Evidence for the contribution of methane in the formation of aromatics and soot in fuel-rich pre-mixed combustion

Article abstract of DOI:10.1016/S0045-6535(00)00256-3

Experimental results from an isothermal laminar flow reactor at atmospheric pressure are presented on the chemical composition in the post-oxidative region of two sooting fuel-rich pre-mixed mixtures diluted in nitrogen. A base case composed of n-heptane and O₂ in N₂ at 1425 K with a C/O of 2.85 was perturbed by substituting 10% of the carbon in n-heptane with carbon as CH₄. While these changes would intuitively reduce aromatics and soot formation by increasing H₂ and decreasing C₂H₂ concentrations, we observe the opposite. The concentrations of individual aromatic species are observed to actually increase by up to 50% and the soot yield increases by 80%.

Full text of DOI:10.1016/S0045-6535(00)00256-3

Chemosphere 42 (2001) 823±826
Evidence for the contribution of methane in the formation of
aromatics and soot in fuel-rich pre-mixed combustion
*
J.F. Roesler , M. Auphan de Tessan, X. Montagne
Institut Fran ßc ais du P eꢀ trole, 1 et 4 av. de Bois Pr eꢀ au, 92852 Rueil-Malmaison Cedex, France
Abstract
Experimental results from an isothermal laminar ¯ow reactor at atmospheric pressure are presented on the chemical
composition in the post-oxidative region of two sooting fuel-rich pre-mixed mixtures diluted in nitrogen. A base case
composed of n-heptane and O in N
n-heptane with carbon as CH . While these changes would intuitively reduce aromatics and soot formation by in-
2 2
creasing H and decreasing C H concentrations, we observe the opposite. The concentrations of individual aromatic
2
2
at 1425 K with a C/O of 2.85 was perturbed by substituting 10% of the carbon in
4
2
species are observed to actually increase by up to 50% and the soot yield increases by 80%. Ó 2001 Elsevier Science
Ltd. All rights reserved.
Keywords: Soot; PAH; Combustion; Methane; Flow reactor
1
. Introduction
showing that its presence in the soot growth environ-
ment contributes to the growth rate of aromatic soot
pre-cursor compounds.
Past experimental evidence has shown that acetylene
and molecular hydrogen are controlling elements in the
reaction bath for polycyclic aromatic hydrocarbon
(
(
PAH) growth in fuel-rich pre-mixed combustion
Bockhorn, 1994). According to the HACA, polymeric
2. Experimental set-up
growth sequence model suggested by Frenklach et al.
1984), acetylene is the carbon-addition element and
The experiments are performed in an isothermal
laminar non-plug-¯ow reactor operating at atmospheric
pressures in a three-zone temperature controlled furnace
described elsewhere (Roesler, 1998; Roesler and Au-
phan-de-Tessan, 1998). Following rapid heat-up, the
98% nitrogen diluted pre-mixed reactants enter a 3 cm
diameter quartz duct and the ¯ow is laminarized by
passing through a porous quartz disk prior to entering
the 30 cm test section. The temperature pro®le measured
under a ¯ow of nitrogen is ¯at within 5 K. A fraction of
the combustion products (20%) are sampled along the
center-line axis and quenched by an oil-cooled probe at
(
molecular hydrogen is a species that tends to reduce the
growth process by competing for radical sites. Methane
is also an important product in fuel-rich combustion
systems whose role in the growth process of high mo-
lecular weight compounds has not been addressed in
much detail as it is the least sooting hydrocarbon. Fur-
thermore, Harris and Weiner (1983) have argued that
methane cannot be an important soot growth species in
pre-mixed atmospheric pressure ¯ames. The present
work provides new insight into the action of methane by
2
00°C. Soot is ®ltered at 200°C and condensable species
are then trapped. In a ®rst pass, the trap is at ambient
temperature and the low molecular weight gas phase
species, up to benzene, are analyzed on line with a FTIR
spectrometer. The higher molecular weight species are
*
Corresponding author. Tel.: +33-1-4752-5811; fax: +33-1-
752-6685.
E-mail address: john.roesler@ifp.fr (J.F. Roesler).
4
0
045-6535/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 0 0 ) 0 0 2 5 6 - 3

8
24
J.F. Roesler et al. / Chemosphere 42 (2001) 823±826
collected on a second pass with the trap at dry-ice
temperature ()70°C). All wetted surfaces (probe Te¯on
liner, soot ®lter and cold trap) are washed ultrasonically
The e€ect of changing the initial mixture composition
on the growth bath composition is only to alter methane
and acetylene concentrations. Methane is relatively slow
to oxidize and its addition yields a 70% self-increase in
BATH2 relative to BATH1. In contrast, the acetylene
concentration is reduced by 20% due to the reduction in
initial n-heptane. The global e€ect is therefore a substi-
2 2 2 2
with CH Cl . The detailed speciation of the CH Cl
soluble fraction is obtained by GC±FID and GC±MS
analysis. The total masses of condensable species (CS) of
molecular weight species greater than acenaphthylene
and the mass of soot (non-dichloromethane-soluble
condensable species) are determined gravimetrically.
tution of C
tions.
2 2
H with methane in equal carbon propor-
Based on the reaction systems described above one
would intuitively argue that, with less C and equal
concentrations, the formation of benzene, PAH and
soot should be reduced. Yet the experimental data in-
dicate just the opposite: the rate of formation of C
2
H
2
3
. Results and discussion
H
2
In order to bear some relation to a diesel fuel, the
initially injected hydrocarbon is n-heptane. Its purpose is
primarily to generate a C ±H (and now CH ) bath
6 6
H
increases by 60%. The concentrations of the PAH and
soot have been measured only at the 20 cm position and
increase in relative amounts that are species-dependent.
Naphthalene increases from 11 to 16 ppmv (‡45%),
acenaphthylene from 13 to 16 ppmv (‡23%), ¯uo-
ranthene from 1.3 to 1.5 ppmv (‡15%), pyrene from 2.1
to 3.2 ppmv (‡50%), CS from 0.29 to 0.33 mg/l in the
hot gases (‡15%), and soot from 0.048 to 0.084 mg/l of
soot (‡80%). The later increase is likely to be the cu-
mulated e€ect of a shortened induction period and faster
growth rate. Thus, globally the e€ect of the added
methane extends throughout the total chain of reactions
leading to soot.
2
H
2
2
4
for PAH and soot formation. Two mixtures are studied
at 1425 K with a bulk mean ¯ow velocity of 61 cm/s. The
®rst, BATH1 contains initial mole fractions of 0.49% n-
heptane and 0.6% O yielding an equivalence ratio of 9.0
2
and a C/O ratio of 2.85. BATH2 has the same total
carbon content and C/O ratio, but 10% of the carbon is
injected as methane instead of n-heptane.
Fig. 1 displays the measured species pro®les. The
decomposition of n-heptane is not observed in these
experiments because it occurs in the pre-heat region on a
time scale that is much shorter than that of PAH and
soot formation. The oxygen is also consumed rapidly in
These results concur with recent experiments that
showed a methane laminar pre-mixed ¯ame to produce
2 2
the pre-heat region. The pro®les of CO, H O and CO
are thus nearly ¯at. The molecular hydrogen content is
obtained by H-atom balancing on the measured species
and on the amount of injected n-heptane. Other species
would contribute negligibly. The rise in molecular hy-
drogen along the reactor tube results from the decay of
ethylene, methane and acetylene that are the three main
hydrocarbon species remaining at the end of the pre-
heat zone.
2 6
more PAH than an equivalent C H ¯ame (Senkan and
Castaldi, 1996). While the ¯ame data provide a global
di€erence between the two fuel types in a complex re-
action system with large temperature gradient, the pre-
sent results isolate the speci®c role of methane by
imposing only a perturbation in an isothermal reaction
bath. Senkan and Castaldi (1996) suggested that the
presence of methyl radicals promotes the formation of
Fig. 1. Mole fraction pro®les measured along the reactor axis. Dashed lines and open symbols are curve ®ts and data for BATH1 (no
added methane). Solid lines and ®lled symbols are curve ®ts and data for BATH2 (added methane).

J.F. Roesler et al. / Chemosphere 42 (2001) 823±826
825
Fig. 2. Calculated species pro®les using the detailed reaction mechanism of Marinov et al. (1996) to show that the observed experi-
mental e€ects are qualitatively reproduced. Calculations were performed assuming plug-¯ow conditions. BATH1 (no added methane):
dashed lines, BATH2 (added methane): solid lines.
hydrocarbon species containing an odd-number of car-
bon atoms, which would then accelerate benzene and
naphthalene production via mechanisms proposed by
Marinov et al. (1996, 1997, 1998). Accordingly, a pos-
sible mechanistic interpretation is that methane in-
creases the methyl radical concentration and
subsequently the following chain of reactions that lead
to benzene:
tatively capable of predicting the observed e€ects as
shown in Fig. 2: adding methane accelerates the growth
of aromatics. The calculations were performed using an
isothermal plug-¯ow reactor approximation with initial
conditions given by the measurements at the 10 cm
position as given in Fig. 1. All other species concen-
trations, including benzene, were initially set to zero.
Calculations with the Wang and Frenklach (1997) re-
action mechanism gave poorer predictions with still a
30% rise in benzene growth rate but with decreases in
naphthalene and pyrene. The later mechanism includes
benzene production via reaction (3), but does not in-
clude growth of PAH via odd-carbon numbered spe-
cies, which is likely to explain its poorer trends
predictions.
CH
3
‡ C
2
H
2
! pC H
4
ꢀpropyne† ‡ H
ꢀ1†
ꢀ2†
ꢀ3†
3
pC H
3
4
‡ H ! C
3
H
3
ꢀpropargyl† ‡ H
2
C
3
H
3
‡ C ! benzene
3 3
H
In conclusion, our experimental results demonstrate
that methane contributes signi®cantly in determining
the growth rate of aromatic species and soot in fuel-
rich pre-mixed combustion. Detailed modeling of
combustion chemistry is suciently well developed
that it can qualitatively predict the present observa-
tions. These models will help identify the implications
of the ®ndings on PAH and soot formation in prac-
tical combustion devices. More detailed experimental
and modeling work is under way to provide deeper
insights into the chemical role of methane identi®ed
here.
The additional benzene itself could drive the formation
of PAH and soot to higher concentrations via HACA
reaction steps (these involve even-carbon numbered
species), or other pathways involving methyl radicals
and odd-carbon numbered species may be involved such
as the formation of naphthalene via:
C
C
C
2
2
3
H
H
H
4
3
6
‡ CH
3
3
! C
3
3
H
H
6
6
‡ H
ꢀ4†
ꢀ5†
ꢀ6†
ꢀ7†
ꢀ8†
‡ CH
! C
! 1-C
‡ C
‡ cC
3
H
5
‡ H
Acknowledgements
1
-C
3
H
5
2
H
2
! cC
5
H
5
ꢀcyclopentadienyl† ‡ H
ꢀnaphthalene† ‡ H ‡ H
We wish to thank the ADEME (Agence De lÕEn-
vironnement et de la Ma ^õ trise de lÕEnergie) for the ®-
nancial support of M. Auphan de Tessan, and P.
Thoral for his technical help in running the experi-
ments.
cC
5
H
5
5
H
5
! C10
H
8
Calculations performed with the detailed kinetic re-
action mechanisms of Marinov et al. (1997) are quali-

8
26
J.F. Roesler et al. / Chemosphere 42 (2001) 823±826
Marinov, N.M., Pitz, W.J., Westbrook, C.K., Lutz, A.E.,
References
Vincitore, A.M., Senkan, S.M., 1998. Chemical kinetic
modeling of a methane opposed-¯ow di€usion ¯ame and
comparison to experiments. In: 27th Symposium (Interna-
tional) on Combustion, The Combustion Institute, Pitts-
burgh, PA, pp. 605±613.
Bockhorn, H., 1994. Soot Formation in Combustion ± Mech-
anisms and Models. Springer Series in Chemical Physics 59,
Springer, Heidelberg.
Frenklach, M., Clary, D.W., Gardiner, Jr., W.C., Stein, S.E.,
984. Detailed kinetic modeling of soot formation in shock-
Roesler, J.F., 1998. An Experimental and two-dimensional
investigation of combustion chemistry in a laminar non
plug-¯ow reactor. In: 27th Symposium (International) on
Combustion, The Combustion Institute, Pittsburgh, PA,
pp. 287±293.
1
tube pyrolysis of acetylene. In: 20th Symposium (Inter-
national) on Combustion, The Combustion Institute,
Pittsburgh, PA, pp. 887±901.
Harris, S.J., Weiner, A.M., 1983. Surface growth of soot
particles in premixed ethylene/air ¯ames. Combust. Sci.
Technol. 31, 155.
Roesler, J.F., Auphan-de-Tessan, M., 1998. Work-in-progress
poster W4A05, In: 27th Symposium (International) on
Combustion, Boulder, CO, USA, 2±7 August.
Marinov, N.M., Pitz, W.J., Westbrook, C.K., Castaldi, M.J.,
Senkan, S.M., 1996. Modeling of aromatic and polycyclic
aromatic hydrocarbon formation in premixed methane and
ethane ¯ames. Combust. Sci. Technol. 116/117, 211±287.
Marinov, N.M., Castaldi, M.J., Melius, C.F., Tsang, W., 1997.
Aromatic and polycyclic aromatic hydrocarbon formation
in a premixed propane ¯ame. Combust. Sci. Tech. 128, 295±
Senkan, S.M., Castaldi, M., 1996. Formation of polycyclic
aromatic hydrocarbons (PAH) in methane combustion:
comparative new results from premixed ¯ames. Combust.
Flame 107, 141±150.
Wang, H., Frenklach, M., 1997. A detailed kinetic modeling
study of aromatics formation in laminar premixed acetylene
and ethylene ¯ames. Combust. Flame 110, 173±221.
3
42.