Separation of hydrophobic polymer additives by microemulsion electrokinetic chromatography
Hilder, EF and Klampfl, CW and Buchberger, W and Haddad, PR, Separation of hydrophobic polymer additives by microemulsion electrokinetic chromatography, Journal of Chromatography A, 922, (1-2) pp. 293-302. ISSN 0021-9673 (2001) [Refereed Article]
Microemulsion electrokinetic chromatography (MEEKC) has been applied to the separation of some phenolic anti-oxidants [Irganox 1024, Irganox 1035, Irganox 1076, Irganox 1010, Irganox 1330, Irgafos 138, Irganox 168 and 2,6-di-tert.-butyl-4-methylphenol (BHT)]. Due to the extremely hydrophobic nature of these analytes, they could not be separated using standard MEEKC conditions and two alternative approaches were investigated. Using an acidic buffer (phosphate, pH 2.5) to effectively suppress the electroosmotic flow, the addition of 2-propanol to the aqueous phase of the microemulsion buffer to improve partitioning of the analytes, and a negative separation voltage, separation of five of the analytes in under 10 min was possible. The second approach, using a basic buffer (borate, pH 9.2) and a positive separation voltage resulted in complete resolution of all eight analytes. A mixed surfactant system comprising the anionic sodium dodecyl sulfate (SDS) and neutral Brij 35 was used to reduce the overall charge and with it the mobility of the droplets, and hence the separation time. Using an optimised MEEKC buffer consisting of 2.25% (w/w) SDS, 0.75% (w/w) Brij 35, 0.8% (w/w) n-octane, 6.6% (w/w) 1-butanol, 25% (w/w) 2-propanol and 64.6% (w/w) 10 mM borate buffer (pH 9.2) the eight target analytes were baseline separated in under 25 min. For these analytes, MEEKC was found to be superior to micellar electrokinetic chromatography in every respect. Specifically, the solubility of the analytes was better, the selectivity was more favourable, the analysis time was shorter and the separation efficiency was up to 72% higher when using the MEEKC method. Detection limits from 5.4 to 26 μg/ml were obtained and the calibration plot was linear over more than one order of magnitude. The optimised method could be applied to the determination of Irganox 1330 and Irganox 1010 in polypropylene.