mobile phase - HPLC

[veronika]

I want to analyse HMF and according to literature there are two ways:
1) Mobile phase is 5% acetonitrile : 95% water. UV detection at 284 nm
2) Mobile phase is 0.08M sodium acetate buffer. UV detection at 280 nm
All analysis are made on ODS column.
As I don't have DAD I can't see UV spectra of HMF, but is it possible that because of mobile phase I have shift in maximum of HMF in UV spectra (stationary phase is the same) ?
If I want to make analysis on the same column using 10% methanol : 90% water, should I measure HMF at 280 or 284 nm ?
Thanks for the answer in advance.
Veronika

[tom jupille]

You don't need a DAD. You probably have access to a spectrophotometer (if not, contact the nearest university undergraduate lab and arrange to use theirs for a half hour). Prepare a solution of HMF (what is HMF, by the way?) in your MeOH/water solution and record the spectrum.

It may not make much difference. UV spectra are usually broad and featureless, and most UV detectors have a fairly wide bandpass, so even approximately right will work in many cases.

[DR]

Or -(no UV spec available, no university nearby)- set up your detector so that it does not rezero on wavelength changes, pump your HMF standard (in mobile phase) w/o column at a moderate flow rate until you're satisfied that the detector cell is uniformly full of the material, stop flow, set and zero the detector at ~275nm and take manual absorbance readings from the detector display as you step through wavelengths 276-290. Your highest reading will correspond to your maximum absorbance for that system.

[adam]

Tom

I have been trying for years to find a reference that does a good job discussing issues such as bandpass, slitwidth, reference wavelength, etc in HPLC with UV-detection. Can you recommend any?

Thanks Adam

[veronika]

Thanks for answer to tom jupille and DR !
There is spectrophotometer in the lab next to mine !
Veronika

P.S.
HMF is 5-hydroxymethylfurfural !

[DR]

uh-oh - someone has been formulating their amine based APIs with reducing sugars?
::quacks quietly::

[tom jupille]

I have been trying for years to find a reference that does a good job discussing issues such as bandpass, slitwidth, reference wavelength, etc in HPLC with UV-detection. Can you recommend any?

I would imagine that a good text on spectrophotometry would be a place to look, but then I haven't looked lately. When I started in grad school (early 70s), my advisor was a solid-state chemist; his working tool(s) were UV spectrophotometry and fluorescence spectroscopy. I got very intimately aquainted with the tradeoffs between bandpass and signal/noise ratio generated as a function of slit width!

That said, if anyone is interested in helping to write a good reference, I've set up a page on the ChromFAQ wiki:
http://www.lcresources.com/wiki/index.p ... VDetectors

It's unprotected (unlike the rest of ChromFAQ), so anyone can edit / update. It will be interesting to see how it works out.

[HW Mueller]

Dr, have you successfully done that determination of the max wave length via stop flow? I have immense problems here with such things (change of absorbance on stopping, drifts...). The only reasonable success was with running a relatively fast spectrum during stop flow. Finding the optimum wavelength can be done without much of a problem with my system via 3 or 4 standard determinations at different wavelength.
Unless I have "slept" through some developments, a conventional spectrophotometer has sensitivity that is at least a hundredfold lower than a UV detector...could give rise to some problems, including linearity.

[tom jupille]

Hans, while the actual absorbance in a UV detector may shift as a result of flow rate changes, I can't visualize a mechanism by which the shape of the absorbance spectrum would change significantly (but then, there are a lot of true things that I can't visualize! ).

The only purpose of the UV scan is to find out λmax approximately. You're right about it not being terribly difficult to do it by successive approximations (that sounds so much more scientific than "trial and error"). If the run times are short, it may even be as quick as a UV scan.

[HW Mueller]

Tom, that´s what I was implying: Getting a wrong max due to an artifactual absorbance (interference by scattering, refractive index...), not by a real shift in lamda max.