The determination of copper in sea water using flow injection with chemiluminescence detection
This thesis describes the design, optimisation and shipboard deployment of a flow injection - chemiluminescence (FI-CL) technique for the determination of labile Cu(II) and total copper (by UV irradiation) in seawater. The operational parameters of the FI manifold in a UHP water sample matrix and the 1,10-phenanthroline CL reaction were rigorously optimised. Interferences to the CL reaction were investigated and the good analytical figures of merit obtained presented. The FI-CL method was modified for the determination of ultra trace levels of Cu(II) in seawater by the incorporation of a new design of micro-column containing 8-hydroxyquinoline (8HQ) resin for in-line matrix separation and preconcentration. Reagent clean-up techniques, blank procedures and a standard addition protocol are detailed. The optimised method is selective for Cu(II) in the linear range 0.1 - 50 nM, with precision of <4% (n=4) for a typical seawater analysis, and a limit of detection (3s) of 25 pM for a loading time of 90 s. The FI-CL analyser was fully automated and then validated by field deployment on the Tamar Estuary, during which its robustness, reliability and stand alone capability were demonstrated. Good accuracy was achieved for a seawater CRM analysed onboard. The near real time Cu data obtained was in good agreement with a comparative voltammetric method. The FI-CL method was further validated by field deployment on the Atlantic Meridional Transect (AMT 9) during which Cu(II) (filtered, acidified (pH 2) HNO3) in the surface waters (<250 m) of the North and South Atlantic (50°N to 50 °S) was mapped. Spatial variation in Cu(II) concentrations was observed (<0.7 to 6.1 nM) through the contrasting biogeochemical provinces encountered that representated coastal, upwelling and oligotrophic regions. Copper (II) enrichments were imposed on a trend of decreasing Cu(Il) concentrations away from European coastal waters (>2.5 nM) to open ocean gyres (< 1 nM). Away from strong input mechanisms, upper water column Cu(II) concentrations were ca. 1.5 nM, being dominated by long range aerosol input mechanisms. Input sources are fingerprinted via correlation with nutrients and hydrographic data, whilst the dominant sinks are active biological uptake and particle reactivity. Cu(II) vertical distributions through the upper mixed layer display strong relationships with chlorophyll α particularly in remote oceanic regimes. An in-line UA photo-oxidation system was constructed and optimised for the digestion of organically complexed Cu to enable near real time determination of total Cu in seawater by FI-CL. It achieved very efficient digestion of DOM (96.3 %) using a short irradiation time (600 s), with good recovery of Cu. Robustness, reproducibilty of irradiation, effective operational life and safety were considerably improved compared to existing systems. DOM rich Tamar Estuary and Celtic Sea samples were in good agreement with Cu(II) results from conventional batch UV digestion and voltammetric detection.