Selective Decarbonylation of Fatty Acid Esters to Linear α-Olefins

Article abstract of DOI:10.1021/acs.organomet.7b00411

Selective decarbonylation of p-nitrophenol esters of fatty acids to the corresponding linear α-olefins (LAOs) was achieved using palladium catalysis. After extensive ligand screening, a mixed-ligand system exploiting the trans-spanning diphosphine XantPhos and an N-heterocyclic carbene (IPr) was identified as the most effective in yielding high α-selectivity and high conversions of the ester (>98% selectivity, >90% conversion using 2.5 mol % of PdCl₂ and 5 mol % of the ligands, 190 °C, 2-2.5 h). On the basis of insights from modeling at the density functional level of theory, we propose that the mixed-ligand set achieves high α-selectivity by promoting olefin dissociation from the palladium center following β-hydride elimination, which is especially facilitated both by the combined steric bias of the mixed-ligand set and by the ability of the XantPhos ligand to coordinate in both mono- and bidentate fashions.

Full text of DOI:10.1021/acs.organomet.7b00411

Article
pubs.acs.org/Organometallics
Selective Decarbonylation of Fatty Acid Esters to Linear α‑Olefins
Alex John,† Busra Dereli,†,‡ Manuel A. Ortuno,†,‡ Hillis E. Johnson,† Marc A. Hillmyer,†
̈
̃
Christopher J. Cramer,*,†,‡ and William B. Tolman*,†
‡
†Department of Chemistry & Center for Sustainable Polymers and Department of Chemistry, Supercomputing Institute, and
Chemical Theory Center, University of Minnesota, 207 Pleasant Street SE, Minneapolis, Minnesota 55455-0431, United States
S
* Supporting Information
ABSTRACT: Selective decarbonylation of p-nitrophenol esters of fatty acids to
the corresponding linear α-olefins (LAOs) was achieved using palladium
catalysis. After extensive ligand screening, a mixed-ligand system exploiting the
trans-spanning diphosphine XantPhos and an N-heterocyclic carbene (IPr) was
identified as the most effective in yielding high α-selectivity and high
conversions of the ester (>98% selectivity, >90% conversion using 2.5 mol %
of PdCl2 and 5 mol % of the ligands, 190 °C, 2−2.5 h). On the basis of insights
from modeling at the density functional level of theory, we propose that the
mixed-ligand set achieves high α-selectivity by promoting olefin dissociation
from the palladium center following β-hydride elimination, which is especially facilitated both by the combined steric bias of the
mixed-ligand set and by the ability of the XantPhos ligand to coordinate in both mono- and bidentate fashions.
carbon atoms, their decarbonylation would yield LAOs bearing
an odd-number carbon skeleton, expensive4c variants that are
complementary to those obtained via traditional routes.
Most of the recent strategies focused on decarbonylation of
carboxylic acids, including fatty acids, require stoichiometric or
greater quantities of a sacrificial anhydride to activate them.2,7
As an alternative strategy, we recently reported a palladium-
catalyzed decarbonylation of p-nitrophenol esters11 that built
upon a previous report of a Heck reaction of such esters.12 The
decarbonylation was promoted by the presence of alkali-/
alkaline-earth-metal halides and exhibited a broad substrate
scope, including fatty acid esters. Low α-selectivity was
observed in these reactions, however. In view of the potential
for such reactions to provide an alternate, sustainable route to
commodity olefins such as LAOs, we have focused on
enhancing the α-selectivity.
INTRODUCTION
■
Dehydrative decarbonylation of aliphatic carboxylic acids to the
corresponding olefins featuring one less carbon atom holds
promise for converting bioderived molecules to commodity
chemicals.1 Several catalytic systems based on Rh,2 Ir,3 Pd,4 and
to a lesser extent base metals such as Ni5 and Fe,6 have shown
promise in effecting such decarbonylation reactions. In
particular, we and others have shown that this strategy is
useful for accessing polymerizable olefins such as styrene,
acrylates, and acrylonitrile4d,7 starting from hydrocinnamic acid,
alkyl succinates, and 3-cyanopropanoic acid (derived from
glutamic acid via oxidative decarboxylation), respectively. Fatty
acids, which are abundantly available from vegetable oils and
animal fats, represent a particularly valuable substrate class for
this reaction, as their decarbonylation could selectively give
access to linear α-olefins (LAOs, Scheme 1).2,8 LAOs find
Scheme 1. Dehydrative Decarbonylation of Fatty Acids
We report here the development of a highly selective,
palladium-catalyzed decarbonylation of p-nitrophenol esters of
fatty acids to their corresponding LAOs. A dual ligand system
involving XantPhos (4,5-bis(diphenylphosphino)-9,9-dimethyl-
xanthene) and IPr (1,3-bis(2,6-diisopropylphenyl)imidazol-2-
ylidene) was found to be optimal for achieving high α-
selectivity (up to >98%) together with high conversion (>90%)
over short reaction times (2 h). Combined experimental and
theoretical studies have unraveled the synergistic roles of the
two added ligands, suggesting that XantPhos dictates the α-
selectivity while IPr promotes high conversions at short
reaction times.
utility in a plethora of industrial applications, as exemplified by
their direct use as comonomers (C4−C8) for the production of
linear low-density polyethylene (LLDPE) and as key precursors
in the synthesis of surfactants, detergents, lubricants, and
plasticizers.9 Current commercial production of LAOs relies on
either ethylene oligomerization or Fischer−Tropsch processes
starting from syngas.10 Fatty acids, on the other hand, being
naturally available and renewable, offer a sustainable resource
for the generation of this important class of chemicals. In
addition, because natural fatty acids feature an even number of
Received: June 5, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
RESULTS AND DISCUSSION
■
Previously, we reported the decarbonylation of p-nitrophenyl
nonanoate (1), an illustrative fatty acid ester, using PdCl2 (5
mol %) and LiCl (100 mol %) in DMPU (N,N′-
dimethylpropyleneurea) at 160 °C in a sealed tube, and
observed 35% conversion to a mixture of octenes (14% 1-
octene).11,13 Working under the assumption that internal
octene isomers were generated by recoordination of 1-octene
to the palladium center and subsequent isomerization, we
hypothesized that distilling 1-octene from the reaction mixture
as it was formed would improve the α-selectivity of the
reaction. However, performing the same reaction under
continuous distillation conditions did not result in significantly
improved 1-octene selectivity (14%), suggesting that the olefin
is not readily released from the palladium center in the first
place.
Surmising that added ligands would facilitate release of 1-
octene from the metal site, we performed a preliminary survey
(Table 1) and found that phosphine ligands were effective, with
Figure 1. Overlay of GC-MS traces for decarbonylation of 1 to 1-
octene (asterisks denote a retention time of 3.56 min) and its isomers
(all other peaks): (a) PdCl2, no ligand; (b) PdCl2 (5 mol
%)/XantPhos (5 mol %) (Table 1, entry 12); (c) PdCl2 (2.5 mol
%)/XantPhos (5 mol %)/IPr (5 mol %) (Table 3, entry 1).
Table 1. Ligand Effects on Decarbonylation of p-Nitrophenyl
Nonanoate
a
substrate, p-nitrophenyl 4-phenylbutanoate (2), that would be
particularly sensitive to isomerization, as the resulting internal
alkene is benzylic in nature.
In the absence of any added ligands or in the presence of
XantPhos, decarbonylation of 2 using PdCl2 at 190 °C almost
exclusively gave β-methylstyrene (2a′, >90%; Table 2 and
b
c
conversion
(%)
α-selectivity
entry
ligand (mol %)
dba (10)
T (οC)
(1-octene %)
d
1
160
160
160
160
>95
>95
<20
<20
14
14
15
15
e
2
3
4
N-Cbz-L-alanine
(10)
Table 2. Ligand Effects on Decarbonylation of p-Nitrophenyl
a
4-Phenylbutanoate
5
IPr (10)
160
160
160
160
190
190
190
190
<20
<20
<20
<20
>95
>95
>95
>95
20
33
61
70
46
35
45
82
6
PPh3 (10)
7
PPh3 (20)
8
DPEPhos (5)
PPh3 (20)
9
10
11
DPEPhos (5)
DPEPhos (10)
b
entry
PdCl2/XantPhos/PPh3 (mol %)
α-selectivity (allylbenzene %)
1
2
3
4
5
6
5/0/0
<5
<5
54
63
28
92
<5
12 XantPhos (5)
10/10/0
10/20/0
5/20/0
5/5/10
5/10/10
5/10/10
a
Reaction conditions unless specified otherwise: a mixture of the ester
(1 equiv), PdCl2 (5 mol %), ligand (5−20 mol %), and LiCl (1 equiv)
in DMPU (ca. 0.5 mL) was heated at 160 or 190 °C for ∼16 h.
b
c
Determined by GC-MS, conversion of ester. Determined by GC-
d
MS; 1-octene vs sum of internal octenes. Reaction carried out in a
closed reactor over 16 h. Reaction carried out in continuous
e
c
7
a
distillation mode for 16 h.
Reaction conditions unless specified otherwise: a mixture of the ester
(1 equiv), PdCl2 (5 mol %), ligand(s) (5−20 mol %), and LiCl (1
equiv) in DMPU (ca. 0.5 mL) was heated at 190 °C overnight (16 h).
up to 70% 1-octene (selectivity) observed with DPEPhos (5
mol %; entries 6−8).14 Unfortunately, the overall conversion at
nearly constant reaction time was found to be significantly
lower (∼20% conversion of ester) in the presence of added
phosphine ligands. Conversions could be enhanced by
increasing the reaction temperature to 190 °C, but at the
expense of 1-octene selectivity (entries 9−11). However,
switching from DPEPhos to its more rigid analogue XantPhos,
shown previously to induce good α-selectivities,4c resulted in a
high proportion of 1-octene (82%) while preserving the high
conversion of the ester (>95%, entry 12, Figure 1b).
Encouraged by this finding, we sought to further improve the
α-selectivity and turned to a test system with a more reactive
b
Determined by GC-MS; allylbenzene vs sum of internal olefins.
Reaction carried out in the absence of LiCl.
c
Figure S1a in the Supporting Information) and very small
amounts of allylbenzene (2a, <5%), suggesting that olefin
isomerization is highly favorable in this case. Finding that
increasing the XantPhos:PdCl2 ratio resulted in improved
selectivity for allylbenzene (54% (2:1), 63% (4:1)), but wishing
to avoid use of large excesses of the expensive XantPhos, we
tested mixtures of XantPhos and PPh3 and found that with
XantPhos (10%)/PPh3 (10%) the selectivity rose significantly
in favor of allylbenzene (92%; Table 2 and Figure S1b). The
B
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
a
Table 3. Decarbonylation of Fatty Acid Esters
b
entry
ester
product
1-octene
α-selectivity (%)
1
2
3
4
5
6
7
8
nonanoic (C9)
4-phenylbutyric
>98
>98
>98
>98
>98
>98
>98
86
allylbenzene
undecylenic (C11)
lauric (C12)
1,9-decadiene
1-undecene
myristic (C14)
palmitic (C16)
stearic (C18)
1-tridecene
1-pentadecene
1-heptadecene
1,9(Z)-heptadecadiene
oleic (C18-9Z)
a
Reaction conditions: a mixture of the ester (1 equiv), PdCl2 (2.5 mol %), XantPhos (5 mol %), IPr (5 mol %), and LiCl (20 mol %) in DMPU (ca.
b
1
0.5 mL) was heated at 190 °C for 2.5 h. Determined by H NMR and confirmed by GC-MS. Conversion of the ester is 85−90% at this point.
LiCl additive was also found to be crucial in curbing
isomerization, as a reaction carried out in its absence resulted
in complete loss of α-selectivity (<5% allylbenzene).
Table 4. Role of XantPhos and IPr in the Dual Ligand
System
a
Employing the dual-ligand PdCl2/XantPhos/PPh3 system in
the decarbonylation of 1 resulted in almost exclusive selectivity
for 1-octene (>98%). However, upon extending the dual ligand
system to heavier fatty acid esters such as stearic (C18) and
those featuring an unsaturated carbon chain such as
undecylenic acid (C11) or oleic acid (9-Z-C18), lower α-
selectivities were observed (30 and 11%, respectively). The
general utility of N-heterocyclic carbenes (NHCs) in organo-
metallic catalysis15 prompted us to explore their efficacy as a
coligand (substitute for PPh3). Significantly, when 1,3-bis(2,6-
diisopropylphenyl)imidazol-2-ylidene (IPr) was used in con-
junction with XantPhos, with a relatively low loading of LiCl
(20%) and with the reactions stopped at 85−90% conversion
(∼2.5 h), very high α-selectivities for a wide range of fatty acid
substrates were observed (Table 3). We did note that oleic acid,
which contains an internal double bond, showed some
isomerization (86% selectivity).16 We were also gratified to
see very high α-selectivity using this system for the decarbon-
ylation of 2 (>98% allylbenzene, Figure S1c in the Supporting
Information).
In order to understand the respective roles of the two ligands
needed to achieve the high α-selectivities for the wide range of
substrates, we systematically varied their relative amounts in the
decarbonylation of palmitic ester (6, Table 4). Using IPr (5 mol
%) in the absence of XantPhos resulted in high conversion but
almost negligible α-selectivity. Conversely, reactions carried out
in the absence of IPr exhibited α-selectivity that scaled with
XantPhos loading, reaching excellent selectivity (>98%) at a
XantPhos:PdCl2 loading of 2:1. However, the overall ester
conversion was observed to be low under these conditions
(46%). The inclusion of IPr (2.5 or 5 mol %) into the PdCl2
(2.5 mol %)/XantPhos (5 mol %) system resulted in superior
conversion of the ester (∼90%) within a 2 h reaction time. On
the basis of these observations, we hypothesized that the
XantPhos ligand controls the α selectivity in these decarbon-
ylation reactions by promoting olefin dissociation after the β-
hydride elimination step, thus lessening any chain-walking
processes involving recoordination of the α-olefin product. On
the other hand, we surmised that the IPr ligand helps maintain
high conversions by facilitating generation of a low-coordinate
entry PdCl2/XantPhos/IPr (mol %) conversion (%) selectivity (% α)
1
2
3
4
5
6
2.5/0/5
95
68
<5
79
2.5/2.5/0
2.5/5/0
46
>98
95
2.5/5/2.5
2.5/5/5
84
>90
71
>98
>98
1.25/2.5/2.5
a
Reaction conditions: ester (0.085 g), LiCl (0.010 g), and the catalyst
mix in DMPU (ca. 0.5 mL) was heated at 190 °C for 2 h. Conversion
1
of the ester and selectivity for α-olefin were determined by H NMR
analysis.
Pd species that performs the initial oxidative addition step, or
otherwise stabilizing other Pd intermediates. To test these
notions, we evaluated relevant reaction pathways using theory.
Density Functional Modeling. To gain mechanistic
insight at the atomic level of detail, we performed density
functional theory (DFT) calculations17 at the M06-L level (see
Computational Details). We focused on the Pd-catalyzed
decarbonylation of 2 in order (i) to fully characterize reaction
coordinates for the formation of terminal 2a and internal 2a′
alkene products and (ii) to explain the roles of the ligands in
promoting α-selectivity.
18
We assume formation of Pd(0) from reduction of PdCl2
and first examine reaction mechanisms for homoligated
catalysts [Pd(IPr)2] and [Pd(XantPhos)2]. Unless otherwise
noted, all ΔG values are 463.15 K free energies (kcal mol−1) in
N,N-dimethylacetamide (ε = 37.781), which we take as an
analogue of DMPU (ε = 36.1219).
Reaction Mechanisms. Scheme 2 illustrates a plausible
mechanism for the Pd-catalyzed decarbonylation of ester 2 to
form either allylbenzene 2a (cycle I) or β-methylstyrene 2a′
(cycle II), where L = IPr, XantPhos.4,5,20 Starting from PdL2 A,
2 is added oxidatively to the Pd center, forming the Pd−acyl
species B. Subsequent acyl decarbonylation gives rise to the
C
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
generically, we use the prefix L to avoid confusion between
computed structures and experimental compounds discussed
above. The numbers in a name L-n reflect chronological
occurrence along overall reaction coordinates as illustrated in
the following figures. When n is preceded by “TS”, this
indicates the transition-state structure leading from structure n
to subsequent intermediate n + 1.
Scheme 2. Reaction Mechanisms Leading to Terminal (2a;
Cycle I) and Internal (2a′; Cycle II) Olefin Products
a
Reaction Coordinates. We first computed the reaction
coordinates for the formation of 2a (cycle I in Scheme 2) with
the ligands IPr and XantPhos. The results are summarized in
Figure 2 (see also Figures S4 and S5 in the Supporting
Information for detailed discussion and structures). The bis-
ligated complexes PdL2 L-1 and the substrate 2 are taken as the
zeroes of energies for their respective coordinates. In each case,
the first step is the loss of one ligand L and coordination of 2 to
form the adduct L-2. The substrate 2 then adds oxidatively to
Pd via L-TS2 to generate acyl intermediate L-3.21 Decarbon-
ylation of the acyl group follows slightly different pathways
according to the denticity of the remaining ligand. IPr complex
C-3 undergoes carbonyl deinsertion (through C-TS3) followed
by release of aryloxide to form cationic C-4. In contrast, the κ2-
XantPhos complex P-3 first loses aryloxide to create a suitable
vacant metal site and then undergoes decarbonylation (through
cationic P-TS3) to form cationic P-4. Subsequent β-hydride
elimination22 via L-TS4 yields the alkene−hydride intermediate
L-5. For the XantPhos species, several isomers for P-TS4 were
considered (Figure S6 in the Supporting Information), and the
most stable is included in Figure 2. Finally, the alkene ligand in
L-5 can either decoordinate to form L-6 and terminal olefin 2a
(as a product) or it can react further to isomerize to internal
olefin 2a′; partitioning between these two paths will be
addressed next. Subsequent deprotonation and replacement of
CO by ligand (involving L-7 species not discussed here; see
Figures S4 and S5) regenerate the initial catalyst L-1.
a
Note that A and B are neutral while C−G are cationic.
Pd−alkyl C, which undergoes β-hydride elimination to yield
key intermediate D, a Pd hydride with bound alkene. Ligand
decoordination from D produces the terminal alkene 2a and,
following loss of CO and deprotonation (by liberated p-
nitrophenolate), Pd0 catalyst is regenerated. Alternatively, the
alkene ligand in D can rotate about the Pd···CC axis, a
necessary first step for entering the isomerization reaction
channel. The resulting intermediate E can then undergo
insertion to generate alkylpalladium intermediate F and
subsequent benzylic β-hydride elimination to produce coordi-
nated internal alkene intermediate G. As for D, this
intermediate can then liberate alkene, in this case 2a′, with
regeneration of the active catalyst following loss of CO and
deprotonation.
Nomenclature. As there are many stationary points to be
considered along several different reaction coordinates, we
define a nomenclature system to make comparison as
straightforward as possible. Intermediates are named with a
prefix and a number. The prefix reflects the ligand set: C = IPr
(carbene), P = XantPhos (phosphine), and later M = mixed
(IPr and XantPhos, together). When referring to any ligand,
The overall reaction is computed to be slightly endergonic
for a standard-state pressure of 1 atm of CO, but irreversible
removal of CO from the reaction medium would be expected
Figure 2. Gibbs energy (463.15 K) reaction coordinate (kcal mol−1) leading to the formation of 2a (cf. Scheme 2, cycle I).
D
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 3. Gibbs energy (463.15 K) reaction coordinate (kcal mol−1) for the formation of 2a′ (cf. Scheme 2, cycle II).
to drive the equilibrium toward products. The most stable
minimum along the reaction coordinate in both cases is the
initial bis-ligated (pre)catalyst L-1. For the IPr-based catalyst,
the rate-determining TS is oxidative addition (C-TS2) with a
free energy of activation of 32.3 kcal mol−1; for the XantPhos
derivative, in contrast, the rate-determining TS is β-hydride
elimination (P-TS4) with a very high free energy of activation
of 42.9 kcal mol−1 (note that these are standard-state free
energies, i.e., the free energies associated with all species being
at 1 M concentrationthe effective free energies for all species
along the coordinate relative to L-1 are lowered under
experimental conditions by the excess of substrate compared
to catalyst). The lower-barrier pathway for the IPr case in
comparison to the XantPhos case is consistent with the higher
yields observed experimentally for the former in comparison to
the latter over identical reaction times.
mol−1 (from C-5 to C-TS5), which is consistent with the
prevalence of isomerization observed experimentally. For
XantPhos, in contrast, while alkene rotation is also facile, the
reinsertion step required to generate P-10 has an activation free
energy of 8.4 kcal mol−1 (from P-8 to P-TS9). The larger
activation free energy for XantPhos as ligand in comparison to
IPr implies an increased lifetime of the original alkene−hydride
intermediate for the former, which increases the likelihood of
alkene decoordination prior to isomerization and thus higher
expected α-selectivity.
Feasibility of Mixed Ligand Species. To assess the
experiments with the dual ligand system (Table 4), we evaluate
the possible formation of complexes bearing both IPr and
XantPhos ligands, i.e., “mixed” species (prefix M). Figure 4
shows selected relevant species for the C, P, and M systems.
Considering next the reaction coordinate for the isomer-
ization of 2a to 2a′ (cycle II in Scheme 2), the results are
summarized in Figure 3 (see Figures S4 and S5 in the
Supporting Information for detailed discussion and structures).
Starting from the alkene−hydride species L-5, an alkene
rotation about the Pd···CC axis23 changes the orientation of
the alkene relative to the hydride so as to permit ultimate
reinsertion and formation of the secondary alkyl compound L-
10. For IPr, the necessary olefin rotation takes place in two
steps via C-TS5 and C-TS8 to generate reactive intermediate
C-9. For XantPhos, olefin rotation proceeds through only a
single step via P-TS5 and the resulting P-8 undergoes
reinsertion directly. From precursors C-9 and P-8, hydride
reinsertion occurs through L-TS9 (we use this label for
consistency of nomenclature, even though no intermediate P-9
exists) to produce the secondary alkyl complex L-10.
Following low-barrier rotations about the Pd−C bond to
produce conformers L-11, β-hydride eliminations via L-TS11
produce intermediates L-12 with a bound internal alkene
ligand.24,25 Subsequent alkene decoordinations, metal deproto-
nations, and substitutions of CO by L (Figures S4 and S5 in the
Supporting Information) again regenerate the initial (pre)-
catalysts L-1. For IPr, the activation free energy to irreversibly
generate C-10, following formation of C-5, is only 5.4 kcal
Figure 4. Structures and Gibbs energy (463.15 K) in kcal mol−1 for
IPr, XantPhos, and mixed-ligand derivatives with respect to the most
stable species P-1.
E
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
For the sake of comparison, all energies in this figure are shown
with respect to the same zero of energy, P-1, with
stoichiometric balance of all other ligands.
We first note that the formation of the Pd(0) species M-1 is
not favored, lying at 11.9 kcal mol−1 above P-1. However, the
cationic Pd(II)−acyl derivative M-3 (14.1 kcal mol−1) is
comparable to C-3 (14.9 kcal mol−1) and even more stable than
P-3 (18.9 kcal mol−1). After decarbonylation, the substantial
steric crowding in the M system should eliminate CO as a
ligand. For the β-hydride elimination step, M-TS4 (47.1 kcal
mol−1)26 is found slightly above P-TS4 (42.9 kcal mol−1).
Nevertheless, there are two relevant factors to discuss. First, the
relative energy from M-1 to M-TS4, 35.2 kcal mol−1, is actually
lower than that from P-1 to P-TS4, 42.9 kcal mol−1. This result
suggests that some fraction of the high energy of M-TS4 likely
comes from the penalty of removing the bidentate XantPhos
ligand (M-1 in comparison to P-1) rather than any other
stereoelectronic effects that disfavor the β-H elimination
process per se. Second, the energies of M-TS4 and M-5 are
computed assuming a standard-state pressure of 1 atm of CO.
As previously discussed, the irreversible release of CO should
stabilize these species. Overall, we consider the formation of
mixed intermediates to be plausible under experimental
conditions.
Figure 5. Energy surface scans leading to α-olefin dissociation along
the Pd···CC distance for C (red), P (blue), and M (green) systems
in comparison to energies of activation associated with rate-
determining steps leading to olefin isomerization (vertically at left).
facile. To address reaction path partitioning, we focus on Pd···
alkene adducts having Pd−C distances of ∼3.7 Å (generically
labeled L-AD, for “alkene decoordination”, in Figure 5).
Dissociation of C-AD to this extent is 17.5 kcal mol−1 uphill on
the potential energy surface, which is much more energy than is
required to instead follow the alternative pathway through C-
TS5 that leads to olefin isomerization (2.7 kcal mol−1). The
computational data are thus consistent with the low selectivity
for α-olefins that is observed when IPr is the only available
ligand. In contrast, structure P-AD at 8.1 kcal mol−1 is only
slightly above P-TS9 at 6.0 kcal mol−1, suggesting that
substantially more olefin dissociation relative to isomerization
might be expected to occur, again consistent with experiment.
Finally, the energy of M-AD follows the same trend as P-AD
until ∼3.3 Å; after that point, the energy drops by ca. 4 kcal
mol−1. This reflects nascent recoordination of the previously
noncoordinated phosphine group of the κ1-XantPhos ligand
that is made possible by the departure of the olefin: the M-AD
Pd···P bond distance changes from 4.110 to 3.596 Å as the
olefin dissociation coordinate goes from 3.316 to 3.416 Å. This
effect should further yield high selectivity toward terminal
olefins, as experimentally observed.
Origin of Enhanced α-Selectivity with Mixed-Ligand Set.
We now examine in more detail the competition between
alkene isomerization and alkene decoordination for C-5 and P-
5. In addition, we consider mixed M complexes which
coordinate both IPr and XantPhos ligands.
First, we recapitulate the activation free energies for the
formation of the secondary alkylpalladium species L-10 (and
ultimately internal olefin) from the relevant preceding
intermediates; the stationary points required for this analysis
are provided in Figure 4. As noted above, for IPr, the most
demanding step is alkene rotation from C-5 through C-TS5
(ΔG⧧ = 5.4 kcal mol−1), while for XantPhos it is hydride
reinsertion from P-8 through P-TS9 (ΔG⧧ = 8.4 kcal mol−1).
For the mixed system, no alkene rotation transition state M-
TS5 could be foundthe coordination sphere about Pd is
evidently too crowded to permit such an isomerization.
The salient question is how these activation energies to
generate L-10 compare to activation energies to decoordinate
α-olefin from L-5. To estimate the alkene decoordination
energy, we computed potential energies as a function of the
distance between Pd and the substituted carbon of the α-olefin,
permitting all other degrees of freedom to relax (Figure 5).
Included in Figure 5, for comparison, are the activation energies
associated with TS structures C-TS5 and P-TS9: i.e., the TS
structures for the respective rate-determining steps for alkene
isomerization with the different ligand sets. Note that the
activation energies in Figure 5 differ somewhat from the
activation free energies in Figure 3 because they do not include
thermal contributions to free energy; however, this approach
facilitates comparison to the olefin dissociation reaction
coordinate since (i) constrained structures along the latter are
not stationary points and (ii) no TS structure for olefin
dissociation exists on the potential energy surface (because it is
entropy gain, not potential energy, that makes the reaction
exergonic with a barrier).
We emphasize again that, on a free energy surface, entropy
will cause the olefin dissociation coordinate ultimately to
proceed downhill, but in the absence of a stationary point on
the potential energy surface, it is not straightforward to define a
quantitative free energy of activation for this process.
A closer inspection of the L-AD species at Pd···C distances of
∼3.4 Å (Figure 6) provides insight into the predicted energetic
trends. For the IPr ligand, alkene removal following C-AD
ultimately generates the low-coordinate T-shaped complex C-
6.27 Owing to this undercoordination, even at long Pd···Cβ
distances (3.365 Å), the relaxed Pd···Cα distance remains
substantially shorter (2.593 Å). The bidentate nature of
XantPhos, by contrast, leaves the more stable tetracoordinate
complex P-6 following alkene decoordination from P-AD. As
for the mixed system M-AD, while alkene decoordination
should nominally lead to a three-coordinate product, the
bidentate nature of the XantPhos ligand stabilizes the transient
low-coordinate species through reversion to a bidentate binding
mode in M-6. Indeed, this ratchet-and-pawl-like behavior of the
The energy for alkene decoordination from the IPr complex
(red line, Figure 5) increases rapidly, reflecting a strong binding
of the alkene to the metal. In contrast, decoordination with the
XantPhos (blue line) and mixed-ligand sets (green line) is more
F
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 6. C-AD (a), P-AD (b), and M-AD (c) structures (Pd···Cβ distance ∼3.4 Å) and relative Gibbs free energies (463.15 K, kcal mol−1) for
alkene decoordination.
XantPhos ligand likely contributes to still higher selectivity, as
recoordination of free alkene and subsequent isomerization are
made energetically more difficult upon the readoption of a
bidentate binding mode by XantPhos.
and irreversible decoordination of the terminal alkene
immediately after its generation, thereby precluding the alkene
isomerization pathway.
The strategic use of multiple ligands to enhance catalyst
lifetime and efficiency as well as to improve reactivity/
selectivity has precedent,28 with work on transition-metal-
catalyzed cross-coupling reactions being of particular signifi-
cance.28f Mixed-ligand catalytic systems have also found utility
in circumventing side reactions in polymerization reactions.28b,d
While in most cases the two ligands are proposed to stabilize
the metal center during the catalytic cycle(s) by means of
equilibration, a dual role for one of the ligands as an
organocatalyst also has been suggested.28c The common
theme is that the best reactivity/selectivity is observed only
in the presence of both ligands. Our findings offer a new
example, applied to the selective production of useful LAOs
from biorenewable fatty acid esters, with DFT calculations
providing insights into the specific roles of the effects of the two
ligands.
Considering other experimental trends as a function of ligand
set, it is difficult for theory to address differences in conversion
with different catalysts, as the characterization of all possible
paths that can lead to inactivation of a particular catalytic
system is generally impractical. Theory also offers no specific
insight into the role of LiCl as an additive, although the
computed mechanisms do permit some speculation. For the
XantPhos/PPh3 system, LiCl critically enhances selectivity
(Table 2, entry 7), while for the mixed XantPhos/IPr ligand set,
its role is less decisive. Thus, it may be that LiCl plays a role in
stabilizing κ1-XantPhos coordination in the less sterically
crowded mixed complex incorporating PPh3; additional
investigations are under way to assess such a possibility.
CONCLUSIONS
■
A screening of various reaction conditions has led to the
discovery of the catalytic decarbonylation of a variety of p-
nitrophenyl esters with high conversions and α-selectivities.
Synergistic use of both XantPhos and IPr ligands was found to
be critical for obtaining a higher selectivity for LAOs.
Experimental data support complementary roles for the two
ligands when both are simultaneously present, such that
XantPhos enhances α-selectivity while IPr enhances conversion
at fixed reaction time. Density functional calculations support
these roles and provide insight into the reaction mechanism at
the atomic level of detail. Through a comparison of competing
alkene dissociation paths, which deliver terminal alkene, and
alkene rotation paths, which are the first steps required to
generate an internal alkene, the calculations suggest that in a
mixed-ligand system the bulkiness of IPr combined with the
flexibly mono- or bidentate nature of XantPhos promote a rapid
EXPERIMENTAL SECTION
■
General Procedures. All manipulations were carried out using
either Schlenk tube techniques under an inert atmosphere or in a
glovebox. Palladium salts, phosphine ligands, fatty acids, p-nitrophenyl
chloroformate, LiCl, N,N-dimethylenepropyleneurea (DMPU), and
triethylamine were purchased from commercial sources. The N-
heterocyclic carbene ligand IPr was synthesized using literature
procedures.29 Mass spectra were acquired in the ESI (+) mode.
Ether and dichloromethane were passed through a purification column
(Glass Contour, Laguna, CA) prior to use. All H, 13C{1H}, and 31P
1
NMR spectra were collected on either a Varian VXR-500 or Varian VI-
300 spectrometer and calibrated to the residual protonated solvent at δ
1
7.27 for deuterated chloroform (CDCl3). H NMR peaks are labeled
as singlet (s), doublet (d), triplet (t), and multiplet (m). All GC-MS
experiments were conducted on an Agilent Technologies 7890A GC
system and 5975C VL MSD. The GC column was a HP-5 ms with
dimensions 30 m × 0.25 mm. The standard method used for all runs
G
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
involved an initial oven temperature of 50 °C (held for 2 min)
followed by a 20 °C min−1 ramp to 70 °C (held for 6 min), followed
by a final 20 °C min−1 ramp to 230 °C (held for 3 min).
All minima and transition states were confirmed by analytical
vibrational frequency calculations at 463.15 K. It was confirmed that
transition-state structures connect with reactants and products by
following those normal modes associated with imaginary frequencies.
All frequencies below 50 cm−1 were replaced by 50 cm−1 when
accounting for thermal contributions to vibrational partition
functions.44
Solvation effects were computed using the SMD continuum
solvation model.45 We used N,N-dimethylacetamide (DMA, ε =
37.781) to mimic the highly polar solvent of the experiments (DMPU,
ε = 36.12), for which more specific parameters are not available.
Single-point calculations were performed on gas-phase geometries
using the SDD basis set with an additional f function46 for Pd and the
6-311++G(d,p) basis set47 for all other atoms. Gibbs energies for all
species (except for CO gas) were corrected by a factor of RT ln 38.0,
corresponding to 3.3 kcal mol−1, for standard-state corrections from 1
atm gas (at 463.15 K) to 1 M solution.
General Procedure for Syntheses of p-Nitrophenyl Esters.
To a stirred solution of the carboxylic acid (5 mmol) in dry CH2Cl2
(ca. 30 mL) under an N2 atmosphere was added NEt3 (0.7 mL, 5.02
mmol) using a syringe, and the mixture was stirred for 10 min. At the
end of this period, p-nitrophenyl chloroformate (1.01 g, 5.01 mmol)
was added to the mixture in portions over a period of 5−10 min. The
reaction mixture showed an immediate color change to pale yellow
along with strong effervescence. The reaction mixture was stirred at
room temperature overnight (∼16 h). During this time, the reaction
mixture would generally show the presence of an off-white precipitate.
Subsequently, the reaction mixture was concentrated under vacuum to
remove all the CH2Cl2 and then EtOAc (ca. 20 mL) was added to the
residue obtained. The resulting mixture was filtered through a pad of
basic alumina, and the filtrate was concentrated on a rotary evaporator.
The residue obtained was dried under high vacuum and was usually
about >95% pure ester product, as judged by 1H NMR spectroscopy. If
the ester obtained was not of substantially high purity, the crude
product was redissolved in EtOAc (ca. 15 mL) and subjected to
another filtration through basic alumina. All the fatty acid esters
reported, 1−8, are known compounds and were confirmed by
comparison to literature data: p-nitrophenyl nonanoate (1),11 p-
nitrophenyl 4-phenylbutyrate (2),11 p-nitrophenyl undecylenate (3),11
p-nitrophenyl dodecanoate (4),30 p-nitrophenyl tetradecanoate (5),31
p-nitrophenyl hexadecanoate (6),31 p-nitrophenyl octadecanoate (7),31
p-nitrophenyl (9Z)-octadecanoate (8).32
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.organo-
met.7b00411.
Spectral data for isolated olefins, representative 1H NMR
spectra, GC-MS overlay plots, and computational
details/plots reported (PDF)
Computed Cartesian coordinates of all of the molecules
Representative Procedure for Decarbonylation Reactions.
Inside a N2-filled glovebox, a Schlenk tube was charged with p-
nitrophenyl nonanoate (0.063 g, 0.226 mmol, 1 equiv), PdCl2 (0.001
g, 0.006 mmol, 2.5 mol %), XantPhos (0.006 g, 0.011 mmol, 5 mol %),
IPr (0.005 g, 0.011 mmol, 5 mol %), and LiCl (0.002 or 0.010 g, 20 or
100 mol %), and DMPU (ca. 0.5 mL) was injected in followed by a
Teflon-coated stir bar. The reaction mixture was light yellow at this
time. The Schlenk tube was sealed, brought outside the box, and set in
an oil bath preheated to 190 °C. The reaction mixture was stirred at
this temperature for the stipulated period of time, during which it
darkened and finally turned red-brown. At the end of the reaction, the
Schlenk tube was pulled out of the oil bath and cooled to room
temperature.
reported in this study (XYZ)
AUTHOR INFORMATION
Corresponding Authors
*E-mail for C.J.C.: cramer@umn.edu.
*E-mail for W.B.T.: wtolman@umn.edu.
ORCID
■
Manuel A. Ortuno: 0000-0002-6175-3941
̃
Marc A. Hillmyer: 0000-0001-8255-3853
Christopher J. Cramer: 0000-0001-5048-1859
William B. Tolman: 0000-0002-2243-6409
GC-MS Analysis. The reaction mixture was transferred to a
separatory funnel, diluted with EtOAc (ca. 5 mL), and washed with
1 M HCl (ca. 5 mL × 2) and brine (ca. 5 mL). The organic layer was
collected, dried over MgSO4, and analyzed by GC-MS using 1,3,5-
trimethoxybenzene as an internal standard.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Olefin Isolation (Reaction Carried Out at 1 mmol Scale). The
reaction mixture was transferred to a separatory funnel and diluted
with 1 M HCl (ca. 10 mL). This mixture was extracted with pentane
(ca. 5 mL × 3). The combined pentane extracts were washed
successively with 1 M HCl (ca. 5 mL), saturated NaHCO3 solution
(ca. 5 mL), and saturated NaCl solution (ca. 5 mL) and then
separated. The combined pentane extract was filtered through a plug
of alumina, dried with anhydrous MgSO4, and then concentrated
under vacuum to remove pentane and obtain the olefin product, which
Funding for this work was provided by the Center for
Sustainable Polymers at the University of Minnesota, a
National Science Foundation (NSF) supported Center for
Chemical Innovation (CHE-1413862). The authors acknowl-
edge the Minnesota Supercomputing Institute (MSI) at the
University of Minnesota for providing resources that
contributed to the research results reported within this paper.
1
was ∼95% pure by H NMR spectroscopy.
REFERENCES
Computational Details. All calculations were performed at the
density functional theory (DFT) level33 using the M06-L density
functional34,35 as implemented in Gaussian 09.36 This functional
performs well for transition-metal chemistry,35,37 nonbonded inter-
actions associated with medium-range electron correlation,35 and bond
dissociation enthalpies.38 Additionally, M06-L was found to be optimal
for a closely related Pd-catalyzed dehydrative decarbonylation in
comparison to DLPNO−CCSD(T)39 benchmark energies.40 All
geometries were optimized in the gas phase with an ultrafine grid
for integral evaluations. Automatically generated density fitting was
employed to speed the evaluation of Coulomb integrals. The 6-
31G(d,p) basis set41 was used for H, C, N, O, and P with an additional
diffuse function for O,42 and the SDD43 effective core potential and its
associated double-ζ basis set was employed for Pd.
■
(1) Dawes, G. J. S.; Scott, E. L.; No
̂
tre, J. L.; Sanders, J. P. M.; Bitter,
J. H. Green Chem. 2015, 17, 3231−3250.
(2) Foglia, T. A.; Barr, P. A. J. Am. Oil Chem. Soc. 1976, 53, 737−741.
́
(3) (a) Ternel, J.; Lebarbe, T.; Monflier, E.; Hapiot, F. ChemSusChem
2015, 8, 1585−1592. (b) Maetani, S.; Fukuyuma, T.; Suzuki, N.;
Ishihara, D.; Ryu, I. Organometallics 2011, 30, 1389−1394.
(4) (a) Chatterjee, A.; Eliasson, S. H. H.; Tornroos, K. W.; Jensen, V.
R. ACS Catal. 2016, 6, 7784−7789. (b) Liu, Y.; Virgil, S. C.; Grubbs,
R. H.; Stoltz, B. M. Angew. Chem., Int. Ed. 2015, 54, 11800−18003.
(c) Liu, Y.; Kim, K. E.; Herbert, M. B.; Fedorov, A.; Grubbs, R. H.;
Stoltz, B. M. Adv. Synth. Catal. 2014, 356, 130−136. (d) Miranda, M.
O.; Pieterangelo, A.; Hillmyer, M. A.; Tolman, W. B. Green Chem.
2012, 14, 490−494. (e) Kraus, G. A.; Riley, S. Synthesis 2012, 44,
H
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
3003−3005. (f) Le No
̂
tre, J.; Scott, E. L.; Franssen, M. C. R.; Sanders,
C. G.; Glorius, F. J. Am. Chem. Soc. 2017, 139, 4443−4451. (d) Iizuka,
E.; Wakioka, M.; Ozawa, F. Macromolecules 2016, 49, 3310−3317.
(e) Panish, R. A.; Chintala, S. R.; Fox, J. M. Angew. Chem., Int. Ed.
2016, 55, 4983−4987. (f) Fan, Y.; Cong, M.; Peng, L. Chem. - Eur. J.
2014, 20, 2698−2702. (g) Maiti, D.; Fors, B. P.; Henderson, J. L.;
Nakamura, Y.; Buchwald, S. L. Chem. Sci. 2011, 2, 57−68. (h) Fors, B.
P.; Buchwald, S. L. J. Am. Chem. Soc. 2010, 132, 15914−15917.
(i) Yang, Q.; Ney, J. E.; Wolfe, J. P. Org. Lett. 2005, 7, 2575−2578.
(j) Moreau, C.; Frost, C. G.; Murrer, B. Tetrahedron Lett. 1999, 40,
5617−5620.
(29) Bantreil, X.; Nolan, S. P. Nat. Protoc. 2011, 6, 69−77.
(30) Kobayashi, S.; Matsubara, R.; Nakamura, Y.; Kitagawa, H.;
Sugiura, M. J. Am. Chem. Soc. 2003, 125, 2507−2515.
(31) Plusquellec, D.; Roulleau, F.; Bertho, F.; Lefeuvre, M.
Tetrahedron 1986, 42, 2457−2467.
(32) Brundiek, H. B.; Evitt, A. S.; Kourist, R.; Bornscheuer, U. T.
Angew. Chem., Int. Ed. 2012, 51, 412−414.
(33) Cramer, C. J.; Truhlar, D. G. Phys. Chem. Chem. Phys. 2009, 11,
10757−10816.
(34) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101.
(35) (a) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167.
(b) Zhao, Y.; Truhlar, D. G. Chem. Phys. Lett. 2011, 502, 1−13.
(36) Frisch, M. J., et al. Gaussian 09, Revision E.01; Gaussian, Inc.,
Wallingford, CT, 2010.
J. P. M. Tetrahedron Lett. 2010, 51, 3712−3715. (g) Gooβen, L. J.;
Rodríguez, N. Chem. Commun. 2004, 724−725. (h) Miller, J. A.;
Nelson, J. A.; Byrne, M. P. J. Org. Chem. 1993, 58, 18−20.
(5) John, A.; Miranda, M. O.; Ding, K.; Dereli, B.; Ortuno, M. A.;
̃
LaPointe, A. M.; Coates, G. W.; Cramer, C. J.; Tolman, W. B.
Organometallics 2016, 35, 2391−2400.
(6) Maetani, S.; Fukuyuma, T.; Suzuki, N.; Ishihara, D.; Ryu, I. Chem.
Commun. 2012, 48, 2552−2554.
̂
(7) Le Notre, J.; Scott, E. L.; Franssen, M. C. R.; Sanders, J. P. M.
Green Chem. 2011, 13, 807−809.
(8) Santillan-Jimenez, E.; Crocker, M. J. J. Chem. Technol. Biotechnol.
2012, 87, 1041−1050.
(9) Vogt, D. In Applied Homogeneous Catalysis with Organometallic
Compounds; Cornils, B., Herrmann, W. A., Eds.; Wiley-VCH:
Weinheim, Germany, 1996; pp 246−258.
(10) (a) Weissermel, K.; Arpe, H.-J. Industrial Organic Chemistry, 3rd
ed.; Wiley-VCH: Weinheim, Germany, 1997; pp 75−83. (b) McGuin-
ness, D. S. Chem. Rev. 2011, 111, 2321−2341. Speiser, F.; Braunstein,
P.; Saussine, L. Acc. Chem. Res. 2005, 38, 784−793. (c) Dry, M. E.
Catal. Today 2002, 71, 227−241.
(11) John, A.; Hogan, L. T.; Hillmyer, M. A.; Tolman, W. B. Chem.
Commun. 2015, 51, 2731−2733.
(12) Gooβen, L. J.; Paetzold, J. Angew. Chem., Int. Ed. 2002, 41,
1237−1241.
(37) Gusev, D. G. Organometallics 2013, 32, 4239−4243.
(38) (a) Sieffert, N.; Buhl, M. Inorg. Chem. 2009, 48, 4622−4624.
̈
(13) Methyl or ethyl esters did not undergo decarbonylation under
similar conditions.
(b) Raju, R. K.; Bengali, A. A.; Brothers, E. N. Dalton Trans. 2016, 45,
13766−13778.
(14) Extensive hydrolysis of the ester was not observed, in contrast to
previous reactions with the p-nitrophenyl ester of hydrocinnamic acid
performed under similar conditions.11 We speculate that this disparity
is due to the activated nature of the latter substrate relative to the
aliphatic substrates used herein.
(15) (a) Hopkinson, M. N.; Richter, C.; Schedler, M.; Glorius, F.
́
Nature 2014, 510, 485−496. (b) Díez-Gonzalez, S.; Marion, N.;
Nolan, S. P. Chem. Rev. 2009, 109, 3612−3676. (c) Herrmann, W. A.
Angew. Chem., Int. Ed. 2002, 41, 1290−1309.
(16) We speculate that the internal double bond could coordinate to
the palladium center during the catalytic cycle, which could facilitate
isomerization processes.
(39) (a) Riplinger, C.; Neese, F. J. Chem. Phys. 2013, 138, 034106.
(b) Riplinger, C.; Sandhoefer, B.; Hansen, A.; Neese, F. J. Chem. Phys.
2013, 139, 134101.
(40) (a) Lee, M.-T.; Lee, H. M.; Hu, C.-H. Organometallics 2007, 26,
1317−1324. (b) Backtorp, C.; Norrby, P.-O. Dalton Trans. 2011, 40,
̈
11308−11314. (c) Menezes da Silva, V. H.; Braga, A. A. C.; Cundari,
T. R. Organometallics 2016, 35, 3170.
(41) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972,
56, 2257−2261. (b) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley,
J. S.; Gordon, M. S.; DeFrees, D. J.; Pople, J. A. J. Chem. Phys. 1982,
77, 3654−3665.
(42) Clark, T.; Chandrasekhar, J.; Spitznagel, G. W.; Schleyer, P. J.
Comput. Chem. 1983, 4, 294−301.
(17) Sperger, T.; Sanhueza, I. A.; Kalvet, I.; Schoenebeck, F. Chem.
Rev. 2015, 115, 9532−9586.
(18) Beletskaya, I. P.; Cheprakov, A. V. Chem. Rev. 2000, 100, 3009−
3066.
(43) Andrae, D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chim. Acta 1990, 77, 123−141.
(44) Ribeiro, R. F.; Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J.
Phys. Chem. B 2011, 115, 14556−14562.
(19) Rosenfarb, J.; Huffman, H. L., Jr.; Caruso, J. A. J. Chem. Eng.
Data 1976, 21, 150−153.
(45) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B
2009, 113, 6378−6396.
(20) Ortuno, M. A.; Dereli, B.; Cramer, C. J. Inorg. Chem. 2016, 55,
̃
4124−4131.
(46) Ehlers, A. W.; Bohme, M.; Dapprich, S.; Gobbi, A.; Hollwarth,
̈
̈
(21) Bao, T.-S.; Zhaorigetu, B.; Agula, B.; Baiyin, M.; Jia, M. J. Org.
Chem. 2014, 79, 803.
A.; Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
̈
Chem. Phys. Lett. 1993, 208, 111−114.
(22) (a) Lee, M.-T.; Lee, H. M.; Hu, C.-H. Organometallics 2007, 26,
(47) (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.
1317−1324. (b) Backtorp, C.; Norrby, P.-O. Dalton Trans. 2011, 40,
̈
11308−11314. (c) Menezes da Silva, V. H.; Braga, A. A. C.; Cundari,
T. R. Organometallics 2016, 35, 3170.
(23) Gordillo, A.; Ortuno, M. A.; Lop
́
ez-Mardomingo, C.; Lledos
́
, A.;
̃
Ujaque, G.; de Jesus, E. J. Am. Chem. Soc. 2013, 135, 13749−13763.
́
(24) Scarso, A.; Colladon, M.; Sgarbossa, P.; Santo, C.; Michelin, R.
A.; Strukul, G. Organometallics 2010, 29, 1487−1497.
(25) (a) Xu, L.-P.; Hilton, M. J.; Zhang, X.; Norrby, P.-O.; Wu, Y.-D.;
Sigman, M. S.; Wiest, O. J. Am. Chem. Soc. 2014, 136, 1960−1967.
(b) Hilton, M. J.; Xu, L.-P.; Norrby, P.-O.; Wu, Y.-D.; Wiest, O.;
Sigman, M. S. J. Org. Chem. 2014, 79, 11841−11850.
(26) The isomer M-TS4′ with the phosphine ligand trans to the
incoming hydogen was found at 52.9 kcal mol−1.
(27) Ortuno, M. A.; Conejero, S.; Lledos
2013, 9, 1352−1382.
́
, A. Beilstein J. Org. Chem.
̃
(28) (a) Rej, S.; Majumdar, M.; Kando, S.; Sugino, Y.; Tsurugi, H.;
Mashima, K. Inorg. Chem. 2017, 56, 634−644. (b) Wakioka, M.;
Takahashi, R.; Ichihara, N.; Ozawa, F. Macromolecules 2017, 50, 927−
934. (c) Guo, C.; Janssen-Muller, D.; Fleige, M.; Lerchen, A.; Daniliuc,
I
DOI: 10.1021/acs.organomet.7b00411
Organometallics XXXX, XXX, XXX−XXX