Products and mechanism of the gas-phase reaction of NO radicals
3
with a-pinene
Torsten Berndt* and Olaf Boge
Institut f ur T ropospharenforschung e.V ., Permoserstr. 15, 04303 L eipzig, Germany
The gas-phase reaction of NO radicals with a-pinene has been studied under Ñow conditions in the pressure range 20 \ p/mbar
3
\
200 at 298 K using gas chromatographyÈmass spectrometry/Ñame ionisation detection (GCÈMS/FID), MS and long-path
FTIR spectroscopy as detection techniques. NO radicals were generated by thermal decomposition of N O . He, N or O ÈN
3
2 5
2
2
2
mixtures served as carrier gas. In the absence of O in the carrier gas, a-pinene oxide (ca. 60%) and an organic nitrate (ca. 40%)
2
were found to be the main products with minor amounts of a-campholene aldehyde. The yields were slightly pressure dependent.
In the presence of O , pinonaldehyde was the predominant product. When NO was added for conversion of the formed peroxyl
2
radicals via RO ] NO ] RO ] NO , the following product yields were obtained at p \ 200 mbar and [O ] [ 1017 molecule
2
2
2
cm~3, pinonaldehyde 75 ^ 6%, a-pinene oxide 15 ^ 3%, organic nitrates (total) 14 ^ 3%, a-campholene aldehyde 3 ^ 1%. In the
absence of NO, nitroperoxy-group-containing substances were observed, arising from the reaction RO ] NO ] RO NO . A
2
2
2
2
reaction mechanism is proposed and a tropospheric application of the results is discussed.
Terpenes, in particular a-pinene, are amongst the most abun-
dant naturally emitted hydrocarbons in the troposphere, with
a globally averaged emission of 550 Tg a~1. On a global scale,
the emission of biogenic hydrocarbons exceeds the emission of
non-methane hydrocarbons with anthropogenic origin, 90 Tg
a~1.1 Thus, biogenic hydrocarbons play an important role in
the chemistry of the troposphere. Their degradation process
can be initiated by reaction with OH radicals, O and NO
pathways leading to the observed products. The sum of all
experimental Ðndings should allow us to postulate a reaction
mechanism applicable to the troposphere. For an extrapo-
lation of experimental results from laboratory studies to the
troposphere, a more detailed knowledge of the elementary
steps of the considered chemical process seems to be neces-
sary. Flow-tube experiments represent an e†ective way of elu-
cidating these processes.
3
3
radicals. Rate constants for most of these reactions at room
temperature are available with high accuracy.2h4 Tropo-
spheric lifetimes calculated with the assumption [OH] \ 106
Experimental
The experimental set-up has been described previously.11,12
Reactions were conducted in a 2.13 cm id quartz Ñow-tube in
molecule cm~3, [O ] \ 7.5 ] 1011 molecule cm~3 and
3
[
NO ] \ 2.5 ] 108 molecule cm~3 5 show that the night-time
the pressure range 20 \ p/mbar \ 200 at 298 K. NO radicals
3
3
reaction of NO radicals with terpenes is an important sink
were generated in a thermal pre-reactor, located in a side tube,
3
for such substances, for example a-pinene: q \ 5.2 h, q
\
by thermal decomposition of N O at ca. 450 K.
OH
O3
2 5
4
.3 h, q
\ 10.8 min. Note that the commonly used night-
NO3
N O ] M ] NO ] NO ] M
(1)
time concentration for NO radicals of 2.5 ] 108 molecule
2 5
3
2
3
cm~3 seems to be inappropriate for situations with relatively
For GC analysis, a GCÈMS/FID system (HP 5890 with
HP-MSD 5971) was connected to the tube ca. 85 cm down-
high terpene concentration levels because the terpenes may
suppress the NO radical concentration.6,7 In general, in spite
stream of the mixing point for the NO radicals by means of a
vacuum system. IdentiÐcation of the products, a-pinene oxide,
3
3
of the importance of terpenes in the troposphere, knowledge
of the product distributions and mechanisms of the radical- or
ozone-initiated degradation is very limited. This fact is valid
a-campholene aldehyde and pinonaldehyde, was achieved
using mass spectra and retention times of authentic samples.
Absolute calibrations have been performed for a-pinene and
a-pinene oxide by means of the FID. According to the method
of “e†ective carbon numberÏ,13 the response factor of a-
campholene aldehyde was assumed to be equal to that from
a-pinene oxide. Because of observed loss processes, especially
in the transfer unit to the GC-system, and problems with the
metering unit (low vapour pressure), a quantitative determi-
nation of pinonaldehyde by this method was found to be
defective.
especially for NO radical reactions.
3
Jay and Stieglitz8 investigated product distributions from
the reaction of NO radicals with a-pinene, b-pinene and
3
sabinene in synthetic air in a 10 l glass Ñask. As well as a large
number of unidentiÐed products (nitrates, hydroxynitrates and
ketonitrates), pinonaldehyde from a-pinene and nopinone
from b-pinene were detected, but no yields were obtained.
From an FTIR chamber study,9 the formation of nitro-
oxynitroperoxy compounds in the initial stage of the reaction
of NO3 radicals with a-pinene, b-pinene, *3-carene and D-
limonene was observed. Because of rapid particle formation,
Ðnal products of the gas-phase reaction could not be identi-
For on-line FTIR detection, a 2050 cm3 cell with a White
mirror system (optical path length 10 m) was attached to the
Ñow tube downstream of the port for the GC sampling. In the
spectral range 1000È4000 cm~1, by co-adding of 200 scans
and an instrumental resolution of 8 cm~1, FTIR analysis was
performed by a Nicolet spectrometer (Magna 750) with a
MCT detector. Spectra were recorded when either only the
Ðed. Recently, product yields from the reaction of NO rad-
3
icals with a-pinene obtained under atmospheric conditions in
a 200 m3 chamber have been published:10 pinonaldehyde
5
8%, organic nitrates (total) 12%, a-pinene oxide \3%.
Here, we investigate the gas-phase reaction of NO radicals
N O Ñow was switched on and the thermal pre-reactor was
3
2 5
with a-pinene under Ñow conditions, with special attention to
the product distribution. By variation of the experimental
conditions (system pressure, composition of the carrier gas), it
was expected to get information concerning the individual
cold or all Ñows (N O and a-pinene) were set and the pre-
2
5
reactor was hot. In the resulting di†erence spectrum, con-
sumed N O appeared as negative bands and all products, as
well as residual a-pinene, as positive bands. Calibration curves
2
5
J. Chem. Soc., Faraday T rans., 1997, 93(17), 3021È3027 3021