Minimization of Amounts of Catalyst and Solvent in NHC-Catalyzed Benzoin Reactions of Solid Aldehydes: Mechanistic Consideration of Solid-to-Solid Conversion and Total Synthesis of Isodarparvinol B

Abstract

Attempts were made to minimize the amounts of catalyst and solvent in the NHC-catalyzed benzoin reactions of solid aldehydes. In some case, solid-to-solid conversions proceeded in the solvent-free NHC-catalyzed benzoin reactions. Even if a mixture of the substrate, N-heterocyclic carbene (NHC) precursor, and inorganic base was initially a powdery solid, the reaction did proceed at reaction temperature lower than the melting points of each compound. The solid mixture partially melted or became a slurry or suspension in the meantime. We call this solid/liquid mixture a semisolid state. The reaction giving an optically active product was faster than that giving a racemic mixture of the same product. Melting-point depression was observed for a series of mixtures of the substrate and product in different substrate/product ratios. Solvent-free solid-to-solid conversions were accelerated by the formation of a semisolid state resulting from the melting-point depression of the solid substrate accompanied by the product formation. In the case of solid substrates with high melting points, melting-point depression was useless, and the addition of a small amount of solvent was needed. The first total synthesis of isodarparvinol B was achieved via the NHC-catalyzed intramolecular benzoin reaction using a small amount of solvent as an additive.

Introduction

Because of excellent organocatalytic performance, N-heterocyclic carbenes (NHCs) have attracted much attention of synthetic chemists.¹⁻⁵ NHCs are generated from precursors, such as thiazolium, imidazolium, and triazolium salts, upon treatment with a base. NHCs promote various C–C bond-forming reactions via polarity inversion (Umpolung) of the formyl group into an acyl anion equivalent. NHC organocatalysts show diverse reactivities, high enantioselectivities, and broad substrate scope. The benzoin reaction,^1a,1b,2,3 the Stetter reaction,^1e,1s,5 and other reactions via homoenolates^1h,1i,1r and acyl azoliums^1k,1n,1o are typical examples of NHC-catalyzed reactions. We synthesized bicyclic and tricyclic compounds with contiguous quaternary stereocenters by the intramolecular crossed benzoin reaction catalyzed by NHC.⁴ Except for a few examples,⁶ most NHC-catalyzed reactions require a catalyst loading as much as 5–20 mol %, which may restrict the large-scale utilization of NHCs. However, methods for reducing the amount of triazolium salts, which are expensive, will enhance the opportunity of the industrial use of NHCs.

Solvent-free reactions are important from the viewpoint of green chemistry.⁷ Recently, we have found that solvent-free NHC-catalyzed reactions proceeded with a small amount of catalyst.⁸ For example, the solvent-free intermolecular benzoin reaction of aromatic aldehydes required only 0.2 mol % catalyst, although 5–10 mol % catalyst is usually used in an organic solvent.² Although the intramolecular benzoin reaction typically uses 20 mol % catalyst,^3,4 0.5 mol % catalyst was found to catalyze highly selective reactions under solvent-free conditions.⁸ The catalyst loading could be reduced to 0.2 mol % in the solvent-free intramolecular Stetter reaction in contrast to a typical catalyst loading of 20 mol % in an organic solvent,⁵ and no intermolecular benzoin or Stetter reactions took place under solvent-free conditions.⁸

In the study of the solvent-free NHC-catalyzed reactions, interestingly, we found solid-to-liquid or solid-to-solid conversions. Even if a mixture of the substrate, NHC precursor, and inorganic base was initially a powdery solid, the reaction did proceed at reaction temperature lower than the melting points of each compound, where a solid mixture partially melted or became a suspension in the meantime. We call this solid/liquid mixture a semisolid state.⁸ The semisolid state was generated in most solid-to-solid conversion reactions. We proposed that this phenomenon might be related to the melting-point depression; as the product accumulates slowly in the powdery state, decreasing the purity of the substrate, a semisolid state appears as a result of the melting-point depression, where the fluidity of substances significantly increases to accelerate the reaction. It is reported by other researchers that a liquid reaction mixture is formed in a reaction of solid substances under solvent-free conditions because a eutectic with a lower melting point is generated from the substrate and product.^9,10 There is also an example of a noncatalytic solid–state reaction accelerated by the vapor of the organic solvent,¹¹ which demonstrates the importance of fluidity of molecules even in the solid state. Only a limited number of catalytic solid-to-solid conversions are known, such as the aldol reaction and ether synthesis reported by Scott⁹ and the Suzuki–Miyaura reaction and hydrogenation reported by Monguchi and Sajiki.¹² In view of the importance of the NHC-catalyzed reactions with high atom efficiency producing no byproducts, which are achieved with low catalyst loading, here, we minimized the amounts of catalyst and solvent in the NHC-catalyzed benzoin reactions of solid aldehydes. To address the mechanistic aspect, we investigated the effects of chirality of NHC catalysts, stirring of the reaction mixture, and a small amount of solvent on the solid-to-solid conversions. In addition, we synthesized a natural product, isodarparvinol B, using the NHC-catalyzed intramolecular benzoin reaction in a key step.

Results and Discussion

Solvent-free Solid-to-Solid Conversions

In our previous study,⁸ solid-to-solid conversions took place when chiral NHC precatalyst B (Figure 1) was used in no solvent, where a powdery solid mixture partially melted and changed into a semisolid state or became a slurry or suspension. This phenomenon might be related to the melting-point depression. Here, we investigated whether chiral or achiral catalysts would make a difference in the outcome of the reactions because it is well known that identical chiral compounds with different enantiomeric purities have different melting points; for example, a racemic mixture has a higher melting point than a nonracemic counterpart in most cases. To a Schlenk flask were added Cs₂CO₃, precatalyst A or B, and solid aldehyde in this order, and the powdery mixture was gently stirred under Ar in a thermostatic bath (procedure A). Otherwise, after a mixture of Cs₂CO₃ and precatalyst A or B in a Schlenk flask was gently stirred under Ar at 30 °C for 1 h to facilitate the generation of NHC, solid aldehyde was added, and the mixture was gently stirred under Ar at constant reaction temperature in a thermostatic bath (procedure B).

Figure 1.

NHC precatalysts A, B, and ent-B.

We conducted the intermolecular benzoin reaction of solid aldehydes 1a–c using achiral precatalyst A or chiral precatalyst B (Table 1). The benzoin reaction of 1a–c with precatalyst B was carried out at reaction temperature lower than the melting points of 1a–c (entries 1, 4, and 8). As a result, despite low catalyst loading (0.5 mol %), solid aldehyde 1a was converted into solid benzoin product 2a in 98% yield with 78% ee (entry 1). The reaction of solid aldehyde 1b at 30 °C also afforded solid product 2b in 99% yield with 79% ee (entry 4). Although solid aldehyde 1c showed poor reactivity at 30 °C, the reaction proceeded at 50 °C (entry 8). We next examined the benzoin reaction of 1a–c using achiral precatalyst A under otherwise the same reaction conditions. As a result, 1a and 1b showed little or no reactivity (entries 2 and 5). In view of the difference in physical properties between the two precatalysts, A and B, we also employed a racemic mixture of chiral precatalysts B and ent-B instead of achiral precatalyst A. As a result, the semisolid state did not appear, giving 2b in 13% yield. Because a racemic mixture of a chiral product has higher crystallinity and a higher melting point than the optically active nonracemic counterpart, a semisolid state is less likely to generate in the former case (racemic mixture). In fact, the yields were improved at a slightly elevated temperature (entries 3, 6, and 7). In the case of aldehyde 1c, the yield obtained with A at 50 °C was comparable to that obtained with B (entry 9).

Table 1. Intermolecular Benzoin Reaction under Solvent-free Conditions^a.

entry	1	NHC	loading (mol %)	T (°C)	yield (%)b	ee (%)c
1d	1a	B	0.5	30	98	78
2	1a	A	0.5	30	4
3	1a	A	0.5	40	94
4d	1b	B	1	30	99	79
5	1b	A	1	30	2
6	1b	A	1	40	75
7	1b	A	1	50	84
8d	1c	B	1	50	87	73
9	1c	A	1	50	85

^aConditions: aldehyde 1 (5.0 mmol), NHC precatalyst (amount indicated above), Cs₂CO₃ (1 equiv with respect to precatalyst), Ar, 12 h (procedure A).

^bIsolated yield.

^cDetermined by chiral HPLC. The (R)-enantiomers were predominant.

^dData taken from ref (8).

We next examined the intramolecular benzoin reaction of solid substrate 3 using NHC precursors A and B (Table 2). In all cases, after stirring a mixture of A or B and Cs₂CO₃ for 1 h, 3 was added (procedure B). When chiral precatalyst B was used at 40 °C for 48 h, solid product 4 was obtained in 76% yield with 53% ee (entry 1). When achiral precatalyst A was used under otherwise the same reaction conditions, a racemic mixture of 4 was obtained in 56% yield (entry 2). The same reaction at 60 °C afforded 4 in 86% yield (entry 3), whereas increasing the reaction temperature to 80 °C decreased the yield to 64%, giving aldol condensation product 4′ in 31% yield (entry 4). Clearly, achiral precatalyst A promoted the solid-to-solid conversions less efficiently than chiral precatalyst B (Tables 1 and 2). This is because a racemic product has higher crystallinity than a nonracemic counterpart; the former has a higher melting point and is difficult to generate a semisolid state. This conclusion was supported by another control experiment; the use of a racemic mixture of B and ent-B at 40 °C resulted in almost no reaction.

Table 2. Intramolecular Benzoin Reaction under Solvent-free Conditions^a.

entry	NHC	T (°C)	yield (%)b	ee (%)c
1d	B	40	76 (16)	53
2	A	40	56 (1)
3	A	60	86 (9)
4	A	80	64 (31)

^aConditions: aldehyde 3 (2.0 mmol), NHC precatalyst (1 mol %), Cs₂CO₃ (1 mol %), Ar, 48 h (procedure B).

^bIsolated yield for 4. The data in parentheses are the isolated yields for byproduct 4′.

^cDetermined by chiral HPLC. (S)-4 was predominant.

^dData taken from ref (8).

Melting-Point Depression

In all the solvent-free solid-to-solid conversions that proceeded efficiently, we observed that a solid mixture partially melted or became a slurry or suspension, which may be related to melting-point depression. To check this possibility, we ground mixtures of the substrate and product in different substrate/product ratios and measured their melting points. As a result, melting-point depression was confirmed in all cases (Supporting Information). Optically active products had lower melting points than the corresponding racemic products, and melting-point depression was more remarkable in the former case than in the latter case; some mixtures started to melt at around reaction temperature. Although it may be difficult to reproduce a real reaction mixture just by mixing the substrate with the product, these data strongly support that a semisolid state induced by the melting-point depression accelerates the solid-to-solid conversions. Indeed, when we monitored the progress of the solvent-free benzoin reaction of solid aldehyde 1c, the reaction proceeded remarkably in 0.5–2 h, during which the reaction mixture became a semisolid (Figure 2).

Figure 2.

Time course of the solvent-free benzoin reaction of 1c.

Effect of Stirring on Solid-to-Solid Conversions

Monguchi, Sajiki, and co-workers have reported that in solvent-free Pd-catalyzed hydrogenation, solid-to-solid conversions proceeded without stirring.¹² Inspired by this surprising report, we examined the solvent-free benzoin reaction of solid aldehyde 1a without stirring to investigate the effect of stirring on the solvent-free solid-to-solid conversion. To a Schlenk flask were added Cs₂CO₃, precatalyst B, and aldehyde 1a in this order, and the powdery mixture under Ar was left at 30 °C without stirring. The benzoin product 2a was obtained in 4% yield with 61% ee after 12 h (not shown). When the same reaction without stirring was carried out for three weeks in duplicate, the appearance of the reaction mixtures gradually changed without generation of the semisolid state (Figure 3), and product 2a was obtained in 26% yield with 72% ee or 31% yield with 75% ee. The reaction without stirring was found to be much slower than that with stirring because the same reaction but with stirring afforded 2a in 98% yield with 78% ee in 12 h (Table 1, entry 1). Clearly, stirring is important for this solid-to-solid conversion even if stirring is gentle. It should be noted that this solid-to-solid conversion without stirring was reproducible. We consider that this reaction without stirring was successful partly because 1a is a sublimable substance. In fact, a needle crystal grew on the solid after two weeks (Figure 3). Because 1a can sublime to diffuse and make a contact with the catalyst, the reaction may proceed even without stirring.

Figure 3.

Photographs of solvent-free benzoin reaction of 1a without stirring.

Effect of a Small Amount of Solvent

Under solvent-free conditions, solid substrates are much less reactive than liquid substrates. Solid substrates with high melting points, such as 1c (mp 62 °C) and 3 (mp 112–114 °C), needed to be heated to make a semisolid state; however, the reaction of 3 at higher temperature resulted in the formation of byproduct 4′ (Table 2). It is therefore ideal to make a semisolid or slurry state without elevating reaction temperature. We decided to investigate whether the reaction is facilitated by the addition of a small amount of solvent.

We examined the effect of a small amount of THF on the benzoin reaction of solid substrate 1a (2 mmol) at 30 °C (Table 3, entries 1 and 2). Although the solvent-free reaction catalyzed by 0.5 mol % catalyst A afforded racemic product 2a in 21% yield, the addition of a small amount of THF (only 100 μL) enhanced the yield to 95%. The effect of a small amount of THF was also examined for solid substrates 1b–c (entries 3–6). The addition of THF (100 μL) remarkably enhanced the yields of a racemic mixture of 2b or 2c. Figure 4a shows photographs of a reaction mixture of 1c in the presence of THF (100 μL). A slurry gradually changed into a solid as the reaction proceeded.

Table 3. Effect of a Small Amount of Organic Solvent on Intermolecular Benzoin Reaction^a.

entry	1	additive	yield (%)b
1	1a		21
2	1a	THF	95
3	1b		7
4	1b	THF	100
5	1c		5
6	1c	THF	93

^aConditions: aldehyde 1 (2.0 mmol), NHC precatalyst A (0.5 mol %), Cs₂CO₃ (0.5 mol %), THF as an additive (0 or 100 μL), Ar, 30 °C, 12 h (procedure A).

^bYield was determined using 2-methoxynaphthalene as an internal standard.

Figure 4.

Photographs of reaction mixtures before and after the benzoin reaction of (a) 1c, (b) 3, and (c) 9 using the organic solvent as an additive.

A small amount of organic solvent was also used in the intramolecular benzoin reaction of solid substrate 3 (2 mmol) (Table 4). Although almost no reaction proceeded at 30 °C with achiral precatalyst A under the solvent-free conditions, the addition of toluene (100 μL) greatly improved the yield of racemic product 4 (entries 1–2). The asymmetric intramolecular benzoin reaction of 3 with chiral precatalyst B was also promoted by adding a small amount of solvent (entries 3–6). A solvent-containing viscous powder changed into a dry solid (Figure 4b). The side reaction giving 4′ was suppressed using chiral precatalyst B and the organic solvent (entries 4–6).

Table 4. Effect of a Small Amount of Organic Solvent on Intramolecular Benzoin Reaction^a.

entry	NHC	additive	yield (%)b	ee (%)c
1	A		trace
2	A	toluene	72 (13)
3	B		63 (10)	50
4	B	THF	95 (3)	51
5	B	1,4-dioxane	90 (2)	53
6	B	CHCl3	87 (2)	41

^aConditions: aldehyde 3 (481 mg, 2.0 mmol), NHC precatalyst (1 mol %), Cs₂CO₃ (1 mol %), additive (0 or 100 μL), Ar, 30 °C, 48 h (procedure B).

^bIsolated yield for 4. The data in parentheses are the isolated yields for byproduct 4′.

^cDetermined by chiral HPLC. (S)-4 was predominant.

Total Synthesis of Isodarparvinol B

Various natural products have been synthesized utilizing the NHC-catalyzed benzoin reaction and other reactions.^13,14 However, about 10 mol % catalyst loading has been used in most cases. We have found that the amount of NHC catalyst could be reduced under solvent-free conditions.⁸ In the case of solid substrates, solid-to-liquid or solid-to-solid conversions took place via a semisolid state; however, solid substrates with high melting points exhibited little or no reactivity. In such a case, a small amount of organic solvent facilitated the benzoin reaction of solid substrates with low catalyst loading (0.5–1 mol %) as described above (Tables 3 and 4). To further demonstrate the usefulness of the abovementioned reaction conditions, we decided to synthesize a natural product containing a benzoin skeleton, isodarparvinol B (5), for the first time, using the NHC-catalyzed intramolecular benzoin reaction in a key step. Isodarparvinol B (5) is an isoflavanonol derivative isolated by Umehara and co-workers from the heartwood of a medicinal plant in Thailand, Dalbergia parviflora.¹⁵

The synthetic route to (−)-5 is shown in Scheme 1. The opposite enantiomer, (+)-5, was also synthesized using the opposite enantiomer of NHC precatalyst B, ent-B (not shown in Scheme 1). Compounds 6 and 8 were prepared from starting materials in one and four steps, respectively, according to the refs (14c) and (16). Ketone 6 was subjected to α-bromination with phenyltrimethylammonium tribromide to give 7 in 66% yield. Compounds 7 and 8 were then connected by the Williamson ether synthesis, and the subsequent deprotection of the acetal group afforded 9 in 62% yield. Aldehyde 9 was subjected to the NHC-catalyzed intramolecular benzoin reaction to furnish optically active 4-chromanone 10 in 99% yield with 86% ee. Finally, hydrogenation of 10 gave (−)-5 in 93% yield with 90% ee (38% total yield from a starting material). ¹H and ¹³C NMR signals for 5 accorded with those reported for natural isodarparvinol B, and the sign of the specific rotation value of the synthetic product, (−)-5, was identical to that of the natural product.^15,17 To determine the absolute configuration of (−)-5, CD spectroscopy was employed (Supporting Information). (−)-5 showed a positive Cotton effect in the carbonyl n−π* transition region around 320–360 nm,¹⁷ which strongly suggests that (−)-5 has the (S)-configuration according to the modified octant rule.¹⁸ This assignment was supported by TD-DFT calculations (Supporting Information).

Scheme 1. Total Synthesis of (−)-Isodarparvinol B, (−)-5.

We searched for reaction conditions suitable for the NHC-catalyzed benzoin reaction of 9. In view of the high melting point of 9 (mp 139–140 °C), we added a small amount of CHCl₃, a good solvent for 9. Several attempts according to procedure B resulted in the formation of both benzoin product 10 and aldol condensation product 10′, the latter of which resulted from the action of the base Cs₂CO₃. To completely generate NHC from precatalyst B and Cs₂CO₃, CHCl₃ was initially added, and the mixture was stirred at 30 °C for 1 h, to which 9 (0.5 mmol) was added (procedure C). As a result, the side reaction could be significantly suppressed. The results of further optimization are summarized in Table 5. The reaction at 30 °C with chiral precatalyst B (2 mol %) gave benzoin product 10 in 59% yield with 82% ee (entry 1). When reaction temperature was elevated to 40 °C, 10 was successfully obtained in 99% yield with 86% ee (entry 2). Photographs of the reaction mixture before and after this reaction are shown in Figure 4c. A wet solid changed into a gummy solid. Further elevation of temperature to 50 °C hampered the selective synthesis of 10 (46% yield, 78% ee), giving byproduct 10′ in 27% yield (entry 3). Catalyst loading could be reduced to 1.5 mol % to give a high yield (93%), whereas a catalyst loading of 1 mol % resulted in a low yield (22%) (entries 4 and 5). It was confirmed that procedure C (entry 2) under the optimized conditions was superior to procedure B (entry 6). When the amount of CHCl₃ was reduced to 200 μL, 10 was obtained in 64% yield (entry 7). Based on these results, entry 2 was the best choice. This heterogeneous reaction using a small amount of solvent as an additive (entry 2) was comparable to the ordinary homogeneous reaction (entry 8).

Table 5. Optimization of Intramolecular Benzoin Reaction^a.

entry	loading (mol %)	T (°C)	time (h)	yield (%)b	ee (%)c
1	2	30	48	59 (9)	82
2	2	40	24	99 (0)	86
3	2	50	12	46 (27)	78
4	1.5	40	24	93 (0)	86
5	1	40	24	22 (0)	89
6d	2	40	24	85 (14)	80
7e	2	40	24	64 (4)	85
8f	2	40	24	96 (0)	90

^aConditions: aldehyde 9 (294 mg, 0.5 mmol), NHC precatalyst B (amount indicated above), Cs₂CO₃ (1 equiv with respect to precatalyst B), CHCl₃ (600 μL), and N₂ (procedure C).

^bIsolated yield for 10. The data in parentheses are the isolated yields for byproduct 10′.

^cDetermined by chiral HPLC.

^dProcedure B.

^eCHCl₃ (200 μL).

^fCHCl₃ (3.6 mL) was added to make a homogeneous solution.

Conclusions

Organic solvents are used in most organic reactions to make a homogeneous solution of compounds. Reduction of the organic solvent will be beneficial for saving fossil fuels, energy, resources, and cost, and development of solvent-free reactions is important from the viewpoint of green chemistry. Recent successful examples of efficient solvent-free catalysis have encouraged further research and development.¹⁹ However, removing the solvent is one of the most difficult things, especially when the substrate is a solid. In our previous study on NHC-catalyzed reactions, the amount of catalyst could be significantly reduced when no solvent was used,⁸ and in the case of solid substrates, solid-to-solid or solid-to-liquid conversions proceeded even at reaction temperature below the melting points. Here, we gained mechanistically useful information: (1) A semisolid or slurry appeared in the middle of the reactions. (2) The reaction giving an optically active product was faster than that giving a racemic mixture of the product. (3) Melting-point depression was observed for a series of mixtures of the substrate and product in different substrate/product ratios. Based on these observations, we concluded that solvent-free solid-to-solid conversion reactions were accelerated by the formation of a semisolid state or a slurry or suspension resulting from the melting-point depression of the solid substrate accompanied by product formation, decreasing the purity of the substrate. In the case of sublimable substrates, solid-to-solid conversions proceeded slowly without stirring because mass transfer occurs via sublimation. In the case of substrates with high melting points, melting-point depression was useless, and the addition of a small amount of solvent for making a pseudo-semisolid state was needed. Isodarparvinol B was synthesized via the NHC-catalyzed intramolecular benzoin reaction using a small amount of solvent as an additive. We expect that catalytic solid-to-solid conversions with or without a small amount of organic solvent may be employed in industrial synthetic processes in future.

Experimental Section

General Methods

NMR spectra were measured on a Varian 400-MR spectrometer or a JEOL JNM-ECS400 spectrometer, and chemical shifts are reported as the delta scale in ppm using an internal reference [δ = 7.26 (CDCl₃) or 3.31 (CD₃OD) for ¹H NMR and δ = 77.16 (CDCl₃) or 49.0 (CD₃OD) for ¹³C NMR]. IR spectra were recorded on a Shimadzu IRAffinity-1 spectrophotometer. HPLC was performed on a Shimadzu LC-20AT/SPD-20A. Optical rotations were measured on a Horiba SEPA-300 polarimeter at the sodium D line. UV and CD spectra were measured on a Shimadzu UV-2600 spectrophotometer and a JASCO J-1500 spectropolarimeter, respectively. Melting points were measured on a Yanaco melting point apparatus (uncorrected). Column chromatography was carried out using Fuji Silysia BW-127 ZH (100–270 mesh), and thin layer chromatography was performed on Merck silica gel 60 F₂₅₄.

Solvent-free Intermolecular Benzoin Reaction

All the NHC precatalysts and aldehydes 1 were purchased. Aldehydes 1a–c were solids in a pure form at room temperature: mp 45–46 °C for 1a, 54–56 °C for 1b, and 57–59 °C for 1c. The enantiomeric purities of 2a–c were determined by HPLC using a chiral column (Daicel), and the absolute configurations were determined by comparison with the signs of the reported specific rotation values.

Typical Procedure (Procedure A)

Cs₂CO₃ (8.15 mg, 0.025 mmol, 0.5 mol %), NHC precatalyst (0.025 mmol, 0.5 mol %), and 1 (5.0 mmol) were added to a Schlenk flask in this order. The flask was quickly evacuated and filled with Ar. The mixture was gently stirred in a thermostatic bath for 12 h. The reaction was quenched with saturated aqueous NH₄Cl or 3% HCl (1 mL). The product was extracted with EtOAc or CHCl₃ (20 mL × 3). The organic layer was dried over Na₂SO₄ and concentrated. Purification by silica gel column chromatography [hexane/EtOAc (4:1)] gave 2.

rac-1,2-Bis(4-chlorophenyl)-2-hydroxyethanone (rac-2a)

659 mg (2.34 mmol, 94% yield, entry 3 in Table 1); white solid; mp 86–88 °C; ¹H NMR (CDCl₃, 400 MHz): δ 4.48 (d, J = 6.0 Hz, 1H), 5.88 (d, J = 5.9 Hz, 1H), 7.25 (d, J = 9.2 Hz, 2H), 7.31 (d, J = 8.5 Hz, 2H), 7.39 (d, J = 8.6 Hz, 2H), 7.83 (d, J = 8.6 Hz, 2H).

(R)-1,2-Bis(4-chlorophenyl)-2-hydroxyethanone ((R)-2a)⁸

691 mg (2.46 mmol, 98% yield, entry 1 in Table 1); pale yellow solid; mp 79–82 °C; [α]_D¹⁸ – 62.6 (c 1.02, CHCl₃), 78% ee, lit.^2e [α]_D²⁵ – 30.5 (c 0.66, CHCl₃) for (R)-2a with 85% ee; HPLC: Chiralpak IA, hexane/i-PrOH = 9:1, 0.5 mL/min, 254 nm, (R) 38.1 min, (S) 44.8 min.

rac-1,2-Bis(4-bromophenyl)-2-hydroxyethanone (rac-2b)

692 mg (1.87 mmol, 75% yield, entry 6 in Table 1); white solid; mp 98–100 °C; ¹H NMR (CDCl₃, 400 MHz): δ 4.47 (d, J = 6.0 Hz, 1H), 5.86 (d, J = 5.9 Hz, 1H), 7.18 (d, J = 8.4 Hz, 2H), 7.46 (d, J = 8.4 Hz, 2H), 7.56 (d, J = 8.5 Hz, 2H), 7.74 (d, J = 8.6 Hz, 2H).

(R)-1,2-Bis(4-bromophenyl)-2-hydroxyethanone ((R)-2b)⁸

915 mg (2.47 mmol, 99% yield, entry 4 in Table 1); pale yellow solid; mp 77–81 °C; [α]_D¹⁹ – 23.3 (c 1.02, CHCl₃), 79% ee, lit.^2e [α]_D²⁵ – 4.4 (c 0.63, CHCl₃) for (R)-2b with 30% ee; HPLC: Chiralpak IA, hexane/i-PrOH = 9:1, 0.5 mL/min, 254 nm, (R) 42.5 min, (S) 49.5 min.

rac-2-Hydroxy-1,2-bis(2-naphthyl)ethanone (rac-2c)

666 mg (2.13 mmol, 85% yield, entry 9 in Table 1); yellow solid; mp 124–126 °C; ¹H NMR (CDCl₃, 400 MHz): δ 4.72 (d, J = 6.3 Hz, 1H), 6.28 (d, J = 6.0 Hz, 1H), 7.42–7.59 (m, 5H), 7.75–7.82 (m, 5H), 7.87 (d, J = 8.2 Hz, 1H), 7.92 (s, 1H), 8.00 (dd, J = 1.6, 8.7 Hz, 1H), 8.50 (s, 1H).

(R)-2-Hydroxy-1,2-bis(2-naphthyl)ethanone ((R)-2c)⁸

679 mg (2.17 mmol, 87% yield, entry 8 in Table 1); yellow solid; mp 107–111 °C; [α]_D¹⁹ + 35.6 (c 1.00, MeOH), 73% ee, lit.^2e [α]_D²⁵ + 52.4 (c 0.37, MeOH) for (R)-2c with 93% ee; HPLC: Chiralpak IA, hexane/i-PrOH = 4:1, 1.0 mL/min, 254 nm, (R) 43.5 min, (S) 73.6 min.

Solvent-free Intramolecular Benzoin Reaction

Aldehyde 3 was prepared and characterized according to the ref (20). Spectroscopic data matched those in the reference. Aldehyde 3 was a solid in a pure form at room temperature: mp 112–114 °C. The enantiomeric purity of 4 was determined by HPLC using a chiral column (Daicel). The absolute configuration of 4 was determined by comparison with the sign of the reported specific rotation value.

Typical Procedure (Procedure B)

A mixture of the NHC precatalyst (0.02 mmol, 1.0 mol %) and Cs₂CO₃ (6.26 mg, 0.02 mmol, 1.0 mol %) was stirred under Ar in a Schlenk flask in a thermostatic bath at 30 °C for 1 h. Aldehyde 3 (480 mg, 2.00 mmol) was added, and the flask was quickly evacuated and filled with Ar. The mixture was gently stirred in a thermostatic bath for 48 h. The reaction was quenched with saturated aqueous NH₄Cl (1 mL). The product was extracted with EtOAc (20 mL × 3). The organic layer was dried over Na₂SO₄ and concentrated. Purification by silica gel column chromatography [hexane/EtOAc (4:1)] afforded 4.

rac-3-Hydroxy-3-phenyl-4-chromanone (rac-4)

411 mg (1.71 mmol, 86% yield, entry 3 in Table 2); white solid; mp 94–98 °C; ¹H NMR (CDCl₃, 400 MHz): δ 4.17 (s, 1H), 4.48 (d, J = 11.8 Hz, 1H), 4.86 (d, J = 11.8 Hz, 1H), 6.97 (d, J = 8.4 Hz, 1H), 7.07 (t, J = 7.5 Hz, 1H), 7.30–7.35 (m, 3H), 7.46–7.48 (m, 2H), 7.52 (dt, J = 1.5, 7.9 Hz, 1H), 7.94 (dd, J = 1.6, 7.9 Hz, 1H).

2-Benzoyl-1-benzofuran (4′)²¹

38 mg (0.17 mmol, 9% yield, entry 3 in Table 2); pale yellow solid; mp 82–84 °C; ¹H NMR (CDCl₃, 400 MHz): δ 7.34 (t, J = 7.6 Hz, 1H), 7.49–7.57 (m, 4H), 7.64 (t, J = 7.6 Hz, 2H), 7.74 (d, J = 7.9 Hz, 1H), 8.05 (d, J = 7.7 Hz, 2H); ¹³C NMR (CDCl₃, 100 MHz): δ 112.7, 116.7, 123.5, 124.1, 127.2, 128.5, 128.7, 129.6, 133.0, 137.4, 152.4, 156.2, 184.6; IR (KBr) 3146, 3130, 3075, 3057, 3036, 3026, 1641, 1614, 1599, 1545, 1474, 1331, 1298, 1219, 1190, 1177, 1121, 972, 899, 750, 743, 723, 694, 675, 575; HRMS (FAB) calcd for C₁₅H₁₁O₂, 223.0759; found, 223.0758 [M + H]⁺.

Solvent-free Benzoin Reaction of 1a without Stirring

Cs₂CO₃ (8.15 mg, 0.025 mmol, 0.5 mol %), NHC precatalyst B (11.7 mg, 0.025 mmol, 0.5 mol %), and 1a (703 mg, 5.0 mmol) were added to a Schlenk flask in this order. The flask was quickly evacuated and filled with Ar. The mixture was left without stirring in a thermostatic bath at 30 °C for 3 weeks. The reaction was quenched with saturated aqueous NH₄Cl (1 mL). The product was extracted with CHCl₃ (20 mL × 3). The organic layer was dried over Na₂SO₄ and concentrated. Purification by silica gel column chromatography [hexane/EtOAc (4:1)] gave 2a as a white solid (183 mg, 0.65 mmol, 26% yield).

Intermolecular Benzoin Reaction of 1 Using the Organic Solvent as an Additive

Cs₂CO₃ (3.26 mg, 0.01 mmol, 0.5 mol %), NHC precatalyst A (3.63 mg, 0.01 mmol, 0.5 mol %), and 1 (2.0 mmol) were added to a Schlenk flask in this order. The flask was quickly evacuated and filled with Ar. A small amount of THF (100 μL) was added, and the mixture was gently stirred in a thermostatic bath at 30 °C for 12 h. The reaction was quenched with saturated aqueous NH₄Cl (1 mL). The product was extracted with CHCl₃ (20 mL × 3). The organic layer was dried over Na₂SO₄ and concentrated. The product was quantified by ¹H NMR after the addition of 2-methoxynaphthalene as the internal standard. The results are shown in Table 3.

Intramolecular Benzoin Reaction of 3 Using the Organic Solvent as an Additive

A mixture of NHC precatalyst A or B (0.02 mmol, 1.0 mol %) and Cs₂CO₃ (6.52 mg, 0.02 mmol, 1.0 mol %) was stirred under Ar in a Schlenk flask in a thermostatic bath at 30 °C for 1 h. Aldehyde 3 (480 mg, 2.00 mmol) was added, and the flask was quickly evacuated and filled with Ar. A small amount of organic solvent (100 μL) was added, and the mixture was gently stirred in a thermostatic bath at 30 °C for 48 h. The product was purified by silica gel column chromatography [hexane/EtOAc (4:1)] to afford 4.

(S)-3-Hydroxy-3-phenyl-4-chromanone ((S)-4)

477 mg (1.98 mmol, 95% yield, entry 4 in Table 4); [α]_D²⁴ – 44.7 (c 1.03, CHCl₃), 51% ee, lit.⁸ [α]_D²⁶ – 38.9 (c 1.10, CHCl₃) for (S)-4 with 53% ee; ¹H NMR (CDCl₃, 400 MHz): δ 4.19 (s, 1H), 4.48 (d, J = 11.8 Hz, 1H), 4.86 (d, J = 11.5 Hz, 1H), 6.97 (d, J = 7.9 Hz, 1H), 7.07 (dt, J = 0.9, 7.5 Hz, 1H), 7.30–7.35 (m, 3H), 7.46–7.48 (m, 2H), 7.52 (ddd, J = 1.6, 7.0, 8.6 Hz, 1H), 7.94 (dd, J = 1.8, 7.8 Hz, 1H); ¹³C NMR (CDCl₃, 100 MHz): δ 73.5, 73.8, 118.0, 119.3, 122.0, 126.1, 127.7, 128.7, 128.8, 136.8, 138.5, 161.5, 194.6; HPLC: Chiralpak AD-H, hexane/i-PrOH = 9:1, 0.5 mL/min, 254 nm, (S) 28.5 min, (R) 35.2 min.

Total Synthesis of Isodarparvinol B (5)

Compounds 6 and 8 were prepared according to refs (14c) and (16), and the purities were checked by ¹H and ¹³C NMR spectra (Supporting Information).

1-(2,3-Bis(benzyloxy)-4-methoxyphenyl)-2-bromoethan-1-one (7)

A solution of phenyltrimethylammonium tribromide (5.22 g, 13.9 mmol) in THF (17 mL) was slowly added to 6 (5.36 g, 14.8 mmol) at 0 °C. After the mixture was stirred at room temperature for 2 h, the reaction was quenched with H₂O. The product was extracted with Et₂O, and the organic layer was washed with H₂O and dried over MgSO₄. Purification by silica gel column chromatography [hexane/CHCl₃/EtOAc (5:4:1)] gave 7 as a pale yellow solid (4.09 g, 9.27 mmol, 66% yield). mp 42–48 °C; ¹H NMR (CDCl₃, 400 MHz): δ 3.92 (s, 3H), 4.43 (s, 2H), 5.05 (s, 2H), 5.20 (s, 2H), 6.77 (d, J = 8.9 Hz, 1H), 7.32–7.45 (m, 10H), 7.56 (d, J = 8.9 Hz, 1H); ¹³C NMR (CDCl₃, 100 MHz): δ 36.9, 56.3, 75.7, 76.8, 107.8, 123.6, 126.9, 128.4, 128.5, 128.68, 128.74, 128.8, 136.4, 137.0, 140.8, 152.9, 158.5, 191.5; IR (KBr) 3009, 2961, 2941, 2887, 1678, 1585, 1439, 1369, 1292, 1206, 1096, 972, 760, 700; HRMS (FAB) calcd for C₂₃H₂₂⁷⁹BrO₄, 441.0701; found, 441.0702 [M + H]⁺.

4-Benzyloxy-2-{2-[2,3-bis(benzyloxy)-4-methoxyphenyl]-2-oxoethoxy}benzaldehyde (9)

A mixture of 7 (1.28 g, 2.90 mmol), 8 (855 mg, 2.99 mmol), and K₂CO₃ (415 mg, 3.00 mmol) in acetone (10 mL) was heated at reflux for 24 h. The mixture was cooled to room temperature, and H₂O (20 mL) was added. The product was extracted with EtOAc (30 mL × 2). The combined organic layers were washed with H₂O, dried over Na₂SO₄, filtered, and evaporated to give a crude product. A solution of the crude product in CH₂Cl₂ and MeOH (1:4, 10 mL) was cooled to 0 °C. HCl (1 N, 1 mL) was added, and the solution was stirred at 0 °C for 1 h. The product, which precipitated, was collected by filtration. The solid was recrystallized from CHCl₃/MeOH to yield 9 as a white solid (1.06 g, 1.80 mmol, 62% yield). mp 139–140 °C; ¹H NMR (CDCl₃, 400 MHz): δ 3.94 (s, 3H), 4.99 (s, 2H), 5.06 (s, 2H), 5.07 (s, 2H), 5.25 (s, 2H), 6.07 (d, J = 2.0 Hz, 1H), 6.61 (dd, J = 1.8, 8.7 Hz, 1H), 6.81 (d, J = 9.0 Hz, 1H), 7.20–7.46 (m, 15H), 7.69 (d, J = 9.0 Hz, 1H), 7.82 (d, J = 8.7 Hz, 1H), 10.37 (s, 1H); ¹³C NMR (CDCl₃, 100 MHz): δ 56.4, 70.5, 73.9, 75.8, 76.7, 100.2, 106.9, 107.9, 119.6, 123.1, 126.5, 127.9, 128.45, 128.54, 128.57, 128.59, 128.7, 128.82, 128.84, 128.9, 130.5, 136.0, 136.4, 137.0, 140.6, 153.5, 158.9, 162.5, 165.1, 188.7, 193.1; IR (KBr) 3034, 2880, 2778, 1692, 1609, 1587, 1441, 1294, 1254, 1186, 1103; HRMS (FAB) calcd for C₃₇H₃₃O₇, 589.2226; found, 589.2225 [M + H]⁺.

(S)-(−)-7-Benzyloxy-3-[2,3-bis(benzyloxy)-4-methoxyphenyl]-3-hydroxychroman-4-one ((S)-(−)-10)

A mixture of NHC precatalyst B (4.67 mg, 0.01 mmol, 2.0 mol %) and Cs₂CO₃ (3.14 mg, 0.010 mmol, 2.0 mol %) in dry CHCl₃ (600 μL) under N₂ was stirred in a round-bottom flask in a thermostatic bath at 30 °C for 1 h. Aldehyde 9 (294 mg, 0.50 mmol) was added, and the mixture was gently stirred under N₂ in a thermostatic bath at 40 °C for 24 h. Purification by silica gel column chromatography [hexane/CHCl₃/EtOAc (5:4:1)] afforded (S)-(−)-10 as a yellow solid (290 mg, 0.493 mmol, 99% yield). mp 39–45 °C; [α]_D²⁵ – 32.7 (c 0.1, CHCl₃), 86% ee; ¹H NMR (CDCl₃, 400 MHz, 40 °C): δ 3.80 (s, 1H), 3.83 (s, 3H), 4.24 (d, J = 11.8 Hz, 1H), 4.96 (d, J = 11.6 Hz, 1H), 4.99 (d, J = 10.6 Hz, 1H), 4.99 (s, 2H), 5.08 (s, 2H), 5.22 (d, J = 10.6 Hz, 1H), 6.44 (d, J = 2.2 Hz, 1H), 6.61 (dd, J = 2.3, 8.8 Hz, 1H), 6.65 (d, J = 8.8 Hz, 1H), 7.14 (d, J = 8.8 Hz, 1H), 7.21–7.41 (m, 15H), 7.75 (d, J = 8.9 Hz, 1H); ¹³C NMR (CDCl₃, 100 MHz): δ 56.1, 70.4, 74.3, 74.4, 75.3, 75.4, 101.8, 107.3, 111.1, 113.8, 122.6, 125.0, 127.6, 127.8, 128.1, 128.29, 128.34, 128.4, 128.5, 128.7, 128.9, 129.8, 136.0, 137.3, 137.5, 141.6, 151.2, 154.7, 163.2, 165.2, 191.7; IR (KBr) 3418, 3030, 2940, 1684, 1609, 1576, 1497, 1456, 1439, 1248, 1171, 1092; HRMS (FAB) calcd for C₃₇H₃₃O₇, 589.2226; found, 589.2224 [M + H]⁺; HPLC: Chiralpak AD-H, hexane/i-PrOH = 3:2, 1.0 mL/min, 254 nm, (S) 40.3 min, (R) 46.0 min.

(R)-(+)-10

(R)-(+)-10 was prepared from 9 using the opposite enantiomer of NHC precatalyst B, ent-B (432 mg, 0.734 mmol, 73% yield). [α]_D¹⁷ + 31.0 (c 0.1, CHCl₃), 82% ee.

6-Benzyloxy-2-[2,3-bis(benzyloxy)-4-methoxybenzoyl]-1-benzofuran (10′)

78 mg (0.137 mmol, 27% yield, entry 3 in Table 5); pale yellow solid; mp 107–109 °C; ¹H NMR (CDCl₃, 400 MHz): δ 3.93 (s, 3H), 5.09 (s, 2H), 5.11 (s, 2H), 5.14 (s, 2H), 6.76 (d, J = 8.6 Hz, 1H), 7.01 (dd, J = 2.1, 8.7 Hz, 1H), 7.08–7.21 (m, 7H), 7.29–7.51 (m, 12H); ¹³C NMR (CDCl₃, 100 MHz): δ 56.3, 70.6, 75.6, 76.6, 97.1, 107.1, 115.0, 117.5, 120.8, 123.8, 125.3, 126.7, 127.6, 128.0, 128.2, 128.3, 128.4, 128.7, 128.76, 128.80, 136.4, 136.9, 137.4, 141.8, 151.7, 152.6, 156.7, 157.6, 160.2, 182.9; IR (KBr) 3107, 3090, 3065, 3030, 3009, 2961, 2940, 2882, 2837, 1636, 1620, 1591, 1543, 1491, 1456, 1443, 1364, 1335, 1302, 1273, 1252, 1225, 1171, 1161, 1115, 1099, 978, 814, 758, 743, 735, 694; HRMS (FAB) calcd for C₃₇H₃₁O₆, 571.2121; found, 571.2121 [M + H]⁺.

(−)-Isodarparvinol B ((S)-(−)-5)

To a solution of (−)-10 (81 mg, 0.14 mmol) in EtOH (2.8 mL) at room temperature was added 10% Pd/C (7 mg). The mixture was stirred under H₂ (1 atm, balloon) at room temperature for 4.5 h. The solution was passed through a Celite pad and washed with EtOAc. Purification by silica gel column chromatography [hexane/EtOAc (1:2)] gave (S)-(−)-5 as an off-white solid (40 mg, 0.13 mmol, 93% yield). mp 125–129 °C; [α]_D²⁰ – 13.3 (c 0.1, MeOH), 90% ee, lit.^15,17 [α]_D²⁵ – 407.8 (c 0.7, MeOH) for the natural product; ¹H NMR (CD₃OD, 400 MHz): δ 3.83 (s, 3H), 4.18 (d, J = 11.9 Hz, 1H), 4.90 (d, J = 11.9 Hz, 1H), 6.32 (d, J = 2.2 Hz, 1H), 6.50 (d, J = 8.8 Hz, 1H), 6.51 (dd, J = 2.3, 8.8 Hz, 1H), 6.94 (d, J = 8.8 Hz, 1H), 7.78 (d, J = 8.7 Hz, 1H); ¹³C NMR (CD₃OD, 100 MHz): δ 56.5, 75.0, 75.7, 103.5, 103.8, 112.1, 113.9, 118.2, 119.7, 131.0, 135.3, 144.3, 149.8, 164.8, 166.5, 192.1; IR (KBr) 3404, 2938, 2839, 1670, 1609, 1510, 1474, 1339, 1290, 1252, 1169, 1090, 1034, 945, 901, 853, 820, 775, 689; HRMS (FAB) calcd for C₁₆H₁₅O₇, 319.0818; found, 319.0819 [M + H]⁺; HPLC: Chiralpak AD-H, hexane/i-PrOH = 2:1, 0.5 mL/min, 254 nm, (S) 21.6 min, (R) 26.9 min.

(+)-Isodarparvinol B ((R)-(+)-5)

(R)-(+)-5 was prepared from (+)-10 (44 mg, 0.14 mmol, 82% yield). [α]_D¹⁶ + 13.0 (c 0.1, MeOH), 93% ee.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.0c01141.

• Melting-point depression experiments, CD and NMR spectra, and DFT calculations (PDF)

The authors declare no competing financial interest.