Thermochemical and Chemical Kinetic Data for Fluorinated Hydrocarbons

D.R.F Burgess, Jr., M.R. Zachariah, W. Tsang
Chemical Science and Technology Laboratory
National Institute of Standards and Technology
Gaithersburg, MD 20899-0001
and

P.R. Westmoreland
Department of Chemical Engineering
University of Massachusetts
Amherst, MA 01003-3110


ABSTRACT


Click HERE to Expand Figure

TABLE OF CONTENTS

1. INTRODUCTION

1.1. Overview . . . Goto 1.2.

The thermochemical and chemical kinetic data presented here were compiled as a part of flame-inhibition modeling work that was part of a large, short-term, intensive effort at NIST (Grosshandler et al., 1994) that evaluated (for the U.S. Air Force, Navy, Army and Federal Aviation Administration) potential fluorinated hydrocarbons and other agents as replacements for Halon 1301 (CF3Br). The modeling effort complemented the many experimental measurements in the overall NIST program, which characterized the effectiveness of various potential replacements. Our work focused on agent chemistry. This report is to provide documentation of the thermochemical and chemical kinetic data used in the modeling work. Further details of the evaluation of the thermochemical and chemical kinetic data and the ab initio calculations will be provided elsewhere, as will the results of the simulations of flame inhibition.

Halon 1301 is very effective as a chemical extinguisher. However, it is also extremely effective for depleting stratospheric ozone. Consequently, its production and use are restricted. The major objective of the modeling part of the overall work was to provide a chemical basis for rationalizing the relative degree of effectiveness of each candidate agent. A fundamental understanding of the chemistry of these agents in hydrocarbon flames should facilitate identification of desired characteristics of effective agents. That is, utilization of simple chemical concepts should enable screening and selection of potential agents with minimal time and human resources.

In order to accomplish this goal, it was necessary to develop a chemical mechanism based on elementary reaction steps for their destruction, their participation in and influence on hydrocarbon flame chemistry, as well as for prediction of potential by-products of incomplete combustion. However, neither a comprehensive mechanism nor a general review of the relevant the relevant chemistry existed prior to this study. Consequently, a significant effort was required in order simply to construct such a comprehensive mechanism prior to its use in any simulations. Further refinement of the mechanism requires experimental validation through comparisons with flame and flow reactor measurements. A few studies have already been completed (Westmoreland et al., 1993; Burgess et al., 1994; Westmoreland et al., 1994; Miziolek et al., 1994; Babushok et al., 1994; Daniel et al., 1994; Burgess et al., 1995; Babushok et al., 1995; Linteris et al., 1995).

The focus of the mechanism development work was restricted to the chemistry involving only fluoromethanes and fluoroethanes. This includes both four of the candidate agents specifically being considered as replacements (i.e., CH2F2, CF3-CH2F, CF3-CHF2, CF3-CF3), as well as all of the other possible fluoromethanes and fluoroethanes. Larger fluorinated hydrocarbon agents (e.g., C3F8) and chlorine-substituted agents (e.g., CHF2Cl) were not explicitly considered in our study, because these additions significantly increase the complexity of the chemistry that must be considered. However, the effectiveness of each can be estimated to some degree by analogy to the other agents that were studied. This can be done by using the qualitative trends observed and the fundamental understanding of the chemistry developed by this and future work.

The complete set of fluoromethanes and fluoroethanes were studied for two basic reasons. These are discussed below.

First, when the four specific candidate agents being considered decompose in the flame, they generate a pool of fluorinated hydrocarbon stable species and radicals, which results in the formation of many of the other fluoromethanes and fluoroethanes. Consequently, in order to describe adequately the decomposition of the four specific agents (and resultant chemistry), it is necessary to describe the chemistry of all of the many intermediates and products that are created, including most of the other fluoromethanes and fluoroethanes. For example, the lowest energy and primary decomposition pathway for one of the candidate agents, CF3-CF3, involves dissociation of the C-C bond to form (two) *CF3 radicals. These *CF3 radicals will then react with methyl radicals, *CH3, which are present in significant concentrations in hydrocarbon flames. This radical-radical combination reaction has two decomposition channels whose relative importance is dependent upon temperature and pressure. One channel results in the formation of a fluoroethylene, CH2=CF2 (and HF by-product). The other channel results in the formation of another fluoroethane, CH3-CF3. It is important that the mechanism adequately predicts the combustion products, the magnitude and rate of heat release in the flame, and ultimately the effectiveness of the added agent. Consequently, this requires that the mechanism also be able to describe correctly the decomposition of these two additional stable fluorinated hydrocarbon species (CH3-CF3, CH2=CF2). When one considers all decomposition channels for both the candidate agents and their decomposition products, one essentially must include all chemistry describing almost all of the fluoromethanes and fluoroethanes.

Second, given that there are to date very little experimental data with which to make comparisons, it is imperative to provide a level of self-consistency to this purely modeling work by considering a range of modeling parameters, including different reactor/flame geometries, different fuels, different (potential) agents, and many other conditions. In doing so, it allowed us to develop confidence in the validity of the qualitative trends that we observed. Quantitative prediction of the absolute or even relative effectiveness of the specific agents will require successful comparison of the simulations with experimental measurements.

1.2. Background . . . Goto 1.3.

There has been a significant amount of work over many years that has investigated the effectiveness of halogenated fire suppressants, as well as other types of fire suppressants (see Bibliography: Section 6.12. Flame Inhibition (halogens) and Section 6.13. Flame Inhibition (non-halogens). We will not review this body of work, but refer the reader to these and other relevant sources. A large part of our work is based on the pioneering work in this area by Biordi and coworkers (e.g., Biordi et al., 1973), Dixon-Lewis (e.g., Dixon-Lewis, 1977), Fristrom (e.g., Fristrom and Van Tiggelen, 1979) and Westbrook (1983). In earlier experiments on a range of candidates, CF3Br was identified as being very effective for extinguishing flames. However, the mechanism for inhibition by CF3Br was not understood. Biordi and coworkers measured both stable and radical species in methane flames doped with CF3Br using a flame-sampling molecular beam mass spectrometer. Many of the relevant elementary reactions that describe the decomposition of CF3Br, its chemistry, and its influence on hydrocarbon flames were determined in this work. Westbrook developed the first comprehensive chemical mechanism to describe in detail the chemistry of CF3Br and modeled inhibition in hydrocarbon flames. As a result of this work, it is generally agreed that flame suppression by bromine-containing compounds is a result of catalytic destruction of H atom species by Br atoms. The ability of bromine to recycle in the chemical system in the flame is directly related to the weak molecular bonds formed by bromine. Thus, bromine when complexed in molecules such as CF3Br, CH3Br, HBr, and Br2, can through a number of reactions regenerate Br atoms. It was also determined in these studies that *CF3, formed by decomposition of CF3Br, also removes H atoms by competition with radicals important to combustion (e.g., H, O, OH, HO2).

There are a variety of ways in which fire suppressants act in inhibiting hydrocarbon flames. Most of these effects are intimately related. For example, a heat loss means a temperature decrease, which causes the chemistry to slow, which means fewer radicals leading to product formation, which means less heat generated, which results in a further temperature decrease and so on. One can separate suppression effects into two general categories: physical and chemical (although there is overlap). Chemical effects are directly related to the characteristics of the specific molecule (e.g., H, F, or Cl substitution), while physical effects, in general, are not. For example, the heat capacity of a molecule is a physical effect, since to a first approximation it is largely a function of the number of atoms in the molecule and their connectivity but not the identity of the molecule (i.e., its chemical composition).

There are a number of physical effects that relate to fluid mechanics, mass transport, and heat transport processes that were not specifically studied as a part of this modeling work. Physical effects that should be taken into account are, for example, PVT properties (pressure-volume-temperature), discharge dynamics, dispersion mechanics, as well as more complex flow reactor geometries than could be considered in this work. We will not discuss these physical effects in any detail here. However, it is necessary to understand the impact of these effects as they relate to our results in order to provide a framework for transfer of the results of our simulations to more realistic conditions. Some of the fluid mechanics and mass transport effects include dilution, turbulent mixing, diffusive mixing, thermal diffusion, and buoyancy. Heat related effects, which cannot directly be addressed in the simulations, include the latent heat of vaporization and non-adiabatic effects, such as radiative cooling or heat losses for flames attached to a cold surface.

There are a number of different types of chemical effects. Most of these involve different competing factors. First, all of the fluorinated hydrocarbons will eventually decompose and then burn (forming CO2, H2O, and HF). This liberates heat and increases flame temperatures (this speeds flame chemistry). On the other hand, the agents considered as replacements are large molecules with many atoms. Consequently, their high heat capacities may result in a decrease in flame temperature prior to complete combustion (this slows flame chemistry). The competition between these two factors is strongly dependent upon conditions; most important of which is the mechanics of mixing of the fuel and the oxidizer. Another set of competing effects involves fluorinated radicals produced by agent decomposition. These radicals are slower to burn than their pure hydrocarbon analogues, because the C-F bond is significantly stronger than the C-H bond. Consequently, reactions involving these radicals may effectively compete with analogous pure hydrocarbon chemistry by creating less "flammable" intermediates, thereby inhibiting combustion of the hydrocarbon fuel. For example, since the agents are added to the air stream, their immediate decomposition products (radicals) are formed in oxygen-rich, relatively cold regions of the flame. Consequently, these radicals may be involved in termination steps, such as *CF3 + HO2* => CHF3 + O2, slowing radical chain reactions and inhibiting the flame. These radicals also compete with hydrocarbon radicals for important H, O, and OH radicals. On the other hand, these radicals also react with stable molecules in the colder air stream (e.g., O2), generating more radicals, such as O atoms, and thereby initiating chemistry or promoting combustion of the fuel.

1.3. Mechanism Development . . . Goto 1.4.

We constructed a large comprehensive reaction set or "mechanism" for fluorinated hydrocarbon chemistry involving C1 and C2 stable and radical hydrocarbon species, including partially oxidized fluorinated hydrocarbons. The mechanism should be considered only a framework for future model development, rather than a finished product. Future refinements will require experimental validation by high-temperature flow reactor, premixed flame, and diffusion flame measurements, as well as measurements of important, yet currently uncertain rate constants.

1.4. Acknowledgments . . . Goto 2. (Thermochemistry)

The authors wish to acknowledge support of the project "Agent Screening for Halon 1301 Aviation Replacement" by the U.S. Naval Air Systems Command, the U.S. Army Aviation and Troop Command, the Federal Aviation Administration Technical Center, and the U.S. Air Force. This program at NIST (Grosshandler et al., 1994) was under the direction of Mr. Michael Bennett at the Wright Patterson AFB Flight Dynamics Laboratory, Vehicle Subsystems Division, Survivability Enhancement Branch. The authors also wish to express their gratitude to Dr. Carl F. Melius of Sandia National Laboratories for use of the BAC (Bond Additivity Corrections) code.

FORWARD TO