Regioselective Gas-Phase n-Butane Transfer Dehydrogenation via Silica-Supported Pincer-Iridium Complexes

Article abstract of DOI:10.1002/cctc.202001399

The production of olefins via on-purpose dehydrogenation of alkanes allows for a more efficient, selective and lower cost alternative to processes such as steam cracking. Silica-supported pincer-iridium complexes of the form [(≡SiO?^R4POCOP)Ir(CO)] (^R4POCOP=κ³-C₆H₃-2,6-(OPR₂)₂) are effective for acceptorless alkane dehydrogenation, and have been shown stable up to 300 °C. However, while solution-phase analogues of such species have demonstrated high regioselectivity for terminal olefin production under transfer dehydrogenation conditions at or below 240 °C, in open systems at 300 °C, regioselectivity under acceptorless dehydrogenation conditions is consistently low. In this work, complexes [(≡SiO?^tBu4POCOP)Ir(CO)] (1) and [(≡SiO?^iPr4PCP)Ir(CO)] (2) were synthesized via immobilization of molecular precursors. These complexes were used for gas-phase butane transfer dehydrogenation using increasingly sterically demanding olefins, resulting in observed selectivities of up to 77 %. The results indicate that the active site is conserved upon immobilization.

Full text of DOI:10.1002/cctc.202001399

The European Society Journal for Catalysis
Supported by
Accepted Article
Title: Regioselective Gas-Phase n-Butane Transfer Dehydrogenation
via Silica-Supported Pincer-Iridium Complexes
Authors: Boris Sheludko, Cristina F. Castro, Chaitanya A. Khalap,
Thomas J. Emge, Alan S. Goldman, and Fuat E. Celik
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.202001399
Link to VoR: https://doi.org/10.1002/cctc.202001399
01/2020

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
Regioselective Gas-Phase n-Butane Transfer Dehydrogenation
via Silica-Supported Pincer-Iridium Complexes
Boris Sheludko,[a],[b] Cristina F. Castro,[b] Chaitanya A. Khalap,[b] Thomas J. Emge,[a] Alan S.
Goldman,*[a] Fuat. E. Celik*[b]
[a]
[b]
Boris Sheludko, Dr. Thomas. J. Emge, Prof. Alan S. Goldman
Department of Chemistry and Chemical Biology
Rutgers, The State University of New Jersey
123 Bevier Road, Piscataway, NJ 08854
E-mail: alan.goldman@rutgers.edu
Boris Sheludko, Cristina F. Castro, Chaitanya A. Khalap, Prof. Fuat E. Celik
Department of Chemical and Biochemical Engineering
Rutgers, The State University of New Jersey
98 Brett Road, Piscataway, NJ 08854
E-mail: fuat.celik@rutgers.edu
Abstract: The production of olefins via on-purpose dehydrogenation
of alkanes allows for a more efficient, selective and lower cost
alternative to processes such as steam cracking. Silica-supported
pincer-iridium complexes of the form [(≡SiO-R4POCOP)Ir(CO)]
We have previously shown that silica-supported pincer-
iridium complexes are effective acceptorless dehydrogenation
catalysts (Figure 1, top).[2i] The stability of such catalysts at
elevated temperatures up to 300 °C arises from the concurrent
generation of CO over these catalysts, which prevents thermal
decomposition.[5] However, the selectivity of the system is
unremarkable, resulting in an equilibrated mixture of product
butenes (ca. 20% 1-butene) even at very low residence times,
possibly due to rapid isomerization over the iridium species.[2i, 6]
Previously, regioselectivity has been achieved using molecular
pincer-iridium species in the heterogeneous (gas-solid) propene-
butane transfer dehydrogenation of linear alkanes (e.g. up to ca.
65% 1-butene[2l]), even in gas-solid heterogeneous reactions,
while also achieving turnovers up to 2000 hr-1 at early reaction
times at 240 °C.[2l] However, at longer reaction times, the fraction
of 1-alkene declines due to isomerization to generate internal
butenes. To harness the intrinsic regioselectivity of such
catalysts, one possible approach is to use a plug-flow reactor to
shorten the residence time of the feed with the catalyst, thus
limiting the amount of isomerization that takes place.
(
R4POCOP = κ3-C6H3-2,6-(OPR2)2) are effective for acceptorless
alkane dehydrogenation, and have been shown stable up to 300 °C.
However, while solution-phase analogues of such species have
demonstrated high regioselectivity for terminal olefin production
under transfer dehydrogenation conditions at or below 240 °C, in
open systems at 300 °C, regioselectivity under acceptorless
dehydrogenation conditions is consistently low. In this work,
complexes [(≡SiO-tBu4POCOP)Ir(CO)] (1) and [(≡SiO-iPr4PCP)Ir(CO)]
(2) were synthesized via immobilization of molecular precursors.
These complexes were used for gas-phase butane transfer
dehydrogenation using increasingly sterically demanding olefins,
resulting in observed selectivities of up to 77%. The results indicate
that the active site is conserved upon immobilization.
Introduction
Olefins are versatile building blocks in the synthesis of
various products spanning from fine chemicals to bulk polymers
to fuels.[1] In contrast, alkanes are among the most inert organic
species. Nevertheless, alkane activation is currently carried out
on an industrial scale, typically at very high temperatures (600 –
1000 °C) that lead to reduced selectivity and formation of coke
which, in between cycles, must be burned off prior to further
reaction. Alkane dehydrogenation under milder conditions would
be beneficial due to reduced energy demand and an opportunity
to modulate the selectivity of the reaction.
Pincer-iridium complexes of the form [(R4PCP)Ir] (R4PCP =
κ3-C6H3-2,6-(XPR2)2; X = CH2, O; R = Bu, Pr) have received
great attention in the context of hydrocarbon functionalization
and have found application in reactions such as alkane
dehydrogenation and alkane metathesis among other C-H
t
i
Figure 1. Silica-supported species 1 and 2 used in this work, prepared via
immobilization of molecular precursors.
activation
reactions,[2]
dehydroaromatization,[3]
and
dehydrogenative C-C coupling.[4] Thermal (acceptorless)
dehydrogenation can be performed using these species at rates
of > 700 hr-1 at 200 °C.[2a-f, 2h-n] However, such processes are
highly endothermic and thermodynamically unfavorable at lower
temperatures. To counter this limitation, a sacrificial hydrogen
acceptor may be added to the system to effect transfer
dehydrogenation which is approximately thermoneutral.[2a-o]
While previous work has made use of silica-tethered [(≡SiO-
tBu4POCOP)Ir(CO)] (1) in such continuous-flow systems,
complexes based on the [(iPr4PCP)Ir] framework are generally
more active dehydrogenation catalysts presumably largely due
to the less sterically hindered iridium environment. It has
previously been shown that differences in the steric environment
1
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
of the iridium center significantly influence the activity of pincer-
iridium species. For example, [(iPr4PCP)Ir]-based catalysts
typically show alkane dehydrogenation activities an order of
magnitude or more greater than [(tBu4PCP)Ir]-based catalysts.[7]
In the model reaction of pentane-propene transfer
dehydrogenation, the former complex achieved an initial rate of
ca. 3800 TO hr-1 as compared with 120 TO hr-1 by the latter
species.[2l] However, decreasing the size of the phosphinoalkyl
substituents in complexes of type [(R4PCP)Ir] (where R is smaller
than i-Pr) can lead to formation of clusters in absence of a
trapping ligand.[7a, 8] The site-isolation intrinsic to silica-supported
pincer-iridium species appears to be a possible strategy that
would allow easier access to such sterically unhindered
iridium(I) species while mitigating intramolecular interactions
leading to their dimerization or oligomerization. In pursuit of this,
we chose to synthesize [(≡SiO-iPr4PCP)Ir(CO)] (2) via tethering[9]
of the molecular precursor [(tBu2PO-iPr4PCP)Ir(CO)] and to
explore the catalytic properties of the former in a model light-
alkane dehydrogenation system at high temperatures (300 °C).
Figure 2. ORTEP diagram of novel pincer-iridium complex [(HO-
iPr4PCP)IrHCl(THF)] with thermal ellipsoids drawn at the 50% probability level.
Only one molecule of the asymmetric unit cell is shown, and hydrogen atoms
apart from ligand hydride are omitted for clarity. Selected bond lengths (Å) and
bond angles (deg): Ir1-Cl1: 2.496(5), Ir1-P1: 2.284(4), Ir1-P2: 2,296(4), Ir1-C1:
2.03(2), Ir1-O2: 2.29(1), O1-C4: 1.41(2), C1-Ir1-Cl1: 175.8(5), P1-Ir1-P2:
162.7(2), C1-Ir1-O2: 90.6(6).
Results and Discussion
Synthesis of [(tBu2PO-iPr4PCP)Ir(CO)]
Catalyst 2, an analogue to [(iPr4PCP)Ir] previously used in
solution and gas/solid[2l] phase alkane dehydrogenation
reactions, was synthesized in an attempt to obtain increased
reactivity compared with 1.[10]
Alkane Transfer Dehydrogenation
Alkane dehydrogenation is a highly endergonic process and
the occurrence of back reaction is a critical consideration in the
experimental design. It is expected that a decreasing residence
time would limit competition from the back reaction and, in the
limit of no back reaction, the rate of the forward reaction may be
measured accurately. Indeed, this ability to measure intrinsic
reaction rates is one benefit of a plug-flow reactor setup.
Surprisingly, the product distribution arising from
acceptorless butane dehydrogenation catalyzed by 1 at 340 °C
was independent of the butane residence time over two orders
of magnitude (τ = 0.0016 – 0.000064 molcatalyst min L-1).[2i]
Isomerization of 1-butene was studied over 1 and 2 in more
detail, revealing rates of several thousand net forward turnovers
per hour (Table S7).
The pre-catalyst [(HO-iPr4PCP)IrHCl] was synthesized via
metalation of the HO-iPr4PCP ligand with [Ir(COD)Cl]2. The ligand
was synthesized via
a
protocol analogous to reported
methods.[11] Recrystallization of the HCl complex from a solution
in THF and hexanes at -30 °C led to formation of colorless X-ray
quality crystals of THF-adduct [(HO-iPr4PCP)IrHCl(THF)], which
enabled verification of the proposed structure (See Supporting
Information section S1[11] and figure 2). CCDC-2023539 contains
the supplementary crystallographic data for this paper. These
data can be obtained free of charge from The Cambridge
Crystallographic
Data
Centre
via
www.ccdc.cam.ac.uk/data_request/cif.
The unit cell contains two conformational isomers. In both
cases the THF and hydride ligands are mutually trans,
consistent with their respective very weak and very strong trans
influences.
In comparing the activity of 1 and 2 under transfer
dehydrogenation conditions, the nature of the hydrogen acceptor
and its partial pressure were varied. The three olefin hydrogen
acceptors tested were ethylene, propene and 3,3-dimethyl-1-
butene (tert-butylethylene, TBE). The carbonyl derivatives 1 and
2 were chosen since it was previously shown that the ethylene-
bound analogues are converted to carbonyl-bound species
The hydride and phosphorus atoms of [(HO-iPr4PCP)IrHCl]
afford signals at  -37.13 and 58.08 ppm, in the 1H and 31P NMR
spectra respectively, essentially equal to those previously
reported for [(MeO-iPr4PCP)IrHCl],[13]  -37.06 and 58.68 ppm,
and not very different from those for the unsubstituted analogue,
[(iPr4PCP)IrHCl],[2l] at  -36.25 and 58.4 ppm. This suggests that
the effect of varying the para substituent on the electronic
environment of the iridium is likely to be relatively minor and the
effect of varying the group bound to the O atom at the para
position even more subtle. Thus replacing the hydroxyl group
with either a phosphinite moiety or even a metal oxide surface
such as silica is not expected to result in a significant change to
the electronics of the metal center, in accord with previous
observations.[14] The conclusion that the active site in the
supported phase is essentially identical to that of solution-phase
analogues is further supported with kinetic evidence below.
under these reaction conditions and that the carbonyl species
5]
remains active for transfer dehydrogenation.[2i,
No leaching
occurs under reaction conditions.[5] The summary of the kinetic
results is presented in Table 1.
In all cases, the addition of a hydrogen acceptor depressed
the activity in comparison to acceptorless dehydrogenation
(Table 1, Figure 3). Given that olefin-bound complexes are not
observed during reaction conditions at 300 °C,[5] this reduction in
activity was likely
a consequence of an increased CO
concentration in the system, itself arising from the conversion of
acceptor olefin used with trace water present.[2i] Although the
rate of CO formation was too small to observe or measure,
catalyst activity is has been demonstrated[5]to be very sensitive
to CO pressure and, therefore, the rate may be expected to
2
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
decrease with increasing olefin content in the gas phase. When
using TBE, the rate suppression was significantly more severe
than with ethylene or propene. The mechanism of olefin
hydration/decarbonylation to give CO is not fully understood but
this may be attributable to more facile hydration of the
intermediate branched olefin (e.g. the rate of isobutene
pathway, product α-olefin reacts directly with the 14e- [(R4PCP)Ir]
species to form an η3-allyl complex, which can eliminate the
internal olefin. The rates of both of these reactions can be
compared to primary transfer dehydrogenation to explain the
expected selectivity.[2l] Because both alkane addition and allyl
formation result from reaction with [(R4PCP)Ir], it is difficult to
hydration is ca. 100x faster than n-butene hydration at 115 °C[15]). influence the rates of these two reactions independently.
Increasing steric crowding at the reaction center impedes both
20
16
12
8
reactions to a similar extent according to DFT calculations.[2l]
The allyl pathway for isomerization therefore cannot be
completely eliminated. The maximum achievable selectivity for a
given catalyst therefore occurs in the absence of any
contribution from the hydride isomerization pathway.
100
a) catalyst 1
80
60
40
20
0
4
0
0.0
0.1
0.2
0.3
0.4
Partial Pressure (atm)
Figure 3. Transfer-dehydrogenation activity of catalyst 2 as a function of
partial pressure of acceptor with various hydrogen acceptors: ethylene ( ),
propene ( ) and TBE (
). represents the total turnover frequency for
formation of butenes under acceptorless conditions. Vtotal = 80 mL min-1, Pbutane
= 0.37 atm, Pacceptor = 7.4 x 10-3 – 0.37 atm, PHe = 1.1 – 0.73 atm, Ptotal = 1.47
atm, T = 300 °C.
̇
0.0
0.1
0.2
0.3
Partial Pressure Propene (atm)
100
80
60
40
20
0
The results of transfer dehydrogenation of butane with
propene as the hydrogen acceptor over catalysts 1 and 2 are
shown in in Figure 4.
b) catalyst 2
Catalyst 1 showed greater selectivity for 1-butene (~64-75%)
production compared with acceptorless dehydrogenation at all
tested pressures of propene. With 2, the selectivity for 1-butene
at low propene pressures was no different than under
acceptorless conditions, but increased with propene pressure,
reaching ~53% at 0.20 atm. Thus the presence of propene was
found to substantially increase selectivity for 1-butene with both
complexes by comparison to acceptorless dehydrogenation.
Steric crowding at the iridium center significantly impacts
catalytic activity.[2l, 7a] t-Butyl substituted pincer ligands are far
0.0
0.1
0.2
0.3
more crowded than the i-propyl analogues, and typically display
Partial Pressure Propene (atm)
7a]
much lower dehydrogenation rates.[2l,
Accordingly, the
Figure 4. Product selectivities of 1-butene ( ), trans-2-butene ( ) and cis-2-
butene ( ) from butane dehydrogenation as functions of propene partial
pressure over (a) 1 and (b) 2. Hollow shapes represent measured selectivities
measured activity of 1 was approximately an order of magnitude
lower than that of 2 at propene pressures of 0.16 – 0.2 atm (0.15
vs 3 TO hr-1) (Table 1).
The observed product distribution in alkane dehydrogenation
by pincer-iridium is affected by two separate isomerization
pathways (Figure 5).[2l] In the hydride isomerization pathway,
2,1-insertion of an α-olefin into an Ir-H bond of intermediate
[(R4PCP)IrH2] leads to an isoalkyl species, which can reductively
eliminate to yield internal olefin. In the allyl isomerization
for 1-butene ( ), trans-2-butene ( ) and cis-2-butene ( ) under acceptorless
dehydrogenation conditions. Vtotal = 80 mL min-1, Pbutane = 0.37 atm, PC3H6 = 7.4
̇
x 10-3 – 0.22 atm, PHe = 1.1 – 0.88 atm, Ptotal = 1.47 atm, T = 300 °C.
3
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
Table 1. Summary of 1-butene selectivity and turnover frequency as a function of the catalyst used, the nature of the olefin acceptor and the partial pressure of
the olefin.
-1
TOF (molbutenes
Catalyst
Acceptor
N/A
Partial Pressure (atm)
1-butene Selectivity (%)
molcatalyst-1 hr-1)
1
2
2
2
2
2
1
1
2
2
N/A
N/A
20
20
48
52
22
53
64
75
21
18
8
19
5.3
2.6
14
N/A
Ethylene
Ethylene
Propene
Propene
Propene
Propene
TBE
0.08
0.31
0.007
0.20
0.007
0.16
0.001
0.23
3
0.54
0.15
6.51
1
TBE
Figure 5. Proposed mechanistic pathways for observed product selectivities arising from butane dehydrogenation using supported pincer-iridium complexes.
Isomerization of 1-olefin by [(R4PCP)IrH2] competes with
hydrogenation of olefin acceptor. It was previously reported that
catalyst 1. In other words, even at the lowest pressures of
propene investigated, isomerization by the dihydride of 1 is
steric effects strongly impact the relative rates of these reactions. almost completely eliminated, but the important allyl pathway is
For example, in the case of [(tBu4PCP)IrH2] the barrier to
isomerization of 1-hexene to 2-hexene was calculated to be 40.2
kJ/mol higher than that of propene hydrogenation. By contrast,
for [(iPr4PCP)IrH2], the isomerization was calculated to have a
barrier only 3.8 kJ/mol higher than propene isomerization. Thus
higher partial pressures of propene inhibit the hydride
isomerization pathway for [(iPr4PCP)Ir], and likewise for its
supported analogue 2, resulting in increased yields of 1-butene
(Figure 4b). For [(tBu4PCP)IrH2], the hydride pathway was
calculated not to play a significant role in the isomerization;
instead the allyl isomerization pathway predominated even at
very low acceptor pressures. The relative rates of the
unaffected by propene at any partial pressure.
When using ethylene for transfer dehydrogenation of butane
over 2, the 1-butene selectivity was similarly found to increase
by comparison to acceptorless dehydrogenation, but it did not
appear to be a strong function of ethylene partial pressure
(Figure 6). The selectivity (~52%) was approximately the same
as at higher propene pressure in Figure 4b as both systems
reached the same maximum. Closer investigation of the
behavior at low partial pressures of ethylene was conducted in a
separate experiment (Figure 7). Butane dehydrogenation over 2
was carried out with ethylene acceptor fed in very low
concentrations. A cylinder containing a mixture of butane and
ethylene was prepared, and this mixture was used as the gas
feed. Due to the difference in vapor pressures, the ethylene was
depleted over time, varying the partial pressure in the feed
stream. This experiment showed that selectivity varied over a
very small range, and by 7 × 10-5 atm has approximately
isomerization pathway and dehydrogenation involve
a
competition for the 14e- pincer-Ir species (between alkane and
1-alkene product), which is unaffected by the concentration of
propene. Likely for this reason, the percent yield of 1-butene is
unaffected by the pressure of propene in the case of the bulky
4
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
reached the maximum value. The near-independence of product
distribution on ethylene pressure follows from the very facile
hydrogenation of ethylene by
calculated to have a barrier 32.6 kJ/mol lower than that of 1-
hexene isomerization by the same species.[2l] Thus at even very
low pressures (ca. 7 × 10-5 atm or above) of ethylene, the butene
product distribution is determined by a competition of butane
dehydrogenation with isomerization by the allyl pathway, which
is unaffected by acceptor.
distribution of butene isomers as is observed in the absence of
acceptor.
[(iPr4PCP)IrH2], which was
To determine whether acceptorless dehydrogenation
contributes to the activity observed in the previous experiments,
a ratio β was devised to compare the rates of butene formation
and acceptor hydrogenation. If the reaction proceeds solely by
transfer dehydrogenation, these amounts are equal and β is
equal to one. A value less than one indicates that consumption
of the acceptor is outpaced by dehydrogenation, so acceptorless
dehydrogenation must also be taking place in the system. The
results are plotted in Figure 9.
In the case of ethylene as acceptor, β was approximately
one at 0.1 atm of ethylene and above, indicating a strong
preference for the reaction to proceed solely via transfer
dehydrogenation. At very low ethylene partial pressures (Figure
S4) both β and 1-butene selectivity were significantly lower.
Propene showed similar behavior. At higher partial pressure, in
the same range where selectivity was maximized, β was
approximately one. Below 0.1 atm propene, β was less than one,
coinciding with lower yields of 1-butene due to isomerization by
dihydride. In the case of TBE as hydrogen acceptor, the
acceptor hydrogenation was never competitive with hydride
isomerization, and acceptorless dehydrogenation dominated the
activity as well, with β < 0.5 at all partial pressures.
100
80
60
40
20
0
0.0
0.1
0.2
0.3
0.4
Partial Pressure Ethylene (atm)
Thus a low rate of acceptor hydrogenation results in a low
value for β and a low observed selectivity for 1-butene
production due to an increased contribution to olefin
isomerization by the iridium dihydride. By extension, it is then
expected that acceptorless dehydrogenation conditions (β = 0),
would show low selectivity for 1-butene production as well,
consistent with our previously reported results with supported
pincer-iridium complexes.[2i, 5]
Figure 6. Product selectivities of 1-butene ( ), trans-2-butene ( ) and cis-2-
butene ( ) from butane dehydrogenation over 2 as functions of ethylene
partial pressure. Hollow shapes represent measured selectivities for 1-butene
(
), trans-2-butene
(
)
and cis-2-butene
(
) under acceptorless
dehydrogenation conditions. Vtotal = 80 mL min-1, Pbutane = 0.37 atm, PC2H4
0.09 – 0.37 atm, PHe = 1.01 – 0.73 atm, Ptotal = 1.47 atm, T = 300 °C.
=
̇
100
80
60
40
20
0
100
80
60
40
20
0
0
2
4
6
8
Partial Pressure Ethylene (10-5 atm)
0.0
0.1
0.2
0.3
Figure 7. Product selectivities of 1-butene ( ), trans-2-butene ( ) and cis-2-
butene ( ) from butane dehydrogenation over 2 as functions of ethylene
partial pressure. Hollow shapes represent measured selectivities for 1-butene
Partial Pressure TBE (atm)
Figure 8. Product selectivities of 1-butene ( ), trans-2-butene ( ) and cis-2-
butene ( ) from butane dehydrogenation over 2 as functions of TBE partial
(
), trans-2-butene
(
)
and cis-2-butene
(
) under acceptorless
dehydrogenation conditions. Vtotal = 80 mL min-1, Pbutane ≈ 0.37 atm, PC2H4 = 6.5
x 10-5 – 9.7 x 10-6 atm, PHe ≈ 1.1 atm, Ptotal = 1.47 atm, T = 300 °C.
̇
pressure from 0.08
– 16 mol%. Hollow shapes represent measured
selectivities for 1-butene ( ), trans-2-butene ( ) and cis-2-butene ( ) under
acceptorless dehydrogenation conditions. Vtotal = 80 mL min-1, Pbutane = 0.37
̇
atm, PTBE = 1.0 x 10-3 – 0.23 atm, PHe = 1.1 – 0.87 atm, Ptotal = 1.47 atm, T =
In contrast with ethylene and propene, TBE showed no
effect on product selectivity over 2 as a function of partial
pressure (Figure 8). Again, this is well rationalized by prior DFT
calculations on the unsupported species. TBE hydrogenation
was calculated to have a barrier 15.9 kJ/mol higher than 1-
hexene isomerization by [(iPr4PCP)IrH2].[2l] Thus extensive
isomerization by the supported analogue of this species (in
addition to isomerization by the supported analogue of
[(iPr4PCP)Ir]) via the allyl pathway) results in a thermodynamic
300 °C.
In the analysis above, we relied on mechanistic studies[2l] of
solution-phase dehydrogenation systems with a number of key
differences from those in the present study. In this work, the
pincer-iridium complexes were CO-ligated and covalently
anchored to a silica surface. Nevertheless, the results were
completely consistent with those calculated by DFT for alkane
dehydrogenation by unsubstituted pincer-iridium complexes,
5
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
unbound to any support and not bound to CO, in vacuo at lower
temperature. This highlights the chemical equivalence of the
present system to more familiar unsupported systems containing
the same active site.
Further, these results show that the immobilization
procedure previously used for bulkier pincer-iridium complexes
may be extended to those species containing less crowded
active sites, informing future catalyst design and allowing for the
study of more active such analogues.
for further use. Supports were calcined to 550 °C with a ramp
rate of 2 °C/min and a hold of five hours at temperature.
Synthesis of [(tBu2PO-iPr4PCP)Ir(CO)]
Synthesis of the precursor ligand HO-iPr4PCP was carried out
by analogy to literature methods.[11] Tetrahydrofuran (THF) and
methanol were distilled over Na/K alloy. All other solvents were
purchased anhydrous and used as received. Cyclooctadiene
was purified by distillation under argon to remove heavy
impurities. Anhydrous triethylamine, di-iso-propylphosphine and
chloro-di-tert-butylphosphine were used as received. Solution-
phase 1H and 31P NMR spectra were collected using either a
1.25
1.00
0.75
0.50
0.25
0.00
1
Bruker Avance III operating at 400 MHz for H NMR or 162 MHz
for 31P{1H} NMR or a Varian VNMRS operating at 500 MHz for
1H NMR or 202 MHz for 31P(1H) NMR.
Synthesis of 3,5-dihydroxymethylphenol[11]
0.0
0.1
0.2
0.3
0.4
b
Partial Pressure Acceptor (atm)
To a round-bottom flask on ice containing a slurry of 2.71 g
Figure 9. Ratio of hydrogenated acceptor to total butenes formed over 2 using
(7.14
tetrahydrofuran was slowly added 5.00 g (2.38 x 10-2 mol) of
dimethyl 5-hydroxyisophthalate as solution in 80 mL
x
10-2 mol) lithium aluminum hydride in 40 mL
ethylene ( ), propene ( ) and TBE ( ). Vtotal = 80 mL min-1, Pbutane = 0.37
̇
atm, Pacceptor = 0 – 0.37 atm, PHe = 1.1 – 0.73 atm, Ptotal = 1.47 atm, T = 300 °C.
a
tetrahydrofuran via cannula needle. The slurry became a yellow-
green color which was then refluxed 16 hours in an oil bath at
90 °C. At the end of this time, the reaction was allowed to cool to
room temperature and then put on ice, at which point it was
quenched with 1M hydrochloric acid to a pH = 3. The reaction
mixture was then filtered of solids and extracted 4 x 50 mL with
ethyl acetate. These fractions were then combined and dried
over sodium sulfate, filtered again and the solvent was removed
in vacuo to yield a viscous yellow oil (1.72 g, 1.11 x 10-2 mol,
46.6 % yield). Identity of the product was confirmed by
comparison to reported NMR spectra.[11b]
Conclusion
In summary, we present here the first report of transfer
dehydrogenation effected by a silica-tethered pincer-iridium
system based on the [(iPr4PCP)Ir] framework.[10] As previously
reported for unsupported complexes, the use of a rapidly
hydrogenated acceptor results in higher selectivity for 1-alkene
formation by preventing olefin isomerization via the pincer-
iridium dihydride intermediate. Dehydrogenation rates were
lower at higher partial pressures of olefinic hydrogen acceptor,
likely caused by an increase in CO production. Terminal olefin
selectivity up to 75% was observed with a (tBu4POCOP)
derivative, and ca. 53% was observed with an (iPr4PCP)
derivative. The experimental results with supported species are
consistent with a mechanistic model previously developed for
solution-phase analogues and provide evidence for equivalence
of the shared active site.
Synthesis of 3,5-dibromomethylphenol[11]
To a Schlenk flask equipped with a stir bar was added 10
mL sulfolane and 2.2 mL (6.27 g, 2.32 x 10-2 mol) phosphorus
tribromide and cooled in an ice bath after addition. To a second
Schlenk flask was added 10 mL sulfolane and 1.717 g (1.11 x
10-2 mol) 3,5-dihydroxymethylphenol. The contents of this flask
were cannula-transferred into the first flask and sealed under
argon for three days. At the end of this time, the reaction mixture
was unsealed and poured into 100 mL of ice water. It was then
saturated with sodium chloride and extracted 3 x 50 mL with
diethyl ether. The organic fractions were combined and washed
2 x 50 mL with brine before being dried over sodium sulfate and
filtered. Finally, the diethyl ether was removed in vacuo to yield a
crude yellow oil. The final product (1.82 g, 46 %) was isolated
Experimental Section
Catalyst Preparation
All chemical syntheses and material preparations were
performed under an air-free argon atmosphere unless otherwise
noted. Complex [(tBu2PO-tBu4POCOP)Ir(CO)] was synthesized,
and supported
1 was prepared, according to previously
established procedures.[2i]
Calcination of support materials was carried out under dry air
in a 6.4 mm outer diameter (OD) quartz tube reactor with an
expanded section of 12.5 mm OD packed with quartz wool to
hold the catalyst powder in place. After calcination, the powder
was flushed of air prior to being brought into an argon glovebox
6
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
via column chromatography using a 3:1 hexanes:diethyl ether
solvent. Identity of the product was confirmed by comparison to
reported NMR spectra.[11b]
dissociated to regenerate
a
solution of 1:1 THF:[(HO-
iPr4PCP)IrHCl] as determined by 1H NMR. Removal of solvent at
this point allowed for isolation of pure final product.
1H NMR (400 MHz, C6D6) δ 6.50 (s, 2H, Ar-H), 4.42 (bs, 1H,
O-H), 2.778 (d of vt, JPH = 4 Hz, JHH = 18.4 Hz, 2H, Ar-CH2P),
2.689 (d of vt, JPH = 4 Hz, JHH = 16.8 Hz, 2H, Ar-CH2P), 2.676 (m,
2H, PCH(CH3)2), 1.98 (m, 2H, PCH(CH3)2), 1.22 (m, 12H,
PCH(CH3)2), 0.92 (app. sext, 7.6 Hz, 12H, PCH(CH3)2), -37.13 (s,
1H, Ir-H). 31P NMR (162 MHz, C6D6) δ 58.08.
Synthesis of 3,5-di-iso-propylphosphinomethylphenol (HO-
iPr4PCP-H)[11]
Synthesis of [(tBu2PO-iPr4PCP)Ir(C2H4)]
In a Schlenk flask equipped with a stir bar was dissolved 1.0
g (3.57 mmol) 3,5-dibromomethylphenol in 40 mL of freshly
dried, degassed methanol. To this solution was added 0.8685 g
di-iso-propylphosphine (7.32 mmol, ca. 1.3 mL) and the reaction
mixture was refluxed for three days, at the end of which the
solution was a clear, amber color. The reaction was cooled to
room temperature and 3 mL triethylamine was added and
allowed to stir for ca. 1 hour, after which the solvent was
removed under vacuum to yield an off-yellow solid. The residue
was washed 1 x 10 mL acetone and then extracted 3 x 10 mL
pentane, with the organic fractions being combined and the
solvent removed to yield crude product. This was then
recrystallized from a highly concentrated diethylether solution
layered with pentane to yield the final product as white crystals
(1.02 g, 3.38 mmol, 94 %). Identity of the product was confirmed
by comparison to reported NMR spectra.[11b]
To a Schlenk flask equipped with a reflux condenser and a
stir bar was added 150.8 mg (0.259 mmol) [(HO-iPr4PCP)IrHCl]
and 58.8 mg (0.524 mol) potassium tert-butoxide. The flask was
put under ethylene flow and, after 5 minutes, 10 mL toluene was
added down the condenser. The solution immediately turned
colorless and then reddish-brown. The solution was allowed to
stir at 30 °C for 30 minutes. At this point, 1 mL (0.79 g, 0.544
mmol) chloro-di-tert-butylphosphine was added followed by ca. 2
mL of toluene and the reaction was allowed to reflux for 16
hours to yield a reddish-green solution. The solvent was then
removed and the residue was extracted with 3 x 5 mL toluene
and filtered through a 0.2 um syringe filter. The organics were
combined and dried in vacuo to yield the pure final product as a
very dark reddish brown powder in quantitative yield.
Synthesis of [(HO-iPr4PCP)IrHCl]
1H NMR (400 MHz, C6D6) δ 7.39 (s, 2H, Ar-H), 3.04 (app. t, 4
Hz, 4H, C2H4), 3.00 (t, 3.2 Hz, 4H, Ar-CH2P), 2.00 (app. tp, 2.4
Hz, 7.2 Hz, 4H, PCH(CH3)2), 1.26 (d, 11.5 Hz, 18H, PC(CH3)3),
1.051 (app. q (dd), 3.6 Hz, 12H, PH(CH3)2), 0.894 (app. q (dd),
6.4 Hz, 12H, PH(CH3)2). 31P NMR (162 MHz, C6D6) δ 148.92,
49.26.
Synthesis of [(tBu2PO-iPr4PCP)Ir(CO)]
In a Schlenk flask charged with a stir bar was dissolved
0.414 g (1.17 mmol) 3,5-di-iso-propylphosphinomethylphenol
and 0.393 g (0.585 mmol) [Ir(COD)Cl]2 in benzene. The flask
was put under a flow of hydrogen and then heated to reflux
temperature, during which the mixture became orange with
some white precipitate. Refluxing for 16 hours resulted in a deep
red solution with some black precipitate present on the flask
walls. The solution was cooled to room temperature and solvent
was removed in vacuo to yield a dark red solid. The residual
solid was washed with toluene and the supernatant was filtered
through a 0.2 m syringe filter. The material was recrystallized
from a concentrated solution of THF layered with pentane at -
30 °C to yield white crystals which were determined to be the
six-coordinate adduct [(HO-iPr4PCP)IrHCl(THF)] via X-ray
diffraction. Upon dissolution in toluene, the THF rapidly
[(tBu2PO-iPr4PCP)Ir(C2H4)] was synthesized as above, after
which the crude product from reaction was transferred to a
Kontes flask, which was then evacuated and charged with
carbon monoxide. The solution was allowed to stir for 10
minutes, during which it turned yellow-brown. All gas was
removed via one freeze-pump-thaw cycle and the flask was
7
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
Brookhart, Organometallics 2004, 23, 1766-1776; g) W. Yao, Y.
Zhang, X. Jia, Z. Huang, Angew. Chem. Int. Ed. 2014, 53, 1390-
1394; h) K. E. Allen, D. M. Heinekey, A. S. Goldman, K. I.
Goldberg, Organometallics 2013, 32, 1579-1582; i) B. Sheludko, M.
T. Cunningham, A. S. Goldman, F. E. Celik, ACS Catal. 2018, 8,
7828-7841; j) A. R. Chianese, M. J. Drance, K. H. Jensen, S. P.
McCollom, N. Yusufova, S. E. Shaner, D. Y. Shopov, J. A. Tendler,
Organometallics 2014, 33, 457-464; k) A. R. Chianese, A. Mo, N. L.
Lampland, R. L. Swartz, P. T. Bremer, Organometallics 2010, 29,
3019-3026; l) A. Kumar, T. Zhou, T. J. Emge, O. Mironov, R. J.
Saxton, K. Krogh-Jespersen, A. S. Goldman, J. Am. Chem. Soc.
2015, 137, 9894-9911; m) A. Kumar, J. D. Hackenberg, G. Zhuo, A.
M. Steffens, O. Mironov, R. J. Saxton, A. S. Goldman, J. Mol. Catal.
A. Chem. 2016; n) Z. Huang, E. Rolfe, E. C. Carson, M. Brookhart,
A. S. Goldman, S. H. El-Khalafy, A. H. R. MacArthur, Adv. Synth.
Catal. 2010, 352, 125-135; o) J. Choi, A. H. MacArthur, M.
Brookhart, A. S. Goldman, Chem. Rev. 2011, 111, 1761-1779; p)
D. Hermann, M. Gandelman, H. Rozenberg, L. J. W. Shimon, D.
Milstein, Organometallics 2002, 21, 812-818; q) E. Ben-Ari, M.
Gandelman, H. Rozenberg, L. J. W. Shimon, D. Milstein, J. Am.
Chem. Soc. 2003, 125, 4714-4715; r) E. Ben-Ari, G. Leitus, L. J. W.
Shimon, D. Milstein, J. Am. Chem. Soc. 2006, 128, 15390-15391;
s) Y. Wang, L. Qian, Z. Huang, G. Liu, Z. Huang, Chin. J. Chem.
2020, 38, 837-841; t) X. Zhang, S.-B. Wu, X. Leng, L. W. Chung, G.
Liu, Z. Huang, ACS Catal. 2020, 10, 6475-6487; u) X. Jia, L.
Zhang, C. Qin, X. Leng, Z. Huang, Chem. Comm. 2014, 50,
11056-11059. v) H. Valdés, E. Rufino-Felipe, D. Morales-Morales,
J. Organomet. Chem 2019, 15, 120864. w) H. Valdés, M. A.
García-Eleno, D. Canseco-Gonzalez, D. Morales-Morales,
ChemCatChem 2018, 10, 3136-3172.
R. Ahuja, B. Punji, M. Findlater, C. Supplee, W. Schinski, M.
Brookhart, A. S. Goldman, Nat. Chem. 2010, 3, 167.
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. 2017, 139, 8977-8989.
B. Sheludko, C. F. Castro, A. S. Goldman, F. E. Celik, ACS Catal.
2020, DOI: 10.1021/acscatal.0c02406.
Y. Wang, C. Qin, X. Jia, X. Leng, Z. Huang, Angew. Chem. Int. Ed.
2017, 56, 1614-1618.
a) S. Kundu, Y. Choliy, G. Zhuo, R. Ahuja, T. J. Emge, R.
Warmuth, M. Brookhart, K. Krogh-Jespersen, A. S. Goldman,
Organometallics 2009, 28, 5432-5444; b) F. C. Liu, A. S. Goldman,
Chem. Comm. 1999, 655-656.
J. M. Goldberg, G. W. Wong, K. E. Brastow, W. Kaminsky, K. I.
Goldberg, D. M. Heinekey, Organometallics 2015, 34, 753-762.
B. C. Vicente, Z. Huang, M. Brookhart, A. S. Goldman, S. L. Scott,
Dalton Trans. 2011, 40, 4268-4274.
again charged with carbon monoxide and allowed to stir a
further 30 minutes. The solution was then dried in vacuo to yield
a yellow-brown powder that was determined to be 70% pure by
31P NMR and used without further purification. The major
impurity present was found to be di-tert-butylphosphine oxide.
1H NMR (500 MHz, C6D6) δ 7.33 (s, 2H, Ar-H), 3.16 (s, 4H,
Ar-CH2P), 1.98 (bs, 4H, PCH(CH3)2), 1.243 (d, 11.5 Hz, 18H,
PC(CH3)3), 1.168 (q, 8 Hz, 12H, PH(CH3)2), 0.918 (q, 7 Hz, 12H,
PH(CH3)2). 31P NMR (202 MHz, C6D6) δ 149.51, 67.39.
Gas-phase Continuous-flow Catalytic Data
Reactions were carried out in a 6.4 mm outer diameter (OD)
quartz tube reactor with an expanded section of 12.5 mm OD
packed with quartz wool to hold the catalyst powder in place.
The reactor was packed with supported catalyst under argon
atmosphere and sealed with valves prior to connection to the
gas-flow manifold. The reactor was placed inside a resistively
heated ceramic furnace with external temperature control, and
the catalyst bed temperature was measured with a K-type
thermocouple placed in direct contact with the catalyst powder.
The tubing upstream of the reactor was purged with helium gas
prior to opening the reactor to the manifold.
Butane (Airgas, 99.99%) was used as received. Helium
(Praxair, 99.999%) was passed through an on-stream oxygen
and moisture trap. Helium was bubbled through a stainless-steel
saturator filled with 3,3-dimethyl-1-butene (tert-butylethylene,
TBE) to provide the desired vapor pressure. Reaction products
were analyzed using an Agilent 7890B GC equipped with a GS-
GASPRO capillary column (0.32 mm x 60 m) fitted with a flame
ionization detector.
[3]
[4]
[5]
[6]
[7]
In a typical experiment, 2 mg of molecular pincer-iridium
[8]
complex
[(tBu2PO-tBu4POCOP)Ir(CO)]
or
[(tBu2PO-
iPr4PCP)Ir(CO)] was supported on 60 mg of silica support via
incipient wetness impregnation and the resulting material was
used without further treatment. Unless stated otherwise, all
experiments were carried out at 1.47 atm total pressure due to
pressure drop through the catalyst bed. Gas flow rates reported
were measured at room temperature and pressure.
[9]
[10]
[11]
O. S. Mironov, R. J. Saxton, A. S. Goldman, A. Kumar,
US10017430B2, February 15, 2018.
a) N. Ashkenazi, A. Vigalok, S. Parthiban, Y. Ben-David, L. J. W.
Shimon, J. M. L. Martin, D. Milstein, J. Am. Chem. Soc. 2000, 122,
8797-8798; b) A. Dauth, U. Gellrich, Y. Diskin-Posner, Y. Ben-
David, D. Milstein, J. Am. Chem. Soc. 2017, 139, 2799-2807.
a) APEX2, SAINT and SADABS. Bruker AXS Inc. Madison, WI,
USA. 2013. b) G. M. Sheldrick, Acta Cryst. 2008, A64, 112-122. c)
A. Bondi, Journal of Physical Chemistry C 1964, 68(3), 441-451.
K. Zhu, P. D. Achord, X. Zhang, K. Krogh-Jespersen, A. S.
Goldman, J. Am. Chem. Soc. 2004, 126, 13044-13053.
Z. Huang, M. Brookhart, A. S. Goldman, S. Kundu, A. Ray, S. L.
Scott, B. C. Vicente, Adv. Synth. Catal. 2009, 351, 188-206.
K. Deguchi, I. Yuhara, S. Ogasawara, Nippon Kagaku Kaishi 1974,
1974, 319-324.
[12]
Selectivity was calculated via:
Si = xi / Σxi ∙ 100%
[13]
[14]
[15]
(xi = mol fraction of product)
Acknowledgements
This work was supported by the National Science Foundation
(CBET-1705746) and the Department of Energy Office of
Science (DE-SC0020139). The authors would like to thank Prof.
Akshai Kumar for valuable discussions, and Chenxu Liu and
Cesar A. Rubio for their help in data processing.
Keywords: iridium • dehydrogenation • hydrogen transfer •
heterogeneous catalysis • supported catalysts
[1]
[2]
J. Sattler, J. Ruiz-Martinez, E. Santillan-Jimenez, B. M.
Weckhuysen, Chem. Rev. 2014, 114, 10613-10653.
a) A. Kumar, T. M. Bhatti, A. S. Goldman, Chem. Rev. 2017, 117,
12357-12384; b) J. Choi, A. S. Goldman, in Iridium Catalysis (Ed.:
P. G. Andersson), Springer Berlin Heidelberg, Berlin, Heidelberg,
2011, pp. 139-167; c) D. Morales-Morales, Pincer Compounds,
Chemistry and Applications, 1st ed., Elsevier, 2018; d) C. M.
Jensen, Chem. Comm. 1999, 2443-2449; e) I. Göttker-
Schnetmann, P. White, M. Brookhart, J. Am. Chem. Soc. 2004,
126, 1804-1811; f) I. Göttker-Schnetmann, P. S. White, M.
8
This article is protected by copyright. All rights reserved.

10.1002/cctc.202001399
ChemCatChem
FULL PAPER
Entry for the Table of Contents
100
Regioselective Alkane
Transfer Dehydrogenation
80
60
40
20
0
0.0
0.1
0.2
0.3
Partial Pressure Propylene (atm)
Silica-immobilized pincer-iridium complexes may be used for alkane-olefin transfer dehydrogenation, reaching terminal olefin
selectivities of 70 % with a sufficiently unhindered hydrogen acceptor. The increased regioselectivity results from inhibition of the
hydride isomerization pathway, previously implicated for analogous, molecular solution-phase species. This result is consistent with
preservation of the iridium active site upon tethering.
9
This article is protected by copyright. All rights reserved.