On the coupling reaction of lithium azaenolates of chiral oxazolines with carbonyl compounds

Article abstract of DOI:10.1016/S0040-4020(01)00611-1

Optically pure chloromethyloxazolines have been prepared, lithiated and then reacted with benzophenone to furnish stereoselectively oxazolinyl oxiranes and formyl oxiranes. The stereoselection observed has been related to the geometry of the intermediate lithium azaenolates. Calculations have been carried out in order to evaluate the relative stability of the E and Z forms of the lithium azaenolates. The reaction of a lithium azaenolate with other ketones leading to oxazolinyl epoxides is also reported.

Full text of DOI:10.1016/S0040-4020(01)00611-1

TETRAHEDRON
Pergamon
Tetrahedron 57 42001) 6775±6786
On the coupling reaction of lithium azaenolates of chiral
oxazolines with carbonyl compounds
Saverio Florio,a,p Vito Capriati,a Renzo Luisi,a Alessandro Abbottob and Daniel J. Pippelc
a
Á
Dipartimento Farmaco-Chimico, Centro C.N.R. `M.I.S.O.', Universita di Bari, Via E. Orabona 4, I-70126 Bari, Italy
Á
b
Dipartimento di Scienza dei Materiali, Universita degli Studi di MilanoÐBicocca, Via Cozzi 53, I-20125 Milano, Italy
cDepartment of Chemistry, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA
Received 12 January 2001; revised 4 May 2001; accepted 31 May 2001
AbstractÐOptically pure chloromethyloxazolines have been prepared, lithiated and then reacted with benzophenone to furnish stereo-
selectively oxazolinyl oxiranes and formyl oxiranes. The stereoselection observed has been related to the geometry of the intermediate
lithium azaenolates. Calculations have been carried out in order to evaluate the relative stability of the E and Z forms of the lithium
azaenolates. The reaction of a lithium azaenolate with other ketones leading to oxazolinyl epoxides is also reported. q 2001 Elsevier
Science Ltd. All rights reserved.
1. Introduction
lithiated 2-chloromethyloxazoline with carbonyl com-
pounds, conversion of the resulting chlorohydrins into the
corresponding oxiranes and deblocking of the oxazoline
moiety.1
We have recently reported an oxazoline-mediated synthesis
of formyl oxiranes which was based on the reaction of
Scheme 1. Reagents and conditions: 4i) t-BuOCl, CCl4; 4ii) LDA, 2788C; 4iii) Ph2CO; 4iv) H1; 4v) NaOH/i-PrOH.
Keywords: lithium and compounds; enolates; oxazolines; epoxides; computer-assisted methods.
Corresponding author. Tel.: 139-805-442749; fax: 139-805-442251; e-mail: ¯orio@farmchim.uniba.it
p
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020401)00 611-1

6776
S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
2. Results and discussion
2.1. Synthesis
Chlorination9 of 2,4-dimethyl-5-phenyl-2-oxazoline 1a10
4Scheme 1) with t-butyl hypochlorite11 in CCl4 at room
temperature afforded oxazoline 2 which was puri®ed by
¯ash chromatography.12 Treatment of 2 4Scheme 1) with
lithium diisopropylamide 4LDA) at 2788C in THF
furnished lithiated derivative 3, which was stable at low
temperature and could be kept there for at least 1 h. The
reaction of 3 with benzophenone provided diastereomeric
chlorohydrins 4a and 4b 474% yield; dr 4a/4bˆ1:1.6) which
could be separated by column chromatography 4Scheme 1).
The con®guration of 4a was proved to be R at Ca on the
basis of the X-ray analysis. Therefore, that of 4b at the same
carbon should be S. Chlorohydrins 4a and 4b, which
presumably exist intramolecularly hydrogen-bond asso-
ciated13,14 4Chart 1), could quantitatively and stereo-
speci®cally be converted into the oxazolinyl oxiranes 5a
and 5b, respectively, upon treatment with NaOH. The
oxazolinyl ring of 5a and 5b was then elaborated according
to a known protocol1 to give highly enantiomerically pure
formyl epoxides 6a 472% yield; S, ee%ˆ91) and 6b 493%
yield; R, ee%.99), respectively 4Scheme 1).
Scheme 2.
The poor stereoselection of the reaction of 3 with benzo-
phenone, which leads at the end to 6, might be explained by
considering that the stereoselective determining step is the
lithium azaenolate formation 4deprotonation of the dia-
stereotopic hydrogens at the chlorine bearing carbon atom
to give the geometric isomers 4E)-3 and 4Z)-3). These
isomers, which are supposed not to interconvert,15 should
be of quite different energy on the basis of semiempirical
calculations16 4see ahead) which indicate that the E isomer is
by far 4about 25 kcal mol21) more stable than the corre-
sponding Z isomer, due to a favorable internal coordination
of lithium between the chlorine and nitrogen atoms. In view
As oxazolinyl and formyl oxiranes are useful intermediates
for the preparation of a large variety of compounds,2±7 we
decided to extend the above synthetic strategy to the prepa-
ration of optically active oxazolinyl and formyl oxiranes
starting from chiral oxazolines.8 We performed semi-
empirical and ab initio calculations on the carbanionic
species in order to rationalize the stereoselective paths.
Both the synthetic and calculation results are presented in
this paper.
Table 1. Reaction of lithium azaenolate 9 with Ph2CO in various solvents
Solvent
Reaction time 4min)
Epoxides 11 4% yield)a
Formyl epoxides 6 4% yield)a
ee 4%)b
Prevalent con®guration of 6
THF
15
15
15
15
15
74
69
65
58
84
21
88
83
83
83
83
83
37
21
12
21
18
R
R
R
R
R
R
R
TBME
Toluene
Hexane
DME
Et2O
3070
15
83
THP
83
25
a
Isolated yields for each individual step. Quantitative conversion from chlorohydrins 10.
ee refers to formyl epoxides 6; determined by GC on a chiral stationary phase 4see Section 4).
b

S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
Table 2. Reaction of 13 with Ph2CO in various solvents
6777
Solvent
Reaction time 4min)
Epoxides 15 4% yield)a
Formyl epoxides 6 4% yield)a
ee 4%)b
Prevalent con®guration of 6
THF
Et2O
THP
Hexane
DME
Toluene
3078
3093
4088
3078
4024
3084
3730
3700
2740
1750
3710
370
S
S
S
S
R
R
a
Isolated yields for each individual step. Quantitative conversion from chlorohydrins 14.
ee refers to formyl epoxides 6; determined by GC on a chiral stationary phase 4see Section 4).
b
of such a computed large energy difference, we presume that
diastereomeric chlorohydrins 10a and 10b 487% yield; dr
10a/10bˆ1:2.2) that could be separated by column chroma-
tography and assigned the con®guration by IR, NMR and
con®rmed by an X-ray analysis carried out on 10a. On this
basis it was assigned the R con®guration at the a-carbon.
Consequently, diastereomer 10b was assigned the S con®g-
uration at the same a-carbon. As in the case of 3, the observed
poor diastereoselection of the reaction of 9 with benzophenone
to give chlorohydrins 10a and 10b could be rationalized by
assuming that the predominant 4E)-9 isomer is attacked by the
carbonyl compound prevalently on the si face 4opposite to the
isopropyl group) for steric reasons 4Scheme 2). Isomers 10a
and 10b were, then, treated with NaOH to give, stereo-
speci®cally, epoxides 11a and 11b, having the S and the R
con®guration, respectively, at the a-carbon.
the E species predominates by far in the solution.17 The poor
diastereoselection of the reaction of 3 with benzophenone to
give chlorohydrins 4a and 4b, therefore, could be explained
by assuming that the attack of the carbonyl compound to
4E)-3, which has a planar arrangement, takes place on either
the re and si faces with a slight predominance of the si face
due to some steric hindrance exerted by the substituents on
the oxazoline ring. The Ph and Me substituents, however,
Ê
are rather far 4.4.5 A) from the carbanionic carbon 4nucleo-
philic site) so that they are not expected to signi®cantly
affect the approach of the carbonyl compound) to give
chlorohydrin 4b 4S,R,S con®guration) as the major product
4Scheme 2).
Lithiated oxazoline 9 4Scheme 1), prepared by lithiation of
oxazoline 8,18 reacted cleanly with benzophenone affording
The stereoselection observed in the reactions of lithiated
Table 3. Reaction of lithium azaenolate 13 with ketones
Compound
15
Methoda
Oxazolinyl epoxideb 4% yield)
drc
RˆPh±
A
B
A
B
A
B
A
B
A
B
78
77
7046/54
4045/55
85
5047/53
73
78
9066/33
78
33/67
45/55
Rˆp-Cl±C6H4±
R,Rˆ±4CH2)5±
17a
17b
44/56
R,Rˆ±4CH2)11
±
17c
56/44
38/62
R,Rˆ2-Adamantylidene
17d
±
a
Method A: A THF solution of 2-4chloromethyl)oxazoline 41 mmol) and ketone 41.3 mmol) was added dropwise at 2908C to a LDA solution 41.3 mmol)
4Barbier's technique); the reaction mixture was quenched with sat. aq. NH4Cl after 30min. Method B: A THF solution of 2-4chloromethyl)oxazoline
41 mmol) was added dropwise at 2908C to a LDA solution 41.3 mmol); the resulting red solution was stirred for 45 min at 2908C and then a THF solution of
ketone 41.3 mmol) was added dropwise. The reaction mixture was ®nally quenched with sat. aq. NH4Cl after 30min.
b
c
The crude mixture of chlorohydrins was quantitatively converted to oxazolinyl epoxides whose isolated yields were reported.
Diastereomeric ratio from H NMR.
1

6778
S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
Table 4. PM3 and DFT Relative energies of isolated lithium azaenolates 3,
9 and 13 and solvated lithium azaenolate 9
Species
PM3
DFT
Planara 4E)-3
0.0
124.9
125.0
0.0
123.2
0.0
126.3
124.7
0
123.2
121.7
Bridgedb 4Z)-3
Planara 4Z)-3
Planara 4E)-13c
4Z)-13
Planara 4E)-9
0.0
16.9
116.4
.0
Bridgedb 4Z)-9
Planara 4Z)-9
Planara 4E)-9´2Me2O
Bridgedb 4Z)-9´2Me2O
Planara 4Z)-9´2Me2O
Single-point B3LYP/6-311Gp/6-31Gp//PM3 calculations, with the
additional keywords 5D and SCFˆTIGHT. Energies in kcal mol21
.
a
The lithium cation is approximately placed in the plane N±C2±Ca±Cl.
The lithium cation is placed out of the plane N±C2±Ca±Cl, on the other
side with respect to substituents in the 4- and 5-position 4Li±N±C2±C
dihedral angles 227.68).
b
c
The case of lithium azaenolate 13 is handled separately due to the
presence of additional intramolecular interactions 4see text and Table 6).
oxazoline 9 with Ph2CO decreased to some extent on going
from ethereal to hydrocarbon solvents which may be due to
different degrees of aggregation of the lithiated species
4Table 1).
Oxazoline 12 4Scheme 1) has been prepared by chlorination
of commercially available oxazoline 1c with t-butyl hypo-
chlorite according to a procedure similar to that followed for
oxazoline 2.19 Compound 12 was substantially pure and did
not need further puri®cation for the metalation reaction 4see
below). In any case, attempted puri®cation by column
chromatography on silica gel caused ring cleavage to give
4quantitative conversion) the hydroxy amide 7b 4one stereo-
1
isomer), whose structure was established by IR, H/13C
NMR and GC-MS analysis 4Chart 1).
Metalation of 12 4Scheme 1) gave a dark red solution of
lithiated oxazoline 13, that was trapped with benzophenone
affording an inseparable mixture of chlorohydrins 14a and
14b 4dr 14a/14bˆ2:1 by 1H NMR). Here again the coupling
reaction of 13 with benzophenone occurred with poor
diastereoselection. The mixture of 14a and 14b was then
reacted with NaOH in i-PrOH to give diastereomeric oxazo-
linyl oxiranes 15a and 15b that could not be separated. On
standing on silica gel for long time oxiranes 15a and 15b
underwent ring cleavage of the oxazolinyl moiety to furnish
oxiranyl hydroxy amides 16a and 16b 4Chart 1) which were
separated by crystallization 4hexane). Their structures were
established by IR, NMR and GC-MS analysis. The con®g-
uration at the oxiranyl carbon of 16a was proved to be R by
X-ray analysis.
Figure 1. Planar lithium azaenolates 4E)-3, 4E)-13a, and lithium azaenolate
4E)-13b showing a MeO±Li coordination 4dihedral angle Li±N±C42)±
C4a)ˆ1258).
that the reaction of 13 with benzophenone, ending up with
the formation of formyl oxiranes 6, proceeded in a stereo-
selective manner 4low) but opposite to that observed in the
reactions of 3 and 9.
The elaboration of the oxazoline moiety of 15a and 15b
provided enantiomerically enriched formyl epoxide 6a
4ee%ˆ33) 4see Section 4 and Table 2). Comparable values
of ee were obtained in solvents such as Et2O and tetra-
hydropyran 4THP). Lower ee values 415%) were observed
in hexane. There was no enantioselection in toluene. In
DME the reaction led, instead, to 6b 4ee%ˆ31) 4see Table
2). Unfortunately, at present we do not have a plausible
explanation for such changing ee values. It is worth noting
A comparable diastereoselectivity was observed in the
reaction of lithium azaenolate 13 with other ketones in
THF 4Table 3). The kind of stereoselection was not time
dependent. Indeed, the same diastereomeric ratio of the
related chlorohydrins was observed when the lithium
azaenolate 13 was reacted with ketones soon after its

S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
Table 5. Solvation energies of lithium azaenolate 9
6779
a
Species
Solvation energy 4kcal mol21
)
Planar 4E)-9
Bridged 4Z)-9
Planar 4Z)-9
216.7
219.7
219.6
PM3.
a
Values refer to the process of solvation of the free lithium azaenolate
species with two molecules of Me2O.
which, however, are rather far from the reactive center to
exert a signi®cant steric effect.
Semiempirical calculations were con®rmed both consider-
ing the role of the solvent and using ab initio energy compu-
tations, as we did for the case of azaenolate 9 4Tables 4
and 5). With reference to lithiated oxazoline 9, PM3 E±Z
relative energies ranged between 24.7 and 26.3 kcal mol21
while the DFT values were computed lower but in the same
direction. The computed higher stability of the E isomer,
with respect to the Z species, was veri®ed in the presence of
the solvent effect.27 Conclusions of computations performed
in absence of solvent are essentially con®rmed, with the E
lithium azaenolate being by far the most stable. As in the
case of lithiated oxazoline 3, the poor stereoselection of the
coupling reaction of 9 with benzophenone has to be put in
relation with the geometry of the lithium azaenolate 4mainly
4E)-9) and the feeble steric effect of the isopropyl group.
PM3 optimizations of the E isomer of lithiated oxazoline 13
resulted in two energy minimum geometries, 4E)-13a and
4E)-13b 4Fig. 1). Both geometries present internal coordi-
nation of lithium between the chlorine and nitrogen atoms,
as already obtained for 4E)-3 and 4E)-9. In the most stable
4PM3) 4E)-13a isomer, the lithium ion is preferably placed
in the plane of the p-system and no additional internal
coordination involving the methoxy group is observed.
The second computed geometry 4E)-13b, less stable by
3.6 kcal mol21, shows MeO±Li coordination. Electron
density calculations con®rm the occurrence of the internal
coordination MeO±Li.28 In this case the lithium ion is
located in the same face occupied by the methoxy group;
actually, a preliminary coordination with the carbonyl group
of the electrophile could lead to some degree of selectivity,
with preferential formation of chlorohydrin 4R)-14a. Since
the semiempirical approach could not be appropriate to
describe adequately the weak intramolecular binding
energies, we decided to calculate energies at the DFT
level 4Table 6). The high-level ab initio computations now
predict that the intramolecular coordinated 4MeO±Li)
isomer 4E)-13b is more stable29 than the 4E)-13a isomer
Figure 2. Planar and bridged lithium azaenolate 4Z)-3.
generation 4Barbier conditions) or after 45 min from its
generation.
2.2. Computational investigations
In an effort to provide an explanation for the modest dia-
stereoselection of the reaction of lithium azaenolates 3, 9
and 13 with benzophenone we decided to do some calcu-
lations.20±25 Table 4 collects PM3 relative energies of the
geometry optimized lithium azaenolates 4E)-3 and 4Z)-3.
PM3 geometry optimization of 4E)-3 shows an almost
planar arrangement along the sequence Li±N±C2±C±Cl
4Fig. 1). In contrast, we found two energy minimum geome-
tries for the corresponding Z isomer: a planar and a bridged
arrangement 4see Fig. 2).26 Semiempirical calculations
indicate that the E isomer is more stable than the corre-
sponding Z isomer. PM3 E±Z relative energies range around
25 kcal mol21. Such a higher stability of the E species has to
be ascribed to a favorable internal coordination of lithium
between the chlorine and the nitrogen atoms. Therefore,
calculations do support the hypothesis made above accord-
ing to which the modest stereoselection of the reaction of 3
with benzophenone is to be ascribed to the fact that the more
abundant geometric isomer 4E)-3, having a planar arrange-
ment with lithium chelated between the nitrogen and
chlorine atoms, can be attacked by the ketone on both
faces 4re and si faces) with only a slight preference of that
opposite to the phenyl and the methyl groups 4si face),
4no MeO±Li coordination) by about 1.2 kcal mol21
.
Table 6. PM3 and DFT Relative energies of 4E)-13 isomers
Compound
PM3
DFT
4E)-13a
4E)-13b
0.0
13.6
11.2
0.0
Single-point B3LYP/6-311Gp/6-31Gp//PM3 calculations, with the
additional keywords 5D and SCFˆTIGHT. In order to properly compute
the non bonding interactions, diffuse functions were included for Li, N and
O atoms 46-311Gp). Energies in kcal mol21
.

6780
S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
The 4Z)-13 isomer, in which the lithium coordination by the
chlorine atom is not feasible, presents a favored MeO±Li
interaction, with the lithium ion placed out of the plane. This
geometry would be the favored one in the absence of the
chlorine atom in the a position. Therefore, should the 4Z)-13
isomer be present in solution, the attack of the carbonyl
compound would occur on the face opposite to the phenyl
group via a preliminary coordination between the oxygen of
the carbonyl and lithium. In principle, a pronounced stereo-
selection in favor of the chlorohydrin 4R)-14a should be
expected and this was not the case.
suitable for a high stereoselection. This is under way in
our lab and results will be reported in due course.
4. Experimental
4.1. General
Tetrahydrofuran 4THF) and diethyl ether 4Et2O) were
freshly distilled under a nitrogen atmosphere over sodium
benzophenone ketyl. Dry tetrahydropyran 4THP), hexane,
1,2-dimethoxyethane 4DME), toluene, tert-butyl methyl
ether 4TBME), CCl4, 41R,2S)-42)-norephedrine, l-2-
amino-3-methyl-1-butanol, ethyl acetimidate hydrochlo-
ride, 44S,5S)-42)-4-methoxymethyl-2-methyl-5-phenyl-2-
oxazoline and all other chemicals were of commercial
grade 4Aldrich) and used without further puri®cation. tert-
Butyl hypochlorite was prepared according to the litera-
ture.11 TMEDA and diisopropylamine were distilled over
®nely powdered calcium hydride. Petroleum ether refers
to the 40±608C boiling fraction. Commercial solution of
n-BuLi 4in hexanes) from Aldrich was titrated by using
On the basis of such indications, in order to explain the low
stereoselection of the conversion of 12 into 6 we assume that
the predominant isomer 4E)-13 is present in solution as a
mixture of the two forms 4E)-13a and 4E)-13b. The carbonyl
compound would attack the 4E)-13a isomer either from the
top or from the bottom thus showing no or very poor stereo-
selection. The 4E)-13b isomer, instead, would be attacked
by the carbonyl compound preferentially on its re face via a
transition state in which the lithium is coordinated by the
methoxy group and the oxygen of the carbonyl 4Scheme 2).
Therefore, the observed preferential formation of the
2R,40S,50S chlorohydrin 4leading to the 2S,40S,50S oxazo-
linyl oxirane and then to the S formyl oxirane) could be
rationalized by assuming the contribution to the stereo-
chemistry of the coupling reaction of both the isomers
4E)-13a and 4E)-13b.
1
N-pivaloyl-o-toluidine prior to use.31 For the H- and 13C
NMR spectra 41H NMR: 90, 200, 300, 500 MHz, 13C NMR:
50.3, 125 MHz), CDCl3 was used as solvent. GC-MS spec-
trometry analyses were performed on a gas chromatograph
HP 5995C 4dimethylsilicon capillary column, 30m,
0.25 mm ID) equipped with a mass selective detector
operating at 70eV 4EI). Melting points were uncorrected.
TLC was performed on Merck silica gel plates with F-254
indicator; visualization was accomplished by UV light
4254 nm). Column chromatography was performed by
using silica gel 470±230 mesh) with petroleum ether/diethyl
ether 4or AcOEt) mixture as the eluent. All reactions
involving air-sensitive reagents were performed under
nitrogen in oven-dried glassware using syringe-septum cap
technique.
Therefore, the semiempirical and DFT investigation have
pointed out that the stereochemistry of the coupling reaction
of lithiated oxazoline 13 4by far E con®gurated) is only in
some extent determined by the steric effect of the C5 phenyl
group and the coordinating ability of the C4 methoxy group.
3. Conclusion
In conclusion, this paper reports some so far undescribed
aspects of the oxazolines chemistry. Indeed, while the
chemistry of 2-alkyloxazolines has been deeply investi-
gated, that of the halogenoalkyl derivatives has not. Our
work has demonstrated that of the two geometrical forms
of lithium azaenolates of chiral chloroalkyloxazolines, those
adopting the E con®gurations are by far more stable than the
Z ones due to the intramolecular chelation of lithium cation
between the chlorine and nitrogen atoms. This sounds inter-
esting if one considers that in the lithiation reaction of
2-alkyloxazolines the two kinetic azaenolates form in
approximately 95:5 ratio in favor of the Z isomer.17 In the
coupling reactions of 3, 9 and 13 with carbonyl compounds,
steric and coordination effect of substituents at the positions
4 and 5 of the oxazoline ring is only modest, mainly due to
their distance from the carbanionic center 4the a-carbon).
With substituents capable only of some steric effect the
attack of the carbonyl compound to the azaenolate prefer-
ably occurs from the opposite side 4si face), while with
coordinating substituents from the same side 4re face).
Although the reactions of the above azaenolates with
carbonyl compounds are not highly stereoselective, chloro-
hydrins can be separated and stereospeci®cally transformed
into optically active formyl oxiranes. More work is worth
doing in order to establish which metal azaenolates30 are
4.2. Enantiomeric purity analysis
Enantiomeric purity assays were carried out with both
racemic and optically active substrates on a GC equipped
with a permethylated b-cyclodextrin 4b-DEX 120, Supelco)
430m £0.25 mm ID£0.25 mm ®lm thickness) fused silica
capillary column as stationary phase. Det.: FID, 3008C.
Column head pressure: 20psi. Oven: 150 8C. Sample:
1 mL, 500 mg mL21 each in methylene chloride. 42S)-46a):
tRˆ83.39 min; 42R)-46b): tRˆ86.61 min.
4.2.1. Preparation of 04S,5R)-02)-2,4-dimethyl-5-phenyl-
2-oxazoline 01a). To a stirred milky solution of ethyl
acetimidate hydrochloride 42.17 g, 17.60mmol) in dry
CCl4 415 mL), at 08C, a solution of 41R,2S)-42)-nore-
phedrine 42.00 g, 13.23 mmol) in dry CCl4 412 mL) was
added dropwise. The resulting mixture was stirred overnight
at room temperature and poured into 50mL of 2% NaHCO 3.
The organic layer was separated and the aqueous layer
extracted with CH2Cl2 43£20mL). The combined organic
extracts were dried 4Na2SO4) and evaporated in vacuo. The
crude product was puri®ed by ¯ash chromatography
4AcOEt) to give 1.97 g 485% yield) of a colorless oil.
[a]Dˆ2246 4c 1 CHCl3) [lit.,10b [a]Dˆ2247 4c 3.43
CHCl3)]. 1H and 13C NMR spectroscopical data are similar

S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
6781
to those reported in Ref. 10b. GC-MS 470 eV); m/z 4%): 175
44.2, M1), 105 46.9), 77 411.0), 69 4100.0), 68 427.1), 42
443.5). FTIR 4®lm, cm21): 3033, 2975, 1678 4CvN), 1385,
1226, 979, 700.
1193, 1091, 1000, 958, 751, 704, 638. Anal. Calcd for
C24H22ClNO2: C 73.56, H 5.66, N 3.57. Found C 73.36, H
5.77, N 3.46.
:2S,40S,50R)-:2)-2-Chloro-1,1-diphenyl-2-:4-methyl-5-
phenyl-2-oxazolin-2-yl)ethan-1-ol 44b): White needles, mp
126±1298C 4hexane). [a]Dˆ2182 4c 1, CHCl3). dr.99%
4from 1H NMR) 1H NMR 4300 MHz): d 0.63 4d, Jˆ7.0Hz,
3H, CH3CH), 4.23±4.33 4m, 1H, CHCH3), 5.48 4d, Jˆ
9.9 Hz, CHPh), 5.56 4s, 1H, CHCl), 5.78 4s, 1H, exchanges
with D2O, OH), 7.03±7.06 4m, 2H, aromatic H), 7.20±7.35
4m, 9H, aromatic H), 7.56±7.604m, 4H, aromatic H). 13C
NMR 4APT, 50.3 MHz): dˆ16.9 4CH3), 57.5 4CHCH3),
64.5 4CHPh), 78.6 4C±OH), 84.5 4CHCl), 125.6 4ArCH),
126.1 4ArCH), 127.2 4ArCH), 127.5 4ArCH), 128.2
4ArCH), 128.4 4ArCH), 128.5 4ArCH), 135.5 4ArC), 144.8
4ArC), 164.5 4CvN). GC-MS 470eV); m/z 4%): 211 43.4,
M1122Ph2CO), 209 411.1, M12Ph2CO), 132 48.9), 105
433.5), 103 480.1), 68 4100.0), 41 422.9). FTIR 4KBr, cm21):
3229 4broad, OH), 1662 4CvN), 1454, 1378, 1089, 953,
750, 706. Anal. Calcd for C24H22ClNO2: C 73.56, H 5.66,
N 3.57. Found C 73.45, H 5.64, N 3.26.
4.2.2. Preparation of 04S,5R)-02)-2-chloromethyl-4-
methyl-5-phenyl-2-oxazoline 02). tert-Butyl hypochlorite
4859 mg, 7.92 mmol) was added dropwise to a stirred
solution of 44S,5R)-42)-2,4-dimethyl-5-phenyl-2-oxazoline
41a) 41.26 g, 7.20mmol) in anhydrous CCl 4 425 mL), under
N2 at an ice-bath temperature. After 30min the ice-bath was
removed and the reaction mixture was stirred at room
temperature for 3 h. The solvent was removed by rotary
evaporator under reduced pressure and the crude product
puri®ed by ¯ash chromatography 4AcOEt/petroleum
ether 8/2). Oxazoline 2 was isolated in 88% yield as a
pale oil. Small amounts of 7a, that could form on
standing on silica gel, can be removed by washing the
product with cold hexane. [a]Dˆ2253.6 4c 1 CHCl3)
[lit.,11 [a]Dˆ2269 4c 3.68 CHCl3)]. 1H and 13C NMR spec-
troscopical data are similar to those reported in Ref. 11. GC-
MS 470eV); m/z 4%): 211 42.7, M112), 209 48.1, M1), 132
47.3), 105 432.6), 103 473.6), 68 4100.0), 41 4234). FTIR
4®lm, cm21): 3032, 2978, 1670 4CvN), 1455, 1344, 978,
746, 701.
4.2.4. Preparation of oxazolinyl epoxides 5a and 5b. The
chlorohydrins 4a and 4b were cyclized quantitatively to
epoxides 5a and 5b, respectively, in NaOH/i-PrOH.32 The
latter were puri®ed by ¯ash chromatography 4silica gel;
petroleum ether/AcOEt: 9:1) and showed the following
data:
4.2.3. Preparation of chlorohydrins 4a and 4b. A solution
of 16.74 mmol of lithium diisopropylamide 4LDA) 4from
2.34 mL of diisopropylamine and 10.84 mL of 1.54 M
n-BuLi in 15 mL of dry THF) was prepared under N2 and
stirred at 08C for 15 min. To this solution, cooled at 21008C
with a methanol±liquid nitrogen bath, a mixture of
44S,5R)-42)-2-chloromethyl-4-methyl-5-phenyl-2-oxazoline
42) 42.50g, 11.96 mmol) and benzophenone 42.39 g,
13.16 mmol) in 50mL of dry THF was added dropwise
over a 2 h period and the resulting yellow solution was
stirred at this temperature for 15 min. Then the reaction
mixture was quenched 4at 21008C), with sat. aq. NH4Cl
and extracted with Et2O 43£30mL). The combined organic
layers were dried 4Na2SO4) and concentrated in vacuo. The
crude mixture was ¯ash-chromatographed 4silica gel;
petroleum ether/AcOEt: 9/1) to give 4a and 4b 474% overall
:2S,40S,50R)-:2)-2-:3,3-diphenyloxiranyl)-4-methyl-5-
phenyl-2-oxazoline 45a): Oil. [a]Dˆ2112.04 c 1 CHCl3). 1H
NMR 490MHz): dˆ0.60 4d, Jˆ7.0Hz, 3H, C H3), 4.13±
4.504m overlapping s at 4.23, 1H, C HCH3), 4.23 4s, 1H,
CH epoxide), 5.66 4d, Jˆ10.0 Hz, 1H, CHPh), 6.50±6.77
4m, 2H, aromatic H), 7.27±7.87 4m, 13H, aromatic H). 13C
NMR 4125 MHz): dˆ17.7 4CH3), 59.9, 65.0, 66.3, 125.7
4ArCH), 126.6 4ArCH), 127.5 4ArCH), 127.9 4ArCH),
128.1 4ArCH), 128.2 4ArCH), 128.3 4ArCH), 128.41
4ArCH), 128.43 4ArCH), 135.6 4ArC), 136.04Ar C), 139.0
4ArC), 161.4 4CvN). GC-MS 470eV); m/z 4%): 355 412.1,
M1), 338 413.3), 248 415.6), 238 428.8), 208 428.1), 165
461.5), 118 4100.0), 105 426.8), 77 422.8), 67 415.4). FTIR
4®lm, cm21): 2975, 1656 4CvN), 1542 41450), 1079, 755,
702. Anal. Calcd for C24H21NO2: C 81.10, H 5.96, N 3.94.
Found C 80.98, H 5.91, N 3.93.
1
and isolated yield; dr 4a/4b 1:1.6, from H NMR) which
showed the following data:
:2R,40S,50R)-:2)-2-Chloro-1,1-diphenyl-2-:4-methyl-5-
phenyl-2-oxazolin-2-yl)ethan-1-ol 44a): Colorless pris-
matical bipyramidal crystals, mp 112±1138C 4hexane).
:2R,40S,50R)-:2)-2-:3,3-diphenyloxiranyl)-4-methyl-5-
phenyl-2-oxazoline 45b): Oil. [a]Dˆ295.04 c 1, CHCl3). 1H
NMR 490MHz): dˆ0.37 4d, Jˆ7.0Hz, 3H, C H3), 4.07±
4.504m overlapping s at 4.20, 1H, C HCH3), 4.204s, 1H,
CH epoxide), 5.37 4d, Jˆ10.0 Hz, 1H, CHPh), 6.87±7.72
43m, 15H, aromatic H). 13C NMR 4APT, 50.3 MHz): dˆ
17.1 4CH3), 58.9, 64.9, 66.8 4CPh2), 84.6, 126.04Ar CH),
127.3 4ArCH), 127.7 4ArCH), 127.9 4ArCH), 128.08
4ArCH), 128.12 4ArCH), 128.2 4ArCH), 128.4 4ArCH),
135.6 4ArC), 136.04Ar C), 138.8 4ArC), 160.6 4CvN).
GC-MS 470eV); m/z 4rel.int.): 355 45.8, M1), 338 48.2),
248 415.8), 238 428.0), 208 426.7), 165 461.4), 118
4100.0), 105 427.5), 77 425.1), 67 416.2). FTIR 4®lm,
cm21): 3408, 1647 4CvN), 1540, 1448, 1219, 1089, 703.
Anal. Calcd for C24H21NO2: C 81.10, H 5.96, N 3.94. Found
C 80.95, H 5.93, N 3.94.
1
1
[a]Dˆ2105.2 4c1 CHCl3). dr 96:4 4from H NMR) H
NMR 4300 MHz): dˆ0.39 4d, Jˆ7.0Hz, 3H, C H3), 4.36±
4.46 4m, 1H, CHCH3), 5.51 4s, 1H, CHCl), 5.54 4d, Jˆ
10.1 Hz, 1H, CHPh), 6.18 4s, 1H, exchanges with D2O,
OH), 6.69±6.71 4m, 2H, aromatic H), 7.18±7.37 4m, 9H,
aromatic H), 7.59±7.64 4m, 4H, aromatic H). 13C NMR
4APT, 50.3 MHz): dˆ16.8 4CH3), 56.1, 64.3, 78.8
4C±OH), 84.7, 125.7 4ArCH), 125.8 4ArCH), 126.1
4ArCH), 126.5 4ArCH), 127.3 4ArCH), 127.5 4ArCH),
127.6 4ArCH), 128.2 4ArCH), 128.3 4ArCH), 128.5
4ArCH), 135.1 4ArC), 142.6 4ArC), 144.3 4ArC), 165.1
4CvN). GC-MS 470eV); m/z 4%): 211 44.2, M11
22Ph2CO), 209 48.1, M12Ph2CO), 132 49.1), 105
431.4), 103 476.0), 68 4100.0) 41 422.4). FTIR 4KBr,
cm21): 3362 4broad, OH), 1667 4CvN), 1450, 1383,

6782
S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
4.2.5. Deblocking of oxazolinyl epoxides 5a and 5b to
02S)-3,3-diphenyl-oxirane-2-carbaldehyde 06a) and 02R)-
3,3-diphenyl-oxirane-2-carbaldehyde 06b). To a solution
of oxazolinyl epoxide 45a or 5b) 4443 mg, 1.25 mmol) in
dry CH2Cl2 48.5 mL), under N2 at 08C, methyl tri¯ate
41.75 mmol, 197.6 mL) was added directly. The resulting
solution was stirred for 30min at 0 8C; then it was allowed
to cool to 2788C with a methanol±liquid nitrogen bath and
a solution of NaBH4 429.5 mg) in dry THF 45.5 mL)/dry
EtOH 41.38 mL) was added dropwise. The reaction mixture
was allowed to warm to room temperature and stirred for
10h in the case of 5a 4with 2 equiv. of 4COOH)2´2H2O 1%
w/w) and for only 30min 4with no oxalic acid) in the case of
5b; then it was poured into 50mL of water, extracted with
Et2O 43£10mL) and the combined organic layers dried
4Na2SO4). Concentration of the ethereal solution left an oil
that was quickly ¯ash-chromatographed 4silica gel; petro-
leum ether/AcOEt: 9:1) 4so removing also little quantities of
benzophenone as degradation product) to give: 42S)-46a)
472% yield) [a]Dˆ238 4c 1 CHCl3). tRˆ83.39 min.
10b, were separated by ¯ash chromatography 4silica gel;
petroleum ether/AcOEt: 9:1) 487% overall and isolated
1
yield; dr 10a/10bˆ1:2.2, from H NMR) and showed the
following data:
:2R,40S)-:1)-2-Chloro-1,1-diphenyl-2-:4-isopropyl-2-oxazo-
lin-2-yl)ethan-1-ol 410a): White needles, mp 138±1398C
1
4hexane). [a]Dˆ137.47 4c 1 CHCl3). dr: 97/3 4from H
NMR). 1H NMR 4500 MHz): dˆ0.49 and 0.54 42£d,
Jˆ6.7 Hz, 2£3H, 2£CH3), 1.25±1.34 4m, 1H), 3.83±3.91
4m, 2H), 4.20±4.23 4m, 1H), 5.38 4s, 1H), 6.26 4br s, 1H,
exchanges with D2O, OH), 7.13±7.24 4m, 2H, aromatic H),
7.26±7.34 4m, 4H, aromatic H), 7.54±7.61 4m, 4H, aromatic
H). 13C NMR 4125 MHz): dˆ17.5, 17.9, 31.9, 56.5, 70.2,
71.5, 78.5, 125.2, 125.4, 127.0, 127.1, 128.0, 128.3, 142.8,
145.4, 165.1. GC-MS 470eV); m/z 4%): 161 [0.73, M12
Ph2CO), 118 4100.0), 92 418.7), 90 457.7), 76 47.4), 55
438.0), 54 438.6), 43 476.5). FTIR 4KBr, cm21): 3284
4broad, OH), 2954, 1661 4CvN), 1454, 1199, 1035, 942,
753, 696. Anal. Calcd for C20H22ClNO2: C 69.86, H 6.45, N
4.07. Found C 69.52, H 6.27, N 4.30.
ee%ˆ91; 42R)-46b) 493% yield) [a]Dˆ144.6 4c
1
CHCl31). tRˆ86.61 min. ee%.99. Waxy solid, mp 70±
728C. H NMR 490MHz): dˆ3.67 4d, Jˆ6.0Hz, 1H, C H
:2S,40S)-:2)-2-Chloro-1,1-diphenyl-2-:4-isopropyl-2-oxazo-
lin-2-yl)ethan-1-ol 410b): White needles, mp 127±1288C
epoxide), 7.23±7.504m, 1H0, aromatic
H), 8.92 4d,
Jˆ6.0Hz, 1H, C HO). 13C NMR 4APT, 50.3 MHz): dˆ
66.2 4CH), 67.5 4CPh2), 126.4 4ArCH), 128.1 4ArCH),
128.6 4ArCH), 128.7 4ArCH), 134.8 4ArC), 138.5 4ArC),
198.4 4CHO). GC-MS 470eV); m/z 4%): 224 43.6, M1),
223 43.4), 207 411.0), 195 460.7), 179 419.5), 165 4100.0),
105 412.2), 77 417.9). FTIR 4®lm, cm21): 3044, 2922, 1711
4HCvO), 1444, 1261, 1083, 756, 694. Anal. Calcd for
C15H12O2: C 80.34, H 5.39. Found C 80.74, H 5.77.
4hexane). [a]Dˆ252.4 4c 1, CHCl3). dr.99% 4from H
1
1
NMR). H NMR 4500 MHz): dˆ0.77 and 0.81 42£d, Jˆ
6.8 Hz, 2£3H, 2£CH3), 1.56±1.64 4m, 1H), 3.71±3.76 4m,
1H), 3.93 4like t, Jˆ8.3 Hz, 1H), 4.13 4dd, Jˆ8.5, 9.8 Hz,
1H), 5.404s, 1H), 5.99 4br s, 1H, exchanges with D 2O, OH),
7.14±7.22 4m, 2H, aromatic H), 7.23±7.32 4m, 4H, aromatic
H), 7.45±7.56 4m, 4H, aromatic H). 13C NMR 4125 MHz):
dˆ18.0, 18.3, 32.8, 57.1, 70.3, 71.4, 78.7, 125.5, 125.7, 127.1,
127.2, 128.1, 128.3, 143.1, 144.8, 164.7. GC-MS 470eV); m/z
4%): 161 40.7, M12Ph2CO), 120 434.2), 118 4100.0), 90
435.8), 83 415.4), 54 48.6), 43 415.1). FTIR 4KBr, cm21):
3222 4broad, OH), 2965, 1651 4CvN), 1447, 1362, 1251,
1170, 1064, 967, 759, 696. Anal. Calcd for C20H22ClNO2: C
69.86, H 6.45, N 4.07. Found C 69.47, H 6.32, N 3.68.
4.2.6. Synthesis of chloromethyloxazolines 8 and 12.
Their preparation was performed in the same manner as
that of 44S,5R)-42)-2-chloromethyl-4-methyl-5-phenyl-2-
oxazoline 42) by using 44S)-4-isopropyl-2-methyl-2-oxazo-
line181b and the commercially available 4Aldrich) 44S,5S)-
42)-2-methyl-4-methoxymethyl-5-phenyl-2-oxazoline 1c,
respectively. They showed the following data:
4.2.8. Preparation of oxazolinyl epoxides 11a and 11b.
The chlorohydrins 10a and 10b were cyclized quantitatively
to epoxides 11a and 11b, respectively, in NaOH/i-PrOH.32
The latter were puri®ed by ¯ash chromatography 4silica gel;
petroleum ether/AcOEt: 7:3) and showed the following data:
:4S)-:2)-2-Chloromethyl-4-isopropyl-2-oxazoline 48): Oil,
50%, bp 758C/0.1 mbar. [a]Dˆ297.8 4c 5 CHCl3). [lit.,11
1
[a]Dˆ297.9 4c 2.69 CHCl3)]. H and 13C NMR data are
similar to those reported in Ref. 11. GC-MS 470eV); m/z
4%): 163 40.2, M112), 161 40.6, M1), 120430.9), 119
429.8), 118 4100.0), 92 417.0), 90 454.5), 55 419.6), 54
422.2), 43 442.2), 41 440.3). FTIR 4®lm, cm21): 2955,
1666 4CvN), 1466, 1355, 1238, 977, 755.
:2S,40S)-:2)-2-:3,3-Diphenyloxiranyl)-4-isopropyl-2-oxazo-
line 411a): White solid, mp 83±848C 4hexane). [a]Dˆ28.67
4c 1 CHCl3). dr 97/3 4from 1H NMR). 1H NMR 4500 MHz):
dˆ0.75 and 0.85 42£d, Jˆ6.8 Hz, 2£3H, 2£CH3), 1.56±
1.62 4m, 1H), 3.46 4like t, Jˆ8.7 Hz, 1H), 3.76±3.81 4m,
1H), 4.06 4dd, Jˆ8.5, 9.7 Hz, 1H), 4.08 4d, Jˆ0.5 Hz, 1H),
7.22±7.34 4m, 8H, aromatic H), 7.42±7.46 4m, 2H, aromatic
H). 13C NMR 4125 MHz): dˆ18.0, 18.7, 32.1, 59.7, 66.2,
70.2, 72.3, 127.89, 127.93, 128.0, 128.2, 128.3, 135.7,
138.9, 161.7. GC-MS 470eV); m/z 4%): 307 48.7, M1),
290 429.6), 278 411.0), 208 4100.0), 165 457.2), 152 45.6),
105 49.4), 105 426.8), 77 47.0). FTIR 4KBr, cm21): 3059,
2955, 1658 4CvN), 1453, 1263, 985, 781, 761, 703. Anal.
Calcd for C20H21NO2: C 78.15, H 6.89, N 4.56. Found: C
78.42, H 7.21, N 4.34.
:4S,5S)-:2)-2-Chloromethyl-4-methoxymethyl-5-phenyl-2-
oxazoline 412): Oil, 60%. [a]Dˆ284.1 4c 12 CHCl3). [lit.,11
1
[a]Dˆ284.6 4c 3.37 CHCl3)]. H and 13C NMR data are
similar to those reported in Ref. 11. GC-MS 470eV); m/z
4%): 241 40.2, M112), 239 40.6, M1), 194 441.3), 146 431.6),
119 454.0), 105 425.5), 104 423.8), 103 420.2), 102 434.2), 98
475.4), 91 456.1), 45 4100.0). FTIR 4®lm, cm21): 3027, 2915,
2814, 1666 4CvN), 1447, 1189, 1122, 976, 757, 695.
4.2.7. Preparation of chlorohydrins 10a and 10b. Their
preparation was performed in the same manner as those of
4a and 4b by using in this case chloromethyloxazoline 8.
The two diastereomeric chlorohydrins so obtained, 10a and
:2R,40S)-:2)-2-:3,3-diphenyloxiranyl)-4-isopropyl-2-oxazo-
line 411b): Oil. [a]Dˆ251.3 4c 0.33 CHCl3). dr 97/3 4from

S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
6783
1
1H NMR). H NMR 4500 MHz): dˆ0.64 and 0.69 42£d,
1H), 2.83 4dd, Jˆ9.4, 2.8 Hz, 1H), 3.00 4s, 3H, CH3O),
3.20±3.40 4br s, 1H, exchanges with D2O), 3.81 4s, 1H,
CHCvO), 3.81±3.88 4m partially overlapping s at 3.81,
1H), 4.81 4d, Jˆ4.8 Hz, 1H, CHOH), 6.51 4d, Jˆ8.9 Hz,
1H, exchanges slowly with D2O), 7.21±7.52 43m, 15H,
aromatic H). 13C NMR 4125 MHz): dˆ53.6, 58.8, 63.7,
67.2, 72.1, 74.0, 126.1, 126.3, 127.7, 128.1, 128.29,
128.33, 128.4, 135.5, 138.6, 140.8, 166.7 4CvO). FTIR
4KBr, cm21): 3457 4broad), 3417 4sharp), 3062, 2980,
1653 4CvO), 1536, 1450, 1122, 1050, 761, 629. Anal.
Calcd for C25H25NO4: C 74.42, H 6.25, N 3.47. Found C
74.15, H 6.47, N 3.24.
Jˆ6.7 Hz, 2£3H, 2£CH3), 1.30±1.40 4m, 1H), 3.73±3.82
4m, 2H), 3.90±3.96 4m, 1H), 4.04 4d, Jˆ0.9 Hz, 1H), 7.25±
7.33 4m, 8H, aromatic H), 7.42±7.48 4m, 2H, aromatic H).
13C NMR 4125 MHz): dˆ18.2, 18.6, 32.5, 59.5, 66.5, 70.6,
72.4, 126.8, 127.98, 128.04, 128.3, 128.4, 135.7, 139.0,
161.4. GC-MS 470eV); m/z 4%): 307 410.9, M1), 290
493.5), 278 410.7), 248 412.8), 208 4100.0), 191 425.4),
165 458.0), 105 410.9), 77 47.8). FTIR 4CHCl3, cm21):
2962, 2928, 1680, 1658, 1657, 1602, 1447, 1262, 1098,
1010, 823. Anal. Calcd for C20H21NO2: C 78.15, H 6.89,
N 4.56. Found: C 78.33, H 7.37, N 4.55.
4.2.9. 01S,2R)-02)-2-Chloro-N-02-hydroxy-1-methyl-2-
phenylethyl)acetamide 07a). White solid, mp 113±1148C
4.2.13. General procedure for the lithiation of the chloro-
methyloxazolines 8 and 12 and reaction with Ph2CO in
various solvents 0Tables 1 and 2). A solution of LDA
41.4 equiv.) in the given solvent 41.6 mL mmol21) was
cooled to 21008C with a methanol±liquid nitrogen bath.
A mixture of oxazoline 8 or 12 41 equiv.) and Ph2CO
41.3 equiv.) in the given solvent 41.6 mL mmol21) was
added dropwise and the resulting solution stirred at
21008C for the ®xed time 4see Tables 1 and 2) under N2.
Then, the reaction mixture was quenched with sat. aq.
NH4Cl and extracted with Et2O. The combined organic
layers were dried over Na2SO4 and concentrated in vacuo.
The resulting crude mixture of chlorohydrins 10a and 10b or
14a and 14b was puri®ed by column chromatography 4silica
gel; petroleum ether/AcOEt 9:1) 4see Tables 1 and 2 for
isolated yields) and subsequently cyclized quantitatively to
a diastereomeric mixture of epoxides 11a and 11b or 15a and
15b with NaOH/i-PrOH.32 Diastereomers 11a and 11b could
be separated and characterized as previously described.
These new compounds showed the following data:
1
4hexane). [a]Dˆ253.9 4c 1 CHCl3). H NMR 4300 MHz)
4as reported in Ref. 11). GC-MS 470eV); m/z 4rel.int.): 228
40.5, M111), 210 41.5), 134 412.6), 121 476.0), 107 445.4),
86 472.6), 77 461.5), 57 427.6), 44 4100.0). FTIR 4KBr,
cm21): 3308, 1646, 4CvO), 1544 4NH), 1264, 1024, 760,
703, 585. Anal. Calcd for C11H14NO2Cl: C 58.03, H 6.20, N
6.15. Found C 58.43, H 6.45, N 6.02.
4.2.10. 01S,2S)-01)-2-Chloro-N-02-hydroxy-1-methoxy-
methyl-2-phenylethyl)acetamide 07b). White solid, mp
1
114±1168C 4Et2O). [a]Dˆ143 4c 5.3 CHCl3). H NMR
4300 MHz): d 1.61 4br s, 1H, exchanges with D2O), 3.37
4s, 3H, CH3O), 3.39±3.49 4br s, 1H, which exchanges with
D2O), 3.54 and 3.59 42£dd, 2H, AB part of an ABMX
2
3
3
system, JABˆ9.6 Hz, JAMˆ3.4 Hz, JBMˆ4.0Hz, C H2O),
3.92 and 3.98 42H, 2£d, AB system, 2JABˆ15.3 Hz, CH2Cl),
4.14±4.21 4m, 1H, M part, CHNH), 5.04 4d, 1H, X part,
3JMXˆ3.7 Hz, CHPh), 7.25±7.35 42m, 5H, aromatic H).
13C NMR 4APT, 50.3 MHz): d 42.6 4CH2Cl), 54.9
4CHNH), 59.4 4CH3O), 73.5 4CH2O), 74.3 4CHOH), 125.9
4ArCH), 127.9 4ArCH), 128.4 4ArCH), 140.7 4ArC), 166.3
4CONH). GC-MS 470eV); m/z 4%): 24040.1, M 1217), 121
430.1), 119 4100.0), 84 430.5), 79 423.6), 77 434.7), 74
429.4). FTIR 4KBr, cm21): 3320±3200 4br, OH), 1644
4CvO), 1527 4NH). Anal. Calcd for C12H16NO3Cl: C
55.93, H 6.26, N 5.43. Found C 56.28, H 6.49, N 5.38.
:2S,40S,50S)-2-:3,3-diphenyloxiranyl)-4-methoxymethyl-5-
phenyl-2-oxazoline :major) 415a): Inseparable mixture of
1
diastereomers 41.35±2:1 ratio by H NMR analysis). Oil,
see Table 2 for isolated yields in various solvents. 1H
NMR 4500 MHz): dˆ3.13 4dd, Jˆ6.8, 9.7 Hz, 1H,
CHaOCH3 minor), 3.304s, 3H, C H3O minor), 3.36 4s, 3H,
CH3O major), 3.41±3.44 4m, 2H, CHbOCH3 minor1
CHaOCH3 major), 3.52 4dd, Jˆ4.3, 9.7 Hz, 1H, CHbOCH3
major), 3.94±4.01 42m, 2H, CHCH2 major1CHCH2
minor), 4.19 4s, 1H, CH epoxide major), 4.23 4s, 1H, CH
epoxide minor), 5.06 4d, Jˆ7.5 Hz, 1H, CHPh major), 5.19
4d, Jˆ7.7 Hz, 1H, CHPh minor), 6.70±6.72 4m, 2H,
aromatic H minor), 6.92±6.94 4m, 2H, aromatic H major),
7.19±7.39 43m, 22H, 11 aromatic H minor111 aromatic H
major), 7.51±7.57 4m, 4H, 2 aromatic H major12 aromatic
H minor). 13C NMR 4125 MHz): dˆ59.0, 59.07, 59.13,
59.4, 66.61, 66.63, 73.6, 73.7, 74.0, 74.1, 84.0, 84.3,
125.2, 125.8, 126.7, 127.1, 127.81, 127.84, 128.02,
128.05, 128.14, 128.18, 128.28, 128.32, 128.35, 128.4,
128.5, 135.5, 135.6, 138.78, 138.84, 139.5, 139.6, 161.9,
162.2. GC-MS 470eV) 4major); m/z 4%): 385 43.4, M1),
368 49.1), 238 435.6), 208 436.7), 165 482.1), 148 4100.0),
147 499.5), 105 442.1), 77 427.2), 45 456.9). GC-MS 470 eV)
4minor); m/z 4%): 385 45.8, M1), 368 49.9), 238 434.6), 208
435.9), 165 481.4), 148 499.6), 147 4100.0), 105 444.3), 77
431.7), 45 464.9). FTIR 4®lm, cm21) 4major1minor): 3061,
2926, 16604C vN), 1679 4CvN), 1603, 1495, 1435, 1449,
1200, 1127, 977, 912, 758, 699. Anal. Calcd for C25H25NO3:
C 77.90, H 6.01, N 3.63. Found: C 78.28, H 6.22, N 3.65.
4.2.11. 02R,10S,20S)-3,3-Diphenyl-N-02-hydroxy-1-meth-
oxymethyl-2-phenylethyl)oxiranecarboxamide
016a).
White solid, mp 152±1538C 4Et2O). [a]Dˆ153.8 4c 0.5
CHCl3). 1H NMR 4300 MHz): dˆ2.00±2.20 4br s, 1H,
exchanges with D2O), 2.85 4dd, Jˆ9.7 and 3.3 Hz, 1H),
3.14 4dd, Jˆ9.7 and 3.8 Hz, 1H), 3.17 4s, 3H, CH3O), 3.94
4s, 1H, CHCvO), 3.90±3.97 4m partially overlapping s at
3.94, 1H), 4.08 4d, Jˆ7.6 Hz, 1H, CHOH), 6.41 4d,
Jˆ8.8 Hz, 1H, which exchanges slowly with D2O), 7.07±
7.54 44m, 15H, aromatic H). 13C NMR 450.3 MHz):
dˆ55.4, 59.0, 64.1, 67.6, 71.5, 74.6, 126.5, 127.9, 128.0,
128.4, 128.47, 128.51, 128.7, 129.0, 135.4, 138.6, 140.1,
167.2 4CvO). FTIR 4KBr, cm21): 3415 4sharp), 3252
4broad), 3033, 1651 4CvO), 1529, 1494, 1448, 1124, 949,
704, 552. Anal. Calcd for C25H25NO4: C 74.42, H 6.25, N
3.47. Found C 74.03, H 6.54, N 3.21.
4.2.12. 02S,10S,20S)-3,3-Diphenyl-N-02-hydroxy-1-meth-
oxymethyl-2-phenylethyl)oxiranecarboxamide
016b).
White solid, mp 132±1338C 4Et2O). [a]Dˆ129.2 4c 1
1
CHCl3). H NMR 4300 MHz): dˆ2.304dd, Jˆ9.4, 4.4 Hz,

6784
S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
4.2.14. 02S,40S,50S)-2-03,3-Bis-p-chlorophenyloxiranyl)-
4-methoxymethyl-5-phenyl-2-oxazoline 017a) 0presumed
con®guration of major isomer, methods A and B, see
Table 3). Inseparable mixture of diastereomers. Oil, see
4s, 3H CH3O minor), 3.44 4d, Jˆ1.1 Hz, 1H, CH epoxide
major), 3.45 4s, 1H, CH epoxide minor), 3.51 4dd, Jˆ5.8,
9.7 Hz, 2H, CHaOCH3 major1CHaOCH3 minor), 3.58±3.63
4m, 2H, CHbOCH3 major1CHbOCH3 minor), 4.16±4.21
4m, 2H, CHCH2 major1CHCH2 minor), 5.35 4d, Jˆ
7.2 Hz, 1H, CHPh major), 5.37 4d, Jˆ7.0Hz, 1H, C HPh
minor), 7.24±7.35 42m, 10H, 5 aromatic H major15
aromatic H minor). GC-MS 470eV) 4major); m/z 4%): 385
45.0, M1), 368 47.3), 340 417.7), 312 410.5), 238 461.9), 205
422.1), 148 4100.0), 118 419.1), 116 434.4), 91 424.3), 83
412.0), 77 410.0), 67 433.2), 55 441.1), 45 457.2), 41
438.6). GC-MS 470eV) 4minor); m/z 4%): 385 45.2, M1),
368 48.1), 340 422.9), 312 410.6), 238 457.4), 205 421.7), 148
4100.0), 116 436.3), 105 414.6), 91 424.2), 77 411.0), 67
1
Table 3 for isolated yields and dr. H NMR 4300 MHz):
dˆ3.23 4dd, Jˆ6.2, 9.7 Hz, 1H, CHaOCH3 minor), 3.30
4s, 3H, CH3O minor), 3.36 4s, 3H, CH3O major), 3.40±
3.52 4m, 3H, CH2OCH3 major1CHbOCH3 minor), 3.94±
4.03 4m, 2H, CHCH2 major1CHCH2), 4.13 4s, 1H, CH
epoxide major), 4.18 4s, 1H, CH epoxide minor), 5.07 4d,
Jˆ7.4 Hz, 1H, CHPh major), 5.18 4d, Jˆ7.8 Hz, 1H, CHPh
minor), 6.67±6.71 4m, 2H, aromatic H minor), 6.86±6.90
4m, 2H, aromatic H major), 7.21±7.45 42m, 22H, 11 aro-
matic H minor111 aromatic H major). 13C NMR
4125 MHz): dˆ59.10, 59.12, 59.15, 59.4, 65.5, 65.6, 73.3,
73.5, 74.0, 74.1, 76.7, 125.2, 125.8, 128.08, 128.11, 128.2,
128.38, 128.43, 128.49, 128.52, 128.58, 128.65, 128.7,
129.47, 129.55, 133.64, 133.65, 134.31, 134.34, 134.5,
134.6, 136.81, 136.83, 139.1, 139.4, 161.4, 161.6. GC-MS
470eV) 4major); m/z 4%): 455 40.8, M112), 453 41.0, M1),
438 47.9), 436 49.7), 199 421.6), 148 4100.0), 116 428.7), 91
419.5), 45 453.0). GC-MS 470 eV) 4minor); m/z 4%): 453
40.7, M1), 438 43.9), 436 45.8), 199 423.7), 148 4100.0),
439.3), 55 448.7), 45 468.0), 41 450.2). FTIR 4KBr, cm21
)
4major1minor): nˆ2930, 1651 4CvN), 1474, 1326, 1131,
1084, 961, 745, 700.
4.2.17. 03R,40S,50S)-2-0Adamantane-2-spiro-20-oxiranyl)-
4-methoxymethyl-5-phenyl-2-oxazoline 017d) 0presumed
con®guration of the major isomer according to Barbier's
technique, see Table 3). Inseparable mixture of dia-
1
stereomers. Oil, see Table 3 for isolated yields and dr. H
116 429.2), 91 411.9), 45 446.9). FTIR 4®lm, cm21
)
4major1minor): 3034, 2926, 1668 4CvN), 1598, 1491,
1400, 1322, 1201, 1198, 1091, 1015, 974, 822, 698.
NMR 4500 MHz): dˆ1.43±2.06 43m, 28H, 14 adamantyl H
major114 adamantyl H minor), 3.33 4s, 3H, CH3O minor),
3.35 4s, 3H, CH3O major), 3.43±3.45 4m overlaps singlets at
3.48 and 3.49, 2H, CHaOCH3 major1CHaOCH3 minor),
3.48 4s, 1H, CH epoxide major), 3.49 4s, 1H, CH epoxide
minor), 3.56±3.62 4m, 2H, CHbOCH3 major1CHbOCH3
minor), 4.10±4.18 4m, 2H, CHCH2 major1CHCH2
minor), 5.32 4d, Jˆ6.9 Hz, 1H, CHPh minor), 5.34 4d, Jˆ
6.9 Hz, 1H, CHPh major), 7.23±7.33 42m, 10H, 5 aromatic
H major15 aromatic H minor). 13C NMR 4125 MHz): dˆ
26.70, 26.73, 26.85, 26.88, 30.7, 30.9, 34.93, 34.97, 34.99,
36.1, 36.52, 36.55, 36.6, 36.9, 37.0, 57.2, 57.3, 59.0, 59.1,
70.2, 70.5, 73.66, 73.73, 74.08, 74.12, 83.9, 84.0, 125.60,
125.61, 128.1, 128.2, 128.57, 128.59, 140.04, 140.06,
163.28, 163.29. GC-MS 470eV) 4major); m/z 4%): 353
42.9, M1), 336 465.7), 308.10 422.0), 252 424.9), 206
455.8), 148 4100.0), 116 445.1), 105 428.8), 91 462.0), 77
434.6), 67 430.6), 45 456.3). GC-MS 470 eV) 4minor); m/z
4%): 353 41.29, M1), 336 427.98), 304 48.77), 252 410.17),
206 445.32), 148 4100), 116 438.36), 105 419.04), 91 448.29),
4.2.15. 02S,40S,50S)-2-01-Oxaspiro[2,5]ottanyl)-4-meth-
oxymethyl-5-phenyl-2-oxazoline 017b) 0presumed con®g-
uration of the major isomer according to Barbier's
technique, see Table 3). Inseparable mixture of dia-
1
stereomers. Oil, see Table 3 for isolated yields and dr. H
NMR 4500 MHz): d 1.44±1.66 42m, 12H, 6 cyclohexyl H
major16 cyclohexyl H minor), 1.70±1.82 4m, 8H, 4 cyclo-
hexyl H major14 cyclohexyl H minor), 3.36 4s, 3H, CH3O
major), 3.37 4s, 3H, CH3O minor), 3.46 4s, 1H, CH epoxide
major), 3.47 4d, Jˆ1.1 Hz, 1H, CH epoxide minor), 3.48±
3.52 4m, 2H, CHaOCH3 major1CHaOCH3 minor), 3.58±
3.63 4m, 2H, CHbOCH3 major1CHbOCH3 minor), 4.15±
4.21 4m, 2H, CHCH2 major1CHCH2 minor), 5.34 4d, Jˆ
7.0Hz, 1H, C HPh major), 5.36 4d, Jˆ6.9 Hz, 1H, CHPh
minor), 7.24±7.36 42m, 10H, 5 aromatic H major15
aromatic H minor). GC-MS 470eV) 4major); m/z 4%): 301
46.3, M1), 284 420.1), 256 429.7), 228 415.7), 200 420.7),
148 476.2), 116 464.8), 105 424.8), 91 438.8), 81 439.4),
67 4100.0), 45 439.2). GC-MS 470 eV) 4minor); m/z 4%):
301 420.5, M1), 256 438.5), 228 419.1), 200 423.4), 148
499.8), 116 470.5), 105 427.5), 91 442.2), 81 442.5), 67
4100.0), 45 477.43). FTIR 4®lm, cm21) 4major1minor):
3033, 2932, 1674 4CvN), 1448, 1270, 1193, 1127, 976,
735, 699.
77 422.55), 67 424.28), 45 462.46). FTIR 4®lm, cm21
)
4major1minor): nˆ3413, 2910, 1674 4CvN), 1450,
1195, 1128, 981, 699.
X-Ray crystallographic data33 4excluding structure factors)
for 4a, 10a and 16a have been deposited at the Cambridge
Crystallographic Data Centre.34
4.2.16. 02S,40S,50S)-2-01-Oxaspiro[2,11]tetradecanyl)-4-
methoxymethyl-5-phenyl-2-oxazoline 017c) 0presumed
con®guration of the major isomer according to method
B, see Table 3). Inseparable mixture of diastereomers.
Acknowledgements
This work was carried out under the framework of the
National Project `Stereoselezione in Sintesi Organica.
Metodologie ed Applicazioni' supported by the
1
Waxy solid, see Table 3 for isolated yields and dr. H
NMR 4500 MHz): dˆ1.34 4br s, 28H, 14 cyclododecyl H
major114 cyclododecyl H minor), 1.50±1.60 4m, 10H, 5
cyclododecyl H major15 cyclododecyl H minor), 1.61±
1.71 4m, 4H, 2 cyclododecyl H major12 cyclododecyl H
minor), 1.81±1.86 4m, 2H, 1 cyclododecyl H major11
cyclododecyl H minor), 3.35 4s, 3H, CH3O major), 3.37
Á
Ministero dell'Universita e della ricerca Scienti®ca e
Tecnologica 4MURST, Rome), and by the University
of Bari and CNR 4Rome). The National Institute of
Health 4GM 18874) for support of this work is also
acknowledged.

S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
6785
decomposition and prolonged contact with silica gel promoted
ring opening to give the hydroxy amide 7a 4Chart 1). Belvisi,
L.; Gennari, C.; Poli, G.; Scolastico, C.; Salom, B.; Vassallo,
M. Tetrahedron 1992, 48, 3945±3960.
References
1. Florio, S.; Capriati, V.; Luisi, R. Tetrahedron Lett. 1996, 37,
4781±4784.
13. PM3 calculations 4Ref. 14) have con®rmed the existence in
compound 4b of a favored intramolecular hydrogen bonding
involving the hydroxy group and the oxazolinyl nitrogen atom
4Chart 1). Some of the different geometries used as starting
structures have led to hydrogen-unbonded higher energy
minima 413.0kcal mol 21). The latter present additional
destabilizing interactions between the oxazolinyl ring and
one of the phenyl groups a to OH. The distance between the
2. Scheeren, J. W.; Lange, J. Tetrahedron Lett. 1984, 25, 1609±
1612.
3. Evans, D. A.; Williams, J. M. Tetrahedron Lett. 1988, 29,
5065±5068.
4. 4a) Takeda, Y.; Matsumoto, T.; Sato, F. J. Org. Chem. 1986,
51, 4728±4731. 4b) Howe, G. P.; Wang, S.; Procter, G.
Tetrahedron Lett. 1987, 28, 2629±2632.
5. Molander, J. A.; Shubert, D. C. J. Am. Chem. Soc. 1987, 109,
576±578.
Ê
phenyl and the oxazolinyl ring is less than 2 A, to be compared
Ê
with a distance of more than 3.4 A in the hydrogen-bonded
6. Parikh, J. R.; Doering, W.vonE. J. Am. Chem. Soc. 1967, 89,
5505±5507.
conformation. Therefore, the hydrogen bond conformation
allows an ideal arrangement of the rings, minimizing steric
interactions. Both stabilizing factors sinergically contribute to
determine the most favored geometry of 4b 4Chart 1). Similar
results were obtained for the diastereoisomer 4a. In this case,
the hydrogen bond is somewhat longer due to an optimal
reciprocal arrangement of the three phenyl rings 4Chart 1). It
is worth pointing out that the X-ray analysis of 4a showed a
7. Mancuso, A. J.; Swern, D. Synthesis 1981, 165±185.
8. Preliminary results presented at 5th International Symposium
on Carbanion Chemistry, Sendai, Japan, August 1-4, 1998, PP
007, p 91.
Á
9. Breton, P.; Andre-Barres, C.; Langlois, Y. Synth. Commun.
1992, 22, 2543±2554.
10. Oxazoline 1a was prepared by the reaction of 41R,2S)-42)-
norephedrine with ethyl acetimidate hydrochloride according
to a procedure reported by: 4a) Meyers, A. I.; Knaus, G.;
Kamata, K.; Ford, M. E. J. Am. Chem. Soc. 1976, 98, 567±
576. Compound 1a showed an optical activity 4[a]Dˆ2246, c
1 CHCl3) quite close to that recently reported by: 4b) Kamata,
K.; Agata, I.; Meyers, A. I. J. Org. Chem. 1998, 63, 3113±
3116.
Ê
lower value of the N±HO bond distance 41.97 A). This must
be ascribed to the known underestimation of non-bonding
forces by semiempirical calculations. 4Ref. 22).
14. MOPAC 6.0: Stewart, J. J. P. QCPE 1990, 455. All of the
calculations were performed using the PM3 Hamiltonian,
with the keywords PM3, EF, and PRECISE in order to
meet rigorous criteria. Geometries were optimized with no
symmetry constraints.
11. Mintz, M.J.; Walling, C. Organic Syntheses; Wiley & Sons:
New York, 1973; Collect. Vol. V, pp 184±187. One equiv. of
tert-butyl hypochlorite has to be used. An excess causes
the undesired formation of 2,2-dichloromethyl-4-methyl-5-
phenyl-2-oxazoline. While this paper was in preparation, a
paper on the synthesis of chiral chloromethyloxazolines has
been published: Kamata, K.; Sato, H.; Takagi, E.; Agata, I.;
Meyers, A.I. Heterocycles 1999, 51 42), 373±378.
15. Evidences have been reported for lithium azaenolates
derived from 2-ethyl-4-methoxymethyl-5-phenyl-2-oxazo-
line: 4a) Hobbler, M. A.; Bergbreiter, D. E.; Newcomb, M.
J. Am. Chem. Soc. 1978, 100, 8182±8185. 4b) Meyers, A. I.;
Snyder, E. S.; Ackerman, J. J. H. J. Am. Chem. Soc. 1978, 100,
8186±8189. 4c) Le Bail, M.; Aitken, D. J.; Vergne, F.; Husson,
H. P. J. Chem. Soc., Perkin Trans. 1 1997, 1681±1689.
16. A multinuclear investigation in THF at low temperature
combined with a study of chemical shift/charge relationships
and semiempirical calculations on the lithium salt of
2-41-chloroethyl)-4,4-dimethyl-2-oxazoline had provided
structural insights into the nature of this anionic system. In
particular, semiempirical calculations had indicated that, of
the two possible geometrical isomers, the E one is the more
stable, as a result of an internal chelation involving the
chlorine atom. See: Abbotto, A.; Bradamante, S.; Florio, S.;
Capriati, V. J. Org. Chem. 1997, 62, 8937±8940.
12. Attempted further puri®cation of 2 by distillation caused
17. Meyers, A. I.; Yamamoto, Y.; Mihelich, E. D.; Bell, R. A.
J. Org. Chem. 1980, 45, 2792±2796.
18. Oxazoline 8 has been prepared by chlorination 4analogously to
the preparation of 2 and 12) of the corresponding 44S)-4-iso-
propyl-2-methyl-2-oxazoline 1b, which was synthesized by
condensation of l-2-amino-3-methyl-1-butanol with ethyl
acetimidate hydrochloride as reported: Kurth, M. J.; Decker,
O. H. W. J. Org. Chem. 1985, 50, 5769±5775.
19. Less advantageously compound 12 can be prepared from
3-methoxy-2-amino-1-phenyl propanol and the ethyl imidate
of chloroacetonitrile. See: Meyers, A. I.; Knaus, G.; Kendall,
P. M. Tetrahedron Lett. 1974, 39, 3495±3498. 44S,5S)-42)-2-
methyl-4-methoxymethyl-5-phenyl-2-oxazoline can also be
prepared as reported in Ref. 10a,b.
20. Structure and energetics of intermediate lithium azaenolates
have been computationally investigated at semiempirical
Chart 1.

6786
S. Florio et al. / Tetrahedron 57 :2001) 6775±6786
stabilized by a p-interaction with the azaenolate anion, has
4Refs. 14,21,22) and ab initio 4Ref. 23) level. In order to
correctly account for weak intramolecular 4intramolecular
coordination) and intermolecular 4coordination with
molecules of solvent) bond forces it is necessary to include
electron correlation. We have applied in the present work the
semiempirical PM3 Hamiltonian and the Density Functional
Theory 4DFT) 4Ref. 24) with the B3LYP functional 4Ref. 25).
It has been shown that the semiempirical PM3 approach is
adequate to correctly compute geometries, whereas it may
have large errors in the energy calculations 4Ref. 22). For
the above reason, geometries were fully optimized at semi-
empirical level but corresponding energies have been
compared with ab initio level 4DFT//PM3) computations.
been often found as a particularly stable species, as in the case
of lithium enolates 4Ref. 22) or aromatic alkyllithium deriva-
tives 4Anders, E.; Opitz, A.; van Eikema Hommes, N. J. R.;
Hampel, F. J. Org. Chem. 1993, 58, 4424±4430).
27. We took into account the role of this effect by explicitly
including two molecules of solvent22 in the PM3 geometry
optimization and energy calculation of the lithium azaenolate
9. As a model of ethereal solvent we chose dimethyl ether.
28. Charges on lithium are 10.32 and 10.21 for 4E)-13a and
4E)-13b, respectively. In the methoxy group, charges on the
oxygen atoms are 20.27 and 20.18, respectively.
29. Meyers had reported that in lithiated 2-ethyl-4-methoxy-
methyl-5-phenyloxazoline the methoxy group provides
coordination to lithium thus determining the stereochemistry
of the reaction with electrophiles. Meyers, A. I. Acc. Chem.
Res. 1978, 11, 375±381. The authors did not mention any
coordination of lithium cation by chlorine in 2-lithiochloro-
methyl-4-methoxymethyl-5-phenyl-2-oxazoline. See Ref. 19.
30. Three preliminary communications on highly diastereo-
selective aldol-type reactions via differently substituted
boron and titanium azaenolates of chiral oxazolines with
ketones and aldehydes, have been recently reported:
4a) Florio, S.; Capriati, V.; Luisi, R.; Abbotto, A. Tetrahedron
Lett. 1999, 40, 7421±7425. 4b) Florio, S.; Capriati, V.; Luisi,
R. Tetrahedron Lett. 2000, 41, 5295±5298. 4c) Capriati, V.;
Florio, S.; Luisi, R. Eur. J. Org. Chem. 2001, 2035±2039.
31. Suffert, J. J. Org. Chem. 1989, 54, 509±512.
21. PM3 lithium parameters: Anders, E.; Koch, R.; Freunscht, P.
J. Comput. Chem. 1993, 14, 1301±1312.
22. The PM3 semiempirical Hamiltonian has been recently shown
to well compare with high level electron-correlated ab initio
computations performed on lithium enolates: Abbotto, A.;
Streitwieser, A.; Schleyer, P. V. R. J. Am. Chem. Soc. 1997,
119, 11255±11268, and referenced cited therein.
23. All ab initio calculations used the GAUSSIAN 94 program
package: Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P.
M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith,
T.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.;
Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman,
J. B.; Cioslowski, J.; Stefanov, B. B.; Nanayakkara, A.;
Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.;
Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker,
J.; Stewart, J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A.
Gaussian 94, Revision E.2; Gaussian, Inc.: Pittsburgh PA,
1995.
32. Florio, S.; Capriati, V.; Russo, V. Gazz. Chim. Ital. 1997, 127,
587±595.
33. The Bruker X-ray structure re®nement package consists of
SAINT Version 5, SHELXTL Version 5, and SMART
Version 5 and is marketed by Bruker AXS, Inc., Madison, WI.
34. Deposition number CCDC-124102 for compound 4a, CCDC
124100 for compound 10a and CCDC-124101 for compound
16a. Copies of the data can be obtained free of charge on appli-
cation to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK 4fax:
4internat.) 144-1223/336-033; e-mail: deposit@ccdc.cam.ac.uk).
24. Seminario, J. M., Politzer, P., Eds.; Modern Density Func-
tional Theory. A Tool for Chemistry, Elsevier: New York,
1995; Vol. 2.
25. 4a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648±5652. 4b) Lee,
C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785±789.
26. The bridged structure, where the charge on the cation is