The GPC analysis of nylons and polyesters have historically been very difficult to do, with m-cresol at 100° C being the solvent choice for many years. There have been a variety of other solvent choices that workers have tried, but one that works quite well is hexafluoroisopropanol, (HFIP). HFIP has an advantage over m-cresol in that nylons and polyesters dissolve at room temperature. One disadvantage is cost: HFIP costs approximately $1,000 per liter. This is the reason we have investigated the GPC analysis of these two popular polymers using HFIP with solvent efficient columns, which are 4.6mm i.d. These are 30cm long columns, but the more narrow i.d. (as compared to conventional 7.8mm columns) allows for a large savings in solvent use (and disposal costs). The flow rate is usually ~0.35 ml/minute, which will give approximately the same eluent linear velocity as 1.0 ml/minute with the 7.8 x 300mm columns we usually use. These solvent efficient HFIP columns are specially packed in methanol for conversion directly to HFIP at 0.05 ml/minute.

For our analysis of nylons and polyesters, 0.05M sodium trifluoroacetic acid was added to the HFIP, to prevent any polar interactions. Nylons in particular will exhibit tailing on the low molecular weight end if the salt is not added to the HFIP. Once again, the Alliance GPC system with dRI detector was used for the analysis. Because of the low system volume (low dispersity) of the Alliance system, excellent resolution may still be obtained with the 4.6mm columns. We used a column designation of HR2, HR3, and HR4, which represents high resolution columns in the 500, 10³, and 10⁴ angstrom range. The RI and the columns were maintained at 30 °C and the injection volume was only 25 µl for the narrow PMMA standards and samples. Polystyrene does not dissolve in HFIP, so the narrow poly(methyl methacrylate) standards are used, and they work very well.

Here we show a third-order calibration curve for the PMMA standards in HFIP (triplicate injections of each standard). The extraordinary retention time reproducibility of the standards is obvious from the curve.

The first set of samples run was Poly(ethylene terephthalate), (PET) and Poly(butyleneterephthalate), (PBT) shown here.

Also shown below is an overlay of 5 molecular weight distributions of Nylon 6/6. A Nylon 6/6 broad standard was used for the calibration, so the molecular weights shown are "accurate" for the nylon sample.

The last work shown in HFIP is for two medical plastic grade Polyether/amide copolymers, used to make catheters. The two samples did not have the same physical properties nor the same "processability", yet FTIR, thermal analysis, rheological measurements, melt flow index, etc., showed no discernible differences between the two samples.

If you look at the two molecular weight distributions individually (shown here as 5 overlays), they indeed do look quite identical to eachother.

However, if you look at the overlays of the 5 MWD's for each here, you can easily see some differences between the two. The reproducibility of the Alliance system gives you the confidence to say that these small MWD differences are indeed real, and not due to injection-to-injection variability.