ABSTRACT CTS-9720705 David A Kofke /SUNY Hydrogen fluoride (HF) plays a key role in many industries, but the study of HF via experiment is greatly hampered by its highly toxic and corrosive nature. Consequently there exists very little published experimental data on which to base the design of processes that use HF. The objective of this work is to improve our understanding of the molecular-level behavior of HF, and thereby develop engineering models that reliably predict the properties of HF and its mixtures from the few experimental data that are available. The approach synthesizes computational chemistry, molecular simulation, and state-of-the-art engineering modeling. Thermodynamic properties of particular interest include vapor-liquid phase equilibria, heat effects, and surface tension. Bulk hydrogen fluoride is a complex substance, and in fact has been described as "the most imperfect gas so far studied". The origin of HF nonideality is in its very strong nature to associate, forming chains of HF molecules. Chain formation manifests itself in (among other things) anomalously large vapor-phase heat capacity, an unusual vapor-pressure curve, and very small surface tension. Experiment has not provided conclusive information about the nature of these chains (i.e., the distribution of chain lengths, and the degree to which they form rings and branched structures), and this ignorance has hampered the formulation of engineering models that can predict HF behavior over a wide range of conditions. Although bulk HF is a complicated substance, the HF molecule itself is a very simple one (unlike, say, polymer molecules, which also form complex substances), and it has been studied very intensively by sophisticated theoretical means. These studies are the source of almost all of our detailed understanding of HF association; however this fundamental progress has not yet been applied in the formulation of engineering models and, moreover, the results pertain mostly to the vapor phase , so they do not provide a suitable basis for understanding vapor-liquid equilibria or surface tension. Our preliminary studies have found none of the existing molecular models to be suitably robust to warrant their use as a basis for constructing engineering models. Exploiting the molecular simplicity to systematically develop accurate molecular and engineering models of the complex bulk phase is the central component of this work. This effort is timely because of the convergence of several recent developments: the formulation of a good understanding of the HF molecule; the greatly advanced state of computation hardware; the development of sophisticated molecular simulation algorithms to aid the engineering-model development; and the formulation of appropriate, theoretically sound engineering methods for associating fluids, methods with which an HF model may be constructed. HF is a key ingredient in the production of hydrofluorocarbons (HFCs), which see application as environmentally benign blowing agents, solvents, refrigerants, and as ingredients in fluoropolymer manufacture; by international agreement entire industries are now switching from ozone-depleting chlorofluorocarbon to HFC technologies. HF is also important to the steel, petroleum, glass, electronic, aluminum and energy industries, and it is central to the continually growing field of fluorine chemistry. Safety is a major concern in HF processing, as accidental releases from production facilities can have disastrous consequences. Particularly troublesome is the ability of HF clouds to persist while traveling over long distances, without being dispersed, diluted, and thereby rendered harmless. This persistence is related to the anomalously low surface tension of HF, and its ability to form small droplets, or aerosols. Improved understanding of HF surface tension would enable the formulation of additives that promote dispersion of dangerous HF clouds. The complete quantitative synthesis of computational chemistry, molecular simulation, and engineering modeling has never before been completed to the extent in this study. HF is a logical starting point for this attempt because it is complex at the bulk scale but relatively simple at the molecular level. Success in this effort would guide the development of first-principles models for other complex materials.
