Microfluidic reactions using [11C]carbon monoxide solutions for the synthesis of a positron emission tomography radiotracer

Article abstract of DOI:10.1039/c0ob00631a

Microfluidic technology has been used to perform [¹¹C] carbonylation reactions using solutions containing [¹¹C]CO in the form of the complex, copper(i)tris(3,5-dimethylpyrazolyl)borate-[ ¹¹C]carbonyl (Cu(Tp*)[¹¹C]CO). The synthesis of the model compound [¹¹C]N-benzylbenzamide and the known tracer molecule [¹¹C]trans-N-[5-(2-flurophenyl)-2-pyrimidinyl]-3-oxospiro[5- azaisobenzofurane-1(3H),1′-cyclohexane]-4′-carboxamide ([ ¹¹C]MK-0233), a ligand for the neuropeptide Y Y5 receptor, have been performed using this technique. Following semi-preparative HPLC purification and reformulation, 1262 ± 113 MBq of [¹¹C]MK-0233 was produced at the end of the synthesis with a specific activity of 100 ± 30 GBq μmol^-1 and a >99% radiochemical purity. This corresponds to a decay corrected radiochemical yield of 7.2 ± 0.7%. Using a 3 mL vial as the reaction vessel, and following semi-preparative HPLC purification and reformulation, 1255 ± 392 MBq of [¹¹C]MK-0233 was produced at the end of the synthesis with a specific activity of 100 ± 15 GBq μmol^-1 and a >99% radiochemical purity. This corresponds to a decay corrected radiochemical yield of 7.1 ± 2.2%.

Full text of DOI:10.1039/c0ob00631a

View Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 3313
www.rsc.org/obc
PAPER
Microfluidic reactions using [11C]carbon monoxide solutions for the synthesis
of a positron emission tomography radiotracer†
Steven Kealey,a,b Christophe Plisson,a T. Lee Collier,c Nicholas J. Long,d Stephen M. Husbands,b
Laurent Martarelloa and Antony D. Geee
Received 26th August 2010, Accepted 6th January 2011
DOI: 10.1039/c0ob00631a
Microfluidic technology has been used to perform [11C]carbonylation reactions using solutions
containing [11C]CO in the form of the complex, copper(I)tris(3,5-dimethylpyrazolyl)borate-
[11C]carbonyl (Cu(Tp*)[11C]CO). The synthesis of the model compound [11C]N-benzylbenzamide and
the known tracer molecule [11C]trans-N-[5-(2-flurophenyl)-2-pyrimidinyl]-3-oxospiro[5-
azaisobenzofurane-1(3H),1¢-cyclohexane]-4¢-carboxamide ([11C]MK-0233), a ligand for the
neuropeptide Y Y5 receptor, have been performed using this technique. Following semi-preparative
HPLC purification and reformulation, 1262 113 MBq of [11C]MK-0233 was produced at the end of
the synthesis with a specific activity of 100 30 GBq mmol-1 and a >99% radiochemical purity. This
corresponds to a decay corrected radiochemical yield of 7.2 0.7%. Using a 3 mL vial as the reaction
vessel, and following semi-preparative HPLC purification and reformulation, 1255 392 MBq of
[11C]MK-0233 was produced at the end of the synthesis with a specific activity of 100 15 GBq mmol-1
and a >99% radiochemical purity. This corresponds to a decay corrected radiochemical yield of 7.1
2.2%.
complexation to CO-binding molecules in solution. This was
Introduction
first demonstrated using a solution of BH3·THF to form a
BH3·[11C]CO adduct at low temperatures,5 and most recently
Positron emission tomography (PET) is a powerful non-invasive
imaging technique that uses radiolabelled compounds as molec-
with a copper(I) tris(pyrazolyl)borate complex to form a copper
ular probes to image biological processes in vivo.1 [11C]Carbon
monoxide is an important 11C-labelling reagent for radiotracer
synthesis as it can provide access to a range of 11C-labelled
molecules through metal-catalysed carbonylation reactions.1,2 The
low solubility of this gas and its delivery at high dilution, as well
as the time restrictions imposed by the short radioactive half-
life of carbon-11 (20 min), has restricted the widespread use
of this radiolabelling reagent. In the past 15 years, significant
advances have been made in this field through the use of micro-
autoclave reactors to perform carbonylation reactions at high
pressures.3,4 We are interested in techniques that enhance the
reactivity of [11C]CO by increasing its solubility through chemical
[11C]carbonyl complex (Cu(Tp*)[11C]CO) at room temperature
(Scheme 1a).6 Both of these systems were found to trap [11C]CO
efficiently (>90%) and release the gas controllably, by heating in
the case of BH3·[11C]CO or by the addition of triphenylphos-
phine in the case of Cu(Tp*)[11C]CO. This led to their success-
ful application as [11C]CO-sources in low-pressure, palladium-
mediated [11C]carbonylation reactions to form the test compound
[11C]N-benzylbenzamide and side-product [11C]dibenzylurea
(Scheme 1b).
aGlaxoSmithKline Clinical Imaging Centre, Hammersmith Hospital, Lon-
don, W12 0NN, UK
bDepartment of Pharmacy and Pharmacology, University of Bath, Bath,
BA2 7AY, UK
cAdvion BioSystems, Inc., Division of Imaging Sciences, Ithaca, NY, 14850,
USA
dDepartment of Chemistry, Imperial College London, London, SW7 2AZ,
UK
eKing’s College London, Rayne Institute, St. Thomas’ Hospital, London,
SE1 7EH, UK. E-mail: antony.gee@kcl.ac.uk
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c0ob00631a
Scheme 1 (a) [11C]CO trapping procedure using BH3·THF (left) and a
copper(I) complex (right). (b) Synthesis of [11C]N-benzylbenzamide and
side-product [11C]dibenzylurea using a [11C]CO solution.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 3313–3319 | 3313
©

View Online
Fig. 1 Schematic of the microfluidic apparatus (HPLC purification apparatus used for the synthesis of [11C]MK-0233).
flow microfluidic device, the Advion NanoTekꢀ,18 was chosen for
these experiments. This integrated device contains syringe pumps
that are connected to a microreactor—a coiled silica capillary—
which is housed in a metal casing and seated in a heater block. The
device was placed inside the hot-cell and connected to a robotic
valve setup (Fig. 1).
R
The ability to handle this useful reagent in solution opens up
a new radiolabelling technique in which solutions of [11C]CO
may be used in continuous flow microfluidic reactions. The use
of microfluidics for radiotracer synthesis has begun to attract
considerable attention in recent years,7–16 as this technology is well-
placed to overcome some of the constraints associated with PET
radiochemistry, including the use of sub-micromolar quantities of
radiolabelling reagents, the requirement for short reaction times
and the need for an automated, reproducible process.
Microfluidic reactions for the synthesis of test compound
[11C]N-benzylbenzamide
While a macroscopic tube reactor has previously been reported
for performing continuous flow [11C]carbonylation reactions,17
microfluidics has not made a significant impact on [11C]CO chem-
istry, despite its enormous potential. This may be a result of the
restrictions caused by the low concentration and solubility of this
gas when using conventional techniques, as well as the mechanistic
difficulties of controlling the [11C]CO/carrier gas flow. The recent
emergence of efficient [11C]CO trapping systems therefore provides
an excellent opportunity to open up the field of [11C]CO chemistry
to microfluidics. It was postulated that microfluidics would be
especially suited to performing [11C]carbonylation reactions, since
the [11C]CO gas that is released from the complex will be
constrained within the channels of the microreactor, thus keeping
it in contact with the reagents. This could be advantageous over
conventional vial-based reactions, where [11C]CO gas is likely to
escape to the headspace of the vessel, thus reducing its interactions
with the substrate in solution.
A significant feature of microfluidics is the ability to perform
reactions using small quantities of reagents. In radiolabelling
chemistry, this could be exploited to allow multiple reactions
to be performed from a single production of radioisotope. In
order to examine the feasibility of this process for microfluidic
[11C]carbonylation reactions, the palladium-mediated coupling of
iodobenzene and benzylamine (Scheme 1b) was chosen as a model
reaction. Having been performed previously in glass vials,6 this
reaction has been shown to yield [11C]N-benzylbenzamide in good
yield in the presence of a palladium(0) catalyst, while the use
of a palladium(II) catalyst results in significant formation of a by-
product, [11C]dibenzylurea. By performing multiple reactions with
microfluidics, it was decided to examine the effect of microreactor
temperature on the [11C]N-benzylbenzamide/[11C]dibenzylurea
product distribution formed in the presence of the palladium(II)
species Pd(dppp)Cl2.
The process began with [11C]CO trapping to form the
Cu(Tp*)[11C]CO solution. This solution was then aspirated into
Results and discussion
R
syringe 1 of the NanoTekꢀ, while a separate solution contain-
In order to test the suitability of the microfluidic approach
to solution-based [11C]carbonylation chemistry, the synthesis of
the test compound [11C]N-benzylbenzamide (and side-product
[11C]dibenzylurea) was initially explored using Cu(Tp*)[11C]CO
solutions. The ease of formation of Cu(Tp*)[11C]CO (trapping
occurs at room temperature in a one-pot process with a high
trapping efficiency) made this reagent the obvious candidate for
performing these reactions. A commercially-available continuous
ing the cross-coupling reagents (triphenylphosphine, palladium
catalyst, benzylamine, iodobenzene) was aspirated into syringe 2.
To perform a reaction, aliquots of both solutions were dispensed
into the microreactor and the crude product that flowed out
was collected in a vial for analysis. The radiolabelled product
distribution was measured by radioanalytical HPLC to give the
radiochemical purity of each component in the mixture.19 100
mL aliquots of the Cu(Tp*)[11C]CO solution were used for each
3314 | Org. Biomol. Chem., 2011, 9, 3313–3319
This journal is
The Royal Society of Chemistry 2011
©

View Online
Table 1 Results of microfluidic [11C]carbonylation reactions for the
synthesis of [11C]N-benzylbenzamide and [11C]dibenzylurea
Microfluidic reactions the for synthesis of [11C]MK-0233
Since the ultimate goal is to apply this technology to the clinical
production of PET radiotracers, it was decided to examine the
suitability of the microfluidic system towards the synthesis of a
biologically-active tracer molecule. The neuropeptide Y Y5 recep-
tor antagonist [11C]MK-0233, previously synthesised by Burns et
al.32 and used to examine the role of this receptor in weight loss,33
was chosen for these studies. This labelling reaction proceeds via
a palladium(0)-mediated ring-closing [11C]carbonylation reaction
between pyridyl-bromide and hydroxyl functional groups of the
starting material, 1, to form a 11C-labelled lactone, [11C]MK-0233
(Scheme 3).
Temp./◦C
RCP [11C]ureaa
RCP [11C]amidea
100
125
150
175
200
73
73
52
32
22
2
6
3
6
1
12
19
44
61
69
6
4
3
5
1
a Average of 2 runs ( standard deviation. Flow rate = 20 mL min-1 per
syringe.
experiment, allowing up to four reactions to be performed from
a single 400 mL batch of Cu(Tp*)[11C]CO. Three productions of
Cu(Tp*)[11C]CO enabled a range of temperatures between 100 and
◦
200 C to be probed with a repeat run at each temperature, the
results of which are summarised in Table 1.
Microfluidics was indeed found to be suitable for performing
[11C]carbonylation reactions, producing the expected 11C-labelled
products in a distribution that appeared to be temperature
dependent. High microreactor temperatures (>150 ◦C) were
found to favour [11C◦]N-benzylbenzamide formation, while lower
temperatures (<150 C) produced [11C]dibenzylurea as the major
product. This phenomenon is believed to arise from the two
distinct catalytic cycles that drive these processes (Scheme 2) and
the effect that the palladium oxidation state has on which path is
favoured. Amide formation is known to occur via a palladium(0)-
mediated process,20 while urea formation is thought, in this case,
to occur via a palladium(II)-mediated oxidative carbonylation
process.21–31 As the starting material used is a palladium(II) species,
at lower temperatures, it is expected that this species will remain
largely unreduced and therefore act as the starting point for
the oxidative carbonylation reaction to yield the 11C-labelled
urea (Scheme 2, RHS). At higher temperatures, reduction of
the palladium(II) by excess phosphine and/or amine may be
accelerated, thus yielding a palladium(0) species that can catalyse
the aminocarbonylation reaction to form the 11C-labelled amide
(Scheme 2, LHS). Since these experiments involve nanomolar
quantities of [11C]CO, any palladium species will be in relative ex-
cess to [11C]CO and, as such, regeneration of the palladium would
not be required in order to convert all of the [11C]CO to 11C-labelled
products.
Scheme
3 a palladium-mediated
Synthesis of [11C]MK-0233 via
[11C]carbonylation reaction following a copper-mediated [11C]CO trapping
process.
Before performing a complete synthesis and purification, the op-
timum reaction conditions were first sought by performing small-
scale reactions using a single batch of Cu(Tp*)[11C]CO. As for the
test reaction, this process began with [11C]CO trapping to form
the Cu(Tp*)[11C]CO solution, which was aspirated into syringe
1. This solution was then dispensed into the microreactor along
with the cross-coupling reagents (triphenylphosphine, palladium
catalyst, 1) from syringe 2 (Fig. 1).
In keeping with the conditions reported by Burns et al.,32
the palladium(0) catalyst Pd(PPh3)4 was used for these studies.
Optimisation reactions were carried out using 30 mL aliquots of
the Cu(Tp*)[11C]CO and cross-coupling reagent solutions to probe
the effects of temperature and flow rate on the radiochemical
purity of the crude (unpurified) product, as measured by analytical
HPLC. These reactions were performed at a range of temperatures
at a fixed arbitrary flow rate of 30 mL min-1 delivered from
each syringe (Table 2, entries 1–6). This showed that negligible
product formation was observed below 130 ◦C, leaving most of the
Cu(Tp*)[11C]CO complex unreacted, while a significant increase
in RCP was observed upon increasing the temperature to 150 ◦C.
Further raising the temperature at 10 ◦C intervals revealed that
the yield reached a maximum value of 81% at 160 ◦C, which was
not improved by further raising the temperature. The combined
flow rate of 60 mL min-1 used in these experiments corresponds to
a residence time of a quarter of a minute for the reagents within
the microreactor. Next, the impact of flow rate (and therefore
residence time of the reagents within the microreactor) on yield
was examined. At a constant temperature of 180 ◦C, experiments
were performed at flow rates of 50, 100 and 150 mL min-1 (Table 2,
entries 7–9). This showed that the greater the flow rate, the lower
Scheme 2 Competing catalytic cycles in the formation of 11C-labelled
amide and 11C-labelled urea using palladium catalysts in the presence of
amine and aryl halide. Ln = 1,3-bis(diphenylphosphino)propane.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 3313–3319 | 3315
©

View Online
with a specific activity of 100 30 GBq mmol-1 and a >99%
radiochemical purity. The total preparation time from the end of
cyclotron bombardment to receiving the reformulated dose was 27
min. This corresponds to a decay-corrected radiochemical yield of
7.2 0.7%, calculated from the total amount of [11C]CO delivered
to the copper(I) trapping solution.
Table 2 Effect of temperature and flow rate on the radiochemical purity
of [11C]MK-0233
Entry Temp./◦C Flow rate/mL min-1 Residence time/sa RCP (%)
1
2
3
4
5
6
7
8
9
120
130
150
160
170
180
180
180
180
30
30
30
30
30
30
50
100
150
31
31
31
31
31
31
19
9
0
9
67
81
79
79
76
73
72
Vial reactions for the synthesis of [11C]MK-0233
While low pressure [11C]carbonylation reactions using
Cu(Tp*)[11C]CO solutions have previously been employed
for the synthesis of the test compound, this method has not
been demonstrated for the synthesis of biologically-active tracer
molecules—for potential use in PET imaging studies. In order
to examine the feasibility of this process and for comparison
with the microfluidic method, the synthesis of [11C]MK-0233 was
performed using this low pressure technique.
The reactions were performed using a 3 mL glass vial fitted with
a rubber septum cap and pierced with needles to allow the transfer
of reagents into and out of the vessel. An array of three-way
valves connected by PTFE tubing were used to control the flow
of gaseous and liquid reagents in this process (Fig. 2). As before,
the apparatus was placed in a lead-shielded fume hood (hot-cell)
and controlled externally by a computer. In these experiments
the [11C]CO trapping process was performed initially to give a
solution of Cu(Tp*)[11C]CO, to which the cross-coupling reagent
solution (PPh3, Pd(PPh3)4, 1) was added using an argon gas sweep.
Following this, the vial was sealed and heated to 150 ◦C for 5 min.
Unlike the microfluidic method, repeated small-scale reactions
using one production of [11C]CO could not easily be performed for
the optimisation process. Instead, a series of [11C]CO productions
were used to perform test reactions, and initially these suffered
difficulties due to radioactive gaseous releases from the apparatus,
6
a 100 mm ¥ 2 m microreactor, volume = 16 mL.
the yield, although the deterioration was only slight, suggesting
that, under these microfluidic conditions, the reaction proceeds so
quickly that reducing the residence time to the order of seconds
does not prevent the reaction from occurring.
For the subsequent full-scale synthesis, aiming to produce
[11C]MK-0233 for potential use in PET experiments, the cyclotron
bombardment time was increased to a ‘full bombardment’, giving
approximately 37 GBq of [11C]CO after the reduction of [11C]CO2.
Following the optimisation process, 400 mL of the Cu(Tp*)[11C]CO
solution was used and, as a compromise between maximising
yields and minimising synthesis time, reactions were performed at
180 ◦C with a flow rate of 100 mL min-1. Following the microfluidic
reaction, the solution was purified by semi-preparative HPLC (Fig.
1, RHS) and then reformulated to give the product in a potentially
injectable ethanol–saline solution. The radiochemical purity and
specific activity of the product was then measured by analytical
HPLC. Using this technique, over three consecutive runs, 1262
113 MBq of [11C]MK-0233 was produced at end of the synthesis
Fig. 2 Valve setup for the reaction and purification of [11C]MK-0233 using the vial technique.
3316 | Org. Biomol. Chem., 2011, 9, 3313–3319 This journal is The Royal Society of Chemistry 2011
©

View Online
◦
especially when heating at temperatures in excess of 150 C due
solution-based carbonylations, we have found the technique to be
suitable for tracer production, with the potential to outperform
the vial method. Microfluidics was found to be advantageous
when using a volatile, radioactive reagent such as [11C]CO as
unwanted radioactive emissions were not observed, most likely
because during the heating process, the gases are held within
the confines of the reactor. Vial reactions, however, are prone
to occasional and unpredictable releases of [11C]CO into the hot
cell as a consequence of elevated pressures within the vial during
heating, potentially lowering the radiochemical yield. As such,
the upper temperature for vial reactions is limited to that of the
boiling point of the solvent, while the increased pressure tolerance
of the microfluidic setup allowed temperatures above the normal
boiling point of the solvent to be used. The automation that can
be achieved with microfluidic devices means that the apparatus
can be easily cleaned between runs using a pre-programmed
macro. Microfluidic technology carries the significant advantage
of allowing reaction optimisation to be performed rapidly using
sub-milligram quantities of reagents and often using only one
production of [11C]CO. The optimisation process for vial reactions,
on the other hand, were more time consuming, requiring multiple
productions of [11C]CO to test a range of temperatures and
alterations in the valve setup.
to pressurisation of the solvents. These releases were attributed
to leaks from the three-way valves that were initially used to
seal the reaction vial during heating. To avoid this problem, a
Rheodyneꢀ HPLC valve and a compressed air-activated two-way
R
valve were used to seal the vial, and these proved to be leak-tight.
To further minimise the release of [11C]CO from the reaction vial,
two rubber septa were placed together in the cap to seal the vial,
and repeated puncturing of the septa with needles was avoided. At
◦
150 C, we found that [11C]CO release was minimised, while the
reaction proceeded cleanly to give crude radiochemical purities of
>90%. The overall radiochemical yields were not measured at this
stage due to variations between runs in the amount of [11C]CO2
produced.
As for the microfluidic process, after optimisation was complete,
the synthesis of [11C]MK-0233 was performed on a larger scale
using full cyclotron bombardment and purification by semi-
preparative HPLC, and reformulation to give the product as a
potentially injectable ethanol–saline solution. The radiochemical
purity and specific activity was then measured by analytical HPLC.
Using this technique, over three consecutive runs, 1255 392
MBq of [11C]MK-0233 was obtained at end of synthesis with a
specific activity of 100 15 GBq mmol-1 and a radiochemical
purity >99%. The total preparation time from the end of cyclotron
bombardment to receiving the reformulated dose was 27 min.
This corresponds to a decay-corrected radiochemical yield of 7.1
2.2%, calculated from the total amount of [11C]CO delivered to
the copper(I) trapping solution.
Conclusion
A copper(I)[11C]carbonyl complex has been successfully used to
perform [11C]carbonylation reactions using microfluidic technol-
ogy or conventional glass reaction vials. Both approaches gave
yields on a scale suitable for performing in vivo PET studies,
despite being unoptimised in terms of catalyst/ligand choice.
Despite the longer optimisation process associated with the vial
technique, reasonable radiochemical yields were obtained for the
tracer compound, indicating that this low-cost approach provides
a viable method for performing [11C]carbonylation reactions for
clinical tracer production.
Microfluidic technology was found to be well-suited to perform-
ing solution-[11C]CO chemistry and enabled reaction optimisation
to be performed rapidly using small quantities of reagents. As well
as performing these test reactions, the process was successfully
scaled-up, as a proof-of-concept, to produce the tracer molecule
on a scale suitable for PET studies. We believe that the combination
of [11C]CO solutions with microfluidics can play an important
role in the future radiolabelling of PET tracers. One such role is
the development of more potent and selective tracer molecules,
where pseudo-combinatorial syntheses could be performed to
label a series of different precursors using a single batch of
Cu(Tp*)[11C]CO. A mini-library of related radiotracers could be
synthesised in this manner and studied in parallel, for example
by tissue section autoradiography on adjacent tissue sections,
thus enabling rapid screening of radioligands to probe structure–
activity relationships. These studies are currently being performed
in our laboratories and will be reported in due course.
Comparison of methods
The radiochemical yields obtained through the microfluidic
method and the vial method were found to be almost identical,
with both processes taking the same amount of time (from the
delivery of [11C]CO2 to receiving the dose of purified [11C]MK-
0233). Despite our predictions, a greater yield was not achieved
through microfluidics, which could result from a number of factors.
Firstly, in the microfluidic experiments, in order to avoid aspirating
air into the system, 700 mL of trapping solution was used, while
only 400 mL of this was used for radiolabelling. We believe
this wastage could be all but eliminated by further reducing the
trapping volume, which would be expected to significantly improve
radiochemical yields. A second factor that is likely to contribute
to the lowering of yields is the use of high flow rates in the
microfluidic process in order to allow the full bolus of reactants
to pass through the reactor in a timeframe similar to the heating
step in corresponding batch reactions. Microfluidic flow rates of
100 mL per syringe correspond to a residence time of the reagents
within the microreactor of under 10 s, and this is likely to result in
incomplete conversion of [11C]CO to radiolabelled product. It is
thought that increasing the reactor length would improve yields by
providing an increase in reagent residence time without sacrificing
reaction time (as would occur if the flow rate was lowered instead).
This will be investigated in future studies.
Experimental section
General procedures
The previous report describing the synthesis of [11C]MK-0233,32
performed at high pressure using a micro-autoclave reactor, does
not state the yield or specific activity of the product, so no
comparison with our methods can be made. For the purposes
of our investigation, to examine the viability of microfluidics for
All chemicals and solvents were purchased from commercial
sources, except for trans-4-(3-bromopyridin-4-yl)-4-hydroxy-N-
[5-(2-fluorophenyl)-2-pyrimidinyl]-cyclohexanecarboxamide (1),
which was synthesised using the procedure described by Burns
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 3313–3319 | 3317
©

View Online
et al.32 Semi-preparative HPLC purification was performed using
a reversed-phase column (Agilent Zorbax SB-C18, 5 mm, 9.4 ¥
250 mm) using a mobile phase of MeCN + 0.1% trifluoroacetic
acid and water using an initial gradient (0–2 min from 5–40%
MeCN) followed by an isocratic 40% MeCN at a flow rate of 8
mL min-1. UV (254 nm) and radioactivity (pin diode detector)
signals were recorded using Chromeleon software. An unlabelled
reference sample of MK-0233 was used to determine the retention
time of this compound (9.3 min). QC of the final reformulated
dose was performed by analytical HPLC using a reversed phase
column (Agilent Zorbax SB-Phenyl, 3.5 mm, 4.6 ¥ 150 mm) and a
60 : 40 ammonium formate (50 mM, pH 4) : MeCN mobile phase
at a flow rate of 1.5 mL min-1 (retention time = 5.6 min).
Microfluidic reactions the for synthesis of [11C]MK-0233
The cross-coupling reagent solution was prepared by the addition
of Pd(PPh3)4 (3.0 mg, 3 mmol), PPh3 (4.2 mg, 16 mmol) and 1 (1.0
mg, 2 mmol) into a 1 mL vial, flushing with nitrogen for 10 min,
followed by the addition of anhydrous dimethylformamide (0.5
mL). This solution was transferred to a 400 mL loop by aspirating
R
syringe 1 of the NanoTekꢀ. The trapping solution was prepared
by the addition of CuCl (1.1 mg, 11 mmol) and K[Tp*] (3.7 mg,
11 mmol) to a 3 mL V-bottomed glass vial, flushing with nitrogen
for 10 min, followed by the addition of anhydrous MeCN (0.7
mL). The solution was then passed through a 0.45 mm filter into
another 3 mL V-bottomed glass vial under a nitrogen atmosphere.
[11C]CO/He was bubbled through the trapping solution and the
radioactivity of the vial monitored using a non-calibrated pin
diode detector. Following complete delivery of the radioactivity
to the vial, corresponding to the formation of Cu(Tp*)[11C]CO,
the solution was transferred into a 400 mL loop by aspirating
[11C]Carbon dioxide was produced using a Siemens Eclipse
HP cyclotron by 11 MeV proton bombardment (45 min at 55
mA) of a target containing nitrogen and 1% oxygen. [11C]CO
was produced using an Eckert and Ziegler reduction module
using the following procedure: [11C]CO2 was delivered from the
cyclotron and trapped at room temperature in a stainless steel
loop containing molecular sieves. The [11C]CO2 was then released
by passing a helium flow through the loop while heating to 400
◦C. The resultant [11C]CO2/He gas stream was passed through a
glass tube packed with molybdenum powder at 850 ◦C, converting
the [11C]CO2 to [11C]CO, with any unreduced [11C]CO2 being
trapped using Ascarite. The resultant [11C]CO/He gas stream
was delivered to the trapping solution at a flow rate of 20 mL
min-1. The time taken from the end of cyclotron bombardment
to complete delivery of [11C]CO to the trapping vial was 5 to
6 min.
R
syringe 2 of the NanoTekꢀ. The cross-coupling reagent solution
(400 mL) and the [11C]CO solution (400 mL) were then dispensed
simultaneously at a flow rate of 100 mL min-1 per syringe int◦o a
100 mm ¥ 2 m microreactor situated in a heater block at 180 C.
The mixture leaving the microreactor was collected in a vial and
injected into the HPLC for semi-preparative purification. Upon
observation of the radioactivity peak at the expected retention
time, the flow was diverted into a vial containing a solution of
K2CO3 (100 mg) in water (30 mL). This solution was then passed
through a C18 Sep-Pak cartridge to trap the product, [11C]MK-
0233, and washed with water (6 mL). The product was eluted from
the Sep-Pak cartridge into a 10 mL dose vial using EtOH (2 mL)
an then saline solution (8 mL). The radioactivity of the dose vial
was measured in a dose calibrator, and the radiochemical purity
and specific activity of the product were measured by analytical
HPLC.
Microfluidic reactions for the synthesis of test compound
[11C]N-benzylbenzamide
The cross-coupling reagent solution was prepared by the addition
of Pd(dppp)Cl2 (1.3 mg, 2 mmol), PPh3 (5.8 mg, 22 mmol),
iodobenzene (2.3 mg, 11 mmol) and benzylamine (0.1 mL) into
a 1 mL vial, flushing with nitrogen for 10 min, followed by the
addition of anhydrous dimethylformamide (0.6 mL). This solution
was transferred to a 400 mL loop by aspirating syringe 1 of the
Vial reaction for the synthesis of [11C]MK-0233
The trapping solution was prepared by the addition of CuCl (1.1
mg, 11 mmol) and K[Tp*] (3.7 mg, 11 mmol) into a 3 mL V-
bottomed glass vial, flushing with nitrogen for 10 min, followed
by the addition of anhydrous THF (1.1 mL). The carbonylation
reagents were prepared by the addition of Pd(PPh3)4 (2.4 mg,
2 mmol), PPh3 (5.8 mg, 22 mmol) and 1 (1.0 mg, 2 mmol) to a
second vial, flushing it with nitrogen for 10 min, followed by the
addition of anhydrous dimethylformamide (0.5 mL). [11C]CO/He
was bubbled through the trapping solution and the radioactivity
of the vial monitored using a non-calibrated pin diode detector.
Following complete delivery of the radioactivity to the vial, the
carbonylation reagents were added using an argon gas push. The
vial was subsequently sealed and heated at 150 ◦C in a heating
block for 6 min. The reaction mixture was then transferred to the
quench vial using a positive pressure of argon gas and injected into
an HPLC for semi-preparative purification. Upon observation of
the radioactivity peak at the expected retention time, the flow was
diverted into a vial containing a solution of K2CO3 (100 mg) in
water (30 mL). This solution was then passed through a C18 Sep-
Pak cartridge to trap the product, [11C]MK-0233, and washed with
water (6 mL). The product was eluted from the Sep-Pak cartridge
into a 10 mL dose vial using EtOH (2 mL) and then saline solution
R
NanoTekꢀ. The trapping solution was prepared by the addition
of CuCl (1.1 mg, 11 mmol) and K[Tp*] (3.7 mg, 11 mmol) to a
3 mL V-bottomed glass vial, flushing with nitrogen for 10 min,
followed by the addition of anhydrous tetrahydrofuran (1.0 mL).
The solution was then filtered into another 3 mL V-bottomed
glass vial under a nitrogen atmosphere. [11C]CO/He was bubbled
through the trapping solution and the radioactivity of the vial
monitored using a non-calibrated pin diode detector. Following
complete delivery of the radioactivity to the vial, corresponding
to the formation of Cu(Tp*)[11C]CO, the solution was transferred
R
into a 400 mL loop by aspirating syringe 2 of the NanoTekꢀ. The
cross-coupling reagent solution (100 mL) and the [11C]CO solution
(100 mL) were then dispensed simultaneously through a 100 mm ¥
4 m microreactor situated in a heater block. The mixture leaving
the microreactor was collected in a 1 mL vial containing water (0.1
mL) and the product mixture analysed by analytical HPLC using
a reversed-phase column (Agilent XDC C18, 5 mm, 4.6 ¥ 150 mm)
and a 60 : 40 water–MeCN mobile phase at a flow rate of 1.5 mL
min-1 (amide retention time = 4.5 min, urea retention time = 4.1
min).
3318 | Org. Biomol. Chem., 2011, 9, 3313–3319
This journal is
The Royal Society of Chemistry 2011
©

View Online
(8 mL). The radioactivity of the dose vial was measured in a dose
calibrator, and the radiochemical purity and specific activity of
the product were measured by analytical HPLC.
12 C. J. Steel, A. T. O’Brien, S. K. Luthra and F. Brady, J. Labelled Compd.
Radiopharm., 2007, 50, 308.
13 H. Audrain, Angew. Chem., Int. Ed., 2007, 46, 1772.
14 P. W. Miller, J. Chem. Technol. Biotechnol., 2009, 84, 309.
15 A. M. Elizarov, Lab Chip, 2009, 9, 1326.
16 J. H. Chun, S. Lu, Y. S. Lee and V. W. Pike, J. Org. Chem., 2010, 75,
3332.
Acknowledgements
17 P. W. Miller, N. J. Long, A. J. De Mello, R. Vilar, H. Audrain, C. Bender,
J. Passchier and A. Gee, Angew. Chem., Int. Ed., 2007, 46, 2875.
18 http://www.advion.com/biosystems/nanotek/nanotek-positron-
emission-tomography.php.
19 Radiochemical purity (RCP) refers to the radioactivity of a particular
component in a solution as a proportion of the total radioactivity of
the solution, as judged by analytical HPLC. This does not account for
unreacted [11C]CO due to its low solubility.
20 C. F. J. Barnard, Organometallics, 2008, 27, 5402.
21 K. Orito, M. Miyazawa, T. Nakamura, A. Horibata, H. Ushito, H.
Nagasaki, M. Yuguchi, S. Yamashita, T. Yamazaki and M. Tokuda, J.
Org. Chem., 2006, 71, 5951.
22 B. Gabriele, G. Salerno, R. Mancuso and M. Costa, J. Org. Chem.,
2004, 69, 4741.
23 I. Chiarotto and M. Feroci, J. Org. Chem., 2003, 68, 7137.
24 F. Shi, Y. Deng, T. SiMa and H. Yang, Tetrahedron Lett., 2001, 42,
2161.
25 H. Yang, Y. Deng and F. Shi, J. Mol. Catal. A: Chem., 2001, 176, 73.
26 G. M. Dahlen and A. Sen, Macromolecules, 1993, 26, 1784.
27 B. M. Choudary, R. K. Koteswara, S. D. Pirozhkov and A. L. Lapidus,
Synth. Commun., 1991, 21, 1923.
28 I. Pri-Bar and H. Alper, Can. J. Chem., 1990, 68, 1544.
29 S. P. Gupte and R. V. Chaudhari, J. Catal., 1988, 114, 246.
30 P. Giannoccaro, J. Organomet. Chem., 1987, 336, 271.
31 J. Tsuji and N. Iwamoto, J. Chem. Soc. D, 1966, 380.
32 H. D. Burns, R. E. Gibson, T. G. Hamill and T. Fukami, PCT Int.
Appl. WO 03/010175 A2, 2003.
33 N. Erondu, I. Gantz, B. Musser, S. Suryawanshi, M. Mallick, C. Addy,
J. Cote, G. Bray, K. Fujioka, H. Bays, P. Hollander, S. M. Sanabria-
Bohorquez, W. Eng, B. Langstrom, R. J. Hargreaves, H. D. Burns,
A. Kanatani, T. Fukami, D. J. MacNeil, K. M. Gottesdiener, J. M.
Amatruda, K. D. Kaufman and S. B. Heymsfield, Cell Metab., 2006, 4,
275.
We thank Valeria Rampazzo for the synthesis of the radiolabelling
precursor and the reference sample of MK-0233. N. J. L. would
like to thank the Leverhulme Trust for the award of a Research
Fellowship.
Notes and references
1 P. W. Miller, N. J. Long, R. Vilar and A. D. Gee, Angew. Chem., Int.
Ed., 2008, 47, 8998.
2 B. La˚ngstro¨m, O. Itsenko and O. Rahman, J. Labelled Compd.
Radiopharm., 2007, 50, 794.
3 E. D. Hostetler and H. D. Burns, Nucl. Med. Biol., 2002, 29, 845.
4 T. Kihlberg and B. Langstrom, PCT Int. Appl. WO 2002102711 A1,
2002.
5 H. Audrain, L. Martarello, A. Gee and D. Bender, Chem. Commun.,
2004, 558.
6 S. Kealey, P. W. Miller, N. J. Long, C. Plisson, L. Martarello and A. D.
Gee, Chem. Commun., 2009, 3696.
7 A. Elizarov, R. M. van Dam, Y. S. Shin, H. C. Kolb, H. C. Padgett,
D. Stout, J. Shu, J. Huang, A. Daridon and J. R. Heath, J. Nucl. Med.,
2010, 51, 282.
8 S. Y. Lu, P. Watts, F. T. Chin, J. Hong, M. J. L. E. Briard and V. W.
Pike, Lab Chip, 2004, 4, 523.
9 C. C. Lee, G. Sui, A. Elizarov, C. J. Shu, Y. S. Shin, A. N. Dooley, J.
Huang, A. Daridon, P. Wyatt, D. Stout, H. C. Kolb, O. N. Witte, N.
Satyamurthy, J. R. Heath, M. E. Phelps, S. R. Quake and H. R. Tseng,
Science, 2005, 310, 1793.
10 J. M. Gillies, C. Prenant, G. N. Chimon, G. J. Smethurst, W. Perrie, I.
Hamblett, B. Dekker and J. Zweit, Appl. Radiat. Isot., 2006, 64, 325.
11 H. M. Wu, G. Sui, C. C. Lee, M. L. Prins, W. Ladno, H. D. Lin, A. S.
Yu, M. E. Phelps and S. C. Huang, J. Nucl. Med., 2007, 48, 837.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 3313–3319 | 3319
©