Rapid multiphase carbonylation reactions by using a microtube reactor: Applications in positron emission tomography 11C-radiolabeling

Article abstract of DOI:10.1002/anie.200604541

Taking the tube: A silica-supported palladium catalyst packed into teflon tubing is a simple, low-cost, and effective method for carrying out carbonylative cross-coupling reactions of arylhalides with amines and radiolabeled carbon monoxide gas. In carbonylation reactions, the microtube reactor displays enhanced yields over a short time period (12 min) compared with batch methods. (Chemical Equation Presented).

Full text of DOI:10.1002/anie.200604541

Angewandte
Chemie
DOI: 10.1002/anie.200604541
Microfluidic Reactors
Rapid Multiphase Carbonylation Reactions by Using a Microtube
Reactor: Applications in Positron Emission Tomography
11C-Radiolabeling**
Philip W. Miller, Nicholas J. Long,* Andrew J. de Mello, Ramon Vilar, HØl›ne Audrain,
Dirk Bender, Jan Passchier, and Antony Gee
The development of new strategies for the synthesis of
positron-emitting radiolabeled compounds is of great current
interest owing to the increased use of positron-emission
tomography (PET) in medical imaging. Short-lived positron-
ing more attractive to synthetic chemists mainly owing to the
ease of purification steps and the cost efficiency of reusable
reagents and precious-metal catalysts. We are interested in
the palladium-catalyzed carbonylative cross-coupling reac-
tions of aryl halides, which can be used as a route to a number
of biologically interesting molecules such as amides, lactams,
esters, and lactones. In the typical synthesis of an amide, an
amine and an aryl halide are coupled together with a molecule
of carbon monoxide in the presence of a palladium catalyst
(Scheme 1). However, owing to the poor solubility of carbon
emitting radionuclides, such as 11C (t1/2 = 20.4 min), 13N (t1/2
=
9.96 min), 15O (t1/2 = 2.07 min), and 18F (t1/2 = 109.7 min),
require specially adapted methods of synthesis owing to
their short half-lives and the submicromolar quantities used in
reactions. [11C]carbon monoxide is becoming a recognized
[1–5]
precursor for 11Clabeling
owing to the potentially wide
ranging number of functional groups that can be incorporated
into target molecules. However, 11CO labeling is not without
its difficulties owing in part to the poor reactivity of carbon
monoxide and the problems associated with trapping enough
11CO in organic solvents to perform the reaction. Herein we
show that a possible solution to these limitations is to use
microfluidic-reactor technology.
Microfluidic-reactor devices are particularly well suited to
conducting multiphase gas–liquid or gas–liquid–solid reac-
tions; this is a result of their increased surface-area-to-volume
ratio that allows enhanced surface contact between different
phases. When compared with conventional laboratory meth-
ods, microfluidic reactors have been demonstrated to improve
reaction rates, yields, and/or selectivities of a number of gas–
liquid[6–10] and gas–liquid–solid reactions.[11–13] These improve-
ments are a direct result of superior heat and mass transfer
within the system and controlled mixing of reagent streams.
The use of supported reagents[14] and catalysts[15,16] is becom-
Scheme 1. Amide formation through the carbonylative cross-coupling
reaction. X=halogen, R=electron-donating or -withdrawing group.
monoxide in common organic solvents,[17] carbonylation
reactions are usually carried out at elevated pressures. Such
methods require high-pressure autoclave reactors that are
inherently expensive and require special safety precautions.
We recently reported the use of a continuous-flow micro-
fluidic reactor[18] for conducting carbonylation reactions that
proved to be superior to corresponding batch-reaction
methods. Herein, we report a simple, low-cost and effective
method for gas–liquid–solid carbonylation reactions employ-
ing a reusable silica-supported palladium catalyst packed into
standard teflon tubing (see Figure S1 in the Supporting
Information). This methodology makes it possible to obtain
good yields of the corresponding amides within very short
reaction times. These results, in turn, prompted us to employ
the microfluidic reactor system for 11C-radiolabeling experi-
ments. More specifically, we present preliminary results on
11CO carbonylative cross-coupling reactions to yield 11C-
labeled amides. To prepare the microtube reactor, a teflon
tube (45 cm in length and 1 mm in diameter) was packed by
injecting a slurry of the supported catalyst (250 mg) in THF
and plugged at both ends with cotton filters. The supported
catalyst (Scheme 2) was prepared by first synthesizing the
palladium–phosphine complex followed by attachment to the
silica-support material; this method differs to that previously
reported in which the ligand was first attached to a silica
support before complexation with PdII.[19] Two equivalents of
bis-diphenylphosphonium chloride were reacted with amino-
[*] Dr. P. W. Miller, Prof. N. J. Long, Prof. A. J. de Mello, Dr. R. Vilar
Department of Chemistry
Imperial College London
London, SW72AZ (UK)
Fax: (+44)207-594-5804
E-mail: n.long@imperial.ac.uk
Dr. H. Audrain, Dr. D. Bender
PET Imaging Division
Aarhus University Hospital
8000 Aarhus(Denmark)
Dr. J. Passchier, Dr. A. Gee
PET Imaging Division
GSK, Academic Centre for Clinical Investigation
Addenbrooke’sHospital, Cambridge (UK)
[**] We thank the NERC for the ICP-MS analysis performed in the School
of Earth Science and Geography at the University of Kingston,
London, by Dr. Kathryn Linge, Dr. Kym Jarvis, and Benoit Disch.
Supporting information for thisarticle isavailable on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 2875 –2878
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2875

Communications
reaction. Iodobenzene, as expected, gave a higher
yield than bromobenzene (Table 1, entries 1 and 2)
owing to the more facile oxidative addition of arylio-
dides to Pd0 in the catalytic cycle.[20] Product yields
were significantly higher over the 12-minute reaction
period when the substrates contained electron-with-
drawing p-trifluoromethyl, p-benzonitrile, or 2-pyridyl
groups (Table 1, entries 4–6). The yields are signifi-
cantly less when an electron-donating methoxy group
is attached to the para position of the aryl ring (Table 1,
entry 3). Corresponding batch reactions were carried
out by using the same palladium-supported catalyst to
Scheme 2. Synthesis of the silica-supported diphosphine catalyst 1c.
propyltriethoxysilane to form the chelating ligand 1a. This
was then reacted with one equivalent of [Pd(cod)Cl2] (cod =
1,5-cyclooctadiene) to form complex 1b. This compound was
then reacted with silica to form the supported catalyst 1c.
Inductively coupled plasma (ICP)-MS analysis of the sup-
ported catalyst gave palladium loadings of 1.5%.
The reactor setup (Figure 1) consists of a precision syringe
pump to control the infusion rate of the solvents and
substrates, an injector port for substrate injection, a mass
Table 1: Silica-supported palladium-catalyzed carbonylative cross-cou-
pling reactionsusing the microtube reactor.
Entry
Substrate
Yield [%][a]
Yield [%][b]
(micro reactions)
(batch reactions)
1
2
3
63
26
23
10
0
0
4
>99
44
5
6
>99
>99
27
43
[a] Average of three runsbased on the aryl halide and calculated by GC
with diphenyl ether asthe internal standard. [b] Average of two runs
based on the aryl halide and calculated by GC with diphenyl ether as the
internal standard.
Figure 1. Schematic representation of the carbonylation microtube
reactor setup.
flow controller to meter the flow of carbon monoxide into the
system, and a mixing T piece to mix the gas and liquid
reagents and which connects directly to the microtube
reactor.
compare the microtube reactor with traditional synthetic
methods that use standard Schlenk glassware. In all cases,
significantly higher yields, over the same 12-minute time
period, were obtained by using the microtube reactor method
(Table 1).
In a typical reaction, a 50-mL aliquot of substrate solution
(aryl halide and benzylamine) was injected through the
injector port into a continuous flow of solvent (THF) where it
was mixed with a steady stream of carbon monoxide, passed
into the microtube reactor, and then heated to 758C. After a
12-minute residence time period in the reactor, the solvent
flow was increased, flushing the product from the reactor for
collection and analysis. We examined six different arylhalide
substrates for the carbonylative cross-coupling reaction with
benzylamine; the results of these reactions are displayed in
Table 1. Three consecutive runs were carried out for each
arylhalide substrate by using the same microtube reactor for
all the carbonylation reactions (a total of 18 separate
reactions); no loss in catalytic activity was observed between
the consecutive runs.
An inherent property of microchannels is the high sur-
face-area-to-volume ratios, ranging from 10000 to
50000 m2 mÀ3, generated as a consequence of their decreased
size. However, when a packing material such as silica is used,
surface-area-to-volume ratios can be increased dramatically
even when compared with microchannels. The surface-area-
to-volume ratio generated within our device was estimated to
be 1.1 ” 109 m2 mÀ3, which is approximately 20000 times
[
]
*
greater than that generated within a microchannel. We
[*] Surface area of silica=750 m2 gÀ1, pore volume=0.68 cm3 gÀ1. Silica-
supported catalyst (0.25 g) was used in the microtube reactor,
therefore total pore volume (taken to be the reactor volu-
me)=0.17 cm3. Surface-area-to-volume ratio=187.5 m2/
Yields of the carbonylated products from the microtube
reactor varied depending on the substrates used in the
1.7”10À7 m3 =1.1”109 m2 mÀ3
.
2876 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2875 –2878

Angewandte
Chemie
Table 2: 11CO carbonylative cross-coupling reactions using the micro-
tube-reactor system.[a]
believe the improved yields for these flow reactions are due
directly to the large surface area provided by the silica-
supported catalyst. The silica support provides a microporous
structure and an extremely large surface area for the liquid
reagents to flow through and spread out. This allows the
liquid reagent to coat the surface and therefore enhance the
contact with carbon monoxide gas. Furthermore, the silica
support should also increase turbulence in the system,
enhancing the mixing of the gas and liquid reagents. Once it
was demonstrated that the microtube reactor gave good
yields of the corresponding amides in short reaction times
(under 12 min), it was of interest to investigate its applic-
ability for radiolabeling experiments with 11CO. As indicated
before, an important limitation when labeling molecules for
Entry
Labeled product
Radiochemical
yield [%][b]
Radiochemical
purity [%][c]
1
run 1
run 2
79
65
96
94
2
3
4
run 1
run 2
67
70
95
95
run 1
run 2
46
68
70
90
PET imaging by using 11CO is the short half-life of 11C (t1/2
=
run 1
run 2
45
33
72
80
20.4 min). A schematic diagram of the experimental setup
and experimental details for the radiolabeling experiment is
shown in the Supporting Information.
11CO was produced on-line by reduction of cyclotron-
generated 11CO2 by using a molybdenum catalyst at 8508C.[21]
A stainless-steel loop packed with molecular sieves and
cooled by using liquid nitrogen was used to trap and
concentrate the 11CO, reaching a maximum radioactivity at
approximately 8 min from the end of cyclotron bombardment
(EOB). The 11CO was released from the trap by warming to
ambient temperatures and forced under a controlled flow of
nitrogen into the microtube reactor where the aryl halide and
benzylamine reagents had been preloaded and heated to
808C. After about 6-min reaction time, under a continuous
flow of nitrogen gas, the catalyst loop was lifted out of the oil
bath and solvent was pumped through for approximately
5 min, flushing the labeled product into a sample vial. The
crude product was filtered and analyzed by using analytical
radio HPLC. Four different aryl halides (iodobenzene, 4-
bromobenzonitrile, 4-bromobenzotrifluoride, and 4-iodoani-
sole) were investigated to test the applicability of the
microtube reactor system to the 11CO labeling for amide
formation. Separate microtube reactors were used for each
different substrate and two consecutive runs were carried out
for each substrate; the results of these reactions are displayed
in Table 2.
The total radioactivity in the system was taken as the sum
of the radioactivity measured in the crude product at the end
of synthesis, unreacted 11CO that swept through the micro-
tube reactor into a sealed bag, and the radioactivity that was
left on the microtube reactor. It was found that typically 10–
15% of the total radioactivity remained on the microtube
reactor at the end of the synthesis, which may be attributed to
traces of labeled products or intermediates sticking to the
silica. Good radiochemical yields (> 64%) and purities (>
93%) of the crude products were obtained for the products
shown in Table 2, entries 1 and 2 within about 10 min of 11CO
release from the trap, whereas modest radiochemical yields
(> 33%) and purities (> 70%) were obtained for Table 2,
entries 3 and 4. Table 2, entry 3 shows lower than expected
radiochemical yields even though this substrate is activated by
the electron-withdrawing para-CF3 group. The lower radio-
chemical yields obtained for Table 2, entry 4 may be due to
the effect of the deactivating para-methoxy group on the aryl
[a] * indicatesthe labeling position. [b] Decay-corrected radiochemical
yields, expressed as a percentage of the total radioactivity delivered to
the microreactor system, are based on the measured radioactivity
trapped in the crude products at the end of synthesis and corrected by
their radiochemical purities. [c] Radiochemical purities determined by
analytical radio HPLC.
halide, which will affect the oxidative addition step of the
catalytic cycle. When multiple-labeling reactions, using differ-
ent substrates, were carried out on the same microtube
reactor, mixtures of radiolabeled products were obtained.
Even after extensive flushing of the microtube reactor with
polar solvent between successive runs, mixtures of labeled
products were still obtained. We believe this to be due to
traces of either unreacted starting material remaining on the
silica or the oxidative addition aryl-Pd species, which remains
adhered to the silica and which cannot simply be removed by
washing with solvent.
In conclusion, we have provided an effective method for
continuous flow carbonylation reactions by using a microtube
reactor packed with palladium-supported catalyst, which
proved to be reusable for a number of runs. We synthesized
a series of amides over a set reaction time and compared this
method with batch reaction conditions; the packed microtube
reactor method proved superior in all cases and provided
excellent yields for three of the substrates in only 12 minutes.
We have successfully applied the methodology towards
radiolabeling by 11CO carbonylative cross-coupling reactions
and gained modest to good radiochemical yields and purities
of labeled amides. It is expected that improved synthetic and
radiochemical yields may be obtained by increasing the
residence time of the substrates within the microtube reactor,
and we are currently in the process of optimizing these
reactions and applying our setup to other carbonylation
reactions.
Experimental Section
General procedure for microtube carbonylation reactions: In a
typical reaction, an aliquot (50 mL) of 1m aryl halide solution in
Angew. Chem. Int. Ed. 2007, 46, 2875 –2878
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2877

Communications
benzylamine was injected into a continuous flow of THF through a
[6] Y. Önal, M. Lucas, P. Claus, Chem. Eng. Technol. 2005, 28, 972.
[7] K. Jähnisch, U. Dingerdissen, Chem. Eng. Technol. 2005, 28, 426.
[8] R. C. R. Wootton, R. Fortt, A. J. de Mello, Org. Process Res.
Dev. 2002, 6, 187.
[9] L. Kiwi-Minsker, A. Renken, Catal. Today 2005, 110, 2.
[10] G. N. Doku, W. Verboom, D. N. Reinhoudt, A. van den Berg,
Tetrahedron 2005, 61, 2733.
[11] J. Kobayashi, Y. Mori, K. Okamoto, R. Akiyama, M. Ueno, T.
Kitamori, S. Kobayashi, Science 2004, 304, 1305.
[12] N. Yoswathananont, K. Nitta, Y. Nishiuchi, M. Sato, Chem.
Commun. 2005, 40.
[13] M. W. Losey, R. J. Jackman, S. L. Firebaugh, M. A. Schmidt,
K. F. Jensen, J. Microelectromech. Syst. 2002, 11, 709.
[14] I. R. Baxendale, J. Deeley, C. M. Griffiths-Jones, S. V. Ley, S.
Saaby, G. K. Tranmer, Chem. Commun. 2006, 2566.
[15] Y. R. de Miguel, A. S. Shearer, Biotechnol. Bioeng. 2000, 71, 119.
[16] Q. H. Fan, Y. M. Li, A. S. C. Chan, Chem. Rev. 2002, 102, 3385.
[17] H. M. Colquhoun, D. J. Thompson, M. V. Twigg, Carbonylation
Direct Synthesis of Carbonyl Compounds, Plenum, New York,
1991, p. 37.
Rheodyne 7725 injector port. The substrate mixture was then
premixed with a continuous flow of carbon monoxide gas by using a
mixing T piece (part no. P-512, Anachem Ltd) before the mixture
entered the microtube reactor, and was then heated to 758Cby using
an oil bath. Liquid flow rates were set at 10 mLminÀ1 and the gas flow
rate was metered at 2 sccm by using a Sierra 100 series Smart-Trak
mass flow controller. After a residence time period of 12 min from the
initial substrate injection, the product was flushed off the microtube
reactor by using a higher flow of THF (200 mLminÀ1). The product
was collected, the solvent was evaporated in vacuo, and a standard
solution of diphenyl ether was added prior to GCanalysis.
Received: November 6, 2006
Published online: March 9, 2007
Keywords: carbonylation · microfluidics· microreactors·
.
palladium · positron emission tomography
[18] P. W. Miller, N. J. Long, A. J. de Mello, R. Vilar, A. Gee, J.
Passchier, Chem. Commun. 2006, 546.
[19] S. Antebi, P. Arya, L. E. Manzer, H. Alper, J. Org. Chem. 2002,
67, 6623.
[1] O. Itsenko, B. Långström, J. Org. Chem. 2005, 70, 2244.
[2] F. Karimi, J. Barletta, B. Långström, Eur. J. Org. Chem. 2005,
2374.
[3] O. Itsenko, T. Kihlberg, B. Långström, Synlett 2005, 3154.
[4] G. Antoni, T. Kihlberg, B. B. Långström, Handbook of Radio-
pharmaceuticals: Radiochemistry and Applications (Eds.: M. J.
Welch, C. S. Redvanly), Wiley, New York, 2003, Chapter 5,
p. 140.
[20] U. Christmann, R. Vilar, Angew. Chem. 2005, 117, 370; Angew.
Chem. Int. Ed. 2005, 44, 366.
[21] S. K. Zeisler, M. Nader, A. Theobald, F. Oberdorfer, Appl.
Radiat. Isot. 1997, 48, 1091.
[5] H. Audrain, L. Martarello, A. Gee, D. Bender, Chem. Commun.
2004, 558.
2878
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2875 –2878