hexene/n-heptane reaction, with no kinetically relevant
interactions observed. While the rate of tantalum-catalyzed
alkene dimerization is positive order in 1-alkene and is relatively
insensitive to increasing temperature, iridium-catalyzed transfer
hydrogenation is inverse order in 1-alkene and is accelerated
substantially by increasing temperature. These opposing
mechanistic features highlight the difficulty in achieving
efficient tandem catalysis with two closely linked catalytic
cycles: the rates of the individual processes must be well-
matched, but each catalyst may operate most effectively under
very different reaction conditions.9 In the present system,
General Procedure for Kinetics Experiments. The following
description for the determination of initial rates of 1-hexene
dimerization is representative (see the Supporting Information for
detailed experimental procedures for all data collected).
Precatalyst 1 (10.0 mg, 0.0240 mmol) was dissolved in 3 mL of a
standard solution of 1-hexene (248, 252, 412, 699, 1010, 2000, or 4000
mM) and adamantane internal standard (∼25 mM) in n-decane in a 4
K
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
mL screw-top vial. The vial was sealed with a Teflon-lined screw cap.
The mixture was heated briefly with a heat gun and stirred vigorously
to dissolve the precatalyst and ensure a homogeneous solution. The
solution was then split into 15 aliquots of 0.2 mL each in 15 separate 4
mL vials containing stir bars. These vials were sealed and stirred at 80,
90, 100, 110, or 125 °C in an aluminum block heater. At specified time
intervals, (1, 2, 3, or 4 min) vials were removed from the heat block
and immersed in a dry ice/acetone bath to rapidly cool the contents,
and the contents were then diluted with dichloromethane to quench
the reaction. These solutions were passed through a short plug of silica
gel into a GC autosampler vial and analyzed by GC.
Data for each time point were collected from 3 or 4 different vials,
giving 12−15 data points for each initial rate determination. The
concentrations of both major and minor product isomers were plotted
versus time, and the kobs values were calculated by linear regression
analysis (Table S1 and Figures S4, S6, S8, S10, and S12 (Supporting
Information)). These data were used to generate double-reciprocal
plots for each product isomer at each temperature (Figures S5, S7, S9,
S11, and S13 (Supporting Information)). These plots were subject to
linear regression analysis to generate slope/intercept values, which
were used to calculate k1, Keq, and k2Keq′ according to eqs 2−9.
General Procedure for Tandem Catalytic Reactions. The
following is a representative example (entry 1, Table 4). 1 (6.6 mg,
0.016 mmol) and 2 (5.9 mg, 0.010 mmol) were dissolved in 2 mL of a
standard solution of 1-hexene (250 mM) and adamantane (25 mM) in
n-heptane in a 4 mL screw-top vial containing a Teflon-coated stir bar.
The vial was sealed with a Teflon-lined screw cap. The mixture was
stirred at 100 °C in an aluminum block heater for 5 h. After it was
cooled to room temperature, the brown solution was diluted with n-
heptane and passed through a short plug of silica gel into a GC
autosampler vial and analyzed by GC.
Procedure for Monitoring Dimerization of Styrene over
Time. Precatalyst 1 (6.6 mg, 0.016 mmol), styrene (53.0 mg, 0.509
mmol), and 1,3,5-trimethoxybenzene (internal standard, 1.6 mg) were
dissolved in C6D6 (2 mL). This solution was then split evenly into four
J. Young NMR tubes. Initial 1H NMR spectra were recorded, and then
the tubes were heated to either 100 °C (two tubes) or 125 °C (two
tubes) in an oil bath. After set time intervals, the tubes were removed
from the oil baths and new spectra acquired until ∼4 half-lives had
passed. The three aromatic protons of the 1,3,5-trimethoxybenzene
internal standard (∼6.25 ppm) were integrated versus the terminal
vinylic proton of styrene that is cis relative to the phenyl group (∼5.6
ppm) to generate concentration versus time plots (see the Supporting
Information for more details).
Procedure for Styrene/1-Heptene Coupling. Precatalyst 1
(20.4 mg, 0.050 mmol), styrene (52.1 mg, 0.500 mmol), 1-heptene
(49.0 mg, 0.499 mmol), and adamantane (internal standard, 10.2 mg,
0.075 mmol) were dissolved in 2 mL of n-heptane in a 4 mL screw-top
vial containing a Teflon-coated stir bar. The vial was sealed with a
Teflon-lined screw cap, and the contents were stirred at 100 °C in an
aluminum block heater. After 18 h, the vial was removed and the
contents were diluted with dichloromethane to a total volume of ∼4
mL. This solution was passed through a short plug of silica gel before
analysis by GC and GC/MS.
Notes
The authors declare the following competing financial
interest(s): a patent application has been filed based partially
on this work.
ACKNOWLEDGMENTS
■
This work was funded by BP through the XC2 program. D.C.L.
thanks the NSERC for a PDF.
REFERENCES
■
(1) (a) Annual Energy Outlook 2013, U.S. Energy Information
Administration (EIA), http://www.eia.gov/forecasts/aeo/index.cfm.
(b) Mango, F. D. Org. Geochem. 1997, 26, 417.
(2) (a) Ancheyta, J.; Rana, M. S.; Furimsky, E. Catal. Today 2005,
109, 3.
(3) Natural gas production, transmission, and consumption 2011, U.S.
EIA, http://www.eia.gov/naturalgas/annual/pdf/table_002.pdf.
(4) (a) Huber, G. W.; Chheda, J. N.; Barrett, C. J.; Dumesic, J. A.
Science 2005, 308, 1446. (b) Huber, G. W.; Iborra, S.; Corma, A. Chem.
Rev. 2006, 106, 4044. (c) Schlaf, M. Dalton Trans. 2006, 4645.
(5) (a) Sadenghbeigi, R. Fluid Catalytic Cracking Handbook, 3rd ed.;
Elsevier: New York, 2012. (b) Speight, J. G. The Chemistry and
Technology of Petroleum, 4th ed.; CRC Press: Boca Raton, FL, 2006;
Chapter 18.
(6) de Klerk, A. Fischer-Tropsch Refining; Wiley-VCH: Weinheim,
Germany, 2011.
(7) (a) Basset, J.-M.; Coperet, C.; Soulivong, D.; Taoufik, M.; Cazat,
́
J. T. Acc. Chem. Res. 2010, 43, 323. (b) Haibach, M. C.; Kundu, S.;
Brookhart, M.; Goldman, A. S. Acc. Chem. Res. 2012, 45, 947.
(c) Chen, C. Y.; O’Rear, D. J.; Leung, P. Top. Catal. 2012, 55, 1344.
(8) For examples of alkane metathesis that do exhibit selectivity for
C2n−2 products, see: (a) Goldman, A. S.; Roy, A. H.; Huang, Z.; Ahuja,
R.; Schinski, W.; Brookhart, M. Science 2006, 312, 257. (b) Bailey, B.
C.; Schrock, R. R.; Kundu, S.; Goldman, A. S.; Huang, Z.; Brookhart,
M. Organometallics 2009, 28, 355. For an example of selective alkane
cross metathesis, see: (c) Dobereiner, G. E.; Yuan, J.; Schrock, R. R.;
Goldman, A. S.; Hackenberg, J. D. J. Am. Chem. Soc. 2013, 135, 12572.
(9) (a) Fogg, D. E.; dos Santos, E. N. Coord. Chem. Rev. 2004, 248,
2365. (b) Wasilke, J.-C.; Obrey, S. J.; Baker, R. T.; Bazan, G. C. Chem.
Rev. 2005, 105, 1001.
(10) Leitch, D. C.; Lam, Y. C.; Labinger, J. A.; Bercaw, J. E. J. Am.
Chem. Soc. 2013, 135, 10302.
(11) (a) Scurrell, M. S. Appl. Catal. 1987, 32, 1. (b) Albright, L. F.
Ind. Eng. Chem. Res. 2009, 48, 1409. (c) For an organometallic
approach to alkane/alkene coupling with a single catalyst, see: Sadow,
A.; Tilley, T. D. J. Am. Chem. Soc. 2003, 125, 7971.
(12) Several transition-metal catalysts can dimerize 1-alkenes to
linear alkenes with up to 97% selectivity. See, for example: (a) Keim,
W.; Hoffmann, B.; Lodewick, R.; Peukert, M.; Schmitt, G.; Fleischauer,
J.; Meier, U. J. Mol. Catal. 1979, 6, 79. (b) Small, B. L.; Marcucci, A. J.
Organometallics 2001, 20, 5738. (c) Small, B. L. Organometallics 2003,
22, 3178. (d) Small, B. L.; Schmidt, R. Chem. Eur. J. 2004, 10, 1014.
(e) Broene, R. D.; Brookhart, M.; Lamanna, W. M.; Volpe, A. F. J. Am.
Chem. Soc. 2005, 127, 17194.
(13) (a) McLain, S. J.; Schrock, R. R. J. Am. Chem. Soc. 1978, 100,
1315. (b) McLain, S. J.; Sancho, J.; Schrock, R. R. J. Am. Chem. Soc.
1979, 101, 5451. (c) McLain, S. J.; Wood, C. D.; Schrock, R. R. J. Am.
Chem. Soc. 1979, 101, 4558. (d) McLain, S. J.; Sancho, J.; Schrock, R.
R. J. Am. Chem. Soc. 1980, 102, 5610.
ASSOCIATED CONTENT
■
S
* Supporting Information
Text giving full experimental procedures, tables giving
expanded data, and figures giving full kinetic data and plots,
spectroscopic data, and representative GC traces. This material
is available free of charge via the Internet at http://pubs.acs.org.
(14) Choi, J.; MacArthur, A. H. R.; Brookhart, M.; Goldman, A. S.
Chem. Rev. 2011, 111, 1761.
(15) (a) Gupta, M.; Hagen, C.; Kaska, W. C.; Flesher, R.; Jensen, C.
M. Chem. Commun. 1996, 2083. (b) Xu, W.-W.; Rosini, G. P.; Krogh-
Jespersen, K.; Goldman, A. S.; Gupta, M.; Jensen, C. M.; Kaska, W. C.
Chem. Commun. 1997, 2273. (c) Jensen, C. M. Chem. Commun. 1999,
2443. (d) Liu, F.; Pak, E. B.; Singh, B.; Jensen, C. M.; Goldman, A. S. J.
Am. Chem. Soc. 1999, 121, 4086. (e) Liu, F.; Goldman, A. S. Chem.
Commun. 1999, 655. (f) Krogh-Jespersen, K.; Czerw, M.; Zhu, K.;
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail for J.A.L.: jal@caltech.edu.
*E-mail for J.E.B.: bercaw@caltech.edu.
L
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Singh, B.; Kanzelberger, M.; Darji, N.; Achord, P. D.; Renkema, K. B.;
Goldman, A. S. J. Am. Chem. Soc. 2002, 124, 10797. (g) Krogh-
Jespersen, K.; Czerw, M.; Summa, N.; Renkema, K. B.; Achord, P. D.;
Goldman, A. S. J. Am. Chem. Soc. 2002, 124, 11404. (h) Zhu, K.;
Achord, P. D.; Zhang, X.; Krogh-Jespersen, K.; Goldman, A. S. J. Am.
Chem. Soc. 2004, 126, 13044. (i) Gottker-Schnetmann, I.; Brookhart,
̈
M. J. Am. Chem. Soc. 2004, 126, 1766. (j) Gottker-Schnetmann, I.;
̈
Brookhart, M. J. Am. Chem. Soc. 2004, 126, 1804. (k) Kundu, S.;
Choliy, Y.; Zhuo, G.; Ahuja, R.; Emge, T. J.; Warmuth, R.; Brookhart,
M.; Krogh-Jespersen, K.; Goldman, A. S. Organometallics 2009, 28,
5432. (l) Huang, Z.; Brookhart, M.; Goldman, A. S.; Kundu, S.; Ray,
A.; Scott, S. L.; Vicente, B. C. Adv. Synth. Catal. 2009, 351, 188.
(m) Punji, B.; Emge, T. J.; Goldman, A. S. Organometallics 2010, 29,
2702. (n) Ahuja, R.; Punji, B.; Findlater, M.; Supplee, C.; Schinski, W.;
Brookhart, M.; Goldman, A. S. Nat. Chem. 2011, 3, 167. (o) Nawara-
Hultzsch, A. J.; Hackenberg, J. D.; Punji, B.; Supplee, C.; Emge, T. J.;
Bailey, B. C.; Schrock, R. R.; Brookhart, M.; Goldman, A. S. ACS Catal.
2013, 3, 2505.
(16) Renkema, K. B.; Kissin, Y. V.; Goldman, A. S. J. Am. Chem. Soc.
2003, 125, 7770.
(17) Biswas, S.; Huang, Z.; Choliy, Y.; Wang, D. Y.; Brookhart, M.;
Krogh-Jespersen, K.; Goldman, A. S. J. Am. Chem. Soc. 2012, 134,
13276.
(18) See the Supporting Information for more details on the
1
estimation of Keq values by H NMR spectroscopy.
(19) (a) The data were fitted using the freeware DynaFit kinetics
program distributed by BioKin: http://www.biokin.com/dynafit/. See
the Supporting Information for more details. (b) Kuzmic, P. Anal.
Biochem. 1996, 237, 260.
(20) Here, we define “cooperativity” as the amount of heptene
generated by dehydrogenation that is incorporated into C13 or C14
dimers. Since the [1-heptene] produced is equal to the [n-hexane]
observed, the cooperativity factor is ([C13] + 2[C14])/[n-hexane]. For
the reactions involving styrene, cooperativity is equal to ([cross-
products] + 2[C14])/[ethylbenzene].
(21) James, D. H.; Castor, W. M. Styrene. In Ullmann’s Encyclopedia
of Industrial Chemistry; Wiley: New York, 2011.
(22) There is also a minor observed product resulting from styrene/
1-nonane coupling (identified by GC/MS). The 1-nonane is formed
by the coupling of ethylene (from the Ta precatalyst) and 1-heptene.
See the Supporting Information for more details.
(23) See the Supporting Information for more details on the isomer
assignments.
(24) (a) Bercaw, J. E.; Marvich, R. H.; Bell, L. G.; Brintzinger, H. H.
J. Am. Chem. Soc. 1972, 94, 1219. (b) Burger, B. J.; Bercaw, J. E. In
Experimental Organometallic Chemistry; Wayda, A. I., Darensbourg, M.
Y., Eds.; American Chemical Society: Washington, DC, 1987; ACS
Symposium Series 357, p 79.
(25) Goldman, A. S.; Ghosh, R. In Handbook of C-H Transformations-
Applications in Organic Synthesis; Dyker, G., Ed.; Wiley-VCH: New
York, 2005; p 616.
M
dx.doi.org/10.1021/om500231t | Organometallics XXXX, XXX, XXX−XXX