Molecular structure impacts on secondary organic aerosol formation from glycol ethers

Abstract

Glycol ethers, a class of widely used solvents in consumer products, are often considered exempt as volatile organic compounds based on their vapor pressure or boiling points by regulatory agencies. However, recent studies found that glycol ethers volatilize at ambient conditions nearly as rapidly as the traditional high-volatility solvents indicating the potential of glycol ethers to form secondary organic aerosol (SOA). This is the first work on SOA formation from glycol ethers. The impact of molecular structure, specifically-OH, on SOA formation from glycol ethers and related ethers are investigated in the work. Ethers with and without-OH, with methyl group hindrance on-OH and with-OH at different location are studied in the presence of NO X and under "NO X free" conditions. Photooxidation experiments under different oxidation conditions confirm that the processing of ethers is a combination of carbonyl formation, cyclization and fragmentation. Bulk SOA chemical composition analysis and oxidation products identified in both gas and particle phase suggests that the presence and location of-OH in the carbon bond of ethers determine the occurrence of cyclization mechanism during ether oxidation. The cyclization is proposed as a critical SOA formation mechanism to prevent the formation of volatile compounds from fragmentation during the oxidation of ethers. Glycol ethers with-CH 2-O-CH 2 CH 2 OH structure is found to readily form cyclization products, especially with the presence of NOx, which is more relevant to urban atmospheric conditions than without NOx. Glycol ethers are evaluated as dominating SOA precursors among all ethers studied. It is estimated that the contribution of glycol ethers to anthropogenic SOA is roughly 1% of the current organic aerosol from mobile sources. The contribution of glycol ethers to anthropogenic SOA is roughly 1% of the current organic aerosol from mobile sources and will play a more important role in future anthropogenic SOA formation.

Potential of select intermediate-volatility organic compounds and consumer products for secondary organic aerosol and ozone formation under relevant urban conditions

“…DEGEE, Propylene Glycol, and DBE-5 show moderate aerosol formation. The SOA formed from DEGEE and DEGBE is observed to decrease slightly during the last 3 h of the experiments, which may indicate that some semi-volatile species partition back from aerosol phase to gas phase and fragmented during further oxidation (Li and Cocker, 2017). The SOA formations from n-Tridecane, DPGMEA, DBE-5, Propylene Glycol, DEGEE and DEGBE are all delayed, which may be due to formation of low-volatility compounds through multigeneration processes.…”

Reactive organic carbon emissions from volatile chemical products

“…S1 in the Supplement ). These include (1) all linear alkanes use a quadratic polynomial fit to the volatility basis set (VBS) data from Presto et al (2010) at 10 µg m −3 ; (2) all cyclic alkanes use linear alkane yields that are three carbons larger in size ( Tkacik et al, 2012 ); (3) all branched alkanes use yields obtained from the Statistical Oxidation Model (SOM; Cappa and Wilson, 2012 ), as reported in McDonald et al (2018) ; (4) benzene and xylenes use the average yields from Ng et al (2007) under high-NO x conditions; (5) toluene uses the average from Ng et al (2007) under high-NO x conditions and the VBS data from Hildebrant et al (2009) at 10 µg m −3 ; (6) all alkenes use yields obtained from SOM, as reported in McDonald et al (2018) ; (7) volatile methyl siloxanes use the two-product model parameters from Janecheck et al (2019) , which includes additional SOA yields from Wu and Johnson (2017), at 10 µg m −3 ; (8) all glycol ethers use chamber results and molecular structure relationships from Li and Cocker (2018) for reported and unreported glycol ethers, respectively; (9) benzyl alcohol uses the average of the lower-bound yields reported by Charan et al (2020) ; (10) all remaining non-cyclic oxygenates, where available, use the arithmetic average of SOM results and a 1-D VBS approach, as reported by McDonald et al (2018) ; (11) all remaining cyclic oxygenates, where available, use yields obtained from SOM, as reported by McDonald et al (2018) ; (12) all halocarbons and compounds with less than five carbons are assigned a yield of zero; and (13) all remaining species are conservatively assigned a yield of zero if the effective saturation concentration (i.e., C ∗ = ( P vap ·MW) / ( R·T )) is ≥3×10 6 µg m −3 and assigned the same yield as n -dodecane if the effective saturation concentration is < 3×10 6 µg m −3 . The MIR of each compound, which measures the formation potential of ozone under various atmospheric conditions where ozone is sensitive to changes in organic compounds ( Carter, 2010b ), is calculated using the SAPRC-07 chemical mechanism ( Carter, 2010a ) and expressed as a mass of additional ozone formed per mass of organic emitted ( Carter, 2010b ).…”

“…Second, adequate chemical mechanism surrogates for species common in VCPs (e.g., siloxanes) are lacking ( Qin et al, 2020 ). As VCPs and their components could have significant SOA potential ( Li et al, 2018 ; Shah et al, 2020 ), revisiting VCP emissions mapping to chemical mechanisms could help reduce modeled bias, which has historically been difficult to resolve ( Baker et al, 2015 ; Ensberg et al, 2014 ; Lu et al, 2020 ; Woody et al, 2016 ). Third, VCPs feature substantial quantities of intermediate-volatility organic carbon (IVOC) compounds ( CARB, 2019 ), and better representing their source strength could help resolve the high IVOC concentrations observed in urban atmospheres ( Lu et al, 2020 ; Zhao et al, 2014 ).…”

“…For example, reducing the magnitude of regulatory VOC emissions from VCPs can be accomplished by reformulating a product with lower-volatility ingredients that are less likely to evaporate ( Ozone Transport Commission, 2016 ). However, if these lower-volatility replacement ingredients eventually evaporate on atmospherically relevant timescales, they could be efficient SOA precursors ( Li et al, 2018 ).…”

Ultrafine particle emission from floor cleaning products

“…The presence of air pollutants related to the use of cleaning products is due to the ozone‐initiated reactions of volatile organic compounds (VOCs), such as terpenes and terpenoids (eg, d‐limonene), 9‐14 and glycol ethers 15 included in the cleaning product composition. The ozone‐VOC interactions produce new oxygenated volatiles, for example, formaldehyde and acetaldehyde, 11,13,16,17 and some multi‐oxygenated compounds (eg, dicarbonyls, peroxides) that may condense forming ultrafine secondary organic aerosols (SOAs) 14,18‐23 .…”