Hydroxyl-radical-initiated oxidation of isobutyl isopropyl ether under laboratory conditions related to the troposphere. Product studies and proposed mechanism

Article abstract of DOI:10.1039/a701766i

The products formed by the hydroxyl-radical-initiated oxidation of the model ether, isobutyl isopropyl ether [(CH₃)₂CHCH₂OCH(CH₃)₂], have been investigated by irradiating synthetic air mixtures containing the substrate, methyl nitrite, and nitric oxide at ppm levels in a Teflon bag reactor at room temperature. The decay of reactant and formation of products were monitored by gas chromatography, mass spectrometry and by HPLC. The molar yields of the major products (mol of product formed/mol of isobutyl isopropyl ether consumed) were as follows: acetone, 0.56 ± 0.04; isopropyl formate, 0.48 ± 0.03; isobutyl acetate, 0.28 ± 0.02; 2-hydroxy-2-methylpropyl acetate [CH₃C(O)OCH₂C(OH)(CH₃)₂], 0.25 ± 0.1. The molar yields of the minor products were as follows: isobutyraldehyde, 0.06 ± 0.05; isopropyl nitrate, 0.09 ± 0.06; 1,1,4-trimethyl-3-oxapentyl nitrate [(CH₃)₂CHOCH₂C(CH₃) ₂(ONO₂)], 0.07 ± 0.02; isopropyl isobutyrate [(CH₃)₂CHC(O)OCH(CH₃)₂] ca. 0.01; and isobutyl formate, ca. 0.01. The major products are explained by a mechanism involving initial OH attack at the -CH- and -CH₂- groups in the alkyl side chains of the ether followed by the subsequent reactions of the resulting carbon-centred, organic peroxy, and organic oxy radicals. The observed products, in conjunction with the proposed reaction pathways, account for a total yield of about 1.15, indicating that all the main routes are accounted for in the degradation of this ether. The major reaction pathways of the three principal organic oxy radicals are summarised as follows (percentage of overall reaction in brackets): (CH₃)₂C(O)OCH₂CH(CH₃) ₂ → CH₃C(O)OCH₂CH(CH₃)₂ + CH₃ (28%) (CH₃)₂CHOCH(O)CH(CH₃)₂ → (CH₃)₂CHOC(O)H + CH(CH₃)₂ (≤48%) (CH₃)₂CHOCH₂C(O)(CH₃) ₂ → (CH₃)₂COCH₂C(OH)(CH₃) ₂ (25%) This study supports the finding that organic oxy radicals generated from ethers and containing the structure RCH(O^.)OR undergo mainly decomposition by C-C bond cleavage, whereas those oxy radicals with the structure RCH(O^.)CH₂OR undergo preferential 1,5-H-atom transfer isomerisation reactions. The following rate coefficients (10^-12 cm³ molecule^-1 s^-1) at room temperature for the reactions of OH radicals with the reactant and products have been determined by the relative rate technique: isobutyl isopropyl ether, 19.5 ± 0.4; isobutyl acetate, 6.0 ± 0.5; isobutyraldehyde, 25.8 ± 0.7; isopropyl formate, 2.1 ± 0.1; isopropyl isobutyrate, 6.5 ± 0.4; 1,1,4-trimethyl-3-oxapentyl nitrate, 16.5 ± 0.7; and 2-hydroxy-2-methylpropyl acetate, 9.5 ± 1.6.

Full text of DOI:10.1039/a701766i

View Article Online / Journal Homepage / Table of Contents for this issue
Hydroxyl-radical-initiated oxidation of isobutyl isopropyl ether under
laboratory conditions related to the troposphere
Product studies and proposed mechanism
Konrad Stemmler, Wolfgang Mengon and J. Alistair Kerr¤*
EAW AG, Swiss Federal Institute for Environmental Science and T echnology, ET H Zurich,
CH-8600 Dubendorf, Switzerland
The products formed by the hydroxyl-radical-initiated oxidation of the model ether, isobutyl isopropyl ether
[(CH ) CHCH OCH(CH ) ], have been investigated by irradiating synthetic air mixtures containing the substrate, methyl
3 2
2
3 2
nitrite, and nitric oxide at ppm levels in a TeÑon bag reactor at room temperature. The decay of reactant and formation of
products were monitored by gas chromatography, mass spectrometry and by HPLC. The molar yields of the major products (mol
of product formed/mol of isobutyl isopropyl ether consumed) were as follows: acetone, 0.56 ^ 0.04; isopropyl formate,
0.48 ^ 0.03; isobutyl acetate, 0.28 ^ 0.02; 2-hydroxy-2-methylpropyl acetate [CH C(O)OCH C(OH)(CH ) ], 0.25 ^ 0.1. The
3
2
3 2
molar yields of the minor products were as follows: isobutyraldehyde, 0.06 ^ 0.05; isopropyl nitrate, 0.09 ^ 0.06; 1,1,4-trimethyl-
3-oxapentyl nitrate [(CH ) CHOCH C(CH ) (ONO )], 0.07 ^ 0.02; isopropyl isobutyrate [(CH ) CHC(O)OCH(CH ) ] ca.
3 2
2
3 2
2
3 2
3 2
0.01; and isobutyl formate, ca. 0.01. The major products are explained by a mechanism involving initial OH attack at the
wCHw and wCH w groups in the alkyl side chains of the ether followed by the subsequent reactions of the resulting carbon-
2
centred, organic peroxy, and organic oxy radicals. The observed products, in conjunction with the proposed reaction pathways,
account for a total yield of about 1.15, indicating that all the main routes are accounted for in the degradation of this ether. The
major reaction pathways of the three principal organic oxy radicals are summarised as follows (percentage of overall reaction in
brackets):
Æ
Æ
(CH ) C(O)OCH CH(CH ) ] CH C(O)OCH CH(CH ) ] CH
3 2 3 2 3 2 3
(28%)
(O48%)
(25%)
2
3
2
Æ
Æ
(CH ) CHOCH(O)CH(CH ) ] (CH ) CHOC(O)H ] CH(CH )
3 2 3 2 3 2 3 2
Æ
Æ
(CH ) CHOCH C(O)(CH ) ] (CH ) COCH C(OH)(CH )
3 2 3 2 3 2 3 2
2
2
This study supports the Ðnding that organic oxy radicals generated from ethers and containing the structure RCH(O~)OR
undergo mainly decomposition by CwC bond cleavage, whereas those oxy radicals with the structure RCH(O~)CH OR undergo
2
preferential 1,5-H-atom transfer isomerisation reactions. The following rate coefficients (10~12 cm3 molecule~1 s~1) at room
temperature for the reactions of OH radicals with the reactant and products have been determined by the relative rate technique:
isobutyl isopropyl ether, 19.5 ^ 0.4; isobutyl acetate, 6.0 ^ 0.5; isobutyraldehyde, 25.8 ^ 0.7; isopropyl formate, 2.1 ^ 0.1; isopro-
pyl isobutyrate, 6.5 ^ 0.4; 1,1,4-trimethyl-3-oxapentyl nitrate, 16.5 ^ 0.7; and 2-hydroxy-2-methylpropyl acetate, 9.5 ^ 1.6.
of the corresponding oxygen-containing alkoxy radicals is still
too limited to develop the empirical approach.3 To provide
more laboratory data on the atmospheric degradations of
oxygen-containing compounds, detailed product and mecha-
nistic studies of the atmospheric degradations of ethers4 and
glycol ethers have been carried out.5,6
The mechanism of elementary reactions for the breakdown of
volatile organic compounds in the troposphere is generally
agreed1,2 to involve the initial formation of carbon-centred or
alkyl radicals, mainly from attack of hydroxyl radicals on the
organic molecule. These carbon-centred radicals are rapidly
oxidised in air to the corresponding alkylperoxy radicals,
In the present study we have extended this approach by
investigating the mechanism of the atmospheric oxidation of a
model ether, isobutyl isopropyl ether. This molecule contains
two readily transferable tertiary CwH bonds together with a
sufficiently large carbon skeleton to enable the resulting oxy
radicals to undergo rapid isomerisation reactions.7 The
detailed quantitative product studies, that we have under-
taken, should provide further insight into the reactions, espe-
cially isomerisation, of oxygen-containing alkoxy radicals
under atmospherically related conditions.
which in a NO -containing atmosphere, principally react to
x
generate NO plus an oxygen-centred or alkoxy radical. The
2
subsequent reactions of the alkoxy radicals, either directly
with O , by decomposition or by isomerisation, give rise to
2
the Ðrst generation of carbonyl compounds such as aldehydes,
ketones, and esters, from the original organic compound. The
relative rates of the reactions of the alkoxy radicals under
atmospheric conditions consequently control the distribution
of initial carbonyl products which are formed.
For hydrocarbon molecules such as alkanes and alkenes,2
there is now a sufficient body of laboratory kinetic and ther-
modynamic data to enable the product distributions to be
estimated via sets of empirical rules for the reactions of the
hydrocarbon alkoxy radicals.3 In the case of oxygen-
containing organic compounds, the data base for the reactions
Experimental
Materials
The source or preparation of synthetic air, methyl nitrite and
nitric oxide have already been described.5 Isobutyl isopropyl
ether was prepared by reacting 0.5 mol sodium isopropylate
¤ Present address: School of Chemistry, University of Birmingham,
Birmingham, UK B15 2TT. Fax: ]44 121 472 8067. E-mail:
a.kerr=bham.ac.uk.
J. Chem. Soc., Faraday T rans., 1997, 93(16), 2865È2875
2865

View Article Online
and 0.5 mol isobutyl bromide at room temperature for three
days. The reaction produced isopropyl alcohol and iso-
butylene as by-products. The reaction mixture was distilled
twice and the remaining isopropyl alcohol was removed by
reÑuxing with sodium for several hours followed by vacuum
distillation. This puriÐcation procedure was repeated three
times. The purity of the resulting isobutyl isopropyl ether was
checked by GCÈFID, GCÈMS and NMR and found to be ca.
99%. Isopropyl isobutyrate was synthesised by reÑuxing a
mixture of 0.4 mol isopropyl alcohol and 0.4 mol isobutyryl
chloride for 3 h and was puriÐed by distillation. The purity
was found to be [99.5% (GCÈFID, GCÈMS). Isopropyl
formate was prepared by reÑuxing a mixture of 0.2 mol iso-
propyl alcohol and 1 mol formic acid overnight. The mixture
was neutralised with sodium hydrogen carbonate. The organic
layer was separated and washed with an aqueous sodium
hydrogen carbonate solution, dried with sodium sulfate and
distilled. The purity of the product was found to be [99.5%
(GCÈFID, GCÈMS). Isobutyl formate (purity [99%) was pre-
pared by the same procedure as for isopropyl formate with
isopropyl alcohol replaced by isobutyl alcohol. Isopropyl
nitrate ([99% purity) was synthesised8,9 by esteriÐcation of
0.05 mol isopropyl alcohol with nitric acid in the presence of
sulfuric acid at temperatures lower than 273 K. 2-Hydroxy-2-
methylpropyl acetate was synthesised as follows: a suspension
of 0.66 mol 1,2-epoxy-2-methylpropane (isobutylene oxide)
and 0.66 mol of ammonium acetate in 30 cm3 water was
stirred under reÑux for 5 h. A further 0.34 mol ammonium
acetate was then added and the two-phase mixture was stirred
for 10 h at 323 K to produce a homogeneous mixture, which
was extracted twice with 200 cm3 diethyl ether. The ether
extracts were washed with saturated aqueous sodium hydro-
gen carbonate solution, dried with sodium sulfate and vacuum
distilled with a 30 cm Vigreux column. The product was
checked with GCÈFID, GCÈMS and NMR and found to be
98% pure. The product 1,1,4-trimethyl-3-oxapentyl nitrate
was produced by the reaction of 0.04 mol 2-hydroxy-1-
isopropoxy-2-methylpropane with N-nitrocollidinum tetra-
Ñuoroborate as described by Olah et al.10 The reaction con-
verted only about 20% of the precursor alcohol into the
nitrate, but the nitrate could be isolated by twice repeated
liquid chromatography through a silica gel 60 (Merck) column
(2 cm ] 40 cm), using gradient mixtures of pentane and
diethyl ether. The product was conÐrmed by high resolution
mass spectrometry and by NMR analysis and the purity was
found to be ca. 97% (GCÈFID). The precursor 2-hydroxy-1-
isopropoxy-2-methylpropane was synthesised by the reaction
of 0.22 mol sodium isopropylate with 0.84 mol 1,2-epoxy-2-
methylpropane in 200 cm3 isopropyl alcohol. The mixture was
reÑuxed for 8 h followed by the addition of 50 cm3 water,
when the organic layer was separated and dried with sodium
sulfate. The isopropyl alcohol was distilled o† and the residue
was fractionated by vacuum distillation with a 25 cm Vigreux
column. The resulting product was [99% pure (GCÈFID,
GCÈMS, NMR). An attempt was made to synthesise 1,1,4-
trimethyl-3-oxapentyl nitrite by esteriÐcation of 2-hydroxy-1-
isopropoxy-2-methylpropane with sodium nitrite in the pres-
ence of sulfuric acid as described by Noyes.11 The product
could not be isolated by vacuum distillation or by liquid chro-
matography owing to the instability of this tertiary nitrite and
the resulting two-component mixture contained about 40% of
the nitrite along with the precursor alcohol, 2-hydroxy-1-
isopropoxy-2-methylpropane. The mixture was analysed by
1H, 13C, 17O NMR spectroscopy and by high resolution mass
spectrometry (GCÈMS) to conÐrm the identity of the 1,1,4-
trimethyl-3-oxapentyl nitrite and was used in a qualitative
photolysis experiment.
acetone (Fluka, [99.5%), dipropyl ether (Fluka, 99%), dibutyl
ether (Fluka, [99.5%), isobutyraldehyde (Fluka, [99%) and
acetic acid (Merck, 99.8%).
Apparatus
The relative rate coefficient measurements and the product
studies were carried out in an experimental system which has
been described previously4,7,12 and which is summarised here.
Hydroxyl radicals were produced from the photolysis of
methyl nitrite in air containing NO:
CH ONO ] hl ] CH O ] NO
(1)
(2)
(3)
3
3
CH O ] O ] CH O ] HO
3
2
2
2
HO ] NO ] OH ] NO
2
2
The experiments were performed at room temperature
(297 ^ 3 K) and at atmospheric pressure (725 ^ 10 Torr) in a
200 dm3 TeÑon bag reactor surrounded by 16 black lamps
(Philips L20/05) which provide UV-radiation in the region of
350È450 nm. Reactant mixtures were prepared by sweeping
measured amounts (capacitance manometer MKS Baraton
220 C) of isobutyl isopropyl ether and methyl nitrite vapours
from a calibrated volume of a vacuum line into the TeÑon bag
with a stream of synthetic air. The bag was Ðlled with up to
200 dm3 of synthetic air using mass Ñow controllers and a
stopwatch and nitric oxide was added to minimise the forma-
tion of O . Once the gas mixture was prepared, the bag was
3
agitated and left to stand for about 0.5È1 h in the dark to
allow the reactants to mix. The bag mixture was irradiated for
up to 1.5 h during which time samples were periodically
removed and analysed.
Relative rate coefficient measurements
The relative rate measurements of OH reactions with isobutyl
isopropyl ether and with some of its observed products (TST)
were each carried out in two separate experiments with either
dipropyl ether or dibutyl ether as the reference substances
(REF):
OH ] TST ] products
OH ] REF ] products
(4)
(5)
Assuming that there are no loss processes of the test (TST)
and reference compound (REF) other than the reaction with
OH radicals, the rate expressions for these processes are given
by:
[d[TST]/dt \ k [OH][TST]
(I)
TST
[d[REF]/dt \ k [OH][REF]
(II)
REF
which can be integrated and combined to give the following
equation:
ln ([TST] /[TST] ) \ (k /k ) ] ln ([REF] /[REF] )
0
t
TST REF
0
t
(III)
where the subscripts 0 and t indicate concentration at the
beginning of the experiment and at time t respectively. A plot
of ln ([TST] /[TST] ) vs. ln ([REF] /[REF] ) thus yields the
0
t
0
t
rate coefficient ratio k /k . The bag mixture, initially con-
TST REF
taining the test and reference compounds, nitric oxide and
The following compounds were used without further puriÐ-
cation, other than degassing by repeated thaw and freeze
cycles under vacuum: isobutyl acetate (Fluka, ca. 99%),
methyl nitrite, was irradiated for up to 2 h during which time
samples of the mixture were periodically removed via a 3 cm3
stainless steel loop and gas-sampling valve. The test and the
2866
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
reference compounds were analysed by GC (Carlo Erba
HRGC 5300) with a FID detector equipped either with a
combination of a 30 m DB-XLB fused silica capillary column
(J&W ScientiÐc, 0.25 mm id, 0.25 lm Ðlm) and a 25 m DB-5
fused silica capillary column (J&W ScientiÐc, 0.25 mm id, 0.25
lm Ðlm) or a 30 m DB-Wax fused silica column (J&W Scienti-
Ðc, 0.32 mm id, 0.25 lm Ðlm) operated with temperature pro-
gramming from 313 to 453 K.
ducts, was analysed in the second GC containing a less polar
DB-5 capillary column. Both GC columns were operated for 1
min at 313 K after injection and were then heated to 453 K
with a heating rate of 20 K min~1. In these three experiments
the mixing ratios of NO and NO were measured using a
2
commercial chemiluminescent analyser (Thermo Environ-
mental Instruments Model 42). The NO mixing ratios were
2
corrected for the contribution of methyl nitrite to the NO
x
signal. The approximate mixing ratios of methyl nitrite were
determined by GCÈFID using the DB-5 capillary column as
described above.
Analyses of products
Not all of the products could be detected and quantiÐed
under the same analytical conditions and consequently several
sets of experiments were carried out. A total of 12 quantitative
experiments to measure the products of the OH-radical-initi-
ated oxidation of isobutyl isopropyl ether were performed.
In experiments no. 1È3, the decay of the starting material,
isopropyl isobutyl ether, and the formation of the major pro-
ducts, acetone, isopropyl formate and isobutyl acetate, were
measured by GCÈFID (Carlo Erba HRGC 5300). Vapour
samples were injected via 3 cm3 gas sample loop on a com-
bination of a 30 m DB-XLB and a 25 m DB-5 fused silica
capillary column maintained at 313 K for 1 min and then
heated to 453 K at the rate of 10 K min~1. This column com-
bination was necessary to resolve the signal peaks of methyl
nitrate, isobutyraldehyde and isopropyl formate. To check
further for interference arising from peak overlaps, this
column combination was tested using a quadrupole mass
spectrometer detector (Fisons MD800) which enables the
measurement of speciÐc signals for each compound. It was not
possible to measure the minor product isobutyraldehyde
under these conditions owing to interference from the methyl
nitrate signal.
In experiments no. 4È7, the starting compound and the
major products were again measured as described above.
Additionally, the minor products isopropyl isobutyrate, iso-
propyl nitrate, and isobutyl formate were analysed by with-
drawing a gas sample from the bag through a 3 cm3 sampling
valve and injecting onto a DB-1701 capillary column (J&W
ScientiÐc, 30 m, 0.25 mm id, 1 lm Ðlm) with temperature pro-
gramming from 313 to 433 K. The products were detected
with a mass spectrometer (Fisons MD800, 70 eV electron
impact ionisation) operated in the single ion mode, recording
ions with m/z \ 71.0 ^ 0.2, and 89.0 ^ 0.2 for isopropyl iso-
butyrate; m/z \ 56.0 ^ 0.2 and 60.0 ^ 0.2 for isobutyl
formate; and m/z \ 90.0 ^ 0.2 and 46.0 ^ 0.2 for isopropyl
nitrate, during the elution period of each compound.
In experiments no. 7È9, isobutyl isopropyl ether and the
major products were measured as described above and the
minor carbonyl compound isobutyraldehyde was quantitat-
ively determined by derivatisation with 2,4-dinitro-
phenylhydrazine (2,4-DNPH) followed by HPLC separation
with a 250 ] 4 mm id Zorbax ODS (5 lm particle size)
column (Saulentechnik Knauer) using a three-component gra-
dient solvent programme and a UVÈVIS detector set to 360
nm.13 Samples of the gas mixture (2 dm3) were withdrawn
from the bag at a Ñow rate of 0.5 dm3 min~1 through an
impinger containing 4 cm3 of acidiÐed 2,4-DNPHÈacetonitrile
solution cooled to 273 K.13 During this sampling procedure it
was not possible to prevent small amounts of acetonitrile from
entering the bag reactor which precluded the analysis of
acetone by gas chromatography owing to co-elution.
In experiments no. 10È12, vapour samples were injected
simultaneously via 3 cm3 sample loops into two gas chro-
matographs equipped with FID detectors. The Ðrst GC was
Ðtted with a 30 m DB-Wax fused silica column to analyse the
polar compounds 1,1,4-trimethyl-3-oxapentyl nitrate and 2-
hydroxy-2-methylpropyl acetate. Isobutyl isopropyl ether,
which had a retention time of only 0.7 min on this column
and was suspected of co-eluting with other rapidly eluted pro-
In all of the above analyses, the compounds were calibrated
by preparing at least four bag mixtures containing standards
in air at mixing ratios comparable to those in the experiments
and by analysing with the same procedures as in the experi-
ments to yield response factors obtained from a linear least-
squares Ðt of the data. For qualitative GCÈMS analysis,
vapour samples were preconcentrated on Tenax TA adsorp-
tion tubes followed by thermal desorption into a GCÈMS as
described previously.6
Results
Rate coefficient measurements
The rate coefficients of the OH radical reactions with isobutyl
isopropyl ether, isobutyl acetate, isobutyraldehyde, isopropyl
formate, and isopropyl isobutyrate were each determined in
two separate relative rate measurements involving dipropyl
ether as the reference compound and the DB-XLBÈDB-5 GC
column combination for chromatographic separation. The
rate coefficients of 1,1,4-trimethyl-3-oxapentyl nitrate and 2-
hydroxy-2-methylpropyl acetate were each determined relative
to dibutyl ether in two separate experiments involving the
DB-Wax capillary column. The initial mixing ratios in the bag
reactor were as follows: isobutyl isopropyl ether, 3.2 and 3.0
ppm; isobutyl acetate, 3.5 and 3.7 ppm; isobutyraldehyde, 3.1
and 3.2 ppm; isopropyl formate, 3.9 and 5.3 ppm; isopropyl
isobutyrate, 3.3 and 2.9 ppm; 1,1,4-trimethyl-3-oxapentyl
nitrate, 0.5 and 1.1 ppm; 2-hydroxy-2-methylpropyl acetate,
1.5 and 1.6 ppm; dipropyl ether, 2.8È5.4 ppm; dibutyl ether,
2.0È4.0 ppm; NO, 8.9È20 ppm; and methyl nitrite, 5.5È29
ppm. Typical lnÈln plots of the present data according to eqn.
(III) are shown in Fig. 1 and 2 and the rate coefficient ratios
derived from the slopes of linear least-square Ðts of the data
are listed in Table 1. The error limits correspond to 95% con-
Ðdence intervals. These rate coefficient ratios were converted
to absolute rate coefficients on the basis of the evaluated
values1 for k(OH ] dipropyl ether) \ 18.5 ] 10~12 cm3
molecule~1 s~1 and k(OH ] dibutyl ether) \ 28.8 ] 10~12
cm3 molecule~1 s~1 at 298 K and are shown in Table 1. In
general, linear lnÈln plots with near zero intercepts and good
reproducibility were observed for these competitive experi-
ments. Both sets of data from the experiments competing 2-
hydroxy-2-methylpropyl acetate against dibutyl ether,
however, revealed distinct curvature in their plots, as shown in
Fig. 2. The possibility of GC interference from the products of
the oxidation of dibutyl ether in the analysis of 2-hydroxy-2-
methylpropyl acetate was excluded, by the irradiation of a
mixture of dibutyl ether, nitric oxide and methyl nitrite and
analysis under the same conditions. However, interference
from its own oxidation products cannot be ruled out. In addi-
tion, in the competitive experiments with 2-hydroxy-2-methyl-
propyl acetate, higher than normal methyl nitrite
concentrations were employed to reduce the importance of
wall loss processes compared to the loss via OH reaction. The
resulting increased removal rates of 2-hydroxy-2-methylpropyl
acetate and dibutyl ether reduced the number of analyses
taken during the run (see Fig. 2). Finally, it should be pointed
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
2867

View Article Online
out that systematic errors in obtaining gas kinetic data for
2-hydroxy-2-methylpropyl acetate are likely to arise from the
low volatility and polar nature of this compound which may
cause it to be adsorbed on the surfaces of the reactor and of
the analytical system.
Product measurements
The observed product mixing ratios were corrected for sec-
ondary removal by the OH radical according to the method
of Atkinson et al.14 using the expressions:
F[prod] \ f*[IBIPE]
(IV)
where
A
k
[ k
B
IBIPE
prod
F \
k
IBIPE
A
1 [ [IBIPE] /[IBIPE]
B
([IBIPE] /[IBIPE] )kprod@kIBtIPE [ [IBIPE] /[IBIPE]
0
]
t
0
t
0
(V)
k
and k
are the room temperature rate coefficients of
IBIPE
prod
the OH radical reactions with isobutyl isopropyl ether and
with the given product and *[IBIPE] \ [IBIPE]
0
[ [IBIPE] , where [IBIPE] refers to the concentration of
t
isobutyl isopropyl ether and the subscripts denote the time at
which it was measured. The fractions of isobutyl isopropyl
ether which form the given product, f, were derived by linear
least-squares Ðts of the slopes of the data plotted according to
eqn. (IV). The rate coefficients of the OH radical reactions
needed to calculate the correction factors, F, which have been
measured in this work are given in Table 1. The remaining
OH rate coefficients are either taken from the literature or
estimated using the structure activity relationship (SAR) of
Fig. 1 Sample plots of ln ([TST] /[TST] ) vs. ln ([dipropyl
0
t
ether] /[dipropyl ether] ) for the reaction of OH with following test
0
t
compounds (TST): (>) isobutyl isopropyl ether, (…) isobutyl acetate,
(+ ) isopropyl formate, (@) isopropyl isobutyrate and (=) iso-
butyraldehyde
Kwok
k(OH ] acetone) \ 2.19 ] 10~13 cm3 molecule~1 s~1 at 298
(ref. 1), k(OH ] isopropyl nitrate) \ 4.9 ] 10~13 cm3
molecule~1 s~1 at 298 (ref. 1), k(OH ] isobutyl
and
Atkinson15
and
are
as
follows:
K
K
formate) \ 4.55 ] 10~12 cm3 molecule~1 s~1 at 298 K (SAR
estimation15). The maximum values of the correction factors,
F, (corresponding to the largest extent of the reaction) are
given in the caption of Fig. 3 for the typical sets of results
shown. In order to obtain an estimation of the error intro-
duced via this correction procedure, the rate coefficients of iso-
butyl isopropyl ether and the products with the OH radical
were varied simultaneously in the range of the errors given in
Table 1 and the e†ects of these variations were included in the
errors of the product yields in Table 2. For the two rate coeffi-
cients measured competitively against dibutyl ether, an addi-
tional variation range of ^20% was employed to account for
systematic errors between the two di†erent reference com-
pounds utilised.
Fig. 2 Sample plots of ln ([TST] /[TST] ) vs. ln ([dipropyl
0
t
ether] /[dipropyl ether] ) for the reaction of OH with the following
0
t
test compounds (TST): (=) 1,1,4-trimethyl-3-oxapentyl nitrate and
(…) 2-hydroxy-2-methylpropyl acetate
Table 1 Room temperature rate coefficients for reaction of OH radicals with oxygenated compounds determined in this study
present study
ka/10~12 cm3 molecule~1 s~1
literature data
compound
(k /k )a
kd/10~12 cm3 molecule~1 s~1
TST REF
isobutyl isopropyl ether
isobutyl acetate
isobutyraldehyde
1.05 ^ 0.02b
0.32 ^ 0.03b
1.39 ^ 0.04b
19.5 ^ 0.4
6.0 ^ 0.5
25.8 ^ 0.7
18.1 ^ 2.1e
27 ^ 5f
24.2 ^ 3.3g
46.3 ^ 7.4h
isopropyl formate
isopropyl isobutyrate
1,1,4-trimethyl-3-oxapentyl nitrate
2-hydroxy-2-methylpropyl acetate
0.11 ^ 0.01b
0.35 ^ 0.02b
0.57 ^ 0.02c
0.33 ^ 0.06c
2.1 ^ 0.1
6.5 ^ 0.4
16.5 ^ 0.7
9.5 ^ 1.6
a Error limits are for 95% conÐdence intervals. b REF \ dipropyl ether. c REF \ dibutyl ether. d Errors are 2p. e Ref. 20. f Ref. 17. g Ref. 18. h Ref.
19.
2868
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Table 2 Correcteda yields of the major products from the photo-
oxidation of isobutyl isopropyl ether
observed product
yieldb,c
acetone
isopropyl formate
isobutyl acetate
isobutyraldehyde
isopropyl isobutyrate
isobutyl formate
isopropyl nitrate
1,1,4-trimethyl-3-oxapentyl nitrate
2-hydroxy-2-methylpropyl acetate
55.8 ^ 4.2
47.7 ^ 3.2
28.3 ^ 1.7
5.5 ^ 4.7
0.84 ^ 0.32
0.59 ^ 0.77
8.70 ^ 5.7
7.1 ^ 1.5
25.2 ^ 9.6
a Corrected for secondary OH radical attack (see text). b Expressed as
a percentage of isobutyl isopropyl ether reacted. c Error limits are for
95% conÐdence intervals and include the errors in the analytical
responses of isobutyl isopropyl ether and the respective product, the
scatter of the results and the worst case estimates of the errors in
correcting for secondary OH reactions.
nitrite and NO were 3.0È4.3 ppm, 11È18 ppm, and 10È17
ppm, respectively. Fig. 3(a) shows typical yields of isobutyl
acetate, isopropyl formate and acetone, corrected for second-
ary OH reactions, vs. the mixing ratios of isobutyl isopropyl
ether removed by OH reaction during the course of the given
experiment. In four experiments (no. 4È7) quantiÐcation of the
minor products isopropyl nitrate, isopropyl isobutyrate and
isobutyl formate was carried out and Fig. 3(b) shows typical
yields of these compounds vs. the mixing ratios of isobutyl
isopropyl ether removed by OH reaction for a given experi-
ment. Three of these experiments (no. 7È9) above included
measurements of the carbonyl product isobutyraldehyde via
derivatisation with 2,4-dinitrophenylhydrazine (2,4-DNPH).
Additionally, acetone was detected as the main carbonyl con-
taining product, but no calibrations were carried out to quan-
tify this compound by this analytical method. No additional
unknown 2,4-dinitrophenylhydrazone derivatives were
observed. For a typical experiment, a plot of the yield of iso-
butyraldehyde, corrected for secondary OH removal, vs. the
amount of isobutyl isopropyl ether reacted is shown in Fig.
3(c).
In experiments no. 10È12 the mixing ratios of isobutyl iso-
propyl ether, 1,1,4-trimethyl-3-oxapentyl nitrate and 2-
hydroxy-2-methylpropyl acetate have been determined.
Additionally several unknown minor products were observed,
which were all thought to be formed by reactions of OH rad-
icals with the Ðrst generation products, since their signals did
not increase linearly with the amount of isobutyl isopropyl
ether consumed. The most abundant of these secondary pro-
ducts was identiÐed to be acetic acid using a liquid standard
of this compound. The sources of acetic acid were found, by
qualitative irradiations of methyl nitriteÈNOÈprimary product
mixtures, to be mainly the OH-initiated oxidation of 2-
hydroxy-2-methylpropyl acetate and, as a minor contribution,
the secondary reactions of isobutyl acetate. The initial mixing
ratios of isobutyl isopropyl ether, methyl nitrite and nitric
oxide in the latter three experiments were 5.7È6.4 ppm, 13È15
ppm and 8È11 ppm respectively. A plot of the mixing ratios of
1,1,4-trimethyl-3-oxapentyl nitrate and 2-hydroxy-2-methyl-
propyl acetate, corrected for secondary OH removal, vs. iso-
butyl isopropyl formate removed during the irradiation for a
typical experiment is shown in Fig. 3(d).
Fig. 3 Plots of the mixing ratios, corrected for secondary reactions
with OH radicals, of (…) acetone (F \ 1.007), (>) isopropyl formate
(F \ 1.08), (=) isobutyl acetate (F \ 1.22), (L) isopropyl nitrate
(F \ 1.005), (K) isopropyl isobutyrate (F \ 1.1, plot displaced by
[0.01 ppm on the vertical axis), (|) isobutyl formate (F \ 1.07), (@)
isobutyraldehyde (F \ 1.43), (+ ) 1,1,4-trimethyl-3-oxapentyl nitrate
(F \ 2.15), (+) 2-hydroxy-2-methylpropyl acetate (F \ 1.61) against
the mixing ratios of isobutyl isopropyl ether reacted with the OH
radical
The overall results of all the experiments are summarised in
Table 2, where the quoted values are weighted mean values of
the results of the various experiments with error limits for
95% conÐdence intervals, derived as described by Cvetanovic
et al.,16 including errors in the analytical responses, the scatter
of the results and worst case estimates of the errors in correct-
ing for secondary OH reactions.
Not all of the identiÐed products could be quantiÐed in any
given experiment. Thus, nine experiments (no. 1È9) were
carried out to measure principally the formation of isobutyl
acetate, isopropyl formate and acetone. In these experiments
the initial mixing ratios of isobutyl isopropyl ether, methyl
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
2869

View Article Online
combined to give an average value of ca. 10.6 ] 10~12 cm3
molecule~1 s~1 at 298 K. For the reaction of diisobutyl ether
with the OH radical, the only reported value is that of Bennett
and Kerr,28 i.e. k \ 26.1 ] 10~12 cm3 molecule~1 s~1. Hence
the average of the rate coefficients of diisobutyl and diiso-
propyl ethers with the hydroxyl radical gives an estimate of
the rate coefficient of isobutyl isopropyl ether of ca.
18.4 ] 10~12 cm3 molecule~1 s~1, which is in good agree-
ment with the measured value reported here. This conclusion
can be interpreted as giving support to the rate coefficient for
diisobutyl ether reported by Bennett and Kerr,28 although
some of the values they reported for other ethers appear to
have been underestimated.1
Discussion
Rate coefficient determinations
To the best of our knowledge, of the OH rate coefficients that
we have measured, there are literature data only for the reac-
tion of isobutyraldehyde as listed in Table 1. The present
room temperature rate coefficient is in good agreement with
that data of Kerr and Sheppard17 and of Semmes et al.,18 and
is in line with the evaluated rate coefficient of Atkinson21 of
27.4 ] 10~12 cm3 molecule~1 s~1. On the other hand the
values of Do be et al.19 and of Audley et al.20 are factors of ca.
1.8 higher and ca. 1.4 lower respectively than our determi-
nation.
The present room temperature rate coefficient of isobutyl
isopropyl ether with the OH radical can be compared to the
corresponding literature data for diisopropyl and diisobutyl
ethers, on the basis that the reactivity of ethers towards OH
attack can be expressed as the sum of the independent contri-
butions of both alkyl side chains.22,23 For diisopropyl ether,
the rate coefficients of McLoughlin et al.24 of 11.1 ] 10~12
cm3 molecule~1 s~1, Mellouki et al.25 of 10.2 ] 10~12 cm3
molecule~1 s~1, Wallington et al.26 of 9.79 ] 10~12 cm3
molecule~1 s~1 and the relative and absolute data of Nelson
et al.27 of 11.3 ] 10~12 cm3 molecule~1 s~1 and 10.7 ] 10~12
cm3 molecule~1 s~1 are in excellent agreement and can be
Product analyses and proposed mechanisms
Reaction Schemes 1, 2 and 3 show proposed mechanisms for
the breakdown of isobutyl isopropyl ether following the initial
H-atom abstraction by the OH radical from the secondary
and the two tertiary CwH groups in the molecule, including
all the possible major channels. The observed products, upon
which the proposed reactions are based, are highlighted in
boxes. In most of these schemes the organic peroxy radicals
produced by the reaction of the carbon-centred radicals with
O
clarity.
and all unobserved alkyl nitrates have been omitted for
2
Scheme 1
2870
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Scheme 1 shows the reactions occurring after H-atom
abstraction by the hydroxyl radical from the tertiary CwH
bond of the isopropyl side chain in isobutyl isopropyl ether.
The intermediate organic oxy radical 2 appears to decompose
mainly via rupture of a CwC bond to form the principal
product, isobutyl acetate and a methyl radical. An additional
decomposition pathway, involving the rupture of the CwO
bond, leads to the formation of acetone and the alkoxy radical
3, which is believed to lead mainly to isobutyraldehyde29,30
via reaction with oxygen. The relatively low yield of iso-
butyraldehyde, which can also be formed from another reac-
tion in the oxidation of isobutyl isopropyl ether (see Scheme
2), indicates that the breakdown via CwO breakage is a
minor channel of oxy radical 2. This is in agreement with the
Ðnding of Wallington et al.,26 that the OH-radical-initiated
oxidation of diisopropyl ether leads to isopropyl acetate
(yield: 105 ^ 6%) as the only observed product, and therefore
the intermediate 1-isopropoxy-1-methylethyl oxy radical
[(CH ) C(O~)OCH(CH ) ] does not react by CwO bond
was observed by qualitative MS analyses of the bag mixtures,
although formic acid was observable under the experimental
conditions. As a kinetically less favourable reaction channel,
oxy radical 7 may decompose to give the hemiacetal 11 and a
methyl radical. Compound 11 would be expected to decom-
pose to form acetone and hydroxyacetone during analysis, but
the latter compound was not observed. Although the pro-
posed reactions in Scheme 1 are complex and the products
proposed following the potential isomerisation reaction are
hypothetical and no standards of such unstable compounds
can be synthesised, the high yield of isobutyl acetate (ca. 28%
of isobutyl isopropyl ether removed) indicates that the major
pathway of oxy radical 2 under atmospheric conditions is
decomposition via CwC bond scission. Assuming that both
the alkyl substituents in an ether molecule react independently
towards OH attack, the contribution of the isopropyl side
chain (q) to the total reactivity of isobutyl isopropyl ether is
given by the expression:
3 2
3 2
0.5 ] k(OH ] diisopropyl ether) ] 100
breaking. As a further possibly important reaction, oxy radical
2 may undergo isomerisation, via a six-membered ring tran-
sition state, involving the relatively weak tertiary CwH atom
to produce the organic radical 6. The resulting tertiary oxy
radical 7 is thought29 to decompose mainly to form acetone
and the organic radical 8, leading to the complex alkoxy
radical 9. By analogy with oxy radicals derived from methyl
ethers,31,32,33 [CH OCH (O~) or (CH ) COCH (O~)], radical
q \
B 27% (VI)
k(OH ] isobutyl isopropyl ether)
This percentage is close to the observed yield of isobutyl
acetate, implying that this compound is the exclusive product
from the reactions following H abstraction from the tertiary
CwH bond in the isopropyl group of the ether.
Scheme 2 shows the proposed reactions following H-atom
3
2
3 3
2
9 is proposed to mainly undergo reaction with oxygen or
CwH bond Ðssion leading to the organic formate 10. This
molecule is a hemiacylal which would certainly be unstable
and might be expected to form acetone and formic acid, at
least under the present analytical conditions. No formic acid
abstraction from the CH group of the isobutyl side chain.
2
The intermediate organic oxy radical 12 decomposes mainly
via CwC bond breakage to form isopropyl formate and the
isopropyl radical 13, which can form isopropyl nitrate and the
2-propoxy radical, 14, leading to acetone by a reaction with
Scheme 2
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
2871

View Article Online
Scheme 3
oxygen.34 Alternatively, oxy radical 12 may decompose via
CwO bond rupture yielding isobutyraldehyde and a 2-
propoxy radical which again forms acetone. The relatively
small amounts of the isobutyraldehyde, which may addi-
tionally be formed by other reactions in the degradation of
this ether (Scheme 1), indicate that CwO bond rupture in oxy
radical 12 is a minor channel, which is in agreement with pre-
vious atmospheric product studies of ether type com-
pounds.4h6,26,31h33,36,37 This is substantiated by empirical
methods for estimating the rate coefficients of alkoxy radical
decomposition reactions,1,38 which relate estimated enthalpy
changes39,40 to the activation energies of such processes. Thus
the two potential pathways leading to the formation of iso-
butyraldehyde above, the decompositions of oxy radical 2
(Scheme 1) and oxy radical 12 (Scheme 2) via CwO bond
rupture are both estimated to be ca. 46 kJ molv1 endothermic,
whereas the enthalpy changes of the decomposition reactions
of these radicals by CwC bond rupture are estimated to be ca.
[23 and ]2.7 kJ mol~1 respectively. Thus, on the basis of
these estimated enthalpy changes, the CwO bond Ðssions
should be insigniÐcant relative to the alternative CwC bond
Ðssions in both oxy radicals. It is consequently unclear which
pathway leads to the observed production of iso-
butyraldehyde. The reaction of the organic oxy radical 12 with
oxygen, leading to a small amount of isopropyl isobutyrate, is
signiÐcantly slower than the decomposition of this radical via
CwC bond breakage, under the experimental conditions. The
relatively high yields of isopropyl nitrate may indicate that
this compound was produced not only by the reaction of the
isopropyl peroxy radical with nitric oxide as shown in Scheme
2, but also by the reaction of the isopropoxy radical with
nitrogen dioxide:
(CH ) CHO~ ] NO ] M ] (CH ) CHONO ] M (6)
3 2
2
3 2
2
This reaction is believed to be insigniÐcant at typical tropo-
spheric NO levels, in relation to the competing reaction of
2
the isopropoxy radical with oxygen:1
(CH ) CHO~ ] O ] (CH ) CxO ] HO
(7)
3 2
2
3 2
2
Under the present experimental conditions, the NO mixing
ratios built up from values near 0 ppm to ca. 10 ppm at the
2
end of an experiment. The yields of isopropyl nitrate by this
pathway were simulated for the conditions of experiments no.
10È12, based on the known rate coefficients of reactions (6)
and (7),34,35 the given O and NO concentrations and esti-
2
2
mated isopropoxy radical yields from the overall reaction (40È
60%). These calculated yields of isopropyl nitrate from
reaction (6) corresponded roughly to half of the observed
build up of isopropyl nitrate under the experimental condi-
tions.
Scheme 3 shows the proposed reactions following a H-atom
abstraction from the tertiary CwH bond of the isobutyl side
2872
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Scheme 4
chain. The initial carbon-centred radical adds oxygen to form
a peroxy radical, which further reacts with nitric oxide. The
latter reaction leads partly to the formation of 1,1,4-trimethyl-
3-oxapentyl nitrate, but the reaction produces mainly the
organic oxy radical 15. This oxy radical is expected to
undergo isomerisation via 1,5-H-atom shift by abstracting a H
atom from the tertiary CwH centre in the isopropyl group of
the ether. The resulting organic radical 16 is, by analogy with
radical 1 (see Scheme 1), most likely to lead to the formation
of 2-hydroxy-2-methylpropyl acetate via breakage of the
CwC bond in the intermediate organic oxy radical 17. As
shown in Scheme 3, oxy radical 15 may additionally decom-
pose to produce acetone and radical 18. This radical leads to
oxy radical 19 which, in line with the reactions31,32,33 of the
oxy radicals CH OCH (O~) or (CH ) COCH (O~), will react
lish the formation of 2-hydroxy-2-methylpropyl acetate by the
above mechanism were carried out in qualitative experiments.
Hydroxyl-radical-initiated reactions of the Ðrst generation
product isobutyl acetate, in the presence of typical concentra-
tions of formaldehyde found in the present experiments, did
not produce the acetate, which could possibly have arisen
from the bimolecular reaction32 of the long-lived tertiary oxy
radical CH C(O)CH C(O~)(CH ) with the labile CwH bonds
in formaldehyde. However, the photolysis of 1,1,4-trimethyl-3-
oxapentyl nitrite in air, which leads7 to the initial formation of
oxy radical 15, produced 2-hydroxy-2-methylpropyl acetate as
a stable end product. The latter experiment was carried out, in
the absence of a pure standard of 1,1,4-trimethyl-3-oxapentyl
nitrite, with a gas mixture containing several ppm of this com-
pound together with 2-hydroxy-1-isopropoxy-2-methyl-
propane (see Materials section) and 10 ppm nitric oxide. In
this qualitative experiment two further products were
observed and identiÐed as acetone and isopropyl formate,
believed to come from the decomposition of oxy radical 15.
The abstraction of H atoms from the four methyl groups in
isobutyl isopropyl ether are expected to be slow processes.
From the structure activity relationship (SAR) of Kwok and
3
2
3 2
3
2
3 3
2
predominantly with oxygen or by CwH bond Ðssion to form
isopropyl formate. The alternative decomposition reaction
would lead to 1-isopropoxypropan-2-one and
a methyl
radical. A standard of 1-isopropoxypropan-2-one was not
available but a reference mass spectrum of this compound is
reported in the NIST mass spectra database and no equiva-
lent mass spectrum was found in our analyses. Tests to estab-
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
2873

View Article Online
Atkinson15 the contribution of the four methyl groups to the
total reactivity of isobutyl isopropyl ether towards OH attack
can be calculated to be ca. 2% and consequently only minor
products were expected from these channels. Scheme 4 shows
the possible reaction sequences after H-atom abstraction from
a methyl group of the isobutyl side chain. The intermediate
primary oxy radical 21 may break down to form formalde-
hyde and the organic radical 22. This radical may be con-
verted to the oxy radical 23 via oxygen addition and reaction
with NO. Oxy radical 23 may isomerise via 1,5-H-atom shift
to form radical 24 and subsequently oxy radical 25, which, by
analogy with oxy radical 2 (see Scheme 1), is expected to
produce mainly 1-acetoxy-2-hydroxypropane via decomposi-
tion. A reference mass spectrum of this compound is available
in the NIST mass spectra database, but no equivalent mass
spectrum was observed in our experiments. Oxy radical 23
may alternatively react with oxygen to produce 1-
isopropoxypropan-2-one, which was not observed by mass
spectral analysis as stated above. Oxy radical 23 could also
undergo two decomposition reactions as shown in Scheme 4.
The decomposition product acetaldehyde was not observed by
analysis of the gas mixtures via derivatisation with 2,4-DNPH
followed by HPLC separation and comparison with a stan-
dard of the 2,4-DNPH hydrazone derivative of acetaldehyde.
It seems unlikely that oxy radical 21 will isomerise via 1,4- or
1,6-H-atom shifts.
formaldehyde and the organic radical 28 which is thought to
be the source of the observed minor product isobutyl formate.
Overall the proposed mechanism gives a reasonably quanti-
tative account of the observed product distribution. The
absence of major unknown peaks in our analyses and the fact
that the observed products, in conjunction with the proposed
mechanism, account for ca. 115 ^ 11% of the isobutyl isopro-
pyl ether removed, indicates that all the main routes are
accounted for in the degradation of this ether. Furthermore,
the proposed reaction mechanism for the formation of the
major products is based entirely on established gas-phase
reactions. Clearly the relatively high errors in the reported
product yields do not rule out the possibility of additional
minor pathways. One such additional source of error seems
likely to involve the loss of the low vapour pressure products
on the TeÑon walls and on aerosols formed in the system.
These loss processes were investigated by irradiating a typical
bag mixture and following the decay of the concentrations of
the products in the dark over a period of over 2 h. The
observed trends in the analyses during this period were small.
For the most critical compound, 2-hydroxy-2-methylpropyl
acetate, a loss rate of ca. 3% h~1 was observed and conse-
quently no corrections were applied for this e†ect, which
introduces a systematic error of \3%. Under the irradiation
conditions of the experiments, only the product 1,1,4-
trimethyl-3-oxapentyl nitrate revealed a measurable decay due
to photolysis [k \ (2.6 ^ 0.6) ] 10~5 s~1] which was accom-
panied by the build-up of an unidentiÐed product. No correc-
tions for this loss process were applied.
In Scheme 5, the OH attack occurs on a methyl group of
the isopropyl side chain. The resulting organic oxy radical 27
may (i) react with oxygen to form the aldehyde 30, (ii) isomer-
ise via a 1,5-H-atom shift to produce in the subsequent reac-
tions mainly the respective formate or (iii) decompose to form
One of the main objectives of the present study was to
investigate isomerisation reactions in organic oxy radicals
Scheme 5
2874
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
8
9
A. P. Black and F. H. Barbers, Org. Synth., 1943, coll. vol. 2, 412.
D. W. Stocker, PhD thesis, University of Birmingham, Birming-
ham, UK, 1985.
formed from ethers. Isopropyl isobutyl ether was chosen as a
model compound for this product study for two reasons.
Firstly, because the two tertiary CwH groups in this molecule
will result in the formation of two structurally di†erent types
10 G. A. Olah, S. C. Narang, R. L. Pearson and C. A. Cupas, Syn-
thesis, 1978, 6, 452.
of oxy radicals; type
I [(CH ) C(O~)OCH CH(CH ) ],
11 W. A. Noyes, Org. Synth., 1943, coll. vol. 2, 108.
12 J. A. Kerr and D. W. Stocker, J. Atmos. Chem., 1986, 4, 253.
13 D. F. Smith, T. E. Kleindienst and E. E. Hudgens, J. Chromato-
gr., 1989, 483, 431.
14 R. Atkinson, S. M. Aschmann, W. P. L. Carter, A. M. Winer and
J. N. Pitts Jr., J. Phys. Chem., 1982, 86, 4563.
15 E. C. Kwok and R. Atkinson, Atmos. Environ., 1995, 29, 1685.
16 R. J. Cvetanovic, D. L. Singleton and G. Paraskevopoulos, J.
Phys. Chem., 1979, 83, 51.
17 J. A. Kerr and D. W. Sheppard, Environ. Sci. T echnol., 1981, 15,
960.
3 2
2
3 2
derived from the isopropyl side chain, has an alkoxy substit-
uent in the a-position relative to the radical centre, whereas
type II [(CH ) HCOCH C(O~)(CH ) ], from the isobutyl side
3 2
2
3 2
chain, does not. Secondly, the two relatively labile tertiary
CwH bonds in the ether are in positions which potentially
facilitate the rapid isomerisation reactions of the resulting oxy
radicals, via 1,5-H-atom transfers. In general, the OH-initiated
oxidation of ethers is believed15 to generate mainly oxy rad-
icals of type I, although we have recently shown6 that oxy
radicals of type II can play an important role in the atmo-
18 D. H. Semmes, A. R. Ravishankara, C. A. Gump-Perkins and
P. H. Wine, Int. J. Chem. Kinet., 1985, 17, 303.
spheric oxidation of ethers with longer ([C ) alkyl side
19 S. Dobe, L. A. Khachatryan and T. Berces, Ber. Bunsen-Ges.
Phys. Chem., 1989, 96, 3312.
3
chains. The present study has conÐrmed that organic oxy
radical types I and II react di†erently under atmospheric con-
ditions. None of the observed products indicated any isomer-
isation reaction of the oxy radical of type I; thus, on the basis
of the high yields of isobutyl acetate, oxy radical 2 undergoes
mainly unimolecular decomposition via rupture of a CwC
bond. Conversely the observed major product, 2-hydroxy-2-
methylpropyl acetate, indicates that for oxy radicals of type II
(oxy radical 15 in Scheme 3) the 1,5-H shift is an important
reaction under atmospheric conditions. The rate coefficient of
this isomerisation reaction cannot be calculated from the
present data, since the competitively produced decomposition
products of oxy radical 15, acetone and isopropyl formate,
also arise from the decay of oxy radical 12 (see Scheme 2). A
recent product study of the OH-initiated photo-oxidation of
2-butoxyethanol6 came to a similar conclusion concerning the
relative importance of isomerisation reactions and decomposi-
tions reactions for oxy radicals with the structures
20 G. J. Audley, D. L. Baulch and I. M. Campell, J. Chem. Soc.,
Faraday T rans. 1, 1981, 77, 2541.
21 R. Atkinson, J. Phys. Chem. Ref. Data, 1989, monograph no. 1.
22 S. Teton, A. Mellouki, G. Le Bras and H. Sidebottom, Int. J.
Chem. Kinet., 1996, 28, 291.
23 T. J. Wallington, P. Dagaut, R. Liu and M. J. Kurylo, Int. J.
Chem. Kinet., 1988, 20, 541.
24 P. McLoughlin, R. Kane and I. Shanahan, Int. J. Chem. Kinet.,
1993, 25, 137.
25 A. Mellouki, S. Teton and G. Le Bras, Int. J. Chem. Kinet., 1995,
27, 791.
26 T. J. Wallington, J. M. Andino, A. R. Potts, S. J. Rudy, W. O.
Siegl, Z. Zhang, M. J. Kurylo and R. E. Huie, Environ. Sci.
T echnol., 1993, 27, 98.
27 L. Nelson, O. Rattigan, R. Neavyn, H. Sidebottom, J. Treacy and
O. J. Nielsen, Int. J. Chem. Kinet., 1990, 22, 1111.
28 P. J. Bennett and J. A. Kerr, J. Atmos. Chem., 1989, 8, 87.
29 A. P. Altshuller, J. Atmos. Chem., 1991, 12, 19.
30 W. P. L. Carter and R. Atkinson, J. Atmos. Chem., 1985, 3, 377.
31 S. M. Japar, T. J. Wallington, J. F. O. Richert and J. C. Ball, Int.
J. Chem. Kinet., 1990, 22, 1257.
RCH(O~)OR@ and RCH(O~)CH OR@, respectively.
2
32 D. F. Smith, T. E. Kleindienst, E. E. Hudgens, C. D. McIver and
J. J. Bufalini, Int. J. Chem. Kinet., 1991, 23, 907.
We thank the Schweizerische Nationalfonds zur Forderung
der wissenschaftlichen Forschung for Ðnancial support and
BASF AG (Schweiz) for supplying a sample of 1,2-epoxy-2-
methylpropane. We also thank Dr. Marc Suter (EAWAG) and
Dr. Roland Hany (EMPA) for carrying out the high
resolution mass spectrometric and NMR measurements,
respectively.
33 E. C. Tuazon, W. P. L. Carter, S. M. Aschmann and R. Atkinson,
Int. J. Chem. Kinet., 1991, 23, 1003.
34 R. J. Balla, H. H. Nelson and J. R. McDonald, Chem. Phys., 1985,
99, 323.
35 L. Batt and R. T. Milne, Int. J. Chem. Kinet., 1977, 9, 141.
36 T. J. Wallington and S. M. Japar, Environ. Sci. T echnol., 1991, 25,
410.
37 D. F. Smith, T. E. Kleindienst, E. E. Hudgens, C. D. McIver and
J. J. Bufalini, Int. J. Chem. Kinet., 1992, 24, 199.
References
38 R. Atkinson and W. P. L. Carter, J. Atmos. Chem., 1991, 13, 195.
39 S. W. Benson, T hermochemical Kinetics, Wiley, New York, 1976.
40 National Institute of Standards Reference Database 25, Struc-
tures and Properties Database and Estimation Program, Version
1.2, S. E. Stein, J. M. Rukkers and R. L. Brown, Chemical
Kinetics Division, NIST, Gaithersburg, MD, USA, 1991.
1
2
3
4
R. Atkinson, J. Phys. Chem. Ref. Data, 1994, monograph no. 2.
R. Atkinson, J. Phys. Chem. Ref. Data, 1997, 26, 215.
R. Atkinson, Int. J. Chem. Kinet., 1997, 29, 99.
J. Eberhard, C. Muller, D. W. Stocker and J. A. Kerr, Int. J.
Chem. Kinet., 1993, 25, 639.
5
6
7
K. Stemmler, W. Mengon and J. A. Kerr, Environ. Sci. T echnol.,
1996, 30, 3385.
K. Stemmler, W. Mengon, D. J. Kinnison and J. A. Kerr, Environ.
Sci. T echnol., 1997, 31, 1496.
J. Eberhard, C. Muller, D. W. Stocker and J. A. Kerr, Environ.
Sci. T echnol., 1995, 29, 232.
Paper 7/01766I; Received 13th March, 1997
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
2875