Relative asymmetric induction in the intramolecular reaction between alkynes and cyclopropylcarbene-chromium complexes: Stereocontrolled synthesis of five-membered rings fused to oxygen heterocycles

Article abstract of DOI:10.1021/jo982144q

Synthesis of cyclopentenone derivatives fused to oxygen heterocycles by means of the intramolecular coupling of alkynes and cyclopropylcarbene- chromium complexes has been examined for a variety of cases in which the tethering chain features a stereogenic

Full text of DOI:10.1021/jo982144q

J . Org. Chem. 1999, 64, 1291-1301
1291
Rela tive Asym m etr ic In d u ction in th e In tr a m olecu la r Rea ction
betw een Alk yn es a n d Cyclop r op ylca r ben e-Ch r om iu m Com p lexes:
Ster eocon tr olled Syn th esis of F ive-Mem ber ed Rin gs F u sed to
Oxygen Heter ocycles
J ingbo Yan, J in Zhu, and J ulius J . Matasi
Department of Chemistry and Biochemistry, University of Maryland, College Park, Maryland 20742-2021
J ames W. Herndon*
Department of Chemistry and Biochemistry, New Mexico State University, Las Cruces, New Mexico 88003
Received October 26, 1998
Synthesis of cyclopentenone derivatives fused to oxygen heterocycles by means of the intramolecular
coupling of alkynes and cyclopropylcarbene-chromium complexes has been examined for a variety
of cases in which the tethering chain features a stereogenic center. In some cases, this process
proceeds with a very high degree of stereoselectivity; however, the extent and direction of relative
asymmetric induction is very dependent upon the position of the chiral atom within the tethering
chain and the length of the tethering chain. In the best case, featuring a two-carbon tether and a
stereogenic center at the homopropargylic position (complex 1N), the heterocyclic ring was produced
with 97:3 selectivity for the cis heterocycle (3N). In the worst case, featuring a three-carbon tether
and a stereogenic center at the homopropargylic position (complexes 1F and 1I), no stereoselectivity
was observed. Improvement in stereoselectivity was noticed when terminal alkynes were replaced
by silylated alkynes and when proton sources were eliminated from the reaction.
Sch em e 1
The intramolecular coupling of alkynes and cyclopro-
pylcarbene-chromium complexes in aqueous solvents
affords cyclopentenone rings fused to oxygen heterocycles
in good-to-excellent yield (Scheme 1).1 Thermolysis of
alkyne-carbene complex 1A initially produces a cyclo-
pentadienone intermediate (2A),2 which is reduced to the
cyclopentenone vinylogous ester derivatives 3A under the
reaction conditions. A new stereocenter is created at the
asterisked carbon atom during the reduction process. In
the example in Scheme 1, the tethering chain of complex
1A features a stereogenic center, and in this case, the
heterocyclic ring of the products was constructed in a 94:6
trans:cis ratio, favoring the diequatorial isomer (hetero-
cyclic ring substituents trans).3 On the basis of the high
degree of diastereoselectivity observed in this system,
coupled with the myriad of methods available for ma-
nipulation of the vinylogous ester functionality,4 a wide
variety of cyclopentanoid ring structures can potentially
be prepared stereoselectively by using cyclopentenones
derived from intramolecular alkyne-cyclopropylcarbene
complex coupling reactions. In this manuscript, ther-
molysis of a variety of alkyne-containing cyclopropylcar-
bene complexes that differ in the length of tether and
Sch em e 2
position of the stereogenic center has been examined. The
effect of position and ring size on the degree and direction
of relative asymmetric induction will be discussed.
(1) Herndon, J . W.; Matasi, J . J . J . Org. Chem. 1990, 55, 786-788.
(2) In some cases, this intermediate can be isolated. (a) Herndon,
J . W.; Patel, P. P. Tetrahedron Lett. 1997, 38, 59-62. (b) Tumer, S.
U.; Herndon, J . W.; McMullen, L. A. J . Am. Chem. Soc. 1992, 114,
8394-8404.
(3) Yan, J .; Herndon, J . W. J . Org. Chem. 1998, 63, 2325-2331.
(4) (a) For a review, see: Schick, H.; Eichhorn, I. Synthesis 1989,
477-492. (b) Bhatia, G. S.; Lowe, R. F.; Stoodley, R. J . J . Chem. Soc.,
Chem. Commun. 1997, 1981-1983. (c) Schinzer, D.; Kalesse, M.
Tetrahedron Lett: 1991, 32, 4691-4694. (d) Ohba, M.; Haneishi, T.;
Fujii, T. Chem. Pharm. Bull. 1995, 43, 26-31. (e) Crimmins, M. T.
Chem. Rev. 1988, 88, 1453-1473. (f) Smith, A. B., III; Dorsey, B. D.;
Ohba, M.; Lupo, A. T., J r.; Malamas, M. S. J . Org. Chem. 1988, 53,
4314-4325. (g) Bunnelle, W. H. Adv. Cycloaddit. 1993, 3, 67-98. (h)
Stork, G.; Danheiser, R. L. J . Org. Chem. 1973, 38, 1775-1777.
Resu lts
Syn th esis of Alk yn e-Con ta in in g Cyclop r op ylca r -
ben e Com p lexes. Alkyne-containing cyclopropylcarbene
complexes were readily prepared from the corresponding
alkynols (5) using acylate salt 4 and acetyl chloride
(Scheme 2). The yield for this process was usually greater
than 70% when a primary alcohol was employed but
10.1021/jo982144q CCC: $18.00 © 1999 American Chemical Society
Published on Web 01/22/1999

1292 J . Org. Chem., Vol. 64, No. 4, 1999
Yan et al.
Sch em e 3
of the multistep synthetic route in eq 4. Alkynol 5I, which
features a homopropargylic phenyl group, was prepared
via propargylation of ethyl phenylacetate, followed by
reduction (eq 5). Alkynols 5K and 5L were prepared from
terminal alkynes via the synthetic route in eq 6. Alkynol
5M was prepared from monoprotected diol 8A (eq 7).
Alkynol 5N was prepared from phenylacetylene and
propylene oxide (eq 8).
Solven t Effects on th e Ster eoselectivity of Het-
er ocyclic Rin g Con str u ction . A variety of alkyne-
carbene complexes (1) have been synthesized and con-
verted to the corresponding heterocycle-fused cyclopenten-
one derivatives (3) upon thermolysis in aqueous solvent;
the results are depicted in Table 1. The effect of solvent
on the stereoselectivity of cyclopentadienone reduction
was examined for two of the table entries (L and N).
Thermolysis of alkyne-carbene complex 1N was tested
in both toluene and dioxane at various temperatures. The
observed stereoselectivity for formation of the cis het-
erocyclic ring was considerably higher when the reaction
was conducted in 1% aqueous toluene (97:3) than in 1%
aqueous dioxane (65:35). A similar study was conducted
on complex 1L, and nearly identical stereoselectivity was
observed in toluene (70:30) and dioxane (75:25).1 Only
the results from thermolyses in toluene are presented in
the table. A noteworthy advantage of the toluene solvent
system is that terminal alkynes are more suitable
substrates for the intramolecular cyclopentannulation
reaction, even in cases where a six-membered heterocyclic
ring is formed as a result of intramolecular carbene-
alkyne coupling. As reported in our earlier work and in
related studies by others,1,8 these complexes cyclize
inefficiently unless the heterocyclic ring under construc-
tion contains at least seven atoms.
Su bstr a te Dep en d en ce on th e Ster eoselectivity
of Heter ocyclic Rin g Con str u ction . Thermolysis of a
variety of alkyne-containing complexes is reported in
Table 1. Five-, six-, and seven-membered heterocyclic
ring-forming reactions were tested. Reduction of the
cyclopentadienone intermediate afforded heterocyclic
rings with a high degree of stereoselectivity for three
substrate classes: entries A-C, which feature a propar-
gylic methyl group and a three-carbon tether; entry N,
which features a homopropargylic stereocenter and a two-
carbon tether; and entry O, which features a propar-
gylic stereocenter and a four-carbon tether. Alkyne-
carbene complexes featuring propargylic stereocenters
afforded trans heterocycles as the major product (entries
A-E, M, O), whereas alkyne-carbene complexes fea-
turing homopropargyl stereocenters (entries F-J , N)
afforded cis heterocycles. In the two cases featuring a
more distant stereocenter (entries K and L), predomi-
nantly the cis heterocycle was obtained. In all cases
involving nonsilylated internal alkynes (entries B-C,
L-M), the trans carbocycle was the major product.
Carbocyclic ring stereoselectivity appears to be unrelated
to heterocyclic ring stereoselectivity, because both ter-
minal and internal alkynes usually afforded nearly
identical levels of stereoselectivity in construction of the
heterocyclic ring (compare entries B and C with A).
Comparison of silylated alkynes, which ultimately af-
forded desilylated cyclopentenones, and nonsilylated
substantially less with a secondary alcohol, although
higher yields were observed when an excess of acylate
salt 4 and acetyl chloride was employed.
Alkynol derivatives were prepared according to one of
the general synthetic protocols in Scheme 3. Synthesis
of 4-methyl-5-pentyn-1-ol derivatives (5A-C; substituent
letters correlate with Table 1 entry letters for compounds
1-3, 5, and 12-17) has previously been discussed.3
Alkynols bearing a propargylic phenyl group (5D, E, and
O) were prepared by treatment of the dianion5 of 3-phen-
yl-1-propyne6 (6) with either ethylene oxide7 or oxetane
(eq 1). Many of the alkynol derivatives were prepared
from monoprotected diols (8) derived from methallyl
alcohol (7A) or 3-methyl-3-buten-1-ol (8A) (eq 2). Alkynol
5F was prepared from monoprotected diol 8A by means
(5) Hommes, H.; Verkruijsse, H. D.; Brandsma, L. Recl. Trav. Chim.
Pays-Bas 1980, 99, 113-114.
(6) This surprisingly expensive compound was readily prepared by
isomerization of 1-phenyl-1-propyne. Mulvaney, J . E.; Folk, T. L.;
Newton, D. J . J . Org. Chem. 1967, 32, 1674-1675.
(7) Funk, R. L.; Vollhardt, K. P. C. J . Am. Chem. Soc. 1980, 102,
5245-5253.
(8) (a) Wulff, W. D.; McCallum, J . S.; Kunng, F.-A. J . Am. Chem.
Soc. 1988, 110, 7419-7434. (b) Semmelhack, M. F.; Bozell, J . J .; Keller,
L.; Sato, T.; Speiss, E. J .; Wulff, W. D.; Zask, A. Tetrahedron 1985,
41, 5803-5812.

Five-Membered Rings Fused to Oxygen Heterocycles
J . Org. Chem., Vol. 64, No. 4, 1999 1293
Ta ble 1. Dep en d en ce of Str u ctu r a l Va r ia bles on th e Ster eoselectivity of Heter ocyclic Rin g Con str u ction
alkynes is not consistent with this observation (see
below). Alkyne-carbene complexes featuring a propar-
gylic phenyl group (entries D-E) afforded cyclopentapy-
ran derivatives with a lower degree of stereoselectivity
than the analogous complexes featuring a propargylic
methyl group (entries A-C).

1294 J . Org. Chem., Vol. 64, No. 4, 1999
Yan et al.
Sch em e 4
Optim ization of Ster eoselectivity for En tr ies F -J .
Some attempts to further optimize the stereoselectivity
in the least selective system, entries F-J , were under-
taken. A variety of additives that might also serve as
ligands or cyclopentadienone reductants were added to
the reaction mixture. Reducing additives included iron
pentacarbonyl, chromium hexacarbonyl, molybdenum
hexacarbonyl, and stannous chloride. Only iron penta-
carbonyl produced a stereoselectivity enhancement; how-
ever, the yield of cyclopentenones was severely reduced.
The only noticeable stereoselectivity enhancement oc-
curred when tris(o-tolyl)phosphine was present during
the reaction; the stereoselectivity increased from 1:1 to
2:1. In the cases involving homopropargylic stereocenters
and three-carbon tethers (entries F-J ), silylated alkynes
afforded heterocyclic rings in greater stereoselectivity
than terminal alkynes. A similar effect was noted for
entries D and E, which feature a propargylic phenyl
group.
Dep en d en ce of Ster eoselectivity on Wa ter Con -
cen tr a tion . The amount of water added to the system
had a dramatic effect on the stereoselectivity of hetero-
cyclic ring formation for entries E, F, and O. In the
absence of any proton source,9 the reaction was com-
pletely stereoselective, forming only the trans heterocycle;
however, as expected the yield was very low (30%). When
only enough water to balance the reaction equation was
used (2 equiv), the stereoselectivity was lower and the
yield was only marginally improved (36%).
stereomers 14 and 15,13,14 which would afford the ob-
served products 3A-trans and 3A-cis after addition of a
second proton. Thus, the more stable stereoisomer (14
vs 16) is the major product of the reaction. For a more
detailed discussion of this theory, see ref 3.
Discu ssion
Ra tion a liza tion of th e Obser ved Ster eoch em is-
tr y. As noted in previous papers,1,2b the trans stereo-
chemistry is anticipated for the carbocyclic ring, and this
seems to be observed in all cases. With the exception of
entries K and L, the heterocyclic ring in which the two
substituents were either diequatorial or pseudodiequa-
torial was the major reaction product.
The major product of entries K and L is the cis
heterocycle,15 which would not be the more stable isomer
on the basis of simple axial vs equatorial arguments.
Because there is substantial resonance interaction in the
vinylogous ester system, there is substantial double bond
character in the sp2-carbon-oxygen bond of the hetero-
cyclic ring. In a simple system featuring similar reso-
nance interaction, δ-valerolactone, the boat conformation
was determined to be slightly more stable than the
chairlike form,16 and substituted derivatives can adopt
either form.17 If the boat conformation is more stable, the
trans isomer of compounds featuring the 1,4-stereochem-
ical relationship would be destabilized because one of the
substituents would be forced to occupy an axial position
at the prow of the boat.
On the basis of previous studies,2b,3 the mechanistic
events in Scheme 4 best account for the observed stereo-
selectivity. Complexation of chromium at the cyclopen-
tadienone oxygen10 affords complex 12. Resonance struc-
ture 13, which features a cyclopentadienyl anion, would
afford the corresponding cyclopentadiene (e.g., 14-16)11
upon monoprotonation. A rapid series of 1,5-hydrogen
shifts12 would afford a net equilibrium between dia-
(9) In the absence of an external proton source, the unreacted
carbene complex might function as the proton source. (a) H/D deute-
rium exchange was observed upon thermolysis of a related cyclopro-
pylcarbene-chromium complex in DMF/D2O. Herndon, J . W.; Mc-
Mullen, L. A. J . Am. Chem. Soc. 1989, 111, 6854-6856. (b) For a more
detailed account of carbene complex acidity and exchange reactions,
see Bernasconi, C. F. Chem. Soc. Rev. 1997, 26, 299-308.
In many cases, silicon-substituted alkynes afforded
different stereoisomeric ratios than terminal alkynes. A
possible reason for this effect is that, under the reaction
conditions, the initially produced R-silyl ketone can
(10) Both oxygen- and η4-diene complexes have been proposed in
the reaction of cyclopentadienones with group VI metal carbonyls.
These results are more consistent with a simple reduction-protonation
sequence, which employs an oxygen-bound complex. (a) Brown, D. A.;
Hargaden, J . P.; McMullin, C. M.; Gogan, N.; Sloan, H. J . Chem. Soc.
1963, 4914-4918. (b) Adams, H.; Bailey, N. A.; Hempstead, P. D.;
Morris, M. J .; Riley, S.; Beddoes, R. L.; Cook, E. S. J . Chem. Soc.,
Dalton Trans. 1993, 91-100.
(13) Compounds 14 and 15 would appear to be the major isomers
at equilibrium since 3-alkoxy-2-cyclopenten-1-one derivatives (vinyl-
ogous esters) are formed almost exclusively and not 4-alkoxy-2-
cyclopenten-1-ones or 3-alkoxy-3-cyclopenten-1-ones. Previous studies2b
have shown that 4-alkoxy-2-cyclopenten-1-ones do not rearrange to
vinylogous esters under the conditions for the cyclopentannulation
reaction.
(14) The scrambling of hydrogen atoms was confirmed through
deuterium labeling studies; see ref 2b.
(15) The stereochemistry depicted for the major isomer of 3L is the
cis heterocycle-trans carbocyle; the original stereochemical assignment
in ref 1 is in error.
(11) In theory, any of the cyclopentadiene regioisomers could be
obtained kinetically upon protonation; however, the reaction temper-
ature is substantially greater than the temperature required for 1,5-
hydrogen shifts in cyclopentadienes and thus rapid equilibration of
all cyclopentadiene isomers is anticipated.
(12) For a recent example of a 1,5-hydrogen shift of a cyclopenta-
dienol anion in a protic medium, see: Clark, W. N.; Tickner-Eldridge,
A. M.; Huang, G. K.; Pridgen, L. N.; Olsen, M. A.; Mills, R. J .; Lantos,
I.; Baine, N. H. J . Am. Chem. Soc. 1998, 120, 4550-4551.
(16) Allinger, N. L.; Chang, S. H. M. Tetrahedron 1977, 33, 1561-
1567.
(17) Axiotis, S.; Dreux, J .; Perrin, M.; Royer, H. Tetrahedron 1982,
38, 499-504.

Five-Membered Rings Fused to Oxygen Heterocycles
J . Org. Chem., Vol. 64, No. 4, 1999 1295
Sch em e 5
undergo thermodynamic equilibration during the desilyl-
ation process, which could proceed through either an
enolate or a silyl enol ether (e.g., 17G, Scheme 5)18 that
can further equilibrate through 1,5-hydrogen shifts.
Thus, the reactions of nonsilylated alkynes might reflect
the thermodynamic preferences of the chromoxycyclo-
pentadiene intermediates 14-16, whereas the silylated
alkynes reflect the thermodynamic preferences of the
siloxycyclopentadiene intermediates (e.g., 17G). Alter-
natively, if the second proton is added prior to complete
equilibration of the cyclopentadiene intermediate, the
silylated systems have a second chance to equilibrate
through the mechanism depicted in Scheme 5.
A correlation between stereoselectivity and amount of
water additive was also noted. Five-membered ring-
containing products are produced in low yield from the
coupling of cyclopropylcarbene-chromium complexes and
alkynes in the absence of water.2b However, superior
relative asymmetric induction was observed under these
conditions for substrates containing propargylic phenyl
groups (entries D, E, and O). A possible origin of this
effect is enhanced equilibration of the cyclopentadiene
intermediates (e.g., 14 and 15). In the absence of an
added proton source, the starting carbene complex could
serve as the proton source. Because there are now
insufficient protons to balance the reaction equation,
there is greater opportunity for the reaction intermedi-
ates 14 and 15 to equilibrate prior to addition of the
second proton. Thermal decomposition of the anionic
carbene complex19 could then account for the low yield
of cycloadducts.
F igu r e 1.
The stereochemistry of compound 3I-cis was assigned
through X-ray analysis; the X-ray structure also showed
that 3I-cis exists in the chair conformation in the solid
state. Stereochemical assignment of the major isomer for
a related compound, 3F -cis, was based on observed
similarities in the proton NMR spectra of compounds 3F -
cis and 3I-cis. Most notably, both compounds feature a
long-range coupling (1.3-1.4 Hz) for the equatorial
proton next to oxygen.
The stereochemistry of 3K-trans (the minor isomer)
was assigned on the basis of a NOESY experiment
(Figure 1); the axial proton â to oxygen (δ 1.65) exhibits
a cross-peak with both the bridgehead hydrogen (δ 2.68)
and the methyl group (δ 1.40), which implies that the
methyl group and the bridgehead hydrogen are cis (i.e.,
the trans heterocycle). Thus, the other isomer, 3K-cis,
was assigned as cis. Assignment of the stereochemistry
for the products of entry L was based upon chemical shift
similarities to the stereoisomers of 3K, in which the
proton next to oxygen is less shielded in the cis isomer
(δ 4.46 in 3K-cis and δ 4.12 in 3K-trans). This same
proton in the products from entry L is less shielded in
the cis heterocycles (δ 4.40 in 3L-cis-trans and d 4.45
in 3L-cis-cis) than in the trans heterocycles (δ 4.10 in
both 3L-trans-trans and 3L-trans-cis). For compounds
featuring three stereocenters, the first cis-trans designa-
tion refers to the stereochemistry in the heterocyclic ring,
and the second designation refers to the stereochemistry
in the carbocyclic ring.
Assign m en t of Ster eoch em istr y. In all of the cases
involving nonterminal alkynes, the carbocyclic ring ster-
eochemistry was assigned on the basis of the coupling
constant between the protons at the 4- and 5-positions
(cyclopentenone numbering) of the cyclopentenone rings.
The couplings between cis protons are usually larger (>7
Hz) than couplings between trans protons (<5 Hz).
Assignment of heterocyclic ring stereochemistry for
compounds featuring a propargylic stereocenter and a
three-carbon tether (entries A-E) have previously been
discussed.3
An additional noteworthy observation concerning six-
membered ring heterocycles deserves comment. If both
of the six-membered ring substituents can be diequato-
rial, the coupling constants of the heterocyclic ring
protons suggest a chair conformation, whereas the ob-
served couplings are inconsistent with a chair conforma-
tion if one of the substituents would be axial in the chair
conformation. Thus, in the chair conformation, each
proton of the heterocyclic ring is either axial (featuring
one large geminal coupling, a similar coupling constant
to the axial protons on the adjacent carbon, and a small
coupling to the equatorial protons on the adjacent carbon)
(18) (a) Brook, A. G.; MacRae, D. M.; Limburg, W. W. J . Am. Chem.
Soc. 1967, 89, 5493-5495. (b) Larson, G. L.; Berrios, R.; Prieto, J . A.
Tetrahedron Lett. 1989, 30, 283-286.
(19) See: Bernasconi, C. F.; Leyes, A. E.; Ragains, M. L.; Shi, Y.;
Wang, H.; Wulff, W. D. J . Am. Chem. Soc. 1998, 120, 8632-8639, and
references therein.

1296 J . Org. Chem., Vol. 64, No. 4, 1999
Yan et al.
or equatorial (featuring one large geminal coupling and
small couplings to all the protons on the adjacent
carbons).
low-yielding cycloaddition performed in anhydrous sol-
vent proceeded with a very high degree of stereoselec-
tivity in the systems examined and may hold the key to
future development of a high-yielding stereoselective
synthesis of cyclopentenones fused to oxygen hetero-
cycles.
The heterocyclic ring stereochemistry for the products
1
of entries L and M was assigned on the basis of their H
NMR spectra coupled with observed and calculated
couplings in a related system, γ-butyrolactone.20 This
compound was chosen as a model because the sp2-
carbon-oxygen bond of 3M and 3N is anticipated to have
substantial double bond character due to resonance
interaction with the carbonyl compound, as is featured
in γ-butyrolactone. The observed coupling constant be-
tween the bridgehead hydrogen and the proton next to
the methyl group in the trans heterocycles (3M-trans-
trans and 3M-trans-cis) is 12.1 and 11.7 Hz. The observed
coupling in the corresponding cis heterocycle is 6.6 Hz.
The analogous calculated couplings in 2,3-dimethyl-γ-
butyrolactone are 10.46 Hz for the trans isomer and 7.44
Hz for the cis isomer. The observed couplings between
the bridgehead protons and the adjacent protons of the
heterocyclic ring of 3N-cis-cis are 13.9 and 7.5 Hz, which
are closer to the values predicted for the cis isomer (12.8
and 8.5 Hz in cis 2,4-dimethyl-γ-butyrolactone and 9.0
and 8.0 Hz in the trans isomer). In compound 3N-trans-
trans, coupling of the proton next to oxygen and the
adjacent protons is 7.8 and 0 Hz, which is very close to
the calculated values of 8.8 and 1.6 Hz in trans 2,4-
dimethylbutyolactone.
The stereochemistry of the major isomer in 3O was
assigned as trans on the basis of trends established in
the five- and six-membered ring heterocycles. The ben-
zylic proton is coupled to protons on adjacent carbons
with coupling constants of 11.2, 11.2, and 3.2 Hz, which
is consistent with an axial or (pseudoaxial) proton
coupling to two other axial protons and an equatorial
proton.21 The same proton in the minor isomer occurs at
δ 3.29, and is coupled to protons on adjacent carbons with
coupling constants totaling 3.7 Hz, which is consistent
with an equatorial (or pseudoequatorial) proton coupling
to three other axial or equatorial protons.
Exp er im en ta l Section
For general experimental procedures, see ref 2b. The
experimental procedure for entry C has been presented in a
previous full paper.3
Gen er a l P r oced u r e I: Syn th esis of Alk yn e-Con ta in in g
Ca r ben e Com p lexes. To a solution of acylate salt 422 (0.40
g, 1.20 mmol) in dichloromethane (20 mL) at 0 °C under
nitrogen was added acetyl chloride (0.08 mL, 1.20 mmol) via
syringe, followed by the immediate addition of alkynol (1.20
mmol). The reaction mixture was stirred at 0 °C for about 1
h, warmed to 25 °C, and stirred for 20 min. The solvent was
removed on a rotary evaporator. Flash chromatography of the
residue after evaporation, with hexane as the eluent, gave the
pure carbene complex after solvent removal.
P r ep a r a tion of Ca r ben e Com p lex 1A. To a solution of
3-methyl-1-(tert-butyldimethylsilyloxy)-4-pentyne (350 mg, 1.81
mmol) in dichloromethane (20 mL) at room temperature was
added boron trifluoride etherate (741 mg, 5.21 mmol). The
reaction was stirred at room temperature for 2 h. The reaction
solution was poured into water (20 mL) and extracted with
ethyl ether (3 × 20 mL). The organic layers were combined
and dried over anhydrous magnesium sulfate. The solvent was
partially removed by distillation until about 20 mL of the
solution remained (because of volatility of the product), and
then General Procedure I was followed using this solution,
acylate salt 4 (643 mg, 1.81 mmol), and acetyl chloride (0. 13
mL, 1.81 mmol) in dichloromethane (20 mL). After chromato-
graphic purification, a yellow oil (245 mg, 40%) identified as
1
carbene complex 1A was obtained. H NMR (CDC13): d 5.03
(t, 2 H, J ) 6.1 Hz), 3.45 (m, 1 H), 2.62 (m, 1 H), 2. 1 0 (d, 1
H, J ) 2.6 Hz), 2.02 (m, 2 H), 1.33 (m, 2 H), 1.26 (d, 3 H, J )
7.0 Hz), 1.23 (m, 2 H). 13C NMR (CDC13): δ 352.0, 223.6, 216.8,
86.5, 78.4, 70.0, 41.4, 36.0, 22.9, 21.0, 17.8. IR (CC14): 3303
(m) 2061 (vs), 1952 (vs) cm-1. MS (CI): m/e 343 (M + 1, 8),
342 (M, 29), 52 (100). HRMS: calcd for C15H14CrO6 342.0196,
found 342.0174.
P r ep a r a tion of Ca r ben e Com p lex 1B. General Procedure
I was followed using a solution of 3-methyl-5-phenyl-4-pentyn-
l-ol (311 mg, 1.79 mmol) in dichloromethane (10 mL), acylate
salt 4 (634 mg, 1.79 mmol), and acetyl chloride (0.13 mL, 1.7
mmol) in dichloromethane (20 mL). After chromatographic
purification, a yellow oil (516 mg, 69%) identified as carbene
Con clu sion
The diastereoselectivity of the intramolecular coupling
of alkynes and cyclopropylcarbene-chromium complexes
has been evaluated for a diverse array of alkyne-carbene
complexes featuring a chiral carbon in the tether. Re-
gardless of ring size, the major product was trans if the
stereocenters of the heterocyclic ring were in a 1,2-
relationship and cis if the stereocenters were in a 1,3- or
1,4-relationship. Excellent diastereoselectivity was ob-
served as a result of the cyclopentadienone reduction in
many but not all cases. The extent of diastereoselectivity
was very dependent on the position of the chiral center
and the ring size of the heterocyclic ring produced. The
1
complex 1B was obtained. H NMR (CDC13): δ 7.24 (m, 5 H),
5.02 (dd, 2 H, J ) 11.8, 6.2 Hz), 3.40 (m, 1 H), 2.78 (m, 1 H),
2.05 (m, 2 H), 1.47 (m, 2 H), 1.30 (d, 3 H, J ) 6.9 Hz), 1.14 (m,
2 H). 13C NMR (CDC13): δ 351.9, 223.5, 216.8, 131.6, 128.2,
127.8, 123.3, 91.8, 82.2, 41.4, 36.2, 23.8, 21.2, 17.8. IR (CC14):
2054 (s), 1936 (vs) cm-1. MS (CI): m/e 419 (M + 1, 1), 418 (M,
2), 226 (100). HRMS: calcd for C2lHl8CrO6 418.0509, found
418.0549.
P r ep a r a tion of Ca r ben e Com p lex 1D. General Procedure
I was followed using a solution of 3-phenyl-4-pentyn-1-ol (240
mg, 1.50 mmol) in dichloromethane (10 mL), acylate salt 4 (506
mg, 1.50 mmol), and acetyl chloride (0.106 mL, 1.50 mmol) in
dichloromethane (20 mL). After chromatographic purification,
a yellow oil (383 mg, 63%) identified as carbene complex 1D
was obtained. 1H NMR (CDCl3): δ 7.4 (m, 5 H), 5.02 (m, 2 H),
3.89 (t, 1 H, J ) 7.5 Hz), 3.50 (m, 1 H), 2.29 (m, 2 H), 1.1-1.5
(m, 4 H), -0.21 (s, 9 H). 13C NMR (CDCl3): δ 224.1, 217.3,
140.8, 129.5, 129.0, 106.0, 89.7, 79.3, 41.9, 39.0, 36.1, 19.2, 0.1.
(20) (a) Experimental studies: Hussain, S. A. M. T.; Ollis, W. D.;
Smith, C.; Stoddart, J . F. J . Chem. Soc., Perkin Trans. 1 1975, 1480-
1492. (b) Theoretical studies: J aime, C.; Ortun˜o, R. M.; Font, J . J .
Org. Chem. 1986, 51, 3946-3951.
(21) (a) Conformational studies of cycloheptane derivatives suggest
that cycloheptane exists largely in a chairlike conformation. Elliel, E.
L.; Wilen, S. H. In Stereochemistry of Organic Compounds; Wiley-
Interscience: New York, 1994; pp 762-765. (b) The chair conformation
is more pronounced in cycloheptene, which might be a better model
due to the double bond character in the sp2 C-O bond. Ermolaeva, L.
I.; Mastryukov, V. S.; Allinger, N. L. Almenningen, A. J . Mol. Struct.
1989, 196, 151-156. c. Menard, D.; St.-J acques, M. Tetrahedron 1983,
39, 1041-1060.
(22) For a procedure for synthesis of this compound, see: Herndon,
J . W.; Tumer, S. U.; McMullen, L. A.; Matasi, J . J .; Schnatter, W. F.
K. In Advances in Metal-Organic Chemistry; Liebeskind, L. S., Ed.;
J AI Press: Greenwich, CT, 1994; Vol. III, pp 51-95.

Five-Membered Rings Fused to Oxygen Heterocycles
J . Org. Chem., Vol. 64, No. 4, 1999 1297
IR (neat): 2061, 1982, 1929 cm-1. An acceptable mass spec-
trum or elemental analysis could not be obtained for this
molecule.23
acylate salt 4 (614 mg, 1.84 mmol), and acetyl chloride (0.13
mL, 1.84 mmol) in dichloromethane (20 mL). After chromato-
graphic purification, a yellow oil (596 mg, 68%) identified as
carbene complex 1I was obtained. 1H NMR (CDCl3): δ 7.26
(m, 5 H), 5.26 (dd, 1 H, J ) 10.4, 5.9 Hz), 5.08 (dd, 1 H,10.3,
8.1 Hz), 3.35 (m, 2 H), 2.65 (d, 2 H, J ) 6.2 Hz), 1.04 (m, 4 H),
0.11 (s, 9 H). 13C NMR (CDCl3): δ 352.3, 223.4, 216.6, 139.9,
128.7, 127.4, 102.8, 87.9, 81.8, 44.0, 41.6, 23.9, 17.9, 17.8,
-0.24. IR (CCl4): 2060 (s), 1923 (vs) cm-1. MS (CI): m/e 477
P r ep a r a tion of Ca r ben e Com p lex 1E. . General Proce-
dure I was followed using a solution of 3-phenyl-5-trimethyl-
silyl-4-pentyn-1-ol (230 mg, 0.99 mmol) in dichloromethane (10
mL), acylate salt 4 (338 mg, 1.00 mmol), and acetyl chloride
(0.071 mL, 1.00 mmol) in dichloromethane (20 mL). After
chromatographic purification, a yellow oil (293 mg, 62%)
identified as carbene complex 1E was obtained. 1H NMR
(CDCl3) δ 7.3 (m, 5 H), 5.04 (m, 2 H), 3.90 (t, 1 H, J ) 7.1 Hz),
3.50 (m, 1 H), 2.32 (m, 2 H), 1.45-1.15 (m, 4 H), 0.21 (s, 9 H).
13C NMR: δ 223.6, 216.6, 140.1, 128.8, 128.4, 127.2, 105.5,
89.0, 78.0, 59.9, 41.4, 37.5, 35.7, 17.8, -0.1. IR (neat): 2178
(w), 2061 (m), 1982 (m), 1929 (vs) cm-1. An acceptable mass
spectrum or elemental analysis could not be obtained for this
molecule.23
P r ep a r a tion of Ca r ben e Com p lex IF . To a solution of
4-methyl-5-(tert-butyldimethylsilyloxy)-1-pentyne (683 mg, 3.53
mmol) in dichloromethane (20 mL) at room temperature was
added boron trifluoride etherate (741 mg, 5.21 mmol). The
reaction was stirred at room temperature for 2 h. The reaction
solution was poured into water (20 mL) and extracted with
ethyl ether (3 × 20 mL). The organic layers were combined
and dried over anhydrous magnesium sulfate. The solvent was
partially removed by distillation until about 20 mL of the
solution remained (because of volatility of the product), and
then General Procedure I was followed using acylate salt 4
(1.18 g, 3.53 mmol) and acetyl chloride (0.21 mL, 3.53 mmol)
in dichloromethane (20 mL). After chromatographic purifica-
tion, a yellow oil (785 mg, 65%) identified as carbene complex
1F was obtained. 1H NMR (CDCl3): δ 4.82 (d, 2 H, J ) 5.6
Hz), 3.48 (m, 1 H), 2.29 (m, 3 H), 2.02 (t, 1 H, J ) 3.9 Hz),
1.38 (m, 2 H), 1.19 (m, 2 H), 1.13 (d, 3 H, J ) 6.8). 13C NMR
(CDC13): δ 352.2, 223.5, 216.7, 83.2, 80.7, 70.6, 41.5, 32.2, 22.5,
17.6, 16.2. IR (CC14): 2060 (s), 1936 (vs) cm-1. MS (CI): m/e
342 (M, 6), 150 (100). HRMS: calcd for C15H14CrO6 342.0196,
found 342.0211.
(M + 1, 2), 476 (M, 4), 104 (100). HRMS: calcd for C23H24
-
CrO6Si 476.0747, found 476.0744.
P r ep a r a tion of Ca r ben e Com p lex 1J . General Procedure
I was followed using a solution of 2-methyl-5-triethylsilyl-4-
pentyn-1-ol (285 mg, 1.34 mmol) in dichloromethane (20 mL),
acylate salt 4 (450 mg, 1.34 mmol), and acetyl chloride (0.10
mL, 1.34 mmol) in dichloromethane (20 mL). After chromato-
graphic purification, a yellow oil (428 mg, 70%) identified as
carbene complex 1J was obtained. 1H NMR (CDCl3): δ 4.82
(d, 2 H, J ) 5.8 Hz), 3.46 (m, 1 H), 2.37 (d, 2 H, J ) 5.6 Hz),
2.27 (m, 1 H), 1.34 (m, 2 H), 1.21 (m, 2 H), 1.15 (d, 3 H, J )
6.6 Hz), 0.96 (t, 9 H, J ) 7.6 Hz), 0.54 (q, 6 H, J ) 7.8). 13C
NMR (CDCl3): δ 351.9, 223.5, 216.7, 104.2, 84.4, 83.2, 41.5,
33.0, 24.0, 17.6, 16.2, 7.3, 4.4. IR (CCl4): 2060 (s), 1941 (vs),
1981 (s) cm-1. MS (CI): m/e 456 (M, 3), 235 (100). HRMS: calcd
for C2lH28CrO6Si 456.1016, found 456.1021.
P r ep a r a tion of Ca r ben e Com p lex 1K. General Procedure
I was followed using a solution of 6-trimethylsilyl-5-hexyn-2-
ol (321 mg, 1.89 mmol) in dichloromethane (20 mL), acylate
salt 4 (1.27 g, 3.78 mmol), and acetyl chloride (0.27 mL, 3.78
mmol) in dichloromethane (20 mL). After chromatographic
purification, a yellow oil (297 mg, 38%) identified as carbene
1
complex 1K was obtained. H NMR (CDCl3): δ 5.66 (m, 1 H),
3.38 (m, 1 H), 2.33 (m, 2 H), 2.00 (m, 2 H), 1.42 (d, 3 H, J )
6.2 Hz), 1.29 (m, 4 H), 0.13 (s, 9 H). 13C NMR (CDCl3): δ 347.8,
223.8, 216.6, 104.8, 87.5, 85.7, 41.3, 34.6, 19.9, 17.5, 17.3, 15.8,
-0.7. IR (CCl4): 2060 (s), 1920 (vs) cm-1. MS (CI): m/e 414
(M, 4), 272 (100). HRMS: calcd for C18H22CrOSi 414.0591,
found 414.0562.
P r ep a r a tion of Ca r ben e Com p lex 1L. General Procedure
I was followed using 6-phenyl-5-hexyn-2-ol (0.196 g, 1.10
mmol), acylate salt 4 (0.767 g, 2.30 mmol), and acetyl chloride
(0.160 mL, 2.20 mmol) in dichloromethane (40 mL). After
purification by flash chromatography on silica gel using pure
hexane as the eluent, a yellow oil (0.267 g, 56%) identified as
carbene complex 1L was obtained. 1H NMR (CDCl3): δ 7.38-
7.24 (m, 5 H), 5.68 (quintet, 1 H, J ) 6.2 Hz), 3.40 (m, 1 H),
2.50 (m, 2 H), 3.40 (m, 1 H), 2.50 (m, 2 H), 2.08 (m, 2 H), 1.44
(d, 3 H, J ) 6.2 Hz), 1.39-1.08 (m, 4 H). 13C NMR (CDCl3): δ
347.9, 223.5, 216.7, 131.5, 128.2, 127.5, 123.5, 88.0, 87.6, 81.5,
41.4, 35.0, 20.2, 17.7, 17.4, 15.4. IR (CH2Cl2): 2061 (s), 1979
(sh), 1938 (vs) cm-1. MS (EI): 418 (M, 2), 226 (100). HRMS:
calcd for C21H18CrO6 418.0508, found 418.0522.
P r ep a r a tion of Ca r ben e Com p lex 1G. General Procedure
I was followed using a solution of 2-methyl-5-trimethylsilyl-
4-pentyn-1-ol (296 mg, 1.74 mmol) in dichloromethane (10 mL),
acylate salt 4 (528 mg, 1.74 mmol), and acetyl chloride (0.12
mL, 1.74 mmol) in dichloromethane (20 mL). After chromato-
graphic purification, a yellow oil (533 mg, 74%) identified as
1
carbene complex 1G was obtained. H NMR (CDC13): δ 4.81
(dd, 2 H, J ) 5.7, 2.0 Hz), 3.47 (m, 1 H), 2.30 (m, 3 H), 1.37
(m, 2 H), 1.19 (m, 2 H), 1.11 (d, 3 H, J ) 6.6 Hz), 0.12 (s, 9 H).
13C NMR (CDC13): δ 352.0, 223.5, 216.8, 103.2, 83.4, 41.6, 33.0,
24.0, 17.7, 16.4, -0.03. IR (CCl4): 2061 (s), 1928 (vs) cm-1
.
-
MS (CI): m/e 414 (M, 3), 272 (100). HRMS: calcd for C18H22
CrO6Si 414.0591, found 414.0604.
P r ep a r a tion of Ca r ben e Com p lex 1H. General Proce-
dure I was followed using a solution of 2-phenyl-4-pentyn-1-
ol (310 mg, 1.94 mmol) in dichloromethane (20 mL), acylate
salt 4 (649 mg, 1.94 mmol), and acetyl chloride (0.14 mL, 1.94
mmol) in dichloromethane (20 mL). After chromatographic
purification, a yellow oil (557 mg, 71%) identified as carbene
P r ep a r a tion of Ca r ben e Com p lex 1M. General Proce-
dure I was followed using a solution of 2-methyl-4-phenyl-3-
butyn-1-ol (298 mg, 1.86 mmol) in dichloromethane (10 mL),
acylate salt 4 (620 mg, 1.86 mmol), and acetyl chloride (146
mg, 1.86 mmol) in dichloromethane (20 mL). After chromato-
graphic purification, a yellow oil (511 mg, 68%) identified as
1
complex 1H was obtained. H NMR (CDCl3): δ 7.27 (m, 5 H),
1
carbene complex 1M was obtained. H NMR (CDCl3): δ 7.28
5.20 (dd, 1 H, J ) 10.4, 6.0 Hz), 5.06 (dd, 1 H, J ) 10.2, 7.6
Hz), 3.35 (m, 2 H), 2.62 (dd, 2 H, J ) 7.1, 2.6 Hz), 2.01 (t, 1 H,
J ) 2.6 Hz), 1.07 (m, 4 H). 13C NMR (CDCl3): δ 352.5, 223.4,
216.6, 139.6, 128.8, 127.6, 127.3, 81.9, 80.5, 71.2, 44.0, 41.7,
22.3, 17.9. IR (CCl4): 2060 (s), 1935 (vs) cm-1. MS (CI): m/e
404 (M, 1), 212 (79), 104 (83). HRMS: calcd for C20H16CrO6
404.0352, found 404.0385.
(m, 5 H), 4.85 (m, 2 H), 3.42 (m, 1 H), 3.15 (m, 1 H), 1.47 (m,
2 H), 1.34 (d, 3 H, J ) 7.0 Hz), 1.15 (m, 2 H). 13C NMR
(CDCl3): δ 352.2, 223.4, 216.7, 131.6, 128.3, 128.1, 123.0, 89.2,
82.6, 41.8, 27.6, 18.1, 17.6. IR (CCl4): 2062 (s), 1936 (vs) cm-1
.
MS (CI): m/e 404 (M, 8), 212 (99), 129 (100). HRMS: calcd for
C
20H16CrO6 404.0352, found 404.0376.
P r ep a r a tion of Ca r ben e Com p lex 1N. General Procedure
P r ep a r a tion of Ca r ben e Com p lex 1I. General Procedure
I was followed using a solution of 2-phenyl-5-trimethylsilyl-
4-pentyn-l-ol (425 mg, 1.84 mmol) in dichloromethane (20 mL),
I was followed using 5-phenyl-4-pentyn-2-ol (0.197 g, 1.25
mmol), acylate salt 4 (0.836 g, 2.50 mmol), and acetyl chloride
(0.180 mL, 2.50 mmol) in dichloromethane (40 mL). After
purification by flash chromatography on silica gel using pure
hexane as the eluent, a yellow oil identified as carbene complex
IN (0.177 g, 36%) was obtained. 1H NMR (CDCl3): δ 7.32-
7.15 (m, 5 H), 5.22 (qt, 1 H, J ) 6.2, 5.4 Hz), 3.27 (m, 1 H),
2.30 (d, 2 H, J ) 5.4 Hz), 1.52 (d, 3 H, J ) 6.2 Hz), 1.48-1.30
(m, 2 H), 1.20-1.05 (m, 2 H). 13C NMR (CDCl3): δ 348.6, 223.5,
(23) The stability of this complex was not noticeably different from
that of the other carbene complexes. We attribute the failure to acquire
these data to decomposition during shipment. Successful acquisition
of mass spectral data for alkyne-carbene complexes was possible at
the University of Maryland as a result of the presence of in-house high-
resolution mass spectrometry facilities.

1298 J . Org. Chem., Vol. 64, No. 4, 1999
Yan et al.
216.7, 131.6, 128.3, 128.1, 123.1, 85.4, 83.7, 41.6, 27.3, 19.9,
3B-trans-trans, see below) coupling constant (7.8 Hz) between
the allylic proton and the proton next to the methyl group.
The compound in the other fraction (110 mg, 57%) was
identified as 3B-trans-trans. 1H NMR (CDCl3): δ 7.21 (m, 5
H), 5.36 (d, 1 H, J ) 1.5 Hz), 4.36 (ddd, 1 H, J ) 4.8, 6.1, 11.3
Hz), 4.13 (ddd, 1 H J ) 11.3, 8.1, 4.5 Hz), 3.28 (d, 1 H, J ) 3.2
Hz), 2.56 (ddd, 1 H, J ) 11.0, 3.2, 1.5 Hz), 2.00 (dddd, J )
13.0, 10.6, 6.1, 4.5 Hz), 2.05-1.6 (m, 2 H), 1.54 (dddd, 1 H, J
) 13.0, 13.0, 8.1, 4.8 Hz), 0.93 (d, 3 H, J ) 6.5). 13C NMR
(CDCl3): δ 203.5, 189.8, 139.5, 128.7, 128.1, 126.9, 106.3, 68.7,
58.7, 52.1, 31.7, 20.6. IR (CCl4): 1679 (s), 1598 (s) cm-1. MS
(EI): m/e 228 (M, 100). HRMS: calcd for C15H16O2 228.1150,
found 228.1144. The relative configuration on the cyclopen-
tenone ring was assigned as trans on the basis of the small
coupling constant (3.2 Hz) between the hydrogen at the
benzylic and allylic positions. The relative configuration of
substituents on the six-membered heterocycle was established
as trans on the basis of the large coupling constant (11.0 Hz)
between the allylic proton and the proton next to the methyl
group (see the text), which is consistent with the coupling of
two axial protons.
18.2, 17.5. IR (CDCl3): 2161 (vs), 1982 (sh), 1926 (vs), cm-1
;
MS (EI): 404 (M, 3), 212 (100), 141 (100). HRMS: calcd for
C20H16CrO6 404.0354, found 404.0352.
P r ep a r a tion of Ca r ben e Com p lex 1O. General Procedure
I was followed using a solution of 4-phenyl-5-hexyn-1-ol (250
mg, 1.02 mmol) in dichloromethane (20 mL), acylate salt 4 (355
mg, 1.05 mmol), and acetyl chloride (0.075 mL, 1.05 mmol) in
dichloromethane (20 mL). After chromatographic purification,
a yellow oil (262 mg, 72%) identified as carbene complex 1O
1
was obtained. H NMR (CDCl3): δ 7.4 (m, 5 H), 4.97 (t, 2 H,
J ) 5.9 Hz), 3.80 (m, 1 H), 3.51 (m, 1 H), 2.40 (d, 1 H, J ) 2.4
Hz), 2.0 (m, 4 H), 1.1-1.5 (m, 4 H). 13C NMR (CDCl3): δ 224.1,
217.3, 141.0, 129.2, 127.7, 85.3, 80.5, 72.3, 41.9, 37.6, 34.9, 27.4,
18.2. IR (neat): 2060, 1982, 1941 cm-1. An acceptable mass
spectrum or elemental analysis could not be obtained for this
molecule.23
Gen er a l P r oced u r e II: In tr a m olecu la r Rea ction s of
Ca r ben e Com p lexes w ith Alk yn es. To a three-neck round-
bottom flask equipped with a reflux condenser and rubber
septum, under nitrogen, was added toluene (100 mL) and
water (1 mL), and the solution was heated to reflux. To this
refluxing solution was added a solution of carbene complex in
toluene (30 mL) via syringe pump over a period of 4 h. After
the addition was complete, the mixture was heated at reflux
for an additional 20 h and then cooled to room temperature.
The resulting green mixture was filtered through Celite, and
The second fraction was identified as a mixture of compound
3B-trans-cis and compound 3B-cis-cis (trace). The further
purification was accomplished by preparative thin-layer chro-
matography (2% methanol in dichloromethane) to give 11.6
1
mg (6%) of compound 3B-trans-cis. H NMR (CDCl3): δ 7.18
(m, 5 H), 5.50 (d, 1 H, J ) 1.9 Hz), 4.38 (ddd, 1 H, J ) 11.1,
4.4, 4.4 Hz), 4.07 (ddd, 1 H, J ) 11.1, 11.1, 5.3 Hz), 3.78 (d, 1
H. J ) 7.0 Hz), 2.74 (ddd, 1 H, J ) 10.8, 7.0, 1.9 Hz), 1.80 (m,
1 H), 1.52 (m, 2 H), 0.65 (d, 3 H, J ) 6.0). 13C NMR (CDCl3):
δ 204.4, 190.6, 137.1, 129.7, 129.3, 128.5, 108.6, 69.1, 55.8, 48.5,
31.9, 26.6, 20.2. The relative configuration on the cyclopen-
tenone ring was assigned as cis on the basis of the large
coupling constant (7.0 Hz) between the hydrogen at the
benzylic and allylic positions. The relative configuration of
substituents on the six-membered heterocycle was established
as trans on the basis of the large coupling constant (10.8 Hz)
between the allylic proton and the proton next to the methyl
group (see the text), which is consistent with the coupling of
two axial protons.
the solvent was removed on
a rotary evaporator. Final
purification was achieved by a flash chromatography on silica
gel using 4:1 hexane/ethyl acetate as eluent unless otherwise
noted. In all cases, a separate experiment was performed in
which the individual isomers of 3 were not separated (i.e.,
isolated as a single fraction); the yields and the ratios (obtained
via relative integration of the alkene protons at δ 5.2-5.5)
quoted in Table 1 were obtained from this experiment.
Th er m olysis of Ca r ben e Com p lex 1A. General Procedure
II was followed using carbene complex 1A (225 mg, 0.66 mmol);
two fractions were isolated.
The compound in the first fraction (53.2 mg, 53%) was
identified as compound 3A-trans. 1H NMR (CDCl3): δ 5.27 (d,
1 H, J ) 1.6 Hz), 4.34 (m, 1 H), 4.10 (m, 1 H), 2.55 (dd, 1 H,
J ) 17.5, 6.7 Hz), 2.44 (m, 1 H), 2.06 (dd, 1 H, J ) 17.5, 3.4
Hz), 1.98 (m, 1 H), 1.56 (m, 1 H), 1.03 (d, 3 H, J ) 6.3). 13C
NMR (CDCl3): δ 203.9, 190.9, 107.6, 68.7, 43.1, 40.2, 32.4, 31.5,
20.6. IR (CCl4): 1699 (s), 1601 (s) cm-1. MS (EI): m/e 152 (M,
24), 91 (100). HRMS: calcd for C9H1202 152.0837, found
152.0842. The trans relative configuration of the substituents
on the six-membered ring was assigned to this compound. This
assignment was based on the large coupling constant (13.4 Hz)
between the allylic hydrogen and the hydrogen next to methyl
group, which is consistent with the coupling of two axial
protons, as exists in the trans heterocyclic ring.
Th er m olysis of Ca r ben e Com p lex 1D. General Proce-
dure II was followed using carbene complex 1D (290 mg, 0.655
mmol); two fractions were isolated.
The compound in the first fraction (44 mg, 31%) was
1
identified as 3D-trans. H NMR (CDCl3): δ 7.4-7.2 (m, 10
H), 5.44 (d, 1 H, J ) 1.5 Hz), 4.60 (ddd, 1 H, J ) 11.3, 5.6, 5.6
Hz), 4.30 (ddd, 1 H, J ) 11.3, 8.2, 4.8 Hz), 3.11 (dddd, 1 H, J
) 12.1, 6.6, 3.6, 1.5 Hz), 2.76 (ddd, 1 H, J ) 12.1, 9.1, 6.2 Hz),
2.40 (dd, 1 H, J ) 18.0, 6.6 Hz), 2.20 (m, 2 H), 2.13 (dd, 1 H,
J ) 18.0, 3.6 Hz). 13C NMR (CDCl3): δ 203.6, 189.8, 142.6,
128.9, 127.3, 126.6, 108.3, 69.0, 43.7, 42.1, 40.0, 31.8. IR
(neat): 1686, 1596 cm-1. Combustion Analysis: calcd for
The compound in the second fraction (9.03 mg, 9%) was
C
14H14O2 C 78.48%, H 6.59%; found C 78.29%, H 6.59%. The
1
identified as compound 3A-cis. H NMR (CDCl3): δ 5.33 (d, 1
compound was assigned as the trans isomer on the basis of
the large coupling constant between the bridgehead and
benzylic hydrogens (12.1 Hz), which is consistent with an
axial-axial coupling as exists in the trans isomer.
H, J ) 1.4 Hz), 4.28 (t, 2 H, J ) 6.4 Hz), 3.13 (dddd, 1 H, J )
7.0, 6.5, 3.7, 1.5 Hz), 2.42 (dd, 1 H, J ) 18.0, 6.5 Hz), 2.28 (dd,
1 H, J ) 18.0, 3.7 Hz), 2.24 (m, 1 H), 1.65 (m, 2 H), 0.93 (d, 3
H, J ) 7.0). 13C NMR (CDCl3): δ 204.7, 190.3, 108.1, 66.2,
39.3, 37.1, 31.1, 25.7, 14.5.
The compound in the second fraction (43 mg, 31%) was
identified as 3D-cis. 1H NMR (CDCl3): δ 7.4-7.0 (m, 10H),
5.40 (br s, 1 H), 4.42 (m, 2 H), 3.46 (m, 2 H), 2.49 (m, 1 H),
2.31 (dd, 1 H, J ) 18.0, 7.1 Hz), 2.21 (m, 1 H), 1.84 (dd, 1 H,
J ) 18.0, 2.2 Hz). 13C NMR (CDCl3): δ 203.9, 190.2, 141.1,
128.7, 128.0, 127.1, 107.5, 65.8, 37.6, 37.3, 31.0. The compound
was assigned as the cis isomer on the basis of similarity to
the NMR spectra of compound 3A-cis, most notably the protons
on the same carbon as oxygen have the same chemical shift
in 3D-cis (δ 4.42 Hz); this was also observed for 3A-cis and
3B-cis-trans.
Th er m olysis of Ca r ben e Com p lex 1E. General Procedure
II was followed using carbene complex 1E (286 mg, 0.600
mmol); two fractions were isolated and identified as 3D-trans
(62 mg, 49%) and 3D-cis (21 mg, 16%). The spectral data were
identical to those previously listed for these compounds, which
was also obtained via thermolysis of carbene complex 1D.
Th er m olysis of Ca r ben e Com p lex 1B. General Procedure
II was followed using carbene complex 1B (355 mg, 0.85 mmol);
three fractions were isolated.
The first fraction was identified as a mixture of compound
3B-trans-trans and compound 3B-cis-trans. The further puri-
fication was accomplished by preparative thin-layer chroma-
tography (2% methanol in dichloromethane) to give 9.7 mg
1
(5%) of compound 3B-cis-trans. H NMR (CDCl3): δ 7.20 (m,
5 H), 5.44 (d, 1 H, J ) 1.6 Hz), 4.31 (dd, 1 H, J ) 6.8, 5.6 Hz),
3.50 (d, 1 H, J ) 4.0 Hz), 3.10 (ddd, 1 H, J ) 7.8, 4.0, 1.6 Hz),
2.39-2.21 (m, 2 H), 1.10 (d, 3 H, J ) 7.2). The relative
configuration on the cyclopentenone ring was assigned as trans
on the basis of the small coupling constant (4.0 Hz) between
the hydrogen at the benzylic and allylic positions. The relative
configuration of substituents on the six-membered heterocycle
was established as cis on the basis of the smaller (relative to

Five-Membered Rings Fused to Oxygen Heterocycles
J . Org. Chem., Vol. 64, No. 4, 1999 1299
Th er m olysis of Ca r ben e Com p lex 1F . General Procedure
II was followed using carbene complex 1F (226 mg, 0.66 mmol);
two fractions were isolated.
2.05 (dd, 1 H, J ) 17.9, 2.8 Hz), 1.86 (ddd, 1 H, J ) 14.0, 11.5,
9.8 Hz). 13C NMR (CDCl3): δ 204.0, 191.0, 140.8, 129.1, 127.5,
127.3, 106.2, 71.5, 42.2, 39.2, 33.2, 33.0. IR (CCl4): 1698 (s),
1600 (vs) cm-1. MS (EI): m/e 214 (M, 83), 116 (100). HRMS:
calcd for C14H14O2 214.0994, found 214.0981.
The compound in the first fraction (34.4 mg, 34%) was
1
identified as 3F -cis. H NMR (CDCl3): δ 5.31 (t, 1 H, J ) 1.4
Hz), 4.35 (ddd, 1 H, J ) 11.1, 4.6, 1.4 Hz), 3.66 (dd, 1 H, J )
11.1, 8.4 Hz), 2.81 (m, 1 H), 2.60 (dd, 1 H, J ) 17.6, 6.8 Hz),
2.22 (m, 2 H), 2.12 (dd, 1 H, J ) 17.6, 3.7 Hz), 1.06 (ddd, 1 H,
J ) 12.5, 11.1, 11.1 Hz), 0.95 (d, 3 H, J ) 6.6 Hz). 13C NMR
(CDCl3): δ 204.3, 190.9, 107.9, 75.1, 41.4, 36.2, 34.5, 29.0, 18.0.
IR (CCl4): 1672 (s), 1600 (vs) cm-1. MS (EI): m/e 152 (M, 100).
HRMS: calcd for C9H12O2 152.0837, found 152.0850. The
compound was assigned as cis on the basis of the long-range
coupling observed for the proton at δ 4.35, which is also present
in a structurally related compound, 3I-cis, whose structure was
confirmed by X-ray.
Th er m olysis of Ca r ben e Com p lex 1J . General Procedure
II was followed using carbene complex 1J (273 mg, 0.57 mmol).
Purification by flash chromatography on silica gel using
hexane/ethyl acetate (3:2) as the eluent provided two fractions.
The compound in the first fraction (49.3 mg, 40%) was
identified as the 3I-cis. The compound in the second fraction
(24.3 mg, 20%) was identified as the other isomer, 3I-trans.
The spectral data were identical to those previously listed for
these compounds, which was also obtained via thermolysis of
carbene complex 1I.
Th er m olysis of Ca r ben e Com p lex IK. General Procedure
II was followed using carbene complex IK (188 mg, 0.45 mmol).
Purification by flash chromatography on silica gel using 2%
methanol in dichloromethane as the eluent provided two
fractions.
The compound in the second fraction (32.9 mg, 33%) was
1
identified as the other isomer, 3F -trans. H NMR (CDCl3): δ
5.29 (d, 1 H, J ) 1.6 Hz), 4.17 (dd, 1 H, J ) 11.1, 5.2 Hz), 3.89
(t, 1 H, J ) 11.1 Hz), 3.05 (m, 1 H), 2.64 (dd, 1 H, J ) 17.8,
6.8 Hz), 2.20 (m, 1 H), 2.05 (dd, 1 H, J ) 17.8, 2.9 Hz), 1.65
(m, 2 H), 0.99 (d, 3 H, J ) 6.9 Hz). 13C NMR (CDCl3): δ 204.7,
190.3, 108.1, 66.2, 39.3, 37.1, 31.1, 25.7, 14.5. IR (CCl4): 1678
(s), 1600 (vs) cm-1. MS (EI): m/e 153 (M + 1, 17), 152 (M, 100).
HRMS: calcd for C9H12O2 152.0837, found 152.0847. This
isomer was assigned as trans on the basis of spectral similarity
to 3I-trans, most notably the absence of the long-range
coupling observed in 3F -cis and 3I-cis.
The compound in the first fraction (27.4 mg, 40%) was
1
identified as 3K-cis. H NMR (CDCl3): δ 5.25 (d, 1 H, J ) 1.4
Hz), 4.47 (m, 1 H), 3.01 (ddddd, 1 H, J ) 11.5, 8.4, 6.8, 2.8, 1.4
Hz), 2.61 (dd, 1 H, J ) 17.8, 6.8 Hz), 2.22 (dddd, 1 H, J )
13.3, 10.6, 8.4, 6.2 Hz), 2.03 (dd, 1 H, J ) 17.8, 2.8 Hz), 1.94
(dddd, 1 H, J ) 14.5, 10.6, 4.5, 3.3 Hz), 1.69 (dddd, 1 H J )
14.5, 10.6, 10.6, 6.2 Hz), 1.42 (dddd, 1 H, J ) 13.3, 11.5, 10.6,
4.5 Hz), 1.35 (d, 3 H, J ) 6.2 Hz); Irradiate δ 1.35: δ 4.47 (dd,
J ) 10.6, 3.3 Hz). 13C NMR (CDCl3): δ 204.3, 192.4, 105.3,
74.0, 42.0, 33.4, 28.9, 24.0, 20.7. IR (CCl4): 1667 (s), 1593 (vs)
cm-1. MS (EI): m/e 152 (M, 100). HRMS: calcd for C9H12O2
152.0837, found 152.0831. This isomer is believed to have a
cis relationship between the methyl group and the carbocyclic
ring as a result of the coupling constants within the hetero-
cyclic ring; the coupling constants for these hydrogens (δ 4.47,
3.01, 2.22, 1.94, 1.69, and 1.42) are less suggestive of a chair
conformation than those of the other isomer (see discussion
for other isomer). Because this isomer would have an axial
substituent in a perfect chair, this is the isomer which is most
likely to deviate from the chair conformation.
Th er m olysis of Ca r ben e Com p lex 1G. General Proce-
dure II was followed using carbene complex 1G (230 mg, 0.56
mmol). Purification by flash chromatography on silica gel using
hexane/ethyl acetate (3:2) as the eluent provided two fractions.
The compound in the first fraction (37.6 mg, 44%) was
identified as 3F -cis. The compound in the second fraction (18.1
mg, 22%) was identified as the other isomer, 3F -trans. The
ratio of 3F -cis to 3F -trans (67:33) was determined by integra-
tion of alkene protons at 5.3 ppm. The spectral data were
identical to those previously listed for these compounds, which
was also obtained via thermolysis of carbene complex 1F .
Th er m olysis of Ca r ben e Com p lex 1H. General Proce-
dure II was followed using carbene complex 1H (230 mg, 0.56
mmol). Purification by flash chromatography on silica gel using
hexane/ethyl acetate (3:2) as the eluent provided two fractions.
The compound in the first fraction (37.6 mg, 44%) was
identified as 3F -cis. The compound in the second fraction (18.
1 mg, 22%) was identified as the other isomer, 3F -trans. The
spectral data were identical to those previously listed for these
compounds, which was also obtained via thermolysis of car-
bene complex 1F .
Th er m olysis of Ca r ben e Com p lex 1I. General Procedure
II was followed using carbene complex 1I (284 mg, 0.70 mmol).
Purification by flash chromatography on silica gel using 2%
methanol in dichloromethane as the eluent provided two
fractions.
The compound in the first fraction (45.1 mg, 30%) was
identified as 3I-cis. 1H NMR (CDCl3): δ 7.45 (m, 5 H), 5.33 (d,
1 H, J ) 1.5 Hz), 4.45 (ddd, 1 H, J ) 11.2, 5.2, 1.3 Hz), 3.95
(dd, 1 H, J ) 11.2, 9.4 Hz), 3.24 (m, 1 H), 2.95 (m, 1 H), 2.61
(dd, 1 H, J ) 17.6, 6.6 Hz), 2.35 (ddd, 1 H J ) 12.7, 5.2, 5.2
Hz), 2.14 (dd, 1 H, J ) 17.6, 3.9 Hz), 1.65 (ddd, 1 H, J ) 12.7,
12.7, 11.0 Hz). 13C NMR (CDCl3): δ 204.1, 190.2, 140.4, 128.9,
127.5, 108.8, 74.5, 41.4, 40.9, 36.8, 33.4. IR (CCl4): 1698 (s),
1613 (s) cm-1. MS (EI): m/e 214 (M, 100). HRMS: calcd for
The compound in the second fraction (18.2 mg, 26%) was
identified as other isomer, 3K-trans. 1H NMR (CDCl3): δ 5.30
(d, 1 H, J ) 1.6 Hz), 4.12 (m, 1 H), 2.68 (ddddd, 1 H, J ) 12.8,
6.7, 5.0, 4.1, 1.6 Hz), 2.56 (dd, 1 H, J ) 17.4, 6.7 Hz), 2.11 (dd,
1 H, J ) 17.4, 4.1 Hz), 2.04 (dddd, 1 H, J ) 12.8, 5.0, 3.2, 3.2
Hz), 1.93 (dddd, 1 H, J ) 13.1, 3.2, 3.2, 3.2 Hz), 1.65 (dddd, 1
H, J ) 13.1, 12.8, 10.4, 3.2 Hz), 1.48 (dddd, 1 H, J ) 12.8,
12.8, 12.8, 3.2 Hz), 1.40 (d, 3 H, J ) 6.4 Hz). Irradiate at δ
4.12: δ 1.93 (ddd, J ) 13.1, 3.2, 3.2 Hz), 1.65 (dddd, 1 H, J )
12.8, 12.8, 10.4, 3.2 Hz), 1.40 (s). Irradiate at δ 1.40: δ 4.12
(dd, J ) 10.4, 3.2 Hz). 13C NMR (CDCl3): δ 204.7, 191.3, 109.2,
78.8, 41.2, 37.0, 32.1, 27.1, 21.8. IR (CCl4): 1661 (s), 1590 (vs)
cm-1. MS (EI): m/e 152 (M, 100). HRMS: calcd for C9H12O2
152.0837, found 152.0830. The compound was assigned as
having the trans relative configuration of the substituents on
the six-membered heterocyclic ring. The 1H NMR data are
suggestive of a chair conformation in which the methyl group
is equatorial. Every proton in a CH2 group of the ring appears
as a quartet of doublets if axial (one geminal, two axial-axial,
and one equatorial axial couplings) or a doublet of quartets
(one geminal and three gauche-type couplings). This assign-
ment was further supported by NOESY experiment, which
showed cross-peaks between the proton at δ 1.65 (axial proton
â to oxygen) and both δ 2.68 (bridgehead H) and δ 1.40 (the
methyl group), implying that the bridgehead hydrogen and the
methyl group are on the same face of the ring.
C
14H14O2 214.0994, found 214.1011. This compound is a solid
with a melting point of 122.5-123.5 °C. Recrystalization was
performed in a two-solvent system of dichloromethane and
methanol. The compound was assigned as having the cis
relative configuration of the substituents on the six-membered
heterocyclic ring on the basis of an X-ray structure determi-
nation.
Th er m olysis of Ca r ben e Com p lex 1L. General Procedure
II was followed using carbene complex 1L (300 mg, 0.700
mmol); two fractions were isolated after chromatographic
separation.
The compound in the second fraction (39.1 mg, 26%) was
identified as the other isomer, 3I-trans. H NMR (CDCl3): δ
7.25 (m, 5 H), 5.3 0 (d, 1 H, J ) 1. 5), 4.25 (t, 1 H, J ) 11.1
Hz), 4.23 (dd, 1 H, J ) 11.1, 6.2 Hz), 3.24 (m, 2 H), 2.63 (dd,
1 H, J ) 17.9, 6.8 Hz), 2.32 (ddd, 1 H, J ) 14.0, 8.5, 6.5 Hz),
The compound in the first fraction was an inseparable
mixture of three compounds (64 mg, 40%). The major compo-
nent (58% of the mixture) was assigned as 3L-cis-trans. 1H
NMR (CDCl3): δ 7.30-7.00 (m, 5 H), 5.34 (d, 1 H, J ) 1.4
Hz), 4.40 (m, 1 H), 3.20 (d, 1 H, J ) 3.2 Hz), 2.99 (dddd, 1 H,
1

1300 J . Org. Chem., Vol. 64, No. 4, 1999
Yan et al.
J ) 8.5, 8.5, 3.2, 1.4 Hz), 2.27 (dddd, 1 H, J ) 13.9, 10.4, 8.5,
6.0 Hz), 1.88 (dddd, 1 H, J ) 14.3, 10.4, 8.2, 4.3 Hz), 1.60-
1.70 (m, 2 H), 1.34 (d, 3 H, J ) 6.2). Irradiate at δ 1.34: δ
4.40 (dd, J ) 10.0, 3.6 Hz). The carbocyclic ring stereochem-
istry was assigned as trans on the basis of the smaller coupling
constant (3.2 Hz) between the benzylic and allylic protons. The
heterocyclic ring stereochemistry was assigned as cis on the
basis of chemical shift similarity to 3K-cis; the proton next to
oxygen features a chemical shift of δ 4.40, which is closer to
the value for the same proton in 3K-cis (δ 4.47) than that for
the same proton in the 3K-trans (δ 4.12). The second most
abundant component (34% of the mixture) was assigned as
substituents on the five-membered heterocycle was established
as cis due to the smaller coupling (relative to 3M-trans-trans,
see below) of 6.6 Hz between the allylic hydrogen and the
hydrogen on the same carbon as the methyl group.
The compound in the second fraction (32.3 mg, 26%) was
1
identified as 3M-trans-trans. H NMR (CDCl3): δ 7.32 (m, 5
H), 5.23 (d, 1 H, J ) 1.0 Hz), 4.67 (dd, 1 H, J ) 18.1, 8.8 Hz),
4.09 (dd, 1 H, J ) 18.1, 9.1 Hz), 3.40 (d, 1 H, J ) 4.9 Hz), 2.81
(ddd, 1 H, J ) 11.7, 4.9, 1.0 Hz), 2.30 (m, 1 H), 1.03 (d, 3 H, J
) 6.6 Hz); Irradiate δ 3.40: δ 2.81 (dd, 1 H, J ) 11.7, 1.0 Hz).
Irradiate δ 1.03: δ 2.30 (ddd, J ) 11.7, 9.1, 8.8 Hz). 13C NMR
(CDCl3): δ 205.6, 194.3, 138.0, 128.8, 128.3, 127.1, 99.5, 83.6,
59.2, 57.4, 38.8, 14.0. IR (CCl4): 1699 (m), 1625 (s) cm -1; MS
(EI): m/e 214 (M, 100). HRMS: calcd for C14H14O2 214.0994,
found 214.0976. The relative configuration of substituents on
the cyclopentenone ring was assigned as trans due to the small
coupling constant (4.9 Hz) between the hydrogen at the
benzylic and allylic positions. The relative configuration of
substituents on the five-membered heterocycle was established
as trans due to the extremely large coupling constant between
the allylic hydrogen and the hydrogen on the same carbon as
the methyl group (11.7 Hz), which is consistent only with
coupling of two pseudoaxial protons, as exists in the trans
heterocyclic ring.
1
3L-trans-trans. H NMR (CDCl3): δ 7.30-7.00 (m, 5 H), 5.38
(d, 1 H, J ) 1.2 Hz), 4.10 (m, 1 H), 3.27 (d, 1 H, J ) 4.7 Hz),
2.66 (m, 1 H), 2.09 (m, 1 H), 1.38 (d, 3 H, J ) 6. 1) (the
remaining protons cannot be reliably assigned because of
overlap with peaks for the other isomers in this fraction).
Irradiate at δ 1.38: δ 4. 10 (dd, J ) 10.7, 3.2 Hz). The
carbocyclic ring stereochemistry was assigned as trans on the
basis of the small coupling constant (4.7 Hz) between the
benzylic and allylic protons. The heterocyclic ring stereochem-
istry was assigned as trans on the basis of chemical shift
similarity to 3K-trans; the proton next to oxygen features a
chemical shift of δ 4.10, which is closer to the value for the
same proton in 3K-trans (δ 4.12) than that for the same proton
in the 3K-cis (δ 4. 47). The minor component (8% of the
mixture) was assigned as 3L-trans-cis. 1H NMR (CDCl3): δ
7.30-7.00 (m, 5 H), 5.56 (d, 1 H, J ) 1. 9), 4.10 (m, 1 H), 3.78
(d, 1 H, J ) 7.3 Hz), 2.78 (m, 1 H) 2.06 (m, 1 H), 1.35 (d, 3 H,
J ) 6.1 Hz), 0.97 (dddd, 1 H, J ) 13.7, 13.7, 13.7. 3.3) (the
remaining protons cannot be reliably assigned because of
overlap with peaks for the other isomers in this fraction).
Irradiate at δ 1.35: δ 4.10 (dd, J ) 10.7, 3.2 Hz). The
carbocyclic ring stereochemistry was assigned as cis on the
basis of the large coupling constant (7.3 Hz) between the
benzylic and allylic protons. The heterocyclic ring stereochem-
istry was assigned as trans on the basis of chemical shift
similarity to 3K-trans; the proton next to oxygen features a
chemical shift of δ 4.10, which is closer to the value for the
same proton in 3K-trans (δ 4.12) than that for the same proton
in the 3K-cis (δ 4. 47).
The compound in third fraction (19.9 mg, 16%) was identi-
1
fied as 3M-trans-cis. H NMR (CDCl3): δ 7.13 (m, 5 H), 5.35
(d, 1 H, J ) 1.9 Hz), 4.54 (dd, 1 H, J ) 8.8, 7.4 Hz), 4.04 (dd,
1 H, J ) 10.6, 8.8 Hz), 3.91(d, 1 H, J ) 7.5 Hz), 3.00 (ddd, 1
H, J ) 12.1, 7.5, 1.9 Hz), 1.75 (m, 1 H), 0.81 (d, 3 H, J ) 6.4
Hz). Irradiate at δ 3.91: δ 3.00 (dd, 1 H, J ) 12.1, 1.9 Hz). 13
C
NMR (CDCl3): δ 206.5, 196.6, 136.6, 128.6, 127.3, 101.1, 83.7,
55.5, 54.0, 33.6, 29.7, 13.8. IR (CCl4): 1699 (m), 1623 (s) cm-1
.
MS (EI): m/e 214 (M, 100). HRMS: calcd for Cl4H14O2
214.0994, found 214.1000. The relative configuration on the
cyclopentenone ring was assigned as cis on the basis of the
large coupling constant (7.5 Hz) between the hydrogen at the
benzylic and allylic positions. The relative configuration of
substituents on the five-membered heterocycle was established
as trans on the basis of the extremely large coupling constant
between the allylic hydrogen and the hydrogen on the same
carbon as the methyl group (12.1 Hz), which is consistent only
with coupling of two pseudoaxial protons, as exists in the trans
heterocyclic ring.
The compound in the second fraction was a identified as
compound 1L-cis-cis (34 mg, 21%). 1H NMR (CDCl3): δ 7.30-
6.95 (m, 5 H), 5.47 (d, 1 H, J ) 1.5 Hz), 4.45 (m, 1 H), 3.85 (d,
1 H, J ) 7.3 Hz), 3.33 (dddd, 1 H, J ) 11.9, 7.3, 7.3, 1.5 Hz),
1.83 (m, 1 H), 1. 35 (d, 3 H, J ) 6.1 Hz), 1.65-0.85 (m, 3 H).
Irradiate at δ 1.35: δ 4.45 (dd, J ) 9.3, 3.7 Hz). 13C NMR
(CDCl3): δ 205.1, 192.2, 137.5, 129.4, 128.5, 127.0, 106.4, 74.5,
Th er m olysis of Ca r ben e Com p lex IN in Tolu en e.
General Procedure II was followed using carbene complex 1N
(112 mg, 0.300 mmol); two fractions were isolated after
chromatographic purification.
The compound in the first fraction was identified as 3N-
55.3, 38.3, 29.0, 20.8, 19.5. IR (CDCl3): 1679 (s), 1598 (vs) cm-1
.
1
cis-trans (27 mg, 48%). H NMR (CDCl3): δ 7.38-7.10 (m, 5
MS (EI): 228 (M, 20), 86 (100). HRMS: calcd for Cl5H16O2
228.1158, found: 228.1150. The carbocyclic ring stereochem-
istry was assigned as cis on the basis of the large coupling
constant (7.3 Hz) between the benzylic and allylic protons. The
heterocyclic ring stereochemistry was assigned as cis on the
basis of chemical shift similarity to 3K-cis; the proton next to
oxygen features a chemical shift of δ 4.45, which is closer to
the value for the same proton in 3K-cis (δ 4.47) than the same
proton in the 3K-trans (δ 4. 12).
H), 5.26 (d, 1 H, J ) 1.1 Hz), 4.92 (m, 1 H), 3.45 (d, 1 H, J )
5. 1 Hz), 3.33 (m, 1 H), 2.46 (ddd, 1 H, J ) 11.7, 8.1, 4.0 Hz),
1.52 (d, 3 H, J ) 6.3 Hz) overlapping with 1.52 (m, 1 H).
Irradiate at δ 1.52: δ 4.65 (d, J ) 7.1 Hz). 13C NMR (CDCl3):
δ 205.6, 193.4, 138.0, 128.8, 127.2, 98.8, 88.5, 59.4, 52.3, 37.7,
20.2. IR (CH2Cl2): 1694 (s), 1623 (s) cm-1. MS (EI): 214 (M,
100). HRMS: calcd for C14H14O2 214.0994, found 214.0989. The
relative configuration of substituents on the cyclopentenone
ring was assigned as trans on the basis of the small coupling
constant (5.1 Hz) between the hydrogen at the benzylic and
allylic positions. The relative configuration of the substituents
on the heterocyclic ring was established as cis on the basis of
Th er m olysis of Ca r ben e Com p lex 1M. General Proce-
dure II was followed using carbene complex 1M (235 mg, 0.58
mmol); three fractions were obtained.
1
The first fraction was identified as a mixture (8:1) of
compound 3M-cis-trans and compound 3M-cis-cis. Further
purification was accomplished by preparative thin-layer chro-
matography (hexane/ethyl acetate, 3:2) to give 22.3 mg (18%)
similarity of the H NMR spectrum to that of 3N-cis-cis; most
notably the chemical shift of the proton on the same carbon
as oxygen (δ 4.92 in 3N-cis-trans) more closely resembles the
value for the same proton in 3N-cis-cis (δ 4.94) than that of
3N-trans-trans (δ 5.14).
1
of compound 3M-cis-trans. H NMR (CDCl3): δ 7.23 (m, 5 H),
5.26 (d, 1 H, J ) 1.7 Hz), 4.64 (dd, 1 H, J ) 9.1, 4.5 Hz), 4.35
(d, 1 H, J ) 9.1 Hz), 3.50 (d, 1 H, J ) 5.2 Hz), 3.38 (ddd, 1 H,
J ) 6.6, 5.2, 1.7 Hz), 2.58 (m, 1 H), 1.02 (d, 3 H, J ) 7.2 Hz).
Irradiate at δ 3.50: δ 3.38 (dd, 1 H, J ) 6.6, 1.7 Hz). 13C NMR
(CDCl3): δ 206.4, 192.8, 138.2, 128.8, 128.4, 127.2, 99.5, 85.1,
55.1, 54.4, 32.4, 13. 1. The relative configuration on the
cyclopentenone ring was assigned as trans on the basis of the
small coupling constant (5.2 Hz) between the hydrogen at the
benzylic and allylic positions. The relative configuration of
The compound in the second fraction was identified as 3N-
1
cis-cis (15 mg, 26%). H NMR (CDCl3): δ 7.28-6.95 (m, 5 H),
5.35 (d, 1 H, J ) 2.0 Hz), 4.94 (m, 1 H), 3.90 (d, 1 H, J ) 7.5
Hz), 3.57 (dddd, 1 H, J ) 12.8, 7.5, 7.5, 1.1 Hz), 1.89 (ddd, 1
H, J ) 12.8, 7.5, 4.2 Hz), 1.36 (d, 3 H, J ) 6.7 Hz), 1.00 (ddd,
1 H, J ) 12.8, 12.7, 10.4 Hz). Irradiate at δ 1.36: δ 4.94 (dd,
1 H, J ) 10.4, 4.2 Hz). The relative configuration on the
cyclopentenone ring was assigned as cis on the basis of the
large coupling constant (7.5 Hz) between the hydrogen at the

Five-Membered Rings Fused to Oxygen Heterocycles
J . Org. Chem., Vol. 64, No. 4, 1999 1301
benzylic and allylic positions. The relative configuration of
substituents in the heterocyclic ring was established as cis on
the basis of the agreement of all of the coupling constants
involving protons in the heteocyclic ring with predicted values
(see text). Most notably a large pseudoaxial-pseudoaxial
coupling exists for both the bridgehead proton at δ 3.57 (12.8
Hz) and the proton on the same carbon as oxygen at δ 4.94
(10.4 Hz).
Hz), 2.10 (m, 2 H), 1.96 (dd, 1 H, J ) 18.4, 2.4 Hz), (1.85, m,
2 H). 13C NMR (CDCl3): δ 203.8, 192.7, 144.5, 128.4, 126.4,
109.3, 71.7, 50.2, 44.3, 42.4, 38.6, 29.5. IR (neat): 1675, 1588
cm-1 ; Combustion Analysis: calcd for C15H16O2 C 78.91%, H
7.06%; found C 79.15%, H 6.85%. The heterocyclic ring was
assigned as trans on the basis of the coupling for the benzylic
proton (δ 2.68), which features two large pseudoaxial-pseudo-
axial couplings (11.2 Hz).
Th er m olysis of Ca r b en e Com p lex 1N in Dioxa n e.
General Procedure II was followed using carbene complex 1N
(112 mg, 0.300 mmol) and dioxane as the solvent. Final
purification was achieved using flash chromatography on silica
gel using 4:1 hexane/ethyl acetate as the eluent.
The product in the second fraction (13 mg, 10%) was
identified as 3O-cis. 1H NMR (CDCl3): δ 7.2 (m, 5 H), 5.40 (d,
1 H, J ) 1.0 Hz), 4.54 (br d, 1 H, J ) 12.6 Hz), 4.34 (t, 1 H, J
) 12.6 Hz), 3.50 (dddd, 1 H, J ) 7.3, 3.7, 3.3, 1.0 Hz), 3.29 (q,
1 H, J ) 3.7 Hz), 2.63 (dd, 1 H, J ) 18.2, 7.2 Hz), 2.08 (dd, 1
H, J ) 18.2, 3.3 Hz) overlapping with 1.75-2.22 (m, 4 H). 13C
NMR (CDCl3): δ 203.7, 192.5, 138.9, 129.4, 128.3, 127.1, 111.4,
72.3, 47.0, 43.1, 40.8, 35.8, 26.1. The heterocyclic ring was
assigned as cis on the basis of the coupling for the benzylic
proton (δ 3.29), which features only small pseudoequatorial-
pseudoequatorial or pseudoequatorial-pseudoaxial couplings
(3.7 Hz).
An additional compound over the previous experiment could
never be isolated in pure form but exhibits the following
1
spectral characteristics. H NMR (CDCl3): δ 7.40-7.10 (m, 5
H), 5.23 (d, 1 H, J ) 1.5 Hz), 5.14 (dq, 1 H, J ) 7.8, 6.7 Hz,
appears as a quintet), 3.39 (d, 1 H, J ) 5.5 Hz), 3.35 (m, 1 H),
2.05 (m, 1 H), 1.36 (d, 3 H, J ) 6.7 Hz), 1.15 (m, 1 H). Irradiate
at δ 1.36: δ 5.14 (d, J ) 7.8 Hz). The carbocyclic ring was
assigned as trans because the coupling of the vicinal protons
on the carbocyclic ring was 5.5 Hz, which was identical to that
for 3N-cis-trans from the previous experiment. The heterocyclic
ring was assigned as trans on the basis of the coupling pattern
for the proton next to oxygen (δ 5.14). Decoupling the methyl
group afforded a doublet with a coupling constant of 7.8 Hz.
Because this proton is vicinal to two protons, the coupling to
one of them is zero; only the trans heterocyclic ring isomer
features a 90° dihedral angle involving the proton at δ 5.14 in
the energy-minimized conformation.
Ack n ow led gm en t. We thank the Petroleum Re-
search Fund, administered by the American Chemical
Society, the NIH (GM-40777), and New Mexico State
University for financial support. We thank Mr. J ose
Morales for his thorough examination of the carbene
complex synthesis with secondary alcohols. We thank
Dr. Yui-Fai Lam for assistance with obtaining 2D-NMR
spectra and Dr. J ames Fettinger for obtaining the X-ray
structure for compound 3I-cis.
Th er m olysis of Ca r ben e Com p lex 1O. General Proce-
dure II was followed using carbene complex 1O (240 mg, 0.560
mmol); two fractions were isolated after chromatographic
purification.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra for all significant compounds in Table 1, X-ray
parameters for compound 3I-cis, and experimental procedures
for the synthesis of alkynols used to synthesize carbene
complexes 1A-O. This material is available free of charge via
the Internet at http://pubs.acs.org.
The product in the first fraction (76 mg, 58%) was identified
1
as 3O-trans. H NMR (CDCl3): δ 7.4-7.1 (m, 5 H); 5.44 (s, 1
H), 4.48 (dddd, 1 H, J ) 12.4, 5.2, 2.0, 2.0 Hz), 4.33 (dd, 1 H,
J ) 12.4, 10.4 Hz), 3.44 (ddd, 1 H, J ) 11.2, 6.8, 2.4 Hz), 2.68
(ddd, 1 H, J ) 11.2, 11.2, 3.2 Hz), 2.38 (dd, 1 H, J ) 18.4, 6.8
J O982144Q