Shock tube studies on the decomposition of 2-butanol

Article abstract of DOI:10.1021/jp306975s

The thermal decomposition of 2-butanol have been studied at temperatures of 1045-1221 K and pressures of 1.5-6 bar using the single pulse shock tube technique. Dilute concentrations of 2-butanol have been decomposed in the presence of large quantities of a radical inhibitor. The mechanism for decomposition involves direct elimination of water producing cis- and trans-2-butene, and 1-butene, and C-C bond fission producing ethylene. Acetaldehyde, propionaldehyde, and propene were also observed in much smaller yields from C-C bond fission. The respective unimolecular rate expressions are as follows: k(C₃H₆(OH)CH₃ → cis-CH ₃CH=CHCH₃ + H₂O) = 10(13.1 ± 0.3) exp(-33414 ± 755 K/T) s^-1; k(C₃H ₆(OH)CH₃ → trans-CH₃CH=CHCH₃ + H₂O) = 10(13.5 ± 0.3) exp(-33820 ± 755 K/T) s ^-1; k(C₃H₆(OH)CH₃ → CH ₃CH₂CH=CH₂ + H₂O) = 10(13.6 ± 0.3) exp(-33002 ± 755 K/T) s^-1; k(C₃H ₆(OH)CH₃ → C₂H₅ ^? + ^?CH(OH)CH₃) = 10(15.9 ± 0.3) exp(-39252 ± 755 K/T) s^-1. These rate expressions are compared with analogous reactions for primary and tertiary butanols. They form a basis for the prediction of those for related systems. Comparison with estimated values used in the simulation of butanol combustion is indicative of the uncertainties in the rate constants that are used in such models. The activation energy of 326 kJ/mol leads to a bond dissociation energy of the CH(OH)CH₃ radical (H-CH(OH)CH₃) of 400 kJ/mol, in excellent agreement with earlier calculated results from theory and disagreement with the experimental results from iodination studies in the expected range.

Full text of DOI:10.1021/jp306975s

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Article
pubs.acs.org/JPCA
Shock Tube Studies on the Decomposition of 2‑Butanol
Claudette M. Rosado-Reyes* and Wing Tsang
National Institute of Standards and Technology, U.S. Department of Commerce, Gaithersburg, Maryland 20899, United States
ABSTRACT: The thermal decomposition of 2-butanol have been studied at temperatures of
1045−1221 K and pressures of 1.5−6 bar using the single pulse shock tube technique. Dilute
concentrations of 2-butanol have been decomposed in the presence of large quantities of a
radical inhibitor. The mechanism for decomposition involves direct elimination of water
producing cis- and trans-2-butene, and 1-butene, and CC bond fission producing ethylene.
Acetaldehyde, propionaldehyde, and propene were also observed in much smaller yields from
C−C bond fission. The respective unimolecular rate expressions are as follows: k(C3H6(OH)-
CH3 → cis-CH3CHCHCH3 + H2O) = 10(13.1 0.3) exp(−33414 755 K/T) s−1; k(C3H6(OH)CH3 → trans-CH3CH
CHCH3 + H2O) = 10(13.5 0.3) exp(−33820 755 K/T) s−1; k(C3H6(OH)CH3 → CH3CH2CHCH2 + H2O) = 10(13.6
•
0.3) exp(−33002 755 K/T) s−1; k(C3H6(OH)CH3 → C2H5 + •CH(OH)CH3) = 10(15.9 0.3) exp(−39252 755 K/T)
s−1. These rate expressions are compared with analogous reactions for primary and tertiary butanols. They form a basis for the
prediction of those for related systems. Comparison with estimated values used in the simulation of butanol combustion is
indicative of the uncertainties in the rate constants that are used in such models. The activation energy of 326 kJ/mol leads to a
bond dissociation energy of the CH(OH)CH3 radical (H−CH(OH)CH3) of 400 kJ/mol, in excellent agreement with earlier
calculated results from theory and disagreement with the experimental results from iodination studies in the expected range.
Thus unambiguous experimental results are important not only
for the purposes of simulations but also as a basis for calibrating
theoretical calculations.
INTRODUCTION
■
There have been many recent studies on the use of butanols as
fuels in combustion processes.1−6 Many of these studies have
attempted to simulate the behavior of such systems using
fundamental single step chemical reactions. This paper is
concerned with the initial unimolecular decomposition of 2-
butanol. It is an extension of our earlier work on n-butanol.7
Our interest is in the rate expressions of the breakdown
processes when 2-butanol is subject to high temperature
combustion environments.
This work is designed to help provide fundamental or
transferable information on the chemical kinetics of biofuels
combustion. This will bring such databases to the same level as
those for the petroleum based fuels that are widely used in
computer simulations. The emphasis on fundamental informa-
tion arises from the practical need of ultimately dealing with
fuel mixtures. Fundamental information represents data that
lead directly to mixing rules that cannot be determined on any
other basis. Any possible use of biofuels will be in combination
with petroleum based substances. The greatest difference in the
fundamental reaction processes between any two fuels is the
initial radical formation and destruction processes. Once the
primary radicals are formed upon initial thermal decomposition,
their subsequent reactions become increasingly similar as
combustion progresses, ultimately leading to the production
of carbon dioxide, water or soot.
For the present study we use a well established methodology
that has been previously employed to study the fundamental
kinetics of the decomposition of many intermediate size organic
fuels8 and more recently the fuel radicals that are the first
products9 from fuel decomposition. These results provide the
basic information from which the rate expressions of such
processes for all petroleum based compounds can be estimated.
These numbers are generally accepted and used in the most
recent combustion kinetics databases such as JetSurf.10 The
present work concentrates on the effect of introducing an OH
group into the basic hydrocarbon framework.
The experiments were carried out in a heated single pulse
shock tube. The key features of these experiments are the dilute
concentrations of the test mixture, of the order of 500 μL/L,
and the presence of large excesses of a chemical inhibitor, 1%
1,3,5-trimethylbenzene. The chemical inhibitor captures all the
reactive radicals that are formed. Thus only the unimolecular
reaction contributes to the product yields. The isolation of the
unimolecular reaction from all other processes means that it is
possible to simultaneously study other unimolecular reactions
whose rate expression for decomposition are well established in
the same system. Thus for the extent of decomposition an
average temperature of the system can be determined using an
internal standard approach makes certain that the nonidealities
in the shock tube study are experienced by both target
molecules. Therefore, results from the shock tube study are
The determination of the rate expressions for the chemical
processes involved in fuel breakdown during combustion has
increasingly dependence on the use of ab initio calculations for
estimations. The uncertainties in these estimations are
somewhat unclear. This is especially the case as the fuel
molecules become larger. An added complication is the many
different theory approaches utilized in making such estimates.
Received: July 13, 2012
Revised: August 31, 2012
Published: September 4, 2012
This article not subject to U.S. Copyright.
Published 2012 by the American Chemical
Society
9599
dx.doi.org/10.1021/jp306975s | J. Phys. Chem. A 2012, 116, 9599−9606

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Physical Chemistry A
Article
Figure 1. Schematic illustration of the shock tube apparatus and components.
traceable and transferable and provide the basic information
necessary for developing the relationships between the
behaviors of all fuels that have been studied.
butanols. These assumptions will be confirmed with the
product distribution obtained from 2-butanol pyrolysis.
We have previously studied the decomposition of n-butanol.7
Ethylene and 1-butene are the main detectable products. The
rate expressions for all the possible decomposition pathways are
EXPERIMENTAL METHODS13
■
The thermal decomposition of 2-butanol has been induced in a
heated single pulse shock tube. A schematic of the instrument
can be found in Figure 1. A brief description of the shock tube
scheme is provided. The shock tube consists of two sections, a
171 cm driven section and 125 cm driver section, built from 3.2
cm2 i.d. aluminum tubing, separated by a cellophane diaphragm.
Upon the rupture of the diaphragm, a shock wave is created,
which subjects the test gas to adiabatic compression and
heating during two temperature steps, first by the incident wave
and second by the reflected wave from the end wall. The shock-
heated gas remains at temperature for about 500 50 μs prior
to rapid quenching by the dump tank. Immediately after, an
undiluted gas sample is extracted into two valves and sampling
loops, where it is compressed to typically 1 bar, and the injected
into two individual GC columns for simultaneous separation of
the mixture components and analysis. Lighter gases (C1−C4)
are more efficiently separated in a Restek 30 m × 0.53 mm i.d.
Rt-Alumina capillary column, coupled with a back-flushable
precolumn to trap heavy species and prevent contamination.
Heavier gas components are well separated in a J&W Scientific
30 m × 0.53 mm i.d. DB-1 fused silica column. Simultaneous
flame ionization and mass spectral detection and analysis are
performed after the separation of the mixture components. FID
peaks areas are converted into molar quantities using molar
responses determined from standard samples.
k(C4H9OH → CH3 + CH2CH2CH2OH)
= 2.5 × 1016 exp(−42410 K/T) s−1
k(C4H9OH → CH3CH2 + CH2CH2OH)
= 2.5 × 1016 exp(−41150 K/T) s−1
k(C4H9OH → CH3CH2CH2 + CH2OH)
= 2.5 × 1016 exp(−41150 K/T) s−1
k(C4H9OH → CH3CH2CHCH2 + H2O)
= 1.0 × 1014 exp(−35089 K/T) s−1
Of particular interest is the rate expression for the direct
elimination of water. This reaction must involve a four-center
transition state. The rate expression for direct water elimination
from 1-butanol is in disagreement with results derived from ab
initio calculations11 and on a per H-atom basis is surprisingly
close to that for tert-butanol.1 This is contrary to what one
would expect on the basis of other four-center elimination
processes involving alkyl halides where elimination of HCl is
drastically affected by alkyl substitution.12 The concerted water
elimination process involving the simultaneous breaking and
formation of number of bonds would seem to provide a crucial
test of the accuracy of appropriate ab initio calculations of
transition state properties. The bond breaking reactions involve
assumptions regarding the distribution of ethylene from the
three possible channels and are based on the effect of OH
substitution from the decomposition of more highly branched
Experimental studies were carried out with a mixture
concentration of 500 μL/L 2-butanol in 1.0% 1,3,5-
trimethylbenzene. The internal temperature standard used in
the present study is 4-methylcyclohexene,14
9600
dx.doi.org/10.1021/jp306975s | J. Phys. Chem. A 2012, 116, 9599−9606

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Physical Chemistry A
Article
Figure 2. Proposed reaction mechanism for the thermal decomposition of 2-butanol.
From this rate expression one obtains the following reaction
temperature as function of the product yield,
information regarding the nature of the transition state for this
type of elimination process. Also found were methane and
ethane, from methyl and H radical recombination and methyl
self-reaction, respectively. H and CH3 fragments are scavenged
by the excess of 1,3,5-trimethylbenzene. A detailed breakdown
of the decomposition of 2-butanol, including water elimination
and C−C bond breaking channels, is given in Figure 2. The
processes that are considered for analysis are those that lead to
the formation of the butenes, ethylene, and propene.
Excellent GC separation was obtained between cis-2-butene,
trans-2-butene, and 1-butene in both GC columns, making
them unique markers for concerted 1,2-elimination of water.
They provide a basis for comparison with the earlier work on 1-
butanol and branched butanols. The overall process is a typical
multichannel unimolecular decomposition reaction. Rate
parameters were derived easily because the formation of all
the hydrocarbon products were clearly monitored by the GC.
The following rate expressions were derived for 1045−1221 K
and 1.5−6 bar, where the degree of conversion of the
temperature standard was up to 61%, and that for the 2-
butanol was 30%. This leads to the following rate expressions,
Texp [K] = (Ea/R)/(ln A − ln kexp)
where the experimental rate for the internal standard
decomposition is defined as
k
exp [s−1] = τ−1 ln([std]0 /[std]τ )
and τ is the residence time of about 500 μs. Note that the
internal standard reaction is purely molecular in nature. Thus
radicals cannot be released into the system. The concentration
of standard used in these studies was 200 μL/L.
The thermal decomposition of 2-butanol was examined in
the absence and presence of the temperature standard. This
permits an estimate to be made of the concentration of propene
formed from 2-butanol decomposition because propene is also
formed from the decomposition of the temperature standard.
Because an equal amount of 1,4-butadiene and propene was
found from the decomposition of the standard we found that a
small amount of propene was indeed formed from the
decomposition of the 2-butanol. The aftershock temperature
was then determined from the formation of 1,3-butadiene and
confirmed by the disappearance of 4-methylcyclohexene.
Shock pressures are calculated from the experimental
temperature and mixture composition via ideal shock equations.
Chemicals were purchased from Sigma-Aldrich Chemicals. Gas
chromatography did not uncover any impurities and the
samples were therefore used without purification.
k(C3H6(OH)CH3 → cis‐CH3CHCHCH3 + H2O)
= 1.41 × 1013 exp(−33414 K/T) s−1
k(C3H6(OH)CH3 → trans‐CH3CHCHCH3 + H2O)
= 2.90 × 1013 exp(−33820 K/T) s−1
Aftershock experimental conditions of 1045−1221 K in
k(C3H6(OH)CH3 → CH3CH2CHCH2 + H2O)
= 3.57 × 1013 exp(−33002 K/T) s−1
temperature and 1.5−6 bar in pressures were achieved.
RESULTS
■
•
•
k(C3H6(OH)CH3 → C2H5 + CH(OH)CH3)
= 7.40 × 1015 exp(−39252 K/T) s−1
The major stable products detected from 2-butanol decom-
position are ethylene, trans-2-butene, cis-2-butene, 1-butene,
propene, and acetaldehyde. There is evidence that propional-
dehyde is also present. Ethylene and propene along with the
aldehydes are the expected products following the initial
cleavage of the C−C bonds. The butenes can only arise from
1,2-elimination of water from the 2-butanol. Note that with the
OH in the secondary position, 1-butene can only be formed
from the abstraction of a primary H-atom whereas the 2-
butenes are formed from secondary hydrogen. The relative
yields and rate expressions should therefore lead to interesting
The respective Arrhenius plots, for the water elimination
channels and C2−C3 during 2-butanol decomposition are
shown in Figures 3 and 4.
The kinetic information for the C2−C3 bond scission
reaction of 2-butanol is based on the yield of ethylene. Note
that ethylene is the only compound that can be formed from
this process. Therefore, there can be little ambiguity regarding
the derived rate expression. This is unlike the situation for 1-
9601
dx.doi.org/10.1021/jp306975s | J. Phys. Chem. A 2012, 116, 9599−9606

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Physical Chemistry A
Article
We have attempted to relate propene yields as a consequence
of propene formation from C3−C4 bond cleavage. As we can be
seen from Figure 2, this can be only formed upon methyl
ejection followed by the very fast unimolecular decomposition
of the 2-hydroxypropyl radical through the β bond cleavage of
the C−OH bond. This is thermodynamically favored over the
breaking of the C−CH3 bond. A key problem in the
interpretation of the results is the small yields of propene or
roughly one tenth of the ethylene formed. The relationship
between ethylene and propene yields can be found in Figure 5.
Following the procedure outlined above we obtain the
following rate expression,
•
k(C3H6(OH)CH3 → CH3• + CH2CH(OH)CH3)
= 1.05 × 1013 exp(−34282 K/T) s−1
Figure 3. Arrhenius plots for the decomposition of 2-butanol into the
different water elimination channels.
Figure 5. Molar yields of ethylene plotted against those of propene, in
the absence of temperature standard.
Figure 4. Arrhenius plots for the decomposition of 2-butanol via C−C
bond fission.
The A-factor and activation energy are far smaller than what
would be expected for the CH3−C bond dissociation energy.
One suspects that there may be other sources of propene from
impurity in the system. A plot of the rate constant can be found
in Figure 4 and shows the relative unimportance of the C1−C2
reaction in comparison to the C2−C3 channel. At the present
time, propene is not considered in the product analysis, along
with ethylene, 1-butene, and 2-butene.
butanol where ethylene can be derived from all the C−C bond
cleavage reactions.
It is assumed that the ethyl radical is quantitatively converted
to ethylene and the hydroxyl ethyl radical is not transformed to
ethylene. These are very valid assumptions. The lifetime of
ethyl radicals under the present conditions is less than a few
microseconds. We assume that 1,2 isomerization of the
hydroxyl ethyl radical is much slower than H-atom ejection.
Note that H-atom ejection will in all cases lead to the formation
of acetaldehyde.
Our analysis of acetaldehyde was not satisfactory. To
establish the relative sensitivity of acetaldehyde to ethylene,
experiments were carried out on the thermal decomposition of
ethyl vinyl ether, which is known to decompose via a retroene
reaction to form equal amounts of ethylene and acetylene.
From these experiments we observed lesser amounts of
acetaldehyde than expected and the amount of acetaldehyde
seems to decrease with the total yields of ether decomposing.
We plan to return to this issue in the subsequent reports. Note
that the absence of pressure dependence in the rate constants in
all the experiments with 2-butanol mean that all the rate
expressions directly determined are true unimolecular rate
expressions.
A final process involves the cleavage of the C1−C2 bond. The
process is
•
k(C3H6(OH)CH3 → CH3 + CH(OH)CH2CH3
Our analytical procedure does not permit us to assess the
yields arising from methyl radicals. The 1-hydroxypropyl radical
can decompose via the following processes,
•CH(OH)CH2CH3 → H + CH3CH2CHO
•CH(OH)CH2CH3 → CH3 + CH3CHO
As noted above, we have been able to detect small amount of
propionaldehyde. Acetaldehyde was also formed in larger
amounts from C2−C3 bond cleavage. Thus there is not
sufficient data to give a quantitative measure of the contribution
from this channel. From earlier correlations it is clear that this
will be substantially smaller than that from the main bond
breaking channel. We have accordingly chosen to ignore
9602
dx.doi.org/10.1021/jp306975s | J. Phys. Chem. A 2012, 116, 9599−9606