Two sides of the same coin: Compositional analysis on polymer mixtures and copolymers

The Malvern Panalytical compositional analysis coupled with GPC can provide the key to characterizing copolymers to the next level. In this application note you will find out how we analytically breakdown copolymers into their component parts and what extra information we get when we do so. The specific examples analyzed in this application note are blends of polystyrene and polymethylmethacrylate, polystyrene and polyvinyl chloride, and polystyrene polybutadiene, as well as the copolymers poly(styrene-co-isoprene), poly(styrene-co-butadiene) and poly(styrene-co-acrylonitrile), but this can be applied to many more copolymers than just these and if you’re interested to know if your polymer is applicable contact us here.

Copolymers diversify the polymer market greatly as they can exert properties unreached by homopolymers. The properties produced by copolymers go beyond the mixture of monomers chosen to make the copolymer, but also how the monomers are distributed in the polymer chain. How monomers are added to a polymer creates different copolymer types such as random, gradient, graft, brush, or block copolymers. Block copolymers feature in this application note as they have properties unique to their random copolymer equivalents.

Characterizing copolymers in terms of their compositional parts is largely done as a bulk measurement where the ratio of the comonomers incorporated into the polymer chain can be determined. GPC provides a critical difference in that we separate our sample before it goes to the detectors, and by getting compositional data for each data slice it provides results that are much more information rich.

Challenges, however, exist when characterizing a copolymer by GPC coupled to differential refractive index (RI) and static light scattering (SLS) detectors. The challenges are how to determine the concentration of one monomer versus another and how to improve the dn/dc applied when calculating the molecular weight. (We have discussed what is a dn/dc value and what it means previously and if it is unfamiliar to you, you can read about it in this blog post) Our approach to this challenge has been to use a second concentration detector, a UV/vis spectrometer, in series with the RI and SLS detectors. The two concentration detectors in tandem are used to determine the concentration of component A and component B in the mixture or copolymer. Shown below, equations 1 and 2 describe how the concentration of each component gives the observed signal intensities in the RI and UV detectors. For this calculation we therefore must know the refractive index increment of components A and B ((𝑑𝑛∕𝑑𝑐)_𝐴, (𝑑𝑛∕𝑑𝑐)_𝐵) as well as the extinction coefficients of components A and B ((𝑑𝐴∕𝑑𝑐)_𝐴 , (𝑑𝐴∕𝑑𝑐)_𝐵). We measure the signal intensity of the RI and UV detectors (I_RI and I­_UV) and the detectors constants of those detectors (K_RI, K_UV) are predetermined in the calibration. With all of these parameters determined it leaves the simultaneous equations that output the concentrations of components A and B (c_A and c_B). The dn/dc of the copolymer may then be calculated using equation 3.

Polymer mixtures

To illustrate the possible results gained when analyzing a formulation, blend, or mixture we applied this analysis to a series of polymer mixtures. Samples were prepared by mixing solutions of polystyrene with polymethylmethacrylate (PMMA), polybutadiene (PBd), and polyvinyl chloride (PVC) of different molecular weights and ratios of the two components. In Table 1, below, we can compare the ratios of component A to component B prepared versus those calculated using a compositional analysis calculation method. It is evident that the calculated results follow closely to the prepared values, with error in calculation ranging from zero up to as far as 4 %.

Table 1. Six mixtures of polymer solutions prepared and analyzed by compositional analysis. Shown are the polymer type, molecular weight (Mw), and ratio of the two polymer types (A,B) used to prepare the mixtures, and finally the calculated ratio of the two polymer types.

	Polymer A	Polymer B	Mw A (kg/mol)	Mw B (kg/mol)	A:B prepared	A:B calculated
Sample 1	PS	PMMA	247	93	52:48	54:46
Sample 2	PS	PMMA	247	93	36:64	36:64
Sample 3	PS	PMMA	247	225	55:45	58:42
Sample 4	PS	PMMA	247	225	80:20	82:18
Sample 5	PS	PVC	46	77	48:52	46:54
Sample 6	PS	PBd	11	8	45:55	49:51

The variation of ratios, polymer types and molecular weights illustrates that we can use this method freely, if we have the dn/dc and dA/dc values for the two components. Although polystyrene is used here in all the mixtures, it is not necessary for it to be in your mixture. When at least one of your components is a chromophore, you can apply this method.

Copolymers

Polymers made of PS, PBd, polyisoprene (PI) and polyacrylonitrile (PAN) were purchased from Sigma Aldrich. Two of the samples were defined as triblock copolymers: PS-PI-PS and PS-PBd-PS, whereas the type of copolymer was not defined for the other two samples: PS-PI and PS-PAN. Two overlays of the acquired chromatograms are shown below in Figure 1, firstly to the left is an overlay of the samples as measured by the RI detector, and to the right as measured by the UV/vis detector at 254 nm. Each sample is a different color: PS-PI-PS - red, PS-PI - purple, PS-PBd-PS - green and PS-PAN – black. Comparing the two samples made of PS and PI, we see that the PS-PI-PS and the PS-PI samples have similar shaped chromatograms and peak intensities in the RI, however relative intensities between the two are quite different in the UV detector indicating a different chromophore content and therefore less polystyrene in the sample PS-PI. As evident in the expanded portion of the RI detector three of the samples contain a separate peak at a higher retention volume, which points towards a lower molecular weight impurity. The PS-PBd-PS sample seems to have an initial peak at 14 mL followed by the main peak, it’s difficult to speculate what this may be, perhaps one might think for example that this part of the sample contains another block in comparison to the main peak. Looking at the chromatograms of sample PS-PAN, it appears as though the sample has a wide molecular weight distribution as it elutes over a large volume range (from circa. 12.5 to 18 mL).

 
Figure 1. Overlay of RI chromatograms (left) and UV/vis chromatograms (at 254 nm) (right) of PS-PI-PS - red, PS-PI - purple, PS-PBd-PS - green and PS-PAN – black. An expanded view of peak 2 is shown in the RI.  

The samples were measured in duplicates and the average results are shown below in Table 2. The weight-average molecular weight (Mw), number-average molecular weight (Mn) and dispersity (Mw/Mn) were calculated using a dn/dc calculated at each point in the sample’s chromatogram to improve the accuracy of the calculations. The weight fraction (Wt Fr.) of each component A and B is also shown in Table 2. The secondary peak evident in the samples PS-PI-PS, PS-PI and PS-PBd-PS eluting at roughly 17-18 mL was characterized separately from the main peak and indicated in the table as peak 2. Some interesting results were observed from this. In all three samples the secondary peak (peak 2) consists of a majority of polystyrene, even when the main peak consisted predominantly of component B. This could indicate the need to further develop the synthesis as low molecular weight species could alter the thermal and physical properties and perhaps even the self-assembly of the copolymer. We saw that the samples PS-PI-PS and PS-PI have similar elution profiles (Figure 1), this is also reflected in the molecular weight, however compositional analysis allows us to see that they contain clearly different quantities of PI in the main peak, 77.6% vs 84.7%.

Table 2. Copolymer samples quantitative results using compositional analysis: Molecular weight (Mw, Mn), dispersity (Mw/Mn) and weight fraction (Wt Fr.) for primary peak (1) and secondary peak (2) of the four samples PS-PI-PS, PS-PI, PS-PBd-PS and PS-PAN.

Sample	Peak	Mw (g/mol)	Mn (g/mol)	Mw/Mn	Wt Fr. (A) (%)	Wt Fr. (B) (%)
PS-PI-PS	1	134100	125200	1.07	22.4	77.6
	2	25600	21900	1.17	90.2	9.8
PS-PI	1	136600	123500	1.11	15.3	84.7
	2	24100	20100	1.20	82.7	17.3
PS-PBd-PS	1	67500	59400	1.14	25.9	74.1
	2	16600	15200	1.09	98.7	1.3
PS-PAN	1	169200	73600	2.30	67.8	32.2

The results can also be shown graphically, where any of Wt fr. A, Wt fr. B, concentration A, concentration B, molecular weight A, or molecular weight B may be presented as a function of time or alternatively, Wt fr. A or Wt fr. B can also be graphed as a function of molecular weight. These graphics can present an interesting insight into how each part of the molecular weight distribution may be composed of the two components. This could tell us for example whether the composition is homogeneous, the weight fraction is steadily changing, or the chemical makeup of certain parts of the population are significantly different. The samples PS-PBd-PS and PS-PAN provide good examples of how these graphs can be very informative. The samples PS-PBd-PS (Figure 2. Left) and PS-PAN (Figure 2. Right) are graphed with an x axis of retention volume, and two y axis: firstly Wt fr. A and secondly RI. Recalling our previous comments on the chromatograms of sample PS-PBd-PS it was unclear what the initial peak eluting at 14 mL is. The Wt fr. A graph indicates that this initial peak contains close to the same fraction of PS as in the rest of the main peak which may indicate that it is a discreet molecular weight population of the triblock copolymer, a product of a coupling reaction, or be a feature of chromatography. However, what we can conclude is that this peak is not a homopolymer and it does not have a different ratio of components than the main peak. The sample PS-PAN presents a different picture. We observe that Wt fr. A varies from 0.66 up to 0.69 from 13 – 16 mL, and then for lower molecular weight it tails and the Wt fr. A increases to 1, indicating the lower molecular weight chains consist more of PS than the main body of the peak.

Figure 2. An overlay of the RI chromatogram and Wt fr. A for samples PS-PBd-PS (left) and PS-PAN (right).

Conclusion

In this application note we have presented the analysis of polymer mixtures and copolymers by multi-detection chromatography on the OMNISEC system. The compositional analysis was shown to be applicable to a range of polymers consisting of polystyrene, polymethylmethacrylate, polybutadiene, polyvinylchloride, polyisoprene, and polyacrylonitrile. The samples analyzed represented a mixture of sample types including blends, random copolymers and block copolymers. Initially we showed that with compositional analysis we could accurately calculate the composition of the polymer mixture into their component homopolymers. Following this we showed the molecular weights calculated for copolymers with improved accuracy due to dn/dc calculation at each data slice. We presented the range of additional results given by compositional analysis and we illustrated the application of weight fraction in depth including the interpretation of two plots of Wt fr. A as a function of retention volume. If you would be interested to discuss this further, please get in touch with us here.