Download files
Abstract
The toxicity of arsenic depends on its chemical form. Organo-arsenic compounds are acknowledged to be less toxic than inorganic arsenic compounds. Therefore, measurement of the concentration of arsenic as element (“total arsenic”) alone is insufficient because organo-arsenicals comprise an important fraction of many environmental samples.
In this thesis several milestones on the path towards the development of a continuous and miniaturised arsenic speciation system based on surface enhanced Raman spectroscopy was achieved.
The well known oxidation reduction cycle (ORC) technique for fabricating polycrystalline gold foils as effective surface enhanced Raman spectroscopy (SERS) substrates was systematically optimised using a central composite experimental design (CCD). A novel one step electrochemical method, which has been developed for fabrication of nano-porous gold films (NPGF), was demonstrated and optimised as an effective alternative to ORC.
Direct-write electron beam lithography and metal-lift-off nanofabrication techniques were also used to create gold nano-pillar systems which serve as SERS substrates. The CCD optimisation model, built from the dimensions of the pillars and the distance between them predicts that higher Raman signal enhancements can be obtained using arrays with 185 nm spacing and 214 nm x 214 nm dimensions.
A Raman spectroscopic study of twelve organo-arsenics was reported. The prediction of the normal mode vibrational frequencies and assignments were based on calculations done at the HF, DFT/B3LYP levels of theory using 6-311++G(3df,3pd) basis set.
The first multivariate detection of the concentration of arsenic species using SERS was demonstrated. The adsorption of arsenic species on the surface has been aided by selecting the arsenic species according to their ionic charges. The species were partitioned into monolayers of charged alkanethiols. Multivariate regression of pure arsenic species over the concentration range of 0 – 5 mM was shown using a PLS1 algorithm. Simultaneous determination of arsenate and arsenobetaine has been demonstrated using a two sensor system. PLS2 was used for calibration of 16 samples containing all the possible combination of the chosen concentration range. The PLS2 model was applied to adequate amount of independent test sets and the actual concentration of the test sets agreed well with the predicted concentration.