Electrochemical Synthesis of Aryl Iodides by Anodic Iododesilylation

Article abstract of DOI:10.1002/anie.201711293

An electrochemical access to iodinated aromatic compounds starting from trimethylsilyl-substituted arenes is presented. By design of experiments, highly efficient and mild conditions were identified for a wide range of substrates. A functional group stability test and the synthesis of an important 3-iodobenzylguanidine radiotracer illustrate the scope of this process.

Full text of DOI:10.1002/anie.201711293

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201711293
German Edition: DOI: 10.1002/ange.201711293
Electrosynthesis
Electrochemical Synthesis of Aryl Iodides by Anodic Iododesilylation
Robert Mçckel, Jessica Hille, Erik Winterling, Stephan Weidemꢀller, Tabea Melanie Faber, and
Gerhard Hilt*
Dedicated to Professor Reinhart Mçckel
Abstract: An electrochemical access to iodinated aromatic
compounds starting from trimethylsilyl-substituted arenes is
presented. By design of experiments, highly efficient and mild
conditions were identified for a wide range of substrates. A
functional group stability test and the synthesis of an important
3-iodobenzylguanidine radiotracer illustrate the scope of this
process.
A
ryl iodides are used as precursors in contemporary carbon–
carbon bond-forming reactions,[1] for the synthesis of hyper-
valent iodine compounds,[2] or as radiopharmaceuticals,[3] and
a variety of approaches have been developed for their
synthesis. Most often, their synthesis is accomplished by
electrophilic aromatic substitution, either by using pregen-
erated electrophilic iodinating reagents such as N-iodosucci-
nimide (NIS) or by in situ oxidation of iodide or iodine to
reactive iodonium ions with various oxidants.[4] The central
problem to all possibilities is the lack of atom efficiency as
pregenerated reagents or oxidants have to be used in
stoichiometric amounts. The electrochemical generation of
electrophilic iodinating species offers a unique solution to
reduce the amount of waste chemicals. A seminal example of
this approach was reported by Yoshida and co-workers who
directly iodinated aromatic substrates by electrochemical
generation of iodonium ions (Scheme 1).[5] The only draw-
back of this and similar methods is the decreased control over
substitution site selectivity when performing the reaction in
a one-pot fashion, and the restriction to electron-rich
substrates.
Scheme 1. Iodination reactions.[5,6]
as mainly electron-rich substrates and preferably ortho-
substituted trimethylsilyl arenes could be employed with
satisfying yields. Nevertheless, the method showed excellent
selectivity, and the substrates were easily accessible from
regioselective cobalt-catalysed Diels–Alder reactions of tri-
methylsilyl-substituted alkynes with 1,3-dienes.[7]
Herein, we describe the combination of our strategy with
sustainable electrochemical methods into a novel protocol.
We sought to introduce, in analogy to our initial iodination
procedure, a trimethylsilyl group as the directing group for an
iododesilylation reaction. The use of a directing group to
enhance selectivity combined with high reactivity and short
reaction times would outweigh the negative impact on atom
economy.
As part of the optimisation, a number of preconditions
were drawn to guarantee the design of a simple, mild, green,
and inexpensive process. The reaction should be conducted at
room temperature, no inert conditions should be necessary,
and the iodine source should be potassium iodide as we were
hoping to thereby obviate the need for an additional
supporting electrolyte. Furthermore, a constant current
density of 6.7 mAcmÀ1 was applied to allow for a chemo-
selective electrochemical reaction and to keep the reaction
times relatively short.
An iodination method reported recently by our group
relied on the generation of reactive tert-butyl hypoiodite by
the reaction of zinc iodide with tert-butyl hydrogen peroxide
(TBHP), which was reacted in situ with trimethylsilyl arenes
(Scheme 1).[6] The introduction of a trimethylsilyl group
allowed for selective iodination. The major limitation was,
apart from low atom economy, a diminished substrate range,
[*] M. Sc. R. Mçckel, J. Hille, E. Winterling, S. Weidemꢀller, T. M. Faber,
Prof. Dr. G. Hilt
Fachbereich Chemie, Philipps-Universitꢁt Marburg
Hans-Meerwein-Straße 4, 35043 Marburg (Germany)
M. Sc. R. Mçckel, Prof. Dr. G. Hilt
Institut fꢀr Chemie, Universitꢁt Oldenburg
Carl-von-Ossietzky-Straße 9–11, 26111 Oldenburg (Germany)
E-mail: gerhard.hilt@uni-oldenburg.de
In a first test reaction with trimethyl(phenyl)silane (1) as
the substrate, no product was detected upon using an
undivided cell (Table 1, entry 1). Therefore, the screening of
commonly used solvents (entry 2–5) and all following reac-
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under:
https://doi.org/10.1002/anie.201711293.
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
Table 1: Electrochemical iododesilylation of trimethyl(phenyl)silane 1.
experiments without an iodonium source in the reaction or
with a substrate without a TMS group (toluene) suggested
that the methoxylation occurs mainly under participation of
iodonium ions and only when TMS-substituted substrates are
present.
Entry
Solvent
KI
[equiv]
Applied charge
Yield
[%]
To improve the yields and decrease the extent of
methoxylation, we investigated the influence of the iodide
loading and the solvent in more detail. To cover the
methoxylation side reaction as well, (para-tolyl)trimethylsi-
lane was used as a test substrate for this series of experiments.
To be able to address all important interactions between the
parameters, we used a design of experiment approach
applying a central composite plan.[9,10] Parameters being
optimized were a) the methanol concentration (1:1–9:1
acetonitrile/methanol (lower methanol concentrations were
excluded because they lead to poor conductivity and there-
fore need an additional electrolyte)), b) the KI stoichiometry
(1.0–1.2 equiv), and c) the applied charge (2.2–2.5 FmolÀ1) as
those three parameters seemed to have the biggest impact on
yields according to our previous optimization. The resulting
plan contained 18 reactions covering one- and two-factor and
quadratic interactions (Table 2).
[FmolÀ1
]
1[a]
2
3
4
5
6
7
8
9
10
11
12[b]
13[c]
14[d]
15[e]
MeCN
MeCN
(MeOCH2)2
CH2Cl2
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.1
1.1
1.2
1.2
1.0
1.1
1.1
1.1
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.2
2.4
2.4
2.6
2.0
2.4
2.4
2.4
0
40
10
37
10
64
80
84
97
86
94
48
MeOH
MeCN/MeOH (1:1)
MeCN/MeOH (7:3)
MeCN/MeOH (7:3)
MeCN/MeOH (7:3)
MeCN/MeOH (7:3)
MeCN/MeOH (7:3)
MeCN/MeOH (7:3)
MeCN/MeOH (7:3)
MeCN/MeOH (7:3)
MeCN/MeOH (7:3)
[f]
–
–
87
[f]
Unless otherwise stated, a divided cell with platinum electrodes with
0.5 mL conc. H2SO4 in the cathode compartment was used. The yields
were determined by GC analysis using mesitylene as the internal
standard (for experimental details, see the Supporting Information).
[a] An undivided cell was used. [b] Graphite electrodes were used.
[c] Instead of H2SO4, acetic acid (1 mL) was used. [d] Instead of H2SO4,
methanesulfonic acid (0.5 mL) was used. [e] Instead of H2SO4, LiClO4
(3 mmol, 0.3m) was used. [f] No conductivity observed.
Table 2: Optimisation by design of experiments.[a]
tions were performed in H-type divided cells. Disappoint-
ingly, the best yield of 2 was quite low (40%) even when
acetonitrile was used as the solvent. When CH2Cl2, methanol,
or dimethoxyethane were used, the yields amounted only to
10–37% (entry 1–5). The low yields were attributed to the
formation of large quantities of the protodesilylation side
product benzene. Only a solvent mixture of acetonitrile/
methanol (1:1) led to an improved yield of 64%, mainly
owing to a significant decrease in the protodesilylation side
reaction.[8] A slight decrease in the methanol content to a 7:3
mixture further increased the yield to 80%. A final improve-
ment was achieved by slightly increasing the KI loading to
1.1 equiv and adjusting the applied charge to 2.4 FmolÀ1,
which resulted in an almost quantitative yield of 97%.
However, a further increase in the amount of the iodine
source or the applied charge (entries 10 and 11) led to
decreases in yield. Unfortunately, changing the electrodes
from platinum plates to graphite decreased the yield to 48%
(entry 12). Finally, the acid or supporting electrolyte that was
used in the cathode compartment to lower the cell potential
was changed from sulfuric acid to acetic acid, methanesul-
fonic acid, or lithium perchlorate, but all of them led to
reduced yields (entry 13–15).
Factor
p Value
MeCN/MeOH ratio
0.0000
0.0001
0.0003
0.0031
(applied electricity)2
amount of KI
applied electricityꢂMeCN/MeOH ratio
[a] All reactions were carried out on 0.5 mmol scale using a divided cell
and platinum electrodes. The yields were determined by GC-FID analysis
using mesitylene as the internal standard (see the Supporting Informa-
tion for details). Numbers in grey indicate the optimal values for the
three parameters and the corresponding yield (with its confidence
interval in brackets). The p values indicate the probability of whether the
respective effect can be explained by the null hypothesis (effects below
0.01 are regarded as significant).[9e]
Four important interactions returning a model with an
R2 value of 0.95 (which confirms that there are only slight
differences between the observed and predicted values) were
found: a) the methanol concentration, b) the quadratic term
of the applied electricity, c) a two-factor interaction between
methanol concentration and charge consumption, and d) the
KI equivalents. According to the obtained model, the optimal
values for the three parameters are a methanol concentration
of 10 vol% in acetonitrile, 1.0 equiv of KI, and an applied
charge of 2.3 FmolÀ1.
To our disappointment, when we applied our optimized
reaction conditions to substrates other than trimethylsilyl
benzene, the yields were only moderate (68–71%; see the
Supporting Information). The main reason was determined
by profound GC/MS analysis of the reaction products:
Significant quantities of the starting materials were found to
be methoxylated at the benzylic position, if present. Control
These observations are in line with the previously
obtained results as a minimal iodide loading and a minimal
2
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
methanol concentration should supress methoxylation. The
the yield of the desired product and the amount of additive
greater applied charge than the theoretically needed
2.0 FmolÀ1 can be ascribed to partial oxidation of the solvent.
This can be seen in the two-factor interaction, which shows
that the necessary applied electricity is higher (and the yields
are lower) for higher methanol concentrations as oxidation of
methanol becomes more decisive.
recovered after the reaction are determined, which gives an
indication for the stability of the respective additional func-
tional group under the applied reaction conditions (Table 3).
Table 3: Functional group tolerance test using (para-tolyl)trimethylsilane
as the substrate.[a]
The optimised reaction conditions were tested on a variety
of ortho-, meta-, and para-functionalized TMS-substituted
arenes (Scheme 2) with electron-donating and electron-with-
Additive
Yield
(starting
material)
Yield
(product)
Yield
(additive)
1-octene
1-dodecyne
benzonitrile
butyl phenyl ketone
methyl benzoate
aniline
71%
62%
5%
0%
4%
13%
4%
89%
86%
82%
0%
0%
13%
96%
78%
81%
0%
5%
nitrobenzene
carbazole
indole
0%
0%
0%
0%
12%
37%
86%
84%
83%
65%
76%
56%
38%
8%
100%
0%
0%
100%
86%[b]
0%
sulfolane
4-FC6H4-CH2-PPh3Br
1,2-epoxy-n-octane
cyclopropylbenzene
32%
[a] If not stated otherwise, the yields were determined by GC-FID
analysis. [b] Determined by 19F NMR analysis using C6F6 as an internal
standard.
In general, functional groups that lack oxidative stability,
such as many heteroaromatic compounds, alcohols, and
primary amines, do not withstand the strong oxidative
reaction conditions of the electrolysis. Likewise, groups that
show reactivity towards iodonium ions, such as alkenes and
alkynes, are not compatible. However, a number of other
functional groups, such as aromatic halides, ester, amide, keto,
nitrile, nitro, or sulfone moieties, as well as phosphonium salts
were stable under these conditions.
To demonstrate an application of our method, we chose
the iododesilylation of Boc-protected 3-trimethysilylbenzyl
guanidine 5 as the Boc-deprotected variant of the iododesi-
lylation product is used as an 121I tracer in SPECT tomog-
raphy and as a tumour therapeutic in its 129I-marked
version.[12] The Boc-protected substrate was employed as we
knew from the compatibility test that primary amines are not
tolerated (Scheme 3). Nevertheless, this structure still has
a high functional group density, which renders it a challenging
Scheme 2. Results of the electrochemical TMS–iodine exchange reac-
tion. [a] Yield determined by GC-FID analysis using mesitylene as the
internal standard. [b] Yield of isolated product.
drawing substituents. Good to excellent yields of 70–97%
were mostly obtained for both electron-deficient substrates,
such as (4-fluorophenyl)trimethylsilane (4c, 88%), and
electron-rich substrates, such as (4-methoxyphenyl)trimethyl-
silane (4b, 73%). A steric influence as in our initial paper
could not be detected as the yields for para- and ortho-
substituted substrates were comparable. Very electron-poor
substrates (3 f) showed diminished reactivity, leading to low
yields or extended reaction times. Furthermore, (3-methoxy-
phenyl)trimethylsilane (3h) gave the corresponding product
in a poor yield of 31%, which, according to GC-MS analysis,
can be explained by direct iodination between the methoxy
and the TMS group without affecting the trimethylsilyl group.
To check for further functional group tolerance, a method
developed by Glorius and co-workers was used.[11] This
method is based on the addition of additives containing
a certain functional group. To test the reaction compatibility,
Scheme 3. Electrochemical TMS–iodine exchange for the synthesis of
the derivative of a tumour therapeutic.
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
substrate for an electrochemical reaction under oxidative
Conflict of interest
conditions. Fortunately, we were able to isolate the product 6
on a 0.25 mmol scale without the need for column purification
in 88%. This example not only demonstrates the overall good
functional group tolerance but furthermore a possible appli-
cation in the synthesis of isotopically labelled iodine mole-
cules for applications in the life sciences. As the reaction time
depends solely on the amount of used starting material and
reactions for isotope labelling are typically conducted on very
small scales, this process should be a highly useful alternative
to existing iodination methods.[12]
The authors declare no conflict of interest.
Keywords: arenes · compatibility test · design of experiments ·
electrochemistry · iododesilylation
[1] a) T. Dohi, Y. Kita in Iodine Chemistry and Applications (Ed.: T.
Kaiho), Wiley, Hoboken, 2014, pp. 303 – 310.
In conclusion, we have developed a new approach for the
synthesis of aryl iodides by electrochemical iododesilylation
that is atom-economic and only requires sulfuric acid as the
supporting electrolyte within the cathode compartment.
Various products were obtained in good to very good yields
upon rational optimization by design of experiments. The
broad functional group tolerance was illustrated with a large
variety of substrates and furthermore by a compatibility test.
Finally, we applied the method for the synthesis of Boc-
protected 3-iodobenzylguanidine, which yielded the product
in 88% yield without the need for further expensive
purification.
[2] a) L. Skulski, Molecules 2000, 5, 1331; b) E. A. Merritt, B.
Olofsson, Angew. Chem. Int. Ed. 2009, 48, 9052; Angew. Chem.
2009, 121, 9214; c) Y. Li, D. P. Hari, M. V. Vita, J. Waser, Angew.
Chem. Int. Ed. 2016, 55, 4436; Angew. Chem. 2016, 128, 4512;
d) A. Yoshimura, V. V. Zhdankin, Chem. Rev. 2016, 116, 3328.
[3] a) M. H. Bourguignon, E. K. J. Pauwels, C. Locꢀh, B. Maziꢁre,
Eur. J. Nucl. Med. 1997, 24, 331; b) S. J. Goldsmith, Semin. Nucl.
Med. 1975, 5, 125; c) M. H. Rønnest, F. Nissen, P. J. Pedersen,
T. O. Larsen, W. Mier, M. H. Clausen, Eur. J. Org. Chem. 2013,
3970.
[4] a) S. Stavber, M. Jereb, M. Zupan, J. Stefan, Synthesis 2008, 1487;
b) S. R. Waldvogel, K. M. Wehming, Sci. Synth. 2007, 31, 235.
[5] a) K. Midorikawa, S. Suga, J.-i. Yoshida, Chem. Commun. 2006,
3794; b) K. Kataoka, Y. Hagiwara, K. Midorikawa, S. Suga, J.-i.
Yoshida, Org. Process Res. Dev. 2008, 12, 1130.
[6] R. Mçckel, G. Hilt, Org. Lett. 2015, 17, 1644.
[7] G. Hilt, J. Janikowski, W. Hess, Angew. Chem. Int. Ed. 2006, 45,
5204; Angew. Chem. 2006, 118, 5328.
[8] Initial isodesmic and mechanistic DFT calculations suggested
that the addition of methanol leads to the formation of MeOI
instead of HSO4I as the iodinating species, which might be the
reason for the reduced extent of protodesilylation owing to the
aprotic character of methyl hypoiodite compared to HSO4I; see:
V. D. Filimonov, O. K. Poleshchuk, E. A. Krasnokutskaya, G.
Frenking, J. Mol. Model. 2011, 17, 2759.
Experimental Section
General procedure for the electrochemical iododesilylation: In
an H-type cell, the anode chamber was charged with the aryl
trimethylsilane (1.00 mmol, 1.00 equiv) and potassium iodide
(1.00 mmol, 1.00 equiv), and the cathode chamber was charged with
sulfuric acid (0.5 mL). The respective solvent was added simulta-
neously to the anode (10 mL) and cathode (9.5 mL + 0.5 mL H2SO4
conc.) chambers. The cell was equipped with platinum electrodes, and
the solutions were stirred until all solids had dissolved. The reaction
was electrolysed under constant current (6.7 mAcmÀ2) at room
temperature until the conversion (GC/MS monitoring; normally
complete conversion could be observed with the naked eye as the
colour of the anode chamber changed from dark red to yellow)
reached completion. The yield was either determined by GC analysis
(mesitylene as the standard) or the solution was diluted with diethyl
ether, washed with potassium carbonate and sodium thiosulfate, dried
over magnesium sulfate, and concentrated under reduced pressure. If
necessary, the crude product was purified by flash chromatography on
silica gel.
[9] JMP, Version 13.1., SAS Institute Inc., Cary, NC, 1989–2013.
[10] a) P. M. Murray, F. Bellany, L. Benhamou, D.-K. Bucar, A. B.
ˇ
Taborb, T. D. Sheppard, Org. Biomol. Chem. 2016, 14, 2373;
b) S. A. Weissman, N. G. Anderson, Org. Process Res. Dev. 2015,
19, 1605; c) E. N. Bess, A. J. Bischoff, M. S. Sigman, Proc. Natl.
Acad. Sci. USA 2014, 111, 14698; d) P. Renzi, M. Bella, Synlett
2017, 28, 306; e) R. Nuzzo, Nature 2014, 506, 150; f) a central
composite design consists of a full factorial plan in which each
factor is measured on two levels and an additional centre point
that is augmented with points on each face of the design (“star
points”) in order to measure curvature (quadratic functions of
the factors).
[11] a) K. D. Collins, F. Glorius, Nat. Chem. 2013, 5, 597; b) K. D.
Collins, A. Rꢂhling, F. Glorius, Nat. Protoc. 2014, 9, 1348; c) T.
Gensch, F. Glorius, Science 2016, 352, 294.
[12] a) G. Vaidyanathan, S. Shankar, M. R. Zalutsky, Bioconjugate
Chem. 2001, 12, 786; b) A. van Berkel, K. Pacak, J. W. M.
Lenders, Clin. Endocrinol. 2014, 81, 329; c) J. J. Mukherjee,
G. A. Kaltsas, N. Islam, P. N. Plowman, R. Foley, J. Hikmat, K. E.
Britton, P. J. Jenkins, S. L. Chew, J. P. Monson, G. M. Besser,
A. B. Grossman, Clin. Endocrinol. 2001, 55, 47.
Acknowledgements
We thank the Hochschulrechenzentrum Marburg and Old-
enburg for providing computer time and for the excellent
service. We thank C. Kohlmeyer for valuable help with the
synthesis of starting materials and for helpful discussions.
R.M. thanks Reinhart Mçckel for an introduction into
“design of experiments”.
Manuscript received: November 3, 2017
Version of record online: && &&, &&&&
4
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Electrosynthesis
Get me out of here: An electrochemical
access to iodinated aromatic compounds
starting from trimethylsilyl-substituted
arenes is presented. By design of experi-
ments, highly efficient and mild condi-
tions were identified for a wide range of
substrates. A functional group stability
test and the synthesis of an important 3-
iodobenzylguanidine radiotracer illus-
trate the scope of this process.
R. Mçckel, J. Hille, E. Winterling,
S. Weidemꢁller, T. M. Faber,
G. Hilt*
&&&& — &&&&
Electrochemical Synthesis of Aryl Iodides
by Anodic Iododesilylation
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!