Glycosylidene carbenes: Part 31. Glycosylidene diaziridines: Stereoselective addition of ammonia and methylamine to lactone oxime sulfonates

1488

Helvetica Chimica Acta Vol. 86 (2003)

Glycosylidene Carbenes

Part 31¹)

Glycosylidene Diaziridines: Stereoselective Addition of Ammonia and Methylamine

to Lactone Oxime Sulfonates

by Bruno Bernet, Sissi E.Mangholz , Karin Briner, and Andrea Vasella*

Laboratorium f¸r Organische Chemie, ETH-Hˆnggerberg, CH-8093 Z¸rich

Dedicated to Professor Jack D. Dunitz on the occasion of his 80th birthday

The diastereoselectivity of the addition of NH₃and MeNH₂to glyconolactone oxime sulfonates and the

structures of the resulting N-unsubstituted and N-methylated glycosylidene diaziridines were

The ¹⁵N-labelled glucono- and galactono-1,5-lactone oxime mesylates 1* and 9* add NH₃mostly axially

(> 3 :1; Scheme 4), while the ¹⁵N-labelled mannono-1,5-lactone oxime sulfonate 19* adds NH₃mostly

equatorially (9 :1; Scheme 7). The ¹⁵N-labelled mannono-1,4-lactone oxime sulfonate 30* adds NH₃mostly from

the exo side (>4 :1; Scheme 9). The configuration of the N-methylated pyranosylidene diaziridines 17, 18, 28,

and 29 suggests that MeNH₂adds to 1, 9, 19, and 23 mostly to exclusively from the equatorial direction (> 7 :3;

Schemes 5 and 8). The mannono-1,4-lactone oxime sulfonate 30 adds MeNH₂mostly from the exo side (85 :15;

Scheme 10), while the ribo analogue 37 adds MeNH₂mostly from the endo side (4 :1; Scheme 10). Analysis of

the preferred and of the reactive conformers of the tetrahedral intermediates suggests that the addition of the

amine to lactone oxime sulfonates is kinetically controlled. The diastereoselectivity of the diaziridine formation

is rationalized as the result of the competing influences of intramolecular H-bonding during addition of the

amines, steric interactions (addition of MeNH₂), and the kinetic anomeric effect.

The diaziridines obtained from 2,3,5-tri-O-benzyl-d-ribono- and -d-arabinono-1,4-lactone oxime meth-

anesulfonate (42 and 48; Scheme 11) decomposed readily to mixtures of 1,4-dihydro-1,2,4,5-tetrazines, pentono-

1,4-lactones, and pentonamides.

The N-unsubstituted gluco- and galactopyranosylidene diaziridines 2, 4, 6, 8, and 10 are mixtures of two

trans-substituted isomers (S/R ca. 19 :1, Scheme 2). The main, (S,S)-configured isomers S are stabilised by a

weak intramolecular H-bond from the pseudoaxial NH to ROÀC(2). The diaziridines 12, derived from

GlcNAc, cannot form such a H-bond; the (R,R)-isomer dominates (R/S 85 :15; Scheme 3). The 2,3-di-O-benzyl-

d-mannopyranosylidene diaziridines 20 and 22 adopt

a

⁴C₁conformation, which does not allow an

intramolecular H-bond; they are nearly 1:1 mixtures of R and S diastereoisomers, whereas the ^OH₅

conformation of the 2,3:5,6-di-O-isopropylidene-d-mannopyranosylidene diaziridines 24 is compatible with a

weak H-bond from the equatorial NH to OÀC(2); the (R,R)-isomer is favoured (R/S ꢀ 7:3; Scheme 6). The

mannofuranosylidene diaziridine 31 completely prefers the (R,R)-configuration (Scheme 9).

Introduction. Glycosylidene carbenes (VI in Scheme 1) are glycosylating agents.

They are particularly useful for the glycosylation of tertiary, sterically hindered, and

poorly nucleophilic hydroxy compounds and for the detection of intramolecular H-

bonds in partially protected monosaccharides (see [2 6] and refs. cit. therein). These

carbenes are generated by thermolysis or photolysis of glycosylidene diazirines V. The

diazirines were prepared from partially protected aldose oximes I by oxidation ( ! II),

1

)

Part 30: [1]

Helvetica Chimica Acta Vol. 86 (2003)

1489

sulfonylation ( ! III), treatment with NH₃in MeOH, and oxidation of the diaziridines

IV with I₂(Scheme 1). These diaziridines are, as a rule, mixtures of trans-configured

diastereoisomers.

Scheme 1

OR

OSO₂R¹

OH

O

OH

RO

N

RO

N

RO

N

OR

I

II

III

NH₃/MeOH

OR

O

N

NH

I₂/Et₃N

MeOH

RO

NH

OR

N

OR

VI

V

IV

The transformation of glyconolactone oxime sulfonates to spirodiaziridines is

initiated by the addition of NH₃to the CN bond, but the structure of the resulting

diastereoisomeric trans diaziridines does not betray the equatorial or axial direction of

this attack. We report on the configuration of the diaziridines and on the direction of

the addition of NH₃and of MeNH₂.

Results and Discussion. 1. d-Gluco- and d-Galactopyranosylidene Diaziridines.

The transformation of the gluco- and the galacto-configured sulfonates 1 [7], 3 [5], 5

[8], 7 [5], and 9 [7] (Scheme 2) into the corresponding diaziridines has already been

1

described. According to the H-NMR spectra of CDCl₃solutions, the diaziridines are

ca. 19 :1 mixtures of two trans-configured diastereoisomers 2S/2R²), 4S/4R, 6S/6R, 8S/

8R, and 10S/10R as evidenced by the vicinal J(NH_e,NH_a) values of 9.4 Hz.

We started our investigation by repeating the addition of NH₃to the benzylated 1

and 9, and found that benzene solutions of the diaziridines 2S/2R and 10S/10R are

more-stable than the previously described CHCl₃solutions, so that solutions in C₆D₆

were used for the more-recent NMR investigations. In either solvent, 2 and 10 were

19:1 mixtures of the S and R diastereoisomers.

¹H-NMR Analysis showed that the (S,S)-diastereoisomers of the C(2)-alkoxy- and

C(2)-acyloxy-diaziridines 2, 4, 6, 8, and 10 dominate in solution. In contradistinction,

the major diastereoisomers of the C(2)-acetamido-diaziridines 12 and 14 possess the

(R,R)-configuration. This difference appears to be mostly due to an intramolecular

NÀH_a¥¥¥ OC(2) H-bond that is possible only for the C(2)-OR diaziridines.

The vicinal coupling constants J(2,3), J(3,4), and J(4,5) evidence a ⁴C₁conformation for 2S, 4S, 4R, 8S, 8R,

and 10S. The configuration of the N-atoms of the tetra-O-benzylated diaziridines 2S and 10S was analysed on the

basis of nuclear Overhauser effects (NOEs; Fig. 1). NOEs of 2.2 3.3% were observed between the NH of 2S

2

)

Throughout the paper, S denotes the (S,S)- and R the (R,R)-configuration at the N-atoms.

1490

Helvetica Chimica Acta Vol. 86 (2003)

Scheme 2

OBn

O

OBn

O

OBn

O

H_e

BnO

NH₃

+

N

OMs

N

BnO

N

95 : 5 (CDCl₃)

95 : 5 (C₆D₆)

1

H_a

2S

H_a

2R

OR

O

OR

O

OR

O

H_e

RO

NH₃

N

+

OTf

N

RO

H_a

RO

N

R = Piv

R = Ac

H_a

97 : 3 (CDCl₃)

3

5

4S

6S

4R

6R

> 97 : 3 (CDCl₃)

H_e

R

O

R

O

R

O

BnO

O

H_e

O

NH₃

O

N

OMs

+

BnO

N

BnO

N

BnO

H_a

BnO

N

H_a

7

R = 4-MeO–C₆H₄

94 : 6 (CDCl₃)

BnO

8S

BnO

8R

BnO

OBn

O

OBn

O

OBn

O

H_e

NH₃

N

BnO

+

OMs

N

BnO

N

95 : 5 (CDCl₃)

95 : 5 (C₆D₆)

H_a

9

10S

10R

resonating at lower field (H_a; 2.68ppm in C ₆D₆) and HÀC(3). A weaker NOE (1.3%) was detected between

NH_aand HÀC(5), but there was no NOE between NH_aand HÀC(2). No NOE×s were detected for the NH of

2S resonating at higher field (NH_e). These observations evidence the (S,S) configuration of the major

diastereoisomer of 2 in solution; the same configuration was established for 2 in the solid state [1]. Similarly, we

observed a NOE of ca. 2% between the low field NH_aof 10S and HÀC(3). A remarkable chemical-shift

difference (Dd) between H_aand H_eof 2S was observed for solutions in C₆D₆(Dd 0.42 ppm) and CDCl₃(Dd

0.30 ppm; Table 2 in Exper. Part). The corresponding Dd values for 10S are even larger by 0.1 ppm. A similar,

large Dd value of 0.52 ppm for the 4,6-O-benzylidenated analogue 8S allows us to unambiguously assign the NH

groups and the configuration. However, only small Dd values were observed for the major isomers S of the

pivaloate 4 (Dd 0.03 ppm) and of the acetate 6 (Dd 0.07 ppm), probably mainly due to the anisotropy effect

of the C(2)OCO (and C(6)OCO?) groups. We tentatively assume that H_ais still more deshielded. The

assignment of the NH of the minor isomers 4R and 8R is based on a comparison with the data of the N-Me

analogues (see below). For the minor isomers (R series), the proximity of the lone pair of the pseudoaxial N-

atom and H-C(3) leads to a downfield shift for HÀC(3) (Dd 0.2 0.3 ppm for 4S/4R and 8S/8R). For the

major isomers (S series), the corresponding lone pair is close to HÀC(5), which is deshielded (Dd 0.74 ppm

for 4S/4R and ca. 0.1 ppm for 8S/8R). The large shift difference for HÀC(5) of the pivaloate 4S/4R correlates

with a different orientation of the (pivaloyloxy)methyl side chain of 4S in C₆D₆as indicated by an inversion of

the ratio of the J(5,6) and J(5,6') values.

The plausibility of the assumed NÀH_a¥¥¥ OC(2) H-bond was checked by AMPAC

calculations (AM1, gas phase [9]) of 6-deoxy-2,3,4-tri-O-methyl-d-glucopyrano-

sylidene diaziridines. According to these calculations, the S isomer is by 3.5 kcal/mol

more stable than the R isomer, and H_aof the S isomer may form a stronger

intramolecular H-bond to MeOÀC(2) (d(H_a¥¥¥ O) 2.49 ä, a(NÀH_a¥¥¥ O) 1038)

than H_eof the R isomer (d(H_e¥¥¥ O) 2.56 ä, a(NÀH_e¥¥¥ O) 1008). This result is in

Helvetica Chimica Acta Vol. 86 (2003)

1491

OBn

BnO

2.26 ppm

H_e

2.24 ppm

H_e

O

BnO

N

BnO

H

BnO

OBn

1.3%

H_a

2.68 ppm

2.2%

N

H

H_a

2.75 ppm

3.3%

2.3%

2.1%

10S

2S

Fig. 1. NOEs between NH and glycosylidene H-atoms of the diaziridines 2S and 10S in C₆D₆solution

accordance with the fact that intramolecular H-bonds between axial and equatorial OH

groups are stronger than intramolecular H-bonds between two equatorial OH groups

[10 12].

Analogues of the above diaziridines possessing an equatorial C(2)ÀNHAc instead

of a C(2)ÀOR group cannot form a H-bond equivalent to the one postulated for the

C(2)ÀOR diaziridines of the S series. AM1 Calculations, indeed, predict the absence of

an intramolecular H-bond in 2-acetamido-2,6-dideoxy-3,4-di-O-methyl-d-glucopyra-

nosylidene diaziridines both in the S isomer (d(H_a¥¥¥ N) 2.81 ä) and in the R isomer

(d(H_e¥¥¥ N) 2.68ä). The R isomer is favoured by only 0.2 kcal/mol.

The preparation of the GlcNAc- and AllNAc-derived diaziridines 12S/12R [5] and

14S/14R [13] (Scheme 3) has already been described. In both cases, a 85 :15 mixture of

diastereoisomers was obtained. A closer analysis of the ¹H-NMR data of these

diaziridines (cf. Table 2 in Exper. Part) allowed determination of the configuration in

the absence of NOE data. The determination is based on the relative chemical shifts of

HÀC(5) of the S/R diastereoisomers. The major isomers possess the (R,R)

configuration, as indicated by the downfield shift for HÀC(5) of the minor isomer

(Dd 0.49 (12S/12R) and 0.14 ppm (14S/14R)), evidencing the proximity to the lone

pair of the pseudoaxial NH group. However, despite the proximity to the N lone pair,

HÀC(3) of 12R (3.62 ppm) is more-shielded than HÀC(3) of 12S (3.87 ppm),

probably due to a different orientation of the AcNH group, as evidenced by the

J(2,HN) values of 9.4 Hz for 12R and of 8.2 Hz for 12S (steric repulsion between

H_aÀN and HÀNAc?), and by the Dd values for HÀC(2), as compared to those of 4S/

4R and 8S/8R (opposite relative shifts). The signals for H_aand H_eof 12S and 12R were

assigned by comparing the d(NH) values with those for 2S/2R (Dd ꢁ 0.04 ppm, except

for Dd 0.19 ppm for NH_aof the R isomer). The small shift difference between NH_a

and NH_eof 14S and 14R (Dd ꢁ 0.11 ppm) prevents an assignment. The isomers 12R,

14S, and 14R adopt a ⁴C₁conformation, whereas 12S exists as mixture of the ⁴C₁chair

and (presumably) a boat conformer (J(2,3) 7.6, J(3,4) 5.5 Hz, J(4,5) not assigned).

The preference for the (R,R)-configuration of the 2-acetamido diaziridines 12 and 14

evidences the effect of the intramolecular H-bond between H_aand ROÀC(2) of the

(S,S)-configured gluco and galacto diaziridines; the origin for the preference of 12R

and 14R is not clear (interaction between H_aÀN and HÀNAc?).

To assess the diasteroselectivity in the transformation of 1 to 2, one needs to

selectively label one N-atom of 2. It appeared convenient to introduce a ¹⁵N label in the

1492

Helvetica Chimica Acta Vol. 86 (2003)

Scheme 3

Fig. 2. Mechanism of the diaziridine formation: pseudoaxial vs. pseudoequatorial attack of NH₃at the anomeric

centre of 1*

oximes I (Scheme 1) using solid ¹⁵NH₂OH ¥ HCl. A preferred labelling of the

pseudoequatorial N-atom of the resulting diaziridines upon addition of NH₃(Fig. 2)

may evidence an axial attack on the C¹⁵N bond of 1*³)⁴), as postulated by the rules of

Deslongchamps [14][15]. To yield the diaziridine with pseudoequatorial ¹⁵N, this attack

must lead, irreversibly or reversibly, to the equatorial mesyloxy amine T2. If T2 is

formed reversibly, ring closure from T2 must be faster than from the axial

diastereoisomer T4. Conversely, labelling of the axial N-atom of 2 upon addition of

NH₃would evidence cyclisation of the axial mesyloxy amine T4. Assuming the validity

of Deslongchamps×s rules, this would mean that T2 is formed reversibly, and that T4

cyclises more rapidly than T2.

The reaction of the unlabelled oximes 1 and 9 with NH₃proceeded in good yields,

speaking against fast equilibrium between T2 and T4 via 1*, and a slow ring closure of

T2 and T4. The initially formed zwitterions T1 and T3 must be transformed into T2 and

T4 before cyclising to the diaziridines. Breakdown of T2 and T4 by elimination of the

3

)

The asterisk denotes ¹⁵N-labelled compounds.

Under the basic conditions, protonation of the ring O-atom, followed by the intermediate opening of the

pyranosylidene ring and (partial) epimerisation at C(1), appears improbable.

4

Helvetica Chimica Acta Vol. 86 (2003)

1493

better leaving group (MsONH₂) is expected to lead to glycosylidene imines, well-

known intermediates of the Fischer-Kiliani cyanhydrin synthesis that are rapidly

hydrolysed or transformed into amides [16].

Oxime formation was optimised for the reaction of 2,3,4,6-tetra-O-benzyl-d-

glucopyranose with ¹⁵NH₂OH ¥ HCl. Reducing the amount of NH₂OH ¥ HCl from 8[17]

to 1.3 equiv. still afforded a nearly quantitative yield of the corresponding crude (E/Z)-

oximes. Oxidation of the crude ¹⁵N-labelled hydroxy oximes derived from 2,3,4,6-tetra-

O-benzyl-d-gluco- and -galactopyranose with MnO₂gave a 61 62% yield of the

lactone oximes 15* and 16* (Scheme 4). Mesylation of 15* and 16* (MsCl, Et₃N) and

crystallisation from Et₂O gave 1* and 9* in 56 and 87% yield, respectively. Treatment of

1* with a saturated solution of NH₃in MeOH, followed by crystallisation at À158, gave

the isotopomers 2Se* and 2Sa*⁵) in a 25 :75 ratio. Repetition of the reaction led to

2Se*/2Sa* 22 :78. Similar transformations of 9* gave the isotopomers 10Se* and 10Sa*

in ratios of 15 :85 and 25 :75. These results are compatible with either an irreversible

pseudoaxial attack of NH₃to 1* and 9*, or with a preferred ring closure of the

equatorial mesyloxy amine.

Scheme 4

OBn

O

OBn

O

OBn

O

H_e

BnO

NH₃

¹⁵N

N

+

N

OR

¹⁵N

BnO

¹⁵N

H_a

15*

1*

R = H

R = Ms

25 : 75 (CDCl₃)

22 : 78 (C₆D₆)

2Se*

2Sa*

BnO

OBn

O

OBn

O

OBn

O

H_e

NH₃

¹⁵N

N

+

N

BnO

OR

¹⁵N

BnO

¹⁵N

H_a

10Sa*

R = H

16*

9*

15 : 85 (CDCl₃)

25 : 75 (C₆D₆)

R = Ms

10Se*

In the ¹⁵N-NMR spectra of the gluconolactone oximes 15* and 1*, a br. d (J 1.7 2.2 Hz) is found at

À65.80 and À60.67 ppm, and, in those of the galactonolactone oximes 16* and 9*, one finds a dd (J 1.2 1.3

and 0.2 Hz) at À77.91 and À80.36 ppm, respectively. The incorporation of ¹⁵N into 15*, 1*, 16*, and 9* leads to

an additional splitting of the signals for HÀC(2) (³J(¹⁵N,H) 1.2 1.6 Hz), HÀC(3) (only of 15* and 1*:

2

1

⁴J(¹⁵N,H) 1.0 1.6 Hz), OH (of 15* and 16*: J(¹⁵N,H) ꢁ 0.8Hz), C(1) (only of 16* and 9*: J(¹⁵N,C) 0.8

1.2 Hz), C(2) (²J(¹⁵N,C) 9.5 11.4 Hz), and C(3) (³J(¹⁵N,C) 1.7 2.6 Hz).

In the ¹H-NMR spectra, the ¹⁵N label of the N-unsubstituted diaziridines 2Se*/2Sa* and 10Se*/10Sa* is

evidenced by additional splitting of only the NH signals, characterized by ¹J(¹⁵N,H) of 56.8 57.3 Hz ⁶) and

²J(¹⁵N,H) of 2.8 3.7 Hz ( cf. Table 2 in Exper. Part). The dd for the NH groups of the (R,R)-configured

diastereoisomers (ca. 5% expected) are probably hidden by noise. The ¹³C-NMR spectra of 2Se*/2Sa* and

10Se*/10Sa* show a single set of signals; the chemical-shift values are identical to those of the unlabelled 2S and

10S (cf. Table 3 in Expert. Part). C(1) and C(2) of 2Se*/2Sa* and 10Se*/10Sa* show couplings of 5.1 6.1 and

5

)

The letters a and e denote the pseudoaxial and the pseudoequatorial position of the NR group stemming

from the amine.

Compare with 56.3 60.7 Hz for ¹⁵N-labelled bis(trifluoromethyl)-diaziridines [18].

6

)

1494

Helvetica Chimica Acta Vol. 86 (2003)

ꢁ 3.9 Hz with ¹⁵N, respectively. The ¹⁵N-NMR spectra of 2Se*/2Sa* and 10Se*/10Sa* in CDCl₃and C₆D₆reflect

1

the same product ratios as in the H-NMR spectra (cf. Table 4 in Exper. Part). The ¹⁵N signals appear as dd×s

1

showing a large J(¹⁵N,H) of 56.5 58.0 ppm and a small ²J(¹⁵N,H) of 2.7 3.6 Hz. The weaker signal for the

pseudoaxial ¹⁵N-atom resonates at higher field (Dd 8.4 10.4 ppm).

To evaluate the scope of the diaziridine formation, we cursorily examined the

addition of MeNH₂, Me₂CHNH₂, Me₃CNH₂, and aniline in MeOH to 9; of these

amines, only MeNH₂reacted to a discernible extent. Treatment of 1 with a 7m solution

of MeNH₂in dry MeOH for 2.5 h at room temperature and purification of the product

by filtration through LiChroPrep-NH₂gave 97% of a 72 :28mixture of 17Se and 17Ra

containing traces (ca. 5%) of an unassigned epimer. Similarly, treatment of 9 led to a

85 :15 mixture of the galacto-configured 18Se and 18Ra. MeNH₂was also added to the

¹⁵N-labelled sulfonates 1* and 9* to facilitate the measurement of ¹⁵N-NMR spectra.

The mixture 18Se/18Ra solidified upon standing. Crystallisation from MeOH afforded

pure 18Se, and recrystallisation in AcOEt/hexane gave crystals suitable for X-ray

analysis.

The solid-state structure of 18Se reveals a pseudoequatorial MeN group, the (S,S)-

4

configuration of the N-atoms, a C₁conformation of the pyranosylidene ring, and a tg

conformation for the BnOCH₂side chain (Fig. 3)⁷). The structure closely resembles

that of the N-unsubstituted 2S [1] (Table 1). Even a (weak) intramolecular H-bond

from N_aÀH to BnOÀ(2) is found in both structures, in keeping with the results of the

AM1 calculations. This H-bond is slightly weaker in 18Se than in 2S as evidenced by

larger N_a¥¥¥ O(2) (Dd 0.05 0.07 ä) and H_a¥¥¥ O(2) distances.

Table 1. Selected Bond Lengths [ä] and Bond Angles [8] for the Diaziridines 18Se and 2S

18Se

2S [1]

18Se

2S [1]

C(5)ÀO(5)

O(5)ÀC(1)

C(1)ÀN_a

C(1)ÀN_e

N_aÀN_e

1.448(3)

1.388(3)

1.445(3)

1.422(3)

1.547(3)

0.94(2)

1.451(3)

1.385(4)

1.436(4)

1.426(4)

1.539(4)

0.90(3)

2.840(3)

2.38(3)

0.92(4)

O(5)ÀC(1)ÀN_a

O(5)ÀC(1)ÀN_e

C(1)ÀN_aÀN_e

C(1)ÀN_eÀN_a

N_aÀC(1)ÀN_e

C(1)ÀN_aÀH

N_eÀN_aÀH

115.8(2)

117.8(2)

56.6(1)

58.1(1)

65.3(1)

109(2)

103(2)

111(2)

115.5(2)

117.7(3)

57.17(19)

57.8(2)

65.0(2)

109(2)

106.2(19)

112(2)

104(2)

N_aÀH_a

N_a¥¥¥ O(2)

H_a¥¥¥ O(2)

N_eÀH_e

2.906(3)

2.43(3)

N_aÀH ¥¥¥ O(2)

N_aÀN_eÀH

N_eÀCH₃

1.448(3)

N_aÀN_eÀCH₃

109.4(2)

The configurations of the N-atoms of 17Se, 17Ra, 18Se, and 18Ra in solution were revealed by NOE

experiments (Fig. 4). NOEs of 2.8 2.9% were observed for H ÀC(3) of the major isomers 17Se and 18Se upon

irradiating the NH signal at 3.02 and 3.08ppm, respectively. This evidences a pseudoaxial NH group and the

(S,S)-configuration of the major isomers. Irradiation of the NMe group of the minor isomers 17Ra and 18Ra

leads to a NOE for HÀC(5) of 2.0 3.3%. In addition, a NOE (3.1%) between NH and HÀC(2) of 17Ra was

observed. Thus, the minor isomers possess a pseudoaxial NMe group and the (R,R) configuration. The

7

)

The crystallographic data have been deposited with the Cambridge Crystallographic Data Centre as

deposition No. CCDC-197608. Copies of the data can be obtained, free of charge, on application to the

CCDC, 12 Union Road, Cambridge CB21EZ, UK (fax: 44(1223)336033; e-mail: deposit@ccdc.cam.a-

c.uk).

Helvetica Chimica Acta Vol. 86 (2003)

1495

Fig. 3. X-Ray structure of 18Se. The H-atoms of the Me and the Bn groups are omitted for clarity.

3.1%

2.84 ppm

H

2.75 ppm

Me

BnO

H

O

N

H

OBn

H

N

H

3.3%

Me

2.65 ppm

2.9%

3.02 ppm

1.5%

17Se

17Ra

2.92 ppm

BnO

OBn

O

2.75 ppm

Me

H

O

N

BnO

H

OBn

H

OBn

N

H

Me

2.68 ppm

2.0%

2.8%

3.08 ppm

2.3%

18Se

18Ra

Fig. 4. NOEs between NH or NMe, and glycosylidene H-atoms of the diaziridines 17Se, 17Ra, 18Se, and 18Ra in

C₆D₆solution

configuration of the diaziridine ring has no influence on the pyranose ring conformation. The same J(2,3),

J(3,4), and J(4,5) (cf. Table 2 in Exper. Part) were observed for the major and the minor isomers, and they

evidence a ⁴C₁conformation of 17Se, 17Ra, 18Se, and 18Ra. As expected, the (S,S)-configuration of the major

isomers correlates with a downfield shift for HÀC(5) of the S isomers (Dd ꢂ 0.3 ppm), and the (R,R)-

configuration of the minor isomers correlates with a downfield shift for HÀC(3) of the R isomers (Dd

0.4 ppm).

1496

Helvetica Chimica Acta Vol. 86 (2003)

The incorporation of ¹⁵N in 17Se*/17Ra* and 18Se*/18Ra* leads to additional splitting of the signals for NH

(¹J(¹⁵N,H) 57.4 58.3 Hz) and NMe (³J(¹⁵N,H) 2.7 2.9 Hz). The same couplings are also visible in the ¹⁵N-

NMR spectra of 17Se*/17Ra* 72 :28and 18Se*/18Ra* 85 :15 where dqs appear at À297.3 to À299.6 ppm (cf.

Table 4 in Exper. Part). In the ¹³C-NMR spectra, the signal for the NMe group of 17Se, 17Ra, 18Se, and 18Ra is

split by coupling with ¹⁵N (²J(¹⁵N,C) 3.6 4.1 Hz). Only the pseudoaxial ¹⁵N-atom of the major isomers 17Se*

and 18Se* couples with C(1) (¹J(¹⁵N,C) 4.1 4.3 Hz; Table 3 in Exper. Part).

Out of the four possible N-Me-diaziridines only the two were detected that can form an intramolecular H-

bond from NÀH to BnOÀC(2). The other isomers are also disfavoured by a 1,5-interaction between the NMe

and BnOÀC(2) groups.

The ratios of the isotopomeric gluco- and galacto-configured diaziridines (2Se*/

2Sa* and 10Se*/10Sa* 22 :78and 25 :75, resp.) and the ratios of the corresponding N-

Me-diaziridines (17Se/17Ra and 18Se/18Ra 72 :28and 85 :15, resp.) characterize a

preferred ring closure from a tetrahedral intermediate with an axial NH₂group

(derived from added NH₃) and an equatorial MeNH group, respectively. This

difference may be due to steric and/or electronic interactions operating either during

the addition of NH₃and MeNH₂to the imino group or during ring closure of the

intermediate mesyloxy amines (see Fig. 2)⁸). As mentioned above, the axial attack of

the nucleophiles on the lactone oxime sulfonates is stereoelectronically favoured. The

axial attack may also be favoured by an (initially intermolecular) H-bond between the

attacking amine and C(2)ÀOR. The strength of such a H-bond should increase parallel

to the C(1)ÀN bond-formation during which the amine is transformed into an

ammonium substituent. That a cis-axial ammonium group may form a stronger H-bond

to C(2)ÀOR than a trans-equatorial one is suggested by the stronger H-bond between

the pseudoaxial NH and C(2)ÀOBn of the C(2)-alkoxylated N-unsubstituted

diaziridines⁹).

To get better insight into the stereoselectivity of the ring-closing step, we modelled

the reactive conformers of the diasteroisomeric 2-amino- and 2-(methylamino)-2-

(mesyloxyamino)-tetrahydropyrans M1/M2 and M3/M4 (AM1 calculation; Fig. 5). In

the reactive conformations, the lone pair of the amino group at C(1) is directed towards

the (mesyloxy)amino substituent. This implies synperiplanar orientation of the HÀN

bonds of M1 and M2 and of the HÀN and MeÀN bonds of M3 and M4 with C(2)ÀC(3)

and C(2)ÀO. The MsO group is oriented away from the amino group, and the N-lone

pair of the MsONH group is antiperiplanar to the C(1)ÀO bond (exo-anomeric effect).

Thus, the reactive conformers M11 and M21 of the epimeric intermediates M1 and M2

have to be compared, and, similarly, the two pairs M31/M32 and M41/M42 of the

epimeric intermediates M3 and M4, respectively. The significantly greater stability of

M11 over M21 shows that the orientation of the MsO group is the main energetic

factor. Among the MeNH isomers, M31 is by far the most favoured (reactive)

conformer. A comparison to M41 and M42 shows that this is due to the equatorial

orientation of the MsONH substituent and to a coplanar arrangement of MeÀN and

OÀC(2), avoiding the interaction with MeOÀC(3) that destabilises M32. The

8

)

The formation of an intermediate nitrene appears improbable, since hydroxylamine O-sulfonic acid does

not form the corresponding nitrene under basic conditions [19].

The relative strength of these H-bonds derives from the ratio of the equilibrating trans-diaziridines. For

the nonequilibrating N-Me-diaziridines, one finds the same relative N ¥¥¥ O distances as for the

unsubstituted analogues, suggesting the same relative strength of their NÀH ¥¥¥ OÀC(2) H-bonds.

9

Helvetica Chimica Acta Vol. 86 (2003)

1497

NHOMs

NH₂

NHOMs

NHMe

O

NHOMs

OMe

M1

M2

M3

M4

OMs

N

H

O

N

H

MeO

MsO

H

N

H

M11

M21

0 kcal/mol

3.6 kcal/mol

OMs

Me

H

O

Me

H

O

N

H

N

H

N

H

MeO

Me

H

MeO

H

Me

MeO

MsO

MeO

N

H

MsO

M42

3.3 kcal/mol

M31

M32

M41

3.0 kcal/mol

0 kcal/mol

2.8 kcal/mol

Fig. 5. Calculated (AM1) relative energies of the reactive intermediates M11, M12, M31, M32, M41, and M42 of

the epimeric 2-aminotetrahydropyrans M1 and M2, and of the epimeric 2-(methylamino)tetrahydropyrans M3

and M4. Destabilizing 1,5-interaction indicated by double-headed arrows.

difference between the coplanar arrangements of MeÀN and C(3)ÀC(2) vs. MeÀN

and OÀC(2) amounts to only 0.3 kcal/mol (compare M41 and M42).

The relative stabilities of these reactive conformers imply a more-facile formation

of the diaziridines derived from the tetrahedral intermediate resulting from an axial

attack of NH₃or MeNH₂. This is not in agreement with the configuration of the N-Me-

diaziridines 17 and 18, although the axial attack is favoured by the kinetic anomeric

effect and by H-bonding of the attacking amine to the cis-oriented ROÀC(2). Steric

interactions in the addition step must then be responsible for the preferred addition of

MeNH₂from the equatorial side. To evaluate the steric interactions during the

addition, we analysed the conformations A1 A6 of the primary addition products

resulting from an equatorial and an axial attack of MeNH₂on 1 (Fig. 6). The

conformers A1 and A4 cannot form an intramolecular H-bond and are disfavoured by a

1,5-interaction between the MeN and the BnOÀC(2) groups. The conformers A2, A3,

A5, and A6 can form an intramolecular H-bond, but conformers A2 and A5 are

disfavoured by a 1,5-interaction between the MeN and the MsO group, and A6 is

disfavoured by 1,5-interactions between MeN and both C(3) and C(5). Only A3 is not

disfavoured by steric interactions. It results from an equatorial addition of MeNH₂to 1.

This means that the kinetic anomeric effect is overruled by the combined influence of

H-bonding and steric interactions that favour the equatorial addition of MeNH₂, a

conclusion that could not be drawn from the addition of NH₃, since, here, both the

kinetic anomeric effect and H-bonding of the cis-axial ammonium group to BnOÀC(2)

favour the axial addition¹⁰).

10

)

For another case, where the interaction of the attacking nucleophile with C(2)ÀOR dominates over the

stereoelectronic control, see [20].

1498

Helvetica Chimica Acta Vol. 86 (2003)

OBn

O

OBn

H

OBn

O

H

N

Me

N

O

BnO

N

H

Me

H

BnO

H

N

OMs

A1

A2

A5

A3

OBn

O

OBn

O

OBn

O

BnO

N

H

N

H

OMs

BnO

H

BnO

H

BnO

Me

N

Me

H

A4

A6

BnO

H

N

BnO

O

N

H

OMs

Me

N

H

OMs

Me

A8

A9

Fig. 6. Intramolecular H-bonds (hashed lines) and destabilizing 1,5-interactions (double-headed arrows) in the

intermediate zwitterionic addition products of MeNH₂to 1 (A1 A6) and to 19 (A8 and A9).

The analysis of the tetrahedral intermediates formed upon equatorial and axial

addition of MeNH₂to the manno-configured analogue 19 of 1 showed that the epimeric

A8 and A9 (Fig. 6) are the preferred conformers, differing essentially by the

intramolecular H-bond of A8. The hypothesis of this dominant influence of the H-

bond predicts a preferred equatorial addition of both NH₃and MeNH₂to 19.

A thermodynamic control of the diastereoselective formation of the diaziridines

was evaluated by calculating the relative energies of the ground-state of M1 M4

(Fig. 5) and of the corresponding diastereoisomers with an axial MeO group. The gas-

phase calculations speak against thermodynamic control; they show higher stability of

the equatorial amines M2 and M4 over M1 and M3 (DE 1.0 and 0.6 kcal/mol, resp.),

and higher stability of the axial amine and methylamine of the diastereoisomers

possessing an axial MeO group (DE 2.8and 0.8kcal/mol, resp.), corresponding to

manno-configured derivatives.

2. d-Mannopyranosylidene-diaziridines. The preparation of lactone oxime sulfo-

nates 19 [7], 21 [5], and 23 [21] has been described (Scheme 6). Reaction of these

sulfonates with NH₃in MeOH gave 20S/20R, 22S/22R, and 24S/24R in ratios of ca. 1 :1

[7], 48:52 [5], and 1:9 [21] (CDCl ₃), respectively. Repetition of the reactions with NH₃

and analysis in C₆D₆led to only slightly different ratios; i.e., 20S/20R 55 :45, 22S/22R

60 :40, and 24S/24R 3 :7, indicating a slightly stronger solvent dependence of the

product ratio in the manno than in the gluco/galacto series.

The assignment of the NH groups is based on NOE experiments for 20S, 20R, 24S, and 24R (Fig. 7). NOEs

of 3.5 5.9% evidence the cis-arrangement of H_aof 20S and H_eof 20R with HÀC(2). The unambiguous

assignment of 20S and 20R is based on a NOE of 3.8% between H_aand HÀC(3) of 20S. The assignment is

corroborated by the NOESY spectrum of 20S/20R 55 :45 in C₆D₆showing strong cross-peaks between H_aand

HÀC(2) of 20S, and between H_eand HÀC(2) of 20R, a weaker cross-peak between H_aand HÀC(3) of 20S,

Helvetica Chimica Acta Vol. 86 (2003)

1499

Scheme 5

H_e

OBn

O

OBn

O

OBn

O

Me

BnO

MeNH₂

N

OMs

+

BnO

N

BnO

N

H_a

Me

1

17Se

17Ra

72 : 28 (C₆D₆)

H_e

OBn

O

OBn

O

OBn

O

Me

BnO

¹⁵N

N

BnO

+

¹⁵N

BnO

¹⁵N

N

H_a

1*

17Se*

72 : 28 (C₆D₆)

17Ra*

H_e

BnO

OBn

O

BnO

OBn

O

BnO

OBn

O

Me

MeNH₂

N

OMs

+

BnO

N

BnO

N

H_a

Me

9

18Se

85 : 15 (C₆D₆)

BnO

18Ra

H_e

BnO

OBn

O

BnO

OBn

O

OBn

O

Me

MeNH₂

¹⁵N

N

+

¹⁵N

BnO

¹⁵N

N

H_a

9*

18Se*

85 : 15 (C₆D₆)

18Ra*

Scheme 6

1500

Helvetica Chimica Acta Vol. 86 (2003)

2.33 ppm

H_e

2.47 ppm

OBn

O

OBn

O

BnO

H_e

N

H

5.9%

H

N

4.1%

H

H_a

3.8%

1.96 ppm

3.5%

1.19 ppm

20S

20R

2.23 ppm

H_e

2.51 ppm

H_e

1.4%

O

N

H

1.8%

N

H

H_a

1.81 ppm

H_a

1.93 ppm

1.0%

24S

24R

Fig. 7. NOEs between NH and glycosylidene H-atoms of the mannopyranosylidene diaziridines 20S, 20R, 24S,

and 24R in C₆D₆solution

and a weak cross-peak between H_aand CH₂(6) of 20R. Additional weak cross-peaks between H_aof 20S and

HÀC(2) of 20R, between H_eand HÀC(2) of 20S, and between H_eof 20R and HÀC(2) of 20S evidence rapid

interconversion of 20S and 20R. Saturation transfer between the NH groups and H₂O prevents quantitative

analysis of the NOESY spectrum. The (R,R)-configuration of the major isomer 24R is established by weak

NOEs (1.4 1.8%) between H_eand HÀC(2), and between H_aand HÀC(5).

The ⁴C₁configuration of 22S and 22R in CDCl₃and C₆D₆, and of 20S in C₆D₆(the glycosylidene H of 20S/

20R ca. 1:1 in CDCl₃were not assigned) is evidenced by J(2,3) 3.0 3.2 and J(3,4) J(4,5) 9.3 9.8Hz ( cf.

Table 5 in Exper. Part). This allows ready assignment of the isomers 22S and 22R according to the relative shifts

of HÀC(3) (downfield shift for HÀC(3) of 22R by 0.2 0.35 ppm) and HÀC(5) (downfield shift for HÀC(5) of

22S by 0.2 0.55 ppm). The isomer 20R does not completely prefer a ⁴C₁conformation, as indicated by J(3,4)

7.1 and J(4,5) 6.1 Hz. Similar values for J(3,4) and J(4,5) suggest a ca. 2 :1 equilibrium between the ⁴C₁

conformer lacking an intramolecular H-bond and the ¹C₄conformer possessing an intramolecular NÀH ¥¥¥ OBn

H-bond, rather than the participation of a twist-boat conformer (J(4,5) of ca. 10 and J(3,4) of ca. 5 Hz are

4

expected for the most probable S₂conformer).

The nearly 1:1 ratio of 20S/20R and 22S/22R is not surprising, since there are no

intramolecular H-bonds in the ⁴C₁conformers. AM1 Calculations for the R isomer of 6-

4

deoxy-2,3,4-tri-O-methyl-d-mannopyranosylidene diaziridine show that H_eof the C₁

conformer does not form an intramolecular H-bond to BnOÀC(2) (d(H_e¥¥¥O) 3.16 ä,

1

a(NÀ H_e¥¥¥O) 838). However, a weak corresponding H-bond is possible in the C₄

conformer (d(H ¥¥¥ O) 2.61 ä, a(NÀH¥¥¥O) 1018). ⁴C₁and the ⁴S₂conformers

(DE 3.7 and 1.8kcal/mol, resp.). AM1 Calculations predict an ^OH₅conformation for

24S and an S₅conformation for 24R, which agrees well with the observed J(2,3), J(3,4),

and J(4,5) values (6.7 8.0, 6.0 7.8, and 10.3 10.4 Hz, resp.; cf. Table 5 in Exper. Part). The

intramolecular H-bond of H_eof 24R to the OC(2) (d(H_e¥¥¥O) 2.36 ä, a(NÀH_eb¥¥¥O)

1058) is responsible for the strong preference of 24R.

Helvetica Chimica Acta Vol. 86 (2003)

1501

The ¹⁵N-labelled methanesulfonate 19* (Scheme 7) was prepared from 25, similarly

as described above for the gluco and galacto isomers 1* and 9*, respectively, and

obtained in an overall yield of 74% from 25. Treatment of 19* with NH₃in MeOH gave

a 50 :41:5 :4 mixture of 20Se*, 20Re*, 20Sa*, and 20Ra*. The product ratio in C₆D₆

1

2

was unambiguously assigned by means of the J(¹⁵N,H) and J(¹⁵N,H) couplings (cf.

Table 5 in Exper. Part). The ratio 20Se*/20Sa* to 20Re*/20Ra* 55 :45 was exactly the

same as for the unlabelled 20S to 20R.

Scheme 7

The reaction of 19 with MeNH₂in MeOH gave 98% of a 85 :15 mixture of 28Se and

28Re (Scheme 8). Similarly, 23 was transformed into a 80 :20 mixture of 29Se and

29Re.

Scheme 8

The presence of ¹⁵N in 27* and 19* is evidenced by odd m/z values for [M H] in the high-resolution

mass spectra (27*: 555.2522, 19*: 633.2282). The ¹⁵N-label leads to additional splitting of the NMR signals for

3

HÀC(2) (³J(H,N) 1.3 Hz), C(2) (²J(C,N) 10.8Hz), and C(3) ( J(C,N) ꢁ 2.6 Hz). The ¹⁵N-NMR spectra of

27* and 19* display a s at À76.4 and À76.6 ppm, resp.

1502

Helvetica Chimica Acta Vol. 86 (2003)

The configuration of 28Se, 28Re, 29Se, and 29Re was established by NOE experiments and by the relative

chemical shifts of HÀC(3) and HÀC(5). The cis-arrangement of H_aand HÀC(2) of 28Se and 29Se is evidenced

by NOEs of 3.6 8.1% (Fig. 8). Similarly, NOEs of 2.3 3.6% reveal the cis-arrangement of MeN and HÀC(2)

of 28Re and 29Re. The unambiguous assignment of the (S,S)-configuration to 28Se and 29Se is based on a NOE

of 3.4% between H_aand HÀC(3) of 28Se, and on NOEs of 0.8 0.9% between MeM and both CH ₂(6) of 29Se.

The assignment of the (R,R) configuration of 28Re and 29Re is based on the downfield shift of HÀC(5) of

28Re and 29Re (Dd 0.42 for 28Se/28Re and 0.49 ppm for 29Se/29Re), and on the downfield shift of HÀC(3)

of 28Se (Dd 0.40 ppm; cf. Table 5 in Exper. Part). Due to the ^OH₅conformation of 29Se and 29Re, HÀC(3)

should be influenced only weakly by the configuration of the pseudoaxial N-atom; this is indeed observed (Dd

0.05 ppm). The pseudoequatorial N-atoms of 28Se and 28Re are methylated, preventing any intramolecular H-

1

4

bonds to BnOÀC(2) also in the inverted C₄conformers, and both diastereoisomers adopt completely the C₁

conformation.

Fig. 8. NOEs between NH or NMe, and glycosylidene H-atoms of the N-methylated mannopyranosylidene

diaziridines 28Se, 28Re, 29Se, and 29Re in C₆D₆solution

In the mannopyranose series, a nearly exclusive ring closure of the mesyloxy amines

obtained by equatorial addition of NH₃(91%) and of MeNH₂(> 95%, limit of

¹H-NMR analysis) was observed. This result agrees with the hypothesis that the amine

adding to to the CN bond forms a H-bond to C(2)ÀOR even in MeOH solution. To

further test the influence of this H-bond, we subjected the 2,3-O-isopropylidene-

protected d-mannono- and d-ribono-1,4-lactone oxime sulfonates 30 and 37 (Schemes 9

and 10) to the action of NH₃and MeNH₂, since formation of a H-bond to C(2)ÀOR is

only possible if the amines add from the sterically disfavoured endo side.

3. d-Pentofuranosylidene-Diaziridines. Reaction of the mannono-1,4-lactone oxime

methanesulfonate 30 (Scheme 9) with NH₃in MeOH at ambient temperature, followed

by crystallisation of the product at 48, gave 70% of 31R; no isomer was detected in the

¹H-NMR spectrum (CDCl₃) [7]. The reaction was repeated, and the product was

analysed in C₆D₆; exclusively 31R was detected. The ribono-1,4-lactone oxime

methanesulfonate 37 (Scheme 10) did not react under these conditions (14 d at

ambient temperature), but was completely consumed within 24 h when the reaction

Helvetica Chimica Acta Vol. 86 (2003)

1503

Scheme 9

was conducted in a closed vessel at a pressure of 5 atm and at ambient temperature.

However, no diaziridine could be isolated; it probably decomposed prior to isolation

[22]. A single diaziridine (d(NH) 2.68and 2.37 ppm, J(NH,NH) 8.0 Hz) was

detected in the crude mixture obtained from the reaction of the corresponding 5-O-

[(phenylamino)carbonyl]ribosylidene methanesulfonate with NH₃, but it decomposed

during attempted purification [23].

The ¹⁵N-labelled hydroximo lactone (Z)-34* (Scheme 9) was prepared from 32

according to [24], but without adapting the reaction conditions to the reduced amount

of ¹⁵NH₂OH ¥ HCl. The hemiacetal 32 was transformed to the (E/Z)-oximes 33* and

then oxidised by MnO₂to (E)-34* and (Z)-34*. Separation of the diastereoisomers by

flash chromatography and crystallisation yielded 35% of (Z)-34* and 5% of (E)-34*.

The minor (E)-34* completely isomerised to (Z)-34* upon standing at room

temperature for 3 d. Mesylation of (Z)-34* gave 30*. It reacted with NH₃in MeOH

to a 8:2 mixture of the isotopomers 31Rx* and 31Rn*¹¹) (57%); repeating the addition

led to a 9 :1 mixture of 31Rx*/31Rn*.

The incorporation of ¹⁵N in (Z)-34* is evidenced by the peak for [M Na] at m/z 297.108in the high-

resolution mass spectrum, the s at À91.1 ppm in the ¹⁵N-NMR spectrum, and by the additional splittings of the

signals of C(1) (¹J(¹⁵N,C) 3.0 Hz) and C(2) (²J(¹⁵N,C) 8.7 Hz). HÀC(2) of (Z)-34* does not show a

³J(¹⁵N,H) coupling, in contradistinction to the pyranose series.

The reaction of 30 with MeNH₂in MeOH, followed by filtration of the crude

through LiChroprep-NH₂, gave 92% of a mixture 35Rx/35Sn, and two unknown

secondary products¹²) (Scheme 10). Their ratio in C₆D₆was 74 :14 :6 :6. The analogous

11

)

The letters x and n denote the exo and endo positions, respectively, of the NR group stemming from the

amine.

The larger values of J(2,3) (8.1 and 7.6 vs. 5.7 5.8Hz) indicate that the secondary products no longer

possess a furanose ring. Four NMe signals at 2.34 2.44 ppm indicate that these minor products are N-

methyl amides (cf. formation of 45 and 46; Scheme 11).

12

1504

Helvetica Chimica Acta Vol. 86 (2003)

Scheme 10

reaction of the ribo methanesulfonate 37 led to the formation of the four possible

isomeric N-Me-diaziridines 38Rn, 38Sn, 38Rx, and 38Sx; their ratio in C₆D₆was

76 :4 :12 :8.

Remarkably, the exo addition is preferred (>4 :1) in the reaction of NH₃and

MeNH₂with the mannono-1,4-lactone oxime methanesulfonate 30, and the endo

addition (4 :1) in the reaction of MeNH₂with the ribono-1,4-lactone oxime

methanesulfonate 37; i.e., both add trans to the substituent at C(4). The stereo-

selectivity of the addition to ^OE conformers is as predicted by the stereoelectronic

effect, favouring exo addition to the mannono-1,4-lactone and endo addition to the

ribono-1,4-lactone derivative¹³). The sterically favoured exo addition to the ribono-1,4-

lactone oxime methanesulfonate is disfavoured both by a pseudoequatorial trajectory

and by an unfavourable interaction of the pseudoaxial (negatively charged) (mesyl-

oxy)imido substituent with OÀC(2). It is not clear to what extent the stereoselectivities

are influenced by different H-bonding to OÀC(2) in the endo addition to the manno-

and ribofuranosylidene derivatives. However, stereoelectronic control, and not H-

bonding, appears to be the decisive factor.

The strikingly different selectivities observed for the mannopyranosylidene and -

furanosylidene derivatives demonstrate the difficulty of predicting the direction of the

addition, and justify a detailed discussion of the configuration of the furanose-derived

diaziridines.

NOE measurements allowed unambiguous assignment of the configuration of the ribose derivatives 38Rn,

38Sn, 38Rx, and 38Sx, and proved helpful for the interpretation of the weak NOEs obtained from the

mannofuranose derivatives 31R and 35Rx. NOEs of 3.4 3.7% were observed between HÀN and HÀC(2) of

38Rn, evidencing the cis-arrangement of H_exoÀN and HÀC(2), and the (R,R)-configuration of the N-atoms

(Fig. 9). Irradiation of NMe of both 38Rn and 38Rx, resonating at 2.79 and 2.81 ppm, respectively, led to signal

enhancements of 1.05 and 0.4% for HÀC(2) and HÀC(3) only of 38Rx. Considering the 76 :12 ratio of 38Rn/

13

)

For similar endo vs. exo selectivity in the radical transfer to 2,3-O-isopropylidene-d-mannofuranose and -

d-ribofuranose derivatives, see [25].

Helvetica Chimica Acta Vol. 86 (2003)

1505

38Rx and the fact that the NOE enhancement is related to the sum of the NMe integrals of 38Rn/38Rx, one

calculates NOEs of 7 and 2.8% for HÀC(2) and HÀC(3) of 38Rx, evidencing the cis-arrangement of MeN,

HÀC(2), and HÀC(3), and the (R,R)-configuration. This conclusion is corroborated by a NOE of 3.6% for

NMe of 38Rx upon irradiation of HÀC(2). Irradiation of NMe of 38Sx at 2.54 ppm led to signal enhancements

of 1.0 and 0.7% for HÀC(2) and HÀC(3) of 38Rx, respectively, evidencing isomerisation of 38Sx into 38Rx.

This indicates the cis-arrangement of NMe, HÀC(2), and HÀC(3), and the (S,S)-configuration of 38Sx. Not

surprisingly, no NOE could be observed upon irradiation of NH and NMe of 38Sn.

Fig. 9. NOEs between NH or NMe, and glycosylidene H-atoms of the furanosylidene diaziridines 38Rn, 38Sn,

38Rx, 38Sx, 31R, 35Rx, and 35Sn in C₆D₆solution. The structures 31S, 35Sx, and 35Rn, diastereoisomers of

31R, 35Rx, and 35Sn, respectively, have been calculated.

In the mannofuranose series (Fig. 9), weak NOEs were observed upon irradiating the low-field NH of the

N-unsubstituted diaziridine 31 at 2.41 ppm and the NH of the major N-Me-diaziridine (35Rx) at 2.85 ppm,

whereas irradiations of the high-field NH of 31 at 2.01 ppm, and of the NMe of 35Rx at 2.46 ppm did not lead to

intensity enhancements for the corresponding HÀC(2) signals. These NOE measurements evidence that

HÀC(2) is on the same side of the diaziridine ring as the low-field NH of 31, and as NH of 35Rx; the assignment

of the structure 31A to 31 and of 35Rx to the major isomer of 35 was based on the NOE intensities and on the

1506

Helvetica Chimica Acta Vol. 86 (2003)

HÀN chemical-shift values¹⁴). The enhancements (1 1.6%) are distinctly smaller than the one observed for

HÀC(2) of 38Rn (3.7%), suggesting a trans-arrangement (relative to the furanose ring) of HÀN and HÀC(2)

of 31R and 35Rx, and thus the (R,R)-configuration at the N-atoms. This interpretation is confirmed by the small

enhancement (1.8%) observed for HÀC(2) of 24R upon irradiating a similarly trans-oriented HÀN (see Fig. 7).

The configurational assignment of 31R and 35Rx is corroborated by the HÀN chemical-shift values.

The N-Me group of the diaziridines leads to a downfield shift for the adjacent HÀN. For pyranosylidene

diaziridines, a downfield shift of 0.31 to 0.36 ppm is observed upon formal N-methylation of 2S to 17Se, 2R to

17Ra, 10S to 18Se, 10R to 18Ra, 20S to 28Se, 20R to 28Re, and 24S to 29Se, with one exception, viz. formal N-

methylation of the isopropylidene-protected 24R to 29Re, which leads only to a downfield shift of 0.15 ppm.

The shift difference between HÀN of 35Rx, resonating at 2.85 ppm, and the high-field HÀN of 31R, resonating

at 2.01 ppm, is 0.84 ppm, and thus too large to be due to the N-methylation. It evidences that HÀN of 35Rx

corresponds to the low-field HÀN of 31R, resonating at 2.41 ppm. The shift difference of 0.44 ppm is slightly

larger than that observed in the pyranose series, perhaps indicating a larger downfield shift upon N-methylation

of furanosylidene diaziridines. This value was used to calculate the d(NH) values of the N-unsubstituted manno-

diaziridines 31R and 31S from the d(NH) values of the ribo-N-Me-diaziridines 38. The calculated d values for

H_exoÀN and H_endoÀN of 31R are 1.94 and 2.29 ppm, respectively, and those for H_exoÀN and H_endoÀN of 31S 1.89

and 1.99 ppm, respectively. The experimental d values of 31 (2.01 and 2.41 ppm) agree distinctly better with the

calculated values for 31R, evidencing the (R,R)-configuration of the N-atoms. d(NH) and d(NMe) of the major

manno-N-Me-diaziridine 35Rx agree only with those of 38Sx, confirming the configurational assignment of

35Rx. The configuration of the minor N-Me-diaziridine 35Sn cannot be determined by such a comparison, since

HÀN and MeN of 38Rn, 38Sn, and 38Rx resonate in the narrow range of 2.33 2.43 and 2.74 2.81 ppm,

respectively. The configurational assignment of 35Sn is based on the observation that 31 completely favours the

(R,R) configuration and on the strong preference of 38Rn over 38Sn.

The strong preference for 31R and 35Rx in solution agrees with AM1 gas-phase

calculations. Compounds 31R and 35Rx are favoured by 1.5 2.1 kcal/mol over 31S and

35Sx, respectively, due to a weak intramolecular H-bond NÀH_endo¥¥¥ OC(2) (d(H_endo¥¥¥

O) 2.63 ä; a(NÀH_endo¥¥¥ O) 988). According to the calculations, the conforma-

tion of the furanose ring is determined by the configuration of C(4) and the annulation

of the 1,3-dioxolane ring; the configuration of the N-atoms is irrelevant. All the

furanosylidene diaziridines mentioned above adopt a southern-type conformation

O

close to a (flat) E.

The isopropylidene groups enhance the stability of the furanosylidene-diaziridines

31, 35, and 38 relative to the corresponding benzylated analogues derived from 42 and

48 (Scheme 11).

The O-benzylated ribono-1,4-lactone oxime methanesulfonate 42 was prepared in

the usual way from 2,3,5-tri-O-benzyl-d-ribofuranose (39) [26][27] by oxime formation

to 40, oxidation to 41, and mesylation. The arabino-methanesulfonate 48 was obtained

from the known hydroximo lactone 47 [28]. Treatment of 42 with NH₃in MeOH gave

the crystalline 1,4-dihydro-1,2,4,5-tetrazine 43 (21%), the ribonolactone 44 [29 31]

(10%), and the hydroxyribonamide 45 (25%). The analogous reaction of 48 afforded

the crystalline 1,4-dihydro-1,2,4,5-tetrazine 49 (21%), the arabinolactone 50 (23%),

and a mixture that was not analysed (ca. 12%). Only the ribonolactone 44 (44%) and

the N-Me-ribonamide 46 (17%) were isolated from the reaction of 42 with MeNH₂in

MeOH.

14

)

Signal overlap and the presence of two side products prevented NOE analysis of minor N-Me-diaziridine

(35Sn).

Helvetica Chimica Acta Vol. 86 (2003)

1507

Scheme 11

OBn

O

OBn

OH

OBn

OH

O

N

BnO OBn

39

40

41

N

NH

BnO

OBn

O

OBn

O

OMs

HN

N

NH₃or

MeNH₂

O

OH

N

+

NHR

OH HO

BnO OBn

OBn BnO

42

43

44

45 R = H

46 R = Me

N

NH

N

BnO

OBn

NH₃

OBn

HN

OBn

OR

O

N

O

OH HO

BnO

OBn BnO

49

50

47 R = H

48 R = Ms

1

A single set of H- and ¹³C-NMR signals reveals the C₂symmetry of 43 and 49. Their 1,4-dihydrotetrazine

structure¹⁵) was evidenced by the [M Na] and [M K] peaks at m/z 887 and 903, respectively, in the mass

spectrum, and the NÀH band at 3250 3260 cm^À1in the IR spectrum (KBr). The OH groups of 43 and 49

resonate in CDCl₃as ds at 2.73 and 2.51 ppm, respectively (cf. Table 8 in Exper. Part). The relatively large

J(H,OH) value of 5.7 and 7.2 Hz indicate intramolecular H-bonds. The HÀN and Ph groups resonate at 7.21

7.34 ppm. The ss for C(3) and C(6) of 43 and 49 appear at 148.3 and 149.5 ppm, respectively (cf. Table 9 in Exper.

Part), these are shift values typical of 1,4-dihydro-1,2,4,5-tetrazines [34 36]. The hydroxy-amide structure of 45

and 46 is revealed by the downfield shift of HÀN (45: 5.49 and 6.65; 46: 6.71 ppm) and C(1) (45: 173.9; 46:

171.2 ppm) and by OÀH, NÀH, and CO bands (45: 3510, 3400, and 1680; 46: 3560, 3430, and 1660 cm^À1).

The 1,4-dihydro-1,2,4,5-tetrazines 43 and 49 derive from the intermediate furano-

sylidene diaziridines¹⁶). Ring opening of the diaziridines obtained by the reaction of 42

and 48 with NH₃or MeNH₂is assumed to lead to lactone hydrazones. These

hydrazones, in part, dimerise (via azomethine imines?) to the 1,4-dihydro-1,2,4,5-

tetrazines 43 and 49, and they also react with NH₃or MeNH₂to generate, via imino

ethers, the lactones 44 and 50, and the hydroxy amides 45 and 46¹⁷).

We thank Matthias Bˆhm, Lukas Frunz, and Patricia Nickut for the preparation of the ¹⁵N-labelled

mannose derivatives, Dr. A. Linden, University Z¸rich, for the X-ray analysis, and the Swiss National Science

Foundation and F. Hoffmann-La Roche AG, Basel, for generous support.

The 1,4-dihydro-1,2,4,5-tetrazines adopt a flattened ^3,6B conformation with the 3- and 6-substituents in

bowsprit position [32]. A 1,2-dihydro-1,2,4,5-tetrazine structure is excluded, since 1,2-dihydro-1,2,4,5-

tetrazines rearrange readily to the more-stable 1,4-dihydro-1,2,4,5-tetrazines [33].

15

)

16

17

)

For the reaction of hexono-1,5-lactone hydrazone to a 4-amino-4H-1,2,4-triazole and a 1,2,4,5-tetrazine via

a 1,2-dihydro-1,2,4,5-tetrazine, see [37].

We cannot exclude the direct transformation of 42 and 48 into imino ethers although lactone oxime

sulfonates that yield stable diaziridines did not give rise to either lactones or open-chain amides.

1508

Helvetica Chimica Acta Vol. 86 (2003)

Experimental Part

General. See [29]. Sat. solns. of NH₃in MeOH were prepared by bubbling NH₃through cooled (08) MeOH.

Normal workup means concentrating below 308 in a B¸chi rotary evaporator, dissolving the residue in Et₂O,

filtering through LiChroprep-NH₂, concentrating, and drying under high vacuum at a pressure below 0.1 mbar

at 208. NMR Spectra: chemical shifts d in ppm relative to TMS (¹H- and ¹³C-NMR) or MeNO₂(¹⁵N-NMR) as an

internal standard, coupling constants J in Hz.

General Procedure for the Preparation of N-Methyldiaziridines. A soln. of the methansulfonate in 7.04m

MeNH₂in anh. MeOH was stirred at r.t. for the indicated period and evaporated. The residue was dissolved in

Et₂O and filtered through LiChroprep-NH₂. Evaporation and drying for 5 h at r.t. gave a mixture of the

diaziridines.

(1'S,2'S)- and (1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-hydrazi-d-glucitol (2S/2R). According to [7].

¹H-NMR (400 MHz, C₆D₆, 2S/2R 95 :5): Table 2; additionally for 2S, 7.30 7.06 (m, 20 arom. H); 4.88 (d, J

11.0), 4.86 (d, J 10.2), 4.77 (d, J 10.8), 4.72 (d, J 11.4), 4.62 (d, J 11.4,), 4.46 (d, J 10.9), 4.38( d, J

12.0), 4.28( d, J 12.0) (8PhC H). ¹³C-NMR (50.3 MHz, C₆D₆, only signals of 2S present): Table 3; additionally,

139.39, 139.15, 138.71, 138.48 (4s); 128.50 127.38 (several d); 75.61 (t, 2 PhCH₂); 75.02, 73.55 (2t, 2 PhCH₂).

(1'S,2'S)- and (1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-hydrazi-d-galactitol (10S/10R). According to

[7]. ¹H-NMR (400 MHz, C₆D₆, 10S/10R 95 :5): Table 2; additionally for 10S, 7.37 7.04 (m, 20 arom. H); 5.03 (d,

J 11.2), 4.74 (d, J 10.8), 4.59 (d, J 11.6), 4.54 (d, J 11.9), 4.47 (d, J 10.8), 4.40 (d, J 11.9), 4.28( d, J

11.8), 4.21 (d, J 11.8) (8 PhCH). ¹H-NMR (400 MHz, CDCl₃, 10S/10R 9 :1): Table 2 and [7]. ¹³C-NMR

(50.3 MHz, C₆D₆, only signals of 10S visible): Table 3; additionally, 138.84, 138.64, 138.28, 138.08 (4s); 128.12

127.12 (several d); 75.41, 74.90, 73.06, 72.82 (4t, 4 PhCH₂).

2,3,4,6-Tetra-O-benzyl-d-(¹⁵N)gluconhydroximo-1,5-lactone (15*). A soln. of EtONa in EtOH (33.55 ml;

1.15 g of Na dissolved in 250 ml of EtOH) was treated with a soln. of ¹⁵NH₂OH ¥ HCl (Cambridge Isotope

Laboratories, >98% of ¹⁵N; 505 mg, 7.17 mmol) in dry MeOH (35 ml), stirred for 5 min, treated with 2,3,4,6-

tetra-O-benzyl-d-glucopyranose (3 g, 5.56 mmol), and stirred at reflux for 35 h. The soln. was cooled to r.t.,

diluted with H₂O (100 ml), and extracted with CH₂Cl₂(3 Â 75 ml). Drying the combined org. layers (MgSO₄)

and evaporation gave crude (E/Z)-2,3,4,6-tetra-O-benzyl-d-(¹⁵N)glucose oxime (3 g, 97%), which was dissolved

in dry MeOH (24 ml), treated with MnO₂(prepared according to [38]; 1.33 g, 16.3 mmol), and stirred at reflux

for 6 h. After filtration through Celite (washing the residue with warm MeOH), evaporation and FC (hexane/

AcOEt 7:3) gave crystalline 15* (1.83 g, 61%), which was recrystallized in AcOEt/hexane 3 :4 (7 ml). R_f

(hexane/AcOEt 3 :2) 0.4. [a]²_D⁵ 44.1 (c 1.0, CHCl₃). M.p. 878. IR (CHCl₃): 3580m, 3450w, 3360w (br.),

3090w, 3060w, 3030w (sh.), 3000m, 2920m, 2860m, 1970w (sh.), 1950w, 1875w, 1810w, 1780w (br.), 1655m (sh.),

1645m, 1635m (sh.), 1620w, 1605w (sh.), 1585w, 1575w (sh.), 1490w, 1450m, 1360m, 1280m, 1260m, 1080s (sh.),

1070s, 1025s, 995m (sh.), 910s, 8 50w. ¹H-NMR (400 MHz, CDCl₃): 7.37 7.23 (m, 18arom. H); 7.18 7.15 ( m, 2

arom. H); 7.05 (d, ²J(H,N) 0.8, exchange with D₂O, OH); 4.73 (d, J 11.9, PhCH); 4.65 (d, J 12.2, PhCH);

4.60 4.59 (m, 2 PhCH); 4.59 (ddd, J 10.1, 4.2, 2.0, HÀC(5)); 4.55 (d, J 11.7), 4.49 (d, J 12.1), 4.47 (d, J

3

11.5), 4.37 (d, J 11.7) (4 PhCH); 4.11 (t, J ꢂ J(H,N) ꢂ 1.5, HÀC(2)); 3.93 (ddd, J 4.5, 2.1, ⁴J(H,N) 1.6,

HÀC(3)); 3.85 (dd, J 11.3, 2.1, HÀC(6)); 3.82 (dd, J 10.1, 4.3, H À C(4)); 3.78( dd, J 11.4, 4.2, H'ÀC(6)).

¹³C-NMR (50.3 MHz, CDCl₃): 151.21 (s, C(1)); 137.90, 137.62, 137.13, 137.05 (4s); 128.55 127.49 (several d);

81.28 (dd, ³J(C,N) 1.9, C(3)); 77.46 (d, C(4)); 75.97 (d, C(5)); 73.43 (t, PhCH₂); 73.09 (dd, ²J(C,N) 10.7,

C(2)); 72.89, 71.54, 70.48 (3t, 3 PhCH₂); 68.10 (t, C(6)). ¹⁵N-NMR (40.6 MHz, CDCl₃): À65.80 (br. d, J 2.2).

Anal. calc. for C₃₄H₃₅¹⁵NO₆(554.66): C 73.63, H 6.36, N 2.70; found: C 73.48, H 6.46, N 2.54.

[2,3,4,6-Tetra-O-benzyl-d-(¹⁵N)glucopyranosylidene]amino Methanesulfonate (1*). A soln. of 15* (1 g,

1.8mmol) in dry CH ₂Cl₂(20 ml) was treated dropwise at 08 with Et₃N (0.6 ml, 4.3 mmol) and then slowly with

MsCl (0.15 ml, 1.96 mmol). The clear soln. was stirred for 30 min, washed with 1m NaHCO₃soln. (2 Â 15 ml) and

H₂O (3 Â 30 ml), dried (MgSO₄), and evaporated. Recrystallizion in Et₂O gave 1* (638mg, 56%). M.p. 63 8. R_f

(hexane/AcOEt 3 :2) 0.55. [a]²_D⁵ 38.8 (c 1.1, CHCl₃). IR (CHCl₃): 3090w, 3060w, 3030w (sh.), 3010w,

2920w (br.), 2870w, 1970w, 1950w, 1870w, 1810w, 1755w, 1635m, 1620w (sh.), 1605w (sh.), 1585w (sh.), 1555w,

1490w, 1450m, 1370s, 1325m, 1290m, 1260m, 1175m, 1095s (sh.), 1070s, 1025m, 1005m (sh.), 1000m (sh.), 965s,

955w, 910w, 8 35s, 8 25s (sh.). ¹H-NMR (400 MHz, CDCl₃): 7.40 7.26 (m, 18arom. H); 7.23 7.16 ( m, 2 arom. H);

4.74 (d, J 12.0, PhCH); 4.65 (d, J 12.4, PhCH); 4.67 4.62 (m, HÀC(5)); 4.59 (d, J 12.3), 4.55 (d, J 12.0),

3

4.53 (d, J 11.8), 4.49 (d, J 11.7), 4.48( d, J 11.9), 4.35 (d, J 11.8) (6 PhCH); 4.16 (t, J ꢂ J(H,N) ꢂ 1.6,

4

HÀC(2)); 3.95 (ddd, J 4.1, 2.0, J(H,N) 1.0, HÀC(3)); 3.86 (dd, J 10.2, 3.9, HÀC(4)); 3.83 (dd, J 11.6,

2.1, HÀC(6)); 3.76 (dd, J 11.6, 3.9, H'ÀC(6)); 3.12 (s, MsO). ¹³C-NMR (50.3 MHz, CDCl₃): 157.32 (s, C(1));

137.74, 137.33, 136.76, 136.29 (4s); 128.66 127.68 (several d); 80.41 (dd, ³J(C,N) 1.7, C(3)); 77.40 (d, C(4));

Helvetica Chimica Acta Vol. 86 (2003)

1509

Table 2. Selected ¹H-NMRChemical Shifts [ppm] and Coupling Constants [Hz] of the Gluco-, Galacto-, and Allopyranosylidene Diaziridines 2, 4, 6, 8,

10, 12, 14, 17, and 18

a

2S/2R [7]

2S/2R

2Se*/2Sa*

4S/4R [5]

7897 :

6S [8]

8S/8R ) [5]

Ratio

95 :

5

b

95 :

C₆D₆

5

b

25 :

75

22 :

C₆D₆

3

> 97%

CDCl₃

94 :

6

Solvent CDCl₃

CDCl₃

C₆D₆

CDCl₃

HÀC(2) 4.11

HÀC(3) 3.68

HÀC(4) 3.91

HÀC(5) 3.82 3.78

HÀC(6) 3.78

)

3.93

3.59

3.94

3.88

3.66

3.56

2.681.79

2.26

9.4

9.1

9.9

3.4

1.7

)

4.11

3.69

3.92

3.82 3.78

3.783.67

3.683.57

3.93

3.60

3.95

3.89

5.87

5.49

5.47

3.89

5.61 5.68 5.63 4.11

3.92

b

)

5.69 5.31 5.26 3.83 3.78 4.13

5.48 5.31 5.26 3.83 3.78 3.85

3.15 4.06

3.83 3.78 3.78 3.74

44..017854.343.28

4.07

4.30

H'ÀC(6) 3.68

3.97 4.11

3.74

3.73

2.84

c

d

H_aN )

2.66

2.36

9.4

1.95

2.32

2.67 (57.9/3.0) 2.69 (57.0/2.9)

2.37 (3.7/57.7) 2.26 (3.7/57.4)

2.13

)

1.74 2.47 )

2.04

2.34

9.0

c

d

H_eN )

2.53

2.10 )

2.22 2.40 )

2.32

8.1

b

J(2,3)

J(3,4)

J(4,5)

J(5,6)

J(5,6')

)

9.5

9.1

9.810.0

3.0

2.81.7

9.4

9.1

9.3

9.5

9.8

1.84.6

1.98.89.8

12.6

9.3

9.5

)

b

9.1

9.9

2.8

)

b

)

9.9

)

b

)

3.4

4.1

4.0

4.4

b

)

4.6 2.2

11.0

9.4/9.4

b

J(6,6') 10.7

J(H_a,H_e) 9.4

)

11.0

9.4

)

10.6

9.4/9.4

12.5 12.810.6

9.4 9.4

10.6

9.4

9.5

5

9.4

a

d

10S/10R [7]

95 :

10S/10R

10Se*/10Sa*

12S/12R )^f) [5] 14S/14R ) [13]

Ratio

5

95 :

15 :

85

25 :

75 15 :

85

15 :

85

Solvent CDCl₃

C₆D₆

CDCl₃

C₆D₆

CDCl₃

b

HÀC(2) 4.50

HÀC(3) 3.64

HÀC(4) 4.07

HÀC(5) 3.96

HÀC(6) 3.58

H'ÀC(6) 3.54

)

4.59

3.39

3.92

4.03

3.76

3.61

1.77

2.24

9.8

2.9

1.2

7.8

5.6

)

4.50

3.64

4.083.92

3.97

3.583.76

3.54

4.59

3.39

4.39

4.6843.8

3.62 4.19

4.70

b

)

3.87

3.93 3.91 8 3.93 3.8

4.11

4.22

b

)

b

)

4.04

3.62 4.47

4.33

3.81

2.08 )

2.15 )

b

)

2

33..784 4.348.38

b

)

3.62

3.62 3.66 3.665 3.77

d

e

H_aN

2.681.91

2.75

(526..698/2.9) 2).76/2(.285.6.28

2.14 2.18

)

e

H_eN

2.25

9.9

2.7

1.1

7.6

5.9

9.2

2.41

2.60

2.24 (3.7/57.6) 2.27 (3.6/57.3)

9.9

2.82.9

1.2

7.6

5.9

9.1

9.4/9.4

2.36

9.43.2 2.8

2.36 2.07 )

b

J(2,3)

J(3,4)

J(4,5)

J(5,6)

J(5,6')

J(6,6')

)

9.87.6

5.5

1.2

7.84.5

5.6

9.0

b

)

9.5

2.0

2.2

9.89.3

5.2

b

)

9.3

b

)

3.9

5.0

b

)

< 1.0

10.811.8

9.1

9.89.8

11.8

8.8

b

)

9.0

9.4

)

10.4

9.3

J(H_a,H_e) 9.4

17Se/17Ra

72 :

9.4

28

9.4

9.4/9.4

9.3

a

17Se*/17Ra*

18Se/18Ra )^g) 18Se*/18Ra*

Ratio

72 :

28

85 :

15

85 :

15

Solvent C₆D₆

C₆D₆

HÀC(2) 3.85

HÀC(3) 3.60

HÀC(4) 3.95

HÀC(5) 3.89

HÀC(6) 3.69

3.73

4.01

3.94

3.85

3.60

3.95

3.73

4.01

3.94

4.49 4.41

3.383.78

3.93 3.93

4.51

4.41

3.93

3.74 3.70

3.39

13.8

3.93

3.64 3.59 3.89

3.66 3.69

3.64 3.59 3.62

3.02 (57.5)

2.84

2.65

9.4

3.64 3.59 4.03 3.75 3.72 4.03

3.66 3.79 3.75 3.72 3.80

3.64 3.59 3.65 3.75 3.72 3.65

3.08 3.09 (57.4)

H'ÀC(6) 3.62

c

H_aN )

3.02

c

H_eN )

2.84 (58.2)

2.75 (2.7) 2.65 (2.9) 2.75 2.682.74 (2.7) (22.7.6)8

9.4

9.0

10.0

3.6

1.7

2.92

2.91 (58.3)

c

MeN )

2.75

9.4

9.0

10.0

3.6

1.7

J(2,3)

J(3,4)

J(4,5)

J(5,6)

J(5,6')

9.4

9.0

9.9

3.0

1.0

7.7

5.7

9.0

9.89.8 9.8

9.0

9.9

3.0

1.0

2.9

1.0

7.8

5.7

9.0

3.0

1.0

b

4.0

)

b

)

b

J(6,6') 11.0

10.5

11.0

10.5

)

a

c

d

) Assignment based on a H,¹³C-COSY spectrum. ^b) Not assigned. ) In parentheses, J(H,¹⁵N) or J(Me,¹⁵N).

)

^e) Tentative

1

f

assignment, may be interchanged. ) Signals for AcNH: 12S: 6.37 (d, J 8.2), 1.83 (s); 12R: 5.53 (d, J 9.4), 1.81 (s); 14S: 5.53

g

(d, J 8.5), 1.78 (s); 14R: 5.91 (d, J 9.4), 1.79 (s). ) Assignment based on a H,¹H-COSY spectrum.

1

1510

Helvetica Chimica Acta Vol. 86 (2003)

Table 3. Selected ¹³C-NMRChemical Shifts [ppm] of the Gluco- and Galactopyranosylidene-diaziridines 2, 10,

17, and 18 (J(C,¹⁵N) in parentheses)

2S [7]

2S

2Se*/2Sa* 2Se*/2Sa* 10S^a) [7] 10S

10Se*/10Sa* 10Se*/10Sa*

Solvent CDCl₃C₆D₆

CDCl₃

C₆D₆

CDCl₃

C₆D₆

CDCl₃

C₆D₆

C(1)

C(2)

C(3)

C(4)

C(5)

C(6)

82.97

76.53

84.29

76.53

77.00

67.77

83.46

77.15

84.68

77.15

77.64

68.72

82.97 (5.1) 83.44 (5.9) 83.28

76.62 (br.) 77.15 (3.5) 74.14

83.46

74.44

81.63

74.23

75.33

68.14

83.30 (5.5)

74.18(3.9)

81.60

83.91 (6.1)

74.93 (br.)

82.10

84.32

76.56

77.00

67.84

84.70

77.1874.17

77.66

81.55

74.77

75.42

67.79

75.30

75.46

67.83

75.42

68.66

68.75

17Se

17Se*

17Ra

17Ra*

18Se^a)

18Se*

18Ra^a)

18Ra*

Solvent C₆D₆

C₆D₆

C(1)

C(2)

C(3)

C(4)

C(5)

C(6)

MeN

85.90

85.94 (4.1) 86.08

86.14

86.37

75.25

82.20

75.25

75.99

68.57

86.36 (4.3) 86.49

86.45

77.51^b) 77.59^b)

77.36^d)

85.47

77.52^b)

85.55

75.27

82.25

75.33

76.02

68.61

974.7874.8

84.71

84.79

83.07

75.25

75.66

69.05

83.06

75.17

75.71

69.08

77.29^b) 77.37^b)

76.80^d)

78.28^d)

68.79

76.90^b)

78.38^b)

68.86

77.74^b) 77.80^b)

68.49

39.11

68.57

39.14 (3.7) 38.32

38.33 (3.9) 39.03

38.99 (3.6) 38.52

38.52 (4.1)

a

) Assignment based upon a H,¹³C-HMQC spectrum. ^b) Assignments may be interchanged.

1

76.87 (d, C(5)); 73.33, 72.98(2 t, 2 PhCH₂); 72.31 (dd, ²J(C,N) 11.4, C(2)); 71.63, 71.07 (2t, 2 PhCH₂); 67.35

(t, C(6)); 36.07 (q, MsO). ¹⁵N-NMR (40.6 MHz, CDCl₃): À60.67 (br. d, J 1.7). Anal. calc. for C₃₅H₃₇¹⁵NO₈S

(632.75): C 66.44, H 5.89, N 2.37, S 5.07; found: C 66.44, H 6.10, N 2.30, S 4.89.

(1R,1'S,2'S)- and (1S,1'S,2'S)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-hydrazi-d-(¹⁵N)glucitol (2Se*/2Sa*).

After dissolution of 1* (400 mg, 0.632 mmol) in a sat. soln. of NH₃in MeOH (7.2 ml) and stirring for 30 h at

r.t., the soln. was half-concentrated and then cooled at À158 overnight. The crystals were filtered off and dried

under high vacuum to give 2Se*/2Sa* (271 mg, 77%). Evaporation of the mother liquor, filtration through

LiChroprep-NH₂(Et₂O), evaporation, and crystallisation from MeOH gave an additional crop of 2Se*/2Sa*

(ca. 7%). R_f(hexane/AcOEt 1:1) 0.49. M.p. 50 518. [a]²_D⁵ 44.1 (c 1.1, CHCl₃). IR (CHCl₃): 3270w, 3090w,

3060w, 3030w (sh.), 3000w, 2960w, 2910m, 2870m, 1955w, 1875w, 1810w, 1605w, 1495w, 1455m, 1400w, 1360m,

1320w, 1285m (sh.), 1260m, 1145m, 1125s, 1085s (br.), 1035s, 1025s, 1015m, 1005m, 950w, 910w, 8 8w5, 8 60w.

¹H-NMR (600 MHz, C₆D₆, 2Sa*/2Se* 78:22): Table 2; additionally, 7.30 7.06 (m, 20 arom. H); 4.89 (d, J

11.3), 4.86 (d, J 11.4), 4.77 (d, J 10.8), 4.73 (d, J 11.3), 4.63 (d, J 11.3), 4.48( d, J 10.9), 4.39 (d, J

1

12.0), 4.29 (d, J 12.0) (8PhC H). H-NMR (400 MHz, CDCl₃, 2Sa*/2Se* 75 :25): Table 2; additionally, 7.37

7.26 (m, 18arom. H); 7.19 7.13 ( m, 2 arom. H); 4.92 (d, J 10.9), 4.87 (d, J 10.7), 4.86 (d, J 10.8), 4.81

(d, J 10.9), 4.68( d, J 10.7), 4.64 (d, J 12.1), 4.55 (d, J 10.6), 4.48( d, J 12.1) (8PhC H). ¹³C-NMR

(150.9 MHz, C₆D₆, only one set of signals for 2Se*/2Sa*): Table 3; additionally, 139.39, 139.15, 138.71, 138.44

(4s); 128.59 127.63 (several d); 75.62, 75.09, 75.03, 73.57 (4t, 4 PhCH₂). ¹³C-NMR (100.6 MHz, CDCl₃, only one

set of signals for 2Se*/2Sa*): Table 3; additionally, 138.39, 138.02, 137.70, 137.64 (4s); 128.36 127.63 (several d);

75.69, 75.52, 75.06, 73.52 (4t, 4 PhCH₂). ¹⁵N-NMR (60.8MHz, C ₆D₆, 2Sa*/2Se* 75 :25): Table 4. ¹⁵N-NMR

(40.6 MHz, CDCl₃, 2Sa*/2Se* 75:25): Table 4.

2,3,4,6-Tetra-O-benzyl-d-(¹⁵N)galactonhydroximo-1,5-lactone (16*). A soln. of EtONa in EtOH (33.55 ml;

1.15 g of Na dissolved in 250 ml of EtOH) was treated with a soln. of ¹⁵NH₂OH ¥ HCl (505 mg, 7.17 mmol) in dry

MeOH (35 ml), stirred for 5 min, treated with 2,3,4,6-tetra-O-benzyl-d-galactopyranose [26][39] (3 g,

5.56 mmol), and stirred at reflux for 35 h. The soln. was cooled to r.t., diluted with H₂O (100 ml), and

extracted with CH₂Cl₂(3 Â 75 ml). Drying the combined org. layers (MgSO₄) and evaporation gave crude (E/

Z)-2,3,4,6-tetra-O-benzyl-d-galactose (¹⁵N)oxime (3 g, 97%), which was dissolved in dry MeOH (24 ml),

treated with MnO₂(1.33 g, 16.3 mmol), and stirred at reflux for 6 h. After filtration through Celite (washing the

residue with warm MeOH), evaporation and FC (hexane/AcOEt 7 :3) gave crude 16* (2.5 g, 81%) as a pale

yellow oil. An additional FC (hexane/AcOEt 85 :15) afforded 16* (1.85 g, 62%) as a colourless oil. IR (CHCl₃):

Helvetica Chimica Acta Vol. 86 (2003)

1511

Table 4. ¹⁵N-NMR( 40.6 MHz) Chemical Shifts [ppm] of the ¹⁵N-Labelled Diaziridines 2*, 10*, 17*, and 18*, and of the

Unlabelled Diaziridines 20, 28, 29, and 38

2Se*/2Sa*

10Se*/10Sa*

17Se*/17Ra*

18Se*/18Ra*

Ratio

25 :

75

25 :

75

15 :

85

17:

83

72 :

28

85 :

15

Solvent CDCl₃

C₆D₆

CDCl₃

C₆D₆

d(N)

À 317.0^a) À 306.9^a) À 312.8 À 304.4 À 314.4 À 304.0 À 313.4 À 304.2 À 297.3 À 297.5 À 298.1 À 299.6

¹J(N,H)

²J(N,H)

³J(N,Me)

58.0

3.6

57.8

3.0

57.1

3.6

57.5

2.7

57.7

2.9

57.7

2.9

56.5

3.0

57.2

3.0

57.5

58.4

57.2

58.2

2.9

3.0

2.6

2.8

20S/20R

28Se/28Re

29Se/29Re

38Rn/38Rx

Ratio

Solvent CDCl₃

55 :

45

85 :

C₆D₆

15 80 :

C₆D₆

20 85 :

C₆D₆

15

d(HN)

À 292.0 to À 290.5

À 294.2 À 294.2 À 291.7 À 275.9 À 284.7 À 275.3

57.1 57.1 59.1 58.0 54.7 59.6

À 293.8 296.7 À 296.2 À 297.9 À 300.4 À 301.9

b

¹J(N,H)

d(HN)

)

c

¹J(N,Me)

)

^a) This spectrum was recorded at SF parameter differing by 0.001 MHz, leading to a shift difference of 24.7 ppm.

The original d values were corrected by this value. Since also a different SR parameter was used, the corrected

shift values may still not be accurate. ^b) Not assigned. ^c) Only line broadening.

3580m, 3350m, 3050m, 2870s, 1960w, 1845w, 1805w, 1645m, 1605m, 1445m, 1350s, 1260 1200s, 1150 1000s,

950 900s, 8 50m. ¹H-NMR (400 MHz, CDCl₃): 7.66 (br. s, NOH); 7.36 7.25 (m, 20 arom. H); 4.78( d, J 11.6),

4.76 (d, J 11.6), 4.67 (d, J 12.0), 4.57 (d, J 12.0), 4.56 (d, J 11.6) (5 PhCH); 4.56 4.45 (m, 3 PhCH,

HÀC(5)); 4.30 (dd, J 5.2, ³J(H,N) 1.2, HÀC(2)); 4.19 (t, J 3.3, HÀC(4)); 3.89 (dd, J 5.2, 3.0, HÀC(3));

3.83 (dd, J 10.4, 7.2, HÀC(6)); 3.77 (dd, J 10.4, 5.2, H'ÀC(6)). ¹³C-NMR (100.6 MHz, CDCl₃): 151.51

(d, ¹J(C,N) 0.8, C(1)); 137.81, 137.68 (2 C), 137.40 (3s); 128.47 127.54 (several d); 78.33 (d, C(4)); 78.23

(dd, ³J(C,N) 2.1, C(3)); 74.16 (dd, ²J(C,N) 9.5, C(2)); 73.47, 73.31 (2t, 2 PhCH₂); 72.49 (d, C(5)); 72.46, 71.72

(2t, 2 PhCH₂); 68.53 (t, C(6)). ¹⁵N-NMR (40.6 MHz, CDCl₃): À77.91 (dd, J 1.2, 0.2).

[2,3,4,6-Tetra-O-benzyl-d-(¹⁵N)galactopyranosylidene]amino Methanesulfonate (9*). At 08, a soln. of 16*

(1 g, 1.8mmol) in dry CH ₂Cl₂(20 ml) was treated dropwise with Et₃N (0.6 ml, 4.3 mmol) and then with MsCl

(150 ml, 1.96 mmol). The clear soln. was stirred for 30 min, washed with 1m NaHCO₃soln. (2 Â 15 ml) and H₂O

(3 Â 30 ml). Drying (MgSO₄), evaporation, and recrystallisation in Et₂O gave 9* (991 mg, 87%). M.p. 768. R_f

(hexane/AcOEt 3 :2) 0.55. [a]²_D⁵ 11.7 (c 1.0, CHCl₃). IR (CHCl₃): 3000w, 2950m, 2930m, 2870m, 1720w,

1

1615m, 1495w, 1450w, 1360m, 1325m, 1170m, 1100 1000s, 970s, 8 20s. H-NMR (400 MHz, CDCl₃): 7.38 7.21

(m, 20 arom. H); 4.86 (d, J 11.3), 4.85 (d, J 11.5), 4.64 (d, J 12.0), 4.59 (d, J 11.4), 4.56 (d,J 12.0), 4.54

(d, J 11.3), 4.53 (d, J 12.0), 4.47 (d, J 11.9) (8PhC H); 4.45 4.41 (m, HÀC(5)); 4.38( dd, J 5.0,

³J(H,N) 1.3, HÀC(2)); 4.14 (dd, J 3.2, 1.8, HÀC(4)); 3.88 (dd, J 5.0, 3.2, HÀC(3)); 3.73 (AB, 2 HÀC(6));

3.00 (s, MsO). ¹³C-NMR (100.6 MHz, CDCl₃): 158.70 (d, ¹J(C,N) 1.2, C(1)); 137.66, 137.43, 137.35, 136.85 (4s);

128.52 127.59 (several d); 80.15 (dd, ³J(C,N) 2.6, C(3)); 78.67 (d, C(4)); 74.65 (dd, ²J(C,N) 10.0, C(2));

74.40, 73.58, 72.35, 72.16 (4t, 4 PhCH₂); 71.93 (d, C(5)); 67.48( t, C(6)); 35.93 (q, MsO). ¹⁵N-NMR (40.6 MHz,

CDCl₃): À80.36 (dd, J 1.3, 0.2). Anal. calc. for C₃₅H₃₇¹⁵NO₈(632.75): C 66.44, H 5.89, N 2.37, S 5.07; found: C

66.70, H 6.15, N 2.62, S 4.89.

(1R,1'S,2'S)- and (1S,1'S,2'S)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-hydrazi-d-(¹⁵N)galactitol (10Se*/

10Sa*). After dissolution of 9* (1.00 g, 1.58mmol) in a sat. soln. of NH ₃in MeOH (40 ml) and stirring for

5 h, the soln. was half-concentrated and then cooled at À158 overnight. The crystals were filtered off and dried

under high vacuum to give 10Se*/10Sa* (718mg, 82%) as white long needles. Evaporation of the mother liquor,

filtration through LiChroprep-NH₂(Et₂O), evaporation, and crystallisation from MeOH gave an additional

crop of 10Se*/10Sa* (ca. 10%). R_f(hexane/AcOEt 1:1) 0.47. M.p. 90 918. [a]²_D⁵ 19.5 (c 1.1, CHCl₃). IR

(CHCl₃): 3250w, 3050w, 3000w, 2905w, 2860m, 1495w, 1450w, 1350w, 1250 1200w, 1090s, 945w, 910w. ¹H-NMR

(600 MHz, C₆D₆, 10Sa*/10Se* 3 :1): Table 2; additionally, 7.37 7.06 (m, 20 arom. H); 5.03 (d, J 11.1), 4.74

(d, J 10.8), 4.59 (d, J 12.3), 4.55 (d, J 11.9), 4.47 (d, J 10.8), 4.41 (d, J 11.8), 4.28 (d, J 11.8), 4.22

(d, J 11.8) (8 PhCH). ¹H-NMR (400 MHz, CDCl₃, 10Sa*/10Se* 85 :15): Table 2; additionally, 7.39 7.26 (m, 20

1512

Helvetica Chimica Acta Vol. 86 (2003)

arom. H); 5.00 (d, J 11.3), 4.84 (d, J 10.8), 4.78 (d, J 11.7), 4.77 4.74 (m), 4.73 (d, J 11.7), 4.64 (d, J

11.3), 4.46 (d, J 11.9), 4.42 (d, J 11.9) (8PhC H). ¹³C-NMR (150.9 MHz, C₆D₆, only one set of signals for

10Se*/10Sa*): Table 3; additionally, 139.32, 139.13, 138.74, 138.56 (4s); 128.60 127.59 (several d); 75.92, 75.85,

73.58, 72.36 (4t, 4 PhCH₂). ¹³C-NMR (100.6 MHz, CDCl₃, only one set of signals for 10Se*/10Sa*): Table 3;

additionally, 138.30, 138.17, 137.92, 137.64 (4s); 128.71 127.43 (several d); 75.78( J(C,N) 0.8), 74.97, 73.42,

73.18( J(C,N) 0.8) (4t, 4 PhCH₂). ¹⁵N-NMR (60.8MHz, C ₆D₆, 10Sa*/10Se* 83 :17): Table 4. ¹⁵N-NMR

(40.6 MHz, CDCl₃, 10Sa*/10Se* 3 : 1): Table 4. Anal. calc. for C₃₄H₃₆¹⁴N¹⁵NO₅(553.68): C 66.44, H 5.89, N 2.37;

found: C 66.70, H 6.15, N 2.62.

(1S,1'S,2'S)- and (1R,1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-(1-methylhydrazi)-d-glucitol (17Se/

17Ra). The reaction of powdered 1 (500 mg, 0.79 mmol) in 7.04m MeNH₂in dry MeOH (20 ml; 2.5 h) gave

17Se/17Ra 5 :2 (436 mg, 97%). Colourless oil. R_f(hexane/AcOEt 1:1) 0.58. [a]²_D⁵ 35.3 (c 0.56, MeOH). IR

(CHCl₃): 3250w, 3050w, 2990w, 2980w, 2920w, 2860m, 1650w (br.), 1495w, 1450m, 1410w (sh.), 1355m, 1305w,

1240w (br.), 1195w (sh.), 1175m (sh.), 1145m (sh.), 1110s (sh.), 1085s, 1060s, 1025m, 995m, 910w. ¹H-NMR

(600 MHz, C₆D₆; 17Se/17Ra 5 :2, ca. 95% pure, assignment based on a ¹H,¹H-COSY spectrum): Table 2;

additionally for 17Se/17Ra, 7.30 7.01 (m, 20 arom. H); additionally for 17Se: 4.90 (d, J 11.3), 4.87 (d, J

11.5), 4.86 (d, J 10.9), 4.73 (d, J 11.5), 4.63 (d, J 11.3), 4.50 (d, J 10.9), 4.39 (d, J 12.1), 4.31 (d, J 12.1)

(8PhC H); additionally for 17Ra, 4.89 (d, J ꢂ 11.2, PhCH); 4.79 (s, PhCH₂); 4.74 (d, J ꢂ 11.8), 4.62 (d, J 11.1),

4.45 (d, J 12.1), 4.33 (d, J 12.2), 4.21 (d, J 11.8) (5 PhCH). ¹³C-NMR (50.3 MHz, C₆D₆, 17Se/17Ra 5 :2):

Table 3; additionally for 17Se/Ra, 128.73 127.53 (several d); additionally for 17Se, 139.38, 139.10, 138.75, 138.61

(4s); 75.70, 75.57, 74.95, 73.36 (4t, 4 PhCH₂); additionally for 17Ra, 139.05, 138.86, 138.69, 138.29 (4s); 75.31,

75.19, 75.07, 73.59 (4t, 4 PhCH₂). CI-MS (NH₃): 568(28), 567 (79, [ M 1] ), 478(13), 477 (44), 459 (19), 448

(38), 447 (100), 445 (10), 444 (37), 436 (20), 431 (11), 430 (34). Anal. calc. for C₃₅H₃₈N₂O₅¥ 0.5 H₂O (575.69): C

73.02, H 6.82, N 4.89; found: C 73.21, H 6.68, N 5.11.

(1R,1'S,2'S)- and (1S,1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-(1-methylhydrazi)-d-(2'-¹⁵N)glucitol

(17Se*/17Ra*). The reaction of 1* (112 mg, 0.18mmol) in 7.04 m MeNH₂in dry MeOH (8ml; 2.5 h) gave

17Se*/17Ra* 5 :2 (101 mg, 98%). R_f(hexane/AcOEt 1:1) 0.58. ¹H-NMR (600 MHz, C₆D₆, 17Se*/17Ra* 5 :2):

Table 2. ¹³C-NMR (150.9 MHz, C₆D₆, 17Se*/17Ra* 5 :2) Table 3; additionally for 17Se*/17Ra*, 139.44 138.45

(several s), 128.80 127.60 (several d); additionally for 17Se*, 75.75, 75.63, 74.99, 73.44 (4t, 4 PhCH₂);

additionally for 17Ra*, 75.35, 75.25, 75.12, 73.69 (4t, 4 PhCH₂). ¹⁵N-NMR (60.8MHz, C ₆D₆): Table 4.

(1S,1'S,2'S)- and (1R,1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-(1-methylhydrazi)-d-galactitol (18Se/

18Ra). The reaction of 9 (500 mg, 0.79 mmol) in 7.04m MeNH₂in dry MeOH (20 ml; 45 min) gave 18Se/18Ra

85 :15 (427 mg, 95%). Crystallization from dry MeOH afforded 18Se (306 mg, 68%) as colourless crystals.

Data of 18Se: R_f(hexane/AcOEt 1 :1) 0.57. M.p. 79 818. [a]²_D⁵ 41.8( c 0.58, MeOH). IR (KBr):

3230w, 3040w, 3020w, 2950w, 2910m, 2860m, 1600w, 1490m, 1465w, 1445m, 1405w, 1395w, 1365m, 1340m, 1300m,

1270w, 1260w, 1250w, 1225m, 1210m, 1150m, 1135m, 1105s, 1070m, 1055m, 1040m, 1020m, 1005m, 98 0m, 955m,

920w, 905w. ¹H-NMR (400 MHz, C₆D₆): Table 2; additionally, 7.37 7.06 (m, 20 arom. H); 5.04 (d, J 11.2), 4.84

(d, J 10.8), 4.59 (d, J 11.2), 4.55 (d, J 11.9), 4.49 (d, J 11.0), 4.41 (d, J 11.9), 4.31 (d, J 11.8), 4.24

(d, J 11.8) (8 PhCH). CI-MS (NH₃): 567 (100, [M 1] ), 108(14), 91 (6).

Data of 18Se/18Ra 85 :15: R_f(hexane/AcOEt 1:1) 0.57. ¹H-NMR (600 MHz, C₆D₆): Table 2; additionally

for 18Ra, 4.99 (d, J 11.3), 4.79 (d, J 11.7), 4.59 (d, J 11.2), 4.46 (d, J 11.8), 4.42 (d, J 11.7), 4.32 (d, J

11.8), 4.24 (d, J 11.8), 4.19 (d, J 11.8) (8 PhCH). ¹³C-NMR (150.9 MHz, C₆D₆, assignment based on a ¹H,¹³C-

COSY spectrum): Table 3; additionally for 18Se/18Ra, 139.32 138.35 (several s); 128.68 127.60 (several d);

additionally for 18Se, 76.02, 75.47, 73.55, 73.36 (4t, 4 PhCH₂); additionally for 18Ra, 75.28, 75.20, 73.69, 72.81 (4t,

4 PhCH₂).

(1R,1'S,2'S)- and (1S,1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-(1-methylhydrazi)-d-(2'-¹⁵N)galactitol

(18Se*/18Ra*). The reaction of 9* (100 mg, 0.16 mmol) in 7.04m MeNH₂in dry MeOH (8ml; 2 h) gave 18Se*/

18Ra* 85 :15 (85 mg, 93%). Crystallization from dry MeOH afforded 18Se* (60 mg, 63%) as colourless crystals.

1

Data of 18Se*: H-NMR (600 MHz, C₆D₆): Table 2; additionally, 7.42 7.02 (m, 20 arom. H); 5.04 (d, J

11.2), 4.84 (d, J 10.8), 4.60 (d, J 11.2), 4.56 (d, J 11.8), 4.50 (d, J 11.1), 4.42 (d, J 11.8), 4.32 (d, J

11.8), 4.24 (d, J 11.8) (8 PhCH). ¹³C-NMR (150.9 MHz, C₆D₆): Table 3; additionally, 139.34, 139.20, 138.98,

138.65 (4s); 128.75 127.48 (several d); 76.01, 75.46, 73.57, 73.41 (4t, 4 PhCH₂).

Data of 18Ra*: ¹H-NMR (600 MHz, C₆D₆, 18Se*/18Ra* 85 :15): Table 2. ¹³C-NMR (150.9 MHz, C₆D₆,

18Se*/18Ra* 85 :15): Table 3. ¹⁵N-NMR (60.8MHz, C ₆D₆): Table 4.

X-Ray Analysis of 18Se⁷). Recrystallization of 18Se in AcOEt/hexane gave crystals suitable for X-ray

analysis: C₃₅H₃₈N₂O₅(566.70); monoclinic P2₁; a 11.851 (3), b 7.836 (1), c 16.652 (3), b 104.18(1) 8;

D_calc. 1.255 Mg/m³; Z 2. From a crystal of size 0.08 Â 0.22 Â 0.38mm 5634 reflections were measured on an

Helvetica Chimica Acta Vol. 86 (2003)

1513

Rigaku AFC5R diffractometer with MoK_aradiation (graphite monochromator, l 0.71073 ä) and a 12-kW

rotating anode generator at 173 K. R 0.0358, R_w 0.0350. The unit-cell constants and an orientation matrix for

data collection were obtained from a least-squares refinement of the setting angles of 25 reflections in the range

488 < 2q < 528. The w/2q scan mode was employed for data collection, where the w scan width was (1.30

0.35 tan q)8 and the w scan speed was 88 min^À1. The structure was solved by direct methods with SHELXS86,

which revealed the positions of all non-H-atoms, which were refined anisotropically. The amine H-atom was

placed in the position indicated by a difference-electron-density map, and its position was allowed to refine

together with an isotropic displacement parameter. All remaining H-atoms were fixed in geometrically

calculated positions (d(CÀH) 0.95 ä). The structure shows no unusual features. The only H-bonding

interaction is a very weak intramolecular interaction between the amine H-atom and the neighbouring O-atom

OÀC(2).

(1'S,2'S)- and (1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-hydrazi-d-mannitol (20S/20R). According to

[7]. The crude product was dissolved in Et₂O and filtered through LiChroprep-NH₂. Evaporation and drying

gave 20S/20R as a clear oil. ¹H-NMR (600 MHz, C₆D₆; 20S/20R 55 :45; assignment based on a ¹H,¹H-COSY

spectrum): Table 5; additionally for 20S/20R, 7.48 7.42 ( m, 2 arom. H); 7.31 7.04 (m, 18arom. H); 4.96 4.26

(m, 8PhC H). ¹³C-NMR (50.3 MHz, C₆D₆; 20S/20R 55 :45): Table 6; additionally for 20S/20R, 139.34, 138.99,

138.96, 138.86, 138.71 (5s); 128.65 127.35 (several d), 73.55 (t, PhCH₂); additionally for 20S, 75.22, 73.39, 71.61

(3t, 3 PhCH₂); additionally for 20R, 73.99, 72.61, 71.85 (3t, 3 PhCH₂). ¹⁵N-NMR (60.8MHz, C ₆D₆): Table 4.

(1'S,2'S)- and (1'R,2'R)-1,5-Anhydro-2,3:4,6-di-O-isopropylidene-1-hydrazi-d-mannitol (24S/24R). Ac-

cording to [21]. Filtration of the crude product through LiChroprep-NH₂gave 24S/24R (95%) as a colourless

foam. Solutions of 24S/24R slowly isomerized and partially decomposed. R_f(hexane/AcOEt 1 :1) 0.18. [a]_D²⁵

22.1 (c 0.43, MeOH). ¹H-NMR (600 MHz, C₆D₆; 24S/24R 3 :7): Table 5; additionally for 24S, 1.50, 1.28, 1.14,

0.95 (4s, 4 Me); additionally for 24R, 1.40, 1.35, 1.20, 1.08(4 s, 4 Me). ¹³C-NMR (50.3 MHz, C₆D₆; 24S/24R 2 :3):

Table 6; additionally for 24S, 112.25, 99.23 (2s, 2 Me₂C); 28.23, 26.77, 26.42, 18.02 (4q, 2 Me₂C); additionally for

24R, 110.10, 99.44 (2s, 2 Me₂C); 28.68, 27.08, 25.13, 18.19 (4q, 2 Me₂C). ¹³C-NMR (50.3 MHz, CDCl₃; 24S/24R

2 :3): Table 6; additionally for 24S, 111.20, 99.82 (2s, 2 Me₂C); 28.82, 27.06, 25.01, 18.82 (4q, 2 Me₂C); additionally

for 24R, 110.98, 99.87 (2s, 2 Me₂C); 28.69, 27.51, 25.49, 18.75 (4q, 2 Me₂C).

(E/Z)-2,3,4,6-Tetra-O-benzyl-d-mannose (¹⁵N)Oxime (26*). A soln. of 0.52n MeONa in MeOH (0.46 ml,

0.24 mmol) was diluted with MeOH (1.5 ml), treated with ¹⁵NH₂OH ¥ HCl (32.4 mg, 0.45 mmol), warmed to 558,

treated dropwise with a soln. of 25 (111 mg, 0.21 mmol) in MeOH (1.5 ml), and stirred at 558 for 20 h. After

evaporation, a soln. of the residue in AcOEt was washed with H₂O (2Â) and brine, dried (MgSO₄), evaporated,

and dried i.v. to affording crude 26* (112 mg, 98%). Yellowish oil. R_f(hexane/1,2-Dimethoxyethane 1:1) 0.21.

(Z)-2,3,4,6-Tetra-O-benzyl-d-(¹⁵N)mannonhydroximolactone (27*). A soln. of 26* (112 mg, 0.2 mmol) and

AcONa (63 mg, 0.76 mmol) in EtOH (7 ml) was warmed to 758, treated dropwise within 1 h with a soln of

NaIO₄(198mg, 0.95 mmol) in 3 ml H ₂O (3 ml), and stirred for 2 h. After evaporation, a soln. of the residue in

AcOEt was washed with H₂O, sat. NaHCO₃soln., H₂O, and brine, dried (MgSO₄), and evaporated. FC (hexane/

AcOEt 2.1) gave 27* (85 mg, 76%). Slightly yellowish oil. R_f(hexane/AcOEt 1 :1) 0.42. IR (CHCl₃): 3380w,

3324w (br.), 3066w, 2910m, 2870m, 1641m, 1612w, 1497m, 1454s, 1364m, 1285m, 1105s, 1071s, 1027s, 945w, 913m,

877w, 8 46w. ¹H-NMR (300 MHz, CDCl₃): 7.42 7.17 (m, 20 arom. H, OH); 4.88 (d, J 10.8), 4.72 (d, J 12.4),

4.62 (d, J 12.1) (3 PhCH); 4.53 (d, J ꢂ 11.8, 2 PhCH); 4.51 (d, J ꢂ 12.5), 4.46 (d, J 12.1), 4.45 (d, J 12.1)

(3 PhCH); 4.30 (t, J 9.0, HÀC(4)); 4.19 (dd, J 3.1, ³J(H,¹⁵N) 1.3, HÀC(2)); 4.00 (dt, J 9.0, 3.6, HÀC(5));

3.79 (d, J 3.7, 2 HÀC(6)); 3.71 (dd, J 9.0, 3.0, HÀC(3)). ¹³C-NMR (100.6 MHz, CDCl₃): 151.80 (s, C(1));

137.79, 137.88, 137.64, 137.26 (4s); 128.42 127.72 (several d); 80.80 (d, C(4)); 79.65 (dd, ³J(C,N) 2.6, C(3));

74.95 (t, PhCH₂); 73.48( d, C(5)); 73.44, 71.60 (2t, 2 PhCH₂); 70.84 (dd, ²J(C,N) 10.8, C(2)); 70.35 (t, PhCH₂);

68.85 (t, C(6)). ¹⁵N-NMR (40.6 MHz, CDCl₃): À76.41 (s). HR-MALDI-MS: 593.2068(5, [ M K] ), 577.2340

(100, [M Na] ; C₃₄H₃₅¹⁵NNaO₆; calc. 577.2332), 555.2522 (39, [M H] ; C₃₄H₃₆¹⁵NO₆; calc. 555.2513),

539.2581 (9).

(Z)-[2,3,4,6-Tetra-O-benzyl-d-(¹⁵N)mannopyranosylidene]amino Methanesulfonate (19*). A suspension of

27* (40 mg, 0.072 mmol) and 4-ä molecular sieves (10 mg) in CH₂Cl₂(0.7 ml) was cooled to 08, treated with

Et₃N (13 ml, 0.093 mmol), stirred for 5 min, treated with a soln. of MsCl (5.6 ml, 0.072 mmol) in CH₂Cl₂(0.3 ml),

and stirred for 10 min. The mixture was diluted with cold AcOEt (10 ml), washed with cold sat. Na₂CO₃soln.,

ice/H₂O, and cold brine, and dried (Na₂SO₄). Evaporation at 258 gave 19* (45 mg, 99%). Colourless oil. R_f

(hexane/AcOEt 3 :2) 0.57. IR (CHCl₃): 3089w, 3066w, 2928w, 2870m, 1627m, 1497m, 1451m, 1371s, 1326m,

1295m, 1103s, 1065s, 1027s, 968s, 911w. ¹H-NMR (300 MHz, C₆D₆): 7.37 (br. d, J 6.5, 2 arom. H); 7.24 7.04 (m,

18arom. H); 4.70 ( d, J 12.1), 4.69 (d, J 11.5) (2 PhCH); 4.46 (t, J 8.5, HÀC(4)); 4.41 (d, J 12.7), 4.36

(d, J 11.5), 4.34 (d, J 12.1), 4.26 (d, J 11.8) (4 PhCH); 4.205 (dd, J 3.1, ³J(H,N) 1.3, HÀC(2)); 4.17

1514

Helvetica Chimica Acta Vol. 86 (2003)

Table 5. Selected ¹H-NMRChemical Shifts [ppm] and Coupling Constants [Hz] of the Mannopyranosylidene

and -furanosylidene-diaziridines 20, 22, 24, 28, 29, 31, and 35

20S/20R^a)

22S/22R

22S/22R [5]

24S/24R [21]

24S/24R^a)

Ratio

Solvent

55 : 45

C₆D₆

48:

CDCl₃

52

60 :

C₆D₆

40

10 :

CDCl₃

90

30 :

C₆D₆

70

b

HÀC(2)

HÀC(3)

HÀC(4)

HÀC(5)

HÀC(6)

H'ÀC(6)

H_aN

H_eN

J(2,3)

J(3,4)

J(4,5)

3.283.81

3.54

3.81

4.40

3.81

4.31

3.88

1.681.67

2.53

3.2

9.89.9

9.4

5.0

10.1

10.4

9.1

3.53

4.02

4.42

3.60

4.30

3.93

3.24

3.56

4.45

3.81

4.11

3.52

3.40

3.91

4.42

3.25

4.04

3.52

)

4.35

4.24 3.99

4.03 4.16

3.72 4.05

3.19 3.27

3.67 3.77

3.39 3.64

b

3.53 3.94

4.46 4.19

3.97 3.90

3.76 3.87

3.64 3.87

1.19 1.96

2.47 2.33

3.0 3.0

9.3 7.1

4.33

3.96

3.55

3.86

3.73

b

1.181.77

2.12

2.101 11..983

2.28

2.09

3.2

2.35

3.2

1.82

3.2

2.55

8.0

6.0

2.23 2.51

b

)

7.4

6.5

6.7

7.8

b

9.89.8

b

9.86.1

9.4

4.9

10.0

10.3

9.2

10.4

10.4 10.3

5.7 5.6

10.1 10.5

11.0 10.9

b

J(5,6)

J(5,6')

J(6,6')

J(H_a,H_e)

4.6

1.5

11.2

)

5.0

10.2

10.4

9.2

4.9

10.2

9.2

5.6

10.5

11.0

9.3

b

9.1 9.4

9.3

9.4

9.5

20S/20R [7] 20Se*/20Re*/20Sa*/20Ra*

Ratio

ca. 1 : 1

50 :

41:

5 :

4

Solvent

CDCl₃

C₆D₆

H_aN^c)

1.45 1.90

2.49 2.04

8.7 8.7

1.16 (56.9) 1.94 (57.4) 1.16 (2.8) 1.94 (3.5)

2.46 (3.8) 2.36 (3.2) 2.46 (58.2) 2.36 (57.2)

H_eN^c)

J(H_a,H_e)

9.2

9.4

9.1

9.4

28Se/28Re

29Se/29Re^a)

31R [7]

31R

31Rx*/31Rn*

35Rx/35Sn^a)

Ratio

Solvent

85 : 15

C₆D₆

80 :

C₆D₆

20

100%

CDCl₃

100%

C₆D₆

4 :

C₆D₆

1

85 :

C₆D₆

15

b

HÀC(2)

HÀC(3)

HÀC(4)

HÀC(5)

HÀC(6)

H'ÀC(6)

3.24 3.915 3.65

3.53 3.925 4.14

4.44 4.44 3.99

3.97 3.45 3.55

3.79 3.73 3.85

3.69 3.61 3.59

4.084.76

4.19

3.93

3.06

3.69

4.34

)

b

4.96

4.37

4.39

b

4.083.65

4.45

3.64

3.58

b

4.46

4.01

3.97

4.46

4.01

3.97

4.51

4.06

4.00

b

4.10

4.01

b

3.61

2.082.21

H_aN or H_exoÀN^c) 1.52

H_eN or H_endoÀN^c)

2.32 2.15

2.01

2.02 (

ca. 3.0/57.2)

2.43 (57.8/ca. 3.0) 2.85

2.52

2.51

2.41

MeN

2.70 2.43 2.57

2.61

5.5

7.9

2.46 2.70

b

J(2,3)

J(3,4)

J(4,5)

J(5,6)

J(5,6')

J(6,6')

J(H_a,H_e)

2.9 < 1

6.3

7.2

5.85.7

3.3

5.7

)

c

b

9.3

)

3.1

7.2

3.1

7.3

3.1

7.4

)

b

9.9 10.0 10.4

10.1

5.7

8.0

4.6

1.6

11.2

4.85.2

1.6 10.4

11.3 10.4

5.9

5.5

5.3

5.2

10.1

10.3

4.3

9.0

9.5

6.4

8.8

9.0

6.5

8.7

9.3/9.3

6.4

8.7

a

c

) Assignment based on a H,¹H-COSY spectrum. ^b) Not assigned. ) In parentheses, J(¹⁵N,H ) .

1

Helvetica Chimica Acta Vol. 86 (2003)

1515

Table 6. Selected ¹³C-NMRChemical Shifts [ppm] of the Mannopyranosylidene and -furanosylidene-diaziridines

20, 22, 24, 28, 29, 31, and 35

20S/20R

22S/22R [5]

24S/24R

Ratio

55 :

45

ca. 1 : 1

1 :

9

55 :

45

Solvent

C₆D₆

CDCl₃

C₆D₆

a

C(1)

C(2)

C(3)

C(4)

C(5)

C(6)

82.18

78.20

81.72

74.53

76.40

69.35

81.47

77.66

81.13

74.53

76.01

69.78

82.53, 82.24

78.62, 77.32

78.14, 78.09

78.19

69.83, 69.00

68.33

)

80.96

81.22

76.06^b)

73.67^b)

71.66

66.81

62.28

76.15^b)

72.24^b)

72.03

67.66

61.58

76.51^b)

71.71^b)

70.10

76.41^b)

75.31^b)

71.86

67.96

60.79

66.65

61.88

28Se/28Re

29Se/29Re

31R

35Rx/35Sn

Ratio

85 :

15

80 :

20

4 :

1

Solvent

C₆D₆

CDCl₃

C₆D₆

a

C(1)

C(2)

C(3)

C(4)

C(5)

C(6)

MeN

84.08

78.50

82.04

74.87

77.88

69.22

38.88

)

84.06

83.96

90.83

91.40

93.69

80.24^b)

79.66^b)

82.58

73.45

66.8

92.74

78.59

84.08

74.30

77.2

69.44

38.80

77.11^b)

76.16^b)

67.70

77.27^b)

75.33^b)

66.94

80.25^b)

79.36^b)

80.95

80.60^b)

79.61^b)

81.42

80.40^b)

79.95^b)

80.74

a

72.87

62.3861.95

38.63

72.87

66.67

41.37

72.97

66.93

73.51

67.056

)

39.13

40.14

^a) Not assigned. ^b) Assignments may be interchanged.

(d, J 12.1, 2 PhCH); 3.77 (dt, J 8.4, 3.1, HÀC(5)); 3.49 (d, J 3.1, 2 HÀC(6)); 3.43 (dd, J 8.4, 2.8,

HÀC(3)); 2.51 (s, MsO). ¹³C-NMR (75.6 MHz, CDCl₃): 157.74 (s, C(1)); 138.52, 138.40, 138.29, 137.49 (4s);

128.7 127.6 (several d); 82.17 (d, C(4)); 79.25 (br. d, C(3)); 74.84 (t, PhCH₂); 73.43 (d, C(5)); 73.31, 72.03, 71.52

(3t, 3 PhCH₂); 68.31 (t, C(6)); 35.73 (q, MsO), dd for C(2) hidden by the noise or other signals. ¹⁵N-NMR

(40.6 MHz, C₆D₆): À76.6 (s). HR-MALDI: 671.1820 (25, [M K] , C₃₅H₃₇K¹⁵NO₈S ; calc. 671.1842), 655.2107

(100, [M Na] ; calc. for C₃₅H₃₇¹⁵NNaO₈S 655.2102), 633.2282 (19, [M H] ; calc. for C₃₅H₃₈¹⁵NO₈S

633.2283), 561 (32), 559 (30), 539 (49), 470 (25), 469 (82), 425 (25), 181 (27).

(1R,1'S,2'S)-, (1R,1'R,2'R)-, (1S,1'S,2'S)-, and (1S,1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-hydrazi-

d-(¹⁵N)mannitol (20Se*/20Re*/20Sa*/20Ra*). A sat. soln. of NH₃in MeOH (NH₃was dumped into MeOH

at 08, 2 ml) was cooled to 08, treated with 19* (10 mg, 15.8 mmol), and stirred for 23 h at 238. After evaporation at

08, the residue was dissolved in Et₂O and filtered through Lichroprep-NH₂(2 ml). Evaporation of the filtrate at

258 gave a 50 :41:5 :4 mixture 20Se*/20Re*/20Sa*/20Ra* (7.5 mg, 86%). Colourless oil. R_f(hexane/AcOEt 1 :1)

1

0.28. H-NMR (300 MHz, C₆D₆; 20Se*/20Re*/20Sa*/20Ra* 50 :41:5 :4): Table 5; additionally, 7.49 (br. d, J

6.8, 2 arom. H); 7.44 7.01 (m, 18arom. H); additionally for 20Se*/20Sa*, 4.62 (d, J 11.8, PhCH); 4.52 4.23

(m, 7 PhCH, HÀC(4)); 3.98( ddd, J 10.0, 4.4,1.6, HÀC(5)); 3.76 (dd, J 11.2, 4.3, HÀC(6)); 3.63 (dd, J 11.2,

1.6, H'ÀC(6)); 3.50 (d, J 9.3, 2.9, HÀC(3)); 3.25 (d, J 3.1, HÀC(2)); additionally for 20Re*/20Ra*, 4.95

(d, J 11.3, PhCH); 4.92 (d, J 10.6, PhCH); 4.56 (d, J 11.5, PhCH); 4.52 4.23 (m, 5 PhCH); 4.19 (t, J ꢂ 6.4,

HÀC(4)); 3.94 (dd, J 6.8, 2.8, HÀC(3)); 3.91 3.84 (m, HÀC(5), 2 HÀC(6)); 3.80 (d, J 3.1, HÀC(2)).

(1S,1'S,2'S)- and (1S,1'R,2'R)-1,5-Anhydro-2,3,4,6-tetra-O-benzyl-1-(1-methylhydrazi)-d-mannitol (28Se/

28Re). The reaction of 19 (515 mg, 0.81 mmol) in 7.04m MeNH₂in dry MeOH (20 ml; 2 h) and drying for 3 h at

08 gave 28Se/28Re 85 :15 (461 mg, 98%). R_f(hexane/AcOEt 1:1) 0.27. [a]²_D⁵ À10.2 (c 0.61, MeOH). IR

(CHCl₃): 3250w, 3050w, 3020w (sh.), 2990m, 2930m, 2860m, 1665w, 1495m, 1450m, 1410w, 1380w, 1360m, 1325w,

1285m, 1110s (br.), 1085s (sh.), 1050s, 1025m, 910w, 8 40w (br.). ¹H-NMR (400 MHz, C₆D₆; 28Se/28Re 85 :15):

Table 5; additionally for 28Se/28Re, 7.49 (d, J 7.2, 2 arom. H); 7.36 7.05 (m, 18arom. H); additionally for

28Se, 4.96 (d, J 11.3), 4.89 (d, J 12.5), 4.73 (d, J 12.4), 4.56 (d, J 11.4), 4.49 (d, J 12.1), 4.36 (d, J

1516

Helvetica Chimica Acta Vol. 86 (2003)

12.1), 4.35 (d, J 11.8), 4.28 (d, J 11.7) (8PhC H); additionally for 28Re, 5.00 (d, J 11.4), 4.55 (d, J 12.0),

4.53 (d, J 12.6), 4.38( d, J 12.7) (4 PhCH). ¹³C-NMR (50.3 MHz, C₆D₆, 28Se/28Re 85 :15): Table 6;

additionally for 28Se/28Re, 139.41 138.94 (several s); 128.62 127.59 (several d); additionally for 28Se, 75.22,

73.43, 71.60, 71.55 (4t, 4 PhCH₂); additionally for 28Re, 73.92, 73.73, 72.92, 71.55 (4t, 4 PhCH₂). ¹⁵N-NMR

(60.8MHz, C ₆D₆): Table 4. ESI-MS : 589 (100, [M Na] ).

(1S,1'S,2'S)- and (1R,1'S,2'S)-1,5-Anhydro-2,3:4,6-di-O-isopropylidene-1-(1-methylhydrazi)-d-mannitol

(29Se/29Re). The reaction of 23 (507 mg, 1.44 mmol) in 7.04m MeNH₂in dry MeOH (20 ml; 1 h) gave 29Se/

29Re 4 :1 (368mg, 83%) as a colourless foam. R_f(hexane/AcOEt 1 :1) 0.26. IR (CHCl₃): 3260w, 2990m,

2930m, 1455w, 1440w, 1380s, 1370m, 1345w, 1295m, 1265m, 1240m, 1195m, 1160m, 1105s, 1090s, 108 0s, 1030m,

995m, 970m, 940m, 8 50m. ¹H-NMR (600 MHz, C₆D₆; 29Se/29Re 4 :1, assignment based on a ¹H,¹H-COSY

spectrum): Table 5 ; additionally for 29Se, 1.54, 1.43, 1.21, 1.19 (4s, 2 Me₂C); additionally for 29Re, 1.50, 1.42,

1.24, 1.11 (4s, 2 Me₂C). ¹³C-NMR (50.3 MHz, C₆D₆, 29Se/29Re 4 :1): Table 6 ; additionally for 29Se, 111.18,

99.63 (2s, 2 Me₂C); 29.15, 27.55, 26.41, 18.68 (4q, 2 Me₂C); additionally for 29Re, 110.45, 99.75 (2s, 2 Me₂C);

29.21, 28.38, 26.41, 18.68 (4q, 2 Me₂C). ¹⁵N-NMR (60.8MHz, C ₆D₆): Table 4. ESI-MS: 325 (92, [M K] ), 309

(45, [M Na] ), 287 (100, [M 1] ). Anal. calc. for C₁₃H₂₂N₂O₅(286.45): C 54.51, H 7.74, N 9.77; found: C

54.27, H 8.00, N 9.85.

(1'S,2'S)-1,4-Anhydro-2,3:5,6-di-O-isopropylidene-1-hydrazi-d-mannitol (31R). According to [7].

¹H-NMR (400 MHz, C₆D₆): Table 5; additionally, 1.41, 1.32, 1.26, 1.08(4 s, 4 Me). ¹³C-NMR (50.3 MHz,

C₆D₆): Table 6; additionally, 113.45, 109.16 (2s, 2 Me₂C); 27.02, 26.27, 25.59, 25.42 (4q, 2 Me₂C).

(E/Z)-2,3:5,6-Di-O-isopropylidene-d-mannofuranose (¹⁵N)Oxime ((E/Z)-33*). A soln. of MeONa in

MeOH (92 mg of Na, 4 mmol; 30 ml of MeOH) was diluted with. MeOH (20 ml), treated with ¹⁵NH₂OH ¥ HCl

(308mg, 4.4 mmol), warmed to 50 8, treated portionwise with 32 (0.75 g, 2.9 mmol), stirred for 9 h at 508, 3 d at

r.t., and 8h at 50 8, and evaporated. A soln. of the residue in AcOEt, was washed with H₂O (2Â) and brine, dried

(MgSO₄), and evaporated. FC (AcOEt/hexane 1 :1) gave (E/Z)-33* (0.4 g, 50%). R_f(AcOEt/hexane 1:1) 0.47.

(E)- and (Z)-2,3:5,6-Di-O-isopropylidene-d-(¹⁵N)mannonhydroximo-1,4-lactone ((E)- and (Z)-34*). A

soln. of (E/Z)-33* (0.4 g, 1.45 mmol) in MeOH (12 ml) was treated with MnO₂(prepared according to [38];

0.3 g, 3.42 mmol), stirred for 3 h, and filtered through Celite. Evaporation of the filtrate, FC (AcOEt/hexane

1:1) and crystallisation from CH₂Cl₂/hexane gave (Z)-34* (140 mg, 35%) und (E)-34* (20 mg, 5%).

1

Data of (Z)-34*: R_f(AcOEt/hexane 3 :2) 0.48. M.p. 1758. H-NMR (300 MHz, CDCl₃): 6.35 (br. s, OH);

5.15 (d, J 5.6, HÀC(2)); 4.88 (dd, J 5.6, 3.7, HÀC(3)); 4.50 (dt, J 8.1, 5.0, HÀC(5)); 4.30 (dd, J 8.4, 3.7,

HÀC(4)); 4.15 (d, J 4.7, 2 HÀC(6)); 1.51, 1.47, 1.42, 1.40 (4s, 2 Me₂C). ¹H-NMR (500 MHz, C₆D₆): 6.75 (br. s,

OH); 4.62 (d, J 5.6, HÀC(2)); 4.39 (ddd, J 7.7, 6.3, 4.9, HÀC(5)); 4.11 (dd, J 5.6, 3.5, HÀC(3)); 4.05

(dd, J 8.9, 4.9, HÀC(6)); 3.92 (dd, J 8.9, 6.3, H'ÀC(6)); 3.65 (dd, J 7.7, 3.5, H À C(4)); 1.38, 1.33, 1.22, 1.09

(4s, 2 Me₂C). ¹³C-NMR (125.8MHz, C ₆D₆): 156.58( d, ¹J(C,N) 3.0, C(1)); 113.66, 109.43 (2s, 2 Me₂C); 82.33

(d, C(4)); 77.89 (dd, ²J(C,N) 8.7, C(2)); 77.71 (br. d, C(3)); 73.14 (d, C(5)); 66.76 (t, C(6)); 26.97, 26.90, 25.81,

25.36 (4q, 2 Me₂C). ¹⁵N-NMR (50.7 MHz, C₆D₆): À91.1 (s). HR-MALDI-MS: 298.111 (12), 297.108 (100, [M

Na] ; C₁₂H₁₈¹⁵NNaO₆; calc. 297.108).

Data of (E)-34*: R_f(AcOEt/hexane 3 :2) 0.75. M.p. 1228. (E)-34* isomerized to (Z)-34* upon storage at r.t.

(Z)-(2,3 :5,6-Di-O-isopropylidene-d-(¹⁵N)mannofuranosylidene)amino Methanesulfonate (30*). At r.t.

and under N₂, a soln. of (Z)-34* (120 mg, 0.43 mmol) in dry CH₂Cl₂(10 ml) was treated with Et₃N (0.14 ml,

1 mmol) and dropwise with MsCl (0.05 ml, 0.65 mmol), stirred for 1 h, diluted with CH₂Cl₂, washed with sat.

NaHCO₃soln. and H₂O (2Â), dried (MgSO₄), and evaporated. FC (AcOEt/hexane 1:1) gave crude 30*

(150 mg) as yellowish oil. Crystallisation from Et₂O/hexane gave 30* (131 mg, 87%). Colourless crystals. R_f

(AcOEt/hexane 1:1) 0.45. M.p. 100 1018.

(1R,1'S,2'S)- and (1S,1'S,2'S)-1,4-Anhydro-1-hydrazi-2,3:5,6-di-O-isopropylidene-d-(¹⁵N)mannitol (31Rx*/

31Rn*). A soln. of 30* (63 mg, 0.22 mmol) in a saturated soln. of NH₃in dry MeOH (4 ml) was stirred for 5 h

under N₂and at r.t. and stored for 10 h at 48. Filtration through LiChroprep-NH₂, evaporation, and drying gave a

colourless resin (46 mg). FC (AcOEt/hexane 1:1, column cooled with acetone/dry ice) afforded 31Rx*/31Rn*

4 :1 (28mg, 57%). R_f(AcOEt/hexane 1:1) 0.24. ¹H-NMR (300 MHz, C₆D₆): Table 5; additionally, 1.42, 1.32,

1.26, 1.08(4 s, 2 Me₂C).

(1S,1'S,2'S)- and (1R,1'R,2'R)-1,4-Anhydro-2,3:5,6-di-O-isopropylidene-1-(1-methylhydrazi)-d-mannitol

(35Rx/35Sn). The reaction of 30 (142 mg, 0.4 mmol) in 7.04m MeNH₂in dry MeOH (8ml; 1.5 h) and drying

at 08 for 4 h gave a 74 :14 :6 :6 mixture of 35Rx, 35Sn, and two secondary products (107 mg, 92%). R_f(hexane/

AcOEt 1:2) 0.32 and 0.16. IR (CHCl₃): 3260w, 2980m, 2950w, 2930m, 2880w, 1665m, 1530w, 1450w, 1415w,

1380m, 1370m, 1250m (br.), 1195m, 1155m, 1145m, 1110m, 1070s, 1045m, 995w, 970m, 950w, 935w (sh.), 885w,

Helvetica Chimica Acta Vol. 86 (2003)

1517

840w. ¹H-NMR (400 MHz, C₆D₆; 35Rx/35Sn 85 :15, assignment based on a ¹H,¹H-COSY spectrum): Table 5;

additionally for 35Rx, 1.41, 1.33, 1.26, 1.11 (4s, 2 Me₂C); additionally for the secondary products, 4.65 (d, J 8.1,

HÀC(2)); 4.23 (d, J 7.6, HÀC(2)); 2.44, 2.43, 2.36, 2.34 (4s, 2 MeN). ¹³C-NMR (50.3 MHz, C₆D₆; 35Rx/35Sn

85 :15): Table 6; additionally for 35Rx, 113.36, 109.19 (2s, 2 Me₂C); 26.90, 26.22, 25.62, 25.38(4 q, 2 Me₂C);

additionally for 35Sn, 113.19, 109.34 (2s, 2 Me₂C). CI-MS (NH₃): 288 (12), 287 (100, [M 1] ).

(Z)-[2,3-O-Isopropylidene-5-O-(triphenylmethyl)-d-ribofuranosylidene]amino Methanesulfonate ((Z)-

37). A soln. of (Z)-36¹⁸) [17] (4.44 g, 10 mmol) and Et₃N (3.0 ml, 21.5 mmol) in dry CH₂Cl₂(100 ml) was

treated dropwise with MsCl (0.85 ml, 10.9 mmol), and stirred for 1 h. Washing with sat. NaHCO₃soln. and H₂O,

drying (MgSO₄), evaporation, and FC (hexane/AcOEt 2 :1) gave (Z)-37 (4.46 mg, 85%). Colourless foam. R_f

(hexane/AcOEt 1:1) 0.56. IR (CHCl₃): 3090w (sh.), 3060w (sh.), 3020w, 3000w (sh.), 2940w, 2880w, 1675m,

1600w, 1490m, 1450m, 1370s, 1325m, 1255m, 1180s, 1150m, 1095s, 1080s (sh.), 1000s, 970s, 960w, 935w, 900w,

870m. ¹H-NMR (400 MHz, CDCl₃): 7.39 7.16 (m, Ph₃C); 5.43 (d, J 5.8, HÀC(2)); 4.84 (t, J ꢂ 1.9, HÀC(4));

4.60 (dd, J 5.9, 1.0, HÀC(3)); 3.73 (dd, J 10.9, 2.5, HÀC(5)); 3.02 (dd, J 10.9, 1.6, H'ÀC(5)); 3.12 (s,

MsO); 1.50, 1.36 (2s, Me₂C). ¹³C-NMR (50.3 MHz, CDCl₃): 164.73 (s, C(1)); 142.82 (3s); 128.36 127.46 (several

d); 114.01 (s, Me₂C); 88.23 (d, C(4)); 87.94 (s, Ph₃C); 80.16 (d, C(3)); 78.52 (d, C(2)); 63.31 (t, C(5)); 36.01

(q, MsO); 26.75, 25.52 (2q, Me₂C). CI-MS: 541 (7, [M NH₄] ), 482 (3), 431 (29), 430 (100, [M À MsO 2] ),

299 (27), 244 (14), 243 (82, Ph₃C ), 188 (13). Anal. calc. for C₂₈H₂₉NO₇S (523.59): C 64.23, H 5.58, N 2.67, S 6.12;

found: C 63.98, H 5.67, N 2.52, S 6.30.

(1R,1'R,2'R)-, (1R,1'S,2'S)-, (1S,1'R,2'R)-, and (1S,1'S,2'S)-1,4-Anhydro-2,3-O-isopropylidene-1-(1-meth-

ylhydrazi)-5-O-ꢀtriphenylmethyl)-d-ribitol (38Rn/38Sn/38Rx/38Sx). The reaction of (Z)-37 (510 mg,

0.97 mmol) in 7.04m MeNH₂in dry MeOH (20 ml; 3.5 h) gave 38Rn/38Sn/38Rx/38Sx 76 :4 :12 :8(418mg,

94%). R_f(hexane/AcOEt 1 :1) 0.25. M.p. 48 50 8. [a]²_D⁵ À36.7 (c 0.53, MeOH). IR (KBr): 3240m, 3040w,

3010w, 2980m, 2920m, 2860w, 1650w (br.), 1595w, 1485m, 1445s, 1410w, 1370s, 1320m, 1250m, 1210s, 1175m,

1150m, 1115s, 1075s, 1040m (sh.), 1025m (sh.), 995m, 975m, 945w, 8 95w, 8 65m, 8 05w. ¹H-NMR (400 MHz, C₆D₆;

38Rn/38Sn/38Rx/38Sx 76 :4 :12 :8, assignment based on a ¹H,¹H-COSY spectrum): Table 7; additionally for

38Rn/38Sn/38Rx/38Sx, 7.52 7.33 (m, 6 arom. H); 7.15 6.97 (m, 9 arom. H); additionally for 38Rn, 1.54, 1.21 (2s,

Me₂C); additionally for 38Sn, 1.41, 1.15 (2s, Me₂C); additionally for 38Rx, 1.48, 1.14 (2s, Me₂C); additionally for

38Sx, 1.36, 1.06 (2s, Me₂C). ¹³C-NMR (50.3 MHz, C₆D₆, 38Rn/38Rx 85 :15): Table 7; additionally for 38Rn/

38Rx, 144.11 (3s); 129.42 127.03 (several d), 87.69 (s, Ph₃C); additionally for 38Rn, 113.06 (s, Me₂C); 27.04,

26.02 (2q, Me₂C); additionally for 38Rx, 26.86, 25.47 (2q, Me₂C). ¹⁵N-NMR (60.8MHz, C ₆D₆): Table 4. CI-MS

(NH₃): 460 (25), 459 (100, [M 1] ), 243 (27, Ph₃C ).

(E/Z)-2,3,5-Tri-O-benzyl-d-ribose Oxime (40) [40]. A soln. of NaOEt (0.92 g of Na, 40.0 mmol) in abs.

EtOH (150 ml) was treated at 558 with NH₂OH ¥ HCl (5.56 g, 80.0 mmol) and portionwise with 39 [26][27]

(4.21 g, 10.0 mmol), stirred for 2.5 h at 558, and evaporated. The residue was dissolved in CH₂Cl₂, washed (2 Â

with H₂O, 1 Â with brine), and dried (MgSO₄). Evaporation and drying gave (E)-40/(Z)-40 4 :1 (4.50 g, quant.).

Colourless oil. R_f(hexane/AcOEt 1:1) 0.42. [a]_D²⁵ 42.0 (c 1.06, CHCl₃). IR (CHCl₃): 3570m, 3340m, 3060w,

3020w (sh.), 2990m, 2910m, 2865m, 1495m, 1455m, 1390w, 1370m, 1350m, 1325m (br.), 1260w, 1090s (br.), 1070s

(sh.), 1025m, 945m, 910m, 8 15w. ¹H-NMR (400 MHz, CDCl₃, (E)-40/(Z)-40 4 :1): 8.89 (br. s, exchange with

D₂O, NOH); 7.52 (d, J 8.4, 0.8 H), 6.98 (d, J 6.3, 0.2 H) (HÀC(1)); 7.38 7.21 ( m, 15 arom. H); 5.15 (dd, J

2.1, 6.3, 0.2 H), 4.47 (dd, J 1.7, 8.3, 0.8 H) (HÀC(2)); 4.88 (d, J 11.4, 0.8H), 4.76 ( d, J 11.4, 0.2 H), 4.76

(d, J 11.4, 0.2 H), 4.70 (d, J 10.9, 0.2 H), 4.64 (d, J 10.3, 0.2 H), 4.61 (d, J 11.7, 0.8H), 4.58( d, J 12.0,

0.8H), 4.60 4.49 ( m, 2 H), 4.45 (d, J 11.7, 0.8H) (6 PhC H); 4.63 4.58(br. s, exchange with D₂O, 0.8H),

4.42 4.37 (br. s, exchange with D₂O, 0.2 H) (OH); 3.93 3.88 (m, 0.2 H), 3.45 (br. dd, J 7.0, 8.9, 0.8 H)

(HÀC(4)); 3.86 (dd, J 2.1, 8.9, 0.2 H), 3.78 (dd, J 2.1, 8.9, 0.8 H) (HÀC(3)); 3.78 3.74 ( m, 1.6 H), 3.67

(dd, J 2.7, 9.6, 0.2 H), 3.62 (dd,J 4.9, 9.7, 0.2 H) (2 HÀC(5)). ¹³C-NMR (50.3 MHz, CDCl₃, (E)-40/(Z)-40

4 :1): (E)-40: 149.03 (d, C(1)); 137.91, 137.83, 136.76 (3s); 128.42 127.21 (several d); 80.64 (d, C(3)); 78.21

(d, C(2); 74.13, 73.57, 72.87 (3t, 3 PhCH₂); 71.14 (t, C(5)); 68.79 (d, C(4)); (Z)-40: 151.70 (d, C(1)); 137.83,

136.04 (2s); 79.65 (d, C(3)); 74.06, 73.12, 72.47 (3t, 3 PhCH₂); 71.99 (d, C(2)); 71.23 (t, C(5)); 69.22 (d, C(4)).

CI-MS (NH₃): 453 (8, [M NH₄] ), 437 (28), 436 (100, [M 1] ). Anal. calc. for C₂₆H₂₉NO₅(435.52): C 71.71,

H 6.71, N 3.22; found: C 71.44, H 6.46, N 3.24.

18

)

The preparation of 36 by oxidation with MnO₂[24] instead of NaIO₄[17] gave (Z)-37/(E)-36 4:1 and,

hence, by mesylation, (Z)-36/(E)-37 4:1. H-NMR data of (Z)-37 (400 MHz, CDCl₃, (Z)-37/(E)-37 4:1):

1

5.58( d, J 5.8, HÀC(2)); 4.78(br. t, J ꢂ 1.9, HÀC(4)); 4.41 (d, J 5.8, HÀC(3)); 3.76 (dd, J 10.9, 2.5,

HÀC(5)); 3.13 (s, MsO); 3.07 (dd, J 10.9, 1.6, H'ÀC(5)); 1.48, 1.33 (2s, 2 Me₂C).

Article Doi

Glycosylidene carbenes: Part 31. Glycosylidene diaziridines: Stereoselective addition of ammonia and methylamine to lactone oxime sulfonates

DOI: 10.1002/hlca.200390132

Source and publish data:

Authors:

Article abstract of DOI:10.1002/hlca.200390132

Full text of DOI:10.1002/hlca.200390132

Products guided by the article

R&D Labs maybe for 620946-65-8

Relevant to this article

Hot Product