**1. Introduction**

The use of renewable energies instead of conventional fossil fuels has become a global trend from the perspective of expected fossil fuel depletion and global climate change. Biodiesel fuel, which is produced by the transesterification of vegetable or animal fats with methanol, has emerged as a promising alternative to petroleum-derived diesel fuel [1,2]. The biodiesel manufacturing process inevitably provides 10 wt% of glycerol (1,2,3-propanetriol) as the main side product, and an oversupply of the crude glycerol would be projected with the growing biodiesel market [3]. Thus, great effort has been devoted to the development of an efficient process for the transformation of glycerol to value-added commodity chemicals, which would indirectly contribute to sustainable biodiesel production [4–7]. Among them, optically active glycerol derivatives would be an attractive target in the field of glycerol valorization. Chiral glycerol derivatives such as glyceraldehyde and glycidyl tosylate are utilized as valuable C3 building blocks in medicinal [8–12] and synthetic organic chemistry [13–18]. A chiral pool approach is a traditional strategy to access enantiopure glycerol derivatives, but the need for multi-step transformations may be a major drawback [19–23]. Asymmetric desymmetrization of glycerol would be one of the most straightforward methods for chiral glycerol derivatives production. Several types of enzymes, i.e., lipase, kinase, and dehydrogenase/oxidase, have been successfully applied to this strategy, affording optically active glycerols with various enantioselectivities [24–29]. On the other hand, despite the recent development of the enantioselective desymmetrization [30,31] of 1,2-diols [32–39] and 1,3-diols [40–48], including C2-substituted glycerols [49–53], the non-enzymatic direct desymmetrization of glycerol is still a challenging task presumably due to an extremely high hydrophilic nature of glycerol. In this context, the use of 2-*O*-protected glycerol derivatives would be the most common strategy for the chemical desymmetrization of glycerol (Scheme 1a) [54–57]. In

**Citation:** Yamamoto, K.; Miyamoto, K.; Ueno, M.; Takemoto, Y.; Kuriyama, M.; Onomura, O. Copper-Catalyzed Asymmetric Sulfonylative Desymmetrization of Glycerol. *Molecules* **2022**, *27*, 9025. https://doi.org/10.3390/ molecules27249025

Academic Editors: Alison Rinderspacher, Mircea Darabantu and Gloria Proni

Received: 26 November 2022 Accepted: 16 December 2022 Published: 18 December 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

2013, Tan et al. developed the first non-enzymatic direct desymmetrization of glycerol through organocatalyzed enantioselective silylation (Scheme 1b) [58]. In their protocol, the high enantioselectivity was achieved through the secondary kinetic resolution on the initially formed monosilylated glycerol. Very recently, the copper-catalyzed sulfonylative desymmetrization of glycerol using a non-commercially available ligand with silver carbonate was described in the Chinese patent [59]. Although the desired product was obtained with high enantioselectivity under copper-catalyzed conditions, the use of a commercially available chiral ligand and a non-precious metal base would be desirable from a practical and economical point of view [60–62]. Herein, we report the asymmetric desymmetrization of glycerol through sulfonylation with a Cu/(*R*,*R*)-PhBOX complex and sodium carbonate affording the optically active monosulfonylated glycerol in an excellent yield and enantioselectivity (Scheme 1c) [63].

$$\begin{array}{cccc} \text{sure}\_{\text{C}} \stackrel{\mathsf{H}}{\underset{\text{C}}{\rightleftharpoons}} & \text{sure} & \text{sued} \\ \text{sure}\_{\text{C}} \stackrel{\mathsf{H}}{\underset{\text{C}}{\rightleftharpoons}} & \text{sowe} & \text{sowe} \end{array} \quad \begin{array}{c} \text{40} \\ \text{sowe} \stackrel{\mathsf{H}}{\underset{\text{C}}{\rightleftharpoons}} & \text{sowe} \end{array}$$

$$\begin{array}{ccccc} \text{s} & \text{ $\upharpoonright$ } & \text{ $\vothlef} \text{$ \upharpoonright $} & \text{$ \vothlef} \text{ $\vothlef} \\ \text{s} & \text{$ \upharpoonright $} & \text{$ \vothlef} \end{array} \quad \begin{array}{ccccc} \text{ $\upharpoonright$ } & \text{ $\upharpoonright$ } & \text{ $\upharpoonright$ } \\ \text{ $\upharpoonright$ } & \text{ $\upharpoonright$ } & \text{ $\upharpoonright$ } \end{array}$$

**Scheme 1.** Asymmetric desymmetrization of glycerols. (**a**) Enantioselective desymmetrization of protected glycerols. (**b**) Organocatalyzed enantioselective silylation. (**c**) This work: Cu-catalyzed enantioselective sulfonylation. The asterisk denotes the chiral center.

#### **2. Results and Discussion**

For the initial attempt to optimize the enantioselective desymmetrization of glycerol, compound **1** was treated with *p*-toluenesulfonyl chloride (TsCl) in the presence of copper trifluoromethanesulfonate (Cu(OTf)2)/(*R*,*R*)-PhBOX and sodium carbonate in acetonitrile. Pleasingly, the desired monotosylated glycerol **2** was obtained in 91% yield with 83% ee (Table 1, entry 1). Using other carbonate salts, i.e., potassium carbonate and cesium carbonate, resulted in a decrease in both yield and enantioselectivity (entries 2 and 3). Organic bases were not suitable for the present transformation (entries 4 and 5). Next, other copper catalysts were examined to evaluate the catalytic activity in this reaction system. While CuCl provided **2** with a slightly lowered yield and enantioselectivity, CuBr and CuI exhibited a similar reactivity compared with Cu(OTf)2 (entries 6–8). The use of CuCN led to the formation of **2** in 83% yield with higher enantioselectivity, and the reaction concentration was able to be doubled without significant changes regarding both yield and enantioselectivity (entries 9–10). Pleasingly, we found that acetone was a better solvent choice to afford the desired product in an excellent yield and enantioselectivity (96% yield, 94% ee), and the concentration of 0.25 M was found to be suitable for the present reaction (entries 11–12). The catalyst loading was able to be reduced to 5 mol% without a significant decrease in the yield and enantioselectivity (entry 13). We also examined the feasibility of the gram-scale preparation of **2**. The reaction with 6.0 mmol of glycerol successfully provided the desired product **2** in 88% yield (1.30 g) with 93% ee (entry 14). Control experiments revealed that both the copper salt and the BOX ligand were essential to promote the tosylation of **1** (entries 15–16).


**Table 1.** Optimization of reaction conditions 1.

 Reaction conditions: **1** (1.0 mmol), TsCl (1.2 mmol), [Cu] (0.1 mmol), (*R*,*R*)-PhBOX (0.1 mmol), base (1.5 mmol), solvent (0.125 M), rt, 3 h. <sup>2</sup> Isolated yield after column chromatography. <sup>3</sup> Determined by chiral HPLC analysis. Reaction time (12 h). <sup>5</sup> Solvent (0.25 M). <sup>6</sup> Solvent (0.5 M). <sup>7</sup> CuCN (0.05 mmol), (*R*,*R*)-PhBOX (0.05 mmol), 10 h. **1** (6.0 mmol), 9 h. <sup>9</sup> The reaction was carried out in the absence of (*R*,*R*)-PhBOX.

In order to gain insight into the chemoselectivity of the present reaction, we performed competition studies with alcohol additives (Table 2). The addition of *n*-propanol (1.0 eq) led to a slight decrease in both yield and enantioselectivity, but the formation of *n*-propyl tosylate was not detected (entry 1 vs. entry 2). Moreover, selective sulfonylation of glycerol (**1**) over 1,2- and 1,3-diols was observed under the present reaction conditions, and the desired monosulfonylated glycerol **2** was obtained without a significant loss of enantioselectivity (entry 1 vs. entries 3–4). In addition, the reaction of 2-*O*-benzylglycerol (**4**) provided the corresponding monotosylated product **5** in a low yield with poor enantioselectivity (Scheme 2). These results indicated that the present reaction system would be highly selective for the glycerol transformation even in the presence of other alcohols, and the presence of a free 2-hydroxy moiety would play a crucial role in accelerating the tosylation with high asymmetric induction.

**Scheme 2.** Enantioselective desymmetrization of 2-*O*-benzylglycerol.


**Table 2.** Competition studies 1.

<sup>1</sup> Reaction conditions: **1** (1.0 mmol), additive (1.0 mmol), TsCl (1.2 mmol), CuCN (0.1 mmol), (*R*,*R*)-PhBOX (0.1 mmol), Na2CO3 (1.5 mmol), acetone (0.25 M), rt, 3 h. <sup>2</sup> Isolated yield after column chromatography. n.d. = not detected. <sup>3</sup> Determined by chiral HPLC analysis.

With successful asymmetric desymmetrization of glycerol achieved, we then investigated the synthetic applications of the obtained optically active glycerol. First, the siteselective protection of the remained hydroxy groups in (*R*)-**2** was examined (Scheme 3). The primary hydroxy moiety was selectively protected by using *tert*-butyldimethylsilyl chloride (TBSCl) with imidazole, affording the corresponding product (*R*)-**6**. Acetylation of the secondary hydroxy group with Ac2O in the presence of a DMAP catalyst provided (*R*)-**7** in an excellent yield. Since each protective group would be removed by different deprotecting protocols, (*R*)-**7** would be potentially useful as a versatile chiral C3 building block.

**Scheme 3.** Synthesis of optically active glycerol derivatives.

Next, we turned our attention to the application of the present transformation in the synthesis of an optically active synthetic ceramide. Ceramides are major components of the lamellar structure in stratum corneum lipids which protect the epidermis from excess transepidermal water loss and from the permeation of pathogens [64]. Interestingly, optically active natural ceramides showed different thermotropic behavior from racemic variants, and the lamellar liquid crystalline system containing optically active natural ceramides improved recovering effects of a water-holding ability and a barrier function of the skin [65]. Sphingolipid E (SLE) was a synthetic ceramide designed and synthesized by Kao Corporation as a structural analog of natural type 2 ceramide and was found to form a stable lamellar structure that exhibits a high water-holding ability [66]. Although SLE has been utilized as a racemic form, optically active SLE might affect the physicochemical property to form a lamellar structure and the interaction mode with water molecules. The synthesis of optically active SLE commenced with the introduction of nitrogen functionality to (*R*)-**2** (Scheme 4). The nucleophilic azide substitution of the tosyloxy group in (*R*)-**2** with sodium azide provided azide diol (*S*)-**8** in an excellent yield in the presence of 15-crown-5. The enantiomeric excess of (*S*)-**8** was determined after the tosylation of the primary hydroxy group, and no obvious racemization was observed in the azide substitution step. The boronic acid-catalyzed site-selective alkylation [67] of (*S*)-**8** with cetyl bromide followed by the Pd/C-catalyzed reduction of the azide group afforded the corresponding aminoalcohol (*S*)-**10**. *N*-Alkylated product (*S*)-**11** was successfully obtained by the reductive amination of (*S*)-**10** with TBS-protected glycolaldehyde using 2-picoline borane as a reductant. Finally,

(*S*)-**11** was transformed into the optically active synthetic ceramide (*S*)-**12** via the amidation with palmitoyl chloride followed by the removal of the TBS group.

**Scheme 4.** Synthesis of an optically active synthetic ceramide.

In conclusion, we have developed the copper-catalyzed asymmetric sulfonylative desymmetrization of glycerol. The reaction smoothly proceeded under mild reaction conditions with a commercially available (*R*,*R*)-PhBOX ligand and an inexpensive inorganic base, providing the optically active monotosylated glycerol derivative in a high yield with high enantiomeric excess. The synthetic utility of the present transformation was demonstrated by the preparation of an enantio-enriched C3 building block with three different types of protective groups. Moreover, the synthesis of the optically active synthetic ceramide was also achieved from the monotosylated glycerol in six steps without a notable loss of enantiopurity.
