Thermolysis of a spiro-fused oxadiazoline: The carbonyl ylide-cyclic carbene-diradical sequence

Article abstract of DOI:10.1139/v02-126

Thermolysis of spiro-fused oxadiazoline 1 in benzene led to loss of N₂ to form a carbonyl ylide intermediate. Most of the ylide fragmented to acetone and 4-phenyl-1,3-dioxane-2-ylidene, which could be trapped with tert-butyl alcohol. In the absence of the trapping agent, the major pathway followed by the carbene was fragmentation to a diradical, 5-phenyl-2-oxa-1-oxo-1,5-pentanediyl. The latter diradical coupled to α-phenyl-δ-butyrolactone and decarboxylated to afford cyclopropylbenzene. Other products from the reaction were those of apparent insertion of the carbene into a C-H bond of the benzene solvent and into a C-H bond of acetone. Such reactions appear to be without precedent - alternative, non-carbene mechanisms are proposed.

Full text of DOI:10.1139/v02-126

1187
Thermolysis of a spiro-fused oxadiazoline: The
carbonyl ylide–cyclic carbene–diradical sequence
Nadine Merkley and John Warkentin
Abstract: Thermolysis of spiro-fused oxadiazoline 1 in benzene led to loss of N2 to form a carbonyl ylide intermedi-
ate. Most of the ylide fragmented to acetone and 4-phenyl-1,3-dioxane-2-ylidene, which could be trapped with tert-
butyl alcohol. In the absence of the trapping agent, the major pathway followed by the carbene was fragmentation to a
diradical, 5-phenyl-2-oxa-1-oxo-1,5-pentanediyl. The latter diradical coupled to ꢀ-phenyl-ꢁ-butyrolactone and
decarboxylated to afford cyclopropylbenzene. Other products from the reaction were those of apparent insertion of the
carbene into a C–H bond of the benzene solvent and into a C–H bond of acetone. Such reactions appear to be without
precedent — alternative, non-carbene mechanisms are proposed.
Key words: dioxacarbene, carbonyl ylide, cyclopropylbenzene, diradical, lactone, oxadiazoline.
Résumé : La thermolyse de l’oxadiazoline à fusion spiro, 1, dans le benzène, conduit à une perte d’azote et à la for-
mation d’un intermédiaire carbonyl-ylure. L’ylure se fragmente principalement en acétone et en 4-phényl-1,3-dioxane-2-
ylidène que l’on peut piéger avec de l’alcool tert-butylique. En absence de piège, la voie réactionnelle principale suivie
par le carbène en est une de fragmentation en biradical, le 5-phényl-2-oxa-1-oxopentane-1,5-diyle. Ce dernier radical
peut se coupler à l’ꢀ-phényl-ꢁ-butyrolactone et il se décarboxyle pour former du cyclopropylbenzène. Parmi les autres
produits de la réaction, on a observé ceux résultant de l’insertion apparente du carbène dans une liaison C–H du
benzène agissant comme solvant et dans une liaison C–H de l’acétone. De telles réactions ne semblent pas avoir été
observées antérieurement et on propose des mécanismes alternatifs n’impliquant pas de carbène.
Mots clés : dioxacarbène, carbonyl-ylure, cyclopropylbenzène, biradical, lactone, oxadiazoline.
[Traduit par la Rédaction] Merkley and Warkentin 1195
Introduction
The hydrazinolysis of 9 was also selective, affording primar-
ily 10, which crystallized from solution during the reaction.
Oxidation of 11 with Pb(OAc)4 led to a mixture of 1 and 12.
The intramolecular acid catalyzed substitution, leading from
12 to 1 to complete the cyclization, was modeled after
intermolecular analogues (6).
Thermolysis of oxadiazolines (1) is now a well-established
method for the generation of alkoxy- or dialkoxycarbenes
(1). Fragmentation of alkoxy- or dialkoxycarbenes to radi-
cals was reported many years ago, primarily for gas phase
conditions (2), and computations for model systems followed
(3). Such fragmentations of dialkoxycarbenes to radical pairs,
in solution under relatively mild conditions, have recently
been studied experimentally (4) and modeled (without solva-
tion) with smaller analogues (3). Fragmentation of cyclic
oxy- or dioxycarbenes to corresponding diradicals appears to
be limited to a theoretical study by Borden and Hoo (5).
We now report the synthesis and thermolysis of oxadia-
zoline 1 in benzene at 110°C. The thermolysis occurs
through a number of intermediates that are formed sequen-
tially including a carbonyl ylide (2), a carbene (3), and
diradicals 4 and 5 (Scheme 1).
A sample of 1 in benzene-d6, in a sealed tube, was heated
1
at 110°C for 72 h before an H NMR spectrum of the crude
mixture of products was acquired. The solvent and volatile
products were then evaporated and the residue was
chromatographed on SiO2.
Results and discussion
Evidence for an ylide intermediate (2) came from the find-
ing that a minor product was ketone 15 (Scheme 3). The yield
of that product was not affected by inclusion of acetone (1.1 M)
in the benzene solution of 1, indicating that it could not have
arisen from insertion of 3 into a C–H bond of acetone.
Moreover, it could not have arisen from any reaction of free
acetone.
Methods
Oxadiazoline 1 was synthesized by means of the sequence
of reactions shown in Scheme 2. Ring opening of 6 was se-
lective, leading primarily to 7, which was saponified to 8.
We suggest that 15 arose from 2 through 13, which came
from a 1,4-sigmatropic H-shift in 2. Such rearrangements of
carbonyl ylides are known (7). Apparently, a fully concerted
Received 2 April 2002. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 16 August 2002.
N. Merkley and J. Warkentin.1 Department of Chemistry, McMaster University, 1280 Main St. W., Hamilton, ON L8S 4M1,
Canada.
1Corresponding author (e-mail: warkent@mcmaster.ca).
Can. J. Chem. 80: 1187–1195 (2002)
DOI: 10.1139/V02-126
© 2002 NRC Canada

1188
Can. J. Chem. Vol. 80, 2002
Scheme 1.
Ph
Ph
Ph
Ph
Ph
-Me2CO
.
O
O
N
O
O
-N2
O
O
.
.
O
O
O
Ϯ
.
..
O
N
3
4
5
1
2
Scheme 2.
Ph
Ph
Ph
Ph
Ac2O
H2CO
KOH
MeOH
O
OAc
O
O
+
H
H2SO4
OH OH
OAc
6
7
8
Ph
Ph
Ph
(Im)2CO
1. H2NNH2
2. Me2CO
OH
LTA
CH2Cl2
TFA
CH2Cl2
8
1
OAc
O
O
OH
O
O
O
N
NH
NR2
10 R=H
11 R2= CMe2
O
O
N
9
12
Scheme 3.
Ph
Ph
Ph
Ph
O
O
O
+
O
H
O
O
H
O
O
O
O
H
_
_
+
O
O
2
13
14
15
turn of the ion pair to 16 presumably scrambles the oxygen
atoms of the acetoxy group whereas the enolate fragment in
Scheme 3 would bond to carbon to form 15 irreversibly.
Scheme 4.
O
The formation of diastereomers 19 from thermolysis of 1
in benzene containing tert-butyl alcohol is evidence for
carbene 3, whereas the formation of lactone 20 and cyclo-
propylbenzene (21) in the absence of tert-butyl alcohol indi-
cates that 3 fragments to diradical 4 (Scheme 5). There is
analogy for the fragmentation of 3 to diradical 4 in the frag-
mentation of cinnamyloxy(methoxy)carbene to cinnamyl and
methoxycarbonyl radicals and of benzyloxy(methoxy)carbene
to benzyl and methoxycarbonyl radicals (4).
Cyclopropylbenzene (21) is an expected product from 5,
which is an intermediate from decarboxylation of 4. We
were unable to detect (by GC, with authentic samples for
reference) either 1- or 3-propenylbenzene, which might also
have arisen from 5 by means of intra- or intermolecular
H-transfer. Reversal of 4 to 3 is speculative but, because 4
must be born with the geometry shown and because the bar-
rier to radical coupling cannot be large, the reversal step
(4aJ3) is a probable, albeit invisible, reaction.
O
O
H
O
O
O
O
+
O
O
H
_
O
O
H
O
16
17
18
fragmentation of the oxadiazoline, to the carbene, N2, and
acetone, as suggested by Smith (8) on the basis of computa-
tion, is not general. Rearrangement of 13 to 15 is unprece-
dented, to the best of our knowledge, but it is reasonable,
given that both fragments of ion pair 14 that are proposed
for the rearrangement are stabilized. A good model system is
the 2-acetoxy acetal 16, thermolysis of which at 120°C af-
fords the mixed diester 18 in 68% yield (9). The reaction
can be understood in terms of reversible formation of an ion
pair and nucleophilic substitution at carbon (Scheme 4). Re-
© 2002 NRC Canada

Merkley and Warkentin
1189
Scheme 5.
Ph
Ph
Ph
Ph
5
.
O
4
.
6
O
t-BuOH
Ph
O
3
1
O
O
O
O
O
.
2
.
8 9
OC(Me)3
..
O
H
O
3
4a
4b
20
19
Ph
.
Ph
Ph
Ph
.
+
X
.
21
5
Scheme 6.
Scheme 7.
Ph
Ph
PhH
O
O
O
O
O
O
..
3
PhH
110°C
O
N
O
O
Ph
22
O
O
+
O
O
H2O
in PhH
O
O
..
H OH
N
PhH
Ph
X
25
28
Ph
OH
O
H
27
SiO2
O
H
O
O
OH
23
24
O
O
H
Ph
The formation of 20 and 21 in roughly equal amounts
means that the rate constants for cyclization and decarbo-
xylation of 4 at 110°C are very similar. The rate constant for
decarboxylation was estimated as follows. From the
Arrhenius equation of Newcomb (10) we obtained the rate
constant kH at 110°C for the reaction of RCH2OCO radicals
with HSnBu3 (kH = 1.77 × 106 M–1 s–1). The ratio kH/kCO2
was available from the product ratio RCH3:RCH2OCHO =
38:25 for the reaction of RCH2OCO radicals with HSnBu3
(average concentration 0.025 M) in toluene at 110°C (11).
26
Additional products, not shown in the previous Schemes,
were diastereomeric 2,4-diphenyl-1,3-dioxanes (22) (12) and
diastereomeric hemiorthoformates (4-phenyl-1,3-dioxan-2-
ols) (23). The latter products are readily accounted for in
terms of capture of carbene 3 by adventitious water. Hemi-
orthoformates (23) were thermally stable, but an attempt at
isolation by chromatography on silica led to 1-phenyl-1,3-
propanediol-3-formate (24) (Scheme 6).
Thus, the rate constant for decarboxylation came to kCO2
~
6.8 × 104 s–1, and the upper limit of the rate constant for
cyclization of 4, to form 20, could be set at about 1 × 105 s–1.
That estimated rate constant is smaller than might have been
expected, given that the process is an intramolecular cou-
pling of a diradical to a five-membered ring. The reason for
the small rate constant must lie in the fact that the initial
conformation 4a is the wrong one for coupling to 20.
Rotation into conformation 4b, over a barrier of about
9 kcal mol–1, is required (12) before lactone 20 can form.
Such a barrier could limit the rate of coupling considerably,
because the conformations 4a and 4b are only two of many
conformations of the diradical. Rotations about the other sin-
gle bonds of 4a have barriers (13) smaller than 4 kcal mol–1
and therefore other conformations of the diradical will be
populated before conformation 4b is reached. Moreover,
when 4b is reached it will have only about a one-in-four
chance of being in the singlet state (14). With these features
in consideration, the relatively slow coupling of 4 to lactone
20 becomes understandable.
We were surprised to find 22 (12%, 6:1 ratio of
diastereomers), because dialkoxycarbenes have never been
shown to insert into a C–H bond of benzene. In fact, the
more nucleophilic imidazolidene-2-ylidene does not appear
to react with toluene, although products from apparent inser-
tions into the acidic C–H bonds of acetylene and of a
sulphone had been observed (15a). 1,3,4-Triphenyl-4,5-
dihydro-1H-1,2,4-triazole-5-ylidene undergoes many reac-
tions but not insertion into C–H bonds (15b). A possible rea-
son for the apparently higher reactivity of 3, relative to that
of acyclic dialkoxycarbenes, was thought to be the enforced
0°, 0° conformation. For allyloxyhydroxycarbene, the 0°, 0°
conformation was calculated to be 11.7 kcal mol–1 higher in
energy than the most stable (180°, 180°) conformation (3c).
A reduction of the stability of the carbene, as a result of rais-
ing the energy of the ground state, would reduce the barrier
to insertion if the transition state for insertion were affected
less, or not at all, by the conformational restriction. This at-
tractive explanation was abandoned when it was found that
© 2002 NRC Canada

1190
Can. J. Chem. Vol. 80, 2002
Scheme 8.
Ph
.
20
O
.
4b
O
21
Ph
20 + 21
Ph
Ph
Ph
Ph
O
N
O
O
.
O
.
.
O . O
H
O
O
O
O
O
PhH
..
.
H
N
.
1
3
4a
.
Ph
29
30
O
O
H
Ph
22
Scheme 9.
Scheme 11.
O
OMe
O
OMe
OH
MeO
O
Scheme 10.
H2O
(MeO)2C:
Ph
Ph
Ph
O
O
+
Me2CO
O
O
compete. The carbene from 25 does not fragment to a radi-
cal pair and attack on benzene was therefore not observed.
Having estimated the rate constant for decarboxylation
and coupling, we were able to estimate the rate constant for
attack on benzene by 4a and (or) 4b as follows. Knowing
that the rates of coupling and decarboxylation are about
twice as large as that of attack on benzene (from yields of
O
O
_
O
..
3
O
31
32
Ph
Ph
the products) we could set kCO = 2kbenzene[benzene]. The
2
O
O
O
O
O
concentration of benzene is about 10 M and kCO has an up-
H2O
2
31 + 32
per limit of 1 × 105 s–1 (above). Thus kbenzene would not need
to have a value greater than about 5 × 104 M–1 s–1 for aro-
matic substitution to compete. Aromatic substitution oc-
curred when methoxycarbonyl radicals, generated in
benzene by thermolysis of methyl azodicarboxylate at
130°C, gave methyl benzoate in 13% yield (19). Support for
the postulate that aromatic substitution with 4 occurred also
came from the discovery of methyl benzoate (3%) in the
product mixture formed when benzyloxy(methoxy)carbene
was generated in benzene (Scheme 9). We had not recog-
nized that minor product previously (4b).
O
O
O
O
O
O
Ph
34
33
thermolysis of 25 in benzene (16) gave the carbene dimer 27
(20%) and 5,5-dimethyl-1,3-dioxan-2-ol (28, 21%), the latter
being the result of reaction of the carbene with adventitious
water (Scheme 7). There were additional products, but 26
could not be detected. With the results in Scheme 7, it is
hard to ascribe 22 (Scheme 6) to insertion of 3 into benzene.
The only alternative mechanism that we could conceive of
is radical substitution on benzene through the sequence of
Scheme 8. Both intra- and intermolecular additions of acyl
radicals to double bonds are known (17), as is radical aro-
matic substitution (18), and relatively slow closure of 4
(above) could permit aromatic substitution, via 29 and 30, to
The cyclization step, 29J30, is a 6-endo process for
which there are now many examples, although we were un-
able to find a very close analogue (20). Subsequent forma-
tion of diastereomers 22 by a radical mechanism is not
surprising, because kinetic control could apply in the migra-
tion step that forms the new C–H bond. There is ample prece-
dent for the postulated H-migration in 30. Analogous
migrations from C-2 of 1,3-diyls are well known (21) al-
though mono-radicals do not rearrange in that way. Synthesis
© 2002 NRC Canada

Merkley and Warkentin
1191
Scheme 12.
Ph
Ph
Ph
Ph
Ph
Ph
Ph
O
O
O
O
O
N
O
O
t-BuOH
PhH
PhH
O
O
O
H
O
+
+
+
+
O
H
O
O
O
Ph
CH2COCH3
O
N
H
OBu
O
21(26%)
20(28%)
15(2%)
22(12%)
34(2%)
1(38%)
19(75%)
of authentic 22 by acid-catalyzed reaction of 1-phenyl-1,3-
propanediol with benzaldehyde gave only one diastereomer,
which we assume was the more stable cis-isomer, formed
under thermodynamic control.
sidual CHCl3 at 7.26 ppm or to the 13C signal at 77.0 ppm.
The coupling constants of the first-order multiplets are
in Hz. Chemical shift assignments are based on analysis of
1
1
1H–1H COSY, H–13C HSQC, and H–13C HMBC spectra.
Connectivities of the atoms were determined with the aid of
Spiro-fused ester 34 was also obtained, in 2% yield. The
structure is not in doubt, but the source of it is not known.
Addition of acetone (initial concentration 1.1 M) before
thermolysis raised the yield of 34 to 28%! We have not de-
veloped an attractive explanation for its formation. A partial
solution to the problem involves attack of the carbene (3) on
acetone to afford 32, which, like 2, should be equilibrated
with the oxirane 31 (Scheme 10). The postulated reaction of
the carbene with acetone is in keeping with the observation
that dimethoxycarbene reacts with cyclohexanone to afford
an oxirane (22). Coupling of two units of 32 is unlikely, be-
cause 32 is a reactive intermediate that must be formed very
slowly from 1. However, a reaction between 31 and 32 is
plausible, given that 31 is expected to be quite polar. Only
two 2,2-dialkoxyoxiranes have been reported to date (22, 23)
and they are sensitive materials that hydrolyse readily, as
shown for one of them in Scheme 11. A bimolecular reac-
tion between 31 and 32 would be expected to lead to 33, and
partial hydrolysis of 33 after the reaction tube was opened
would yield the observed 34 (Scheme 10). Water in the
added acetone cannot be a major contributor because the ac-
etone was dried carefully and because the yield of
hemiorthoformate (23) did not rise as a result of adding ace-
tone to the benzene solvent prior to thermolysis.
1
1
both H–13C HMBC and H–13C HSQC 2-D NMR experi-
ments. Assignments of signals to axial (ax) and equatorial
1
(eq) H atoms are based on good analogy (24).
4-Phenyl-1,3-dioxane (6) (25)
Styrene (100 g, 0.96 mol) was dissolved in 200 mL (2.5 mol)
of 37% formaldehyde. Concentrated H2SO4 (8.0 mL, 0.14 mol)
was added, the round-bottom flask was fitted with a reflux
condenser, and the mixture was refluxed for 7 h. After cool-
ing, the solution was washed twice with 160 mL of benzene
and the combined benzene layers were washed twice with
250 mL of water. Most of the benzene was removed by dis-
tillation (80°C) at atmospheric pressure. The pressure was
then reduced to 2 mm Hg (1 mm Hg = 133.322 Pa), and
4-phenyl-1,3-dioxane was collected as a clear oil (83.9 g,
1
0.51 mol, 57%), boiling between 96 and 130°C. H NMR
3
(200 MHz, CDCl3) ꢂ: 1.71 (m, 1H), 2.10 (ddd, J = 4.9 Hz,
3
2
3J = 12.2 Hz, J = 12.4 Hz, J = 12.4 Hz, 1H), 3.87 (ddd,
3J = 2.5 Hz, J = 11.8 Hz, J = 11.8 Hz, 1H), 4.19 (dd, J =
4.8 Hz, J = 11.4 Hz, 1H), 4.64 (dd, J = 2.5 Hz, J =
11.2 Hz, 1H), 4.89 (d, 2J = 6.3 Hz, 1H), 5.22 (d, 2J = 6.3 Hz,
1H), 7.26–7.39 (m, 5H).
3
2
3
2
3
2
1,5-Diacetoxy-5-phenyl-2-oxapentane (7) (26)
In summary, the thermolysis chemistry of oxadiazoline 1
is exceptionally rich. It loses N2 to form carbonyl ylide 2
that undergoes 1,4-sigmatropic rearrangement to afford 15
via 13. The major fate of 2 is fragmentation to carbene 3
that can be trapped efficiently in a bimolecular reaction with
tert-butyl alcohol. If 3 is not intercepted, it undergoes ring
opening to diradical 4, which undergoes three reactions at
comparable rates. The diradical attacks benzene through a
sequence of unprecedented reactions that lead to substitution
product 22. Diyl 4 also cyclizes to lactone 20 and it
decarboxylates to afford phenylcyclopropane via cyclization
of diyl 5. Finally, the formation of 34 suggests that oxirane
31 is an undetected intermediate from the reaction of 3 with
acetone. The products, accounted for individually above, are
listed in Scheme 12.
Concentrated HCl (1 mL) was added to 4-phenyl-1,3-
dioxane (6) (80 g, 0.49 mol) dissolved in acetic anhydride
(138 mL, 1.46 mol) and the solution was heated to 80°C for
20 h. The solution was cooled, neutralized with 10% NaOH
to pH 7 (as determined with universal litmus paper), and
then washed twice with ether (50 mL). The ether layer was
dried with magnesium sulfate. After vacuum filtration the
ether was removed with a rotary evaporator leaving 7
(108.4 g, 0.41 mol, 83%) as an orange oil. 1H NMR
(500 MHz, CDCl3) ꢂ: 1.87 (s, 3H), 1.90–1.96 (m, 1H), 1.99
(s, 3H), 2.05–2.12 (m, 1H), 4.01–4.06 (m, 1H), 4.13–4.18
3
(m, 1H), 4.65 (m, 1H), 5.01 (d, J = 2.6 Hz, 1H), 5.28 (d,
3J = 2.6 Hz, 1H), 7.24–7.32 (m, 5H). 13C NMR (125 MHz,
CDCl3) ꢂ: 20.49, 20.60, 36.70, 60.80, 78.01, 86.81, 126.36,
127.82, 128.35, 140.81, 170.17, 170.58. EI-MS m/z: 206 (1),
177 (4), 163 (3), 146 (17), 133 (6), 117 (29), 107 (25), 91
(6), 77 (8), 43 (100). CI-MS (NH3) m/z: 284 ([M + NH4]+,
28), 177 (100), 146 (17), 117 (95), 75 (4).
Experimental
General
NMR spectra were collected with Bruker AV-300 or AV-
500 NMR spectrometers, with samples in CDCl3. The chem-
ical shifts, in ppm, are referenced to the H signal from re-
1-Phenyl-1,3-propanediol (8) (27)
To a solution of 7 (15.62 g, 0.058 mol) in 10 mL of ether
was slowly added a saturated solution of KOH in methanol
1
© 2002 NRC Canada

1192
Can. J. Chem. Vol. 80, 2002
(30 mL). The solution was stirred for 3 h before it was
neutralized with 5% HCl. The aqueous–methanol layer was
extracted twice with 30 mL of ether. Evaporation of most of
the ether and subsequent distillation of the residue gave 1-
phenyl-1,3-propanediol (4.7 g, 0.031 mol, 54%), boiling at
119°C, 0.5 mm Hg (1 mm Hg = 133.322 Pa) (lit. (28) value
3.1 mmol, 60%); mp 109–110°C. 1H NMR (300 MHz,
CDCl3) ꢂ: 1.82 (s, 3H), 2.03 (s, 3H), 2.07–2.11 (m, 2H),
4.24–4.27 (m, 1H), 4.45–4.53 (m, 1H), 4.81 (dd, 3J =
3
4.2 Hz, J = 4.5 Hz, 1H), 7.27–7.38 (m, 5H). 13C NMR
(75 MHz, CDCl3) ꢂ: 16.24, 25.49, 38.62, 63.30, 71.36,
125.89, 127.69, 128.64, 144.10, 151.37, 154.45. EI-MS m/z:
251 ([M + H]+, 1), 144 (15), 117 (100), 91 (17), 72 (38), 56
(31), 43 (38). CI-MS (NH3) m/z: 251 ([M + H]+, 29), 233
(23), 193 (48), 117 (100), 99 (12), 72 (19). HR-MS calcd.
for C13H19N2O3: 251.1395; found: 251.1382.
1
bp 117 to 118°C at 0.25 mm Hg). H NMR (200 MHz,
CD3OD) ꢂ: 1.68–1.91 (m, 2H), 3.62–3.41 (m, 2H), 4.83 (dd,
3
3J = 2.6 Hz, J = 5.4 Hz, 1H), 7.22–7.52 (m, 5H). 13C NMR
(50 MHz, CD3OD) ꢂ: 42.71, 65.50, 75.92, 126.73, 128.20,
1
129.40, 146.33. H NMR (200 MHz, CDCl3) ꢂ: 1.86 (m,
3
3
2H), 3.76 (m, 2H), 4.83 (dd, J = 4.7 Hz, J = 7.9 Hz, 1H),
7.22–7.52 (m, 5H). 13C NMR (50 MHz, CDCl3) ꢂ: 40.34,
60.44, 73.08, 125.55, 127.25, 128.25, 144.19. EI-MS m/z:
152 ([M]+, 14), 133 (9), 105 (50), 107 (100), 91 (16), 79
(78), 51 (23), 43 (16). CI-MS (NH3) m/z: 170 ([M + NH4]+,
21), 152 (100), 135 (56), 117 (86), 91 (16).
3,4-Diaza-2,2-dimethyl-7-phenyl-1,6,10-trioxaspiro[5.4]dec-
3-ene (1)
To an ice-cooled solution of lead(IV) acetate (4.43 g, 0.010 mol)
in dichloromethane (25 mL) under nitrogen was added,
dropwise, over a period of 1 h, a solution of 11 (1 g,
0.004 mol) in dichloromethane (5 mL). After addition, stir-
ring was continued for 2 days. The solution was filtered
through Celite, washed four times with 25 mL of 5% NaHCO3
4-Phenyl-1,3-dioxan-2-one (9) (29)
1
1,1>-Carbonyldiimidazole (3.51 g, 0.022 mol) in 100 mL
of dichloromethane was added dropwise (ca. 1 h) to a solu-
tion of 8 (3.0 g, 0.02 mol) and pyridine (4.4 mL, 0.054 mol)
in 180 mL of dichloromethane. After stirring overnight the
solution was washed three times with 5% HCl (250 mL),
and once each with water (250 mL) and brine (250 mL). The
dichloromethane layer was dried with MgSO4 and, after fil-
tration, the solvent was removed with a rotary evaporator to
leave 9 as white crystals (2.49 g, 0.014 mol, 70%) after
solution and dried over Na2SO4. H NMR spectroscopy of
an aliquot indicated that the product was probably a mixture
of 1 and 12. The solvent was evaporated and the residue was
dissolved in dichloromethane (20 mL) before trifluoroacetic
acid (0.01 mL) was added. After 90 min the cyclization of
12 to 1, monitored with TLC, was complete. The solution
was washed with aqueous NaOH and dried over MgSO4.
Evaporation of the solvent and radial chromatography of the
residue (Chromatotron, 10% EtOAc – hexanes) gave 1, a
white solid, as a mixture of two diastereomers in 1:1 ratio
1
recrystallization from hot ethanol; mp 53–54°C. H NMR
1
(300 MHz, CDCl3) ꢂ: 2.20–2.38 (m, 2H), 4.46–4.51 (m,
(0.38 g, 1.5 mmol, 38%). H NMR (composite spectrum,
3
3
2H), 5.51 (dd, J = 3.8 Hz, J = 9.6 Hz, 1H), 7.26–7.44 (m,
5H). 13C NMR (75 MHz, CDCl3) ꢂ: 29.42, 66.87, 80.15,
125.67, 127.83, 129.0 137.90, 148.77 (C=O). EI-MS m/z:
178 ([M]+, 12), 117 (82), 104 (100), 91 (11), 77 (51), 56
(48), 51 (36). CI-MS (NH3) m/z: 196 ([M + NH4]+, 100),
179 ([M + H]+, 30), 134 (9), 117 (89), 104 (8). HR-MS
calcd. for C10H10O3: 178.0629; found: 178.0657.
300 MHz, CDCl3) ꢂ: 1.48 (s, 3H), 1.52 (s, 3H), 1.53 (s, 6H),
1.90–2.03 (m, 2H), 2.18–2.39 (m, 2H), 4.23–4.31 (m, 2H),
3
3
2
4.35 (ddd, J = 3.5 Hz, J = 11.3 Hz, J = 11.3 Hz, 1H) 4.81
(ddd, 3J = 2.8 Hz, 3J = 11.8 Hz, 2J = 11.9 Hz, 1H), 5.24 (dd,
3J = 3.1 Hz, J = 11.6 Hz, 1H), 5.73 (dd, J = 2.7 Hz, J =
11.6 Hz, 1H), 7.23–7.40 (m, 10H). 13C NMR (composite
spectrum, 75 MHz, CDCl3) ꢂ: 24.49, 24.64, 31.74, 31.92,
63.02, 65.16, 74.62, 77.09, 119.81, 120.44, 125.92, 128.18,
128.54, 140.21 (analogous, but not spirocyclic, oxadia-
zolines typically have a C-5 signal at ꢂ 118.7–119.4 and a
C-2 signal at ꢂ 136.7–137.9) (30). EI-MS m/z: 162 ([M–N2–
Me2CO]+, 1), 116 (72), 117 (100), 104 (12), 91 (19), 65 (3),
43 (19).
3
3
3
Synthesis of 10 by hydrazinolysis of 9
Hydrazine monohydrate (7.1 mL, 0.15 mol) was added,
with stirring, to 4-phenyl-1,3-dioxan-2-one (9) (3.96 g,
0.022 mol) in 70 mL of dichloromethane under nitrogen. In
about 2 h a white precipitate had formed. The heterogeneous
mixture was filtered leaving the product as white crystals
1
(3.63 g, 0.017 mol, 77%). H NMR (200 MHz, CD3OD) ꢂ:
Thermolysis of 1 in the presence of tert-butyl alcohol
A solution of oxadiazoline 1 (27.4 mg, 0.11 mmol) and
tert-butyl alcohol (82.2 mg, 1.1 mmol) in 2 mL of benzene
was flame sealed into an NMR tube. The tube was heated
for 72 h at 110°C with a constant-temperature oil bath. Two
diastereomers of 2-tert-butoxy-4-phenyl-1,3-dioxane (19)
(75% combined yield) were obtained in a ratio of 3:1, as de-
termined by integration of clean singlets at ꢂ 5.64 and 6.03.
The crude sample was not purified but the signals corre-
sponding to the major diastereomer could be assigned. Some
integrations are approximate, because of partial overlap with
signals from the minor diastereomer.
1.92 (m, 2H), 4.00ꢃ4.12 (m, 2H), 4.66 (m, 1H), 7.17–7.26
(m, 5H). 13C NMR (50 MHz, CD3OD) ꢂ: 39.50, 63.39,
71.73, 126.93, 128.38, 129.38, 145.96, 160.69 (C=O). EI-
MS m/z: 211 ([M + H]+, 1), 192 (7), 134 (9), 118 (11), 117
(100), 91 (11), 79 (16), 57 (11), 43 (12). CI-MS (NH3) m/z:
211 ([M + H]+, 3), 193 (100), 152 (13), 135 (8), 117 (100),
105 (8), 76 (11), 69 (19), 43 (52).
3-Isopropylidene carbazic acid 3-hydroxy-3-phenylpropyl
ester (11)
To a solution of 10 (1.10 g, 5.23 mmol) in dichloro-
methane (10 mL) was added 1 mL of acetone. After the 10
had dissolved, Na2SO4 (0.24 g, 1.4 mmol) was added, and
the mixture was stirred for 2 h. The Na2SO4 was removed by
filtration and the solvent was removed with a rotary evapora-
tor. Compound 11 was obtained as a white solid (0.78 g,
1H NMR (major diastereomer, 500 MHz, C6D6) ꢂ: 1.59 (s,
9H, CH3), 1.48–1.73 (dm, 1H, H5-eq), 1.96 (ddd, 3J =
3
3
2
5.0 Hz, J = 12.6 Hz, J = 12.4 Hz, J = 12.4 Hz, 1H,
3
H5-ax), 3.36–3.50 (m, 1H, H6-ax), 3.57 (dd, J = 1.4 Hz,
3J = 4.0 Hz, J = 11.0 Hz, 1H, H6-eq), 5.32 (dd, J =
2
3
© 2002 NRC Canada

Merkley and Warkentin
1193
3
2
3
2.4 Hz, J = 11.7 Hz, 1H, H4-ax), 6.03 (s, 1H), 7.03–7.37
3J = 11.7 Hz, J = –11.9 Hz, 1H, H6-ax), 4.20 (ddd, J =
(m, 5H). 13C NMR (125 MHz, C6D6) ꢂ: 28.74 (C9), 33.77
(C5), 58.32 (C6), 69.44 (C4), 77.13 (C8), 105.21 (C2),
125.91, 128.08, 128.32, 142.90. EI-MS m/z: 152 (6), 134
(31), 107 (100), 79 (84), 59 (14), 51 (30). CI-MS
(NH3) m/z: 196 (8), 179 (8), 152 (24), 134 (17), 117 (100),
105 (19).
1.3 Hz, J = 4.9 Hz, J = –11.5 Hz, 1H, H6-eq), 4.72 (dd,
3
2
3J = 2.5 Hz, J = 11.4 Hz, 1H, H4-ax), 5.17 (t, J = 5.1 Hz,
1H, H2), 7.28–7.4 (m, 5H). 13C NMR (125 MHz, CDCl3) ꢂ:
31.51, 33.17, 49.33, 67.02, 78.93, 99.06, 125.95, 127.98,
128.61, 141.49, 205.21. CI-MS (NH3) m/z: 238 ([M + NH4]+,
25), 221 ([M + H]+, 58), 187 (31), 152 (30), 135 (44), 117
(100), 91 (14), 78 (10), 43 (14).
3
3
Thermolysis of 1 in benzene
8,8,11,11-Tetramethyl-2-phenyl-1,5,7,10-tetraoxa-9-
oxospiro[5.5]undecane (34)
Oxadiazoline 1 (0.1515 g, 0.61 mmol), and the internal
standard 1,4-dimethoxybenzene (0.0205 g, 0.148 mmol)
were dissolved in 6 mL of benzene and sealed into a
thermolysis tube after three cycles of freeze-pump-thaw de-
gassing. The tube was then kept in an oil bath at 110°C for
72 h. The crude mixture of products was injected into the
GC–FID to determine the product yields. The peak areas
were corrected for the detector response and products were
identified by coinjection of authentic compounds. The re-
mainder of the mixture was freed from solvent and the resi-
due was separated by radial chromatography (Chromatotron,
100% hexanes) to afford the following products.
2%. Connectivity determined by means of ROESY 2-D
NMR spectroscopy. IR (NaBr, neat) (cm–1): 1739. 1H NMR
3
2
(500 MHz, C6D6) ꢂ: 1.15 (dddd, J = 1.5, 2.6, 2.6 Hz, J =
13.4 Hz, 1H, H3-eq), 1.48 (s, 3H, H11>), 1.50 (s, 3H, H11>>),
3
1.53 (s, 3H, H8>), 1.56 (s, 3H, H8>> ), 1.62 (dddd, J = 4.9,
11.7, 12.9 Hz, 2J = 13.1 Hz, 1H, H3-ax), 3.41 (ddd, 3J = 1.6,
4.9 Hz, 2J = 11.1 Hz, 1H, H4-eq), 3.97 (ddd, 3J = 2.5,
12.8 Hz, 2J = 11.1 Hz, 1H, H4-ax), 5.00 (dd, 3J = 2.7,
11.7 Hz, 1H, H10-ax), 7.08–7.12 (m, 5H). 13C NMR
(125 MHz, C6D6) ꢂ: 24.07 (C11>), 24.20 (C11>), 28.58 (C8>),
28.72 (C8>), 32.47 (C3), 60.34 (C4), 71.65 (C2), 76.74
(C11), 84.16 (C8), 109.07 (C6), 128.85, 128.39, 128.15,
141.42, 171.06 (C=O). EI-MS m/z: 204 (4), 179 (4), 117
(100), 104 (18), 70 (70), 43 (37). CI-MS (NH3) m/z: 307
([M + H]+, 18), 134 (18), 117 (100), 70 (32). The yield of
this product rose to 28% in an experiment in which the start-
ing solvent was 1.1 M in acetone.
Cyclopropylbenzene (21) (31)
1
26%. H NMR (200 MHz, CDCl3) ꢂ: 0.85 (m, 2H), 1.11
(m, 2H), 2.02 (m, 1H), 7.19–7.43 (m, 5H). EI-MS m/z: 118
([M]+, 54), 117 (100), 115 (48), 91 (74), 77 (22).
4-Phenyl-1,3-dioxan-2-one (9) (29)
5%.
2,4-Diphenyl-1,3-dioxane (22)
ꢀ-Phenyl-ꢁ-butyrolactone (20) (32, 33)
12%. Diastereomers in 6:1 ratio (by NMR, after chroma-
tography, which failed to separate them and may have al-
tered the initial diastereomer ratio). 1H NMR (major
diastereomer, 500 MHz, CDCl3) ꢂ: 1.81 (ddd, 3J = 2.5,
4.0 Hz, 2J = 13.4 Hz, 1H, H5-eq), 2.14 (dddd, 3J = 4.9, 12.2,
28%. 1H NMR (300 MHz, CDCl3) ꢂ: 2.40 (dddd, 3J = 8.2,
2
3
9.2, 10.1 Hz, J = 12.6 Hz, 1H), 2.71 (dddd, J = 3.4, 6.8,
2
3
3
9.1 Hz, J = 12.6 Hz, 1H), 3.80 (dd, J = 9.7 Hz, J =
3
2
9.9 Hz, 1H), 4.34 (ddd, J = 6.7, 9.1 Hz, J = 9.1 Hz, 1H),
3
2
2
3
4.46 (ddd, J = 3.4, 8.2 Hz, J = 9.1 Hz, 1H), 7.27–7.40 (m,
5H). 13C NMR (75 MHz, CDCl3) ꢂ: 31.62, 45.54, 66.56,
127.71, 127.98, 129.00, 136.79, 177.43. EI-MS m/z: 162
([M]+, 22), 117 (100), 103 (16), 91 (48), 77 (24), 63 (14), 51
(15). HR-MS calcd. for C10H10O2: 162.0681; found: 162.0675.
11.9 Hz, J = 13.4 Hz, 1H, H5-ax), 4.14 (ddd, J = 2.5,
2
3
11.9 Hz, J = 11.9 Hz, 1H, H6-ax), 4.37 (ddd, J = 1.3,
4.9 Hz, 2J = 11.5 Hz, 1H, H6-eq), 4.92 (dd, 3J = 2.6,
11.4 Hz, H4-ax), 5.73 (s, 1H, H2-ax), 7.28–7.58 (m, 10H).
13C NMR (125 MHz, CDCl3) ꢂ: 33.63 (C5), 67.48 (C6),
79.29 (C4), 101.79 (C2), 126.03, 126.38, 127.89, 128.38,
128.59, 128.95, 131.04, 138.88, 141.86. EI-MS m/z: 239
([M – H]+, 1), 134 (7), 117 (61), 105 (100), 78 (35), 77 (82),
51 (32). CI-MS (NH3) m/z: 241 ([M + H]+, 24), 152 (16),
134 (40), 117 (100), 91 (24). The minor diastereomer was
inferred from the presence of a singlet at 5.23 ppm, upfield
from that of the major diastereomer by 0.55 ppm. In the two
diastereomers formed from trapping of the carbene with tert-
butyl alcohol the minor diastereomer gave rise to a singlet
0.39 ppm upfield from the signal for the orthoformyl H of
the major diastereomer.
4-Phenyl-1,3-dioxan-2-ol (23)
19% (not isolable, structure assignment tentative, dia-
stereomer ratio not known). EI-MS m/z: 163 ([M – OH]+, 1),
133 (2), 117 (100), 105 (30), 91 (20), 77 (39).
1-Phenyl-1,3-propanediol-3-formate (24) (34)
1H NMR (300 MHz, CDCl3) ꢂ: 2.02–2.19 (m, 2H), 4.23
3
3
3
2
(ddd, J = 5.9 Hz, J = 5.7 Hz, J = 5.7 Hz, J = 11.3 Hz,
3
3
3
2
1H), 4.43 (ddd, J = 5.9 Hz, J = 5.7 Hz, J = 5.7 Hz, J =
3
3
11.3 Hz, 1H), 4.83 (dd, J = 5 Hz, J = 7.6 Hz, 1H), 7.35–
7.36 (m, 5H), 8.08 (s, 1H). 13C NMR (75 MHz, CDCl3) ꢂ:
37.93, 61.27, 71.31, 125.85, 128.04, 128.80, 143.90, 161.22.
EI-MS m/z: 180 ([M]+, 3), 163 (4), 152 (7), 134 (52), 107
(96), 91 (20), 79 (100), 51 (21), 43 (17). HR-MS calcd. for
C10H12O3: 180.0786; found: 180.0778.
Thermolysis of 1 in the presence of added acetone
A solution of oxadiazoline 1 (0.1684 g, 0.679 mmol) and
acetone (0.5 mL) in benzene (5.5 mL) was sealed into a
thermolysis tube after three cycles of freeze–pump–thaw de-
gassing. The tube was kept in an oil bath at 110°C for 72 h.
After thermolysis the internal standard, 1,4-dimethoxy-
benzene (6.8 mg, 0.1218 mmol), was added and the crude
solution was injected into the GC–FID for determination of
product yields. Detector response factors were applied to ob-
tain the corrected yields below.
1-(4-Phenyl-1,3-dioxan-2-yl)-2-propanone (15)
2%. Apparently only one diastereomer was isolated.
1H NMR (500 MHz, CDCl3) ꢂ: 1.70 (ddd, J = 2.5 Hz, J =
3
3
2
3
3.5 Hz, J = –13.4 Hz, 1H, H5-eq), 1.99 (dddd, J = 4.6 Hz,
3
3
2J = –13.3 Hz, J = 11.9 Hz, J = 11.9 Hz, 1H, H5-ax), 2.22
3
3
(s, 3H), 2.84 (d, J = 5.1 Hz, 2H), 3.94 (ddd, J = 2.6 Hz,
© 2002 NRC Canada

1194
Can. J. Chem. Vol. 80, 2002
Louw. Tetrahedron Lett. 4941 (1972); (e) A.M. Foster and
W.C. Agosta. J. Am. Chem. Soc. 95, 608 (1973); ( f ) G.B.
Merrill and H. Schechter. Tetrahedron Lett. 4527 (1975);
(g) W.T. Borden, P.W. Concannon, and D.L. Phillips. Tetrahe-
dron Lett. 3161 (1973); (h) W.T. Borden and L.H. Hoo. J. Am.
Chem. Soc. 100, 6274 (1978); (i) J.M. Cummins, T.A. Porter,
and M. Jones, Jr. J. Am. Chem. Soc. 120, 6473 (1998).
Cyclopropylbenzene (21) (4%), ꢀ-phenyl-ꢁ-butyrolactone
(20) (5%), 2-hydroxy-4-phenyl-1,3-dioxane (23) (21%), 1-
(4-phenyl-1,3-dioxan-2-yl)-2-propanone (15) (2%), 8,8,11,11-
tetramethyl-2-phenyl-1,5,7,10-tetraoxa-9-oxospiro[5.5]undecane
(34) (28%), 2,4-diphenyl-1,3-dioxane (diastereomers) (22)
(13%), 4-phenyl-1,3-dioxan-2-one (9) (7%).
3. (a) D. Feller, W.T. Borden, and E.R. Davidson. J. Comput.
Chem. 1, 158 (1980); (b) D.L. Reid, J. Hernandez-Trujillo, and
J. Warkentin. J. Phys. Chem. A, 104, 3398 (2000); (c) D.L.
Reid and J. Warkentin. J. Chem. Soc. Perkin Trans. 2, 1980
(2000).
4. (a) P.C. Venneri and J. Warkentin. J. Am. Chem. Soc. 120,
11 182 (1998); (b) N. Merkley, M. El-Saidi, and J. Warkentin.
Can. J. Chem. 78, 356 (2000).
Synthesis of cis-2,4-diphenyl-1,3-dioxane
The procedure was modeled after one (35) for making an
analogue. To 1-phenyl-1,3-propanediol (5 g, 0.032 mol) and
benzaldehyde (3.4 g, 0.036 mol) in dichloromethane
(50 mL) was added p-toluenesulfonic acid (0.5 g, 2.9 mmol).
The solution was stirred overnight before it was washed with
10 mL of NaHCO3 and dried over Na2SO4. The residue ob-
tained upon removal of the solvent was purified by means of
column chromatography (10% EtOAc – hexanes) yielding
22 (5.65 g, 73%). The cis configuration was assigned on the
basis of the known preference for the C-2 phenyl group to
be in the equatorial position (36). 1H NMR (200 MHz,
5. W.T. Borden and L.H. Hoo. J. Am. Chem. Soc. 100, 6274
(1978).
6. K. Kassam, D.L. Pole, M. El-Saidi, and J. Warkentin. J. Am.
Chem. Soc. 116, 1161 (1994).
7. (a) A.P. Hitchcock, S. Zweep, T. Steel, M. Békhazi, and J.
Warkentin. Can. J. Chem. 60, 2914 (1982); (b) A.C. Lottes,
J.A. Landgrebe, and K. Larsen. Tetrahedron Lett. 30, 4089
(1989); (c) M.W. Majchrzak and J. Warkentin. Can. J. Chem.
67, 1753 (1989); (d) T. Golz, S. Hammes and F.-G. Klärner.
Chem. Ber. 126, 485 (1993).
3
CDCl3) ꢂ: 1.83 (m, 1H, H5-eq), 2.18 (dddd, J = 4.9, 12.2,
2
3
12.2 Hz, J = 12.3 Hz, 1H, H5-ax), 4.16 (ddd, J = 2.3,
2
3
11.8 Hz, J = 11.8 Hz, 1H, H6-ax), 4.40 (ddd, J = 1.0,
4.5 Hz, 2J = 11.4 Hz, 1H, H6-eq), 4.98 (dd, 3J = 2.4,
11.2 Hz, 1H, H4-ax), 5.73 (s, 1H, H2), 7.33–7.66 (m, 10H).
8. W.B. Smith. J. Org. Chem. 60, 7456 (1995).
9. A.P.M. van der Veek and F.H. van Putten. Tetrahedron Lett.
3951 (1970).
3,4-Diaza-2,2,8,8-tetramethyl-1,6,10-trioxaspiro[5.4]dec-
3-ene (25)
10. P.A. Simakov, F.N. Martinez, J.H. Horner, and M. Newcomb.
The synthesis and properties of 25 have been described
J. Org. Chem. 63, 1226 (1998).
11. J. Pfenniger, C. Heuberger, and W. Graf. Helv. Chim. Acta, 63,
2328 (1980).
12. A preliminary communication has been published. N. Merkley,
D.L.Reid, and J. Warkentin. Tetrahedron Lett. 43, 1927 (2002).
13. J.P. Lowe. In Progress in physical organic chemistry. Vol. 6.
Edited by A. Streitwieser, Jr. and R.W. Taft. Interscience Pub-
lishers, New York. 1968.
(16).
5,5-Dimethyl-1,3-dioxan-2-ol (28)
Unstable material, isolated in 21% yield by preparative
GC (OV-17, 6> × 1/4>>, He at 40 mL min–1) and tentatively
1
identified as 28. H NMR (200 MHz, C6D6) ꢂ: 1.02 (s, 3H),
1.76 (s, 3H), 3.13 (d, 2J = 10.6 Hz, 2H), 3.81 (d, 2J =
10.6 Hz, 2H), 5.97 (s, 1H).
14. L. Salem and C. Rowland. Angew. Chem. Int. Ed. Engl. 11, 92
(1972).
Carbene dimer (27)
15. (a) A.J. Arduengo, III, J.C. Calabrese, F. Davidson, H.V.R.
Dias, J.R. Goerlich, R. Krafczyk, W.J. Marshall, M. Tamm,
and R. Schmutzler. Helv. Chim. Acta, 82, 2348 (1999); (b) D.
Enders, K. Breuer, J. Runsink, and J.R. Teles. Liebigs Ann.
2019 (1996).
16. N. Merkley and J. Warkentin. Arkivoc [online], 2, 7 (2001).
(OT-243C). Available from http://www.arkat-usa.org/ark/jour-
nal/volume2/part3/tee/001/.
Yield 20%, isolated by preparative GC (conditions above).
1H NMR (300 MHz, C6D6) ꢂ: 0.067 (s, 12H), 3.43 (s, 8H).
13C NMR (75 MHz, C6D6) ꢂ: 21.86, 30.95, 78.36, 136.69.
EI-MS m/z: 229 ([M + H]+, 4), 191 (8), 161 (13), 132 (32),
115 (92), 69 (100), 56 (78). HR-MS calcd. for C12H21O4
[M + H]+: 229.1439; found: 229.1453.
17. D.L. Boger and R.J. Mathvink. J. Am. Chem. Soc. 112, 4003
(1990).
18. C. Chatgilialoglu, D. Crich, M. Komatsu, and I. Ryu. Chem.
Rev. 99, 1921 (1999).
19. M. Fiorentino, L. Testafarri, M. Tiecco, and L. Troisi. J. Org.
Chem. 41, 173 (1976).
20. (a) M. Breithor, L. Herndon, and H.M.R. Hoffmann. Tetrahe-
dron, 53, 8401 (1997); (b) P. Double and G. Pattenden. J.
Chem. Soc. Perkin Trans. 1, 2005 (1998); (c) E.W. Della, C.
Kostakis, and P.A. Smith. Org. Lett. 1, 363 (1999); (d) S.
Handa and G. Pattenden. J. Chem. Soc. Perkin Trans. 1, 843
(1999); (e) I. Ryu, S. Ogura, S. Minakata, and M. Komatsu.
Tetrahedron Lett. 40, 1515 (1999).
21. J.E. Baldwin and R. Shukla. J. Phys. Chem. A, 103, 7821
(1999).
22. M. Dawid, P.C. Venneri, and J. Warkentin. Can. J. Chem. 79,
110 (2001).
Acknowledgements
The authors are grateful for an operating grant from the
Natural Sciences and Engineering Research Council of Can-
ada (NSERC), from which this work was financed. Helpful
discussions with Professor R.A. Bell and the assistance of
Dr. D. Hughes with the acquisition and interpretation of
NMR spectra are also gratefully acknowledged.
References
1. J. Warkentin. J. Chem. Soc. Perkin Trans. 1, 2161 (2000).
2. (a) R.J. Crawford and R. Raap. Proc. Chem. Soc. London, 370
(1963); (b) R.W. Hoffmann. Angew. Chem. Int. Ed. Engl. 10,
529 (1971); (c) R.W. Hoffmann, R. Hirsch, R. Fleming, and
M.T. Reetz. Chem. Ber. 105, 3532 (1972); (d) P.C. Oele and R.
© 2002 NRC Canada

Merkley and Warkentin
1195
28. H.C. Brown and S. Kim. J. Org. Chem. 49, 1064 (1984).
23. M.S. Newman and E. Kilbourn. J. Org. Chem. 35, 3186 (1970).
24. (a) R.J. Abraham and P. Loftus. In Proton and carbon-13 NMR
spectroscopy: an integrated approach. Heydon, London. 1979.
29. M. Fujiwara, A. Baba, and H. Matsuda, J. Heterocycl. Chem.
26, 1659 (1989).
30. M. El-Saidi, K. Kassam, D.L. Pole, T. Tadey, and J. Warkentin.
J. Am. Chem. Soc. 114, 8751 (1992).
31. C. Galli. Magn. Reson. Chem. 27, 214 (1989).
32. Y. Ishii, T. Yoshida, K. Yamawaki, and M. Ogawa. J. Org.
Chem. 53, 5549 (1998).
1
pp. 44ꢃ46; (b) H. Booth. In Application of H NMR spectros-
copy to the conformational analysis of cyclic compounds.
Prog. NMR spectroscopy. Vol 5. Edited by J.W. Emsley, J.
Feeney, and L.H. Sutcliffe. Pergamon Press, New York. 1969.
pp. 149–381; (c) F.A. Bovey. In Nuclear magnetic resonance
spectroscopy. 2nd ed. Academic Press, New York. 1988;
(d) R.K. Harris. In Nuclear magnetic resonance spectroscopy.
Pitman Publishing Inc., Marshfield, Mass. 1983. p. 223.
25. (a) R.L. Shriner and P.R. Ruby. In Org. Synth. Coll. Vol. IV.
Edited by N. Rabjohn. John Wiley & Sons, New York. 1960. p.
786; (b) J. Delmau, J.-C. Duplan, and M. Davidson. Tetrahe-
dron, 23, 4371 (1967).
33. S. Karlsson, A. Hallberg, and S. Gronowitz. J. Organomet.
Chem. 403, 133 (1991).
34. O. Meresz, K.P. Leung, and A.S. Denes. Tetrahedron Lett.
2797 (1972).
35. V.V. Chistyakov and L.N. Pospelova. Dokl. Akad. Nauk SSSR,
296, 496 (1987).
36. E. Eliel, W.F. Bailey, K.B. Wiberg, H. Connon, and F.W.
Nader. Liebigs Ann. Chem. 2240 (1976).
26. H. Moe and B.B. Corson. J. Org. Chem. 24, 1768 (1959).
27. T. Sakomoto, Y. Kondo, and H. Yamanaka. Tetrahedron Lett.
33, 6845 (1992).
© 2002 NRC Canada