An amusing story (only in hindsight!) from the annals of Organic Process Research and Development [1]:
Scale-up is a minefield! Impurities from somewhere contaminated the product. In GMP philosophy, impurities on an industrial scale must be lower than at the laboratory level. The reason is that raw materials, solvents, and reagent qualities are checked, and reactor vessels, pipelines and centrifuges are cleaned with great care prior to the GMP production. The contamination from the facilities, consequently, might be slight. However, we found unexpected impurities in the case of scale-up. They could not be completely anticipated in the laboratory....The authors describe noting turbidity in typically clear product solutions and unknown peaks in their HPLCs. An investigation ensued:
From these results (the wet cake showed turbidity), the product did not become contaminated with turbidity compounds before the activated carbon treatment or after drying. Therefore, all of the equipment, crystallizer, feed pipe, pressure filter, pumps, clothes, gloves, and storage bags were checked. That is, those components were rinsed with MeOH, and the MeOH was analyzed by HPLC. After that, no equipment, except for the natural rubber gloves, contained the peak B material, and the polyethylene bags used in this process contained peak C and peak D material! [snip]They found the impurities to be BHT, dioctylphthalate and other polymer stabilizers:
...During runs 2 and 3 in Table 1, the turbidity was caused by the 200-L polyethylene bag. It was caused by natural rubber gloves during run 4 (Tables 1 and 3). To prevent the turbidity, the process has been changed as follows: the material of the gloves was returned to polyethylene. In this case, the contact time with the wet cake is very short; therefore, there is no worry about turbidity. The containers for the wet cake storage were changed from polyethylene bags to stainless steel buckets. After these modifications, no turbidity was observed on the plant scale (Table 5).
Murphy is a jerk, let me tell ya.
1. Sano, T.; Senzaki, H.; Sugaya, T.; Kasai, M. "Contamination of Dipeptide by Polymer Stabilizers Leached from Gloves and Packaging during Scale-up." Organic Process Research & Development 2000, 4, 349−352.
Murphy Law or a Japanese Jerk of a boss? There have been cases of long-suffering employees dropping worse stuff than pair of gloves into the reactor....
ReplyDeleteAs I often have felt stalked by Murphy (and would use stronger term than jerk) in lab work and process execution would definitely agree with sentiment in OPRD article/comment above. He appears to spend a disproportionate amount of time is pilot plant/manufacturing facilities. Having been haunted with leaky hoses/pipes, rusty drums/reactors, frayed filter clothes, stuck valves, deteriorated or incompatible gaskets and other out of left field occurrences in well run/maintained US and EU Pharma facilities it makes one wonder greatly about potential problems when outsourcing to other parts of the world (do Chindia have Murphy counterparts?).
ReplyDeleteI dispute milkshake that misbehavior is likely a wide-spread issue. Although I am sure there are incidents of sabotage due to disgruntled operators I would suggest such would be much more difficult to do and get away with in manufacturing setting. Majority of "less educated" operators seem to care more about the products they make knowing aimed towards patients than typical PhDs perhaps because the latter want to get back to the lab. Do not recall if ChemJobber ever got his proof/awarded the Duck but I personally suspect higher ratio of sabotage in lab than in plants.
CMCguy