France fS*- A111Q3 D72b7M

cations

-Tor-BatiotoWMof Standards

? Otoary. t-Ol Admin. Bldg.

1«IAY 14 1981

1180855NBS SPECIAL PUBLICATION 531

*»«BlU Of* Li2ClOO

)51

U.S. DEPARTMENT OF COMMERCE / National Bureau of Standards

Summary Reportthe WorkshopHigh TemperaiChemical Kinel

Application:

Combustion Resec

NATIONAL BUREAU OF STANDARDS

The National Bureau of Standards' was established by an act of Congress March 3, 1901. The

Bureau's overall goal is to strengthen and advance the Nation's science and technology and

facilitate their effective application for public benefit. To this end, the Bureau conducts

research and provides: (I) a basis for the Nation's physical measurement system, (2) scientific

and technological services for industry and government, (3) a technical basis for equity in

trade, and (4) technical services to promote public safety. The Bureau's technical work is

performed by the National Measurement Laboratory, the National Engineering Laboratory,

and the Institute for Computer Sciences and Technology.

THE NATIONAL MEASUREMENT LABORATORY provides the national system of

physical and chemical and materials measurement; coordinates the system with measurement

systems of other nations and furnishes essential services leading to accurate and uniform

physical and chemical measurement throughout the Nation's scientific community, industry,

and commerce; conducts materials research leading to improved methods of measurement,

standards, and data on the properties of materials needed by industry, commerce, educational

institutions, and Government; provides advisory and research services to other Government

Agencies; develops, produces, and distributes Standard Reference Materials; and provides

calibration services The Laboratory consists of the following centers:

Absolute Physical Quantities' — Radiation Research — Thermodynamics and

Molecular Science — Analytical Chemistry — Materials Science.

THE NATIONAL ENGINEERING LABORATORY provides technology and technical

services to users in the public and private sectors to address national needs and to solve

national problems in the public interest; conducts research in engineering and applied science

in support of objectives in these efforts; builds and maintains competence in the necessary

disciplines required to carry out this research and technical service; develops engineering data

and measurement capabilities; provides engineering measurement traceability services;

develops test methods and proposes engineering standards and code changes; develops and

proposes new engineering practices; and develops and improves mechanisms to transfer

results of ils research to the utlimate user. The Laboratory consists of the following centers:

Applied Mathematics — Electronics and Electrical Engineering 2 — Mechanical

Engineering and Process Technology 2 — Building Technology — Fire Research —Consumer Product Technology — Field Methods.

THE INSTITUTE FOR COMPUTER SCIENCES AND TECHNOLOGY conducts

research and provides scientific and technical services to aid Federal Agencies in the selection,

acquisition, application, and use of computer technology to improve effectiveness and

economy in Government operations in accordance with Public Law 89-306 (40 U.S.C. 759),

relev ant Executive Orders, and other directives; carries out this mission by managing the

Federal Information Processing Standards Program, developing Federal ADP standards

guidelines, and managing Federal participation in ADP voluntary standardization activities;

provides scientific and technological advisory services and assistance to Federal Agencies; andprovides the technical foundation for computer-related policies of the Federal Government.The Institute consists of the following divisions:

Systems and Software — Computer Systems Engineering — Information Technology.

Headquarters and Laboratories at Gaithersburg, Maryland, unless otherwise noted;mailing address Washington, D.C. 20234.

Some divisions within the center are located at Boulder, Colorado, 80303.

The National Bureau of Standards was reorganized, effective April 9, 1978.

Bureau of Standfi*

1379 Nummary Report on the Workshop on HighTemperature Chemical Kinetics:

Applications to Combustion Research

[j

D. Garvin, R. L. Brown, R. F. Hampson,M. J. Kurylo, and W. Tsang

National Measurement Laboratory

National Bureau of Standards

Washington, D.C. 20234

Sponsored by

Department of Energy

Washington, D.C. 20545 and

Office of Standard Reference DataNational Bureau of Standards

Washington, D.C. 20234

U.S. DEPARTMENT OF COMMERCE, Juanita M. Kreps, Secretary

Jordan J. Baruch, Assistant Secretary for Science and Technology

NATIONAL BUREAU OF STANDARDS, Ernest Ambler, Director

Issued December 1978

Identification of commercial products does

not imply recommendation or endorsement by

the National Bureau of Standards nor that the

product is necessarily the best available for the

purpose.

Library of Congress Catalog Card Number 78-600125

National Bureau of Standards Special Publication 531Nat. Bur. Stand. (U.S.), Spec. Publ. 531,89 pages(Dec. 1978)

CODEN: XNBSAV

U.S. GOVERNMENT PRINTING OFFICE

WASHINGTON: 1978

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402

Stock No. 003-003-02005-0 Price $2.75

(Add 25 percent additional for other than U.S. mailing).

Table of Contents

Abstract 1

1. Introduction 1

2. Program of the Meeting 4

3. Commentary on the Meetingand Suggestions 6

4. Abstracts of Invited Papers 13

5. Discussion of Needs in HighTemperature Kinetics Research 28

5.1 Outline for the Discussion 285.2 Summary of the Discussion 31

6. Discussion of Acquisition of HighTemperature Kinetic Data 37

6.1 Outline for the Discussion 376.2 Summary of the Discussion 42

7. Contributions Submitted by Participants 49

8. List of Registrants 80

iii

Summary Report on the Workshop on High Temperature ChemicalKinetics: Applications to Combustion Research

D. Garvin, R.L. Brown, R.F. Hampson, M.J. Kurylo, andW. Tsang

Abstract

The proceedings at a workshop on the applications ofhigh temperature chemical kinetics to combustion researchare summarized. This workshop, held Dec. 12-13, 1977 atthe National Bureau of Standards, provided a forum for theexchange of views on the needs for kinetics research duringthe next five to ten years. Experimental techniques,measurements, theoretical developments and data evaluationwere treated in four review papers and two discussionsessions.

This report contains the program of the meeting,abstracts of the review papers, a commentary by theorganizers of the meeting, summaries of the discussions,formal comments submitted by the participants and anattendance list.

Keywords: chemical kinetics, combustion, high temperaturechemistry, reaction rate constants, measurementtechniaues, assessment of needs.

1 . Introduction

On December 12 and 13, 1977, a workshop was held at theNational Bureau of Standards, Gaither sburg , MD on "HighTemperature Chemical Kinetics: Applications to CombustionResearch". This workshop was sponsored by the Division ofBasic Energy Sciences, Department of Energy and the Officeof Standard Reference Data, NBS. It was attended by onehundred persons, mainly chemical kineticists interested incombustion, but also including researchers concerned withthe modeling of practical combustion systems and thedevelopment of diagnostic techniques for probing suchsystems

.

The objective of the workshop was to provide a forumfor an exchange of views on the needs for high temperaturechemical kinetics research during the next five to tenyears, research that would support the development ofimproved chemical theoretical developments, experimentalmeasurements and data evaluation.

1

Speakers and discussion leaders were instructed toattempt to answer the following questions for each .topicd iscussed

:

Why it is important in combustion research?What is the current state of our knowledge?What do we need to know, and how accurately?What speculative areas should be explored?What do you see as the research priorities?

Four invited review papers provided orientation for theparticipants. These papers covered topics in modeling ofcombustion, the application of laser diagnostic techniques,combustion chemistry and mechanisms, and the availabilityof high temperature rate data. Two extensive discussionperiods were directed toward needs for high temperaturekinetics research and the acquisition of high temperaturekinetics data.

This report contains the program of the meeting,abstracts of the invited papers, summaries of the generaldiscussions, written contributions submitted by theparticipants and a commentary by the organizers of themeeting. An attendance list is appended.

The association of chemical kinetics and combustion isnot new. Almost from the start the programs of theInternational Symposia on Combustion, organized by theCombustion Institute, have featured kinetic studies ofindividual reactions and of oxidation systems. The journal"Combustion and Flame" regularly features articles onkinetics

.

Nor are discussions on the interrelationship ofkinetics and combustion research novel, although they havegrown in importance recently. In 1974 the National ScienceFoundation sponsored a "Workshop on Energy-related BasicCombustion Studies" (Glassman and Sirirnano, 1974) thatincluded an analysis of combustion related kinetics.Problems discussed then are still with us today. InSeptember 1977, Project Squid held a "Workshop on:Alternate Hydrocarbon Fuels for Engines: Combustion andChemical Kinetics" at Columbia, Maryland. It was sponsoredjointly by the Department of Energy and the Department ofDefense. The kinetics topics covered at that meeting andat the present one overlap extensively. The workshopreported on here has been followed by a "Workshop onModeling of Combustion in Practical Systems", held at LosAngeles in January 1978, also sponsored by the Departmentof Energy.

2

These interdisciplinary meetings indicate a high levelof concern for the solution of combustion problems on thepart of the sponsors. They have enhanced the knowledge ofthe participants by exposing them to the problems and needsof workers in disciplines other than their own. Problemsrequiring early solution become apparent. Continued cross-discipline communication should improve our understandingof the complex phenomenon of chemical combustion.

3

2. Program of the Meeting

Monday , December 12 , 1977

9:15 Welcome and Introductory Remarks

E. Horowitz, Deputy DirectorInstitute for Materials Research, NBS

J. S. Kane, DirectorDivision of Basic Energy Sciences, DOE

9:30 INVITED PAPERS

Session Chairman: L. H. Gevantman, NBS

Recent developments in computer modelling ofchemical combustion systems.J. S. Chang and C. K. Westbrook,Lawrence Livermore Laboratory

10:30 Coffee break

11:00 Advances in laser diagnostics for combustionresearch

.

D. L. Hartley,Sandia Livermore Laboratory

12:00-12:45 DISCUSSION: Needs for High Temperatureand Kinetics Research

2:00-5:15 Discussion Leaders: I. Glassman and F . L. DryerPrinceton UniversityReporters: R. F. Hampson and R. L. Brown,National Bureau of Standards

Tuesday, December 13, 1977

9:15 INVITED PAPERS

Session Chairman: 0. W. Adams, DOE

Combustion Chemistry and Mechanisms.D. M. Golden,Stanford Research Institute

10:15 High Temperature Chemical Kinetics Data.D. Garvin and W. Tsang,National Bureau of Standards

11:15 Coffee Break

11:45-12:45 DISCUSSION: Acquisition of High Temperatureand Kinetic Data

1:45-4:45 Discussion Leaders: S. W. Benson,University of Southern CaliforniaD. Gutman,Illinois Institute of TechnologyReporters: W. Tsang and M. J. Kurylo,National Bureau of Standards

4:45 Adjournment

Organizing Committee: 0. W. Adams (DOE);D. Garvin, L. H. Gevantman, R. F. Hampson,and W. Tsang (NBS) .

5

3. Commentary on the Meeting and Suggestions

W. Tsang and D. Garvin

This section is concerned with the role of chemicalkinetics in combustion research. Topics considered are thetypes of contribution that kinetics can make , interactionsbetween technological disciplines involved in the study ofpractical combustion systems , identification of somegeneral problems and a summary of the state of the art.

The statements made here are the personal opinions ofthe authors. They are not a summary of agreed-uponconclusions reached at the workshop. No such summary wasattempted, the belief being that information transfer amongthe participants should be a primary goal at this stage.But the statements have been stimulated by and draw heavilyon ideas expressed at the workshop. They also incorporatematerial presented elsewhere. Other persons undoubtedlywill wish to supplement, revise and refocus these remarks.

Combustion research and development involves much morethan combustion chemistry, which, in turn, is a broadertopic than combustion kinetics . The conceptualorganization of the field of combustion has been analyzedby Essenhigh (ref . 1). His summary is shown in Figure 1.Although this was developed for industrial furnacetechnology, the translation to other combustion systems issimple. Closely analogous analyses undoubtedly hold forother practical systems that include kinetics, e.g.,industrial chemical processing and air oollution control.

Three successive stages of activity are shown in Figure1: basic sciences, technological interactions of basicsciences, and engineering systems application anddevelopment. Studies of reaction kinetics and mechanismsfall in the basic sciences stage and only begin to beapplied to combustion at the technological stage, where

j

they are combined with the outputs from other sciences.Direct interaction of kinetics with the engineering stageis not shown. Instead, the scientific results areprocessed via major new concepts introduced at the otherstages. Inevitably, the details considered important atthe basic sciences level are hidden (or lost) in the finaloutputs.

6

This leads to an appearance of strong decoupling of thesciences from the final outputs and makes it very difficultto attribute improvements at the engineering level toadvances in the sciences. That this decoupling is not realis becoming apparent as detailed modeling is applied topractical combustion systems. Two conclusions from a

recent modeling study bear on this point. Griffin et al

.

(ref. 2) modeled the fluid dynamics of the internalcombustion engine. Their model included a detailed kineticdescription of the chemistry. They stated:

"The combustion process in the IC engine can berepresented by finite-rate chemical reactions.The solutions are subject to inaccuracies inthe rate data."

They also concluded:"A partial mechanism can simulate the heatprocess due to combustion, but the completemechanism is required to compute the pollutantsaccurately. n

The first quotation provides evidence of a direct influenceof the scientific research on the final output. Thatinfluence would be there, although perhaps diluted andcertainly hidden, in design studies that did not useexplicit modeling techniques. The second speaks to anever-present problem: simplified chemistry can be used forstudies of limited scope, but different chemistry isimportant in different problems.

Within the discipline of chemical kinetics there is a

widely held belief that detailed study of the chemistrythat occurs in oxidation and pyrolysis systems (mechanisms)and quantitative data on the important elementary steps(rate constants) will be required for a satisfactoryspecification of combustion. That belief is valid and issupported by experience with partial solutions. It helpsexplain why the study of combustion kinetics is a goal initself, independent of the consideration of practicalsystems. Indeed, oxidation kinetics are sufficientlycomplex to require the full attention of the specialists inthe art. The same is true for other scientific disciplinespertinent to combustion. Thus there is a self-imposeddisciplinary decoupling, in the interest of solving the"local" problems, that may be a major barrier to ,the

development of coherent combustion research programs.

7

Although a detailed understanding of kinetics may benecessary for the advancement of combustion research (itcertainly is not sufficient) , it is not apparent that thisgeneralized approach can be pursued rapidly enough for theresults to be widely useful in technological studies. Theproblem is that there are many practical combustion systemsand the study of them is going on in parallel with the workon kinetics. This indicates a need for establishment ofpriorities for kinetics studies and a need for moreinterdisciplinary interactions than in the past.

Nobody has solved the problem of establishment ofpriorities within a research community that consists ofmany independent small research groups, and we do notattempt to do so here. Clearly any priorities must berelated to the applied problems that will be important inthe near future (research takes time, impact on immediateproblems is likely to be small) . But priorities forscientific studies that are stated in terms of theperformance of engine X or furnace Y are inappropriate andare, themselves, decoupled from the sciences. More usefulpriorities would cover classes of behavior applicable toseveral combustion systems. They could be stated in termsof needs for basic studies that would aid in the solutionof identified technological problems, e.g., quench-layerkinetics in internal combustion engines, soot formation,pyrolysis in fuel droplets, pollutant formation in theoxidation of alternative fuels or ignition in aromatichydrocarbon-air mixtures. It is then up to the kineticiststo translate these needs into possible experiments, andexplain what parts of the problem can be solved.

The situation is more promising with respect tointerdisciplinary interaction. Most important is thedirect feedback from modeling studies. Because models usedetailed kinetics, and because sensitivity analyses of themcan point out areas of weakness, they can indicate researchneeds. In addition, the explicit display of the chemistryused in the models permits the kineticist to determinewhether or not that chemistry is consistent wih fundamentalstudies. We suggest that such analyses should be given thestatus of a branch of combustion kinetics.

Interpretation and simplification of kinetics for thebenefit of others is another promising area ofinterdisciplinary work. The combustion scientist mayproperly ask: "which aspects of the chemistry areimportant in my problem and which can be ignored for thepresent?" The person studying spray combustion, however

8

expert in his own field, should not be expected to beknowledgeable about reaction mechanisms or estimating rateconstants. Guidance from the kineticist is desirable. Oneform of this is the development of simplified ("global")mechanisms. But extensive collaboration probably is

required, as opposed to general guidance. We suggest thatsuch collaborations should be promoted.

The discussion above covers general aspects of the roleof kinetics in combustion. We conclude this section bylisting some specific conclusions. Most of these areconcerned with the state of the art within the disciplineof combustion kinetics, but some touch on the generaltopics considered above.

1. Chemical kinetics of high temperature transformationsinvolving hydrocarbons and related compounds areessential for the conceptualization, understanding,control and optimization of combustion devices.

2. Types of needed information range from general mechanismsto detailed single step rate constants. They coverheterogeneous as well as homogeneous phenomena andinvolve the oxidation and pyrolysis of aliphatic,aromatic and heterocyclic compounds. Data on pollutantformation is particularly important. These reactionscannot be decoupled from the basic combustion process.

3. Global rates constants are needed for the modeling ofspecific devices. They are not transferable and can beentirely empirical but preferably should be traceableto the pertinent set of single step rate constants.

4. Traditional engineering usage has place greater relianceon general mechanisms as opposed to the detailedkinetics necessary for computer simulations. Thelatter represents a growing research area and hasenormous potentials for providing complete and generalsolutions and, ultimately, as an alternative tophysical testing.

5. There is a serious lack of mechanistic information onthe pyrolysis and oxidation of aromatic andheterocyclic systems. Thus the information for thedesign of combustion devices using various coal derivedfuels even on a traditional basis does not exist.

6. Although the mechanisms of the oxidation and pyrolysisof aliphatic hydrocarbons are fairly well understood,large numbers of the needed rate constants remainundetermined. There is a particular need for data onhydrocarbons larger than methane.

7 There exists a vast arsenal of kinetic tools for thedetermination of the mechanism and rates of hightemperature transformation in the gas phase. • Allreactions involving ground state molecules probably areaccessible. There has been however no systematiceffort (similar to that for stratospheric pollution) inthis direction.

8. Transition state theory provides a possible frameworkfor the extrapolation, prediction and correlation ofrate parameters. Critical tests are still necessary.Particular information needs include thermal propertiesof unstable species as well as rate determinations.

9. In all kinetic studies proper understanding of mechanismsmust precede accurate rate determinations. Particularlyimportant is the application of a variety of experimentalmethodologies to the same system. Laser basedtechniques have qreat potential but there are severe

limitations especially in complex systems withmultiplicity of species (as in combustion) . Moretraditional techniaues, with capabilities for totalanalysis, e.g. gas chromatography and massspectrometry will continue to play key roles withrespect to unraveling mechanisms.

10. Computer modeling of combustion or other reactive systemsis still in its infancy. There appears to be a directneed for the establishing of proper methodology in thedevelopment and extraction of information from models.Algorithms for carrying out sensitivity analysis andapplication to combustion systems must be developed.

11. There is essentially no capability in the combustionkinetics community for supplying detailed single stepquantitative data on condensed media or multiphasereaction processes. These are research frontier areasand careful studies on model systems, with extensiveaid from computer simulations will be of more lastingvalue than direct studies in combustion environments.At present the best that can be expected from such workis qualitative mechanistic information. Soot formationand reactions in or on soot also falls into thiscategory. Homogeneous gas phase reactions are ingeneral important components of these processes.Proper understanding of these processes may be a

prereauisite for advances.12. It is likely that there are too many physical properties

of interest for combustion studies (rates,thermodynamic data) for all to be measured. Theappropriate tactics are: (a) measure the most *

important, and (b) estimate the remainder.

10

References

1. Essenhigh, Robert H. , "Research Opportunities forUniversities in Combustion of Coal", Proceedings ofConference on "Research Opportunities for Universitieson Coal" OCR/NSF-RANN, SUNY, Buffalo, NY, October,1974 .

2. Griffin, M. D. , Diwaker, R. , Anderson, J. D. , andJones, F. , "Computational Fluid Dynamics Applied toFlows in an Internal Combustion Engine", AIAA 16thAerospace Sciences Meeting, 78-57. Huntsville,Alabama, January 16-18, 1978.

1

ii

12

4. Abstracts of Invited Papers

COMPUTER MODELING OF COMBUSTION CHEMICAL KINETICS

J. S. Chang and C. K. Westbrook

Lawrence Livermore Laboratory

P.O. Box 808, Livermore, California 94550

ABSTRACT

Computer modeling of chemical kinetics in combustionsystems is a complex chemical, physical, and mathematicalproblem. In most practical applications, chemicalreactions are closely coupled to other physical processes,adding further difficulties to the analysis of the chemicalkinetics features of the combustion. In this paper we willtry to indicate the degree of sophistication which iscurrently possible in numerical models of detailed chemicalreaction mechanisms. We will also describe some of themethods used for simplifying or idealizing real combustionsystems in order to make their analysis possible, usingavailable numerical methods and computers. Finally, wedescribe what we feel are the most probable directions inwhich kinetic modeling for combustion applications islikely to proceed.

In addition to chemical kinetics, most combustionsystems involve a variety of other physical processes,including mass and energy transport by means of eithermolecular diffusion or turbulent transfer. In some systemsmultiphase flows occur, with liquid atomization andvaporization processes in many of these situations.Radiative heat transfer may be important, and surfaceeffects, including wall heat losses and catalyticcombustion, may occur. In each large class of realcombustion systems a somewhat different and distinct set ofphysical operators contribute to the overall systemperformance. Many, if not all of these physicaloperators are themselves extremely difficult to modelcomputationally. In view of this degree of complexity, itis desirable to place our initial emphasis on physicalsystems in which chemical kinetics is the dominant physicaloperator. Some such systems are one dimensional laminarflames, shock tubes, and the turbulent flow reactor. Theseconfigurations are often either time-independent orspatially invariant, greatly simplifying the computational

13

analysis. We will show results of several computationsmade for laminar flame structure and propagation,, as wellas results of turbulent flow reactor experiment modelingfor a variety of fuels.

It is our opinion that for a pure kinetics problem, a

mechanism of any reasonable size can be handled numericallyby available methods and computers. This capability wouldextend to kinetic mechanisms involving up to severalhundred chemical species and many hundreds of chemicalreactions. Since known reaction mechanisms are muchsmaller than these sizes, this statement includes anycurrently proposed reaction mechanisms. The computationaldifficulties begin to arise when spatial variations areintroduced. While the pure kinetics problem might include20 equations (species) , a two-dimensional axisymmetr iccombustion chamber with a resolution determined by a 50 x

50 computational grid would require the solution ofapproximately 50 x 50 x (5 + 20) or 62500 equations foreach time step. Since the computational time is required

for a problem of this type depends roughly on n , whereis the number of equations, the two-dimensional problemquickly becomes intractable. Inclusion of detailedreaction mechanisms for problems with one space dimensioncan be rather expensive but is possible, and some exampleswill be shown of stratified charge internal combustionengine studies in one space variable.

Since two- and three-dimensional geometry precludes theinclusion of detailed reaction mechanisms, it is necessaryto find alternative kinetic models for these problems. Wewill briefly discuss a number of global, quasi-global, andsemi-global reaction mechanisms in present use, some oftheir properties, and some examples of their use inturbulent flow reactor simulations. Several other schemeswill also be discussed, including species "lumping 58

,

partial equilibrium assumptions, and partial decoupling ofportions of the kinetic mechanism for species purposes,,

We will discuss a number of combustion areas in whichchemical kinetics are either dominant or present specialproblems, either in practice or with respect to theirmodeling. These include NOx , SOx , and soot formation,ignition processes, and catalytic combustors. In addition,we will describe a series of numerical studies of reactionquenching by a variety of physical mechanisms. Theseinclude quenching by a thermal boundary layer (wallquench) , by a sharp gradient in equivalence ratio (volumequenching) , and by a rapid expansion of the combustionchamber (bulk quenching)

.

14

We have attempted to indicate areas in which combustionmodeling has specific needs for which kineticists may helpprovide solutions. Most obviously, there are a multitudeof elementary chemical reactions for which rate data areunavailable or unreliable. Thermochemical data andreaction mechanisms are also needed for most practicalcombustion fuels. In many fuel oxidation mechanisms,kineticists could provide insight into identifying theimportant and rate-controlling steps. This would help intwo major ways. First, it would help to validate thedetailed mechanisms used in thermodynamic (0-D) andone-dimensional models. Also, it would motivate theselection of semi-global or overall steps which could beused to simplify reaction mechanisms for use in largermodels. Another important function would be to indicatethe degree of detail necessary , including species,elementary reactions, and reaction rates, for differentappl ications

.

We will briefly discuss coordination between modelingand diagnostics. It is important for the modeler torecognize the limitations and possibilities of experiments,including concentration, temperature, and velocitymeasurements. Problems exist for both types of researchwith respect to physical resolution, accuracy, and otherproperties. We see a definite value in designing idealizedexperiments to emphasize kinetic effects in much the samemanner as noted earlier for modeling purposes. And modelvalidation by comparison with a variety of well documentedexperiments is very important. We also recommend that, aspart of its data base development, NBS provide criticalevaluation of selected experiments which emphasize kineticeffects. These experiments can then be very useful asmodel validation vehicles.

We find that reaction rate data is best represented byArrhenius coefficients. However, temperature regimes incombustion environments can vary from room temperature toover 3000 K. Many important reactions display significantnon-Ar rhenius behavior over this range, and it would behelpful for a data base to find a convenient way to handlethis behavior within a basically Arrhenius framework. Wealso believe it would be very useful to maintain data forheats of formation, equilibrium coefficients of formationin both tabular and Arrhenius form, reverse reaction rates,and all manner of thermochemical data. While we realizesome of this information is redundant and perhaps evenformally inconsistent, it would greatly assist manymodeling programs.

15

ADVANCES IN LASER DIAGNOSTICS FOR COMBUSTION RESEARCH

D. L. Hartley

Combustion Sciences Department 8350

Sandia Laboratories

Livermore, CA 94550

ABSTRACT

Numerous advances have been realized in recent years indeveloping and applying new measurement techniques forexperimental studies of combustion processes. Intrusiveprobes such as thermocouples and sampling probes haveprovided a wealth of information on combustion systems andundoubtedly will continue to do so. More recently,however, with the advent of electro-optic devices such aslasers and the rapidly developing array of detectors,techniques for measuring properties in combustion systemsare being developed which do not require physically probingthe experiment with a perturbing probe, but rather use someform of optical scattering to measure flow conditions.

The importance of these optical scattering techniquesarises from their ability to provide excellent timeresponse from essentially a point measurement in space aswell as to overcome the limitations of physical probes.These limitations are largely related to the disturbancethat they impose upon the system under study, thedifficulty in surviving the hostile environment presentedby combustion systems, and the difficulty in many cases ofrelating the measurement to properties of interest.Optical scattering techniques can be used to overcome manyof these limitations for a large number of researchproblems

.

During the last few years, several studies have beenconducted by various segments of the scientific communityto assess the status of optical scattering methods and toidentify their limitations and the potential for extendingtheir capabilities to problems in combustion research. In1974, the American Physical Society conducted a summerstudy on the role of physics in combustion and highlightedoptical diagnostics (Ref. 1). In 1975, Project SQUIDconducted a workshop focused on combustion diagnostics injet propulsion systems (Ref. 2). In 1976, the AIAAemphasized combustion diagnostics in their annual National

16

meeting and published special proceedings containing mostof these papers (Ref . 3) . In 1977 the EPA sponsored a

study at United Technologies to provide a detailed reviewof laser techniques for practical combustion diagnostics(Ref. 4) and ERDA sponsored a study at Sandia addressingthe application of these techniques to measurements incoal-fired gas turbines (Ref. 5)

.

To underscore the importance of optical diagnostics incombustion research, the DOE has approved the constructionof a Combustion Diagnostics Facility at Sandia inLivermore, California, where several specially designeddiagnostic systems will be made available for qualifiedusers. This facility will be operational in mid-1980.

The purpose of this presentation is therefore twofold.On the one hand, this paper will introduce these techniquesand their current status to the chemical kineticists in theaudience who may be interested in potentially using them intheir own research. On the other hand, feedback from thisworkshop will help formulate decisions concerningdiagnostic equipment choices at the Combustion DiagnosticsFacil ity

.

Hopefully reflecting the special interests ofresearchers in high temperature chemical kinetics, I willlimit the discussion of diagnostics to those which arecapable of measuring species concentration and temperatureover regimes of their interest. The techniques I willdiscuss are elastic Rayleigh scattering; linear inelasticscattering such as spontaneous Raman; resonant Paman andfluorescence; line-of-sight absorption; and severalnonlinear optical processes including Coherent Anti-StokesRaman Spectroscopy (CARS) , Stimulated Anti-Stokes RamanSpectroscopy (SARS) , Raman Induced Kerr Effect Spectroscopy(RIKES) , and High Order Raman Spectral ExcitationSpectroscopy (HORSES)

.

A brief description of field image techniques; lasershear inter ferometry and Ramanogr aphy , will also bepresented

.

In this abstract I shall discuss only a few principalscattering techniques, namely Rayleigh, Spontaneous Raman,Fluorescence and some of the nonlinear optical techniques.

Rayleigh Scattering is an elastic scattering of photonsby the irradiated molecules. Being elastic, the scatteredlight experiences no characteristic shift in frequency,

17

thus it can only be used to measure total number densityor, by resolving the Doppler line width, may be used tomeasure temperature. Although of limited versatility,Payleigh scattering has been used recently to measuretemperature profiles in diffusion flames and to measuredensity profiles through a moving flame front. The highscattering cross-section makes Rayleigh scattering easilyapplied by laboratory researchers.

Raman Scattering is an inelastic process in which thescattered light has undergone a change in frequencycharacteristic of the internal energy levels of theirradiated molecule. For the applications discussed here,these internal modes are the vibrational and rotationalenergy level manifold. The intensity of light scattered ata given Raman wavelength shift (small shifts forrotational, larger shifts for vibrational) can be relateddirectly to the number density of molecules of thatparticular species in the sample volume. For high-temperature gases, high molecular energy levels arepopulated increasingly, and the resulting distribution ofintensity among Raman scattering lines is alteredaccordingly. Measurement of this redistribution of energy(i.e., shifts in intensity of ground-state and excited-state vibrational Raman lines) can be used to determine gastemperatures

.

In compensation for the low signal strength of Ramanscattering, its virtues include (1) specificity of speciesidentity, (2) time resolution not limited by scatteringprocess (10-14 sec) , (3) accessibility of temperatureinformation, (4) capability of probing systems not inchemical or thermal equilibrium, and (5) relative lack ofinterferences.

Numerous applications of spontaneous Raman scatteringto combustion research have emerged in the last two orthree years. Laboratory systems, including flat flames anddiffusion flames, and practical systems from internalcombustion engines to gas turbines have been probed withRaman systems with varying degrees of success. Theseresults shed considerable light on the potential range ofapplication of spontaneous Raman scattering to problems inchemical kinetics. "Typical" conclusions for Ramanscattering studies at a gas turbine inlet (Ref. 5) arethat in the absence of limitations due to particulates orlaser- induced spectral background (both problems yet to becompletely evaluated) , 2 a errors of less than 10% are

18

predicted for total measurement times of 0.02 to 0.5seconds for all major species (N2, 02, C02, H20 and 0.8 to80 seconds for minor species (NO, CO, S02) using a

conventional cw laser or advanced high- energy pulsed laser(one joule per pulse, 10 pulses per second).

Laser Fluorescence like Paman scattering, can be usedas a species-sensitive optical detection method.Fluorescence techniques, using cw or pulsed tunable lasersources in principle offer the possibility of extremelysensitive detection of both atomic and molecular species.However, since fluorescence is an absorption andre-emission process, quenching and pressure broadening makedata interpretation difficult. For this reason,fluorescence techniques involving saturation show greaterpromise since such approaches minimize the problem ofquenching. Typical molecule s of interest, which have anabsorption wavelength that is accessible to tunable lasersources (roughly 0.2-1.5 microns) are C2, CH, CN, CS, NH

,

NH2, NO, OH, S02, HCN , CH20, C6H6, and C10H8. Since thefluorescence cross-section s are many orders of magnitudegreater than those for Raman scattering, fluorescencedetection of minor species concentration is currentlyreceiving considerable attention for combustiond iagnost ics

.

CAPS is a nonlinear optics process that recently hasbeen shown to be of great promise as a combustion probe.CARS is a three-wave mixing process., requiring two lightsources, one of adjustable frequency. For the purpose ofexplanation, one beam will be referred to as the pump beamat frequency w and the other as Stokes beam wa . These twobeams are focused colinearly into the sample volume. Whenthe frequency difference between the two beams is equal toa Raman shift for a gas in the sample volume, a coherentanti-Stokes Raman beam of frequency w

ff' is produced by the

interaction with that gas.

The scattered beam is almost parallel to the incidentbeam and is many orders of magnitude more intense thanspontaneous Raman scattering. The CARS signal, however, isno longer a simple linear dependence on laser power andnumber density, as is spontaneous Raman scattering.Instead, the signal is proportional to Stokes laser power,the square of the pump laser power, and the square of thedensity. This nonlinearity in functional dependence oninput conditions places potential limitations on theapplication of CARS to many practical combustionenvironments. The work by Taran and his coworkers has

19

addressed many of these limitations and capabilities ofCARS and has been the pioneering effort to make CARS a

practical system for combustion research. CARS appears tobe very good for thermometry and for major speciesdetection (>1000 ppm) . For minor species, however, theprobability of success is much poorer.

SARS and RIKES are other nonlinear techniques whichoffer relief from the intrinsic background problems ofCARS. Also in contrast to CARS, these techniques do notrequire phase matching and the spectral output is similarto spontaneous Raman scattering and thus relatively easy tointerpret. However, for both of these systems thedetection apparatus can be very complex. These techniquesare still in their infancy and need considerabledevelopment before their potential can be betterunderstood

.

For each of the techniques mentioned above, limitationsexist which are yet to be completely resolved.Nonetheless, over the last few years, their range ofapplicability has been demonstrated over an increasingnumber of problems, and these gains are illuminated in therecent reviews referenced herein. In general, thesesystems are expensive ($50K to 200K) and require a traineduser. The techniques are not yet developed to the point ofcasual application and this may never be the case.However, for the interested researcher in chemicalkinetics, the problems (and risks) of setting up one ofthese techniques for his use are significantly lower than a

few years ago. Today a number of integrated systems are inuse which can be used as models and the number ofknowledgeable users is greatly increased.

The Combustion Diagnostics Facility at Sandia is beingorganized and equipped to provide training on these opticaltechniques, to improve and simplify integrated diagnosticsystems, and to make various techniques andstate-of-the-art equipment available to users from thecombustion research community.

2 0

REFERENCES

1. Hartley, D. L. , Hardesty, D. R. , Lapp, M. , Dooher , J.,and Dryer, F., "The Role of Physics in Combustion,"Efficient Use of Energy , AIP ConferenceProceedings , No. 25, 1975, New York.

2. Goulard, R. (ed.), Combustion Measurements , ModernTechniques and Instrumentation , a Project SquidWorkshop, Academic Press, 1976.

3. Zinn, B. (ed.), Experimental Diagnostics in GasPhase Combustion , Progress in Astronautics andAeronautics, Vol. 53, New York, 1977.

4. Eckbreth, A. C. , Bonczyk, P. A., and Verdieck, J. R.

,

Review of Laser Raman and Fluorescence Techniquesfor Practical Combustion Diagnostics , UnitedTechnologies Research Center Report UTRC-R77

,

952665-5, 1977.

5. Coleman, H. W. , Hardesty, D. R., Cattolica, R. J.,Pohl , J. H., Rahn, L. A., Hill, R. A., andHartley, D. L., Interim Report: DiagnosticsAssessment for Advanced Power Systems , SandiaReport SAND77-8216, 1977.

21

COMBUSTION CHEMISTRY AND MECHANISMS

David M. Golden

SRI International

333 Ravenswood Avenue, Menlo Park, CA 94025

ABSTRACT

Understanding of the chemistry involved in combustionrequires the general understanding of mechanisms ofpyrolysis and oxidation. The object is to write a

mechanism and to evaluate individual rate constants withinsome framework. Hopefully, this will eventually facilitatethe generalization of such mechanisms for efficacy inmodeling

.

The framework most often employed is that of transitionstate theory. In this paper we will explore the validityand limitations of the use of transition state theory as aframework and suggest some possible experimental probes ofthe questions raised.

Experiments have been performed for many years todetermine the temperature (and pressure) dependence of rateconstants for elementary chemical reactions that arepostulated to be important in pyrolysis and oxidation ofhydrocarbons. Even though more sophisticated experimentaltechniques continue to become available, it will never bepossible to measure all possible reactions under allpossible values of temperature and pressure. Thisimpossibility will not present us with insurmountabledifficulties if we have a fundamental understanding of thenature of thermal elementary reactions which allows us toextrapolate data for specific reactions to temperature (andpressure) ranges which have not been measured, as well asto other similar reactions.

A good deal of success in establishing a framework forcodification and extrapolation of rate data has beenachieved with transition state theory. I will review someof the many examples where transition state theory servesas an adequate framework for rate data analysis. This willinclude the prediction of rate constants and their pressureand temperature dependence. We will see that critical dataanalysis based on transition state theory considerationshas been helpful in calling our attention to systematic

22

problems with certain experiments. Particular attentionwill be paid to entropy and heat capacity of activation asa criterion for understanding the details of a chemicalreaction.

Thus, for example, we will review the fact that the"A-factors" for atom metathesis reactions, often availableover a modest (~300-800 K) temperature range decrease for a

given substrate as the reactive radical becomesprogressively larger, reflecting the loss of entropy due toexternal degrees of freedom.

Reaction log [A/M-ls-1]

H + C2H6 ~ 11

OH + C2H6 ~ 9.5

CH3 + C2H6 ~ 8.5

We will also recall that abstraction of an H-atom toform a stabilized species reflects the smaller value ofentropy in the transition state.

CH3 + C3H6; log [A/M-ls-1] ~8.0

We will discuss other examples of validity of thisapproach, which will include some discussion of pressuredependence of rate constants as well.

Having established the validity of the thermochemicalkinetics approach, we will explore some limitations.Limitations include the fact that reported temperaturedependencies for simple H-atom metathesis reactions oftenexceed those predicted at temperatures greater than 1000 K.On the other hand, some specific examples have recentlybeen published of reactions with temperature dependenciesthat are less than predicted. Since some of these data areobtained in fairly complex experiments, there exists astrong need for direct measurements of selected reactionsover a long temperature range. Experiments should givedata about products as well as rate parameters becauseoften reactions may proceed by multiple pathways. This isparticularly possible at higher temperatures. Specificsuggestions as to needed data will be made.

In addition, many simple reactive systems have beenshown to proceed to specific product states. If this is

accepted as a general characteristic of elementaryprocesses, a legitimate question about the validity of a

23

statistical theory, such as transition state theory, israised. This question will also be discussed.

Finally, I will address the question of generalmechanisms for oxidation and pyrolysis of aliphatic andaromatic hydrocarbons.

24

HIGH TEMPERATURE CHEMICAL KINETICS DATA

David Garvin and Wing Tsang

National Bureau of Standards, Washington, DC 20234

ABSTRACT

A data base of reaction rate constants is needed foruse in combustion research. This data base must as a startconsist of rate constants for elementary thermal reactions.Data on single step reactions are applicable to all systemsthat may be studied whereas global rate constants areuseful only in specific systems and are not transferable.Global rate constants can and should be developed from theelementary rate data base and the applicable mechanisms.State-to-state and molecular beam experiments can givevaluable mechanistic information, but, today, crosssections derived from them are not easily usable incombustion problems. In time they can be folded in andguide the study of non-equilibrium aspects of combustion.

There has been a modest decline in the total number ofgas kinetics measurements in recent years. But hightemperature measurements (above 500 K) have decreased moresharply, limiting the growth of the combustion data base.Low temperature measurements have increased in recent yearsas pollution and atmospheric chemistry problems have cometo the fore. One important result is that the hightemperature data base is old and items in it only rarelyare based on the modern higher precision experimentalmethods. Adaptation of these methods to high temperaturekinetics is needed.

Substantial numbers of rate constants gas phasereactions have been reviewed and recommended rate constantshave been published. Most of the reactions in the H2/02system have been reviewed. The CH4/02 system is alsoreasonably well covered.

Rate constants for the overall reactions of H , 0, OH,H02 and CH3 with organic molecules have been codified, butin some cases the data are sparse and only rarely have thebranching ratios been considered. Rate constants for theinitial steps of pyrolysis reactions are reasonably wellunder control although much of the data base is old. Thereviews should be updated. In contrast several classes ofreactions need extensive evaluation work. Reactions ofmolecular oxygen have not been evaluated systematically.Processes leading to carbon formation are poorly

25

understood. Radical-radical combination reactions deservemore work and are getting it - both experimentally andtheoretically. The outlook is promising. Radicald ispropor tionation reactions are much less well understoodand have not been systematized. The combustion scientistfaces three problems in using these codified data: theyare reported in many sources, they are often based on datathat have large error ranges and, inevitably, exactly thereaction for which data are needed today are missing fromthe collections. Programs are needed that will pulltogether the scattered recommendations, that will providehigher precision measurements on key (sensitive) reactionsand that will fill gaps in the data base by the use ofestimation techniques.

It is difficult for the kineticist to determine whichmeasurements should have high priority. Several approachesare desirable. Reactions that delineate combustionmechanisms should be studied. Even qualitative mechanisticinformation is very valuable. Reactions that make possiblethe development of correlations for classes of reactionsshould be emphasized. Rate constants for reactions towhich models are very sensitive should be measuredcarefully and remeasured if older data have large errors.

The usefulness of guidance from modelling studies isindicated above. But modelling of laboratory experimentson reaction kinetics also should be fostered. Most hightemperature experiments involve many elementary reactionsand require elaborate interpretation of data. This can bemade more reliable by modelling. The technique ofsensitivity analysis should make it possible to set morerealistic error limits for rate constants.

The art of data evaluation needs further development ifits output - recommended rate constants - is to be reliableenough for application to combustion models. Guidance isneeded from theory in order to improve judgements on thevalidity of data. Transition state theory and unimolecularrate theories have already proven their usefulness in theinterpretation of rate constants and should be pushed todevelop a framework for the techniques for extrapolation ofdata to different temperature and pressure regimes. Theyalso are the starting points for the establishment ofcorrelations among closely related reactions. Gaskineticists also have much to learn from the physicalorganic chemists.

26

Empirical and semiempir ical techniques for estimatingrate constants should be the subject of a major effort.They may be the answer to rapid expansion of the presentdata bank. The development of estimation techniques hasbeen an active although small part of kinetics for decades.Twenty five methods have been identified. The morepromising ones should be studied, validated, and theirranges of applicability determined.

While a kinetics data base for combustion studies willemphasize oxidation and pyrolysis reactions, these classesare too restrictive for present needs. Combustion kineticsis high temperature kinetics _in toto . One must be able tohandle energy release steps, to treat pollutant formationand control, to explain the behavior of additives and tomeet as yet unseen needs. In addition a wide range of hightemperature measurements - accurate ones - is needed toguide the theoretical developments and correlation workthat will bolster the art of data evaluation.

27

5. Discussion of Needs in High Temperature Kinetics Research

Discussion Leaders: F . L. Dryer and I. GlassmanPrinceton University, Princeton, NJ 08540

5.1 Outline for the Discussion

Many topics in this area that need to be considered bythe kinetics community are listed below. This outline isintended to serve as a guide for the discussion.

A. Aromatics Pyrolysis and Oxidation Kinetics Mechanisms

1. How are these different for benzene, alkylatedbenzene and polynuclear aromatics? The samequestion must be asked about the specific problemsin the following subsections.

2. Initiation and attack on aromatic ring (leading tofragmentation) . Different for premixed anddiffusion flames?

3. Chain branching steps - do the H2-02 branchingsteps play same role as in reactions of aliphatics.

4. Production of major oxidation products - Carboxylgroups form, how? How do they break away fromfragmented ( ?) ring, what are their fate?

5. What nitrogen fragments form and how, particularlyfrom coal at fluidized bed temperature, what is thefate of these nitrogen fragments, what are the othercompetitive steps for oxygen atoms?

6. Same as 5, above, for sulfur.

B. Soot and PCAH (polycyclic aromatic hydrocarbon)Formation

1. What are the elementary processes in the gas-phase,including nucleation.

2. Does the process change with the character of theflame (pre-mixed or diffusion) for aromatics?

3. What radicals attack soot precursors most readilyand how?

4. Kinetic mechanisms of soot inhibitors and promoters.

28

5. Is surface oxidation of soot and carbon (graphite)vastly different? Coal char? Does C02 form atall on carbonaceous surfaces? Can it be due tocatalytic effects?

6. What are the liquid-phase pyrolysis reactionsleading to cenospheres? How are they affected byimpurities and dissolved oxygen?

C. Overall Rate Expressions

1. Which systems hold promise for these short-cutprocedures?

2. What are the limitations of this approach?

D. Sulfur Oxidation Kinetics

1. In particular, what is the mechanism of oxidativeattack on organic sulfur leading to S02?

2. Why is S03 formation in real systems so sensitiveto stoichiometry?

E. Heterogeneous Kinetics

1. Quench layer kinetics - completely obscure orwell-known? Is it the region of odor (aldehyde)formation in diesels? What problems really existin sampling probes?

2. Of f-stoichiometr ic catalytic combustion fuel-oxygenreaction on surfaces - simply a mechanism forsupply of radicals to gas phase or a mechanismfor complete conversion to major products?

3. Catalytic surface combustion related to sulfur andnitrogen containing fuels.

F. Aliphatic Oxidation

1. Importance of H02 reactions in aliphatic and COoxidation

.

2. Rate data for H02 reactions.

3. More work on acetylene and olefin oxidation kineticmechanism at high temperatures.

4. Oxidation in very rich mixtures - effect of 02 andH20 on pyrolysis.

29

G. NOx Formation and Reduction

1 . Do we already know enough about NO formationreactions?

2. Should we not concentrate on mechanisms and ratesof NO reduction reactions?

H. Thermal and Oxidation Stability of Alternative Fuels

1, A very important medium temperature range kineticsproblem.

2. Kinetic mechanisms of in-situ combustion recoveryof shale oil and other thermal and chemicalsecondary and tertiary recovery schemes*

I. Chemical (Homogeneous and Heterogeneous) Inhibiton ofFlame Propagation

1. Lean-limit kinetics (fuel safety problems).

2. Anti-knock kinetics.

3. Effect of large halocarbons molecules on lean-limitand stoichiometric combustion.

30

5.2 Summary of the Discussion

P.L. Brown and R.F. Hampson

The session on the first day began with thepresentation of two invited papers [Chang and Westbrook,and Hartley] and a survey [Dryer and Glassman] on the needsin high temperature kinetics research. Aspects of all ofthese are pertinent to an understanding of the generaldiscussion and are reviewed here.*

In the first paper Westbrook reviewed the subject ofcomputer modeling of combustion systems. He pointed outthat chemical kinetics is but one subsystem of a combustionsystem involving a large variety of other closely coupledphysical subsystems. Modeling can be done on three levelsof sophistication - subsystem modeling, parameterizedmodeling, and system modeling. Combustion systems arecharacterized by energy production by chemical reactionsand the most distinctive feature is that chemical kineticsdrives the system dynamics. For some systems such as one-dimensional laminar flames, shock tubes and turbulent flowreactors, chemical kinetics is the dominant physicaloperator. He discussed the modeling of results reported byDryer and Glassman in the Fourteenth Symposium onCombustion for the oxidation of CO and CH4 using aturbulent flow reactor. (Reference 1) The agreement withexperimental data was excellent. He also discussed themodeling of recent results on CH30H oxidation. Theconclusion reached from this purely kinetics modeling isthat any reasonably sized mechanism (up to 500 reactionsinvolving 200 species) can be handled with present daycomputers and computer techniques. The number ofdifferential equations one must solve depends on the numberof species and not the number of reactions. Thecomputational problem increases dramatically when oneattempts to couple the kinetics with fluid mechanics. Forexample, a 2-dimensional flow problem using a 50 by 50computational grid, 5 hydrodynamic variables, and 5

chemical species requires 25000 differential equations tobe solved at each time step.

Statements made during the discussion are attributed toindividuals only if they submitted written comments. Forthose cases, references are given to numbered comments in

Section 7.

31

Some kind of kinetic simplification is needed to keepthe computer costs within reasonable bounds. Westbrookdiscussed a number of alternatives to the use of detailedchemical kinetic mechanisms. These include the use ofglobal type mechanisms. Other techniques use "lumping"methods, radical pool concepts and partial equilibriumcalculations. The principle under-lying these techniquesis the attempt to reduce the number of equations to besolved by replacing a differential equation by an algebraicequation. Another technique to reduce the number ofequations is partial decoupling of the mechanism for tracespecies. He discussed a number of representative problemswhere chemical kinetics are crucial to an understanding ofthe physical system and where the solution to the problemcan be said to be "kinetics information limited". Theseinclude NOx and SOx formation, soot formation, catalyticcombustion, spark ignition processes, and quenching. Forthe application of modeling to practical combustion systemsa large extension of the data base is needed. There aremany important fuels, and the only one we have a goodmechanism for is methane. We need detailed chemicalkinetic data, reaction mechanisms, and thermo- chemicaldata for other fuels. Combustion problems cover anextremely wide temperature range; modelers much preferusing functional forms for temperature dependence ratherthan tabular data. Thermochemical data are often thehardest to obtain. In the general discussion, Westbrook(7-1) referred to the question of the coupling betweenkinetics and fluid flow and recommended including transportproperties in a kinetics data base.

Following this talk, an opinion was expressed thatmost chemical kinetics shown in detailed mechanisms isirrelevant, with most systems being sensitive to a very fewnumber of reactions. However, the majority feeling wasthat although a simplified mechanism can be devised todescribe many systems over narrow ranges of experimentalparameters, to describe them over extended ranges requiressome form of detailed mechanism. There was considered tobe a real need to understand chemical mechanisms and tomeasure rate constants for elementary reactions.

The relationship of elementary reaction rate data tothe modeling of practical combustion systems receivedfurther consideration during the general discussion period.Edelson (7-2) questioned what would be the impact of thebetter measurement of a particular rate constant in acomplex reaction scheme for a combustion model in which thechemistry is summarized in a few reactions. Engleman (7-3)

32

pointed out three areas in which chemical kinetics cancontribute to combustion research: defining the problem,expanding the data base, and arriving at a solution. Keck(7-4) recommended that laminar flame speed data be used asa critical test of the predictive capability of combustionmodels. It was pointed out that flame speed is directlyrelated to flame thickness and to quench layer thickness,and that flame speed measurements are needed for practicalfuels at high temperatures and high pressures. For thistopic also, the view was reiterated that we must understandthe basic mechanism and know the rate constants for theelementary reactions.

Dean (7-6) and Fontijn (7-8) addressed the generalquestion of the usefulness of kinetics research. They wereoptimistic. Kinetics data can find quick practicalapplication, as has been shown in rocket exhaust combustionstudies, and in atmospheric pollution. The kineticist hasa definite role to play. He studies the complexities ofthe system in order to guide the modeler in developingsimplified mechanisms. Also, with an effectivemodeler-kineticist feedback loop emphasis can be placed onimproving the critical kinetic data and the essentialmechanism

.

In the second invited talk, Hartley discussed advancesin laser diagnostics for combustion research. He pointedout that optical scattering techniques are useful for theirability to make time dependent measurements on smallregions of space without the disturbance normallyintroduced by physical probes. A number of techniques werediscussed to introduce chemical kineticists to thepossibilities of using them in their own research.

After this talk there was some discussion of whetheror not the information that could be obtained by thediagnostic techniques planned for the CombustionDiagnostics Facility would be sufficient to unravel thechemistry occurring in the system. Myerson (7-5)recommended that the techniques to be used in this facilityshould not be restricted to those using lasers, but thatother techniques, especially fast broad-spectrum infrareddiagnostics should also be used.

The general discussion led by Dryer and Glassmanfocussed on the research needs in high temperature kineticsresearch. Glassman opened the discussion by noting that we

know enough of the overall mechanistic characteristics ofmost fuels to be able to burn them efficiently. The newdimension arises from the environmental constraints that

33

now exist. Factors such as sulfur and nitrogen emission,and soot formation have become major concerns in combustionresearch. A discussion outline was then presented in anattempt to concentrate attention on combustion problemsconsidered by the discussion leaders to be important. (Thedetailed outline is printed immediately before thissummary.) These current combustion problems werecategorized as follows;

1. Aromatics - pyrolysis and oxidation.2. Soot and polycyclic aromatic hydrocarbon formation.3. Heterogeneous kinetics, quenching, probes, catalysis4. Alternative fuels - thermal and oxidative stability.5. Sulfur oxidation kinetics.6. NOx formation - reactions involving fuels.7. Aliphatic oxidation - mechanism and rates.8. Ignition, initiation and induction periods.

Apart from the topics of NOx chemistry and aliphaticoxidation, these are topics that have not been studied indetail by many kineticists. The main result of thediscussion was to alert the kinetics community to the needfor work in these areas.

It was generally considered that, at present, it isnot possible to predict the kinetics of aromatic systemsand that working out the details of the mechanism andkinetics will take perhaps a decade. However, it was feltthat sufficient experimental tools are available to beginattacking the problem.

A generally similar situation was considered to existin the area of soot formation. There are not reliable ratedata for nucleation reactions. But there are strongsimilarities in the oxidation of soot and graphite. It waspointed out that in the combustion of some polymers, sootformation is reduced upon addition of transition metaloxides. The mechanism is unknown, but is beinginvestigated. Fontijn (7-7) pointed out that sootformation is being studied and listed severalinvestigators

.

There was very little discussion of the subject ofheterogeneous kinetics, although it was felt to beimportant to study heterogeneous reactions so that theireffects could be differentiated from homogeneous reactionsin experimental chemical systems and practical combustors.

Also there was very little discussion of the subjectof alternative fuels. The point raised by the discussionleaders was that fuels derived from oil should show changesin viscosity and even solidification during storage.

The current status of our knowledge of the lowtemperature sulfur oxidation reactions was summarized. Itwas pointed out that the oxidation of divalent sulfur by 0or OH begins with an addition mechanism which is known onlyat low temperatures and cannot be extrapolated to e.g.2500 C. It was pointed out that the conversion of S02 toS03 is not yet understood.

NOx formation and reduction were the subject ofconsiderable discussion. Bowman (7-9) emphasized that NOxemissions . are kinetically limited and discussed the presentstate of knowledge of NO formation kinetics. He statedthat we know well enough the rate constants for thereactions forming NO from atmospheric N2 (the Zeldovichmechanism) and that any inadequacy in our ability topredict the formation of NO by this mechanism is related toour inability to specify adequately the environment(temperature, 0 atom concentration, etc.). However, thismechanism does not predict the behavior of NO formation infuel- rich flames or in fuels containing organic nitrogencompounds. He stressed the need for a coordinated researchprogram in the chemical kinetics of conversion of fuelnitrogen to NOx and pointed out as specific research needsthe determination of the mechanism and kinetics of thefollowing reactions: reactions of hydrocarbon fragments(e.g., CH, CH2) with N2, the initial breakdown offuel-nitrogen species, reactions of N-containingintermediate species (HCN, NH, NH2, CN) found in flames,reactions of NO with N-containing intermediates and fuelfragments, and lastly the suggested interaction between NOxand SOx in flames.

It was suggested that the possible attack onatmospheric N2 by the radical C2H must also be consideredon the basis of a revised value of its heat of formationand newer data on the concentration of C2H2.

A recent study correlating the concentration of HCN in

fuel-rich flames with the concentration of CH and CH2 wasmentioned. McLean (7-10) noted that studies of theproduction of NO from fuel nitrogen have shown NO formationand destruction reactions to be simultaneously important.

35

Myer son (7-11) pointed out that the rapid reactions ofthese same hydrocarbon fragments with the NO produced mustalso be considered. Laufer (7-12) discussed results of a

recent study of the reaction of CH2 with N2 indicating thatformation of the spectroscopically observed HCN mustproceed through a "long-lived" intermediate and is not theresult of a simple reaction scheme involving N2 and eitherCH or CH2 . In this connection, Lin (7-13) pointed out thatalthough the direct reaction of CH with N2 to give HCN + Nis "spin-hindered", this hindrance is expected to beconsiderably less if the intermediate complex is stable andlong-lived , analogous to some 0 (ID) quenching reactions.Dorko (7-14) stated that research on organic nitrogencompounds is needed in order to devise methods forbypassing purification procedures.

It was generally agreed that the subject of aliphaticoxidation is not well understood; we don't even completelyunderstand methane oxidation. In particular, the nature offormaldehyde dissociation, and methyl radical reactionswith molecular oxygen needs clarification. Gelinas (7-15)felt that there is a need for simultaneous measurement ofall possible flame species to yield sufficient informationto allow discrimination along various chemical mechanisms.

Gann (7-16) discussed the effects of additives oncombustion systems with emphasis on flame inhibitionstudies involving halogens and metal atoms.

In summary the main result of the first day'sdiscussion was to point out to the kinetics community theneed for fundamental research in areas not yet studied indetail

.

References

1. Dryer, F.L. and Glassman, I. "High-Temperature OxidationCO and CH4 " Symp. Comb. 14th (Combustion Institute,Pittsburgh, 1973) 987-1001.

36

6. DISCUSSION OF ACQUISITION OF HIGH TEMPERATURE KINETIC DATA

Discussion Leaders: S.W. Benson, Univ. of So. CaliforniaD. Gutman, Illinois Institute of Tech.

6 . 1 Outline for the Discussion

The combustion of fossil fuels probably involves everycategory of elementary gas-phase reactions known as well asnumerous heterogeneous processes. The successful modelingof practical combustion systems requires three kinds ofinformation. First, a broad pool of general knowledge onmany classes of elementary reactions is essential to have a

reasonable basis for deciding which kinds of reactionsshould be included in the mechanism of the process understudy. Second, thermochemical information on reactants,intermediates, and products is needed to calculateequilibrium constants and estimate rate constants. Thisinformation can be used to identify those reactions in themechanism of the process under study for which accurateArrhenius parameters are needed for reliable predictivepurposes. Thirdly, for key reactions, accurate rateconstants are needed at all temperatures occurring duringthe combustion process. Knowledge of mechanism changeswith temperature as well as pressure and temperaturechanges in the rate constants of important unimolecularprocesses must also be available.

As a framework for the discussion on Acquisition ofHigh Temperature Kinetic Data it would be useful to discusschemical reactions important in combustion by classesdefined more or less by the methods used to study themsince future studies of reactions within each group willface similar experimental problems and limitations.Questions which will be raised and hopefully answered foreach category of reaction should include the following:

1. Among the current means of isolating the reaction ofinterest for study, which provide the most useful andreliable rate constants and mechanistic information?

2. How can current methods of studying these elementaryreactions be improved to be either more reliable ormore accurate?

3. How can new and emerging analytical methods andexperimental procedures be used to develop fresh approachesto studying these gas kinetic steps?

37

4. What thermochemical information is most needed tocalculate Ar rhenius parameters for this kind of reactionand how can it be obtained?

5. What specific reactions might have a higher priorityfor study in order to obtain the most useful informationfor practical modeling of combustion processes?

6. What value do studies performed below 500 K have forpredicting mechanisms and rate constants at combustiontemperatures (say 2500 K)

?

7. What are "areas of exploration" which might offer newinsights into various aspects of combustion includingignition, rate of heat release, formation of majorpollutants (such as NO, S02 , and soot) and the formationof trace amounts of highly undesirable products?

Classes of Reactions Important in Combustion Processes

1. HIGH ACTIVATION ENERGY REACTIONS (EA>15 kcal/mole) -

reactions with high activation energies controlignition delays, rates of branching, and changes ofmechanism with temperature,, In spite of theirimportance, little is known about many because suchreactions must be studied at very high temperatures,often under nonisothermal conditions, and usually insystems where secondary and tertiary reactions cannotbe suppressed . These reactions can be subdivided asfollows

:

A. PRIMARY INITIATION - In an oxygen rich environmentprimary production of free-radical intermediatesprobably proceeds to a great extent by the reaction,

Fuel + 02 -> 2R. (e.g. CH4 + 02 -> CH3 + H02)

It has not yet been possible to study such reactionsdirectly. What information is available is largelyderived from induction-time measurements inshock-heated fuel + 02 mixtures*

B. PYROLYSIS OF MOLECULAR REACTION INTERMEDIATES -

The pyrolysis of peroxides, aldehydes, and otheroxygen-containing products play key rate-determiningroles in low temperature combustion and in thetransiton to high-temperature burning. A considerable

33

amount of information is available from bulbexperiments on the pyrolysis of larger organicmolecules, and from shock-tube studies on thepyrolysis of H202 and simple hydrocarbons. Gaps inour knowledge should be identified, and the role ofabsolute rate theory calculations of rate constantsshould receive special attention.

C. PYROLYSIS OF FREE-RADICALS - The pyrolysis ofchain- carriers is largely responsible for thechange in combustion mechanism with temperature.Although there is some knowledge of the mechanisms,branching ratios and rate constants for thepyrolysis of alkyl radical little quantitativeinformation is available on the rate constants forthe pyrolysis of peroxy radicals (e.g. CH302 ->

CH3 + 02) , acyl radicals and alkoxy radicals. Formany of these intermediates internal rearrangementsoccur at elevated temperatures complicating thekinetics picture.

Virtually all our limited knowledge of the rateconstants for these processes come from indirectobservations. Should special attention be directedat developing methods of detecting and directlymeasuring the rate constants of reactions involvingpolyatomic free radicals?

D. ATOM TRANSFER REACTIONS INVOLVING ALKYL RADICALSwith 02 - Although reactions of the type,

R. + 02 -> olefin + H02, and

R. + 02 -> RO. + 0

are known to be important in high-temperaturecombustion, and it is known that rate constants forthese reactions are essential for combustionmodeling, still essentially no quantitativeinformation exists for either of these processes.What methods might be applied to isolate theseprocesses for quantitative study?

2. MODERATE ACTIVATION ENERGY REACTIONS (2 kcal/mole<EA<10kcal/mole - This category includes the chain-carryingtransfer reactions responsible for heat release, somechain branching, and product formation.

3^

A considerable body of information exists onatom-exchange involving H, and 0 atoms due to the easein generating them under controlled conditions, theability to detect these atoms at low concentrations andthe fact that they can be studied in pyrex or quartzvessels or in flow reacttors. An increasing body ofequally accurate information on OH-r ad icals is becomingavailable

.

Most rate constant measurements do not extendabove 500 K. Should more attention be directed atdeveloping extrapolation methods to obtainhigh-temperature rate constants? How can accurate rateconstants be measured above 500 K.

There is still a dearth of quantitative knowledgeon the atom-transfer reactions of the more complexchain carriers present in the low-temperaturecombustion zone, reactions of alkoxy, alkylperoxy, andacyl radicals and to a lesser extent the reactions ofalkyl and hydroperoxy radicals. The problem is, ofcourse, that it is difficult to produce theseintermediates under controlled conditions and also todetect them at low concentrations. Again an importantquestion is what special methods can be developed tostudy reactions of polyatomic free radicals moredirectly and more quantitatively?

ADDITION AND RECOMBINATION REACTIONS (0 kcal/mole)<EA<5kcal/mole) - addition reactions such as,

R + 02 -> R02, and

H;0;OH;R + alkenes , alkynes <-> Adduct -> Products

are important radical-chain processes and arerelatively easy to study even at room temperaturebecause of their low activation energies. Studies of 0and OH reactions with unsaturated hydrocarbons revealthe energy rich adduct can decompose in many ways,making identification of reactive routes anddetermination of branching ratios important. ChangingArrhenius parameters with temperature suggest mechanismchanges as the temperature is raised. It would beuseful to study the mechanism and rate constant changesof representative reactions from room temperature to2500 K to provide accurate data on such transitions.

40

Hopefully the behavior of analogous reactions couldthen be calculated from the knowledge gained from suchstud ies

.

There is a large body of knowledge on therecombination rates of many free radicals, but lesserknowledge of disproportionation branching ratios andtheir temperature dependencies. Here again newexperimental approches are needed to extend the body ofknowledge in this area.

HETEROGENEOUS REACTIONS - The loss of free-radicals onaerosol particles early in combustion and on sootduring late combustion deserves discussion. If theseprocesses are important, how can they be quantitativelystudied?

41

6.2 Summary of the Discussion

M . J. Kurylo and W. Tsang

The general tenor of the discussion during the secondday on aquisition of kinetics data was dictated by the twoinvited papers [Golden, and Garvin and Tsang] and a

detailed preview by the discussion leaders [Benson andGutman] on the important classes of combustion reactionswhich must be attacked, In the following we summarizethese talks and follow with the main points raised duringthe discussions.*

Golden began his review by demonstrating the immensenumber of reaction processes that must be known if a

complete kinetic model of combustion processes is to beconstructed. Exact measurement of all the reactions isclearly impossible within realistic time constraints. Thusalgorithms for scaling, extrapolation, or prediction ofrate parameters across reaction conditions and classes ofreactants provide the only possible means of satisfying theambitious data needs. Transition state theory usingthermochemical kinetics as a calculational tool appears atpresent to be the only method which gives a prescriptionfor constructing such rules. Golden illustrated itsutility with applications to systems of importance incombustion contexts. He emphasized the need for supportingthis approach through improved measurements onthermochemical properties of free radicals, radicaltermination processes, NOx formation reactions etc. Goldenpointed out that although transition state theory is anequilibrium kinetic theory, PPKM calculations provide abasis for estimating departures from equilibrium behaviorwhen energy-exchange effects are manifested in fall-offsfrom unimolecular behavior or in the reverse bimolecularprocess. He demonstrated the power of this approach withexamples from atmospheric reactions and indicated that atthe higher temperatures in combustion systems such effectscould be expected to be much more important. Keyparameters for such calculations were third bodyefficiencies and the structure of the transition states.Developments in these areas were summarized.

Statements made during the discussion are attributed toindividuals only if they submitted written comments. Forthose cases, references are given to numbered comments inSection 7.

42

In the subsequent discussions the equilibrium nature oftransition state theory was reiterated by several speakers.Kwei (7-17, 7-18) pointed out that inasmuch as combustionsystems frequently display large departures from thermalequilibrium the entire basis for the use of suchcalculations under these conditions must be regarded withcaution. Furthermore he noted that although transitionstate theory can frequently fit experimental data,questions regarding applicability and uniqueness remain.Thus uncertainties regarding predictive and extr apolativepowers must still be considered. He suggested molecularbeam experiments as a means of settling ambiguous points.In a similar vein the possibility of checking transitionstate predictions with the results of quantum mechanicalcalculations was presented and mention was made of suchefforts on the reactions 0 + CH4 and OH + CO. Shaw (7-19)on the basis of his experience in correlating data onradical abstraction process recommended an empirical raterelation. Unfortunately, this formulation was at variancewith transition state predictions. Skinner (7-20) wasdoubtful of the usefulness of the transition state approachas a predictive tool. Instead he saw a more limited roleas an interpola tive tool, especially among compounds of aparticular class. As such it might play a particularlyuseful role as a basis for characterizing complex systemsin terms of the reactivity of a small number of activegroups. For the advocates of transition state theory theclaim was made that entropies of activation could be soaccurately calculated that, given a rate constant, theactivation energy might frequently be calculated quiteaccurately. With respect to curvature in the Ar rheniusplot, the suggestion was made that, since rate constants atlower temperatures are more sensitive to this effect,measurement under such conditions (< ambient) shouldprovide a more unambiguous basis for extrapolation tohigher temperatures. This is undoubtedly true, however forthe large temperature ranges to be covered changes inreaction mechanisms could bring in other processes.

Finally, there was a discussion on the format forpresenting the results of RRKM calculations. The generalconclusion was that multiparameter fits can be easily made.However modelers indicated a preference for simple functionstrictly defined within limits of applicability.

Garvin and Tsang (in their paper) defined the mostimportant elements of a combustion kinetic data base as therate constants for elementary gas phase reactions involvingthe pyrolysis and oxidation of hydrocarbons and assortedpollutant related compounds at elevated temperatures.

43

Concentration on this type of data was due to thetransferability of the information, the ubiquity of theprocesses in all combustion contexts and the existence ofconsiderable information. They recognized that a completedata base should also contain state to state, heterogeneousand condensed phase data. However the underdevelopednature of these subjects rendered their inclusiondifficult. Global reactions should be derived fromelementary rates. Their non-tr ansfer abil ity suggested thatinclusion in a general data base would not be worthwhile.Statistics on the production of high temperature kineticsdata showed a constant or slightly declining rate duringthe past decade. This was in contrast with the rapidincrease in measurements near ambient conditions (frompollution concerns) . The successes in this area suggestedthat applications of the newer techniques to hightemperature kinetics with a corresponding increase inactivity would yield important information. Sources ofevaluated kinetics data were listed and the advantages of a

sole-source data base indicated. In agreement with Golden,they noted the immense volume of data needed and suggestedstrong experimental efforts on those reactions ofimportance in predictive and correlative schemes. Avariety of these schemes was summarized. With respect todata accuracy, the importance of base point reactions inrelative rate determinations was recognized andrecommendations were made for definitive studies of theserates. The possible contributions from chemical kineticsmodeling toward the building of the data base werediscussed. These involved the identification of keyprocesses and the unraveling of complex experiments. Inboth cases sensitivity analysis would play a key role.

The subsequent discussions focused principally on theproper use of chemical kinetics data in modeling. Many ofthe comments were based on participants' experiences withmodeling the methane oxidation system. Some specificproblems with respect to rates and mechanisms werementioned. There was disappointment with past progress inthis area and agreement that the blind use of extendedlists of rate constants would not not be useful[Bauer (7-21), Hyerson (7-22) and others]. Keck (7-4)suggested that useful results are not achievable in

the near future, Bauer (7-21) and Gelinas (7-36) outlinedpossible procedures that might reduce the number ofreactions considered to those which were truly meaningful.The need for proper procedures for sensitivity analysiswas made clear. Gelinas (7-23) warned that reactionmechanisms may change with time regions. Indritz (7-24)pointed out that in eliminating a reaction one has set itsrate constant at zero. Myerson (7-22) deplored the all too

44

easy tendency to simply fit the data over limited ranges.Engleman (7-3) suggested that the kineticist's role shouldbe more than supplying rate constant and that they may beable to make important contributions to problemdefinitions. Thus there was general agreement that the"present Laissez-faire approach which required at least twodecades to home in on a reasonable mechanism for methane"Bauer (7-21) was unsatisfactory and there is a need for

procedures in data analysis.

With respect to methane - oxidation the possibleconsequence of Golden's statement in his earlierpresentation that he was unable to detect products from theCH3 + 02 reaction under Very Low Pressure Pyrolysisconditions at 1000 K was challenged by several participantson the basis of shock tube data. Possible means ofreconciling the apparent conflicts were suggested. Otherproblem areas in the data base for methane oxidationinclude formaldehyde decomposition and formyl radicalreactions. It was made clear that the quantitative detailsof the methane oxidation process remain uncertain. FinallyWestley (7-25) mentioned the uncertainty in data onpyrolysis brought about by possible presence of tracequantities of oxygen.

Prior to the discussion session on high temperaturedata acquisition, Benson gave a general review of theinterrelation of chemical kinetics, combustion and theenergy crisis. In particular he highlighted the possiblekey roles of alkylperoxy radicals in flame propagation.

Gutman continued the discussion preview by presenting a

systematic picture of the various classes of combustionrelated reactions that must be studied (i.e. initiation,branching, propagation, and termination). In each of theclasses, the existing uncertainties far outweigh the knownswhen we depart from the old familiar oxygen-hydrogen andoxygen- methane systems. He suggested, however, a varietyof experimental tools which may prove useful for solvingthese difficult problems. Particular emphasis was placedon laser-based techniques and their possible role in thethe generation, detection, and kinetic measurement ofunstable polyatomic species. He challenged theexperimentalists to apply the new methodology not only torate constant measurement but to mechanistic delineation aswell. Here again particular emphasis was placed on thepresent lack of information (or means to obtain it) onorgano-peroxy and organo-oxy radicals.

45

Gutman stressed the importance of a careful mechanisticsensitivity analysis in the interpretation of complexexperimental results. The use of a common data base forsuch calculation was mentioned as being a particularnecessity.

Workshop participants took up the challenge of Bensonand Gutmann and presented a variety of technologies whichthey were presently using or proposed as useful forperforming high temperature combustion diagnostics and/orkinetic measurements. Many of these replies appear aswritten comments by the participants and are cited below.

Bauer (7-26) described some recent experimentsemploying a laser- Schlieren technique (post shock frontdensity gradient technique) for making measurements onshock front gradients during the dark (induction) periodsof hydrocarbon pyrolysis and oxidation. Using this method,one examines the exo- , endo- thermic balance immediatelyafter the shock has passed through the system. Acautionary note was added that one should be concerned withgas dynamic instability due to very exothermic processes.Other comments affirmed the presently accurate temperaturecharacterization in shock heated systems. Keck suggested a

similar type of measurement facility in the sphericalcombustion bomb. He outlined the use of pressuretransducers, ionization probes and laser diagnostics toobserve flame front profiles following ignition. Hefurther stressed (7-27) the importance of makingmeasurements at high gas densities where many of the ratecontrolling processes change from tr imolecular tobimolecul ar in character.

Tsang's recommendation (7-37) that attempts be made atminimizing the number of ongoing reactions in such systemsto simplify interpretation underscored his advocacy of thesingle pulse shock tube technique. With such an apparatus,all but the initial reaction can be eliminated (by lowreactant densities) thereby yielding very detailedmechanistic information. He suggested that under welldefined initial conditions, one can "bootstrap" his waythrough the various stages of complexity.

These same preferences were echoed by Myer son (7-28)who envisioned the single pulse shock tube as an excellenttechnique for obtaining the needed kinetic data onpyrolysis and combustion of aromatic and coal- relevant

46

molecules. In addition, he cautioned the researchers notto abandon these older methods of combustion analysis forthe alluring charm of total laser diagnostics, but ratherapproach the problems with the full armament of measuringtechniques

.

Kern offered still another shock tube-associatedtechnique in the use of time of flight mass spec trometr icsampling of reflected shock zones. Such experiments havehad applicability in the measurement of precursors to sootformation and the identification of chemi-ions.

Skinner (7-29, 7-30) presented an extensive list ofmeasurement techniaues and discussed their applicabilityfor combustion studies. More specifically he, too,advocated the use of shock tube studies in order to work upgradually in the complexity of systems thereby allowingmodelers to check out a complete kinetic subsystem apartfrom transport phenomena.

Eogan (7-31) and Seery (7-32) pointed out theversatility of molecular beam sampling mass spectrometryfor probing both stable and unstable species in hightemperature combustion systems. Such an apparatus coupledwith optical diagnostic techniques appears useful inprobing complex systems as well as studying isolatedreaction processes.

Benson briefly outlined recent experiments performedusing a very low pressure pyrolysis reactor for controlledfree radical production. While limited in temperaturerange, this technique promised to be useful in getting somefirst data on a number of radical-radical interactions aswell as radial heats of formation.

Brink's comment (7-33) solicited the inclusion of dyelaser intracavity absorption as possible diagnostic forstudying combustion reactions. Its high sensitivity andgood spectral fidelity make it a viable candidate.

A still newer laser technique was described by Huie(7-34) who has utilized infrared mul tiphotonphotodissociation to produce cleanly organic radicals (viz.CF2) . Combination of this production technique with mass

47

spectrome tr ic detection should prove useful in planningcontrolled radical kinetic studies at high temperature.

Livingston related some recent sucess in using ESR todetect short lived free radicals produced thermally influids. Present capabilities allow for pressures up to 75atmospheres and temperatures approaching 800 K.

A planned extension of the flash photolysis resonancefluorescent techniques to nearly 2000 K was described byFontijn (7-35) . The original technique has been proven tobe quite versatile by many workers in measuring radical andatom reactions from 170 K to 500 K using real timefluorescence detection. Incorporation of this typefacility in a furnace should permit for the first time thedirect measurement of many elementary reactions attemperatures of interest to combustion modeling.

Along similar lines, the usefulness of pulsedradiolysis techniques to produce free radicals incontrolled kinetics experiments was related andmeasurements made using this technique were summarized(7-38). Present capabilities allow pressures of up toseveral atmospheres. It is planned to extend thetemperature capability to at least 1000 K.

48

7 . Contributions Submitted by Participants

The participants at the meeting were encouraged towrite down comments they made at the meeting and were giventhe opportunity to submit others afterwards. Many did so.These comments are printed in this section, aproximately inthe order in which they are referred to in the summaries ofthe discussions, sections 5.2 and 6.2. In general, thismeans that comments on particular topics appear near eachother. Comment references in the Summaries lead the readerto this section. An author index for this section followsthe comments.

49

7-1

Topic: Necessary Elementary Data for Modeling.

C. Westbrook, Lawrence Livermore Laboratory. Many kineticsphenomena may be modeled from elementary rate data alone,but there are many more which are characterized by a closecoupling between kinetic and spatial transport processes.In such environments observable results depend upon singleparticle transport rates through mul ticomponent gasmixtures. There is a wealth of kinetic data, which isavailable from good flame experimental data, but to beapproached by modeling these transport coefficients must beknown to within the same uncertainties as the rate data, ifnot better. I suggest that a proper function of a database in this area would be to include data and evaluationson binary diffusion coefficients for different chemicalspecies. Modelers can use this data to estimatemulticomponent diffusion coefficients if necessary, but thebinary diffusion coefficients into a variety of componentspecies are needed in these transport models. Althoughthese coefficients can be estimated from kinetic theoryarguments, it is likely that molecules with many internaldegrees of freedom may not diffuse in a simple way, so thatexperimental data are needed, and these data should beevaluated by someone who can do so properly.

7-2

Topic: Needs for High Temperature Kinetics Research.

D. Edelson , Bell Laboratories . One of the major areas ofuncertainty is the relationship between fundamentalchemical kinetics on the one hand and the state of the artin computer simulation of combustion systems on the other.At the present time the latter is limited to theconsideration of "global" (lumped) chemical models becauseof the prior necessity of treating the very complex fluidand thermal transport of any real combustion system.

Because the chemistry provides the driving force for thefluid dynamics, it is not clear whether a separation of thechemical and transport computations is possible, sinceneither can be considered a small perturbation on theother .

In this light, it is very difficult to assess the impactwhich the detailed study of any particular phase of thefundamental chemistry can have on the overall problem.Beyond a qualitative grasp of the important features of thechemical mechanism, i.e., whether a class of reactions doesor does not take place, it becomes questionable whether

50

efforts at quantifying particular steps is worthwhile.

Advances in the techniques of modeling these complicatedprocesses are required if these fragments are to beassembled into a coherent picture. New machines, and newmethods in numerical mathematics are becoming availablewhich could profitably assist in resolving thesedifficulties. Efforts should be made at graduallysupplanting the present use of simplification andapproximation methods with more exact calculations. Asthis is done, then a complementary requirement for improvedkinetic data will become apparent, and will point thedirection for future research.

7-3

Topic: Needs in High Temperature Kinetics Research.

V. Engleman , Science Appl ications Inc . ChemicalKTneticists can make valuable contributions tocombustion research in three areas:

(1) Defining the problem - identifying the keyphenomena and establishing a plausible mechanism.

(2) Expanding the data base - sensitivity studiesand obtaining improved kinetic information on key reactions.

(3) Arriving at a solution - narrowing down themechanism to a tractable size and perhapsarriving at meaningful global or semi-globalrate expressions, recognizing that this step isproblem dependent .

Specific combustion problems are at different stages ofneed. Coal combustion/gasification is still at stage 1.

Methane/air combustion is generally at an advanced stage 2.

And portions of the methane/air combustion problem areready for stage 3.

The kineticist should play a key role in the problemdefinition (stage 1) and should work closely with themodeler, especially at that stage.

It should be recognized that stage 2 is a continuing effort(improving accuracy of rate constants) and can belong-term. Stage 3 should proceed during that process.

51

The chemical kinetics problem is not solely in determiningreaction rates even though that is an important activity.Overall problem definition and problem resolution need theguidance that can be provided by the kineticist.

7-4

Topic: Combustion Modelling

J. C. Keck , Massachusetts Institute of Technology. Amongthe most important combustion properties required forpractical applications in engines and combustors burningpremixed charges are the laminar flame speed, laminar flamethickness, wall quenching distance, ignition energy andspontaneous ignition limits at densities from roughly 1 to100 atmospheres. The prediction of those properties fromfirst principle is an exceedingly complex problem requiringboth a knowledge of a large number of ill defined reactionrates and the numerical solution of a large system ofcoupled rate and diffusion equations. The probability thatsuch a synthesis will lead to quantitatively useful (+ 20%)results in the near future is extremely remote. It istherefore important that the direct measurement of thefundamental combustion properties listed above be includedas an integral part of any basic chemical kinetics programrelated to combustion. This would serve the dual purposeof providing a critical test of the predictive capabilityof the synthetic approach as well as making immediatelyavailable essential information required for practicalapplications. In this regard it could be noted thatlaminar combustion properties are also useful in turbulentcombustion for predicting turbulent flame thickness,ignition energy and wall quenching distance, in cases wherethe turbulence scale is large compared to the laminar flamethickness.

7-5

Topic: Combustion Diagnostic Facility - Infrared et al

.

A. L. Myer son , Exxon . It is rather perplexing tounderstand why this facility which I understood was to befor general use, seems to be completely dominated byscattering techniques and/or those which use lasers. Withthe advent of a major stress on more complex hydrocarbons(aromatics, et al . , from coal etc.), and the availabilityof fast, broad spectrum infrared diagnostics, at least thelatter techniques should be a significant part of thefacility. Other techniques should include: ARAS, laserdensitometry, single-pulse shock tube analysis, etc.

52

7-6

Topic: Impact of Chemical Kinetics on Combustion Research.

A. M. Dean , University of Missour i-Columbia

.

Thekineticist can provide information about the elementarychemical reactions and the associated rate constants for a

particular combustion system. This information is vital tothe combustion modeler even though it is often too detailedfor practical combustor calculations. The importance ofthe mechanistic information is that, for a sufficientlydetailed mechanism, it should be applicable over a widerange of conditions. The modeler can then make thenecessary simplifications for his particular applicationand check the accuracy of his approximations against thecomplete mechanism. In this respect the role of thekineticist can be viewed as one who attempts to understandall the complexities of the system so that modelers canhave a guide as to what constitutes a valid simplificationfor a given situation.

7-7

Topic: Soot Formation.

A. Fontijn , AeroChem Research Labs . Activity in this fieldis too low relative to its importance, but it is incorrectto state that no kinetic studies are performed.

At AeroChein Calcote and Miller are studying ionicmechanisms in soot formation. Other important activitiesare going on; Dr. Jack Howard at MIT and Dr. BillShipman at Cabot Corporation (carbon black) can giveinformation on some of these.

7-8

Topic: When will the new kinetic data that we can generatebecome of practical use?*

A. Fonti jn , AeroChem Research Labs . Many pessimisticremarks on this subject have been made this afternoon.More optimistic notes were sounded by those involved in thestratospheric pollution problem. However, there is directevidence as well in the combustion area that the right typeof kinetic data can find quick practical application.Prime examples here are rocket exhaust combustion problems,a field of major interest ten to twenty years ago and on a

lesser scale still today. Modeling studies are a common•approach here and the interaction between the modelers andkineticists has paid off handsomely. The type of

53

interaction required is, of course: (i) kineticists maketheir best guesses as to rate coefficients (ii) modelersincorporate these and establish which reactions are themost important, (iii) kineticists measure these, and (iv)the improved model is used by design engineers and bykineticists to determine if, and which, further reactionsneed study.

7-9

Topic: Chemical Kinetics of NO Formation in Combustion

C. T. Bowman , Stanford University. Environmentalconsiderations place significant constraints on nitrogenoxide emissions from combustion systems. The trend towarduse of fuels containing significant quantities of organicnitrogen compounds will make control of NOx emissions moredifficult. To support development of NOx controltechniques, including combustion modification andcombustion gas clean-up by injection of reactive additives,a co-ordinated research program in chemical kinetics ofconversion of fuel nitrogen to nitrogen oxides is required.Some specific research needs include:

1. Determination of the mechanisms of reactions betweenhydrocarbon fragments and molecular nitrogen.What are the reaction products? What are therelevant rate coefficients?

2. Determination of the mechanisms of initialbreakdown of fuel nitrogen species. What are thereaction products? Does the relative importance ofvarious reaction paths change in the presence ofoxygen? Are there common intermediate species?

3. Determination of mechanisms and rates of reaction ofnitrogen-containing intermediate species found inflames (HCN, NH, NH2, CN) . These species appear toplay a significant role in the NO formation process.

4. Determination of mechanisms and rates of reactionsbetween NO and nitrogen-containng intermediatesand fuel fragments. These reactions play a majorrole in NO removal in the flame zone and post-combustion gas.

5. Study possible interactions between NOx - SOx in flames.There is some evidence to suggest that the presence ofsulfur compounds reduces NO formation from organicnitrogen compounds in flames.

54

7-10

Topic: NOx Formation and Reduction

W. J . McLean , Cornell University . The argument that in thefuel nitrogen problem we should emphasize studies of theNOx reduction reactions rather than the formation reactionsneeds some clarification. While it is true that NOx formsquite fast from fuel nitrogen, it is not necessarily truethat the formation and reduction processes can be separatedand studied independently. A number of studies reported inthe literature indicate that the fuel nitrogen mechanisminvolves NO in a dual competitive reaction path. That is,NO is simultaneously being formed from some nitrogeneousintermediate and converted in part to N2 by reaction withother single-N compounds. Our own studies of NO productionduring HCN oxidation also indicate this general behavior,and we find NO formation and destruction to besimultaneously important in determining the eventual fateof the nitrogen.

Studies of the combined mechanisms for NO production andremoval can assist combustion technologists in choosingoperating stoichiometr ies and temperatures which maximizeconversion of fuel nitrogen to N2.

7-11

Topic: Nitric Oxide Reactions for NO control.

A. L . Myerson , Exxon . Although we have discussed theimportance of measuring the chemical kinetics of nitricoxide formation and destruction, as needed for NO pollutioncontrol, no mention has been made of one of the mostimportant set of reactions in this respect. There is everyevidence that no model can be made of steady-state ornon-steady-state NO concentrations in hydrocarboncombustion, without considering the fast reaction of NOwith hydrocarbon free radicals characteristic of therelevant flames and explosion, viz.

CH + NO = CHO + NCH + NO = HCN + 0

These are exothermic, fast, are present both early and latein hydrocarbon combustion flames and depend on theeguivalence ratio for their degree of importance. They arealso of considerable importance in the internal combustionengines when NO reactions with wall hydrocarbon combustionfilms take place reducing the NO.

55

7-12

Topic: Formation of NOx in Fuel-Rich Systems.

A. Laufer}National Bureau of Standards . We have examined

the reaction of CH2 with N2 and directly observed theformation of HCN for which we obtained a temporal profile.The experiment involved the flash photolysis of CH2C0 andanalysis by vacuum ultraviolet absorption spectroscopy.HCN has a well-characterized spectrum in this region.

The major experimental observation is that HCN appears onlyafter several hundred milliseconds following the flash.The role of CH may be eliminated both on energetic andkinetic grounds. Energetically, by the incidentwavelength. Kinetically, if we use the slower of the twomeasured values for CH + N2, we find a 1/e time for theproduction of HCN of about 10 m sec - in disagreement withobservation. In a like manner, we may calculate thelifetime of either 1CH2 of 3CH2 in this system to be lessthan 1 millisecond. Since neither of the obviousprecursors can lead to HCN through a simple reactionscheme, we are required to postulate a "long-lived"intermediate which can slowly re-arrange qr decompose toHCN. A complete description of this wo ok will appearshortly in "Combustion and Flame".7-13

Topic: Mechanism of NOx Formation.

M. C . Lin , Naval Research Laboratory . With regard to themechanism of the CH + N2 reaction related to NOx formation,

I have the following comments: (a) Although the CH(X 2n) +

N 2(X 1

2g+

) -> HCN(X 12+

) + N

(

4S3 / 2 ) reaction is

"spin-hindered", the hindrance (or forbiddenness) isexpected to be considerably less if the reactionintermediate HC=N=N is stable and long-lived. The bestexamples of this are the relaxation of O^D^) atoms by N2and CO. Both reactions are known to occur readily (1)

,

although they are apparently "spin-forbidden". (b) We havepreviously shown (via laser emission and reaction productanalyses) that the flash photolysis of CHBr 3 below 200 nmgenerated CH in significant quantities . (2) We plan to studythe reactions of CH and C2H (from Br C2H ) with variousmolecular species of importance to combustion chemistry.An ArF laser (193 nm) will be used as a photolysis sourceand a dye laser for the detection of transient species.The conventional flash photolysis method is probably tooslow for the CH radical because of its great reactivity andthe initial long, blinding interference due to the flash.

56

1. R. F. Hampson and D. Garvin (Eds.)./ "Chemical Kineticand Photochemical Data from Modelling AtmosphericChemistry," NBS Technical Note No. 866, 1975.

2. M. C. Lin, Int. J. Chem. Kinet. 6, 1 (1974); J. Phys.Chem. 77, 2726 (1973); J. Chem. Phys. 61, 1835 (1974).

Topic: NOx Formation and Reduction.

E . A. Dor ko , Air Force Institute of Technology . Nitrogen-containing compounds . The mechanism, temperature, andpressure dependence of decomposition fragmentation andoxidation of nitrogen-containing organic compounds needsstudy to understand what occurs, in order to eliminate NOxpollutants without necessarily passing through a

purification procedure (which is expensive).

In addition in the study of oxidation of aromatics,pyridine needs to be studied as an analogous, aromatic,nitrogen- containing system. At this point very little isknown of these matters.

Topic: Simultaneous Measurements in Flames.

R. J . Gelinas , Sc ience Appl ications . Among very highpriority needs in combustion science are simultaneousmeasurement of major, intermediate, and minor species inflames, as a function of space and time. The emphasisshould be on tractable fuels in highly interpr etable simpleflame systems. I am aware of only the work of Biordi andof the group at Louvain approaching this objective. Suchwork is necessary to provide distinctive levels ofinformation content which will allow elementary physicaland chemical mechanisms to be studied in morediscriminating detail. In the absence of simultaneousmeasurements a superabundance of free parameters isunavoidable and the information content yields minimaldiscrimination of important detailed mechanisms.

7-14

7-15

57

7-16

Topic: Chemical Inhibition of Flame Propagation.

R. G« Gann , National Bureau of Standards

.

There is a needto know where the "soft" spots are in the combustionsequence, i.e., those places where a small quantity of anadditive can drastically affect the global rate of fuelconsumption, heat produced, etc. There is current evidencethat halogen or metal atoms can be additives of this type.

If we knew how they worked, we would know how to optimizethe selection of perturbing species. Thus the problemrequires bulk experiments to determine specificsuper-additives (for a given desired effect) and todetermine some character of the reactive processesinvolved: e.g., is the effect temperature or pressuredependent? What is the result of systematically varyingthe additive species, such as substituting a chlorine atomfor a bromine atom or a glycol for an alcohol? With theevidence, possible competing reactions can be identified,the kinetics measured, and the magnitude of the effectcorroborated (or not) by insertion into a detailed kineticmodel or again by checking relative effects of two or moreadditives. A specific example of this involves bromineinhibition of flames. Several studies have led to thehypothesis that termolecular recombination rates arecritical, and that H + Br + M -> HBr + M is a specific rateconstant for which we need better data to proceed further.Clearly more of this specific isolation study is needed.

In the flame inhibition field, numerous halogens, metalsand organometallics have been studied and the "winners"identified. Further such studies are in progress.

7-17

Topic: Role of non-thermal process in combustion.

G . H. Kwei , Los Alamos Scientific Laboratory

.

I would liketo offer a word of caution on the use of equilibrium rateconstants in the modeling of combustion processes. Sincedense mixtures of gases, many of which are reactive, areinvolved, it is not at all clear whether or notthermalization will precede reaction. Even ifthermalization is rapid, minor channels involvingnon-thermal processes could be very important in pollutantformation, soot formation, and efficient combustion; suchprocesses are clearly important in luminescence andion-formation in flames.

58

Although it would be difficult to incorporate morecomplicated kinetics in present modeling calculations, abetter understanding of the competition between energyexchange and chemical reaction and of the role ofnon-thermal processes is absolutely essential to ourunderstanding of combustion processes. If we do not havethis kind of information, we may be modeling the wrongsystem.

In order to meet this objective, we should identify a fewkey reactions and study (1) the dependence of the rate orcross section on reactant excitation (i.e.) undernon-thermal conditions) and (2) the rates or cross sectionsfor de-excitation in collisions with molecules that arepresent in the combustion system.

7-18

Topic: Reliability of "Thermochemical Kinetics".

G. H. Kwei , Los Alamos Scientific Laboratory. Although Dr.Golden's inclusion of heat capacity terms in thepr e-exponential factors has provided good fits to theexperimental data for several reactions, it certainly isnot clear that "thermochemical kinetics" would provideequally good fits for all chemical reactions or that thesetechniques are reliable in the extrapolation of lowtemperature data.

Dryer and Glassman have obtained an equally good fit forthe reaction of OH radicals with CO molecules by assumingthat the transition state has the same structure andvibrational frequencies as the stable HONO molecule. Thereare two problems with this treatment: (1) Both intuitiveelectronic structure theory and experimental data on thepressure dependence of the rate constant suggest that thisreaction proceeds via the formation and subsequentdecomposition of a long-lived complex and (2) ourqualitative understanding of the electronic structures ofHONO and HOCO suggest that neither the geometry nor thevibrational frequencies will be the same. Item (1)

suggests that transition state theory is not applicable anditem (2) suggests that even if it were applicable thequality of the fit will be affected. This raises thequestion of whether or not we are covering-up interestingand important features of the reaction dynamics by fittingthe data to transition state models.

59

We are currently planning a crossed molecular beam study ofthis reaction as part of a larger program to study thedynamics of reactions important in combustion processes andin atmospheric chemistry. Since both the collision energydependence and the reactant vibrational state dependence(for both reactants) of the reaction cross section will bedetermined, a much clearer understanding of this reactionwill soon be available.

7-19

Topic: Hydrogen Transfer Reactions.

R. Shaw , Lockheed Missiles and Space Company .

Kinetics Measurements .

Rate constants for the following reactions need to bemeasured above 2000K:

R + CH4 = RH + CH3

where R is H, O, HO, and 02.

Kinetics Estimates .

Existing, measured, rate constants need to be fittedto k = AT2 exp (-c/t) for the reactions:

R + R'H = RH + R' (2)

where R is H, 0, HO, aliphatic and aromatic and whereR' is aliphatic or aromatic.

Thermochemistry.

We need more work on the heat of formation of CH3.

7-20

Topic: Correlation of Theoretical and Experimental Studies.

G. Skinner , Wright State University . At our present stateof chemical knowledge it is unwise to use theoreticalmethods, such as Thermochemical Kinetics, RRKM theory, ormore detailed theories to make long extrapolations ofkinetic data (such as from experimental data at 300-600K upto 2000K) unless there are some experimental data at thehigh temperatures on reactions of a similar chemical typeagainst which the theoretical method can be tested. In

'

60

general, we should try to get experimental data on at leastone and preferably a few specific reactions which arerepresentative of various types (such as reactions of 0atoms with secondary hydrogen atoms on fairly largehydrocarbon molecules). Once this has been done, theorycan be used to obtain rate constants for other similarreactions, to fill in any unstudied temperature ranges, andto make moderate temperature extrapolations (such as 1500 -

2000K) .

This approach enables us to work on the most easily studiedmolecules of a particular type, and therefore, to get themost accurate data, even if the easy or least difficultmolecules are not those of greatest direct interest incombustion, since the theoretical methods can then fill inother molecules as needed.

It may well be possible to extend this method even furtherin computer modeling. If, as may be expected, the rateconstants of all reactions of a given type for largerhydrocarbon molecules are about the same, the modeler maynot have to consider concentrations of the large number ofdifferent molecules that may be present, but only thenumber of reaction centers of a given type. This maygreatly reduce the number of "species" and reactions. Anadditional parameter such as the average number of C atomsper molecule would complete an adequate kinetic descriptionof a collection of hundreds of different hydrocarbonmolecules, such as one finds in gasoline.

7-21

General Comment , S. H. Bauer .

There is agreement that were one to model combustionprocesses utilizing only one or two global reactions he mayarrive at misleading conclusions. At the other extreme,incorporating all of the conceivable reactions (hundreds)is definitely not practical. Professor J. Keek's rationalapproach merits careful scrutiny. One cannot justify onoperational grounds the use of extended lists of reactionswhen the amount of information provided by our experimentalprobes is inherently limited, so that there is no feedbackfrom the laboratory to the model builder. I find itdifficult to justify (relative to combustion theory) theexpenditure of great effort in obtaining for some selectedreactions very precise rate constants over limited rangesof pressure and temperature, particularly when thereactions and the conditions of operation are selected on

the basis of convenience. One should demonstrate that the

predictions for the system's behavior are sensitive to the

61

precise values. What is often missing is the priorselection of the reactions which should be included oromitted for each range of operations of the combustor.The crucial question which we should try to answer, isthere a better strategy available than the present laissezfaire approach, which required at least two decades tohome-in on a reasonable mechanism for methane? In myopinion the basic components for such a strategy areavailable but further work is required for theirintegration and optimization. I submit the following;

1. For any system it is possible to list all thesignificant participating species—atoms, radicals, andmolecules which could be present in substantial, small, orminute concentrations. This is based on our presentchemical knowledge, and general "intuition".

2. For most species one can trust currently availableempirical relations for estimating their enthalpies offormation and entropies.

3 . It is then a straightforward calculation to estimatethe equil ibr ium concentrations of the species listed in (1)[for any range of temperature and pressurej

.

4. An algorithm has been written (W. M. Shaub and S. H.Bauer: will appear in Combustion and Flame) for a computerto list all possible pairwise combinations (reactions) ofthe components on a list, so as to generate other membersin that list, subject to atom conservation. [Ternarycombinations of atoms or radicals appear as the inverse ofbinary dissociations]

.

5. For each reaction the corresponding AS,AH , AG arecalculated

.

6. All reactions for whichAG > a (some rational positivevalue) are removed from the list.

7. Other simple criteria for removing unlikely reactionshave been programmed, such asAS either very large or verynegative, etc.

However, beyond this stage we encounter the practicalcrunch; How to program a computer so as to reduce thenumber of reactions (on the basis of rational kineticcriteria ) from the remaining several hundred by a factor of10. Hence at this stage we need:

(a) Empirical procedures for estimating rate constants.These must include estimates of ranges of acceptablevalues. Professor Benson has proposed a rational approach.

62

(b) Selection of critical observables for each system.

(c) Mathematical procedures for testing the sensitivity ofthe observables to the reactions which remain afterapplying steps (6) and (7), subject to the estimates in(a) . For instance one may delete any reaction wherein theproduct of the rate constant by the equilibriumconcentrations of the reactants is below some lower setlimit. This can be justified on the basis of the generalobservation that the very reactive, unstable radicalsrapidly attain local-equilibrium levels, due to themultitude of chain reactions, and then drift along with theconcentrations of the major species.

Dr. Gelinas mentioned three procedures now available formaking such sensitivity tests. These should be explored indetail and further developed. Since I have not used them,it is not obvious to me that any of these would be suitablewhen the specified error limits are large for a substantialfraction of the rate constants as listed in (a).Nevertheless, I do believe that the problem is notinsoluble, provided each of us takes a hard boiled attituderegarding our favorite reactions: for any specified testwe should be ready to omit any reaction which does notsignificantly affect the predicted observables.

7-22

Topic: Computer Modelling of Chemical Combustion Systems.

A. L . Myerson , Exxon . I believe it is incumbent upon us toencourage those who mathematically model complex chemicalmechanisms of combustion to maintain a more restrictive andrealistic attitude. There is absolutely no justificationfor including 50, 100 or even more chemical mechanisticsteps, if 50% of them are uncertain by an order ofmagnitude, 25% by 1/2 an order of magnitude and perhaps 10%by only 50% or 100%. It is too easy to matchtime-dependent phenomena at one temperature. The rules ofthe game must be made much more stringent than that. Tosome extent, a more precise outlook might be engendered byadopting a "watch-dog" attitude in refereeing and reviewingprocesses.

6 3

7-23

Topic: Species Equilibration

R. J. Gel inas , Sc ience Appl ications . The question of whenindividual chemical species are in local equilibrium (at a

single local temperature) can be addressed by examiningindividual species time derivatives, y^(x,t), which are

described by differential equations of the form:

ka y2, etc.) for individual reaction processes which produce

and destroy the ith chemical species at (x,t); and T. is

th 1the local rate of transport of the i chemical species at(x,t) .

For the present discussion, allow Ti to be zero orconstant,, One finds that, as equilibrium is approached,the individual forward and reverse reactions contained inthe sets {p[(y) } and {L*(y) y^} come into balance. This

occurs over relatively long time scales. When localequilibrium is finally approached, detailed balancing isachieved between each forward and reverse process, reactionby reaction, at the local temperature 0(x,t).

Practically, the accession to local equilibrium can bereadily observed by simply rank-ordering the P°(y) and

L*(y)Yj_ for all i species, as a function of time. At

"early" times, individual forward and reverse reactions arescrambled in their order, and only a few such terms tend toconstitute the y^'s. At "intermediate" times some forward

and reverse reactions are seen to approach each other inrank-ordering and in individual magnitudes, and the numberof processes constituting the y^ ' s is increasing. At large,

asymptotic times, all reaction pairs (forward and reverse)balance each other and y i

->0.

In this classic accession to equilibrium the magnitude ofy^ relative to individual Pj°(y) and L

ia(y) y j_

terms also

undergoes a systematic progression. At "early" times y^ is

of the same magnitude as a few individual Pj°(y) and

L1

a(y) y_L

terms. At "intermediate" times, y Lis becoming

small relative to individual reaction terms (P^y) and

64

L^(y)y^) . In this interval y^ is generally determined by

the difference of an increasing set of competing P^y) and

L*(y)y^) terms. At this stage detailed balancing has been

achieved by perhaps only one or two (if any at all)individual forward and reverse reaction terms. The otherreaction terms have sizeable, but offsetting, magnitudeswhich are still scrambled in rank-ordering. At localeguilibrium y^ becomes vanishingly small relative to the

individual P^a and L*(y)y±, and detailed balancing isapproached, reaction by reaction, at the local temperature,9(x,t). Even the weakest individual reactions are requiredto establish the final eguilibrium state.

To answer the original question, it turns out for mostsimple combustion systems near atmospheric pressure(.01-10atm), local equilibrium is approached only at timeswhich are very long compared to flame induction,propagation, energy deposition, and final product formationtimes. Thus, equilibrium rarely exists at a singletemperature for most species at interesting times of flameevolution in the examples discussed today. Depending uponspecific problem circumstances a fairly small number ofreactions can suffice, or as many as 20-30 reactions may beneeded to properly constitute the y^'s. Short-cuts are

ill-advised. As a final note, it is quite possible forquasi-steady states to occur for some species atinteresting flame evolution times; but this need not besynonymous with the existence of equilibrium at a singletemperature. Quasi-steady state can, and does, occur whensome y^ is "sufficiently small" relative to some larger

scale process over specified time scales, say, tq« In such

cases, Tq may generally correspond to "intermediate" times

discussed above, for which eguilibrium has not yet beenestablished at a single temperature. Again, the number

mof

individual reactions needed to properly constitute the y^ '

s

over variable time scales is very much a problem-dependentmatter

.

As a footnote, the concept of multiple temperatures,corresponding to effective equilibria for differentspecies, has often been suggested in non-equilibriumphysics and chemistry, but with little practical advantagebeing realized either conceptually or computationally.

65

7-24

Topic: Use of estimated rate constants in modeling.

D. Indr itz , Princeton University . There is frequentlycriticism of the use of rate parameters that are uncertainin computer models. A point to remember is that omissionof a reaction is orders of magnitude more incorrect thanany reasonable estimate would be. Omission of a reactionbecause its rate is uncertain is effectively including arate constant of zero - orders of magnitude off.

7-25

Topic: Aliphatic/Oxidation in Very Rich Mixtures - Effectof 02 and H20 on Pyrolysis.

E

.

Westley , National Bureau of Standards . I think that the"Outline of Needs" should include a statement showingclearly and explicitly the difference between non-oxidativepyrolysis and oxidative pyrolysis (= combustion) . Thisdifferentiation is implicit in paragraph F4 , but no wherein the outline is it explicit.

Such a differentiation raises the question: How accurateis our knowledge about rate constants resulting in a closedsystem , completely free of any trace of 02 , H20 , or anyother oxygenated compound ? The knowledge of non-oxidativepyrolysis is necessary to study efficiently the oxidativepyrolysis (= combustion)

.

7-26

Topic: CH4 Pyrolysis and Oxidation; use of Post Shock FrontDensity Gradient Technique.

S . H . Bauer , Cornell University . Following my briefdescription of the early stages of the methane pyrolysisand oxidation (fuel-rich regime):

1) Professor F . Kaufman excitedly pointed to one of myslides which showed:

CH4 + OH -> CH3 + H20 k = 3.00 x 10 13 exp (-60 , 000/RT)

This is clearly a typographical error, for which I

apologize. The value we used in our model was: 3.00 x

10 13 exp(-6, 000/RT) , based on the work of Peeters andMahnen, 14th Combustion Symposium.

66

2) Professor A. M. Dean questioned our use of thereaction:

CH20 + M = CO + H2 + M k = 2.10 x 10 16 exp (-35 , 000/RT)

This was also taken from Peeters and Mahnen's paper.Indeed, in our complete analysis we tested the alternateroute for the decomposition of formaldehyde:

CH20 + M -> CHO + H + MCHO + M -> CO + H + MCHO + 02 -> CO + H02CHO + H -> CO + H2

When we inserted a suitable set of rate constants for thesesteps, essentially in agreement with the work of Scheckerand Jost, which Dean favors, we found that all the computeddensity gradient profiles were essentiallyindistinguishable from those based on the single directdecomposition step. In other words, our experiments cannotdiscriminate between these two mechanisms.

In my brief description of the use of the psfdg technique I

attempted to show its utility as a diagnostic for detailedstudies of induction periods prior to the onset ofignition. I am not aware of any other experiments whichprovide this type of information. However, I should alsoadd, that the conclusions derived from the density gradientprofiles are limited (as is any other single diagnostic)

.

We can make statements only with respect to reactions whicheither directly or indirectly (through the production ofreactive intermediates) affect the thermal balance in theshock heated sample. Thus, on the basis of our data alone,we could derive no information on the role of CH2 in theCH4 pyrolysis, nor the detailed path for the decompositionof CH20. On the other hand, the need for the highlyexothermic step:

CH3 + 02 -> CH20 + OH

is loud and clear for T>2000 K. We underscore the need tocouple several types of diagnostics for providing sharpercriteria for reaction mechanisms.

67

7-27

Topic: High Density Kinetics.

J. C. Keck , Massachusetts Institute of Technology . In theinvestigation of individual reaction rates related tocombustion, it is very important to include measurements athigh densities, 1 to 100 atmospheres. This is the range inwhich the order of many of the rate controlling reactionsmay be expected to change from trimolecular to unimolecularin character and the assumption of bimolecular scaling ishighly questionable.

Topic: Relation Between Combustion of Benzene et al . to Coal

A. L. Myerson , Exxon . It would certainly seem that we areall basically in agreement concerning the serious need forchemical kinetic data on the combustion and pyrolysis ofaromatics and other coal-relevant molecules. One techniquewhich seems to possess considerable advantage for suchresearch is the single-pulse shock tube.

This leads to a major point' in our long-range researchneeds. If we are to expend considerable effort on thestudy of these coal-relevant molecules, a major effort willhave to be made on correlating the behavior of these moretractable molecules with that of coal itself. This meansthat many of us will have to make side-by-side or parallelstudies of coal and coal-relevant molecules under asrelatable conditions as feasible.

Topic: Applicability of Various Kinetic Techniques.

C. B Skinner , Wr ight State University . It seems useful tomake a concise list of some of the principal techniques inkinetics, with their typical applications. Seventechniques are listed below without meaning to imply thatthere are no others.

7-28

7-29

Technique Application

.1

.

Molecular beams. Elementary reactions, oftenwith state selection.

2. Low-pressure flowreactors

.

Often useful for elementaryreactions - sometimes withcompl ications

.

68

3. Static reactors. Same as 2. Mainly usedrecently for unimolecularreactions

.

4. Photochemistry. Used to produce active speciesor study reactions of highlyenergized molecules.

5. Shock tubes. Elementary reactions infavorable cases - more oftenfor study of groups of reactions.A good point to model the kineticsubsystem of a complex combustionreaction

.

6. High Pressure Flow Usually involves groups ofReactors. reactions. Transport properties

often of considerable importance.

7. Laminar flames. Same as 6 - Transport propertiesalways important, but not veryhard to handle.

The methods are listed in decreasing order of thesimplicity of understanding the final results (but not ofcarrying out the experiment). All are valuable inproviding data of different types to use in combustionstudies

.

7-30

Topic: Role of Shock Tubes in Kinetics Research.

G. B. Skinner , Wr ight State University . Single elementaryreactions may be studied in a shock tube under favorablecircumstances. Perhaps even more important for combustionresearch, this technique can be used to study the neteffect of groups of reactions (such as in hydrocarbonoxidation) under temperature and pressure conditionscommonly reached in practical combustion processes, withminimal influences of heat and mass transfer effects andminimum wall catalytic effects. For modelers, simulationof shock tube data provides an opportunity to check out a

complete kinetics sub-system before combining it withtransport property sub-systems. Under some conditions, theshock tube can also be used to study wall quenching ofreactions under conditions close to those in practicalcombustion systems.

69

7-31

Topics Monitoring Radical Concentrations.

D. Bogan , Naval Research Laboratory . Development of laserassisted optical diagnostic techniques is receivingconsiderable attention. When these methods are (become)available for a particular species they are (will be) themethods of choice because they cause the minimumperturbation in complex chemical systems. Nevertheless,the beam sampling differentially pumped mass spectrometeris probably still the most generally useful diagnostic toolfor gas phase kinetics. The most serious problem of themass spec, i.e. extensive ion fragmentation, can beovercome by using photo ionization

.

High power lasers in the UV and near UV have recentlybecome widely available. These offer the possibility ofcombining laser multiphoton ionization with the mass spec.Most free radicals have low lying electronically excitedstates and will absorb at longer wavelength than stablespecies which are simultaneously present in kineticstudies. Hence sensitive detection of free radicals shouldbe obtainable with species selectivity in both thewavelength of the multiphoton process and in the mass tocharge ratio of the ionized species.

7-32

Topic? 1) Needs for High Temperature Kinetics Data.2) Acquisition of High Temperature Kinetic Data.

D. J. Seery , United Technologies Resear ch Center .

1) One clear consensus, that seemed evident to me, was,the need for high temperature chemical kinetics studies offuel-rich hydrocarbon oxidation. These studies ofreactions such as R. + R! and R. 4- R ' H should includeboth aromatic and aliphatic species.

2) The one experimental technique capable of measuringboth stable and unstable species at high temperatures ismolecular beam sampling - mass spectrometry. Consideringthe potential of this technique for studying reactivesystems, it is disappointing that so few research labs arepursuing this path.

70

7-33

Topic: Dye laser intracavity absorption and enhancement.

G. 0. Brink , SUNY-Buf falo « The possibility of the use oftunable dye laser" intracavity absorption or enhancement asa diagnostic tool in combustion research should not beoverlooked in the design of kinetic experiments or in theplanning of the new laser diagnostic facility. We havebeen examining the application of this technique to thedetection of atomic sodium in an atomic beam, and have

detected as few as 10 6 atoms in absorption and less than

10^ in enhancement.

At the present time the intracavity process is notunderstood and additional work will be required before itcould be used as a routine diagnostic. However, the highsensitivity and good spectral fidelity that may be obtainedmake the technique potentially valuable for combustionresearch

.

7-34

Topic: Radical Generation by C02-TEA Laser Photolysis

R. Huie , National Bureau of Standards . Recently, J. T.Herron, W. Braun, W. Tsang and I have been working on anew method for the generation of high concentration of freeradicals in a controlled manner: infrared laserphotolysis. Specifically, we have used a C02-TEA laser toproduce CF2 from the photolysis of CF2HC1. The reactionCF2 + CF2 + M -> C2F4 + M was monitored using a guadrupolemass spectrometer and a rate constant derived which is inexcellent agreement with the literature value. From theamount of C2F4 produced, we estimated that 140 mtorr of CF2was produced using laser fluence of 8 J cm-2.

Not only do we feel that this method will be of greatutility for the study of reactions of a large number ofradicals near room temperature, we envision that thetechnique may allow the generation of non-equilibriumconcentrations of selected radicals in flames, where theirinfluence on flame chemistry may then be determined.

71

7-35

Topic: New Experimental Techniques.

A. Fonti jn , AeroChem Research Labs . The need forelementary reaction kinetic data of realistic temperaturesand wide temperature ranges has been implicitly emphasizedfrequently in this meeting. Dr. Garvin this morningmentioned our HTFFR (high-temperature fast-flow reactor)technique in which we have studied metal atom oxidationkinetics at temperatures from 300 to 1900K and pressuresfrom 3-150 Torr. As in many hydrocarbon combustionreactions Arrhenius behavior is often not obeyed in theseprocesses.

Under such conditions extrapolations over wide temperatureranges are not meaningful; in the A1/C02 reactionextrapolation of the 300-700K data to 1900 K would havegiven a rate coefficient an order of magnitude too low(J. Chem. Phys. 67/ 1561 (1977)).

The HTFFR as such appears unsuited for similar hydrocarboncombustion reaction studies because of pyrolysis problems.However, with support from DOE, Dr. Felder and I have nowadapted another previously essentially low temperaturetechnique, pulsed photolysis resonance fluorescence/absorption, to be useful for the same wide temperaturerange. With such an HTP (high temperature photolysis)reactor we have now initiated study of the 0/CH4 reaction,to be followed by OH or 0 with C6H6 and 0/CH3. The rangeof pressures which can be covered with the HTP is probablyeven larger than indicated above for the HTFFR; 3-760 Torris anticipated. We suggest that once our initial studieshave demonstrated the practicality of the HTP technique, itshould be widely applied.

72

7-36

Topic: Kinetic Sensitivity Testing - A Quantitative Basis.

R. J . Gelinas , Science Applications . A recurrent phrase inthis meeting has been, "What is important (chemically)?" Inpartial answer, I will briefly indicate an increasinglyquantitative basis for importance assessments that iscurrently under development.

Three general axioms are well-advised in testing theimportance of individual chemical reactions in flames;they are:

1. Emphasize elementary physical and chemical processes tothe greatest practical extent.

2. Retention or rejection of elementary reactions is verymuch a problem-specific matter, in which physical andchemical mechanisms play off, relatively, against eachother

.

3. Resist the urge to reject candidate processesprematurely.

Each of these axioms is worthy of discussion at anothertime. To test kinetics mechanisms two quantitative levelsof analysis are important:

.Purely kinetics sensitivity analysis•Coupled kinetics-fluid dynamics sensitivity analysis.

Both are essential to validating a kinetics mechanism forspecific applications.

The governing equations for chemical species densities,y i

(x,t) (particles cm-3) , are:

yi(x,t) = E|p"(y) - ^(yjy.j +TA , i=l,N

where Pj and L^y^ are rate expressions (of the form ka y

kay^, etc.) for individual reaction processes which produce

and destroy the ith chemical species at (x,t); and T^ is

the local rate of transport of the ith chemical species at(x,t)

.

PURELY KINETICS ANALYSISIn purely kinetics analysis T^ is neglected. Two

quantitative stages of importance testing can be exercised.The first simply rank-orders all the individual reaction

terms P°(y) and L*{y)yiwhich constitute the y^'s at all

interesting problem times. Reaction rank-ordering can alsobe extended to energy conservation equations. It isgenerally found that reactions shift about in rank-orderingas time evolves. Reactions that are not important overshort time scales may become important over longer time

73

scales, etc. Examples are contained in references 1 and 2.The second quantitative stage evaluates sensitivityvariables, <^y^ , dk , which are a measure of the resultant

variation in the solutions, y i(x , t) , with respect to

intrinsic variations (uncertainties) in the rateparameters. The time-honored method of evaluation dy/^kintroduces an arbitrary variation, Ak , into a given rateparameter, k, and solves the perturbed kinetic mechanism.The sensitivity is obtained from the difference quantity[y (k+Ak ,t) - y(k,t)]/Ak. Several new computational methods(3-7) are now appearing which attempt to evaluate dy/dkwithout repeated solutions of the kinetics mechanism. Thenew methods all have economical or logical problems, whichhave impeded widespread practical applications to thepresent time. Further research should resolve many ofthese problems.COUPLED KINETICS-FLUID DYNAMICS ANALYSISFor coupled analysis the magnitudes of the transport rates,T^ (x,t) , may be highly variable from one point to another

in space and in time for flame systems. The transportrates, of course, depend upon gradients which are driven bythe energy and momentum equations, coupled to the chemicalspecies. Depending upon local, problem-specificconditions, fluid dynamics coupling can markedly alter thesensitivity results that would be obtained from a purelykinetics analysis. Reaction rank-ordering vis a vis, Ti

,

can be readily accomplished in 1-D and 2-D transientanalysis operating on both species and energy equations, asa function of space and time. Only the time-honored methodis currently practical for evaluating sensitivityvariables, dy^/dk, in 1-D or 2-D transient systems.

REFERENCES

1. Gelinas, R.J., "Methane Combustion Kinetics", SAI ReportNo, SAI/PL-NI-77-02 (October 18, 1977).

2. Gelinas, R.J. and Skewes-Cox, P.D. "Tropospher icPhotochemical Mechanisms", Symposium on ReactionMechanisms, Models, and Computers, American ChemicalSociety 173rd National Meeting, New Orleans, LA, (March20-25, 1977). J. Phys. Chem. 81, 2468 (1977).

3. R. I. Cukier, CM. Fortuin, K. E. Schuler, A. G. Petscheand J. H. Schaibly, J. Chem. Phys. 59 , 3873, (1973).

4. J. H. Schaibly and K. E. Schuler, J. Chem. Phys. 59 ,

3879, (1973).5. R. W. Atherton, R. B. Schainker , and E. R. Ducot , AIChE

(March 1975) .

6. Dickinson, R. P. and Gelinas, R. J., "Sensitivity Analysof Ordinary Differential Equation Systems ", J

.

Computational Physics , 21, No. 2, 123-143, (June 1976).7. Stolarski, R. , NASA Goddard Reports X-624-77-122 and 123

(1977) (on Stochastic Sensitivity Methods)

.

74

7-37

Topic: Acquisition and Needs for Kinetic Data

W. Tsang , National Bureau of Standards . Single Pulse Shocktube experiments can be an excellent means of determiningthe mechanisms and rates of initiation in pyrolytic andmany oxidative hydrocarbon decomposition systems. This isbecause the short reaction times < 1 msec and the lowdilutions at which experiments are carried out (due to thesensitivity of gas chromatographic detectors) prevent allexcept the initial unimolecular decomposition processesfrom occurring. These same conditions permit the use of aninternal standard, [a unimolecular decomposition whose rateparameters are well established] thus guaranteeing data ofthe highest precision and accuracy. It should be notedthat these measurements are particularly suited for thederivation of correlations in rate parameters since theseQuantities are always compared with one another. Atpresent sufficient data have been accumulated so as topermit derivation of the rate parameters for the initiatingprocess (bond-breaking) in the decomposition of practicallyall aliphatic hydrocarbons.

With this information as a base, and with suitableadjustment of reactant concentrations measurements on therates of many of the chain processes can and are presentlybeing carried out. These include H and CH3 radical attackon aliphatic and unsaturates as well as branching ratiosfor the decomposition of alkyl radicals. It appears thatthe complete set of rate constants for the decomposition ofaliphatic hydrocarbons to small unsaturates such asethylene or acetylene are accessible. This in turn willset the stage for decomposition studies in the presence ofincreasing quantities of oxygen. The analogy here is withthe work of Baldwin and coworkers where completeunderstanding of the H2, 02 system permit studies on theeffect of added hydrocarbons.

The strategy outlined above can be easily applied tostudies on aromatic and heterocyclic systems. The onefundamental problem is that of analysis. Specifically, cantrace quantities of such compounds and related products bereproducibly introduced and withdrawn from a single pulseshock tube? The most obvious solution is to maintain theshock tube at several hundred degrees C. We do not believethat there are any intrinsic problems. Furthermore in viewof the need for the data as outlined by Drs. Dryer andGlassman and that the same apparatus may be of use fordetermining the mechanism of soot- formation (throughanalysis of the "pr e-soot"compounds that may be formed)

75

experiments of this type would certainly appear to beworthwhile.

Finally, we note that in a more general sense the importantadvances in instrumentation in recent years are such thatmeasurements can be made at increasingly shorter times,lower concentrations and on more exotic or unstablespecies. Thus virtually all reactions involving thermallyequilibrated ground state molecules are accessible. Thesingle pulse shock tube measurements mentioned above are areflection of these advances. The main need is for carefulsystematic work with emphasis on the elementary processes.

76

7-38

Topic: Reactions of OH and HO2

S. Gordon , Argonne National Laboratory .

It is clear that combustion using coal and oil as wellas new fuels used in conventional and new combustiontechniques is of increasing importance. Knowledge of thecombustion process is therefore important in increasing theefficiency of these techniques and in designing newcombustion concepts. It is well known that combustionresults from a chain process involving OH and H02 radicalswhich are initially produced by thermal process at ratherhigh temperatures. The condition for ignition in theburning of fuels are such that very complex reactionsinvolving competing, sequential, and chain branchingreactions, occur in which the OH and H02 radicals play arole

.

We have the capability of studying the individualelementary reactions of these radicals under controlledexperimental conditions. To date we have determined rateconstants for the reactions of the OH radical with a largenumber of simple hydrocarbons including saturated,olefinic, and aromatic compounds as well as a number ofsimple inorganic compounds such as NH3, N20, NO, N02, S02,and C02. These studies were carried out at temperaturesranging from 300 to 440 K and at total pressures ofreactant plus inert gas of about 1 atmosphere. We areextending our work to 600-700 K. In addition we havestudied the rate constants for the following reaction:

H02 + H02 -> H202 + 02.

(k = 1.9 x 10 9 M-l sec-1 in absence of H20, Hamilton, J.Chem. Phys. 6_3 , 3682 (1975).

Knowledge of the reaction rate constants for thisprocess is vital to the understanding of combustion andatmospheric reactions involving H02. Many rate constantshave been determined for H02 reactions in competition withthe above recombination reaction. There has been a greatvariation in the literature on the above rate constants.We have shown that the reported discrepancies can beexplained by the presence of water vapor, since the H02 andwater form a relatively stable complex which influences therate of recombination of the H02 radicals. We have foundthat NH3 has even a greater effect on this reaction. Ourstudies indicate that as much as 35% of atmospheric H02

77

could be complexed at relatively high humidities. Anotherimportant reaction is the cross reaction between OH andH02. There is presently an approximately 30-folduncertainty in the rate constant for this reaction. Withour 2 MeV Febetron and associated gas phase-pulseradiolysis apparatus, we have completed the experimentalportion of the work on this reaction. H20-02-Ar (or N2)mixtures at one atmosphere were pulse radiolyzed. Thedecays of OH (309 nm) and H02 (230 nm) were followed as afunction of reactant concentration, dose, dose rate, totalpressure and diluent gas. Our work will eliminate or sub-stantially reduce the uncertainty in the rate constantresulting from the previously reported studies. Researchhas also been initiated on the reaction of H02 with S02.We are also extending our program to determine thereactivity of OH and H02 with organic compounds todetermine the products of these reactions. [Commentabridged, tables of rate constants omitted, Ed.]

78

Author Index to Comments

Bauer , S . H

.

7—21j

7-26

Bogan, D. 7-31

Bowman, C. T. 7-9

Br ink , G . 0

.

7—33

Dean , A . M

.

7—6

Dorko, E. A. 7-14

Edelson, D. 7-2

Engleman, V. 7-3

Fontijn, A. 7-7 , 7-8, 7-35

Gann, R. G. 7-16

Gel inas , R. J

.

7-15 7-23, 7-36

Gordon, S. 7-38

Huie, R. 7-34

Indr i tz , D

.

7-24

Keck , J . C

.

7-4

,

7-27

Kwe i , G . H

.

7-17 7-18

Lauf er , A

.

7-12

Lin, M. C. 7-13

McLean, W. J. 7-10

Myerson, A. L. 7-5

,

Til -i "7007-11, 7-22, 7-28

Seery, D. J. 7-32

Shaw, R. 7-19

Skinner, G. 7-20 7-29, 7-30

Tsang, W. 7-37

Westbrook, C. 7-1

Westley, F. 7-25

79

8. List of Registrants

Stanley AbramowitzNational Bureau of StandardsBuilding 222, Room B350Washington, DC 20234

0. William AdamsDepartment of EnergyDiv. Basic Energy SciencesWashington, DC 20545

David L. AllaraBell Laboratories7F-212Murray Hill, NJ 07974

Pierre AusloosNational Bureau of StandardsBldg. 222, Room A243Washington, DC 20234

Donald L. BallAFOSR/NCBoiling AFBWashington, DC 20332

A. M. BassNational Eureau of StandardsBldg. 222, Room B162Washington, DC 20234

E . Karl BastrassDepartment of Energy20 Massachusetts Ave., NWWashington, DC 20545

S. H . BauerCornell UniversityDepartment of ChemistryIthaca, NY 14853

Sidney W. BensonDepartment of ChemistryUniv. of Southern CaliforniaLos Angeles, CA 90009

Joan C. BiordiU.S. Bureau of Minesc/o David Forshey2401 E Street, NWWashington, DC 20241

J. W. BirkelandDepartment of Energy20 Massachusetts Ave., NWWashington, DC 20545

Denis J. BoganNaval Research Lab.Code 6180Washington, DC 20375

Craig T. BowmanMech. Eng. Dept.Stanford UniversityStanford, CA 94305

C. 0. BrinkDept. Physics & AstronomyFronczak HallSUNY - BuffaloAmherst, NY 14250

Robert L. BrownNational Bureau of StandardBldg. 222, Room B152Washington, DC 20234

Julius S. ChangLawrence Livermore LabP.O. Box 808Livermore, CA 94550

Norm CohenAerospace CorporationBuilding 130, Boom 150P.O. Box 92957Los Angeles, CA 90009

80

Anthony M . DeanUniv. Missouri - ColumbiaChemistry DepartmentColumbia, MO 65201

Ernest A. DorkoAF Inst, of TechnologyAttn: AFIT/ENYWright-Patterson AFB, OH 45433

Frederick L. DryerPrinceton UniversityRoom D-315A, Engr. Quad.Princeton, NJ 08540

David EdelsonBell Laboratories600 Mountain AvenueMurray Hill, NJ 07974

Victor S. EnglemanScience Applications, Inc.1200 Prospect StreetLa Jolla, CA 92037

Richard C. FarmerDept. of Chem. Engr.Louisiana State UniversityBaton Rouge, LA 70803

William Felder! AeroChem Research Labs.P.O. Box 12Princeton, NJ 08540

Paul R. FieldsArgonne National LaboratoryChemistry Division9700 S. Cass AvenueArgonne, IL 60439

[A. FontijnAeroChem Research Lab., Inc.P.O. Box 12Princeton, NJ 08540

Robert M. FristromJohns Hopkins UniversityJohns Hopkins RoadLaurel, MD 20810

Richard G. GannNational Bureau of StandardsBldg. 225, Room B06Washington, DC 20234

W. C. GardinerUniv. of TexasWEL 103WAustin, TX 78712

David GarvinNational Bureau of StandardsBldg. 222, Room B346Washington, DC 20234

Robert J. GelinasScience Applications, Inc.80 Mission DrivePleasanton, CA 94566

Lewis H. GevantmanNational Bureau of StandardsBldg. 101, Room A529Washington, DC 20234

I. GlassmanPrinceton UniversityPrinceton, NJ 08540

David M. GoldenSRI International333 Ravenswood AvenueMenlo Park, CA 94025

Robert GoulardGeorge Washington Univ.SEAS-CMEEWashington, DC 20052

David GutmanIllinois Institute of TechnologyDepartment of Chemistry3300 S. Federal St.Chicago, IL 60616

Joshua HalpernHoward UniversityDepartment of ChemistryWashington, DC 20059

81

Robert F. Hampson, Jr.National Bureau of StandardsBldg. 222, Room A165Washington, DC 20234

Ronald K. HansonStanford UniversityMechanical EngineeringStanford, CA 94305

Don HardestySandia LabsLivermore, CA 94550

Charles E. HunterHampton Inst. /Ford Motor Co.Hampton, VA 23668

Robert E. HuieNational Bureau of StandardsBldg. 222, Room A145Washington, DC 20234

Doren IndritzPrinceton UniversityFrick Chemistry LabPrinceton, NJ 08540

Dan L. HartleyDepartment 8350Sandia LaboratoriesLivermore, CA 94550

John W. HastieNational Bureau of StandardsBldg. 223, Room A307Washington, DC 20234

Edward F. HayesNational Science FoundationChemistry DivisionWashington, DC 20550

J. HerronNational Bureau of StandardsBldg. 222, Room A145Washington, DC 20234

Jimmie HodgesonNational Bureau of StandardsBldg. 222, Room A345Washington, DC 20234

E. HorowitzNational Bureau of StandardsBldg. 223, Room B362Washington, DC 20234

Jeffrey W. HudgensNaval Research LabCode 6110Washington, DC 20375

Jerome HudisBrookhaven National Lab.Chemistry DepartmentUpton, NY 11973

Marilyn E. JacoxNational Bureau of StandardsBldg. 222, Room A243Washington, DC 20234

Richard J. KandelDepartment of EnergyDivision of Basic Energy SciencesWashington, DC 20545

James S. KaneDepartment of Energy20 Massachusetts Ave., NWWashington, DC 20545

Takashi KashiwagiNational Bureau of StandardsBldg. 225, Room A07Washington, DC 20234

Frederick KaufmanDepartment of ChemistryUniversity of PittsburghPittsburgh, PA 15260

James C. KeckMassachusetts Inst, of Tech.Room 3-342Cambridge, MA 02139

George E. KellerUS Army Ballistic Research LabAberdeen Proving Ground, MD 210

R. D. KernDepartment of ChemistryUniversity of New OrleansNew Orleans, LA 70122

82

John H. KieferUniv. of Illinois, ChicagoChicago, IL 60680

Nathan KleinBallistic Research LabAberdeen Proving Ground, MD 21005

R. Bruce KlemmBrookhaven National LabBldg. 527Upton, NY 11973

Walter KoskiDepartment of ChemistryJohns Hopkins UniversityCharles and 34th StreetsBaltimore, MD 21218

Michael J. KuryloNational Eureau of StandardsBldg. 222, Room A145Washington, DC 20034

George H. KweiLos Alamos Sci. Lab.MS 732, Los Alamos Sci. Lab.Los Alamos, NM 87545

Allan H. LauferNational Bureau of StandardsBldg. 222, Room A145Washington, DC 20234

Yuan T. LeeLawrence Berkeley Lab.Bldg. 70A, Room 4414Univ. of CaliforniaBerkeley, CA 94720

Robert S. LevineNational Bureau of StandardsBldg. 225, Room B64Washington, DC 20234

Arthur LevyBattelle Columbus Lab.505 King AvenueColumbus, OH 43201

M . C. LinNaval Research Lab.Washington, DC 20375

Ralph LivingstonOak Ridge National LabP.O. Box XOak Ridge, TN 37830

Andrej MacekDepartment of Energy20 Massachusetts Ave., NWWashington, DC 20545

William G. MallardNational Bureau of StandardsBldg. 225, Room B06Washington, DC 20234

Michael J. MankaNational Bureau of StandardsBldg. 225, Room B06Washington, DC 20234

R. A. MarcusUniversity of Illinois174 Noyes LaboratoryUrbana, IL 61801

Richard I. MartinezNational Bureau of StandardsBldg. 222, Room A243 .

Washington, DC 20234

Joseph. F. MasiAF Office of Sci. Res.Building 410Boiling Air Force BaseWashington, DC 20332

Max S. MathesonArgonne National Lab9700 South Cass AvenueArgonne, IL 60439

William J. McLeanCornell University284 Grumman HallIthaca, NY 14853

83

Albert L. MyersonExxon Res. & Engr. CompanyP.O. Box 8

Linden, NJ 07036

Cheuk-Yiu NgAmes LaboratoryMetallurgy Room 106Iowa State UniversityAmes, IA 50011

Robert C. OliverInst, for Defense Analyses400 Army-Navy DriveArlington, VA 22202

Elaine OranNaval Research LaboratoryCode 7750Washington, DC 20375

Howard B. PalmerPenn State University319 M . I. Bldg.University Park, PA 16802

Elliot S. PierceDiv. of Basic Energy SciencesDepartment of EnergyWashington, DC 20545

Herschel RabitzDepartment of ChemistryPrinceton UniversityPrinceton, NJ 08540

R. F. SawyerDept. Mechanical Engr.University of CaliforniaBerkeley, CA 94720

Daniel J. SeeryUnited Tech. Res. Cntr.Silver LaneEast Hartford, CT 06108

Robert ShawLockheed Missiles & Space Co.3251 Hanover StreetPalo Alto, CA 94304

David M. SilverJohns Hopkins UniversityApplied Physics LaboratoryJohns Hopkins RoadLaurel, MD 20810

Walter ShaubNaval Research LaboratoryWashington, DC 20375

Gordon B. SkinnerWright State UniversityDepartment of ChemistryDayton, OH 45435

Michael W. SlackGrumman AerospaceResearch Department, Pit. 5

Bethpage, NY 11714

William S. SmithFederal Aviation Agency800 Independence Ave., SWWashington, DC 20591

Kermit C. SmythNational Bureau of StandardsBldg. 225, Room B06Washington, Dc 20234

8H

William SpindelNational Academy of SciencesNational Research Council2101 Constitution Ave., NWWashington, DC 20418

David R. SquireUS Army Research OfficeP.O. Box 12211Research Triangle Park, NC 27709

F. Dee StevensonDepartment of EnergyDiv. Basic Energy SciencesWashington, DC 20545

Wing TsangNational Bureau of StandardsBldg. 222, Room A145Washington, DC 20234

Albert F. WagnerArgonne National LaboratoryChemistry Division9700 S. Cass AvenueArgonne, IL 60439

James A. WalkerNational Bureau of StandardsBldg. 222, Room A145Washington, DC 20234

Karl WestbergThe Aerospace CorporationP.O. Box 92957Los Angeles, CA 90274

Charles K. WestbrookLawrence Livermore LaboratoryP.O. Box 808 L-71Livermore, CA 94550

Francis WestleyNational Bureau of StandardsBldg. 222, Room A145Washington, DC 20234

Bernard T. WolfsonAFOSRAerospace Sciences DirectorateBoiling Air Force BaseWashington, DC 20234

85

HBS-1I4A (REV. 5-78)

J.S. DEPT. OF COMM.BIBLIOGRAPHIC DATA

SHEET

4. TITLE AND SUBTITLE

1. PUBLICATION OR REPORT NO.

NBS SP-531

Summary Report on the Workshop on High Temperature Chemical

Kinetics: Applications to Combustion Research

3. Recipient's AC c<

5. Publication Date

December 19786. Performing Organiz

7. AUTHOR(S)D. Garvin, R. L. Brown, R. F. Hampson, M. J . Kury to

fW. Tsang

8. Performing Organ. Report No.

9. PERFORMING ORGANIZATION NAME AND ADDRESS

NATIONAL BUREAU OF STANDARDSDEPARTMENT OF COMMERCEWASHINGTON, D.C 20234

:t/Task/Work Unit No.

11. Contract/Grant No.

EA-77-A -01 -06 10

12. Sponsoring Orgar mplete Addre > (Sir H, City, State, ZIP)

Department of Energy, Washington, D.C. 20545 andOffice of Standard Reference Data, National Pureau ofStandards, Washington, D.C. 20234

14. Sponsoring Agency Code

IS. SUPPLEMENTARY NOTES

Library of Congress Catalog Card Number 78-600125

16. ABSTRACT (A 200-word or less factual summary ot most significant information. If document includes a significant

bibliography or literature survey, mention it here.)

The proceedings at a workshop on the applications of high temperature chemicalkinetics to combustion research are summarized. This workshop, held Dec. 12-13, 1977at the National Bureau of Standards, provided a forum for the exchange of views onthe needs for kinetics research during the next five to ten years. Experimentaltechniques, measurements, theoretical developments and data evaluation were treatedin four review papers and two discussion sessions.

This report contains the program of the meeting, abstracts of the review papers,a commentary by the organizers of the meeting, summaries of the discussions, formal

comments submitted by the participants and an attendance list.

17. KEY WORDS (six to twelve entries; alphabetical order; capit,

eparated by semicolons

)

nly the first letter of the first key word unlei

18. AVAILABILITY Unlimited 19. SECURITY CLASS(THIS REPORT)

21. NO. OF PAGES

J For Official Distribution. Do Not Release to NTISUNCL ASSIFIED

89

[X1 Order From Sup. of Doc, U.S. Government Printing OfficeWashington. D.C. 20402. SD Stock No. SN0O3-003 ,

20. SECURITY CLASS(THIS PAGE)

22. Price

$2.751 !

Order From National Technical Information Service (NTIS)Springfield, Virginia 22161 UNCLASSIFIED

chemical kinetics; combustion; high temperature chemistry; reaction rate constants; j*'^

measurement techniques; assessment of needs

ir U.S. GOVERNMENT PRINTING OFFICE: 1978 O—281-067 (337)

view copy, write Journal of

i, National Bureau of Standards,

PARTMENT OF COMMERCEton, D.C. 20234

Subscribe now-IhenewNationalBureauof Standards

JournalThe expanded Journal of Research of the National

Bureau of Standards reports NBS research anddevelopment in those disciplines of the physical

and engineering sciences in which the Bureau is

active. These include physics, chemistry, engineer-

ing, mathematics, and computer sciences. Paperscover a broad range of subjects, with majoremphasis on measurement methodology, and the

basic technology underlying standardization. Also

included from time to time are survey articles on

topics closely related to the Bureau's technical

and scientific programs. As a special service to

subscribers each issue contains complete citations

to all recent NBS publications in NBS and non-

NBS media. Issued six times a year. Annual sub-

scription: domestic $17.00; foreign $21.25. Single

copy, $3.00 domestic; $3.75 foreign.

• Note : The Journal was formerly published in

two sections: Section A "Physics and Chem-istry" and Section B "Mathematical Sciences."

NBS Board of Editors

Churchill Eisenhart,

Executive Editor (Mathematics)

John W. Cooper (Physics)

Donald D. Wagman (Chemistry)

Andrew J. Fowell (Engineering)

Joseph O. Harrison (Computer Science)

Stephen J. Smith (Boulder Labs.)

>tion Order Form

ly subscription to NBS Journal of Research

). Add $4.25 for foreign mailing. No additional postage is

1 for mailing within the United States or its f

* File Code 2Q)

Name-First, Last

I I 1 I 1 II 1 1 1 1 1 I 1 1 1 1 1 1 1 1 1 1 1 1

Company Name or Additional Address Line

1 I I 1 1 1 1 II 1 1 1 1 I I 1 1 1 1 1 1 1 1

Street Address

1 1 1 1 1 1 I I I 1 1 I I I I 1 1

City

I I I I

State

_L_Zip Code

I 1 1 1

Remittance Enclosed

(Make checks payable

to Superintendent of

Documents)

Charge to my Deposit

Account No.

MAIL ORDER FORM TO:Superintendent of Documents

Government Printing Office

Washington, D.C. 20402

NBS TECHNICAL PUBLICATIONS

PERIODICALS

•URNAL OF RESEARCH—The Journal of Research

the National Bureau of Standards reports NBS research

d development in those disciplines of the physical and

gineering sciences in which the Bureau is active. These

:lude physics, chemistry, engineering, mathematics, and

Tiputer sciences. Papers cover a broad range of subjects,

;h major emphasis on measurement methodology, and: basic technology underlying standardization. Also in-

<ded from time to time are survey articles on topics closely

ated to the Bureau's technical and scientific programs. Aspecial service to subscribers each issue contains complete

ations to all recent NBS publications in NBS and non-

iS media. Issued six times a year. Annual subscription:

mestic $17.00; foreign $21.25. Single copy, $3.00 domestic;

75 foreign.

•te: The Journal was formerly published in two sections:

.tion A "Physics and Chemistry" and Section B "Mathe-tical Sciences."

MENSIONS/NBSis monthly magazine is published to inform scientists,

y'neers, businessmen, industry, teachers, students, andlsumers of the latest advances in science and technology,

h primary emphasis on the work at NBS. The magazine

|

hlights and reviews such issues as energy research, fire

[itection, building technology, metric conversion, pollution

i itement, health and safety, and consumer product per-

mance. In addition, it reports the results of Bureau pro-

!ms in measurement standards and techniques, properties

j

matter and materials, engineering standards and services,

irumentation, and automatic data processing.

Vnnual subscription: Domestic, $1 1.00; Foreign $13.75

NONPERIODICALSnographs—Major contributions to the technical liter-

re on various subjects related to the Bureau's scientific

technical activities,

adbooks—Recommended codes of engineering and indus-

1 practice (including safety codes) developed in coopera-

i with interested industries, professional organizations,

regulatory bodies,

rial Publications—Include proceedings of conferences

nsored by NBS, NBS annual reports, and other special

lications appropriate to this grouping such as wall charts,

ket cards, and bibliographies.

ilied Mathematics Series—Mathematical tables, man-i, and studies of special interest to physicists, engineers,

mists, biologists, mathematicians, computer programmers,others engaged in scientific and technical work,

ional Standard Reference Data Series—Provides quanti-

se data on the physical and chemical properties of

erials, compiled from the world's literature and critically

uated. Developed under a world-wide program co-

llated by NBS. Program under authority of Nationalidard Data Act (Public Law 90-396).

NOTE: At present the principal publication outlet for these

data is the Journal of Physical and Chemical Reference

Data (JPCRD) published quarterly for NBS by the Ameri-can Chemical Society (ACS) and the American Institute of

Physics (AIP). Subscriptions, reprints, and supplementsavailable from ACS, 1155 Sixteenth St. N.W., Wash., D.C.20056.

Building Science Series—Disseminates technical information

developed at the Bureau on building materials, components,systems, and whole structures. The series presents research

results, test methods, and performance criteria related to the

structural and environmental functions and the durability

and safety characteristics of building elements and systems.

Technical Notes—Studies or reports which are complete in

themselves but restrictive in their treatment of a subject.

Analogous to monographs but not so comprehensive in

scope or definitive in treatment of the subject area. Oftenserve as a vehicle for final reports of work performed at

NBS under the sponsorship of other government agencies.

Voluntary Product Standards—Developed under procedures

published by the Department of Commerce in Part 10,

Title 15, of the Code of Federal Regulations. The purposeof the standards is to establish nationally recognized require-

ments for products, and to provide all concerned interests

with a basis for common understanding of the characteristics

of the products. NBS administers this program as a supple-

ment to the activities of the private sector standardizing

organizations.

Consumer Information Series—Practical information, basedon NBS research and experience, covering areas of interest

to the consumer. Easily understandable language andillustrations provide useful background knowledge for shop-

ping in today's technological marketplace.

Order above NBS publications from: Superintendent of

Documents, Government Printing Office, Washington, D.C.20402.

Order following NBS publications—NBSIR's and FIPS fromthe National Technical Information Services, Springfield,

Va. 22161.

Federal Information Processing Standards Publications

(FIPS PUB)—Publications in this series collectively consti-

tute the Federal Information Processing Standards Register.

Register serves as the official source of information in the

Federal Government regarding standards issued by NBSpursuant to the Federal Property and Administrative Serv-

ices Act of 1949 as amended, Public Law 89-306 (79 Stat.

1127), and as implemented by Executive Order 11717

(38 FR 12315, dated May 11, 1973) and Part 6 of Title 15

CFR (Code of Federal Regulations).

NBS Interagency Reports (NBSIR)—A special series of

interim or final reports on work performed by NBS for

outside sponsors (both government and non-government).

In general, initial distribution is handled by the sponsor;

public distribution is by the National Technical Information

Services (Springfield, Va. 22161) in paper copy or microfiche

form.

BIBLIOGRAPHIC SUBSCRIPTION SERVICES

I following current-awareness and literature-survey bibli-

• iphies are issued periodically by the Bureau:I ogenic Data Center Current Awareness Service. A litera-

( re survey issued biweekly. Annual subscription: Domes-I c, $25.00; Foreign, $30.00.I lined Natural Gas. A literature survey issued quarterly.

1 ual subscription: $20.00.

Superconducting Devices and Materials. A literature survey

issued quarterly. Annual subscription: $30.00. Send subscrip-

tion orders and remittances for the preceding bibliographic

services to National Bureau of Standards, Cryogenic Data

Center (275.02) Boulder, Colorado 80302.

U.S. DEPARTMENT OF COMMERCENational Bureau of StandardsWashington. D.C. 20234

. POSTAGE AND FEES PAID

OFFICIAL BUSINESS ° 5 DEPARTMENT ^COMMERCE

Penalty for Private Use, $300

SPECIAL FOURTH-CLASS RATEBOOK