Insights into regioselective oxy (-O-) and imino (-NH-) group insertions of 3-nortricyclanone

Article abstract of DOI:10.1021/jo0159587

Lactam 4-azatricyclo[3.2.1.0^2,7]octan-3-one (16a) was proven to be formed in a previously reported reaction that claimed production of lactam 3-azatricyclo[3.2.1.0^2,7]octan-4-one (17a). In a related reaction, bicyclo[3.1.0]hex-2-ene-endo-6-carbonitrile (15), lactam (16a), and novel hydroxycarbonitriles 19-21 were selectively formed when 3-nortricyclanone (1) was treated with aqueous hydroxylamine-O-sulfonic acid (HOSA). Since nitrile 15 neither hydrolyzed nor underwent intramolecular Ritter reaction under these conditions, mechanisms involving Beckmann rearrangement of 3 to nitrilium ion 9 (normal) and Beckmann fragmentation of 3 to cation 8 (abnormal) were investigated using semiempirical calculations. When alkaline HOSA was employed, lactams 16a and 17a were formed in a 1:2 ratio, perhaps via oxaziridine 6a. A similar selectivity was observed using an NH₃/NaOCl reagent solution, which afforded lactone 4-oxatricyclo[3.2.1.0^2,7]octan-3-one (16b) in addition to both lactams. Consequently, the Baeyer-Villiger oxidation of 1 with NaOCl gave 16b exclusively. Finally, the Schmidt reaction of ketone 1 gave only the lactam 17a, via cyclopropyl migration, and the same fragmentation products obtained from the acidic HOSA reaction. Migration selectivities are rationalized in terms of nucleofugacity, electronic effects, cyclopropyl regulation, and MO theory.

Full text of DOI:10.1021/jo0159587

450
J . Org. Chem. 2002, 67, 450-456
In sigh ts in to Regioselective Oxy (-O-) a n d Im in o (-NH-) Gr ou p
In ser tion s of 3-Nor tr icycla n on e
Murray G. Rosenberg,† Ulrike Haslinger, and Udo H. Brinker*
Institut fu¨r Organische Chemie, Universita¨t Wien, Wa¨hringer Strasse 38, A-1090 Wien, Austria
Udo.Brinker@univie.ac.at
Received J uly 20, 2001
Lactam 4-azatricyclo[3.2.1.02,7]octan-3-one (16a ) was proven to be formed in a previously reported
reaction that claimed production of lactam 3-azatricyclo[3.2.1.02,7]octan-4-one (17a ). In a related
reaction, bicyclo[3.1.0]hex-2-ene-endo-6-carbonitrile (15), lactam (16a ), and novel hydroxycarboni-
triles 19-21 were selectively formed when 3-nortricyclanone (1) was treated with aqueous
hydroxylamine-O-sulfonic acid (HOSA). Since nitrile 15 neither hydrolyzed nor underwent
intramolecular Ritter reaction under these conditions, mechanisms involving Beckmann rearrange-
ment of 3 to nitrilium ion 9 (normal) and Beckmann fragmentation of 3 to cation 8 (abnormal)
were investigated using semiempirical calculations. When alkaline HOSA was employed, lactams
16a and 17a were formed in a 1:2 ratio, perhaps via oxaziridine 6a . A similar selectivity was
observed using an NH3/NaOCl reagent solution, which afforded lactone 4-oxatricyclo[3.2.1.02,7]octan-
3-one (16b) in addition to both lactams. Consequently, the Baeyer-Villiger oxidation of 1 with
NaOCl gave 16b exclusively. Finally, the Schmidt reaction of ketone 1 gave only the lactam 17a ,
via cyclopropyl migration, and the same fragmentation products obtained from the acidic HOSA
reaction. Migration selectivities are rationalized in terms of nucleofugacity, electronic effects,
cyclopropyl regulation, and MO theory.
In tr od u ction
rearrangement of oxime derivatives proceeds stereo-
specifically whereby the neighboring group lying anti to
the N-nucleofuge assists the ionization during its migra-
tion.9 The stereoconfiguration of the migrating group is
retained. Retention of configuration is also observed for
the Baeyer-Villiger oxidation, for which general migra-
tory aptitudes within unsymmetrical ketones have been
ascertained: 3°-alkyl > 2°-alkyl > benzyl > phenyl > 1°-
alkyl > cyclopropyl > methyl.10 The migrating substitu-
ent is the electrofuge that can better stabilize a partial
positive charge. Thus, cyclopropyl migration is normally
observed only with methyl-substituted ketones, giving
acetic esters.11 However, exceptions can be found. Treat-
ment of bicyclo[3.1.0]hexan-2-one with peroxy acid m-
CPBA led only to the lactone 2-oxabicyclo[4.1.0]heptan-
3-one, via cyclopropyl migration.12 Consequently, experi-
mentation is warranted since other factors like steric
effects may be operating.
The transformation of a strained cyclic ketone into its
corresponding spiroannular diaziridine is plagued by ring
expansion when ketimine formation is inefficient.1 There-
fore, the conversion of highly strained 3-nortricyclanone
(1) into 3-hydrazinortricyclane (6) is difficult (eq 1,
Scheme 1).2 Nevertheless, new insights regarding the
migration selectivities within ketone 1 have been gained.
Skeletal rearrangements involving migration to an
electron-deficient carbon include the homologation of
ketones by diazomethane,3 the Tiffeneau-Demjanov ring
expansion,4 and the pinacol rearrangement.5 Similarly,
migrations to electron-deficient heteroatoms constitute
a synthetically useful class of organic reactions worthy
of attention.6 The Beckmann rearrangement, among
others, involves migration to an electron-deficient nitro-
gen,7 while the Baeyer-Villiger oxidation features mi-
gration to an electron-deficient oxygen.8 The Beckmann
Cyclopropylmethyl cations are stabilized by conjuga-
tion between the cyclopropane ring’s π-type Walsh orbit-
als and the vacant p-orbital on the vicinal carbocation
center.13,14 Thus, 3-nortricyclanone (tricyclo[2.2.1.02,6]-
* Corresponding author. Phone: +43-1-4277-52121. Fax: +43-1-
4277-52140.
† Department of Chemistry, State University of New York, Bing-
hamton, NY 13902-6016.
(1) (a) Kupfer, R. Ph.D. Dissertation, State University of New York
at Binghamton, Binghamton, New York, 1995; pp 18, 154. (b) Kupfer,
R.; Brinker, U. H. Unpublished results.
(2) Rosenberg, M. G.; Brinker, U. H. J . Org. Chem. 2001, 66, 1517-
1522.
(3) Gutsche, C. D. Org. React. (NY) 1954, 8, 364-429.
(4) Smith, P. A. S.; Baer, D. R. Org. React. (NY) 1960, 11, 157-188.
(5) Collins, C. J . Q. Rev., Chem. Soc. 1960, 14, 357-377.
(6) Smith, P. A. S. In Molecular Rearrangements; de Mayo, P., Ed.;
Wiley: New York, 1963; pp 457-591.
(9) Nevertheless, it is the aryl group of alkyl aryl oximes that
typically migrates regardless of the initial configuration.
(10) (a) House, H. O. Modern Synthetic Reactions, 2nd ed.; W. A.
Benjamin: Menlo Park, CA, 1972; pp 321-329. (b) Plesnicˇar, B. In
Oxidation in Organic Chemistry; Trahanovsky, W. S., Ed.; Academic:
New York, 1978; Part C, pp 254-267.
(11) Emmons, W. D.; Lucas, G. B. J . Am. Chem. Soc. 1955, 77,
2287-2288.
(12) Wiberg, K. B.; Snoonian, J . R. J . Org. Chem. 1998, 63, 1390-
1401.
(7) (a) Donaruma, L. G.; Heldt, W. Z. Org. React. (NY) 1960, 11,
1-156. (b) Gawley, R. E. Org. React. (NY) 1988, 35, 1-420.
(8) (a) Krow, G. R. Org. React. (NY) 1993, 43, 251-798. (b) Krow,
G. R. Tetrahedron 1981, 37, 2697-2724. (c) Lee, J . B.; Uff, B. C. Q.
Rev., Chem. Soc. 1967, 21, 429-457. (d) Hassall, C. H. Org. React.
(NY) 1957, 9, 73-106.
(13) (a) Advanced Organic Chemistry, 4th ed.; March, J ., Ed.;
Wiley: New York, 1992; pp 152, 169, 323-324. (b) J orgensen, W. L.;
Salem, L. The Organic Chemist’s Book of Orbitals; Academic: New
York, 1973; pp 35-37, 206-209. (c) Wilcox, C. F.; Loew, L. M.;
Hoffmann, R. J . Am. Chem. Soc. 1973, 95, 8192-8193. (d) Macomber,
R. S. J . Chem. Educ. 1998, 75, 1346-1350.
10.1021/jo0159587 CCC: $22.00 © 2002 American Chemical Society
Published on Web 12/15/2001

Regioselective Oxy and Imino Group Insertions of 3-Nortricyclanone
J . Org. Chem., Vol. 67, No. 2, 2002 451
Sch em e 1. Dia zir id in e F or m a tion Wa s Not Obser ved Wh en 3 Wa s Qu en ch ed by Am m on ia
heptan-3-one)15 (1) is an ideal system with which to study
this effect because its three-membered ring is locked in
an endocyclic configuration that keeps the C1-C6 bond
in 1 parallel to the electrophilic p-orbital on C3 (eq 1).16
Indeed, this perfect alignment is the main feature
responsible for the inexorable cyclopropylcarbene frag-
mentation of the corresponding singlet carbene, 3-nortri-
cyclanylidene.2,17 With this in mind, the lactamization of
1,18 reportedly giving “3-azatricyclo[3.2.1.02,7]octan-4-one
(17a )” in 37.9% yield,15d is surprising for two reasons: (1)
the yield of 17a is rather high given the tricyclic system’s
propensity to relieve strain via fragmentation, and (2)
alleged lactam 17a derives from an unpreferred cyclo-
propyl migration.19 So, if imino (-NH-) group insertion
is truly permissible for 1, the migration selectivity is still
questionable.
thermore, a facile way to perform the Beckmann re-
arrangement directly upon cyclic ketones makes use of
HOSA.21 This eliminates the need for preparing the
oxime, e.g., 3-nortricyclanone oxime (7),15d,23 in a separate
step because the rearrangement proceeds through oxime-
O-sulfonic acids.24
Thus, when ketone 1 was treated with aqueous HOSA
and either subsequently quenched with concentrated
aqueous NH3 or not, the major product formed was
bicyclo[3.1.0]hex-2-ene-endo-6-carbonitrile (15) (Scheme
2).15d,25 This demonstrates that 3 is truly unstable and
ionizes prior to the addition of NH3. Indeed, in a follow-
up experiment employing a different weak base quencher
(saturated aqueous K2CO3), the potassium salt of 3
(cf. 4) could not be precipitated from the aqueous
solution.22a
In addition, three novel “hydrolysis” products of 15
were also formed during the reaction, regardless of
quenching (Scheme 3). These were 3-hydroxybicyclo-
[3.1.0]hexane-endo-6-carbonitrile (19),26 2-hydroxybicyclo-
[3.1.0]hexane-endo-6-carbonitrile epimer 1 (20), and
2-hydroxybicyclo[3.1.0]hexane-endo-6-carbonitrile epimer
2 (21). But control experiments showed that nitrile 15
was unaffected by the original acidic reaction conditions.
Thus, hydroxycarbonitriles 19-21 were formed contem-
poraneously with 15.
Resu lts a n d Discu ssion
Elusive diaziridines are sometimes accessible by the
action of ammonia upon oxime-O-sulfonic acids or es-
ters.20 Therefore, 3-nortricyclanone oxime-O-sulfonic acid
(3) was immediately sought for this transformation
because such diaziridine precursors are easily formed by
the action of H2NOSO2OH, hydroxylamine-O-sulfonic
acid (HOSA), upon cyclic ketones (eq 1).21
(19) “Since the structure 50 [i.e., 17a ] was assigned [in ref 15d] on
the assumption of cyclopropane ring migration, it should be considered
unproven without confirmatory spectral data.” See p 1291 in Krow, G.
R. Tetrahedron 1981, 37, 1283-1307.
(20) (a) Schmitz, E. In Chemistry of Diazirines; Liu, M. T. H., Ed.;
CRC Press: Boca Raton, FL, 1987; Vol. 1, pp 57-98. (b) Schmitz, E.
In Methoden der Organischen Chemie (Houben-Weyl); Klamann, D.,
Ed.; Thieme: Stuttgart, 1992; Vol. E16c; pp 678-728. (c) Schmitz, E.
Adv. Heterocycl. Chem. 1979, 24, 63-107. (d) Makhova, N. N.;
Petukhova, V. Y.; Khmel’nitskii, L. I. Bull. Acad. Sci. USSR, Div.
Chem. Sci. 1982, 31, 1858-1862. (e) Mikhailyuk, A. N.; Makhova, N.
N.; Bova, A. E.; Khmel’nitskii, L. I.; Novikov, S. S. Bull. Acad. Sci.
USSR, Div. Chem. Sci. 1978, 27, 1367-1371. (f) Briner, K.; Vasella,
A. Helv. Chim. Acta 1989, 72, 1371-1382. (g) Vasella, A.; Witzig, C.;
Waldraff, C.; Uhlmann, P.; Briner, K.; Bernet, B.; Panza, L.; Husi, R.
Helv. Chim. Acta 1993, 76, 2847-2875. (h) Brunner, J .; Senn, H.;
Richards, F. M. J . Biol. Chem. 1980, 255, 3313-3318. (i) Nassal, M.
Liebigs Ann. Chem. 1983, 1510-1523. (j) Schustov, G. V.; Tavakalyan,
N. B.; Kostyanovsky, R. G. Angew. Chem., Int. Ed. Engl. 1981, 20,
200-201.
(21) (a) Sanford, J . K.; Blair, F. T.; Arroya, J .; Sherk, K. W. J . Am.
Chem. Soc. 1945, 67, 1941-1943. (b) Novoselov, E. F.; Isaev, S. D.;
Yurchenko, A. G.; Vodichka, L.; Trshiska, Y. J . Org. Chem. USSR 1981,
17, 2284-2289. (c) Olah, G. A.; Fung, A. P. Synthesis 1979, 537-538.
(d) Encyclopedia of Reagents for Organic Synthesis; Paquette, L. A.,
Ed.; Wiley: New York, 1995; Vol. 4, pp 2764-2767.
(22) (a) Smith, P. A. S. J . Am. Chem. Soc. 1948, 70, 323-326. (b)
Smith, R. A. G.; Knowles, J . R. J . Chem. Soc., Perkin Trans. 2 1975,
686-694.
It was anticipated that the quenching of in situ 3 with
NH3 would lead to the planned diaziridine, 3-hydrazino-
tricyclane (6) (Scheme 1).2 However, none was formed.
Historically, oxime-O-sulfonic acids are unstable spe-
cies that, on occasion, have been isolated as their potas-
sium or ammonium salts (e.g., 4 in Scheme 1).22 Fur-
(14) For related examples involving the activation of σ-bonds vicinal
to cyclopropane rings, see: (a) Xu, L.; Smith, W. B.; Brinker, U. H. J .
Am. Chem. Soc. 1992, 114, 783-784. (b) Banwell, M. G.; Haddad, N.;
Huglin, J . A.; MacKay, M. F.; Reum, M. E.; Ryan, J . H.; Turner, K. A.
J . Chem. Soc., Chem. Commun. 1993, 954-957. (c) Halton, B.; Battiste,
M. A.; Rehberg, R.; Deyrup, C. L.; Brennan, M. E. J . Am. Chem. Soc.
1967, 89, 5964-5965.
(15) (a) Meinwald, J .; Crandall, J . K. J . Am. Chem. Soc. 1966, 88,
1292-1301. (b) Meinwald, J .; Crandall, J .; Hymans, W. E. Org. Synth.
Coll. 1973, 5, 866-868. (c) Upadek, H. Ph.D. Dissertation, Ruhr-
Universita¨t Bochum, Germany, 1977; pp 156-159. (d) Hall, H. K., J r.
J . Am. Chem. Soc. 1960, 82, 1209-1215.
(16) The carbonyl carbon of ketone 1 is partially positive (i.e.,
+0.2558) according to AM1 calculations. See Supporting Information.
(17) Herges, R. Angew. Chem., Int. Ed. Engl. 1994, 33, 255-276.
(18) Via in situ 3-nortricyclanone oxime-O-benzensulfonic ester.
Bicyclo[3.1.0]hex-2-ene-endo-6-carbonitrile (15) is also reported, likely
formed via δ-elimination (i.e., 1,4-conjugate elimination) of in situ
3-nortricyclanone oxime O-benzensulfonic ester with the excess NaOH
via the E2′ mechanism.
(23) (a) Hawkes, G. E.; Herwig, K.; Roberts, J . D. J . Org. Chem.
1974, 39, 1017-1028. (b) Botto, R. E.; Westerman, P. W.; Roberts, J .
D. Org. Magn. Reson. 1978, 11, 510-515. (c) Chizhov, A. O.; Zefirov,
N. S.; Zyk, N. V. J . Org. Chem. 1987, 52, 5647-5655.
(24) (a) Gregory, B. J .; Moodie, R. B.; Schofield, K. J . Chem. Soc., B
1970, 338-346. (b) Kim, S.-G.; Kawakami, T.; Ando, T.; Yukawa, Y.
Bull. Chem. Soc. J pn. 1979, 52, 1115-1120. (c) Yukawa, Y.; Ando, T.
J . Chem. Soc., D 1971, 1601-1602. (d) Pearson, D. E.; Ball, F. J . Org.
Chem. 1949, 14, 118-131.
(25) Lu¨bbe, F.; Sustmann, R. Chem. Ber. 1979, 112, 57-70.
(26) Steric hindrance by the cyano group of cation 8 likely precludes
the formation of endo-3-hydroxybicyclo[3.1.0]hexane-endo-6-carbo-
nitrile.

452 J . Org. Chem., Vol. 67, No. 2, 2002
Rosenberg et al.
Sch em e 2. Dep iction of th e Nor m a l a n d Abn or m a l Beck m a n n Rea r r a n gem en ts fr om 3, a n d
Sh ech ter -Sch m id t-Typ e Rea ction fr om 2a , Lea d in g to La cta m s a n d Nitr ilesa
a
AM1 enthalpies are in kcal/mol.
Sch em e 3. F or m a tion of Hyd r oxyca r bon itr iles
Lactam 16a likely derives from the normal Beckmann
rearrangement syn -3f9. But 16a could also stem from
a tandem reaction involving the abnormal Beckmann
rearrangement syn -3f8, called Beckmann fragmenta-
tion,29 and the subsequent Ritter reaction 8f9,21b,30
even though intramolecular Ritter reactions are rare.31
Still, lactam 16a might arise from the path 2a f13a ,
a mechanistic variant of the well-known Schmidt re-
action for ketones.32 Therefore, theoretical chemistry
was engaged to help elucidate the mechanism. The
results of AM1 calculations are also summarized in
Scheme 2.33
19-21
Nitrile 15 results from Beckmann fragmentation. But
C-N rupture can occur in either one (syn -3f8) or two
(syn -3f9f8) steps. The calculated dihedral angle (ω)
depicted for syn -3 is (169°. Thus, the pertinent Cδ-H
σ-orbital is not parallel (i.e., (180°) to the NR-O bond.
Yet substantial overlap between the filled Cδ-H σ-orbital
and the empty Câ-Cγ σ*-orbital could still develop. It is
therefore reasonable that 15 is generated by a concerted
heterolytic fragmentation (syn -3f8) because syn -3 es-
sentially meets the rigorous stereoelectronic require-
ments outlined by Grob’s antiparallelism principle.34 But
the activation enthalpy (∆H‡) would be somewhat ele-
vated, making the transformation less rapid.35
Finally, the lactam 4-azatricyclo[3.2.1.02,7]octan-3-one
(16a ) was also formed from 1, but in low yield (Table 1).
In fact, when we repeated a published procedure on a
small scale,15d we obtained nitrile 15 and lactam 16a ,18
not 17a as reported. So, imino group insertion is truly
permissible for 1 but the usual methylidyne migration
preference is observed (vide supra).7b Curiously, the 13C
NMR spectra for both lactams 16a and 17a have been
reported but neither preparative details nor allusion to
the earlier structural misassignment are mentioned.27 We
herein report both the separation of lactams 16a and 17a
via column chromatography and their structural verifica-
tion using heteronuclear correlation (HETCOR) spectros-
copy.28
(29) Ferris, A. F. J . Org. Chem. 1960, 25, 12-18.
(30) (a) Hill, R. K.; Conley, R. T.; Chortyk, O. T. J . Am. Chem. Soc.
1965, 87, 5646-5651. (b) Palmere, R. M.; Conley, R. T.; Rabinowitz,
J . L. J . Org. Chem. 1972, 37, 4095-4097. (c) Korsloot, J . G.; Keizer,
V. G. Tetrahedron Lett. 1969, 3517-3520.
(31) Van Emelen, K.; De Wit, T.; Hoornaert, G. J .; Compernolle, F.
Org. Lett. 2000, 2, 3083-3086.
(32) Fikes, L. E.; Shechter, H. J . Org. Chem. 1979, 44, 741-744.
(33) Austin Model 1: Dewar, M. J . S.; Zoebisch, E. G.; Healy, E. F.;
Stewart, J . J . P. J . Am. Chem. Soc. 1985, 107, 3902-3909.
(34) Grob, C. A. Angew. Chem., Int. Ed. Engl. 1969, 8, 535-546.
(27) Barfield, M.; Babaqi, A. S. Magn. Reson. Chem. 1987, 25, 443-
451.
(28) See Experimental Section and Supporting Information.

Regioselective Oxy and Imino Group Insertions of 3-Nortricyclanone
J . Org. Chem., Vol. 67, No. 2, 2002 453
Ta ble 1. Rela tive Yield s (%) of Heter oa tom In ser tion P r od u cts of Keton e 1
reagent
HOSA
HOSA/NaOH
NH3/NaOCl
NaOCl
inserting group(s)
leaving group
16a
17a
16b
15, 19-21
othera
-NH-
-NH-
H2SO4(aq)
Na2SO4(aq)
NaCl(aq)
NaCl(aq)
N2(g)
8.1
18.6
17.3
90.4
1.6
44.1
11.4
4.1
37.3
43.6
-NH-, -O-
-O-
27.7
95.9
HN3
-NH-
11.2
45.7
43.0
a
See Experimental Section.
F igu r e 2. Participation of π-type Walsh orbitals might
regulate migration selectivity by either competing against or
cooperating with group X.
F igu r e 1. Degree to which Grob fragmentation is allowed is
a function of the pertinent torsional angle (ω) and is related
to the activation enthalpy (∆H‡).
real intermediate. Indeed, geometry optimization of 10
also leads to 8. Hence, this common “intermediate” from
syn -3 and a n ti-3 could explain how ionization of the
latter diastereomer also leads to the elimination, substi-
tution, and rearrangement products stemming from
cation 8.
Regarding a n ti-3, concerted heterolytic fragmentation
(i.e., a n ti-3f12) leading to 18 is truly doubtful because
its ω ) (147°. This transformation is Grob-forbidden and
would have a large ∆H‡ (Figure 1).36 Besides, 12 is
avoided because it has a highly angle-strained cationic
sp2 center on cyclopropyl C6 and is thus more than 20
kcal higher in energy than 8.37
However, it should be emphasized that both diastereo-
mers of 3 are expected to be formed, especially under
acidic conditions where isomerization occurs readily.38 So,
the concerted anti migration/ionization steps from each
diastereomer should yield the respective nitrilium ion
intermediates 9 and 11, which differ by ca. 2 kcal.
However, since no products deriving from 11 were
observed,39 it is likely that cyclopropyl migration leading
to 11 does not take place due to kinetic factors. Stabiliza-
tion of 9 by the cyclopropane ring’s π-type Walsh orbitals
could be relevant (Figure 2).
Note that nitrilium ions have linear sp centers, both
on carbon and on nitrogen. However, AM1 calculations
show that 9 and 11 are bent. Interestingly, imposing a
linear constraint on C3 of 9 led to cation 8 after geometry
optimization. Thus, the increased ring strain that linear-
ity would incur is simply avoided via C-N rupture.
Finally, it must be mentioned that expulsion of the H2SO4
nucleofuge from protonated 3 without skeletal rearrange-
ment would give nitrenium cation 10. But 10 is not a
The Shechter-Schmidt-type intermediate,32 i.e., 2a (eq
1), can give both 13a and 14a , which are O-protonated
tricyclic lactams.40 These cations have essentially the
same energy, so there would be no thermodynamic origin
of selectivity. Remarkably, migratory aptitudes during
the Schmidt reaction tend somehow to conflict with those
of the Beckmann rearrangement.41 This could be due to
the nature of the N-nucleofuge (Table 1; vide infra).
So, to elucidate matters, the N-(sulfooxy)carbinolamine
intermediate 2a (eq 1) was again targeted.42 But now,
its formation within an alkaline environment should
thwart dehydration to 3 and lend insight into lactam
formation. Remarkably, when ketone 1 was introduced
into an alkaline solution of HOSA, both lactams 16a and
17a were formed with a 2-fold preference for 17a (Scheme
4). However, in contrast to the acidic HOSA reaction, the
alkaline HOSA reaction would not go to completion and
was also hampered by side reactions.43
Schmitz has shown that cyclization of N-(sulfooxy)-
carbinolamine intermediates, which yields the oxa-
(40) They are also N-protonated lactims. Note that strained lactams
with pyramidal N atoms, like R-lactams, are normally N-protonated
in acidic media.
(35) Hence, nitrile 15 formation from syn -3 is an example of
δ-elimination (i.e., 1,4-conjugate elimination) by the E1′ mechanism.
[cf. ref 18].
(36) Also, see Figure S3 in Supporting Information for AM1-
calculated energies.
(37) The cyclopentane ring in 8 adopts a flat conformation after
geometry optimization regardless of the starting geometry. For a
conformational analysis, see Figure S4 in Supporting Information.
Cation 12 is more boatlike. However, the large energy difference of
22.4 kcal is mostly due to cation 12 being very strained.
(38) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in
Organic Chemistry; Harper & Row: New York, 1987; p 491.
(39) The hypothetical elimination of HCN by nitrile 18 to give
benzene was not investigated.
(41) (a) Krow, G. R.; Szczepanski, S. W.; Kim, J . Y.; Liu, N.; Sheikh,
A.; Xiao, Y.; Yuan, J . J . Org. Chem. 1999, 64, 1254-1258. (b) Krow,
G. R.; Cheung, O. H.; Hu, Z.; Lee, Y. B. J . Org. Chem. 1996, 61, 5574-
5580. (c) Krow, G. R. Tetrahedron 1981, 37, 1283-1307. (d) Elderfield,
R. C.; Losin, E. T. J . Org. Chem. 1961, 26, 1703-1710. (e) Potti, N.
D.; Nobles, W. L. J . Pharm. Sci. 1968, 57, 1785-1787. (f) Shechter,
H.; Kirk, J . C. J . Am. Chem. Soc. 1951, 73, 3087-91.
(42) Urbansky, E. T. J . Chem. Educ. 2000, 77, 1644-1647.
(43) Azine is formed from the side reaction of ketone 1 and in situ
hydrazine: (a) 3H2NOSO2OH(aq) + 6OH-(aq) f NH3(aq) + N2(g) + 6H2O(l)
2-
+ 3SO4 (aq), (b) H2NOSO2OH(aq) + 3NH3(aq) f NH2NH2(aq) + 2NH4+(aq)
2-
+ SO4 (aq) (cf. Raschig hydrazine synthesis: (i) Raschig, F. Chem. Ber.
1907, 40, 4580-4588. (ii) Adams, R.; Brown, B. K. Org. Synth. Coll.
1941, 1, 309-310.), and (c) 2(1) + NH2NH2(aq) f azine + 2H2O(l)
.

454 J . Org. Chem., Vol. 67, No. 2, 2002
Rosenberg et al.
Sch em e 4. Rea ction of Keton e 1 w ith Alka lin e
Aqu eou s HOSA Lea d s to Both La cta m s 16a a n d
17a
Sch em e 5. Ba eyer -Villiger Oxid a tion of Keton e 1
w ith Aqu eou s Hyp och lor ite Lea d s On ly to La cton e
16b
ziridine system (e.g., 6a in Scheme 4),44 can occur when
ketones are reacted with HOSA under alkaline condi-
tions.20c,45 Nevertheless, the intermediacy of oxaziridine
6a can currently only be inferred from the reaction
conditions, which give 1-oxa-2-azaspiro[2.5]octane from
cyclohexanone.45a Yet the direct ring expansion of Na-
2a to lactams 16 and 17 may be operative. Finally, as
suggested by AM1, it is doubtful that protonation and
direct rearrangement of N-(sulfooxy)carbinolamine 2a to
either cation 13a or cation 14a occur during the acidic
HOSA reaction (Scheme 2) because no 17a had been
formed.
Next, diaziridine 6 was sought from an N-chloro-gem-
diamine intermediate,46 which is related to the N-
(sulfooxy)-gem-diamine intermediate 5a (Scheme 1). But
when dilute NaOCl, sodium hypochlorite,47 was cau-
tiously added to a stirred mixture of ketone 1 and
aqueous NH3, no diaziridine 6 was formed. Instead, both
lactams 16a and 17a were again produced (16a :17a )
1:2.5), perhaps via the same oxaziridine 6a (vide
supra).44d,48 In addition, the lactone 4-oxatricyclo[3.2.1.02,7]-
octan-3-one (16b) was formed.49,50
6b, but no evidence for the latter was apparent.53
Nonetheless, both intermediates should, in principle,
form both lactones 16b and 17b. Hence, the absence of
lactone 17b is intriguing since both lactams 16a and 17a
were formed under analogous conditions (cf. Scheme 4).
It is noteworthy, when N-(sulfooxy)-gem-diamine 5a is
compared to O-chloro-gem-diol 2b, that our attempt to
form the diaziridine 6 (Scheme 1) mimicked the classical
attempt of Baeyer and Villiger to form the isoelectronic
dioxirane ring system (e.g., 6b in Scheme 5).53 They also
observed migration to an electron-deficient heteroatom
in lieu of spiroannulation.54
Finally, to determine whether the migratory aptitude
during the Schmidt reaction of ketone 1 would conflict
with that of the Beckmann rearrangement,41 ketone 1
was reacted with in situ HN3, hydrazoic acid.55,56 Surpris-
ingly, in addition to nitrile 15 and hydroxycarbonitriles
19-21, lactam 17a was formed exclusively. This unex-
pected result clearly demonstrates that cation 8 is not a
precursor of lactam 16a .
Accordingly, when ketone 1 was treated with just
dilute NaOCl, the alkaline Baeyer-Villiger oxidation
gave lactone 16b exclusively, in 52% overall yield (Scheme
5).51,52 This selective lactonization is in agreement with
a previous account of the customary acidic Baeyer-
Villiger oxidation of 1.49b Lactone 16b may arise directly
from O-chloro-gem-diol 2b or indirectly from dioxirane
Notwithstanding, the homologation of ketone 1, either
with diazomethane or by the Tiffeneau-Demjanov ring
expansion,3,4 is reported to give mostly tricyclo[3.2.1.02,7]-
octan-4-one (17c) and only minor amounts of tricyclo-
(44) (a) Aube´, J .; Hammond, M.; Gherardini, E.; Takusagawa, F. J .
Org. Chem. 1991, 56, 499-508, 4086. (b) Smith, B. T.; Gracias, V.;
Aube´, J . J . Org. Chem. 2000, 65, 3771-3774. (c) Oliveros-Desherces,
E.; Rivie`re, M.; Parello, J .; Lattes, A. Tetrahedron Lett. 1975, 851-
854. (d) Schmitz, E.; Ohme, R.; Schramm, S.; Striegler, H.; Heyne, H.-
U.; Rusche, J . J . Prakt. Chem. 1977, 319, 195-200. (e) Streith, J .; Fizet,
C.; Fritz, H. Helv. Chim. Acta 1976, 59, 2786-2792.
(45) (a) Schmitz, E.; Ohme, R.; Schramm, S. Chem. Ber. 1964, 97,
2521-2526. (b) Schmitz, E. Adv. Heterocycl. Chem. 1963, 2, 83-130.
(c) Schmitz, E. Angew. Chem., Int. Ed. Engl. 1964, 3, 333-341.
(46) (a) Kuznetsov, V. V.; Makhova, N. N.; Khmel’nitskii, L. I. Russ.
Chem. Bull. 1997, 46, 1354-1356. (b) Kuznetsov, V. V.; Makhova, N.
N.; Strelenko, Y. A.; Khmel’nitskii, L. I. Bull. Acad. Sci. USSR, Div.
Chem. Sci. 1991, 40, 2496-2505. (c) Makhova, N. N.; Petukhova, V.
Y.; Kuznetsov, V. V.; Khmel’nitskii, L. I.; Lebedev, V. P.; Pepekin, V.
I. Int. Annu. Conf. ICT 1998, 29th (Energetic Materials), 10.1-10.13;
Chem. Abstr. 1998, 129, 138094.
(51) Na+OCl- is rarely used to conduct the Baeyer-Villiger oxida-
tion. Moreover, Lieben hypohalite oxidation predominates with aryl
methyl ketones, giving arenecarboxylates (i.e., haloform reaction): (a)
Lieben, A. Ann. Chem. Pharm. (Suppl.) 1870, 7, 218-236. (b) Lieben,
A. Ann. Chem. Pharm. (Suppl.) 1870, 7, 377-378. (c) Hassner, A.;
Stumer, C. Organic Syntheses Based on Name Reactions and Unnamed
Reactions; Pergamon: Tarrytown, New York, 1994; p 235). Yet the
Baeyer-Villiger oxidation can compete, giving methyl esters stemming
from unpreferred methyl migration. For an example, see: (d) Kawano,
N.; Okigawa, M.; Hasaka, N.; Kouno, I.; Kawahara, Y.; Fujita, Y. J .
Org. Chem. 1981, 46, 389-392.
(52) Compare to Dakin reaction with Na+OOH-. See: Hocking, M.
B.; Bhandari, K.; Shell, B.; Smyth, T. A. J . Org. Chem. 1982, 47, 4208-
4215.
(53) Murray, R. W.; Singh, M.; J eyaraman, R. J . Am. Chem. Soc.
1992, 114, 1346-1351. Note that Murray et al. use the term “Criegee
intermediate” when referring to their R-hydroxyperoxysulfuric ester
intermediate 6. But this term is usually reserved for carbonyl-O-oxides,
though both can form dioxiranes. Since Criegee devised the mecha-
nisms for the Baeyer-Villiger oxidation of ketones, the Criegee
rearrangement of peroxycarboxylic esters and hydroperoxides, and the
ozonolysis of alkenes, confusion may arise.
(47) (a) Encyclopedia of Reagents for Organic Synthesis; Paquette,
L. A., Ed.; Wiley: New York, 1995; Vol. 7, pp 4580-4585. (b)
Chakrabartty, S. K. In Oxidation in Organic Chemistry; Trahanovsky,
W. S., Ed.; Academic: New York, 1978; Part C; pp 359-363.
(48) The potential formation of transient 3-nortricyclanone N-
chloroimine (cf. 3) cannot be refuted a priori; see refs 44d and 66.
(49) (a) Sustmann, R.; Lu¨bbe, F. Chem. Ber. 1976, 109, 444-454.
(b) Sauers, R. R. Tetrahedron Lett. 1962, 1015-1017. (c) Olah, G. A.;
Prakash, G. K. S.; Rawdah, T. N.; Whittaker, D.; Rees, J . C. J . Am.
Chem. Soc. 1979, 101, 3935-3939.
(54) Baeyer, A.; Villiger, V. Ber. Dtsch. Chem. Ges. 1899, 32, 3625-
3633.
(55) Wolff, H. Org. React. (NY) 1946, 3, 307-336.
(50) For details about the parent system, tricyclo[3.2.1.02,7]octan-
3-one (16c), see: Moore, W. R.; Moser, W. R.; LaPrade, J . E. J . Org.
Chem. 1963, 28, 2200-2205.
(56) Notably, HN3 has also been used to form diaziridines, but from
thioketones. See: Middleton, W. J . U.S. Patent 3 226 439, 1965; Chem.
Abstr. 1966, 64, 9597. See ref 20c.

Regioselective Oxy and Imino Group Insertions of 3-Nortricyclanone
J . Org. Chem., Vol. 67, No. 2, 2002 455
[3.2.1.02,7]octan-3-one (16c).57 Like lactam 17a , ketone
17c also stems from the ostensibly unpreferred cyclo-
propyl migration.
a Hewlett-Packard HP 1090 Liquid Chromatograph, with a
Hewlett-Packard 35900E interface, equipped with a Spheri-
sorb S5W 5-µm column (250 × 4 mm). Isocratic elution giving
a 70% hexanes and 30% EtOAc mixture was employed at 1.0
mL/min. Product signals were observed using a HP 1047A RI
detector.
From AM1 calculations, the 4-azatricyclo[3.2.1.02,7]oct-
3-en-3-yl cation (9) and the 3-hydroxy-4-azatricyclo-
[3.2.1.02,7]oct-3-yl cation (13a ) benefit little thermody-
namically from delocalization of their cyclopropane rings
with their p-orbitals on C3.58 This may be due to an
overwhelming +R effect of N4 (Figure 2, compounds 9
and 13a ).59 Conversely, the 3-hydroxy-4-oxatricyclo-
[3.2.1.02,7]oct-3-yl cation (13b), which is putatively formed
during the acidic Baeyer-Villiger oxidation, is more
stable than the 4-hydroxy-3-oxatricyclo[3.2.1.02,7]oct-4-
yl cation (14b) by ca. 7 kcal.58 Perhaps this is due to the
-I effect of O4 (Figure 2, compound 13b).60 Accordingly,
lactone 16b is exclusively produced. In the same vein,
the 3-hydroxytricyclo[3.2.1.02,7]oct-3-yl cation (13c), which
is generated during the Tiffeneau-Demjanov ring expan-
sion, is more stable than the 4-hydroxytricyclo[3.2.1.02,7]-
oct-4-yl cation (14c) by ca. 5 kcal.58,61 However, ketone
17c is preferentially formed. As astutely stated by a
reviewer of this report, “It is a puzzle why the Baeyer-
Villiger and Demjanov reactions, which are mechanisti-
cally similar, give different regiochemistry.”
Rea ction of 3-Nor tr icycla n on e (1) w ith Acid ic HOSA.
Neat ketone 1 (0.541 g, 5.0 mmol) was injected into a solution
of HOSA (0.595 g, 5.0 mmol, 5.0 mL H2O) whereupon homog-
enization occurred within 1 min of swirling. A pH measure-
ment of the aqueous solution demonstrated that it was acidic
(pH ) 1), as expected from the liberation of H2SO4.63 After 1
h, the solution became turbid, and after 24 h, oil globules were
observed. Next, the mixture was extracted with CHCl3 (4 × 5
mL). The combined extracts were dried over anhydrous
MgSO4, filtered, analyzed by GC and GC-MS, and rotary
evaporated to give 0.493 g of a clear, uncolored liquid that
turned slightly orange upon standing. The product mixture
contained 3.1% of unreacted 1. Relative yield of products: 15
(43.1%), 16a (8.1%), 19 (17.7%), 20 (12.4%), 21 (17.2%), and
3-nortricyclanol (1.6%).64
Bicyclo[3.1.0]h ex-2-en e-en d o-6-ca r bon itr ile (15): Rf 0.59
(alkaline KMnO4; hexanes/EtOAc, 2:3); 1H NMR (CDCl3) δ 1.68
(1 H, dd, J ) 7.4, J ) 7.0), 2.08 (1 H, ddm, J ) 13.9, J ) 7.0),
2.43-2.47 (1 H, m), 2.51 (1 H, dm, J )18.8), 2.74 (1 H, ddm,
J ) 18.8, J ) 6.7), 5.78-5.84 (2 H, m); 13C NMR (CDCl3) δ
10.7 (d, J ) 173), 22.2 (d, J ) 171), 30.8 (d, J ) 177), 33.9 (t,
J ) 128), 117.6 (s), 128.2 (d, J ) 165), 132.4 (d, J ) 167); νjmax
/
A possible solution to this enigma may depend on the
nucleofugacity of the leaving group. Even though the
Baeyer-Villiger oxidation is mechanistically similar to
the Tiffeneau-Demjanov ring expansion, different leav-
ing groups are expelled (Table 1). In contrast, the
Schmidt rearrangement and the Tiffeneau-Demjanov
ring expansion both feature cyclopropyl migration via
cm-1 (neat) 3065, 2911, 2839, 2232, 1677, 802, 764; m/z (EI)
105 (M+, 67), 104 (100), 79 (14), 78 (74), 77 (42), 66 (5), 65
(31), 52 (16), 51 (16).
4-Aza tr icyclo[3.2.1.02,7]octa n -3-on e (16a ): Rf 0.18 (alka-
1
line KMnO4; CHCl3/MeOH, 97:3); mp 173-180 °C; H NMR
(400.1 MHz, CDCl3)65 δ 1.52 (2 H, d, J ) 11.9), 1.72-1.81 (3
H, m), 1.85 (2 H, d, J ) 7.6), 3.52-3.58 (1 H, m), 7.01 (1 H, br
s); 13C NMR (100.6 MHz, CDCl3)65 δ 17.2 (d, J ) 174), 21.8 (d,
departure of N2(g)
.
J ) 169), 31.7 (t, J ) 132), 46.3 (d, J ) 151), 171.7 (s); νjmax
/
cm-1 (KBr) 3162, 3052, 2933, 2858, 1670, 1654, 1625, 1475,
1321, 1272, 1089, 808; m/z (EI) 123 (M+, 100), 122 (5), 96 (5),
95 (41), 94 (58), 82 (8), 81 (18), 80 (42), 79 (47), 77 (24), 68
(30), 67 (23), 53 (23), (found M+ 123.0687, calcd for C7H9NO
123.0684). Found: C, 68.35; H, 7.44; N, 11.33%. Anal. Calcd
for C7H9NO: C, 68.27; H, 7.37; N, 11.37%.
Exp er im en ta l Section
Gen er a l In for m a tion . 3-Nortricyclanone (1) was prepared
according to the literature.15b,62 Hydroxylamine-O-sulfonic acid
(HOSA, 95%, Fluka) was used as purchased. FT-NMR spectra
were recorded on a Bruker Avance DPX-250 spectrometer (Bo
) 5.875 T) at the following Larmor frequencies: νo(1H) ) 250.1
MHz and νo(13C) ) 62.9 MHz. NMR coupling constants (J ) are
given in Hertz. IR absorption spectra were recorded on a
Perkin-Elmer 1600 FT-IR spectrometer by applying samples
as neat liquids onto a Si wafer or by pressing dilute KBr
pellets. Analytical GC analyses were conducted using a Fisons
3-H yd r oxyb icyclo[3.1.0]h e xa n e -en d o-6-ca r b on it r ile
(19): Rf 0.29 (alkaline KMnO4; hexanes/EtOAc, 1:2); 1Η ΝΜR
(CDCl3) δ 1.45 (1 H, tm, J ) 7.8), 1.85-2.01 (2 H, m), 2.08-
2.16 (1 H, m), 1.99 (2 H, ddm, J ) 14.4, J ) 5.6), 2.26 (2 H,
1
dd, J ) 14.4, J ) 6.8), 4.38-4.49 (1 H, m); H NMR (C6D6) δ
0.48 (1 H, tm, J ) 7.9), 0.98 (2 H, ddm, J ) 7.9, J ) 3.5),
1.04-1.08 (1 H, m), 1.52-1.64 (2 H, m), 1.93 (2 H, dd, J )
14.3, J ) 6.8), 4.11-4.23 (1 H, m); 13C NMR (CDCl3) δ 10.1 (d,
J ) 170), 24.2 (d, J ) 173), 36.1 (t, J ) 131), 73.1 (d, J ) 148),
119.3 (s); νjmax/cm-1 (neat) 3418, 3040, 2936, 2233, 1079; m/z
(EI) 123 (M+, 2), 122 (10), 105 (11), 104 (14), 96 (9), 95 (38),
94 (49), 83 (9), 82 (55), 81 (18), 80 (96), 79 (11), 67 (100), 57
(29).
8000 series GC instrument outfitted with
a 30 m poly-
(dimethylsiloxane) capillary column (Perkin-Elmer, PE-1, 0.32
mm i.d., and a 0.25 µm film thickness) and a flame-ionization
detector. The following conditions and temperature program
were used: Toven,i ) 50 °C (5 min), ramp ) +10 °C/min, Toven,f
) 230 °C (7 min), Tinjector ) 185 °C, TFID ) 260 °C, and a flow
) 2.1 (mL He)/min. GC-MS analyses were conducted using a
Hewlett-Packard 6890 series GC instrument outfitted with a
30 m poly(methylphenylsiloxane) capillary column (Hewlett-
Packard, HP-5MS (95% dimethyl and 5% diphenyl), 0.25 mm
i.d., and a 0.25 µm film thickness) and a Hewlett-Packard 5973
MSD instrument. The following GC conditions and tempera-
ture program were used: Toven,i ) 80 °C (7 min), ramp ) +10
°C/min, Toven,f ) 270 °C (5 min), Tinjector ) 200 °C, and an initial
flow ) 0.8 (mL He)/min. Analytical HPLC was performed using
2-H yd r oxyb icyclo[3.1.0]h exa n e-en d o-6-ca r b on it r ile,
Ep im er 1 (20): Rf 0.27 (alkaline KMnO4; hexanes/EtOAc,
1:2); 1H NMR (CDCl3) δ 1.49 (1 H, tm, J ) 7.5), 1.74 (1 H,
ddm, J ) 13.2, J ) 10.8), 1.85-2.11 (4 H, m), 2.15-2.35 (2 H,
m), 4.41 (1 H, dm, J ) 5.5); 1H NMR (C6D6) δ 0.53 (1 H, tm, J
) 8.0), 0.92-0.98 (1 H, m), 1.03-1.11 (1 H, m), 1.09-1.16 (1
H, m), 1.33 (1 H, ddm, J ) 14.1, J ) 9.0), 1.48-1.59 (1 H, m),
1.74-1.88 (1 H, m), 1.85-2.01 (1 H, m), 3.91-3.98 (1 H, m);
13C NMR (CDCl3) δ 6.1 (d, J ) 168), 24.0 (t, J ) 134), 25.0 (d,
J ) 176), 31.7 (d, J ) 173), 32.9 (t, J ) 134), 73.3 (d, J ) 151),
119.0 (s); νjmax/cm-1 (neat) 3418, 3040, 2936, 2233, 1079; m/z
(57) (a) Krow, G. R. Tetrahedron 1987, 43, 3-38. (b) Lumb, J . T.;
Whitham, G. H. Tetrahedron 1965, 21, 499-501. (c) Sauers, R. R.;
Beisler, J . A.; Feilich, H. J . Org. Chem. 1967, 32, 569-575.
(58) See Supporting Information.
(63) When performed, quenching with concentrated aqueous NH3
(10.0 mL, ca. 150 mmol) was conducted at this juncture.
(64) Hydroxycarbonitriles 19 and 21 gave a combined signal on both
GC columns. However, 19 and 21 gave well-separated signals on the
HPLC column, and thus their ratio was determined by HPLC integra-
tion.
(59) νjCO(16a ) ) νjCO(17a ) ) 1654 cm-1
(60) νjCO(16b) ) 1725 cm-1
(61) νjCO(16c) ) 1710 cm-1 (see ref 50); νjCO(17c) ) 1730 cm-1 (see
ref 57b).
.
.
(62) Schmerling, L.; Luvisi, J . P.; Welch, R. W. J . Am. Chem. Soc.
1956, 78, 2819-2821.
(65) Bruker Avance DRX-400 spectrometer (Bo ) 9.398 T).

456 J . Org. Chem., Vol. 67, No. 2, 2002
Rosenberg et al.
(EI) 123 (M+, 9), 122 (16), 105 (7), 104 (10), 96 (17), 95 (17),
94 (34), 83 (13), 82 (68), 81 (20), 80 (87), 79 (24), 67 (100), 57
(20).
5.0 mmol) was slowly dripped in. The ice bath twice needed
replenishing. Stirring was continued for 1 h (pH ) 11). The
mixture was refrigerated overnight (4 °C). The yellow solution
was then administered 5 M NaOH (1.0 mL, 5.0 mmol) and
refrigerated for another 2 days to destroy any remaining
NH2Cl, chloramine. Since no diaziridine 6 had precipitated,
the dark yellow mixture was neutralized with 6 M HCl (ca. 1
mL) to pH ) 7, warmed to 30 °C, and diluted with MeOH (40
mL). The mixture was placed in the freezer (-25 °C) for 2 days,
but still no precipitate was observed. So, the mixture was
rotary evaporated (35 °C) to dryness. The tan residue was
triturated with CH2Cl2 (3 × 8 mL), and the resulting light
yellow filtrate was analyzed by GC and GC-MS. (The residue
tested positive for NH4+ cation.) Relative yield of products: 16a
(17.3%), 17a (43.6%), 16b (27.7%), di(3-nortricyclanylidene)-
hydrazine (3.9%),43 unknown a (m/z ) 123, 3.6%), and un-
known b (m/z ) 124, 3.9%). Evaporation of the solvent left
0.200 g of yellow residue.
2-H yd r oxyb icyclo[3.1.0]h exa n e-en d o-6-ca r b on it r ile,
Ep im er 2 (21): Rf 0.23 (alkaline KMnO4; hexanes/EtOAc,
1:2); mp 48-54 °C; 1H NMR (CDCl3) δ 1.48 (1 H, dd, J ) 7.8,
J ) 7.5), 1.60-1.78 (1 H, m), 1.78-1.87 (1 H, m), 1.97-2.16
(1 H, m), 2.09 (2 H, dm, J ) 11.5), 2.16-2.30 (1 H, m), 4.80 (1
H, td, J ) 8.3, J ) 4.8); 13C NMR (CDCl3) δ 5.1 (d, J ) 169),
24.2 (t, J ) 134), 24.9 (d, J ) 172), 30.2 (t, J ) 132), 30.7 (d,
J ) 175), 74.5 (d, J ) 141), 120.1 (s); νjmax/cm-1 (neat) 3284,
3035, 2968, 2934, 2233, 1039, 822; m/z (EI) 123 (M+, 8), 122
(9), 105 (3), 104 (9), 96 (11), 95 (26), 94 (23), 91 (17), 83 (52),
82 (87), 81 (27), 80 (95), 79 (14), 67 (100), 57 (80), (found [M -
H]+ 122.0601, calcd for C7H8ON 122.0606).
3-Nor t r icycla n ol: Rf 0.36 (alkaline KMnO4; hexanes/
1
EtOAc, 2:3); mp 97-101 °C; H NMR (CDCl3) δ 0.98-1.04 (1
H, m), 1.08-1.18 (3 H, m), 1.21 (1 H, dm, J ) 10.6), 1.30 (1 H,
dm, J ) 10.6), 1.71-1.79 (2 H, m), 2.21-2.27 (1 H, m), 3.74-
3.80 (1 H, m); 13C NMR (CDCl3) δ 10.5 (d, J ) 177), 13.3 (d, J
) 176), 16.0 (d, J ) 177), 29.2 (t, J ) 133), 30.4 (t, J ) 133),
35.4 (d, J ) 148), 77.2 (d, J ) 150); νjmax/cm-1 (neat) 3330, 3062,
2986, 2940, 2868, 1454, 1307, 1290, 1134, 1077, 1049, 944, 908,
803; m/z (EI) 110 (M+, 38), 109 (21), 95 (55), 92 (33), 91 (70),
82 (21), 81 (63), 79 (100), 77 (46), 69 (20), 68 (20), 67 (73), 66
(87), 65 (22), 55 (43), 53 (25), 51 (13), (found M+ 110.0734, calcd
for C7H10O 110.0732).
Rea ction of 3-Nor tr icycla n on e (1) w ith Na OCl. Neat
ketone 1 (0.541 g, 5.0 mmol) was injected into a 0.57 M NaOCl
solution (27.0 mL, 15.3 mmol) that was submerged in an ice
bath. The reaction mixture was stirred for 2 h at 0 °C and
then for 46 h at room temperature in the dark. It was then
poured into 50 mL of H2O and extracted with CH2Cl2 (3 × 30
mL). The combined organic extracts were washed with 45 mL
of 0.67 M Na2S2O3 (to reduce unreacted hypochlorite) and then
with 45 mL of H2O. The organic layer was dried over
anhydrous MgSO4, filtered, and shown via GC and GC-MS to
contain 16b (95.9%) and 3-nortricyclanol (4.1%). After rotary
evaporation, the residue was free of 3-nortricyclanol giving
Rea ction of 3-Nor tr icycla n on e (1) w ith Acid ic HOSA
a n d Sa tu r a ted K2CO3. Neat ketone 1 (0.541 g, 5.0 mmol) was
injected into a solution of HOSA (0.595 g, 5.0 mmol, 5.0 mL
H2O) whereupon homogenization occurred within 1 min of
swirling. After 2 min, 8.1 m K2CO3 (1.31 g, 5.0 mmol) was
swirled in and the mixture was set in an ice bath. However,
crystal formation could not be initiated by scratching the cold
inner walls with a glass rod.
1
pure 16b (0.320 g, 51.6%) as a white solid: mp 68-69 °C; H
NMR (CDCl3) δ 1.74-1.88 (5 H, m), 2.00 (2 H, dm, J ) 7.5),
4.52-4.58 (1 H, m); 13C NMR (CDCl3) δ 18.5 (d, J ) 179), 21.3
(d, J ) 173), 30.7 (t, J ) 134), 73.5 (d, J ) 164), 170.9 (s);
νjmax/cm-1 (KBr) 3080, 3019, 2962, 2941, 2862, 1725, 1217,
1196, 1066, 811; m/z (EI) 124 (M+, 69), 96 (15), 81 (41), 80
(70), 79 (100), 77 (31), 68 (56), 67 (37), 54 (30), 53 (52), (found
M+ 124.0524, calcd for C7H8O2 124.0524).
Reaction of 3-Nor tr icyclan on e (1) with Alkalin e HOSA.
HOSA (5.95 g, 50 mmol) was dissolved in ice cold 1.0 M NaOH
(100 mL, 100 mmol). Neat ketone 1 (5.41 g, 50 mmol) was
immediately stirred into the effervescing solution, which was
subsequently stoppered and rapidly stirred in an ice bath.
Stirring was continued at room temperature for 48 h. The
mixture was subsequently extracted with CH2Cl2 (20 × 20
mL). The combined organic extracts were washed with H2O,
dried over anhydrous MgSO4, suction filtered, and analyzed
by GC and GC-MS. The product mixture contained 65.9% of
unreacted 1. Relative yield of products: 16a (18.6%), 17a
(37.3%), 3-nortricyclanol (9.1%), di(3-nortricyclanylidene)hy-
drazine (19.0%),43 unknown a (m/z ) 123, 9.3%), and unknown
b (m/z ) 124, 6.7%). After rotary evaporation, the residual
crystalline solid (1.17 g) was chromatographed (silica gel 60,
230-400 mesh) using a 97:3 mixture of CHCl3 and MeOH to
yield lactams 16a (0.392 g, 6.4%) and 17a (0.590 g, 9.6%) as
separate fractions.
Rea ction of 3-Nor tr icycla n on e (1) w ith HN3. CAUTION!
A three-necked round-bottom flask was equipped with a
stirbar, thermometer, balloon, and twistable solids addition
flask and set in an ice bath. It was charged with a mixture of
ketone 1 (0.541 g, 5.0 mmol, 10.0 mL CHCl3) and 50% H2SO4
(10.0 mL) at 0 °C. Sodium azide (0.390 g, 6.0 mmol) was slowly
“twisted” into the mixture whereupon N2 evolution was
observed. Afterward, stirring was continued for 2 h at at 0 °C
and then overnight at room temperature. The reaction mixture
was set in an ice bath, slowly neutralized with ca. 100 mL of
3 M NaOH, and then extracted with CH2Cl2 (2 × 30 mL). The
combined organic extracts were washed with H2O, dried over
anhydrous MgSO4, suction filtered, analyzed by GC and GC-
MS, and rotary evaporated to give 0.351 g of a brown oil.
Relative yield of products: 15 (13.1%), 17a (11.2%), 19+21
(18.4%), 20 (14.2%), unknown c (m/z ) 123, 3.8%), unknown
d (m/z ) 123, 5.3%), and unknowns e-j (m/z ) 228, 33.9%).
3-Aza tr icyclo[3.2.1.02,7]octa n -4-on e (17a ): Rf 0.27 (alka-
1
line KMnO4; CHCl3/MeOH, 97:3); mp 160-161 °C; H NMR
(400.1 MHz, CDCl3)65 δ 1.45 (2 H, dm, J ) 6.8), 1.54 (2 H, d,
J ) 12.4), 1.83 (2 H, dd, J ) 5.1, J ) 12.4), 2.56-2.62 (1 H,
m), 2.78 (1 H, dd, J ) 6.8, J ) 11.8), 7.58 (1 H, br s); 13C NMR
(100.6 MHz, CDCl3)65 δ 14.0 (d, J ) 172), 27.0 (t, J ) 133),
28.9 (d, J ) 178), 40.4 (d, J ) 145), 174.5 (s); νjmax/cm-1 (KBr)
3180, 3048, 2938, 2862, 1670, 1654, 1312, 1168, 1128, 846, 790;
m/z (EI) 123 (M+, 100), 96 (8), 95 (17), 94 (49), 82 (16), 81 (5),
80 (60), 79 (34), 77 (8), 69 (28), 68 (28), 67 (44), 56 (12), 55
(32), (found M+ 123.0686, calcd for C7H9NO 123.0684).
Found: C, 68.17; H, 7.32; N, 11.29%. Anal. Calcd for C7H9-
NO: C, 68.27; H, 7.37; N, 11.37%.
Ack n ow led gm en t . We thank the Fonds zur Fo¨r-
derung der wissenschaftlichen Forschung in O¨ sterreich
(Project P12533-CHE) for financial support. We are also
grateful to Dr. D. Krois, Mag. W. Knoll, and the late
Dr. A. Werner for their assistance. The reviewers of this
report were exceptionally helpful.
Su p p or tin g In for m a tion Ava ila ble: HETCOR spectra
of lactams 16a and 17a ; reaction paths for 9f8, syn -3f8, and
a n ti-3f12; computed conformational analysis of 8; and Z-
matrixes for 1, syn -3, a n ti-3, syn -7, a n ti-7, 8, iso-8, 9-12,
13a , iso-13a , 13b, 13c, 14a , iso-14a , 14b, 14c, 16a -c, and
17a -c with computed total energies. This material is available
free of charge via the Internet at http://pubs.acs.org.
Reaction of 3-Nor tr icyclan on e (1) with NH3 an d NaOCl.
CAUTION! Neat ketone 1 (0.541 g, 5.0 mmol) was injected
into a rapidly stirred 2.33 M NH3 solution (4.5 mL, 10.5 mmol)
that was submerged in an ice bath. The mixture became peach
colored (pH ) 9).66 Next, a 0.56 M NaOCl solution (8.9 mL,
(66) This might be ascribed to 3-nortricyclanone imine or its trimer
(see ref 2, footnote 48).
J O0159587