Quantification of VOC Emissions from Carbonized Refuse-Derived Fuel Using Solid-Phase Microextraction and Gas Chromatography-Mass Spectrometry

In this work, for the first time, the volatile organic compound (VOC) emissions from carbonized refuse-derived fuel (CRDF) were quantified on a laboratory scale. The analyzed CRDF was generated from the torrefaction of municipal waste. Headspace solid-phase microextraction (SPME) and gas chromatography-mass spectrometry (GC-MS) was used to identify 84 VOCs, including many that are toxic, e.g., derivatives of benzene or toluene. The highest emissions were measured for nonanal, octanal, and heptanal. The top 10 most emitted VOCs contributed to almost 65% of the total emissions. The VOC mixture emitted from torrefied CRDF differed from that emitted by other types of pyrolyzed biochars, produced from different types of feedstock, and under different pyrolysis conditions. SPME was a useful technology for surveying VOC emissions. Results provide an initial database of the types and relative quantities of VOCs emitted from CRDF. This data is needed for further development of CRDF technology and comprehensive assessment of environmental impact and practical storage, transport, and potential adoption of CRDF as means of energy and resource recovery from municipal waste.


Introduction
Biochar is a fine-grained product characterized by a high content of organic carbon and low susceptibility to decomposition. It is obtained in the process of torrefaction, pyrolysis, or gasification of plant biomass, biodegradable waste, and sewage sludge [1]. The European Biochar Certificate [2] defines the carbon content above 50% of dry matter as the main requirement for biochar classification. Biochar has a wide range of applications with more than 50 already documented [3]. Biochars' intended use depends on the production process characteristics, primarily calorific value and the specific surface area [3]. The substrates used in the production of biochar include [4]: wood biomass, agricultural biomass (e.g., crop residues), energy crops (e.g., Miscanthus, energetic willow, Virginia mallow), organic waste including: organic fraction of municipal waste [5,6], waste from agro-food processing

Results
In this work, for the first time, emissions of VOCs from CRDF was studied qualitatively and quantitatively. Qualitative analysis consisted of identifying compounds based on MS spectral database and available literature (Kovats Retention C7-40 Index). Table 1 shows the VOCs emitted from the analyzed CRDF with the GC column retention time and the coefficient both in the literature and with the GC software presented in the database (Kovats Retention C7-40 Index). Also included was the internal standard (2-undecanone) added during analyzes (compound #80).
The largest (by number) group were derivatives of benzene and naphthalene (e.g., tetralin). The highest density of peak elution of VOCs from the chromatographic column occurred between 7 to 12 min (Table 1). Most of the identified compounds had boiling points between 100 and 240 • C; i.e., the typical range of VOCs [16]. One compound was classified as very volatile (VVOCs) and one as a semi-VOC (Table 2). Among the identified compounds, many have been known to have a negative impact on human health and the natural environment, including mutagenic and carcinogenic aromatic compounds, e.g., toluene, benzene, ethylbenzene or cumene, and PAH, e.g., naphthalene.
The total mass of VOCs emitted from CRDF was 16.4 mg/kg (Table 2) based on 7 days of accumulation in the headspace of a sealed storage vessel. The top 10 compounds with the highest emissions were as follows: nonanal, octanal, heptanal, butylbenzene, hexanal, 1-methyl-4-prop-1-en-2-ylcyclohexene, benzaldehyde, decanal, toluene, and hexylbenzene. Among the analyzed compounds, the highest emission (as a group) from the CRDF was determined for aldehydes: nonanal, followed by octanal, and heptanal ( Table 2). The top 10 of the most emitted VOCs consisted almost 65% of total emissions.  Table 2. VOCs emissions (accumulated in a headspace of sealed vessel over 7 days of storage) from (torrefied) carbonized refuse-derived fuel ordered from the highest (µg of VOC per kg of CRDF) to lowest; % of total emissions, boiling point, VOC classification, and a comparison with VOCs emitted from other types of (pyrolyzed) biochar (woody biomass, algal biochar, and municipal solid waste (compost), respectively) [11,12,14].

Discussion
The determined composition of the VOCs mixture emitted from CRDF stored in a sealed vessel (this research) is unique because it was likely driven by the type of municipal waste and the process parameters used for its production. However, for illustrative purposes, it is useful to compare with VOCs emitted from other types of biochar. Spokas et al. [11] reported 140 different compounds, 74 were identified in all studied biochars, generated from 77 different materials; but without municipal solid waste and without fuels derived from municipal waste. Spokas et al. [11] have not found clear feedstock dependencies to the adsorbed VOC composition, suggesting a stronger linkage with biochar production conditions coupled with post-production handling and processing. Lower pyrolytic temperatures (≤350 • C) produced biochars with adsorbed VOCs consisting of short carbon chain aldehydes, furans, and ketones; elevated temperature biochars (>350 • C) typically were dominated by adsorbed aromatic compounds and longer carbon chain hydrocarbons.
In the present work, only eight compounds were also reported by Spokas et al. [11] ( Table 2). This relatively small number of common VOCs corroborates the unique influence of feedstock type -CRDF (in this research), and torrefaction process (a lower temperature process different to pyrolysis, and gasification) on VOCs formation during waste/biomass thermal treatment. Similarly, to present studies [11] aldehydes were identified in biochars ( Table 2).
Buss et al. [12] analyzed VOCs emitted from three algal biochars, including two contaminated by re-condensates during pyrolysis. Buss et al. [12] identified numerous compounds from phenol groups mainly methylated and ethylene (25 compounds, but only phenol was common with present study) and acids such as acetic, formic or propionic. Taherymoosavi et al. [14] used municipal waste (compost) for the production of biochar and thus, was closest (as a source) to this work. Taherymoosavi et al., [14] analyzed biochar formed in the pyrolysis process at temperatures from 105 to 650 • C and reported the presence of alkylbenzenes, methoxy alkylphenols, organic compounds containing nitrogen, furans, and aromatic compounds. However, only phenol was a common compound identified in the present study ( Table 2). Compared results show that only two compounds acetic acid and phenol were identified in the present study and [11,12], and [12,14] respectively.
There is little research in literature related to the subject of qualitative and quantitative identification of VOCs emitted from the surface of biochar, especially from biochar produced from municipal solid waste such as CRDF. This is a relatively new topic related to the trend of using torrefaction, and low-temperature pyrolysis of municipal solid waste in recent years. These new trends in municipal solids treatment are being sought as an alternative to both energy production and 'Waste to Carbon' utilization (e.g., CRDF). Thus the interest in identifying and mitigating VOC emissions from biochar will likely increase. As biochar VOCs are still not deeply explored, it is required to continue research on the effects of feedstock type and thermal treatment conditions on VOCs formation and emission, especially in the contest on potential harmful effect to workers during biochar storage and transportation and end users.

CRDF Used in the Experiment
CRDF was produced in the torrefaction process at 260 • C and a 50 min retention time in a batch reactor, according to the procedure described by [5]. The analyzed CRDF from the torrefaction of municipal waste at 260 • C and 50 min of retention time was characterized by physicochemical properties similar to those described in the literature. CRDF with a lower heating value (LHV) of 25.95 MJ/kg was similar to CRDF obtained in earlier studies [5] and to biochar from grass produced in a similar temperature range (250 to 350 • C) by Weber and Quicker [17], which had a calorific value of 25 to 30 MJ/kg. The higher heating value (HHV) of CRDF used in this experiment (27.315 MJ/kg) could define it as a 'hard coal' (HHV > 23.9 MJ/kg), according to the IEA's classification [18]. The moisture content of the analyzed material (1.54%) was in the 1 to 6% range [19]. The proximate and ultimate properties of the CRDF used were summarized by Białowiec et al. [9].

Qualitative and Quantitative Analyses of VOC Emitted from CRDF
Measurements of VOCs were made using headspace (HS) solid-phase microextraction (SPME) technology for gas extraction and gas chromatography coupled with mass spectrometry (GC-MS) (Palo Alto, CA, USA) for analyses. SPME technology combines sampling and sample preparation and is suited for exploratory qualitative and quantitative work on VOC emissions from a wide range of sources such as contaminated soils [20,21], decaying animal carcasses [22,23], fermentation by-products in beverages and aromas in wines [24,25], biological fluids and gases [26][27][28][29][30]. A comprehensive review of SPME applications to food and environmental analysis was published by Merkle et al. [31]. The apparatus and reagents were as follows: the internal standard-a solution of 2-undecanone at a ratio of 20 µg compound per 20 mL of distilled water; 2) water bath with a temperature of 40 • C with glycol; 3) manual holder for SPME; 4) universal SPME fiber 3-component DVB/CAR/PDMS 50/30 µm coating (Supelco Inc., Bellefonte, PA, USA); 5) 10 µL syringe for internal standard addition; 6) a laboratory incubator (Thermo Fisher Scientific Inc., Waltham, MA, USA) with a constant temperature of 23 • C.

Preparation of CRDF Samples
To prepare the samples for VOCs emission analysis, the CRDF was pre-treated and ground in a 2SIEL90L2 grinding mill (Celia Indukta, Bielsko-Biała, Poland) to homogenize the sample to size <0.5 mm. Next, 10 g of bulk 3 subsamples were placed in a sealed 1000 mL glass vessels. An internal standard, 10 µg of 2-undecanone (Sigma-Aldrich, St. Louis, MO, USA), was added to the vessels to account for the variability in emissions and to aid VOC quantification. Each sealed sample was stored in a laboratory incubator at a constant temperature of 23 • C for 7 days, after which it was removed for sampling. The VOCs extraction was carried out from the headspace of sealed vessel, by the SPME.

Solid-Phase Microextraction
After placing the sealed vessel with the sample in a water bath with glycol preheated to 40 • C, a 3-component universal fiber coating (DVB/CAR/PDMS 50/30 µm) was introduced into the vessel headspace. The SPME exposure lasted 20 min, similarly to the types of coatings and extraction times used for VOC emissions from solid, porous matter. The DVB/CAR/PDMS 50/30 µm SPME coating is often recommended and used for exploratory work on VOC emissions from unknown sources [25,26,28]. The coating represents a mixture of polymers capable of extracting VOCs with a wide range of properties, i.e., suitable for the work with CRDF. No specific optimization was made on sampling time. However, it was chosen based on practical considerations and preliminary trials aiming at reliably extracting the greatest number of VOCs in a relatively short extraction.

Gas Chromatography with Mass Spectrometry
The separation, identification and quantification of VOCs adsorbed on the fiber was conducted using a GC coupled to a MS detector (Saturn 2000 MS Varian Chrompack, Palo Alto, CA, USA) with ZB-5 (Phenomenex, Torrance, CA, USA) column (30 m × 0.25 µm film × 0.25 mm i.d.). Chromatographic conditions were performed according to Calin-Sanchez et al. [32]. Scanning (1 scan/s) was performed in the range of 35-400 m/z using electron impact ionization at 70 eV [33]. The analyses were performed using helium as a carrier gas at a flow rate of 1.0 mL/min, in splitless mode in SPME, and with the following program for the oven temperature: 50 • C at the beginning; 4 • C/min to 130 • C; and 10 • C/min to 180 • C and 20 • C/min to 280 • C with a hold for 4 min. The injector was held at 220 • C.

Data Analysis
The VOCs emitted from CRDF samples were identified using three independent analytical methods: retention indices (RI), GC-MS retention times of authentic chemical standards, mass spectra of compounds [34] and comparison with authentic standards, if possible.
The retention index standards used in this study consisted of a mixture of aliphatic hydrocarbons ranging from C-7 through C-40 dissolved in hexane [34].
The use of internal standard enabled quantitative analysis of VOCs. It was carried out using the Mnova MS 12.0.1 software (Mestrelab Research, S.L., Santiago de Compostela, Spain) based on the retention time of individual compounds, through the integration of the peak area of the chromatogram. The percentage ratio of individual VOC was determined. VOC emissions (on per mass of CRDF basis) were estimated based on the recovered internal standard. All raw data were shown as Supplementary Materials.

Conclusions
In the analyzed CRDF (biochar) from municipal waste, 84 VOCs have been identified, including many that are toxic, e.g., derivatives of benzene or toluene. The highest emission was measured for nonanal, octanal, heptanal. The top 10 of the most emitted VOCs consisted almost 65% of total emissions. The mixture of emitted from CRDF VOCs differed from those emitted by other types of biochars, produced from different types of feedstock, and under different pyrolysis conditions. SPME provided a useful tool for characterizing VOC emissions from CRDF, a new potential fuel exemplifying the 'Waste to Carbon' concept in a circular, zero-waste economy.
Supplementary Materials: The following files have been submitted as supplementary materials in zipped folder "supplementary materials.zip": explanatory file "readme.docx", raw data in files "CRDF MS raw data.jdx; CRDF MS raw data.csv; CRDF peaks raw data.xlsx" and tables (Tables S1 and S2) in the file "Tables.xlsx". Funding: Authors would like to thank the Fulbright Foundation for funding the project titled "Research on pollutants emission from Carbonized Refuse Derived Fuel into the environment", completed at the Iowa State University. In addition, this project was partially supported by the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa. Project no. IOW05400 (Animal Production Systems: Synthesis of Methods to Determine Triple Bottom Line Sustainability from Findings of Reductionist Research) is sponsored by Hatch Act and State of Iowa funds."