Chromatography Forum

Dear Chromatographer

I would like to know why ceffine only to be used to calibrate HPLC.

If you mean the use of caffeine to check UV wavelength accuracy, lots of chromatographers prefer to use it over holmium oxide filters (or use it in addition to) because caffeine has both a local min & max in the UV range that's at least close to the wavelengths commonly used on a daily basis. HOs wavelengths tend to be much higher.

Caffeine is cheap, stable, safe and can be shipped without restrictions. It is not the best choice for wavelength calibration, but for injector precision and detector linearity it works well.

We usually use anthracene for wavelength accuracy test, it has very sharp spectrum.
Mark, what is the best choice for wavelength calibration?

I don't really know why this was chosen, but we use pyrene in methanol for wavelength verification. You can also use solutions of rare earth ions like erbium perchlorate. Most modern UV detectors can check themselves. Some use holmium glass, others use prominent emission lines in the light source. With older models you need to use an external standard.

Thank you all for u r reply.

If detector don't have scanning option then How to check wavelength accuracy in variable wavelength detector & with what ?

Thanks in advanced

You may found many reference about this. This is one example:
G Hall and JW Dolan, LCGC North America, 20(9), 842-848 (2002).
It uses anthracene solution and the test is performed by programming 5 separate run with wavelengths of 249, 250, 251, 252, and 253 nm. If a secondary wavelength is required, you may add three additional injections of the test solution at 348, 350, and 352 nm. Measure the area of anthracene for each run. The wavelenght for maximum response should be 251 nm ± 2 nm. Anthracene has a secondary maximum at 350 nm.

A programmable variable wavelength detector is capable of scanning over a narrow range, just not quickly enough to be useful for real-time spectra collection. The firmware uses a special mode to do the scanning during the wavelength check.

At the risk of sounding stupid , if you don't have a scanning detector you could make a "sufficiently concentrated" solution and manually fill the detector cell with it. You can then manually change the detector wavelength and record the absorbance. While the detector might have an autozero function whenever the wavelength is changed, there might be a way to switch this function off while you are performing the check. Somewhat tedious and time consuming, but perhaps quicker than perfoming multiple runs. Not sure if there is some problem with this approach.

Can I used mixture of known compunds who having maxima at different wavelength (for example Comp. A have maxima at 254nm and Comp B have maxima at 275nm). This mixture of compound can inject at different wavelegth i.e 254nm & 275nm ? What I will achiceve is at 254nm run comp.A will show higher response and at 275nm run Comp B will show higher response. Is it OK ?

Thanks

At the risk of sounding stupid , if you don't have a scanning detector you could make a "sufficiently concentrated" solution and manually fill the detector cell with it. You can then manually change the detector wavelength and record the absorbance. While the detector might have an autozero function whenever the wavelength is changed, there might be a way to switch this function off while you are performing the check. Somewhat tedious and time consuming, but perhaps quicker than perfoming multiple runs. Not sure if there is some problem with this approach.

That's how it's done w/ older Shimadzus. You prepare a potassium permangenate solution series, shut off the detector's "autozero on wavelength change" function and step through different wavelengths at each solution concentration. From these numbers, molar absorbtivities are calculated...

DR, why use different concentrations?

Russ,
The problem with stationary sample (no flow) measurements is scattering, or whatever, due to air in the sample. Air can also produce a hell of a drift at stop flow.

How do you introduce the sample to the flow cell? I have had some reasonable success in pushing most of the content of a 5mL syringe, filled with solutions, through as fast as possible.
This might have been mentioned before: When I check UV absorbance of HPLC water, etc., I evakuate 100mL of it in an empty prep column and apply N2 pressure to force the liquids through the detector and a 34 bar resistor. Still there is a small drift, but if you did this with a wavelength standard this shouldn´t hurt wavelength checks too much.

HWM: I also use a syringe filled with the test solution to fill the flow cell. I disconnect the back pressure regulator, connect the syringe to the flow cell through a ZDV fitting and after a "sufficient" amount (5 to 10 mL) of the solution is pushed through, cap the detector outlet fitting. I generally allow a couple of minutes for any equilibration before measurements. We check going up in wavelength then recheck coming back down and do not notice any major drift problems. Our test solution is made in methanol, which may not have as much dissolved air in it as an aqueous one. Perhaps I have just been lucky. The saying is that God looks after little kids, drunks, and fools. Since I am well past being a little kid and rarely drink, I guess that lets me know what category I fall into.

DR, why use different concentrations?

It's not my procedure, but I would guess that the multiple concentrations serves to demonstrate linearity (across different wavelengths). If each concentration were prepared from a different weighing of the solid, you would also have 4 different molar absorbtivity values (per wavelength tested) to average. In theory, this minimizes the chance of systematic error skewing the values.

P.S. If we do the Shimadzu test, we prepare 500mL of each solution and use the pumps (ZDV in place of column) to introduce the solution to the flow cell. We pump lots of it before taking a reading and we only take readings if they're constant for 60 seconds (if we can't get that after pumping 30-40mL, there's another problem).

Russ and DR.,
good to hear that it takes some effort, I thought my equipment is particularily strange in this regard.
Also, MeOH should dissolve air a bit better than water, some tests here show that it has less tendency to degas, so we sometimes use it to flush out air in columns, etc.