An efficient and practical synthesis of [2-11C]indole via superfast nucleophilic [11C]cyanation and RANEY Nickel catalyzed reductive cyclization

Article abstract of DOI:10.1039/c5ob01654a

A rapid method for the synthesis of carbon-11 radiolabeled indole was developed using a sub-nanomolar quantity of no-carrier-added [¹¹C]cyanide as radio-precursor. Based upon a reported synthesis of 2-(2-nitrophenyl)acetonitrile (2), a highly reactive substrate 2-nitrobenzyl bromide (1) was evaluated for nucleophilic [¹¹C]cyanation. Additionally, related reaction conditions were explored with the goal of obtaining of highly reactive 2-(2-nitrophenyl)-[1-¹¹C]acetonitrile ([¹¹C]-2) while inhibiting its rapid conversion to 2,3-bis(2-nitrophenyl)-[1-¹¹C]propanenitrile ([¹¹C]-3). Next, a RANEY Nickel catalyzed reductive cyclization method was utilized for synthesizing the desired [2-¹¹C]indole with hydrazinium monoformate as the active reducing agent. Extensive and iterative screening of basicity, temperature and stoichiometry was required to overcome the large stoichiometry bias that favored 2-nitrobenzylbromide (1) over [¹¹C]cyanide, which both caused further alkylation of the desired nitrile and poisoned the RANEY Nickel catalyst. The result is an efficient two-step, streamlined method to reliably synthesize [2-¹¹C]indole with an entire radiochemical yield of 21 ± 2.2% (n = 5, ranging from 18-24%). The radiochemical purity of the final product was >98% and specific activity was 176 ± 24.8 GBq μmol^-1 (n = 5, ranging from 141-204 GBq μmol^-1). The total radiosynthesis time including product purification by semi-preparative HPLC was 50-55 min from end of cyclotron bombardment.

Full text of DOI:10.1039/c5ob01654a

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
An efficient and practical synthesis of [2-11C]indole
via superfast nucleophilic [11C]cyanation and
RANEY® Nickel catalyzed reductive cyclization†
Cite this: DOI: 10.1039/c5ob01654a
So Jeong Lee,a,b Joanna S. Fowler,a David Alexoff,a Michael Schueller,a
Dohyun Kim,a Alexander Nauth,a,c Carina Weber,a,c Sung Won Kim,a
Jacob M. Hooker,a,d Ling Mad and Wenchao Qu*a
A rapid method for the synthesis of carbon-11 radiolabeled indole was developed using a sub-nanomolar
quantity of no-carrier-added [11C]cyanide as radio-precursor. Based upon a reported synthesis of 2-(2-
nitrophenyl)acetonitrile (2), a highly reactive substrate 2-nitrobenzyl bromide (1) was evaluated for
nucleophilic [11C]cyanation. Additionally, related reaction conditions were explored with the goal of
obtaining of highly reactive 2-(2-nitrophenyl)-[1-11C]acetonitrile ([11C]-2) while inhibiting its rapid conver-
sion to 2,3-bis(2-nitrophenyl)-[1-11C]propanenitrile ([11C]-3). Next, a RANEY® Nickel catalyzed reductive
cyclization method was utilized for synthesizing the desired [2-11C]indole with hydrazinium monoformate
as the active reducing agent. Extensive and iterative screening of basicity, temperature and stoichiometry
was required to overcome the large stoichiometry bias that favored 2-nitrobenzylbromide (1) over [11C]
cyanide, which both caused further alkylation of the desired nitrile and poisoned the RANEY® Nickel cata-
lyst. The result is an efficient two-step, streamlined method to reliably synthesize [2-11C]indole with an
entire radiochemical yield of 21 2.2% (n = 5, ranging from 18–24%). The radiochemical purity of the final
product was >98% and specific activity was 176 24.8 GBq µmol−1 (n = 5, ranging from 141–204 GBq
µmol−1). The total radiosynthesis time including product purification by semi-preparative HPLC was
50–55 min from end of cyclotron bombardment.
Received 6th August 2015,
Accepted 18th September 2015
DOI: 10.1039/c5ob01654a
www.rsc.org/obc
The synthesis of indole and its derivatives has played a promi- help us to answer some core questions, such as the relative
nent role in organic chemistry because of the importance of importance of different auxin biosynthesis pathways in various
natural products containing the indole ring, such as alkaloids plant tissues as well as the role of auxin biosynthesis in plant–
and other biologically active molecules.1 One of the most microbe interactions. In this regard, we set out to synthesize
important plant hormones, indole-3-acetic acid (auxin or IAA), carbon-11 (half life: 20.4 min) radiolabeled indole as a tool for
is such an example and it plays a pivotal role in plant growth improved mapping and understanding of both biosynthetic
and development.2 Indole is known to be the core precursor and metabolic pathways of auxin in whole plants in vivo using
for all five known auxin biosynthetic pathways including positron emission tomography (PET) at tracer levels.
tryptophan-dependent
and
tryptophan-independent
Although the synthesis of molecules containing an indole
pathways.3–6 Tracking the downstream flow of indole could ring as a core structure has been extensively investigated,7,8
there are only limited reports related to the incorporation of
stable and radioactive isotopes into the indole ring. In 2002,
aBiological, Environmental & Climate Sciences Department, Brookhaven National
Czeskis and colleagues reported an elegant method for syn-
thesis of 4-hydroxy [2-14C]indole.9 Starting from Na14CN, they
Laboratory, Upton, NY 11973, USA. E-mail: wqu@bnl.gov
bDepartment of Chemistry, Stony Brook University, Stony Brook, NY 11794, USA
first synthesized p-chlorophenoxy [1-14C]acetonitrile which was
cInstitut für Kernchemie, Johannes Gutenberg-Universität, D-55128 Mainz, Germany
reacted with benzyl-protected 3-nitrophenol via a vicarious
nucleophilic substitution reaction to afford a 2-(2-nitrophenyl)
acetonitrile derivative, which was reduced and cyclized to form
14C-labeled 4-hydroxyindole. Four years later, Pedras and
Okinyo reported a synthesis of stable isotope deuterium
labeled [5,6,7,8-2H4]indole using chloroacetonitrile direct alkyl-
ation and reductive cyclization reactions.10 More recently,
dAthinoula A. Martinos Center for Biomedical Imaging, Department of Radiology,
Massachusetts General Hospital, Harvard Medical School, Charlestown,
Massachusetts 02129, USA
†Electronic supplementary information (ESI) available: Diagrams of [11C]HCN
production procedure, detailed experimental procedure, representative HPLC
analysis results of intermediate and product for measuring radiochemical
purities as well as semi-prep HPLC profile for purification of final product, etc.
See DOI: 10.1039/c5ob01654a
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 1 Previous reports for synthesis of 11C-labeled indole.
Czeskis incorporated carbon-14 into the 2-position on the
indole ring through the reaction of 2-nitrobenzyl bromide with
Results and discussion
Na14CN to form 2-(2-nitrophenyl)-[1-14C]acetonitrile followed We started by testing the feasibility of synthesizing the inter-
by reductive cyclization.11
mediate nitrile, 2. One report by Kalir and Mualem in 1987 18
Additionally, two methods for synthesis of 11C-labeled disclosed a reliable method using NaCN plus HCN (for pH
indole, [2-11C]indole, were reported in 1998 and 1999 adjustment) using bromide, 1 as starting material, which pro-
(Scheme 1):12,13 in the first report, the radioprecursor [11C] vided the nitrile, 2 with yields up to 90%. A later report
CH3NO2 (synthesized from [11C]CH3I) was condensed with showed that this compound was also synthesized in 52% yield
2-nitrobenzaldehyde to form β,2-dinitro-[β-11C]styrene which in aqueous DMSO at near neutral conditions for 1 h at
when treated with titanium(III) chloride underwent reductive 0–5 °C.11 Following these reports, we initiated the “cold” test
cyclization to form desired [2-11C]indole. In the second report for synthesis of nitrile, 2 intermediate and we found, with
[11C]CH3I was reacted with triphenylarsine followed by butyl slightly modified conditions (Scheme 3), that a 10–60% yield
lithium to form triphenylarsonium [11C]methylide which was of nitrile, 2 could be synthesized by reacting bromide, 1 with
then reacted with 2-aminobenzaldehyde to give [2-11C]indole. either KCN or a mixture of KCN/HCN.
Although both reports showed the feasibility of synthesizing
Next, we introduced radioactive [11C]HCN into the reaction
[2-11C]indole, they only provided the analytical HPLC data for system and tested a carrier-added (CA) [11C]cyanation reaction
the reaction mixture and there were no details on either the (Scheme 3). The radio-HPLC analysis was very encouraging. It
purification method or isolated yield of [2-11C]indole.
showed that 55% of the radioactivity converted to the desired
Given the availability of [11C]HCN and a long history of uti- [11C]nitrile, [11C]-2 in the presence of KCN in EtOH for 10 min
lizing [11C]HCN as a radiolabeling precursor to synthesize at 100 °C. We then eliminated the carrier KCN from the reac-
various 11C-labeled bioactive radiotracers,14–17 as well as above- tion to test the conditions for the synthesis of the no-carrier-
mentioned successful examples of cyanide based methods for added (NCA) [11C]-2 (Scheme 4). Under slightly acidic con-
synthesizing various isotopically labeled indole derivatives,9–11 ditions, there was no reaction. Increasing the reaction temp-
we envisioned the synthesis of [2-11C]indole using [11C]HCN as erature (from 50 °C to 150 °C) initiated the [11C]cyanation
the radiolabeling precursor (Scheme 2): starting from 2-nitro- reaction but gave only 2,3-bis(2-nitrophenyl)-[1-11C]propane-
benzyl bromide (referred to hereafter as bromide, 1), nucleo- nitrile ([11C]dimer, [11C]-3), which was formed by the further
philic cyanation with [11C]cyanide ([11C]CN−) could provide us alkylation (here called “dimerization”) of [11C]-2 with starting
with 2-(2-nitrophenyl)-[1-11C]acetonitrile ([11C]nitrile, 2). Next, material bromide, 1. When KHCO3 was used to trap the [11C]
reductive cyclization, if successful, would give us desired radio- HCN and the reaction mixture was maintained at slightly basic
tracer, [2-11C]indole. Herein, we describe our recent explora- pH, radio-HPLC analysis showed that the NCA [11C]-2 was
tions in developing a rapid method for synthesis of [2-11C] sometimes formed, but these results were not reproducible
indole based upon a [11C]CN− nucleophilic cyanation followed and, in most cases, only a small amount of [11C]-2 could be
by reductive cyclization.
detected and the majority of radioactivity in the reaction
Scheme 2 Our design of synthesis [2-11C]indole using [11C]HCN as radiolabeling precursor.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 3 Results for the synthesis of nitrile, 2 and carrier added [11C]-2.
Scheme 4 Preliminary results for the radiosynthesis of [11C]-2.
mixture was [11C]CN−, [11C]-3 and an unidentified low polarity greatly favors the nucleophilic cyanation reaction, it also favors
radioactive compound(s) (Scheme 4).
further alkylation once a certain amount of [11C]-2 has accu-
These initial results were not surprising nor were they un- mulated in the reaction mixture (Scheme 4). At this stage, we
precedented. Kalir and Mualem also reported the formation of speculated that it was necessary to carefully examine the NCA
dimer, 3 from the reaction of bromide, 1 with cyanide. They [11C]cyanation conditions in order to out-compete the very
minimized its formation by carefully controlling the reaction rapid further alkylation and provide a sufficient amount of
stoichiometry with excess NaCN to obtain a ratio of 3 : 1 [11C]-2 for the next reductive cyclization step.
(NaCN : bromide, 1) and conducting the reaction under acidic
Because we anticipated that careful pH adjustment would
conditions.18 Czeskis also reported dimer, 3 formation and be necessary to avoid further alkylation of the [11C]-2, we envi-
minimized it by using aqueous DMSO and quenching the reac- sioned that it was necessary to remove ammonia and water
tion after a 1 h reaction at 0–5 °C (Scheme 4).11 Adapting these which were carried in the gas phase production of our radio-
conditions to [11C]cyanation with no-carrier-added [11C] precursor [11C]HCN.19 For this purpose, a [11C]HCN purifi-
cyanide was challenging because the chemical quantity of cation method, first introduced by Långström,20 was adapted
cyclotron-produced [11C]cyanide is fixed and very small (0.03 for our [11C]HCN delivery system: the helium gas stream con-
µmol–0.06 µmol) whereas milligram quantities (10 mg, 46.3 taining radioactivity was first purged through a 50% H2SO4
µmol) of the bromide, 1 is typically used to drive the nucleo- acid bath heated at 65 °C to remove NH3, and then passed
philic substitution reaction. This results in a ratio of [11C] through a freshly packed P2O5 drying tube for further removal
cyanide : bromide, 1 of ∼1 : 700–1500 in a typical radiosyn- of trace H2O. With this purified [11C]HCN in hand, we started
thesis. Though this low ratio of radioprecursor to substrate to screen nucleophilic [11C]cyanation conditions for the syn-
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
thesis of NCA [11C]-2 using CsHCO3/18-crown-6/DMF which [11C]-2 intermediate. Though the longer reaction time (from
was first introduced in our recent paper on the optimized syn- 8 min to 25 min) provided higher ORCY (58%), further alkyl-
thesis of carbon-11 labeled L-glutamine.21 Results are reported ation occurred at the expense of [11C]-2. Only 9% of the total
in Table 1.
radioactivity was [11C]-2 and the rest was [11C]-3. Based upon
We also reduced the mole ratio of [11C]cyanide : bromide to these results (Table 1, entries 1 to 4), it was clear that bromide,
∼1 : 150 (by reducing the amount of bromide, 1 from 10 mg to 1 is a highly reactive substrate, and the ORCY reached 90%
1 mg for this set of experiments) to minimize the further alkyl- after an 8 min reaction time even at room temperature. It was
ation to form [11C]-3. Although directly duplicating the also evident that conditions that favored a high ratio of [11C]-2
L-[5-11C]-glutamine synthesis conditions gave a 79% overall to [11C]-3 also dramatically reduced the ORCY. It was definitely
radiochemical yield (ORCY, includes total radioactivity of necessary to adjust other reaction parameters to obtain a high
[11C]-2 and [11C]-3, decay corrected), only undesired [11C]-3 was ORCY while avoiding consumption of the intermediate [11C]-2.
detected using radio-HPLC analysis (Table 1, entry 1). Clearly,
We next set out to investigate the effect of basicity on ORCY
milder reaction conditions were needed to avoid consumption and radiolabeled product distribution. The results listed in
of radioactive intermediate [11C]-2. Reducing the reaction Table 1, entries 5 and 6 showed that no [11C]cyanation reaction
temperature from 60 °C to 30–35 °C gave similar results (entry occurred when the pH was less than 8. Neither the direct use
2). Further decreasing the temperature to 18–20 °C not only of NCA [11C]HCN (pH = 5) nor the addition of near neutral
provided us with 90% ORCY but also 4% of the desired [11C]-2 form salt KBr (pH = 7.7) initiated the [11C]cyanation reaction.
(entry 3). For the next step, the reaction mixture was cooled With the introduction of more basic K2CO3 into the reaction
with an ice bath and the [11C]cyanation reaction rate was (entries 7 and 8 compared with entry 6), the [11C]cyanation
clearly reduced. Although the ORCY was only 22% with an reaction rate was dramatically improved. Adjusting the pH to
8 min reaction time at 0 °C (entry 4), the results were quite 8.4 resulted in a 26% ORCY with a mixture of [11C]-2 and
encouraging since all detected radioactivity was the desired [11C]-3 (ratio at 23 : 77) favoring the formation of [11C]-3 (entry 7).
Table 1 Exploratory results of [11C]cyanation based upon conditions reported in the recent update of the synthesis of L-[5-11C]-glutamine21
Entry
Cation sourcea
pHb
Time (min)
Temp (°C)
ORCYc
Radioactivity ratio ([11C]-2 : [11C]-3 d)
1
2
3
4
CsHCO3
CsHCO3
CsHCO3
CsHCO3
8.5
8.5
8.5
8.5
8
8
8
8
25
8
10
10
10
0.2
0.6
2
60
79
73
90
22
58
0
0 : 100
0 : 100
4 : 96
100 : 0
9 : 91
nd
nd
23 : 77
0 : 100
92 : 8
71 : 29
21 : 79
16 : 84
7 : 93
85 : 15
49 : 51
28 : 72
78 : 22
78 : 22
39 : 61
28 : 72
22 : 28
30–35
18–20
0
5e
6
7
8
9
None
KBr
K2CO3/KBr (1/98)
K2CO3/KBr (5/90)
K2CO3/KHCO3 (1/4)
5
60
18
18
18
0
7.7
8.4
8.7
9.6
0
26
67
73
72
76
77
69
68
81
70
76
77
77
74
64
5
15
0.6
2
10
0.2
0.6
2
10
K2CO3
11.2
0
0
11f
K2CO3/KHCO3 (1/4)
9.6
5
15
a All the experiments were performed with 2-nitrobenzyl bromide (1 mg, 4.63 μmol) with different cation sources (15 μmol) in presence of 18-C-6
(30 μmol) dissolved in 0.5 mL of DMF; if two different cation sources were used, the numbers listed in parenthesis were the mole ratios. b pH was
measured with the cation source dissolved in aqueous solution (give the volume of water because pH is concentration dependent). c Overall
radiochemical yield (%, ORCY, decay-corrected) of [11C]cyanation was determined by analytical radio-HPLC and included both [11C]-2 and [11C]-3
(entries 1–11). d This number is the radioactivity ratio of [11C]-2 : [11C]-3 (decay-corrected), nd: not detected. e A solution of H11CN in DMF was
weakly acidic without addition of a cation source and 18-C-6. f Reaction solvent was changed to N,N-dimethylacetamide (DMA).
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 2 Exploration of reductive cyclization method to convert [11C]-2
to [2-11C]indole
Increasing the pH to 8.7 increased the ORCY to 60% but only
[11C]-3 was formed (entry 8). These experiments showed that
basic conditions were required to initiate the [11C]cyanation
reaction but confirmed that the formation of [11C]-3 was
favored. Encouraged by the increase in ORCY, we further
reduced the temperature to 0 °C, modified the cation source
and increased the pH to 9.6, and then carefully monitored the
reaction by analytical radio-HPLC (entry 9). The results again
showed that this reaction was very fast. The ORCY reached to
73% after only 0.2 min with a product distribution favoring
[11C]-2 (92 : 8). Longer reaction times reversed the ratio to favor
the undesired [11C]-3. When K2CO3 was chosen as a cation
source and the pH was increased to 11.2 (entry 10), the radio-
HPLC analysis results of samples taken at different time points
showed a similar trend as the results listed in entry 9: the
short reaction time (0.6 min) provided the higher [11C]-2 yield
but that the longer reaction times favored further alkylation.
Finally, the solvent N,N-dimethylacetamide (DMA) was tested
and the results also showed that the ORCY plateaued at 76%
even after 0.2 min with a ratio of [11C]-2 : [11C]-3 at 78 : 22
(entry 11). We noted that further alkylation is comparatively
slower with DMA as the reaction solvent when comparing the
results from entry 11 with the results listed in entry 9.
This series of experiments showed that it was difficult to
avoid consumption of the [11C]-2 with the excess bromide, 1.
However, by carefully controlling several reaction parameters,
i.e., reaction temperature, basicity as well as reaction time, it is
possible to obtain a good conversion of [11C]cyanide to
product [11C]-2, while minimizing formation of side product
[11C]-3. Eventually, to achieve better reproducibility and to
reach higher yields of [11C]-2, the [11C]cyanation reaction con-
ditions listed in entry 11 (pH 9.6, bromide, 1 1.0 mg, DMA as
solvent, reaction at 0 °C for 0.6 min) were selected for the syn-
thesis of intermediate [11C]-2. We next turned our attention to
the exploration of the reductive cyclization step.
We initially tested several reducing agents for the reductive
cyclization of [11C]-2 to [2-11C]indole including: Pd–C/H2 in
ethyl acetate22,23 and in alcoholic solvents;24,25 Pd–C/H2 in
ethanol/acetic acid;9,26,27 Pd–C in the presence of in situ
H2;28,29 RANEY® Nickel or Pd–C in the presence of NH4CO2H/
HCO2H;30 Al–NiCl2·6H2O;31 Zn/acetic acid.32 Results were gen-
erally disappointing: although most worked under carrier-
added conditions with or without carbon-11, they either pro-
vided only a trace of desired [2-11C]indole or none at all under
no-carrier-added (NCA) conditions. Later, a report by Gowda
et al. using RANEY® Nickel based rapid reduction (both nitro
and nitrile groups) caught our attention.33 This method
quickly and selectively reduced various compounds containing
nitro or nitrile groups to the corresponding amines in high
yield (70–94%) with hydrazinium monoformate as hydrogen
donor and RANEY® Nickel as catalyst at r.t. in 2–6 min. Given
the time limitations of carbon-11 radiochemistry, i.e.
“Working Against Time”,34 this method was especially promis-
ing. We confirmed that this method rapidly converted the
nitrile, 2 to indole under “cold” conditions and set out to
adapt it to our NCA [2-11C]indole synthesis process (Table 2).
Time
Entry (min)
N2H4·HCO2Ha
(mmol)
RANEY®
Nib (mg)
Tempc
(°C)
RCYd
(%)
12
13
14
15
16
17
18
19
20
21
22
23
24e
5
10
15
5
15
15
15
5
10
15
10
10
10
5
5
5
10
10
2.5
2.5
5
5
5
48
48
48
48
48
48
96
48
48
48
48
38
48
r.t
r.t
r.t
r.t
r.t
r.t
r.t
40
40
40
40
40
40
25
26
33
9
19
38
10
18
30
24
35
41
nd
0.5
0.5
0.5
a The mole ratio of hydrazine monohydrate and formic acid was 1 : 1.
b RANEY® Nickel slurry in water (50%) was used for reactions. c Room
temperature indicates 18–20 °C. d RCY (decay-corrected) (%) was
determined by radio-HPLC analysis of filtered crude mixture. e 2 mg of
bromide, 1 was used for synthesis of [11C]-2; after the Sep-Pak
purification, the ORCY was 51% and 42% of radioactivity was [11C]-2;
nd: no [2-11C]indole was detected.
The initial test result was quite promising (Table 2, entry
12). We obtained 25% of desired [2-11C]indole by using
5 mmol N2H4·HCO2H and 48 mg RANEY® Ni at r.t. for 5 min.
The results from next two experiments (entries 13 and 14)
showed that a longer reaction time of 15 min resulted in a
moderate increase in yield to 33%. Surprisingly, when the
amount of N2H4·HCO2H, was increased to 10 mmol, the yield
was decreased (entries 15 and 16) and only afforded less than
20% of [2-11C]indole even after 15 min. In contrast, reducing
the N2H4·HCO2H from 5 mmol to 2.5 mmol increased the reac-
tion yield to 38% (entry 17). A two-fold increase in the
RANEY® Ni catalyst negatively affected the yield (entry 18).
Moreover, a large amount of radioactivity remained in the fil-
tration solid cake after the filtering off the catalyst from the
reaction mixture and only 10% of [2-11C]indole was detected in
the filtrate. When the reaction temperature was raised to
40 °C, we found that the reaction yield was reached a peak at
10 min (30%) and further extension of reaction time decreased
the reaction yield to 24% (entries 19–21). Next, a sharp
reduction of N2H4·HCO2H from 5 mmol to 0.5 mmol clearly
improved the reaction and gave a 35% yield (entry 22). A 20%
reduction of RANEY® Ni catalyst also showed positive impact
and increased reaction yield to 41% (entry 23) which was the
highest reductive cyclization yield obtained for this series of
experiments.
Since it appeared that we had reached a ceiling for the
reductive cyclization reaction, we tried to further improve the
[11C]cyanation yield by simply increasing the amount of start-
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 5 Reductive cyclization inhibition test.
ing material bromide, 1 thereby helping to increase overall
yields for the two steps. We doubled the amount of bromide, 1
for the [11C]cyanation and 42% of [11C]-2 was obtained after
the Sep-Pak purification. However, when this reaction mixture
was submitted to reductive cyclization, we were disappointed
to find that no [2-11C]indole was detected (entry 24). Using
2 mg of bromide, 1 for the [11C]cyanation totally inhibited the
following reductive cyclization reaction leading us to speculate
that the presence of bromide, 1 in the reaction mixture may be
responsible for the yield ceiling of ∼40% which we observed
on the reductive cyclization.
For all of our reactions, we incorporated a Sep-Pak solid
phase extraction (SPE) method to quickly remove the reaction
solvent DMA, unreacted [11C]cyanide ion, inorganic salts,
phase transfer catalyst (PTC) 18-crown-6 as well as H2O after
the [11C]cyanation step. After that, the ethanol wash process
not only eluted radioactive [11C]-2 and [11C]-3 but also a large
amount of the unreacted bromide, 1 into the reduction vessel
from this Sep-Pak cartridge. To test our hypothesis that the
residual bromide, 1 would interfere or even totally inhibit the
reductive cyclization, a “cold” test for reductive cyclization step
was tested again with the addition of extra bromide, 1 (2 mg)
and no indole was detected confirming that the presence of
the residual bromide, 1 which was carried over to the second
vessel poisoned the catalyst (Scheme 5).
Fig. 1 The amount of residual bromide, 1 after purification of “cold”
cyanation reaction mixture by SPE method.
1 (Table 3). By decreasing the bromide, 1 from 1 mg to 0.1 mg,
the reaction rate was clearly decreased (entry 25). After a
0.6 min reaction, only 25% of the radioactivity was converted
to desired intermediate [11C]-2. Longer times only decreased
To determine how much of the excess bromide, 1 starting
material was carried over into the reductive cyclization reac-
tion, we next carried out a series of “cold” cyanation reactions
with different starting amount of bromide, 1. After the follow-
ing Sep-Pak SPE purification process, the ethanol elute was
analyzed with analytical HPLC and the results showed that
55–78% of the bromide, 1 which was used for the cyanation
reaction was carried over into the second reaction vessel after
the Sep-Pak SPE purification (Fig. 1). For example, when
1.0 mg of bromide, 1 was used for cyanation reaction, 0.65 mg
of it was eluted from the Sep-Pak and introduced into the next
reaction vessel.
Table 3 Further exploration of the
reduced amount of bromide, 1
[
11C]cyanation reaction with
Ratio of [11C]-2 : [11C]-3 (ORCY)a
Bromide, 1 Temp
Entry (mg)
(°C)
0.6 min 2 min
5 min
15 min
25
26
27
28
29
0.1
0.1
0.1
0.1
0.2
0
100 : 0
(25)
97 : 3
(37)
96 : 4
(48)
96 : 4
(47)
100 : 0
(12)
96 : 4
(26)
100 : 0
(6)
93 : 7
(15)
100 : 0
(3)
83 : 17
(6)
20
40
60
40
91 : 9
(43)
98 : 2
(41)
95 : 5
(19)
97 : 3
(33)
91 : 9
(11)
94 : 6
(16)
Based upon the high reactivity of the bromide, 1 to SN2
[11C]cyanation (Table 1), we predicted that it would be possible
to get comparable RCY of [11C]-2 using less bromide, 1 by
simply modifying the [11C]cyanation reaction conditions. We
next designed a series of experiments to measure RCY and
77 : 23
(78)
72 : 28
(78)
42 : 58
(71)
27 : 73
(62)
a The ratio of [11C]-2 : [11C]-3 was calculated based upon radio-HPLC
analysis of reaction solution samples taken at different time points
(decay-corrected); ORCY (%, overall radiochemical yield, decay-
product distribution when reducing the amount of bromide, corrected) included radioactive peaks of both [11C]-2 and [11C]-3.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
formed in all experiments. Decreasing the amount of bromide,
1 from 1.0 mg to 0.1 mg, which is similar to the strategy that
Kalir and Mualem used to increase the yield of nitrile, 2,18
clearly decreased the formation of undesired by-product
[11C]-3. To see if the yield of [11C]-2 could be further improved,
the [11C]cyanation reaction was further tested at 40 °C with a
slight increase in the amount of bromide, 1 from 0.1 mg to
0.2 mg. The results were delightful: the ORCY reached 78% with
ratio of [11C]-2 : [11C]-3 at 77 : 23 (entry 29, Fig. 2). Further exten-
sion of reaction time did not show any benefits, it not only
reduced the ORCY but also decreased the ratio of [11C]-2 : [11C]-3.
With the above [11C]cyanation results in hand, we re-investi-
gated the synthesis of [2-11C]indole with a decreased amount
of bromide, 1 starting material. Once a superfast (0.6 min)
[11C]cyanation was performed, the reaction mixture was
quickly purified with a SPE Sep-Pak. Following elution with
EtOH, the intermediate [11C]-2 was subjected to a reductive
cyclization reaction using our optimized reaction conditions
(Table 2, entry 23). When 0.1 mg of bromide, 1 was used for
[11C]cyanation reaction, only 21% of desired [11C]-2 was
obtained (Table 4, entry 30). The subsequent reductive cycliza-
tion followed by semi-prep HPLC purification process yielded
57% of [2-11C]indole and the entire two-step preparative RCY
for this synthesis was 12% (entry 30). When the 0.2 mg of
bromide was used, 48% of the desired [11C]-2 was obtained
after the SPE purification. The next reductive cyclization step
yielded 54% of [2-11C]indole and the entire two-step prepara-
tive RCY was 26% (Table 4, entry 31). These two experiments
showed that the reductive cyclization reaction was clearly
improved by decreasing the amount of bromide, 1 and that the
preparative (isolated) yield of the reductive cyclization step was
over 50%. In addition, increasing the amount of bromide,
1 from 0.1 to 0.2 mg doubled the yield of [11C]-2 and also
increased the entire radiochemical yield of [2-11C]indole from
12% to 26% (Table 4, entry 31).
Fig. 2 Analytical radio-HPLC profiles of two reaction samples from the
[
UV-HPLC profile of sample co-injected with nitrile, 2 and dimer, 3;
bottom: radio-HPLC profile.
11C]cyanation reaction with 0.1 mg and 0.2 mg of bromide, 1; top:
the amount of [11C]-2. However, increasing the reaction temp-
erature clearly improved the RCY (entries 26–28). The
optimum conditions were found at 40 °C with 0.6 min reaction
time (entry 27, Fig. 2). Further increasing the reaction tempera-
ture to 60 °C did not improve the reaction. One important
finding was that only a negligible amount of [11C]-3 was
With the combination of a reduced amount of starting
bromide, 1 (0.2 mg), strictly controlled and also superfast reac-
tion time (0.6 min), mild temperature (40 °C) as well as a polar
aprotic solvent DMA, the [11C]cyanation reaction provided us a
Table 4 Further exploration of hydrogenation with reduced starting amount of bromide, 1 a
Entry
Bromide, 1 (mg)
1st step RCY (%)
2nd step RCY (%)
Entire RCYb (%)
30
31
0.1
0.2
21
48
57
54
12
26
a Both 1st step and 2nd step RCYs were preparative results (purified by SPE and semi-prep HPLC method, respectively). b Entire RCY (%), entire
radiochemical yield (decay-corrected) for synthesis of [2-11C]indole.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 6 Optimized reaction parameters for synthesis of [2-11C]indole.
nearly 50% yield of [11C]-2 intermediate. Next, the reductive sumed the desired [11C]-2 by further alkylation and also poi-
cyclization of [11C]-2 under selected conditions (0.5 mmol soned the reductive cyclization step. An iterative systematic
N2H4·HCO2H, 38 mg RANEY® Ni in EtOH/H2O solution at investigation and optimization of each step, however, resulted
40 °C for 10 min) yielded over 50% of [2-11C]indole. For the in a robust, rapid and reproducible method for the radiosyn-
whole synthetic process, 26% of [2-11C]indole (calculated from thesis of [2-11C]indole. Currently, we are working on plant and
starting [11C]HCN, decay-corrected) was obtained in 54 min, bacteria imaging studies using the [2-11C]indole tracer and
calculated from the time of end of bombardment (EOB) to the these results will be reported separately.
completion of the semi-prep HPLC purification.
To confirm the robustness of this newly developed synthetic
method, we repeated the synthesis of [2-11C]indole five times
with the optimized reaction conditions described above. The
Acknowledgements
results show the high reliability of this synthetic process
This manuscript has been co-authored by employees of Broo-
(Scheme 6). With a six-min cyclotron beam time, which gene-
khaven Science Associates, LLC under Contract DE-SC0012704
rates ∼24.4 GBq (660 mCi) of [11C]CO2, 0.17–0.40 GBq
with the U. S. Department of Energy, Office of Biological and
(4.7–10.9 mCi) of [2-11C]indole product was obtained at the
Environmental Research within the Office of Science.
end of synthesis (EOS). The radiochemical yield for both steps
Additional support was provided by a Mesoscale Pilot project
was 21
2.2% (calculated from [11C]HCN radioactivity col-
(PS005) from the Department of Energy, Office of Biological
and Environmental Research, as well as the German Academic
Exchange Service (Deutscher Akademischer Austauschdienst,
DAAD) Bonn, which supported an internship for Carina Weber
and Alex Nauth. We also thank Richard Ferrieri for advices
and encouragement, as well as Youwen Xu and Colleen Shea
for technical and manuscript edit assistance.
lected in the first reaction vessel) and was 8.2 2.0% (calcu-
lated from [11C]CO2 radioactivity generated from a six-min
cyclotron beam). The radiochemical purity of final product was
>98% and the specific activity of final product was 176 24.8
GBq μmol−1 (4.8 0.7 Ci µmol−1). The complete processing
time, calculated from EOB to the EOS, ranged from 50 to
55 min.
References
Conclusion
1 N. K. Kaushik, N. Kaushik, P. Attri, N. Kumar, C. H. Kim,
A. K. Verma and E. H. Choi, Molecules, 2013, 18, 6620–
6662.
2 A. W. Woodward and B. Bartel, Ann. Bot., 2005, 95, 707–
735.
3 N. D. Tivendale, N. W. Davies, P. P. Molesworth,
S. E. Davidson, J. A. Smith, E. K. Lowe, J. B. Reid and
J. J. Ross, Plant Physiol., 2010, 154, 1957–1965.
4 J. Normanly, Cold Spring Harbor Perspect. Biol., 2010, 2,
a001594.
In summary, we successfully developed
a practical and
efficient method for synthesis of a biologically interesting PET
tracer [2-11C]indole from NCA [11C]HCN with high specific
activity. The overall process illustrates the challenges in deter-
mining optimum conditions for controlling a superfast [11C]
cyanation reaction that yielded a highly reactive intermediate
[11C]nitrile, [11C]-2. More specifically, under NCA conditions,
the amount of [11/12C]HCN is very low (0.03–0.06 µmol) and
fixed, biasing the stoichiometry in favor of the highly reactive
bromide, 1. In this case, the amount of bromide, 1 initially
used in the formation of the intermediate [11C]-2 both con-
5 K. A. Phillips, A. L. Skirpan, X. Liu, A. Christensen,
T. L. Slewinski, C. Hudson, S. Barazesh, J. D. Cohen,
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
S. Malcomber and P. McSteen, Plant Cell, 2011, 23, 550– 21 T. Gleede, B. Riehl, C. Shea, L. Kersting, A. S. Cankaya,
566.
D. Alexoff, M. Schueller, J. S. Fowler and W. Qu, Amino
Acids, 2015, 47, 525–533.
6 P. McSteen, Cold Spring Harbor Perspect. Biol., 2010, 2,
a001479.
22 H. R. Snyder, E. P. Merica, C. G. Force and E. G. White,
J. Am. Chem. Soc., 1958, 80, 4622–4625.
23 G. N. Walker, J. Med. Chem., 1965, 8, 583–588.
7 G. R. Humphrey and J. T. Kuethe, Chem. Rev., 2006, 106,
2875–2911.
8 G. Shukla, R. K. Verma, G. K. Verma and M. S. Singh, Tetra- 24 J. E. Macor, J. T. Froman, R. J. Post and K. Ryan, Tetrahe-
hedron Lett., 2011, 52, 7195–7198. dron Lett., 1997, 38, 1673–1676.
9 B. A. Czeskis, D. K. Clodfelter and W. J. Wheeler, J. Labelled 25 H. Sajiki, T. Ikawa and K. Hirota, Org. Lett., 2004, 6, 4977–
Compd. Radiopharm., 2002, 45, 1143–1152. 4980.
10 M. S. C. Pedras and D. P. O. Okinyo, J. Labelled Compd. 26 G. N. Walker, J. Am. Chem. Soc., 1955, 77, 3844–3850.
Radiopharm., 2006, 49, 33–45.
11 B. A. Czeskis, J. Labelled Compd. Radiopharm., 2012, 55,
171–179.
12 J. Zessin and J. Steinbach, J. Labelled Compd. Radiopharm.,
1998, 41, 669–676.
13 J. Zessin, J. Steinbach and B. Johannsen, J. Labelled Compd.
Radiopharm., 1999, 42, 725–736.
14 J. F. Lamb, R. W. James and H. S. Winchell, Int. J. Appl.
Radiat. Isotop., 1971, 22, 475–479.
15 D. R. Christman, R. D. Finn, K. I. Karlstrom and A. P. Wolf,
J. Nucl. Med., 1973, 14, 864–866.
16 P. W. Miller, N. J. Long, R. Vilar and A. D. Gee, Angew.
Chem., Int. Ed., 2008, 47, 8998–9033.
27 G. Groszek, M. Bednarski, M. Dybala and B. Filipek,
Eur. J. Med. Chem., 2009, 44, 809–817.
28 P. K. Mandal and J. S. McMurray, J. Org. Chem., 2007, 72,
6599–6601.
29 O. Khdour, A. Ouyang and E. B. Skibo, J. Org. Chem., 2006,
71, 5855–5863.
30 R. A. W. Johnstone, A. H. Wilby and I. D. Entwistle, Chem.
Rev., 1985, 85, 129–170.
31 B. K. Sarmah and N. C. Barua, Tetrahedron, 1991, 47, 8587–
8600.
32 J. C. Reader, T. P. Matthews, S. Klair, K.-M. J. Cheung,
J. Scanlon, N. Proisy, G. Addison, J. Ellard, N. Piton, S. Taylor,
M. Cherry, M. Fisher, K. Boxall, S. Burns, M. I. Walton,
I. M. Westwood, A. Hayes, P. Eve, M. Valenti, A. de Haven
Brandon, G. Box, R. L. M. van Montfort, D. H. Williams,
G. W. Aherne, F. I. Raynaud, S. A. Eccles, M. D. Garrett and
I. Collins, J. Med. Chem., 2011, 54, 8328–8342.
33 S. Gowda and D. C. Gowda, Tetrahedron, 2002, 58, 2211–
2213.
17 H. G. Lee, P. J. Milner, M. S. Placzek, S. L. Buchwald and
J. M. Hooker, J. Am. Chem. Soc., 2015, 137, 648–651.
18 A. Kalir and R. Mualem, Synthesis, 1987, 514–515.
19 D. Kim, D. Alexoff, S. W. Kim, J. Hooker and R. A. Ferrieri,
US 20130045151A1, 2013.
20 G. Antoni, H. Omura, M. Bergstrom, Y. Furuya, R. Moulder,
A. Roberto, A. Sundin, Y. Watanabe and B. Langstrom, 34 J. S. Fowler and A. P. Wolf, Acc. Chem. Res., 1997, 30, 181–
Nucl. Med. Biol., 1997, 24, 595–601.
188.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.