The talks started in earnest today, with five concurrent sessions. Most were dedicated to pharmaceutical applications, so it was “interesting” to try to cover many of them. One session, in particular, caught my attention: "New Directions and Modeling". Since it made numerous references to chemometrics/multivariate modeling, I had to check it out. In the interest of brevity, I will merely highlight a few of the better talks throughout the day.
Sharmista Chatterjee (FDA) spoke about the value of multivariate models for real-time release testing (RTRT). He pointed out (as is generally know to us NIR-types) that MV models are the heart of design space and understanding the interaction of all parameters and how they affect the “quality” of the final product.
Phil Nethercote (GSK) spoke to using QbD in analytical methods. This has been mentioned at other meetings, but needs to be emphasized. In HPLC, it has been a practice to “stress” the method (miniature variations in flow, temperature) and show that it still gives correct answers. This approach may be spread to all methods: design space for SOPs and operating methods.
Theodora Kourti (GSK) spoke about expanding the MV approach to stability monitoring. GSK’s major thesis is that close monitoring of everything from raw material parameters through operating parameters through final product qualities may be “crunched” into a formula to predict and correct stability of a product.
Bob Mattes from FOSS NIRSystems showed more applications of NIR for bioreactor monitoring. As familiar as I am with NIR, I am always pleasantly surprised at what may be measured in a complex aqueous mixture.
Process Raman was also the topic of several papers. Andrew Whitly (HORIBA) disclosed the company’s new transmission Raman (I will investigate this further).
Ian Lewis (Kaiser Optical Systems) spoke about several newer approaches, including a convertible probe (large area to small area) for various process measurements.
These were the presentations that I felt elaborated on new ideas and technologies. A number of other talks, while good, were similar to those given in the past.