Ortho-selectivity in the nucleophilic aromatic substitution (S NAr) reactions of 3-substituted, 2,6-dichloropyridines with alkali metal alkoxides

Article abstract of DOI:10.1016/j.tetlet.2011.06.007

3-Substituted, 2,6-dichloropyridines have featured in the syntheses of small molecule inhibitors of a wide variety of biological targets. Hence, the regioselective displacement of the chlorines is of significant interest. Through conducting an extensive solvent study, we have found that non-polar, aprotic solvents of low hydrogen bond basicities favour substitution of the chlorine ortho to the 3-substituent by alkali metal alkoxides. We present convincing evidence that coordination of the alkali metal counter-ion to the 3-substituent (nitro, ester, amide) is the origin of the ortho-selectivity to give a cyclic, six-membered transition state. Excellent ortho-selectivities (≥98:2) for secondary and tertiary alkoxides were realized with the sodium counter-ion, whereas the more reactive primary alkoxides required the harder, more Lewis acidic lithium counter-ion.

Full text of DOI:10.1016/j.tetlet.2011.06.007

Tetrahedron Letters 52 (2011) 4172–4176
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Ortho-selectivity in the nucleophilic aromatic substitution (SNAr) reactions
of 3-substituted, 2,6-dichloropyridines with alkali metal alkoxides
⇑
Jeremy L. Yap, Kellie Hom, Steven Fletcher
Department of Pharmaceutical Sciences, University of Maryland School of Pharmacy, 20 N Pine Street, Baltimore, MD 21201, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
3-Substituted, 2,6-dichloropyridines have featured in the syntheses of small molecule inhibitors of a wide
variety of biological targets. Hence, the regioselective displacement of the chlorines is of significant inter-
est. Through conducting an extensive solvent study, we have found that non-polar, aprotic solvents of
low hydrogen bond basicities favour substitution of the chlorine ortho to the 3-substituent by alkali metal
alkoxides. We present convincing evidence that coordination of the alkali metal counter-ion to the 3-sub-
stituent (nitro, ester, amide) is the origin of the ortho-selectivity to give a cyclic, six-membered transition
state. Excellent ortho-selectivities (P98:2) for secondary and tertiary alkoxides were realized with the
sodium counter-ion, whereas the more reactive primary alkoxides required the harder, more Lewis acidic
lithium counter-ion.
Received 19 April 2011
Revised 31 May 2011
Accepted 2 June 2011
Available online 12 June 2011
Keywords:
Ortho
Regioselective
Nucleophilic aromatic substitution (SNAr)
2,6-Dichloro-3-nitropyridine
Coordination
Ó 2011 Elsevier Ltd. All rights reserved.
Coordination of Lewis basic groups to alkyllithiums has been
exploited in the ortho-lithiation of arenes bearing a directing meta-
lation group (DMG).1 These intermediates can then be reacted with
a variety of electrophiles in what is overall an adaptation of elec-
trophilic aromatic substitution. Until very recently,2,3 this directing
effect had been somewhat overlooked in the nucleophilic aromatic
substitution (SNAr) reactions of 1-substituted, 2,4-dihaloaromatic
compounds with metal alkoxides. In those publications,2,3 how-
ever, there were no detailed discussions on the Lewis basicities
of the reaction solvent and the electron-withdrawing/directing
group (at the 1-position) and their relative abilities to promote
ortho attack. In the more recent of the two papers, the scope of
alkoxide nucleophile was not investigated. Furthermore, a compre-
hensive study of analogous reactions with the related 3-substi-
tuted, 2,6-dichloropyridines, in which metal cation coordination
to the pyridine nitrogen could compromise the directing effect of
the 3-substituent, has never been reported in the literature. This
is surprising given their use in the development of a large number
of small-molecule inhibitors of a variety of biological targets, as
well in the synthesis of the pharmaceutical compound flupirtine.4
En route to the preparation of a new
a-helix mimetic, com-
Scheme 1. Reagents and conditions: (a) NH3, EtOH; (b) NaNO2, H2SO4, H2O;
(c) Ag2CO3, (CH3)2CHI; (d) Bu3SnCH@CH2, Pd(PPh3)4; (e) (1) O3, AcOH; (2) AgNO3,
KOH; (f) (CH3)2CHOH, NaH, toluene, 0 °C?rt, 16 h, 93%.
pound 6 was required as a particular subunit. Previously, its syn-
thesis has been accomplished via the lengthy fashion a through e
in Scheme 1 in research that led to low micromolar antagonists
of the Bak–Bcl-xL complex.5 This route is further hampered by
the non-trivial regioselective alkylation of the pyridone oxygen,
which requires the use of expensive silver carbonate.6 We consid-
ered if compound 4 could be obtained from 1 in a single step by a
regioselective SNAr reaction. In this Letter, we reveal that non-polar
solvents of low hydrogen bond basicities promote ortho-selective
attack of 3-substituted, 2,6-dichloropyridines by alkali metal alk-
⇑
Corresponding author. Tel.: +1 410 706 6361; fax: +1 410 706 5017.
E-mail address: sfletche@rx.umaryland.edu (S. Fletcher).
0040-4039/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2011.06.007

J. L. Yap et al. / Tetrahedron Letters 52 (2011) 4172–4176
4173
oxides through the formation of a cyclic, six-membered transition
state containing the metal counter-ion, reminiscent of the ortho-
lithiation of arenes. The effects of the solvent, metal counter-ion
and directing group are explored, as is the scope of the alkoxide
nucleophile.
room temperature for all starting material to be consumed, a ratio
of 4a:4b of 98:2 was achieved, with no detectable 4c. Whilst it is
anticipated that sodium isopropoxide is poorly soluble in toluene,
it is envisaged that its solubility would be increased upon coordi-
nation of the sodium cation to the polar nitro group of 1, bringing
the associated nucleophilic isopropoxide anion into close proxim-
ity to the activated ortho-carbon. Thus, attack ortho to the nitro
group would be favoured and would lead to an electrostatically
neutral transition state, which would be favoured in solvents of
low dielectric constant, such as toluene. This concept of ‘built-in
solvation’ was first proposed by Bunnett for a similar system,9
and would explain the ortho regiochemical outcome of this reac-
tion. Confirmation of the ortho-selective attack of isopropoxide
was acquired through 1H and NOESY NMR analysis of 4a and of a
mixture of 4a/4b (Supplementary data).
Since the reaction of 1 with ammonia (and amines) to afford 2 is
known to proceed with excellent ortho-selectivity,7 we considered
that this may be due to the formation of a stabilizing hydrogen
bond in the transition state between the ammonia nucleophile
and the nitro group of the electrophilic aryl chloride. Indeed, it
has recently been shown that solvent hydrogen bond basicity
(SHBB) has a dramatic effect on the regiochemistry of SNAr reac-
tions of electron deficient polyfluoroarenes with secondary amines
through its effect on amine hydrogen bonding in the transition
state.8 2,4-Difluoroacetophenone gives excellent ortho-selectivities
in solvents of poor hydrogen bond basicities owing to the forma-
tion of a hydrogen-bonded, six-membered transition state that is
not accessible upon para attack.8 In turn, we speculated that
ortho-selectivity with metal alkoxides might likewise be effected
through coordination of the metal counter-ion with the nitro group
to form an analogous six-membered transition state, and thus that
we might be able to achieve the transformation of 1 into 4 in a sin-
gle step. In addition to the obvious economical benefits, this would
accelerate the development of small molecules incorporating this
motif.
A variety of solvents was investigated to ascertain their effects
on regioselectivities; the results of our findings are given in Table 1,
and they are presented in order of decreasing ortho-selectivity.
There is a correlation between the regiochemical outcome of the
reaction and both the solvent’s dielectric constant (er) and its Taft
pKHB value.10 The latter parameter can be used as a measurement
of SHBB, which, given that hydrogen bonding is a special class of
the Lewis acid-base interaction, presumably correlates well with
Lewis basicity towards a given metal cation (Lewis acid). Solvents
of lowest er and pKHB values furnished the highest regioselectivi-
ties. Conversely, solvents of highest er and pKHB values gave the
lowest regioselectivities, and also led to greater amounts of bis-
substituted product. Of the solvents studied, toluene afforded the
best ortho-selectivity, whilst DMSO afforded the worst. It is note-
worthy that for CH3CN, DMF and DMSO, all of which have similar
dielectric constants, a clear trend exists: the greater the pKHB, the
more reduced was the amount of ortho product. These results are
consistent with our hypothesis that ortho-selectivity is achieved
The conversion of substrate 1 into product 4 with in situ-gener-
ated sodium isopropoxide was initially conducted in DMF, a dipo-
lar, aprotic solvent that is known to accelerate SNAr reactions, and
in the non-polar, aprotic solvent toluene. For the DMF reaction, all
starting material was consumed within a matter of hours at room
temperature to give an inseparable and approximate 2:1 mixture
of 4a and 4b, as well as 8% of the bis-substituted by-product 4c.
Conversely, for the reaction in toluene, which required 16 h at
Table 1
Solvent and counter-ion effects on the regioselectivity of the reactiona
Yieldd (%)
Ratio 4a:4b:4ce
c
Entry
Solvent
erb
pKHB
Base
1
2
3
4
5
6
7
8
Toluene
Dioxane
CH2Cl2
THF
CH3CN
Isopropanol
DMF
DMSO
Toluene
THF
2.38
2.21
8.93
À0.363
1.033
À0.3f
1.28
0.913
0.82
2.103
2.583
2.38
NaH
NaH
NaH
NaH
93
82
90
82
98:2:0
93:6:1
92:7:1
85:13:2
72:26:2
67:33:0
59:33:8
49:40:11
66:34:0
94:6:0
7.58
35.94
18.23
36.71
46.45
2.38
7.58
7.58
7.58
NaH
50, 91g
57, 98g
70
NaOiPr
NaH
NaH
NaH
LiHMDS
NaHMDS
KHMDS
22
9h
10
11
12
51, 97g
ND
1.28
1.28
1.28
THF
THF
ND
ND
86:14:0
82:9:9
a
Reaction conditions: 2,6-dichloro-3-nitropyridine (1 mmol) and isopropyl alcohol (1.2 mmol) were dissolved in the solvent (10 mL). The reaction was cooled to 0 °C (the
DMSO reaction was performed at room temperature). NaH (1.3 mmol) was added under N2. After 30 min, the reaction flask was removed from the ice bath and stirred for 16 h
at rt.
b
Relative dielectric constant.
A scale for hydrogen bond basicity.
c
d
Isolated yield after purification by silica gel flash column chromatography.
e
Determined by 1H NMR.
f
Value for generic chloroalkane.
Yield based on recovered starting material.
Reaction performed with 1.2 mmol of 15-crown-5; NaH was added last. ND = not determined.
g
h

4174
J. L. Yap et al. / Tetrahedron Letters 52 (2011) 4172–4176
Table 2
through coordination of the sodium cation to the nitro group, since
the lower the solvent’s polarity, the more poorly the sodium iso-
propoxide will be solvated and the lower the SSHB, the weaker will
be its ability to coordinate (solvate) the sodium ions in particular.
To acquire further proof that coordination of the sodium counter-
ion is the ‘key’ to ortho-selectivity, we repeated the reaction in
toluene but this time in the presence of 15-crown-5. Through
comparing entries 1 and 9 in Table 1, it can be seen that the crown
ether significantly reduced the regioselectivity of the SNAr reaction,
affording a similar result as that with dipolar, aprotic solvents like
DMF (entry 7). Finally, we investigated the effect of the metal
counter-ion on the regioselectivity of the reaction. Employing the
lithium, sodium and potassium salts of hexamethyldisilazane
(HMDS) as the base in THF, we observed that the identity of the
metal counter-ion played a notable role (entries 10–12), with the
amount of ortho product decreasing in the order Li+ > Na+ > K+
(although the K+ counter-ion led to a greater amount of bis-substi-
tuted product). This result makes sense, given that Li+ is the hard-
est of all the alkali metal cations and can therefore form the
strongest bonds with an oxygen of the nitro group. A plausible
transition state for the ortho regiochemical outcome of this reac-
tion is shown in Figure 1, and this is consistent with the data pre-
sented in Table 1.
The alkoxide substrate scope for this reaction was next ex-
plored. As can be seen in Table 2, primary, secondary and tertiary
alkoxides all reacted in high yields and with excellent ortho/para
regioselectivities of P93:7. The hindered tert-butoxide anion de-
manded the use of an excess of the nucleophile as well as a reac-
tion temperature of 60 °C; the yield and regioselectivity were not
compromised by these more forcing conditions (entry 8). The high-
est regioselectivities were observed with the sodium salts of sec-
ondary and tertiary alkoxides. Given our findings in Table 1 on
the effect of the metal counter-ion, we were able to boost the
ortho-selective attack of primary alkoxides by invoking the lithium
salt (compare entries 2 and 3). Thus, for complete formation of the
ortho regioisomer, when the alkoxide is primary, the lithium salt
should be used, whilst the sodium salt is suitable for secondary
and tertiary alkoxides.
We next investigated the scope of this regioselective SNAr reac-
tion with 2,6-dichloropyridines carrying different substituents at
the 3-position. To ensure the reaction still proceeded, electron-with-
drawing groups were employed: nitrile, ester and amide, as well as
nitro (data from Table 1). This time, sodium isopropoxide was added
directly rather than being prepared in situ; notably, this led to more
sluggish and incomplete reactions. Thus, we have presented both
yields and yields based on recovered starting material. For each of
the four 2,6-dichloropyridine substrates, the highest ortho-selectiv-
ity was obtained in toluene (Table 3), and the regioselectivity
dropped as the solvent became progressively more polar and more
able to solvate the sodium counter-ion (higher SHBB), consistent
with the data in Table 1. When X was nitro, ester or amide, excellent
ortho-selectivity (P97:3) was observed in all cases in toluene. How-
ever, a notable exception is 2,6-dichloronicotinonitrile, which gave a
substantially reduced regioselectivity in toluene (entry 4) and no
Substrate scope for the reaction of 2,6-dichloro-3-nitropyridine with a variety of
alkali metal alkoxidesa
NO2
NO2
NO2
Cl
OR
Cl
ROH, base
+
N
Cl
N
Cl
N
OR
B
toluene, RT, 16 h
1
A
Entry
Product
Base
NaH
Yieldb (%)
Ratio, A:Bc
NO2
OMe
OEt
OEt
OBn
O
1d,e
81
93:7
N
Cl
NO2
2
NaH
76
91
94:6
99:1
93:7
96:4
98:2
99:1
100:0
N
Cl
NO2
3
LiHMDSf
N
Cl
NO2
4
NaH
68, 97g
N
Cl
NO2
5
NaH
94
N
Cl
NO2
OiPr
6
NaH
93
N
Cl
NO2
O
7
NaH
84
N
Cl
NO2
OtBu
8h
NaH
95
N
Cl
a
Reaction conditions: 2,6-dichloro-3-nitropyridine (1 mmol) and the alcohol
(1.2 mmol) were dissolved in toluene (10 mL). The reaction was cooled to 0 °C (the
DMSO reaction was performed at room temperature). NaH (1.3 mmol) was added
under N2. After 30 min, the reaction flask was removed from the ice bath and stirred
for 16 h at rt.
b
Isolated yield after purification by silica gel flash column chromatography.
c
Determined by 1H NMR.
d
To minimize bis-substitution, sodium methoxide was first generated in situ and
then the pyridine was added.
e
Approximately 7% of bis-substituted by-product was detected.
3 equiv of LiHMDS used.
Yield based on recovered starting material.
Modified reaction conditions: 2.6 equiv tert-butanol, 2.8 equiv NaH, 60 °C,
f
g
h
30 min.
regioselectivity in CH3CN (entry 6). These results were rather sur-
prising at first, since we expected no regioselectivity would be ob-
served in any solvent, given the digonal, sp-hybridized nature of
Figure 1. Built-in solvation of sodium isopropoxide by the nitro group accounts for
the ortho-selectivity of the SNAr reaction in non-polar solvents of poor hydrogen
bond basicities (pKHB), such as toluene.

J. L. Yap et al. / Tetrahedron Letters 52 (2011) 4172–4176
4175
Table 3
Substrate scope for the reaction of sodium isopropoxide with 3-substituted, 2,6-dichloropyridinesa
Entry
2,6-Dichloropyridine
Approximate pKb of X12
Solvent
Yieldb (%)
Ratio, A:Bc
1d
2d
Toluene
Dioxane
CH3CN
93
82
98:2
94:6
74:26
3d
50, 91e
26
4
5
Toluene
Dioxane
69, 93e
95
72:28
66:34
6
CH3CN
92
53:47
24
21
7
8
9
Toluene
Dioxane
CH3CN
53, 83e
67, 82e
98:2
94:6
79:21
78, 85e
10
11
12
Toluene
Dioxane
CH3CN
51, 92e
60, 92e
97:3
95:5
88:12
71, 87e
15
a
Reaction conditions: 2,6-dichloro-3-X-pyridine (0.5 mmol) was dissolved in the solvent (5 mL), then sodium isopropoxide (1.5 mmol) was added. The reaction was stirred
at room temperature for 16 h.
b
Isolated yield after purification by silica gel flash column chromatography.
c
Determined by 1H NMR.
d
Data from Table 1 with sodium isopropoxide generated in situ.
Yield based on recovered starting material.
e
the nitrile functionality which directs the N lone pair away from a
favoured, six-membered transition state. Although less common,
solvent study, experiments with 15-crown-5 and performing the
SNAr reaction on 2,6-dichloropyridines with varying 3-substitu-
ents, we have provided evidence that coordination of the metal
counter-ion to the 3-substituent is the origin of the regioselectivi-
ty. Given that 3-substituted, 2,6-dichloropyridines have featured in
the design of a large number of small-molecule inhibitors, we feel
the work reported herein may be of significant interest.
nitriles are known to engage in ‘side-on’
and metal cations in which the C–N triple bond acts as the Lewis
base,11 in contrast to ‘end-on’
-coordination through the nitrogen
p-coordination with metals
r
lone pair. Side-on coordination would afford a transition state more
akin, but less optimal, to that presented in Figure 1, and would
explain the less pronounced ortho-selectivity.
The stronger a base (lower pKb), the more thermodynamically
favourable is its reaction with an acid. Invoking the broader Lewis
definition of acids and bases, it follows that more basic (electron
pair-donating) functional groups will coordinate more strongly to
a given metal cation (Lewis acid). Excepting the data for 2,6-dichlo-
ronicotinonitrile for the reason just described, there is a clear effect
of the basicity (pKb) of the X group, which correlates well with its
Lewis basicity, on the regiochemical outcome of the reaction in
CH3CN (entries 3, 9 and 12). An increase in ortho product was ob-
served in the order nitro < ester < amide, in line with the increasing
basicities12 of these groups.
In conclusion, we have demonstrated that SNAr reactions of
3-substituted, 2,6-dichloropyridines with alkali metal alkoxides
proceed with excellent ortho-selectivities (ortho with respect to
the 3-substituent). A broad range of alkoxides was examined:
complete, or almost complete (P98:2), ortho-selectivity with
secondary and tertiary alkoxides was realized with the sodium
counter-ion, whilst the more reactive primary alkoxides required
the more Lewis acidic lithium counter-ion. Through conducting a
Acknowledgements
The authors wish to thank the University of Maryland School of
Pharmacy and the American Association of Colleges of Pharmacy
(AACP) for financial support of this work, and Dr. Andrew Coop
for helpful discussions.
Supplementary data
Supplementary data (1H NMR spectrum of 4a, and 1H and
NOESY NMR spectra of a mixture of 4a/4b) associated with this
article can be found, in the online version, at doi:10.1016/
j.tetlet.2011.06.007.
References and notes
1. Schlosser, M. Angew. Chem., Int. Ed. 2005, 44, 376–393.
2. Ouellet, S. G.; Bernardi, A.; Angelaud, R.; O’Shea, P. D. Tetrahedron Lett. 2009, 50,
3776–3779.

4176
J. L. Yap et al. / Tetrahedron Letters 52 (2011) 4172–4176
3. Whilst this work was in progress, a similar manuscript was published: Wendt,
M. D.; Kunzer, A. R. Tetrahedron Lett. 2010, 51, 3041–3044.
7. (a) Choi, Y. M.; Kucharczyk, N.; Duane, R. J. Labelled Compd. Radiopharm. 1987,
24, 1–14; (b) Oguchi, M.; Wada, K.; Honma, H.; Tanaka, A.; Kaneko, T.;
Sakakibara, S.; Ohsumi, J.; Serizawa, N.; Fujiwara, T.; Horikoshi, H.; Fujita, T. J.
Med. Chem. 2000, 43, 3052–3066.
8. Wang, X.; Salaski, E. J.; Berger, D. M.; Powell, D. Org. Lett. 2009, 11, 5662–5664.
9. Bunnett, J. F.; Morath, R. J. J. Am. Chem. Soc. 1955, 77, 5051–5055.
10. (a) Taft, R. W.; Gurka, D.; Joris, L.; Schleyer, P. R.; Rakshys, J. W. J. Am. Chem. Soc.
1969, 91, 4801–4808; (b) Laurence, C.; Brameld, K. A.; Graton, J.; Le Questel, J.-
Y.; Renault, E. J. Med. Chem. 2009, 52, 4073–4086.
11. (a) Solomun, T.; Christmann, K.; Baumgärtel, H. J. Phys. Chem. 1989, 93, 7199–
7208; (b) Chen, L.-Z.; Miller, J. M. Can. J. Chem. 1991, 69, 2002–2007.
12. pKb values determined from entering pKa values (taken from http://
evans.harvard.edu/pdf/evans_pKa_table.pdf) of the protonated X groups into
the equation pKw = pKa + pKb, where pKw = 14.
4. As of April 2011, SciFinder identified more than 5,000 reactions involving 2,6-
dichloro-3-nitropyridine and 2,6-dichloronicotinic acid/2,6-dichloronicotinates.
For example: (a) Yoon, Y. A.; Kim, D. H.; Lee, B. M.; Kim, T. K.; Cha, M. H.; Sim, J.
Y.; Kim, J. G. Bioorg. Med. Chem. Lett. 2010, 20, 5237–5240; (b) Tian, X.; Qin, B.;
Wu, Z.; Wang, X.; Lu, H.; Morris-Natschke, S. L.; Chen, C. H.; Jiang, S.; Lee, K.-H.;
Xie, L. J. Med. Chem. 2010, 53, 8287–8297; (c) Renneberg, D.; Dervan, P. B. J. Am.
Chem. Soc. 2003, 125, 5707–5716; (d) Colbry, Norman L.; Elslager, Edward F.;
Werbel, Leslie M. J. Heterocycl. Chem. 1984, 21, 1521–1525; (e) Seydel, J. K.;
Schaper, K.-J.; Coats, E. A.; Cordes, H. P. J. Med. Chem. 1994, 37, 3016–3022.
5. Ernst, J. T.; Becerril, J.; Park, H. S.; Yin, H.; Hamilton, A. D. Angew. Chem., Int. Ed.
2003, 42, 535–539.
6. Hopkins, G. C.; Jonak, J. P.; Minnemeyer, H. J.; Tieckelmann, H. J. Org. Chem.
1967, 32, 4040–4044.