A POCO type pincer complex of iridium: Synthesis, characterization, and catalysis

Article abstract of DOI:10.1016/j.poly.2018.12.001

A convenient synthesis of a new POCO-type pincer ligand, “^tBuPOCOH” (3-(di-tert-butylphosphinito)acetophenone is reported. Metallation using [Ir(COD)Cl]₂ provides the dimeric species (μ-Cl-[^tBuPOCOIrHCl]₂) (1a) as a major product, along with isomer 1b. Though not fully characterized, 1b is shown to be chemically equivalent to 1a by a series of experiments with AgOTf and CO which lead to formation of a single product, ^tBuPOCOIr(CO)HOTf (3). The 1a/b mixture gives inferior results to ^iPrPCPIr when used as pre-catalyst in the dehydrogenative coupling of vinyl arenes, though olefin isomerization activity is enhanced. This system was also evaluated in the Geurbet conversion of ethanol to n-butanol and higher alcohols. The CO adducts of 1a/b, cis/trans-2, were found to give the best results as pre-catalyst, achieving a 33% yield of n-butanol and an overall 47% yield of n-alcohols with a catalyst loading of 0.5% when heated at 150 °C for 4 hours. This represents the first example of a pincer ligated iridium complex as catalyst in the Geurbet reaction of ethanol.

Full text of DOI:10.1016/j.poly.2018.12.001

Polyhedron 160 (2019) 83–91
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
A POCO type pincer complex of iridium: Synthesis, characterization, and
catalysis
b
Miles Wilklow-Marnell a, , William W. Brennessel
⇑
a Department of Chemistry, State University of New York at New Paltz, New Paltz, NY 12561, USA
b Department of Chemistry, University of Rochester, Rochester, NY 14627, USA
a r t i c l e i n f o
a b s t r a c t
A convenient synthesis of a new POCO-type pincer ligand, ‘‘tBuPOCOH”, (3-(di-tert-butylphosphinito)ace-
Article history:
Received 4 August 2018
Accepted 1 December 2018
Available online 13 December 2018
tophenone is reported. Metallation using [Ir(COD)Cl]2 provides the dimeric species
[
(l-Cl-
tBuPOCOIrHCl]2) (1a) as a major product, along with isomer 1b. Though not fully characterized, 1b is
shown to be chemically equivalent to 1a by a series of experiments with AgOTf and CO which lead to for-
mation of a single product, tBuPOCOIr(CO)HOTf (3). The 1a/b mixture gives inferior results to iPrPCPIr
when used as pre-catalyst in the dehydrogenative coupling of vinyl arenes, though olefin isomerization
activity is enhanced. This system was also evaluated in the Geurbet conversion of ethanol to n-butanol
and higher alcohols. The CO adducts of 1a/b, cis/trans-2, were found to give the best results as pre-cat-
alyst, achieving a 33% yield of n-butanol and an overall 47% yield of n-alcohols with a catalyst loading of
0.5% when heated at 150 °C for 4 hours. This represents the first example of a pincer ligated iridium com-
plex as catalyst in the Geurbet reaction of ethanol.
Keywords:
Pincer complex
Iridium
Catalysis
Geurbet reaction
Dehydrogenative coupling
Ó 2018 Elsevier Ltd. All rights reserved.
1. Introduction
facilitated by iridium pincer complexes including the efficient
transfer dehydrogenation of cycloalkanes [20,21], the acceptorless
The use of pincer ligated transition metal complexes has
become widespread in the realm of organometallic transforma-
tions [1–7]. Earlier work focused mainly on second and third row
metal complexes and helped to elucidate many of the fundamental
aspects of pincer complex reactivity. Notable findings include the
addition of strong CAH and CAC bonds to tBuPCP rhodium com-
plexes [8,9], a stable ruthenium agostic CAH complex [10], and
efficient palladium catalysts for Heck and Suzuki coupling reac-
tions using POCOP type pincer ligands [11–13]. More recently, pin-
cer chemistry has ventured to the first row transition metals and
now includes impressive examples such as the cobalt catalyzed
stereospecific hydrogenation of alkynes to Z and E-alkenes [14],
the reversible iron catalyzed acceptorless dehydrogenation/hydro-
genation of alcohols and ketones [15], and the manganese cat-
alyzed hydrogenation of amides to the corresponding alcohols
and amines [16].
dehydrogenation of cyclic and n-alkanes by a homogenous catalyst
[22,23], and the catalytic dehydrogenation of ammonia borane
[24], an important candidate in hydrogen storage technology.
While these examples utilized symmetric pincer ligands, an
increasing number of reported pincer complexes are asymmetric
which vastly increases the variability/tunability of steric and elec-
tronic factors in these systems, sometimes providing advantages
over related symmetric species [5]. For example, iPrPOCOPIr has
been reported to achieve 13 turn-overs minꢀ1 (by inference
ꢁ780 turn-over hourꢀ1
) in the transfer dehydrogenation of
cyclooctane using tert-butylethylene as acceptor [25], but the anal-
ogous iPrPSCOPIr complex with one phosphinite and one phos-
phinothious group is reported to give 2649 turn-overs hourꢀ1
under nearly identical conditions, correlating to a more than three
times increase in activity [26].
As part of an ongoing program to explore and expand the scope
of iridium pincer chemistry, a new asymmetric POCO type pincer
ligand, ‘‘tBuPOCOH” (3-(di-tert-butylphosphinito)acetophenone),
has been synthesized and metallated using [Ir(COD)Cl]2 (Scheme 1).
The product formed, tBuPOCOIrHCl, is obtained as a mixture of two
species, 1a and 1b in a 1–0.35 ratio. Herein, the reactivity of these
compounds with AgOtf and carbon monoxide is reported, as well
With respect to iridium a rich body of pincer chemistry exists,
primarily exploiting the facile addition and elimination of
element-hydrogen bonds to iridium centers in these complexes
[17–19]. Accordingly, a number of important discoveries have been
⇑
Corresponding author.
E-mail address: marnellm@newpaltz.edu (M. Wilklow-Marnell).
https://doi.org/10.1016/j.poly.2018.12.001
0277-5387/Ó 2018 Elsevier Ltd. All rights reserved.

84
M. Wilklow-Marnell, W.W. Brennessel / Polyhedron 160 (2019) 83–91
Scheme 1. Synthesis of tBuPOCOH ligand and metallation using [Ir(COD)Cl]2.
as their use as pre-catalysts in the dehydrogenative coupling of
styrene, and Geurbet conversion of ethanol to n-butanol.
of tBuPOCOIrHCl (Fig. 2). Surprisingly, dissolution of these crystals
revealed identical 31P and 1H NMR spectra to that obtained before
recrystallization. However, close inspection of the crystals under a
microscope showed that while the bulk of the material was hexag-
onal platelets of 1a, all of the crystals had a coating of fine poly-
crystalline material, likely 1b. NMR analysis of the supernatant
also revealed identical spectra. Apparently, 1a and 1b crystallize
separately, but concomitantly. The identity of 1b is not completely
defined, but it is almost certainly an isomer or monomeric form of
1a, which is corroborated by reaction studies with AgOTf and CO
(vide infra).
As in the solid state, 1a most likely remains in dimeric form in
solution as well. While tBuPOCOIrHCl may exist as two enan-
tiomers, this cannot account for the two species observed, because
if all tBuPOCOIrHCl assumed monomeric form in solution only one
species would be detected by NMR. One possibility for the identity
2. Results and discussion
The tBuPOCOH ligand is readily prepared by refluxing 1 equiv. of
sodium hydride with 3-hydroxyacetophenone in THF followed by
addition of 1.1 equiv. di-tert-butylchlorophosphine, briefly reflux-
ing, and then allowing to stir overnight at room temperature.
Removal of volatiles from this mixture and extraction of the resul-
tant residue with THF, followed by concentration under vacuum
provides the ligand in 96% yield as a slightly yellow oil. The pro-
duct appears as a singlet in the 31P{1H} NMR spectrum at d 155.0
in C6D6. Over several weeks after synthesizing, a number of large
crystals of tBuPOCOH were found to have spontaneously formed
from the oil allowing for structure analysis by X-ray diffraction
(Fig. 1).
of 1b is the l-H analog of 1a which would be consistent with the
somewhat more upfield resonance of the 1b hydride, however, this
would likely present a triplet hydride resonance in the 1H NMR
spectrum which is not observed. Since dimer 1a is formed from
enantiomeric pairs of tBuPOCOIrHCl, and formation of a racemate
from the reaction of tBuPOCOIrH and [Ir(COD)Cl]2 would likely lead
solely to 1a, the remaining possibilities are that one enantiomer
forms in excess and may either remain in monomeric form, or form
Refluxing tBuPOCOH and [Ir(COD)Cl]2 in toluene for 24 h fol-
lowed by concentration to dryness, and washing of the solids
obtained with pentane, provides an orange solid material in 84%
yield formulated as tBuPOCOIrHCl. Indeed, the elemental analysis
results are consistent with this formulation. However, inspection
of the 31P{1H} NMR spectrum revealed two species (1a, 1b) as sin-
glets at d 156.5 and 154.7 in a 1–0.35 ratio, respectively. The
hydride region of the 1H NMR displays two doublets at d ꢀ27.3
(major) and ꢀ27.6 (minor) in the same ratio observed by 31P
NMR spectroscopy. These species do not appear to be in equilib-
rium as the ratio of 1a and 1b does not change with variable tem-
perature (0–70 °C), or solvent (benzene, toluene, chloroform, THF).
Recrystallization from toluene provided crystals suitable for X-ray
a sterically strained
l-Cl dimer with phosphinite groups facing
each other. Of these species, the
l
-Cl dimer seems the most likely
candidate in view of the very similar chemical shifts of the hydride
signals for 1a and 1b. Related monomeric PCPIrHCl and POCO-
PIrHCl species with a hydride trans to a vacant site display excep-
tionally upfield hydride resonances, as far as d ꢀ45 [27,21]. While
the possibility that 1b is actually monomeric with a solvent mole-
cule occupying the ‘‘vacant” site cannot be completely ruled out, it
seems quite unlikely in light of the fact that 1H NMR resonances for
coordinated solvent could not be located upon addition of excess
toluene, CHCl3, or THF to solutions of 1a/b in CDCl3. Furthermore,
since a significant difference in molecular weight (ꢁ500 amu),
and likely solution diffusion rate, would exist between monomeric
and dimeric forms, a DOSY-NMR experiment was conducted on a
sample of 1a/b in CDCl3 (see SI). Based on the hydride region of
the 1H NMR spectrum, diffusion rates for 1a and 1b were found
to be nearly identical (0.7% difference based on longest rate),
whereas under identical conditions ferrocene and acetonitrile
(ꢁ145 amu difference) were found to have a an approximately
50% difference in diffusion rate. These findings indicate that 1a
and 1b are both either monomeric or dimeric, of which dimeric
seems more likely. In order to ascertain the chemical equivalency
of 1a and 1b, a series of experiments reacting the as synthesized
1a/b mixture with AgOTf and CO were conducted, which led to
convergence to a single product, confirming the overall formula-
tion of 1a/b as tBuPOCOIrHCl.
diffraction which showed the major species 1a to be the l-Cl dimer
Fig. 1. ORTEP diagram of the tBuPOCOH ligand with thermal ellipsoids at 50%
probability level.
In the solid state, 1a/b is found to be quite air stable. Samples
left open to air for one month were found to provide the same

M. Wilklow-Marnell, W.W. Brennessel / Polyhedron 160 (2019) 83–91
85
Fig. 2. ORTEP diagram of [tBuPOCOIrHCl]2, 1a, with thermal ellipsoids at 50% probability.
31P and 1H NMR spectra as freshly synthesized material. Further-
more, although all solvents were dried prior to use for the sake
of consistency, no water sensitivity has been found with this com-
plex, and removing oxygen by freeze–pumpthaw cycles or purging
with argon is the only pretreatment necessary for solvents in prac-
tical use. In solution, 1a/b decomposes slowly over 16 h to several
days in solvents which have not been degassed, dependent on the
solvent used.
These results indicate that the formation of cis-2 and trans-2 from
1a/b results for kinetic reasons due to steric/structural differences
in these isomers, and that the thermodynamic product is trans-2,
which is not in equilibrium with cis-2.
When solutions of trans-2, or mixed cis-2 and trans-2, in aro-
matic solvent or CH2Cl2 are treated with a stoichiometric amount
of AgOTf, quantitative conversion to a single new product with a res-
onance at d 162.7 in the 31P{1H} NMR spectrum is observed after stir-
ring overnight. A single hydride resonance (doublet, JPH = 21.2 Hz) is
found in the 1H NMR spectrum at d ꢀ26.0, consistent with a hydride
in an apical position trans to OTf. Removal of AgCl by filtration, con-
centration, and recrystallization by vapor diffusion of pentane into a
concentrated benzene solution of this product and storing at ꢀ17 °C
provided crystals suitable for X-ray diffraction. The structure deter-
mined confirmed formation of the expected product, tBuPOCOIr(CO)
(H)OTf (3) (Fig. 3, Scheme 3), though the crystals obtained contained
½ co-crystallized benzene per iridium complex.
This series of reactions may also be conducted in the opposite
order, first reacting 1a/b with AgOTf, ostensibly forming
tBuPOCOIrHOTf (4), followed by addition of CO to form 3. The tri-
flate ligand in 4 appears to be highly labile, and exceptionally
broad resonances are observed in the 31P{1H} NMR spectra of 4
in arene solvents. However, upon addition of ꢂ1 equiv. AgOTf to
toluene or benzene solutions of 4, a single sharp resonance is
observed, accompanied by an accordingly well resolved 1H NMR
spectrum displaying a single hydride signal (see SI). NaOTf may
also be used, though due to poorer solubility it was found
Assuming that 1a and 1b are dimeric isomers of each with the
molecular formula [tBuPOCOIrHCl]2, it was expected that addition
of carbon monoxide would generate a single species by cleavage
of these dimers into their coordinatively saturated monomeric
form. However, when a saturated solution of 1a/b in toluene is
exposed to 1 atm CO, accompanied by a color change from orange
to pale yellow, two new species are generated, again in a 1–0.35
ratio as detected by 31P{1H} NMR spectroscopy as singlets at d
164.0, and 162.3, respectively. Informatively, the hydride reso-
nances of these compounds appear as doublets at d ꢀ9.4 (minor)
and ꢀ19.8 (major). These chemical shifts are consistent with
assignment as the cis (cis-2) and trans (trans-2) H-Cl isomers of
tBuPOCOIr(CO)HCl (Scheme 2), and are in good agreement with
tBu
reported data for related species [28,29]. The analogous
-
POCOPIr(CO)HCl complex is reported to exist as both the cis and
trans forms in equilibrium, favoring the trans form, as determined
by thermolysis studies [29]. When the as prepared mixture of cis-2
and trans-2 is heated at 60 °C for 6 min (after removal of excess
CO) complete, irreversible, conversion to trans-2 is observed.
Scheme 2. Reactivity of 1a and 1b with carbon monoxide to form cis-2 and trans-2, and thermal conversion of cis-2 to trans-2.

86
M. Wilklow-Marnell, W.W. Brennessel / Polyhedron 160 (2019) 83–91
a 20% yield of (E,E)-1,4-diphenyl-1,3-butadiene were observed by
GCMS based on the stoichiometry in Scheme 5. Unlike the iPrPCPIr
catalyzed reaction, no evidence of (E,Z)-1,4-diphenyl-1,3-butadi-
ene was found, likely due to the greater isomerization activity of
tBuPOCOIr (vide infra). However, substantial amounts of two prod-
ucts with m/z 208, presumably (E) and (Z) 1,4-diphenyl-1-butene,
were detected, representing a further ꢁ13% yield of coupled prod-
ucts. When the reaction was run for 48 hours, no improvement in
yield of (E,E)-1,4-diphenyl-1,3-butadiene was observed, though the
amount of product with m/z 208 increased somewhat to ꢁ18%.
Despite the reduced activity of 1a/b with respect to iPrPCPIr, it
was considered that the reduced steric bulk of the tBuPOCOIr frag-
ment might provide some advantage in the coupling of sterically
demanding substrates such as
a-methylstyrene. However, only
6% conversion was observed, compared to 9% reported for
iPrPCPIrHCl.
Allylbenzene was also tested as a substrate and >99% conver-
sion was observed. However, 97% of the products resulted from
isomerization to cis and trans-1-phenylpropene in
a 1:30
ratio, respectively. This is similar to iPrPCPIr, although in this case
tBuPOCOIr is more reactive (82% conversion with iPrPCPIr). Propyl-
benzene and 2 peaks with masses consistent with coupled allyl-
benzene isomers account for the remaining 3% of products. It
would seem that the less encumbered sterics of tBuPOCOIr leads
to improved performance in olefin isomerization, while providing
no beneficial effects in dehydrogenative coupling, likely largely
influenced by electronic differences between the POCO and PCP
systems. These results also suggest that tBuPOCOIr operates by
the same mechanism as iPrPCPIr in dehydrogenative coupling. This
requires a vinyl arene substrate which forms a metalloindene
intermediate, and addition of a second vinyl arene, which must
undergo CAH oxidative addition of a terminal vinylic proton form-
ing a congested transition state complex, before hydride insertion
and ultimately CAC reductive elimination. In this mechanism the
addition of the second vinyl arene molecule is highly inhibited,
Fig. 3. ORTEP diagram of tBuPOCOIr(CO)(H)OTf (3) with thermal ellipsoids at 50%
probability.
necessary to generate a saturated solution of NaOTf by sonication/
heating before adding to 4. Likewise, when a 30% v/v solution of
THF in benzene is used, a single defined product is observed by
31P{1H} NMR spectroscopy, likely due to formation of [tBuPOCOIr
(THF)H][OTf]. These proposed equilibria are shown in Scheme 4.
Immediately after addition of 1 atm CO to a freshly prepared
toluene solution of 4, formation of 3 is observed by 31P{1H} NMR
spectroscopy, representing 68% of total products. Another reso-
nance at d 166.8, representing 27% of total products, is also
observed with an associated hydride doublet at d ꢀ8.66 in the 1H
NMR spectrum. The remaining 5% of products is accounted for by
another unidentified species with a 31P{1H} NMR resonance at d
164.4, which also presents a hydride signal at d ꢀ10.4. Concentra-
tion of this solution to dryness in vacuo, followed by addition of
fresh solvent, effects quantitative conversion to 3. These results
are consistent with the formation of 3 and the cis isomer of 3 as
the major products of addition of CO to tBuPOCOIrHOTf.
The dehydrogenative coupling of vinyl-arenes to form (E,E)-1,4-
diaryl-1,3-butadienes has recently been reported using iPrPCPIrHCl
and iPrPCPIr(C2H4) as pre-catalysts (Scheme 5) [30]. Under identical
conditions to those reported for iPrPCPIrHCl, 1a/b was evaluated as
pre-catalyst in this reaction. It was anticipated that the reduced
steric congestion around the active site in the tBuPOCOIr fragment,
in comparison to iPrPCPIr, might correlate with increased activity.
However, using 5 mole% 1a/b, 2 equiv. of KOtBu (based on Ir),
and 654 mM styrene in toluene heated at 150 °C for 24 h in a
sealed ampoule under argon, only 40% conversion of styrene and
or prevented, by
a and b substitutions on the incoming vinyl arene
(Scheme 6). Also, instead of invoking loosely defined ‘‘electronic
differences”, it is possible that reduced activity of tBuPOCOIr results
from deactivation by cyclometalation of a tert-butyl group through
CAH addition to iridium, which has been reported to occur in
related tBuPOCOPIrHCl species in the presence of protic acids
[31]. Another deactivation pathway, perhaps more likely, results
from the dehydrogenative coupling of a phosphine tert-butyl
methyl group to a cyclometallated styrene molecule (iridaindene
moiety). Formation of such a complex was found to prevent any
turnovers for butadiene formation in attempts at dehydrogenative
coupling of vinyl arenes using tBuPCPIr, while iPrPCPIr efficiently
catalyzed the process [30]. Additionally, although the metal center
of iPrPCPIr is sterically congested in comparison to tBuPOCOIr, this
may actually hold the initial cyclometallated vinyl arene molecule
in an optimal geometry for addition of a second molecule, which
may be deviated from in tBuPOCOIr.
Scheme 3. Reaction of AgOTf with cis-2 and trans-2 to form 3.

M. Wilklow-Marnell, W.W. Brennessel / Polyhedron 160 (2019) 83–91
87
Scheme 4. Proposed equilibria of fluxional species 4 in solution.
Catalysis was first attempted in neat ethanol using a loading of
1% 1a/b, and 50 equiv. KOtBu (based on Ir), at 150 °C for 4 hours.
Under these conditions, 64% conversion was observed with a 27%
yield of n-butanol in 68% selectivity along with 8% and 2% yield
of n-hexanol and n-octanol, respectively (Table 1, entry 1). The
remaining conversion not accounted for by >C4 n-alcohols may
be mainly attributed to branched C6 and C8 alcohols, formation
of MOEt species by protonation of base (in this case protonation
gives tert-butanol), and somewhat surprisingly, only traces of ethyl
acetate (<1%). Selectivity for n-butanol is calculated as the percent
of n-butanol out of conversion products not resultant from base
protonation. Lowering of the catalyst loading was investigated
next. Using 0.1% 1a/b provided a 16% yield of n-butanol in 73%
selectivity, as well as 4% and 1% yields of n-hexanol and n-octanol
(Table 1, entry 2). Unlike many of the reported systems to date, this
catalyst does not present particularly high selectivity for n-butanol
at low conversions. Increasing the loading to 0.2% gave only 22%
yield of n-butanol, and conducting the reaction for 24 h gave neg-
ligible improvement (Table 1, entries 3 and 4). At 0.5% loading the
best yield of n-butanol using 1a/b was achieved, giving 31% yield in
64% selectivity (Table 1, entry 5). The yield of n-butanol actually
decreased to 26% when the reaction was conducted for 72 h,
though the yield of n-hexanol and n-octanol remained essentially
constant (Table 1, entry 6). This may be the result of competitive
dehydrogenation of C6 and C8 alcohols at higher conversions lead-
ing to even higher chain alcohols, or possibly insoluble oligo/poly-
meric side products not detected by GC analysis. Using lower
loadings of base proved ineffective (Table 1, entries 7 and 8), and
using more than 50 equiv. at 0.5% loading of 1a/b tended to give
heterogeneous reaction mixtures, and made post reaction analysis
especially difficult due to slurry or gel-like mixtures obtained upon
cooling.
Scheme 5. Iridium catalyzed dehydrogenative coupling of vinyl arenes.
Scheme 6. Proposed metalloindene intermediate showing steric interactions of
and b substitutions preventing iridium catalyzed dehydrogenative coupling of vinyl
arenes.
a
An iridium complex has also recently been reported as catalyst
in the Geurbet conversion of ethanol to n-butanol with remarkable
selectivity (>99%) utilizing bulky transition-metal hydroxides as
bases [32]. This reaction involves the dehydrogenation of a primary
alcohol to the corresponding aldehyde, base catalyzed aldol con-
densation to form an
a,b-unsaturated aldehyde, and subsequent
hydrogenation using the hydrogen removed in the dehydrogena-
tion step. Ideally, the overall reaction provides the coupling of
two alcohols into a larger chain alcohol with water as the sole
byproduct (Scheme 7). This reaction has garnered substantial
attention in recent years due to the enhanced fuel characteristics
of n-butanol and higher alcohols in comparison to ethanol
[33–35], and a number of advances have been made in the
homogenous catalysis of this process, mainly based on ruthenium
complexes [36–39], as well as two examples employing man-
ganese PNP pincer complexes [40,41]. However, to the knowledge
of the authors, no iridium pincer complexes have been reported as
catalysts for the Geurbet reaction of ethanol to form longer chain
alcohols. To begin filling this gap, tBuPOCOIr based pre-catalysts
have been evaluated in this process for formation of primary linear
alcohols, and found to give comparable yields to some of the best
reported systems in terms of n-butanol produced.
Pre-catalyst 1a/b was next assessed using other bases. Unex-
pectedly, sodium ethoxide, which is commonly used as a base in
the Geurbet process, gave nearly negligible conversion and no
detectable amounts of higher alcohols (Table 1, entry 9). Potassium
hydroxide performed better than sodium ethoxide, but gave a
lower yield of n-butanol compared to using KOtBu under otherwise
identical conditions, as well as producing a larger amount of
unidentified side-products (Table 1, entry 10). In light of the
unmatched selectivity for n-butanol reported using Cp*Ir[(2-OH-
6-phenyl)pyridine]Cl and bulky Ni or Cu transition metal
hydroxide bases [32], [Tp0Ni(
l
ꢀ OH)]2 was also tested as a base
in the tBuPOCOIr catalyzed Geurbet system. However, using 1a/b
as pre-catalyst no formation of n-butanol, or conversion of ethanol,
was observed and a coating of Ni metal was seen to form on the
inner surface of the reaction vessel (Table 1, entry 11). The rela-
tively bulky organic bases sodium triphenylmethoxide and sodium
tricyclohexylmethoxide were also tested, providing diminished or
negligible yields, respectively (Table 1, entries 12 and 13). To date,
the most active homogenous catalyst reported for the Geurbet pro-
cess of ethanol is a Ru PNP pincer complex which is believed to
Scheme 7. Overall stoichiometry in the Geurbet reaction of primary alcohols.

88
M. Wilklow-Marnell, W.W. Brennessel / Polyhedron 160 (2019) 83–91
Table 1
Catalytic conversion of ethanol to higher alcohols using tBuPOCOIr based catalysts.
Entry Catalyst loading (%) Base (equiv. based on Ir) Time (h) Conversion (%)a Yield n-BuOH (%)b/Selectivity (%)c Yield n-HexOH (%) Yield n-OctOH (%)
1
2
3
4
5
6
7
8
1
0.1
0.2
0.2
0.5
0.5
0.5
0.5
0.5
0.2
0.2
0.5
KOtBu (50)
KOtBu (50)
KOtBu (50)
KOtBu (50)
KOtBu (50)
KOtBu (50)
KOtBu (1)
KOtBu (10)
NaOEt (50)
KOH (50)
[TpNiOH]2 (50)
Ph3CONa
(C6H11)3CONa (50)
KOtBu (50)
KOtBu (50)
4
4
4
24
4
72
4
4
4
4
64
23
38
35
62
75
<1
20
3
41
0
63
23
69
64
27/68
16/73
22/79
23/71
31/64
26/52
n/a
14/75
n/a
14/60
n/a
8
4
6
6
9
9
n/a
3
n/a
4
n/a
4
n/a
11
8
2
1
1
1
2
2
n/a
<1
n/a
1
n/a
<1
n/a
3
9
10
11
12
13
14
15
4
4
4
4
17/46
<1/n/a
33/62
26/65
0.5
0.5 (Ir(CO)HCl)
0.5 (1 atm H2)
4
2
Conditions: 1a/b used as pre-catalyst unless otherwise noted, 150 °C, neat ethanol, argon atmosphere.
a
Disappearance of ethanol.
Determined by GC, tridecane used as standard.
Percent out of conversion not accounted for by base protonation.
b
c
retain a CO ligand during catalysis [38]. To see what effect the pres-
ence of a CO ligand might have on the tBuPOCOIr catalyzed reaction,
a mixture of cis/trans-2 was generated in situ by the addition of
1 atm CO to 1a/b in toluene, removal of volatiles, and then addition
of base and ethanol. Indeed, the highest yield of n-butanol was
obtained using cis/trans-2 as pre-catalyst (33%), as well as the
highest overall yield of n-alcohols (47%) (Table 1, entry 14). The
highest reported yield of n-butanol by a homogenous catalyst in
this process is 38.4% [38]. Lastly, the reaction was conducted under
1 atm of hydrogen in the hopes that selectivity might be increased
due to re-hydrogenation of C4 and greater aldehydes as they are
formed, and the statistically greater probability of ethanol dehy-
drogenation at moderate conversion levels would be predominant
(Table 1, entry 15). Unfortunately, the yield of >C4 alcohols was
barely effected, while the yield of n-butanol was reduced 16% com-
pared to identical conditions without H2 added. However, this
modest inhibition in the presence of a significant amount of H2
would seem to indicate that the tBuPOCOIr catalyst may be a candi-
date for the acceptorless dehydrogenation of alcohols given an effi-
cient means of expelling hydrogen from the reaction system.
In light of these findings, the mechanism of this reaction came
in to question, and in this interest the stoichiometric reactivity of
tBuPOCOIr with ethanol was investigated. Upon dehydrohalogena-
tion of 1a/b by KOtBu in C6D6 a mixture of products with broad res-
onances is observed by 31P NMR spectroscopy. The associated 1H
NMR spectrum presents extensively overlapped and broadened
peaks quite unaccommodating to characterization, likely due to
reversible dimerization and solvent coordination. Formation of
dinuclear clusters, and difficult to characterize mixtures, have been
reported for variants of tBuPCPIr where the steric congestion
around iridium is reduced by replacing only two phosphine tert-
butyl groups with methyl substituents [42]. Only very slight traces
of 2 hydride resonances were observed. Addition of ethanol only
slightly improves matters and addition of 1 equiv. to this mixture
in C6D6 gives rise to two definite major products (out of >10 seen)
in an approximately 1:6 ratio as observed by 31P NMR spectroscopy
(d 190.6, 167.9). What appears to be a quartet signal for the CH2
protons of coordinated, or OAH activated, ethanol (d 3.83,
JHH = 7.1 Hz) is also observed in the 1H NMR spectrum. The iridium
alkoxide Cp*IrPPh3H(OEt) is reported to display a OCH2CH3 reso-
nance at d 3.79 (JHH = 6 Hz) [43]. Furthermore, the signal for the
OH proton of ethanol, which was observed at d 4.00 in the 1H
NMR spectrum of a sample of pure ethanol in C6D6, is completely
absent. These data are consistent with formation of an Ir-ethoxide
complex. However, it is difficult to determine conclusively if this
species is resultant from -O coordination, or OAH addition as sus-
j
pected, because of the inability to reliably assign a hydride signal
to the species responsible for the quartet observed due to the myr-
iad hydride resonances found in the 1H NMR spectrum, as well as
extensive signal overlap preventing location of the putative Ir-
OCH2CH3 protons. After addition of 10 equiv. EtOH, and heating
for ꢁ10 min at 150 °C, a quartet consistent with the C(sp2)-H pro-
ton of acetaldehyde is also noted at d 9.05 (JHH = 2.8 Hz). Attempts
to crystallize potential intermediates from these mixtures were
unsuccessful.
Given these results, a plausible, though certainly highly simpli-
fied, mechanism for ethanol dehydrogenation by tBuPOCOIr may be
proposed which follows a ‘‘classic” pathway of oxidative addition,
b-hydride elimination, dissociation of product, and reductive elim-
ination of hydrogen or, most likely in this case, hydrogenation of
another substrate molecule such as the crotonaldehyde formed
by aldol condensation of acetaldehyde (Scheme 8). While extensive
Scheme 8. A plausible, though simplified, catalytic cycle for the Geurbet reaction of
ethanol using tBuPOCOIr.

M. Wilklow-Marnell, W.W. Brennessel / Polyhedron 160 (2019) 83–91
89
speciation though dimeric/oligomeric cluster formation, as well as
facile reversible coordination/activation of solvent or substrate, is
readily apparent in this system, it is possible that these equilibria
represent unproductive pathways outside of the catalytic cycle,
and present a potential means for catalytic improvement by pre-
vention of their formation by steric or electronic modifications.
For example, the good performance of tBuPOCOIr(CO) (generated
from cis/trans-2 and base) indicates that the presence of an L-type
ligand may help to stabilize the active catalytic species with
respect to non-productive side equilibria or deactivation pathways,
and careful selection of the appropriate ligand might further
improve efficiency.
are added followed by 50 mL THF. Rapid bubbling is noted and
the mixture stirred at room temperature for 30 min. The flask is
then fit with a septum sealed condenser and refluxed for 1 h under
argon. After this period the reaction mixture is allowed to cool to
room
temperature
and
1.00 g
(5.54 mmol)
di-tert-
butylchlorophosphine dissolved in 20 mL THF is then added by syr-
inge. The mixture is then refluxed an additional 1 h, before stirring
overnight at room temperature. All volatiles are then removed
under reduced pressure and the residue extracted with THF
(3 ꢃ 15 mL), which is filtered and then concentrated in vacuo to
provide 1.342 g (97% yield) of a viscous slightly yellow oil. Com-
plete removal of excess di-tert-butylchlorophosphine was found
to require repeated dissolution in THF followed by concentration
in vacuo. Crystals suitable for X-ray diffraction were found to spon-
taneously form from this oil after several weeks of storage. 31P{1H}
NMR (C6D6): d 155.0 (s). 1H NMR (C6D6): d 8.03 (vq, 1H, aryl), 7.36
(d, JHH = 8.4 Hz, 2H, meta-aryl), 7.02 (t, JHH = 7.9 Hz, 1H, para-aryl),
3. Conclusions
In conclusion, a new POCO-type pincer ligand has been synthe-
sized, characterized, and metallated using [Ir(COD)Cl]2 to form 1a
t
2.11 (s, 3H, CH3), 1.09 (d, JPH = 12 Hz, 18H, Bu-H). Anal. Calc. for
(l
-Cl-[tBuPOCOIrHCl]2) and isomer 1b. The chemical equivalency
C16H25O2P: C, 68.55; H, 8.99. Found: C, 68.37; H, 8.94%.
of 1a and 1b has been demonstrated by reaction of 1a/b with AgOTf
and CO, irrespective of order of addition, resulting in convergence
to 3 as the sole product. The performance of 1a/b as a pre-catalyst
for the dehydrogenative coupling of vinyl arenes has been investi-
gated, giving poor yields in comparison to iPrPCPIr, demonstrating
that reduced steric hindrance does not inherently correlate with
increased activity in this process, though performance in olefin iso-
merization was enhanced with respect to iPrPCPIr. Both 1a/b and
cis/trans-2 were evaluated in the Geurbet conversion of ethanol
to linear alcohols, and optimal results were obtained using a 0.5%
catalyst loading of cis/trans-2 in conjunction with 50 equiv. KOtBu
as base, giving 33% yield of n-butanol and an overall 47% yield of n-
alcohols. Further investigation is currently underway in terms of
mechanistic understanding of the tBuPOCOIr catalyzed Geurbet
reaction, reaction condition and ligand modifications aimed at
increased Geurbet performance, and other catalytic processes
involving H-element bond manipulation.
4.2. Synthesis of 1a/b
To a 50 mL round-bottom Schlenk flask, 0.25 g (0.37 mmol) of
[Ir(COD)Cl]2 is added along with 20 mL toluene providing a clear
orange solution. To this, 0.208 g (0.74 mmol) tBuPOCOH dissolved
in 5 mL toluene is added, accompanied by a color change to deep
reddish brown. The flask is then fit with a septum sealed con-
denser and the reaction solution is refluxed for 24 h under argon
atmosphere. After this period the reaction mixture is allowed to
cool to room temperature, and an orange-red supernatant is seen
over a large amount of orange solid. Concentration of this mix-
ture to dryness in vacuo, and washing of the obtained solids with
pentane (3 ꢃ 5 mL), provides 0.315 g (84% yield) of an orange
solid after drying under reduced pressure. Dissolving in hot
toluene followed by slow cooling provided single crystals suitable
for X-ray diffraction which proved to be 1a, though they were
coated in a fine polycrystalline material. The solubility of this
4. Experimental
product is rather poor, and
a large number of scans were
required for acceptable NMR spectra. Best results were obtained
in CDCl3. 31P{1H} NMR (CDCl3): d 156.5 (s, 1a), 154.7 (s, 1b). 1H
NMR (CDCl3) All downfield peaks are found to overlap, ranges
reported: d 7.28–7.1 (m, aryl H’s), 6.81–6.61 (m, aryl H’s), 2.71
(2 overlapping bs, CH3), 1.61 (m, Bu-H), 1.26 (m, Bu-H), ꢀ27.3
(d, JPH = 24.4 Hz, Ir-H 1a), ꢀ27.5 (d, JPH = 26.8 Hz, Ir-H, 1b).
Anal. Calc. for [C16H25ClIrO2P]2: C, 37.83; H, 4.96. Found: C,
38.42; H, 4.92%.
All manipulations were carried out under argon atmosphere
either in a Vacuum Atmospheres glove box, or by modified Schlenk
techniques. All NMR spectra were collected on a Bruker AMX
400 MHz, or JEOL JNM-ECZS 400 MHz spectrometer. All 31P NMR
spectra were referenced to external H3PO4. Proton NMR spectra
were referenced to residual deuterated solvent signal. DOSY-NMR
were collected using a 16 point linear array starting at a gradient
of 3.0 mT/m, final gradient of 0.27 T/m, and with a diffusion time
of 0.125 s. All aromatic, alkane, and ether solvents were dried over
sodium/benzophenone, distilled from the resultant purple solution
prior to use, and stored over 3 Å molecular sieves. CDCl3, CHCl3,
CH2Cl2, and EtOH were dried/stored with 3 Å molecular sieves acti-
vated by heating at 250 °C under vacuum until a constant pressure
of ꢁ10 mTorr was reached. All other reagents were used as
received from commercial sources without further purification.
X-ray structure collection was conducted on a Bruker SMART APEX
II CCD platform diffractometer. GC was carried out on a Shimadzu
t
t
4.3. Formation of cis/trans-2 and isomerization to solely trans-2
A saturated solution of 1a/b is generated in arene solvent or
chloroform by heating ꢁ15 mg 1a/b in 1 mL of solvent at 80–
130 °C in a sealed ampoule under argon followed by filtration.
Then, 0.5 mL of this solution is added to a J-Young NMR tube which
is degassed by 3 freeze–pumpthaw cycles prior to addition of
1 atm CO. A color change from orange to pale yellow is observed
within <1 min after mixing. 31P NMR spectroscopy of this solution
revealed formation of cis/trans-2 in a 1–0.35 ratio. Subsequent
removal of CO by concentration in vacuo, addition of fresh solvent,
and heating at 60 °C for 6 min provides trans-2 in quantitative
yield. Removal of solvent gives a yellow solid. Analytical data
reported for trans-2. 31P{1H} NMR (CDCl3): d 163.9 (s). 1H NMR
(CDCl3): d 7.30 (d, JHH = 7.2 Hz, 1H, aryl), 6.95 (m, 2H, aryl), 2.70
(s, 3H, CH3), 1.39 (d, JPH = 16.0 Hz, 9H, tBu-H), 1.22 (d, JPH = 15.6 Hz,
9H, tBu-H), ꢀ19.81 (d, JPH = 20.4 Hz, 1H, Ir-H). Anal. Calc. for
GC-2010 with a DB-WAXetr column (30 m ꢃ 0.25 mm ID, 0.50
lm
film) at 50–250 °C, 4.76 mL/min flow. All GCMS utilized a Thermo
Fisher Scientific Focus-GC and DSQ-II MS with a TR-5MS column
(30 m ꢃ 0.25 mm ID ꢃ 0.1
lm film) at 40–260 °C, 3 mL/min flow.
4.1. Synthesis of tBuPOCOH (3-(di-tert-butylphosphinito)
acetophenone)
To a 100 mL round-bottom Schlenk flask, 0.675 g (4.96 mmol)
3-hydroxyacetophenone, 120 mg (5.00 mmol) NaH, and a stir bar
C17H25ClIrO3P: C, 38.09; H, 4.70. Found: C, 37.65; H, 4.63%.

90
M. Wilklow-Marnell, W.W. Brennessel / Polyhedron 160 (2019) 83–91
4.4. Synthesis of 3
Acknowledgements
To
a 25 mL round-bottom ampoule, a stir bar, 25 mg
This paper is submitted in honor of the 65th birthday of William
D. Jones, whom the authors gratefully thank for thoughtful discus-
sions, guidance, and friendship. This work was partially supported
by the NSF under the CCI Center for Enabling New Technology
through Catalysis (CENTC), CHE-1205189, and by SUNY New Paltz.
(0.0492 mmol by monomer) 1a/b, and 20 mL CH2Cl2 are added.
The sealed vessel is sonicated for 1 hr to dissolve all solids. The
solution is then degassed by 3 freeze–pumpthaw cycles and
1 atm CO applied, providing a light yellow solution after briefly
mixing. The solution is then subject to an additional freeze–
pumpthaw cycle before adding 13 mg (0.0506 mmol) AgOTf and
stirring overnight. After this time, the mixture is filtered to remove
AgCl and any excess AgOTf and then concentrated to dryness in
vacuo. Pentane is then added and the resultant mixture re-filtered
to further ensure removal of AgOTf. After evaporation of solvent
30.7 mg of an orange-yellowish solid is obtained (96% yield). Vapor
diffusion of pentane into a concentrated benzene solution of this
product over 2 days, followed by storing at ꢀ17 °C afforded crys-
tals suitable for X-ray diffraction which contained ½ equivalent
co-crystallized benzene per Ir. 31P{1H} NMR (THF-d8): d 162.9 (s).
1H NMR (THF-d8): d 7.59 (m, 1H, aryl), 7.19 (m, 2H, aryl), 2.8 (s,
Appendix A. Supplementary data
CCDC 1859925–1859927 contains the supplementary crystallo-
graphic data for tBuPOCOH, and complexes 1a and 3. These data can
be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/
retrieving.html, or from the Cambridge Crystallographic Data Cen-
tre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223 336
033; or e-mail: deposit@ccdc.cam.ac.uk. Supplementary data to
this article can be found online at https://doi.org/10.1016/j.poly.
2018.12.001.
t
3H, CH3), 1.41 (d, JPH = 16 Hz, 9H, Bu-H), 1.34 (d, JPH = 16 Hz, 9H,
tBu-H), ꢀ26.0 (d, JPH = 21.2 Hz, 1H, Ir-H). 19F NMR (C6D6): d ꢀ77.2
(s). Anal. Calc. for C18H25F3IrO6PSꢄ0.5 C6H6: C, 36.62; H, 4.10. Found:
C, 35.95; H, 3.93%.
References
[1] M. Albrecht, G. van Koten, Angew. Chem., Int. Ed. 40 (2001) 3750.
[2] M.E. van der Boom, D. Milstein, Chem Rev. 103 (2003) 1759.
[3] D. Morales-Morales, Rev. Soc. Quim. Méx. 48 (2004) 338.
[4] J.T. Singelton, Tetrahedron 59 (2003) 1837.
[5] M. Asay, D. Morales-Morales, Dalton Trans. 44 (2015) 17432.
[6] K.J. Szabó, O.F. Wendt (Eds.), Pincer and Pincer-Type Complexes: Applications
in Organic Synthesis and Catalysis, Wiley-VCH Verlag GmbH & Co. KGaA, 2014,
https://doi.org/10.1002/9783527681303.
[7] M.A.W. Lawrence, K. Green, P.N. Nelson, S.C. Lorraine, Polyhedron 143 (2018)
11.
[8] S. Nemeh, C. Jensen, E. Binamira-Soriaga, W.C. Kaska, Organometallics 2 (1983)
1442.
[9] M. Gozin, A. Welsman, Y. Ben-David, D. Milstein, Nature 364 (1993) 699.
[10] D.G. Gusev, M. Madott, F.M. Dolgushin, K.A. Lyssenko, M.Y. Antipin,
Organometallics 19 (2000) 1734.
[11] F. Miyazaki, K. Yamaguchi, M. Shibasaki, Tetrahedron Lett. 40 (1999) 7379.
[12] D. Morales-Morales, R. Redón, C. Yung, C.M. Jensen, Chem. Commun. (2000)
1619.
[13] R.B. Bedford, S.M. Draper, P.N. Scully, S.L. Welch, New J. Chem. 24 (2000) 745.
[14] S. Fu, N. Chen, X. Liu, Z. Shao, S. Luo, Q. Liu, J. Am. Chem. Soc. 138 (2016) 8588.
[15] S. Chakraborty, P.O. Lagaditis, M. Förster, E.A. Bielinski, N. Hazar, M.C.
Holthausen, W.D. Jones, S. Schneider, ACS Catal. 4 (2014) 3994.
[16] V. Papa, J.R. Cabrero-Antonino, E. Alberico, A. Spanneberg, K. Junge, H. Junge,
M. Beller, Chem. Sci. 8 (2017) 3576.
4.5. Synthesis of 4
To a small Erlenmeyer flask, 25.0 mg (0.0492 mmol) 1a/b, a stir
bar, and 10 mL toluene are added. This mixture is stirred, creating a
suspension, and 12.6 mg (1 equiv.) AgOTf is then added followed
by stirring overnight. After this period the mixture is filtered with
a fine frit to remove AgCl, and concentrated in vacuo. Once ‘‘dry” a
hot tap water bath was applied for ꢁ45 min while still under vac-
uum. This provided 30.0 mg of an orange-yellow solid (98% yield).
Due to the highly fluxional nature of this complex spectral data are
reported for 4 in the presence of 1.2 equiv. AgOTf. 31P{1H} NMR
(C6D6): d 154.7 (s). 1H NMR (C6D6): 6.8 (d, JHH = 7.6, 1H, aryl),
6.65 (d, JHH = 7.6 Hz, 1H, aryl), 6.51 (t, JHH = 7.8 Hz, 1H, aryl), 2.3
(s, 1H, CH3), 1.36 (d, JPH = 14.8 Hz, 9H, tBu-H), 1.19 (d, JPH = 15.2 Hz,
t
9H, Bu-H), ꢀ31.30 (vd, JPH = 23.6 Hz, 1H, Ir-H). 19F NMR (C6D6): d
ꢀ77.09 (bs), ligated and free OTf are in exchange and likely
[17] J. Choi, A.H.R. MacArthur, M. Brookhart, A.S. Goldman, Chem. Rev. 111 (2011)
1761.
overlap.
[18] M. Albrecht, D. Morales-Morales, Pincer-type iridium complexes for organic
transformations, in: L.A. Oro, C. Claver (Eds.), Iridium Complexes in Organic
Synthesis, Wiley-VCH Verlag GmbH & Co. KGaA, 2009, pp. 299–323.
[19] C.M. Jensen, Chem. Commun. (1999) 2443.
4.6. General procedure for dehydrogenative coupling
[20] M. Gupta, C. Hagen, R.J. Flesher, W.C. Kaska, C.M. Jensen, Chem. Commun.
(1996) 2083.
[21] I. Gӧttker-Schnetmann, P. White, M. Brookhart, J. Am. Chem. Soc. 126 (2004)
1804.
[22] W.-W. Xu, G.P. Rosini, M. Gupta, C.M. Jensen, W.C. Kaska, K. Krogh-Jespersen,
A.S. Goldman, Chem. Commun. (1997) 2273.
[23] F. Liu, A.S. Goldman, Chem. Commun. (1999) 655.
[24] M.C. Denney, V. Pons, T.J. Hebden, D.M. Heinekey, K.I. Goldberg, J. Am. Chem.
Soc. 128 (2006) 12048.
[25] D. Morales-Morales, R. Redón, C. Yung, C.M. Jensen, Inorg. Chim. Acta. 357
(2004) 2953.
[26] W. Yao, Y. Zhang, X. Jia, Z. Huang, Angew. Chem., Int. Ed. 53 (2014) 1390.
[27] M. Kanzelberger, B. Singh, M. Czerw, K. Krogh-Jespersen, A.S. Goldman, J. Am.
Chem. Soc. 122 (2000) 11017.
[28] C.J. Moulton, B.L. Shaw, J. Chem. Soc., Dalton Trans. (1976) 1020.
[29] J.M. Goldberg, G.W. Wong, K.E. Brastow, W. Kaminsky, K.I. Goldberg, D.M.
Heinekey, Organometallics 34 (2015) 753.
[30] M. Wilklow-Marnell, B. Li, T. Zhou, K. Krogh-Jespersen, W.W. Brennessel, T.J.
Emge, A.S. Goldman, W.D. Jones, J. Am. Chem. Soc. 139 (2017) 8977.
[31] A.V. Polukeev, S.A. Kuklin, P.V. Petrovskii, A.S. Peregudov, F.M. Dolgushin, M.G.
Ezernitskaya, A.A. Koridzea, Russ. Chem. Bull., Int. Ed. 59 (2010) 745.
[32] S. Chakraborty, P.E. Piszel, C.E. Hayes, R. Tom Baker, W.D. Jones, J. Am. Chem.
Soc. 137 (2015) 14264.
[33] P. Durre, Biotechnol. J. 2 (2007) 1525.
[34] S. Atsumi, A.F. Cann, M.R. Connor, C.R. Shen, K.M. Smith, M.P. Brynildsen, K.J.Y.
Chou, T. Hanai, J.C. Liao, Metab. Eng. 10 (2008) 305.
To
a 4 mL glass ampoule with Teflon closure, 9.0 mg
(0.0177 mmol) 1a/b, 4 mg (0.0356 mmol) potassium tert-butoxide,
a stir bar, and 0.5 mL toluene are added. This is followed by addi-
tion of 0.354 mmol substrate. The sealed vessel is then heated,
with stirring, at 150 °C for a given amount of time in an aluminum
heating block. After this period the reaction is cooled to room tem-
perature, and then 80
mixture 1.5 L is withdrawn and diluted with THF to 2 mL in a vol-
umetric flask for GCMS analysis.
lL dodecane is added and mixed. From this
l
4.7. General procedure for Geurbet reaction of ethanol
To a 4 mL glass ampoule with Teflon closure, a given amount of
catalyst and base are added along with a stir bar. Then, 0.3 mL dry,
argon purged, ethanol is added by syringe under positive argon
pressure and the vessel is quickly sealed. This is then heated at
150 °C, with stirring, for a prescribed amount of time. After the
reaction time is complete and reaction cooled to room tempera-
ture, 1–3 mL THF is added and mixed, followed by filtration
through a short plug of celite. Then, 120
to 6 mL THF, along with 40 L tridecane, for GC analysis.
lL of the filtrate is added
[35] B.G. Harvey, H.A. Meylemans, J. Chem. Technol. Biotechnol. 86 (2011) 2.
[36] R.L. Wingad, P.J. Gates, S.T.G. Street, D.F. Wass, ACS Catal. 5 (2015) 5822.
l

M. Wilklow-Marnell, W.W. Brennessel / Polyhedron 160 (2019) 83–91
91
[37] R.L. Wingad, E.J.E. Bergstrӧm, M. Everett, K.J. Pellow, D.F. Wass, Chem.
Commun. 52 (2016) 5202.
[38] Yinjun Xie, Y. Ben-David, L.J.W. Shimon, D. Milstein, J. Am. Chem. Soc. 138
(2016) 9077.
[39] H. Aitchison, R.L. Wingad, D.F. Wass, ACS Catal. 6 (2016) 7125.
[40] S. Fu, Z. Shao, Y. Wang, Q. Liu, J. Am. Chem. Soc. 139 (2017) 11941.
[41] N.V. Kulkarni, W.W. Brennessel, W.D. Jones, AC S Catal. 8 (2018) 997.
[42] S. Kundu, Y. Choliy, G. Zhuo, R. Ahuja, T.J. Emge, R. Warmuth, M. Brookhart, K.
Krogh-Jespersen, A.S. Goldman, Organometallics 28 (2009) 5432.
[43] D.S. Glueck, L.J. Newman Winslow, R.G. Bergman, Organometallics 10 (1991)
1462.