This dissertation presents a newly established, detailed model of elementary reactions of soot oxidation. Surface oxidation is the primary mechanism by which carbon is removed from soot. Previously, soot oxidation has been modeled using a one-step global mechanism based on chemical analogy to phenyl oxidation. The one-step model, however, is limited to simple descriptions of surface geometry. Studies of soot growth revealed that chemical analogy alone is insufficient to describe the reactions taking place at the particle surface. Steric effects, neighboring sites, and substrate size must also be considered. For soot growth, studies performed by the Frenklach group found a much richer chemistry than previously thought to exist, most notably the incorporation of five-member rings into the substrate leading to curvature of the initially planar aromatics. Recent experimental studies have found that the soot oxidation rate depends on soot particle curvature, so a one-step global mechanism is insufficient to explain the experimental observations. The purpose of developing a detailed kinetic model is to gain a qualitative and quantitative understanding of oxidation processes of soot particles, large polycyclic aromatics, and graphene edges.

To build the oxidation model, the energetics and kinetics of key elementary oxidation reactions were investigated. Reaction rate coefficients and product branching ratios were calculated for the oxidation of six- and five-member ring graphene-edges by molecular oxygen at temperatures and pressures relevant to combustion. The reactions and their corresponding rates were added to an existing model of soot surface growth along with reactions for oxidation of six-member rings by OH, thermal decomposition of oxyradicals, and O, H, and OH abstraction and addition.

Detailed kinetic Monte Carlo (KMC) simulations of high-temperature oxidation by molecular oxygen of a graphene sheet were performed using the newly established model of graphene-edge reactions. The KMC results revealed two principal pathways for an oxyradical site: oxidation and regeneration of an aromatic radical site. The overall oxidation rate is computed to be time-dependent, with reactivity decreasing over time as the ratio of reactive edge sites decreases relative to the number of basal-plane carbon atoms. At the same time, the oxidation rate was found to be higher for graphene with a higher initial curvature. Both results are in accord with experimental observations. Analysis showed that distinct aspects of graphene-edge morphology are responsible for curvature either raising or reducing the oxidative reactivity of the graphene edge.

Oxidation pathways for graphene-edge five-member rings reacting with atomic oxygen were investigated. The rate coefficient for oxidation by atomic oxygen exceeded that of for oxidation by molecular oxygen by several orders of magnitude. The detailed surface oxidation model was augmented to include this reaction along with several more abstraction and oxidation reactions.

KMC simulations were then performed in an evolving gas phase environment that was coupled to the surface chemistry for conditions analogous to those of shock tube experiments of soot oxidation at high temperatures. The KMC results showed that the oxidation rate was dependent on several factors. CO concentration profiles calculated from the KMC model were found to be in agreement with experimental measurements.

Abstract

University of California, Merced

Return on Investment for the Merced California Accountable Communities for Health Initiative

A dissertation submitted in satisfaction of the requirements

for the degree of Doctor of Philosophy in Public Health

by Ravi Singh

2020

The use of return on investment (ROI) analysis in public health is steadily increasing in recent years. Local health departments (LHDs) have started using ROIs more frequently to secure funding for their program and initiatives, as ROIs are viewed to be more attractive to potential funders compared to other evaluations. The Merced California Accountable Communities for Health Initiative (MCACHI) focused on reducing diabetes, heart disease, and associated depression in Merced County by implementing the National Diabetes Prevention Program (NDPP). The purpose of this three-paper dissertation was to estimate the ROI for the MCACHI, while providing a framework for researchers and LHDs for using ROIs for chronic disease prevention programs working closely with a community advisory group (CAG) .The reason the MCACHI was interested in estimating the ROI was to use it as a way of presenting to potential funders in the region that the program is a good investment, hoping the project can become sustainable. The second chapter aims to place a value by the general population for public health programs done in their communities, using a discrete choice experiment (DCE). The third chapter aims to cost out the entire MCACHI, estimating costs for the actual program as well as costing out the program if more people completed it and if it was implemented more efficiently. The fourth chapter of this dissertation aimed to calculate the ROI for the MCACHI, using information from the second and third chapter, while also providing a framework for using two methodologies for placing monetary values on health outcomes. The first method used information from the DCE in chapter two to estimate the willingness to pay (WTP) for the diabetes prevention program, and the second method used quality adjusted life years (QALYs) to place a monetary value on the health outcomes. This study found that the ROI is sensitive to many parameters and has the potential to show the program was a good investment over a longer time horizon.

Keywords: LHD, QALYs, ROI, WTP

Abstract:

Introduction: The demand for organs currently far exceeds supply. Understanding individuals’ motivations for deciding whether to donate an organ from a deceased relative would guide outreach efforts.

Methods: Focus group participants and literature were used to identify attributes to create a discrete choice experiment (DCE). Participants (N=86 ages 18 to 31 [mean=20.5]) were presented with 16 choices sets and asked to choose whether they would agree to donate a deceased relative’s organs. The choices contained attributes of the recipient (age, kidney’s lifespan, reason for failure, impact if not transplanted), the deceased donor (donor’s wishes and relationship to decision maker), and monetary incentives (amount, payer, payee). Conditional logit analysis was used to estimate the model, and latent class analysis identified two distinct groups of respondents.

Results: The results suggest a strong preference for donating organs, with the age of the recipient, reason for recipient’s need, and impact of not receiving the organ emerging as important factors. The financial incentive was not important. Latent class analysis suggested the two groups: Respondents in Class 1 placed relatively more importance on the wishes of the donor and having a financial incentive, while those in Class 2 placed relatively more weight on the impact should the recipient not receive the organ. Membership is the groups was predicted by gender and reported risk aversion.

Conclusion: DCE proved to be a useful tool for evaluating important factors in organ donation. Future studies can expand with evaluation of a larger sample representative of general population.