*2.1. Data Sources*

Hourly VOCs and ozone concentration observation data throughout 2019 were from the Shenyang ecological environment monitoring center of Liaoning Province. VOCs' monitoring is based on a GC5000 online gas chromatograph produced by the AMA Instrument Company (Ulm, Germany). The instrument consists of two subsystems (a GC5000 VOC analyzer and a GC5000 BTX analyzer), a calibration module (a DIM200 VOC calibration instrument) and other auxiliary equipment. A type 49i ozone analyzer produced by Thermo Fisher China Co., Ltd. (Shanghai, China) was used for ozone monitoring. The monitoring data of wind direction and speed used to analyze the distribution of wind roses in Shenyang are from the Shenyang National Basic Meteorological Station (NBMS).

The three VOC environmental monitoring sites in Shenyang are Huagongyuan (HGY), Danan Street (DNS), and Tianzhushan Road (TZSR), and the 11 state-controlled ambient air quality monitoring sites are Dongling Road (DLR), Hunnan East Road (HNR), Jingshen Street (JSS), Dongling Street (LDS), Senlin Road (SLR), Shenliao West Road (SLXR), Taiyuan Street (TYS), Wenhua Road (WHR), Xiaoheyan (XHY), Xinxiu Street (XXS), and Yunong Road (YNR). The locations of each air quality monitoring site and Shenyang NBMS are shown in Figure 1. The characteristics and levels of ozone, NOx, and PM2.5 are listed in Table 1. HGY, DNS, LDS, TYS, WHR, XHY, and NBMS are urban sites; TZSR, DLR, HNR, XXS, YNR, SLXR, and JSS are suburban sites; and SLR is a rural site due to its remote location. The land-use at the suburban sites is a mixture of city and cropland, whereas the rural site, SLR, is primarily covered by forest and a reservoir. HGY is located in an economic-technical development zone with many enterprises, including chemical, electrical, and pharmaceutical factories, etc., which leads to a large number of industrial source emissions, while the transport of freight vehicles also emits vehicle exhaust. In addition, HNR and XXS are accessible via a highway around the city area of Shenyang.

**Figure 1.** Map of Liaoning Province and locations of the ambient air quality monitoring sites in Shenyang.

Regarding the key pollutants, the ozone level at the suburban sites is higher than that at the urban sites in general, among which the maximum appears at YNR, a downwind suburban site located in the direction of the prevailing wind from Shenyang. Spatial variation in Shenyang has been reported based on the observations from 2013 to 2015 as well [25]. However, the levels in NOx and PM2.5 are significantly higher at urban sites than at suburban sites, with a peak level in NOx appearing at HGY and high levels at the other two VOC environmental monitoring sites. Compared to pollutant levels in Beijing in 2016 [26], PM2.5 in Shenyang during 2019 was relatively close, whereas NOx in Shenyang was around 10 ppbv higher, meaning that the current high-level of NOx in Shenyang still cannot be negligible.

*Atmosphere* **2021**, *12*, 1240


#### *2.2. Data Analysis Methods*

In 2019, 27 types of alkanes, 14 types of aromatic hydrocarbons, 10 types of alkenes and one type of alkyne were detected at the three VOCs environmental monitoring sites in Shenyang, with a total of 52 VOC species. The original concentration unit of each VOC species was μg/m3. First, according to the method of Liu et al. [25], the VOC concentrations were converted into volume mixture ratios (with the units of ppbv) at a standard temperature and pressure. Then, the diurnal and seasonal variation characteristics of VOC concentrations in Shenyang were analyzed using the mean hourly concentrations of the three monitoring sites.

#### 2.2.1. Aerosol Formation

Based on the smokebox experiment of Grosjean and Seinfeld [19], this study used FAC, the fraction of VOC converted into aerosol, called aerosol yield, whereby the change in the amount of SOA formed is divided by the change of emissions of individual VOCs in order to estimate the SOA formation potential of atmospheric VOCs. According to Grosjean's hypothesis, the formation of SOA only takes place during the day (from 8 a.m. to 5 p.m. Beijing time), and VOCs only react with hydroxyl radicals to form SOA. The FAC and FVOCr (the fraction of VOC consumed by gas-phase chemical reactions with the unit of %) used in the formula are obtained by the smokebox experiment [27]. According to Grosjean [27], the estimations of FVOCr for each VOC were completed for a number of scenarios that specify the concentrations of the electrophiles, e.g., ozone, hydroxyl radicals, and NO3, as well as the airmass transportation time. The daytime scenario selected here was relevant to southern Californian smog episodes, which was set at O3 = 100 ppbv, OH = 1.0 × 10<sup>6</sup> molecules cm<sup>−</sup>3, and NO3 = 0, with a chemical reaction time of 6 h. For this scenario, Grosjean calculated for each alkene the fraction that was consumed by its reaction with ozone and by its reaction with hydroxyl radicals, respectively (these calculations were not necessary for alkanes, aromatics, and saturated oxygenates, which reacted only with hydroxyl radicals). He also assumed that aerosol production from alkene was via their reaction with ozone. The assumption was supported by results from a number of experimental studies on cyclic alkene, which was shown to produce aerosol by reaction with ozone and little or no aerosol by reaction with hydroxyl radicals [19,28].

Components that can form an SOA can be defined as:

$$\text{FAC}\_{i} = \text{SOAp} / \text{VOCs}\_{0} \tag{1}$$

In the formula, FACi is the FAC of the ith VOC as a dimensionless quantity. SOAp is the formation potential of SOA with the unit of ppbv. VOCs0 should have been the initial emission amount quantified in mass, moles, or flux units for a given region or for a unit area. However, due to the limited technology, it was almost impossible to obtain flux data for so many VOC components until now. Therefore, here, an "approximate initial concentration" of VOCs0 is used to represent the levels of emission source with the unit of ppbv. The FACs are used to calculate the SOA formation potential:

$$\text{SOAPp} = \text{VOCs}\_0 \times \text{FAC} \tag{2}$$

Considering that the VOCs measured at the receptor point (VOCst) are usually the concentrations after oxidation, the relationship between VOCst and the initial concentration of VOCs0 can be expressed by the following formula:

$$\text{VOCst} = \text{VOCs}0 \times (1 - \text{FVOCr}/100) \tag{3}$$

#### 2.2.2. Ozone Formation

The types of VOC in the atmosphere are very complex, with the resulting amount of ozone generated from the reaction dependent on the levels, rates of oxidation, oxidation mechanism, and concentrations of NOx. The MIR method considers the impacts of different

reaction mechanisms and VOC/NOx ratios on ozone formation. Incremental reactivity (IR) is the change in ozone divided by the change in the emissions of individual VOCs. It is a quantity term used to calculate the OFP (ozone formation potential) of an individual VOC compound, and MIR explains the maximum reactivity condition of individual VOC species in a high NOx environment, in which ozone formation is most sensitive to VOCs. The calculation formula of OFPi is as follows:

$$\rm{OFP\_i} = \rm{VOCs\_0} \times \rm{MIR\_i} \tag{4}$$

In the formula, OFPi is the OFP of the ith VOC with the unit of ppbv. MIRi is the maximum incremental reactivity of the ith VOC in ozone as a dimensionless quantity. Similarly, the relationship between VOCst and VOCs0 can refer to formula (3), which converts measured concentrations into the approximate initial concentrations [29].
