Viscosity in HPLC SEC with Polysaccharides samples

[kurian]

They have a very poor intensity, that's reason we have excluded overloading (axe Y : 0 to 20 µRIU)
This chromatogram is for the good system, the other show another tailing with more difference (several waves after the apex)

If it was overloading, the tailing should be the appearance of high molecular weight released by the column.
However the MALLS shows a very good molecular distribution throughout the elution.

Waters system has probably a 200 µl loop and my system has a direct injection. For me, for me it correlates with a greater diffusion of PS with high viscosity in the system without loop.

We are not at the first transfer between the two systems for PS samples and we have habitually no problem.
Difference between old transfer and this one ? the viscosity of PS...

[danko]

Hi again,

First of all; column overload has nothing to do with intensity (i.e. detector overload). So, I still think that’s the problem.
Regarding the loop thing: If the waters system I Alliance and nobody installed another loop, then the loop I 100 µL.
Finally, how do the chromatograms from the good system look like?

Best Regards

[kurian]

Profiles showing here are from the good system.

[kurian]

Here is the chromatogram from Waters Alliance 2695 with RI detector 410 or 2414

[kurian]

Here is chrom from Merck-Hitachi.
3 different batch column TSK5000PWxl
Dextran showing is the 3 800 000 Da and we see difference between one batch and the two others.

Though difference between sample profile, reproductibility is good but not perfect.

Overloading column is the result of an high viscosity PS.
I think both are dependent, overloading is the consequence.
And if i have a greater overload on Alliance i have a sample more concentrated because the PS have not time to distribute (diffuse) in dead volume.

Is it correct ?

[danko]

I see the same tendency i.e. the more analyte the more peak deformation. And if you’re still in doubt you can just try and inject increasing injection volumes of the 2 mg/mL solution. F. ex. inject 25, 50, 100, 150 and 200 µL and you’ll see the same peak shapes as you see when you inject 200 µL of different analyte concentrations i.e. no, or negligible effect from sample solution viscosity.

Best Regards

[danko]

Kuran,

I had only seen your previous past i.e. the Waters chromatograms at the time I responded. Now seeing you last post, I can see it even clearly. If you compare the Y-axes of the two systems you’ll see that the Waters’ chroms peak heights are approx./roughly half the highs of the Merck-Hitachi’s. That indicates that you’re injecting (unintentionally) half the volume on the Waters system, which “thinksâ€

[kurian]

Both system inject 100 µl thus we have 200 µg injected on a RUN sample with a concentration at 2 mg/ml.

If we have half the volume on Waters than Merck, why they have more overload ?

[danko]

Sorry about the 200 µL misunderstanding. I thought you injected 200 µL, but you obviously didn’t.
Never the less you have much lower signal on the Waters sytem, which of course does not necessarily mean that you’re loading a smaller amount of the sample, but I would investigate that matter if I were you. As I wrote earlier I regarded that fact merely as an indication of a smaller load on the Waters system and it still could be the case – for one or another reason.
Anyway the most plausible explanation of the distorted peaks is still a column overload and as I mentioned earlier you can convince yourself of that simply by injecting increasing volumes of the same sample solution (i.e. 2 mg/mL) which means the viscosity is kept constant.

Best Regards

[kurian]

Of course, thanks a lot for your answers. It's very constructive for me.
I start this different volume injection for this week-end.

[HW Mueller]

Whew, and I don´t follow this any more even without weekend experiments.

[Uwe Neue]

Are you still using the buffer? And, if yes, are you making up your samples in the buffer?

[kurian]

Yes, for more facilities we keep using buffer but not for preparation of samples.

Other buffer give the same.

[Uwe Neue]

Unless you fix this problem of differences between the mobile phase and the injection volume, you will lead yourself and everybody else astray.

[kurian]

I have launched several injection volumes at 2 mg/ml and result was good to 50 µl (equal to 100µg) and gives me the same results as 100 µl at 1mg/ml.
So i think it was overloading like you.
But if viscosity play a game with dead volume, your injection volume have a greater diffusion if you decrease it.
Thus i finally injected 200 µl at 1mg/ml (equal to 200µg) and normally, if i have an overloading, it's expected to have a tailing.

But no...

Here is an abstract to a study PS viscosity (french to english traduction by Google)
We have shown that the effects of concentration below the critical concentration for the recovery of channels linked to the product of the intrinsic viscosity by the concentration of injection and that these effects must be minimized to obtain a reliable calibration columns