The contribution of chaotic dynamics on the genesis of Chapecó silicic rocks from Paraná-Etendeka Magmatic Province: an experimental approach

• Ano de defesa: 2023
• Autor(a) principal: Vicentini, Caio Morelli
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Chaotic Dynamics at High-Temperature; Chapecó-type Silicic Rocks; Contaminação Mútua em Sistemas Máfico-Félsico; Dinâmica Caótica a Altas Temperaturas; Experimental Petrology at High-Temperature; Mafic-Felsic Mutual Contamination; Magma Mixing; Mistura de Magmas; Paraná-Etendeka Magmatic Province; Petrologia Experimental em Altas Temperaturas; Província Magmática Paraná-Etendeka; Rochas Silicáticas do tipo Chapecó
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/14/14132/tde-17112023-151156/

Mixing dynamics is thought to decisively influence volcanism on Earth. Magma mixing is pointed to occur under chaotic dynamics in nature based on several field observations of fractal patterns (i.e., scale invariant) emerging on geological sites. Using a chaotic mixing approach, this work is the first attempt to experimentally study the mutual contamination between mafic and felsic phases using natural samples from the Paraná-Etendeka Magmatic Province (PEMP) as end-members. Our final aim is to unravel the origin of the high-Ti Chapecó dacites from PEMP, distinguishing between Chapecó-Guarapuava (CGD) and Chapecó-Ourinhos dacites (COD). A campaign of four chaotic mixing experiments was performed using an apparatus that operates at high temperatures (i.e., Texp > 1,000 °C) under 1 atm pressure to mix silicate melts with stark contrasting rheologies (i.e., viscosities) by means of controlled chaotic dynamics. Experiments 1, 2 and 3 consist in the mixing of 80% of an identical basaltic glass (high-Ti Pitanga type from PEMP) with 20% of rhyolitic glasses that vary according to the experiment (rocks from the Paraná Basin basement) at a constant temperature Texp = 1,350 °C. Experiment 4 is the mixing of 80% of a dacitic glass (high-Ti CGD from PEMP) with 20% of a rhyolitic glass (the identical used in Exp2) at Texp = 1,500 °C. After experiments, morphological aspects of the experimental products include stretched and folded filaments of alternating phases, confirming that chaotic dynamics was achieved. In the experiments 1, 2 and 3, orbicular structures containing dendritic crystals and remnant portions of glass in the basaltic area were described for the first time. The observed dendritic areas of the basaltic regions point towards an early crystallization process during the initial quenching, and that the crystallization process occurs heterogeneously. In Exp4, sections develop fractal structures that are expected theoretically (i.e., Poincaré sections). The fractal dimension Dbox = 1.60(3) calculated for a representative section is similar to literature data. Chemical transects along contact zones between the interconnected end-members confirm chemical exchanges (or contamination) between the melts, which occur by diffusion. The highest contamination degrees in the mafic phases are observed for SiO2, K2O, Cs, Rb, U, Th and Pb (trace elements were not determined in Exp4). As a consequence, these elements depict strong non-linear curves when plotted against other elements. Regarding trace analyses, the group of elements V, Sc, Sr, Cs, Rb, U and Th (G1) show normal diffusion profiles, whilst the elements Ga, Nb, Zr, Y, Ba, REE, Ta, Hf and Pb (G2) show uphill diffusion. It causes anomalous concentrations along chemical transects and a more marked non-linear behaviour in inter-elemental plots associated to G2. For all experiments, the elemental mobility along the contact areas was quantified calculating the normalized variance (2n), which consistently indicates the connection between viscosity and mobility for major/minor (all experiments) and for G1 elements (experiments 1, 2 and 3). Moreover, it was identified that the most mobile elements present low (< 1.0) or high field strength Z/r2 (> 4.0). In the case of G2 elements, the most expressive values of (2n) are associated to the smallest initial gradients (e.g., Ga, Nb and REEs). Considering data from G1 and G2, it is proposed that differences up to 30% in the initial gradient enhance considerably the probability of uphill diffusion. In respect to the Exp4, expected (2n) patterns emerge in comparison with the values computed using available data from Chapecó dacites. It supports the generation of COD from the interaction of pre-existing CGD melts with crustal material. In respect to the linear mixing model LM, initial calculations are in disagreement with our experimental results regarding the best contaminant candidate on the formation of the Chapecó dacites. Taking into account the same degree of evolution derived from Ti contents, different pairs of trace elements lead to distinct contamination degrees f. The intervals of f are: i) 0.2 to 0.5 to generate CGD from Exp2 results; and ii) 0.3 to 0.7 to generate COD from Exp1 results. The best fits (i.e., similar f) are found for elements of G1 and Ti (such as Rb, Sr, U and Th) although REE and Pb are in good agreement as well. Data from Exp4 well reproduce the COD chemical behaviour of major compounds that usually act as network formers in melts (Fe, Al, Ti and Si). The experimental data produced in this thesis suggest that magma mixing might develop a central role on Chapecó dacites genesis by means of a chaotic dynamics. Reproduced features, such as morphological, chemical and frequency patterns, are similar to the PEMP dacitic rare outcrops, particularly for elements that present expressive initial gradients in the mixing system. It points towards short interaction times, low convective forces, and a predominance of density driven separation of contrasting melts (i.e., simulated conditions) as possible mechanisms involved in the genesis of these rocks. Further studies on trace elements (e.g., Exp4) and isotopic systems in the hybrid experimental glasses are necessary and may shed more light on the genesis of such relevant silicic magmatism.