Further characterisation and applications of the diffusive gradients in thin films technique : In situ measurements of anions and cations in environmental waters

As both the toxicity and the mobility of trace elements are related to chemical forms present, robust methods for element speciation analysis are of great interest. During the last 15 years, hundreds of scientific articles have been published on the development and applications of the diffusive gradients in thin films (DGT) passive sampling technique. The aim of this thesis was to explore new application areas as well as carry out further characterisation of DGT-adsorbents already on the market. The commercially available DGT containing ferrihydrite adsorbent, currently in use for the determination of phosphate and inorganic arsenic, was characterised with respect to anionic arsenate, molybdate, antimonate, vanadate and tungstate determination. Tests were performed in the laboratory as well as in the field. Diffusion coefficients were determined for the anions using two different methods with good agreement. Simultaneous measurements of arsenate were conducted as quality control to facilitate comparison of the performance with previous work. The ferrihydrite-backed DGT was concluded useful for application over the pH-range 4 to 10 for vanadate and tungstate, and 4 to <8 for molybdate and antimonate. At pH values ≥8, deteriorating adsorption was observed. Further investigations of the ferrihydrite-DGT device were done with respect to organic arsenic species. From previous research it is understood that the two most prevalent forms of organic arsenic in natural waters, monomethylarsinate (MMA) and dimethylarsonate (DMA), adsorb to ferrihydrite. It was concluded that MMA and under some conditions DMA are accumulated and might therefore be included in total arsenic measurements. A method for speciation of inorganic As, DMA and MMA was described. DGT sampling was applied at three stations, with different salinities, in the brackish Baltic Sea. Time series as well as vertical profiles were taken and complementary membrane- (<0.22 μm) and ultrafiltrations (<1 kDa) were conducted on discrete samples collected at 5 m depth. A combination of a restricted pore (RP) version of DGT and the normal open pore (OP) DGT, both loaded with Chelex cation exchanger, was used for speciation of copper and nickel. Due to minimal differences in results between the OP- and RP-DGTs it was suggested that the complexes were smaller than the pore size of the RP gel (~1 nm) resulting in both DGTs accumulating essentially the same fraction. Furthermore, there seemed to be a trend in copper speciation indicating a higher degree of strong complexation with increasing salinity. The low salinity stations are more impacted by fluvial inputs which will likely affect the nature and composition of the organic ligands present. Assuming that copper forms more stable complexes with ligands of marine rather than terrestrial origin would be sufficient to explain the observed trend. Additionally, uranium results from the same sampling tours were used to evaluate OP-DGT for in situ uranium measurements. Previous research has focused on laboratory studies for characterisation of a range of suitable uranium adsorbents, including Chelex. From the Baltic Sea study, a strong correlation between DGT-labile uranium and pH was revealed. The correlation could not be associated to organic complexation, since the ultrafiltration results implicated that uranium was truly dissolved. Instead it must be attributed to the formation of stable inorganic anionic uranyl-carbonate complexes, the stability of which increases with increasing numbers of carbonates and pH. Transplanted aquatic moss has commonly been used to monitor bioavailable trace metal contaminations in freshwater. Like the DGT technique the method has the advantage of generating time-weighted averaged concentrations. The DGT technique has in several previous studies been suggested to mimic biological uptake of trace metals. Four speciation/fractionation techniques – DGT, transplanted aquatic moss, ultrafiltration (1 kDa) and membrane filtration (0.22 μm) – were used in parallel for measurements of Al, Cd, Co, Cu, Fe, Mn, Ni and Zn in a contaminated freshwater stream in northern Sweden. Differences and similarities between the methods were investigated and how these depend on geochemical water quality. Strong correlations between DGT-results and the concentrations in the filtrate (