Gel permeation / size exclusion chromatography (GPC/SEC) is the most well-known method for polymer and macromolecular analysis. As the technique has become more pervasive, there has been a push toward greater efficiency, by way of reduced run times and minimal solvent and sample consumption. The development of Advanced Polymer Chromatography (APC™) systems has accommodated these issues, by employing low volume columns and pumps with the ability to handle higher pressure generated by these columns. These features allowed for more efficient polymer analyses, but the technique still relies on the need for column calibration to provide relative molecular weight values.
While APC represented the evolution of GPC/SEC to a quicker, more efficient method, the development of advanced detectors provided an orthogonal means of improvement. Increasingly sophisticated detectors, including light scattering and viscometer detectors, have been successfully incorporated in GPC/SEC systems to offer complete characterization of a sample, culminating in the OMNISEC REVEAL detector suite within Malvern Panalytical’s OMNISEC system.
Recently, Malvern Panalytical has introduced a version of OMNISEC REVEAL that is compatible with high pressures and minimal band-broadening conditions and, through a collaboration with Waters and their ACQUITY APC system, has brought advanced detection to APC analysis.
Figure 1: The ACQUITY APC from Waters and OMNISEC REVEAL from Malvern Panalytical.
To determine if the efficiency of the ACQUITY APC unit and the power of the advanced detection of OMNISEC REVEAL were additive, two polymer samples with different structural features were analyzed. The goal of this experiment was to obtain enough characterization data to differentiate between linear and branched polystyrene samples – all in a matter of minutes.
Two polystyrene samples A and B were prepared and analyzed in THF at a flow rate of 1.0 mL/min using a set of three 150 mm ACQUITY APC XT columns: 45, 125, and 450 Å, respectively, in series. The column and detector temperatures were set to 40 °C. The concentrations of the polystyrene samples were 1.31 and 1.92 mg/mL, of which 19 and 14 µL injections were made for samples A and B, respectively. The mass introduced to the system was 24.9 and 26.9 µg, respectively; much less than a typical GPC/SEC analysis which might use 300-500 µg per injection.
Regular GPC/SEC analysis requires more sample per injection because the column volumes are much larger, which results in broad peaks as compared to the sharper peaks generated with APC columns. To compare the efficiency of the ACQUITY APC-REVEAL instrument to that of typical GPC/SEC analysis, samples A and B were also analyzed using two 30 cm mixed-bed analytical GPC/SEC columns. The refractive index chromatograms from both GPC/SEC and APC using the ACQUITY APC-REVEAL system for samples A and B are overlaid below in Figure 2. The difference in mobile phase and analysis time required is visibly obvious, and the smaller sample peaks produced using the ACQUITY APC-REVEAL setup indicate less sample was used. Even with the reduced signal intensity, the sensitivity of the OMNISEC REVEAL detectors eliminate any potential compromise in signal quality. While more than 20 mL of solvent is required to produce the broad peaks for samples A and B eluting from the pair of 30 cm mixed-bed analytical columns, the samples elute from three APC columns in less than 6 mL!
Figure 2: Refractive index chromatograms of the linear (red) and branched (purple) samples on two analytical 30 cm GPC/SEC columns overlaid with the same linear (green) and branched (black) samples analyzed using three APC columns.
The triple detector chromatograms for samples A and B, including refractive index, right angle light scattering, and viscometer detector signals, are presented in Figures 3 and 4. The data indicates good signal to noise for all detectors, as well as quality chromatography, exemplified by the return to baseline after the peak for all detector signals. The molecular weight of each sample is plotted over the top of each peak, indicating a gradual decrease in molecular weight with later elution volume, and thus molecular size.
Figure 3. Triple detector chromatogram of polystyrene sample A; refractive index (red), right angle light scattering (green), viscometer (blue) detectors and molecular weight (gold) are presented.
Figure 4: Triple detector chromatogram of polystyrene sample B; refractive index (red), right angle light scattering (green), viscometer (blue) detectors and molecular weight (gold) are presented.
Samples A and B both elute with similar peak shapes from about 2.75 – 4.5 mL. A calibration curve analysis method using only a refractive index detector would determine that these two samples have a similar molecular weight, based on their comparable retention volume. The addition of light scattering and viscometer detectors offer more accurate and further insight to the samples’ molecular structure. The molecular characterization data for samples A and B is summarized in Table 1.
Table 1. Molecular characterization data for polystyrene samples A and B using multi-detector analysis to generate absolute molecular weight.
The similar retention volume (RV) for samples A and B is consistent with the comparable hydrodynamic radius (Rh) values observed. In fact, the samples are so similar in size that a conventional calibration analysis (Table 2) yields a relative molecular weight of around 340,400 Da for sample B, which is only slightly larger than that of sample A and in line with the Rh change. When compared with the light scattering data, this is clearly in significant error.
Table 2. Molecular characterization data for polystyrene samples A and B using conventional calibration analysis to generate relative molecular weight.
In fact, the light scattering data reveals that samples A and B have significantly different molecular weight (Mw) values, 321,000 Da and 427,500 Da, respectively, which is interesting considering their similar molecular size. This discrepancy is resolved when the intrinsic viscosity (IV) values of the samples are considered.
The IV values are almost the same – 0.956 dL/g for sample A and 1.03 dL/g for sample B – yet the two samples have vastly different molecular weights. Since sample B has the higher molecular weight, it possesses more mass within the same molecular volume. Not only does this mean the molecular density of sample B is greater than that for sample A, but also that for any given molecular weight, the IV for that slice of sample B will be less than that of sample A.
This relationship between IV and molecular weight is illustrated in a Mark-Houwink plot, which positions a sample’s IV on the y-axis against its molecular weight on the x-axis. When the plots of multiple samples are overlaid, slight variations in structure between two samples become notable. The Mark-Houwink plots for two injections each of samples A and B are shown in Figure 5.
Figure 5: Mark-Houwink plots for two injections each of sample A (red & purple) and sample B (green & black).
Polymers with consistent structures throughout their molecular weight range have Mark-Houwink plots that appear as straight lines, and samples with similar structures will have plots that overlay. Samples with different molecular densities will appear “stacked,” with the densest material situated lowest in the plot (as molecular density is inversely related to IV).
From the data in Figure 5, it is clear from the red and purple plots that sample A is linear in nature, as there is a consistent relationship between IV and molecular weight. The lower molecular weight portion of sample B, in green and black, displays the same type of linear profile up until 400,000-500,000 Da. At this point, the plot for sample B exhibits some curvature downward as it deviates from the linear plot of sample A. This change means that sample B has a lower IV than sample A, indicating it has a denser molecular structure. The most common molecular feature that leads to this dense structure is branching.
The Mark-Houwink plot, combined with the numerical data differences between samples A and B, provides evidence that sample A is linear polystyrene and sample B is branched polystyrene.
This is an excellent demonstration of the shortcomings of conventional calibration measurements. Sample A is a linear polystyrene (and matches the conventional calibration standards), therefore its relative and absolute molecular weights are in agreement. However, in the branched sample B, the two disagree because conventional calibration does not account for the changes in molecular structure.
The combination of the Waters ACQUITY APC with the advanced detectors of OMNISEC REVEAL allowed the structural differentiation of polystyrene samples A and B discussed herein. The samples were characterized individually, and while both have similar retention volumes, the data obtained by the light scattering and viscometer detectors provided further insight. The differences that were imperceptible based on retention volume, hinted at in the numerical data, became fully illustrated in the Mark-Houwink plot. The minimal quantity of sample and solvent and reduced time required to obtain in-depth characterization of samples using the ACQUITY APC-REVEAL system are advantageous to researchers and manufacturers seeking to fully understand their samples.