7
036 J. Phys. Chem. A, Vol. 108, No. 34, 2004
Baker et al.
rate constant ratio k2/k1 as k2/k1 decreases below ∼0.5, with the
maximum sensitivity occurring when k2/k1 ≈ 1.0. While
measurements of the absolute concentrations of the hydroxy-
carbonyls would provide additional data concerning the rate
constant ratios k2/k1 [see Figure 1 (top)], the uncertainties in
the hydroxyaldehyde concentrations arising from the estimated
OH radicals,22 with O3 and NO3 radicals potentially being
formed as the irradiations proceed and NO is converted to NO2
(by reactions of HO2 and organic peroxy radicals with NO) and
NO2 is formed by the photolysis of methyl nitrite. However,
the measured NO concentrations at the end of the experiments
1
3
-3
were in the range (0.94-1.9) × 10 molecule cm for the
experiments with 2-methyl-3-buten-2-ol and in the range (0.97-
5,13
SPME/GC-FID analysis response factors
preclude this.
1
3
-3
Hence, while the nonlinear least-squares fits lead to values of
A in eq II, these are in arbitrary units and give no information
concerning the values of Rk1/(k2 - k1).
The rate constant ratios, k2/k1, given in Table 1 are placed
on an absolute basis by use of rate constants k1 (in units of
1.7) × 10 molecule cm for those with cis-3-hexen-1-ol.
These final NO concentrations were sufficiently large that
formation of O3 and, hence, of NO3 radicals was of no
importance, and hence, losses of 2-methyl-3-buten-2-ol and cis-
3-hexen-1-ol due to reactions with O3 and NO3 radicals could
be neglected.
-
12
3
-1 -1
1
0
cm molecule s ) for the precursor alcohol at 296 K
18
18
of 1,2-butanediol, 25.1 ( 1.3; 1,3-butanediol, 30.9 ( 1.0;
The rate constants, k2, derived in this work for the hydroxy-
aldehydes CH3CH2CH(OH)CHO, CH3CH(OH)CH2CHO, CH3-
CH(OH)CHO, (CH3)2C(OH)CHO, and HOCH2CH2CHO are the
first reported for these compounds. Our present rate constant,
k2, for the hydroxyketone CH3CH2C(O)CH2OH is in good
and 2-methyl-2,4-pentanediol, 25.8 ( 2.218 (all three being
-
12
3
reevaluated with k(OH + n-octane ) 8.07 × 10
cm
-
1
-1
25
26,27
molecule
s
at 296 K ); 2-methyl-3-buten-2-ol, 60 ( 6;
28
and cis-3-hexen-1-ol, 108 ( 4, and the resulting rate constants,
k2, are also given in Table 1.
29
agreement with our previous, more-direct measurement. While
our present rate constant, k2, for the hydroxyketone CH3C(O)CH2-
Discussion
CH2OH is a factor of 1.8 higher than our previous and more-
2
9
direct relative rate measurement, our present data for
CH3C(O)CH2CH2OH are subject to significant scatter (Figure
Dark losses of 1,2- and 1,3-butanediol, 2-methyl-2,4-pen-
tanediol, 2-methyl-3-buten-2-ol, and cis-3-hexen-1-ol in the
Teflon chamber have been shown in recent studies conducted
3). Our present upper limit to the rate constant k2 for 4-hydroxy-
1
3,18,26
4-methyl-2-pentanone [(CH3)2C(OH)CH2C(O)CH3] (Table 1)
in this laboratory
to be of no importance, being <5% over
2
1,30
is consistent with the two literature values,
which are in
periods of 1.5-4.5 h, time periods comparable to those between
the first sample being collected for analysis and the last sample
collection which averaged 3.9 h and had a range of 2.0-5.1 h
good agreement. Furthermore, our present upper limit of
k2(OH + 4-hydroxy-4-methyl-2-pentanone)/k1(OH + 2-methyl-
2
,4-pentanediol) e 0.15 is consistent with the ratio of 0.14 (
(except for one experiment in which this time was 6.5 h).
38
0.04 obtained from our previous measurements of the rate
constants for the reactions of OH radicals with 4-hydroxy-4-
Aliphatic alcohols do not absorb below ∼200 nm, and hence,
photolysis by blacklamps at >300 nm was of no importance,
3
0
18
2
6
methyl-2-pentanone and 2-methyl-2,4-pentanediol. However,
as we have previously verified for 2-methyl-3-buten-2-ol for
light intensities and irradiation times comparable to those
employed here.
21
the rate constants measured recently by Magneron et al. at
2
2
98 ( 3 K for 4-hydroxy-4-methyl-2-pentanone and 2-methyl-
-
12
3
-1 -1
,4-pentanediol, of (3.6 ( 0.6) × 10
cm molecule
s
We have previously observed29 that the hydroxyketones for
which rate constants were measured in this study [i.e., CH3-
CH2C(O)CH2OH, CH3C(O)CH2CH2OH, and (CH3)2C(OH)-
CH2C(O)CH3] show no losses (<2%) over time periods of 5 h,
including 60 min of photolysis at the same light intensity as
-
11
3
-1 -1
and (1.5 ( 0.4) × 10 cm molecule s , respectively, lead
to a rate constant ratio of k2(OH + 4-hydroxy-4-methyl-2-
pentanone)/k1(OH + 2-methyl-2,4-pentanediol) ) 0.24 ( 0.08,
significantly higher than our present upper limit (Table 1). This
discrepancy suggests that the rate constant for 2-methyl-2,4-
that used here. Furthermore, we have observed in recent
2
1
studies13,18
pentanediol measured by Magneron et al. is too low, possibly
because of wall adsorption/desorption problems in the 140 L
that dark losses of the hydroxyaldehydes CH3CH2-
CH(OH)CHO, HOCH2CHO, CH3CH(OH)CH2CHO, CH3CH-
OH)CHO, (CH3)2C(OH)CHO, and HOCH2CH2CHO (formed
2
1
volume Teflon chamber used and/or unrecognized analytical
problems (the rate constant reported by Magneron et al.21 was
obtained using in situ Fourier transform infrared absorption
spectroscopy, and GC-FID analyses were stated21 to have
resulted in higher, but more scattered, rate constants).
(
from CH3ONO-NO-air irradiations of their precursor alcohol
or diol) in our Teflon chambers were <5% over time periods
of 1.2-3.2 h (again reasonably comparable to the time periods
between the first irradiation and the last sample collection, which
averaged 3.4 h and had a range of 1.3-4.4 h (except for one
The two values of the rate constant for the reaction of OH
radicals with glycolaldehyde obtained here are at the lower end
1
6
experiment in which this time was 6.3 h)). Bacher et al. have
shown that glycolaldehyde absorbs out to ∼340 nm with a
quantum yield for photolysis which may be close to unity and
calculated a tropospheric lifetime of glycolaldehyde due to
photolysis of >2.5 days for summertime mid-latitude condi-
1
5-17
of the three literature values
(Table 1). The rate constant
obtained here for propanal is in good agreement with the
1
7,31-36
literature data
(Table 1).
The room-temperature rate constants obtained here for the
hydroxyaldehydes CH CH CH(OH)CHO, CH CH(OH)CH -
1
6
tions. Using this upper limit to the photolysis rate of
glycolaldehyde as representative of those for the hydroxyalde-
hydes studied here, photolysis of the hydroxyaldehydes at the
light intensities and irradiation times used during the experiments
3
2
3
2
CHO, CH CH(OH)CHO, (CH ) C(OH)CHO, and HOCH CH -
3
3 2
2
2
CHO are consistent with expectations. Thus, the rate constants
-
12
3
-1 -1
(in units of 10
2
cm molecule s ) increase from HOCH -
1
7
(
e40 min at ∼50% of the 12 h average mid-latitude solar
2 2 3
CHO (13) to HOCH CH CHO (20) and from CH CH(OH)-
intensity) is expected to be of no importance (<2% loss).
Therefore, dark losses and photolysis of the alcohols and the
hydroxycarbonyl products during the experiments were <5%
and essentially within the analytical measurement uncertainties.
However, 2-methyl-3-buten-2-ol and cis-3-hexen-1-ol react with
O3 and with NO3 radicals in addition to their reactions with
3 2 3
CHO (17) to CH CH CH(OH)CHO (24) and CH CH(OH)-
CH2CHO (30), consistent with the presence of the additional
-
11
3
CH2 group. The rate constants for HOCH2CHO (1.3 × 10
3
-1 -1 17
-11
cm molecule s ), CH3CH(OH)CHO (1.7 × 10
molecule s ), and (CH3)2C(OH)CHO (1.4 × 10
cm
cm
-
1
-1
-1
-11
3
-
1
molecule s ) are similar, suggesting that the majority of the