Abstract
Diesel exhaust emissions were introduced into an atmospheric simulation chamber and measured using thermal desorption (TD) comprehensive two-dimensional gas chromatography coupled to a flame ionisation detector (GC x GC-FID). An extensive set of measurements were performed to investigate the effect of different engine conditions (i.e. load, speed, "driving scenarios") and emission control devices (with or without diesel oxidative catalyst, DOC) on the composition and abundance of unregulated exhaust gas emissions from a light-duty diesel engine, fuelled with ultra-low sulfur diesel (ULSD). A range of exhaust dilution ratios were investigated (range = 1 : 60 to 1 : 1158), simulating the chemical and physical transformations of the exhaust gas from near to downwind of an emission source. In total, 16 individual and 8 groups of compounds (aliphatics and single-ring aromatics) were measured in the exhaust gas ranging from volatile to intermediate volatility (VOC-IVOC), providing both detailed chemical speciation and groupings of compounds based on their structure and functionality. Measured VOC-IVOC emission rates displayed excellent reproducibility from replicate experiments using similar exhaust dilution ratios. However, at the extremes of the investigated exhaust dilution ratios (comparison of 1 : 60 and 1 : 1158), measured VOC-IVOC emission rates displayed some disagreement owing to poor reproducibility and highlighted the importance of replicate sample measurements. The investigated DOC was found to remove 43 +/- 10% (arithmetic mean +/- experimental uncertainty) of the total speciated VOC-IVOC (Sigma SpVOC-IVOC) emissions. The compound class-dependant removal efficiencies for the investigated DOC were 39 +/- 12% and 83 +/- 3% for the aliphatics and single-ring aromatics, respectively. The DOC aliphatic removal efficiency generally decreased with increasing carbon chain length. The Sigma SpVOC-IVOC emission rates varied significantly with different engine conditions, ranging from 70 to 9268 mg kg(-1) (milligrams of mass emitted per kilogram of fuel burnt). Sigma SpVOC-IVOC emission rates generally decreased with increasing engine load and temperature, and to a lesser degree, engine speed. The exhaust gas composition changed considerably as a result of two influencing factors: engine combustion and DOC hydrocarbon (HC) removal efficiency. Increased engine combustion efficiency resulted in a greater percentage contribution of the C-7 to C-12 n-alkanes to the Sigma SpVOC-IVOC emission rate. Conversely, increased DOC HC removal efficiency resulted in a greater percentage contribution of the C-7 to C-12 branched aliphatics to the Sigma SpVOC-IVOC emission rate. At low engine temperatures (<150 degrees C, below the working temperature of the DOC), the contribution of n-alkanes in the exhaust gas increased with increasing combustion efficiency and may be important in urban environments, as n-alkanes are more efficient at producing secondary organic aerosol (SOA) than their branched counterparts. At very high engine temperatures (maximum applied engine speed and load, engine temperature = 700 degrees C), the n-alkane contribution increased by a factor of 1.6 times greater than that observed in the cold-start experiment (most similar to unburnt fuel) and may suggest liquid-fuel-based estimates of SOA yields may be inconsistent with exhaust SOA yields, particularly at high engine speeds and loads (i.e. high engine temperatures).
Emission rates were found to be 65 times greater from a cold-start experiment than at maximum applied engine speed and load. To our knowledge, this is the first study which uses an atmospheric simulation chamber to separate the effects of the DOC and combustion efficiency on the exhaust gas composition.