Highly active and recyclable heterogeneous iridium pincer catalysts for transfer dehydrogenation of alkanes

Article abstract of DOI:10.1002/adsc.200800615

Pincer-ligated iridium complexes have proven to be highly effective catalysts for the dehydrogenation and transfer-dehydrogenation of alkanes. Immobilization onto a solid support offers significant potential advantages in the application of such catalysts particularly with respect to catalyst separation and recycling. We describe three approaches toward such immobilization: (i) covalent attachment to a Merrifield resin, (ii) covalent bonding to silica via a pendant alkoxysilane group, and (iii) adsorption on γ-alumina (γ-Al₂O₃), through basic functional groups on the para-position of the pincer ligand. The simplest of these approaches, adsorption on γ-Al₂O₃, is also found to be the most effective, yielding catalysts that are robust, recyclable, and comparable to or even more active than the corresponding species in solution. Spectroscopic evidence (NMR, IR) and studies of catalytic activity support the hypothesis that binding occurs at the para-substituent and that this has only a relatively subtle and indirect influence on catalytic behavior.

Full text of DOI:10.1002/adsc.200800615

FULL PAPERS
DOI: 10.1002/adsc.200800615
Highly Active and Recyclable Heterogeneous Iridium Pincer
Catalysts for Transfer Dehydrogenation of Alkanes
Zheng Huang,a Maurice Brookhart,a,* Alan S. Goldman,b,* Sabuj Kundu,b
Amlan Ray,b Susannah L. Scott,c,* and Brian C. Vicentec
a
Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA
Fax : (+1)-919-962-2388; e-mail: brookhar@email.unc.edu
Department of Chemistry and Chemical Biology, Rutgers University, Piscataway, NJ 08854, USA
b
Fax : (+1)-732-445-5312; e-mail: agoldman@rutchem.rutgers.edu
Department of Chemical Engineering, University of California, Santa Barbara, CA 93106, USA
c
Fax : (+1)-805-893-4731; e-mail: sscott@engineering.ucsb.edu
Received: October 8, 2008; Published online: January 2, 2009
Abstract: Pincer-ligated iridium complexes have ligand. The simplest of these approaches, adsorption
proven to be highly effective catalysts for the dehy- on g-Al2O3, is also found to be the most effective,
drogenation and transfer-dehydrogenation of al- yielding catalysts that are robust, recyclable, and
ACHTUNGTRENNUNGkanes. Immobilization onto a solid support offers sig- comparable to or even more active than the corre-
nificant potential advantages in the application of sponding species in solution. Spectroscopic evidence
such catalysts particularly with respect to catalyst (NMR, IR) and studies of catalytic activity support
separation and recycling. We describe three ap- the hypothesis that binding occurs at the para-sub-
proaches toward such immobilization: (i) covalent at- stituent and that this has only a relatively subtle and
tachment to a Merrifield resin, (ii) covalent bonding indirect influence on catalytic behavior.
to silica via a pendant alkoxysilane group, and (iii)
adsorption on g-alumina (g-Al2O3), through basic Keywords: alkanes; alumina; dehydrogenation; iridi-
functional groups on the para-position of the pincer um; pincer ligands; supported catalysts
Introduction
Goldman,[3] while the bis(phosphinite) (t-BuPOCOP)
complexes, 2a–e, have been studied by Brookhart.[4,5]
Catalytic alkane dehydrogenation is carried out indus- Other related iridium pincer systems have also shown
trially on an enormous scale for the conversion of in- good activity in alkane dehydrogenation.[6] Operating
expensive saturated hydrocarbon feedstocks to under relatively mild conditions (100–2008C), these
higher-value olefins and arenes. Heterogeneous cata- iridium pincer complexes can effect the transfer-dehy-
lysts are used in these cracking and reforming pro- drogenation of alkanes, using sacrificial olefins as hy-
cesses which typically operate at high temperatures drogen acceptors [as in Eq. (1)], or dehydrogenation
(400–6008C), resulting in low product selectivities and in the absence of sacrificial acceptors under condi-
poor energy efficiency.[1]
tions where H2 is allowed to escape from the system.
By comparison with conventional heterogeneous
While numerous homogeneous catalytic alkane de-
hydrogenation systems have been reported,[2] iridium dehydrogenation systems, the iridium pincer catalysts
pincer complexes have been shown to be the most (like other homogeneous dehydrogenation catalysts)
À
productive. Two examples of such systems are shown show excellent chemoselectivity: “cracking” (C C
in Figure 1. The iridium bis(phosphine) (t-BuPCP) com- bond cleavage side reactions) has never been report-
plex, 1a,[3] has been investigated by Jensen, Kaska and ed. Even more notably, these complexes have been
188
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
activity. The silica-supported iridium POCOP catalyst
shows good transfer-dehydrogenation activity at early
stages but poor lifetimes. Notably, the g-Al2O3-sup-
ported iridium complexes are recyclable and highly
active, and in some cases even much more productive
than the analogous homogeneous system for transfer
dehydrogenation.
Results and Discussion
Alumina-Supported Iridium Pincer Catalyst Systems
Figure 1. Structures of Ir-PCP and Ir-POCOP complexes.
The iridium PCP dihydride complexes (1a–d) and
POCOP ethylene complexes (3–5) shown in Figure 1
and Figure 2 were used in the alumina adsorption
reported to catalyze dehydrogenation with high regio- studies. Complexes 1a, 1b, 1c and 3 have been previ-
selectivity for the terminal positions of n-alkanes or ously prepared. Complexes 1d, 4 and 5, previously un-
n-alkyl groups. Moreover, in addition to simple al- known, contain polar groups in the para position and
ACHTUNGTRENNUNGkanes, iridium pincer complexes have been reported were thought to be good candidates for adsorption.
to catalyze dehydrogenation of numerous other sub-
strates including amines,[7] alcohols,[8] and polyole-
fins.[9] Mechanistic details of the dehydrogenations by Syntheses of Complexes 1d, 4 and 5
both iridium PCP and POCOP pincer catalysts have
been disclosed.[10]
The synthesis of the bis(phosphine) PCP complex 1d
While the iridium pincer systems are quite robust, is outlined in Scheme 1.[12] Dimethylation of the com-
the utility of these homogeneous catalysts is limited mercially available dimethyl 5-aminoisophthalate (6)
due to problems of catalyst recyclability and product was achieved by adaptation of a method developed
separation. Heterogeneous iridium pincer systems by Borch and Hassid.[13] The diester 7 was then re-
therefore hold promise for a broad range of applica- duced to 5-dimethylamino-1,3-benzenedimethanol (8)
tions. In particular, our groups desired such catalysts by lithium aluminum hydride in THF.[14] Diol 8 was
to combine with heterogeneous olefin metathesis cat- then treated with PBr3 in acetonitrile and the result-
alysts to generate a fully heterogeneous catalyst ing dibromide 9 was recrystallized in 77% yield from
system for alkane metathesis.[11]
acetonitrile/water[15] before conversion to Me2N-PCP-
In this paper we report the results of three strat- H (10) by an adaptation of the method of Moulton
egies for preparing supported iridium pincer com- and Shaw.[16] Metalation of the free ligand could only
plexes: (i) covalent attachment of an iridium complex be achieved under an H2 atmosphere; the resulting
containing a phenoxide functionality to a Merrifield iridium hydrido chloride was reduced to the iridium
resin through an SN reaction with the chlorobenzyl dihydride (1d) and tetrahydride in analogy with the
2
moieties; (ii) covalent bonding to silica of iridium synthesis of 1a and the corresponding tetrahydride.
pincer complexes containing a pendant alkoxysilane Although we were not successful in obtaining crystals
group; (iii) adsorption of Ir pincer complexes (partic- of either (Me2N-PCP)IrH2 or (Me2N-PCP)IrH4, the
ularly those containing basic functional groups) on g- reaction of this mixture with CO, in analogy with the
Al2O3 through a Lewis acid/Lewis base interaction. reaction of the unsubstituted PCP hydrides, gave
The Merrifield resin-supported iridium POCOP cata- clean conversion to (Me2N-PCP)Ir(CO) which was
lyst was found to have low transfer-dehydrogenation crystallographically characterized.[17] For catalytic
Figure 2. Structures of iridium POCOP complexes used for absorption on g-alumina.
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
189

Zheng Huang et al.
FULL PAPERS
Scheme 1.
Scheme 2.
runs, 1d was generally used as a mixture of (Me2N- (12) which then reacts with ethylene to form {p-KO-
PCP)IrH2 and (Me2N-PCP)IrH4 (the iridium PCP and C6H2-2,6-[OP(t-Bu2)]2}Ir(C2H4) (4) in 70% yield.
A
ACHTUNGTRENNUNG
POCOP tetrahydrides readily convert to the dihy-
drides under catalytic conditions).
The preparation of the tris(phosphinite) Ir pincer
complex (5) is outlined in Scheme 3. The tris(phos-
The synthesis of complex 4 is outlined in Scheme 2. phinite) POCOP ligand (13) was synthesized in 89%
Deprotection of the methoxy group of the previously yield by reaction of 1,3,5-benzenetriol with di-tert-bu-
reported[4] {p-OMe-C6H2-2,6-[OP
ACHTUNGTRENNUNG
with
9-I-BBN
A
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
190
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
Scheme 3.
Support of Bisphosphinite (POCOP) Complexes 3
and 4 on g-Alumina; Transfer-Dehydrogenation
Activity
g-Al2O3 has been widely used to support metals or
metal oxides for heterogeneous catalysis, with most
applications being in the petroleum industry.[18] For
example, a g-Al2O3-supported cobalt catalyst has
been used in the Fisher–Tropsch process for decades.
However, g-Al2O3 is rarely used to support organo-
metallic catalysts. Since alumina contains Lewis acidic
sites we considered that iridium pincer complexes
containing basic and/or polar functional groups could
be strongly adsorbed on this support. We visually ex-
amined the effects of adding alumina to cyclooctane
Figure 3. Transfer dehydrogenation of COA and TBE with
alumina-supported Ir catalysts.
solutions of the pincer catalysts, which are orange in TBE couple. Results are summarized in Table 1. A
color. When 50 mg of g-Al2O3 are added to 1.0 mL of system containing 2.5 mmol of the iridium catalyst 4
a COA solution containing 2.5 mmol of phenoxide supported on 280 mg of g-Al2O3, 7.5 mmol of COA
complex 4, the solution is completely decolorized and (3000 equivalents relative to Ir), and 7.5 mmol of
the alumina acquires the orange color of the pincer TBE, was heated at 2008C under argon in a sealed
complex. In contrast, a similarly treated solution of 3 glass vessel for 40 h. This process converted 1990
retains an orange color. Addition of a further 50 mg
alumina results in a nearly colorless solution, suggest-
ing that the parent complex 3 can be adsorbed but
that the interaction is weaker than with 4.
Table 1. Transfer dehydrogenation of COA/TBE using 4 (so-
lution phase) or 4/g-Al2O3 at 2008C.[a]
The transfer of hydrogen from COA to tert-butyl-
ethylene (TBE) as acceptor to yield cyclooctene
(COE) and 2,2-dimethylbutane (TBA) [Eq. (1)] is a
commonly used “benchmark” reaction;[19] this reac-
tion is thermodynamically favorable by 6 kcal
molÀ1.[20] Using the iridium PCP complex 1a, TONs
up to 1000 can be achieved at 2008C but require por-
tionwise addition of TBE which is a strong inhibitor
(Figure 3). The iridium POCOP systems, 2a–2f, are
much less subject to inhibition by TBE; TONs up to
2200 have been observed at 2008C with substrate:ca-
talyst ratios of 3300:1.[4]
Time 4/solution
4/g-Al2O3
(heterogeneous)
equiv. TBA DM2B DM1B TBE
(homogeneous)
equiv. TBA
0.5 h 450
3 h 1450
15 h 1800
800
280
530
690
810
0
0
40
70
1920
1180
670
1290
1600
1990
40 h 1970
130
1st recycle
40 h No activity
1420
1200
1150
1200
200
580
230
20
2nd recycle
40 h No activity
[a]
1.23 mM iridium catalyst in solution or equivalent
amount in the case of supported systems; [COA]o =
[TBE]o =3.7M; 280 mg g-Al2O3.
The g-Al2O3-supported complex 4 was examined
for transfer-dehydrogenation activity using the COA/
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
191

Zheng Huang et al.
FULL PAPERS
equivalents (relative to Ir) of TBE to TBA, and COA
to COE and 1,3-cyclooctadiene (COE/COD ~5/1).[21]
The activity is similar to the homogeneous system
under identical conditions (1970 TONs, 40 h, Table 1).
In the heterogeneous system, part of the TBE was
isomerized to form 2,3-dimethyl-2-butene (DM2B,
810 equivalents, 40 h) and 2,3-dimethyl-1-butene
(DM1B, 70 equivalents, 40 h). These isomerized ole-
fins were shown to be poor acceptors. For the homo-
geneous system, no TBE isomerization products were
observed.
The homogeneous catalyst 4 gradually decomposed
and unidentified black solids precipitated during the
transfer dehydrogenation reaction. The catalytic activ-
ity was completely lost after evaporation of volatiles
and addition of fresh COA and TBE solutions
(Table 1). However, when supported on g-Al2O3, 4
can be recycled with only modest loss of activity. Sur-
prisingly, the extent of isomerization of TBE in-
creased after each recycle. After the first run (show-
ing 1990 TONs), the catalyst was isolated and re-
charged with COA and TBE (3000 equivalents; re-
Figure 4. a): Solid-state 31P MAS NMR of fresh 4/g-Al2O3
(bottom) and b): solid-state 31P MAS NMR of 4/g-Al2O3
after use for catalytic transfer-dehydrogenation (1090 TOs)
(top). 12 kHz MAS, 208C, referenced to 85% H3PO4. Star-
red signals represent spinning sidebands.
spectively). After 40 h at 2008C, 1420 equivalents of trast to the neutral complex 3, the basic phenoxide
TBE were converted to TBA, with the formation of complex 4 must, to a certain extent, “neutralize” the
1150 equivalents of DM2B and 200 equivalents of acidity of g-Al2O3, thus decreasing the TBE isomeri-
DM1B. A second recycle exhibited 1200 turnovers zation rate. Since most of the TBE was isomerized to
(TONs) with 1200 equivalents of DM2B and 580 form DM2B and DM1B with 3/g-Al2O3, the transfer
equivalents of DM1B formed as isomerization prod- dehydrogenation efficiency is very low.
ucts.
Solid-state 31P MAS NMR analysis of the g-Al2O3-
The isomerization of TBE decreases the transfer supported 4 (5.0 mmol of 4 on 150 mg of g-Al2O3)
dehydrogenation efficiency since DM2B and DM1B shows a single species with a 31P NMR shift at
are inefficient hydrogen acceptors. (No 2,3-dimethyl- 177 ppm which is close to that of complex 4 in solu-
butane, the hydrogenation product of DM2B or tion (170 ppm). After 1090 turnovers in the hydrogen
DM1B, was observed in the transfer dehydrogenation transfer reaction, a new minor (ca. 10%) species with
reaction catalyzed by 4/g-Al2O3). In control experi-
a
31P NMR shift of 66 ppm appears (Figure 4.). This
ments with 280 mg of g-Al2O3 (no Ir), 7.5 mmol of shift corresponds to one for the species generated
COA, and 7.5 mmol of TBE, 98% of TBE was iso- when the supported catalyst is exposed to oxygen and
merized at 200 8C to form DM2B (78%) and DM1B suggests that catalyst decay occurs through oxidation,
(22%) after 6 h. No transfer dehydrogenation prod- probably at phosphorus.
ucts TBA or COE were observed. This result indi-
cates that the isomerization of TBE is catalyzed by
the g-Al2O3, presumably by acidic sites on the alumi- Preparation of Basic Alumina and Transfer-
na surface. DM1B and DM2B are well known prod- Dehydrogenation Activity of Basic Alumina-
ucts of acid-catalyzed rearrangement of TBE.[22]
Supported 4 and 5
The supported parent complex 3 (3/g-Al2O3; 2.5
mmol of 3 on 280 mg of g-Al2O3) was also investigated As discussed above, the acid-catalyzed isomerization
for transfer dehydrogenation under similar conditions. of TBE by g-Al2O3 results in decreased efficiency of
Compared to 4/g-Al2O3, the transfer reaction cata- COA/TBE transfer dehydrogenation. A variety of
lyzed by 3/g-Al2O3 shows a higher TBE isomerization commercially available aluminas, including acidic,
rate and poorer dehydrogenation activity. After 15 h neutral, and basic were screened as the supports for
at 2008C, 2300 and 480 equivalents of DM2B and the iridium catalyst. However, all of these aluminas
DM1B were produced with the 3/g-Al2O3, and only were found to isomerize TBE rapidly. To decrease the
190 equivalents of TBA were produced as the transfer acidity of alumina, an Na2O-modified g-Al2O3 (2.7
dehydrogenation product. The reaction with 4/g- wt% of Na2O) solid support was synthesized by
Al2O3 under the same reaction conditions produced adding an aqueous solution of NaOH or Na2CO3 to g-
690 equivalents of DM2B, 40 equivalents of DM1B, Al2O3. The solid was calcined at 5508C for 18 h under
and 1600 equivalents of TBA (See Table 1). In con- a flow of O2. A control experiment with 310 mg of
192
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
Table 2. Transfer dehydrogenation of COA/TBE with solution-phase or supported 4 and 5.
Time
Homogeneous
Heterogeneous
4[a] (2008C)
TBA (equiv.)
5
[a] (2008C)
4/Al2O3/Na2O[a] (200 8C)
5/Al2O3/Na2O[a] (200 8C)
TBA
4/Al2O3/Na2O[b] (240 8C)
TBA
TBA
DM2B
DM2B
TBA
DM2B
0.5 h 270
1260
2110
2540
2660
1220
2220
3490
4140
30
80
160
180
980
0
0
30
40
2170
3620
6240
7010
0
40
130
180
3 h
1380
1830
1860
2520
3730
4310
15 h
40 h
recycle, 2008C recycle, 2008C recycle, 2008C
No activity No activity 660 60
recycle, 2008C
1520
recycle, 2408C
940 50
15 h
0
[a]
0.49 mM iridium catalyst in solution or equivalent amount in the case of supported systems (1.34 mmol supported iridium
catalyst); 10.05 mmol COA and TBE, [COA]o =[TBE]o =3.7M; 310 mg g-Al2O3/Na2O.
1.34 mmol supported iridium catalyst; 14.74 mmol COA and TBE, [COA]o =[TBE]o =3.7M; 310 mg g-Al2O3/Na2O.
[b]
the Na2O-modified alumina (Al2O3/Na2O), 7.5 mmol 150.8 ppm. The significant upfield shift indicates a re-
of COA, and 7.5 mmol of TBE, resulted in isomeriza- action between the uncoordinated phosphinite and
tion of only 1% of the TBE to form DM2B after 48 h the alumina. There is a third, minor 31P NMR signal
at 2008C.
observed at 151 ppm. This signal position is close to
The Al2O3/Na2O solid (310 mg) was used to support the uncoordinated phosphine in solution (150.8 ppm)
catalyst 4 (1.3 mmol) and was screened for transfer de- and is ascribed to some of 5 in which the free phos-
hydrogenation activity (Table 2). Transfer dehydro- phinite group is not interacting with alumina.
genation of COA and TBE (7500 equivalents relative
To investigate the interaction of uncoordinated
to Ir) catalyzed by 4/Al2O3/Na2O results in 4140 turn- phosphinite with alumina, the free 1,3,5-tri(di-tert-bu-
overs and only 2.4% isomerization of TBE after 40 h tylphosphinite)benzene ligand and the model com-
at 2008C. Recycle showed that only 660 TONs were pound phenyl di-tert-butylphosphinite were supported
obtained after 15 h (Table 2). It should be noted that on Al2O3/Na2O. Analysis of the Al2O3/Na2O-support-
the transfer-dehydrogenation activity with the hetero-
geneous system 4/Al2O3/Na2O is much higher than
the activity with the homogeneous catalyst 4 which
exhibited only 1860 TONs after 40 h under the same
reaction conditions (Table 2). The homogeneous cata-
lyst 4 decays faster than the heterogeneous system,
and no catalytic activity was observed after evapora-
tion of volatiles and addition of fresh COA and TBE
solutions (Table 2). Heating at a higher temperature,
2408C, with the heterogeneous system and 11000
equivalents of COA and TBE gave 7010 turnovers
after 40 h, and only 1.6% of TBE was isomerized to
form DM2B (Table 2).
One major disadvantage of catalyst 4 is its lengthy
and involved synthesis. Complex 5, readily prepared
from 1,3,5-benzenetriol, was found to adsorb strongly
on alumina. Analysis of the Al2O3/Na2O-supported 5
(7.6 mmol of 5 on 150 mg of Al2O3/Na2O) by solid-
state 31P MAS NMR shows two major phosphorus en-
vironments (see Figure 5). The peak at 177 ppm is as-
signed to the intact, coordinated phosphinite group
with a spinning sideband at 275 ppm (the spinning
sideband expected at 78 ppm is obscured by the
second peak in the spectrum). This constitutes a slight
upfield shift from the solution state chemical shift of
181.1 ppm and is a result of complex-support interac-
tions. The peak at 78 ppm is attributed to the phos-
phinite group not coordinated to Ir, which is shifted
Figure 5. a): Solid-state 31P MAS NMR of phenyl di-tert-bu-
tylphosphinite/Al2O3/Na2O (bottom), b): triphosphinite-ben-
zene/Al2O3/Na2O (middle), and c): 5/Al2O3/Na2O (top).
12 kHz MAS, 208C, referenced to 85% H3PO4. Starred sig-
far upfield from its solution state chemical shift of nals represent spinning sidebands.
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
193

Zheng Huang et al.
FULL PAPERS
ed phenyl di-tert-butylphosphinite by solid-state 31P Catalytic Activity of Bisphosphine (PCP) Complexes
MAS NMR shows a large upfield shift from the solu- Supported on g-Alumina
tion-state chemical shift of 153 ppm to 69 ppm (see
Figure 5). Supporting the 1,3,5-tri(di-tert-butylphos- Unlike the bisphosphinite analogues, (PCP)IrH2 (1a)
phinite)benzene results in two major phosphorus en- and derivatives react with TBE, if present in high
vironments. The sharp peak at 68.5 ppm is consistent concentration, to give an unidentified, catalytically in-
with a phosphinite reacting with the alumina surface active, decomposition product. Thus we conducted
as in the phenyl di-tert-butylphosphinite sample. The our study of alumina-supported PCP complexes under
broad peak at 154 ppm is assigned to the phosphinite conditions somewhat different than those applied in
groups not directly interacting with the surface. The studies of the alumina-supported POCOP complexes.
results indicate the uncoordinated phosphinite in
Complexes (MeO-PCP)IrH2 (1b) and [MeOC(O)-
complex 5 can indeed react with the alumina surface PCP]IrH2 (1c) have been previously synthesized.[23]
and this results in a 31P upfield shift by about 70– The methoxy-substituted complex 1b was previously
80 ppm in the 31P MAS NMR spectrum. The nature reported to be a more robust alkane dehydrogenation
of this interaction is currently under study.
catalyst than the parent complex 1a,[24] while giving
The Al2O3/Na2O-supported 5 (1.3 mmol 5 on slightly higher rates of acceptorless dehydrogenation
310 mg Al2O3/Na2O) mirrors the activity of Al2O3/ (of cyclodecane) but slightly lower rates of n-octane/
Na2O-supported 4 but recycle of 5/Al2O3/Na2O is NBE transfer-dehydrogenation. As seen in Table 3
much more efficient. For example, transfer-dehydro- (solution phase), turnover frequencies (TOFs) for
genation of COA and TBE (7500 equivalents each COA/TBE transfer-dehydrogenation by 1b are also
relative to Ir) catalyzed by 5/Al2O3/Na2O results in somewhat lower than are found for 1a. The ester-sub-
4310 turnovers and less than 1% isomerization of stituted complex 1c is found to afford slightly greater
TBE after 40 h at 2008C. The first recycle exhibited initial rates for catalytic COA/TBE transfer dehydro-
1520 turnovers after 15 h (Table 2). For the homoge- genation than either 1b or 1a (Table 3). However 1c
neous system, the conversion at early time (30 min, apparently undergoes significant decomposition under
1260 TONs) is greater than the heterogeneous system the catalytic conditions as indicated by a decrease in
1
(30 min, 980 TONs), however; after 40 h, the homoge- catalytic activity. Accordingly, 31P and H NMR spec-
neous system exhibited only 2660 TONs relative to troscopy independently reveal that in the presence of
4310 TONs for the supported system. Attempted re- TBE 1c reacts to give six-coordinate iridium hydride
cycle of the solution-phase catalyst 5 by evaporation complexes. This decomposition is attributable to inter-
of volatiles and addition of fresh COA and TBE re-
sulted in no additional turnovers, indicating the com-
plete decomposition of 5 under transfer-dehydrogena-
tion condition. Thus the alumina-supported 5 has a
much longer lifetime than catalyst 5 in solution.
Table 3. COA/TBE transfer-dehydrogenation catalyzed by
solution-phase and g-alumina-supported (PCP)Ir-based com-
plexes at 1258C.[a]
Catalyst
Time
[min]
Homogeneous Heterogeneous
(solution
(g-alumina)
Leaching Experiments
phase)
AHCTUNGTRENNUNG
AHCTUNGTRENN[GUN COE] (mM)
COA suspensions of 4/Al2O3/Na2O and 5/Al2O3/Na2O
were filtered at 2008C. The solid material was extract-
ed twice more with COA at 2008C and the colorless
filtrates were combined and analyzed for iridium con-
tent by ICP-MS. The analysis of solutions indicated
that only 0.02% of Ir had leached into the solution
from the Al2O3/Na2O-supported 4 and only 0.007%
from the Al2O3/Na2O-supported 5. The solutions
showed no activity for COA/TBE transfer-dehydro-
genation. These results, in combination with the activ-
ity of the supported systems being comparable to or
even greater than that of the catalysts in solution
phase, clearly indicate that the active catalytic species
remains bound to the alumina surface during the
transfer-dehydrogenation reaction.
(PCP)IrH2 (1a)
15
60
240
61
3
3
4
28
164
368
36
A
(1b)
60
240
15
115
352
73
60
84
49
(MeO2C-
PCP)IrH2 (1c)
60
240
155
258
20
119
354
42
AHCTUNGTREN(GNUN Me2N-PCP)IrH2 15
(1d)
60
240
68
200
111
283
[a]
5 mM iridium catalyst in solution or equivalent amount
in the case of supported systems; [TBE]o =0.4M, 1 mL
COA and 100 mg g-Al2O3.
194
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
Table 4. COA/TBE transfer dehydrogenation: recycling catalysts 1c and 1d.[a]
Catalyst (5 mM equivalent)
Time
[h]
1st
2nd
3rd
4th
5th
6th
7th
8th
cycle cycle
cycle cycle cycle cycle cycle cycle
(MeO2C-PCP)IrH2 (1c) heterogeneous (g-alumi-
na-supported)
1
4
8
1
4
8
1
4
8
117
331
440
75
281
465
115
314
464
101
188
259
67
222
339
91
27
41
49
N
56
47
30
70
114
46
74
15
41
65
16
43
65
9
4
154
246
66
135
216
116
161
61
119
197
20
30
10
23
31
11
14
6
12
15
ACHTUNGTRENNUNG(Me2N-PCP)IrH2 (1d) heterogeneous (g-alumina-
supported)
173
315
117
[a]
5 mM iridium catalyst in solution or equivalent amount in the case of supported systems. [TBE]o =0.6M, 1 mL COA and
100 mg g-Al2O3.
À
molecular addition of a C H bond ortho to the ester lution was removed from the solid, which was then
functionality, in accord with the previously reported washed two times with COA (2ꢁ2 mL) and a fresh
reaction of (PCP)Ir with acetophenone.[25]
TBE/COA solution was then added. The subsequent
Upon addition of g-alumina to a COA (100 mg g- catalytic runs showed significantly decreased reactivi-
Al2O3 +1 mL COA) solution of unsubstituted PCP ty (Table 4).
iridium complex 1a (5 mM) the red solution turned
Very promising results were obtained with the new
clear and the solid acquired the characteristic red catalyst (Me2N-PCP)IrH2 (1d). Of the four X-PCP iri-
color of the complex. Upon heating to 1258C the red dium catalysts used in this study, 1d gave the lowest
solid rapidly turned orange, suggesting that decompo- TOFs for COA/TBE transfer-dehydrogenation in so-
sition had occurred. Accordingly, very little COA/ lution (Table 3). Thus it is found that initial COA/
TBE transfer-hydrogenation occurred in the presence TBE solution-phase transfer-dehydrogenation rates
of alumina at 1258C (less than 5 mM COE formed; increase with decreasing electron-donating ability of
Table 3).
the group X: Me2N<MeO<H<CO2Me. However,
Attempts to support (PCP)Ir-based catalysts on alu- when 1d was adsorbed on g-alumina, initial rates of
mina were more promising with MeO-PCP complex COA/TBE transfer-dehydrogenation were greater
1b than with 1a, but were still not satisfactory. As in than obtained by solution-phase 1d. This is consistent
the case of 1a, upon addition of alumina the solution with the correlation with electron-withdrawing ability
lost its red color which was acquired by the alumina, of X; binding of the Me2N group to a Lewis acidic
but, in contrast to alumina-supported 1a, no color surface site would indeed be expected, based on this
change was observed even upon heating. After 15 min correlation, to increase catalytic activity.
at 1258C, 1b (5 mM) afforded product yields in the
In addition to the increased TOFs observed upon
presence of alumina only slightly less than in the ab- binding 1d to alumina, the total TONs effected by the
sence of alumina (28 mM vs. 36 mM). But after 1d/g-alumina system after 4 h were significantly great-
240 min the total yield was substantially less than was er than achieved with the homogeneous system (283
obtained in the absence of alumina (84 vs. 352 mM).
vs. 200 mM; Table 3). The system proved to be robust
In contrast to results with 1a and 1b, the catalyst and recyclable (Table 4). The solution was removed
lifetime and total turnovers effected by ester-substi- after 8 h of catalysis at 1258C and the remaining solid
tuted complex 1c were increased in the presence of was washed two times with COA; upon addition of
alumina. As with all the bound iridium PCP catalysts, fresh TBE/COA solution to the solid, each subse-
adsorption of 1c visually appeared to be complete. Al- quent run showed only a relatively small decrease in
though initial rates of COA/TBE transfer-dehydro- catalytic activity. This process involves extensive ex-
genation were slightly lowered by the presence of alu- posure of the catalyst (which is sensitive to O2, H2O
mina (49 mM vs. 73 mM after 15 min), the yield of and even N2) to an imperfect glove-box atmosphere;
COE was appreciably greater after 240 min than was thus the observed decrease in TOF for each cycle rep-
obtained with the solution-phase catalyst (354 mM vs. resents only an upper limit of the degree of decompo-
258 mM). This effect can be rationalized by assuming sition that occurred during the actual catalytic run.
that adsorption of the catalyst to alumina inhibits the Recycling of the solution-phase catalyst necessarily
intermolecular catalyst de-activation reaction noted involves a different protocol, namely, removal of sol-
above. However, attempts to recycle the 1c/g-alumina vent under vacuum before adding fresh solution.
catalyst system met with only partial success. The so- While this presumably involves less exposure to im-
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
195

Zheng Huang et al.
FULL PAPERS
Table 5. n-Octane/TBE transfer-dehydrogenation by (Me2N-
PCP)IrH2 (1d) at 1258C.[a]
purities, the loss of activity with each cycle is approxi-
mately the same as that observed in the case of the
alumina-supported system (Table 4).
The 1d/g-alumina system was also effective for the
transfer dehydrogenation of n-octane, and as was
found with COA, more active than solution-phase 1d
(Table 5).
Time
[min]
Homogeneous
(solution phase)
Heterogeneous
(g-alumina-supported)
1-Octene Total octene 1-Octene Total octene
(mM)
(mM)
(mM)
(mM)
15
30
60
120
240
4
9
15
16
16
7
2
3
3
4
4
15
30
59
99
130
The yield of 1-octene from n-octane with this
system is much lower than with solution-phase 1d. We
initially assumed that this lower apparent selectivity
was due to the isomerization of 1-octene by g-alumi-
na. Control experiments with g-alumina, with no iridi-
um present, do indeed show that 1-octene is isomer-
ized under these conditions (Table 6). For example,
after ca. 60 min, with an initial 1-octene concentration
of 29 mM, isomerization is ca. 50% complete with cis-
and trans-2-octene being the only major products.
However, it does not seem that this level of isomeri-
zation activity (half-life of ca. 60 min), by itself, could
account for the much lower yields of 1-octene ob-
tained from 1d/g-alumina vs. solution phase 1d (e.g.,
3 mM 1-octene out of 30 mM total octene product vs.
9 mM 1-octene out of 22 mM total octene, after only
30 min of catalysis). Further work is ongoing to eluci-
date the reason for the relatively low yield of terminal
alkene, but possible explanations include formation of
a minor decomposition product on alumina that acts
as a highly active catalyst for isomerization, or per-
haps simply increased isomerization activity from 1d
upon binding to alumina (possibly due to decreased
electron-density at Ir). It should be noted, however,
that even the small yields of 1-octene observed at
early reaction times indicate that at least partial selec-
tivity for dehydrogenation at the terminal position is
22
43
73
98
[a]
5 mM iridium catalyst in solution or equivalent amount
in the case of supported systems. [TBE]o =0.4M, 1 mL n-
octane and 100 mg g-Al2O3.
Table 6. Control experiment: isomerization of 1-octene by g-
alumina (no iridium present, 100 mg g-Al2O3 +1 mL n-oc-
AHCTUNGTRENNUNG
tane).[a]
Initial [1-
octene]
Time
[min]
1-Octene
(mM)
trans-2-
Octene
(mM)
cis-2-
Octene
(mM)
29 mM
5
27
1
1
10
15
30
60
5
10
15
30
60
26
24
20
14
413
404
393
366
307
2
3
6
8
1
2
3
5
427 mM
8
6
14
22
46
88
12
15
24
40
retained upon binding to alumina (even the low 1- the methoxy group does not bind as strongly as the
octene concentrations observed at early reaction ester or dimethyl amino groups of 1c or 1d, respec-
times are much greater than equilibrium values). Fur- tively. Electron-withdrawing ability of the para-sub-
thermore, the predominant internal octenes observed stituent is correlated with catalytic TOFs in solution;
are cis- and trans-2-octene, with much lower concen- consistent with the presumed binding of the Me2N
trations of 3- and 4-octene (e.g., >80% 2-octene after group to a Lewis acidic surface site, the Me2N-PCP
60 min); this is indicative of selectivity for the termi- catalyst 1d affords increased TOFs when bound to g-
nal position followed by rapid a-b isomerization and alumina. The 1d/g-alumina system is also found to be
slower further internal isomerization.
quite stable under catalytic conditions, and even toler-
To summarize the results in this section, (PCP)IrH2 ates multiple cycles of solvent removal, washing, and
(1a) shows very rapid loss of catalytic activity in the reuse. In the case of n-alkane, the product distribution
presence of g-alumina. All three para-substituted from the 1d/g-alumina system is predominantly 2-
complexes investigated in this study underwent de- octene, but this is likely due to increased rates of a-b
composition in the presence of g-alumina far more isomerization rather than selectivity for dehydrogena-
slowly than 1a, if at all. Thus para-substituent binding tion the 2-position of the alkane chain.
appears to inhibit a decomposition reaction of the iri-
dium center with alumina. In the case of complex 1c
the presence of alumina also appears to inhibit the in- Quantifying the Strength of Binding of the
termolecular decomposition reaction that is observed Me2N-PCP Unit to Alumina
in the solution phase; this effect is similar to that ob-
served for complexes 4 and 5 as discussed in the pre-
ACHTUGNERTN(NUNG Me2N-PCP)IrH2 (30 mg) was dissolved in 10 mL
ceding section. In the case of catalyst 1b, deactivation COA in the presence of 1 g g-alumina. The mixture
by alumina still occurs, albeit slowly, suggesting that was stirred for 15 min and then filtered; the filtrate
196
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
was then evaporated under vacuum. The residue was that 1b-CO binds to g-alumina with the p-methoxy
analyzed by X-ray fluorescence and the iridium con- group acting as a Lewis base toward a Lewis acid sur-
tent was found to be below the detection limit of this face site; the bound methoxy group is then electron-
method, which is estimated as <1ꢁ10À6 g, or 0.003% withdrawing, as might be expected. Likewise, the nCO
of the initial amount of iridium.
value of complex 1d-CO in solution is 7 cmÀ1 less
The iridium carbonyl complexes are much more than that of 1a-CO; but upon binding to alumina, nCO
robust than the catalytically active hydrides. For this is blue-shifted by 10.5 cmÀ1 and is then 3 cmÀ1 higher
reason, (Me2N-PCP)Ir(CO) (1d-CO) was used to help than that of 1a-CO/g-alumina. Thus, as indicated by
quantify, by UV-vis spectroscopy, the strength of bind- the relative binding-induced blue shifts, the Me2N
ing of the (Me2N-PCP)Ir unit to alumina. An n- group of 1d-CO apparently donates significantly more
hexane solution of 1d-CO (2.5 mM; 3 mg in 2.0 mL) electron density to the alumina than does the MeO
has an absorbance of 1.58 at l=493 nm (1.0 cm path- group of 1b-CO; this is in accord with results of the
length). When 2.0 mL of the same solution was stirred catalytic runs which indicated that 1d binds more
in the presence of 50 mg g-alumina for 15 min and strongly to g-alumina than does 1b.
then filtered, the absorbance at l=493 nm was found
Ester-substituted complex 1c-CO shows no signifi-
to be <1ꢁ10À4. The concentration of 1d-CO in solu- cant change in nCO upon binding to alumina. This may
tion under these conditions is therefore <0.006% of be attributable to a weak interaction with alumina; al-
that present prior to the addition of alumina.
ternatively, the binding may proceed via a trans-
esterification type reaction that does not result in a
large change in electron density at the iridium center.
Likewise, the binding of (KO-POCOP)Ir(CO) to alu-
mina yields no significant change in nCO; this suggests
that the alumina surface exerts an electronic effect
indi- similar to that of the K+ counterion in solution. Final-
Infrared Spectroscopic Characterization of the
Iridium Complexes Supported on g-Alumina
[26]
À
C O stretching frequencies act as a valuable
cator of small changes in electron density at the metal ly, [(t-Bu)2PO-POCOP]Ir(CO) shows a significant
center of transition metal carbonyl complexes. In blue shift upon binding to alumina (1934 cmÀ1 to
order to probe the nature of the binding of the X- 1945 cmÀ1). While this large shift is consistent with a
PCP complexes to g-alumina, we prepared the corre- strong Lewis acid-base interaction, it may also be
sponding (X-PCP)Ir(CO) complexes. All complexes noted that the resulting nCO value is equal to that of
appeared, visually, to be fully adsorbed by g-alumina. alumina-bound (KO-POCOP)Ir(CO). This may sug-
After removal of solvent, nujol mulls of the material gest that the binding involves cleavage of the (t-
À
were prepared for IR analysis.
Bu)2P O bond to give a species very similar to that
Adsorption of (PCP)Ir(CO) (1a-CO) on alumina obtained from the binding of (KO-POCOP)Ir; this is
À
results in a negligible change in C O stretching fre- currently under further investigation.
quency as compared with 1a-CO dissolved in nujol
hydrocarbon (1925.9 cmÀ1 vs. 1925.3 cmÀ1; Table 7).
The value of nCO of (MeO-PCP)Ir(CO) in Nujol is ca. Synthesis of a Merrifield Resin-Supported Iridium
3 cmÀ1 red-shifted versus solution-phase 1a-CO, con- Pincer Complex and Transfer-Dehydrogenation
sistent with the electron-donating properties of the p- Activity
methoxy group. When bound to alumina however,
nCO of 1b-CO is 5 cmÀ1 greater than solution-phase Complex 4, which contains a phenoxide functionality,
1b-CO and ca. 2 cmÀ1 greater than either solution- was attached to a Merrifield resin through an SN re-
2
phase 1a-CO or 1a-CO/g-alumina. These data indicate action with the chlorobenzyl moieties (Scheme 4).
The Merrifield resin (2% cross-linked, 200–400 mesh,
2.25 mmol Cl/g) was swollen in THF-d8 for 1 h before
Table 7. C-O stretching frequencies of complexes (X- the coupling reaction. To prevent the possible dissoci-
PCP)Ir(CO) in solution and adsorbed on g-alumina.
ation of the ethylene ligand, reactions were conducted
under ca. 1.5 atmosphere of ethylene. Protonation of
the phenoxide group of 4 occurred during the reac-
tion although the THF solvent and Merrifield resin
were dried before use. With the addition of KH (2–3
equivalents relative to Ir), the conversion to phenol
was avoided. When a red THF-d8 solution of 4
(18 mg, 27 mmol) was treated with a stoichiometric
amount of Merrifield resin (12 mg, 27 mmol)
Compound
nCO [cmÀ1]
(nujol)
nCO [cmÀ1]
(g-alumina,
nujol)
AHCTUNGTRENNUNG
A
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
1925.3
1922.6
1918.4
1931.0
1945
1925.9
1927.5
1928.9
1930.5
1945
ACHTUNGTRENNUNG
(Ir:chloro
ACHTUNGTRENNbUNG enzyl moiety=1:1) for 15 days at 658C,
A
1934
1945
the THF solution was still red indicating incomplete
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
197

Zheng Huang et al.
FULL PAPERS
Scheme 4.
1
reaction. Analysis of the THF-d8 solution by H NMR Covalent Attachment of Iridium Pincer Complexes
spectroscopy indicated that only about 45% of 4 was Containing Pendant Alkoxy Silane Groups to Silica
immobilized on the resin. When 4 was treated with a and Transfer-Dehydrogenation Activity
7 molar excess of Merrifield resin (Ir:chlorobenzyl
moiety=1:7) in THF, after 7 days at 658C the original Iridium pincer complexes containing either a pendant
red solution of complex 4 faded to colorless and the -SiACHTNUTRGNEUNG(OMe)3 group, 15, or -Si(Me)2OMe group, 16, were
original white resin was now red. Analysis of the prepared by treating 4 with either 3-iodopropyltrime-
THF-d8 solution by NMR indicated that no 4 re- thoxysilane or 3-iodopropyldimethylmethoxysilane,
mained in solution. The solids were filtered and respectively, in THF under an ethylene atmosphere
washed 3 times with THF. To remove the excess KH (Scheme 5).
in the solid mixture, excess t-BuOH was added to the
Attachment of 15 to silica was achieved by heating
solid mixture in THF-d8 under an ethylene atmos- 300 mg silica (Grace XPO 2402) with 16 mg (20
phere (t-BuOH has proven to be compatible with the mmol) 15 in toluene-d8 at 1208C under an atmosphere
parent complex 3). Reaction of t-BuOH with KH pro- of ethylene to prevent ethylene loss from the Ir
duced KO-t-Bu and H2 which then hydrogenated eth- center.[27] Periodic analysis of the solution by
ylene to form ethane (ethane was observed by 1H NMR showed that as the concentration of 15 de-
1H NMR spectroscopy). The red resin was filtered, creased, methanol and ethane formed and increased
washed 3 times with THF and H2O (to remove KCl in concentration. After 2 days, the original red solu-
and KO-t-Bu) under argon and then dried under high tion became colorless and the silica acquired a pink
vacuum overnight.
color. No detectable 15 remained in solution and ca.
The polymer-supported iridium catalyst shows poor two equivalents of methanol were produced (relative
activity and short lifetimes for transfer dehydrogena- to Ir) which indicated that on average two methoxy
tion of COA by TBE. A system containing 5.4 mmol groups of 15 reacted with the silanol groups on silica
of the polymer-supported iridium catalyst, 240 mmol surface to produce a siloxane linkage and methanol.
of COA (44 equivalents relative to Ir), 240 mmol of Excess trimethylsilyldimethylamine was added to cap
TBE, and 0.3 mL of mesitylene-d12 was heated at the remaining silanol groups.[28] This supported cata-
1758C under argon in a J. Young NMR tube. After 2 lyst, which contained 63 mmol Ir/g, was isolated,
days, 85% conversion of TBE to TBA (37 TONs) was washed with pentane, toluene, and THF three times,
1
observed by H NMR. The supported catalyst was re- respectively, and dried under high vacuum.
covered and recharged with same amount of COA,
Transfer dehydrogenation of COA/TBE with this
TBE, and mesitylene. Heating the sample for 2 days silica-supported catalyst (50 mg, 3.15 mmol Ir) showed
at 1758C resulted in 20% conversion (9 TONs). After low activity. Reaction with 750 equivalents of COA
the second cycle, the catalyst lost activity completely and TBE at 2008C gave only 135 turnovers after 24 h.
and the support changed in color from red to black. Heating for longer times resulted in no additional
The decomposition product is unidentified. One possi- turnovers, indicating the decomposition of the cata-
ble decomposition pathway could result from reaction lyst.
of the remaining chlorobenzyl moieties of the Merri-
The formation of ethane during the immobilization
field resin with the Ir center. In light of the low activi- process indicates that hydrogenation of ethylene oc-
ty of this system, further experiments were not pur- curred. A control experiment was conducted by heat-
sued.
ing homogeneous complex 3 with 3 equivalents of
methanol under an ethylene atmosphere in toluene at
1208C. Iridium carbonyl complex 17 and ethane were
formed as observed by NMR spectroscopy. Produc-
tion of CO from methanol will produce the hydrogen
198
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
Scheme 5.
necessary for the hydrogenation of ethylene Ir center (we assume under these conditions the eth-
(Scheme 6). The formation of ethane during the im- ylene complex will be converted to the COE com-
mobilization process implies conversion of a signifi- plex). After reaction of 16 with silica for 2 days at
cant fraction of the Ir ethylene complex to an Ir car- 1508C in COE, the original red solution became col-
bonyl complex which we have shown is inactive for orless and the silica particles were now red. Excess di-
dehydrogenation.
methyl(trimethylsilyl)amine was then added to cap
In order to avoid extensive production of an iridi- the remaining silanol groups on silica. The solid mate-
um carbonyl complex, reaction conditions and starting rial was collected and washed with pentane, toluene,
material were modified (Scheme 5). The monome- and THF three times, respectively, and was dried
thoxy complex 16 was employed to reduce the pro- under high vacuum. The catalyst made by this modi-
duction of methanol to one equivalent and cyclooc- fied procedure showed improved activity for transfer
tene (COE) was used as the reaction solvent to pro- dehydrogenation. After 15 h, transfer dehydrogena-
vide a high concentration of alkene to “protect” the tion of COA/TBE (3000 equivalents of COA and
Scheme 6.
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
199

Zheng Huang et al.
FULL PAPERS
mesh, 2.25 mmol Cl/g) was purchased from Aldrich and
dried under high vacuum at 408C for 20 h. Silica (Grace
XPO 2402) was supplied by DuPont and used as received.
All other reagents were purchased from Sigma–Aldrich or
Strem and used as received.
TBE) at 200 8C yielded 790 turnovers. However, heat-
ing for longer times did not increase the TON (790
after 39 h), which indicates that this silica-supported
catalyst is not stable and loses activity at a relatively
fast rate.
NMR spectra were recorded on DRX-400, VANCE-400,
1
Varian-400 and Varian-500 spectrometers. H and 13C NMR
spectra were referenced to residual protio solvent peaks.
31P NMR chemical shifts were referenced to an external
H3PO4 standard. The samples for solid-state 31P MAS NMR
spectra were packed into 4 mm zirconia solid-state NMR
rotors under an argon atmosphere and sealed with tight-fit-
ting rotor caps. Solid-state 31P MAS NMR spectra were re-
corded on a Bruker Avance DSX300 spectrometer operating
at 121.49 MHz with magic angle spinning (MAS) of 12 kHz.
The rotors were spun with N2 in order to keep the samples
air-free. All spectra were recorded at room temperature and
are referenced to an aqueous 85% H3PO4 solution.
UV-visible spectra were recorded on a Varian Cary-50
spectrophotometer. Infrared spectra were recorded on a
Thermo Nicolet 360-FT-IR instrument. Elemental analyses
were carried out by Robertson Microlit Laboratories, NJ.
GC analyses (FID detection) were performed according
to the following methods:
Method A: Agilent 6850 Series GC System fitted with an
Agilent HP-1 column (100% dimethylpolysiloxane, 30 mꢁ
0.32 mm i.d., 0.25 mm film thickness). Typical temperature
program: 5 min isothermal at 338C, 20 8CminÀ1 heat up,
10 min isothermal at 3008C. Flow rate: 1 mLminÀ1 (He).
Split ratio: 400. Inlet temperature: 2508C. Detector temper-
ature: 2508C.
Method B: Thermo Electron Corporation Focus GC in-
strument fitted with an Agilent HP-1 column (100% methyl
silicone gum: 100 mꢁ0.25 mm IDꢁ0.5 mm film thickness).
Typical temperature program: Starting temperature 1008C,
58CminÀ1 up to 2308C with hold time 10 min, then
58CminÀ1 up to ending temperature 2508C. Flow rate:
1 mLminÀ1 (He). Split ratio: 25. Inlet temperature: 2308C.
Detector temperature: 2508C.
Conclusions
Three approaches are reported here for constructing
iridium pincer-based transfer dehydrogenation cata-
lysts. POCOP iridium catalysts were covalently at-
tached to a Merrifield resin and to silica. These sys-
tems showed low to moderate transfer dehydrogena-
tion activity. A third method was developed in which
iridium pincer complexes bearing basic functional
groups in the para-position bind to g-alumina through
a Lewis acid/Lewis base interaction. These alumina-
supported complexes have been characterized by
solid state 31P NMR and IR spectroscopies which
yield information concerning the nature of surface
binding. The catalysts are thermally robust and recy-
clable and display high activities with turnover num-
bers up to 7000 for transfer of hydrogen from cyclo-
octane to tert-butylethylene. ICP-MS experiments
with para-phosphinite- and para-oxide-supported sys-
tems show negligible leaching from alumina under re-
action conditions. A number of applications can be
envisioned for these catalysts including use in alkane
metathesis.[11]
Experimental Section
General Considerations
All manipulations were carried out using standard Schlenk,
high-vacuum and glove-box techniques. Tetrahydrofuran
(THF) was distilled under a nitrogen atmosphere from
sodium benzophenone ketyl prior to use. Pentane and tolu-
ene were passed through columns of activated alumina.
Water was degassed by purging with argon. Benzene, THF-
d8, and toluene-d8 were dried with 4 ꢂ molecular sieves and
degassed by freeze-pump-thaw cycles. Acetone was dried
with 3 ꢂ molecular sieves for 7 h and degassed by freeze-
pump-thaw methods. Cyclooctane (COA), 3,3’-dimethyl-1-
butene (TBE), p-xylene and mesitylene were purchased
from Aldrich, dried with LiAlH4 or Na/K, and vacuum
Method C: As in Method B but with typical temperature
program:Às1tarting temperature 608C with hold time 70 min,
108CminÀ1 up to ending temperature 2508C.
108Cmin
up to 2008C with hold time 10 min, then
ICP-MS Analysis was performed on a Varian 820-MS.
Sample preparation for ICP-MS: The COA (2 mLꢁ3) sus-
pension of alumina-supported iridium complex (2.50 mmol Ir
on 310 mg of alumina) was filtered through a frit separated
double-cell glassware. The COA solution was combined and
dried under high vacuum and the residue was digested by a
69.5% HNO3 solution at 1008C for 2 h. Finally, the sample
was diluted with HNO3 at 2%. The quantification was car-
ried out using external calibration curves from dilution of a
certified ICP-MS Ir standard (Varian). The calibration
curves were made as follows: 191Ir and 193Ir: 1000, 500, 250,
50, and 5 ppb. The certified standard was diluted with 2%
HNO3 solution. The quadratic correlation coefficient ob-
tained in the regression line was 0.99997. 115In was used as
the internal standard for correcting possible instrumental
drifts. Both 191Ir and 193Ir were analyzed. Five replicates
were carried out for each isotope analyzed. The RSD mean
obtained was 3.0%.
transferred into sealed flasks. Complexes 1a, 1b, 1c, 3,[11] and
[29]
[Ir
G
were synthesized as previously reported.
KH was purchased from Aldrich as 30 wt% in oil and
washed with hexanes five times prior to use. I(CH2)3Si-
(OMe)3 and Cl(CH2)3Si(Me)2A(OMe) were purchased from
Gelest and used as received. I(CH2)3Si(Me)2A(OMe) was pre-
pared through the Finkelstein reaction by stirring the mix-
ture of NaI and Cl(CH2)3Si(Me)2A(OMe) in dried acetone
AHCTUNGTRENNUNG
A
R
CHTUNGTRENNUNG
A
CHTUNGTRENNUNG
A
CHTUNGTRENNUNG
overnight at 608C. g-Al2O3, acidic g-Al2O3, neutral g-Al2O3,
and basic g-Al2O3 were purchased from Strem and calcined
as noted below. Merrifield resin (2% cross-linked, 200–400
200
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
for 2 h. Volatiles were removed under high vacuum, and the
red solid was dried under high vacuum for 4 h to give the
pure product; yield: 105 mg (0.156 mmol, 72%). 1H NMR
(400 MHz, 238C, THF-d8): d=1.26 (virtual triplet, apparent
J=6.2 Hz, 36H, 4ꢁt-Bu), 2.64 (s, 4H, C2H4), 5.67 (s, 2H, 3-
and 5-H); 31P{1H} NMR (162 MHz, 238C, THF-d8): d=170.3
(s); 13C{1H} NMR (100.6 MHz, 238C, THF-d8): d=26.5
Synthesis of Iridium Complex {p-OH-C6H2-2,6-[OPACTHUNTGRNEUNG(t-
Bu)2]2}IrHI (11)
[CH3, m, 2ꢁP
ACHTUNGTRENNUNG
11.1 Hz, 2 ꢁ PACHTUNGTRENNNUG
J=5.5 Hz, C3 and C5), 134.5 (Cq, m, C1), 168.8 (Cq, virtual
triplet, apparent J=7.0 Hz, C2 and C6), 174.4 (Cq, s, C4);
elemental analysis calculated for C24H42O3P2IrK (672.19): C
42.90, H 6.30; found: C 42.66, H 6.31.
{p-OMe-C6H2-2,6-[OP
(t-Bu)2]2}IrHCl
(2b)
(200 mg,
Synthesis of Phenyl Di-tert-butylphosphinite
0.305 mmol) was dissolved in benzene (6 mL) in a flame-
dried Schlenk flask and put under a flow of argon. 9-I-BBN
(1M in hexanes, 0.63 mL) was added, and the solution was
stirred for 2 h at room temperature. The solvent was re-
moved at room temperature under high vacuum and then
the by-product 9-Cl-BBN and the extra 9-I-BBN were re-
moved at 808C under high vacuum. A mixture of benzene
(3 mL) and degassed water (7 mL) was added into the resi-
due, and the solution was stirred at room temperature over-
night. Volatiles were removed under high vacuum. The resi-
due was washed with pentane (3ꢁ6 mL) and the resulting
red solid was dried under high vacuum overnight to give the
pure product; yield: 192 mg (0.262 mmol, 86%). 1H NMR
A solution of 10.63 mmol of phenol (1.0 g) in 25 mL of THF
was slowly added via syringe to a suspension of 10.8 mmol
of NaH (259 mg) in 25 mL of THF under a flow of argon
(caution: hydrogen evolution). The mixture was heated to
reflux for 2 h, di-tert-butylchlorophosphine (10.63 mmol,
2.0 g) was then added via syringe, and the mixture was re-
fluxed for an additional 8 h. After evaporation of the sol-
vent under high vacuum, the residue was extracted with 3ꢁ
40 mL of pentane, and the extract was cannula transferred
and filtered through a pad of Celite. After removal of pen-
tane under high vacuum, the flask was heated to 558C for
3 h under high vacuum to remove residual amounts of di-
tert-butylchlorophosphine. Pure product as clear oil was col-
2
(400 MHz, 238C, CDCl3): d=À42.11 (t, JP, H =13.0 Hz, 1H,
IrH), 1.36 (virtual triplet, apparent J=7.2 Hz, 36H, 4ꢁt-
Bu), 4.49 (s, 1H, OH), 6.19 (s, 2H, 3- and 5-H); 31P{1H}
NMR (162 MHz, 238C, CDCl3): d=181.3; 13C{1H} NMR
(100.6 MHz, 238C, CDCl3): d=28.0 [CH3, virtual triplet, ap-
1
lected; yield: 2.334 g (9.8 mmol, 92%). H NMR (400 MHz,
238C, CDCl3): d=1.13 (d, 3JP, H =11.6 Hz, 18H, 2ꢁt-Bu),
3
3
6.90 (t, JH,H =7.2 Hz, 1H, 4-H), 7.11 (d, JH,H =7.6 Hz, 2H,
2- and 6-H), 7.20 (m, 2H, 3- and 5-H); 31P{1H} NMR
(162 MHz, 238C, CDCl3): d=153.12 (s). 13C{1H} NMR
parent J=2.8 Hz, P
ent J=2.6 Hz, P(t-Bu)2], 40.1 [Cq, virtual triplet, apparent
J=12.6 Hz, P(t-Bu)2], 43.4 [Cq, virtual triplet, apparent J=
11.5 Hz, 2ꢁP(t-Bu)2], 93.3 (CH, virtual triplet, apparent J=
ACHTUNGTRENNUNG(t-Bu)2], 28.1 [CH3, virtual triplet, appar-
ACHTUNGTRENNUNG
2
(100.6 MHz, 238C, CDCl3): d=27.4 [CH3, d, JP, C =15.1 Hz,
ACHTUNGTRENNUNG
PACHTUNTRGENN(UG t-Bu)2], 35.6 [Cq, d, JP, C =25.1 Hz, PCAHTUNGTRNEN(UGN t-Bu)2], 118.4 (CH, d,
ACHTUNGTRENNUNG
3JP, C =10.1 Hz, C2 and C6), 121.3 (CH, s, C4), 129.3 (CH, s,
C3 and C5), 159.9 (Cq, d, 2JP, C =9.0 Hz, C1).
6.0 Hz, C3 and C5), 111.7 (Cq, m br, C1), 155.3 (Cq, s, C4),
166.9 (Cq, virtual triplet, apparent J=6.0 Hz, C2 and C6);
elemental analysis calculated for C22H40IO3P2Ir (734.11): C
36.02, H 5.50; found: C 36.87, H 5.31.
Synthesis of Pincer Ligand 1,3,5-Tri(di-tert-
butylphosphinite)benzene (13)
Synthesis of Iridium Complex {p-OK-C6H2-2,6-[OP
Bu)2]2}Ir(C2H4) (4)
ACHTUNGERTN(NUNG t-
A solution of 10 mmol of 1,3,5-trihydroxybenzene (1.261 g)
in 25 mL of THF was slowly added via syringe to a suspen-
sion of 31 mmol of NaH (782 mg) in 25 mL of THF under a
flow of argon (caution: hydrogen evolution). The mixture
was heated to reflux for 2 h, di-tert-butylchlorophosphine
(31.0 mmol, 5.834 g) was then added via syringe, and the
mixture was refluxed for additional 2 h. After evaporation
of the solvent under high vacuum, the residue was extracted
with 3ꢁ40 mL of pentane, and the extract was cannula
transferred and filtered through a pad of Celite. After re-
moval of pentane under high vacuum, the flask was heated
to 558C for 3 h under high vacuum to remove residual
amounts of di-tert-butylchlorophosphine. Pure product as
clear a waxy solid was collected; yield: 4.972 g (8.9 mmol,
ACHTUNGTRENNUNG
Complex 11 (160 mg, 0.218 mmol) and KH (22 mg,
0.548 mmol) were weighed into a flame-dried Kontes flask
under an argon atmosphere. THF (5 mL) was added to the
flask via syringe and the resulting suspension was stirred for
2 h at room temperature. The solution was filtered into a
Schlenk flask through a 0.2 mm pore size syringe filter (Nal-
gene 199–2020). Ethylene was bubbled through the solution
1
3
89%). H NMR (400 MHz, 238C, CDCl3): d=1.13 (d, JP, H
=
11.6 Hz, 54H, 6ꢁt-Bu), 6.58 (m, 3H, 2-, 4-, and 6-H);
31P{1H} NMR (162 MHz, 238C, CDCl3): d=151.22 (s);
13C{1H} NMR (100.6 MHz, 238C, CDCl3): d=27.4 [CH3, d,
2JP, C =16.1 Hz, 3ꢁP
ACHTNUGTNRENU(GN t-Bu)2], 35.6 [Cq, d, JP, C =26.1 Hz, 3ꢁP-
3
AHCTUNGTRENNUNG
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
201

Zheng Huang et al.
FULL PAPERS
vacuum, and the red brown solid was dried under high
vacuum overnight to give the pure product; yield: 371 mg
(0.477 mmol, 75%). 1H NMR (400 MHz, 238C, C6D6): d=
Synthesis of Iridium Vomplex {p-OPACHTNUTRGNEUNG(t-Bu)2-C6H2-2,6-
[OPACHTUNGTRENNUNG(t-Bu)2)]2}IrHCl (14)
3
1.10 (d, JP, H =8.0 Hz, 18H, 2ꢁt-Bu), 1.23 (virtual triplet, ap-
parent J=6.8 Hz, 36H, 4ꢁt-Bu), 3.11 (s, 4H, C2H4), 7.06 (s,
2H, 3- and 5-H); 31P{1H} NMR (162 MHz, 238C, C6D6): d=
150.8 [s, uncoordinated P
ACHTUNGTNER(NUNG t-Bu)2], 181.1 [s, coordinated 2ꢁP-
AHCTUNGTRENNUNG
2
[CH3, d, JP, >C =15.8 Hz, uncoordinated P
virtual triplet, apparent J=3.0 Hz, coordinated 2ꢁP
35.6 [Cq, d, JP, C =26.8 Hz, uncoordinated P(t-Bu)2], 35.8 (s,
C2H4), 41.5 [Cq, virtual triplet, apparent J=11.0 Hz, coordi-
nated 2ꢁP(t-Bu)2], 94.7 (CH, dvt, apparent J=5.5 and
11.1 Hz, C3 and C5), 139.7 (Cq, m, C1), 161.5 (Cq, d, JP, C
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
=
A Schlenk flask was charged with 1.1 mmol of 13 and
0.5 mmol of [(COD)IrCl]2 and put under a flow of argon.
Toluene (5 mL) was added via syringe, and the solution was
stirred in an oil bath for 15 h at 1508C. The reaction mixture
was cooled to room temperature. Volatiles were removed
under high vacuum and the residue was extracted with
10 mL of pentane under ultrasound (10 min). After filtration
(no inert gas required), the solid was washed with 3ꢁ10 mL
of pentane and dried under high vacuum overnight to afford
10.1 Hz, C4) 168.8 (Cq, virtual triplet, apparent J=8.0, C2
and C6); elemental analysis calculated for C32H60O3P3Ir
(778.34): C 49.40m H 7.77; found: C 49.60, H 7.60.
Calcination of Alumina
g-Al2O3, acidic g-Al2O3, neutral g-Al2O3, and basic g-Al2O3
were calcined at 5508C for 2 h under a flow of O2 and
cooled to 1358C under O2, then cooled to room temperature
under high vacuum. The solids were brought into the
drybox under high vacuum and stored under argon.
of pure red purple product; yield: 645 mg (0.82 mmol, 82%).
2
1H NMR (400 MHz, 238C, CDCl3): d=À41.91 (t, JP, H
=
3
13.2 Hz, 1H, IrH), 1.16 (d, JP,H =12 Hz, 18H, 2ꢁt-Bu), 1.33
(m, 36H, 4ꢁt-Bu), 6.43 (s, 2H, 3- and 5-H); 31P{1H} NMR
Synthesis of Na2O-Modified Alumina
(162 MHz, 238C, CDCl3): d=154.2 [s, uncoordinated P
Bu)2], 175.8 [s, coordinated 2ꢁP
(t-Bu)2]; 13C{1H} NMR
(100.6 MHz, 238C, CDCl3): d=27.5 [CH3, d, JP, C =16.1 Hz,
uncoordinated P(t-Bu)2], 27.6 [CH3, br, coordinated P(t-
Bu)2], 27.8 [CH3, br, coordinated P(t-Bu)2], 35.6 [Cq, d,
P, C =25.1 Hz, uncoordinated P(t-Bu)2], 39.5 [Cq, virtual trip-
let, apparent J=12.6 Hz, coordinated P(t-Bu)2], 43.1 [Cq,
virtual triplet, apparent J=11.1 Hz, coordinated P(t-Bu)2],
ACHTUNGERTN(NUNG t-
In
a vial, 235 mg (2.22 mmol) of Na2CO3 or 178 mg
AHCTUNGTRENNUNG
2
(4.45 mmol) of NaOH were dissolved in 10 mL of distilled
water. The solution was added to 5 g of g-Al2O3. The sus-
pension was stirred at room temperature until all of the
water was absorbed by the alumina. The solid was dried in a
1208C oven overnight, then calcined at 5508C for 17 h
under a flow of O2 and cooled to 1358C under O2, and then
cooled to room temperature under high vacuum. The solid
was brought into the drybox under high vacuum and stored
under argon.
A
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
J
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
96.2 (CH, dvt, apparent J=4.0 and 9.0 Hz, C3 and C5),
109.3 (Cq, m br, C1), 159.4 (Cq, m br, C4), 166.9 (Cq, virtual
triplet, apparent J=5.5, C2 and C6); elemental analysis cal-
culated for C30H57O3P3ClIr (786.28): C 45.82, H 7.31; found:
C 45.58, H 7.04.
Synthesis of Alumina-Supported Iridium Pincer
Complexes
Synthesis of Iridium Complex {p-OP
[OP(t-Bu)2]2}Ir(C2H4) (5)
ACHTUNGTRNE(UNNG t-Bu)2-C6H2-2,6-
The alumina-supported pincer Ir ethylene complexes can be
prepared either in situ or by using pentane as solvent.
Method A (in situ): Ir complex (1.34–2.50 mmol) was dis-
solved in alkane (1–3 mL) (cyclooctane or linear alkanes).
The solution was added to 280–310 mg (2.74–3.04 mmol) of
g-Al2O3. The suspension was stirred at room temperature.
After 10–20 min, the original red solution turned colorless
and the alumina acquired a rust-red color. The suspension
continued to stir for 2–4 h.
N
ACHTUNGTRENNUNG
Method B (pentane): Ir complex (1.34–2.50 mmol) was dis-
solved in pentane (1.5 mL). The solution was added to 280–
310 mg (2.74–3.04 mmol) of g-Al2O3. The suspension was
stirred at room temperature for 2–4 h. The original red solu-
tion turned colorless and the alumina acquired a rust-red
color. The solvent was removed by syringe in the glove-box.
The solid was washed with pentane three times. Pentane
was evaporated in the glove-box (no vacuum applied) and
the rust-red solid was collected.
Complex 14 (500 mg, 0.636 mmol) and NaO-t-Bu (64 mg,
0.666 mmol) were weighed into a Schlenk flask and put
under a flow of argon. Toluene (20 mL) was added to the
flask via syringe and ethylene was bubbled through the solu-
tion for 2 h. After evaporation of the solvent under high
vacuum, the residue was extracted with 3ꢁ20 mL of pen-
tane, and the extract was cannula transferred and filtered
through a pad of Celite. Pentane was removed under high
202
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
Synthesis of Alumina-Supported 13 and Phenyl Di-
tert-butylphosphinite
Isomerization of TBE or 1-Octene by Alumina
(Control Experiments)
Compound 13 (26 mmol), or phenyl di-tert-butylphosphinite,
was dissolved in pentane (1.5 mL). The solution was added
to 200 mg (1.96 mmol) of g-Al2O3. The suspension was
stirred at room temperature for 2–4 h. The solvent was re-
moved by syringe in the glove-box. The solid was washed
with pentane three times. Volatile was evaporated in the
glove-box and the white solid was collected.
A flask was charged with 280–310 mg of g-Al2O3 or Al2O3/
Na2O, 0.842 g of COA (7.50 mmol), and 0.664 g of TBE
(7.50 mmol). In the case of 1-octene, a flask was charged
with 100 mg of g-Al2O3, COA (1 mL), and 1-octene
(4.55 mL, 29 mmol or 67 mL, 427 mmol). The flask was
sealed tightly with a Teflon plug under an argon atmos-
phere, and the solution was stirred in an oil bath at 1258C
(1-octene) or 2008C (TBE). Periodically, the flask was re-
moved from the bath and cooled in an ice bath. An aliquot
was removed from the flask, and analyzed by GC.
Hydrogen Transfer from Cyclooctane (COA) to 3,3-
Dimethyl-1-butene (TBE) Catalyzed by g-Al2O3-
Supported Iridium Pincer Complexes
Synthesis of a Merrifield Resin-Supported Iridium
Pincer Complex
Complexes 1a–d: the iridium complex (5 mmol) was dis-
solved in COA (1 mL) in a Kontes flask. g-Al2O3 (100 mg,
0.98 mmol) was added to the solution and the suspension
was stirred at room temperature for 20 min. TBE (95%,
70 mL, 0.54 mmol) was added to the suspension. The flask
was sealed tightly with a Teflon plug under an argon atmos-
phere, and the suspension was stirred in an oil bath at
1258C. Periodically, the flask was removed from the bath
and cooled in an ice bath. An aliquot was removed from the
flask, and analyzed by GC (method B). Turnover numbers
were calculated for each aliquot using mesitylene as a GC
standard.
Complexes 3–5: the iridium complex (1.34–2.50 mmol) was
dissolved in COA (0.842–1.655 g, 7.50–14.74 mmol) in a
Kontes flask. g-Al2O3 (280–310 mg, 2.74–3.04 mmol) was
added to the solution and the suspension was stirred at
room temperature for 2–4 h. TBE (95%, 0.664–1.306 g,
7.50–14.74 mmol) was added to the suspension. The proce-
dure was otherwise as described above for complexes 1a–d
but reactions were conducted at 200 or 2408C and the solu-
tion was analyzed by GC using method A. Turnover num-
bers were calculated for each aliquot. Results are summar-
ized in the text.
The Merrifield resin (12 mg, 27 mmol chlorobenzyl moiety;
or 84 mg, 189 mmol chlorobenzyl moiety when Merrifield
resin was used in excess) was swollen in THF-d8 (0.5 mL)
for 1 to 2 h in a thick-walled J. Young tube in the glove-box.
Complex 4 (18 mg, 26.8 mmol) and KH (3 mg, 75 mmol)
were added to the tube. The THF-d8 solution was degassed
by freeze-pump-thaw cycles. The tube was refilled with eth-
ylene gas at À788C. The suspension was heated at 658C.
1
The solution was monitored by H and 31P{1H} NMR spec-
troscopy periodically. For the sample with Ir:chlorobenzyl
moiety ratios of 1:1, after 15 days, about 45% of 4 was im-
mobilized on the resin (determined by NMR). For the
sample with a 7 molar excess of Merrifield resin, after 7
days, no 4 was observed in the solution. The solid was col-
lected by filtration and washed with 3 times of THF (0.5 mL
each) in the glove-box. The solid was reloaded into a thick-
walled J. Young tube and THF-d8 (0.5 mL) was added to the
tube. Degassed t-BuOH (ca. 0.2 mL) was added into the J.
Young tube by vacuum transfer method (caution, hydrogen
evolution). The red resin was filtered, washed with 3 times
of THF and H2O (0.5 mL each) under an argon atmosphere,
and then dried under high vacuum overnight.
The heterogeneous catalysts can be recycled. After each
cycle, the solution was syringed out and the solid was
washed 2–3 times with COA. Fresh COA and TBE were
then added.
Hydrogen Transfer from COA to TBE Catalyzed by
Merrifield Resin-Supported Iridium Pincer Complex
A J. Young tube was charged with 8 mg of the Merrifield
resin supported iridium catalyst (5.4 mmol of Ir), 32 mL of
COA (240 mmol), 33 mL of TBE (240 mmol), and 0.3 mL of
mesitylene-d12. The tube was sealed tightly under an argon
atmosphere, and then heated in an oil bath at 1758C. The
sample was analyzed by NMR spectroscopy periodically.
After each cycle, the catalyst was filtered, washed with pen-
tane and dried under high vacuum. Fresh COA, TBE and
mesitylene-d12 were then added. Results are summarized in
the text.
Hydrogen Transfer from COA or n-Octane to TBE
Catalyzed by Solution-Phase Iridium Pincer
Complexes
A flask was charged with iridium pincer complex 1a, 1b, 1c
or 1d (5 mmol), COA or n-octane (1 mL), and TBE (95%,
70 mL, 0.54 mmol). Respective values in the case of complex
4 or 5: iridium complex, 1.34–2.50 mmol; COA, 0.842–
1.655 g, 7.50–14.74 mmol; TBE (95%), 0.664–1.306 g, 7.50–
14.74 mmol. The flask was sealed tightly with a Teflon plug
under an argon atmosphere, and the solution was stirred in
an oil bath at 1258C (complexes 1a–d) or 2008C (complexes
4 and 5). Periodically, the flask was removed from the bath
and cooled in an ice bath. An aliquot was removed from the
flask, and analyzed by GC (method A, complexes 1a–d;
method B, complexes 1a–d). Turnover numbers were calcu-
lated for each aliquot. Recycle of the homogeneous catalysts
was obtained by evaporation of under high vacuum and ad-
dition of fresh COA and TBE. Results are summarized in
the text.
Synthesis of Dimethyl 5-Dimethylaminoisophthalate
(7)
The compound was prepared via the literature procedure
for reductive alkylation of aromatic amines. Sodium cyano-
borohydride (5 g, 75.6 mmol) was added to a stirred solution
of 5 g (23.4 mmol) of dimethyl 5-aminoisophthalate (6)
(98%) and 20 mL (269 mmol) of 37% aqueous formalde-
hyde in 150 mL acetonitrile. This was followed by slow addi-
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
203

Zheng Huang et al.
FULL PAPERS
tion (over a period of 20 min) of 3 mL of glacial acetic acid
to adjust the pH at 5–6. The resulting solution was stirred at
room temperature for 8 h and the solvent was removed
under reduced pressure. The wet-solid obtained thereby was
washed thoroughly with distilled water and air-dried to give
7 as a light yellowish-white solid; yield: 5.39 g (97%).
1H NMR (CDCl3): d=8.01 (s, 1H, Ar), 7.56 (s, 2H, Ar),
3.94 (s, 6H, CO2CH3), 3.05 [s, 6H, (CH3)2N].
under vacuum, giving the ligand 10 as a white solid; yield:
1
1.03 g (72%). 31P{1H} NMR (C6D6): d=31.03 (s); H NMR
2
(C6D6): d=6.98 (s, 1H, Ar), 6.79 (s, 2H, Ar), 2.87 (d, JH,P
=
=
3
2.4 Hz, 4H, CH2), 2.78 [s, 6H, (CH3)2N], 1.18 [d, JH,P
10.8 Hz, 36H, CACTHNUTRGNEUNG(CH3)3].
Synthesis of (Me2N-PCP)IrHCl
To 0.51 g of 10 (1.16 mmol) in 30 mL of toluene was added
0.38 g of [Ir(COD)Cl]2 (0.57 mmol) at room temperature
AHCTUNGTRENNUNG
Synthesis of 5-Dimethylamino-1,3-benzenedimethanol
(8)
and stirred for 30 min under a hydrogen atmosphere. (note:
the solution changes color from yellow to deep red under
the hydrogen atmosphere at room temperature). This mix-
ture was refluxed for three days under the hydrogen with
stirring, and the solvent was removed under vacuum giving
4b as dark-red solid; yield: 0.78 g (94%). 31P{1H} NMR
To a stirred suspension of 2.69 g (67.4 mmol) of lithium alu-
minum hydride (95%) in THF (50 mL) at 08C under an
argon atmosphere was slowly added a THF (100 mL) solu-
tion of 7 (5 g, 21.1 mmol). After the addition was complete,
the resultant suspension was refluxed for 18 h, diluted with
100 mL tetrahydrofuran and cooled to 08C. Excess LiAlH4
was quenched by slow addition of a saturated sodium sulfate
solution followed by distilled water and the suspension was
stirred at 08C - 58C for 1 h (until the gray color of LiAlH4
disappeared completely). The suspension was filtered
through a pad of anhydrous magnesium sulfate and subse-
quently washed with ethyl acetate (3ꢁ50 mL). The com-
bined filtrates were concentrated under reduced pressure to
give a clear colorless oil that crystallized upon standing. The
product 8 was recrystallized from THF/heptane system as a
white powder; yield: 3.49 g (91%). 1H NMR (DMSO-d6):
d=6.58 (s, 1H, Ar), 6.56 (s, 2H, Ar), 5.03 (t, 2H, OH), 4.41
(d, 4H, CH2), 2.87 [s, 6H, (CH3)2N].
1
(C6D6): d=67.03 (s); H NMR (C6D6): d=6.65 (s, 2H, Ar),
3.16 (dvt, the left part of ABX2
3.9 Hz, 2H, CH2), 3.06 (dvt, the right part of ABX2 , JH,H
,
2JH,H =17.7 Hz, JH,P
=
=
2
17.7 Hz, apparent J=3.9 Hz, 2H, CH2), 2.77 [s, 6H,
(CH3)2N], 1.34 [virtual triplet, apparent J=6.9 Hz, 18H, C-
AHCTUNGTERN(GNUN CH3)3], 1.29 [virtual triplet, apparent J=6.9 Hz, 18H, C-
2
N
Synthesis of (Me2N-PCP)IrH4 and (Me2N-PCP)IrH2
(1d)
A stream of hydrogen was passed through a solution of
0.73 g of (Me2N-PCP)IrHCl (1.1 mmol) in 300 mL anhy-
drous pentane for about 30 min. This was followed by a
slow dropwise addition of 1.1 mL of 1M LiBEt3H in THF
(1.1 mmol) to this solution with continuous stirring under a
hydrogen atmosphere. The solution turned nearly colorless
and some white precipitate was formed at the bottom of the
flask. After the addition of LiBEt3H was complete, stirring
was continued for 1 h and finally the solution was filtered
under argon atmosphere. (note: on changing from H2 to
argon atmosphere the solution rapidly turned deep red).
The solvent was removed under vacuum, giving 1d as red-
dish brown crystals containing ca. 10% of (Me2N-PCP)IrH4;
yield: 0.55 g (79%). (All the PCP-Ir tetrahydride is quickly
converted to the corresponding dihydride under catalytic
conditions; accordingly, for catalytic runs, the dihydrides are
frequently used containing varying amounts of tetrahy-
dride.)
Synthesis of 1,3-Bis(bromomethyl)-5-
dimethylaminoACHTUNGTRENNUNGbenzene (9)
PBr3 (10.4 mL, 110 mmol) was added dropwise over a 30-
minute period to a stirred solution of 8 (5.0 g, 27.6 mmol) in
140 mL of anhydrous acetonitrile at 08C under aqn argon
atmosphere. The solution was stirred at room temperature
for 2 h and then heated to 708C for an additional 4 h. The
reaction was quenched by pouring the solution over ice fol-
lowed by slow addition of a saturated NaHCO3 solution to
adjust pH to ~7. The solution was filtered and the precipi-
tated product was dissolved in acetonitrile. Pure 9 was re-
crystallized out from acetonitrile/water system to give a
1
white powder; yield: 6.52 g (77%). H NMR (CDCl3): d=
NMR data for (Me2N-PCP)IrH4: 31P{1H} NMR (C6D6):
6.78 (s, 1H, Ar), 6.66 (s, 2H, Ar), 4.45 (s, 4H, CH2), 2.99 [s,
6H, (CH3)2N].
1
d=72.42 (s); H NMR (C6D6): d=6.73 (s, 2H, Ar), 3.32 (vir-
tual triplet, apparent J=3.9 Hz, 4H, CH2), 2.79 [s, 6H,
(CH3)2N], 1.24 [virtual triplet, apparent J=6.9 Hz, 36H, C-
(CH3)3], À9.09 (t, JH,P =9.9 Hz, 4H, IrH4).
Synthesis of 1,3-Bis[di(tert-butyl)phosphinomethyl]-5-
dimethylaminobenzene (Me2N-PCP-H) (10)
2
Synthesis of this ligand and its corresponding iridium hydri-
do chloride were based on reported syntheses by Shaw for
the parent ligand.[16] To 1.0 g of 9 (3.25 mmol) in 20 mL of
degassed acetone was added 1.36 mL (7.2 mmol) of di-tert-
butylphosphine (98%) (Strem) at room temperature. The
mixture was heated under reflux with stirring for 24 h under
an argon atmosphere, and the solvent was removed under
vacuum. The solid was dissolved in degassed deionized
water (15 mL) and treated with a solution of potassium car-
bonate (2.7 g, 19.5 mmol) in degassed deionized water
(10 mL). The diphosphine ligand was extracted with de-
gassed n-hexane (3ꢁ20 mL) and the solvent was evaporated
1H NMR (C6D6): d=6.84 (s, 2H, Ar), 3.62 (virtual triplet,
apparent J=3.6 Hz, 4H, CH2), 2.76 [s, 6H, (CH3)2N], 1.33
[virtual triplet, apparent J=6.9 Hz, 36H, C
(t, JH,P =8.7 Hz, 2H, IrH2).
(CH3)3], À19.99
2
Synthesis of {p-OCAHTUNGTREN(GNUN CH2)3SiACHUTNRTGEGN(UNN OMe)3-C6H2-2,6-[OPACHTUNGTRENNUNG(t-
Bu)2]2}Ir(C2H4) (15)
AHCTUNGTRENNUNG
Complex 4 (80 mg, 0.119 mmol), IACTHNUTRGENN(UG CH2)3SiACHTUGNTERN(NUGN OMe)3 (138 mg,
0.476 mmol), and THF (10 mL) were added to a Kontes
flask. The THF solution was degassed by freeze-pump-thaw
cycles. The flask was refilled with ethylene gas at À788C.
204
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206

Highly Active and Recyclable Heterogeneous Iridium Pincer Catalysts
FULL PAPERS
sion was degassed by freeze-pump-thaw cycles. The flask
was refilled with ethylene gas at À788C and the suspension
was stirred at 1208C for 2 days. The flask was cooled to
room temperature and ethylene gas was removed by freeze-
pump-thaw cycles. Excess trimethylsilyldimethylamine
(3 mL, 18.7 mmol) was added in the glove-box and the flask
was degassed and refilled with ethylene gas at À788C. The
suspension was stirred at room temperature for 2 days. This
supported catalyst was filtered under argon, and washed
with pentane, toluene and THF three times (5 mL each), re-
spectively. The orange solid was dried under high vacuum
overnight to give the product; yield: 1.46 g.
The mixture was heated at 658C for 2 h. NaOMe (22 mg,
0.407 mmol) was then added to the Kontes flask in the
glove-box and the flask was refilled with ethylene gas at
À788C. The mixture was stirred at room temperature for 2
Synthesis of Silica-Supported Iridium Pincer Complex
16
days. NaOMe reacted with the excess of I
G
ACHTUNGERTN(NUNG OMe)3
(808C/2 mm) to produce NaI and CH2 =CHCH2Si
ACHTUNGTRENNUNG
Complex 16 (48 mg, 0.062 mmol), silica (1.05 g), and COE
(5 mL) were added to a Kontes flask. The COE suspension
was stirred at 1508C for 2 days. The flask was cooled to
room temperature and excess trimethylsilyldimethylamine
(3 mL, 18.7 mmol) was added in the glove-box. The suspen-
sion was stirred at room temperature for 2 days. This sup-
ported catalyst was filtered under argon, and washed with
pentane, toluene and THF three times (5 mL each), respec-
tively. The light orange solid was dried under high vacuum
overnight to give the product; yield: 0.86 g.
which is relatively more volatile (bp, 1468C/760 mm) and
easier to remove. Volatiles were then removed under high
vacuum. The residue was extracted with 3ꢁ10 mL of pen-
tane, and the extract was filtered through a 0.2 mm pore size
syringe filter (Nalgene 199–2020) into a Schlenk flask. Re-
moval of the solvent under high vacuum afforded of a red
waxy solid which contained ca. 95% of 11 by NMR and was
used without further purification in the next step; yield:
1
69 mg (0.086 mmol, 73%). H NMR (400 MHz, 238C, C6D6):
d=0.75 ((m, 2H, CH2Si), 1.27 (virtual triplet, apparent J=
6.6 Hz, 36H, 4ꢁt-Bu), 1.88 (m, 2H, OCH2CH2), 3.14 (t,
3JP, H =2.7 Hz, 4H, C2H4), 3.40, (s, 9H, 3ꢁOMe), 3.71 (t,
2JH,H =6.5 Hz, 2H, OCH2CH2), 6.65 (s, 2H, 3- and 5-H);
31P{1H} NMR (162 MHz, 238C, C6D6): d=181.4.
Hydrogen Transfer from COA to TBE, Catalyzed by
Silica-Supported Iridium Pincer Complexes
The silica-supported complex 15 or 16 (2.0–3.15 mmol),
COA (0.265–0.674 g, 2.36–6.0 mmol), and TBE (95%, 0.209–
0.532 g, 2.36–6.0 mmol) were added to a Kontes flask. The
flask was sealed tightly with a Teflon plug under an argon
atmosphere and the suspension stirred in an oil bath at
2008C. Periodically, the flask was removed from the bath
and cooled in an ice bath. An aliquot was removed from the
flask, and analyzed by GC (method A). Turnover numbers
were calculated for each aliquot. Details are summarized in
the text.
Synthesis of {p-OACHTUNGTRENNUNG(CH2)3Si(Me)2AHCTUNTGERN(NUGN OMe)-C6H2-2,6-
[OP(t-Bu)2]2}Ir(C2H4) (16)
A
ACHTUNGTRENNUNG
Acknowledgements
The same synthetic procedure used for 15 was used except
that (CH2)3Si(OMe)3 was replaced by (CH2)3Si(Me)2
ACHTUNGTRENNUNG(OMe) (123 mg, 0.476 mmol), giving a red waxy solid which
This work is financially supported by the NSF under the aus-
pices of the Center for Enabling New Technologies through
Catalysis (CENTC).
I
N
G
IACHTUNGTRENNUNG
contained ca. 95% of 12 by NMR and was used without fur-
ther purification in the next step; yield: 63 mg (0.082 mmol,
1
69%). H NMR (400 MHz, 238C, C6D6): d=0.09 [s, 6H, Si-
ACHTUNGTRENNUNG
References
3JP, H =2.7 Hz, 4H, C2H4), 3.17, (s, 3H, OMe), 3.71 (t, JH,H
=
2
[1] a) P. Wieseman, Petrochemicals, Ellis Horwood, Chi-
chester, England, 1986, pp 90–91; b) K. Weissermel,
H.-J. Arpel, Industrial Organic Chemistry, Wiley-VCH,
Weinheim, Germany, 2003, pp 59–89.
6.4 Hz, 2H, OCH2CH2), 6.68 (s, 2H, 3- and 5-H); 31P{1H}
NMR (162 MHz, 238C, C6D6): d=181.5.
[2] For recent reviews on this topic, see: a) F. Kakiuchi, N.
Chatani, Adv. Synth. Catal. 2003, 345, 1077–1101;
b) J. A. Labinger, J. E. Bercaw, Nature 2002, 417, 507–
514; c) R. H. Crabtree, J. Chem. Soc. Dalton Trans.
2001, 2437–2450; d) C. M. Jensen, Chem. Commun.
Synthesis of Silica-Supported Iridium Pincer Complex
15
Complex 15 (50 mg, 0.062 mmol), silica (1.5 g), and toluene
(10 mL) were added to a Kontes flask. The toluene suspen-
Adv. Synth. Catal. 2009, 351, 188 – 206
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
205

Zheng Huang et al.
FULL PAPERS
1999, 2443–2449; e) R. G. Begman, Nature 2007, 446,
391–393.
[12] A. Ray, PhD thesis, Rutgers University, 2007.
[13] R. F Borch, A. I. Hassid, J. Org. Chem. 1972, 37, 1673.
[14] S. L. Stang, R. Meier, C. Rocaboy, J. A. Gladysz, J. Flu-
orine Chem. 2003, 119, 141.
[15] Y. Mao, V. Boekelheide, J. Org. Chem. 1980, 45, 2746.
[16] C. J. Moulton, B. L. Shaw, J. Chem. Soc. Dalton Trans.
1976, 1020.
[17] CCDC 712608 contains the supplementary crystallo-
graphic data for 1d-CO reported in this paper. These
data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.
cam.ac.uk/data_request/cif.
[18] M. Trueba, S. P. Trasatti, Eur. J. Inorg. Chem. 2005,
3393–3403.
[19] R. H. Crabtree, J. M. Mihelcic, J. M. Quirk, J. Am.
Chem. Soc. 1979, 101, 7738–7739.
[20] NIST Standard Reference Database Number 69
[3] a) M. Gupta, C. Hagen, R. J. Flesher, W. C. Kaska,
C. M. Jensen, Chem. Commun. 1996, 2083–2084; b) M.
Gupta, C. Hagen, C. W. Kaska, R. E. Cramer, C. M.
Jensen, J. Am. Chem. Soc. 1997, 119, 840–841; c) W.-W.
Xu, G. P. Rosini, M. Gupta, C. M. Jensen, W. C. Kaska,
K. Krogh-Jespersen, A. S. Goldman, Chem. Commun.
1997, 2273–2274.
[4] I. Gçttker-Schnetmann, P. S. White, M. Brookhart, J.
Am. Chem. Soc. 2004, 126, 1804–1811.
[5] Jensen et al. prepared the analogous i-Pr-POCOP iridi-
um complex and studied its alkane dehydrogenation ac-
tivity. D. Morales-Morales, R. Redꢃn, C. Yung, C. M.
Jensen, Inorg. Chim. Acta 2004, 357, 2953–2956.
[6] a) M. W. Haenel, S. Oevers, K. Angermund, W. C.
Kaska, H.-J. Fan, M. B. Hall, Angew. Chem. 2001, 113,
3708–3712; Angew. Chem. Int. Ed. 2001, 40, 3596–
3600; b) S. A. Kuklin, A. M. Sheloumov, F. M. Dolgush-
in, M. G. Ezernitskaya, A. S. Peregudov, P. V. Petrov-
skii, A. A. Koridze, Organometallics 2006, 25, 5466–
5476.
[7] a) X. Zhang, A. Fried, S. Knapp, A. S. Goldman,
Chem. Commun. 2003, 2060–2061; b) W. H. Bernskoet-
ter, M Brookhart, Organometallics 2008, 27, 2036–
2045.
[8] D. Morales-Morales, R. Redon, Z. Wang, D. W. Lee, C.
Yung, K. Magnuson, C. M. Jensen, Can. J. Chem. 2001,
79, 823–829.
–
March, 2003 Release, http://webbook.nist.gov/chemis-
try/.
[21] TONs were extracted from the ratio of TBE, TBA,
DM1B, and DM2B analyzed by GC. The ratio of COE
and 1,3-COD was extracted from the olefinic signals of
COE and COD in 13C[1H] NMR (COE and COD have
identical retention times in GC). Determined by NMR,
the sum of COE and COD double bonds equals TON
of TBE within 5% difference.
[22] W. O. Haag, H. Pines, J. Am. Chem. Soc. 1960, 82,
2488–2494.
[23] K. Krogh-Jespersen, M. Czerw, K. Zhu, B. Singh, M.
Kanzelberger, N. Darji, P. D. Achord, K. B. Renkema,
A. S. Goldman, J. Am. Chem. Soc. 2002, 124, 10797–
10809.
[9] A. Ray, K. Zhu, Y. V. Kissin, A. E. Cherian, G. W.
Coates, A. S. Goldman, Chem. Commun. 2005, 3388–
3390.
[10] a) I. Gçttker-Schnetmann, M. Brookhart, J. Am. Chem.
Soc. 2004, 126, 9330–9338; b) K. B. Renkema, Y. V.
Kissin, A. S. Goldman, J. Am. Chem. Soc. 2003, 125,
7770–7771; c) K. Krogh-Jespersen, M. Czerw, N.
Summa, K. B. Renkema, P. D. Achord, A. S. Goldman,
J. Am. Chem. Soc. 2002, 124, 11404–11416; d) K.
Krogh-Jespersen, M. Czerw, K. Zhu, B. Singh, M. Kan-
zelberger, N. Darji, P. D. Achord, K. B. Renkema, A. S.
Goldman, J. Am. Chem. Soc. 2002, 124, 10797–10809;
e) S. Li, M. B. Hall, Organometallics 2001, 20, 2153–
2160; f) D. Morales-Morales, D. W. Lee, Z. Wang,
C. M. Jensen, Organometallics 2001, 20, 1144–1147;
g) M. Kanzelberger, B. Singh, M. Czerw, K. Krogh-Jes-
persen, A. S. Goldman, J. Am. Chem. Soc. 2000, 122,
11017–11018; h) F. Liu, E. B. Pak, B. Singh, C. M.
Jensen, A. S. Goldman, J. Am. Chem. Soc. 1999, 121,
4086–4087; i) S. Niu, M. B. Hall, J. Am. Chem. Soc.
1999, 121, 3992–3999; j) F. Liu, A. S. Goldman, Chem.
Commun. 1999, 655–656; k) D. W. Lee, W. C. Kaska,
C. M. Jensen, Organometallics 1998, 17, 1–3.
[24] K. Zhu, P. D. Achord, X. Zhang, K. Krogh-Jespersen,
A. S. Goldman, J. Am. Chem. Soc. 2004, 126, 13044–
13053.
[25] X. Zhang, M. Kanzelberger, T. J. Emge, A. S. Goldman,
J. Am. Chem. Soc. 2004, 126, 13192–13193.
[26] A. S. Goldman, K. Krogh-Jespersen, J. Am. Chem. Soc.
1996, 118, 12159–12166.
[27] Covalent attachment of palladium pincer complexes to
silica has been reported by Weck and Jones, and
van Koten: a) K. Yu, W. J. Sommer, J. M. Richardson,
M. Weck, C. W. Jones, Adv. Synth. Catal. 2005, 347,
161–171; b) W. J. Sommer, K. Yu, J. S. Sears, Y. Ji, X.
Zhang, R. J. Davis, C. D. Sherrill, C. W. Jones, M.
Weck, Organometallics 2005, 24, 4351–4361; c) N. C.
Mehendale, J. R. A. Sietsma, K. P. de Jong, C. A. van
Walree, R. J. M. Klein Gebbink, G. van Koten, Adv.
Synth. Catal. 2007, 349, 2619–2630.
[28] K. Yu, W. J. Sommer, M. Weck, C. W. Jones, J. Catal.
2004, 226, 101–110.
[29] J. L. Herde, J. C. Lambert, C. V. Senoff, Inorg. Synth.
1974, 15, 18–19.
[11] A. S. Goldman, A. H. Roy, Z. Huang, R. Ahuja, W.
Schinski, M. Brookhart Science 2006, 312, 257–261.
206
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 188 – 206