Drift in GPC

[kac]

Hi

I have a problem with drift of retention times in a GPC method.

The method in question is used to measure pore sizes of silica gels. A gel is packed in a column 50 cm long and 4 mm in diameter. Seven solutions of polystyrene standards are chromatographed in sequence on the column at room temperature using dichloromethane as mobile phase. The total run time is nearly 1 hour.

The pore size of the gel is calculated from the retention times of the polystyrene standards. The result is very sensitive to drift in the retention times and many runs must be rejected because of a drift that is too high (a drift of 1 % can be accepted, but less is desirable; a drift of 10 % has been seen).

We don’t know what causes the drift but we do not believe the HPLC equipment is the problem. Could it be poor temperature control? Alterations of the room temperature seldom exceed ± 2 °C, I think.

[Rolandas Plausinaitis]

Hi,

I understood You use "new" column each time. Does retention time change between columns (RT will always depend on how the column was packed - the only way to improve measurements reproducibility is to standardize packing procedure as much as possible) or on the same column between measurement series?

[Uwe Neue]

This should be size-exclusion chromatography, and column equilibration of a silica-based column should not be an issue. I would think that the problem may very well be the flow rate. In SEC/GPC you need high precision and a high reliability of your pump. MeCl2 has a low boiling point, and your pump may not like it.

A few thing to check to verify that it is the pump: always inject a marker for the total permeating volume such as toluene together with your polystyrene markers. If this peak wiggles around, you have a problem with the pump.

Another possibility could be the age of the standard, breakdown of the standards, or related issues. What polystyrene molecular weights are you using? Are they fresh, or have they been sitting around for some time? If the latter, for how long? Are all MWs showing this problem of changes in retention, or only the very large ones?

[unmgvar]

Kac,

Like Uwe suggested your problem might be more related to the pump or solvent.
do you have an online degasser?
Are you using an isocratic pump or a regular LPG pump?
Some pumps need to have their pump seals changes when dealing with the organic solvents used in SEC. have you checked that?
Are you using the pump for regular liquid chromatography? if yes do you put buffers in there?
Do you sonicate your MeCL2?
Do you helium sparge the mobile phase? Do you vacuum degas prior to putting it in the system?

Their might be also something esle related to the mobile phase and column. do you use that column for different applications with different type of organic solvents? This can cause swelling or shrinking of the sorbent phase. This could require longer equilibration times for your column and it could when the transition is not done correctly between the solvents simply kill the column.

[HW Mueller]

Did anybody ever get a silica column to swell appreciably? Did the silica crack? Maybe afterwards it could be used for UPLC?

[kac]

Thank you for your interest and quick response.

Five of the seven solutions we inject in a series contain toluene, and it is from the retention times of the toluene peaks that we discover the drift. We may run three series in succession, the first two showing drift, the third showing no drift.

The stability of our polystyrene standards is tested regularly and is very fine. The retention times of the peaks are very stable and do not wiggle around. When we experience drift we see a steady, slow increase of the retention times of both the standard with the highest molecular weight (182100 Da) and of toluene. The drift is, of course, most apparent at the retention times of toluene.

[unmgvar]

Kac,

just a small interlude regarding SEC columns, for HW.

the particules are with pores and some of them of course are relatively big ones. this geometry is different then the regular silca gel which are more like hard balls in comparison.
when you have to switch between solvents you many times have to use in-between solvents to make the transition. some of them like methanol, acetone can cause those behaviours of swelling or shrinking.
for that reason manufacturers sell prepack SEC columns that are also pre-washed and shipped in different solvents. for example:

http://www.phenomenex.com/Phen/Doc/zhsdx.pdf

as seen in page 2 and 3

then of course it is not advised to switch between solvents with SEC columns

[unmgvar]

Kac,

can you go through my questions please, they go through some of the more common reasons that there is a RT drift in SEC.

from what you show it looks like you might have an equilibration matter of the column.
do you turn off the pump?
SEC columns are still more sensitive to pressure changes then LC columns. with them it is still advised to increase and decrease flow slowly.
if you can don't stop the flow, recycle the solvent then you get ride of the problem all together

[Uwe Neue]

If all peaks drift or shift, the most likely cause of the problem is the pump.

The second possibility is that you dissolve something out of your pores. This would agree also with the idea that two injections show a drift, and the third one does not show this any more. What dissolves is not the silica. Silica is insoluble in MeCl2, but it might be water or other strongly adsorbed stuff on your silica. Check the history of the silica, including the solvents and reagents used for column packing.

[HW Mueller]

unmgvar, your link , p. 2 & 3 seems to be about polymer columns. So I stiill havn´t seen anything on silica swelling.

[unmgvar]

thanks HW, you are right. can't believe i missed it twice.
i have to stop logging in so late, it will bite me one day.

kac as HW points out i was talking for polymer SEC columns all the time. i missed the fact that you were using silica. sorry.

regarding Uwe's last post. is your MeCl2 dry? have you tried using a "half dried" solution? according to my "Uwe's troubleshooting guide" this does shorten considerably the equilibration neede for normal phase columns.

also Kac if what Uwe suggest is right you are doing SEC with also an effect of absorbtion chromatography, hence the high influence on the toluene.
i am curious in knowing do you calculate somehow the ratio between the 2 for your chromatography? can it be done. doesn't this effect the reliability of the SEC separation if i am using different STD types of polymer in conventional calibration?