Research in this field started in the wake of the reports of SL-PC. Consisting of a catalyst-containing supported liquid layer for CF reactions in the gas phase, the concept was transferred to batch reactions, using a catalyst dissolved in a supported aqueous phase. This was first referred to as supported aqueous-phase catalysis (SAPC) by Davis in an article published in Nature in 1989.8 Later, the concept was extended, using a variety of names, but the essence has remained the same: a supported catalyst-philic phase.
In SAPC, a hydrophilic support such as silica is contacted with a water-soluble organometallic complex by aqueous-phase impregnation. After evacuation of the water phase, the organometallic complex becomes distributed on the support. Exposure to water vapor for a fixed time allows precise amounts of water to condense on the solid surface. The solid, coated by an aqueous film of catalyst, is placed in an immiscible organic phase that contains the reagents. The reactants diffuse from the organic phase into the porous solid, where they react at the water-organic interface, and the products diffuse back to the bulk organic phase. Along with the advantage of immobilization of the organometallic species, SAPC offers a high surface area for support, which translates into a high interfacial area, and the possibility of selectivity variations from bulk equilibrium product distribution through the effect of the interface.8,9 One of the differences between SAPC and SL-PC is that, while SL-PC is designed for gas-phase reagents, SAPC is very efficient for liquid-phase reagents. Figure 6.4 shows schematically an SAPC.
The prototype reaction was the hydroformylation of oleyl alcohol (water insoluble) with a water-soluble rhodium complex, HRh(CO)[P(m-C6H4SO3Na)3]3 (Figure 6.5). Oleyl alcohol was converted to the aldehyde (yield = 97%) using 2 mol % Rh with respect to the substrate and cyclohexane as the solvent, at 50 atmospheres CO/H2, and 100°C. The SAPCs were shown to be stable upon recycling, and extensive work proved that Rh is not leached into the organic phase. Since neither oleyl alcohol nor its products are water soluble, the reaction must take place at the aqueous -organic interface where Rh must be immobilized. Also, if the metal catalyst was supported on various controlled pore glasses with
different surface areas, the resulting conversions would be proportional to the interface area. It is noteworthy that the reaction did not proceed in liquid water as solvent with an unsupported catalyst. But, if the support was added, the reaction started, implying that the components self-assembled to yield the active SAPC. Rates were higher than with the corresponding homogeneous system.10,11
Analogously, the SAPC catalyzed hydroformylation reaction was carried out using other water-soluble metal complexes of Pt and Co. Pt complexes in the presence of an Sn co-catalyst underwent hydrolysis of the Pt-Sn bond, which led to lower reaction selectivity.12 With the corresponding Co catalyst, good hydrofor-mylation selectivities and conversions could be achieved, provided excess phos-phine was used.13 Other authors performed hydrogenation of a,b-unsaturated aldehydes using SAPC, and Ru and Ir water-soluble complexes.14
SAPC was then applied to asymmetric hydrogenation catalysis using the chiral Ru catalyst [Ru(BINAP-4SO3Na)(C6H6)Cl]Cl.15 17 The immobilization technique was the same as that used earlier. In this case, however, the presence of water caused the cleavage of the Rh-Cl bond, whose presence is critical for asymmetric induction. Therefore, a different nonvolatile hydrophilic liquid—inert with respect to the metal-halogen bond—was needed to replace water. The highly polar ethyl-ene glycol was chosen as the catalyst-philic phase, coupled with a nonpolar mixture of cyclohexane and chloroform as the hydrophobic organic phase. The asymmetric reduction shown in Figure 6.6 proceeded with 96% enantiometric excess (ee) at 100% conversion in the system—now called supported liquid-phase
SAPC Rh catal.
Figure 6.5 SAPC hydroformylation of oleyl alcohol.
catalysis (SL-PC)—without loss of Ru at a detection limit of 32 ppb. The SLP-Ru-BINAP-4SO3Na was found to be at least 50 times more active than its two-phase (EtOAc/H2O) analog, and only slightly less active than the homogeneous counterpart. In the absence of the silica support conversion did not exceed 2%. In this case, the stability of the SLPC catalyst also was confirmed by adding all the individual components of the SL-PC system separately, and by observing that they self-assemble and are more stable in the heterogeneous configuration than separated, which implies that the reverse process (i.e., separation of the SL-PC components) is unlikely.
A very similar kind of system was also described by Naughton and Drago shortly after.18 They described supported homogeneous film catalysts (SHFCS), where water-soluble Rh catalysts were dissolved in PEG catalyst-philic phases immobilized on a silica gel support, and used for hydroformylation reactions. What was different in this case was that the catalyst-philic phase was modified by inserting a nonionic surfactant—Surfynol 485—that enhanced catalytic activity, presumably by solubilizing the substrate in the catalyst-philic PEG phase, and increasing the concentration of the alkene available for hydroformylation. Other hydrophilic polymers, such as polyvinylpyrrolidinone, polyethelene oxide, and polyvinyl alcohol, were also tested as catalytic films. Finally, high-boiling polar compounds such as formamide and glycerol were also employed; however, all showed lower activity.
Analogously, over the years, Arai and co-workers have investigated silica-supported ethylene glycol as a catalyst-philic phase, which contained a metal precursor, for C-C bond-forming reactions, such as the Heck reaction. They describe a multiphasic system with an organic phase (solvent) that contains only reactants and products without any catalyst. The products could be recovered by simple filtration, and the catalyst recycled many times without deactivation, since it did not precipitate, thus making the catalytic system stable and reusable (Figure 6.7).19,20
The method of catalyst immobilization is one for the reasons for the success of the SAPC approach. Rather than covalently linking an organometallic complex to a support—which usually leads to loss of catalytic efficiency and leaching of the metal—it is the catalyst-philic phase that is immobilized.
Horvath recognized that SAPC solved the problem posed by the solubility of lypophilic substrates in aqueous biphasic catalysis with water-soluble homogeneous catalysts.21 He compared biphasic aqueous-organic catalysis with SAPC, in order to clarify whether in SAPC the catalyst remained dissolved in the
Organic phase Organic phase
Solvent, Reactants Solvent, Products
KOAc. HI solid
aqueous phase, or if it works at the aqueous-organic interface. High-pressure infrared (IR) studies indicated that water acts as an immobilization agent rather than a solvent. This was apparent from the fact that in SAPC water (not the catalyst) leached from the support in an amount that left only two monolayers of water on the hydrophilic support. This led to the theory that the water layer holds the water-soluble phosphines by hydrogen bonding the hydrated Na-sulphonate groups to the surface (Figure 6.8). The metal coordinated by the phosphines is therefore found precisely at the interface between the supported aqueous phase and the bulk organic phase (Figure 6.9), and is readily available for hydrophobic substrates.
M S = substrate
Supported solvent | ; L = ligand
Figure 6.9 Schematic coordination of water-soluble phosphines with a metal at the water-organic interface.
Since these first reports on the use of SAPC, the concept has been applied to a large number of reactions, with different metals and ligands. It is peculiar that the technology was renamed "glass bead technology" in a review on the topic.22 The investigated reactions range from hydroformylation, to hydrogenation, to Wacker oxidation, to Heck couplings,23 to Suzuki couplings, to allylic substitution; using Rh, Pd, Co, Pt, and in the presence of supported phases, water, PEGs, ethylene glycol.
A recent technological modification of SLPC and SAPC introduced supercritical carbon dioxide (scCO2) as the "organic" phase.24 In a SAPC-scCO2 system, using Ru[P(m-C6H4SO3Na)3]3, it was possible to efficiently reduce cinnamalde-hyde to the corresponding unsaturated alcohol with high selectivity (96%). Here the high solubility of reactant gases in scCO2 overcomes G-L-L mass-transfer limitations.
Recently the concept has been reformulated and applied to a new class of solvents: ionic liquids. These were supported on silica—covalently anchored or adsorbed—and used as catalyst-philic phases for metal complexes in hydroformylation and hydrogenation reactions. The concept is identical to that of SAPC, but the acronym was modified to SILC (supported ionic liquid catalysis). The first type of reaction that was reported was the hydroformylation of 1-hexene, using Rh as catalyst.25 The ionic liquid phase was made by butylmethylimidazolium hexafluorophosphate ([bmim][PF6]) supported on silica, modified by covalently anchoring ionic liquid fragments (Figure 6.10). What was obtained was a system where [bmim][PF6] was supported on silica and contained the active catalytic species HRh(CO)[P(m-C6H4SO3Na)3]3 plus an excess of free phosphine. The catalyst was a free-flowing powder that was made to react with the substrate and CO/H2. The activity of this SILC system was slightly higher than conventional biphasic catalysis; however, leaching became significant unless free phosphine ligand was added in the ionic liquid layer.
SILC was also used without covalently anchoring the ionic liquid fragment to the silica support. In this case, [bmim][PF6] was simply added to silica in acetone together with the catalyst.26 [Rh(norbornadiene)(PPh3)2]PF6 and the solvent evaporated to yield the supported catalyst-philic phase. Catalyst evaluation on the hydrogenation of model olefins showed enhanced activity in comparison to homogeneous and biphasic reaction systems, in analogy to Davis's observations.9 Also
Figure 6.10 Immobilization of an ionic liquid used for SILC hydroformylation.
Figure 6.10 Immobilization of an ionic liquid used for SILC hydroformylation.
in this case, a concentration effect was invoked to explain the better performance, as most of the reaction occurs at the interphase (Figure 6.11). The SILC showed good stability. The same catalyst system could be used 18 times without loss of activity. As far as leaching of the metal is concerned, Rh remained below the 33-ppb (parts per billion) detection limit, and the residual organic phase did not show catalytic activity. In addition, scanning electron microscopy (SEM) and transmission electron microscope (TEM) images of used and fresh catalysts were identical, indicating no clustering of the metal.
An immediate extension of SAPC and SILC was the continuous-flow analogue performed by Fehrmann and co-workers, called supported ionic liquid phase (SILP).27 The catalyst was prepared by impregnation of the rhodium precursor,
PPh2 PPh2 Sulfoxantphos
PPh2 PPh2 Sulfoxantphos
Figure 6.12 Structure of the ligands sulfoxantphos and NORBOS-Cs3.
the sulfonated biphosphine ligand (sulfoxantphos, Figure 6.12) on silica, in the presence of the ionic liquids [bmim][PF6], and halogen-free [bmim][n-C8Hi7_ OSO3]. The continuous-flow gas-phase hydroformylation of propene was demonstrated at 10 bar and 100°C. It was shown that the reaction proceeds with a ligand/Rh ratio between 10 and 20, indicating that the active ligand-containing species are formed in situ. What was also evident was that the catalytic performance was scarcely influenced by the anion of the ionic liquid. Finally, it was observed that deactivation of the catalyst could be prevented by tuning the ligand/Rh ratio.
The use of a catalyst-philic phase made by [bmim][n-C8H17OSO3] also addresses environmental issues and should be noted. In fact, while the PF6 anion is certainly useful for exploratory studies, it easily hydrolyzes and generates HF.
The scope of SILP was extended by investigating charged monophosphine ligands, as well as liquid-phase CF hydroformylation.28 The latter was demonstrated on 1-octene using the SILP Rh-(NORBOS-Cs3)/ [bmim][PF6]/silica catalyst. The authors recognize that the supported catalyst-philic phase offers the significant advantage of very efficient ionic liquid use.
Both SILC and SILP offer the advantage over SAPC of using ionic liquids instead of water. The low vapor pressure ensures that the supported phase remains liquid under the reaction conditions, and that it is retained during continuous flow operation.
Onium salts, such as tetraethylammonium bromide (TEAB) and tetra-n-butylammonium bromide (TBAB), were also tested as PTCs immobilized on clay. In particular, Montmorillonite K10 modified with TBAB efficiently catalyzed the substitution reaction of a-tosyloxyketones with azide to a-azidoketones, in a biphasic CHCl3/water system (Figure 6.13).29,30 The transformation is a PTC reaction, where the reagents get transferred from the liquid to the solid phase. The authors dubbed the PTC-modified catalyst system "surfactant pillared clay" that formed a "thin membrane-like film at the interface of the chloroform in water emulsion," that is, a third liquid phase with a high affinity for the clay. The advantages over traditional nucleophilic substitution conditions were that the product obtained was very pure under these conditions and could be easily recovered without the need for dangerous distillation steps.
Surfactant pillared clay
R1 + NaOTs
Figure 6.13 Substation of a-tosyloxyketones with azide to give a-azidoketones.
Was this article helpful?