Generation and reactivity of aza-oxyallyl cationic intermediates: Aza-[4 + 3] cycloaddition reactions for heterocycle synthesis

Article abstract of DOI:10.1021/ja201901d

Aza-[4 + 3] cycloadditions of putative aza-oxyallyl cationic intermediates and cyclic dienes are reported. The intermediate is generated by the dehydrohalogenation of α-haloamides. The reaction is general to a variety of α-haloamides and is diastereoselective. Computational and experimental data suggest that an N-alkoxy substituent stabilizes the aza-oxyallyl cationic intermediate.

Full text of DOI:10.1021/ja201901d

COMMUNICATION
pubs.acs.org/JACS
Generation and Reactivity of Aza-Oxyallyl Cationic Intermediates:
Aza-[4 þ 3] Cycloaddition Reactions for Heterocycle Synthesis
Christopher S. Jeffrey,* Korry L. Barnes, John A. Eickhoff, and Christopher R. Carson
Department of Chemistry, University of Nevada—Reno, Reno, Nevada 89557, United States
S Supporting Information
b
ABSTRACT: Aza-[4 þ 3] cycloadditions of putative aza-
oxyallyl cationic intermediates and cyclic dienes are re-
ported. The intermediate is generated by the dehydrohalo-
genation of R-haloamides. The reaction is general to a
variety of R-haloamides and is diastereoselective. Computa-
tional and experimental data suggest that an N-alkoxy
substituent stabilizes the aza-oxyallyl cationic intermediate.
he [4 þ 3] cycloaddition reaction of oxyallyl cationic inter-
T
mediates (e.g., 2, Figure 1) with dienes has been established
as a premier method for the construction of seven-membered
carbocycles 3.1 Our interest in the reactions of electrophilic
nitrogen species led us to consider aza-oxyallyl cationic inter-
mediates 7 for the synthesis of seven-membered azacycles 8, a
widely represented heterocycle in natural products, pharmaceu-
ticals, peptidomimetics, and monomers for polymerization. This
communication reports the design and generation of aza-oxyallyl
cationic intermediates 7 and their aza-[4 þ 3] cycloaddition with
cyclic dienes.
Figure 1. Aza-[4 þ 3] cycloaddition reaction of aza-oxyallyl cations
with dienes.
starting material, failing to provide the desired cycloadduct 10a
under a variety of conditions.
Oxygen,9 nitrogen,10 and sulfur11 substituents are established
to stabilize oxyallyl cationic intermediates1a,b and acyl nitrenium
ions.12 An N-alkoxy bromoamide13 9b was prepared, and its
dehydrohalogenation in the presence of furan was studied. Treat-
ment of the R-bromohydroxamate 9b to the F€ohlisch conditions
(CF3CH2OH/Et3N)14 in the presence of furan provided the
desired cycloadduct 10b in a 38% yield. The trifluoroethylether
11b (56%) was formed as the major product of this reaction,
presumed to be the result of solvolysis of the intermediate or the
N-benzyloxyaziridinone.5,10
The aza-oxyallylic cation has been primarily discussed in the
context of the synthesis, reactions, and rearrangements of R-lactams
6. Sheehan and Lengyl suggested that aza-oxyallyl cationic inter-
mediates 7 could be relevant to the regioselectivity trends of the
nucleophilic ring-opening reactions of R-lactams 6.2 Stang and
Anderson proposed that the conversion of an alkylidene oxazirine
5 to an R-lactam 6 proceeds through an aza-oxyallylic cation 7.3
Tuscano and co-workers have computationally identified an aza-
oxyallyl cationic transition state in their studies of the isomeriza-
tion of an alkylidene oxazirine 5 to an R-lactam 6.4 Kikgugawa
proposed an aza-oxyallyl cationic intermediate in the solvolysis
of N-chloro-N-alkoxyphenylacetamides.5 However, theoretical6
and stereochemical7 studies of the nucleophilic ring-opening
reactions of alkyl-substituted R-lactams 6 and attempts to trap
the proposed intermediate with alkene reactants have failed to
provide any evidence of its involvement in these reactions.8 The
possibility of intercepting an aza-oxyallyl cationic intermediate in
a [4 þ 3] cycloaddition reaction led us to pursue studies of its
generation and its reaction with dienes.
The choice of base and solvent was found to directly influence the
yield of the cycloadduct. Changing the solvent from trifluoroethanol
(TFE) to hexafluoroisopropanol (HFIP) significantly improved the
yield of the cycloadduct 10b and slowed the formation of the
solvolysis product 11b (cf. entry 2 with 3ꢀ6, Table 1).15 Carbonate
bases provided the desired product in comparable yields (entries
4ꢀ6, Table 1) to the reaction using triethylamine. The reaction
could be effected in ether by using triethylamine with lithium
perchlorate as a Lewis acid additive,16 but these conditions also
provided a methacrylamide as the major product from the elimina-
tion of the cationic intermediate or a transient R-lactam.
Base mediated dehydrohalogenation of R-haloketones 1 is a
common method for the generation of oxyallylic cations 2.1
Similarly, a dehydrohalogenation of an R-haloamide 4 could
provide the desired intermediate 7 in situ. The dehydrohalogena-
tion of 2-bromo-N-benzylbutyramide (9a) in the presence of
furan was studied. This reaction resulted in the recovery of the
Various haloamides underwent the aza-[4 þ 3] cycloaddition
with furan (Table 2). Monoalkyl substituted bromoamides (entries
Received: March 2, 2011
Published: May 03, 2011
r
2011 American Chemical Society
7688
dx.doi.org/10.1021/ja201901d J. Am. Chem. Soc. 2011, 133, 7688–7691
|

Journal of the American Chemical Society
COMMUNICATION
Table 1. Solvent, Base, and Substituent Effects on the Yield of
the Aza-[4 þ 3] Reaction of 9b with Furan
Table 2. Evaluation of the Substrate Scope of the Aza-[4 þ 3]
Reaction of Various Halo-Benzyloxyamides 12 with Furan or
Cyclopentadienea
Entry
R
Solvent
Base
Et3N
Yielda %
1
2
3
4
5
6
7
8
9
Bn
TFE
no reaction
38%b
OBn
OBn
OBn
OBn
OBn
OBn
OBn
OBn
TFE
Et3N
HFIP
HFIP
HFIP
HFIP
TFE
Et3N
78%
Cs2CO3
K2CO3
58%
67%
Na2CO3
imidazole
pyridine
Et3N, LiClO4
74%
decomp.
no reaction
See text
TFE
Et2O
a Isolated yield of 10b. b Provided a 56% yield of 11b.
2ꢀ5) provided the highest yields of the cycloadduct. All monoaryl
and monoalkyl haloamides (entries 2ꢀ5 and 7) selectively provided
the endo-diastereoisomer (g19:1).17 In the case of entry 6, it was
observed that the ratio of diastereoisomers at early conversion
(g19:1, endo:exo at ca. 40% conversion) was greater than at full
conversion (2:1, endo:exo). Monoalkyl bromoamides reacted
slower than aryl and dialkyl substituted haloamides. R-Chloroa-
mides reacted slower than R-bromoamides (cf. entries2 and5). The
bromoacetamide (X = Br, R1 = R2 = H, entry 1) was unreactive
under these conditions, resulting in the recovery of the starting
material. R-Chloromethoxyacetamide (R1 = OCH3) was found to
be difficult to handle and resulted in decomposition during the
cycloaddition reaction. Dialkyl substituted and aryl substituted
haloamides provided the cycloadduct in moderate yields when
run in HFIP (entries 8, 9, and 12). The aza-[4þ 3] cycloaddition of
bromoamides with cyclopentadiene afforded the cycloadducts in
comparable yields to the corresponding furan reactions (entries
10ꢀ12, Table 2) with less selectivity for the endo adduct.
a Conditions: solvent was furan or cyclopentadiene (1:1 v/v, 0.25 M) at
0 to 25 °C with Et3N (2.0 equiv). Diastereoisomeric ratio (dr) was
determined from crude 1H NMR analysis. b g 19:1 dr indicates that the
minor diastereoisomer was not detected. c Isolated yield of both
diastereoisomers, 13.
A computational analysis of the isomerization of R-lactams 14
and 15 to their corresponding aza-oxyallylic cation was conducted
at the B3LYP/6-31G* level of theory with a conductor polarized
continuum model (cpcm)18 for solvation in methanol (Figure 2).19,20
No stationary point could be identified for the aza-oxyallylic cation
(R = Et), which collapsed to the R-lactam 14 during the geometry
minimization. Stationary points for both aza-allylic cation 16 (R =
OCH3) and the corresponding R-lactam 15 were located indicating a
stabilizing effect of this group on the dipolar intermediate. The polar-
ity of the medium had an effect on the depth of the well correspond-
ing to the aza-oxyallylic cation 16 (cf. dashed line vs red line,
Figure 2), consistent with the dipolar character of the intermediate.
The experimental and computational data support a mecha-
nism where dehydrohalogenation of the haloamide 12 generates
an aza-oxyallyl cationic intermediate i that reacts as a dienophile in
the aza-[4 þ 3] cycloaddition reaction (Scheme 1). An electron-
donating group (OBn) is essential for permitting the aza-[4 þ 3]
cycloaddition to take place and was computationally found to
stabilize the proposed intermediate. Aryl substituents accelerate
the overall rate of conversion. Much like the all-carbon [4 þ 3]
cycloaddition with cyclic dienes, the aza-[4 þ 3] cycloaddition
demonstrates a preference for the formation of the endo cycload-
duct. The observation that the diastereoisomeric ratio of cycload-
ducts (entry 6, Table 2) was high at early conversion and that the
purified endo-adduct isomerized to a ca. 1:1 ratio of diastereoisomers
(13-endo to 13-exo, R = Cl, Scheme 1) under the reaction conditions
suggests that there is a kinetic preference for the endo-cycloadduct.
The precise nature of this diastereoselectivity is currently under
further investigation.
In summary, N-benzyloxy R-haloamides react under basic
conditions with cyclic dienes to provide bicyclic lactams in good
yield and with diastereoselectivity. Computational and experi-
mental data support that this reaction proceeds through an aza-
oxyallyl cationic intermediate. The N-alkoxy substituent was
essential for this reaction, and computational analysis revealed
that it acts to stabilize the cationic intermediate. Studies focused
7689
dx.doi.org/10.1021/ja201901d |J. Am. Chem. Soc. 2011, 133, 7688–7691

Journal of the American Chemical Society
COMMUNICATION
We thank Dr. Stephen M. Spain for assistance with mass
spectrometry.
’ REFERENCES
(1) (a) Lohse, A. G.; Hsung, R. P. Chem.—Eur. J. 2011, 17, 3812–
3822. (b) Harmata, M. Chem. Commun. 2010, 8904–8922. (c) Harmata,
M. Chem. Commun. 2010, 8886–8903. (d) Huan, J.; Hsung, R. P.
Chemtracts 2005, 18, 207–214. (e) Harmata, M. Adv. Synth. Catal.
2006, 2297–2306. (f) Harmata, M. Acc. Chem. Res. 2001, 34, 595–605.
(g) Cha, J. K.; Oh, J. Curr. Org. Chem. 1998, 2, 217–232.(h) Harmata, M.
In Advances in Cycloaddition; Lautens, M., Ed.; JAI: Grennwich, 1997;
Vol. 4, pp 41ꢀ86. (i) West, F. G. In Advances in Cycloaddition; Lautens,
M., Ed.; JAI: Grennwich, CT, 1997; Vol. 4, pp 1ꢀ40. (j) Harmata, M.
Tetrahedron 1997, 53, 6235–6280.(k) Padwa, A.; Schoffstall, A. In
Advances in Cycloaddition; Curran, D. P., Ed.; JAI Press: Greenwich,
CT, 1990; Vol. 2, pp 1ꢀ89. (l) Harmata, M. Recent Res. Dev. Org. Chem.
1997, 1, 523–535. (m) Rigby, J. H.; Pigge, F. C. Org. React. 1997,
51, 351–478. (n) Mann, J. Tetrahedron 1986, 42, 4611–4659.
(o) Hoffmann, H. M. R. Angew. Chem., Int. Ed. Engl. 1984, 23, 1–19.
(p) Hoffmann, H. M. R. Angew. Chem., Int. Ed. Engl. 1973, 12, 819–835.
(2) For reviews on the reactivity and synthesis of R-lactams, see:
(a) Lengyel, I.; Sheehan, J. C. Angew. Chem., Int. Ed. Engl. 1968, 7, 25–36.
(b) L’Abbe, G. Angew. Chem., Int. Ed. Engl. 1980, 19, 276–289.
(3) Stang, P. J.; Anderson, G. H. Gazz. Chim. Ital. 1995, 125,
329–331.
Figure 2. Relaxed potential energy scans along the C(3)ꢀN(1) co-
ordinate of R-lactams 14 and 15 (10ꢀ13 steps at 0.1 Å intervals) at a
B3LYP/6-31G* level of theory. Green line/O = 14 in methanol, red
line/9 = 15 in methanol; dashed-line/ꢁ = 15 in the gas phase.
Stationary points are indicated by [.
(4) Cohen, A. D.; Showalter, B. M.; Toscano, J. P. Org. Lett. 2004,
6, 401–403.
(5) Kikugawa, Y.; Shimada, M.; Kato, M.; Sakamoto, T. Chem.
Pharm. Bull. 1993, 41, 2192–2194.
(6) Tantillo, D. J.; Houk, K. N.; Hoffman, R. V.; Tao, J. J. Org. Chem.
Scheme 1. Proposed Mechanism for the Aza-[4 þ 3] Cy-
cloaddition Reaction with Furan (X = O) and Cyclopenta-
diene (X = CH2)
1999, 64, 3830–3837.
(7) For a review of the stereochemical studies of the nucleophilic
ring opening of R-lactams, see: Hoffman, R. V. In The Amide Linkage:
Selected Structural Aspects in Chemistry Biochemistry and Material Science;
Greenberg, A., Breneman, C. M., Liebman, J. F., Eds.; John Wiley & Sons
Inc.: New York, 2000; p 137.
(8) Lengyel, I.; Mark, R. V. Acta Chim. Acad. Sci. Hung. 1974,
81, 475–479.
(9) For examples of [4 þ 3] cycloaddition reactions of nitrogen-
stabilized oxyallylic cations, see: (a) Antoline, J. E.; Hsung, R. P. Synlett
2008, 739–744. (b) Antoline, J. E.; Hsung, R. P.; Huang, J.; Song, Z.; Li,
G. Org. Lett. 2007, 9, 1275–1278. (c) MaGee, D. I.; Godineau, E.;
Thornton, P. D.; Walters, M. A.; Sponholtz, D. J. Eur. J. Org. Chem.
2006, 3667–3680. (d) Huang, J.; Hsung, R. P. J. Am. Chem. Soc. 2005,
127, 50–51. (e) Rameshkumar, C.; Hsung, R. P. Angew. Chem., Int. Ed.
2004, 43, 615–618. (f) Xiong, H.; Huang, J.; Ghosh, S. K.; Hsung, R. P.
J. Am. Chem. Soc. 2003, 125, 12694–12695. (g) Harmata, M.; Ghosh,
S. K.; Hong, X.; Wacharasindhu, S.; Kirchhoefer, P. J. Am. Chem. Soc.
2003, 125, 2058–2059. (h) Xiong, H.; Hsung, R. P.; Berry, C. R.;
Rameshkumar, C. J. Am. Chem. Soc. 2001, 123, 7174–7175. (i) Walters,
M. A.; Arcand, H. R. J. Org. Chem. 1996, 61, 1478–1486.
(10) For recent examples of cycloaddition reactions of oxygen-
stabilized oxyallylic cations, see: (a) Harmata, M.; Huang, C. Tetrahedron
Lett. 2009, 50, 5701–5703. (b) Armstrong, A.; Dominguez-Fernandez, B.;
Tsuchiya, T. Tetrahedron 2006, 62, 6614–6620. (c) Aungst, R. A., Jr.;
Funk, R. L. Org. Lett. 2001, 3, 3553–3555. (d) Stark, C. B. W.; Pierau,
S.; Wartchow, R.; Hoffmann, H. M. R. Chem.—Eur. J. 2000,
6, 684–691. (e) Misske, A. M.; Hoffmann, H. M. R. Chem.—Eur. J.
2000, 6, 3313–3320. (f) Lee, J. C.; Cha, J. K. Tetrahedron 2000,
56, 10175–10184. (g) Harmata, M.; Sharma, U. Org. Lett. 2000,
2, 2703–2705.(h) Harmata, M.; Rashatasakhon, P. Synlett 2000,
1419–1422 and references cited within.
on the detailed mechanistic aspects, the synthetic scope, and the
applications of this reaction are underway.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures, tabu-
b
lated characterization data, complete ref 19, tabulated computational
data, and copies of 1H and 13CNMR spectra for all new compounds
are provided in the Supporting Information. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
cjeffrey@unr.edu
(11) For examples of cycloaddition reactions of sulfur-stabilized
oxyallylic cations, see: (a) Harmata, M.; Kahraman, M.; Adenu, G.;
Barnes, C. L. Heterocycles 2004, 62, 583–618. (b) Masuya, K.; Domon,
K.; Tanino, K.; Kuwajima, I. J. Am. Chem. Soc. 1998, 120, 1724.
(c) Harmata, M.; Carter, K. W. ARKIVOC 2002, 62–70. (d) Harmata,
’ ACKNOWLEDGMENT
Startup funding for this work was provided by the University
of Nevada, Reno. We thank Prof. Thomas Bell for chemicals and
Profs. Benjamin King and Robert Sheridan for helpful discussions.
7690
dx.doi.org/10.1021/ja201901d |J. Am. Chem. Soc. 2011, 133, 7688–7691

Journal of the American Chemical Society
COMMUNICATION
M.; Jones, D. E. Tetrahedron Lett. 1996, 37, 783–786. (e) Harmata, M.;
Fletcher, V. R.; Claassen, R. J., II. J. Am. Chem. Soc. 1991, 113, 9861–
9862. (f) Ogura, K.; Ishida, M.; Fujita, M. Bull. Chem. Soc. Jpn. 1989,
62, 3987–3993.
(12) Kikugawa, Y. Heterocycles 2009, 78, 571–607 and references
cited within.
(13) N-Methoxy R-bromo amides were unstable and readily decom-
posed to a methacrylamide. R-Halobenzyloxyamides were solids that
were much more stable than their N-methoxy counterparts.
(14) F€ohlisch, B.; Gehrlach, E.; Herter, R. Angew. Chem., Int. Ed.
1982, 21, 137.
(15) Hexafluoroisopropanol improves the yield of reactions pro-
ceeding through reactive oxyallyl cationic intermediates; see: (a) Harmata,
M.; Huang, C.; Rooshenas, P.; Schreiner, P. R. Angew. Chem., Int. Ed. 2008,
47, 8696–8699. (b) Myers, A. G.; Barbay, J. K. Org. Lett. 2001, 3, 425–428.
(16) Herter, R.; F€ohlisch, B. Synthesis 1982, 976–979.
(17) The relative configuration of 15 was assigned by coupling
constant analysis [J ∼ 5 Hz C(5)ꢀC(4) for the endo vs J < 3 Hz
C(4)ꢀC(5) for the exo] and NOE correlations.
(18) (a) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995–
2001. (b) Barone, V.; Cossi, M.; Tomasi, J. J. Comput. Chem. 1998,
19, 404–417.
(19) Frisch, M. J. Gaussian 03, revision C.02; Gaussian, Inc.: Wall-
ingford, CT, 2004.
(20) Density functional theory (DFT) methods are demonstrated to
accurately model various aspects of the [4 þ 3] cycloaddition of
oxyallylic cations: (a) Krenske, E. H.; Houk, K. N.; Lohse, A. G.;
Antoline, J. E.; Hsung, R. P. Chem. Sci. 2010, 1, 387–392. (b) Krenske,
E. H.; Houk, K. N; Harmata, M. Org. Lett. 2010, 12, 444–447.
(c) Domingo, L. R.; Arno, M.; Saez, J. A. J. Org. Chem. 2009,
74, 5934–5940. (d) Arno, M.; Picher, M. T.; Domingo, L. R.; Andres,
J. Chem.—Eur. J. 2004, 10, 4742–4749. (e) Saez, J. A.; Arno, M.;
Domingo, L. R. Org. Lett. 2003, 5, 4117–4120. (f) Cramer, C. J.;
Barrows, S. E. J. Phys. Org. Chem. 2000, 13, 176–186. (g) Cramer,
C. J.; Barrows, S. E. J. Org. Chem. 1998, 63, 5523–5532.
7691
dx.doi.org/10.1021/ja201901d |J. Am. Chem. Soc. 2011, 133, 7688–7691