J . Org. Chem. 1986,51, 3479-3485 column chromatography with hexane as eluant, giving 100 mg of clear oil (99 % yield): 'H NMR (CDC13)6 1.90-1.60 (m, 13 H), 1.47 (d, J = 11.7 Hz, 2 H), 1.11(d, J = 10.4 Hz, 1H), 0.86 (d, J = 6.6 Hz, 6 H); 13C NMR (CDC13)6 51.7 (d), 39.5 (t),38.5 (t),31.9 (t), 29.4 (d), 28.1 (d), 27.9 (d), 27.1 (d), 20.9 (9). Dihydroaromatic Trapping of B +from Biadamantylidene (1) (Table 11) and from Dioxetane 2. The trapping experiments were performed by adding concentrated solutions of 9" in CH2C12 (0.02 M) dropwise to solutions of 1 or 2 in oxygen-saturated CH2C12 at -78 "C containing 5.0 equiv of 11 or 12. The reactions were quenched with Et3N, warmed to room temperature, and evaporated. The crude materials were tritrated thoroughly with hot
3479
pentane, filtered, and evaporated. Analysis of the product mixtures by 'H NMR was facilitated by comparison to spectra of known materials, and product ratios were measured by integration.
Acknowledgment. We thank the donors of Petroleum Research Fund, administered by the American Chemical Society, for partial financial support of this work, and Amoco Oil Co. for a fellowship t o M.F.T. Registry No. 1, 30541-56-1; 2, 35544-39-9; 7, 73321-28-5; 8, 73679-39-7; 11, 628-41-1; 12, 613-31-0; 13, 53862-33-2; 15, 15914-95-1;16, 20441-18-3; 19, 29186-07-0.
Halogenation of 1,5-Anhydrohex-l-enitols(Glycals). Influence of the C-6 Substituent Derek Horton,* Waldemar. Priebe, and Oscar Varela Department of Chemistry, T h e Ohio State University, Columbus, Ohio 43210
Received February 12, 1986 The stereochemistry of addition of chlorine and bromine to 3,4-di-O-acetyl-~-rhamnal(I), 3,4-di-O-acetyl-~-fucal (2), and other glycals has been investigated. Variations in the reaction conditions lead to dihalides having different
configurations at C-1 and C-2. Chlorination in nonpolar solvents appears to proceed via stabilized syn ion pairs, selectively affording cis addition products. The product distribution from the bromination reactions suggests the participation of different ionic intermediates. In both chlorination and bromination, the product distribution is affected by the polarity of the solvent, the structure of the enol ether, and the nature of the halogen. The product distribution in the bromination reactions depends on the electron-withdrawing or -donating effect of the substituent at C-6. This result was interpreted in terms of the effect that the 6-substituent may exert on a nonbonding electron pair of the ring oxygen atom, affecting the stabilization of the carbocation at the anomeric center. The C-6 substituent exerts no significant effect on the chlorination 'reaction. Changes in the configuration of the substituents of the glycal may also modify the stereochemical course of the reaction. Thus, a change in the orientation of the acetoxy group at C-4 from equatorial (1) to axial (2) influenced the side of attack by halogen upon the double bond, leading to different ratios of cis and trans addition products from 1 and 2. T h e addition of halogens to cyclic enol ethers was earlier investigated by Lemieux and Fraser-Reid,2v3who proposed a general mechanism involving polar attack of halogen on the double bond, resulting in formation of carbonium ions, which upon attack by halide ion lead principally t o products of thermodynamic control. However, Igarashi e t al.4 established t h a t product formation is under kinetic not thermodynamic control and t h a t the stereoselectivity of t h e addition is dependent on the polarity of t h e solvent. Boullanger a n d Descotes5 studied comparatively t h e addition of chlorine and bromine t o acetylated a n d benzylated derivatives of D-glued. T h e product distribution was explained on t h e basis of Igarashi's mechanism a n d a quantitative correlation established between the polarity of the solvents a n d the stereospecificity of the addition of chlorine. A general mechanism of chlorination of 3,4-dihydro-2H-pyran in several solvents, consistent with t h e observed solvent dependency, was proposed by Stone and Daves? In the present work, chlorination a n d bromination of
3,4-di-O-acetyl-~-rhamnal(I), 3,4-di-O-acetyl-~-fucal(2), a n d other substituted glycals has been studied as part of
from
(1) Supported, in part, by Grant No. GM-11976 the National Institute of General Medical Sciences. For a preliminary report, see: Horton, D.; Priebe, W.; Varela, 0. Abstr. Pap.-Am. Chem. SOC.1982, 184th, CARB-33. (2) Lemieux, R. U.; Fraser-Reid, B. Can. J.Chem. 1964,42,532-538. (3) Lemieux, R. U.; Fraser-Reid,B. Can. J. Chem. 1965,43,1460-1478. (4) Igarashi, K.; Honma, T.;Imagawa, T . J. Org. Chem. 1970, 35, 610-616. (5) Boullanger, P.; Descotes, G. Carbohydr. Res. 1976, 51, 55-63. (6)Stone, T. E.; Daves, G. D. J. Org. Chem. 1977, 42, 2151-2154.
0022-3263/86/1951-3479$01.50/0
a synthetic program targeted toward 2'-halo derivatives of anthracycline antibiotics.' T h e 1,2-dihalides constitute useful synthetic intermediates or starting m a t e r i a k 8 Product distribution obtained by halogenation of glycals under controlled conditions is discussed in terms of solvent polarity, the nature of the halogen, and t h e influence of steric effects in the alkene. T h e factors discussed here as being responsible for the product distribution during the halogenation reactions may help t o explain the course of other electrophilic additions t o alkenes.
Results and Discussion Structural Assignments for the Products. Halogenation of the glycal derivatives was performed in t h e dark a t 0 "C. T h e composition of t h e mixtures was determined by 'H NMR spectroscopy and, in some instances, confirmed by analytical LC. T h e main products formed by the addition of chlorine or bromine (Scheme I) t o the double bond of 3,4-di-O-acetyl-~-rhamnal(1) and 3,4-di0-acetyl-L-fucal (2) were isolated by column chromatography and the minor ones purified by semipreparative LC. T h e structures of the resultant 2,6-dideoxy-2-haloglycosyl halides were established on the basis of their lH N M R spectra and optical rotations. It was carefully verified that (7) Horton, D.; Priebe, W. U S . Patent 4427664, Jan. 24, 1984. Horton, D.; Priebe, W.; Varela, 0. Abstr. Pap. Int. Carbohydr. Symp., I l t h , 1982, Abstr. 1-41. Horton, D.; Priebe, W. Carbohydr. Res. 1985, 136, 391-396. Horton, D.; Priebe, W.; Varela, 0. Carbohydr. Res. 1985,144, 305-315. (8) Bock, K.; Lundt, I.; Pederson, C. Corbohydr. Res. 1984, 130, 125-134.
0 1986 American Chemical Society
Horton et al.
3480 J . Org. Chem., Vol. 51, No. 18, 1986 Scheme I H
X
R@
F
x
+ "R l
X
+
x&;R
R'
R2
OAc
AcO 1, R' = OAc;
OAC
OAc
X
R2
R2
R2
+ RFTx 'R OAc
X
a-L-gluco
P-L-gluCo
a-L-manno
P-L-manno
la le
lb
IC
Id
a-1.-galacto
(3-L-galacto
U-L-talo
6-L-talo
2a 2e
2b
2c 2f
2d
RZ = H
X = c1 X = Br 2, R' = H; R 2 = OAc x = e1 X = Br
If
the a anomer. This conclusion was additionally confirmed no anomerization or other isomerization of the products by the long-range coupling constants J1,5 (0.7 Hz) and J1,3 took place during the isolation procedures. (0.5 Hz).The specific rotations of -78' for IC and + 5 8 O Derivatives of L-rhamnal diacetate (la, lb, IC, Id, le, for Id provide classical confirmation of these anomeric and If) showed large J4,5 values (9.5-10.0 Hz), indicating assignments. the trans disposition of these protons in the favored ' C 4 ( ~ ) In the dihalo derivatives 2a, 2b, 2c, and 2d, the l C 4 ( ~ ) conformation. A small value of J2,3(3.3-4.5 Hz) indicates conformation is also to be expected," and this was cona cis-equatorial-axial relationship for these two protons firmed by the large J2,3 (10.7) values for 2a and 2b. The and thus the L-manno configuration for IC, Id, and If; trans-diaxial orientation of H-2 and H-3 is only possible larger values of J2,3(-10.0 Hz) denote a trans-diaxial for the galacto isomers in the anticipated ' C 4 ( ~confor) disposition for H-2 and H-3 and, therefore, the L-gluco mation. The configuration at C-1 of compound 2b was configuration for la, lb, and le. Thus, from the four established on the basis of the large Jl,2value (9.1 Hz), isomers obtained by chlorination of L-rhamnal diacetate which indicates that 2b (["ID -32') is the /3 anomer of (l),two of them (la and lb) have the L-gluco configuration 3,4-di-O-acetyl-2-chloro-2,6-dideoxy-~-galactopyranosyl and the two others the L-manno (IC and la). chloride. The smaller value of J1,2 (3.7 Hz) and the higher The configuration at the anomeric center was established chemical shift for H-1 and H-5 (A6H-5*-b0.59 Hz) than in by 'H NMR by considering a combination of the following compound 2b support the assignment of 2a ([cf]D -230') factors: (a) the Jl,2coupling constant; (b) the chemical as the a anomer. Furthermore, compound 2a, as for the shift of H-1; (c) the variation of chemical shift of H-3 and other a anomers, shows long-range coupling between H-1 H-5; and (d) long-range coupling constants. Optical roand H-5 (J1,5 = 0.7 Hz). The similar values of Jl,zfor 2c tation was used to confirm proposed assignments. The and 2d do not permit configurational assignment at the orientation of H-1 and H-2 in l b was readily established anomeric center, but the chemical shifts of H-1 (6.30 ppm as trans-diaxial because of the large value of Jl,z (9.1 Hz), for 2c and 5.55 for 2d) and H-5 (AcJHW5&= 0.58 ppm) demonstrating that compound l b is the /3 anomer (["ID -45O) of 3,4-di-0-acetyl-2-chloro-2,6-dideoxy-~-gluco-indicate compound 2c to be the a anomer of 3,4-di-Oacetyl-2-chloro-2,6-dideoxy-~-talopyranosyl chloride, and pyranosyl chloride and the other gluco isomer (la) is the 2d ( [ a ] D + 4 O ) the p anomer. The J1,5 coupling of 0.8 Hz a anomer ([cfI2OD -219'). The chemical shift of H-1 (6.08 in 2c supports these assignments. ppm for la and 5.26 for lb) confirmed these assignments, Bromination of L-rhamnal diacetate (1) afforded two because equatorial protons at C-1 are generally shifted to compounds, l e and If. The J2,3 coupling constant of 10.7 lower field than axial o n e ~ . ~Furthermore, *l~ the difference Hz for le indicates the L-gluco configuration. The value in chemical shift of H-3 (A6,-,.-b) and H-5 ( A 6 ~ . . ~ a - b ) beof J2,3(3.8 Hz) for If dictates a cis-equatorial-axial oritween la and l b is, respectively, 0.27 and 0.58 ppm, inentation for H-2 and H-3 and the L-manno configuration. dicating a downfield shift of H-3 and H-5 in l a because The L-gluco isomer l e ([a]D -283') is the a anomer because of the parallel 1,3-interaction with axial chlorine at C-1. value of 3.6 Hz demonstrates a cis relationship the Jl,2 Finally it was observed that all a anomers show a longbetween H-1 and H-2; the p anomer should display a much range Jl,5 coupling constant of 0.5-0.8 Hz (in some inlarger coupling constant, similar to that in l b (9.1 Hz). stances also J1,3 of 0.5 Hz was observed); this last finding The a configuration assigned at C-1 for l e was confirmed constitutes one more line of proof for the assigned conby the chemical shift of H-1 (6.39 ppm), which shows figuration at the anomeric center. Compound l a showed downfield shielding of 0.31 ppm because of the effect of J1,5 = 0.7 Hz and J1,3 = 0.5 Hz. bromine, and by the chemical shift of H-5 in le, which is The J192coupling constants for the L-manno anomers very similar to that of l a (4.30 and 4.29 ppm, respectively). IC and Id are very similar and give no information about The chemical shift of H-1 in If, downfield by 0.47 ppm the configuration at C-1. However, on the basis of the with respect to H-1 of IC, suggests the a configuration for difference of chemical shifts of H-1, compound IC ( [ a ] D If (["ID -125O). Both compounds (IC and If) have iden-78O) was established as the a anomer (H-1,6.13 ppm) of 3,4-di-0-acetyl-2-chloro-2,6-dideoxy-~-mannopyranosyl tical chemical shifts for H-5 (4.21 ppm). Finally, the dibromides le and If show long-rangeJ1,5 coupling constants chloride and Id ( [ a ] D + 5 8 O ) as the anomer (H-1, 5.57 of 0.5 and 0.7 Hz. These couplings are observed only for ppm). In support of this, the A6H-5rd value of 0.57 ppm the a anomers of the 1,2-dichlorides,and this observation indicates that isomer IC, exhibiting a downfield shift, is constitutes yet one more proof for the anomeric assignments. (9) Bundle, D. R.; Lemieux, R. U. Methods Carbohydr. Chem. 1976, 7,79-86. (10)Descotes, G.;Chizat, F.; Martin, J. C. Bull. SOC.Chim. Fr. 1970, 2304-2309.
(11)Durette, P.L.;Horton, D. Adu. Carbohydr. Chem. Biochem. 1971, 26,49-125.
J . Org. Chem., Vol. 51, No. 18, 1986 3481
Halogenation of 1,5-Anhydrohex-l-enitols
Table I. Chemical Shifts (6, ppm) and Coupling Constants (Hz) for the Dihalides la-d, 2a-d, le, If, 2e, and 2f H-1 compd
H-2 J1,5
J1,Z
J2,3
6.08
la
0.7
0.7
1.5
Id 2a
4.57 0.7
2b
0.8
1.1
2d
4.4 3.3 4.11
6.39 3.6
If
0.5
10.7
0.7
3.8
2e
0.7
11.2
0.7
4.5
0.4
9.9
0.5
9.6
0.4
1.3
0.5
1.7
5.45 5.41
2.10, 2.09
3.64
1.31
2.10, 2.06
4.53
1.20
2.16, 2.05
3.94
1.26
2.19, 2.07
4.49
1.27
2.17, 2.09
3.91
1.30
2.10, 2.04
4.30
1.25
2.07, 2.06
4.21
1.31
2.10, 2.09
4.53
1.20
2.16, 2.05
4.47
1.28
2.18, 2.09
6.2 6.3 0.4
6.5
0.4
6.5
5.32
3.6
1.28
6.4
5.31
5.59 1.0
4.21
6.5
5.31
3.2 4.65
6.80 1.0
0.4 4.83
9.8 4.37
6.52 3.4
2f
5.51 9.2
4.87
6.60 1.3
5.27
1.6
2.09, 2.04
6.4
0.5 1.8 5.14-5.21 1.5
3.6
1.29
6.5
5.25
5.62 1.1
0.4
1.1
4.34
5.55 1.6
le
5.05
4.40
6.30
2C
1.2
3.3
10.7
3.71
6.2 5.33
0.5
3.2
4.09
5.28 9.1
5.21 9.5
5.39
10.7
CH3CO 2.07, 2.05
6.3
9.8
5.01
H-6 1.24
6.2 5.25
0.5
4.29
J5,6
6.2
9.8
9.8 4.41
6.19 3.7
10.0
5.56
3.6
J1,4
4.83
9.9
3.8
5.57 1.2
0.5
9.4 4.64
6.13
IC
J4,5
5.20
10.1
H-5
4.82
9.3 3.91
9.1
J1,3
5.47
10.5
5.26
H-4
J3,4
J2,4
4.13
3.7
lb
H-3
Table 11. Product Distribution in the Chlorination of 1 and 2 in Various Solvents4
solvent
b
CC14 (CHzC1)z CHBNOz
2.23 10.37 38.57
E - L-gIUCO
1.
76 55 35 ~ - L - ~ Q ~ Q 2ac ~
CC14 (CHzCUz CH3NOz
2.23 10.37 38.57
65 55 53
8-L-gluco lb
3 10 23 oP-L-galacto 2b 17 30 35
E-
L-mmno
B- L-manno ld
IC
4 14 24 a-L-talo 2c 6 10 12
17 21 18 P-L-talo 2d 12 5
