Jonathan A. Iggo

A Multilateral Mechanistic Study into Asymmetric Transfer Hydrogenation in Water

The mechanism of aqueous-phase asymmetric transfer hydrogenation (ATH) of acetophenone (acp) with HCOONa catalyzed by Ru-TsDPEN has been investigated by stoichiometric reactions, NMR probing, kinetic and isotope effect measurements, DFT modeling, and X-ray structure analysis. The chloride [RuCl(TsDPEN)(p-cymene)] (1), hydride [RuH(TsDPEN)(p-cymene)] (3), and the 16-electorn species [Ru(TsDPEN-H)(p-cymene)] (4) were shown to be involved in the aqueous ATH, with 1 being the precatalyst, and 3 as the active catalyst detectable by NMR in both stoichiometric and catalytic reactions. The formato complex [Ru(OCOH)(TsDPEN)(p-cymene)] (2) was not observed; its existence, however, was demonstrated by its reversible decarboxylation to form 3. Both 1 and 3 were protonated under acidic conditions, leading to ring opening of the TsDPEN ligand. 4 reacted with water, affording a hydroxyl species. In a homogeneous DMF/H(2)O solvent, the ATH was found to be first order in the concentration of catalyst and acp, and inhibited by CO(2). In conjunction with the NMR results, this suggests that hydrogen transfer to ketone is the rate-determining step. The addition of water stabilized the ruthenium catalyst and accelerated the ATH reaction; it does so by participating in the catalytic cycle. DFT calculations revealed that water hydrogen bonds to the ketone oxygen at the transition state of hydrogen transfer, lowering the energy barrier by about 4 kcal mol(-1). The calculations also suggested that the hydrogen transfer is more step-wise in nature rather than concerted. This is supported to some degree by the kinetic isotope effects, which were obscured by extensive H/D scrambling.

Direct Acylation of Aryl Bromides with Aldehydes by Palladium Catalysis

A new protocol for the direct acylation of aryl bromides with aldehydes is established. It appears to involve palladium-amine cooperative catalysis, affording synthetically important alkyl aryl ketones in moderate to excellent yields in a straightforward manner, and broadening the scope of metal-catalyzed coupling reactions.

Hydrogen-Bonding-Promoted Oxidative Addition and Regioselective Arylation of Olefins with Aryl Chlorides

The first, general, and highly efficient catalytic system that allows a wide range of activated and unactivated aryl chlorides to couple regioselectively with olefins has been developed. The Heck arylation reaction is likely to be controlled by the oxidative addition of ArCl to Pd(0). Hence, an electron-rich diphosphine, 4-MeO-dppp, was introduced to facilitate the catalysis. Solvent choice is critical, however; only sluggish arylation is observed in DMF or DMSO, whereas the reaction proceeds well in ethylene glycol at 0.1-1 mol % catalyst loadings, displaying excellent regioselectivity. Mechanistic evidence supports that the arylation is turnover-limited by the oxidative addition step and, most importantly, that the oxidative addition is accelerated by ethylene glycol, most likely via hydrogen bonding to the chloride at the transition state as shown by DFT calculations. Ethylene glycol thus plays a double role in the arylation, facilitating oxidative addition and promoting the subsequent dissociation of chloride from Pd(II) to give a cationic Pd(II)-olefin species, which is key to the regioselectivity observed.

Branching out with aminals: microporous organic polymers from difunctional monomers

Microporous organic polymers (MOPs) have been prepared via one-pot polycondensation reactions between aldehydes and amines. Primary amines were reacted with imines to produce porous polymers from A2 + B2 monomer combinations. The resulting networks exhibit BET surface areas in the range 500–600 m2 g−1. This approach opens up the possibility of synthesising MOPs using readily-available and inexpensive precursors.

Phosphine Ligands in the Palladium‐Catalysed Methoxycarbonylation of Ethene: Insights into the Catalytic Cycle through an HP NMR Spectroscopic Study

Novel cis-1,2-bis(di-tert-butyl-phosphinomethyl) carbocyclic ligands 6-9 have been prepared and the corresponding palladium complexes [Pd(O(3)SCH(3))(L-L)][O(3)SCH(3)] (L-L=diphosphine) 32-35 synthesised and characterised by NMR spectroscopy and X-ray diffraction. These diphosphine ligands give very active catalysts for the palladium-catalysed methoxycarbonylation of ethene. The activity varies with the size of the carbocyclic backbone, ligands 7 and 9, containing four- and six-membered ring backbones giving more active systems. The acid used as co-catalyst has a strong influence on the activity, with excess trifluoroacetic acid affording the highest conversion, whereas excess methyl sulfonic acid inhibits the catalytic system. An in operando NMR spectroscopic mechanistic study has established the catalytic cycle and resting state of the catalyst under operating reaction conditions. Although the catalysis follows the hydride pathway, the resting state is shown to be the hydride precursor complex [Pd(O(3)SCH(3))(L-L)][O(3)SCH(3)], which demonstrates that an isolable/detectable hydride complex is not a prerequisite for this mechanism.

Preparation and characterisation of solvent-stabilised nanoparticulate platinum and palladium and their catalytic behaviour towards the enantioselective hydrogenation of ethyl pyruvate

Solvent-stabilised Pt and Pd nanoparticles, of size range 2.3–2.8 nm and 2.7–3.8 nm, respectively, have been prepared by metal vapour synthesis routes, characterised by transmission electron microscopy (TEM), and their behaviour as catalysts for the enantioselective hydrogenation of ethyl pyruvate (EP) investigated; comparisons have been effected with the performance of standard supported Pt and Pd catalysts. Cinchona alkaloid-modified Pt nanoparticles display parallel behaviour to that exhibited by their conventional supported counterparts both in terms of the sense of the enantioselectivity in the ethyl lactate product and in the acceleration in reaction rate relative to the unmodified system. With Pd, however, significant differences are noted. Here, the sense of the enantioselectivity relative to that reported previously over conventional supported catalysts is reversed, i.e., an (R)- vs. (S)-enantiomer switch occurs, and a rate acceleration rather than retardation is noted on cinchona alkaloid modification. The Pt particle size distribution shows a higher degree of monodispersity after use in catalysis, although the average particle size remains essentially unchanged, whereas the behaviour of the Pd nanoparticles shows evidence of concentration dependence, lower concentrations showing Pt-like behaviour but more highly concentrated preparations showing evidence of significant aggregation during catalysis. With Pt catalysts, the presence of water as a component of the ketonic solvent system is shown to result in a significant acceleration in overall reaction rate with both conventional supported catalysts and their solvent-stabilised counterparts. In sharp contrast, totally aqueous-based colloidal platinum preparations, obtained by conventional salt reduction, display very low reaction rates and enantioselectivities.

Cooperative Catalysis through Noncovalent Interactions

Noncovalent interactions, such as hydrogen bonding, electrostatic, p–p, CH–p, and hydrophobic forces, play an essential role in the action of nature s catalysts, enzymes. In the last decade these interactions have been successfully exploited in organocatalysis with small organic molecules. In contrast, such interactions have rarely been studied in the wellestablished area of organometallic catalysis, where electronic interactions through covalent bonding and steric effects imposed by bound ligands dictate the activity and selectivity of a metal catalyst. An interesting question is: What happens when an organocatalyst meets an organometallic catalyst? This unification has already created an exciting new space for both fields: cooperative catalysis, where reactants are activated simultaneously by both types of catalyst, thereby enabling reactivity and selectivity patterns inaccessible within each field alone. However, the mechanisms by which the two catalysts cooperatively effect the catalysis remain to be delineated. We recently found that combining an achiral iridium catalyst with a chiral phosphoric acid allows for highly enantioselective hydrogenation of imines (Scheme 1). To gain insight into the mechanism of this metal–organo cooperative catalysis, we studied the catalytic system with a range of techniques, including high pressure 2D-NMR spectroscopy, diffusion measurements, and NOEconstrained computation. Herein we report our findings. To evaluate the mechanism, a simplified achiral complex C was used, which leads to [C][A ] upon mixing, in situ or ex situ, with the chiral phosphoric acid HA through protonation at the amido nitrogen (Scheme 1). In the asymmetric hydrogenation of the model ketimine 1a, [C][A ] afforded 95% ee and full conversion. On the basis of related studies, the hydrogenation can be broadly explained by the catalytic cycle shown in Scheme 1, that is, [C][A ] activates H2 to give the hydride D and protonated 1a, which forms an ion pair with the phosphate affording [1a][A ]; hydride transfer furnishes the amine product 2a while regenerating [C][A ]. Questions pertinent to possible iridium–phosphate cooperation then arise: 1) How does the chiral phosphoric acid induce asymmetry in the hydrogenation? and 2) Does the enantioselectivity result from D being formed enantioselectively from [C][A ], from the phosphate salt [1a][A ], or from interactions involving all three components? We looked first at how the formation of hydride D and its transfer into the substrate are influenced by the chiral acid HA. The studies were carried out in CH2Cl2 or CD2Cl2 owing to the low solubility of the various metal complexes in toluene. The catalytic hydrogenation is feasible in both solvents, giving a 95% ee in toluene and 85 % ee in CH2Cl2 in the case of hydrogenation of 1a with C and HA under the conditions given in Scheme 1. The solution NMR studies show that the ionic complex [C][A ] is formed instantly on protonation of C (0.05 mmol) with one equivalent HA in CD2Cl2 (0.5 mL). Under H2 pressure (> 1 bar), proton transfer from a [C]–H2 dihydrogen intermediate (not observed) to 1a converts [C] into the hydride D and affords the salt [1a] [A ]. Formation of D took place instantly even at 78 8C, and it is observed during catalytic turnover, thus indicating that the hydrogenation is rate-limited by the hydride transfer step. Scheme 1. Hydrogenation of imine with achiral C and chiral acid HA (PMP = p-methoxyphenyl, Ar= 2,4,6-triisopropylphenyl, Ts = tosyl, Bn = benzyl).

Analysis of the mesh size in a supramolecular hydrogel by PFG-NMR spectroscopy

Pulsed field gradient NMR (PFG-NMR) spectroscopy has been used to determine the network mesh size in stable hydrogels formed upon addition of Ca2+ to solutions of naphthalene diphenylalanine (2FF). At pH 12, the solutions at 0.55 wt% 2FF comprise worm-like micelles. Addition of Ca2+ results in cross-linking of these micelles. The self-diffusion of dextran guests of nominal 6, 40, 70, 100, 500, 670, 1400 and 2000 kDa, possessing hydrodynamic diameters (2Rh) similar to the expected pore sizes in these systems, was studied both in the precursor micellar solutions and in the hydrogels. The diffusivity of probes with 2Rh < 40 nm is restricted to a similar extent in both types of network with diffusion coefficients scaling as ∼Mr−0.5, where Mr is the nominal mass of the probe, consistent with relatively unrestricted diffusion. Diffusion coefficients fit well the equation Dn/Do = exp(−Rh/ξ), where Dn and Do are the diffusion coefficients in the presence and absence of network respectively and ξ is the mesh size, giving a mesh size of approximately 40 nm. The heaviest ca. 10% of the nominal 2000 kDa dextran fraction having approximate mass and hydrodynamic diameter 3300 kDa and 84 nm respectively was almost immobilised by the gel, consistent with this estimate of the mesh size. The restriction was much weaker in the micellar solution, which is attributed to the transient nature of this micellar network in the absence of Ca2+. Finally, the mesh size for micellar solutions prepared at 1.1 wt% 2FF is smaller than that of micellar solutions prepared at lower concentrations of 2FF. However, the corresponding gels have a larger mesh size than those prepared at lower concentrations of 2FF. We attribute this to increased fibre aggregation at the higher 2FF concentration. This correlates with lower rheological moduli at higher 2FF concentrations.

Magnetically Aligned Supramolecular Hydrogels

The magnetic-field-induced alignment of the fibrillar structures present in an aqueous solution of a dipeptide gelator, and the subsequent retention of this alignment upon transformation to a hydrogel upon the addition of CaCl2 or upon a reduction in solution pH is reported. Utilising the switchable nature of the magnetic field coupled with the slow diffusion of CaCl2, it is possible to precisely control the extent of anisotropy across a hydrogel, something that is generally very difficult to do using alternative methods. The approach is readily extended to other compounds that form viscous solutions at high pH. It is expected that this work will greatly expand the utility of such low-molecular-weight gelators (LMWG) in areas where alignment is key.

The Synthesis of, and Characterization of the Dynamic Processes Occurring in, Pd(II) Chelate Complexes of 2-pyridyldiphenylphosphine

Pd(II) complexes in which 2-pyridyldiphenylphosphine (Ph(2)Ppy) chelates the Pd(II) centre have been prepared and characterized by multinuclear NMR spectroscopy and by X-ray crystallographic analysis. trans-[Pd(kappa(1)-Ph(2)Ppy)(2)Cl(2)] is transformed into [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)Ppy)Cl]Cl by the addition of a few drops of methanol to dichloromethane solutions, and into [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)Ppy)Cl]X by addition of AgX or TlX, (X = BF(4)(-), CF(3)SO(3)(-) or MeSO(3)(-)). [Pd(kappa(1)-Ph(2)Ppy)(2)(p-benzoquinone)] can be transformed into [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)Ppy)(MeSO(3))][MeSO(3)] by the addition of two equivalents of MeSO(3)H. Addition of further MeSO(3)H affords [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)PpyH)(MeSO(3))][MeSO(3)](2). Addition of two equivalents of CF(3)SO(3)H, MeSO(3)H or CF(3)CO(2)H and two equivalents of Ph(2)Ppy to [Pd(OAc)(2)] in CH(2)Cl(2) or CH(2)Cl(2)-MeOH affords [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)Ppy)X]X, (X = CF(3)SO(3)(-), MeSO(3)(-) or CF(3)CO(2)(-)), however addition of two equivalents of HBF(4).Et(2)O affords a different complex, tentatively formulated as [Pd(kappa(2)-Ph(2)Ppy)(2)]X(2). Addition of excess acid results in the clean formation of [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)PpyH)(X)]X(2). In methanol, addition of MeSO(3)H and three equivalents of Ph(2)Ppy to [Pd(OAc)(2)] affords [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)Ppy)(2)][MeSO(3)](2) as the principal Pd-phosphine complex. The fluxional processes occuring in these complexes and in [Pd (kappa(1)-Ph(2)Ppy)(3)Cl]X, (X = Cl, OTf) and the potential for hemilability of the Ph(2)Ppy ligand has been investigated by variable-temperature NMR. The activation entropy and enthalpy for the regiospecific fluxional processes occuring in [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)Ppy)(2)][MeSO(3)](2) have been determined and are in the range -10 to -30 J mol(-1) K(-1) and ca. 30 kJ mol(-1) respectively, consistent with associative pathways being followed. The observed regioselectivities of the exchanges are attributed to the constraints imposed by microscopic reversibility and the small bite angle of the Ph(2)Ppy ligand. X-Ray crystal structure determinations of trans-[Pd(kappa(1)-Ph(2)Ppy)(2)Cl(2)], [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)Ppy)Cl][BF(4)], [Pd(kappa(1)-Ph(2)Ppy)(2)(p-benzoquinone)], trans-[Pd(kappa(1)-Ph(2)PpyH)(2)Cl(2)][MeSO(3)](2), and [Pd(kappa(1)-Ph(2)Ppy)(3)Cl](Cl) are reported. In [Pd(kappa(2)-Ph(2)Ppy)(kappa(1)-Ph(2)Ppy)Cl][BF(4)] a donor-acceptor interaction is seen between the pyridyl-N of the monodentate Ph(2)Ppy ligand and the phosphorus of the chelating Ph(2)Ppy resulting in a trigonal bipyramidal geometry at this phosphorus.