Dimethyl ether (DME) oxidation is
a model chemical system
with
a small number of prototypical reaction intermediates that also has
practical importance for low-carbon transportation. Although it has
been studied experimentally and theoretically, ambiguity remains in
the relative importance of competing DME oxidation pathways in the
low-temperature autoignition regime. To focus on the primary reactions
in DME autoignition, we measured the time-resolved concentration of
five intermediates, CH3OCH2OO (ROO), OOCH2OCH2OOH (OOQOOH), HOOCH2OCHO (hydroperoxymethyl
formate, HPMF), CH2O, and CH3OCHO (methyl formate,
MF), from photolytically initiated experiments. We performed these
studies at P = 10 bar and T = 450–575
K, using a high-pressure photolysis reactor coupled to a time-of-flight
mass spectrometer with tunable vacuum-ultraviolet synchrotron ionization
at the Advanced Light Source. Our measurements reveal that the timescale
of ROO decay and product formation is much shorter than predicted
by current DME combustion models. The models also strongly underpredict
the observed yields of CH2O and MF and do not capture the
temperature dependence of OOQOOH and HPMF yields. Adding the ROO +
OH → RO + HO2 reaction to the chemical mechanism
(with a rate coefficient approximated from similar reactions) improves
the prediction of MF. Increasing the rate coefficients of ROO ↔
QOOH and QOOH + O2 ↔ OOQOOH reactions brings the
model predictions closer to experimental observations for OOQOOH and
HPMF, while increasing the rate coefficient for the QOOH →
2CH2O + OH reaction is needed to improve the predictions
of formaldehyde. To aid future quantification of DME oxidation intermediates
by photoionization mass spectrometry, we report experimentally determined
ionization cross-sections for ROO, OOQOOH, and HPMF.