Influence of Field Sampling Methods on Measuring Volatile Organic Compounds in a Swine Facility Using SUMMA Canisters

Abstract

Volatile organic compounds (VOCs) play a crucial role in emission control, being one of the most important sources of odor while also serving as significant precursors to secondary organic aerosols and ozone formation. Appropriate sampling methods are essential for accurately assessing the concentration and composition of VOCs within swine barns. In this study, the effects of both passive air sampling and active air sampling on VOCs were evaluated, and the influence of storage time on the VOC stability in sampling canisters for both methods was investigated. SUMMA canisters, which are electropolished and passivated with silanization, offer excellent corrosion protection and resistance to high pressure and temperature and were used in this study. The predominant component categories prevailing within the pig house were found to be oxygenated VOCs (OVOCs) and volatile sulfur compounds (VSCs), with ethanol emerging as the most abundant component of VOCs detected. Notably, the statistical analysis results revealed no significant differences between passive and active sampling regarding the impact of storage time on substance concentration. Changes in canister pressure also did not significantly affect substance stability. The results showed that the C2–C3 compounds remained relatively stable, especially within 3 days, with recoveries above 80% within 20 days. Methyl sulfide, dimethyl disulfide, and ethanol were more stable within the first week, but their recoveries significantly dropped by day 20, with methyl sulfide and dimethyl disulfide at 62.3% and 65.3%, respectively. This study contributes to the development of a foundation for selecting appropriate VOC sampling methods in swine facilities for conducting a rational analysis of VOC samples.

1. Introduction

Odor pollution has become a major challenge for ecological improvement [1]. Intensive swine feeding operations are notably associated with unpleasant odors, prompting numerous complaints from nearby residents [2,3,4,5]. The predominant constituents contributing to the malodorous nature of swine feeding operations encompass ammonia (NH₃), hydrogen sulfide (H₂S), and volatile organic compounds (VOCs). While the issues of odor pollution associated with H₂S and NH₃ have attracted considerable attention, the detection and mitigation of VOCs pose formidable challenges, characterized by their vast variety, low concentrations, and high associated costs, thus resulting in a paucity of investigations in this domain [6,7,8].

VOCs are a key contributor to these odors, not only polluting the air with unpleasant smells but also posing risks to human and animal health, such as headaches, allergic skin, nose and throat soreness, fatigue, nausea, dizziness, and other symptoms [9]. Moreover, some VOCs can interact with highly reactive components such as hydroxyl radicals to form secondary organic aerosols (SOAs), a major component of fine particulate matter (PM_2.5), and ozone (O₃), which possesses strong oxidizing properties, further jeopardizing human health [10,11].

Swine operations generate a diverse mix of VOCs. Presently, researchers have identified over 500 distinct VOCs emanating from swine farms, encompassing a wide spectrum of chemical classes such as alcohols, esters, ketones, alkenes, aromatics, phenols, volatile fatty acids (VFAs) [12], volatile sulfur compounds (VSCs), and various other compounds [9,13,14]. Interestingly, some of the VOCs released by swine farms have a low olfactory threshold, which means that scents can be generated even at low concentrations, significantly contributing to malodors [15,16]. Studies indicate that 1,3-butadiene, toluene, and benzene pose cancer risks [17], while alkenes and aromatics are primary contributors to the formation of O₃ and SOAs, respectively [18]. Additionally, VSCs and p-cresols have the most significant impact on odor [19]. China’s emission standards for offensive odors, as stipulated in GB 14554-93, categorize specific volatile organic compounds (VOCs) as malodor-controlled substances. This includes compounds such as trimethylamine, methanethiol, methyl sulfide, dimethyl disulfide, carbon disulfide, and styrene [20]. Furthermore, in the “National Advanced Pollution Prevention and Control Technology Catalog” released by China’s Ministry of Environmental Protection in 2016, VOC pollution prevention and control is listed as one of the five major key emission reduction projects.

The detection of VOCs primarily involves three key steps: sample collection, sample pretreatment, and sample analysis [21]. Typically, VOCs are gathered through the utilization of sorbent tubes containing adsorbents or through the use of containers such as sampling bags and sampling canisters [22]. Sampling bags, such as Tedlar bags, Teflon bags, and Fluode bags, are cost-effective, versatile, and easy to use. However, they are prone to sample volatilization and contamination [23,24,25]. Sorbent tubes, typically made of glass or stainless steel, contain adsorbents such as activated carbon or molecular sieves. They offer high sensitivity for specific VOCs but have a short lifespan and are limited in their ability to detect a wide range of VOCs [26,27,28,29,30]. For effective VOC sampling, the use of SUMMA canisters has proven to be a reliable method. These canisters feature polished and inert surfaces and joints, minimizing the adsorption and chemical reactions of chemically active substances on the inner walls. This design ensures the accurate preservation of samples, guaranteeing the reliability of the collected data [31]. Despite its advantages, the SUMMA canister has several drawbacks. Firstly, it is expensive to manufacture and maintain, and its large size and weight make it cumbersome to transport and operate. Proper storage and transport are crucial to avoid contamination or damage, and the canister must be thoroughly cleaned and calibrated. Additionally, in low concentration environments, the detection sensitivity of the SUMMA canister may be inferior to other methods, leading to less accurate measurements [32,33,34].

SUMMA canister sampling is divided into two types: passive and active sampling [35,36,37]. Active sampling uses a sampling pump and flow controller to deliver the sample into the canister, allowing for the collection of large volumes of air, often up to twice the canister’s capacity. However, the sampling pump must be certified as free of analyte residues or losses to ensure sample integrity. Additionally, the need for a power supply can make active sampling challenging in remote areas. In contrast, passive sampling mainly involves transient sampling and constant flow sampling. Transient sampling utilizes the vacuum inside the canister to draw in an air sample over a short period. This method is simple to operate, requires no power supply, and is suitable for field sampling in various locations [38,39,40]. The composition of VOCs in pigsties is complex, and their stability within SUMMA canisters has not been studied. Therefore, it is crucial to comprehensively assess the feasibility and limitations of these two sampling methods in monitoring pig farm environments. Comparing and analyzing the storage stability of VOCs in SUMMA canisters is essential for ensuring the scientific accuracy and sustainability of VOC detection in pig farms.

Both SUMMA canister sampling methods utilize the same type of container, are suitable for ambient air sampling, and share an identical cleaning process. These methods do not alter the chemical properties of the samples collected. Stringent quality control is essential for both methods, involving the use of calibrated flow controllers, leak detection procedures, and the assurance of the cleanliness of the canisters. However, there are differences in the sample volume and sampling duration between the two methods. Active sampling typically collects larger sample volumes and requires more time compared to the instantaneous sampling associated with passive sampling.

VOCs are significant pollutants in pig farm environments, directly impacting both environmental quality and the health of humans and animals, and pose potential health risks to people. Choosing the appropriate sampling method is crucial for ensuring the effectiveness and accuracy of VOC detection. SUMMA canisters have unique characteristics and significant advantages in monitoring and analyzing environmental VOCs. They effectively prevent sample contamination and external interference, and samples stored in SUMMA canisters remain relatively stable. The complex VOC composition in pigsties has not yet been studied for its stability within SUMMA canisters. Therefore, comprehensively evaluating the feasibility and limitations of these two sampling methods in pig farm environment monitoring and comparing the storage stability of VOCs in SUMMA canisters are of great significance for the scientific and sustainable detection of VOCs in pig farms.

The primary objectives of this study are to comprehensively investigate variations in VOC concentrations, specifically focusing on the influence of different canister sampling methods employed during real time in a pig house, and to evaluate the impact of storage time on VOC component concentrations by analyzing various canister sampling methods. By examining these data, we aim to enhance the precision and robustness of VOC analysis in livestock environments and to provide guidelines to mitigate the impact of VOC emissions and foster sustainable practices in livestock production.

2. Materials and Methods

2.1. Study Site Description

The swine facility encompasses 16 fattening pig houses, 4 nursery houses, and a manure processing area (Figure 1). For this study, the focus was on a specific fattening barn (410 m × 160 m), housing approximately 600 pigs, and equipped with a total of 6 fans: 1 inverter fan and 5 temperature-controlled fans. During the experiment, only the inverter fan was operating, while the other fans remained inactive. The pigs received automatic feed and water delivery, and a mechanical manure removal system, featuring a flat-shape design, was employed to manage slurry on the floor. The selected sampling site was positioned in the middle of the fattening pig house, approximately 1.0 m above the floor.

2.2. Sampling Method

The samples were collected using 6 L SUMMA canisters (Entech, Simi Valley, CA, USA) with a maximum pressure capacity of 275 kPa. Following EPA Method TO-15, the canisters were cleaned to achieve a vacuum below 0.03 kPa and subsequently evacuated to a sub-ambient pressure lower than 0.006 kPa prior to sampling [41]. The cleaning process involved evacuating the canisters 3–6 times using an automated canister cleaning system (Entech 3100A, Simi Valley, CA, USA). This system efficiently loaded and pumped high-purity chromatographic nitrogen using both a rough pump and a molecular drag pump to remove any contaminants present in the canisters. After cleaning, the container valve was tightly closed and clearly labelled for sampling.

Air samples were collected using two methods: passive air sampling, which relies on the pressure differential, and active air sampling, which uses a positive pressure sampler (Figure 2). The passive sampling period was 10 min. Following the completion of sampling, the inside of the sampling canister was at atmospheric pressure, with a pressure of 100 kPa. The active air sampling period was 20 min, and the pressure of the gas collected in the canister could be 241 kPa. During sampling, the air temperature and relative humidity were monitored using a portable hot-wire anemometer (VelociCalc^® Multi-Function Ventilation Meter 9565-A, TSI, Shoreview, MN, USA) in the pig house. Subsequently, the collected samples were transported back to the laboratory for further analysis.

2.3. Analysis Method

2.3.1. Standard Curve Creation

Before analyzing the samples, a standard curve was created to establish the relationship between the concentration of standard gases and the chromatographic peak area. The standard gases used in the analysis contained 112 compositions from US EPA PAMS, US EPA TO-15, and National Standard Gas in China for organic sulfur. The standard gas was diluted to 20 ppb and 2.0 ppb using a gas configurator (Entech 4700, Simi Valley, CA, USA), and six gradients (0.5 ppb, 1.0 ppb, 2.0 ppb, 5.0 ppb, 7.5 ppb, 10 ppb) of the standard samples were measured by the sample analyzing system. The retention time (RT), qualitative ions (Ion1, Ion2, Ion3), CAS NO., standard curve correlation factor (R²), and relative standard deviation (RSD) of the standards are shown in

Table A1. Standard curve-related parameters.

Number	Standard Compound	RT	Ion1	Ion2	Ion3	CAS NO.	R2	RSD
1	Ethylene	12.456	28	27	26	74-85-1	0.999	0.052
2	Acetylene	12.710	26	25	24	74-86-2	0.999	0.053
3	Ethane	13.864	28	27	29	74-84-0	0.999	0.046
4	Propylene	21.325	41	42	39	115-07-1	0.999	0.051
5	Propane	21.987	44	43	39	74-98-6	0.999	0.050
6	Bromochloromethane (is)	28.744	130	128	93	74-97-5	-	0.090
7	Dichlorodifluoromethane	10.620	85	87	101	75-71-8	0.999	0.010
8	Methyl chloride	11.817	50	52	-	74-87-3	0.999	0.020
9	Isobutane	12.761	43	42	41	75-28-5	0.999	0.049
10	1,2-Dichlorotetrafluoroethane	12.637	85	135	137	76-14-2	0.999	0.012
11	Vinyl chloride	13.439	62	64	63	75-01-4	0.998	0.021
12	1-Butylene	14.179	41	39	56	106-98-9	0.997	0.004
13	1,3-Butadiene	14.386	54	53	39	106-99-0	0.998	0.021
14	n-Butane	14.749	58	42	41	106-97-8	0.999	0.007
15	trans-2-Butene	15.388	56	41	50	624-64-6	0.997	0.036
16	Bromomethane	16.009	94	96	93	74-83-9	0.998	0.008
17	cis-2-Butene	16.344	56	55	41	590-18-1	0.997	0.026
18	Chloroethane	16.995	64	66	49	75-00-3	0.999	0.029
19	Ethyl alcohol	17.505	45	46	-	64-17-5	0.987	0.075
20	Acrolein	19.147	56	55	38	107-02-8	0.985	0.013
21	Isopentane	19.937	57	43	42	78-78-4	0.997	0.004
22	Acetone	19.649	43	58	-	67-64-1	0.999	0.103
23	Trichlorofluoromethane	20.504	101	103	-	75-69-4	0.999	0.007
24	1-Pentene	20.981	42	55	70	109-67-1	0.996	0.040
25	Propan-2-ol	20.571	45	43	-	67-63-0	0.990	0.015
26	n-Pentane	21.821	43	42	41	109-66-0	0.997	0.019
27	2-Methyl-1,3-Butadiene	22.181	55	70	39	78-79-5	0.996	0.077
28	trans-2-Pentene	22.338	67	53	39	646-04-8	0.997	0.041
29	Vinylidene chloride	22.766	61	96	98	75-35-4	0.999	0.055
30	cis-2-Pentene	22.920	55	70	42	627-20-3	0.996	0.078
31	Dichloromethane	23.105	49	86	84	75-09-2	0.999	0.029
32	Carbon disulfide	24.170	76	78	77	75-15-0	0.998	0.083
33	1,1,2-Trichlorotrifluoroethane	23.921	151	101	-	76-13-1	0.999	0.017
34	2,2-Dimethyl butane	24.460	57	71	43	75-83-2	0.997	0.006
35	trans-Dichloroethylene	25.812	96	98	61	156-60-5	0.997	0.080
36	1,1-Dichloroethane	26.337	63	65	98	75-34-3	0.999	0.018
37	Methyl tert-Butyl Ether	26.360	73	57	-	1634-04-4	0.995	0.078
38	Cyclopentane	26.551	42	55	70	287-92-3	0.999	0.072
39	2,3-Dimethylbutane	26.535	42	43	71	79-29-8	0.998	0.066
40	2-Methylpentane	26.735	43	71	55	107-83-5	0.998	0.003
41	Ethenyl ethanoate	26.512	43	86	-	108-05-4	0.995	0.027
42	2-Butanone	27.138	43	72	57	78-93-3	0.995	0.057
43	3-Methylpentane	27.715	57	56	41	96-14-0	0.995	0.036
44	1-Hexene	28.032	56	41	84	592-41-6	0.997	0.029
45	1,2-cis-Dichloroethene	28.336	96	98	61	156-59-2	0.996	0.096
46	n-Hexane	28.750	57	41	86	110-54-3	0.998	0.008
47	Ethyl acetate	28.563	43	61	45	141-78-6	0.996	0.093
48	Trichloromethane	28.992	83	85	-	67-66-3	0.999	0.016
49	Tetrahydrofuran	29.855	40	71	72	109-99-9	0.996	0.116
50	Methylcyclopentane	30.699	57	43	85	96-37-7	0.998	0.066
51	2,4-Dimethylpentane	30.665	56	69	41	108-08-7	0.997	0.100
52	1,2-Dichloroethane	30.641	62	64	49	107-06-2	0.999	0.141
53	1,1,1-Trichloroethane	31.233	97	61	117	71-55-6	0.999	0.073
54	1,4-Difluorobenzene (is)	32.690	114	63	88	540-36-3	-	0.177
55	Benzene	32.210	78	77	52	71-43-2	1.000	0.141
56	Carbontetrachloride	32.528	117	119	121	56-23-5	0.999	0.114
57	Cyclohexane	32.814	56	69	84	110-82-7	1.000	0.093
58	2-Methylhexane	32.813	43	57	85	591-76-4	0.999	0.078
59	2,3-Dimethylpentane	33.070	56	43	71	565-59-3	1.000	0.065
60	3-Methylhexane	33.341	57	43	70	589-34-4	1.000	0.069
61	1,2-Dichloropropane	33.821	63	62	76	78-87-5	1.000	0.067
62	2,2,4-Trimethylpentane	34.231	57	56	41	540-84-1	0.999	0.085
63	1,4-Dioxane	34.108	88	58	-	123-91-1	0.998	0.084
64	Bromodichloromethane	34.190	83	129	47	75-27-4	0.999	0.061
65	Trichloroethylene	34.257	130	132	95	79-01-6	0.999	0.086
66	Methyl methacrylate	34.268	69	41	39	80-62-6	0.998	0.011
67	Heptane	34.580	43	57	71	142-82-5	0.999	0.044
68	Methyl isobutyl ketone	35.756	43	58	85	108-10-1	0.999	0.072
69	cis-1,3-Dichloropropylene	35.863	75	110	-	10061-01-5	1.000	0.012
70	Methylcyclohexane	36.270	83	98	55	108-87-2	1.000	0.065
71	trans-1,3-Dichloropropene	36.831	75	110	39	10061-02-6	0.998	0.051
72	1,1,2-Trichloroethane	37.316	97	83	61	79-00-5	0.999	0.088
73	2,3,4-Trimethylpentane	37.506	43	71	55	565-75-3	1.000	0.090
74	Toluene	37.931	57	43	70	108-88-3	0.999	0.005
75	2-Methylheptane	37.960	91	92	-	592-27-8	0.999	0.060
76	2-Hexanone	38.092	58	100	-	591-78-6	0.999	0.050
77	3-Methylheptane	39.529	85	57	43	589-81-1	1.000	0.045
78	Chlorodibromomethane	38.896	129	127	-	124-48-1	1.000	0.060
79	1,2-Dibromoethane	39.453	107	109	-	106-93-4	0.998	0.064
80	Octane	39.529	85	71	57	111-65-9	1.000	0.084
81	Tetrachloroethene	40.302	166	131	94	127-18-4	0.998	0.081
82	Chlorobenzene-d5 (is)	41.652	117	82	119	3114-55-4	-	0.100
83	Chlorobenzene	41.753	112	77	114	108-90-7	0.999	0.033
84	Ethylbenzene	42.413	91	106	-	100-41-4	0.998	0.042
85	(o + p)-Xylene	42.761	91	106	104	108-38-3	0.998	0.101
86	Tribromomethane	43.248	173	171	175	75-25-2	0.999	0.063
87	Styrene	43.600	104	78	51	100-42-5	0.998	0.096
88	1,2-Dimethylbenzene	43.873	91	106	104	95-47-6	0.998	0.071
89	1,1,2,2-Tetrachloroethane	43.832	83	95	131	79-34-5	0.999	0.043
90	n-Nonane	43.891	91	106	-	111-84-2	0.999	0.063
91	4-Bromofluorobenzene (is)	44.941	174	176	-	460-00-4	-	0.081
92	Cumene	45.144	105	120	-	98-82-8	1.000	0.085
93	n-Propylbenzene	46.376	91	120	-	103-65-1	1.000	0.046
94	3-Ethyltoluene	46.579	105	120	-	620-14-4	1.000	0.081
95	4-Ethyltoluene	46.706	105	120	91	622-96-8	0.998	0.096
96	Mesitylene	46.851	105	120	77	108-67-8	0.999	0.019
97	2-Ethyltoluene	47.451	105	120		611-14-3	1.000	0.064
98	1,2,4-Trimethylbenzene	47.996	105	120	77	95-63-6	0.999	0.063
99	Decane	47.834	105	120	-	124-18-5	0.998	0.055
100	Benzyl chloride	48.492	91	126	-	100-44-7	0.999	0.045
101	1,3-Dichlorbenzene	48.621	146	111	148	541-73-1	0.999	0.078
102	1,4-Dichlorobenzene	48.784	146	111	148	106-46-7	0.999	0.071
103	1,2,3-Trimethylbenzene	49.249	105	120	-	526-73-8	0.998	0.078
104	1,2-Dichlorobenzene	49.830	146	111	148	95-50-1	0.999	0.003
105	1,3-Diethylbenzene	49.957	119	105	134	141-93-5	0.999	0.039
106	p-Diethyl benzene	50.272	119	105	134	105-05-5	0.999	0.057
107	n-Hendecane	51.599	57	71	156	1120-21-4	0.999	0.092
108	1,2,4-Trichlorobenzene	56.085	180	145	182	120-82-1	0.998	0.052
109	Dodecane	55.760	57	71	120	112-40-3	0.999	0.053
110	Naphthalene	56.730	128	64	-	465-73-6	0.995	0.046
111	Hexachloro-1,3-butadiene	58.169	225	190	118	87-68-3	0.999	0.051
112	Methanethiol	15.314	47	48	45	74-93-1	0.994	0.029
113	Ethyl mercaptan	21.358	62	47	45	75-08-1	0.989	0.042
114	Dimethyl sulfide	22.571	62	47	45	75-18-3	0.998	0.061
115	Methylthioethane	29.032	61	76	48	624-89-5	0.999	0.042
116	Thiophene	32.513	84	58	45	110-02-1	1.000	0.036
117	Diethyl sulfide	34.141	75	90	61	352-93-2	0.998	0.039
118	Dimethyl Disulfide	36.567	94	79	45	624-92-0	0.995	0.031

. To calibrate the system, four known VOC concentrations were used as internal standards for samples: bromochloromethane, 1,4-difluorobenzene, chlorobenzene-d5, and 4-bromofluorobenzene.

2.3.2. Instrument Analysis

The samples collected in pressurized canisters at the swine houses were transported back to the laboratory of the Zhejiang Ecological and Environmental Monitoring Center for GC-MS analysis. The canister samples were concentrated using a pre-concentrator (Entech 7200A Instruments Inc., Simi Valley, CA, USA) and then were analyzed on a gas chromatography-mass selective detector/flame ionization detector (GC-MS/FID, Agilent 7890B GC/5977B MS/FID, Wilmington, NC, USA). Flame ionization detector (FID) is used for C2–C3 target compounds, and mass selective detector (MS) is used for the rest of the target compounds [42,43,44,45].

The pre-concentrator has a three-stage trapping system to enhance the effect of separation that can remove the H₂O, CO₂, and other impurities. Firstly, 400 mL air samples in each canister and 40 mL of standard samples at 100 ppb concentration were injected into the cool trap Module 1 (M1) of the pre-concentrator through the autosampler. H₂O and inert gases, such as O₂ and N₂, can be removed through M1. VOCs were then transferred to cool trap Module 2 (M2), which contains Tenax to capture VOCs as well as to remove CO₂ and other air components. Lastly, VOCs were re-focused in the transfer line of cool trap Module 3 (M3) to make the chromatographic peak sharper. The operational procedures of the three-stage trapping system are presented in

Table 1. The operational procedures of the three-stage trapping system.

Zone	Trap Temperature (°C)	Preheat (°C)	Transfer (°C)		Precool (°C)	Inject (°C)	Bakeout (°C)	Transfer Line (°C)
			M1→M2	M2→M3
M1	−40	10	10				120
M2	−100	−40	−40	235			220
M3				−180	−185	70	110
Sample								20
GC								110

.

VOCs were transferred from M3 to the GC-MS/FID for identification and quantification. The samples were divided by GC, and then the divided portion was analyzed by MS, resulting in more precise detection. FID was used to quantify C2–C3 hydrocarbons containing ethylene, acetylene, ethane, propylene, and propane, while MS was used to quantify other compounds. In GC-MS analysis, different detectors are used for different compounds based on their sensitivity, selectivity, and applicability. Specifically, C2–C3 compounds typically use FID, while other substances are analyzed with MS. FID is highly sensitive to small hydrocarbon molecules such as C2–C3, producing strong signals with low detection limits, thus allowing the accurate and sensitive detection of these small hydrocarbons. Although MS has high sensitivity, it is less responsive to small hydrocarbons compared to FID. MS is better suited for analyzing complex mixtures and compounds with characteristic fragmentation patterns. By providing structural information through mass spectra, MS helps distinguish and identify components in complex mixtures. Both the calibrations of C2–C3 compounds detected by FID and other compounds detected by MS are performed using standards. MS calibration involves using standards of known mass and concentration to calibrate the mass axis and response factor of the mass spectrometer. FID typically uses a standard gas containing known concentrations of C2–C3 compounds for calibration. In both MS and FID, the signal intensity of the standards with known concentrations is measured to establish a linear relationship between the concentration and signal intensity for quantitative analysis.

The GC oven temperature was originally set to 25 °C for 12 min before increasing to 250 °C for 10 min and operating for 59 min. The inlet of GC was 150 °C, and the total flow rate was 34.5 mL min⁻¹. The operating parameters for the FID are as follows: heater temperature of 250 °C, air flow of 400 mL min⁻¹, hydrogen gas flow of 30 mL min⁻¹, and tail gas flow by He of 25 mL min⁻¹. The MS was run in full scan mode, and all scanned data was analyzed using MSD chemstation software. To assure continuous usage of the samples, the pressure in the tank was kept constant by utilizing high-purity nitrogen gas before each use, and the following formula was used to calibrate the dilution ratio.

$$C_x={\displaystyle \frac{A_x\times I_{is}\times V_{is}\times M\times f}{A_{is}\times V_x\times \overline{RF}\times V_m}}$$

(1)

Equation (1): Used to determine the target concentration of the sample, where C_x is the concentration of target sample x, μg m⁻³; A_x is the peak area of the target sample; A_is is the peak area of the standard sample; I_is is the mole fraction of the standard sample, nmol mol⁻¹; V_is and V_x are the injection volume of the standard sample and target sample, respectively; M is the molar mass of the target sample, g mol⁻¹; V_m is the molar volume of the gas, L mol⁻¹; f is the dilution ratio of the target sample; and RF is the average relative response factor of the calibration curve.

3. Results and Discussion

3.1. Main Compositions of VOCs at Swine House

In the experimental swine house, 62 VOCs were detected using quantification analysis, including 14 alkanes, 6 olefins, 17 halogenated hydrocarbons, 15 aromatic hydrocarbons, 7 oxygenated VOCs (OVOCs), and 3 volatile sulfur compounds (VSCs). All of these compounds were determined by retention time and fragment ions and were confirmed by GC-MS. The temperature and humidity in the swine house were 26.1 °C and 70%, respectively.

Both the passive air sampling and active air sampling methods revealed OVOCs as the most abundant compounds among all categories of VOCs, constituting 86.5% and 77.9% of the total VOC concentration, respectively. This finding aligns with Blunden’s observations, indicating that OVOCs are the predominant compounds in swine houses [41]. The primary cause of this condition is the high concentration of ethanol, which makes up more than 85% of OVOCs.

Figure 3 depicts the VOC concentration ranges observed in this study. In the active air sampling method, the main VOC categories collected were OVOCs, alkanes, and halogenated hydrocarbons. In the passive air sampling method, the main VOC categories collected were OVOCs, VSCs, and alkanes. Notably, the variation in sampling methods had a significant effect on the concentrations of OVOCs and VSCs, while it had no significant impact on other VOCs (Figure 4). This discrepancy primarily arises from variations in the ethanol concentration, constituting more than half of the total concentration. The variance in the VSC and OVOC concentrations between the sampling methods may be attributed to the wall effect adsorption of the sampling pump during positive pressure sampling, as well as the potential differences in flow rates between the two methods [46,47,48].

In this study, the results indicated that ethanol, acetone, methyl sulfide, dimethyl disulfide, isobutene, dichloromethane, 2-butanone, toluene, and vinyl acetate were the main compounds present in the swine house. Rumsey [49] conducted a similar study using GC-MS analysis with samples collected in SUMMA canisters, and they identified methanol, ethanol, acetaldehyde, acetone, 2,3-butanedione, and p-cresol as the primary VOCs in swine barns. It is noteworthy that all of these compounds, except methanol, were also found in our study.

In addition to the compounds mentioned above, eight VOCs were not included in the standard samples. These VOCs were identified through qualitative analysis of the chromatograms and compared with the NIST14 library, a standard reference database developed by the National Institute of Standards and Technology. The compound names and emissions of these VOCs are presented in

Table 2. The compounds and concentrations of qualitatively evaluated VOCs.

Compound	Concentration (μg m−3)
	Passive	Active
trimethylamine	67.44 1 (9.76) 2	7.51 (1.32)
methyl acetate	5.36 (0.31)	4.85 (0.64)
2,3-butanedione	5.34 (1.02)	3.23 (0.27)
1-butanol	5.13 (0.97)	2.23 (0.46)
butyl acetate	0.89 (0.04)	0.79 (0.09)
benzaldehyde	1.87 (0.05)	1.39 (0.12)
dimethyl trisulfide	7.73 (2.13)	0.58 (0.09)
p-cresol	6.65 (1.04)	0.87 (0.23)

¹ Mean value. ² Standard deviation.

.

Among the VOCs identified, dimethyl trisulfide and p-cresol, known as major sources of odor, exhibited different concentrations in the two sampling methods. The passive air sampling method was found to be more effective in capturing odor-causing compounds, including methanethiol, methyl sulfide, dimethyl disulfide, dimethyl trisulfide, and p-cresol. Passive sampling methods eliminate the need for mechanical pumps or other power equipment, reducing the potential for interference with the sampling process and loss of compounds. These findings highlight the suitability of the passive air sampling method for accurately assessing and quantifying odor-related compounds in the swine house environment.

3.2. Odorous VOC Emissions

VOCs with lower olfactory thresholds, which are sensory thresholds that detect the presence of odors, as well as recognition thresholds that determine the characteristics of odors, play a crucial role in the intensity of odors emanating from swine feeding operations. The primary odorous VOCs include VSCs, indoles, phenols, and VFAs [7].

In this study, the main odorous compounds identified were methyl sulfide, dimethyl disulfide, dimethyl trisulfide, trimethylamine, and p-cresol, with concentrations ranging from 0.87 μg m⁻³ to 35.38 μg m⁻³. Except for dimethyl trisulfide, for which no olfactory threshold was found, the remaining four substances all exceeded their respective olfactory thresholds. Among the odorous compounds, five odorous compounds (trimethylamine, carbon disulfide, methyl sulfide, dimethyl disulfide, and styrene) were determined to be under control in China, and seven odorous compounds (ethyl acetate, p-xylene, m-xylene, toluene, styrene, methyl sulfide, and dimethyl disulfide) were found to be under control in Japan. While the other controlled odorants do not individually exceed the olfactory threshold, they may still interact to produce odors. These findings underscore the importance of monitoring and controlling the emissions of odorous VOCs, particularly those with low olfactory thresholds, to mitigate the impact of odor nuisance from swine feeding operations on local communities.

3.3. The Influence of Storage Time on VOC Concentrations

3.3.1. Compounds Determined by FID

Pre-concentrated C2–C3 target compounds are separated by gas chromatography and detected by FID after being pre-concentrated to remove H₂O and inert gas in a cold trap. The specific compounds measured in this analysis included ethylene, acetylene, ethane, propylene, and propane. Figure 5 illustrates the recovery of C2–C3 compounds. This study analyzed the sample alterations under both constant pressure and pressure fluctuations since the sample consumption in the container during the multiple injection analyses caused pressure changes inside the container. Constant pressure in the canister was kept by filling it with nitrogen.

We divided the experimental samples into three groups, which were (a) passive air sampling, (b) active air sampling with constant pressure in the canister, and (c) active air sampling with pressure changes in the canister. Three replicates were taken for each method. All three groups had excellent storage stability, with more than 95% recovery after four days of storage. However, the recovery was significantly less within a week, followed by gradual stability. Therefore, it is recommended to analyze the collected samples within a timeframe of four days. The storage stability of ethylene was the best among the samples obtained by passive air sampling, with 90% recovery after one week. The ethane and propylene recoveries were over 85%. Acetylene and propane were about 80%. Acetylene and propane had relatively poor storage stability under active air sampling. Positive air sampling can support repeat sample analysis since pressure variations inside the container have less impact on the overall recovery pattern.

3.3.2. Compounds Determined by MS

Quantification of VOCs was carried out using internal standard analytical methods on GCMS. This study revealed a noteworthy reduction in the recovery rate of several compounds, namely, ethanol, methyl sulfide, and dimethyl disulfide. Ethanol, on the other hand, exhibited the highest concentration of compounds found in swine barns and demonstrated the ability to dissolve a wide range of VOCs. Additionally, methyl sulfide and dimethyl disulfide, both malodorous compounds, were detected, which may cause discomfort in both humans and animals. The influence of storage time on the concentration of these three compounds is illustrated in Figure 6.

In this study, it was observed that all compounds showed higher recovery rates within the first week of storage. During the initial 12-day period, no significant differences were detected among the three storage strategies. The recovery of these compounds remained over 80% throughout this period. However, on Day 20, a more noticeable decline in recovery was observed, with the atmospheric pressure canister showing a recovery rate above 60% and the positive pressure canister showing a recovery rate above 70%. Specifically, methyl sulfide and dimethyl disulfide exhibited less stability by Day 20, with recovery rates of 62.3% and 65.3%, respectively, in the atmospheric pressure canister and 71.3% and 69.1%, respectively, in the positive pressure canister. On the other hand, the recovery rate of other compounds remained at 80% for the entire 20-day period. As a result, it is recommended that all compounds be analyzed within 1 week to ensure accurate results. Furthermore, the effect of storage time on the stability of compounds appeared to be consistent across both the passive air sampling and active air sampling methods, indicating that the storage behavior was independent of the sampling approach.

4. Conclusions

The aim of this study was twofold: to investigate the variations in VOC concentrations resulting from different canister sampling methods during real time in a pig house and to evaluate the impact of storage time on VOC component concentrations by analyzing various canister sampling methods. The study revealed that OVOCs and VSCs were the predominant classes of compounds found in the swine barn, with ethanol, acetone, methyl sulfide, and dimethyl disulfide being the major VOCs identified. Among these, OVOCs have the highest concentration in pig houses, while VSCs are the primary sources of odor.

Both passive air sampling and active air sampling showed no significant difference in the number of detected VOCs, while active air sampling demonstrated lower errors between parallel samples. Nevertheless, noteworthy differences were observed in the sampling of OVOCs and VSCs across the various methods, which needs further investigation. The qualitative analysis detected eight VOCs, with dimethyl trisulfide and p-cresol identified as major odor sources exhibiting varying concentrations among these two sampling methods. The storage time had minimal effects on the stability of compounds in the different sampling methods, and all compounds should be analyzed within four days and no later than one week at the maximum, especially for unstable compounds such as VSCs.