Long term monitoring of dissolved inorganic nutrient concentrations such as silicate is the key to improve our understanding of biogeochemical cycles in the ocean. Electrochemical methods are proposed to measure in situ silicate concentration as they allow to develop miniaturized, reagentless sensor. We report here on the development and the optimization of the in situversion of our silicate sensor.

Silicate is a non-electroactive specie and its determination is performed by measuring its corresponding silicomolybdic complex formed in situ at acidic pH after the oxidation of a molybdenum electrode while the counter electrode is isolated behind a Nafion^® membrane [1].

The mechanical design of the sensor has been adapted from the laboratory prototype containing two connected cells: one for the complexation and one for the detection on a gold electrode (Æ=2mm) by cyclic voltammetry. All the electrochemical and chemical parameters, such as the Mo oxidation, the complexation time, the silicomolybdic complex detection as well as the complex transfer from one cell to the other and the cleaning protocol have been optimized to detect the smallest silicate concentration possible with the shortest time. Reproducible calibrations performed with this sensor provided linear regressions between 2.5 µmol L^-1 and 70 µmol L^-1 and a limit of detection of 1 µmol L^-1.

The 1st version of the sensor has been validated so far down to 50 meters using a pressure column and deployed in the Kiel fjord (Baltic Sea, Germany) down to 25 meters. The electronics (including datalogger) was also validated in situ as well as the solenoid pump placed into an equi-pressure compartment filled with dielectric oil. Some mechanical improvements have been done and the 2^ndversion of the silicate sensor is currently in fabrication. This new version will be deployed off Chile (South of Coquimbo) in April 2017 and results obtained will be discussed.

[1] Lacombe et al., Talanta 77 (2008) 744-750.