Page 355 - Multidimensional Chromatography
P. 355

346                                     Multidimensional Chromatography

                           therefore a suitable technique for efficient on-line clean-up procedures. Some exam-
                           ples of various applications are shown in Table 13.1.
                             Some of the different methods which demonstrate the suitability of this technique
                           for determining single analytes will now be described in greater detail.
                             Ethylenethiourea (ETU) is a highly polar metabolite of the ethylenebisdithiocar-
                           bamate (EBDC) fungicides. This is a relatively stable degradation product which is
                           also present in EBDC formulations (at concentrations 0.02–5%) (48).  ETU is
                           mainly determined by GC using a precolumn derivatization, which is time consum-
                           ing, or by LC using UV or electrochemical detection with a clean-up step. Both tech-
                           niques are very laborious and  are also not sensitive enough. LC–LC is therefore an
                           attractive alternative to these  methods and has been used to determine ETU in aque-
                           ous samples (48). The experimental conditions are given in Table 13.1. The authors,
                           after studying different packing materials of the alkylbonded columns, decided to
                           work with a 5  m Hypersyl ODS for better separation of the interfering compounds
                           and peak compression. Ammonium acetate (pH 7.5) was also added to the mobile
                           phase since this gave a better peak profile than with pure water at pH 3; the flow rates
                                                          1
                           were 1.0 (M-1)  and 1.1 (M-2) ml min . Experiments showed that the injection vol-
                           umes on C-1 should not exceed 200 l and these were eluted with 2.6 ml of mobile
                           phase. C-1 was switched on-line with C-2, and the ETU-containing fraction trans-
                           ferred to C-2 by using 0.44 ml of mobile phase (40 s). The retention time of ETU on
                           C-2 was about 3 min, while the total analysis took less than 10 min.
                             From Figure 13.6, which compares the direct LC analysis and the LC–LC analy-
                           sis, we can see that there is a significant increase in selectivity.
                             With this method, levels of 0.1  g l  1  can be detected in ground water and, if an off-
                           line liquid–liquid extraction step is added, levels of 0.1 g l  1  can be detected (48).
                             Another example is the determination of bentazone in aqueous samples.
                           Bentazone is a common medium-polar pesticide, and is an acidic compound which
                           co-elutes with humic and/or fulvic acids. In this application, two additional bound-
                           ary conditions are important. First, the pH of the M-1 mobile phase should be as
                           low as possible for processing large sample volumes, with a pH of 2.3 being about
                           the best that one can achieve when working with alkyl-modified silicas. Secondly,
                           modifier gradients should be avoided in order to prevent interferences caused by
                           the continuous release of humic and/or fulvic acids from the column during the
                           gradient (46).
                             With bentazone, small changes in the composition of the mobile phase have a dra-
                           matic effect on the final results (see Figure 13.7).
                             The methanol gradient from 50 to 60% releases quite a lot of interfering compo-
                           nents. Omitting the step gradient does not provide enough selectivity and so the best
                           conditions were obtained with a pH gradient. The experimental conditions are shown
                           in Table 13.1.
                             This method can quantify levels of 0.1  g l  1  in real samples and reproducibil-
                           ity values are good; the total analysis time was 8 min. Figure 13.8 compares the
                           chromatogram obtained by using this method with one obtained without column
                           switching.
   350   351   352   353   354   355   356   357   358   359   360