Knowledge Center

Article / Jan 01, 2011

Organic solvent nanofiltration: A platform for removal of genotoxins from active pharmaceutical ingredients

  • Gyoergy Szekely
  • William Heggie
  • Boerje Sellergren
  • Frederico Ferreira
Journal of Membrane Science (2011), 381(1-2), 21-33

This paper discusses the use of org. solvent nanofiltration (OSN) as a general platform for the removal of genotoxic impurities (GTI) used as reactants or formed during the synthesis of active pharmaceutical ingredients (API) or their intermediates.  The analyses carried out takes into account simple and easy available inputs, such as the mol. size of the API and GTIs.  Results are discussed in terms of API losses and GTI removal, as well as final contamination levels.  Boundaries are established for a quick assessment of the API/GTI sepn. in three categories (A) OSN provides a straight forward route for removal of GTIs from API post reactions, (B) OSN can be applied for a useful removal of GTI from API post reactions, and (C) OSN alone is not adequate removal of GTI form API and other technologies or new OSN membranes are required. For the intermediate category, case B, higher diavolumes may be required to reach low GTI toxicity thresholds leading to API losses with significant economic impact.  Therefore, for this case, feasibility of OSN purifn. is also discussed on the basis of daily dose of an API and the corresponding GTI threshold limit.  Nine APIs and eleven GTIs were selected as model compds. to be representative of the diversity of chem. reactions and compd. structures produced by the Pharmaceutical Industry. The soly. of APIs in several solvents was evaluated and Me Et ketone (MEK) and THF were selected as model solvents for this study.  These two polar solvents are usually used in industry due to their ability to dissolve APIs at high loads.  However the very same dissoln. power makes it difficult to develop polymeric OSN membranes stable to such solvents.  Two available OSN membranes were selected on which the literature data available is scarce in order to complement the known data.  Membrane rejections at two applied pressures were exptl. estd. and diafiltration was used to illustrate API/GTI potential sepn.  The methodol. proposed here can be applied to other membranes, solvents, APIs and GTIs.