C. Guo et al. / Tetrahedron Letters 52 (2011) 1692–1696
1693
Table 1
Model system for evaluation of reaction conditions
H
H
N
O
N
O
N
N
Cl
N
H
N
N
N
HN
O
SNAr
HN
N
N
HN
N
N
+
+
HN
N
HN
conditions
N
N
H2N
O
7
5a
6A
acylated product 7'
N
Conditions
Basic
Reaction conditions
Et3N/DMF, wave, 180 °C
Et3N/iPrOH, wave, 140 °C
Et3N/NMP, wave, 200 °C
TMG/DMF, 50–100 °C
NH4Cl, 25–100 °C
Observations
l
l
l
Decomp., no product
Decomp., trace of product
Decomp., no product
Decomp., no product
Decomp., no product
Decomp., no product
Decomp., no product
Decomp., no product
No solvent, 140–180 °C
Silver salt
AgBF4/NMP,
lwave, 150 °C
Ag2CO3/NMP,
l
wave, 100 °C
Pd
Acidic
Xantphos/Pd(OAc)2/Cs2CO3/dioxane/ 80 °C
HCl/dioxane/iPrOH, 140 °C
1:1 HOAc/dioxane, lwave, 200°C
3 equiv HOAc/dioxane, lwave, 200 °C
TFA, 100 °C
Undesired regioisomer-addition to pyrazole ring N
Decomp. trace of product
Decomp. trace of product
70 + ꢀ5% product
70 + trace product
Me3CCO2H, neat, 100–160 °C
1:1 HOAc/H2O, 100 °C
70 + trace product
ꢀ50% (33%) product + trace 70
trace of product 7 and significant decomposition of the starting
materials. Changing the solvent to DMF or NMP led only to decom-
position, and changing the base strength from strong (tetramethyl-
guanidine, TMG) to weak (aminopyrazole 6A alone) to buffered
(ammonium chloride) had no effect. Likewise, varying the reaction
temperature and heat source (50–180 °C, microwave vs thermal
heating) provided no improvement. Addition of silver salts (to acti-
vate the chloropyrimidine) only led to more decomposition.
Slightly better results were obtained using acidic conditions. Heat-
ing in the presence of excess HCl, TFA, or acetic acid in the organic
solvent (or acid itself as solvent) gave trace amounts of the desired
product 7. Attempts to drive the reaction to completion by raising
the temperature led to decomposition or, when using TFA or acetic
acid, acylation of the aminopyrrolopyrazole. Since some of the de-
sired product 7 was formed in the acetic acid reactions, they were
the focus of further optimization. To avoid the acyl aminopyrro-
lopyrazole side product 70, reaction conditions were changed to
suppress the acylation. Reducing the concentration of acetic acid
(from co-solvent down to only 3 equiv) led to a slight increase in
product yield (ꢀ5%). Replacing acetic acid with the more sterically
hindered pivalic acid (Me3CCO2H) did not prevent acylation. We
hypothesized that an ethylidyneoxonium intermediate (formed
from acetic acid under elevated temperature) might be the precur-
sor to the acyl side products 70, and that water might be able to
scavenge that reactive species. Indeed, substituting water for
dioxane as a co-solvent minimized the formation of the side prod-
uct, affording an improved yield of the desired product 7 (50% by
LC–MS, 33% isolated yield after preparative HPLC purification). In
contrast, palladium-catalyzed coupling afforded regioisomeric
products resulting from addition to the pyrazole ring nitrogens
rather than to the exocyclic primary amine. Shen et al.5 report sig-
nificant variation in the regioselectivity of Buchwald-type reac-
tions of aminopyrazoles depending on the combination of a
catalyst, ligand, and base. Many combinations were selective for
addition to the ring nitrogens.
The substrates of interest here are not very reactive so the
addition step may be the rate-determining step. Basic conditions
do not activate the chloropyrimidine (3), which by itself is not
electrophilic enough to undergo the addition of the weakly
nucleophilic aminopyrazole (4). A silver cation may activate aryl-
halide (3) via its strong coordination with the halide, but causes
decomposition of the aminopyrazole (4) at the elevated temper-
ature required for the reaction. Weak acids, such as acetic acid
(pKa = 4.75), are especially effective in facilitating the addition
step between these two partners: the chloropyrimidine is acti-
vated by protonation (3–30), while a substantial fraction of the
weakly basic aminopyrazole (estimate pKa of 4 is ꢀ4.8)15 re-
mains as the more reactive free base. In contrast, stronger acids
such as hydrochloric acid or TFA fully protonate the aminopyrz-
ole, and, therefore, hinder the addition step. When hydrochloric
acid solution is used as the reaction solvent, the excess nucleo-
philic anions may accelerate the acid-catalyzed decomposition
of the substrate, particularly at elevated temperatures. Aqueous
acetic acid appears to be an optimal solvent to promote the
SNAr reaction between the chloropyrimidine and the amino-
pyrazole.
Table 2 illustrates the broad scope of the SNAr reaction in aque-
ous acetic acid. A number of diverse pyrimidine analogs (7–30)
were successfully prepared from aminopyrrolopyrazole intermedi-
ates 6A–C.1 Most of the 4-chloropyrimidine reagents (5a–s) re-
acted to give a moderate yield of product after isolation by
preparative HPLC. The actual conversion is likely better, but the
product molecules are highly soluble in water, so significant loss
of the product could occur during work-up. The SNAr conditions
are mild enough that sensitive functional groups, such as cyano
(h, i), heteroarylfluoride (e, j, k, f) and benzylchloride (r, s) survive
the reaction—no hydrolysis products were observed. At the same
time, the method is effective at activating the less electrophilic
chloropyrimidines to SNAr displacement. For example, pyrimidines
e and j are deactivated by alkyoxy groups, but afford yields of the
desired products comparable to b, k, or o, which lack the alkoxy
groups.
The SNAr reaction mechanism consists of two steps, an addition
followed by an elimination as shown in Scheme 2.