Silacarboxylic acids as efficient carbon monoxide releasing molecules: Synthesis and application in palladium-catalyzed carbonylation reactions

Article abstract of DOI:10.1021/ja208652n

Silacarboxylic acids have been demonstrated to be easy to handle, air-stable carbon monoxide precursors. Different silacarboxylic acids were synthesized from the corresponding chlorosilanes and carbon dioxide, and their decarbonylation, upon treatment with an array of activators, was evaluated. The release of CO from crystalline MePh₂SiCO₂H proved to be highly efficient, and it was successfully applied in a selection of palladium-catalyzed carbonylative couplings using near-stoichiometric quantities of carbon monoxide precursor. Finally, the synthesis of MePh₂Si ¹³CO₂H and its application in carbonyl labeling of two bioactive compounds was demonstrated.

Full text of DOI:10.1021/ja208652n

COMMUNICATION
pubs.acs.org/JACS
Silacarboxylic Acids as Efficient Carbon Monoxide Releasing
Molecules: Synthesis and Application in Palladium-Catalyzed
Carbonylation Reactions
Stig D. Friis, Rolf H. Taaning, Anders T. Lindhardt, and Troels Skrydstrup*
Center for Insoluble Protein Structures (inSPIN), Department of Chemistry, and the Interdisciplinary Nanoscience Center (iNANO),
Aarhus University, Langelandsgade 140, 8000 Aarhus C, Denmark
S Supporting Information
b
Scheme 1. General CO Precursor Motif
ABSTRACT: Silacarboxylic acids have been demonstrated
to be easy to handle, air-stable carbon monoxide precursors.
Different silacarboxylic acids were synthesized from the
corresponding chlorosilanes and carbon dioxide, and their
decarbonylation, upon treatment with an array of activators,
was evaluated. The release of CO from crystalline MePh2-
SiCO2H proved to be highly efficient, and it was successfully
applied in a selection of palladium-catalyzed carbonylative
couplings using near-stoichiometric quantities of carbon
monoxide precursor. Finally, the synthesis of MePh2Si-
13CO2H and its application in carbonyl labeling of two
bioactive compounds was demonstrated.
alternative crystalline CO-releasing molecules (CORMs) that
can readily be prepared and activated to release CO upon treatment
with a simple non-transition-metal activator. In this communica-
tion, we demonstrate the possibility of using silacarboxylic acids
as easy-to-handle substrates for this purpose.
Considering earlier work on boranocarbonates as CO-releas-
ing agents, we speculated whether silacarboxylic acids and deriva-
tives thereof could also represent interesting CORMs, as they
would potentially be easier to synthesize and handle than the
boron-containing analogues (Scheme 1).10 The limited work
previously published demonstrated their CO releasing proper-
ties, but only after they were heated at high temperature
(>125 °C) or in a nonquantitative manner under basic condi-
tions.11À13 On the other hand, with the inherent high fluorophi-
licity of the silicon atom, we were compelled to examine whether
decarbonylation could possibly be achieved at ambient tempera-
ture with a fluoride source. To this end, several silacarboxylic
acids were prepared and their CO-releasing abilities examined.
The silacarboxylic acids were synthesized from their corre-
sponding chlorosilanes via initial reduction using metallic lithium
and subsequent reaction with carbon dioxide. With this metho-
dology, three silacarboxylic acids were synthesized in good yields
(Table 1). Reduction of the equivalent hydridosilanes with metallic
lithium or lithium naphthalide was also examined, but this re-
sulted in less clean conversions and poorer yields.14,15
ransition-metal-catalyzed carbonylations represents an im-
T
portant class of reactions for the effective preparation of an
array of carbonyl-containing compounds. Characteristic of many
of these reactions are the mild reaction conditions and the high
functional group tolerance. This allows the introduction of a
carbonyl group in an advanced intermediate, and hence, much
effort has been initiated to improve existing reactions and identify
new carbonylation transformations.1À3
There is nevertheless a major drawback in the use of transi-
tion-metal-catalyzed carbonylations that is related to the nature
of the reagents chosen. In most cases, the highly toxic, tasteless,
odorless, and colorless gas carbon monoxide is required, and it is
often exploited in large excess and in many instances at pressures
greater than atmospheric. The need for specially facilitated
laboratories equipped with CO detectors and possibly high-
pressure reactors limits the use of this small diatomic reagent.
Various solids and liquids that are capable of releasing CO upon
external stimulation have therefore been developed as alterna-
tives. One example, Mo(CO)6, has been studied and used ex-
tensively by Larhed and others, and despite its toxicity, it has
proven its worth as a CO surrogate. However, general and mild
conditions initiating the decarbonylation steps from these CO
precursors are still needed, and most are not feasible for isotopic
labeling.4À7
Of the three silacarboxylic acids synthesized, methyldiphenyl-
silacarboxylic acid (2) and tert-butyldiphenylsilacarboxylic acid
(3) proved to be highly crystalline (recrystallization from n-heptane),
hence facilitating gram-scale synthesis. Attempts to access the
triphenylsilacarboxylic acid were also made, resulting in a highly
insoluble colorless solid that in the ensuing studies did not
release CO upon treatment with a fluoride source.
We recently reported the use of a two-chamber system that
allows several different Pd-catalyzed carbonylation transforma-
tions to be conducted in a simple manner.8,9 The method applies
stoichiometric or substoichiometric amounts of a crystalline acid
chloride that liberates CO via a Pd-catalyzed decarbonylation
process. To simplify this procedure, we have been investigating
The decarbonylation of the silacarboxylic acids was evaluated
as shown in Table 2. The three precursors were first subjected to
Received: September 14, 2011
Published: October 20, 2011
r
2011 American Chemical Society
18114
dx.doi.org/10.1021/ja208652n J. Am. Chem. Soc. 2011, 133, 18114–18117
|

Journal of the American Chemical Society
COMMUNICATION
Table 1. Synthesis of Silacarboxylic Acidsa
Table 3. Decarbonylation of MePh2SiCO2H (2) Employing
Different Activators and Solvents
entry solventa
activator
KF
temp [°C] time [min] gas yield [%]
1
2
DMF
rt
1
5
94
96
THF
KF
KF
40
40
70
70
rt
a Lithium pieces and chlorosilane were stirred in THF for 5 h before the
solution was transferred to another flask via cannula, reacted with CO2 at
À78 °C for 20 min, and then allowed to warm to rt.
3
dioxane
10
20
1200
1
95
4b
PEG5000 KF
94
5
water
KF
41
6c
7
dioxane
dioxane
dioxane
dioxane
dioxane
TBAF
CsF
100
94
rt
7
Table 2. Decarbonylation of Silacarboxylic Acids Using
Potassium Fluoride
8
MeOK
t-BuOK
MePh2SiOK
40
40
40
7
97
9
5
86
10
10
99
a Concentration 0.25 M. b 0.50 mmol in 1.0 g poly(ethylene glycol)
methyl ether. c 1 M in THF, concentration 0.20 M.
Table 4. Decarbonylation of MePh2SiCO2H (2) Using Sub-
stoichiometric Potassium Fluoridea
entry
equiv of KF
time [h]
gas yield [%]
1
2
3b
0.5
0.33
0.5
2
94
94
92
78
0.25
0.1
4
0.05
48
a Concentration 0.25 M. b 95% was isolated from an aminocarbonylation
producing 4 using 0.67 equiv of MePh2SiCO2H and 0.07 equiv of KF.
The results of an evaluation of different solvents and activators
for the decarbonylation of 2 are revealed in Table 3. After ad-
dition of a slight excess of the given activator and then the solvent,
the gas evolution was monitored over time until completion.16
The decarbonylation of 2 proceeded smoothly in a range of
solvents with KF (entries 1À4), with only water giving poor gas
evolution (entry 5), which may be explained by the hydration of
the fluoride anion and hence modulation of its nucleophilicity. A
number of other fluoride activators (entries 6 and 7) and oxygen
nucleophiles/bases were also evaluated (entries 8À10).17 The
bulky oxygen-based activators possibly promote decarbonylation
through a 1,2-Brook rearrangement rather than nucleophilic
addition to silicon by operating as a base with formation of the
potassium carboxylate.18 Whether this is also the operating mechan-
ism in the fluoride-promoted decarbonylation is a matter of ongoing
investigations in our group.
a Chamber A: silacarboxylic acid (0.33 mmol) and KF (0.36 mmol) in
dioxane (3 mL). Chamber B: 4-iodoanisole (0.5 mmol), hexylamine
(1.0 mmol), TEA (1.0 mmol), Pd(dba)2 (5 mol %), and PPh3 (10 mol %)
in dioxane (3 mL). Both chambers were heated to 80 °C for 16 h
b Isolated by flash column chromatography. c At 60 °C
KF in dioxane, and the gas evolution was monitored over time.16
To determine the nature and equivalents of the gas released, a
substoichiometric amount of each silacarboxylic acid was applied
in a Pd-catalyzed aminocarbonylation between p-iodoanisole and
n-hexylamine in a setup involving the two-chamber system as
described previously.8 Both 2 and 3 exhibited near-quantitative
CO release at ambient temperature, and excellent yields were ob-
tained for the synthesis of amide 4(entries 2 and 3). Nevertheless, as
silacarboxylic acid 2 was easier to purify by recrystallization and
the decarbonylation proved to be the fastest, it was chosen for the
subsequent studies.
The limited solubility of KF in dioxane is in sharp contrast to
the fast CO release observed. Additionally, the hydroxide ion
leaving the silacarboxylic acid after decarbonylation should be an
activator equivalent to potassium methoxide. These considerations
indicate that only a substoichiometric quantity of the activator is
required. Excellent gas yields were observed even with a loading
18115
dx.doi.org/10.1021/ja208652n |J. Am. Chem. Soc. 2011, 133, 18114–18117

Journal of the American Chemical Society
COMMUNICATION
Scheme 2. Carbonylative Couplings Using MePh2SiCO2H as the CO Sourcea
a See Supporting Information for exact reaction conditions.
of 0.1 equiv of KF with only a moderate increase in reaction time
(Table 4). This slower CO release could be desirable in a number
of cases, as shown by Leadbeater and others.19À21
Six different Pd-catalyzed carbonylative couplings were per-
formed with silacarboxylic acid 2 in a two-chamber system
(Scheme 2). All of the products were synthesized by applying
only a slight excess of the CO precursor at ambient pressure. In
general, the CO precursor 2 was dissolved in the same solvent
used for the carbonylative coupling, and 1.1 equiv of KF was used
as the activator.
In the first example, amide 5 was synthesized using the same
conditions as for 4 (Table 2) with a slightly lower temperature,
which provided the amide product in 82% yield. Alkoxycarbo-
nylation of 8-tosylquinoline with cyclopropylmethanol afforded
the desired ester 6 in 75% isolated yield using slightly modified
literature conditions, with only 1.5 equiv of 2 in contrast to the
large excess of CO supplied by a balloon.22
Double carbonylations of aryl iodides can be conducted at
room temperature, as shown by Kondo.23 Only a few CO pre-
cursors, however, have the ability to release quantitative amounts
of CO quickly at this temperature.4À7 As illustrated in the
synthesis of compound 7, silacarboxylic acid 2 proved to be very
effective for these types of transformations. When a 3:1 ratio of
2 to aryl iodide was employed, the desired product 7 was isolated
in 91% yield after a reaction time of only 2 h.
To reduce the speed of decarbonylation, 0.1 equiv of KF was used
for this reaction, resulting in the formation of 8 in 77% yield.
Subjecting the same aryl iodide to slightly modified carbonylative
Sonogashira reaction conditions with 1-hexyne provided product
9 in 82% isolated yield employing water as solvent.26 CO release
from 2 is, however, very slow in water, and the precursor was
therefore dissolved in dioxane using CsF as the activator to ensure
prompt decarbonylation.
As a last example, β-lactam 10 was synthesized according to a
literature procedure, thereby displaying the mildness and versa-
tility of the carbonylative conditions using silacarboxylic acids. 10
was isolated in a satisfactory 62% yield.27
Finally, we examined the potential of this methodology for 13
C
labeling experiments (Scheme 3). The reaction of a slight excess
of methyldiphenylsilyllithium with 13CO2 furnished [13C]methyl-
diphenylsilacarboxylic acid (11) in an excellent yield of 99% after
recrystallization. Next, 11 was used to provide a substoichio-
metric quantity of 13CO in the synthesis of piclamilast (12).28
This potent cyclic AMP phosphodiesterase inhibitor was isolated
in 74% yield. Compound 13 belongs to a group of HIV-1
inhibitors targeting the glycoprotein gp120 situated in the viral
envelope.29 It was synthesized in a doubly labeled version using a
minor excess of 13CO generated from 11 at room temperature
and then isolated in 68% yield after deprotection.
In summary, silacarboxylic acids have been presented as versatile,
benchtop-stable, easy-to-handle, metal-free CO-releasing mole-
cules.30 The two-step synthesis of three silacarboxylic acids was
demonstrated, and conditions for near-quantitative CO release
Our group recently published carbonylative Heck reaction
conditions using near- and substoichiometric CO generated from
a precursor that generally exhibits slower CO release than 2.24,25
18116
dx.doi.org/10.1021/ja208652n |J. Am. Chem. Soc. 2011, 133, 18114–18117

Journal of the American Chemical Society
COMMUNICATION
Scheme 3. Preparation and Utilization of 13C-Labeled Sila-
carboxylic Acida
(7) Morimoto, T.; Fuji, K.; Tsutsumi, K.; Kakiuchi, K. J. Am. Chem.
Soc. 2002, 124, 3806.
(8) Hermange, P.; Lindhardt, A. T.; Taaning, R. H.; Bjerglund, K.;
Lupp, D.; Skrydstrup, T. J. Am. Chem. Soc. 2011, 133, 6061.
(9) Patent application currently proceeding: Taaning, R. H.; Hermange,
P.; Lindhardt, A. T.; Friis, S. D.; Skrydstrup, T. (Aarhus University,
Denmark). System Providing Controlled Delivery of Gaseous Co for
Carbonylation Reactions. PA 2010 70546, Dec 15, 2010.
(10) Alberto, R.; Ortner, K.; Wheatley, N.; Schibli, R.; Schubiger,
A. P. J. Am. Chem. Soc. 2001, 123, 3135.
(11) Brook, A. G. J. Am. Chem. Soc. 1955, 77, 4827.
(12) Brook, A. G.; Gilman, H. J. Am. Chem. Soc. 1955, 77, 2322.
(13) Gilman, H.; Trepka, W. J. J. Org. Chem. 1960, 25, 2201.
(14) Hernꢀandez, D.; Mose, R.; Skrydstrup, T. Org. Lett. 2011, 13, 732.
(15) Nielsen, L.; Skrydstrup, T. J. Am. Chem. Soc. 2008, 130, 13145.
(16) See the Supporting Information for details concerning gas
measurements, calculation of gas yields, and graphs and raw data for
gas release over time.
(17) The results support earlier speculations by Igawa et al. that
activated silacarboxylic acid derivatives decarbonylate upon treatment
with a nucleophile. See: Igawa, K.; Kokan, N.; Tomooka, K. Angew.
Chem., Int. Ed. 2010, 49, 728.
(18) Brook, A. G. Acc. Chem. Res. 1974, 7, 77.
(19) Kormos, C. M.; Leadbeater, N. E. Synlett 2007, 2006.
(20) Gavi~no, R.; Pellegrini, S.; Castanet, Y.; Mortreux, A.; Mentrꢀe,
O. Appl. Catal., A 2001, 217, 91.
a See the Supporting Information for the exact reaction conditions.
(21) A high partial pressure of CO can cause precipitation of various
palladium carbonyl species, bringing the catalytic cycle to a halt.
(22) Munday, R. H.; Martinelli, J. R.; Buchwald, S. L. J. Am. Chem.
Soc. 2008, 130, 2754.
(23) Iizuka, M.; Kondo, Y. Chem. Commun. 2006, 1739.
(24) Hermange, P.; Gøgsig, T. M.; Lindhardt, A. T.; Taaning, R. H.;
Skrydstrup, T. Org. Lett. 2011, 13, 2444.
(25) Using identical conditions but with 1.1 equiv of KF afforded an
isolated yield of only 51%.
(26) Liang, B.; Huang, M.; You, Z.; Xiong, Z.; Lu, K.; Fathi, R.; Chen,
J.; Yang, Z. J. Org. Chem. 2005, 70, 6097.
from these derivatives were identified. Lastly, near-stoichiometric
quantities of MePh2SiCO2H were utilized in a number of Pd-
catalyzed carbonylative couplings that resulted in good to
excellent yields. An application in 13C-isotope labeling was also
successfully demonstrated via the synthesis of two labeled
bioactive compounds. Efforts to examine the application of these
CO-releasing molecules for other carbon-isotope-labeling stud-
ies and the challenge of in situ CO generation are now underway.
(27) Zhang, Z.; Liu, Y.; Ling, L.; Li, Y.; Dong, Y.; Gong, M.; Zhao, X.;
Zhang, Y.; Wang, J. J. Am. Chem. Soc. 2011, 133, 4330.
(28) Ashton, M. J.;Cook, D. C.; Fenton, G.; Karlsson, J.-A.; Palfreyman,
M. N.; Raeburn, D.; Ratcliffe, A. J.; Souness, J. E.; Thurairatnam, S.; Vicker,
N. J. Med. Chem. 1994, 37, 1696.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures and
b
spectroscopic data. This material is available free of charge via the
Internet at http://pubs.acs.org.
(29) Wang, J.; Le, N.; Heredia, A.; Song, H.; Redfield, R.; Wang,
L.-X. Org. Biomol. Chem. 2005, 3, 1781.
’ AUTHOR INFORMATION
(30) Benchtop storage of 2 over a 2 month period did not give rise to
any change in appearance or the 1H NMR spectrum.
Corresponding Author
ts@chem.au.dk
’ ACKNOWLEDGMENT
We are deeply appreciative of generous financial support from
the Danish National Research Foundation, the Danish Natural
Science Research Council, the Carlsberg Foundation, the Lund-
beck Foundation, and Aarhus University.
’ REFERENCES
(1) Brennf€uhrer, A.; Neumann, H.; Beller, M. Angew. Chem., Int. Ed.
2009, 48, 4114.
(2) Martinelli, J. R.; Clark, T. P.; Watson, D. A.; Munday, R. H.;
Buchwald, S. L. Angew. Chem., Int. Ed. 2007, 46, 8460.
(3) Zeng, F.; Alper, H. Org. Lett. 2011, 13, 2868.
(4) Wie-ckowska, A.; Fransson, R.; Odell, L. R.; Larhed, M. J. Org.
Chem. 2011, 76, 978.
(5) Wannberg, J.; Larhed, M. In Modern Carbonylation Methods;
Kollꢀar, L., Ed.; Wiley-VCH: Weinheim, Germany, 2008; pp 93À114.
(6) Elmore, C. S. J. Labelled Compd. Radiopharm. 2011, 54, 59.
18117
dx.doi.org/10.1021/ja208652n |J. Am. Chem. Soc. 2011, 133, 18114–18117