Soot formation in shock-tube oxidation of hydrocarbons

Twentieth Symposium (International) on Combustion/The Combustion Institute. 1984/pp. 871-878

SOOT FORMATION IN SHOCK-TUBE OXIDATION OF HYDROCARBONS M. FRENKLACH, M. K. RAMACHANDRA AND R. A. MATULA Department of Chemical Engineering Louisiana State University Baton Rouge, Louisiana, 70803, USA

Soot formation during oxidation of acetylene, allene, 1,3-butadiene, toluene, benzene and chlorobenzene was studied in argon-diluted mixtures behind reflected shock waves. The formation of soot in pyrolysis of benzene, chlorobenzene and chlorobenzene-acetylene argondiluted mixtures was also investigated. The experiments were conducted at temperatures from 1430 to 3490 K, pressures from 0.20 to 3.14 bar, and carbon atom concentrations of approximately 2 x 101: and 5 x 101= atoms/cm ~. Soot formation was monitored by measuring the attenuation of a laser beam at 639.8 nm. Comparing soot-yield dependencies obtained in pyrolysis and oxidation at similar conditions, it was observed that oxygen, added in moderate amounts, promotes the formation of soot at lower temperatures and retards it at higher ones. The promoting effect of oxygen was particularly pronounced in acetylene mixtures. Investigating further the effect of fuel molecular structure, it was noted that soot formation from benzene is very similar to that from toluene; however, the chlorobenzene soot-yield bell is shifted to lower temperatures. The addition of acetylene to chlorobenzene pyrolysis decreased the amount of soot produced. All of the observed phenomena can be rationalized within the conceptual model of soot formation suggested recently for hydrocarbon pyrolysis.

Introduction Understanding soot formation in the combustion of hydrocarbons, remains a challenging problem of combustion science. 1-3 Recently, 4-6 we analyzed time, concentration, pressure, extinction mode, and fuel type dependencies of soot production in hydrocarbon pyrolysis under shock tube conditions. The results obtained in these studies were explained in light of conceptual models developed for pyrol[sis of aromatic4'5 and non-aromatic hydrocarbons; the underlying principle for these models being an interaction between aromatic rings and acetylenic species suggested by Howard and coworkers. 7.s In the present work we investigate soot formation in shock-tube oxidation of various hydrocarbons. It was recently reported by Wang et al. 9 and Rawlins et al. I~ that oxygen suppresses soot formation from toluene. The results obtained in the present work indicate that, depending upon experimental conditions, oxygen also promotes the production of soot. In this study we primarily concentrate on soot formation in oxidation of acetylene, allene, 1,3-butadiene and toluene, the species for which soot formation dependencies were established in the previous pyrolysis studies. 4'~ The oxidation was conducted at the conditions of the pyrolysis experiments, in order to compare the results of both studies. 871

To further investigate the effect of fuel structure on soot formation in shock tubes, the pyrolysis and oxidation of benzene and chlorobenzene were also conducted. The interest in chlorobenzene is motivated by the recent results of Smith and Johnson u and Singh and Kern 12 that significantly larger concentrations of phenyl radicals are observed in pyrolysis of chlorobenzene than in that of benzene.

Experimental The experimental apparatus and procedures used in this study were similar to those described in our previous works on hydrocarbon pyrolysis. 4'6 Briefly, the experiments were conducted behind reflected shock waves in a 7.62 em i.d. shock tube. The test gas mixtures were prepared manometrieally. The stated purities of the gases were: argon-99.995%, oxygen-99.5%, acetylene-99.6%, allene-97% and 1,3butadiene-99%. These gases were used without further purification. The toluene (Reagent, Baker), benzene (Speetranalyzed, Fisher) and ehlorobenzene (Reagent, Baker) were purified by repeated freezing and evacuation. The shock tube was cleaned after every run. Fourteen different mixtures were tested during the course of this study. The experimental conditions are summarized in Table I. The series of experiments were designed so that the effects of tern-

872

COMBUSTION-GENERATED PARTICULATES TABLE I Summary of experimental conditions behind reflected shock waves

Series A B C D E F G H I J K L M N

Composition (% vol. in Argon) Fuel 02 0.311, CrH8 1.75, CvHs 4.65, C~H2 4.65, C2H2 20.00, CzH2 20.00, Call2 20.00, C2H2 0.726, C3H 4

0.54, C4H6 0.311, C6H6

0.311, C6H6 0.311, CsH~CI 0.311, C6H5C1 0.311, C6H5C1/ 1.09, CzHz J

T (K)

P (bar)

[Carbon] • 10 -17 (atoms/cm 3)

0.311 -0.311

1496-2391 1668-2932 1532-3490 1429-2259 1379-2623 1431-2758 1578-2997 1497-2041 1561-2083 1561-2273 1516-2243 1447-2201 1433-2184

1.87-3.08 0.33-0.68 1.10-3.14 1.08-1.70 0.22-0.37 0.20-0.40 0.26-0.50 1.93-2.56 2.02-2.68 1.94-2.87 1.98-2.83 1.89-2.81 1.81-2.77

1.97-2.03 1.69-2.33 4.43-6.19 4.89-5.23 4.16-5.96 4.01-5.83 4,09-5.68 1.98-2.06 1,96-2.04 1,67-1.75 1,70-1.76 1,70-1.78 1.66-1.77

--

1475-2238

1.92-2.82

3.66-3.84

0.311 1.75 4.65 1.5 6.45 4.30 2.15 0.726 0.54 --

perature, pressure and oxygen concentration could be clearly observed, To estimate the temperature change as a result of reaction, we simulated the oxidation of toluene, using the mechanism of Jachimowski and coworkers, 13'14 and the oxidation of acetylene, using the mechanism of Gardiner et al. 15 The details of these computations can be found elsewhere. 16 The numerical results indicated that for a reaction time of 1 ms, the time at which the soot-yield dependencies are compared in this paper (this time is chosen arbitrarily--qualitatively similar results were obtained at all observation times), the temperature changes in all the acetylene series, C through G, and the toluene series at higher pressures, A, are less than 20 K and the temperature changes in the toluene series at lower pressures, B, are below 60 K. The soot conversion was determined by measuring the attenuation of a He-Ne 632.8 nm laser beam. As in our previous studies, the soot yields were calculated according to Graham's modeP 7 b u t leaving out the quantity E(m) = -Im[(m 2 - l)/(m z + 2)], where m is the complex refractive index of soot particles. This arbitrary form of reporting the results was chosen to emphasize the ambiguity in the value of m and in the laser-extinction model itself.4'18 An additional source of uncertainty is the absorption of a laser beam by molecular species rather than soot particles; light at 632.8 nm is already absorbed by compounds with as few as six aromatic rings. 19 However, Rawlins and coworkers 1~176concluded from their recent optical studies that light extinction at 632.8 nm is primarily caused by soot particles.

Results and Discussion

Toluene Mixtures The temperature dependency, i.e. the existence of a "soot-yield bell," and the time development of this bell in oxidation of toluene are qualitatively similar to those observed in toluene pyrolysis,4 The quantitative effect of oxygen on soot yields appears to be dependent on the pressure range. At higher pressures oxygen suppresses the formation of soot within the entire temperature range tested (Fig. 1). The effect is similar to that resulting from a reduczl tion in the initial concentration of toluene. At lower pressures (Fig. 2), the addition of oxygen not only reduces the maximum soot yield but also shifts the maximum to a lower temperature. The shift is too large to be explained by a temperature change as a result of chemical reaction. Thus we conclude that at lower pressures and lower temperatures oxygen actually promotes soot formation. The observed phenomenon can also be interpreted as follows: the shift of the soot yield maximum to higher temperatures when pressure is lowered is smaller when oxygen is present. This observation indicates that addition of oxygen causes reactions to occur that compete with pressuredependent processes. In our recent analysis of soot formation during pyrolysis of aromatic hydrocarbons, a it was suggested that pressure-dependent fragmentation of an aromatic ring initiates the production of soot. Hence, oxidation reactions must compete with ring fragmentation. Furthermore, the promoting effect of oxygen on soot formation ob-

SOOT FORMATION IN OXIDATION OF HYDROCARBONS g

tion products CO and COz and small inefficient fragments, active intermediates for the soot-forming route. At low temperatures, when thermal ring fragmentation is slow, the formation of these intermediates by oxidative reactions enhances soot production. At higher temperatures, when thermal decomposition is predominant, additional removal of aromatic rings and active intermediates by oxidative agents inhibits formation of soot.

to_ /k

D . 3 1 1 Z E 1 H e + ~r-

+

[3.311 7 . E T H a + 0 . 3 1 1 Z O z * R r "

,.~.

A

Y b2

=, N~ ).-

Acetylene Mixtures

~.

g

OO.

16go.

1800. 2OOO. TEMPERtqTURE (KI

2200.

2v,OO.

FIG. I. Comparison of soot yields in pyrolysis (Mixture 1 in Ref. 4) and oxidation (Series A) of toluene at higher pressures for a reaction time of 1 IllS.

served below 2100 K at low pressures and the suppression of soot above this temperature indicate that oxidizing agents (such as O2, O, OH) attack the aromatic ringzz producing, besides the oxida-

I . 75 ZCrflll + Fir c~

+

1" 75 Z C ~ H I * I " 75 Z 0 a * R r

m_

A

L~ X

9~

o

tJ 1.-. E:I

,4'

~300.

Large amounts of oxygen, such as in Series C and E, completely suppress the production of soot from acetylene. The addition of oxygen in smaller quantities" significantly promotes soot formation at lower temperatures, while suppressing it at higher temperatures (Figs. 3 and 4). At higher pressures (Fig. 3), this actually results in a shift of the sootyield bell to a lower temperature by as much as 500 K with only a slight effect on the maximum value itself. The observed phenomena can be rationalized within the conceptual model of soot formation suggested recently for acetylene pyrolysis.6 At low temperatures, the rate limiting step of soot formation is the production of aliphatie intermediates. Addition of relatively small amounts of oxygen resuits in formation of reactive radicals (e.g. O and OH), which prmnote the pyrolysis reactions. Under these conditions, the rate of formation of aromatic intermediates is enhanced, which, in turn, increases the rate of soot production. At high temperatures, a partial equilibrium is quickly established and the fragmentation of aromatic rings becomes a dominant factor in soot formation process. As was postulated earlier, the oxidizing agents remove aromatic rings and so inhibit the formation of soot. Thus, the overall effect of oxygen on soot formation in acetylene pyrolysis should be a shift of the soot yields to lower temperatures, as was observed experimentally. The larger effect of oxygen on soot production for acetylene compared to that for toluene can also be explained now. At low temperatures, oxygen very efficiently promotes the formation of aromatics from acetylene compared to a relatively marginal contribution to the formation of reactive ahphatic species in the case of toluene. At high temperatures, oxygen effectively diverts reaction intermediates from the route to soot for both cases.

g 1700.

2t00. 2500, TEt~PER~TURE [KI

2g00.

3300,

El(;. 2. Comparison of soot yields in pyrolysis (Mixture 2 in Ref. 4) and oxidation (Series B) of toluene at .lower pressures for a reaction time of 1 InS.

873

AUene and 1,3-Butadiene Mixtures Figures 5 and 6 present the results obtained with aIlene and 1,3-butadiene mixtures. Analysis of these figures indicates that the effect of oxygen on soot tbrmation from allene and butadiene is mtermedi-

874

COMBUSTION-GENERATED PARTICULATES

..:A

u,.65 ZCaHz + Ar

+

L&. 65 ZCaHz+I.50 ZOa+Ar

&

O. 7 2 6 ZC:H= + Rr

+

O. 7 2 6 ZC=H~+O, 726 ZO~+Rr

-_

o

.,

~

X

~"

+

A

=o

).-

~-

o

=.. o

E2~

d. %;00.

+ ~

a~/~ :]0.

1500,

' - 5,;~.

=soo.

zdoo.

=2oo.

2~oo.

TEMPERRTURE (K]

o

--

+

,

1900.

+ ;~300,

;~'/00.

TEMPERATURE (K)

FIG. 5. Comparison of soot yields in pyrolysis (Series I in Ref. 6) and oxidation (Series It) of allene at a reaction tilne of 1 ms.

FIC. 3. Comparison of soot yields in pyrolysis (Series D in Ref. 6) and oxidation (Series D) of acetylene at higher pressures for a reaction time of

,2"

1 ms.

ate between the effects observed for acetylene and toluene. The promotion effect at lower temperatures and the suppression at higher temperatures can be explained as previously by interaction between oxidizing agents and active intermediates.

A

0 . 5 u, ZC=H e § R~

-t-

O. 50, ZCtHs+O. 50, ZOa+Rr

E

~

A

AA

g

.~ ,4.

N5 ~'-

g

o ~;"

N

0

20. OO ZCaHz + R r

A

2 0 . 0 0 7.CaHa+2. 15 7.0a+gr 2 0 . 0 0 XC2Hz+t~.30 ZO2*Rr

+

d

(~

AA

e~ o

~09 ~

1700.

1900 9

2100,

2300.

2500,

TEHPERIqTURE (KI

I-J'

~2 ~--

FI(;. 6. Comparison o f soot yields in pyrolysis (Series J in Ref. 6) and oxidation (Series 1) of 1,3hutadiene at a reaction time of 1 ms.

,2.

g m N d

%00.

Benzene and Chlorobenzene Mixtures

14oo.-'~'-" 21'oo. 2~oo. T E H P E R A T U R E (KJ

2~oo.

33oo.

FIe;. 4. Comparison of soot yields in pyrolysis (Series G in Ref. 6) and oxidation (Series F and G) of acetylene at lower pressures for a reaction time of 1 ms.

Figures 7 and 8 present the soot yields obtained in pyrolysis and oxidation of benzene and chlorobenzene mixtures, where they are compared to the results obtained at similar conditions in the toluene mixture. As can be seen in these figures, soot formation in pyrolysis and oxidation of benzene is very similar, both qualitatively and quantitatively, to that of toluene. This fact provides fimrther support for

SOOT FORMATION IN OXIDATION OF HYDROCARBONS

(~

Q.311ZCTH e + Rr

/~

O. 311%CeHs + g r

"F

0.311%CeHsC attsC1 +

~'

+

~

+.

g~

875

Examining Figs. 7 and 8, one observes that the effect of oxygen on soot yields is smaller for chlorobenzene than for benzene and toluene. More interestingly, however, the soot-yield bell is shifted to lower temperatures by about 150 K compared to benzene. The shift and its value can be explained by the difference between the cleavage of C - - H and C--CI bonds in benzene and chlorobenzene molecules, respectively. The position of the soot-yield maximum in pyrolysis of aromatic hydrocarbons is determined 4'~ by the ratio

E)

)--

g

N

r = kf/kp[A]o,

@ +

tdoo.

oo,

tdoo.

~doo.

2~oo.

TEHPERRTURE (K)

where kf is the rate constant of ring fragmentation, [A]o is the initial hydrocarbon concentration, and kp is the rate constant of the polymerization process 2400. forming soot. Assuming the maximum appears at the same value of r, one can write

Fla. 7. Comparison of soot yields in pyrolysis of toluene (Series A in Bef. 4), benzene (Series J) and chlorobenzene (Series L) at a reaction time of 1 ins.

O I 3 | I Z C T H e + O o 3 I I ZOz+Rr

=o

(1)

k~/k.[A]o = k}/k;,[A'lo,

(2)

where the left-hand side of Eq. (2) is ratio (1) for benzene and the right-hand side is that for chlorobenzene. For the conditions used in this study [A]o = [A']o (see Table I) and assuming kp = k~,, Eq. (2) can be rewritten as

0,311ZCBHs+O.311ZOz+Rr 0.31IZCaHsCI+O.311ZOz+Ar

ky = k'f.

(3)

Considering the energetics 23'24 of the elementary reactions which may constitute the fragmentation path for benzene molecule, the following sequence

w

(bH---> ~b + H )--

~-

AH~

= 474 kJ/mol; AS~

= 153 J/mol" K, (R1)

(b ~ HC~C--CH~---CH--CH~(~H O

o.+ 9Joo.

-

t6oo.

t~oo.

2doo.

2~oo.

AH~

r 2~00,

TEMPERRTURE (K}

FIG. 8. Comparison of soot yields in oxidation of toluene (Series A), benzene (Series K) and chlorobenzene (Series M) at a reaction time of I ins. our conceptual model because one would expect the rate constants of the aromatic ring fragmentation, the suggested initiation of soot formation from aromatic hydrocarbons under shock tubes conditions, 4'6 to be of the same order of magnitude for homologous hydrocarbons. A slight decrease in the amount of soot produced from benzene compared to that from toluene may suggest that soot formation should be analyzed on a total-carbon-atoms basis rather than on a molecular one.

= 247 kJ/mol; hS~

= 65 J/mol" K, (R2)

where (b denotes phenyl radical, is suggested for the experimental conditions used in this study. The rate limiting step of the sequence, particularly in the initial phase of pyrolysis, is reaction (R1). Hence, the rate constant of this reaction, kh should be considered for the left-hand side of Eq. (3). By analogy, the right-hand side of Eq. (3)is set equal to the rate constant of the following reaction ~bCl ~ (b + C1 AH~ hS~

= 365 kJ/mol; = 123 J/mol. K. 25'26 (R3)

Assuming the Arrhenius form for the rate constant expressions and assuming the ratio of the rate con-

876

COMBUSTION-GENERATED PARTICULATES o

stants for the reverse directions of reactions (R1) and (B3) k_/k" = (ix,/p.)~/2, where tx and Ix' are the reduced masses for the corresponding recombinations, Eq. (3) can be rewritten as R

_-en(IX'/IX) + (AS 2

-

E) =o

"

d_

+

1.09 ZCzH2 + Rr 0.311ZCaHsC1 + fir I. 09 XC2H~+O.311ZC6HsCI+RP

AS')

A

w

= AH/Tm - A H ' / T "

(4)

where Tm and T" are the temperatures of soot yield maxima in mixtures J and L, respectively. For a reaction time of 1 ms, Tm is approximately equal to 1900 K (Fig. 7). Using the above thermodynamic data for 1800 K, the approximate midpoint of the temperature interval of interest, we obtain from Eq. (4) T" = 1772 K, so that the temperature shift, T,, - T'm, equals approximately 130 K. This is in good agreement with the corresponding experimental value. The above discussion is in disagreement with the conclusion of Singh and Kern 12 that "the phenyl radical is not the primary product of benzene decomposition . . . [and] . . . fragmentation of benzene [presumably by an elementary reaction] is the dominant pathway of the decomposition in the temperature range of 1400-2200 K." Their conclusion is based on the observation, previously noted by Smith and Johnson, ll that a larger concentration of phenyl radicals is observed in the pyrolysis of chlorobenzene than in the pyrolysis of benzene at similar conditions. This result can be explained, however, by a simple kinetic argument of consecutive reactions. Indeed, in a two-step sequence kl

A

k2

~bH --->~b--~ products, the concentration of the intermediate, ~b, increases with the ratio kl/k2. Based on the previous discussion, for chlorobenzene this ratio at 1800 K is approximately 8 times larger than a similar ratio for benzene, which should result in a higher concentration of phenyl radicals for the former case. Further support for our argument is provided by the results of additional: experiments, in which acetylene was added to chlorobenzene. One would expect that addition of acetylene to chlorobenzene would promote or, at least, not affect the production of soot. However, experimental results (Fig. 9) indicate that acetylene suppresses soot formation. This phenomenon can again be rationalized within the conceptual model discussed earlier. The reaction between phenyt radical and acetylene, being fast and producing a relatively stable species, competes with fragmentation (R2.). Since the radical H C ~ C - - C H ~ C H - - C H = C H , formed in Reaction (R2), and immediate products of its decomposition are presumably more efficient "building blocks" than acetylene, s the removal of phenyl rad-

.5 ~o

i" ~, ~doo. '

q,Boo. ~ ~ t ~ o . C "-%doo. TEHPERRTURE (K}

2200. z

FIG. 9. Comparison of soot yields ill pyrolysis of acetylene (Series A in Ref. 6), chlorobenzene (Series L) and acetylene-chlorobenzene mixture (Series N) at a reaction time of 1 ms. ical from the fragmentation route may explain the observed reduction in the soot yields.

Conclusion

The main conclusion to be drawn from the resuits of this work is that the mechanism of soot formation in shock tubes is probably the same for both pyrolysis and oxidation of hydrocarbons. That is, the addition of oxygen does not alter the soot route but rather promotes or inhibits this route by means of competitive reactions. Acknowledgment

We wish to thank Professor R. D. Kern for providing us with the preprint of his paper. The work was supported by NASA Contract No. NAS 3-23542 and NASA Grant No. NAG 3-477. REFERENCES 1. HAYNES, B.

S. AND WAGNER, H. GG.: Prog. Energy Combust. Sci. 7, 229 (1981).

2. Particulate Carbon: Formation During Combustion (D. C. Siegla and G. W. Smith, Eds.),

Plenum, 1981. 3. Soot in Combustion Systems and Its Toxic

SOOT FORMATION IN OXIDATION OF HYDROCARBONS

Properties (J. Lahaye and G. Prado, Eds.), 4. 5. 6.

7.

Plenum, 1983. FRENKLACIt, M., TAKI, S. AND MATULA, R. A.: Combnst. Flame 49, 275 (1983). FRENKLACH, n . AND CLARY, D.: Ind. Eng. Chem. Fundam. 22, 433 (1983). FRENKLACti, n . , TAKI, S., DURGAPRASAD, M. B. AND MATULA, R. A.: Combust. Flame 54, 81 (1983). BITTNER, J. D. AND HOWARD, J. B.: Eighteenth

Symposium (International) on Combustion, p. 1105, Tile Combustion Institute, 1981. 8. BITTNER, J. D., HOWARD, J. B. AND PALMER, H. B.: p. 95 in Ref. 3. 9. WANG, T. S., MATULA, R. A. AND FARMER, R. C.: Eighteenth Symposium (International) on Combustion, p. 1149, The Combustion Institute, 1981. 10, RAWLINS, W. T., TANZAWA, T,, SCHERTZER, S. P. AND KRECH, R. I-I.: Synthetic Fuel Combus-

877

Poitiers, France, July, 1983. 16. FRENKLACH, M.: Shock Tube Study of the Fuel

Structure Effects on the Chemical Mechanisms Responsible for Soot Formation, NASA Report CR-174661, November 1983. 17. GRAHAM, S. C., HOMER, J. B. AND ROSENFELD, J. L. J.: Proc. R. Soc. London A344, 259 (1975). 18. FRENKLACH, M., TAKI, S., Ll KWOK CHEONG, C. K. AND MATULA, R, A.: Combust. Flame 51, 37 (1983). 19. BAUER, S. H. AND ZHANG, L. M.: 14th Inter-

national Symposium on Shock Tubes and Waves, Sydney, 1983.

20. RAWLINS,W. T., COWLES, L. M. AND KRECH, R. H.: Optical Signatures of Soot Formation in the Pyrolysis of Toluene Near 2000 K. Paper presented at the Fall Technical Meeting of the Eastern Section of the Combustion Institute, Providence, Rhode Island, November, 1983. 21. FRENKLACH, M., CLARY, D. AND MATULA, R. A.:

tion: Pollutant Formation. Soot Initiation Mechanisms in Burning Aromatics, Physical

Empirical Modeling of Soot Formation in Pyrolysis of Aromatic Hydrocarbons. Paper pre-

Sciences Report TR-361, 1983. 11. SMITH, R. D. AND JOttNSON, A. L.: Combust. Flame 51, 1 (1983). 12. S1NGH, n . J. AND KERN, R. D.: Colnbust. Flame 54, 49 (1983). 13. MCLAIN, A. G., JACHIMOWSKI, C. J. AND WILSON, C. H.: Chemical Kinetics Modeling of

sented at the Fall Technical Meeting of the Eastern Section of the Combustion Institute, Providence, Rhode Island, November, 1983. 22. LITZINGER, T. A., BREZINSKY, K. AND GLASSMAN,

I. : Some Further Results on the Toluene Oxidation Mechanism. Paper presented at the Fall

Benzene and Toluene Oxidation Behind Shock Waves, NASA Technical Paper 1472, 1979. 14. JACItIMOSKI, C. J.: Combust. Flame 23, 233

23.

(1974); 29, 55 (1977); 31, 102 (1978). 15. GARDINER, W. C., JR., HWANG, S. M, AND WARNATZ, J. : Combustion Mechanism of Acet-

24. 25.

ylene Flames and Ignition Using a Truncated Reaction Mechanism. Paper presented at the 9th International Colloquium on Dynamics of Explosions and Reactive Systems, Universite de

26.

Technical Meeting of the Eastern Section of the Combustion Institute, Atlantic City, New Jersey, December, 1982. FRENKLACH, M., CLARY, D., GARDINER, W. C., JR. AND STEIN, S. E.: this symposium. STEIN, S. E.: to be published. STULL, D. R., WESTRUM, E. F., JR. AND SINKE, G. C.: The Chemical Thermodynamics of Organic Compounds, p. 532, Wiley, 1969. BENSON, S. W.: T h e r m o c h e m i c a l Kinetics, Wiley, 1976.

COMMENTS R. D. Kern, Univ. of New Orleans, USA. The assumption of equating the rate constants for the polymerization processes occurring in the benzene and chlorobenzene should be tested by adding a source of C1 atoms to the benzene mixture. In our studies of chlorobenzene pyrolysis, we detect one mole of HC1 produced for each mole of ~b-C1 decomposed. Furthermore, the rate constants for the decomposition of c h l o r o b e n z e n e (Skinner) and B e n z e n e (Kiefer) have been reported at this Symposium, The ratio of these rate constants at 1800 K is 1.3 instead of the value of 8 calculated from your argument. The temperature shift of the soot bell is less than 10 K.

Authors" Reply. W h e t h e r the presence of C1 atoms affects the polymerization process of soot formation is an interesting question and the issue is currently under experimental investigation in our laboratory. The assumption of equating the rate coeflqcients for polymerization processes in the benzene and chloro-benzene mixtures of equal initial reactant concentrations is based on the experimental observation (Figs. 7 and 8 of the paper) that whereas the soot-yield bell for chlorobenzene is shifted to a lower temperature compared to that of benzene, the maximum value of soot yield remains practically the same. Regarding the question raised on the rate constant values, it would be more ap-

878

COMBUSTION-GENERATED PARTICULATES

propriate and consistent at this time t to compare values obtained using the same technique. Thus, the ratio of the decomposition of chloro-benzene and benzene based on the ARAS technique, 2 i.e. kzo/ k~,, is 10.6 at 1800 K or the temperature shift of the soot bell using these rate coefficients is 166 K, in good agreement with our estimate and experimental results.

REFERENCES 1. J. H. KIErER'S comment to Ref. 2 of this paper. 2. KERX, R. D. ET AL.: This symposium.

tt. Gg. Wagner, Univ. Gottingen, W. Germany. Could you please give the influence of 02 on the induction periods for soot formation in your systems?

Authors" Reply. Induction times for soot appearance and the influence of 02 on them are reported and discussed in Ref. 16 of the paper.