JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 20 February 2008
Published in J Environ Qual 37:521-534 (2008)
DOI: 10.2134/jeq2006.0382
© 2008 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lo, Y.-C. M.
Right arrow Articles by Xin, H.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lo, Y.-C. M.
Right arrow Articles by Xin, H.
Agricola
Right arrow Articles by Lo, Y.-C. M.
Right arrow Articles by Xin, H.
Related Collections
Right arrow Volatile Organic Compounds
Right arrow Air Pollution
Right arrow Animal Waste

Simultaneous Chemical and Sensory Characterization of Volatile Organic Compounds and Semi-Volatile Organic Compounds Emitted from Swine Manure using Solid Phase Microextraction and Multidimensional Gas Chromatography–Mass Spectrometry–Olfactometry

Yin-Cheung M. Loa, Jacek A. Koziela,*, Lingshuang Caia, Steven J. Hoffa, William S. Jenksb and Hongwei Xina

a Dep. of Agricultural and Biosystems Engineering, Iowa State Univ., Ames, IA 50011
b Dep. of Chemistry, Iowa State Univ., Ames, IA 50011

* Corresponding author (koziel{at}iastate.edu).

Received for publication September 20, 2006.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
Swine manure is associated with emissions of odor, volatile organic compounds (VOCs) and other gases that can affect air quality on local and regional scales. In this research, a solid phase microextraction (SPME) and novel multidimensional gas chromatography–mass spectrometry–olfactometry (MDGC-MS-O) system were used to simultaneously identify VOCs and related odors emitted from swine manure. Gas samples were extracted from manure headspace using Carboxen/polydimethylsiloxane (PDMS) 85-µm SPME fibers. The MDGC-MS-O system was equipped with two columns in series with a system of valves allowing transfer of samples between columns (heartcutting). The heartcuts were used to maximize the isolation, separation, and identification of compounds. The odor impact of separated compounds was evaluated by a trained panelist for character and intensity. A total of 295 compounds with molecular weights ranging from 34 to 260 were identified. Seventy one compounds had a distinct odor. Nearly 68% of the compounds for which reaction rates with OH·radicals are known had an estimated atmospheric lifetime <24 h.

Abbreviations: BP, boiling point • DT, detection threshold • FID, flame ionization detector • GC, gas chromatograph • HAP, hazardous air pollutant • HC, heartcut • HS, headspace • MSD, mass selective detector • O, olfactometry • PDMS, polydimethylsiloxane • RT, column retention time • SPME, solid phase microextraction • TIC, total ion chromatogram • VOC, volatile organic compound • VP, vapor pressure


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
SWINE OPERATIONS are sources of aerial emissions of odors, volatile organic compounds (VOCs), particulate matter, and other gases including NH3, H2S, and methane (NRC, 2003). Several research groups have investigated odor emissions from swine operations (Jacobson et al., 2003; Lim et al., 2003). The main source of malodor is microbial degradation in the anaerobic environment of manure storage. A number of studies have been conducted to identify compounds emitted from swine manure. However, a limited number of studies have attempted to determine the relationship between VOCs and corresponding odor. Schaefer (1977) identified 20 VOCs from liquid swine manure. More than 30 VOCs were identified as being responsible for malodor in fresh and rotten swine manure in the research conducted by Yasuhara et al. (1984). Twenty five odorous compounds were identified by Kai and Schafer (2004). Zahn et al. (1997) and Clanton and Schmidt (2000) have also reported recognition and identification of odorous VOCs from swine manure. To date, Schiffman et al. (2001) has by far the most comprehensive list of VOCs and fixed gases in air around swine production facilities that has been published with 324 compounds being tentatively identified. Comparison of sampling locations, methods, and VOCs related to swine manure and swine operations is presented in Table 1 .


View this table:
[in this window]
[in a new window]

  		Table 1. Comparison of analytical methods used to characterize volatile organic compounds (VOCs) emitted from swine operations.

 
Volatile organic compounds identified in and around swine operations can be classified into different chemical groups including acids, alcohols, aldehydes, amines, volatile fatty acids (VFAs), hydrocarbons, ketones, indoles, phenols, nitrogen-containing compounds, sulfur-containing compounds, and other compounds (Schiffman et al., 2001). Thus, the chemical and sensory characterization of these compounds is quite challenging due to the wide range of physicochemical properties. Organic compounds with molecular weights from C2 to C9 were recognized as having the most impact to air quality at swine operations (Zahn et al., 1997). The boiling point (BP) of the key odorous VOCs such as indoles and VFAs ranged from less than 10°C to greater than 250°C and vapor pressures (VP) ranged from less than 10 Pa to greater than 100,000 Pa (Willig et al., 2004). Published detection thresholds (DTs) of VOCs identified in and around swine operations in North Carolina ranged from greater than 100 µL L–1 to less than 0.001 µL L–1 (Schiffman et al., 2001). Large uncertainties are associated with published DTs (Devos et al., 1990; AIHA, 1989; Rychlik et al., 1998). For many compounds, the estimates of DTs continue to be adjusted or are not known.

Sample collection methods include the use of sorbent traps (e.g., Tenax, Tenax-TA), Tedlar bags, whole air sampling, and solid phase microextraction (SPME). Several standard methods based on sorbent tubes (USEPA TO-17) and whole canisters (USEPA TO-15) were developed for VOCs in ambient air in typical urban and less polluted rural environments (Woolfenden and McClenny, 1999; McClenny and Holdren, 1999). However, no standard method is available for odor-causing VOCs in livestock environments. This is because these compounds are often polar and reactive, can undergo reactions with themselves and interact with sampling lines and containers, and can be affected by the presence of moisture (Keener et al., 2002; Koziel et al., 2005). Thus, it has been challenging to develop robust sampling and analysis methods for these compounds. Some have attempted to modify existing TO-15 and TO-17 to sample the VOCs in and around swine operations (Schiffman et al., 2001; Zahn et al., 2001; Blunden et al., 2005). However, caution should be exercised when standard methods are modified for other compounds. SPME is an alternative for air sampling (Pawliszyn, 1997; Koziel and Pawliszyn, 2001).

SPME combines sampling and sample preparation into one step, reducing the sampling/sample preparation time. No solvent or pump is needed with SPME, and it is a reusable sampling technique suitable for laboratory and field work. SPME extractions are facilitated on a polymeric coating that has a high affinity for organic compounds. One potential useful characteristic of SPME fibers is that the polymeric coatings used are not affected by the presence of moisture when long sampling times are used (Miller and Woodbury, 2006). SPME has been used for sampling of volatile compounds in air (Koziel and Pawliszyn, 2001; Begnaud et al., 2003). Quantitative sampling of airborne compounds such as alkanes, aromatic hydrocarbons, and formaldehyde is possible with SPME (Martos and Pawliszyn, 1997; Martos and Pawliszyn, 1997; Koziel and Pawliszyn, 2001; Koziel et al., 2001). Larreta et al. (2006) quantified VOCs in cow slurry and Cai et al. (2007) used SPME to evaluate the effectiveness of zeolite to control VOCs and odors emissions from poultry manure. To date, relatively little progress has been made with SPME applications for the quantification of odorous gases in and around livestock operations. However, SPME has been useful for qualitative characterization and screening of livestock gases. Sampling of livestock VOCs and odorants with SPME has been used to characterize swine dust odorants (Cai et al., 2006), downwind odor impact of beef cattle feedlots (Wright et al., 2005), and downwind odor impact of swine finisher operations (Bulliner et al., 2006; Koziel et al., 2006).

Gas Chromatography–Olfactometry (GC-O)
Gas chromatography coupled with a flame ionization detector (FID) is often used for chemical separation and analysis of VOCs emitted from livestock operations (Schiffman et al., 2001; Begnaud et al., 2003; Kai and Schafer, 2004). The addition of a sniff port enables simultaneous chemical and olfactometry analysis of livestock odor (Kai and Schafer, 2004; Wright et al., 2005; Cai et al., 2006). To date, most GC-O applications are related to agriculture, food chemistry, and the consumer products industry. Another option is to combine olfactometry with GC-mass spectrometry (MS) for compound identification (Pollien et al., 1997). Cai et al. (2006) reported partitioning of odorants to various fractions of swine dust. Koziel et al. (2006) used GC-MS-O to analyze air samples downwind from swine finisher barns. Separation of livestock VOCs is typically accomplished on a single-column GC. However, a single column does not always have the capability to completely separate a complex air sample. Thus, multidimensional GC provides a new powerful way to resolve the complex livestock air.

Multidimensional Gas Chromatography
Multidimensional GC (MDGC) utilizing multiple columns represents the state-of-the-art refinement for the separation of VOCs and semi-VOCs. Most compounds emitted from livestock manure are polar and many characteristic odorants are semi-volatile. The use of a single column without enough resolution power may result in incomplete chromatographic separation. Co-elution of two or more compounds is a critical obstacle in GC analysis (Bertsch, 1999) and it is even more critical for odor characterization. To date, conventional MDGC with heartcut (HC) has been used in the tobacco industry, the petroleum industry, and in food chemistry studies. Recently, the multidimensional gas chromatography–mass spectrometry–olfactometry (MDGC-MS-O) approach was used to analyze complex odorous samples from swine operations (Bulliner et al., 2006).

The overall objective of this research was to identify and characterize VOCs emitted from swine manure using SPME and MDGC-MS-O. To date, no study related to VOCs emitted from livestock operations has been conducted using the MDGC-MS-O system. The advantage of using MDGC-MS-O is the enhanced VOC separation and the simultaneous odor identification. It is critical to characterize swine manure VOCs and malodorous gases to improve the understanding of the environmental impact of swine operations. This knowledge is also needed to develop and evaluate odor and gas mitigation strategies and technologies.


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results and Discussion
 Conclusions
 REFERENCES
 
Manure samples were collected from the nursery pit, the finisher pit and the outside storage pit at the Swine Nutrition and Management Research Farm (Ames, IA). Manure samples were then transported to the laboratory and dispensed 15 mL into 40 mL screw-capped vials with a polytetrafluoroethylene (PTFE)-lined silicone septum (Supelco, Bellefonte, PA). All vials were stored in a fume hood at room temperature up to 4 d before sampling.

Sampling and Sample Preparation of Swine Manure Headspace with Solid Phase Microextraction
Four commercially available SPME fibers, including Carboxen/PDMS 85 µm, PDMS 100 µm, Polyacrylate 85 µm, and PDMS/DVB 65 µm (Supelco, Bellefonte, PA), were first used to select the most efficient fiber coating in extracting VOCs and semi-VOCs emitted from swine manure. All SPME fibers were used for extracting headspace samples from replicates. All fibers were conditioned before first use according to the manufacturer's instruction. SPME fibers were inserted into each vial through the septum exposing the SPME fiber to the vial headspace. Preliminary experiments were conducted to determine the optimal SPME sampling time. Exposure times between 10 s and 24 h were examined. The number of compounds and the intensity of odor events increased with sampling time. Thus, samples with longer extraction times had a greater potential to reveal new compounds and odors. For practical reasons, a 24 h SPME extraction at room temperature was selected. It was assumed that long storage simulated anaerobic environment associated with higher malodor. With each extraction, the SPME fiber was removed immediately from the vial and was inserted into the GC injection port for analysis. All GC injector conditions were the same as reported in Cai et al. (2006).

Analyses of Swine Manure Headspace Samples with Multidimensional Gas Chromatography–Mass Spectrometry–Olfactometry
Simultaneous chemical and sensory analyses of gases emitted from swine manure were completed using the MDGC-MS-O (Microanalytics, Round Rock, TX) system. The system GC–MS-O components, the software, and basic GC oven programs are described in detail in Cai et al. (2006). The GC was operated in a constant pressure mode where the mid-point pressure, i.e., pressure between pre-column and analytical column (Fig. 1 ), was always at 5.8 psi and the HC sweep pressure was 5.0 psi. Multidimensional capability was used for better separation of gases and odors of compounds associated with swine manure. The MDGC-MS-O system was capable of working in three modes, i.e., GC-FID only, GC-FID-O, and GC-MS-O. The MultiTrax (Microanalytics, Round Rock, TX) software was used to control the timing of valves and HC for each mode.


Figure 1
View larger version (37K):
[in this window]
[in a new window]

  		Fig. 1. Schematic of multidimensional gas chromatography–mass spectrometry–olfactometry (MDGC-MS-O).

 
Heartcut is as a fraction of a sample "cut" from the non-polar pre-column and transferred to the polar column with a Dean's switch between the two columns. Compounds were further separated on the polar column and then simultaneously analyzed on the mass selective detector and the sniff port. When the HC valve was opened (GC-MS-O mode), the gas sample was transferred from the pre-column into the polar column. When the HC valve was closed (GC-FID or GC-FID-O modes), the sample was not transferred into the polar column. Compounds were further separated on the polar analytical column and then simultaneously analyzed on the mass selective detector (MSD) and by the panelist at the sniff port. Only one trained panelist was responsible for odor determination throughout the entire study. A series of 30-s-wide HCs starting from 0.05 to 24 min (e.g., 0.05 min–0.5 min, 0.5 min–1 min, and so on) were used to methodically expand the chromatographic separation and enhance identification of new compounds. The total number of separate HCs was therefore 48.

Data Analysis and Interpretation
Three sets of signals were generated for each sample including the total ion chromatogram (TIC), the FID signal, and the aromagram (Bulliner et al., 2006). Data were analyzed using the AromaTrax, BenchTop/PBM (Palisade Corp., Ithaca, NY) and MSD ChemStation (Agilent Inc., Wilmington, DE) software as well as information from the NIST library (NIST, 2005). Compounds recorded in TICs for each HC were recognized and identified according to criteria: match of sample mass spectrum, retention time, and odor. Based on the recorded odor events, TIC were analyzed around the time in which the odor event began and the compound responsible for the odor was recognized. Whenever feasible, compounds were positively identified by matching the retention time and its mass spectra with that of a pure standard (Sigma-Aldrich, St. Louis, MS). Odor character related to a particular compound was recorded and compared with the odor databases at LRI & Odour Database (2005) and Flavornet (Acree and Arn, 2004).

Physicochemical properties for the identified compounds were selected to better characterize compounds emitted from swine manure. These physical and chemical properties included carbon number, BP, VP, water solubility (S), Henry's law constant (Hc), octanol-water partitioning coefficient (logKow), atmospheric lifetime ({tau}), and the liquid- and gas-phase molecular diffusion coefficient (Dl and Dg, respectively). Examination of these properties is useful to develop air sampling methods and to model a chemical's fate once emitted from manure. Recently, the reactivity of VOCs released from livestock operations has been of interest to many researchers and regulatory agencies. In this research, {tau} was estimated based on the reaction with OH radicals using the following formula:

Formula 1[1]
where k is a rate constant at 298 K (cm3 molecule–1sec–1) and [OH·] is an average atmospheric OH concentration of 1.0 x 106 molecule cm–3. Values of k were obtained from Syracuse Research Corporation (2004) and ranged from 0 (H2S) to 2.27 E-10 (Tropex). The estimation of Dg was more accurate using the Wilke and Lee method (Lyman, 1982). The estimation of Dl was made using the Hayduk and Laudie method (Lyman, 1982). Detection threshold for odor was based on Devos et al. (1990), AIHA (1989), and Rychlik et al. (1998).


   Results and Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
Conclusions
 REFERENCES
 
Selection of Solid Phase Microextraction Fiber Coating
Four commercially available SPME fibers (85 µm Carboxen/PDMS, 100 µm PDMS, 85 µm Polyacrylate, and 65 µm PDMS/DVB) were used for selecting the most efficient coating. All four fibers were exposed to manure headspace for 24 h. Comparison of chromatograms obtained in GC-FID mode showed that all four SPME fibers were very efficient. However, the Carboxen/PDMS 85 µm coating was capable of extracting compounds in a wide range of column retention times (RTs) from very volatile to semi-volatile compounds (Lo, 2006). The Carboxen/PDMS 85 µm coating was more efficient in extracting very volatile compounds eluting with RTs up to 8 min. The second most efficient fiber in this region was the PDMS/DVB 65 µm, followed by the Polyacrylate 85 µm and the PDMS 100 µm coatings, respectively. Both the Carboxen/PDMS 85 µm and the PDMS/DVB 65 µm were also very efficient for midrange VOCs eluting between 8- and 23-min RT. The PDMS 100 µm and the PDMS/DVB 65 µm were the most efficient for semi-VOC compounds eluting later than 23 min. These observations were consistent with general guidelines for fiber selection (Pawliszyn, 1997).

Comparison of TIC signals of air samples collected in headspace of swine manure with four SPME fiber coatings and analyzed using the MDGC-MS-O mode was then completed (Lo, 2006). The identical 30 s heartcut between 11 and 11.50 min representing midrange volatility compounds was selected where all four fibers were expected to be very effective (Pawliszyn, 1997). The Carboxen/PDMS 85 µm coating was the most efficient in extracting the largest number of detectable (above the TIC baseline) compounds. The number of compounds identified with the same detection criteria were 9, 6, 6, and 3 for the Carboxen/PDMS 85 µm, PDMS/DVB 65 µm, PA 85 µm, and the PDMS 100 µm fiber coating, respectively. Also, no additional compounds were found in the TICs collected with SPME fibers other than those collected with the Carboxen/PDMS 85 µm coating.

The use of different SPME fiber coatings and its effects on odor were also examined. The comparison of aromagrams of air samples collected with four SPME fiber coatings and analyzed in the MDGC-MS-O mode resulted in concluding that the Carboxen/PDMS 85 µm coating was more effective in capturing odorous compounds. The number of odor events recorded using AromaTrax were 7, 4, 4, and 2 for Carboxen/PDMS 85 µm, PDMS/DVB 65 µm, Polyacrylate 85 µm, and the PDMS 100 µm fiber coating, respectively. In addition to number of odor events detected, odor intensity and odor event area were also compared (Lo, 2006). Odor event areas were calculated by the software using the following equation:

Formula 2[2]
where odor duration = odor event end time – event start time.

Carboxen/PDMS 85 µm and PDMS/DVB 65 µm coatings had the highest average odor intensities (22.3 and 22.8%, respectively), while Polyacrylate 85 µm coating ranked third (14.25%) and the PDMS 100 µm fiber coating yielded the lowest odor intensity recorded (7.5%). Among these four different SPME coatings, the Carboxen/PDMS 85 µm coating had the highest total peak area count while the PDMS 100 µm coating ranked last. Carboxen-containing coatings were very efficient in extracting VFAs and sulfur compounds. Fibers with a divinylbenzene (DVB) phase were generally more efficient in extracting phenolics.

Based on the comparison of both the chemical and olfactometry data analysis, Carboxen/PDMS fiber coating was the most effective coating for VOCs/semi-VOCs extraction in terms of number of compounds extracted (both in GC-FID mode and MDGC-MS-O mode), number of odor events recorded, odor intensity, and odor event peak area count. Therefore Carboxen/PDMS 85 µm coating was selected for subsequent extractions of gases from swine manure headspace.

Solid Phase Microextraction Replications
Three replications of the same 30-s-wide HC using three different Carboxen/PDMS 85 µm fibers were analyzed using MDGC-MS-O mode to determine variability between the TIC signals and odor events. All three fibers were exposed to manure headspace for 24 h under room temperature. Five of the most significant peaks from these three replicates were selected for qualitative comparisons using peak area integrations with MSD ChemStation. The relative standard deviation (RSD) of the peak area count were 40.7, 20.7, 27.3, 1.3, and 3.9% for 5-undecene, 1-octanethiol, benzeneethanol, phenol, and 4-methylphenol, respectively. These apparent variations are likely due to uncontrolled microbial activity in manure replicates. These variations did not have a significant effect on the detection and identification of compounds.

Aromagrams recorded using AromaTrax from the three replicates analyzed in the MDGC-MS-O mode were compared for odor intensities. Seven odor events were consistently detected and recorded for the HC 11 to 11.50 min. Average odor intensities varied from 11.3 to 29.0% on 100% relative scale. The RSD for these odor intensities varied from 6.0 to 54.4% (mean = 25.2%). Besides the variability introduced by the panelist, the results were also affected by uncontrolled microbial activity and resulting variability in gaseous emissions in each vial.

The comparison of both chemical and olfactometry data obtained from experiments comparing SPME fiber coatings and comparing the three replicates using the same fiber coating (i.e., Carboxen/PDMS 85 µm) were satisfactory and suitable to address the objectives of this research. Identification of compounds and their odor character was consistent in all replicates, and aroma peaks analysis was consistent in all three replicates, with an acceptable variation in odor intensity and peak area count. Therefore Carboxen/PDMS 85 µm SPME fiber for swine manure headspace extractions was selected for all remaining air sampling for the chemical and sensory analysis in this research.

Multidimensional Gas Chromatography–Mass Spectrometry–Olfactometry
Headspace samples were analyzed with the MDGC-MS-O mode utilizing 30-s-wide HCs. During those HCs, the midpoint valve was opened for only 30 s allowing a narrow range of separated compounds from the pre-column to be transferred to the analytical column for better separation. An example of a 30-s HC is shown in Fig. 2 . The aromagram recorded was only for the odors detected from the 30-s-wide HC and sent from the pre-column to the polar column. It is interesting to note that only 23 odor events were recorded, much less than the events from the full HC (Lo, 2006), which allowed for easier matching of odor events and chemical compounds. The chromatographic separations were improved. Sample background from co-eluting compounds was also lower, allowing for improved spectral matches. Note that only one panelist was responsible for odor determination in this study. If multiple panelists were used, it is natural to expect some variations related to odor character and intensity between panelists due to subjective human response. Discussion of RSDs related to human perception of p-cresol for three panelists using MDGC-MS-O was reported by Bulliner et al. (2006).


Figure 2
View larger version (29K):
[in this window]
[in a new window]

  		Fig. 2. Comparison of aromagram, flame ionization detector (FID), total ion chromatogram (TIC) signals with heartcut (HC) between 15.50 and 16.00 min.

 
Identification of Volatile Organic Compounds and Semi-Volatile Organic Compounds
A summary of all compounds identified in swine manure headspace is shown in Table 2 . Table 2 contains the chromatographic retention time, compound name, the CAS number and the heartcut timing for which a compound was first detected. Some compounds were identified in several HCs. This was due to the insufficient chromatographic separation on the 12 m non-polar precolumn where the peaks were often wide. Thus, HCs set for equal 30-s intervals were transferring them to the column several times. It is interesting to mention that the resulting retention time of the compound shifted slightly toward longer times if the compound was heartcutted from a wide peak. This was due to the change of mid-point pressure during each heartcut. For semi-VOC, the shift of RTs was greater than the RT shift for VOCs.


View this table:
[in this window]
[in a new window]

  		Table 2. Summary of compounds identified from swine manure headspace.{dagger}

 
Whenever applicable, published or detected odor character (CambridgeSoft Corporation, 2006; NIST, 2005), detection threshold (Devos et al., 1990; AIHA, 1989; Rychlik et al., 1998), presented in µL L–1, and estimated atmospheric lifetime (using Eq. [1]) are also presented in Table 2. Odor descriptors recorded by the panelist in this study were presented, along with odor descriptors obtained from Flavornet (Acree and Arn, 2004) and LRI & Odour Database (2006).

A total of 295 compounds emitted from swine manure were identified. These compounds can be classified into 12 chemical classes with the numbers of compounds in each class, i.e., acids (9), alcohols (33), aldehydes (4), aromatics (32), esters (6), ethers (10), fixed gases (2), hydrocarbons (36), ketones (71), nitrogen-containing compounds (35), phenols (19), and sulfur-containing compounds (38). Molecular weights ranged from 34 to 260. Of these 295 compounds, 113 were positively confirmed with pure standards. Approximately 25% of the total compounds had distinct odor. Approximately 107 compounds had been reported in previous studies (Table 1) (Yasuhara et al., 1984; Clanton and Schmidt, 2000; Zahn et al., 2001; Schiffman et al., 2001; Begnaud et al., 2003; Willig et al., 2004; Kai and Schafer, 2004). The total of 188 compounds was reported here for the first time (Tables 1 and 2) and 26 of them had a distinct odor. The difference in chemical/odor distribution between the three different locations (nursery pit, finisher pit, and outside storage) where manure was collected was not the scope in this study. However, it is reasonable to assume that some differences in emission rates of the majority of VOCs emitted from manure are associated with variables such as the diet, manure origin and age, manure management, season, and location. Zahn et al. (1997) reported some evidence that there is a link between the source and the specific VOCs related to swine manure. However, more work is needed to provide such chemical and odor profiles as a function of aforementioned variables.

To date, the most comprehensive list of VOCs associated with swine manure was published by Schiffman et al. (2001) with 324 compounds listed. However, close inspection of the data shows that this list was made based on chemical analysis of swine manure and air samples collected around swine operations in North Carolina. If only the compounds found in air samples are considered, the list is reduced to 251 compounds (Schiffman et al., 2001). In this research, a total of 295 compounds were found in the headspace of swine manure. In addition, Schiffman et al. (2001) quantified 81 compounds based on surrogate calibrations for a subset of 14 compounds only. There is no information about the number of compounds that were confirmed with pure standards except the listed 14 (for calibrations). In this research, 113 compounds were positively confirmed with standards.

Sixteen compounds found in this research are listed as hazardous air pollutants (USEPA, 2002). Note that quantification of chemicals emitted from swine manure was not part of the objective of this research, thus the concentrations of these 16 compounds were not estimated. Future research is warranted to determine concentrations, emission rates, and fate of these compounds. The 16 compounds classified as hazardous air pollutants (HAPs) were: carbon disulfide, 2-butanone, benzene, 4-methyl-2-pentanone, toluene, ethyl-benzene, 1,4-dimethyl-benzene, 1,3-dimethyl-benzene, 1,2-dimethyl-benzene, ethenyl-benzene, naphthalene, quinoline, phenol, 2-methyl-phenol, and 4-methyl-phenol.

Only 77 compounds identified had DTs published in previous studies (AIHA, 1989; Devos et al., 1990; Rychlik et al., 1998). Detection thresholds are summarized in Table 2 and Fig. 3 where the DTs are presented in a frequency distribution chart with ranges from 0.001 nL L–1 to 1 µL L–1. The majority (~80%) of compounds had their DTs between 1 nL L–1 and 1 µL L–1. The six compounds with a DT less than 1 nL L–1 (i.e., most odorous compounds) were 2-bromo-phenol, indole, 2,4-hexadienal, skatole, 2-chloro-phenol, and 2,6-dimethyl-phenol. Approximately 47% of compounds summarized in Fig. 3 had odor character that can be considered "offensive," 31% had odor character that can be considered "neutral," and 22% had "pleasant" odor character. This summary was based on the assessment of odor descriptors summarized from this research as well as from Flavornet (Acree and Arn, 2004) and LRI & Odour Database (2006).


Figure 3
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 3. Distribution of odor threshold for 77 out of 295 compounds emitted from swine manure.

 
Characterization of Physicochemical Properties of Volatile Organic Compounds
Physicochemical properties of VOCs are useful for the comprehensive characterization, including analytical methods development, measurements, modeling, fate, and development, of emissions abatement approaches. These parameters were obtained from several databases including the Syracuse Research Corp. (2004), NIST WebBook (National Institute of Standard and Technology, 2005) and CambridgeSoft Corp. (2006), while the gas phase and liquid phase diffusion coefficients were calculated based on the methods from Lyman (1982). In general, VOCs emitted from manure represent very wide ranges of physicochemical properties. A summary of all findings was presented by Lo (2006). Approximately 89% of the compounds identified fell into the category of VOCs and semi-VOCs (<12C) (De Nevers, 1995; Peterson, 2005) with about 74% within C5 to C10. The compound identified with the highest carbon number was heptadecane (C17H36), and the compounds with the lowest carbon number included methanethiol (CH4S) and carbon disulfide (CS2). It is interesting to mention that as many as 31 of the identified compounds (~11%) had a carbon number ≥12. In a straight sense these compounds were not classified as VOCs (De Nevers, 1995) because of the carbon number threshold (<12). These compounds also tended to have high BPs and low VP, which made them difficult to sample and analyze. It is remarkable that the Carboxen/PDMS SPME fiber was capable of extracting these compounds.

The BP of 215 compounds (for which BPs were known or published) ranged from –60.3 to 322°C with a mean BP of 168°C. As many as 66% of the compounds had a BP between 120 and 220°C. Approximately 17% of the compounds had a BP below 120°C and 18% were above 220°C. Vapor pressure and water solubility are important in determining the emission of VOCs in swine manure. A compound with a high VP and low solubility is considered as more volatile than those with low VP and high water solubility (Verscherene, 2001). Vapor pressure and solubility of 219 compounds (for which VP was known or published) were summarized (Lo, 2006). As many as 63% of the compounds had a VP greater than 69.33 Pa with a range from 4.6E–5 Pa Hg to 2.08E + 6 Pa. About 68% of the 219 compounds had solubility between 100 and 1.00E + 5 mg L–1 (for which water solubility was known or published), which are considered very soluble in water. As many as 86% of compounds had a Henry's law constant, defining the solubility of gases in pure water <0.01 (atm-m3mole–1). Water-octanol partitioning coefficients (for which were known or published) (logKow) ranged from –2.2 to 9.26, with about 82% out of 228 volatiles having a logKow of less than 4. The compound with the highest logKow was 1-heptanethiol, with a logKow = 9.26. The estimated Dg and Dl ranged from 0.043 to 0.18 cm2 s–1 and 4.54E–06 to 1.93E–05 cm2 s–1, respectively. As many as 76% of the 216 VOCs fell between Dg 0.06 to 0.09 cm2 s–1, and 82% of the 217 VOCs fell between Dl 6.00E–06 to 1.00E–05 cm2 s–1 (for those VOCs with chemical structure and BPs known or published).

Many VOCs emitted into the atmosphere take part in the degradation/transformation reactions (Atkinson et al., 1999). The estimation of atmospheric lifetime of VOCs depends on the reaction with hydroxyl radical (OH·) as this compound dominates the daytime atmospheric reactions. The frequency distribution of the estimated atmospheric lifetime for 210 VOCs (for which atmospheric OH·rate constants used to calculate the atmospheric lifetime were known or published) is presented in Fig. 4 . More than half of compounds (approximately 68%) had an estimated {tau} less than 24 h, which are considered as very reactive, with dimethyl disulfide ({tau} = 1.224 h) being the most reactive compound emitted from swine manure. In general, the shorter the atmospheric lifetime, the more reactive a compound is. This, in turn, could be detrimental to the chain of reactions leading to the net production of ozone. Reactivity causes transformations into new compounds and potentially affects the odor. Thus, atmospheric reactivity may also affect the overall odor. This reactivity-odor link has not yet been characterized.


Figure 4
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 4. Distribution of estimated {tau} for 210 compounds emitted from swine manure.

 

   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
REFERENCES
 
The following conclusions can be drawn from this research:
(1) SPME combined with MDGC-MS-O was a powerful tool used to extract and separate VOCs and gases emitted from swine manure, to identify compounds, and to determine their odor characteristics. The use of heartcut improved chromatographic separations and compound identification.
(2) A wide range of VOCs and gases were emitted from swine manure. As many as 295 compounds were identified from the gas samples using MDGC-MS-O. Seventy one compounds were recognized as odorous compounds. Sixteen of the compounds identified were listed as HAPs. The six compounds with DT less than 1 nL L–1 were 2-bromo-phenol, indole, 2,4-hexadienal, skatole, 2-chloro-phenol, and 2,6-dimethyl-phenol.
(3) Among the 295 compounds identified, 188 were not reported in previous studies. This total number (295) also represents an improvement by 44 of the total number of compounds listed in the most comprehensive summary of compounds present in swine manure and/or air around swine operations (Schiffman et al., 2001).
(4) Physical and chemical properties of the compounds emitted from swine manure were studied and summarized. The 295 compounds identified were classified into 12 chemical classes: acids (9), alcohols (33), aldehydes (4), aromatics (32), esters (6), ethers (10), fixed gases (2), hydrocarbons (36), ketones (71), nitrogen-containing compounds (35), phenols (19), and sulfur-containing compounds (38).
(5) Nearly 68% of the total 210 compounds (for which the reaction rate with OH was known) had an estimated atmospheric lifetime ({tau}) <24 h, with dimethyl disulfide ({tau} = 1.22 h) being the most reactive.
(6) Measurement of actual concentrations and emission rates for specific VOCs of interest is warranted.


   ACKNOWLEDGMENTS
 
We would like to acknowledge the help and support from Dan Johnson, manager of the Swine Nutrition and Management Research Farm, as well as the staff members of the farm. We would also like to acknowledge Iowa Pork Producers Association and Iowa State Univ. for funding this research project.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 J ANIM SCIHome page
A. Shabtay, U. Ravid, A. Brosh, R. Baybikov, H. Eitam, and Y. Laor
Dynamics of offensive gas-phase odorants in fresh and aged feces throughout the development of beef cattle
J Anim Sci, May 1, 2009; 87(5): 1835 - 1848.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lo, Y.-C. M.
Right arrow Articles by Xin, H.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lo, Y.-C. M.
Right arrow Articles by Xin, H.
Agricola
Right arrow Articles by Lo, Y.-C. M.
Right arrow Articles by Xin, H.
Related Collections
Right arrow Volatile Organic Compounds
Right arrow Air Pollution
Right arrow Animal Waste

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome