The purpose of this PhD thesis is to describe the processes that take place where iron-containing minerals are present. The work was especially focused on the ferro-ferric hydroxide minerals, the green rust series, and their possible role in the geochemical iron-cycle. The goal of the first study was to describe the iron-containing minerals in the granitic fractures at Äspö, Sweden. The structure, chemical composition and stable iron isotope ratio of the ferric oxides allowed reconstruction of the paleo-redox conditions in the granitic fractures. I found 3 types of formation environments: 1) course-grained hydrothermal hematites, 2) very fine-grained amorphous iron-oxides that had precipitated during drilling and 3) crystalline iron-oxides of intermediate size which we interpreted to form at low temperatures (< 10° C) at a depth of less than ~100 m below surface.

When pH is neutral to slightly basic Fe(II)-containing solutions oxidise, green rust precipitates. In the second study, X-ray diffraction allowed identification of green rust in groundwater from areas with medium to relatively high iron concentration. The method I developed is relatively simple and is based on the observation that green rust is stabilised, when it attaches to substrates such as glass and mica. The sampling was done from the site close to the airport of Bornholm and in the Äspö hard rock laboratory tunnel. The samples were transported to the laboratory under nitrogen where the X-ray diffraction showed that the material contained GR_CO3. The presence of green rust was verified by a drastic decrease in the intensity of the GR peaks when the samples had been exposed to air for a couple of days.

Knowledge about the chemical composition of minerals and their crystal structures is essential for use in databases and, because composition determines the behaviour of a solid. A literature review showed that previous reports of sulphate-containing green rust have presented structural and compositional parameters that have been erroneous or lacking. The third study was a detailed examination of sulphate-containing green rust by X-ray diffraction, Rietveld analysis and Mössbauer spectroscopy. Using newly determined data for chemical composition from pure samples, we were able to determine details in the structure and composition of the solid. The new results proved that sodium is essential for sulphate-containing green rust, GR_Na,SO4. The chemical composition is NaFe(II)₆Fe(III)₃(SO₄)₂(OH)₁₈12H₂O, its space group is P-3, and its cell parameters are: a = 9.528(6) Å, c = 10.968(8) Å and Z = 1.

The last chapter describes how GR_Na,SO4 reacts with other redox-sensitive elements. Selenium and neptunium are typical daughter products of spent radioactive fuel rods. These and other radioactive elements pose a serious environmental risk for long term storage, if leaks develop in the waste repository. To better understand the behaviour of these radioactive elements in the natural environment and their reactivity with the repository iron structures, I examined how selenite (SeO₃^2-) and neptunyl (NpO₂⁺) react with GR_Na,SO4. Selenite was reduced readily by reaction at the edge of the GR particles. It converted to trigonal elemental Se(0), while the GR transformed to goethite. When neptunyl reacted with GR, it was reduced to tetravalent Np, and the GR also oxidised to goethite. Following complete oxidation of the green rust suspension slurry, approximately 40% of the Np remained in the tetravalent redox state. The results indicate that Np(IV) is either incorporated in the goethite structure, or it exists as discrete Np(IV)-oxides. The presence of Np(IV) oxides is less likely, because neptunium dioxide are readily oxidise to pentavalent Np.

In this thesis, I have shown that Fe-containing minerals can be used as indicators of redox conditions and that green rust sulphate contains a monovalent cation in the interlayer, namely sodium. Thus, green rust is not solely an anionic exchanger, but can also exchange cations with the surrounding solutions, similar to what is observed for clays. In groundwater transport modelling, parameters describing the behaviour of green rust under aquifer conditions must be included because the compounds react readily with a number of toxic components. In some cases, they are immobilised whereas in others it is possible that they are mobilised by colloidal transport. From my research, it is now even clearer that models intended to simulate or predict groundwater transport of toxic elements should include thermodynamic and kinetic parameters for green rust and its reactivity with dissolved compounds.