Direct dehydrative n-pyridinylation of amides

Article abstract of DOI:10.1021/jo802355d

Electrophilic activation of secondary amides with trifluo-romethanesulfonic anhydride in the presence of 2-fluoropy-ridine followed by introduction of a pyridine N-oxide derivative and warming affords the corresponding N-pyridi-nyl tertiary amide derivati

Full text of DOI:10.1021/jo802355d

TABLE 1. Optimization of Reaction Conditionsa
Direct Dehydrative N-Pyridinylation of Amides
Jonathan William Medley and Mohammad Movassaghi*
Department of Chemistry, Massachusetts Institute of
Technology, Cambridge, Massachusetts 02139
moVassag@mit.edu
entry
base additive
2-Cl-pyridine
2-Cl-pyridine
none
pyridine
Et3N
2-Br-pyridine
3-Cl-pyridine
ethyl nicotinate
3-Br-pyridine
2,6-lutidine
2-F-pyridine
2-F-pyridine
2-F-pyridine
DIPEA
base equiv N-oxide x equiv yield (%)b
ReceiVed October 21, 2008
1
2
3
4
5
6
7
8
1.2
1.2
1.1
2.0
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
77
81
16
10
0
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
2.0
5.0
1.2
1.2
1.5
58
74
73
72
55
99
88
76
48
15
72
9
10
11
12
13
14
15
16
Electrophilic activation of secondary amides with trifluo-
romethanesulfonic anhydride in the presence of 2-fluoropy-
ridine followed by introduction of a pyridine N-oxide
derivative and warming affords the corresponding N-pyridi-
nyl tertiary amide derivatives. A mechanism supported by
in situ monitoring and deuterium labeling experiments is
discussed.
2,6-dichloropyridine
Hendrickson reagentc
a Conditions: Amide 1a, Tf2O (1.1 equiv), base additive, CH2Cl2,
-78f0 °C; isoquinoline N-oxide, 0f23 °C, 3 h. b Isolated yield.
c Hendrickson reagent ((Ph3P+)2O·2TfO-) was prepared (ref 5) and used
in place of Tf2O without base additive.
N-vinyl amides to enable nucleophilic addition and annulation.
We envisioned that these conditions would serve well for a
modified Abramovitch reaction. Amide 1a and isoquinoline
N-oxide (2a) served as substrates for our early exploration of
this chemistry (Table 1). Interestingly, the use of 2-fluoropy-
ridine (2-FPyr, 1.2 equiv) as a base additive afforded a
significant improvement in the reaction yield as compared to
2-ClPyr, furnishing amide 3aa in 99% yield (compare entries
1 and 11, Table 1). More nucleophilic and stronger base
additives generally gave poorer yields as compared to base
additives with attenuated nucleophilicity and basicity. These
observations suggest that optimal conditions provide a balance
between the need for a base additive to promote electrophilic
activation of the amide substrate while avoiding nucleophilic
inhibition of this reaction. Consistent with our earlier findings,7,8
the presence of the optimal base additive in excess or its absence
led to a marked decrease in the yield of the desired product
(entries 3, 12, and 13, Table 1). The use of the Hendrickson
reagent ((Ph3P+)2O·2TfO-)5 as the activating agent for this
dehydrative N-pyridinylation reaction proved less effective as
compared to the optimal conditions described above (entry 16,
Table 1). The overall yield can be improved by use of excess
2a due to competitive N-oxide decomposition (entry 2, Table
1, vide infra).
Pyridinylated, isoquinolinylated, and quinolinylated amides
are key structural motifs in a wide variety of pharmaceutical
compounds and natural products.1 Abramovitch and co-workers
reported a fascinating methodology for the synthesis of N-
pyridinyl amide derivatives via the nucleophilic addition of
heteroaromatic N-oxides to imidoyl chlorides followed by a
thermal rearrangement.2 Recently, alternative conditions have
been reported for the in situ activation of primary3 and
secondary4 amides for nucleophilic addition of N-oxides to
furnish the N-pyridinylated amides. The development of mild
reaction conditions and expansion of the substrate scope for
this transformation is of interest given the importance of the
products. Herein we describe the N-pyridinylation, N-isoquino-
linylation, and N-quinolinylation of various secondary amides
and discuss a plausible mechanism supported by deuterium
labeling and in situ monitoring experiments.
We have reported the use of 2-chloropyridine (2-ClPyr) with
trifluoromethanesulfonic anhydride6 (Tf2O) as a versatile reagent
combination for the synthesis of pyrimidine7 and pyridine
derivatives.8 These methodologies provide the desired azahet-
erocycles via electrophilic activation of secondary N-aryl or
(1) (a) Jones, G. In ComprehensiVe Heterocyclic Chemistry II; Katritzky,
A. R., Rees, C. W., Scriven, E. F. V., McKillop, A., Eds.; Pergamon: Oxford,
1996; Vol. 5, p 245. (b) Henry, G. D. Tetrahedron 2004, 60, 6043.
(2) (a) Abramovitch, R. A.; Singer, G. M. J. Am. Chem. Soc. 1969, 91, 5672.
(b) Abramovitch, R. A.; Singer, G. M. J. Org. Chem. 1974, 39, 1795. (c)
Abramovitch, R. A.; Rogers, R. B. J. Org. Chem. 1974, 39, 1802. (d)
Abramovitch, R. A.; Rogers, R. B.; Singer, G. M. J. Org. Chem. 1975, 40, 41.
(e) Abramovitch, R. A.; Rogers, R. B. Tetrahedron Lett. 1971, 22, 1951.
(3) Couturier, M.; Caron, L.; Tumidajksi, S.; Jones, K.; White, T. D. Org.
Lett. 2006, 8, 1929.
(6) For elegant prior studies on amide activation, see: (a) Charette, A. B.;
Grenon, M. Can. J. Chem. 2001, 79, 1694. Review: (b) Baraznenok, I. L.;
Nenajdenko, V. G.; Balenkova, E. S. Tetrahedron 2000, 56, 3077. For
2-chloropyridine as a base additive, see: (c) Myers, A. G.; Tom, N. J.; Fraley,
M. E.; Cohen, S. B.; Madar, D. J. J. Am. Chem. Soc. 1997, 119, 6072.
(7) (a) Movassaghi, M.; Hill, M. D. J. Am. Chem. Soc. 2006, 128, 14254.
(b) Movassaghi, M.; Hill, M. D. Nat. Protoc. 2007, 2, 2018.
(4) Manley, P. J.; Bilodeau, M. T. Org. Lett. 2002, 4, 3127.
(5) Hendrickson, J. B.; Hussoin, M. D. J. Org. Chem. 1987, 52, 4137.
(8) Movassaghi, M.; Hill, M.; Ahmad, O. K. J. Am. Chem. Soc. 2007, 129,
10096.
10.1021/jo802355d CCC: $40.75
Published on Web 12/29/2008
2009 American Chemical Society
J. Org. Chem. 2009, 74, 1341–1344 1341

TABLE 2. Direct Dehydrative N-Pyridinylation of Amidesa
to 2a) under the electrophilic activation reaction conditions.10
Interestingly, in reactions employing N-oxide 2c, amides that
exhibited high reactivity in isoquinolinylation reactions gave
higher yields when 2,6-lutidine4 was used in place of 2-FPyr
as the base additive (products 3cc, 3dc, and 3hc, Table 2), likely
owing to slight suppression of N-oxide decomposition. However,
the use of 2,6-lutidine in place of 2-fluoropyridine with less
reactive N-aryl benzamides (1b and 1e, Table 2) resulted in
significantly lower yields of the desired products.
Attempts to N-quinolinylate amide 1a under our standard
conditions gave no detectable amount of the desired product
3ab. This is consistent with poor nucleophilic addition of 2b to
the activated intermediate, allowing a competitive decomposition
of amide 1a.11 Given that nucleophilic base additives inhibit
the desired reaction (Table 1), we conjectured that the N-
pyridinylated products formed may also play an inhibitory role.
Activation of amide 1f followed by sequential addition of
N-pyridinylated amide 3fc (1.00 equiv) and pyridine N-oxide
(2c) gave a low 33% yield of the desired amide 3fc,12 which is
less than half the expected yield. This suggests that product
inhibition can be significant in these pyridinylation reactions.
However, activation of amide 1f under optimized conditions,
followed by sequential addition of product 3fa (1.00 equiv) and
isoquinoline N-oxide (2a), gave 92% yield of the desired
N-isoquinolinylated amide 3fa,12 indicating no significant
product inhibition in this reaction. Furthermore, sequential
activation of an enantiomerically enriched amide 1h8 under
optimized conditions followed by introduction of isoquinoline
N-oxide (2a) provided the optically active N-isoquinolinylated
amide (+)-3ha without erosion of optical activity (eq 1).
Interestingly, while N-pyridinylation of amides was generally
less efficient as compared to N-isoquinolinylation, the use of
both electron-rich and electron-poor 4-substituted pyridine
N-oxides 2d and 2e, respectively, gave modest yield of the
desired products (eqs 2 and 3). The successful N-pyridinylation
of amide 1d with 4-nitropyridine N-oxide (2e) is notable, as its
use as a nucleophile for the Abramovitch reaction was previ-
ously reported to be unsuccessful,2a,b suggesting greater elec-
trophilicity of the intermediate under the conditions described
here.
a Isolated yields of products 3xy. Average of two experiments.
Conditions: Amide 1x (1 equiv), Tf2O (1.1 equiv), 2-FPyr (1.2 equiv),
CH2Cl2, -78f0 °C; N-oxide 2y (2.0 equiv), 0f23 °C, 4 h. b N-Oxide
2a (1.1 equiv),
d 2,6-Lutidine used as base. e Low yield due to product decomposition.
3 4 h.
h. c Decomposition of 1a observed over
We next examined the optimal conditions with a range of
amide substrates with three representative heteroaromatic N-
oxides (Table 2). Isoquinolinylation of amides under our
conditions is highly efficient, giving good to excellent yields
in all cases examined. Isoquinolinylation of N-alkyl benzamides
(3da, 3fa, and 3ga, Table 2), in addition to N-aryl and N-vinyl
amides (3ca and 3ha, Table 2), was achieved in high yields
under our standard reaction conditions. Electron-rich benzamides
notwithstanding (3aa, Table 2), the least efficient substrates in
this series were N-aryl benzamides (3ba and 3ea, Table 2). In
all cases, completely regioselective isoquinolinylations pro-
ceeded at the 1-position of the isoquinoline ring.2-4
The use of quinoline N-oxide (2b, Table 2) and pyridine
N-oxide (2c) as substrates also gave completely regioselective
acylamination, however, with reduced overall efficiency for the
formation of the desired products. This is due in part to the
faster decomposition2b,e,9 of N-oxides 2b and 2c (as compared
To gain better understanding of the intermediates involved
in this transformation, a series of in situ IR and NMR monitoring
(9) Yin, J.; Xiang, B.; Huffman, M. A.; Raab, C. E.; Davies, I. W. J. Org.
Chem. 2007, 72, 4554.
(10) This observation is supported by control experiments involving the
exposure of N-oxides 2a-c to trifluoromethanesulfonic anhydride and 2-fluo-
ropyridine under standard reaction conditions.
(11) Control experiments revealed that activation of amide 1a in the absence
of a competent nucleophile led to decomposition.
(12) Please see the Supporting Information for details.
1342 J. Org. Chem. Vol. 74, No. 3, 2009

SCHEME 1
experiments were performed. The conversion of amide 1d to
N-isoquinolinylated amide 3da under optimized conditions was
monitored by in situ IR analysis. Addition of Tf2O to a mixture
of amide 1d and 2-FPyr resulted in complete consumption of
the amide absorption band (1668 cm-1) and appearance of a
persistent absorption at 2370 cm-1, suggestive of a nitrilium
ion intermediate.2b,13 Addition of isoquinoline N-oxide (2a)
resulted in immediate disappearance of the absorption at 2370
cm-1 and appearance of a persistent absorption at 1691 cm-1
,
These differences in amide reactivity were further substantiated
by in situ H NMR monitoring of the electrophilic activation
1
which was due to the protonated product 3da. Interestingly, the
activation of N-(4-methoxyphenyl)benzamide (1i) with the
reagent combination of 2-ClPyr and Tf2O did not lead to an
observable absorption corresponding to a nitrilium ion, but
step. Interestingly, amides that demonstrated the least propensity
to form a nitrilium ion upon activation under the optimal reaction
conditions also gave the lowest yields in reactions with
isoquinoline N-oxide (e.g., entry 3ba, Table 2). Furthermore,
reduced yield of the desired product upon addition of excess
base additive (or use of nucleophilic bases, Table 1) is consistent
with the observed disappearance of the nitrilium species during
in situ monitoring experiments.
instead gave rise to a persistent absorption at 1600 cm-1
,
suggestive of an amidinium intermediate.8 These observations
suggest that while electrophilic activation of 1d using 2-FPyr
results in 5d (Scheme 1), similar activation of 1i using 2-ClPyr
leads to predominant formation of 4i rather than 5i.
To determine the degree to which the formation of a nitrilium
ion depends on the nature of the base additive and the amide
structure itself, a series of in situ IR monitoring experiments
were carried out.12 For comparison, while activation of N-alkyl
benzamide 1d under optimal conditions resulted in an absorption
suggesting a nitrilium ion (2370 cm-1),14 the activation of N-aryl
benzamides 1b and 1i under the same conditions led to no
detection of an IR absorption consistent with a nitrilium ion,
but instead resulted in the appearance of an IR absorption
suggesting an amidinium ion (1621 cm-1 in both cases).8
However, Tf2O activation of the electron-rich N-aryl benzamide
1a in the presence of either 2-FPyr or 2-ClPyr (1.2 equiv) indeed
resulted in an absorption at 2312 cm-1, suggesting a persistent
nitrilium ion intermediate. Interestingly, addition of extra
equivalents of 2-ClPyr resulted in complete disappearance of
this absorption band and appearance of a persistent absorption
at 1594 cm-1, consistent with the formation of the previously
observed amidinium ion.8 Even the electron-poor N-alkyl
Additional mechanistic insight was obtained using deuterated
substrates 2a-d2, 2c-d2, and 2c-d1 (eqs 4 and 5). Electrophilic
activation of N-alkyl benzamide 1f under optimal conditions
followed by introduction of excess17 isoquinoline N-oxide (2a)
and 1,3-dideuteroisoquinoline N-oxide (2a-d2) provided a
mixture of N-isoquinolinylated products 3fa and 3fa-d1 corre-
sponding to kH/kD ) 1.0 in 86% combined yield (eq 4).12 The
same outcome was observed in a similar experiment using
excess 2c-d2 and 2c, resulting in a mixture of the N-pyridinylated
products 3fc and 3fc-d1 corresponding to kH/kD ) 1.0 in a
combined yield of 59% (eq 4). As another mechanistic probe,
activation of 1f under optimal conditions and the use of excess
2-deuteropyridine N-oxide (2a-d1) provided the expected N-
pyridinylated amide 3fc as a mixture of nondeuterated and
monodeuterated derivatives (eq 5).12 Importantly, the ratio of
3fc-d0 and 3fc-d1 was found to be 1.0:2.0, reflecting an
observable primary kinetic isotope effect (kH/kD ) 2.0).18 These
observations suggest that addition19 of the imidate nitrogen onto
the pyridinium ring is reversible, whereas nucleophilic addition
of the N-oxide 2 to the nitrilium ion 5 (or another electrophilic
variant) is irreversible (Scheme 1).
benzamide 1f resulted in a lasting nitrilium ion (2354 cm-1
)
upon electrophilic activation in the presence of either 2-FPyr
or 2-ClPyr (1.2 equiv), although the presence of excess 2-ClPyr
resulted in disappearance of the absorption at 2354 cm-1 and
the appearance of an absorption at 1609 cm-1.15 These observa-
tions suggest that activation of N-alkyl amides under these
conditions more readily results in persistent nitrilium ion
formation, while N-aryl amides show reluctance to form the
corresponding nitrilium ion, likely owing to the inductive effect
of the nitrogen substituent.16 Only the particularly electron-rich
N-aryl benzamide 1a resulted in any observable nitrilium ion,
perhaps due to greater stabilization by resonance contribution.
We describe a direct method for the dehydrative N-pyridi-
nylation of amides under electrophilic activation by the reagent
(17) Excess N-oxides were used to minimize any effect due to change in
concentration during the reaction. Deuterated pyridine N-oxides were particularly
prone to decomposition as compared to deuterated isoquinoline N-oxide
derivatives.
(18) (a) Zollinger, H. In AdVances in Physcial Organic Chemistry; Gold,
V., Ed.; Academic Press: London, 1964; Vol. 2, pp 163-196. (b) Berliner, E.
In Progress in Physical Organic Chemistry; Cohen, S. G., Streitweiser, A., Taft,
R. W., Eds.; Interscience: New York, 1964; Vol. 2, pp 253-321. (c) Jackson,
A. H.; Lynch, P. P. J. Chem. Soc., Perkin Trans. 2 1987, 1483. (d) Maresh,
J. J.; Giddings, L.-A.; Friedrich, A.; Loris, E. A.; Panjikar, S.; Trout, B. L.;
Sto¨ckigt, J.; Peters, B.; O’Connor, S. E. J. Am. Chem. Soc. 2008, 130, 710.
(19) Scheme 1 only depicts an intramolecular pathway in the conversion of
7 to 3. Given the range of reactivity observed, we do not rule out an
intermolecular C-N bond forming step. For representative related studies, see:
(a) Pachter, I. J. J. Am. Chem. Soc. 1953, 75, 3026. (b) Vozza, J. F. J. Org.
Chem. 1962, 27, 3856. (c) Oae, S.; Kitao, T.; Kitaoka, Y. J. Am. Chem. Soc.
1962, 84, 3359. (d) Oae, S.; Kitaoka, Y.; Kitao, T. Tetrahedron 1964, 20, 2685.
(e) Bodalski, R.; Katritzky, A. R. Tetrahedron Lett. 1968, 257. (f) Kozuka, S.;
Tamagoki, S.; Negoro, T.; Oae, S. Tetrahedron Lett. 1968, 923.
(13) For IR characterization of isolated nitrilium salts, see: (a) Booth, B. L.;
Jibodu, K. O.; Proenc¸a, M. F. J. Chem. Soc., Chem. Commun. 1980, 1151. (b)
Carrier, A. M.; Davidson, J. G.; Barefield, E. K.; Van Derveer, D. G.
Organometallics 1987, 6, 454.
(14) Addition of either 2-FPyr or 2-ClPyr resulted in an increase in the
intensity of this absorption band.
(15) When these conditions (excess of 2-ClPyr) were used for the transforma-
tion of amide 1f to amide 3fa and 3fc, a significant decrease in the yields (70
and 23%, respectively) was observed.
(16) Ugar, I.; Beck, F.; Fetzer, U. Chem. Ber. 1962, 95, 126.
J. Org. Chem. Vol. 74, No. 3, 2009 1343

residue was purified by flash column chromatography on silica gel
(30 f 50% ethyl acetate in hexanes) to afford the amide 3da (79.0
1
mg, 100%): H NMR (300 MHz, CDCl3, 20 °C) δ 8.31 (d, 1H, J
) 5.5 Hz), 7.99 (d, 1H, J ) 8.5 Hz), 7.78 (d, 1H, J ) 8.5 Hz),
7.65-7.50 (m, 3H), 7.40-7.20 (m, 2 H), 7.15-6.90 (m, 3H), 3.63
(s, 3H); 13C NMR (75 MHz, CDCl3, 20 °C) δ 171.8, 155.9, 141.5,
138.2, 136.2, 130.8, 130.0, 128.4, 128.1, 127.8, 127.3, 125.0, 124.5,
121.0, 37.2; FTIR (neat) cm-1 3058 (m), 2936 (w), 1651 (s), 1560
(s), 1363 (s); HRMS (ESI) calcd for C17H15N2O [M + H]+
263.1179, found 263.1179; TLC (50% ethyl acetate in hexanes),
Rf 0.33 (UV).
N-(4-Methoxypyridin-2-yl)-N-methylbenzamide (3dd, eq
2): Trifluoromethanesulfonic anhydride (46.0 µL, 0.273 mmol, 1.10
equiv) was added via syringe to a solution of N-methylbenzamide
(1d, 33.6 mg, 0.249 mmol, 1 equiv) and 2-fluoropyridine (25.6 µL,
0.298 mmol, 1.20 equiv) in dichloromethane (1.5 mL) at -78 °C.
After 2 min, the reaction mixture was allowed to warm to 0 °C.
After 5 min, 4-methoxypyridine N-oxide (62.2 mg, 0.497 mmol,
2.00 equiv) was added as a solid under an argon atmosphere. After
5 min, the resulting mixture was allowed to warm to 23 °C. After
4 h, triethylamine (100 µL) was added to quench the trifluo-
romethanesulfonate salts, and the mixture was diluted by the
addition of dichloromethane (10 mL). The reaction mixture was
washed with brine (10 mL), and the layers were separated. The
aqueous layer was extracted with dichloromethane (2 × 5 mL).
The combined organic layers were dried over anhydrous sodium
sulfate, filtered, and concentrated under reduced pressure. The
residue was purified by flash column chromatography on alumina
gel (10 f 20% ethyl acetate in hexanes) to afford the amide 3dd
combination of Tf2O and 2-FPyr. This methodology allows for
a highly effective activation of a variety of amide substrates,
including N-aryl amides,4 without requiring the isolation of
sensitive intermediates or the use of heavy metal Lewis acid
additives, and proceeds in shortened reaction times without the
need for elevated temperatures.2 Our in situ monitoring experi-
ments suggest greater propensity for the formation of persistent
nitrilium ion intermediates when N-alkyl amide substrates are
used as compared to N-aryl amides. Our studies with deuterated
N-oxide substrates suggest an irreversible nucleophilic addition
step and a plausible interconversion of intermediates 7 and 8
based on the observed kinetic isotope effect. The activation
conditions described here allow the trapping of highly electro-
philic intermediates with weakly nucleophilic N-oxides.
1
(44.9 mg, 75%): H NMR (500 MHz, CDCl3, 20 °C) δ 8.22 (d,
1H, J ) 6.0 Hz), 7.38-7.28 (m, 3H), 7.26-7.20 (m, 2H), 6.57
(dd, 1H, J ) 6.0, 2.5 Hz), 6.28 (d, 1H, J ) 2.5 Hz), 3.56 (s, 3H),
3.55 (s, 3H); 13C NMR (125 MHz, CDCl3, 20 °C) δ 171.2, 166.6,
158.4, 149.6, 136.3, 130.3, 128.5, 128.2, 108.4, 107.3, 55.4, 36.1;
FTIR (neat) cm-1 3061 (w), 2941 (w), 1652 (s), 1595 (s), 1362
(s); HRMS (ESI) calcd for C14H15N2O2 [M + H]+ 243.1128, found
243.1133; TLC (50% ethyl acetate in hexanes), Rf 0.28 (UV).
Experimental Section
N-(Isoquinolin-1-yl)-N-methylbenzamide (3da, Table 2): Tri-
fluoromethanesulfonic anhydride (55.7 µL, 0.331 mmol, 1.10 equiv)
was added via gastight syringe to a solution of N-methylbenzamide
(1d, 40.7 mg, 0.301 mmol, 1 equiv) and 2-fluoropyridine (31.0 µL,
0.361 mmol, 1.20 equiv) in dichloromethane (1.5 mL) at -78 °C.
After 2 min, the reaction mixture was allowed to warm to 0 °C.
After 5 min, isoquinoline N-oxide (2a, 87.4 mg, 0.602 mmol, 2.00
equiv) was added as a solid under an argon atmosphere. After 5
min, the resulting mixture was allowed to warm to 23 °C. After
4 h, triethylamine (100 µL) was added to quench the trifluo-
romethanesulfonate salts, and the mixture was diluted by the
addition of dichloromethane (10 mL). The reaction mixture was
washed with brine (10 mL), and the layers were separated. The
aqueous layer was extracted with dichloromethane (2 × 5 mL).
The combined organic layers were dried over anhydrous sodium
sulfate, filtered, and concentrated under reduced pressure. The
Acknowledgment. M.M. is a Beckman Young Investigator
and an Alfred P. Sloan research fellow. J.W.M. is grateful for
a NDSEG fellowship. We thank Dr. M.D. Hill for early
contributions and helpful discussions. We are grateful to Dr.
Li Li for obtaining mass spectrometric data at the MIT
Department of Chemistry Instrumentation Facility. This study
was supported by NSF (547905). We acknowledge supplemen-
tary support from Amgen, AstraZeneca, Boehringer Ingelheim,
DuPont, GlaxoSmithKline, and Lilly.
Supporting Information Available: Experimental proce-
dures, spectroscopic data for products. This material is available
free of charge via the Internet at http://pubs.acs.org.
JO802355D
1344 J. Org. Chem. Vol. 74, No. 3, 2009