NEWS & RESOURCES

This study was small, but it is the research I am most proud of. After working on lithium geochemistry for several years, I realized that peer colleagues (including myself) had been taking something important for granted: we assumed we understood how lithium interacts with iron oxides, common minerals in laterite. In reality, almost no one had ever tested this idea properly.

For more than 20 years, studies had assumed that Li is adsorbed to iron oxides and even fractionates its isotopic signature when it does, despite the lack of direct experimental evidence. In our work, my colleagues and I decided to go back to basics and test this assumption in the laboratory.

Our experiments showed that, under normal natural conditions, Li is very unlikely to be taken up by Fe oxides. This only happens in extreme environments with pH at 12 and when the minerals are poorly crystallized. This finding overturns a long-standing idea that has influenced how scientists interpret Li signals in soils and rivers.

Along the way, we made an unexpected discovery: not only does the type of mineral matter, but so does how “mature” it is. Minerals that are newly formed behave very differently from older, well-crystallized ones. This has largely been ignored in geochemistry, where the focus has traditionally been on mineral type rather than mineral crystallinity.

This insight has wide implications. It helps explain why landscapes such as river floodplains can behave like hot spots for chemical weathering and challenges current models that treat the same mineral as if it always behaves the same way. Our results show that ignoring mineral crystallinity can lead to huge errors in predicting how chemical signals are recorded in the environment. In other words, current geochemical models often treat fluids as variables while assuming solids are constant. Our findings show that solids must also be treated as variables.