Dear Community,
I would like to develop and validate a qualitative NMR method for the analysis of a specific category of chemicals, and my question is the following: What are the criteria that I must assess?
I found a publication that states that the Limit of detection (LOD), the specificity and selectivity (NMR is inherently specific though), and Robustness must be investigated for a qualitative NMR method. But then what would be the experimental protocol to assess those criteria? For example: is there a need to perform the same experiment multiple times (on the same sample) and calculate the standard deviation for the estimation of the LOD, or just one experiment would be enough?
I would be more than grateful if someone experimented with NMR method development could provide more details on the subject.
Thank you.
The common practice is to examine at least 3 samples and measure/analyse each sample at least 3 times. This is the minimum for making a statistical evaluation about the variability. You can read more about the current status of qNMR monographs in the following publications:
https://qnmr.usp.org/t/monographs-using-nmr-and-or-qnmr-in-usp-documentary-standards/75
https://www.pmda.go.jp/files/000241922.pdf
However, when it comes to automation, they are all missing the criteria for evaluating the reliability of the analysis even though NMR provides an ingenious method to measure the reliability of the automatic interpretation. This can be done directly by comparing the experimental data to a theoretically spectrum calculated based on this automatic interpretation. The difference between the two and some other consistency measures allow for establishing diagnostic scores, which must be within predefined ranges.
Dear Matthias Niemitz
Thank you for your answer. My concern is the following: is it necessary to perform a statistical study, when the only retained criteria are: the Limit of detection (LOD), Selectivity & specificity, and Robustness, which can be assessed through just one NMR analysis according to the following document : Technical Report Guide to NMR Method Development and Validation – Part I: Ide...
PS: I am dealing with qualitative NMR analysis
Hi,
What is the goal of your NMR measurement?
Are you acquiring a spectrum of certain compounds and that has to resemble that of a reference compound? This could imply that you can convert a qualitative assessment to a relative quantitative value (score).
Or are you looking at the presence/absence of a certain NMR signal at a given chemical shift? Is that an on/off situation (black or white)
Depending on the attribute you're trying to determine, the approach to validate a method can differ, but it seems you mainly have to focus on specificity, which means analyzing at least some positive and negative control samples.
Hope this helps!
Hi Eef Dirksen
It is more the second scenario, I will be looking for the presence of specific analytes (which are in the number of millions) by combining, 1D and 2D, 1H and 31P based NMR experiments in an attempt to reveal their signals.
samples can be anything from aqueous to organic solvents with some typical contaminants (eg: diesel,PEGs, other non-pertinent chemicals ...etc).
Thank you.
- Badges
- Science topic
- Similar topics
- Filing