Divergent Access to (1,1) and (1,2)-Azidolactones from Alkenes using Hypervalent Iodine Reagents

Article abstract of DOI:10.1002/chem.201702599

A versatile synthesis of azidolactones through azidation and cyclization of carboxylic acids onto alkenes has been developed. Based on either photoredox or palladium catalysis, (1,1) and (1,2) azido lactones can be selectively synthesized. The choice of catalyst and benziodoxol(on)e reagent serving as azide source was essential to initiate either a radical or Lewis acid mediated process with divergent outcome. These transformations were carried out under mild conditions using a low catalyst loading and gave access to a large scope of azido lactones.

Full text of DOI:10.1002/chem.201702599

A Journal of
Accepted Article
Title: Divergent Access to (1,1) and (1,2)-Azidolactones from Alkenes
using Hypervalent Iodine Reagents
Authors: Sébastien Alazet, Franck Le Vaillant, Stefano Nicolai, Thibaut
Courant, and Jerome Waser
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201702599
Link to VoR: http://dx.doi.org/10.1002/chem.201702599
Supported by

10.1002/chem.201702599
Chemistry - A European Journal
COMMUNICATION
Divergent Access to (1,1) and (1,2)-Azidolactones from Alkenes using
Hypervalent Iodine Reagents
Sébastien Alazet, Franck Le Vaillant, Stefano Nicolai, Thibaut Courant and Jerome Waser[*]
Dedication
Abstract: A versatile synthesis of azidolactones through azidation
and cyclization of carboxylic acids onto alkenes has been
developed. Based on either photoredox or palladium catalysis,
(1,1) and (1,2) azido lactones can be selectively synthesized. The
choice of catalyst and benziodoxol(on)e reagent serving as azide
source was essential to initiate either a radical or Lewis Acid
mediated process with divergent outcome. These transformations
were carried out under mild conditions using a low catalyst
loading and gave access to a large scope of azido lactones.
for the transfer of functional groups via non-conventional
reactivity.[7]
AzidoBenziodoXolone (ABX, 3a) and
AzidoDimethyl-BenziodoXole (ADBX, 3b) (Scheme 1) have been
first described by Zhdankin and co-workers[8] and are stable solids
decomposing at temperature higher than 100 °C.[9] They have been
used as azide transfer reagents in presence of metal catalysts,
relying either on radical-based[8,10] or Lewis-acid mediated
pathways.[11] In particular, the azidation of styrene-type double
bonds has recently attracted much interest and ABX (3a) and
ADBX (3b) have emerged as good sources of azide radicals under
reductive conditions.[10a,h,j,l] In 2015, Greaney and co-workers
reported the photoredox-catalyzed nucleoazidation of styrenes
using the Sauvage/McMillin catalyst Cu(dap)2Cl[12] and ABX
reagent 3a, resulting in particularly mild reaction conditions.[10h]
On the other hand, the azidation of ketoesters and silyl enol ethers
has been successful in the presence of Lewis acids.[11]
Amino lactones have found widespread applications in natural
product synthesis and in medicinal chemistry.[1] Furthermore, they
serve as versatile starting materials for accessing other important
building blocks, such as amino alcohols.[2] As non-basic precursors
of amines, azides are highly useful synthetic intermediates on the
way to nitrogen-rich compounds. They can also be transformed
easily into various other nitrogen-containing functional groups.[3]
Therefore, azidolactones are versatile and convenient precursors of
both amino lactones and alcohols. Nevertheless, the preparation of
azidolactones is usually based on multi-step protocols and lacks
efficiency. The installation of the azido group commonly proceeds
through substitution reactions on pre-functionalized substrates,
Nevertheless,
to
the
best
of
our
knowledge,
azidobenziodoxol(on)es have never been used in the synthesis of
azidolactones.
such as halides.[4] Very recently,
a
copper catalyzed
enantioselective radical azidocarboxylation of alkenes was
developed by Buchwald and co-workers using phenyliodine
diacetate (PIDA) and TMSN3 as azide source, giving access to
(1,2)-azidolactones 2 (Scheme 1, A).[5] However, this method is
based on the in situ formation of highly reactive and potentially
explosive hypervalent iodine reagents for the generation of the
needed azide radical. Furthermore, this approach is limited to the
synthesis of (1,2)-azidolactones: access to the (1,1) regioisomers, if
possible from the same starting materials, would be highly
attractive to extend the range of accessible azidolactones. (1,1)-
Azidolactones are a largely unexplored class of compounds, which
have been reported only rarely in the literature.[6]
In recent years, benziodoxol(on)es, a class of stable cyclic
hypervalent iodine reagents, have emerged as privileged reagents
[*]
Dr. S. Alazet, F. Le Vaillant, Dr. S. Nicolai, Dr. T. Courant[+] and
Prof. Dr. J. Waser
Scheme 1. Synthesis of Azidolactones from Alkenes.
Laboratory of Catalysis and Organic Synthesis
Ecole Polytechnique Fédérale de Lausanne
EPFL SB ISIC LCSO, BCH 4306, 1015 Lausanne (CH)
Fax: (+)41 21 693 97 00
E-mail: jerome.waser@epfl.ch
Current address: Laboratoire COBRA – Bâtiment IRCOF UMR 6014
CNRS-INSA- Université de Rouen 1,
Herein, we report the first synthesis of azidolactones based on the
use of azidobenziodoxol(on)es (Scheme 1, B). The diverging
behavior of ABX (3a) and ADBX (3b) in presence of metal
catalysts led to either a radical-based or Lewis Acid mediated
pathway, allowing the selective formation of (1,2)- or (1,1)-
azidolactones 2 and 4 starting from styrene derivatives. The
former was achieved under mild photoredox conditions using
visible light irradiation and only 0.5 mol% Cu(dap)2Cl as catalyst,
[+]
Rue Tesnière, 76281 Mont-Saint-Aignan Cedex (France)
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702599
Chemistry - A European Journal
COMMUNICATION
whereas the later was realized by Lewis acid activation of the
hypervalent iodine reagent, involving an aryl migration step.
Furthermore, the synthesis of (1,2)-azidolactones starting from
olefins without arene substituent was possible using Lewis acid
activation.
In contrast, when ADBX (3b) was used instead of ABX (3a) under
the optimized photoredox conditions, only traces of (1,2)-azido-γ-
butyrolactones 2a were observed. The major product was 4-oxo-5-
phenylpentanoic acid (5), which was isolated in 40% yield.
(Equation 1) We speculated that 5 could originate from the
hydrolysis of a (1,1)-azidolactone 4a formed from reaction of the
hypervalent iodine reagent with the double bond followed by 1,2-
phenyl migration. In fact, such rearrangements have been observed
for oxygenation or fluorination,[14] but are unprecedented for
azidation with benziodoxole reagents. In this case, the copper
catalyst probably acted as a Lewis acid to activate ADBX (3b) and
not as a redox active catalyst able to generate the azide radical after
a SET event.
Based on the impressive work of Greaney and co-workers,[10h] we
started our investigations with the photoredox catalysis strategy for
the formation of (1,2)-azidolactone 2 from unsaturated carboxylic
acid 1 (Scheme 2). Only ABX (3a) was successful in this
transformation. Upon
a
fine adjustment of the reaction
conditions,[13] the desired azidolactone 2a could be isolated in 81%
yield with 0.5 mol% of the photoredox catalyst under blue LED
irradiation. Interestingly, while the reaction also worked in the
absence of light, conversion to product 2a dropped to 20%. This
indicated that a radical chain process was possible, but less
efficient. Using the combination of PIDA and TMSN3 for azide
radical formation,[5] only traces of 2a were detected.
With this simple protocol for azidation in hands, we investigated
the scope of the reaction (Scheme 2). A broad series of unsaturated
carboxylic acids
1 containing substituted aryl groups was
converted into the corresponding azidolactones 2 in good to
excellent yields. Electron rich (2b, 2c) as well as halogen-
substituted arenes (2d-f) were well tolerated. A more rigid benzene
backbone led to excellent yields of (1,2)-azido-phthalide
derivatives 2g-i. Dimethyl-substituted lactone 2j was also obtained
in 90% yield. A thiophene heterocycle was compatible with the
reaction conditions (2k, 87%). An aromatic ring bearing a group in
the ortho position was also tolerated (2l, 71%). Tricyclic (1,2)-
azido-γ-butyrolactone 2m was obtained in 65% yield. When enoic
acid 1n containing a 1,3-enyne motif was used, azidation product
2n was formed in 47% yield. Importantly, (1,2)-azido-δ-lactones
could be also obtained under the same reaction conditions,
although lower yields were observed (product 2o-q). Scale up of
the reaction demonstrated that even lower catalyst loadings were
possible: 1.0 g of 2e was isolated in 82% yield using only 0.05
mol% catalyst.
Equation 1. Reaction of 1a with ADBX (3b).
In fact, in presence of several Lewis acid catalysts such as
Zn(OTf)2, In(OTf)3 and Sn(OTf)2, (1,1)-azidolactone 4a could be
isolated as the major product.[13] Pd(hfacac)2 (2 mol%) was
identified as the best catalyst for this transformation leading to 4 in
75% yield by NMR. Sn(OTf)2 provided similar results, but with
lower reproducibility. The low catalyst loading is noteworthy, as
most reported oxidative rearrangements required stoichiometric
amounts of Lewis or Brønsted acids.[14] Although 4a was sensitive
to hydrolysis, it was still possible to purify it by column
chromatography (52% pure 4a isolated).
Scheme 3. Scope of (1,1)-azidolactonization of 1. (isolated yield
given, NMR yield in brackets). a) Isolated as a mixture with 2-(2-
iodophenyl)propan-2-ol.
The reaction worked efficiently in the presence of halogen and
methyl substituents on the aromatic ring, affording products 4b-f in
55-80% yield. A thiophenyl group was also able to migrate,
affording 4g in 52% isolated yield. A geminal dimethyl substituted
substrate 1h gave the more stable azido lactone 4h in 77% isolated
Scheme 2. Scope of the (1,2)-azidolactonization.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702599
Chemistry - A European Journal
COMMUNICATION
yield. Finally, (1,1)-azido-phthalide derivatives 4i, 4j, 4k were also
generated in high yields.
Due to the high sensitivity of the obtained (1,1)-azidolactones, the
direct modification of the crude products was then investigated
(Scheme 4). Copper catalyzed cycloaddition starting from 4k and
phenylacetylene gave triazole 6 as a stable crystalline solid in 65%
yield over two steps starting from the olefin. Hydrogenation of the
crude reaction mixture containing 4a in EtOAc, using 5 mol% of
Pd/C, afforded ketone 5 in 75% yield over two steps.[15] The γ-
ketoester could be selectively generated through the hydrogenation
of the crude material of 4a in methanol in 70% yield over two
steps. When using 10 mol% of Pt black as catalyst with (1,1)-
azido-γ-butyrolactone 4d in the presence of Boc2O, the N-Boc
protected Gamma AminoButyric Acid (GABA) derivative 8 was
isolated in 48% yield. The GABA motif is commonly found in
many bioactive compounds and is an important target in synthetic
chemistry.[16] In the case of (1,1)-azido-phthalide derivative 4k,
only azide reduction was observed under Pt/C catalysis and the
Scheme 5. Lewis acid catalyzed synthesis of (1,2)-azidolactones. a)
20 h of reaction. b) 10 mol% of Pd(hfacac)2 were used.
(1,1)-amino-phthalide
9 was isolated in 80% yield. These
successful transformations demonstrated the potential of (1,1)-
azidolactones as useful building blocks.
A proton transfer can also be expected from the carboxylic acid to
the more basic alkoxide, a key step to promote catalyst turnover
only possible with 3b. Ring opening through the attack of the
carboxylic acid delivers then the highly electrophilic intermediate
III. In case of aliphatic substituents, direct substitution with the
azide would occur, leading to (1,2)-azidolactones 2. In case of aryl
substituents, 1,2-migration via a phenonium ion or a non-classical
carbocation IV is faster and gives product 4a. Further works will
be needed to support this speculative mechanism.
Scheme 4. Product modifications. *Isolated yields over two steps
from 1. Reaction conditions: a) CuI (10 mol%), Et3N (3.0 equiv.),
phenylacetylene (2.0 equiv.), THF, 25°C; b) H2, Pd/C (5 mol%),
EtOAc, 25 °C; c) H2, Pd/C (2.5 mol%), MeOH, 25 °C; d) H2, Pt black
(10 mol%), Boc2O (2.0 equiv.), THF, 25 °C.
In order to investigate the generality of the 1,2-shift, we then
replaced the aryl substituent by a simple alkyl group (Scheme 5).
No (1,1)-azidolactone was obtained with a benzyl group: only
traces of (1,2)-azidolactonization product 2r could be observed.
1,2-Azidolactone 2s could be obtained in 40% with a phenyl group
in position to the alkene. Interestingly, compounds 2s is not
formed under photoredox conditions, as an aryl group on the
alkene is required for oxidation of the carbon centered radical
intermediate. Better yields were obtained with more rigid
substrates derived from benzoic acid (products 2g and 2t-v). -
lactone 2g can be synthesized in better yield under photoredox
conditions, but -lactone 2t-v can be accessed only with Lewis acid
catalysis, showing that the two methods are highly complementary.
Based on the precedence in literature on radical and Lewis acid
pathways with hypervalent iodine reagents,[10,14] a speculative
mechanism can be proposed for the developed transformations
(Scheme 6). The better results obtained with ABX (3a) under
photoredox conditions could be due to its expected stronger
oxidant character when compared to ADBX (3b).[17] In contrast,
the Lewis acid catalytic cycle would be initiated by the activation
of ADBX (1b).[18] The resulting more electrophilic complex I will
react with alkene 1 to give iodocyclopropylium cation II.
Scheme 6. Speculative reaction mechanism.
In conclusion, we have reported a general and versatile
synthesis of azidolactones starting from alkene-containing
carboxylic acids. A large range of (1,1) and (1,2)-azidolactones
were obtained selectively in high yields and with broad functional
group tolerance. The fine modulation of the reactivity possible for
benziodoxolone reagents was key to enable either a photoredox or
a Lewis acid pathway to access (1,1)- and (1,2)-azidolactones
respectively. These results further establish the exceptional
potential of cyclic hypervalent iodine reagents for group transfer
reactions and will allow a broader use of azidolactones as useful
building blocks in synthetic and medicinal chemistry.
This article is protected by copyright. All rights reserved.

10.1002/chem.201702599
Chemistry - A European Journal
COMMUNICATION
Acknowledgements
X. W. Dong, C. Fang, J. J. Liu, C. H. Hou, B. Tan, X. Y. Liu,
Nat. Commun. 2016, 7, 13852.
[11] a) Q.-H. Deng, T. Bleith, H. Wadepohl, L. H. Gade, J. Am.
Chem. Soc. 2013, 135, 5356; b) M. V. Vita, J. Waser, Org.
Lett. 2013, 15, 3246; c) M. V. Vita, P. Caramenti, J. Waser,
Org. Lett. 2015, 17, 5832;
[12] a) B. T. Ahn, D. R. McMillin, Inorg. Chem. 1978, 17, 2253;
b) C. O. Dietrichbuchecker, P. A. Marnot, J. P. Sauvage, J. R.
Kirchhoff, D. R. McMillin, J. Chem. Soc., Chem. Commun.
1983, 513.
We thank ERC (European Research Council, Starting Grant
iTools4MC, number 334840) and EPFL for financial support.
Keywords: Azides, lactones, hypervalent iodine, photoredox, 1,2
shift.
[1] a) G. Blaskó, D. J. Gula, M. Shamma, J. Nat. Prod. 1982, 45,
105; b) M. Bös, H. Stadler, J. Wichmann, F. Jenck, J. R.
Martin, J.-L. Moreau, A. J. Sleight, Helv. Chim. Acta 1998,
81, 525; c) C. Kaiser, C. J. Spagnuolo, T. C. Adams, V. H.
Audia, A. C. Dupont, H. Hatoum, V. C. Lowe, J. C. Prosser,
B. L. Sturm, L. Noronha-Blob, J. Med. Chem. 1992, 35,
4415; d) Y. Shimojima, H. Hayashi, J. Med. Chem. 1983, 26,
1370.
[2] a) S. C. Bergmeier, Tetrahedron 2000, 56, 2561; b) T. J.
Donohoe, C. K. A. Callens, A. Flores, A. R. Lacy, A. H.
Rathi, Chem. Eur. J. 2011, 17, 58.
[3] S. Bräse, C. Gil, K. Knepper, V. Zimmermann, Angew. Chem.
Int. Ed. 2005, 44, 5188.
[4] a) Z. Ding, R. B. Silverman, J. Med. Chem. 1992, 35, 885; b)
A. Gross, D. R. Borcherding, D. Friedrich, J. S. Sabol,
Tetrahedron Lett. 2001, 42, 1631; c) D. Scarpi, L. Bartali, A.
Casini, E. G. Occhiato, Eur. J. Org. Chem. 2013, 2013, 1306;
d) R. Surmont, G. Verniest, J. W. Thuring, P. ten Holte, F.
Deroose, N. De Kimpe, Org. Biomol. Chem. 2010, 8, 4514;
e) C. K. Tan, C. Le, Y.-Y. Yeung, Chem. Commun. 2012, 48,
5793.
[5] R. Zhu, S. L. Buchwald, J. Am. Chem. Soc. 2015, 137, 8069.
[6] a) M. V. Bhatt, K. S. Rao, G. V. Rao, J. Org. Chem 1977, 42,
2697; b) R. Guilhemat, M. Pereyre, M. Petraud, Bull. Soc.
Chim. Fr. 1980, 2, 334; c) H. Takeuchi, S. Eguchi, J. Chem.
Soc., Perkin Trans. 1 1988, 2149.
[13] See Supporting Information for a complete list of tested
reaction conditions.
[14] Selected examples: a) A. C. Boye, D. Meyer, C. K. Ingison,
A. N. French, T. Wirth, Org. Lett. 2003, 5, 2157; b) U. Farid,
F. Malmedy, R. Claveau, L. Albers, T. Wirth, Angew. Chem.,
Int. Ed. 2013, 52, 7018; c) M. Brown, R. Kumar, J. Rehbein,
T. Wirth, Chem. Eur. J. 2016, 22, 4030; d) N. O. Ilchenko, B.
O. A. Tasch, K. J. Szabó, Angew. Chem. Int. Ed. 2014, 53,
12897; e) J. Zhang, K. J. Szabó, F. Himo, ACS Catal. 2017, 7,
1093; f) G. C. Geary, E. G. Hope, A. M. Stuart, Angew.
Chem., Int. Ed. 2015, 54, 14911; g) L. Liu, T. Zhang, Y.-F.
Yang, D. Zhang-Negrerie, X. Zhang, Y. Du, Y.-D. Wu, K.
Zhao, J. Org. Chem. 2016, 81, 4058; h) A. Ulmer, C.
Brunner, A. M. Arnold, A. Pöthig, T. Gulder, Chem. Eur. J.
2016, 22, 3660; i) B. Zhou, T. Yan, X.-S. Xue, J.-P. Cheng,
Org. Lett. 2016, 18, 6128; For a review, see: j) F. V. Singh, T.
Wirth, Synthesis 2013, 45, 2499.
[15] In principle, a simple hydrolysis would also lead to the same
compound. Nevertheless, better yields were obtained under
hydrogenation conditions, probably because prior reduction
of the azide facilitates hydrolysis under mild conditions.
[16] a) M. Ordonez, C. Cativiela, Tetrahedron-Asymmetry 2007,
18, 3; b) P. G. Vasudev, S. Chatterjee, N. Shamala, P.
Balaram, Chem. Rev. 2011, 111, 657.
[17] The oxidation potential of azidobenziodoxol(on)e is not yet
known, but a similar trend as observed with Togni reagents
can be expected: Y. Yasu, T. Koike, M. Akita, Angew. Chem.,
Int. Ed. 2012, 51, 9567.
[7] a) Y. Li, D. P. Hari, M. V. Vita, J. Waser, Angew. Chem. Int.
Ed. 2016, 55, 4436; b) A. Yoshimura, V. V. Zhdankin, Chem.
Rev. 2016, 116, 3328.
[8] a) V. V. Zhdankin, A. P. Krasutsky, C. J. Kuehl, A. J.
Simonsen, J. K. Woodward, B. Mismash, J. T. Bolz, J. Am.
Chem. Soc. 1996, 118, 5192; b) V. V. Zhdankin, M.
McSherry, B. Mismash, J. T. Bolz, J. K. Woodward, R. M.
Arbit, S. Erickson, Tetrahedron Lett. 1997, 38, 21; c) S. Akai,
T. Okuno, M. Egi, T. Takada, H. Tohma, Y. Kita,
Heterocycles 1996, 42, 47.
[18] At this stage, a redox catalytic cycle involving different
oxidation states of palladium cannot be excluded, but it
appears less probable as several Lewis acids were able to
catalyze the reactions.
[9] M. V. Vita, J. Waser, Angew. Chem., Int. Ed. 2015, 54, 5290.
Caution: Reagent 3a displays an explosive decomposition
behaviour.
[10] a) B. Zhang, A. Studer, Org. Lett. 2013, 15, 4548; b) H. Yin,
T. Wang, N. Jiao, Org. Lett. 2014, 16, 2302; c) Y. Fan, W.
Wan, G. Ma, W. Gao, H. Jiang, S. Zhu, J. Hao, Chem. Comm.
2014, 50, 5733; d) A. Sharma, J. F. Hartwig, Nature 2015,
517, 600; e) R. R. Karimov, A. Sharma, J. F. Hartwig, ACS
Central Sci. 2016, 2, 715; f) N. Fuentes, W. Kong, L.
Fernández-Sánchez, E. Merino, C. Nevado, J. Am. Chem. Soc.
2015, 137, 964; g) W. Kong, N. Fuentes, A. García-
Domínguez, E. Merino, C. Nevado, Angew. Chem. Int. Ed.
2015, 54, 2487; h) G. Fumagalli, P. T. G. Rabet, S. Boyd, M.
F. Greaney, Angew. Chem. Int. Ed. 2015, 54, 11481; i) P. T.
G. Rabet, G. Fumagalli, S. Boyd, M. F. Greaney, Org. Lett.
2016, 18, 1646; j) M.-Z. Lu, C.-Q. Wang, T.-P. Loh, Org.
Lett. 2015, 17, 6110; k) Y. Shinomoto, A. Yoshimura, H.
Shimizu, M. Yamazaki, V. V. Zhdankin, A. Saito, Org. Lett.
2015, 17, 5212; l) Y.-A. Yuan, D.-F. Lu, Y.-R. Chen, H. Xu,
Angew. Chem., Int. Ed. 2016, 55, 534; m) Y. Wang, G.-X. Li,
G. Yang, G. He, G. Chen, Chem. Sci. 2016, 7, 2679; n) T.
Yang, H. Zhu, W. Yu, Org. Biomol. Chem. 2016, 14, 3376;
o) L. Li, Z. L. Li, F. L. Wang, Z. Guo, Y. F. Cheng, N. Wang,
This article is protected by copyright. All rights reserved.

10.1002/chem.201702599
Chemistry - A European Journal
COMMUNICATION
Entry for the Table of Contents
COMMUNICATION
Sébastien Alazet, Franck Le Vaillant,
Stéfano Nicolai, Thibaut Courant and
Jerome Waser*
Page No. – Page No.
From one to both: A versatile synthesis of azidolactones from alkenes and carboxylic
acids has been developed based on photoredox and palladium catalysis. (1,1) and
(1,2)-azido lactones can be selectively synthesized through the choice of benziodoxole
reagent and catalyst. These transformations have been carried out under mild
conditions using low catalyst loading and give access to a large scope of
azidolactones.
This article is protected by copyright. All rights reserved.