Chemical reactivity hazards have posed a significant challenge for industries that manufacture, store, and handle reactive chemicals. Without proper management and control of reactivity, numerous incidents have caused tremendous loss of property and human lives. The U.S. Chemical Safety and Hazard Investigation Board (CSB) reported 167 incidents involving reactive chemicals that occurred in the U.S. from 1980 to 2001. According to the report, 35 percent of the incidents were caused by thermal runaway reactions, such as incidents that involved hydroxylamine and hydroxylamine nitrate. The thermal stability of hydroxylamine system under various industrial conditions was studied thoroughly to develop an understanding necessary to prevent recurrence of incidents. The macroscopic runaway reaction behavior of hydroxylamine system was analyzed using a RSST (Reactive System Screening Tool) and an APTAC (Automatic Pressure Tracking Calorimeter). Also, computational chemistry was employed as a powerful tool to evaluate and predict the measured reactivity. A method was proposed to develop a runaway reaction mechanism that provides atomic level ofinformation on elementary reaction steps, in terms of reaction thermochemistry, activation barriers, and reaction rates. This work aims to bridge molecular and macroscopic scales for process safety regarding reactive chemicals and to understand macroscopic runaway reaction behaviors from a molecular point of view.

Kyoto University (京都大学) / 0048 / 新制・課程博士 / 博士(工学) / 甲第14167号 / 工博第3001号 / 新制||工||1445(附属図書館) / 26477 / UT51-2008-N484 / 京都大学大学院工学研究科分子工学専攻 / (主査)教授 榊 茂好, 教授 田中 一義, 教授 田中 庸裕 / 学位規則第4条第1項該当

This dissertation is mainly focused on charge assisted noncovalent interactions specially hydrogen and halogen bonds. Generally, noncovalent interactions are only weak forces of interaction but an introduction of suitable charge on binding units increases the strength of the noncovalent bonds by a several orders of magnitude. These charge assisted noncovalent interactions have wide ranges of applications from crystal engineering to drug design. Not only that, nature accomplishes a number of important tasks using these interactions. Although, a good number of theoretical and experimental studies have already been done in this field, some fundamental properties of charge assisted hydrogen and halogen bonds still lack molecular level understanding and their electronic properties are yet to be explored. Better understanding of the electronic properties of these bonds will have applications on the rational design of drugs, noble functional materials, catalysts and so on. In most of this dissertation, comparative studies have been made between charge and neutral noncovalent interactions by quantum mechanical calculations. The comparisons are primarily focused on energetics and the electronic properties. In most of the cases, comparative studies are also made between hydrogen and halogen bonds which contradict the long time notion that the H-bond is the strongest noncovalent interactions.Besides that, this dissertation also explores the long range behavior and directional properties of various neutral and charge assisted noncovalent bonds.

In dieser Arbeit wurden die Molekülstrukturen und die elektronischen Eigenschaften von Vanadium(IV)- und Oxo-Rhenium(V)-Chelaten mit einem kombinierten molekülmechanisch-quantenchemischen Ansatz untersucht, um sterische und elektronische Effekte der Komplexierung mit einem theoretischen Modell zu quantifizieren. Es konnte gezeigt werden, daß auf diese Weise detaillierte Aussagen zu den Bindungsverhältnissen der Metallchelate getroffen werden können. Die Berechnung der Molekülstrukturen gelingt mit exzellenter Übereinstimmung mit den Kristallstrukturen der Komplexe. Die molekülmechanischen Berechnungen erfolgen auf der Grundlage des Extensible Systematic Force Field ESFF und des Consistent Force Field 91 (CFF91). Dabei konnte die hohe Flexibilität und Zuverlässigkeit des regelbasierten ESFF für eine Vielzahl verschiedenster Metallchelate nachgewiesen werden. Aufgrund der mangelhaften Ergebnisse für trigonal-prismatische Komplexgeometrien mit dem ESFF wurden eine Anpassung des CFF91 für derartige Vanadiumkomplexe vorgenommen. Auf Grundlage von theoretischen Ergebnissen wurden die alternativen Strukturen von isoelektronischen Vanadiumkomplexen berechnet und in Übereinstimmung mit experimentellen Daten, theoretischen Modellen der Komplexchemie und empirischen Fakten eine Hypothese für die Ursache der strukturellen Differenzen erarbeitet.<br> Der hier vorgestellte, kombinierte Algorithmus aus kraftfeldbasierter Geometrieoptimierung und single-point-Rechnung an diesen Strukturen ist ein zuverlässiger und relativ schneller Weg Molekülgeometrien von Metallkomplexen zu berechnen. Er kann somit zur Voraussagen von Komplexstrukturen und zur gezielten Modellierung definierter Koordinationsgeometrien verwendet werden. / In this work the molecular structures and the electronic properties of Vanadium(IV)- and Oxo-Rhenium(V)-chelates have been investigated to quantify steric and electronic effects of complexation. It has been shown, that in this way detailed insight can be gained into the bonding conditions of that metal complexes. Molecular mechanic calculations based on the Extensible Systematic Force Field (ESFF) and the Consistent Force Field 91 (CFF91) have been carried out. High flexibility and reliability of the rule based ESFF has been proven for a large variety of different metal chelates. Due to the poor ESFF-results for trigonal-prismatic complex geometries, a fit of the CFF91 for that species was done. Based on the theoretical results the alternative structure of isoelectronical vanadium(IV)- complexes have been calculated and a hypothesis on the reason for the structural differnces have been stated in accordance with experimental results, theoretical models of complex chemistry, and empirical facts. This combined approach of force field based geometry optimization and single point calculation at these structures has been proven to be a reliable and fast way to get molecular structures of metal complexes. It can be used to predict complex structures for modelling destinct coordination geometries.

Emerging contaminants are defined as synthetic or naturally occurring chemicals or microorganisms that are not currently regulated but have the potential to enter the environment and cause adverse ecological and/or human health effects. With recent development in analytical techniques, emerging contaminants have been detected in wastewater, source water, and finished drinking water. These environmental occurrence data have raised public concern about the fate and ecological impacts of such compounds. Concerns regarding emerging contaminants and the many chemicals that are in use or production necessitate a task to assess their potential health effects and removal efficiency during water treatment. Advanced oxidation processes (AOPs) are attractive and promising technologies for emerging contaminant control due to its capability of mineralizing organic compound via reactions with highly active hydroxyl radicals. However, the nonselective reactivity of hydroxyl radicals and the radical chain reactions make AOPs mechanistically complex processes. In addition, the diversity and complexity of the structure of a large number of emerging contaminants make it difficult and expensive to study the degradation pathways of each contaminant and the fate of the intermediates and byproducts. The intermediates and byproducts that are produced may pose potential effects to human and aquatic ecosystems. Consequently, there is a need to develop first-principle based mechanistic models that can enumerate reaction pathway, calculate concentrations of the byproducts, and estimate their human effects for both water treatment and reuse practices. This dissertation develops methods to predict reaction rate constants for elementary reactions that are identified by a previously developed computer-based reaction pathway generator. Many intermediates and byproducts that are experimentally identified for HO* induced reactions with emerging contaminants include common lower molecular weight organic compounds on the basis of several carbons. These lower carbon intermediates and byproducts also react with HO* at relatively smaller reaction rate constants (i.e., k < 109 M-1s-1) and may significantly affect overall performance of AOPs. In addition, the structures of emerging contaminants with various functional groups are too complicated to model. As a consequence, the rate constant predictors are established based on the conventional organic compounds as an initial approch. A group contribution method (GCM) predicts the aqueous phase hydroxyl radical reaction rate constants for compounds with a wide range of functional groups. The GCM is a first comprehensive tool to predict aqueous phase hydroxyl radical reaction rate constants for reactions that include hydrogen-atom abstraction from a C-H bond and/or a O-H bond by hydroxyl radical, hydroxyl radical addition to a C=C unsaturated bond in alkenes and aromatic compounds, and hydroxyl radical interaction with sulfur-, nitrogen-, or phosphorus-atom-containing compounds. The GCM shows predictability; factor of difference of 2 from literature-reported experimental values. The GCM successfully predicts the hydroxyl radical reaction rate constants for a limited number of emerging contaminants. Linear free energy relationships (LFERs) bridge a kinetic property with a thermochemical property. The LFERs is a new proof-of-concept approach for Ab initio reaction rate constants predictors. The kinetic property represents literature-reported and our experimentally obtained hydroxyl radical reaction rate constants for neutral and ionized compounds. The thermochemical property represents quantum mechanically calculated aqueous phase free energy of activation. Various Ab initio quantum mechanical methods and solvation models are explored to calculate the aqueous phase free energy of activation of reactantas and transition states. The quantum mechanically calculcated aqueous phase free energies of activation are within the acceptable range when compared to those that are obtained from the experiments. These approaches may be applied to other reaction mechanisms to establish a library of rate constant predictions for the mechanistic modeling of AOPs. The predicted kinetic information enables one to identify important pathways of AOP mechanisms that are initiated by hydroxyl radical, and can be used to calculate concentration profiles of parent compounds, intermediates and byproducts. The mechanistic model guides the design of experiments that are used to examine the reaction mechanisms of important intermediates and byproducts and the application of AOPs to real fields.