Studies on the regioselective nucleophilic aromatic substitution (S NAr) reaction of 2-substituted 3,5-dichloropyrazines

Article abstract of DOI:10.1021/ol4006695

Differences in regioselectivity were observed during the S_NAr reaction of amines with unsymmetrical 3,5-dichloropyrazines. This study revealed that when the 2-position of the pyrazine was occupied with an electron-withdrawing group (EWG), nucle

Full text of DOI:10.1021/ol4006695

ORGANIC
LETTERS
2013
Vol. 15, No. 9
2156–2159
Studies on the Regioselective
Nucleophilic Aromatic Substitution
(SNAr) Reaction of 2‑Substituted
3,5-Dichloropyrazines
Stephanie Scales,* Sarah Johnson, Qiyue Hu, Quyen-Quyen Do, Paul Richardson,
Fen Wang, John Braganza, Shijian Ren,† Yadong Wan,† Baojiang Zheng,†
Darius Faizi, and Indrawan McAlpine
La Jolla Laboratories, Pfizer, Inc., 10770 Science Center Drive, San Diego, California
92121, United States, and WuXi App Tec Co., Shanghai, China
stephanie.scales@pfizer.com
Received March 12, 2013
ABSTRACT
Differences in regioselectivity were observed during the SNAr reaction of amines with unsymmetrical 3,5-dichloropyrazines. This study revealed that
when the 2-position of the pyrazine was occupied with an electron-withdrawing group (EWG), nucleophilic attack occurred preferentially at the
5-position. When the 2-position was substituted with an electron-donating group (EDG), nucleophilic attack occurred preferentially at the 3-position.
These results are reported along with a computational rationale for the experimental observations based on the Fukui index at the reacting centers.
Substituted pyrazines containing diverse functional groups
are found in numerous naturally occurring compounds,
many of which have important pharmacological activities.1
Scheme 1. First Attempts at Unsymmetrical Pyrazine Targets
The ability to install amines in a regioselective manner to an
unsymmetrical pyrazine core was needed. Predicting this
positional selectivity would be valuable in designing synthetic
routes and would enable the SAR of a lead series. After
making a small variety of 2-substituted 3,5-dichloropyra-
zines, a contrast in reactivity under standard SNAr conditions
using the same series of amines across different substrates
was observed.
In the initial case (Scheme 1, eq 1), addition of an amine
to 2-methyl-3,5-dichloropyrazine in DMSO with CsF
gave regioselective substitution at the 3-position. A second
† WuXi App Tec Co.
(1) (a) Barlin, G. B. The Chemistry of Heterocyclic Compounds;
Wiley-Interscience: New York, 1982; p 655. (b) Garg, N. K.; Stoltz,
B. M. Tetrahedron Lett. 2005, 46, 2423–2426. (c) Guo, C.; Bhandaru,
S.; Fuchs, P. L. J. Am. Chem. Soc. 1996, 118, 10672–10673. (d) Hirsh,
prompted an investigation into the regioselective SNAr
A. J.; Molino, B. F.; Zhang, J.; Astakhova, N.; Geiss, W. B.; Sargent,
reaction utilizing 3,5-dichloropyrazine-2-carbonitrile un-
der identical conditions gave preferential substitution at
the 5-position (Scheme 1, eq 2).2 These intriguing results
reaction of unsymmetrical pyrazines, and the exploitation
B. J.; Swenson, B. D.; Usyatinsky, A.; Wyle, M. J.; Boucher, R. C.;
Smith, R. T.; Zamurs, A.; Johnson, M. R. J. Med. Chem. 2006, 49, 4098–
115. (e) Wailzer, B.; Klocker, J.; Buchbauer, G.; Ecker, G.; Wolschann,
P. J. Med. Chem. 2001, 44, 2805–13. (f) Walker, J. A., II; Liu, W.; Wise,
D. S.; Drach, J. C.; Townsend, L. B. J. Med. Chem. 1998, 41, 1236–41.
(2) Ninkovic, S., Braganza, J. F., Collins, M. R., Kath, J. C., Li, H.,
Richter, D. T. PCT/IB2009/053389, Nov 2, 2010.
r
10.1021/ol4006695
Published on Web 04/19/2013
2013 American Chemical Society

of a computational method to reliably predict this selectivity
in subsequent reactions.
anion was quenched with the appropriate electrophile
(Scheme 3). Quenching the anion with methyliodide gave
a 45% yield of 2-methyl-3,5-dichloropyrazine. Lower yields
were obtained when iodine was used as the electrophile,
giving a 27% yield of 3,5-dichloro-2-iodopyrazine. The
methyl ester was synthesized using CO2 to form the car-
boxylic acid which was converted directly to the methyl ester
in 49% yield over the two steps.
The first example of a regioslective SNAr reaction with un-
symmetrical pyrazines was reported by Camerino in 1960.3
Nitrogen nucleophilies were added to 2-amino-3,5-dibromo-
pyrazine, and attack occurred preferentially at the 3-position
(Scheme 2). Numerous examples of this regioselective SNAr
reaction have been reported in the literature since.4 Regio-
chemical selectivity was also reported with 6,8-dibromoimi-
dazo[1,2-a]pyrazine where substitution at the 8-bromo posi-
tion occurred preferentially over the 6-bromo position.5 The
opposite regioselectivity was observed during the reaction
with 3,5-dichloropyrazine-2-carbaldehyde and sodium meth-
oxide in which substitution occurred preferentially at the
5-position.6
Scheme 3. Metalation of 2,6-Dichloropyrazine To Form
Unsymmetrical Pyrazines
Scheme 2. Examples of Regioselectivity in the Literature
Synthesis of the remaining starting materials is shown
in Scheme 4. 2,6-Dichloropyrazine was converted to the
carboamide in 35% yield using the Minisci reaction8
followed by dehydration with POCl3 to give 3,5-dichloro-
pyrazine-2-carbonitrile (4) in 66% yield (Scheme 4, eq 1).
2-Methoxy-3,5-dichloropyrazine (5) was synthesized in
three steps beginning with NCS-mediated chlorination of
2-aminopyrazine (Scheme 4, eq 2). Diazotization of the
amino group and subsequent hydrolysis provided the
phenol which was converted to the methyl ether in 5%
yield over three steps.
This survey of the selective SNAr reaction began with a
series of pyrazines that featured both electron donating
and electron withdrawing functionalities. Gram quantities
of 2-substituted 3,5-dichloropyrazines were synthesized by
metalation of commercially available 2,6-dichloropyrazine.7
The symmetrical dichloropyrazine was deprotonated with
lithium tetramethylpiperidide (LiTMP) at ꢀ78 °C and the
Scheme 4. Synthesis of the Cyano- and Methoxydichloropyrazines
(3) (a) Camerino, B.; Palamidessi, G. Gazz. Chim. Ital. 1960, 90,
1807–1814. (b) Camerino, B.; Palamidessi, G. Gaz. Chim. Ital. 1960, 90,
1815–1820.
(4) (a) Bonnet, P. A.; Michel, A.; Laurent, F.; Sablayrolles, C.;
Rechencq, E.; Mani, J. C.; Boucard, M.; Chapat, J. P. J. Med. Chem.
1992, 35, 3353–8. (b) Henderson, A. J.; Hadden, M.; Guo, C.; Douglas,
N.; Decornez, H.; Hellberg, M. R.; Rusinko, A.; McLaughlin, M.;
Sharif, N.; Drace, C.; Patil, R. Bioorg. Med. Chem. Lett. 2010, 20,
1137–40. (c) Jiang, B.; Yang, C. G.; Xiong, W. N.; Wang, J. Bioorg. Med.
Chem. 2001, 9, 1149–54. (d) Whelligan, D. K.; Solanki, S.; Taylor, D.;
Thomson, D. W.; Cheung, K. M.; Boxall, K.; Mas-Droux, C.; Barillari,
C.; Burns, S.; Grummitt, C. G.; Collins, I.; van Montfort, R. L.; Aherne,
G. W.; Bayliss, R.; Hoelder, S. J. Med. Chem. 2010, 53, 7682–98.
(5) (a) Lumma, W. C., Jr.; Randall, W. C.; Cresson, E. L.; Huff, J. R.;
Hartman, R. D.; Lyon, T. F. J. Med. Chem. 1983, 26, 357–63. (b) Vitse,
O.; Laurent, F.; Pocock, T. M.; Benezech, V.; Zanik, L.; Elliott, K. R.;
Subra, G.; Portet, K.; Bompart, J.; Chapat, J. P.; Small, R. C.; Michel,
A.; Bonnet, P. A. Bioorg. Med. Chem. 1999, 7, 1059–65.
(6) Xie, L., Ghosh, M., Maynard, G. US 2003/0220348 A1, November
27, 2003.
(8) (a) Duncton, M. A. J. Med. Chem. Comm. 2011, 2, 1135–1161. (b)
Minisci, F.; Gardini, G. P. Tetrahedron Lett. 1970, 1, 15–16. (c) Minisci,
F.; Bernardi, R.; Bertini, F.; Galli, R.; Perchinummo, M. Tetrahedron
1971, 27, 3575–3580. (d) Minisci, F.; Citterio, A.; Vismara, E. Tetra-
hedron 1985, 41, 4157–4170.
(7) (a) Sloss, M., Mckenna, J., Yoon, W. H., Norris, S., Robinson,
D., Parnes, J., Shevlin, G. PCT/US2009/000123, 2009. (b) Turck, A.;
Trohay, D.; Mojovic, L.; Ple, N.; Queguiner, G. J. Organomet. Chem. 1991,
412, 301–310.
Org. Lett., Vol. 15, No. 9, 2013
2157

Withthe pyrazinesinhand, weselecteda seriesof amines
with varying degrees of steric bulk and nucleophilicity to
investigate the regioselective SNAr reaction. The amines
used for this study were isobutylamine, morpholine, ben-
zylamine, diethylamine, and trifluoroethylamine. General
SNAr reaction conditions were used with a 1:1 stoichio-
metric ratio of pyrazine to amine, 3 equiv of CsF, and
0.3 M concentration in DMSO. Reactions were initially
carried out at room temperature and monitored by LCMS
after 2 h. If no desired product or minimal amount of product
(<10%) had been formed, the reactions were heated to 75 °C
until all starting materials were consumed. For the pyrazines
bearing electron-withdrawing substituents (CN, CO2Me, I),
reactions were usually complete within 2ꢀ5 h at room
temperature, while heat was required for reactions with
electron-donating functionalities (OMe, Me).
examples shown, the nucleophilicity and steric bulk of the
amine played a minimal role in the selectivity, while pyrazine
substitution had a significant effect on regiochemical out-
come of the reaction. When the 2-position was substituted
with an electron-donating group, nucleophilic attack oc-
curred preferentially at the 3-position. The methoxy group
provided the greatest selectivity with none of the minor
regioisomer being observed for any of the amines studies,
whereas reactions with 2-methyl-3,5-dichloropyrazine typi-
cally gave 8ꢀ14% of the minor isomer. When the 2-position
of the pyrazine was occupied with an electron-withdrawing
group, nucleophilic attack occurred preferentially at the
5-position. The cyano and methyl ester functionalities gave
similar selectivities in which 3ꢀ8% of the minor isomer was
formed in most cases. Reactions with 3,5-dichloro-2-iodo-
pyrazine were less selective and favored the 3-position despite
the electron-withdrawing character usually associated with
an iodine substituent. This may be explained by the increased
electron density of the pyrazine ring through resonance
effects of the iodo lone pair electrons.
Table 1. Results of the SNAr Reaction with Unsymmetrical
Pyrazines
With these results, we considered possible explanations for
the observed selectivities and sought to identify a computa-
tional method for predicting positional selectivity of the
reaction. Atomic charge distribution, solvent accessible sur-
face area, sterics, the relief of A strain and Fukui function
were considered as possible factors for the observed results.
Calculated CM1A atomic charge9 and solvent accessible
surface area10 values showed little differentiation between
the C3 and C5 positions of the pyrazine as shown in Figure 1.
entry
X
RNH2
product A (%) product B (%)
1
CH3
CH3
i-BuNH2
morpholine
BuNH2
6A (62)
7A (63)
8A (46)
9A (59)
þþ
6B (10a)
7B (14)
8B (8a)
2
3
CH3
b
4
CH3
Et2NH
ꢀ
5
CH3
CF3CH2NH2
i-BuNH2
morpholine
BuNH2
þþ
6
OCH3
OCH3
OCH3
OCH3
OCH3
CO2Me
CO2Me
CO2Me
CO2Me
10A (75)
11A (82)
12A (85)
13A (46)
14A (46)
15A (8)
16A (7)
17A (5)
18A (3)
19A (3)
ꢀ
ꢀ
7
ꢀ
8
ꢀ
9
Et2NH
ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
CF3CH2NH2
i-BuNH2
morpholine
BuNH2
ꢀ
15B (6)
16B (65)
17B (74)
18B (59)
19B (53)
20B (75)
21B (69)
22B (50)
23B (56)
24B (62)
25B (20)
26B (19)
27B (21)
28B (20)
29B (11)
Et2NH
CO2Me CF3CH2NH2
CN
CN
CN
CN
CN
I
i-BuNH2
morpholine
BuNH2
21A (3)
22A (2)
ꢀ
Et2NH
CF3CH2NH2
i-BuNH2
morpholine
BuNH2
ꢀ
25A (60)
26A (59)
27A (54)
28A (51)
29A (42)
I
I
I
Et2NH
I
CF3CH2NH2
a Ratiodeterminedby1HNMR. b ꢀ no product detected. þþ no product
Figure 1. Calculated atom-based Fukui indices.
isolated.
Given the A-values are ranked CH3 (1.7) > CO2CH3
(1.27) > OCH3 (0.6) > I (0.43) > CN (0.17), this would
All reactions were purified by silica gel column chroma-
tography, and the regiochemistry of the products con-
firmed by NOE NMR experiments, or in some cases
X-ray crystal structures.
The experiments revealed clear trends for selectivity of
the unsymmetrical pyrazines shown in Table 1. In the
(9) CM1A atomic charge: The input molecules were first minimized
using OPLS2005 FF with water as solvent in MacroModel and then
further geometry optimized at the AM1 level. The CM1A (AM1þCM1)
atomic charges were calculated using the AMSOL program.
2158
Org. Lett., Vol. 15, No. 9, 2013

predict that methyl and ester substituents would provide
greater selectivity at the 3-position over methoxy, iodo,
and cyano, and as such, A strain values did not correlate
with the observed experimental trends. On the other hand,
Fukui indices did show differentiation between the C3 and
C5 positions of the 3,5-dichloropyrazines and correlated
well with the observed experimental trends (Figure 1). The
Fukui function11 is a computational model that has been
used to predict regioselectivity in organic reactions involv-
ing nucleophiles and electrophiles such as dipole addition
reactions12 and Michael reactions.13 The Fukui function
represents the changes in the molecular electron density
upon addition or removal of charge. For a given reference
atom in a series of related molecules, one can rank them
according to the values of a particular Fukui index for that
atom. High positive values (fþ) indicate reactivity toward
nucleophilic attack while high negative values (fꢀ) indicate
reactivity toward electrophilic attack. Atom-based Fukui
indices were calculated for the five unsymmetrical pyrazines
in Figure 1 using Jaguar from Schrodinger14 and were in good
agreement with the observed experimental regioselectivites.
The preferred site of reactivity is rendered as a solid gray
ball. The calculated Fukui indices (fþ) agree with our
experimental results where the larger (fþ) is the preferred
site for nucleophilic attack. In addition, there is little
differentiation between the calculated C3 and C5 values
(fþ) for the iodo pyrazine, which correlates well with our
observations that iodo pyrazines showed reduced selectiv-
ities in the SNAr reaction.
In conclusion, we have shown that the addition of
nitrogen nucleophilies to unsymmetrical 3,5-dichloropyr-
azine is both regioselective and occurs in modest chemical
yields. Clear trends exist in which electron withdrawing
groups direct substitution to the 5-chloro position of the
pyrazinewhileelectrondonating groupsdirectsubstitution
to the 3-chloro position. In addition, calculated Fukui
indices for the unsymmetrical 3,5-dichloropyrazines cor-
relate well with experimental regioselectivities and can be
used to reliably predict the preferred site of reactivity.
Acknowledgment. This work was supported by senior
management at Pfizer La Jolla. Wegratefully acknowledge
their support to conduct this work. We thank Prof. Arnold
Rheingold and Dr. Curtis Moore from the UCSD Crystal-
lography Laboratory for providing crystal structures. We
also thank our Pfizer La Jolla colleagues Jason Ewanicki
for NMR support and Alex Yanovsky for crystallography
support.
(10) (a) Solvent-accessible surface area: The input molecules were
first minimized using OPLS2005 FF with water as solvent in Macro-
˚
Model with probe radius 1.4 A and the rest of the parameters as default
from MSMS: (b) Sanner, M. F.; Spehner, J. C.; Olson, A. J. Biopolymers
1996, 38, 305–320.
(11) (a) Fukui, K. Science 1982, 218, 747–54. (b) Parr, R. G.; Yang,
W. J. Am. Chem. Soc. 1984, 106, 4049–4050.
(12) Madjarova, G.; Tadjer, A.; Cholakova, T. P.; Dobrev, A. A.;
Mineva, T. J. Phys. Chem. A 2005, 109, 387–93.
ꢀ
(13) Domingo, L. R.; Chamorro, E.; Perez, P. Eur. J. Org. Chem.
2009, 3036–3044.
Supporting Information Available. Experimental pro-
cedures and spectroscopic data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(14) The Fukui index is calculated in the following way: The input
molecules were first minimized using OPLS2005 FF with water as
solvent in MacroModel. The minimized conformations were then
geometry optimized using B3LYP/6-31G** basis set. The Fukui indices
were then calculated based on the optimized geometry at the same level.
All QM calculations were performed using Jaguar from Schrodinger.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 9, 2013
2159