Substituent effects on intramolecular epoxide cyclizations that can competitively occur at aromatic or double bond positions

Article abstract of DOI:10.1021/jo951945f

The effects of substituents on epoxides that can competitively cyclize at either a double bond or aromatic position were determined. Adding a group that would stabilize the transition state of the double bond cyclization of an epoxide that underwent predominantly aromatic cyclization increased the relative amount of the former pathway, but not dramatically. Adding a strong activating substituent to the aromatic group of an epoxide that underwent predominantly double bond cyclization increased the relative amount of aromatic cyclization, but a complex product mixture was obtained. An explanation of the behavior observed is presented.

Full text of DOI:10.1021/jo951945f

J . Org. Chem. 1996, 61, 2075-2080
2075
Su bstitu en t Effects On In tr a m olecu la r Ep oxid e Cycliza tion s Th a t
Ca n Com p etitively Occu r a t Ar om a tic or Dou ble Bon d P osition s1
Stephen K. Taylor,* Scott A. May, and Eric S. Stansby
Department of Chemistry, Hope College, Holland, Michigan 49422-9000
Received November 1, 1995X
The effects of substituents on epoxides that can competitively cyclize at either a double bond or
aromatic position were determined. Adding a group that would stabilize the transition state of
the double bond cyclization of an epoxide that underwent predominantly aromatic cyclization
increased the relative amount of the former pathway, but not dramatically. Adding a strong
activating substituent to the aromatic group of an epoxide that underwent predominantly double
bond cyclization increased the relative amount of aromatic cyclization, but a complex product mixture
was obtained. An explanation of the behavior observed is presented.
Ta ble 1. Cycliza tion Su bstr a tes a n d Ma jor P r od u cts
In tr od u ction
Biomimetic cyclizations of epoxides, especially epoxy-
ene cyclizations, have been investigated for some time.2
Although epoxide cyclizations to aromatic positions (epoxy-
arene cyclizations) is a newer area of study,2-4 these
reactions have already been used to make important
natural and medicinal products in high yields.2,5,6
In previous studies comparing these two related areas
of chemistry,3 we determined the relative facility of
epoxide cyclizations to intramolecular aromatic and
double bond positions (Table 1). When an exo versus
endo cyclization7 can competitively occur, be it aromatic
versus double bond or vice versa,3 the exo process
predominates. When both aromatic and double bond
cyclization pathways are endo, the aromatic cyclization
process occurrs (1 f 2). On the other hand, when both
ring-forming pathways could be exo, the double bond
cyclization process is favored (7 f 8).
In this paper, we report studies on compounds which
have activating substituents added to the previous
compounds to see if the ring formation preferences could
be reversed. For example, if a methoxy group was added
to the aromatic ring of 3, would this lead to epoxy-arene
cyclization rather than the previously observed epoxy-
ene cyclization? Alternatively, would a methyl group
attached to the internal carbon of the double bond of 1
promote double bond cyclization since the resulting
reactive species would resemble a tertiary cation?
X Abstract published in Advance ACS Abstracts, March 1, 1996.
(1) A preliminary account of some of this work has been published:
Taylor, S. K.; May, S. A.; Hopkins, J . A. Tetrahedron Lett. 1993, 34,
1283.
(2) For a recent review see (a) Taylor, S. K. Org. Prep. Proc. Int.
1992, 24, 245. For earlier reviews see (b) van Tamelen, E. E. Acc. Chem.
Res. 1968, 1, 111. (c) Sutherland, J . K. Chem. Soc. Rev. 1980, 9, 265.
(d) J ohnson, W. S. Bioorg. Chem. 1976, 5, 51. (e) Goldsmith, D.
Fortschr. Chem. Org. Naturst. 1971, 29, 363.
(3) Taylor, S. K.; Bischoff, D. S.; Blankespoor, C. L.; Deck, P. A.;
Harvey, S. M.; J ohnson, P. L.; Marolweski, A. E.; Mork, S. W.; Motry,
D. G.; Van Eenenaam, R. J . Org. Chem. 1990, 55, 4202. See correc-
tion: Ibid. 1991, 56, 5736.
(4) (a) Taylor, S. K.; Hockerman, G. H.; Karrick, G. L.; Lyle, S. B.;
Schramm, S. B. J . Org. Chem. 1983, 48, 2449. (b) Taylor, S. K.;
Davisson, M. E.; Hissom, B. R., J r.; Brown, S. L.; Pristach, H. A.;
Schramm, S. B.; Harvey, S. M. Ibid. 1987, 52, 425. (c) Taylor, S. K.;
Bankespoor, C. L.; Harvey, S. M.; Richardson, L. J . J . Org. Chem. 1988,
53, 3309.
Resu lts
Compounds 1 and 3 (Table 1) underwent exclusively
epoxy-arene cyclizations. Compounds 9 (eq 1) and 13 (eq
2) differ from those compounds by having a methyl group
on the olefin. These changes should favor epoxy-ene
cyclization since the resulting double bond cyclization
intermediate should resemble a tertiary cation. How-
ever, when 9 was treated with BF3‚OEt2 in dichlo-
(5) (a) Broka, C. A.; Chan, S.; Peterson, B. Ibid. 1988, 53, 1586. (b)
Burnell, R. H.; Dufour, J . M. Can. J . Chem. 1987, 65, 21.
(6) (a) Tanis, S. P.; Herrington, P. M. J . Org. Chem. 1983, 48, 4572.
(b) Tanis, S. P.; Raggon, J . W. Ibid. 1987, 52, 819.
(7) Baldwin, J . E. J . Chem. Soc., Chem. Commun. 1976, 734. For
epoxide cyclizations, the exo cyclization results in ring formation with
the resulting OH group outside the ring formed. Endo cyclizations
result in the OH in the ring formed.
0022-3263/96/1961-2075$12.00/0 © 1996 American Chemical Society

2076 J . Org. Chem., Vol. 61, No. 6, 1996
Taylor et al.
Sch em e 1
romethane, an epoxy-arene product still predominated:
10 and 11 were isolated in 53% and 17% yields, respec-
tively.8 In this and the other cyclizations, the reactions
were conducted under very dilute conditions to minimize
intermolecular reactions.
Epoxide 21 (eq 4) cyclizes to a complex mixture of
products which were exceptionally difficult to separate
and identify. HPLC and independent synthesis of some
compounds were necessary to identify the products.
Equation 4 shows the major products in their relative
abundances. These compounds account for 80% of the
volatile product composition with only one of the other
unidentified compounds consisting of >2% of the mixture.
Although a single step could be envisioned for the two
rings formed to give 10 (Scheme 1, pathway B), we
showed they were formed in a stepwise fashion from an
alcohol precursor (12) that we prepared independently
(Scheme 1, pathway A). This compound was subjected
to the reaction conditions, and it quickly formed 10 in
>90% yield. Also, when the reaction was followed by
NMR, the aromatic region changed rapidly, well before
any changes occurred in the vinyl region. When the
reaction was done using shorter reaction times or the
weaker Lewis acid SnCl4, we detected 12 in significant
quantities by GCMS. These observations support the
stepwise pathway A.
Compounds 14 and 15 resulted in 28% and 21%
isolated yields (eq 2) when 13 was treated with BF3‚OEt
similarly.9 Compound 14 can be formed in a pathway
similar to that shown in Scheme 1 pathway A. Here, the
double bond cyclization product 15 is formed in signifi-
cant but not predominant quantity.
13
The epoxides used in this study were synthesized by
the Wittig and epoxidation procedures developed earlier,3
as shown in eqs 5 and 6. The aryl dienes used to make
epoxides 16 and 21 gave >90% internal epoxidation
products. The small amounts of side products were
removed by HPLC as described below. The aryl diene
precursors to epoxides 9 and 13 gave nearly equal
amounts of terminal and internal epoxidation products.
The two epoxides resulting from each epoxidation pro-
cedure were separated by HPLC. The resulting terminal
epoxides were briefly tested for cyclization reactions, but
no interesting results were observed.
Compound 3 (Table 1) underwent exclusively double
bond cyclization.3 Compound 16 is similar to 3, but it
has a CH3O activating group added to it. This compound
cyclizes to a complex mixture of products, with the major
four shown in their relative quantities (eq 3). The
reaction was followed over time, and the products shown
are those which result immediately after the epoxide is
consumed.
The activating substituent makes a difference, but not
as much as one might expect: the double bond and
aromatic cyclization pathways compete almost equally
well.
(8) Products 10 and 11 account for 81% of the volatile product
distribution of eq 2. Products analogous to those represented by the
structure below account for another 9% as shown by GCMS and an
NMR of the mixture of isomers. We were unable to separate these
isomers, but they would be expected to accompany 10 and 12.
Discu ssion
In the previous study,3 epoxides 1 and 3 (Table 1) gave
exclusively aromatic cyclization and yielded one product
each. Although 3 and 7 gave almost exclusively epoxy-
ene cyclization, the product mixture was complex. This
contrast between cyclization pathways was attributed to
the reactive species involved in the two different reac-
tions. The epoxy-arene pathway would go through an
arenium ion whose primary subsequent step would be
the loss of a proton to rearomatize the ring (see Scheme
(9) The yields of these products were undoubtedly higher, but these
are the yields of pure, isolated compounds. Impure chromatographic
fractions containing these compounds were not counted in the yields.

Substituent Effects on Intramolecular Epoxide Cyclizations
J . Org. Chem., Vol. 61, No. 6, 1996 2077
a 3 × 6 cm column of alumina with 200 mL of 9:1 hexane:
ether. The concentrated organic layer was carefully distilled
giving 1.33 g (41%, >97% GC purity)) of 28 (a later fraction
had 10% trans isomer present), bp 74-76 °C (0.1 mm); 1H
NMR (CDCl3) δ 1.7 (s, 3 H), 1.9-2.1 (m, 4 H), 2.3 (m, 2 H), 2.6
(t, 3 H), 4.7 (s, 1 H), 4.74 (s, 1 H), 5.4 (m, 2 H), 7.2-7.4 (m, 5
H): 13C NMR (CDCl3) δ 23.0, 26.1, 29.9, 36.6, 38.2, 110.6,
126.3, 126.4, 128.9, 129.1, 130.6, 142.7, 146.1; IR (NaCl disks)
1649 (m), 698 (s) cm-1; mass spectrum m/ e (relative intensity)
200 (5), 104 (19), 91 (100), 67 (24), 65 (21). Anal. Calcd for
C15H20: C 89.91; H, 10.29. Found: C, 89.83; H, 10.11.
2-Meth yl-9-p h en yl-cis-1,5-n on a d ien e (29) was made as
above by combining the triphenylphosphonium salt3 of 1-bromo-
4-phenylbutane (2.13 g, 4.5 mmol) with 2.2 mL of 2.2 M n-BuLi
and 0.55 g of 4-methyl-4-pentenal11 (5.6 mmol) to give 29 (cis:
trans 9:1) in 45% yield, bp 78-79 °C (0.05 mm): 1H NMR
(CDCl3) δ 1.8 (s, 3 H), 1.7-2.0 (m, 2 H), 2.1-2.5 (m, 6 H), 2.7
(t, 2 H), 4.7 (s, 1 H), 4.8 (s, 1 H), 5.5 (m, 2 H), 7.2-7.5 (m, 5
H); 13C NMR (CDCl3) δ 23.2, 26.3, 27.6, 32.2, 36.2, 38.5, 110.8,
126.4, 129.0, 129.2, 130.4, 131.0, 143.2, 146.2; IR (NaCl disks)
1650 (m), 760 and 700 (s) cm-1. Anal. Calcd for C16H22: C,
89.66; H, 10.34. Found: C, 89.30; H, 10.41.
9-(m -Meth oxyp h en yl)-cis-1,6-n on a d ien e (30) was made
by adding 4.4 mL of 2.5 M (11 mmol) n-BuLi dropwise to 8.2
mL of DMSO. After 0.5 h, a mixture of 17 mL of DMSO and
5.0 g (10 mmol) of the triphenylphosphonium bromide salt (mp
127.5-128.5 °C) of 3-(m-methoxyphenyl)-1-bromopropane12
was added over 35 min. The resulting mixture was stirred 4
h and worked up by adding the mixture to 100 mL of cold
water and extracting the mixture three times with 50 mL
portions of 4:1 hexane:ether. The combined organic layers
were washed with water and run through alumina with 4:1
hexane:ether. The concentrated organic product was distilled,
1.41 g (60%), bp 90-92 °C (0.025 mm); NMR (CDCl3) δ 1.5 (p,
J ) 7.3 Hz, 2 H), 2.1 (m, 4 H), 2.5 (m, 2 H), 2.7 (t, J ) 7.3 Hz,
2 H); 3.9 (s, 3 H), 5.1 (m, 2 H), 5.5 (m, 2 H), 5.9 (m, 1 H), 6.9
(m, 3 H), 7.3 (t, J ) 7 Hz, 1 H); 13C NMR (CDCl3) δ 27.4, 29.6,
29.9, 34.1, 36.8, 55.7, 111.8, 115.1, 115.2, 121.7, 129.8, 129.9,
131.0, 139.5, 144.4, 160.4; IR (NaCl disks) 1250 (s), 720 and
770 (s, meta) cm-1; mass spectrum, m/ e (relative intensity)
230 (10), 122 (69), 121 (100), 91 (38). Anal. Calcd for
C16H22O: C, 83.43; H, 9.63. Found: C, 82.52; H, 9.38. The
13C and 1H NMR showed the compound was very pure.
10-(m -Meth oxyp h en yl)-cis-1,6-d eca d ien e (31) was made
from the triphenylphosphonium salt (mp 130-132 °C)3 of 4-(m-
methoxyphenyl)-1-bromobutane12 and 5-hexenal by the Wittig
procedure above in 56% distilled yield (cis:trans 9:1), bp 95-
100 °C (0.01 mm); NMR (CDCl3) δ 1.46 (p, J ) 7 Hz, 2 H), 1.7
(m, 2 H), 2.1 (m, 6 H), 2.6, (m, 2 H), 3.8 (s, CH3O), 5.0 (m, 2
H), 5.4 (m, 2 H), 5.8 (m, 1 H), 6.8 (m, 3 H), 7.2 (m, 1 H); 13C
NMR (CDCl3) δ 27.4, 27.5, 29.6, 32.0, 34.0, 36.2, 55.8, 111.6,
114.9, 115.1, 121.6, 129.9, 130.3, 130.7, 139.5, 144.9; IR (NaCl
disks) 1260 (s, CH3O-Ar), 910 (s), 780 and 695 (s); mass
spectrum m/ e (relative intensity) 244 (4) 161 (23), 122 (8), 121
(21), 44 (22), 32 (100). Anal. Calcd for C17H24O: C, 83.55; H,
9.90. Found: C, 83.10; H, 9.82.
1, pathway A). The double bond cyclization would involve
a carbocation-like species which could undergo proton
migrations and eliminations almost isoenergetically. The
result would be a complex product mixture, the behavior
observed. In this work, our results were consistent with
this conclusion. The cleanest reaction mixture was the
one shown in eq 1, which gave the highest percentage of
epoxy-arene ring formation (where we could account for
90% of the product mixture8).
Although the products obtained in this work are
interesting and pertinent to biomimetic epoxide cycliza-
tions,2,3 it is unfortunate that the product compositions
are not simpler. The methoxy group is known to activate
the aromatic ring by three orders of magnitude10 in
certain cases, so it might be thought that the use of this
group would have completely switched the epoxy-ene
cyclizations to epoxy-arene ones. This probably did not
occur because the rate-determining step involved in these
reactions is not highly arenium-ion like.
It is noteworthy that in the cyclizations of 9 and 13,
the epoxide acts as the source of initiation and termina-
tion steps. This is somewhat rare2a-e and it may also be
a reason for fewer products in these reactions: the oxygen
termination step probably prevents further rearrange-
ments.
Exp er im en ta l Section
The equipment used has been described elsewhere.3 The
solvents used were dried immediately before use by distilling
them from a drying agent under N2. THF and ether were
distilled from sodium/benzophenone, and CH2Cl2 was distilled
from CaH2. Boiling points and melting points are uncorrected.
Semipreparative HPLC was done using a 25 cm × 10 mm silica
gel column. 13C NMR spectra were supplied for referee review
to prove product purities (g96% purity unless specified
otherwise), and are referenced to the center peak of CDCl3 set
at 77.6 ppm. Elemental analyses were determined by Gal-
braith Labs, Inc. or Dornis and Kolbe. BF3‚OEt2 was used
from a bottle sealed under N2.
2-Meth yl-8-p h en yl-cis-1,5-octa d ien e (28). Under dry N2,
7.3 mL of 2.4 M n-BuLi in hexane (17 mmol) was added
dropwise to 14 mL of dry DMSO. After stirring 0.5 h, 7.9 g of
the triphenylphosphonium bromide salt of 1-bromo-3-phenyl-
propane (17 mmol, made exactly as before3) in 45 mL of dry
DMSO was added over 1.5 h. The resulting red-orange
solution was stirred 0.5 h, and then 2.10 g (21 mmol) of
4-methyl-4-pentenal11 in 5 mL of dry DMSO was added
dropwise. After stirring 4 h, the reaction mixture was added
to 85 mL of ice-water, and the resulting mixture was
extracted with 85 mL of 9:1 hexane:ether. The organic layer
was dried (MgSO4), concentrated to 50 mL, and run through
Ep oxid a tion P r oced u r e. The epoxides were prepared by
the following procedure.
cis-3,4-Ep oxy-7-m eth yl-1-p h en yl-7-octen e (9). A solu-
tion of 1.72 g of 85% pure mCPBA (8.5 mol) in 40 mL of CH2-
Cl2 was added over 1 h to 1.33 (8.6 mmol) of 28 in 21 mL of
CH2Cl2. The mixture was refluxed for 1.5 h and stirred for 6
h. The mixture was cooled and 25 mL of petroleum ether was
added. After approximately 1 h, a white precipitate formed,
and the mixture was suction filtered. The filtrate was washed
twice with 20% NaHSO3 and twice with 5% NaHCO3. After
drying (MgSO4) and rotary evaporation of the solvent the
product was short-path distilled, 83-86 °C (0.015 mm).
Spinning band distillation was used in an unsucccessful
attempt to separate the internal and terminal (∼1:1) epoxides.
HPLC (10 mm × 20 cm 10 µm silica gel column) with 90:10
trace hexane:CH2Cl2:EtOH gave >99% pure 9 (71 mg): 1H
NMR (CDCl3) δ 1.6 (m, 4 H), 1.7 (s, 3 H), 1.9 (m, 2 H), 2.2 (m,
(10) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in
Organic Chemistry, 3rd ed.; Harper and Row: New York, 1987; pp
541-545.
(12) Trost, B. M.; Crimmin, M. J .; Butler, D. J . Org. Chem. 1978,
43, 4449.
(11) Vig, O. P.; Vig, B.; Anand, R. J . Ind. J . Chem. 1979, 7, 1111.

2078 J . Org. Chem., Vol. 61, No. 6, 1996
Taylor et al.
2 H), 2.7-3.4 (2 overlapping multiplets, 4 H), 4.70 (s, 1 H),
4.74 (s, 1 H), 7.2-7.4 (m, 5 H); 13C NMR (CDCl3) δ 23.2, 26.7,
30.6, 33.6, 35.2, 57.4, 57.8, 111.2, 126.8, 129.2, 142.1, 145.5;
IR (NaCl disks) 1650 (m), 890 (m), 750 (m), 699 (m); mass
spectrum, m/ e (relative intensity): 216 (0.4), 118 (14), 117 (35),
112 (17), 104 (19), 92 (23), 91 (100), 67 (36); exact mass m/ e
calcd for C15H200 216.1514, found 216.1528. 1H and 13C NMR
and GC showed the compound was 98% pure.
cis-4,5-Ep oxy-8-m eth yl-1-p h en yl-8-n on en e (13) was pre-
pared as described above using 391 mg of 29 (1.7 mmol) in 10
mL of CH2Cl2 and 510 mg of 65% mCPBA (1.9 mmol) in 5 mL
of CH2Cl2. Reflux was conducted for 3 h, and the solution was
stirred at room temperature for 6 h. Workup was as above.
The crude product (0.4 g) was run through a plug of silica gel
with 60:40 hexane:CH2Cl2 (0.266 g of product) and then
purified by semipreparative HPLC (89:11 trace hexane:CH2-
Cl2:EtOH) to give 99% pure 13. 1H NMR (CDCl3) δ 1.5-2.0
(m, 6 H), 1.7 (s, 3 H), 2.0-2.1 (m, 2 H), 2.6-2.8 (m, 2 H), 2.8-
3.0 (m, 2 H), 4.71 (s, 1 H), 4.74 (s, 1 H), 7.2-7.4 (m, 5 H); 13C
NMR (CDCl3) δ 23.0, 26.8, 28.0, 29.0, 35.2, 36.3, 57.3, 57.6,
111.1, 126.5, 129.0, 142.7, 145.4; IR (NaCl disks) 1649 (m),
891 (s), 749 (s), 699 (s); mass spectrum, m/ e (relative intensity)
230 (6), 131 (50), 104 (100), 91 (100); exact mass m/ e calcd for
C16H22O 230.1670, found 230.1700.
1-(m -m eth oxyp h en yl)-cis-3,4-ep oxy-8-n on en e (16) was
prepared by dropwise addition of a solution of 1.09 g of 55%
pure mCPBA (4.05 mmol) and 23 mL of CH2Cl2 to 0.9 g (3.9
mmol) of 30 in 13 mL of CH2Cl2. The mixture was refluxed
1.5 h and stirred 4.5 h at room temperature. To the ice-cooled
mixture was added 13 mL of petroleum ether. The resulting
mixture was filtered, and the solid was washed with hexane.
The combined organic wash and filtrate was extracted with
40 mL of 20% NaHSO3, 5% NaHCO3 (three times), and brine
and dried (MgSO4). The product left after solvent removal was
column chromatographed (80:20:trace hexane:EtOAc:EtOH),
giving 0.60 g (62%) of product. This compound was purified
by HPLC (85:15 trace hexanes:EtOAc:EtOH) for the analyti-
cally pure compound used in cyclization studies: 1H NMR
(CDCl3) δ 1.4-1.7 (m, 4 H), 1.8 (q, J ) 7 Hz, 2 H), 2.1 (m, 2
H), 2.7-3.0 (2 overlapping multiplets, 4 H), 3.8 (s, 3 H), 5.0
(m, 2 H), 5.7 (m, 1 H), 6.8 (m, 3 H), 7.2 (m, 1 H); 13C NMR
(CDCl3) δ 26.4, 27.9, 30.4, 33.6, 34.2, 55.8, 57.2, 57.9, 111.9,
114.9, 115.6, 121.5, 130.1, 138.9, 143.7, 160.4; IR (NaCl disks)
1250 (s), 900 (s), 760 (s), 690 (s) cm-1; mass spectrum m/ e
(relative intensity) 246 (15), 134 (88), 121 (97), 91 (85), 78 (49),
65 (41), 55 (69), 43 (14), 41 (100). Anal. Calcd for C16H22O2:
C, 78.01; H, 9.00. Found: C, 77.93; H, 8.69.
1-(m -Meth oxyp h en yl)-cis-4,5-ep oxy-9-d ecen e (21) was
prepared as above using 1.05 g of 85% pure mCPBA (4 mmol)
in 22 mL of CH2Cl2 and 0.92 g (3.8 mmol) of 31 in 12 mL of
CH2Cl2. Workup was conducted as described above. Flash
chromatography (85:14:1 hexane:CH2Cl2:EtOH) gave fractions
that were 92-95% pure (0.15 g, 66%). One of the fractions
was chromatographed (HPLC) (70:20:1 hexane:CH2Cl2:EtOH)
to afford 97% (by GC) pure compound: 1H NMR (CDCl3) δ 1.5-
1.7 (m, 6 H), 1.7-2.0 (m, 2 H), 2.0-2.2 (m, 2H), 2.6-2.8 (m, 2
H), 2.9-3.0 (m, 2 H), 3.8 (s, 3 H), 5.0 (m, 2 H), 5.8 (m, 1 H),
6.7-6.8 (m, 3 H), 7.2 (m, 1 H); 13C NMR (CDCl3) δ 26.5, 28.0,
28.1, 29.0, 34.2, 36.4, 55.8, 57.6 (2 C), 111.8, 114.9, 115.6, 121.5,
130.0, 139.0, 144.3, 160.3; IR (NaCl disks) 1250 (s), 910 (s),
760 and 690 (s); mass spectrum m/ e (relative intensity) 260
(14), 161 (60), 134 (100), 122 (38). Anal. Calcd for C17H24O2:
C, 78.42; H, 9.29. Found: C, 78.38; H, 9.46.
verified by GCMS comparison with the authentic sample. The
products were identified as described below.
Cycliza tion s of 16 a n d 21. A solution of 0.10 g of 16 in
0.5 mL of CH2Cl2 was added dropwise to a solution of 6 µL of
BF3‚OEt2 in 18 mL of CH2Cl2. The reaction progress was
followed by GC by taking aliquots (10 µL) of the reaction
solution at 5, 15, and 25 min. At 1 h, 1.6% epoxide was left
whereas 30% of 16 was present at 5 min. The reaction was
stopped after 1 h, and the mixture was washed three times
with 25 mL of 5% NaHCO3 and once with brine and dried
(MgSO4). Rotary evaporation left 96 mg of the product
distribution shown in eq 3. Products were separated by HPLC
(85:15 trace hexane:EtOAc:EtOH).
Epoxide 21 was treated similarly by adding 0.12 g (49 mmol)
of 31 in 0.5 mL of CH2
Cl2 to a solution of 7 µL of BF3‚OEt3 in
21.5 mL of CH2Cl2. After 0.5 h, the reaction mixture was
worked up as above giving 0.11 g of crude product (eq 4).
HPLC using 60:40 trace hexane:CH2Cl2:EtOH gave the prod-
ucts indicated.
3,3-Dim e t h yl-2,3,4a ,5,6,10b -h e xa h yd r o-1H -b e n zo[f]-
ch r om en e (10) had a 22 mL retention volume on HPLC (98%
GC purity): 1H NMR (CDCl3) δ 1.1 (s, 3 H), 1.3 (s, 3 H), 1.3 (2
H under singlet), 1.8 (m, 1 H), 2.0-2.2 (m, 2 H), 2.3 (m, 1 H),
2.6 (m, 1 H), 2.8 (d, J ) 2.5 Hz, 1 H), 3.1 (q of d, J ) 16.1, 12.7
and 5.4 Hz), 4.1 (t, J ) 5 Hz, 1 H), 7.1-7.3 (m, 4 H). The
spectrum is very complex, and we can provide a complex
analysis of all the spin systems and 2D NMR upon request.
13C NMR (CDCl3) δ 23.3, 24.9, 25.6, 29.4, 31.8, 32.3, 36.8, 67.4,
71.7, 126.0, 126.4, 127.3, (2 C), 129.4, 129.5; IR (NaCl disks)
no CdO or OH bands, 1229 (m), 737 (ortho) cm-1; mass
spectrum m/ e (relative intensity) 216 (39), 129 (100), 115 (56),
43 (55) exact mass m/ e calcd for C15H20O 216.1509, found
216.1514.
1-M e t h y l-3-(2-p h e n y le t h y l)-7-o x a b i c y c lo [2.2.1]-
h ep t a n e (11) had a 35 mL retention volume on the above
HPLC: 1H NMR (CDCl3) δ 1.1-1.9 (m, 9 H), 1.5 (s, 3 H), 2.6
(t, J ) 7 Hz, 2 H), 4.2 (d, J ) 5 Hz, 1 H), 7.2-7.4 (m, 5 H); 13
C
NMR (CDCl3) δ 21.6, 32.2, 34.5, 36.3, 38.0, 44.9, 45.3, 81.3,
84.4, 126.3 (2 C) 128.0, 129.1 (2 C), 143.1; IR (NaCl disks) 1207
(m), 748 and 699 (monosubst benzene) mass spectrum m/ e
(relative intensity) 216 (2), 198 (43), 104 (211), 91 (100), 43
(94), exact mass m/ e calcd for C15H20O 216.1509, found
216.1481.
r-(3-Met h yl-3-b u t en -1-yl)-1,2,3,4-t et r a h yd r o-2-n a p h -
th ol (12). A solution of 1.61 g (11 mmol) of â-tetralone in 3
mL of THF was added dropwise to a suspension of 0.48 g (12
mmol) of 60% NaH (12 mmol, in mineral oil) at -5 to -10 °C.
The suspension was stirred 15 min, and 1.34 g (9 mmol) of
4-bromo-3-methyl-1-butene in 3 mL of THF was added drop-
wise to the mixture at -10 °C. After 45 min at -8 °C, the
mixture was stirred 1 h at room temperature and refluxed 4.5
h. The mixture was diluted with 25 mL of ether, and 20 mL
of saturated NH4Cl was added. The organic layer was washed
with brine, dried (MgSO4), and rotary evaporated, leaving 1.8
g of crude product.
One gram of the product was distilled, bp 92-95 °C (0.01
mm), giving 0.17 g of 78% pure product. This compound was
purified further by rotary thin-layer chromatography (80:20
trace hexane:CH2Cl2:EtOH) giving 63 mg of 96% pure com-
pound 1-(3-methyl-3-buten-1-yl)-3,4-dihydro-1H-naphthalen-
2-one (32): 1H NMR (CDCl3) δ 1.7 (s, 3 H), 1.9-2.1 (m 4 H),
2.4-2.7 (m, 2 H), 2.9-3.1 (m, 1 H), 3.1-3.2 (m, 1 H), 3.35-
3.45 (m, 1 H), 4.68 (s, 1 H), 4.73 (s, 1 H), 7.1-7.3 (m, 4 H); 13
C
Cycliza tion s of 9 a n d 13. In a typical reaction, 71 mg
(0.33 mmol) of 9 in 2 mL of dry methylene chloride was added
dropwise under N2 to 9 µL (0.073 mmol) of BF3‚OEt2 in 15
mL of dry methylene chloride. The solution was stirred at
room temperature for 3.5 h, and then it was washed with 5%
NaHCO3 and saturated NaCl and dried (MgSO4). The prod-
ucts were purified directly by semipreparative HPLC using a
92:8 mixture of hexane:ethyl acetate (trace of EtOH). The
yield of the reaction was determined by GC using 12 (as a
standard) that had been prepared independently. Compound
13 was treated similarly only the reaction time was 50 min,
and yields were of isolated, pure compounds. Subsequent
reactions were followed by GC, and the existence of 12 was
NMR (CDCl3) δ 23.1, 28.6, 30.4, 35.5, 38.2, 53.9, 111.1, 127.4,
127.5, 128.6, 128.7, 137.2, 138.0, 145.6, 212.9; mass spectrum
m/ e (relative intensity) 214 (30), 158 (23), 146 (100); exact
mass calcd for C15H18O 214.1353, found 214.1353. This
compound was reduced with LiAlH4 to give 12: 32 (90 mg,
4.2 mmol) in 5 mL of ether was added to a slurry of 2.10 mg
of LiAlH4 (excess) and 15 mL of ether. Workup includes
adding 0.4 mL of 15% NaOH and then 1 mL of H2O to form a
solid. The solid was vacuum filtered, and the filtrate was
column chromatographed (hexanes:8-20% EtOAc:trace EtOH)
gave a fraction that was predominantly one stereoisomer
(12): 1H NMR (CDCl3) δ 1.5 (broad s, 1 H), 1.8 (s, 3 H), 2.0
(m, 4 H), 2.2 (m, 2 H), 2.8 (m, 2 H), 3.0 (m, 1 H), 4.2 (m, 1 H),

Substituent Effects on Intramolecular Epoxide Cyclizations
J . Org. Chem., Vol. 61, No. 6, 1996 2079
4.748 (s, 1 H), 4.757 (s, 1 H), 7.1-7.2 (m, 5 H); 13C NMR
(CDCl3) δ 23.1, 27.1, 28.5, 36.5, 44.2, 69.4, 110.8, 126.3, 126.8,
129.3 (2 C), 129.4, 136.2, 146.6; IR (NaCl disks) 3600-3200,
1047 (s), 886 (s), 752 (s, ortho); mass spectrum m/ e (relative
intensity) 216 (14), 160 (56), 142 (100), 129 (70), 117 (60), 115
(67); exact mass calcd for C15H20O, 216.1509, found 216.1514.
Alcohol (13.1 mg) 12 was added slowly to a solution of 3 µL of
BF3‚OEt2 in 3 mL of dry CH2Cl2, and the solution was stirred
35 min. The solution was washed three times with 5%
NaHCO3 and once with brine and dried (MgSO4). A GC yield
of product 10 was 83%.
2,2-Dim eth yl-5-(1,2,3,4-tetr a h yd r on a p h th a len -1-yl)tet-
r a h yd r ofu r a n (14) was isolated by HPLC (30 mL retention
volumn) as described above: 1H NMR (CDCl3) δ 1.25 (s, 3 H),
1.29 (s, 3 H), 1.5-2.0 (m, 6 H), 2.7 (t, J ) 6.4 Hz, 2 H), 3.0 (q,
J ) 6.4 Hz, 2 H), 4.4 (q, J ) 6.8 Hz, 1 H), 7.0-7.5 (m, 5 H);
13C NMR (CDCl3) δ 22, 25, 27, 28, 30, 39, 42, 53, 81, 82, 126.2,
126.3, 129.0, 129.1, 129.2, 138; IR (NaCl disks) 1144 (s), 1052
(s), 740 (s, ortho); mass spectrum m/ e (relative intensity) 230
(1), 131 (29), 99 (100), 81 (84), 43 (38); exact mass calcd for
C16H22O 230.1665, found 230.1662.
(s), 770 and 690 (s, meta) cm-1; mass spectrum m/ e (relative
intensity) 246 (36), 135 (100), 121 (45), 83 (37), 55 (49). Anal.
Calcd for C16H22O2; C, 78.01; H, 9.00. Found: C, 77.98; H,
9.07.
1-(m-Meth oxyph en eth yl)cycloh exan ecar baldeh yde (20)
was identified based on NMR comparison with
a para-
substituted analog independently synthesized, and on GCMS
data: 1H NMR (CDCl3) δ 1.1-1.9 (m, 10 H), 2.1 (d, J ) 13 Hz,
2 H), 2.7 (m, 2 H), 3.8 (s, 3 H), 6.8 (m, 3 H), 7.2 (m, J ) 8 Hz,
1 H) m/ e (relative intensity); IR (NaCl disks) 1702 (s), 1260
(s); mass spectrum m/ e 246 (13), 134 (26), 122 (100), 121 (44).
See below for analysis of the analog.
1-(6-Meth oxy-1,2,3,4-tetr a h yd r on a p h th a len -1-yl)h ex-5-
en -1-ol (22): 1H NMR (CDCl3) δ 1.4-2.2 (m, 11 H), 2.3 (m, 3
H), 3.8 (s, 3 H), 3.8 (overlapping multiplet, 1 H), 5.0 (m, 2 H),
5.8 (m, 1 H), 6.7 (m, 2 H), 7.1 (d, J ) 8 Hz, 1 H); 13C NMR
(CDCl3) δ 20.7, 25.6, 26.1, 30.4, 33.9, 34.4, 43.8, 55.8, 75.3,
112.1, 114.8, 115.2, 129.8, 131.1, 139.5, 140.1, 158.5; IR (NaCl
disks) 3600-3200, 1260 (s), 1044 (s) 817 and 875 (1,2,4-
trisubstituted benzene); mass spectrum m/ e (relative inten-
sity) 260 (3) 162 (33), 161 (100), exact mass calcd for C17H24O2
260.1776, found 260.1782. Anal. Calcd for C17H24O2: C, 78.42;
H, 9.29. Found: C, 77.48; H, 9.29.
1-Me t h y l-3-(3-p h e n y lp r o p y l)-7-o x a b ic y c lo [2.2.1]-
h ep t a n e (15) was isolated by HPLC (36 mL retention
volume): 1H NMR (CDCl3) δ 1.55 (s, 3 H), 1.2-1.8 (m, 11 H),
2.6 (t, J ) 6 Hz, 2 H), 4.1 (d, J ) 5 Hz, 1 H), 7.1-7.3 (m, 5 H);
13C NMR (CDCl3) δ 21.6, 30.2, 30.3, 32.2, 36.0, 36.1, 36.7, 44.8,
45.7, 81.3, 84.3, 126.3, 128.9, 129.0 (2 C), 143.3; IR (NaCl disks)
1090 (m), 747 and 699 (monosubst benzene); mass spectrum
m/ e (relative intensity) 230 (4), 212 (38), 111 (32), 108 (36),
1-(8-Meth oxy-1,2,3,4-tetr a h yd r on a p h th a len -1-yl)h ex-5-
en -1-ol (23): 1H NMR (CDCl3) 1.2-2.2 (m, 11 H), 2.8 (m, 2
H), 3.2 (1 H), 3.7 (t, J ) 7 Hz, 1 H), 3.8 (s, 3 H), 5.0 (m, 2 H),
5.8 (m, 1 H), 6.70 and 6.76 (doublets, J ) 8 Hz, 2 H), 7.1 (t, J
) 8 Hz, 1 H); 13C NMR (CDCl3) δ 18.5, 24.8, 25.1, 28.8, 29.7,
33.8, 35.7, 37.5, 55.4, 74.8, 107.4, 114.3, 122.5, 126.6, 127.2,
138.9, 139.1, 157.2; IR (NaCl disks) 3600-3200, 1254 (s), 770
and 744 (1,2,3-trisubstituted benzene) cm-1; mass spectrum
m/ e (relative intensity) 191 (3), 162 (100), 161 (64), 115 (31).
2-P en t-4-en yl-2a ,3,4,5-tetr a h yd r o-2H-n a p h th o[1,8-bc]-
104 (66), 91 (100), 43 (85); exact mass calcd for C16H22
230.1671, found 230.1664.
O
1-(4-P en ten -1-yl)-6-m eth oxy-1,2,3,4-tetr a h yd r o-2-n a p h -
th ol (17), isolated by HPLC (85:15 trace hexane:EtOAc:EtOH),
gave a fraction of 17 that was 93% pure by GC: 1H NMR
(CDCl3) δ 1.5-2.2 (m, 2 H) 2.5-3.1 (m, 3 H), 3.7 (s, 3 H), 4.2
(m, 1 H), 5.0 (m, 2 H), 5.8 (m, 1 H), 6.6 (s, 1 H), 6.7 (m, 2 H),
7.1 (d, J ) 8 Hz, 1 H); 13C NMR (CDCl3) δ 28.1, 28.2, 28.3, 31,
37, 45, 57, 61, 113, 114 (2 C), 115, 132, 140; IR (NaCl disks)
3600-3200 (s), 1260 (s), 1038 (s), 870 (m), 810 (m); mass
spectrum m/ e (relative intensity) 246 (33), 238 (48), 177 (66),
44 (100) (The aromatic region was exactly the same as that of
1
fu r a n (25): H NMR (CDCl3) δ 1.2-2.2 (m, 8 H), 2.5-2.9 (m
5 H), 4.2-4.3 (ddd, J ) 10.8, 7.3 and 5.4 would suggest a cis
stereochemistry, 1 H), 5.0 (m, 2 H), 5.8 (m, 1 H), 6.7 (d, J ) 7
Hz, 1 H), 6.9 (d, J ) 7 Hz, 1 H), 7.0 (t, J ) 7 Hz, 1 H); 13C
NMR (CDCl3) δ 23.8, 25.8, 26.0, 27.1, 34.4, 34.5, 45, 77.2, 93.6,
106.4, 115.5, 120.0 (2 C), 128.6, 139.1; IR (NaCl disks) 1250
(s), 1230 (s), 760 and 730 (m; 1,2,3-trisubstituted) cm-1; mass
spectrum m/ e (relative intensity) 228 (54), 171 (100), 147 (62),
(1-(m-methoxyphenyl)-1-methanol made earlier.4b
)
145 (69), 131 (45), 115 (44); exact mass calculated for C16H20
228.1514, found 228.1519.
O
4-(4-P en t en -1-yl)-5-m et h oxy-1,2-d ih yd r on a p h t h a len e
(18) could not be purified by HPLC and was tentatively
identified based on the similarity of its GCMS data to that of
26: 230 (69), 134 (100).
1-Cycloh exyl-4-(m -m et h oxyp h en yl)-1-b u t a n on e (26):
1H NMR (CDCl3) δ 1.1-2.0 (m, 10 H), 2.3 (m, 1 H), 2.3 (t, J )
7 Hz, 2 H), 2.6 (t, J ) 7 Hz), 3.8 (s, 3 H), 6.7 (m, 3 H), 7.2 (t,
J ) 8 Hz, 1 H); 13C NMR (CDCl3) δ 25.7, 26.3, 26.5, 29.1, 35.9,
40.4, 51.5, 55.7, 111.9, 114.8, 121.6, 130.0, 144.1, 214.6; IR
(NaCl disks) 1710 (s), 1260 (s), 780 and 700 (m, meta
substituted) cm-1; mass spectrum m/ e (relative intensity) 260
(23), 134 (100). Anal. Calcd for C17H24O2; C, 78.20; H, 9.29.
Found: C, 78.38; H, 9.37.
1 -[3 -(m -M e t h o x y p h e n y l )p r o p y l ]c y c l o h e x a n e -
ca r ba ld eh yd e (27) could not be isolated pure as it was
present in small amounts and also air-oxidized very rapidly.
It was identified by comparison with the para isomer made
by exactly the procedures described earlier3 for the preparation
of 1-(3-phenylpropyl)cyclohexanecarbaldehyde (27 - CH3O).
This entails reacting the enolate of cyclohexanocarbonitrile
with 3-(3-methoxyphenyl)-1-bromopropane12 and then reducing
the product with diisobutylaluminum hydride (43% yield);
flash chromatography (80:20 trace hexane:CH2Cl2:EtOH) yielded
analytically pure 27: 1H NMR (CDCl3) 1.1-2.0 (m, 12 H), 2.5
(t, J ) 7 Hz, 2 H), 3.8 (s, 3 H), 6.8 (d, J ) 8 Hz, 2 H), 7.06 (d,
J ) 8 Hz, 2 H), 9.3 (s, 1 H); 13C NMR δ 23.3, 26.1, 26.5, 26.6,
31.7, 36.1, 36.6, 50.3, 55.9, 114.4, 129.02, 129.9, 134.6, 158.4,
207.9; IR (NaCl disks) 1710 (s), 1250 (s), 740 and 690 (m,
meta); mass spectrum m/ e (relative intensity) 260 (22), 121
(100). Anal. Calcd for C17H24O2: C, 78.42; H, 9.21. Found:
C, 78.52; H, 9.50.
1-Cycloh exyl-3-(m -m eth oxyp h en yl)-1-p r op a n on e (19)
was isolated by HPLC and also prepared independently.
A
Grignard reagent was prepared from 1-(m-methoxyphenyl)-
2-bromoethane4b (4.6 g, 21 mmol) and 0.52 g of Mg turnings
(21 mmol) in 10 mL of ether and 10 mL of THF. A solution of
2.4 g (21 mmol) of cyclohexanecarbaldehyde and 10 mL of ether
was added dropwise, and the mixture was stirred overnight.
The mixture was cooled with an ice bath, and 10% H2SO4 was
added slowly to neutralize the mixture. After 5% NaHCO3
and aqueous washes, the dried (MgSO4) organic layer was
rotary evaporated. The product was triturated and recrystal-
lized using hexane, giving 3.68 g (64%) of 1-cyclohexyl-3-(m-
methoxyphenyl)-1-propanol (33): mp 54-55.2 °C; 1H NMR
(CDCl3) δ 1-2.0 δ (m, 14 H), 2.6 (m, 1 H), 2.8 (m, 1 H), 3.8 (m,
1 H), 3.8 (s, 3 H), 6.8 (m, 3 H), 7.2 (m, 1 H); 13C NMR (CDCl3)
δ 26.1, 27.0, 27.2, 28.5, 29.8, 33.1, 36.5, 44.5, 55.8, 76.3, 111.7,
114.9, 121.5, 130.0, 144.8, 160.3; IR (melt) 3600-3200 (s), 1250
(s), 770 and 680 (s, meta). Anal. Calcd for C17H24O2; C, 77.38;
H, 9.74. Found: 77.72; H, 9.74.
33 (2.15 g, 8.6 mmol) was oxidized with pyridinium dichro-
mate (5.6 g, 26 mmol) in 21 mL of dry CH2Cl2 using standard
methods.3,13 The product was distilled 112-116 °C (0.01 mm),
giving 0.93 g of 19 (44% yield, 96% GC pure): 1H NMR (CDCl3)
δ 1.1-1.5 (m, 5 H), 1.5-2.0 (m, 5 H), 2.3 (m, 1 H) 2.7-2.9 (m,
4 H), 3.7 (s, 3 H), 6.7 (m, 3 H), 7.2 (t, J ) 7 Hz, 1 H); 13C NMR
δ 26.3 (2 C), 26.5, 29.1 (2 C), 30.4, 42.8, 51.5, 55.7, 112.0, 114.7,
121.3, 130.1, 143.7, 160.4, 213.5; IR (NaCl disks) 1700 (s), 1250
Ack n ow led gm en t is made to the donors of the
Petroleum Research Fund, administered by the Ameri-
can Chemical Society, for partial support of this work
and to the National Science Foundation (CHE-8804803
(13) Corey, E. J .; Schmidt, G. Tetrahedron Lett. 1979, 399.

2080 J . Org. Chem., Vol. 61, No. 6, 1996
Taylor et al.
Su p p or tin g In for m a tion Ava ila ble: Copies of 13C NMR
spectra of compounds 9-15, 17, 23, and 25 and 1H NMR of
20 and 25 reported in the Experimental Section (12 pages).
This material is contained in libraries on microfiche, im-
mediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
and 8807450) for the purchase of an NMR spectrometer
and for support for students. We thank J ohn Grutzner
of Purdue University for help with NMR interpretations
and J oe Lazar for help with MS interpretations and
analyses. This article is dedicated to Williams S.
J ohnson, whose work in part inspired our own and who
gave personal and professional encouragement to me
(S.K.T.) in his latter days.
J O951945F