How to choose the most representive wavelength?

[IUpuppy]

Hi,

I have been running HPLC for almost 2 years but still I don't know the rule of choosing the right wavelength.

Most of the time we use 254 to begin with and sometimes we modify the wavelength either by quantitative analysis or comparison with NMR etcs..

I have been monitoring a reaction recently and found that different wavelength results different level of purity for the compound we are interested. For example, the PDA area percentage at 254nm of the peak we are interested is 67%, and at 296nm is 93%.

The only way I could think of choosing a more representative wavelength is to do a quantitative analysis, or maybe using more than 1 detector?

I wonder what people usually do to choose the right wavelength. any suggestions will be appreciated!

Thank you all

[dhobbs]

Peak area percentage isn't a reliable measurement of purity by weight regardless of what wavelength you choose. Many impurities methods use low wavelengths such as 205 nm because lots of impurities won't be detected at higher wavelengths. If you want a true percent purity value, though, the best approach is to use the lambda max of your compound as the analysis wavelength and determine quantity of compound in your sample compared to a standard of known purity.

[KM-USA]

We tend to use higher wavelengths than 254nm whenever possible, typically better specificity. 254nm is mainly a holdover from decades ago when UV detectors had that wavelength only.

For example, if my lambda max was 296nm, I'd typically use that, unless there were matrix interferences there. If there were, I'd choose a wavelength of less matrix interference. Oftentimes we just run a UV scan on the spectrophotometer for a new analyte and start there.

[Hollow]

As already said before, UV purity tests at a single wavelength aren't too reliable.
some compounds just don't absorb at the selected wavelength or don't do at all (e.g. salts, sugars, olefines etc).
Even if they do, the specifc coefficient of extinction can vary a lot, so a big peak can be only traces or vice versa.
Keeping this mind, if you want to do some purity tests by UV it would be best to use as PDA detector and looking on a PDA Max plot over a wide wavelength range.

[IUpuppy]

Thanks for giving me some insight into choosing wavelength everyone.

Is it okay to use a fixed wavelength for reaction monitoring? I thought that the PDA percentage of the product becomes constant or even decreasing(maybe due to decomposition) is a sign of reaction completion.

Thanks again for everyone's help!

Hyuna

[dhobbs]

You can certainly use a fixed wavelength for monitoring a reaction, and yes, the product should generally become constant as the reaction goes to completion, but I don't see what you would gain by monitoring area percentage rather than just the area of the product peak.

[IUpuppy]

Hi, thank you Dhobbs for your reply.

Actually there is no particular reason that I choose to read percentage rather than the absolute area- I was taught to monitor reaction by reading PDA area percentage when I joined in the lab. I just didn't know which one is more accurate in determining the progress of the reaction, given that UV detection is that quantitative.

Thanks so much for everyone's help!