Stereoselective organocatalyzed glycosylations – thiouracil, thioureas and monothiophthalimide act as Brønsted acid catalysts at low loadings† †Electronic supplementary information (ESI) available: Detailed experimental procedures, characterisation and copies of NMR spectra. See DOI: 10.1039/c8sc027

Thiouracil catalyzes stereoselective glycosylations with galactals in loadings as low as 0.1 mol%.

S4 hydride (60 wt% in mineral oil, 2.4 g, 60 mmol) was added portionwise with vigorous stirring; upon complete addition the reaction mixture was allowed to return to room temperature and then it was left to stir for 30 minutes. The mixture was re-cooled to 0 °C and freshly distilled allyl bromide (5.2 mL, 60 mmol) was added dropwise with stirring. The mixture was left to stir at ambient temperature for 16 h, becoming dark yellow-brown in appearance. Methanol (5 mL) was added slowly to the reaction to quench the reaction and the mixture was concentrated in vacuo to a yellow residue, which was taken up in dichloromethane (20 mL) and washed with water (3 × 20 mL) and brine (3 × 20 mL). The organic phase was dried with magnesium sulfate and concentrated to a pale yellow oil in vacuo (approx. 2 g). Purification via silica gel flash chromatography (6:1 to 4:1 cyclohexane:EtOAc, R f = 0.41) afforded 1,2-dideoxy-3,4,6-tri-O-allyl-D-lyxo-1hexenpyranose 3b as a slightly pale yellow oil (1.50 g, 56%). 1 -6). NMR data were consistent with literature data. 5

3,4-O-Dibenzyl-L-fucal 3h
Following a literature procedure, 7 3,4-O-diacetyl-L-fucal (250 mg, 1.17 mmol) was dissolved in a solution of MeOH (8 mL), Et 3 N (1 mL) and H 2 O (1 mL) and stirred at room temperature. TLC analysis (2:1, cyclohexane/ethyl acetate, H 2 SO 4 stain) after 24 h showed complete deacetylation of the starting material. The reaction mixture was concentrated in vacuo to afford L-fucal as a white solid. Under a N 2 atmosphere, L-fucal was dissolved in anhydrous THF (3 mL, 0.4 M). The flask was cooled to 0 ºC and NaH (60% dispersion in mineral oil) (187 mg, 4.67 mmol) was added to the solution. The ice-bath was removed and the reaction was stirred at room temperature for 30 minutes. The reaction mixture was again cooled to 0 ºC and treated slowly with BnBr (0.50 mL, 4.2 mmol). The ice-bath was removed and the reaction mixture was left to stir at room temperature. TLC analysis (3:1, cyclohexane/ethyl acetate, H 2 SO 4 stain) after 18 h showed complete consumption of L-fucal. The reaction mixture was quenched with MeOH (2 mL), diluted with DCM (40 mL), washed with 1 M HCl (15 mL) followed by saturated NaHCO 3 (15 mL) and brine (15 mL 16.7 (C-6). NMR data were in agreement with literature. 9
NaOH (12.4 ml, 1mM). The mixture was stirred for 30 min at room temperature. BnBr (1.4 ml, 12 mmol) was then added to the reaction mixture. The mixture was heated at reflux for 3 days (time unoptimised). TLC analysis of the reaction mixture (ethyl acetate) showed a small amount of starting material remained (R f = 0.35) and three new spots appeared in the reaction mixture (R f = 0.85, 0.75 and 0.65). 1 H NMR analysis of the reaction mixture showed that a small amount of starting material remained (~5%) and that the desired products had formed (3-OH/2-OH; 2.4:1 based on integrations of H-7). The aqueous and organic layers were separated and the aqueous layer was extracted with CH 2 Cl 2 (3 × 50 ml). The combined organic layers were washed with NaHCO 3 (100 ml), brine (100 ml), dried over MgSO 4 and filtered using Büchner filtration. The solvent was removed from the filtrate using rotary evaporation which gave a brown oil (9 g). Column chromatography was performed (75:25; cyclohexane/ethyl acetate) but did not lead to the isolation of pure products (R f = 0.24  and 0.14 (2-OH)). Concentration of the solution of the mixed fractions containing the 3-OH product and other impurities, but free of 2-OH product, allowed the desired product 6b to crystallise from the remaining ethyl acetate as a white solid (1.48 g, 37% yield). Mixed fractions containing the 2-OH product and other impurities, but free of the 3-OH product, were concentrated using rotary evaporation and dissolved using a mixture of CH 2 Cl 2 /cyclohexane/MeOH (4.75:0.25:5). Removal of the CH 2 Cl 2 by rotary evaporation allowed the desired product 6a to crystallize from the remaining cyclohexane/methanol as a white solid (0.58 g, 15% yield). 0.5 g (13% yield) remained as mixed fractions containing the two desired products plus unidentified impurities.

Methyl 2,3-di-O-benzyl-4,6-O-benzylidene-α-D-glucopyranoside 15
Under a N 2 atmosphere, 14 (1.99 g, 7.1 mmol) as weighed into the flask and dissolved in anhydrous DMF (70 ml). The solution was cooled to 0 °C (50:50; ice/water). NaH (60% dispersion in mineral oil) (1.19 g, 29.7 mmol) was added to the reaction mixture. The icebath was removed and the reaction mixture was left to stir at room temperature for 30 min. The reaction was again cooled to 0 °C and BnBr (3.5 ml, 29 mmol) was added dropwise to the reaction. The reaction was left stirring at room temperature, under a N 2 atmosphere, for 16.5 h (time unoptimised). TLC analysis (7:3; cyclohexane/ethyl acetate; H 2 SO 4 stain (10-15% EtOH)) against a pure sample of product showed that the starting material was consumed in the reaction and the desired product had formed (R f = 0.73). MeOH (1 ml) was added to the reaction mixture to quench the reaction. The solvents were removed using rotary evaporation which gave a yellow solid. The solid obtained was dissolved in CH 2 Cl 2 (75 ml) and washed with deionised water (2  75 ml). The aqueous layer was then extracted with CH 2 Cl 2 (3  75 ml). The organic layers were combined and washed with brine (150 ml) and then dried with MgSO 4 , filtered and concentrated in vacuo. Purification by column chromatography (85:15; cyclohexane/ethyl acetate) gave a white solid (2.71 g, 82% yield). A small impurity (~10%) was present in this sample following column chromatography. 1.51 g of this sample was recrystallised using n-hexanes which gave 15 as white solid (1.28 g, 39% yield).

Methyl 2,3,6-tri-O-benzyl-α-D-glucopyranoside 6c
Based on literature procedures, 14,15 4Å molecular sieves (powdered) (450 mg) were weighed into a three-necked round bottom flask. This flask was flame-dried using a propane torch, allowed to cool under vacuum and then switched to a N 2 atmosphere. The flask was equipped with a stir-bar, gas inlet and thermometer. Pyranoside 15 (1 g, 2 mmol) was added, followed by NaCNBH 3 (1.0M in THF, 35 ml, 35 mmol). The mixture was cooled to 0 °C (50:50; ice/water;). HCl (4.0M in dioxane, 10 ml) was added to the reaction mixture until the pH reached 1-2 and no more fizzing was observed upon addition. The reaction mixture was left to warm to room temperature. TLC analysis (4:1; cyclohexane/ethyl acetate) showed that the sugar starting material (R f = 0.26) had been consumed in the reaction and two new spots appeared (R f = 0.11 and a baseline spot). The reaction mixture was diluted with EtOAc (40 ml) and deionised H 2 O (40 ml) and then filtered using Büchner filtration. The filtrate was transferred to a separating funnel and sat. NaHCO 3 solution (40 ml) was added. The aqueous layer was extracted with EtOAc (3  40 ml). The combined organic layers were washed with brine (80 ml). The organic layer was dried using Na 2 SO 4 , filtered and concentrated in vacuo. Purification by column chromatography (7:3; cyclohexane/ethyl acetate) gave 6c as a colourless oil (0.76 g, 76% yield).

Phenyl 2,3,4-tri-O-benzyl-β-O-D-thioglucopyranoside 6d
Following modified literature procedures, 5,16 16a (3.9 g, 5.7 mmol) was dissolved in methanol (150 ml) and THF (30 ml). 3M NaOH aq (24 ml) was added and the reaction was stirred for 16 h. TLC (cyclohexane:EtOAc; 2:1) showed full consumption of starting material (R f = 0.5) and a new product at R f = 0.0. The reaction was concentrated under reduced pressure to remove the THF and MeOH. The resulting solution was neutralised (pH 7) with 2M HCl. The aqueous layer was extracted with EtOAc (15 ml × 15) and solvent removed under reduced pressure to give crude 16b as an off-white solid (1.32 g, 88%). Crude 16b (960 mg, 3.53 mmol) and imidazole (490 mg, 7.2 mmol) were charged to a RBF and dried under high vacuum for 3 h. Anhydrous DMF (12.5 ml) was added to the reaction and when everything was in solution TIPSCl (2 ml, 9.4 mmol) was added to the RBF under N 2 and stirred for 18 h. The solution was concentrated under reduced pressure and the residue was dissolved in CH 2 Cl 2 (50 ml), washed with water (35 ml), brine (30 ml), dried over The intermediate 16c was then dissolved in anhydrous DMF (16 ml) and the solution was cooled in an ice bath. NaH (60% dispersion in mineral oil) (790 mg, 19.8 mmol) was charged to the flask and the reaction was stirred for 1 h at RT. Then the reaction was cooled in an ice bath before benzyl bromide (2.3 ml, 19 mmol) was added dropwise. The reaction was left stirring under N 2 at RT for 72 h. TLC analysis (cyclohexane:EtOAc; 2:1; H 2 SO 4 (10-15% EtOH) stain) showed that intermediates were still present. NaH (60% dispersion in mineral oil) (20 mg, 5.0 mmol) was charged to the flask and the reaction was stirred for 1 h. Benzyl bromide (0.6 ml, 5 mmol) was added dropwise and the reaction stirred overnight (time unoptimised). The reaction was quenched with MeOH (5 ml) and solvent concentrated under reduced pressure. The residue was then dissolved in CH 2 Cl 2 (50 ml), washed with water (35 ml), brine (30 ml), dried over MgSO 4 , filtered and concentrated under reduced pressure to give 16d. Following purification by column chromatography (cyclohexane:EtOAc; 98:2) phenyl 2,3,4-tri-O-benzy-6-O-triisopropylsilyl-β-O-D-thio glucopyranoside 16d was obtained as a white solid (1.48 g, 60%).  Acceptors 6e-h were purchased from commercial suppliers and used without further purification.

Thiouracil-catalysed Glycosylations
General Procedure 1 A 5 ml RBF equipped with a stir-bar, gas-inlet and reflux condenser was set up under a N 2 atmosphere. The glycal donor (1.2 eq) was weighed into the round-bottomed flask and placed under vacuum for ca. 30 min. An anhydrous solution of acceptor (0.8M in CH 2 Cl 2 ) was made by charging a known quantity of acceptor to a flask under N 2 . Anhydrous CH 2 Cl 2 was added to make up a 0.8M solution. The stock solution was dried by adding MgSO 4 (1:1 mol/mol w.r.t. acceptor) and let sit for 30 min. The stock solution of acceptor in CH 2 Cl 2 (1 eq) was then decanted by syringe and added to the glycal donor under N 2 . After everything was in solution, 2-thiouracil (1 mol%) was added giving a suspension. The reaction was refluxed under N 2 for 18 h or until TLC or NMR analysis showed the reaction was complete. Some solvent loss was noted over the course of the reaction and the higher concentration that results is believed to be beneficial. The solution was then concentrated under reduced pressure and purified by column chromatography.

General Procedure 2
The glycal donor (0.6 mmol) and acceptor (0.5 mmol) were weighed into a round bottomed flask equipped with a stirring bar under N 2 and put under vacuum for 40 minutes. The flask was refilled with N 2 before anhydrous CH 2 Cl 2 (0.6 ml) was added to make a 0.8M solution wrt the acceptor. After everything was in solution 2-thiouracil (1 mol%) was added to the solution under N 2 . The reaction was refluxed under N 2 for 18 h or until TLC or NMR analysis showed the reaction was complete. Some solvent loss was noted over the course of the reaction and the higher concentration that results is believed to be beneficial. The solution was then concentrated under reduced pressure and purified by column chromatography. Table 1

Investigation of catalyst solubility
Using a volumetric flask, 2-thiouracil 2 (1.1 mg  0.3, 8.6 μmol) was dissolved in MeOH (10 ml; 0.86mM). Serial dilutions were made by removal of a known aliquot from the stock solution (using a pipette) and making up to a known volume using a volumetric flask. The following serial dilutions were obtained:
We note that 9a underwent anomerisation under extended reaction times with the amount of , increasing relative to ,ß over time. Thus after 112 h at reflux the ,:,ß ratio was determined to be 13:1 by 1 H NMR spectroscopy. Some degradation was also noted. A similar time-dependent anomerisation was previously observed by Yoshimura and co-workers in their synthesis of α,α-trehalose. 29

Procedures for Benzyl group removal
General Procedure for removal of benzyl protecting groups by hydrogenation: The protected 2-deoxyglycoside was weighed into a round bottom flask and dissolved in a mixture of methanol/ethyl acetate (9:1). Pd (10% on carbon) (10 mol% for each benzyl group to be removed) was then added to the solution. The atmosphere was changed to hydrogen first by placing the reaction solution under house vacuum (200 mbar), closing the reaction flask to house vacuum and then purging the flask with hydrogen (using a hydrogen balloon). The cycle was repeated three times and on the final purge with hydrogen the hydrogen balloon was left in place. The reaction was monitored using TLC. The reaction mixture was filtered using Celite ® . The solution was concentrated using rotary evaporation and purified using column chromatography (8:2; dichloromethane/methanol; 2% H 2 O). For further purification the product isolated from column chromatography was passed through a plug of Octadecyl-C18-Silica (8:2; methanol/H 2 O).

Mechanistic Studies
Glycosylation reactions in the absence of the catalyst

Control reaction in reflux condenser
In a slight modification to General Procedure 1, galactal donor 3a (251 mg, 0.6 mmol) and an anhydrous CH 2 Cl 2 solution of galactose acceptor 4 (0.8M 0.6 ml, 0.5 mmol) were used. Analysis of the 1 H NMR spectrum of the crude mixture after 20 h showed only the starting materials. An analogous reaction with glucal donor 3f also showed no reaction in the absence of catalyst.

Control reaction with uracil instead of thiouracil as catalyst
In a slight modification to General Procedure 1, galactal donor 3a (200 mg, 0.48 mmol) and an anhydrous CH 2 Cl 2 solution of galactose acceptor 4 (0.83M 0.48 ml, 0.40 mmol) and uracil (0.4 mg, 4 mol) were used. Analysis of the 1 H NMR spectrum of the crude mixture after 19 h did not detect any disaccharide product.

Control reaction in sealed vessel
Following General Procedure 3, galactal donor 3a (201 mg, 0.48 mmol) and an anhydrous CH 2 Cl 2 solution of galactose acceptor 4 (1.8M (w.r.t. acceptor and solvent total volume), 0.25 ml, 0.5 mmol) were used. Analysis of the 1 H NMR spectrum of the crude mixture after 18 h showed only the starting materials.

No reaction of Galactal with Water in the absence of catalyst
In a slight modification to the General Procedure 1, galactal donor 3a (255 mg, 0.6 mmol), deionised H 2 O (9 μl, 0.5 mmol), and anhydrous CH 2 Cl 2 (0.8M) were used. Analysis of the 1 H NMR spectrum of the crude reaction mixture after 16 h showed only the starting materials.

No reaction of galactal and p-toluenesulfonamide in the absence of catalyst:
In a slight modification to General Procedure 1, galactal 3a (200 mg, 0.48 mmol), ptoluenesulfonamide (68 mg, 0.40 mmol) and anhydrous CH 2 Cl 2 (0.48 ml) were used. TLC analysis during the reaction showed no consumption of 3a. NMR analysis of the crude reaction material after 20 h showed only unreacted starting materials.

Disaccharide Mixture Submitted to Standard Reaction Conditions
In a slight modification to the general procedure 1, the α/β-5a (53 mg, 0.08 mmol) was submitted to the standard 2-thiouracil-catalysed glycosylation conditions. 2-Thiouracil (0.1 mg, 1 mol%) and anhydrous CH 2 Cl 2 (0.31M, 0.26 ml) were used. Analysis of the 1 H NMR spectrum of the crude mixture after 16 h showed only the unchanged starting material α/β-5a (no change in the α/β ratio was observed).

3,4,6-Tri-O-benzyl-2-deoxy-/β-D-galactopyranosyl p-toluenesulfonamide 12
Following a literature procedure, 25 the sulfonamide 12 was prepared as an / mixture (: = 8:92). The title compound was subjected to the standard 2-thiouracil-catalysed conditions (General Procedure 3). The reaction mixture was heated to reflux for 18 h. Analysis of the 1 H NMR spectrum of the crude reaction mixture showed no significant change in the / ratio.

1,1'-Linked disaccharides
We note that 1,1'-linked disaccharides did undergo anomerisation under extended reaction times with the amount of , increasing relative to ,ß over time.

Control Reaction of galactal 3a with CD 3 OD in the absence of catalyst
In a slight modification to General Procedure 3, galactal donor 3a (250 mg, 0.59 mmol) and CD 3 OD (20 μl, 0.49 mmol) were weighed into a crimp top vial. The vial was sealed and the atmosphere was changed to nitrogen. The mixture was dissolved in anhydrous CH 2 Cl 2 (2.5M w.r.t. acceptor, 0.2 ml) and heated at reflux for 24 h. TLC and 1 H NMR analysis of the crude reaction mixture showed the desired product had not formed and some breakdown of the galactal donor had occurred (aldehyde peak at 10.1 ppm in the 1 H NMR spectrum). The mixture was heated at reflux for a further 24 h, however TLC and 1 H NMR showed no desired product was formed and further breakdown of the galactal donor was observed.   The proton and carbon NMR data were consistent with the data provided above when 2thiouracil was used as a catalyst for the reaction. Probing syn/anti addition in the synthesis of 1,1'-linked disaccharides

Synthesis of Monothiophthalimide 13
Following the literature procedure, 34 a round bottom flask equipped with a magnetic stir-bar, a condenser and gas inlet (all flame-dried) was placed under a N 2 atmosphere. Phthalimide (3.00 g, 20.4 mmol) was weighed into the reaction flask and dissolved in anhydrous THF (45 ml). Lawesson's reagent (8.34 g, 20.6 mmol) was added and the mixture was heated to 60 °C. The mixture was heated at 60 °C for 11 h. A strong colour change from yellow (time = 0 h) to dark purple (time = 11 h) was observed. TLC analysis (4:1; cyclohexane/ethyl acetate) showed that the phthalimide starting material (R f = 0.14) remained and two new phthalimidederived spots appeared (R f = 0.25, 0.18). Several other spots related to Lawesson's reagent were also observed (R f = 0.51, 0.41, 0.07 and a baseline spot). The reaction mixture was concentrated using a rotary evaporator. The dark solid obtained was purified by column chromatography (95:5; cyclohexane/ethyl acetate) which gave the desired product as a pink solid (1.56 g, 47% yield) and dithiophthalimide as a brown solid (0.93 g, 50% yield). Analysis of the 1 H NMR spectra showed low levels of impurities still present in both the mono-and di-thiophthalimide. The solids were recrystallised using toluene. Monothiophthalimide 13 was obtained as a pink solid (0.3 g, 9% yield). Dithiophthalimide 20 was obtained as a brown solid (0.15 g, 8% yield).

Control reaction with monothiophthalimide 13 instead of thiouracil as catalyst
In a slight modification to General Procedure 1, galactal donor 3a, an anhydrous CH 2 Cl 2 solution of galactose acceptor 4 and monothiophthalimide 13 (1 mol%) were used. The disaccharide 5a was obtained in 84% yield after purification.

Probing Catalyst-Substrate Interactions using 1 H NMR spectroscopy
Under a nitrogen atmosphere, a solution of monothiophthalimide 13 (4.9 mg, 0.03 mmol) was prepared in CD 2 Cl 2 (1.5 ml, concentration 0.02M) and dried over 4Å molecular sieves. An aliquot (0.7 ml) of this solution was analysed by 1 H NMR spectroscopy, the NH proton was observed at  8.84 ppm. A CD 2 Cl 2 solution of galactal donor 3a (24.0 mg, 0.058 mmol) and 13 was prepared from the parent stock solution (0.8 ml, 0.072M w.r.t. 3a, and 0.02M w.r.t. 13) and dried over 4Å molecular sieves. The mixture of 3a and 13 was analysed by 1 H NMR spectroscopy. In the presence of galactal 3a (3.7 equiv.), a small downfield shift for the NH proton of 13 was observed in the 1 H NMR spectrum ( 8.84 changed to 8.94 ppm). Under a nitrogen atmosphere, a solution of monothiophthalimide 13 (8.3 mg, 0.051 mmol) was prepared in CD 2 Cl 2 (2.5 ml, concentration 0.02M) and dried over 4Å molecular sieves. An aliquot (0.7 ml) of this solution was analysed by 1 H NMR spectroscopy, the NH proton was observed at  8.86 ppm. A CD 2 Cl 2 solution of diacetone galactose 4 (43.3 mg, 0.166 S52 mmol) and 13 was prepared from the parent stock solution (1.0 ml, 0.166M w.r.t. 4, and 0.02M w.r.t. 13) and dried over 4Å molecular sieves. The solution of 13 and 4 was added (0.6 mL) to sample of 13 and the mixture was analysed by 1 H NMR spectroscopy. In the presence of alcohol 4 (3.7 equiv.), a downfield shift for the NH proton of 13 was observed in the 1 H NMR spectrum ( 8.86 changed to 9.26 ppm). The OH proton also sharpened and showed a shift from 2.11 to 2.15 ppm.