linearity of UV detector issue

[jiang295]

hi I just wonder what determines the linearity range of UV detector. is the conc. of analyte or signal intensity? This question pop up when I change analytical cell to prep flow cell.
thanks if you can clarify for that.

[Uwe Neue]

The signal of the UV detector is A = E*d*c, where E is the extinction coefficient of your analyte, d is the length of the path over which the signal is measured, and c is the analye concentration. Prep UV cells have a shorter path length, and the signal becomes lower when you change to a prep cell.

In prep chromatography, you should stay away from the absorption maxima to keep the signals within the (linear) range of the detector.

[jiang295]

thanks Uwe Neue, so are you saying that linear range is determined by conc. of analyte rather than types of UV detector?
th

The signal of the UV detector is A = E*d*c, where E is the extinction coefficient of your analyte, d is the length of the path over which the signal is measured, and c is the analye concentration. Prep UV cells have a shorter path length, and the signal becomes lower when you change to a prep cell.

In prep chromatography, you should stay away from the absorption maxima to keep the signals within the (linear) range of the detector.

[bisnettrj2]

Someone correct me if I'm wrong here, but I think the answer is 'It depends'. The extinction coefficient refers to how strongly your analyte absorbs UV at the wavelength you're using - hence Uwe's advice to not use the absorption maxima of your analyte in preparative chromatography. Therefore, the two factors that become important in Uwe's equation are the extinction coefficient and the concentration, when the path length is fixed.

For example, you could have a low to moderate concentration of an analyte saturating your detector when you are at the absorption maxima of that analyte, or you could have a high concentration of your analyte saturating your detector when you're at a wavelength that is not the absorbance maxima of your analyte.

[Uwe Neue]

Not sure why you care about the "linear range" of a detector when you are doing preparative chromatography. When people do prep and exchange an analytical flow cell to a prep flow cell, they want to reduce the signal and are not worried about sensitivity. "Linear range" takes both the upper limit and the noise into account.

For an in-depth understanding of the subject of UV/Vis detectors, I recommend A. C. Gilby, Encyclopedia of Separation Science, Chromatography:Liquid/Detectors:Ultraviolet and Visible Detection.

[jiang295]

because i want to do gradient linearity test by putting acetone in methanol.
instead of putting 0.1% acetone as usually recommend, i use 1% acetone based on the ratio of path length of analytical and prep cell.
but I did not get a straight line, so I wonder 1% acetone will go beyond linearity range of UV detector though intensity is well below 1Au.

Not sure why you care about the "linear range" of a detector when you are doing preparative chromatography. When people do prep and exchange an analytical flow cell to a prep flow cell, they want to reduce the signal and are not worried about sensitivity. "Linear range" takes both the upper limit and the noise into account.

For an in-depth understanding of the subject of UV/Vis detectors, I recommend A. C. Gilby, Encyclopedia of Separation Science, Chromatography:Liquid/Detectors:Ultraviolet and Visible Detection.

[jiang295]

so you think as long as you did not saturate detector, it will keep in the linear range of UV detector?
I do not whether I understand you right or not?

Someone correct me if I'm wrong here, but I think the answer is 'It depends'. The extinction coefficient refers to how strongly your analyte absorbs UV at the wavelength you're using - hence Uwe's advice to not use the absorption maxima of your analyte in preparative chromatography. Therefore, the two factors that become important in Uwe's equation are the extinction coefficient and the concentration, when the path length is fixed.

For example, you could have a low to moderate concentration of an analyte saturating your detector when you are at the absorption maxima of that analyte, or you could have a high concentration of your analyte saturating your detector when you're at a wavelength that is not the absorbance maxima of your analyte.

[bisnettrj2]

Well, like Uwe said, you're doing preparative chromatography, but still talking about the 'linear range' of the detector - why? The linear range is from some defined lower-limit of S/N ratio to the point where an increase in concentration will cause the detector response to stop being linear. You can 'extend' the linear range by working at a wavelength that is off the absorbance maxima of the peak - but you sacrifice lower-end sensitivity. You also sacrifice sensitivity with the shorter path-length of the preparative cell.

[danko]

I don’t think Jiang295 cares about the linearity in connection with the purification target’s amount.
I think the goal in this context is to examine the system’s ability to perform a gradient.
So, if I were correct in understanding the case properly, I’d recommend a repetition of the experiment using 0.5% acetone and comparing the outcome to the result obtained with 1 % acetone. If the traces are parallel then you don’t have a problem with the dynamic range of the detector. If not parallel then you’ll know what’s the problem.
As for the reason why the dynamic range could be reduced even though the abs. max is below 1 AU, there could be other explanations than analyte saturation. Typical examples are dirty flow cell window, low lamp energy etc.

Best Regards

[jiang295]

Hi danko,
yes, I want to use it for gradient performance.
so are you saying below 1Au will be uv detector dynamic range (linear range) regardless how long flow path I use if nothing wrong with uv detector.

I don’t think Jiang295 cares about the linearity in connection with the purification target’s amount.
I think the goal in this context is to examine the system’s ability to perform a gradient.
So, if I were correct in understanding the case properly, I’d recommend a repetition of the experiment using 0.5% acetone and comparing the outcome to the result obtained with 1 % acetone. If the traces are parallel then you don’t have a problem with the dynamic range of the detector. If not parallel then you’ll know what’s the problem.
As for the reason why the dynamic range could be reduced even though the abs. max is below 1 AU, there could be other explanations than analyte saturation. Typical examples are dirty flow cell window, low lamp energy etc.

Best Regards

[danko]

Exactly!
But as you mention: If nothing wrong with the detector!
A special case (exemption) would be dirty (energy absorbing) mobile phase. In your case it could be polluted water.


Best Regards

[JGK]

If you look at the absorbance v transmittance table here http://en.wikipedia.org/wiki/Absorbance ... smitttance you can see that at 1 AU only 10% of the source light is reaching the detector, at 2 AU it's 1% and at 3AU 0.1%.

Consequently, small deviations at high AU levels could easily throw your linearity out especcially on unweighted regressions where the highest concentration dominates the calculation.

I've always tried to avoid any analysis where the maximum concentration exceeds a 1 AU absorbance for this reason.

[Alex Buske]

You ran a linear gradient with water and 1% acetone - right?
And the detector output didn't schow a straight line- right?
The max absorbance was well below 1 AU - right?
And now you don't know wether your pumps underperform or detector isn't for some reason linear in that region - right?

So why dont you simply check linearity of the detector? Prepare acetone solutions 0.5, 1, 1.5, 2, 2.5, 3, 4, 5 % and injet the directly with a syring in the detector^, read the output absorbtion value and put it in a spreadsheet.
Alternatively you could inject the solutions with the autosampler.

Alex

[jiang295]

that is also good idea. much quicker.

You ran a linear gradient with water and 1% acetone - right?
And the detector output didn't schow a straight line- right?
The max absorbance was well below 1 AU - right?
And now you don't know wether your pumps underperform or detector isn't for some reason linear in that region - right?

So why dont you simply check linearity of the detector? Prepare acetone solutions 0.5, 1, 1.5, 2, 2.5, 3, 4, 5 % and injet the directly with a syring in the detector^, read the output absorbtion value and put it in a spreadsheet.
Alternatively you could inject the solutions with the autosampler.

Alex