With HP resolution was at least 0.9 but with Waters resolution is less that 0.9; consistantly obtained resolution of 0.5 (same solvents ACN:Water and the same gradient method). Any ideas?

Any other parameters eg tailing factor, peak height, ...

I've about 20 peaks,other parameters are OK and resolution is OK for all peaks except this one one pair.

Are you using the pre-columns?

Hi Lilly,

If your affected peaks are early or late in the chromatogram check for threads discussing dwell volume. A difference in dwell volume between two systems can produce changes in gradient resolution. Good Luck.

When you use a gradient, your begin time for the gradient should be the same at the two systems: I will explaine this:

If you start with a gradient at the begin of the run on a system with a dwell volume of 1.5 ml, at a flowrate of 1.0 ml/min gives this a chromatogram where the first 1.5 min are isocratic and then starts the gradient.

If you use a system now where you have got a dwell volume of 1.0 ml, at the same flowrate,
this gives a chromatogram where the first 1.0 min is isocratic,

This will lead to a 0.5 min earlier start of you gradient, wich can cause different resolutions.

Yes I'm using a pre-column, I will check the dwell volumes.

Any other ideas?

Thank you all!

Lilly

The column heaters between the 2 systems (A1100 and Alliance) are different and can give different column temperatures. Typically, when the 1100 heater is set at a given temperature, the actual column temp is somewhat lower than the setpoint temperature. On the Alliance, the setpoint is much closer to the actual column temp. Differences in column temperature will certainly affect your separation. I have seen this problem several times when tranfering methods back and forth between the 2 systems.

A.A.

During our calibrations of both Alliance and Agilent, the Alliance system has consistently had a greater difference between the set temperature and the actual temperature. This can be a real problem if your method is run a 25 C since the Alliance system contains a column heater and the Agilent system use both a heater and a chiller.

Column heater is set at 40 degrees, I will investigate this!

Thanks,

Lilly

Just to clarify, if you take a temperature probe an place it against a column in an A1100 column oven at 40 C, you will read a temperature between 32 and 39 C depending on where you measure. Do the same thing with an Alliance and you get readings between 39-40. Also, an Alliance heater needs to be set at 5 C above ambient temp for optimum performance, so as anon points out, at 25 C there could be an issue. During some method transfer experiments, we found that to transfer a method from the 1100 at 40 C, we needed to set the Alliance at 35 to get about the same separation.

A.A.

If you go to the Waters web site and do a search (maybe look for dwell volume) you should see an article on gradient robustness. I think this will give you the answers that you need. The variables are many and include the slope of the gradient the expansion volume available and many other things that are included in an application note. Also remember that HP is a low volume system and Waters is not.

Lilly,
Another check on the dwell problem. Find the dwell problem on each manufacturer's equipment, start the gradient, and then inject the sample when the dwell volume has passed the injector, sort of catching the wave. If that does bring the two resolutions closer together in value, the problem lies someplace else.
Jim

Danny,

The volume of a standard A1100 is between 743-847
uL (it is variable because of the pluse dampener). The volume of a standard Alliance 2690/2695 is 650 uL. I am interested in how you conclude that "HP is a low volume system and Waters is not".

A.A.

Another possible problem is the difference in mobile phase preheaters in the Alliance and Agilent systems. The Alliance preheater is a small coil of tubing prior to the column and the heating is provided by the air ciculation in the oven while the Agilent uses a coil of tubing which is part of the heating element. In our work, we have found that for some separations this difference can effect the peak shape and hence the resolution

I don't know about this, Andy. In my Alliance system, the preheat coil is buried in between metal plates that are on top of the heater.

Lilly, I' m going to jump back a bit here and ask some questions:

- is the difference in resolution caused by a difference in retention times or a difference in peak widths? (i.e., are the peaks coming out at the "wrong" time, or are they at the "right" time but just too fat.)

- have you tried swapping the columns between the two systems? (to eliminate the possibility that you just have a bad column).

Assuming you have exonerated the column, many of the things to look at have already been suggested.

1. If the peak widths are OK, but retention times are off:

- is the temperature actually the same on both systems (not just set the same). If you're using the same column, look at the comparative back pressures for a clue; higher temperature usually results in lower pressure, so if the pressure is significantly different, this suggests temperature as the culprit.

- if the temperature is the same on both systems, then I would concur with Jim Gorum that dwell volume is the next likely suspect. Do *not* rely on the manufacturers's specs for this; the plumbing may have been changed since it was manufactured (changing the loop size in the injector can change dwell volume). Checking dwell volume takes about an hour, but it's well worth it. I can give you the procedure if you e-mail me privately.

2. If the retention times are OK but the peak widths are bad, there are another whole range of possibilies:

- is your sample dissolved in something significantly stronger than your initial mobile phase? If so, you may be seeing a "smearing" of the sample as it sticks to the column. Ironically, "better" HPLC hardware (less extra-column volume) can have more serious problems in this regard (pre-column mixing volume allows the sample to be somewhat diluted by mobile phase before it hits the columns).

- same size sample loops on both instruments?


-- Tom Jupille / LC Resources Inc.

Tom,

Peak width is the problem (too wide)r.t are ok. I tried a few columns, instrument conditions are the same. Can the legnth of the tubing from the column to the detector have effect?

One thing I forgot to mention, two detectors are used UV and Fluorescence; UV first then fluorescence. UV results are the same for both instrument but for Waters 2690 fluorescence results are not acceptable.

Lilly

Now we are getting somewhere. It appears that all your peaks are getting broader, and that the problem is with the second detector in series.
If this is correct, I would check out the outlet tubing of the first detector that connects to the second detector. It is also possible that a fair amount of bandspreading happens already before the first detector, but you should see it. The peaks should be broader already in the first detector.
Most of the time, the outlet tubing of a detector has a large i.d.. See if you can replace it with a tubing with a smaller i.d. If nothing else is possible, just cut it as short as you can.

Uwe;

You saved my job:-) It was the tubing! Replaced it with a smaller one and BINGO got a resolution of 4.8!!!

THANK YOU ALL!!

Lilly