A novel approach to 3-methylindoles by a heck/cyclization/isomerization process

Article abstract of DOI:10.1021/ol902636v

Chemical Equation presented A novel synthesis of 3-methyllndoles from chlorotrlflates through a Heck reaction, carbamate/aryl chloride coupling, and Isomerlzatlon sequence Is presented. The three-step sequence Is highly efficient and general, enabling the

Full text of DOI:10.1021/ol902636v

ORGANIC
LETTERS
2010
Vol. 12, No. 4
668-671
A Novel Approach to 3-Methylindoles by
a Heck/Cyclization/Isomerization Process
Carl A. Baxter,* Ed Cleator,* Mahbub Alam, Antony J. Davies, Adrian Goodyear,
and Michael O’Hagan
Department of Process Research, Merck Sharp and Dohme Research Laboratories,
Hertford Road, Hoddesdon, Hertfordshire, EN11 9BU, United Kingdom
carl_baxter2@merck.com; edward_cleator@merck.com
Received November 16, 2009
ABSTRACT
A novel synthesis of 3-methylindoles from chlorotriflates through a Heck reaction, carbamate/aryl chloride coupling, and isomerization sequence is
presented. The three-step sequence is highly efficient and general, enabling the regiocontrolled synthesis of substituted indoles in short order.
Functionalized indoles are prevalent structural motifs that
are found in many natural products and drug molecules, and
as such there has been a considerable synthetic effort directed
toward their synthesis. These endeavors have provided the
synthetic community with a variety of methods for the
synthesis of variously substituted indoles with the subject
extensively covered in a number of review articles.1
exemplified in the Larock2 indole synthesis as well as the
cyclization of 2-alkynylaniline derivatives (Scheme 1).3 Even
Scheme 1. Larock, Barluenga, and Willis Approaches to Indoles
Despite the existence of many classical routes to indoles, most
of these have been superseded by transition metal-catalyzed
methods, which afford the indole in a more selective fashion,
while tolerating a wide variety of functional groups. Extensive
work has demonstrated the efficiency of the palladium-catalyzed
synthesis of indoles.1d Most of these reported methods for the
preparation of indoles have concentrated on building the five-
membered ring of the indole onto a suitably substituted benzene
ring; usually the indole nitrogen is present as an aniline and
the heterocycle is then formed via carbon-carbon (C-C) bond
construction.1b Alternatively, C-C bond formation may be
followed by cyclization onto a suitably positioned aniline as
though this chemistry is widely applicable, some highly
substituted halo-anilines are not easily accessible, and the alkyne
partner can be expensive. A less developed strategy toward
indoles, involving C-C bond formation prior to cyclization in
a carbon-nitrogen (C-N) bond forming process onto an aryl
halide, has been the focus of recently published work. The
Barluenga group demonstrated the synthesis of indoles through
(1) (a) Humphrey, G. R.; Kuethe, J. T. Chem. ReV. 2006, 106, 2875–
2911. (b) Cacchi, S.; Fabrizi, G. Chem. ReV. 2005, 105, 2873–2920. (c)
Thansandote, P.; Lautens, M. Chem.sEur. J. 2009, 15, 5874–5883. (d) Li,
J. J.; Gribble, G. W., Eds. Palladium in Heterocyclic Chemistry; Pergamon:
Oxford, UK, 2000.
(2) (a) Larock, R. C.; Yum, E. K. J. Am. Chem. Soc. 1991, 113, 6689–
6690. (b) Larock, R. C.; Yum, E. K.; Refvik, M. D. J. Org. Chem. 1998,
63, 7652–7662.
(3) (a) Sakamoto, T.; Kondo, Y.; Yamanaka, H. Heterocycles 1986, 24,
31–32. (b) Kakamoto, T.; Kondo, Y.; Yamanaka, H. Heterocycles 1986,
24, 1845–1847.
10.1021/ol902636v  2010 American Chemical Society
Published on Web 01/14/2010

a sequence of C-arylation and intramolecular N-arylation when
dihalobenzene derivatives were reacted with anions derived
from imines.4a While this provided a new approach to 2,3-
substituted indoles, dihalobenzene precursors are not readily
available, and the imine precursors also required preparation.
During the course of our studies the Barluenga group extended
this work to o-chlorosulfonates; however, the efficiency of the
indole formation was highly dependent on the aryl substitution
pattern, and chlorononaflates were in some cases required to
obtain good yields.4b Another example of indole formation by
amination of a suitably functionalized benzene ring has been
disclosed by Willis et al. This group employed a palladium-
catalyzed double amination of bis-activated styrene with primary
amines to generate the corresponding sterically demanding
N-substituted indoles.5 Again, while this approach provided a
new entry into indole systems the more elaborate styrenyl
precursors can be challenging to prepare.6
blocks. Hence, a Heck reaction between 3 and N-Boc allylamine
would give 4 (Scheme 3). Cyclization (Cu or Pd mediated using
the o-halide) would then give 5; it was unknown whether the
double bond would isomerize during the cyclization or remain
exo and require a separate isomerization. Jørgensen has shown
that 2-iodobromobenzenes react with allylamine in a palladium-
catalyzed aryl amination-Heck cyclization cascade to give
indoles directly, with amination occurring prior to the Heck
cyclization.8 However, most dihalogenated reagents are not
commercially available and would require multistep sequences
for their production, dictating that this method would not lead
to a general process for indole construction.
Scheme 3. Proposed Indole Synthesis
During a recent program we were challenged to develop a
reliable synthesis of a highly substituted 3-methylindole frag-
ment. The substitution pattern of the indole fragment often
dictates which particular indole synthesis will be suitable, with
starting material availability and functional group tolerance
being two major considerations. Most currently available indole
syntheses showed limitations, such as poor selectivity or the
requirement of lengthy syntheses of suitably functionalized aryl
precursors when applied to our target. This led us to initiate a
fundamentally new strategy for the synthesis of our desired indole.
During an exhaustive examination of the current literature,
we were drawn to the work of Hallberg, who has demon-
strated a regioselective Heck reaction between aryl triflates
such as 1 and N-Boc allylamine to give the corresponding
2-aryl-substituted allylcarbamate 2 (Scheme 2).7 Overall the
method lacked generality, while the yield for the o-chloro
example we required was poor under thermal conditions; the
reaction required microwave irradiation to obtain a 63%
yield, and these conditions are difficult to achieve on scale.
Cognizant that both the accessibility of the reagents and the
simplicity of the indole synthesis contribute to the overall
effectiveness of the transformation, our investigations began by
studying chlorotriflates 6 (prepared in one step from the
corresponding commercially available 2-chlorophenols).9 The
chlorotriflate was chosen as chlorophenols are widely available
from commercial sources.10 Subjecting both 6a and 6b to the
Heck reaction conditions developed by Hallberg, we found that
considerable amounts of phenol 8 or des-chloro 9 were
generated, especially with 6b (Scheme 4). As well as the
problematic hydrolysis, poor conversions of 6a (46%) were
observed. While this reaction provided the key carbamates 7a
and 7b for cyclization studies, to support our novel indole
preparation, the Heck reaction required optimization.
Scheme 4. Heck Reaction on Chlorotriflate 6
Scheme 2. Regioselective Heck Reaction, Hallberg et al.
We reasoned that if the Heck reaction could be improved
and conditions for the cyclization of this carbamate onto the
o-aryl chloride could be identified, then the indole could be
constructed in short order from commercially available building
It was believed that the choice of base could be critical in
minimizing the formation of the phenol and hence provide
a more efficient Heck reaction. A screen of inorganic and
organic bases in the Heck reaction with aryl triflate 6b was
performed.11 From this screen NaOAc was identified as a
more efficient base for the transformation, giving 87%
conversion, with 79% 7b and 8% des-Cl 9b. The amount of
N-Boc allylamine was also reduced from 3 equiv to 1.2
equiv, further improving the original Hallberg work.
(4) (a) Barluenga, J.; Jime´nez-Aquino, A.; Valde´s, C.; Aznar, F. Angew.
Chem., Int. Ed. Engl. 2007, 46, 1529–1532. (b) Barluenga, J.; Jime´nez-Aquino,
A.; Aznar, F.; Valde´s, C. J. Am. Chem. Soc. 2009, 131, 4031–4041.
(5) (a) Willis, M. C.; Brace, G. N.; Holmes, I. P. Angew. Chem., Int.
Ed. Engl. 2005, 44, 403–406. (b) Fletcher, A. J.; Bax, M. N.; Willis, M. C.
Chem. Commun. 2007, 4764–4766. (c) Hodgkinson, R. C.; Schulz, J.; Willis,
M. C. Org. Biomol. Chem. 2009, 7, 432–434.
(6) Tadd, A. C.; Matsuno, A.; Fielding, M. R.; Willis, M. C. Org. Lett.
2009, 11, 583–586.
Org. Lett., Vol. 12, No. 4, 2010
669

These new conditions were employed with chlorotriflate 6a,
giving 7a in 89% yield after chromatography (Scheme 5). Our
attention next turned to the novel cyclization of 7a to 10
(methylene-substituted indoline assumed). Various Pd/ligand
and Cu/ligand catalysis systems were screened for the cycliza-
tion 7a-10, and our first hit was discovered when we attempted
the C-N bond-forming reaction in DMF.12 This formed the
indoline 10 in 21 LCAP (liquid chromatography area percent).
chlorotriflates were also used (entries 7 and 8), further showing
the scope of this chemistry to prepare functionalized Boc-
protected 3-methylindoles. The three-step sequence to substi-
tuted indoles gave yields ranging from 54% to 80%, which
corresponds to a yield of 81% to 92% over each step in the
sequence. All the examples in Table 1 have been demonstrated
on between 1 and 120 g of starting chlorotriflate, showing that
this chemistry is reliable and scalable. What is also striking is
the efficiency of the Heck reaction, which can now be utilized
as a general method to prepare a range of 2-aryl-substituted
allylcarbamates. Additionally, the intermediate methylene-
substituted indolines are also useful structures in their own right.
Scheme 5. Demonstration of the Indole Synthesis
Table 1. The Preparation of Various 3-Methylindoles
Optimizing this further we found XPhos was the preferred
ligand for the cyclization of 7a to 10, with K2CO3 in DMF at
80 °C.13 These conditions for the cyclization gave 7a-10 in
55% isolated yield after chromatography. The exo-orientation
1
of the double bond in 10 was proven by H NMR analysis,
and the double bond was isomerized efficiently with camphor-
sulfonic acid (CSA)14 to give the Boc-protected 3-methylindole
11. This result successfully demonstrated our original hypothesis
drawn up in Scheme 3. As each reaction was sufficiently clean
the route shown in Scheme 5 could be streamlined; hence at
the end of the Heck reaction a solvent switch to DMF was
performed, followed by addition of the cyclization catalyst,
ligand, and base. At the end of the cyclization reaction the
methylene-substituted indoline could be isomerized by treating
with CSA, which gave the Boc-indole 11.15 This three-step
sequence avoided unnecessary operations and gave 11 in 60%
yield, which corresponds to yields of >84% per step.
We next showed the generality of this process toward the
synthesis of Boc-protected 3-methylindoles (Table 1). A variety
of substitution on the chlorotriflate was tolerated including
electron-withdrawing groups (entries 2 and 8), an electron-
donating group (entry 5), as well as methyl and fluoro
substitution (entries 1, 3, 4, and 7). This method is noteworthy
in the generality to prepare 4-, 5-, 6-, and 7-substituted indoles
with high efficiency and in short order from commercially
available chlorophenols. Additionally, more highly substituted
(7) (a) Olofsson, K.; Sahlin, H.; Larhed, M.; Hallberg, A. J. Org. Chem.
2001, 66, 544–549. (b) For a discussion of mechanism and regioselectivity in
the Heck reaction, see: Cabri, W.; Candiani, I. Acc. Chem. Res. 1995, 28,
2-7.
(8) Jensen, T.; Pedersen, H.; Bang-Andersen, B.; Madsen, R.; Jørgensen,
M. Angew. Chem., Int. Ed. Engl. 2008, 47, 888–890.
(9) See the Supporting Information for the synthesis of the chlorotriflates
from the corresponding chlorophenols.
(10) o-Bromo triflates were found to not be as efficient in the Heck
reaction.
a Reagents and conditions: (i) N-Boc allylamine (1.2 equiv), Pd(OAc)2
(3 mol %), dppf (10 mol %), NaOAc (1.2 equiv), MeCN, 80 °C, 15-24 h.
(ii) Pd(OAc)2 (3 mol %), XPhos (10 mol %), K2CO3 (1.5 equiv), DMF, 80
°C, 15-18 h. (iii) CSA (20 mol %), CH2Cl2, 25 °C, 48 h. b Isolated yield
after column chromatography of the indole product. c K2CO3 (1.5 equiv)
was used as the base in the Heck reaction.
(11) For the full table from the base screen, please see the Supporting
Information.
670
Org. Lett., Vol. 12, No. 4, 2010

Current syntheses to these intermediates involve numerous steps,
which are often nonselective and require functionalized aryl
precursors.16 The outlined method for their preparation signifi-
cantly simplifies the synthesis of these compounds.
The Boc-indoles could be deprotected in a straightforward
manner by treatment with NaOEt in EtOH17 to provide the
free indoles for further elaboration at the N-1 position,
Scheme 6.
In summary, we have demonstrated a novel and scalable
synthesis of 3-methylindoles from chlorotriflates through a
Heck reaction, amide/aryl chloride coupling, and isomeriza-
tion sequence. The 2-chlorophenols are commercially avail-
able, and the three-step sequence to substituted indoles is
highly efficient, enabling the regiocontrolled synthesis of
substituted indoles in short order. Additionally, the Heck
reaction of aryl triflates with protected allylamines has been
optimized and is now efficient for the preparation of
synthetically useful 2-aryl-N-Boc allylamines. The utilization
of these 3-methyl indoles as well as chemistry using the
methylene-substituted indoline intermediates is the focus of
further investigations and will be reported in due course.
Scheme 6. Removal of the Boc Group To Give Free Indoles
Acknowledgment. We wish to thank Simon Hamilton and
David Murray (Analytical Research MSD, Hoddesdon) for
mass spec analysis.
Supporting Information Available: Experimental pro-
cedures and full spectroscopic data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Benzenesulfonyl allylamine was also found to be a viable
allyl source in this reaction, generating the corresponding
N-sulfonate indole with the same efficiency as the Boc
allylamine through the three steps, further demonstrating the
utility of this indole synthesis, Scheme 7. This is of particular
interest as N-sulfonyl indoles are often used in active drug
molecules.18
OL902636V
(12) CuI (10 mol %) with L-proline (20 mol %) also gave a small
amount of the desired methylene-substituted indoline (11%). Attempts
to optimize this proved fruitless with messy reaction profiles, forming
multiple products.
(13) Investigations into a single catalyst system for the Heck reaction
and cyclization have so far proved elusive.
(14) Tietze, L. F.; Buhr, W. Angew. Chem., Int. Ed. Engl. 1995, 34,
1366–1368.
(15) While it was possible to directly charge CSA at the end of the
cyclization reaction, the process benefitted from a quick aqueous workup
prior to addition of CSA.
Scheme 7. Use of Benzenesulfonyl Allylamine
(16) (a) Lin, H.; Kazmaier, U. Eur. J. Org. Chem. 2009, 1221–1227.
(b) Ma, J.; Yin, W.; Zhou, H.; Liao, X.; Cook, J. M. J. Org. Chem. 2009,
74, 264–273. (c) Tercel, M.; Denny, W. A. J. Chem. Soc., Perkin Trans. 1
1998, 509–519. (d) Tidwell, J. H.; Buchwald, S. L. J. Am. Chem. Soc. 1994,
116, 11797–11810. (e) Sakamoto, T.; Kondo, Y.; Uchiyama, M.; Yamanaka,
H. J. Chem. Soc., Perkin Trans. 1 1993, 1941–1942. (f) Boger, D. L.;
Coleman, R. S. J. Am. Chem. Soc. 1988, 110, 4796–4807.
(17) Qiu, X.-L.; Zhu, J.; Wu, G.; Lee, W.-H.; Chamberlain, A. R. J.
Org. Chem. 2009, 74, 2018–2027.
(18) Tsai, Y.; Dukat, M.; Slassi, A.; MacLean, N.; Demchyshyn, L.;
Savage, J. E.; Roth, B. L.; Hufesein, S.; Lee, M.; Glennon, R. A. Bioorg.
Med. Chem. Lett. 2000, 10, 2295–2299.
Org. Lett., Vol. 12, No. 4, 2010
671