Mechanisms in Cyclopropylcarbene Reactions
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6475
singlet carbene. Crossover experiments have shown that
aromatization of cyclohexadienylidenes proceeds by initial
fragmentation into a pair of radicals, which then diffuse apart,
finally to recombine to give the product.16 Even here, the
identity of the reacting spin state is not known with certainty.
lecular chemistry of cyclopropyldiazomethanes does not involve
carbenes.
It appears that the diradical mechanism for conjugated
cyclopropylcarbene rearrangements, proposed by us long ago,
is correct. It remains to be seen if more simple cyclopropyl-
carbenes react in similar fashion.
Experimental Section
General. Photolysis was carried out with a 450-W medium-pressure
mercury arc (Hanovia lamp). Melting points were determined on a
Thomas-Hoover Uni-melt capillary melting point appartus and are
uncorrected. 1H NMR spectra were obtained on a General Electric
QE 300 spectrometer at 300 MHz. Gas chromatographic/mass
spectrometric analyses were performed on a Hewlett-Packard 5890/
5971 Series II gas chromatograph/mass spectrometer with a HP-1701
capillary column (30 m × 0.25 mm i.d., 0.25 mm film thickness).
Preparative gas chromatography was performed on a Gow-Mac 580
Despite seemingly ample opportunity, rearranged products
do not appear often. Thus the chrysanthemylcarbenes (19) give
none of the cyclohexadienes that might have been anticipated
from a diradical intermediate.3 However, in an acyclic system
such as this one, it seems likely that the allylic radical, if formed
at all, would be primarily in the trans arrangement that is unable
to close to the six-membered ring.
1
gas chromatograph with an aluminum column (6 ft × /4 in.) packed
with 10% OV-101 on Chromosorb WHP. Precise masses were
measured on a KRATOS MS50 RFA high-resolution mass spectrom-
eter. GC/FTIR measurements were made on a Hewlett-Packard HP
5898 gas chromatograph (30 m × 0.32 mm id, 0.5 mm film thickness)
with a Restek rtx-1 column connected to a Nicolet 730 FTIR
spectrometer. Ethyl diazoacetate, 2-vinylnaphthalene, p-toluenesulfon-
hydrazide, PCC, and NaH were purchased from Aldrich Chemical Co.
Solvents (Et2O, CH2Cl2, THF) were dried and distilled prior to use.
Ethyl 1-Benzonorcaradienecarboxylate.17 To molten naphthalene
(750 g, 5.78 mol) and anhydrous CuSO4 (140 g, 880 mmol), at 150 °C
under argon, was added dropwise ethyl diazoacetate (100 g, 867 mmol)
over 1 h with mechanical stirring. The dark slurry was stirred for 4 h
at 150 °C and then cooled to ambient temperature. The mixture was
distilled under reduced pressure to remove naphthalene: bp 110-130
°C (25 Torr). The ester was collected as a yellow oil: bp 140-160
°C (2.5 Torr). Redistillation of this material yielded 37 g (20%) of
ethyl 1-benzonorcaradienecarboxylate as a pale yellow oil: 1H NMR
(300 MHz, CDCl3) δ 0.87 (q, 2H), 1.30 (t, 3H), 2.65 (m, 1H), 3.08
(m, 1H), 4.2 (d, 1H), 6.32 (m, 1H), 6.41 (d, 1H), 7.0-7.5 (m, 4H).
1-(Hydroxymethyl)benzonorcaradiene. Ethyl 1-benzonorcaradi-
enecarboxylate (37 g, 173 mmol) was dissolved in anhydrous Et2O
(35 mL) and added dropwise to a mixture of LiAlH4 (10 g, 263 mmol)
in refluxing anhydrous Et2O (70 mL) over 1.5 h. Reaction was followed
by TLC (50/50 EtOAc/hexanes). Reflux was maintained for another
1.5 h, at which time 10% NH4Cl (10 mL) was added to quench the
reaction. The mixture was filtered to remove the product (a white
precipitate) which was washed with excess ether and CH2Cl2. The
aqueous layer was extracted with Et2O (30 mL), and the combined
organic layers were dried over Na2SO4. The product was concentrated
to yield 25.5 g (mp 64-67 °C, 86.6%). This material was used without
further purification in the next step: 1H NMR (300 MHz, CDCl3) δ
0.34 (m, 1H), 1.17 (m, 1H), 1.39 (s, 1H), 1.86 (m, 1H), 2.27 (m, 1H),
By contrast, in all the cases in which rearrangement is
observed, the allylic portion of the putative diradical must be
generated in the productive cis arrangement:
More puzzling is the absence of rearrangement in carbenes
such as 20 in which a rearrangement-capable allylic radical is
possible.1 Perhaps the transition state for formation of the
diradical requires more delocalization than the single double
bond in 20 can provide; perhaps small amounts of rearranged
products, or vinyl compounds, have been overlooked.
3.68 (m, 2H), 6.18 (m, 1H), 7.0-7.5 (m, 4H); IR (neat) 3300 (br) cm-1
.
1-Benzonorcaradienecarboxaldehyde. Crude 1-(hydroxymethyl)-
benzonorcaradiene (25.5 g, 148 mmol) dissolved in dry CH2Cl2 (150
mL) was rapidly charged into a heterogeneous mixture of PCC (47.9
g, 1.5 equiv) in dry CH2Cl2 (300 mL). The mixture immediately turned
black. The reaction was followed by TLC (50/50 EtOAc/hexanes) and
diluted with 3 × 400 mL of Et2O upon completion (1 h). Crude product
was filtered through Flourisil and concentrated under reduced pressure
to yield 22 g (87%) of a black liquid. A sample of 5 g of this crude
material was purified by flash column chromatography (90/10 hexanes/
EtOAc). A total of 3.5 g of pure material, mp 34-36 °C, was
recovered: 1H NMR (300 MHz, CDCl3) δ 1.18 (m, 1H), 2.86 (m, 1H),
3.28 (m, 1H), 6.33 (q, 1H), 6.47 (d, 1H), 7.0-7.5 (m, 4H), 9.91 (s,
1H). 13C NMR (75 MHz, CDCl3) d 202.7, 132.4, 131.0, 128.7, 128.2,
128.0, 127.2, 126.7, 125.9, 32.7, 31.3, 29.8. HRMS (EI) calcd for
C12H10O 170.07315, found 170.07341.
Finally, these days it is obligatory to bring up the question
of whether carbenes are really involved at all. This problem is
especially acute when intramolecular chemistry is at issue. There
is ample evidence that cyclobutene formation from decomposi-
tion of cyclopropyl diazomethanes involves either complete5b
or substantial5a direct rearrangement of the diazo compound.
Cyclobutene 12 is surely involved in the overall reactions of
10 and 11, particularly in the formation of benzocyclooctatet-
raene. It is very unlikely that 12 is the source of the vinyl
compounds or benzobarrelene, which do seem to derive from
cyclopropylcarbenes. Indeed, one of the reasons these stepwise
reactions have been so difficult to find is that most of
“cyclopropylcarbene” chemistrysring expansion to cyclobuteness
is in reality diazo compound chemistry. Much of the intramo-
1-Benzonorcaradienecarboxaldehyde Tosylhydrazone. The al-
dehyde (3.5 g, 20.6 mmol, 1.0 equiv) dissolved in a minimal amount
(16) Berdick, T. E.; Levin, R. H.; Wolf, A. D.; Jones, M., Jr. J. Am.
Chem. Soc. 1973, 95, 5087.
(17) Doering, W. v. E.; Goldstein, M. J. Tetrahedron, 1959 5, 53.