for the mixtures nearest to tropospheric composition (142
Torr O , 562 Torr N ). Results for both OH-initiated and
Cl-initiated oxidations are shown, each being the mean of six
runs. Conclusions of previous investigations are included for
comparison.
qualitatively by adding to the original mechanism reaction (8),
the analogue of reaction (6) but with the formation of CH O
2
2
3 2
instead of HO .
2
C H OCH(O)CH ] O ] HCOOC H ] CH O
(8)
2
5
3
2
2 5 3 2
The product yields in this work are in satisfactory agree-
ment with previous studies notwithstanding the far higher
reactant pressures used. It should be noted that the values
quoted for ref. 5 are the fractions of the decomposition of 1-
ethoxyethoxy proceeding by routes (4), (5), and (6), rather than
the yields as deÐned above. Since the C H O radical produc-
Quantitatively, however, this is unsatisfactory. A mechanism
involving (4), (5), (6) and (8) would lead to a linear dependence
of ethyl formate/ethyl acetate on the reciprocal of the oxygen
pressure and this is not observed. Moreover, at very high
oxygen pressures the ratio ethyl formate/ethyl acetate
approaches k /k and the value of about 6 seems improbable.
2
5
ed in reaction (5) is partly converted to more CH CHO, the
8 6
A more probable explanation is that there is a component
of ethyl acetate formation which is independent of oxygen.
The most likely process is the decomposition step (9), in which
a hydrogen atom is ejected from 1-ethoxyethoxy.
3
experimental yield of acetaldehyde was reduced in ref. 5 to
obtain the fraction given.
E†ect of variation of oxygen pressure
C H OCH(O)CH ] CH COOC H ] H
2 5
(9)
The mechanism shown earlier implies that the yield of ethyl
acetate should depend on the pressure of oxygen. Product
yields were measured in runs with various pressures of oxygen
but with the same total pressure of 703 ^ 3 Torr.
2
5
3
3
Step (9) is an alternative to step (4) which produces the major
product ethyl formate by Ðssion of a carbonÈcarbon bond.
For simple alkoxy radicals such as ethoxy and isopropoxy
CÈH Ðssion is about 25 kJ mol~1 more endothermic than
CÈC Ðssion.6 Assuming similar Arrhenius A factors this would
lead to CÈC Ðssion being more favourable by a factor in rate
of 2.4 ] 104.
For the CÈC Ðssion step (4) yielding ethyl formate however,
the estimated enthalpy change is [2.5 kJ mol~1 and that for
step (9) indicates it to be thermoneutral.7 A di†erence of 2.5 kJ
mol~1 would correspond to a factor of 2.74 in favour of step
4) over step (9).
The mechanism involving steps (4), (5), (6), and (9) predicts a
The results are shown in Table 2, in which each yield is the
mean of the results of six runs. As can be seen in Fig. 1,
increasing the oxygen pressure has little e†ect on the yield of
the major product, ethyl formate, but causes some increase in
acetaldehyde and in ethyl acetate. The increase in acetal-
dehyde can be attributed to an increase in the fractional con-
version of C H O to CH CHO with increasing O . The
2
5
3
2
increase in ethyl acetate is due to reaction (6). However,
according to the mechanism given above, the ratio of the
yields of ethyl acetate and ethyl formate should be directly
proportional to the oxygen pressure. This ratio is shown in
Fig. 2, and its weak dependence on oxygen pressure is clearly
at odds with the mechanism.
(
linear dependence of ethyl acetate/ethyl formate on oxygen
pressure in satisfactory agreement with the results in Fig. 2.
The intercept is k /k \ 0.11 at 293 K, and k /k \ 1.9
To account for the discrepancy the mechanism requires
modiÐcation either to make ethyl acetate production less
dependent on oxygen, or the formation of ethyl formate more
oxygen-dependent. The latter alternative might be achieved
9 4
6 4
]
10~21 cm3 molecule~1 is derived from the slope.
Reaction (9) should therefore be added to the original oxi-
dation mechanism. The results indicate that under atmo-
spheric conditions about 90% of the ethyl acetate is produced
by reaction (9) rather than reaction (6). Support for this pro-
posal is provided by evidence for the occurrence of the corre-
sponding H-atom ejection from the methoxymethoxy radical
reported by Jenkin et al.8
Estimates of the heats of formation of the methoxymethoxy
and 1-ethoxyethoxy radicals using BensonÏs additivity rules
suggest that both these radicals eliminate hydrogen with a low
endothermicity. A further example is provided by the decom-
position of the HOCH O radical for which an enthalpy
2
change of 13.4 kJ mol~1 has been given.9 This decomposition
has been invoked to explain the chain process producing H
in the photo oxidation of formaldehyde. Clearly the elimi-
2
nation of hydrogen atoms from these oxygenated alkoxy rad-
icals may have a wider signiÐcance which cannot be ignored.
It has been suggested by a referee that the apparent weak
dependence of the ethyl acetate/ethyl formate ratio on oxygen
pressure could be due to the relatively high NO pressure used
in this work and the consequent dominance of reactions (10)
and/or (11) over reaction (6).
Fig. 1 E†ect of oxygen pressure on product yields. EF, ethyl
formate; EA, ethyl acetate; AA, acetaldehyde.
CH CH OCH(O)CH ] NO ] CH CH OCH(ONO)CH
3
2
3
3
2
3
(
10a)
]
CH CH OC(O)CH ] HNO
3
2
3
(
10b)
CH CH OCH(O)CH ] NO ] CH CH OCH(ONO )CH
3
2
3
2
3
2
2
3
(
11a)
]
CH CH OC(O)CH ] HONO
3
2
3
Fig. 2 E†ect of oxygen pressure on the ratio of ethyl acetate to ethyl
formate. EA, ethyl acetate; EF, ethyl formate.
(11b)
3244
Phys. Chem. Chem. Phys., 1999, 1, 3243È3245
Cheema
