On the stereoselectivity of γ-lactol substitutions with allyl- and propargylsilanes - Synthesis of disubstituted tetrahydrofuran derivatives

Article abstract of DOI:10.1002/1099-0690(200012)2000:23<3893::AID-EJOC3893>3.0.CO;2-E

Monosubstituted γ-lactols 1a-1c, 3a-3c and 4a-4c, as well as disubstituted γ-lactol 5 and the γ-hydroxy-substituted γ-lactone 6, were transformed into disubstituted tetrahydrofuran derivatives by treatment with allyl- and propargylsilanes in the presence of Lewis acids. The diastereoselectivities were moderate to excellent and are interpreted by application of the Felkin-Anh model to cyclic oxocarbenium ions. The effects of the equilibria between conformers of the intermediates are discussed. The surprisingly high diastereoselectivity of 4-substituted γ-lactols 3 can be explained on this basis.

Full text of DOI:10.1002/1099-0690(200012)2000:23<3893::AID-EJOC3893>3.0.CO;2-E

FULL PAPER
On the Stereoselectivity of γ-Lactol Substitutions with Allyl- and
Propargylsilanes ؊ Synthesis of Disubstituted Tetrahydrofuran Derivatives
Andreas Schmitt[a][°] and Hans-Ulrich Reißig*[a]
Dedicated to Prof. Helmut Vorbrüggen on the occasion of his 70th birthday
Keywords: Carbocations / FelkinϪAnh model / Lewis acids / Silanes / Tetrahydrofuran
Monosubstituted γ-lactols 1a−1c, 3a−3c and 4a−4c, as well
as disubstituted γ-lactol 5 and the γ-hydroxy-substituted γ-
tion of the Felkin−Anh model to cyclic oxocarbenium ions.
The effects of the equilibria between conformers of the inter-
lactone 6, were transformed into disubstituted tetrahydrofu- mediates are discussed. The surprisingly high diastereose-
ran derivatives by treatment with allyl- and propargylsilanes
lectivity of 4-substituted γ-lactols 3 can be explained on this
in the presence of Lewis acids. The diastereoselectivities basis.
were moderate to excellent and are interpreted by applica-
Introduction
Substituted tetrahydrofuran derivatives are important
synthetic intermediates[2] and occur as structural units in
many interesting natural products, such as pheromones[3] or
polyether antibiotics.[4] Thus, their diastereoselective syn-
thesis is currently receiving much attention.[5] In our group
we became interested in Lewis acid-promoted substitution
at the anomeric centre of γ-lactols.[6] Although widely em-
ployed in carbohydrate chemistry,[7] this methodology is
only rarely applied to the synthesis of simple tetrahydrofu-
ran derivatives.[8] This might be due to the lack of informa-
tion about the stereochemical outcome of this reaction,
which makes it difficult to predict the diastereoisomeric dis-
tribution. In this paper we wish to report our results regard-
ing the Lewis acid-induced reactions of γ-lactols with allyl-
and propargylsilanes.[9] We also propose a model to explain
the diastereoselectivities observed in these reactions. For
this purpose, γ-lactols 1Ϫ5 and γ-hydroxy-substituted γ-lac-
tone 6, bearing substituents of varying steric demand, were
treated with a broad variety of silylated nucleophiles under
Lewis acid conditions.
ous workup, 2,3-disubstituted tetrahydrofurans 11Ϫ14 were
isolated as mixtures of cis and trans isomers (Table 1).
The stereochemical outcome of the reactions depends
considerably on the nature of the γ-lactol substituent and
on the substitution pattern of the nucleophiles. The influ-
ence of substituent R is exemplified by reactions with allyl-
silane 8 (entries 2Ϫ4). A methyl group (1a) or a phenyl
group (1b) both show a similar influence on the diastereose-
lectivity of the reaction, R ϭ Me being slightly more select-
ive. Not unexpectedly, the tert-butyl group (1c) directs the
attack of the nucleophile exclusively in the trans direction
(entry 4).
Results
With respect to the nucleophile, substituents at the react-
γ-Lactols 1aϪ1c were treated with propargylsilane 7 and ive centre influence the diastereoselectivity of the reaction
allylsilanes 8Ϫ10 in the presence of BF3·OEt2. After aque- strongly, while a substituent remote from the carbon in-
volved in bond formation shows no effect. Thus, going from
propargylsilane 7 to prenylsilane 10 (entries 1 and 6), the
[a]
Freie Universität Berlin, Institut für Chemie Ϫ Organische
diastereoselectivity of the reactions increases from trans/
cis ϭ 60:40 to 95:5, while reaction of allylsilane 8 and of
1,1-dimethylallylsilane 9 results in equal diastereomer ratios
(trans/cis ϭ 68:32, entries 2 and 5).
Chemie,
Takustrasse 3, 14195 Berlin, Germany
Fax: (internat.) ϩ49-30/838 55 367
E-mail: Hans.Reissig@Chemie.FU-Berlin.de
See ref.[1]
[°]
Eur. J. Org. Chem. 2000, 3893Ϫ3901
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1434Ϫ193X/00/1223Ϫ3893 $ 17.50ϩ.50/0
3893

A. Schmitt, H.-U. Reißig
FULL PAPER
at a higher level of selectivity Ϫ ranging from trans/cis ϭ
86:14 with propargylsilane 7 (entry 1) to exclusive forma-
tion of trans-19 when using prenylsilane 10 (entry 5).
Table 2. BF3-promoted reactions between 4-substituted γ-lactols
3aϪ3c and silanes 7, 8 and 10
Table 1. BF3-promoted reactions between 3-substituted γ-lactols
1aϪ1c and silanes 7Ϫ10
Entry
Lactol
Silane
Product
Yield
(%)
trans:cis
Entry
Lactol
Silane
Product
Yield
(%)
trans:cis
1
2
3
4
5
3a
3a
3b
3c
3a
7
17
93
68
80
84
95
86:14
95:5
1
2
3
4
5
6
1a
1a
1b
1c
1a
1a
7
11
65
53
91
74
67
50
60:40
68:32
63:37
Ն 98:2
68:32
95:5
8
18a
18b
18c
19
8
12a
12b
12c
13
8
95:5
8
8
Ն 98:2
Ն 98:2
8
10
9
10
14
Reactions of 5-substituted γ-lactols 4aϪ4c with organosi-
lanes seem to be the least predictable transformations. They
can lead to preferential formation of either cis isomers or
trans ones, depending on the nature of the substituent at
the γ-lactol ring and the steric demand of the attacking nu-
cleophile. Combination of a small substituent (R ϭ Me,
4a) and low organosilane steric demand leads to good cis
selectivity (Table 3, entry 1, 21, cis/trans ϭ 80:20). Increas-
Access to the cis series of 2,3-disubstituted tetrahydrofu-
ran derivatives can also be gained by reduction of a 2,3-
disubstituted γ-lactol such as
2
with triethylsilane
(15).[6aϪ6c,10] As in the other examples, the nucleophile at-
tacks trans to the substituent in the 3-position, thus re-
sulting with moderate selectivity (trans/cis ϭ 21:79) in cis-
disubstituted tetrahydrofuran 16.
Compared with those of 3-substituted γ-lactols 1aϪ1c,
the reactions of γ-lactols 3aϪ3c proceed with significantly
higher diastereoselectivity. The nature of the substituent at
the γ-lactol ring has a less pronounced influence (entries
2Ϫ4 of Table 2). In reaction with allylsilane 8, a methyl
group (3a) and a phenyl group (3b) induce the same excel-
lent diastereoselectivity (trans/cis ϭ 95:5), while a tert-butyl
substituent (3c) again directs the attack of the nucleophile
exclusively in the trans direction (trans/cis Ն 98:2). The ef-
fect of the substitution pattern in the silane is similar to
that observed in reactions of 3-substituted γ-lactols 1, but
3894
Eur. J. Org. Chem. 2000, 3893Ϫ3901

Synthesis of Disubstituted Tetrahydrofuran Derivatives
FULL PAPER
Table 3. BF3-promoted reactions between 5-substituted γ-lactols similar diastereoselectivities. For the reaction of the 5-hy-
4aϪ4c and silanes 7, 8, 10 and 20
droxylactone 6, product 26 was obtained in higher yield but
with similar diastereoselectivity when titanium tetrachloride
was employed as the Lewis acid, instead of boron trifluor-
ide.
Entry
Lactol
Silane
Product
Yield
(%)
trans:cis
1
2
3
4
5
6
7
4a
4a
4b
4c
4a
4b
4c
7
21
82
20:80
40:60
47:53
88:12
83:17
77:23
90:10
8
22a
22b
22c
23a
23b
24
70
Table 4. Lewis acid dependence of reactions between γ-lactols and
silanes 8 and 10
8
80
8
72
10
10
20
93
Lactol
Silane
Lewis Acid
Product
Yield
(%)
trans:cis
84
61[a]
[a]
1a
1a
1a
1a
4a
4a
4a
8
BF3·Et2O
ZnCl2
12a
12a
14
53
79
50
79
79
Ϫ
34
76
80
Ϫ
68:32
69:31
95:5
95:5
40:60
Ϫ
40:60
40:60
47:53
Ϫ
Purity only 83% according to GC.
8
10
10
8
BF3·Et2O
ZnCl2
14
BF3·Et2O
TiCl4
SnCl4
EtAlCl2
BF3·Et2O
TiCl4
BF3·Et2O
TiCl4
22a
Ϫ[a]
22a
22a
22b
Ϫ[a]
26
ing the steric demand of the nucleophile in reaction with γ-
lactols bearing small substituents (R ϭ Me, Ph) signific-
antly decreases the cis selectivity (entries 2 and 3), while 4a
8
8
8
4b
8
reaction of strongly hindered nucleophiles such as prenylsil-
4b
8
ane 10 and the silylated propargylsilane 20 leads to prefer-
ential formation of the trans isomers (entries 5Ϫ7). γ-Lactol
4c, with the large tert-butyl substituent, reacted with good
selectivity with allylsilane 8, to form the trans isomer (entry
4, trans/cis ϭ 88:12).
6
6
8
74
99
74:26
71:29
8
26
[a]
Decomposition.
The effect of the leaving group on the diastereoselectivity
Reaction of the trans-4,5-dimethyl-substituted γ-lactol 5
with allylsilane 8 furnished an 86:14 mixture of two diaster-
eomeric tetrahydrofuran derivatives 25. The major isomer
was assigned as 2,4-trans-substituted 25. Thus, the 4-methyl
group had the expected effect on the outcome of this reac-
tion, although this was modulated slightly by the 5-methyl
substituent.[11]
of the γ-lactol substitution was also examined. Thus, the
reactions of γ-lactol 1a and of 2-phenylselenenyltetrahydro-
furan 27[12] with allylsilane 8 were studied. In both reac-
tions, the allyl-substituted derivative 12a was formed in
comparable yields and, within the limits of NMR accuracy,
identical diastereomer ratios. Thus, the nature of the leaving
group showed no influence on the stereochemical out-
come.[13]
As well as simple γ-lactols, 5-hydroxy-3,4-dihydrofuran-
2-ones such as 6 can also be employed in Lewis acid-pro-
moted reaction with nucleophiles. The corresponding γ-lac-
tone 26 was obtained with moderate trans selectivity (trans/
cis ϭ 74:26).
1
Assignment of the diastereomers, based on H and 13C
NMR spectroscopic data, was unambiguous in most cases.
The measurements were supplemented by NOE experi-
ments for some characteristic examples. A more detailed
discussion of the configurational assignment has been pre-
sented previously.[1,6d]
We also studied the influence on the diastereoselectivity
of γ-lactol reactions exerted by the Lewis acid and by the
leaving group. As shown in Table 4, the nature of the Lewis
acid employed to promote the reaction of γ-lactols with or- Discussion
ganosilanes had no significant influence on the stereochem-
ical outcome. Strong Lewis acids capable of chelate forma-
Two borderline mechanisms for Lewis acid-induced reac-
tion result in reduced yields or complex product mixtures, tions of acetals with nucleophiles are discussed in the liter-
probably due to the formation of open-chain by-products. ature. The reaction may involve an oxocarbenium ion in an
On the other hand, all monocoordinating Lewis acids em- SN1-type process, or it may occur by preliminary formation
ployed promoted the reaction equally well, providing fairly of an acetal/Lewis acid complex, from which the product is
Eur. J. Org. Chem. 2000, 3893Ϫ3901
3895

A. Schmitt, H.-U. Reißig
FULL PAPER
formed in an SN2-type reaction. In their investigations on Table 5. MMX-89 force-field calculations on methyl-substituted
oxocarbenium ions
Lewis acid-induced substitutions in open-chain acetals,
Denmark and co-workers[14] found that the Lewis acids em-
Oxocarbenium Ion
∆Hf
∆∆Hf
K[a]
ployed exerted a strong influence on the stereochemical out-
come, as did the nature of the leaving group. The authors
concluded that the reaction mechanism depends on the sub-
strate and the Lewis acid employed. Results for substitution
[kcal/mol]
[kcal/mol]
3-Meax
3-Meeq
4-Meax
153.01
153.42
152.95
152.20
153.15
152.79
0.41
0.75
0.36
2.9
6.7
2.5
in cyclic acetals reconfirmed these findings.[15] On the other 4-Meeq
5-Meax
5-Meeq
hand, the reaction mechanism for the Lewis acid-promoted
reaction of γ-lactols with nucleophiles presented in this re-
[a]
port seems unambiguous. Since there is no significant de-
pendence of diastereoselectivity either on the Lewis acid
employed or on the leaving group, it seems that neither of
them participates in the product-determining step. Hence,
the reaction should follow the SN1-mode, via an oxocarben-
ium ion intermediate. For such a mechanism there are still
two possible pathways, since either a cyclic oxocarbenium
ion or an open-chain intermediate may be involved. Reac-
tion via the cyclic oxocarbenium ion would lead directly to
the isolated tetrahydrofuran derivatives, while addition of
the nucleophile to the carbonyl/Lewis acid complex would
yield the corresponding 1,4-diol, which could cyclize in the
presence of Lewis acids. This pathway can be ruled out,
since a smooth cyclization is rather unlikely under the rela-
tively mild conditions employed.[16] Moreover, the high dia-
stereoselectivities recorded for 4-substituted γ-lactols 3 or
(in part) for 5-substituted γ-lactols 4 cannot plausibly be
explained on the basis of an open-chain intermediate.[17]
Thus, to interpret the observed diastereoselectivities, we as-
sume the formation of the cyclic oxocarbenium ion by
Lewis acid-assisted dissociation of the hydroxyl group, fol-
lowed by reaction with the silanes.
K was calculated from ∆∆Hf by neglecting ∆∆Sf.
stability for the conformer with the substituents in pseudo-
equatorial position, but the energy differences are consider-
ably larger (∆∆Hf ϭ 0.86 to 2.32 kcal/mol). The reliability
of these calculations is demonstrated by comparison of the
bond lengths obtained for the CϭOϩ moiety (1.28 A) with
˚
those calculated by ab initio methods for similar ions
[21]
˚
(1.26 A).
We explain the diastereoselectivities by assuming a
FelkinϪAnh-type approach of the nucleophile to the cyclic
oxocarbenium ion. Thus, the silane attacks the electrophilic
carbon opposite to the pseudoaxial (perpendicular) sub-
stituent at C-3. For the 3-substituted systems, this stereo-
electronic effect should strongly favour formation of
trans products from A, since the alternative attack should be
hampered by the equatorial substituent R in conformer B.
However, this kinetic effect on the diastereoselectivity[22]
must be modified by the slight thermodynamic preference
of B over A. Thus, small nucleophiles such as 7 and 8 may
also attack B with small substituents (Me, Ph), giving trans/
cis mixtures of the order of 2:1. Only if the nucleophile is
bulky (e.g., 10) or if substituent R is sterically demanding
(tert-butyl) is almost exclusive formation of trans products
observed (Table 1).
The oxocarbenium ion may exist in two distinct half-
chair conformations, which are in rapid equilibrium
through pseudorotation.[18] To estimate this conformer
equilibrium, we performed MMX-89 force-field calcula-
tions,[19] checked by the PIMM-88 force-field as developed
by Lindner and co-workers.[20] These calculations sup-
ported the assumption that conformers with their substitu-
ents in pseudoequatorial position are more stable than
those with these substituents in pseudoaxial position. The
data obtained for the methyl-substituted oxocarbenium ion
are presented in Table 5.
While kinetic (stereoelectronic) and thermodynamic ef-
fects are opposed for 3-substituted oxocarbenium ions A
and B, they appear to cooperate for 4-substituted oxocar-
Similarly, the corresponding calculations[1] for phenyl- benium ions. In a FelkinϪAnh approach to the more stable
and tert-butyl-substituted oxocarbenium ions show higher conformer C, the pseudoequatorial substituent R is in a
more remote position than the pseudoaxial substituent R
3896
Eur. J. Org. Chem. 2000, 3893Ϫ3901

Synthesis of Disubstituted Tetrahydrofuran Derivatives
FULL PAPER
in the less stable conformation D. In accordance with this a highly electrophilic, cyclic, O-acylated oxocarbenium ion.
interpretation, we obtain very high trans selectivities in this The stereoselectivity of the reaction between β-formyl ester
series, even with slim nucleophiles and small R substitu- 28 and allylsilane 8 is highly dependent on the Lewis acid
ents (Table 2).
employed;[25] this is not the case for the reaction of 6, which
gives nearly identical trans/cis ratios of γ-lactone 26. There-
fore, we assume that this transformation also proceeds via
the cyclic intermediate, as demonstrated for the other γ-
lactols in this report.
For 5-substituted oxocarbenium ions, conformer F, with
the pseudoequatorial substituent, is slightly preferred over
E, but the R substituent is positioned rather remote from
the reaction centre. Thus, the thermodynamic effect should
gain in importance, compared to the case of the 3-substi-
tuted oxocarbenium ions A/B.
The reaction between the oxocarbenium ion bearing the
smallest substituent (R ϭ Me) and the slim (and fast) silane
7 seems to reflect the approximate ratio of the two con-
formers F and E (K ഠ 2.5 according to MMX-89, Table 5),
thus leading to considerable cis selectivity (entry 1, Table 3).
Conclusions
These reactions between monosubstituted γ-lactols
1aϪ1c, 3aϪ3c and 4aϪ4c and a variety of allyl- and pro-
pargylsilanes in the presence of boron trifluorideϪdiethyl
ether proceed in good yields and Ϫ strongly dependent on
the substitution pattern Ϫ with good to excellent diastereo-
selectivity. Surprisingly, the 4-substituted γ-lactols 3 furnish
their corresponding tetrahydrofuran derivatives with higher
diastereoselectivity than do the 3-substituted or 5-substi-
tuted γ-lactols 1 and 4. This behaviour may be explained
by the converging or opposing influences of thermodyn-
amic and kinetic (stereoelectronic) effects.[26] Our model
study makes disubstituted tetrahydrofurans available in a
predictable manner; the results may be applied to the syn-
theses of more complex products. The interplay of two sub-
stituents has been demonstrated by the conversion of disub-
stituted γ-lactol 5 into trisubstituted tetrahydrofuran deriv-
atives 25. A further report will deal with Lewis acid-pro-
moted reactions between γ-lactols 1, 3 and 4 and silyl enol
ethers.[27]
An increase in steric effects initially leads to the loss of
any stereoselectivity (entries 2 and 3, Table 3). However,
with bulky groups either on the oxocarbenium ion or in the
silane, it is possible to achieve even moderate to good trans
selectivity (entries 4Ϫ7, Table 3). These steric effects seem
to promote reaction via conformer E.
The stereoselectivity of the reaction between trans-4,5-
dimethyl-substituted γ-lactol 5 and allylsilane 8 can be ra-
tionalized smoothly by comparison of the associated
monosubstituted oxocarbenium ions. Thus, the 4-methyl
group should induce excellent trans selectivity (see con-
former C and entry 2, Table 2) but the effect is opposed by
that due to the 5-methyl group, which exhibits a weak cis
preference (see conformer F and entry 2, Table 3). As would
be expected from our interpretation discussed above, the
effect due to the 4-methyl group dominates.[23]
Experimental Section
General Remarks: See ref.[6d] NMR spectroscopic data of mixtures
of diastereomers are given in the following manner: signal of trans
isomer/signal of cis isomer. The tetrahydrofuran derivatives pre-
The reaction between 5-hydroxylactone 6 and allylsilane
8 may proceed via an open-chain β-formylcarboxylic acid,
as discussed for related examples in the literature,[24] or via pared are very volatile and rather sensitive to oxygen. Therefore,
Eur. J. Org. Chem. 2000, 3893Ϫ3901
3897

A. Schmitt, H.-U. Reißig
FULL PAPER
for a number of products (in particular with 1,2-propadienyl sub-
stituents) no correct elemental analysis could be obtained.
4.0, 9.0, 13.5 Hz, 0.5 H, 3-H), 1.27/1.16 (d, J ϭ 6.0 Hz, 3 H, 5-
Me), 0.99/0.96 (d, J ϭ 6.0 Hz, 3 H, 4-Me); the ratio of the two
anomers is strongly dependant on the concentration of the solu-
tion. Ϫ 13C NMR (CDCl3): δ ϭ 98.7/97.4 (d, C-2), 79.8/82.8 (d,
C-5), 41.4/42.4 (t, C-3), 38.2/40.7 (d, C-4), 18.3/20.4, 15.8 (2 d, 4-
Me, 5-Me).
Starting Materials: Preparations of γ-lactols 1a, 1b, 1c, 3a, 3b, 3c,
4a, 4b and 4c were described with all details in ref.[6d] Compound
6 was prepared according to a known procedure.[28]
3-Methyl-2-phenyltetrahydrofuran-2-ol
(2):
3-Methyldihydro-
General Procedure for the Treatment of γ-Lactols with Silanes: To
a solution of the corresponding γ-lactol in dry dichloromethane
(2 mL/mmol) was added 2 equivalents of the corresponding silane,
and the mixture was cooled to Ϫ78 °C. Then, 3 equivalents of the
Lewis acid were added via syringe, and the mixture was allowed to
warm up to room temperature over 16 h. After addition of water
(2 mL/mmol) and extraction with dichloromethane (3 ϫ 20 mL),
the organic layers were dried (MgSO4) and concentrated in vacuo.
The residue was carefully distilled in a Kugelrohr oven. Boron tri-
fluoride and zinc dichloride were used as etherates. For the
amounts of starting materials used and the resulting products, see
Table 6.
2(3H)furanone (500 mg, 5.00 mmol) was dissolved in tetrahydrofu-
ran (10 mL) at Ϫ78 °C, and a phenyllithium solution (2 , 2.50 mL,
5.00 mmol) in tetrahydrofuran was added. After stirring for 4.5 h
at Ϫ78 °C and 30 min at Ϫ30 °C, sat. aqueous ammonium chloride
solution (10 mL) was added, the mixture was extracted with methyl
tert-butyl ether, and the organic layers were dried (MgSO4) and
concentrated in vacuo. The resulting crude product (1.00 g, 76:24
mixture of anomers) was used without purification. Ϫ 1H NMR
(CDCl3): δ ϭ 8.00Ϫ7.50 (m, 5 H, Ph), 4.25 (dt, J ϭ 1.5, 9.0 Hz,
0.24 H, 5-H), 4.10 (dt, J ϭ 7.0, 9.0 Hz, 0.24 H, 5-H), 3.75Ϫ3.57 (m,
2.28 H, 5-H, OH), 2.60Ϫ2.45, 2.40Ϫ2.28, 2.15Ϫ2.00, 1.90Ϫ1.85,
1.85Ϫ1.77, 1.70Ϫ1.58 (6 m, 3 H, 3-H, 4-H), 1.18/1.22 (d/d, J ϭ
7.0 Hz, 2.28 H and 0.72 H, 3-Me); several signals may be assigned
to the acyclic tautomer of 2: δ ϭ3.17 (s), 2.95 (broad s), 1.16 (s),
1.06 (d, J ϭ 6.0 Hz). Ϫ 13C NMR (CDCl3): δ ϭ 141.1Ϫ125.8 (sev-
eral s and d, Ph and C-2 of both anomers and the acyclic tauto-
mer), 60.2 (t, C-5), 37.2/34.1 (d, C-3), 36.0/30.6 (t, C-4), 17.3/15.1
(q, 3-Me).
3-Methyl-2-phenyltetrahydrofuran (16): According to the general
procedure, unpurified 2 (890 mg, 5.00 mmol) was treated with trie-
thylsilane (15) (1.16 g, 10.0 mmol) and boron trifluorideϪdiethyl
ether (2.12 g, 15.0 mmol). After extractive workup, the crude prod-
uct (1.10 g) was purified by Kugelrohr distillation (100 °C/0.37
Torr) to afford 16 (638 mg, 79%) as an almost pure mixture of the
two diastereomers (trans/cis ϭ 21:79).[16a] The two diastereomers
may be separated by HPLC (Hygrosyl 5µ column, hexane/ methyl
tert-butyl ether ϭ 98:2). Ϫ 1H NMR (CDCl3): cis-16 δ ϭ
7.33Ϫ7.18 (m, 5 H, Ph), 4.95 (d, J ϭ 6.5 Hz, 1 H, 2-H), 4.17 (dt,
J ϭ 6.0, 8.0 Hz, 1 H, 5-H), 3.91 (dt, J ϭ 7.0, 8.0 Hz, 1 H, 5-H),
trans-4,5-Dimethyl-tetrahydrofuran-2-ol (5): According to the gen-
eral procedure as described in ref.[6d] trans-4,5-dimethyldihydro-
2(3H)-furan-2-one (trans/cis ϭ 95:5, 1.14 g, 10.0 mmol) was treated
with diisobutylaluminum hydride (12.0 mmol) in toluene to give
0.766 g (66%) of 5 as a 53:47 mixture of anomers. Ϫ 1H NMR 2.50 (broad sept, J ഠ 6.5 Hz, 1 H, 3-H), 2.15 (dddd, J ϭ 6.0, 7.0,
(CDCl3): δ ϭ 5.50Ϫ5.35 (m, 1 H, 2-H), 4.52/4.50Ϫ4.40 (d/m, J ϭ
3.0 Hz, 1 H, OH), 3.76/3.53 (qd, J ϭ 6.0, 9.0 Hz/6.0, 8.5 Hz, 1 H, H, 4-H), 0.60 (d, J ϭ 7.0 Hz, 3 H, 3-Me). Ϫ trans-16 δ ϭ 7.33Ϫ7.20
5-H), 2.50Ϫ1.90, 1.75Ϫ1.50 (2 m, 2.5 H, 3-H, 4-H), 1.44 (ddd, J ϭ (m, 5 H, Ph), 4.27 (d, J ϭ 8.0 Hz, 1 H, 2-H), 4.09 (dt, J ϭ 7.0,
8.0, 12.0 Hz, 1 H, 4-H), 1.70 (dddd, J ϭ 6.0, 7.0, 8.0, 12.0 Hz, 1
Table 6. Reactions between γ-lactols and silanes in the presence of Lewis acids, according to the general procedure
γ-Lactol
[mg (mmol)]
Silane
[2 equiv.]
Lewis acid
[3 equiv.]
Product
Yield
[mg (%)]
trans/cis
b.p.
[°C/Torr]
1a [400 (3.92)]
1a [300 (2.94)]
1b [300 (1.83)]
1c [450 (3.13)]
1a [465 (4.56)]
1a [300 (2.94)]
3a [400 (3.92)]
3a [459 (4.50)]
3b [400 (2.44)]
3c [450 (3.13)]
3a [400 (3.92)]
4a [400 (3.92)]
4a [292 (2.86)]
4b [400 (2.44)]
4c [330 (2.29)]
4a [400 (3.92)]
4b [400 (2.44)]
4a [400 (3.92)]
7
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
TiCl4
11
318 (65)
195 (53)
312 (91)
388 (74)
472 (67)
225 (50)
454 (93)
388 (68)
368 (80)
440 (84)
574 (95)
400 (82)
252 (70)
365 (80)
275 (72)
562 (93)
442 (84)
533 (61)[a]
175 (91)
335 (74)
560 (99)
270 (79)
235 (79)
126 (34)
124 (76)
189 (60)
60:40
68:32
63:37
Ͼ 98:2
68:32
95:5
50/5.5
8
12a
12b
12c
13
30/6.0
8
55/0.02
60/0.01
79/6.0[29]
45/4.5
40/3.0
50/6.0
70/0.02
60/0.75
45/7.5
40/1.0
45/5.0
100/0.04
60/0.07
45/5.0
95/0.01
80/0.75
40/9.0
8
9
10
7
14
17
86:14
95:5
8
18a
18b
18c
19
8
95:5
8
Ն 98:2
Ն 98:2
20:80
40:60
47:53
88:12
83:17
77:23
90:10
86:14
74:26
71:29
69:31
95:5
10
7
21
8
22a
22b
22c
23a
23b
24
8
8
10
10
20
8
5
6
6
[160 (1.38)]
[400 (2.25)]
[500 (2.81)]
25
8
26
120/0.01
120/0.01
30/6.0
8
26
[b]
1a [276 (2.71)]
1a [197 (1.93)]
4a [300 (2.95)]
4a [143 (1.40)]
27 [600 (2.49)]
8
ZnCl2[b]
12a
14
8
ZnCl2
SnCl4
EtAlCl2
BF3·OEt2
45/4.5
8
22a
22a
12a
40:60
40:60
70:30
45/5.0
8
45/5.0
8
30/6.0
[a]
[b]
Purity according to GC and NMR ഠ83%; unknown impurities could not be removed by chromatography. Ϫ Lewis acid added at
0 °C.
3898
Eur. J. Org. Chem. 2000, 3893Ϫ3901

Synthesis of Disubstituted Tetrahydrofuran Derivatives
FULL PAPER
8.5 Hz, 1 H, 5-H), 4.02 (dt, J ϭ 4.5, 8.5 Hz, 1 H, 5-H), 2.20 (dtd, 3-Methyl-2-prenyltetrahydrofuran (13): 1H NMR (CDCl3): δ ϭ
J ϭ 4.5, 7.0, 12.0 Hz, 1 H, 4-H), 2.06 (mc, 1 H, 3-H), 1.69 (tdd, 5.30Ϫ5.10, 2.35Ϫ2.05, 1.63Ϫ1.47, 1.73, 1.64 (3 m, 2 s, 1 H, 0.68
J ϭ 8.5, 9.0, 12.0 Hz, 1 H, 4-H), 1.07 (d, J ϭ 6.5 Hz, 3 H, 3-Me). H, 1.32 H, 3 H, 3 H, 2-prenyl), 3.98Ϫ3.64 (m, 2.32 H, 2-H, cis, 5-
13C NMR (CDCl3): cis-16 δ ϭ 140.8, 127.8, 126.8, 126.2 (s, 3 d, H), 3.38 (dt, J ϭ 5.0, 7.0 Hz, 0.68 H, 2-H, trans), 2.35Ϫ2.05 (m,
Ph), 83.6 (d, C-2), 67.2 (t, C-5), 37.3 (d, C-3), 33.7 (t, C-4), 15.4 2.64 H, 4-H, 3-H, cis), 1.92Ϫ1.79 (m, 0.68 H, 3-H, trans),
Ϫ
(q, 3-Me). Ϫ trans-16 δ ϭ 142.2, 128.2, 127.4, 126.1 (s, 3 d, Ph), 1.63Ϫ1.47 (m, 0.32 H, 4-H, cis), 1.04/0.94 (d, J ϭ 7.0 Hz, 3 H, 3-
87.9 (d, C-2), 67.8 (t, C-5), 42.8 (d, C-3), 34.9 (t, C-4), 16.4 (q,
3-Me).
Me). Ϫ 13C NMR (CDCl3): δ ϭ 132.8/132.6, 120.5/120.7 (s, d, Cϭ
CH), 88.0/81.4 (d, C-2), 68.5/65.8 (t, C-5), 38.3/34.9 (d, C-3), 34.5/
33.6, 32.5 (2 t, C-4, 2-CH2), 25.8, 17.4 (2 q, ϭCMe2), 17.9/15.2 (q,
3-Me). Ϫ For further analytical data see ref.[29]
Analytical and Spectroscopic Data
3-Methyl-2-(1,2-propadienyl)tetrahydrofuran (11): 1H NMR (CDCl3):
δ ϭ 5.18/5.09 (q, J ϭ 7.0 Hz, 1 H, ϭCH), 4.85Ϫ4.75/4.42Ϫ4.36
(m, 2 H, ϭCH2), 3.97 (dt, J ϭ 5.0, 8.0 Hz, 0.6 H, 5-H, trans),
3.89Ϫ3.73 (m, 2.4 H, 2-H, 5-H), 2.35 (broad sept, J ഠ 7 Hz, 0.4
H, 3-H, cis), 2.20Ϫ2.14 (m, 1.6 H, 3-H, 4-H), 1.67Ϫ1.52 (m, 1 H,
4-H), 1.06/1.00 (d, J ϭ 7.0 Hz, 3 H, 3-Me). Ϫ 13C NMR (CDCl3):
δ ϭ 208.6/208.5, 89.3/91.4, 76.1/75.8 (s, d, t, CHϭCϭCH2), 83.5/
80.2 (d, C-2), 67.1 (t, C-5), 38.3/37.0 (d, C-3), 34.3 (t, C-4), 16.5/
14.7 (q, 3-Me). Ϫ IR (neat): ν˜ ϭ 3000Ϫ2840 cmϪ1 (CH), 1955 (Cϭ
CϭC), 1630 (CϭC). Ϫ MS (70 eV, EI): m/z (%) ϭ 124 (Mϩ, 3),
123 (Mϩ Ϫ 1, 6), 85 (Mϩ Ϫ [CH2ϭCϭCH], 24), 69 (15), 56
([cpr·ϩ], 15), 55 ([cprϩ], 20), 53 (10), 43 ([CH3COϩ], 53), 41 ([CH2ϭ
CHϪCH2ϩ], 65), 39 ([CH2ϭCϭCHϩ], 42), 32 (40). Ϫ C8H12O
(124.2): calcd. C 77.38, H 9.74; found C 76.29, H 9.78.
2-(1,1-Dimethylallyl)-3-methyltetrahydrofuran (14): 1H NMR
(CDCl3): δ ϭ 5.87Ϫ5.78, 4.97Ϫ4.91, 0.97/0.90, 0.96/0.88 (2 m, 2 s,
1 H, 2 H, 3 H, 3 H, H2CϭCHϪCMe2), 3.79Ϫ3.64/3.61Ϫ3.55 (m,
2 H, 5-H), 3.15 (d, J ϭ 5.5 Hz, 0.95 H, 2-H, trans), 2.11Ϫ1.84/
1.35Ϫ1.09 (m, 2 H, 4-H), 1.43 (broad sept, J ഠ 6.0 Hz, 0.95 H, 3-
H), 0.99/0.87 (d, J ϭ 7.5/7.0 Hz, 3 H, 3-Me). Ϫ 13C NMR (CDCl3):
δ ϭ 145.0, 111.6 (d, t, CHϭCH2), 93.1 (d, C-2), 67.3/66.0 (t, C-5),
40.6 (s, 2-CMe2), 35.5/34.7 (t, C-4), 33.9 (d, C-3), 23.3, 23.1/15.0,
20.5/13.7 (3 q, 3-Me, 2-CMe2). Ϫ IR (neat): ν˜ ϭ 3080 cmϪ1
(ϭCH), 3000Ϫ2810 (CH), 1635 (CϭC). Ϫ C10H18O (154.3): calcd.
C 77.86, H 11.77, found C 77.26, H 11.80.
4-Methyl-2-(1,2-propadienyl)tetrahydrofuran (17): 1H NMR
(CDCl3): δ ϭ 5.21, 5.05Ϫ4.09 (mc, m, 3 H, CHϭCϭCH2),
4.57Ϫ4.49/4.45Ϫ4.38 (m, 1 H, 2-H), 4.02/3.92 (dd, J ϭ 7.0, 8.0/7.5,
8.0 Hz, 1 H, 5-H), 3.31/3.38 (dd/t, J ϭ 8.0, 7.0/8.0 Hz, 1 H, 5-H),
2.43Ϫ2.31 (m, 1 H, 4-H), 1.95, 1.43 (ddd, td, J ϭ 5.5, 7.5, 12.5 and
8.0, 12.5 Hz, 0.86 H each, 3-H, trans), 1.05 (d, J ϭ 7.0 Hz, 2.58 H,
4-Me). Ϫ 13C NMR (CDCl3): δ ϭ 207.7, 93.0/92.8, 76.6 (s, d, t,
CHϭCϭCH2), 76.4/76.6, (d, C-2), 74.9/74.6 (t, C-5), 39.8/40.7 (t,
C-3), 33.1/34.4 (d, C-4), 17.9/17.5 (q, 4-Me). Ϫ IR (neat): ν˜ ϭ 3080
cmϪ1 (ϭCH), 2980Ϫ2800 (CH), 1955 (CϭCϭC), 1635 (CϭC). Ϫ
MS (70 eV, EI): m/z (%) ϭ 124 (Mϩ, 3), 85 (Mϩ Ϫ [CH2ϭCϭCH],
100), 79 (Mϩ Ϫ [C2H5O], 12), 77 (10), 67 (19), 57 (41), 55 ([cprϩ],
14), 43 ([CH3COϩ], 23), 41 ([CH2ϭCHϪCH2ϩ], 71), 39 ([CH2ϭ
CϭCHϩ], 14).
2-Allyl-3-methyltetrahydrofuran (12a): 1H NMR (CDCl3): δ ϭ
5.92Ϫ5.80, 5.15Ϫ5.04, 2.35Ϫ1.48 (3 m, 1 H, 2 H, 3.68 H, 2-allyl,
4-H), 3.93 (td, J ϭ 6.5, 8.0 Hz, 0.32 H, 5-H, cis), 3.88Ϫ3.76 (m,
1.68 H, 5-H, trans, 2-H, cis), 3.73 (td, J ϭ 6.0, 8.0 Hz, 0.32 H, 5-
H, cis), 3.38 (dt, J ϭ 4.5, 7.5 Hz, 0.68 H, 2-H, trans), 1.55Ϫ1.35
(m, 0.32 H, 4-H), 1.78/1.46 (broad sept, J ഠ 7.0/8.0 Hz, 0.32/0.68
H, 3-H). Ϫ 13C NMR (CDCl3): δ ϭ 135.2/135.6 (d, ϭCH), 116.4/
116.2 (t, ϭCH2), 85.0/80.9 (d, C-2), 66.7/66.0 (t, C-5), 38.5/35.1 (t,
2-CH2), 38.3/35.2 (d, C-3), 34.6/33.8 (t, C-4), 17.1/14.2 (q, 3-Me).
Ϫ IR (neat): ν˜ ϭ 3035 cmϪ1 (ϭCH), 3000Ϫ2800 (CH), 1640
(CϭC). Ϫ C8H14O (126.2): calcd. C 76.14, H 11.18; found C 75.74,
H 11.37.
2-Allyl-3-phenyltetrahydrofuran (12b): 1H NMR (CDCl3): δ ϭ
2-Allyl-4-methyltetrahydrofuran (18a): 1H NMR (CDCl3): δ ϭ
5.90Ϫ5.72, 5.15Ϫ5.00, 2.40Ϫ2.15 (3 m, 1 H, 2 H, 2 H, 2-allyl), 4.03
(qd, J ϭ 6.5, 8.5 Hz, 0.95 H, 2-H, trans), 4.00/3.88 (t/dd, J ϭ 8.0/
8.0, 7.0 Hz, 1 H, 5-H), 3.26/3.36 (dd, J ϭ 7.0, 8.0 Hz, 1 H, 5-H),
2.26 (mc, 1 H, 4-H), 1.74, 1.57 (2 ddd, J ϭ 6.5, 8.5, 12.5 and 6.5,
7.0, 12.5 Hz, 0.95 each, 3-H, trans), 1.02/1.04 (d, J ϭ 7.0 Hz, 3 H,
4-Me). Ϫ 13C NMR (CDCl3): δ ϭ 135.1, 116.7 (d, t, CHϭCH2),
78.0/79.4 (d, C-2), 75.1/74.6 (t, C-5), 40.5/40.4, 39.0 (2 t, 2-CH2, C-
3), 33.2/34.3 (d, C-4), 18.0/17.9 (q, 4-Me). Ϫ IR (neat): ν˜ ϭ 3035
cmϪ1 (ϭCH), 3000Ϫ2800 (CH), 1640 (CϭC). Ϫ C8H14O (126.2):
calcd. C 76.14, H 11.18; found C 75.12, H 11.38.
7.32Ϫ7.17/7.31Ϫ7.14 (m,
5 H, 3-Ph), 5.88Ϫ5.75/5.79Ϫ5.65,
5.10Ϫ5.01/4.98Ϫ4.88, 2.39Ϫ2.17/2.00Ϫ1.78 (3 m, 1 H, 2 H, 2 H,
2-allyl), 4.18 (dt, J ϭ 5.0, 8.5 Hz, 0.37 H, 5-H, cis), 4.03Ϫ3.94 (m,
1.26 H, 5-H, trans), 3.98 (dt, J ϭ 5.5, 8.5 Hz, 0.37 H, 2-H, cis),
3.87 (dt, J ϭ 7.0, 8.5 Hz, 0.37 H, 5-H, cis), 3.86 (ddd, J ϭ 4.0, 7.0,
8.0 Hz, 0.63 H, 2-H, trans), 3.35 (td, J ϭ 5.5, 8.5 Hz, 0.37 H, 3-H,
cis), 2.94 (q, J ഠ 8.5 Hz, 0.63 H, 3-H, trans), 2.40 (dtd, J ϭ 5.0,
8.5, 13.0 Hz, 0.37 H, 4-H, cis), 2.39Ϫ2.30 (m, 0.63 H, 4-H, trans),
2.14Ϫ1.98 (m, 0.63 H, 4-H, trans), 2.12 (dddd, J ϭ 5.5, 7.0, 8.5,
13.0 Hz, 0.37 H, 4-H, cis). Ϫ 13C NMR (CDCl3): δ ϭ 141.7/141.9,
128.6/128.5, 127.7/128.2, 126.6/126.4 (s, 3 d, Ph), 134.9/135.4 (d,
ϭCH), 116.9 (t, ϭCH2), 85.3/82.1 (d, C-2), 67.7/67.0 (t, C-5), 50.3/
47.4 (d, C-3), 35.9 (t, 2-CH2), 35.4/33.2 (t, C-4). Ϫ IR (neat): ν˜ ϭ
3080, 3020 cmϪ1 (ϭCH), 2980Ϫ2920 (CH), 1640, 1600, 1490 (Cϭ
C). Ϫ C13H16O (188.3): calcd. C 82.94, H 8.57; found C 82.08,
H 8.64.
2-Allyl-4-phenyltetrahydrofuran (18b): 1H NMR (CDCl3): δ ϭ
7.28Ϫ7.07 (m, 5 H, Ph), 5.82Ϫ5.72, 5.10Ϫ4.89, 2.39Ϫ2.18 (3 m, 1
H, 2 H, 2 H, 2-allyl), 4.58Ϫ4.21 (m, 0.1 H, 2-H, 5-H, cis), 4.15 (dd,
J ϭ 7.5, 8.5 Hz. 0.95 H, 5-H, trans), 4.13 (quint, J ϭ 6.5 Hz, 0. 95
H, 2-H, trans), 3.72 (t, J ϭ 8.5 Hz, 0.05 H, 5-H, cis), 3.63 (dd, J ϭ
8.0, 8.5 Hz, 0.95 H, 5-H, trans), 3.33 (broad quint, J ഠ 8.0 Hz,
0.95 H, 4-H, trans), 2.83Ϫ2.74 (m, 0.05 H, 4-H, cis), 2.58Ϫ2.49,
2.08Ϫ1.91, 1.70Ϫ1.56 (3 m, 0.05 H, 1.9 H, 0.05 H, 3-H). Ϫ 13C
NMR (CDCl3): δ ϭ 142.5/142.0, 128.5, 127.2, 126.5/126.6 (s, 3 d,
Ph), 134.8, 117.0 (d, t, CHϭCH2), 78.7/79.6 (d, C-2), 74.6/74.3 (t,
C-5), 44.5/45.4 (d, C-4), 40.4, 39.0/40.0 (2 t, 2-CH2, C-3). Ϫ IR
(neat): ν˜ ϭ 3080, 3020 cmϪ1 (ϭCH), 3000Ϫ2820 (CH), 1640, 1490
(CϭC). Ϫ C13H16O (188.3): calcd. C 82.94, H 8.57, found C 82.37,
H 8.58.
2-Allyl-3-tert-butyltetrahydrofuran (12c): 1H NMR (CDCl3): δ ϭ
5.92Ϫ5.78, 5.11Ϫ5.01, 2.34Ϫ2.17 (3 m, 1 H, 2 H, 2 H, 2-allyl), 3.85
(ddd, J ϭ 4.5, 5.5, 8.0 Hz, 1 H, 2-H), 3.80Ϫ3.69 (m, 2 H, 5-H),
1.91Ϫ1.80, 1.80Ϫ1.69 (2 m, 1 H, 2 H, 4-H, 3-H), 0.89 (s, 9 H, tBu).
Ϫ
13C NMR (CDCl3): δ ϭ 135.6 (d, ϭCH), 116.4 (t, ϭCH2), 79.6
(d, C-2), 67.2 (t, C-5), 54.0 (d, C-3), 40.9 (t, 2-CH2), 31.9, 27.8/29.5
(s, q, 3-tBu). Ϫ IR (neat): ν˜ ϭ 3060 cmϪ1 (ϭCH), 2980Ϫ2820
(CH), 1635 (CϭC). Ϫ C11H20O (168.3): calcd. C 78.51, H 11.98;
found C 78.80, H 11.54.
Eur. J. Org. Chem. 2000, 3893Ϫ3901
3899

A. Schmitt, H.-U. Reißig
2-Allyl-4-tert-butyltetrahydrofuran (18c): 1H NMR (CDCl3): δ ϭ 86.7/87.2 (d, C-5), 78.7/78.0 (d, C-2), 40.1/40.8 (t, 2-CH2), 33.9/
FULL PAPER
5.90Ϫ5.70, 5.15Ϫ5.00, 2.45Ϫ2.16, 2.11 (3 m, ddd, J ϭ 8.0, 9.5,
33.3, 25.5/24.7 (s, q, 5-tBu), 31.9/30.5, 26.8/25.7 (2 t, C-3, C-4). Ϫ
17.5 Hz, 1 H, 2 H, 1 H, 1 H, 2-allyl), 3.95Ϫ3.85 (m, 2 H, 2-H, 5- IR (neat): ν˜ ϭ 3060 cmϪ1 (ϭCH), 3000Ϫ2860 (CH), 1635
H), 3.46 (t, J ϭ 9.0. Hz, 1 H, 5-H), 2.45Ϫ2.16 (m, 1 H, 4-H), 1.76, (CϭC). Ϫ C11H20O (168.3): calcd. C 78.51, H 11.98; found C
1.54 (td, ddd, J ϭ 7.5, 12.5 and 6.5, 9.5, 12.5 Hz, 1 H each, 3-H), 79.00, H 11.72.
0.87 (s, 9 H, tBu). Ϫ 13C NMR (CDCl3): δ ϭ 134.8, 116.4 (d, t,
2-(1,1-Dimethylallyl)-5-methyltetrahydrofuran (23a): 1H NMR
CHϭCH2), 78.5 (d, C-2), 69.1 (t, C-5), 49.0 (d, C-4), 40.4, 32.1 (2
t, 2-CH2, C-3), 31.0, 27.4 (s, q, 4-tBu). Ϫ IR (neat): ν˜ ϭ 3060 cmϪ1
(ϭCH), 3000Ϫ2800 (CH), 1635 (CϭC). Ϫ C11H20O (168.3): calcd.
C 78.51, 11.98; found C 78.76, H 11.59.
(CDCl3): δ ϭ 5.97Ϫ5.86, 5.05Ϫ4.98, 1.05, 0.99 (2 m, 2 s, 1 H, 2
H, 3 H, 3 H, CH2ϭCHϪMe2), 4.06Ϫ3.94 (m, 1 H, 5-H), 3.80/3.65
(dd/t, J ϭ 6.5, 9. 0 and 7.5 Hz, 0.83/0.17 H, 2-H), 2.02Ϫ1.26 (m, 4
H, 3-H, 4-H), 1.23/1.22 (d, J ϭ 6.0 Hz, 3 H, 5-Me). Ϫ 13C NMR
(CDCl3): δ ϭ 145.1/145.3, 112.0/111.8 (d, t, CHϭCH2), 86.0/86.5
(d, C-5), 75.8/75.3 (d, C-2) 40.8/40.0 (s, 2-CMe2), 34.5/34.2, 27.9/
26.7 (2 t, C-3, C-4), 23.6, 23.0, 21.2/21.0 (3 q, 2-CMe2, 5-Me). Ϫ
IR (neat): ν˜ ϭ 3080 cmϪ1 (ϭCH), 3000Ϫ2800 (CH), 1635
(CϭC). Ϫ C10H18O (154.3): calcd. C 77.86, H 11.77; found C
77.96, H 11.99.
2-(1,1-Dimethylallyl)-4-methyltetrahydrofuran (19): 1H NMR
(CDCl3): δ ϭ 5.94Ϫ5.84, 5.05Ϫ4.99, 1.04, 1.00 (2 m, 2 s, l H, 2 H,
3 H, 3 H, H2CϭCHϪCMe2), 3.94 (dd, J ϭ 6.5, 8.0 Hz, 1 H, 5-H),
3.77 (dd, J ϭ 7.5, 8.0 Hz, 1 H, 2-H), 3.31 (dd, J ϭ 6.5, 8.0 Hz, 1
H, 5-H), 2.28Ϫ2.17 (m, 1 H, 4-H), 1.82, 1.42 (td, ddd, J ϭ 8.0,
12.5 and 5.5, 7.5, 12.5 Hz, 1 H each, 3-H), 1.03 (d, J ϭ 6.5 Hz, 3
H, 4-Me). Ϫ 13C NMR (CDCl3): δ ϭ 145.0, 112.1 (d, t, CHϭCH2),
85.7 (d, C-2), 75.4 (t, C-5), 40.4 (s, 2-CMe2), 35.0 (t, C-3), 33.7 (d,
C-4), 23.7, 22.7 (2 q, 2-CMe2), 16.0 (q, 4-Me). Ϫ IR (neat): ν˜ ϭ
3080 cmϪ1 (ϭCH), 2980Ϫ2800 (CH), l635 (CϭC). Ϫ C10H18O
(154.3): calcd. C 77.86, H 11.77; found C 77.59, H 11.97.
2-(1,1-Dimethylallyl)-5-phenyltetrahydrofuran (23b): 1H NMR
(CDCl3): δ ϭ 7.41Ϫ7.18/7.33Ϫ7.13 (m, 5 H, Ph), 5.98Ϫ5.89,
5.03Ϫ4.91, 1.04/1.02, 0.99/0.98 (2 m, 2 s, 1 H, 2 H, 3 H, 3 H, CH2ϭ
CHϪCMe2), 4.84/4.66 (dd, J ϭ 6.0, 8.5 and 6.5, 9.0 Hz, 0.77/0.23
H, 5-H), 3.92/3.68 (dd/t, J ϭ 6.5, 9.0 and 7.0 Hz, 0.77/0.23 H, 2-
H), 2.35Ϫ2.20/2.22Ϫ2.12 (m, 1 H, 4-H), 2.00Ϫ1.60, 1.53Ϫ1.41 (2
m, 2.77 H, 0.23H, 3-H, 4-H). Ϫ 13C NMR (CDCl3): δ ϭ 143.9/
142.9, 128.2/127.1, 127.0/125.8, 125.4/125.5 (s, 3 d, Ph), 144.9/
145.0, 117.4 (d, t, CHϭCH2), 87.1/86.4 (d, C-5), 81.2/80.8 (d, C-
2), 40.1/40.3 (s, 2-CMe2), 35.8/34.8, 28.0/27.0 (2 t, C-3, C-4), 23.9/
23.6, 23.0/23.3 (2 q, 2-CMe2). Ϫ IR (neat): ν˜ ϭ 3080, 3020 cmϪ1
(ϭCH), 2990Ϫ2840 (CH), 1635, 1595, 1480 (CϭC). Ϫ C15H20O
(216.3): calcd. C 83.29, H 9.32; found C 82.64, H 9.46.
5-Methyl-2-(1,2-propadienyl)tetrahydrofuran (21): 1H NMR
(CDCl3): δ ϭ 5.21/5.23, 4.88Ϫ4.72 (q, J ϭ 6.5 Hz, m, 1 H, 2 H,
CHϭCϭCH2), 4.60Ϫ4.50/4.43Ϫ4.34 (m, 0.2/0.8 H, 2-H),
4.20Ϫ4.08/4.06Ϫ3.95 (m, 0.2/0.8 H, 5-H), 2.15Ϫ1.95 (m, 2 H, 3-
H), 1.85Ϫ1.70, 1.60Ϫ1.45 (2 m, 1.6 H, 4-H, cis), 1.45Ϫ1.35 (m, 0.4
H, 4-H, trans). Ϫ 13C NMR (CDCl3): δ ϭ 208.0, 93.0/93.1, 76.4 (s,
d, t, CHϭCϭCH2), 77.1, 75.8 (2 d, C-2, C-5), 32.7, 32.2 (2 t, C-3,
C-4), 21.2/21.3 (q, 5-Me). Ϫ IR (neat): ν˜ ϭ 3080 cmϪ1 (ϭCH),
3000Ϫ2800 (CH), 1955 (CϭCϭC), 1640 (CϭC). Ϫ MS (70 eV, EI):
m/z (%) ϭ 124 (Mϩ, 14), 109 (Mϩ Ϫ CH3, 8), 85 (Mϩ Ϫ [CH2ϭ
CϭCH], 100), 83 (10), 79 (Mϩ Ϫ [C2H5O], 20), 67 (20), 57 (25), 55
([cprϩ], 14), 43 ([CH3COϩ], 44), 41 ([CH2ϭCHϪCH2ϩ], 51), 39
([CH2ϭCϭCHϩ], 32).
5-Methyl-2-(3-methyl-1-trimethylsilyl-1,2-butadienyl)tetrahydro-
furan (24): 1H NMR (CDCl3): δ ϭ 4.51/4.31 (t, J ϭ 6.0 Hz, 0.1 H,
0.9 H, 2-H), 4.15Ϫ4.00/3.95Ϫ3.80 (m, 0.1 H, 0.9 H, 5-H),
1.60Ϫ1.59 (m, 4 H, 3-H, 4-H), 1.56, 1.55 (2 s, 6 H, Me), 1.13 (d,
J ϭ 6.0 Hz, 3 H, 5-Me), 0.00 (s, 9 H, SiMe3). Ϫ 13C NMR (CDCl3):
δ ϭ 203.6/203.7 (s, ϭCϭ), 98.8 (s, ϭCSiMe3), 90.4 (s, ϭCMe2),
78.9/77.9, 75.2/75.9 (2 d, C-2, C-5), 31.4/32.2, 31.2/32.1 (2 t, C-3,
C-4), 23.5/26.5, 21.1/20.9, 19.6 (3 q, Me), 0.6/Ϫ0.3 (q, SiMe3). Ϫ IR
(neat): ν˜ ϭ 3000Ϫ2800 cmϪ1 (CH), 1920 (CϭCϭC), 1625 (CϭC).
2-Allyl-5-methyltetrahydrofuran (22a): 1H NMR (CDCl3): δ ϭ 5.82,
5.07, 2.42Ϫ2.16 (mc, mc, m, 1 H, 2 H, 2 H, 2-allyl), 4.09 (mc, 1.4
H, 5-H, 2-H, trans), 3.91 (mc, 0.6 H, 2-H, cis), 2.08Ϫ1.87, 1.61 (m,
mc, 3 H, 3-H, 4-H), 1.24/1.21 (d, J ϭ 6.0 Hz, 3 H, 5-Me). Ϫ 13C
NMR (CDCl3): δ ϭ 135.1, 116.7 (d, t, CHϭCH2), 77.9/78.7 (d, C-
2), 74.8/75.4 (d, C-5), 40.5 (t, 2-CH2), 33.9/32.8, 31.7/30.7 (2 t, C-
3, C-4), 21.4 (q, 5-Me). Ϫ IR (neat): ν˜ ϭ 3070 cmϪ1 (ϭCH),
3000Ϫ2800 (CH), 1640 (CϭC). Ϫ C8H14O (126.2): calcd. C 76.14,
H 11.18; found C 75.62, H 11.39.
2-Allyl-4,5-dimethyltetrahydrofuran (25): 1H NMR ([D6]acetone):
δ ϭ 5.88Ϫ5.70, 5.58Ϫ4.93 (2 m, 1 H, 2 H, CHϭCH2), 4.05Ϫ3.90
(m, 0.14 H, 2-H), 3.87 (qd, J ϭ 6.0, 7.5 Hz, 0.86 H, 2-H), 3.43/3.30
(qd/qd, J ϭ 6.0, 9.0 and 6.0, 8.0 Hz, 0.14 H and 0.86 H, 5-H),
2.30Ϫ2.10 (m, 2 H, 5-CH2), 1.80Ϫ1.60 (m, 0.42 H, 3-H, 4-H), 1.78
(ddd, J ϭ 5.5, 8.5, 11.5 Hz, 0.86 H, 3-H), 1.73Ϫ1.58 (m, 0.86 H,
4-H), 1.55 (td, J ϭ 8.5, 11.5 Hz, 0.86 H, 3-H), 1.13/1.10 (d/d, J ϭ
6.0 Hz, 2.58 H and 0.42 H, 5-Me), 0.97/0.95 (d/d, J ϭ 6.0 and
6.5 Hz, 0.42 H and 2.58 H, 4-Me). Ϫ 13C NMR (CDCl3): δ ϭ
134.9, 116.6 (d, t, CHϭCH2), 82.0/80.9, 77.0/77.4 (2 d, C-2, C-5),
40.9, 39.1 (2 t, C-3, C-6), 40.5/41.9 (d, C-4), 19.2, 16.7/16.0 (2 q, 4-
Me, 5-Me). Ϫ IR (neat): ν˜ ϭ 3055 cmϪ1 (ϭCH), 2980Ϫ2810 (CH),
1640 (CϭC), 1080Ϫ1040 (CϪO). Ϫ MS (70 eV, EI): m/z (%) ϭ 99
(Mϩ Ϫ[C3H5], 100), 76 (42) 55 (66), 43 (83), 41 ([C3H5ϩ], 31), 39
(19). Ϫ C9H16O (140.2): calcd. C 77.09, H 11.50; found C 75.25,
H 11.51.
2-Allyl-5-phenyltetrahydrofuran (22b): 1H NMR (CDCl3): δ ϭ
7.39Ϫ7.16/7.43Ϫ7.21 (m,
5
H, Ph), 5.93Ϫ5.79/5.98Ϫ5.85,
5.15Ϫ5.03/5.19Ϫ5.07, 2.57Ϫ2.49, 2.47Ϫ2.25 (4 m, 1 H, 2 H, 1 H,
l H, 2-allyl), 4.99/4.88 (dd/t, J ϭ 6.5, 8.0 and 7.0 Hz, 0.47/0.53 H,
5-H), 4.29Ϫ4.21/4.11 (m/broad quint, J ഠ 6.5 Hz, 0.47/0.53 H, 2-
H), 2.38Ϫ2.25, 2.32Ϫ2.20, 2.12Ϫ2.00, 1.89Ϫ1.62 (4 m, 4 H, 3-H,
4-H). Ϫ 13C NMR (CDCl3): δ ϭ 142.4/143.0, 127.4/128.3, 125.7/
127.3, 125.8 (s, 3 d, Ph), 134.4, 117.3 (d, t, CHϭCH2), 81.3/80.7
(d, C-2), 79.5/79.6 (d, C-5), 40.0/40.1 (t, 2-CH2), 35.1/34.2, 31.5/
30.4 (2 t, C-3, C-4). Ϫ IR (neat): ν˜ ϭ 3080, 3020 cmϪ1 (ϭCH),
3000Ϫ2800 (CH), 1640, 1600, 1480 (CϭC). Ϫ C13H16O (188.3):
calcd. C 82.94, H 8.57; found C 82.56, H 8.59.
2-Allyl-5-tert-butyltetrahydrofuran (22c): 1H NMR (CDCl3): δ ϭ
5.96Ϫ5.74, 5.29Ϫ4.90, 2.42Ϫ2.33, 2.23Ϫ2.13 (4 m, 1 H, 2 H, 1 H,
1 H, 2-allyl), 3.97Ϫ3.88 (m, 1 H, 2-H), 3.65/3.53 (dd, J ϭ 6.0, 9.0
and 7.0, 8.0 Hz, 0.88/0.12 H, 5-H), 2.00Ϫ1.90, 1.87Ϫ1.78,
4-Allyl-3-phenyl-γ-butyrolactone (26): For 1H NMR spectroscopic
data and other analytical data see ref.[30]
Ϫ
13C NMR (CDCl3):
δ ϭ 175.3/176.4 (s, C-2), 132.1/132.7, 118.7/117.8 (d, t, CHϭCH2),
1.70Ϫ1.41 (3 m, 1 H, 1 H, 2 H, 3-H, 4-H), 0.88/0.84 (s, 7.9/1.1 H, 138.7/137.6, 128.9/128.6, 127.6/127.7, 127.0 (s, 3 d, Ph), 85.5/83.1
tBu). Ϫ 13C NMR (CDCl3): δ ϭ 135.0, 116.2 (d, t, CHϭCH2),
(d, C-5), 46.1/44.1 (d, C-4), 37.4/35.5, 37.1/35.4 (2 t, C-3, 5-CH2).
3900
Eur. J. Org. Chem. 2000, 3893Ϫ3901

Synthesis of Disubstituted Tetrahydrofuran Derivatives
FULL PAPER
only be obtained as an enriched mixture with 5 as a minor
component. However, its reaction with 8 in the presence of
boron trifluoride is highly stereoselective, giving only the 2,4-
trans-2,5-trans diastereomer (in admixture with the two isomers
of 25). In this case, the two methyl groups cooperate, resulting
in excellent diastereoselectivity.
A. Schmitt, H.-U. Reißig, manuscript in preparation; for the
preparation of (phenylselenenyl)tetrahydrofuran derivatives see
this ref.
Unfortunately, the acetates derived from γ-lactols used in this
study were rather unstable. Thus, we could not investigate ste-
reoselectivity with these substrates.
S. E. Denmark, T. W. Willson, J. Am. Chem. Soc. 1989, 111,
3475Ϫ3476.
Acknowledgments
Generous support by the Deutsche Forschungsgemeinschaft and
the Fonds der Chemischen Industrie is most gratefully acknow-
ledged. We thank Dr. R. Zimmer for his help during the prepara-
tion of this manuscript.
[12]
[13]
[1]
A. Schmitt, Dissertation, Technische Hochschule Darmstadt,
1992.
[2]
Selected examples: Y. Guindon, M. Therien, Y. Girard, C.
Yoakim, J. Org. Chem. 1987, 52, 1680Ϫ1686. Ϫ D. J. Goldsch-
mith, E. Kennedy, R. G. Cambell, J. Org. Chem. 1975, 40,
3571Ϫ3574. Ϫ H. Frauenrath, T. Phillipps, Liebigs Ann. Chem.
1985, 1951Ϫ1961.
[14]
[15]
S. E. Denmark, T. W. Willson, N. G. Almstead, J. Am. Chem.
Soc. 1989, 111, 9258Ϫ9260. Ϫ S. E. Denmark, N. G. Almstead,
J. Org. Chem. 1991, 56, 6458Ϫ6467. Ϫ S. E. Denmark, N. G.
Almstead, J. Am. Chem. Soc. 1991, 113, 8089Ϫ8110. Ϫ T. Sam-
makia, R. S. Smith, J. Org. Chem. 1992, 57, 2997Ϫ3000. Ϫ H.
Matsutani, S. Ichikawa, J. Yaruva, T. Kusumoto, T. Hiyama, J.
Am. Chem. Soc. 1997, 119, 4541Ϫ4542.
[3]
Review: K. Mori, Tetrahedron 1989, 45, 3233Ϫ3298.
[4]
J. W. Westley, Polyether Antibiotics: Naturally Occurring Acid
Ionophores, Marcel Dekker, New York, 1982.
[5]
Reviews: T. L. B. Boivin, Tetrahedron 1987, 43, 3309Ϫ3362. Ϫ
G. Cardillo, M. Orena, Tetrahedron 1990, 46, 3321Ϫ3408. Ϫ
`
J.-C. Harmange, B. Figadere, Tetrahedron: Asymmetry 1993, 4,
[16]
[17]
Cyclizations under mild conditions are known for benzylic 1,4-
1711Ϫ1754. Ϫ U. Koert, Synthesis 1994, 115Ϫ132. Ϫ M. C.
[16a]
diols:
G. Dana, J.-P. Gireault, Bull. Soc. Chim. Fr. 1972,
Elliot, J. Chem. Soc., Perkin Trans. 1 1998, 4175Ϫ4200.
[16b]
1650Ϫ1656. Ϫ
M. Lj. Mihailovic, R. I. Mamuzic, Lj.
[6] [6a]
C. Brückner, H. Lorey, H.-U. Reißig, Angew. Chem. 1986,
Zigic-Mamuzic, J. Bosniak, Z. Cekovic, Tetrahedron 1967, 23,
98, 559Ϫ560; Angew. Chem. Int. Ed. Engl. 1986, 25, 556Ϫ557.
Ϫ
[16c]
215Ϫ226. Ϫ
1970, 209Ϫ224.
M. Lj. Mihailovic, Z. Cekovic, Synthesis
[6b]
C. Brückner, H. Holzinger, H.-U. Reißig, J. Org. Chem.
[6c]
1988, 53, 2450Ϫ2456. Ϫ
H.-U. Reißig, H. Holzinger, G.
Glomsda, Tetrahedron 1989, 45, 3139Ϫ3150. Ϫ [6d] A. Schmitt,
H. Angert, T. Kunz, H.-U. Reißig, Tetrahedron 1992, 48,
5681Ϫ5690. Ϫ H.-U. Reißig, H. Angert, T. Kunz, A. Janowitz,
G. Handke, E. Bruce-Adjei, J. Org. Chem. 1993, 58,
6280Ϫ6285; see these references for other examples; see ref.[8a]
for low diastereoselectivities of related examples.
H.-U. Reißig, Chem. Ber. 1995, 128, 871Ϫ876. Ϫ For recent
[6e]
contributions of other groups:
Y. Nishiyama, T. Katoh, K.
Deguchi, Y. Morimoto, K. Itoh, J. Org. Chem. 1997, 62,
9339Ϫ9341. Ϫ [6f] E. Alonso, D. J. Ramon, M. Yus, Tetrahedron
´
[18]
[19]
B. Fuchs, Top. Stereochem. 1978, 10, 1Ϫ94.
1997, 53, 2641Ϫ2652.
[7]
[8]
K. E. Gilbert, J. J. Gajewski, program MMX, version 89.000,
MM2 & P1, QCPE 395 ϩ318 (P1) in program package PCMO-
DEL, Serena Software.
S. A. Barbirad, Y. Wang, Y. Kishi, J. Org. Chem. 1987, 52,
1370Ϫ1372. Ϫ Review: R. R. Schmidt in Comprehensive Or-
ganic Synthesis (B. M. Trost, E. Winterfeldt, Eds.), Pergamon
Press, Oxford 1991, Vol. 6, chpt. 1.2. Ϫ M. H. D. Postema,
Tetrahedron 1992, 48, 8545Ϫ8599.
[20]
[21]
[22]
H. J. Lindner, program PIMM, version 9/1987. Ϫ A. E. Smith,
H. J. Lindner, J. Comput.-Aided Mol. Des. 1991, 5, 235Ϫ262.
G. Rauhut, A. Schmitt, unpublished ab initio 6Ϫ31G* calcula-
tions, Technische Hochschule Darmstadt, 1991.
For reactions of free γ-lactols see: [8a] K. Tomooka, K. Matsuz-
awa, K. Suzuki, G.-i. Tsuchihashi, Tetrahedron Lett. 1987, 28,
6339Ϫ6342. Ϫ [8b] P. A. Wender, F. E. McDonald, J. Am. Chem.
The problem may be regarded as a CurtinϪHammett system
with “feed in” mechanism, since the low activation barrier be-
tween the two conformers guarantees that their equilibrium is
always established. For a discussion see: J. I. Seeman, Chem.
Rev. 1983, 83, 83Ϫ134.
[8c]
Soc. 1990, 112, 4956Ϫ4958. Ϫ
G. H. Posner, S. R. Haines,
Tetrahedron Lett. 1985, 26, 1823Ϫ1826; see this ref. for reac-
tions of activated or protected γ-lactols. Ϫ P. de Shong, G. A.
Slough, V. Elango, G. L. Trainor, J. Am. Chem. Soc. 1985, 107,
[8d]
[23]
7788Ϫ7790. Ϫ
K. Utimoto, T. Horiie, Tetrahedron Lett.
For recent reactions of related disubstituted oxocarbenium ions
see: [23a] J. Ishihara, J. Miyakawa, T. Tsujimoto, A. Murai, Syn-
1982, 23, 237Ϫ238. Ϫ K. Utimoto, Y. Wakabayshi, T. Horiie,
[23b]
M. Inoue, Y. Shishiyama, M. Obayashi, H. Nozaki, Tetrahed-
lett 1997, 1417Ϫ1419. Ϫ
J. T. Shaw, K. A. Woerpel, Tetra-
[8e]
ron 1983, 39, 967Ϫ973. Ϫ
M. C. Pirrung, C. V. de Amicis,
hedron 1999, 55, 8747Ϫ8756.
P. Bernardelli, L. A. Paquette, J. Org. Chem. 1997, 62,
[8f]
Tetrahedron Lett. 1988, 29, 159Ϫ162. Ϫ
K. Hori, K. No-
[24]
[25]
mura, H. Kazuno, E. Yoshii, Chem. Pharm. Bull. 1990, 38,
1778Ϫ1780. Ϫ K. Hori, K. Nomura, N. Hikage, E. Yoshii,
Chem. Pharm. Bull. 1990, 38, 1781Ϫ1783. Ϫ K. Hori, N. Hik-
age, A. Inagaki, S. Mori, K. Nomura, E. Yoshii, J. Org. Chem.
1992, 57, 2888Ϫ2902. Ϫ [8g] S. Murata, R. Noyori, Tetrahedron
8284Ϫ8285.
T. Kunz, H.-U. Reißig, Angew. Chem. 1988, 100, 297Ϫ298; An-
gew. Chem. Int. Ed. Engl. 1988, 27, 268Ϫ270. Ϫ T. Kunz, A.
Janowitz, H.-U. Reißig, Chem. Ber. 1989, 122, 2165Ϫ2175.
C. H. Larsen, B. H. Ridgway, J. T. Shaw, K. A. Woerpel, J.
Am. Chem. Soc. 1999, 121, 12208Ϫ12209. See this ref. for a
recent study proposing a favoured “inside” attack on cyclic ox-
ocarbenium ions. The interpretation given is almost identical
with the model proposed by us in our preliminary publication[9]
and in this report.
A. Schmitt, H.-U. Reißig, manuscript submitted to Eur. J.
Org. Chem.
J. J. Bourguignon, C. G. Wermuth, J. Org. Chem. 1981, 46,
4889Ϫ4894. Ϫ J. J. Bourguignon, A. Schoenfelder, M. Schmitt,
C. G. Wermuth, U. Hechler, B. Charlier, M. Maitre, J. Med.
Chem. 1988, 31, 893Ϫ897.
[26]
Lett. 1982, 23, 2601Ϫ2602. Ϫ S. Murata, M. Suzuki, R.
[8h]
Noyori, Tetrahedron 1988, 44, 4259Ϫ4275. Ϫ
M. T. Reetz,
Angew. Chem. 1982, 94, 97Ϫ109; Angew. Chem. Int. Ed. Engl.
1982, 21, 96Ϫ108. Ϫ M. T. Reetz, H. Müller-Starke, Liebigs
[8i]
Ann. Chem. 1983, 1726Ϫ1738. Ϫ
D. S. Brown, M. Bruno,
R. J. Davenport, S. V. Ley, Tetrahedron 1989, 45, 4292Ϫ4308.
Ϫ D. S. Brown, P. Charreau, S. V. Ley, Synlett 1990, 749Ϫ750;
see these refs. for related reactions of substituted 2-phenylsul-
fonyltetrahydrofurans.
[27]
[28]
[9]
A. Schmitt, H.-U. Reißig, Synlett 1990, 40Ϫ42; see this ref. for
preliminary communication of our results.
[10] [10a] G. A. Kraus, K. A. Frazier, B. D. Roth, M. J. Taschner, K.
[29]
[30]
[10b]
C. F. Seidel, D. Felix, A. Eschenmoser, K. Biemann, E. Palluy,
M. Stoll, Helv. Chim. Acta 1961, 44, 598Ϫ606.
Neuenschwander, J. Org. Chem. 1981, 46, 2417Ϫ2419. Ϫ
G. A. Kraus, M. T. Molina, J. A. Walling J. Chem. Soc., Chem.
Commun. 1986, 1568Ϫ1569. Ϫ G. A. Kraus, M. T. Molina, J.
Org. Chem. 1988, 53, 752Ϫ753.
T. Kunz, H.-U. Reißig, Liebigs Ann. Chem. 1989, 891Ϫ893.
Received June 5, 2000
[O00290]
[11]
The corresponding cis-4,5-dimethyl-substituted γ-lactol could
Eur. J. Org. Chem. 2000, 3893Ϫ3901
3901