The chemistry of the N-methyl-3-dehydropyridinium ylid

Article abstract of DOI:10.1021/ja9639168

The reaction of atomic carbon with N-methylpyrrole, 5b, at 77 K generates the N-methyl-3-dehydropyridinium ylid, 6b, which can be trapped with added hydrogen halides or CO₂. The addition of CO₂ is strong evidence for the ylid 6b rather than cumulene 7. Deuterium and ¹³C labeling studies demonstrate that 6b rapidly rearranges to the N-methyl-2-dehydropyridinium ylid, 13, by an intermolecular mechanism. Ylid 13 can be trapped with added acids or with O₂ to form 1-methyl-2-pyridone.

Full text of DOI:10.1021/ja9639168

J. Am. Chem. Soc. 1997, 119, 5091-5094
5091
The Chemistry of the N-Methyl-3-dehydropyridinium Ylid
Weitao Pan and Philip B. Shevlin*
Contribution from the Department of Chemistry, Auburn UniVersity, Alabama 36849-5310
ReceiVed NoVember 12, 1996. ReVised Manuscript ReceiVed March 31, 1997X
Abstract: The reaction of atomic carbon with N-methylpyrrole, 5b, at 77 K generates the N-methyl-3-dehydropyr-
idinium ylid, 6b, which can be trapped with added hydrogen halides or CO2. The addition of CO2 is strong evidence
for the ylid 6b rather than cumulene 7. Deuterium and 13C labeling studies demonstrate that 6b rapidly rearranges
to the N-methyl-2-dehydropyridinium ylid, 13, by an intermolecular mechanism. Ylid 13 can be trapped with added
acids or with O2 to form 1-methyl-2-pyridone.
One of the interesting ways in which to introduce strain into
organic molecules is to place cumulative double bonds in small
rings. It has long been recognized that there are several possible
structures for these highly strained cyclic cumulenes.1 In
addition to a nonplanar allene, 1, these include the zwitterionic
species 2 or 3 and biradical 4 (either singlet or triplet). Although
there is no doubt that the accessible 1,2-cycloalkadienes are
at 77K.4,5 The intermediacy of 6b in the 77 K cocondensation
of arc generated carbon with N-methylpyrrole, 5b, is implied
by the fact that addition of methanolic HCl to the cold
condensate generates the N-methylpyridinium ion, 8.
The absence of acidic protons in 5b was expected to increase
the lifetime of 6b and permit the observation of additional
reactions. In particular, reactions with electrophiles other than
the proton were anticipated. In an attempt to observe such
reactions and to probe the nucleophilicity of 6b, methyl iodide
was added to the 77 K matrix of C + 5b. However upon
subsequent addition of HCl, this reaction gave only 8 and none
of the anticipated N-methyl-3-methylpyridinium ion. When an
attempt was made to alkylate 6b as it is formed by condensing
C + 5b + CH3I at 77K, the products were 8 and, quite
surprisingly, the 3-iodo-N-methylpyridinium ion (9a, eq 1) in
a 4:1 ratio. The condensation of other alkyl halides invariably
results in the formation of 3-halo-N-methylpyridinium ions (eq
1).
nonplanar allenes, it is of interest to design systems in which a
planar zwitterionic species such as 2 may be stabilized. Our
recently reported reaction of atomic carbon with pyrrole, 5a,
which appears to proceed via the novel dehydropyridinium ylid
6a, is an example of such a system.2 In this case, we propose
that the aromaticity of the pyrydinium ion favors the planar ylid
6 over the nonplanar cumulene 7.3
However the formation of 6a in the presence of 5a with its
acidic hydrogen, limits the chemical reactivity which can be
observed for 6 to intermolecular proton transfer to generate
pyridine.2 In order to further probe the reactivity of 6 and to
gain additional chemical evidence for its structure, we now
report an investigation of the N-methyl-3-dehydropyridinium
ylid, 6b, in which novel chemistry is observed.
Although we were first tempted to attribute the formation of
9a to a halogen abstraction by a biradical form of 6b,6 we were
forced to revise our ideas when the use of CF3Cl and FCCl3
incorporated both fluorine and chlorine into the 3-position of
the N-methylpyridinium ion. In the case of CF3Cl, the 9b:9c
ratio was 1:1.5, while it was 6.8:1 in the case of FCCl3. Since
abstraction of fluorine by biradicaloid 6b is unlikely, we feel
that the 3-halo-N-methylpyridinium ions result from halometh-
Results and Discussion
Reactions were carried out in the standard carbon arc reactor
in which atomic carbon is generated by striking an arc between
two high purity graphite electrodes and condensed with substrate
(4) For reviews of the chemistry of atomic carbon, see: (a) Skell, P. S.;
Havel J.; McGlinchey, M. J. Acc. Chem. Res. 1973, 6, 97. (b) MacKay, C.
In Carbenes; Moss, R. A., Jones, M., Jr., Eds.; Wiley-Interscience: New
York, 1975; Vol. II, pp l-42. (c) Shevlin, P. B. In ReactiVe Intermediates;
Abramovitch, R. A., Ed.; Plenum Press: New York, 1980; Vol. I, pp 1-36.
(5) Skell, P. S.; Wescott, L. D., Jr.; Golstein, J. P.; Engel, R. R. J. Am.
Chem. Soc. 1965, 87, 2829.
X Abstract published in AdVance ACS Abstracts, May 1, 1997.
(1) Johnson, R. A. Chem. ReV. 1989, 89, 1111.
(2) Emanuel, C. J.; Shevlin, P. B. J. Am. Chem. Soc. 1994, 116, 5991.
(3) Although we have not yet carried out detailed calculations of the
energy surfaces involved in these reactions, we note that geometry
optimization at the MP2/6-31G* level demonstrates that planar 6a and 6b
are energy minima. All attempts to find energy minima for cumulenes 7
gave 6 instead.
(6) For a discussion of the various electronic states possible in cyclic
cumulenes see ref 1.
S0002-7863(96)03916-9 CCC: $14.00 © 1997 American Chemical Society

5092 J. Am. Chem. Soc., Vol. 119, No. 22, 1997
Pan and SheVlin
ylidynes formed directly in the reaction of atomic carbon with
the alkyl halide as shown in eq 2. We have reported the rapid
addition of CF and CCl to double bonds to generate R-halo-
cyclopropyl radicals,7ab and work in our laboratory indicates
that CI is generated and reacts in the same manner.7c In this
These interesting results suggest that initially formed 6b
rearranges to the 2-dehydropyridinium ylid 13 in a reaction in
which the thermodynamic driving force may be the fact that
the negative charge moves closer to the electronegative nitrogen.
There are a number of mechanistic possibilities for this
interesting rearrangement. An intramolecular rearrangement by
a simple 1,2-hydrogen migration seems unlikely to occur on a
77 K surface. Rearrangements such as this in which a group
migrates to an adjacent filled orbital generally have a high barrier
due to the fact that the four-electron system requires that an
case the exothermicity of the addition reaction to form R-ha-
locyclopropyl radical 10 drives the ring opening to 11 which
subsequently generates 9 in an electron transfer. In an analogous
reaction, we have observed the reaction of CF with cyclopen-
tadiene to give fluorobenzene and with pyrrole to generate
3-fluoropyridine.7d Although the above observations nicely
rationalize the formation of 9, they do not fully explain the
failure of 6b to undergo alkylation.
In our search for nucleophilic reactions of 6b, we next turned
to carbon dioxide, an electrophile not expected to generate
another highly energetic species upon reaction with carbon.
Cocondensation of 5b + C + CO2 at 77 K followed by addition
of methanolic HCl generates 8 and N-methylpyridinium 3-car-
boxylic acid 128 in a 1.2:1 ratio (eq 3). The observation of 12
electron pair occupy a high energy (usually) antibonding orbital.
However, this rearrangement could proceed by an intramolecular
skeletal rearrangement of the type that occur in the phenyl9 and
pyridyl10 carbene rearrangements. Equation 5 demonstrates that
this skeletal rearrangement would reveal itself upon addition
of 13C atoms to 5b. Thus, 13C enriched carbon vapor was
reacted with 5b and methanolic HCl added to the low temper-
ature matrix. However, this reaction gave 8 with 87% of its
excess 13C in the 3-position ruling out the intramolecular
rearrangement in eq 5 as a major process.
Since both experimental evidence and experience appear to
eliminate an intramolecular mechanism as the major pathway
in the rearrangement of 6b to 13, we turned our attention to
intermolecular mechanisms. Such mechanisms would most
likely involve simple acid-base chemistry of the type illustrated
in eq 6. In this case, the acid is probably the N-methylpyrrole,
provides strong chemical evidence in favor of ylid structure 6
rather than cumulene 7 which would not be expected to react
with CO2. However, 12 is not generated when CO2 is added
to the 77 K matrix formed upon reaction of C with 5b (eq 3).
In order to attempt to understand why 6b could seemingly
be trapped by methanolic HCl added to the low temperature
matrix of C + 5b after reaction but not by carbon dioxide added
to this same matrix, we have used CH3OD as the trapping agent.
Surprisingly, addition of CH3OD to the low temperature matrix
2
after reaction results in the formation of 8 whose H NMR
spectrum shows deuterium only in the 2-position. However
when the CH3OD is added during the condensation, the
deuterium is found exclusively at the 3-position in 8 (eq 4). A
third experiment, in which a layer of CH3OD is first condensed
followed by cocondensation of C + 5b and the process repeated
many times, leads to 8 with D in both the 2- and 3-positions in
a 3:1 ratio.
5b. An alternative in which two molecules of 6b find each
other on the low temperature matrix seems unlikely when one
considers that highly reactive 6b is generated in the presence
of a large excess of 5b.
If 5b does participate in the rearrangement of 6b to 13 via
acid-base chemistry, this fact may be uncovered by appropriate
“crossover” experiments using deuterium labeled 5b. We have
(7) (a) LaFrancois, C. J.; Shevlin, P. B. J. Am. Chem. Soc. 1994, 116,
9405. (b) Sztyrbicka, R.; Rahman, M. M.; D’Aunoy, M. E.; Shevlin, P. B.
J. Am. Chem. Soc. 1990, 112, 6712. (c) LaFrancois, C., J. Ph. D. Thesis,
Auburn University, 1996. (d) Sztyrbicka, R. Ph. D. Thesis, Auburn
University, 1992.
(9) Gaspar, P. P.; Hsu, J.-P.; Chari, S.; Jones, M., Jr. Tetrhedron 1985,
41, 1479 and references therein.
(10) (a) Wentrup, C. Top. Curr. Chem. 1976, 62, 173. (b) Wentrup, C.
ReactiVe Molecules; Wiley-Interscience: New York, 1984.
(8) Spa¨th, B. Chem. Ber. 1944, 77, 362.

Chemistry of the N-Methyl-3-dehydropyridinium Ylid
J. Am. Chem. Soc., Vol. 119, No. 22, 1997 5093
carried out two such experiments. In the first, N-trideutero-
methylpyrrole, 14, was reacted with C atoms, and methanolic
HCl was subsequently added to generate N-trideuteromethyl-
pyridinuim ion, 15, with only very minor traces of deuterium
on the ring carbons (eq 7). Thus the methyl group in 5b is not
a significant source of acidic protons in the intermolecular
rearrangement.
Although we cannot trap 13, with CO2, we have found that
it can be trapped as 1-methyl-2-pyridone, 16, by addition of
oxygen to the matrix resulting from the reaction of C + 5b.
Although the mechanistic details of this trapping reaction are
not clear, it seems likely that it proceeds Via stabilized carbonyl
oxide, 17.
Conclusions
However a second experiment in which a 1.2:1 mixture of
5b and 5b-1,2,3,4-d4 is reacted with C atoms reveals, after
addition of methanolic HCl, 8-d0, 8-d1, 8-d3, 8-d4, and 8-d5 in
a 6.8:3.5:1:5.1:2.2 ratio as determined by electrospray ms. If
the rearrangement of 6b to 13 were intramolecular, we would
expect only 8-d0 and 8-d4 after addition of HCl. The fact that
appreciable amounts of 8-d1, 8-d3, and 8-d5 are formed
implicates an intermolecular mechanism involving acid-base
equilibria in which 6b is protonated by 5b and a subsequent
proton transfer generates 13 as shown in eq 8. The 8-d5 formed
in this reaction results from a D transfer to 8-d4 which fails to
undergo a second D transfer to give 13-d4 and remains as 8-d5
(eq 8). It is interesting that the acidic hydrogens which protonate
6b are the vinyl hydrogens in 5b rather than the supposedly
more acidic methyl hydrogens.11 However, it has been observed
that the C2-position and not the methyl in 5b is lithiated upon
treatment with alkyllithiums.12 In the present case, it appears
that a kinetically controlled transfer of a vinylic proton in 5b
to 6b occurs on the matrix at low temperature.
These investigations provide interesting examples of the
ability of atomic carbon to channel its high energy into the
production of novel reactive intermediates. In particular, we
have demonstrated the generation of highly reactive 3-dehy-
dropyridinium ylids whose chemistry indicates an ylid rather
than a cumulene structure. These 3-dehydropyridinium ylids
are rather elusive species which, if not trapped immediately,
will undergo an intermolecular rearrangement to the 2-dehy-
dropyridinium ylid. The reaction conditions, which permit these
species to be generated at low temperatures in the absence of
complicating factors such as counter ions and strong acids and
bases, allow mechanistic studies which would not be possible
with these intermediates generated by other methods.
Experimental Section
General Methods. Unless otherwise specified, all reagents were
purchased from commercial suppliers and used without further purifica-
tion. NMR spectra were recorded on a Bruker 250 MHz spectrometer.
Mass spectra were determined on a TRIOS 2000 mass spectrometer
operating in the electrospray mode. HPLC was performed using a
Shimadzu 10-A instrument.
General Procedure for the Reaction of Carbon Vapor with
Substrate. The carbon arc reactor was modeled after that described
by Skell and co-workers.5 Carbon was vaporized by striking an
intermittent arc between two graphite rods attached to water-cooled
brass electrodes and condensed on the walls of the reactor at 77 K at
5 × 10-2 Torr. The quantities of carbon given below are calculated
from the loss of weight of the electrodes. Since most of the carbon is
lost in macroscopic pieces, yields based on carbon are not meaningful.
Carbon was either cocondensed with degassed substrate or condensed
on a layer of substrate. In the latter case, a number of condensation
cycles were generally run. At the conclusion of the reaction, the liquid
nitrogen bath was removed, and the reactor was allowed to warm to
room temperature. The reactants and some volatile products were
pumped to a trap at 77 K. The residue in the reactor bottom was first
extracted twice with 20 mL of CH2Cl2 containing several drops of
concentrated HCl and further extracted three times with 20 mL of
MeOH containing several drops of concentrated HCl. The combined
CH2Cl2 extracts were generally found to contain mainly starting
material. The combined MeOH extract was filtered, the solvent was
evaporated, and the residue was analyzed by NMR (1H, 2H, and 13C as
appropriate) and electrospray ms. With the exception of 3-iodo-N-
methylpyridinium chloride, 9a, all products were identified by compar-
ing their spectra with those of authentic samples synthesized by
literature procedures.14
It is now clear that attempts to observe the chemistry of 6b
are complicated by its rapid intermolecular rearrangement to
13. Only electrophiles added during the formation of 6b whose
rate of trapping is competitive with rearrangement, such as CO2
and acids, may be used as traps. The fact that 13 can be
protonated to a pyridinium ion but fails to undergo nucleophilic
addition to alkyl halides or carbon dioxide may reflect the
enhanced stability of 13 over 6b. In this connection, we note
that decarboxylation of N-methylpyridinium 2-carboxylic acid
is 500 times faster than that of 12 at 196°.13
The Reaction of C with N-Methylpyrrole, 5b. Carbon (1.3 g,
108.3 mmol) and 5b (3.00 g, 37.0 mmol) were cocondensed at 77K.
The MeOH-HCl extract contained N-methylpyridinium chloride, 8
(0.19 mmol as determined by 1H NMR). 8: 1H NMR (250 MHz, D2O)
δ 4.27 (s, 3H), 7.92 (bt, J ) 7.5 Hz, 2H), 8.41 (t, J ) 7.8 Hz, 1H),
(11) DePuy, C. H.; Kass, S. H.; Bean, G. P. J. Org. Chem. 1988, 53,
1427.
(14) Zoltewicz, J. A.; Cross, R. E. J. Chem. Soc., Perkin Trans. II 1974,
(12) (a) Shirley, D. A.; Gross, B. H.; Roussel, P. A. J. Org. Chem. 1955,
20, 225. (b) Chadwick, D. J.; Cliffe, I. A. J. Chem. Soc., Perkin Trans. I
1979, 2845.
1363.
(15) The deuterated N-methylpyrroles were prepared using the procedure
for the unlabeled compound: Santaniello, E.; Farachi, C.; Ponti, F. Synthesis
1979, 617.
(13) Hake, P.; Manteco´n, J. J. Am. Chem. Soc. 1964, 86, 5230.

5094 J. Am. Chem. Soc., Vol. 119, No. 22, 1997
Pan and SheVlin
8.66 (d, J ) 6.0 Hz, 2H); 13C NMR (62.5 MHz, D2O) δ 48.50, 127.97,
145.00, 145.27; ES-MS (E+) m/z (rel intensity), 94.3(100), 95.3(8.9).
The Reaction of C with N-Methylpyrrole, 5b, and CH3I. Carbon
(1.2 g, 100 mmol), 5b (3.66 g, 47.5 mmol), and CH3I (6.6 g 46.5 mmol)
were cocondensed at 77 K. The MeOH-HCl extract contained 8 and
3-iodo-N-methylpyridinium chloride, 9a, in a 4:1 ratio as determined
13C is incorporated into the methyl of 8 and conclude that 87% of the
excess 13C is on C3 with 12.5% on C2 and 0.5% on C1.
The Reaction of C with 5b and CH3OD. Carbon (0.75 g, 62.5
mmol), 5b (1.7 g, 21 mmol), and CH3OD (7.8 g, 236 mmol) were
cocondensed at 77 K. To the cold condensate was added 2 mL of 1 N
HCl-30 mL of MeOH. Analysis of 8 by 2H NMR revealed deuterium
only in the 3 position, 1H NMR showed H2:H3:H4 in a 2.0:1.5:1 ratio.
ES-MS m/e (rel intensity) 94.3 (31), 95.3 (100), 96.3 (17).
1
by H NMR. The 3-iodo-N-methylpyridinium chloride (1.6 × 10-5
m) was separated from 8 by RP-HPLC. 9a: 1H NMR (250 MHz, D2O)
δ 4.22 (s, 3H), 7.68 (dd, J ) 6.2 Hz, 8.4 Hz, 1H), 8.75 (d, J ) 8.4 Hz,
1H), 8.69 (d, J ) 6.2 Hz, 1H), 9.06 (s, 1H); 13C NMR (62.5 MHz,
D2O) δ 49.3, 128.3, 138.0, 144.2, 151.1,153.8; ES-MS (E+) m/z (rel
intensity), 220.1(100), 221.1(7.7), ES-MS (E-) m/z, 127.
The Reaction of C with 5b followed by CH3OD. Carbon (1.40 g
116.7 mmol), and 5b (2.5 g, 21 mmol) were cocondensed at 77 K. To
the cold condensate was added CH3OD (8.8 g, 267 mmol). After
warming to room temperature was added 2 mL of 1 N HCl-20 mL of
The Reaction of C with 5b and CH3Cl. Carbon (1.0 g, 83.3 mmol),
5b (2.31 g, 28.2 mmol) and CH3Cl(2.0 g, 38.9 mmol) were cocondensed
at 77 K. The MeOH-HCl extract contained 8 and 3-chloro-N-
2
CH3OH. Analysis of 8 by H NMR revealed deuterium only in the 2
position. 1H NMR showed H2:H3:H4 in a 0.89:1.98:1 Ratio. ES-MS
m/e (rel intensity) 94.3 (100), 95.3 (91), 96.3 (28).
1
methylpyridinium chloride, 9b, in a 3.3:1 ratio as determined by H
The Reaction of C with 5b and CH3OD Layered on the Reactor
Bottom. A layer of CH3OD was condensed on the cold reactor bottom.
Carbon vapor and 5b were then cocondensed in ten short bursts of the
arc and the process was repeated until the supply of 5b was exhausted.
A total of 1.0 g (83.3 mmol) of carbon, 2.5 g (30.9 mmol) of 5b, and
3 mL (80.7 mmol) of CH3OD was added at 77 K. To the cold reactor
NMR. 9b: 1H NMR (250 MHz, D2O) δ 4.29 (s, 3H), 7.96 (dd, J )
8.2, 6 Hz, 1H), 8.48 (dt, J ) 8.2, 0.6 Hz, 1H), 8.68 (d, J ) 6 Hz, 1H),
8.94 (s, 1H); 13C NMR (62.5 MHz, D2O) δ 48.6, 128.18, 134.71,
143.36, 141.244, 144.9; ES-MS (E+) m/z (rel intensity), 128.3 (100),
129.3 (8.6), 130.3 (44.7).
The Reaction of C with 5b and CFCl3. Carbon (1.1 g, 91.7 mmol),
5b (2.4 g, 29.6 mmol), and CFCl3 (50.0 mmol) were cocondensed at
77 K. The MeOH-HCl extract contained 8, 9b, and 3-fluoro-N-
methylpyridinium chloride, 9c, in a 2.8:6.8:1 ratio as determined by
1H NMR. 9c: 1H NMR (250 MHz, D2O) δ 4.32 (s, 3H), 7.97-8.04
(m, 1H), 8.32 (bt, J ) 5.5 Hz, 1H), 8.62 (d, J ) 6.3 Hz, 1H), 8.94 (bs,
1H); 13C (62.5 MHz, D2O) δ 48.7, 129.2 (d, J ) 8 Hz), 132.85 (d, J
) 19 Hz), 135.15 (dt, J ) 18 Hz), 142.19 (bs), 160.1 (d, J ) 250 Hz);
ES-MS (E+) m/z (rel intensity), 112.3 (100), 113.3(9.0).
2
bottom was added 30 mL of 1 N HCl. Analysis of 8 by H NMR
revealed deuterium in the 2 and 3 positions in a 3:1 ratio. 1H NMR
showed H2:H3:H4 in a 0.91:1.73:1 ratio. ES-MS m/e (rel intensity)
94.3 (100), 95.3 (91), 96.3 (28).
Reaction of C with N-Tridueteromethylpyrrole, 14. Carbon (0.66
g, 55 mmol) and 1415 (1.1 g, 13 mmol) were cocondensed at 77 K. To
the cold condensate was added 2 mL of 1 N HCl-20 mL of MeOH.
Analysis of 8 by 2H NMR revealed deuterium only on the methyl group.
ES-MS m/e (rel intensity) 97.3 (100), 98.3 (19.5).
The Reaction of C with 5b and CClF3. Carbon (1.4 g, 116.7
mmol), 5b (2.6 gram, 32.1 mmol), and CClF3 (50.0 mmol) were
cocondensed at 77 K. The MeOH-HCl extract contained 8, 9b, and
The Reaction of C with a Mixture of 5b and 5b-1,2,3,4-d4. Carbon
(0.26 g, 21.6 mmol), 5b (0.35 g, 4.3 mmol), and 5b-1,2,3,4-d415 (0.30
g, 3.5 mmol) were cocondensed at 77 K. To the cold condensate was
1
9c in a 2:1:1.5 ratio as determined by H NMR.
2
added 1 mL of 12 N HCl-10 mL of MeOH. Analysis of 8 by H
Reaction of 5b with CO2. Carbon (0.75 g, 62.5 mmol), 5b (1.7 g,
21 mmol), and CO2 (7.8 g, 236 mmol) were cocondensed at 77 K. The
MeOH-HCl extract contained 8 and 1-methyl-3-pyridinium carbox-
NMR revealed deuterium in the 2, 3, and 4 positions in a 1.88:1.77:1
ratio. The 1H NMR showed H2:H4:H3: N-Me in a 0.36:0.367:0.188:1
ratio. ES-MS m/e (rel intensity) 94.3 (100), 95.3 (51.5), 96.3 (7.7),
97.3 (14.7),98.3 (74.7), 99.3 (31.8).
1
ylate, 12, in a 1.2:1 ratio as determined by H NMR. 12: 1H NMR
(250 MHz, D2O) δ 4.35 (s, 3H), 8.06 (t, J ) 7 Hz, 1H), 8.94-8.85
(bm, 2H), 9.28 (s, 1H); 13C NMR (62.5 MHz, D2O) δ 48.9, 128.3,
134.7, 145.5, 146.6, 147.3, 166.5; ES-MS m/e (rel intensity) 138.3 (100),
139.3 (12).
Reaction of 13 with O2. Carbon (0.65 g, 54.2 mmol) and 5b (2.63
g, 32.5 mmol) were cocondensed at 77 K. To the cold condensate
was added 5 Torr O2. The ether extract of the reactor bottom was
purified by flash silica gel column chromatography (CH2Cl2-Et2O-
MeOH) and found to contain 1-methyl-2-pyridone, 16, which was
identified by comparison of its 1H NMR with that of an authentic
sample. 16: 1H NMR (250 MHz, D2O) δ 3.55 (s, 3H), 6.18 (dt, J )
6.75, 1.34 Hz, 1H), 6.54 (dt, J ) 8.75, 0.92 Hz, 1H), 7.31-7.38 (m,
2H); 13C NMR (62.5 MHz, D2O) δ 37.56, 105.90, 120.58, 138. 38,
139.56, 163.12.
Reaction of 13C Enriched Carbon Vapor with 5b. A 0.125 × 1
in. hole was drilled in a 0.25 in. diameter high purity graphite rod.
The hole in this 3.0 g rod was filled with 0.129 g amorphous carbon-
13 (99.4% 13C). The rod was placed in the reactor and resistively heated
under vacuum at a current of 100 A by keeping it in touch with an
unlabeled graphite rod for 2 min. The heating cycle was repeated five
times. The rod was placed in the reactor, and an arc was struck between
it and an unlabeled rod. Enriched carbon vapor (1.01 g, 86.2 mmol)
was reacted with 0.8 g (9.8 mmol) 5b as described above. The 13C
NMR spectrum of 8 showed peaks at δ ) 48.5 (Me), 127.97 (C3), 145
(C4), 145.27 (C2) in a 1:30.7:2.1:3.8 intensity ratio. ES-MS m/e (rel
intensity) 94.3 (100), 95.3 (24). The spectrum of unlabeled 8 showed
these same peaks in a 1:5.6:2.3:2.0 intensity ratio. We assume that no
Acknowledgment. Support of this work through National
Science Foundation Grant CHE-9508570 is gratefully
acknowledged.
JA9639168