Characterizing Wall Loss Effects of Intermediate-Volatility Hydrocarbons in a Smog Chamber with a Teflon Reactor

Abstract

Intermediate-volatility organic compounds (IVOCs) serve as pivotal precursors to secondary organic aerosol (SOA). They are highly susceptible to substantial wall losses both in indoor environments and within smog chambers even with Teflon walls. Accurately characterizing the wall loss effects of IVOCs is thus essential for simulation studies aiming to replicate their atmospheric behaviors in smog chambers to ensure precise modeling of their physical and chemical processes, including SOA formation, yet a comprehensive understanding of the wall loss behavior of IVOCs remains elusive. In this study, we conducted a thorough characterization of wall losses for typical intermediate-volatility hydrocarbon compounds, including eight normal alkanes (n-alkanes) and eight polycyclic aromatic hydrocarbons (PAHs), using the smog chamber with a 30 m³ Teflon reactor. Changes in the concentrations of gaseous IVOCs with the chamber were observed under dark conditions, and the experimental data were fitted to the reversible gas–wall mass transfer theory to determine the key parameters such as the wall accommodation coefficient (α_w) and the equivalent organic aerosol concentration (C_w) for different species. Our results reveal that C_w values for these hydrocarbon IVOCs range from 0.02 to 5.41 mg/m³, which increase with volatility for the PAHs but are relative stable for alkanes with an average of 3.82 ± 0.92 mg/m³. α_w span from 1.24 × 10⁻⁷ to 1.01 × 10⁻⁶, with the values for n-alkanes initially showing an increase followed by a decrease as carbon numbers rise and volatility decreases. The average α_w for n-alkanes and PAHs are 3.34 × 10⁻⁷ and 6.53 × 10⁻⁷, respectively. Our study shows that IVOCs exhibit different loss rates onto clean chamber walls under dry and dark conditions, with increasing rate as the volatility decreases. This study demonstrates how parameters can be acquired to address wall losses when conducting smog chamber simulation on atmospheric processes of IVOCs.

1. Introduction

Smog chambers are important tools for investigating chemical and physical processes in the atmosphere by controlling specific compounds, emission sources, and environmental conditions [1,2,3,4,5,6,7,8]. Teflon film has been widely used as a reactor material in such simulation studies because of its chemically inert properties and light transmittance. It has been noted that both gases and particulate matters can be deposited on the surface of the chamber wall, which may have an effect on both gas-phase reactivity and aerosol-phase reaction, resulting in an underestimated formation of particles and reaction products, as well as an overestimated consumption of reaction precursors [9,10,11,12,13,14,15,16].

The wall effect of chemicals is challenging during all chamber experiments. For particulate matters, the rate of concentration decay in a closed container varies with particle size and is affected by the shape of the chamber, the state of mixing, the nature of the particulate itself, and the electrostatic state of the wall [8,17,18,19,20,21,22]. A number of particle wall loss correction methods have been developed as a function of particle concentration and size distribution by determining the first-order particle wall loss rate constant, k [1,10,23,24,25,26,27]. For gases, losses on the chamber wall can be dominated by either the diffusion process to the wall surface or the molecules’ adsorption on the wall, which depends on the mass accommodation coefficient (α_w) [28]. While volatile organic compounds (VOCs), such as n-butane, n-pentane, toluene, and methyl nitrate, have shown no or negligible wall effects [29], studies have indicated that intermediate- and low-volatility compounds can be lost substantially to the chamber wall surfaces [30,31,32,33,34,35,36]. In comparison to low-volatility compounds, which have near zeroth-order and quasi-irreversible wall loss rates, IVOCs have reversible and more complicated dynamic processes [33,35].

Intermediate-volatility organic compounds (IVOCs) are a group of organic compounds with effective saturation concentrations between 10³ and 10⁶ μg/m³, corresponding roughly to the volatility range from C₁₂ to C₂₂ n-alkanes [37,38,39,40]. In this volatility range, several species, including C₁₂–C₂₂ n-alkanes, naphthalene, methylnaphthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, alkylcyclohexane, and certain phenolic compounds such as cresol, catechol, guaiacol, syringol and so on, have been observed in numerous field campaigns [41,42,43,44,45,46] and showed higher potential to form secondary organic aerosols (SOAs) [47,48,49,50,51,52,53,54,55]. For example, SOA yield from the reaction of n-octane with an OH radical in the presence of NO_x is 0.5%, while the yield of n-tetradecane is 55.3% [52]. To better understand the potential for SOA production from IVOCs, numerous chamber experiments have been employed to study oxidation processes and product compositions [51,52,56,57,58,59,60,61]. Due to the relatively low volatility, IVOCs could be adsorbed significantly on Teflon walls according to the theory of gas/particle partition [62]. Neglecting the impact of wall loss of IVOCs may lead to an underestimation of yields for secondary products and an inaccurate representation of aging processes [14,15,63,64]. Previously, we performed a series of experiments to evaluate the wall loss rates of both particle and VOC species inside the smog chamber at the Guangzhou Institute of Geochemistry, Chinese Academy Sciences (GIG-CAS) [65,66,67,68,69,70]. However, the loss of IVOC species on the GIG-CAS smog chamber walls is not yet well characterized. In this study, we selected typical intermediate volatile hydrocarbon compounds such as n-alkanes and PAHs to investigate their wall loss behavior in the smog chamber, which will lay a solid basis for subsequent chamber simulation work on IVOCs and provide valuable implications for the indoor behavior of IVOCs as well.

2. Materials and Methods

2.1. Experimental Setup

Experiments were carried out in the GIG-CAS smog chamber with a 30 m³ Teflon reactor, utilizing two fans installed inside for active mixing. The reactor is surrounded by temperature-controlled enclosure equipment with a series of sensors and monitors, which measure and show the evolution of temperature, pressure, and relative humidity during the experiments. The fans are coated with inert Teflon material, ensuring that the chemicals introduced are well mixed for 120 s. Further detailed information regarding the chamber setup can be found in previous studies [65,66,67,68,69,70]. Prior to wall loss experiments, the chamber was flushed for at least 24 h and filled with clean, dry air at approximately 298 K to ensure that the background levels of potential oxidants are sufficiently low. The relative humidity was kept below 10% and particle concentrations were virtually negligible (particle number concentration below 1 particle cm⁻³) throughout the wall loss experiment of IVOCs, thus mitigating the influence of the water layer on the Teflon surface and minimizing mass transfer from gas to particles. The chamber was maintained in a light-free environment during the experiment to prevent potential oxidation of light-absorbing PAHs as well as the generation of OH radicals resulting from Teflon wall exposure to irradiation, effectively minimizing any additional chemical losses that could occur. About 2.3 ppm of CO was introduced into the chamber and monitored continuously to test for dilution effects (Figure S1).

2.2. Reagents and Instrumentation

The purchased standards were as follows: ammonium sulfate (99% purity, Aladdin, Shanghai, China); eight n-alkanes (all above 99% in purity, Sigma Aldrich, St. Louis, MO, USA) including n-dodecane, n-tridecane, n-tetradecane, n-pentadecane, n-hexadecane, n-heptadecane, n-octadecane, and n-nonadecane; and eight PAHs (all above 99% in purity, Sigma Aldrich, St. Louis, MO, USA) including naphthalene, 1-methyl-naphthalene, 2-methyl-naphthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, and anthracene.

For the particle wall loss experiment, the ammonium sulfate was dissolved in ultrapure water at a concentration of 0.36 mol/L. The solution was then transferred to an atomizer (AMT 220, Topas GmbH, Dresden, Germany), and aerosols were generated and introduced into the chamber with dry air. The purge gas was halted once the particle concentration in the chamber reached 1 × 10⁴ particles cm⁻³. The concentration and size distribution of particles was measured by a scanning mobility particle sizer (SMPS, Model 3081 classifier, Model 3775 CPC, TSI Inc., Shoreview, MN, USA).

2.3. Analysis of IVOCs

For the IVOC wall loss experiment, n-alkanes and PAHs were flushed with a stream of clean air using an injector, which utilized a 12 cm long × 0.375 cm O.D. stainless steel tube extending into the chamber while controlling the internal temperature of the injector at 473 K. This relatively short tube length and elevated temperature were employed to minimize the loss of organic compounds (OCs) to the tube’s inner wall and to facilitate evaporation. The sampling gas stream was conveyed through a 100 cm long × 0.375 cm O.D. stainless steel tube, which was maintained at a temperature of ~40 °C [71]. Sorbent tubes (Tenax TA/Carbograph 5TD, Marks International Ltd., Bridgend, UK) were used to collect samples at a flow rate of 200 sccm for 5 min every 30 min during the experiments. Prior to each sampling event, 20 L of air was drawn from the chamber using the same sampling line to ensure that equilibrium was reached between the gas phase and the tube’s inner wall, thereby minimizing the impact of sorption loss on the tube [35]. The detailed description of tube sample analysis has been described in our previous studies [44,45,72]. Briefly, sample tubes were desorbed using a thermal desorber (TD-100, Marks International Ltd., Bridgend, UK) at 320 °C for 20 min. All desorbed compounds were cryogenically trapped at −10°C and then rapidly desorbed again at 325 °C to transfer the IVOCs into the gas chromatography/mass selective detector (7890 GC/5975 MSD, Agilent Technologies, Santa Clara, CA, USA) equipped with a capillary column (HP-5MS, 30 m × 0.25 mm × 0.25 µm, Agilent Technologies, Santa Clara, CA, USA). The GC oven temperature was initially at 40 °C for 3 min, then raised at 5 °C min⁻¹ to 290 °C and held for 10 min. The MSD was operated in the scan mode with electron impact (EI) ionization at an ionization energy of 70 eV.

2.4. Gas–Wall Transfer

The fundamental framework of the mass transfer of organic vapor between gas and wall includes two processes: deposition from gas to wall and desorption from wall to gas. The rates of these reversible processes can be expressed with Equations (1) and (2), respectively [28,35,73,74]:

$$r_{g2w}=k_{g2w}\times O{C}_g={\displaystyle \frac{A}{V}}{({\displaystyle \frac{\pi }{2\sqrt{(k_e{D}_g)}}}+{\displaystyle \frac{4}{{\alpha }_w\omega }})}^{-1}\times O{C}_g$$

(1)

$$r_{w2g}=k_{w2g}\times O{C}_s={\displaystyle \frac{A}{V}}{({\displaystyle \frac{\pi }{2\sqrt{(k_e{D}_g)}}}+{\displaystyle \frac{4}{{\alpha }_w\omega }})}^{-1}\times {\displaystyle \frac{1}{K_w}}\times O{C}_s$$

(2)

where r_g2w is the transfer rate of OC from bulk gas to the wall, while r_w2g is the transfer rate from the wall to bulk gas. Both k_g2w and k_w2g denote the first-order rate constant (s⁻¹), which can be calculated from a range of parameters. OC_g represents the concentration of OC in the bulk gas, while OC_s denotes the concentration of OC on the surface of the wall. A (m²) is the surface area of the chamber, and V (m³) is the volume of the chamber. k_e (s⁻¹) is the eddy diffusion coefficient, which will be addressed in detail in Section 3.1 in order to determine its value in this study. D_g (m² s⁻¹) is gas molecular diffusivity, and for simplicity of the calculation, its value was set to 5 × 10⁻⁶ m² s⁻¹, which corresponds to the diffusion coefficients of n-dodecane and anthracene in air at 298 K and 1 atmosphere [75], without taking into account the effect of the temperature and compound properties on D_g. α_w (dimensionless) is the wall accommodation coefficient, which expresses the efficiency of the Teflon wall in trapping OC and determines that the concentration decrease rate is due to wall losses. ω (m s⁻¹) is the mean thermal speed of the gas molecules [40]. K_w is the ratio of k_g2w to k_w2g, which determines the magnitude of wall loss at equilibrium. According to gas–particle partitioning theory, K_w can be expressed as Equation (3) [32,74]:

$$K_w={\displaystyle \frac{k_{g2w}}{k_{w2g}}}={\displaystyle \frac{RT{C}_w}{M_w{\gamma }_{w}P°}}$$

(3)

where R is the gas constant (8.314 Pa m³ mol⁻¹ K⁻¹). T is the temperature in Kelvin (K). In this study, the Teflon wall is regarded as an organic aerosol with a certain mass concentration of C_w, thereby facilitating the calculation of the desorption rate of OC from the wall to the bulk gas. Therefore, C_w is the equivalent organic aerosol concentration (g/m³). M_w is the molecular mass of the Teflon wall and is assumed to be 200 g/mol in previous studies [76]. P° (Pa) is the saturation vapor pressure. In our study, the vapor pressure values of n-alkanes are obtained using the EVAPORATION model [77], and the values of PAHs are obtained using the modified grain method [78]. γ_w is the activity coefficient of OCs in the surface layer of Teflon, and it is assumed to be 1 [35,73]. Direct measurement of the γ_w is challenging since γ_w is typically a function of the mole fractions of all substances in aerosol, as well as temperature and pressure [40]. For wall loss characterization of organic vapor, several studies have already calculated compound-dependent C_w values [34,35,36,73], while there are also researchers using the thickness of the Teflon surface layer, density, and chamber size to estimate a constant C_w for a specific chamber and fitting γ_w as an indicator of wall loss magnitude [32,74]. Whichever method is employed, the effect of the interaction between specific organic compounds and the Teflon “solution” on the degree of wall loss must be taken into account. In our study, we have acquired C_w for various compounds, in consideration of the inherent physical meaning of γ_w and the discrepancy of Teflon material between chambers. For obtaining C_w and α_w, Equations (1) and (2) will be converted into integral form to facilitate fitting, and they are shown in Section 3.2 in detail.

3. Results and Discussion

3.1. Fitting of k_e

The values of k_e are various in different chambers and can be estimated from measured particle size deposition rates [14,28,34]. In our chamber, we assume that the Teflon wall resembles organic aerosols, and therefore, k_e was determined by minimizing the difference between the measured and theoretical distributions of particle loss rate constants with respect to particle sizes, as illustrated in Figure 1. The theoretical loss rate is widely recognized to be primarily influenced by three processes: eddy diffusion, electrostatic force, and gravitational settling. The impact of electrostatic force mainly arises from the interaction between the charge carried by the surface of particles and the electric field generated by the Teflon wall, particularly within the particle size range of 100–1000 nm [28,79]. However, given that the measured data focus on smaller particles and we do not observe a significant effect from the electrostatic force in this study, we do not consider their influence here [80]. Equation (4) below is used for quantifying the particles’ loss rate coefficient in the chamber [17,18]:

$$\alpha ={\displaystyle \frac{1}{L}}\left[{\displaystyle \frac{8\sqrt{D{k}_e}}{\pi }}+V_s\coth \left({\displaystyle \frac{V_s\pi }{4\sqrt{D{k}_e}}}\right)\right]$$

(4)

where α (s⁻¹) represents the particles’ loss rate coefficient describing the first-order process. L (m) represents the side length of the chamber. D (m² s⁻¹) represents the Brownian diffusion coefficient, and V_s (m s⁻¹) represents the terminal settling velocity of the particles. Detailed calculation methods for both D and V_s can be found in Seinfeld and Pandis [40]. In order to obtain the optimal value of k_e, a customized variable J can be constructed with Equation (5), where α(D_p,i) (s⁻¹) represents the particles’ loss rate coefficient of a specific size bin i [19]:

$$J=\sum {\left(\alpha {\left(D_{p,i}\right)}_{observed}-\alpha {\left(D_{p,i}\right)}_{theoretical}\right)}^2$$

(5)

The MATLAB (v9.13.0(R2022b), The MathWorks Inc., Natick, MA, USA) function fminsearch is utilized to calculate the minimum value of the unconstrained function J using the derivative-free method. The best-fitting curve and the observed particle loss rate constants are compared in Figure 1 for the smallest value of J with the k_e = 1.6863 s⁻¹.

3.2. Fitting of C_w and α_w

The deposition and desorption of OC on the wall surface are mainly controlled by the processes represented by Equations (1) and (2), which are reversible one-order processes. Therefore, Equations (1) and (2) can be transformed into an integral form of Equation (6), which represents the concentration of OC in the gas phase as a function of time [8]. Using Equation (6) to fit the temporal concentration of measured IVOCs, we can determine the optimal values for intermediate parameters (i.e., k_g2w, k_w2g, and OC_∞,g). Consequently, we can then derive the final values of C_w and α_w:

$$OC{\left(t\right)}_{simulated}=\left(O{C}_0-O{C}_{\infty ,g}\right)e^{-\left(k_{g2w}+k_{w2g}\right)t}+O{C}_{\infty ,g}$$

(6)

where OC(t) represents the vapor concentration of OC at time point t, OC₀ indicates the initial vapor concentration, and OC_∞,g refers to the equilibrium OC concentration in bulk gas. In this manner, the optimal values of the parameters k_g2w + k_w2g and OC_∞,g in Equation (6) can be derived. Consequently, OC_∞,w, as OC concentration on the wall at equilibrium can be found in Equation (7):

$$O{C}_{\infty ,w}=O{C}_0-O{C}_{\infty ,g}$$

(7)

In practice, the K_w shown in Equation (3) is equivalent to the equilibrium ratio of OC between the wall phase and the bulk phase, as indicated in Equation (8). k_g2w and k_w2g can be calculated using Equations (3) and (6), C_w can be calculated using Equation (3), and α_w can be derived from k_g2w or k_w2g on the basis of Equations (1) and (2).

$$K_w={\displaystyle \frac{O{C}_{\infty ,w}}{O{C}_{\infty ,g}}}$$

(8)

Figure 2 illustrates the concentration variations and fitting curves of some studied species, spanning from n-tetradecane to n-nonadecane, along with three PAHs, such as fluorene, phenanthrene, and anthracene. It is noteworthy that n-dodecane, n-tridecane, naphthalene, 1-methyl-naphthalene, 2-methyl-naphthalene, acenaphthene, and acenaphthylene do not exhibit a clear decreasing trend due to their relatively high volatility. Wall loss for other species selected in our study was too small to observe (see Figure S2 in the Supporting Information); thus, fittings could not be carried out for these compounds to obtain reasonable curves and parameters. The ratios of OC_∞,g to OC₀ indicate the fraction of OC remaining in the gas phase at equilibrium, and its value significantly decreases with the reduced volatility in the n-alkane or PAH class. For n-nonadecane, the loss of its organic vapor to the wall can reach up to 75.7% upon achieving gas–wall partition equilibrium. Despite the higher extent of wall loss observed for n-nonadecane, achieving equilibrium between the gas phase and wall phase requires a significantly longer time, with a characteristic timescale τ_gw (equal to 1/(k_g2w + k_g2w)) reaching up to 7.88 h. In contrast, n-tetradecane, which exhibits minimal wall loss, has a τ_gw value of only 34.35 min.

Wall loss parameters calculated for these compounds are present in

Table 1. The wall loss parameters and volatility for IVOC species. C_w and α_w are determined by fitting based on measurements conducted in this study. α_w1 and α_w2 are calculated with equations proposed by Zhang et al. [73] and Huang et al. [74], respectively. Saturation mass concentration is an indicator of the volatility of different species, where the C° of n-alkanes is predicted from the EVAPORATION model [77], while the C° of PAHs is predicted by the modified grain method provided by EPI Suite™ software (v4.11, United States Environmental Protection Agency, Washington, DC, USA).

Compounds	Cw (mg/m3)	C° (μg/m3)	αw	αw1	αw2
n-Dodecane	/	1.83 × 106	/	/	/
n-Tridecane	/	6.42 × 105	/	/	/
n-Tetradecane	5.41	2.24 × 105	1.24 × 10−7	3.96 × 10−8	5.53 × 10−7
n-Pentadecane	3.87	7.89 × 104	1.53 × 10−7	4.73 × 10−8	1.10 × 10−6
n-Hexadecane	4.08	2.74 × 104	1.69 × 10−7	5.65 × 10−8	2.20 × 10−6
n-Heptadecane	3.84	9.71 × 103	7.38 × 10−7	6.74 × 10−8	4.35 × 10−6
n-Octadecane	2.31	3.39 × 103	4.67 × 10−7	8.05 × 10−8	8.67 × 10−6
n-Nonadecane	3.39	1.19 × 103	3.54 × 10−7	9.61 × 10−8	1.72 × 10−5
Naphthalene	/	2.78 × 105	/	/	/
1-Methyl-Naphthalene	/	2.82 × 105	/	/	/
2-Methyl-Naphthalene	/	2.63 × 105	/	/	/
Acenaphthene	/	1.07 × 104	/	/	/
Acenaphthylene	/	1.03 × 104	/	/	/
Fluorene	1.22	2.95 × 103	1.08 × 10−6	/	/
Phenanthrene	0.40	4.14 × 102	3.60 × 10−7	/	/
Anthracene	0.03	2.08 × 101	5.21 × 10−7	/	/

For species that have no discernible declining trend, their C_w, α_w, α_w1, and α_w2 are represented as /.

. Typically, the values of C_w for most n-alkanes are close to 2 mg/m³, which are close to the values reported by Matsunaga and Ziemann [35], but they exhibit a decreasing trend with decreasing volatility for three PAHs with an average of 0.55 mg/m³. As for α_w, their values for C₁₄–C₁₇ n-alkanes increase with decreasing volatility, consistent with findings by Matsunaga and Ziemann [35], as well as Huang et al. [74]. However, a downward trend in α_w is observed with further decreases in volatility within other n-alkanes. The average αw for alkanes and PAHs are 3.34 × 10⁻⁷ and 6.53 × 10⁻⁷, respectively. Based on the functional relationship between volatility and α_w proposed by Zhang et al. [73] and Huang et al. [74], we could calculate the n-alkanes’ accommodation coefficient in our experiments, denoted as α_w1 and α_w2, respectively. The range of α_w/α_w1 is 0.09–0.33, with an average of 0.25. The range of α_w/α_w2 is 4.45 to 48.62, with an average of 16.29.

Table 1 lists the volatility and accommodation coefficients of the organic compounds investigated in this study. The relationship between volatility and wall loss behavior is obvious within each OC class. For n-alkanes with higher volatility than n-tetradecane, and PAHs with higher volatility than fluorene, their wall loss can be negligible in the GIG-CAS chamber. Similar findings were reported by Matsunaga and Ziemann, who did not observe a significant decrease (a fraction of the concentration in the gas phase, OC_g/OC₀, decreased by <5%) in gas-phase concentration for n-alkanes, 1-alkenes, 2-alcohols, and 2-ketones with higher volatility than n-tetradecane, 1-tetradecene, 2-decanol, and 2-decanone, respectively. It can serve as a threshold to determine whether significant wall loss of a specific kind of organic vapor exists in our chamber. However, predicting the presence of significant wall loss based solely on volatility may be insufficient without considering the class difference in compounds. For instance, although n-heptadecane and acenaphthylene exhibit similar volatility, n-heptadecane can experience up to 32.3% loss of its total concentration to the Teflon wall, whereas acenaphthylene shows no significant trend in wall loss.

3.3. Comparison of C_w and K_w in Different Studies

The comparison of C_w within different studies is depicted in Figure 3a. Zhang et al. [73] and Kretchmer et al. [34] demonstrated that the C_w of oxidized products (with C° ranging from 10⁰ to 10⁶ μg/m³) from toluene, n-dodecane, α-pinene, isoprene, and C₆–C₁₂ 1-alcohol significantly increased with elevated volatility under both high and low NO_x conditions. Similar trends were observed for PAHs in our study and for C₁₄–C₁₉ alkanes in the research by He et al. [32]. However, there is no significant trend for C₁₄–C₁₉ alkanes in our study, with an average C_w of 3.82 ± 0.92 mg/m³, despite n-tetradecane exhibiting the highest volatility and C_w value. Furthermore, discrepancies in C_w for the same compound persist across different studies. The C_w values measured by He et al. [32] for C₁₄–C₁₉ alkanes are higher than those obtained from Matsunaga and Ziemann [35] and our work. This discrepancy may be attributed to the surface area concentration of the chamber (defined as the ratio of surface area to volume, m²/m³) used by He et al. [32], which is 7.26 m²/m³. While for other chambers, the ratio of surface area to volume was approximately 2 to 3 m²/m³. At equivalent levels of volatility, there are also discrepancies in C_w. As shown in Figure 3a, in the lower volatility range (C° < 10² μg/m³), Krechmer et al. [34] reported higher C_w compared to those obtained by Zhang et al. [73], whereas in the higher volatility range (C° > 10⁴ μg/m³), the C_w measured from the Ziemann Group and our work are lower than those of most compounds reported by Zhang et al. [73]. It suggests that structural differences among compounds may contribute to various degrees of wall loss even when volatility is consistent.

The relationships between the logarithm of gas–particle partitioning coefficient (ln K_p) and that of subcooled saturation vapor pressure (ln P_L°) are frequently utilized to assess the partitioning behavior of organic contaminants in field studies in the atmosphere [82,83]. Goss et al. [84] proposed that when the slope between ln K_p and ln P_L° of the pollutant exceeds −0.6, the gas–particle partitioning process is predominantly governed by chemical adsorption, indicating involvement with a “solvent” possessing high cohesive energy. Conversely, when the slope falls below −1, the partitioning process is primarily governed by physical adsorption. As depicted in Figure 3b, the slope of n-alkanes is between −0.6 and −1 in our study. The slopes obtained from fitting the data of Matsunaga and Zieman [35] and He et al. [32] exceed −0.6, suggesting that the depletion of n-alkanes to the Teflon wall is primarily governed by adsorption and adheres to Raoult’s law, thus aligning with the conclusions drawn by Matsunaga. The deviation of the ln K_w–ln P_L° slope from the ideal value of −1 towards shallower slopes is attributed to the relationship between activity coefficients and volatility, representing changes in molecular structure [84]. Particularly, the presence of polar or low-polar molecules like water or fluorinated ethylene propylene (FEP) as the solvent renders non-polar alkane molecules “uncomfortable”, resulting in an increase in their activity coefficients with longer alkane chains, leading to decreased solubility. Hence, the diminished slope of ln K–ln P_L° observed in this study could stem from the retention of non-polar molecules within the Teflon film over the chamber’s historical usage, thereby attenuating the overall polarity of FEP and shifting the slope between ln K_w and ln P_L° in a “steeper” direction.

4. Conclusions

We investigated the wall loss behavior of eight n-alkanes and eight PAHs that are classified as IVOCs within an indoor chamber situated at GIG-CAS, and we procured pivotal parameters suitable for this chamber that delineate the wall loss of these species, encompassing the eddy diffusion coefficient, wall accommodation coefficient, and equivalent organic aerosol concentration. In addition, the findings revealed that α_w spanned from 1.24 × 10⁻⁷ to 1.01 × 10⁻⁶, and C_w varied from 0.02 to 5.41 mg/m³ for C₁₄ to C₁₉ n-alkanes and three PAHs (fluorene, phenanthrene, and anthracene). It was noteworthy that the C_w of PAHs demonstrated a distinct escalation with elevated volatility. Nevertheless, the C_w of C₁₄–C₁₉ n-alkanes remained relatively stable. The α_w for C₁₄–C₁₇ increases with decreasing volatility, but a downward trend is observed with further decreases in volatility within the other n-alkanes. The relationship between K_w and P_L° suggested that the gas–wall partitioning mechanism in this study might diverge from the previous literature, but the deviation of the slope may originate from the influence of non-polar compounds within the Teflon film on “solution” activity. Here, we only focused on hydrocarbon–IVOC studies at dry conditions, for other IVOC species with different structures; their wall loss properties may be more complex due to the presence of different groups and a higher number of oxygen atoms. It is worth noting that, apart from the nature of IVOCs, the wall losses of IVOCs are impacted by factors such as temperature, RH, and the chamber’s wall material, as well as the continued use of the chamber. Consequently, parameters obtained for one chamber may not apply to another, and those derived under specific environmental conditions may not be applicable to others. In this study, we demonstrate the method of obtaining these parameters through experiments. The parameters presented in this study, derived at ~25C and RH < 10%, are not necessarily suitable for simulations under different conditions. Therefore, parameters should be determined under identical conditions to those intended for real chamber simulation. Once the conditions for the chamber simulation are defined, repeating the wall loss experiments under those conditions would enable a more accurate assessment of the parameter uncertainties.