Iridium-catalyzed dehydrogenative decarbonylation of primary alcohols with the liberation of syngas

Article abstract of DOI:10.1002/chem.201202631

A new iridium-catalyzed reaction in which molecular hydrogen and carbon monoxide are cleaved from primary alcohols in the absence of any stoichiometric additives has been developed. The dehydrogenative decarbonylation was achieved with a catalyst generated in situ from [Ir(coe)₂Cl]₂ (coe=cyclooctene) and racemic 2,2'-bis(diphenylphosphino)-1,1'-binaphthyl (rac-BINAP) in a mesitylene solution saturated with water. A catalytic amount of lithium chloride was also added to improve the catalyst turnover. The reaction has been applied to a variety of primary alcohols and gives rise to products in good to excellent yields. Ethers, esters, imides, and aryl halides are stable under the reaction conditions, whereas olefins are partially saturated. The reaction is believed to proceed by two consecutive organometallic transformations that are catalyzed by the same iridium(I)-BINAP species. First, dehydrogenation of the primary alcohol to the corresponding aldehyde takes place, which is then followed by decarbonylation to the product with one less carbon atom.

Full text of DOI:10.1002/chem.201202631

FULL PAPER
DOI: 10.1002/chem.201202631
Iridium-Catalyzed Dehydrogenative Decarbonylation of Primary Alcohols
with the Liberation of Syngas
Esben P. K. Olsen and Robert Madsen*[a]
Abstract: A new iridium-catalyzed re-
action in which molecular hydrogen
and carbon monoxide are cleaved from
primary alcohols in the absence of any
stoichiometric additives has been de-
veloped. The dehydrogenative decar-
bonylation was achieved with a catalyst
amount of lithium chloride was also
added to improve the catalyst turnover.
The reaction has been applied to a va-
riety of primary alcohols and gives rise
to products in good to excellent yields.
Ethers, esters, imides, and aryl halides
are stable under the reaction condi-
tions, whereas olefins are partially satu-
rated. The reaction is believed to pro-
ceed by two consecutive organometal-
lic transformations that are catalyzed
by the same iridium(I)–BINAP species.
First, dehydrogenation of the primary
alcohol to the corresponding aldehyde
takes place, which is then followed by
decarbonylation to the product with
one less carbon atom.
generated in situ from [IrACHTUNGRTENUNG(coe)2Cl]2
(coe=cyclooctene) and racemic 2,2’-
bis(diphenylphosphino)-1,1’-binaphthyl
(rac-BINAP) in a mesitylene solution
Keywords: alcohols · decarbonyla-
tion · dehydrogenation · iridium ·
syngas
saturated with water.
A
catalytic
Introduction
rhodium catalyst to shorten an aldose chain by one carbon
atom.[10] The decarbonylation has also been used on several
occasions in conjunction with the Pauson–Khand reaction,
in which the aldehyde serves as an in situ carbon monoxide
source and the metal promotes two different transforma-
tions in the same pot.[11] Recently, cinnamyl alcohol was em-
ployed as the carbon monoxide source in a rhodium-cata-
lyzed Pauson–Khand reaction.[12] Although several alcohols
were examined in this transformation, cinnamyl alcohol
gave the best results, presumably because it can also act as a
hydrogen acceptor. Very recently, [Cp*IrCl2]2 was shown to
catalyze the Guerbet reaction of primary alcohols in the
presence of potassium tert-butoxide; the product alcohols
subsequently underwent a dehydrogenation with concomi-
tant extrusion of carbon monoxide under the reaction condi-
tions.[13] Furthermore, the iridium–PNP pincer complex
Ir[N(2-PiPr2-4-MeC6H3)2][14] and the ruthenium complex
The dehydrogenation of an alcohol constitutes the key step
in a number of homogeneous transition-metal-catalyzed
transformations, such as the transfer hydrogenation of ke-
tones with isopropanol,[1] the alkylation of amines and C-nu-
cleophiles with alcohols,[2] and the dehydrogenative synthesis
of amides and imines from alcohols and amines.[3] In addi-
tion, the direct conversion of alcohols into aldehydes/ke-
tones and hydrogen gas can be accomplished with a low cat-
alyst loading at temperatures between 90 and 1308C.[4,5]
These hydrogen transfer reactions are most effectively cata-
lyzed by various ruthenium and iridium complexes, although
other platinum metal catalysts have also been employed. In
hydrogen-transfer reactions with primary alcohols, a com-
peting decarbonylation with the release of carbon monoxide
has sometimes been observed.[3b,4] This side reaction occurs
from the intermediate aldehyde and may lead to poisoning
of the catalyst.[4,5c]
[RuH4ACHTNUGRTNEUNG
(PPh3)3][15] have been reacted stoichiometrically with
ethanol to yield hydrogen gas, methane, and the correspond-
ing metal–CO complexes.
The direct decarbonylation of aldehydes has also been de-
scribed with several platinum metal catalysts,[6] and the most
effective are based on rhodium,[7] iridium,[8] and palladium.[9]
In the absence of a carbon monoxide scavenger the catalytic
decarbonylation can be achieved at temperatures ranging
from 100 to 1608C. The reaction has been conducted with a
The mechanism for the dehydrogenation of alcohols to
carbonyl compounds has been extensively investigated with
many different catalysts,[16,17] whereas the mechanism for the
decarbonylation of aldehydes has only been thoroughly
studied with a rhodium catalyst.[18] We speculated whether a
platinum metal catalyst would be able to mediate both the
dehydrogenation of a primary alcohol and the decarbonyla-
tion of the resulting aldehyde, and in this way liberate both
hydrogen gas and carbon monoxide in the absence of a scav-
enger. This transformation would be a useful new tool in or-
ganic synthesis and could also find applications for degrad-
ing biomass-derived alcohols. Herein, we describe a new iri-
dium-catalyzed dehydrogenative decarbonylation reaction
[a] E. P. K. Olsen, Prof. Dr. R. Madsen
Department of Chemistry, Building 201
Technical University of Denmark
2800 Kgs. Lyngby (Denmark)
Fax : (+45)4593-3968
E-mail: rm@kemi.dtu.dk
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202631.
Chem. Eur. J. 2012, 18, 16023 – 16029
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
16023

À
Table 1. Dehydrogenative decarbonylation of 2-naphthylmethanol with
different ligands.
of primary alcohols in which a C C bond is cleaved and
syngas is liberated.
Results and Discussion
For the initial screening, 2-naphthylmethanol was selected
as the substrate because this would give the stable conjugat-
ed 2-naphthaldehyde in the process and make the decarbon-
ylation step more demanding.
The task was to find a catalyst that would perform both
the dehydrogenation and the decarbonylation without being
deactivated by hydrogen or carbon monoxide. Complexes
Entry
Ligand
PPh3
PPh3
PtBu3·HBF4
PtBu3·HBF4
Yield [%][a]
A
17[b]
6
5
[Ru
(PPh3)3Cl], [Cp*IrCl2]2, and [Ir
in refluxing mesitylene with PPh3 and BINAP as the ligands.
Only [Ir(coe)2Cl]2 gave a good conversion to the desired
A
N
ACHTUNGTERN(UNGN ethene)2Cl]2, [Rh-
B
A
ACHTUNGTRENNUNG
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
34[b]
44
41
43
nd[e]
27
28
31
27
42
46
49
36
35
20
27
35
42
39
45
32
42
36
23
[c]
ACHTUNGTRENNUNG
[d]
product and was therefore selected as the precursor for a
more detailed ligand screening (Table 1). To achieve a good
comparison, the reactions were all stopped prematurely
after 8 h and yields were determined by GC. Although the
[c,d]
6
[c,d]
PCy3·HBF4
nd[e]
10
8
SIpTol·HCl[d]
SIpTol·HCl[c,d]
1,10-phenanthroline
2,2’-bipyridyl
dppe
4
3
9
12
9
[IrACHTUNGTRENNUNG(cod)Cl]2/PPh3 system has shown excellent results in the
decarbonylation step,[8b] PPh3 and other monodentate li-
gands were poor ligands for the dehydrogenative decarbony-
lation (Table 1, entries 1–7). Curiously, a substantial amount
of the deoxygenated side product 2-methylnaphthalene was
dppp
dppb
1,5-dpp-pentane
1,6-dpp-hexane
rac-BINAP
7
7
observed when a 1:2 ratio of [IrACHTNUTRGNEUNG(coe)2Cl]2 and PPh3 was ap-
61
59
27
16
17
18
21
13
17
5
plied (Table 1, entry 1). With the use of PCy3·HBF4 and
K2CO3 as the base, the Friedel–Crafts product 2-(2,4,6-tri-
methylbenzyl)naphthalene was formed as the major product
and isolated in 42% yield (Table 1, entry 5). The same side
product was formed in a substantial amount together with 2-
methylnaphthalene when the reaction was performed in the
absence of a phosphine ligand. Except for these cases, the
reactions in Table 1 were all very clean and showed a mix-
ture of only unreacted 2-naphthylmethanol with the naph-
thalene and 2-naphthaldehyde products.
Bidentate pyridyl ligands also provided limited amounts
of the desired product (Table 1, entries 8 and 9). The same
was observed for bidentate phosphine ligands with a flexible
aliphatic backbone (Table 1, entries 10–14). However, im-
proved yields were obtained when the bidentate phosphine
ligands contained a more rigid backbone (Table 1, en-
tries 15–22). In this category, rac-BINAP with a biaryl back-
bone and a bite angle[19] of around 938 gave the highest yield
(Table 1, entry 15). A similar result was observed with the
use of BIPHEP (Table 1, entry 16), whereas substituted ana-
logues of BIPHEP gave lower yields (Table 1, entries 17 and
18). Rigid ligands with a smaller or a larger bite angle than
BINAP (Table 1, entries 19–22), as well as the ligand Dave-
phos (Table 1, entry 23) and the pincer ligand 1,3-C6H4-
BIPHEP[f]
3,3’-dpp-H8-[2,2’]binaphthalene
6,6’-dpp-H4-[5,5’]biisobenzofuran
cis-1,2-dpp-ethylene
1,2-dpp-benzene
dppf
Xantphos[g]
Davephos[h]
1,3-C6H4ACTHNUTRGENNUG(CH2PCAHTUNGTREN(NNGU tBu)2)2
[a] Determined by GC. [b] 2-Methylnaphthalene (41%) was also formed.
[c] 10.0 mol%. [d] A 1:1 ratio of ligand and base was applied. [e] nd=not
determined, but 42% of 2-(2,4,6-trimethylbenzyl)naphthalene was isolat-
ed. [f] 2,2’-Bis(diphenylphosphino)-1,1’-biphenyl. [g] 4,5-Bis(diphenyl-
phosphino)-9,9-dimethylxanthene.
(N,N-dimethylamino)-1,1’-biphenyl. Tol=toluene, dpp=diphenylphos-
phino, dppf=1,1’-bis(diphenylphosphino)ferrocene.
[h] 2-(Dicyclohexylphosphino)-2’-
lower boiling points (Table 2, entries 1–4). Polar solvents
with high boiling points also gave lower conversions, which
is presumably due to coordination and inactivation of the
iridium complexes in the catalytic cycle (Table 2, entries 5–
7). As a result, mesitylene seems to be the solvent of choice
for the dehydrogenative decarbonylation and the reaction
should be carried out at reflux.
Different iridium-dimer complexes were also investigated,
as depicted in Table 3. Hardly any difference in reactivity
could be determined between the cyclooctadiene and the cy-
clooctene complex, although the cyclooctadiene ligand was
hydrogenated to a larger extent than the cyclooctene ligand
by the molecular hydrogen developed during the reaction
A
ACHTUNGTRENNUNG(tBu)2)2 (Table 1, entry 24), also gave lower product
By using rac-BINAP as the preferred ligand, a variety of
high-boiling solvents were screened, as illustrated in Table 2.
A lower yield of the desired product was obtained from re-
actions in xylene and toluene than in mesitylene due to the
(Table 3, entries 1 and 2). A comparison of [IrACTHNURTGNENG(U cod)Cl]2, [Ir-
16024
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16023 – 16029

Iridium-Catalyzed Liberation of Syngas
FULL PAPER
Table 2. Dehydrogenative decarbonylation of 2-naphthylmethanol in
high boiling solvents.
150 ppm or 0.07 mol% of water under these reaction condi-
tions. Initially, water was believed to act as a proton source
and the weak Brønsted acids benzoic acid, 2,4,6-trichloro-
benzoic acid, and 3-fluorobenzoic acid were therefore also
investigated as additives, but they all gave rise to a slower
reaction. The same result was obtained with bases such as
K2CO3, KOtBu, imidazole, DMAP, and 2,6-diisopropylani-
line.
Entry
Solvent
T [8C]
Yield [%][a]
A
B
1
2
3
4
5
6
7
mesitylene
p-xylene
p-xylene (H2O, LiCl)
toluene
diglyme
DMA
164
138
138
110
162
165
170[b]
61
25
33
7
24
24
15
20
31
30
24
25
20
19
The decarbonylation step is the reverse of a migratory
CO insertion reaction, which can be accelerated by nucleo-
philic solvents.[21] Thus the effect of water could originate
from its nucleophilicity. DMSO and triphenylphosphine
oxide were therefore also investigated as additives. The ex-
periments were performed by taking samples every hour
and analyzing them by GC. In addition, LiBr and NaCl
were included to elucidate the effect of LiCl. The result of
the comparison is shown in Figure 1.
NMP
[a] Determined by GC. [b] Reaction was run at lower temperature than
the boiling point of NMP. DMA=dimethylacetamide, NMP=N-methyl-
pyrrolidinone.
It is clear that DMSO, LiBr, and NaCl do not show any
notable rate enhancements. Interestingly, triphenylphos-
phine oxide accelerated the reaction, but not as much as
H2O combined with LiCl. An experiment with 100 mol% of
D2O was performed and resulted in deuterium being scram-
bled into the cyclooctene ligand, probably catalyzed by an
iridium–hydride species. Some incorporation of deuterium
was also found in the naphthalene product according to GC-
MS analysis. D2O has previously been observed to scramble
hydrogen with deuterium in iridium–dihydride complexes.[22]
This result indicates that a proton exchange between water
and an iridium–hydride complex can happen during the de-
hydrogenative decarbonylation, but the exact reason for the
rate enhancement with water remains unknown.
The effect of lithium chloride is also not known in detail.
Under hydrogenation conditions, iridium(I)–phosphine com-
plexes can form inactive bridging hydride species[23] depend-
ing on the conditions and the steric demand of the phos-
phine ligand.[24] It may be that the presence of lithium chlor-
ide in our case hampers the formation of these rather stable
dimeric and trimeric clusters, and in this way leads to an im-
proved catalyst turnover.
Table 3. Dehydrogenative decarbonylation of 2-naphthylmethanol by iri-
dium dimer complexes.
Entry
Ir dimer
Yield [%][a]
1
2
3
4
5
[Ir
[Ir
[Ir
[Ir
[Ir
N
61
62
38
30
nd[b]
[a] Determined by GC. [b] Not determined due to the significant forma-
tion of 2-(2,4,6-trimethylbenzyl)naphthalene as a side product.
ACHTUNGTRENNUNG(cod)Br]2, and [IrACHTUNGTRENNUNG(cod)OMe]2 clearly revealed that use of
the chloride complex gave a higher yield than the bromide
and the methoxide analogues (Table 3, entries 2–4). Interest-
ingly, the Friedel–Crafts side product was observed when
À
the weakly coordinating BF4 was applied as the counterion
(Table 3, entry 5).
The initial screenings with 2-naphthylmethanol have led
To evaluate the effect of additives on the iridium-cata-
lyzed transformation, different Lewis acids were added to
to an optimized catalyst system consisting of [Ir
rac-BINAP, and LiCl in mesitylene saturated with water.
ACHTUNGTERN(NUNG coe)2Cl]2,
the mixture of [Ir
rac-BINAP, and 2-naphthylme-
thanol. Bi(OTf)3, FeCl3, TeCl4,
ACHTUNGTERN(NUNG coe)2Cl]2,
ACHTUNGTRENNUNG
MgBr2·OEt2, and BF3·OEt2 all
retarded the reaction, whereas
LiCl seemed to enhance the
rate of the transformation. Fur-
thermore, it was discovered
that water also had a remarka-
ble effect on the progress of
the reaction.[20] The optimal
amount of water needed for a
rate acceleration was found to
be a saturated solution in mesi-
tylene, which corresponds to
Figure 1. Dehydrogenative decarbonylation of 2-naphthylmethanol with different additives.
Chem. Eur. J. 2012, 18, 16023 – 16029
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16025

R. Madsen and E. P. K. Olsen
With this system in hand, the scope of the reaction was in-
vestigated with different alcohols. First, a variety of benzylic
alcohols were transformed under the optimized conditions
(Table 4). Ether groups were tolerated and the correspond-
ing products could be isolated in good yields (Table 4, en-
Even though the catalytic system was optimized for ben-
zylic alcohols, the dehydrogenative decarbonylation also
worked smoothly with nonbenzylic substrates (Table 5).
Table 5. Dehydrogenative decarbonylation of non-benzylic alcohols.
Table 4. Dehydrogenative decarbonylation of benzylic alcohols.
Entry Substrate
1
t
Product
Isolated
yield [%]
[h]
Entry Substrate
1
t [h] Product
Isolated yield [%]
92
16
91
20
2
3
4
16
18
24
66
90
96
2
3
4
20
16
18
93
94
91
5
6
18
60
91
79
5
6
18
24
92
87
7[a]
16
95
7
8
20
16
93
76
[a] A 10:1 inseparable mixture of the saturated to unsaturated product
was obtained, as determined by NMR.
2-(6-Hydroxyhexyl)isoindoline-1,3-dione was converted
cleanly into the corresponding 2-pentylisoindoline, which
was isolated in 91% yield (Table 5, entry 1). Silyl ethers
were tolerated under the reaction conditions and the more
stable tert-butyldiphenyl silyl (TBDPS) ether gave the best
yield (Table 5, entries 2 and 3). The sulfonamide needed a
slightly longer reaction time before full conversion was ob-
tained, but then an excellent 96% yield was isolated
(Table 5, entry 4). The diol derived from a 2,2-disubstituted
malonate was completely transformed within the same time
as a single primary alcohol, despite the turnover number
being double (Table 5, entry 5). This example represents a
useful synthetic tool for the double alkylation of a CH2
group, with a possibility to vary the two alkyl groups. Unex-
pectedly, hydroxymethylbenzodioxane reacted quite slowly
(Table 5, entry 6). The reason is still unclear, but the alde-
hyde does not accumulate, which may indicate that the de-
hydrogenation is the rate-determining sequence for this type
of substrate. 1,4-Benzodioxane was isolated in 79% yield
9
48
97
tries 1 and 2). The same was observed with the tosyl sub-
stituent, the reaction of which showed no sign of side prod-
uct formation (Table 4, entry 3). A chlorinated benzyl alco-
hol (Table 4, entry 4) and an aryl bromide (Table 4, entry 5)
were also converted in good yields. The ester functionality
was not affected (Table 4, entries 6 and 7), although a lower
yield of the methyl ester was observed, which is probably
due to the low boiling point of the product. A thioether
could be isolated in an acceptable yield and again the low
boiling point may account for the slightly lower yield be-
cause the alcohol was completely consumed and no side
products were observed (Table 4, entry 8). A methoxy group
in the ortho position slowed the reaction down, but full con-
version was obtained after 48 h (Table 4, entry 9).
16026
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16023 – 16029

Iridium-Catalyzed Liberation of Syngas
FULL PAPER
after 60 h and the catalyst was still active, despite the pro-
longed reaction time. Conversion of the cinnamyl alcohol
shown in Table 5, entry 7 gave the hydrogenated product in
a 10:1 ratio to the unsaturated product.
When the double bond was moved further away from the
hydroxy group, as in oleyl alcohol, the olefin was only parti-
ally hydrogenated under the reaction conditions (Scheme 1).
Oleyl alcohol was converted into heptadecane in 68% yield,
as determined by NMR, whereas the remaining substrate
was converted to a mixture of various other heptadecenes.
reaction. Because only two gaseous molecules are formed
per molecule of alcohol, there is no indication of a compet-
ing water–gas shift reaction under these conditions. This is
not surprising; only a catalytic amount of water is present in
the mixture and the water–gas shift reaction is known to be
slow at 1648C and atmospheric pressure.[26] This was con-
firmed by allowing the liberated gas from the reaction to
pass through a solution of calcium hydroxide. This experi-
ment only yielded 2% of calcium carbonate, which is within
the experimental error of zero for this setup.
It is clear that the transformation involves both a dehy-
drogenation and a decarbonylation, but it is not known
whether the two steps operate independently or in conjunc-
tion with each other. However, with the use of 5 mol% of
the iridium catalyst 2-naphthaldehyde was converted into
naphthalene in less than 5 h, which shows that the decarbon-
ylation step can be achieved in the absence of the dehydro-
genation step. When the conversion of 2-naphthylmethanol
with 5 mol% of iridium catalyst was monitored over time,
accumulation of the aldehyde, up to 30%, was observed in
the mixture (Figure 3). This illustrates that the aldehyde is
able to dissociate from iridium and suggests that the trans-
formation proceeds by two catalytic cycles.
Scheme 1. Dehydrogenative decarbonylation of oleyl alcohol.
The gas evolution was measured by reacting 1.0 mmol of
2-naphthylmethanol in a Schlenk tube connected to a bu-
rette filled with water. A total gas volume of 42.4 mL was
collected, corresponding to approximately 1.8 mmol, which
confirms that two gaseous molecules were formed in the re-
action (Figure 2). Notably, the substrate was converted with
the same reaction rate in this experiment as that observed
above in Figure 1, which shows that the continuous removal
of the liberated gas by a flow of nitrogen is not necessary.
Figure 3. Composition of the reaction mixture as a function of time.
A proposed mechanism is illustrated in Scheme 2. The
catalytically active component is believed to be the iridium
Figure 2. Gas formation as a function of time.
À
complex 1, which is expected to oxidatively add to the O H
bond in the substrate. This oxidative addition has been thor-
oughly studied with similar iridium(I) phosphine complexes
and is known to be favored by nonpolar solvents such as
benzene.[27,28] The addition is reversible and is not affected
by the addition of lithium chloride.[27] The oxidative addition
of water to iridium(I) complexes can also occur, but this re-
action does not seem to be faster than the addition of alco-
hols.[27] Furthermore, rapid exchange between alcohols and
the alkoxide ligands has been observed in these iridium
complexes[27] and a similar exchange between a hydroxide
ligand and an alcohol should be possible.
In another experiment, hydrogen gas was detected by
adding diphenylacetylene to the reaction mixture, which led
to the formation of stilbene as a mixture of the E and Z iso-
mers. Carbon monoxide was detected in a two-chamber
system that was made by combining two Schlenk tubes. One
Schlenk tube contained the reaction mixture and the other
contained [Ir
progressed in the first tube, [Ir(CO)2Cl
in the latter and was transformed into Vaskaꢁs complex,
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
[Ir(CO)ClACHTUNGTRENNUNG
(PPh3)2], under vacuum.[25] The formation of
Vaskaꢁs complex was verified by IR and 31P NMR spectro-
scopy, and confirms that carbon monoxide is liberated in the
In the next step, one of the hydrogens in the b-position is
extracted by iridium to give the iridium–dihydride complex
Chem. Eur. J. 2012, 18, 16023 – 16029
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16027

R. Madsen and E. P. K. Olsen
The decarbonylation is ex-
pected to be initiated by an ox-
idative addition to give the iri-
dium monohydride–acyl com-
plex 5, followed by a migratory
extrusion of CO to provide the
fully saturated complex
6
(Scheme 2).[6] The reductive
elimination is expected to
happen prior to the dissocia-
tion of CO, as observed for an
acyl–Cp*IrIII complex[32] and
for the rhodium(I)–dppp-cata-
lyzed decarbonylation.[18] To
Scheme 2. Proposed catalytic cycles.
investigate whether
a b-hy-
dride elimination followed by
hydrogenation of the resulting
3. This reaction can be achieved because the square pyrami-
dal complex 2 has a free cis coordination site that makes it
possible to form the octahedral complex 3. This direct b-hy-
dride elimination has also been carefully investigated with
alkene could compete with the direct reductive elimination
from complex 6, the deuterated alcohol in Scheme 3 was
subjected to the reaction conditions. In case of a b-hydride
elimination route, scrambling of deuterium and hydrogen
would occur. This, however, was not observed and under-
lines the chemoselectivity of the transformation.
analogous iridiumACHTUNGTRENNUNG(III) complexes and nonpolar solvents
were shown to benefit the reaction.[29] The mechanism of
the b-hydride elimination has also been investigated with
square planar iridium(I)–alkoxide complexes and, in this
case, the reaction was shown to be reversible.[30] Due to the
strong trans influence of the hydride ligands, complex 3 is
expected to be destabilized by the trans-dihydride configura-
tion,[31] which forces the aldehyde to dissociate. This fast dis-
sociation of the aldehyde has been observed in a number of
similar iridium–dihydride complexes.[27] As the release of hy-
drogen from iridium increases the entropy of the reaction, it
is a matter of heat to reductively eliminate hydrogen gas.
Convincing evidence about the nature of ligand L has not
yet been obtained. However, initial rate studies with [Ir-
Scheme 3. Dehydrogenative decarbonylation of a g-deuterated alcohol.
Conclusion
We have described a new iridium-catalyzed reaction for the
consecutive dehydrogenation and decarbonylation of pri-
mary alcohols. The conditions have been optimized and the
reaction has been applied to a variety of different alcohols.
A mechanism with two catalytic cycles is proposed, in which
the same catalytically active species is responsible for both
cycles. This reaction presents a new pathway for transform-
ing one of the key functional groups in organic chemistry.
ACHTUNGTRENNUNG(cod)Cl]2 and [IrACHTUNGTRENNUNG(cod)Br]2 in the absence of lithium chloride
provided very similar conversions during the first 20 min,
which suggests that the halide is dissociated from the cata-
lytically active species and therefore the ligand L is not
chloride. Interestingly, after exactly 2 turnovers a notable
decrease in the reaction rate was detected with both com-
plexes. This observation might indicate the presence of a
CO ligand on the catalytically active species. The higher
conversion that occurs over time when applying [Ir
ACHTUNGTNER(NUNG cod)Cl]2
compared to [Ir(cod)Br]2 suggests that the presence of
N
Experimental Section
chloride improves the catalyst longevity, which might be the
same effect that is observed upon the addition of lithium
chloride.
Attempts were made to measure the kinetic isotope effect
with 2-a,a-[D2]-naphthylmethanol, but with [IrACTHNUGRTNEUNG(coe)2Cl]2 as
the iridium source a rapid scrambling of hydrogen and deu-
terium occurred between the alcohol and the cyclooctene
ligand. This is the same effect as observed in the experiment
with D2O mentioned previously, and indicates that the first
two steps are reversible. In an experiment that was conduct-
ed in fully deuterated mesitylene, no scrambling between
the solvent and the alcohol or the alkene was observed.
General procedure: Alcohol (1 mmol), [IrACHTUNTGRNEUNG(coe)2Cl]2 (22.4 mg,
0.025 mmol), rac-BINAP (31.1 mg, 0.050 mmol), LiCl (4.2 mg, 0.1 mmol),
and 2.0 mL of mesitylene saturated with H2O (approx. 150 ppm) were
added to a Schlenk tube and heated to 1708C with stirring until a TLC
analysis revealed complete conversion of the alcohol and the correspond-
ing aldehyde. The crude mixture was purified directly by column chroma-
tography to provide the product.
16028
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16023 – 16029

Iridium-Catalyzed Liberation of Syngas
FULL PAPER
[12] J. H. Park, Y. Cho, Y. K. Chung, Angew. Chem. 2010, 122, 5264–
5267; Angew. Chem. Int. Ed. 2010, 49, 5138–5141.
Acknowledgements
[13] Y. Obora, Y. Anno, R. Okamoto, T. Matsu-ura, Y. Ishii, Angew.
Chem. 2011, 123, 8777–8781; Angew. Chem. Int. Ed. 2011, 50, 8618–
8622.
[14] J. G. Melnick, A. T. Radosevich, D. Villagrꢃn, D. G. Nocera, Chem.
Commun. 2010, 46, 79–81.
Financial support from The Danish Council for Strategic Research and
Nordic Energy Research (N-INNER II) is gratefully acknowledged. We
thank Samuel Gilbert Elliot for performing the experiment shown in
Scheme 3.
[15] L. S. Van Der Sluys, G. J. Kubas, K. G. Caulton, Organometallics
1991, 10, 1033–1038.
[16] a) M. C. Warner, C. P. Casey, J.-E. Bꢄckvall, Top. Organomet. Chem.
2011, 37, 85–125; b) J. S. M. Samec, J.-E. Bꢄckvall, P. G. Andersson,
P. Brandt, Chem. Soc. Rev. 2006, 35, 237–248.
[17] For recent examples, see: a) P. Fristrup, M. Tursky, R. Madsen, Org.
Biomol. Chem. 2012, 10, 2569–2577; b) H. Li, X. Wang, F. Huang,
G. Lu, J. Jiang, Z.-X. Wang, Organometallics 2011, 30, 5233–5247;
c) A. J. Johansson, E. Zuidema, C. Bolm, Chem. Eur. J. 2010, 16,
13487–13499; d) J. Bosson, A. Poater, L. Cavallo, S. P. Nolan, J.
Am. Chem. Soc. 2010, 132, 13146–13149; e) N. Sieffert, M. Bꢅhl, J.
Am. Chem. Soc. 2010, 132, 8056–8070.
[18] P. Fristrup, M. Kreis, A. Palmelund, P.-O. Norrby, R. Madsen, J. Am.
Chem. Soc. 2008, 130, 5206–5215.
[19] M.-N. Birkholz, Z. Freixa, P. W. N. M. van Leeuwen, Chem. Soc.
Rev. 2009, 38, 1099–1118.
[20] The effect of water was discovered in an experiment where a flow
of N2 was compared to a flow of dry N2 (dried by passing through a
column of KOH pellets).
[21] S. L. Webb, C. M. Giandomenico, J. Halpern, J. Am. Chem. Soc.
1986, 108, 345–347.
[1] a) O. Saidi, J. M. J. Williams, Top. Organomet. Chem. 2011, 34, 77–
106; b) C. Wang, X. Wu, J. Xiao, Chem. Asian J. 2008, 3, 1750–1770;
c) S. Gladiali, E. Alberico, Chem. Soc. Rev. 2006, 35, 226–236.
[2] a) Y. Obora, Y. Ishii, Synlett 2011, 30–51; b) G. Guillena, D. J.
Ramꢂn, M. Yus, Chem. Rev. 2010, 110, 1611–1641; c) G. E. Dober-
einer, R. H. Crabtree, Chem. Rev. 2010, 110, 681–703; d) T. D.
Nixon, M. K. Whittlesey, J. M. J. Williams, Dalton Trans. 2009, 753–
762.
[3] For recent examples, see: a) B. Gnanaprakasam, E. Balaraman, C.
Gunanathan, D. Milstein, J. Polym. Sci. Part A 2012, 50, 1755–1765;
b) A. Maggi, R. Madsen, Organometallics 2012, 31, 451–455; c) B.
Gnanaprakasam, E. Balaraman, Y. Ben-David, D. Milstein, Angew.
Chem. 2011, 123, 12448–12452; Angew. Chem. Int. Ed. 2011, 50,
12240–12244; d) C. Chen, Y. Zhang, S. H. Hong, J. Org. Chem.
2011, 76, 10005–10010; e) N. D. Schley, G. E. Dobereiner, R. H.
Crabtree, Organometallics 2011, 30, 4174–4179; f) M. A. Esteruelas,
N. Honczek, M. Olivꢃn, E. OÇate, M. Valencia, Organometallics
2011, 30, 2468–2471; g) J. H. Dam, G. Osztrovszky, L. U. Nordstrøm,
R. Madsen, Chem. Eur. J. 2010, 16, 6820–6827; h) Y. Zhang, C.
Chen, S. C. Ghosh, Y. Li, S. H. Hong, Organometallics 2010, 29,
1374–1378; i) B. Gnanaprakasam, J. Zhang, D. Milstein, Angew.
Chem. 2010, 122, 1510–1513; Angew. Chem. Int. Ed. 2010, 49, 1468–
1471.
[22] V. M. Iluc, A. Fedorov, R. H. Grubbs, Organometallics 2012, 31, 39–
41.
[23] a) Y. Xu, M. A. Celik, A. L. Thompson, H. Cai, M. Yurtsever, B.
Odell, J. C. Green, D. M. P. Mingos, J. M. Brown, Angew. Chem.
2009, 121, 590–593; Angew. Chem. Int. Ed. 2009, 48, 582–585; b) A.
Dervisi, C. Carcedo, L. Ooi, Adv. Synth. Catal. 2006, 348, 175–183;
c) S. P. Smidt, A. Pfaltz, E. Martinez-Viviente, P. S. Pregosin, A. Al-
binati, Organometallics 2003, 22, 1000–1009.
[4] T. C. Johnson, D. J. Morris, M. Wills, Chem. Soc. Rev. 2010, 39, 81–
88.
[5] For recent examples, see: a) R. Kawahara, K. Fujita, R. Yamaguchi,
J. Am. Chem. Soc. 2012, 134, 3643–3646; b) M. Nielsen, A.
Kammer, D. Cozzula, H. Junge, S. Gladiali, M. Beller, Angew.
Chem. 2011, 123, 9767–9771; Angew. Chem. Int. Ed. 2011, 50, 9593–
9597; c) S. Musa, I. Shaposhnikov, S. Cohen, D. Gelman, Angew.
Chem. 2011, 123, 3595–3599; Angew. Chem. Int. Ed. 2011, 50, 3533–
3537; d) W. Baratta, G. Bossi, E. Putignano, P. Rigo, Chem. Eur. J.
2011, 17, 3474–3481.
[24] D.-S. Wang, J. Zhou, D.-W. Wang, Y.-L. Guo, Y.-G. Zhou, Tetrahe-
dron Lett. 2010, 51, 525–528.
[25] M. J. Burk, R. H. Crabtree, Inorg. Chem. 1986, 25, 931–932.
[26] G. Jacobs, B. H. Davis, Catalysis 2007, 20, 122–285.
[27] a) O. Blum, D. Milstein, J. Am. Chem. Soc. 2002, 124, 11456–11467;
b) K. Tani, K. Nakajima, A. Iseki, T. Yamagata, Chem. Commun.
2001, 1630–1631; c) O. Blum, D. Milstein, Angew. Chem. 1995, 107,
210–212; Angew. Chem. Int. Ed. Engl. 1995, 34, 229–231.
[28] a) K. Tani, A. Iseki, T. Yamagata, Angew. Chem. 1998, 110, 3590–
3592; Angew. Chem. Int. Ed. 1998, 37, 3381–3383; b) F. T. Ladipo,
M. Kooti, J. S. Merola, Inorg. Chem. 1993, 32, 1681–1688.
[29] a) O. Blum, D. Milstein, J. Organomet. Chem. 2000, 593–594, 479–
484; b) O. Blum, D. Milstein, J. Am. Chem. Soc. 1995, 117, 4582–
4594.
[30] a) S. A. Macgregor, P. Vadivelu, Organometallics 2007, 26, 3651–
3659; b) J. Zhao, H. Hesslink, J. F. Hartwig, J. Am. Chem. Soc. 2001,
123, 7220–7227.
[31] B. Rybtchinski, Y. Ben-David, D. Milstein, Organometallics 1997,
16, 3786–3793.
[6] M. A. Garralda, Dalton Trans. 2009, 3635–3645.
[7] a) M. Kreis, A. Palmelund, L. Bunch, R. Madsen, Adv. Synth. Catal.
2006, 348, 2148–2154; b) C. M. Beck, S. E. Rathmill, Y. J. Park, J.
Chen, R. H. Crabtree, L. M. Liable-Sands, A. L. Rheingold, Organo-
metallics 1999, 18, 5311–5317; c) D. H. Doughty, L. H. Pignolet, J.
Am. Chem. Soc. 1978, 100, 7083–7085.
[8] a) J. J. Adams, N. Arulsamy, D. M. Roddick, Organometallics 2012,
31, 1439–1447; b) T. Iwai, T. Fujihara, Y. Tsuji, Chem. Commun.
2008, 6215–6217.
[9] A. Modak, A. Deb, T. Patra, S. Rana, S. Maity, D. Maiti, Chem.
Commun. 2012, 48, 4253–4255.
[10] R. N. Monrad, R. Madsen, J. Org. Chem. 2007, 72, 9782–9785.
[11] a) K. Ikeda, T. Morimoto, K. Kakiuchi, J. Org. Chem. 2010, 75,
6279–6282; b) T. Shibata, N. Toshida, M. Yamasaki, S. Maekawa, K.
Takagi, Tetrahedron 2005, 61, 9974–9979; c) F. Y. Kwong, Y. M. Li,
W. H. Lam, L. Qiu, H. W. Lee, C. H. Yeung, K. S. Chan, A. S. C.
Chan, Chem. Eur. J. 2005, 11, 3872–3880; d) T. Shibata, N. Toshida,
K. Takagi, J. Org. Chem. 2002, 67, 7446–7450; e) T. Morimoto, K.
Fuji, K. Tsutsumi, K. Kakiuchi, J. Am. Chem. Soc. 2002, 124, 3806–
3807.
[32] J. G. Cordaro, R. G. Bergman, J. Am. Chem. Soc. 2004, 126, 3432–
3433.
Received: July 24, 2012
Published online: October 29, 2012
Chem. Eur. J. 2012, 18, 16023 – 16029
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16029