Development of Tetranuclear Zinc Cluster-Catalyzed Environmentally Friendly Reactions and Mechanistic Studies

Abstract

Our studies on tetranuclear zinc cluster-catalyzed environmentally friendly reactions are presented here. The newly developed μ-oxo-tetranuclear zinc cluster is a highly efficient catalyst for the direct formation of oxazolines from esters, carboxylic acids, lactones, and nitriles; for the transesterification of various methyl esters including α-amino esters and β-keto esters; for the acetylation of alcohols in EtOAc; and for the deacylation of esters in MeOH. A unique hydroxy group-selective acylation in the presence of inherently much more nucleophilic amino groups was also achieved by this zinc cluster. Zinc cluster-catalyzed transesterification, in particular, was drastically accelerated by the addition of alkyl amine and N-heteroaromatic ligands, which coordinate with the metals, stabilize the clusters with lower nuclearities, and enhance catalytic activity for the transesterification. We also performed several mechanistic studies which revealed that alkoxide metal complexes are the active species in this catalytic cycle, and that the Michaelis−Menten behavior of the complexes through an ordered ternary complex mechanism is similar to that of dinuclear metalloenzymes. The deprotonation of nucleophiles was the most important step in this process, not only for achieving high catalytic activity but also for determining chemoselectivity, resulting in the chemical differentiation of alcohols and amines.

1. Introduction

The dramatic evolution of organic chemistry in the 20th century has enabled the synthesis of many complex molecules, including highly functionalized natural products.^1–4) Increasing the supply of scarce or inaccessible natural products is essential for the production of more sophisticated pharmaceutical agents and biological tools. Practical synthesis of such complex molecules, however, remains quite difficult, even with modern organic chemistry techniques, and thus the development of catalyst-promoted atom-economical⁵⁾ and environmentally benign processes is in high demand.^6,7) Among these processes, ester bond formation is a fundamental and well-studied methodology because it is a ubiquitous chemical bond, abundant in both natural and synthetic organic compounds.^1–4) Ester synthesis is commonly performed in synthetic organic chemistry using carboxylic acid with alcohol under stoichiometric amounts of condensation reagent, or by treatment with a highly reactive acylating reagent, such as acyl chloride and acid anhydride, due to the proven reliability of these methods. However, the inevitable formation of more than a stoichiometric amount of co-product has been a drawback. Although catalyst-promoted ester formation between carboxylic acid and alcohol is an ideal method in terms of atom-economy, there remains much room for improvement in terms of practicality due to the use of acidic conditions at a high reaction temperature, resulting in narrow functional group compatibility. An attractive alternative method of ester synthesis is transesterification promoted by a catalytic amount of metal reagent.^8–11) This process generates only nontoxic lower alcohols, and can be performed under almost neutral conditions, allowing for high functional group compatibility. Furthermore, ease of handling, as well as the high stability and high solubility of the esters in most organic solvents compared with the corresponding carboxylic acids, makes them advantageous as a starting material. Several metal-catalyzed transesterification reactions have recently been reported.^12–27) Most of these methods, however, require the use of toxic metal salts or tediously strict reaction conditions. In addition, the application of sterically demanding alcohols to catalytic transesterification remains difficult. Therefore, the development of highly active nontoxic metal-catalyzed transesterification reactions is highly desirable for practical utility. Recently, we developed a direct conversion of carboxylic acids, esters, and lactones with β-amino alcohols to oxazolines catalyzed by a μ₄-oxo-tetranuclear zinc cluster, Zn₄(OCOCF₃)₆O (1b)²⁸⁾ (Fig. 1). This zinc cluster 1b also efficiently catalyzed the transesterification of various methyl esters under mild conditions, and exhibited high tolerance for various functional groups.^29–38) Moreover, we successfully developed an O-selective acylation in the presence of more nucleophilic primary and secondary aliphatic amino groups using 1b-catalyzed transesterification.²⁹⁾ This review presents our studies on tetranuclear zinc cluster-catalyzed environmentally friendly reactions,^28–40) including the unprecedented O-selective acylation of aminoalcohols,^29.37) and their mechanistic studies.^36,38)

Fig. 1. The Structures of μ₄-Oxo-Tetranuclear Zinc Cluster: Zn₄(OCOR)₆O (1)

2. Tetranuclear Zinc Cluster

In 1924, Auger and Robin synthesized an acetate-bridged tetranuclear zinc cluster Zn₄(OCOCH₃)₆O (1a) for the first time by vacuum distillation of zinc acetate hydrate Zn(OCOCH₃)₂·xH₂O (2a).⁴¹⁾ Its μ₄-oxo structure was later established by X-ray crystallographic analysis by Wyart⁴²⁾ and by Koyama and Saito.⁴³⁾ Other tetranuclear zinc clusters with different carboxylate ligands, such as Zn₄(OCOR)₆O (R=Et, n-Pr, t-Bu, Ph, etc.), were subsequently prepared using this synthetic method or similar pyrolytic methods.^44–48) The tetranuclear zinc structures of these clusters were highly stabilized by six carboxylates and one oxo ligand; therefore, such a tetranuclear zinc cluster unit was used as the key component of a metal organic framework.^49,50) The high stability of the clusters, however, includes a negative aspect as well. “Stable” complexes are often “inactive” as a catalyst. In fact, acetate-bridged tetranuclear zinc cluster 1a was an inefficient catalyst in almost all the reactions we examined. Fortunately, during our studies on catalytic oxazoline formation, we found that changing the ligand from an acetate to a trifluoroacetate greatly improved the catalytic activity for oxazoline formation²⁸⁾ and transesterification^29–38) (vide infra). Similarly to 1a, a trifluoroacetate-bridged tetranuclear zinc cluster Zn₄(OCOCF₃)₆O (1b) was originally synthesized by heating a zinc trifluoroacetate hydrate, Zn(OCOCF₃)₂·xH₂O, at 320°C under high vacuum (<0.02 mmHg) to produce the solidified zinc cluster 1b on a glass wall, followed by collection of the resulting white solid under argon atmosphere, avoiding contact with moisture, to give pure 1b in 84% yield.²⁸⁾ We also synthesized a series of tetranuclear zinc clusters, Zn₄(OCOR)₆O (1c: R=CF₂CF₃, 1d: R=CF₂CF₂CF₃, and 1e: R=CF₂CH₂CH₃), as white solids in good yield (70–84%) by pyrolysis of the corresponding zinc carboxylates and purification by vacuum distillation (<0.02 mmHg).³⁸⁾ These preparation methods can be performed in gram-scale.

Although such sublimation provides highly pure zinc clusters, the high temperature synthesis and moisture-free handling required for the creation of 1 decrease its accessibility to organic chemists. Thus, we aimed to develop a more practical synthetic method of 1b, which is the best catalyst of transesterification. After intensive studies, we revealed that construction of the μ₄-oxo-tetranuclear zinc trifluoroacetate motif proceeded smoothly even at 110°C, a much milder condition than that used in the preceding synthetic method (sublimation at 320°C). Heating zinc trifluoroacetate hydrate under toluene reflux conditions, and simple filtration of the resulting precipitates, afforded a white solid, which was a mixture of trifluoroacetate-bridged tetranuclear zinc cluster Zn₄(OCOCF₃)₆O (1b) and its trifluoroacetic acid (TFA) adduct Zn₄(OCOCF₃)₆O·CF₃CO₂H (2b)³⁴⁾ (Chart 1). Because this method does not require high temperature conditions or a sublimation process, kilogram scale synthesis of the zinc cluster proceeded with the same efficiency. The catalytic activity and functional group tolerance of the TFA adduct 2b in transesterification and oxazoline formation were almost identical to those of 1b. Another advantage of 2b is its low hygroscopic nature, which allows handling without the concern of introducing moisture. Currently, the TFA adduct 2b is commercially available under the name of ZnTAC24™, and more than 200 kg of 2b has been produced to date.

Chart 1. Synthesis of Zinc Clusters and Probable Structure of Zn₄(OCOCF₃)₆O·CF₃CO₂H (2b)

3. Direct Catalytic Conversion of Esters, Lactones, and Carboxylic Acids to Oxazolines

Oxazolines are widely present in natural products, and their use in pharmaceutical drug discovery has stimulated great interest in recent years.^51–54) Moreover, enantiomerically pure oxazolines are commonly present in ligands used for asymmetric catalysis.^55–58) A common route to creating oxazolines is the reaction of an acid chloride with a β-amino alcohol; the corresponding hydroxyamide is then treated with thionyl chloride and cyclized with a base via inversion of the configuration. Several milder approaches have been developed for the cyclization of a hydroxyamide. In terms of atom-economy, direct conversion of carboxylic acid equivalents to oxazolines in a catalytic manner is highly desirable. Zinc(II) chloride-catalyzed direct conversion of nitriles with β-amino alcohols to oxazolines was developed and widely utilized for the syntheses of a variety of chiral bis(oxazoline) ligands.^59,60) Despite the easier accessibility of esters compared to nitriles, the direct use of esters as a substrate for oxazoline synthesis is rarely discussed.^61,62) To this end, we developed a direct catalytic conversion of esters and lactones to oxazolines by a tandem condensation–cyclodehydration reaction for the first time.²⁸⁾ Although Zn(OCOCH₃)₂ gave unsatisfactory results, the use of tetranuclear zinc cluster Zn₄(OCOCH₃)₆O (1a) improved the yield from 40% (Table 1, Entry 5) to 55% (Entry 7). Further improvement was achieved by the use of Zn₄(OCOCF₃)₆O (1b) (Entry 8, 83%).

Table 1. Catalyst Screening for Direct Conversion of Ester to Oxazoline

Entry	Catalyst	Yield (%)
1	—	0
2	ZnCl2	46
3	Cd(OAc)2	43
4	ZnO	19
5	Zn(OCOCH3)2	40
6	Zn(OCOCF3)2	65
7	Zn4(OCOCH3)6O (1a)	55
8	Zn4(OCOCF3)6O (1b)	83

Only 1.25 mol% of zinc cluster 1b was needed to catalyze the reactions of various aromatic esters with electron-donating and electron-withdrawing substituents (Table 2). Because zinc cluster 1b is less acidic than Zn(OCOCF₃)₂, acid sensitive methoxyethoxymethyl (MEM) ether and tert-butyldimethylsilyl (TBDMS) ether functionality remained intact under this catalytic reaction.

Table 2. Zinc Cluster-Catalyzed Direct Catalytic Conversion of Ester to Oxazoline

Entry	Ester (R1)	β-Aminoalcohola)	Yield (%)
1	3b: C6H4-4-Br	4a: R2=i-Pr, R3=H	73
2	3c: C6H4-4-NO2	4a: R2=i-Pr, R3=H	71
3b)	3d: C6H4-4-OMEM	4a: R2=i-Pr, R3=H	74
4b)	3e: C6H4-4-OTBDMS	4a: R2=i-Pr, R3=H	70
5b)	3t: C6H4-4-CH2OMEM	4a: R2=i-Pr, R3=H	75
6b)	3g: C6H4-4-CH2OTBDMS	4a: R2=i-Pr, R3=H	87
7	3h: (CH2)3CH3	4a: R2=i-Pr, R3=H	>99
8	3i: (CH2)16CH3	4a: R2=i-Pr, R3=H	80
9	3j: CH2CH2Ph	4a: R2=i-Pr, R3=H	94
10	3j: CH2CH2Ph	4b: R2=t-Bu, R3=H	>99
11	3j: CH2CH2Ph	4c: R2=H, R3=Ph	77
12	3j: CH2CH2Ph	4d: R2=R3=Me	97
13	3J: CH2CH2Ph	4e:	79
14	3j: CH2CH2Ph	4t:	>99
15	3j: CH2CH2Ph	4g:	52

a) R⁴=R⁵=H for 4a–d. b) Reaction time was 36 h.

This catalysis was also applicable to the reaction of a variety of lactones, carboxylic acids, and nitriles.²⁸⁾ Zinc cluster 1b promoted the reaction of 5- to 7-membered lactones to afford the corresponding hydroxy oxazolines, with good to excellent yield. The obtained hydroxy oxazolines should be highly useful for the construction of new functionalized oxazoline libraries. This is the first example of a direct catalytic conversion of esters and lactones to oxazolines. Furthermore, the optimized conditions using 1b are applicable to the reaction of carboxylic acids with higher efficiency. The combination of a variety of carboxylic acids and β-aminoalcohols, including a sterically hindered tert-alcohol derivative, afforded the corresponding oxazolines in high yield. Probably due to the high thermodynamic stability of oxazolines, this procedure does not require azeotropic dehydration, and even the presence of excess water or methanol resulted in only a slight decrease in catalytic efficiency, reflecting the stability of the active species against oxygen nucleophiles.

The reactivities of ester and carboxylic acid are almost the same, while that of nitrile is slightly lower, but not significantly different. On the other hand, when two of these functionalities exist at the same time, we found remarkable chemoselectivity.²⁸⁾ Under 1b-catalyzed conditions, carboxylic acid reacted much faster than nitrile or ester, allowing for easy access to a variety of C₁-symmetric bis(oxazoline) ligands that contain different oxazoline moieties (Chart 2). It is important to know that the dehydrative condensation of carboxylic acids with β-aminoalcochols proceeds much faster than the transesterification of esters. Later, Rh-complexes derived from these C₁-symmetric Phbox ligands were found to be highly efficient catalysts for the asymmetric alkynylation of α-ketoester⁶³⁾ and α-ketiminoester.⁶⁴⁾

Chart 2. Zinc Cluster-Catalyzed Chemoselective Reactions of Carboxylic Acids with β-Amino Alcohols over Nitriles or Esters and Their Application for the Synthesis of C₁-Symmetric Bis(oxazoline) Ligands

At this stage, we propose that this zinc catalysis proceeds through the following mechanism: after dissociation of one of the trifluoroacetates from the cluster, one zinc ion acts as the Lewis acid to activate the ester, and the other zinc ion forms zinc alkoxide 12 or zinc amide 14 to increase the nucleophilicity of the hydroxyl group or amino group similarly to the action of dinuclear metallo-enzymes (Chart 3). The resulting zinc alkoxide 12 or zinc amide 14 then attacks the activated ester in an intramolecular fashion to afford the corresponding aminoester 13 or hydroxyamide 15, which is subsequently converted to the oxazoline 5 through condensation reaction. This cooperative mechanism prompted us to examine the possibility of 1b-catalyzed transesterification (12→13), and ester-amide exchange reactions (14→15).

Chart 3. Plausible Reaction Pathway for Oxazoline Synthesis

4. Transesterification

4.1. Introduction

Because transesterification is an equilibrium reaction, it is difficult to attain high conversions. The following methods have been used to force the reaction toward the product side: (i) the use of excess amounts of either of the reactants, (ii) the use of an enol ester as a reactant, accompanied by the formation of the corresponding aldehyde or ketone, and (iii) the removal of the resulting lower alcohol by molecular sieves or continuous distillation. The last approach is the most ideal method, and several catalytic transesterifications at high temperature using esters of lower alcohols were developed using this approach.^8–11) There is great demand, however, for the development of a versatile transesterification under mild and harmless conditions to produce highly functionalized compounds such as pharmaceutical agents.

4.2. Optimization of the Reaction Conditions

Using zinc cluster 1b as a catalyst, we first examined solvent effects at 80°C or lower³⁰⁾ (Table 3, Entries 1–10). Toluene, chlorobenzene, octane, chloroform, cyclopentyl methyl ether (CPME), and methyl tert-butyl ether (TBME) gave low to moderate yields (Entries 1–5). Coordinating solvents, such as tetrahydrofuran (THF), acetonitrile, dimethyl sulfoxide (DMSO), and N,N-dimethylformamide (DMF) were not effective for this reaction, probably because active site of zinc is occupied by the solvent molecules (Entries 7–10). Finally, we found that diisopropyl ether was the best solvent (Entry 6, 79% yield). Diisopropyl ether forms a roughly 1 : 1 azeotrope with methanol, which boils at ca. 10°C lower than pure diisopropyl ether (bp=68°C) at atmospheric pressure, making it easy to remove the resulting MeOH from the reaction mixture. CAUTION: Diisopropyl ether can form explosive peroxides upon standing in air for long periods. For safety, toluene or CPME can be used as an alternative solvent; these solvents usually give a higher yield, although they also require a higher reaction temperature (Entries 11, 12).

Table 3. Optimization of Reaction Conditions for 1b-Catalyzed Transesterification

Entry	Solvent	Temp. (°C)	Yield (%)
1	Toluene	80	48
2	PhCl	80	41
3	CHCl3	Reflux (bp 61°C)	14
4	CPME	80	38
5	TBME	Reflux (bp 55°C)	58
6	i-Pr2O	Reflux (bp 68°C)	79
7	THF	Reflux (bp 66°C)	20
8	CH3CN	80	16
9	DMF	80	22
10	DMSO	80	34
11	Toluene	Reflux (bp 110°C)	95
12	CPME	Reflux (bp 106°C)	98

4.3. Scope and Limitations (1)

Under the optimized conditions, the scope and limitations of alcohols 16b–i were examined³⁰⁾ (Table 4). The transesterification of 3j with a variety of primary and secondary aliphatic alcohols 16b–i, including benzylic alcohols and allylic alcohols, was efficiently catalyzed by 1b to afford the corresponding esters 17 in high yield (up to 99% yield). Because the reaction conditions are almost neutral, the acid sensitive TBDMS ether of phenol persisted (Entry 4), and neither the isomerization nor cyclization of geraniol (16h) occurred (Entry 7). In contrast, transesterification with tertiary alcohols did not proceed due to steric hindrance. Acidic alcohols, such as phenol (pK_a 9.95) and 1,1,1,3,3,3-hexafluoro-2-propanol (pK_a 9.3), did not succeed in achieving the reaction, probably due to the lower nucleophilicity of the corresponding zinc alkoxides. Taking advantage of this chemoselectivity, then, aliphatic alcohols can be selectively acylated in the presence of acidic alcohols (vide infra).

Table 4. Transesterification of Various Alcohols

Entry	Alcohol (HOR)	Time (h)	Yield (%)
1	16b: HO(CH2)3CH3	18	92
2	16c: HO(CH2)5CH3	24	94
3	16d: HO(CH2)17CH3	18	99
4	16e:	24	92
5	16f:	40	76
6	16g:	40	91
7	16h:	40	91
8	16i:	40	96

Next, we estimated the substrate generality of the ester component³⁰⁾ (Table 5). Aromatic esters with various substituents at the para-position were converted to the corresponding butyl esters 17 in good to excellent yield (Entries 1–7). As shown above, nitrile is a good substrate for zinc cluster 1b-catalyzed oxazoline formation. Moreover, oxazoline reacts with alcohol in the presence of Lewis acid. Under the current reaction conditions, however, neither functional group reacted with the alcohol (Entries 4, 7). Due to the large reactivity difference between aliphatic and aromatic alcohols (vide supra), the reaction of 4-hydroxybenzoate 3m provided the corresponding product 17mb in good yield without the formations of a phenol ester (Entry 6). When methyl cinnamate (3o) was used as a substrate, 1,4-addition did not proceed, and the desired product 17ob was obtained in 86% yield (Entry 8). Various aliphatic esters with higher reactivities can also be used (Entries 9–11) without the loss of the highly acid-sensitive tetrahydropyranyl ether functional group (Entry 10). Dimethyl esters were also converted to the corresponding dibutyl esters in excellent yield, accompanied by only trace amounts of monobutyl esters (Entries 12, 13).

Table 5. Transesterification of Various Methyl Esters

Entry	Ester (RCO2Me)	Time (h)	Yield (%)
1	3a: R′=H	40	96
2	3k: R′=Cl	24	91
3	3b: R′=Br	40	90
4	3l: R′=CN	24	77
5	3c: R′=NO2	40	>99
6	3m: R′=OH	40	76
7	3n:	44	93
8	3o:	24	86
9	3i: CH3(CH2)16CO2Me	24	92
10	3p: THPO(CH2)9CO2Me	44	87
11	3q: c-Hex-CO2Me	40	97
12	3r:	40	>99
13	3s:	40	97

4.4. Scope and Limitations (2) α-Amino Esters

α-Amino esters are widely present in natural and unnatural bioactive compounds, and play an important role in the development of pharmaceuticals. Therefore, the transesterification of α-amino acid esters with a broader scope of N-protective groups, alcohols, and side chains were investigated.^30,35)

First, various N-protected glycine methyl esters 18 were examined (Table 6, Entries 1–8). All reactions proceeded smoothly, to afford the corresponding butyl esters 19 in high yield without any deprotection. Noteworthy was that a variety of carbamate groups, such as benzyloxycarbonyl (Cbz) (18a), tert-butoxycarbonyl (Boc) (18b), 9-fluorenylmethyloxycarbonyl (Fmoc) (18c), allyloxycarbonyl (Alloc) (18d), and 2,2,2-trichloroethyl chloroformate (Troc) (18e), did not decompose during the reaction (Entries 1–5). Because the transesterification conditions were completely neutral, Fmoc (18c) and 2-nitrobenzenesulfonyl (Ns) (18f) groups, which are sensitive to basic conditions and nucleophiles, did not decompose at all (Entries 3, 6). The scope and limitations of alcohols 16 were also explored (Entries 9–16). Similarly to the simple esters, the reactions of Cbz-Gly-OMe (18a) with various primary and secondary alcohols worked well to give the corresponding esters 19 in high yield, while the reaction with D-menthol (16n) was rather slow, presumably due to steric hindrance; it was completed by a prolonged reaction time (48 h) to afford Cbz–glycine D-methyl ester (19an) in 92% yield (Entry 13).

Table 6. Transesterification of Various N-Protected Glycine Methyl Esters and Alcohols

Entry	NPP′	Alcohol	Yield (%)
1	18a: NHCbz	16b: HOBu	97
2	18b: NHBoc	16b: HOBu	99
3	18c: NHFmoc	16b: HOBu	95
4	18d: NHCO(allyl)	16b: HOBu	93
5	18e: NHCO(CH2Cl3)	16b: HOBu	98
6	18f: NHNs	16b: HOBu	84
7	18g: NHBz	16b: HOBu	93
8a)	18h: NPht	16b: HOBu	95
9	18a: NHCbz	16j: HOC18H37	90
10	18a: NHCbz	16k: HOCH2-t-Bu	99
11	18a: NHCbz	16l: HOCHPr2	79
12	18a: NHCbz	16m: HO-c-Hex	99
13b)	18a: NHCbz	16n: D-Menthol	92
14	18a: NHCbz	16f: Allyl alcohol	98
15	18a: NHCbz	16g: Cinnamyl alcohol	99
16	18a: NHCbz	16a: HOBn	96

a) Reaction time was 24 h. b) Reaction time was 48 h.

The same transesterification was applied to the conversion of various methyl esters of N-Cbz-protected α-amino acids to the corresponding butyl esters, in which steric congestion at the α-position and applicability of the functional group were systematically assessed³⁵⁾ (Table 7). Transesterification of the methyl esters of alanine (18i), leucine (18l), and phenylalanine (18m) proceeded smoothly to give the corresponding butyl esters in high yield (Entries 1, 4, 5). In contrast, the reaction rates of methyl esters of valine (18j) and isoleucine (18k), both of which have congested secondary aliphatic side chains, were retarded, and the transformations were suppressed, resulting in rather low yields (Entries 2, 3). These reactions were later improved by the addition of N,N-dimethyl-4-aminopyridine (DMAP) (vide infra). N-Cbz-protected tyrosine methyl ester (18n) reacted without the protection of the hydroxy group (Entry 6). This catalyst system was applied to the transesterification of other N-Cbz-protected amino acid methyl esters bearing an additional protective group, such as methoxy methyl (MOM) ether (18p) (Entry 8), thioether (18q, 18r) (Entries 9, 10), a trityl group on imidazole (18v) (Entry 14), and those with different functionalities such as cyano (18t) (Entry 12) and indole (18u) (Entry 13) groups. Double transesterification of Cbz-aspartic acid α,β-dimethyl ester (18s) with 4.0 equiv. of 16b afforded the corresponding α,β-dibutyl ester 19sb in 91% (Entry 11). Phenylglycine derivatives are racemized much more easily than any natural amino acid esters in peptide coupling reactions conducted under basic conditions because of the higher acidity of the proton at the benzylic chiral center. Zinc catalyst 1b was advantageous in that no racemization was observed for the smooth transesterification of N-Cbz-protected phenylglycine methyl ester 18x with 16b to give 19xb (Entry 16). The reactions of dipeptides also proceeded in good yield with no epimerization of the stereocenters (Entries 17, 18).

Table 7. Transesterification of Various N-Protected α-Amino Esters

Entry	Ester		Yield	Entry	Ester		Yield
1		18i	95%	9		18q	84%
2		18j	41%	10		18r	84%
3		18k	44%	11a)		18s	91%
4		18l	97%	12		18t	79%
5		18m	96%	13		18u	99%
6		18n	87%	14		18v	99%
7		18o	90%	15		18w	97%
8		18p	87%	16		18x	90%
17						18y	92%
18						18z	82%

a) 4 eq of 16b was used.

4.5. Scope and Limitations (3) β-Keto Esters

β-Keto esters are highly useful for various transformations, such as condensation reactions and alkylation, because of their electrophilic and nucleophilic nature; they are thus used as organic building blocks for the synthesis of complex bioactive natural products.⁶⁵⁾ For the synthesis of a wide variety of β-keto esters, transesterification of various combinations of β-keto esters (especially methyl and ethyl esters) with alcohols has received considerable attention. However, the transesterification of β-keto esters is sluggish and requires an excess of β-keto ester and high boiling alcohols.^8–11) In addition, unlike simple esters, the chelation nature of β-keto esters suppresses the activity of metal catalysts through the formation of a coordinate bond to metal ions, and the acidic nature of β-keto esters (pK_a ca. 14 in DMSO) suppresses the activity of the base catalyst. To date, several acidic and basic catalysts have been used for the transesterification of β-keto esters, but the substrate generality of the β-keto esters has much room for improvement. Almost all of the substrates used for the reactions were α-unsubstituted simple β-keto esters. The reactions of sterically more congested α-substituted β-keto esters have not been well studied.

Fortunately, zinc cluster 1b was highly effective for the transesterification of β-keto esters as well.³⁸⁾ Under toluene reflux conditions, the reactions of various primary alcohols, including sterically congested neopentyl alcohol (16k), gave the desired products in high yield (Table 8, Entry 1–3). β-Keto esters of allylic alcohols such as 21ag are difficult to prepare because they readily proceed to Carroll rearrangement. Although the reaction of cinnamyl alcohol (16g) under toluene refluxed conditions led to the formation of the desired allylic ester 21ag in moderate yield, along with 1,3-diphenyl-1-oxo-4-pentene, such a side reaction was efficiently prevented by lowering the reaction temperature under i-Pr₂O refluxed conditions to give 8ae in 86% yield (Entry 4). Propargylic alcohol (16o) and alcohols with a pivaloyl ester (Entry 6) and silylether groups (Entry 7) have not been used for the transesterification of β-keto esters, but they were compatible with this zinc catalysis, affording the products in high yield. The highly acid-sensitive tetrahydropyranyl (THP) ether group was partially decomposed due to the acidity of β-keto ester 20a, but the addition of 10 mol% of DMAP efficiently suppressed the decomposition of THP ether to provide 21ar in 94% yield (Entry 8). Notably, the reactions of sterically more congested secondary alcohols, such as menthol (16n) and (+)-borneol (16s), and the tertiary alcohol adamantanol (16t) also proceeded smoothly (Entries 9–13). Transesterification of tert-butanol (16u) is a difficult task, not only because of its steric bulkiness, but also because it is prone to undesired elimination. Although the reaction was sluggish under the standard reaction conditions, the use of 5 equiv of 16u greatly improved the yield of 21au to 82% (Entry 14).

Table 8. Transesterification of β-Keto Ester with Various Alcohols

Entry	Alcohol (HOR)	Time (h)	Yield (%)
1	16a: HOBn	45	89
2	16c: HOHex	48	86
3	16k: HOCH2-t-Bu	44	89
4a)	16g:	65	86
5	16o:	48	82
6	16p: HO(CH2)6OPiv	60	89
7	16q: HO(CH2)6OTBDMS	48	86
8b)	16r: HO(CH2)6OTHP	48	94
9	16l: HOCHPr2	44	89
10	16m: HO-c-Hex	45	95
11	16n: D-Menthol	44	93
12c)	16s:	44	97
13d)	16t:	72	85
14e)	16u: HO-t-Bu	72	82

a) i-Pr₂O was used as the solvent. b) 10 mol% of DMAP was added. c) 1.5 eq of 16s was used. d) 2.0 eq of 16t was used. e) 5.0 eq of 16u was used.

Next, the substrate scope of β-keto esters 20 was examined³⁸⁾ (Table 9). β-Alkyl substituted β-keto esters, such as methyl acetoacetate (20b) and sterically demanding methyl pivaloylacetate (20c), were also good substrates (Entries 1, 2). While α-methyl substituted β-keto ester 20d gave the product 21dc in only moderate yield, even after 48 h (Entry 3, vide infra), β-keto esters derived from 5- and 6-membered cyclic ketones were efficiently transesterified to 21ec and fc, respectively, in high yield (Entries 4, 5). This zinc catalysis was highly effective for β-keto esters as well as for other active β-dicarbonyl derivatives. In the case of β-diesters, dimethyl malonate (20g) and Meldrum’s acid (20h), both ester groups were transesterified with 2.6 eq of alcohol 16c to give dihexyl malonate (21gc) in good yield (Entries 6, 7). When methyl-N-hexyl malonamide (20i) and trimethyl phosphonoacetate (20j) were used, only the ester groups reacted, leaving the remaining amide and phosphate groups intact (Entries 8, 9).

Table 9. Transesterification of Various β-Dicarbonyl Compounds

Entry	β-Dicarbonyl compound	Time (h)	Yield (%)
1	20b:	48	81
2	20c:	48	81
3	20d:	48	60
4	20e:	42	87
5	20f:	48	97
6a)	20g:	68	98
7a)	20h:	68	73
8	20i:	60	77
9b)	20j:	48	87

a) 2.6 eq of 16c was used. b) 16a was used instead of 16c.

4.6. Catalytic Acetylation Using Ethyl Acetate as the Acetyl Donor

There is another way to look at zinc cluster-catalyzed transesterification: We used ethyl acetate as an acetylation reagent.³¹⁾

The acetylation of alcohols is one of the most important and fundamental reactions in organic synthesis because the O-acetyl moiety is ubiquitous, not only in synthetic intermediates but also in various biologically active natural products and pharmaceutical compounds. In general, the acetylation of hydroxyl groups is conducted with acetyl chloride or acetic anhydride in the presence of greater than stoichiometric amounts of a base, resulting in the formation of greater than stoichiometric amounts of unwanted chemical waste. Alternatively, catalytic transesterification has been applied to acetylation. High conversion, however, is difficult to attain with the direct use of methyl acetate and ethyl acetate for acetylation because of the low electrophilicity of these simple acetates and the existence of a reverse reaction.^66–72) Thus, most reported acetylations by transesterification use enol esters as the acetyl donor to improve reactivity and prevent the reverse reaction. Direct use of methyl acetate and ethyl acetate is limited to only a few examples.^73–88)

We initiated our studies by catalytic acetylation using various acetates (AcOR) as the acetyl donor, and finally decided to use ethyl acetate as the acetyl donor based on its high stability, accessibility, and economic advantages.

The scope and limitations of alcohols 16 were then examined³¹⁾ (Table 10). Benzyl alcohols with various functionalities in the para-position were successfully converted to the corresponding acetates in good to excellent yield (Entries 1–8). Of particular note is that highly acid-sensitive triethyl silyl (TES) ether survived under the reaction conditions (Entry 5). Moreover, the benzoyl (Entry 7) and pivaloyl (Entry 8) groups were not scrambled. Other primary aliphatic alcohols were also suitable substrates for this catalysis (Entries 9–12). Acetylation of the D-glucose derivative 16β, which has both isopropylidene and benzylidene acetal functionalities, proceeded in quantitative yield without cleavage of the acetal protecting groups (Entry 12). The present catalysis was also applicable to the acetylation of secondary alcohols, including those on various steroid compounds 22γ–ζ in which triene and enone functionalities remained intact (Entries 14–17). The reaction of β-estradiol (16ε) resulted in the exclusive formation of monoacetate 22ε in greater than 99% yield (Entry 16). To the best of our knowledge, this is the first example of a highly chemoselective catalytic acetylation of secondary aliphatic alcohols over that of aromatic alcohols through transesterification.

Table 10. Catalytic Acetylation of Various Alcohols

Entry	Alcohol (HOR)	Time (h)	Yield (%)
1	16a: R′=H	18	98
2	16v: R′=Cl	38	76
3	16w: R′=Br	38	81
4	16e: R′=OTBDMS	24	97
5	16x: R′=CH2OTES	18	89
6	16y: R′=CH2OMEM	36	96
7	16z: R′=CH2OBz	40	83
8	16α: R′=CH2OPiv	36	89
9	16j: HOC18H37	38	>99
10	16g:	38	94
11	16h:	24	97
12	16β:	18	>99
13	16n:	18	75
14	16γ:	40	>99
15	16δ:	40	>99
16	16ε:	40	>99a)
17	16ζ:	40	93

a) Yield of the corresponding 17-OAc product.

4.7. Catalytic Deacylation of Acetates and Benzoates with MeOH

Acyl groups, especially acetyl groups, are one of the most common and useful protecting groups in organic synthesis.⁸⁹⁾ Consequently, a number of acylation and deacylation methods have been developed. Although several efficient transesterification catalysts for acetylation have been reported, deacylation still relies largely on classical basic hydrolysis. The disadvantage of this procedure, however, is the low tolerance to many functional groups and the potential occurrence of undesired side reactions, such as elimination and epimerization. Although catalytic cleavage of ester bonds using simple alcohol nucleophiles, such as methanol and ethanol, is a more desirable method, it is not a trivial task; for example, the catalytic activity of most of Lewis acid catalysts decreases in the presence of stoichiometric or excess amounts of alcohol. Low conversion of sterically congested acetates of secondary and tertiary alcohol is another problem.

Because the zinc cluster 1b retains high catalyst abilities even in the presence of excess amounts of alcohol,²⁸⁾ we anticipated that deacetylation, which is the reverse reaction of the above-mentioned acetylation,³¹⁾ would also be catalyzed by 1b simply by changing the solvent from ethyl acetate to methanol. In the presence of 1.25 mol% of 1b, deacetylation of 22η was completed within 2 h under methanol (bp=65°C) reflux conditions³²⁾ (Table 11, Entry 1). This process can be conducted very easily and safely; the deacetylation proceeded simply by heating the acetate in commercial methanol in the presence of a catalytic amount of 1b. Thus, nearly pure product 16η was obtained by simple evaporation and filtration through a short-pad silica gel; methyl acetate was the only co-product of this process. Subsequently, other benzyl acetates with various substituents at the para-position were successfully converted to the corresponding alcohols 16 in good to excellent yield (Entries 2–6). MEM (Entry 2), TBDMS (Entry 3), and THP (Entry 4) ethers were all tolerated, and neither deprotection nor scrambling of the pivaloyl group was observed (Entry 5) due to the steric hindrance of the pivaloyl group. In addition, the reactions of acetates bearing N-Boc (Entry 7) and N-Cbz (Entry 8) groups afforded the corresponding alcohols in quantitative yield without loss of the carbamate-protecting groups. While deacetylation of benzylic acetates often causes elimination problems, the present zinc catalysis was also applicable to 1-indanyl acetate (22ν) (Entry 12) and α-methylbenzyl acetate (22ξ) (Entry 13). Moreover, this zinc catalysis was also effective for debenzoylation of benzoate (BzOR) in the same efficiency as deacetylation.³²⁾

Table 11. Catalytic Deacetylation of Various Acetates

Entry	Acetate (AcOR)	Time (h)	Yield (%)
1	22η: R′=NO2	6	95
2	22y: R′=CH2OMEM	12	97
3	22θ: R′=CH2OTBDMS	12	98
4	22τ: R′=CH2OTHP	12	96
5	22α: R′=CH2OPiv	12	92
6	22κ: R′=CH2NHAc	12	98
7	22λ:	18	>99
8	22μ:	18	>99
9	22j: AcOC17H37	12	96
10	22g:	12	92
11	22h:	12	96
12	22ν:	12	85
13	22ξ:	12	85
14	22n:	96	81
15a)	22γ:	18	83

a) EtOH was used instead of MeOH to be homogeneous.

Deacetylation of tertiary acetates is more challenging. Although we previously found that 1b-catalyzed acetylation of highly congested 1-adamantanol (16t) with ethyl acetate did not give the desired product,³¹⁾ the reverse reaction, namely deacetylation of 22t to 16t with methanol, readily proceeded (86% yield)³²⁾ (Chart 4). Deacetylation of 22π is another challenging reaction because elimination of the acetoxy moiety of 22π readily proceeds. In contrast, under the mild conditions of our proposed zinc catalysis, the Zn cluster 1b catalyzed the deacetylation of 22π without any undesired elimination reaction, leading to the formation of 16π in 57% yield (or 93% yield of the recovered starting material). It is noteworthy that, although the catalytic activity for sterically congested substrates is not yet highly satisfactory, the present reaction is, to our knowledge, the most effective catalytic deacetylation of tertiary acetates through transesterification.

Chart 4. Catalytic Deacetylation of Tertiary Acetates

4.8. Practical Large-Scale Reactions

The applicability of 1b-catalyzed transesterification to large-scale synthesis was demonstrated by synthesizing isopentyl pentanoate (17tσ), which is well-known to have an apple flavor, under solvent-free conditions³⁰⁾ (Chart 5). In the presence of 0.8 mol% of the zinc cluster 1b, the reaction of 133 mL (1.0 mol) of methyl pentanoate (3t) and 131 mL (1.2 mol) of isopentanol (16σ) at 100°C under solvent-free conditions was completed within 43 h, and the resulting mixture was directly purified by distillation to give the pure product 17tσ in 80% yield. The E-factor value,^90–93) an assessment of the waste generation and environmental impact of chemical manufacturing processes, of this reaction was only 0.66, indicating the high environmental and economical advantage of the catalytic transesterification presented here.

Chart 5. Catalytic Deacetylation of Tertiary Acetates

4.9. Chemoselective Acylation of Alcohols over Amines

Based on our proposed mechanism of catalytic oxazoline formation, we expected that zinc cluster 1b would also catalyze transesterification and ester-amide exchange reactions. Indeed, as mentioned above, 1b was found to be a highly efficient catalyst for transesterification. On the other hand, when we applied this zinc catalysis condition to amidation using amine as a nucleophile instead of alcohol, surprisingly, almost no reaction proceeded. This was an unsatisfactory result, but it inspired us to examine the following chemoselective reactions.

Esters and amides are ubiquitous functional groups in natural and synthetic organic compounds, and they are commonly synthesized by acylation of the corresponding alcohol and amine, respectively, with carboxylic acid, acid chloride, or acid anhydride. As the nucleophilicity of the amino group is much greater than that of the hydroxyl group, the amine can be selectively acylated to give the corresponding amide, even in the presence of excess alcohol and/or water. This chemoselectivity has been well utilized for several highly efficient amidation reactions, such as the Schotten–Baumann reactions⁹⁴⁾ (Chart 6, path a, 23→24). On the other hand, selective O-acylation (path b, 23→25) is quite difficult to perform under ordinary organic reactions. When aminoester 25 was targeted, the only feasible route was an indirect protection-deprotection process including in situ protection with acid (path c). The requirement for such a multistep transformation decreases the atom-economy of this process. To comply with the demand for an environmentally benign process, reversing the normal chemoselectivity of path a and minimizing waste are very important. The ideal method leading to the development of a new transformation without using protecting groups is a direct catalytic conversion of aminoalcohol 23 to aminoester 25 in a highly chemoselective manner (path b); however, there were no examples of such a reaction using an artificial catalyst.

Chart 6. Acylation of Amino Alcohol 23

We first investigated selective O-acylation using a 1 : 1 mixture of cyclohexanol (16m) and cyclohexylamine (28m)²⁹⁾ (Chart 7). When PhCOCl or (PhCO)₂O was used as an acylation reagent with triethylamine, the acylation of amine 28m proceeded exclusively to give the corresponding N-cyclohexylbenzamide (29am) in >99% yield, and cyclohexyl benzoate (17am) was not detected (Eqs. 1, 2), consistent with normal chemoselectivity. To tackle the issue of chemoselectivity, we focused on transesterification using the tetranuclear zinc cluster 1b as a catalyst. Although so-called monomeric zinc complexes Zn(OCOR)₂ showed only moderate reactivity, when 1b was used for this chemoselective acylation reaction, alcohol 16m was selectively acylated to afford ester 17am in 96% yield, along with only 1% of amide 29am (Eq. 3).

Chart 7. Chemoselective Acylation Using a 1 : 1 Mixture of Alcohol and Amine

We next performed the selective O-acylation using various combinations of alcohols and amines. All of the reactions that we evaluated using primary amines proceeded in a highly chemoselective manner.²⁹⁾ When the reaction was performed in the presence of cyclic secondary amines, such as pyrolidine, we observed some amide formation due to the higher nucleophilicity of the secondary amines, but the desired ester was still obtained in reasonably high yield. In addition, under the optimized condition, aromatic esters with electron-donating and electron-withdrawing substituents, α, β-unsaturated esters, and aliphatic esters were selectively converted to the corresponding cyclohexyl esters in high yield (94–99%), accompanied by only trace amounts of amides. To demonstrate the usefulness and effectiveness of this zinc catalysis in organic synthesis, we performed the O-selective acylation of amino alcohols 23 (Table 12). When β-amino alcohol 23a was used as a substrate, hydroxyamide 24aa was obtained in 77% yield along with diacylation products 30aa in 23% yield (Entry 1). The product 24aa is produced through O-acylation (23a→25aa), with a subsequent complete O→N acyl transfer reaction (25aa→24aa) due to the instability of the resulting amino ester 25aa. When amino alcohols 23b–d tethered by long alkyl chains were treated, we obtained amino esters 25b–d in good yield (82–90%) (Entries 2–4). Furthermore, the reaction of trans-4-aminocyclohexanol (23e) provided amino ester 25ae exclusively (99%), presumably due to trans-stereochemistry preventing the intramolecular O→N acyl transfer reaction. Even when amino alcohols 23f and g with highly nucleophilic secondary amino groups (piperidine unit) were used, the reactions proceeded in an O-acylation selective manner to give the corresponding amino esters 25 in high yield (88, 92%).

Table 12. Chemoselective Acylation of Various Amino Alcohols

Entry	Aminoalcohol	25 (%)	24 (%)	30 (%)
1	23a:	ND	41	10
2	23b: H2N-(CH2)6-OH	82	ND	18
3	23c: H2N-(CH2)8-OH	90	ND	7
4	23d: H2N-(CH2)10-OH	90	ND	7
5a)	23e:	99	ND	ND
6a)	23f:	88	ND	17
7a)	23g:	92	ND	7

a) Toluene was used as the solvent instead of i-Pr₂O.

We also examined the O-selective acetylation reactions of amino alcohols.³¹⁾ Current acetylation conditions require a higher level of chemoselectivity because excess ethyl acetate is used as both the acetyl donor and solvent (17 equiv). To our surprise, O-acetylation proceeded quite rapidly compared with N-acetylation. For example, O-selective acetylation of 4-piperidinemethanol (23f), bearing both primary hydroxyl and secondary amino groups, with ethyl acetate (17 equiv) was efficiently catalyzed by the zinc cluster 1b to exclusively provide the corresponding O-acetylated product 31f (Chart 8). It is noteworthy that such high chemoselectivities were realized even when a large excess of the acetyl donor was used.

Chart 8. Chemoselective Acetylation of β-Amino Alcohol

Based on these results, we anticipated that this selectivity allowed for the transesterification of amino acid esters bearing a primary or secondary aliphatic amino group. In fact, transesterification of N-protection-free valine methyl ester (33) with 1-butanol (16b) afforded the butyl ester 34 in 69% yield upon treatment with CbzCl³⁵⁾ (Chart 9). As a substrate with a secondary aliphatic amino group, the reaction of N-benzyl glycine methyl ester (36) gave the corresponding butyl ester 37 in 89% yield. Acylation of the amino group was not observed in either reaction.

Chart 9. Transesterification of α-Amino Esters Bearing a Primary or Secondary Aliphatic Amino Group

5. Additive Effect of N-Heteroaromatics

5.1. DMAP

Transesterification of sterically demanding substrates remains a difficult and challenging task. During the course of our studies on the above-mentioned chemoselective acylation of hydroxyl groups over amino groups, we discovered important clues to solving this problem. In the presence of an equivalent amount of cyclohexylamine (28m), for example, the acylation of cyclohexyl alcohol (16m) with methyl 3-phenylpropanoate (3j) provided the corresponding ester 17jm in 94% yield after refluxing for 18 h, whereas in the absence of 28m, the reaction provided the same product but in only 22% yield, even after refluxing for 48 h³³⁾ (Chart 10). These results clearly indicate that under this acylation condition, highly nucleophilic alkylamine, which is in general acylated in preference to alcohol, greatly accelerates the acylation of alcohol without being converted to the corresponding amide. Further, the addition of only a catalytic amount of alkylamine (20 mol%) was sufficient to achieve satisfactory catalyst activity.

Chart 10. Effects of Amine 28m on Transesterification

Encouraged by such drastic additive effects of 28m, we examined a variety of amine additives.³³⁾ Other sterically-unhindered primary amines, such as 1-hexylamine, also had significantly positive effects. Secondary alkylamines were also good additives, but tertiary and aromatic amines had only limited effect. Based on the fact that metal ions in the active site of some metalloenzymes, such as aminopeptidase, are supported by both carboxylate and imidazole ligands, we further examined heteroaromatics as additives, and found that N-methylimidazole (NMI), DMAP and 4-pyrrolidinopyridine were highly effective.³³⁾ We performed the transesterification with various amounts of DMAP; yield of the product dramatically increased with an increase in DMAP in the range of 0 to 20 mol%. Although the reaction with 40 mol% of DMAP gave a slightly better result, we determined 20 mol% loading of DMAP to be an optimized condition in terms of atom economy. The time–course for the reaction in the presence of the best alkylamine-type additive 28m and the best N-heterocycle-type additive DMAP revealed that the addition of 20 mol% of these additives increased the initial rate of the reactions more than 3- and 15-fold, respectively (Fig. 2). The addition of these additives at any stage of the reaction sufficiently accelerated the reaction.

Fig. 2. Time–Course for the 1b-Catalyzed Transesterification of 3a with 16a in the Presence of 20 mol% of 28m (▲), 20 mol% of DMAP (■) or in the Absence of Additives (●)

With the best additive, DMAP, in hand, we investigated the substrate generality of the transesterification of comparatively low reactive esters, as well as alcohols.³³⁾ As shown in Table 13, DMAP had drastic positive effects on the reaction, resulting in great improvement of the chemical yield (up to 98%). The reactions of various methyl esters, including highly congested methyl 1-adamantanecarboxylate (3w), with 1-butanol (16b) in the presence of 20 mol% DMAP resulted in satisfactory yields, though sluggish reactions were observed for the corresponding substrates in the absence of the additive (Entries 1–4). The DMAP additive system was also superior for the less reactive alcohols (Entries 5–8). Notably, the addition of DMAP not only accelerated the catalytic rate but also suppressed undesirable side reactions because of the neutral reaction conditions; the reaction of 3a with 1-indanol (16ν) afforded the product 1-indanyl 3-phenylpropanoate (17aν) in 88% yield despite observations that the same reaction conducted in the absence of DMAP resulted in a complex mixture of decomposed products from 16ν via a carbocation intermediate (Entry 7).

Table 13. Additive Effects of DMAP on 1b-Catalyzed Transesterification: Scope and Limitations

Entry	Ester	Alcohol	Solvent	Time (h)	Yield (%) x=0	Yield (%) x=20
1	3u	16b	Toluene	48	2	89
2	3v	16b	Toluene	18	7	89
3	3q	16b	i-Pr2O	18	25	98
4	3w	16b	Toluene	90	<1	94
5	3a	16a	Toluene	18	58	92
6	3a	16l	i-Pr2O	48	38	93
7	3a	16ν	i-Pr2O	18	5	88
8	3a	16n	i-Pr2O	72	23	83

The addition of DMAP (20 mol%) also accelerated the transesterification of less reactive α-amino esters, increasing the yield of 19jb from 41% (Table 7, Entry 2) to 95%, and that of 19kb from 44% (Entry 3) to 90% yield, respectively.

5.2. Bis(imidazole) Ligand

Very recently, we developed a highly stable but reactive bis(imidazole)/zinc catalyst 39 (Fig. 3) for transesterification based on the results of these additive effects.³⁷⁾ This new zinc complex 39 catalyzed transesterification reactions with enhanced efficiency (Table 14). Tertiary alcohol, previously not an applicable substrate for the transesterification of simple esters, could also be used, and the corresponding product was isolated in high yield. Labile acrylate and methacrylate could be transesterificated in preference to 1,4-addition, indicating that the present zinc complex catalysis could be useful for the synthesis of various monomers using readily available methyl acrylate and methacrylate.

Fig. 3. Highly Stable Crystalline Zinc Complex 39a and X-Ray Crystallographic Structure

Table 14. Transesterification of Various Methyl Esters

Structures of Products 17 94% yield (Condition A, PhCl)		99% yield (Condition A, PhCl)
90% yield (Condition A, PhCI)	90% yield (Condition A, PhCI)		98% yield (Condition A, PhCI)
98% yield (Condition B, xylene)	82% yielda) (Condition B, HCO2Et)		94% yield (Condition B, toluene)
93% yield (Condition B, toluene)		93% yield (>99% ee) (Condition B, i-Pr2O)
86% yield (Condition B, i-Pr2O)		76% yield (Condition B, i-Pr2O)

a) HCO₂Et was used as an electrophile instead of HCO₂Me.

Table 15. Transesterification Using Dimethyl or Diethyl Carbonates

Zinc complex 39a also catalyzed the transesterification of carbonates, and various linear carbonates 41, cyclic carbonates 43, and cyclic oxazolidinones 45 were synthesized in high yield using dimethyl or diethyl carbonate 40³⁷⁾ (Table 15). In addition, zinc complex 39a also catalyzed the chemoselective transesterification of unprotected-amino alcohols using dimethyl carbonate with better chemoselectivity than Zn₄(OCOCF₃)₆O (1b) (O/N=ca. 2/1) (Chart 11). To the best of our knowledge, this is the first example of chemoselective transesterification using dimethyl carbonate.

Chart 11. Chemoselective Transesterification of Amino Alcohol Using Dimethyl Carbonate

Zinc complex 39a can be prepared in multigram scale using a simple procedure. A water addition experiment revealed the high stability of zinc complex 39a compared to 1b. The stable nature of zinc complex 39a allowed for its recovery and reuse in transesterification reactions³⁷⁾ (Chart 12). Using methyl benzoate (3a) and benzyl alcohol (16a) as model substrates, the transesterification reaction was performed in gram scale. After refluxing the reaction mixture for 5 h under ambient air atmosphere, volatiles, including the desired product 17aa, were removed under reduced pressure. The residue was washed with n-hexane, and the remaining zinc complex was subjected to the next reaction cycle. The recovered zinc complex was reused five times without a significant loss of catalytic activity (average over 95% yield). Even after five runs, zinc complex 39a could be recovered in 91% yield without any decomposition. Alternatively, zinc complex 39a can be recovered and reused using simple deposition and decantation techniques.

Chart 12. Recovery and Reuse Experiment Using Zinc Complex 39a

6. Mechanistic Studies

6.1. Effects of DMAP Additive

To gain insight into the mechanism of 1b-catalyzed transesterification, the effects of additives such as DMAP, and the origin of chemoselectivity, we performed several mechanistic studies.^33,36,38)

First, kinetic data, electrospray ionization (ESI)-MS analysis, and NMR studies on 1b-catalyzed transesterification indicated that DMAP coordinates with zinc ions up to a certain ratio of DMAP to Zn, and stabilizes more catalytically active clusters with lower nuclearities.³³⁾ During these studies, we also discovered a drastic cluster effect involved in accelerating the reaction rate using 1b. To obtain high catalytic activity, therefore, we need a cluster structure, trifluoroacetate ligands, and DMAP additive.³³⁾

6.2 Electronic Effects of Tetranuclear Zinc Cluster

Next, to gain further insight into the unique catalyst systems of 1b and to develop more efficient catalysts, we intensively studied the electronic effects of zinc cluster catalysts by changing their carboxylate ligands. Thus, we synthesized a series of tetranuclear zinc clusters, Zn₄(OCOR)₆O (1a: R=CH₃, 1b: R=CF₃, 1c: R=CF₂CF₃, 1d: R=CF₂CF₂CF₃, 1e: R=CF₂CH₂CH₃)³⁸⁾ (Fig. 1). Zinc clusters 1a–e were then applied to the catalytic transesterification of several combinations of methyl esters 3 and alcohols 16. Depending on the pK_a values of the corresponding carboxylic acids, the Lewis acidities of the zinc clusters 1 should increase in the following order: 1a<1e<1b<1c<1d. The results of all examined substrate combinations revealed that the Lewis acidity of zinc ions severely affected the catalytic activity of the zinc clusters in transesterifications. Yields of the products tended to initially increase and then decrease with an increase in the Lewis acidity of 1, and moderately Lewis acidic 1b (for aromatic ester) and/or 1c (for aliphatic ester) afforded the best results. In general, an increase in Lewis acidity improves the catalytic activity of Lewis acid catalysts. In the zinc cluster system, however, either an increase or decrease in Lewis acidity from 1b and c decreased the catalytic activity. These findings suggest that a balance between Lewis acidity and Brønsted basicity of the catalyst is essential to achieving high catalyst activity, consistent with the dual activation of the electrophile (ester) and nucleophile (alcohol) by the cooperative zinc centers (Fig. 4). Higher Lewis acidity increases the electrophilicity of the coordinated carbonyl group of esters, thereby facilitating the transesterification, whereas the nucleophilicity of the zinc alkoxide moieties is reduced.

Fig. 4. Dual Activation of Ester and Alcohol by Zinc Cluster 1

In contrast to transesterification, the catalytic activity of zinc cluster 1 in ester-amide exchange reactions had a somewhat opposite tendency; moderately Lewis acidic zinc clusters 1b and c afforded a lower yield, while less acidic 1a and more acidic 1d afforded a higher yield, indicating that zinc cluster 1-catalyzed amidation proceeded through a reaction mechanism that differed from the cooperative mechanism of the 1-catalyzed transesterification.³⁸⁾

The results shown in Table 9, as well as in previous reports, highlight the issue of the low reactivity of sterically congested α-substituted and α,α-disubstituted β-keto esters in transesterification. To address this issue, we further optimized the reaction conditions based on the carboxylate ligand effects. The slightly more Lewis acidic catalyst 1c, compared with 1b, afforded better results for both β-keto esters 20d and k³⁸⁾ (Chart 13). Thus, this new zinc catalyst 1c is especially useful for the reaction of β-keto esters with an acidic α-proton such as 20d, whereas a basic additive/catalyst was not effective for these substrates.

Chart 13. Transesterification of Less Reactive β-Keto Esters

6.3. Mechanistic Studies Using Cobalt Complexes

After screening various metal carboxylates for chemoselective acylation of alcohols over amines, we found that a wide variety of first-row transition metal carboxylates serve as O-selective catalysts, and identified a new catalyst system based on Co(II)-carboxylate clusters that function as O-selective acylation catalysts.³⁶⁾ We selected Co(II) as the best metal to study the mechanistic details of O-selective acylation, not only because of its high catalytic activity and chemoselectivity, but also its potential advantages based on UV-Vis spectroscopic characterization. In addition, a pivalate bridged octanuclear cluster, Co₈(OCO-t-Bu)₁₂O₂ (46),⁹⁵⁾ has the dimeric structure of a μ-oxo-tetranuclear unit analogous to zinc clusters 1. Based on various experimental results, UV-Vis spectroscopic analysis, kinetic studies, and density functional theory (DFT) calculations, a schematic of the possible catalytic cycle for transesterification using cobalt cluster 46 is shown in Chart 14 (figures in parentheses are energies calculated using the B3LYP-D method in kcal/mol). Transesterification catalyzed by the combined system of 46 and 2,2′-bipyridine proceeded through the formation of the key intermediate alkoxide-bridged cobalt dinuclear complex 47. The isolated 47 exhibited excellent catalytic activity (TON=1120 and TOF=224 h⁻¹), suggesting that alkoxide cobalt complex 47 is the active species in this catalytic cycle and performs enzyme-like catalysis following the Michaelis–Menten mechanism (K_m,alcohol=4.51×10⁻² M and K_m,ester=1.09×10⁻¹ M) via a ternary-complex formation similar to that of dinuclear metallo-enzymes. We also confirmed that the advantages of cluster complexes as catalyst precursors include the deprotonation ability of basic μ-oxo moieties in the cobalt cluster, and stabilization of the alkoxy ligand by bridged dinuclear complexation. Deprotonation of nucleophiles was thus the most important step, not only for achieving high catalytic activity but also for determining chemoselectivity, resulting in the chemical differentiation of alcohols and amines.

Chart 14. Possible Reaction Mechanism for Transesterification Catalyzed by Cobalt Cluster

7. Conclusion

As presented here, we developed several environmentally friendly reactions using μ-oxo-tetranuclear zinc cluster as the catalyst. This zinc cluster efficiently catalyzed the direct formation of oxazolines from esters, carboxylic acids, lactones, and nitriles; the transesterification of various methyl esters including α-amino esters and β-keto esters; the acetylation of alcohols in EtOAc; and the deacylation of esters in MeOH. Because the reaction conditions of this method are almost neutral, this zinc catalysis has broad substrate generality wherein even in acid sensitive functional groups, such as THP ether and TES ether, catalysis persisted, and side reactions such as isomerization, cyclization, 1,4-addition, did not occur. Moreover, we achieved unprecedented O-selective acylations in the presence of much more nucleophilic primary and secondary alkyl amino groups.

In general, to reverse the innate reactivity of amino and hydroxyl groups, the protection of more reactive amino groups has been necessary, resulting in the formation of more than stoichiometric amounts of chemical waste, which reduces the reaction efficiency. The catalytic O-selective acylation reactions described here clearly indicate that such innate chemoselectivity is reversible in a catalytic manner. Our intensive mechanistic studies revealed that the transesterification proceeds with Michaelis–Menten behavior through an ordered ternary complex mechanism similar to dinuclear metallo-enzymes, suggesting that the formation of metal alkoxides, followed by coordination of the ester, is responsible for the unique O-selective acylation. Because of the high efficiency in terms of atom economy and step economy, these direct new catalyses will be a powerful tool for organic synthesis, and I hope that the findings discussed herein will facilitate the development of new environmentally friendly reactions, including new catalyst-controlled chemoselective reactions.⁹⁶⁾

Acknowledgment

First of all, I would like to express my heartfelt respect and gratitude to Prof. K. Mashima of Osaka University for his kind and valuable advice and encouragement during these studies. The research reviewed in this paper was possible only through the dedication, enthusiasm, and creativity of scores of co-workers: Dr. T. Iwasaki, Dr. Y. Maegawa, Miss A. Yoshiyama, Dr. Y. Hayashi, Prof. R. Dembinski (Oakland University), Dr. K. Agura, Miss Y. Fujii, Dr. Y. Matsushima (Takasago Co.), Dr. S. Santoro (Stockholm University), Prof. F. Himo (Stockholm University), Mr. D. Nakatake, Miss Y. Yokote, Dr. R. Yazaki, and Miss M. Wada. These works received the support from the following: Encouragement of Young Scientists (A) from the Japan Society for the Promotion of Science (JSPS), a Grant-in-Aid for Science Research in a Priority Area (Chemistry of Concerto Catalysis) from Japan’s Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, a Grant-in-Aid for Scientific Research (B) (21390003) from MEXT, Core Research for Evolutional Science and Technology (CREST) from The Japan Science and Technology Agency (JST), Research for Promoting Technological Seeds (JST), a Grant-in-Aid for Scientific Research and Scientific Research on Innovative Areas (Integrated Organic Synthesis) from MEXT, a Grant-in-Aid for Scientific Research (B) (#24390004), Scientific Research on Innovative Areas (Middle molecular strategy), Platform for Drug Discovery, Informatics, and Structural Life Science, all from MEXT; and from the following foundations: the Uehara Memorial Foundation, Sumitomo Foundation, Hoh-ansha Foundation, Mitsubishi Chemical Corporation Foundation, and Kurata Grants.

Conflict of Interest

The author declares no conflict of interest.

References

• 1) Nicolaou K. C., Sorensen E. J., “Classics in Total Synthesis,” Wiley-VCH, Weinheim, 1996.
• 2) Nicolaou K. C., Snyder S. A., “Classics in Total Synthesis II,” Wiley-VCH, Weinheim, 2003.
• 3) Nicolaou K. C., Montagnon T., “Molecules That Changed the World,” Wiley-VCH, Weinheim, 2008.
• 4) Nicolaou K. C., Chan J. S., “Classics in Total Synthesis III,” Wiley-VCH, Weinheim, 2010.
• 5) Trost B. M., Science, 254, 1471–1477 (1991).
• 6) Jørgensen L., McKerrall S. J., Kuttruff C. A., Ungeheuer F., Felding J., Baran P. S., Science, 341, 878–882 (2013).
• 7) Mukaiyama T., Ishikawa H., Koshino H., Hayashi Y., Chemistry, 19, 17789–17800 (2013).
• 8) Otera J., Chem. Rev., 93, 1449–1470 (1993).
• 9) Hoydonckx H. E., De Vos D. E., Chavan S. A., Jacobs P. A., Top. Catal., 27, 83–96 (2004).
• 10) Grasa G. A., Singh R., Nolan S. P., Synthesis, 2004, 971 (2004).
• 11) Enders D., Niemeier O., Henseler A., Chem. Rev., 107, 5606–5655 (2007).
• 12) Higgins N. A., US-Patent 2676945 (1954).
• 13) Mehrotra R. C., J. Am. Chem. Soc., 76, 2266–2267 (1954).
• 14) Chimura K., Takashima S., Kawashima M., Shimoshinbara Y., JP-Patent 48049740 (1973).
• 15) Seebach D., Hungerbühler E., Naef R., Schnurrenberger P., Weidmann B., Züeger M., Synthesis, 1982, 138–141 (1982).
• 16) Otera J., Yano T., Kawabata A., Nozaki H., Tetrahedron Lett., 27, 2383–2386 (1986).
• 17) Otera J., Danoh N., Nozaki H., J. Org. Chem., 56, 5307–5311 (1991).
• 18) Otera J., Dan-Oh N., Nozaki H., J. Chem. Soc., Chem. Commun., 1991, 1742 (1991).
• 19) Hobbs L. A., Smith P. J., Appl. Organomet. Chem., 6, 95–100 (1992).
• 20) Krasik P., Tetrahedron Lett., 39, 4223–4226 (1998).
• 21) Chen J.-W., Chen L.-W., J. Polym. Sci. A Polym. Chem., 37, 1797–1803 (1999).
• 22) Xiang J., Toyoshima S., Orita A., Otera J., Angew. Chem. Int. Ed., 40, 3670 (2001).
• 23) Chen C.-T., Kuo J.-H., Ku C.-H., Weng S.-S., Liu C.-Y., J. Org. Chem., 70, 1328–1339 (2005).
• 24) Magens S., Plietker B., J. Org. Chem., 75, 3715–3721 (2010).
• 25) Weng S.-S., Ke C.-S., Chen F.-K., Lyu Y.-F., Lin G.-Y., Tetrahedron, 67, 1640–1648 (2011).
• 26) Hatano M., Furuya Y., Shimmura T., Moriyama K., Kamiya S., Maki T., Ishihara K., Org. Lett., 13, 426–429 (2011).
• 27) Hatano M., Ishihara K., Chem. Commun., 49, 1983 (2013).
• 28) Ohshima T., Iwasaki T., Mashima K., Chem. Commun., 2006, 2711 (2006).
• 29) Ohshima T., Iwasaki T., Maegawa Y., Yoshiyama A., Mashima K., J. Am. Chem. Soc., 130, 2944–2945 (2008).
• 30) Iwasaki T., Maegawa Y., Hayashi Y., Ohshima T., Mashima K., J. Org. Chem., 73, 5147–5150 (2008).
• 31) Iwasaki T., Maegawa Y., Hayashi Y., Ohshima T., Mashima K., Synlett, 10, 1659 (2009).
• 32) Iwasaki T., Agura K., Maegawa Y., Hayashi Y., Ohshima T., Mashima K., Chemistry, 16, 11567–11571 (2010).
• 33) Maegawa Y., Ohshima T., Hayashi Y., Agura K., Iwasaki T., Mashima K., ACS Catal., 1, 1178–1182 (2011).
• 34) Hayashi Y., Ohshima T., Fujii Y., Matsushima Y., Mashima K., Catal. Sci. Technol., 1, 230 (2011).
• 35) Maegawa Y., Agura K., Hayashi Y., Ohshima T., Mashima K., Synlett, 23, 137 (2012).
• 36) Hayashi Y., Santoro S., Azuma Y., Himo F., Ohshima T., Mashima K., J. Am. Chem. Soc., 135, 6192–6199 (2013).
• 37) Nakatake D., Yokote Y., Matsushima Y., Yazaki R., Ohshima T., Green Chem., 18, 1524–1530 (2016).
• 38) Agura K., Hayashi Y., Wada M., Nakatake D., Mashima K., Ohshima T., Chem. Asian J., accepted. doi: 10.1002/asia.201600062.
• 39) Sniady A., Durham A., Morreale M. S., Marcinek A., Szafert S., Lis T., Brzezinska K. R., Iwasaki T., Ohshima T., Mashima K., Dembinski R., J. Org. Chem., 73, 5881–5889 (2008).
• 40) Yang Y., Hayashi Y., Fujii Y., Nagano T., Kita Y., Ohshima T., Okuda J., Mashima K., Catal. Sci. Tech., 2, 509–513 (2012).
• 41) Auger V., Robin I., Compt. Rend., 178, 1546 (1924).
• 42) Wyart J., Bull. Soc. Fr. Mineral., 49, 148 (1926).
• 43) Koyama H., Saito Y., Bull. Chem. Soc. Jpn., 27, 112–114 (1954).
• 44) Gordon R. M., Silver H. B., Can. J. Chem., 61, 1218–1221 (1983).
• 45) Clegg W., Harbron D. R., Homan C. D., Hunt P. A., Little J. R., Straughan B. P., Inorg. Chim. Acta, 186, 51–60 (1991).
• 46) Ming-cai Y., Chi-wei M., Chang-chun A., Liang-jie Y., Ju-tang S., Wuhan University Journal of Natural Sciences, 9, 939–942 (2004).
• 47) Hermes S., Witte T., Hikov T., Zacher D., Bahnmüller S., Langstein G., Huber K., Fischer R. A., J. Am. Chem. Soc., 129, 5324–5325 (2007).
• 48) Berkesi O., Dreveni I., Andor A., Goggin P. L., Inorg. Chim. Acta, 181, 285–289 (1991).
• 49) Eddaoudi M., Kim J., Rosi N., Vodak D., Wachter J., O’Keeffe M., Yaghi O. M., Science, 295, 469–472 (2002).
• 50) Yaghi O. M., O’Keeffe M., Ockwig N. W., Chae H. K., Eddaoudi M., Kim J., Nature (London), 423, 705–714 (2003).
• 51) Bergeron R. J., Chem. Rev., 84, 587–602 (1984).
• 52) Roy R. S., Gehring A. M., Milne J. C., Belshaw P. J., Walsh C. T., Roy R. S., Gehring A. M., Milne J. C., Belshaw P. J., Walsh C. T., Nat. Prod. Rep., 16, 249–263 (1999).
• 53) Lewis J. R., Nat. Prod. Rep., 19, 223–258 (2002).
• 54) Jin Z., Nat. Prod. Rep., 20, 584–605 (2003).
• 55) Reuman M., Meyers A. I., Tetrahedron, 41, 837–860 (1985).
• 56) Gant T. G., Meyers A. I., Tetrahedron, 50, 2297–2360 (1994).
• 57) Gómez M., Muller G., Rocamora M., Coord. Chem. Rev., 193-195, 769–835 (1999).
• 58) Meyers A. I., J. Org. Chem., 70, 6137–6151 (2005).
• 59) Witte H., Seeliger W., Angew. Chem. Int. Ed. Engl., 11, 287–288 (1972).
• 60) Bolm C., Weickhardt K., Zehnder M., Ranff T., Chem. Ber., 124, 1173–1180 (1991).
• 61) Lowenthal R. E., Abiko A., Masamune S., Tetrahedron Lett., 31, 6005–6008 (1990).
• 62) Zhou P., Blubaum J. E., Burms C. T., Natale N. R., Tetrahedron Lett., 38, 7019–7020 (1997).
• 63) Ohshima T., Kawabata T., Takeuchi Y., Kakinuma T., Iwasaki T., Yonezawa T., Murakami H., Nishiyama H., Mashima K., Angew. Chem. Int. Ed., 50, 6296–6300 (2011).
• 64) Morisaki K., Sawa M., Nomaguchi J.-y., Morimoto H., Takeuchi Y., Mashima K., Ohshima T., Chemistry, 19, 8417–8420 (2013).
• 65) Benetti S., Romagnoli R., De Risi C., Spalluto G., Zanirato V., Chem. Rev., 95, 1065–1114 (1995).
• 66) Nishiguchi T., Taya H., J. Am. Chem. Soc., 111, 9102–9103 (1989).
• 67) Iranpoor N., Shekarriz M., Bull. Chem. Soc. Jpn., 72, 455–458 (1999).
• 68) Orita A., Sakamoto K., Hamada Y., Mitsutome A., Otera J., Tetrahedron, 55, 2899–2910 (1999).
• 69) Ranu B. C., Dutta P., Sarkar A., J. Chem. Soc., Perkin Trans. 1, 2000, 2223–2225 (2000).
• 70) Habibi M. H., Tangestaninejad S., Mirkhani V., Yadollahi B., Tetrahedron, 57, 8333–8337 (2001).
• 71) Singh R., Kissling R. M., Letellier M.-A., Nolan S. P., J. Org. Chem., 69, 209–212 (2004).
• 72) Tayebee R., Alizadeh M. H., Monatsh. Chem., 137, 1063–1069 (2006).
• 73) Kita Y., Maeda H., Omori K., Okuno T., Tamura Y., Synlett, 1993, 273–274 (1993).
• 74) Ishii Y., Takeno M., Kawasaki Y., Muromachi A., Nishiyama A. R., Sakaguchi S., J. Org. Chem., 61, 3088–3092 (1996).
• 75) Dinh P. M., Howarth J. A., Hudnott A. R., Williams J. M. J., Harris W., Tetrahedron Lett., 37, 7623–7626 (1996).
• 76) Orita A., Mitsutome A., Otera J., J. Org. Chem., 63, 2420–2421 (1998).
• 77) Ilankumaran P., Verkade J. G., J. Org. Chem., 64, 9063–9066 (1999).
• 78) Lin M.-H., RajanBabu T. V., Org. Lett., 2, 997–1000 (2000).
• 79) Grasa G. A., Kissling R. M., Nolan S. P., Org. Lett., 4, 3583–3586 (2002).
• 80) Grasa G. A., Güveli T., Singh R., Nolan S. P., J. Org. Chem., 68, 2812–2819 (2003).
• 81) Bosco J. W. J., Saikia A. K., Chem. Commun., 2004, 1116 (2004).
• 82) Shirae Y., Mino T., Hasegawa T., Sakamoto M., Fujita T., Tetrahedron Lett., 46, 5877–5879 (2005).
• 83) Bosco J. W. J., Agrahari A., Saikia A. K., Tetrahedron Lett., 47, 4065–4068 (2006).
• 84) Kobayashi J., Mori Y., Kobayashi S., Chem. Commun., 2006, 4227 (2006).
• 85) Mino T., Hasegawa T., Shirae Y., Sakamoto A., Fujita T., J. Organomet. Chem., 692, 4389–4396 (2007).
• 86) Magens S., Ertelt M., Jatsch A., Plietker B., Org. Lett., 10, 53 (2008).
• 87) Wong C.-H., Whitesides G. M., “Enzymes in Synthetic Organic Chemistry,” Pergamon, Oxford, 1994.
• 88) Faber K., “Biotransformations in Organic Chemistry,” Springer, Berlin, 2000.
• 89) Green T. W., Wuts P. G. M., “Protective Groups in Organic Synthesis,” 4th ed., John Wiley & Sons, New Jersey, 2006.
• 90) Sheldon R. A., Chem. Ind. (London), 1992, 903 (1992).
• 91) Sheldon R. A., Chem. Ind. (London), 1997, 12 (1997).
• 92) Sheldon R. A., Pure Appl. Chem., 72, 1233 (2000).
• 93) Poliakoff M., Fitzpatrick J. M., Farren T. R., Anastas P. T., Science, 297, 807–810 (2002).
• 94) Sonntag N. O. V., Chem. Rev., 52, 237–416 (1953).
• 95) Sidorov A. A., Fomina I. G., Ponina M. O., Aleksandrov G. G., Nefedov S. E., Eremenko I. L., Moiseev I. I., Russ. Chem. Bull. Int. Ed., 49, 958–959 (2000).
• 96) Uesugi S., Li Z., Yazaki R., Ohshima T., Angew. Chem. Int. Ed., 53, 1611–1615 (2014).