• Replies 3 replies
  • Subscribers 8 subscribers
  • Views 414 views
  • Users 0 members are here
Related Content

Monitoring of a Chemical Reaction through PAT

Rodolfo Romanach

The monitoring of a chemical reaction through PAT was a wonderful learning exercise. The tranesterification reaction to form biodiesel was monitored through near infrared and Raman spectroscopy. It turned out that what appeared to be a simple chemical reaction (on paper), was a reaction of multiples stages. Principal Component Analysis was performed for different time periods of the reaction to better understand it. Even though this was not a reaction for a pharmaceutical process, it could provide helpful insights for future studies. The doi link for the study is: doi.org/.../12-0672

The full citation is:  Fontalvo-Gomez, M., et al., In-Line Near-Infrared (NIR) and Raman Spectroscopy Coupled with Principal Component Analysis (PCA) for In Situ Evaluation of the Transesterification Reaction. Applied Spectroscopy, 2013. 67(10): p. 1142-1149.

 

  • Thanks for sharing, Rodolfo. In the API synthesis, the “Monitoring of a Chemical Reaction through PAT”, selection of the proper PAT, and its implementation come into three aspects. Here is a short list to kick off the discussion:

    - Molecule/reaction:

    **What proper PAT could be used for the molecule/reaction of interest?

    **Is the molecule (product or starting material) Raman active?

    **Is there water present or generated during the reaction that causes problems with FTIR?

    **Is the molecule conductive, and can a conductivity meter be used?

    **Do we have response factors (calibration) for all molecules and picks in the spectrum?

     

    - Equipment:

    **Is it CSTR or PFR equipment? E.g. For the CSTR, a React-IR and for PFR, a Flow-IR can be used.

    **Where/how do I place the probe? Can it be inside the reactor (inline) or is a flow cell (online) required around the CSTR.

    **For on-line PAT, how should the sampling be? Would there be any clogging or homogeneity issues, or any quenching or dilution required?

    **For multiphase systems (S-L, L-L), which phase is being measured?

    **How would the PAT maintenance, calibration, fouling, cleaning, .?

     

    - Application:

    **Is the intended application for just monitoring, or is the PAT connected to a process control scheme?

    **If data is used for active process control, then the process control is its own can of worm! Such as the data frequency required, delay/lag considerations, …

    **Is there an orthogonal test required? If so, how would the data reconciliation be?

    **How do I develop models for the PAT (PCA, PLS, Regression, …)?

    **If the production is GMP, the PAT can be part of different CMC modules, such as analytical methods, process control strategy, data integrity, QMS, data historian, …

     

    The list goes on and on, and cases can be fairly simple, or very complicated. I encourage interested parties to consult with SMEs (all aspects).

     

    I suggest continuing the discussion on each of these technical topics individually.

    Thanks again.

  • As Rodolfo points out there is a lot pharmaceuticals can learn from our colleagues in the heavier chemical industries.  Online sensing, even using instruments we would consider offline such as gas chromatography, have been used for decades now.  

    Nima, thanks for your comments as well.  With your knowledge, and a few collaborators, a textbook is possible.  

    We should also keep in mind the value of at-line analysis during development and keeping online techniques where feasible only as necessary to control the process.  Some programs have benefitted from measuring everything and had commercial success from removing all but the essential for commercial manufacture. 

     

  • I would say that the use of PAT for process understanding in Pharmaceutical R&D is fairly generalized and the benefits are well understood, especially when we are talking about flow chemistry - FTIR, NMR, HPLC are probably the most general/informative, but others can also be useful.

    When the process moves to a production setting, then the discussion on the use of PAT really needs to address all the considerations raised by Nima, and probably others. Looking at cGMP manufacturing in drug substance, the inclusion of PAT is far from being a simple yes / no decision.