TOC analysis validation

GOOD VALIDATION PRACTICE (cGVP) LABORATORY, METHOD AND ANALYTICAL VALIDATION
#1

Dear all, I hope that you can help me with this problem:

Im trying to validate a TOC method for traces of productos for cleaning validation. To do this I spike some surface with a known quantity of the product, let it dry and then rinse it to determinate recuperation level.

However after the first test, while the samples got an expected value for TOC, the “Blank” sample (prepared with the same water used to rinse the surface) had about 2 or 3 times higher TOC. Thinking it was a procedural error on my part I tryed adding more “Blanks” from the same water, to discard contamination on the vial, and one more “Blank” made form low TOC water (from Shimadzu), to discard some cotamination on my part. The test got the same results: the “Blanks” made from the rinsing water had higher TOC than the “Spiked” samples, while both the samples and the Shimadzu Blank had expected values.

This seems so illogical to me since there is no reason for the blanks to be more “contaminated” than spiked samples, and even in the case that the water used for rinsing were dirty, the samples would be with higher or around the range TOC levels. Also the periodic analysis for water quality for the water I am using for rinsing in the tests always give good values (0.200-0.300 or so ppm).

So what I am doing wrong? This are the conditions of the experiment:

Water used: Inyectable grade (0.350 ppm TOC mean)
Method used: NPOC by Combustion and NDIR
Method parameters:
150 mcl inyected
1.5 % acid
1.5 min sparging
3/4 inyections
Product concentration (after rinse recovery): around 0.150 ppm (50% of it being carbon).

So while the samples get me TOCs of around 0.350 to 0.450, the blanks are around 1.000 to 1.500 ppm

Thanks for your help

#2

Hi

Does my problem is too strange or everybody thinks its my fault?

I really had tried everything to know the origin of this. I know its not me, because I use gloves and mask to prepare and manipulate samples. Its not the water, since it all the sampls (not only the blanks) would be high. It arent the vials since they are chosen at random.

It can be that the product Im using to spike the surfaces is an oftalmic solution of gramicidine, neomicine and polymixine B?

#3

[quote=AiC]Hi

Does my problem is too strange or everybody thinks its my fault?

I really had tried everything to know the origin of this. I know its not me, because I use gloves and mask to prepare and manipulate samples. Its not the water, since it all the sampls (not only the blanks) would be high. It arent the vials since they are chosen at random.

It can be that the product Im using to spike the surfaces is an oftalmic solution of gramicidine, neomicine and polymixine B?[/quote]

Dear AiC,

without being present at the sampling of the surface, it is nearly impossible to pinpoint potential mistakes. From the data provided, it does not seem to be an issue from the blanks but rather from the sample itself, because if the water used has a bias TOC of 350 ppm, the encountered blank values of 1000 ppm seem high but at least consistent, but TOC of a spiked surface with less TOC (150 ppm) than the water used for sampling is clearly illogical. I would try to clean all the surfaces with sodium percarbonate solution prior to the recovery test (rinse well, last rinse with water for TOC), to eliminate organic residue, and sample with water for TOC analysis (which has typically less than 0,1 ppm TOC) instead of WFI, to eliminate TOC variability in the sampling water as a source of error.

Best regards

Alfred