Chapter 1
Lewis Acid-Brønsted Base Catalysis

Masakatsu Shibasaki and Naoya Kumagai

1.1 Introduction

From the synthetic point of view, organic synthesis via catalytic processes offers many benefits. Catalysis frequently obviates the excessive use of the activating reagents and associated tedious purification processes, thereby offering more environmentally benign synthetic processes. Furthermore, the specific activation mode of a catalyst allows for highly chemoselective transformations that are seldom achieved by noncatalytic processes. Over the past two decades, the concept of cooperative catalysts has evolved and subsequently rapidly advanced as the most finely refined class of artificial catalysts for preparative chemistry [1]. The cooperative catalysts exhibit two catalytic functions simultaneously to achieve a dual activation mode to specific substrate(s) (Figure 1.1). The obvious advantage of this activation strategy is not only the significant enhancement of the reaction rate due to intramolecularity or a proximity effect but also the broadened scope of the applicable reactions following the synergistic activation of otherwise unreactive substrate sets.

Figure 1.1 Schematic representation of the Lewis acid-Brønsted base cooperative catalysts.

In this chapter, cooperative catalysts that exhibit Lewis acid and Brønsted base activation modes are reviewed. While recent interest in artificial catalysts focuses on the efficient production of enantioenriched building blocks [2], herein only asymmetric Lewis acid-Brønsted base cooperative catalysts are covered. Metal-based asymmetric cooperative catalysts that display transition-metal catalysis are described in other chapters [3]. In this chapter, the focus is on the reactions promoted by the effective coupling of an in situ generated active nucleophile by a Brønsted base and an electrophile activated by a Lewis acid.

1.2 Lewis Acid-Brønsted Base Catalysis in Metalloenzymes

The essence of Lewis acid-Brønsted base catalysis is the manifestation of two different catalytic functions in a synergistic manner. This often occurs via two different catalytic sites in near proximity - referred to as two-center catalysis. Two-center catalysis involving a Lewis acid and a Brønsted base is largely exploited in metalloenzyme reactions [4, 5]. A typical biological degradation reaction, such as urea hydrolysis promoted by urease, utilizes dinickel two-center cooperative catalysis (Figure 1.2) [4b, 6]. Two Ni(II) cations are located in near proximity at the active site of urease, and one Ni(II) cation is coordinated by urea to electrophilically activate the urea carbonyl. Another Ni(II) cation (Ni hydroxide) functions as a Brønsted base with the aid of the adjacent histidine side chain to produce a nucleophilically active Ni hydroxide. The synergistic activation of both the nucleophile and electrophile provides significantly accelerated hydrolysis. Urea generally does not readily undergo simple basic hydrolysis in organic synthesis, but with the cooperative catalysis of a dinickel active site the reaction rate is enhanced by a factor of 1014. An artificial model of this cooperative hydrolysis has been achieved with a dicopper catalyst comprising a low molecular weight ligand and Cu(II) cations [7].

Figure 1.2 Proposed activation mode in urease.

This type of Lewis acid-Brønsted base cooperative catalysis is operative also in enantioselective carbon-carbon bond-forming processes in biological contexts. Class II aldolase, a Zn-dependent metalloenzyme, illustrates this (Figure 1.3). The aldolase efficiently promotes the enantioselective aldol reaction of dihydroxyacetone phosphate (DHAP) and various aldehydes under virtually neutral conditions [8]. DHAP coordinates to a Zn(II) cation in a bidentate manner to increase the acidity of the a-proton, which is deprotonated by the adjacent glutamic acid-73 residue as a Brønsted base. This cooperation enables the catalytic generation of an active Zn-enolate, which is integrated into the following aldol addition to an aldehyde that is activated by the tyrosine-113 residue by hydrogen bonding. These naturally occurring macromolecular catalytic machineries have inspired chemists to mimic the cooperative activation strategy in artificial catalyst design.

Figure 1.3 Proposed activation mode in Zn-dependent class II aldolase.

Obviously, an inevitable drawback in enzymatic catalysis is its strict substrate specificity at the expense of extraordinary rate enhancement. Artificial cooperative catalysts follow a somewhat loose three-dimensional design of two catalytic functions to acquire both rate enhancement through synergistic activation and sufficient substrate generality to showcase the synthetic utility.

1.3 Hard Lewis Acid-Brønsted Base Cooperative Catalysis

1.3.1 Cooperative Catalysts Based on a 1,1´-Binaphthol Ligand Platform

1.3.1.1 Heterobimetallic Catalysts

A series of hard Lewis acid-Brønsted base cooperative heterobimetallic catalysts utilizing 1,1´-binaphthol and its derivatives as a chiral bidentate ligand were developed by Shibasaki et al. [9] (Figure 1.4). Depending on the nature of the central metal cation [rare earth metal (RE) or group 13 metal (M(13))], two general types of cooperative catalysts are generated [10]. By combining RE and alkali metals (M(1)), heterobimetallic catalysts of the general formula RE-M3-tris(1,1´-binaphthoxide) (type 1) are formed. Following the initial identification of La-Li3-tris(1,1´-binaphthoxide) (RE = La, M(1) = Li, abbreviated as LLB) in the first report on the catalytic asymmetric nitroaldol reaction [10a-12] (Scheme 1.1), several heterobimetallic catalysts emerged by changing the combination of RE (Y, La, Pr, Sm, Yb) and M (Li, Na, K) to promote a wide range of catalytic asymmetric transformations (Figure 1.5) [13-26].1 Irrespective of the combination, a highly symmetrical architecture of RE-M3-tris(1,1´-binaphthoxide) is maintained (based on laser desorption/ionization time-of-flight mass spectrometry data). Some of the heterobimetallic catalysts, such as LSB (RE = La, M(1) = Na), PrSB (RE = Pr, M(1) = Na), NdSB (RE = Nd, M(1) = Na), and EuSB (RE = Eu, M(1) = Na), were unequivocally characterized by X-ray crystallographic analysis [10b, 13, 27].

Figure 1.4 Two types of Lewis acid-Brønsted base cooperative heterobimetallic catalysts based on 1,1´-binaphthol and its derivatives as a chiral ligand platform.

Scheme 1.1 Seminal nitroaldol reaction promoted by the heterbimetallic catalyst LLB.

Figure 1.5 Schematic representation of the utility of RE-M(1)3-tris(1,1´-binaphthoxide) cooperative catalysts in catalytic asymmetric transformations.

Although these complexes have a chiral center at the central RE, a 1,1´-binaphthol unit existed only in the ? configuration, presumably because of the higher thermodynamic stability. Biphenyldiols were also exploited to constitute similar catalyst architecture for some reactions. The essence of this catalytic system is the cooperative function of RE as the Lewis acid to activate electrophiles and M(1)-1,1´-binaphthoxide as the Brønsted base to activate pronucleophiles, allowing for the subsequent facilitated bond formation in the chiral environment. The coordination number of RE generally ranges from 6 to 12 [28]. Hence, the central RE of these complexes is not coordinatively saturated, and it is anticipated that it accepts the additional coordination of electrophiles. Coordination to the RE center of these complexes has been of interest [29], and direct evidence to prove the coordination of Lewis basic electrophiles to RE has been reported by Walsh et al. in a series of NMR and crystallographic studies [30]. Differences in RE-M(1) combinations lead to a series of complexes with slightly different metal-oxygen bond lengths, covering a broad range of catalytic asymmetric transformations (Figure 1.5). La is most frequently identified as the best RE, presumably because La has the largest ionic radius and is prone to functioning more as a Lewis acid to activate electrophiles. The exceptionally wide variety of reactions presented in Figure 1.5 is indicative that these heterobimetallic cooperative catalysts are one of the most successful classes of asymmetric catalysts known. A reaction mechanism based on Lewis acid-Lewis acid cooperative catalysis in which M(1) serves as a Lewis acid has also been proposed for the aza-Michael reaction, Corey-Chaykovsky epoxidation, and cyclopropanation [21, 25].

It is worth highlighting the direct aldol reaction with LLB (RE = La, M(1) = Li) because this specific reaction was the first to be demonstrated by this heterobimetallic cooperative catalyst and because of the sustained topic in the field of Lewis acid-Brønsted base cooperative catalysis. In 1997, Shibasaki et al. reported the first example of the direct aldol reaction, in which nucleophilically active enolate...