11B
10A
10A
10B
instrument used a 25 m capillary column with 0.3 mm internal
diameter, coated with OV351 or with FFAP. Preparative separ-
10B
O
11A
1
1
O
2
6
2
6
ations were carried out on a Perkin-Elmer F21 chromatograph
3
3
1
9
9
–
using a glass column of inch internal diameter containing
7A
7B
4A
7A
7B
4A
Carbowax 20M coated on4a Celite support. Mass spectra were
measured on a Fisons Trio 1000 spectrometer, coupled to a
Fisons 3800 gas–liquid chromatograph.
8
8
4B
4B
5
5
11
10
Preparation of materials
Samples of nopol and myrtenal were obtained from Aldrich,
and were used without further purification.
formation of 1-(cyclopropyl)ethanol by deamination of pent-3-
enylamine.9
The interesting question is why the double bond assisting
departure of a toluene-p-sulfonate group gives rise to a
cyclobutyl system (which rapidly rearranges) while the double
bond assisting departure of a diazo group gives rise to a spiro-
cyclopropyl system. Hanack and Schneider10 have shown that
both reactions can give both systems; the spirocyclopropyl
system is often produced on its own10,11 but cyclobutyl systems
are rarely the sole product, and then usually result from deamin-
ation rather than the toluene-p-sulfonate acetolysis. We are
reluctant to propose a ‘low energy’ carbocation, and prefer the
theory of Kirmse and Voigt,12 who pointed out that heterolysis
of a toluene-p-sulfonate to yield a carbocation involves an
appreciable change in the geometry of the reactant; in other
words, the transition date for the reaction is ‘late’ on the reac-
tion coordinate. Decomposition of a diazonium ion, on the
other hand, involves an excellent leaving group, and therefore
passes its transition state early on the reaction coordinate and
without significant distortion of the geometry of the substrate.
Consequently, deamination can produce carbocations which
are bypassed in solvolysis.
Our results provide an excellent illustration of this theory.
Nitrogen is a very good leaving group, so deamination proceeds
with loss of nitrogen at an early point on the reaction coordin-
ate along the lines suggested in Scheme 2. The toluene-p-
sulfonate group is an inferior leaving group, and leaves at a
much later point on the reaction coordinate, not, in fact, until
substantial electron shift has occurred, as a primary alkyl car-
bocation would be too unstable to be formed. Consequently,
shift of the methylene bridge from C-1 to C-2 commences
before separation of the leaving group is accomplished. Thus,
the pinane cyclobutyl carbocation is not formed; the rearrange-
ment sweeps through to yield the product 2.
An alternative view of the situation is that nitrogen is a very
good leaving group, so is easily displaced by shift of the elec-
trons of the double bond. The toluene-p-sulfonate is much
more difficult to displace with a nucleophile1 and the electrons
of the double bond are insufficient to accomplish this task until
reinforced by the extra electrons supplied by the shift of the
methylene bridge from C-1 to C-2.
Removal of the toluene-p-sulfonate group by a trialkyl alu-
minium compound would probably come between these two
reactions, and would permit ion 7 to be formed without partici-
pation of a methylene bridge shift. Since the reactions took
place in benzene or dichloromethane, the ion would probably
have sufficient lifetime to equilibrate to yield all possible prod-
ucts, as was observed.
Nopyl chloride 8
Nopol (49.8 g), triphenylphosphine (86.5 g) and carbon tetra-
chloride (175 ml) were refluxed together for 1 h.13 In this time, a
white solid precipitated out of solution. The flask was allowed
to cool, and the solid filtered off. The solvent was removed to
leave a colourless oil (45.7 g, 83%), which was purified by distil-
lation at 59 ЊC at 0.8 Torr to give nopyl chloride (lit.,14 bp 56–
60 ЊC at 0.8 Torr), δH 0.92 and 1.36 (both 3H, s); νmax/cmϪ1 3000,
2900, 2810, 1645, 1460, 1440, 1425, 1375, 1360, 1285, 1260,
1235, 1225, 1195, 1175, 1130, 1110, 1090, 1075, 1035, 1020, 950,
900, 880, 800, 785, 770, 730 and 655; δC(CDCl3) 144.3 (s, C-2),
119.1 (d, C-3), 45.5 (d, C-1), 42.4 (t, C-11), 40.6 (d, C-5), 40.0 (t,
C-10), 37.8 (s, C-6), 31.5 (t, C-4), 31.2 (t, C-7), 28.1 (q, C-8) and
21.0 (q, C-9). Assignments of C-4 and C-7 could be reversed.
N-Nopylphthalimide
Nopyl chloride (45 g) was converted into N-nopylphthalimide
by the method of Cope and Burrows,14 giving a dark oil (82.5
g), νmax/cmϪ1 3000, 2900, 2810, 1765, 1700, 1605, 1460, 1440,
1425, 1390, 1350, 1330, 1300, 1260, 1210, 1195, 1180, 1165,
1120, 1090, 1080, 1000, 950, 935, 880, 860, 785, 750, 710 and
655; δC(CDCl3) 166.2, 133.5, 131.9, 118.6, 144.3 (s, C-2), 122.8
(d, C-3), 45.3 (d, C-1), 40.4 (d, C-5), 37.6 (s, C-6), 36.0 (t, C-11),
35.1 (t, C-10), 31.4 (t, C-4), 31.0 (t, C-7), 25.9 (q, C-8) and 20.7
(q, C-9). Assignments of C-4 and C-7 could be reversed.
This was prepared from N-nopylphthalimide by the published
route.14 N-Nopylphthalimide (82.5 g) gave a yield of 27.2 g
(55%) of nopylamine hydrochloride, mp 231–239 ЊC; δH 0.76
(3H, s), 1.20 (3H, s), 5.3 (1H, br) and 8.18 (3H, br); νmax(KBr
disc)/cmϪ1 2900, 1570, 1465, 1375, 1355, 1255, 1210, 1195, 1135,
1120, 1085, 1070, 1035, 1005, 950, 940, 920, 880, 840, 800, 770,
700 and 600.
2-(1-Acetoxyethyl)-6,6-dimethylbicyclo[3.1.1]hept-2-ene 9
Myrtenal (7.5 g) in anhydrous diethyl ether (20 ml) was added
dropwise to an ethereal solution of methylmagnesium iodide
made from magnesium (1.7 g) and methyl iodide (10 g). After
refluxing for 1 h, the mixture was poured onto an ice–dilute
sulfuric acid mixture, then extracted with diethyl ether. The
diethyl ether extracts were washed with water, dried over
magnesium sulfate, and the diethyl ether removed to give the
alcohol as a yellow oil (7.9 g), νmax/cmϪ1 3350, 2900, 1460, 1440,
1375, 1360, 1200, 1060, 1040, 1000, 980, 905, 875 and 800.
Acetylation with acetyl chloride in pyridine gave the acetate,
νmax/cmϪ1 2980, 2810, 1720, 1440, 1360, 1230, 1035, 1010, 935
and 800.
Experimental
NMR spectra were recorded on a Bruker AMX 400 spec-
trometer, the 1H spectra at 400 MHz and the 13C spectra at 100
MHz. Spectra were recorded in CDCl3 or C6D6 with SiMe4 as
internal standard. COSY, DFTP and decoupling experiments
were used to aid assignment.
Infrared spectra were recorded on a Perkin-Elmer 1320 spec-
trometer or a Perkin-Elmer Paragon 1000 FT-IR spectrometer.
Unless otherwise stated, solid samples were run as a Nujol mull,
and liquid samples as neat liquid. Reaction mixtures were ana-
lysed on a Dani 3800 gas–liquid chromatograph with a flame
ionisation detector using nitrogen as the carrier gas. The
Deamination with nitrous acid
The amine, or its hydrochloride, was dissolved in dry acetic
acid, and to the stirred solution at room temperature was added
sodium nitrite over a period of 30 min, and the solution stirred
for 2 h. The solution was poured into water and extracted with
diethyl ether. The combined extracts were washed with satur-
ated aqueous sodium hydrogen carbonate then water and dried
over magnesium sulfate. The volume of solvent was reduced,
and the products examined by GLC.
J. Chem. Soc., Perkin Trans. 1, 1998
645