This paper describes a conceptual process design for the production of the model component phenol by a recombinant strain of the micro-organism Pseudomonas putida S12. The (bio)production of the inhibiting component phenol in a bioreactor is combined with direct product removal by membrane extraction (pertraction). Continuous fermentation with the pertraction unit inside the reactor (integrated process), is compared to fed-batch fermentation with the pertraction unit outside the reactor (nonintegrated process). In the nonintegrated process, the bioprocess is completely separated from the pertraction process. An extended model for fermentation, consisting of biomass growth and phenol production combined with product inhibition, cell removal, pertraction with 1-octanol, and regeneration of the solvent by distillation is described with the help of experimental and theoretical data. Running the fermentation process at a lower product concentration results in a more efficient substrate utilization into biomass and phenol. The disadvantage of the integrated process is the need for large distillation columns and a high energy input for solvent regeneration due to the low product concentration in the solvent and the high solvent fluxes. Economic evaluation of the two processes show that to obtain a return of investment of 15%, the product cost price of the integrated process is a factor three lower as compared to that of the nonintegrated process. The benefits of an integrated process will pay off even more for very toxic and inhibiting products that do not allow for a high concentration in the (bio)reactor as compared to phenol.

In situ phenol pertraction with 1-octanol has been experimentally studied to improve the production of the model component phenol by a recombinant strain of Pseudomonas putida S12. When the phenol concentration in the reactor reaches 2mM, the cells in fermentations without phenol removal are inhibited in growth and phenol production. Growth and phenol production stop after approximately 80h at a phenol concentration in the reactor of 3.8mM. When phenol is removed from the fermentation broth by pertraction, a lower maximum aqueous phenol concentration of 2.6mM is achieved, while the total phenol production increases to 132%, as compared to the fermentation without pertraction. There are indications that the volumetric productivity (mmolL^-1h^-1) increases slightly in the fermentations with in situ pertraction compared to the reference experiments. As expected, the amount of phenol produced per gram biomass (the specific productivity, mmolg^-1L^-1) remains constant in time for all fermentations. The use of pertraction for in situ phenol removal is compared to in situ second phase extraction, in situ solvent impregnated resins and in-stream pertraction. Although the system shows promising results, further modifications such as using a solvent with a higher partition coefficient can improve the overall performance.

This paper evaluates the applicability of aqueous solutions of Pluronics for the removal of phenol in a separation and regeneration process. Experimental results show that Pluronic P103 micelles allow extraction of phenol from aqueous solutions at 30°C. The phenol can be released due to the transition of the Pluronic micelles into unimers with a mild temperature switch from 30 to 8°C. Ultrafiltration membranes provide a barrier between the aqueous Pluronic stripping solution and the aqueous solution in a (bio)reactor containing the desired product. Additionally, a similar UF membrane is used to separate the micelles and unimers from water. Steady state model analysis of the proposed separation and regeneration process are performed to obtain a phenol mass flow rate in the product stream equal to the phenol production rate in the (bio)reactor. Furthermore, the process is analyzed for different process configurations and a cost estimation is made. The results show that for the model product phenol, the process costs are mainly determined by the required membrane area. The proposed process can be suited for products that allow for a higher critical concentration in the (bio)reactor as compared to phenol. The resulting higher driving force for membrane extraction will result in a decrease of the overall process costs. For products with a lower solubility in water, recovery is easy after regeneration of the micellar solvent.

The applicability of aqueous solutions of Pluronics for the removal of the model product phenol was evaluated. Phenol is a chemical that can be produced by a recombinant strain of the solvent tolerant bacterium Pseudomonas putida S12. However, the growth of the micro-organisms and the phenol production is inhibited at low phenol concentrations. The solubility of phenol in aqueous solutions containing micelles of Pluronics F108, F68, P105, L64, P104, P85, P103 and L122 was determined. Additionally, the regeneration behavior of the Pluronics was studied by determining the solubility of phenol in Pluronic unimers after lowering the temperature of the solution to below critical micelle temperature. The phase behavior of Pluronics in the presence of fermentation broth and the growth of P. putida S12 in the presence of Pluronics were studied. Finally, the results obtained in this study were used to estimate the amount of cooling energy needed to obtain 1 g of phenol using Pluronics P85 and P103.

A membrane reactor is presented for homogeneous catalysis in supercritical carbon dioxide with in situ catalyst separation. This concept offers the advantages of benign high-density gases, i.e., the possibility of achieving a high concentration of gaseous reactants in the same phase as the substrates and catalyst as well as easy catalyst localization by means of a membrane. For the separation of the homogeneous catalyst from the products an inorganic microporous membrane is used. The concept is demonstrated for the hydrogenation of 1-butene using a fluorous derivative of Wilkinson's catalyst [RhCl{P-(C ₆H₄-p-SiMe₂CH₂CH₂C ₈F₁₇)₃}₃]. The size of Wilkinson's catalyst, 2-4 nm, is clearly larger than the pore diameter, 0.5-0.8 nm, of the silica membrane. The membrane will, therefore, retain the catalyst, while the substrates and products diffuse through the membrane. Stable operation and continuous production of n-butane has been achieved at a temperature of 353 K and a pressure of 20 MPa. A turnover number of 1.2×10⁵ has been obtained during 32 h of reaction. The retention of the catalyst was checked using UV-vis spectroscopy and ICP-AAS; no rhodium or phosphorous species were detected at the permeate side of the membrane.

This paper describes a number of developments leading to sustainable catalytic and multi-phase reactors. For this purpose, traditional solvents are replaced by clean solvent alternatives such as supercritical fluids and aqueous surfactant systems, and processes are intensified using membrane reactors and contacting reactants at a molecular level.

The one-component transient permeation of carbon dioxide through an alumina-supported silica membrane is studied for pressures up to 2.0 × 10⁷ Pa, at temperatures of 296, 313, and 358 K. The permeation is obtained for gaseous, liquid, and supercritical carbon dioxide. For all conditions high carbon dioxide fluxes were obtained. The membrane has been tested with a maximum pressure difference between the feed and the permeate side of about 2.5 × 10⁶ Pa. At a feed-side pressure of 2.0 × 10⁷ Pa and at 358 K the permeance is equal to 8.0 × 10^-8 mol m^-2 s^-1 Pa^-1. The transient transport through the microporous membrane can be described with a single mass-transfer coefficient. For a feed-side pressure between 1.0 × 10⁵ and 200 × 10⁵ Pa and a temperature of 296 K the mass-transfer coefficient increases by a factor of 80. The fact that carbon dioxide has a high flux through this type of silica membrane opens the way for regeneration of carbon dioxide at supercritical conditions.

Membrane separation technology is successfully applied for the immobilization of a homogeneous catalyst (a (1H, 1H, 2H, 2H-perfluoroalkyl)-dimethylsilyl-substituted derivative of Wilkinson's catalyst) in a continuous process that uses supercritical carbon dioxide as solvent. The catalyst is separated from the products by a microporous silica membrane (see scheme).

Perfluoro-functionalized poly(propylene imine) dendrimers have been used as reactive extractants for anionic species and as phase transfer catalysts for two types of reactions. Different generations of dendrimers have been used for applications in carbon dioxide. First, the reactive extraction of the inorganic ions permanganate and dichromate, from an aqueous phase to a supercritical carbon dioxide phase, has been studied using the functionalized dendrimers. The yield for these extractions is rather low, indicating that the perfluoro-functionalized poly(propylene imine) dendrimers are not suited for oxidation reactions with these ions in carbon dioxide. Second, the poly(propylene imine) dendrimers have been used as a phase transfer catalyst for the halogen exchange reaction of benzyl chloride into benzyl bromide and for the esterification of oxalic acid with pentafluorobenzylbromide. In contrast with the low extraction yield for the inorganic permanganate and dichromate ions, a high reaction rate was obtained for the halogen exchange reaction with the functionalized dendrimers as phase transfer catalysts. It is found that the rate of reaction for both cases depends on the generation number of the dendrimer. For the higher-generation dendrimers, it is more difficult for the substrate to migrate from the bulk carbon dioxide phase to the interior of the dendrimer. This indicates that the mass transfer of the reagents from the organic phase into the interior of the dendrimer determines the reaction rate for the fourth and higher generation dendrimers. A model is proposed to describe the extraction and the phase transfer catalysis based on the fact that the modified dendrimers combine micellar properties with those of standard phase transfer catalysts. From the different experiments it can be concluded that the perfluoro-functionalized poly(propylene imine) dendrimers have great potential as multisite phase transfer catalysts.

Liquid and supercritical carbon dioxide can be used as a replacer for organic solvents, which potentially enables the development of clean processes. A general disadvantage of CO₂, however, is that it is a very poor solvent for high molecular weight or hydrophilic molecules. By using reverse micelles to overcome this problem, a wide variety of novel processes can be thought of. In this review, it is shown that the use of micellar systems in supercritical CO₂ leads to processes generating a minimum amount of waste and with a low energy requirement. It is also shown that especially downstream processing of biochemicals and polymerization reactions can benefit from this technique, as the number of consecutive process steps can be reduced significantly.