(71e) Pilot Plant Development of Molecular Separations by Organic Solvent Nanofiltration

Organic Solvent Nanofiltration (OSN) separates molecules present in solutions of organic solvents. It is the same in principle as the well known RO/nanofiltration used in water treatments, except that the separations are based in organic solvents. This opens up a whole range of separations applications in organic processes – concentration, purification, catalyst retention and solvent recovery are just a few of these.

In OSN, a pressure gradient is applied across a solvent stable nanoporous film, inducing transport of liquid across the film. Depending on relative permeation rates, molecular species in a feed liquid stream can be concentrated and separated. Typically OSN will use a small fraction (less than 10%) of the energy of thermal methods such as evaporation and distillation, and can work at temperatures around ambient. Challenges for application of OSN technology in the refining, chemicals and pharmaceutical industry have been: (i) the lack of membranes with wide solvent resistance, and which retain their nanostructure with exposure to aggressive liquids such as dipolar aprotic solvents (DMF, DMSO, THF); and (ii) acceptance of the introduction of a new process technology to large scale chemical processes.

With respect to the first challenge, this presentation will describe the recent development of two new classes of commercially available OSN membranes and modules, and their applications to processes typically used in chemical manufacturing routes. The DuraMem^TM series OSN membranes are made through post-formation processing including crosslinking of polymer membranes, which makes them stable in polar solvents such as acetone, DMF and THF. The PuraMem^TM S series are thin film composite membranes which show exceptional performance in apolar solvents. Both these membranes are available for lab testing as flat sheets, and as spiral wound cartridges for scale up.

To meet the second challenge, the development of organic processes where OSN is used as a molecular separation step proceeds from lab bench scale to kilo/pilot scale, and then to process scale. The laboratory bench testing programme must yield key performance data to show that OSN is a viable technical solution, and that the technology is economically viable. Suitable process models are employed to predict the performance of the process, and to decide the size of the appropriate pilot scale tests that will provide convincing data for full scale plant design. Next, it is essential to move to a pilot plant testing programme utilising spiral wound membrane modules. The spiral modules must remain stable in the organic solution being processed, and so it is very important that the construction of these modules is the same for the pilot programme as will be used for the scaled up process. Membrane lifetime is a key factor that must be estimated from the data obtained in the pilot testing programme. Typically, at Evonik we have used spiral wound modules of 1.8” x 12” (0.2m2) through to 2.5” x 40” (2 m2) for the testing programmes; sometimes we also employ process size 4” x 40” modules for advanced tests. Where the application of OSN is in the pharmaceutical industry, compliance with GMP requirements and the need for demonstrating that leachable materials from the spiral element will not affect the overall process, is also essential.

The paper will describe examples of pilot testing programmes that have been carried out and which have led to process scale installations of OSN technology. OSN membrane modules are being successfully applied in cGMP manufacturing processes in the pharmaceutical industry (GSK, Johnson and Johnson), and for larger scale chemicals processing including homogenous catalyst retention at Evonik, and these applications will be described.