Synthesis of 13N-labelled radiotracers by using microfluidic technology

Article abstract of DOI:10.1002/jlcr.2946

Microfluidics has recently emerged as a useful alternative to traditional methods for the preparation of radiotracers labelled with positron emitters. Up to date, microfluidics technology has been applied to the radiosynthesis of ¹⁸F-labelled and ¹¹C-labelled compounds; however, application to other shorter-lived positron emitters has not been investigated. The radiosynthesis of S-[¹³N]nitrosothiols and N-[ ¹³N]nitrosamines was approached in microfluidic system by reaction of the corresponding thiol or secondary amine, respectively, with [ ¹³N]NO₂^- in the presence of mineral acid. The radiosynthesis of azo compounds was carried out by reaction of the same labelling agent with primary aromatic amines in acidic media to yield the corresponding diazonium salts, which were further reacted with aromatic amines and alcohols to yield the corresponding ¹³N-labelled azo compounds. Radiochemical conversion values for S-[¹³N]nitrosothiols and ¹³N-labelled azo compounds calculated from chromatographic profiles improved our previous results by using conventional methods. The formation of N-[¹³N]nitrosamines could not be detected, independently of experimental conditions. In conclusion, the preparation of S-[ ¹³N]nitrosothiols and ¹³N-labelled azo compounds was successfully achieved by using microfluidics technology. Higher radiochemical conversions than those previously reported using conventional synthetic strategies have been obtained. Copyright

Full text of DOI:10.1002/jlcr.2946

Research Article
Received 17 April 2012,
Revised 6 June 2012,
Accepted 23 June 2012
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.2946
Synthesis of 13N-labelled radiotracers by using
microfluidic technology
Vijay Gaja,a Vanessa Gómez-Vallejo,a Mar Cuadrado-Tejedor,b
c
a
*
José I. Borrell, and Jordi Llop
Microfluidics has recently emerged as a useful alternative to traditional methods for the preparation of radiotracers
labelled with positron emitters. Up to date, microfluidics technology has been applied to the radiosynthesis of 18F-labelled
and 11C-labelled compounds; however, application to other shorter-lived positron emitters has not been investigated. The
radiosynthesis of S-[13N]nitrosothiols and N-[13N]nitrosamines was approached in microfluidic system by reaction of the
corresponding thiol or secondary amine, respectively, with [13N]NO2À in the presence of mineral acid. The radiosynthesis
of azo compounds was carried out by reaction of the same labelling agent with primary aromatic amines in acidic media
to yield the corresponding diazonium salts, which were further reacted with aromatic amines and alcohols to yield the
corresponding 13N-labelled azo compounds. Radiochemical conversion values for S-[13N]nitrosothiols and 13N-labelled azo
compounds calculated from chromatographic profiles improved our previous results by using conventional methods. The
formation of N-[13N]nitrosamines could not be detected, independently of experimental conditions. In conclusion, the
preparation of S-[13N]nitrosothiols and 13N-labelled azo compounds was successfully achieved by using microfluidics
technology. Higher radiochemical conversions than those previously reported using conventional synthetic strategies have
been obtained.
Keywords: nitrogen-13; microfluidics; nitrosothiols; nitrosamines; azo; PET
fallypride,6 [18F]FIAU7 and [18F]EtDT, [18F]PrDT and [18F]CB-102.5
Introduction
In all cases, the cyclotron-produced 18FÀ was trapped in an
Positron emission tomography (PET) is a non-invasive in vivo
anionic exchange resin cartridge and eluted with Kryptofix K2.2.2/
imaging technique that produces a three-dimensional image of
K2CO3 into a conical shape vial where azeotropic evaporation of
the solvent and reconstitution with acetonitrile left the 18F-fluoride
functional processes in a living organism, allowing the study of
in vivo biochemistry and biology underlying disease and
solution ready for multiple batch production by using the
therapeutic intervention.1,2 Despite the increasing number of
microfluidics system. More recently, the same system has been
biomedical cyclotrons installed all over the world3 (estimated
over 650 in 2010 according to the International Atomic Energy
Agency), the possibility to perform studies with PET is limited
by the production capabilities of radiochemistry laboratories,
where the complexity of the processes associated to the
preparation of the radiotracers (including radionuclide production,
labelling, purification and formulation) traditionally collides with
the real-time flexibility required to perform clinical or preclinical
PET studies. Because of this fact, faster, more efficient and reliable
radiosynthesis procedures are continuously developed, and in this
scenario, the microfluidic technology offers a very promising
approach for fast labelling optimization and dose-on-demand
implementation.4,5
used to perform [11C]carbonylation reactions by using solutions
containing [11C]CO in the form of the copper(I)tris(3,5-dimethylpyr-
azolyl)borate-[11C]carbonyl complex (Cu(Tp*)[11C]CO).8
Carbon-11 and fluorine-18 have been historically the most
widely used positron emitters, and because of this fact, first
applications of microfluidics technology have been focused in
the preparation of 18F-labelled and 11C-labelled radiotracers.
Although the use of other radioisotopes such as 13N (half-life of
9.97 min, maximum positron energy of 1.19 MeV) might be very
advantageous for the development of alternative synthetic
strategies to label molecules in different positions, the short
Among all microfluidic alternatives, the continuous flow
concept was introduced in the market in a commercially
available system by Advion (Figure 1 for schematic flowchart),
being the main advantages of this system the elimination of
temperature gradients within the reaction mixture and the
possibility to perform reactions at high pressure, which lead to
faster incorporation rates of the labelling agent into the final
radiotracer while the amount of precursor can be decreased.
Up to date, such system has been used for the preparation
of 18F-labelled radiotracers, including (among others) [18F]
aRadiochemistry Department, CIC biomaGUNE, Paseo Miramón 182, San
Sebastián 20009, Spain
bDivision of Neuroscience, CIMA, University of Navarra, Pamplona, Spain
cGrup d’Enginyeria Molecular, Institut Químic de Sarrià, Universitat Ramon Llull,
Via Augusta 390, 08017 Barcelona, Spain
*Correspondence to: Jordi Llop, Radiochemistry Department, CIC biomaGUNE,
Paseo Miramón 182, San Sebastián, 20009, Spain.
E-mail: jllop@cicbiomagune.es
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.

V. Gaja et al.
Figure 1. Flowchart of the automated system for the synthesis of 13N-labelled radiotracers. V1–V8, two-way to three-way electrovalves; P1–P3, syringe pumps; L1–L3,
loops; R1 and R2, reactors.
half-life of this isotope requires preparation of single dose p. a., >99%, GC quality), trifluoroacetic acid (TFA) (puriss. p. a., for
batches, as well as the development of efficient synthetic HPLC, >99%) and piperidine (puriss. p. a., >99%, GC quality)
strategies. In this context, the use of microfluidics could provide were purchased from Sigma-Aldrich Química S.A. (Madrid, Spain)
a very interesting alternative to facilitate the use of 13N-labelled and used without further purification. Ethanol, acetonitrile and
radiotracers in preclinical and clinical practice.
methanol were of HPLC grade and purchased from Scharlab
Recently, we have developed strategies for the generation of the (Sentmenat, Barcelona, Spain). Water (ultrapure, Type I water, ISO
labelling agent [13N]NO2À starting from cyclotron-produced [13N] 3696) was obtained from a Milli-QW Integral system (Millipore
NOÀ3 and its incorporation into bioactive molecules such as N-[13N] Iberica S.A.U., Madrid, Spain).
nitrosamines,9 S-[13N]nitrosothiols10,11 and azo compounds.12
Despite the short half-life of nitrogen-13, these 13N-labelled
compounds have been prepared with good radiochemical Production of nitrogen-13
yields in short times. In the present paper, we describe for the first
Nitrogen-13 was produced in an IBA Cyclone 18/9 cyclotron via
time the radiosynthesis of 13N-labelled S-[13N]nitrosothiols and azo
the 16O(p,a)13N nuclear reaction. The target system consisting
compounds by using microfluidics technology, with improved
of an aluminium body with aluminium vacuum foil (thickness
radiochemical conversion (RCC) results compared with previously
25 mm, ∅ 23 mm) and havar target foil (thickness 25 mm, ∅
published data. Attempts to synthesize 13N-labelled nitrosamines
are also reported and briefly discussed. The potential application
29 mm) containing 1.75 mL of water was irradiated with
18 MeV protons at a beam current of 20 mA for 20 s (integrated
of microfluidics to the preparation of 13N-labelled radiotracers
current of 0.1 mA h). The irradiated solution was collected in a
and improvements to be implemented are also mentioned.
vial and passed through a glass column filled with cadmium
to quantitatively reduce [13N]NO3À into [13N]NOÀ2 . The cadmium
Materials and methods
column was further rinsed with 1 mL of water, and the final
solution was transferred to a conical shape vial to be used in
General
the microfluidics system. A small fraction of this solution
Aniline (ACS reagent), 4-aminobenzenesulfonic acid (ACS (20 mL) was submitted to HPLC (see succeeding discussions for
reagent), phenol (ReagentPlusW, ≥99.5%), 4-phenylazophenol conditions) to determine the percentage of radioactivity
(98%), 4-phenylazoaniline, 4-(4-hydroxy-phenylazo)-benzenesul- as [13N]NO2À.
fonic acid sodium salt, ammonium formate (AMF) (reagent
A set of 10 preliminary experiments was carried out in order to
grade, 97%), formic acid (ACS reagent, >96%), acetic acid determine the amount of radioactivity generated at the end of
(purum, >99.0%), sodium acetate (ACS reagent, ≥99.0%), cyclo- the irradiation process (381 Æ 15 MBq) and the percentage of
hexanethiol (97%), 1-propanethiol (99%), glutathione (>98%), [13N]NOÀ3 , [13N]NOÀ2 and [13N]NH4+ before the reduction step
1-adamantanethiol (>99%, GC quality), diisopropylamine (puriss. (86%, 5% and 9%, respectively).
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

V. Gaja et al.
The microfluidic platform
The radiosyntheses under microfluidic conditions were conducted
using a NanoTek Microfluidic Synthesis System (Advion, USA),
which consisted of two blocks: a reactor module and a pump
module (Figure 1). The pump module block consisted of two
high-pressure syringe pumps (P1 and P2), both connected to an
eight-way bridged valve. The reactor module comprises a high-
pressure syringe pump (P3) connected to an eight-way bridged
valve and four thermostated slots where microreactors can be
hosted. P1 and P2 were used to dispense the precursors, and P3
was used to dispense [13N]NOÀ2 . Reagents were loaded in loops
1–3 (L1, L2 and L3 in Figure 1) before being pushed to the reactor.
For one-step reactions (preparation of S-[13N]nitrosothiols and
N-[13N]nitrosamines), only P1 and P3 were used, whereas P1, P2
and P3 were used for conducting two-step reactions (preparation
of 13N-labelled azo compounds). Microreactors were made of
fused silica tubing (100 mm ID; length= 2 m) rolled and held by a
brass ring casted with a thermoresistant silicone polymer.
Scheme 2. Radiosynthesis of N-[13N]-nitrosamines.
approach, in which the precursor was loaded in L1 and the acid
was added to the labelling agent solution and loaded in L3, was
also assayed.
13N-Labelled azo compounds
For the synthesis of 13N-labelled azo compounds, a two-step
reaction was used (Scheme 3). The labelling agent ([13N]NO2À)
was loaded in storage loop L3 (Figure 1), the aromatic amine
(precursor A, Scheme 3) was loaded in L1, and the phenol or
aromatic amine (precursor B, Scheme 3) was loaded in L2. The
reactions were conducted using discrete bolus of reagents
driven by the pertinent pumps (P3 for labelling agent, P1 and
P2 for precursors for steps 1 and 2, respectively) in microreactors
R1 (step 1) and R2 (step 2). Different relative ratios between
bolus volumes, flow rates and reactor temperatures were
assayed. Pure water was used to push the reaction crude to
the final vial after reaction was finished and to clean the reaction
pathway between consecutive runs.
Synthesis of S-[13N]nitrosothiols
The synthesis of S-[13N]nitrosothiols was approached by direct
nitrosation of the corresponding thiols with [13N]NO2À in the pre-
sence of mineral acid (Scheme 1). The labelling agent ([13N]NO2À)
was loaded in storage loop L3 (Figure 1) whereas the precursor
was dissolved in acidic acetonitrile solution and loaded in L1.
The reactions were conducted using discrete bolus of reagents
driven by the pertinent pumps (P3 for labelling agent, P1 for
the precursor, Figure 1) in microreactor 1 (R1). Different relative
ratios between bolus volumes and flow rates were assayed. Pure
solvent was used to push the reaction crude to the final vial after
reaction was finished and to clean the reaction pathway
between consecutive runs.
Determination of radiochemical conversion
Synthesis of N-[13N]nitrosamines
The determination of the RCC was carried out by HPLC. An
Agilent 1200 series HPLC equipped with a quaternary pump, a
variable wavelength detector (l = 220 nm) and a radiometric
detector (Gabi, Raytest) was used in all cases.
The synthesis of N-[13N]nitrosamines was approached by direct
nitrosation of the corresponding secondary amines with [13N]
NOÀ2 in the presence of mineral acid (Scheme 2). The labelling
agent ([13N]NOÀ2 ) was loaded in storage loop L3 (Figure 1)
whereas the precursor was dissolved in acidic solution and
loaded in L1. The reactions were conducted using discrete bolus
of reagents driven by the pertinent pumps (P3 for labelling
agent, P1 for precursor, Figure 1) in microreactor 1 (R1). A second
Chromatographic conditions for [13N]NO2À determination:
stationary phase: HP Asahipak ODP-50 column (4.0 Â 125 mm,
5 mm); mobile phase: solution containing additive for ionic
chromatography (P/N 5062–2480, Agilent Technologies, 15 mL)
in a mixture water/acetonitrile (86/14), adjusted to pH = 8.6 with
1 M NaOH solution at a flow rate of 1 mL/min.
Chromatographic conditions for S-[13N]nitrosothiols: station-
ary phase: Mediterranea Sea18 column (4.6 Â150 mm, 5 mm);
mobile phase: water/methanol/acetonitrile (10/15/75) at a flow
rate of 1 mL/min for compounds 1–3 and 5; aqueous 0.1% TFA
solution/acetonitrile (95/5) at a flow rate of 1 mL/min for
compound 4.
Chromatographic conditions for S-[13N]nitrosamines: stationary
phase: YMC J’sphere ODS-H80 column (4.6Â120 mm, 4 mm);
mobile phase: water/acetonitrile/methanol (28/18.5/3.5) at a flow
rate of 1 mL/min.
Chromatographic conditions for 13N-labelled azo compounds:
stationary phase: Zorbax Eclipse XDB C18 column (4.6 Â150 mm,
5 mm); mobile phase: AMF 0.1 M aqueous solution (pH = 3.9)/
methanol/acetonitrile (40/15/45) at a flow rate of 1 mL/min for
compound 8; AMF 0.1 M aqueous solution (pH = 3.9)/methanol/
acetonitrile (40/15/45) at a flow rate of 1 mL/min for compound
9; AMF 0.1 M aqueous solution (pH = 3.9)/acetonitrile (83/17) at
a flow rate of 1 mL/min for compound 10.
Scheme 1. Radiosynthesis of S-[13N]-nitrosothiols.
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

V. Gaja et al.
Scheme 3. Radiosynthesis of 13N-labelled azo compounds.
Radiochemical conversion values were calculated in all cases and reaction temperatures were used. However, only HCl (which
from chromatographic profiles, referred to the percentage of could not be employed in solid phase support synthesis because
[13N]NOÀ2 in the solution resulting from the reduction step.
of displacement of the labelling agent) was used as acid.
For optimization of the experimental conditions, several acid
concentrations and flow rates were assayed. A significant
dependence of RCC with acid concentration was found, as
expected, in all cases. Almost 100% RCC values (Table 1) were
obtained at room temperature for compounds 2, 3 and 4 at
HCl concentrations of 0.05 M (compounds 2 and 3) and 0.1 M
(compound 4). These are milder conditions than the ones used
in solid phase support synthesis (acid concentration = 1 M). Flow
rates of 20, 40 and 60 mL/min (P1 + P3) were assayed, but effects
on RCC could not be observed, probably because of the fast
reaction rate. When the acid concentration was decreased by a
factor of 10, significantly lower RCCs, which were dependent
on flow rate (the lower the flow rate, the higher the RCC) were
obtained (results not shown). Optimal conditions offered better
RCC values (99.5 Æ 0.2, 98.4 Æ 1.1 and 99.6 Æ 0.3 for 2–4) than
those obtained using solid phase support synthesis (60.3 Æ 5.5,
74.5 Æ 2.1 and 48.7 Æ 6.3, respectively), thus confirming the
faster incorporation rates of the labelling agent into the final
radiotracer by using the microfluidics approach (Table 1).
Results and discussion
Synthesis of S-[13N]nitrosothiols
The synthesis of S-[13N]nitrosothiols was first reported by our
research group in 2009. The first attempts consisted of synthesizing
[13N]GSNO by dissolving glutathione in aqueous hydrochloric acid,
which was reacted with [13N]NO2À in a reaction vial.10 In a more
recent work,11 the labelling strategy was based on trapping [13N]
NOÀ2 in an anion exchange resin, which was further reacted (in
the solid support) with a group of thiols to yield the desired
S-[13N]nitrosothiols in short reaction times (60 s) with good RCCs
(48.7–74.5% with respect to [13N]NOÀ2 , depending on the starting
thiol). All reactions were assayed at room temperature and
with the same acid concentration. Because the formation of
S-nitrosothiols is a very straightforward reaction that works very well
both in non-radioactive and in radioactive conditions, our first
attempts to implement microfluidics technology to the preparation
of 13N-labelled radiotracers were performed for the synthesis of
S-[13N]nitrosothiols. In order to obtain comparative results to
previously published data, the same concentration of precursors
Lower RCC values were obtained in the case of compounds 1
and 5. In both cases, no conversion was observed when HCl was
not added to the precursor solution (Figure 2); an increase in acid
Table 1. Optimized experimental conditions and radiochemical conversion for the preparation of 13N-labelled nitrosothiols by
microfluidics and by solid phase support synthesis
Solid phase support
RCC (%)f
Microfluidics
Compound
Acid
Acid
RCC (%)f
1
2
3
4
5
TFAa
TFAa
TFAa
HCla
TFAa
72.5 Æ 3.9
60.3 Æ 5.5
74.5 Æ 2.1
48.7 Æ 6.3
66.2 Æ 5.8
HClb
HClc
HClc
HCld
HCle
53.6 Æ 8.8
99.5 Æ 0.2
98.4 Æ 1.1
99.6 Æ 0.3
57.3 Æ 6.5
Solid phase support results have been obtained from Gómez-Vallejo et al.11 RCC, radiochemical conversion.
aConcentration: 1.0 M.
bConcentration: 0.01 M.
cConcentration: 0.05 M.
dConcentration: 0.1 M.
eConcentration: 0.5 M.
fCalculated as the ratio between the amount of radiotracer before purification and the amount of [13N]NOÀ2 obtained after the
reduction step.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

V. Gaja et al.
was anticipated to be not convenient; therefore, the approach
based on direct reaction of the amines with [13N]NO2À in acidic
media was assayed using piperidine as secondary amine. In first
instance, a solution of the amine in aqueous acidic solution was
loaded in L1, and the labelling agent ([13N]NO2À) was loaded in
L3. Different acid concentrations (0.001–1 M) and reaction
temperatures (60–120 ꢀC) were assayed while the flow rate was
maintained at 40 mL/min. Unfortunately, the presence of N-[13N]
nitrosopiperidine could not be detected in any case; thus, a
second approach in which the acid was added to the labelling
agent and the resulting solution was loaded in L3 while the
solution of the amine was loaded in L1 was assayed with
identical negative results, independently of the acid concentration
(0.001–1 M) and reaction temperature (60–120 ꢀC).
Despite that unsuccessful results could be expected,
13N-nitrosation of diisopropylamine was also tested here by using
microfluidics conditions. No N-[13N]nitrosodiisopropylamine could
be detected, irrespective of the experimental conditions used.
These results suggest that the preparation of 13N-labelled nitrosa-
mines under microfluidic conditions might require activation of
the secondary amine via similar strategies to those previously
reported by our research group.
Synthesis of 13N-labelled azo compounds
The synthesis of 13N-labelled azo compounds, which could be
potentially used as in vivo b-amyloid markers, was reported
recently by our research group for the first time.12 The synthetic
approach consisted of a two-step process; the first step involved
the reaction of anion exchange resin trapped [13N]NO2À with
primary aromatic amines in acidic conditions to yield the
corresponding 13N-labelled diazonium salt, which was then
pushed to a reactor pre-charged with a solution of an aromatic
amine or phenol to yield the 13N-labelled azo compound. In
the present work, a parallel strategy was used, although both
steps were performed under microfluidic conditions; the diazo-
nium salt was formed in reactor R1 (Figure 1) and was mixed
with the corresponding aromatic amine or phenol (Scheme 3)
in reactor R2 (Figure 1). The formation of the diazonium salt
takes place under acidic conditions (pH < 2), whereas the
formation of the azo compound gives better yields when
carried out under moderate acidic or basic conditions (pH > 4,
depending on the structure of the final tracer). In order to obtain
comparative results to those previously reported, the same
media and precursor concentrations were used; thus, a solution
of the aromatic amine (precursor A) in 1 M (compounds 8 and
9) or 0.1 M (compound 10) aqueous HCl was loaded in L1, the
labelling agent ([13N]NO2À, aqueous solution) was loaded in L3
and a solution of the aromatic amine or phenol (precursor B,
Scheme 3, see Table 2 for solvent) was loaded in L2.
Figure 2. Radiochemical conversion values for compounds 1 (a) and 5 (b) as a
function of HCl concentration and flow rate.
concentration yielded higher RCCs up to a certain level; in the
case of compound 1 (Figure 2(a)), RCC increased up to
61.8 Æ 2.3% for HCl concentration of 10 mM. However, at higher
HCl concentrations, the RCC reached a plateau (65.3 Æ 3.4% at
HCl concentration of 1 M, result not shown in the figure). A clear
effect of the flow rate on the RCC was observed, independently
of the HCl concentration (lower flow rates yielding higher RCC
values). Interestingly, measurable values of RCC were only
observed for compound
5 when HCl concentration was
≥50 mM (Figure 2(b)). A maximum in RCC was obtained for HCl
concentration of 0.5 M at a flow rate of 20 mL/min (64.1 Æ 4.8%),
which decreased at higher acid concentrations. This result
suggests the decomposition of the S-[13N]nitrosothiol in strong
acidic conditions. For both compounds, the results obtained in
our previous work (solid support synthesis, RCC values of
72.5 Æ 3.9 and 66.2 Æ 5.8% for 1 and 5, respectively) could not
be reached, although lower flow rates at the optimal HCl concen-
tration (0.5 M) might lead to higher RCC values; however, lower
flow rates require longer overall reaction times and could thus
compromise decay-corrected radiochemical yield.
Synthesis of N-[13N]nitrosamines
First assays were performed on compound 8. As expected,
The synthesis of N-[13N]nitrosamines has been also previously because of the aforementioned dependence of reaction rate
reported by our research group.9 In our previous work, our first with pH, the ratio between the volumes of precursors A and B
attempts consisted of the reaction of nitrous acid (generated had a big effect on RCC (Figure 3). Almost quantitative RCC
from nitrite solution and mineral acid in water) with secondary values were obtained when ratios (precursor B)/(precursor A) = 2/1
amines. In this previously reported work, the presence of the were used, with minor effect of flow rate (RCC = 91.6 Æ 1.7,
corresponding N-[13N]nitrosamine could not be detected 93.1 Æ 2.3 and 93.9 Æ 4.1% for flow rates of 40, 60 and 80 mL/min
irrespective of the starting secondary amine. Thus, a second in step 1). When the (precursor B)/(precursor A) ratio was decreased
combined approach (resin-supported NOÀ2 + Ph3P/Br2/amine to 1 and 0.5, lower RCC values (around 35% and 5%, respectively)
strategy) was assayed on a series of aliphatic secondary amines were obtained, with again minor effect due to modifications
with good results (RCC values in the range 45.6 Æ 7.4 to in the flow rate. Interestingly, under the same experimental
53.4 Æ 1.3). In the current work, the use of the latter strategy conditions as described in our previous work12 (same precursor
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

V. Gaja et al.
Table 2. Optimized experimental conditions and radiochemical conversion for the preparation of 13N-labelled azo compounds
by microfluidics and by solid phase support synthesis
CA (M)a
SolventA
CB (M)c
SolventB
RCC (%)e
RCC (%)f
b
d
Compound
8
9
10
0.25
0.25
0.25
1.0 M aqueous HCl
1.0 M aqueous HCl
0.1 M aqueous HCl
0.45
0.60
0.45
1.0 M aqueous NaOH
2.5 M AMF/HCOOH (pH = 4.0)
1.0 M aqueous NaOH
58.3 Æ 5.5g
40.9 Æ 11.0h
40.3 Æ 1.5h
93.9 Æ 4.1g
i
57.0 Æ 2.3
49.3 Æ 6.0j
Solid phase support results have been obtained from Gómez-Vallejo et al.12
aConcentration of precursor A.
bSolvent used for precursor A solution.
cConcentration of precursor B.
dSolvent used for precursor B solution.
eRCC with solid phase support approach.
fRCC with microfluidics approach.
gSteps 1 and 2 at room temperature.
hStep 1 at room temperature, step 2 at 90 ꢀC.
iStep 1 at room temperature, step 2 at 80 ꢀC.
jSteps 1 and 2 at 80 ꢀC.
latter. Interestingly, although RCC values obtained using micro-
fluidics at 40 and 80 ꢀC for steps 1 and 2, respectively, are lower
than those obtained using solid phase support synthesis, the use
of microfluidics allowed heating in step 1 (more difficult to
implement in solid phase support synthesis) and thus RCC values
could be improved under optimal conditions (Table 2).
In view of the results reported here, microfluidics is a promising
tool for the synthesis of 13N-labelled radiotracers. In all cases,
equivalent or higher RCC values were obtained when the microflui-
dics technology was applied to the preparation of 13N-labelled
nitrosothiols and azo compounds in comparison with traditional
(in vial or solid support) synthetic methods. However, further inves-
tigation has to be carried out in order to establish microfluidics as a
routine tool for the preparation of 13N-labelled radiotracers in the
preclinical and/or clinical environments. As a matter of fact, two
different issues still remain to be solved. First, a purification step
Figure 3. Radiochemical conversion values for compound 8 as a function of volume
of precursor B (volume of precursor A = 45mL/min in all cases) and flow rate in step 1.
concentrations, reaction temperature for steps 1 and 2), higher RCC should be included in the whole preparation flowchart. As small
values were obtained (93.9 Æ 4.1% vs. 58.3 Æ 5.5%), which is a clear volumes are usually handled, the reaction mixture could be
indication of the advantages of using microfluidics in this particular potentially directed into a loop, and HPLC might be performed
reaction (Table 2).
under analytical conditions. Eventually, UPLC could be applied with
For compound 9, low RCC values were obtained when steps 1 the subsequent reduction in chromatographic retention times. Sec-
and 2 were performed at room temperature, independent of the ond and more important, a pre-concentration step for the labelling
flow rate of steps 1 and 2 (range: 3.9 Æ 0.4% to 20.0 Æ 3.6%). An agent ([13N]NO2À) should be developed. Currently, 1.75 mL of water
increase in the temperature of step 1 to 80 ꢀC had moderate is irradiated in the cyclotron, and after the reduction step, the cad-
negative effects on RCC (probably due to decomposition of the mium column is further eluted with 1 mL of water (total starting
intermediate diazonium salt), whereas an increase in the volume = 2.75 mL); amounts of activity in the range 150–200 MBq
temperature of step 2 offered optimal RCC values (52.0 Æ 4.8 to of [13N]NO2À are typically obtained with 0.1 mA h integrated current,
57.0 Æ 2.3%, depending on the flow rates of steps 1 and 2). which means an activity concentration around 60 MBq/mL. Taking
Again, significantly higher RCC values than those obtained using into account that 20–100 mL/min reaction flows are used in the
the solid phase support synthesis were obtained (Table 2).
microfluidics system, a maximum of 1.2–6.0 MBq of radiotracer can
A different trend was observed for the preparation of be produced per minute (assuming 100% RCC), which is clearly
compound 10. Moderate RCC values were obtained when room insufficient for further application to imaging studies. In order to
temperature was used for steps 1 and 2 (29.7 Æ 3.8 to obtain starting volumes more appropriate for microfluidics, the
30.1 Æ 3.0% depending on flow rates). These values were slightly pre-concentration step should be implemented after reduction of
improved when the temperature for step 2 was increased up to [13N]NO3À to [13N]NO2À; such procedures might be based on ion
80 ꢀC (32.6 Æ 2.0 to 34.2 Æ 3.3%) and reached optimal values exchange resins or, as recently published for the pre-concentration
(44.9 Æ 5.8 to 49.3 Æ 6.0%) when the temperature of step 1 was of 18F, on electrochemical methods13; in this scenario, the improved
also increased to 80 ꢀC. This increase in the RCC when T1 was RCC obtained under microfluidics conditions would be translated
increased suggests a slower reaction rate in the formation of into higher overall radiochemical yields. Implementation of these
the corresponding diazonium salt and higher stability of the two key steps is currently under development in our laboratory.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

V. Gaja et al.
de Industria, Comercio y Turismo of the Basque Government for
financial support.
Conclusions
In this study, the procedures for the preparation of S-[13N]nitro-
sothiols and 13N-labelled azo compounds have been implemen-
ted using a microfluidics system for the first time. In general
terms, under similar or identical experimental conditions, higher
RCC values were obtained when compared with our previously
reported solid phase support synthesis procedure11,12 in similar
overall reaction times. The preparation of N-[13N]nitrosamines
was also approached under several experimental conditions,
without positive results. However, activation of the secondary
amine9 could also be assayed in the microfluidics system as an
alternative method to obtain significant RCC values.
The here reported method, coupled to an adequate
purification system (currently under development in our labora-
tory), should allow the preparation of a wide range of S-[13N]
nitrosothiols and 13N-labelled azo compounds on a dose-on-
demand basis, although a pre-concentration step of the labelling
agent might be required. This dose-on-demand concept
becomes relevant in the case of nitrogen-13, in which because
of its short half-life, multi-dose batch preparation is out of the
scope of research and clinical PET centres.
Conflict of Interest
The authors did not report any conflict of interest.
References
[1] S. S. Gambhir, Nat. Rev. Cancer 2002, 2, 683–693.
[2] T. F. Massoud, S. S. Gambhir, Genes Dev. 2003, 17, 545–580.
[3] International Atomic Energy Agency. Radioisotope products and
their availability. Nuclear technology Review. 2010; p.36.
[4] G. Lucignani, Eur. J. Nucl. Med. Mol. Imaging 2006, 33, 849–851.
[5] G. Pascali, G. Mazzone, G. Saccomanni, C. Manera, P. A. Salvadori,
Nucl. Med. Biol. 2010, 37, 547–555.
[6] S. Lu, A. M. Giamis, V. W. Pike, Curr. Radiopharm. 2009, 2,
nihpa81093.
[7] H. Anderson, N. Pillarsetty, M. Cantorias, J. S. Lewis, Nucl. Med. Biol.
2010, 37, 439–442.
[8] S. Kealey, C. Plisson, T. L. Collier, N. J. Long, S. M. Husbands, L. Martarello,
A. D. Gee, Org. Biomol. Chem. 2011, 9, 3313–3319.
[9] V. Gómez-Vallejo, K. Kato, M. Hanyu, K. Minegishi, J. I. Borrell, J. Llop,
Bioorg. Med. Chem. Lett. 2009, 19, 1913–1915.
[10] J. Llop, V. Gómez-Vallejo, M. Bosque, G. Quincoces, I. Peñuelas, Appl.
Radiat. Isot. 2009, 67, 95–99.
[11] V. Gómez-Vallejo, K. Kato, I. Oliden, J. Calvo, Z. Baz, J. I. Borrell, J. Llop,
Tetrahedron Lett. 2010, 51, 2990–2993.
[12] V. Gómez-Vallejo, J. I. Borrell, J. Llop, Eur. J. Med. Chem. 2010, 45,
5318–5323.
Acknowledgements
The authors would like to thank Mikel González for the technical
support in the radiochemistry laboratory and the Departamento
[13] R. Wong, R. Iwata, H. Saiki, S. Furumoto, Y. Ishikawa, E. Ozeki, Appl.
Radiat. Isot. 2012, 70, 193–199.
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org