Received 8 May 2012
aSchool of Chemical and Material Engineering, Jiangnan University, Wuxi, Jiangsu 214122, People's Republic of China, and bKey Laboratory of Nuclear Medicine, Ministry of Health, Jiangsu Key Laboratory of Molecular Nuclear Medicine, Jiangsu Institute of Nuclear Medicine, Wuxi 214063, People's Republic of China
The title compound, C16H24O10·0.11H2O, is a key intermediate in the synthesis of 2-deoxy-2-[18F]fluoro-D-glucose (18F-FDG), which is the most widely used molecular-imaging probe for positron emission tomography (PET). The crystal structure has two independent molecules (A and B) in the asymmetric unit, with closely comparable geometries. The pyranose ring adopts a 4C1 conformation [Cremer-Pople puckering parameters: Q = 0.553 (2) Å, = 16.2 (2)° and = 290.4 (8)° for molecule A, and Q = 0.529 (2) Å, =15.3 (3)° and = 268.2 (9)° for molecule B], and the dioxolane ring adopts an envelope conformation. The chiral centre in the dioxolane ring, introduced during the synthesis of the compound, has an R configuration, with the ethoxy group exo to the mannopyranose ring. The asymmetric unit also contains one water molecule with a refined site-occupancy factor of 0.222 (8), which bridges between molecules A and B via O-HO hydrogen bonds.
To date, 2-deoxy-2-[18F]fluoro-D-glucose (18F-FDG) [(III); see Scheme] has been the most widely used molecular-imaging probe in positron emission tomography (PET) (Alavi & Reivich, 2002; Nutt, 2002). Due to its structural similarity to glucose, 18F-FDG is used to measure the cellular uptake and metabolic rates of local glucose in physiological or pathophysiological conditions. Nowadays, 18F-FDG-PET is used not only in research laboratories to address questions of scientific interest, but also in hospitals as a routine clinical diagnostic tool in cardiology, neurology and oncology (Gambhir et al., 2001).
The title compound, (I), is one of the key intermediates during the synthesis of 18F-FDG. Hydrolysis of the 1,2-orthoacetate groups of (I) with 1 M aqueous HCl, followed by triflation with Tf2O-pyridine (Tf is triflate or trifluoromethanesulfonate), yields 1,3,4,6-tetra-O-acetyl-2-O-trifluoromethanesulfonyl--D-mannopyranose (mannose trifluoromethanesulfonate), (II). Nucleophilic [18F] fluorination and acidic/basic deprotection of (II) then provides 18F-FDG, (III) (see Scheme) (Toyokuni et al., 2004). Although hydrolysis of (I) gives the desired 1,3,4,6-tetra-O-acetyl--D-mannopyranose product, the reaction also produces an undesired 2,3,4,6-tetra-O-acetyl--D-mannopyranose by-product, and the yieldcan help us to understand not only its chemical properties but also the mechanism of the hydrolysis reaction.
The crystal structure of (I) has two independment molecules in the asymmetric unit (molecules A and B; Fig. 1) and a water molecule with a refined site-occupancy factor of 0.222 (8). In molecule A, the C-C bonds within the pyranose ring have bond lengths ranging between 1.510 (3) and 1.523 (3) Å, in accordance with those in a survey of 27 pyranoid rings reported by Arnott & Scott (1972). The exocyclic C1-O1 bond has a length of 1.398 (3) Å which is the shortest of the C-O bonds in the pyranose ring. The C1-O5 and C5-O5 bond lengths are 1.437 (2) and 1.422 (3) Å, respectively, suggesting that the C1-O1 bond adopts an equatorial orientation according to Arnott & Scott (1972). The bond angles at the C atoms of the ring vary between 107.28 (15) and 114.27 (17)°, with the angles at C1 and C2 larger than the expected values of 109.2 and 110.5°, respectively, and the angle at C5 smaller than the expected value of 110.0°. Although the O5-C1-O1 angle [106.47 (18)°] appears normal for equatorial C1-O1, the C5-O5-C1 angle [113.45 (17)°] is somewhat larger than the expected value of 112.0°. These data indicate that C1-O1 is distorted from an idealized equatorial orientation and is bisectional. This could be explained by the presence of the ethoxyethylidene group at the 1,2-positions. The pyranose ring of molecule B exhibits similar structural features to those of molecule A.
The endocyclic torsion angles in the pyranose rings vary from 40.2 (3) to 65.3 (2)° (absolute values), indicating that the six-membered rings are distorted from an idealized 4C1 chair conformation. The conformation in (I) is different from the skew conformation (3S5) reported for 3,4,6-tri-O-acetyl-l,2-O-[1-(exo-ethoxy)ethylidene]--D-glucopyranose (Heitmann et al., 1974). Further insight into the distortions is obtained from the Cremer-Pople puckering parameters (Cremer & Pople, 1975): for molecule A, Q = 0.553 (2) Å, = 16.2 (2)° and = 290.4 (8)°; for molecule B, Q = 0.529 (2) Å, = 15.3 (3)° and = 268.2 (9)°. The extent of the ring distortion, embodied in the value of , is slightly greater for molecule A than for molecule B. The direction of the ring distortion, embodied in the value of , shows a distortion towards a BC2,C5 conformation for molecule A and C1SC5 for molecule B (Jeffrey & Yates, 1979).
In the five-membered ring of molecule A, the C2-O2-C7-O1 torsion angle [5.2 (2)°] indicates an approximately planar structure for these four atoms (r.m.s. deviation from the mean plane = 0.023 Å). Atom C1 deviates from the mean plane by 0.535 (3) Å. In molecule B, the corresponding values are 3.3 (2)°, r.m.s. deviation = 0.015 Å and deviation of C1' = 0.525 (3) Å. Thus, the conformation of the dioxolane ring is an envelope, with Cremer-Pople parameters of Q = 0.349 (2) Å and = 43.2 (3)° for molecule A, and Q = 0.339 (2) Å and = 40.7 (4)° for molecule B.
The bond lengths and angles in the acetoxy groups and acetoxymethyl substituent all appear normal, although the bond angles at C6 [107.47 (16)°] and C6' [107.26 (18)°] are, respectively, 4.3 and 4.5° smaller than the average bond angle (111.8°) found in 23 other pyranose derivatives (Arnott & Scott, 1972). The three substituents all adopt the same equatorial orientation as in the parent compound, -D-mannose, due to the 4C1 conformation of the mannopyranose ring. Similar to what is observed for many acetylated sugars, the acetyl groups directly connected to the sugar ring [i.e. C(ring)-O-C(=O)] are close to coplanar with the H atom bound to the same ring C atom (Haines & Hughes, 2007). The O5-C5-C6-O6 [-71.0 (2)°] and O5'-C5'-C6'-O6' [-63.2 (2)°] torsion angles indicate a gg arrangement of the exocyclic C6-O6 (C6'-O6') bond (i.e. H5 anti to O6 and H5' anti to O6').
There are some structural differences on the O1 (O1') and O2 (O2') sides of the five-membered dioxolane ring. In molecule A, the C1-O1 bond [1.398 (3) Å] is the shortest of the four C-O bonds involving O1 and O2, and the other three bonds are almost identical in length [average 1.435 (2) Å]. The O1-C7-C8 angle [113.10 (18)°] is larger than O2-C7-C8 [110.08 (18)°] and the O1-C7-O9 angle [108.90 (17)°] is smaller than O2-C7-O9 [112.19 (17)°]. The C7-O9 bond [1.380 (3) Å] is the shortest of the three ether bonds at C7, with the C7-O1 and C7-O2 bonds effectively identical [average 1.435 (2) Å]. The same structural features are found in molecule B. This might give an indication of why hydrolysis of (I) yields almost equivalent amounts of the two major isomeric products (Toyokuni et al., 2004).
Atom C7 (C7') is the only new chiral centre formed during the synthesis of (I) (Toyokuni et al., 2004). It has an R configuration, with the ethoxy group exo to the mannopyranose ring. The 1H NMR spectrum (CDCl3) gives a single peak at 1.75 p.p.m. for the C8 methyl group, consistent with the conclusion of earlier 1H NMR studies of glucopyranose 1,2-(alkyl orthoacetates) (Lemieux & Morgan, 1965), that the diastereomer for which the C8 methyl H atoms resonate at a lower field has the configuration in which the alkoxy group is exo to the pyranose ring.
The bicyclic ring structure of (I) has five pendant groups, each of which terminates in a methyl group. In the packing of the crystal structure, the pendant groups of one molecule lie between pendant groups from other molecules (Fig. 2). The partially occupied water molecules bridge between molecules A and B, forming hydrogen bonds (Table 1). There is no direct hydrogen bonding between molecules A and B. Thermogravimetric analysis was carried out for (I), but it is difficult to detect any water loss prior to decomposition of the compound, since the solvent water accounts for only approximately 0.52% of the mass.
| || Figure 1 |
The asymmetric unit of (I), showing the atom-numbering scheme. Displacement ellipsoids are drawn at the 50% probability level.
| || Figure 2 |
The molecular packing of (I), viewed along the a axis. Dashed lines indicate hydrogen bonds.
The title compound was synthesized according to the literature procedure of Toyokuni et al. (2004). The crude product was recrystallized from ethanol (m.p. 375.0-375.8 K; literature value 374-376 K). 1H NMR (400 MHz, CDCl3): 1.18 (t, J = 7.1 Hz, 3H, MeCH2O), 1.75 (s, 3H, MeCO2), 2.05 (s, 3H, Ac), 2.07 (s, 3H, Ac), 2.12 (s, 3H, Ac), 3.56 (m, 2H, MeCH2O), 3.68 (ddd, J = 9.5, 4.9 and 2.7 Hz, 1H, H5), 4.14 (dd, J = 12.1 and 2.6 Hz, 1H) and 4.24 (dd, J = 12.1 and 4.9 Hz, 1H, 2 H6), 4.60 (dd, J = 3.9 and 2.7 Hz, 1H, H2), 5.15 (dd, J = 9.9 and 4.0 Hz, 1H, H3), 5.30 (t, J = 9.7 Hz, 1H, H4), 5.48 (d, J = 2.5 Hz, 1H, H1).
All H atoms bound to C atoms were placed geometrically and refined using a riding model, with C-H = 0.98-1.00 Å and Uiso(H) = 1.2Ueq(C). The H atoms of the water molecule were placed so as to form hydrogen bonds to atoms O62 and O62', with O-H = 0.85 Å, then refined as riding with Uiso(H) = 1.2Ueq(O). In the absence of significant anomalous scattering, the absolute structure could not be determined. Friedel pairs were not merged. The absolute structure is assigned on the basis of unchanging chiral centres in the synthetic procedure (Toyokuni et al., 2004).
Data collection: CrystalClear (Rigaku, 2008); cell refinement: CrystalClear; data reduction: CrystalClear; program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: DIAMOND (Brandenburg, 1999); software used to prepare material for publication: SHELXL97.
Supplementary data for this paper are available from the IUCr electronic archives (Reference: BI3042 ). Services for accessing these data are described at the back of the journal.
This research was supported by Jiangsu Province Science and Technology Support Program for Social Development (grant No. BE2010623).
Alavi, A. & Reivich, M. (2002). Semin. Nucl. Med. 32, 2-5.
Arnott, S. & Scott, W. E. (1972). J. Chem. Soc. Perkin Trans. 2, pp. 324-335.
Brandenburg, K. (1999). DIAMOND. Crystal Impact GbR, Bonn, Germany.
Cremer, D. & Pople, J. A. (1975). J. Am. Chem. Soc. 97, 1354-1358.
Gambhir, S. S., Czernin, J., Schwimmer, J., Silverman, D. H., Coleman, R. E. & Phelps, M. E. (2001). J. Nucl. Med. 42, 1S-93S.
Haines, A. H. & Hughes, D. L. (2007). Carbohydr. Res. 342, 2264-2269.
Heitmann, J. A., Richards, G. F. & Schroeder, L. R. (1974). Acta Cryst. B30, 2322-2328.
Higashi, T. (1995). ABSCOR. Rigaku Corporation, Tokyo, Japan.
Jeffrey, G. A. & Yates, J. H. (1979). Carbohydr. Res. 74, 319-322.
Lemieux, R. U. & Morgan, A. R. (1965). Can. J. Chem. 43, 2199-2204.
Nutt, R. (2002). Mol. Imaging Biol. 4, 11-26.
Rigaku (2008). CrystalClear. Rigaku Corporation, Tokyo, Japan.
Sheldrick, G. M. (2008). Acta Cryst. A64, 112-122.
Toyokuni, T., Kumar, J. S. D., Gunawan, P., Basarah, E. S., Liu, J., Barrio, J. R. & Satyamurthy, N. (2004). Mol. Imaging. Biol. 6, 324-330.