Synthesis of carba analogs of 6-O-(benzyl)-d-allal- and -d-galactal-derived allyl epoxides and evaluation of the regio- and stereoselective behavior in nucleophilic addition reactions

Article abstract of DOI:10.1016/j.tet.2011.04.036

The new racemic diastereoisomeric epoxides 6α and 6β, the carba analogs of the corresponding d-galactal- and d-allal-derived allyl epoxides have been synthesized and their regio- and stereoselective behavior examined in addition reactions with model O-, C-, N-, and S-nucleophiles. The results have indicated that epoxide 6β has a pronounced tendency toward anti-1,2-addition, whereas epoxide 6α shows interesting levels of syn- and/or anti-1,4-addition processes. A chiral recognition process found with epoxide 6β, turned out to be consistently reduced in epoxide 6α. All the results have been rationalized on the basis of conformational, steric, and stereoelectronic effects.

Full text of DOI:10.1016/j.tet.2011.04.036

Tetrahedron 67 (2011) 4696e4709
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Synthesis of carba analogs of 6-O-(benzyl)-
D
-allal- and - -galactal-derived allyl
D
epoxides and evaluation of the regio- and stereoselective behavior in nucleophilic
addition reactions
,
a
a
a
a
Valeria Di Bussolo a , Ileana Frau , Lorenzo Checchia , Lucilla Favero , Mauro Pineschi ,
*
Gloria Uccello-Barretta b, Federica Balzano b, Graziella Roselli c, Gabriele Renzi c, Paolo Crotti a
Dipartimento di Scienze Farmaceutiche, Universita di Pisa, Via Bonanno 33, I-56126 Pisa, Italy
Dipartimento di Chimica e Chimica Industriale, Universita di Pisa, Via Risorgimento 33, I-56126 Pisa, Italy
Scuola di Scienze e Tecnologie-Sezione Chimica, Universita di Camerino, Via S. Agostino 1, I-62032 Camerino, Italy
,
*
a
ꢀ
b
ꢀ
c
ꢀ
a r t i c l e i n f o
a b s t r a c t
Article history:
The new racemic diastereoisomeric epoxides 6a and 6b, the carba analogs of the corresponding D-gal-
Received 20 November 2010
Received in revised form 22 March 2011
Accepted 11 April 2011
actal- and
havior examined in addition reactions with model O-, C-, N-, and S-nucleophiles. The results have
indicated that epoxide 6 has a pronounced tendency toward anti-1,2-addition, whereas epoxide 6
shows interesting levels of syn- and/or anti-1,4-addition processes. A chiral recognition process found
with epoxide 6 , turned out to be consistently reduced in epoxide 6 . All the results have been ratio-
nalized on the basis of conformational, steric, and stereoelectronic effects.
Ó 2011 Elsevier Ltd. All rights reserved.
D-allal-derived allyl epoxides have been synthesized and their regio- and stereoselective be-
b
a
Available online 16 April 2011
b
a
This work is dedicated to the memory of
Professor David Y. Gin, outstanding scientist
and dear friend
Keywords:
Allyl epoxides
Glycal carba analogs
Glycomimetics
Nucleophilic addition
Regioselectivity
1. Introduction
or aziridine, were obtained in a completely stereoselective way by
means of a new, uncatalyzed, completely stereoselective, directly
substrate-dependent glycosylation process.2,3
The structural resemblance to the parent sugars makes carba-
sugars attractive compounds for acting as carbohydrate mimetics
with a possible recognition by enzymes or other biological systems
in place of the related ‘true’ sugars. For this reason, carba analogs
with similar or even improved biological properties and/or with
increased stability toward endogenous degradation, compared
with those of the parent carbohydrate structures, have been pre-
pared with interesting biological properties, in particular as enzy-
matic inhibitors.1
O
O
BnO
BnO
X
X
1β
2β
1α
2α
X = O
X = N-Ns
The versatility and the efficiency of the new glycosylation pro-
cess made its application in a reiterative version possible on ep-
Epoxides 1a and 1b and related N-nosyl aziridines 2a and 2b
oxide 1
b for the synthesis of 2,3-unsaturated-1,6-di- 3 and
were found to be excellent glycosyl donors, particularly in glyco-
sylation of O-nucleophiles, as alcohols, where corresponding alkyl
O-glycosides, having the same configuration as the starting epoxide
trisaccharides 4.4 Subsequently, disaccharide 3 was dihydroxylated
to the corresponding fully-OH substituted disaccharide 5 by
a completely stereofacial selective process (Scheme 1).5
On the basis of these results, we thought it interesting to
examine the regio- and stereoselective behavior in nucleophilic
* Corresponding authors. Tel.: þ390502219690; fax: þ390502219660; e-mail
addresses: valeriadb@farm.unipi.it (V. Di Bussolo), crotti@farm.unipi.it, paolo.
crotti@farm.unipi.it (P. Crotti).
addition reactions of the new diastereoisomeric allyl epoxides 6
a
and 6 , the carba analogs of glycal-derived epoxides 1 and 1b,
b
a
0040-4020/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2011.04.036

V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
4697
O
O-t-Bu
O
O
O
O
BnO
HO
HO
HO
O
O-t-Bu
4
O
O
O
BnO
BnO
HO
O
O-t-Bu
OH
O
O
HO
BnO
O
HO
OH
HO
1
3
5
OH
OH
Scheme 1. 2,3-Unsaturated-1,6-di- and trisaccharides from epoxide 1b.
with a view toward their subsequent use in the synthesis of carba
oligosaccharides, analogs of 3e5.
means of the bulky TBSCl, affording the corresponding O-TBS de-
rivative 10. Mesylation (MsCl/Py) of the residual secondary hy-
droxyl group of 10 yielded mesylate 11. Deprotection of mesylate 11
by the TBAF/THF protocol provided trans hydroxy mesylate 12,
which on treatment under alkaline conditions (t-BuOK), yielded the
BnO
BnO
O
O
desired cis epoxide 6b through a completely regio- and stereo-
6α
6β
selective process (66% overall yield starting from trans diol 9, four
steps).
Because of some difficulties initially encountered in the mono-
benzylation, triol 7 was initially protected at the primary hydroxyl
functionality as corresponding O-trityl ether 9-Tr. Application to 9-
Tr of the same protocol previously described for the monobenzyl
ether 9 led, through the formation of the corresponding O-trityl
derivatives 10e12-Tr, to the synthesis of epoxide 6b-Tr (Scheme 4).
Whilst useful for some nucleophilic addition reactions under al-
kaline conditions, the O-trityl protecting group present in epoxide
2. Results and discussion
2.1. Synthesis of epoxides 6
a
and 6b
For the synthesis of racemic epoxide 6b 6
, the known racemic
c-3-hydroxymethyl-5-cyclohexen-r-1,t-2-diol 7,7a obtained by sa-
ponification (catalytic MeONa in MeOH) of the corresponding tri-
acetate 8,7 was considered an appropriate precursor. Subsequently,
6b
-Tr turned out, unfortunately, not to be stable even under the
epoxide 6
b
could be the precursor of the diastereoisomeric carba-
(Scheme 2).
mild acid conditions necessary for epoxide ring opening. As a con-
epoxide 6
a
Scheme 2. Retrosynthetic analysis for epoxides 6a and 6b.
Following the pioneering procedure described by Ogawa,8a,b
triacetate 8 was prepared starting from DielseAlder reaction be-
tween furane and methyl acrylate,8c as shown in Scheme 3.
Scheme 4. Synthesis of O-benzyl- and O-trityl-protected epoxide 6b and 6b-Tr.
sequence, we decided to go back to the O-benzyl protection, which
was finally realized, as described above (Scheme 4).
The synthesis of diastereoisomeric epoxide 6
diol 13, which is reasonable to expect may be prepared by acid
or alkaline hydrolysis of epoxide 6 . Actually, application of the
well-known KOH/DMSO/90 ꢀC protocol9 to epoxide 6
led to
desired trans diol 13, which was unexpectedly obtained in an un-
satisfactory low yield (50%). trans Diol 13 was subjected to the usual
protocol: regioselective protection of the allyl hydroxyl group to
give the corresponding O-TBS derivative 14, which was trans-
formed (MsCl/Py) into mesylate 15, and subsequently deprotected
(TBAF/THF) with the formation of trans hydroxy mesylate 16. Base-
a starts from trans
b
b
Scheme 3. Ogawa’s synthesis of triol 7.
The regioselective benzylation of the primary alcoholic func-
tionality of triol 7 by means of BnBr in the presence of hindered
LHMDS as the base, afforded the monobenzyl derivative 9, which
was subjected to the usual protection/deprotection protocol, as
shown in Scheme 4. Following this protocol, monobenzyl derivative
9 was selectively protected at the secondary allyl hydroxyl group by
catalyzed (t-BuOK/MeCN) cyclization of 16 afforded epoxide 6
a,
which was obtained with 27% overall yield (five steps, Scheme 5).

4698
V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
with the exclusive and quantitative formation of trans-hydroxy ac-
etate 18 (99% yield),11 sufficiently pure to be directly transformed
(MsCl/Py) into the corresponding trans mesyloxy acetate 20, the ul-
timate precursor of epoxide 6a. Actually, the treatment of 20 under
basicconditions (t-BuOK/MeCN) resulted in the alkalinehydrolysisof
the acetate group, followed by intramolecular SN2-type reaction of
the intermediate alcoholate on the vicinal carbon, with theformation
of the desired epoxide 6
Inthisway, startingfromdiastereoisomericepoxide6
was obtained with a good overall yield (62%) through a reduced (three
a (Scheme 7).
b, epoxide6a
Scheme 5. Synthesis of epoxide 6a.
steps) synthetic sequence. Epoxides 6a, 6b, and 6b-Tr turned out to be
sufficiently stable, and could be stored for a long time at ꢁ15 ꢀC.
The unsatisfactory overall yield obtained in the synthesis of
epoxide 6 , mostly due to the low yield (50%) observed in the first
step prompted us to check other opening reaction conditions of
epoxide 6 in order to obtain trans diol 13, or a corresponding
a
2.2. Nucleophilic addition reaction to epoxides 6a and 6b
b
synthetically useful derivative, with a decidedly better yield. Un-
fortunately, other opening procedures initially taken into consid-
eration were unsatisfactory too (Scheme 6); (a) the acid hydrolysis
Simple O-, N-, S-, and C-nucleophiles were taken as appropriate
models in order to check the regio- and stereoselective behavior of
diastereoisomeric epoxides 6a and 6b in nucleophilic addition re-
actions. In all cases, the addition products obtained were generi-
cally indicated as anti-1,2-addition products (trans-1,2-derivatives)
and/or syn- and anti-1,4-addition products (cis- and trans-1,4-de-
rivatives), and for the sake of convenience an arbitrary nomencla-
ture and numbering was used, as shown in Scheme 8. Protocol A
of epoxide 6
b (0.01 N TsOH in THF/H2O) turned out not to be
regioselective, and the desired trans diol 13 was accompanied by
a substantial amount of the corresponding 1,4-addition products
(30% in a mixture of cis and trans diastereoisomers 17), (b) the use
of commercially available Me3SiOꢁKþ, a KOH synthetic equivalent,
was completely ineffective, whereas (c) the use of the AcONa/H2O/
DMF (100 ꢀC) or tetrabutylammonium acetate (TBAAc, room tem-
perature) led to the desired trans diol 13, still with an unsatisfactory
yield for such a simple transformation (65e70%) (Scheme 6).10
At this point, we thought that the use, under acid conditions, of
a nucleophile weaker than H2O in order to minimize the formation of
regioisomers might represent the solution of the problem. A first
(nucleophile as the solvent) and/or protocol
B (nucleophile,
3e6 equiv, in a non-nucleophilic solvent) reaction conditions were
used, depending on the type of the nucleophile.
2.2.1. O-Nucleophiles. The regio- and stereoselective behavior of
epoxides 6a and 6b with O-nucleophiles was examined by means
of simple and synthetically significant MeOH and AcOH, used under
different reaction conditions (acid or basic conditions with differ-
ent epoxide/nucleophile ratio). The results obtained are collected in
Tables 1 and 2.
attempt based on the acetolysis ofepoxide 6b by means of0.2 N TsOH
in AcOH led, once again, to a non-regioselective result, affording
a 77:23 mixture of trans-1,2-hydroxy acetate 18 and trans-1,4-hy-
droxy acetate 19 (Table 1 and Scheme 7). However, when the same
reactionwas repeated by means of an AcOH solution in CH2Cl2, in the
presence of 0.01 N TsOH, in such a way as to have a 1:3 epoxide/AcOH
ratio, a complete 1,2-regio- and anti stereoselectivity was observed,
The methanolysis of epoxide 6
b carried out under alkaline
conditions (MeONa/MeOH, protocol A) is completely 1,2-regiose-
lective and anti stereoselective, with the exclusive obtainment of
trans methoxy alcohol 21 (entry 1, Table 1 and Scheme 9).
Scheme 6. Synthesis of trans diol 13 from epoxide 6b.
Scheme 7. Acetolysis of epoxide 6b.

V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
4699
Scheme 8. anti-1,2- and syn- and anti-1,4-addition products from epoxides 6a and 6b.
Table 1
Regio- and stereoselectivity of solvolysis reactions of epoxide 6b with O-nucleophiles
anti-1,2-addition
product
syn- and anti-1,4 addition
product
Nu
BnO
HO
BnO
BnO
NuH
HO
O
Nu
6β
13
18
21
Nu = OH
Nu = OAc
Nu = OMe
17
19 (trans)
NuH = MeOH, AcOH, H2O
22 (trans), 23 (cis)
Entry
Reagents
T (ꢀC)
anti-1,2-Addition product
syn- and anti-1,4-Addition product
1
2
3
4
5
6
7
8
2 N MeONa/MeOH
MeOH/0.2 N H2SO4
MeOH/LiClO4
80
rt
90
rt
rt
rt
rt
rt
90
90
90
rt
>99
80
77
75
<1
<1
20 (trans 22/cis 23¼2:1)
23a
MeOH/Cu(OTf)2
25a
D3þ/MeOH in D2 (gas-phase)
MeOH/CH2Cl2/0.01 N TsOHb
AcOH/0.2 N TsOH
>99 (trans 22/cis 23¼4:1)
>99
77
>1
23 (trans 19)
AcOH/CH2Cl2/0.01 N TsOHc
AcONa/DMF/H2O
>99
>99d
>99d
>99
70
<1
<1
<1
<1
30a
9
10
11
12
TBAAc/DMF
2 N aq KOH/DMSO
H2O/THF/0.01 N TsOH
a
syn/anti Ratio not determined.
Epoxide/MeOH ratio 1:6.
Epoxide/AcOH ratio 1:3.
trans Diol 13.
b
c
d
Table 2
Regio- and stereoselectivity of solvolysis reactions of epoxide 6a with O-nucleophiles
syn-1,2-addition
product
anti-1,2-addition
product
syn-1,4-addition
product
anti-1,4-addition
product
Nu
OMe
BnO
BnO
HO
BnO
HO
BnO
HO
BnO
HO
NuH
O
OAc
Nu
NuH = MeOH, AcOH
Nu = OMe
6α
26
-
24
28
25
29
Nu = OAc
-
27
Entry
Reagents
T (ꢀC)
syn-1,2-Addition product
anti-1,2-Addition product
syn-1,4-Addition product
anti-1,4-Addition product
1
2
3
4
5
6
2 N MeONa/MeOH
MeOH/0.2 N H2SO4
80
rt
rt
rt
rt
rt
d
d
d
d
28
30
>99
32
d
<1
44
76
47
16
32
<1
24
24
10
d
D3þ/MeOH in D2 (gas-phase)
MeOH/CH2Cl2/0.01 N TsOHa
AcOH/0.2 N TsOH
43
56
38
AcOH/CH2Cl2/5ꢂ10ꢁ3 N TsOHb
d
a
Epoxide/MeOH ratio 1:6.
Epoxide/AcOH ratio 1:3.
b

4700
V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
Scheme 9. Methanolysis of epoxide 6b under alkaline and acidic conditions.
When the methanolysis reaction is carried out under acid con-
leading to a mixture of the corresponding anti-1,2-addition product
(trans 1,2-methoxy alcohol 24, 32%) and 1,4-addition products (a
2:1 mixture of syn- and anti-1,4-addition products, the cis- 25, and
trans-1,4-methoxy alcohol 26), but the significant difference is that
the 1,4-addition is now the main addition process (68%). Also in this
case, the application of protocol B reaction conditions determines
a slight decrease in the overall 1,4-regioselectivity (57%), which is
accompanied by an increase in the syn-1,4-/anti-1,4-addition
product stereoselectivity (from 2:1 to 4.7:1, entry 4, Table 2).
ditions (MeOH/H2SO4 0.2 N, protocol A), the reaction is not regio-
selective, and leads to an 80:20 mixture of the corresponding
1,2- and 1,4-addition products. Inside this mixture, while the 1,2-
addition process is completely anti stereoselective and leads to the
trans methoxy alcohol 21, the 1,4-addition process is not stereo-
selective, leading to a 1:2 mixture of the corresponding syn (the
cis-1,4-methoxy alcohol 23) and anti-1,4-addition product (the
trans-1,4-methoxy alcohol 22) (entry 2, Table 1 and Scheme 9).
Considering that the 1,4-addition of an appropriate O-nucleo-
Analogously to observations with epoxide 6
b, the methanolysis of
phile to epoxide 6
b
is necessary for the synthesis of ‘carba’ oligo-
epoxide 6 under gas-phase-operating conditions led to a com-
a
saccharides, the present results obtained in the methanolysis were
not encouraging, seeing the low amount of 1,4-regioselectivity
obtained in these reactions. For this reason, we tried to increase the
desired 1,4-regioselectivity by carrying out the same reactions
under different reaction conditions (Scheme 9).
pletely 1,4-regioselective result, but now the reaction was largely
syn stereoselective and the syn-1,4-addition product (the cis-1,4-
methoxy alcohol 25) was the main reaction product (76%).
Decidedly different is the behavior of epoxide 6a in acetolysis.
Actually, under these conditions a substantial amount of syn-1,2-
addition product, the cis-1,2-acetoxy alcohol 27, was obtained, to
the point that an almost 1:1:1 mixture of 27, trans-1,2-acetoxy
alcohol 28, and cis-1,4-acetoxy alcohol 29 was obtained under
protocol B reaction conditions. Corresponding syn-1,2-addition
products were never observed in the opening reactions of epoxide
In view of our continuing interest in the regio- and stereo-
selectivity of the opening reactions of epoxides under gas-phase
reaction conditions, we thought it interesting to repeat the meth-
anolysis reactions of epoxide 6
b
also in the gas-phase.12 Following
in the gas-
the typical protocol,13 the methanolysis of epoxide 6
b
phase turned out to be completely 1,4-regioselective, with the
exclusive formation of a 4:1 mixture of the corresponding anti-
(trans-1,4-methoxy alcohol 22) and syn-1,4-addition product (cis-
1,4-methoxy alcohol 23).14
6
b
with O-nucleophiles (Tables 1 and 2).
The sufficiently satisfactory 1,4-regioselective result observed in
the methanolysis, under protocol B reaction conditions, indicated
epoxide 6 as a potentially useful candidate for the construction of
O-linked carba oligosaccharides. This possibility was checked by
examining the addition reaction of epoxide 6 to 3-cyclohexene-1-
a
On the basis of this result, we repeated the acid methanolysis of
epoxide 6
b
in the condensed-phase, trying to adopt reaction con-
a
ditions similar to the gas-phase by using a 0.01 N TsOH/CH2Cl2
solution containing a drastically reduced amount of methanol, in
such a way that the epoxide/MeOH ratio was 1:6. Unfortunately,
methanol, taken as a simplified model of a carba-monosaccharide
with a free primary alcoholic functionality (eCH2OH). The addition
reaction was carried out in CH2Cl2 in the presence of TsOH (0.005 N)
with a 1:3 epoxide/nucleophile ratio. After two days stirring at room
temperature, the crude reaction mixture showed the presence of
three addition products, the corresponding anti-1,2-addition prod-
uct 30 and the regioisomeric anti- 31 and syn-1,4-addition product
32 in a 30:28:42 ratio. Separation of this mixture by preparative TLC
afforded the syn-1,4-addition product 32 pure, whereas the
remaining anti-1,2- 30 and anti-1,4-addition product 31 were still
obtained in a mixture. The structure of the obtained pure addition
also in these conditions (protocol B), the methanolysis of epoxide 6
b
turned out to be completely 1,2-regioselective with the exclusive
formation of the corresponding anti-1,2-addition product, the
trans-1,2-methoxy alcohol 21 (Scheme 9 and Table 1).
The completely or highly 1,2-regioselective behavior found in
the methanolysis of epoxide 6
b was also found in the corre-
sponding acetolysis reaction (see Scheme 7 and Table 1). Under
protocol A reaction conditions (0.2 N TsOH in AcOH), a 77:23
mixture of the corresponding anti-1,2-addition product (trans-1,2-
hydroxy acetate 18) and anti-1,4-addition product (trans 1,4-hy-
droxy acetate 19) was obtained, whereas under protocol B reaction
conditions (CH2Cl2 as the solvent, epoxide/AcOH/TsOH¼1:3:0.05),
the reaction was completely 1,2-regioselective and the trans-hy-
droxy acetate 18 turned out to be the only reaction product.15
product 32 corresponds tothat of a simplified a-O-linked-1,6-carba-
disaccharide, as desired (Scheme 10).
As a control, the same reaction was repeated also on the di-
astereoisomeric epoxide 6 . As expected a complete 1,2-regio- and
b
anti stereoselectivity was obtained and the corresponding anti-1,2-
addition product 33 was the only reaction product (Scheme 11).
As indicated by a theoretical study carried out on corresponding
simplified, 6-CH2OMe-substituted models (see Supplementary
data) and confirmed by 1H NMR conformational analysis, epox-
The methanolysis of the diastereoisomeric carba-epoxide 6
indicated that the behavior of this epoxide is similar to that of
epoxide 6 , but significant and synthetically useful differences are
present (Table 2).
As with 6 , the alkaline methanolysis (MeONa/MeOH) of ep-
oxide 6 is completely 1,2-regio- and anti stereoselective and the
acid methanolysis (MeOH/0.2 N H2SO4) is not regioselective,
a
b
ides 6
>99% in the case of 6
a
and 6
b
exist almost exclusively (96% in the case of 6
a and
b
b
) as the corresponding conformer 6
a
0 and 6b0
a
with the side chain axial and equatorial, respectively (Schemes 12
and 13).16

V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
4701
Considering that stereoelectronic factors associated with the
opening processes of three-membered rings make trans diaxial
opening of a cycloaliphatic oxirane the favored opening process, the
conformational population inside 6
opinion, for the opposite regioselective behavior found for these
epoxides in methanolysis reactions. Actually, in epoxide 6 , nu-
cleophilic attack on the C(2) allyl oxirane carbon, from the -side, of
the only existing protonated conformer 6
0-H (route a) corresponds
a and 6b is responsible, in our
b
a
b
to the requirement for a trans diaxial opening process, and does not
show any particular steric hindrance. Accordingly, the anti-1,2-ad-
dition process is correctly the main opening process with this ep-
oxide (Scheme 12 and Table 1), whereas the 1,4-addition process
(routes b and c) is decidedly less important.
Scheme 10. Addition of 3-cyclohexene-1-methanol to epoxide 6a.
In diastereoisomeric epoxide 6
opening of the oxirane ring can only occur by nucleophilic attack
from the -side of the largely existing corresponding protonated
conformer 6
0-H, but in this case the attack is subjected to an un-
a, the corresponding trans diaxial
b
a
favorable 1,3-diaxial interaction (torsional strain) with the C(6)e
C(7) bond of the axial side chain (route a, Scheme 13). As a conse-
quence, anti-1,2-addition process is sufficiently slowed down to
make the 1,4-addition process [routes b and c from more stable
protonated conformer 6
protonated conformer 6
a
0-H and routes d and e from less stable
a
00-H] competitive to the point of becoming
Scheme 11. Addition of 3-cyclohexene-1-methanol to epoxide 6b.
the main addition process. Inside the 1,4-addition pathway, in
agreement with the experimental results, the syn-1,4-addition
process (routes b and e), which is not subjected to any particular
strain (route b) or corresponds to a pseudoaxial attack (route e),
prevails over the alternative anti-1,4-addition process (routes c and
d) for which a 1,3-diaxial interaction with the axial side chain is
present (route c) or it corresponds to a less favored pseudoequa-
torial attack (route d) (Scheme 13). In this framework, considering
that the 1,2-addition process with retention of configuration is fa-
vored by the axial opening of the oxirane ring and formation of
a discrete (not free) allyl carbocationic species (possible only in
protonated conformer 6
chain toward nucleophilic attack from the
a
0-H),17a the slowing steric effect of the side
face of 6
0-H, could be
b
a
responsible of the formation of a substantial amount (27%) of the
corresponding syn-1,2-addition product (cis derivative 27) with the
less nucleophilic AcOH. Evidently, in these conditions, a syn-1,2-
addition process, which necessarily develops entirely on the less
hindered
A comparison of the regio- and stereoselective behavior of ep-
oxides 6 and 6 with the corresponding glycal-derived epoxides
a
face, can become competitive (route f, Scheme 13).17b
Scheme 12. Methanolysis of epoxide 6
b.
a
b
Scheme 13. Methanolysis of epoxide 6a.

4702
V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
1
a
and 1
b
in their reactions with MeOH2 appears particularly in-
and 1 react with MeOH
previously stated reduced reactivity at the C(2) oxirane carbon for
this epoxide.
TMSN3 (3 equiv) in MeCN reacts with epoxides 6
the presence of a Lewis acid [Sc(OTf)3, (0.15 equiv)]. In these con-
ditions, epoxide 6 leads to a complex reaction mixture mostly
consisting of trans-1,2-azido alcohol 40 (anti-1,2-addition product),
as the only addition product, whereas epoxide 6 , in accordance
teresting. Glycal-derived epoxides 1
a
b
under practically neutral reaction conditions at room temperature,
in the absence of any catalyst, affording only corresponding 1,4-
addition products, as an almost 1:1 syn-/anti-1,4-addition product
mixture, under protocol A, and only the corresponding syn-1,4-ad-
dition product, under protocol B (Scheme 14).2,18 On the contrary,
a and 6b only in
b
a
with the present epoxides 6
a
and 6
b
, decidedly alkaline (MeONa/
with its lower reactivity, leads only to a non-addition product, the
MeOH) or acidic conditions (MeOH/H2SO4 or TsOH) are necessary
for the opening process to occur, and the reactions are completely
(basic conditions) or consistently (acid conditions, particularly with
unsaturated ketone 39 (Scheme 16).20
2.2.3. S-Nucleophiles. PhSH was used as typical S-nucleophile and
epoxide 6
b) 1,2-regio- and anti stereoselective (Tables 1 and 2).
no substantial differences were observed in the behavior of either
Scheme 14. Methanolysis of epoxides 1a and 1b, under protocol B reaction conditions.
We think that the decidedly different, almost opposite, regio-
and stereoselective behavior observed in the methanolysis of
epoxides 1a and 1b with respect to epoxides 6a and 6b can be
attributed to the presence in epoxides 1a and 1b of the endocyclic
oxygen, which facilitates by conjugation the development of
a partial positive charge on C(4) during the opening process. In
association with the hypothesized epoxide/nucleophile co-
ordination (hydrogen bond),2 this makes the attack of the poorly
nucleophilic MeOH more favorable only at the more electrophilic
C(4) carbon, further from the partially negative oxirane oxygen.2
It is interesting to note that a regioselective behavior very similar
to that observed with glycal-derived epoxides 1
condensed-phase (complete 1,4-regioselectivity) is obtained in the
acid methanolysis of epoxides 6 and 6 in the gas-phase. Evidently,
a and 1b in the
a
b
under these particular reaction conditions, characterized by the
absence of any solvent and acid counterion, the developing of
a partially positive charge on the C(4) carbon atom, then attacked by
the nucleophile by a SN20 process, is particularly favored.19
Scheme 15. Aminolysis of epoxides 6a and 6b with diethylamine.
epoxides. Actually, the reactions of epoxides 6
a and 6b with
PhSH/NEt3 (protocol B) are completely 1,2-regio- and anti stereo-
selective, with the exclusive formation of the corresponding anti-
1,2-addition product, trans-1,2-phenylthio alcohol 42 and 43 from
2.2.2. N-Nucleophiles. As examples of N-nucleophiles, diethyl-
amine and an azide ion (from NaN3 and TMSN3) were considered in
their reactions with epoxides 6a and 6b.
6a and 6b, respectively (Scheme 17). Due to the high nucleophi-
Epoxides 6 and 6 did not react with diethylamine under
a
b
licity of PhSH, both reactions are very fast (30 min) and do not need
the presence of any Lewis acid catalyst to occur.
protocol A; even after several days, the starting epoxides were re-
covered completely unreacted. However, if the reactions were re-
peated in the same conditions but in the presence of a Lewis acid
(LA), such as Sc(OTf)3 (0.1e0.2 equiv) completely 1,2-regio- and anti
stereoselective reactions were observed, with the exclusive for-
2.2.4. C-Nucleophiles. As the addition of organocopper reagents to
allylic epoxides is an important method for regio- and stereo-
selective carbonecarbon bond formation,21 the behavior of epox-
mation of trans-1,2-N,N-diethylamino alcohols 35, from 6
from 6 (anti-1,2-addition products). Even if the regio- and ster-
eoselectivity observed is the same, the corresponding reactivity is
decidedly different: epoxide 6 completely reacts in 30 h, whereas
for the complete conversion of epoxide 6 , extremely long reaction
a, and 36,
ides 6
a and 6b with some methyl-based organocopper reagents
b
was examined (Tables 3 and 4 and Fig. 1).
The addition reaction of Me2CuLi to epoxide 6b is not regiose-
b
lective, but completely anti stereoselective, and leads to a 57:43
mixture of the corresponding anti-1,2-addition product (trans-1,2-
methyl alcohol 45) and anti-1,4-addition product (trans-1,4-methyl
alcohol 46) (entry 1, Table 3). On the contrary, the corresponding
a
times (10 days) were necessary. The previously discussed stereo-
electronic factors associated with the trans diaxial opening of the
oxirane ring of epoxide 6
sponsible for the different reactivity observed (Scheme 15).
The azidolysis of epoxides 6 and 6 , carried out with NaN3
a, absent in epoxide 6b, could be re-
reaction of epoxide 6a turned out to be neither regio- nor stereo-
selective, affording a mixture of anti-1,2-addition product (trans-
1,2-methyl alcohol 47, 26%), anti-1,4-addition product (trans-1,
4-methyl alcohol 48, 23%), and the syn-1,4-addition product (cis-
1,4-methyl alcohol 49, 25%), accompanied by a non-addition
product, ketone 39 (26%) (entry 1, Table 4).
a
b
(5 equiv)/NH4Cl in MeOH/H2O turned out not to be regioselective,
but completelyanti stereoselective. Actually, 8:2 and 55:45 mixtures
of the corresponding anti-1,2-addition products (trans-1,2-azido
alcohols 37 and 40) and anti-1,4-addition products (trans-1,4-azido
The complete anti stereoselectivity observed in the addition of
alcohols 38 and 41) were obtained from 6
(Scheme 16). The larger amount of the corresponding 1,4-addition
products found for epoxide 6 (45%) is the consequence of the
a and 6b, respectively
Me2CuLi to epoxide 6b is in agreement with the typical behavior of
cuprates with allyl oxirane systems.21 In this framework, the initial
a

V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
4703
Scheme 16. Azidolysis of epoxides 6a and 6b by the NaN3/NH4Cl and TMSN3/Sc(OTf)3 protocol.
(Scheme 18). A competitive isomerization of 51, through a Cu(III)
-allyl system as 52, to the regioisomeric anti- -Cu(III) intermediate
p
a
53 and subsequent reductive elimination, could be responsible of the
formation of the anti-1,2-addition product 45. A comparison of the
present result with the one obtained by Marino in the addition of
Me2CuLi to 1,3-cyclohexadiene monoepoxide, where an almost 1:1
mixture of the corresponding anti-1,2- and anti-1,4-addition product
was obtained,22 would indicate that the eCH2OBn side chain, present
in epoxide 6
In epoxide 6
fashion to the oxirane ring and subsequent oxidative addition to
anti-
-Cu(III) intermediate 55 (route a, X¼Me, Scheme 19), which
can lead to both anti-1,4- 48 (by reductive elimination) and
anti-1,2-addition product 47 (through corresponding -allyl system
and anti- -Cu(III) intermediate 56) suffers from the presence of the
b
, has a modest influence on the addition process.
a
, the formation of a complex as 54 in an anti
p
g
p
a
Scheme 17. Addition reaction of PhSH to epoxides 6a and 6b.
axial side chain (steric hindrance). As a consequence, a corre-
sponding, not similarly hindered, syn-1,4-addition process (route b)
leading to the syn g-Cu(III) intermediate 58, through p complex 57,
formation of a
as 50 is reasonably followed by an oxidative addition to the allylic
system to give the anti -Cu(III) intermediate 51, which can directly
p complex, from the opposite side of the oxirane ring,
becomes highly competitive and a substantial amount of the syn-
1,4-addition product (cis-1,4-methyl alcohol 49, 25%) is obtained,
after the reductive elimination step (Scheme 19).23
g
undergo reductive elimination to anti-1,4-addition product 46
Table 3
Ring opening reaction of epoxide 6b with methyl-based organocopper reagents
anti-1,2-addition
product
anti-1,4-addition
product
Me
BnO
BnO
HO
BnO
HO
O
Me
45
6β
46
Entry
Reagents
Me2CuLi
MeMgBr/CuCN
T (ꢀC)
anti-1,2-Addition product
anti-1,4-Addition product
1
2
3
4
ꢁ15
ꢁ40
ꢁ78
ꢁ78
57
7
43
93
Me2Zn (0.6 equiv)/Cu(OTf)2/(R,R,R)-44a
Me2Zn (1.5 equiv)/Cu(OTf)2/(R,R,R)-44b
>99 (79% ee)
72 (45% ee)
28 (98% ee)
a
At 26% conversion, unreacted epoxide 24% ee.
Conversion (100%).
b

4704
V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
Table 4
Ring opening reaction of epoxide 6
a
with methyl-based organocopper reagents
anti-1,2-addition
anti-1,4-addition
product
syn-1,4-addition
product
non-addition
product
product
Me
Me
BnO
HO
BnO
HO
BnO
HO
BnO
BnO
O
O
Me
47
6α
48
49
39
Entry
Reagents
Me2CuLi
MeMgBr/CuCN
T (ꢀC)
47
48
23
90
>99 (24% ee)
86 (16% ee)
49
39
1
2
3
4
ꢁ15
ꢁ40
26
10
d
d
25
d
d
d
26
d
d
d
Me2Zn (1.5 equiv)/Cu(OTf)2/(R,R,R)-44a
Me2Zn (1.5 equiv)/Cu(OTf)2/(R,R,R)-44b
ꢁ78/ꢁ50
ꢁ78/rt
a
Conversion (40%). Under the same reaction conditions and 71% conversion, compound 48 showed 16% ee.
Conversion (100%): a non-addition by-product was also present (14%).
b
The formation of ketone 39 in this reaction is a further dem-
onstration of the reduced tendency of epoxide 6 to addition re-
a
actions, to the point that also an LA-catalyzed isomerization process
becomes competitive. In this case, the reagent (e.g., Me2CuLi$LiI)
behaves as the LA catalyst, too (Scheme 20).
Even if modest as regards the regio- and/or stereoselectivity, the
addition reaction of Me2CuLi to epoxides 6
a and 6b is synthetically
useful, because it turned out to be the only protocol available for
Fig. 1. Chiral ligand (R,R,R)-44 used in Me2Zn/Cu(OTf)2 nucleophilic addition protocol.
Scheme 18. Formation of anti-1,2- and anti-1,4-addition product from epoxide 6b.
Scheme 19. Formation of anti-1,2-, anti-1,4-, and syn-1,4-addition product from epoxide 6a.

V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
4705
Scheme 20. Formation of ketone 39.
a substantial formation of the corresponding anti-1,2-addition
products from both epoxides and the corresponding syn-1,4-addi-
4-methyl alcohol 48), showed a low enantiomeric excess (24% ee,
entry 3, Table 4). Even forcing the reaction to completion, by op-
erating at room temperature and in the presence of an excess of
Me2Zn, the result did not change and the trans-1,4-methyl alcohol
48 obtained in these conditions showed a lower enantiomeric ex-
cess (16% ee), and was accompanied by a certain amount (14%) of
a non-addition by-product, not further examined (entry 4, Table 4).
Evidently, the approaching and the subsequent coordination of the
tion product from epoxide 6
As for the regio- and stereoselectivity, decidedly better results
were obtained in the addition reactions of MeMgBr to epoxides 6
and 6 in the presence of catalytic amounts of CuCN. Actually, with
a (Tables 3 and 4).
a
b
this reagent, only mixtures of the corresponding anti-1,2-addition
products and anti-1,4-addition products were obtained from both
epoxides, with the corresponding anti-1,4-addition product, the
chiral copper catalyst to the double bond of epoxide 6a, which
trans-1,4-methyl alcohol 48 (90%), from epoxide 6
a
, and 46 (93%),
would normally proceed from the opposite side of the oxirane ring
is sterically hindered by the presence of the axial benzyloxymethyl
moiety on C(6) (see intermediates 54 and 55, X¼phosphoramidite,
Scheme 19), with a subsequent detrimental effect on the chiral
recognition process.
fromepoxide 6 , as the main reactionproducts (entry 2, Tables 3 and
b
4). With regard to this, it is widely accepted that the electron-
withdrawing nature of eCN group destabilizes the initially formed
anti-
g-Cu(III) intermediates 51 (from epoxide 6
b) and 55 (from ep-
oxide 6
a) accelerating the reductive elimination step from these
intermediates with a subsequent marked increase of the 1,4-regio-
selectivity, as experimentally found (X¼CN, Schemes 18 and 19).24
A complete or highly regioselective formation of anti-1,4-addi-
tion products 46 and 48 was obtained, under kinetic resolution
3. Structures and configurations
The regioisomeric 1,2- and 1,4-addition product structure of the
products obtained in the opening reactions of epoxides 6a and 6b
conditions, in the addition reactions of Me2Zn to allyl oxiranes 6
b
has been simply determined by means of 1H NMR COSY and NOESY
experiments. An appropriate NMR conformational analysis has
indicated that all the obtained 1,2- and 1,4-addition products from
and 6a, respectively, in the presence of a catalytic amount of copper
complex with chiral phosphoramidite 44 (entry 3, Tables 3 and 4).25
Even if apparently similar, these reactions have shown significant
epoxide 6a and 6b preferentially exist as the corresponding con-
differences between the two epoxides 6
a
and 6
b
, which are par-
former with the side chain equatorial, respectively. In particular,
axial arrangement of H(6) proton was established on the basis of
the dipolar interaction detected between H(6) and only one of
adjacent methylene protons, H(50). Relative arrangement of protons
ticularly interesting. With epoxide 6 , the complete 1,4-regiose-
a
lectivity and anti stereoselectivity initially obtained at ꢁ78 ꢀC (40%
conversion), with the exclusive formation of trans-1,4-methyl al-
cohol 48 (anti-1,4-addition product), is maintained also when the
reaction temperature is allowed to rise to room temperature in
order to have a complete conversion of the starting epoxide into 48
H(1) and H(6) of products from epoxide 6
the basis of coupling constants pattern for H(1) and its dipolar in-
teraction with proton H(5). In the case of products from epoxide 6
a was assessed both on
b
,
(entries 3 and 4, Table 4). Epoxide 6
b
turned out to be more reactive
H(1) and H(6) were, respectively, in equatorial and axial arrange-
ment as shown by coupling constants values and detection of inter-
NOE H(1)eH(6). On this basis, considering that the configuration at
C(1) necessarily corresponds to that of the starting epoxide, the
relative configuration at the remaining C(2) or C(4) carbons in each
1,2- and 1,4-addition product, respectively, has been firmly estab-
lished on the basis of dipolar interactions involving protons H(2),
H(1), and H(6) in 1,2-addition products and protons H(4), H(5),
H(50), and H(6) in 1,4-addition products and by examination of the
coupling constants associated to these protons. When Nu¼Me,
OMe, also NOE between the protons of these groups and H(6), H(5),
and H(50) have been considered (Scheme 21).
than 6 , and a complete 1,4-regioselectivity and anti stereo-
a
selectivity, with the exclusive formation of trans-1,4-methyl alcohol
46 (anti-1,4-addition product), with a consistent 79% ee, was ob-
served at low temperature (ꢁ78 ꢀC) and in the presence of a low
amount of Me2Zn (0.6 equiv) (entry 3, Table 3). When the amount
of organometallic reagent was increased (1.5 equiv) and the re-
action temperature was maintained at ꢁ78 ꢀC, there was a loss of
regioselectivity. Actually, under these conditions, the anti-1,4-ad-
dition product (trans-1,4-methyl alcohol 46) was now obtained
with a reduced enantiomeric excess (45% ee), accompanied by the
corresponding anti-1,2-addition product (trans-1,2-methyl alcohol
45) in an almost 3:1 ratio (entry 4, Table 3). Interestingly, trans-1,2-
methyl alcohol 45 was obtained in a low yield (28%), but with a high
enantiomeric excess (98% ee). These data clearly indicate chiral
recognition of the chiral catalyst complex by the enantiomers of
4. Conclusions
In nucleophilic addition reactions with O-, N-, and S-nucleo-
epoxide 6
b
. However, this recognition is inferior to that previously
philes, under basic and/or acid conditions, epoxides 6
a regioselectivity favoring 1,2- rather than 1,4-addition process. The
only significant exception is offered by epoxide 6 with O-nucleo-
a and 6b show
observed by us, under the same reaction conditions, for the
enantiomers of 1,3-cyclohexadiene monoepoxide.25 In that case
a
the
increased
regiodivergency
(anti-1,4-/anti-1,2-addition
philes under protocol B reaction conditions and in the acid azidolysis,
product¼60:40) and enantiomeric excess found (97% and 64% ee
for the corresponding anti-1,2- and anti-1,4-addition product, re-
spectively) indicated the complementary reactions of the enan-
tiomers of the starting allyl epoxide.
where consistent amounts of corresponding 1,4-addition products
are obtained. These results make epoxide 6a a potentially useful
candidate for the construction of O-and N-linked carba oligosac-
charides. With methyl-based C-nucleophiles (Me2CuLi and
MeMgBr/catalytic CuCN), the regioselectivity is generally shifted
A corresponding chiral recognition is consistently reduced in the
case of epoxide 6
a
, where at 40% conversion, the only addition
toward 1,4-addition processes by both epoxides 6a and 6b. In the
product, the corresponding anti-1,4-addition product (trans-1,
addition reactions of Me2Zn, under kinetic resolution conditions in

4706
V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
Scheme 21. NOE in anti- and syn-1,2- and -1,4-addition products from epoxides 6a and 6b.
the presence of a catalytic amount of a copper complex with chiral
phosphoramidite 44, a complementary reaction of the enantiomers
NMR spectra were recorded at 250 and 62.5 MHz, respectively. 1H
NMR COSY and NOESY experiments were performed on a spec-
trometer operating at 600 MHz. The enantiomeric excesses for al-
of the starting allyl epoxide was observed with epoxide 6
the same conditions, a corresponding chiral recognition is com-
pletely absent in the case of epoxide 6 . The presence of the axial
eCH2OBn side chain in the more stable conformer 6a0 of epoxide 6
b. Under
cohols 45, 46, and 48 and epoxide 6b were determined by HPLC
a
Daicel Chiracel OD-H column. Triacetate 8 and triol 7 were pre-
a
,
pared as previously described.7
besides having, in general, a slowing effect toward nucleophilic at-
tack, is, in our opinion, responsible, in this epoxide, for the com-
petitive formation of 1,4-addition products, syn-1,2-addition
product (O-nucleophiles) and, in some cases, of non-addition
products (N- and C-nucleophiles) and for the absence of chiral rec-
ognition by the chiral Cu(OTf)2/(R,R,R)-phosphoramidite 44 com-
plex in addition reactions of Me2Zn.
5.2. Addition reaction of MeOH and AcOH (O-nucleophiles) to
epoxides 6b and 6a
5.2.1. Reaction of epoxide 6
Typical procedure. Epoxide 6
b
with 0.2 N H2SO4/MeOH (protocol A).
(0.080 g, 0.37 mmol) was added to
b
a 0.2 N H2SO4/MeOH (1.5 mL) and the resulting reaction mixture
was stirred 24 h at room temperature. After dilution with CH2Cl2,
solid NaHCO3 was added. Evaporation of the filtered organic solu-
tion afforded a crude product (0.087 g, 95% yield) consisting of
trans- 1,2-methoxy alcohol 21 (80%) and a 1:2 mixture of cis- and
trans-1,4-methoxy alcohol 23 and 22 (20%) (1H NMR), which was
subjected to preparative TLC, using an 8:2 hexane/AcOEt mixture as
the eluant. Extraction of the most intense bands afforded trans-1,2-
methoxy alcohol 21 (anti-1,2-addition product, 0.045 g, 49% yield)
and 1:2 mixture of 1,4-methoxy alcohols 23 and 22 (0.014 g). This
mixture was subjected to preparative TLC, using a 9:1 CH2Cl2/i-Pr2O
mixture as the eluant. Extraction of the two most intense bands
afforded cis-1,4-methoxy alcohol 23 (syn-1,4-addition product,
0.003 g, 3% yield) and trans-1,4-methoxy alcohol 22 (anti-1,4-ad-
dition product, 0.009 g, 9% yield).
5. Experimental
5.1. General
All reactions requiring anhydrous conditions were performed in
a flame-dried modified Schlenk (Kjeldahl shape) flasks fitted with
a glass stopper or rubber septa under a positive pressure of argon.
Anhydrous benzene, toluene, Et2O, and THF were obtained by
distillation from sodium/benzophenone. Flash column chroma-
tography was performed employing 230e400 mesh silica gel
(MachereyeNagel). Analytical TLC was performed on Alugram SIL
G/UV254 silica gel sheets (MachereyeNagel) with detection by 0.5%
phosphomolybdic acid solution in 95% EtOH. Routine 1H and 13C

V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
4707
5.2.1.1. (1R
hexen-1-ol (22). A liquid, Rf¼0.06 (9:1 CH2Cl2/i-Pr2O); FTIR (neat)
3445, 1449, 1095 cmꢁ1 1H NMR (CDCl3)
7.40e7.28 (m, 5H, Ph),
*
,4S
*
,6R
*
)-6-(Benzyloxymethyl)-4-methoxy-2-cyclo-
(0.011 g, 24% yield), and trans-1,4-methoxy alcohol 26 (anti-1,4-
addition product) (0.004 g, 9% yield).
n
.
d
6.07e5.94 (m, 2H, CH]CH), 4.55 (d, 1H, J¼12.0 Hz, CHaHbPh), 4.53
(d, 1H, J¼12.0 Hz, CHaHbPh), 4.36e4.27 (m, 1H, CHOH), 3.76e3.68
(m, 1H, CHOMe), 3.66e3.56 (m, 2H, CH2OBn), 3.35 (s, 3H, OMe),
2.43 (d, 1H, J¼4.6 Hz, OH), 2.41e2.23 (m, 1H, C(6)H), 1.88e1.73 (m,
1H, C(5)HaHb), 1.72e1.62 (m, 1H, C(5)HaHb). 13C NMR (CDCl3)
5.2.4.1. (1S
hexen-1-ol (25). A liquid, Rf¼0.13 (8:2 hexane/AcOEt); FTIR (neat)
3460, 1447, 1370, 1097 cmꢁ1. 1H NMR (CDCl3)
7.40e7.28 (m, 5H,
*
,4S
*
,6R )-6-(Benzyloxymethyl)-4-methoxy-2-cyclo-
*
n
d
Ph), 5.93e5.81 (m, 2H, CH]CH), 4.58 (d, 1H, J¼12.0 Hz, CHaHbPh),
4.53 (d, 1H, J¼12.0 Hz, CHaHbPh), 4.09 (d, 1H, J¼8.8 Hz, CHOH),
3.70e3.65 (m, 1H, CHOMe), 3.64e3.58 (m, 1H, CHaHbOBn), 3.48
(dt, 1H, J¼8.8, 2.6 Hz, CHaHbOBn), 3.36 (s, 3H, OMe), 2.21e2.03 (m,
1H, C(6)H), 1.86e1.74 (m, 1H, C(5)HaHb), 1.42e1.27 (m, 1H, C(5)
d
138.2, 132.3, 128.9, 128.7, 127.9, 127.8, 77.4, 73.4, 72.6, 71.9, 56.6,
34.8. Anal. Calcd for C15H20O3: C, 72.55; H, 8.12. Found: C, 72.14; H,
7.76. MS (m/z) 77, 79, 91, 109, 124, 157 (Mþ).
HaHb). 13C NMR (CDCl3)
d 138.0, 134.9, 128.7, 127.9, 127.8, 126.3,
5.2.1.2. (1R
hexen-1-ol (23). A liquid, Rf¼0.04 (9:1 CH2Cl2/i-Pr2O); FTIR (neat)
3431, 1452, 1097 cmꢁ1 1H NMR (CDCl3)
7.40e7.28 (m, 5H, Ph),
*
,4R
*
,6R
*
)-6-(Benzyloxymethyl)-4-methoxy-2-cyclo-
75.1, 73.7, 72.1, 71.8, 56.7, 37.1, 28.3. Anal. Calcd for C15H20O3: C,
72.55; H, 8.12. Found: C, 72.64; H, 7.59. MS (m/z) 77, 79, 91, 109,
124, 157 (Mþ).
n
.
d
5.60e5.87 (m, 2H, CH]CH), 5.54 (s, 2H, CH2Ph), 4.27e4.16 (m, 1H,
CHOH), 3.80 (dd, 1H, J¼10.0, 5.8 Hz, CHOMe), 3.67 (dd, 1H, J¼9.2,
5.8 Hz, CHaHbOBn), 3.51 (dd, 1H, J¼9.2, 5.8 Hz, CHaHbOBn), 3.38 (s,
3H, OMe), 1.96e1.87 (m, 2H, C(6)H and C(5)HaHb), 1.51e1.42 (m, 1H,
5.2.4.2. (1S
hexen-1-ol (26). A liquid, Rf¼0.15 (8:2 hexane/AcOEt); FTIR (neat)
3458, 1445, 1370, 1098 cmꢁ1. 1H NMR (CDCl3)
7.40e7.28 (m, 5H,
*
,4R
*
,6R
*
)-6-(Benzyloxymethyl)-4-methoxy-2-cyclo-
n
d
C(5)HaHb). 13C NMR (CDCl3)
d
138.4, 132.6, 130.1, 128.7, 127.9, 76.1,
Ph), 5.81 (dd, 1H, J¼10.3, 1.4 Hz, CH]C(3)H), 5.70e5.77 (m, 1H, C(2)
H]CH), 4.53 (s, 2H, CH2Ph), 4.25e4.16 (m, 1H, CHOH), 4.00e3.89
(m, 1H, CHOMe), 3.62 (dd, 1H, J¼8.9, 4.4 Hz, CHaHbOBn), 3.53 (t, 1H,
J¼8.9 Hz, CHaHbOBn), 3.35 (s, 3H, OMe), 2.05e1.85 (m, 2H, C(6)H
and C(5)HaHb), 1.23e1.15 (m, 1H, C(5)HaHb). 13C NMR (CDCl3)
73.5, 72.0, 64.8, 56.0, 37.8, 29.9 (1 aromatic signal overlapped). Anal.
Calcd for C15H20O3: C, 72.55; H, 8.12. Found: C, 72.24; H, 7.89. MS
(m/z) 77, 79, 91, 109, 121, 157 (Mþ).
5.2.2. Reaction of epoxide 6
(protocol B). Typical procedure. Epoxide 6
b
with 0.01 N TsOH in MeOH/CH2Cl2
(0.025 g, 0.116 mmol)
d 137.8, 132.3, 129.7, 128.7, 128.1, 127.9, 75.9, 75.2, 73.8, 72.1, 55.8,
b
41.7, 30.2. Anal. Calcd for C15H20O3: C, 72.55; H, 8.12. Found: C,
was added to a CH2Cl2 solution (1.2 mL) containing MeOH
(0.030 mL, 0.696 mmol, 6.0 equiv), TsOH$H2O (0.002 g,
0.012 mmol, 0.1 equiv) (epoxide/TsOH/MeOH¼1:0.1:6) and the
resulting mixture was stirred 24 h at room temperature. Dilution
with CH2Cl2 and evaporation of the washed (saturated aqueous
NaHCO3 and saturated aqueous NaCl) afforded a crude product
consisting of trans 1,2-methoxy alcohol 21 (0.026 g, 90% yield),
practically pure.
72.84; H, 7.88. MS (m/z) 77, 79, 91, 109, 121, 157 (Mþ).
5.2.5. Reaction of epoxide 6
(protocol B). Proceeding as previously described for the corre-
sponding reaction of epoxide 6 , the treatment of epoxide 6
a with 0.005 N TsOH in AcOH/CH2Cl2
b
a
(0.072 g, 0.333 mmol) with a CH2Cl2 solution (3.2 mL) containing
AcOH (0.06 mL, 0.999 mmol, 3.0 equiv) and TsOH$H2O (0.003 g,
0.017 mmol, 0.05 equiv) (epoxide/TsOH/AcOH¼1:0.05:3) for 18 h at
room temperature (two additional portions of TsOH$H2O (0.001 g,
0.004 mmol) were added after 3 h and 14 h) afforded a crude re-
action product (0.075 g, 81% yield) consisting of a 38:30:32 mixture
of trans-1,2-hydroxy acetate 28 (anti-1,2-addition product), cis-1,2-
hydroxy acetate 27 (syn-1,2-addition product), and cis-1,4-hydroxy
acetate 29 (syn-1,4-addition product) (1H NMR), which was sub-
jected to preparative TLC (a 9:1 CH2Cl2/(i-Pr)2O) mixture was used
as the eluant. Extraction of the three most intense bands (the faster
moving band contained 29 and the slower moving band contained
27) afforded pure cis-1,2-hydroxy acetate 27 (0.008 g, 8% yield),
trans-1,2-hydroxy acetate 28 (0.018 g, 20% yield), and cis-1,4-hy-
droxy acetate 29 (0.014 g, 15% yield).
5.2.3. Reaction of epoxide 6
(protocol B). Typical procedure. Epoxide 6
b
with 0.01 N TsOH in AcOH/CH2Cl2
(0.150 g, 0.694 mmol)
b
was dissolved in a CH2Cl2 solution (6.8 mL) containing AcOH
(0.12 mL, 2.08 mmol, 3.0 equiv) and TsOH$H2O (0.006 g,
0.034 mmol, 0.05 equiv) (epoxide/TsOH/AcOH¼1:0.05:3) and the
reaction mixture was stirred for 6 h at room temperature, adding
two portions of TsOH$H2O: after 1.5 h (0.005 g, 0.027 mmol) and
3.5 h (0.003 g, 0.016 mmol). The reaction mixture was diluted with
CH2Cl2 and solid NaHCO3 was added. Evaporation of the filtered
organic solution afforded a crude product consisting of (1S
*
,2S
*
,6R )-
*
2-(acetoxy)-6-(benzyloxymethyl)-3-cyclohexen-1-ol (18) (0.199 g,
99% yield), practically pure as a liquid, which was used in the next
step without any further purification: Rf¼0.12 (8:2 hexane/AcOEt);
5.2.5.1. (1R
hexen-1-ol (27). Rf¼0.13 (9:1 CH2Cl2/(i-Pr)2O); FTIR (neat)
1730, 1367, 1234, 1170, 1091 cmꢁ1. 1H NMR (CDCl3)
7.39e7.28 (m,
*
,2S
*
,6R
*
)-6-(Benzyloxymethyl)-2-acetoxy-3-cyclo-
FTIR (neat)
n
3481, 1735, 1368, 1230, 1170, 1089, 972, 935 cmꢁ1. 1H
n
3489,
NMR (CDCl3)
d
7.40e7.28 (m, 5H, Ph), 6.04e5.94 (m, 1H), 5.73e5.62
d
(m,1H), 5.18e5.09 (m,1H), 4.58e4.47 (m, 2H, CH2Ph), 4.06e3.97 (m,
5H, Ph), 5.98e5.87 (m,1H), 5.82e5.73 (m,1H), 5.31 (t,1H, J¼4.5 Hz),
4.55 (s, 2H, CH2Ph), 3.90e3.81 (m, 1H), 3.66e3.59 (m, 2H),
2.37e2.18 (m, 2H), 2.09 (s, 3H, MeC]O), 2.08e2.06 (m, 1H). 13C
1H), 3.72e3.06 (m, 2H), 2.34e2.08 (m, 3H), 2.06 (s, 3H, MeC]O). 13C
NMR (CDCl3) d 170.7, 137.9, 132.4, 128.7, 128.1, 127.9, 122.7, 73.8, 72.7,
70.9, 70.7, 35.1, 24.8, 21.4. Anal. Calcd for C16H20O4: C, 69.55; H, 7.30.
Found: C, 69.34; H, 7.19.
NMR (CDCl3) d 171.3, 138.1, 132.7, 128.9, 127.8, 123.4, 122.7, 73.9,
72.7, 71.6, 69.0, 35.6, 28.8. Anal. Calcd for C16H20O4: C, 69.55; H,
7.30. Found: C, 69.94; H, 7.01.
5.2.4. Reaction of epoxide 6
A). Proceeding as previously described for the corresponding re-
action of epoxide 6 , the reaction of epoxide 6 (0.040 g,
0.185 mmol) with 0.2 N H2SO4/MeOH (1.0 mL) afforded a crude
reaction product (0.040 g, 87% yield) consisting of a 32:44:24
mixture of methoxy alcohols 24, 25, and 26 (1H NMR), which was
subjected to preparative TLC (an 8:2 hexane/AcOEt mixture was
used as the eluant). Extraction of the more intense bands afforded
trans-1,2-methoxy alcohol 24 (anti-1,2-addition product) (0.007 g,
15% yield), cis-1,4-methoxy alcohol 25 (syn-1,4-addition product)
a with 0.2 N H2SO4/MeOH (protocol
5.2.5.2. (1R
1-ol (28). Rf¼0.16 (9:1 CH2Cl2/(i-Pr)2O); FTIR (neat)
1370,1235,1171,1090 cmꢁ1. 1H NMR (CDCl3)
7.39e7.28 (m, 5H, Ph),
*
, 2R
*
,6R
*
)-6-(Benzyloxymethyl)-2-acetoxy-3-cyclohexen-
b
a
n
3475, 1732,
d
5.82e5.72 (m, 1H), 5.54e5.45 (m, 1H), 5.38e5.29 (m, 1H), 4.56 (s,
2H, CH2Ph), 3.81 (dd, 1H, J¼10.6, 7.9 Hz), 3.71e3.56 (m, 3H),
2.26e2.14 (m, 2H), 2.12 (s, 3H, MeC]O), 2.08e1.92 (m, 1H). 13C NMR
(CDCl3) d 171.8, 137.9, 129.2, 128.7, 128.0, 127.7, 125.3, 77.4, 77.0, 73.7,
72.9, 29.8, 28.0. Anal. Calcd for C16H20O4: C, 69.55; H, 7.30. Found: C,
69.83; H, 6.97.

4708
V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
5.2.5.3. (1S
hexen-1-ol (29). Rf¼0.19 (9:1 CH2Cl2/(i-Pr)2O); FTIR (neat)
1732, 1368, 1229, 1170, 1092 cmꢁ1. 1H NMR (CDCl3)
7.39e7.28 (m,
*
,4S
*
,6R
*
)-6-(Benzyloxymethyl)-4-acetoxy-2-cyclo-
4-Methyl alcohol 46 showed 79% ee, whereas the starting epoxide
6b showed 24% ee (determined on the crude reaction mixture). The
n
3485,
d
crude reaction product was subjected to preparative TLC with an
8:2 hexane/AcOEt mixture as the eluant. Extraction of the most
intense band afforded pure trans-1,4-methyl alcohol 46 (0.006 g,
5H, Ph), 5.94 (dd, 1H, J¼10.1, 1.8 Hz), 5.85e5.77 (m, 1H), 5.22e5.15
(m, 1H), 4.57 (s, 2H, CH2Ph), 4.13 (d, 1H, J¼10.1 Hz), 3.64 (dd, 1H,
J¼9.0, 4.4 Hz), 3.52e3.49 (m, 1H), 3.48 (t, 1H, J¼9.0 Hz), 2.24e2.06
(m, 1H), 2.04 (s, 3H, MeC]O), 1.79e1.57 (m, 2H). 13C NMR (CDCl3)
11% yield), [
a
]D25 þ26.9 (c 0.3, CHCl3).
d
171.1, 137.7, 136.4, 128.7, 128.1, 127.9, 124.7, 75.1, 73.8, 71.9, 65.8,
5.3.3. Reaction of epoxide 6b with Me2Zn/Cu(OTf)2/(R,R,R)-phos-
37.2, 29.9, 29.0. Anal. Calcd for C16H20O4: C, 69.55; H, 7.30. Found: C,
69.79; H, 6.88.
phoramidite 44. (R,R,R)-Phosphoramidite 44 (0.004 g, 0.006 mmol,
0.03 equiv) was added to a solution of Cu(OTf)2 (0.0012 g,
0.004 mmol, 0.015 equiv) in anhydrous toluene (0.8 mL) and the
reaction mixture was stirred 20 min at room temperature, then
cooled at ꢁ78 ꢀC and treated in succession with a solution of ep-
5.3. Addition reactions of C-nucleophiles to epoxides 6
and 6
b
a
oxide 6b (0.050 g, 0.231 mmol) in anhydrous toluene (0.4 mL) and
5.3.1. Reaction of epoxide 6
b
with Me2CuLi. Typical procedure. A so-
2.0 M Me2Zn in toluene (0.175 mL, 0.35 mmol, 1.5 equiv). After 1 h
stirring at the same temperature, dilution with Et2O and evapora-
tion of the washed (10% aqueous NH4Cl and saturated aqueous
NaCl) organic solution afforded a crude reaction product (0.050 g,
93% yield) consisting of a 28:72 mixture of trans- 1,2-methyl alco-
hol 45 (anti-1,2-addition product) and trans- 1,4-methyl alcohol 46
(anti-1,4-addition product) (1H NMR, entry 4, Table 3). trans-1,2-
Methyl alcohol 45 showed 98% ee (determined on the crude re-
action mixture), whereas trans-1,4-methyl alcohol 46 showed 45%
ee (determined on the crude reaction mixture). The crude reaction
mixture was subjected to preparative TLC with an 8:2 hexane/
AcOEt mixture, as the eluant. Extraction of the two most intense
bands (the faster moving band contained 45) afforded pure
lution of CuI (0.055 g, 0.291 mmol, 3.0 equiv) in anhydrous Et2O
(2.6 mL) was treated, under argon at ꢁ15 ꢀC, with 1.6 M MeLi
(0.42 mL, 0.677 mmol, 6.0 equiv). After few minutes, a solution of
epoxide 6b (0.021 g, 0.046 mmol) in anhydrous Et2O (0.3 mL) was
added and the resulting reaction mixture was stirred at 0 ꢀC for 1.5 h.
Dilution with Et2O and evaporation of the washed (10% aqueous
NH4Cl and saturated aqueous NaCl) organic solution afforded a crude
reaction product (0.021 g, 94% yield) consisting of a 57:43 mixture of
trans-1,2-methyl alcohol 45 and trans- 1,4-methyl alcohol 46 (1H
NMR, entry 1, Table 3), which was subjected to preparative TLC (an
8:2 hexane/AcOEt mixture was used as the eluant). Extraction of the
more intense bands yielded trans-1,2-methyl alcohol 45 (anti-1,2-
addition product, 0.009 g, 40% yield) and trans- 1,4-methyl alcohol
46 (anti-1,4-addition product, 0.006 g, 27% yield).
(þ)-trans-1,2-methyl alcohol 45 (0.008 g, 15% yield), [
a
]D25 þ86.4 (c
0.36, CHCl3), and trans- 1,4-methyl alcohol 46 (0.028 g, 52% yield).
5.3.1.1. (1R
hexen-1-ol (45). A liquid, Rf¼0.21 (8:2 hexane/AcOEt); FTIR (neat)
3390, 1350, 1225, 1089 cmꢁ1. 1H NMR (CDCl3)
7.40e7.28 (m, 5H,
*
,2S
*
,6R
*
)-6-(Benzyloxymethyl)-2-methyl-3-cyclo-
5.4. Reactions in the gas-phase
n
d
5.4.1. Materials. Oxygen and trimethylamine were high-purity
gases from Matheson Gas Products Inc., deuterium (99.98%) was
purchased from Aldrich and all were used without further purifi-
cation. The purity of starting epoxides 6a and 6b was checked by
analytical gas chromatography on the same columns employed for
the analysis of their gas phase products.
Ph), 5.67e5.56 (m, 1H), 5.55e5.46 (m, 1H), 4.53 (s, 2H, CH2Ph),
3.79e3.64 (m, 2H), 3.59 (dd, 1H, J¼9.1, 4.0 Hz), 3.01 (br s, 1H),
2.17e2.31 (m, 1H), 2.16e1.99 (m, 3H), 1.04 (d, 3H, J¼7.2 Hz, Me). 13C
NMR (CDCl3)
d
138.1, 130.4, 128.7, 127.9, 127.8, 124.8, 74.5, 73.7, 73.4,
37.2, 35.1, 25.6, 19.2. Anal. Calcd for C15H20O2: C, 77.55; H, 8.68.
Found: C, 77.69; H, 8.29.
5.4.2. Procedure. The gaseous mixtures were prepared by in-
troducing fragile ampoules, containing weighed amounts of
selected epoxide (0.0076e0.0080 mmol) and methanol
(0.0231e0.0237 mmol), into 250-mL Pyrex bulbs, equipped with
a break-seal arm, and connected to a greaseless vacuum line. Fol-
lowing the introduction of the gaseous components (deuterium,
oxygen, and trimethylamine) at the desired partial pressures into the
carefully evacuated and outgassed vessels, the latter were then
allowed to come to room temperature, the fragile ampoules broken,
and the gaseous components allowed to mix before being subjected
to the irradiation. The gaseous mixtures were submitted to irradia-
tion at a constant temperature (37.5 ꢀC) in a 60Co 220 Gammacell
5.3.1.2. (1R
hexen-1-ol (46). A liquid, Rf¼0.18 (8:2 hexane/AcOEt); FTIR (neat)
3398, 1347, 1221, 1084 cmꢁ1. 1H NMR (CDCl3)
7.40e7.28 (m, 5H,
*
,4S
*
,6R )-6-(Benzyloxymethyl)-4-methyl-2-cyclo-
*
n
d
Ph), 5.81e5.67 (m, 2H, CH]CH), 4.53 (s, 2H, CH2Ph), 4.27 (br s, 1H,
CHOH), 3.69 (dd, 1H, J¼9.1, 3.0 Hz, CHaHbOBn), 3.52 (dd, 1H, J¼9.1,
5.0 Hz, CHaHbOBn), 2.48e2.34 (m, 1H, CHMe), 2.33e2.12 (m, 2H,
C(6)H and C(5)HaHb), 1.83e1.68 (m, 1H, C(5)HaHb), 0.98 (d, 3H,
J¼7.2 Hz, Me). 13C NMR (CDCl3)
d 138.3, 135.9, 128.6, 127.8, 73.5,
72.0, 66.1, 35.7, 28.8, 28.6, 20.4 (two aromatic signals overlapped).
Anal. Calcd for C15H20O2: C, 77.55; H, 8.68. Found: C, 77.91; H, 8.46.
5.3.2. Reaction of epoxide 6
b
with Me2Zn/Cu(OTf)2/(R,R,R)-phos-
from Nuclear Canada Ltd. (dose: 1.5ꢂ104 Gy; dose rate: 1ꢂ104 Gy hꢁ1
,
phoramidite 44 under kinetic resolution conditions. (R,R,R)-Phos-
phoramidite 44 (0.004 g, 0.007 mmol, 0.03 equiv) was added, at
room temperature, to a solution of Cu(OTf)2 (0.6 mg, 0.0035 mmol,
0.015 equiv) in anhydrous toluene (0.4 mL) and the resulting re-
action mixture was stirred 20 min at the same temperature, then
determined with a Fricke dosimeter). Control experiments, carried
out at doses ranging from 1ꢂ104 to 1ꢂ105 Gy, showed that the rel-
ative yields of products are largely independent of the dose. In order
to verify the ‘stability’ of the reactions products, the methoxy alco-
hols 21e26 were placed with the gaseous members (D2, O2, and
NMe3) into Pyrex bulbs and irradiated at the same experimental
cooled at ꢁ78 ꢀC and treated with a solution of epoxide 6
b (0.050 g,
0.231 mmol) in anhydrous toluene (0.4 mL) and subsequently with
2.0 M Me2Zn in toluene (0.069 mL, 0.139 mmol, 0.6 equiv). After
35 min, HPLC analysis showed 26% conversion. The reaction mix-
ture was diluted with Et2O: evaporation of the washed (10%
aqueous NH4Cl and saturated aqueous NaCl) organic solution
afforded a crude reaction product (0.050 g) consisting of a mixture
of trans-1,4-methyl alcohol 46, as the only addition product (entry
conditions adopted for epoxides 6a and 6b
(37.5 ꢀCddose:
1.5ꢂ104 Gy). In all the cases the methoxy alcohols were recovered
unchanged and no trace of isomerization products was found.
5.4.3. Product analysis. The analysis of the products was performed
by injecting measured portions of the homogeneous reaction mix-
ture into a HewlettePackard 5890 series II gas chromatograph,
equipped with a flame ionization detection unit. In order to prevent
3, Table 3), and starting, unreacted, epoxide 6b
(1H NMR). trans-1,

V. Di Bussolo et al. / Tetrahedron 67 (2011) 4696e4709
4709
which affect, to a variable extent, the corresponding processes carried out in
solution. Accordingly, this information can often be useful for the ration-
alization of the regio- and stereoselective behavior of the same epoxides in
opening reactions carried out in the condensed phase. See, as an example:
Crotti, P.; Di Bussolo, V.; Macchia, F.; Favero, L.; Pineschi, M.; Lucarelli, L.;
Roselli, G.; Renzi, G. J. Phys. Org. Chem. 2005, 18, 321e328 and references
therein.
selective loss of the reaction products by adsorption on the glass of
the reaction bulb (and to obtain reproducible and meaningful re-
action yields), the analysis was repeated after careful washing of the
bulb walls with anhydrous ether. Satisfactory agreement between
the results of the gaseous mixture and the ether solution analysis
was found in all runs. The products were identified by comparison of
their retention volumes with those of authentic standard com-
pounds on the following columns: (i) a 30 m long, 0.25 mm i.d.
HP5MSÔ fused silica capillary column, operating at temperatures
ranging from 100 to 210 ꢀC, 5 ꢀC minꢁ1; (ii) a 30 m long, 0.32 mm i.d.
Supelcowax 10Ô fused silica capillary column, operating at 190 ꢀC.
The identity of the products was further confirmed by GLC/MS, using
a HewlettePackard 5890A gas chromatograph in line with an HP
5971A quadrupole mass spectrometer. The yields of the products
were measured, using the internal standard method and individual
calibration factors to correct for the detector response. The results
given in Tables 1 and 2 (see alsoTable 1, Supplementary data) are the
average of at least three measurements taken on at least two dif-
ferent runs for each point.
13. The methanolysis of epoxide 6
b
in the gas-phase was carried out in a D2 at-
mosphere containing epoxide 6
b and only a small excess of MeOH (3 equiv) in
the presence of Me3N. The irradiation of the reaction mixture with a 60Co 220
Gammacell generates the gaseous Bronsted acid D3þ, the necessary gaseous
€
catalyst of the reaction. The crude reaction product was examined by GC/MS
(see Experimental section).
14. The stereoselectivity toward trans-1,4-methoxyalcohol 22 is similar tothat found
in the small amount (20%) of 1,4-addition products obtained from the corre-
sponding reaction carried out in the condensed phase (0.2 N H2SO4/MeOH).
15. (a) These results would indicate that in epoxide 6b, the C(2) oxirane carbon is
largely the more reactive center of the allyl system, to the point that in the
presence of a low amount of O-nucleophile, it becomes the only reactive one.
(b) The result obtained by AcOH/TsOH under protocol B reaction conditions
(entry 8, Table 1) turned out to be particularly useful for the synthesis of ep-
oxide 6
16. The 96:4
a
(see Scheme 7).
equilibrium given for epoxide 6
a
is the average of the
6α’
6α”
values obtained by means of MP2 and B3LYP calculations when CHCl3 is
considered as the solvent (Supporting data). This value is in accordance with
the corresponding result obtained by 1H NMR conformational analysis. A
similar result was previously obtained by a theoretical conformational anal-
ysis carried out on the corresponding 6-methyl substituted epoxides, see:
Crotti, P.; Di Bussolo, V.; Pomelli, C. S.; Favero, L. Theor. Chem. Account 2009,
122, 245e256.
Acknowledgements
ꢀ
This work was supported by the Universita di Pisa and MIUR
17. (a) The possibility that a corresponding syn-1,2-addition process of epoxide 6
can occur by an equatorial opening of the oxirane ring through the less stable
conformer 6
00-H cannot reasonably be ruled out (Scheme 13). However, it was
a
ꢀ
(Ministero dell’Istruzione, della Universita e della Ricerca), Roma.
P.C. gratefully acknowledges Merck Research Laboratories for the
generous financial support deriving from the 2005 ADP Chemistry
Award. Professor Renato Noto of the University of Palermo is kindly
acknowledged for helpful discussions.
a
demonstrated that, in nucleophilic opening reactions, the ‘axial cleavage’ of 1,2-
epoxycyclohexanes is easier than the corresponding ‘equatorial cleavage’ and
favours the syn stereoselectivity of the addition process. See: Battistini, C.;
Crotti, P.; Damiani, D.; Macchia, F. J. Org. Chem. 1979, 44, 1643e1647 and ref-
erences therein. (b) As previously demonstrated in the case of 2-aryl-oxiranes,
we think that the syn-1,2-addition process leading to cis-2-acetoxy alcohol 27
derives from a nucleophile-separated ionedipole pair, such as 34, followed by
nucleophilic pseudoequatorial attack on the partially developed C(2) carboca-
tion with retention of configuration (Scheme 13). See Ref. 12 and Crotti, P.;
Dell’Omodarme, G.; Ferretti, M.; Macchia, F. J. Am. Chem. Soc. 1987, 109,
1463e1469.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tet.2011.04.036.
18. No corresponding anti-1,2-addition products were ever observed in any case
under these conditions. Under acid conditions (0.2 N H2SO4/MeOH/toluene),
References and notes
epoxides 1a and 1b lead to mixtures of the corresponding methyl a- and b-O-
glycosides,2 which are rapidly transformed into the furane derivative 1-F
ꢁ
ꢁ
1. (a) Arjona, O.; Gomez, A. M.; Lopez, J. C.; Plumet, J. Chem. Rev. 2007, 107,
1919e2036; (b) Marco-Contelles, J.; Molina, M. T.; Anjum, S. Chem. Rev. 2004,
104, 2857e2900; (c) Cumpstey, I. Carbohydr. Res. 2009, 344, 2285e2310.
2. (a) Di Bussolo, V.; Caselli, M.; Romano, M. R.; Pineschi, M.; Crotti, P. J. Org. Chem.
2004, 69, 7383e7386; (b) Di Bussolo, V.; Caselli, M.; Romano, M. R.; Pineschi,
M.; Crotti, P. J. Org. Chem. 2004, 69, 8702e8708. For the corresponding results
(unpublished results from our laboratory; for the sake of simplicity, only epoxide
1b
is shown). For the formation of 1-F, see: Zamojski, A.; Chmielewski, M.;
Konowal, A. Tetrahedron 1970, 26, 183e189.
OH
O
OMe
O
BnO
HO
BnO
from 6-deoxy derivatives of 6a and 6b, see: Di Bussolo, V.; Favero, L.; Romano,
O
HO
0.2 N H2SO4/MeOH
toluene
M. R.; Pineschi, M.; Crotti, P. Tetrahedron 2008, 64, 8188e8201.
3. Di Bussolo, V.; Romano, M. R.; Pineschi, M.; Crotti, P. Tetrahedron 2007, 63,
2482e2489 and references therein.
4. Di Bussolo, V.; Checchia, L.; Romano, M. R.; Pineschi, M.; Crotti, P. Org. Lett.
2008, 10, 2493e2496. Corresponding 4-amino-substituted and ‘mixed’ oligo-
saccharides were also synthesized by reiterative application of the same gly-
O
methyl
1-F
1β
α- and -β-O-glycosides
19. The 1,4-regioselectivity (SN20 process) observed in the acid methanolysis
cosylation process to aziridine 2
5. Unpublished results from our laboratory.
6. Epoxides 6 and 6 were prepared as corresponding racemic mixtures, con-
a.
of epoxides 6a and 6b under gas-phase operating conditions is in accordance
with the slightly preferred SN20 pathway, previously observed in the
gas-phase-acid-induced methanolysis of allyl alcohols. See: Renzi, G.; Lom-
bardozzi, A.; Dezi, E.; Pizzabiocca, A.; Speranza, M. Chem. Eur. J. 1996, 2,
316e322.
a
b
sidering that it was not necessary to have them as pure enantiomers for these
preliminary studies on their reactivity.
7. (a) Ogawa, S.; Nishi, K.; Shibata, Y. Carbohydr. Res. 1990, 206, 352e360; (b)
Tsunoda, H.; Ogawa, S. Liebigs Ann. 1995, 267e277.
20. The formation of ketone 39 is due to the presence of Sc(OTf)3. Actually, even
after long reaction times (3 days) in the presence of TMSN3, epoxide 6a could
be recovered completely unreacted. Only when Sc(OTf)3 was added, was a fast
formation of ketone 39 observed.
8. (a) Ogawa, S.; Kasahara, I.; Suami, T. Bull. Chem. Soc. Jpn. 1979, 52, 118e123; (b)
Ogawa, S.; Toyokuni, T.; Kondoh, T.; Hattori, Y.; Iwasaki, S.; Suetsugu, M.; Suami,
T. Bull. Chem. Soc. Jpn. 1981, 54, 2739e2746; (c) Kotsuki, H.; Asao, K.; Ohnishi, H.
Bull. Chem. Soc. Jpn. 1984, 57, 3339e3340.
9. Balsamo, A.; Crotti, P.; Macchia, B.; Macchia, F. Tetrahedron 1973, 29, 2183e2188.
10. Evidently, the corresponding allyl trans-hydroxy acetate 18, the primary re-
action product, is hydrolyzed in the reaction mixture to trans diol 13.
11. It is interesting to note that when the same opening protocol is applied to the
corresponding hydrolysis (H2O, 3 equiv, in THF) and methanolysis reaction
(MeOH, 3 equiv, in CH2Cl2) in the presence of TsOH, the reactions are not re-
gioselective, affording a 7:3 (hydrolysis) and 8:2 mixture (methanolysis) of the
corresponding anti-1,2-addition product and 1,4-addition products (vide infra
and see Table 1).
21. Crotti, P.; Di Bussolo, V.; Bertolini, F.; Pineschi, M. Curr. Org. Synth. 2009, 6,
290e324.
22. Marino, J. P.; Floyd, D. M. Tetrahedron Lett. 1979, 20, 675e678.
23. Actually, epoxide 6a .
can react also by means of the less stable conformer 6a00
However, in the discussion, the attention was directed toward the more stable
conformer 6a0 because only in this conformer a more favorable axial opening of
the oxirane ring can occur (see Ref. 17a).
24. (a) Yoshikai, N.; Zhang, S.-L.; Nakamura, E. J. Am. Chem. Soc. 2008, 130,
12862e12863; (b) Langlois, J.-B.; Alexakis, A. Adv. Synth. Catal. 2010, 352,
447e457.
25. Bertozzi, F.; Crotti, P.; Macchia, F.; Pineschi, M.; Feringa, B. L. Angew. Chem., Int.
Ed. 2001, 40, 930e932. For a review on this topic, see also: Pineschi, M. New J.
Chem. 2004, 28, 657e665.
12. Obviously, the opening reactions of epoxides carried out in the gas-phase have
no synthetic valence. They can only give information about the intrinsic re-
activity of epoxides unperturbed by the effects of solvation and ion-pairing