Lewis acid-promoted reactions of γ-lactols with silyl enol ethers - Stereoselective formation of functionalized tetrahydrofuran derivatives

Article abstract of DOI:10.1002/1099-0690(200103)2001:6<1169::AID-EJOC1169>3.0.CO;2-L

The monosubstituted γ-lactols 1a, 1b, 1c, and 1d and the disubstituted γ-lactol 1e were converted into tetrahydrofuran derivatives by reaction with typical silyl enol ethers in the presence of Lewis acids. Although the most suitable Lewis acid appears to be zinc chloride, BF₃·Et₂O or diethylaluminium chloride are also suitable under appropriate conditions. The stereoselectivities of these substitution reactions are similar to those observed with other silylated nucleophiles; however, there are several important differences. A comparison of the diastereoselectivities of different γ-lactols and of various silylated nucleophiles and organometallic compounds will also be presented in this paper.

Full text of DOI:10.1002/1099-0690(200103)2001:6<1169::AID-EJOC1169>3.0.CO;2-L

FULL PAPER
Lewis Acid-Promoted Reactions of γ-Lactols with Silyl Enol Ethers —
Stereoselective Formation of Functionalized Tetrahydrofuran Derivatives
Andreas Schmitt[a][‡] and Hans-Ulrich Reißig*[a]
Dedicated to Professor Horst Kunz on the occasion of his 60th birthday
Keywords: Asymmetric synthesis / Lewis acids / Tetrahydrofuran / Enols / FelkinϪAnh model / Carbocations
The monosubstituted γ-lactols 1a, 1b, 1c, and 1d and the di-
The stereoselectivities of these substitution reactions are sim-
substituted γ-lactol 1e were converted into tetrahydrofuran ilar to those observed with other silylated nucleophiles; how-
derivatives by reaction with typical silyl enol ethers in the
presence of Lewis acids. Although the most suitable Lewis
ever, there are several important differences. A comparison
of the diastereoselectivities of different γ-lactols and of vari-
acid appears to be zinc chloride, BF3·Et2O or diethylalumi- ous silylated nucleophiles and organometallic compounds
nium chloride are also suitable under appropriate conditions.
will also be presented in this paper.
Introduction
Since the stereoselective synthesis of tetrahydrofurans is of
considerable interest for the preparation of natural products
and other biologically active compounds, our approach to
this product class should have some practical relevance.[6]
In preceding publications we reported the Lewis acid-in-
duced reactions of substituted γ-lactols 1 with silylated nu-
cleophiles (e.g. allyl- and propargylsilanes), which furnished
di- and trisubstituted tetrahydrofuran derivatives such as 3
in good yields and very often excellent diastereoselectivities
(Scheme 1).[1] Also, substitution reactions with suitable or-
ganometallic compounds affording products 4 with good
selectivity are possible.[2] These systematic studies with sev-
eral monosubstituted (and one disubstituted) γ-lactols sup-
plemented our earlier investigations,[3] which employed
highly substituted γ-lactols. In all these experiments good
evidence was gained that oxocarbenium ions of type 2 are
intermediates in these transformations. The observed ste-
reoselectivities could plausibly be explained by a modified
Felkin-Anh model.[4]
Results
Treatment of γ-lactol 1a with an excess of boron trifluor-
ide (3 equivalents) in the presence of silyl enol ether 5 (2
equivalents) provided the expected disubstituted tetrahy-
drofuran 6 (Scheme 2) with very high trans selectivity, albeit
in only moderate yield. In the course of these investigations
we found that zinc chloride appears to be the Lewis acid of
choice when silyl enol ethers are employed as nucleo-
philes.[7] This may be due to the sensitivity of the enol ethers
and of the products to strong Lewis acids. However, in our
earlier studies we demonstrated that the influence of the
Scheme 1
In this final paper of the series we present our experi-
ments with silyl enol ethers as nucleophiles, which should
provide tetrahydrofuran derivatives with a carbonyl group
in the side chain introduced by this substitution reaction.[5]
[a]
Freie Universität Berlin, Institut für Chemie Ϫ Organische
Chemie,
Takustr. 3, 14195 Berlin, Germany
Fax: (internat.) ϩ49-30/8385-5367
E-mail: hans.reissig@chemie.fu-berlin.de
[‡]
See Ref.[8]
Scheme 2
Eur. J. Org. Chem. 2001, 1169Ϫ1174
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/0306Ϫ1169 $ 17.50ϩ.50/0
1169

A. Schmitt, H.-U. Reißig
FULL PAPER
Lewis acid employed on the diastereoselectivity is only mar-
ginal. Thus, the diastereoselectivities as observed in the sub-
sequent experiments should be similar even when per-
formed with different Lewis acids.
The reaction of γ-lactol 1a with the prochiral propio-
phenone-derived silyl enol ether 7 in the presence of ZnCl2
gave compound 8 in good yield (Scheme 2). Here we recog-
nized that reactions performed with this mild Lewis acid
required temperatures higher than Ϫ20 °C. Product 8 was
formed as a mixture of four diastereomers; however, the
trans:cis (diastereofacial) selectivity was very high (91:9),
whereas the simple diastereoselectivity was rather low
(60:40). An attempted equilibration of this mixture with
DBU did not change the initial ratio.
The β-unsubstituted silyl enol ether 9 was also combined
with 1a in the presence of diethylaluminium chloride as
Lewis acid (Scheme 2). In this case it is important to use
only one equivalent of the aluminium reagent since an ex-
cess seems to cause addition of the ethyl group to the car- Scheme 4
bonyl function of 10.[8] However, under appropriate condi-
tions this reaction produces a very good yield of disubsti-
tuted tetrahydrofuran 10, but the diastereoselectivity (trans-
As expected from our experience[1,2] with other nucleo-
philes, the γ-lactol 1d gave only poor diastereoselectivities.
Its substitution reaction with the silyl enol ether 9
(Scheme 5) furnished a 42:58 mixture of two diastereomers
15 in moderate yield. In this case the configuration of the
major product is not definite, since the criteria we use for
structural assignments of the other tetrahydrofuran derivat-
ives in this publication as well as in our earlier work[1,2,8]
do not allow an unambiguous decision. However, due to
the very low diastereoselectivity observed in the formation
of 15 this is of no concern.
:cis ϭ 61:39) was rather disappointing.
The second model compound studied was phenyl-substi-
tuted γ-lactol 1b. In general, it seems to behave like 1a, thus
giving a fairly high trans-selective reaction with silyl enol
ether 11 in the presence of zinc chloride (Scheme 3).[9] Com-
pound 12 was obtained in good yield with a trans:cis ratio
of 91:9. Under similar conditions lactol 1b was also treated
with the parent silyl enol ether, trimethylsiloxyethene.[8]
However, only oligomeric material was obtained, which is
probably due to subsequent reactions of the resulting prod-
uct containing a sterically unhindered aldehyde function.
Scheme 5
We also introduced the trans-4,5-disubstituted γ-lactol 1e
into the reaction under investigation. To our pleasure the
substitution of 1e with the silyl enol ether 9 (Scheme 6) af-
Scheme 3
γ-Lactol 1c was the most selective model substrate in re- forded the tetrahydrofuran derivative 16 in excellent yield,
actions with other nucleophiles providing excellent trans:cis even with BF3 as the Lewis acid, and a moderate diastereo-
ratios of at least 95:5.[1] Thus, we were rather surprised (and selectivity (79:21) in favour of 16a was observed. However,
disappointed) that 1c reacts with silyl enol ethers with only again the reaction of 1e with allyltrimethylsilane had given
moderate to good stereoselectivity. The transformation of a significantly higher selectivity (86:14). The influence on
the silyl enol ether 11 into substitution product 13 the stereoselectivity of the 4-methyl group in 1e seems to be
(Scheme 4) proved again that the diastereoselectivity of the higher than the effect of the 5-methyl group, in accordance
γ-lactol reactions is independent of the Lewis acid em- with our earlier published examples.[1]
ployed. With BF3·Et2O and with ZnCl2 almost identical
Finally, we examined a one-pot procedure which com-
trans:cis ratios were observed, but the milder zinc Lewis bines generation of the γ-lactol and its substitution reaction
acid afforded a considerably higher yield. Whereas the β- with a silyl enol ether. For this purpose the γ-lactone 17 was
disubstituted silyl enol ether 11 still gave a rather good first treated with diisobutylaluminium hydride, then with
trans-selectivity, it was remarkably lower with the β-unsub- BF3·Et2O, and finally nucleophile 9 was added (Scheme 7).
stituted nucleophile 9, furnishing the tetrahydrofuran deriv- Product 14 was obtained in a similar trans:cis ratio of iso-
ative 14 in a trans:cis selectivity of only 73:27.
mers (72:28) as observed in the two-step sequence (see
1170
Eur. J. Org. Chem. 2001, 1169Ϫ1174

Lewis Acid-Promoted Reactions of γ-Lactols with Silyl Enol Ethers
FULL PAPER
actions employing different γ-lactols 1 and the various (sily-
lated) nucleophilic components.[1,2] Scheme 8 reveals that
the tert-butyl-substituted lactols 1f and 1g are the most
trans selective as to be expected with this bulky group. Sur-
prisingly, the 4-substituted lactols 1c and 1h are very close
with respect to their trans selectivity. At the end of this se-
quence we find the 5-substituted compounds 1d and 1j,
whereas the γ-lactols 1a and 1b bearing the substituent rela-
tively close to the reacting anomeric centre generally
showed only moderate selectivity.
Scheme 6
above). The yield was rather low, and therefore the one-pot
procedure seems not to be superior to the normal protocol.
However, we did not try the one-pot reaction with other
(milder) Lewis acids which might give better results.
Scheme 7
Discussion
The reactions of silyl enol ethers with γ-lactols generally
proceed with lower yield than those of 1 with allylsilanes.
This may be due to the higher sensitivity of silyl enol ethers
and their reaction products to Lewis acids. The observed
diastereoselectivities do not completely fit in the pattern ob-
served for reactions of 1 with allylsilanes or propargylsil-
anes.[1] They are generally slightly higher for 3-substituted
γ-lactols like 1a and 1b, but are significantly lower for the
4-substituted γ-lactol 1c. The one example presented here
with the 5-substituted γ-lactol 1d demonstrates that with
this electrophile the selectivity is low (and relatively unpre-
dictable); with other nucleophiles we have also observed low
diastereoselectivities or moderate cis- or trans-selective reac-
tions depending on the size of the nucleophile and the sub-
stituent at the γ-lactol.[1,2]
We do not want to repeat our interpretation of γ-lactol
substitutions as discussed earlier.[1,2] The Felkin-Anh
model[4] should also explain the results observed here, al-
though no obvious arguments can be given for deviations
Scheme 8
The selectivity pattern with respect to the nucleophiles
in selectivity. The lower selectivities of silyl enol ethers as employed is more difficult to understand and should be di-
observed in the majority of examples may be due to their vided into three parts according to the γ-lactol type (see
higher nucleophilicity,[10] but we hesitate to discuss further Scheme 9). Thus, for reactions of 3-substituted γ-lactols the
this apparently delicate balance between steric and elec- β-disubstituted nucleophiles are most trans selective,
tronic effects due to the relatively small number of experi- whereas slim silylated components such as propargylsilane
mental results available. However, there is a general trend are the least trans-selective reagents. For the 4-substituted
with β-disubstituted silyl enol ethers as 5 and 11 are the γ-lactols, the sequence is roughly similar although there are
most diastereoselective species, whereas the β-unsubstituted deviations in detail (e.g. the surprisingly low trans selectivity
silyl enol ethers such as 9 are relatively unselective.[11] Of of the β-disubstituted silyl enol ether). The reactions of the
more importance may be the qualitative selectivity pattern 5-substituted γ-lactols are the least predictable transforma-
which can be gained from comparison of all our model re- tions: sterically hindered nucleophiles are trans selective, the
Eur. J. Org. Chem. 2001, 1169Ϫ1174
1171

A. Schmitt, H.-U. Reißig
FULL PAPER
Scheme 9
reagents of apparently moderate steric demand are unselect- Experimental Section
ive, whereas silylated components with little hindrance at
General Remarks: See ref.[2] 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-
pared are rather sensitive to oxygen. Therefore, for most com-
pounds no correct elemental analysis could be obtained (see com-
pound 12); instead, they were converted into crystalline 2,4-dinitro-
phenylhydrazones by a standard procedure which, after recrystal-
lization, provided correct elemental analyses.
the reacting carbon are slightly cis selective.[12] As men-
tioned above, it is difficult to discuss these effects in detail.
Nevertheless, the pattern as presented as result of our
model reactions may be used as a guide for predicting the
stereochemical outcome of other reactions at the anomeric
centre of γ-lactols or similar cyclic acetals.
Starting Materials: The preparation of the γ-lactols 1a, 1b, 1c, and
1d is described in ref.[2] Compound 1e was described in ref.[1]
Conclusions
General Procedure for the Reactions of γ-Lactols with Silyl Enol
Ethers: To a solution of the corresponding γ-lactol 1 in dry
dichloromethane (2 mL/mmol) was added 2 equivalents of the cor-
responding silyl enol ether and the mixture was cooled to Ϫ78 °C.
Then, 3 equivalents of the Lewis acid were added from a syringe
In our model reactions we could demonstrate that a vari-
ety of silylated nucleophiles or organometallic compounds
can be employed for direct substitution of the HO group at
the anomeric centre of γ-lactols. By use of the appropriate and the mixture was allowed to warm to room temperature within
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 dis-
tilled in a kugelrohr oven. Boron trifluoride and zinc dichloride
were used as their etherates. For the amounts of starting materials
used and the resulting products see Table 1.
2-Methyl-2-(3-methyl-2-tetrahydrofuranyl)propanal (6): 1H NMR
(CDCl3): δ ϭ 0.99, 1.01 (2 s, 3 H each, Me), 1.02 (d, J ϭ 7.0 Hz,
3 H, 3Ј-Me), 1.42Ϫ1.52, 1.84Ϫ2.17 (2 m, 1 H, 2 H, 3Ј-H, 4Ј-H),
3.46 (d, J ϭ 6.0 Hz, 1 H, 2Ј-H), 3.60Ϫ3.80 (m, 2 H, 5Ј-H), 9.52 (s,
1 H, 1-H); the signals of cis-6 are not detectable. Ϫ 13C NMR
(CDCl3): δ ϭ 16.8/17.7, 18.7/18.5, 19.7/20.8 (3 q, Me), 33.8/33.3
(d, C-3Ј), 34.9/35.0 (t, C-4Ј), 49.4 (s, C-2), 67.2/67.3 (t, C-5Ј), 89.3/
90.7 (d, C-2Ј), 205.5 (d, C-1). Ϫ IR (neat): ν˜ ϭ 2980Ϫ2830 cmϪ1
(CH), 1710 (CϭO).
Lewis acid — in most cases BF3·Et2O is sufficient — this
nucleophilic substitution is possible without conversion of
the hydroxy group into a better leaving group; the involve-
ment of an oxocarbenium ion as intermediate is plausible
and supported by experimental facts. We could gain in-
formation about the diastereoselectivity of this process and
its dependence on the substitution of the γ-lactol and the
type of nucleophile employed. The most surprising result
was the very high trans selectivity observed with the 4-sub-
stituted γ-lactols, which is often higher than those recorded
for the corresponding 3-substituted γ-lactols. A mechanistic
rationalization for this effect has been presented. Our re-
sults obtained with model compounds may help to design
syntheses of advanced tetrahydrofuran derivatives.
1172
Eur. J. Org. Chem. 2001, 1169Ϫ1174

Lewis Acid-Promoted Reactions of γ-Lactols with Silyl Enol Ethers
Table 1. Reactions of γ-lactols 1aϪ1e with silyl enol ethers in the presence of Lewis acids according to the general procedure
FULL PAPER
γ-Lactol
[mg (mmol)]
Silyl enol ether
[2 equiv.]
Lewis acid
[3 equiv.]
Product
Yield (%)
[mg (%)]
trans:cis
b.p.
[° C/Torr]
1a [500 (4.90)]
1a [420 (4.12)]
1a [190 (1.86)]
1b [260 (1.59)]
1c [350 (3.43)]
1c [400 (3.92)]
1c [300 (1.83)]
1d [250 (2.45)]
1e [135 (1.16)]
5
BF3·OEt2
ZnCl2
6
327 (43)[a]
687 (76)
250 (86)[d]
288 (83)
107 (20)
374 (61)
187 (29)
241 (53)
212 (92)
95:5
70/4.5
90/0.01
90/0.01
95/0.01
70/0.4
70/0.4
70/0.4
70/0.4
75/0.4
7
8
(56 ϩ 35): (5 ϩ 4)[b]
9
Et2AlCl[c]
ZnCl2
10
12
13
13
14
15
16
61:39[e]
91:9
88:12
86:14
73:27
42:58
79:21[f]
11
11
11
9
BF3·OEt2
ZnCl2
BF3·OEt2
BF3·OEt2
ZnCl2
9
9
[a]
[b]
[c]
[d]
[e]
Product contains tert-butyldimethylsilanol. Ϫ Four diastereomers. Ϫ One equivalent. Ϫ Product was impure. Ϫ Determined
by 13C NMR of the distilled sample. Ϫ Trans:cis with respect to substituents at C-2 and C-4.
[f]
2,4-Dinitrophenylhydrazone of
6
(m.p. 110Ϫ112 °C).
Ϫ
C 77.03, H 8.31; found C 74.94, H 8.15. No correct elemental ana-
C15H20N4O5 (336.3): calcd. C 53.57, H 5.99, N 16.66; found C lysis could be obtained of this compound.
53.39, H 5.92, N 16.78.
2-Methyl-2-(4-methyl-2-tetrahydrofuranyl)propanal (13): 1H NMR
2-(3-Methyl-2-tetrahydrofuranyl)-1-phenylpropanone (8) (four dia-
stereomers): 1H NMR (CDCl3): δ ϭ 1.04 (d, J ϭ 6.0 Hz, 0.11 H,
3Ј-Me), 1.04 (d, J ϭ 6.5 Hz, 1.69 H, 3Ј-Me), 1.07 (d, J ϭ 6.5 Hz,
1.06 H, 3Ј-Me), 1.16 (d, J ϭ 7.0 Hz, 0.14 H, 3Ј-Me), 1.20 (d, J ϭ
7.0 Hz, 0.14 H, 3-Me), 1.21 (d, J ϭ 7.0 Hz, 1.06 H, 3-Me), 1.29 (d,
J ϭ 7.0 Hz, 1.69 H, 3-Me), 1.38 (d, J ϭ 7.0 Hz, 0.11 H, 3-Me),
1.45Ϫ1.60, 1.94Ϫ2.24 (2 m, 3 H, 3Ј-H, 4Ј-H), 3.50Ϫ3.90 (m, 4 H,
2Ј-H, 5Ј-H, 3-H), 7.40 Ϫ7.60, 7.90Ϫ8.00 (2 m, 5 H, Ph). Ϫ 13C
NMR (CDCl3): δ ϭ 13.7* ϩ 13.0 (q, C-3), 18.4* ϩ 19.1 (q, 3Ј-
Me), 35.0* ϩ 34.9/31.1# ϩ 33.0 (t, C-4Ј), 36.9* ϩ 36.0/33.8# ϩ33.6
(d, C-3Ј), 44.4* ϩ 45.2/41.9# ϩ 41.5 (d, C-2), 66.7/65.7# ϩ 65.8 (t,
C-5Ј), 86.7* ϩ 87.0/83.4# ϩ 83.8 (d, C-2Ј), 128.2* ϩ 127.9/127.7,
128.3* ϩ 128.1/128.5, 132.7* ϩ 132.5/132.9, 136.8* ϩ 137.1 (3 d,
s, Ph), 202.5 ϩ 202.6 (s, C-1); *major trans isomer/#major cis iso-
mer. Ϫ IR (neat): ν˜ ϭ 3080Ϫ3000 cmϪ1 (ϭCH), 3000Ϫ2800 (CH),
1670 (CϭO). Ϫ C14H18O2 (218.3): calcd. C 77.03, H 8.31; found C
77.10, H 8.33.
(CDCl3): δ ϭ 0.97, 0.99 (2 s, 2 H each, Me), 0.97 (d, J ϭ 7.0 Hz,
3 H, 4Ј-Me), 1.48 (td, J ϭ 8.0, 13.0 Hz, 1 H, 3Ј-H), 2.14Ϫ2.28/
2.30Ϫ2.40 (m, 1 H, 4Ј-H), 2.80 (td, J ϭ 8.0, 13.0 Hz, 1 H, 3Ј-H),
3.23/3.20, 3.88/3.95 (dd/t, dd, J ϭ 6.5, 8.5/8.5 Hz, J ϭ 6.5, 8.5 Hz/
6.0, 8.5 Hz, 1 H each, 5Ј-H), 4.01 (t, J ϭ 7.5 Hz, 1 H, 2Ј-H), 9.53/
9.54 (s, 1 H, 1-H). Ϫ 13C NMR (CDCl3): δ ϭ 18.0, 18.2, 19.2 (3
q, 3 Me), 33.2/34.2 (d, C-4Ј), 34.5/35.4 (t, C-3Ј), 49.3/49.1 (s, C-2),
75.3/75.1 (t, C-5Ј), 82.4 (d, C-2Ј), 205.9/205.2 (d, C-1). Ϫ IR (neat):
ν˜ ϭ 2980Ϫ2940 cmϪ1 (CH), 1720 (CϭO).
2,4-Dinitrophenylhydrazone of 13 (m.p. 116Ϫ117 °C).
Ϫ
C15H20N4O5 (336.3): calcd. C 53.57, H 5.99, N 16.66; found C
53.72, H 5.90, N 16.54.
3,3-Dimethyl-1-(4-methyl-2-tetrahydrofuranyl)-2-butanone
(14): 1H NMR (CDCl3): δ ϭ 1.04 (d, J ϭ 6.5 Hz, 3 H, 4Ј-Me),
1.13/1.14 (s, 9 H, tBu), 1.65, 1.75 (2 ddd, J ϭ 7.0, 8.5, 12.5 Hz, J ϭ
6.0, 7.0, 12.5 Hz, 1 H each, 3Ј-H), 2.24Ϫ2.39 (m, 1 H, 4Ј-H), 2.53/
2.58, 2.92/2.97 (2 dd, J ϭ 7.0, 17.0/6.5, 17.0 Hz, J ϭ 6.0, 17.0 Hz,
1 H each, 1-H), 3.25/3.33, 3.97/3.87 (dd/t, dd/t, J ϭ 7.0, 8.5 Hz/
7.5 Hz, J ϭ 7.0, 8.5/8.0 Hz, 1 H each, 5Ј-H), 4.38/4.23- 4.33 (tt/m,
J ϭ 6.0, 7.0 Hz, 1 H, 2Ј-H). Ϫ 13C NMR (CDCl3): δ ϭ 17.9/17.8
(q, 4Ј-Me), 26.2, 44.2 (q, s, tBu), 33.2/34.3 (d, C-4Ј), 39.8/41.0 (t,
C-3Ј), 42.9/43.0 (t, C-1), 74.8/74.3 (t, C-5Ј), 74.8/75.9 (d, C-2Ј),
214.1 (s, C-2). Ϫ IR (neat): ν˜ ϭ 2980Ϫ2820 cmϪ1 (CH), 1705
(CϭO).
3,3-Dimethyl-1-(3-methyl-2-tetrahydrofuranyl)-2-butanone
(10): 1H NMR (CDCl3): δ ϭ 1.05/0.87 (d, J ϭ 7.0 Hz, 3 H, 3Ј-
Me), 1.14/1.15 (s, 9 H, tBu), 1.48Ϫ1.65, 2.00Ϫ2.20 (2 m, 1 H each,
4Ј-H), 1.85/2.32Ϫ2.45 (sept/m, J ϭ 7.0 Hz, 1 H, 3Ј-H), 2.53/2.50
(dd, J ϭ 4.5, 16.5 Hz/6.5, 17.5 Hz, 1 H, 1-H), 2.80/2.87 (dd, J ϭ
7.5, 16.5 Hz/7.0, 17.5 Hz, 1 H, 1-H), 3.80Ϫ3.90/3.75Ϫ3.90 (m, 1 H,
2Ј-H), 3.80Ϫ3.90/3.70, 4.23 (m/dt, q, J ϭ 6.0, 8.0 Hz, J ϭ 6.0 Hz,
2 H, 5Ј-H). Ϫ 13C NMR (CDCl3): δ ϭ 14.4/17.0 (q, 3Ј-Me), 25.9,
44.0/26.2, 43.9 (q, s, tBu), 34.1/33.6 (t, C-4Ј), 39.0/34.6 (d, C-3Ј),
41.0/37.3 (t, C-1), 66.5/65.7 (t, C-5Ј), 77.3/81.3 (d, C-2Ј), 213.6/
213.7 (s, C-2). Ϫ IR (neat): ν˜ ϭ 2990Ϫ2810 cmϪ1 (CH), 1700
(CϭO).
2,4-Dinitrophenylhydrazone of 14 (m.p. 129Ϫ131 °C).
Ϫ
C17H24N4O5 (364.4): calcd. C 56.03, H 6.64, N 15.37; found C
55.89, H 6.59, N 15.31.
3,3-Dimethyl-1-(5-methyl-2-tetrahydrofuranyl)-2-butanone
1
(15): H NMR (CDCl3): δ ϭ 1.14/1.13 (s, 9 H, tBu), 1.21/1.23 (d,
2,4-Dinitrophenylhydrazone of 10 (m.p. 122Ϫ125 °C).
Ϫ
J ϭ 6.0 Hz, 3 H, 5Ј-Me), 1.34Ϫ1.56, 1.94Ϫ2.26 (2 m, 2 H each, 3Ј-
H, 4Ј-H), 2.57/2.54, 2.94/2.98 (2 dd, J ϭ 7.5, 17.0 Hz, J ϭ 5.5,
C17H24N4O5 (364.4): calcd. C 56.03, H 6.64, N 15.37; found C
55.82, H 6.65, N 15.50.
17.0 Hz,
1 H each, 1-H), 4.03Ϫ4.18/3.88Ϫ3.97, 4.36Ϫ4.45,
2-Methyl-2-(3-phenyl-2-tetrahydrofuranyl)propanal (12): 1H NMR
(CDCl3): δ ϭ 0.79, 0.89 (2 s, 3 H each, Me), 1.88, 2.19/1.56Ϫ1.80
(qd, dddd/m, J ϭ 8.0, 12.5 Hz, J ϭ 5.0, 7.0, 8.0, 12.5 Hz, 2 H, 4Ј-
H), 3.04/3.00Ϫ3.10 (q/m, J ϭ 8.0 Hz, 1 H, 3Ј-H), 3.78 (dt, J ϭ 7.0,
8.0 Hz, 1 H, 5Ј-H), 3.90 (dt, J ϭ 5.0, 8.0 Hz, 1 H, 5Ј-H), 3.98 (d,
J ϭ 8.0 Hz, 1 H, 2Ј-H), 7.00Ϫ7.40 (m, 5 H, Ph), 9.36 (s, 1 H, 1-
H). Ϫ 13C NMR (CDCl3): δ ϭ 17.5/18.5, 19.0/19.7 (2 q, 2 Me),
37.4/32.6 (t, C-4Ј), 46.5/46.4 (d, C-3Ј), 49.4 (s, C-2), 68.5/67.6 (t, C-
5Ј), 89.0/88.9 (d, C-2Ј), 126.5, 127.5/127.0, 128.6/128.1, 142.7/142.8
4.19Ϫ4.26 (2 m, 1 H each, 5Ј-H, 2Ј-H). Ϫ 13C NMR (CDCl3): δ ϭ
21.4/21.3 (q, 5Ј-Me), 26.2, 44.2 (q, s, tBu), 32.6/31.6, 33.8/32.8 (2 t,
C-3Ј, C-4Ј), 42.9 (t, C-1), 74.7/74.6, 75.6/75.4 (2 d, C-5Ј, C-2Ј),
214.1/214.2 (s, C-2). Ϫ IR (neat): ν˜ ϭ 3000Ϫ2840 cmϪ1 (CH), 1700
(CϭO).
2,4-Dinitrophenylhydrazone of 15 (m.p. 121Ϫ122 °C).
Ϫ
C17H24N4O5 (364.4): calcd. C 56.03, H 6.64, N 15.37; found C
55.85, H 6.54, N 15.33.
(3 d, s, Ph), 205.5 (d, C-1). Ϫ IR (neat): ν˜ ϭ 3080Ϫ3000 cmϪ1 (ϭ 3,3-Dimethyl-1-(4,5-dimethyl-2-tetrahydrofuranyl)-2-
CH), 2980Ϫ2840 (CH), 1725 (CϭO). Ϫ C14H18O2 (218.3): calcd. butanone (16): 1H NMR (CDCl3): δ ϭ 0.76/0.77, 1.14 (2 d, J ϭ
Eur. J. Org. Chem. 2001, 1169Ϫ1174
1173

A. Schmitt, H.-U. Reißig
FULL PAPER
hedron 1989, 45, 3139Ϫ3150; for recent contributions of other
6.5 Hz, J ϭ 6.0 Hz, 3 H each, 4Ј-Me, 5Ј-Me), 0.95/0.96 (s, 9 H,
tBu), 1.40Ϫ1.70/2.17 (m/broad quint, J ഠ 6.0 Hz, 2 H, 3Ј-H), 2.28/
2.27, 2.83/2.81 (2 dd, J ϭ 6.0, 17.0 Hz/6.0, 16.5 Hz, 1 H each, 1-
H), 3.27/3.45 (qd, J ϭ 6.0, 7.0 Hz/6.0, 9.0 Hz, 1 H, 5Ј-H), 4.49/4.59
(tt/dq, J ϭ 6.0, 7.5 Hz/3.0, 6.0 Hz, 1 H, 2Ј-H). Ϫ 13C NMR
(CDCl3): δ ϭ 16.8/16.2, 19.3/19.2 (2 q, 5Ј-Me, 4Ј-Me), 26.2, 44.1
(q, s, tBu), 40.2/42.0 (d, C-4Ј), 40.2/42.0 (t, C-3Ј), 43.5 (t, C-1),
73.7/74.0 (d, C-2Ј), 82.0/80.9 (d, C-5Ј), 214.2 (s, C-2). Ϫ IR (neat):
ν˜ ϭ 2980Ϫ2820 cmϪ1 (CH), 1700 (CϭO).
´
groups: E. Alonso, D. J. Ramon, M. Yus, Tetrahedron 1997, 53,
2641Ϫ2652; Y. Nishiyama, T. Katoh, K. Deguchi, Y. Morim-
oto, K. Itoh, J. Org. Chem. 1997, 62, 9339Ϫ9341; R. A. Pili,
V. B. Riatto, Tetrahedron: Asymmetry 2000, 11, 3675Ϫ3686.
M. Cherest, H. Felkin, N. Prudent, Tetrahedron Lett. 1968,
2199Ϫ2204; M. Cherest, H. Felkin, Tetrahedron Lett,. 1968,
2205Ϫ2208; see also: C. H. Larsen, B. H. Ridgway, J. T. Shaw,
K. A. Woerpel, J. Am. Chem. Soc. 1999, 121, 12208Ϫ12209.
S. Murata, R. Noyori, Tetrahedron Lett. 1982, 23, 2601Ϫ2602;
M. T. Reetz, Angew. Chem. 1982, 94, 97Ϫ109; Angew. Chem.
Int. Ed. Engl. 1982, 21, 96Ϫ108; M. T. Reetz, H. Müller-Starke,
Liebigs Ann. Chem. 1983, 1726Ϫ1738; T. Mukaiyama, S. Ko-
bayashi, S. Shoda, Chem. Lett. 1984, 1529Ϫ1530; K. Narasaka,
Y. Ichikawa, H. Kubota, Chem. Lett. 1987, 2139Ϫ2142; S. Mu-
rata, M. Suzuki, R. Noyori, Tetrahedron 1988, 44, 4259Ϫ4275;
U. Koert, M. Stein, K. Harms, Tetrahedron Lett. 1993, 34,
[4]
[5]
2,4-Dinitrophenylhydrazone of 16 (m.p. 134Ϫ138 °C).
Ϫ
C18H26N4O5 (378.4): calcd. C 57.13, H 6.92, N 14.81; found C
56.92, H 6.87, N 14.76.
Preparation of 14 by a One-Pot Procedure: γ-Lactone 17 (500 mg,
5.00 mmol) was dissolved in 10 mL of toluene and cooled to Ϫ80
to Ϫ90 °C. Then, diisobutylaluminium hydride (1  solution in
toluene, 6.0 mL, 6.0 mmol) was added within 30 min. The resulting
solution was stirred for 30 min. at Ϫ78 °C and then boron trifluor-
ide etherate (2.11 g,15.0 mmol) and, after 15 min., 9 (1.72 g,
10.0 mmol) were added. The mixture was warmed to Ϫ30 °C within
2 h, then hydrolysed with 10 mL of water and warmed to room
temperature. After precipitation of the aluminium hydroxides the
mixture was filtered through a pad of Celite and the filtrate was
extracted with tert-butyl methyl ether. The combined organic
phases were dried (MgSO4) and concentrated. The residue was dis-
tilled by kugelrohr (70 °C/0.4 Torr) distillation to give 300 mg
(38%) of pure 14 (trans:cis ϭ 72:28).
`
´
2299Ϫ2302; B. Figadere, C. Chaboche, J.-F. Peyrat, A. Cave,
Tetrahedron Lett. 1993, 34, 8093Ϫ8096; J. Uenishi, A. Sohma,
O. Yonemitsu, Chem. Lett. 1996, 595Ϫ596; Reviews: R. R.
Schmidt, in Comprehensive Organic Synthesis (Eds.: B. M.
Trost, E. Winterfeldt), Pergamon Press, Oxford 1991, Vol. 6,
chapter 1.2; M. H. D. Postema, Tetrahedron 1992, 48,
8545Ϫ8599. Review on the related reactions of N-acyliminium
ions: W. N. Speckamp, M. J. Moolenaar, Tetrahedron 2000,
56, 3817Ϫ3856.
[6]
Reviews: K. Mori, Tetrahedron 1989, 45, 3233Ϫ3298; J. W.
Westley, Polyether Antibiotics: Natural Occurring Acid Ionoph-
ores; Marcel Dekker, New York, 1982; T. L. B. Boivin, Tetra-
hedron 1987, 43, 3309Ϫ3362; G. Cardillo, M. Orena, Tetrahed-
`
ron 1990, 46, 3321Ϫ3408; J.-C. Harmange, B. Figadere, Tetra-
hedron: Asymmetry 1993, 4, 1711Ϫ1754; U. Koert, Synthesis
1995, 115Ϫ132; M. C. Elliot, J. Chem. Soc., Perkin Trans. 1
1998, 4175Ϫ4200.
[7]
[8]
[9]
One side reaction may be β-elimination under tetrahydrofuran
ring opening followed by Michael-addition.
A. Schmitt, Dissertation, Technische Hochschule Darmstadt,
1992.
In several experiments no noteworthy differences in the diaster-
eoselectivities were found when trimethylsilyl enol ethers were
replaced by dimethyl-tert-butylsilyl enol ethers.[8]
Review: H. Mayr, M. Patz, Angew. Chem. 1994, 106,
990Ϫ1010; Angew. Chem. Int. Ed. Engl. 1994, 33, 938.
A related effect has recently been observed for the reactions of
a γ-lactol with titanium enolates: R. A. Pilli, V. B. Riatto, I.
Vencato, Org. Lett. 2000, 2, 53Ϫ56.
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 preparation
of this manuscript.
[10]
[11]
[1]
A. Schmitt, H.-U. Reißig, Eur. J. Org. Chem. 2000, 3893Ϫ3901;
for a preliminary communication of our results see: A. Schmitt,
H.-U. Reißig, Synlett 1990, 40Ϫ42.
A. Schmitt, H.-U. Reißig, Chem. Ber. 1995, 128, 871Ϫ876.
[2]
[12]
A similarly low selectivity has been reported for a rearrange-
ment of an anomerically linked silyl enol ether: D. J. Dixon, S.
V. Ley, E. W. Tate, Synlett 1998, 1093Ϫ1095.
Received October 6, 2000
[3]
C. Brückner, H. Lorey, H.-U. Reißig, Angew. Chem. 1986, 98,
559Ϫ560; Angew. Chem. Int. Ed. Engl. 1986, 25, 556Ϫ557; C.
Brückner, H. Holzinger, H.-U. Reißig, J. Org. Chem. 1988, 53,
2450Ϫ2456; H.-U. Reißig, H. Holzinger, G. Glomsda, Tetra-
[O00518]
1174
Eur. J. Org. Chem. 2001, 1169Ϫ1174