A low temperature experimental alteration of a rhyolitic obsidian

Experimental alteration of obsidian by distilled water in an open system (T = 82 +/- 5 degrees C; pH = 5.5 - 6) was carried out to investigate on the mechanism of leaching and on the mobilization of major elements as well as some trace elements, including rare earths. Nine samples were altered at di fferent times (up to 1717 hours) and products of leaching were studied by S EM, HRTEM, and XPS,and chemically analysed by XRF and ICP-MS. The most evident chemical change occurring in the rhyolitic obsidian after the experimental leaching is the loss of alkalis and the gain of H2O. Stati stical analysis revealed that inward diffusion of molecular water into the bulk glass takes place during the early stage of alteration, followed by ca tion(+)-(HO+)-O-3 or cation(+)-H+ interdiffusion and changes within the gla ss structure as well as disruption of the silica network at the glass-solut ion interface. The structural reorganization of the glass is accompanied by the formation of flake-leaf-needle morphologies as evidenced and observed by HRTEM. These morphologies might represent three steps of local crystalli zation within the glass, which culminates in the formation of an acicular m ineral, probably halloysite in the interior of the glass. The specific weight-loss leach rate of the rhyolitic obsidian under study i s very slow, i.e. 1.5 +/- 0.51 x 10(-8) (g)glass m(-2) s(-1). Leach rates, estimable only for some of the detected elements show the following order a s regards mobility: Na>Si approximate to K>Ca>Ti approximate to Al for majo r elements, and Pb>Lu>Er>Zn>La>Rb>Ce for trace elements. Values calculated for trace elements are higher than for major elements. There is no simple e xplanation at hand. Finally, XPS analysis clearly showed that the formation of a leached layer did not take place. However, the very small variation of Na/Al, Mg/Al, and Ca/Al ratios with leaching time might be indicative of the existence of an ultrathin layer, even lower than 30 Angstrom. Discrepancies between the exp erimental findings reported in literature are explainable by structural dif ferences of the obsidians.