M. Michalska, G. Snatzke

Circular Dichroism of Monosaccharide Dichalcogenides Maria Michalska and Giinther Snatzke"

Laboratory of Organic Chemistry, Institute of Chemistry, Medical Academy, PL-90151 Lodi (Poland)

Lehrstuhl fur Strukturchemie der Universitat Bochum D-4630 Bochum (F.R. Germany)

Received July 21, 1986

The CD spectra of four diselenides (1,2,4,5) and one ditelluridc (3) with sugar end groups have been measured under different conditions. One additional transition of diselenides has been iden- tified around 360 to 380 nm, which has analogy io the spectrum of the ditelluride. The assignment of the two Cotton dects below 250 om has been reversed, Exchange of the SeZ moiety against Te2 shifts all Cotton effects bathochromically, but they retain their signs.

The preferred torsional angle' -3) R - Z - Z - R (Z = 0, S, Se, Te) of a dichalcogenide is approximately go", each molecule is, therefore, inherently chiral; if R is an achiral moiety then the two entities are isoenergetic, and the sub- stance is optically inactive because it is racemic (50% P and 50% M rotamer). If, however, R is a chiral substituent, then the P and M form are diastereomers and in equilibrium their ratio will in principle deviate from 1 : 1.

Circular dichroism (CD) is a bulk property of a huge ensemble of molecules and, therefore, a CD curve of a sub- stance consisting of flexible molecules is a superposition of the CD curves of the individual stereoisomers. Even for a rigid molecule for practical reasons it is useful to factorize each Cotton effect into several contributions, as those de- riving from the inherently chiral chromophore, those from a chiral perturbation of the chromophore by its chiral en- vironment (through-bond/through-space), etc. '). In general, contributions of the first type are quite strong and override the others, but in the case of dichalcogenides the first two transitions become degenerate5) for a torsional angle R - Z - Z - R equal to go", thus the associated two strong Cotton effects, which have opposite signs, should compen- sate each other, effects of the second-mentioned type becom- ing then dominant. These latter can be described by appro- priate sector rules, whereas those of the first type are gov- erned by helicity rules 'I.

Whereas optically active disulfides are known since a long time (best-known example: cystine), corresponding diselen- ides or ditellurides are scarcely known7,*). Recently in L6di we succeeded in the synthesis of several such monosaccha- ride dichalcogenides by treating sugar epoxides with 0,O- dialkylphosphoroselenoates (2,4, 5)9) or by methanolysis of 5,5-d~methyl-2-oxo-2-[(2,3,4,6-tetra-O-acetyl-~-~-glucopyr- anosy1)selenol- or -telluro]-1,3,2-dioxaphosphorinanes (1, 3)103"). The CD data of five examples are presented in Table 1. In three of them (1, 2: Sez; 3: Te2) the Z-Z bridge

179

Circabud&&oismus VOD M o m s c c h & ' Dkbakogenideo Die CD-Spektren von vier Dwleniden (1, 2, 4, 5) und einem Ditellurid (3) mit Zucker-Endgruppen wurden unter verschiede- nen Bedingungen gemessen. En zusiitzlicher obergang van Di- seleniden konnte urn 360 bis 380 om identifiziert werden, der seine Analogie in dcm Spektrum des Ditellurides Cindet. Die Zuordnung der zwei Cotton-Effekte unterhalb von 250 om wurde umgekehn. Austausch der &Einheit gem Te2 verschiebt alle Cotton-Ef- fekte bathochrom, sie behalten aber ihre Vomichen.

is between the anomeric positions of two P-glucose moieties, whereas in two (4, 5: Sez) this bridge is attached to the C-6 atoms of glucofuranose.

1: 2 =Se, R = A c

' 2 3: Z = T e l R = A c

\ ' 2

Laur7) has observed four Cotton effects for an open-chain optically active diselenide in similar wavelength regions as we have found for 1, 2, 4, and 5 only the position of the Cotton effect at shortest wavelengths depends significantly on the partial structure of these compounds: if the Se2 bridge is connected to a CH2 group (ref.7), 4, 5), then the CD maximum is around 210 nm or at even longer wavelengths, whereas if it is next to the ring oxygen, then this Cotton effect is extremely large and is centered around 195 nm (Figure 1). Obviously this hemiseleno acetal moiety forms already one single chromophore by combination of the AO's from 0 and Se as, e.g., is well-demonstrated by PES for dithioacetals12). The CD curve at longest wavelengths is very broad and can be fitted one single Gaussian-shaped curve, which would have, however, an unusual large half-band width') (appr. 6200 cm-'), so we ascribe it rather to the presence of two Cotton effects of the same sign.

Liebigs Ann. Chem. 1987, 179- 181 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1987 0170-2041/87/0303 -0179 $ 02.50/0

180 M. Michalska, G. Snatzke

Table 1. CD data of diselenides and ditellurides

1 Cnml (A&nlax) Sub- Solvent")

stance Temperature

1 AC, 24°C EPA, 20°C

EPA, -140°C

2 ET, 24°C

AC, 24°C

3 AC, 24°C EPA, +2O"C EPA, - 180°C

4 AC. 24°C

5 AC, 24°C

315 (-0.54), 273 (+0.17), 194 (-36.6), 320 (-0.81), 275 (+0.48), not measured at shorter wavelengths

317 (-0.59), 270 (+0.43), negative below 250 nm 311 (-0.66), negative minimum 272

312 (-l.lO), 275 (+0.73)

(-0.07), 196 (-23.1) 386 (-0.09), 252 sh (-4), 218 (-17.2) 392 (-0.45), 248 sh (-4.4) 418 sh (-0.23), 358 (- 1.44), 274 (+ 1.90), 250 (-9.22), negative below 240 nm 331 (+0.63), 277 (-0.45), 229 sh (+2.1), 210 (+5.90) 330 (+0.59), 279 (-0.22), 229 sh (+1.93), 211 (+4.3)

a) AC = acetonitrile. - EPA = diethyl ether/isopentane/ethanol. 5:5:2. - ET = ethanol.

negative for 1 and 2, but positive for 4 and 5. The assignment of the proposed additional transition around 360 to 380 nm remains uncertain (maybe singlet ---$ triplet?).

The next two Cotton effects around 230 and 200 nm have been tentatively assigned7) to the n- +o*(C - Se) and o(Se - Se)+ o*(Se - Se) transition, respectively. We would like to reverse this assignment, however, since the excitation from the n- orbital into the energetically lowest combina- tion of the C-Se o* orbitals (A symmetry) is associated with a stronger magnetic transition moment, and it is always the fourth CD band at shortest wavelengths, which is by far the largest one.

The ditelluride 3, which is the strict isologue of 1, shows - again in agreement with the literature7) - more Cotton effects than the latter, although we could not observe all of them in the same solvent or at room temperature alone (cf. Figure 2). As for the diselenides the CD band at longest wavelengths is very broad, but here below - 140°C clearly two separate Cotton effects (both of negative sign) are seen, vic. around 410 and 358 nm. This can be taken as an indi- cation that also in the case of the diselenides the broad first CD band consists actually of two Cotton effects of the same sign. It is interesting to note that this first CD maximum is

For the conformationally rigid trans-2,3-diselenadecalin Law7) has observed two Cotton effects at 364 and 275 nm, which can be associated with the n--+o*(Se- Se) and the n + -+ o*(Se - Se) transitions. Their signs are in agreement with the prediction from the sign of the torsional angle C - Se - Se - C (if the same rule holds as for isologous di- sulfides). Since the AE,,, values of these CD bands of 1, 2, 4, and 5 are still relatively large (appr. 1/4 of those found for the diselenadecalin), with the first band somewhat shifted hypsochromically, one has to conclude that the Se - Se tor- sional angle is widened, but not yet around 90 ', in which case the CD should be much smaller. From the signs of the Cotton effects follows then that these torsional angles are

Figure 1. CD of 1 (-----) and 4 (-) in acetonitrile. Curves 5 times enlarged above 250 nm, 5 times compressed for 1 below 250 nm.

shifted from 393 nm at room temperature to 358 nm at - 180°C. This is much too large for an equilibrium between differently solvated species and must, therefore, be ascribed to a change of the R-Te-Te-R torsional angle or/and the equilibrium between two species of opposite Cotton ef- fects lying very close to each other. Such a strong depend- ence of the position of the longest UV and CD maximum on the respective torsional angle is well-known from experiments 13) and understood from theory 5, (e. g. maxi- mum absorption for disulfides with torsional angle C-S-S-C % 0" at 334 nm, with ~ 9 0 at 250 nm).

The comparison of the CD curves of 1 and 2 with that of 3 (of the same relative and absolute configuration) shows

Liebigs Ann. Chem. 1987, 179-181

Circular Dichroism of Monosaccharide Dichalcogenides 181

2. a

Figure 2. CD of 3 in acetonitrile (-----) and in EPA (5:5:2) (see Table 1) at +20"C (-) and at - 18O'C ( 0 0). Curves above approximately 300 nm 10 times enlarged.

furthermore, that all Cotton effects are shifted bathochrom- ically, but retain their signs. This strongly indicates that all corresponding transitions have analogous parentages. Sim- ilar results have been found when the chiroptical properties of corresponding diselenides and disulfides of the same ab- solute configuration are compared, so that most probably f and higher orbitals are not involved in these transitions.

1) s - s: e.g, B, Beagley, K, T, McAloon, T ~ ~ ~ ~ . ~~~~d~~ sot, 67 (1971) 3216. Se-Se: e.g. P. DAntonio, C. George, A. H. Lowrey, J. Karle, J , Chem, Phys, 55 (,971) 1071, Te-Te: e.g. M. R. Spirlet, G. V. d. Bossche, 0. Dideberg, L. DuPont, A m CrYstallogr., Sect. B, 35 (1979) 1727.

dl E. Charney, The Molecular Basis of Optical Activity, J. Wiley, N~~ York 1979,

'' Cf. R. W. Woody, Tetrahedron 29 (1973) 1273 and references cited there.

6J Cf. G. Snatzke, Angew. Chem. 91 (1979) 380, Angew. Chem. Znt. Ed. End. 18 (1979) 363; Chem. Unserer Zeit 15 (1981) 78; 16 (1982) 160.

7) p, H, A, Laur in Proceedings of the Third International sym- posium on Organic Selenium and Tellurium Compounds (D. Cagniant, G. Kirsch, Ed.), p. 217, Universite de Metz 1981.

*) A few data have been mentioned: G. Snatzke in The Chemistry of Organic Selenium and Tellurium Compounds (S. Patai, Z. Ra- poport, Ed.), Vol. 1, p. 667 (1986).

91 M, Michalska, W, ~ ~ d ~ l ~ k ~ , Tetrahedron 37 (1981) 2989. lo) M. Michalska, J. Michalski, I. Orling-Krgzel, Pol. J . Chem. 53

(1979) 253. M. Michalska, J. Czyzewska-Chlebny, J. Chem. Soc., Chem. Commun. 1985, 693. R. Gleiter, J. Spanget-Larsen, Top. Curr. Chem. 86 (1979) 139.

M. M. thanks the Heinrich-Hertz-Stiftung for a grant, G. S. the Deutsche Forschungsgemeinscha~ and the Fonds der Chemischen Zndustrie for financial support, Mr. U. Wagner for skilful measure- ments.

Experimental The CD-spectra were recorded with an ISA-Jobin-Yvon Dichro-

graph Mark 111 in concentrations of approximately 0.1 to 0.3 mmOl/l. Curves were smoothened by the Golay-Savitzky algorithm with a PDP-8/e computer.

CAS Registry Numbers

5: 105139-50-2 [116/86]

13) G. Bergson, Ark. Kemi 12 (1958) 233; 18 (1962) 409. 1: 103730-31-0 2: 105122-22-3 ,I 3: 105122-23-4 14: 105122-24-5 1

Liebigs Ann. Chem. 1987, 179-181