Stability of arsenic peptides in plant extracts: off-line versus on-line parallel elemental and molecular mass spectrometric detection for liquid chromatographic separation.

The instability of metal and metalloid complexes during analytical processes has always been an issue of an uncertainty regarding their speciation in plant extracts. Two different speciation protocols were compared regarding the analysis of arsenic phytochelatin (As(III)PC) complexes in fresh plant material. As the final step for separation/detection both methods used RP-HPLC simultaneously coupled to ICP-MS and ES-MS. However, one method was the often used off-line approach using two-dimensional separation, i.e. a pre-cleaning step using size-exclusion chromatography with subsequent fraction collection and freeze-drying prior to the analysis using RP-HPLC-ICP-MS and/or ES-MS. This approach revealed that less than 2% of the total arsenic was bound to peptides such as phytochelatins in the root extract of an arsenate exposed Thunbergia alata, whereas the direct on-line method showed that 83% of arsenic was bound to peptides, mainly as As(III)PC(3) and (GS)As(III)PC(2). Key analytical factors were identified which destabilise the As(III)PCs. The low pH of the mobile phase (0.1% formic acid) using RP-HPLC-ICP-MS/ES-MS stabilises the arsenic peptide complexes in the plant extract as well as the free peptide concentration, as shown by the kinetic disintegration study of the model compound As(III)(GS)(3) at pH 2.2 and 3.8. But only short half-lives of only a few hours were determined for the arsenic glutathione complex. Although As(III)PC(3) showed a ten times higher half-life (23 h) in a plant extract, the pre-cleaning step with subsequent fractionation in a mobile phase of pH 5.6 contributes to the destabilisation of the arsenic peptides in the off-line method. Furthermore, it was found that during a freeze-drying process more than 90% of an As(III)PC(3) complex and smaller free peptides such as PC(2) and PC(3) can be lost. Although the two-dimensional off-line method has been used successfully for other metal complexes, it is concluded here that the fractionation and the subsequent freeze-drying were responsible for the loss of arsenic phytochelatin complexes during the analysis. Hence, the on-line HPLC-ICP-MS/ES-MS is the preferred method for such unstable peptide complexes. Since freeze-drying has been found to be undesirable for sample storage other methods for sample handling needed to be investigated. Hence, the storage of the fresh plant at low temperature was tested. We can report for the first time a storage method which successfully conserves the integrity of the labile arsenic phytochelatin complexes: quantitative recovery of As(III)PC(3) in a formic acid extract of a Thunbergia alata exposed for 24 h to 1 mg As(v) L(-1) was found when the fresh plant was stored for 21 days at 193 K.