A joint experimental and ab initio study on the reactivity of several hydroxy selenides. Stereoselective synthesis of cis-disubstituted tetrahydrofurans via seleniranium ions

Article abstract of DOI:10.1016/S0040-4020(01)00618-4

The reactivity of several hydroxy selenides bearing an ethereal chain with catalytic amounts of perchloric acid in dichloromethane was investigated. Results showed that the position of the oxygen atom with respect to the seleniranium ring was crucial in o

Full text of DOI:10.1016/S0040-4020(01)00618-4

TETRAHEDRON
Pergamon
Tetrahedron 57 12001) 6815±6822
A joint experimental and ab initio study on the reactivity
of several hydroxy selenides. Stereoselective synthesis
of cis-disubstituted tetrahydrofurans via seleniranium ions
Michelangelo Gruttadauria,p Carmela Aprile, Francesca D'Anna, Paolo Lo Meo,
Serena Riela and Renato Noto
Á
Dipartimento di Chimica Organica `E. Paterno', Viale delle Scienze, Parco d'Orleans II, 90128 Palermo, Italy
Received 6 April 2001; revised 9 May 2001; accepted 31 May 2001
AbstractÐThe reactivity of several hydroxy selenides bearing an ethereal chain with catalytic amounts of perchloric acid in dichloro-
methane was investigated. Results showed that the position of the oxygen atom with respect to the seleniranium ring was crucial in order to
get a good yield of the cyclized product. The factors on which yields of the 5-endo cyclization of the seleniranium ions depend were analysed
by ab initio 1HF/3-21Gp) studies. An explanation of the different coordinating ability, towards the positively charged selenium atom, of the
allylic OMe and homoallylic OH-2groups was given. q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
of the appropriate unsaturated alcohol with a source of the
electrophilic seleno species PhSe1, such as PhSeCl.12±18
Both the methods lead to oxygenated heterocyclic rings
via the intermediate seleniranium ion. Recently we studied
the behaviour of the intermediate seleniranium ion having
two hydroxyl groups and an aliphatic R chain 1Fig. 1).17 We
found that the 5-endo-tet19 cyclization occurs only at high
acid concentration or when the primary hydroxy group is
protected.
Structurally complex tetrahydrofuran units are often found
in many natural products such as polyether antibiotics and a
particularly challenging aspect is the stereocontrolled syn-
thesis of the substituted tetrahydrofuran unit, especially
those in which there is a cis relationship between sub-
1
stituents at the 2and 5 positions. During the last decade
several authors have prepared in a synthetically useful
manner substituted tetrahydrofurans via 5-endo-cyclization
of intermediate seleniranium,2±8 thiiranium9 or iodonium
ions.10,11 In the last years our efforts have been devoted to
the stereoselective synthesis of oxygenated heterocyclic
rings from mixtures of hydroxy selenides or by treatment
2. Results and discussion
In order to gain deeper information about the in¯uence of
Figure 1. Different ring closure of the intermediate seleniranium ion.
Keywords: cyclizations; oxygen heterocycles; selenium; theoretical studies.
Corresponding author. Tel.: 139-091-596919; fax: 139-091-596825; e-mail: mgrutt@unipa.it
p
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020101)00618-4

6816
M. Gruttadauria et al. / Tetrahedron 57 .2001) 6815±6822
products from which we were able to isolate the tetrahydro-
furan product again in low yield 130%). In our opinion, there
could be two reasons for this low yield: 1i) a coordination on
the positively charged phenylselenyl moiety by the oxygen
atom of the benzyloxy group that stabilises the seleniranium
ion, increasing in this way the activation energy of the
reaction;21 1ii) some electronically unfavourable factor1s)
introduced to the carbon atom that undergoes the endo
attack. The partial rupture of C±Se bond necessary to
reach the transition state should increase the partial positive
charge at the carbon atom. Then, the 5-endo-tet cyclization
should be favoured if the partial positive charge can be
supported in this position. Indeed, when the carbon atom
was substituted with a phenyl2,3 or a furan-3-yl4,5 group the
reaction took place.22 Also with an aliphatic chain the reac-
tion proceeded with good yield.5±7 The CH2OBn group, in
our case, could represent the electronically unfavourable
factor that does not allow the cyclization to proceed in
good yield.23 Because of the electronegativity of the oxygen
atom the partial positive charge at C-5 cannot be supported.
Moreover we prepared the hydroxy selenides 6c,d, in which
the more distant position of the oxygen atom could play a
role both in the coordination and in the stabilization of a
partial positive charge. The acid treatment of 6c gave the
tetrahydrofuran ring after 30 min with a good yield 160%). It
should be noticed that being the triisopropylsilyl protecting
group a non-chelating group, the coordination of the
positively charged phenylselenyl moiety should be ham-
pered. However, the product obtained showed the
hydroxyl function deprotected, i.e. the labile triisopropyl-
silyl protecting group was removed. This could occur prior
to or after the cyclization step. We found that the triiso-
propylsilyl group was removed after the cyclization took
place, indeed when we treated the hydroxy selenides 6e
with perchloric acid no reaction was observed. The inertness
of 6e may be ascribed to the existence of an effective intra-
molecular hydrogen bond network in the molecule.
Scheme 1. Reagents: 1a) BuLi, BF3´OEt2, THF, 2788C, 69±85%; 1b) H2,
Lindlar, EtOH, rt, 89±96%; 1c) t-BuOOH, VO1acac)2, CH2Cl2, rt, 72±84%;
1d) 1PhSe)2, NaBH4, EtOH, rt, 73±80%; 1e) HClO4 1cat.), CH2Cl2, rt,
30±72%.
the R group in the 5-endo-tet cyclization, we have analysed
the reactivity of hydroxy selenides 6 bearing, as R
group, an oxygenated aliphatic chain 1RvCH2OPG;
RvCH2CH2OPG). These compounds were prepared as
outlined in Scheme 1. Epoxidation of homoallylic alcohols
4 using tert-butyl-hydroperoxide and VO1acac)2 gave the
syn-epoxy-alcohols with good stereoselectivity 1.93:7),
the con®guration of the major epoxide being assigned by
analogy with the literature.20 The hydroxy selenides were
usually obtained as a ca. 80/20 mixture and we have not
determined which was which. Previously we studied the
cyclization of 6a, obtaining a complex mixture of products
from which we could isolate the tetrahydrofuran 7f with
low yield 127%).14
The next step was the cyclization reaction of hydroxy
selenides 6d. Also in this case we obtained a good yield
172%). The benzyloxy, in contrast to the triisopropyloxy,
is a chelating group able to coordinate the phenylselenyl
moiety. Then, the good yields for 6c,d should be ascribed
only to the more distant position of the oxygen atom.
The reaction was complicated by the fact that in the
operating conditions the labile triisopropylsilyl protecting
group was removed then allowing also the cyclization in
the exo mode and, possibly, the rearrangement.16 The result
was a complex mixture of products. In order to avoid this,
we protected both the primary hydroxyl groups as benzyl
ethers. The hydroxy selenides 6b were then treated with
perchloric acid, but also in this case the reaction was
sluggish and after 24 h we obtained a complex mixture of
In order to estimate the importance of point 1i) and 1ii) in the
cyclization of seleniranium ions we performed ab initio
calculations at the HF/3-21Gp level of theory on the
model species I±VI and IV´H1±VI´H1 1Fig. 2).
In a previous paper we reported that HF/3-21Gp calculations
allow a detailed analysis of the behaviour of I.17 Now we
extended our attention to the species II and III, in order to
model the seleniranium ions deriving from 6a,b and 6c,d,
respectively. In order to save computational time we choose
to replace the Bn or TIPS groups in R 1see Scheme 1) with a
methyl group, while the R0 groups were replaced by a
hydrogen atom. A preliminary search had shown that for I
at least six different conformations could be taken in con-
sideration.17 We restricted our analysis for II and III only to
three possible conformers, namely at, f2and c2, having
respectively the chain carrying the OH groups completely
Figure 2. Models submitted to calculations.

M. Gruttadauria et al. / Tetrahedron 57 .2001) 6815±6822
6817
Figure 3. Conformers examined for model species I±III 1input structures for geometry optimization).
unfolded, or folded in such a way to present the OH-2group
towards the selenium atom or towards the backside of the
C±Se bonds of the seleniranium ring 1see Fig. 3). At the same
time in all cases the conformation of the ±CH2OCH3 or
±CH2CH2OCH3 moieties was accounted to present the
±OCH3 group directed towards the positively charged Se
atom.
The extent of the interaction between the Se atom and the
OMe group can be easily evaluated by means of the NBO
population analysis,24 a tool which has been proven
effective in describing pseudo high valent Se species.25±27
The oxygen atom is able to delocalise part of its electron
density in the sp
antibonding orbital, with an energy
gain given by the NBO deletion energy 1Table 1).
Se±Ph
Table 1. Calculated parameters for the 5-endo-type closure of seleniranium models I,II and III
I
II
III
Ef 1hartree)a
Ef 1hartree)c
Ef 1hartree)
22999.08971b
22999.08268b
22999.08800b,d
±
23112.34535
23112.33885
23112.33312e
6.90
23151.16248
23151.15988
23151.15461e
1.82
a
)
a
Edel 1kcal mol21
Ê
d Se´´´OMe 1A)
±
2.773
3.069
Esp1C±Se) 1hartree)c
20.00712
20.94392
0.93680
10.01887
20.93420
0.95307
10.00476
20.92942
0.93418
ELP1O) 1hartree)c
DEsp 1C±Se)±LP1O) 1hartree)c
TS &I)
22999.07269b
10.68
TS &II)
23112.31762
17.40
TS &III)
23151.14683
9.82
Ef 1hartree)
Ring closure DH± 1kcal mol21
qC 1e)f
)
10.207
49
56
10.084
46
65
10.210
39
58
% O´´´C bond formationg
% C´´´Se bond formationg
IV´H1
22999.10882b
22.67b
V´H1
23112.36274
28.31
VI´H1
23151.20422
36.01
Ef 1hartree)
Ring opening DH± 1kcal mol21
)
IV
22998.73751b
V
23111.98542
VI
23150.80064
Ef 1hartree)
a
For the at conformer.
Values from Ref. 17.
For the c2conformer.
For the f2conformer.
For conformer at0.
b
c
d
e
f
Charge on the carbon atom as calculated by the NBO analysis.28
See Ref. 29.
g

6818
M. Gruttadauria et al. / Tetrahedron 57 .2001) 6815±6822
is lost on reaching the transition state for the ring closure
process 1see Fig. 4).
Nonetheless NBO analysis for the c2conformers, which are
the actual reactive species, shows that the energy gap
between the sp
antibonding orbital and the OH-2lone
Se±C
pair is very similar in I and III, but strongly increases on
passing to II.31 The two combined effects contribute to make
the activation energy for II much higher than that for I and
III. Furthermore, calculations for the TS leading to V
predict the development of only a small positive charge
on the attacked C atom, while a higher positive charge is
predicted in the TSs leading to IV and VI. At the same time
bond formation percentage shows that the degree of dis-
ruption of the C±Se bond in the TS for V is anomalously
low with respect to the contemporary formation of the C±O
bond, differently from the TS leading to IV and VI that
appear as pure SN2processes. As a ®nal remark, data
collected in table con®rm that the 5-endo mode cyclization
is an irreversible process. Indeed, activation barriers for ring
opening are much higher than for closure, ranging largely
Figure 4. Optimised structures 1HF/3-21Gp) for transition states TS&I),
TS&II) and TS&III).
over the value of 22 kcal mol21 32
.
Surprisingly, for both models II and III we found that the f2
form is not a stable minimum, but collapses to the at form,
which is in turn the absolute energy minimum.
3. Conclusion
The collapse found by us of conformers IIf2and IIIf2to IIat
and IIIat, respectively, while by contrast If2and Iat are
distinct minima,17 may be attributed to the presence of
Se±OMe interaction. Thus OMe group seems better coordi-
The low yield for the cyclization of 6a,b could be ascribed
both to the coordination of the positively charged phenyl-
selenyl moiety by the benzyloxy group in the seleniranium
ions and to the presence of the oxygen atom 1of the benzyl-
oxy group) that hampers the development of a partial posi-
tive charge at C-5 and then the rupture of the C´´´Se bond.
Both these factors make the activation energy of the ring
closure process high. In the cyclization of 6c,d both the
coordination and some electronic unfavourable factors
introduced to the carbon atom that undergo the endo attack
should be negligible. These results also explain the observed
regioselectivity in the hydroxyselenylation of allylic alco-
hols.33 The allylic OMe group was found better coordinating
than the homoallylic OMe group to the spSe±Ph antibonding
orbital. The different coordination ability between the
homoallylic OH and the allylic OMe groups is probably
ascribed to the strain introduced by the chain folding.
Further evidences of Se´´´O interaction have been recently
reported by us.18
nating than the OH-2to the sp
antibonding orbital.
Se±Ph
Furthermore, the higher Edel1II) value than Edel1III) 1see
Table 1) shows that the formation of a pseudo-4-membered
ring in II leads to a more favourable Se±O stabilizing inter-
action than the formation of a pseudo-5-membered ring in
III, as evidenced also by the shorter Se±O distance.
It may be argued that the lack in stability for IIf2and IIIf2is
due to steric strain factors. As a matter of fact, If2, unless
showing the deletion energy equal to 10.65 kcal mol21 for
the Se±OH-2interaction, is less stable than Iat for about
1 kcal mol21 17
So we may evaluate the strain induced by
.
folding of the chain to be about 11.7 kcal mol21. In IIf2the
additional Se±OMe interaction reduces the extent of the
Se±OH-2one and makes it unable to compensate the fold-
ing strain. To get more detailed insights we studied also a
further conformation, namely at0, for II and III, similar to
at, but bearing the ±OMe group unfolded far away from the
Se atom. Model IIat0 is found less stable than IIat by about
7.7 kcal mol21. Thus the Se±OMe interaction largely over-
whelms the folding strain both in terms of orbitalic and
coulombic interactions. Similar competing coordinations
have already been recognized.30
4. Experimental
Anhydrous solvents were distilled as follows: Tetrahydro-
furan and diethyl ether were distilled under nitrogen from
sodium benzophenone immediately prior to use. Dichloro-
methane was distilled under nitrogen from calcium hydride
and used immediately. 1H and 13C NMR spectra were
recorded on a Bruker AC-E series 250 MHz spectrometer.
IR spectra were recorded on a Perkin±Elmer infrared
spectrophotometer 1model 1310) using KBr cells. Flash
chromatography was carried out using Macherey±Nagel
silica gel 10.04±0.063 mm). Light petroleum refers to the
fraction boiling in the range 40±608C. Ab initio calculations
were performed with the GAUSSIAN98 program
distribuited by Gaussian Inc.34 Full geometry optimisation
was performed for each model species examined. Minimum
This preliminary conformational search was needed in order
to rationalise the behaviour of molecules 6b,d, comparing
the 5-endo cyclization of models I, II and III. Our calcula-
tions predict similar activation barriers for the formation of
IV and VI from I and III respectively, but a higher barrier
for the formation of V and thus a lack in reactivity for II, in
good agreement with experimental results. NBO population
analysis was very useful to rationalise these ®ndings.
Models predict that the Se±OMe interaction for II and III

M. Gruttadauria et al. / Tetrahedron 57 .2001) 6815±6822
6819
and transition states structures were con®rmed by inspection
of the hessian matrix eigenvalues. Compounds 3±6a,7e
were described elsewhere.14
4.2.1. &^)&Z,SR)-1,6-&Dibenzyloxy)-hex-4-en-2-ol &4b). IR
1
1liquid ®lm) nmax 3448, 1445, 1205, 1072, 736 cm21. H
NMR 1250 MHz, CDCl3): dˆ2.28±2.33 1m, 2H), 2.74 1s,
1H), 3.40 1dd, 1H, Jˆ9.6, 5.0 Hz), 3.48 1dd, 1H, Jˆ9.6,
3.6 Hz), 3.81±3.90 1m, 1H), 4.08±4.17 1m, 2H), 4.53±
4.59 1m, 4H), 5.67±5.82 1m, 2H), 7.27±7.39 1m, 10H).
13C NMR 1CDCl3): dˆ31.7, 65.4, 69.7, 72.2, 73.3, 73.7,
127.5, 127.6, 127.7, 127.8, 128.2, 128.3, 128.6, 129.0,
137.8, 138.0. Anal. Calcd for C20H24O3: C, 76.89; H, 7.74.
Found: C, 76.90; H, 7.74.
4.1. General procedure for the synthesis of alkynes 3b±d
Butyllithium 11.60 M in hexane; 5.7 mL, 9.12mmol) was
added dropwise to a solution of the alkyne 1RvCH2OBn,
1.34 g, 9.15 mmol) in anhydrous tetrahydrofuran 120 mL) at
2788C. After 20 min, BF3±Et2O 10.75 mL, 6.1 mmol) was
added followed, after 5 min, by a solution of the epoxide
1R0ˆBn., 0.93 mL, 6.1 mmol) in anhydrous tetrahydrofuran
13.3 mL). After 2h at 2788C, saturated aqueous NaHCO3
110 mL) was added, the mixture allowed to warm to room
temperature and added to water. The mixture was extracted
with Et2O and the combined organic extracts were washed
with brine, dried 1Na2SO4) and the solvent evaporated in
vacuo. Puri®cation of the crude product by ¯ash chromato-
graphy 1light petroleum±Et2O 7/1) gave 3b 11.39 g, 73%) as
an oil.
4.2.2. &^)&Z,SR)-1-&Benzyloxy)-7-&triisopropylsilyloxy)-
hept-4-en-2-ol &4c). Oil, yield 189%) from light petro-
leum±Et2O 7/1; IR 1liquid ®lm) nmax 3440, 1450, 1380,
1
1245, 1100, 880 cm21. H NMR 1250 MHz, CDCl3): dˆ
1.08 1m, 21H), 2.28±2.38 1m, 4H), 2.56 1d, 1H, Jˆ
3.3 Hz), 3.40 1dd, 1H, Jˆ9.3, 7.2Hz), 3.52 1dd, 1H, Jˆ
9.3, 3.6 Hz), 3.71 1t, 2H, Jˆ6.8 Hz), 3.84±3.91 1m, 1H),
4.57 1s, 2H), 5.49±5.65 1m, 2H), 7.27±7.38 1m, 5H). 13C
NMR 1CDCl3): dˆ11.9, 17.9, 31.2, 31.4, 62.9, 70.1, 73.3,
73.9, 126.3, 127.6, 128.3, 129.0, 138.0. Anal. Calcd for
C23H40O3Si: C, 70.36; H, 10.27. Found: C, 70.50; H, 10.35.
4.1.1. &^)&SR)-1,6-&Dibenzyloxy)-hex-4-yn-2-ol &3b). IR
1liquid ®lm) nmax 3440, 1454, 1355, 1207, 1070,
4.2.3. &^)&Z,SR)-1,7-&Dibenzyloxy)-hept-4-en-2-ol &4d).
Oil, yield 192%) from light petroleum±Et2O 7/1; IR 1liquid
1
738 cm21. H NMR 1250 MHz, CDCl3): dˆ2.52±2.56 1m,
1
®lm) nmax 3440, 1450, 1365, 1100, 740 cm21. H NMR
2H), 2.82 1s, 1H), 3.53 1dd, 1H, Jˆ9.7, 4.0 Hz), 3.64 1dd,
1H, Jˆ9.7, 6.5 Hz), 3.95±4.04 1m, 1H), 4.18±4.20 1m, 2H),
4.60 1s, 4H), 7.27±7.39 1m, 10H). 13C NMR 1CDCl3): dˆ
23.9, 57.6, 66.9, 71.4, 72.9, 73.4, 78.2, 82.6, 127.7, 127.8,
128.0, 128.3, 128.4, 137.5, 137.8. Anal. Calcd for C20H22O3:
C, 77.39; H, 7.14. Found: C, 77.50; H, 7.14.
1250 MHz, CDCl3): dˆ2.31±2.48 1m, 4H), 2.81 1s, 1H),
3.39±3.53 1m, 4H), 3.84±3.921m, 1H), 4.54 1s, 2H), 4.57
1s, 2H), 5.56±5.61 1m, 2H), 7.28±7.37 1m, 10H). 13C NMR
1CDCl3): dˆ27.9, 31.4, 69.3, 70.0, 72.8, 73.2, 73.9, 126.8,
127.4, 127.5, 127.6, 128.2, 128.3, 128.6, 138.0, 138.2. Anal.
Calcd for C21H26O3: C, 77.27; H, 8.03. Found: C, 77.30; H,
8.10.
4.1.2.
&^)&SR)-1-&Benzyloxy)-7-&triisopropylsilyloxy)-
hept-4-yn-2-ol &3c). Oil, yield 185%) from light petro-
leum±Et2O 9/1; IR 1liquid ®lm) nmax 3420, 1455, 1380,
4.3. General procedure for the synthesis of epoxides 5b±d
1
1100, 910, 880 cm21. H NMR 1250 MHz, CDCl3): dˆ
1.08 1m, 21H), 2.38±2.45 1m, 4H), 2.67 1d, 1H, Jˆ
5.2Hz), 3.50 1dd, 1H, Jˆ9.0, 6.3 Hz), 3.61 1dd, 1H, Jˆ
9.0, 3.6 Hz), 3.77 1m, 2H), 3.89±3.96 1m, 1H), 4.57 1s,
2H), 7.27±7.36 1m, 5H). 13C NMR 1CDCl3): dˆ11.1,
17.9, 23.1, 23.8, 62.3, 69.0, 72.9, 73.2, 76.7, 79.6, 127.6,
128.3, 137.8. Anal. Calcd for C23H38O3Si: C, 70.72; H, 9.80.
Found: C, 70.90; H, 9.85.
To a solution of 4b 11.12g, 3.6 mmol) in anhydrous
dichloromethane 134 mL) at 08C, VO1acac)2 117 mg,
0.064 mmol) was added, then tert-butyl hydroperoxide
15.5 M in decane; 0.98 mL, 5.40 mmol) was added drop-
wise. The reaction mixture was stirred at 08C for 10 min
then allowed to warm to room temperature. After 20 h the
reaction was quenched with saturated aqueous sodium thio-
sulfate. The mixture was washed with water, brine, dried
1Na2SO4) and the solvent evaporated in vacuo. Puri®cation
of the crude product by ¯ash chromatography 1light petro-
leum±ethyl acetate 3/1) gave 5b 1855 mg, 72%) as an oil.
4.1.3. &^)&SR)-1,7-&Dibenzyloxy)-hept-4-yn-2-ol &3d).
Oil, yield 169%) from light petroleum±Et2O 7/1; IR 1liquid
1
®lm) nmax 3440, 1450, 1360, 1100, 740 cm21. H NMR
1250 MHz, CDCl3): dˆ2.39±2.48 1m, 4H), 2.83 1d, 1H,
Jˆ3.9 Hz, OH), 3.43±3.56 1m, 4H), 3.85±3.97 1m, 1H),
4.50±4.53 1s, 4H), 7.26±7.34 1m, 10H). 13C NMR
1CDCl3): dˆ20.0, 23.7, 68.4, 68.9, 72.7, 72.8, 73.1, 76.8,
79.1, 127.4, 127.5, 128.2, 137.8, 137.9. Anal. Calcd for
C21H24O3: C, 77.75; H, 7.46. Found: C, 77.60; H, 7.50.
4.3.1. &^)&2SR,4RS, 5SR)-1,6-&Dibenzyloxy)-4,5-epoxy-
hexan-2-ol &5b). IR 1liquid ®lm) nmax 3446, 1496, 1454,
1
1365, 1205, 1100 cm21. H NMR 1250 MHz, CDCl3): dˆ
1.64±1.86 1m, 2H), 2.80 1s, 1H), 3.16±3.25 1m, 2H), 3.44±
3.61 1m, 3H), 3.68 1dd, 1H, Jˆ11, 4.4 Hz), 4.02±4.13 1m,
1H), 4.53 and 4.63 1d, each 1H, Jˆ11.8 Hz), 4.56 1s, 2H),
7.27±7.40 1m, 10H). 13C NMR 1CDCl3): dˆ31.6, 53.3,
54.4, 68.0, 68.9, 73.3, 73.4, 73.7, 127.7, 127.8, 128.4,
137.7, 137.8. Anal. Calcd for C20H24O4: C, 73.15; H, 7.37.
Found: C, 73.25; H, 7.44.
4.2. General procedure for the synthesis of alkenes 4b±d
A suspension of Lindlar catalyst 1126 mg) in a solution of
the alkyne 3b 11.26 g, 4.06 mmol) in ethanol 120 mL) was
stirred vigorously under hydrogen for 210 min. The mixture
was ®ltered through Celite and the ®ltrate evaporated in
vacuo. Puri®cation of the crude product by ¯ash chroma-
tography 1light petroleum±Et2O 7/1) gave 4b 11.22 g, 96%)
as an oil.
4.3.2. &^)&2SR,4RS, 5SR)-1-&Benzyloxy)-7-&triisopropyl-
silyloxy)-4,5-epoxy-heptan-2-ol &5c). Oil, yield 175%)
from light petroleum±ethyl acetate 7/1; IR 1liquid ®lm)
1
nmax 3420, 1460, 1100, 880 cm21. H NMR 1250 MHz,

6820
M. Gruttadauria et al. / Tetrahedron 57 .2001) 6815±6822
CDCl3): dˆ1.08 1m, 21H), 1.58±1.93 1m, 4H), 2.77 1d, 1H,
Jˆ3.2 Hz), 3.10±3.20 1m, 2H), 3.46±3.56 1m, 2H), 3.85±3.90
1m, 2H), 4.07±4.11 1m, 1H), 4.57 1s, 2H), 7.27±7.34 1m, 5H).
13C NMR 1CDCl3): dˆ11.8, 17.9, 31.4, 31.6, 54.0, 54.2, 60.6,
69.1, 73.3, 73.8, 127.6, 127.7, 128.4, 137.8. Anal. Calcd for
C23H40O4Si: C, 67.60; H, 9.87. Found: C, 67.70; H, 9.85.
3.621m, 1H), 3.72±3.78 1m, H2 ), 3.81±3.93 1m, 1H),
4.00±4.07 1m, 1H), 4.50 1s, 2H, minor regioisomer), 4.53
1s, 2H, major regioisomer), 4.82 1d, 1H, Jˆ4.7 Hz, minor
regioisomer), 4.86 1d, 1H, Jˆ5.4 Hz, major regioisomer),
4.94 1d, 1H, Jˆ4.9 Hz, major regioisomer), 5.09 1d, 1H,
Jˆ4.1 Hz, minor regioisomer), 7.26±7.41 1m, 8H), 7.55±
7.61 1m, 2H). 13C NMR 1DMSO-d6): dˆmajor regioisomer
11.6, 18.0, 36.3, 37.3, 50.6, 60.5, 67.5, 69.8, 72.4, 75.1,
126.6, 127.5, 127.6, 128.3, 129.1, 131.0, 132.7, 138.7.
Anal. Calcd for C29H46O4SeSi: C, 61.57; H, 8.20. Found:
C, 61.70; H, 8.30.
4.3.3. &^)&2SR,4RS,5SR)-1,7-&Dibenzyloxy)-4,5-epoxy-
heptan-2-ol &5d). Oil, yield 184%) from light petroleum±
ethyl acetate 3/1; IR 1liquid ®lm) nmax 3440, 1450, 1365,
1
1100, 740 cm21. H NMR 1250 MHz, CDCl3): dˆ1.65±
1.90 1m, 4H), 2.99 1s, 1H), 3.09±3.20 1m, 2H), 3.48±3.52
1m, 2H), 3.62±3.67 1m, 2H), 4.10±4.20 1m, 1H), 4.54 1s,
2H), 4.56 1s, 2H), 7.29±7.37 1m, 10H). 13C NMR 1CDCl3):
dˆ28.4, 31.4, 53.9, 54.0, 67.2, 68.7, 72.9, 73.1, 73.7, 127.4,
127.5, 127.6, 128.2, 128.3, 137.7, 137.9. Anal. Calcd for
C21H26O4: C, 73.66; H, 7.65. Found: C, 73.60; H, 7.60.
4.4.3.
&^)&2SR,4SR,5SR)-1,7-&Dibenzyloxy)-4-phenyl-
selenyl-heptan-2,5-diol and &^)&2SR,4RS,5RS)-1,7-&di-
benzyloxy)-5-phenylselenyl-heptan-2,4-diol &6d). Pale-
yellow oil, yield 173%) from light petroleum±ethyl acetate
3/1; regioisomers ratio ca. 80/20 1determined by 1H NMR);
IR 1liquid ®lm, mixture of isomers) nmax 3410, 1575, 1490,
1
1450, 1360, 1090, 740 cm21. H NMR 1250 MHz, DMSO-
4.4. General procedure for the synthesis of hydroxy
selenides 6b±d
d6): dˆ1.72±2.03 1m, 4H), 3.35±3.70 1m, 5H), 3.85±3.93
1m, 1H), 4.02±4.10 1m, 1H), 4.46 1s, 2H), 4.54 1s, 2H), 4.87
1d, 1H, Jˆ5.5 Hz, OH, major regioisomer), 5.06 1d, 1H,
Jˆ4.9 Hz, OH, major regioisomer), 5.121d, 1H, Jˆ
4.5 Hz, OH, minor regioisomer), 7.27±7.41 1m, 13H),
7.58±7.65 1m, 2H). 13C NMR 1DMSO-d6): dˆmajor regio-
isomer: 34.4, 36.8, 50.8, 67.3, 67.6, 70.2, 72.0, 72.4, 75.1,
126.7, 127.4, 127.5, 127.6, 128.4, 129.2, 131.1, 132.8,
138.7, 138.8. Anal. Calcd for C27H32O4Se: C, 64.92; H,
6.46. Found: C, 65.05; H, 6.50.
Diphenyl diselenide 1189 mg, 0.61 mmol) was dissolved in
absolute ethanol 11.8 mL), sodium borohydride 167 mg,
1.77 mmol) was added in batches and the mixture was
stirred until the bright yellow solution turned colourless.
Compound 5b 1361 mg, 1.10 mmol) was dissolved in
absolute ethanol 11.4 mL) and added via cannula. The reac-
tion mixture was stirred for 16 h then concentrated under
reduced pressure. Dichloromethane was added and the
mixture was washed with water. The organic phase was
washed with brine, dried 1Na2SO4) and the solvent evapo-
rated in vacuo. Puri®cation of the crude product by ¯ash
chromatography 1light petroleum±ethyl acetate 4/1) gave
the title compounds 6b 1427 mg, 80%) as pale-yellow oil.
4.4.4. &^)&2SR,4SR,5SR)-1-&Benzyloxy)-4-phenylselenyl-
heptan-2,5,7-triol and &^)&2SR,4RS,5RS)-1-&benzyloxy)-
5-phenylselenyl-heptan-2,4,7-triol &6e). A solution of
tetrabutylammonium ¯uoride 11.03 g, 3.28 mmol) in
anhydrous tetrahydrofuran 15 mL) was added dropwise to
a solution of the silyl ethers 6c 1927 mg, 1.64 mmol) in
tetrahydrofuran 15 mL) at 08C, and the mixture allowed to
warm to room temperature and stirred for 18 h. The solution
was concentrated under reduced pressure, then dissolved in
ethyl acetate and extracted with water. The organic phase
was washed with brine, dried 1Na2SO4) and the solvent
evaporated in vacuo. Puri®cation of the crude product by
¯ash chromatography 1light petroleum±ethyl acetate 1/1)
gave the title compounds 6e 1648 mg, 95%) as oil. Regio-
isomers ratio ca. 80/20 1determined by 1H NMR); IR 1liquid
®lm, mixture of isomers) nmax 3370, 1475, 1490, 1475,
4.4.1.
&^)&2SR,3SR,4SR)-1,6-&Dibenzyloxy)-3-phenyl-
selenyl-hexan-2,5-diol and &^)&2RS,3RS,4SR)-1,6-&Di-
benzyloxy)-2-phenylselenyl-hexan-3,5-diol &6b). Regio-
isomers ratio ca. 85/15 1determined by 1H NMR); IR 1liquid
®lm, mixture of isomers) nmax 3418, 1577, 1454,
1075 cm21
.
1H NMR 1250 MHz, DMSO-d6): dˆ1.75±
1.87 1m, 1H), 1.92±2.06 1m, 1H), 3.32±3.54 1m, 3H),
3.64±3.78 1m, 2H), 3.85±3.94 1m, 1H), 3.98±4.10 1m,
1H), 4.35 and 4.43 1d, each 1H, Jˆ12Hz), 4.51 1s, 2H),
4.77 1d, 1H, Jˆ5.6 Hz, OH, major regioisomer), 4.10 1d,
1H, Jˆ4.0 Hz, OH, minor regioisomer), 5.25 1d, 1H, Jˆ
4.6 Hz, OH, major regioisomer), 5.40 1d, 1H, Jˆ4.1 Hz,
OH, minor regioisomer), 7.23±7.42 1m, 13H), 7.58±7.63
1m, 2H). 13C NMR 1CDCl3): dˆmajor regioisomer 36.5,
47.8, 68.9, 72.5, 73.0, 73.2, 73.3, 74.4, 127.4, 127.6,
127.7, 128.3, 128.4, 129.1, 129.4, 134.1, 137.8. Anal.
Calcd for C26H30O4Se: C, 64.32; H, 6.23. Found: C,
64.40; H, 6.32.
1
1450, 1180, 740 cm21. H NMR 1250 MHz, DMSO-d6):
dˆ1.70±1.95 1m, 4H), 3.35±3.65 1m, 4H), 3.69±3.97 1m,
1H), 4.01±4.10 1m, 1H), 4.41±4.47 1m, 1H, major regio-
isomer), 4.50 1s, 2H, minor regioisomer), 4.52 1s, 2H, major
regioisomer), 4.61±4.70 1m, 1H, minor regioisomer), 4.96
1d, 1H, Jˆ5.2Hz, OH, major regioisomer), 5.10 1d, 1H,
Jˆ4.6 Hz, OH, major regioisomer), 7.26±7.41 1m, 8H),
7.56±7.63 1m, 2H). 13C NMR 1DMSO-d6): dˆmajor regio-
isomer 37.2, 37.7, 51.1, 58.5, 67.6, 70.6, 72.4, 75.1, 126.6,
127.5, 127.6, 128.4, 129.2, 131.2, 132.6, 138.72. Anal.
Calcd for C20H26O4Se: C, 58.68; H, 6.40. Found: C,
58.75; H, 6.32.
4.4.2. &^)&2SR,4SR,5SR)-1-&Benzyloxy)-7-&triisopropyl-
silyloxy)-4-phenylselenyl-heptan-2,5-diol and &^)&2SR,4RS,
5RS)-1-&Benzyloxy)-7-&triisopropylsilyloxy)-5-phenyl-
selenyl-heptan-2,4-diol &6c). Pale-yellow oil, yield 179%)
from light petroleum±ethyl acetate 5/1; regioisomers ratio
ca. 80/20 1determined by 1H NMR); IR 1liquid ®lm, mixture
of isomers) nmax 3400, 1578, 1470, 1460, 1100, 880,
4.5. General procedure for the synthesis of tetrahydro-
furans 7b,d,e
1
738 cm21. H NMR 1250 MHz, DMSO-d6): dˆ1.02±1.09
1m, 21H), 1.70±1.90 1m, 4H), 3.34±3.49 1m, 2H), 3.54±
To a solution of hydroxy selenides 6c 1566 mg, 1.00 mmol)

M. Gruttadauria et al. / Tetrahedron 57 .2001) 6815±6822
6821
in anhydrous dichloromethane 150 mL) at room temperature
References
were added three drops 115 mL) of HClO4 170%). The reac-
tion mixture was vigorously stirred for 30 min then
quenched with saturated aqueous NaHCO3 and extracted
with water. The organic phase was washed with brine,
dried 1Na2SO4) and the solvent evaporated in vacuo.
Puri®cation of the crude product by ¯ash chromatography
1light petroleum±ethyl acetate 5/1) gave 7e 1228 mg 60%)
as an oil.
1. Harmange, J. C.; Figadere, B. Tetrahedron: Asymmetry 1993,
4, 1711±1754.
2. Andrey, O.; Ducry, L.; Landais, Y.; Planchenault, D.; Weber,
V. Tetrahedron 1997, 53, 4339±4352.
3. Kang, S. H.; Hwang, T. S.; Kim, W. J.; Lim, J. K. Tetrahedron
Lett. 1991, 32, 4015±4018.
4. Kang, S. H.; Lee, S. B. Tetrahedron Lett. 1993, 34, 1955±
1958.
4.5.1. &^)&2SR,3SR,5SR)-2-Hydroxyethyl-3-phenylselenyl-
5-benzyloxymethyl-tetrahydrofuran &7e). IR 1liquid ®lm)
5. Lipshutz, B. H.; Gross, T. J. Org. Chem. 1995, 60, 3572±
3573.
nmax 3340, 1575, 1480, 1455, 1100 cm21 1H NMR
.
6. Mihelich, E. D.; Hite, G. A. J. Am. Chem. Soc. 1992, 114,
7318±7319.
1250 MHz, CDCl3): dˆ1.82±2.02 1m, 3H), 2.47 1ddd, 1H,
Jˆ13.4, 7.4 and 7.4 Hz), 3.00 1s, 1H, OH), 3.50±3.621m,
2H), 3.77 1t, 3H, Jˆ5.6 Hz), 3.50±3.90 1m, 1H), 4.06±4.15
1m, 1H), 4.18±4.26 1m, 1H), 4.53 and 4.59 1d, each 1H,
Jˆ12.1 Hz), 7.22±7.36 1m, 8H), 7.49±7.53 1m, 2H). 13C
NMR 1CDCl3): dˆ35.5, 36.2, 44.8, 60.6, 72.3, 73.1, 77.1,
80.5, 127.1, 127.4, 127.5, 128.1, 128.9, 129.6, 133.3, 137.7.
Anal. Calcd for C20H24O3Se: C, 61.38; H, 6.18. Found: C,
61.40; H, 6.22.
7. Mihelich, E. D. J. Am. Chem. Soc. 1990, 112, 8995±8997.
8. Bravo, F.; Castillon, S. Eur. J. Org. Chem. 2001, 507±516.
9. House, D.; Kerr, F.; Warren, S. J. Chem. Soc., Chem.
Commun. 2000, 1779±1780.
10. Kang, S. H.; Lee, S. B. Tetrahedron Lett. 1993, 34, 7579±
7582.
11. Barks, J. M.; Weingarten, G. G.; Knight, D. W. J. Chem. Soc.,
Perkin Trans. 1 2000, 3469±3476.
12. Arista, L.; Gruttadauria, M.; Thomas, E. J. Synlett 1997, 62 7±
628.
4.5.2.
&^)&2SR,3SR,5SR)-2-Benzyloxyethyl-3-phenyl-
selenyl-5-benzyloxymethyl-tetrahydrofuran &7d). Oil,
yield 172%) from light petroleum±ethyl acetate 7/1; IR
1liquid ®lm) nmax 1575, 1490, 1470, 1450, 1360,
13. Arista, L.; Gruttadauria, M.; Noto, R. Heterocycles 1998, 48,
1325±1330.
1100 cm21
.
1H NMR 1250 MHz, CDCl3): dˆ1.97±2.11
14. Gruttadauria, M.; Lo Meo, P.; Noto, R. Tetrahedron 1999, 55,
4769±4782.
1m, 3H), 2.57 1ddd, 1H, Jˆ13.5, 7.6, 7.6 Hz), 3.54±3.70
1m, 4H), 3.87 1ddd, 1H, Jˆ7.5, 5.6, 5.6 Hz), 4.14±4.22
1m, 2H), 4.51 and 4.56 1d, each 1H, Jˆ11.9 Hz), 4.60 and
4.66 1d, each 1H, Jˆ12.2 Hz), 7.28±7.41 1m, 13H), 7.53±
7.57 1m, 2H). 13C NMR 1CDCl3): dˆ33.5, 36.8, 45.4, 67.5,
72.8, 73.2, 77.0, 78.8, 127.1, 127.4, 127.5, 127.6, 127.7,
128.2, 129.0, 133.5, 138.1, 138.4. Anal. Calcd for
C27H30O3Se: C, 67.35; H, 6.28. Found: C, 67.35; H, 6.30.
15. Gruttadauria, M.; Lo Meo, P.; Noto, R. Tetrahedron 1999, 55,
14097±14110.
16. Gruttadauria, M.; Noto, R. Tetrahedron Lett. 1999, 40, 8477±
8481.
17. Gruttadauria, M.; Lo Meo, P.; Noto, R. Tetrahedron 2001, 57,
1819±1826.
18. Gruttadauria, M.; Aprile, C.; Riela, S.; Noto, R. Tetrahedron
Lett. 2001, 42, 2213±2215.
19. Following Warren's nomenclature this should be a 6-endo/
5-exo cyclization: McIntyre, S.; Warren, S. Tetrahedron
Lett. 1990, 31, 3457±3460.
4.5.3. &^)&2SR,3SR,5SR)-2,5-Dibenzyloxymethyl-3-phenyl-
selenyl-tetrahydrofuran &7b). To the solution of hydroxy
selenides 6b were added 20 mL of HClO4 170%). The reac-
tion mixture was vigorously stirred for 20 h. Oil, yield
130%) from light petroleum±ethyl acetate 8/1; IR 1liquid
20. Mihelich, E. D.; Daniels, K.; Eickhoff, D. J. J. Am. Chem. Soc.
1981, 103, 7690±7692.
1
®lm) nmax 1577, 1477, 1454, 1205, 1074 cm21. H NMR
21. Toshimitsu, A.; Terao, K.; Uemura, S. J. Org. Chem. 1987, 52,
2018±2026.
1250 MHz, CDCl3): dˆ2.06 1ddd, 1H, Jˆ8.5, 8.0, 8.0 Hz),
2.48 1ddd, 1H, Jˆ12.8, 7.1, 7.1 Hz), 3.55 1dd, 1H, Jˆ10.1,
4.6 Hz), 3.65 1dd, 1H, 10.1 and 5.9 Hz), 3.73±3.77 1m, 2H),
3.82±3.91 1m, 1H), 4.13±4.19 1m, 1H), 4.30±4.37 1m, 1H),
4.52and 4.58 1d, each 1H, Jˆ12Hz), 4.58 and 4.67 1d, each
1H, Jˆ9.8 Hz), 7.27±7.41 1m, 13H), 7.54±7.58 1m, 2H).
13C NMR 1CDCl3): dˆ37.2, 43.0, 72.1, 72.5, 73.2, 77.6,
80.3, 126.8, 127.4, 127.5, 127.6, 127.7, 128.2, 128.3,
128.4, 129.0, 130.9, 133.4, 138.0, 138.1. Anal. Calcd for
C26H28O3Se: C, 66.80; H, 6.04. Found: C, 66.76; H, 6.02.
22. However in these cases the reaction should be considered as a
5-exo cyclization.
23. For a similar case see: Haraguchi, K.; Hosoe, M.; Tanaka, H.;
Tsuruoka, S.; Kanmuri, K.; Miyasaka, T. Tetrahedron Lett
1998, 39, 5517±5520.
24. Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys.
1985, 83, 735±746.
25. Iwaoka, M.; Tomoda, S. J. Org. Chem. 1995, 60, 5299±5302.
26. Iwaoka, M.; Tomoda, S. J. Am. Chem. Soc. 1996, 118, 8077±
8084.
27. Komatsu, H.; Iwaoka, M.; Tomoda, S. J. Chem. Soc., Chem.
Commun. 1999, 205±206.
Acknowledgements
28. We choose to calculate charges by NBO analysis rather than
by Mulliken population analysis, as proposed by Reed et al.,
see Ref. 24.
Financial support from the University of Palermo 1funds
for selected research topics) and Italian MURST within
the National Research Project, `Non-aromatic hetero-
cycles in stereocontrolled processes' is gratefully
acknowledged.
29. The percentage of bond formation in the TS for the C´´´Se
bond may be evaluated as the ratio100 1dP2dTS)/1dP2dI)
where dI, dTS and dP are the C´´´Se distances in the selen-
iranium ions I±III, in the TSs and in the related cyclization

6822
M. Gruttadauria et al. / Tetrahedron 57 .2001) 6815±6822
products 1the conjugate acid form) respectively. In the same
way the percentage of formation for the O´´´C bond may be
estimated as 100 1d0I2d0TS)/1d0 2d0P) where d0 , d0TS and d0P
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Mont-
gomery, Jr., J. A.; Stratmann, R. E.; Burant, J. C.; Dapprich,
S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.;
Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.;
Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski,
J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.;
Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill,
P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.;
Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A.
Gaussian 98, Revision A.6; Gaussian Inc: Pittsburgh, PA,
11998).
I
I
are the O´´´C distances in I±III, in the TSs and in the related
products respectively. The values for dI and d0I were deduced
from the `apt for closure c2' conformers.
30. Spichty, M.; Fragale, G.; Wirth, T. J. Am. Chem. Soc. 2000,
122, 10914±10916.
31. From the frontier orbitals interaction point of view, reactivity
should have an inverse dependence from such an energy gap.
32. In order to calculate the DH0 for the proton exchange process
the following values of Ef were used. For H2O:
275.58596 hartree.17 For H3O1: 275.89123.
33. Cooper, M. A.; Ward, A. D. Tetrahedron Lett. 1995, 36,
2327±2330.
34. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;