Experimental Setup

Emission characterization and photochemical aging experiments were conducted at the California Air Resources Board’s (CARB) Haagen-Smit Laboratory in El Monte, California. A schematic of the test set-up is shown in Figure 2.2. Detailed descriptions of these facilities have been published previously [20]. The focus of this report is the results from 29 smog chamber

22

experiments, but extensive additional primary only tests were performed with these vehicles and are described in a forthcoming paper [21].

Sixty-four light duty gasoline vehicles (LDGV) were recruited from the California in-use fleet for primary emissions testing and 15 of these vehicles were also tested using the smog chamber (see Table 2.3 for a detailed description of the smog chamber test vehicles and the initial conditions in all 29 chamber experiments). All vehicles tested in the chamber were operated on the same CA summertime commercial gasoline; details of its composition are provided in Table 4.2 below and Table 2.4 in the SI. We did not attempt to control for the greater wear and tear on older vehicles (although mileage and other indicators of use did not correlate with vehicle age in all cases).

Table 2.1. Fuel analysis of gasoline used in all LDGV experiments. The method of analysis and the relevant ASTM are also provided. Further compositional analysis and an elemental mass balance are provided in Table 2.4 of the SI.

6.08 2.11 0.56 23.8 5.20 12.38 0.08 6.8 135 212 313 8.80 0.7410

ASTM 4815, GC/FID

ASTM D5580,

GC/FID ASTM D86

The vehicles were driven on a Clayton (Model AC-48) 48” single roll electric chassis dynamometer, and most experiments used the cold start Unified Cycle (UC) driving schedule (details shown in Figure 2.1). The UC has three phases (bags) and a soak period: Bag 1 is a 1.2 mile cold start phase lasting 300 seconds; Bag 2 is a 8.6 mile trip phase lasting 1135 seconds;

these are run consecutively and followed by a 600 second hot soak (engine off); Bag 3 is a hot start phase and duplicates the specifications of Bag 1. Four hot start UC tests were also run (one pre-LEV, one LEV-1 and two LEV-2 experiments) to investigate the importance of the cold start phase on vehicle emissions. In the hot start experiments the vehicle was warmed up (emissions

23

were not sampled) using the Bag 1 portion of the cycle; immediately after finishing the warm-up the vehicle was driven according to the normal UC.

Figure 2.1. All vehicles were operated on a chassis dynamometer according to the Unified Cycle (UC), which consists of three phases, simulating urban stop-and-go driving conditions.

The UC is a more aggressive driving cycle than the federal FTP-75, featuring higher speeds, higher accelerations, fewer stops per mile, and less idle time. Emissions from vehicles operated over the UC are generally higher then when operated over the FTP—a cycle that has also been used in emission regulations. The two cycles are compared in Table 2.5 of the SI.

Tail pipe emissions were sampled using a Horiba constant volume sampling (CVS) system. Primary PM measurements were made by drawing a sample from the CVS through a pre-baked (to remove adsorbed carbonaceous contamination) 47 mm quartz filter and a combined Teflon filter + quartz filter (to correct for sampling artifacts) at 47°C following

procedures described in CFR 1065 [22]. In brief, the pre-fired quartz-fiber filters were collected

24

and analyzed with a Sunset Laboratory Organic Carbon/Elemental Carbon (OC/EC) Analyzer using the IMPROVE protocol [23].

Figure 2.2. Test set-up used for characterizing the photo-oxidized motor vehicle emissions (not to scale).

For the chamber experiments emissions were transferred from the end of the CVS via electrically heated (47 °C) 0.5” O.D. Silcosteel (i.e., passivated internal bore) tubing to a 7 m³ Teflon smog chamber where they were photochemically aged [18]. Before each experiment the chamber was cleaned by flushing with HEPA- and activated carbon-filtered air overnight. The chamber was clean before each experiment: less than 10 particles per cm³, NOx concentration <5 ppb. For experiments with low black carbon concentrations (listed in Table 2.3) the chamber was seeded to mitigate nucleation; approximately 10 μg/m³ ammonium sulfate was injected into the chamber just before the vehicle was started. The smog chamber was located indoors, in a large air conditioned space; its temperature and humidity varied between 25°C-30°C and 30%-50%.

Vehicle emissions were added to the approximately half-filled chamber over the 38 minute UC (but not during the 10 minute hot soak period); thus, these experiments represent trip

25

average emissions. The chamber was covered (dark) during filling to prevent photo-chemistry.

The exhaust inside the chamber was diluted by a factor of 200-300 compared to the tailpipe in three stages: first, it was diluted approximately 10:1 with ambient temperature HEPA-filtered air in the CVS; it was then diluted another 8-10:1 with 47°C HEPA- and activated carbon-filtered air using Dekati ejector diluters in the transfer line; finally, it was diluted another 2-3:1 in the smog chamber. After filling, PM concentrations inside the chamber were within the range of urban ambient conditions (~0.5 - 15 μg m^-3) for all of the vehicles. The mixing ratios of individual VOCs were typically less than 1 ppb, but were as high as 20 ppb for the highest emitting vehicle. NOx concentrations after injecting emissions were between 0.1 and 2.4 ppm.

After adding exhaust, HONO was introduced into the chamber as an OH radical source by bubbling dry air into a 1:2 solution (volume) of 0.1M NaNO2 and 0.05 M H2SO4 for ~30 minutes and forcing the resulting gaseous mixture into the chamber. VOC/NOx ratios were adjusted to approximately 3:1 (typical of many urban environments) by adding propene (0.0-1.00 ppm), which is not considered to be a SOA precursor [16]. One high NOx experiment

(VOC/NOx = 0.25) was also performed to investigate the impact of this parameter on SOA formation. After ~45 minutes of characterization of the primary emissions in the dark, the emissions were photo-oxidized by exposing them to UV lights (Model F40BL UVA, General Electric) continuously for 3 hours.