This study qualifies and quantifies the effects of pH, hardness, alkalinity, salinity and bone calcination temperature related with the adsorption of As(V) onto biogenic hydroxyapatite (HAPb) obtained from cow-charred bones. Arsenic contamination of surface and subsurface waters is widely extended in Argentina. It is a problem of major concern, particularly in rural and suburban areas where there are not water treatment plants for supplying of drinking water. HAPb is a natural material, whose absorbent properties can be used for the design of low-cost technologies for As(V) abatement in water. In this work HAPb has been characterized by physical and chemical analysis (XRD, SEM, EDAX, BET, and electrophoretic mobility). A Plackett–Burman screening experimental design allowed us to determine the main variables affecting the efficiency of As(V) sorption onto HAPb. Based on these variables and with a design of higher order we developed a model of the system to study its behaviour. Data collection was planned through a Doehlert experimental design and a back propagation artificial neural network was used to work it out. Results showed that salinity is the major variable affecting the efficiency of the As(V) immobilization process but pH and hardness should be taken into account because of associations among them.

The potential for colloid-facilitated migration in contaminated sites is well known, and remediation techniques such as ultrafiltration are often considered for contaminant removal. Although this approach could be successful, the stability of the contaminant species in groundwater and the removal efficiency need to be investigated to ensure proper decontamination of moving aquifers. In our study, we have sampled contaminated groundwater near a former radioactive liquid disposal area at Chalk River, Ontario, Canada. Samples were taken in 2002 and 2004 to determine the behaviour of radiocontaminants by size fractionation using ultrafiltration, with emphasis on ⁶⁰Co and ¹³⁷Cs. The contaminant concentrations varied significantly for both contaminants in the two samples (34.5 and 25.5 Bq/l for ⁶⁰Co, 25.5 and 97.2 Bq/l for ¹³⁷Cs). On the other hand, the size fractionation (5,000 Daa nominal cut-off) remained consistent between the 2002 and 2004 samples, as most of the ⁶⁰Co (72%-83%) remained in the filtrate, while almost all of the ¹³⁷Cs (>98%) was retained along with the colloidal-sized material. Release of ⁶⁰Co and ¹³⁷Cs from the colloidal material yielded desorption coefficients (K D₋des) of 7.8 x 10⁵ and 1.7 x 10⁸ ml/g for ⁶⁰Co and ¹³⁷Cs, respectively.