Regio- and stereoselective allylation and crotylation of indoles at C2 through the use of potassium organotrifluoroborate salts

Article abstract of DOI:10.1002/anie.201207978

A practical method for the allylation, prenylation, propargylation, and diastereoselective crotylation of indoles has been developed using air- and moisture-stable potassium organotrifluoroborate reagents (see scheme). Lewis acids such as BF₃×Et₂O promote addition to afford 2-allyl- and 2-crotylindolines in high yields and diastereoselectivities. Copyright

Full text of DOI:10.1002/anie.201207978

.
Angewandte
Communications
DOI: 10.1002/anie.201207978
Indole Functionalization
Regio- and Stereoselective Allylation and Crotylation of Indoles at C2
Through the Use of Potassium Organotrifluoroborate Salts**
Farhad Nowrouzi and Robert A. Batey*
Dedicated to Professor William B. Motherwell on the occasion of his 65th birthday
The site- and stereoselective functionalization of the indole
ring system is an area of considerable recent interest, because
of the importance of indoles as natural products, and as
a result of their privileged status in biologically active targets
ment of more convenient protocols for the allylation and
prenylation of the C2 position of indoles are therefore of
considerable interest. We now report a general and selective
C2 allylation of indoles using allylic trifluoroborate salts.
Over the last decade, several groups have reported the
utility of organotrifluoroborate salts[10] as air- and moisture-
stable reagents that act as synthetic equivalents to boronic
acids.[11,12] While most reports of these salts have focused on
their use for cross-coupling chemistry,[10] other synthetically
useful applications have also emerged. For example, additions
to electrophiles, including carbonyl groups and imines, under
a variety of conditions, including the use of Lewis acid, phase
transfer catalysis, montmorillonite K10 clay, indium and
rhodium(I) catalysis, have been reported.[13] In each of these
cases, the abstraction of fluoride ions presumably occurs to
initially generate an allyldifluoroborane species, which can
either react directly with the electrophile or be converted into
another reactive allylation agent. We considered that the
intermediacy of an in situ generated allyldifluoroborane
species might similarly allow the addition to indoles via
their 3H-indole tautomers. Initial reaction optimizations were
carried out using potassium allyltrifluoroborate 1a and indole
2a. Activation by K10 led to only moderate yields of indoline
3a because of competing side reactions of 2a, while the use of
an indium-based protocol, which had been successfully
developed for additions to ketones,[13g] did not give the
desired product 3a (Table 1, entries 1 and 2). The use of
a stoichiometric amount of BF3·Et2O afforded 3a in good
yield after 30 h (Table 1, entry 3). Use of a catalytic amount of
BF3·Et2O (15 mol%) led to side reactions (indole dimeriza-
tion[14]) at higher concentrations, while conversion was poor at
lower concentrations. Catalytic Yb(OTf)3 was also capable of
promoting addition of 1a to 2a (Table 1, entry 4). Application
of some of these conditions to reactions of indoles bearing
substituents at the 2- and 3-positions and E- and Z-crotyltri-
fluoroborate salts 1b and 1c generally gave lower conversions
with catalytic amounts of Lewis acids, while good conversions
and yields were achieved using stoichiometric amounts of
BF3·Et2O (Table 1, entries 5–15). Crotylations of 2a occurred
stereospecifically, giving 3b and 3c from reactions of 1b and
1c, respectively. The use of 2-methylindole required longer
reaction times in order to achieve full conversion, while
reactions of 1c occurred more slowly than the corresponding
reactions using 1a or 1b. N-Methylindole failed to give the
desired product, even with a stoichiometric amount of
BF3·Et2O. Finally, reactions using the pinacol ester of
allylboronic acid rather than 1a gave poor results.
and pharmaceuticals.[1] Approaches for selective C C bond
À
functionalization of indole rings include metal-catalyzed and
organocatalyzed reactions,[2–5] which allow direct formation of
both 3- and 2-substituted indoles or their indoline (2,3-
dihydro-1H-indole) counterparts. Despite the obvious syn-
thetic potential of the direct addition of organometallic
nucleophiles to the C2 position of indoles, such strategies are
only rarely encountered,[1] in large part because the indole
À
ring itself is nucleophilic, and because the indole N H bond is
prone to deprotonation and consequent deactivation of the
indole core for nucleophilic attack. In this context, the studies
of Bubnov and co-workers on the addition of triallylborane
and triprenylborane to indoles and other heterocycles are
particularly noteworthy.[6,7] However, the low reactivity of the
indole ring toward nucleophilic addition necessitates the use
of the highly reactive triallylborane reagent at elevated
temperatures, with addition occurring through reaction with
the 3H-indole imine tautomer. The major disadvantages of
this approach are the high reactivity of the borane reagents
toward other functional groups, and the highly oxygen-
sensitive and pyrophoric nature of these reagents, which
necessitate specific handling techniques and preparation
generally immediately prior to use. This requirement has
limited the utility of this approach, and less-reactive nucle-
ophiles, such as allylboronate esters, unfortunately do not
undergo addition to indoles. The closely related reverse-
prenylation protocol of Danishefsky and co-workers has
utilized the more stable prenyl-9-BBN reagent.[8] However,
this reagent must generally be used in excess to achieve
acceptable yields in the addition reaction, and is inconvenient
to prepare, as it requires sequential distillations for purifica-
tion. Nevertheless, despite the problems associated with C2
allylation and prenylation methods, they have found utility in
numerous complex natural product syntheses.[9] The develop-
[*] Dr. F. Nowrouzi, Prof. Dr. R. A. Batey
Department of Chemistry, University of Toronto
80 St. George St., Toronto, ON, M5S 3H6 (Canada)
E-mail: rbatey@chem.utoronto.ca
Homepage: http://www.chem.utoronto.ca/staff/RAB/index.html
[**] The Natural Science and Engineering Research Council (NSERC) of
Canada supported this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201207978.
892
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 892 –895

Angewandte
Chemie
Table 1: Optimization of the reactions of indoles 2 with allyltrifluorobo-
rate salt 1a and crotyltrifluoroborate salts 1b and 1c.[a]
Table 2: Substrate scope for the allylation of indoles using potassium
allyltrifluoroborate 1a.[a]
Entry
X
2
3
Yield
[%][b]
1
2
3
4
5
6
7
8
H
5-Br
5-Cl
5-F
5-OH
5-CO2Me
7-Me
3-Me
2-Me
2-Ph
2a
2d
2e
2 f
2g
2h
2i
2b
2c
2j
3a
3 f
3g
3h
3i
92
90
92
79
90
86
Entry
2
1
Cond.[a]
Product
Conv.[b] [%]
(Yield)
1[c]
2
3
2a
2a
2a
2a
1a
1a
1a
1a
A
B
C
D
3a
3a
3a
3a
50
ꢀ5
3j
100 (92)
100 (79)
3k
3d
3e
3l
87[c]
87[d]
91
4
9
10
11
5
6
2a
2a
1b
1b
A
C
3b
3b
100 (72)
100 (88)
85[c]
38[e]
2k
3m
7
8
9
2a
2a
2a
1c
1c
1c
A
C
D
3c
3c
3c
<10
100 (89)
30
12
2l
3n
88[f]
10
11
12
2b
2b
2b
1a
1a
1a
A
C
D
3d
3d
3d
ꢀ5
100 (83)[d]
65[e]
[a] Reactions were carried out on a 0.50 mmol scale with respect to the
indole substrates at a concentration of 0.125m for 6-72 h. [b] Isolated
yields. [c] 3.0 equiv of 1a and BF3·Et2O were required over 62 h to achieve
full conversion. [d] d.r.=2.1:1, based on 1H NMR analysis of the crude
reaction mixture. [e] 52% conversion after 74 h. [f] 4.0 equiv of 1a was
used.
13
14
15
2c
2c
2c
1a
1a
1a
A
C
D
3e
3e
3e
ꢀ5
91
95
[a] Reactions were carried out for 30 h using substrates 1a and 1b and
for 72 h using substrate 1c. Conditions: A) montmorillonite K10 clay,
CH2Cl2, H2O; B) In (1.0 equiv), CH2Cl2, H2O; C) BF3·Et2O (100 mol%),
CH2Cl2; D) Yb(OTf)3 (15 mol%), CH2Cl2. [b] Conversion was monitored
by 1H NMR analysis of the crude products. [c] The reaction was
performed under an inert atmosphere for 48 h. [d] The d.r. was 2.1:1,
based on 1H NMR analysis of the crude product. [e] The d.r. was 1.7:1,
based on 1H NMR analysis of the crude product.
unsuccessful, but the reaction using an excess of 1a
(4.0 equiv) led to the formation of a,a-diallylated indole 3n
as the sole product (Table 2, entry 12).
The scope of the procedure with other salts was also
investigated (Table 3). Diastereospecific crotylation of indole
occurred in good yields and diastereoselectivities using E- and
Z-crotyltrifluoroborates (Table 3, entries 1–2). Reaction of
the prenyltrifluoroborate salt 1d also occurred in good yield
to give the reverse-prenylated indoline 3o (Table 3, entry 3).
Reaction of the allenyltrifluoroborate salt 1e under identical
conditions led to propargylated indole 3p as the sole product
(Table 3, entry 4). This is the first example of a propargylation
of indole using an allenylboron reagent. Finally, reaction of 2c
with 1b also occurred to give 3q with a good d.r.
The 2-substituted indoline products 3 are valuable syn-
thetic intermediates.[1,3] A short metathesis-based synthetic
route served to illustrate this, affording the diastereoisomeric
dihydropyridoindolone products 5a and 5b in good yields
from 3b and 3c, respectively (Scheme 1). The relative
stereochemical assignment of 3b and 3c was also established,
based upon the rigid cyclic structures of 5a and 5b, through
a comparative analysis of 3J and 4J (1H, 1H) coupling constants
and NOE correlations.[15]
Reactions of 1a with a variety of substituted indoles were
surveyed under the optimized conditions with reaction times
of 6–72 h (Table 2). Excellent yields were obtained using 5-
substituted indoles, including those with electron-donating
and electron-withdrawing substituents (Table 2, entries 2–6).
Steric effects of substituents at the 2-, 3- and 7-positions of the
indole ring were detrimental to the reaction progress, and
longer reaction times were required to achieve full conver-
sion. For example, full conversion was observed only after
72 h in reactions of the methyl-substituted indoles (Table 2,
entries 7–9). Addition of 1a to 3-methylindole gave 3d in
good yield, but with modest diastereoselectivity that favoured
the trans product (Table 2, entry 8). Reactions of 2-substi-
tuted indoles required a longer reaction time and/or larger
quantities of reagents to achieve full conversions (Table 2,
entries 9 and 10). For example, 2-phenyl indole required
3.0 equivalents of 1a and BF3·Et2O over 74 h to afford 3l in
85% yield of isolated product. Allylation of the sterically
demanding substrate 2k was slow (Table 2, entry 11).
Attempts to selectively monoallylate 3-chloroindole were
The observed results are consistent with additions of
allylic boron difluoride intermediates to the Z-configured
cyclic imine 3H-indole tautomers. Addition was not observed
with N-methylindole, a substrate that cannot form the
Angew. Chem. Int. Ed. 2013, 52, 892 –895
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
893

.
Angewandte
Communications
Table 3: Substrate scope with respect to the allylating agent 2.[a]
Entry
1
2
1
3
Yield [%][b]
(d.r.)[c]
2a
1b
3b
95 (ꢁ95:5)
2
3
4
2a
2a
2a
1c
1d
1e
3c
3o
3p
83[d] (ꢁ95:5)
89
83
Scheme 2. Plausible transition states for the crotylation of indole.
In summary, a general method for regio- and diastereo-
selective allyl-, crotyl-, allenyl-, and prenylation at the 2-
position of unprotected indoles has been developed. The
stability of the organotrifluoroborate salts that are used
provides an operationally straightforward protocol for indo-
line synthesis, and avoids the use of sensitive triallylboron,
prenyl-9-BBN, or related borane-based reagents. The syn-
thetic significance of the reaction derives from the importance
of indoles as synthetic targets, the ready availability of indole
precursors, and the mild conditions employed. This study also
illustrates the significant potential for identifying new reac-
tivity profiles of organotrifluoroborate salts beyond their
widespread application for cross-coupling chemistry. Further
studies on the addition reactions of these salts to indoles,
heterocycles, and other electrophiles will be reported in due
course.
5
2c
1b
3q
91 (95:5)
[a] Reactions were carried out on a 0.50 mmol scale with respect to the
indole substrates using the conditions shown in Table 2. [b] Yields of
isolated products. [c] d.r. determined by 1H NMR analysis of the crude
products through comparison of resonances of the methyl groups with
the most-upfield chemical shift. [d] Full conversion achieved using
trifluoroborate salt 1c (3.0 equiv) and BF3·Et2O (2.0 equiv) over 74 h.
Received: October 3, 2012
Published online: December 6, 2012
Keywords: addition reactions · heterocycles · indoles ·
.
indolines · organotrifluoroborate
Scheme 1. Synthesis of 5a and 5b from 3b and 3c, respectively.
[1] For general reviews of indole chemistry, see: a) “Heterocyclic
Scaffolds II: Reactions and Applications of Indoles”: Topics in
Heterocyclic Chemistry, Vol. 26 (Ed.: G. W. Gribble), Springer,
Berlin, 2010; b) R. J. Sundberg, Indoles, Academic Press,
London, 1996; c) D. S. C. Black, R. J. Sundberg, G. W. Gribble
in Comprehensive Heterocyclic Chemistry II (Eds.: A. R.
Katritzky, C. W. Rees, E. F. V. Scriven), Elsevier, Oxford, 1996,
pp. 39 – 257; d) H. Dçpp, D. Dçpp, U. Langer, B. Gerding,
Houben-Weyl, Vol. E6b, Thieme, Stuttgart, 1994, pp. 546 – 1336.
[2] For selected review references, see: a) G. Bartoli, G. Bencivenni,
R. Dalpozzo, Chem. Soc. Rev. 2010, 39, 4449 – 4465; b) E. M.
Beck, M. J. Gaunt, Top. Curr. Chem. 2009, 292, 85 – 121; c) M.
Bandini, A. Eichholzer, Angew. Chem. 2009, 121, 9786 – 9824;
Angew. Chem. Int. Ed. 2009, 48, 9608 – 9644; d) I. V. Seregin, V.
Gevorgyan, Chem. Soc. Rev. 2007, 36, 1173 – 1193; e) G. R.
Humphrey, J. T. Kuethe, Chem. Rev. 2006, 106, 2875 – 2911; f) S.
Cacchi, G. Fabrizi, Chem. Rev. 2005, 105, 2873 – 2920.
requisite imine tautomer. The stereospecific crotylation of
indole using the E- and Z-crotyltrifluoroborate salts to afford
the anti and syn products 3b and 3c, respectively, is consistent
with the addition occurring via a Zimmerman–Traxler-like
transition state (Scheme 2). Thus, reactions via chair-like
transition states 6a and 6d are favored over reactions via
boat-like transition states 6b and 6c as a result of reduced
syn-pentane-like interactions. The lower reactivity of the Z-
crotyltrifluoroborate relative to its E-isomer can be attributed
to greater destabilization based on the 1,3-diaxial-like steric
interaction that occurs between the methyl group of the
trifluoroborate and the indole ring in transition state 6d
relative to 6a. The lower reactivity of C7-substituted indole
derivatives can also be rationalized as a result of unfavourable
steric interactions in the transition states.
[3] For selected recent examples of indoline syntheses, see: a) T.
Saget, S. J. Lemouzy, N. Cramer, Angew. Chem. 2012, 124, 2281 –
2285; Angew. Chem. Int. Ed. 2012, 51, 2238 – 2242; b) G. He, C.
Lu, Y. Zhao, W. A. Nack, G. Chen, Org. Lett. 2012, 14, 2944 –
894
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 892 –895

Angewandte
Chemie
2947; c) K. H. Kang, J. Do, Y. S. Park, J. Org. Chem. 2012, 77,
808 – 812; d) S. Anas, A. Cordi, H. B. Kagan, Chem. Commun.
2011, 47, 11483 – 11485; e) A. Baeza, A. Pfaltz, Chem. Eur. J.
2010, 16, 2036 – 2039; f) E. E. Maciver, S. Thompson, M. D.
Smith, Angew. Chem. 2009, 121, 10164 – 10167; Angew. Chem.
Int. Ed. 2009, 48, 9979 – 9982; g) T.-S. Mei, X. Wang, J.-Q. Yu, J.
Am. Chem. Soc. 2009, 131, 10806 – 10807; h) P. H. Fuller, J.-W.
Kim, S. R. Chemler, J. Am. Chem. Soc. 2008, 130, 17638 – 17639.
[4] For a review of methods for indoline synthesis, see: D. Liu, G.
Zhao, L. Xiang, Eur. J. Org. Chem. 2010, 3975 – 3984.
[9] See for example: a) C. A. Kuttruff, H. Zipse, D. Trauner, Angew.
Chem. 2011, 123, 1438 – 1441; Angew. Chem. Int. Ed. 2011, 50,
1402 – 1405; b) J. M. Finefield, R. M. Williams, J. Org. Chem.
2010, 75, 2785 – 2789.
[10] For reviews of organotrifluoroborates, see: a) G. A. Molander,
N. Ellis, Acc. Chem. Res. 2007, 40, 275 – 286; b) H. A. Stefani, R.
Cella, A. S. Vieira, Tetrahedron 2007, 63, 3623 – 3658; c) S.
Darses, J.-P. GenÞt, Chem. Rev. 2008, 108, 288 – 325.
[11] E. Vedejs, S. C. Fields, M. R. Shrimpf, J. Am. Chem. Soc. 1993,
115, 11612 – 11613.
[5] There are many examples of allylation reactions of the C3
position of indoles, see for example: a) M. Bandini, A. Melloni,
A. Umani-Ronchi, Org. Lett. 2004, 6, 3199 – 3202; b) B. M. Trost,
J. Quancard, J. Am. Chem. Soc. 2006, 128, 6314 – 6315; c) N.
Kagawa, J. P. Malerich, V. H. Rawal, Org. Lett. 2008, 10, 2381 –
2384.
[12] For general reviews of organoboron chemistry, see: a) D. S.
Matteson, Stereodirected Synthesis with Organoboranes,
Springer, Berlin, 1995; b) Boronic Acids: Preparation and
Applications in Organic Synthesis Medicine and Materials,
Vol. 1 and 2, 2nd ed. (Ed.: D. G. Hall), Wiley-VCH, Weinheim,
2011.
[6] a) Y. N. Bubnov, V. C. Bogdanov, B. M. Mikhailov, Zh. Obshch.
Khim. 1968, 38, 260 – 270; b) Y. N. Bubnov, Pure Appl. Chem.
1994, 66, 235 – 244; c) Y. N. Bubnov in Advances in Boron
Chemistry (Ed. W. Siebert), The Royal Society of Chemistry,
Cambridge 1997, pp. 123 – 138; d) N. Y. Kuznetsov, V. N. Khrus-
talev, I. A. Godovikov, Y. N. Bubnov, Eur. J. Org. Chem. 2006,
113 – 120; e) Y. N. Bubnov, E. V. Klimkina, I. V. Zhun, F. V.
Pastukhov, I. V. Yampolsky, Pure Appl. Chem. 2000, 72, 1641 –
1644; f) Y. N. Bubnov, N. Y. Kuznetsov, M. E. Gurskii, A. L.
Semenova, G. D. Kolomnikova, T. V. Potapova, Pure Appl.
Chem. 2006, 78, 1357 – 1368.
[13] a) R. A. Batey, A. N. Thadani, D. V. Smil, Org. Lett. 1999, 1,
1683 – 1686; b) R. A. Batey, A. N. Thadani, D. V. Smil, A. J.
Lough, Synthesis 2000, 990 – 998; c) A. N. Thadani, R. A. Batey,
Org. Lett. 2002, 4, 3827 – 3830; d) A. N. Thadani, R. A. Batey,
Tetrahedron Lett. 2003, 44, 8051 – 8055; e) S. W. Li, R. A. Batey,
Chem. Commun. 2004, 1382 – 1383; f) F. Nowrouzi, A. N.
Thadani, R. A. Batey, Org. Lett. 2009, 11, 2631 – 2634; g) F.
Nowrouzi, J. Janetzko, R. A. Batey, Org. Lett. 2010, 12, 5490 –
5493.
[14] H. Ishii, E. Sakurada, K. Murakami, J. Chem. Soc. Perkin Trans.
1 1988, 2387 – 2395.
[7] T. R. Ramadhar, R. A. Batey, Synthesis 2011, 1321 – 1346.
[8] J. M. Schkeryantz, J. C. G. Woo, P. Siliphaivanh, K. M. Depew,
S. J. Danishefsky, J. Am. Chem. Soc. 1999, 121, 11964 – 11975.
[15] NOE correlations between C9a-H and C9-Me were strong for
5a, but absent for 5b.
Angew. Chem. Int. Ed. 2013, 52, 892 –895
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
895