Selective synthesis of 1 -functionalized-alkyl-1H-indazoles

Article abstract of DOI:10.1021/ol902050m

An efficient method for the selective "N1" alkylation of indazoles is described. Use of a-halo esters, lactones, ketones, amides, and bromoacetonitrile provides good to excellent yield of the desired N1 products.

Full text of DOI:10.1021/ol902050m

ORGANIC
LETTERS
2009
Vol. 11, No. 21
5054-5057
Selective Synthesis of
1-Functionalized-alkyl-1H-indazoles§
Kevin W. Hunt,* David A. Moreno, Nicole Suiter, Christopher T. Clark, and
Ganghyeok Kim
Array BioPharma Inc., 3200 Walnut Street, Boulder, Colorado 80301
keVin.hunt@arraybiopharma.com
Received September 3, 2009
ABSTRACT
An efficient method for the selective “N1” alkylation of indazoles is described. Use of r-halo esters, lactones, ketones, amides, and
bromoacetonitrile provides good to excellent yield of the desired N1 products.
The indazole ring is a commonly utilized template in the
discovery of pharmacologically active compounds. The
ability to manipulate the substitution pattern is therefore
crucial to successful drug research. Classically used as a
mimic for the more electron-rich indole moiety, the indazole
has come into fashion as an effective hinge binding motif in
the kinase inhibitor literature.1 In terms of substitution
Figure 1. Indazole ring system and alkylation products.
reactions, the indazole ring (Figure 1) possesses two nitrogens
capable of reacting with electrophiles. In the absence of base,
the N2 product is often the major, if not exclusive, product.2
The addition of aqueous base, heating the reaction, and
altering the nature of the electrophile are all published
strategies to provide a greater proportion of the N1 product.3
However, the selective alkylation of the N1 position remains
a challenge, specifically for C3-unsubstituted indazoles.4
Presented herein is a method for selective N1 alkylation of
indazoles5 utilizing R-halo esters, ketones, lactones, amides,
and bromoacetonitrile.
The initial alkylations of commercially available 6-ni-
troindazole with common bromides provided an expected
mixture of N1 and N2 alkylation products (Table 1, entries
1-3). However, the use of methyl bromoacetate furnished
the N1 product in good yield and excellent selectivity (entry
4). A review of the literature offered no explanation for this
result despite a previous report of the alkylation of 6-ni-
troindazole with ethyl bromoacetate.6 Since the addition of
an electron-withdrawing ester provided high N1 selectivity,
§ Dedicated to Professor Paul A. Grieco on the occasion of his 65th
birthday.
(1) A SciFinder search of the indazole core structure returned 57,227
individual hits. A text search of indazole returned 4,364 references.
(2) Cheung, M.; Boloor, A.; Stafford, J. A. J. Org. Chem. 2003, 68,
4093.
(5) For brevity, we refer to the “1-alkyl-1H” and “2-alkyl-2H” products
as the “N1” and “N2” products, respectively.
(3) (a) Von Auwers, K.; Allardt, H.-G. Ber. 1924, 57B, 1098. (b) Jaffari,
G. A.; Nunn, A. J. J. Chem. Soc. Perkin Trans. 1 1973, 2371. (c) Morel,
S.; Boyer, G.; Coullet, F.; Galy, J.-P. Synth. Commun. 1996, 26, 2443. (d)
Boyer, G.; Galy, J.-P.; Barbe, J. Heterocycles 1995, 41, 487.
(4) For a selective N1 substitution via a reversible 1,4 addition, see:
Saenz, J.; Mitchell, M.; Bahmanyar, S.; Stankovic, N.; Perry, M.; Craig-
Woods, B.; Kline, B.; Yu, S.; Albizati, K. Org. Process Res. DeV. 2007,
11, 30–38.
(6) (a) The high yield of the isolated product implies good selectivity:
Muri, E. M. F.; Mishra, H.; Avery, M. A.; Williamson, J. S. Synth. Commun.
2003, 33, 1977. (b) The reaction of methyl bromoacetate with 6-nitroindazole
was reported in 34% yield: Han, Q.; Dominguez, C.; Stouten, P. F. W.;
Park, J. M.; Duffy, D. E.; Galemmo, R. A., Jr.; Rossi, K. A.; Alexander,
R. S.; Smallwood, A. M.; Wong, P. C.; Wright, M. M.; Leuttgen, J. M.;
Knabb, R. M.; Wexler, R. R. J. Med. Chem. 2000, 43, 4398.
10.1021/ol902050m CCC: $40.75
Published on Web 10/05/2009
 2009 American Chemical Society

If 6-nitroindazole is treated with 1 equiv each of methyl
bromoacetate and cesium carbonate, the ratio of N1:N2 is
3:1 after 1 h and remains unchanged after 16 h. However,
treatment with an additional 0.5 equiv of methyl bromoac-
etate and cesium carbonate rescues the selectivity, providing
>130:1 ratio of the N1:N2 products after 2 h. These results
suggest thermodynamic control for the alkylation with methyl
bromoacetate, in contrast to the other electrophiles that lead
to kinetic product distribution.
Table 1. Electron-Withdrawing Group Improves N1 Selectivitya
The effect of the leaving group was next examined.
Employing the commercially available ethyl esters of chloro-,
bromo-, and iodoacetate, the leaving group showed minimal
effect. All three reagents worked effectively, providing good
to excellent N1 selectivity. Although ethyl and methyl
bromoacetate are employed in most of this work, ethyl
chloroacetate works equally well, is a more potent reagent
(mol/g), and is a weaker lachrymator.
Seeing little effect of the halide leaving group, a series of
common solvents were screened for the conversion of 1a to
5a. DMF, DMA, acetone, and THF provided the best
conversion (75-100%), N1 selectivity (63:1 to >100:1), and
yield (78-85%). In contrast, dichloromethane, p-dioxane,7
and toluene as solvent resulted in poor N1 selectivity (<3:
1). Overall, DMF provided the best results across a range of
electrophiles and indazoles.
a Reaction conditions: 0.3 M, 2 equiv R-Br, 3 equiv Cs2CO3.
whereas the inductively withdrawing methoxy group did not
(entry 3), a series of variables were investigated to determine
the source of selectivity.
First, the reaction of 6-nitroindazole with methyl bro-
moacetate was followed by HPLC to determine overall
conversion of starting material and N1 versus N2 selectivity.
Remarkably, the selectivity for alkylation of 6-nitroindazole
with methyl bromoacetate greatly increased over time. Table
2 shows that 6-nitroindazole is converted to a mixture of
Having found DMF as an optimal solvent, the importance
of the base was explored (Table 3). Good to moderate
Table 3. Effect of Base on Conversion and Selectivitya
Table 2. Selectivity versus Timea
entry
base
Cs2CO3
K2CO3
DIEA
DBU
proton sponge
NaH
conversion (%) a:b ratio yield of 5a (%)
1
2
3
4
5
6
7
100
100
100
81
131:1
92:1
2.2:1
9:1
19:1
1.9:1
63:1
87
80
ND
ND
44
entry
time (h)
5a:5b ratio
conversion (%)
68
1
2
3
4
5
0.5
3
6
16
48
3:1
9:1
25:1
131:1
135:1
90
98
99
100
100
99
77
ND
61
CsOH
a Reaction conditions: 0.3 M, 2 equiv R-Br, 3 equiv base.
a Reaction conditions: 0.3 M, 2 equiv R-Br, 3 equiv Cs2CO3.
selectivity was observed with cesium carbonate, potassium
carbonate, cesium hydroxide, DBU, and proton sponge.8 The
wide range in N1 selectivity from <2:1 (sodium hydride) to
131:1 (cesium carbonate) was unexpected. In addition to the
importance of base for N1 selectivity and overall yield, a
few other observations deserve comment. First, the concen-
N1 and N2 products rapidly at rt. However, the N1:N2 ratio
gradually increased from 3:1 at 0.5 h to >130:1 at 16 and
48 h without substantial change in yield. Given the similarity
of the 16 and 48 h time points, all further experiments were
evaluated at 16 h. In reactions of 6-nitroindazole with allyl
bromide, benzyl bromide, or 1-bromo-2-methoxyethane, no
increase of the N1:N2 ratio was seen with prolonged reaction
times.
(7) The poor selectivity with dioxane (∼1:1) allowed for isolation of
the pure N2 product for mechanistic studies.
(8) Proton sponge: 1,8-bis(dimethylamino)naphthalene.
Org. Lett., Vol. 11, No. 21, 2009
5055

tration of the reaction mixture had a measurable effect on
some of the less selective bases (all reactions herein were
0.3 M relative to the indazole). For example, the N1
selectivity with sodium hydride varied nonreproducibly from
∼1:1 to as high as 8:1 with slight variations in concentration
(0.1 to 0.4 M). Second, the particle size and grade of the
carbonate bases proved important for reproducibility across
different reaction scales (10 mg to multikilogram). Dry, fine
mesh cesium or potassium carbonate reliably provided N1
products selectively and in good to excellent yield, whereas
granulated forms were less reproducible. Decreasing the
equivalents of base from 3 to 2 was tolerated in many cases,
as was lowering the halo acetate equivalents to 1.5. The effect
of decreasing the reagent equivalents was empirically
determined (often necessary for large-scale reactions and
multifunctional indazoles) and studied on a case by case
basis, but the higher reagent loading consistently provided
good yields and selectivity. Finally, pouring the reaction
mixtures directly into a rapidly stirred solution of 2% acetic
acid in water often yielded a workable solid, eliminating a
traditional workup and/or column purification.
appreciable loss in selectivity is observed with simple
changes at the alcohol portion of the ester (entries 1-3).
Substitution in the R position is tolerated, although with loss
in selectivity (entries 4 and 5). The R-bromo amides (entries
6 and 7) and bromoacetonitrile serve as selective electrophiles
providing moderate yields of the N1 product. Chloroacetone
is an extremely selective and effective electrophile, providing
95% of the isolated N1 product (entry 12). Bromoacetophe-
none and methyl bromoacetoacetate (entries 9 and 11) were
selective electrophiles, although product purification proved
difficult as a result of side reactions (aldol reactions, multiple
alkylation, etc.).9 Reducing the loading of base and electro-
phile did not, unfortunately, allow for clean isolation of the
desired products. In general, products containing readily
enolizable ketones were difficult to purify (substitution of
the aryl group did not improve the purification difficulties).
In contrast, the product of bromomalonate was readily
isolated, providing moderate selectivity and yield of the
desired product (entry 10).
The final factors investigated were the substitution and
electron density effects of the indazole substrates (Table 5).
The 6- and 5-nitrated indazoles provide a similar high yield
of selectivity, in sharp contrast to 7-nitroindazole10 (entries
1-3). The peri-interaction between the 7-nitro and the N1
substituent is the likely factor disfavoring the N1 selectivity.
The more electron-rich 5-methoxy and unsubstituted inda-
zoles (entries 4 and 5) provide the corresponding N1 products
in good yield with high selectivity. The Boc protected
indazole anilines (entries 6 and 7) alkylated with modest
selectivity at the N1 position only if the number of
equivalents of cesium carbonate was lowered to 1.1 (use of
3 equiv led to N-alkylation of the carbamate). The bromi-
nated indazole and the 5-ethoxycarbonylindazole were
moderately selective for N1, still providing a good yield of
the isolated N1 products. Finally, introducing a C-3 sub-
stituent predictably increases the N1 selectivity (>200:1,
entry 11).
To expand the synthetic utility of this reaction, other
electron-withdrawing groups were examined (Table 4). No
Table 4. R-Halo Ester, Amide, Nitrile, and Ketone
Electrophilesa
The observed equilibration between the N1 and N2
products is precedented by acylated indazoles,11 indazole
nucleosides,12 and N-pyridinium indazoles.13 Further, the
equilibration of the N2 to N1 product is consistent with
greater calculated (>4.1 kCal/mol)14 and measured15 stability
of 1H-indazole versus 2H-indazole. Since the N2 product
equilibration is unique to the R-electron-withdrawing group
halides (see Table 1), the mechanism was not immediately
obvious.
(9) Under solvent-free conditions, microwave radiation cleanly provides
the N1 product of indazole and 2-bromo-1-(4-bromophenyl)ethanone: Pe´rez,
E.; Sotelo, E.; Loupy, A.; Mocelo, R.; Suarez, M.; Pe´rez, R.; Autie´, M.
Heterocycles 1996, 43, 539.
(10) Atwell, G. J.; Sykes, B. M.; O’Conner, C. J.; Denny, W. A. J. Med.
Chem. 1994, 37, 371.
(11) (a) Von Auwers, K.; Ernecke, A.; Wolter, E. Ann. 1930, 478, 154.
(b) Yamazaki, T.; Baum, G.; Shechter, H. Tetrahedron Lett. 1974, 15, 4421.
(12) Boryski, J. Pol. J. Chem. 1999, 73, 1019.
(13) Isin, E. M.; de Jonge, M.; Castagnoli, N., Jr. J. Org. Chem. 2001,
66, 4220.
(14) Catalan, J.; de Paz, J. L. G.; Elguero, J. J. Chem. Soc. Perkin Trans.
2 1996, 57.
a Reaction conditions: 0.3 M, 2 equiv R-X, 3 equiv Cs2CO3.
(15) Elguero, J. J. Phys. Chem. 1994, 98, 10606.
(16) See Supporting Information.
5056
Org. Lett., Vol. 11, No. 21, 2009

selectivity points to the formation of a ꢀ-indazole succinate
capable of elimination to the unalkylated indazole starting
material (Figure 2). This ꢀ-elimination step is similar to the
Table 5. Indazole Variationa
Figure 2. C-Alkylation of N2 ester.
mechanism invoked for the equilibration of N2 to N1
indazole 1,4-addition products in the reaction of an advanced
indazole intermediate with ethyl acrylate or ethyl 3-bro-
mopropanoate.4
In summary, we disclose a convenient, reliable method
for the selective N1 alkylation of indazoles with appropriate
electrophiles. R-Halo esters, lactones, ketones, amides, and
bromoacetonitrile provide good to excellent yields of the
desired N1 products. The reaction proceeds equally well with
both electron-rich and -deficient indazoles, except in the case
when a C7-N1 peri-interaction is introduced. The unique
N1 selectivity of this reaction is likely due to the N2 to N1
equilibration via a ꢀ-indazole succinate (or equivalent in the
amide, nitrile, and ketone examples).
Acknowledgment. We thank Andrew Allen (Array Bio-
Pharma Inc.) for NMR analysis.
Supporting Information Available: General synthetic
1
procedure, mechanism support, H NMR of N1 products,
nuclear Overhouser enhancement correlation, and total
correlation spectroscopy (TOCSY) experiment. This ma-
terial is available free of charge via the Internet at
http://pubs.acs.org.
a Reaction conditions: 0.3 M, 2 equiv R2-X, 3 equiv Cs2CO3. b Reaction
conditions for entries 6 and 7: 0.3 M, 2 equiv R2-X, 1.1 equiv Cs2CO3.
The necessity of excess alkylating reagent (or the addition
of the likely byproduct diethyl fumarate)16 for greater
OL902050M
Org. Lett., Vol. 11, No. 21, 2009
5057