[11C]Carbon monoxide in the palladium-mediated synthesis of 11C-labelled ketones

Article abstract of DOI:10.1039/a703062b

[¹¹C]Carbon monoxide has been used in the palladium-mediated synthesis of [carbonyl-¹¹C]ketones. Methyl iodide, vinylic and arylic halides and trifluoromethanesulfonates (triflates) have been coupled with tin reagents with insertion of [¹¹C]carbon monoxide at very low concentrations (10-100 nmol[¹¹C]CO in a total volume of 10 ml). The labelled products are obtained in 36-62% isolated decay-corrected radiochemical yields within 30 min of the end of radionuclide production. In order to use [¹¹C]carbon monoxide efficiently, a gas handling system has been developed which allows the radioactive gas to recirculate through the reaction media. The reactions are performed using a one pot procedure. The best results are achieved with mixed tin reagents containing an unsaturated transferable substituent and Pd(AsPh₃)₄. In a typical experiment starting from 25 GBq of [¹¹C]carbon dioxide, 4.2 GBq (47%) of [carbonyl-¹¹C]acetophenone 1 is obtained 30 min after the end of radionuclide production. The specific radioactivity of 1 is 91 GBq μmol^-1. [carbonyl-¹³C]Benzophenone 6 has been synthesised using the same approach to verify the position of the label.

Full text of DOI:10.1039/a703062b

View Article Online / Journal Homepage / Table of Contents for this issue
1
1
[
C]Carbon monoxide in the palladium-mediated synthesis of
1
1
C-labelled ketones
a
a,b
a,b,c
Pelle Lidström, Tor Kihlberg and Bengt Långström
a
Department of Organic Chemistry, Institute of Chemistry, Uppsala University, Box 531,
S-751 21 Uppsala, Sweden
Uppsala University PET Centre, UAS, S-751 85 Uppsala, Sweden
b
c
The Subfemtomol Biorecognition Project, Japan Science and Technology Corporation and
Uppsala University, UAS, S-751 85 Uppsala, Sweden
1
1
11
[
C]Carbon monoxide has been used in the palladium-mediated synthesis of [carbonyl- C]ketones.
M ethyl iodide, vinylic and arylic halides and trifluoromethanesulfonates (triflates) have been coupled
1
1
11
with tin reagents with insertion of [ C]carbon monoxide at very low concentrations (10–100 nmol [ C]CO
in a total volume of 10 ml). The labelled products are obtained in 36–62% isolated decay-corrected
1
1
radiochemical yields within 30 min of the end of radionuclide production. In order to use [ C]carbon
monoxide efficiently, a gas handling system has been developed which allows the radioactive gas to
recirculate through the reaction media. The reactions are performed using a one pot procedure. The best
results are achieved with mixed tin reagents containing an unsaturated transferable substituent and
1
1
Pd(AsPh ) . In a typical experiment starting from 25 G Bq of [ C]carbon dioxide, 4.2 G Bq (47%) of
3
4
1
1
[
carbonyl- C]acetophenone 1 is obtained 30 min after the end of radionuclide production. The specific
radioactivity of 1 is 91 G Bq ìmol . [carbonyl- C]Benzophenone 6 has been synthesised using the same
Ϫ1 13
approach to verify the position of the label.
1
1
Introduction
Recently, a palladium-promoted approach to C-labelled
ketones was presented where halides were coupled with differ-
Substances labelled with short-lived positron emitting nuclides
can be used with positron emission tomography (PET) and are
11
3d
ent tin reagents with the insertion of [ C]carbon monoxide
Scheme 1). By this method different ketones labelled in the
(
1
useful tools in basic and clinical sciences. The radionuclides
1
1
15
18
most frequently used are C, O and F with half-lives of 20.3,
.07 and 110 min respectively. In order to increase the number
L
Pd
L
L
Pd
L
O
2
RX
1
11CO
2
1
1
PdL4
X
R
X
C
R
of available radiotracers and the possibilities of performing
multi-positional labelling, there is a need for new, general label-
ling methods. Considering the short half-lives of the nuclides
used and radiation safety aspects, synthetic radiochemical
methods need to be rapid, reliable and suitable for automation.
3
R'SnR''3
1
1
O
L
O
Although [ C]carbon monoxide was one of the first tracers
1
1
C
11
2
to be used in human studies, it has been used rarely in labelling
R'
R
R' Pd
L
C R
3
4
reactions and only a few examples are found in the literature.
Carbon monoxide usually has a low reactivity in organic syn-
thesis and rather harsh conditions, including high pressures,
high temperatures, long reaction times and highly reactive
Scheme 1 A mechanistic scheme for the synthesis of [carbonyl-
C]ketones from [ C]carbon monoxide, including the four basic steps
11
11
a
in the carbonylative cross coupling reaction: 1, oxidative addition; 2,
4
11
11
reagents often have to be employed. Thus, [ C]carbon mon-
[ C]CO insertion; 3, transmetallation; 4, reductive elimination
oxide has not gained much attention as a radioactive precursor
1
1
for the synthesis of C-labelled tracers.
carbonyl moiety were synthesised. An interesting application
Development of palladium-catalysed carbonylative coupling
reactions has provided a mild and efficient tool for the trans-
formation of carbon monoxide into different carbonyl com-
could be the labelling of the 17β-methyl ketone side chain of
1
1
the D-ring of progestins. [21- C]Progesterone has earlier been
1
1
synthesised using [ C]methyl iodide in a carbon alkylation
5
11
pounds, e.g. ketones, aldehydes, esters and amides. These reac-
reaction and recently from the 17-acid chloride using a [ C]me-
6
tions are especially interesting since they can be performed
under low carbon monoxide pressure and without the use of
polycarbonyl complexes. On account of the low mass involved
thyl cuprate. The synthesis of the 17β-methyl ketone side chain
of progestins via the palladium-catalysed carbonylation reac-
tion has been conducted on the macroscopic scale using halides
1
1
7
11
when high specific radioactivity C is used,† a high pressure of
as substrates. Carbonylation with [ C]carbon monoxide would
1
1
[
C]carbon monoxide is difficult to obtain. The use of polycar-
provide a route to carbonyl labelled compounds and may be
1
1
11
bonyl complexes in C-labelling reactions is unsuitable since it
advantageous in the C-labelling of more sensitive substrates.
1
1
would afford a low yield of labelled product. Furthermore, C-
polycarbonyl complexes would be difficult to synthesise with-
out isotopic dilution, leading to low specific radioactivities
which would be unsuitable for many PET applications.
Results and discussion
1
1
As model compounds, [carbonyl- C]acetophenone 1, [car-
1
1
11
bonyl- C]benzophenone 2, [carbonyl- C]acetone 3, [carbonyl-
1
1
11
C]cyclohexenyl methyl ketone 4 and [carbonyl- C]cyclo-
†
5
Specific radioactivity = radioactivity/mass, typically in the range of
Ϫ1
pentenyl methyl ketone 5 were synthesised via a palladium-
promoted coupling of a halide or trifluoromethanesulfonate
0–500 GBq µmol . In a reaction using 1 GBq, the amount of
radioactive precursor would be approximately 2–20 nmol.
J. Chem. Soc., Perkin Trans. 1, 1997
2701

View Article Online
1
1
11
a
Table 1 Palladium-mediated synthesis of some [carbonyl- C]ketones from [ C]carbon monoxide
Palladium
complex
RY b
(%)
Isolated
RY (%)
c
Entry
1
Substrate
Tin reagent
Sn(CH3)4
T/ ЊC
Product
I
Pd(PPh3)4
130
O
*
45–46
36–37
(2)
(2)
1
O
2
SnPhBu3
Pd(AsPh3)4
80
53–71
(6)
38–61
(3)
CH3I
*
1
I
O
*
3
4
5
SnPh(CH3)3
Sn(CH3)4
Sn(CH3)4
Pd(dppf)2
Pd(PPh3)4
Pd(AsPh3)4
130
130
130
60–69
52–58
(2)
(2)
2
O
*
62–70
3)
52–62
(3)
CH3I
(
3
OTf
O
*
37–65
34–42
(3)
(3)
4
O
OTf
OTf
6
7
Sn(CH3)4
Sn(CH3)4
Pd(dppf)2
130
130
9
—
—
*
5
5
5
Pd(AsPh3)4
O
*
6
O
*
Sn(CH3)3
Pd(AsPh3)4
80
69–70
(2)
53–57
(2)
8
CH3I
a
4
5–85 µmol substrate, 15–45 µmol tin reagent and 1.4 µmol of the palladium complex in 300 µl NMP was used, the reaction mixture was heated for 5
b
min. Radiochemical yield (RY) as determined by HPLC from a sample withdrawn from the reaction mixture. The range is based on (n)
experiments. Decay corrected isolated yield based on the total amount of radioactivity in the recirculating system.
c
1
1
(
[
triflate) with the appropriate tin reagent with insertion of
reduced to [ C]carbon monoxide in a small zinc filled tube
1
1
11
11
C]carbon monoxide (Table 1). [ C]Carbon monoxide was
(85 × 2 mm, 0.65 g Zn) at 400 ЊC. The [ C]carbon monoxide
was transferred to the recirculation unit via ascarite (to trap
unconverted [ C]carbon dioxide, usually less than 5%), valve 3
was switched and the pump turned on. The gas was recirculated
1
1
obtained from [ C]carbon dioxide by reduction over a zinc
catalyst at 400 ЊC. This was conducted either using the Scand-
itronix RNP-17 radionuclide production system or in a newly
developed gas handling system. Nitrogen was used as carrier
gas to the gas handling system where it was replaced by helium.
The reactions were performed in one pot procedures and analy-
ses were performed by analytical radio-LC and radio-GC.
1
1
Ϫ1
at 25–30 ml min and the total volume of the recirculation unit
was 10 ml. After several minutes of recirculation the remaining
gaseous radioactivity was transferred to a balloon and the
radioactivity of the reaction vial, the silica solid-phase extrac-
tion cartridge (SPE) and the balloon was measured. The silica
SPE was used to catch volatile products and to protect the
pump from solvents.
To evaluate the benefit of the recirculation unit, the synthesis
of 1 from [ C]carbon monoxide, phenyl iodide and tetramethyl-
tin was investigated. In these experiments, [ C]carbon mon-
The gas handling system
A major problem in the previously published investigation was
the low trapping efficiency‡ (~10%) of [ C]carbon monoxide in
the reaction media. Carbon monoxide has a low solubility in
most organic solvents. Furthermore, when the [ C]carbon
monoxide was produced in the Scanditronix system it had to be
delivered from the cyclotron vault to the chemistry laboratory
1
1
3
d
11
4
11
11
oxide was produced using the Scanditronix system, trapped on
a short plug of silica at Ϫ196 ЊC and released in a slow stream
of helium upon heating. The drawback with this approach
Ϫ1
in a stream of nitrogen at high flow rates (100–200 ml min ),
1
1
which caused the radioactive gas to pass through the reaction
media under conditions of low trapping efficiency. To increase
the trapping efficiency a gas handling system was developed
where [ C]carbon dioxide was locally converted to [ C]carbon
monoxide which was delivered at lower flow rates and
recirculated through the reaction media (Fig. 1).
In this system [ C]carbon dioxide was trapped and concen-
trated on a column (Porapak Q, 50/80 mesh, 25 × 2 mm) at
Ϫ196 ЊC. Upon heating, the radioactive gas was released in a
(compared to the local conversion of [ C]carbon dioxide to
11 11
[ C]carbon monoxide) was that large amounts of the [ C]car-
1
1
bon monoxide (~20–60%) were reoxidised to [ C]carbon
dioxide when heated on the silica surface. This prompted the
development of the local, small zinc oven.
1
1
11
The recirculation unit increased the trapping efficiency as
well as the radiochemical yield as determined by analytical LC
from a sample withdrawn from the reaction mixture. This led to
an improved radiochemical yield calculated from the total
1
1
Ϫ1
11
stream of helium (4–6 ml min ). Attempts to use a small piece
of steel tubing for the trapping of [ C]carbon dioxide resulted
amount of [ C]carbon monoxide (Fig. 2). The results demon-
1
1
11
strated that with the recirculation unit, [ C]carbon monoxide
in an uncontrolled release of the radioactive gas due to the
evaporation of condensed nitrogen. [ C]Carbon dioxide was
could be efficiently used as a radioactive precursor in the syn-
thesis of C-labelled radiotracers.
1
1
11
Reaction conditions
‡
Trapping efficiency, i.e. the fraction of radioactivity remaining in the
reaction media.
The encouraging results on the recirculation unit prompted a
2
702
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
1
1
11
Fig. 1 A gas handling system for the preconcentration of [ C]carbon dioxide followed by conversion to and recirculation of [ C]carbon monoxide
1
1
Fig. 2 Effect of the use of a recirculating gas system on the radio-
Fig. 3 Temperature dependence of [carbonyl- C]acetophenone 1
1
1
chemical yield of [carbonyl- C]acetophenone 1. The reactions were
performed with Pd(PPh3)4 (1.4 µmol), phenyl iodide (45 µmol) and
radiochemical yield. The reactions were performed with Pd(PPh ) (1.4
µmol), phenyl iodide (45 µmol) and tetramethyltin (45 µmol), in 300 µl
NMP. The [ C]carbon monoxide was introduced into the reaction mix-
ture at room temperature and the reactions were heated for 5 min.
Fraction of the radioactivity trapped in the reaction media. Residual
[ C]carbon monoxide after the reaction. Radiochemical purity of a
sample withdrawn from the reaction mixture. Radiochemical yield
3
4
1
1
11
tetramethyltin (45µmol)in 300µlDMSO. The[ C]carbon monoxidewas
introduced into the reaction mixture at 90 ЊC and the reaction held at
a
a
b
that temperature for 5 min. Performed with the use of the whole
1
1
c
recirculating gas system, including both the preconcentration of the
1
1
d
[
C]carbon monoxide on the silica loop and the recirculation of the gas
b
11
phase. Performed only with the preconcentration of the [ C]carbon
described as Trap × RCP
monoxide on the silica loop and the release with a helium flow of 4 ml
Ϫ1
c
11
min
.
Direct delivery of the [ C]carbon monoxide from the Scand-
8
reactions is usually the transmetallation (Scheme 1, step 3).
d
itronix production system. Fraction of the radioactivity trapped in the
reaction media. Radiochemical purity of a sample withdrawn from the
reaction mixture. Radiochemical yield described as Trap × RCP
e
While CO insertion has been reported to be facile at temperat-
ures as low as Ϫ20 ЊC, the transmetallation step often requires
temperatures above 100 ЊC and long reaction times when
unreactive tin reagents, i.e. tetraalkyltin reagents, are employed.
f
re-examination of the factors governing the outcome of the
1
1
3d
11
palladium-promoted C-carbonylation reaction. As reported,
the use of polar aprotic solvents, e.g. N-methylpiperidin-2-one
NMP) and dimethyl sulfoxide (DMSO), gave the best results
The temperature dependence of the [carbonyl- C]-
acetophenone 1 synthesis from phenyl iodide and tetramethyltin
was examined (Fig. 3). The results indicated that temperatures
higher than 150 ЊC might be used when thermally stable sub-
strates and palladium complexes are employed. However, in
order to enable evaluation of the effect of changes in param-
eters other than the temperature on the radiochemical yield, a
reaction temperature of 130 ЊC was selected. The use of
reagents that promote a faster transmetallation enabled the use
of lower reaction temperatures. Hence, the formation of 1 from
methyl iodide and phenyl(tributyl)tin was rapid at 80 ЊC while
(
while the use of ethers, e.g. tetrahydrofuran (THF ) and 1,4-
dioxane, gave lower radiochemical yields. These results concur
8
with those reported by Farina on the Stille coupling. Although
1
1
the use of DMSO resulted in a higher consumption of [ C]car-
bon monoxide, it also resulted in more by-products compared
to NMP. Furthermore, NMP could be used at higher temperat-
ures and was the solvent selected.
The rate-determining step in palladium-catalysed coupling
J. Chem. Soc., Perkin Trans. 1, 1997
2703

View Article Online
µmol (67 m) phenyl iodide was used, the radiochemical yield
of 1 from phenyl iodide and trimethyltin was diminished. The
amount of tin reagent added did not seem to have a large
impact on the radiochemical yield of 5 from cyclopentenyl-
(trimethyl)tin and methyl iodide, e.g. 15–45 µmol (0.05–0.15 )
cyclopentenyl(trimethyl)tin was used without affecting the
radiochemical yield. The amount of catalyst could be varied
from 0.5 to 6 µmol (1.7–20 m) without any effect on the
radiochemical yield of
tetramethyltin.
1
from phenyl iodide and
Synthetic strategy
1
1
In the synthesis of C-labelled methyl ketones there are alterna-
tive routes available (Scheme 2). Either the methyl group
1
1
Fig. 4 Radiochemical yields of [carbonyl- C]acetophenone 1 using
different ligands. The reactions were performed with in situ generation
of the palladium complex from Pd dba ؒCHCl (0.7 mg, 0.7 µmol), and
2
3
3
O
ligand (4 equiv., 5.6 µmol) in 300 µl NMP. An exception was the use of
commercially available Pd(PPh ) (preformed, 1.4 µmol). Phenyl iodide
OTf
O
*
3
4
1
1
(
45 µmol), tetramethyltin (45 µmol) and [ C]carbon monoxide were
Route A
3
1
introduced into the reaction mixture at room temperature and the reac-
tion mixtures were heated for 5 min at 130 ЊC. Radiochemical yields are
described as trap × radiochemical purity (RCP) of a sample withdrawn
from the reaction mixture.
2
the corresponding transformation from phenyl iodide and
tetramethyltin required a longer reaction time and a reaction
temperature of 130 ЊC.
O
SnMe3
*
Route B
4
Choice of catalyst
Oxidative addition–transmetallation sequences have previously
been reported to be rather tolerant with regard to the choice of
9
11
catalyst. This was also experienced in C carbonylations using
tetraalkyltin compounds. In the present investigation the active
catalyst was usually generated in situ from tris(dibenzyl-
ideneacetone)dipalladium chloroform adduct (Pd dba ؒCHCl )
*
O
Route C
5
2
3
3
and four equivalents of ligand. In the synthesis of 1, five out of
six catalysts gave similar results (Fig. 4). The residual amount
1
1
11
of unreacted [ C]carbon monoxide varied from 1% [Pd-
AsPh ) ] to 34% [preformed Pd(PPh ) ]. In addition to 1,
Scheme
2
Alternative routes for the synthesis of
O, 2,6-dimethylpyridine,
CH Cl ; 2, Sn Me , Pd(PPh ) , THF; 3, [ C]CO, SnMe , Pd(dppf) ,
C-labelled
(
cyclopentenyl methyl ketone. Reagents: 1, Tf
2
3
4
3 4
1
1
1
1
formation of [carbonyl- C]benzophenone was observed. The
additional phenyl group probably originated from the phos-
phine or arsine used as ligand. This scrambling, which was
observed in all experiments, is a phenomenon which was
2
2
2
6
3
4
4
2
1
1
11
NMP; 4, [ C]CO, CH I, Pd(AsPh ) , NMP; 5, [ C]CH I, CO,
3
3
4
3
Pd(PPh ) , 1,4-dioxane.
3
4
originates from the tin compound, i.e. tetramethyltin (route A);
or the methyl group originates from methyl iodide (route B).
Since the rate limiting step usually is the transmetallation, any
factor that facilitates this step may allow the reaction to be run
under milder conditions. In the synthesis of 5 (Table 1, entries
1
0
recently reported by Morita et al. and Segelstein et al. The
radiochemical yield of the by-product was typically 10–25%.
An increased amount of added catalyst, as well as elevated reac-
tion temperature, was found to increase the relative amount of
this by-product.
6
–8), route A (tetramethyltin and cyclopentenyl triflate) was
The choice of catalyst was found to be crucial when triflates
were used as substrates. The radiochemical yield of 4 was
unable to provide the labelled product in a radiochemical yield
exceeding 10%, whereas route B [cyclopentenyl(trimethyl)tin
and methyl iodide] resulted in a radiochemical yield of 70%
(55% isolated). The transfer of a cyclopentenyl group instead of
a methyl group from tin to palladium enhanced the rate of the
transmetallation and the reaction could be performed at 80 ЊC.
The transfer rate for different groups from tin to palladium
is in the order: alkynyl > alkenyl > aryl > benzyl > alkyl.
Attempts to facilitate the transmetallation by the use of
Pd(AsPh ) failed (entry 7), probably due to thermal instability
doubled with the use of AsPh as the ligand as compared to
3
other ligands. AsPh3 has been used to facilitate the rate-
8
determining transmetallation in the Stille coupling. This ligand
enhances the rate of the reaction and simultaneously reduces
the free ligand inhibition observed with more electron donating
ligands. The increased radiochemical yield probably reflected
faster product formation, allowing a higher conversion of
cyclohexenyl triflate to 4 before the triflate and/or the catalyst
decomposed.
9,11
3
4
of the triflate.
Labelling in different positions is a tool to get more detailed
Stoichiometry
12
information on various physiological processes with PET. The
On account of the high specific radioactivity, 10–100 nmol of
11
procedure presented here resulted in C-carbonyl labelled
1
1
[
C]carbon monoxide in a total volume of the recirculation
11
ketones, whereas the use of [ C]methyl iodide, carbon mon-
unit of 10 ml was used in this study. The maximum partial
11
oxide and an organotin reagent would produce the [ C]methyl
1
1
pressure of [ C]carbon monoxide was calculated to be 2.5–25
13
ketone instead (route C, Scheme 2). This approach was stud-
Ϫ4
Ϫ3
Pa (2.5 × 10 to 2.5 × 10 atm). Although many palladium-
catalysed carbonylation reactions are performed under atmos-
pheric or elevated pressure of carbon monoxide, synthesis of
ied in a few experiments, yielding 15% of the cyclopentenyl
11
[
C]methyl ketone using preformed Pd(PPh ) in dioxane.§
3 4
1
1
C-labelled ketones could be efficiently conducted using this
1
1
§
1.3 mg (1.2 µmol) Pd(PPh ) , 300 µl 1,4-dioxane and [ C]CH I in 100 µl
3 4 3
1
1
low pressure of [ C]carbon monoxide. Due to limited avail-
ability of precursor and a desire for high specific radioactivity
product, small scale synthesis is desirable. When less than 20
THF. Sparged with CO for 30 s, 2µl (10 µmol) Sn(cyclopentenyl)(CH ) ,
3
3
1
20 ЊC, 8 min. The radiochemical yield is given as the radiochemical
purity of a sample withdrawn from the reaction mixture.
2
704
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
All three routes are available from the same starting material.
and analysed with Beckman system Gold software. Radioactiv-
ity was measured in an ion chamber, Veenstra Instrumenten bv,
VDC-202.
1
1
Route B may be the first choice where C-carbonylations are
concerned. The use of milder reaction conditions and the pos-
sibility of using less of the substrate are good arguments for
this route. Route A saves one synthetic step in the laboratory
and can be considered when robust substrates are employed.
Phenyl iodide was distilled and methyl iodide was used as
received (Aldrich). Cyclohexenyl and cyclopentenyl triflates
were synthesised from the corresponding cycloalkanones
1
1
15
Route A and B involve the use of [ C]carbon monoxide which
according to published procedures. Tetramethyltin and phe-
could give a higher specific radioactivity compared to the use of
nyl(tributyl)tin were purchased from Aldrich and used as
received. Cyclopentenyl(trimethyl)tin was synthesised from
1
1
[
C]methyl iodide which is employed in route C.
1
6
cyclopentenyl triflate as described. The palladium complexes
were generated in situ from tris(dibenzylideneacetone)-
dipalladium chloroform adduct (Aldrich) and the appropriate
ligand. Ligands were purchased from Lancaster or Aldrich and
used as received. Tetrakis(triphenylphosphine)palladium was
purchased from Lancaster.
N-Methylpiperidin-2-one, NMP, was distilled over calcium
hydride before use. Anhydrous dimethyl sulfoxide, DMSO, was
purchased (Aldrich Sure/Seal) and used as received. Ethereal
solvents were freshly distilled before use.
Characterisation 1
1
The identities of C-labelled products were assessed with ana-
lytical reversed phase LC before and after addition of
unlabelled reference material. The radioactivity that remained
in the gaseous phase was analysed using Radio-GC with
1
1
unlabelled reference and found to be unreacted [ C]carbon
1
1
monoxide. [carbonyl- C]Benzophenone 2 was also character-
1
3
13
ised by C-NMR analysis of [carbonyl- C]benzophenone 6
synthesised by the same method as that used for [carbonyl-
1
1
13
Acetophenone, benzophenone and acetone were purchased,
while cyclohexenyl methyl ketone and cyclopentenyl methyl
ketone were synthesised via Friedel–Crafts acylations on the
C]benzophenone with the simultaneous addition of [ C]car-
1
1
13
bon monoxide and [ C]carbon monoxide. The C NMR signal
1
3
at 196.7 ppm corresponded to the C signal from the carbonyl
carbon of authentic benzophenone. The material 6 was further
analysed on LC-MS and the molecular mass (m/z 184, M ϩ 1)
was found to be one mass unit over the molecular mass of
authentic benzophenone (m/z 183, M ϩ 1).
1
7
corresponding cycloalkenes.
Identities of synthesised materials were determined using H
1
1
3
and C NMR spectroscopy and GC–MS. NMR spectra were
recorded on a Varian XL 300 (300 MHz) or a Varian Gemini
2
spectrometer (200 MHz). Tetramethylsilane or chloroform[ H]
was used as internal standard. GC–MS was performed with GC
separations on a Varian 3400 GC. Mass spectra were recorded
on a Finnigan MAT INCOS 50 instrument with electron
impact ionisation at 70 MeV. LC–MS was performed using a
Micromass VG Quattro with positive atmospheric pressure
Conclusions
1
1
[
C]Carbon monoxide at very low concentrations was used in
the palladium-promoted carbonylative coupling of organic hal-
ides and triflates with organotin compounds. This is a versatile
ϩ
chemical ionisation (APcI ). A Beckman 126 pump, a CMA
240 autosampler and a Beckman Ultrasphere ODS C18 (5 µm,
1
1
method for the production of ketones C-labelled in the car-
1
1
bonyl moiety. Using this approach vinylic or arylic C-labelled
250 × 4.6 mm id) column were used. Mobile phase was water–
1
1
Ϫ1
methyl ketones, e.g. [carbonyl- C]acetophenone
1
and
methanol (20:80), 1 ml min .
1
1
11
[
carbonyl- C]cyclopentenyl methyl ketone 5, as well as C-
labelled ketones containing two unsaturated groups, e.g.
carbonyl- C]benzophenone 2, can be synthesised in good
radiochemical yields. The presented method is rapid, mild, gen-
eral and conducted in a one pot procedure suitable for automa-
The gas handling system was built using Valco AC4W valves,
VICI AG, a Leister typ-700 heat gun and a KNF Neuberger
NMP 30 KVDC pump. Porapak Q 50/80 was purchased from
Supelco and silica gel 100/120 from Alltech.
1
1
[
1
1
11
tion. The method holds promise for routine production of C-
Methyl [carbonyl- C]ketones from tetramethyltin
1
3
labelled ketones. The use of [ C]carbon monoxide in the same
Pd dba ؒCHCl (0.7mg, 0.7µmol)and ligand (4equiv., 5.6µmol)
2
3
3
1
3
method enables the synthesis of different carbonyl C-labelled
ketones.
were dissolved in 300 µl NMP at room temperature. The solution
was conditioned by degassing with helium and kept under a
helium atmosphere during the entire reaction. The halide or
1
1
triflate (45–85 µmol) was added and the [ C]carbon monoxide
was introduced into the recirculation unit. When the radio-
active gas had been trapped, the system was closed and the
pump turned on. Tetramethyltin (45 µmol) was added and the
reaction mixture was heated for 5 min at 130 ЊC while recirculat-
Experimental
General
[
1
1
C]Carbon dioxide production was performed using a Scand-
14 11
itronix MC-17 cyclotron by the N(p,α) C nuclear reaction. A
nitrogen (AGA Nitrogen 6.0) gas target containing 0.1% oxy-
gen (AGA Oxygen 4.8) was used.
1
1
ing the [ C]carbon monoxide. At the end of recirculation, the
heating bath was removed and radioactivity remaining in the
gas phase was transferred into a balloon using a stream of
helium. The radioactivity in the gaseous phase, the silica guard
column and the reaction mixture was measured. The reaction
mixture was purified by preparative LC: solvent A–C (50:50)
isocratic elution 0–5 min, linear gradient to 0:100, 5–10 min,
Analytical LC was performed on a Beckman system
equipped with a Beckman 126 pump, a Beckman 168 UV
ϩ
detector in series with a β -flow detector and a Beckman C-18
ultrasphere ODS column, 5 µm, 250 × 4.6 mm (id). A Gilson 231
was used as autoinjector.
Preparative LC was performed on an identical Beckman sys-
tem, equipped with a Beckman ultrasphere ODS column,
Ϫ1
flow 5 ml min . The identity and radiochemical purity of the
collected fractions were assessed by analytical LC: solvent A–C
2
50 × 10 mm (id). Synthia, a robotic system for production of
(60:40) isocratic elution 0–5 min, linear gradient to 10:90, 5–8
1
4
Ϫ1
radiopharmaceuticals, was used for injection and fraction col-
lection. Mobile phases used for analytical and preparative LC
were: 0.05  ammonium formate pH 3.5 (A), 0.01  potassium
dihydrogen phosphate (B) and acetonitrile–water (50:7) (C).
Radio-GC was performed on a Shimadzu GC-A14 with
min, flow 2 ml min , wavelength 254 nm, retention times (t ) of
r
1, 4 and 5 were 4.6, 6.7 and 3.9 min respectively.
1
1
Methyl [carbonyl- C]ketones from methyl iodide
The syntheses were carried out as described for the synthesis
ϩ
11
TCD detection in series with a β -flow detector. An Alltech
of methyl [carbonyl- C]ketones from tetramethyltin with the
CTR column was used. Conditions were as follows: column
following changes: methyl iodide (85 µmol) and tin reagent (15–
45 µmol) were used and the reaction mixture was heated at
80 ЊC.
temp. 35 ЊC, injection temp. 50 ЊC, TCD temp. 70 ЊC, TCD cur-
Ϫ1
rent 150 mA, carrier gas He at 60 ml min . Data were collected
J. Chem. Soc., Perkin Trans. 1, 1997
2705

View Article Online
1
1
[
carbonyl- C]Benzophenone 2
2 C. A. Tobias, J. H. Lawrence, F. J. W. Roughton, W. J. Rooth and
M. I. Gregerson, Am. J. Physiol., 1945, 145, 253.
The synthesis was carried out as described for the synthesis of
methyl [carbonyl- C]ketones from tetramethyltin with the fol-
lowing changes: diphenylphosphinoferrocene (dppf) (1.6 mg,
.8 µmol) was used as ligand. Phenyl iodide (5 µl, 45 µmol) and
phenyl(tributyl)tin (8 µl, 45 µmol) were used and the reaction
mixture was heated at 80 ЊC. The reaction mixture was purified
by preparative LC: solvent A–C (50:50) isocratic elution
1
1
3 (a) D. Y. Tang, A. Lipman, G.-J. Meyer, C.-N. Wan, A. P. Wolf,
J. Labelled Compd. Radiopharm., 1979, 16, 435; (b) P. J. Kothari,
R. D. Finn, M. M. Vora, T. E. Booth, A. M. Emran and G. W.
Kabalka, Int. J. Appl. Radiat. Isot., 1985, 36, 412; (c) M. R.
Kilbourn, P. A. Jerabek and M. J. Welch, J. Chem. Soc., Chem.
Commun., 1983, 861; (d) Y. Andersson and B. Långström, J. Chem.
Soc., Perkin Trans. 1, 1995, 287.
2
4
H. M. Colquhoun, D. J. Thompson and M. V. Twigg, Carbonylation,
Plenum Press, New York, 1991.
0
–5 min, linear gradient to 0:100, 5–10 min, flow 5 ml
Ϫ1
min . The identity and radiochemical purity of the collected
fractions were assessed by analytical LC: solvent A–C (60:40)
isocratic elution 0–5 min, linear gradient to 10:90, 5–8 min,
5
(a) A. C. Gyorkos, J. K. Stille and L. S. Hegedus, J. Am. Chem. Soc.,
1
990, 112, 8465; (b) G. T. Crisp, W. J. Scott and J. K. Stille, J. Am.
Chem. Soc., 1984, 106, 7500; (c) A. M. Echavarren and J. K. Stille,
J. Am. Chem. Soc., 1988, 110, 1557.
Ϫ1
flow 2 ml min , wavelength 254 nm, retention time of 2 was
6
7
(a) W. Vaalburg, J. W. Terpstra, T. Wiegman, K. Ishiwata, A. M. J.
Paans and M. G. Woldring, J. Labelled Compd. Radiopharm. 1986,
9
.6 min.
2
3, 1422; (b) P. Lidström, H. Neu and B. Långström, J. Labelled
11
[
carbonyl- C]Acetone 3
Compd. Radiopharm., 1997, 39, 695.
R. Skoda-Föles, Z. Csákai, L. Kollár, J. Horváth and Z. Tuba,
Steroids, 1995, 60, 812.
The synthesis was carried out as described for the synthesis of
methyl [carbonyl- C]ketones from tetramethyltin with the fol-
1
1
lowing changes; PPh (1.5 mg, 5.6 µmol) was used as ligand and
8 (a) V. Farina and G. P. Roth, Adv. Met-Org. Chem., 1996, 5, 1;
b) V. Farina, Pure Appl. Chem., 1996, 68, 73.
3
(
methyl iodide (85 µmol) as substrate. Most of the product 3
distilled from the reaction mixture and was caught in the silica
guard column, which after the reaction was eluted with 2 ml of
LC eluent into the reaction mixture. The resulting solution was
purified on preparative LC: Solvent B–C (98:2) isocratic elu-
tion 0–5 min, linear gradient to 0:100, 5–10 min, flow 5 ml
9
L. S. Hegedus, Transition metals in the synthesis of complex organic
molecules, University science books, Mill Valley, 1994.
1
0 (a) D. K. Morita, J. K. Stille and J. R. Norton, J. Am. Chem. Soc.,
995, 117, 8576; (b) B. E. Segelstein, T. W. Butler and B. L. Chenard,
J. Org. Chem., 1995, 60, 12.
1
11 J. W. Labadie and J. K. Stille, J. Am. Chem. Soc., 1983, 105, 6129.
12 (a) T. Kihlberg, S. Valind and B. Långström, Int. J. Appl. Radiat.
Isot., Nucl. Med. Biol., 1994, 21, 1053; (b) T. Kihlberg, S. Valind and
B. Långström, Int. J. Appl. Radiat. Isot., Nucl. Med. Biol., 1994, 21,
Ϫ1
min . The identity and radiochemical purity of the collected
fractions were assessed by analytical LC: solvent B–C (95:5)
Ϫ1
isocratic elution 0–5 min flow 1 ml min , linear gradient to
1
067; (c) J. R. Grierson, J. E. Biskupiak, J. M. Link and K. A.
Ϫ1
5
:95, 5–15 min, flow 2 ml min , wavelength 260 nm, retention
Krohn, Appl. Radiat. Isot., 1993, 44, 1449.
time of 3 was 4.9 min.
13 (a) Y. Andersson, Ph.D. Thesis, University of Uppsala 1995;
(
b) Y. Andersson and B. Långström, in Synthesis and applications of
1
3
isotopically labelled compounds, eds. J. Allen and R. Voges, John
Wiley and Sons, 1995.
4 P. Bjurling, R. Reineck, G. Westerberg, A. D. Gee, J. Sutcliffe and
B Långström, Proceedings of the VIth workshop on targetry and
target chemistry, Vancouver, Canada, 1995, 282.
15 (a) P. J. Stang and W. Treptow, Synthesis, 1980, 283; (b) P. Stang and
T. E. Deuber, Org. Synth. 1974, 54, 79.
6 W. D. Wulff, G. A. Peterson, W. E. Bauta, K.-S. Chan, K. L. Faron,
S. R. Gilbertson, R. W. Kaesler, D. C. Yang and C. K. Murray,
J. Org. Chem., 1986, 51, 277.
[
carbonyl- C]Benzophenone 6
11
As described for [carbonyl- C]benzophenone 2 with the follow-
ing changes: Pd dba ؒCHCl (1.4 mg, 1.4 µmol), AsPh (3.4 mg,
.6µmol), phenyliodide(20µl, 180µmol)and phenyl(tributyl)tin
40 µl, 225 µmol) were used. [ C]Carbon monoxide (5 ml, 225
µmol) was slowly added via syringe after the addition of the tin
1
2
3
3
3
5
(
13
1
reagent.
1
7 (a) N. Jones and H. T. Taylor, J. Chem. Soc., 1959, 4017; (b)
E. E. Royals and C. M. Hendry, J. Org. Chem., 1950, 15, 1147.
Acknowledgements
Financial support was provided by the Swedish Natural
Sciences Research Council (K3463).
References
Paper 7/03062B
Received 6th May 1997
Accepted 6th June 1997
1
H. N. Wagner, Z. Szabo and J. W. Buchanan, Principles of Nuclear
Medicine, W. B. Saunders Company, Philadelphia, 1995.
2
706
J. Chem. Soc., Perkin Trans. 1, 1997