CH2402 Homogenous Catalysis Catalytic Cycles Important Terms

precatalyst

Figure 1 - The general form of a catalytic cycle

The overall rate of a catalytic process is controlled by the turnover limiting step. The concentration of active catalyst refers to the sum of the concentrations of all catalytic species within the cycle. The catalyst resting state refers to the species of highest concentration, off-cycle this inhibits the catalysis, within the cycle this defines the turnover limiting step. The turnover number is the moles of product per moles of catalyst or precatalyst and the turnover frequency is the turnover number per unit time, the distinction between these terms is important to consider when reading literature. A large number of catalytic cycles consist of; Oxidative addition of a substrate One or more migratory insertion steps Termination via reductive elimination or protonolysis/hydrolysis or likewise

Hydroamination Reactions A hydroamination reaction is the formal addition of either ammonia, a primary amine (RNH 2) or a secondary amine (R2NH) to an unsaturated carbon-carbon bond. The reaction may performed both inter- and intramolecularly (Figure 2).

Figure 2 - Hydroamination of alkenes and alkynes

There are three groups of catalyst and three methods of activation possible for hydroamination reactions; 1. Deprotonation (NH activation) usually performed using Lewis acidic, early transition metals in high oxidation states, e.g. Zr(IV), and lanthanides (Figure 3) The ligands used in these reactions are often based on the cyclopentadienyl ligand (though there are examples of non-metallocene systems (Figure 4)), since the reaction requires coordination of both the amine and the unsaturated carbon one of the most important aspects of optimisation is ensuring there is enough of the metal surface available for both to approach. This can be achieved by using a larger metal or through use of ansa-metallocenes, in which the cyclopentadienyl ligands are held open by a bridging moiety, often SiR2. For intramolecular hydroamination the length of the chain will also have an impact on the stearic arrangement of the reaction components, it is also found that alkynes are easier to hydroaminate than alkenes most likely due to the presence of additional bulk in the alkene providing further stearic hindrance. In some literature cases it is found that smaller metals perform better hydroaminations of alkynes cf. alkenes, this is a case in which the electronic benefits outweigh the stearic negatives. The -system of the alkyne is more accessible, with a greater donating capacity and so a greater electronic stabilisation of the four-centre transition state this is a thermodynamic effect.

Figure 3 Hydroamination by NH activation (deprotonation) using an organolanthanide catalyst

Figure 4 - Examples of non-metallocene Ln catalysts

Enantiomeric control has proven elusive with this method. There is a relationship between the ionic radii of the rare earth metal and the enantiomeric excess with the maximum enantioselectivity observed with samarocene derivatives (samarium, Sm, has a radii of ~1.08, cf.~ 0.97 for lutetium and 1.16 for lanthanum). One cause of poor selectivity is the facile epimerisation (Figure 5) of the cyclopentadienyl ligand possible when using racemic precatalyst, rotation about the Si-cp bond leads to loss of any enantioselectivity.

Figure 5 - Epimerisation of a cyclopentadienyl ligand

Hydroamination of alkynes with group IV metals (Figure 6) occurs via formation of an imido complex (cf. amido for lanthanide systems), since the group IV metals are more Lewis acidic they are better able to deprotonate the amine and stabilise the resulting imido moiety. Once the imido complex is formed the alkyne is introduced via 2+2 cycloaddition.

Figure 6 - Hydroamination of alkynes using a group IV metallocene

It is also possible to perform combined hydroamination/bicyclisation in which a C-N bond is formed followed by a C-C bond (Figure 7).

Figure 7 - Combined hydroamination/bicyclisation

Intermolecular hydroaminations are much more difficult to achieve using rare earth catalysis, this is mostly due to inefficient competition between strongly binding amines and weakly binding alkenes (the rate for these reactions is essentially zero-order with respect to the amine and first-order with respect to the alkene and lanthanide complex) for the free coordination sites, as a result a large excess of alkene is often necessary which contradicts the atom-efficient objective of hydroamination reactions. The rates for intermolecular hydroaminations are up to 350 times slower for alkenes and 1400 times slower for alkynes cf. intramolecular reactions. 2. Oxidative addition (NH activation) typically achieved with later transition metals in lower oxidation states, e.g. Rh(I) (Figure 8) This approach is appealing due to the higher tolerance for functionality, as well as lower air and moisture sensitivity, however, substrates must be activated to account for the weaker reactivity and so example substrates are alkynes, allenes, styrenes and strained alkenes.

Figure 8 - Hydroamination by oxidative addition

3. Alkene activation (-coordination) achieved using mildly Lewis acidic middle-late transition metals, e.g. Pd(II) (Figure 9)

[M] NHR

[M]

R NH2 [M]
Figure 9 - Hydroamination by alkene activation

RNH2

There are two possible cycles for this reaction (Error! Reference source not found.), the first involves oxidative amination and requires a co-oxidant, the second is a protonolysis and requires an acidic N-H bond.

Hydrophosphination Hydrophosphination reactions are almost identical to hydroaminations bar some differences in catalytic cycle thermodynamics. In Hydrophosphination the initial protonolysis of precatalyst is very slow (approx. 104 times slower than for hydroaminations)

Figure 10 - Comparison of organolanthanide mediated hydroamination and Hydrophosphination

Catalytic C-C Bond Forming Reactions Carbon-carbon bond formation is important for the conversion of small molecules into larger fragments, in particular polymer chemistry has grown in importance in recent decades due to move away from steel and timber as construction materials. The three most important types of carbon-carbon bond forming catalysis are; 1. Cross-coupling reactions RM + RX RR + MX This technique is most commonly employed in laboratories and the fine chemical sector, C-C bonds are formed under very mild conditions with very high functional group tolerance however due to the formation of waste products they are less commercially viable. There are several different approaches to cross coupling (Figure 11) and all except Heck coupling follow the same general mechanism; oxidative addition, transmetallation then reductive elimination (Figure 12). The rate determining step of these reactions is usually the oxidative addition and so ligand design Is focussed on net donors in an effort to maximise the electron density at the metal centre and stabilise the low-coordinate reactive species.

Figure 11 - Different types of cross-coupling reaction

Figure 12 - The general mechanism of a cross-coupling reaction

Palladium is most often used, although nickel, platinum, rhodium and copper have also been used. The palladium operates as Pd(0)/Pd(II), the former being a pre-requisite for the initial oxidative addition step. Precatalyst is often Pd(II) metal and must first be reduced to form the active Pd(0) catalyst, this is achieved by one of several routes; a) Reduction by an organometallic reagent LnPdX2 + 2 RM LnPdR2 L2Pd0+R-R

b) Reduction by an amine (often employed as a base) LnPdX2 + Et3N[Cl(L)2Pd(NEt3)]+X-HPdL2XPdL2+HX c) Reduction by a phosphine ligand Pd(OAc)2Pd(OAc)PPh3(PPh3)+Pd(PPh3)(OAc)PPh3OAcPdPPh3+Ac2O+Ph3PO The dibenzylideneacetone (dba) (Figure 13) ligand is weakly binding and are readily replaced by stronger donors, Pd(dba) complexes are often used as a precursor species for this reason.

Figure 13 - The structure of dibenzylideneacetone (dba)

Also used are phosphines (Figure 14), perhaps the most widely so, especially arylphosphines. The main advantage of phosphines is the tenability, the stearic and electronic properties are readily altered by changing the substituents at the phosphine, however, they are often air-sensitive and the catalytic species may not be robust.

Figure 14 - Example phosphine ligands used in carbon-carbon coupling

Stearically demanding, electron-rich phosphines (Figure 15) are increasingly popular as they enable transformations from aryl chlorides.

Figure 15 - Stearically demanding, electron-rich phosphine ligands

N-heterocylic carbenes (NHCs) (Figure 16) are becoming the preferred ligand choice as they are stronger donors, and so promote oxidative addition, this leads to air-stable, robust complexes that are easier to handle than the phosphine counterparts.

Figure 16 - N-heterocyclic carbene (NHC) ligands

Bulkier ligands are preferable as there is no requirement for a pre-equilibrium to form catalyst (i.e. there is a higher concentration of active catalyst at any given time in the reaction mixture than for smaller ligands) , they also produce complexes that are more reactive. Upon addition of the first substrate the smaller ligand complexes will form trans PdL2ArX (for Suzuki coupling) as this is more stable, addition of ArB-(OH)3M+1 then gives trans PdL2ArR which is less stable than the necessary cis form for successful reductive elimination. With bulkier ligands addition of ArX2 to the catalyst will give a T-shaped PdArLX complex (held in shape by agnostic interactions) which then gives cis PdLArR upon addition of ArB-(OH)3M+, this reactive form is capable of reductive elimination without a need for equilibria, another reason bulkier ligands are preferred. Heck coupling (Error! Reference source not found.) is unique of the above reactions as the carboncarbon bond forming step is a migratory insertion and not reductive elimination. There are two mechanistic pathways that both end by There are two mechanistic pathways that both end by hydride elimination, usually giving the stable trans product.

For Suzuki coupling, ArBX2 is a poor nucleophile that does not undergo transmetallation readily, hence conversion to the boronate, [RBX2(OH)]2 ArX rates are greatest when X=I and lowest when X=Cl

Figure 17 - The two mechanistic pathways for Heck coupling reactions

2. Alkene metathesis 3. Alkene dimerization/oligomerisation

Carbon-carbon coupling reactions have proven very difficult with alkyl halides, this is due to the decreased rate of oxidative addition (oxidative addition of an alkyl halide occurs by a radical process) and reductive elimination as well as the presence of -hydrides that compete through -hydride elimination. In aryl systems any -hydride present is unlikely to be capable of attaining the necessary square-planar transition state necessary for elimination to occur. Carbonylation Reactions The incorporation of CO as an aldehyde, ketone or carboxylic acid as part of a larger organic molecule can also be achieved via three routes; 1. Palladium catalysed carbonylation (using C-C coupling chemistry) RX + CO + NuH + Et3N RCONu + NHEt3+XThe catalytic cycle (Figure 18) is very similar to the previous C-C coupling chemistry but instead of transmetallation there is insertion of a carbon monoxide unit. The R group is either aryl or vinyl, for reasons previously discussed there are difficulties incorporating alkyl units with the nucleophile either an amine or an alcohol. The nucleophillic attack may be oxidative addition to give Pd IV(H)(Nu) intermediate or direct attack at the Pd bound carbon. This approach is very effective for intramolecular carbonylation.

Figure 18 - The catalytic cycle of Pd catalysed carbonylation

Pd0 is almost always ligated, often by phosphines or NHCs

Figure 19 - Formation of aldehydes via H2

Figure 20 - Carbonylation in conjunction with transmetallation

2. Polyketone formation (ethane/CO polymerisation)

Figure 21 - The general scheme of polyketone formation

Polyketones are high performance thermoplastics with many desirable properties; High strength High stiffness Impact and chemical resistance High melting points Ease of fabrication

The polymer is produced by sequential migratory insertions of the carbon monoxide and alkene substrate units (Figure 22). In non-protic solvents a methylating agent, usually SnMe4, is necessary to generate the active catalyst. The alkene insertion appears to be the rate limiting step but the kinetics are not always clear cut and in some cases CO insertion can be rate determining. Bidentate phosphine ligands are preferred as monodentate phosphines prefer the formation of alkyl propanoate, rates are greatest for diphosphines with butyl and propyl backbones as opposed to ethyl backbones. The larger backbone forces the substrate ligands closer together, increasing the overlap and improving the efficiency of the migratory insertion step.

Figure 22 - The catalytic cycle of polyketone formation

Chain termination is achieved by -hydride elimination (Figure 23), protonation of an alkyl palladium species (Figure 24) or nucleophillic attack at an acylpalladium complex (Figure 25).

Figure 23 - -hydride elimination during carbonylation

Figure 24 - Protonation of the alkyl palladium species

Figure 25 - Nucleophilic attack at an acylpalladium complex

3. Hydroformylation

Figure 26 Hydroformylation

Hydroformylation is the largest homogenous catalytic process in industry with over 15 billion of aldehydes (and subsequently, alcohols) are synthesised every year. Thermodynamically the formation of an alkane, without insertion of carbon monoxide, is preferred due largely to entropy, however the aldehyde is produced through kinetic control. The reaction is highly exothermic and is conducted in adiabatic conditions.

Using cobalt catalysts (Figure 27) it is possible to achieve 60-80% selectivity. Higher pressures of carbon monoxide decrease the reaction rate but increase the linear:branched ratio. Addition of PR 3 will stabilise the catalyst so that lower CO pressures and higher temperatures can be used without decomposition of the catalytic species. PR3 modified systems can also perform further reduction to give the alcohol directly.

Figure 27 - Cobalt catalysed hydroformylation

Rhodium catalysed hydroformylation (Figure 28) is also usually carried out in the presence of PPh3, this decreases the rate of the reaction but increases the selectivity for linear product. Unlike in the cobalt catalysed system, decreased CO pressure enhances the selectivity due to the preference for the diequatorial (ee) 4 form of [RhH(P2)(CO)2)]. The rate determining step is believed to be the oxidative addition of H2 in the absence of PPh3 but in the typical phosphine modified systems it is more likely the rate determining step is the alkene association and/or the following migratory insertion. Under standard conditions, in the phosphine modified model, the kinetics conform to a 1st order dependence on the concentration of alkene, a 0th order dependence on the concentration of H2 and negative order with respect to the concentrations of carbon monoxide and phosphine. 5

4 5

The diequatorial form is believed to be the reason for linear selectivity There is some disagreement in the literature, but this kinetic model is thought to be generally true

Figure 28 - Rhodium catalysed hydroformylation

The use of more basic trialkylphosphines will lead to less reactive systems whereas more electronwithdrawing ligands, e.g. phosphites, will reduce the electron density available at the metal, thus weakening CO binding and providing greater concentrations of the four-coordinate Rh(I) species in the catalytic cycle. Bulkier ligands will also promote loss of CO, thus enhancing rates. Phosphites are weaker -donors and stronger -acceptors than phosphines and have smaller cone angles. Generally the rates of hydroformylation and the selectivity for the linear product increase with increasing -acceptance by the phosphine or phosphite. Very bulky phosphites are extremely active but selectivity is dramatically reduced, the bulk prevents formation of HRh(L) 2(CO)2 and so HRh(L)(CO)2 is the active pre-catalyst, the ready loss of CO leads to high activity but can also lead to overconsumption of CO (the rate exceeds the rate of mass transport) leading to reactive, lowcoordinate Rh species that isomerise 1-alkenes rapidly. Diphosphites are a better route to linear products. In a move towards environmentally benign systems there is a push to discover alternative ligands to phosphines that will perform hydroformylation with a similar success. A water soluble system has been proposed using 3,3,3-phosphinidynetris(benzenesulfonic acid) trisodium salt (TPPTS), the resulting catalyst HRh(CO)[TPPTS Na3]3 has a -9 charge in aqueous solution. Ethene, propene and butene are all soluble enough that migration into the aqueous catalyst phase can occur and due to the catalysts total insolubility in organic solvents it is easy to separate the product from the catalyst by remigration. Rates are slower than with conventional rhodium phosphine systems due to the presence of lower concentrations in the water phase.