Transition metal mediated synthesis using [11C]CO at low pressure - A simplified method for 11C-carbonylation

Article abstract of DOI:10.1002/jlcr.2930

Transition metal mediated carbonylation with [¹¹C]CO has proven a useful method to label a wide array of compounds in the carbonyl position. However, the general use in radiopharmaceutical synthesis has been hampered by the low solubility of carbon monoxide in most solvents and the resulting challenge to confine [¹¹C]CO in low volume reaction vessels. This paper introduces a method that utilises xenon to transfer pre-concentrated [¹¹C]CO to a sealed disposable glass vial containing carbonylation reagents. The high solubility of xenon in the organic solvent made it possible to confine the [¹¹C]CO without utilising a pressure autoclave or chemical trapping additives. The utility of the method in ¹¹C- carbonylation was investigated by conducting three model reactions, where [ ¹¹C-carbonyl]N-benzylbenzamide, [¹¹C-carbonyl]triclocarban and [¹¹C-carbonyl]methyl nicotinate were afforded in decay corrected radiochemical yields of 71 ± 6%, 42 ± 15% and 29 ± 10%, respectively. These promising results and the straight forward technical implementation suggest that ¹¹C-cabonylation can become a viable mean to provide labelled carbonyl functionalities in routine radiopharmaceutical synthesis. Compounds labelled with short lived positron emitters are used in Positron Emission Tomography, a molecular imaging technology with applications in clinical diagnostics, clinical research and basic biomedical research. Xenon was used to transfer pre-concentrated [¹¹C]CO to a sealed disposable glass vial containing carbonylation reagents. The high solubility of xenon in the organic solvent made it possible to confine the [¹¹C]CO without utilising high pressure or chemical trapping additives. The utility of the method in ¹¹C-carbonylation was investigated by conducting three model reactions, where [¹¹C-carbonyl]N-benzylbenzamide, [ ¹¹C-carbonyl]triclocarban and [¹¹C-carbonyl]methyl nicotinate were successfully synthesised. Copyright

Full text of DOI:10.1002/jlcr.2930

Research Article
Received 15 February 2012,
Revised 27 March 2012,
Accepted 14 April 2012
Published online 11 May 2012 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.2930
Transition metal mediated synthesis using
[11C]CO at low pressure – a simplified method
11
for C-carbonylation
*
Jonas Eriksson, Johan van den Hoek, and Albert D. Windhorst
Transition metal mediated carbonylation with [11C]CO has proven a useful method to label a wide array of compounds in the
carbonyl position. However, the general use in radiopharmaceutical synthesis has been hampered by the low solubility of
carbon monoxide in most solvents and the resulting challenge to confine [11C]CO in low volume reaction vessels. This paper
introduces a method that utilises xenon to transfer pre-concentrated [11C]CO to a sealed disposable glass vial containing
carbonylation reagents. The high solubility of xenon in the organic solvent made it possible to confine the [11C]CO without
utilising a pressure autoclave or chemical trapping additives. The utility of the method in 11C-carbonylation was investigated
by conducting three model reactions, where [11C-carbonyl]N-benzylbenzamide, [11C-carbonyl]triclocarban and [11C-carbonyl]
methyl nicotinate were afforded in decay corrected radiochemical yields of 71 ꢀ 6%, 42 ꢀ 15% and 29 ꢀ 10%, respectively.
These promising results and the straight forward technical implementation suggest that 11C-cabonylation can become a
viable mean to provide labelled carbonyl functionalities in routine radiopharmaceutical synthesis. Compounds labelled with
short lived positron emitters are used in Positron Emission Tomography, a molecular imaging technology with applications
in clinical diagnostics, clinical research and basic biomedical research.
Keywords: carbon-11; carbon monoxide; carbonylation; palladium; rhodium; Positron Emission Tomography; PET
use in radiopharmaceutical synthesis has been hindered by poor
Introduction
solubility and the resulting difficulty to confine [11C]CO in low
Carbon monoxide can be efficiently reacted with organic
substrates by transition metal-catalysed carbonylation to form
amides, carboxylic acids, ketones, ureas and other carbonyl
volume reaction vessels. When bubbled conventionally into
reaction solutions, the trapping efficiency of [11C]CO has
been reported to be very low, for example, 1–10% in THF and
containing compounds.1 This makes carbon monoxide an impor-
tant feedstock for both industrial bulk processes as well as small
scale synthesis of fine chemicals. Carbonylation have also shown
to be a versatile method for the labelling of compounds in the
carbonyl position by use of [11C]CO (b+, t1/2 = 20.4 min).2–10
Compounds labelled with short lived positron emitters are
utilised in positron emission tomography (PET), a molecular
imaging technology which can examine pharmacokinetics and
biological functions based on the in vivo distribution of the posi-
tron emitter. PET applications often require that the labelled
compound has high specific activity (GBq/mmol), something that
favours the use of [11C]CO in labelling synthesis as the relatively
low abundance of carbon monoxide in air reduces the risk for
isotopical dilution. The utility and use of [11C]CO in carbonylation
reactions resembles to a large extent, the general scheme of
transition-metal-catalysed carbonylation with isotopically
unmodified carbon monoxide. The most important differences
are that minute amount of [11C]CO is used (10–100 nmol) and
that the conversion of [11C]CO to product is optimised while all
other reagents are employed in large excess including transition
metal complexes. The use of large stochiometric excess rather
than strictly catalytic conditions often facilitate rapid reactions
(<5 min), an important requirement because of the short
physical half-life of 11C. Although [11C]CO can provide efficient
routes to compounds labelled in the carbonyl position, routine
4.5–6.2% in 1,2-dimethoxyethane/water.11,12 Hence, the overall
yields were severely reduced although the trapped [11C]CO was
converted to product in good yields. A breakthrough came with
a method that reacted [11C]CO in a small autoclave (200 mL)
under high solvent pressure.7,13 Near quantitative transfer of
[11C]CO to the autoclave was achieved by concentrating the
[11C]CO on a few miligramme of silica gel, cooled at ꢁ196ꢂC,
followed by release and transfer of the [11C]CO to the autoclave
in a stream of helium which pressurised the autoclave to 3–5 bar.
The carbonylation reaction was performed after adding a
reagent solution to the autoclave at a pressure of 350 bar to
force the [11C]CO and the helium transfer gas into solution. A
high pressure system has also been presented where [11C]CO
and the reagent solution were simultaneously added to a
loop-reactor by flooding the CO-trap.14 Enhanced trapping of
[11C]CO at atmospheric pressure was later reported by reversible
Department of Nuclear Medicine & PET Research, VU University Medical Center,
Amsterdam, The Netherlands
*Correspondence to: Jonas Eriksson, Department of Nuclear Medicine & PET
Research, VU University Medical Center, The Netherlands.
E-mail: paj.eriksson@gmail.com
J. Label Compd. Radiopharm 2012, 55 223–228
Copyright © 2012 John Wiley & Sons, Ltd.

J. Eriksson et al.
Figure 1. Instrumental set-up used in the 11C-carbonylation reactions with [11C]CO at near ambient pressure.
switched so that only the column effluent containing [11C]CO2 was
transferred over the reductor column (400ꢂC, zinc, 14–50 mesh, Merck).
The reductor column converted the [11C]CO2 on-line to [11C]CO and
any unreacted [11C]CO2 was subsequently removed by the Ascarite
column (0.5 g). The [11C]CO was then passed to the CO-trap (ꢁ196ꢂC,
anhydrous silica gel, 1 mg) where it was concentrated to a very small
volume. The transfer gas flowing through the CO-trap was changed
from helium to xenon (Xe ≥99.995%, 1.5 mL/min, Fluka) by switching
valve V3, and the trap was heated to release the [11C]CO, which was
bubbled via a needle into a solution contained in a septum equipped
conical vial (1 mL). The vial was completely sealed except for the
transfer needle. After the transfer was completed, the needle was
removed and the vial with [11C]CO was ready for use. All experiments
were carried out with the same vial type (1 mL, clear crimp vial,
10 ꢄ 32 mm, type 27333, Supelco) and septum (11-mm aluminium
crimp cap, 1.5-mm silicone/PTFE seal, Grace Alltech). An in-house
machined aluminium cap was fitted over the septum when performing
reactions at elevated temperatures, Figure 2.
The CO-trap, CO2-trap and the gas purifier column contained silica gel
(100/80 mesh, GC-stationary phase, type 5655, Alltech). The traps
contained anhydrous silica gel, whereas the purifier column contained
moist silica gel, the latter obtained by adding water (5 mass%) to
silica gel dried at 120ꢂC for 16 h. The CO-trap was devised by
enlarging the inner diameter at one side of a stainless steel tubing
(o.d. 1/16′′, i.d. 0.02′′), filling the resulting cavity (1 ꢄ 1.2 mm) with silica
gel and then connecting the tubing to a union (1/16′′, ZU1M, VICI Valco)
equipped with a screen (1/16′′, 10 m, 10SR1-10, VICI Valco), as depicted
in Figure 3. The newly assembled CO-trap was conditioned at
approximately 100ꢂC while flushing the silica gel with helium gas to
remove moisture.
complexation with BH3ꢃTHF in the reaction solution.15 More
recently, a method was presented that achieved near quantita-
tive trapping by use of a copper complex that reversibly bound
[11C]CO bubbled conventionally into the solution.9 The [11C]CO
was released from the copper complex when the palladium
feedstock and ligands used for the carbonylation reaction were
added. Hence, the catalytically active complex was added or
formed in-situ in the presence of [11C]CO and the copper
complex. Another notable development was the use of micro-
fluidics to perform 11C-carbonylation reactions.16,17
The aim of the current work was to investigate if a soluble
transfer gas enabled [11C]CO to be added and utilised in
carbonylation reactions at near ambient pressure. This would from
a technical perspective greatly simplify [11C]CO carbonylation
compared with the high pressure method. It also opens up for
exploration of more sensitive catalyst systems, as the reagent
solution can be carefully prepared before adding [11C]CO in a
one-pot fashion without the use of complexation additives.
Ultimately, this method may allow [11C]CO to be used in radiophar-
maceutical synthesis with the same ease as [11C]methyl iodide.
Experimental
General
All radiochemical yields given are decay corrected and based on
activity measurements and analytical HPLC analysis. The identities
of products were assessed by co-elution with isotopically unmodified
reference compounds. The HPLC analysis included a wash of the
column with acetonitrile/water 90/10 (v/v), performed after elution
of the desired product. Measurements with a Geiger Mueller detector
confirmed that no activity remained on the column. Activity
measurements were performed in
a dose calibrator (VDC-405,
Veenstra). Samples withdrawn from reaction mixtures were analysed
by using analytical radio/UV-HPLC (PU-2089 plus, UV-2075 plus, Jasco).
Marvin was used for drawing chemical structures (Marvin 5.8, 2012,
ChemAxon). Reagents and reference compounds were obtained from
Sigma-Aldrich and used without further purification.
Production and handling of [11C]CO
11C was prepared by the 14N(p,a)11C nuclear reaction and obtained as
[
11C]CO2 by proton irradiation (18 MeV, Cyclone 18/9, IBA) of nitrogen
gas containing 0.5% oxygen (Linde gas). The [11C]CO2 was then trans-
ferred to the experimental set-up and concentrated at the CO2-trap
(ꢁ196ꢂC, anhydrous silica gel, 50 mg), Figure 1.
The CO2-trap was then heated and valve V1 was switched to transfer
the activity in helium (He 6.0, 30 mL/min, Praxair) over a gas purifier
column (400 ꢄ 4 mm, silica gel, ~5 mass% water). Valve V2 was then
Figure 2. Septum equipped crimp top vial (1 mL) with an additional aluminium
cap to support the septum. The cap enabled reaction temperatures significantly
higher than the boiling point of the solvent.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 223–228

J. Eriksson et al.
75/25/0.1(v/v/v) eluent at a flow of 1 mL/min, 254 nm, tR =4.9min. The Radio-
chemical yield was 29 ꢀ 10% (n=3).
Results and discussion
The aim of this work was to develop a method to perform car-
bonylation reactions with [11C]CO at near ambient pressure
and without complexing additives, as this would greatly simplify
applications in radiopharmaceutical synthesis. The low pressure
makes possible to use disposable reaction vials hence eliminat-
ing carry over issues and reducing the need for cleaning valida-
tion in a GMP setting (good manufacturing practice). Although
[11C]CO has too low solubility to be efficiently trapped when
bubbled conventionally through reaction mediums, [11C]CO
could instead be confined in a sealed reaction vial with only
minor pressure increase by using a soluble transfer gas. Xenon
was selected as transfer gas in this investigation because it is
relatively inert and has high solubility in many organic solvents.
A reliable production method was needed to perform experi-
ments with [11C]CO. Initially, the reduction of [11C]CO2 over
heated zinc18 did not proceed in our lab, as the fresh zinc was
immediately and permanently deactivated without generating
useful amounts of [11C]CO. Gas chromatographic purification of
[11C]CO2 was then investigated to learn why the reduction failed.
Figure 3. The CO-trap consisted of a miniature silica gel column that was cooled
with liquid nitrogen. A cavity for the silica was created at the end of a stainless steel
tubing (1/16′′) by enlarging the inner diameter by drilling. The tubing was then
connected to a standard Valco union where it was pressed against a screen filter.
The arrow shows the direction of the transfer gas flow when trapping and releasing
the [11C]CO.
Synthesis of [11C-carbonyl]N-benzylbenzamide
To tris(dibenzylideneacetone)dipalladium(0) (Pd2(dba)3) (0.8mg, 0.9mmol)
and triphenylphosphine (2.9mg, 11mmol) was added THF (0.7 mL),
followed by iodobenzene (2mL, 18 mmol) and benzylamine (4mL, 37 mmol).
The crimp vial was fitted with a septum and vent needle, after which the
solution was degassed by bubbling with helium. The vial was then trans-
ferred to the experimental set-up where the [11C]CO transfer needle was
pierced through the septum and the gas phase was briefly purged with
xenon gas (1.5 mL/min). The vent needle was removed and the transfer The resulting method to purify the [11C]CO2 over silica gel
needle was pushed down into the reaction solution prior to transfer of
[
enabled the zinc to be reused typically 7–10 times with full
[11C]CO2 conversion before being depleted. A more detailed
description of the investigation will be given elsewhere (Verbeek
et al., [11C]phenytoin revisited, manuscript in preparation). The
second important step was to concentrate the [11C]CO to a small
volume to facilitate transfer to the low volume reaction vial.
Quantitative trapping at the CO-trap was achieved by using
anhydrous silica gel cooled at liquid nitrogen temperature. This
has already been described in the literature together with
11C]CO. After the transfer was completed, the needle was removed and
the activity in the vial was measured. The vial was then heated at 100ꢂC
in a heating block for 5min to yield [11C-carbonyl]N-benzylbenzamide.
The activity in the vial was measured, followed by degassing and a
repeated activity measurement. A sample was withdrawn from the
reaction mixture and analysed by using radio/UV-HPLC with a Grace
Platinum 100A 5u C18 column and acetonitrile/water 45/55 (v/v) eluent
at a flow of 1 mL/min. The identity of the product was confirmed by co-
elution of the radiolabelled product with authentic reference
compound monitored at 254 nm, retention time (tR) = 6.4 min. The the high pressure 11C-carbonylation method.13 However, the
Radiochemical yield was 71 ꢀ 6% (n = 3).
CO-trap used in the current work was devised slightly different
to minimise the silica volume while simplifying the assembly
Synthesis of [11C-carbonyl]triclocarban (3-(4-chlorophenyl)-
by using 1/16′′ fittings, Figure 3.
1-(3,4-dichlorophenyl)[11C]urea)
Transfer of [11C]CO in xenon into THF
The synthesis followed the same general procedure as used in the
labelling of the amide. Chloro(1,5-cyclooctadiene)rhodium(I) dimer
Tetrahydrofuran is a suitable solvent for many types of carbony-
(0.7mg, 1.6mmol) and triphenylphosphine (1.75mg, 6.7mmol) were placed
lation reactions. At equilibrium, the solubility of xenon in THF
in a 1mL vial. THF (0.5mL) was added followed by 3,4-dichloroaniline
corresponds to 3.9 mL gas per 1.0 mL of solvent (at 1.0 atm
(10.6 mg, 93 mmol) and 1-azido-4-chlorobenzene in tert-butyl methyl ether
partial pressure and 25ꢂC).19 In comparison, this is 160 times
(20 mL, 0.5M, 10 mmol). An aluminium cap was fixed on top of the septum.
After handling and addition of [11C]CO as described earlier, the vial was
heated at 140ꢂC for 5 min to yield [11C-carbonyl]triclocarban. HPLC analysis
was performed on a Phenomenex Luna 5u C18 column with acetonitrile/
water 80/20 (v/v) eluent at a flow of 1mL/min, 254 nm, tR = 7.7min. The
Radiochemical yield was 42 ꢀ 15% (n = 3).
higher solubility than that of helium. In the first experiments,
[11C]CO was transferred into THF (700 mL). The carbonylation
reagents were omitted to exclude effects from chemical
trapping, for example, by the transition metal complex. The
transfer from the CO-trap to the vial took less than 10 s,
facilitated by the small volume in the CO-trap and stainless steel
capillary tubings. Tiny bubbles were observed on the surface of
the solution during the transfer, indicating that not all xenon
was dissolved. However, the pressure build-up in the vial was
not high enough to cause noticeable leakage of radioactive
material when removing the transfer needle. Activity measure-
ments showed that the transfer was quantitative with no remain-
ing activity in the CO-trap. Although xenon enabled transfer to
the vial at low pressure, it was assumed that the choice of
transfer gas would not influence the physical solubility of [11C]CO
in THF. This was confirmed when repeating the experiments with
Synthesis of [11C-carbonyl]methyl nicotinate
The synthesis followed the same general procedure as used in the
labelling of the amide. Pd2(dba)3 (0.8mg, 0.9 mmol) and triphenylphosphine
(2.7mg, 10.3mmol) were placed in a 1-mL vial. THF (0.5mL) was added
followed by 3-iodopyridine (2.1 mg, 10.2 mmol). The vial was heated at
70ꢂC for 2 min then removed from heating, and trimethylphenylammonium
hydroxide in methanol (0.5 M, 50 mL, 25 mmol) was added. An aluminium
cap was fixed on top of the septum. After handling and addition of [11C]
CO as described previously, the vial was heated at 160ꢂC for 5min to
yield
[
11C-carbonyl]methyl nicotinate. HPLC analysis was performed
on a Phenomenex Luna 5u C18 column with acetonitrile/water/TFA the addition of a vent needle inserted through the septum of the
J. Label Compd. Radiopharm 2012, 55 223–228
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

J. Eriksson et al.
O
PPh3
I
Pd2(dba)3
NH2
N
H
*
+
*
+
CO
THF
100°C
5 min
Scheme 1. Synthesis of [11C-carbonyl]N-benzylbenzamide. Position of the 11C label = *.
Cl
PPh3
N3
RhCl(cod)2
H2N
Cl
Cl
Cl
Cl
O
+
+
*
CO
THF
140°C
5 min
Cl
N
H
*
N
H
Scheme 2. Synthesis of [11C-carbonyl]triclocarban. Position of the 11C label = *.
subsequently reacts with amines.2–4,21 The key advantage of
using [11C]CO over amination of [11C]phosgene is that
unsymmetrical [11C]ureas can be readily obtained. In the second
model reaction was [11C-carbonyl]triclocarban synthesised, an
unsymmetrical urea derivative with antimicrobial properties,
Scheme 2.22 A Rh(I) complex was prepared in THF (500 uL), after
which the azide and aniline were added. An aluminium cap, with
a bore to enable insertion of the transfer needle, was fixed on
the vial prior to the addition of [11C]CO, Figure 2. The structural
support on top of the septum made it possible to heat the reac-
tion mixture significantly above the boiling point of the solvent,
without tearing the septum or leakage of activity. The [11C]CO
conversion was 95 ꢀ 5% and the selectivity towards the
desired product was 44 ꢀ 18%, giving a radiochemical yield of
42 ꢀ 15% (n = 3). The radiochemical yield was not significantly
affected when the reaction volume was reduced to 250 mL while
keeping the same reagent concentration (32%, n = 1). The high
conversion of [11C]CO was in line with what generally has been
reported for Rh(I) mediated 11C-carbonylations conducted with
the high pressure method.2–4,21 The [11C]CO conversion still
remained high (90%, n = 1) when the reaction temperature was
lowered to 100ꢂC, but the product selectivity dropped signifi-
cantly (7%, n = 1), possibly because of the reduced reactivity of
the deactivated 3,4-dichloroaniline at lower temperature and
consequently more remaining rhodium-[11C]carbonyl species at
the end of reaction. Hence, the yield may be improved by shift-
ing the chloro-substituents so that the azide has two and the
aniline has one; this was however not investigated.
vial. In this open system, the trapping of activity in the vial was 26%
when xenon was used as transfer gas and 16% when helium was
used. After rapid purge of the gas phase with 5 mL of air, 10%
and 8% of the activity remained, respectively. Hence,
approximately, the same amount of [11C]CO was dissolved with
both transfer gases. The higher activity in the vial immediately after
transfer with xenon was most likely because of solvation of
the transfer gas and hence less efficient purge of the gas phase
during transfer.
11C-carbonylation
The applicability of using the method in 11C-carbonylation
was tested with three model reactions. In the first reaction,
[11C-carbonyl]N-benzylbenzamide was synthesised according to
Scheme 1. The transition metal complex was prepared from
Pd2(dba)3 and triphenylphosphine in THF (700 uL) followed by
the addition of the amine, organohalide and [11C]CO.
After the 5-min reaction time, the crude mixture was transferred
to a 2-mL vial and purged with helium to remove unreacted [11C]
CO and potential volatile by-products. On the basis of the activity
at the start of the carbonylation reaction, 88 ꢀ 4% still remained
after transfer and purge, defined here as the conversion of [11C]
CO to non-volatile products. HPLC analysis of the purged-reaction
mixture showed that 81ꢀ 3% of the remaining activity was in
the form of [11C-carbonyl]N-benzylbenzamide, defined here as
the reaction selectivity towards desired product. The radiochemical
yield was 71 ꢀ 6% (n = 3) as calculated from the conversion and
selectivity according to Equation [1].
In the third model reaction, [11C-carbonyl]methyl nicotinate
was synthesised by methoxycarbonylation of 3-iodopyridine
according to Scheme 3. The solution of Pd(0) and iodopyridine
in THF (500 mL) was preheated at 70ꢂC to facilitate rapid oxidative
addition of the pyridine ring before trimethylphenylammonium
hydroxide (0.5 M) in methanol and [11C]CO were added. Again,
Conversion of
½11CꢅCO to nonꢁ
volatile productsð%Þ
Reaction selectivity
towards desired
product ð%Þ
Radiochemical
yield
¼
x
(1)
The obtained radiochemical yield was similar to the 67%
reported by Kealey et al., who utilised a complexation addi-
tive in the synthesis of the same compound.9 In a second
comparison, the yield was identical with that reported by
Karimi et al. for the analogous compound [11C-carbonyl]N-
benzyl-4-methoxybenzamide, which was synthesised by use
of the high pressure method.20
PPh3
Pd2(dba)3
TMPAH
O
I
*
O
+
+
CH3OH
*
CO
THF
160°C
5 min
N
N
Ureas labelled in the carbonyl position can be efficiently
synthesised by 11C-carbonylation of organoazides to form
isocyanate intermediates, as proposed by Doi et al., which
Scheme 3. Synthesis of [11C-carbonyl]methyl nicotinate. Position of the 11C label = *.
Copyright © 2012 John Wiley & Sons, Ltd. J. Label Compd. Radiopharm 2012, 55 223–228
www.jlcr.org

J. Eriksson et al.
Although not optimised, the radiochemical yields in the three
model reactions proved that the method to confine [11C]CO in
disposable glass vials at low pressure was applicable for use in
11C-carbonylation reactions. The high conversion of [11C]CO
showed that the reactivity was seemingly unhampered by
the presence of xenon and by the equilibrium distribution of
[11C]CO in the gas phase and the THF solution. Hence, the
method can be expected to give similar reaction efficiency as
the high pressure method in which the gas phase was removed
by filling the autoclave with solvent at high pressure. The
experiments also showed that structural support to the septum
enabled the use of disposable glass vials at reaction
temperatures considerably higher than the boiling point of
the solvent.23
Figure 4. Plot showing the pressure increase when adding transfer gas to a selection
of solvents (0.5 mL) contained in capped vials (1mL). The gases were fed by a mass
flow controller set at 1.5sccm. The experiment was repeated twice; the average value
is presented. To be noted, compared with the set-up used for the model reactions, the
dead volume in the set-up was significantly enlarged by the addition of a mass flow
controller, tubing and a pressure sensor hence resulting in a slower pressure increase.
Xenon solubility in various solvents
A series of solvents were tested to assess the utility of using
xenon as transfer gas with other solvents than THF. [11C]CO
was omitted in these experiments. Instead, the pressure in the
vial was monitored while adding xenon or helium. The pressure
sensor was connected to a coarse needle inserted into the
septum. A mass flow controller set at 1.5 sccm (standard cubic
centimetres/min) was used to provide reproducible data. The
mass flows of helium and xenon were confirmed to be equal by
comparing the pressure profile when adding the gases separately
into empty vials. The pressure profiles obtained with the different
solvents are presented in Figure 4. The data, which was intended
for qualitative comparison of the pressure build-up between the
different solvents, showed a trend that corresponded reasonably
well with the literature data on the solubility of the transfer gases,
Table 1. As expected, the pressure increase was moderate when
adding xenon to THF whereas water and some of the polar
organic solvents provided lower solubility and hence faster pres-
sure increase. The poor solubility of helium in THF was confirmed
by the pressure profile, which matched confinement of helium in
the gas phase volume only, as verified with control experiments
made with an empty vial.
Table 1. Ostwald solubilities
Ostwald
Solubility system
(solute in solvent)
solubility
Temperature
(ꢂC)
(L)
Reference
19
24
25
26
25
27
28
27
19
He in THF
Xe in water
0.024*
0.118
1.61*
2.20
2.41*
2.42
3.08
3.59
3.87*
25
20
25
20
25
20
25
20
25
Xe in acetonitrile
Xe in methanol
Xe in acetone
Xe in dioxane
Xe in benzene
Xe in toluene
Xe in THF
25
19
*Data from reference
and
were converted to Ostwald
solubility coefficients (L), defined as the ratio between the
concentration of gas molecules in the solvent and the
concentration of gas molecules in the gas phase, at equilibrium.
the aluminium cap was used to support the septum, as the
reaction was performed at 160ꢂC because of the poor
Proposed design of a simplified [11C]CO dispensing module
nucleophilicity of the alcohol. The [11C]CO conversion was During this work, we choose to use gas purification to enable
68 ꢀ 6% and the selectivity towards formation of the labelled zinc reduction of [11C]CO2 to [11C]CO, as this was a part of an
methyl ester was 42 ꢀ 11%, giving a radiochemical yield of ongoing investigation. However, it should be noted that a
29 ꢀ 10% (n = 3).
more simplified set-up can be designed at facilities where
Figure 5. Proposed simplified design of a [11C]CO dispensing module.
J. Label Compd. Radiopharm 2012, 55 223–228
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

J. Eriksson et al.
[2] H. Doi, J. Barletta, M. Suzuki, R. Noyori, Y. Watanabe, B. Långström,
Org. Biomol. Chem. 2004, 2, 3063–3066.
[3] J. Barletta, F. Karimi, B. Långström, J. Labelled Comp. Rad. 2006, 49,
429–436.
[4] O. Åberg, B. Långström, J. Labelled Comp. Rad. 2011, 54, 38–42.
[5] J. Eriksson, O. Åberg, B. Långström, Eur. J. Org. Chem. 2007, 2007,
455–461.
zinc reduction proceeds without gas purification or where
molybdenum29 is successfully used for the reduction, Figure 5.
Conclusions
By the use of pre-concentration and transfer in xenon gas, it was
possible to confine [11C]CO in a standard glass vial containing
reagent solution, without the use of high pressure or complexing
additives. The [11C]CO was utilised in carbonylation reactions
where 11C-labelled amide, urea and methyl ester were obtained
in yields comparable with those previously reported using high
pressure or complexing additives. From a technical perspective, this
method greatly simplifies implementation of 11C-carbonylation
compared with the high pressure method. From a synthetic
perspective, the method allows for careful preparation of reagent
solutions before adding [11C]CO to the same vial, hence opening
doors for exploration of sensitive carbonylation reagents and for
multi-step one-pot procedures. The possibility to use disposable
reaction vials eliminates carry over issues and reduces the need
for cleaning validation in a GMP setting. Ultimately, this method
may demonstrate the potential to perform carbonylation with
[11C]CO in routine radiopharmaceutical synthesis with similar ease
as methylation with [11C]methyl iodide.
[6] J. Eriksson, G. Antoni, B. Långström, J. Labelled Comp. Rad. 2006, 49,
1105–1116.
[7] T. Kihlberg, B. Långström, J. Org. Chem. 1999, 64, 9201–9205.
[8] B. Långström, O. Itsenko, O. Rahman, J. Labelled Comp. Rad. 2007, 50,
794–810.
[9] S. Kealey, P. W. Miller, N. J. Long, C. Plisson, L. Martarello, A. D. Gee,
Chem. Commun. 2009, 3696–3698. DOI: 10.1039/B906166E.
[10] S. Lu, J. Hong, T. Itoh, M. Fujita, O. Inoue, R. B. Innis, V. W. Pike, J.
Labelled Comp. Rad. 2010, 53, 548–551.
[11] P. Lidström, T. Kihlberg, B. Långström, J. Chem. Soc., Perkin Trans. 1
1997, 2701–2706. DOI: 10.1039/A703062B.
[12] M. H. Al-Qahtani, V. W. Pike, J. Labelled Comp. Rad. 2000, 43,
825–835.
[13] T. Kihlberg, B. Långström, T. Ferm, J. Eriksson, WO/2006/008,603
[14] E. D. Hostetler, H. D. Burns, Nucl. Med. Biol. 2002, 29, 845–848.
[15] H. Audrain, L. Martarello, A. Gee, D. Bender, Chem. Commun. 2004,
558–559. DOI: 10.1039/B314982J.
[16] P. W. Miller, N. J. Long, A. J. de Mello, R. Vilar, H. Audrain, D.
Bender, J. Passchier, A. Gee, Angew. Chem. Int. Ed. 2007, 46,
2875–2878.
[17] P. W. Miller, H. Audrain, D. Bender, A. J. de Mello, A. D. Gee, N. J. Long,
R. Vilar, Chem. Eur. J. 2011, 17, 460–463.
[18] M. J. Welch, M. M. Ter-Pogossian, Radiat. Res. 1968, 36, 580–587.
[19] F. Gibanel, M. C. Lopez, F. M. Royo, J. Santafe, J. S. Urieta, J. Solution.
Chem. 1993, 22, 211–217.
[20] F. Karimi, B. Långström, Eur. J. Org. Chem. 2003, 2003, 2132–2137.
[21] O. Ilovich, O. Åberg, B. Långström, E. Mishani, J. Labelled Comp. Rad.
2009, 52, 151–157.
[22] R. U. Halden, D. H. Paull, Environ. Sci. Tech. 2005, 39, 1420–1426.
[23] M. Damm, C. O. Kappe, J. Combin. Chem. 2009, 11, 460–468.
[24] G. L. Pollack, Science 1991, 251, 1323.
[25] N. Segebarth, L. Aïtjeddig, E. Locci, K. Bartik, M. Luhmer, J. Phys.
Chem. A 2006, 110, 10770–10776.
Acknowledgements
The authors would like to acknowledge F.L. Buijs, A. de Nooij,
P.E. Schollema, W. Verhoeff and L.F. van Rooij for the assistance
when building the experimental set-up and BV Cyclotron VU
for providing [11C]CO2.
Conflict of Interest
The authors did not report any conflict of interest.
[26] G. L. Pollack, J. F. Himm, J. J. Enyeart, J. Chem. Phys. 1984, 81, 3239.
[27] F. Cejnar, L. Wilhelmova, P. Vrzala, Int. J. Appl. Radiat. Isot. 1977, 28,
281–284.
References
[28] L. H. Clever, J. Phys. Chem. 1958, 62, 375–376.
[1] A. Brennführer, H. Neumann, M. Beller, Angew. Chem. Int. Ed. 2009, [29] S. K. Zeisler, M. Nader, A. Theobald, F. Oberdorfer, Appl. Radiat. Isot.
48, 4114–4133. 1997, 48, 1091–1095.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012, 55 223–228