Shock Tube Study of Thermal Rearrangement of 1,5-Hexadiyne over Wide Temperature and Pressure Regime

Article abstract of DOI:10.1021/jp037310z

The pyrolysis of 1,5-hexadiyne has been studied in a high-pressure single pulse shock tube to investigate the mechanisms involved in the production of benzene from propargyl radicals. Analysis of the reaction products by gas chromatography and matrix isolation Fourier transform infrared spectroscopy has positively identified six linear C₆H₆ species and two cyclic C₆H₆ species. Of these species cis-1,3-hexadien-5- yne and trans-1,3-hexadiene-5-yne have been unambiguously identified for the first time and provide vital information concerning a low-temperature route to benzene that does not involve the formation of fulvene; however, the data also provide support for two high-temperature paths from propargyl radicals to benzene via fulvene. Thus experimental evidence has been gained that supports two different routes to benzene formation. The mechanisms and rate coefficients that have been obtained in this work are discussed.

Full text of DOI:10.1021/jp037310z

3406
J. Phys. Chem. A 2004, 108, 3406-3415
Shock Tube Study of Thermal Rearrangement of 1,5-Hexadiyne over Wide Temperature
and Pressure Regime
Robert S. Tranter,§,| Weiyong Tang,§ Ken B. Anderson,†,‡ and Kenneth Brezinsky*,§
Departments of Mechanical Engineering and Chemical Engineering, UniVersity of Illinois at Chicago,
842 West Taylor Street, M/C 251, Chicago, Illinois 60607, and Chemistry DiVision, Argonne National
Laboratory, Argonne, Illinois 60439
ReceiVed: October 31, 2003; In Final Form: January 30, 2004
The pyrolysis of 1,5-hexadiyne has been studied in a high-pressure single pulse shock tube to investigate the
mechanisms involved in the production of benzene from propargyl radicals. Analysis of the reaction products
by gas chromatography and matrix isolation Fourier transform infrared spectroscopy has positively identified
six linear C6H6 species and two cyclic C6H6 species. Of these species cis-1,3-hexadien-5-yne and trans-1,3-
hexadiene-5-yne have been unambiguously identified for the first time and provide vital information concerning
a low-temperature route to benzene that does not involve the formation of fulvene; however, the data also
provide support for two high-temperature paths from propargyl radicals to benzene via fulvene. Thus
experimental evidence has been gained that supports two different routes to benzene formation. The mechanisms
and rate coefficients that have been obtained in this work are discussed.
Introduction
SCHEME 1: Chemical Reaction Pathways Involving
Stable Species for the Thermal Isomerizations of
1,5-Hexadiynea
Understanding the key routes to benzene formation in sooting
flames1-6 is important because it is via growth from small
aromatic species that the production of soot is thought to occur.
The recombination of propargyl radicals, reaction 1, is purported
to be a primary route to benzene formation.
The experimental7-11 and theoretical12-19 studies that have
been made on propargyl recombination show that reaction 1 is
a gross simplification of a complex series of simultaneous and
consecutive reactions that are predominantly isomerizations of
C6H6 species. Initially, two propargyl radicals can recombine
to form three stable linear C6H6 species, as shown below.
a The length of the arrow does not indicate the significance of the
path. Abbreviations: 15HD, 1,5-hexadiyne; 1245HT, 1,2,4,5-hexa-
tetraene; 12HD5Y, 1,2-hexadiene-5-yne; 34DMCB, 3,4-dimethylene-
cyclobutene; 13HD5Y, 1,3-hexadiene-5-yne; 2E13BD, 2-ethynyl-1,3-
butadiene.
2HCtCCH2 f
HCtCCH2CH2CtCH (1,5-hexadiyne, 15HD) (2)
2HCtCCH2 f
H2CdCdCHCHdCdCH2 (1,2,4,5-hexatetraene, 1245HT)
theoretical studies suggest that at very high temperatures one
of the metastable intermediates involved in benzene formation
can decompose to phenyl + H without first forming benzene.19
The major stable species involved in the various reaction paths
from propargyl radicals to benzene are summarized in Scheme
1, which is based on the extant theoretical and experimental
studies of propargyl recombination and the related pyrolysis of
1,5-hexadiyne.
(3)
2HCtCCH2 f
H2CdCdCHCH2CtCH (1,2-hexadiene-5-yne, 12HD5Y)
(4)
These species then isomerize to ultimately form benzene, which
at high enough temperatures and long enough reaction times
can decompose to a phenyl radical and an H-atom. Additionally,
The experimental investigations of propargyl recombination
have employed a variety of techniques and sources for producing
propargyl radicals. There is general agreement that the products
of reactions 2-4 are the primary recombination products and
that these species can interconvert and isomerize to secondary
products that ultimately lead to benzene. Alkemade and
Homann7 found 12HD5Y, reaction 4, to be the major recom-
* Corresponding author. E-mail: Kenbrez@UIC.edu.
† Present address: Department of Geology, Southern Illinois University,
Carbondale, IL, 62901.
‡ Argonne National Laboratory.
§ University of Illinois at Chicago.
| Present address: Argonne National Laboratory.
10.1021/jp037310z CCC: $27.50 © 2004 American Chemical Society
Published on Web 03/30/2004

Thermal Rearrangement of 1,5-Hexadiyne
J. Phys. Chem. A, Vol. 108, No. 16, 2004 3407
bination product and only a few percent of 15HD, reaction 1,
at 623-673 K and 2-4 Torr, and are the only group to have
directly observed and identified 1245HT, reaction 3, as a
recombination product of propargyl radicals. In addition to the
primary products, Alkemade and Homann observed 1,3-hexa-
diene-5-yne (13HD5Y) and benzene as secondary products.
Benzene appeared to be formed consecutively from a linear
C6H6 species, and Alkemade and Homann did not observe
fulvene, which is often postulated as a stable intermediate
between linear C6H6’s and benzene, Scheme 1. Fahr and Nayak8
found 15HD to account for 60% of their initial products at room
temperature and 50 Torr with 25% 12HD5Y and 15% of an
unidentified linear C6H6 species, which by comparison with
Alkemade and Homann may be 1245HT. It is clear that Fahr
and Nayak observed a primary product distribution different
from that of Alkemade and Homann. This may be explained
by Alkemade and Homann’s suggestion that under their
experimental conditions the 15HD is chemically activated when
formed and isomerizes to 1245HT and 12HD5Y. Shock tube
ARAS studies by Scherer et al.10 on propargyl recombination
were supplemented by GC-MS analysis of post shock gas
samples that indicated the presence of five C6H6 isomers of
which they were only able to positively identify benzene.
Scherer et al. noted that as the benzene concentration increased
with increasing reaction temperature the other species concen-
trations diminished. Recently, Shafir et al.9 used a flow system
at temperatures from 500 to 1000 K to study propargyl
recombination and observed 15HD as the sole linear recombina-
tion product. They also observed two other C6H6 species, which
were not identified, and benzene, which became the only product
at 900 K. In addition, small quantities of fulvene were seen.
and benzene are formed simultaneously, i.e., fulvene is not an
intermediate to benzene, or more than one route to benzene
exists. The only experimental study of fulvene to benzene
isomerization28 indicates that, under the conditions of the prior
15HD experiments, fulvene to benzene conversion is too slow
to be significant. This idea is also supported by a theoretical
study of fulvene isomerization.29 Conversely, most theoretical
studies of propargyl recombination suggest that benzene is
formed sequentially from fulvene. A very recent theoretical
paper19 may have solved this problem and includes routes to
benzene that treat fulvene as an intermediate to benzene and a
route to benzene that bypasses fulvene entirely.
Hence there would appear to be some disagreement between
experimental and theoretical studies over the paths from
propargyl to benzene and it is the intent of this work to address
the disagreement by using the very clean, well characterized
reaction environment of the high-pressure shock tube coupled
with sensitive analytical techniques capable of unambiguously
identifying C6H6 isomers to investigate 1,5-hexadiyne pyrolysis.
Experimental Section
The high-pressure shock tube is operated as a single pulse
shock tube with tailored driven and driver section lengths to
produce optimal quenching of the reactive gases at the end of
the reaction period. The shock tube and the sampling techniques
used in these studies have been described in detail in recent
publications30-32 and only a brief description will be given here.
Shock waves are generated by spontaneously bursting pre-
scored diaphragms between the driver and driven sections of
the shock tube, and the experiments are performed in the stable,
isothermal zone behind the reflected shock wave where the
temperature and pressure are T5 and P5, respectively. In this
work experiments were performed at nominal P5 values of 25,
50, 300, and 500 bar. The 300 and 500 bar experiments utilized
prescored diaphragms made of soft brass that were 0.032 in./
0.010 in. (thickness/score depth) and 0.050 in./0.016 in.,
respectively. For the lower pressure experiments aluminum
diaphragms were used, 0.025 in./0.010 in. for 25 bar and 0.025
in./0.005 in. for 50 bar. The appropriate diaphragms were
determined on the basis of prior experience and experiment.
Pressures behind the reflected shock wave and reaction times
were obtained from the pressure profiles measured by a
piezoelectric pressure transducer mounted axially in the end wall
of the driven section. The reaction time was calculated on the
basis of the method of Hidaka et al.,33 i.e., the time between
the arrival of the incident shock and P5 falling to 80% of the
reaction pressure. In this work the reaction times were 1.2-1.5
ms for 300 and 500 bar and 1.6-1.9 ms for 25 and 50 bar.
Reaction temperatures ranged from 780 to 1400 K. For the
300 and 500 bar experiments real gas effects are significant31,34
and the temperatures were determined by using chemical
thermometers,31 with the lower temperatures being determined
by extrapolation of the temperature calibration curves. The
reaction temperatures in the 25 and 50 bar experiments were
calculated from the incident shock velocity, and for these
experiments errors in the temperature due to deviations from
ideal gas behavior are expected to be small.34 Shock velocities,
required for determining temperatures both from chemical
thermometers and by calculation, for all experiments were
obtained by the time taken for the incident shock wave to travel
between piezoelectric pressure transducers mounted along the
side wall of the driven section near the end of the shock tube.
For each experiment a sample of the reagent gas was taken
prior to firing the shock tube, preshock sample, and a sample
The potential energy surface for reaction 1 is complex and
has been described in a series of papers,12-19 with two of the
most recent treatments being by Dean and Carstensen18 and
Miller and Klippenstein.19 The main reaction paths from these
publications are summarized in Scheme 1, and the reader is
referred to ref 19 for a detailed energy level diagram. Note that
in the majority of these studies the route to benzene from any
of the three propargyl recombination products involves the
sequential formation of fulvene and then benzene. In a very
recent paper, Miller and Klippenstein19 have suggested an
additional route to benzene that does not involve fulvene but
occurs via cyclization of cis-1,3-hexadiene-5-yne, which is
formed from 1245HT. Alkemade and Homann found 1,3-
hexadiene-5-yne in their experimental work on propargyl
recombination. However, they did not specify if they had
observed the cis or trans form and it is likely that they had not
separated 1,3-hexadiene-5-yne into the cis and trans isomers
chromatographically.
The theoretical studies all indicate that a variety of stable
species, shown in Scheme 1, should be observed during
propargyl recombination, with benzene being the ultimate
product except at high temperatures where benzene also
decomposes. The earlier experimental work indicates that 1,5-
hexadiyne should be the dominant primary product from
propargyl recombination and this species provides a convenient
entry point for experimentally examining the isomerization
chemistry of the propargyl recombination products, particularly
if a technique is used that employs sensitive analytical tech-
niques that can discriminate C6H6 isomers.
The earlier experimental studies20-27 on 1,5-hexadiyne ther-
molysis and theoretical work on propargyl recombination raise
an interesting problem concerning the formation of fulvene and
benzene. The experimental work tends to indicate that fulvene

3408 J. Phys. Chem. A, Vol. 108, No. 16, 2004
Tranter et al.
of the gas near the end wall of the driven section of the shock
tube was collected shortly after the arrival of the expansion wave
generated from bursting the diaphragm, postshock sample. The
timing for collecting the postshock sample was determined
empirically.30 Both the preshock and postshock samples are
collected in electropolished stainless steel vessels stored in an
oven at 50 °C prior to and after sampling. Samples could be
stored for a couple of days with no discernible change in
composition. Over the course of a couple of weeks a slow
isomerization of fulvene to 3,4,-dimethylenecyclobutene was
observed in a sample stored at 50 °C.
Reagents. The reagent mixtures were prepared manometri-
cally in 50 L high-pressure vessels that were maintained at 45
°C and allowed to stand overnight before use. The mixtures
contained 40-55 ppm 1,5-hexadiyne (>99.9%, see below), 200
ppm xenon (AGA, 99.99%) with the balance argon (BOC,
99.999%). The argon was passed over Oxisorb (Messer-
Griesheim) to remove traces of O2 prior to admission to the
mixing vessel, and xenon was used as supplied. Xenon is used
as an internal standard for the analytical work and to account
for any dilution of the postshock sample by the driver gas, He.
Figure 1. Total ion chromatogram from GC-mi-FTIR-MS analysis
of a post shock sample from 1,5-hexadiyne pyrolysis: (A) 2-ethynyl-
1,3-butadiene; (B) 1,5-hexadiyne; (C) 3,4-dimethylenecyclobutene; (D)
fulvene; (E) cis-1,3-hexadiene-5-yne; (F) trans-1,3-hexadiene-5-yne;
(G) benzene; (H) 1,2-hexadene-5-yne.
1,5-Hexadiyne was obtained from GFS Chemicals in 50%
solution with pentane. Prior to use the 1,5-hexadiyne was
purified by a two-step distillation process. In the first step a
crude separation was effected by an atmospheric pressure
distillation and the fraction that distilled over up to 70 °C was
collected. The distillate was almost pure pentane (bp 36 °C,
mp -130 °C), and the residue in the distillation flask was about
98-99% 1,5-hexadiyne (bp 86 °C, mp -6 °C) with the
remainder pentane, confirmed by GC-MS. Next a low-
temperature (approximately -30 °C) vacuum distillation was
performed with the 98-99% 1,5-hexadiyne to remove the
remaining pentane. After this step a purity of better than 99.9%
was obtained for the 1,5-hexadiyne (confirmed by GC-FID and
GC-MS). The purified 1,5-hexadiyne could be stored refriger-
ated for extended periods without noticeable degradation.
Analytical Techniques. A variety of analytical techniques
have been used in the course of this work. The bulk of the
analyses were performed using our standard two-column, GC-
TCD and GC-FID method with GC-MS being used to confirm
molecular formulas and where possible species identities. In
this technique gas samples are introduced via a sample/pressure
reduction rig to two Valco gas sampling valves (GSVs) mounted
on the same GC (HP 6890). The sample loops on the valves
are filled simultaneously, and injections are made onto two
columns maintained in the same oven. An HP-Molseive 5A
column (30 m, 0.32 mm, 12 µm) is connected to a TCD and
used to detect xenon, which acts as an internal standard. The
second column, HP-1MS (30 m, 0.032 mm, 3 µm), is connected
to an FID and was used to detect stable hydrocarbon species.
In this work all species were baseline resolved and in samples
from experiments where the reaction temperature was above
920 K, up to eight species were observed. Of these, 1,5-
hexadiyne and benzene were positively identified by the
injection of authentic standards. GC-MS analyses, using the
HP1-MS column, indicated that all the observed species had
the formula C6H6 and due to a lack of authentic standards for
the analytes it was not possible to make unambiguous identifica-
tions based solely on the mass spectra. Consequently, a novel
technique, GC-matrix isolation-FTIR-MS (GC-mi-FTIR-
MS),35 which was available at Argonne National Laboratories
was used to obtain simultaneous mass and FTIR spectra for
species eluting from the GC column. The FTIR spectra allow
unambiguous assignment of species identities. In the GC-mi-
Figure 2. Major species observed from 1,5-hexadiyne pyrolysis at 25
bar ([1,5-hexadiyne]0 ) 42 ppm): (0) 1,5-hexadiyne; (O) 3,4-
dimethylenecyclobutene; (4), fulvene; (3) benzene.
FTIR-MS work a 60 m ZB-5 capillary column was used, which
under the analytical conditions performs very similarly to the
HP-1MS column, i.e., baseline separation achieved, same
number of peaks observed and elution order is the same. The
GC-mi-FTIR-MS portion of this work, including details of
the technique, has been reported in a separate publication where
the FTIR and MS spectra are available along with the results
of density functional theory calculations that were used as aids
in assigning structural identities.35 A significant aspect of the
FTIR work was that both the cis and trans isomers of
1,3-hexadiene-5-yne, 1,2-hexadiene-5-yne, and 2-ethynyl-1,3,-
butadiene were positively identified. Previously, Alkemade and
Homann7 had observed what was probably a mixture of cis-
and trans-1,3-hexadiene-5-yne but had not separated it into the
distinct isomers. A sample total ion chromatogram is shown in
Figure 1.
Results
The species concentration profiles with respect to temperature
are shown for the 25, 50, 300, and 500 bar experiments in
Figures 2-9, note concentrations are shown as mole fractions
relative to initial 15HD concentration. Figure 10 shows that over
the variation in pressure of a factor of 20 in these experiments

Thermal Rearrangement of 1,5-Hexadiyne
J. Phys. Chem. A, Vol. 108, No. 16, 2004 3409
Figure 3. Minor species observed from 1,5-hexadiyne pyrolysis at 25
bar ([1,5-hexadiyne]0 ) 42 ppm): (0) cis-1,3-hexadiene-5-yne; (O)
trans-1,3-hexadiene-5-yne; (4) 2-ethynyl-1,3-butadiene.
Figure 6. Major species observed from 1,5-hexadiyne pyrolysis at 300
bar. Closed symbols: [1,5-hexadiyne]0 ) 40 ppm. Open symbols: [1,5-
hexadiyne]0 ) 46 ppm. Key: (9, 0) 1,5-hexadiyne; (b, O) 3,4-
dimethylenecyclobutene; (2, 4) fulvene; (1, 3) benzene.
Figure 4. Major species observed from 1,5-hexadiyne pyrolysis at 50
bar ([1,5-hexadiyne]0 ) 42 ppm): (0) 1,5-hexadiyne; (O) 3,4-
dimethylenecyclobutene; (4) fulvene; (3) benzene.
Figure 7. Minor species observed from 1,5-hexadiyne pyrolysis at
300 bar ([1,5-hexadiyne]; ) 46 ppm): (0) cis-1,3-hexadiene-5-yne;
(O) trans-1,3-hexadiene-5-yne; (4) 2-ethynyl-1,3-butadiene.
Figure 5. Minor species observed from 1,5-hexadiyne pyrolysis at 50
bar ([1,5-hexadiyne]0 ) 42 ppm): (0) cis-1,3-hexadiene-5-yne; (O)
trans-1,3-hexadiene-5-yne; (4) 2-ethynyl-1,3-butadiene.
Figure 8. Major species observed from 1,5-hexadiyne pyrolysis at 500
bar. Closed symbols: [1,5-hexadiyne]0 ) 55 ppm. Open symbols: [1,5-
hexadiyne]0 ) 48 ppm. Key: (9, 0) 1,5-hexadiyne; (b, O) 3,4-
dimethylenecyclobutene; (2, 4) fulvene; (1, 3) benzene.
there are no significant differences for each experimental range
in the 15HD profiles and similar results are obtained for all the
other species observed.
At temperatures below about 900 K, 1,5-hexadiyne is
quantitatively converted to 3,4,-dimethylenecyclobutene
(34DMCB) and no other species are observed. Around 950 K
the concentration of 34DMCB reaches a maximum and small
amounts of fulvene are formed. When the reaction temperature
reaches 1000 K, approximately 50% of the 34DMCB has been
consumed, significant concentrations of fulvene are present and
traces of benzene, cis-1,3-hexadiene-5-yne (cis-13HD5Y) and
trans-1,3-hexadiene-5-yne (trans-13HD5Y) are formed. About

3410 J. Phys. Chem. A, Vol. 108, No. 16, 2004
Tranter et al.
Figure 11. Carbon balance at 50 bar ([1,5-hexadiyne]0 ) 42 ppm):
(0) postshock; (O) preshock.
Figure 9. Minor species observed from 1,5-hexadiyne pyrolysis at
500 bar ([1,5-hexadiyne]0 ) 48 ppm): (0) cis-1,3-hexadiene-5-yne;
(O) trans-1,3-hexadiene-5-yne; (4) 2-ethynyl-1,3-butadiene.
Figure 12. Carbon balance at 300 bar. Closed symbols: [1,5-
hexadiyne]0 ) 40 ppm. Open symbols: [1,5-hexadiyne]0 ) 46 ppm.
(9, 0) postshock; (b, O) preshock.
Figure 10. Effect of reaction pressure on 1,5-hexadiyne consump-
tion: (0) 25 bar; (O) 50 bar; (4) 300 bar; (3) 500 bar.
50 K after cis- and trans-13HD5Y appear, 2-ethynyl-1,3-
butadiene (2E13BD) is formed. The two rotamers of 13HD5Y
are formed in almost equal proportions with about 10-20%
more of the cis form than the trans form and the maximum
mole fraction of the cis form is around 0.07 at 1100 K. Both
the cis and trans rotamers are completely consumed by 1200
K. The mole fraction of 2E13BD builds slowly from 1000 K
to a maximum of 0.02 at 1200 K and is completely consumed
by 1350 K. The benzene concentration increases steadily from
its appearance at 1000 K to a maximum at 1400 K where it is
the only stable species present.
ments. Consequently, there should be no low activation barrier
process such as radical recombination reactions that can occur
during the quenching of the reactive gases by the rarefaction
waves in the shock tube and the reaction times can be obtained
with confidence by the normal method mentioned above.
The experimental datasets are available in the Supporting
Information.
Discussion
A very small peak in the chromatographic analysis that
corresponded to 1,2-hexadiene-5-yne was observed and in the
samples analyzed by GC-mi-FTIR-MS gave good quality
FTIR spectra; see ref 35 for details. However, due to the low
concentration, ng/L level, it was not possible to accurately
quantify this peak. 1,2,4,5-Hexatetraene was not observed at
all in the kinetic or FTIR analyses.
Carbon balances have been calculated for every experiment
and are presented for the 50 and 300 bar experiments in Figures
11 and 12. It is quite clear that the carbon balances at 50 and
300 bar are good although there may be a small deficit at the
very highest temperatures of this work and similar carbon
balances are observed for the 25 and 500 bar experiments. The
good carbon balances indicate that all species are being
recovered. The lack of species with formulas other than C6H6
clearly indicates that no fragmentation and hence no chain
processes are occurring under the conditions of these experi-
Thermal Rearrangement Rate Coefficients. 1,5-Hexadiyne
to 3,4-Dimethylenecyclobutene. Because 3,4-dimethylenecyclo-
butene is the only significant product from the thermal rear-
rangement of 1,5-hexadiyne at relatively low temperatures from
780 to 950 K, it is possible to determine the rate coefficients of
reaction 5 at different reaction pressures if it is assumed that
the hexatetraene intermediate rapidly cyclizes to the cyclobutene,
represented by ka in Scheme 2. The mechanism of this process
is discussed below.
From the experimental data, ka’s at different reaction pressures
were calculated by least-squares fit, as shown in Figure 13. The
individual rate expressions for each pressure range and the
literature expressions for ka are presented in Table 1. Also shown
in Figure 13 is the computational result of Melius et al. for
ka,14 which predicts rate coefficients that are about 3 orders of
magnitude larger than the experimental values. Miller and

Thermal Rearrangement of 1,5-Hexadiyne
J. Phys. Chem. A, Vol. 108, No. 16, 2004 3411
SCHEME 2: 1,5-Hexadiyne Conversion to
3,4-Dimethylenecyclobutene and Fulvenea
a Dashed lines represent overall processes for which rate coefficients
have been extracted. See text for details.
Figure 14. Arrhenius plots for reaction 3,4-dimethylenecyclobutene
f fulvene. In the figure symbols represent experimental data, and lines
represent linear fits to the data. Key: solid line and 9, 25 bar; dashed
line and b, 50 bar; dotted line and 2, 300 bar; dashed-dotted line and
1, 500 bar. Data from Miller and Klippenstein,19 Melius et al.,14 and
Stein et al.25 See Table 2 for detailed rate parameter values.
TABLE 2: Rate Coefficients for Conversion of
3,4-Dimethylenecyclobutene to Fulvenea
pressure/bar
temperature/K
log(A)
E/kcal
source
25
50
300
500
1.013
b
780-950
780-950
780-950
780-950
523-723
13.03
12.38
14.43
12.92
12.9
13.63
13.07
50.22
47.12
54.83
48.74
50.00
49.26
50.08
p.w.
p.w.
p.w.
p.w.
25
14
19
c
Figure 13. Arrhenius plots for reaction 1,5-hexadiyne f 3,4-
dimethylenecyclobutene. In the inset symbols represent experimental
data and lines represent linear fits to the data. Key: solid line and 9,
25 bar; dashed line and b, 50 bar; dotted line and 2, 300 bar; dashed-
dotted line and 1, 500 bar. Data from Miller and Klippenstein,19 Melius
et al.,14 Stein et al.,25 and Huntsman and Wristers.20 See Table 1 for
detailed rate parameter values.
a See text for discussion of reaction path. b Theoretical computation.
c Theoretical computation. A and E were obtained by extrapolation from
Figure 8 in ref 19.
Dimethylenecyclobutene to FulVene. At temperatures above
about 950 K in the current work all the 1,5-hexadiyne has been
consumed, 3,4-dimethylenecyclobutene reaches a maximum, and
fulvene begins to appear. By assuming at temperatures above
950 K that the reaction system can be modeled by the formation
of 3,4-dimethylenecyclobutene being reversible to 1,2,4,5
hexatetraene and the 1245HT isomerizing to fulvene, Scheme
2, one can obtain kb in Scheme 2. Obviously such a process is
a simplification of the sequence of elementary reactions
involved12,17 in converting 3,4-dimethylenecyclobutene to ful-
vene; it assumes that all intermediate species are highly reactive
and that additional channels are negligible. Using such a scheme,
rate coefficients, kb, have been obtained and are shown in Figure
14 where they are compared with values from Stein et al.25 and
theoretical calculations by Miller and Klippenstein,19 and Melius
et al.14 The current work has yielded rate coefficients, Table 2,
that are in good agreement with the low-temperature values from
Stein et al. and the extrapolated values from Miller and
Klippenstein that have been estimated from Figure 8 of ref 19.
Once again Melius et al. overpredict the measured rate co-
efficients over the whole temperature range, this time by a factor
of 3-4. It should be noted that in the present work the
temperature range where 3,4-dimethylenecyclobutene and ful-
vene are the only products is limited; however, the other species,
cis- and trans-1,3 hexadien-5-yne, that could be formed from
thermolysis of 3,4-dimethylenecyclobutene are only ever present
in very low concentrations and the formation of these species
instead of fulvene should lead to only minor errors in the
deduced rate coefficients.
TABLE 1: Rate Coefficients for Conversion of
1,5-Hexadiyne to 3,4-Dimethylenecyclobutenea
pressure/bar
temperature/K
log(A)
E/kcal
source
25
50
300
500
1.013
1.013
b
780-950
780-950
780-950
780-950
523-723
473-563
10.82
10.98
10.22
10.12
11.7
11.4
12.28
12.28
33.36
33.57
30.11
29.50
35.50
34.40
28.22
36.61
p.w.
p.w.
p.w.
p.w.
25
21
14
c
19
a See text for discussion of reaction path. b Theoretical computation.
c Theoretical computation. A and E were obtained by extrapolation from
Figure 6 in ref 19.
Klippenstein19 modified the barrier heights for 1,5-hexadiyne
going to 1,2,4,5-hexatetraene and 1,2,4,5-hexatetraene to 3,4-
dimethylenecyclobutene by 1.0 kcal/mol and 0.71 kcal/mol,
respectively, and obtain good fits to the Stein et al.25 and
Huntsman and Wristers20 data. An Arrhenius expression for
reaction 5 has been estimated from Figure 6 in ref 19, and the
rate coefficients have been extrapolated to the high temperatures
of the current work. The extrapolated expression overpredicts
the current data by about a factor of 3 to 4 but maintains good
agreement with the Stein et al. data at low temperatures. Figure
13 also shows a small, essentially negligible, increase in the
rate coefficients from the current work as the pressure increases
from 25 to 500 bar. For a 20-fold increase in pressure, k
increases by about a factor of 1.5.
Mechanism of Formation of 3,4-Dimethylenecyclobutene.
The thermal rearrangement of 1,5-hexadiyne has been studied

3412 J. Phys. Chem. A, Vol. 108, No. 16, 2004
Tranter et al.
by Huntsman and Wristers20 in a static system (210-232 °C)
and a flow system (250-550 °C). In both sets of experiments
they observed 3,4-dimethylenecyclobutene as the sole product.
Huntsman and Wristers have suggested that 3,4-dimethylene-
cyclobutene is formed from 1,5-hexadiyne by a two-step process
(see Scheme 2) involving a Cope rearrangement to 1,2,4,5-
hexatetraene followed by rapid cyclization of 1245HT to 3,4-
dimethylenecyclobutene. A direct route from 15HD to 34DMCB
has also been suggested by Huntsman and Wristers.20 However,
on the basis of the large negative entropy of activation observed
(-9.4 e.u. at 300 °C) and consideration of the Woodward-
Hoffmann rules, Jones et al.,23 Huntsman and Wristers, and
others20-22 concur that 3,4-dimethylenecyclobutene is formed
via a Cope rearrangement of 1,5-hexadiyne followed by a rapid
cyclization of 1,2,4,5-hexatetraene.
Henry and Bergman24 also report that at 250 °C they observed
only 3,4-dimethylenecyclobutene in the thermolysis of 1,5-
hexadiyne; however, at 620 °C in a vacuum flow tube they
observed a mixture of 3,4-dimethylenecyclobutene, fulvene, and
benzene. In atmospheric flow reactor studies of 1,5-hexadiyne
pyrolysis Stein et al.25 observed 3,4-dimethylenecyclobutene as
the only product for temperatures less than 460 °C and residence
times of approximately 30 s. However, in very low-pressure
Figure 15. Benzene:fulvene ratios observed in 1,5-hexadiyne pyrolysis
at 50 bar postshock pressure. Similar trends were found for the 25,
300, and 500 bar experiments.
results of other investigators, suggests, as discussed in the
following paragraphs, that in fact both processes are occurring
and the dominant route is dependent on temperature.
25
pyrolysis (VLPP) studies on the same system, Stein et al.
found that benzene and fulvene were formed along with 3,4-
dimethylenecyclobutene at temperatures from 250 °C to around
700 °C. In the current work 34DMCB is the only product formed
from 15HD and at 950 K about 90% of the initial 15HD has
been converted to 34DMCB, with 10% 15HD remaining in the
post shock gases, Figures 2, 4, 6, and 8. It may be expected
that 1,2,4,5-hexatetraene should be observed in the product
analysis; however, in the current and previous studies of 1,5-
In Figures 2, 4, 6, and 8 it is quite clear, from our work, that
fulvene begins to be formed at around 920-950 K which is
about the same temperature that 34DMCB reaches its maximum
concentration. Benzene first appears at 1000-1050 K. The
benzene:fulvene ratios for the 50 bar experiments are shown in
Figure 15 and it is obvious that there is a dramatic steady
increase in this ratio from the onset of benzene formation up to
1400 K where benzene is the sole product; similar trends were
observed for the 25, 300, and 500 bar experiments.
26
hexadiyne pyrolysis 1245HT has not been found. Hopf has
studied the thermal isomerization of 1245HT to 34DMCB and
subsequent analysis by Huntsman showed the isomerization to
be considerably faster than the formation of 1245HT from
15HD.21 Thus the lack of 1,2,4,5-hexatetraene in the present
studies is not a surprise as 1245HT is most likely converted
effectively instantaneously into 34DMCB. The only experi-
mental observation of 1245HT in propargyl recombination
studies was made by Alkemade and Homann7 where it is formed
as a primary recombination product of two propargyl radicals
and accounts for 7-19% of the total products.
The theoretical papers by Miller and Klippenstein and
others12-19 permit 15HD to isomerize to 1245HT and then to
34DMCB, which is consistent with the experimental work
described above. In the latest publication on propargyl recom-
bination by Miller and Klippenstein19 they have simulated the
experimental data from Stein et al.,25 and the results of their
calculations give good agreement with the observed 3,4-
dimethylenecyclobutene concentrations. Thus, the present work
is consistent with previous observations; it can be concluded
that the mechanism for the formation of 3,4-dimethylene-
cyclobutene from 1,5-hexadiyne involves the formation of
1,2,4,5-hexatetraene, which isomerizes quickly on the time scale
of all extant experiments to 3,4-dimethlylenecyclobutene.
This large change in the benzene:fulvene ratio appears to
indicate that benzene is formed from fulvene and seems to be
in direct contrast to the lower temperature studies of Stein et
al. who observed only relatively small changes in the benzene:
fulvene ratio for large extents of consumption of the initial
15HD. However, Stein et al.’s data extend only to temperatures
that are slightly higher than the maximum in their fulvene
concentration whereas in the current work the reaction was
followed to a temperature where benzene was the only product.
Over the extended temperature range of the current work the
fulvene maximum is found to occur between 1090 and 1110 K
and at the fulvene maximum the benzene:fulvene ratio is about
2:1 to 2.6:1, which is very similar to the values observed by
Stein et al. At their highest reaction temperatures, beyond the
temperature of the fulvene maximum, Stein et al. noticed a slight
decrease in the fulvene concentration with a corresponding
increase in the benzene concentration that corresponds to the
trend in the current work and again appears to support the
contention that benzene is formed form fulvene. However, in
Stein et al.’s work, due to the small changes involved, they could
not conclusively state that fulvene was being converted to
benzene.
A direct experimental investigation of the fulvene to benzene
isomerization by Gaynor et al.28 has generated an Arrhenius
rate expression that permits simulations of the present data using
a simple model, Table 3, to determine if fulvene to benzene
isomerization is possible, Figures 16 and 17. The simulations
indicate that fulvene to benzene isomerization does take place
at our conditions but should not be significant until ap-
proximately 1200 K. Because our experiments indicate that
significant amounts of benzene are observed at temperatures
Mechanism of Formation of Fulvene and Benzene. The
formation of fulvene and benzene in propargyl recombination
and 1,5-hexadiyne experiments are complex processes and there
has been some debate as to whether fulvene is formed before
benzene and then isomerizes to benzene or if fulvene and
benzene are formed simultaneously. Prior experimental evidence
favors simultaneous formation and theoretical studies favor the
consecutive route. Our experimental work combined with the

Thermal Rearrangement of 1,5-Hexadiyne
J. Phys. Chem. A, Vol. 108, No. 16, 2004 3413
from a path not involving fulvene. Further support for the
simultaneous, parallel formation of fulvene and benzene in
15HD pyrolysis can be gained from the experimental work by
Henry and Bergmann on 3,4-dimethylenecyclobutene pyrolysis
at 890 K,24 in which they observe a higher benzene:fulvene ratio
than found by Coller et al.22 in the direct pyrolysis of 15HD at
the same temperature. Alkemade and Homann7 interpret the
relatively large difference in benzene:fulvene ratios from
pyrolysis of 34DMCB and 15HD to indicate an additional path
to benzene not involving fulvene that is accessible in the 15HD
experiments. Thus it would appear that fulvene can be converted
to benzene at high temperatures but that there is also the
possibility of the existence of a lower temperature path directly
to benzene.
The possibility of a low-temperature benzene route that does
not involve fulvene was addressed in part by a recent reexami-
nation of the propargyl recombination potential energy surface,
PES, by Miller and Klippenstein,19 who have simulated the
experimental data of Stein et al.25 They have shown that their
model, which includes a nonfulvene route to benzene via cis-
1,3-hexadiene-5-yne and some small modifications to barrier
heights, fits the experimental work of Stein et al. well. Miller
and Klippenstein also show that not including this route as in
the original Miller and Melius model13,14 overpredicts fulvene
and underpredicts benzene and 3,4-dimethylenecyclobutene. The
new model of Miller and Klippenstein based on their calculated
PES predicts that the nonfulvene path to benzene should include
the formation of cis-1,3-hexadiene-5-yne and trans-1,3-hexa-
diene-5-yne; see Scheme 1. In our current work we were able
to quantitatively measure the formation of cis-1,3-hexadiene-
5-yne and trans-1,3-hexadiene-5-yne, confirming the theoretical
predictions of Miller and Klippenstein. It is clear from our
experimental measurements shown in Figures 3, 5, 7, and 9
that these species are observed, albeit at low concentrations
(maximum mole fractions are 0.07 cis and 0.06 trans), in a
relatively narrow temperature range between about 950 and 1180
K and that there is a slight excess of the cis to trans isomer.
The low concentrations of cis and trans 13HD5Y at temperatures
below 1180 K, where benzene formation is significant, indicate
that cis-13HD5Y, as postulated by Miller and Klippenstein, is
efficiently converted to benzene. By 1180 K benzene accounts
for 40% of the species in the current work even though fulvene
to benzene isomerization should be insignificant at these
temperatures. Supplemental confirmation of the postulated path
involving cis-13HDY in Scheme 1 is found in the results of
Hopf and Muso36 who have studied the thermolysis of both cis-
and trans-13HD5Y and found that both isomers and benzene
are generated regardless of which isomer is used as the starting
material.
Figure 16. Simulation of 25 bar experimental data using the three-
step model. Detailed rate parameter values are shown in Table 3.
Symbols represent experimental data, and lines represent simulation
results. Key: solid line and 9, 1,5-hexadiyne; dashed line and b, 3,4-
dimethylenecyclobutene.
Figure 17. Simulation of 25 bar experimental data using the three-
step model. Detailed rate parameter values are shown in Table 3.
Symbols represent experimental data, and lines represent simulation
results. Key: solid line and 9, fulvene; dashed line and b, benzene.
TABLE 3: Simple Model Used to Simulate 25 Bar
Experimental Data. 15HD ) 1,5-hexadiyne, 34DMCB )
3,4-dimethylenecyclobutene
reaction
log(A)
E/kcal
source
15HD )> 34DMCB
34DMCB )> Fulvene
Fulvene )> Benzene
10.82
13.03
13.49
33.36
50.22
68.2
p.w.
p.w.
28
Thus it appears from the current work that benzene and
fulvene are formed simultaneously at temperatures below around
1180 K with benzene being predominantly formed by the
isomerization of cis-1,3-hexadiyne-5-yne, which is in accord
with the recent theoretical prediction by Miller and Klippen-
stein.19 At higher temperatures the isomerization of fulvene to
benzene, indicated by the growth of benzene as the fulvene
concentration decreases, should be important, and if it is
assumed that this becomes so at the fulvene maximum, i.e., 1100
K, then some adjustment to the fulvene to benzene rate
expression of Gaynor et al. may be warranted.
as low as 1050 K either there is another route to benzene or the
rate coefficients generated by Gaynor may be suspect. This latter
point was addressed by Madden et al. 29 who have used ab initio
calculations to investigate fulvene to benzene isomerization.
They concluded that the activation energy obtained by Gaynor
et al. may be too low and if this is the case the onset of fulvene
to benzene conversion should not occur until even higher
temperatures than the ones observed in our experiment. Con-
sequently, both Gaynor et al.’s rate coefficient and an even more
refined one from Madden et al. suggest that the lower temper-
ature formation of benzene observed in our work may result
Finally, fulvene can also be formed from 2-ethynyl-1,3-
butadiene, Scheme 1, which has been observed in these
experiments in small quantities. The exact importance of this

3414 J. Phys. Chem. A, Vol. 108, No. 16, 2004
Tranter et al.
route to the total fulvene concentration will have to be
determined from detailed modeling.
Eight stable C6H6 isomers have been positively identified in
this work, including the important observation for the first time
of cis-1,3-hexadiene-5-yne and trans-1,3-hexadiene-5-yne. These
two species and their appearance temperatures conclusively
demonstrate the existence of a low-temperature route to benzene
that does not require the formation and subsequent isomerization
of fulvene. Thus, for the lower temperatures in this work, fulvene
and benzene are formed simultaneously by separate reaction
paths. Recent theoretical analyses of the likely reaction paths
appear to be completely consistent with this observation.
Furthermore, 1,2-hexadiene-5-yne and its isomerization product
2-ethynyl-1,3-butadiene were observed at reaction temperatures
that are probably too low for dissociation of 1,5-hexadiyne back
to two propargyl radicals. On the current potential energy
surfaces the only route to 12HD5Y from 15HD is via dissocia-
tion and recombination, and the most recent work indicates that
the PES may need to be modified to include a direct route
between 15HD and 12HD5Y. Last, a high-temperature fulvene
to benzene route, predicted by the PES, is confirmed by our
experimental evidence.
Mechanism of Formation of 2-Ethynyl-1,3-butadiene.
2-ethynyl-1,3-butadiene appears between 1150 and 1300 K with
a maximum mole fraction of 0.02 in these experiments. This
species can only be formed from 1,2-hexadiene-5-yne, a species
that is inaccessible on the Miller and Klippenstein PES when
starting with 1,5-hexadiyne without first dissociating 15HD to
propargyl radicals. In the low-concentration kinetic work, 1,2-
hexadiene-5-yne was observed but the concentrations were too
low to measure quantitatively. In the higher concentration
samples prepared for GC-mi-FTIR-MS analysis a small but
well formed peak was observed whose spectrum confirmed that
the species was 1,2-hexadiene-5-yne, and comparison between
the FTIR analyses and the GC-MS, GC-FID analyses indicates
that trace amounts of 1,2-hexadiene-5-yne are present in the
kinetic samples. A second potential route to 2-ethynyl-1,3-
butadiene not involving dissociation of 15HD is from the
isomerization of benzene through fulvene, and this has been
observed by Nakashima et al.37 in very vibrationally hot
benzene. However, in the present thermally excited work it is
very unlikely that the temperatures are high enough to drive
benzene back to 2-ethynyl-1,3-butadiene. The presence of trace
amounts of 1,2-hexadiene-5-yne tend to suggest that some of
the original 1,5-hexadiyne has either decomposed to propargyl
radicals that have then recombined with a fraction forming
12HD5Y or that there is a direct isomerization path from 15HD
to 12HD5Y that is not included on the Miller and Klippenstein
PES.
Acknowledgment. We thank Dr L. B. Harding of Argonne
National Laboratories for performing the DFT calculations that
were used in interpreting the FTIR spectra. R.S.T., W.T., and
K.B. are grateful to the National Science Foundation for support
under contract CTS 0109053. K.B.A. gratefully acknowledges
the support of the Office of Basic Energy Sciences, Division
of Chemical Sciences, Geosciences, and Biosciences, under
contract number W-31-109-ENG-38.
Some preliminary simulations have been performed to try to
assess the relative importance of dissociation of 15HD to two
propargyl radicals, reverse of reaction 2, and isomerization of
15HD to 12HD5Y. A rate coefficient for dissociation of 15HD
to two propargyls, k-2 (reaction 2), was obtained from a QRRK
calculation by Dean.38 Rates for reactions 2 and 3 were taken
from ref 18. There does not appear to be a published rate
coefficient expression for reaction 4, and this was estimated to
be half of that for reaction 2 on the basis of Alkemade and
Homann’s product distributions. For the isomerization of 15HD
to 12HD5Y we have used the rate coefficient expression for a
1,3 H-atom shift in propyne to give allene.39 The results of this
preliminary and necessarily crude modeling indicate that at
temperatures above 1100 K dissociation of 15HD to propargyl
can be significant but that recombination to 15HD, reaction 2,
is the favored fate of the propargyl radicals with small amounts
forming 1245HT and 12HD5Y in approximately 1.5:2 propor-
tions, which is roughly similar to the distribution seen by
Alkemade and Homann. The model also indicates that direct
isomerization of 15HD to 12HD5Y is negligible. However, these
results should be treated with caution. If, as the model predicts,
1245HT and 12HD5Y are formed in 1.5:2 ratio then 1245HT
should have been observed in the analysis of the postshock
samples and there is absolutely no indication that this species
is present even in the concentrated samples prepared for GC-
mi-FTIR-MS analysis. It is therefore suggested that the
isomerization of 15HD to 12HD5Y may be more significant
than indicated by this preliminary modeling work and a thorough
theoretical investigation of the mechanism may be warranted.
Supporting Information Available: Tables of experimental
data containing reaction conditions and species concentrations.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References and Notes
(1) Miller, J. A. Proc. Combust. Inst. 1996, 20, 461.
(2) Miller, J. A. Faraday Discuss. 2001, 119, 461.
(3) Richter, H.; Howard, J. B. Prog. Energy Combust. Sci. 2000, 26,
565.
(4) D’Anna, A.; Violi, A.; D’Allessio, A. Combust. Flame 2000, 121,
418.
(5) Kern, R. D.; Chen, H.; Kiefer, J. H.; Mudipalli, P. S. Combust.
Flame 1995, 100, 177.
(6) Lindstedt, P. Proc. Combust. Inst. 1998, 27, 269.
(7) Alkemade, U.; Homann, K. H. Z. Phys. Chem. 1989, 161, 19.
(8) Fahr, A.; Nayak, A. Int. J. Chem. Kinet. 2000, 32, 118.
(9) Shafir, E. V.; Slagle, I. R.; Knyazev, V. D. J. Phys. Chem. A 2003,
107, 8893.
(10) Scherer, S.; Just, T.; Frank, P. Proc. Combust. Inst. 2000, 28, 1511.
(11) DeSain, J. D.; Taatjes, C. A. J. Phys. Chem. A 2003, 107, 4843.
(12) Thomas S. D.; Communal F.; Westmoreland, P. R. ACS, DiV. Fuel
Chem., 1991, 1449.
(13) Miller, J. A.; Melius, C. F. Combust. Flame 1992, 91, 21.
(14) Melius, C. F.; Miller, J. A.; Evleth, E. M. Proc. Combust. Inst.
1992, 24, 621.
(15) Miller, J. A.; Klippenstein, S. J. J. Phys. Chem. A 2001, 105, 7254.
(16) Klippenstein, S. J.; Miller, J. A. J. Phys. Chem. A 2002, 106, 9267.
(17) Carstensen, H.-H., Dean, A. M. Proceedings of the Third Joint
Meeting of the U.S. Sections of The Combustion Institute, Chicago, March,
16-18, 2003.
(18) Carstensen, H.-H.; Dean, A. M. Private communication.
(19) Miller, J. A.; Klippenstein, S. J. J. Phys. Chem. A 2003, 107, 7783.
(20) Huntsman, W. D.; Wristers, H. J.J. Am. Chem. Soc. 1967, 18, 342.
(21) Huntsman, W. D. Intra-Sci. Chem. Rep. 1972, 6, 151.
(22) Coller, B. A. W.; Heffernan, M. L.; Jones, A. J. Aust. J. Chem.
1968, 21, 1807.
Conclusions
An extensive experimental study of the pyrolysis of 1,5-
hexadiyne has been performed at four reaction pressures of 25,
50, 300, and 500 bar and temperatures from 780 to 1400 K.
(23) Kent, J. E., Jones, A. J. Aust. J. Chem. 1970, 23, 1059.
(24) Henry, T. J.; Bergman, H. G. J. Am. Chem. Soc. 1972, 9, 5103.

Thermal Rearrangement of 1,5-Hexadiyne
J. Phys. Chem. A, Vol. 108, No. 16, 2004 3415
(25) Stein, S. E.; Walker, J. A.; Suryan, M.; Fahr, A. Proc. Combust.
Inst. 1990, 23, 85.
(32) Tranter, R. S.; Sivaramakrishnan, R.; Brezinsky, K.; Allendorf, M.
D. Phys. Chem. Chem. Phys. 2002, 4, 2001.
(26) Hopf, H. Chem. Ber. 1971, 104, 1499.
(33) Hidaka, Y.; Shiba, S.; Takuma, H.; Suga, M. Int. J. Chem. Kinet.
1985, 17, 441.
(34) Davidson, D. F.; Hanson, R. K. Isr. J. Chem. 1996, 36, 321.
(35) Anderson, K. B.; Tranter, R. S.; Tang, W.; Brezinsky, K.; Harding
L. J. Phys. Chem. A 2004, 108, 3403.
(36) Hopf, H.; Musso, H. Angew. Chem. Int. Ed. 1969, 8, 680.
(37) Nakshima, N., Yatsuhasi, T. Bull. Chem. Soc. Jpn. 2001, 74, 579-
593.
(27) Tang, W.; Tranter, R. S.; Brezinsky, K.; Raju, A. S. K. Proceedings
of the Third Joint Meeting of the U.S. Sections of The Combustion Institute,
Chicago, March, 16-18, 2003.
(28) Gaynor, B. J.; Gilbert, R. G.; King, K. D.; Harman, P. J. Aust. J.
Chem. 1981, 34, 449.
(29) Madden, L. K.; Mebel, A. M.; Lin, M. C.; Melius, C. F. J. Phys.
Org. Chem. 1996, 9, 801.
(30) Tranter, R. S., Fulle, D.; Brezinsky, K. ReV. Sci. Instrum. 2001,
72, 3046.
(38) Dean, A. M. J. Phys. Chem. 1985, 89, 4600
(31) Tranter, R. S.; Sivaramakrishnan, R.; Srinivasan, R.; Brezinsky,
K. Int. J. Chem. Kinet. 2001, 33, 722.
(39) Kiefer, J. H.; Mudipalli, P. S.; Sidhu, S. S.; Kern, R. D.; Jursic, B.
S.; Xie K.; Chen, H. J. Phys. Chem. A 1997, 101, 4057.