Chromatography Forum

Hi All

We have a HPLC Method which we usually run on a 2695 Waters HPLC & 2487 Detector. At the hold up time in the chromatogarm we get a solvent peak which is usually about 0.004AU.

We recently run this same sample/Method on an Agilient 1100 DAD, we get the solvent peak at the same retention time but it is 0.40AU.

Basically the solvent peak (which is Chloroform) is massive in the Agilient Chromatogram compared to the chromatogram obtained with the Waters HPLC.

Any ideas what can cause this?

Thanks in advance

How do the analyte peaks compare?

Are you using a wider band width on the Agilent versus the Waters? Is the peak narrower?

Several things can affect how high the peak is between two different detectors. Sometimes even the scale that the different companies use in their software.

Additional possibilities:
- differences in extra-column band broadening (would affect the peak width, per James Ball suggestion)
- flow cell alignment; if the flow cell is out of alignment, there can be more off-axis light and thus more sensitivity to RI shifts.

How do the analyte peaks compare?

The Analyte response on the Agilent is about half that observed on the Waters for both height and Area

Are you using a wider band width on the Agilent versus the Waters? Is the peak narrower?

On the Agilent the detector settings are as below.

Peak Width: greater than 0.1min(2s)
Margin for negative absorbance mAU: 100
Slit: 1nm

Analog A
Attenuation mAU: 1000
Zero offset: 5%
Voltage Range: 1.0v full scale

Channels
Bandwidth: 100nm

The Waters Detector setting are as below
Data mode: Absorbance
Filter type: Hammering
Ratio minimum(Au) 0.1
Time Constant(sec): 1.0

Analog A
AUFs: 2.000
Absorbance off set: (AU) 0.00
Voltage offset (mV): 0
Polarity: +

Should the Voltage range on the Agilent match the AUFs on the Waters?
Waters Bandwith in the specifications just states 5nm, were the Agilent Bandwidth is set to 100nm
Can the slit be set on the Waters as it is on the Agilent?

Hope this makes sense.

The obvious likely culprit is the band-width.
On the Agilent system if your selected wavelength is 280nm, for example, then you are actually sensitive to changes in absorbance anywhere from 230 to 330, and at the low-wavelength end there is probably substantial absorbance by injection solvent. On the Waters system it looks like you are only looking at 278-283 approx (again, for example; I don't know what wavelength you actually set), where there may be little or no absorbance by injection solvent.

You need to make sure that you've selected not only the same wavelength, but also the same bandwidth on both instruments. Also check that you haven't got a reference wavelength specified on one system but not the other.

While I agree with the above, I also suspect that, given the 1AU Agilent setting vs. 2 for the Waters, you should actually get more peak area from the Agilent (all other things being equal - which is seldom the case).

Since you are not, I suspect that Tom is also correct in that you have an alignment problem with the Agilent (lamp, flow cell are the most likely culprits).

While I agree with the above, I also suspect that, given the 1AU Agilent setting vs. 2 for the Waters, you should actually get more peak area from the Agilent (all other things being equal - which is seldom the case).

Thanks DR why is this the case what effected does the AUFs have on the Area?

Kind Regards

Let's assume that both detectors have a 0 to 1V analog output range.

Setting a detector to 1AUFS would make it more sensitive than setting it to 2AUFS. All things being equal, the 1 setting should net you twice the area of the 2 setting (assuming you are not going off scale).

Let's assume that both detectors have a 0 to 1V analog output range.

Setting a detector to 1AUFS would make it more sensitive than setting it to 2AUFS. All things being equal, the 1 setting should net you twice the area of the 2 setting (assuming you are not going off scale).

On most Agilent systems the AUFS only affects output to an analog device such as a chart recorder, the data system used a digital output in mAU and ignores the Analog signal, so it should not matter if using a data system. Not sure how the Waters instrument works.

The obvious likely culprit is the band-width.
On the Agilent system if your selected wavelength is 280nm, for example, then you are actually sensitive to changes in absorbance anywhere from 230 to 330, and at the low-wavelength end there is probably substantial absorbance by injection solvent. On the Waters system it looks like you are only looking at 278-283 approx (again, for example; I don't know what wavelength you actually set), where there may be little or no absorbance by injection solvent.

You need to make sure that you've selected not only the same wavelength, but also the same bandwidth on both instruments. Also check that you haven't got a reference wavelength specified on one system but not the other.

I agree here, the bandwidth can make a huge difference on the Agilent DAD detector depending on what your target wavelength is.

On the Agilent DAD set the bandwidth for DADa to 4 nm and switch off the reference! Then run again. If the reference wavelength is e.g. 330 nm and the bandwidth is 100 nm is potentially might overlap your analysis wavelength and suppress the signal.
Then get back to us.

Thanks for your help with this? The change in bandwidth has increased the Area.

Next step is now to investigate the max allowable bandwidth on your analysis wavelength, increase to e.g. 8 nm and see is that increased the area (as expected) without compromising the s/n ratio.