Compatibility of Various Carbanion Nucleophiles with Heteroaromatic Nucleophilic Substitution by the SRN1 Mechanism

Article abstract of DOI:10.1021/jo962353f

Carbanions generated from 2,4,4-trimethyl-2-oxazoline (1a), 2-benzyl-4,4-dimethyl-2-oxazoline (1b), 2,4-dimethylthiazole (13a), 2-benzyl-4-methylthiazole (13b), N,N-dimethylacetamide (17a), tert-butyl acetate (17b), ethyl phenylacetate (17c), N-methyl-N-phenyl-2-butenamide (22), tert-butyl 3-butenoate (25), and dimethyl methylphosphonate (29a) by means of KNH₂ in liquid NH₃ all reacted with 2-bromopyridine (2) via photoassisted reactions that exhibited characteristics of the S_RN1 mechanism. Similar results were obtained in reactions of these carbanions with other substrates, including 2-chloroquinoline (6), iodobenzene (9), bromobenzene (10), and bromomesitylene (11).

Full text of DOI:10.1021/jo962353f

6152
J . Org. Chem. 1997, 62, 6152-6159
Com p a tibility of Va r iou s Ca r ba n ion Nu cleop h iles w ith
Heter oa r om a tic Nu cleop h ilic Su bstitu tion by th e SRN1 Mech a n ism
J im-Wah Wong, Kenneth J . Natalie, J r., Godson C. Nwokogu, J yothi S. Pisipati,
Patrick T. Flaherty, Thomas D. Greenwood, and J ames F. Wolfe*
Department of Chemistry and the Harvey W. Peters Research Center for the Study of Parkinson’s Disease
and Disorders of the Central Nervous System, Virginia Polytechnic Institute and State University,
Blacksburg, Virginia 24061-0212
Received December 16, 1996 (Revised Manuscript Received May 23, 1997X
)
Carbanions generated from 2,4,4-trimethyl-2-oxazoline (1a ), 2-benzyl-4,4-dimethyl-2-oxazoline (1b),
2,4-dimethylthiazole (13a ), 2-benzyl-4-methylthiazole (13b), N,N-dimethylacetamide (17a ), tert-
butyl acetate (17b), ethyl phenylacetate (17c), N-methyl-N-phenyl-2-butenamide (22), tert-butyl
3-butenoate (25), and dimethyl methylphosphonate (29a ) by means of KNH2 in liquid NH3 all reacted
with 2-bromopyridine (2) via photoassisted reactions that exhibited characteristics of the SRN1
mechanism. Similar results were obtained in reactions of these carbanions with other substrates,
including 2-chloroquinoline (6), iodobenzene (9), bromobenzene (10), and bromomesitylene (11).
Sch em e 1
In connection with our continuing interest in the scope
and synthetic utility of heteroaromatic SRN1 reactions1,2
(Scheme 1), we undertook a study of the participation of
several synthetically useful but previously untested
classes of carbanion nucleophiles as reactants in such
radical chain processes involving π-deficient halohetero-
cycles. We were especially interested in nucleophiles in
which the carbanion-stabilizing functional groups were
readily amenable to synthetic transformations following
attachment to the heterocylic substrate. Therefore, the
nucleophiles chosen for study included carbanions de-
rived from lateral deprotonation of the synthetically
versatile 2,4,4-trialkyl-2-oxazolines 1a ,b3 and 2,4-dialkyl-
thiazoles 13a ,b,4 as well as certain carboxamide and ester
enolates and dienolates, along with the carbanion of
dimethyl methylphosphonate (29a ).
gest type of base available in liquid NH3. The necessity
for photoassistance in successful substitution reactions
was tested by attempting analogous reactions in the dark,
while the radical chain nature of productive reactions was
probed using the radical scavenger, di-tert-butyl nitroxide
(DTBN).6 Since pyridyl halides generally represent the
least reactive class of π-deficient substrates among
typical heteroaromatic azines in both SRN1 and SNAr
reactions,7 2-bromopyridine (2) was chosen as the model
substrate. In cases where a particular nucleophile had
not been tested previously for SRN1 reactivity with
carboaromatic substrates, iodobenzene (9), bromobenzene
(10), and 1-bromomesitylene (11) were used as prototypi-
cal aryl halides.
Owing to the established superiority of liquid NH3 as
a medium for SRN1 reactions,1,5 and in keeping with our
aim to minimize the chemical complexity of the reaction
mixtures, we limited our experiments to this solvent,
using near-UV light (350 nm) to promote the substitution.
The respective carbanions were generated with an alkali
metal amide, usually KNH2, which represents the stron-
There have been no reports to date of carbanions
derived from 2-alkyl-2-oxazolines and 2-alkylthiazoles
participating in SRN1 reactions. In addition, the extent
to which such carbanions can be generated by means of
alkali metal amides in liquid NH3 is not well established.8
When we began this study, examples of SRN1 reactions
of carboxamide and ester enolates with π-deficient het-
eroaromatic halides were rare, although Rossi had re-
X Abstract published in Advance ACS Abstracts, August 1, 1997.
(1) For reviews on SRN1 reactions, see: (a)Rossi, R. A.; de Rossi, R.
H. Aromatic Substitution by the SRN1 Mechanism; ACS Monograph
178; American Chemical Society: Washington, DC, 1983. (b) Bowman,
W. R. Chem. Soc. Rev. 1988, 17, 283. (c) Rossi, R. A.; Pierini, A. B.;
Palacios, S. M. In Advances in Free Radical Chemistry; Tanner, D. D.,
Ed.; J AI Press: Greenwich, CT, 1990; Chapter 5, p 193; J . Chem. Educ.
1989, 66, 720. (d) Norris, R. K. In Comprehensive Organic Synthesis;
Trost, B. M., Ed.; Pergammon: New York, 1991; Vol. 4, p 451. (e)
Saveant, J .-M. Tetrahedron 1994, 50, 10117. (f) Rossi, R. A., Pierini,
A. B., Penenory, A. B. In The Chemistry of Functional Groups, Suppl.
D, The Chemistry of Halides, Pseudo-halides and Azides; Patai S.,
Rappoport, Z., Eds.; Wiley: New York, 1995; Chapter 24, p 1395.
(2) (a) Dillender, S. C.; Greenwood, T. D.; Hendi, M. S.; Wolfe, J . F.
J . Org. Chem. 1986, 51, 1184. (b) Hermann, C. K. F.; Sachdeva, Y. P.;
Wolfe, J . F. J . Heterocycl. Chem. 1987, 24, 1016.
(6) (a) Hoffman, A. K.; Feldman, A. M.; Geblum, E.; Hodgson, W.
G. J . Am. Chem. Soc. 1964, 86, 639. (b) Nelson, S. F.; Bartlett, P. D.
J . Am. Chem. Soc. 1966, 88, 143. (c) Carver, D. R.; Hubbard, J . S.;
Wolfe, J . F. J . Org. Chem. 1982, 47, 1036.
(7) (a) Komin, A. P.; Wolfe, J . F. J . Org. Chem. 1977, 42, 2481. (b)
Carver, D. R.; Komin, A. P., Hubbard, J . S.; Wolfe, J . F. J . Org. Chem.
1981, 46, 294. (c) Borosky, G. L.; Pierini, A. B.; Rossi, R. A. J . Org.
Chem. 1992, 57, 247.
(8) Treatment of oxazolines 1a and 1b with KNH2 or NaNH2 in
liquid NH3 followed by propyl nitrate affords the corresponding 2-(R-
nitroalkyl) derivatives in yields of 13 and 75-84%, respectively: Feuer,
H.; Bevinakatti; Luo, X.-G. J . Heterocycl. Chem. 1986, 23, 825.
(9) (a) Rossi, R. A.; Alonso, R. A. J . Org. Chem. 1980, 45, 1239. (b)
Palacios, S. M.; Asis, S. E.; Rossi, R. A. Bull. Soc. Chim. Fr. 1993,
130, 111.
(3) For a review of the synthetic utility of carbanions derived from
such oxazolines, and for other applications of 2-oxazolines in synthesis,
see: (a) Meyers, A. I.; Mihelich, E. D. Angew. Chem., Int. Ed. Engl.
1976, 15, 270. (b) Reuman, M.; Meyers, A. I. Tetrahedron 1985, 41,
837. (c) Gant, T. G., Meyers, A. I. Tetrahedron 1994, 50, 2297.
(4) (a) Altman, L. J .; Richeimer, S. L. Tetrahedron Lett. 1971, 4709.
(b) Dondoni, A. In Modern Synthetic Methods; Scheffold, R., Ed.; Verlag
Helvetica Chimica Acta: Basel, 1992; pp 377-437. (c) Dondoni, A. In
New Aspects of Organic Chemistry II; Yoshida, Z., Ohshiro, Y., Eds.;
Kodansha: Tokyo, and VCH: Weinheim, 1992; pp 105-128.
(5) (a) Bunnett, J . F.; Scamehorn, R. G., Traber, J . Org. Chem. 1976,
41, 3677. (b) Moon, M. A.; Wolfe, J . F. J . Org. Chem., 1979, 44, 4081.
(10) Alonso, R. A.; Rodriguez, C. H.; Rossi, R. A. J . Org. Chem. 1989,
54, 5983.
S0022-3263(96)02353-5 CCC: $14.00 © 1997 American Chemical Society

Compatibility of Various Carbanion Nucleophiles
J . Org. Chem., Vol. 62, No. 18, 1997 6153
Ta ble 1. Rea ction s of 2,4,4-Tr ia lk yl-2-oxa zolin e Ca r ba n ion s Der ived fr om 1a ,b w ith Heter oa r om a tic a n d Ca r boa r om a tic
Ha lid es
carbanion
from
reaction
time (min)
reaction
conditions
products
expt
cation
substr
(yield, %)a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1b
1b
1b
1b
1b
1b
1b
K
K
Li
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
2
2
2
2
2
6
6
6
9
10
9
9
11
11
11
11
2
2
2
2
9
5
60
60
5
5
5
5
5
5
120
5
5
5
60
5
5
5
120
5
5
5
hν
3a (26),c 4 (50),c 5 (1)c
3a (33), 4 (59)d
3a (33), 4 (45)d
3a (4), 4 (64)
3a (<1), 4 (76)
7 (41), 8 (16)
7 (45), 8 (21)
6 (37), 7 (13), 8 (22)
1b (56),c 12a (28),c 5 (3)c
1b (63), 12a (28)
1b (28)
hν
hν
dark
hν, DTBNb
hν
dark
hν, DTBNb
hν
hν
dark
hν, DTBNb
1b (24)
hν
hν
dark
12b (50)d
12b (94)
s.m.e
hν, DTBNb
hν
s.m.e
3b (70)
hν
dark
3b (88)
s.m.e
hν, DTBNb
hν
1b (66),c 3b (34)c
12a (43)
9
9
9
60
5
5
hν
12a (62)
dark
s.m.e
hν, DTBNb
12a (26)
a
b
Unless indicated otherwise yields are those of isolated products. 10 mol % based on substrate. c Yields determined by 1H NMR.
Compound 5 was observed by 1H NMR in the crude reaction mixture, but not isolated. e Only starting materials (s.m.) could be detected
d
in the 1H NMR spectrum of the crude reaction mixture.
Sch em e 2
ported on the successful application of photostimulated,
intermolecular SRN1 reactions of amide9 and lactam10
enolates with several carboaromatic halides. Our re-
port11 of the intramolecular photocyclization of the R-eno-
late of 3-acetamido-2-chloropyridine was the only pub-
lished example of a putative heteroaromatic SRN1 reaction
involving a carboxamide enolate. Reports12 of reactions
of ester enolates with carboaromatic halides were scarce
as well, and the only example with a halogenated
heterocycle involved the reaction of 2,6-dibromopyridine
with the potassium enolate of ethyl phenylacetate.13
More recently, McKillop and van Leeuwen published a
thorough study of ferrous ion-promoted SRN1 reactions
of the sodium enolate of tert-butyl acetate with carboaro-
matic and heteroaromatic halides.14 In the same paper,
the sodium enolates of N-acylmorpholines were found to
react with several carboaromatic substrates, but halo-
heterocycles were not studied. In spite of the synthetic
versatility of phosphonate-stabilized carbanions,15 there
appear to be no literature examples of these or other
phosphorous-stabilized carbanions participating in either
carboaromatic or heteroaromatic SRN1 reactions.
Resu lts a n d Discu ssion
Ca r ba n ion s Der ived fr om 2,4,4-Tr ia lk yl-2-oxa zo-
lin es. These reactions are illustrated in Scheme 2, and
the results of specific experiments appear in Table 1.
Treatment of 2,4,4-trimethyl-2-oxazoline (1a ) with 1
equiv of KNH2 in liquid NH3 followed by addition of
(11) Goehring, R. R.; Sachdeva, Y. P.; Pisipati, J . S.; Sleevi, M. C.;
Wolfe, J . F. J . Am. Chem. Soc. 1985, 107, 435.
(12) (a) Semmelhack, M. F.; Bargar, T. J . Org. Chem. 1977, 42, 1481.
(b) Semmelhack, M. F.; Bargar, T. J . Am. Chem. Soc. 1980, 102, 7765.
(13) Carver, D. R.; Greenwood, T. D.; Hubbard, J . S.; Komin, A. P.;
Sachdeva, Y. P.; Wolfe, J . F. J . Org. Chem. 1983, 48, 1180.
(14) van Leeuwen, M.; McKillop, A. J . Chem. Soc., Perkin Trans. 1
1993, 2433.
(15) (a) Cadogen, J . I. G., Ed. Organophosphorous Reagents in
Organic Synthesis; Academic Press: London, 1979. (b) Maryanoff, B.
E.; Reitz, A. B. Chem. Rev. 1989, 89, 863.
2-bromopyridine (2) with irradiation of the reaction
mixture for 5 min at -33 °C afforded substitution product
3a (26%) along with 50% of 2-aminopyridine (4) and

6154 J . Org. Chem., Vol. 62, No. 18, 1997
Wong et al.
Sch em e 3
oxazoline dimer 5 (1%). Similar results were obtained
after 60 min of irradiation using either KNH2 or LiNH2
(Table 1 expts 1-3). Yields of 3a declined dramatically
after 5 min of illumination in the presence of 10 mol %
of DTBN, or when the reaction was carried out in the
dark. In both cases (expts 4 and 5) yields of 4 increased
sharply, while dimer 5 could not be detected.
When 2-chloroquinoline (6) was employed as substrate
with the carbanion of 1a under illumination, a rapid
reaction occurred to consume 6 and give 2-(2-quinolyl-
methyl)oxazoline 7 (41%) along with a 16% yield of
2-aminoquinoline (8) (expt 6). Similar yields of 7 and 8
were obtained from reactions conducted in the dark (expt
7). However, irradiation in the presence of 10 mol % of
DTBN (expt 8) resulted in much lower yields (13%) of 7
and recovery of more starting materials when compared
with both photostimulated and dark reactions conducted
without DTBN.
A series of photostimulated, dark, and DTBN-inhibited
reactions of oxazoline 1a in the presence of KNH2 in
liquid NH3 with the carboaromatic substrates iodoben-
zene (9), bromobenzene (10), and 1-bromomesitylene (11)
produced mono- (1b and 12b) and diarylated (12a )
products as summarized in Table 1 (expts 9-16).
The results obtained with 1a and 2-bromopyridine (2)
are consistent with photoinduced, radical chain pyridy-
lation of the lateral carbanion of 1a to give 3a . Forma-
tion of 2-aminopyridine (4) in photostimulated reactions
(expts 1-3), and to a greater extent in dark and in
DTBN-inhibited reactions (expts 4 and 5), indicates that
1a is incompletely deprotonated by KNH2 in liquid NH3.
Thus, amide ion competes via an ionic SNAr reaction for
substrate 2 with the carbanion of 1a , which appears to
react with 2 mainly by the SRN1 mechanism. When the
SRN1 pathway is impeded by conducting the reaction in
the dark, or in the presence of DTBN, the ionic amination
reaction dominates.6c In the reactions of the carbanion
of 1a with 2-chloroquinoline (6) the yield of 7 was
diminished only in the presence of DTBN (expt 8),
near-UV irradiation for 1 h, followed by careful anaerobic
quenching with solid NH4Cl to avoid possible aerial
oxidative dimerization of the carbanion (vide infra),
afforded 5 in 30% yield. Without illumination, 5 was
formed in only trace amounts under otherwise identical
experimental conditions or when the dark reaction
mixture was exposed to air during neutralization with
NH4Cl.
2-Benzyl-4,4-dimethyl-2-oxazoline (1b) was readily
converted to its lateral carbanion8 by means of 1 equiv
of KNH2 in liquid NH3 as evidenced by facile, photoin-
duced reactions with 2 to form 2-pyridyl derivative 3b
in yields of 70-88% (expts 17 and 18). When the reaction
mixture was denied illumination, only starting materials
were present after 5 min (expt 19) and DTBN exerted a
strong inhibitory effect on the photostimulated substitu-
tion reaction (expt 20). None of these reactions produced
detectable (1H NMR) amounts of 2-aminopyridine (4) or
the phenylated analog of dimer 5. Iodobenzene (9) proved
to be less reactive than 2 as an electron acceptor in
photostimulated reactions (expts 21 and 22) and exhib-
ited a complete lack of reactivity in the dark, along with
diminished photoreactivity in the presence of DTBN
(expts 23 and 24).
implicating the occurrence of a thermally initiated SRN
process in the dark reaction (expt 7).
1
In contrast to reactions involving 2-bromopyridine (2)
(expts 4 and 5) when reactions of 1a with iodobenzene
(9) were conducted in the dark (expt 11) or with irradia-
tion in the presence of DTBN (expt 12), appreciable (24-
28%) amounts of the phenylated product 1b were formed.
It seems unlikely that the carbanion of 1a would react
with 9 by either an ionic SNAr mechanism or by an SRN
1
Ca r ba n ion s Der ived fr om 2,4-Dia lk ylth ia zoles.
Results of reactions of the carbanions derived from 2,4-
dimethylthiazole (13a ) and 2-benzyl-4-methylthiazole
(13b) with 2-bromopyridine (2), iodobenzene (9), and
bromobenzene (10) are illustrated in Scheme 3 and
summarized in Table 2.
As with oxazoline 1a , deprotonation at the 2-methyl
group of 13a by KNH2 is incomplete as evidenced by
competitive formation of 2-aminopyridine (4) when 2 was
employed as substrate (expts 25-27). Amination of
bromobenzene (10) was not observed, but this substrate,
like 2, displayed the dependency on near-UV irradiation
and susceptibility to DTBN (expts 28-30) expected for a
photostimulated SRN1 reaction.
pathway that was more resistant to inhibition or more
prone to thermal initiation than the analogous radical
chain process involving 2. Instead, it is more reasonable
to assume that amide ion in equilibrium with 1a effects
benzyne phenylation of the carbanion of 1a to produce
1b. When the possibility of aryne formation was pre-
cluded by using 1-bromomestiylene (11) as substrate,
photostimulated reactions of the carbanion of 1a to form
substitution product 12b proceeded quite well (expts 13
and 14), while dark and DTBN-inhibited reactions with
11 gave only recovered starting materials (expts 15 and
16).
Although it is tempting to assume that dimer 5 is
formed by coupling of 2-oxazolinylmethyl radicals gener-
ated in the initiation step of the radical chain substitution
involving the carbanion of 1a (Scheme 1, step 1), this is
not the only pathway available for dimerization. Thus,
treatment of 1a alone with KNH2 in liquid NH3 under
Reactions of the carbanion of 13a with substrates 2
and 10 were accompanied by formation of thiazole dimer
15 in yields of 12-18% under photostimulation (with or
without DTBN) and in the dark (expts 25 and 28-30).
Dimer 15 was also produced from 13a in 20% yield in

Compatibility of Various Carbanion Nucleophiles
J . Org. Chem., Vol. 62, No. 18, 1997 6155
Ta ble 2. Rea ction s of 2,4-Dia lk ylth ia zole Ca r ba n ion s Der ived fr om 13a ,b w ith Heter oa r om a tic a n d Ca r boa r om a tic
Su bstr a tes
carbanion
from
reaction
time (min)
reaction
conditions
products
expt
cation
substr
(yield, %)a
25
26
27
28
29
30
31
32
33
34
35
36
37
13a
13a
13a
13a
13a
13a
13b
13b
13b
13b
13b
13b
13b
K
K
K
K
K
K
K
K
K
K
K
K
K
2
2
2
10
10
10
2
2
2
9
100
100
100
100
100
100
100
100
100
100
100
100
100
hν
14a (37), 4 (37), 15 (12)
4 (89)
4 (89)
13b (59), 15 (14)
13b (12), 15 (14)
15 (18)
dark
hν, DTBN
hν
dark
hν, DTBN
hν
dark
hν, DTBN
hν
hν
dark
hν, DTBN
14b (94)
s.m.b
s.m.b
16 (39)
16 (57)
10
10
10
16 (9)
s.m.b
Unless indicated otherwise, yields are those of isolated products. Only starting materials (s.m.) could be detected in the 1H NMR
a
b
spectrum of the crude reaction mixture.
Sch em e 4
the absence of substrate by irradiating a mixture of 13a
and KNH2 in liquid NH3 for 1 h, followed by an anaerobic
quench with NH4Cl. Without irradiation, <2% of 15 was
formed under these conditions. Thiazole 13a afforded
18% of dimer 15 when a similar dark reaction mixture
was poured onto solid NH4Cl, thereby exposing the
carbanion to air. This is in contrast to oxazoline 1a ,
which failed to yield dimer 5 upon similar treatment.
Thus, while the carbanions of both 1a and 13a undergo
photoinduced dimerization, only the latter is appreciably
susceptible to aerobic oxidative dimerization. The anaer-
obic dimerization of these carbanions may result from a
redox process involving photostimulated electron transfer
from the carbanion to the respective neutral precursors,
1a and 13a , which are present owing to their incomplete
deprotonation.
Deprotonation of 2-benzyl-4-methylthiazole (13b) with
KNH2 appears complete as evidenced by nearly quantita-
tive, photoassisted 2-pyridylation of the resulting car-
banion to afford 14b (expt 31). This reaction, as well as
the analogous phenylations with 9 and 10, required
irradiation and was strongly inhibited by DTBN (expts
32-37). The lower reactivity of the carbanion of 2-ben-
zylthiazole 13b toward carboaromatic halides 9 and 10
versus heterocyclic substrate 2 may be the result of the
greater rate at which this more stable (lower pKa)
carbanion transfers an electron to the more easily
reduced pyridyl halide 2 than to halobenzenes 9 and
10.7,16
Ta ble 3. Rea ction s of th e P ota ssiu m En ola tes of
N,N-Dim eth yla ceta m id e (17a ), ter t-Bu tyl Aceta te (17b),
a n d Eth yl P h en yla ceta te (17c) w ith Su bstr a tes 2 a n d 6
reaction
reaction
products
expt enolate substr time (min) conditions
(yield, %)a
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
17a
17a
17a
17a
17a
17a
17a
17a
17a
17b
17b
17b
17c
17c
17c
2
2
2
2
6
6
6
6
6
2
2
2
2
2
2
5
20
5
5
5
60
5
5
5
5
hν
18a (74),b 19a (5)b
18a (84), 19a (7)
18a (17)
18a (2)
20 (87)
Ca r boxa m id e a n d Ester En ola tes. As anticipated
from earlier reports of carboxamide enolate ion reactivity
toward carboaromatic halides under photostimulation9
or in the presence of ferrous ion,14 the potassium enolate,
17a , of N,N-dimethylacetamide reacted smoothly with
heterocyclic halides 2 and 6 under illumination to afford
good yields of 2-pyridyl (18a ) and 2-quinolyl (20) prod-
ucts, accompanied by minor amounts of disubstitution
products 19a and 21, respectively (Scheme 4). These
results are summarized in Table 3 (expts 38-46), where
it may be seen that the radical chain character of the
reactions was indicated by their relatively sluggish
nature in the dark and the strong inhibitory action of
catalytic amounts of DTBN. It appears that the dark
reaction of amide enolate 17a with 2-chloroquinoline (6)
to give 20 (expt 44) is a thermally initiated SRN1 process,
hν
dark
hν, DTBN
hν
hν
dark
20 (80), 21 (14)
20 (41)
dark, DTBN s.m.c
hν, DTBN
hν
dark
hν, DTBN
hν
dark
s.m.c
18b (44), 19b (21)
18b (5)
s.m.c
5
5
5
5
18c (77)
s.m.c
5
hν, DTBN
s.m.c
a
Unless indicated otherwise, yields are those of isolated
products. Yields determined by GC. c Only starting materials
(s.m.) could be detected in the 1H NMR spectrum of the crude
reaction mixture.
b
since DTBN, which would not be expected to inhibit an
ionic SNAr reaction, prevents formation of 20 in the dark
(expt 45).
(16) Borosky, G. L.; Pierini, A. B.; Rossi, R. A. J . Org. Chem. 1992,
57, 247.

6156 J . Org. Chem., Vol. 62, No. 18, 1997
Wong et al.
Sch em e 5
Reactions of the potassium enolates of tert-butyl ac-
etate (17b) and ethyl phenylacetate (17c) with substrate
2 (expts 47-52) exhibited mechanistic features similar
to those of the carboxamide enolates. Illumination
facilitated the substitution process, while experiments
involving near-UV light together with DTBN (expts 49
and 52), or those conducted in the dark (expts 48 and
51), failed to afford products 18b,c and 19b.
Although dienolate ions derived from R,â-unsaturated
carbonyl compounds have not been unequivocally identi-
fied as participating nucleophiles in either carboaromatic
or heteroaromatic SRN1 reactions,11,17 we have now found
that the potassium dienolates of (E)-N-methyl-N-phenyl-
2-butenamide (22) and tert-butyl 3-butenoate (25) both
react smoothly and regioselectively with substrates 2 and
9 under photostimulation to give exclusively γ-substi-
tuted products (Scheme 5). Thus, reaction of the dieno-
late of amide 22 with 2 gave 23a in 50% isolated yield,
with no detectable (1H NMR) amounts of products arising
from pyridylation at the position R to the carbonyl group.
The analogous reaction of this dienolate with 9 similarly
resulted in exclusive γ-phenylation, but was complicated
by formation of both mono- (23b) and diphenylated (23c
and 24) products.
22, phenylation of the ester dienolate mainly resulted in
monosubstitution at the γ-position to produce both 26b
(42%) and diphenyl derivatives 26c (8%) and 27 (8%).
Photoassisted reactions of the potassium dienolate of
ester 25 with both 2 and 9 were accompanied by forma-
tion of ester self-condensation product 28a , the origin of
which can be traced to a Michael addition of the original
dienolate anion to the neutral R,â-unsaturated isomer of
25, produced by proton-metal exchange between excess
dienolate present in the reaction mixture and the acidic
R-hydrogens of the SRN1 product.18 All of the substitution
reactions shown in Scheme 5, including formation of
diester 28a , failed to proceed in the dark.
Ca r b a n ion of Dim et h yl Met h ylp h osp h on a t e.
Treatment of dimethyl methylphosphonate (29a ) with 1
equiv of KNH2 in liquid NH3 generated R-carbanion 29b
in sufficient equilibrium concentration to afford R-(2-
pyridyl) phosphonate 30a in 68% yield after 1 h of
illumination (expt 53). The accompanying formation of
2-aminopyridine (4) in 10% yield implicated competitive
SNAr amination of 2 (Table 4, expt 53). This was further
Pyridylation of the dienolate from â,γ-unsaturated
ester 25 gave 26a (61%) unaccompanied by either R-cou-
pled or R,â-unsaturated products. As in the case of amide
supported by the results of expts 54 and 55, where
amination of 2 became the major substitution pathway
when electron transfer was suppressed without a light
source or by adding DTBN to the illuminated reaction
mixture. Phenylation of carbanion 29b with iodobenzene
(9) to form 30b,c responded to photostimulation (expt 56),
(17) (a) A report that the dienolate of cyclohex-2-en-1-one failed to
react with bromobenzene under near-UV irradiation appears to be the
only published example of an attempt to effect an intermolecular
aromatic SRN1 reaction with the dienolate of an R,â-unsaturated
carbonyl compound: Bunnett, J . F.; Sundberg, J . E. J . Org. Chem.
1976, 41, 1702. (b) Our earlier finding11 that carbanions from a series
of R,â-unsaturated o-haloanilides underwent photoinduced cyclizations
prompted us to intiate the present study of intermolecular reactions
with dienolates from R,â-unsaturated anilides. (c) For a description of
SRN1 reactions involving carbanions from R,â-unsaturated nitriles,
see: Alonso, R. A.; Austin, E.; Rossi, R. A. J . Org. Chem. 1988, 53,
6065.
(18) (a) Munch-Petersen, J . J . Org. Chem. 1957, 22, 170. (b) Shabtai,
J .; Ney-Igner, E., Pines, H. J . Org. Chem. 1981, 46, 3795. (c) Shabtai,
J .; Pines, H. J . Org. Chem. 1965, 30, 3854.

Compatibility of Various Carbanion Nucleophiles
J . Org. Chem., Vol. 62, No. 18, 1997 6157
Ta ble 4. Rea ction s of Dim eth yl Meth ylp h osp h on a te
Ca r ba n ion 29b w ith Su bstr a tes 2 a n d 9
Galbraith Laboratories, Inc., Knoxville, TN, or by Atlantic
Microlab, Inc., Norcross, GA.
Gen er a l P r oced u r e for P h otostim u la ted Rea ction s.
To a solution of KNH2 (2 or 3 equiv) in liquid NH3 (15 mL/
mmol substrate) was added the nucleophile precursor (2 or 3
equiv) via syringe. After stirring at -33 °C for 15 min, the
lights of the photoreactor were turned on and the haloaromatic
substrate (1 equiv) was immediately added. Irradiation of the
resulting solution was continued for the specified time (see
tables) after which the reaction mixture was quenched by
pouring onto excess solid NH4Cl in a beaker. The reaction
vessel was rinsed with 2 × 100 mL portions of ether, and the
rinses were added slowly to the NH3 solution. After evapora-
tion of the NH3 on a steam bath, the ethereal solution was
decanted and the solid residue was washed with 2 × 50 mL
portions of ether. The combined ethereal solutions were dried,
filtered, and concentrated on a rotary evaporator to yield the
crude product which was purified as further described in this
section.
reaction
reaction
products
expt K salt substr time (min) conditions
(yield, %)a
53
54
55
56
57
58
29b
29b
29b
29b
29b
29b
2
2
2
9
9
9
60
25
25
60
25
25
hν
dark
30a (68), 4 (10)
2 (40), 4 (51)
hν, DTBN 30a (4)b
hν
dark
30b (47), 30c (18)
s.m.c
hν, DTBN 30b (30)d
a
Unless indicated otherwise, yields are those of isolated
b
products. The crude reaction mixture also contained 2 and 4 as
determined by qualitative GC and 1H NMR analysis. c Only
starting materials (s.m.) could be detected in the 1H NMR
d
spectrum of the crude reaction mixture. Unreacted 9 was also
present.
failed to occur in the dark (expt 57), but was surprisingly
resistant to inhibition by DTBN (expt 58).
The same procedure was used in dark reactions except the
reaction flask was wrapped in a black cloth before addition of
the substrate. In inhibited reactions, 10 mol % of DTBN was
added to the solution of the nucleophile just prior to the
addition of the substrate. The results of photostimulated
reactions are described below. In all cases the molar quantity
of KNH2 was the same as that of the carbanion precursor.
2-Ben zyl-4,4-d im eth yl-2-oxa zolin e (1b) a n d 2-Ben zh y-
d r yl-4,4-d im eth yl-2-oxa zolin e (12a ). Irradiation of 2,4,4-
trimethyl-2-oxazoline (1a ) (3.39 g, 30.0 mmol) and 9 (2.04 g,
10.0 mmol) for 5 min afforded 1.54 g of a yellow oil, the 1H
NMR spectrum of which showed three components: mono-
substitution product 1b,19 disubstitution product 12a ,8 and
oxazoline dimer, 1,2-bis(4′,4′-dimethyl-2′-oxazolin-2′-yl)ethane
(5),23 in a molar ratio of 15:4:1, respectively. Calculation of
weight ratios gave 1.06 g (56%) of 1b, 0.39 g (28%) of 12a ,
and 0.09 g (3%) of 5 (based on total oxazoline 1a ). Separation
of this mixture was accomplished by MPLC using EtOAc to
elute first 12a (0.34 g) followed by 1b (0.98 g). Dimer 5 (0.06
g) was then eluted with 10% MeOH-EtOAc. The 1H NMR
spectra of these compounds, all of which were colorless oils,
were in complete agreement with those reported earlier.
2-((4,4-Dim et h yloxa zolin -2-yl)m et h yl)p yr id in e (3a ).
Photostimulated reaction of 1a (5.77 g, 51.0 mmol) and 2 (2.69
Con clu sion
Carbanions derived from alkyl oxazolines (1a ,b), thia-
zoles (13a ,b), and phosphonate 29a by means of KNH2
in liquid NH3 are shown here for the first time to
participate in photostimulated SRN1 reactions with car-
boaromatic and heteroaromatic halides. These reactions
proceed with sufficient generality and efficiency to be
synthetically useful, with the caveat that incomplete,
equilibrium deprotonation of the respective carbanion
precursors may result in competitive SNAr amination of
certain heterocyclic substrates.
The present results involving the potassium enolates
of N,N-dimethylacetamide (17a ), tert-butylacetate (17b),
and ethyl phenylacetate (17c) in SRN1 reactions with
heterocyclic substrates 2 and 6 complement the results
of earlier studies of reactions of these and related
carboxamide9-11,14 and ester12,14 enolates with carboaro-
matic halides and, in the process, establish these enolates
as versatile nucleophiles in aromatic SRN1 chemistry.
Successful SRN1 reactions of dienolates derived from
carboxamide 22 and ester 25, respectively, represent the
first examples of enolates of R,â-unsaturated carbonyl
compounds participating in such radical chain processes
with levels of reactivity and regioselectivity that make
them preparatively useful and predictable.
1
g, 17.0 mmol) after 5 min gave 1.64 g of a brown oil, H NMR
analysis of which indicated a 1:1.9 mixture of 3a and 2-ami-
nopyridine (4), respectively. Crude 3a , 0.85 g, was isolated
by MPLC using 10% MeOH-CH2Cl2 as eluent. GC-MS
analysis of this colorless oil showed 3a contaminated with ca.
12 mol %, 0.12 g, of dimer 5. 1H NMR for 3a : δ 1.29 (s, 6H),
3.82 (s, 2H), 3.94 (s, 2H), 7.15-7.18 (m, 1H), 7.19-7.32 (m,
1H), 7.61-7.68 (m, 1H), 8.55 (m, 1H). MS(EI): m/ z (relative
intensity) 190 (M+, 10), 189 (10), 175 (100), 118 (80), 93 (70),
78 (30). HRMS: calcd for C11H14N2O 190.1106. Found:
190.1097.
Exp er im en ta l Section
Gen er a l. Photostimulated reactions were performed under
N2 using a Rayonet RPR-240 or Rayonet RPR-100 photochemi-
cal reactor equipped with lamps emitting maximally at 350
nm. Commercial anhydrous NH3 (Matheson) was used directly
from the tank. 2-Benzyl-4,4-dimethyl-2-oxazoline (1b) was
prepared by the reaction of phenylacetic acid with 2-amino-
2-methylpropanol.19 2-Benzyl-4-methylthiazole (13b) was ob-
tained from the reaction of phenylthioacetamide and chloro-
acetone.20 (E)-N-Methyl-N-phenyl-2-butenamide (22)21 and
tert-butyl 3-butenoate (24)22 were synthesized via reactions of
but-2-enoyl chloride with N-methylaniline and 2-methyl-2-
propanol, respectively. All other reagents were obtained
commercially and distilled or recrystallized before use. 1H
NMR spectra were determined using CDCl3 as the solvent
unless otherwise indicated. Medium pressure liquid chroma-
tography (MPLC) was carried out with 230-400 mesh E.
Merck silica gel. Elemental analyses were performed by either
2-((4,4-Dim et h yloxa zolin -2-yl)b en zyl)p yr id in e (3b ).
From the reaction of 2-benzyl-4,4-dimethyl-2-oxazoline (1b)
(3.92 g, 20.7 mmol) and 2 (1.09 g, 6.90 mmol) after 2 h of
irradiation was obtained 3.51 g of a crude product mixture
which was chromatographed first with 1:1 hexane-ether to
remove excess 1b and then with ether to give 1.01 g of 3b as
a light yellow solid. The eluent was changed to 10% MeOH-
ether, and 0.60 g of a mixture of 3b and its enamine tautomer
3c was eluted. Analysis of this mixture by GC indicated a
2:1 ratio of 3b:3c. Preparative GC afforded 3c as a white solid.
The total yield of 3b,c amounted to 1.61 g (88%). 1H NMR for
3b: δ 1.29 (s, 3H), 1.32 (s, 3H), 3.97 (s, 2H), 5.26 (s, 1H), 7.12-
7.16 (m, 1H), 7.26-7.64 (m, 7H), 8.56 (m, 1H). MS(EI): m/ z
(relative intensity) 266 (M+, 25), 251 (27), 167 (100). 1H NMR
for 3c: δ 1.36 (s, 3H), 1.40 (s, 3H), 4.05 (s, 2H), 7.38 (m, 8H),
8.57 (m, 1H). Anal. (mixture 3b,c). Calcd for C17H18N2O: C,
76.66; H, 6.81; N, 10.52. Found: C, 76.41; H, 6.86; N, 10.39.
2-((4,4-Dim et h yloxa zolin -2-yl)m et h yl)q u in olin e (7).
From 1a (3.00 g, 26.6 mmol) and 6 (1.44 g, 8.86 mmol) was
(19) Hansen, J . F.; Wang, S. J . Org. Chem. 1976, 41, 3635.
(20) Clarke, G. M.; Sykes, P. J . Chem. Soc. C 1967, 1269.
(21) Maxim, N. N.; Ioanid, N. Bull. Soc. Chim. Romania 1930, 12,
28; Chem. Abstr. 1931, 25, 488.
(23) Furuta, K.; Ikeda, N.; Yamamoto, H. Tetrahedron Lett. 1984,
(22) Boots, S. G.; Boots, M. R. J . Pharm. Sci. 1975, 64, 1262.
675.

6158 J . Org. Chem., Vol. 62, No. 18, 1997
Wong et al.
obtained 2.19 g of an orange oil which was chromatographed
with 10% EtOH-EtOAc to give 0.21 g (16%) of 2-aminoquino-
line (8) followed by 0.87 g (41%) of 7 as a yellow oil which
solidified when triturated with hexane. Recrystallization from
hexane afforded pure 7 as a pale yellow solid, mp 45 °C. 1H
NMR: δ 1.31 (s, 6H), 3.95 (s, 2H), 4.01 (s, 2H), 7.44-7.54 (m,
2H), 7.67-7.73 (m, 1H), 7.80 (d, 1H), 8.06-8.14 (m, 2H). Anal.
Calcd for C15H16N2O: C, 74.97; H, 6.71; N, 11.66. Found: C,
74.72; H, 6.70; N, 11.61.
2-Ben zh yd r yl-4,4-d im eth yl-2-oxa zolin e (12a ). Irradia-
tion of 2-benzyl-4,4-dimethyl-2-oxazoline (16) (3.83 g, 20.2
mmol) and 10 (1.06 g, 6.73 mmol) yielded after 1 h 1.11 g (62%)
of 12a as a light yellow oil, lit.8 mp 57-58 °C, following MPLC
with 2:1 hexane-ether to remove excess 16. 1H NMR: δ 1.29
(s, 6H), 3.96 (s, 2H), 5.10 (s, 1H), 7.24-7.31 (m, 10H).
MS(CI): m/ z (relative intensity) 265 (M+, 100), 167 (80), 104
(30), 76 (30). HRMS: calcd for C18H19NO 265.1466. Found:
265.1421.
2-(Mesitylm eth yl)-4,4-d im eth yl-2-oxa zolin e (12b). Ir-
radiation of 1a (2.44 g, 21.6 mmol) and 11 (1.43 g, 7.2 mmol)
for 1 h gave 1.56 g (94%) of 12b as a light yellow oil by MPLC
with 1:1 hexane-ether. 1H NMR: δ 1.23 (s, 6H), 2.23 (s, 3H),
2.32 (s, 6H), 3.59 (s, 2H), 3.84 (s, 2H), 6.83 (s, 2H). Anal. Calcd
for C15H21NO: C, 77.88; H, 9.15; N, 6.05. Found: C, 77.71;
H, 9.22; N, 6.06.
2-Ben zyl-4-m et h ylt h ia zole (13b ). Reaction of 2,4-di-
methylthiazole (13a ) (3.61 g, 31.8 mmol) with 10 (1.67 g, 10.6
mmol) under photostimulation for 100 min afforded 3.76 g of
a dark brown oil. Chromatography with 2:1 hexane-EtOAc
gave, following elution of excess 13a , 1.33 g of crude 13b as
an orange oil and 0.50 g (14%) of 1,2-bis(4-methylthiazol-2-
yl)ethane (15) on the basis of 13a , mp 64-66 °C, lit.24 mp 68-
69 °C. Distillation of the crude oil gave 1.19 g (59%) of 13b
as a pale yellow liquid, bp 105 °C/0.50 mm, lit.20 bp 90 °C/0.20
mm. 1H NMR spectral data for 13b are in agreement with
those reported.25 1H NMR for 15: δ 2.42 (s, 6H), 3.45 (s, 4H),
6.72 (s, 2H).
2-((4-Meth ylth ia zol-2-yl)m eth yl)p yr id in e (14a ). From
13a (6.45 g, 57.0 mmol) and 2 (3.00 g, 19.0 mmol) was obtained
2.99 g of a dark brown oil after 100 min of irradiation. Column
chromatography with 10:1 CH2Cl2-MeOH afforded 1.34 g
(37%) of 14a as a yellow oil, 0.66 g (37%) of 4, and 0.77 g (12%)
of dimer 15. 1H NMR for 14a : δ 2.36 (s, 3H), 4.29 (s, 2H),
6.73 (s, 1H), 7.10-7.40 (m, 3H), 8.70 (d, 1H). MS(CI): m/ z
(relative intensity) 189 (M+, 100), 145 (20), 118 (25). HRMS:
calcd for C10H10N2S 190.0565. Found: 190.0565.
2-[r-(4-Meth ylth ia zol-2-yl)ben zyl]p yr id in e (14b). Ir-
radiation of 2-benzyl-4-methylthiazole (13b) (3.19 g, 16.9
mmol) and 2 (0.89 g, 5.63 mmol) for 100 min gave 3.74 g of a
dark red oil. After chromatographic separation of excess 13b
by elution with 2:1 hexane-EtOAc, 1.41 g (94%) of 14b was
obtained as a pale yellow solid. Recrystallization from hexane
afforded an analytical sample, mp 69 °C. 1H NMR: δ 2.43 (s,
3H), 5.93 (s, 1H), 6.81 (s, 1H), 7.13-7.39 (m, 7H), 7.60 (t, 1H),
8.31 (d, 1H). MS(CI): m/ z (relative intensity) 265 (M+, 100),
167 (80). Anal. Calcd for C16H14N2S: C, 72.15; H, 5.30; N,
10.52. Found: C, 72.17; H, 5.33; N, 10.62.
NH4Cl (2.00 g, 37.0 mmol) was added under N2 to the dark
green reaction mixture. The resulting colorless solution was
stirred for 15 min and then poured into a beaker. A 100 mL
ether rinse of the reaction flask was added cautiously, and the
NH3 was evaporated. The ether solution was decanted,
combined with an additional 100 mL of ether wash of the solid
residue, and concentrated on a rotary evaporator to remove
the solvent and unreacted 1a . The pale yellow oily residue of
dimer 5 solidified on standing, 0.68 g (30%). Similar reactions
carried out in the dark followed by quenching of the reaction
mixture either by the addition of NH4Cl (anaerobic quench)
or by pouring the reaction mixture onto NH4Cl in a beaker
(aerobic quench) gave only 0.02 g of a violet-colored residue,
the 1H NMR spectrum of which indicated the absence of 5.
Dim er iza tion of th e Ca r ba n ion of 2,4-Dim eth ylth ia -
zole (13a ). Irradiation of a solution containing 13a (2.26 g,
20.0 mmol) and 20.0 mmol of KNH2 in 200 mL of liquid NH3
for 1 h, followed by the anaerobic workup described previously
for the dimerization reaction of 1a , gave 0.60 g of crude dimer
15 as a dark red oil which solidified on standing. Purification
of 15 by MPLC using 1:1 CH2Cl2-EtOAc as eluent yielded 0.45
g (20%) of 15 as a yellow solid, mp 65-67 °C. A similar dark
reaction of 13a gave only 0.08 g of a dark red oily mixture
1
which contained 15 in ca. 50% purity by H NMR. However,
an aerobically quenched dark reaction of 13a afforded 0.40 g
(18%) of essentially pure 15 as an orange-red solid.
N,N-Dim eth ylp yr id in -2-yla ceta m id e (18a ) a n d N,N-
Dim eth yld ip yr id in -2-yla ceta m id e (19a ). N,N-Dimethyl-
acetamide (17a ) (1.31 g, 15.0 mmol) and 2 (0.79 g, 5.00 mmol)
after photostimulation for 5 min gave 1.19 g of crude product
whose GC analysis (I.S.: dimethyl phthalate) showed 74% of
18a and 5% of 19a .27 Analytical samples of 18a and 19a were
obtained as light yellow oils by preparative GLC. 1H NMR
for 18a : δ 2.97 (s, 3H), 3.09 (s, 3H), 3.92 (s, 2H), 7.17 (t, 1H),
7.33 (d, 1H), 7.62 (t, 1H), 8.54 (d, 1H). Anal. Calcd for
C9H12N2O: C, 65.83; H, 7.37; N, 17.06. Found: C, 65.68; H,
7.50; N, 16.90. 1H NMR for 19a : 3.05 (s, 6H), 5.74 (s, 1H),
7.18 (t, 2H), 7.32 (d, 2H), 7.62 (t, 2H), 8.56 (d, 2H). Anal. (as
HCl salt, mp 180 °C). Calcd for C14H16N3OCl: C, 60.54; H,
5.81; N, 15.13. Found: C, 60.30; H, 5.85; N, 15.03.
N,N-Dim eth ylqu in olin -2-yla ceta m id e (20) a n d N,N-
Dim eth yld iqu in olin -2-yla ceta m id e (21). Irradiation of the
carbanion derived from 17a (2.56 g, 29.3 mmol) and 6 (1.59 g,
9.78 mmol) for 60 min gave 2.20 g of a crude orange oily
mixture which was separated on an acidic alumina column
with EtOAc. Compound 20 was isolated as an orange oil, 1.68
g (80%) which crystallized upon trituration with ether, mp 99-
100 °C, followed by 21 as a red solid, 0.45 g (14%). 1H NMR
for 20: δ 2.98 (s, 3H), 3.13 (s, 3H), 4.11 (s, 2H), 7.52 (t, 2H),
7.69 (t, 1H), 7.82 (d, 1H), 8.03 (d, 1H), 8.12 (d, 1H). MS(EI):
m/ z (relative intensity) 214 (M+, 35), 170 (30), 143 (100), 72
(65). Anal. Calcd for C13H14N2O: C, 72.87; H, 6.59; N, 13.07.
Found: C, 72.90; H, 6.59; N, 13.09. 1H NMR for 21: 3.13 (s,
3H), 3.19 (s, 3H), 7.64 (bs, 1H), 7.97 (m, 2H), 8.09 (t, 4H), 8.24
(d, 2H), 8.32 (d, 2H), 8.95 (d, 2H). Anal. (as HCl salt, mp 111
°C). Calcd for C22H20N3OCl: C, 69.93; H, 5.34; N, 11.12.
Found: C, 70.14; H, 5.56; N, 11.14.
2-Ben zh yd r yl-4-m eth ylth ia zole (16). Reaction of 13b
(2.93 g, 15.5 mmol) and 10 (0.81 g, 5.16 mmol) after 100 min
gave 2.36 g of a red oil from which 0.78 g (57%) of 16 was
obtained by MPLC with 5:1 hexane-EtOAc as eluent. Crys-
tallization from hexane gave white crystals of 16, mp 60-63
°C, lit.26 mp 63-65 °C. 1H NMR: δ 2.49 (s, 3H), 5.80 (s, 1H),
6.79 (s, 1H), 7.27 (m, 10H).
P h otostim u la ted Dim er iza tion of th e Ca r ba n ion of
2,4,4-Tr im eth yl-2-oxa zolin e (1a ). To a solution of KNH2
prepared from potassium (0.78 g, 20.0 mmol) in 200 mL of
liquid NH3 was added a solution of 1a (2.26 g, 20.0 mmol) in
15 mL of anhydrous ether. After irradiation for 1 h, solid
ter t-Bu tyl P yr id in -2-yla ceta te (18b) a n d ter t-Bu tyl
Dip yr id in -2-yla ceta te (19b). From tert-butyl acetate (17b)
(2.59 g, 22.4 mmol) and 2 (1.17 g, 7.45 mmol), 1.00 g of crude
product was obtained, which by 1H NMR analysis consisted
of a 2:1 molar ratio of 18b:19b. Column chromatographic
separation using 2:1 hexane-ether led to the isolation of 0.64
g (44%) of 18b as a colorless liquid and 0.21 g (21%) of 19b as
a colorless solid, mp 90-92 °C after recrystallization from
hexane. The 1H NMR and mass spectral data of 18b are
consistent with the values reported.14,28 1H NMR for 19b: δ
1.45 (s, 9H), 5.35 (s, 1H), 7.18 (m, 2H), 7.39 (d, 2H), 7.64 (t,
2H), 8.67 (d, 2H). MS(EI): m/ z (relative intensity) 169 (M+
- CO2But, 70), 57 (100). Anal. Calcd for C16H18N2O2: C,
71.09; H, 6.71; N, 10.36. Found: C, 71.08, H, 6.71, N, 10.27.
(24) Lehr, H.; Guex, W.; Erlenmeyer, H. Helv. Chim. Acta 1945, 28,
1281.
(25) Bordwell, F. G.; Satish, A. V.; J ordan, F.; Rios, C. B.; Chang,
A. C. J . Am. Chem. Soc. 1990, 112, 792.
(26) Davis, M. A.; Campbell, D. J . Benzhydrylthiazole Derivatives
U.S. Patent 3,391,151, 1968; Chem. Abstr. 1968, 69, P67368u.
(27) Shuler, W. A.; Gross, A. Basic Unsaturated Carboxylic Acid
Amides. U.S. Patent 2,953,562, 1960; Chem. Abstr. 1961, 55, P4552f.
(28) Kiselyov, A. S.; Strekowski, L. J . Org. Chem. 1993, 58, 4476.

Compatibility of Various Carbanion Nucleophiles
J . Org. Chem., Vol. 62, No. 18, 1997 6159
Eth yl P h en ylp yr id in -2-yla ceta te (18c). Ethyl phenyl-
acetate (17c) (3.83 g, 23.3 mmol) and 2 (1.23 g, 7.78 mmol)
gave 1.45 g (77%) of 18c as a light yellow oil after MPLC with
2:1 hexane-ether. 1H NMR: δ 1.22-1.27 (t, 3H), 4.19-4.27
(q, 2H), 5.22 (s, 1H), 7.13-7.41 (m, 7H), 7.57-7.63 (m, 1H),
8.56 (d, 1H). Anal. Calcd for C15H15NO2: C, 74.38; H, 6.24;
N, 5.81. Found: C, 74.47; H, 6.26; N, 5.82.
(E )-N -Me t h y l-N -p h e n y l-4-(p y r id in -2-y l)-3-b u t e n a -
m id e (23a ). Irradiation of N-methyl-N-phenyl-2-butenamide
(22) (3.50 g, 20.0 mmol) and 2 (1.58 g, 10.0 mmol) for 1 h gave
4.4 g of a brown oil which was chromatographed using CH2Cl2
to remove 1.52 g (43%) of a mixture of 22 and its isomer,
N-methyl-N-phenyl-3-butenamide. The eluent was changed
to 10% MeOH-CH2Cl2, and 2.58 g of a yellow oil was collected.
Trituration with ether afforded 1.25 g (50%) of 23a as a light
yellow solid, an analytical sample of which was obtained by
recrystallization from ether-hexane as colorless crystals, mp
79-80 °C. 1H NMR: δ 3.10 (d, 2H, J ) 6.8 Hz), 3.30 (s, 3H),
6.37 (d, 1H, J ) 16 Hz), 6.68-6.79 (dt, 1H, J ) 16 Hz, J ′ )
6.8 Hz), 7.07-7.12 (m, 1H), 7.22-7.31 (m, 3H), 7.36-7.47 (m,
3H), 7.56-7.63 (m, 1H), 8.50 (m, 1H). MS(CI): m/ z (relative
intensity) 253 (M+ + 1100), 174 (5), 160 (7), 146 (38). Anal.
Calcd for C16H16N2O: C, 76.16; H, 6.39; N, 11.10. Found: C,
76.31; H, 6.46; N, 11.12.
(E)-N-Meth yl-4,N-d ip h en yl-3-bu ten a m id e (23b). Reac-
tion of 22 (3.50 g, 20.0 mmol) and 9 (2.04 g, 10.0 mmol) after
photostimulation for 1 h gave 4.29 g of a yellow oil. Unreacted
22 and its isomer, total 1.66 g (48%), were removed by vacuum
distillation at 70-75 °C/0.01 mm, and the residue was chro-
matographed using CH2Cl2 as the eluent. A 1.6:1 E:Z mixture
of disubstitution product 24, 0.13 g (8%), was eluted as the
first major component. Further elution gave 0.17 g (10%) of
disubstitution product 23c and then crude 23b, 1.14 g, as a
pale yellow oil which was vacuum distilled at 144-145 °C/
0.01 mm to afford 0.95 g (38%) of 23b. 1H NMR for 23c: δ
2.96 (d, 2H, J ) 7.2 Hz), 3.26 (s, 3H), 6.24 (t, 1H, J ) 7.2 Hz),
7.00-7.35 (m, 15H). HRMS: calcd for C23H21NO 327.1623.
Found: 327.1630. 1H NMR for 23b: δ 3.05 (d, 2H, J ) 6.1
Hz), 3.30 (s, 3H), 6.19 (d, 1H, J ) 16 Hz), 6.21-6.32 (dt, 1H,
J ) 16 Hz, J ′ ) 6.1 Hz), 7.19-7.47 (m, 10H). HRMS: calcd
for C17H17NO 251.1310. Found: 251.1305. 1H NMR for 24,
(E) isomer: δ 3.29 (s, 3H), 4.38 (d, 1H, J ) 8.8 Hz), 6.20 (d,
1H, J ) 16 Hz), 6.63-6.69 (m, 1H), 6.97-7.43 (m, 15H). 1H
NMR for 24, (Z) isomer: δ 3.21 (s, 3H), 4.42 (d, 1H, J ) 10
Hz), 6.59 (d, 1H, J ) 8.4 Hz), 6.63-6.69 (m, 1H), 6.97-7.43
(m, 15H).
(E)-ter t-Bu tyl 4-(P yr id in -2-yl)-3-bu ten oa te (26a ). Fol-
lowing 1 h of irradiation, tert-butyl 3-butenoate (25) (2.84 g,
20.0 mmol) and 2 (1.58 g, 10.0 mmol) yielded 3.69 g of a
reddish-orange liquid, the 1H NMR of which indicated the
complete absence of both 25 and 2. Chromatography with
CH2Cl2 resulted in the elution of a pale yellow liquid, subse-
quently identified as the dimer, tert-butyl 3-methyl-4-(tert-
butoxycarbonyl)-4-hexenoate (28a ), 1.07 g (38%). Further
elution with 10% EtOAc-CH2Cl2 afforded 1.68 g of crude
product as a yellow oil. Vacuum distillation at 98-100 °C/
0.10 mm gave 1.34 g (61%) of colorless 26a . For characteriza-
tion, 26a and 28a were converted to the corresponding acids
by refluxing their hexane solutions with 2.0 mol % of p-
toluenesulfonic acid for 5 h. (E)-2-Ethylidene-3-methylglutaric
acid (28b) was thus obtained in 88% yield from 28a as colorless
crystals from CH2Cl2, mp 128-128.5 °C, lit.18c mp 126-127
°C. (E)-4(Pyridin-2-yl)-3-butenoic acid, derived from 26a ,
crystallized from ether as colorless tiny needles, mp 94-95
°C with CO2 evolution. 1H NMR: δ 3.34-3.37 (dd, 2H, J )
7.0 Hz, J ′ ) 1.2 Hz), 6.63 (d, 1H, J ) 16 Hz), 6.85-6.96 (dt,
1H, J ) 16 Hz, J ′ ) 7.0 Hz), 7.21-7.33 (m, 2H), 7.68-7.75
(m, 1H), 8.66-8.68 (m, 1H), 10.70 (br s, 1H). MS(EI): m/ z
(relative intensity) 163 (M+, 34), 119 (34), 117 (100), 91 (51).
Anal. Calcd for C9H9NO2: C, 66.24; H, 5.56; N, 8.58. Found:
C, 66.37; H, 5.61; N, 8.59.
(E)-ter t-Bu t yl 4-P h en yl-3-b u t en oa t e (26b ). From 25
(2.84 g, 20.0 mmol) and 9 (2.04 g, 10.0 mmol) after irradiation
for 1 h was obtained 3.19 g of a yellow liquid, 1H NMR analysis
of which indicated mostly a mixture of monosubstitution
product 26b, disubstitution products 26c and 27, and dimer
28a . MPLC using 1:1 hexane-benzene eluted 0.40 g of a crude
1:1 mixture of 26c and 27 first as a colorless oil which
crystallized on standing. Recrystallization from hexane af-
forded 0.22 g (15%) of a pure 1:1 mixture of 26c and 27 as
colorless crystals, melting range 80-98 °C.29 Compound 25b
was eluted next as a colorless oil, 0.92 g (42%). The eluent
was changed to benzene and dimer 28a , 0.96 g (34%), was
1
obtained as a nearly colorless liquid. The H NMR spectrum
of 26b was consistent with that reported.30 1H NMR for 26c:
δ 1.44 (s, 9H), 3.07 (d, 2H, J ) 7.4 Hz), 6.24 (t, 1H, J ) 7.4
1
Hz), 7.20-7.35 (m, 10H). H NMR for 27: δ 1.44 (s, 9H), 4.35
(d, 1H, J ) 7.7 Hz), 6.46 (d, 1H, J ) 16 Hz), 6.52-6.61 (dd,
1H, J ) 16 Hz, J ′ ) 7.7 Hz), 7.20-7.35 (m, 10H). Anal.
(mixture of 26c:27). Calcd for C20H22O2: C, 81.60; H, 7.53.
Found: C, 81.46; H, 7.57.
Dim eth yl (P yr idin -2-ylm eth yl)ph osph on ate (30a). Pho-
tostimulated reaction of dimethyl methylphosphonate (29a )
(11.17 g, 90.0 mmol) and 2 (4.74 g, 30.0 mmol) after 1 h gave
12.49 g of a dark orange liquid. Unreacted 29a was removed
by distillation at 41-43 °C/1.0 mm leaving 5.66 g of a red-
brown, more viscous oil which was chromatographed with
EtOAc. A trace of 29a was removed first, followed by 2-ami-
nopyridine (4) 0.28 g (10%). The eluent was changed to 10%
MeOH-EtOAc, and 30a was eluted as a pale yellow liquid,
4.13 g (68%), which solidified when refrigerated. Compound
30a was vacuum distilled at 100-103 °C/0.10 mm, lit.31 bp
96-97 °C/0.2 mm. 1H NMR: δ 3.43 (d, 2H, J ) 22 Hz), 3.73
(d, 6H, J ) 11 Hz), 7.22 (t, 1H), 7.38 (d, 1H), 7.65 (t, 1H), 8.54
(d, 1H). MS(EI): m/ z (relative intensity) 201 (M+, 35), 108
(55), 93 (100), 65 (36).
Dim eth yl Ben zylp h osp h on a te (30b) a n d Dim eth yl
Ben zh yd r ylp h osp h on a te (30c). Reaction of 29a (11.17 g,
90.0 mmol) and 9 (6.12 g, 30.0 mmol) under photostimulation
for 1 h gave 12.48 g of a dark yellow liquid which was
fractionally distilled. Excess 29a distilled first at 40-42 °C/
1.0 mm followed by substitution product 30b as 2.57 g of a
colorless liquid, bp 95-100 °C/0.10 mm, lit.32 bp 85 °C/0.15
mm. Continued distillation afforded a third fraction, 1.12 g
of a pale yellow oil, 120-140 °C/0.10 mm, whose 1H NMR
spectrum indicated it to be mostly a 1:3 mixture of 30b and
30c. Separation of this mixture was achieved with MPLC by
gradient elution using CH2Cl2-EtOAc. Disubstitution product
30c was eluted first as a colorless oil, 0.75 g (18%), which
crystallized from ether-hexane, mp 96-97 °C, lit.33 mp 96-
97 °C. Further elution afforded 0.26 g of additional 30b, total
1
yield 2.83 g (47%). The H NMR spectra were consistent with
those reported for 30b32 and 30c.34
Ack n ow led gm en t. We are pleased to acknowledge
generous financial support from the Harvey W. Peters
Research Center for Parkinson’s Disease and Disorders
of the Central Nervous System Foundation and from
the National Science Foundation through Grant
CHE-8513216.
J O962353F
(29) Ester 26c has been reported as a clear oil and characterized
by hydrolysis to the known acid, see: Flechtner, T. W.; Szabo, L. J .;
Koenig, L. J . J . Org. Chem. 1976, 41, 2038.
(30) Bunnage, M. E.; Davies, S. G.; Goodwin, C. J .; Ichihara, O.
Tetrahedron 1994, 50, 3975.
(31) Maruszewska-Wieczorkowska, E.; Michalski, J .; Skowronska,
A. Rocz. Chem. 1956, 30, 1197; Chem. Abstr. 1957, 51, 11347a.
(32) Crenshaw, M. D.; Zimmer, H. J . Org. Chem. 1983, 48, 2782.
(33) Smith, B. E.; Burger, A. J . Am. Chem. Soc. 1953, 75, 5891.
(34) Ogata, Y.; Yamashita, M.; Mizutani, M. Tetrahedron 1974, 30,
3709.