1Oxidation reactionsMany of the transformations covered here involve oxidation of alkenes. A recent review assesses and compares several methods for enantioselective alkene oxidation, including epoxidation, dihydroxylation and aminohydroxylation.81.1Alkene epoxidationMethyl (trioxo)rhenium (MTO) has been studied extensively in recent years as a catalyst for alkene epoxidation by various oxidants, including hydrogen peroxide,9which is attractive since the by-product is water.10Use of hexafluoroisopropanol as solvent has been reported to allow efficient alkene epoxidation with 30% H2O2in the presence of low MTO loadings (0.1 mol%).11Polymer-supported variants of the MTO catalyst (poly-4-vinylpyridine/polystyrene) as recyclable epoxidation catalysts have been described.12In a non-metal mediated alternative, the fluoroketone1(1–5 mol%) is reported to effect alkene epoxidation using Oxone®when hexafluoroisopropanol (HFIP) is used as solvent.13H2O2in HFIP–CH3CN can also be used as co-oxidant, but the reaction is less efficient. Urea- H2O2adduct in HFIP has also been studied, and reactive alkenes are epoxidised without the need for a ketone catalyst.14Catalytic asymmetric epoxidation of alkenes continues to be intensively studied. In the landmark Sharpless asymmetric epoxidation of allylic alcohols mediated by Ti(OPr)4–diethyl tartrate, the use of a tertiary furyl hydroperoxide as co-oxidant which can be recycled has been reported.15Williams has studied the extension of his interesting concept of “catalytic electronic activation” to Sharpless epoxidation.16The idea was that reversible hydrocyanation of an enal would lead to an allylic cyanohydrin which would undergo Sharpless kinetic resolution, followed by regeneration of the aldehyde (Scheme 1). The process would therefore effect indirect asymmetric epoxidation of α,β-unsaturated aldehydes. In the event, the authors were able to demonstrate the first Sharpless kinetic resolution of allylic cyanohydrins as well as conditions for the reversible hydrocyanation of enals. However, the higher temperatures needed for the hydrocyanation resulted in disappointing selectivities for the kinetic resolution. The complete catalytic cycle (i.e.in situregeneration of the aldehyde) remains to be accomplished.For so-called “unfunctionalised” alkenes, chiral Mn(salen) complexes as pioneered by Jacobsen and Katsuki are the best known reagents, and interest in these systems is unabated.17,18Several groups have reported attachment of salen ligands to solid supports.19–24Novel chiral Mn(iii)salen catalysts with built-in phase-transfer capabilities have been described.25Of particular interest are studies on the mechanism of the Mn(salen) catalysed epoxidation, and two substantial contributions have appeared. The fact that the stereospecificity of the epoxidation ofcis-stilbenes (i.e. ratios ofcis/transepoxide that are obtained) varies with the co-oxidant and counterion has led Roschmann and co-workers to put forward an explanation based on a bifurcation step in the catalytic cycle.26In this model, reaction of the Mn(iii) catalyst with the oxygen donor leads to formation of an adduct Mn(iii)OLG (LG = leaving group) as well as the established Mn(v)(oxo) intermediate. The latter epoxidises olefins through a stepwise, non-stereospecific process, while the former acts as a Lewis-acid activated epoxidant, effecting concerted, stereospecific epoxidation. Based on the observedcis/transepoxide ratios, the authors suggest that PhIO, C6F5IO, Bu4NIO4and O3afford predominantly the Mn(v)(oxo) species, leading to extensive loss of stereospecificity; while Bu4NHSO5, NaOCl and dimethyl dioxirane result largely in concerted epoxidationviaMn(iii)OLG. A Hammett study by Linde and co-workers also supports the operation of two separate reaction pathways.27Cr(salen) complexes can give complementary substrate scope to their Mn counterparts—for example, they afford good selectivity forE-alkenes, which generally give poor ee with Mn(salen) catalysts. Additionally, and, again in contrast to the Mn(oxo) species, the Cr(v)&z.dbd;O oxidants can be isolated and employed in stoichiometric epoxidations, thus allowing issues of catalysis and enantioselectivity to be separated. Gilheany and co-workers have reported the effect of added phosphoryl ligands on this process.28Triarylphosphine additives were most effective in increasing enantioselectivity, with a ceiling effect being observed in that the effect of additive was least significant with the more enantioselective salen ligands. Up to 93% ee was obtained for epoxidation ofE-β-methylstyrene.Highly promising results continue to emerge from the study of chiral ketones as epoxidation catalysts,viadioxirane intermediates, usually with Oxone®as the stoichiometric oxidant. Competing Baeyer–Villiger reaction of the ketone generally means that high catalyst loadings are required. Armstrong has reported further details of the use of the fluorotropinone2as catalyst; this compound is stable and so can be used in relatively low loadings.29Up to 83% ee is obtained at room temperature for epoxidation ofE-1,2-disubstituted and trisubstituted aromatic alkenes. Roberts has reported the preparation and testing of some 2-substituted-2,4-dimethyloxabicyclo[3.2.1]octanones, including the fluoroketone3.30The methyl substituents generally lower the catalyst reactivity. Denmark has reported full details of the use of the fluorotropinone4, as well as the biaryl difluoroketone5.31This ketone gives very good enantioselectivity for the epoxidation of some substrates (94% ee forE-stilbene), but its synthesis is more lengthy. Behar screened several fluorinated biaryl ketones, with the best being6(up to 86% ee for epoxidation of β-methylstyrene, but at −15 °C).32Shing described a series ofd-glucose-derived ulose catalysts, with the best being7, which affordedtrans-stilbene oxide of up to 71% ee.33Tomioka has developed the novel chiral ketone8and itsC2-symmetric analogue9, which epoxidisesE-stilbene with up to 64% ee, and gives 82% ee for epoxidation of 1-phenylcyclohexene.34In general, the aforementioned ketone catalysts work best fortrans- and disubstituted alkenes. Asymmetric epoxidation of terminal alkenes is still generally an unsolved problem. In continuation of his outstanding results with chiral ketone catalysis, Shi has recently described the novel ketone10which affords very promising results (Scheme 2).1This catalyst, which is prepared in 9 steps fromd-glucose, also gives good results for somecis-alkenes. Shi has also reported another new ketone catalyst, diacetate11, which effects highly enantioselective epoxidation of α,β-unsaturated esters (Scheme 3).35Good enantioselectivities have also been reported for the epoxidation of cinnamic acid derivatives with dioxiranes derived from keto bile acids.36A related non-metal mediated epoxidation system employs iminium salts as catalysts, which are convertedin situby Oxone®to oxaziridinium salts. Page has described two novel chiral dibenzazepinium catalysts12and13.37Relative to the six-membered ring iminium salt14previously reported by Page, these compounds show greater reactivity. Iminium salt12gives lower ee than the other catalysts. The best ee obtained for13(60% ee with 1-phenylcyclohexene) is similar to that with14(59% with phenylstilbene), but the selectivity pattern is different. In this iminium-catalysed Oxone®epoxidation system, Lacour has reported that the use of the interesting TRISPHAT counteranion15he has developed, in the presence of 18-crown-6, allows the use of a two-phase CH2Cl2–H2O solvent system which can afford higher enantioselectivities than the usual homogeneous CH3CN–H2O conditions.38In the best case, using Page's iminium cation13with counteranion15, epoxidation of 1-phenyldihydronaphthalene increased from 41 to 76% ee using the new solvent system.Asymmetric variants of nucleophilic epoxidation reactions are also making excellent progress. Roberts has reviewed the use of (polyamino)acids for the epoxidation of α,β-unsaturated ketones.39This approach has been used for the epoxidation of chalcone in a continuously operated membrane reactor.40Asymmetric phase transfer catalysis is another important method.41In an interesting variant of this, Lygo has described a double oxidation process in which conversion of an allylic alcohol substrate to an enone is combined with asymmetric epoxidation using the modified cinchona alkaloid phase transfer catalysts he has developed (Scheme 4).42Since the starting enones are often preparedviathe allylic alcohols, this procedure saves one synthetic step.1.2Alkene dihydroxylationAlkene dihydroxylation (AD) mediated by osmium tetroxide is one of the most reliable and general organic transformations. The reaction is given added importance by the success and generality of the Sharpless asymmetric variant. Certain substrates,e.g. electron-poor or hindered alkenes, sometimes exhibit poor turnover and/or enantioselectivities in catalytic dihydroxylations. By employing a high-throughput screening protocol, Sharpless has evaluated over 500 additives in Os-catalysed dihydroxylation withN-methylmorpholine-N-oxide (NMO) and has discovered that electron-poor alkenes react much more efficiently in acidic media.43Citric acid is the additive of choice, leading to simple practical procedure where mixing of reagents intBuOH–H2O is often followed simply by filtration of the reaction product (Scheme 5). A mechanistic rationale is presented in terms of the “two-cycle” catalytic mechanism (Scheme 6).44Generally, the second cycle is considered deleterious to enantioselectivity in the AD since the chiral ligand is not involved. Indeed, the conventional ferricyanidetBuOH–H2O co-oxidant system was developed to remove this second cycle. Lower ee is often obtained using the UpjohnN-methylmorpholine-N-oxide co-oxidant since the second cycle operates in a homogeneous acetone–water solvent system. Recently, however, Sharpless has started to explore the possibility of exploiting the second cycle for asymmetric catalysis by employing chiral ligands that take the place of one of the diol ligands in the intermediate16and remain attached to the osmium species throughout the process, thus rendering the second cycle enantioselective. Promising results have been obtained withN-tosyl-1,2-amino alcohol ligands (Scheme 7).45Recent trends over the last few years in studies of the Sharpless AD have included the investigation of more environmentally friendly co-oxidant systems, including molecular oxygen. From an industrial viewpoint, the high cost of the osmium catalyst and chiral ligand as well as the toxicity and volatility of OsO4are obstacles to large-scale application. Several approaches are currently under investigation to allow easier recycling of osmium and/or ligand. Zhang has reported a novel modified ligand17which is soluble in thetBuOH–H2O solvent system usually used for the Sharpless AD reaction, but which can be recovered by extraction with CH2Cl2followed by precipitation with ether.46This ligand can be employed in lower loadings than polymer-supported variants reported previously. Immobilisation of osmium is also an important goal. A detailed study of recoverable catalysts based on OsO42−on ion exchange resins has been reported.47The current interest in the use of ionic liquids as reaction solvents is reflected by reports from several groups of their use in osmium-catalysed dihydroxylation reactions. The diol products can be extracted with an immiscible organic solvent, leaving the osmium reagent in the ionic liquid layer, which can be re-used. Two groups reported this concept for non-asymmetric dihydroxylation using the Upjohn co-oxidant system (N-methylmorpholine-N-oxide, NMO).48,49Interestingly, Yao observed that the presence of DMAP was needed to ensure “anchoring” of the OsO4in the ionic liquid [[bmim]PF6] (bmim = 1-butyl-3-methylimidazolium).49A further two reports have successfully accomplished Sharpless asymmetric dihydroxylation in ionic liquids. Afonso employed K2OsO2(OH)4and ferricyanide co-oxidant in [bmim]PF6and water, with or withouttBuOH.50Song also used [bmim][PF6], but studied the Upjohn NMO process and found that leaching of the catalyst system out of the ionic layer could be avoided by using 1,4-bis(9-O-quininyl)phthalazine [(QN)2PHAL] as ligands rather than the usual dihydroquininyl (DQN) series.51Presumably the alkenes in the (QN)-ligands undergo initial dihydroxylation.Donohoe's work on directed dihydroxylationsyn-to hydroxyl groups using OsO4with diamine ligands [N,N,N′,N′-tetramethylethylenediamine (TMEDA)] has solved a long-standing problem, allowing complementary stereocontrol to that usually observed (Scheme 8). An account of the development of this work has recently appeared,52along with a full paper on the directed dihydroxylation of cyclic allylic alcohols and trichloroacetamides.53While osmium catalysis is by far the most widely studied, there are several interesting recent reports of alternative metals for alkene dihydroxylation. Que has continued his studies on non-heme iron oxidation catalysts.54Hage and Feringa have observed diol formation (in competition with epoxidation) in a Mn-based system with H2O2as co-oxidant.55Brown has reported an alternative to the Sharpless AD reaction which uses stoichiometric permanganate in the presence of a chiral phase-transfer reagent (18).56Good enantioselectivities were obtained for dihydroxylation of enones (Scheme 9).1.3Alkene aminohydroxylationcis-Aminohydroxylation of alkenes, involving osmium catalysis in the presence of a nitrogen source, is a powerful approach to amino alcohol synthesis. It is fair to say, however, that this process has yet to reach the practicality of the closely related dihydroxylation reaction, due to problems with regiocontrol and competing diol formation. The Sharpless asymmetric aminohydroxylation has been thoroughly reviewed very recently.57Sharpless has studied the use of ligands that promote “second cycle” reaction, as discussed above for dihydroxylation, in the aminohydroxylation process.45Ojima has studied the regiochemistry of aminohydroxylation of several protected 4-hydroxy-2-butenoates and has put forward a model based on the interaction of the substrate with the chiral ligand's binding pocket.58,59In an interesting approach to regiocontrol, Donohoe has developed an intramolecular aminohydroxylation of allylic carbamates and has now applied this to cyclic allylic carbamates (Scheme 10).60The reaction works for five membered rings (as long as the alkene is exocyclic), as well as seven- and eight-membered.1.4Alcohol oxidationAs one of the most widely employed reactions in synthesis, there is currently tremendous interest in “clean” alcohol oxidation using environmentally friendly reagents and reaction media.61The Swern reaction employing DMSO activated by oxalyl chloride is one of the most popular oxidation methods, particularly for complex substrates, and an odourless version using dodecyl methyl sulfoxide has been described.62Another oxidant that has proved its worth in complex molecular environments is TPAP (tetra-N-propylammonium perruthenate), which can be employed catalytically with, for example, NMO as co-oxidant. The groups of Ley63and Welton64have independently described catalyst recovery by use of tetralkylammonium salts or room-temperature ionic liquids. Sheldon has reported a two-phase [methyltert-butyl ether (MTBE)–H2O], buffered (pH 8.5) catalytic co-oxidant system for TPAP with NaOCl, from which the catalyst is recoverable.65A buffered Ru–NaOCl system also effects α-oxidation of ethers to esters.66Molecular oxygen is a particularly attractive oxidant from an environmental viewpoint.61In 2001, Sigman67and Stoltz68independently reported a Pd-catalysed asymmetric aerobic oxidation of secondary alcohols in the presence of sparteine, allowing kinetic resolution. Further mechanistic studies have appeared, indicating that sparteine plays the roles both of ligand on Pd and base.69Replacement of sparteine with triethylamine allows (non-asymmetric) alcohol oxidation to take place at room temperature with 3 mol% Pd(OAc)2and oxygen.70Cinnamyl alcohol does not work well; this is believed to be because the product α,β-unsaturated carbonyl compound undergoes coordination to Pd. In contrast, another interesting aerobic oxidation system that has appeared recently offers an alternative to manganese dioxide, the standard reagent for selective allylic oxidation (Scheme 11). This chemistry involves an Os–Cu–co-catalyst system.71Despite the cost of Os, the catalytic nature means that it is competitive with the use of stoichiometric MnO2. From a practical viewpoint, the fact that the reactions work at ambient temperatures and atmospheric pressure is highly attractive. Air can be used in place of oxygen as long as it is bubbled through the reaction mixture. Sulfide and acetal functional groups are tolerated.IBX19, the synthetic precursor to the popular Dess–Martin reagent, is receiving increasing attention for alcohol oxidation. One perceived limitation has been the need to use DMSO as reaction solvent in order to dissolve the IBX. Finney has developed a practical protocol for the use of IBX in EtOAc or dichloroethane as solvent.72The trick is to heat the reaction mixture to 80 °C, at which temperature the reagent dissolves. On cooling, the reduced form of IBX precipitates out and can be removed by filtration. The reaction works for conversion of a range of primary and secondary alcohols to aldehydes and ketones respectively; exceptions noted are benzyl alcohol, which is converted to benzoic acid, and Fmoc-phenylglycinol. These are isolated problem substrates, however, since oxidations of other benzylic alcohols stopped at the aldehyde stage, and Fmoc-aminoethanol was oxidised cleanly (Scheme 12).1.5Dehydrogenation of carbonyl compoundsFollowing the discovery of novel reactivity modes during total synthesis programmes, Nicolaou and co-workers have extensively exploited IBX19as a reagent in organic synthesis. One important application is the dehydrogenation of carbonyl compounds, and a comprehensive full paper on this chemistry has appeared, containing a wealth of synthetic and mechanistic detail.2These reactions are believed to proceedviaSET mechanisms. Selected examples are shown inSchemes 13–15. Cycloheptanone20can be converted to either the mono- (21) or bisenone (22) depending on the number of equivalents of IBX employed and the reaction temperature (Scheme 13). Indeed, cycloheptanol may be used as starting material instead, since IBX will oxidise it to the ketone. The reaction tolerates a wide range of functional groups, including acetates, methoxy groups and primary nitriles. Other notable examples include the successful dehydrogenation of23, with an alkene five carbon atoms away (Scheme 14), and the amino acid derivative24(Scheme 15).Nicolaou has discovered some interesting ligand acceleration effects in this reaction. In particular, complexation of IBX with 4-methoxypyridine-N-oxide (MPO) results in the new reagent25(Scheme 16) which is sufficiently reactive to effect dehydrogenation of carbonyl compounds at room temperature.73Other oxidation processes such as benzylic oxidation (vide infra) do not compete at this temperature, and reactive or volatile aldehydes can now be employed. As a further example of the increased substrate scope, the first example of a substrate containing an alkylamine was achieved (Scheme 17). In a complementary study, the same workers have also shown that IBX and IBX·MPO will also effect dehydrogenation of silyl enol ethers to enones.74Nicolaou has begun to explore less expensive variants of IBX,i.e. reagents lacking the aromatic backbone. The iodine complexes26and27derived from DMSO and iodic acid or iodine pentoxide, respectively (Scheme 18), were found to be mild and effective oxidants for dehydrogenation of aldehydes and ketones, with the former generally preferred due to its lower cost. Cyclohexene was added as co-solvent in order to scavenge any iodine by-product. The scope of the reaction was found to be similar to the IBX reaction, with the indanone28performing better with the new reagent (Scheme 19). Additionally, alcohols are inert to26and27, so carbonyl dehydrogenation may be performed in their presence.1.6Benzylic and allylic oxidationAs a further aspect of the SET-chemistry mediated by IBX, introduced above, full details have appeared of the use of this reagent for benzylic oxidation (Scheme 20).2In contrast to conventional oxidants, the presence of alkenes,N-heterocycles, amides and aldehydes does not interfere. The reaction does not proceed with electron-poor substrates such asp-cyano- andp-acetyltoluene, however.The asymmetric variant of the Kharasch–Sosnovsky allylic oxidation reaction, involving the use of Cu-catalysis, chiral bisoxazoline ligands and peroxy esters, was reported simultaneously by several groups in the early 1990s. The area has been reviewed recently by Andrus and Lashley.75The Andrus group has also reported that higher enantioselectivities can be obtained iftert-butylp-nitrobenzoate is used. Screening of a panel of bisoxazolines allowed highly enantioselective allylic oxidation of cyclopentene, cyclohexene, cycloheptene and cyclooctadiene to be accomplished. High ee (94–99%) was obtained, but reaction times are still long (often several days), and yields are often low.76