Author |
Message |
Erik Skibsted
| Posted on Friday, January 30, 2004 - 7:28 am: | |
Hi Again (seems I have tons of questions these days :o)) My question is: We have a powder formulation with app. 10 w/w % analyte and we use the same formulation to manufacture 3 different tablets i.e. 110 mg, 220 mg and 330 mg. Normally when I make a NIR calibration of a tablet product we start with manufacturing 5 bathces with varying analyte concentration i.e. 75%, 90%, 100%, 110% and 125% of the target analyte concentration. But now when we have the same formulation but three different tablet sizes I am worried that I need to make 3 x 5 batches. I was thinking that I could meaby make some experimental design around the different parameters i.e. varying analyte concentration and tablet sizes, compression forces and make a model that predicted the concentration in a tablet which I then could multiply with the tablet weight to get the analyte content ?? Well, I would appriciate any comments on this ! Thank you very much Erik Skibsted University of Amsterdam and Novo Nordisk |
hlmark
| Posted on Friday, January 30, 2004 - 8:13 am: | |
Erik - are you measuring by transmission or reflection? Transmission is generally considered to be better than reflection for tablets, but if you're using reflection then it seems to me that you should be able to combine the different weight tablets into one calibration, in any case. Caveats are that you should make sure that your optics are not viewing outside the tablet (with a little margin, too), for the smallest tablets. Then all the tablets will be filling the field of view and should behave the same way, since they are all the same. Another caveat is: do the tablets behave the same way? If compression varies between the different sizes then that may affect the readings too. If you're using transmission then combining the different weights won't work. In that case you will need some sort of fractional factorial design to reduce the number of batches. Howard \o/ /_\ |
Erik
| Posted on Friday, January 30, 2004 - 8:52 am: | |
I have the ability to measure both reflectance and transmission. I like transmission because the increased information and higher robustnes against inhomogeneous tablets. Well so you suggest I make a fractional factorial design using i.e. analyte concentration, tablet size, compression force as inputs? I believe that varying compression force will have a more severe effect on reflection spectra then transmission spectra? does anyone have experience with that? Erik |
|