2,3,5- Tri - O -benzyl- D -xylofuranose

: The synthesis and crystallization of 2,3,5- tri - O -benzyl- D -xylofuranose permitted us to isolate the alpha anomer with a small contamination of the beta form (ca 10%), whose ﬁrst crystallographic structure obtained in the P 2 1 2 1 2 1 space group was determined at 100 K up to a resolution of sin θ max / λ = 0.71 Å − 1 and reﬁned to an R 1 value of 0.0171 with a Hirshfeld atom reﬁnement (HAR) approach. − Heteronuclear Single (HSQC)] experiments. 1 H NMR (400 MHz) chemical shift values are listed in parts per million (ppm), relative to the corresponding nondeuterated solvent. Data are reported as follows: chemical shift (ppm on the δ scale), multiplicity (s = singlet, d = doublet, and po = partially overlapped), coupling constant J (Hz), and integration. Acquisition of the 13 C NMR (101 MHz) spectrum of compound 1 ( α / β ca. 2:3, CDCl 3 solution) was performed on a broad-band decoupled mode. Chemical shift values are given in ppm, and are related to the corresponding nondeuterated


Introduction
Carbohydrates are ubiquitous key biomolecules that are involved in numerous fundamental biological events. They are an integral part of living cells acting as a vital source of energy, structural building blocks, and cell surface receptors or mediators [1].
Given the pivotal roles of mono-, oligo-and polysaccharides as well as glycoconjugates in the process of life, carbohydrate-mediated processes have therefore progressively been labeled as promising targets in drug discovery [2]. However, sugar derivatives are typically not ideal candidates for therapeutic purposes, as they are mostly unstable and susceptible to hydrolysis [e.g., hemiacetals (aldoses), hemiketals (ketoses) or acetals and ketals (O-glycosides)] [2].
As potential drug candidates, carbohydrate mimics have accordingly gained increased interest, because of their ability to modulate the activity of carbohydrate-processing enzymes, while circumventing the enzymatic and chemical instability of sugars in mammals [3][4][5]. Efforts have thus been devoted to using carbohydrates as synthetic intermediates to generate glycomimetics via chiron approaches.
One class of saccharide analogues that is currently attracting a surge of interest is that of iminosugars. These small molecules, also known as azasugars, are a class of compounds capable of acting as inhibitors or enhancers of many enzymes that act on glycosides, prominently on glycosidases [6][7][8], and glycosyltransferases [9]. They have become the most popular class of glycomimetics reported to date, being able, for instance, to regulate the folding and transport of glycoproteins (e.g., chaperone effect), to alter the glycosylation profile of eukaryotic cells, to interfere in the carbohydrate and glycoconjugate metabolism, and to stop virus attachment and the infection of host cells [6][7][8]10,11].

Synthesis
Our investigation commenced from D-xylose, which was converted into 2,3,5-tri-Obenzyl-α,β-D-xylofuranose 1 in 29% overall yield over three steps with only one purification (Scheme 1). Following typical conditions, D-xylose was inserted in a mixture of dry methanol and acetyl chloride, and the reaction mixture was heated at 30 • C for 3.5 h. The crude mixture was neutralized by addition of a basic resin. Compound 2 was then reacted with an excess of benzyl bromide and NaH in DMF at 20 • C for 20 h to afford methyl 2,3,5-tri-O-benzyl-α,β-D-xylofuranoside 3. Compound 1 (α/β ca. 2:3, CDCl 3 solution, 25 • C) was obtained through subsequent hydrolysis by treatment with a mixture of glacial acetic acid and 1 M aq. HCl solution at 80 • C for 17 h, followed by 4 h at 100 • C. enantiomers have not so far been determined. Herein, we report the synthesis of 2,3,5-tri-O-benzyl-α,β-D-xylofuranose (with a different method, see Section 2.1 below, and experimental part), and the crystallographic description (Ortep structure, 3-D Hirshfeld surface and crystal packing) of 2,3,5-tri-O-benzyl-D-xylofuranose mixing both anomers.

Synthesis
Our investigation commenced from D-xylose, which was converted into 2,3,5-tri-Obenzyl-α,β-D-xylofuranose 1 in 29% overall yield over three steps with only one purification (Scheme 1). Following typical conditions, D-xylose was inserted in a mixture of dry methanol and acetyl chloride, and the reaction mixture was heated at 30 °C for 3.5 h. The crude mixture was neutralized by addition of a basic resin. Compound 2 was then reacted with an excess of benzyl bromide and NaH in DMF at 20 °C for 20 h to afford methyl 2,3,5tri-O-benzyl-α,β-D-xylofuranoside 3. Compound 1 (α/β ca. 2:3, CDCl3 solution, 25 °C) was obtained through subsequent hydrolysis by treatment with a mixture of glacial acetic acid and 1 M aq. HCl solution at 80 °C for 17 h, followed by 4 h at 100 °C. Scheme 1. Synthesis of 2,3,5-tri-O-benzyl-α,β-D-xylofuranose.

Structural Commentary
The structure of D-xylofuranose 1 was confirmed beyond doubt by single crystal Xray diffraction studies performed upon crystals belonging to non-centrosymmetric Sohncke space group, P2 1 2 1 2 1 , at the Mo Kα radiation and at low temperature (100 K) with refined Flack parameter close to 0. Additional redundant measurements on a copper source but at room temperature and limited to a resolution of (sinθ/λ) = 0.6 Å -1 confirmed the significance of the Flack parameter in line with the starting D-xylose derivative.
The molecule found to occupy the asymmetric unit consists of a furanose ring substituted by two O-benzyl groups at C2 and C3, one Met-OBn group in position 4 and one hydroxyl group in position 1. Completion of the independent atom model (IAM) refinement with all hydrogen atoms constrained to geometric positions suggested slight disorder of the central O-benzyl group at C3 treated as static over two close sites and the hydroxy group at C1 in an equatorial position. Nevertheless, at that stage, the first residual peak in the Fourier difference synthesis-albeit low (<0.3e.Å 3 ) but almost twice the value of the second peak-was located in the axial position of C1 in the vicinity of the disordered O-benzyl group. The possibility that the two anomeric forms could be present inside the crystal was therefore checked. We ended up with the confirmation that crystallization of the neat mixture of alpha and beta anomers (unknown ratio) trapped the α-form in large majority over the β form (ca 0.89(1):0.11(1)). Carbohydrate hemiacetals can obviously exist in two anomeric isoforms, either in the solid or liquid state. In the case of compound 1, the α anomer is heavily favored in the crystal state, whereas in CDCl3 at 25 °C, the equilibrium is shifted towards the β form. One may therefore advocate for the selective crystallization of the α-over the β-anomer (e.g., optical resolution). However, there is no evidence that the neat amorphous solid did not already contain the two forms in an α/β ratio of 89:11, and its dissolution in CDCl3 slowly shifted the mutarotation equilibrium towards

Structural Commentary
The structure of D-xylofuranose 1 was confirmed beyond doubt by single crystal X-ray diffraction studies performed upon crystals belonging to non-centrosymmetric Sohncke space group, P2 1 2 1 2 1 , at the Mo Kα radiation and at low temperature (100 K) with refined Flack parameter close to 0. Additional redundant measurements on a copper source but at room temperature and limited to a resolution of (sinθ/λ) = 0.6 Å −1 confirmed the significance of the Flack parameter in line with the starting D-xylose derivative.
The molecule found to occupy the asymmetric unit consists of a furanose ring substituted by two O-benzyl groups at C2 and C3, one Met-OBn group in position 4 and one hydroxyl group in position 1. Completion of the independent atom model (IAM) refinement with all hydrogen atoms constrained to geometric positions suggested slight disorder of the central O-benzyl group at C3 treated as static over two close sites and the hydroxy group at C1 in an equatorial position. Nevertheless, at that stage, the first residual peak in the Fourier difference synthesis-albeit low (<0.3e.Å 3 ) but almost twice the value of the second peak-was located in the axial position of C1 in the vicinity of the disordered O-benzyl group. The possibility that the two anomeric forms could be present inside the crystal was therefore checked. We ended up with the confirmation that crystallization of the neat mixture of alpha and beta anomers (unknown ratio) trapped the α-form in large majority over the β form (ca. 0.89(1):0.11(1)). Carbohydrate hemiacetals can obviously exist in two anomeric isoforms, either in the solid or liquid state. In the case of compound 1, the α anomer is heavily favored in the crystal state, whereas in CDCl 3 at 25 • C, the equilibrium is shifted towards the β form. One may therefore advocate for the selective crystallization of the αover the β-anomer (e.g., optical resolution). However, there is no evidence that the neat amorphous solid did not already contain the two forms in an α/β ratio of 89:11, and its dissolution in CDCl 3 slowly shifted the mutarotation equilibrium towards a ca. 2:3 ratio in favor of the β form. Further diffraction measurement at low temperature and at extended resolution allowed us to perform a Hirshfeld atom refinement (HAR) to document in greater detail the structural parameters of 1 including those of the H atoms (Figures 1, 2, and 5, plus Tables 1-3, and Figure S1 and Table S1 in the supporting information (SI)).
The furanose group flexibility was found uncorrelated to the disorder of the central O-Bn group, a disorder that could be considered as essentially dynamic. While the major α form can be seen to be stabilized by an intramolecular hydrogen bond with O2, graph-set symbol S(5), (see Table 2) and benefit from developing a weak intermolecular H-bond (see below Figure 1, and crystal packing analysis in Figure 4), the minor β-form can find in O3 a 'lifeline' in a hydrophobic environment to make an intramolecular h-bond (graph-set, symbol S(6)).
Molbank 2022, 2022, x FOR PEER REVIEW 3 of 11 a ca. 2:3 ratio in favor of the β form. Further diffraction measurement at low temperature and at extended resolution allowed us to perform a Hirshfeld atom refinement (HAR) to document in greater detail the structural parameters of 1 including those of the H atoms (Figures 1, 2, and 5, plus Tables 1-3, and Figure S1 and Table S1 in the supporting information (SI)). The furanose group flexibility was found uncorrelated to the disorder of the central O-Bn group, a disorder that could be considered as essentially dynamic. While the major α form can be seen to be stabilized by an intramolecular hydrogen bond with O2, graphset symbol S(5), (see Table 2) and benefit from developing a weak intermolecular H-bond (see below Figure 1, and crystal packing analysis in Figure 4), the minor β-form can find in O3 a 'lifeline' in a hydrophobic environment to make an intramolecular h-bond (graphset, symbol S(6)). The conformational analysis of the furanose ring ( Figure 2) based on its internal dihedral angles and its deviation from planarity showed that the pseudorotational phase angle P ≅ 11° and the maximum puckering amplitude νmax ≅ 35 (see Table 1) [25,26]. Thus, this ring adopts a conformation close to 3 T2, where C2 and C3 deviate by 0.116 and -0.443 Å, (by 0.467 and -0.058), respectively, from the plane through atoms C1(B)/C4/O4(B).  (10) 3.3601 (7) 142.0(8) The conformational analysis of the furanose ring ( Figure 2) based on its internal dihedral angles and its deviation from planarity showed that the pseudorotational phase angle P ∼ = 11 • and the maximum puckering amplitude ν max ∼ = 35 (see Table 1) [25,26]. Thus, this ring adopts a conformation close to 3 T 2 , where C2 and C3 deviate by 0.116 and −0.443 Å, (by 0.467 and −0.058), respectively, from the plane through atoms C1(B)/C4/O4(B).

General Remarks
Unless otherwise stated, all reagents and starting materials were purchased from commercial sources and used as received. Methanol (anhydrous, 99.8%) was purchased

General Remarks
Unless otherwise stated, all reagents and starting materials were purchased from commercial sources and used as received. Methanol (anhydrous, 99.8%) was purchased from Sigma-Aldrich Chimie S.a.r.l-38297 Saint-Quentin-Fallavier CEDEX, France. N,N-Dimethylformamide (ACS reagent, ≥99.8%) was purified by passage through a column containing activated alumina under nitrogen pressure (Dry Solvent Station GT S100, Glass Technology, Geneva, CH). Amberlite®IRA-400 was prepared in its OH − form by passing 1 M KOH until the effluent was free of chloride ions, it was then washed with distilled H 2 O until neutral and then with MeOH. NMR spectra were recorded at 298 K with a Bruker Avance III HD nanobay 400 MHz spectrometer equipped with a BBO probe-Brucker Chemical shift values are given in ppm, and are related to the corresponding nondeuterated solvent. High-resolution mass spectra were recorded with a Bruker maXis ESI qTOF ultrahigh-resolution mass spectrometer coupled to a Dionex Ultimate 3000 RSLC system. MS data were acquired in positive mode and were processed using Data Analysis 4.4 software (Bruker). The infrared spectrum of compound (α,β)-1 was recorded neat with a Thermo Scientific Nicolet IS10 FTIR spectrometer using diamond ATR golden gate sampling (Thermo Fisher Scientific, 28199 Bremen, Germany), and is reported in wave numbers (cm −1 ). Analytical thin-layer chromatography (TLC) was performed with Merck Silica Gel 60 F254 precoated plates-VWR Avantor, F-93114 Rosny-sous-Bois CEDEX, France. Visualization of the developed chromatogram was performed under ultraviolet light (254 nm) and on staining by immersion in aqueous, acidic ceric ammonium molybdate followed by charring at ca. 150 • C. Column chromatography was performed in air on Silica Gel 60 (230-400 mesh) with petroleum ether (bp 40-65 • C) and ethyl acetate (EtOAc) as eluents. Organic solutions were concentrated under reduced pressure with a Büchi rotary evaporator-BÜCHI SARL, 91140 Villebon-sur-Yvette, France. Conformation, crystal packing and geometrical parameters for the LUHROX D-arabinofuranose hemiacetal analogue were obtained from the Cambridge Structural Database (CSD, Version 5.42; Nov 2021, Cambridge, UK [27]). Hirshfeld surface of (α,β)-1 was drawn using CrystalExplorer 17.5-f4e298a, University of Western, Australia [29]. X-ray structure determination at rt and 100K was performed on a Rigaku rotating anode-ELEXIENCE SA, 91371 Verrières-le-Buisson CEDEX, France. X-ray data were then processed using CrystalClear 2.0, Tokyo, Japan [30] or CrysAl-isPro 1.171.41.121a, Yarnton, UK [31]. In both cases, the structure was solved by intrinsic phasing methods (SHELXT Version 2018/2, University of Göttingen, Germany [32]). Refinement was done subsequently by full-matrix least-squares against F 2 (SHELXL Version 2018/3, University of Göttingen, Germany) [33]). Ultimately, it was pursued with the Hirshfeld atom refinement (HAR; [34]) with the NoSpherA2 [35] implementation in OLEX2-1.5, Durham University, UK [36]. Bijvoet analyses were performed with Platon (version170914, Utrecht University, The Netherlands).

Procedures and Characterization Data
3.2.1. Synthesis of 2,3,5-Tri-O-benzyl-α,β-D-xylofuranose (α,β)-1 A single-necked round-bottomed flask under argon atmosphere was charged with AcCl (2.5 mL, ca. 35.0 mmol) and dry MeOH (300 mL), and the solution was stirred at 20 • C for 30 min. D-xylose (5.0 g, 33.3 mmol) was then added and the reaction mixture was stirred for 3.5 h at 30 • C. Resin Amberlite IRA-400 (OH − form) was added until pH 8, and the solution was filtered through a cotton plug and concentrated under vacuum. The crude product (Rf 0.6; SiO 2 , CH 2 Cl 2 /MeOH 8:2, v/v) was obtained as a light-yellow oil, and used in the next step without further purification.

Supplementary Materials:
The following are available online at www.mdpi.com/xxx/s1, Table S1: Experimental details for XRD performed at room temperature using copper radiation, Figure S1: Bijvoet pair analyses to assess the absolute configuration of compound 1 (α/β ca. 89:11), Figure S2

Supplementary Materials:
The following are available online, Table S1: Experimental details for XRD performed at room temperature using copper radiation, Figure S1: Bijvoet pair analyses to assess the absolute configuration of compound 1 (α/β ca. 89:11), Figure S2