Photodecomposition of iodopentanes in air: Product distributions from the self-reactions of n-pentyl peroxyl radicals

Article abstract of DOI:10.1002/kin.10031

Product distributions from the 254-nm photooxidation of the three iodopentane isomers were explored as a technique for studying the self-reactions of individual pentyl peroxyl radicals (in air at ambient temperature and pressure). Pentanols and the associated carbonyl compounds (pentanal or pentanones) were major products as expected. Other major products resulted from the isomerization of pentan-1-oxyl and pentan-2-oxyl radicals, but their nature could not be identified. Minor products were alcohols and carbonyl compounds arising from the decomposition of pentoxyl radicals. Diols and mixed hydroxycarbonyl compounds from cross-combination reactions were essentially absent, in contrast to expectation. The observed product distributions were evaluated to derive branching ratios for the radical-preserving pathways of the self-reactions, 0.42 ± 0.17, 0.46 ± 0.10, 0.39 ± 0.08, for pentan-1-yl peroxyl, pentan-2-yl peroxyl, and pentan-3-yl peroxyl, respectively. Rate coefficients derived for the decomposition of the corresponding pentoxyl radicals, relative to their reaction with oxygen, are (5.1 ± 0.5) × 10¹⁸, (1.0 ± 0.2) × 10¹⁸, and (3.2 ± 0.3) × 10¹⁸ molecule cm^-3, respectively. Rate constants for the isomerization of pentan-1-oxyl and pentan-2-oxyl were estimated from the contributions of isomerization products to the total amounts of products as (4.0 ± 1.1) × 10⁵ s^-1 and (1.0 ± 2.0) × 10⁵ s^-1, respectively.

Full text of DOI:10.1002/kin.10031

Photodecomposition
of Iodopentanes in Air:
Product Distributions
from the Self-Reactions
of n-Pentyl Peroxyl Radicals
1
2
1
GERALD HEIMANN,1,3 HEINZ-JURGEN BENKELBERG, OLAF BOGE, PETER WARNECK
¨
¨
1
¨
Max-Planck-Institut fur Chemie, 55020 Mainz, Germany
2
3
¨
¨
Institut fur Tropospharenforschung, 04303 Leipzig, Germany
Verein Deutscher Ingenieure, 40239 Dusseldorf, Germany
¨
Received 5 April 2001; accepted 8 October 2001
ABSTRACT: Product distributions from the 254-nm photooxidation of the three iodopentane
isomers were explored as a technique for studying the self-reactions of individual pentyl per-
oxyl radicals (in air at ambient temperature and pressure). Pentanols and the associated car-
bonyl compounds (pentanal or pentanones) were major products as expected. Other major
products resulted from the isomerization of pentan-1-oxyl and pentan-2-oxyl radicals, but
their nature could not be identified. Minor products were alcohols and carbonyl compounds
arising from the decomposition of pentoxyl radicals. Diols and mixed hydroxycarbonyl com-
pounds from cross-combination reactions were essentially absent, in contrast to expectation.
The observed product distributions were evaluated to derive branching ratios for the radical-
preserving pathways of the self-reactions, 0.42 0.17, 0.46 0.10, 0.39 0.08, for pentan-1-yl
peroxyl, pentan-2-yl peroxyl, and pentan-3-yl peroxyl, respectively. Rate coefficients derived
for the decomposition of the corresponding pentoxyl radicals, relative to their reaction with
oxygen, are (5.1 0.5) 1018, (1.0 0.2) 1018, and (3.2 0.3) 1018 molecule cm 3, re-
spectively. Rate constants for the isomerization of pentan-1-oxyl and pentan-2-oxyl were es-
timated from the contributions of isomerization products to the total amounts of products as
(4.0 1.1) 105 s 1 and (1.0 2.0) 105 s 1, respectively. 2001 John Wiley & Sons, Inc. Int
C
J Chem Kinet 34: 126–138, 2002
oxidation of hydrocarbons in general, and of alkanes in
particular [1]. In the atmosphere, the oxidation occurs
primarily in the gas phase, and that of alkanes is initi-
ated by reaction with OH or NO3 radicals. Mechanisms
of alkane oxidation under atmospheric conditions have
been extensively reviewed [2–4], and general reaction
schemes have been established. Alkyl peroxyl radicals
are formed by the addition of O2 to the site where the
OH (or NO3) radical had removed a hydrogen atom.
The predominant fate ofRO2 radicalsinthe atmosphere
INTRODUCTION
Alkyl peroxyl radicals, RO2, and alkoxyl radicals,
RO, are essential intermediates in the low temperature
Correspondence to: Peter Warneck; e-mail: biogeo@mpch-
mainz.mpg.de.
Contract grant sponsor: German Federal Ministry for Research
and Technology.
c
2001 John Wiley & Sons, Inc.
DOI 10.1002/kin.10031

PHOTODECOMPOSITION OF IODOPENTANES IN AIR
127
is reaction with NO.
photolysis of iodopentanes in air as a means of generat-
ing individual pentylperoxyl radicals. Iodoalkanes ex-
cited to the first higher electronic state are well known
to undergo predissociation. The major products are an
excited iodine atom and the corresponding alkyl rad-
ical [10]. In the case of secondary iodoalkanes one
observes also the elimination of HI with alkenes as
additional products. This process is not expected to
complicate the system. The primary aim of our study
was to determine from the product distributions the
branching ratios of reaction (IV) for the self-reactions
of the three pentylperoxyl isomers and to learn more
about the products resulting from the decomposition
and isomerization processes of the associated pentoxyl
radicals.
The reaction mechanisms induced by the self-
reactions of individual pentylperoxyl radicals are still
complex because additional alkylperoxyl radicals are
produced by the decomposition and isomerization pro-
cesses. Figure 1 shows, as an example, the basic mech-
anism expected to follow from the self-reaction of
pentan-2-yl peroxyl. Pentan-2-one and pentan-2-ol are
the stable products from reaction (IV). Decomposition
of pentan-2-oxyl yields propylperoxyl and 1,5 H-atom
shift yields 4-hydroxypentan-1-yl peroxyl, respec-
tively, as secondary peroxyl radicals. Isomerization
ultimately leads to the formation of 1-hydroxypentan-
4-one, 4-hydroxypentanal, and 1,4-pentandiol as sta-
ble products. Most of these products should be iden-
tifiable by gas chromatography. The relative yields
are expected to indicate the extent of alkoxyl radi-
cal decomposition and isomerization. Thus, another
aim of this study was to confirm and quantify the ex-
pected decomposition and isomerization products and
compare the results with other data [11–13] reported
previously.
OH + HR → R + H2O
R + O2 → RO2
RO2 + NO → RO + NO2 (major) (IIIa)
→ RONO2 (minor) (IIIb)
(I)
(II)
This reaction sequence converts most of the alkyl
peroxyl radicals to alkoxyl radicals. Under laboratory
conditions, in the absence of NO, the RO2 radicals react
mainly with each other
RO2 + RO2 → 2RO + O2
→ alcohol + aldehyde/ketone + O2
(IVb)
(IVa)
where alkoxyl radicals are again produced in the
first channel and stable products in the second [5,6].
Alkoxyl radicals react with oxygen, and they undergo
decomposition and isomerization processes.
RO + O2 → HO2 + aldehyde/ketone
(Va)
RO → R0 + aldehyde (decomposition) (Vb)
RO → HOR00 (isomerization)
(Vc)
Here, R is the parent alkyl radical, R0 is an alkyl
radical of lower carbon number, and HOR00 is a hy-
droxyalkyl radical resulting from internal hydrogen ab-
straction by the alkoxyl group. The HOR00 radical sub-
sequently attaches O2 to form a hydroxyperoxyl rad-
ical, HOR00O2, that undergoes further reactions. Iso-
merization by internal H-atom abstraction is less well
understood than the other processes, but the ring strain
involved suggests that isomerization via 1,5 H-atom
transfer is structurally and energetically favored over
1,4- or 1,6-H-atom shift [7,8]. Butane is the first mem-
ber of n-alkane homologues allowing 1,5 H-atom shift
to take place.
Details of the oxidation mechanisms for C1–C4 hy-
drocarbons are well established [4–6]. The oxidation
of higher alkanes involves an increasing number of
isomeric RO2 radicals that can complicate the oxida-
tion pathways. The oxidation of n-pentane is already
so complex that it has not yet been fully elucidated.
Three types of RO2 radicals are involved: pentan-1-yl
peroxyl, pentan-2-yl peroxyl, and pentan-3-yl peroxyl.
The latter two are formed with greater probability than
the first because the abstraction of a secondary H-atom
from an alkane is energetically favored over that of
a primary H-atom [4,9]. In this study, we examine the
EXPERIMENTAL
Apparatus and experimental techniques were similar to
those described earlier [14,15]. Reactions were carried
out in 2-l spherical glass bulbs. The interior walls were
treated with dimethyldichlorosilane followed by heat-
ing to 200 C to minimize the surface activity toward
polar compounds. Each bulb was fitted with Teflon-
stoppered shut-off valves, a Teflon-coated silicone rub-
ber septum, and a quartz finger reaching into the cen-
ter of the bulb, into which a Penray mercury lamp was
placed. Specialquartzwasusedtoblockozone-forming
radiation ( < 235 nm), whereas emissions at 254 nm
and at longer wavelengths were fully transmitted. Com-
pressed nitrogen served as a coolant of the lamp as well
as to flush out ozone formed in the quartz tube. A bulb
was filled with a mixture of 0.01–0.1% iodopentane in

128
HEIMANN ET AL.
Figure 1 Scheme of reactions resulting from the self-reaction of pent-2-yl peroxyl radicals. Reactions following decomposition
and 1,5 H-atom shift isomerization of pent-2-oxyl radical were discussed by Atkinson [4], 1,4 H-atom shift, which should be
less important, is not included.
synthetic air, usually to a pressure slightly above at-
mospheric. A thin Teflon tube was pushed through a
hole pierced in the rubber septum toward the center of
the bulb, the other end was joined to the sampling loop
of a gas chromatograph. The inlet tubes were made of
quartz-lined stainless steel, and the sampling valve was
kept at 160 C. The Penray lamp was activated for about
10 min to stabilize its output before the lit lamp was
inserted in the quartz finger. Irradiation times were 30–
120 s. Subsequently, a sample was transferred from the
reaction vessel to the sampling loop of the gas chro-
matograph and was injected into the nitrogen carrier
gas flow (3 cm3 min 1). The products were separated
on a capillary column and detected with a flame ion-
ization detector. This procedure was repeated at least
three times in order to check the stability and precision
of the analysis.
65 C at a rate of 8 C min 1, further heating to 200 C
at a rate of 30 C min 1, and finally constant for 3 min
at 200 C. The temperature program used with the CP-
WAX 52 column was 40 C isothermal for 2 min, heat-
ing to 160 C at a rate of 20 C min 1, constant for
10 min at 160 C, further heating to 210 C at a rate
of 30 C min 1, and finally constant for 2 min at
210 C. Although temperature programs were chosen
so as to achieve an optimal peak separation, a full
separation of all products was not always possible.
Products were identified, as far as possible, by com-
parison of retention times with authentic samples of
alcohols, aldehydes, and ketones. Table I provides a
reference list of retention times on the CP-WAX 52 col-
umn for several compounds that were either expected
as products or served as markers for product identi-
fication. Peak area calibration of product compounds
was performed with authentic samples in air prepared
by successive dilution of starting mixtures of known
composition.
Two columns (50 m long, 0.32 mm i.d.) were
mainly used: a CPSil 76 column, coated with
dimethylpolysiloxane (0.34 ␮m film thickness), and
a CP-Wax 52 column, coated with polyethylene glycol
(1.2 ␮m film thickness). The temperature program in
the first case was 30 C isothermal for 3 min, heating to
Two of the iodopentanes, namely 1-iodopentane
and 3-iodopentane were commercially available. Both
contained unknown but photochemically inactive

PHOTODECOMPOSITION OF IODOPENTANES IN AIR
129
Table I Approximate Retention Times (min) of Several Compounds on the CP-WAX 52 Gas Chromatographic
Columna
Compound
Retention Time (min)
Compound
Retention Time (min)
Acetaldehyde
Propanal
Butanal
Ethanol
Pentanal
Pentane-2-one
Pentane-3-one
Propanol
Pentan-3-ol
Pentan-2-ol
Butanol
1.75
2.45
3.35
3.90
4.50
4.50
4.50
5.10
5.80
5.90
6.00
7.00
Pentanol
1-Hydroxypropanone
Glycolaldehyde
2-Hydroxy-2-methylpentan-4-one
1-Hydroxy-3-methylbutan-2-one
1-Hydroxybutan-2-one
Methyl glyoxal
Pentan-2,4-diol
Butan-1,2-diol
1-Hydroxypentan-4-one
Pentan-1,2-diol
Pentan-1,4-diol
7.15
7.88
8.00
8.40
8.40
8.56
9.10
12.3
12.4
13.5
14.8
18.4
Pentan-2,4-dione
aFor operating conditions see Experimental section.
impurities. These compounds were used as received.
2-Iodopentane, which was not commercially available,
was synthesized following the procedure of Brown
and Wheeler [16]. Sodium iodide was reacted with
pentan-2-yl methane sulfonate, which was prepared
from pentan-2-ol and methane sulfonyl chloride in the
presence of pyridine according to Williams and Mosher
[17], followed by distillation from the extract. The pu-
rity obtained, determined by mass spectrometry, was
better than 95%. The main impurity presumably was
pent-2-ene resulting from a partial decomposition of
the methane sulfonate.
of butanol was estimated to be about 33% of that of
butanal. The yield of butanal relative to that of pen-
tanol was determined to be 0.102 0.008. As shown
in Fig. 2, a signal at the retention time of 1,4 dihydrox-
ypentane (18.4 min, compare Table I) was present, but
the associated product 1-hydroxypentan-4-one, which
should have eluted after about 13.5 min, was absent.
If 1-hydroxypentan-4-one is not formed, the identifi-
cation of the peak at 18.4 min as 1,4 dihydroxypen-
tane will also be in doubt. Partly, this peak may be
due to 1-hydroxypentan-4-hydroperoxide rather than
the corresponding diol. Two peaks at retention times
of about 8.1 and 8.6 min must be isomerization prod-
ucts, but their nature could not be identified. The cali-
bration for mixed hydroxycarbonyl compounds such as
1-hydroxypentan-4-one was used to obtain an estimate
RESULTS
1-Iodopentane
Table II presents a suggested reaction mechanism for
the photooxidation of 1-iodopentane. The major prod-
ucts expected are pentanal, pentanol, and mixed 1,4 hy-
droxycarbonyl compounds. Butanal and butanol are as-
sumed to result from the decomposition of the pentan-
1-oxyl radical, which leads to the formation of butyl
peroxyl radicals. Acetaldehyde would be a product
of the decomposition of the 1-hydroxypentan-4-oxyl
radical. Figure 2 shows a chromatogram of irradiated
1-iodopentane obtained with the CP-Wax-52 column.
Pentanal, pentanol, and butanal were identified as prod-
ucts. Butanol, which was expected to occur in con-
junction with butanal, was hidden underneath an im-
purity peak, but it showed up weakly on the CPSil 76
column. Figure 3a shows that the yield of pentanal
rises linearly with irradiation time and that the ratio
of the yields for pentanal and 1-pentanol is constant
with time. The average ratio (given by the slope of
the straight line in Fig. 3b) is 1.14 0.06. The yield
Figure 2 Chromatograms (CP-WAX 52 column) of 1-iodo-
pentane in air at atmospheric pressure before and after irra-
diation (lower and upper traces). Identified product peaks
are accentuated: (1) butanal, (2) pentanal, (3) 1-iodopentane,
(4) pentanol, (5) pentan-1,4-diol.

130
HEIMANN ET AL.
Table II Reaction Scheme for the Photooxidation of 1-Iodopentanea
˙
C4H9CH2I + h␯ → C4H9 CH2 + I
kphot
I + I + M → I2 + M
˙
C4H9CH2 + O2 → C4H9CH2OO
rapid
kpp
(1a)
(1b)
(2a)
(2b)
(2c)
(3)
(4)
(5)
(5a)
(5b)
(5c)
2C4H9CH2OO → 2C4H9CH2O + O2
→ C4H9CH2OH + C4H9CHO + O2
(1
)kpp
C4H9CH2O + O2 → C4H9CHO + HO2
1kO
2
C4H9CH2O → C3H7 CH2 + HCHO
1kdec
1kiso
rapid
rapid
˙
˙
C4H9CH2O → HOCH2(CH2)2 CHCH3
˙
C3H7CH2 + O2 → C3H7CH2OO
˙
HOCH2(CH2)2 CHCH3 + O2 → HOCH2(CH2)2CH(OO )CH3
HOCH2(CH2)2CH(OO )CH3 + C4H9CH2OO
→ HOCH2(CH2)2CH(O )CH3 + C4H9CH2O + O2
→ HOCH2(CH2)2CH(OH)CH3 + C4H9CHO + O2
→ HOCH2(CH2)2COCH3 + C4H9CH2OH + O2
fksp
0.5(1 f )ksp
0.5(1 f )ksp
1
2
(6a)
(6b)
C3H7CH2OO + C4H9CH2OO → C3H7CH2O + C4H9CH2O + O2
kpp
1
4
1
→ C3H7CHO + C4H9CH2OH + O2
→ C3H7CH2OH + C4H9CHO + O2
kpp
4
(6c)
(7a)
(7b)
(9a)
(9b)
(9c)
(10)
(11)
(12)
kpp
2HOCH2(CH2)2CH(OO )CH3 → 2HOCH2(CH2)2CH(O )CH3 + O2
f kss
→ HOCH2(CH2)2CH(OH)CH3 + HOCH2(CH2)2COCH3 + O2
(1 f )kss
HOCH2(CH2)2CH(O )CH3 + O2 → HOCH2(CH2)2COCH3 + HO2
2kO
2
2kdec
2kiso
rapid
˙
HOCH2(CH2)2CH(O )CH3 → HOCH2CH2CH2 + CH3CHO
HOCH2(CH2)2CH(O )CH3 → OCH2(CH2)2CH(OH)CH3
˙
HOCH2CH2CH2 + O2 → HOCH2CH2CH2OO
OCH2(CH2)2CH(OH)CH3 + O2 → OCH(CH2)2CH(OH)CH3 + HO2
HOCH2(CH2)2CH(OO )CH3 + C3H7CH2OO
1kO
2
1
2
1
(12a)
(12b)
(12c)
(13a)
(13b)
(14)
→ HOCH2(CH2)2CH(O )CH3 + C3H7CH2O + O2
→ HOCH2(CH2)2CH(OH)CH3 + C3H7CHO + O2
→ HOCH2(CH2)2COCH3 + C3H7CH2OH + O2
2C3H7CH2OO → 2C3H7CH2O + O2
ksp
4
1
ksp
4
1
ksp
2
1
kpp
2
→ C3H7CH2OH + C3H7CHO + O2
C3H7CH2O + O2 → C3H7CHO + HO2
HOCH2CH2CH2OO + C4H9CH2OO
kpp
1kO
2
(15)
1
2
1
(15a)
(15b)
(15c)
(16)
→ HOCH2CH2CH2O + C4H9CH2O + O2
→ HOCH2CH2CH2OH + C4H9CHO + O2
→ HOCH2CH2CHO + C4H9CH2OH + O2
HOCH2CH2CH2OO + HOCH2(CH2)2CH(OO )CH3
kpp
4
1
kpp
4
kpp
1
2
1
(16a)
(16b)
(16c)
(17)
→ HOCH2CH2CH2O + HOCH2(CH2)2CH(O )CH3 + O2
→ HOCH2CH2CH2OH + HOCH2(CH2)2COCH3 + O2
→ HOCH2CH2CHO + HOCH2(CH2)2CH(OH)CH3 + O2
HOCH2CH2CH2OO + C3H7CH2OO
ksp
4
1
ksp
4
ksp
1
2
1
(17a)
(17b)
(17c)
(18a)
(18b)
(19)
(20a)
(20b)
(20c)
(20d)
(21)
→ HOCH2CH2CH2O + C3H7CH2O + O2
→ HOCH2CH2CH2OH + C3H7CHO + O2
→ HOCH2CH2CHO + C3H7CH2OH + O2
2HOCH2CH2CH2OO → 2HOCH2CH2CH2O + O2
kpp
4
1
kpp
4
1
kpp
2
1
kpp
2
→ HOCH2CH2CH2OH + HOCH2CH2CHO + O2
HOCH2CH2CH2O + O2 → HOCH2CH2CHO + HO2
C4H9CH2OO + HO2 → C4H9CH2OOH + O2
kpp
1kO
2
kHO
2
OHCH2(CH2)2CH(OO )CH3 + HO2 → OHCH2(CH2)2CH(OOH)CH3 + O2
C3H7CH2OO + HO2 → C3H7CH2OOH + O2
kHO
2
kHO
2
HOCH2CH2CH2OO + HO2 → HOCH2CH2CH2OOH + O2
kHO
2
HO2 + HO2 → H2O2 + O2
kH O
2
2
aRate constants used in the calculations: photolysis rate (s 1), kphot 1.3 10 3; bimolecular rate coefficients (cm3 molecule 1 s 1) kpp
=
=
14
15
1
2
1
10 13, kss = 2 10 15, ksp = 2 10
,
1kO = 9.5 10
,
2kO = 8 10 15; Decomposition constants (s
)
1kdec = 4.8 104,2kdec
2
2
104, Isomerisation constants (s 1) 1kiso = 4.0 105, 2kiso = 9 106; branching ratios: = 0.42, f = 0.5–0.8.

PHOTODECOMPOSITION OF IODOPENTANES IN AIR
131
Figure 3 Left: Rise of pentanal with time in the irradiation of 1-iodopentane in air at near atmospheric pressure (open points
for a mixing ratio of 200 ppm, solid points for a mixing ratio of 0.1% scaled by a factor of five). Right: Yield of pentanol versus
that of pentanal (the solid line shows the ratio 0.88 0.05).
of the yields. Acetaldehyde was not observed, so that
decomposition of the 1-hydroxypentan-4-oxyl radical,
if it were involved, must be unimportant compared to
competing reactions.
Table III summarizes relative yields for the observed
products. The data can be used to estimate the branch-
ing ratio = k1a/(k1a + k1b) for the self-reaction of
pentan-1-ylperoxylradicals(seeTableIIfortheassign-
ment of rate coefficients), and values for the decom-
position and isomerization constants for the pentan-1-
oxyl radical. The first quantity is given by
correspondingly
iso/kO [O2] =
[isom.products]
k
2
[pentanal] [pentanol]
= 8.1 2.2
If, as recommended by Atkinson [4], the rate coef-
ficient for the reaction of pentan-1-oxyl radicals with
oxygen is 9.5 10 15 cm3 molecule
s
1, values for
1
the decomposition and isomerization constants will be
kdec (4.8 0.5) 104 and kiso (4.0 1.1) 105
s
1, respectively. Here, the uncertainty of kO is not
2
included in the error limits.
[pentanal] [pentanol] + [other products]
The FACSIMILE computer program [18] was used
to calculate relative product yields resulting from the
reaction mechanism presented in Table II. For this pur-
=
[pentanal] + [pentanol] + [other products]
= 0.42 0.17
pose, the above values for , kO , kdec, and kiso were
2
employed. Rate coefficients for reactions of the per-
oxyl radicals involved are largely unavailable. The ex-
isting data [5,6] show, however, that rate coefficients
for the self-reactions of primary alkyl peroxyl radi-
The relatively large statistical uncertainty of the
branching ratio is caused mainly by the uncertainty
in the yield of the isomerization products. The rate of
decomposition of the pentan-1-oxyl radical relative to
its reaction with oxygen is given by
13
1
cals, kpp, are of the order of 10
cm3 molecule 1 s
,
whereas rate coefficients for the self-reactions of
secondary alkylperoxyl radicals, kss, are of the or-
der of 10 15, cm3 molecule 1 s 1. Rate constants of
cross-combination reactions, ksp, have intermediate
values approximated by the root over the product
of the individual rate coefficients [19]. Accordingly
we have used kpp = 2 10 13, kss = 2 10 15, and
ksp = 2 10 14 cm3 molecule 1 s 1. Branching ratios
[butanal] + [butanol]
k
dec/kO [O2] =
2
[pentanal] [pentanol]
= 0.97 0.10
whereas the rate of isomerization of the pentan-1-
oxyl radical relative to its reaction with oxygen is

132
HEIMANN ET AL.
Table III Photooxidation of 1-Iodopentane: Observed
and Calculated Relative Product Distribution
(b) f = 0.8, to allow comparison with the experimental
data. In case (a) 1,4-pentandiol and 1-hydroxypentan-
4-one should be produced in amounts sufficient for ob-
servation. Since they were not observed, some of the
underlying assumptions must be wrong. In case (b),
which is based on f = 0.8, the production rate of 1,4-
pentandiol and 1-hydroxypentan-4-one is considerably
reduced. In both cases, 4-hydroxypentanal is an impor-
tant product. It is likely that the unidentified peak elut-
ing after 8.6 min is associated with this product, but we
were unable to prove or disprove it for the lack of an
authentic sample. In both cases, decomposition of the
1-hydroxypentan-4-oxyl radical is less important than
its isomerization, which may explain why acetaldehyde
was not observed.
Relative Yield
Calculateda
Product
Experimental Case (a) Case (b)
Pentanol
Pentanal
Butanal
Butanol
Pentan-1,4-diol
1-Hydroxypentan-
4-one
1.0 0.09
1.14 0.06
0.102 0.008
0.034 0.003
0.080 0.025
0.02
1.0
1.14
0.100
0.033
0.194
0.195
1.0
1.16
0.104
0.035
0.077
0.080
Acetaldehyde
Unidentified productsc
4-Hydroxypentanal
Propan-1,3-diol
3-Hydroxypropanal
Formaldehyde
Hydroperoxidesd
Hydroxy-
0.02
0.91 0.07
0.008
–
0.001
–
ndb
nd
nd
nd
nd
nd
0.375
0.002
0.006
0.141
0.152
0.439
0.559
0.003
0.009
0.151
0.214
0.580
2-Iodopentane
Figure 4 shows a chromatogram of the product spec-
trum obtained following irradiation of 2-iodopentane.
While pentan-2-one and pentan-2-ol are significant
products as expected, only minute amounts of 1,4-
diol were formed in contrast to expectation. The cor-
responding mixed hydroxycarbonyl compounds were
entirely absent. A prominent peak in the gas chro-
matogram eluting after 11.8 min indicated an isomer-
ization product. This product was found to be unstable
and nonlinear with irradiation time. Simultaneously, in
the range of retention times 8–10 min, a greater number
of not fully resolved peaks appeared that grew at the
expense of the peak at 11.8 min. The reaction thus pro-
duces at least one unstable substance that decomposes
towardmultiple products. Althoughnone of these prod-
uct could be identified, an estimate of the total yield was
made by summing gas chromatographic peak areas and
applyinganaverageresponsefactor. Thetotalyieldwas
hydroperoxidesd
a (a) and (b): assuming different values for the branching ratio of
hydroxypentylperoxy radical reactions, see Table II and text.
b nd = not determined.
c Occurring mainly in the peak at 8.51 min retention time (see
Fig. 2).
d Predominantly pentan-1-hydroperoxide and 1-hydroxypentan-
4-hydroperoxide.
for peroxyl radical reactions were largely taken to be
0.5 in favor of the radical-preserving pathway and 0.25
for each alcohol-producing channel in the case of cross-
combination reactions. The branching ratios for reac-
tions of the 1-hydroxypentan-4-peroxyl radical with
itself and with pentan-1-yl peroxyl radicals was varied
between f = 0.5 and f = 0.8. Rate coefficients for re-
actions of HO2 with alkyl- and hydroxyalkyl peroxyl
11
1
1
radicals were set to 1.5 10
cm3 molecule
s
in accordance with recommendations for other such
cases [5]. The rate constant for the decomposition
channel of the 1-hydroxy-pentan-4-oxyl radical was
set to 2kdec = 9 103 s 1, similar to the value for the
decomposition of the pentan-2-oxyl radical (see fur-
ther below). The second possible decomposition path-
way leading to a CH3 radical and 4-hydroxybutanal
should have a lower probability [8] and was ignored.
2
The rate constant kiso for isomerization of the 1-
hydroxypentan-4-oxyl radical, leading to the forma-
1
tion of 4-hydroxypentanal, was set to 9 106
s
,
Figure 4 Chromatograms (CP-WAX 52 column) of 2-
iodopentane in air at atmospheric pressure before and after
irradiation (lower and upper traces). Identified product peaks
are accentuated: (1) acetaldehyde, (2) propanal, (3) pentan-
2-one, (4) propanol, (5) pentan-2-ol, (6) pentan-1,4-diol.
following the recommendation of Atkinson [4]. Iso-
merisation and decomposition of the butoxyl radical
was neglected. The results of the calculations are in-
cluded in Table III for two cases, (a) f = 0.5 and

PHOTODECOMPOSITION OF IODOPENTANES IN AIR
133
proportional to irradiation time. In evaluating the data,
we assumed that the peaks appearing at intermediate
retention times are associated with products resulting
from the isomerization of pentan-2-oxyl radicals. In the
lower range of retention times, acetaldehyde, propanal,
and propanol were identified as products. These are
expected products arising from decomposition of the
pentan-2-oxyl radical. Table IV summarizes the rela-
tive product yields. The yield of acetaldehyde is higher
than that for the sum of propanal and propanol, indicat-
ing an additional source of acetaldehyde in this system.
The data were used to estimate the branching ratio
for the self-reaction of pentan-2-yl peroxyl radicals,
and values for the decomposition and isomerization
constants for the pentan-2-oxyl radical relative to that
with oxygen
decomposition and isomerization of the pentan-2-oxyl
1
radical are kdec = (8.4 1.7) 103
s
and kiso
=
(9.5 2.0) 104 s 1, respectively. Again, the uncer-
tainty of kO is not included in the error limits.
2
Computer calculations for the photooxidation of
2-iodopentane were carried out on the basis of the
mechanism shown in Fig. 1. The mechanism is sim-
ilar in many respects to that of peroxyl radicals re-
sulting from 1-iodopentane, so a detailed list of reac-
tions is not presented. Rate coefficients and branch-
ing ratios chosen for the interactions of primary and
secondary alkyl- and hydroxyalkyl peroxyl radicals
were largely identical to those used for modeling the
photooxidation of 1-iodopentane. For the self-reaction
of propylperoxyl radicals the known rate constant
[pentan-2-one] [pentan-2-ol] + [acetaldehyde] + [isom.products]
=
= 0.46 0.10
[pentan-2-one] + [pentan-2-ol] + [acetaldehyde] + [isom.products]
[propanal] + [propanol]
k
dec/kO [O2] =
= 0.20 0.04
2
[pentan-2-one] [pentan-2-ol]
[propanal] + [propanol]
k
iso/kO [O2] =
= 2.27 0.47
2
[pentan-2-one] [pentan-2-ol]
where the uncertainty indicates again the accumu-
lated statistical uncertainty of the measurements. As-
kpp = 3.9 10 13 cm3 molecule 1 s and branching
ratios [5] were used. The rate coefficient for isomer-
ization of the 4-hydroxypentan-1-oxyl radical was set
1
15
1
suming kO
8
10
cm3 molecule 1 s as rec-
2
ommended by Atkinson [4], the rate coefficients for
to kiso = 1.5 107 s 1, following the suggestion of
2
Atkinson [4]. The results of the calculations are in-
cluded in Table IV.
Table IV Photooxidation of 2-Iodopentane: Observed
and Calculated Relative Product Distribution
As in the case of 1-iodopentane, the calculations
predict diols and mixed hydroxycarbonyl compounds
to occur in higher yields than those that were actually
observed. It is again possible to reduce the calculated
yields by adjusting the branching ratios for the re-
actions of the 4-hydroxypentan-1-yl peroxyl radical
with itself and with pentan-2-yl peroxyl. These re-
sults will not be presented in view of our failure to
account for the many other products appearing in the
chromatograms.
Relative Yield
Calculateda
Product
Experimental Case (a) Case (b)
Pentan-2-ol
Pentan-2-one
Propanal
1.0 0.11
1.47 0.18
0.073 0.012
0.023 0.004
0.015 0.01
nd
0.16 0.02
1.07 0.22
0.02
1.0
1.0
1.51
0.073
0.024
0.27
0.28
0.10
–
1.49
0.072
0.024
0.24
0.24
0.16
–
Propanol
Pentan-1,4-diol
4-Hydroxypentanal
Acetaldehyde
Unidentified products
1-Hydroxypentan-
4-one
Hydroperoxidesc
Hydroxy-
hydroperoxidesc
Another problem is the insufficient yield of
acetaldehyde resulting from the calculation. If in
addition to 1,5 H-atom shift the pentan-2-oxyl radical
underwent 1,4 H-atom shift, decomposition of the re-
sulting 2-hydroxypentan-4-oxyl radical would provide
a natural route to acetaldehyde formation. Baldwin et
al. [7] estimated that the ring strain energy imposed
on 1,4 H-shift (involving a five-membered ring) is
0.54
0.46
ndb
nd
1.06
0.96
0.042
0.085
a(a) According to the mechanism shown in Fig. 1. (b) Including
additional reactions given in the text.
1
about 25 kJ mol greater than that on 1,5 H-shift
(proceeding via a six-membered ring), which is 2.5 kJ
mol 1. By combining the activation energy for internal
abstraction of a secondary hydrogen atom given by
bnd = not determined.
cPredominantly pentan-2-hydroperoxide and 4-hydroxypentan-
1-hydroperoxide.

134
HEIMANN ET AL.
Atkinson [4] and the different ring strain energies, one
The rate coefficients required to generate suffi-
1
2
estimates a rate constant of about 1 102 s for 1,4
cient amounts of acetaldehyde were kiso = 1.5
5
H-atom shift. This is greatly insufficient to generate
the excess acetaldehyde observed. Another possible
pathway for acetaldehyde formation follows from
isomerization of the 4-hydroxypentan-1-oxyl radical.
Atkinson [4] assumed that isomerization proceeds only
by abstraction of the hydrogen atom located on the car-
bon skeleton opposite the hydroxyl group, but transfer
of the H-atom from the hydroxyl group may also
occur, even though it would require a seven-membered
intermediate ring structure resulting in an increased
ring strain energy. We have explored the following
mechanism by means of computer simulation:
107 s 1, 4kiso = 5 106 s 1, kiso = 5 106 s 1, and
3kdec = 4 106 s 1. The results are shown in the last
column in Table IV. They indicate that apart from rais-
ing the yield of acetaldehyde the distribution of the
other products is not greatly affected. In order for this
mechanism to explain the formation of excess alde-
hyde, it will be necessary that hydrogen abstraction
from the OH-group is nearly as rapid as that of the
neighboring H-atom. The computer simulation sug-
gests a ratio of 1:3 for the rate coefficients involved.
3-Iodopentane
The 3-pentoxyl radical is not expected to undergo iso-
merization reactions, and this simplifies the oxidation
mechanism. Table V shows the corresponding reac-
tion scheme. The principal products were pentan-3-
one and pentan-3-ol, in accordance with expectation.
Additional products were propanal, acetaldehyde, and
small amounts of ethanol. These products result from
the decomposition of the pentan-3-oxyl radical. The
combined yields of acetaldehyde and ethanol are about
equal to that of propanal in accordance with the mecha-
nism. The CPSil 76 column was required for the quan-
tificationofpentan-3-ol, becauseinchromatogramsob-
tained with the CP-Wax-52 column the alcohol was
obscured by 3-iodopentane. The product distribution
obtained after 90-s irradiation of 3-iodopentane in air
at near atmospheric pressure is given in Table VI. A
number of experiments were made in which the mix-
ing ratio of oxygen was varied, keeping the total pres-
sure of the oxygen–nitrogen mixture nearly constant
CH3CHOH(CH2)2CH2O
˙
→ CH3C OH(CH2)2CH2OH
2kiso
4kiso
CH3CHOH(CH2)2CH2O
→ CH3CH(O )(CH2)2CH2OH
˙
CH3C OH(CH2)2CH2OH + O2
→ CH3CO(CH2)2CH2OH + HO2
CH3CH(O )(CH2)2CH2OH
→ CH3CHOH(CH2)2 C HOH
CH3CH(O )(CH2)2CH2OH
5kiso
3kdec
˙
˙
→ CH3CHO + HOCH2CH2CH2
˙
CH3CHOH(CH2)2C HOH + O2
→ CH3CHOH(CH2)2CHO + HO2
˙
HOCH2CH2CH2 + O2
→ HOCH2CH2CH2OO
Table V Reaction Scheme for the Photooxidation of 3-Iodopentanea
˙
C2H5CH(I)C H5 + h␯ → C2H5CHC2H5 + I
kphot
I + I + M →2I2 + M
˙
C2H5CHC2H5 + O2 → C2H5CH(OO )C2H5
rapid
kss
)kss
10
kdec
(1a)
(1b)
(2a)
(2b)
(3)
(4a)
(4b)
(5)
2C2H5CH(OO )C2H5 → 2C2H5CH(O )C2H5 + O2
→ C2H5CHOHC2H5 + C2H5COC2H5 + O2
(1
8
15
C2H5CH(O )C2H5 + O2 → C2H5COC2H5 + HO2
˙
C2H5CH(O )C2H5 → C2H5CHO + CH3CH2
˙
CH3CH2 + O2 → C2H5OO
rapid
4.35 10
2.45 10
14
2C2H5OO → 2C2H5O + O2
→ C2H5OH + C2H5CHO + O2
C2H5O + O2 → C2H5CHO + HO2
14
15
9.5 10
1
2
1
(6a)
(6b)
(6c)
(7a)
(7b)
(7c)
C2H5OO + C2H5CH(OO )C2H5 → C2H5O + C2H5CH(O )C2H5 + O2
→ C2H5OH + C2H5COC2H5 + O2
ksp
4
1
ksp
4
→ CH3CHO + C2H5CH(OH)C2H5 + O2
ksp
11
12
12
C2H5CH(OO )C2H5 + HO2 → C2H5CH(OOH)C2H5 + O2
C2H5OO + HO2 → C2H5OOH + O2
1.5 10
5.8 10
2.5 10
HO2 + HO2 → H2O2 + O2
a Rate coefficients used in the calculations: kss = 1 10 15, ksp = (2–8) 10 cm3 molecule
s
1, kphot
1
10
s .
15
1
3
1

PHOTODECOMPOSITION OF IODOPENTANES IN AIR
135
Table VI Experimentally Observed and Calculated
Relative Product Distributions for the Photooxidation of
3-Iodopentane
The rate of decomposition of the 3-pentoxyl radical
relative to its reaction with oxygen is
[propanal]
Calculatedb
k
dec/kO [O2] =
2
[pentan-3-one] [pentan-3-ol]
Product
Experimentala
(a)
(b)
(c)
= 0.61 0.06
Pentan-3-ol
Pentan-3-one
Propanal
Ethanal
Ethanol
Hydroperoxidesc
1.00 0.03
1.80 0.04
0.49 0.02
0.33 0.04
0.10 0.04
nd
1.0
1.0
1.0
Assuming again kO = 8 10 15 cm3 molecule 1 s
,
1
1.80
0.49
0.36
0.09
1.09
1.83
0.51
0.37
0.11
1.10
1.84
0.51
0.38
0.12
1.10
2
one obtains kdec (2.6 0.2) 104 s 1. The recipro-
cal slope of the line in Fig. 6 leads to kdec (2.5
0.14) 104 s 1, which is in good agreement. Again,
the uncertainty of kO is not included in the error limits.
2
aAbout 800 ␮mol mol 1 in air at atmospheric pressure; nd = not
The branching ratio = 0.39 for pentan-3-yl per-
oxyl radicals is lower than the values derived for the
two other pentylperoxyl radicals. All the products from
pentan-3-yl peroxyl radicals were fully quantified, in
contrast to the other systems studied, where isomer-
ization products were incompletely characterized and
the values are more uncertain. A possible reason for
the apparently low value might be the occurrence of
side reactions of pentan-3-oxyl radicals with iodopen-
tane. The pentan-3-oxyl radical would be more read-
ily affected by such reactions than the corresponding
other two species because the absence of isomerization
processes reduces the absolute loss rate of the pentan-3-
oxyl radical and makes it more susceptible to side reac-
tions. However, we found no difference in the product
distribution when the mixing ratio of 3-iodopentane in
air was raised from 0.1 to 1 mmol mol 1. Accordingly,
thereisnoevidencefortheoccurrenceofsidereactions.
As in other cases, computer simulations of the
oxidation of 3-iodopentane were carried out. Be-
cause pentan-3-yl peroxyl is a secondary radical,
determined.
bThe rate constant used for the peroxyl radical cross-combination
15
reaction in cases (a), (b), and (c) were ksp = 2 10 15, 8 10
,
14
1
and 1.2 10 cm3 molecule
s
1, respectively.
cPredominantly 3-pentyl-hydroperoxide.
at 103 kPa. Figure 6 shows a plot of ([pentan-3-one]
[pentan-3-ol])/[propanal] as a function of the O2
concentration. The straight line obtained demonstrates
the competition between decomposition of the pentan-
3-oxyl radical and its reaction with oxygen. The slope
of the line, which should be given by kO /kdec, has a
2
1
value of (3.15 0.13) 10 19 cm3 molecule
.
The branching ratio for the self-reaction of 3-
pentylperoxyl radicals was obtained from the relative
product yields shown in Table VI as
= kssa/kss
[pentan-3-one] [pentan-3-ol] + [propanol]
=
[pentan-3-one] + [pentan-3-ol] + [propanal]
= 0.39 0.08
Figure 5 Chromatograms (CP-WAX 52 column) of 3-
iodopentane in air at atmospheric pressure before and after
irradiation (lower and upper traces). Identified product peaks
are accentuated: (1) acetaldehyde, (2) propanal, (3) ethanol,
(4) pentan-3-one, (5) pentan-3-ol is hidden underneath the
peak of 3-iodopentane.
Figure 6 Difference in the yield of 3-pentanone and 3-
pentanol, relative to the yield of propanal, versus oxygen
concentration. The solid point is obtained from the experi-
mental data in Table VI. The straight line demonstrates the
competition between decomposition of pentan-3-oxyl and its
reaction with oxygen.

136
HEIMANN ET AL.
the rate coefficient for its self-reaction was assumed
to be 1 10 15 cm3 molecule 1 s 1. The rate coef-
ficient and branching ratio for the self-reaction of
radicals were not observed in the expected amounts.
Eberhard et al. [20] have used derivatization with 2,4-
dinitrophenylhydrazine and liquid chromatography to
identify 2-hydroxyhexan-5-one produced in the pho-
tolysis of hexan-2-nitrite. Kwok et al. [21] have used
atmospheric pressure ionization triple quadrupole mass
spectrometry to characterize hydroxycarbonyl com-
pounds derived from the reaction of OH with several
n-alkanes including n-pentane. The presence of NO in
both studies ensured that alkylperoxyl radicals were
rapidly converted to the corresponding alkoxyl radi-
cals, so that the concentrations of peroxyl radicals were
kept low. The experimental conditions of this study did
not allow the addition of NO in amounts sufficient to
convert all of the peroxyl radicals to alkoxyl radicals.
Thus, high concentrations of peroxyl radicals had to
be tolerated. In the case of 1-iodopentane, the calcu-
lations indicate that 4-hydroxypentanal should be the
dominant isomerization product. Although we have ob-
served a signal at a retention time characteristic of this
product, nopositiveidentificationcouldbemadeforthe
lack of an authentic sample. In contrast to the photooxi-
dation of 1-iodopentane, which is still reasonably well
behaved, that of 2-iodopentane has remained largely
ambiguous because we were unable to assign a greater
number of products at retention times characteristic
of mixed hydroxycarbonyl compounds, which appear
to arise from the decomposition of an unstable inter-
mediate. Our assumption that these products originate
from isomerization of pentan-2-oxyl may be incorrect.
In view of this difficulty, only a few exploratory ex-
periments were made with 2-iodopentane. Much more
analytical work will be required to enable analysis of
these data.
ethylperoxyl radicals are [5] kpp = 6.8 10 14 cm3
molecule 1 s
,
= 0.64. The rate coefficient ksp
1
for the cross-combination reaction between primary
ethylperoxyl and secondary pentan-3-yl peroxyl rad-
icals was taken to be an adjustable parameter. The
results of the calculations are included in Table VI.
The product distribution is best represented with ksp
=
(2 8) 10 15 cm3 molecule 1 s 1. The calculations
further demonstrate that pentan-3-yl hydroperoxide,
which cannot be analyzed by gas chromatography,
should be formed with a fairly high yield (21%).
DISCUSSION
The results indicate that 254-nm photolysis of 1-
iodopentane and 3-iodopentane in air predominantly
generate pentan-1-yl peroxyl and pentan-3-yl per-
oxyl radicals, respectively. The photooxidation of
2-iodopentane gave rise to a greater number of uniden-
tified products that may partly have resulted from un-
known photodecomposition processes other than clean
dissociation to form an iodine atom and pentan-2-yl.
If so, this would contrast with 2-iodobutane, where
elimination of HI occurs in addition to cleavage of
the carbon–iodine bond, but not any other notice-
able fragmentation [10]. Hence, we have assumed
that the unidentified peaks observed in the case of
2-iodopentane are products resulting from the isomer-
ization of pentan-2-oxyl radicals. Relatively high con-
centrations of iodopentanes were applied, without any
indicationthattheconcentrationrangewasdetrimental.
In addition to the carbonyl compounds and alcohols re-
sulting from the self-reactions of the three peroxyl rad-
icals studied, products arising from the decomposition
pathways of the corresponding alkoxyl radicals could
be quantified, and the relative yields were determined.
It is disappointing to note that the diols and hydroxy-
carbonyl compounds predicted to arise from isomeriza-
tion reactions of the pentan-1-oxyl and pentan-2-oxyl
Table VII summarizes our results for the branching
ratios of the self-reactions of the three pentylperoxyl
radicals studied. The values for pent-1-yl peroxyl and
pent-2-yl peroxyl are affected by the extent of isomer-
ization of the corresponding pentoxyl radicals, so that
they carry a larger range of uncertainty than the branch-
ing ratio for pent-3-yl peroxyl. All three values are
rather similar indicating that 40–45% of the reactions
enter into the radical propagating channel and 55–60%
Table VII Summary of Branching Ratios for the Self-Reaction of Pentyl Peroxyl Radicals and Approximate Rate
Coefficients for the Decomposition and Isomerization of Pentoxyl Radicals Relative to Their Reaction with Oxygen
k
dec/k
kiso/k
O2
(1019 molecule cm
O2
3
a
1
3
a
1
Radical Source
(1018 molecule cm
)
kdec (104 s
)
)
kiso (105 s
)
1-Iodopentane
2-Iodopentane
3-Iodopentane
0.42 0.17
0.46 0.10
0.39 0.08
5.09 0.52
1.05 0.21
3.20 0.31
4.8 0.5
0.84 0.17
2.6 0.3
4.3 1.2
1.2 0.3
–
4.0 1.1
1.0 0.2
–
15
1
15
1
aAssumptions for pent-1-oxyl kO = 9.5 10 cm3 molecule
s
1; in the other two cases, kO = 8 10 cm3 molecule
s
1, [O2] =
2
2
3
5.2 1018 molecule cm
.

PHOTODECOMPOSITION OF IODOPENTANES IN AIR
137
al. [12], kiso/kO = 3.1 1019 molecule cm 3, which
produce alcohols and carbonyl compounds. Similar
ratioshavebeenobservedforseveralotheralkylperoxyl
radicals such as methyl peroxyl 0.35 [5], neo-pentyl
peroxyl 0.40 [5], primary 2,3-dimethyl-butyl peroxyl
0.44 [14], propan-2-yl peroxyl 0.39 [14]. In the latter
case there exists an alternative value of 0.58 [5].
2
leads to kiso 2.5 105 s 1. The value reported by
Dobe et al. [11], kiso
5
103 s 1, is much too low.
In the case of 1-iodopentane, where only two sig-
nificant isomerization products appeared in our chro-
matograms, the isomerization constant obtained should
be more reliable. however, the value obtained here,
Table VII includes relative rates for the decompo-
sition and isomerization pathways of pentoxyl radi-
cals. As these data were derived from measurements of
the decomposition products, they presumably involve
less uncertainties than the isomerization constants. The
present result for pentan-2-oxyl, 8.4 103 s 1, which
is based on an assumed rate coefficient for the reac-
tion of the pentan-2-oxyl radical with oxygen, is in
good agreement with that obtained by Dobe et al. [11],
who have studied the decomposition of pentan-2-oxyl
at elevated temperatures relative to reaction with NO.
Their Arrhenius expression extrapolated to 295 K gives
kdec = 9.4 103 s 1. Atkinson [4] used a revised rate
coefficient for the reaction of alkoxyl radicals with
NO to obtain a corrected value 9.0 103 s 1. Our own
value would have been higher, had we used the yield
for acetaldehyde rather than the sum of the yields of
propanal and propanol. Our value for the decompo-
sition constant of pent-3-oxyl is in excellent agree-
kiso
4
105 s 1, is significantly smaller than that
calculated. Hein et al. [13] have produced pentan-1-
oxyl radicals by laser photolysis of 1-bromo-pentane
at reduced pressure, but they were only able to show
1
that kiso
1
105 s
.
In summary, it has been shown that the photolysis
of iodopentanes provides a useful technique for gener-
ating individual pentylperoxyl isomers. The observed
product distributions were evaluated to determine
the branching ratios for the radical-preserving path-
ways in the self-reactions of the individual pentylper-
oxyl radicals. Diols and mixed hydroxycarbonyl com-
pounds that should be formed by cross-combination
reactions of hydroxyalkyl peroxyl radicals produced
from pentan-1-oxyl and pentan-2-oxyl radicals after
isomerization were essentially absent, in contrast to
expectation. Although the chromatograms indicated
isomerization products, we were unable to identify
any of them under the conditions of our experi-
ments. On the other hand, the decomposition prod-
ucts of pentan-2-oxyl and pentan-3-oxyl radicals could
be quantified and the decomposition constants were
determined.
ment with that of Atkinson et al. [12] who found
3
k
dec/kO = 3.3 1018 molecule cm using a differ-
2
ent method of analysis. It is of interest to note that
the decomposition constant for pentan-3-oxyl is higher
than that of pentan-2-oxyl, by a factor of about three,
although the heat of decomposition is essentially the
same in both cases. The surprisingly high rate of de-
composition of pentan-1-oxyl disagrees with the esti-
mate of Atkinson [8] for the release of formaldehyde
from the hydrocarbon chain, kdec = 24 s 1. As we did
not analyze for formaldehyde, we had to assume that
the source of butanal is the butyl radical formed in the
decomposition of the pentan-1-oxyl radical. This as-
sumption may be wrong. Our sample of 1-iodopentane
also contained unidentified impurities, which may have
provided a source of butanal.
BIBLIOGRAPHY
1. Alfassi, Z. B. (Ed.). PeroxylRadicals;Wiley:Chichester,
UK, 1997.
2. Carter, W. P. L.; Atkinson, R. J Atmos Chem 1985, 3,
377–405.
3. Atkinson, R. Atmos Environ 1990, 24A, 1–41.
4. Atkinson, R. J Phys Chem Ref Data 1997, 26, 215–290.
5. Lightfoot, P. D.; Cox, R. A.; Crowley, J. N.; Destriau, M.;
Hayman, G. D.; Jenkin, M. E.; Moortgat, G. K.; Zabel,
F. Atmos Environ 1992, 26A, 1805–1964.
6. Wallington, T. J.; Dagaut, P.; Kurylo, M. J Chem Rev
1992, 92, 667–710.
7. Baldwin, A. C.; Barker, J. R.; Golden, D. M.; Hendry,
D. G. J Phys Chem 1977, 81, 2483–2492.
8. Atkinson, R. Int J Chem Kinet 1997, 29, 99–111.
9. Kwok, E. S. C.; Atkinson, R. Atmos Environ 1995, 29,
1685–1696.
10. Ross, P. L.; Johnston, M. V. J Phys Chem 1995, 99,
4078–4085.
11. Dobe, S.; Berces, T.; Marta, F. Int J Chem Kinet 1986,
18, 329–344.
12. Atkinson, R.; Kwok, E. S. C.; Arey, J.; Aschmann, S.
Faraday Discuss 1995, 100, 23–37.
Inviewoftheinadequateproductcharacterization, it
will be clear that the results for the isomerization path-
ways included in Table VII carry large uncertainties,
especially since we assumed that the unidentified prod-
uctsresultedpredominantlyfromisomerization. Atkin-
son[4,8]hasestimatedratecoefficientsof1,5hydrogen
shiftisomerization, usingrevisedArrheniusparameters
compared to those originally given by Baldwin et al. [7]
The derived values (at 298 K) for pent-2-oxyl and pent-
1
1-oxyl are 2.0 105 s and 1.6 106 s 1, respec-
tively. In the first case the value obtained here, 1.0
105 s 1, is in reasonable agreement with prediction,
but still lower than the value obtained by Atkinson et

138
HEIMANN ET AL.
13. Hein, H.; Hoffmann, A.; Zellner, R. Phys Chem Chem
Phys 1999, 1, 3743–3752.
14. Heimann, G.; Warneck, P. J Phys Chem 1992, 96, 8403–
8409.
15. Benkelberg, H.-J.; Bo¨ge, O.; Seuwen, R.; Warneck, P.
Phys Chem Chem Phys 2000, 2, 4029–4039.
16. Brown, H. C.; Wheeler, O. H. J Am Chem Soc 1956, 78,
2199–2202.
18. Curtis, A. R.; Sweetenham, W. P. FACSIMILE/CHECKMAT
Users Manual; Aere-R-12805; Her Majesty’s Stationary
Office: London, 1988.
19. Madronich, S.; Calvert, J. G. J Geophys Res 1990, 95,
5697–5715.
20. Eberhard, J.; Mu¨ller, C.; Stocker, D. W.; Kerr,
J. A. Environ Sci Technol 1995, 29, 232–
241.
17. Williams, H. R.; Mosher, H. S. J Am Chem Soc 1954,
76, 2987–2990.
21. Kwok, E. S.; Arey, J.; Atkinson, R. J Phys Chem 1996,
100, 214–219.