Development of Novel Catalytic Direct Syntheses of N-Unsubstituted Ketimines and Their Applications to One-Pot Reactions

Abstract

Herein, we summarize our recent efforts toward developing catalytic methods for the synthesis of N-unsubstituted ketimines and their applications to one-pot reactions for producing various nitrogen-containing compounds. This account provides detailed background, optimization, scope, and mechanistic information. We hope this work will stimulate future studies on the implementation of N-unsubstituted ketimines.

1 Introduction

2 Scandium-Catalyzed Synthesis of N-Unsubstituted Ketimines

3 Tetra-n-butylammonium Fluoride Catalyzed Synthesis of N-Unsubstituted Ketimines

4 Conclusion

Biographical Sketches

Takashi Ohshima received his Bachelor’s degree from The University of Tokyo in 1991 under the direction of Professor Masaji Ohno and his Ph.D. from The University of Tokyo in 1996 under the direction of Professor Masakatsu Shibasaki. In the following year, he joined Otsuka Pharmaceutical Co., Ltd. After two years as a Postdoctoral Fellow at The Scripps Research Institute with Professor K. C. Nicolaou (1997–1999), he returned to Japan as an Assistant Professor at The University of Tokyo. He was appointed as an Associate Professor at Osaka University in 2005. In 2010, he was promoted to Full Professor at Kyushu University. Since 2021, he has also served as Principal Investigator of the Grant-in-Aid for Transformative Research Areas ‘Digitalization-Driven Transformative Organic Synthesis (Digi-TOS)’.

Hiroyuki Morimoto received his Bachelor’s degree in 2004 and his Ph.D. in 2009 from the University of Tokyo under the supervision of Professor Masakatsu Shibasaki. After working as a Postdoctoral Fellow at the University of Illinois at Urbana-Champaign with Professor John F. Hartwig (2009–2010), he returned to Japan and joined Professor Ohshima’s group at Kyushu University as an Assistant Professor and was promoted to lecturer in 2017. In 2023, he moved to Kyushu Institute of Technology as an Associate Professor. His research interests include the development of atom-economical and environmentally benign catalytic reactions.

Yuta Kondo received his bachelor’s degree in 2017 and his Ph.D. in 2023 from Kyushu University under the supervision of Professor Takashi Ohshima. His research interests include the development of new reactions with N-unprotected ketimines. He is currently a pharmacist at Kyushu University Hospital.

Introduction

N-Unsubstituted ketimines are valuable compounds as synthetic intermediates for the production of nitrogen-containing compounds, and their use in various reactions has been actively studied in recent years.[2] [3] [4] [5] [6] Although several useful reactions using N-unsubstituted ketimines have been developed, the synthesis of N-unsubstituted ketimines is rarely discussed and methods have remained undeveloped (Scheme [1]). Conventional methods for synthesizing N-unsubstituted ketimines require stoichiometric amounts of organometallic reagents and harsh reaction conditions, leading to a narrow substrate scope and poor environmental profile.[7–17] To develop the field of N-unsubstituted ketimines, more efficient methods for supplying N-unsubstituted ketimines must be established. With this background, we aimed to develop new reactions for synthesizing N-unsubstituted ketimines under more moderate conditions. This account summarizes our recent work on the catalytic syntheses of N-unsubstituted ketimines.[18] [19] [20]

Scandium-Catalyzed Synthesis of N-Unsubstituted Ketimines

We selected ketones as starting materials to develop a new synthetic method for N-unsubstituted ketimines. Ketones are stable and commercially available compounds. Generally, because N-unsubstituted ketimines are thermodynamically more unstable than the corresponding ketones (Scheme [2]), converting ketones into N-unsubstituted ketimines requires excess reagents and harsh reaction conditions such as high temperature and pressure.

The reaction should ideally be thermodynamically favorable for synthesizing N-unsubstituted ketimines under milder reaction conditions. We focused on the Noyori method, known for carbonyl group protection (Scheme [3]).[21] The Noyori method protects the carbonyl group as a ketal structure. Acetalization is generally reversible, and an excess of reactant is required to push the chemical equilibrium toward the product. The Noyori method overcomes this problem by taking advantage of the stability and low reactivity of the co-product. The reaction of a silyl ether with a carbonyl compound in the presence of a catalytic amount of trimethylsilyl triflate realizes both protection of the carbonyl group and the formation of hexamethyldisiloxane as a co-product. Because hexamethyldisiloxane is very stable and relatively unreactive, the reverse reaction does not occur and the desired acetalization proceeds irreversibly. We hypothesized that Noyori’s strategy of pushing the equilibrium toward the product through low-reactive co-products could be applied to the synthesis of N-unsubstituted ketimines. In other words, if ketones react with bis(trimethylsilyl)amines to form stable and low-reactive hexamethyldisiloxane with N-unsubstituted ketimines, the reaction may be thermodynamically favorable.

Optimization

We selected benzophenone (1a) as a substrate for optimization (Table [1]). First, stirring 1a and bis(trimethylsilyl)amine (2) under heating without a catalyst did not give the desired benzophenone imine (3a); only raw material was recovered. Next, we tested trimethylsilyl triflate, used as a catalyst in the Noyori method, and the desired 3a was obtained. Increasing the reaction temperature and catalyst loading gave 3a almost quantitatively.

Table 1 Initial Optimization of Reaction Conditions

Entry	TMSOTf (mol%)	Temp (°C)	Time (h)	3a (%)a
1	10	70	24	17
2	10	90	24	49
3	10	90	48	92
4	20	90	24	97

^a Determined by ¹H NMR analysis of the crude mixture.

The above studies were conducted under solvent-free conditions, so we next attempted to apply solvents to a broader range of ketones because the nitrogen source 2 and co-product hexamethyldisiloxane are both low-polarity liquids in which 1 has very low solubility. Dilution with solvent drastically reduced the reaction rate, however, and the reaction was not completed even at a 2.0 M concentration (Table [2]).

Table 2 Effects of Solvents

Entry	Solvent (M)	Time (h)	3a (%)a
1	neat	24	97
2	PhCl (0.50)	24	trace
3	PhCl (2.0)	24	59
4	PhCl (2.0)	48	76
5	toluene (2.0)	48	43

^a Determined by ¹H NMR analysis of the crude mixture.

Table 3 Optimization of Lewis Acid Catalyst

Entry	Catalyst	3a (%)a	Entry	Catalyst	3a (%)a
1	none	n.r.	16	Yb(OTf)3	88
2	TMSOTf	20	17	Bi(OTf)3	>99
3	Sc(OTf)3	>99	18	TfOH	11
4	Fe(OTf)2	n.r.	19	TfOHb	14
5	Fe(OTf)3	84	20	Sc(NO3)3	37
6	Cu(OTf)2	n.r.	21	Sc(NTf2)3	59
7	Zn(OTf)2	n.r.	22	Sc(ONf)3	98
8	Y(OTf)3	85	23	Sc(OAc)3	n.r.
9	In(OTf)3	44	24	Sc2O3	n.r.
10	Sn(OTf)2	>99	25	YF3	n.r.
11	La(OTf)3	n.r.	26	YCl3	n.r.
12	Sm(OTf)3	65	27	BiCl3	16
13	Eu(OTf)3	94	28	BiBr3	24
14	Gd(OTf)3	52	29c	Sc(OTf)3	>99
15	Er(OTf)3	68	30c	Bi(OTf)3	96

^a Determined by ¹H NMR analysis of the crude mixture.

^b 15 mol%.

^c For 2 h.

Therefore, metal Lewis acids were investigated as more active Lewis acid catalysts (Table [3]). We investigated metal triflates as metal Lewis acids and found that scandium triflate was much more effective than trimethylsilyl trifluoromethanesulfonate. The reaction did not proceed with iron(II) triflate, while iron(III) triflate gave good results. Copper(II) triflate and zinc(II) triflate, frequently used as Lewis acid catalysts, did not provide the desired product. Yttrium is a rare earth element like scandium, and its triflate salt also gave good results. Indium(III) and tin(II) also exhibited good catalytic activity, especially tin(II). We then examined lanthanoids, which are also commonly used as Lewis acids.[22] [23] [24] While no reaction proceeded with lanthanum(III) triflate, the reaction proceeded with other lanthanides. Bismuth(III) triflate also had high catalytic activity, giving N-unsubstituted ketimine quantitatively. Trifluoromethanesulfonic acid, which may be generated from metal triflates,[25] was also tested as a catalyst, but the desired N-unsubstituted ketimine 3a was obtained in only low yield. A control experiment confirmed that metal Lewis acids functioned as catalytically active species in this system. Next, we investigated metal salts other than triflate. Scandium(III) nitrate, scandium(III) trifluoromethanesulfonimide (Sc(NTf₂)₃), and scandium(III) nonaflate (Sc(ONf)₃) gave the desired product, but all were less active than scandium(III) triflate. Neither scandium(III) acetate nor scandium(III) oxide exhibited any catalytic activity. Metal halides of yttrium and bismuth were also tested, but the reaction did not proceed at all with yttrium salts, and only low yields were obtained with bismuth salts. In the above screening, scandium(III) triflate, tin(II) triflate, and bismuth(III) triflate showed the best results. We excluded tin due to its toxicity. Scandium(III) triflate and bismuth(III) triflate were compared by shortening the reaction time to 2 h, and scandium(III) triflate was found to be more active. Thus, we selected scandium(III) triflate as the optimal catalyst.

Next, we optimized the reaction solvent for the Sc(OTf)₃ catalyst (Table [4]). Aromatic solvents such as chlorobenzene, toluene, and fluorobenzene gave the desired product in high yields, with chlorobenzene affording the best result. Ethereal solvents such as 1,4-dioxane and tetrahydrofuran (THF) were also available. The halogenated solvent 1,2-dichloroethane (DCE) was promising, while acetonitrile gave a low yield. In the case of ethanol, only a tiny amount of the desired product was observed because the ethanol solvent reacted with the bis(trimethylsilyl)amine, thereby preventing the reaction. The desired reaction did not proceed when N,N-dimethylformamide (DMF) or dimethyl sulfoxide (DMSO) were used as solvents, presumably because DMF and DMSO coordinated to the scandium catalyst, which deactivated its catalytic ability. Based on these results, we selected chlorobenzene as the optimal solvent.

Table 4 Optimization of Solvent for Sc(OTf)₃ Catalyst

Entry	Solvent	3a (%)a
1	PhCl	>99 (94)b
2	toluene	95
3	PhF	78
4	1,4-dioxane	95
5	THF	83
6	DCE	95
7	MeCN	33
8	EtOH	6
9	DMF	n.r.
10	DMSO	n.r.

^a Determined by ¹H NMR analysis of the crude mixture.

^b Isolated yield.

Benzophenone imine was somewhat unstable on silica gel and partially hydrolyzed during silica gel column chromatography. We investigated various purification methods and found that hydrolysis could be suppressed by using hexane/triethylamine (9:1) as eluent. This discovery allowed us to isolate the desired N-unsubstituted ketimine in 94% yield.

Substrate Scope

With the optimum conditions in hand, the ketone scope was examined (Scheme [4]). Both electron-withdrawing and electron-donating groups were acceptable for this reaction, giving the corresponding N-unsubstituted ketimines in high yields. In particular, N-unsubstituted ketimines with halogen groups, which are difficult to synthesize by the conventional method using Grignard’s reagent, were efficiently supplied by this method. Ketimine 3k, with an azide group, could also be synthesized, whereas its synthesis using the previous method with azides as the starting materials is difficult.[17] Pyridine and thiophene were also tolerated in this reaction. Phenolic hydroxyl groups of 2-hydroxybenzophenone were not silylated by bis(trimethylsilyl)amine and converted into the corresponding unsubstituted ketimines. The method was also applied to pharmaceuticals such as ketoprofen and fenofibrate, and the corresponding N-unsubstituted ketimines were obtained in high yield. Ketoprofen and fenofibrate have carboxy and ester structures, respectively, and these functional groups are not tolerated under the conditions of conventional methods using strongly basic organometallic reagents. Tricyclic ketones such as fluorenone, dibenzosuberenone, and xanthone could also be applied with this method. N-Unsubstituted ketimines derived from isatin, which are suitable substrates for nucleophilic addition reactions,[6] [26] were efficiently supplied by this method. The reaction applied not only to diaryl ketones but also to alkyl-substituted ketones. Some N-unsubstituted ketimines derived from alkyl-substituted ketones were purified as hydrochloride salts because they were unstable on silica gel, and hydrolysis could not be suppressed even by a triethylamine-containing eluent. We also tried to apply the method to α-iminoesters and α-iminoamides, which are useful as amino acid precursors,[27] but the reactions did not provide promising results.

Next, we attempted to scale the reaction (Scheme [5]). Unlike conventional methods, this method did not require highly reactive hazardous reagents and did not produce difficult-to-remove byproducts. In addition, this method did not require special reactors and could be performed using standard laboratory equipment. The operation was also straightforward – stirring the reagents with heating. Given these features, this method could be easily applied to large-scale synthesis. In a 10-mmol scale reaction, catalyst loading was successfully reduced to 0.25 mol% and the desired N-unsubstituted ketimine was synthesized on a gram scale (Scheme [5a]). Investigation of a 65-mmol scale reaction revealed that the catalyst loading could be reduced to 0.20 mol% while providing more than 10 g of the N-unsubstituted ketimine (Scheme [5b]). As this large-scale synthesis was performed under solvent-free conditions, waste derived from the reaction solvent was eliminated. The larger reaction scale allowed for purification of the product by distillation, which also reduced the amount of waste in the purification process. As a result, we achieved a very environmentally friendly reaction with less waste compared with the conventional method. The E-factor of this reaction was calculated to be 1.3, which is very small compared with the E-factors (5–50) of common fine chemical reactions.[28] [29] N-Unsubstituted ketimines derived from trifluoroacetophenone are relatively stable because the CF₃ group reduces the nucleophilicity on the nitrogen atom of ketimine. In addition, this unsubstituted ketimine is frequently used as a good electrophile activated by the CF₃ group.[6] [30] [31] Therefore, this N-unsubstituted ketimine is a valuable compound. Gram-scale synthesis of this ketimine was also achieved with 1.0 mol% Sc(OTf)₃ and fluorobenzene as the solvent (Scheme [5c]).

Application to One-Pot Synthesis

Having developed a novel catalytic approach, we next aimed to apply our synthetic method to a one-pot synthesis (Scheme [6]). In multi-step organic synthesis, purification is performed at each step to remove excess reagents and byproducts. On the other hand, in a one-pot reaction, the second-step reaction is performed continuously by adding the additional reagent without isolation after the first step. Because purification is performed only after the final step, the process is greatly simplified. In addition, minimizing the purification process, which generally requires large amounts of organic solvents, leads to environmentally friendly reactions.[32] N-Unsubstituted ketimines are generally more unstable than the corresponding ketones and N-protected ketimines and thus require careful handling, including purification. Therefore, a one-pot reaction that does not require the purification of intermediates is considered an ideal reaction for using N-unsubstituted ketimines. A disadvantage of one-pot reactions is that the reaction system becomes more complex with each reaction, and unexpected side reactions may occur. In our method, however, the only co-product, hexamethyldisiloxane, has very low reactivity, and no side reactions of concern occurred. We successfully applied one-pot decarboxylative Mannich-type reactions to isatin-derived nitrogen-free ketimines developed in our laboratory (Scheme [6a]).[26] We also achieved a one-pot Strecker reaction, a highly typical reactions of imines (Scheme [6b]). The product aminonitrile is an intermediate of a bioactive compound reported as an antagonist for the NMDA receptor.[33] We also developed one-pot reactions other than nucleophilic additions. Glycine Schiff base is an essential compound for synthesizing optically active amino acids and was synthesized in one pot on a gram scale (Scheme [6c]).[34] [35] Furthermore, enantioselective alkylation of glycine Schiff base was also performed in one-pot to obtain optically active amino acids in high yield with high enantioselectivity in a three-step, one-pot process from benzophenone (Scheme [6d]).[36] A one-pot synthesis of the fluorenylidene-protected/activated aminoalkane, which is reportedly a carbon pronucleophile,[37] was also achieved (Scheme [6e]). One-pot Buchwald–Hartwig amination gave the aniline derivatives (Scheme [6f]).[38] [39]

Mechanistic Studies

Several mechanistic experiments were performed to gain insight into the reaction mechanism of this catalytic reaction (Scheme [7]). First, ¹H NMR analysis of the crude mixture confirmed the formation of hexamethyldisiloxane as designed (Scheme [7a]). This experimental result confirmed that bis(trimethylsilyl)amine functions as both a nitrogen source and an oxygen scavenger. Second, adding desiccants to the standard conditions reduced yields in all cases (Scheme [7b]). All liquid reagents used were pre-dried with a desiccant, and hygroscopic scandium(III) triflate was dried using a heat gun under reduced pressure before use. Scandium(III) triflate was difficult to dry completely, however, and it is assumed that trace amounts of H₂O are present in the reaction system under standard conditions. The experimental results suggested that trace amounts of H₂O were actually crucial for the acceleration of the reaction.

As it was suggested that trace amounts of H₂O may be involved in the reaction, we performed additional experiments with H₂O. First, when one equivalent of H₂O was added to the standard conditions, the yield was significantly decreased (Scheme [8a]). Next, we evaluated the reaction of H₂O with bis(trimethylsilyl)amine. When H₂O and bis(trimethylsilyl)amine were stirred at 70 °C without a catalyst, the formation of small amounts of trimethylsilanol and hexamethyldisiloxane was observed. This reaction was accelerated by the addition of scandium(III) triflate, and most of the bis(trimethylsilyl)amine was converted into hexamethyldisiloxane (Scheme [8b]). Having found that the reaction of H₂O with bis(trimethylsilyl)amine generated trimethylsilanol, we next investigated the reaction of trimethylsilanol with bis(trimethylsilyl)amine. The reaction of silanol with bis(trimethylsilyl)amine gave hexamethyldisiloxane just by stirring under heat without any catalyst. The scandium catalyst accelerated this reaction (Scheme [8c]). These results suggested that bis(trimethylsilyl)amine reacted with H₂O and gave hexamethyldisiloxane via trimethylsilanol. Conversion of bis(trimethylsilyl)amine and H₂O into hexamethyldisiloxane indicates that ammonia is generated simultaneously. Therefore, we examined the application of ammonia gas instead of bis(trimethylsilyl)amine as a nitrogen source. The desired N-unsubstituted ketimine was obtained in low yield even with ammonia gas (Scheme [8d]).

In addition, we performed reaction-tracking experiments (Figure [1]). The reaction was too fast to follow at 90 °C and therefore we lowered the reaction temperature to 50 °C. At 50 °C, the reaction did not occur for more than 2 h, then gradually proceeded (Figure [1], exp 1). This observation suggested that the reaction had an induction period in the early stage. A similar experiment was then conducted by adding a catalytic amount of trimethylsilanol (Figure [1], exp 2). In exp 2, reaction progress was already observed after 1 h, suggesting that the silanol eliminated the induction period of the reaction. Finally, we performed the same reaction under an ammonia atmosphere (Figure [1], exp 3). Similar to exp 2, elimination of the induction period was confirmed in exp 3. These results suggested that the true nitrogen nucleophile was ammonia or ammonia-like compounds and that silanol contributed to generating the active species from bis(trimethylsilyl)amine. On the other hand, in exp 3, the reaction rate slowed down after 3 h, and the yield at 12 h was lower than that in the argon atmosphere (exp 1), probably due to deactivation of the scandium catalyst by the excess amount of ammonia. This result suggested that gradually generating nitrogen nucleophiles is effective.

Proposed Reaction Mechanism

The estimated reaction mechanism based on previous experimental results is shown in Scheme [9]. First, bis(trimethylsilyl)amines are bulky amines and nucleophilic addition to ketones is expected to be complicated. For example, lithium bis(trimethylsilyl)amide is frequently used as a non-nucleophilic base. Therefore, for bis(trimethylsilyl)amine to react with ketones, a more active nucleophile must be generated by removing the trimethylsilyl group. The above mechanistic experiments suggest that bis(trimethylsilyl)amine can react with trace amounts of H₂O and trimethylsilanol to generate trimethylsilylamine and ammonia. Therefore, we assumed that the true nitrogen nucleophile in this reaction is less bulky trimethylsilylamine or ammonia. In the initial stage of the reaction, the true nitrogen nucleophile generated by the small amount of H₂O present in the reaction system adds to the activated ketone intermediate I, resulting in hemiaminal intermediate II. Trimethylsilanol is eliminated from hemiaminal intermediate II to give intermediate III. A scandium catalyst is exchanged between intermediate III and the unreacted ketone, closing the catalytic cycle while yielding the desired N-unsubstituted ketimine. The observation of the induction period in the tracking experiments was attributed to the slow generation of the true nitrogen nucleophile at 50 °C. Adding trimethylsilanol may promote the generation of sufficient amounts of true nitrogen nucleophiles from the initial phase of the reaction to eliminate the induction period.

Tetra-n-butylammonium Fluoride Catalyzed Synthesis of N-Unsubstituted Ketimines

As noted above, the scandium method exhibited tolerance to a wide range of functional groups. It provided a variety of N-unsubstituted ketimines in high yields, especially for non-enolizable aryl ketones. On the other hand, alkyl-substituted ketones were limited to tertiary alkyl ketones. When cyclohexyl phenyl ketone, a secondary alkyl ketone, was applied to the scandium method, enamine isomers and enamine-derived byproducts were generated with the desired N-unsubstituted ketimines and the reaction mixture was complex (Scheme [10]). In general, N-unsubstituted ketimines derived from alkyl-substituted ketones are readily isomerized to enamines. In addition, N-unsubstituted ketimines derived from alkyl ketones are less resistant to hydrolysis than those derived from diaryl ketones. Therefore, developing an efficient synthesis of N-unsubstituted ketimines derived from enolizable alkyl ketones is more complicated. Thus, we aimed to establish a method for synthesizing alkyl ketone-derived N-unsubstituted ketimines to improve their usefulness and utilization value. With this background, we aimed to develop an improved method for synthesizing N-unsubstituted ketimines.

In control experiments for the scandium method, we found that the reaction proceeded well after adding alcohol, even when the reaction temperature was lowered to 50 °C (Scheme [11]). We assumed this was because bis(trimethylsilyl)amine, a bulky amine and less reactive nucleophile, was desilylated by alcohol, and a more active nitrogen source was generated. Therefore, we hypothesized that if Lewis base reagents could activate bis(trimethylsilyl)amine, the reaction would proceed more efficiently at lower temperatures. In particular, we speculated that fluorine ions, which bind strongly to silicon, would be a good candidate. Furthermore, as trimethylsilanolate generated in the ketimine formation could function as a Lewis base, we predicted that a catalytic amount of Lewis base would be sufficient to complete the reaction.

Optimization

Cyclohexyl phenyl ketone was selected as a model substrate for developing the improved method because of its structural similarity to benzophenone. First, according to our hypothesis, we tested various fluorinated reagents as catalysts (Table [5]). Metal fluorides such as sodium fluoride, potassium fluoride, and silver(I) fluoride did not give the desired product. We next examined the use of tetra-n-butylammonium fluoride (TBAF), a deprotecting reagent for the trimethylsilyl group. Fortunately, we obtained the desired unsubstituted ketimine in high yield using a commercially available TBAF solution in THF. Interestingly, however, tetramethylammonium fluoride (TMAF), the same quaternary ammonium salt as TBAF, did not give the desired product. Diethylaminosulfur trifluoride (DAST), used as a nucleophilic fluorinating reagent, also did not give the desired compound. The fact that only TBAF gave the desired product led us to wonder whether the tetrabutylammonium cation was necessary, and we investigated various tetrabutylammonium salts. Tetra-n-butylammonium chloride (TBACl), the halide analogue of TBAF, and tetra-n-butylammonium difluorotriphenylsilicate (TBAT), the same nucleophilic fluorine reagent as TBAF, did not give N-unsubstituted ketimine. Oxoanion salts such as TBAOAc and TBAOH gave the same results. Only in the case of TBAH₂F₃ did the reaction proceed slightly. TBAF-H₂O also gave N-unsubstituted ketimine but in lower yields than the TBAF solution in THF. We tried in-situ generation of TBAF from TBACl and silver(I) fluoride, but the intended reaction was not observed. Finally, various alkoxide salts were also examined, but none served as a catalyst. Based on these results, TBAF was selected as the optimal catalyst.

Table 5 Optimization of Lewis Base Catalyst

Entry	Catalyst	3y (%)a	1y (%)a
1	NaF	n.d.	>99
2	KF	n.d.	>99
3	AgF	n.d.	>99
4	TBAFb	81	10
5	TMAF	n.d.	98
6	DAST	n.d.	98
7	TBACl	n.d.	>99
8	TBAT	n.d.	98
9	TBAOAc	n.d.	97
10	TBAOHc	n.d.	>99
11	TBAH2F3	11	87
12	TBAF·H2O	53	48
13	TBACl + AgF	n.d.	94
14	NaOMe	n.d.	>99
15	NaOEt	n.d.	>99
16	NaOtBu	n.d.	98

^a Determined by ¹H NMR analysis of the crude mixture.

^b Commercially available TBAF solution (1 M in THF) was used without additional THF.

^c Commercially available TBAOH solution (10% in MeOH) was used.

We investigated reaction temperatures with a TBAF catalyst (Table [6]). Yields increased as the reaction temperature was decreased from 70 °C to 25 °C (room temperature), but decreased significantly when the reaction temperature was further decreased to 0 °C. Thus, we determined that 25 °C was the optimum reaction temperature. Therefore, we reduced the temperature from 90 °C, the optimum temperature for the scandium method, to 25 °C, but the ketone was not consumed even after extending the reaction time, and further optimization was necessary.

Table 6 Effect of Reaction Temperature for TBAF in THF^a

Entry	Temp (°C)	3y (%)b	1y (%)b
1	70	10	88
2	60	26	69
3	50	50	49
4	40	63	31
5	25	81	10
6	0	16	83

^a Commercially available TBAF solution (1 M in THF) was used.

^b Determined by ¹H NMR analysis of the crude mixture.

In the catalyst screening, TBAF-H₂O also gave N-unsubstituted ketimine 3y. Thus, several solvents were examined with TBAF-H₂O because this reagent could be handled as a solid (Table [7]). When TBAF-H₂O was used in THF solvent, the yield was lower than that of commercially available TBAF solution in THF. Interestingly, the use of DMF as a solvent significantly increased the yield. In the case of acetonitrile, no unsubstituted ketimine was observed, and the reaction mixture became complex. Chlorobenzene, the optimal solvent for the scandium method, gave almost the same results as THF.

Table 7 Effects of Solvent for TBAF·H₂O

Entry	Solvent	3y (%)a	1y (%)a
1	THF	53	48
2	DMF	89	15
3	MeCN	n.d.	(complex)
4	PhCl	54	49

^a Determined by ¹H NMR analysis of the crude mixture.

The above results suggested that DMF positively affects the reaction. Because TBAF-H₂O is not easy to handle due to its hygroscopicity, we used TBAF solution in THF as the optimal catalyst and screened various Lewis basic additives (Table [8]). Adding 0.10 equiv of DMF improved the yield, so the amount of DMF added was optimized. Yields improved as the amount of additive was increased but peaked at 1.0 equiv DMF. Next, we screened several compounds used as polar aprotic solvents and DMF, and 1,3-dimethyl-2-imidazolidinone (DMI) was found to be the most effective additive. Furthermore, quantitative conversion was achieved by reducing the nitrogen source to 1.5 equiv. As a result of these optimizations, we have developed an improved method requiring no strong metal Lewis acid reagents. High temperatures were no longer required under the reaction conditions, which suppressed undesired isomerization of the enamines.

Table 8 Effect of Lewis Base Additive on Reactivity^a

Entry	Additive (equiv)	3y (%)b
1	none	45
2	DMF (0.10)	52
3	DMF (0.20)	63
4	DMF (1.0)	68
5	DMF (2.0)	66
6	DMSO (1.0)	59
7	DMP (1.0)	80
8	NMP (1.0)	79
9	DMPU (1.0)	91
10	DMI (1.0)	98
11c	DMI (1.0)	>99

^a Commercially available TBAF solution (1 M in THF) was used.

^b Determined by ¹H NMR analysis of the crude mixture.

^c 2 (1.5 equiv) was used.

Benzophenone was almost quantitatively converted into N-unsubstituted ketimine 3a by stirring with a 10 mol% TBAF solution and 2.0 equiv of bis(trimethylsilyl)amine for 6 h at 25 °C. For further improvement, we screened the reaction solvents (Table [9]). When the amount of catalyst was first reduced to 2.5 mol% and the reaction time shortened to 2 h, only a 25% yield was obtained. Adding DMF solvent improved the yield dramatically, however, giving almost quantitative amounts of N-unsubstituted ketimine. Encouraged by this result, we investigated various solvents. Regarding other amide solvents such as DMF, DMA gave high yields of unsubstituted ketimine but NMP did not improve the yields. DMSO drastically reduced the yields. Urea derivatives DMPU and DMI were effective as additives for alkyl ketones but unsuitable for benzophenone. Acetonitrile, dichloromethane, chlorobenzene, diethyl ether, and 1,4-dioxane were also considered but did not afford good results. From these results, DMF was found to be the best solvent. The reaction concentration was also optimized, and 1.0 M was optimal. The optimal amount of the nitrogen source was 1.5 equivalents. Based on these results, the optimal conditions for diaryl ketones were established as those detailed in Table [9], entry 17.

Table 9 Effect of Solvents for Benzophenone as Substrate^a

Entry	Solvent	Conc. (M)	Time (h)	3a (%)b
1c	none	–	6	>99
2	none	–	2	25
3	DMF	1.0	2	>99
4	DMA	1.0	2	82
5	NMP	1.0	2	21
6	DMSO	1.0	2	3
7	DMPU	1.0	2	5
8	DMI	1.0	2	n.d.
9	MeCN	1.0	2	n.d.
10	CH2Cl2	1.0	2	n.d.
11	PhCl	1.0	2	4
12	Et2O	1.0	2	8
13	1,4-dioxane	1.0	2	9
14	DMF	0.50	2	99
15	DMF	2.0	2	96
16	DMF	1.0	1	98
17d	DMF	1.0	1	>99

^a Commercially available TBAF solution (1 M in THF) was used.

^b Determined by ¹H NMR analysis of the crude mixture.

^c TBAF (10 mol%) was used.

^d 2 (1.5 equiv) was used.

Substrate Scope

With the above optimizations determined, we next examined the substrate scope (Scheme [12]). Alkyl-substituted unsubstituted ketimines are unstable and difficult to purify, and therefore yields are given in NMR yields based on internal standards unless otherwise mentioned. Ketones with cyclic alkyl groups reacted well and gave the corresponding N-unsubstituted ketimines almost quantitatively. Ketones with linear alkyl groups required extended reaction times. Notably, N-unsubstituted ketimines derived from cyclohexyl phenyl ketone or butyrophenone were isolated as hydrochloride salts, confirming that truly unsubstituted ketimines were formed. The reaction proceeded well with non-linear alkyl groups such as isopropyl and tert-butyl. Dicyclopropyl ketone gave the corresponding dialkyl N-unsubstituted ketimine in high yield, although the reaction took 72 h. To evaluate functional group compatibility, electron-withdrawing and -donating groups were applied to the reaction. The free hydroxy group was tolerated, giving the corresponding N-unsubstituted ketimine in high yield by increasing the reaction temperature. N-Unsubstituted ketimines 3x and 3ak, frequently used as substrates for nucleophilic addition reactions,[5] [6] [30] [31] [40] were also synthesized in high yields by fine-tuning the reaction conditions. For diaryl ketones, conditions optimized for benzophenone were applied, and the corresponding unsubstituted ketimines were successfully obtained almost quantitatively. In particular, 3ai and 3am have ester and amide structures, respectively, indicating that these functional groups are tolerated in this reaction. Benzophenone imine was isolated by silica gel column chromatography after removing DMF from the crude mixture by extraction.

Application to One-Pot Synthesis

The TBAF method could also be applied to one-pot reactions. First, the one-pot Strecker reaction was investigated. The classical Strecker reaction is a three-component reaction of an aldehyde (or ketone), an amine, and a cyano source, and N-unprotected aminonitriles can be directly synthesized using ammonia as the amine source, The three-component Strecker reaction is widely used in the case of aldehydes. In the case of ketones, however, the reported examples are limited, probably due to the slow formation of ketimine intermediates.[14] [16] ^, [41] [42] [43] [44] Therefore, we hypothesized that if our TBAF method could be applied to one-pot reactions, the problem of the formation of ketimine intermediates in conventional methods could be solved, and the Strecker reaction of ketones would proceed more efficiently. Based on the above hypothesis, we performed one-pot Strecker reactions using alkyl ketones or benzophenone (Scheme [13], Method A). As a comparison, the results of a conventional three-component Strecker reaction using the same ketones are also shown (Method B). In both ketones, the one-pot reactions (Method A) proceeded well, and the corresponding α,α-disubstituted-α-aminonitriles were obtained in high yields. On the other hand, the conventional three-component reaction (Method B) gave lower yields. These results indicated that our one-pot synthesis is more effective than the conventional method.

Because this one-pot reaction was more efficient than the conventional method, we examined the range of substrates applicable to the one-pot Strecker reaction (Scheme [14]). All reaction conditions for the first step of the one-pot reaction are shown in Scheme [12]. The one-pot reaction proceeded well for cyclic and linear alkyl groups, and α-tetrasubstituted-α-amino acid derivatives were obtained in high yields. Not only linear chains but also secondary and tertiary alkyl groups and dialkyl ketones could be converted into the corresponding aminonitriles in high yields. In particular, aminonitrile 12u was a contiguous tetrasubstituted-α-amino acid derivative, a bulky amino acid unit that has recently attracted attention in medicinal chemistry. Both electron-withdrawing and electron-donating groups were permissible in the one-pot reaction. The product aminonitrile 12aj, with a free hydroxy group at the α-position, decomposed in the retro-aldol reaction, but the decomposition was minimized by shortening the reaction time to obtain aminonitrile in good yield. Electrophilic ketones such as trifluoroacetophenone and isatin were also converted into aminonitriles. We further studied one-pot reactions with benzophenone-type ketones. Benzophenone-type imines are relatively inactive ketimines rarely used as electrophiles. In this one-pot reaction, the addition reaction proceeded well, and α,α-diaryl substituted α-amino acid derivatives were synthesized efficiently and in high yield.

As this method was found to apply to one-pot reactions, we next examined the one-pot reaction of hydrophosphonylation (Scheme [15]). Hydrophosphonylation of the imine with phosphite is a typical example of nucleophilic addition reaction to imines, and its products are also essential compounds such as α-amino acid analogues.[45] [46] [47] [48] This one-pot reaction proceeded well, and the synthesis of various α,α-disubstituted-α-aminophosphonates was achieved. The resulting aminophosphonates were less tolerant to silica gel, but cooling the eluent to –50 °C prevented their degradation. Product precipitation inhibited the reaction; therefore, an additional solvent was used in the second step of the reaction when necessary.

One-pot hydrophosphonylation was also possible with the scandium-catalyzed method (Scheme [16]). In particular, diaryl ketones were more compatible with the scandium method, and the synthesis of various α,α-diaryl-substituted α-aminophosphonates was achieved.

Transformations of the Products

The aminonitrile and aminophosphonate derivatives synthesized in our one-pot reaction could be converted into various compounds (Scheme [17]). The nitrile moiety of the Strecker product aminonitrile was successfully converted into an amide 14 and a hydantoin 15.[49] [50] The hydantoin moiety is also found in pharmaceuticals and biologically active compounds. In particular, hydantoin 15 is an analogue of the antiepileptic drug phenytoin. The products of nucleophilic addition reactions to N-unsubstituted ketimines are N-unprotected amines that can be directly converted without a deprotection process. Aminonitrile 12y was condensed directly with Boc-protected glycine to give the dipeptide derivative 16 in high yield. Aminophosphonate 13ak was similarly condensed directly with glycine to yield peptidyl phosphonates 17 and 18. Peptidyl diaryl phosphonates are reported as specific inhibitors of serine proteases.[51]

Mechanistic Studies

To clarify the reaction mechanism of the TBAF method, we performed a series of control experiments (Scheme [18]). For simplicity, the control experiments were performed with benzophenone and without additives. First, NMR analysis of the reaction mixture revealed that hexamethyldisiloxane was generated as a co-product, as expected (Scheme [18a]). Next, adding desiccants such as MS 4A and MgSO₄ decreased the yield slightly (Scheme [18b]), suggesting that a small amount of H₂O has a positive effect. The desired N-unsubstituted ketimine was also generated when catalytic amounts of tetrabutylammonium chloride (TBACl) and potassium trimethylsilanolate (KOTMS) were used together instead of a TBAF catalyst (Scheme [18c]). Interestingly, the reaction did not proceed when TBACl or KOTMS was used alone.

Because the reaction proceeded with the combination of TBACl and KOTMS, we investigated how the length of the alkyl chain of tetraalkylammonium chloride (R₄NCl) affected the yield (Table [10]). The desired reaction did not proceed with the short alkyl chains methyl and ethyl, but the three-carbon isopropyl gave a small amount of the N-unsubstituted ketimine. The use of longer butyl and hexyl groups further improved the yield. When the reaction solution was left to stand, it was found to separate into two layers. NMR analysis of each layer revealed that the upper layer contained relatively low-polarity compounds such as hexamethyldisiloxane, while the lower layer contained relatively high-polarity compounds such as ketimine. From these results, it is considered that tetraalkylammonium salts serve as phase-transfer catalysts between the low- and high-polarity layers and that the length of the alkyl chain is essential for solubility in the low-polarity layer.

Table 10 Effects of Alkyl Chain Length of the Tetraalkylammonium Catalyst

Entry	R	3a (%)a
1	CH3 (Me)	n.r.
2	C2H5 (Et)	n.r.
3	C3H7 (nPr)	11
4	C4H9 (nBu)	58
5	C6H13 (nHex)	60

^a Determined by ¹H NMR analysis of the crude mixture.

Proposed Reaction Mechanism

Based on the results of the above mechanism experiments, a proposed reaction mechanism is shown in Scheme [19]. The commercially available TBAF solution in THF could contain a small amount of H₂O. First, alkoxide intermediate I is formed from TBAF, H₂O, and bis(trimethylsilyl)amine. Alkoxide intermediate I reacts with unreacted bis(trimethylsilyl)amine to form amide intermediate II, yielding hexamethyldisiloxane. Amide intermediate II is an active nitrogen source that reacts with ketones to give hemiaminal intermediate III. After silyl group transfer gives intermediate IV, intermediate IV releases the desired N-unsubstituted ketimine and regenerates intermediate I. Because intermediate I is also generated by the reaction of TBACl with KOTMS, we assume that intermediate I is the catalytically active species in this reaction.

Conclusion

We developed new catalytic synthetic methods for N-unsubstituted ketimines. These methods do not require strongly basic metal reagents or an excess of the nitrogen source, as used in conventional methods. In addition, the starting materials and catalysts were stable and readily available, making them much simpler than conventional methods that require unstable or explosive reagents. These methods provided facile access to various N-unsubstituted ketimines, and several previously inaccessible functional groups were tolerated. Furthermore, applications to one-pot reactions allowed us to synthesize important nitrogen-containing compounds without isolation of unstable N-unsubstituted ketimines, thereby streamlining the synthetic process. We hope that the detailed information described in this account will stimulate further development of related processes in the near future.

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

The computations were carried out using the computer resources offered under the category of General Projects by the Research Institute for Information Technology at Kyushu University.

• References

• 1 Present address: Department of Pharmacy, Kyushu University Hospital, 3-1-1 Maidashi, Higashi-ku, Fukuoka, 812-8582, Japan
• 2 Morisaki K, Morimoto H, Ohshima T. ACS Catal. 2020; 10: 6924
  CrossrefSearch in Google Scholar
• 3 Yin Q, Shi Y, Wang J, Zhang X. Chem. Soc. Rev. 2020; 49: 6141
  CrossrefPubMedSearch in Google Scholar
• 4 Kondo Y, Morimoto H, Ohshima T. Chem. Lett. 2020; 49: 497
  CrossrefSearch in Google Scholar
• 5 Kadota T, Sawa M, Kondo Y, Morimoto H, Ohshima T. Org. Lett. 2021; 23: 4553
  CrossrefPubMedSearch in Google Scholar
• 6 Yamada K, Kondo Y, Kitamura A, Kadota T, Morimoto H, Ohshima T. ACS Catal. 2023; 13: 3158
  CrossrefSearch in Google Scholar
• 7 Cornell EF. J. Am. Chem. Soc. 1928; 50: 3311
  CrossrefSearch in Google Scholar
• 8 Hou G, Gosselin F, Li W, McWilliams JC, Sun Y, Weisel M, O’Shea PD, Chen C, Davies IW, Zhang X. J. Am. Chem. Soc. 2009; 131: 9882
  CrossrefPubMedSearch in Google Scholar
• 9 Gosselin F, O’Shea PD, Roy S, Reamer RA, Chen C, Volante RP. Org. Lett. 2005; 7: 355
  CrossrefPubMedSearch in Google Scholar
• 10 Harris GH, Harriman BR, Wheeler KW. J. Am. Chem. Soc. 1946; 68: 846
  CrossrefPubMedSearch in Google Scholar
• 11 Brenner DG, Cavolowsky KM, Shepard KL. J. Heterocycl. Chem. 1985; 22: 805
  CrossrefSearch in Google Scholar
• 12 Strain HH. J. Am. Chem. Soc. 1930; 52: 820
  CrossrefSearch in Google Scholar
• 13 Verardo G, Giumanini AG, Strazzolini P, Poiana M. Synth. Commun. 1988; 18: 1501
  CrossrefSearch in Google Scholar
• 14 Sato N, Jitsuoka M, Ishikawa S, Nagai K, Tsuge H, Ando M, Okamoto O, Iwaasa H, Gomori A, Ishihara A, Kanatani A, Fukami T. Bioorg. Med. Chem. Lett. 2009; 19: 1670
  CrossrefPubMedSearch in Google Scholar
• 15 Yamashita Y, Matsumoto M, Chen Y.-J, Kobayashi S. Tetrahedron 2012; 68: 7558
  CrossrefSearch in Google Scholar
• 16 Koos M, Mosher HS. Tetrahedron 1993; 49: 1541
  CrossrefSearch in Google Scholar
• 17 Lee JH, Gupta S, Jeong W, Rhee YH, Park J. Angew. Chem. Int. Ed. 2012; 51: 10851
  CrossrefPubMedSearch in Google Scholar
• 18 Kondo Y, Morisaki K, Hirazawa Y, Morimoto H, Ohshima T. Org. Process Res. Dev. 2019; 23: 1718
  CrossrefSearch in Google Scholar
• 19 Kondo Y, Kadota T, Hirazawa Y, Morisaki K, Morimoto H, Ohshima T. Org. Lett. 2020; 22: 120
  CrossrefPubMedSearch in Google Scholar
• 20 Kondo Y, Hirazawa Y, Kadota T, Yamada K, Morisaki K, Morimoto H, Ohshima T. Org. Lett. 2022; 24: 6594
  CrossrefPubMedSearch in Google Scholar
• 21 Noyori R, Murata S, Suzuki M. Tetrahedron 1981; 37: 3899
  CrossrefSearch in Google Scholar
• 22 Kobayashi SIn. Lanthanides: Chemistry and Use in Organic Synthesis . Springer; Berlin/Heidelberg: 1999: 63
  CrossrefSearch in Google Scholar
• 23 Shibasaki M, Yoshikawa N. Chem. Rev. 2002; 102: 2187
  CrossrefPubMedSearch in Google Scholar
• 24 Mikami K, Terada M, Matsuzawa H. Angew. Chem. Int. Ed. 2002; 41: 3554
  CrossrefPubMedSearch in Google Scholar
• 25 Dumeunier R, Markó IE. Tetrahedron Lett. 2004; 45: 825
  CrossrefSearch in Google Scholar
• 26 Sawa M, Miyazaki S, Yonesaki R, Morimoto H, Ohshima T. Org. Lett. 2018; 20: 5393
  CrossrefPubMedSearch in Google Scholar
• 27 Yasukawa N, Nakamura S. Chem. Commun. 2023; 59: 8343
  CrossrefPubMedSearch in Google Scholar
• 28 Sheldon RA. Green Chem. 2007; 9: 1273
  CrossrefSearch in Google Scholar
• 29 Sheldon RA. Chem. Commun. 2008; 3352
  CrossrefPubMed
• 30 Morisaki K, Morimoto H, Ohshima T. Chem. Commun. 2017; 53: 6319
  CrossrefPubMedSearch in Google Scholar
• 31 Sawa M, Morisaki K, Kondo Y, Morimoto H, Ohshima T. Chem. Eur. J. 2017; 23: 17022
  CrossrefPubMedSearch in Google Scholar
• 32 Hayashi Y. Chem. Sci. 2016; 7: 866
  CrossrefPubMedSearch in Google Scholar
• 33 Gonzalez J, Carroll FI. Tetrahedron Lett. 1996; 37: 8655
  CrossrefSearch in Google Scholar
• 34 Shirakawa S, Maruoka K. Angew. Chem. Int. Ed. 2013; 52: 4312
  CrossrefPubMedSearch in Google Scholar
• 35 O’Donnell MJ. Tetrahedron 2019; 75: 3667
  CrossrefSearch in Google Scholar
• 36 Kitamura M, Shirakawa S, Maruoka K. Angew. Chem. Int. Ed. 2005; 44: 1549
  CrossrefPubMedSearch in Google Scholar
• 37 Chen Y.-J, Seki K, Yamashita Y, Kobayashi S. J. Am. Chem. Soc. 2010; 132: 3244
  CrossrefPubMedSearch in Google Scholar
• 38 Wolfe JP, Åhman J, Sadighi JP, Singer RA, Buchwald SL. Tetrahedron Lett. 1997; 38: 6367
  CrossrefPubMedSearch in Google Scholar
• 39 Mann G, Hartwig JF, Driver MS, Fernández-Rivas C. J. Am. Chem. Soc. 1998; 120: 827
  CrossrefSearch in Google Scholar
• 40 Yonesaki R, Kusagawa I, Morimoto H, Hayashi T, Ohshima T. Chem. Asian J. 2020; 15: 499
  CrossrefPubMedSearch in Google Scholar
• 41 Steiger RE. Org. Synth. 1944; 24: 9
  CrossrefSearch in Google Scholar
• 42 Kuethe JT, Gauthier DR, Beutner GL, Yasuda N. J. Org. Chem. 2007; 72: 7469
  CrossrefPubMedSearch in Google Scholar
• 43 Palacios F, Ochoa de Retana AM, Pascual S, Fernández de Trocóniz G. Tetrahedron 2011; 67: 1575
  CrossrefSearch in Google Scholar
• 44 Islam S, Bučar D.-K, Powner MW. Nat. Chem. 2017; 9: 584
  CrossrefSearch in Google Scholar
• 45 Bera K, Namboothiri IN. N. Asian J. Org. Chem. 2014; 3: 1234
  CrossrefSearch in Google Scholar
• 46 Ordóñez M, Viveros-Ceballos JL, Cativiela C, Sayago FJ. Tetrahedron 2015; 71: 1745
  CrossrefSearch in Google Scholar
• 47 Chen L. Synthesis 2018; 50: 440
  Thieme ConnectSearch in Google Scholar
• 48 Maestro A, del Corte X, López-Francés A, Martínez de Marigorta E, Palacios F, Vicario J. Molecules 2021; 26: 3202
  CrossrefPubMedSearch in Google Scholar
• 49 Chitale S, Derasp JS, Hussain B, Tanveer K, Beauchemin AM. Chem. Commun. 2016; 52: 13147
  CrossrefPubMedSearch in Google Scholar
• 50 Wynands L, Delacroix S, Nguyen Van Nhien A, Soriano E, Marco-Contelles J, Postel D. Tetrahedron 2013; 69: 4899
  CrossrefSearch in Google Scholar
• 51 Maślanka M, Mucha A. Pharmaceuticals 2019; 12: 86
  CrossrefPubMedSearch in Google Scholar

Corresponding Authors

Publication History

Received: 26 May 2023

Accepted after revision: 17 July 2023

Accepted Manuscript online:
17 July 2023

Article published online:
18 September 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany

• References

• 1 Present address: Department of Pharmacy, Kyushu University Hospital, 3-1-1 Maidashi, Higashi-ku, Fukuoka, 812-8582, Japan
• 2 Morisaki K, Morimoto H, Ohshima T. ACS Catal. 2020; 10: 6924
  CrossrefSearch in Google Scholar
• 3 Yin Q, Shi Y, Wang J, Zhang X. Chem. Soc. Rev. 2020; 49: 6141
  CrossrefPubMedSearch in Google Scholar
• 4 Kondo Y, Morimoto H, Ohshima T. Chem. Lett. 2020; 49: 497
  CrossrefSearch in Google Scholar
• 5 Kadota T, Sawa M, Kondo Y, Morimoto H, Ohshima T. Org. Lett. 2021; 23: 4553
  CrossrefPubMedSearch in Google Scholar
• 6 Yamada K, Kondo Y, Kitamura A, Kadota T, Morimoto H, Ohshima T. ACS Catal. 2023; 13: 3158
  CrossrefSearch in Google Scholar
• 7 Cornell EF. J. Am. Chem. Soc. 1928; 50: 3311
  CrossrefSearch in Google Scholar
• 8 Hou G, Gosselin F, Li W, McWilliams JC, Sun Y, Weisel M, O’Shea PD, Chen C, Davies IW, Zhang X. J. Am. Chem. Soc. 2009; 131: 9882
  CrossrefPubMedSearch in Google Scholar
• 9 Gosselin F, O’Shea PD, Roy S, Reamer RA, Chen C, Volante RP. Org. Lett. 2005; 7: 355
  CrossrefPubMedSearch in Google Scholar
• 10 Harris GH, Harriman BR, Wheeler KW. J. Am. Chem. Soc. 1946; 68: 846
  CrossrefPubMedSearch in Google Scholar
• 11 Brenner DG, Cavolowsky KM, Shepard KL. J. Heterocycl. Chem. 1985; 22: 805
  CrossrefSearch in Google Scholar
• 12 Strain HH. J. Am. Chem. Soc. 1930; 52: 820
  CrossrefSearch in Google Scholar
• 13 Verardo G, Giumanini AG, Strazzolini P, Poiana M. Synth. Commun. 1988; 18: 1501
  CrossrefSearch in Google Scholar
• 14 Sato N, Jitsuoka M, Ishikawa S, Nagai K, Tsuge H, Ando M, Okamoto O, Iwaasa H, Gomori A, Ishihara A, Kanatani A, Fukami T. Bioorg. Med. Chem. Lett. 2009; 19: 1670
  CrossrefPubMedSearch in Google Scholar
• 15 Yamashita Y, Matsumoto M, Chen Y.-J, Kobayashi S. Tetrahedron 2012; 68: 7558
  CrossrefSearch in Google Scholar
• 16 Koos M, Mosher HS. Tetrahedron 1993; 49: 1541
  CrossrefSearch in Google Scholar
• 17 Lee JH, Gupta S, Jeong W, Rhee YH, Park J. Angew. Chem. Int. Ed. 2012; 51: 10851
  CrossrefPubMedSearch in Google Scholar
• 18 Kondo Y, Morisaki K, Hirazawa Y, Morimoto H, Ohshima T. Org. Process Res. Dev. 2019; 23: 1718
  CrossrefSearch in Google Scholar
• 19 Kondo Y, Kadota T, Hirazawa Y, Morisaki K, Morimoto H, Ohshima T. Org. Lett. 2020; 22: 120
  CrossrefPubMedSearch in Google Scholar
• 20 Kondo Y, Hirazawa Y, Kadota T, Yamada K, Morisaki K, Morimoto H, Ohshima T. Org. Lett. 2022; 24: 6594
  CrossrefPubMedSearch in Google Scholar
• 21 Noyori R, Murata S, Suzuki M. Tetrahedron 1981; 37: 3899
  CrossrefSearch in Google Scholar
• 22 Kobayashi SIn. Lanthanides: Chemistry and Use in Organic Synthesis . Springer; Berlin/Heidelberg: 1999: 63
  CrossrefSearch in Google Scholar
• 23 Shibasaki M, Yoshikawa N. Chem. Rev. 2002; 102: 2187
  CrossrefPubMedSearch in Google Scholar
• 24 Mikami K, Terada M, Matsuzawa H. Angew. Chem. Int. Ed. 2002; 41: 3554
  CrossrefPubMedSearch in Google Scholar
• 25 Dumeunier R, Markó IE. Tetrahedron Lett. 2004; 45: 825
  CrossrefSearch in Google Scholar
• 26 Sawa M, Miyazaki S, Yonesaki R, Morimoto H, Ohshima T. Org. Lett. 2018; 20: 5393
  CrossrefPubMedSearch in Google Scholar
• 27 Yasukawa N, Nakamura S. Chem. Commun. 2023; 59: 8343
  CrossrefPubMedSearch in Google Scholar
• 28 Sheldon RA. Green Chem. 2007; 9: 1273
  CrossrefSearch in Google Scholar
• 29 Sheldon RA. Chem. Commun. 2008; 3352
  CrossrefPubMed
• 30 Morisaki K, Morimoto H, Ohshima T. Chem. Commun. 2017; 53: 6319
  CrossrefPubMedSearch in Google Scholar
• 31 Sawa M, Morisaki K, Kondo Y, Morimoto H, Ohshima T. Chem. Eur. J. 2017; 23: 17022
  CrossrefPubMedSearch in Google Scholar
• 32 Hayashi Y. Chem. Sci. 2016; 7: 866
  CrossrefPubMedSearch in Google Scholar
• 33 Gonzalez J, Carroll FI. Tetrahedron Lett. 1996; 37: 8655
  CrossrefSearch in Google Scholar
• 34 Shirakawa S, Maruoka K. Angew. Chem. Int. Ed. 2013; 52: 4312
  CrossrefPubMedSearch in Google Scholar
• 35 O’Donnell MJ. Tetrahedron 2019; 75: 3667
  CrossrefSearch in Google Scholar
• 36 Kitamura M, Shirakawa S, Maruoka K. Angew. Chem. Int. Ed. 2005; 44: 1549
  CrossrefPubMedSearch in Google Scholar
• 37 Chen Y.-J, Seki K, Yamashita Y, Kobayashi S. J. Am. Chem. Soc. 2010; 132: 3244
  CrossrefPubMedSearch in Google Scholar
• 38 Wolfe JP, Åhman J, Sadighi JP, Singer RA, Buchwald SL. Tetrahedron Lett. 1997; 38: 6367
  CrossrefPubMedSearch in Google Scholar
• 39 Mann G, Hartwig JF, Driver MS, Fernández-Rivas C. J. Am. Chem. Soc. 1998; 120: 827
  CrossrefSearch in Google Scholar
• 40 Yonesaki R, Kusagawa I, Morimoto H, Hayashi T, Ohshima T. Chem. Asian J. 2020; 15: 499
  CrossrefPubMedSearch in Google Scholar
• 41 Steiger RE. Org. Synth. 1944; 24: 9
  CrossrefSearch in Google Scholar
• 42 Kuethe JT, Gauthier DR, Beutner GL, Yasuda N. J. Org. Chem. 2007; 72: 7469
  CrossrefPubMedSearch in Google Scholar
• 43 Palacios F, Ochoa de Retana AM, Pascual S, Fernández de Trocóniz G. Tetrahedron 2011; 67: 1575
  CrossrefSearch in Google Scholar
• 44 Islam S, Bučar D.-K, Powner MW. Nat. Chem. 2017; 9: 584
  CrossrefSearch in Google Scholar
• 45 Bera K, Namboothiri IN. N. Asian J. Org. Chem. 2014; 3: 1234
  CrossrefSearch in Google Scholar
• 46 Ordóñez M, Viveros-Ceballos JL, Cativiela C, Sayago FJ. Tetrahedron 2015; 71: 1745
  CrossrefSearch in Google Scholar
• 47 Chen L. Synthesis 2018; 50: 440
  Thieme ConnectSearch in Google Scholar
• 48 Maestro A, del Corte X, López-Francés A, Martínez de Marigorta E, Palacios F, Vicario J. Molecules 2021; 26: 3202
  CrossrefPubMedSearch in Google Scholar
• 49 Chitale S, Derasp JS, Hussain B, Tanveer K, Beauchemin AM. Chem. Commun. 2016; 52: 13147
  CrossrefPubMedSearch in Google Scholar
• 50 Wynands L, Delacroix S, Nguyen Van Nhien A, Soriano E, Marco-Contelles J, Postel D. Tetrahedron 2013; 69: 4899
  CrossrefSearch in Google Scholar
• 51 Maślanka M, Mucha A. Pharmaceuticals 2019; 12: 86
  CrossrefPubMedSearch in Google Scholar