Universität zu Köln

Chattopadhyay, Rajorshi ORCID: 0009-0008-3812-3731 (2025). Computational study of aqueous geochemical fluids at high temperature and pressure. PhD thesis, Universität zu Köln.

PDF (PhD Thesis) Rajorshi_PhD_Thesis_final.pdf - Accepted Version Bereitstellung unter der CC-Lizenz: Creative Commons Attribution. Download (11MB)

Abstract

Aqueous fluids play a critical role in the evolution of the Earth - starting from the atmosphere to the deep Earth. They are important carriers of heat energy and matter thereby playing an important role in several chemical reactions that determine the geochemistry of the Earth's interior. Hydrothermal solutions in the Earth reduce the melting point of rocks leading to the formation of volcanoes and plate tectonics. Lower water activity compared to CO2 in the Venusian crust leads to an increase in the melting point of rocks at elevated pressure which hinders plate tectonics. In a way, it is because of the aqueous fluid induced dynamic nature that several lifeforms have sustained on our planet. Hydrothermal fluids also play a dominant role in mobilising several strategically important elements followed by their precipitation resulting in the formation of ore deposits. Elements mobilised and deposited by aqueous fluids depend on the chemical composition and the prevalent thermodynamic conditions. As a result detailed understanding of the chemical structure and composition of hydrothermal fluids under geological conditions is warranted. Several spectroscopic studies have been conducted over the years to delineate the speciation of aqueous fluids. These have been accompanied by high temperature and high pressure experiments with devices like hydrothermal autoclaves, pressure vessels or heated/hydrothermal diamond anvil cells that aim to determine different physical and transport properties of aqueous fluids. However, these experiments are limited in their temperature, pressure and composition ranges. Ever since computers became available for unclassified scientific research, consistent effort has been made to implement fundamental laws of physics (like Newton's Laws of Motion, Schrödinger Equation) thereby providing insights into the structure and dynamics of natural systems on the atomistic scale. Development of more advanced and efficient hardware has made use of MD simulations realizable in studying different fluid rock interaction processes. These studies not only complement available experimental data but also provide an alternate framework to verify various assumptions and simplifications in existing theoretical models of fluids. In addition to these, MD simulations can provide insights into the structure and composition of aqueous geochemical fluids at conditions difficult to simulate in laboratories. In this thesis different types of MD simulations have been used to study geochemical fluids with different compositions at supercritical conditions. Different target properties are calculated and compared against available experimental data using DFT based AIMD simulations and CMD simulations. This provides discernment in the precision and accuracy of these techniques at geological conditions. In the different chapters of the thesis, we study properties of important solvents as well as solutes and the impact of different simulation parameters on them. In Chapter 3, we show that the GGA XC functional like BLYP and meta-GGA XC functional like r2SCAN predicts similar structural properties of aqueous H2SO4 and Na2SO4 solutions at supercritical conditions. Using r2SCAN gives better estimates of the different vibrational modes of oxidized sulfur species as compared to the BLYP functional. However, the predicted vibrational frequencies are still underestimated in comparison to experimental data. Both the functionals reproduce the temperature-induced shifts in different vibrational modes. Electrical conductivity of NaCl solutions in supercritical conditions are calculated with ReaxFF and pairwise SPC/E models in Chapter 4. We show that although ReaxFF is a polarizable, dissociative forcefield fitted on lots of QC data, its accuracy is not good in predicting a property (in this case electrical conductivity) not explicitly present in the training dataset. In Chapter 5, a new ab initio based PIM interaction potential is developed at supercritical conditions. We discuss fundamental issues related to transferability of interaction potentials and establish the superiority of the newly developed interaction potential in predicting structural and thermodynamic properties of La3+ in chloride bearing hydrothermal fluids. Overall, although state of the art techniques are precise and give important information about the geochemistry of supercritical fluids, significant scope of improvement exists. The thesis also serves to give an overview of fundamental issues related to the 'art' of interaction potential development that has to be kept in mind in future projects. This includes cardinal questions like XC functional to be used in generation of training data, types of data to be included in the training set and functional form of the potential to be fitted.

Item Type:	Thesis (PhD thesis)
Creators:	Creators Email ORCID ORCID Put Code Chattopadhyay, Rajorshi rajorshichat@gmail.com orcid.org/0009-0008-3812-3731 UNSPECIFIED
URN:	urn:nbn:de:hbz:38-788072
Date:	22 August 2025
Language:	English
Faculty:	Faculty of Mathematics and Natural Sciences
Divisions:	Faculty of Mathematics and Natural Sciences > Department of Geosciences > Institute of Geology and Mineralog
Subjects:	Natural sciences and mathematics Earth sciences
Uncontrolled Keywords:	Keywords Language Hydrothermal fluids English Thermodynamics and speciation UNSPECIFIED Molecular dynamics simulations UNSPECIFIED
Date of oral exam:	7 February 2025
Referee:	Name Academic Title Jahn, Sandro Prof. Dr.
Refereed:	Yes
URI:	http://kups.ub.uni-koeln.de/id/eprint/78807