Quinidine Thiourea-Catalyzed Aldol Reaction of Unactivated Ketones: Highly Enantioselective Synthesis of 3-Alkyl-3-hydroxyindolin-2-ones

Aldol reaction is one of the most important carbon-carbon bond formation reactions¹ and, therefore, many asymmetric variants of this reaction have been developed in the past.¹ Since List and Barbas discovered the proline-catalyzed cross aldol reaction of ketones and aldehydes,² many amine derivatives, mainly proline derivatives, have been developed for the asymmetric cross aldol reactions.³ Mechanistically, these catalysts activate the ketones or enolizable aldehydes through the formation of an enamine intermediate (Scheme 1, upper equation).³ The mechanism is usually a combination of both covalent and noncovalent catalyses since most of these catalysts also contain a hydrogen bonding moiety to direct the approach of the enamine acceptor. In contrast, although enol and enolate (Scheme 1, lower equation) are the active intermediates in the original aldol reactions,⁴ organocatalyzed enantioselective direct aldol reaction of unactivated ketones via the enolate mechanism is very difficult, because the acidity of the α-proton in these ketones is very low. To our knowledge, there has been no such report.^5,6,7 Nevertheless, the enolate mechanism does have certain advantages, especially when the formation of an enamine is difficult. During our study of using activated ketone compounds as the enamine acceptors for the organocatalyzed aldol reactions,⁸ we envisioned such an organocatalyzed enolate-mediated aldol reaction should be possible if the enolate acceptor is sufficiently activated, because the equilibrium favors the formation of the product with such a substrate. Herein we wish to report the first enantioselective aldol reaction of unactivated ketones involving the enolate mechanism, which may be used for the high enantioselective synthesis of 3-alkyl-3-hydroxyindolin-2-ones.⁹

Scheme 1.

Aldol Reactions Involving the Enamine or the Enolate Intermediates

The enamine-mediated aldol reaction of isatins and ketones or enolizable aldehydes have been used in recent years for the synthesis of 3-alkyl-3-hydroxyindolin-2-ones,¹⁰ which are important biologically active natural products and medicinal compounds.¹¹ Since isatin contains an activated ketone group, we tested our hypothesis with isatin (10a) and acetone (11a) as the model substrates, and some tertiary amines (1–8) as the catalysts (Figure 1). The results of the catalyst screening and optimizations are presented in Table 1.

Figure 1.

Catalysts Screened for the Cross Aldol Reaction of Isatin and Acetone [Ar = 3,5-(CF₃)₂C₆H₃]

Table 1.

Catalyst Screening and Reaction Condition Optimizations^[a]

Entry	Catalyst	Solvent	10a (equiv)	Time (d)	Yield (%)[b]	ee (%)[c]
1	1	acetone	---	1	78	---
2	2	acetone	---	1	84	2
3	3	acetone	---	1	80	61
4	4	acetone	---	1	98	68
5	5	acetone	---	1	99	57[d]
6	6	acetone	---	1	98	57[d]
7	7	acetone	---	1	trace	---
8	8	acetone	---	1	98	40
9	9	acetone	---	1	0	---
10	4	benzene	7	4	56	56
11	4	CH2Cl2	7	4	46	53
12	4	THF	7	4	58	82
13	4	Et2O	7	4	60	76
14	4	dioxane	7	4	64	70
15	4	CH3CN	7	4	22	67
16	4	AcOEt	7	4	trace	---
17	4	DMSO	7	4	trace	---
18	4	hexane	7	4	trace	---
19	4	DME	7	4	15	77
20	4	MeOH	7	4	49	31
21	4	THF	30	4	98	80
22	4	THF	30	4	23	86[e]
23	4	THF	70	4	73	86[e]
24	4	THF	70	6	97	85[e]

^[a]Unless otherwise indicated, all reactions were carried out with isatin (0.10 mmol), acetone, and the catalyst (0.01 mmol, 10 mol %) in the specified solvent (1.0 mL for acetone; 2.0 mL for THF and other solvents) at rt.

^[b]Yield of the isolated product after column chromatography.

^[c]Determined by HPLC analysis. Absolute configuration was assigned according to ref. 10c.

^[d]The S enantiomer was obtained.

^[e]Carried out at 5 °C.

As shown in Table 1, with acetone as the solvent, the reaction catalyzed by quinuclidine (1) yielded a good yield of the desired 12a (entry 1). Quinidine (2) is similarly reactive, but the ee value obtained was poor (entry 2). In contrast, quinidine-derived thioureas¹² 3 and 4 gave much improved ee values for the R-enantiomer (61% and 64%, respectively, entries 3–4). As expected, quinine thiourea (5) and cinchonidine thiourea (6) led to the formation of the S-enantiomer (57% ee, entries 5–6). High yields of 12a were also obtained for these four catalysts. Nonetheless, with thiourea 7, only trace amount of 12a was formed (entry 7). This was probably due to the lower basicity of 7. Takemoto thiourea 8 is more reactive, but the ee value obtained was lower (entry 8). However, a simple thiourea 9 that has no basic moiety does not catalyze the reaction (entry 9). This result excludes the possibility of the involvement of an enol intermediate in this reaction. Since the reaction is catalyzed by 1 (entry 1) and not by 9 (entry 9), it most likely works through the enolate mechanism. If this is the case, the reaction must be a complete noncovalent catalysis. With the best catalyst 4, further optimizations (entries 10–20) identified THF as the best solvent for this reaction, in which the ee value was improved to 81% (entry 12). All the other solvents screened led to less satisfactory results. The product yield could be improved to 98% by using more acetone (30 equiv., entry 21). The ee value was improved to 86% when the reaction was carried out at 5 °C, at the cost of the reactivity (entry 22).¹³ Luckily, the lost reactivity could be remedied by employing more acetone (70 equiv.) and prolonging the reaction time to 6 days, without affecting the enantioselectivity (entries 23–24).

Next the scope of this reaction was studied and the results are summarized in Tables 2 and 3. As shown in Table 2, with acetone as the substrate, various substituted isatins may be applied in this reaction. Excellent yields and high ee values (≥79% ee) of the expected products were obtained (entries 1–6). The electronic nature of the substituent and its position on the isatin phenyl ring show some influences on the enantioselectivities. For example, the ee value of 12c of 4-bromoisatin (entry 3) is much higher than that of 12e of 6-bromoisatin (entry 5). Acetophenone (11b) is a tough substrate for the enamine mechanism due to its low electrophilicity¹⁴ and has never been used in the asymmetric aldol reaction with isatin. However, as is evident from the results in Table 2, 11b is even more reactive than acetone under our conditions (only 10 equiv. ketone is necessary), and the ee values obtained for the products are also higher as compared to those of the corresponding acetone products (entries 7–12). The observed higher reactivity is in accord with the increased acidity of the acetophenone α-proton. Other aryl methyl ketones, such as acetonaphthones 11c and 11d, also give very good results (entries 13–14). Excellent results were also obtained for α,β-unsaturated ketones (E)-3-penten-2-one (11e) and (E)-4-phenyl-3-buten-2-one (11f) (entries 15–16). Beside ketones, acetaldehyde (11g) may also be applied in this reaction. Since the corresponding aldol product 12q is not very stable, it was reduced in situ with NaBH₄ to give the corresponding diol in 82% yield. The ee value of this diol was determined to be 73% (entry 17).

Table 2.

Cross Aldol Reaction of Isatins and Ketones^[a]

Entry	R1/10	R2/11	Time (d)	12/Yield (%)[b]	ee (%)[c]
1	H/a	Me/a	6	a/97	85
2	4-Cl/b	Me/a	4	b/98	87
3	4-Br/c	Me/a	4	c/99	91
4	5-F/d	Me/a	5	d/99	85
5	6-Br/e	Me/a	4	e/99	79
6	5,7-Br2/f	Me/a	5	f/99	83
7	H/a	Ph/b	6	g/75	90
8	4-Cl/b	Ph/b	5	h/98	95
9	4-Br/c	Ph/b	5	i/98	97
10	5-F/d	Ph/b	4	j/99	86
11	6-Br/e	Ph/b	5	k/98	84
12	4,7-Cl2/g	Ph/b	5	l/98	92
13	H/a	1-Np/c	6	m/70	86
14	H/a	2-Np/d	5	n/98	90
15	H/a	/e	6	o/76	87
16	H/a	/f	4	p/98	91
17	H/a	H/g	3	q/82[d]	73[e]

^[a]All reactions were carried out with 9 (0.10 mmol), the ketone 11 (7.0 mmol for acetone, and 1.0 mmol for acetophenone), and catalyst 4 (0.01 mmol, 10 mol %) in THF (2.0 mL) at 5 °C.

^[b]Yield of isolated product after column chromatography.

^[c]Determined by HPLC analyses.

^[d]Yield of the corresponding diol after reduction of the primary product with NaBH₄.

^[e]Determined by HPLC analysis of the corresponding diol.

Table 3.

Cross Aldol Reaction of Activated Carbonyl Compounds and Ketones^[a]

Entry	13	11	14/Yield (%)[b]	ee (%)[c]
1	/a	b	a/60	90
2	/b	a	b/90	61
3		b	c/78	76
4	/c	b[d]	d/41	74
5	PhCOCHO·H2O /d	a[e]	e/46[f]	56
6	PhCOCHO·H2O /d	/h	f/54[g,h]	93[i]

^[a]Unless otherwise indicated, all reactions were carried out with 13 (0.10 mmol), 14 (7.0 mmol for 11a, and 1.0 mmol for 11b), and catalyst 4 (0.01 mmol, 10 mol %) in THF (2.0 mL) at 5 °C for 6 days.

^[b]Yield of isolated product after column chromatography.

^[c]Determined by HPLC analyses.

^[d]The reaction was carried out at rt.

^[e]The reaction was carried out for 4 d with THF (0.2 mL) as the solvent.

^[f]The product of this reaction is 2-hydroxy-1-phenylpentane-1,4-dione (14e).

^[g]The reaction was carried out in cyclohexanone (11h, 1.0 mmol) as the solvent for 2 days; the product of this reaction is 2-(1-hydroxy-2-oxo-2-phenylethyl)cyclohexanone (14f).

^[h]Total yield of two inseparable diastereomers; diastereomeric ratio (anti/syn) was determined to be 86:14 according the ¹H NMR analysis of the crude product. The relative stereochemistry of these products was assigned according to ref. 15.

^[i]Value of the major anti diastereomer.

In addition to isatins, other activated carbonyl compounds may also be used as the substrates in this reaction. As is evident from the results collected in Table 3, the aldol reactions of 7-azaisatin (13a) with acetophenone (11b) gives the expected 14a in 60% yield and 90% ee (entry 1). These results are comparable to those of isatin (Table 2, entry 7). Similarly, the aldol reaction of 1-methyl-7-azaisatin (13b) with acetone (11a) and acetophenone (11b) yield the desired products 14b and 14c in 90% and 78% yields, respectively. The ee values of these two products were determined to be 61% and 76% ee, respectively (entries 1 and 2). The lower ee values obtained with 13b (entry 3) as compared to 13a (entry 1) was most likely due to the steric effects of the N-methyl group in 13b, which may be easily rationalized with our proposed mechanism in Scheme 2 (see below). Other active carbonyl derivatives, such as, 4,4-dimethyldihydrofuran-2,3-dione (13c) and phenylglyoxal hydrate (13d), may also be applied in this reaction. The reaction of 13c with 11b leads to the desired product 14d in 41% yield and 74% ee (entry 4). Compound 13d has both a ketone and an aldehyde group; nonetheless, only the aldehyde group reacts in the cross aldol reactions with acetone (11a) and cyclohexanone (11h). With acetone (11a) as the substrate, the product 2-hydroxy-1-phenylpentane-1,4-dione (14e) was obtained in 46% yield and 56% ee (entry 5). When cyclohexanone (11h) was used as the substrate, the reaction gives 14f as a mixture of anti and syn diastereomers¹⁵ in 54% yield, with an anti/syn ratio of 86:14. The major anti diastereomer was obtained in a high ee value of 93% (entry 6). To our knowledge, this is first example of an enantioselective cross aldol of phenylglyoxal hydrate that yields the anti product as the major product.¹⁶

Scheme 2.

Proposed Transition States for the Aldol Reaction of Isatin and Acetone

A plausible mechanism of this reaction is proposed to account for the observed enantioselectivity in the isatin-acetone cross aldol reaction. As shown in Scheme 2, acetone is deprotonated by the tertiary amine in the quinidine thiourea catalyst backbone. After deprotonation, the enolate associates closely with the catalyst due to ionic interactions. On the other hand, two hydrogen bonds are formed between the isatin carbonyl groups and the thiourea moiety of the quinidine thiourea catalyst. Besides activating the ketone group for the enolate attack, these hydrogen bonds also direct the approach of isatin. Among the two possible orientations of isatin, the re face orientation (Scheme 2, top left structure) is favored since the unfavorable interaction between the isatin benzene ring and the enolate is avoided. Attacking of the enolate to the si face of the isatin ketone group leads to the observed major R enantiomer. These proposed transition states predict that an increase of the steric bulkiness at position 1 of isatin will lead to lower enantioselectivity, while an increase of the steric bulkiness at position 4 will lead to higher enantioselectivity. As shown by the results (Table 3, entry 3 vs. entry 1; Table 2, entries 2 and 3 vs. entry 1), this is indeed the case, which lends some support for this mechanism.

To show the utility of the current method, we developed the first enantioselective synthesis of 3-hydroxyindolin-2-one 12r, which is the lead compound for treating Ewing’s sarcoma discovered most recently by Toretsky and co-workers.¹⁷ As shown in Eq. 1, when isatin 10g and 4′-methoxyacetophenone (11i, 20 equiv.) were allowed to react in the presence of catalyst 4 (20 mol %) under the optimized conditions, the desired product 12r was obtained in 85% yield and 91% ee.¹⁸

(1)

In summary, we have developed the first organocatalyzed asymmetric aldol reaction of unactivated ketones and activated carbonyl compounds on the basis of a new noncovalent catalysis mechanism involving the enolate intermediate by using quinidine thiourea as the catalyst. The method is complementary to the enamine-based organocatalyzed aldol reactions and works well where the enamine mechanism has difficulties. The reaction may be utilized for the highly enantioselective synthesis of biologically significant 3-alkyl-3-hydroxyindolin-2-ones.

Experimental Section

General procedure for the aldol reaction of isatins with ketones

The organocatalyst 4 (5.9 mg, 0.01 mmol, 10 mol %) and isatin 9 (0.10 mmol) were stirred in THF (2.0 mL) for 10 min at 5°C. The corresponding ketone 10 (7.0 mmol for acetone or 1.0 mmol for other methyl ketones) was added and the mixture was stirred at 5°C for the time as given in the tables. The solvent was then removed under reduced pressure and the mixture was purified by flash column chromatography on silica gel (1:1 hexane/ethyl acetate) to give the desired aldol product.