Home » Meet the Experts

Top 10 Tips for GPC Sample Preparation

15 November 2013 No Comment

Written by:

GPC-QnA-arrowsLos Angeles, CA, 12 November 2013. In case you missed it, Matt McGann gave a well-attended webinar where he shared his ‘Top Ten’ on sample preparation for GPC measurements. Since there were several questions during the live event we would like to share the answers with you here. As a highlight, here is Matt’s list of his recommendations upfront:

Top 10 Tips for GPC Sample Preparation

  1. Prepare a fresh sample
  2. Be sure sample and solvent are compatible
  3. Do not use samples or standards over extended time periods
  4. Measure your sample out accurately
  5. Filter the sample before injection
  6. Choose an appropriate concentration
  7. Keep Solvent Enhanced Light Scattering (SELS) in mind
  8. Temperature control of samples is important
  9. Degas samples if necessary
  10. Is your sample ready for injection?

Examples for each of these were shown and discussed in the presentation (i.e. included in the recording). Since this topic resonated with the audience, here are some questions and answers provided during the discussion session after the formal talk ended.

Question: Would the dn/dc of the PLA be calculated for acetone?

Matt: I believe this question is referring to Tip 7 – Using Solvent Enhanced Light Scattering. This technique dissolved a polymer in its most soluble solvent, and eluted the material in a solvent more appropriate for light scattering detection by improving the dn/dc. If PLA was dissolved in Chloroform, and eluted using acetone, when the sample reaches the detector, the chloroform will have separated from the PLA on the column, so the detector will therefore observe the PLA in Acetone. If the dn/dc value is set to zero in the method in the OmniSEC software, the software will automatically calculate the dn/dc of PLA in Acetone.

Question:  In Solvent Enhanced Light Scattering experiments, which dn/dc do we use for quantitative analysis?  dn/dc for the eluent solvent or dn/dc relative for the dissolution solvent?

Matt: When setting up the parameters for a SELS experiment, the dn/dc entered would be that of the polymer dissolved in the eluent. As the sample and its dissolution solvent will be separated on the column the dissolution solvent should elute with the void volume towards the end of the chromatogram, well after the polymer.

Question: When I use different solvents, how can I care for my columns?

Matt:  It is recommended by Malvern that the Viscotek size exclusion columns be maintained in a single solvent. Changing of the mobile phase in a column is always challenging. The change in polarity, viscosity, density and pressure can cause the structure of a column to compress, and crush leading to poor performance. Changing solvents can also cause the substrate to swell, essentially blocking the column and crushing the structure. All of these outcomes will essentially mean the death of your column. Malvern have a range of Viscotek columns which are shipped in specific solvents from THF, and Chloroform to DMF and HFIP.

If it is essential to change the solvent in a column, the rules are to ensure that no salt is present in the column by flushing it with 2 column volumes of pure solvent. Solvent exchange should then be completed at a very low flow (~0.1mL/min), and performed outside of the instrument. All solvent should be diverted directly to waste, and never allowed to flow into the detectors. This process should be performed as little as possible on any column. Further useful information can be found in our technical notes on “General care for a GPC system”, including columns & detectors, and on “Guidelines for aqueous mobile phases”.

Question: Can you advise what a “reasonable” concentration range would be for a polymer to inject ?

Matt: The chosen concentration for any polymer is related to its dn/dc value. The lower the dn/dc the higher the concentration required to achieve a reasonable signal. With that value in mind, the starting concentration range of any material is recommended to be between 1 and 5 mg/mL for most polymers (for example a chromatogram of PS in THF). These values are also appropriate for protein samples (see for example chromatogram of BSA).

Question:  How often do you recommend to calibrate with the standard if you don’t make any changes to the system? 

Matt: This question is one of the most commonly asked questions by my customers, and is a very interesting one to answer. Calibration in the Viscotek system can be performed with every sample, or performed once and that calibration applied to every sample you measure. It’s entirely up to the operator how often this is performed. The question that I would ask the operator is how sure can you be your system has not changed between yesterday and today? Has your solvent composition changed? Is your lamp intensity identical? There are many small changes from day to day which can lead to large changes in instrument response. All of this culminates in the answer you produce being error prone. It is with these issues in mind an operator should decide on the need for a new calibration.

Question:  ­I have a question regarding the solvent selection. It is recommended that we use chloroform for the columns which we have in our lab. So we can only measure the MW of polymers which are only soluble in Chloroform?­

Matt:  As discussed in a previous response above, the changing of solvent in a column is not highly recommended. It can be performed, but is often troublesome to do. With that in mind, my recommendation would be for a system in chloroform, measuring samples which are soluble in chloroform only. It is possible to use the SELS technique by dissolving a material in another more suitable solvent, and eluting with chloroform. However, this technique would still require good solubility in chloroform to be successful. Please also take a look at MRK1662 on the general care for a GPC system, where the issue of different solvents within the same chromatography system is addressed.

Question: ­The standard PS (=polystyrene standard) is available in solid form in bottles. For preparation we have to add 10 mL to each bottle but for actual measurement we only use less than one mL. So, we can’t use the rest for the next time?­

Matt: A Viscotek standard bottle is made by adding 10mL of solvent to the bottle as you note. For most systems a single injection of 100μL is sufficient for calibration. This leaves a substantial volume of material left at the end of a sample analysis. Although I still encourage making and using fresh standard especially if your data is valuable or critical it is possible to save and store your PS standards. I would recommend dividing the sample out into 10 different vials, sealing every vial tightly and wrapping each of them up separately using parafilm or similar laboratory wrap. Finally, placing these vials at low temperature in a fridge or freezer (<4 °C) will not only prevent the solvent from evaporating but also prevent the sample from degrading. A Standard can be stored in this manner for 3 – 6 months.

Question:  I also have a question regarding the baseline. It is recommended to have a straight baseline but if you zoom in, you can see some noise, similar to the noise that you showed when we have aggregates. My question is what amplitude of noise is accepted? ­

Matt:  In order to discuss what a good versus a bad baseline is, we need to first understand what the expected noise levels are for each of our detectors. The short term noise level for a good clean system should generally be <2-4mV for each of the three detectors in the TDA. The long term noise should be <10-15mV per hour. Finally with that in mind, it is important to keep the total scale of expected signal in perspective. If your detector produces an absolute response of 10-20mV to your sample, but the noise level was 5mV to begin with, then any amount of additional noise is going to show up significantly in your data. If your signal response was 500mV however, a small amount of noise caused by incomplete dissolution may simply disappear into your baseline.
Ultimately, there is no simple response to this question, as it is impossible to put a definitive number of what is good and bad. The most straightforward answer I could give is to keep things in perspective. If you suspect that a result shows signs of incomplete dissolution change your dissolution process and try again.


If you have any questions, please email me at ulf.nobbmann@malvern.com. Thanks!