Abstract : Arsenic is a toxic trace element occurring in natural waters in a variety of forms including soluble, particulate and organicbound,
but mainly as inorganic trivalent As(III) and pentavalent As(V) oxidation states. In many parts of the world,
groundwater is polluted with arsenic. This pollution can be caused by human activities (mining, pesticides...) but usually, the
main source of arsenic is geogenic. Epidemiological studies have demonstrated a significant increase in the risks of cancers
associated with high levels of arsenic in drinking water. Consequently, in the case of arsenic, the European standard level in
drinking water has been lowered to 10 µg/L and similar reductions in arsenic levels have been adopted elsewhere, including
the USA. The aim of this study was to develop new and reliable methods to analyse arsenic even at low concentrations, and
simple removal techniques, easy to handle and to apply to low-flow drinking production plants (flow rate < 10 m3/h).
As(III) and As(V) adsorption was studied, first on iron (oxy)hydroxides then on pillared clays : a montmorillonite modified
with iron, titanium and aluminium polycations. Adsorption was carried out under various experimental conditions. It appeared
that arsenic was better adsorbed on iron (oxy)hydroxide. Yet, iron pillared clay was the only media which could be
Speciation being a preponderant factor in adsorption, As(III) oxidation study is of great importance. Different common
reagents used for As(III) oxidation were studied: H2O2, NaOCl, FeCl3, KMnO4 and MnO2(s). In order to test their efficiencies
through As(V) determination, a colorimetric method was developed, based on phosphate measurement. The results showed
that the oxidants which could easily be applied to low-flow drinking production plants were FeCl3 and KMnO4. Thereafter, a
polystyrene resin loaded with manganese oxide was synthesised. This solid presents simultaneous oxidation and adsorption
behaviour, and its adsorption capacities towards As(III) and As(V) were above many studied adsorbents.
The last part of this study allowed us to work closer to natural conditions, through the preparation of an artificial water of
granitic type, such are waters usually polluted with arsenic. The compilation of major ions common concentrations led to the
preparation of a model water, spiked with As(III) or As(V), and used to validate our analysis methods and removal
mechanisms. The results proved that the major ions did not have any influence on these processes, showing their possible use
in a low-flow drinking production plant.