3-Deoxy-1,2-di-O-isopropylidene-5-O-tosyl-d-threo-pentofuranose

In the crystal structure of the title compound, C15H20O6S, the two independent molecules crystalllize in a chiral setting with two different conformations, twisted 4 T 3 and envelope 4 E, for the furanose rings. Weak C—H⋯O contacts strengthen the crystal structure.

In the crystal structure of the title compound, C 15 H 20 O 6 S, the two independent molecules crystalllize in a chiral setting with two different conformations, twisted 4 T 3 and envelope 4 E, for the furanose rings. Weak C-HÁ Á ÁO contacts strengthen the crystal structure.
We wanted to firmly establish their structures, due to a possibility of enolization of the ulose and concomitant inversion of configuration at the C4 position during formation of the tosylhydrazone.
A correct absolute structure of the title compound was important for the further synthetic work. Because of that, we have selected Cu Kα radiation to ensure unambigous determination of the absolute structure.
The six-membered phenyl rings in both molecules are flat within 0.01 Å.
It is visually obvious ( Fig. 3 and Fig. 4) that the conformations of the five-membered rings differs in the two independent molecules A and B. A quantitative analysis of the ring conformations was performed using the method of Cremer and Pople (Cremer & Pople, 1975;Boeyens & Dobson, 1987) for the calculation of parameters of puckering. In molecule A, the polar parameters for the furanose ring and adjacent five membered ring are Q = 0.289 (3) and 0.312 (2) Å, Φ = 122.9 (5)° and 119.7 (5)°, respectively. These suggest a twisted 4 T 3 conformation for the furanose ring (ideal Φ = 126°), slightly distorted towards envelope (Φ = 108°). The substituent ring also has a twisted conformation (Fig. 3).
In the structure of 1,2-di-O-isopropylidene-5-O-tosyl-D-xylofuranose which differs from the title compound in one hydroxy group, the polar parameters are Q = 0.352 (3) Å, Φ = 288.8 (5)°; see refcodes RUWDES and RUWDES01 (Cox et al., 1997). This makes the conformation an almost exact 3 E envelope, but with a different carbon atom in the corner than in the case described here. Obviously, the furanose ring conformation is highly flexible and is easily influenced even by weak intermolecular interactions. A short intramolecular contact is present between sulfonyl O atoms O5 and O15 and neighboring hydrogen atoms of the adjacent respective phenyl rings (see Table 1). This is quite common for aryl sulfonyls and the majority of these compounds exhibit these intramolecular interactions (mean H···O distance is 2.533 Å for more than 2500 analogous structures listed in the Cambridge Structural Database (Allen, 2002)). It may additionaly stabilize the conformation of the molecule. Only weak intermolecular C-H···O contacts (Table 1) exist between neighboring molecules.

Refinement
Final refinement was performed using TWIN/BASF type resulting in BASF = 0.00458. Analysis of the absolute structure using likelihood methods (Hooft et al., 2008) was performed using PLATON (Spek, 2009); 2059 Bijvoet pairs were employed. The results confirmed that the absolute structure had been correctly assigned: the probability that the structure is inverted and probability of racemic twinning being statistically zero. All H atoms were positioned geometrically with C -H =0.95-1.00 Å and U iso (H) = 1.2 or 1.5 U eq (C). Rotating group refinement (AFIX 137) was employed for all methyl groups.

Figure 4
Envelope conformations of the five-membered rings in molecule B. Mean planes through atoms O11, C1, C2, C4 (yellow) and C1, C2, O2, and O3 (green).  View of the title compound showing displacement ellipsoids at the 50% probability level.  Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.