Highlights

Markus Hölscher reviews some of the recent literature in green chemistry

---

Structure of ionic liquids

Room temperature ionic liquids (RTILs) have been established during recent years as “green” solvents in catalytic applications due to their recyclability and their advantageous physical and chemical properties. Owing to their hygroscopic nature RTILs can absorb substantial amounts of water, which not only changes their properties but makes them depend crucially on the water content. In a detailed NMR investigation Mele et al. have collected information on the microscopic structures of the well known RTIL 1-n-butyl-3-methylimidazolium tetrafluoroborate [BMIm]⁺[BF₄]⁻ by evaluating the role of cation–cation, cation–water, and cation–anion interactions (Angew. Chem., Int. Ed., 2003, 42, 4364–4366).

The authors found the imidazolium ring protons to interact with the anion, when no water is present in the RTIL. With increasing amounts of water the C(sp²)–H⋯F interactions are progressively replaced by water contacts, with the water acting as an acceptor toward the cation and a donor towards the anion. Also the imidazolium–imidazolium interaction present in the water free RTIL becomes looser with increasing water content, shifting the rings away from each other. In the pure liquid the cations and anions are associated as tight ion pairs, and this structure is preserved at low water contents and even in DMSO solution. The short range structure of this RTIL is governed by C(sp²)–H⋯F interactions between cation and anion.

Heterogeneous catalysts for Diels–Alder and aldol reactions

Heterogeneous Lewis acid catalysts often suffer from inferior activities and leaching of active components. Homogeneous catalysts are highly selective but difficult to recover. A novel catalyst design may help to solve these problems, by combining the advantages of homogeneous and heterogeneous catalysis while avoiding their drawbacks. Kaneda et al. from the Graduate School of Engineering Science, Osaka University have exploited the unique features of hydroxyapatite (HAP) as a macroligand for Ru³⁺ ions (J. Am. Chem. Soc., 2003, 125, 11460–11461). Simple cation exchange of Ca²⁺ by Ru³⁺ yields cationic RuHAP catalysts which surprise with high selectivities in Diels–Alder reactions and no observable trend for product inhibition as known from other Al-, Ti- or B-based Lewis acid catalysts. Also the less reactive methyl acrylate dienophile undergoes the reaction smoothly with cyclopentadiene as the diene (yield 82%, endo : exo = 91 : 9).

The RuHAP catalysts also tolerate the presence of water in aldol reactions of nitriles and carbonyl compounds, yielding α,β-unsaturated nitriles in high yields. No leaching was detected over a reaction period of 24 h by ICP analysis and the recovered catalyst could be reused without loss of activity.

Asymmetric heterogeneous catalysis

Another example of successful heterogenization of homogeneous catalysts for asymmetric hydrogenation has recently been accomplished using new Rh- and Pd catalysts by the groups of Raja, Thomas, Lewis and Harris (Angew. Chem., Int. Ed., 2003, 42, 4326–4331). Chiral pyrrolidine and amine ligands, which are anchored to porous silica supports with concave and convex surfaces comprise cationic Rh- and Pd catalysts, which efficiently and selectively catalyse the hydrogenation of E-α-phenylcinnamic acid and methyl benzoylformate.

Most striking is the fact that these heterogenized catalysts outperform their homogeneous counterparts, and the authors also stress the fact that it is the concavity of the pores which is superior to supports with convex surfaces. Product selectivities (up to 98%) and enantioselectivities (up to 99% ee) reach high values with conversions between 74 and 100%. Leaching tests showed the catalytic process to be truly heterogeneous and up to the second recycle the Rh catalysts showed no marked loss of conversion, selectivity and ee value.

Kinetic resolution of secondary alcohols

Dynamic kinetic resolution (DKR) is a useful tool for the conversion of racemic substrates to single enantiomeric products. DKR relies on a metal complex, which acts as a racemizing catalyst and an enzyme as a resolving catalyst. So far enzyme/metal-catalyzed DKR systems employed lipases as resolving catalysts, which in the case of simple secondary alcohols have thus yielded only products with (R)-configuration. This has now been changed by Park, Kim and coworkers of the Laboratory of Chirotechnology at the University of Science and Technology, Korea (J. Am. Chem. Soc., 2003, 125, 11494–11495). They chose subtilisin as the resolving catalyst and aminocyclopentadienylruthenium complexes as racemizing catalyst.

The problem of low subtilisin activity in organic solvents was overcome by using a nonionic polyoxyethylene(10) cetyl ether surfactant increasing the activity by a factor of 3 orders of magnitude. On the basis of test experiments with 1-phenylethanol THF was chosen as a suitable solvent and TFEB as acyl donor for the resolution of various secondary alcohols. As an example p-chlorophenyl methyl carbinol transformed with 92% yield and 99% ee.

Epoxidation of alkenes in ionic liquids

The search for environmentally friendly methods for the synthesis of epoxides via alkene epoxidation has recently been stimulated by the findings of Burgess et al. who reported on the catalytic activity of manganese sulfate and sodium bicarbonate for this reaction in aqueous media with hydrogen peroxide as the oxidant (Chem. Rev., 2003, 103, 2457). This approach has been modified by Chan et al. (Org. Lett., 2003, 5, 3423–3425), who succeeded in employing an ionic liquid instead of an organic cosolvent for the water-insoluble lipophilic substrates.

The combination of 1-butyl-3-methylimidazolium tetrafluoroborate [BMIm]⁺[BF₄⁻] as the solvent for the alkene and an aqueous phase with catalytic amounts of manganese sulfate and tetramethylammonium hydrogen carbonate (TMAHC) yields an active catalytic system, which converts various internal alkenes readily into the corresponding epoxides upon addition of hydrogen peroxide. Conversions and yields were 99% or more in many cases and the ionic liquid can directly be reused at least 10 times after extraction of the product. Even though manganese sulfate must be added after a few cycles to retain activity it is a cheap, recyclable, catalytic and environmentally benign reaction.

Photochemical water splitting—summary of a recent debate

As early as 1972 Fujishima and Honda had reported about n-TiO₂ electrodes splitting water photoelectrochemically into hydrogen and oxygen (Nature, 1972, 238, 37). This important finding has stimulated research over the past thirty years. Most recently Khan et al. reported about a new n-type TiO₂ splitting water with a maximum photoconversion efficiency of 8.35%—a result that would have to be considered a major progress in the development of these materials (Science, 2002, 297, 2243). However, several researchers have strongly challenged this report and a considerable debate has emerged only a few days after the publication of these results. In a first answer by Lackner it was argued that Khan et al. had not supplied sufficient data to reliably judge the given efficiency (Science, 2003, 301, 1673c). Even more pertinently Lackner hinted at the possibility of the external power supply used by Khan et al. being the only source of electrolysis. The debate went on when Fujishima, who claimed to have invented the same type of material already in 1975 (only three years after his original publication) argued also that there were not data presented which evidenced hydrogen generation induced by visible light (Science, 2003, 301, 1673a). In a further comment by Hägglund, Grätzel and Kasemo, who approached Khan’s work from different angles, it was shown on theoretical grounds that the calculation method used by Khan et al. was misleading (Science, 2003, 301, 1673b). The authors came to the conclusion that the reported efficiency is by far too large.

Khan et al. responded quickly. They defended their work on technical and ethical grounds and dealt with every single comment in detail (their answer to the comments is longer than the original article), and found their work to be justified. They admitted, however, that they should have cited the 1975 publication by Fujishima (Science, 2003, 301, 1673d). It looks as if not all arguments were settled by this “battle” and more work on these interesting materials might help to clarify the situation in the future.

---