Two contrasting asymmetric approaches to muscarine based on 5-endo-trig cyclisations

Article abstract of DOI:10.1002/ejoc.200300647

5-endo-trig cyclisation of the (Z)-hydroxyalkenoate 17 using iodine as the electrophile gave a good yield of the β-hydroxytetrahydrofuran 18, probably via the corresponding iodohydrin. A variety of one-carbon degradation methods were then used to generate precursors to (-)-muscarine (25d). An alternative strategy featured control of a 5-endo-trig iodocyclisation by an allylic hydroxyl group, which can be used for the highly stereocontrolled synthesis of hydroxy-iodotetrahydrofurans 28. Application of this strategy to the (Z)-alkenediol derivative 38 led to an excellent yield of the tetrahydrofuran 39 when iodine monobromide was used as the electrophile. Two simple and efficient transformations then gave the (+)-muscarine precursor 40b. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200300647

FULL PAPER
Two Contrasting Asymmetric Approaches to Muscarine Based on 5-endo-trig
Cyclisations
David W. Knight,*[a] Duncan E. Shaw,[b] and Emily R. Staples[a]
Keywords: Cyclisation / Muscarine / Oxygen heterocycles / Enantioselectivity / Natural products
5-endo-trig cyclisation of the (Z)-hydroxyalkenoate 17 using
trahydrofurans 28. Application of this strategy to the (Z)-alk-
iodine as the electrophile gave a good yield of the β-hydroxy- enediol derivative 38 led to an excellent yield of the tetrahy-
tetrahydrofuran 18, probably via the corresponding iodohyd-
drofuran 39 when iodine monobromide was used as the elec-
rin. A variety of one-carbon degradation methods were then
trophile. Two simple and efficient transformations then gave
used to generate precursors to (−)-muscarine (25d). An al- the (+)-muscarine precursor 40b.
ternative strategy featured control of a 5-endo-trig iodo-
cyclisation by an allylic hydroxyl group, which can be used ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
for the highly stereocontrolled synthesis of hydroxy-iodote- Germany, 2004)
Introduction
(ϩ)-Muscarine (1) is a major toxic principle of the well-
known Fly agaric mushroom, Amanita muscaria, a common
and spectacular (if dangerous) feature of autumnal wood-
lands.[1] It is, however, not the only biologically active alka-
loid present in this species, because it occurs alongside mus-
cimol (2). This latter, seemingly simple compound is re-
sponsible for the hallucinogenic effects of the mushroom, a
property used in ancient rituals by shamen and in South
American Indian cults of mushroom worship.[2] Muscarine
alkaloids are found in many other species of mushroom in-
cluding other members of the genus Amanita, such as the
Death cap, A. phalloides, along with many Inocybes and Cli-
tocybes species.[1] Many of these compounds are stereoiso-
mers of the muscarine structure 1 and include (Ϫ)-
(2S,3R,5R)-allomuscarine (3), (ϩ)-epimuscarine [the
Early work on muscarine (1) was plagued by purification
problems, and many samples were contaminated with chol-
ine and acetylcholine, which were also present in the fungal
source.[6] Pure material was eventually secured in 1954[7]
and its structure firmly established by X-ray crystallo-
graphic analysis,[8] thus opening up a new era of investi-
gations into its synthesis and pharmacological activity. The
action of muscarine upon smooth muscle so resembles that
of acetylcholine (4) that direct action on cholinergic recep-
tors in the autonomic nervous system has come to be
known as ‘‘muscarinic’’ action.[3]
(2S,3S,5S)-isomer],
and
(ϩ)-epiallomuscarine
[the
(2S,3S,5R) isomer].[3] Muscarine itself has a long and noble
history stretching back to 1811, when the first recorded at-
tempts at its structural determination and synthesis were
made.[4] Despite its structure being uncertain at the time, it
even achieved literary fame in 1930, when the difference
between ‘‘synthetic’’ and natural muscarine held the key to
a murder mystery.[5]
Many muscarinic subtypes have subsequently been ident-
ified,[9] and its acetylcholine agonistic activity has many po-
tential applications in a variety of medicinal therapies, no-
tably neurodegenerative conditions such as Parkinson’s and
Alzheimer’s diseases.[10] Amongst the known muscarine iso-
mers (see above), muscarine (1) itself displays the most po-
tent biological activity and can adopt a conformation close
to that preferred by acetylcholine (4) in solution.[3,8,11]
These potent biological activities, together with its relatively
simple but arguably somewhat awkward structure, have re-
[a]
Chemistry Department, Cardiff University,
P. O. Box 912, Cardiff, CF10 3TB, UK
E-mail: knightdw@cf.ac.uk
[b]
School of Chemistry, University of Nottingham, University
Park,
Nottingham, NG7 2RD, UK
Eur. J. Org. Chem. 2004, 1973Ϫ1982
DOI: 10.1002/ejoc.200300647
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1973

D. W. Knight, D. E. Shaw, E. R. Staples
FULL PAPER
sulted in muscarine becoming something of a test bed for applications of overall 5-endo-trig cyclisations recently high-
new synthetic methodologies or strategies and hence a num- lighted by us as useful synthetic methods. In our earlier
ber of total syntheses have been published to date. Perhaps model studies,[32] we had established that (E)-homoallylic
inevitably, the presence of a number of CϪO bonds in mus- alcohols (5; R1, R2 ϭ alkyl, aryl) undergo very smooth 5-
carine suggests carbohydrate precursors. Examples of this endo-trig iodocyclisations when treated with three equiva-
idea include the first, short (but non-regioselective) ap- lents each of iodine and sodium hydrogencarbonate in
proach from -glucosamic acid[12] and a subsequent syn- acetonitrile under strictly anhydrous conditions to give ex-
thesis by the same group starting from 2-deoxy--ribose.[13] cellent yields of the iodotetrahydrofurans 7. The transition
An alternative began with 2-deoxy--ribose;[14] approaches state conformation 6 appears to account for this stereoch-
from -mannitol,[15] isopropylidene glyceraldehyde,[16] - emical outcome. Examples of the corresponding (Z)-homo-
glucose,[17] and -mannonolactone[18] each have notable allylic alcohols similarly gave the all-cis-iodotetrahydrofur-
features and some the flexibility to prepare other muscarine ans 8, albeit in much poorer yields and much more slowly,
isomers or homologues. Arguably, one of the most practical probably due to a more crowded and hence less favourable
is a more recent approach from -rhamnose.[19] Various transition state. However, a surprise came with the finding
contrasting cyclisation methods to establish the tetrahydro- that, in the special case of the (Z)-unsaturated β-hydroxy
furan ring have also been reported, including diazo dis- esters 9, the products were largely the hydroxytetrahydrofu-
placement by a suitably positioned hydroxyl group,[20] chel- rans 10, rather than the expected iodotetrahydrofurans. An
ation-controlled Grignard addition to an α-acetoxy-alde- explanation for this[33] features participation by the ester
hyde and tosylate displacement[21] and very short ap- group and subsequent regio- and stereoselective formation
proaches based on 5-exo-trig iodocyclisations,[22] which of the iodohydrins 11 and a final, slow cyclisation by iodide
were also a key feature of Mulzer’s methodology.[16] A very displacement. Certainly, iodohydrins can be isolated during
recent synthesis of some muscarine isomers features [3ϩ2] the early stages of the reaction, but full characterisation was
cycloadditions between crotylsilanes and α-silyloxyal- precluded by their propensity to undergo cyclisation. This
dehydes.[23] Other approaches have utilised cyclobutanone relatively rapid and efficient formation of the hydroxytetra-
ring expansion,[24] dihydrofuran hydroboration,[25] and car- hydrofurans 10 suggested that this methodology could be
benoid insertion into a CϪH bond α to an ether oxygen[26] suitable for a relatively efficient synthesis of muscarine
to establish the necessary tetrahydrofuran ring. Lengthier (1).[34]
routes begin with (Z)-5-hydroxymethyl-2(5H)-furanone[27]
or
a
4-oxo-tetrahydrofuran-2-carboxylate.[28] Stereoiso-
mers[29] and homologues,[30] some of which possess high
levels of bioactivity, have also been prepared. Finally, Over-
man has applied his elegant acid-catalysed rearrangement
Results and Discussion
For reasons of economy, we chose to use (S)-malic acid
of allylic diols to a synthesis of a key diol precursor to mus- (12) as our starting material, which we planned would lead
carine,[31] which forms the subject of the latter part of this to (Ϫ)-muscarine (13). The diacid 12 was efficiently con-
paper.
verted into the epoxybutanoate 14 using the method of Lar-
[35]
cheveque.
Subsequent alkylation of lithiopropyne[36] by
ˆ
this epoxide under YamaguchiϪHirao conditions[37] then
delivered an excellent yield of the hydroxyheptynoate 15,
with the only significant impurity being 5Ϫ10% of the eas-
ily separated double addition product 16 (Scheme 1).
Lindlar reduction proceeded smoothly to give the neces-
sary (Z)-hydroxyalkenoate 17, which proved to be rather
photosensitive with respect to isomerisation to the corre-
sponding (E)-isomer. As expected, the key ‘‘iodocyclisa-
tion’’ step then gave a respectable 63% isolated yield of the
desired hydroxytetrahydrofuran 18, whose structure and
stereochemistry were established as described pre-
viously.[32,33] Also isolated (in an 11% yield) was the epim-
eric hydroxytetrahydrofuran 19, which is presumably
formed by non-selective iodohydrin generation. Traces of
iodotetrahydrofuran(s) were also detected which probably
arose due to a small amount of (Z)Ǟ(E) isomerisation
prior to or during the cyclisation step. The hydroxyl group
of the major product 18 was then protected as a robust
TIPS ether, and the resulting derivative 20a saponified to
give the corresponding carboxylic acid 20b. Earlier experi-
Herein, we give a full account of two contrasting asym- ments with the corresponding TBDMS derivative indicated
metric approaches to muscarine (1), both of which feature that this group was not sufficiently stable during the ester
1974
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 1973Ϫ1982

Two Asymmetric Approaches to Muscarine
FULL PAPER
carboxylic acid 20b was converted into the corresponding
acid chloride 22 using oxalyl chloride and catalytic DMF
[for these experiments, racemic samples were used, derived
for convenience from more readily available (Ϯ)-methyl 3,4-
epoxybutanoate] (Scheme 3). Simply adding the acid chlo-
ride 22 to a refluxing suspension of the sodium salt of 2-
mercaptopyridine N-oxide led to the rapid formation of the
intermediate thiohydroxamic ester 23, and thence to the de-
sired chloride 24 in very good overall yield. Subsequent de-
protection using fluoride proceeded smoothly to give the
penultimate muscarine precursor 25a. Unfortunately, this
proved to be insufficiently reactive with trimethylamine;
heating in a sealed tube at 130 °C resulted in extensive de-
composition. Presumably, this is a manifestation of the β-
halo ether deactivating effect, exacerbated by the presence
of the hydroxyl group, which renders the substrate more
thermally sensitive. This is in contrast to the parent tetra-
hydrofurfuryl chloride, which does react smoothly with tri-
methylamine at 130 °C.[41] We therefore returned to the
BartonϪHunsdiecker step and used iodoform as the initial
radical source, with cyclohexene as both a solvent and hal-
ogen trap.[40] This proved to be successful, if not quite so
efficient, and delivered the known hydroxy iodide 25c, fol-
lowing fluoride-induced desilylation of the initial product
25b. We were pleased to find that the hydroxy iodide 25c
exhibited analytical and spectroscopic data which were
identical to those previously reported,[16,22] except for the
sign of rotation. Finally, exposure to trimethylamine in
ethanol at 70 °C delivered, as expected,[16] a sample of -
(Ϫ)-muscarine iodide (25d), which was also identical to the
material prepared previously, except for the sign of rotation.
Scheme 1. i) Lithiopropyne, BF3·OEt2, THF, Ϫ78 °C, 4 h, 80% 15
ϩ 5Ϫ10% 16; ii) H2 (1 atm.), 5% Pd-BaSO4, quinoline, EtOAc,
0.75 h, 87%; iii) I2, NaHCO3, MeCN, 0Ϫ5 °C, 72 h, 63% 18 ϩ ca.
11% 19
saponification step to be of use. At no stage was any epi-
merisation observed, which could conceivably occur by a
retro-Michael ring opening-reclosure mechanism.
To complete the synthesis, it was necessary to carry out
a one-carbon degradation of the acetic acid side chain. For
this, a most convenient and efficient procedure appeared to
be Curtius degradation which was expected to provide the
aminomethyl derivative 21 (Scheme 2). Although the reac-
tion sequence using diphenylphosphoranyl azide[38] fol-
lowed by isocyanate hydrolysis and N-methylation was ap-
parently successful, it gave a sample of O-TIPS muscarine
which proved difficult to purify, despite a report[24] that this
conversion sequence could be achieved in good yield in the
cases of some related derivatives.
Scheme 3. i) (COCl2), DMF (cat.), pyridine, C6H6, 0 °C, 3 h; ii) 2-
mercaptopyridine N-oxide sodium salt, DMAP (cat.), CCl4; iii) 65
°C, 1.5 h, 73% from 20b; iv) TBAF, THF, 20 °C, 40 h, 60%; v) as
ii) but CHI3 in place of CCl4 and reflux in cyclohexene for 15 h,
60% from 20b; vi) Me3N, EtOH, 70 °C, 4 h (sealed tube), 75%
Scheme 2. i) TIPSOTf, 2,6-lutidine, CH2Cl2, 0 °C, 3 h, 74%; ii) -
KOH, MeOH/H2O, 20 °C, 16 h, 97%; iii) (PhO)2P(O)N3, Et3N,
toluene, 80 °C, 2 h then H2O, 20 °C, 16 h, 90%
Because other N-methylation methods (e.g. NaBH3CN/
formaldehyde)[39] were less efficient, we turned to an
alternative one-carbon degradation method
—
the
Overall, this approach is a relatively short one, which cer-
BartonϪHunsdiecker reaction[40] — which, if successful, tainly has some potential for the synthesis of analogues by,
would allow introduction of the required amine function in for example, using homologous 1-alkynes for the formation
the final step, and lead directly to a muscarine salt without of the precursor hydroxy esters (cf. 15). Clearly, the natural
the need for a potentially difficult purification. Hence, the (ϩ)-stereoisomer could be obtained simply by starting with
Eur. J. Org. Chem. 2004, 1973Ϫ1982
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1975

D. W. Knight, D. E. Shaw, E. R. Staples
FULL PAPER
(R)-malic acid or the enantiomer of the epoxy ester 14, ob-
tained by a different route. However, the lack of complete
stereocontrol in the key cyclisation step leading to hydroxy-
tetrahydrofuran 18, together with a relatively poor yield of
the iodide 25b and the five-step synthesis of epoxy ester
14,[35] led us to contemplate an alternative approach based
on the 5-endo-trig cyclisation strategy. The precise tactics
for this were evident from our more recent findings regard-
ing the control of such cyclisations by hydroxyl groups. Of
particular relevance to the present target was the obser-
vation that iodocyclisations of the (Z)-anti-alkene diols 26
delivered excellent yields of (very largely) the iodotetrahyd-
rofurans 28, presumably via the transition state confor-
mation 27.[42] Such efficient cyclisations are in direct con-
trast to the much poorer returns from similar reactions of
simpler (Z)-homoallylic alcohols, to give the all-cis iodote-
trahydrofurans 8. As our model studies had featured only
simple alkyl and aryl substituents, a concern was whether
this methodology could be extended to cases where the
products were tetrahydrofuran-2-methanol derivatives (i.e.
28; R2 ϭ CH2OR), and the related issue of defining suitable
protecting groups. Indeed, our first attempt to apply this
methodology avoided the latter consideration by having
R2 ϭ Ph, in the hope that a late oxidative degradation
might deliver the required one-carbon side chain.
Scheme 4. i) Lithiophenylacetylene, 12-crown-4, THF, Ϫ78 °C to 0
°C, 4 h, 76% (ϩ 11% of the syn-isomer); ii) TBAF, THF, 20 °C,
2 h, 93%; iii) H2 (1 atm.), 5% Pd-CaCO3, quinoline, hexane/2%
MeOH, 20 °C, 2 h; iv) I2, NaHCO3, MeCN, 0 °C, 4 h, 51% from
31b; v) H2 (1 atm.), 10% Pd-C, Et3N, EtOAc, 20 °C, 16 h, 98%;
vi) Ac2O, pyridine, 20 °C, 16 h, 97%
of triethylamine, followed by a simple filtration, resulted in
an essentially quantitative yield of the de-iodinated product
34a. Conversion into the corresponding acetate 34b pro-
ceeded smoothly. Sadly, however, exposure of the latter to
ruthenium tetroxide using the now standard Sharpless con-
ditions[46] led to a high yield of the dione 35 (Scheme 5).
No products arising from attack on the phenyl ring were in
evidence.[47] Ozonolysis under various conditions destroyed
the molecule. Clearly, oxidation of the α-positions of the
tetrahydrofuran ring takes precedence over cleavage of the
phenyl group.
Fortunately, a suitable starting material for this approach
was cheap methyl (S)-lactate (29) which, it was hoped,
would deliver (ϩ)-muscarine (1). Thus, ester 29 was con-
verted in two very efficient steps into the known O-silyl lac-
taldehyde 30.[43] Crucially, non-chelation controlled ad-
dition[44] of lithiated phenylacetylene in the presence of 12-
crown-4 gave an excellent yield of the anti-stereoisomer 31a,
containing approximately 11% of the corresponding syn-
isomer (Scheme 4). Desilylation and Lindlar reduction then
gave the (Z)-alkene diol 32, which subsequently underwent
iodocyclisation to give a respectable yield of stereopure
iodotetrahydrofuran 33. The alkene diol 32 proved to be
quite water-soluble, and much material was lost during the
removal of residual quinoline (used in the reduction step)
by an aqueous acid wash.
Scheme 5
We therefore examined the prospects for employing a
propargyl alcohol derivative,[48] and were glad to find that a
non-chelation controlled addition[44] of lithiated O-TBDPS
propargyl alcohol to O-silyl lactaldehyde 36[43] gave the de-
Fortunately, the presence of this base did not appear to sired anti adduct 37 in excellent yield and with a very usable
have a deleterious effect upon the subsequent iodocyclisa- level of stereoselection (ca. 9:1 anti/syn) (Scheme 6). Sub-
tions, thereby obviating the need for an acid wash. The sequent Lindlar reduction then gave the corresponding (Z)-
stereochemistry of the heterocycle 33 was secured as de- alkene 38 in excellent yield. Optimisation experiments re-
scribed previously, and determined by comparisons with re- vealed that, although this derivative could be selectively de-
lated data.[32,33] Initially, the necessary iodine removal was protected, this was unnecessary as iodocyclisation with loss
carried out using the tri-n-butyltin hydride-AIBN meth- of the O-TBS group took place readily when the (Z)-alkene
od[40b] but the inevitably tedious purification led us to seek 38 was exposed to three equivalents of iodine as usual. Even
alternatives, despite obtaining workable 60Ϫ70% yields of better yields (up to 70%) of the desired iodotetrahydrofuran
the desired product 34a. Fortunately, we found that hydro- 39 were secured by direct treatment of 38 with two equiva-
genolysis[45] of the iodotetrahydrofuran 33 in the presence lents of iodine monobromide in acetonitrile at Ϫ10 °C. For-
1976
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 1973Ϫ1982

Two Asymmetric Approaches to Muscarine
FULL PAPER
ing points were determined using a Kofler hot stage apparatus and
tunately, the product(s) arising from syn-38 could be sepa-
rated easily at this stage. Completion of the synthesis then
followed the pathway described above; hydrogenolysis pro-
vided an excellent yield of the de-iodinated tetrahydrofuran
40a, desilylation of which gave the diol 40b. This latter com-
pound, which showed analytical, rotational and spectro-
are uncorrected. Elemental analyses were obtained using
a
PerkinϪElmer 240C Elemental Microanalyser.
All reactions were conducted in oven-dried apparatus under an at-
mosphere of dry nitrogen unless otherwise stated. All organic solu-
tion from aqueous workups were dried by brief exposure to dried
scopic data identical to those previously reported[19] has magnesium sulfate, followed by gravity filtration. Column chroma-
tography was carried out using Merck Silica Gel 60 (230Ϫ400
mesh). TLC analyses were carried out using Merck silica gel 60
F254 pre-coated, aluminium-backed plates, which were visualised
using potassium permanganate or ammonium molybdenate sprays
or ultraviolet light.
previously been converted into (ϩ)-muscarine (1) in two
straightforward steps. Hence, this final approach represents
a formal synthesis of the natural target.
Methyl (R)-3-Hydroxyhept-5-ynoate (15): Propyne (ca. 8 mL) was
condensed under an atmosphere of dry nitrogen, dissolved in anhy-
drous THF (60 mL) and the solution cooled to Ϫ78 °C before the
addition of nBuLi (21.5 mL of a 1.6  solution in hexanes,
34.4 mmol). The resulting solution was stirred at this temperature
for 0.5 h, and boron trifluoride diethyl etherate (2.83 mL, 23 mmol)
added. After a further 10 min, methyl (S)-3,4-epoxybutanoate
(14)[35] (2.64 g, 23 mmol) was added. After stirring for a further
4 h at Ϫ78 °C, the solution was poured into saturated aqueous
ammonium chloride (200 mL) and the organic layer separated. The
aqueous layer was extracted with ethyl acetate (3 ϫ 50 mL) and
the combined organic solutions dried and evaporated to leave a
pale yellow oil which was chromatographed over silica gel (15%
diethyl ether in petroleum ether) to give the alkynol 15 (2.90 g, 80%)
as a colourless oil. [α]2D4 ϭ Ϫ13.2 (c ϭ 7.1, CH2Cl2). 1H NMR
(250 MHz, CDCl3): δ ϭ 1.79 (t, J ϭ 2.6 Hz, 3 H, 7-CH3),
2.38Ϫ2.40 (m, 2 H, 4-CH2), 2.47 (dd, J ϭ 16.2, 8.6 Hz, 1 H, 2-Ha),
2.71 (dd, J ϭ 16.2, 3.8 Hz, 1 H, 2-Hb), 3.71 (s, 3 H, OCH3),
4.12Ϫ4.20 (m, 1 H, 3-H) ppm. 13C NMR (67.5 MHz, CDCl3): δ ϭ
13.2 (7-CH3), 26.5 (4-CH2), 39.9 (2-CH2), 51.6 (OCH3), 66.7 (3-
CH), 74.3 (C), 78.4 (C), 172.7 (C ϭ O) ppm. IR (film): ν˜ ϭ 3452,
2221, 1736, 1672, 1439, 1162, 1059 cmϪ1. MS (EI): m/z ϭ 138 [Mϩ
Ϫ H2O, 7%], 103 (100), 81 (15), 71 (56), 61 (22). HRMS: calcd. for
C8H10O2 138.0681, found 138.0685.
Scheme 6. i) LiCϵCCH2OTBDPS, 12-crown-4, THF, Ϫ78 °C to 0
°C, 4 h, 65% (ϩ 9% of the syn isomer); ii) H2 (1 atm.), 5% Pd-
CaCO3, quinoline, MeOH, 20 °C, 0.25 h, 90%; iii) IBr, NaHCO3,
MeCN, Ϫ10 °C, 3.5 h, 68%; iv) H2 (1 atm.), 10% Pd-C, Et3N,
EtOAc, 20 °C, 16 h, 91%; v) NH4F, MeOH, 20 °C, 16 h, 56%
Conclusions
Methyl (3R,5Z)-3-Hydroxyhept-5-enoate (17): Quinoline (58 µL)
was added to a suspension of 5% palladium on barium sulfate
Three relatively short approaches to enantiomers of mus- (63 mg) in anhydrous ethyl acetate (2 mL). The suspension was
stirred for 10 min before addition of the alkynol 15 (0.33 g,
2.1 mmol) in ethyl acetate (2 mL). Stirring was continued under a
hydrogen atmosphere, with complete protection from light, until
gas absorption virtually ceased (ca. 0.75 h). The mixture was fil-
tered through Celite, the combined filtrate and washings washed
with 1  hydrochloric acid (10 mL), saturated aqueous sodium hy-
drogen carbonate (10 mL) and brine (10 mL), dried, and the sol-
vents evaporated to leave the (Z)-alkene 17 (0.29 g, 87%) as an oil.
[α]2D3 ϭ Ϫ22.7 (c ϭ 2.7, CH2Cl2). 1H NMR (250 MHz, CDCl3):
δ ϭ 1.84 (dd, J ϭ 7.3, 0.5 Hz, 3 H, 7-CH3), 2.45Ϫ2.35 (m, 2 H, 4-
CH2), 2.52 (dd, J ϭ 16.2, 8.7, 2-Ha), 2.76 (dd, J ϭ 16.2, 3.9 Hz, 2-
Hb), 3.25 (d, J ϭ 4.2 Hz, 1 H, OH), 3.73 (s, 3 H, OCH3), 4.12Ϫ4.21
(m, 1 H, 3-H), 5.40Ϫ5.46 (m, 1 H), 5.61Ϫ5.67 (m, 1 H) ppm. 13C
NMR (67.5 MHz, CDCl3): δ ϭ 13.0 (7-CH3), 34.2 (4-CH2), 40.7
(2-CH2), 51.8 (OCH3), 68.0 (3-CH), 125.4 (5-CH), 127.2 (6-CH),
173.3 (C ϭ O) ppm. IR (film): ν˜ ϭ 3417, 1732, 1438, 1371, 1259,
1202, 1163, 1058 cmϪ1. MS (EI): m/z ϭ 140 [Mϩ Ϫ H2O] (11%),
103 (100), 81 (22), 71 (95), 61 (41). HRMS: calcd. for C8H12O2
140.0837, found 140.0846.
carine (1) have been developed. The shortest begins with
methyl (S)-lactate and is complete in 9 steps, many of which
are simple and high yielding.
Experimental Section
General Remarks: NMR spectra were recorded using Bruker WM
or DPX spectrometers, operating at 250 MHz or 400 MHz, respec-
1
tively for H spectra and at 67.5 MHz or 100.6 MHz, respectively
for 13C spectra. Unless stated otherwise, NMR spectra were meas-
ured using dilute solutions in CDCl3. All NMR measurements were
carried out at 30 °C, and chemical shifts are measured relative to
tetramethylsilane (δ ϭ 0.00 ppm) or to the resonances of CDCl3
(δ ϭ 7.27 ppm for 1H and δ ϭ 77.0 ppm (the central line of the
triplet) for 13C). Low resolution mass spectra were obtained using a
VG Platform II Quadrupole spectrometer operating in the electron
impact (EI; 70 eV) or atmospheric pressure chemical ionisation
(ApcI) modes, as stated. High resolution mass spectrometry was
performed by the EPSRC Mass Spectrometry Service, University Methyl (2R,4S,5R)-(4-Hydroxy-5-methyltetrahydrofuran-2-yl)acet-
College, Swansea, using the ionisation modes specified. Optical ro-
tations were measured using a JASCO DIP 370 polarimeter. Melt-
ate (18): Sodium hydrogen carbonate (0.81 g, 9.6 mmol) was added
to an ice-cold solution of the (Z)-alkene 17 (0.45 g, 3.2 mmol) in
Eur. J. Org. Chem. 2004, 1973Ϫ1982
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1977

D. W. Knight, D. E. Shaw, E. R. Staples
FULL PAPER
anhydrous acetonitrile (7 mL) and the suspension stirred for 5 min.
Solid iodine (2.44 g, 9.6 mmol) was added and the resulting mixture
stirred at 0Ϫ5 °C for 72 h, with the exclusion of light. Diethyl ether
(30 mL) was added and the mixture washed with saturated aqueous
ϫ CH), 20.4 (5-CH3), 40.9 (3-CH2), 41.7 (1-CH2), 74.5 (2-CH),
78.6 (4-CH), 84.3 (5-CH), 176.0 (C ϭ O) ppm. IR (film): ν˜ ϭ 3368,
1717, 1507, 1419, 1163, 1109, 1040 cmϪ1. MS (EI): m/z ϭ 316 [Mϩ]
(11%), 273 [Mϩ Ϫ iPr] (24), 187 (100), 131 (58), 75 (64), HRMS:
sodium thiosulfate (120 mL). The separated aqueous phase was ex- calcd. for C13H25O4Si 273.1522, found 273.1517. C16H32O4Si:
tracted with diethyl ether (3 ϫ 50 mL). The combined organic solu-
tions were dried and evaporated and the residue filtered through a
plug of silica with the aid of diethyl ether to give the hydroxytetra-
hydrofuran 18 (0.35 g, 63%) as a colourless oil. [α]2D2 ϭ ϩ14.8 (c ϭ
0.4, CH2Cl2). 1H NMR (250 MHz, CDCl3): δ ϭ 1.27 (d, J ϭ
6.4 Hz, 3 H, 5-CH3), 1.87 (ddd, J ϭ 13.2, 9.4, 6.4 Hz, 1 H, 3-Ha),
2.02 (ddd, J ϭ 13.2, 6.0, 2.6 Hz, 1 H, 3-Hb), 2.52 (dd, J ϭ 15.6,
6.2 Hz, 1 H, 1Ј-Ha), 2.66 (dd, J ϭ 15.6, 6.9 Hz, 1Ј-Hb), 3.69 (s, 3
H, OCH3), 3.87 (qd, J ϭ 6.4, 3.4 Hz, 1 H, 5-H), 4.00Ϫ4.03 (m, 1
H, 2-H), 4.47 (ddd, J ϭ 9.4, 6.0, 3.4 Hz, 1 H, 4-H) ppm. 13C NMR
(67.5 MHz, CDCl3): δ ϭ 19.8 (5-CH3), 40.5 (3-CH2), 40.7 (1Ј-
CH2), 51.7 (OCH3), 74.0 (2-CH), 77.2 (4-CH), 82.1 (5-CH), 173.3
(C ϭ O) ppm. IR (film): ν˜ ϭ 3402, 1734, 1440, 1371, 1163, 1078
cmϪ1. MS (EI): m/z ϭ 156 [Mϩ Ϫ H2O] (5%), 101 (51), 98 (100), 74
(57), 57 (61). HRMS: calcd. for C8H12O3 156.0786, found 156.0791.
C8H14O4: calcd. C 55.17, H 8.05; found C 55.44, H 8.16.
calcd. 60.76, H 10.13; found C 60.93, H 10.21.
(2R,4S,5R)-2-Aminomethyl-5-methyl-4-(triisopropylsilyloxy)-
tetrahydrofuran (21): Triethylamine (0.06 mL, 0.41 mmol) was ad-
ded to a solution of the acid 20b (0.13 g, 0.41 mmol) in anhydrous
toluene (8 mL), followed by diphenylphosphoranyl azide
(0.088 mL, 0.41 mmol). The resulting solution was slowly warmed
to 80 °C, and then stirred at this temperature for 2 h. The cooled
solution was diluted with water (5 mL), stirred at ambient tempera-
ture for 16 h and extracted with chloroform (3 ϫ 10 mL). The com-
bined extracts were dried and the solvents evaporated to leave a
brown oil, chromatography of which over silica gel (30% EtOAc in
1
hexanes) gave the amine 21 (0.106 g, 90%) as a pale yellow oil. H
NMR (250 MHz, CDCl3): δ ϭ 1.10 (m, 21 H), 1.27 (d, J ϭ 6.4 Hz,
3 H, 5-CH3), 1.73 (ddd, J ϭ 13.1, 9.4, 6.2 Hz, 1 H, 3-Ha), 1.82
(ddd, J ϭ 13.1, 6.1, 2.7 Hz, 1 H, 3-Hb), 3.18 (dd, J ϭ 14.0, 6.1 Hz,
1 H, 2Ј-Ha), 3.49 (dd, J ϭ 14.0, 6.4 Hz, 1 H, 2Ј-Hb), 3.66 (qd, J ϭ
6.4, 3.1 Hz, 1 H, 5-H), 4.02 (ddd, J ϭ 6.2, 6.1, 3.1 Hz, 1 H, 4-H),
4.16 (dddd, J ϭ 9.4, 6.4, 6.1, 2.7 Hz, 1 H, 2-H) ppm. 13C NMR
(67.5 MHz, CDCl3): δ ϭ 12.0 (5-CH3), 17.9 (6 ϫ CH3), 19.8 (3 ϫ
CH), 38.0 (3-CH2), 44.4 (CH2NH2), 76.5 (CH), 78.1 (CH), 83.9 (5-
Methyl (2R,4S,5R)-[5-Methyl-4-(triisopropylsilyloxy)tetrahydrofu-
ran-2-yl]acetate (20a): 2,6-Lutidine (0.154 g, 1.44 mmol) was added
to a stirred solution of the hydroxytetrahydrofuran 18 (0.10 g,
0.57 mmol) in anhydrous dichloromethane (1.6 mL) at 0 °C. Triiso-
propylsilyl trifluoromethanesulfonate (0.23 mL, 0.86 mmol) was
added dropwise and the resulting mixture stirred at this tempera-
ture for 3 h. Brine (20 mL) was then added and the resulting mix-
ture extracted with dichloromethane (3 ϫ 25 mL). The combined
organic extracts were dried, the solvents evaporated, and the resi-
due chromatographed on silica gel (7% EtOAc in petroleum ether)
CH) ppm. IR (film): ν˜ ϭ 3430 cmϪ1. MS (EI): m/z ϭ 234 [Mϩ
iPr] (5%), 143 (100), 81 (93), 57 (70).
Ϫ
(2RS,4SR,5RS)-2-Chloromethyl-5-methyl-4-(triisopropylsilyloxy)-
tetrahydrofuran (24): A sample of the racemic acid (Ϯ)-20b (0.10 g,
0.32 mmol) was dissolved in anhydrous benzene (5 mL) and the
resulting solution stirred in an ice bath for 10 min before the ad-
dition of pyridine (1 drop) and N,N-dimethylformamide (1 drop).
Oxalyl chloride (0.15 mL, 0.32 mmol) was then added dropwise
and stirring continued for 3 h. The solvent was evaporated, the resi-
due dissolved in anhydrous benzene (5 mL) and the solution once
more evaporated to dryness. The crude acid chloride 22 was dis-
solved in anhydrous carbon tetrachloride (3 mL) and the resulting
solution added dropwise to a stirred and refluxing suspension of
sodium 2-mercaptopyridine N-oxide (57 mg, 0.38 mmol) in carbon
tetrachloride (15 mL) containing 4-(dimethylamino)pyridine
(3 mg). After 1.5 h at reflux, the suspension was cooled and filtered
through Celite. The combined filtrate and washings were evapo-
rated and the residue subjected to silica gel chromatography (5%
EtOAc in petroleum ether) to give the racemic chloride 24 (72 mg,
to give the silyl ether 20a (0.14 g, 74%) as a colourless oil, [α]2D0
ϭ
ϩ16.2 (c ϭ 1.6, CH2Cl2). 1H NMR (250 MHz, CDCl3): δ ϭ
1.09Ϫ1.11 (m, 21 H), 1.27 (d, J ϭ 6.4 Hz, 3 H, 5-CH3), 1.80 (ddd,
J ϭ 13.6, 9.3, 6.4 Hz, 1 H, 3-Ha), 2.05 (ddd, J ϭ 13.6, 5.9, 2.7 Hz,
1 H, 3-Hb), 2.54 (dd, J ϭ 15.3, 6.6 Hz, 1 H, 2-Ha), 2.65 (dd, J ϭ
15.3, 6.4 Hz, 1 H, 2-Hb), 3.67 (s, 3 H, OCH3), 3.85 (qd, J ϭ 6.4,
3.1 Hz, 1 H, 5-H), 4.05 (ddd, J ϭ 6.4, 3.1, 2.7 Hz, 1 H, 4-H), 4.50
(dddd, J ϭ 9.3, 6.6, 6.4, 5.9 Hz, 1 H, 2 -H) ppm. 13C NMR
(67.5 MHz, CDCl3): δ ϭ 12.0 (6 ϫ CH3), 17.9 (3 ϫ CH), 19.9 (5-
CH3), 40.6 (3-CH2), 41.2 (1Ј-CH2), 51.6 (OCH3), 74.1 (2-CH), 78.2
(4-CH), 83.4 (5-CH), 171.5 (C ϭ O) ppm. IR (film): ν˜ ϭ 1729,
1464, 1382, 1268, 1163, 1108, 1040 cmϪ1. MS (EI): m/z ϭ 287 [Mϩ
Ϫ iPr] (14%), 187 (100), 113 (21), 103 (31), 75 (35). HRMS: calcd.
for C14H27O4Si 287.1679, found 287.1669.
1
73%) as a clear, colourless oil. H NMR (250 MHz, CDCl3): δ ϭ
1.05Ϫ1.07 (m, 21 H), 1.20 (d, J ϭ 6.4 Hz, 3 H, 5-CH3), 1.93Ϫ1.96
(m, 2 H, 3-CH2), 3.59 (d, J ϭ 4.8 Hz, 2 H, CH2Cl), 3.95 (qd, J ϭ
6.4, 3.1 Hz, 1 H, 5-H), 4.07 (ddd, J ϭ 4.1, 3.6, 3.1 Hz, 1 H, 4-H),
4.39 (dtd, J ϭ 6.2, 4.8, 2.1 Hz, 1 H, 2-H) ppm. 13C NMR
(67.5 MHz, CDCl3): δ ϭ 12.0 (6 ϫ CH3), 17.9 (3 ϫ CH), 20.5 (5-
CH3), 38.8 (3-CH2), 47.2 (CH2ClO), 77.3 (CH), 77.9 (CH), 83.5 (5-
CH) ppm. IR (film): ν˜ ϭ 1402, 1371, 1249, 1106, 1058 cmϪ1. MS
(EI): m/z ϭ 265 [Mϩ (37Cl)-iPr] (8%) 263 [Mϩ (35Cl)-iPr] (29), 187
(77), 131 (100), 97 (40), 75 (33). HRMS: calcd. for C12H2435ClO2Si
263.1234, found 263.1225.
(2R,4S,5R)-(5-Methyl-4-(triisopropylsilyloxy)tetrahydrofuran-2-yl)-
acetic Acid (20b): The silyl ether 20a (1.50 g, 4.4 mmol) was added
to a cooled solution of potassium hydroxide (2.04 g, 36.4 mmol) in
methanol (18 mL). The resulting solution was stirred at ambient
temperature overnight, and the bulk of the solvent evaporated. The
residue was dissolved in water (20 mL) and the solution washed
with diethyl ether (25 mL). This mixture was then acidified with
ice-cold 2  hydrochloric acid (16 mL) and extracted with chloro-
form (3 ϫ 50 mL). The combined extracts were dried and the sol-
vents evaporated to leave the acid 20b (1.40 g, 97%) as a clear,
colourless oil. [α]2D0 ϭ ϩ17.1 (c ϭ 1.0, CH2Cl2). 1H NMR
(2RS,4SR,5RS)-2-Chloromethyl-4-hydroxy-5-methyltetrahydrofuran
(250 MHz, CDCl3): δ ϭ 1.03Ϫ1.05 (m, 21 H), 1.20 (d, J ϭ 6.2 Hz, (25a): Tetrabutylammonium fluoride (0.6 mL of a 1  solution in
3 H, 5-CH3), 1.80 (ddd, J ϭ 12.8, 9.6, 6.0 Hz, 1 H, 3-Ha), 2.02
(ddd, J ϭ 12.8, 5.8, 2.3 Hz, 1 H, 3-Hb), 2.65 (d, J ϭ 6.3 Hz, 2 H,
tetrahydrofuran, 0.60 mmol) was added to a solution of the racemic
chloride 24 (96 mg, 0.30 mmol) in anhydrous tetrahydrofuran
1Ј-CH2), 3.91 (qd, J ϭ 6.2, 2.9 Hz, 1 H, 5-H), 4.06 (ddd, J ϭ 6.0, (2 mL). The resulting solution was stirred for 40 h at ambient tem-
2.9, 2.3 Hz, 1 H, 4-H), 4.48 (dtd, J ϭ 9.6, 6.3, 5.8 Hz, 1 H, 2-H) perature; the solvent was then evaporated. The residue was sepa-
ppm. 13C NMR (67.5 MHz, CDCl3): δ ϭ 12.5 (6 ϫ CH3), 18.5 (3 rated by chromatography on silica gel (20% EtOAc in petroleum
1978
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 1973Ϫ1982

Two Asymmetric Approaches to Muscarine
FULL PAPER
ether) to give the racemic chloro alcohol 25a (30 mg, 60%) as a
clear, colourless oil. 1H NMR (250 MHz, CDCl3): δ ϭ 1.25 (d, J ϭ
(3R,4S)-4-(tert-Butyldiphenylsilyloxy)-1-phenyl-1-pentyn-3-ol (31a):
nBuLi (5.24 mL of a 2.5  solution in hexanes, 13 mmol) was ad-
6.2 Hz, 3 H, 5-CH3), 2.03Ϫ2.08 (m, 2 H, 3-CH2), 3.59 (d, J ϭ ded to anhydrous tetrahydrofuran (50 mL) and the stirred solution
5.0 Hz, 2 H, CH2Cl), 4.02Ϫ4.05 (m, 2 H, 4-and 5-H), 4.39 (tdd, cooled to Ϫ78 °C. 12-Crown-4 (2.29 g, 13 mmol) was added fol-
J ϭ 5.0, 2.8, 2.0 Hz, 1 H, 2-H) ppm. 13C NMR (67.5 MHz, lowed by the dropwise addition of a solution of phenylacetylene
CDCl3): δ ϭ 19.5 (5-CH3), 38.1 (3-CH2), 47.0 (CH2Cl), 77.2 (CH),
(1.34 g, 13 mmol) in tetrahydrofuran (2 mL). After 20 min at Ϫ78
77.3 (CH), 82.9 (5-CH) ppm. IR (film): ν˜ ϭ 3430, 1645, 1444, 1350 °C, a solution of the (S)-aldehyde 30[43] (3.00 g, 9.62 mmol) in tetra-
cmϪ1. MS (EI): m/z ϭ 115 [Mϩ Ϫ Cl] (17%), 101 (71), 71 (53), 70 hydrofuran (10 mL) was slowly added, the resulting solution slowly
(32), 57 (100). HRMS: calcd. for C6H11O2 115.0759; found
115.0757.
warmed to ambient temperature over 4 h, and then quenched by
the addition of saturated aqueous ammonium chloride (50 mL).
The resulting mixture was extracted with diethyl ether (3 ϫ 50 mL),
the combined extracts dried, and the solvents evaporated. Chroma-
tography of the residue (10% Et2O in petroleum ether) gave the
(3R,4S)-alkynol 31a. [α]2D0 ϭ Ϫ9.1 (c ϭ 1.0, CH2Cl2). 1H NMR
(400 MHz, CDCl3): δ ϭ 1.01 (s, 9 H, tBu), 1.11 (d, J ϭ 6.2 Hz, 3
H, 5-CH3), 2.55 (d, J ϭ 6.1 Hz, 1 H, OH), 3.99 (qd, J ϭ 6.2,
3.5 Hz, 1 H, 4-H), 4.41 (dd, J ϭ 6.1, 3.5 Hz, 1 H, 3-H), 7.17Ϫ7.21
(m, 3 H), 7.28Ϫ7.33 (m, 8 H), 7.62Ϫ7.65 (m, 4 H) ppm. 13C NMR
(100.6 MHz, CDCl3): δ ϭ 15.8 (5-CH3), 19.5 (CSi), 27.4 (3 ϫ CH3),
68.4 (2-C), 69.8 (3-CH), 72.9 (4-CH), 87.7 (1-C), 123.3 (C), 128.6
(2 ϫ CH), 130.4 (CH), 132.1 (2 ϫ CH), 136.3 (CH), 136.4 (2 ϫ
CH), 136.6 (2 ϫ C), 136.6 (2 ϫ C) ppm. IR (film): ν˜ ϭ 3308 (br.),
3070, 2857, 2361, 1619, 1443, 1259, 1112, 822 cmϪ1. MS (APcI):
m/z ϭ 415 [Mϩ ϩ H] (50%), 397 [Mϩ Ϫ H2O] (100%). HRMS
calcd. for C27H31O2Si 415.2093, found 415.20946.
(2R,4S,5R)-2-Iodomethyl-5-methyl-4-(triisopropylsilyloxy)-
tetrahydrofuran (25b): The acid (ϩ)-20b (82 mg, 0.26 mmol) was
dissolved in anhydrous benzene (3 mL) containing pyridine (1
drop) and N,N-dimethylformamide (1 drop). The resulting stirred
solution was cooled to 0 °C and treated dropwise with oxalyl chlo-
ride (0.12 mL, 0.26 mmol). After 3 h at 0 °C, the solvent was evapo-
rated, the residue dissolved in anhydrous benzene (5 mL) and the
solution evaporated. The residue was dissolved in anhydrous cyclo-
hexene (4 mL) and the solution added to refluxing cyclohexene
(10 mL) containing iodoform (112 mg, 0.28 mmol), 2-mercaptopyr-
idine N-oxide sodium salt (44 mg, 0.28 mmol), and 4-(dimethylami-
no)pyridine (3 mg). After 15 h at reflux, the resulting suspension
was cooled and filtered through Celite. The filtrate and washings
were combined and the solvents evaporated. Chromatography of
the residue on silica gel (5% EtOAc in petroleum ether) gave the
iodide 25b (54 mg, 60%) as a clear colourless oil. [α]1D8 ϭ ϩ24.6 (c ϭ
8.7, CH2Cl2). 1H NMR (250 MHz, CDCl3): δ ϭ 1.00Ϫ1.02 (m, 21
H), 1.25 (d, J ϭ 6.3 Hz, 3 H, 5-CH3), 1.74 (ddd, J ϭ 12.7, 6.5,
2.9 Hz, 1 H, 3-Ha), 1.93 (ddd, J ϭ 12.7, 6.0, 2.9 Hz, 1 H, 3-Hb),
3.22 (d, J ϭ 4.8 Hz, 2 H, CH2I), 3.92 (qd, J ϭ 6.3, 3.2 Hz, 1 H, 5-
H), 4.02 (ddd, J ϭ 6.0, 3.2, 2.9 Hz, 1 H, 4-H), 4.39 (dtd, J ϭ 6.5,
4.8, 2.9 Hz, 1 H, 2-H) ppm. 13C NMR (67.5 MHz, CDCl3): δ ϭ
11.0 (CH2I), 12.0 (6 ϫ CH3), 17.9 (3 ϫ CH), 19.9 (5-CH3), 41.8
(3-CH2), 77.0 (CH), 78.1 (CH), 83.9 (5-CH) ppm. IR (film): ν˜ ϭ
1445, 1250, 1156, 1071 cmϪ1. MS (EI): m/z ϭ 355 [Mϩ Ϫ iPr] (3%),
187 (100), 141 (35), 127 (40), 75 (15). HRMS: calcd. for
C12H24IO2Si 355.0590, found 355.0582.
Approximately 11% of the (3S,4S)-isomer was present in the
1
sample, as determined by resonances in the H NMR spectrum at
δH ϭ 3.94 (4-H) and 4.33 (3-H) ppm.
(2S,3R)-5-Phenyl-4-pentyne-2,3-diol (31b): Tetrabutylammonium
fluoride (TBAF; 16.4 mL of a 1  solution in tetrahydrofuran,
16.4 mmol) was added to a stirred solution of the alkynol 31a
(3.40 g, 8.21 mmol) in tetrahydrofuran (100 mL). The resulting
solution was stirred for 2 h at ambient temperature, diluted with
dichloromethane (150 mL) and washed with brine (60 mL). The
resulting organic solution was dried and the solvents evaporated to
leave a brown oil which was purified by column chromatography
(EtOAc/petroleum ether, 1:1) to give the alkynediol 31b (1.34 g,
93%) as a colourless solid, m.p. 65Ϫ68 °C. 1H NMR (400 MHz,
CDCl3): δ ϭ 1.15 (d, J ϭ 6.4 Hz, 3 H, 1-CH3), 2.00 (d, J ϭ 6.7 Hz,
1 H, OH), 2.44 (d, J ϭ 6.7 Hz, 1 H, OH), 3.92 (app. quintd, J ഠ
6.4, 3.7 Hz, 1 H, 2-H), 4.47 (dd, J ϭ 6.7, 3.7 Hz, 1 H, 3-H),
7.08Ϫ7.12 (m, 3 H), 7.24Ϫ7.27 (m, 2 H) ppm. 13C NMR
(100.6 MHz, CDCl3): δ ϭ 23.1 (1-CH3), 68.0 (3-CH), 70.8 (2-CH),
86.5 (4-C), 86.9 (5-C), 122.5 (C), 128.7 (2 ϫ CH), 129.1 (CH), 132.2
(2 ϫ CH) ppm. IR (nujol): ν˜ ϭ 3243 (br.), 2348, 1489, 1426, 1084,
996, 914, 755, 689 cmϪ1. MS (ApcI): m/z ϭ 159 [Mϩ ϩ H Ϫ H2O]
(100%). The minor (2S,3S)-isomer (ca. 12%) was evident from the
1H NMR spectrum: δH ϭ 2.38 (OH), 3.88 (2-H), and 4.27 (3-H).
(2R,4S,5R)-4-Hydroxy-2-iodomethyl-5-methyltetrahydrofuran (25c):
A solution of the iodide 25b (64 mg, 0.20 mmol) in tetrahydrofuran
(1 mL) was stirred with tetrabutylammonium fluoride (0.4 mL of a
1  solution in tetrahydrofuran, 0.40 mmol) at ambient tempera-
ture for 40 h. The solvent was evaporated and the residue chroma-
tographed on silica gel (20% EtOAc in petroleum ether) to give the
iodo alcohol 25c (23 mg, 69%) as a clear, colourless oil. [α]1D8
ϭ
ϩ20.5 (c ϭ 2.0, CHCl3) [ref.[16] [α]D ϭ Ϫ26.5 (c ϭ 1.8, CHCl3)].
1H NMR (250 MHz, CDCl3): δ ϭ 1.26 (d, J ϭ 6.4 Hz, 3 H, 5-
CH3), 1.91 (ddd, J ϭ 13.3, 8.6, 6.2 Hz, 1 H, 3-Ha), 2.05 (ddd, J ϭ
13.3, 6.2, 3.0 Hz, 1 H, 3-Hb), 3.34 (dd, J ϭ 10.2, 6.2 Hz, 1 H,
CHaHbI), 3.41 (dd, J ϭ 10.2, 4.9 Hz, 1 H, CHaHbI), 3.97 (qd, J ϭ
6.4, 3.2 Hz, 1 H, 5-H), 4.05 (ddd, J ϭ 8.6, 6.2, 3.2 Hz, 1 H, 4-H),
4.16 (tdd, J ϭ 6.2, 4.9, 3.0 Hz, 1 H, 2-H) ppm. 13C NMR
(67.5 MHz, CDCl3): δ ϭ 10.4 (CH2I), 19.8 (5-CH3) 41.1 (3-CH2),
77.2 (CH), 77.6 (CH), 83.3 (5-CH) ppm. MS (EI): m/z ϭ 242 [Mϩ]
(2%), 115 (31), 101 (100), 71 (70).
(2S,3S,4R,5R)-4-Iodo-2-methyl-5-phenyltetrahydro-3-furanol (33):
A stirred solution of the alkyne diol 31b (2.11 g, 12 mmol) in hex-
ane (60 mL) containing methanol (1 mL), quinoline (0.5 mL) and
5% palladium on calcium carbonate (300 mg) was stirred under one
atmosphere of hydrogen for 2 h, by which time 270 mL of gas had
been absorbed. The mixture was filtered through Celite and the
solid washed with dichloromethane. Evaporation of the combined
filtrates gave (2S,3R,4Z)-5-phenyl-4-pentene-2,3-diol (32) contami-
(؊)-Muscarine Iodide (25d): Trimethylamine (0.05 mL) was added
to a solution of the iodo alcohol 25c (10 mg) in ethanol (0.2 mL), nated with quinoline which was immediately used in the following
the resulting solution heated at 70 °C for 4 h in a sealed tube, and
then allowed to cool slowly overnight. The solvent was evaporated
to leave a solid residue which was crystallised from acetone to give
a pure sample of (Ϫ)-muscarine iodide (25d) as an unstable colour-
less solid, m.p. 135Ϫ137 °C [ref.[16] m.p. 138Ϫ142 °C].
iodocyclisation without further purification. 32: 1H NMR
(400 MHz, CDCl3): δ ϭ 1.06 (d, J ϭ 6.4 Hz, 3 H, 1-CH3), 3.84
(qd, J ϭ 6.4, 3.6 Hz, 1 H, 2-H), 4.45 (dd, J ϭ 9.5, 3.6 Hz, 1 H, 3-
H), 5.69 (dd, J ϭ 11.7, 9.5 Hz, 1 H, 4-H), 6.56 (d, J ϭ 11.7 Hz, 1
H, 5-H), 7.13Ϫ7.20 (m, 3 H), 7.21Ϫ7.23 (m, 2 H) ppm. 13C NMR
Eur. J. Org. Chem. 2004, 1973Ϫ1982
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1979

D. W. Knight, D. E. Shaw, E. R. Staples
FULL PAPER
(100.6 MHz, CDCl3): δ ϭ 19.5 (1-CH3), 74.5 (3-CH), 75.6 (2-CH), m/z ϭ 221 [Mϩ ϩ H] (80%), 161 (100). HRMS: calcd. for C13H17O3
129.0 (CH), 129.3 (2 ϫ CH), 130.1 (2 ϫ CH), 134.0 (4-CH), 136.9 221.1178, found 221.1181.
(5-CH), 142.5 (C) ppm. MS (APcI): m/z ϭ 179 [Mϩ ϩ H] (20%),
(4R,5S)-5-(tert-Butyldimethylsilyloxy)-1-(tert-butyldiphenyl-
160 [Mϩ Ϫ H2O] (100%). A proton resonance at δ ϭ 5.55 (4-H) was
silyloxy)hex-2-yn-4-ol (37): nBuLi (5.3 mL of a 2.5 solution in hex-
used to show the expected presence of ca. 11% of the (3S)-isomer.
anes, 13 mmol) was added to anhydrous tetrahydrofuran (50 mL),
stirred and maintained at Ϫ78 °C. 12-Crown-4 (2.29 g, 13 mmol)
was then added followed by the dropwise addition of a solution
of O-tert-butyldiphenylsilylpropargyl alcohol (3.89 g, 13 mmol) in
The crude (Z)-alkenediol 32 (ca. 12 mmol) was dissolved in anhy-
drous acetonitrile (40 mL) maintained at 0 °C. Sodium hydrogen
carbonate (3.02 g, 36 mmol) was then added and the resulting sus-
pension stirred for 5 min before the addition of solid iodine (9.13 g,
tetrahydrofuran (2 mL). After 20 min at Ϫ78 °C, the (S)-O-silyllac-
taldehyde 36[43] (1.80 g, 9.56 mmol) was added and the resulting
36 mmol) in one portion. After 4 h at 0 °C, the mixture was de-
solution slowly warmed to ambient temperature over 4 h. The reac-
colourised by the addition of saturated aqueous sodium thiosulfate
tion mixture was quenched with saturated aqueous ammonium
(40 mL) and extracted with diethyl ether (2 ϫ 80 mL). The com-
bined extracts were washed with 1  hydrochloric acid (90 mL) and
chloride (50 mL) and extracted with diethyl ether (3 ϫ 50 mL). The
combined extracts were dried and the solvents evaporated to leave
water (40 mL), dried, and the solvents evaporated. The brown oily
a yellow oil which was purified by chromatography (10% EtOAc in
residue was separated by chromatography (10% EtOAc in petro-
petroleum ether) to give the alkynol 37 (2.98 g, 65%) as a pale yel-
leum ether) to give the iodotetrahydrofuran 33 (1.86 g, 51% for the
1
low oil. H NMR (400 MHz, CDCl3): δ ϭ 0.02 (s, 6 H), 0.81 (s, 9
two steps) as a light-sensitive orange oil. 1H NMR (400 MHz,
H), 0.97 (s, 9 H), 1.10 (d, J ϭ 6.2 Hz, 3 H, 6-CH3), 2.21 (d, J ϭ
CDCl3): δ ϭ 1.52 (d, J ϭ 6.5 Hz, 3 H, 2-CH3), 3.22 (br. s, 1 H, 3-
5.0 Hz, 1 H, OH), 3.77 (qd, J ϭ 6.2, 3.9 Hz, 1 H, 5-H), 4.16 (br.
OH), 3.88 (qd, J ϭ 6.5, 2.8 Hz, 1 H, 2-H), 4.27 (dd, J ϭ 4.4, 2.8 Hz,
signal, 1 H, 4-H), 4.28 (d, J ϭ 1.4 Hz, 2 H, 1-CH2), 7.33Ϫ7.35 (m,
1 H, 4-H), 4.46 (m, 1 H, 3-H), 4.56 (d, J ϭ 4.4 Hz, 1 H, 5-H),
6 H), 7.61Ϫ7.64 (m, 4 H) ppm. 13C NMR (100.6 MHz, CDCl3):
7.17Ϫ7.24 (m, 3 H), 7.33Ϫ7.37 (m, 2 H) ppm. 13C NMR
δ ϭ Ϫ1.0 (2 ϫ CH3), 18.5 (6-CH3), 19.6 (C), 20.3 (C), 26.2 (3 ϫ
(100.6 MHz, CDCl3): δ ϭ 18.9 (2-CH3), 39.2 (4-CH), 79.6 (5-CH),
CH3), 27.0 (3 ϫ CH3), 53.1 (1-CH2), 67.4 (5-CH), 71.4 (4-CH),
80.8 (2-CH), 84.5 (3-CH), 124.5 (CH), 126.3 (2 ϫ CH), 126.5 (2 ϫ
83.3 (3-C), 84.8 (2-C), 128.1 (4 ϫ CH), 130.2 (2 ϫ CH), 135.2 (2
CH), 139.1 (C) ppm. IR (film): ν˜ ϭ 3391 cmϪ1. MS (APcI): m/z ϭ
ϫ C), 136.0 (4 ϫ CH) ppm. IR (film): ν˜max ϭ 3444, 2360, 1472,
305 [Mϩ ϩ H] (100%).
1428, 1255, 1112, 1007 cmϪ1. MS (APcI): m/z ϭ 483 [Mϩ ϩ H]
(50%), 465 [Mϩ ϩ H Ϫ H2O] (100). HRMS (NH4CI): calcd. for
(2S,3S,5S)-2-Methyl-5-phenyltetrahydro-3-furanol (34a): The iodo-
C28H46NO3Si 500.3016 (Mϩ ϩ NH4), found 500.3012. The minor
tetrahydrofuran 33 (0.44 g, 1.46 mmol) and 10% palladium on car-
bon (130 mg) were stirred in ethyl acetate (10 mL) containing tri-
ethylamine (0.40 mL, 2.89 mmol) under an atmosphere of hydrogen
(4S,5S)-isomer (ca. 9%) was visible at δH ϭ 3.70 (5-H) and 3.96
(4-H).
for 16 h, by which time 32.7 mL of gas had been absorbed. The
resulting suspension was filtered through a mixture of Celite and
silica and the solid washed with ethyl acetate. Evaporation of the
(4R,5S,2Z)-5-(tert-Butyldimethylsilyloxy)-1-(tert-butyldiphenyl-
silyloxy)hex-2-en-4-ol (38): The alkynol 37 (1.22 g, 2.52 mmol) and
quinoline (0.09 mL, 0.76 mmol) were stirred in methanol (10 mL)
containing 5% palladium on calcium carbonate (0.05 g) under hy-
drogen until ca. 56 mL (2.52 mmol) had been absorbed (0.25 h).
The resulting suspension was filtered through Celite and the solid
washed with diethyl ether. The combined filtrates were washed with
1  hydrochloric acid (40 mL) and brine (40 mL), dried, and the
combined filtrates left the tetrahydrofuranol 34a (0.26 g, 98%) as a
colourless solid, m.p. 80Ϫ81 °C. [α]2D0 ϭ Ϫ62.0 (c ϭ 2.0, CH2Cl2).
1H NMR (400 MHz, CDCl3): δ ϭ 1.15 (d, J ϭ 6.4 Hz, 3 H, 2-
CH3), 1.65 (d, J ϭ 3.8 Hz, 1 H, 3-OH), 1.84 (ddd, J ϭ 13.3, 10.0,
3.1 Hz, 1 H, 4-Ha), 2.02 (ddd, J ϭ 13.3, 5.8, 2.2 Hz, 1 H, 4-Hb),
3.85 (qd, J ϭ 6.4, 2.9 Hz, 1 H, 2-H), 3.93 (app. br. d, J ϭ ca.
solvents evaporated to give the (Z)-alkene 38 (1.09 g, 90%) as a
3.0 Hz, 1 H, 3-H), 4.94 (dd, J ϭ 10.0, 5.8 Hz, 1 H, 5-H), 7.04Ϫ7.09
1
colourless oil. H NMR (400 MHz, CDCl3): δ ϭ Ϫ0.04 (s, 3 H),
(m, 1 H), 7.12Ϫ7.15 (m, 4 H) ppm. 13C NMR (100.6 MHz,
Ϫ0.03 (s, 3 H), 0.85 (s, 9 H), 1.01 (d, J ϭ 6.2 Hz, 3 H, 6-CH3),
CDCl3): δ ϭ 20.3 (2-CH3), 44.1 (4-CH2), 78.4 (3-CH), 80.2 (5-CH),
1.05 (s, 9 H), 3.67 (qd, J ϭ 6.2, 3.8 Hz, 1 H, 5-H), 4.08 (br. dd,
83.6 (2-CH), 126.3 (2 ϫ CH), 127.9 (CH), 128.8 (2 ϫ CH), 142.5
J ϭ 8.2, 3.8 Hz, 4-H), 4.21 (dd, J ϭ 13.3, 5.4 Hz, 1 H, 1-Ha), 4.32
(C) p.p.m. IR (nujol): ν˜ ϭ 3363, 1488, 1452, 1356, 1323 cmϪ1
.
(dd, J ϭ 13.3, 6.8 Hz, 1 H, 1-Hb), 5.44 (dd, J ϭ 8.3, 8.2 Hz, 1 H,
3-H), 5.79 (ddd, J ϭ 8.3, 6.8, 5.4 Hz, 1 H, 2-H), 7.36Ϫ7.45 (m, 6
H), 7.71Ϫ7.74 (m, 4 H) ppm. 13C NMR (100.6 MHz, CDCl3): δ ϭ
Ϫ1.0 (2 ϫ CH3), 15.7 (6-CH3), 18.4 (C), 19.5 (C), 26.2 (3 ϫ CH3),
27.0 (3 ϫ CH3), 61.0 (1-CH2), 71.6 (5-CH), 72.1 (4-CH), 128.0 (4
ϫ CH), 130.1 (2 ϫ CH), 132.6 (2-CH), 135.1 (2 ϫ C), 135.1 (2 ϫ
C), 135.2 (3-CH), 136.0 (4 ϫ CH) ppm. IR (film): ν˜ ϭ 3444, 1590,
1472, 1428, 1255, 1112 cmϪ1. MS (APcI): m/z ϭ 485 [Mϩ ϩ H]
(10%), 467 [Mϩ ϩ H Ϫ H2O] (100). HRMS: calcd. for C28H45O3Si2
485.2907, found 485.2904. C28H44O3Si2: calcd. C 69.37, H, 9.15;
found C 68.94, H 9.12.
HRMS: calcd. for C11H15O2 [Mϩ ϩ H] 179.1072, found 179.1071.
C11H14O2: calcd. C 74.12, H 7.92; found C 74.07, 7.73.
(2S,3R,5S)-3-Acetyloxy-2-methyl-5-phenyltetrahydrofuran
(34b):
Acetic anhydride (0.14 mL, 1.53 mmol) was added to a stirred solu-
tion of the tetrahydrofuranol 34a (0.31 g, 1.40 mmol) in anhydrous
pyridine (1 mL). The resulting solution was stirred at ambient tem-
perature overnight, diluted with diethyl ether (20 mL), washed with
2  aqueous copper() sulfate (20 mL) and brine (20 mL), dried,
and the solvents evaporated to leave the acetate 34b (0.30 g, 97%)
1
as a colourless oil. H NMR (400 MHz, CDCl3): δ ϭ 1.33 (d, J ϭ
6.5 Hz, 3 H, 2-CH3), 1.86Ϫ2.01 (m, 1 H, 4-Ha), 2.05 (s, 3 H), 2.21 (2R,3R,4S,5S)-2-O-(tert-Butyldiphenylsilylmethyl)-3-iodo-5-methyl-
(app. dd, J ϭ 13.7, 5.3 Hz, 1 H, 4-Hb), 4.07 (qd, J ϭ 6.5, 2.3 Hz, tetrahydrofuran-4-ol (39): Sodium hydrogen carbonate (0.46 g,
1 H, 2-H), 4.91 (app. br. d, J ϭ ca. 6.0 Hz, 3-H), 4.99 (dd, J ϭ
5.45 mmol) was added to a stirred solution of the (Z)-alkene 38
10.6, 5.3 Hz, 1 H, 5-H), 7.19Ϫ7.22 (m, 1 H), 7.26Ϫ7.29 (m, 4 H) (0.88 g, 1.8 mmol) in anhydrous acetonitrile (40 mL) maintained at
ppm. 13C NMR (100.6 MHz, CDCl3): δ ϭ 18.9 (2-CH3), 20.2 Ϫ10 °C. After 5 min, iodine monobromide (0.75 g, 3.63 mmol) was
(CH3), 39.7 (4-CH2), 78.7 (3-CH), 80.0 (2-CH), 82.1 (5-CH), 124.9 added in one portion and stirring continued at Ϫ10 °C for 3.5 h
(2 ϫ CH), 126.7 (CH), 127.4 (2 ϫ CH), 140.2 (C), 169.9 (C ϭ O) (TLC monitoring). Saturated aqueous sodium thiosulfate (20 mL)
ppm. IR (film): ν˜ ϭ 1724, 1444, 1377, 999, 912 cmϪ1. MS (APcI):
and dichloromethane (40 mL) were added and the resulting layers
1980
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 1973Ϫ1982

Two Asymmetric Approaches to Muscarine
FULL PAPER
ˆ
separated. The aqueous layer was extracted with dichloromethane
(40 mL), the combined organic solutions washed with brine
(40 mL), dried, and the solvents evaporated. Chromatography (10%
EtOAc in petroleum ether) of the residue gave the iodotetrahydrofu-
ran 39 (0.45 g, 68%) as a pale yellow oil. [α]2D0 ϭ ϩ18.2 (c ϭ 1.2,
and the EPSRC, Rhone-Poulenc-Rorer, and Cardiff University for
financial support.
[1]
M. McKenny, The Savory Wild Mushroom, University of
Washington Press, Seattle, 1976.
P. E. Furst, Mushrooms: Encyclopedia of Psychoactive Drugs
1
[2]
CH2Cl2). H NMR (400 MHz, CDCl3): δ ϭ 0.93 (s, 9 H), 1.34 (d,
J ϭ 6.6 Hz, 3 H, 5-CH3), 1.55 (app. br. s, 1 H, OH), 3.65 (dd, J ϭ
10.5, 5.8 Hz, 1 H, 1Ј-Ha), 3.69Ϫ3.73 (m, 1 H, 2-H), 3.70 (qd, J ϭ
6.6, 3.6 Hz, 5-H), 3.82 (dd, J ϭ 10.5, 4.3 Hz, 1 H, 1Ј-Hb), 4.19 (dd,
J ϭ 4.3, 3.6 Hz, 1 H, 3-H), 4.34 (app. br. t, J ca. 3.9 Hz, 1 H,
4-H), 7.27Ϫ7.38 (m, 6 H), 7.60Ϫ7.68 (m, 4 H) ppm. 13C NMR
(100.6 MHz, CDCl3): δ ϭ 19.7 (5-CH3), 20.5 (C), 27.3 (3 ϫ CH3),
36.0 (3-CH), 68.3 (1Ј-CH2), 77.8 (4-CH), 79.4 (5-CH), 82.1 (2-CH),
128.1 (4 ϫ CH), 130.2 (2 ϫ CH), 134.7 (2 ϫ C), 136.1 (4 ϫ CH)
ppm. IR (film): ν˜ ϭ 3425, 1471, 1428, 1261 cmϪ1. MS (APcI):
m/z ϭ 419 (Mϩ Ϫ Ph, 40%), 291 (Mϩ Ϫ Ph-HI, 100%). HRMS
(NH4CI): calcd. for C22H33INO3Si 514.1274 [M ϩ NH4ϩ], found
514.1271. C22H29IO3Si: calcd. C 53.23, H 5.89; found C 53.08, H
6.12.
(Ed.: S. H. Snyder), Burke Publishing, London, 1986.
[3]
´
P. C. Want, M. M. Joullie, The Alkaloids (Ed.: A. Brossi), Aca-
demic Press, 1984, vol. 23, p. 327.
[4]
[5]
H. Braconnot, Ann. Chim. (Paris) 1811, 79, 265.
D. L. Sayers, R. Eustace, The Documents in the Case, Harper
and Row, New York, 1930.
[6]
[7]
[8]
[9]
K. Bowden, G. A. Mogey, J. Pharm. Pharmacol. 1958, 10, 145.
C. H. Eugster, P. G. Waser, Experientia 1954, 10, 298.
F. Jellnick, Acta Crystallogr. 1957, 10, 277.
M. P. Caulfield, Pharmacol. Ther. 1993, 319; M. De Amici, C.
Pallgroce, P. Angeli, G. Marucci, F. Cantalaness, C. D. Micheli,
Il Farmco 2000, 55, 535.
Y. Karton, B. J. Bradbury, J. Baumgold, R. Peak, K. A. Ja-
cobsen, J. Med. Chem. 1991, 34, 2133; R. Quirion, I. Aubert,
P. A. Lapchak, P. Schaum, S. Teolis, S. Gauthier, D. M. Araujo,
Trends Pharmacol. Sci. 1989, Suppl. S, 80.
P. J. Pauling, F. G. Canepa, Nature 1966, 210, 904; C. C. J.
Culvenor, N. S. Ham, J. Chem. Soc., Chem. Commun. 1966,
537; D. B. Davies, S. S. Danylak, D. M. Brown, Tetrahedron
1982, 38, 41.
E. Hardegger, F. Lohse, Helv. Chim. Acta 1957, 40, 2383; H.
C. Cox, E. Hardegger, F. Kögl, P. Liechti, F. Lohse, C. A. Sale-
mink, Helv. Chim. Acta 1958, 41, 229.
E. Hardegger, H. Furter, J. Kiss, Helv. Chim. Acta 1958, 41,
2401.
S. Pochet, T. Huynh-Dinh, J. Org. Chem. 1982, 47, 193.
A. M. Mabarak, D. M. Brown, J. Chem. Soc., Perkin Trans. 1.
1982, 809.
J. Mulzer, A. Angermann, W. Münch, G. Schlichthörl, A.
Hentzschel, Liebigs Ann. Chem. 1987, 7.
P. Popsavin, M. Popsavin, L. Radic, O. Beric, C. Cirin-Novta,
Tetrahedron Lett. 1999, 40, 9305.
A. Bandzouzi, Y. Chapleur, J. Chem. Soc., Perkin Trans. 1
1987, 661.
S. J. Mantell, G. W. J. Fleet, D. Brown, J. Chem. Soc., Perkin
Trans. 1. 1992, 3023.
J. Whiting, Y. K. Au-Yang, B. Belleau, Can. J. Chem. 1972,
50, 3322.
W. C. Still, J. A. Schneider, J. Org. Chem. 1980, 45, 3375. See
also: T. Matsumotu, A. Ichihara, N. Ito, Tetrahedron 1969,
25, 5889.
R. Amouroux, B. Gerin, M. Chastrette, Tetrahedron Lett. 1982,
23, 4341; T. H. Chan, C. J. Li, Can. J. Chem. 1992, 70, 2726.
S. R. Angle, N. A. Said, J. Am. Chem. Soc. 2002, 124, 3608.
M. C. Pirrung, C. V. De Amicis, Tetrahedron Lett. 1988, 29,
159.
S. Takano, Y. Iwabuchi, K. Ogasawara, J. Chem. Soc., Chem.
Commun. 1989, 1371.
J. Adams, M. A. Poupart, L. Genier, Tetrahedron Lett. 1989,
30, 1753.
[10]
[11]
(2S,4R,5S)-2-O-(tert-Butyldiphenylsilylmethyl)-5-methyltetrahydro-
furan-4-ol (40a): The iodotetrahydrofuran 39 (0.45 g, 1.22 mmol),
triethylamine (0.34 mL, 2.44 mmol), and 10% palladium on carbon
(0.11 g) were stirred in ethyl acetate (10 mL) at ambient tempera-
ture under an atmosphere of hydrogen for 16 h. The resulting slurry
was filtered through Celite and the solid washed with ethyl acetate.
The combined filtrate and washings were diluted with diethyl ether
(20 mL), the solution washed with 1  hydrochloric acid (40 mL)
and brine (20 mL), dried, and the solvents evaporated to give the
tetrahydrofuran 40a (0.33 g, 91%) as a colourless oil. [α]2D0 ϭ ϩ4.3
(c ϭ 1.5, CH2Cl2). 1H NMR (400 MHz, CDCl3): δ ϭ 0.98 (s, 9 H),
1.15 (d, J ϭ 6.4 Hz, 3 H, 5-CH3), 1.81 (ddd, J ϭ 10.2, 6.7, 3.5 Hz,
1 H, 3-Ha), 2.04Ϫ2.12 (m, 1 H, 3-Hb), 3.59 (dd, J ϭ 10.8, 6.7 Hz,
1Ј-Ha), 3.64 (dd, J ϭ 10.8, 4.2 Hz, 1 H, 1Ј-Hb), 3.83 (qd, J ϭ 6.4,
3.6 Hz, 1 H, 5-H), 3.94 (app. br. td, J ϭ 6.7, 3.6 Hz, 1 H, 4-H),
4.15Ϫ4.21 (m, 1 H, 2-H), 7.29Ϫ7.37 (m, 6 H), 7.60Ϫ7.66 (m, 4 H)
ppm. 13C NMR (100.6 MHz, CDCl3): δ ϭ 18.0 (C), 18.2 (5-CH3),
25.8 (3 ϫ CH3), 35.4 (3-CH2), 65.0 (1Ј-CH2), 75.7 (4-CH), 76.5 (2-
CH), 81.3 (5-CH), 126.7 (4 ϫ CH), 128.6 (2 ϫ CH), 133.8 (2 ϫ
C), 134.6 (4 ϫ CH) ppm. IR (film): ν˜ ϭ 3397, 1472, 1428, 1390,
1260, 1111 cmϪ1. MS (APcI): 293 (Mϩ Ϫ Ph, 100%). HRMS
(NH4CI): calcd. for C22H34NO3Si 388.2308 (M ϩ NH4ϩ), found
388.2312.
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
(2S,4R,5S)-(4-Hydroxy-5-methyltetrahydrofuran-2-yl)methanol
(40b): Ammonium fluoride (0.27 g, 7.17 mmol) was added to a
solution of the tetrahydrofuran 40a (0.29 g, 1.19 mmol) in meth-
anol (5 mL), the resulting solution stirred at ambient temperature
for 16 h, and then evaporated. The resulting brown solid was sepa-
rated by column chromatography (10% MeOH in CH2Cl2) to give
the diol 40b (0.09 g, 56%) as a colourless oil. [α]2D0 ϭ Ϫ5.8 (c ϭ 0.2,
CHCl3), (ref.[19] [α]2D0 ϭ Ϫ6.0 (c ϭ 0.5, CHCl3). 1H NMR
(400 MHz, CDCl3): δ ϭ 1.17 (d, J ϭ 6.4 Hz, 3 H, 5-CH3), 1.54 (br.
s, 2 H, 2 ϫ OH), 1.76 (ddd, J ϭ 13.2, 6.5, 3.5 Hz, 1 H, 3-Ha), 1.96
(ddd, J ϭ 13.2, 9.1, 6.4 Hz, 3-Hb), 3.43 (dd, J ϭ 11.9, 4.9 Hz, 1 H,
1Ј-Ha), 3.70 (dd, J ϭ 11.9, 3.0 Hz, 1 H, 1Ј-Hb), 3.85 (qd, J ϭ 6.4,
3.5 Hz, 1 H, 5-H), 3.93 (m, 1 H, 4-H), 4.17Ϫ4.21 (m, 1 H, 2-H)
ppm. 13C NMR (100.6 MHz, CDCl3): δ ϭ 18.9 (5-CH3), 35.1 (3-
CH2), 63.7 (1Ј-CH2), 76.6 (4-CH), 77.7 (2-CH), 81.8 (5-CH) ppm.
HRMS: calcd. for C6H13O3 [Mϩ ϩ H] 133.0865, found 133.0864.
These data correspond to those previously reported.[19]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
K. H. Kang, M. Y. Cha, A. N. Pae, K. I. Choi, Y. S. Cho, H.
Y. Koh, B. Y. Chung, Tetrahedron Lett. 2000, 41, 4137.
`
M. De Amici, C. De Micheli, G. Molteni, D. Pitre, G. Carrea,
S. Riva, S. Spezia, L. Zetta, J. Org. Chem. 1991, 56, 67.
G. Fronza, C. Fuganti, P. Grasseli, Tetrahedron Lett. 1978,
3941.
G. Shapiro, D. Benchler, S. Hennet, Tetrahedron Lett. 1990,
31, 5733.
M. E. Hopkins, L. E. Overman, G. M. Rishton, J. Am. Chem.
Soc. 1991, 113, 5354.
S. B. Bedford, K. E. Bell, F. Bennett, C. J. Hayes, D. W. Knight,
D. E. Shaw, J. Chem. Soc., Perkin Trans. 1 1999, 2143.
F. Bennett, S. B. Bedford, K. E. Bell, G. Fenton, D. W. Knight,
D. E. Shaw, Tetrahedron Lett. 1992, 33, 6507.
Acknowledgments
We thank the EPSRC Mass Spectrometry Service, University Col-
lege, Swansea for recording high resolution mass spectrometric data
Eur. J. Org. Chem. 2004, 1973Ϫ1982
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1981

D. W. Knight, D. E. Shaw, E. R. Staples
FULL PAPER
[34]
[43]
[44]
[45]
[46]
[47]
D. W. Knight, D. E. Shaw, G. Fenton, Synlett 1994, 295.
H. Hayakawa, M. Ohmon, K. Takamichi, I. Matsuda, M. Mi-
yashita, Chem. Commun. 1997, 1219.
A. Alami, B. Crousse, G. Linstrumelle, L. Mambu, M. Larch-
eveque, Synlett 1993, 217.
P. J. Garett, B. Samuelsson, J. Chem. Soc., Perkin Trans. 1
[35]
ˆ
M. Larcheveque, S. Henrot, Tetrahedron 1990, 46, 4277.
[36]
[37]
For an alternative method of generation from 1-bromo-1-pro-
pene, see: D. Toussaint, J. Suffert, Org. Synth. 1998, 76, 214.
M. Yamaguchi, I. Hirao, Tetrahedron Lett. 1983, 24, 391. See
also: A. B. Evans, D. W. Knight, Tetrahedron Lett. 2001, 42,
6947.
ˆ
1980, 2866.
P. H. J. Carlsen, T. Katsuki, V. S. Martin, K. B. Sharpless, J.
Org. Chem. 1981, 46, 3936.
The dione 35 was not characterised beyond NMR spectro-
scopic data: H NMR (400 MHz, CDCl3): δ ϭ 2.06 (s, 3 H),
2.27 (s, 3 H), 3.44 (dd, J ϭ 17.8, 4.5 Hz, 1 H, 4-Ha), 3.50 (dd,
J ϭ 17.8, 6.2 Hz, 1 H, 4-Hb), 5.53 (dd, J ϭ 6.2, 4.5 Hz, 1 H,
3-H), 7.37Ϫ7.43 (m, 2 H), 7.53Ϫ7.57 (m, 1 H), 7.87 (d, J ϭ
8.5 Hz, 2 H) ppm. 13C NMR (100.6 MHz, CDCl3): δ ϭ 19.7
(CH3), 25.7 (CH3), 38.8 (4-CH2), 73.1 (3-CH), 127.4 (2 ϫ CH),
127.6 (2 ϫ CH), 132.7 (CH), 134.9 (C), 169.3 (CO), 195.0
(CO), 205.0 (CO) ppm.
[38]
[39]
T. Shiori, K. Ninomiya, S. Yamada, J. Am. Chem. Soc. 1972,
94, 6203.
1
S. Kim, C. H. Oh, J. S. Ko, K. H. Ahn, Y. J. Kim, J. Org.
Chem. 1985, 50, 1927.
[40] [40a]
D. H. R. Barton, Aldrichimica Acta 1990, 23, 1. [40b] W. B.
Motherwell, D. Crich, Free Radical Chain Reactions in Organic
Synthesis, Academic Press, London, 1991.
Cf.: M. Pigini, M. Gianella, F. Gualtieri, Synth. Commun.
1980, 10, 725.
S. P. Bew, J. M. Barks, D. W. Knight, R. J. Middleton, Tetra-
hedron Lett. 2000, 41, 4447.
[41]
[42]
[48]
D. W. Knight, E. R. Staples, Tetrahedron Lett. 2002, 43, 6771.
Received October 13, 2004
1982
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 1973Ϫ1982