Wednesday, October 5, 2011

Process Wednesday: getting accurate In-Process Controls

Getting reliable in-process controls can be quite difficult; checking on a reaction solution that is heterogeneous can be tricky. Neal Anderson talks about this in Practical Process Research and Development: 
A heterogeneous reaction is often more difficult to sample than a reaction that is a solution. A solution can be sampled from any point in a reactor, and all portions of the sample are expected to give identical assays. With heterogeneous reactions, care must be taken to ensure that the sample withdrawn can assayed to truly indicate the contents of the reactor.  
For a heterogeneous reaction, sampling is easier if the phases are not viscous. If the reaction is dispersion of two liquids, samples can be taken from the top and bottom of the reactor, or increasing the agitation may provide a suspension of both phases that can be readily withdrawn. For a reaction that forms a suspension of a liquid and a solid, samples can be withdrawn as well-dispersed suspensions, or either phase may be withdrawn for assay.  
...Reactions at very low temperatures also pose sampling difficulties: by warming a reaction aliquot to room temperature, the reaction may progress significantly beyond the extent of the reaction in the reactor, providing erroneous analyses. Under these conditions it may be necessary to study the temperature and time constraints for sample preparation. 
I suspect that many chemists in graduate school discover the pitfalls of warming IPCs when they're pulling aliquots out of their -78°C baths to perform TLCs. 


  1. If I had a reaction I thought was warming in my TLC spotter and screwing up my TLC analysis, I would spot 4 or 5 'reaction' lanes. I would pull the spotter out of the reaction, then, as quick as I could I'd spot the 1st rxn lane, then immediately spot the 2nd, then the 3rd, etc. If the reaction was progressing while it warmed, I could usually see this time delay as the SM tapered off in my 4 or 5 rxn lanes.

  2. Amazing that so very few grad students or their advisors, at least at the institutions I worked at, gave this sort of thing much thought. I would have loved to have been a process chemist at some point so all my "cooky ideas" would have been common knowledge among my peers.