Membrane catalyst retention

The catalysis was performed batch-wise (Figure 4.36). After reaching ca. 90% conversion, the bulk phase was replaced and similar turnover frequencies (TOF) of about 25 h"1 were obtained in the following three runs 2, 3 and 4. When the catalyst capsule was removed, no further activity was detected. Furthermore, the Ru content in the bulk phase was always below the detection limit of AAS, which shows good catalyst retention by the membranes used. 

Methods for (bio)catalyst retention include (i) heterogenization on supports (ii) recovery through phase change, such as precipitation or extraction of the catalyst and (iii) membrane filtration of a homogeneous catalyst. All methods, in principle, enable repeated use of a chiral catalyst without much loss of activity or selectivity. In a recent review (Kragl, 2001), examples are given from laboratory and... 

Recovery of catalyst from converted oil. Another way to process the residues is to add hydrogen to effect hydroconversion which avoids the formation of a large quantity of asphalt Solid catalyst is formed afterward by reaction. Membrane filtration is used to separate the converted oil from the catalyst This makes it possible to partially recycle the catalyst to the reactor. Alumina and zirconia membranes with pore diameters ranging from 30 to 600 nm have been tested for this application. The membrane with a pore diameter of 30 nm yields a stable flux and a catalyst retention better than 98% [Deschamps et al., 1989). Concentration polarization is significant and requires a high crossflow velocity and temperature to overcome it. 

Figure 9 Proportion of a soluble catalyst remaining in the reactor vs. the number of exchanged reactor volumes in a continuously operated membrane reactor for different catalyst retentions (R = 99.99%, 99.994, 99%, and 95%).

Even when the inert membrane is used just for catalyst retention in ultra- or nano-filtration membrane reactors, the fine control of membrane porosity is a strong advantage of polymeric membranes when compared with inorganic ones. 

After reaching its maximum productivity (after ca. 8 hours.) the [Gl]-Nii2 showed a fast deactivation when applied in continuous catalysis performed in a membrane reactor (Figure 4.12). The fast loss of activity cannot be due to a lack of retention of the catalyst. Due to the high retention measured, this process should be much slower. A model study revealed that this deactivation process probably takes place by the formation of insoluble Ni(III) species (see Section 4.5 for further details). 

Application of the largest dendritic catalyst 8 (Figure 4.15) in a continuous process showed activity over 15 exchanged reactor volumes (Figure 4.16). The decrease in activity caused by wash out was calculated to be only 25% (retention of ligand 98.1%). The drop in activity was therefore ascribed to the decomposition of the palladium catalyst. Addition of membrane material to batch catalysis experiments did not change the conversion showing that this was not the cause of decomposition. 

After an induction period of ca. 9 hours, the maximum productivity was reached. This was followed by a decrease in activity, which cannot solely be explained by the lack of retention. Calculations showed that at least 20% of the catalyst should still remain in the reactor after 80 h. Additional to this wash-out effect, a deactivation process took place, visible by precipitation of palladium black on the surface of the membrane. Although the catalytic system suffered from deactivation, its selectivity towards 3-phenylbut-l-ene was excellent, being 98% and 85% for the G0- and Gr catalysts respectively. 

After an activation period of 4 h, the conversion showed a maximum of 40% followed by a steady decrease in conversion (Figure 4.38). Overnight, the pressure was decreased to 6 MPa and the needle valve on the permeate side was closed. This shutdown procedure caused the catalyst to precipitate and no reaction occurred anymore. The precipitated catalyst can be used for a new cycle by pressurization of the membrane reactor, redissolving the catalyst. At the end of the third run the conversion had dropped to 33%. A TON of 1.2xl05 in 32 h (t 62 min) was obtained. ICP-AAS analysis of the permeate stream indicated complete retention of the catalyst. The authors propose possible traces of oxygen as the cause of the decrease in activity of the catalyst. 

Membrane filtration using a polyaramide membrane [56] showed a retention of more than 99.8%. Application of this catalyst in a continuously operated membrane reactor showed conversion for more than 150 h. The ee dropped from 80% in the beginning (non-bonded analogue 97%) to about 20% after 150 h. The average ee for the first 80 h was 50%. 

The reduction of acetophenone was carried out at r.t. giving 86% yield with an ee of 97%. This is similar to the ee obtained with unbound analogues. A limited study was conducted on the retention of the catalyst by nanofiltration. It was found that the compound could be retained in the membrane reactor but no specific details were given about these measurements. 

After a hydroformylation run, the reaction solution was subjected to ultrafiltration using an asymmetric polyethersulfone membrane (MWCO 50 kDa) supplied by Sartorius. A retention of 99.8% was found. When the catalyst solution was recycled, virtually the same catalytic activity was observed again (165 TO h 1). Repetitive recycling experiments resulted in 2-7% loss of rhodium, which was subscribed to partial oxidation of the phosphine ligand. 

Cross-flow filtration systems utilize high liquid axial velocities to generate shear at the liquid-membrane interface. Shear is necessary to maintain acceptable permeate fluxes, especially with concentrated catalyst slurries. The degree of catalyst deposition on the filter membrane or membrane fouling is a function of the shear stress at the surface and particle convection with the permeate flow.16 Membrane surface fouling also depends on many application-specific variables, such as particle size in the retentate, viscosity of the permeate, axial velocity, and the transmembrane pressure. All of these variables can influence the degree of deposition of particles within the filter membrane, and thus decrease the effective pore size of the membrane. 

The globular shape of dendritic macromolecules with a persistent nanosize and radius should allow easy separation or retention by ultra- or nanofiltration membranes. This concept of separating the catalysts from the product/substrate... 

Ultrafiltration has been used for the separation of dendritic polymeric supports in multi-step syntheses as well as for the separation of dendritic polymer-sup-ported reagents [4, 21]. However, this technique has most frequently been employed for the separation of polymer-supported catalysts (see Section 7.5) [18]. In the latter case, continuous flow UF-systems, so-called membrane reactors, were used for homogeneous catalysis, with catalysts complexed to dendritic ligands [23-27]. A critical issue for dendritic catalysts is the retention of the catalyst by the membrane (Fig. 7.2b, see also Section 7.5). 

Kragl 13) pioneered the use of membranes to recycle dendritic catalysts. Initially, he used soluble polymeric catalysts in a CFMR for the enantioselective addition of Et2Zn to benzaldehyde. The ligand a,a-diphenyl-(L)-prolinol was coupled to a copolymer prepared from 2-hydroxyethyl methyl acrylate and octadecyl methyl acrylate (molecular weight 96,000 Da). The polymer was retained with a retention factor > 0.998 when a polyaramide ultrafiltration membrane (Hoechst Nadir UF PA20) was used. The enantioselectivity obtained with the polymer-supported catalyst was lower than that obtained with the monomeric ligand (80% ee vs 97% ee), but the activity of the catalyst was similar to that of the monomeric catalyst. This result is in contrast to observations with catalysts in which the ligand was coupled to an insoluble support, which led to a 20% reduction of the catalytic activity. 

Using the addition of diethylzinc to benzaldehyde catalyzed by an a,a-diphenyl-L-prolinol bound to a methacrylate polymer 1 (M = 96 000 g moT ) (Fig. 3.1.3, left) the ttn for the catalyst could be increased 10-fold (from 50 to 500) in comparison with the free catalyst via membrane retention [9]. The product was obtained in 80% enantiomeric excess (ee). 

While a number of dendritic catalysts have been described, catalyst recyclization in most cases is an unsolved problem. Diaminopropyl-type dendrimers bearing Pd-phosphine complexes have been retained by ultra- or nanofiltration membranes, and the constructs have been used as catalysts for allylic substitution in a continuously operating chemzyme membrane reactor (CMR) (Brinkmann, 1999). Retention rates were found to be higher than 99.9%, resulting in a sixfold increase in the total turnover number (TTN) for the Pd catalyst. 

Retention of Heterogenized Chiral Chemical Catalysts in a Membrane Reactor... 

Retention of homogeneous catalysts can be achieved by binding the low-molecular-weight catalyst to a dendrimer, to an already formed polymeric backbone, or to a polymerizable monomeric unit which is polymerized subsequently. Currently, the disadvantages concern the durability of the nanofiltration or ultrafiltration membranes, which, after all, in most cases have not been slated for use in... 

U. Kragl, Polymer enlarged oxazaboro-lidines in a membrane reactor enhandng effectivity by retention of the homogeneous catalyst, Tetrahedron Asymm. 1998, 9, 691-696. 

In the preceding section, we analyzed an immobilized enzyme process and calculated some important parameters such as productivity. In this section, we investigate another process configuration for retaining biocatalysts, the membrane reactor. The advantages and disadvantages of immobilization and membrane retention have already been discussed in Chapter 5. As in the case of immobilization, retention of catalyst by a membrane vastly improves biocatalyst productivity, a feature important on a processing scale but usually not on a laboratory scale. 

We will calculate the reactor performance itself as well as the productivity over time we will see that productivity is influenced by retention of the enzyme catalyst as much as by its deactivation behavior. In the schematic of an enzyme membrane reactor... 

---