We measured a series of spectra of individual compounds and got a good separability with PCA Analysis. The questions risig are as follows: What happens if we ad mictures of the individual compounds ? Do we expect to get a cluster in the middle of / between the pure compounds then ?
What can we expect if we mix one compound with a series of stuff that also is IR absorbing ? We first thought that the resulting PCA clusters for the spectra might then be totally random. However we surely still have the information of the well characterized pures substances. Should we somehow fix the components of the PCA ? And is there some trick how to train the system best to achieve the identification of single compounds in mixtures after training the system with pure substances ?
IR is usually not used to quantitate a molecule's concentration because of the weak response (thus UV). In theory for quantitation the main peak must be unique, strong, sharp, and well separated from other peaks. Therefore, in a mixture you would see 3 peaks. However, there will be other peaks as well due to as an example a Methyl rotation....
Rather than have a useless discussion on RG why not just do it and perform an IR scan of a mixture!
Hi. Thanks for the info. We will surely do so. However it will take some time and I just wanted to explore a bit what we might expect.
That is a really perfect answer. I try that. Thank you very much for your help.
Best wishes !
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training