14Carbon monoxide made simple - Novel approach to the generation, utilization, and scrubbing of 14carbon monoxide

Article abstract of DOI:10.1002/jlcr.2962

A new method is reported for the efficient generation of ¹⁴CO that can be applied in transition metal-catalyzed carbonylation reactions. ¹⁴CO is produced by palladium-catalyzed decarbonylation of the stable acid chloride, ¹⁴COgen. When combined in a two-chamber system, the produced ¹⁴CO is incorporated into the target molecule with high efficiency. As the carbonylation chemistry is performed under mild conditions, this allows the ¹⁴C isotope to enter the synthesis into an advanced stage intermediate. The presented work includes two aminocarbonylations, one amidocarbonylation, and one carbonylative Suzuki-Miyaura coupling, all installing the ¹⁴C isotope in the final step of the synthesis. Finally, the identification of a highly efficient scrubber for the safe trapping and removal of any leftover ¹⁴CO is also disclosed. A new method for the simple generation of ¹⁴CO from a stable, easy to handle CO precursor ¹⁴COgen is presented. When combined with a two-chamber system, COware, the generated ¹⁴CO is successfully applied as the limiting reagent in any carbonylation reaction. Owing to the mild conditions, the desired labeled compounds could be obtained in the final synthetic step of a linear synthesis. Finally, the development of a highly efficient scrubber, capable of removing any remaining or excess ¹⁴CO, is presented. This method provides radiochemists with a complete solution for performing ¹⁴C-carbonylation chemistry. Copyright

Full text of DOI:10.1002/jlcr.2962

Research Article
Received 7 June 2012,
Revised 9 July 2012,
Accepted 23 July 2012
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.2962
14Carbon monoxide made simple – novel
approach to the generation, utilization,
14
and scrubbing of carbon monoxide†
a
b
a
*
*
Anders T. Lindhardt, Roger Simonssen, Rolf H. Taaning,
Thomas M. Gøgsig,a Göran N. Nilsson,b Gunnar Stenhagen,b
Charles S. Elmore, and Troels Skrydstrup
b
a
*
A new method is reported for the efficient generation of 14CO that can be applied in transition metal-catalyzed carbonylation
reactions. 14CO is produced by palladium-catalyzed decarbonylation of the stable acid chloride, 14COgen. When combined in
a two-chamber system, the produced 14CO is incorporated into the target molecule with high efficiency. As the carbonylation
chemistry is performed under mild conditions, this allows the 14C isotope to enter the synthesis into an advanced stage
intermediate. The presented work includes two aminocarbonylations, one amidocarbonylation, and one carbonylative
Suzuki–Miyaura coupling, all installing the 14C isotope in the final step of the synthesis. Finally, the identification of a highly effi-
cient scrubber for the safe trapping and removal of any leftover 14CO is also disclosed.
Keywords: 14carbon monoxide; isotope labeling; carbonylation; two chamber; scrubber
produced radiochemical waste. With respect to 14C-labeled
Introduction
drug metabolism studies, the installed 14C-carbonyl group is
Since the discovery of palladium-catalyzed carbonylation reac-
typically bound to the parent structure by a carbon–carbon
tions with carbon monoxide (CO) in the 1970s, such reactions
bond. This is advantageous because the metabolic stability of a
have proved highly useful in fine chemical synthesis. In particu-
carbon–carbon bond is considerably higher than that of
lar, the last 10 years have resulted in the development of a wide
carbon–heteroatom bonds.
variety of these transformations.1 The classical carbonylation
reaction applies aryl or alkenyl halides or pseudohalides with
either heteroatom or carbon nucleophiles, and examples on
amino-1,2 and alkoxycarbonylations1,3 as well as carbonylative
versions of the Suzuki–Miyaura,1,4 Mizoroki–Heck,1,5 Sonogashira
couplings1,6, and so on are found in the literature. All of these
reactions add a carbonyl group to the structure, which repre-
sents a highly common motif in biologically active structures
and a viable structural platform for further functional group
transformations. Potential products obtained directly from
this carbonylation chemistry include amides, esters, ketones,
aldehydes, and so on, all structural motifs of high importance
in biology and synthesis.1c,7
There are a few 14C-labeling methods and reagents available
as alternatives to carbonylation chemistry. These include carbox-
ylation, formylation, or cyanation reactions. All of them are
well-established methods but still hold significant drawbacks
when compared with the Pd-catalyzed carbonylation chemistry.
Carboxylation with 14CO2 or formylation using 14C-DMF requires
a metalated species as the nucleophile, typically a lithium or
Grignard reagent, which induces significant restraints to the
functional group compatibility.8,9 The transition metal-catalyzed
14C cyanations typically occur under mild conditions with excel-
lent functional group tolerance. However, the installed cyanide
would have to be hydrolyzed or reduced to obtain the desired
CO is a highly versatile reagent, and upon combination with
transition metal catalysis, the overall functional group tolerance of
the reaction becomes excellent. This allows CO to be installed onto
the growing molecule at almost any step in the linear sequence.
Finally, CO can be applied to connect two advanced intermediates
making the overall synthetic strategy highly convergent.
aCenter for Insoluble Protein Structures (inSPIN), Department of Chemistry and
the Interdisciplinary Nanoscience Center (iNANO), Aarhus University, Langelands-
gade 140, 8000 Aarhus C, Denmark
bGlobal DMPK, AstraZeneca R&D Mölndal, SE-431 83 Mölndal, Sweden
With these arguments in hand, it becomes obvious that
transition metal-catalyzed carbonylation chemistry, using 13CO
or 14CO, is a nearly ideal method for carbon isotope labeling.
Installing the isotope at a late stage increases the overall
efficiency of the carbon isotope incorporation. Furthermore,
when working in the field of 14C-labeling, elimination of linear
*Correspondences to: Anders T. Lindhardt, Rolf H. Taaning 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.
E-mail: lindhardt1@gmail.com (A.T. Lindhardt); rtaaning@gmail.com (RF Taaning);
ts@chem.au.dk (T. Skrydstrup)
steps in a synthesis significantly reduces the amount of †Supporting information may be found in the online version of this article.
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.

A. T. Lindhardt et al.
carboxylic acid, amide, or aldehyde, once again, requiring harsh concentrated sulfuric acid.14 Combining sodium [14C]-formate
conditions.8a,10 Because of the reduced functional group tolerance, with sulfuric acid in a round bottom flask and heating it to
all these methods require that the isotope is installed early in the 100 ꢀC for 2 h will result in the formation of 14CO. The formed
linear synthesis resulting in low isotope efficiency. Finally, additional 14CO can then be applied in a carbonylation reaction. However,
transformations are required after the installation of the isotope to to reduce the overall headspace in the reaction setup, relatively
obtain the desired structural motifs adding time-consuming and large amounts of concentrated sulfuric acid are required. Despite
waste-producing steps to the synthesis. By applying a carbonyla- the cumbersome setup, this method has actually performed well
tive approach, these issues are addressed directly and the target in several carbonylation reactions.8a,11,15 Nevertheless, there are
functionality installed in one step under mild conditions.
several major drawbacks to this method to be considered prior
The use of 14CO does, however, not come without drawbacks. to using this approach. Safety is not to be ignored when handling
The venting of radioactive gasses such as 14CO2 or 14CO to the radioactively contaminated concentrated sulfuric acid at 100ꢀC.
surroundings may be subject to severe regulations. Whereas Upon completion, one is left with large amounts of concentrated
working with nonvolatile compounds is a straightforward task, acid that can be costly to dispose of. Finally, an excess of 14CO
application of a radioactive gas as a reagent becomes somewhat (two to three equivalents) has typically been applied to obtain a
more troublesome. The chemist would have to ensure that proper yield in the carbonylation reaction.
nearly the entire gaseous component is consumed in the chemi-
During work in our group related to the safe and efficient
cal transformation. However, without the ability to test the handling of CO, a new two-chamber method for carbonylation
progress of the reaction by methods such as thin layer chroma- chemistry was developed.16 This method is based on the appli-
tography, NMR, and so on, this is not trivial. Furthermore, what cation of solid CO precursors capable of releasing CO in a highly
if the chemical reaction fails completely, and none of the radioac- controlled manner through a Pd-catalyzed decarbonylation
tive gas is consumed? A method is therefore required to trap the reaction. It is important to note that the CO precursors will not
excess radioactivity. This could typically be achieved by conden- release CO by themselves but only when subjected to the correct
sing the gas through cooling to its boiling point and transferring catalytic conditions. To ensure 100% chemical compatibility,
it to a new batch of reagents for consumption. This approach is CO is produced externally (ex situ), not in situ, in a secondary
viable for gasses with relatively high boiling or melting points such chamber. The produced CO then diffuses to the carbonylation
as carbon dioxide (mp. À78 ꢀC), which is condensed with liquid chamber where it is incorporated into the desired structure. So
nitrogen. On the other hand, CO has a boiling point of À191.5 ꢀC, far, CO release has been found to occur in all aprotic solvents,
which makes its complete condensation practically impossible being independent of the choice of the palladium-precursor
with only a mere 4.5ꢀC boiling point difference to liquid nitrogen. and base, but requires P(tBu)3 as the ligand or its air-stable
Hence, to have full access to all of the aforementioned advantages precursor HBF4P(tBu)3. During this work, we became interested
of applying 14CO as a reagent, this would require the development in applying the system for carbon isotope labeling, and to test
of a scrubber, that upon completion of a chemical reaction, could this hypothesis, a simple aminocarbonylation was performed
be activated or connected to the system in a simple manner and applying CO as the limiting reagent (Scheme 1).
that would effectively consume all leftover 14CO.
As shown in Scheme 1, the acid chloride CO precursor
The first report on 14CO dates back to 1947 and since then, the (COgen or 13COgen) is catalytically decarbonylated, releasing
gas has been utilized in numerous occasions.11 However, the CO and 9-methylene fluorene.17 Only the CO will diffuse to the sec-
application of the gas in synthetic purpose is relatively scarce ond chamber where the palladium-catalyzed aminocarbonylation
when compared with its 12CO and 13CO counterparts. This is of iodoanisole takes place. Applying a substoichiometric amount
mainly due to the following reasons: 14CO is not stable in its of CO, this experiment resulted in an excellent 96% isolated
gaseous form leading to significant radiolysis.12 Furthermore, yield of the formed amide (1), a yield based on COgen. Next, a
the majority of published carbonylation chemistry is performed simple change in the CO precursor from COgen to 13COgen
using CO pressures ranging from 5–100 bars, which obviously resulted in the formation of the 13C-labeled version of 1 in the
are not compatible when working with 14CO.1 The literature does same high 96% isolated yield. Notably, the aforementioned
hold a few methods for the generation of 14CO on relatively aminocarbonylation occurs under low-pressure conditions, as
small scales. Apart from the zinc-bed reductions of 14CO2 or the partial pressure of CO does not exceed 1 atm. during the reac-
Ca14CO3 requiring temperatures of 400 or 700 ꢀC, respectively, tion.18 This patented two-chamber method has since been applied
in combination with the need for mercury-based Toepler to numerous carbonylation reactions including carbonylative
vacuum pump, only one practical method has been applied.13 Mizoroki–Heck,19 formation of primary amides20 and tert-butyl
This method takes advantage of the dehydrative nature of esters,21 N-acylation of ureas22, and in carbonylative synthesis
Scheme 1. Aminocarbonylation applying CO as the limiting reagent.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

A. T. Lindhardt et al.
of 1,3-diketones.23 Furthermore, in all of the mentioned reac- of 14COgen. An ester-free batch of 14COgen was synthesized
tions, the method proved very effective for 13C-isotope labeling for the remainder of the experiments. 14COgen was stored as a
using 13COgen.24
In this paper, we wish to report on the application of the two-
215 MBq/mL stock solution in toluene for a stability test.
To benchmark the reactivity of the system, the aminocarbony-
chamber system in combination with 14COgen for 14C-labeling lation illustrated in Scheme 1 was first attempted. Once again,
using the generated 14CO. This patented method represents a 14CO was applied as the limiting reagent, and a 0.1 mmol scale
novel technique for the production and efficient application of (214 MBq) was chosen to test the strength of the technique. A
14CO in carbonylation chemistry. The method is based on a catalytic system based on a palladium(II) source was applied
simple piece of glassware, namely COware, allowing access to for the 14CO release. Both Pd(0) and Pd(II) sources are useful as
perform simple labeling with 14CO. The aim of this work is to catalysts, but CO release will commence relatively fast (30–60 s)
introduce the 14C-label in the final step attaining the target struc- even at room temperature using a Pd(0) precursor such as Pd
ture in only one radiochemical transformation. Furthermore, 14CO (dba)2. The activation of PdCl2(COD) has been shown to be con-
is applied as the limiting reagent in all couplings securing siderably slower, few minutes at elevated temperatures, and for
maximum utility of the carbon isotope. Finally, we also present this reason, this palladium precursor was chosen in the following
work towards the identification of a suitable scrubber, capable of experiments. Running the reaction in the two-chamber glassware
covalently and hence irreversibly trapping any unreacted 14CO.
apparatus (COware) using dioxane in the aminocarbonylation
chamber and 1 mL of the toluene stock solution in the 14CO-
releasing chamber at 80 ꢀC overnight afforded 14C-1 in a good
72% radiochemical yield (155 MBq – 99.8% radiochemical purity)
Results and discussion
The technology behind the CO precursors such as COgen and after column chromatography (Scheme 3). Interestingly, analysis
13COgen is based on solid and bench stable acid chloride pre- of the CO-releasing chamber revealed that a roughly 10% of the
cursors, which by weighing out the correct amount directly activity was retained. This was measured by diluting the crude
allows the chemist to safely control how much CO is ultimately reaction mixture of the CO-releasing chamber and assaying by
produced. The carbon source for the CO precursor originates scintillation counting. From the literature, it is known that Pd will
from CO2 via trapping of the lithiated 9-methyl fluorene forming form Pd–CO clusters, at high partial pressures of CO, which is
the carboxylic acid (Scheme 2). This acid is then transformed into inactive in catalysis.25 However, such complexes would release
the COgen by treatment with oxalyl chloride and catalytic CO upon dilution, which was not the case here even after being
DMF.16 Because all the 14C is derived from 14CO2 (or Ba14CO3), left overnight. Hence, the 14C activity must be covalently bound
the same approach seemed feasible allowing a high efficiency in the chamber. This effect was observed in all the 14C reactions
for obtaining the 14C-labeled compound. Performing the performed, and none of them showed signs of releasing gaseous
reaction with 14CO2 (2.15 GBq, 1 mmol), applying an excess of or other volatile forms of 14C products.
lithiated 9-methyl fluorene in THF, generated the desired
Next, we turned our attention towards the synthesis of a more
9-methylfluoren-9-[14C]-carboxylic acid in an excellent 94% complex structure in olaparib (2) that is a PARP-inhibitor with
radiochemical yield (99% radiochemical purity, >95% pure by potential applications for the treatment of ovarian cancer. It
1H-NMR analysis) after workup. 14COgen was obtained after was envisioned that by applying a CO strategy, this molecule
oxalyl chloride/DMF treatment in toluene in a 94% radiochemical could be assembled in one final step. To this end, the aryl
yield. A slightly lower radiochemical purity was obtained than bromide 3 and the amine 4 fragments were synthesized. After
was expected because of methanol contamination in the rotary a small screening of the reaction conditions, it was found that
evaporator leading to methyl ester formation. This 14COgen a catalytic system composed of Pd(dba)2 in combination with
batch was nevertheless used without further purification in one the bidentate ligand Xantphos provided olaparib in a nearly
of the carbonylation reactions, as the presence of the inactive quantitative yield when performing the reaction with a 0.5 mmol
methyl ester was not expected influence the releasing abilities loading of CO from COgen (Table 1, entry 1). Although a drop in
Scheme 2. Synthesis of COgen, 13COgen, and 14COgen derivatives.
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

A. T. Lindhardt et al.
Scheme 3. Aminocarbonylation using 14COgen as the 14CO precursor.
amount of radioactive material be required, the yield improves
significantly as seen in Table 1, implying that the carbonylative
approach is viable on all scales.
Table 1. Synthesis of olaparib at different CO loadings
Having completed the studies with olaparib, the carbonylative
amidation forming thalidomide (5) was undertaken starting with
the aryl bromide 6. Again, simply by screening the reaction
conditions using COgen or 13COgen, viable coupling conditions
were quickly identified. The same Xantphos/Pd(dba)2 catalytic
system in dioxane proved useful when switching to DMAP as
the stoichiometric base. Leaving the reaction overnight at
100 ꢀC led to the isolation of thalidomide with a good 70%
radiochemical yield (149 MBq, 99% radiochemical purity) after
silica plug filtration and preparative HPLC (Scheme 4).
Finally, the synthesis of the benzophenone derivative fenofi-
brate (7), by a carbonylative Suzuki–Miyaura coupling, was
investigated (Scheme 5). This would position the isotope label
in the metabolically stable benzophenone core of the structure
with two carbon–carbon bonds. On the basis of a report by Bha-
nage et al., conditions facilitating carbonylative Suzuki–Miyaura
couplings in combination with the CO generator were devel-
oped in our group.26 Simply loading PdCl2 (1 mol%) as the
catalyst under ligandless conditions with K2CO3 as the base
proved highly effective. Importantly, the reaction is performed
Entry CO loading (mmol) 3,4 loading (mmol) Yield (%)
1
2
3
4
0.50 COgen
0.25 COgen
0.10 COgen
0.10 14COgen
0.60
0.35
0.20
0.20
95
67
47
37 RCY
RCY: Radiochemical yield
the yield was observed when the CO loading decreased from in anisole, a solvent that turned out to be decisive for CO
0.5 mmol to 0.1 mmol, still a reasonable yield of 47% of olaparib insertion, as other solvents predominantly provided the direct
could be secured after column chromatography (entries 2 and 3). Suzuki–Miyaura coupling products (Scheme 5).
Finally, switching COgen to 14COgen led to a comparable 37%
Application of these optimized Suzuki–Miyaura conditions in
radiochemical yield (99% radiochemical purity) of olaparib (2) the two-chamber glassware at 100 ꢀC overnight afforded 7 in
after a silica plug filtration and preparative HPLC.
72% radiochemical yield (153 MBq, 98% radiochemical purity)
Although a 37% yield may seem poor, the ability to readily after column chromatography.
produce small amounts of a rather complex and highly func-
Having established the utility of 14COgen as a reliable source
tionalized C-14 labeled compound in a single step makes this of the synthetically useful gas 14CO and its application in four
an extremely attractive synthesis. Furthermore, should a larger examples in carbonylative reactions, attention was next turned
Scheme 4. 14C radiolabeling of thalidomide.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

A. T. Lindhardt et al.
Scheme 5. 14C radiolabeling of fenofibrate.
Scheme 6. Scrubber based on the formation of metal-carbonyl complexes.
to identifying a potential scrubber. Even though the aforemen- reliable in the trapping of CO, and the 1H-NMR analysis of the crude
tioned carbonylations were all performed using 14CO as the reaction mixture showed the formation of two species that were
limiting reagent, there is still the possibility that not all of this identified as the ketone 8 and the tertiary alcohol 9 (Scheme 7).
radioactive gas is consumed. The question is then how to
With these experiments in hand, this approach was tested as a
convert this excess 14CO into something that is safely handled potential scrubber for 14CO. Once again, the same amount
and disposed of. A patent from 1981, on the removal of CO by of 14CO was released from 14COgen (214 MBq, 0.1 mmol), and
means of a copper-based ligation system in an acidic medium, the scrubber chamber was loaded with 2 mL n-BuLi (1.6 M in
caught our attention.27 It was decided to test this mixture as a hexanes). The CO-releasing chamber was heated to 100 ꢀC
potential scrubber system as it had shown promising results keeping the scrubber chamber at room temperature, and the
during initial attempts using 12CO. The test was performed in system was left overnight. Dilution of both chambers followed
the two-chamber glassware releasing the same amount of by a scintillation count revealed an excellent 97% of the total
14CO from 14COgen (214 MBq, 0.1 mmol) and leaving the reac- activity could be accounted for when combining the total
tion overnight (Scheme 6). However, it was quickly realized that activity in the two chambers (Scheme 8).
not all gaseous 14CO had been trapped from the headspace.
In practice, this scrubber setup requires the addition of a
Furthermore, upon dilution of the scrubber mixture to perform sealable chamber to COware, containing BuLi, and which can
a scintillation count, the majority of the metal-coordinated be connected to the reaction chambers at any given time. This
14CO dissociated. This occurrence is in stark contrast to the activ- glassware has been developed (14COware), and a schematic
ity that was left behind in the CO-producing chamber, which did drawing is depicted in Figure 1.
not escape the solution either upon dilution or standing open to
Although all reactions discussed so far were performed using
air overnight. This led us to abandon the metal coordination- CO as the limited reagent, we decided to test the scrubber
based scrubber and turned our attention towards trapping the system in combination with the standard aminocarbonylation,
CO covalently through an alternative approach.
depicted in Schemes 1 and 3, applying a twofold excess of CO
After a thorough literature search, a paper from 1970 by (1 mmol scale) generated from COgen (Scheme 9). The amino-
Whitesides and coworkers on the reaction of CO with phenyl carbonylation was run with 4-iodoanisole (0.5 mmol), and the
lithium provided the next lead.28 Whitesides reported that applica- scrubber system was charged to account for the full amount
tion of phenyl lithium or alkyl lithium under a CO atmosphere led of CO applied (10 mmol BuLi, 10 equivalents). A real time
to the formation of several species based on nucleophilic addition manometer was connected to the 14COware to track the
to CO. These systems were tested with both 12CO and 13CO in the reaction pressure profile as a function of time.
presence of 3–10 equivalents of Ph-Li and n-BuLi. In particular, the
From Scheme 9, it is seen that a maximum pressure of
system based on n-BuLi in hexanes proved highly efficient and 2.6 bar is reached after roughly 10–15 min (T1) corresponding
Scheme 7. Reaction of CO with excess n-BuLi.
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

A. T. Lindhardt et al.
Scheme 8. 14CO efficiently trapped by 1.6 M n-BuLi.
to a 1.6 bar partial pressure of CO.29 The aminocarbonylation
then consumes approximately 0.5 mmol of CO causing a pres-
sure decrease to 1.7 bar (T2); after which, the pressure profile
stayed constant for 45 min proving 14COware to be a gas-
tight system. This also provides a 108 min total reaction time
for the aminocarbonylation. At (T3), the connection to the
scrubber system was opened causing an immediate pressure
drop to 1.4 bar because of the increase in overall headspace.
To ensure that the overall pressure would return to 1 bar
upon completion of the scrubbing process, the oil-bath
heater was turned off at (T4) causing the increased rate of
pressure drop seen on the graph. All CO was consumed
within 3 h after the scrubber was connected (T5). Thus an
aminocarbonylation followed by removal of excess CO was
complete within 6 h.
As the final part of the study on the applicability of the
14COgen system for carbonylation chemistry, a stability test
was performed on this 14CO precursor. A portion of the
Figure 1. Schematic drawing of 14COware.
Scheme 9. Aminocarbonylation followed by CO scrubber with pressure profile.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

A. T. Lindhardt et al.
14
215 MBq/mL COgen in toluene was stored at À20 ꢀC. To
2010, 2, 509-513; (c) J. Albaneze-Walker, C. Bazaral, T. Leavey, P. G.
Dormer, J. A. Murry, Org. Lett. 2004, 6, 2097-2100; (d) R. H. Munday,
J. R. Martinelli, S. L. Buchwald, J. Am. Chem. Soc.2008, 130, 2754-2755;
(e) D. A. Watson, X. Fan, S. L. Buchwald, J. Org. Chem. 2008, 73,
7096-7101; (f) W. Magerlein, A. F. Indolese, M. Bellers, Angew. Chem.
Int. Ed. 2001, 40, 2856-2859; (g) W. Magerlein, A. F. Indolese, M.
Beller, J. Organomet. Chem. 2002, 641, 30-40; (h) M. Arthuis,
A. Lecup, E. Roulland, Chem. Commun. 2010, 46, 7810-7812.
[4] For selected examples see: (a) M. V. Khedkar, P. J. Tambade, Z. S. Qureshi,
B. M. Bhanage, Eur. J. Org. Chem. 2010, 6981–6986; (b) M. B. Andrus, Y. Ma,
Y. Zang, C. Song, Tetrahedron Lett. 2002, 43, 9137–9140; (c) B. M. O´
Keefe, N. Simmons, S. F. Martin, Org. Lett. 2008, 10, 5301–5304; (d) Z.
S. Qureshi, K. M. Deshmukh, P. J. Tambade, B. M. Bhanage,
Synthesis 2011, 243–250; (e) P. J. Tambade, Y. P. Patil, A. G. Panda, B.
M. Bhanage, Eur. J. Org. Chem. 2009, 3022–3025; (f) C. Pirez, J. Dheur,
M. Sauthier, Y. Castanet, A. Mortreux, Synlett 2009, 1745–1748.
[5] For selected examples see: (a) X. F. Wu, H. Neumann, M. Beller,
Angew. Chem. Int. Edt. 2010, 49, 5284–5288; (b) X. F. Wu,
H. Neumann, A. Spannenberg, T. Schulz, H. Jiao, M. Beller, J. Am.
Chem. Soc. 2010, 132, 14596–14602; (c) X. Wu, P. Nilsson, M. Larhed,
J. Org. Chem. 2005, 70, 346–349.
[6] For selected examples see: (a) X.-F. Wu, B. Sundararaju, H. Neumann,
P. H. Dixneuf, M. Beller, Chem. Eur J. 2011, 17, 106–110; (b) J. Liu,
J. Chen, C. Xia, J. Catal. 2008, 253, 50–56; (c) M. Lizuka, T. Kondo, Eur.
J. Org. Chem. 2007, 5180–5183; (d) M. T. Rahman, T. Fukuyama,
N. Kamata, M. Sato, I. Ryu, Chem. Commun. 2006, 2236–2238; (e)
A. Fusano, T. Fukuyama, S. Nishitani, T. Inouye, I. Ryu, Org. Lett.,
2010, 12, 2410–2413; (f) T. Fukuyama, R. Yamaura, I. Ryu, Can. J.
Chem., 2005, 83, 711–715; (g) X.-F. Wu, M. Neumann, M. Beller,
Chem. Eur. J., 2010, 16, 12104–12107.
test the stability,
a small fraction was quenched with
N-methylbenzylamine, and a radio HPLC trace was performed. To
date, no visible decomposition of 14COgen has been detected
providing a stability of at least 6 months of this material stored in
toluene at À20 ꢀC.
Conclusion
The formation of 14CO has been revisited based on a novel
CO-releasing system. Combination of the 14CO-precursor
14COgen with a two-chamber glassware system (COware)
afforded an easy to handle approach to perform carbonylation
chemistry using 14CO as the limiting reagent. Four different
carbonylation reactions were tested including two amino-
carbonylations, an amidocarbonylation and a carbonylative
Suzuki–Miyaura coupling. Three of these reactions introduced
the 14C-label in the final synthetic step producing olaparib,
thalidomide, and fenofibrate. Being able to introduce the
14C-isotope late in the linear sequence significantly increases
the usefulness of this approach thereby limiting the required
number of synthetic steps with radioactive material. Because
the generation of 12CO, 13CO, and 14CO from their COgen pre-
cursors is identical, the same set of carbonylation conditions
can be applied for the different isotopes. Finally, to make this
technique generally applicable, the development of a scrubber,
capable of trapping any leftover 14CO, was developed. With this
tool in hand, the handling of 14CO has been made safe, simple,
and readily available to the radiochemist.
[7] D. J. Constable, P. J. Dunn, J. D. Hayler, G. R. Humphrey, J. L. Jr. Leazer,
R. J. Linderman, K. Lorenz, J. Manley, B. A. Pearlman, A. Wells, A. Zaks,
T. Y. Zhang, Green Chem. 2007, 9, 411–420.
[8] For examples on carboxylations with 14CO2, see: (a) Dorff, P. N.; Powell,
M. E.; Elmore, C. S. J. Label. Compd. Radiopharm. 2010, 53, 787–792; (b)
Dorff, P. N.; Powell, M. E.; Elmore, C. S. J. Label. Compd. Radiopharm.
2010, 54, 123–125; (c) C. S. Elmore, J. E. Hall, X. Ye, J. R. Heys, J. Label.
Compd. Radiopharm. 2007, 50, 412–413; (d) P. W. K. Woo, Y.-M. Pu,
C. C. Huang, J. Label. Compd. Radiopharm. 1999, 42, 827–834; (e) R.
Johansson, T. Werner, J. Label. Compd. Radiopharm. 2000, 43,
1265–1273; (f) U. Pleiss, D. Seidel, J. Label. Compd. Radiopharm. 2002
45, 1172–1174; (g) L. Gawell, J. Label. Compd. Radiopharm. 2003, 46,
131–138; (h) S. L. Kitson, S. Jones, W. Watters, F. Chan, D. Madge,
J. Label. Compd. Radiopharm. 2010, 53, 140–146.
Acknowledgements
We are deeply appreciative to the generous financial support
from the Danish National Research Foundation, the Danish
Natural Science Research Council, the Carlsberg Foundation,
the Lundbeck Foundation, and Aarhus University.
[9] For examples on formylations, see: (a) J. Z. Ho, C. S. Elmore, M. A.
Wallace, D. Yao, M. P. Braun, D. C. Dean, D. G. Melillo, C-Y. Chen, Helv.
Chim. Acta 2005, 88, 1040–1047; (b) C. S. Elmore, D. C. Dean, K. Liu,
A. B. Jones, D. G. Melillo, in Synthesis and applications of isotopically
Labelled Compounds, Vol. 7 (Eds.: U. Pleiss, R. Vogens), Wiley, 2001,
224–227; (c) J. T. Kendall, J. Label. Compd. Radiopharm. 1999, 42,
477–484; (d) M. Almeida, A. Boman, T. Lundstedt, J. Label. Compd.
Radiopharm. 2002, 45, 371–377; (e) I. Parveen, M. D. Threadgill,
J. Label. Compd. Radiopharm. 2000, 43, 883–889; (f) D. Zeng, Q. Mi,
H. Sun, H. Wang, J. Label. Compd. Radiopharm. 2004, 47, 167–174.
[10] For examples hydrolysis or reduction of 14C-cyanides, see: (a) T.
Werner, S. Berg, R. Johannsson, J. Label. Compd. Radiopharm. 2004,
47, 175–180; (b) C. S. Elmore, P. N. Dorff, M. E. Powell, J. E. Hall,
T. R. Simpson, J. Label. Compd. Radiopharm. 2011, 54, 239–246; (c)
S. C. Schou, J. Label. Compd. Radiopharm. 2009, 52, 173–176; (d) S.
J. Baker, Y.-K. Zhang, T. Akama, C. Wheeler, J. J. Plattner, R. M. Rosser,
R. P. Reid, N. S. Nixon, J. Label. Compd. Radiopharm. 2007, 50,
245–250; (e) N. Shipley, K. W. M. Lawrie, J. Label. Compd. Radiopharm.
2007, 50, 457–458; (f) Y. Gong, D. C. Hoerr, L. E. Weaner, R. Lin, J.
Label. Compd. Radiopharm. 2008, 51, 268–272.
[11] C. S. Elmore, J. Label. Compd. Radiopharm. 2011, 54, 59–64.
[12] For a review on 14CO see: H. E. Hargraves, A. G. Lashford, A. T. Rees, B.
S. Roughley, J. Label. Compd. Radiopharm. 2007, 50, 435–436.
[13] (a) S. Weinhouse, J. Am. Chem. Soc. 1948, 70, 442–443; (b) J. L. Huston,
T. H. Norris, J. Am. Chem. Soc. 1948, 70, 1968–1969; (c) J. G. Jr. Burr,
W. G. Brown, H. E. Heller, J. Am. Chem. Soc. 1950, 72, 2560–2562.
[14] D. B. Melville, J. G. Pierce C. W. H. Partrigde, J. Biol. Chem. 1949, 180,
299–305.
Conflict of Interest
The authors did not report any conflict of interest.
References
[1] For reviews on CO in transition metal chemistry see: (a) R. Grigg, S. P.
Mutton, Tetrahedron 2010, 66, 5515–5548; (b) A. Brennführer, H.
Neumann, M. Beller, Angew. Chem. Int. Ed. 2009, 48, 4114–4133; (c)
C. F. Bernard, Organomet. 2008, 27, 5402–5422; (d) I. Omae, Coord.
Chem. Rev. 2011, 255, 139–160; (e) T. Morimoto, K. Kakiuchi, Angew.
Chem. Int. Ed. 2004, 43, 5580–5588; (f) A. Yamamoto, Y. Kayaki, K.
Nagayama, I. Shimizu, Synlett 2000, 7, 925–937; (g) J.-J. Brunet, R.
Chauvin, Chem. Soc. Rev. 1995, 24, 89–95; (h) R. Skoda-Földes,
L. Kollár, Curr. Org. Chem. 2002, 6, 1097–1119; (i) M. Beller, B. Cornils,
C. D. Frohning, C. W. Kohlpainter, J. Mol. Catal. A. 1995, 104, 17–85; (j) A.
Brennführer, H. Neumann, M. Beller, ChemCatChem, 2009, 1, 28–41; (k)
I. Ryu, N. Sonoda, Angew. Chem. Int. Ed. 1996, 35, 1050–1066.
[2] For selected examples see: (a) J. R. Martinelli, D. A. Watson, D. M. M.
Freckmann, T. E. Barder, S. L. Buchwald, J. Org. Chem. 2008, 73,
7102–7107; (b) J. R. Martinelli, D. A. Watson, R. H. Munday, S. L.
Buchwald, Angew. Chem. Int. Ed. 2007, 46, 8460–8463; (c) Q. Yang,
H. Cao, A. Robertson, H. Alper, J. Org. Chem. 2010, 75, 6297–6299;
(d) A. C. Tadd, A. Matsuno, M. R. Fielding, M. C. Willis, Org. Lett.
2009, 11, 583–586; (e) H. Cao, H. Alper, Org. Lett. 2010, 12,
4126–4129; (f) A. Deagostino, P. Larini, G. O. Ernesto, L. Pizzuto, C.
Prandi, P. Venturello, J. Org. Chem. 2008, 73, 1941–1945; (g) X.-F.
Wu, H. Neumann, M. Beller, Chem. Eur. J. 2010, 16, 9750–9753.
[3] For selected examples see: (a) A. Zapf, M. Beller, M. Top. Catal. 2002,
19, 101-109; (b) X.-F. Wu, H. Neumann, M. Beller, ChemCatChem
[15] For selected examples see: (a) C. S. Elmore, K. Brush, M. Schou, W.
Palmer, P. N. Dorff, M. E. Powell, V. Hoesch, J. E. Hall, T. Hudzik, C. Hall-
din, C. L. Dantzman, J. Label. Compd. Radiopharm. 2011, 54, 847–854;
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

A. T. Lindhardt et al.
(b) J. P. Simeone, M. P. Braun, L. Liu, S. R. Natarajan, J. Label. Compd.
Radiopharm. 2010, 53, 511–516; (c) C. S. Elmore, D. C. Dean, D. G.
Melillo, J. Label. Compd. Radiopharm. 2000, 43, 1135–1144.
[16] P. Hermange, A. T. Lindhardt, R. H. Taaning, K. M. Bjerglund, D. Lupp,
T. Skrydstrup. J. Am. Chem. Soc. 2011, 133, 6061–6071.
of gaseous CO for carbonylation reactions. PA 2010 70546,
December 15, 2010.
[25] (a) A. G. Sergeev, A. Spannenberg, M. Beller, J. Am. Chem. Soc., 2008,
130, 15549–15563; (b) J. R. Martinelli, D. A. Watson, D. M. M. Freckmann,
T. E. Barder, S. L. Buchwald, J. Org. Chem. 2008, 73, 7102–7107.
[26] M. V. Khedkar, P. J. Tambade, Z. S. Qureshi, B. M. Bhanage, Eur. J. Org.
Chem. 2010, 6981–6986.
[27] K. Kanehori, S. Horigome, M. Katsumoto, Y. Arikawa. U.S. Patent
4,277,452, 1981.
[28] L. S. Trzupek, T. L. Newirth, E. G. Kelly, N. E. Sbarbati, G. M. Whitesides,
J. Am. Chem. Soc. 1973, 95, 8118–8133.
[29] COware is pressure tested to 16 bar by connecting COware to a
pressurized cannister of nitrogen through a manometer and a
reduction valve. However, operating glassware at high pressures
is strongly discouraged, and COware should not be operated at a
total pressure above 5 bar. In any case, extreme caution and
proper safety measures should always be applied when operating
glass equipment at elevated pressures.
[30] G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E. Gottlieb, A.
Nudelman, B. M. Stoltz, J. E. Bercaw, K. I. Goldberg, Organomet.
2010, 29, 2176–2179.
[17] COgen, 13COgen, 14COgen, COware and 14COware are commercially avail-
able through SyTracks A/S. For more information see, www.sytracks.com.
[18] CO is released at a 0.33 mmol scale corresponding to a room
temperature volume of approximately 8.1 mL into a total headspace
of 14 mL (COware Vtot = 20 mL, 6 mL of solvents).
[19] P. Hermange, T. M. Gøgsig, A. T. Lindhardt, R. H. Taaning, T.
Skrydstrup. Org. Lett. 2011, 13, 2444–2447.
[20] D. U. Nielsen, R. H. Taaning, A. T. Lindhardt, T. M. Gøgsig, T.
Skrydstrup, Org. Lett. 2011, 13, 4454–4457.
[21] Z. Xin, T. M. Gøgsig, A. T. Lindhardt, T. Skrydstrup, Org. Lett. 2012, 14,
284–287.
[22] K. Bjerglund, A. T. Lindhardt, T. Skrydstrup, J. Org. Chem. 2012, ASAP.
[23] T. M. Gøgsig, R. H. Taaning, A. T. Lindhardt, T. Skrydstrup, Angew.
Chem. Int. Ed. 2012, 51, 798–801.
[24] R. H. Taaning, P. Hermange, A. T. Lindhardt, T. Skrydstrup,
(Aarhus University, DK) System providing controlled delivery
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012