Chromatography Forum

Does any one in here had any experience in GPC measuring high MW polymers. I am trying to determine the MW of high MW polyethylene oxides and i am getting pretty good results in the low end ( till 2 Million). Anything over that i am getting unreproducible results. I tried running some 4Million and 5 Million MW polyethylene oxide and i get relative low numbers ( ~2M) and also unreproducible results. Can anyone please shed some light on this issue. I would greatly appreciate any help..Thanks.

I am using good Ultrahydrogel column 250, 1000 and 2000 in a series.

I can think of a few things:

1. You may be pushing the exclusion limit on the column.

2. Your preparation could be knocking down the MW of your PEO.

3. The MW of your PEO could have been determined via a different method - viscometry or light scattering - which may not correlate perfectly with GFC at the high end.

Side note: I've found that for molecules such as hyaluronic acid, GFC results (with Ultrahydrogel columns using narrow MWD PEO standards) correlate well with mol wts measured via flow viscometry.

Good luck!

The MW of the high-end analytes is still within the range of the Ultra-Hydrogel 2000 column. So this is not the problem.

I would first look into details of dissolution and injection. Compounds with such a large molecular weight are fragile. They can break down due to shear in the syringe, for example, or if you are trying to dissolve them by shaking the vial. You need to treat such samples really carefully. If you can exclude this as the problem, I would next look for plugged frits in any column in front of the 2000, including precolumn filters, if you use such a thing.

Your calibration can also be an issue. The standards you use to set up your curve must bracket the retention range of your sample. If not, the extrapolated curve can do some strange things, leading to odd results. I have seen products go in and out of specification, simply by changing the calibration range. And it is the portions eluting at the extremes of the retention window that cause the problem (high MW in your case).

I have run high MW PEO up to 13 million, what injection volume and concentration do you use? I find that we need to lower the concentration to less than 0.25 mg/ml and dissolve for 24-48 hours and then inject 50 - 100 micro liters.

[quote="juddc"]I can think of a few things:

1. You may be pushing the exclusion limit on the column.

I am using ultrahydrogel columns in series. 2000,1000 and 250

2. Your preparation could be knocking down the MW of your PEO.

The same preparation has been used for Viscosity tests and it yields
higher MW

3. The MW of your PEO could have been determined via a different method - viscometry or light scattering - which may not correlate perfectly with GFC at the high end.
Side note: I've found that for molecules such as hyaluronic acid, GFC results (with Ultrahydrogel columns using narrow MWD PEO standards) correlate well with mol wts measured via flow viscometry.

We are using Viscosity as an ulternative way to test, but I would like to measure them through GPC as well.

When you say viscosity results are you referring to a batch mode viscosity measurement, like a cannon fenske solution viscosity, or a viscometer detector for GPC? I have found that the concentration of the solution viscosity measurement is usually higher than for a GPC experiment. Our High Mw PEO samples were run using GPC with very reproducible results using light scattering detection. If you need more details, please let me know.

I have run high MW PEO up to 13 million, what injection volume and concentration do you use? I find that we need to lower the concentration to less than 0.25 mg/ml and dissolve for 24-48 hours and then inject 50 - 100 micro liters.

Wow 13 million!!!!! I am using 100ul and concentration is 0.2mg/ml and we dissolved it for 24 hours and still i cant get higher than 2.5M MW. What kind of columns do you use and what mobile phase?

When you say viscosity results are you referring to a batch mode viscosity measurement, like a cannon fenske solution viscosity, or a viscometer detector for GPC? I have found that the concentration of the solution viscosity measurement is usually higher than for a GPC experiment. Our High Mw PEO samples were run using GPC with very reproducible results using light scattering detection. If you need more details, please let me know.

We are a using cannon fenske solution viscosity not a viscometer detector for GPC. For GPC we are using a RI detector. I am definitely interested in more details, i would greatly appreciate any help you can offer. Thanks.

Our work was done on a Polymer Standards Service Suprema Aqueous GPC column with a 30,000 angstrom pore, 10 micron particle, 8 x 300 mm column. Flow rate was 0.3 ml/min, injection volume was 50 microliters, concentration and preparation of sample was as in my previous post. We found that if you tried to run the higher Mw PEO materials at flow rates greater than 0.3 ml/min they tended to shift to lower molecular weights, i.e. they all appeard to be near 2 million. Slowing the flow rate allowed the bigger molecules sufficient time to diffuse though the pores. These big molecules need tome to relax and flow through the GPC column. Also, we used a large inlet fritte than standard at the column ends in order to avoid shearing at the inlet and outlet of the column. Results were consistent with ultracetrification results. Link for the column is here:

http://pssgpcshop.com/mm5/merchant.mvc? ... de=suprema

Let me emphasize again that the way that you are injecting the sample can cause them to break down and create a lower-MW profile than anticipated. For both manual and automated injection, you need to slow down the speed with which the sample is either sucked into the syringe, or loaded into the autoinjector.

A lower concentration helps also to reduce the viscosity of the sample. For this molecular weight of this polymer, concentrations under 0.01% are recommended.

The lower flow rate might also work with the Waters columns you currently use, I only was letting you know what I used in my own work. Uwe's comments on the injector has some validity, but your symptoms are more like what we saw when developing the method for Ultra High Mw polymer analysis we do by SEC/GPC. For example, we have seen in sample preparation that if you stir the solutions using a stir bar the high molecular weight can be reduced through chain breakage by the drag of the polymer against the glass beaker. We find the key is concentration of the sample, flow rate on the column, dissolution time, I have done work with both auto samplers and manual injectors with success. Dissolution is done over 24 - 48 hours and just letting the sample sit at room temperature and gently moving the solution periodically to get a homogeneous solution.

Our work was done on a Polymer Standards Service Suprema Aqueous GPC column with a 30,000 angstrom pore, 10 micron particle, 8 x 300 mm column. Flow rate was 0.3 ml/min, injection volume was 50 microliters, concentration and preparation of sample was as in my previous post. We found that if you tried to run the higher Mw PEO materials at flow rates greater than 0.3 ml/min they tended to shift to lower molecular weights, i.e. they all appeard to be near 2 million. Slowing the flow rate allowed the bigger molecules sufficient time to diffuse though the pores. These big molecules need tome to relax and flow through the GPC column. Also, we used a large inlet fritte than standard at the column ends in order to avoid shearing at the inlet and outlet of the column. Results were consistent with ultracetrification results. Link for the column is here:

http://pssgpcshop.com/mm5/merchant.mvc? ... de=suprema

What range of standards did you use for you calibration curve. We are currently using a narrow range of standards ranging from 500,000 to 1,250,000M MW. If I use this curve, I would have to extrapolate when analysing higher MW, so I am wondering if at high MW would the calibration curve still be valid.
Also I just ran a blank sample with an internal standard at .3ml/min and my retention time jumped to 2 hours ( previously was 55 min). Now I am thinking since PEO has a wide MW distribution, at this low flow rate is it just going to spread out. I mean would i be able to differentiate between a 4M MW vs 5M MW PEO or is just going to overlap? What you guys think? I really do appreciate all the help you guys have given me. Thanks!

Anytime you go outside the calibration range, you are risking error. But the situation is more important in GPC because the curves are not linear (e.g., 5th order polynomial). So, the shape of the curve "outside" the calibration range will depend very much on the nearest standard, and could be much different from reality. You could do some plots in a spreadsheet to see what the curve does, but remember that you are nearing the exclusion limit for the column, and so the curve is running up against a fixed value (the exclusion limit). The GPC curves will show a steeper negative slope in this region, and if your extrapolation doesn't do that, then beware.

Can you get a wider range of standards?

A MW determination that covers a response outside the range of the calibration curve is simply invalid. You absolutely need to bracket the range of your analysis.

Your overall results in SEC improve at lower velocity, but only slightly. You will be able to resolve narrow MW standards better at lower flow rate. If you can resolve a 5M from a 4M standard depends not only on the flow rate, but also if you have chosen the correct pore-size packing that gives you maximum resolution for this range.