the oxidation of fuel radicals - nist...the oxidation of fuel radicals wing tsang, iftikhar awan and...

The Oxidation of Fuel Radicals

Wing Tsang, Iftikhar Awan and Jeffrey A. ManionNational Institute of Standards and Technology

Gaithersburg, MD 20899

Multi Agency Coordination Committee for Combustion Research [MACCCR]

Fuel Summit; Princeton, New JerseySeptember 20, 2010

Supported by the AFOSR (Julian Tishkoff)

AIM OF WORK

Develop a quantitative understanding of the processes responsible for the degradation of real fuels during combustion

By building a database of fundamental and transferable information base of rate expressions and thermodynamic and physical properties of

reactants and intermediates

Providing information necessary for the simulation of combustion processes and hence serve as a complement to physical testing

APPROACH

Combine existing literature (experiments and theory) with experiments in single pulse shock tube

Highlights importance of fundamental approach

permits interpolation and extrapolation

use of data in mixtures

calibration of theory

Focus on the decomposition and isomerization reactions of fuel and fuel radicals characteristic of real fuel mixtures

pyrolysis (C,H)

oxidation (C, H, O)

PREMISE OF WORK

There are simple correlations between structure of fuel molecules and their rate constants for decomposition and isomerization

These correlations can be uncovered from an

examination of the literature

suitable experiments

theoretical treatments

Kinetics Modules in Databases

GRIMECH - methane (light hydrocarbon) combustion

Pyrolysis of fuels

Oxidation of larger fuels

Soot formation

Recent NIST experimental and data

evaluation programTarget of most models Focus of current

NIST work

Widely used Problems with rich

mixtures

Many models, begin with unsaturates

7

SPECIAL ROLE OF UNIMOLECULAR REACTIONS

Bimolecular reactions shuffle atomsOH + RH = H2O + R*

Rate constants are the same for similar groupings

Only unimolecular and bimolecular reactions in simulation database

Unimolecular reactions reduces size of molecule

Reduces fuel to small unsaturates and radical found in existing databases

Unimolecular rate expressions are fundamental properties of any polyatomic molecule and hence should be calculable

How accurate are the calculations?

GENERIC TYPES OF UNIMOLECULAR REACTIONS

Reaction from Boltzmann Distribution

limiting high pressure rate constant, no pressure dependence

High pressure rate expresion; All experimental accessible properties of transition state

Reaction from truncated Boltzmann Distribution: reaction from higher energy levels faster than equilibration by collisions

Pressure dependence of rate constants “fall-off”

Reaction from arbitrary initial distribution function: Chemical activation

Branching ratios for reactions and stabilization: pressure dependence

DETERMINATION OF MECHANISMS AND RATE CONSTANTS

Need for clearly defined experimental systems

understand the role of interfering processes

Ideal conditions most easily realizable in shock tubes

well defined boundary conditions

very little possibility of surface reactions

short reaction times

Detection of practically all stable products

(particularly suitable for large molecules)

SINGLE PULSE SHOCK TUBE AND ASSOCIATED WAVE PROCESSES

Log A-factor15.6 15.8 16.0 16.2 16.4 16.6 16.8

Num

ber o

f com

poun

ds

0

2

4

6

8

10

12

14

16

18

C

alkanes

alkenes

EXPERIMENTAL A-FACTORS ON FUEL MOLECULE DECOMPOSITION FROM SINGLE PULSE SHOCK TUBE

STUDIES AT 1050 K

When combined with activation energies lead to present day accepted bond energies

PAST WORK: RATE CONSTANTS FOR DODECANE DECOMPOSITION

1000/T0.5 0.6 0.7 0.8 0.9

Log

k

0

2

4

6

1000/T0.5 0.6 0.7 0.8 0.9

0

2

4

6

Log

k

per Cn-Cn bond for n=2 to 10

per C-C11 bond

100 BAR

10 BAR

1 BAR

100 BAR

10 BAR

1 BAR

HPHP

RADICALS STUDIED

C H3CH2CH2CH2CH2*

CH3CH2CH2CH2CH2CH2*

CH3CH2CH2CH2CH2CH2CH2*

CH3CH2CH2CH2CH2CH2CH2CH2*

CH2=CHCH2CH2CH2*

CH2=CHCHCH2CH2CH2*

CYCLOHEXYL

CYCLOPENTYL

CH3CH(CH3)CH2CH2CH2*

CH3CH(CH3)CH2CH2CH2CH2*

1 2 3 5 4 3 2 1C C C C C C C C

*CH2(CH2 )6CH3 (1-octyl)

CH3CH*(CH2)5 CH3 (2-octyl) CH3CH2CH*(CH2)4CH3(3-octyl)

CH3(CH2)2CH*(CH2)3CH3 (4-octyl) CH3(CH2)2CH*(CH2)3CH3 (4-octyl)

1,2 (8) 1,3 (7)

1,5 (5) 1,4 (6)

2,3 (6)

2,5 (5) CH2=CH2 + 1-C6H11

1-C5H11 + CH2=CHCH3

CH3 + 1-C7H141-C4H8 + 1-C4H9

1-C6H12 + C2H51-C5H10+ 1-C3H7

MECHANISM AND BRANCHING RATIOS FOR THE DECOMPOSITION OF OCTYL RADICALS

4 isomers undergoing 6 beta bond scissions and 6 reversible isomerizations

1000/T1.00 1.02 1.04 1.06 1.08 1.10 1.12 1.14

Log

[ole

fin b

ranc

hing

ratio

]

-2.0

-1.8

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0C2H4

C3H6(square)

C7H14-1

C4H8-1

C6H12-1(diamond) C5H10-1(circle}

OXIDATION OF FUEL RADICALS: PAST WORKNo direct studies leading to high pressure unimolecular rate expressions

Products from flame and cool flames sampling

Oxygenated organics

Unsaturated organics

Cyclic ethers

Thermal rate constants used in models; no consideration of chemical activation processes

OH and HO2 from smaller fuel molecules and modeling

Sandia (Taatjes)

Ab-initio calculations

Green et al

Merle et al

Bozzellie et al

Sandia (Klippenstein)

Rate Constants from Chemical Activation Processes

Radical + O2 , <=> RadicalOO* => Variety of products

RadicalOO

M

STRATEGY FOR STUDYING RADICAL OXIDATION

Generate radicals as in pyrolysis experiments

Dilute concentrations of radicals

Vast excess of chemical inhibitor

Add sufficient amount of oxygen molecules to change reaction direction from pyrolysis to oxidation

Determine cracking patterns as function of oxygen concentration

Reproduce oxidative cracking pattern through solution of the master equation for chemical activation process

SOURCES OF n-BUTYL RADICALS

n-butyl-I = n-butyl + I

but potential interfering process n-butyl-I = a-butene + HI

general source of radicals for pyrolysis experiments

custom synthesis

N-pentyl-O-NO = n-pentoxy + NO

n-pentoxy = n-butyl + H2CO

potential interfering process: isomerization and decomposition processes of pentoxy

home synthesis

CH3CH2CH2CH2 + O2 CH3CH2CH2CH2OO CH3CH2CH=CH2+H

CH2CH2CH2CH2OOH CH3CHCH2CH2OOH

oHO + C2H4 + CH2CH2OOH C2H4 + HO2

C3H6+ CH2OOH

CH2O + O

Oxidative Decomposition of n-Butyl Radical

CH3CH2CH2CH2ICH3CH2CH=CH2 +HI CH3CH2CH2CH2

CH3CH2 + C2H4

C2H4 + H

Pyrolytic Decomposition of n-Butyl Radical (through n-Butyl Iodide)

8 0 0 8 5 0 9 0 0 9 5 0 1 0 0 0 1 0 5 0 1 1 0 0

LOG

[MO

L-FR

ACTI

ON

]

-1 .0

-0 .5

0 .0

0 .5

1 .0

8 0 0 8 5 0 9 0 0 9 5 0 1 0 0 0 1 0 5 0 1 1 0 0

-3 .0

-2 .5

-2 .0

-1 .5

-1 .0

-0 .5

8 0 0 8 5 0 9 0 0 9 5 0 1 0 0 0 1 0 5 0 1 1 0 0

-2 .5

-2 .0

-1 .5

-1 .0

-0 .5

0 .0

1 50 to rr O % O 21 50 to rr 1 .54 % O 2 1 50 to rr 4 .07 % O 21 50 to rr 9 .21 % O 23 00 to rr 1 0 .3 5% O 26 00 to rr 1 0 .1 5% O 2

8 0 0 8 5 0 9 0 0 9 5 0 1 0 0 0 1 0 5 0 1 1 0 0

-3 .0

-2 .5

-2 .0

-1 .5

-1 .0

8 0 0 8 5 0 9 0 0 9 5 0 1 0 0 0 1 0 5 0 1 1 0 0-3 .5

-3 .0

-2 .5-2 .0

-1 .5

-1 .0

-0 .5

LOG

[MO

L-FR

ACTI

ON

]

LOG

[MO

L-FR

ACTI

ON

] LO

G [M

OL-

FRAC

TIO

N]

E th yle n e

P ro p e n e

1 -B u te n e

B u tan a l

T H F

T em p [K } T e m p [K }

T e m p [K } T e m p [K }

T e m p [K }

PRODUCT YIELDS FROM THE DECOMPOSITION OF 100 PPM IODOBUTANE IN 1% MESITYLENE

1 0 0 0 /T0 .9 0 0 .9 5 1 .0 0 1 .0 5 1 .1 0 1 .1 5 1 .2 0

(pro

duct

s+re

acta

nt] fi

nal/r

eact

ant in

ital

0 .7

0 .8

0 .9

1 .0

1 .1

n o o x y g e n 1 5 0 to rr in it ia l p re s s u re9 .1 5 % o x y g e n 1 5 0 to rr in it ia l p re s s u re1 0 .1 5 % o x y g e n 6 0 0 to rr in it ia l p re s s u re1 0 .3 5 % o x y g e n 3 0 0 to rr in it ia l p re s s u re

a ll m ix tu re s c o n ta in in g 1 0 0 p p m n -b u ty l- I a n d 1 % m e s ity le n e

MASS BALANCE

850 900 950 1000 1050

Log

[mol

e-fr

actio

n]

-1 .0

-0.5

0.0

0.5

1.0

850 900 950 1000 1050

-2.0

-1.5

-1.0

-0.5

0.0

850 900 950 1000 1050-2.5

-2.0

-1.5

-1.0

-0.5

850 900 950 1000 1050

-2.5

-2.0

-1.5

-1.0

850 900 950 1000 1050

-3.0

-2.5

-2.0

-1.5

-1.0

150 torr 0% O 2150 torr 1 .25% O 2150 torr 3 .13% O 2150 torr 6 .05% O 2150 torr 10.1% O 2150 torr 22.6% O 2

Ethylene

Log

[mol

e-fr

actio

n]Lo

g [m

ole-

fract

ion]

Log

[mol

e-fr

actio

n]Lo

g [m

ole-

frac

tion]

Tem p [K ]

Propene

1-B utene

B utanal

TH F

Tem p [K ] Tem p [K ]

PRODUCT DISTRIBUTION FROM THE DECOMPOSITION OF N-PENTYL NITRITE 100 ppm IN 1% MESITYLENE

ETHYLENE/THF RATIOS AS A FUNCTION OF TEMPERATURE AND OXYGEN CONCENTRATION

1/O2 [ mol/cc]0 1e+5 2e+5 3e+5 4e+5 5e+5

C2H

4/TH

F

0

100

200

300

400

500

1050

1020

990

960

930900

870

Ethylene/THF Yields as a Function of 1/O2: Extrapolation to Infinite Oxygen Concentration

Log [C2H4/(2xTHF)] = 7.35 - 6000/T =1.1 at 930 K

1000 /T0 .90 0 .95 1 .00 1 .05 1 .10 1 .15 1 .20

Log[

C3H

6/TH

F]

0 .0

0 .5

1 .0

1 .5

2 .0

150 , 1 .54%

150 .4 .04%

150 , 9 .15%300 ,10 .35% 600 .10 .15%

PROPENE/THF RATIOS AS A FUNCTION OF TEMPERATURE AND OXYGEN CONCENTRATION

Log [C3H6/THF] = 2.92 -2120/T= .64 (930K)

THEORETICAL BASIS FOR PRESENT WORK

Need for fundamental molecular properties of molecules, intermediates and transition states

None of the required information is available from experiments

Reliance on ab initio calculations

Bozzilli

Hadad

Green

Uncertainties

< 1.5 kcals/mol for molecules using isodesmic reactions

no tests for transition states

MASTER EQUATION SOLVER: PROGRAM TO DETERMINE MOLECULAR DISTRIBUTION FUNCTION

RELEVANT RATE CONSTANTS RESPONSIBLE FOR PRODUCT YIELDS

1-Butene propene ethylene

Log [time.s]-12 -10 -8 -6 -4

Log

[bra

nchi

ng ra

tio]

-6

-5

-4

-3

-2

-1

0n-butyl + O2

Ethylene + CH2CH2OOH

Propene + CH2OOH

1-Butene + OOH

THF + OH

1-Butanal + OH

C4H9OO

1-C4H9OO

2-C4H9OO

3-C4H9OO

CALCULATED BRANCHING RATIOS FOR THE CHEMICALLY ACTIVATED DECOMPOSITION OF PROPYLPEROXY

RADICALS: 900 K, 3 BAR

Log [time,s]-12 -10 -8 -6 -4

Log

[bra

nchi

ng ra

tio]

-5

-4

-3

-2

-1

0nC4H9 + O2

Ethylene +CH2CH2OOHPropene +CH2OOH

1-Butene +HO2

OH + THF

1-Butanal +OH

BRANCHING RATIOS WITH SOME RATE EXPRESSIONS ADJUSTED TO FIT OBSERVATIONS 900 k 3 BAR

SUMMARY

Single pulse shock tube has been used to study the oxidation of n-butyl radical

In contrast to the pyrolytic situation many new channels are opened

The necessity of treating the reactions as chemical activation processes

The problems involved in converting results to fundamental unimolecular high pressure rate expressions are desc ribed

Pyrolysis decomposition extends to regions of high oxygen concentration and lower temperatures