This review article summarizes our research related to the initial stages of phase transformations in minerals. We have shown that these can be identified through the formation of chemically induced planar defects, such as twin, antiphase, or inversion boundaries in the affected crystals. Based on the principles of crystal chemistry, early researchers suggested that the atomic structure of a twin boundary might be related to the existing polymorphic modification of that phase. The confirmation of this hypothesis was not possible until the development of modern electron microscopy techniques that enabled a direct insight into the atomic structure of the twin boundaries. Our studies of twins and other translational defects in many natural minerals have indicated that their formation is chemically induced by the presence of dopants that stabilize the particular polytypic structure. The inherent anisotropy imposed by the chemically induced transformation fault (e.g. growth twin) triggers exaggerated growth of the crystal parallel to the fault plane as long as geochemical and thermodynamic conditions favour the formation of such a faulted stacking. These initial growth stages dictated by the growth of chemically induced fault are clearly reflected in a final morphology of the crystals and can be recognized by the characteristic twin-plane reentrant angles and additional symmetry elements that can be observed on the crystals. Because of commonly complex geochemical conditions, natural minerals incorporate an assortment of foreign elements that are present during crystal growth, but only one of these elements, however, is responsible for the formation of the faulted stacking (e.g. twinning). In order to identify the elements that trigger twinning in minerals we developed, in cooperation with University of Bonn and University of Oxford, novel analytical approaches enabling atomic-scale determination of the interfacial crystal chemistry. Our studies of twinning in various minerals, presented in this review article, deal with one of the fundamental scientific challenges that offers the possibilities for true understanding of the basic building principles of solids and the fundamentals of phase transformations in minerals and their successful reproduction under the controlled laboratory conditions. (111) twins in spinel (MgAl2O4) are the first case, where we have undoubtedly proven the principle of chemically induced twinning. The formation of these twins was previously explained in the scientific literature as accidental attachment of two crystallites in twinned orientation in the nucleation state. Based on our atomic-scale studies of (111) twins in natural spinels from Mogok locality we have shown that, on the contrary, twinning is related to the presence of a small amount of beryllium in crystal nucleation stage. This hypothesis was successfully verified with the synthesis of spinel twins in the presence of a small amount of BeO in flux. While the formation of twins was not observed in the pure MgO-Al2O3 samples, already small additions of BeO have a dramatic influence on the growth of MgAl2O4 crystals. The addition of BeO triggers the formation of numerous simple and complex spinel twins as a result of chemically induced twinning in MgAl2O4. The presence of Be in the nucleation state triggers the formation of hcp stacking on the surface of the octahedral MgAl2O4 crystals. We have further shown that the formation of various modulated structures in the MgAl2O4-BeAl2O4 system also known as the taaffeite-type compounds, where the ccp and hcp sequences are interchanging is possible according to this principle. › Daneu N, Rečnik A, Yamazaki T, Dolenec T. Structure and chemistry of (111) twin boundaries in MgAl2O4 spinel crystals from Mogok (Burma). Phys. Chem. Minerals 34 (2007) 233-247. [615262] › Drev S, Rečnik A, Daneu N. Twinning and epitaxial growth of taaffeite-type modulated structures in BeO-doped MgAl2O4. CrystEngComm 15 (2013)
COBISS.SI-ID: 25754407
In this work we described the synthesis of {111} twins and taaffeites in BeO-doped spinel (MgAl2O4). The results are the first proof of our hypothesis that twinning in spinel is chemically induced and not a consequence of accidental attachment of crystals in the nucleation stage. This hypothesis was based on our previous analyses of natural {111} twins from Mogok (Myanmar), where we indirectly showed that beryllium is the most likely dopant responsible for twinning in MgAl2O4, however the presence of beryllium at the twin boundary was not proven: › Daneu N, Rečnik A, Yamazaki T, Dolenec T. Structure and chemistry of (111) twin boundaries in MgAl2O4 spinel crystals from Mogok (Burma). Phys. Chem. Minerals 34 (2007) 233-247. [615262] In order to prove our work hypothesis, we decided to grow spinel twins in flux or in the presence of liquid phase (PbF2) from initial oxides (MgO, Al2O3, BeO). While no twins were observed in the binary MgO-Al2O3 system, twinning was abundant in the samples with BeO addition. The presence of Be in the initial stage of crystal growth leads to the formation of hcp stacking on the surface of octahedral MgAl2O4 crystals (ccp) leading to twin formation. The growth and development of composite twinned crystals is described in detail in the paper. In the first stage, i.e. until beryllium is available for the formation of the hcp stacking, such grain grows very fast (exaggeratedly) in the direction of the twin boundary. This leads to the formation of plate-like composite grains. When the twin-forming dopant is no longer present, the growth in this direction is stopped and the crystal starts to thicken according to the Ostwald ripening law. Electron microscopy analyses have shown that the samples contain simple twins, complex twins and complex modulated taaffeite BexMgyAl2(x+y)O4(x+y) phases, where the ccp and hcp sequences are interchanging. The taaffeite phases may form separate grains or epitaxial layers with the spinel. XRD analyses have revealed also that the spinel unit cell is significantly smaller in syntheses in the presence of BeO indicating that beryllium might be incorporated in the spinel crystal structure in the form of soil solubility. This was not confirmed yet however, if the assumption is correct, then this will be the first system, where we observe that the same dopant can form tropochemical defects and at the same time also solid solution with the main phase. The results are described in the following scientific paper: › Drev S, Rečnik A, Daneu N. Twinning and epitaxial growth of taaffeite-type modulated structures in BeO-doped MgAl2O4. CrystEngComm 15 (2013) 2640-2647. [26530343] One of the most important conclusions of this research work is the fact that BeO doping can be used to influence microstructure development in MgAl2O4 ceramics by influencing the grain growth rate and especially their shape because otherwise cubic (isometric) grains start to grow as platelets. The described principle can be applied to other materials with the cubic spinel structure, where anisotropic grain growth (twinning) could be triggered by suitable doping and enables processing of textured microstructures for special applications (thermoelectric, platelike magnetic particles, materials for batteries,...).
COBISS.SI-ID: 26530343
Our preliminary investigations of twins and related 2D structures have revealed that these defects represent a preparatory stage of phase transformations in minerals. In effect, this led to the problem of epitaxial and topotaxial overgrowths in minerals. Such complex situations may occur when a primary mineral is exposed to a specific geochemical environment where it recrystallizes into a new mineral or mineral assembly in a specific coherent (epitaxial) orientation relationship. Our first study of complex topotaxial recrystallisation in minerals were twins in rutile from different localities. In nature, rutile can be found in the form of morphologically well-defined contact twins on {101} and {301} planes. These twins may form complex sagenite crystal networks or cyclic twins. In our preliminary study on {301} twins from Diamantina in Brazil (Daneu et al, 2007) we found that these twin boundaries regularly contain a coherent ilmenite layer with varying up to a few nanometer thickness. The ilmenite layer is heavily twinned on basal planes and contains nanosized twin domains. The nanostructural features of the ilmenite layere have indicated that the crystals formed oriented topotaxial recrystallization (oxidation) of an oxyhydroxide (Fe, Ti, Al)-tivanite precursor. These results were the basis for our next study in the frame of this project and where we analyzed another type of twins from the same locality, the {101} twins, described in the paper: › Daneu N, Rečnik A, Mader W. Atomic structure and formation mechanism of (101) rutile twins from Diamantina (Brazil). American Mineralogist 99 (2014) 612-624. [27599399] The {101} twin boundaries in rutiles from the same locality have a different composition and contain coherent nanosized corundum inclusions. Based on the characteristics of the rutile/corundum contact and the density of misfit dislocations at the interface we found that they formed by a similar principle as the (301) twins, by dehydration form a hydroxide precursor mineral, in this case diaspore (Al-hydroxide). In the concluding part of the paper we describe the complex mechanism of sagenite rutile twin formation - combined formation of both twin types. Based on the results of our analyses we suggest that rutile twins form as a result of Al-rich Fe-Ti oxyhydroxide (hydroxylian pseudorutile, HPR) dehydration by the diffusion of Ti4+ ions through the matrix crystal lattice. Rutile starts to crystallize epitaxially on the surface of the precursor in different structurally dictated directions. Domains with different orientation form at different nucleation sites. During dehydration of the HPR precursor, the excess Fe and Al atoms that cannot be incorporated into the rutile structure, exsolve as oxyhydroxides at the boundaries between the forming rutile domains. Each phase is exsolved at the structurally best fitting interface; the (Fe,Ti)-oxyhydroxide at the {301} twin boundaries and the Al-hydroxide at the {101} twin boundaries. In the last recrystallisation stage, these two phases oxidize to ilmenite and corundum. In the process of structural and chemical characterization of the {301} twin boundaries that contain coherent lamellae of ilmenite, we used our special technique for the determination of their chemical composition, known as the Concentric Electron Probe or the CEP technique. The method was developed in the cooperation with dr. Thomas Walther (now Sheffield, UK). It enables the determination of the chemical composition of planar defects and other interfaces of subnanometer dimensions with high precision and accuracy. It is generally applicable and gibes at least two orders of magnitude better results in comparison with other analytical techniques of transmission electron microscopy. Using the CEP technique, we were able to determine the chemical composition of the interface lamellae at the {301} twin boundaries in rutile as ilmenite (FeTiO3) instead of hematite (Fe2O3). The phases have iron in different o
COBISS.SI-ID: 27599399
The results described in this achievement were obtained in cooperation with our Hungarian partners from the Pannonian University in Veszprém. We have studied growth defects and epitaxial layers in nanocrystalline magnetite (Fe3O4) and its oxidation product, maghemite (gamma-Fe2O3). Two types of planar defects are identified in magnetite, (111) spinel-law twin boundaries and (110) stacking faults (SF). We have found that in contrast to our theory on chemically induced twinning, twinning in magnetite it is related to a simple magnetic-field-assisted self-assembly and the growth of octahedral nanocrystals throughout their crystallization period. Simple contact twins of crystals sharing common octahedral faces, or even platelike twins develop when two crystals are joined in twinned orientation in the beginning of grain growth and continue their growth as a unit. Crystallographically, twinned domains are related by 180° rotation about the [111]-axis and with the (111) plane as the interface, producing local hcp stacking in the oxygen sub-lattice. SFs are present in both single and twinned magnetite crystals, where they are pinned to (111) twin boundaries and are present only in one domain. The displacement vector corresponding to the observed translation was determined. After thermal treatment at 250°C, both types of planar defects are retained. In addition to both types of planar defects, originating from magnetite, we identified a new formation of few nanometers- thick epitaxial layers, of a hexagonal Fe(III)-oxide–hydroxide, feroxyhyte (delta-FeOOH), covering the octahedral faces of the maghemite crystals. In the paper we set the atomic structural models of twins and stacking faults in magnetite and also of the magnetite- feroxyhyte epitaxial contact. › Nyirõ-Kósa I, Rečnik A, Pósfai M: Novel methods for the synthesis of magnetite nanoparticles with special morphologies and textured assemblages. Journal of nanoparticle research 14 (2012) 1150-1-1150-10. [26132775] After heat-treatment at 250 °C, the magnetite crystals are topotaxially transformed into a fully oxidized product - maghemite (gama-Fe2O3) while retaining the primary morphology. Many vacancy-ordered variants of maghemite were identified, the most common of which had a tetragonal unit cell with dimensions 1x1x3 of that of magnetite. During the experimental TEM work we noticed that the vacancies become disordered when the crystals are exposed to the electron beam for prolonged time. We found that the direction of magnetocrystalline anisotropy does not change during the structural transition. Growth defects that formed already during the growth of magnetite are {111} twin boundaries and {110} stacking faults are also preserved. The twin boundaries are produced by 180° rotation about ‹111› axes, leading to a local hexagonal stacking within the otherwise cubic spinel structure of magnetite, while stacking faults (SFs) are produced by a ¼∙‹110› translation, which reflects the cation position across the {110} interface and the oxygen sublattice remains unchanged. Based on our observation of many crystal clusters, we conclude with confidence that magnetite twins form by magnetic-field-assisted self-orientation and attachment in either identical or twin orientations in which the crystals share a common octahedral face. After being attached, the twins continue to grow and develop characteristic plate-like morphologies. Stacking faults, present in single as well as twinned crystals are a consequence of defective stacking in the cation sublattice. Twin boundaries as well as stacking faults are of growth origin and are present already in magnetite crystals, and their structures remain unchanged after oxidation of magnetite to maghemite. In addition to inherited planar defects, maghemite displays another peculiarity that is not observed in magnetite. We have shown that after conversion of magnetite into maghemite a few atomic layers of an iron(III)-oxide-hydroxide, ferroxyhyte (
COBISS.SI-ID: 26941991
Despite many studies on the formation FeS and phase transformations in Fe-sulphides, the structure and physical properties of first precipitates are still not well understood. In the article we studied the influence of Cu-addition on the first precipitate, a near-amorphous, mackinawite-type FeS and subsequent phase transformations in the Fe-S system. To study the reaction products, iron and copper chlorides were mixed in different ratios with sulfur and diethanolamine (DEA), and sonicated to achieve efficient mixing. Already at this stage, the first FeS precipitate is formed. TEM investigations of Cu-doped mackinawite-like FeS showed enhanced crystallinity accompanied with expansion of the unit cell along the c-axis, proportional to the amount of Cu adsorbed between the (001) layers of the mackinawite structure. The subsequent solvothermal treatment and sulfurization of pure FeS resulted in formation of pyrite, at low doping Cu-rich mackinawite and cubic (Fe,Cu)S with a sphalerite-type structure were formed, while at higher Cu concentrations the end-products were chalcopyrite and bornite, corresponding to the amount of added Cu. Investigations belong to the field of phase transformations in iron sulphides: › Zavašnik J, Stanković N, Arshad MS, Rečnik A. Sonochemical synthesis of mackinawite and the role of Cu addition on phase transformations in the FeS system. Journal of nanoparticle research 16/2:2223 (2014) 13 pages. [27373607] To understand the structural transitions in iron sulfide during aging we synthesized and characterized samples that were prepared at room temperature and aged for various periods of time in the supernatant. The goal of this work was to characterize the first precipitate of iron sulfide. It was found to consist of a few atomic layers thick curved sheets with mackinawite-like (FeS) structure. We have observed that during aging for several months the primary sheet-like sulfide transforms into the spinel-type greigite (Fe3S4) structure. Interestingly, during this phase transition the grain size decreased while the crystallinity increased. To investigate the effect of transition metal incorporation Cu, Zn, and Ni were added to the FeS precipitate and studied whether these metals remain in the solid state during the mackinawite-greigite transition. Cu and Zn were found to form their independent mineral phases and did not enter the mackinawite structure. Ni, however, distributed uniformly within the mackinawite crystals, indices the likely substitution of Fe by Ni. The scientific significance of our results is that a better understanding can be obtained about nanocrystalline iron monosulfides that form in large quantities in the anoxic layers of ocean and some freshwater sediments, as well as in soils and can potentially immobilize several transition metals in their structures: › Csákberényi-Malasics D, Rodriguez-Blanco JD, Kovács Kis V, Rečnik A, Benning LG, Pósfai M. Structural properties and transformations of precipitated FeS. Chemical geology 294-295 (2012) 249-258. [25570343] The formation sequence of Fe-sulphides was also studied by the chemical vapor transport (CVT) method at 600°C in an evacuated quartz tube. Fe-sulphides were synthesized using Fe- and Cu-halides and elementary S as the reaction precursors. Depending on the mobility of elements different Fe-sulphides crystallized through gradually decreasing temperature zones. At the higher temperature zones, where metal ions are present in abundance, the main reaction product is FeS in form of pyrrhotite-3 T with {111} layers cubic FeS coherently intergrown with the {0001} layers of the hosting pyrrhotite. HRTEM analysis has shown that cubic FeS has a high density of twins and stacking faults in the {111} planes. Pyrite crystallized in the temperature zone between 500 and 450°C, where the concentration of metal ions is depleted. With a decreasing temperature its morphology changes from {111} ) {210} ) {100}. Doping with Cu did not result in t
COBISS.SI-ID: 27373607