Palladium catalyzed decarboxylative rearrangement of N-alloc indoles

Article abstract of DOI:10.1021/ol400110c

A highly efficient palladium catalyzed decarboxylative allylic rearrangement of alloc indoles has been developed. This can also be combined with a Suzuki-Miyaura cross-coupling reaction in a single pot transformation. Substituted alloc groups and benzylic variants have also been demonstrated alongside promising initial results on the enantioselective variant.

Full text of DOI:10.1021/ol400110c

ORGANIC
LETTERS
2013
Vol. 15, No. 5
1088–1091
Palladium Catalyzed Decarboxylative
Rearrangement of N‑Alloc Indoles
Jun Chen and Matthew J. Cook*
School of Chemistry and Chemical Engineering, Queen’s University Belfast, Belfast BT9
5AG, Northern Ireland
m.cook@qub.ac.uk
Received January 14, 2013
ABSTRACT
A highly efficient palladium catalyzed decarboxylative allylic rearrangement of alloc indoles has been developed. This can also be combined with a
SuzukiÀMiyaura cross-coupling reaction in a single pot transformation. Substituted alloc groups and benzylic variants have also been
demonstrated alongside promising initial results on the enantioselective variant.
Since its inception in the 1950s, π-allyl chemistry has
developed into many synthetically useful variants.1 A large
number of these processes have been further developed
into enantioselective processes with asymmetric allylic
alkylation reactions being particularly widely used.2
The use of indoles as nucleophilic partners in π-allyl
chemistry has been demonstrated by several groups. Bill-
ups first demonstrated the use of indoles in π-allyl chem-
istry using allylic acetates in 1980.3 Kimura then modified
this with π-allyl species generated from allylic alcohols also
reacting with indole to afford C-3 substitution.4 Trost later
developed this into an enantioselective variant whereby
C-3 substituited indoles react with allylic alcohol derived
π-allyl species to afford products with good enantio-
selectivity.5 Rawal demonstrated that 2,3-disubstituted
indoles could react with allyl carbonates in the presence
of a palladium(0) catalyst.6 The use of iridium catalysts has
also been shown to afford similar reactions, albeit with the
opposite regiochemistry.7
We speculated that indoles containing carbamate nitro-
gen-protecting groups could be used as π-allyl precursors.
Upon treatment with a palladium(0) catalyst these could
decarboxylate to afford a π-allyl intermediate and an
indole anion. The two could recombine at the C-3 position
affording an allylated indoline imine. This rearrangement
would effectively be an aromatic aza-Tsuji allylation
reaction.8 Herein we report the development of the dec-
arboxylative allyl rearrangement of N-allyl indoles.
We first began by looking at carbazole 1a and screened a
number of conditions including palladium sources, sol-
vents, and the use of additives (Table 1). Gratifyingly we
found that Pd(PPh3)4 did catalyze the reaction; however,
the reaction was slow with a significant amount of simple
alloc deprotection being observed. A screen of solvents
found that non-polar aprotic solvents such as toluene and
methylene chloride resulted in much faster reaction times
with less deprotection observed affording the C-3 allylated
product 2a in 90% yield. The mass balance of the reaction
was made up of small amounts of deprotected carbazole as
a byproduct. Both Kimura and Trost had used BEt3 as an
(1) Seminal papers: (a) Smidt, J.; Hafner, W. Angew. Chem. 1959, 71,
284. (b) Tsuji, J.; Takahashi, H.; Morikawa, M. Tetrahedron Lett. 1965,
6, 4387. (c) Hata, G.; Takahashi, K.; Miyake, A. J. Chem. Soc., Chem.
Commun. 1970, 1392. (d) Trost, B. M.; Dietsch, T. J. J. Am. Chem. Soc.
1973, 95, 8200.
(2) For reviews of asymmetric allylic alkylations, see: (a) Tsuji, J.
Pure Appl. Chem. 1999, 71, 1539. (b) Lu, Z.; Ma, S. Angew. Chem., Int.
Ed. 2008, 47, 258. (c) Trost, B. M.; Lee, C. Catalytic Asymmetric
Synthesis, 2nd ed; Ojima, I., Ed.;Wiley-VCH: New York, 2000; pp
593À649. (d) Pfaltz, A.; Lautens, M. Comprehensive Asymmetric Cata-
lysis; Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer: Heidelberg,
1999; pp 833À886. (e) Trost, B. M. Org. Process Res. Dev. 2012, 16, 195.
(3) Billups, W. E.; Erkes, R. S.; Reed, L. E. Synth. Commun. 1980, 10,
147.
(6) Kagawa, N.; Malerich, J. P.; Rawal, V. Org. Lett. 2008, 10, 2381.
(7) Lui, W.-B.; He, H.; Dai, L.-X.; You, S.-L. Org. Lett. 2008, 10,
1815.
(8) For a review of Tsuji reaction, see: Mohr, J. T.; Stoltz, B. M.
Chem.;Asian J. 2007, 2, 1476.
(4) Kimura, M.; Futamata, M.; Mukai, R.; Tamaru, Y. J. Am. Chem.
Soc. 2005, 127, 4592.
(5) Trost, B. M.; Quancard, J. J. Am. Chem. Soc. 2006, 128, 6314.
r
10.1021/ol400110c
Published on Web 02/19/2013
2013 American Chemical Society

additive in their reports, and the use of this was explored.
When BEt3 was used there was no observed deprotection;
however the reactionproceededmoreslowly. This could be
circumvented byincreasing the temperature to50°C which
restored the reaction to a yield of 90% with no observed
deprotection. We screened various other palladium/ligand
combinations, including those used by Rawal and found
that these were inferior for this particular reaction.
Scheme 1. Substrate Scopea
Table 1. Optimization Studies
entry
catalyst
Pd(PPh3)4
solvent
additive
yield (%)b
1
2
3
4
5
6
7
THF
none
none
none
BEt3
BEt3
BEt3
BEt3
12
80
90
45
90c
38
n.r.
Pd(PPh3)4
toluene
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd2dba3/P(2-Fur)3
Pd(OAc)2/PPh3
a Reactions carried out using 1 mol % of catalyst and 1 equiv of
additive (when used) at 0.2 M concentration. b Isolated yields. c Reaction
carried out at 50 °C.
With the optimized conditions in hand, we next explored
the substrate scope (Scheme 1). A variety of carbocyclic
variants were used, and the reaction was tolerant of
differing ring sizes from five to eight affording 77À90%
yields. It was necessary to reduce the five-membered indole
product 2b with NaBH3CN prior to purification, as this
was not stable to silica gel chromatography. We examined
whether an aryl bromide could be tolerated, and gratify-
ingly, the use of bromo carbazole 1e afforded the corre-
sponding rearranged product 2e in 85% yield with no sign
of insertion into the carbonÀbromine bond. Noncyclic
dimethyl derivative 1f could also be rearranged with 2f
being formed in 79% yield.
a Reaction conditions: 1 mol % Pd(PPh3)4, 1 equiv of BEt3 (1 M in
hexane), CH2Cl2 (0.2 M), 50 °C, 16 h. b Isolated yield. c Reduced with
NaBH3CN for isolation.
We next began to examine installing substitution and
in particular the installation of heteroatom substituents.
We discovered that the reaction was indeed tolerant of a
variety of groupsand difluoro-substrate2giswell tolerated
giving a yield similar to that of the nonfluorinated analog.
A piperidine can also be used with little effect on the yields,
as can ketal 2i and the tetracyclic substrate 2j. Once again
the five-membered substrate was unstable to silica gel and
therefore had to be reduced prior to purification. It is
important to mention that the functionalized substrates
required the use of BEt3 as an additive to obtain good
yields. In the absence of this additive a very slow reaction is
observed.
We next examined the use of substituted alloc groups.
These could be easily prepared using Sarpong’s protocol
without the need for phosgene based reagents.9 2-Methyl
substrate 3 rearranged under slightly modified conditions
toprovide goodyields ofproduct 4 (eq1). A higher catalyst
loading was required (5 mol % Pd(PPh3)4), and the addi-
tion of BSA10 as an additional additive gave optimal
results. Under the modified conditions, the reaction of
2,3-dimethyl alloc substrate 5 also proceeded well to give
(9) Heller, S. T.; Schultz, E. E.; Sarpong, R. Angew. Chem., Int. Ed.
2012, 51, 8304.
(10) BSA = N,O-Bis(trimethylsilyl)acetamide.
Org. Lett., Vol. 15, No. 5, 2013
1089

the linear substitution product 6 (eq 2). Likewise, cinnamyl
derivative 7 also rearranged in good yield to afford the
linear product 8 (eq 3). As Rawal had recently reported the
use of benzyl carbonates in a π-allyl substitution reaction
with 2,3-substituted indoles,11 we investigated whether
Cbz indoles could undergo a similar transformation. With
the conditions used for the substituted alloc groups, no
reaction was observed, and utilizing Rawal’s conditions, a
moderate 45% yield of benzyl migrated product 10 was
obtained. The combination of [Pd(allyl)Cl]2 and Xantphos
provided the optimal conditions for this rearrangement
with 10 being produced in 72% yield (eq 4) .
Scheme 2. RearrangementÀSuzuki Processa
a Reaction conditions: 1 mol % Pd(PPh3)4, 1 equiv of BEt3 (1 M in
hexane), CH2Cl2 (0.2 M), 50 °C, 16 h then 1.5 equiv of boronic acid, 2 M
Na2CO3, EtOH, toluene, 100 °C, 3À8 h. b Isolated yield.
already demonstrated the use of NaBH3CN to reduce
themasothergroupshadpreviously;6 thereforewefocused
on carbon-based nucleophiles. Initial attempts with
Grignard reagents proved unsuccessful; however the use
of organolithium reagents proved effective.12 Both butyl
lithium and phenyl lithium gave the addition products 12a
and 12b respectively as a single diastereoisomer (eq 5). This
was tentatively assigned as the syn-diastereoisomer based
on literature precedent.12
As the decarboxylative rearrangement process was tol-
erant of the aromatic bromide present in 2e, we examined
whether a second cross-coupling step could be performed
in the same reaction vessel. We decided to examine the
SuzukiÀMiyaura reaction due to the wide availability of
aryl boronic acids. Indeed this was possible with the
addition of a boronic acid, base, and ethanol following
the rearrangement step (scheme 2). This allowed for the
formation of C-3 allylated biaryl indolines in good to
moderate yields. Alongside the phenyl substituent 11a, a
range of electronically differentiated 4-substituted boronic
acids could be used including methyl 11b and trifluoro-
methyl11c. The 2-position could be substituted with fluoro
11d and 2,6-dimethyl groups 11e being incorporated.
Other tolerated groups were 3-cyano 11f and heterocycles
including pyridine 11g and furan 11h.
With the rearranged imines in hand we examined
whether we could add nucleophiles to these. We had
Finally we have briefly examined the enantioselective
variant of this reaction. Trost’s conditions that were
(11) Zhu, Y.; Rawal, V. J. Am. Chem. Soc. 2012, 134, 111.
(12) Rodriguez, J. G.; Urrutia, A. Tetrahedron 1998, 54, 15613.
Org. Lett., Vol. 15, No. 5, 2013
1090

previously reported used 9-BBN-Hex as an additive in the
place of BEt3 to increase enantioselectivity.5 Little or no
reactivity was observed when 9-BBN-Hex was used as an
additive, although Trost reported no examples of 2,3-
substituted indoles. Gratifyingly, the use of BEt3 and
Trost’s ANDEN ligand with ketalcarbazole 1i did yield a
rearranged product and afforded this product 2l in 62%
yield and 63% ee (eq 6). By contrast, Trost’s intermole-
cular allylic alcohol conditions gave similar results with
66% ee being obtained (eq 7). While far from being
optimized these are encouraging initial results suggesting
that an enantioselective variant can be developed.
In conclusion, we have developed the first example of a
decarboxylative rearrangement of N-alloc indoles. This
reaction is tolerant of functional groups and can also be
used for the rearrangement of substituted alloc groups as
well as Cbz protecting groups. A one-pot rearrange-
mentÀSuzuki coupling has also been developed, and the
use of the products has been explored. Finally we have
demonstrated that this reaction can be rendered enantio-
selective, and our current efforts are focused on improving
the enantioselectivity to synthetically useful levels.
Acknowledgment. We gratefully acknowledge Queen’s
University Belfast and Chirotech, Dr. Reddy’s Labora-
tories, Cambridge, UK for support. We also thank Glax-
oSmithKline and AstraZeneca for the generous donation
of laboratory equipment.
Supporting Information Available. Detailed experimen-
tal procedures, full characterization of all compounds
reported, and copies of NMR spectra (1H, 13C, and 19F
were applicable) are provided. This material is available
free of charge via the Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 5, 2013
1091