Wednesday, August 10, 2011

Process Wednesday: Are you being fooled by your IPCs?

From our mentor-by-literature Neal Anderson, a comment on in-process controls (IPCs):
A simple hypothetical example will serve to demonstrated the value of IPC. Suppose that a reactor is charged with a hydrogenation catalyst, an unsaturated compound and solvent. Air is then replaced with H2, using a suitable evaucation protocol. Then the reactor is charged with H2 to the desired pressure, and the reactor is sealed (the pressure is "locked in"). When the starting material is reduced and H2 is consumed, the pressure is expected to decrease. For any given run, however, an anticipated drop in pressure does not guarantee that the desired reaction is complete. A leak may have occurred, enabling a loss of H2. The wise chemist notes the drop in reaction pressure, then confirms by a second assay that the reduction is complete. [CJ's emphasis] Since processing time and materials are very valuable, IPC is used to ensure that the desired processing endpoint has been reached before proceeding to the next step. Choosing the appropriate IPC and collecting dependable data are challenging, often unappreciated aspects of process development. 
There's something quite tempting about using one data point to tell you what you think you already know. It's probably a good idea to use a second, different method to confirm the first one, especially when you have many man-hours and costly material riding on the consequences. 

No comments:

Post a Comment