Palladium-Catalyzed Oxidation of Organic Chemicals with O2

See allHide authors and affiliations

Science  16 Sep 2005:
Vol. 309, Issue 5742, pp. 1824-1826
DOI: 10.1126/science.1114666

Atmospheric oxygen maintains a tenuous existence on Earth, far from chemical equilibrium with its surroundings. This thermodyamic instability has destructive potential, manifested in accidental fires, explosions, and corrosion, for example, but it also can be harnessed for beneficial purposes. Aerobic organisms produce energy via respiration, which involves the complete oxidation of glucose and other organic molecules to carbon dioxide and water, and fossil fuel combustion supplies the majority of worldwide energy demands. The recent Technology Vision 2020 report published by the Council for Chemical Research highlights selective oxidation of organic chemicals as one of the most critical challenges facing the chemical industry (1), and molecular oxygen embodies the quintessential oxidant for chemical synthesis. This oxidant is available at virtually no cost and produces no environmentally hazardous by-products. But how can chemical reactions between organic molecules and oxygen gas be controlled to produce useful, selectively oxidized products without resulting in complete combustion of the starting materials?

The answer lies in the development of catalysts to guide the chemical reaction toward kinetically favored products. A number of industrial processes feature catalytic methods for aerobic oxidation, but their scope is limited, and chemical reagents such as transition metal oxides and chlorine-based oxidants remain in common use. However, recent developments in homogeneous palladium catalysis point toward new opportunities for selective aerobic oxidation chemistry (2).

The importance of these results must be considered within the context of other homogeneous aerobic oxidation reactions. Among the most useful are autoxidations, radical-chain reactions that find application in the production of numerous large-volume commodity chemicals. Examples include terephthalic acid, a principal component of plastic soda bottles, and cyclohexanone, an important precursor to nylon fabrics and materials. Unfortunately, however, autoxidation reactions are only compatible with substrates that undergo selective radical chemistry. Enzymes and biomimetic catalysts that mediate aerobic oxidation have been classified as oxygenases or oxidases, depending on their catalytic mechanisms. Oxygenases mediate the transfer of one or both oxygen atoms from O2 to the organic molecule. Oxidases mediate substrate oxidation without oxygen-atom transfer and couple this process to the reduction of dioxygen to hydrogen peroxide or water. The energy transduction mechanism used by oxidases is reminiscent of fuel cells (see the figure), wherein the electrode-mediated redox reactions occur without a direct interaction between the substrates, molecular hydrogen and oxygen. Palladium-catalyzed aerobic oxidation reactions are similar to enzymatic oxidases and fuel cells in that the catalytic mechanism can be separated into two independent half-reactions: palladium(II)-mediated oxidation of the organic substrate (SubH2) and dioxygen-coupled oxidation of palladium(0). This mechanism has important implications for the development of new catalytic oxidation reactions. Autoxidation and oxygenase reactions are limited to oxygen-atom-transfer methods; however, an oxidasestyle mechanism is formally compatible with any oxidative transformation. This feature is particularly important because numerous oxidation reactions do not involve the incorporation of oxygen into the substrate, for example, the oxidative amination of alkenes and the dehydrogenation of alcohols or alkanes. The diversity of palladium-mediated organometallic transformations substantially broadens the scope of possible reactivity.

Catalysis revival.

Palladium-catalyzed aerobic oxidation processes, which have recently been the subject of renewed interest, display mechanistic similarities to reactions in fuel cells and oxidase enzyme catalysis. (SubH2 is the organic substrate, Subox is the oxidized organic substrate, L is the neutral donor ligand for palladium, and X is the anionic ligand for palladium.)

Palladium-mediated oxidation reactions are not new, of course. Stoichiometric examples were first identified in the 1800s, but the explosive growth in the study of palladium catalysis originated with development of the Wacker process in the late 1950s (3). In this industrial method for the production of acetaldehyde (C2H4 + ½O2 → CH3CHO), palladium(II) salts promote the oxidative coupling of ethylene and water, and copper(II) cocatalysts mediate aerobic oxidation of the reduced palladium catalyst. This process remains a landmark success in the use of homogeneous catalysis for commodity chemical synthesis.

Despite this early achievement, subsequent advances in palladium catalysis were dominated by nonoxidative coupling reactions. Although palladium oxidation catalysis also progressed, certain features of the Wacker process proved to be quite restrictive. For example, many organic molecules are only sparingly soluble in water, the industrial reaction medium, and the copper cocatalysts that mediate dioxygen-coupled turnover are generally less effective in organic solvents. Consequently, many of the new applications of palladium-catalyzed oxidation required alternative oxidants such as benzoquinone or stoichiometric copper(II) salts. These modified conditions generally limited the applications to laboratory-scale reactions because of their added cost, increased reaction waste, and more complicated product isolation.

The recent revival in palladium oxidation chemistry coincides with the growing recognition that efficient dioxygen-coupled catalysis can be achieved in the absence of copper cocatalysts or related redox mediators (2). In these reactions, catalyst regeneration occurs by direct reaction of molecular oxygen with the reduced palladium species. A prominent feature of the more recent discoveries is the use of oxidatively robust ligands to stabilize the palladium catalyst, promote catalytic turnover, and foster unprecedented reaction selectivity, including asymmetric transformations.

Early reports of direct dioxygen-coupled catalysis appeared in the late 1960s, but the recent surge of research activity in this area originated with codiscovery of the palladium acetatedimethyl sulfoxide [Pd(OAc)2/DMSO] catalyst system in the mid-1990s by Larock and Hightower (4) and by Hiemstra and colleagues (5). This extremely simple catalyst system has been used in a variety of oxidative transformations, including alcohol oxidation, intramolecular hetero- and carbocyclization of alkenes, and dehydrosilylation of silyl enol ethers (2, 6). The metal-coordinating properties of DMSO serve to stabilize the catalyst and promote aerobic oxidation of the reduced catalyst, but mechanistic studies reveal that inefficient capture of reduced palladium by molecular oxygen results in competitive catalyst decomposition via metal aggregation (2). These observations account for the relatively low catalyst lifetime and turnover rates.

The field of homogeneous catalysis owes much of its success to the role of organic ligands that bind to the metal and modulate catalyst stability, reactivity, and selectivity. Palladium oxidation chemistry, however, has been dominated by ligand-free reaction conditions, perhaps because many common ligands for late transition metals, such as phosphines and related soft donors, are susceptible to oxidative decomposition. In the late 1990s, several research groups independently discovered that nitrogen-donor ligands, including aromatic imines and tertiary alkyl amines, confer benefits on catalytic reactivity (7-12). These reactions take place in diverse conditions, ranging from aqueous to nonpolar organic solvents, and exhibit catalytic turnover rates as much as two to three orders of magnitude higher than that of the Pd(OAc)2/DMSO system.

The nitrogen-ligated catalyst systems have been applied to traditional palladium-catalyzed reactions and in the discovery of novel oxidative transformations. Sheldon et al. used a water-soluble phenanthroline ligand with Pd(OAc)2 to achieve aqueous, cocatalyst-free conditions for alcohol oxidation and Wacker-type oxidation of terminal alkenes (7). A group at Enichem S.p.A. (Novara, Italy) has used a related organic-soluble ligand, bathocuproine, to investigate direct routes for hydrogen peroxide synthesis from molecular oxygen (8). Sacrificial substrates, alcohols or carbon monoxide, reduce the ligated palladium(II) catalyst, and hydrogen peroxide forms via reaction of dioxygen and two proton equivalents with palladium(0). A palladium-dioxygen adduct, directly relevant to the latter reaction, has been crystallographically characterized, and studies of its reactivity provide key insights into the mechanism for aerobic oxidation of palladium(0) (2).

The impact of new aerobic oxidation methods will probably be experienced first in small-scale chemical synthesis—for example, in pharmaceutical discovery and academic research labs. In this context, the Pd(OAc)2/pyridine catalyst reported by Nishimura and Uemura (9) represents one of the most convenient and synthetically versatile catalyst systems to date. It has been used by several groups to perform a variety of reactions, including alcohol oxidation, oxidative C-C cleavage reactions with tertiary alcohols, and oxidative C-O, C-N, and C-C coupling reactions with alkenes (2, 9). Inspiration from this catalyst system also led to the discovery of asymmetric catalytic reactions. The groups of Stoltz and Sigman independently reported that the use of (-)- sparteine, a chiral naturally occurring diamine, in the oxidation of secondary alcohols enables preferential reaction of a single enantiomer of the substrate (10-11).

Practical applications of palladium-catalyzed aerobic oxidation for commercial use will require the development of even better catalysts and the discovery of new chemical transformations. Recently, catalysts have been discovered that are active at room temperature and that operate effectively with ambient air rather than pure oxygen or elevated gas pressures (12-15). These results, made possible by the identification of new ligands for palladium, bode well for future catalyst development efforts. Prospects for new reaction development also appear promising. The oxidative coupling of simple alkenes and amine derivatives appears to be a straightforward analog of the Wacker process, but this reaction has eluded researchers for decades. A recent solution to this problem derives its success from the identification of palladium-catalyzed conditions compatible with direct dioxygen-coupled turnover (16). Aerobic oxidation chemistry is a subject of critical importance, and new opportunities afforded by the advances outlined above suggest that the field of palladium oxidation catalysis is poised for a comeback.


View Abstract

Stay Connected to Science

Navigate This Article