Unexpected Behavior of the Heaviest Halogen Astatine in the Nucleophilic Substitution of Aryliodonium Salts

Article abstract of DOI:10.1002/chem.201600922

Aryliodonium salts have become precursors of choice for the synthesis of¹⁸F-labeled tracers for nuclear imaging. However, little is known on the reactivity of these compounds with heavy halides, that is, radioiodide and astatide, at the radiotracer scale. In the first comparative study of radiohalogenation of aryliodonium salts with¹²⁵I^?and²¹¹At^?, initial experiments on a model compound highlight the higher reactivity of astatide compared to iodide, which could not be anticipated from the trends previously observed within the halogen series. Kinetic studies indicate a significant difference in activation energy (E_a=23.5 and 17.1 kcal mol^?1with¹²⁵I^?and²¹¹At^?, respectively). Quantum chemical calculations suggest that astatination occurs via the monomeric form of an iodonium complex whereas iodination occurs via a heterodimeric iodonium intermediate. The good to excellent regioselectivity of halogenation and high yields achieved with diversely substituted aryliodonium salts indicate that this class of compounds is a promising alternative to the stannane chemistry currently used for heavy radiohalogen labeling of tracers in nuclear medicine.

Full text of DOI:10.1002/chem.201600922

DOI: 10.1002/chem.201600922
Full Paper
&
Astatine |Hot Paper|
Unexpected Behavior of the Heaviest Halogen Astatine in the
Nucleophilic Substitution of Aryliodonium Salts
FranÅois Guꢀrard,*[a, b] Yong-Sok Lee,[c] Kwamena Baidoo,[a] Jean-FranÅois Gestin,[b] and
Martin W. Brechbiel[a]
Abstract: Aryliodonium salts have become precursors of
choice for the synthesis of 18F-labeled tracers for nuclear
imaging. However, little is known on the reactivity of these
compounds with heavy halides, that is, radioiodide and asta-
tide, at the radiotracer scale. In the first comparative study
of radiohalogenation of aryliodonium salts with 125Iꢀ and
211Atꢀ, initial experiments on a model compound highlight
the higher reactivity of astatide compared to iodide, which
could not be anticipated from the trends previously ob-
served within the halogen series. Kinetic studies indicate
a significant difference in activation energy (Ea =23.5 and
17.1 kcalmolꢀ1 with 125Iꢀ and 211Atꢀ, respectively). Quantum
chemical calculations suggest that astatination occurs via
the monomeric form of an iodonium complex whereas iodi-
nation occurs via a heterodimeric iodonium intermediate.
The good to excellent regioselectivity of halogenation and
high yields achieved with diversely substituted aryliodonium
salts indicate that this class of compounds is a promising al-
ternative to the stannane chemistry currently used for heavy
radiohalogen labeling of tracers in nuclear medicine.
In contrast to 18F, the other halogens have received limited
attention for use in the reaction with aryliodonium salts at the
radiotracer level. For instance, only one study to date has been
reported on the reactivity of radiochloride,[12] or of radioiodide
(in a patent).[13] Moreover, to our knowledge, there have been
no reports on the reactivity of radioactive bromide or astatide.
However, a number of radioisotopes of the two heaviest halo-
gens have played important roles in nuclear medicine, includ-
ing 123I, 124I, 125I, 131I, and 211At as the most relevant radionu-
clides for use in imaging and/or therapy.[14] Many of the radio-
iodinated and astatinated compounds of interest are aromatic.
They are generally obtained by conventional methods, such as
SNAr by halogen exchange, or electrophilic approaches, such
as halodemetalation or direct substitution.[15,16] However, these
reactions are often associated with issues related to low radio-
chemical yields (RCYs), low specific activities, or concerns
about the toxicity of the precursors and side products when
considering human use. Consequently, the development of
novel methodologies to improve the radiolabeling of arenes
with heavy halogens remains an active area of research.[17,18]
Compared to iodine, the chemistry of astatine remains large-
ly unexplored despite more than 70 years having passed since
the discovery of this element.[19] Its chemical behavior is not
well understood and can differ significantly from the trends for
the other halogens. Good reasons exist for this deficiency and
are primarily due to the fact that no stable isotope exists (for
the most stable, 210At, t1/2 =8.1 h), making difficult the charac-
terization of species by using conventional chemical and spec-
troscopic methods. For instance, the first ionization potential
of At was only very recently accurately determined by using
laser ionization spectroscopy, allowing refinement of previous
estimations.[20] Nonetheless, despite noticeable divergences be-
Introduction
Aryliodonium salts are attractive precursors for arylation of nu-
cleophiles, owing to their low toxicity, the high regioselectivity
they confer, and the mild reaction conditions they require
compared to conventional methods.[1,2] Halogens were
amongst the first nucleophiles to be closely investigated for
the nucleophilic aromatic substitution (SNAr) of aryliodonium
salts more than 50 years ago,[3,4] and they are still considered
for new methodologies in conventional organic synthesis.[5]
However, it is only recently that applications related to halo-
genation with these precursors have been developed, especial-
ly in the field of radiochemistry with the radiolabeling of
arenes with fluorine-18 for positron emission tomography
(PET),[6–9] and improved preparation of important 18F-labeled ra-
diotracers, such as [18F]flumazenil[10] or [18F]mGlu5,[11] both for
brain imaging.
[a] Dr. F. Guꢀrard, Dr. K. Baidoo, Dr. M. W. Brechbiel
Radioimmune & Inorganic Chemistry Section, Radiation Oncology Branch
National Cancer Institute, National Institutes of Health
Bethesda, Maryland 20892 (USA)
[b] Dr. F. Guꢀrard, Dr. J.-F. Gestin
Inserm U892, CNRS UMR6299, Universitꢀ de Nantes
Nantes (France)
E-mail: francois.guerard@univ-nantes.fr
[c] Dr. Y.-S. Lee
Center for Molecular Modeling, Division of Computational Bioscience
Center for Information Technology, National Institutes of Health
Bethesda, Maryland 20892 (USA)
Supporting information and the ORCID identification number for one of the
authors of this article can be found under http://dx.doi.org/10.1002/
chem.201600922.
Chem. Eur. J. 2016, 22, 1 – 9
1
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
tween the chemical behaviors of iodine and astatine, most of
the organic reactions that are applicable to iodine, such as
electrophilic and nucleophilic substitutions, have also been
performed with astatine. This partial understanding of astatine
chemistry has allowed the preparation of various molecules of
biological interest and has led to a number of pre-clinical and
clinical trials, which demonstrated over the last 20 years the
high potential of the 211At isotope (t1/2 =7.2 h) for the treat-
ment of cancers by targeted a-particle therapy.[16,21,22]
In light of the latest advances in astatination methodologies,
electrophilic substitution of arylstannane remains the preferred
method because of the RCYs generally obtained under smooth
conditions.[23,24] However, since astatine can potentially adopt
several oxidation states (-I, +I, +III, +V, and +VII), obtaining
pure At+ for electrophilic reactions is difficult to control, and
partial over-oxidation to unwanted species can be difficult to
avoid. This difficulty is mostly due to the extremely high dilu-
tion of 211At (cyclotron produced and distilled from a bismuth
target) available in solution (generally available at picomolar or
sub-picomolar concentrations), with tiny traces of impurity
being able to perturb its reactivity and lead to side reactions
and inconsistent reaction yields.[25] Additionally, the +I oxida-
tion state is not stable over time, owing to the inherent radia-
tion arising from the decay of 211At to which it is exposed in
solution, leading progressively to changes in its oxidation
state, as reported by Pozzi and Zalutsky.[26]
In contrast, formation of astatine as pure Atꢀ in reducing
media is considerably easier, since only one reduced species of
the halogen is accessible.[27] Therefore, it appears preferable to
have access to alternative methods of astatination that involve
the use of Atꢀ. Complementary to this status, the SNAr of ary-
liodonium salts appears to be an interesting alternative to ex-
plore.
Scheme 1. a) Radioiodination and astatination of bis(4-tert-butylphenyl)iodo-
nium tosylate; b) formation of [211At]4-tert-butylastatobenzene by the con-
ventional astatodestannylation approach.
yields, a parameter that requires additional drying steps of the
radionuclide.
In contrast, the nucleophilicity of the larger halides, iodide
and astatide, was expected to be considerably less affected by
the presence of polar protic solvents such as water due to the
lower strength of hydrate shell of larger anions. Consequently,
available aqueous [125I]NaI and [211At]NaAt were used without
drying to avoid loss of activity during the process. Radiolabel-
ing was performed by heating a 95 vol.% mixture of a 5 mm
iodonium salt in the desired solvent (MeCN, DMF, MeOH, or
a 4:1 H2O/MeCN mixture) with 5 vol.% of the aqueous radioio-
dide solution from 20 to 1408C for 30 min. All reactions were
analyzed chromatographically (HPLC), using the UV trace of 4-
tert-butyliodobenzene as a reference for comparison with the
radioactive trace of the reaction solutions (see the Supporting
Information, Figure S1). The retention time of the astatination
product was nearly identical to 4-tert-butyliodobenzene, sug-
gesting formation of the expected [211At]4-tert-butylastatoben-
zene. To further confirm the identity of the astatinated prod-
uct, [211At]4-tert-butylastatobenzene was also prepared orthog-
onally by electrophilic astatodestannylation of (4-tert-butylphe-
nyl)tributylstannane with 211At oxidized in the +I oxidation
state by N-chlorosuccinimide (Scheme 1b), giving a congruent
HPLC retention time (Figure S1).
In this study, we aimed to probe the reactivity of iodide and
astatide towards aryliodonium salts. 125Iꢀ and 211Atꢀ were inves-
tigated in parallel and their reactivity compared, with the addi-
tional purpose of expanding the understanding of the enig-
matic chemistry of the heaviest of the halogens. We unveiled
an unexpectedly higher reactivity of astatide over iodide that
nucleophilicity difference alone cannot explain, but which in-
volves two possible intermediates, as supported by quantum
chemical calculations.
The results of temperature and solvent impact studies show
a distinct difference in reactivity between iodide and astatide
(Figure 1a,b). In the case of radioiodination, use of MeCN was
by far the optimal choice as this solvent provided RCYs up to
93% at 1208C, whereas the other solvents resulted in RCYs
below 25%, even up to 1408C. In contrast, all solvent condi-
tions gave RCYs greater than 80% for the astatination reaction;
MeOH proved an even better solvent than MeCN, even though
this solvent gave poor RCYs for radioiodination.
Results and Discussion
Reaction parameters
Starting from the model compound bis(4-tert-butylphenyl)io-
donium tosylate (Scheme 1a), the influences of essential pa-
rameters were investigated by adaptation from radiofluorina-
tion of aryliodonium. DMF is generally the preferred solvent
for radiofluorination, although the use of acetonitrile (MeCN) is
also widely reported. This choice is dictated by the necessity of
an aprotic polar solvent to perform the SNAr reaction. The reac-
tivity of the fluoride ion being dramatically reduced in the
presence of water,[28] it is usually necessary to perform the re-
action in anhydrous media to obtain satisfactory radiochemical
An important parameter that can dramatically influence the
reactivity of iodonium salts is the nature of the counterion. In
radiofluorination reactions, sulfonates have generally been pre-
ferred, since they are good leaving groups and are also inert in
comparison to halides, which are able to react with the iodoni-
um by thermal decomposition. We chose to study the influ-
ence of the counterion for radioiodination and astatination by
using the same symmetric iodonium as above associated to
sulfonates (tosylate and triflate) or halides (chloride and bro-
&
&
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
2
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 1. Influence of solvent and temperature on the radiochemical yield of [125I]iodination (a) and [211At]astatination (b) of bis(4-tert-butylphenyl)iodonium
tosylate after 30 min reaction, and influence of counterion in MeCN (c, d). Reactions were performed with a 5 mm iodonium salt concentration and 1.5 MBq
of 125Iꢀ or 5 MBq 211Atꢀ (n=3).
mide). Radioiodination performed at 908C indicated a signifi-
cantly higher rate of reaction with sulfonate salts with RCYs
ꢁ70% after 45 min, compared to the halide salts which gave
RCYsꢁ45% (Figure 1c). In the case of astatination, the same
reactivity sequence was observed, albeit with a smaller differ-
ence between the counterion types and with excellent reactivi-
ty. For instance, the reaction rate was too fast at 908C to dis-
tinguish differences between the anions, with all iodonium
salts giving RCYs of >95% within less than 10 min. When per-
forming the reaction at a lower temperature (658C) the se-
quence OTsꢁOTf>Cl>Br was observed with all compounds
giving RCYs of >50% after 15 min and nearly quantitative
RCYs with all counterions within 45 min (Figure 1d).
The regioselectivity for SNAr towards non-equivalent aryl
groups on diaryliodonium salts can be controlled by different
factors, including the difference in electron density and steric
hindrance between the two rings[29,30] or the presence of a cata-
lyst such as copper salts.[31]
As a model, the electron rich 4-methoxyphenyl was chosen
as directing group, since it was shown to provide high regiose-
lectivity toward the opposite aromatic ring with a wide range
of electron-rich and electron deficient aryl counterparts in pre-
vious studies with 18F.[32,33] A similar preference of radioiodide
and astatide for the most electron-deficient ring was thus also
expected. The 4-methoxyphenyl moiety was also chosen for its
simplicity in this exploratory study although new directing
groups have emerged, such as the electron-rich 2-thienyl[7] or
highly sterically hindered cyclophane derivative.[34] A selection
of variously substituted aryl-4-methoxyphenyl iodonium tosy-
lates were thus prepared (for synthetic details, see the Sup-
porting Information) and investigated in radioiodination and
astatination reactions. As expected, higher RCYs (RPhX and
MeOPhX combined) were obtained when electron-withdrawing
groups were present, with RCYs varying from 46 to 99% in
a 30 min reaction (Table 1).
Overall, this preliminary study guided us to the selection of
MeCN as the preferred solvent to continue the comparative
study between astatide and iodide, with tosylate as counter-
ion.
Influence of the substitution pattern of unsymmetrical iodo-
nium salts
To demonstrate the potential of this reaction, we investigated
the reactivity of unsymmetrically substituted iodonium salts.
Unsymmetrical aryliodonium salts are generally preferred for
radiohalogenation and other synthetic purposes since they are
generally easier and more economical to synthesize, especially
if complex aromatic structures are desired. In the context of ra-
dioiodination for biomedical use, it also appears essential to
avoid the use of symmetrical iodonium salts since each arylio-
donium molecule involved in the SNAr would release an insep-
arable cold equivalent of the radioiodinated compound of in-
terest by reductive elimination and which would contribute to
decreasing the specific activity.
In the case of astatination, the reactivity was so high that
RCYs were quantitative with all substituents under these reac-
tion conditions. The regioselectivity also clearly correlated with
the electronic effects, with high selectivity towards the forma-
tion of the product of interest (RPhX) when R substituents
were the most electron-withdrawing. Additionally, the effect of
steric hindrance was evaluated when R was a methyl group in
the ortho position. Similarly to what is usually observed in fluo-
rination reactions, halogenation was favored on the ortho-sub-
stituted moiety in both radioiodination and astatination, which
can be explained by the transition states formed as demon-
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
3
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
was obtained between the Hammett constants and rate con-
stants with both radiohalides. The reaction constants (1) values
were in line with the significant difference in the influence of
the substituents observed between iodination and astatination
(1=1.82 and 1.34, respectively) considering that the higher
the 1, the more the reaction is sensitive to the electronic effect
of the substituent. A comparison of the regioselectivities for
the SNAr of the lighter radiohalides (18Fꢀ and 38/39Clꢀ) previously
described for identical iodonium salts shows a global decrease
in regioselectivity with increasing halide size (see the Support-
ing Information, Table S2).[12,28] Consequently, it can be fore-
seen from our observations that the design of unsymmetrical
iodonium salts for radiohalogenation will require more optimi-
zation with heavier radiohalogens to obtain optimal regioselec-
tivity and consequently high RCYs and specific activities, partic-
ularly when electron rich products are desired.
Table 1. Influence of the substituent R on the RCY and the selectivity for
the target product of radioiodination and astatination (n=3).
125I
211At
[a]
[a]
R
H
RCY(I+II) [%]
57ꢂ2
(I)/(II) ratio
4.8:1
1.5:1
4.4:1
24:1
10:1
38:1
>50:1
28:1
>50:1
RCY(I+II) [%]
97ꢂ1
(I)/(II) ratio
4.2:1
2:1
3.7:1
8.1:1
5.3:1
8.2:1
16:1
4-Me
3-Me
2-Me
4-Cl
4-CO2Et
4-CN
3-NO2
4-NO2
46ꢂ6
97ꢂ1
61ꢂ1
99ꢂ1
98ꢂ1
98ꢂ1
68ꢂ2
98ꢂ1
92ꢂ1
98ꢂ1
97ꢂ1
99ꢂ1
67ꢂ4[b]
90ꢂ2
99ꢂ1
24:1
29:1
99ꢂ1
Thermochemical and quantum chemical studies
[a] Decay corrected; [b] detected decomposition products.
To better understand the gap in reactivity between the two
halogens, we investigated the reaction further in terms of ki-
netics and thermodynamics. Using the experimental conditions
determined above on the symmetric iodonium model, the rate
constants of the radiohalogenation reactions were determined
at various temperatures (see the Supporting Information, Fig-
ure S2). Rate constants ki’ differed significantly between the
two halogens. For instance, astatination occurred nearly 50
times faster than iodination at 808C (Table 2), reflecting the
higher reactivity of astatide. These rate constants were then
used in the Arrhenius plot for determination of Ea (Figure 3).
An Ea of 23.5 kcalmolꢀ1 was obtained for the radioiodination
whereas astatination occurred with a significantly lower Ea of
17.1 kcalmolꢀ1. The higher nucleophilicity of the astatide anion
is not sufficient to explain the sharp increase in reactivity ob-
served as compared to the iodide. Whereas nucleophilicity in-
creases in SNAr reactions with increasing size of halide, this
was previously shown to have little influence on the Ea of halo-
strated in previous studies.[8] Notably, regioselectivity was
higher for iodination than for astatination, especially with
stronger electron-withdrawing substituents. To rationalize the
observed impact of the substituents, a kinetic study was per-
formed. The reaction was pseudo-first order, considering the
concentration variation of the iodonium salt in large excess
was negligible during the reaction (see the Supporting Infor-
mation, Table S1). The results were then plotted in a Hammett
diagram, using the s value for each substituent and their reac-
tion rate constants ki’ (Figure 2). An excellent linear correlation
genation
of
bis(4-tert-butylphenyl)iodonium
(Clꢀ: 32.1;
Brꢀ: 32.2; Iꢀ: 32.2 kcalmolꢀ1 in DMF).[4]
Table 2. Pseudo-first order reaction rates of the 125I and 211At substitution
on bis(4-tert-butylphenyl)iodonium tosylate in MeCN (n=3).
T [8C]
ki’ [minꢀ1
]
125I
211At
45
50
55
60
65
70
75
80
82.5
85
90
95
100
–
–
–
–
–
–
–
0.0231ꢂ0.0030
0.0368ꢂ0.0015
0.0494ꢂ0.0038
0.0742ꢂ0.0072
0.0992ꢂ0.0097
0.1498ꢂ0.0109
0.2226ꢂ0.0349
0.0082ꢂ0.0011
0.0122ꢂ0.0009
0.0160ꢂ0.0047
0.0225ꢂ0.0015
0.0355ꢂ0.0021
0.0535ꢂ0.0040
0.3770ꢂ0.0281
–
–
–
–
–
Figure 2. Hammett plot for the [125I]iodination (at 908C) and
[
211At]astatination (at 658C) of unsymmetrical iodonium tosylates in MeCN
(n=3).
&
&
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
4
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Iodonium salts are known to exist as dimers in the solid and
gaseous phase, as well as in solution.[39,40] Similarly, we found
that bis(4-tert-butylphenyl)iodonium iodide [(Ar2I)I] adopts a di-
meric structure in the solid state (see the Supporting Informa-
tion, Figure S3).[41] Thus, we aimed to probe the influence of
the monomeric and dimeric intermediates on the reaction by
quantum chemical calculations. The X-Ray structure of iodoni-
um iodide was used to construct monomeric (Ar2I)I, (Ar2I)At,
and (Ar2I)OTs, as well as homodimeric [(Ar2I)OTs]2. Given the ex-
tremely low concentrations of radioiodide (nanomolar to pico-
molar) or astatide (picomolar to femtomolar) involved in the
reaction at radiotracer level, the probability for the iodonium
radiohalide (125I or 211At) to form a homodimer is extremely
low. Rather, given the much higher concentration of (Ar2I)OTs
in solution (5 mm), it would be as the heterodimer [(Ar2I)X]
[(Ar2I)OTs] [Equation (1)].
Figure 3. Arrhenius plot for the determination of Ea of the 125I (~) and 211At-
substitution (*) on bis(4-tert-butylphenyl)iodonium tosylate in MeCN (n=3)
Despite the lack of spectroscopic methods to probe possible
mechanisms with astatine species at the radiotracer level, we
tested several hypotheses to provide a rationale for the differ-
ence in reactivity. Several previously reported potential reac-
tion courses were considered and tested. These included the
presence of metallic impurities in the astatine solution, which
could catalyze the reaction, as seen in the enhanced reactivity
of fluoride on aryliodonium salts in the presence of copper
salts.[9] However this possibility was ruled out based on our ob-
servation that the outcome of the radioiodination reaction (ki-
netics and RCY) was unaffected when performed in the pres-
ence of astatine stock solution after the decay of the 211At ac-
tivity. This test also rules out the sulfite ion as a potential
cause of the reactivity change, Na2SO3 being otherwise absent
in radioiodination solution, whereas astatine stock solution
contains this reducing agent to produce the reactive 211Atꢀ
species. The formation of a benzyne intermediate is a well-
known pathway of aryliodonium salt reactivity in the presence
of a strong base. Whereas the formation of benzyne in the
presence of the basic fluoride anion has been reported,[35]
iodide and astatide are such weak bases that they cannot be
invoked as able to form such benzyne intermediates.
ðAr2IÞOTs þ ðAr2IÞOTs Ð ½ðAr2IÞOTsꢃ2 Ð ½ðAr2IÞOTsꢃ½ðAr2IÞXꢃ
ð1Þ
Energy calculations (see the Supporting Information) indeed
indicate that (Ar2I)OTs is more stable in its dimeric form by
DG=ꢀ0.7 kcalmolꢀ1. Our calculations also indicate that the ex-
change of a tosylate in [(Ar2I)OTs]2 with iodide from NaI or with
astatide from NaAt to form the heterodimeric structures is
thermochemically favorable by DG=ꢀ14.0 kcalmolꢀ1 and
ꢀ15.1 kcalmolꢀ1, respectively (see the Supporting Information,
Figure S4).
Ground states (GS) and transition states (TS) were then cal-
culated, taking into account two possible pathways: either
from the heterodimer or from the monomer in equilibrium (ex-
emplified for iodination in Figure 4).
The calculated Ea, corresponding to the difference of the
sum of the electronic energy and zero-point vibrational energy
between the GS and the TS, is 17.9 kcalmolꢀ1 for iodination
and 17.2 kcalmolꢀ1 for astatination from the monomeric com-
plex, whereas it is 23.2 kcalmolꢀ1 and 22.6 kcalmolꢀ1 from the
dimeric complex for iodination and astatination, respectively.
The extra charge interaction provided by (Ar2I)OTs is likely re-
sponsible for the increase of Ea in the dimer pathway. For ex-
ample, the nucleophilic iodide in the GS1 has to overcome the
charge attraction provided by the central iodine atom of
(Ar2I)OTs to form a bond with the Cipso. The good agreement
between the calculated Ea for iodination via the heterodimer
(23.2 kcalmolꢀ1) and the experimental Ea (23.5 +/ꢀ1.1 kcal
molꢀ1) suggests that the radioiodination occurs in a dimeric
environment. In contrast, the calculated Ea for astatination via
the monomer (17.2 kcalmolꢀ1) agrees with the experimental Ea
(17.1 +/ꢀ0.6 kcalmolꢀ1) which indicates that the astatination
likely occurs in a monomeric environment. It is noted that the
calculated difference of Ea for iodination and astatination is
only 0.6–0.7 kcalmolꢀ1 in either the monomeric or heterodi-
meric form, which suggests that the difference in nucleophilici-
ty between astatide and iodide plays a minor role in the reac-
tion. The proposed mechanism (Figure 4), another representa-
tion of the Curtin–Hammett principle,[42] further indicates that
the relative stability of monomers over dimers modulates the
A more plausible explanation for the reactivity difference is
the possibility of a radical mechanism, as previously proposed
for a similar reaction with the astatination of diazonium
salts.[36] Moreover, radical pathways have also been implicated
in reactions involving iodonium with nucleophiles such as 2-ni-
tropropanate anion via a single electron transfer mechanism.[37]
It has also been suggested that, under specific conditions, ary-
liodonium iodide decomposed in part via a radical path-
way.[13,38] However, the hypothesis of a free radical reaction
was ruled out by comparing the reaction rate in the presence
of atmospheric oxygen (standard conditions of this study), in
the absence of oxygen (solvent degassed under argon), and in
the presence of a radical scavenger (TEMPO). In all three cases,
the reaction rates and products formed were essentially identi-
cal (see the Supporting Information, Table S3), which indicates
that, if it ever occurs, such a radical mechanism is only a minor
pathway in our reaction conditions to influence significantly
the reaction at the radiotracer scale. Consequently, given the
Hammett correlation obtained above, the reaction for both
halides essentially follows a SNAr-type mechanism.
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
5
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 4. Reaction diagram of radioiodination depicting both possible transition states (TS), either via a monomeric iodonium iodide or by a heterodimeric in-
termediate. The heterodimer exists in equilibrium with its monomers, and interconversion between the two is likely much faster than the iodination. Accord-
ingly, depending on the energy of both TS1 and TS2, the reaction follows either a dimeric or a monomeric pathway. The TS of the monomeric (Ar2I)I is charac-
terized by the sp3 hybridization of the Cipso due to the incoming iodide. The distance between the two atoms at the TS is considerably shorter (2.92 ꢂ) than
that in the ground state (GS; 3.90 ꢂ). The IꢀCipso distances of the heterodimer are 2.90 ꢂ in the TS and 4.01 ꢂ in the GS. Note that the TS2 energy is the sum
of the Ea of the monomer and the dimer-dissociation energy (D), which is the difference between the energy of both monomers (GS2) and that of the dimer
(GS1). Product does not show Ar2I-OTs. Atoms are colored as follows: green, carbon; red, oxygen; yellow, sulfur; dark green, iodine. Hydrogen atoms are not
shown. All distances are in ꢂ. A similar diagram can be constructed for astatination. Ea-At values are provided for comparison and optimized structures with
bond lengths are given in Figure S5 (see the Supporting Information).
energy barrier of halogenation in a monomer, which in turn
dictates whether iodination or astatination occurs within
a monomeric or a dimeric complex.
structure. Further investigation will be necessary to corrobo-
rate this point.
The DG value for dimerization (ꢀ2.0 kcalmolꢀ1 at 298.15 K)
in both cases, calculated by subtracting the sum of the free
energy of each monomer from that of the dimer, favors the
formation of the heterodimers. However, the contribution of
translational and rotational entropy of each monomer to dime-
rization in the liquid phase is known to be much lower than in
the gas phase, due to solvent-restricted movement of mono-
mers.[43] Accordingly, the above DG value, including both trans-
lational and rotational entropy based on the ideal gas approxi-
mation, can be more negative than ꢀ2.0 kcalmolꢀ1. Therefore,
this DG alone may not be a reliable marker for determining
whether the reaction follows either a dimeric or a monomeric
pathway.
Conclusion
In summary we have explored the reactivity of heavy halides
towards aryliodonium salts at the radiotracer scale. We un-
veiled an unexpected and significantly higher reactivity of asta-
tide as compared to the trend observed for other halides.
Quantum chemical calculations led us to propose that astatina-
tion occurs via the monomeric form of an iodonium complex
whereas iodination occurs via a more stable heterodimeric io-
donium intermediate. Overall, it appears that iodonium salts
are a promising alternative class of precursor to organotin-
based chemistry for radioiodination and astatination. Iodonium
salts achieved more consistent and nearly quantitative astati-
nation RCYs under mild conditions by using the ꢀI oxidation
state, which is simpler to control in organic media at the radio-
tracer level than the +I oxidation state needed for electrophil-
ic substitutions. As expected, the regioselectivity of SNAr of
non-symmetrical diaryliodonium salts was governed by elec-
tronic and steric effects. However, the lower regioselectivity
that is observed in iodination and, to a greater extent, in astati-
nation, as compared to previously reported radiofluorinations,
suggests that carefully designed diaryliodonium salts will be
required for efficient synthesis and optimal specific activity of
aryl radioiodide and aryl astatide compounds, especially in the
In the case of astatination, it is likely that, due to its higher
polarizability, astatide forms a stronger complex than iodide
with the iodonium, similarly to the reported case of diazonium,
which also exhibits greatly enhanced reactivity with astatide as
compared to iodide.[36] The higher polarizability of astatide in-
duces a higher delocalization of the charge from the halide to
the iodonium compared to lighter halogens, leading to a tight-
er ion pair by virtue of electrostatic interactions. In this con-
text, it can be expected that the extra charge interaction of
(Ar2I)OTs with astatide is considerably lower than with iodide,
resulting in a thermodynamically less favorable heterodimeric
&
&
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
6
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Coordinates of the calculated structures are given in Table S4 (see
the Supporting Information).
case of aryl substituted by electron donating groups. Ultimate-
ly, we anticipate that this radiolabeling approach will become
a highly useful tool for the radiolabeling of biomolecules with
heavy radiohalogens with applications in nuclear imaging and
targeted radiotherapy. Through the development of specifically
designed iodonium salts, such efforts are underway in our lab-
oratory.
Acknowledgements
This research was supported by the Intramural Research Pro-
gram of the NIH, National Cancer Institute, Center for Cancer
Research and Center for Information Technology and in part by
grants from the French National Agency for Research, called
“Investissements d’Avenir” IRON Labex (no. ANR-11-LABX-0018-
01) and ArronaxPlus Equipex (no. ANR-11-EQPX-0004). The
quantum chemical study utilized PC/linux clusters at the
center for Molecular Modeling of the National Institutes of
Health (http://cit.nih.gov). Lawrence Szajek, NIH, CC is thanked
for his cyclotron operations and irradiation of bismuth targets
for the production of 211At. Professor Vanelle is thanked for the
input on radical mechanism investigations.
Experimental Section
Radiochemistry
[
125I]NaI was obtained commercially in 10ꢀ5 m NaOH solution with
a volumic acitivity of 50 mCimLꢀ1 (1.85 MBq/mL) and was diluted
12 times in deionized water before use. 211At was produced by
using the 209Bi(a,2n)211At reaction by bombarding a disposable in-
ternal bismuth target with a-particles from the Cyclotron Corpora-
tion CS-30 cyclotron in the National Institutes of Health Positron
Emission Tomography Department. 211At was recovered from the ir-
radiated target in acetonitrile by using a previously described dry-
distillation procedure.[44] Before use, the 211At solution was diluted
twice in a 10 mgmLꢀ1 aqueous solution of Na2SO3, resulting in
a 1:1 MeCN/water solution of sodium astatide. Iodonium salts were
obtained commercially or synthesized as described in the Support-
ing Information.
Keywords: astatine
·
halogenation
·
iodonium salts
·
nucleophilic substitution · radiopharmaceuticals
[1] E. A. Merritt, B. Olofsson, Angew. Chem. Int. Ed. 2009, 48, 9052–9070;
Angew. Chem. 2009, 121, 9214–9234.
[2] V. V. Zhdankin, P. J. Stang, Chem. Rev. 2008, 108, 5299–5358.
[3] F. M. Beringer, E. J. Geering, I. Kuntz, M. Mausner, J. Phys. Chem. 1956,
60, 141–150.
Reaction of iodonium salts with 125Iꢀ and 211Atꢀ
[4] F. M. Beringer, M. Mausner, J. Am. Chem. Soc. 1958, 80, 4535–4536.
[5] B. Hu, W. H. Miller, K. D. Neumann, E. J. Linstad, S. G. DiMagno, Chem.
Eur. J. 2015, 21, 6394–6398.
[6] M. Tredwell, V. Gouverneur, Angew. Chem. Int. Ed. 2012, 51, 11426–
11437; Angew. Chem. 2012, 124, 11590–11602.
[7] T. L. Ross, J. Ermert, C. Hocke, H. H. Coenen, J. Am. Chem. Soc. 2007,
129, 8018–8025.
[8] J.-H. Chun, S. Lu, Y.-S. Lee, V. W. Pike, J. Org. Chem. 2010, 75, 3332–
3338.
[9] N. Ichiishi, A. F. Brooks, J. J. Topczewski, M. E. Rodnick, M. S. Sanford,
P. J. H. Scott, Org. Lett. 2014, 16, 3224–3227.
[10] B. S. Moon, H. S. Kil, J. H. Park, J. S. Kim, J. Park, D. Y. Chi, B. C. Lee, S. E.
Kim, Org. Biomol. Chem. 2011, 9, 8346–8355.
To iodonium salt (950 nmol) in the selected solvent (190 mL) equili-
brated at the appropriate temperature of reaction was added
[
125I]NaI or [211At]NaAt solution (10 mL, typically 1.5 MBq), prepared
as described above. At desired times, aliquots were withdrawn and
deposited on a silica gel TLC plate and eluted with the appropriate
solvent, and/or diluted in a 1:1 water/MeCN mixture and analyzed
by reverse-phase HPLC using the appropriate elution system. Re-
tention indices and elution systems used for all compounds of this
study are given in the Supporting Information. Aromatic 125I and
211At species were identified by comparison of the retention index
of authentic samples of the non-radioactive iodinated compound.
[11] S. Telu, J.-H. Chun, F. G. Simꢀon, S. Lu, V. W. Pike, Org. Biomol. Chem.
2011, 9, 6629–6638.
[12] M.-R. Zhang, K. Kumata, M. Takei, T. Fukumura, K. Suzuki, Appl. Radiat.
Isot. 2008, 66, 1341–1345.
[13] S. Dimagno, B. Hu, Radioiodinated Compounds, 2015, WO2015147950
(A2).
[14] M. J. Adam, D. S. Wilbur, Chem. Soc. Rev. 2005, 34, 153–163.
[15] H. H. Coenen, S. M. Moerlein, G. Stçcklin, Radiochim. Acta 1983, 34, 47–
68.
Astatination via electrophilic destannylation
To a 2.5 mgmLꢀ1 solution of tributyl(4-tertbutylphenyl)-stannane in
99:1 MeCN/AcOH (20 mL) was added 2 mgmLꢀ1 N-chlorosuccini-
mide in MeCN (5 mL), followed by the 211At (5 MBq activity) in
MeCN (75 mL). The solution was heated for 30 min at 608C and
after cooling down to room temperature, the solution was diluted
with water (100 mL) and an aliquot was analyzed by HPLC, using
the same chromatographic system as described above.
[16] F. Guꢀrard, J.-F. Gestin, M. W. Brechbiel, Cancer Biother. Radiopharm.
2013, 28, 1–20.
[17] A. A. Cant, S. Champion, R. Bhalla, S. L. Pimlott, A. Sutherland, Angew.
Chem. Int. Ed. 2013, 52, 7829–7832; Angew. Chem. 2013, 125, 7983–
7986.
Quantum chemistry
The X-ray structure of bis(4-tert-butylphenyl)iodonium iodide was
used as a template to construct both the monomeric and heterodi-
meric structures of Ar2I+ complexed with iodide or astatide. Quan-
tum chemical calculations were carried out with the density func-
tional theory at the B3LYP level with small-core pseudopotentials[45]
for the I and At atoms with the cc-pVDZ basis set for their 25 va-
lence electrons, together with the all-electron cc-PVDZ basis set for
the rest of the atoms. All calculations were carried out in the reac-
tion field of MeCN with the polarizable continuum model, together
with the UAKS parameters set as implemented in Gaussian 09 soft-
ware.[46] For each TS, a single imaginary frequency was obtained.
[18] R. Yan, K. Sander, E. Galante, V. Rajkumar, A. Badar, M. Robson, E. El-
Emir, M. F. Lythgoe, R. B. Pedley, E. ꢂrstad, J. Am. Chem. Soc. 2013, 135,
703–709.
[19] D. S. Wilbur, Nat. Chem. 2013, 5, 246–246.
[20] S. Rothe, A. N. Andreyev, S. Antalic, A. Borschevsky, L. Capponi, T. E. Co-
colios, H. De Witte, E. Eliav, D. V. Fedorov, V. N. Fedosseev, D. A. Fink, S.
Fritzsche, L. Ghys, M. Huyse, N. Imai, U. Kaldor, Yuri Kudryavtsev, U.
Kçster, J. F. W. Lane, J. Lassen, Nat. Commun. 2013, 4, 1835.
[21] Y.-S. Kim, M. W. Brechbiel, Tumor Biol. 2012, 33, 573–590.
[22] G. Vaidyanathan, M. R. Zalutsky, Curr. Radiopharm. 2008, 1, 177–196.
[23] H. Rajerison, D. Faye, A. Roumesy, N. Louaisil, F. Boeda, A. Faivre-Chau-
vet, J.-F. Gestin, S. Legoupy, Org. Biomol. Chem. 2016, 14, 2121–2126.
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
7
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
ˇˇ
[24] E. Aneheim, A. Gustafsson, P. Albertsson, T. Bꢃck, H. Jensen, S. Palm, S.
Svedhem, S. Lindegren, Bioconjugate Chem. 2016, 27, 688–697.
[25] G. W. Visser, Radiochim. Acta 1989, 47, 97–103.
[26] O. R. Pozzi, M. R. Zalutsky, J. Nucl. Med. 2007, 48, 1190–1196.
[27] D.-C. Sergentu, D. Teze, A. Sabatiꢀ-Gogova, C. Alliot, N. Guo, F. Bassal,
I. D. Silva, D. Deniaud, R. Maurice, J. Champion, N. Galland, G. Monta-
von, Chem. Eur. J. 2016, 22, 2964–2971.
[40] Y.-S. Lee, M. Hodoscek, J.-H. Chun, V. W. Pike, Chem. Eur. J. 2010, 16,
10418–10423.
[41] CCDC 1452802 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre.
[42] J. I. Seeman, Chem. Rev. 1983, 83, 83–134.
[43] Y. B. Yu, P. L. Privalov, R. S. Hodges, Biophys. J. 2001, 81, 1632–1642.
[44] S. Lindegren, T. Bꢃck, H. J. Jensen, Appl Rad Isot 2001, 55, 157–160.
[45] K. A. Peterson, D. Figgen, E. Goll, H. Stoll, M. Dolg, J. Chem. Phys. 2003,
119, 11113–11123.
[28] J.-H. Chun, S. Telu, S. Lu, V. W. Pike, Org. Biomol. Chem. 2013, 11, 5094–
5099.
[29] D. E. Hill, J. P. Holland, Comp. Theor. Chem. 2015, 1066, 34–46.
[30] J. Malmgren, S. Santoro, N. Jalalian, F. Himo, B. Olofsson, Chem. Eur. J.
2013, 19, 10334–10342.
[31] N. Ichiishi, A. J. Canty, B. F. Yates, M. S. Sanford, Org. Lett. 2013, 15,
5134–5137.
[46] Gaussian 09, revision A.02, J. Frisch G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢄ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, and D. J. Fox, Gaussian Inc., Wallingford CT, 2009.
[32] J.-H. Chun, V. W. Pike, Chem. Commun. 2012, 48, 9921–9923.
[33] J.-H. Chun, V. W. Pike, Org. Biomol. Chem. 2013, 11, 6300–6306.
[34] B. Wang, J. W. Graskemper, L. Qin, S. G. DiMagno, Angew. Chem. Int. Ed.
2010, 49, 4079–4083; Angew. Chem. 2010, 122, 4173–4177.
[35] T. Kitamura, M. Yamane, K. Inoue, M. Todaka, N. Fukatsu, Z. Meng, Y. Fu-
jiwara, J. Am. Chem. Soc. 1999, 121, 11674–11679.
[36] G. J. Meyer, K. Roessler, G. Stoecklin, J. Am. Chem. Soc. 1979, 101, 3121–
3123.
[37] P. R. Singh, R. K. Khanna, Tetrahedron Lett. 1982, 23, 5355–5358.
[38] M. S. Ermolenko, V. A. Budylin, A. N. Kost, Chem. Heterocycl. Compd.
1978, 14, 752–754.
[39] N. W. Alcock, R. M. Countryman, J. Chem. Soc. Dalton Trans. 1977, 217–
219.
Received: February 26, 2016
Published online on && &&, 0000
&
&
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
8
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
Wherever I lay my At: Radiohalogena-
tion of diaryiodonium salts has been in-
vestigated and a significantly higher re-
activity of astatide has been observed.
Regioselectivity and reaction kinetics
can be controlled by varying the nature
of substituents. The high reactivity of
these precursors makes them an attrac-
tive alternative to the conventional aryl-
stannane chemistry for radiolabeling
with iodine and astatine.
&
Astatine
F. Guꢀrard,* Y.-S. Lee, K. Baidoo,
J.-F. Gestin, M. W. Brechbiel
&& – &&
Unexpected Behavior of the Heaviest
Halogen Astatine in the Nucleophilic
Substitution of Aryliodonium Salts
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
9
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ