Substituent Effects in the Intramolecular Diels-Alder Reaction of 6-Furylhexenoates

Article abstract of DOI:10.1021/jo9721554

The series of furyl enone esters, 4a-c, were synthesized from furan or furfural by straightforward routes. Their thermal intramolecular Diels-Alder reactions to give the tricyclic ketones 5a-c were studied in acetonitrile and toluene at two temperatures and the kinetics of the reactions determined. A comparison of these data with that obtained for the corresponding esters 2 to give the lactones 3 indicates that the rate enhancements seen for the esters (rate of dimethyl 310 times that of monomethyl) are much larger than those seen for the ketones (rate of dimethyl 6.8 times that of monomethyl). Thus, this is additional evidence for the earlier hypothesis that the presence of the oxygen atom in the tether is a factor responsible for the larger than normal rate enhancements.

Full text of DOI:10.1021/jo9721554

2968
J . Org. Chem. 1998, 63, 2968-2974
Su bstitu en t Effects in th e In tr a m olecu la r Diels-Ald er Rea ction of
6-F u r ylh exen oa tes
Michael E. J ung* and Mehrak Kiankarimi
Department of Chemistry and Biochemistry, University of California, Los Angeles,
Los Angeles, California 90095-1569
Received November 25, 1997
The series of furyl enone esters, 4a -c, were synthesized from furan or furfural by straightforward
routes. Their thermal intramolecular Diels-Alder reactions to give the tricyclic ketones 5a -c were
studied in acetonitrile and toluene at two temperatures and the kinetics of the reactions determined.
A comparison of these data with that obtained for the corresponding esters 2 to give the lactones
3 indicates that the rate enhancements seen for the esters (rate of dimethyl 310 times that of
monomethyl) are much larger than those seen for the ketones (rate of dimethyl 6.8 times that of
monomethyl). Thus, this is additional evidence for the earlier hypothesis that the presence of the
oxygen atom in the tether is a factor responsible for the larger than normal rate enhancements.
In tr od u ction
Previous work in these laboratories has demonstrated
that alkyl substitution on the tether of N-benzyl-N-
furfuryl-â-chloroacrylamides appreciably accelerates the
intramolecular Diels-Alder (IMDA) cyclization of these
compounds.1 Further extensive studies with furfuryl
F igu r e 1. Thorpe-Ingold effect (angle compression).
fumarates have revealed unexpectedly large rate en-
hancements with geminal disubstitution of alkyl groups
on the tether connecting the diene and dienophile.2 The
present work aims at a further understanding of the
accelerating effect on the rates of IMDA reactions of alkyl
substituents on the chain connecting the furan diene and
the dienophilic unit.
The contributions of the Thorpe-Ingold effect, namely
the effect of internal angle compression,3 and the reactive
rotamer effect to the overall observed rate acceleration
in IMDA reactions of furan dienes were first determined
by J ung and Gervay to be strongly in favor of the reactive
rotamer effect.2c Thus, the compression of the central
angle on alkyl substitution (Figure 1) was not as impor-
tant as the reactive rotamer effect (Figure 2). In order
for cyclization between the reactive units X and Y in
compound A to occur, rotation about the central C-C
bonds from the most stable and, therefore, more highly
populated, anti conformation to the gauche conformation
F igu r e 2. Reactive rotamer effect.
is required. But if one of the two carbons of this C-C
bond is substituted with one or two alkyl groups, the
energetic picture changes noticeably since the anti form
is now essentially equienergetic to the gauche form as
shown in B. Thus, the dialkyl-substituted cyclization
substrate can more easily attain the required transition
state for cyclization as compared to the unsubstituted
substrate and therefore would be expected to cyclize more
rapidly. With gem-dialkyl substitution, both the Thorpe-
Ingold effect of angle compression and the reactive
rotamer effect are operative. To distinguish between
these two possible sources of rate enhancement, rate
constants for the IMDA reaction of substituted furfuryl
fumarates 2 were determined. The choice of substituents
on the tether connecting diene and dienophile provided
a range of angles (θ) at the substituted carbon that
allowed for an analysis of the extent of the importance
of the angle compression (Thorpe-Ingold effect) in this
system.
(1) J ung, M. E.; Street, L. J . J . Am. Chem. Soc. 1984, 106, 8327.
(2) (a) J ung, M. E.; Gervay, J . J . Am. Chem. Soc. 1989, 111, 5469.
(b) J ung, M. E.; Gervay, J . Tetrahedron Lett. 1988, 29, 2429. (c) J ung,
M. E.; Gervay, J . J . Am. Chem. Soc. 1991, 113, 224. (d) J ung, M. E.
Synlett 1994, 186.
(3) Beesley, R. M.; Ingold, C. K.; Thorpe, J . F. J . Chem. Soc. 1915,
107, 1080.
S0022-3263(97)02155-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/11/1998

Diels-Alder Reaction of 6-Furylhexenoates
J . Org. Chem., Vol. 63, No. 9, 1998 2969
Sch em e 1. Ra tes of Cycliza tion of F u r fu r yl
Meth yl F u m a r a tes
Sch em e 2
Sch em e 3
As shown in Scheme 1, the unsubstituted precursor
2a cyclizes slowly and is the basis of comparison for other
substrates. With the substitution of one methyl on the
tether (2b), we observed a compression in the angle θ
and an 8-fold increase in rate constant.
With the addition of a second methyl (2c), we noted
further compression of the angle to the normal tetrahe-
dral value and another rate increase, this time by a factor
of 2000. Thus far, all these results could be explained
by the Thorpe-Ingold effect, although the magnitude of
the rate increase for 2c is decidedly larger than the
reported enhancements due to angle compression. The
other important results are the entries for compound 2d
and 2e. Where the internal angle is approximately the
same, as is the case for 2a and 2d , the Thorpe-Ingold
effect would predict a similar rate of cyclization, but the
results in Scheme 1 show a 200-fold increase for the
cyclobutyl-substituted substrate 2d . With the absence
of angle differences, the only factor at work in this
example would seem to be the reactive rotamer effect.
Similarly, in the cyclopropyl case 2e, where the angle is
even larger than the unsubstituted case 2a , the rate of
cyclization is about 10 times faster. Clearly the widening
of the angle does have a retarding effect (compare 2d )
but the net effect is one of acceleration relative to the
unsubstituted case, which directly contradicts the Thorpe-
Ingold postulate. These results argue strongly that angle
compression is not an important factor in the gem-dialkyl
effect in cyclizations giving five-membered rings. The
major factor on cyclization must therefore be a confor-
mational effect like the reactive rotamer effect. J ung and
Gervay also reported activation parameters for the
cyclization of 2a -2e2c which indicated that the large rate
acceleration seen in this system, namely the significant
lowering of the ∆G‡, is due almost entirely to a lowering
of the enthalpy of activation (∆H‡) and not to a difference
in the entropy of activation (∆S‡). These results are in
accord with most of the studies of the gem-dialkyl effect,4
although there is some disagreement.5
Accordingly, the oxygen of the fumarate was replaced by
a carbon, and the cyclization rates were determined for
the unsubstituted and mono- and dimethyl-substituted
derivatives 4a -c to give 5a -c. Comparison of the
results of this kinetic experiment with those obtained by
Gervay would clarify the role of oxygen in this unprec-
edented rate acceleration (Scheme 2). The work herein
describes the kinetics of cyclization of the furylethyl keto
esters 4a -c to give the cyclized ketoesters 5a -c and the
effect of solvent and temperature on their IMDA reac-
tions.
Resu lts a n d Discu ssion
A. Syn th esis. Compounds 4a -c were prepared with
the target 3-keto enoate 4a being synthesized in four
steps from furfural as shown in Scheme 3. The known
(E)-butenone 6,6 prepared by condensation of furfural
with acetone, was selectively hydrogenated using pal-
ladium on carbon to give the saturated ketone 7. Aldol
condensation of the ketone 7 with ethyl glyoxalate7 gave
the â-keto alcohol 8 as well as some starting material
and the product 9 of trapping of the thermodynamic
enolate with the glyoxalate. Tosylation and elimination
of the aldol product 8 proceeded slowly over 2 days to
afford our first precursor 4a as well as some unreacted
starting material.
The more challenging mono- and dimethyl substrates
4b,c were synthesized as follows. Methyl â-(2-furanyl)-
acrylate 10 was prepared in high yield from the reaction
of furfural with trimethyl phosphonoacetate (Scheme 4).
The subsequent cuprate addition proceeded in moderate
yield to give the â-methyl ester 11.8 Reduction of this
To determine the reason for the great acceleration
observed in the case of furfuryl fumarates, we postulated
a possible contribution from the presence of the oxygen
in the connecting chain to the overall rate enhancement.
(4) (a) Allinger, N. L.; Zalkow, V. J . J . Org. Chem. 1960, 25, 701.
(b) Schleyer, P. v. R. J . Am. Chem. Soc. 1961, 83, 1361. (c) Bruice, T.
C.; Pandit, K. P. J . Am. Chem. Soc. 1960, 82, 5858. (d) Danforth, C.;
Nicholson, A. W.; J ames, J . C.; Loudon, G. M. J . Am. Chem. Soc. 1976,
98, 4275.
(6) Leuck, G. J .; Cejka, L. Organic Syntheses; Wiley: New York,
1956: Collect. Vol. I, p 283.
(7) J ung M. E.; Shishido, K.; Davis, L. J . Org. Chem. 1982, 47, 891.
(8) Gustafsson, B.; Nilsson, M.; Ullenius, C. Acta Chem. Scand. B
1977, 31, 667.
(5) Parrill, A. L.; Dolata, D. P. Tetrahedron Lett. 1994, 35, 7319.

2970 J . Org. Chem., Vol. 63, No. 9, 1998
J ung and Kiankarimi
Sch em e 4
Sch em e 5
ester with diisobutylaluminum hydride (DIBAL) at low
temperature gave the desired aldehyde 12 as well as
some starting material and some of the corresponding
alcohol due to overreduction. The iodoalkene 13 was
prepared by addition of hydroiodic acid to propiolic acid
to form the corresponding acid, followed by esterification
with ethanol in benzene.9 The initially formed cis acid
was thermally isomerized to the desired trans isomer in
the presence of a catalytic amount of iodine.10 Coupling
of the iodoolefin 13 with the aldehyde 12 in DMSO under
the Nozaki conditions11 afforded allylic alcohol 14 in 52%
yield. The presence of nickel(II) chloride in catalytic
amounts was essential to the success of the reaction as
noted both by Nozaki and Kishi.12 PCC oxidation of the
alcohol 14 resulted in the formation of the target monom-
ethyl substrate 4b.
precursor and furnished the alcohol 19 in 60% yield.
Finally, Swern oxidation afforded the dimethyl hep-
tenoate 4c, the final cyclization substrate. With these
three cyclization substrates in hand, the kinetic study
of the IMDA reaction could be undertaken.
B. Kin etics. The kinetic experiments were conducted
in sealed 5 mm NMR tubes containing known concentra-
tions of the substrates 4a -c. For each substrate, the
experiment was carried out in deuterioacetonitrile and
deuteriotoluene, both at 25 and at 80 °C. The progress
of the reactions in the NMR tubes was monitored at
irregular intervals by high-field proton NMR. The
samples for high-temperature experiments were kept in
an 80 °C oil bath with the exception of substrate 4c for
which the entire high-temperature kinetic experiments
were performed in a preheated (80 °C) 500 MHz NMR
probe. The ratio of integrals for the shrinking furan peak
(δ 6.0 ppm) in the starting material and the growing H-6
peak in the product (δ 5.1 ppm) were used to determine
the rate constants.
The reversible nature of IMDA reactions of this type
was previously determined by Gervay2c and our findings
confirmed this observation, since reduction of the product
to starting material ratio was observed in several experi-
ments where the sample was heated after having been
initially kept at room temperature. To determine the
rate constants for these reactions, we made use of eq 1,
In preparing the third cyclization substrate, we used
the findings of Hoffmann and Ismail,13 who reported that
tertiary alkyl groups could be introduced on to furans
via a Friedel-Crafts alkylation. Specifically, 4-methyl-
2-oxo-3-pentenenitrile (15) was shown to react success-
fully with furan in the presence of aluminum trichloride
to give a mixture of the mono- and dialkylated products
16 and 17 as shown in Scheme 5. The acyl cyanide
intermediate formed after the first step could be isolated
on treatment with water and quick extraction with ether.
For our purposes, however, the more stable ester would
be more useful. The acyl cyanide (15) was prepared in
good yield (85%) from the corresponding commercially
available acyl chloride by heating with cyanotrimethyl-
silane (TMSCN) in the presence of catalytic zinc iodide.
Following the procedure outlined by Hoffmann and
Ismail, furan and the acyl cyanide 15 were reacted in
benzene to give, after workup and distillation, the two
products 17 and 16 in a 1:1 ratio, each in a maximum
18% yield. This alkylation step was incorporated into
the synthesis despite its low yield because the starting
materials were inexpensive. Reduction of the ester 16
to the aldehyde 18 proceeded in 50% yield. Aldehyde 18
was joined to the iodoolefin 13 by a Nozaki coupling in a
manner similar to that described for the monomethyl
k1
A { } C
k-1
[At] - [Aeq]
ln
) kapp
t
(1)
[Ao] - [Aeq]
the rate expression for reversible first-order reactions.
The kapp refers to the apparent rate constant, and [Ao]
and [Aeq] are the initial and equilibrium concentrations
of the acyclic starting material A, respectively. The
initial concentration [Ao] was known, and the concentra-
tion of the acyclic compound at selected times t ([At])
could be calculated from the ratio of the integral (in the
proton NMR spectra) of the indicated peaks of the
product vs those of the starting material and the initial-
concentration of the starting material, [Ao] (eq 2). The
(9) (a) Bowden, K.; Price, M. J . J . Chem. Soc. B 1970, 1466. (b) J ung,
M. E.: Hagenah, J . A.: Zeng, L.-M. Tetrahedron Lett. 1983, 24, 3973.
(10) Topek, K.; Vsetecka, V.; Prochazka, M. Collect. Czech. Chem.
Commun. 1978, 43, 2395.
(11) Takai, K.; Kimura, K.; Kuroda, T.; Hiyama, T.; Nozaki, H.
Tetrahedron Lett. 1983, 24, 5281.
(12) J in, H.; Uenishi, J .; Christ, W. J .; Kishi, Y. J . Am. Chem. Soc.
1986, 108, 5644 and references therein.
[Ao] ) [At] + [Ct]
[Ct] ) [Ao] - [At]
(13) Ismail, Z. N.; Hoffmann, H. M. R. Angew. Chem. 1983, 95, 737.

Diels-Alder Reaction of 6-Furylhexenoates
[Ct] [Ao] - [At]
J . Org. Chem., Vol. 63, No. 9, 1998 2971
)
) R
[At]
[At]
(experimentally determined ratio of integrals)
[Ao] - [At] ) [At]R
[Ao] ) [At](R + 1)
[Ao]
[Ct]
[At]
[At] )
where R )
(2)
(R + 1)
equilibrium concentration [Aeq], however, could not be
determined experimentally since in most cases the reac-
tion times were long and equilibrium was not achieved
during the experiment time. Therefore, we needed a
method for estimating [Aeq]. There are several available
for this purpose, the most commonly used one being the
Guggenheim approximation.14 For our purposes, the
method published by Moore in 197215 was particularly
appealing since it did not require data collection at
regular intervals and allowed for estimation of [Ao] if this
concentration was not accurately known. Given that eq
1 is not a linear equation, solution by the method of least
squares required an iterative procedure starting from
approximate values of unknown parameters. Since use
of the experimentally known concentration for [Ao] al-
lowed for no error in this value, the problem was treated
as a three-parameter problem, and the three normal
equations involved were solved for δkapp, δ[Ao], and δ-
[Aeq], using estimated values for kapp, [Ao], and [Aeq]. The
improved values for the three parameters above were
obtained by adding the δ values to the previous values
(eq 3). The improved values were then used to solve the
F igu r e 3. Cyclization of 4b at 25 °C in acetonitrile.
F igu r e 4.
Ta ble 1
kapp
substrate
solvent
T (°C)
k1 (s-1
)
k1 (s-1
)
4a
4b
4c
4c
4c
4c
acetonitrile 25 ∼1.6 × 10-6 ∼4.8 × 10-9 ∼1.6 × 10-6
acetonitrile 25
acetonitrile 25
3.4 × 10-6 7.5 × 10-7 2.6 × 10-6
6.2 × 10-6 5.1 × 10-6 1.1 × 10-6
4.9 × 10-6 3.0 × 10-6 1.9 × 10-6
5.0 × 10-3 1.8 × 10-3 3.2 × 10-3
4.0 × 10-3 8.0 × 10-4 3.2 × 10-3
toluene
acetonitrile 80
toluene 80
25
could be detected. The first entry in Table 1 is based
solely on two data points and, therefore, is no more than
an approximation of the rate constant. At 80 °C, the
cyclization was slightly faster for 4a but was accompa-
nied by simultaneous isomerization of the olefin from the
E- to the Z-isomer. Indeed, in all cases, this isomeriza-
tion was observed, but for the instances listed in Table
1, the data points were collected before noticable quanti-
ties of the isomerized alkene were observed. The isomer-
ized alkene also underwent a Diels-Alder cyclization as
indicated by the appearance of a second peak for the
bridgehead (H-6) proton of that adduct in the 5 ppm
region. Thus, the kinetic picture is more complex than
that of a simple unimolecular process. As shown in
Figure 4, there are three equilibria occurring simulta-
neously, the cyclization of the trans substrate, its isomer-
ization to its stereoisomers, and the cyclization of the cis
substrate. This series of reversible reactions and their
corresponding forward and reverse rate constants com-
plicated the process of kinetic analysis for these experi-
ments beyond the scope of our study. It is worth
mentioning, however, that isomerization of the olefin was
consistently faster in toluene than acetonitrile for all
three substrates at both temperatures.
k(improved) ) k(previous) + δk
(3)
three normal equations for the second set of δ values,
which were once again added to the previous parameters.
This process was continued until the δ values were small
enough and no improvement in the value of the unknown
parameters could be achieved. Values thus obtained for
[Ao] and [Aeq] were used to graphically determine kapp
from the slope of a plot of eq 1. A representative plot is
shown in Figure 3.
Using the values obtained for kapp, [Ao], and [Aeq], the
forward and reverse rate constants k1 and k-1 could be
easily determined using eqs 4 and 5.
kapp ) k1 + k-1
(4)
(5)
k1 [Ceq] [A0] - [Aeq]
Keq
)
)
)
-1
[Aeq]
[Aeq]
Table 1 lists the rate constants calculated for cycliza-
tion of compounds 4a -c at 25 °C in acetonitrile and for
4c in toluene at both 25 and 80 °C and in acetonitrile at
80 °C. For the reasons given below, additional kinetic
studies for the cyclizations of 4a and 4b could not be
carried out.
The IMDA reaction of the unsubstituted precursor 4a
at room temperature was extremely slow, and after
months only negligible amounts of the cyclized material
Comparison of the rate constants for the cyclization of
the keto substrates 4a -c with those for the IMDA
reaction of the furfuryl fumarates obtained by Gervay2c
is outlined in Table 2. The absolute values for the forward
rate constants of the fumarate series are larger, as
expected, but the relative increase in rate of cyclization
with substitution of one or two methyls is similar for the
(14) Guggenheim, E. A. Philos. Mag. 1926, 2, 528.
(15) Moore, P. J . Chem. Soc., Faraday Trans. 1 1972, 68, 1890.

2972 J . Org. Chem., Vol. 63, No. 9, 1998
J ung and Kiankarimi
Ta ble 2. Com p a r ison of F or w a r d Ra te Con sta n ts in
ambient temperature unless otherwise specified, and the
chemical shifts are reported in parts per million relative to
the lock of the solvent used. The infrared (IR) spectra were
recorded on an FTIR spectrometer. High-resolution mass
spectra (HRMS) were recorded on an Autospec instrument.
4-(2-F u r a n yl)-2-bu ta n on e (7). A solution of the enone 66
(3.4 g, 25.0 mmol) in ethyl acetate (150 mL) containing 5%
palladium on carbon (150 mg) was hydrogenated at atmo-
spheric pressure until the reaction was complete (TLC). The
suspension was filtered through Celite and concentrated under
reduced pressure. Flash column chromatography on silica gel
(hexane/ethyl acetate 2:1) yielded the desired product 7 as a
colorless oil (2.36 g, 68%). 1H NMR (400 MHz, CDCl3) δ: 7.29
(1H, dd, J ) 1.8, 0.6 Hz), 6.27 (1H, dd, J ) 3.1, 1.9 Hz), 5.99
(1H, dd, J ) 3.1, 0.8 Hz), 2.91 (2H, t, J ) 7.3 Hz), 2.78 (2H, t,
J ) 7.3 Hz), 2.16 (3H, s). 13C NMR (100 MHz, CDCl3) δ:
207.33, 154.51, 141.10, 110.22, 105.20, 41.70, 29.94, 22.18. FT-
IR (neat): 2920.6, 1720.7, 1363.8, 1165.1, 1008.9 cm-1. MS
(EI) m/z (rel intensity) 138.1 ([M]+, 65), 123.10 (10), 95.1 (66).
High-resolution EI MS (m/z): 138.0683, calcd for C8H10O2
138.0681.
E t h yl 6-(2-F u r a n yl)-2-h yd r oxy-4-oxoh exa n oa t e (8).
Lithium diisopropylamide was prepared as follows: to a
solution of diisopropylamine (0.76 mL, 6.6 mmol) in tetrahy-
drofuran (40 mL) at -78 °C was added dropwise n-butyl-
lithium (3.02 mL, 1.82 M). After 30 min, a solution of the
furylbutanone 7 (0.69 g, 5.00 mmol) in THF (5 mL) was added
via syringe over 10 min. After the reaction stirred for 1 h,
ethyl glyoxalate7 (1.02 g, 10 mmol) was added, and after 20
min aqueous ammonium chloride (40 mL) was added. The
organiclayer was separated and the aqueous layer extracted
twice with ether (2 × 10 mL). The combined organic layers
were dried over MgSO4, and the solvent was evaporated.
Purification by flash chromatography (hexane/ethyl acetate
2:1) on silica gel yielded the desired product 8 (0.39 g, 32%)
as well as the diastereomeric aldol products 9, resulting from
the thermodynamic enolate (0.138 g, 11%) and unreacted
starting material. Major Product (8). 1H NMR (400 MHz,
CDCl3) δ: 7.28 (1Η, m), 6.26 (1H, dd, J ) 3.0, 1.9 Hz), 5.99
(1H, dd, J ) 3.1, 0.7 Hz), 4.48 (1H, dd, J ) 5.3, 4.9 Hz), 4.25
(2H, q, J ) 7.1 Hz), 3.17 (1H, d, J ) 5.3 Hz), 2.94 (4H, m),
2.83 (2H, m), 1.28 (3H, t, J ) 7.1 Hz). 13C NMR (100 MHz,
CDCl3) δ: 206.79, 173.66, 154.15, 141.17, 110.25, 105.35, 66.97,
67.98, 45.96, 41.50, 21.84, 14.13. FT-IR (neat): 3489.7, 1736.2,
1720.7, 1097.6 cm-1. MS (EI) m/z (rel intensity) 240.1 ([M]+,
100), 222.1 (8), 94.0 (66). High-resolution EI MS (m/z):
240.0997, calcd for C12H16O5 240.0998. Minor Product (9). 1H
NMR (200 MHz, CDCl3) δ: 7.35 (0.28 H, br s), 7.30 (0.72 H,
br s), 6.25-6.31 (1H, m), 6.11-6.13 (0.28 H, m), 6.0-6.04 (0.72
H, m), 4.14-4.44 (2H, m), 2.74-3.38 (4H, m), 2.17 (2.16 H, s),
2.12 (0.84 H, s), 1.29 (2.16 H, t, J ) 7.3 Hz), 1.28 (0.84 H, t, J
) 7.2 Hz).
Aceton itr ile a t 25 °C
substrate
k1
k1 (rel) substrate
k1
k1 (rel)
4a
4b
4c
4.8 × 10-9
7.5 × 10-7
5.1 × 10-6
1
156
1062
2a
2b
2c
2.0 × 10-7
3.4 × 10-4 1700
1
5.5
1.1 × 10-6
two series as compared to the desmethyl compounds 2a
and 4a . The rate increase for the monomethyl precursor
4b is somewhat larger than the fumarate counterpart 2b,
while for the dimethyl substrate the relative increase is
greater in the fumarate series (2c) than in the keto series
(4c). However, given the inaccuracy of the rate constant
measured for 4a , it is not reasonable to compare the
increases on methyl substitution to 4b or 4c with those
in the ester series 2a -c. But the rate increase in going
from monomethyl to dimethyl is reliable in both systems
and therefore can be compared. Here, the increase in
the keto series (going from 4b to 4c) is 6.8, while the
corresponding increase in the ester series (going from 2b
to 2c) is 310. This is a significant difference and lends
further evidence to our earlier hypothesis,2c namely that
the presence of the ring oxygen is the major factor
responsible for the great rate enhancement observed in
IMDA reaction of these furan dienes with gem-dimethyl
substitution on the tether. Further experiments on
similar substrates not containing a furan will have to be
carried out to verify the importance of this specific diene
on the rate.
Con clu sion
In conclusion, the three cyclization substrates 4a -c
have been prepared from furan derivatives. The IMDA
reaction of these substrates in toluene and acetonitrile
both at 25 and at 80 °C have been studied. Comparison
of the calculated rate constants for substrates 4a -c at
25 °C in acetonitrile with the rate constants reported by
J ung and Gervay for the IMDA reaction of furfuryl
fumarates 2a -c under the same conditions indicates that
the oxygen in the tether of the fumarate is a factor
responsible for the uncharacteristically large rate en-
hancements observed for the cyclization of the mono- and
dimethyl substrates.
(E)-Eth yl 6-(2-F u r a n yl)-4-oxo-2-h exen oa te (4a ). The
â-keto alcohol 8 (0.39 g, 1.62 mmol) and p-toluenesulfonyl
chloride (0.62 g, 3.24 mmol) were stirred in pyridine (10 mL)
for 2 d. The reaction mixture was poured onto ice-water and
extracted twice with ether (2 × 10 mL). The organic extracts
were washed successively with two portions of 50% hydrochlo-
ric acid and water (3 × 5 mL). After drying over MgSO4 and
rotary evaporation of the solvent, the brown residue was
subjected to flash chromatography to give the desired elimina-
tion product 4a (170 mg, 47%) along with unreacted starting
material 8 (111 mg, 28%). 1H NMR (500 MHz, CDCl3) δ: 7.34
(1H, m), 7.10 (1H, d, J ) 16.0 Hz), 6.73 (1H, d, J ) 16.0 Hz),
6.31 (1H, br d, J ) 3.1 Hz), 6.05 (1H, br d, J ) 3.1 Hz), 4.31
Exp er im en ta l Section
Gen er a l Meth od s. All reactions were carried out under
argon with the exclusion of moisture. The following solvents
and reagents were distilled from the indicated agent under
argon: tetrahydrofuran (THF) and diethyl ether from sodium
benzophenone ketyl; dichloromethane, benzene, and toluene
from calcium hydride, and triethylamine from potassium
hydroxide. Flash column chromatography was carried out in
the indicated solvent system (in the percentage of volume) on
230-400 mesh silica gel.
(2H, q, J ) 7.1 Hz), 3.04 (4H, br s), 1.36 (3H, t, J ) 7.1 Hz).13
C
NMR (100 MHz, CDCl3) δ: 198.06, 165.32, 153.81, 141.12,
138.87, 131.01, 110.13, 105.38, 61.34, 39.44, 21.87, 13.99. FT-
IR (neat): 2922.5, 2851.1, 1724.6, 1703.4, 1367.7, 1304.0,
1184.4 cm-1
.
MS (EI) m/z (rel intensity): 222.1 ([M]+, 62),
193.0 (53), 149.1 (57), 95.0 (59). High-resolution EI MS (m/z):
222.0889, calcd for C12H14O4 222.0892.
Meth yl 3-(2-F u r a n yl)p r op en oa te (10). Sodium hydride
(0.84 g, 21 mmol) was suspended in dry THF and the mixture
cooled to 0 °C. Trimethyl phosphonoacetate (3.82 g, 21 mmol)
The proton nuclear magnetic resonance (1H NMR) spectra
were recorded at 200 or 400 MHz (as noted), and the carbon
nuclear magnetic resonance (13C NMR) spectra were recorded
at 100 MHz. Spectra were taken in the indicated solvent at

Diels-Alder Reaction of 6-Furylhexenoates
J . Org. Chem., Vol. 63, No. 9, 1998 2973
was added to the suspension with vigorous stirring. The white
gelatinous mixture was next treated with a THF (25 mL)
solution of freshly distilled furfural (1.92 g, 20 mmol). The
reaction was shown to be complete in 10 min by TLC, and then
saturated aqueous ammonium chloride solution (15 mL) was
added. The organic layer was separated and the aqueous layer
extracted with ethyl acetate (2 × 30 mL). The combined
organic materials were dried over MgSO4, and the solvent was
evaporated. Vacuum distillation (bp 95-98 °C, 2 mmHg)
afforded the R,â-unsaturated ester 10 as a yellow oil (2.74 g,
90%). The spectral data for this compound were consistent
those reported in the literature.16 1H NMR (200 MHz, CDCl3)
δ: 7.18 (1H, d, J ) 15.8 Hz), 7.00 (1H, br s), 6.35 (1H, d, J )
3.3 Hz), 6.21 (1H, dd, J ) 3.4, 1.8 Hz), 6.06 (1H, d, J ) 15.8
Hz), 3.53 (3H, s).
Meth yl 3-(2-F u r a n yl)bu ta n oa te (11). This compound
was synthesized from the furylpropenoate 10 and lithium
dimethylcuprate following the procedure of Gustafsson and co-
workers.8 To a suspension of copper(I) iodide (2.36 g, 12.4
mmol) in anhydrous ether (4 mL) cooled to 0 °C was added
methyllithium (27.7 mL, 0.9 M). After the solution stirred for
30 min at this temperature, a solution of the enoate 10 (1.39
g, 9.14 mmol) in ether (7 mL) was added dropwise. The
reaction mixture was allowed to warm to room temperature
overnight. A solution of saturated ammonium chloride (15 mL)
was then added to the reaciton. The two layers were sepa-
rated, and the aqueous layer was extracted with ether (2 × 5
mL). The combined organic extracts were dried over MgSO4
and evaporated. Flash chromatography on silica gel (hexane/
ether 5:1) gave the desired product 11 as a colorless oil (843
mg, 54%). The spectral data for this compound were consistent
with those reported in the literature.8 1H NMR (500 MHz,
CDCl3) δ: 7.31 (1Η, dd, J ) 1.8, 0.7 Hz), 6.27 (1H, dd, J )
1.8, 3.2 Hz), 6.01 (1H,br d, J ) 3.2 Hz), 3.67 (3H, s), 3.36 (1H,-
br sextet, J ) 7.1 Hz), 2.73 (1H, dd, J ) 15.4, 6.4 Hz), 2.46
(1H, dd, J ) 15.4, 8.2 Hz), 1.30 (3H, d, J ) 6.9 Hz).
3-(2-F u r a n yl)bu ta n a l (12). The methyl ester 11 (112 mg,
0.67 mmol) was dissolved in anhydrous toluene (3 mL) and
cooled to -78 °C. A solution of diisobutylaluminum hydride
(0.74 mL, 1.0 M in hexanes) was added dropwise. After 30
min, an additional amount of hydride (0.1 mL) was added to
reduce the residual ester. After 2 h, 2 N HCl (3 mL) was
added, and then the reaction was diluted with ether. The
organic layer was separated and the aqueous layer extracted
with ether (2 × 3 mL). The combined organic layers were dried
over MgSO4, and the solvent was evaporated. The colorless
oil was purified by flash chromatography on silica gel (hexane/
ether 4:1) to give the slightly impure aldehyde 12 (52 mg, 56%)
along with some starting material and another more polar
product (possibly the corresponding alcohol). The aldehyde
proved unstable to purification and was carried on as crude
material in later experiments. The spectral data for this
compound were consistent with the literature data.17 1H NMR
(400 MHz, CDCl3) δ: 9.76 (1Η, t, J ) 1.9 Ηz), 7.31 (1H, br s),
6.28 (1H, dd, J ) 3.2, 1.9 Hz), 6.02 (1H, d, J ) 3.2 Hz), 3.43
(1H, br sextet, J ) 7.0 Hz), 2.80 (1H, ddd, J ) 16.8, 6.6, 1.9
Hz), 2.59 (1H, ddd, J ) 16.8, 7.2, 1.9 Hz), 1.32 (3H, d, J ) 6.9
Hz).
as white crystals (2.56 g, 90%). The spectral data for this
compound were consistent with those reported in the litera-
ture.9 1H NMR (400 MHz, CDCl3) δ: 7.99 (1H, d, J ) 14.8
Hz), 6.87 (1H, d, J ) 14.8 Hz).
(E)-Eth yl 3-Iod op r op en oa te (13). (E)-3-Iodopropenoic
acid (2.56 g, 12.9 mmol) was dissolved in a 3:1 mixture of
absolute ethanol and benzene (40 mL) and treated with a
catalytic amount (two drops) of concentrated sulfuric acid. The
solution was refluxed for 14 h. A solution of saturated sodium
bicarbonate (20 mL) was added, and the mixure was extracted
with dichloromethane (4 × 10 mL). The combined organic
extracts were dried and evaporated to give the desired ethyl
ester 13 as a pale yellow oil (2.44 g, 84%) that was used
without further purification. The spectral data for this
compound were consistent with the literature data.9 1H NMR
(400 MHz, CDCl3) δ: 7.87 (1H, d, J ) 14.8 Hz), 6.87 (1H, d, J
) 14.8 Hz), 4.20 (2H, q, J ) 7.1 Hz), 1.29 (3H, t, J ) 7.1 Hz).
(E)-Eth yl 6-(2-F u r a n yl)-4-h yd r oxy-2-h ep ten oa te (14).
The furyl aldehyde 12 (76 mg, 0.55 mmol) and the trans
iodoolefin 13 were dissolved in dry dimethyl sulfoxide (13 mL)
with careful exclusion of moisture under an atmosphere of
argon. Chromium(II) chloride (405 mg, 3.3 mmol) and nick-
el(II) chloride (0.4 mg, 0.003 mmol) were added, and the
resulting deep green solution was stirred under argon for 2 d.
To the reaction were then added chloroform and saturated
ammonium chloride solution (10 mL each). The organic layer
was separated, and the aqueous layer was extracted with ethyl
acetate (3 × 5 mL). The combined organic layers were dried
over MgSO4, and the solvent was evaporated. Flash chroma-
tography on silica gel (hexane/ether 4:1) provided the desired
allylic alcohol 14 as a colorless oil (68 mg, 52%). Note that a
1:1 mixture of diastereomers is formed. 1H NMR (500 MHz,
CDCl3) δ: 7.32 (0.5 H, dd, J ) 1.8, 0.8 Hz), 7.31 (0.5 H, dd, J
) 1.8, 0.8 Hz), 6.92 (1H, dd, J ) 15.6, 5.1 Hz), 6.90 (1H, dd, J
) 15.6, 5.1 Hz), 4.32-4.36 (0.5 H, m), 4.17-4.22 (0.5 H, m),
4.20 (1H, q, J ) 7.1 Hz), 4.19 (1H, q, J ) 7.1 Hz), 3.12-3.16
(0.5 H, m), 3.03 (0.5 H, sextet, J ) 7.1 Hz), 1.97 (0.5 H, dt, J
) 13.9, 7.7 Hz), 1.88 (0.5 H, ddd, J ) 13.9, 10.1, 3.6 Hz), 1.79
(0.5 H, d, J ) 4.8 Hz), 1.70-1.77 (1H, m), 1.65 (0.5 H, d, J )
4.8 Hz), 1.29 (6H, m). 13C NMR (100 MHz, CDCl3) δ: 166.47
(2 C’s), 159.38, 159.02, 149.97, 149.64, 141.09, 140.98, 120.41,
120.00, 110.05 (2 C’s), 104.46, 104.01, 69.39, 69.24, 60.48,
60.45, 42.89, 42.64, 29.87, 29.79, 20.07, 19.22, 14.24 (2 C’s).
FT-IR (neat): 3440.5, 2971.7, 1719.8, 1701.0, 1278.0 cm-1. MS
(EI) m/z (rel intensity): 238.1 ([M]+, 19), 220.1 (31), 191.1 (38),
130.1 (40), 95.0 (100). High-resolution EI MS (m/z): 238.1214,
calcd for C13H18O4 238.1205.
(E)-Eth yl 6-(2-F u r a n yl)-4-oxo-2-h ep ten oa te (4b). The
allylic alcohol 14 (114 mg, 0.48 mmol) was added to a
suspension of Celite (200 mg) and pyridinium chlorochromate
(207 mg, 0.96 mmol) in dichloromethane (4 mL), and the dark
suspension was stirred overnight. Upon completion of the
reaction (TLC), the mixture was diluted with ether and filtered
through a pad of silica gel using dichloromethane (20 mL) to
wash the solids. After evaporation of the solvent, the yellow
oil was flash chromatographed on silica gel (hexane/ether 4:1)
to afford the pure product 4b as a clear oil (42 mg, 37%). 1H
NMR (400 MHz, CDCl3) δ: 7.29 (1H, dd, J ) 1.8, 0.8 Hz), 7.02
(1H, d, J ) 16.0 Hz), 6.65 (1H, d, J ) 16.0 Hz), 6.26 (1H, dd,
J ) 3.2, 1.8 Hz), 6.00 (1H, dd, J ) 3.2, 0.7 Hz), 4.26 (2H, q, J
) 7.1 Hz), 3.46 (1H, m), 3.07 (1H, dd, J ) 16.6, 6.0 Hz), 2.77
(1H, dd, J ) 16.6, 7.8 Hz), 1.30 (6H, m). 13C NMR (100 MHz,
CDCl3) δ: 198.23, 165.50, 158.36, 141.11, 139.29, 131.01,
110.07, 104.16, 61.43, 47.00, 28.98, 18.87, 14.13. FT-IR
(E)-3-Iod op r op en oic Acid . A mixture of propiolic acid
(1.01 g, 14.4 mmol) and aqueous hydriodic acid (47%, 3.92 g,
14.4 mmol) was heated to 90 °C for 4 h. Evaporation of the
water under reduced pressure gave a yellow oil that crystal-
lized on cooling. This crude cis iodoolefin was dissolved in
dioxane (10 mL) and treated with a few crystals of iodine. The
(neat): 2972.7, 2936.0, 1726.5, 1703.4, 1298.3, 1184.4 cm-1
.
1
solution was heated to 60 °C and was monitored by H NMR.
MS (EI) m/z (rel intensity): 236.1 ([M]+, 46), 128.0 (73), 95.0
(100). High-resolution EI MS (m/z): 236.1048, calcd for
After the solution was heated for 6 h, the heat was removed
and a 10% solution of Na2S2O3 (5 mL) was added. This
aqueous mixture was extracted with dichloromethane(4 × 5
mL). The combined organic extracts were dried over MgSO4
and the solvent evaporated. The title compound was obtained
C
13H16O4 236.1048.
4-Meth yl-2-oxo-3-p en ten en itr ile (15). To the suspension
of dry zinc iodide (13 mg, 0.042 mmol) in cyanotrimethylsilane
(843 mg, 8.5 mmol) at room temperature was added 3-methyl-
2-butenoyl chloride (1.00 g, 8.5 mmol) all at once. The mixture
was refluxed under argon for 2 h. Vacuum distillation of the
solution (bp 64-67 °C, 21 mmHg) afforded the desired acyl
cyanide 15 as a clear liquid (787 mg, 85%). The spectral data
(16) Ashby, E. C.; Lin, J . J .; Watkins, J . J . Tetrahedron Lett. 1977,
1709.
(17) J enner, G.; Rimmelin, J .; Antoni, F.; Libs, S.; Schleiffer, E. Bull.
Soc. Chim. Fr. 1981, II-65.

2974 J . Org. Chem., Vol. 63, No. 9, 1998
J ung and Kiankarimi
for this compound were consistent with the literature data.18
1H NMR (400 MHz, CDCl3) δ: 6.05 (1H, s), 2.18 (3H, s), 1.98
(3H, s).
δ: 166.58, 161.57, 150.25, 141.25, 119.58, 110.11, 104.13,
68.86, 60.35, 48.61, 35.02, 27.78, 27.23, 14.25. FT-IR (neat):
3445.3, 2970.7, 1709.8, 1705.3, 1279.0, 1159.4 cm-1. MS (EI)
m/z (rel intensity): 252.1 ([M]+,17), 109.1 (100). High-resolu-
tion EI MS (m/z): 252.1361, calcd for C14H20O4 252.1361.
Meth yl 3-(2-F u r a n yl)-3-m eth ylbu ta n oa te (16). This
ester was prepared by an application of the method of
Hoffmann and Ismail.13 A solution of the acylnitrile 15 (560
mg, 5.14 mmol) was dissolved in anhydrous benzene (15 mL)
and treated with aluminum trichloride (171 mg, 1.28 mmol).
After the solution stirred at room temperature for 30 min,
furan (384 mg, 5.65 mmol) was added all at once. A gummy
precipitate was formed. The reaction mixture was stirred at
room temperature for 14 h and then treated with methanol
(15 mL). The amber-colored solution was stirred at room
temperature overnight. The solution was next diluted with
ether (20 mL) and washed successively with water, saturated
sodium bicarbonate solution, and water again (20 mL each).
The combined organic layers were dried over MgSO4, and the
solvent was evaporated. Flash column chromatography on
silica gel (hexane/ether 9:1) gave the desired alkylated furan
16 (172 mg, 18%) as well as the dialkylated product 17 (274
mg, 18%). The spectral data for these compounds were
consistent with the literature data.13 Compound 16. 1H NMR
(400 MHz, CDCl3) δ: 7.32 (1H, dd, J ) 1.8, 0.8 Hz), 6.26 (1H,
dd, J ) 3.2, 1.8 Hz), 6.00 (1H, dd, J ) 3.2, 0.8 Hz), 3.58 (3H,
s), 2.61 (2H, s), 1.39 (6H, s). Compound 17. 1H NMR (400
MHz, CDCl3) δ: 5.86 (2H, s), 3.58 (6H, s), 2.58 (4H, s), 1.39
(12H, s).
3-(2-F u r a n yl)-3-m eth ylbu ta n a l (18). The methyl ester
16 (460 mg, 2.53 mmol) was taken up in toluene (10 mL) and
cooled to -78 °C. A solution of diisobutylaluminum hydride
(2.78 mL, 1.0 M in hexanes) was added slowly via syringe.
After the solution stirred for 2 h at -78 °C, 1 N HCl (5 mL)
was added. The organic layer was separated and the aqueous
layer extracted with ether (2 × 5 mL). The combined organic
extracts were washed with saturated sodium bicarbonate
solution (10 mL) and dried over MgSO4. After evaporation of
the solvents, the residue was subjected to flash chromatogra-
phy to give the desired aldehyde 18 as a colorless oil (190 mg,
50%). The aldehyde proved unstable and was used in the next
step promptly. 1H NMR (400 MHz, CDCl3) δ: 9.59 (1H, t, J )
2.4 Hz), 7.34 (1H, dd, J ) 1.4, 0.5 Hz), 6.29 (1H, dd, J ) 2.5,
1.5 Hz), 6.04 (1H, dd, J ) 2.5, 0.5 Hz), 6.21 (2H, d, J ) 2.4
Hz), 1.40 (6H, s).
(E)-E t h yl 6-(2-F u r a n yl)-4-oxo-6-m et h yl-2-h ep t en oa t e
(4c). Oxalyl chloride (69 µL, 0.79 mmol) was dissolved in
dichloromethane (3 mL) and cooled to -78 °C. Dimethyl
sulfoxide (0.11 mL, 1.59 mmol) was added to this mixture as
a solution in dichloromethane (1 mL). After the solution was
stirred for 15 min at this temperature, a solution of the allylic
alcohol 19 (100 mg, 0.40 mmol) in dichloromethane was added
slowly. After the solution was for 50 min, triethylamine (0.44
mL, 3.18 mmol) was added. The mixture was stirred at -78
°C for an additional 5 min and was then allowed to gradually
warm to room temperature. Sodium bisulfate (3 mL) was
added to the organic mixture, and the whole mixture was
extracted with ether (4 × 3 mL). The ethereal extracts were
dried over MgSO4 and evaporated. Flash chromatography on
silica gel (hexane/ether 3:1) afforded a mixture of the desired
product 4c and an unknown compound. A second chromatog-
raphy (hexane/ether 15:1) afforded the pure enone 4c (30 mg,
30%). 1H NMR (400 MHz, CDCl3) δ: 7.31 (1H, dd, J ) 1.8,
0.8 Hz), 6.68 (1H, d, J ) 15.9 Hz), 6.43 (1H, d, J ) 15.9 Hz),
6.24 (1H, dd, J ) 3.2, 1.8 Hz), 5.98 (1H, dd, J ) 3.2, 0.8 Hz),
4.21 (2H, q, J ) 7.1 Hz), 2.89 (2H, s), 1.38 (6H, s), 1.29 (3H, t,
J ) 7.1 Hz). 13C NMR (100 MHz, CDCl3) δ: 198.78, 165.56,
160.34, 140.97, 139.54, 130.07, 110.14, 104.09, 61.21, 52.39,
35.60, 26.98, 14.12. FT-IR (neat): 2963.0, 2924.5, 1728.4,
1695.6 cm-1. MS (EI) m/z (rel intensity): 250 ([M]+, 27), 222.1
(6), 127.0 (17), 109.1 (100). High-resolution EI MS (m/z):
250.1202, calcd for C14H18O4 250.1205.
(()-(4R,6S,7S,7a S)-Eth yl 2,3,3a ,6,7,7a -Hexa h yd r o-3,3-
d im et h yl-3a ,6-ep oxy-1-oxoin d en e-7-ca r b oxyla t e (5c).
Known amounts of the enoate 4c were placed in four NMR
tubes. Acetonitrile-d3 (0.5 mL) was added to two of the tubes,
while toluene-d8 (0.5 mL) was added to the other two. The
NMR tubes containing solutions of the enoate were then sealed
under vacuum. One each of the samples containing acetoni-
trile and toluene was stored at room temperature and moni-
tored periodically by proton NMR. Each of the other two
samples was placed in a preheated (80 °C) NMR probe and
monitored by scanning at regular intervals. The integration
data thus obtained was processed using an Excel spread sheet
program. After the conclusion of the kinetic experiment, the
sealed tubes were reopened and the material in them combined
and evaporated. Upon chromatography, the cyclized Diels-
Alder product 5c was obtained as a clear oil. 1H NMR (400
MHz, CDCl3) δ: 6.61 (1H, d, J ) 5.9 Hz), 6.33 (1H, dd, J )
5.9, 1.5 Hz), 5.19 (1H, dd, J ) 5.0, 1.5 Hz), 4.11 (2H, q, J )
7.1 Hz), 3.40 (1H, dd, J ) 5.0, 3.1 Hz), 2.85 (1H, br d, J ) 3.1
Hz), 2.48 (1H, d, J ) 18.9 Hz), 2.36 (1H, dd, J ) 18.9, 1.1 Hz),
1.36 (3H, s), 1.24 (3H, t, J ) 7.1 Hz), 1.21 (3H, s). 13C NMR
(100 MHz, CDCl3) δ: 215.31, 170.58, 135.66, 133.96, 101.12,
80.03, 61.12, 55.37, 52.40, 51.11, 35.90, 26.76, 22.91, 14.21.
FT-IR (neat): 2965.0, 1742.0, 1194.1 cm-1. MS (EI) m/z (rel
intensity) 250.1 ([M]+, 18), 109.1 (100). High-resolution EI MS
(m/z): 250.1209, calcd for C14H18O4 250.1205.
(E)-Eth yl 6-(2-F u r a n yl)-4-h yd r oxy-6-m eth yl-2-h ep ten -
oa te (19). In a manner analogous to that used in the
preparation of compound 14, the furyl butanal 18 (100 mg,
0.66 mmol) was coupled to the vinyl iodide 13 (447 mg, 1.98
mmol) using chromium(II) chloride (489 mg, 3.98 mmol) and
catalytic nickel(II) chloride (0.5 mg, 0.004 mmol). After
workup and flash column chromatography (hexane/ether 3:1),
the desired allylic alcohol 19 was obtained as a slightly impure
colorless oil (100 mg, 60%). 1H NMR (400 MHz, CDCl3) δ: 7.34
(1H, dd, J ) 1.8, 0.6 Hz), 6.78 (1H, dd, J ) 15.6, 4.8 Hz), 6.28
(1H, dd, J ) 3.2, 1.9 Hz), 6.06 (1H, dd, J ) 3.2, 0.6 Hz), 5.93
(1H, dd, J ) 15.6, 1.7 Hz), 4.27 (1H, m), 4.17 (2H, q, J ) 7.1
Hz), 1.87 (2H, m), 1.52 (1H, d, J ) 4.1 Hz), 1.37 (3H, s), 1.34
(3H, s), 1.27 (3H, t, J ) 7.1 Hz). 13C NMR (100 MHz, CDCl3)
(18) Hoffmann, H. M. R.; Haase, K.; Ismail, Z. M.; Preftitsi, S.;
Weber, A. Chem. Ber. 1982, 115, 3880.
J O9721554