Proceedings of the Combustion Institute, Volume 28, 2000/pp. 2585–2592
FORMATION PATHWAYS OF ETHYNYL-SUBSTITUTED AND CYCLOPENTAFUSED POLYCYCLIC AROMATIC HYDROCARBONS
NATHAN D. MARSH and MARY J. WORNAT
Princeton University
Department of Mechanical and Aerospace Engineering
Princeton, New Jersey 08544-5263, USA
Two novel classes of polycyclic aromatic hydrocarbons (PAH), those with ethynyl substituents (ethynylPAH) and those with externally fused five-membered rings (cyclopenta-fused PAH or CP-PAH), have
recently been identified in the products of a variety of fuels and combustion/pyrolysis environments.
However, the recently developed capacity for identifying these compounds has raised new questions about
preferential reaction pathways. Specifically, across various fuels and operating conditions, experimentally
observed products are (1) CP-PAH, which result from C2H2 addition to an aryl radical, followed by cyclization to a cyclopenta ring and (2) ethynyl-PAH, which result from C2H2 addition to locations on the aryl
radical where cyclization is not possible. We have never observed ethynyl-PAH resulting from C2H2 addition to an aryl radical at a point where cyclization into a five-membered ring is possible.
To explain this behavior, we have performed AM1 semiempirical quantum chemical computations with
group correction in order to examine the potential energy surfaces of the reaction pathways that lead to
ethynyl-PAH and CP-PAH. We have performed computations for the parent aryl radical, possible ethynylPAH products, possible CP-PAH products, as well as intermediates and transition states, for C2H2 addition
to naphthalene, anthracene, phenanthrene, acenaphthylene, fluoranthene, and pyrene. Possible CP-PAH
products are acenaphthylene, aceanthrylene, acephenanthrylene, pyracylene, cyclopenta[cd]fluoranthene,
and cyclopenta[cd]pyrene.
In all cases, we have found that, although energy differences between ethynyl-PAH isomers are very
small (⬃1 kcal/mol), the experimentally observed ethynyl-PAH is always the lowest energy isomer. Furthermore, the observed preference for cyclization to CP-PAH over formation of an ethynyl-PAH can be
explained by the significantly lower energy barrier (23 vs. 36 kcal/mol) for the cyclization reactions. Finally,
we have determined that, while not prohibited, the isomerization of ethynyl-PAH to CP-PAH requires
significantly higher energy than the aryl-vinyl cyclization reactions, and therefore is not expected to make
a significant contribution to the product distribution. These results are sufficiently consistent that the
computation of reaction pathway energy surfaces can be used to identify likely ethynyl-PAH and CP-PAH
products from the addition of C2H2 to much larger parent PAH.
Introduction
Several investigations [1–12] of the composition of
combustion and pyrolysis products from a variety of
fuels and experimental conditions have led to the
identification of a number of polycyclic aromatic hydrocarbons (PAH) with peripherally fused fivemembered rings and/or ethynyl substituents,
termed cyclopenta-PAH (CP-PAH) and ethynylPAH, respectively. The identifications of many of
these CP-PAH and ethynyl-PAH, however, have only
recently been made possible, due to the recent syntheses of authentic reference standards of such PAH
[3,7,11,13–16]. Because the means of unequivocally
identifying CP-PAH and ethynyl-PAH did not previously exist, these two classes of PAH, for the most
part, have gone unnoticed as fuel products and thus
are noticeably absent from theoretical models for
PAH growth and soot formation.
Despite their elusiveness, CP-PAH and ethynylPAH are very important, due to their biological and
chemical activities. Several studies [11,17–19] suggest that CP-PAH, as a class, are generally more mutagenic than other classes of PAH. In fact, Howard
et al. [20] and Durant et al. [17] have shown that the
CP-PAH cyclopenta[cd]pyrene is at least as biologically active as the well-known benzenoid PAH mutagen, benzo[a]pyrene. The enhanced biological activity of CP-PAH is apparently due to the relatively
facile metabolic epoxidation of the cyclopenta ring
[18], which results from the more localized electronic structure in this ring [19]. The electronic
structure of the cyclopenta ring may also account for
why CP-PAH are more susceptible to atmospheric
oxidation than benzenoid PAH [21].
CP-PAH are also of interest for their chemical behavior in combustion and pyrolysis environments.
For example, an increase in temperature brings
2585
2586
SOOT FORMATION AND DESTRUCTION
Fig. 1. Competing paths for the formation of ethynylPAH and CP-PAH. (i) C2H2 addition adjacent to a valley,
followed by H loss from the vinyl radical, resulting in 1ethynylnaphthalene; (ii) C2H2 addition adjacent to a valley,
followed by cyclization and H loss, resulting in acenaphthylene; (iii) C2H2 addition without an adjacent valley, leading only to 2-ethynylnaphthalene.
about an increase in both soot yield and in the CPPAH fraction of the total PAH [22]—a behavior that
suggests an interlinked chemistry between CP-PAH
and soot. The chemistry of CP-PAH also appears to
be linked to the formation of fullerenes, as shown by
Lafleur et al. [9]: fuels and reactor configurations
which produce fullerenes also produce corranulene,
a possible fullerene precursor, to the exclusion of
CP-PAH; however, fuels and reactor configurations
in which fullerenes are not detected produce an array of CP-PAH. This behavior is evidence of competing chemical pathways that are influenced by the
boundary conditions and/or the initial fuel.
The literature [23,24] suggests that CP-PAH form
from the addition of C2H2 to aryl radicals abundantly
present in combustion and pyrolysis environments.
However, as these works [23,24] show, C2H2 addition does not necessarily result in cyclopenta fusion;
it can also yield ethynyl-substituted PAH. Fig. 1 illustrates the possible results [24] of C2H2 addition
to the two aryl radicals that come from H abstraction
of naphthalene, the 1-naphthyl and 2-naphthyl radicals. In row (i) of Fig. 1, C2H2 addition to the 1naphthyl radical is followed by loss of H, resulting
in 1-ethynylnaphthalene. Alternatively, as shown in
row (ii) of Fig. 1, the same sequence of C2H2 addition and H loss can lead to cyclization across a “valley” to produce acenaphthylene. In contrast, as
shown in row (iii) of Fig. 1, C2H2 addition to the 2naphthyl radical can only yield 2-ethynylnaphthalene, as there is no valley-adjacent carbon available
for cyclization. Fig. 1 thus demonstrates that C2H2
addition to aryl radicals has the potential of producing a mixture of CP-PAH and isomeric ethynylPAH—the complexity of this mixture depending on
the number of possible aryl radicals to which the
C2H2 can add.
Fig. 2. Segment of a reverse-phase HPLC chromatogram of products of brown coal pyrolysis at 1000 ⬚C. Data
extracted from Refs. [1] and [5]. Structures of identified
components shown here (in order of elution, from left to
right) are acenaphthylene, 2-ethynylnaphthalene, 1-ethynylacenaphthylene, acephenanthrylene, aceanthrylene (coeluting with fluoranthene, not shown), 2-ethynylphenanthrene, 2-ethynylanthracene, cyclopenta[cd]fluoranthene,
and cyclopenta[cd]pyrene. Other components of this product sample are identified in Refs. [28–31].
The multiplicity of isomeric ethynyl-PAH and CPPAH structures, of course, poses serious challenges
to the separation and identification of constituents
of fuel product mixtures, since most analytical techniques rely on chromatographic and spectrometric
methods that may not be capable of distinguishing
between isomers. Nevertheless, we have shown that
reverse-phase high-pressure liquid chromatography
(HPLC) coupled with ultraviolet-visible absorption
spectroscopy is capable of distinguishing isomers,
both CP-PAH from ethynyl-PAH, and ethynyl-PAH
from each other [3,25–27]. As we have recently observed [1–5], the application of this analytical technique to fuel products yields results that expose a
whole new set of questions as to why certain products are common to a variety of fuels and systems,
while others prove never to be observed.
Figure 2 presents the portion of an HPLC chromatogram depicting the two- to five-ring PAH produced from the pyrolysis of a brown coal at 1000 ⬚C
[1,5,28–30]. Although over 50 PAH product species
have been unequivocally identified [5,28–31] in this
product sample, the only product structures drawn
in Fig. 2 are those of the CP-PAH and ethynyl-PAH
discussed in this paper. Fig. 2 shows that—of the
three possible products of C2H2 addition to naphthalene’s aryl radicals (shown in Fig. 1)—only acenaphthylene and 2-ethynylnaphthalene are present
among the brown coal pyrolysis products; 1-ethynylnaphthalene is not [5]. This observation suggests
that the reactions depicted in rows (ii) and (iii) of
Fig. 1 are active; the one in row (i) is not. Similarly,
Fig. 2 reveals that for C2H2 addition to anthracene’s
FORMATION PATHWAYS OF EXOTIC PAH
2587
aryl radicals, aceanthrylene and 2-ethynylanthracene
are present; 1-ethynylanthracene is not. For phenanthrene’s aryl radicals, acephenanthrylene and 2ethynylphenanthrene are present; 9-ethynylphenanthrene is not [5].
All of these cases—and the other products illustrated in Fig. 2—adhere to the following two rules:
(1) the only ethynyl-PAH observed are those resulting from C2H2 addition at a position that does not
allow for cyclization to a five-membered ring; and
(2) in any case where C2H2 is added at a position
that does allow for cyclization, the cyclization product is observed, and the ethynyl-PAH is not. Not only
are these two rules exhibited by the products of
brown coal pyrolysis in Fig. 2, but they are also consistent with the CP-PAH and ethynyl-PAH product
distributions obtained from the other fuels we have
examined—a high-rank coal [4,5], anthracene [2],
and benzene [3]—all of which have been pyrolyzed
or combusted in different reactor systems, spanning
a wide range of reaction conditions. This uniformity
of behavior suggests that there is some chemical
preference for the formation of the cyclopenta ring
over that of the ethynyl group. Sarobe et al. [32–35]
have investigated various isomerizations between
ethynyl-PAH and CP-PAH. However, with the exception of acenaphthylene [36], the potential energies of the formation paths of cyclopenta-PAH and
ethynyl-PAH via C2H2 addition have never, until
now, been examined.
In the following, we perform semiempirical quantum chemical computations of the heats of formation of intermediates, products, and transition states,
in order to examine the competing pathways for the
formation of several CP-PAH—acenaphthylene,
aceanthrylene, acephenanthrylene, pyracylene,
cyclopenta[cd]fluoranthene, and cyclopenta[cd]pyrene—and the corresponding ethynyl-PAH. Identification of the theoretically preferred pathways may
permit explanation of the qualitatively uniform CPPAH and ethynyl-PAH product distribution patterns
we observe experimentally for a wide range of fuels
and combustion/pyrolysis conditions.
keyword. Dissociation and recombination reactions
are optimized to transition state (keyword TS) by
defining the bond to be broken as the reaction coordinate.
Wang and Frenklach [39] showed that semiempirical computations introduce increasingly large errors as aromatic rings are added, resulting from approximations of the interatomic potential profile.
This error is particularly problematic in the case of
peripheral five-membered rings, exactly the compounds we examine here. For acenaphthylene, the
simplest CP-PAH, the DfH⬚ is overpredicted by 25%.
Fortunately, Wang and Frenklach [39] also show that
these errors are systematic and can be corrected to
within a few kilocalories for the species they study.
Their methodology is used here.
The method is based on the group additivity
method of Benson [40] but makes the improvement
of using group characteristics for corrections to the
computed DfH⬚, rather than simply summing group
contributions to achieve the final value. In the
method of Wang and Frenklach [36], each benzenoid C atom is classified as either A, peripheral and
bonded to H; B, peripheral but bonded to two type
As and one non-A; C, peripheral but bonded to one
type A and two non-As; or D, internal and not
bonded to any type As. Wang and Frenklach developed correction parameters for these atom types, as
well as for numerous substituent groups and radicals, so that
Theoretical Approach
Results and Discussion
Because of the large number and sizes of species
examined here, we have chosen to compute properties using semiempirical quantum methods. Heats
of formation are determined for all species and transition states using the AM1 Hamiltonian [37] of the
MOPAC 97 [38] program. Species are first optimized for lowest-energy geometries using the keyword GNORM ⳱ 0.1. More stringent criteria including keywords GNORM ⳱ 0.01 and PRECISE
were examined but did not produce significant improvements in the calculated energies. Transition
states of unimolecular reactions are determined using the SADDLE keyword and refined with the TS
Figures 3–8 show the computed potential energy
paths for the addition of C2H2 to naphthalene, anthracene, phenanthrene, acenaphthylene, fluoranthene, and pyrene, respectively. As shown in the figures for each of these parent PAH, the initial step of
the reaction is the addition of C2H2 to selected aryl
radicals of the parent PAH, forming PAH with terminal vinyl radicals. Once formed, these aryl-vinyl
radicals can then follow any one of the following
three routes: Pathway A, H removal from the vinyl
radical to form the ethynyl-PAH; Pathway B, cyclization [36] of the vinyl radical (where a “valley”-adjacent carbon is available) to form an aryl radical with
DfH⬚298 ⳱ DfH⬚298(AM1) ⳮ RGCi
(1)
where GCi is the correction parameter for that
group. We have devised an additional group correction of 14.1 kcal/mol to account for internal fivemembered rings like that in fluoranthene, based on
the measured DfH⬚ of fluoranthene of 69.2 kcal/mol
[41]. The remaining group correction parameters
used in this study, along with additional parameters
for other groups, can be found in Ref. [39]. Discussion of the uncertainties in this method can be found
in Refs. [36] and [39].
2588
SOOT FORMATION AND DESTRUCTION
Fig. 3. Potential energy surface of the formation of ethynylnaphthalenes and acenaphthylene from naphthyl radicals. Solid line, 1-naphthyl parent; dotted line, 2-naphthyl
parent. Species in bold (2-ethynylnaphthalene and acenaphthylene) have been identified in fuel products. Pathways A, B, and C (defined in text) are indicated.
Fig. 4. Potential energy surface of the formation of ethynylanthracenes and aceanthrylene from anthryl radicals.
Solid line, 1-anthryl parent; dotted line, 2-anthryl parent;
dotted-dashed line, 9-anthryl parent. Species in bold (2ethynylanthracene and aceanthrylene) have been identified
in fuel products. Pathways A, B, and C (defined in text)
are indicated.
an ethylene bridge, followed by H loss from the base
of the ethylene bridge to form the cyclopenta-PAH;
or Pathway C, hydrogen migration to the vinyl group
[42], followed by cyclization to the just-vacated radical site, and then H loss from the saturated corner
of the five-membered ring. Ethynyl-PAH result from
Pathway A; CP-PAH result from Pathways B and C.
We now examine each of these steps in turn for their
significance in the formation of ethynyl-substituted
and cyclopenta-fused PAH.
As suggested by their starting points, the first step
in all of the reaction schemes of Figs. 3–8 is actually
the formation of the PAH aryl radicals. We have chosen to begin our analyses with the radicals already
Fig. 5. Potential energy surface of the formation of ethynylphenanthrenes and acephenanthrylene from phenanthryl radicals. Solid line, 1-phenanthryl parent; dotted line,
2-phenanthryl parent; dotted-dashed line, 9-phenanthryl
parent. Species in bold (2-ethynylphenanthrene and acephenanthrylene) have been identified in fuel products.
Pathways A, B, and C (defined in text) are indicated.
Fig. 6. Potential energy surface of the formation of ethynylacenaphthylenes and pyracylene from acenaphthyl radicals. Solid line, 5-acenaphthyl parent; dotted line, 1-acenaphthyl parent; dotted-dashed line, 4-acenaphthyl parent.
Species in bold (1-ethynylacenaphthylene) have been identified in fuel products. Pathways A, B, and C (defined in
text) are indicated.
in place, however, because the radicals can be
formed in a variety of ways: abstraction by H atoms
in a pyrolysis environment, OH attack in a combustion environment, or even the artifact of a previous
cyclization-type step, for example, the formation of
naphthalene from the cyclization of 1-ethynyl-2-vinylbenzene. Of primary interest are the energies of
the radicals themselves, particularly any differences
resulting from radical location. In fact, radical location has very little effect on the energy of the radical
PAH. The 1-, 2-, and 9-phenanthryl radicals (Fig. 5)
are virtually indistinguishable, while in the cases of
FORMATION PATHWAYS OF EXOTIC PAH
Fig. 7. Potential energy surface of the formation of ethynylfluoranthenes and cyclopenta[cd]fluoranthene from
fluoranthyl radicals. Solid line, 3-fluoranthyl parent; dotted
line, 2-fluoranthryl parent. Species in bold (cyclopenta [cd]
fluoranthene) have been identified in fuel products. Pathways A, B, and C (defined in text) are indicated.
Fig. 8. Potential energy surface of the formation of ethynylpyrenes and cyclopenta[cd]pyrene from pyrenyl radicals. Solid line, 1-pyrenyl parent; dotted line, 2-pyrenyl parent; dotted-dashed line, 4-pyrenyl parent. Species in bold
(cyclopenta [cd] pyrene) have been identified in fuel products. Pathways A, B, and C (defined in text) are indicated.
the naphthyl (Fig. 3) and fluoranthyl (Fig. 7) radicals, the 2-isomer, which can only lead to the ethynyl-substituted isomer, has a negligibly lower energy.
The 9-anthryl radical (Fig. 4) has a marginally higher
energy, likely the result of “crowding” by the adjacent H atoms on the zigzag edge. Only acenaphthyl
(Fig. 6) exhibits a significant energy difference, with
1-acenaphthyl lying 8 kcal/mol above the other acenaphthyl radicals. This is most likely a result of the
more localized electronic structure of the ethylene
bridge [19,25], resulting in a slightly higher bond
dissociation energy at this site.
The fact that the energies of the various isomers
of the aryl radicals are nearly indistinguishable
means that the preferential formation of certain
2589
ethynyl-PAH isomers does not result from a preferential formation of particular aryl radicals. In fact,
examination of the rest of Pathway A, the addition
of C2H2 to form aryl-vinyl radicals followed by loss
of H from the vinyl radical to produce the ethynylPAH, shows that the entire path is energetically similar for different substituent locations, as can be seen
in Figs. 3–8. However, despite the small differences
in energy, a common trend arises: the aryl-vinyl radicals that have the opportunity to cyclize into a CPPAH have slightly higher energies than those that
lack a cyclization route (likely a result of the more
proximate H atom at the valley-adjacent site). For
example, in Fig. 3, 1-vinylnaphthyl lies 2 kcal/mol
higher than 2-vinylnaphthyl; in Fig. 4, 1- and 9-vinylanthryl lie 2 and 3 kcal/mol higher than 2-vinylanthryl. This trend is repeated for the ethynyl-PAH
themselves; for example, 1-ethynylnaphthalene has
a slightly higher energy than 2-ethynylnaphthalene
(Fig. 3), and 1- and 9-ethynylanthracene have
slightly higher energies than 2-ethynylanthracene
(Fig. 4). Furthermore, all of the ethynyl-PAH that
have been experimentally identified in fuel products
are those with the marginally lower energy: 2-ethynylnaphthalene [1,3,4,12,28], 2-ethynylanthracene
[2,5], 2-ethynylphenanthrene [5], and 1-ethynylacenaphthylene [1–4,28]. However, the small energy
difference is not sufficient to explain the complete
lack of slightly higher-energy ethynyl-PAH isomers.
The energies of the cyclization routes provide a
more convincing explanation. We have examined
two possible paths for this process. The first, Pathway B [36], is a two-step process where the vinyl
radical displaces the adjacent H atom and closes the
ring, resulting in a relatively low-energy aryl radical,
followed by a loss of the displaced H atom to form
the CP-PAH. The second path, Pathway C [42], involves a shift of the adjacent H atom to the terminal
end of the vinyl radical, resulting in a radical at the
ring-closure site. The terminal vinyl group then
closes on the new radical site, resulting in a partially
hydrogenated ethylene bridge. Loss of the extra H
atom again results in the CP-PAH.
In all the cases examined here, the largest energy
barrier to the formation of the Pathway B intermediate is little more than half that of the barrier to
form the ethynyl-PAH, by Pathway A. As shown by
Frenklach et al. [42], the barrier for the H migration,
Pathway C, is significantly smaller than that of the
two-step process, nearly half again as much. Furthermore, the barrier for the ring closure step of
Pathway C is nearly identical to that of the ringclosure step in Pathway B. Geometry minimizations
of Pathway B show that the displaced H atom is ultimately located perpendicular to the plane of the
PAH, often with a significant amount of twisting to
the entire molecule. Despite this strained configuration, however, the Pathway B intermediate has a
relatively low energy. Finally, we observe that the
2590
SOOT FORMATION AND DESTRUCTION
Fig. 9. Potential energy surface of the formation of acenaphthylene from 1-ethynylnaphthalene via a carbene biradical intermediate, compared with the path via a vinyl
radical intermediate. Species in bold (acenaphthylene)
have been identified in fuel products. Pathways B and C
(defined in text) are indicated.
product of the H migration step of Pathway C, an
aryl-ethylene, is usually of significantly higher energy
than that of the ring closure intermediate in Pathway
B, except in the cases of acenaphthylene (Fig. 6) and
fluoranthene (Fig. 7).
The acenaphthylene and fluoranthene cases are
also the exceptions in the final step, H loss resulting
in the CP-PAH, which have barriers that are 6 and
4 kcal/mol higher, respectively, than any other step
in Pathway B or C, the formation of the cyclopenta
ring. For the other CP-PAH, Figs. 3–5 and Fig. 8,
the barriers for the final H loss are 6–10 kcal/mol
lower than the highest barriers in the paths to CPPAH. It should be noted here that pyracylene, the
predicted cyclization product of C2H2 addition to
acenaphthylene, has not been observed as a fuel
product in any work that has been able to distinguish
it from the ethynylacenaphthylene isomers. The high
barriers to the formation of pyracylene, about the
same as the barriers to form ethynylacenaphthylene,
might indicate the difficulty of this step. On the
other hand, pyracylene is known to readily decompose [34], so its absence as a product might be explained by destruction routes not examined here.
The DfH⬚298 of pyracylene has been measured experimentally as 98 kcal/mol [43], which agrees within a
few kcal/mol with ab initio results [44] and the value
used here (103 kcal/mol).
Finally, we examine an alternate and frequently
invoked path for the formation of cyclopenta rings
[13,32–34,45,46], shown (for the formation of acenaphthylene only) in Fig. 9. The pathways depicted
in Fig. 2 are repeated here for comparison. The additional path allows for conversion of ethynylnaphthalene into acenaphthylene via a carbene biradical, formed by a 1,2-shift of hydrogen on the
appropriate ethynyl-PAH. The carbene biradical
then inserts into the adjacent C–H bond, forming a
five-membered cyclopenta ring. This reaction path
is invoked most often for the unimolecular isomerization of ethynyl-PAH into CP-PAH in synthetic
preparatory schemes [13,32–34,45,46], for example,
flash vacuum thermolysis. As can be seen in Fig. 9,
the DfH⬚298 of the biradical intermediate is significantly larger than the barriers for the paths examined
previously. However, all points on this path include
an additional 52.1 kcal/mol to account for the free
H atom, to remain consistent with the energies given
for the other paths. In other words, in systems where
free H is not significant (for example, synthetic preparation via unimolecular isomerization), formation of
the biradical is not unreasonable. Even in combustion environments, the unimolecular reaction should
not be prohibited, but the energy is sufficiently high
that it should not be significant when compared with
the other routes examined here.
Summary and Conclusions
To better understand the mechanisms of formation of ethynyl-PAH and CP-PAH, we have performed AM1 semiempirical quantum chemical computations with group correction on paths of the
potential energy surfaces of these compounds. We
have focused on groups of isomers that we have previously identified in real combustion and pyrolysis
products, and have included related species that
were not evident in those products, in order to understand why certain compounds appear preferentially over others. For each potential energy surface,
we have examined the precursor aryl radicals, the
ethynyl-PAH and CP-PAH isomers, and the vinyl
radical species and intermediates, as well as the transition states for each step. We have employed the
group correction method particularly to avoid the
large errors in the AM1 computation of cyclopenta
rings; this method is sufficiently robust that the
DfH⬚298 found for the single dicyclopenta species,
pyracylene, agrees exactly with ab initio calculations
and is within a few kcal/mol of the experimental
value.
We have found that the location of the initial radical site makes little difference in the energy of the
reaction path, suggesting that the formation of the
initial radical site does not control the nature of the
product of the reaction. Rather, the energy barrier
for the formation of the ethynyl-substituted isomers
is always much larger than that for the formation of
intermediates to the CP-PAH—strong evidence that
the CP-PAH should be preferentially formed when
that path is available. In only two cases, pyracylene
and cyclopenta[cd]fluoranthene, is the barrier for
the formation of the CP-PAH anywhere close to the
barrier for the formation of the ethynyl-PAH, and of
FORMATION PATHWAYS OF EXOTIC PAH
these two, only the one with the lower barrier, cyclopenta[cd]fluoranthene, is ever found in any products. Finally, we have found that the isomerization
of ethynyl-PAH to CP-PAH is a relatively high-energy process, suggesting that it is not the primary
source of CP-PAH when lower-energy routes are
available.
In our examination of C2H2 addition to six different parent PAH, we have found that the potential
energy surfaces developed by AM1 computation
with group correction accurately predict the preferred isomers experimentally observed in the products of a variety of fuels and configurations. Because
we have evidence that ethynyl-PAH and CP-PAH
are not limited to the small parent PAH examined
here [47], these kinds of calculations will provide a
useful tool for determining which larger ethynylPAH and CP-PAH isomers are most likely to occur.
In conjunction with analytical methods such as spectroscopy [25,26] and chromatography [27], the identification of previously unknown compounds of these
classes can be greatly facilitated. Furthermore, the
elucidation of the participation of ethynyl substitution and cyclopenta fusion in PAH growth will improve detailed kinetic models which attempt to accurately reflect real PAH yields and distributions, by
adding appropriate product species which would
otherwise go unnoticed.
Acknowledgments
We gratefully acknowledge the National Science Foundation and the Princeton University Research Board for
financial support of portions of this work.
REFERENCES
1. Wornat, M. J., Vernaglia, B. A., Lafleur, A. L., Plummer, E. F., Tagizadeh, K., Necula, A., and Scott, L. T.,
Proc. Combust. Inst. 27:1677–1686 (1998).
2. Wornat, M. J., Vriesendorp, F. J. J., Lafleur, A. L.,
Plummer, E. F., Necula, A., and Scott, L. T., Poly.
Arom. Comp. 13:221–240 (1999).
3. Marsh, N. D., Wornat, M. J., Scott, L. T., Necula, A.,
Lafleur, A. L., and Plummer, E. F., Poly. Arom. Comp.
13:379–402 (2000).
4. Ledesma, E. B., Kalish, M. A., Nelson, P. F., Wornat,
M. J., and Mackie, J. C., Energy Fuels 13:1167–1172
(1999).
5. Wornat, M. J., and Ledesma, E. B., Poly. Arom.
Comp., in press (1999).
6. Grieco, W. J., Lafleur, A. L., Swallow, K. C., Richter,
H., Taghizadeh, K., and Howard, J. B., Proc. Combust.
Inst. 27:1669–1675 (1998).
7. Lafleur, A. L., Howard, J. B., Plummer, E. F., Taghizadeh, K., Necula, A., Scott, L. T., and Swallow, K. C.,
Poly. Arom. Comp. 12:223–237 (1998).
2591
8. Benish, T. G., Lafleur, A. L., Taghizadeh, K., and Howard, J. B., Proc. Combust. Inst. 26:2319–2326 (1996).
9. Lafleur, A. L., Howard, J. B., Taghizadeh, K., Plummer, E. F., Scott, L. T., Necula, A., and Swallow, K.
C., J. Phys. Chem. 100:17421–17428 (1996).
10. Lafleur, A. L., Taghizadeh, K., Howard, J. B., Anacleto,
J. F., and Quilliam, M. A., J. Am. Soc. Mass Spectrom.
7:276–286 (1996).
11. Lafleur, A. L., Longwell, J. P., Marr, J. A., Monchamp,
P. A., Plummer, E. F., Thilly, W. G., Mulder, P. P. Y,
Boere, B. B., Cornelisse, J., and Lugtenburg, J., Environ. Health Perspect. 101:146–153 (1993).
12. Griesheimer, J., and Homann, K.-H., Proc. Combust.
Inst. 27:1753–1759 (1998).
13. Scott, L. T., and Necula, A., J. Org. Chem. 61:386–388
(1996).
14. Crisp, G. T., and Jiang, Y.-L., Synth. Commun.
28:2571–2576 (1998).
15. Hopkins, N. E., Foroozesh, M. K., and Alworth, W. L.,
Biochem. Pharmacol. 44:787–796 (1992).
16. Hall, M., Parker, D. K., Grover, P. L., Lu, J.-Y. L.,
Hopkins, N. E., and Alworth, W. L., Chem.-Biol. Interact. 76:181–192 (1990).
17. Durant, J. L., Busby Jr., W. F., Lafleur, A. L., Penman,
B. W., and Crespi, C. L., Mutat. Res. 371:123–157
(1996).
18. Nesnow, S., Leavitt, S., Easterling, R., Watts, R., Toney, S. H., Claxton, L., Sangaiah, R., Toney, G. E.,
Wiley, J., Fraher, P., and Gold, A., Cancer Res.
44:4993–5003 (1984).
19. Fu, P. P., Beland, F. A., and Yang, S. K., Carcinogenesis
1:725–727 (1980).
20. Howard, J. B., Longwell, J. P., Marr, J. A., Pope, C. J.,
Busby Jr., W. F., Lafleur, A. L., and Taghizadeh, K.,
Combust. Flame 101:262–270 (1995).
21. Hites, R. A., in Atmospheric Aerosols: Source/Air
Quality Relationships, ACS Symposium Series 167,
American Chemical Society, Washington, DC, 1981,
pp. 187–196.
22. Wornat, M. J., Sarofim, A. F., and Lafleur, A. L., Proc.
Combust. Inst. 24:955–963 (1992).
23. Bockhorn, H., Fetting, F., and Wenz, H. W., Ber. Bunsen-Ges. Phys. Chem. 87:1067–1073 (1983).
24. Frenklach, M., Clary, D. W., Gardiner Jr., W. C., and
Stein, S. E., Proc. Combust. Inst. 20:887–901 (1984).
25. Marsh, N. D., Mikolajczak, C. J., and Wornat, M. J.,
Spectrochimica Acta, Part A 56:1499–1511 (2000).
26. Marsh, N. D., and Wornat, M. J., Poly Arom. Comp.,
in press (1999).
27. Ledesma, E. B., and Wornat, M. J., Anal. Chem.
72:5437–5443 (2000).
28. Miliou, A., “The Effects of Ion-Exchanged Metals on
the Pyrolysis of Coal,” B.S.E. thesis, Department of
Mechanical and Aerospace Engineering. Princeton
University, Princeton, NJ, 1998.
29. Vernaglia, B. A., “The Effects of Temperature and IonExchanged Metals on the Composition of Brown Coal
Pyrolysis Tars,” M.S.E. thesis, Department of Mechanical and Aerospace Engineering Princeton University, Princeton, NJ, 1997.
2592
SOOT FORMATION AND DESTRUCTION
30. Vernaglia, B. A., Wornat, M. J., Li, C.-Z., and Nelson,
P. F., Proc. Combust. Inst. 26:3287–3294 (1996).
31. Wornat, M. J., Mikolajczak, C. J., Vernaglia, B. A., and
Kalish, M. A., Energy Fuels 13:1092–1096 (1999).
32. Sarobe, M., Zwikker, J. W., Snoeijer, J. D., Wiersum,
U. E., and Jenneskens, L. W., J. Chem. Soc., Chem.
Commun. 1994:89–90 (1994).
33. Sarobe, M. Snoeijer, J. D., Jenneskens, L. W., Slagt,
M. Q., and Zwikker, J. W., Tetrahedron Lett. 36:8489–
8492 (1995).
34. Sarobe, M., Flink, S., Jenneskens, L. W., Zwikker, J.
W., and Wesseling, J., J. Chem. Soc., Perkin Trans.
2:2125–2131 (1996).
35. Sarobe, M., Jenneskens, L. W., Wesseling, J., Snoeijer,
J. D., Zwikker, J. W., and Wiersum, U. E., Liebigs
Ann./Recueil 1997:1207–1213 (1997).
36. Wang, H., and Frenklach, M., J. Phys. Chem.
98:11465–11489 (1994).
37. Dewar, M. J. S., Zoebisch, E. G., Healy, F., and Stewart, J. J. P., J. Am. Chem. Soc. 107:3902–3909 (1985).
38. Stewart, J. J. P., Quantum Chemistry Exchange Program, No. 455, Indiana University, Bloomington, IN,
1983; CS Chem3D Pro 4.0, CambridgeSoft, Cambridge, MA, 1997.
39. Wang, H., and Frenklach, M., J. Phys. Chem. 97:3867–
3874 (1993).
40. Benson, S. W., Thermochemical Kinetics, 2nd ed., Wiley, New York, 1976.
41. Lias, S. G., Bartmess, J. E., Liebman, J. F., Holmes,
J. L., Levin, R. D., and Mallard, W. G., J. Phys. Chem.
Ref. Data 17 (Suppl. 1) (1988).
42. Frenklach, M., Moriarty, N. W., and Brown, N. J.,
Proc. Combust. Inst. 27:1655–1661 (1998).
43. Diogo, H. P., Persy, G., de Piedade, M. E. M., and
Wirz, J., J. Org. Chem. 61:6733–6734 (1996).
44. Schulman, J. M., and Disch, R. L., J. Mol. Struct. Theochem. 259:173–179 (1992).
45. Brown, R. F. C., and Eastwood, F. W., Synlett 1993:9–
19 (1993).
46. Brown, R. F. C., Eastwood, F. W., and Jackman, G. P.,
Aust. J. Chem. 30:1757–1767 (1977).
47. Wornat, M. J., Vriesendorp, F. J. J., Necula, A., Scott,
L. T., Lafleur, A. L., Plummer, E. F., Taghizadeh, K.,
and Swallow, K. C., “The Identification of Two New
Dicyclopenta-Fused Polycyclic Aromatic Hydrocarbons in the Pyrolysis and Combustion Products of a
Variety of Fuels,” paper no. 166, Seventeenth International Symposium on Polycyclic Aromatic Compounds, Bordeaux, France, October 25–29, 1999.