Abstract:
The endemic South African plant species, Aspalathus linearis (rooibos) and Cyclopia spp.
(honeybush) are consumed worldwide as popular herbal teas. In addition, recent trends in the
consumption of these products as ready-to-drink commodities and the use of extracts as ingredients in
food, beverage, cosmetic and nutraceutical products are partially ascribed to their potential healthpromoting
effects. Both rooibos and honeybush teas contain relatively high levels of rare antioxidative
phenolic compounds. The health properties of these plants are largely associated with their
content of phenolic compounds. The analysis of phenolics in these herbal teas is therefore important,
not only in support of fundamental research, but also for quality control and marketing purposes.
High performance liquid chromatography (HPLC) is the method of choice for the analysis of phenolic
compounds in natural products, due to the proven performance of the technique in terms of robustness
and sensitivity. However, HPLC also suffers from some drawbacks such as relatively high solvent
consumption, long analyses and challenges to obtain sufficient resolution for highly complex samples
such as these. Capillary electrophoresis (CE) is a potentially promising alternative technique for the
separation of phenolics. CE offers cheap routine analyses, high speed and high efficiencies and an
alternative separation mechanism compared to HPLC. The aim of this study was to evaluate CE as an
alternative to HPLC for the analysis of rooibos and honeybush tea phenolics.
Following extensive method optimisation, two capillary zone electrophoresis (CZE) methods which
provided efficient separation of the principal rooibos (15) and honeybush (8) tea phenolics, were
successfully developed for the first time. Experimental parameters evaluated and optimised include
the background electrolyte (BGE) composition and pH, capillary dimensions, analysis temperature,
applied voltage and injection volumes. A borate buffer was found to be essential due to the capability
of this BGE to form anionic complexes with phenolics containing vicinal diols, thereby affecting the
selectivity of the separation. The optimised BGE pHs for the rooibos and honeybush methods were
8.80 and 9.25, respectively, resulting in total analysis times (including conditioning of capillary) of 31
(honeybush) and 42 minutes (rooibos). Analysis times were slightly shorter than the routine HPLC
methods. The optimised CZE methods were characterised by alternative selectivity compared to
HPLC, and higher resolution of critical compound pairs was obtained. The methods were validated in
terms of repeatability of electrophoretic mobilities, linearity of calibration curves and sensitivity
(limits of detection (LODs) and limits of quantification (LOQs)). Sensitivity and repeatability of the
CZE methods were however not comparable with HPLC. Finally, the developed methods were
applied to the analysis of fermented and unfermented rooibos (Aspalathus linearis) and honeybush tea
(C. subternata and C. maculata samples). Quantitative data obtained for 10 samples of fermented and
unfermented rooibos and C. subternata and 9 of C. maculata each were compared to those obtained
by routine HPLC methods, and were found to be statistically comparable for the majority of
Stellenbosch University https://scholar.sun.ac.za
iii
compounds, with a few exceptions. The two quantitative CZE methods demonstrated their utility for
the routine quantitative analysis of phenolics in rooibos and honeybush teas, respectively, thereby
confirming the potential of CE as an alternative to HPLC for the routine analysis of these samples.
