1
Application of Stimuli-Responsive and "Non-innocent" Ligands in Base Metal Catalysis

Andrei Chirila Braja Gopal Das Petrus F. Kuijpers Vivek Sinha and Bas de Bruin

University of Amsterdam (UvA), Van 't Hof Institute for Molecular Sciences (HIMS), Homogeneous, Supramolecular and Bio-Inspired Catalysis (HomKat), Science Park 904, 1098 XH Amsterdam, The Netherlands

1.1 Introduction

The development of efficient and selective catalysts is an important goal of modern research in chemistry - the science of matter and its transformations. Our society needs new catalysts to become more sustainable, and a desire for selectivity and efficiency in the preparation of medicines and materials has boosted our interest in developing new methods based on homogeneous catalysis, particularly on the development of new ligands that can be fine-tuned to specific needs. The properties of a metal complex as a whole are the result of the interaction between the metal center and its surrounding ligands. In traditional approaches, the steric and electronic properties of the spectator ligand are used to control the performance of the catalyst, but most of the reactivity takes place at the metal. Recent new approaches deviate from this concept and make use of ligands that play a more prominent role in the elementary bond activation steps in a catalytic cycle [1, 2]. The central idea is that the metal and the ligand can act in a synergistic manner to facilitate a chemical process. In this light, complexes based on the so-called "non-innocent" ligands offer interesting prospects and have attracted quite some attention.

The term "non-innocent" is broadly used, and diverse authors give different interpretations to the term. It was originally introduced by Jørgensen [3] to indicate that assigning metal oxidation states can be ambiguous when complexes contain redox-active ligands. As such, ligands that get reduced or oxidized in a redox process of a transition metal complex are often referred to as "redox non-innocent." [4, 5] With modern spectroscopic techniques, combined with computational studies, assigning metal and ligand oxidations states has become less ambiguous, and hence, many authors started to use the term "redox-active ligands" instead. Gradually, many authors also started to use the term "non-innocent" for ligands that are more than just an ancillary ligand, frequently involving ligands that have reactive moieties that can act in cooperative (catalytic) chemical transformations, act as temporary electron reservoirs, or respond to external triggers to modify the properties or reactivity of a complex. A common objective of many of these investigations is to achieve better control over the catalytic reactivity of first-row transition metal complexes, with the ultimate goal to replace the scarce, expensive noble metals currently used in a variety of catalytic processes by cheap and abundant first-row transition metals. Instead of providing a comprehensive overview of redox non-innocent [6, 7] and cooperative ligands [1, 8, 9], this chapter is intended to provide a conceptual introduction into the topic of achieving control over the catalytic reactivity of non-noble metals using non-innocent ligands on the basis of recent examples.

Noble metals are frequently used in several catalytic synthetic methodologies and many industrial processes [10]. Their catalytic reactivity is most frequently based on their well-established "two-electron reactivity," involving typical elementary steps such as reductive elimination and oxidative addition. These elementary steps easily occur for late (mostly second and third rows) transition metals having two stable oxidation states differing by two electrons. However, most noble metals are scarce and are therefore expensive (and sometimes toxic [11]). Therefore, it is necessary to reinvestigate the use of cheaper, abundant, and benign metals to arrive at cost-effective alternatives. This is not an easy task, as base metals (Fe, Co, Cu, Ni, etc.) often favor one-electron redox processes, and typical elementary steps commonly observed in noble metal catalysis are only scarcely observed for base metals. As such, the unique properties of non-innocent ligands are advantageous to gain better control over the reactivity of base metals. In some cases, this leads to reactivity comparable to that of noble metal complexes (but more cost-effective and benign), whereas in other cases, the combination of a base metal with a "non-innocent" ligand can actually give access to unique new types of reactivity.

This chapter has four parts. In Section 1.2, the concept of responsive ligands is discussed, giving examples of a series of ligands that can be tuned using external stimuli such as light, pH, or ligand-based redox reactions. These can trigger a change in the properties of the ligand, thereby modifying the reactivity of the metal. Section 1.3 deals with redox-active ligands that behave as electron reservoirs. In the examples provided, this feature enables oxidative addition and reductive elimination steps for first-row transition metal complexes that, without the aid of redox-active ligands, are less inclined to undergo these catalytically relevant elementary steps. Section 1.4 focuses on recent examples of cooperative catalysis, in which non-noble metal reactivity is combined with ligand-based reactivity in key substrate activation steps. The last part (Section 1.5) deals with examples in which the coordinated substrate itself acts as a redox-active moiety in key elementary steps of a catalytic reaction. More specifically, these coordinated substrates get oxidized or reduced by the metal by a single electron, thus creating "substrate radicals," which play an important role in catalytic radical-type transformations.

1.2 Stimuli-Responsive Ligands

Common ancillary (innocent) ligands in homogeneous catalysis typically control the activity and selectivity of the catalyst by affecting the steric and electronic properties around the reactive metal center. As such, changing the reactivity of the active metal center usually requires the synthesis of new ligands, which is often associated with elaborate synthetic procedures [6]. However, the electronic and steric properties of ligands can sometimes be influenced in an easier manner by using external stimuli, involving, for example, ligand protonation/deprotonation, ligand oxidation/reduction, or (reversible) light-induced ligand transformations (Scheme 1.1) [12].

Scheme 1.1 Switching catalytic properties of a catalyst using external stimuli.

When using such responsive ligands, the metal oxidation state is typically unaffected, but its reactivity is nonetheless influenced by the new electronic and steric properties of the ligand. Furthermore, the solubility of the metal complex can sometimes be significantly influenced by such external stimuli. In most current literature, these ligands are nevertheless considered to be "innocent" ligands as they are not directly involved in substrate bond making/breaking processes nor lead to ambiguities in assigning the metal oxidation state. Stimuli-responsive ligands are particularly useful to influence the catalyst during a catalytic reaction and are therefore mainly applied to develop switchable catalytic systems.

1.2.1 Redox-Responsive Ligands

Oxidation or reductionof a complex containing one or more redox-active ligands can lead to oxidation or reduction of the ligand rather than the metal. As such, the ligand can switch between one or multiple oxidized and reduced states, by which the electronic properties of the ligand (and thereby the metal) change. These redox processes can be triggered either chemically or electrochemically [13]. Often metallocenes such as ferrocene or cobaltocene are used because of their reversible oxidation and reduction cycles [14]. In other cases, the redox-active part of the ligand of interest is actually a metallocene moiety [15]. Upon oxidation of a ferrocenyl to a ferrocenylium group attached to the ligand, the electron density of the donor ligand decreases and thereby also that of the metal bound to this ligand, as can be observed in a shift of the CO stretch frequency to higher wavenumbers for carbonyl complexes [16]. Recently, a review appeared reporting a variety of chemical oxidants and reductants that allow the design of new catalysts with switchable ligands at a specific desired potential [17]. Examples of the use of redox-active ligands in catalysis frequently involve redox processes that partly occur at the redox-active ligand and partly at the catalytic metal center (see Section 1.3). Examples of redox-responsive ligands in catalysis wherein ligand-based redox processes affect the metal center and its catalytic properties indirectly are rare, especially for base metals. The main application of such reported examples is in the field of switchable catalysis. Furthermore, the solubility of the ligand can change significantly because of charge buildup, thus enabling separation of the catalyst from the reaction mixture after a catalytic...