Download PDF
Synthesis 2024; 56(06): 966-974
DOI: 10.1055/a-2149-4586
paper
Emerging Trends in Glycoscience

C-5 Epimerisation of d-Mannopyranosyl Fluorides: The Influence of Anomeric Configuration on Radical Reactivity

Nicholas W. See 
a   School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, Queensland 4072, Australia
,
Gregory K. Pierens 
b   Centre for Advanced Imaging, The University of Queensland, Brisbane, Queensland 4072, Australia
,
Elizabeth H. Krenske 
a   School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, Queensland 4072, Australia
,
Vito Ferro  ∗
a   School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, Queensland 4072, Australia
› Author Affiliations

We thank the Australian Research Council for financial support (DP220102493). N.W.S. is grateful to The University of Queensland (UQ) for a PhD scholarship.
› Further Information
 
 PDF Download Permissions and Reprints All articles of this category


Abstract

The fluorine-directing effect has so far been exploited to provide short and efficient synthetic routes to rare l-ido sugars. However, the importance of anomeric configuration to its success has remained experimentally unverified. We now report on the synthesis of α- and β-configured per-O-benzoylated mannopyranosyl fluorides and initially show that their reactivity towards photo-bromination is strongly dependent on the anomeric configuration. The stereochemical basis of the fluorine-directing effect is then validated by revealing the striking difference in stereoselectivity observed for the free-radical reductions of the isolated 5-C-bromo sugars. This work importantly provides a synthetic route to a donor-functionalised derivative of l-gulose and reveals new insights into the behaviour of glycosyl radicals.


#

Key words

glycosyl radicals - l-hexoses - photochemistry - halides - rearrangements

Glycosyl radicals represent an attractive and versatile set of tools for editing the structures of carbohydrates.[1] [2] [3] [4] [5] In line with the growth of modern photocatalysis,[6] there is renewed interest in harnessing these reactive intermediates for the regio- and stereoselective epimerisation of stereogenic centres within pyranose scaffolds.[7–9] Indeed, Wendlandt has recently demonstrated that glycosyl radical chemistry can be leveraged to access libraries of rare sugars from simple building blocks.[10]

Previously, we reported on the discovery of a fluorine-directing effect and its ability to provide synthetic access to key members of the rare but biologically pertinent family of l-hexoses.[11] [12] [13] [14] The fluorine-directing effect enables the stereoselective free-radical reduction of 5-C-bromo-d-hexoses (Scheme [1]).[15] [16] The requisite substrates for this protocol are functionalised at the anomeric position with a β-fluoride. Through density functional theory (DFT) calculations, we showed that the fluorine imparts a stereoelectronic bias for l-hexose formation by engaging the endo-anomeric effect (LPO → σ*C–F).

Zoom Image
Scheme 1 Previous fluorine-directed syntheses of l-iduronic acid (a) and l-idose (b), and the current application of this chemistry towards l-gulose

We also demonstrated that the C-6 substituent plays an influential role in this strategy.[17] While a C-6 methoxycarbonyl group is best suited for the synthesis of l-iduronic acid (Scheme [1a]), it was concluded that a benzoyloxymethylene group at the same position provides optimal stereoselectivity for the synthesis of l-idose. This understanding led us to develop a short and efficient synthetic route to glycosyl donor 1 (Scheme [1b]).[17]

While DFT calculations have provided useful insights into the stereo-directing nature of the β-F in this context, the importance of anomeric configuration to l-hexose selectivity has remained experimentally unverified. We now report on the syntheses of α- and β-configured d-mannopyranosyl fluorides. The importance of β-configuration to l-hexose selectivity is then revealed through the photo-bromination and subsequent free-radical reduction of both anomers. In this way, further understanding about the stereochemical basis of the fluorine-directing effect is achieved. Concurrently, we provide a short synthetic route to a derivative of l-gulose 2 – a reasonably expensive sugar[18] of high commercial and biological significance, e.g., as a component of the antitumour antibiotics, the bleomycins.[19] [20] [21]

With the view of obtaining preparative amounts of both α- and β-mannosyl fluorides 6α and 6β for this study, we commenced their joint synthesis with the global O-benzoylation of d-mannose (3) (Scheme [2]). Routine treatment of azeotropically-dried 3 with benzoyl chloride and catalytic 4-(dimethylamino)pyridine (DMAP) in anhydrous pyridine smoothly delivered mannose pentabenzoate 4 in high yield (83%). Upon crystallisation from ethanol, an anomerically pure sample of 4 was obtained as a white solid whose melting point and spectroscopic data agreed with those of Williams.[22]

Zoom Image
Scheme 2 Chemical synthesis of α- and β-mannopyranosyl fluorides 6α and 6β

We next explored two traditional approaches to anomeric fluorination. Through the action of HBr,[23] 4 was rapidly converted into α-mannosyl bromide 5 in quantitative yield which was then vigorously stirred in the dark with AgF[15] and freshly activated molecular sieves to effect fluoride displacement at the anomeric position. A single diastereoisomer was obtained accordingly in moderate yield (52%) after purification on silica gel, whose NMR spectroscopic data matched those reported by Ziegler et al. [24] for the β-mannosyl fluoride. According to Ziegler et al., the reaction of 5 with KHF2 as the source of fluoride stereoselectively yields β-mannosyl fluoride 6β. However, both our and Ziegler’s 13C NMR data matched those of Miethchen and Kolp[25] for the α-anomer 6α. Miethchen reported that reaction with Et3N·3HF as the fluoride source proceeds predominantly via neighbouring group participation (NGP) through which a transient acyloxonium ion shields the β-face of the ring to generate α-mannosyl fluoride 6α. While Miethchen reported the isolation of a mixture of 6α and 6β, only 13C NMR data for the former were provided.

While we preferred the mechanistic rationale of Miethchen and Kolp, we noted this discrepancy and set out to provide clarification on our own accord. In the absence of information derived from the chemical shift of C-1 and the vicinal 3 J 1,2 (H–H) and geminal 1 J C1,F coupling constants, whose magnitudes are collectively similar between α- and β-mannosides, we resorted to performing a series of nuclear Overhauser effect (NOE) experiments. Disappointingly, we failed to observe definitive correlations between H-1 and other ring protons due to signal crowding. The considerable splitting of the H-1 signal imposed by the fluorine nucleus (2 J 1,F = 48.8 Hz) further complicated matters.

Having failed to obtain direct experimental evidence, we elected to take a theoretical approach to investigating the anomeric stereochemistry. DFT calculations were performed on mannosyl fluorides 7α and 7β which are per-O-acetylated analogues of 6α and 6β (Figure [1]). Following geometry optimisation of low-energy conformers (details in Supporting Information) with B3LYP/6-31G(d,p) in vacuum, 13C–19F coupling constants for 7α and 7β were calculated at the BHandH/6-311++G(2d,p)-PCM(CHCl3) level of theory.

Zoom Image
Figure 1 Mannosyl fluorides 7α and 7β are truncated derivatives of 6α and 6β, respectively.

In Table [1], the results of these calculations are displayed adjacently to the experimentally measured J C–F values. The most distinguishing data point in the set was the 2 J C2,F value whose magnitude displayed the strongest dependence on anomeric configuration. Indeed, a difference of 25 Hz between 7α (42 Hz; cf. 6α: 39 Hz) and 7β (17 Hz) was calculated. When the experimental and theoretical 2 J C2,F values were compared, it became clear that the HBr/AgF sequence from 4 had stereoselectively generated α-anomer 6α. These results corroborated the findings of Bock and Pedersen[26] and suggested that Ziegler et al. had also prepared 6α.

Table 1 13C–19F Coupling Constants Measured for 6α and DFT-Calculated Values for Model Fluorides 7α and 7β with Literature[26] Comparison

Calcd (Lit.[26])

Exptl (6α)

7α

7β

C

J C–F (Hz)

J C–F (Hz)

J C–F (Hz)

1

222.4

250.2 (223.1)

239.5 (223.6)

2

39.2

42.2 (39.7)

17.0 (25.0)

3

–

1.6 (1.8)

9.7 (5.9)

4

–

0.4 (n.m.)a

0.4 (n.m.)a

5

2.7

2.6 (3.1)

5.8 (4.4.)

a n.m. = not measured.

Following this, it made sense on thermodynamic grounds that upon treatment of pentabenzoate 4 with Olah’s reagent (HF·pyridine),[27] a spectroscopically identical sample of α-mannosyl fluoride 6α was generated exclusively in high yield (93%). With 6α in hand, we directed our attention towards accessing 6β. At the outset, we acknowledged the difficulties inherent to the construction of β-mannosides. In the absence of β-stereo-directing structural features and conditions, these compounds are kinetically and thermodynamically disfavoured to their α-configured congeners. The use of a participating O-acyl group at C-2 also erodes β-stereoselectivity;[25] however, the importance of benzoyloxy groups to l-hexose selectivity[17] and their inherent compatibility with all chemistries in the scheme is once more emphasised.

Extending upon chemistry explored previously in our laboratory[28] and others,[29] we reasoned that anomeric deoxyfluorination would provide an avenue to 6β. The requisite hemiacetal 8 was accessed via a chemoselective, anomeric deacylation of pentabenzoate 4 with 3-(dimethylamino)-1-propylamine (DMAPA) in tetrahydrofuran (THF).[30] After aqueous workup and flash chromatography, the desired hemiacetal 8 was obtained in excellent yield (86%) as the α-anomer. All characterisation data agreed closely with those of Brimble et al. [31] We then derived a set of deoxyfluorination conditions for the synthesis of 6β after considering the reaction mechanism (Scheme [3]).

Zoom Image
Scheme 3 Postulated reaction pathways to anomeric fluorides 6α and 6β from 8. Reagents and conditions: (a) XtalFluor-M®, Et3N, Et3N·3HF, CH2Cl2, –78 °C to r.t. Counterions have been omitted for clarity.

The pathways leading to mannosyl fluorides 6α and 6β are proposed to jointly commence from oxonium ion 9. Two different reactions of 9 could proceed via transition states (TSs) TS6β and TS6α . Respectively, these TSs feature (i) direct substitution with fluoride to yield 6β and (ii) NGP to generate the α-stereo-directing acyloxonium ion 10. Both reactions are assumed to be irreversible[29a] and SN2-like in nature. Understanding that TS6α is independent of fluoride concentration, we reasoned that the formation of 6β, which occurs via fluoride-dependent TS6β , could be optimised by using an excess of fluoride and by minimising the solvent volume, with careful control of temperature.

With this in mind, 8 was taken up in dry CH2Cl2 and treated with an excess of XtalFluor-M® and Et3N·2HF, the latter of which was generated in situ from the action of Et3N on commercially available Et3N·3HF. The reaction was initiated at low temperature to optimise β-stereoselectivity in the usual manner, and to limit competing anomerisation to more stable 6α through TS6β → 6α (Scheme [3]). Collectively, these conditions yielded a readily separable mixture of mannosyl fluorides 6α and 6β in a ratio of 6:1 and in high yield (91%, combined). This represents an improvement on the Miethchen protocol (refluxing of 5 with Et3N·3HF in CCl4) in terms of yield (Lit. 62%).[25] Conveniently, we were able to spectroscopically diagnose the anomeric configuration of 6β with NOESY experiments (spectra in Supporting Information). We also highlight that the 2 J C2,F coupling constant measured for 6β (20 Hz) was consistent both with the value calculated for 7β by DFT methods (17 Hz; Table [1]) and with that reported by Bock and Pedersen.[26]

Having obtained anomerically pure samples of 6α and 6β, we photo-brominated both compounds according to the conditions of Stick and co-workers (Scheme [4]).[32] Photolysis of 6β with Br2 in refluxing CCl4 regio- and stereoselectively afforded 5-C-bromide 11β in moderate yield (40%). Intriguingly, the conversion of 6β into 11β ceased after 6 hours and could not be encouraged with neither additional reaction time nor Br2. A reasonable quantity of 6β (30%) was consequently recovered from the product mixture (total yield brsm: 56%). Compared with 6β, α-mannosyl fluoride 6α was significantly less reactive under the same conditions. Close monitoring by TLC revealed the sluggish conversion of 6α into a single spot of higher retardation factor (Rf ) over an extensive period (48 h); simultaneous and progressive degradation was also observed. When the behaviours of 6α and 6β towards photo-bromination are compared, it becomes clear that when α-configured, an anomeric fluoride tempers the reactivity of the C-5–H bond towards homolysis. We attribute this to a competition between the hyperconjugative stabilisation of the radical forming at C-5 and the anomeric fluoride.

Zoom Image
Scheme 4 Mannosyl fluorides 6β and 6α displayed substantially different reactivities towards photo-bromination.

During C-5–H bond homolysis, a lone pair on the endocyclic oxygen (LPO) in 6α acts as a σ-donor and donates simultaneously into the anti-oriented σ*C-5–H and σ*C-1–F orbitals. The latter interaction would be expected to dominate.[33] In 6β, the σ*C-1–F orbital is likely gauche to the LPO in the relevant TS and is therefore unable to suppress C-5 radical formation to the same extent. While we propose that these stereoelectronic effects are principally responsible, a 1,3-diaxial interaction between the anomeric fluoride and the homolytic substitution taking place at C-5 may also retard C–H abstraction in 6α.

Following chromatographic purification of the product mixture yielded by the photo-bromination of 6α, NMR analysis unveiled a mixture of two compounds. The major component of this mixture was identified as the desired 5-C-bromide 11α which had formed in parallel with 6-C-bromide 12 in a ratio of 66:34. In both products, and indeed 11β, the H-4 resonance was resolved to a doublet and the diastereotopic H-6 protons gave rise to an AB spin system. Accordingly, this provided direct evidence for substitution at C-5.

The generation of 12 was unexpected and we reasoned that it had occurred via the mechanistic pathway shown in Scheme [5]. The key substituent in this mechanism is the C-6 benzoyloxy group. In a process similar to NGP,[33] this substituent is proposed to capture the relevant C-5 radical at a rate deemed competitive to the intermolecular reaction with a bromine radical.

Zoom Image
Scheme 5 Mechanistic proposal for the generation of 12

The resulting product of the eventuating cyclisation is bicyclic radical 13. The stability of this derivative is underpinned by (i) delocalisation of the radical onto the π-system of the adjacent phenyl ring and (ii) hyperconjugation (LPO → σ*C-5–O). The latter of these features also encourages an axial preference for the new C-5–O bond and explains the C-5 stereochemistry of 12. Upon subsequent colligation with bromine, 13 is proposed to spontaneously collapse to 12. To confirm that 12 had formed via a radical-based mechanism, we repeated the reaction in the presence of excess butylated hydroxytoluene (BHT) and monitored its progress by 19F NMR spectroscopy. Indeed, inhibition was observed which provided evidence in favour of a radical-based reaction pathway.

Finally, we examined the dependence of l-hexose stereoselectivity on anomeric configuration by performing free-radical reductions of 11β and 11α (Scheme [6]). Treatment of 11β with n-Bu3SnH and Et3B in toluene at room temperature generated a readily separable mixture of α-l-gulosyl fluoride 2 and β-d-mannosyl fluoride 6β in a ratio of 53:47 and in reasonable yield (65%, combined). Fortuitously, 2 is equipped with glycosyl donor capability.[34] [35] [36] In contrast to 11β, the free-radical reduction of 11α (as an inseparable 66:34 mixture with 12) was completely stereoselective for α-d-mannosyl fluoride 6α; no trace of β-l-gulosyl fluoride 14 was detected in the product mixture. This result provided critical validation for the importance of β-configuration to the success of the fluorine-directing effect. Compound 15 was also isolated from this reaction and is self-evidently the debromination product of 12. An NOE correlation between the C-6 methyl group and H-4 was observed for 15 (see Supporting Information). This confirmed the configuration at C-5 proposed for 15. Indeed, by chemical correlation, it also validated the C-5 stereochemistry of 12.

Zoom Image
Scheme 6 n-Bu3SnH reductions of 11β and the mixture of bromides 11α and 12

In conclusion, we have shown that the reactivity of mannopyranosyl fluorides towards photo-bromination at C-5 is dependent on the anomeric configuration. Specifically, photolysis of α-mannosyl fluoride 6α proceeded at an appreciably slower rate compared to the same reaction of 6β. Conceivably, the strongly electron-withdrawing nature of fluoride combined with its axial configuration had limited the stabilisation of the forming C-5 radical in 6α. As a consequence, competitive participation of the C-6 benzoyloxy group provided 6-C-bromide 12 as a rearrangement byproduct. In this study, we have provided proof-of-concept for rapid access to l-gulose via exploitation of the fluorine-directing effect. Alternative substrate designs could be explored which would conceivably provide a higher yield of 6β.[28] [29] Through the free-radical reductions of 11α and 11β, we also demonstrated that an α-fluoride is inferior to β-fluoride in terms of its l-hexose stereo-directing ability. This result extends the range of hexoses in which β-configuration is key to the success of the fluorine-directing effect.

Reagents were purchased from Merck and Co. and were used without further purification. All reactions were performed in oven-dried glassware under inert atmosphere (Ar or N2) and were monitored by TLC using silica gel F254 aluminum-backed sheets. Compounds were visualised using p-anisaldehyde/H2SO4 dip. Photo-brominations were performed with an Arlec 500 W tungsten lamp fitted with a linear halogen tube. The lamp was strictly held at a 20 cm distance from the reaction vessel. NMR spectra were measured on Bruker Avance 500 and 700 MHz spectrometers and referenced to the residual solvent peaks (δH = 7.26 ppm, δC = 77.0 ppm). 19F NMR spectra were externally referenced to monofluorobenzene (δF = –113.5 ppm). 2D NMR experiments including 1H–1H COSY, 1H–19F COSY, HSQC and NOESY aided with structure elucidations. Optical rotations were measured on a Jasco P-2000 polarimeter. Melting points were measured on a Digimelt MPA161 apparatus. Low- and high-resolution electrospray ionisation mass spectrometry (LRMS/HRMS) data were obtained with Bruker HCT and Bruker micrOTOFQ spectrometers, respectively, in positive ionisation mode. Flash chromatography was performed on silica gel (230–400 mesh, Grace) under pressure with the specified eluants. Reaction stereoselectivities were measured by 1H NMR analysis of the post-workup product mixture, prior to purification by flash chromatography.


#

1,2,3,4,6-Penta-O-benzoyl-α-d-mannopyranose (4)

An azeotropically-dried sample of d-mannose (3; 5.49 g, 30.5 mmol) was combined with DMAP (10 mg) in anhydrous pyridine (130 mL). After cooling in an ice–water bath (external Dewar temperature: 4 °C), benzoyl chloride (21 mL, 183 mmol) was added dropwise to the stirred mixture over a 10 min period. The reaction was held at the same temperature for an additional 10 min before it was warmed to r.t. and stirred o/n, during which a pink colour and a white precipitate evolved simultaneously. The reaction was quenched in an ice–water bath with anhydrous MeOH (30 mL) after which the bulk of the solvent was removed under reduced pressure. The crude product residue was then taken up in CHCl3 (200 mL) and washed with ice-cold 1.0 M HCl (3 × 20 mL), sat. NaHCO3 (20 mL) and brine (20 mL). The product was then dried (MgSO4), filtered and azeotropically dried several times with anhydrous PhMe to deliver pentabenzoate 4 as a white solid. The product was crystallised from EtOH to yield a single anomer (17.8 g, 83%).

Rf = 0.57 (PhMe/EtOAc, 10:1; stained blue with p-anisaldehyde/H2SO4 dip).

Mp 152–154 °C (Lit.[37] 151–153 °C); [α]D 23 –19.4 (c 0.72, CHCl3) (Lit.[37] –19.3).

1H NMR (500 MHz, CDCl3): δ = 8.22–8.19 (m, 2 H, Ph), 8.11–8.07 (m, 4 H, Ph), 7.98–7.95 (m, 2 H, Ph), 7.88–7.84 (m, 2 H, Ph), 7.71–7.66 (m, 1 H, Ph), 7.65–7.60 (m, 1 H, Ph), 7.60–7.55 (m, 3 H, Ph), 7.54–7.49 (m, 1 H, Ph), 7.48–7.35 (m, 7 H, Ph), 7.31–7.27 (m, 2 H, Ph), 6.63 (d, J 1,2 = 2.1 Hz, 1 H, H-1), 6.28 (dd, J 3,4 = 10.3 Hz, J 4,5 = 10.2 Hz, 1 H, H-4), 6.07 (dd, J 2,3 = 3.3 Hz, 1 H, H-3), 5.91 (dd, 1 H, H-2), 4.70 (dd, part A of ABX, J 5,6a = 2.5 Hz, J 6a,6b = 12.4 Hz, 1 H, H-6a), 4.57 (ddd, J 5,6b = 3.8 Hz, 1 H, H-5), 4.50 (dd, part B of ABX, 1 H, H-6b).

13C NMR (125 MHz, CDCl3): δ = 166.0, 165.6, 165.2, 165.1, 163.8 (5 × C=O), 134.0, 133.6, 133.5, 133.3, 133.0, 130.1, 129.9, 129.8, 129.7, 128.9, 128.8, 128.7, 128.6(9), 128.6(5), 128.5, 128.4(1), 128.3 (Ph), 91.3 (C-1), 71.1 (C-5), 69.9 (C-3), 69.3 (C-2), 66.1 (C-4), 62.3 (C-6).

The 1H and 13C NMR spectra were in accordance with the literature.[38]

LRMS: m/z = 723.2 [M + Na]+.


#

2,3,4,6-Tetra-O-benzoyl-α-d-mannopyranosyl Bromide (5)

Pentabenzoate 4 (1.11 g, 1.58 mmol) was stirred under dry Ar(g) with pulverized, acid-washed 4 Å molecular sieves (100 mg) in dry CH2Cl2 (6 mL) for 45 min. The mixture was submerged in an ice–water bath (external Dewar temperature: 4 °C) after which excess HBr (3 mL, 33% w/w in AcOH) was added dropwise over the course of 5 min. The reaction was then allowed to attain r.t. and left to stir o/n. The product was poured into a vigorously stirred mixture of CHCl3 (50 mL) and ice–water (150 mL). After 10 min, the phases were separated and the aqueous phase was extracted with CHCl3 (3 × 20 mL). The combined organic phases were washed with sat. NaHCO3 (2 × 50 mL) and brine (2 × 50 mL). The product was dried (MgSO4), filtered and concentrated under reduced pressure to furnish the glycosyl bromide 5 as a white foam (1.04 g, quant.).

Rf = 0.38 (n-hexane/PhMe/EtOAc, 10:10:1; stained brown with p-anisaldehyde/H2SO4 dip).

[α]D 23 +10.1 (c 0.50, CHCl3) (Lit.[39] +10.8).

1H NMR (500 MHz, CDCl3): δ = 8.12–8.09 (m, 2 H, Ph), 8.05–8.01 (m, 2 H, Ph), 8.00–7.96 (m, 2 H, Ph), 7.86–7.82 (m, 2 H, Ph), 7.63–7.57 (m, 2 H, Ph), 7.56–7.51 (m, 1 H, Ph), 7.47–7.37 (m, 7 H, Ph), 7.30–7.26 (m, 2 H, Ph), 6.59 (d, J 1,2 = 1.8 Hz, 1 H, H-1), 6.29 (dd, J 2,3 = 3.1 Hz, J 3,4 = 10.2 Hz, 1 H, H-3), 6.24 (dd, J 4,5 = 9.3 Hz, 1 H, H-4), 5.91 (dd, 1 H, H-2), 4.74 (dd, part A of ABX, J 5,6a = 2.7 Hz, J 6a,6b = 12.7 Hz, 1 H, H-6a), 4.66 (ddd, J 5,6b = 3.7 Hz, 1 H, H-5), 4.51 (dd, part B of ABX, 1 H, H-6b).

13C NMR (125 MHz, CDCl3): δ = 165.9, 165.4, 165.3, 164.9 (4 × C=O), 133.7, 133.6, 133.4, 133.2, 129.8(8), 129.8(7), 129.7(9), 129.7(6), 129.6, 128.7(6), 128.7(1), 128.6(8), 128.6(4), 128.5, 128.4 (Ph), 83.3 (C-1), 73.1 (C-5), 72.9 (C-2), 69.1 (C-3), 65.9 (C-4), 61.7 (C-6).

The 1H and 13C NMR spectra were in accordance with the literature.[38]

LRMS: m/z = 683.0 [M + Na]+.


#

2,3,4,6-Tetra-O-benzoyl-α-d-mannopyranosyl Fluoride (6α)


#

Fluorination with Olah’s Reagent

Pentabenzoate 4 (230 mg, 0.33 mmol) was combined with excess Olah’s reagent (HF·pyridine, 1 mL) at –10 °C in a plastic Falcon tube. The suspension was stirred under Ar(g) for 10 min before it was slowly allowed to attain r.t. after which it was stirred o/n. The resulting tan solution was then diluted with EtOAc (20 mL) before it was poured into ice-cold sat. NaHCO3 (20 mL). The mixture was shaken vigorously for 5 min before the phases were allowed to separate and the pH of the organic phase was neutral. The product was extracted from the aqueous phase with EtOAc (10 mL) before the combined organic phases were washed successively with water (20 mL) and brine (2 × 20 mL). The product was then dried (MgSO4), filtered and concentrated under reduced pressure to deliver α-fluoride 6α as a white foam (185 mg, 93%).

Rf = 0.50 (PhMe/EtOAc, 30:1; stained light blue with p-anisaldehyde/H2SO4 dip).

[α]D 23 –79.7 (c 0.58, CHCl3) (Lit.[22] –86.2).

1H NMR (500 MHz, CDCl3): δ = 8.14–8.10 (m, 2 H, Ph), 8.05–8.01 (m, 2 H, Ph), 7.98–7.93 (m, 2 H, Ph), 7.86–7.81 (m, 2 H, Ph), 7.64–7.25 (m, 12 H, Ph), 6.21 (dd, J 3,4 = 10.2 Hz, J 4,5 = 10.0 Hz, 1 H, H-4), 5.92 (ddd, J 2,3 = 3.3 Hz, J 3,F = 1.6 Hz, 1 H, H-3), 5.87 (dd, J 1,2 = 1.9 Hz, J 1,F = 48.8 Hz, 1 H, H-1), 5.86 (dd, 1 H, H-2), 4.78 (dd, part A of ABX, J 5,6a = 2.5 Hz, J 6a,6b = 12.4 Hz, 1 H, H-6a), 4.61 (ddd, J 5,6b = 3.8 Hz, 1 H, H-5), 4.49 (dd, part B of ABX, 1 H, H-6b).

13C NMR (125 MHz, CDCl3): δ = 166.0, 165.3, 165.2, 165.0 (4 × C=O), 133.7, 133.6, 133.3, 133.1, 129.9, 129.8(2), 129.8(0), 129.7, 129.0, 128.7, 128.6, 128.5, 128.4, 128.3, 128.2 (Ph), 104.9 (d, J 1,F = 222.4 Hz, C-1), 71.2 (d, J 5,F = 2.7 Hz, C-5), 69.1 (C-3), 68.5 (d, J 2,F = 39.2 Hz, C-2), 65.7 (C-4), 62.1 (C-6).

19F NMR (470 MHz, CDCl3): δ = –138.2 (ap. d, J 1,F = 48.8 Hz).

The 1H and 13C NMR spectra were in accordance with the literature.[22] [25]

LRMS: m/z = 616.0 [M + NH4]+.


#

Fluorination with AgF

Glycosyl bromide 5 (290 mg, 0.44 mmol) was stirred with freshly activated, pulverized 4 Å molecular sieves (100 mg) in dry acetonitrile (10 mL) for 1 h. To this suspension was added dry AgF (195 mg, 1.5 mmol) after which vigorous stirring was undertaken in the dark for 24 h. The product mixture was concentrated under a stream of N2(g), suspended in EtOAc and filtered through silica gel. Following concentration under reduced pressure, purification by flash chromatography (PhMe/EtOAc, 30:1) gave a sample of α-fluoride 6α which was identical in all aspects to that produced by the above method (137 mg, 52%).


#

2,3,4,6-Tetra-O-benzoyl-α-d-mannopyranoside (8)

Pentabenzoate 4 (4.1 g, 5.8 mmol) and 3-(dimethylamino)-1-propylamine (DMAPA; 2.2 mL, 17.5 mmol) were combined in dry THF (10 mL). The amber solution was stirred at r.t. under N2(g) for 19 h, during which a red colour progressively developed. The product mixture was concentrated under reduced pressure before the crude residue was dissolved in EtOAc (100 mL) and washed with 0.1 M HCl (4 × 20 mL), sat. NaHCO3 (20 mL) and brine (2 × 20 mL). The product was dried (MgSO4), filtered and concentrated under reduced pressure. Purification by flash chromatography (PhMe/EtOAc, 5:1) furnished hemiacetal 8 as a white foam (2.99 g, 86%).

[α]D 23 –77.0 (c 0.6, CHCl3) (Lit.[31] –79).

1H NMR (500 MHz, CDCl3): δ = 8.14–8.10 (m, 2 H, Ph), 8.05–8.01 (m, 2 H, Ph), 7.99–7.95 (m, 2 H, Ph), 7.87–7.84 (m, 2 H, Ph), 7.60–7.55 (m, 2 H, Ph), 7.53–7.48 (m, 1 H, Ph), 7.45–7.39 (m, 3 H, Ph), 7.39–7.34 (m, 4 H, Ph), 7.29–7.26 (m, 2 H, Ph), 6.18 (dd, J 3,4 = 10.2 Hz, J 4,5 = 9.9 Hz, 1 H, H-4), 6.01 (dd, J 2,3 = 3.3 Hz, 1 H, H-3), 5.75 (dd, J 1,2 = 1.9 Hz, 1 H, H-2), 5.54 (dd, J 1,OH = 4.2 Hz, 1 H, H-1), 4.77 (dd, part A of ABX, J 5,6a = 2.6 Hz, J 6a,6b = 12.4 Hz, 1 H, H-6a), 4.68 (ddd, J 5,6b = 3.8 Hz, 1 H, H-5), 4.45 (dd, part B of ABX, 1 H, H-6b), 3.68 (d, 1 H, OH).

13C NMR (125 MHz, CDCl3): δ = 166.4, 165.6, 165.5 (3 × C=O), 133.4(4), 133.4(1), 133.2, 133.1, 129.8(4), 129.8(0), 129.3, 129.1, 128.9, 128.6, 128.4, 128.3 (Ph), 92.4 (C-1), 70.9 (C-2), 69.8 (C-3), 68.9 (C-5), 66.8 (C-4), 62.7 (C-6).

The 1H and 13C NMR data were in accordance with the literature.[31]

LRMS: m/z = 1215.7 [2 M + Na]+.


#

2,3,4,6-Tetra-O-benzoyl-β-d-mannopyranosyl Fluoride (6β)

To a stirred solution of Et3N (310 μL, 2.2 mmol) and Et3N·3HF (720 μL, 4.4 mmol) in dry CH2Cl2 (4.5 mL) was added XtalFluor-M® (810 mg, 3.3 mmol). The resulting red solution was submerged in a dry ice/acetone bath (external Dewar temperature: –78 °C) before a solution of the hemiacetal 8 (1.3 g, 2.2 mmol) in dry CH2Cl2 (4.5 mL) was added dropwise with stirring under Ar(g). The reaction was held at –78 °C for 30 min before it was slowly allowed to attain r.t. Stirring was continued for a further 90 min before the reaction was quenched with sat. NaHCO3 (10 mL). The product was diluted with EtOAc (150 mL) and washed with sat. NaHCO3 (30 mL), water (30 mL) and brine (2 × 30 mL) before it was dried (MgSO4), filtered and concentrated under reduced pressure. Purification by flash chromatography (PhMe/EtOAc, 50:1) gave β-fluoride 6β as a white foam (178 mg, 13%) and α-fluoride 6α (1.03 g, 78%).

Data for 6β: Rf = 0.33 (PhMe/EtOAc, 30:1; stained light blue with p-anisaldehyde/H2SO4 dip).

[α]D 23 –94.3 (c 0.7, CHCl3).

1H NMR (700 MHz, CDCl3): δ = 8.11–8.08 (m, 2 H, Ph), 8.02–7.98 (m, 4 H, Ph), 7.95–7.92 (m, 2 H, Ph), 7.59–7.54 (m, 3 H, Ph), 7.53–7.50 (m, 1 H, Ph), 7.46–7.40 (m, 4 H, Ph), 7.40–7.35 (m, 4 H, Ph), 5.96 (dd, J 3,4 = 7.5 Hz, J 4,5 = 7.2 Hz, 1 H, H-4), 5.94–5.91 (m, 1 H, H-2), 5.88 (dd, J 1,2 = 1.5 Hz, J 1,F = 50.9 Hz, 1 H, H-1), 5.77 (dd, J 2,3 = 3.5 Hz, 1 H, H-3), 4.85 (dd, part A of ABX, J 5,6a = 2.7 Hz, J 6a,6b = 11.6 Hz, 1 H, H-6a), 4.70 (dd, part B of ABX, J 5,6b = 6.3 Hz, 1 H, H-6b), 4.41–4.38 (m, 1 H, H-5).

13C NMR (125 MHz, CDCl3): δ = 166.0, 165.5, 165.3, 165.0 (4 × C=O), 133.7, 133.6(1), 133.6(0), 133.2, 130.0, 129.9, 129.9, 129.6, 128.9, 128.7, 128.6, 128.5, 128.4 (Ph), 104.3 (d, J 1,F = 223.8 Hz, C-1), 72.7 (d, J 5,F = 2.9 Hz, C-5), 68.9 (d, J 3,F = 4.9 Hz, C-3), 67.4 (d, J 2,F = 20.4 Hz, C-2), 66.9 (C-4), 63.0 (C-6).

19F NMR (470 MHz, CDCl3): δ = –141.5 (dd, J 1,F = 50.2 Hz, J 2,F = 11.8 Hz).

LRMS: m/z = 621.3 [M + Na]+.

HRMS: m/z calcd for C34H27FO9Na: 621.1537; found: 621.1524 [M + Na]+.

As discussed, the data we report for 6β are inconsistent with those reported by Ziegler et al., who instead report data consistent with 6α.[24] Of further note, Ziegler et al. report a geminal 2 J H1,F coupling constant of 26.7 Hz which is unusually low for glycosyl fluorides (typical range: 43–59 Hz).[40]


#

2,3,4,6-Tetra-O-benzoyl-5-C-bromo-β-d-mannopyranosyl Fluoride (11β)

A vigorously stirred suspension of β-fluoride 6β (170 mg, 0.28 mmol), dried, finely powdered K2CO3 (150 mg) and Br2 (73 μL, 1.4 mmol) in dry CCl4 (15 mL) was irradiated at reflux for 6 h. After this time, the irradiation was ceased and the product mixture was cooled to r.t. The product was then diluted with EtOAc (150 mL) and washed successively with sat. Na2S2O3 (30 mL), sat. NaHCO3 (30 mL) and brine (30 mL) before it was dried (MgSO4), filtered and concentrated to dryness (rotary evaporator water bath <30 °C). Purification by flash chromatography (PhMe/EtOAc, 40:1) delivered bromide 11β as a yellow oil (75 mg, 40%). Unreacted β-fluoride 6β was also recovered (52 mg, 30%; yield brsm: 56%).

[α]D 23 –110.5 (c 0.5, CHCl3).

1H NMR (500 MHz, CDCl3): δ = 8.21–8.18 (m, 2 H, Ph), 8.02–7.99 (m, 2 H, Ph), 7.95–7.93 (m, 2 H, Ph), 7.83–7.80 (m, 2 H, Ph), 7.68–7.63 (m, 1 H, Ph), 7.57–7.53 (m, 2 H, Ph), 7.51–7.43 (m, 3 H, Ph), 7.42–7.38 (m, 2 H, Ph), 7.30–7.26 (m, 4 H, Ph), 6.29 (d, J 3,4 = 10.2 Hz, 1 H, H-4), 6.21–6.19 (m, 1 H, H-2), 6.15 (dd, J 1,2 = 1.4 Hz, J 1,F = 47.8 Hz, 1 H, H-1), 5.94 (ddd, J 2,3 = 3.1 Hz, J 3,F = 0.9 Hz, 1 H, H-3), 5.17, 4.67 (ABq, J 6a,6b = 12.2 Hz, 2 H, H-6a, H-6b).

13C NMR (125 MHz, CDCl3): δ = 165.3, 165.1, 165.0, 164.7 (4 × C=O), 133.9, 133.6, 133.5, 133.4, 130.1, 130.0, 129.9, 129.8, 129.3, 129.0, 128.7, 128.6(2), 128.6(1), 128.6(0), 128.4, 128.2 (Ph), 104.8 (d, J 1,F = 223.5 Hz, C-1), 97.0 (d, J 5,F = 7.8 Hz, C-5), 69.7 (d, J 3,F = 8.9 Hz, C-3), 67.7 (d, J 2,F = 18.0 Hz, C-2), 66.2 (C-6), 66.1 (C-4).

19F NMR (470 MHz, CDCl3): δ = –153.0 (ap. d, J 1,F = 48.0 Hz).

LRMS: m/z = 699.6 [M + Na]+.

HRMS: m/z calcd for C34H26BrFO9Na: 699.0642; found: 699.0647 [M + Na]+.


#

2,3,4,6-Tetra-O-benzoyl-5-C-bromo-α-d-mannopyranosyl Fluoride (11α) and 2,3,4-Tri-O-benzoyl-5-C-benzoyloxy-6-bromo-6-deoxy-α-d-mannopyranosyl Fluoride (12)

A vigorously stirred suspension of α-fluoride 6α (126 mg, 0.21 mmol), dried, finely powdered K2CO3 (100 mg) and Br2 (50 μL, 1.0 mmol) in dry CCl4 (15 mL) was irradiated at reflux under Ar(g) for 24 h. After this time, the prescribed conditions were maintained while additional Br2 (20 equiv) was added in four portions over a 24 h period. The product mixture was then cooled to r.t. and irradiation was ceased before CHCl3 (40 mL) was added. The product was washed with 1 M Na2S2O3 (3 × 20 mL), sat. NaHCO3 (20 mL) and brine (20 mL) before it was dried (MgSO4), filtered and concentrated to dryness (rotary evaporator water bath <30 °C). 1H NMR spectroscopy revealed a product ratio of 66:34 (11α/12). Purification by flash chromatography (n-hexane/PhMe/EtOAc, 20:20:1 → 2:2:1 → 0:0:1) delivered an inseparable mixture of bromides 11α and 12 as a yellow oil (85 mg, 59% combined).

Key data for 11α: Rf = 0.34 (n-hexane/PhMe/EtOAc, 20:20:1; stained brown with p-anisaldehyde/H2SO4 dip).

1H NMR (500 MHz, CDCl3): δ = 6.40 (d, J 3,4 = 10.6 Hz, 1 H, H-4), 6.27 (dd, J 2,3 = 3.2 Hz, 1 H, H-3), 6.00 (dd, J 2,F = 3.2 Hz, 1 H, H-2), 5.97 (dd, J 1,2 = 1.8 Hz, J 1,F = 49.1 Hz, 1 H, H-1), 5.12, 4.55 (ABq, J 6a,6b = 12.5 Hz, 2 H, H-6a, H-6b).

13C NMR (125 MHz, CDCl3): δ = 105.9 (d, J 1,F = 233.1 Hz, C-1), 94.4 (C-5), 69.0 (d, J 2,F = 18.4 Hz, C-2), 67.2 (C-3), 66.7 (C-6), 65.7 (C-4).

19F NMR (470 MHz, CDCl3): δ = –135.1 (dd, J 1,F = 49.1 Hz, J 2,F = 3.2 Hz).

Key data for 12: Rf = 0.34 (n-hexane/PhMe/EtOAc, 20:20:1; stained brown with p-anisaldehyde/H2SO4 dip).

1H NMR (500 MHz, CDCl3): δ = 6.68 (d, J 3,4 = 10.8 Hz, 1 H, H-4), 6.27 (dd, J 2,3 = 3.2 Hz, 1 H, H-3), 5.97–5.95 (m, 1 H, H-2), 5.93 (dd, J 1,2 = 1.7 Hz, J 1,F = 49.4 Hz, 1 H, H-1), 4.80, 3.85 (ABq, J 6a,6b = 11.6 Hz, 2 H, H-6a, H-6b).

13C NMR (125 MHz, CDCl3): δ = 105.5 (d, J 1,F = 228.9 Hz, C-1), 102.6 (C-5), 68.7 (d, J 2,F = 9.2 Hz, C-2), 68.0 (C-4), 67.3 (d, J 3,F = 1.9 Hz, C-3), 30.8 (C-6).

19F NMR (470 MHz, CDCl3): δ = –132.2 (dd, J 1,F = 49.6 Hz, J 2,F = 1.9 Hz).

Data for 11α and 12:

1H NMR (500 MHz, CDCl3): δ = 8.19–8.12 (m, 2 H, Ph), 8.05–8.03 (m, 1 H, Ph), 8.01–7.98 (m, 1 H, Ph), 7.95–7.91 (m, 1 H, Ph), 7.85–7.80 (m, 2 H, Ph), 7.63–7.52 (m, 4 H, Ph), 7.48–7.35 (m, 6 H, Ph), 7.31–7.27 (m, 3 H, Ph).

13C NMR (125 MHz, CDCl3): δ = 165.4, 165.1, 165.0(8), 165.0(4), 165.0(1), 164.9, 164.8, 163.7 (8 × C=O), 140.0, 133.9(2), 133.8(8), 133.8(6), 133.8(1), 133.7, 133.6, 133.5, 133.4, 133.3, 130.2, 130.1, 130.0, 129.9, 129.8(9), 129.8(4), 129.7(9), 129.7(6), 128.8, 128.7(7), 128.7(4), 128.6(9), 128.6(5), 128.6(2), 128.6(0), 128.5(5), 128.5(0), 128.4, 128.3, 128.2 (Ph).

LRMS: m/z = 699.9 [M + Na]+.

HRMS: m/z calcd for C34H26BrFO9Na: 699.0642; found: 699.0644 [M + Na]+.


#

2,3,4,6-Tetra-O-benzoyl-α-l-gulopyranosyl Fluoride (2)

To a solution of bromide 11β (70 mg, 0.10 mmol) and Et3B (10 μL, 1.0 M in hexanes, 0.01 mmol) in anhydrous PhMe (3 mL, containing dissolved O2) under Ar(g) was added n-Bu3SnH (155 μL, 1.0 M in hexanes, 0.15 mmol) dropwise over the course of 5 min. Stirring was continued at r.t. for 4 h before additional Et3B (0.1 equiv) and n-Bu3SnH (1.0 equiv) were added in succession. After a further 24 h, the solvent was removed under reduced pressure. The crude product residue was taken up in acetonitrile (50 mL) and washed with n-hexane (3 × 20 mL). The acetonitrile phase was then concentrated to dryness. 1H and 19F NMR analysis revealed the presence of α-l-gulosyl fluoride 2 and β-d-mannosyl fluoride 6β in an isomeric ratio of 53:47 (2/6β). Subsequent purification by flash chromatography (PhMe/EtOAc, 40:1) gave an analytical sample of 2 as a white foam (combined yield 2 + 6β: 41 mg, 65%).

Data for 2: Rf = 0.47 (PhMe/EtOAc, 30:1; stained dark green with p-anisaldehyde/H2SO4 dip).

[α]D 24 –32.4 (c 0.9, CHCl3).

1H NMR (500 MHz, CDCl3): δ = 8.18–8.16 (m, 2 H, Ph), 8.14–8.11 (m, 2 H, Ph), 8.04–8.00 (m, 2 H, Ph), 7.91–7.88 (m, 2 H, Ph), 7.66–7.61 (m, 2 H, Ph), 7.57–7.48 (m, 6 H, Ph), 7.43–7.39 (m, 2 H, Ph), 7.36–7.31 (m, 2 H, Ph), 6.05 (dd, J 1,2 = 1.9 Hz, J 1,F = 52.9 Hz, 1 H, H-1), 5.93 (dd, J 2,3 = 2.9 Hz, J 3,4 = 4.1 Hz, 1 H, H-3), 5.67 (dd, J 4,5 = 4.1 Hz, 1 H, H-4), 5.58 (ddd, J 2,F = 25.9 Hz, 1 H, H-2), 5.02 (ddd, J 5,6a = 6.1 Hz, J 5,6b = 5.0 Hz, 1 H, H-5), 4.65 (dd, part A of ABX, J 6a,6b = 12.4 Hz, 1 H, H-6a), 4.50 (dd, part B of ABX, 1 H, H-6b).

13C NMR (125 MHz, CDCl3): δ = 166.0, 165.4, 165.0, 164.8 (4 × C=O), 133.9, 133.7, 133.6, 133.2, 130.1, 130.0, 129.9, 129.8, 129.4, 129.2, 128.7, 128.6, 128.5, 128.4 (Ph), 103.9 (d, J 1,F = 233.0 Hz, C-1), 68.3 (C-3), 66.4 (d, J 5,F = 2.3 Hz, C-5), 66.2 (d, J 2,F = 23.5 Hz, C-2), 66.0 (C-4), 62.3 (C-6).

19F NMR (470 MHz, CDCl3): δ = –147.5 (dd, J 1,F = 55.3 Hz, J 2,F = 26.8 Hz).

LRMS: m/z = 1219.7 [2 M + Na]+.

HRMS: m/z calcd for C34H27FO9Na: 621.1537; found: 621.1556 [M + Na]+.


#

2,3,4-Tri-O-benzoyl-5-C-benzoyloxy-6-deoxy-α-d-mannopyranosyl Fluoride (15)

To a solution of bromides 11α and 12 (65 mg, 0.095 mmol) and Et3B (10 μL, 1.0 M in hexanes, 0.01 mmol) in anhydrous PhMe (2 mL, containing dissolved O2) under Ar(g) was added n-Bu3SnH (143 μL, 1.0 M in hexanes, 0.14 mmol) dropwise over the course of 5 min. Stirring was continued at r.t. for 4 h before the solvent was removed under reduced pressure. The crude product residue was taken up in acetonitrile (40 mL) and washed with n-hexane (3 × 20 mL). The acetonitrile phase was then concentrated to dryness. 1H and 19F NMR analysis revealed the presence of α-d-mannosyl fluoride 6α and 6-deoxymannosyl fluoride 15 in an isomeric ratio of 68:32 (6α/15). Subsequent purification by flash chromatography (n-hexane/PhMe/EtOAc, 20:20:1 → 10:10:1 → 0:0:1) gave an analytical sample of 15 as a colourless oil (combined yield: 46 mg, 80%).

Data for 15: Rf = 0.31 (n-hexane/PhMe/EtOAc, 20:20:1; stained blue with p-anisaldehyde/H2SO4 dip).

[α]D 24 –55.1 (c 0.2, CHCl3).

1H NMR (500 MHz, CDCl3): δ = 8.20–8.16 (m, 2 H, Ph), 8.13–8.09 (m, 2 H, Ph), 8.03–7.99 (m, 2 H, Ph), 7.84–7.80 (m, 2 H, Ph), 7.68–7.62 (m, 2 H, Ph), 7.57–7.51 (m, 5 H, Ph), 7.47–7.42 (m, 2 H, Ph), 7.41–7.37 (m, 3 H, Ph), 6.34 (dd, J 2,3 = 3.3 Hz, J 3,4 = 10.7 Hz, 1 H, H-3), 6.05 (d, 1 H, H-4), 5.94 (dd, J 1,2 = 1.6 Hz, 1 H, H-2), 5.87 (dd, J 1,F = 48.4 Hz, 1 H, H-1), 2.11 (s, 3 H, CH3).

13C NMR (125 MHz, CDCl3): δ = 165.7, 165.3, 165.1, 163.9 (4 × C=O), 133.9, 133.7, 133.4(3), 133.4(1), 130.6, 129.9, 129.8(5), 129.8(0), 129.7, 128.7, 128.6(8), 128.6(6), 128.5, 128.3, 105.0 (d, J 1,F = 226.3 Hz, C-1), 103.4 (C-5), 71.5 (C-4), 69.1 (d, J 2,F = 39.8 Hz, C-2), 65.5 (d, J 3,F = 2.0 Hz, C-3), 23.0 (C-6).

19F NMR (470 MHz, CDCl3): δ = –131.8 (dd, J 1,F = 48.4 Hz, J 2,F = 2.2 Hz).

LRMS: m/z = 620.9 [M + Na]+.

HRMS: m/z calcd for C34H27FO9Na: 621.1537; found: 621.1532 [M + Na]+.


#
#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

Dr Tri Le (UQ) and Dr Desmond Sim (UQ) are thanked for their assistance with NMR experiments.

Supporting Information

    Supporting information for this article is available online at https://doi-org.accesdistant.sorbonne-universite.fr/10.1055/a-2149-4586.
  • Supporting Information

Corresponding Author

Vito Ferro
School of Chemistry and Molecular Biosciences, The University of Queensland
Brisbane, Queensland 4072
Australia   

Publication History

Received: 10 July 2023

Accepted after revision: 07 August 2023

Accepted Manuscript online:
07 August 2023

Article published online:
14 September 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 Previous fluorine-directed syntheses of l-iduronic acid (a) and l-idose (b), and the current application of this chemistry towards l-gulose
Zoom Image
Scheme 2 Chemical synthesis of α- and β-mannopyranosyl fluorides 6α and 6β
Zoom Image
Figure 1 Mannosyl fluorides 7α and 7β are truncated derivatives of 6α and 6β, respectively.
Zoom Image
Scheme 3 Postulated reaction pathways to anomeric fluorides 6α and 6β from 8. Reagents and conditions: (a) XtalFluor-M®, Et3N, Et3N·3HF, CH2Cl2, –78 °C to r.t. Counterions have been omitted for clarity.
Zoom Image
Scheme 4 Mannosyl fluorides 6β and 6α displayed substantially different reactivities towards photo-bromination.
Zoom Image
Scheme 5 Mechanistic proposal for the generation of 12
Zoom Image
Scheme 6 n-Bu3SnH reductions of 11β and the mixture of bromides 11α and 12