684
C. Borriello et al. / Tetrahedron: Asymmetry 15 (2004) 681–686
using the Jacobsen catalyst with (S,S)-II containing the
same substituents on the aryl rings.10 In the case of
polymeric matrix or improving the solubility in alter-
native solvents.
D
-
mannose, the inversion of configuration at C(2) pro-
vided a different local geometry (Fig. 5b), which is
apparently less well suited to promote asymmetry in the
reaction. The behaviour of glucose and mannose also
exhibited another notable difference. In fact, the cationic
complex [Mn(Ma)]PF6 displayed a moderate selectivity
(entry 4), while the neutral [Mn(Ma)OH] afforded only
racemic product (entry 5). On the other hand, no such
difference was found amongst the corresponding glucose
complexes (entries 1 and 2). Also these findings may be a
consequence of the different local symmetry generated
by the two sugars around the metal. In the pseudo-C2
symmetry created by glucose, the active axial sites (ax1
and ax2 in Fig. 5a) become nearly equivalent and, hence,
the enantioselectivity did not change if one apical posi-
tion was occupied by the OHꢀ ligand. This did not hold
true in complexes containing mannose ligands, because
the absence of a local C2 symmetry generated a sub-
stantial difference between the axial positions. In this
case, the low enantioselectivity observed when using
[Mn(Ma)(OH)] may be due to the preferential coordi-
nation of the OHꢀ ligand to the apical site where chiral
discrimination is more effective.
4. Experimental
4.1. General methods
NMR spectra were recorded in C6D6 with a 200 or a 300
MHz spectrometer (Varian Model Gemini). The fol-
lowing abbreviations were used for describing the NMR
multiplicities: s, singlet; d, doublet; t, triplet; dd, double
doublet; m, multiplet. Benzyl-4,6-O-2,3-diamino-2,3-
dideoxy-a-
2,3-diamino-2,3-dideoxy-a-
D
-glucoside6a and methyl-4,6-O-benzylidene-
-mannoside6b were pre-
D
pared according to literature methods.
4.2. Synthesis of ligands H2-G and H2-M
The appropriate aldehyde (2 mmol) was added to a stir-
red solution of the 2,3-hexapyranosediamine 3 (1 mmol)
in 4 mL of dry toluene under nitrogen. After stirring for
1 h at 60 ꢁC, the solvent was removed under vacuum. The
addition of methanol to the crude reaction yielded a
yellow microcrystalline solid, which was separated,
washed with methanol and dried under vacuum (yield
The epoxidation of styrene (entries 6–10) gave in all
cases high conversions (99%). The highest ee [54% of
(S)-enantiomer] can be achieved by using [Mn(Gb)]PF6,
while with [Mn(Ga)OH] and [Mn(Ga)]PF6 the enantio-
selectivity was slightly lower (eeÕs 50% and 45%,
respectively). As expected on the grounds of the above
results, the mannose complexes were less selective, and
induced the preferential formation of the other enan-
tiomer (ee 30% and 32%, respectively, for [Mn(Ga)OH]
and [Mn(Ga)]PF6). These findings are in agreement with
the literature,2 because the oxidation of terminal olefins
promoted by Mn salen complexes is known to afford
only moderate eeÕs. For instance, under the same
experimental conditions, the Jacobsen catalyst with
(S,S)-II (R ¼ R0 ¼ t-Bu) promotes the epoxidation of
styrene in 56% ee.10;13 In any case, the results described
herein largely update those obtained14 with the recently
described Mn(III) salen complex derived from a carbo-
1
>75%). H2-Ga: Selected H NMR data (200 MHz, d,
C6D6): d 14.30 (s, 1H, OH), 13.90 (s, 1H, OH), 8.15 (s,
1H, N@CH), 7.75 (s, 1H, N@CH), 5.30 (s, 1H, H7), 4.82
3
(d, 1H, H1, JH1–H2 ¼ 3.6 Hz), 4.75 (d, 1H, CHHPh,
3Jgem ¼ 12.2 Hz), 4.45 (d, 1H, CHHPh), 4.35 (m, 2H, H5
3
3
and H6eq, JH5–H4 ¼ 9.3 Hz, JH5–H6ax ¼ 3JH6eq–H6ax
¼
3
11.8 Hz), 3.90 (t, 1H, H3, JH3–H2 ¼ 3JH3–H4 ¼ 9.3 Hz),
3.60 (m, 2H, H4 and H6ax), 3.45 (dd, 1H, H2) ppm.
Selected 13C NMR data (75.5 MHz, d, C6D6): d 170.8,
169.2, 159.1–118.3 (20C, aromatics), 101.7, 98.2, 79.9,
71.6, 69.9, 69.3, 67.8, 63.9 ppm. IR (Nujol, KBr): m
1628 cmꢀ1. Anal. Calcd for C50H64N2O6: C, 76.11; H,
8.18; N, 3.55. Found: C, 75.87; H, 8.33; N, 3.56. H2-Gb:
1
Selected H NMR data (200 MHz, d, C6D6): d 14.55 (s,
1H, OH), 14.05 (s, 1H, OH), 8.05 (s, 1H, N@CH), 7.75 (s,
1H, N@CH), 5.03 (s, 1H, H7), 4.75 (d, 1H, H1,
3JH1–H2 ¼ 3.6 Hz), 4.70 (d, 1H, CHHPh, 3Jgem ¼ 12.2 Hz),
4.45 (d, 1H, CHHPh), 4.30 (m, 2H, H5 and H6eq,
hydrate, that is, by suitable modification of b-L-ido-
furanose. In that study, styrene was oxidised in 13% ee
and 75% yield.
3JH5–H4 ¼ 9.4, JH5–H6ax ¼ 3JH6eq–H6ax ¼ 11.2 Hz), 3.95 (t,
3
1H, H3, 3JH3–H2 ¼ 3JH3–H4 ¼ 9.8 Hz), 3.65 (m, 2H, H4 and
H6ax), 3.40 (dd, 1H, H2) ppm. Selected 13C NMR data
(75.5 MHz, d, C6D6): d 170.2, 168.7, 161.3–118.4 (20C,
aromatics), 101.7, 98.1, 79.8, 71.5, 69.9, 69.3, 67.8,
63.7 ppm. IR (Nujol, KBr): m 1629 (C@N) cmꢀ1. Anal.
Calcd for C42H48N2O6: C, 74.53; H, 7.15; N, 4.14. Found:
3. Conclusion
We have shown that with the appropriate modification
of common carbohydrates we can afford ligands useful
for asymmetric synthesis. More precisely, Mn(III) cat-
1
C, 74.71; H, 7.04; N, 4.29. H2-Ma: Selected H NMR
alysts with salen ligands derived from 2,3-
D
-glucosedi-
data (300 MHz, d, C6D6): d 14.05 (s, 1H, OH), 13.73 (s,
1H, OH), 8.09 (s, 1H, N@CH), 8.08 (s, 1H, N@CH), 4.97
(s, 1H, H7), 4.51 (d, 1H, H1), 4.16 (m, 4H, H3, H5, H6ax
amine have been found to be as active as those based on
1,2-cyclohexanediamine. These promising results de-
serve further investigation with the aim of exploiting the
presence of other functional groups naturally present on
the carbohydrate skeleton. For instance, –OH groups
not involved in metal coordination (C1, C4 and C6),
may be used for other purposes, such as anchoring to a
and H6eq, JH3–H2 ¼ 4.0 Hz, JH3–H4 ¼ 3JH5–H4 ¼ 9.4 Hz),
3.65 (t, 1H, H4), 3.48 (dd, 1H, H2) ppm. Selected 13C
NMR data (75.5 MHz, d, C6D6): d 169.7, 169.3, 158.9–
118.5 (16C, aromatics), 101.4, 77.3, 72.0, 69.0, 67.2, 65.2,
3
3
59.8, 54.7 ppm. IR (Nujol, KBr): m 1628 (C@N) cmꢀ1
.