Thermal decomposition products of butyraldehyde
Courtney D. Hatten, Kevin R. Kaskey, Brian J. Warner, Emily M. Wright, and Laura R. McCunn
Citation: The Journal of Chemical Physics 139, 214303 (2013); doi: 10.1063/1.4832898
View online: http://dx.doi.org/10.1063/1.4832898
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/139/21?ver=pdfcov
Published by the AIP Publishing
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
THE JOURNAL OF CHEMICAL PHYSICS 139, 214303 (2013)
Thermal decomposition products of butyraldehyde
Courtney D. Hatten, Kevin R. Kaskey, Brian J. Warner, Emily M. Wright,
and Laura R. McCunna)
Department of Chemistry, Marshall University, Huntington, West Virginia 25755, USA
(Received 12 August 2013; accepted 8 November 2013; published online 2 December 2013)
The thermal decomposition of gas-phase butyraldehyde, CH3 CH2 CH2 CHO, was studied in the 1300–
1600 K range with a hyperthermal nozzle. Products were identified via matrix-isolation Fourier
transform infrared spectroscopy and photoionization mass spectrometry in separate experiments.
There are at least six major initial reactions contributing to the decomposition of butyraldehyde:
a radical decomposition channel leading to propyl radical + CO + H; molecular elimination
to form H2 + ethylketene; a keto-enol tautomerism followed by elimination of H2 O producing
1-butyne; an intramolecular hydrogen shift and elimination producing vinyl alcohol and ethylene, a β–C–C bond scission yielding ethyl and vinoxy radicals; and a γ –C–C bond scission
yielding methyl and CH2 CH2 CHO radicals. The first three reactions are analogous to those observed in the thermal decomposition of acetaldehyde, but the latter three reactions are made
possible by the longer alkyl chain structure of butyraldehyde. The products identified following thermal decomposition of butyraldehyde are CO, HCO, CH3 CH2 CH2 , CH3 CH2 CH=C=O,
H2 O, CH3 CH2 C≡CH, CH2 CH2 , CH2 =CHOH, CH2 CHO, CH3 , HC≡CH, CH2 CCH, CH3 C≡CH,
CH3 CH=CH2 , H2 C=C=O, CH3 CH2 CH3 , CH2 =CHCHO, C4 H2 , C4 H4 , and C4 H8 . The first ten
products listed are direct products of the six reactions listed above. The remaining products
can be attributed to further decomposition reactions or bimolecular reactions in the nozzle.
© 2013 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4832898]
I. INTRODUCTION
Aldehydes are commonly formed when hydrocarbons are
present in environments of high temperatures and low oxygen
content. Examples include the production and combustion of
biofuels as well as cigarette smoking and coffee roasting.1–5
Low molecular weight aldehydes such as butyraldehyde
(CH3 CH2 CH2 CHO) are known byproducts of the pyrolysis
of biomass which occurs in the initial steps in the production of some biofuels.1 Trace amounts of butyraldehyde have
been identified in bio-oil using gas chromatography-mass
spectrometry.2 Butyraldehyde has also been detected via gas
chromatography and mass spectrometry following the pyrolysis of cigarette ingredients such as n-butyl alcohol and white
sugar.3, 4 The pyrolytic environment of a lit cigarette can cause
further breakdown of the aldehyde byproducts found in tobacco. Coffee brewing has been proven to emit aldehydes as
odorous fumes.5 The trace amounts of butyraldehyde emitted
during this process can thermally decompose during the completion of brewing. During the pyrolysis of biomass, cigarette
ingredients, and coffee, as well as the combustion of bio-oil,
aldehyde byproducts are exposed to high temperatures and
can undergo thermal decomposition reactions. Understanding
the thermal decomposition pathways of butyraldehyde would
facilitate the prediction of pollutants or contaminants in the
processes described above.
Decomposition of various aldehydes has been studied
previously with the use of shock tubes and flow reactors.6–8 To
a) Author to whom correspondence should be addressed. Electronic mail:
[email protected]
0021-9606/2013/139(21)/214303/9/$30.00
our knowledge, no thermal decomposition studies have been
performed on butyraldehyde, but insight can be gained by examining the known thermal decomposition pathways of simpler aldehydes such as acetaldehyde. Vasiliou and co-workers
identified the products of acetaldehyde pyrolysis using
matrix-isolation Fourier transform infrared spectroscopy.9, 10
They also presented complimentary experiments using photoionization mass spectrometry (PIMS) to detect the pyrolysis products. The unimolecular pathways for acetaldehyde
that were originally proposed are
CH3 CHO + → CH3 + [HCO]
→ CH3 + H + CO(radical decomposition),
CH3 CHO + → H2 + CH2 =C=O (elimination),
(1)
(2)
CH3 CHO + → [CH2 =CH − OH]
→ HC≡CH + H2 O(isomerization/elimination).
(3)
While products from all three pathways were observed, Vasiliou and co-workers discovered that some of the products came
from bimolecular reactions.10 They repeated the pyrolysis experiments on a 50:50 mixture of isotopically labeled acetaldehyde (CD3 CDO) and unlabeled acetaldehyde (CH3 CHO) and
detected the products using PIMS. This experiment resulted
in isotopically scrambled products, containing both deuterium
and hydrogen. If the products were formed from a unimolecular pathway, then they would have either deuterium
or hydrogen, not both. After ruling out wall reactions within
the hyperthermal nozzle, it was concluded that bimolecular
139, 214303-1
© 2013 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
214303-2
Hatten et al.
J. Chem. Phys. 139, 214303 (2013)
reactions must occur during pyrolysis under their experimental conditions.
Pyrolysis reactions of butyraldehyde can be hypothesized
based on the reactions observed in acetaldehyde by Vasiliou
and co-workers.9 The thermal decomposition mechanisms of
acetaldehyde and butyraldehyde at high temperatures should
be similar, but it is conceivable that butyraldehyde may have
additional reaction pathways due to its longer alkyl chain. For
example, there is a known mechanism11–13 for butyraldehyde
forming a cyclic intermediate where the oxygen abstracts a
γ -hydrogen from the terminal carbon, followed by the elimination of ethylene. Small aldehydes such as propionaldehyde
and acetaldehyde cannot access such a reaction pathway because their alkyl chains are too short to form the cyclic intermediate and bring the oxygen atom in close proximity to
a terminal hydrogen. Also, the C–C bonds in the β and γ
positions relative to the carbonyl group have bond energies
similar to the α C–C bond.14 While a literature search has
not indicated any previous observation of reactions involving
such bond ruptures, they should be considered in the potential
mechanism. In summary, the main reactions expected in the
thermal decomposition of butyraldehyde are
CH3 CH2 CH2 CHO + → CH3 CH2 CH2 + CO + H (radical decomposition), (4)
CH3 CH2 CH2 CHO + → CH3 CH2 CH=C=O + H2 (elimination),
(5)
CH3 CH2 CH2 CHO + → CH3 CH2 C≡CH + H2 O (isomerization/elimination),
(6)
CH3 CH2 CH2 CHO + → CH2 =CH2 + CH2 =C(H)OH (H−shift/elimination),
(7)
CH3 CH2 CH2 CHO + → CH3 CH2 + CH2 CHO (β−C − C scission),
(8)
CH3 CH2 CH2 CHO + → CH3 + CH2 CH2 CHO (γ −C − C scission).
(9)
Comparing the decomposition of butyraldehyde to that of acetaldehyde may show the influence of the size of the alkyl
chain on the reaction pathways.
The purpose of the experiments presented here is to
identify the products of butyraldehyde pyrolysis by using a
hyperthermal nozzle (Chen nozzle)15–17 and matrix-isolation
Fourier transform infrared (FTIR) spectroscopy. The hyperthermal nozzle provided an oxygen-free environment for thermal decomposition of gas-phase butyraldehyde, while matrixisolation FTIR was used to detect the products. Additionally,
PIMS of the pyrolysis products were measured in order to
complement the FTIR spectra and identify products that are
difficult to detect by FTIR. An advantage of the experiments
presented here is that they probe reactions occurring relatively
early in the pyrolysis process. The typical residence time for
a butyraldehyde molecule in the high-temperature region of
the pyrolyzer is on the order of 50–100 μs.9, 15 The expansion of pyrolysis products leaving the high-temperature tube
rapidly cools the products and quenches any reactions. Therefore, it is possible to detect and identify radical intermediates
or products that occur in the pyrolysis mechanism.
II. EXPERIMENTAL METHODS
Thermal decomposition of butyraldehyde (purity
≥99.5%; Sigma Aldrich) was accomplished via a hyperthermal nozzle that has been described elsewhere in the
literature.15 A matrix mixture (0.2%–0.33%) of butyraldehyde in argon totaling approximately 700 Torr was prepared
using standard manometric techniques. The mixture was
expanded from a pulsed valve (General Valve Series 9)
operating at 40 Hz into the 1.5 in. × 1 mm silicon carbide
(SiC) tube which was resistively heated. The temperature of
the SiC tube was controlled using a Series 16A temperature
controller made by Love controls. The pyrolysis products
were sprayed out of the tube and onto a cesium iodide (CsI)
window mounted in a cryostat (Janis Research) with a base
pressure of 1.0 × 10−6 Torr. The CsI window was cooled
to 15 K by a closed-cycle helium refrigerator (Sumitomo
Heavy Industries Ltd.) and regulated by a Lake Shore 331
Temperature Controller. The pyrolysis products were isolated
in an Ar matrix and then cooled to 4 K prior to FTIR analysis.
FTIR spectra were collected for 120 scans with 0.5 cm−1
resolution with a Bruker Vertex 70 spectrometer that was
purged with nitrogen gas.
Photoionization mass spectrometry was also performed
following the pyrolysis of butyraldehyde in separate experiments. A pyrolysis nozzle identical to the one described
above was mounted on a photoionization mass spectrometer
which has been described previously.15 A mixture of 0.3%
butyraldehyde in 2 atm of helium was expanded through the
pyrolysis nozzle (1300–1600 K) into a 10−5 Torr vacuum
chamber and passed through a skimmer (3 mm diameter).
The resultant beam was crossed with a vacuum ultraviolet
(VUV) (118.2 nm) laser operating at 30 Hz, ∼0.5 μJ/pulse.
The VUV light was generated by tripling the 3rd harmonic of
an Nd:YAG in an argon cell. Products from thermal decomposition of butyraldehyde were ionized and then mass analyzed
by a reflectron time of flight (TOF) mass spectrometer. Ions
were detected by a channeltron and the signal was collected
by a Tektronix digital oscilloscope.
III. RESULTS AND DISCUSSION
A list of all the products observed following pyrolysis
of butyraldehyde appears in Table I. PIMS analysis was performed following pyrolysis of butyraldehyde at various SiC
tube temperatures in order to confirm the presence of the
species found in the FTIR spectra, as well as to detect species
that were not discernible by matrix-isolation FTIR. The PIMS
spectra at 300 K, 1300 K, 1500 K, and 1600 K are shown in
Figure 1. The use of two separate detection techniques proved
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
214303-3
Hatten et al.
J. Chem. Phys. 139, 214303 (2013)
TABLE I. Summary of all products observed following the pyrolysis of butyraldehyde at 1300–1600 K, listed in order of molecular weight. Checkmarks indicate evidence observed for each product. A question mark indicates an assignment that is ambiguous or lacks sufficient signal to be certain.
Product
CH3
H2 O
HCCH
CH2 CH2
CO
HCO
CH2 CCH
CH3 CCH
CH3 CHCH2
H2 CCO
H2 CCHO
CH3 CH2 CH2
CH2 CHOH
CH3 CH2 CH3
C4 H2
C4 H4
CH3 CH2 C≡CH
CH2 CHCHO
C4 H8
CH3 CH2 CHCO
CH3 CH2 CHCHOH
PIMS
Matrixisolation FTIR
Comments
?
?
?
?
?
IE prohibits PIMS detection
IE prohibits PIMS detection
IE prohibits PIMS detection
?
?
?
?
IE prohibits PIMS detection
No literature comparison
for MI-FTIR
advantageous in several cases, for example, detection of the
methyl radical (ionization energy, IE = 9.8).18 CH3 + appears
at m/z = 15 in Figure 1. The methyl radical is expected to
exhibit the highest FTIR absorption at 603 cm−1 , with other
notable features at 1398 cm−1 and 3150 cm−1 .19 Bands were
observed at 1399 cm−1 and 3149 cm−1 in the spectra presented here, although there was no appreciable signal at 603
cm−1 . However, the mass spectra provide sufficient evidence
of the assignment.
The thermal decomposition of butyraldehyde led to the
production of several species, which were identified in the
FIG. 1. PIMS spectra following pyrolysis of butyraldehyde at 300 K,
1300 K, 1500 K, and 1600 K obtained with a 10.5 eV photoionization laser.
Pyrolysis was performed on 0.3% samples of butyraldehyde in 2 atm of He
carrier gas.
matrix-isolation FTIR spectra and the PIMS spectra. Many
of these were predicted from reactions (4)–(9). Others can
be explained by further decomposition reactions of the initial
products of butyraldehyde pyrolysis. Finally, some can best
be attributed to bimolecular reactions occurring within the hyperthermal nozzle. The assignment of the various products is
presented in the discussion below, which is divided into sections based on the purported reaction leading to the products
being discussed.
A. Products of radical decomposition
Radical decomposition of butyraldehyde produces the
propyl radical and HCO, which predominantly decomposes
to H + CO (reaction (4)). Figure 2 shows the appearance
of carbon monoxide at 2138 cm−1 .20 The CO product cannot be detected by PIMS because its ionization energy18 is
14.0 eV, above that of the photoionization laser. The propyl
radical should exhibit its strongest absorption at 530 cm−1
in the FTIR spectrum,19 but there is no evidence of that feature or any of the C–H stretches. A small peak at m/z = 43
is present in the PIMS spectrum shown in Figure 1 and suggests the presence of the propyl radical (IE = 8.1 eV),21
although the signal could also be an isotopic variant of
m/z = 42, where very strong PIMS signal occurs. The low
PIMS signal at m/z = 43 and imperceptible FTIR absorption
are likely due to secondary decomposition of the propyl radical, discussed in Sec. III G. Incidentally, the radical decomposition channel is thought to proceed by an initial C–C bond
fission, producing propyl radical + HCO. The HCO radical
dissociates to H + CO, although some HCO may be retained
at lower temperatures or short residence times.9 The mass
spectrum obtained following 1300 K pyrolysis does show evidence of HCO (IE = 8.12 eV)18 at m/z = 29, but the feature disappears at higher temperatures, indicating that a small
fraction of the radicals are indeed stable at lower temperatures. The FTIR experiment was repeated with 1300 K pyrolysis, and barely perceptible bands were observed at 1861,
1093/1089, and 2488 cm−1 (not shown here), which provides
FIG. 2. Matrix infrared absorption spectrum of products from the 1475 K
pyrolysis of a mixture of 0.2% butyraldehyde in argon.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
214303-4
Hatten et al.
FIG. 3. Matrix infrared absorption spectrum of products from the 1475 K pyrolysis of a mixture of 0.2% butyraldehyde in argon. The symbols indicate
bands that belong to unreacted butyraldehyde.
at least a hint of confirmation of HCO. To be thorough,
m/z = 29 also corresponds to CH3 CH2 + , and ethyl radical
(IE = 8.1 eV)22 could be anticipated from reaction (8). However, there was absolutely no evidence19 in the FTIR spectra
for ethyl radical, while there was (weak) spectral evidence
for HCO. This is somewhat surprising, as HCO has a lower
dissociation barrier and dissociation energy23 than the ethyl
radical.24
B. Products of elimination
The elimination reaction of butyraldehyde produces ethyl
ketene and H2 (reaction (5)). Ethylketene is evidenced by
bands in the FTIR spectrum at 1141 cm−1 (Figure 3),
2135 cm−1 , and 2130 cm−1 (Figure 2), matching bands
observed for matrix-isolated ethylketene in the literature.25
There is also a tiny feature at m/z = 70 (CH3 CH2 CH=C=O+ )
observable in the mass spectrum following 1000 K pyrolysis.
The coproduct of elimination, H2 (IE = 15.4 eV),18 cannot be
observed in the mass spectra.
J. Chem. Phys. 139, 214303 (2013)
FIG. 4. Matrix infrared absorption spectrum of products from the 1475 K
pyrolysis of a mixture of 0.2% butyraldehyde in argon. The band labeled
“enol” corresponds to the OH stretch of vinyl alcohol, although 1-buten-1-ol
is also expected to appear in the spectrum, and would likely exhibit an OH
stretch nearby.
elimination of water from acetaldehyde is preceded by a ketoenol tautomerism of acetaldehyde to vinyl alcohol. If a similar mechanism governs butyraldehyde, then 1-buten-1-ol, the
enol tautomer of butyraldehyde, may also be detected in these
experiments. While there are no vibrational bands in the literature for matrix-isolated 1-buten-1-ol for the purpose of comparison to the FTIR spectra presented here, there is an OH
stretch present at 3620 cm−1 (Figure 4) and a C=C stretch at
1663 cm−1 (Figure 5), very close to those observed in vinyl
alcohol at 3619 cm−1 and 1662 cm−1 .10 However, these may
very well belong to vinyl alcohol itself (vide infra), which is
expected following reaction (7).
Evidence of water and possibly 1-buten-1-ol in the spectra led to a search for 1-butyne, the other product of elimination in reaction (6). There is evidence at m/z = 54 for
C. Products of isomerization/elimination
The isomerization/elimination reaction of butyraldehyde
produces 1-butyne and water (reaction (6)). Water is commonly found as a contaminant in matrix-isolation FTIR spectra as trace amounts of water vapor in the vacuum chamber
can condense on the window during deposition. However,
the intensity of the OH stretches and bends observed here
(Figures 4 and 5) exceed what would be expected from contamination over the experimental deposition time, so it is concluded that H2 O must be a product of the pyrolysis of butyraldehyde. This was confirmed experimentally by passing
pure argon through the heated pyrolysis nozzle and comparing the resultant intensities of the vibrational bands of water to those observed following the pyrolysis of butyraldehyde. The ionization energy18 of water (12.6 eV) prohibits
detection via PIMS. Vasiliou and co-workers10 proposed that
FIG. 5. Matrix infrared absorption spectrum of products from the 1475 K
pyrolysis of a mixture of 0.2% butyraldehyde in argon. The symbols indicate bands that belong to unreacted butyraldehyde. The band labeled “enol”
corresponds to the C=C stretch of vinyl alcohol, although the same mode for
1-buten-1-ol would be expected in this region.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
214303-5
Hatten et al.
FIG. 6. Matrix infrared absorption spectrum of products from the 1475 K pyrolysis of a mixture of 0.2% butyraldehyde in argon. The symbols indicate
bands that belong to unreacted butyraldehyde.
1-butyne (IE = 10.2 eV)18 in the PIMS spectrum (Fig. 1) obtained following 1300 K pyrolysis, albeit a diminutive peak.
A recent, unpublished FTIR spectrum of 1-butyne isolated
in an argon matrix measured by Ellison and co-workers26
shows that the bands of highest intensity occur at 3329 cm−1 ,
1623 cm−1 , 632 cm−1 , and 626 cm−1 . The FTIR spectra measured here contain all of these bands, as well as many of the
weaker bands observed in Ellison’s spectrum. Observed features of 1-butyne are labeled in Figures 2, 6, and 7. The experimental results are also consistent with the published gasphase spectrum27 which shows strong bands at 635 cm−1 ,
1250 cm−1 , 2970 cm−1 , and 3320 cm−1 .
The mass spectra reveal that 1-butyne disappears with increasing temperatures, suggesting that it is undergoing secondary dissociation. The thermal decomposition pathways of
1-butyne are well-established in the literature. A very low
pressure pyrolysis study of 1-butyne at 1052–1152 K revealed
propargyl and methyl radicals as the primary products.28
This is noteworthy because the propargyl radical is a known
precursor to polycyclic aromatic hydrocarbons (PAHs) and
FIG. 7. Matrix infrared absorption spectrum of products from the 1475 K
pyrolysis of a mixture of 0.2% butyraldehyde in argon.
J. Chem. Phys. 139, 214303 (2013)
FIG. 8. Matrix infrared absorption spectrum of products from the 1475 K pyrolysis of a mixture of 0.2% butyraldehyde in argon. The symbols indicate
bands that belong to unreacted butyraldehyde.
soot.29, 30 A shock-tube study of 1-butyne in the 1100–1600 K
range resulted in the detection of methane, acetylene, ethylene, ethane, allene, propyne, C4 H2 , vinylacetylene, benzene, and 1,2- and 1,3-butadienes.31 As will be discussed in
Secs. III D–III G, there is spectral evidence consistent with
all of these products, except for methane, ethane, allene, benzene, and the butadienes.
D. Products of the hydrogen shift/elimination reaction
Reaction (7) shows the formation of ethylene and
vinyl alcohol, both of which are observed in these experiments. Ethylene is clearly evidenced in the FTIR spectra at
3083 cm−1 , 1441 cm−1 , 1343 cm−1 , and 947 cm−1 (Figures 3,
7, and 8), matching literature32 values. CH2 CH2 + appears at
m/z = 28 in the mass spectra shown in Figure 1. While ethylene’s ionization energy (10.5 eV)18 is very close to the energy of the ionization laser, observation of CH2 CH2 + is not
unreasonable as neutral ethylene may have significant internal energy, enhancing the probability of ionization. CO+ was
ruled out for the m/z = 28 peak in the mass spectra because
the ionization energy of carbon monoxide18 (14.0 eV) far
exceeds the energy of the ionization laser. As discussed in
Sec. III C, vibrational bands observed at 3620 cm−1 (Figure 4)
and 1663 cm−1 (Figure 5) are identical to those of vinyl alcohol. These bands were initially suspected as belonging to
1-buten-1-ol, but a definitive assignment of that enol cannot
be made for lack of a literature comparison. The PIMS data
prompted the assignment of these vibrational features to vinyl
alcohol. The peak at m/z = 44 in the mass spectra could belong to CH2 CHOH+ , CH3 CHO+ , or CH3 CH2 CH3 + . However, the ionization energy of propane18 (10.9 eV) is above
the photoionization laser energy, while that of vinyl alcohol22
is only 9.3 eV. The presence of acetaldehyde (CH3 CHO), the
keto tautomer of vinyl alcohol, would be expected, yet attempts to identify it in the FTIR spectra were futile because
many of the bands33 overlap with those of butyraldehyde.
Regardless, the bands observed for vinyl alcohol perfectly
matched those in the literature. Therefore, both products
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
214303-6
Hatten et al.
J. Chem. Phys. 139, 214303 (2013)
of reaction (7) can be assigned in this experiment with
confidence.
The revelation of reaction (7) enables an interesting connection to the pyrolysis of acetaldehyde. Now that vinyl alcohol, the enol tautomer of acetaldehyde, is recognized as a
product of butyraldehyde pyrolysis, it is conceivable that every product observed in the pyrolysis of acetaldehyde ought
also to appear in these experiments. Vasiliou et al.9 observed the following products following the thermal decomposition of acetaldehyde: CH3 , CO, H2 C=C=O, CH2 CHOH,
HC≡CH, and H2 O. Several of these products have been identified here already, and evidence for the remaining products
will be presented in Secs. III E–III G.
E. Possibility of β–C–C scission
Reaction (8) shows the breakage of the C–C bond in the
β position relative to the carbonyl of butyraldehyde, yielding ethyl and vinoxy radicals. This reaction has not been
observed in the limited number of experiments concerning
butyraldehyde in the literature, but it should be considered
important to the mechanism because the energy of the β–C–
C bond is 82.5 kcal/mol, slightly lower than the α–C–C bond
(84.1 kcal/mol) that is broken in reaction (4).14 As discussed
in Sec. III A, there is no strong evidence for CH3 CH2 from
these experiments, but that could be explained by secondary
dissociation of the ethyl radical:
CH3 CH2 → CH2 CH2 + H.
(10)
There is some evidence suggesting the presence of the cofragment vinoxy radical, CH2 CHO, in the matrix-isolation FTIR
spectra presented here. Vibrational bands are observed at
692, 723, 765, 1375, 1525, 1542, and 1558 cm−1 , matching
literature19 bands. Several of these bands are quite weak or
partially overlap with other bands, so the assignment is admittedly ambiguous. If vinoxy (IE = 10.2 eV)34 survives the
high-temperature nozzle, it may appear at m/z = 43 in the
PIMS spectra. However, observed signal at m/z = 43 was
very low and could also be attributed to propyl radical, or the
isotopic influence of m/z = 42. If the vinoxy has sufficient
internal energy to dissociate, possible product channels are
H + H2 C=C=O and CH3 + CO. The branching between
these channels is dependent upon the electronic state of vinoxy. In the ground state, the H + ketene channel is suppressed and virtually all vinoxy isomerizes to acetyl, then
proceeds to CH3 + CO.34, 35 This is contrary to vinoxy dissociation in the à and B̃ states, where both channels are observed and H + ketene are the major products.36, 37 Of course,
the products CH3 , CO, and ketene detected in these experiments could all be attributed to other reaction pathways of
butyraldehyde. While there is evidence suggesting that reaction (8) does occur, there is not absolute confidence in the
assignment.
F. Possibility of γ –C–C bond scission
While CH3 was described in Sec. III E as a possible
end result of reaction (8) via dissocation of vinoxy, it could
also come directly from scission of the γ –C–C bond in butyraldehyde, shown in reaction (9). The energy for this reaction is only a few kcal/mol higher than those of reactions
(5)–(8).14 The presence of CH3 , however, is not sufficient evidence for reaction (9) as there are many different sources of
methyl plausible in the reaction mechanism. The cofragment,
ĊH2 CH2 CHO, would almost certainly dissociate to CH2 CH2
+ CO + H (which are products of other reaction channels)
or to acrolein + H.35, 38 Acrolein (CH2 CHCHO) seems to be
a unique product, the presence of which would lend strong
support for assigning reaction (9). The FTIR spectra display
bands associated with both cis- and trans-acrolein,39, 40 including 2834, 2808, 1157, 1051, 999, 917, and 905 cm−1 .
Other observed bands overlap with those of unreacted butyraldehyde. Figure 3 highlights the bands that are clearly
seen. Acrolein (IE = 10.1 eV)41, 42 may also be detected by
PIMS (Figure 1), although m/z = 56 could also correspond to
1- or 2-butene.
G. Products from secondary reactions
With reasonable evidence suggesting six main decomposition pathways (reactions (4)–(9)) in butyraldehyde, this section turns to the identification of products from secondary reaction pathways. The propyl radical is a product of reaction
(4), but there is only a hint of evidence for it in the mass
spectrum collected following 1300 K pyrolysis, and none at
higher temperatures. Propyl radical is not easily detected in
these experiments because it is undergoing secondary dissociation. The propyl radical can undergo both C–C and C–H
beta scission43 in the following reactions:
n − C3 H7 → CH3 + CH2 =CH2 ,
(11)
n − C3 H7 → CH3 CH=CH2 + H.
(12)
There is evidence for the products of both reactions (11) and
(12) in the FTIR and PIMS data. CH3 + appears at m/z = 15 in
Figure 1. While not all of the FTIR bands corresponding to the
methyl radical were discernible, the mass spectra are convincing enough. A similar situation was encountered by Vasiliou
et al. in their attempts to detect methyl radicals from the pyrolysis of acetaldehyde.9 Ethylene was already identified in
Sec. III D in both matrix-isolation FTIR and PIMS experiments. While it is certainly reasonable to assume that reaction (11) is a main source of methyl radical and ethylene, it
bears repeating that these products could also originate from
reactions (7), (9), or (10), or decomposition of products vinoxy and 1-butyne.28, 31 Propene, likely from reaction (12),
is evidenced in Figures 3 and 7 at 998 cm−1 , 909 cm−1 , and
3083 cm−1 .44 C3 H6 + appears at m/z = 42 in Figure 1, although this feature could be attributed to H2 C=C=O+ as
well.
While the dissociation of propyl radical is a plausible
source of methyl radicals, the dissociation of 1-butyne, produced by reaction (6) in the pyrolysis of butyraldehyde, makes
methyl in conjunction with the propargyl radical, CH2 C≡CH.
There is evidence for the propargyl radical in these experiments at m/z = 39 (C3 H3 + ) in the mass spectra in Figure 1.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
214303-7
Hatten et al.
The characteristic vibrational frequencies of matrix-isolated
propargyl19 were not observed in the FTIR spectra presented
here, but that may be a function of pyrolysis temperature. The
feature at m/z = 39 is not evident following 1300 K pyrolysis, appearing only at 1500 K and 1600 K. The vibrational
spectra presented here were measured following pyrolysis at a
maximum temperature of 1475 K, which may have been just
under the threshold needed produce sufficient propargyl for
detection via FTIR.
Ketene, CH2 =C=O, is evidenced by FTIR bands at
3063 cm−1 and 2142 cm−1 , and by the PIMS peak at m/z
= 42. (Figures 1, 2, and 7) C2 H2 O+ and C3 H6 + overlap at
m/z = 42, and both ketene and propene have ionization energies near 9.7 eV, so the infrared bands are necessary to confirm the presence of ketene. As discussed in Sec. III E, Hloss from vinoxy radical can produce ketene, although this is
a suppressed channel in the ground state. Another plausible
source of ketene is the secondary reaction of vinyl alcohol
produced in reaction (7). Vinyl alcohol may tautomerize to
acetaldehyde, which is known to produce ketene via H2 elimination under pyrolysis conditions.9
Acetylene is shown in Figures 6 and 7 at 736 cm−1 ,
3302 cm−1 , and 3288 cm−1 .9 Its ionization energy is too high
(11.4 eV)18 to be detected via PIMS here. There are several
processes that could lead to acetylene during thermal decomposition of butyraldehyde. Elimination of H2 O from vinyl alcohol produces acetylene.9 If there is sufficient time for the
vinyl alcohol from reaction (7) to react prior to deposition
in the cold argon matrix, this secondary reaction could account for the presence of acetylene. Acetylene is also one of
the known products of thermal decomposition of 1-butyne,31
which is produced by reaction (6). Finally, acetylene is a primary product of ethylene pyrolysis,45 and there are several
reactions of butyraldehyde that lead to ethylene.
As mentioned above, the thermal decomposition of
1-butyne leads to myriad chemical species. The decomposition of 1-butyne following reaction (6) may account for
the detection of propyne, C4 H4 , and C4 H2 . Propyne is evidenced by FTIR bands46 at 3330, 2962, 2874, 1448, 634, 630,
and 624 cm−1 in Figures 6–8. PIMS signal at m/z = 40 in
Figure 1 could originate from propyne or allene, which have
somewhat similar ionization energies: 10.36 eV for propyne,
9.7 eV for allene. However, there is no evidence in the FTIR
spectra for allene product. C4 H4 and C4 H2 appear in the
mass spectrum following pyrolysis at 1600 K exclusively.
Pyrolysis was conducted at lower temperatures prior to the
matrix-isolation FTIR analysis, so the vibrational modes cannot be used to identify the specific isomers of C4 H4 and C4 H2
produced.
The presence of C4 H4 and C4 H2 is noteworthy,
because it suggests the capacity for bimolecular carbonbuilding reactions. Isomers of C4 H4 are suspected
precursors to benzene formation in certain flame conditions. Vinylacetylene can react with the vinyl radical to
form benzene.47 Vinylacetylene or butatriene can react with
H2 CCCCH to form aromatics via phenylacetene, pentalene,
and benzocyclobutene.48 These reactions are unlikely at low
temperatures but could be significant in certain combustion
or pyrolysis conditions. While it is thought that reactions of
J. Chem. Phys. 139, 214303 (2013)
even-numbered carbon species are less likely to contribute to
aromatic formation than odd-carbon species such as propargyl radical,49, 50 the presence of C4 H4 and C4 H2 suggests
another route to the eventual formation of polycyclic aromatic
hydrocarbons and soot in the pyrolysis of butyraldehyde.
The PIMS spectrum following 1300 K pyrolysis contains
signal at m/z = 56. This likely includes acrolein, as discussed
in Sec. III F, but C4 H8 + was also considered as a possible
product. The FTIR spectra presented here do not show signatures of 1-butene or 2-butene, likely because they were collected following 1475 K pyrolysis, and the presumed butene
appears to undergo subsequent decomposition at temperatures
above 1300 K. Evaluation of FTIR spectra collected following 1300 K pyrolysis (not shown here) did reveal features at
3037, 2930, 2871, 2864, 1444, 1408, and 1384 cm−1 . These
bands are very close to the reported values44 for cis-2-butene,
although the intensities are low and a few of the bands from
the literature were not observed.
H. The role of bimolecular reactions
While the reactions presented in this paper reasonably
explain the appearance of nearly all of the chemical species
observed in the experiments and listed in Table I, the detailed
mechanisms cannot be determined. What appears to be a simple unimolecular or concerted reaction, such as the elimination of H2 to form ethylketene in reaction (5), may involve
bimolecular processes such as abstraction by hydrogen atoms
that are liberated in reaction (4). In fact, Vasiliou and coworkers10 determined that both ketene and acetylene, shown
in reactions (2) and (3), are produced by bimolecular reactions in the pyrolysis of acetaldehyde. In their work, pyrolysis
of a 50:50 mixture of CH3 CHO and CD3 CDO resulted in production of H2 , D2 , and HD. The presence of HD means that
a bimolecular reaction must be involved in the elimination to
form ethylketene (reaction (2)). The formation of the vinyl
alcohol (enol) intermediate prior to elimination in reaction
(3) appears to be a unimolecular process, but Vasiliou et al.
suggest that a hydrogen atom then attacks the vinyl alcohol
to make CH2 CH2 OH, which dissociates to CH2 CH2 + OH.
H-atom abstraction from ethylene leads to acetylene, while
the OH radical can then attack other acetaldehyde molecules,
abstracting a hydrogen atom to form water.
In an effort to determine the molecularity of reactions
(4)–(9) for butyraldehyde, the matrix-isolation FTIR experiments were repeated by carrying out pyrolysis on a 1:7000
butyraldehyde:argon mixture. Lowering the concentration of
butyraldehyde should stifle, or at least severely limit, bimolecular reactions. The drawback to this approach is that it also
reduces the intensity of FTIR signal. Therefore, there is some
uncertainty in attributing the absence of a particular product
in the FTIR to the dependence upon a bimolecular process.
However, the experiments here do provide clues to the mechanisms at play. Table II shows a summary of products of various thermal decomposition reactions, whether they appear at
high and low matrix ratios, and conclusions from the observations. These results indicate that reaction (4) is likely unimolecular, while reactions (5) and (6) are likely bimolecular. However, reaction (6) is attributed to a two-step process
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
214303-8
Hatten et al.
J. Chem. Phys. 139, 214303 (2013)
TABLE II. Selected products of thermal decomposition of butyraldehyde examined following pyrolysis of 1:250 and 1:7000 mixtures in argon. Checkmarks
indicate the positive identification of the product via matrix-isolation FTIR.
Product
CH3 CH=CH2
CH2 =CH2
CH3 CH2 CH=C=O
CH3 CH2 C≡CH
CH2 =CHOH
Reaction(s) of origin
1:250
1:7000
Conclusions/notes
12 (via 4)
7, 10 (via 8), or 11 (via 4)
5
6
7
Reactions (4) and (12) are likely unimolecular.
At least one of reactions (7) and (10), and (11) is unimolecular.
Reaction (5) is likely bimolecular.
Reaction (6) is likely bimolecular, with the caveat that it could
proceed via an enol intermediate. Either step could be
bimolecular.
Reaction (7) may be bimolecular.
(keto-enol tautomerism followed by H2 O elimination) and it
is impossible to speculate on the molecularity of each step
without the ability to assign definitively the 1-buten-1-ol intermediate. Reaction (7) is ambiguous, but may be bimolecular.
This would be surprising, as it has previously been suggested
to be a unimolecular elimination.11 Further study of this reaction may be worthwhile.
A survey of the list of products in Table I also hints at the
existence of bimolecular reactions. For example, propane is
evidenced51 by FTIR in Figures 3, 6, and 8, but not PIMS, due
to its 10.9 eV ionization energy.18 It is difficult to conceive a
unimolecular path from butyraldehyde to propane. The combination of propyl radical and a hydrogen atom seems more
plausible. Similarly, the presence of products C4 H2 and C4 H4
cannot easily be explained by unimolecular processes. If these
products originate from the decomposition of 1-butyne, as observed in a published shock tube study,31 then bimolecular
reactions are certainly involved. It is also conceivable that reactions of acetylene could lead to these products via a mechanism that must include bimolecular reactions.52
IV. CONCLUSIONS
The thermal decomposition of gas-phase butyraldehyde
in argon has been studied in the 1300–1600 K range with
a hyperthermal nozzle. The products detected via matrixisolation FTIR and/or PIMS were: CO, HCO, CH3 CH2 CH2 ,
CH3 CH2 CH=C=O, H2 O, CH3 CH2 C≡CH, CH2 CH2 ,
CH2 =CHOH, CH2 CHO, CH3 , HC≡CH, CH2 CCH,
CH3 C≡CH, CH3 CH=CH2 , H2 C=C=O, CH3 CH2 CH3 ,
C4 H2 , C4 H4 , CH2 =CHCHO, and C4 H8 . At least six initial
reactions are associated with the thermal decomposition
mechanism, plus numerous subsequent reactions. Several
of the reactions appear to be bimolecular in nature. The
presence of propargyl radical, C4 H2 , and C4 H4 suggests
that the thermal decomposition of butyraldehyde could
lead to formation of polycyclic aromatic hydrocarbons,
an important consideration when evaluating the sooting
potential of industrial processes where aldehydes are exposed
to high temperatures. It is of note that these experiments are
qualitative in that they identify the products of butyraldehyde
pyrolysis. This paper’s comprehensive picture of the products
of pyrolysis combined with the consideration of possible
reactions at play in the mechanism should provide a natural
stepping stone to future studies modeling the kinetics and
establishing an overall mechanism for pyrolysis.
ACKNOWLEDGMENTS
This work was supported by an award from Research
Corporation for Science Advancement and a Faculty Startup Award from The Camille and Henry Dreyfus Foundation.
This material is also based upon work supported by MUADVANCE via the National Science Foundation under Grant
Nos. 0548113 and 0929997. Any opinions, findings, and conclusions or recommendations expressed in this material are
those of the author(s) and do not necessarily reflect the views
of the National Science Foundation. C.D.H. acknowledges a
fellowship from the SURE Program funded through the West
Virginia Research Challenge Fund, and administered by the
West Virginia Higher Education Policy Commission, Division of Science and Research, Grant No. HEPC.dsr.11.24;
AMEND 1. The authors thank Barney Ellison, Jong Hyun
Kim, and Jessie Porterfield for technical assistance in collecting the mass spectra presented in this paper.
1 C.
Lievens, D. Mourant, M. He, R. Gunawan, X. Li, and C.-Z. Li, Fuel 90,
3417 (2011).
2 C. Tessini, N. Mueller, C. Mardones, D. Meier, A. Berg, and D. von Baer,
J. Chromatogr. A 1219, 154 (2012).
3 R. R. Baker and L. J. Bishop, J. Anal. Appl. Pyrolysis 74, 145 (2005).
4 R. R. Baker and L. J. Bishop, J. Anal. Appl. Pyrolysis 71, 223 (2004).
5 E. Kabir and K.-H. Kim, J. Hazard. Mater. 188, 443 (2011).
6 M. B. Colket III, D. W. Naegeli, and I. Glassman, Int. J. Chem. Kinet. 7,
223 (1975).
7 T. Bentz, F. Striebel, and M. Olzmann, J. Phys. Chem. A 112, 6120 (2008).
8 K. Yasunaga, S. Kubo, H. Hoshikawa, T. Kamesawa, and Y. Hidaka, Int. J.
Chem. Kinet. 40, 73 (2008).
9 A. Vasiliou, K. M. Piech, X. Zhang, M. R. Nimlos, M. Ahmed, A. Golan,
O. Kostko, D. L. Osborn, J. W. Daily, J. F. Stanton, and G. B. Ellison, J.
Chem. Phys. 135, 014306 (2011).
10 A. K. Vasiliou, K. M. Piech, B. Reed, X. Zhang, M. R. Nimlos, M. Ahmed,
A. Golan, O. Kostko, D. L. Osborn, D. E. David, K. N. Urness, J. W. Daily,
J. F. Stanton, and G. B. Ellison, J. Chem. Phys. 137, 164308 (2012).
11 Y. Koga, T. Nakanaga, K. Sugawara, A. Watanabe, M. Sugie, H. Takeo, S.
Kondo, and C. Matsumura, J. Mol. Spectrosc. 145, 315 (1991).
12 J. M. Tadić, G. K. Moortgat, P. P. Bera, M. Loewenstein, E. L. Yates, and
T. J. Lee, J. Phys. Chem. A 116, 5830 (2012).
13 J. Tadić, I. Juranić, and G. K. Moortgat, J. Photochem. Photobiol., A 143,
169 (2001).
14 G. da Silva and J. W. Bozzelli, J. Phys. Chem. A 110, 13058 (2006).
15 X. Zhang, A. V. Friderichsen, S. Nandi, G. B. Ellison, D. E. David, J. T.
McKinnon, T. G. Lindeman, D. C. Dayton, and M. R. Nimlos, Rev. Sci.
Instrum. 74, 3077 (2003).
16 D. W. Kohn, H. Clauberg, and P. Chen, Rev. Sci. Instrum. 63, 4003 (1992).
17 J. A. Blush, H. Clauberg, D. W. Kohn, D. W. Minsek, X. Zhang, and P.
Chen, Acc. Chem. Res. 25, 385 (1992).
18 S. G. Lias, “Ionization energy evaluation” in NIST Chemistry WebBook,
NIST Standard Reference Database Vol. 69, edited by P. J. Linstrom and W.
G. Mallard (National Institute of Standards and Technology, Gaithersburg,
MD 20899, accessed 2013), see http://webbook.nist.gov.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29
214303-9
19 M.
Hatten et al.
E. Jacox, J. Phys. Chem. Ref. Data Monogr. 3, 1 (1994).
Givan, A. Loewenschuss, and C. J. Nielsen, J. Chem. Soc., Faraday
Trans. 92, 4927 (1996).
21 S. G. Lias, J. E. Bartmess, J. F. Liebman, J. L. Holmes, R. D. Levin, and
W. G. Mallard, “Ion energetics data” in NIST Chemistry WebBook, NIST
Standard Reference Database Vol. 69, edited by P. J. Linstrom and
W. G. Mallard (National Institute of Standards and Technology, Gaithersburg, MD 20899, accessed 2013), see http://webbook.nist.gov.
22 S. G. Lias and J. F. Liebman, “Ion energetics data” in NIST Chemistry
WebBook, NIST Standard Reference Database Vol. 69, edited by P. J. Linstrom and W. G. Mallard (National Institutes of Standards and Technology,
Gaithersburg, MD 20899, 2013), see http://webbook.nist.gov.
23 L. N. Krasnoperov, E. N. Chesnokov, H. Stark, and A. R. Ravishankara, J.
Phys. Chem. A 108, 11526 (2004).
24 J. M. Hostettler, A. Bach, and P. Chen, J. Chem. Phys. 130, 034303 (2009).
25 N. Y. Dugarte, M. F. Erben, R. M. Romano, M.-F. Ge, Y. Li, and V. C. O.
Della, J. Phys. Chem. A 114, 9462 (2010).
26 G. B. Ellison, private communication (2013).
27 Coblentz Society, “Evaluated infrared reference spectra” in NIST Chemistry Webook, NIST Standard Reference Database Vol. 68, edited
by P. J. Linstrom and W. G. Mallard (National Institute of Standards and Technology, Gaithersburg, MD 20899, accessed 2013), see
http://webbook.nist.gov.
28 K. D. King, Int. J. Chem. Kinet. 10, 545 (1978).
29 J. A. Miller and C. F. Melius, Combust. Flame 91, 21 (1992).
30 J. A. Miller and S. J. Klippenstein, J. Phys. Chem. A 107, 7783 (2003).
31 Y. Hidaka, T. Higashihara, T. Oki, and H. Kawano, Int. J. Chem. Kinet. 27,
321 (1995).
32 M. G. K. Thompson and J. M. Parnis, J. Phys. Chem. A 109, 9465 (2005).
33 R. J. H. Clark and J. R. Dann, J. Phys. Chem. 100, 532 (1996).
34 J. L. Miller, L. R. McCunn, M. J. Krisch, L. J. Butler, and J. Shu, J. Chem.
Phys. 121, 1830 (2004).
20 A.
J. Chem. Phys. 139, 214303 (2013)
35 B.
L. FitzPatrick, B. W. Alligood, L. J. Butler, S.-H. Lee, and J. J.-M. Lin,
J. Chem. Phys. 133, 094306 (2010).
36 M. L. Morton, D. E. Szpunar, and L. J. Butler, J. Chem. Phys. 115, 204
(2001).
37 D. L. Osborn, H. Choi, D. H. Mordaunt, R. T. Bise, D. M. Neumark, and
R. C. McMichael, J. Chem. Phys. 106, 3049 (1997).
38 B. FitzPatrick, J. Phys. Chem. A 115, 1701 (2011).
39 C. E. Blom, R. P. Mueller, and H. H. Guenthard, Chem. Phys. Lett. 73, 483
(1980).
40 Y. N. Panchenko, A. V. Abramenkov, and C. W. Bock, J. Struct. Chem. 52,
42 (2011).
41 C.-G. Zhan, J. A. Nichols, and D. A. Dixon, J. Phys. Chem. A 107, 4184
(2003).
42 K. Ohno, K. Okamura, H. Yamakado, S. Hoshino, T. Takami, and M.
Yamauchi, J. Phys. Chem. 99, 14247 (1995).
43 J. A. Miller and S. J. Klippenstein, J. Phys. Chem. A 117, 2718 (2013).
44 A. J. Barnes and J. D. R. Howells, J. Chem. Soc., Faraday Trans. 2 69, 532
(1973).
45 T. Tanzawa and W. C. Gardiner, Jr., Combust. Flame 39, 241 (1980).
46 C. Araujo-Andrade, A. Gomez-Zavaglia, I. D. Reva, and R. Fausto, J. Phys.
Chem. A 116, 2352 (2012).
47 H. Richter and J. B. Howard, Phys. Chem. Chem. Phys. 4, 2038
(2002).
48 A. Landera, A. M. Mebel, and R. I. Kaiser, Chem. Phys. Lett. 459, 54
(2008).
49 P. Osswald, U. Struckmeier, T. Kasper, K. Kohse-Hoeinghaus, J. Wang, T.
A. Cool, N. Hansen, and P. R. Westmoreland, J. Phys. Chem. A 111, 4093
(2007).
50 H. Wang and M. Frenklach, Combust. Flame 110, 173 (1997).
51 M. A. Goodman, R. L. Sweany, and R. L. Flurry, Jr., J. Phys. Chem. 87,
1753 (1983).
52 R. S. Becker and J. H. Hong, J. Phys. Chem. 87, 163 (1983).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
206.212.5.173 On: Fri, 03 Jan 2014 17:55:29