Odd looking Spectrum...pure?!

[LC_addict]

Hello again all,

Recently i have been performing some analysis. My method is RP and runs on a Gemini NX 5u using a gradient elution of KH2PO4 buffer/MeCN. Diluent is MeCN. I have injected a standard (pure reference solution) and had a look at the UV-Spec/purity of the peak. It seems as thought (even though i know my peak is 'pure') i still seen a difference in spectrum 50% +- of the peak height,and even more so 10% +- height. I have reduced injection volume thinking it may be a UV overload issue and it does improve the Spectra but still isnt perfect...
WL is set at 220nm and i feel it may have a negative effect on analysis...

Any ideas? Cheers again..... see image...

Cheers

[tom jupille]

You might try doing an isocratic run and see if it shows the same effect. If it does, thst suggests that your your standard is not as pure as you hoped *or* that there are two forms of the compound.

If the problem does not show up in an isocratic run, then the possibility is that your compound's UV spectrum shifts with solvent composition. Probably the best-known example of this is TFA (see this article: http://tinyurl.com/kaquphs ).

[mattmullaney]

Hi, LC Addict,

I agree with Mr. Jupille--is there a possibility that there could be rotational conformers such as that which occur with Schiff bases or amide functionalities within the substance you're studying? Also, strangely, UV spectra have at times been observed to change as a function of analyte concentration, in addition to analyte-solvent composition. I've had to acquire spectra for all analytes within a series of chromatograms as a function of analyte concentration on several occasions to provide the best opportunity for accurate spectral matching.

In my past, I have also observed that generally that as long as the absorbance maximum (maxima...) of the eluted peak(s) are less than 0.8 AU, that spectral matching seems to work a bit more smoothly.

Best Wishes!

[Hollow]

couldn't there (also) be some normalization error included?

the SW does the normalization on something that is not compound related but from the eluent.
Your spectra seems to be going down to about 190nm? Which wavelenght is the 100%?

If possible, try to set the normalization point to about 199nm (or 271nm) and see how the overlay look like. Maybe also try to narrow the spectral range e.g. only 195-300nm.
The other mentioned things could still apply.

Btw: Is there any strong reason to record your chromatogramm at 220nm, which is an very steep flank of the spectra. Why not use 199/200nm or if sensitivity is not that important 271nm