A tricycloheptane product in cationic rearrangements

Article abstract of DOI:10.1039/b309329h

The bicycloheptene 6 rearranged in acid to give the tricycloheptane 9, as shown by an X-ray crystal structure determination of the p-nitrobenzoate 10 derived from it. Earlier results in the literature had already indicated that this isomer was the thermodynamic sink. This apparently crowded structure, with its three contiguous quaternary centres, is, nevertheless, lower in energy than other accessible but less crowded structures, because of electronic stabilisation of the more substituted cyclopropane ring conjugated to the ester group.

Full text of DOI:10.1039/b309329h

View Article Online / Journal Homepage / Table of Contents for this issue
A tricycloheptane product in cationic rearrangements
John E. Davies, Ian Fleming* and Jonathan M. Goodman
Department of Chemistry, Lensfield Road, Cambridge, UK CB2 1EW.
E-mail: if10000@cam.ac.uk; Fax: ϩ44 (0)1223 336362; Tel: ϩ44 (0)1223 336372
Received 5th August 2003, Accepted 2nd September 2003
First published as an Advance Article on the web 19th September 2003
The bicycloheptene 6 rearranged in acid to give the tricycloheptane 9, as shown by an X-ray crystal structure
determination of the p-nitrobenzoate 10 derived from it. Earlier results in the literature had already indicated that
this isomer was the thermodynamic sink. This apparently crowded structure, with its three contiguous quaternary
centres, is, nevertheless, lower in energy than other accessible but less crowded structures, because of electronic
stabilisation of the more substituted cyclopropane ring conjugated to the ester group.
In 1964, we did not know the structure of the ester 6, and
Introduction
consequently had no idea what the structure of its rearrange-
ment product might be. We investigated the structure at that
time only briefly. It was an isomer, it had no protons attached to
trigonal centres, it was optically inactive, and reduction with
lithium aluminium hydride gave an alcohol, in which the methyl-
ene protons showed an AB pattern with no further coupling.
Not knowing the structure of its precursor 6, and with no fur-
ther information easily available, the structure of this com-
pound remained a mystery until now, when the ease with which
an X-ray crystal structure can be determined has made it
possible for us to assign the structure 9, derived from that
of its derivative, the p-nitrobenzoate 10. The crystal structure
also showed that, in addition to being optically inactive, the
p-nitrobenzoate 10 and, therefore. the ester 9 were racemic.
Two questions presented themselves: what is the pathway for
In 1964, at an early stage in Eschenmoser’s and Woodward’s
synthesis of vitamin B12, one of us prepared the ester 7 by the
sequence illustrated in Scheme 1.1 The bicyclic ester 6 was a by-
product derived from the diazoalkane 2, before the addition
of the sulfuric acid. The structure of this by-product and the
synthetic possibilities it suggested were the subjects of a sub-
sequent investigation,2 which identified it as the first example of
a reaction between a diazoalkane and an ester that matched the
better known reaction of diazoalkanes with ketones.
the cationic rearrangement 6
9, and why is the ester 9 the
thermodynamic product, when several less crowded isomers are
reasonably accessible? Hoyer reported in 1954 that the corre-
sponding acid 15, which he called dicadisäure, is the product
of acid-catalysed rearrangements of any of the acids 11–14
(Scheme 2).3 The ester 9 and the acid 15 are evidently sitting in
deep wells on the energy surface, in spite of their having three
contiguous quaternary centres.
Scheme 1 Reagents and conditions: i, NaOMe, MeOH, 0 ЊC, 1.25 h; ii,
H2SO4, H2O; iii, spontaneously over the 1.25 h in MeOH, before the
mixture was added to the H2SO4; iv, NaOMe, MeOH, reflux, 48 h;
v. TsOH, C6H6, reflux, 21 h; vi, LiAlH4. Et2O, rt, 1 h; vii, ClCOC6H4-
NO2-p, Py, 0 ЊC, 1 h.
Scheme 2 Reagents and conditions: i, 10% H2SO4 in H2O, reflux, few h.
A short and reasonable pathway for the formation of the
enantiomer of the ester 9 from the ester 6 is shown in Scheme 3,
which identifies a well-precedented 1,3-hydride shift to account
for the formation of both enantiomers of the ester 9.
The fate of the trans ester 8 during the treatment with tolue-
nesulfonic acid is uncertain. GC analysis indicated that it was
largely consumed during the 21 hour reflux, and the mass bal-
ance was closer to suggesting that, rather than being converted
into the desired ester 7, it was being converted into the ester 9.
A reasonable pathway exists for this transformation, including
the racemisation, but we were unable to isolate pure ester 8, and
could not check the conversion directly.
In the original work, the mixture of all four products 3–6 was
refluxed with sodium methoxide in methanol, in order to con-
vert the cis α,β-unsaturated esters 3–5 into their trans esters,
and then refluxed with toluenesulfonic acid in benzene for 21
hours, during which treatment the trans ester derived from the
cis ester 4 gave the desired ester 7. During this reflux with the
toluenesulfonic acid, the bicyclic ester 6 rearranged into a tri-
cyclic ester, the structure of which is the subject of this paper.
The esters were separated by fractional distillation on a spin-
ning-band column to give pure samples of the major product 7
and the new ester derived from the bicyclic ester 6.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 5 7 0 – 3 5 7 1
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
3570

View Article Online
Experimental
Methyl 2,7,7-trimethyltricyclo[2.2.1.02,6]heptan-2-carboxylate 9
The tricyclic ester 9 (8.4 g)1 was redistilled on the spinning band
column (bath temperature 127 ЊC, column temperature 85 ЊC,
drip ratio 1 : 20), collecting fractions over 12 h. The second
fraction (3.45 g) was a pure sample of the ester, bp 101–103 ЊC/
16 mmHg (lit.3 100–101 ЊC/17 mmHg); νmax(CC14)/cmϪ1 1720
(CO); δH(60 MHz; CCl4) 3.55 (3 H, s, OMe), a methylene envel-
ope, 1.12 (3 H., s, Me) and 0.83 (6 H, s, 2 × Me).
2,7,7-Trimethyltricyclo[2.2.1.02,6]heptan-2-carboxylic acid 15
Scheme 3
That leaves the intriguing question of why the congested ester
9 should be thermodynamically favoured over several plausible,
but superficially less congested, isomers 16–20. We suggest that
it is related to the well known property of cyclopropanes in
which they resemble alkenes. We calculated the relative energies
for six tricyclic isomers at the B3LYP/6-31G** level,4 with the
results shown in Fig. 1. The isomer 9, in the conformation 9a
with the carbonyl group pointing up, reassuringly has the
lowest energy.
The ester (2.6 g) was refluxed in methanol (30 ml) with aqueous
sodium hydroxide (10%, 10 cm3) for 3 h, and the methanol
distilled off under reduced pressure. The residue was cooled to
room temperature and washed with ether. The aqueous layer
was acidified with hydrochloric acid (3 mol dmϪ3) and extracted
with ether. The ether was dried (Na2SO4) and evaporated, to give
the acid (1.93 g, 83%) as prisms, mp 135–136Њ (from n-hexane)
(lit.3 mp 139–140 ЊC, from AcOH); νmax(CC14)/cmϪ1 3300–2500
(br, OH) and 1675; δH(60 MHz; CCl4) no absorption at lower field
than δ 2.1, 1.3 (3 H, s, Me) and 0.9 (6 H, s, 2 × Me); [α]25(c. 10 in
EtOH) no rotation between 365 and 589 nm (Found: C, 73.33; H,
8.99. C11H16O2 requires C, 73.30; H, 8.95%).
2,7,7-Trimethyltricyclo[2.2.1.02,6]heptan-2-ylmethyl
p-nitrobenzoate 10
The acid (100 mg) was added to a solution of diazomethane in
ether (25 cm3) and the ether evaporated. The residue, which
showed negligible rotation, was stirred with lithium aluminium
hydride (0.1 g) in ether (10 cm3) at room temperature for l h.
Water was added until the inorganic material settled, the ether
was decanted off, and evaporated to give the alcohol as an oil;
δH(60 MHz; CCl4) 3.65 (2 H, s CH2O), 2.78 (1 H, s, OH), 1.03
(3 H, s, Me), 0.83 (6 H, s, 2 × Me) and 0.70 (1 H, m, cyclo-
propane-H), which was stirred with p-nitrobenzoyl chloride
(0.45 g) in pyridine at 0 ЊC for l h. Water and ether were added,
the aqueous layer was washed with aqueous acid, and the
organic layer evaporated to give the p-nitrobenzoate (0.58 g,
89%), as leaflets, mp 125–127 ЊC (from MeOH); νmax(CC14)/
cmϪ1 1720 (CO), 1610 (Ar) and 1535 (NO2); δH(60 MHz; CCl4)
8.21 (4 H, m. ArHs); 4.8 (1 H, d, J 11, OCHAHB), 4.3 (1 H, d,
J 11, OCHAHB), 1.15 (3 H, s, Me), 0.91 (3 H, s, Me) and 0.90
(3 H, s, Me); [α]23 (c. 1.4 in CHCl3) 0 at all wavelengths between
436 and 589 nm (Found: C, 66.71; H, 6.63; N, 4.52. C18H21NO4
requires: C, 68.55; H, 6.71; N, 4.44%).
Fig. 1 Relative energies in kJ molϪ1 of isomeric tricyclic esters
C12H18O2.
This result is consistent with the three membered ring having,
like a double bond, a thermodynamic preference to be both
conjugated and substituted. This is illustrated by the isodesmic
reactions A–F (Fig. 2). A and B show that esters have a thermo-
dynamic preference to be adjacent to double bonds and cyclo-
propanes, and that the effect is larger for cyclopropanes. Simi-
larly, C and D show both double bonds and cyclopropanes
benefit from methyl substitution. This preference is slightly
increased by the presence of an ester on the far side of the ring
(E), and reduced if the ester is on the same side (F). These
results are all consistent with the thermodynamic preference
calculated for the isomer 9a.
Crystal data for 10. C18H21NO4, M=315.36, triclinic, space
¯
group P1, a = 7.0666(4), b = 7.0975(4), c = 18.3293(13) Å, U =
816.98(9) Å3, Z = 2, µ(Mo-Kα) = 0.091 mmϪ1, 7013 reflections
measured at 180(2) K using an Oxford Cryosystems Cryo-
stream cooling apparatus, 2762 unique (Rint = 0.068); R1 = 0.11,
wR2 = 0.267 [I>2σ(I )]. The structure was solved with SHELXS-
97 and refined with SHELXL-97.5 CCDC reference number
216824. See http://www.rsc.org/suppdata/ob/b3/b309329h/ for
crystallographic data in .cif or other electronic format.
References
1 I. Fleming and R. B. Woodward, J. Chem. Soc., Perkin Trans. 1, 1973,
1653.
2 E. H. Billett and I. Fleming, J. Chem. Soc., Perkin Trans. 1, 1973,
1658; E. H. Billett, I. Fleming and S. W. Hanson, J. Chem. Soc.,
Perkin Trans. 1, 1973, 1661; I. Fleming and S. W. Hanson, J. Chem.
Soc., Perkin Trans. 1, 1973, 1669.
3 H. L. Hoyer, Chem. Ber., 1954, 87, 1849.
4 Jaguar version 4.2: Schrodinger, Inc., Portland, Oregon, 2000.
5 G. M. Sheldrick, SHELXS-97/SHELXL-97 Programs for solution/
refinement of crystal structures, University of Göttingen, Germany,
1997.
Fig. 2 Isodesmic reactions; energies in kJ molϪ1
.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 5 7 0 – 3 5 7 1
3571