There is increasing interest in the usage of biogas produced from wastewaters, anaerobic digesters and landfill sites as a source of green energy. Biogas generated from these types of sites require process monitoring due to contamination with siloxanes derived from hygiene products, detergent, antifoams, etc. Siloxanes are troublesome impurities in biogases in engine sources.

   The combustion of biogas can lead to deposits of silicon dioxide particles which can cause problems and damage different kind of gas engines or turbines through their build up and via corrosion.The objective of this study was to develop a method on a thermal desorption unit coupled to a gas chromatography and mass spectrometer (TD-GC-MS) to identify and quantify siloxanes present in a representative biogas sample with a low detection limit (µg/m³). Enabling the occurrence of siloxanes in a biogas matrix and efficiency of the removal processes present in the industry to be monitored.

E. Dooms^1*, K. Haerens¹, P. Segers¹

¹OLFASCAN, a brand of Milvus Consulting NV, Wondelgemkaai 159, 9000 Ghent, Belgium. *elisabeth.dooms@olfascan.com

  Competing interests: The author has declared that no competing interests exist.

   Academic editor: Carlos N Díaz.

   Content quality: This paper has been peer reviewed by at least two reviewers. See scientific committee here

   Citation: E. Dooms, K. Haereens, P. Segers, 2021, Laboratory and field study on the analyses of siloxanes in biogas by TD-GC-MS, 9th IWA Odour& VOC/Air Emission Conference, Bilbao, Spain, Olores.org.

   Copyright: 2021 Olores.org. Open Content Creative Commons license. It is allowed to download, reuse, reprint, modify, distribute, and/or copy articles in olores.org website, as long as the original authors and source are cited. No permission is required from the authors or the publishers.

   ISBN: 978-84-09-37032-0

   Keywords: TD-GC-MS; Carbon Graphitized sorbent; canister; Nalophane sampling bag; volatile methyl siloxanes; raw biogas.

Download Presentation

Video

Abstract

   There is increasing interest in the usage of biogas produced from wastewaters, anaerobic digesters and landfill sites as a source of green energy. Biogas generated from these types of sites require process monitoring due to contamination with siloxanes derived from hygiene products, detergent, antifoams, etc. Siloxanes are troublesome impurities in biogases in engine sources. The combustion of biogas can lead to deposits of silicon dioxide particles which can cause problems and damage different kind of gas engines or turbines through their build up and via corrosion.

   The objective of this study was to develop a method on a thermal desorption unit coupled to a gas chromatography and mass spectrometer (TD-GC-MS) to identify and quantify siloxanes present in a representative biogas sample with a low detection limit (µg/m³). Enabling the occurrence of siloxanes in a biogas matrix and efficiency of the removal processes present in the industry to be monitored.

   In collaboration with MARKES INTERNATIONAL, a method on the TD-GC-MS was evaluated with analytical standard over a range of 10 – 800 µg/mL prepared in methanol (RSD<10%). To retain the different kinds of heavier molecules in the best manner, different adsorption tubes were monitored and the Carbon Graphitized adsorption tubes were found to have the best reproducibility (RSD<10%). Safe sampling volume on Graphitized Carbon tubes was analysed and set at 2500 ml with a flowrate of maximal 100 ml/min. In addition, the most common sampling methods were evaluated in the OLFASCAN laboratory and in the field to analyse the interference of siloxanes in these receptacles and with other compounds present in the biogas matrix. Canisters, sampling bags and direct adsorption on Carbon Graphitized tubes were tested. The predominant siloxanes present in landfill biogas were D4, L2 and D5 and in digester biogas were D5 and D4. Less volatile siloxanes tend to adsorb onto the walls of the sampling bag and canister which results in a lower intensity.

1. Introduction

   Biogas is an environmentally friendly energy which gives the opportunity to decrease the use of greenhouse gas emissions. Biogas formation can occur spontaneously in landfills or under controlled conditions in digestion reactors by degrading organic material with microbes. The biogas produced during anaerobic digestion contains about 50-75% methane (CH₄), 20-40% carbon dioxide (CO₂) and other small variable quantities of other compounds such as water vapor (H₂O), oxygen (O₂), nitrogen (N₂), hydrogen sulphide (H₂S), ammonia (NH₃) and volatile organic compounds (VOC) depending of the composition of the feeds and conditions at which the digestion was carried out. In order to obtain a greater quality of the crude biogas and to be able to use it for electricity and heat production, this biogas should be purified from unwanted substances to avoid corrosion and mechanical wear of the equipment (Roels, 2010).

   The presence of some trace impurities, like siloxanes, ensure that continuous monitoring is necessary during biogas generation. Siloxanes are present in fuel, hygiene products, detergent, antifoams and biomedical products, therefore they’re found in waste flows hence they are frequently disposed in landfills. Siloxanes are organometallic compounds that consist of silicon, oxygen and methyl groups. The more volatile methylsiloxanes are released into the atmosphere through volatilization, but less volatile siloxanes remain in the sludge. The major siloxanes found in landfill biogas are in following order D4, L2, D5 and L3 (Table 1). In active landfills their levels are usually higher than in closed landfills. From all trace compounds in biogas, siloxanes are found to be difficult to control. The combustion of biogas can lead to deposits of silicon dioxide (SiO₂) particles which can cause problems and damage different kind of gas engines or turbines through their build up, via corrosion and can act as a thermal and electric insulator (Narros et al., 2009; Nyamukamba et al., 2020).

Table 1: Physical properties of typical siloxane compounds in biogas (Nyamukamba et al., 2020)

   For this reason the analysis of siloxanes in biogas can assist in determining whether a removal system is operating correctly. Collected samples from the biogas can by analysed by a thermal desorption unit coupled to a gas chromatography and mass spectrometer (TD-GC-MS) to identify and quantify siloxanes present with a low detection limit (µg/m³). In collaboration with MARKES INTERNATIONAL, a method on the TD-GC-MS was evaluated with analytical standard prepared in methanol. To retain the different kinds of heavier molecules in the best manner, Tenax TA and Carbon Graphitized sorbent tubes were monitored. In addition, the most common sampling methods were evaluated in the laboratory and in the field to analyse the interference of siloxanes with these receptacles and/ or other compounds present in the biogas matrix. Canisters, Nalophane sampling bags and direct adsorption on sorbent tubes were tested. The indirect impinger method is not being tested as this method requires complicated sampling protocol, it can take to up to 3 hours to analyse, it can be a source of sample contamination and artifacts can be produced and seen in the chromatogram (Kaszubska and Wzorek, 2018; Kim et al., 2013; Narros et al., 2009).

2. Materials and methods

1. Chemicals

   L2, L3, D3, L4, D4, L5 and D5 were supplied by Sigma-Aldrich with a purity of more than 97%. Toluene (99+% Acros Organics) was used as internal standard.

2. Preparation techniques

   In the laboratory the analyses for optimisation of the TD-GC-MS method were done by preparing solutions of siloxanes and toluene in methanol ranging from 10 – 800 µg/ml). The standard solution can be injected onto a sorbent tube or in a nalophane bag halfway filled with nitrogen (99,996% Air Products) by connecting the tube or bag to a sampling system. Through the sampling system a continuously flow of nitrogen (99,996% Air Products) at 50 ml/min is present. 1 µl of a solution can be injected with a microsyringe through a septum into this nitrogen flow. The solution evaporates and the components present adsorb onto the sorbent or end up in the sampling bag. After injection the continuous flow of nitrogen through the sorbent is held for 5 minutes. When a gaseous sample is made, the bag can be further filled with nitrogen (99,996% Air Products) after injection. Afterwards the sample in the sampling bag can be loaded onto a sorbent tube. For this, the tube is connected to the bag and the sample is drawn onto the sorbent with a GilAir pump with a flow rate of 100 ml/min.

3. Sampling techniques

   Three different sampling methods for sampling landfill gas are compared. The goal is to select the most efficient and reproducible methodology to analyse siloxanes on site.

• Canister: a clean Silonite coated canister from Entech was used to collect a whole air grab sample. The canister was put on vacuum and via a tubing into the feed line. The canister can fill itself with air when the valve is opened. In the laboratory the sample can be drawn over a Carbon Graphitised sorbent tube by pressurizing the canister with nitrogen (99,996% Air Products) to 15 psi. The tube can than be connected to the canister with a restrictor. To draw the biogas over the tube the canister valve is opened and the flow through the adsorption tube is measured with a Restek proflow 6000 as well as the sampling time to calculate the sampling volume. The tubes can then be further analysed with the TD-GC-MS in the laboratory.

• Sampling bag: a sampling bag made from Nalophane was tested. The bag is mounted into an airtight receptacle (barrel) and the bag fills itself with air by creating an under pressure in the receptacle according to the lung principle. After sampling the sampled air can be analysed in the laboratory by connecting a Carbon Graphitised sorbent tube to the bag and drawing the biogas onto tube with a GilAir sampling pump. A flow rate of 100 ml/min was used and the sampling volume was noted. After sampling, the sorbent tube can be analysed directly with the TD-GC-MS in the laboratory.

• Sorbent tube: a Carbon Graphitised sorbent tube was used with a GilAir sampling pump to draw biogas directly onto the sorbent. A flow rate of 100 ml/min was used and the sampling volume was noted. After sampling, the sorbent tube can be analysed directly with the TD-GC-MS in the laboratory.

4. Analysis

   Analyses were carried out using the MARKES INTERNATIONAL Ultra-xr (U-T6SUL-2S general purpose trap) coupled to the Shimadzu Nexis GC-2030 gas chromatogram with a BPX-Volatiles column (30 m x 0.25 mm, 1.4 µm film thickness) and Shimadzu GCMS-QP2020 NX mass spectrometer. Tubes were desorbed at 320°C during 10 minutes and transferred onto trap. The trap was heated to 330°C during 3 minutes. Split ratio was 4:12:54. The column oven temperature ramped from 35°C (3 minutes) to 280°C at 15 °C/min and held temperature for 6 minutes. Helium (99,99990% Air Products) is used as carrier gas. The mass range is set at 29 – 450 m/z. The compounds were identified by comparing the mass spectra with the selected ions obtained from the standard.

   To minimize contamination of siloxanes on trap and/or on column, after each batch a blank was analysed with the same settings as the samples. Tubes were cleaned with TC-20 (MARKES) at 335°C during 60 minutes with a flow of 50 ml/min of nitrogen (99,9999% Air Products). To control reproducibility of the analysis toluene (99+% Acros Organics) was used as internal standard. The relative standard deviation (RSD) of the area values was determined and should not exceed 10% for the siloxanes and 5% for toluene.

   A multipoint calibration was obtained by injecting a standard solution of 800 µg/ml in methanol on three different Carbon Graphitized tubes. The tubes are connected to a sampling system with a continuous flow of 50 ml/min of nitrogen (99,996% Air Products). 1 µl of the 800 µg/ml solution is injected with a microsyringe through a septum into this nitrogen flow where the solution evaporates and the compounds adsorb onto the sorbent. After injection, the flow of nitrogen through the sorbent is held for 5 minutes. The three spiked tubes are analysed with five repetitions of recollection on the TD-GC-MS. Recollection was done on the same tube.

   The safe sampling volume on Carbon Graphitized sorbent of the siloxanes is determined by analysing the breakthrough volume. Therefore different volumes, ranging from 100 – 5000 ml, of gaseous sample is drawn onto two linked Carbon Graphitized tubes and analysed with the TD-GC-MS.

5. Results and discussion

1. Calibration with recollection on TD-GC-MS

   The repeatability and linearity of the multipoint calibration was determined. The RSD of the peak area of each concentration was under 10% for siloxanes and under 5% for the internal standard. D3 and L3 are the more difficult siloxanes to analyse and quantify and D3 is expected to be unstable as described by Nyamukamba et al. (2020). Thence the RSD cannot be reduced to less than 5% for these components, but the linearity of all the siloxanes and the internal standard can be held at R²>0.999.

   The limit of detection (LOD) can be calculated for each component individually and is variable with the sampling volume used. Due to the low detection limit (µg/m³) small concentrations of the siloxane trace impurities can be detected with this method. This is a more sensitive technique than the on-line FTIR spectroscopy research of Hepburn et al. (2015) where the LOD is in mg/m³.

2. Sampling method

   Both Tenax as Carbon Graphitized sorbents were evaluated and the Carbon Graphitized tubes are found to have the most reproducible results for siloxanes (RSD<10%). Accordingly the breakthrough volume was determined for the seven siloxanes on Carbon Graphitized sorbent. For every volume a breakthrough of <5% was achieved and the safe sampling volume can be set at 2500 ml. The breakthrough volume and safe sampling volume will probably be higher, but in practice a higher volume than 2500 ml will not be used for sampling. During this evaluation different pump flows were used varying between 50 – 200 ml/min. Some unreproducible results (RSD>5%) were obtained when using 200 ml/min for 5000 ml and not with 100 ml/min. Further analyses are therefor performed with a flow rate of maximal 100 ml/min.

   Three different collection method were first analysed in the laboratory were a spiked sorbent tube is compared to a loaded tube from a gaseous sample in a Nalophane bag and in a canister (Fig. 1). A difference of intensity is detected for the less volatile compounds were D4, L4, L5 and D5 have a poorer recovery for the Nalophane sampling bag and canister. The results from Watson et al. have a similar observation with canisters were the response of L4, L5 and D5 decreases. This can be troublesome when sampling is done for biogas at landfills were D4 and D5 are expected to be more abundant.

Fig. 1.: Chromatogram comparing sampling method. Black: tube spiked with siloxanes and toluene; pink: gaseous sample in sampling bag of siloxanes and toluene loaded onto tube; blue: gaseous sample in canister loaded onto tube

   Fig.2 shows the concentration (µg/m³) of siloxanes measured of crude biogas taken at a landfill in Belgium in a zone that is closed since the early 2000, comparing the different sampling techniques used. Predominant siloxanes in this sample are D4, L2 and D5 as expected by Nyamukamba et al. (2020). Other siloxanes have a concentration that lies under the average LOD of 75 µg/m³, equivalent to 23 µg Si/m³. Samples taken with adsorption tubes in situ are 10% of the concentration detected with the sampling bag and canister. This may have been caused by the underpressure that was present at the sampling point causing loss during sampling. D4 and D5 have a higher intensity and L2 a lower intensity in the canister than those sampled in a nalophane bag and can be a result of adsorption of the less volatile compounds onto the wall of the nalophane bag. An average total siloxane concentration of 2883 µg/m³ was measured, equivalent to a total of 708 µg Si/m³. Arnold and Kajolinna (2008) reports that in raw biogas the total concentration of siloxanes can go up to 400 mg/m³ depending on site, but that an average less than 10 mg/m³ can be expected. The total concentration of siloxanes measured in both sampling bag as in canister are similar (RSD<10%). Sampling with a canister takes the preference due to the ease of sampling, the long shelf life of the sample and the safety of the employee when high concentrations of H­_2­S are present and it cannot diffuse through the wall. The results from Hayes et al. (2003) confirms that sampling with a canister has the easiest setup and shortest sampling time and doesn’t require an additional pump for sampling bags and sorbent tubes or an ice bath in the field for the impinger technique. Hayes et al. (2003) and Hepburna et al. (2015) notes that a canister is used to take a grab sample and will not map fluctuation of the concentration of siloxanes in the process. This can be prevented when a flow regulator is used which can be set to take a sample for several hours. Measuring with canister and TD-GC-MS will not give a real time result as seen with the on-line FTIR spectroscopy, but by optimizing the method a quick result should be obtained to follow up activated carbon filters.

Fig. 2.: Comparison of sampling method with biogas from landfill. Blue: sampling directly onto Carbon Graphitized adsorption tube; Orange: sampling via Nalophane sampling bag; Grey: Sampling via Silonite canister.

   In Fig.3 the results of raw biogas from a digester in Belgium can be observed. Here an overpressure was present in the pipeline of the sampling point. Therefore direct sampling onto tube was possible and gives a better result in comparison with previous observations. In this sample D5 and D4 are present and a less variable (RSD<10%) result between the sampling techniques is measured. The effect of siloxanes sticking to the wall of the canister is also not observed in the research of Piechota (2021). Therefore the easiest sampling technique, which is with canister, takes preference when further samples will be taken. The linear siloxanes and D3 detected have a concentration that lies under the average LOD of 8 µg/m³, equivalent to 2 µg Si/m³, and will not be further discussed. The total siloxane mean concentration measured is 930 µg/m³, equivalent to a total of 23 µg Si/m³.

Fig. 3.: Comparison of sampling method with biogas from digester. Blue: sampling directly onto Carbon Graphitized adsorption tube; Orange: sampling via Nalophane sampling bag; Grey: Sampling via Silonite canister.

6. Conclusions

   Quantifying siloxanes in an air sample with a TD-GC-MS is possible by using a multipoint calibration with recollection. The siloxanes L3 and D3 give a lower reproducibility with RSD<10%.

   When sampling onto Carbon Graphitized adsorption tubes, a flow of maximal 100 ml/min should be used and a safe sampling volume of 2500 ml was obtained.

   At site sampling with adsorption tubes is favourable in comparison with sampling bags and canister when the sampling point is in over pressure. D4 and D5 are found to be the most abundant siloxanes in different types of biogas and are expected to adsorb onto the walls of the bag and canister as seen in the laboratory. Therefor quantification of less volatile siloxanes can be difficult. But during the analysis with raw biogas, the concentration of these less volatiles where similar to the concentrations measured with direct adsorption (RSD<10%). Also the pressure present at the sampling point can have an influence on the sampling method and taking samples on site with adsorption tubes is more inconvenient. The preferred method to sample is with canister for various reasons, such as the ease of sampling and safety because high concentrations of H­_2­S cannot diffuse.

   Further studies will follow around the interference of moisture with these methods.

7. References

Arnold M. and Kajolinna T.. 2008. On-line measurement and removal of biogas trace compounds. VTT Technical Research Centre of Finland.

Hayes H.C., Graening G.J., Saeed S., Kao S.. 2003. A summary of available analytical methods fot the dermination of siloxanes in biogas.

Hepburna C.A., Valeb P., Brownc A.S., Simmsd N.J., McAdama E.J.. 2015. Development of on-line FTIR spectroscopy for siloxane detection in biogas to enhance carbon contactor management. Talanta 2015. 141, 128-136.

Kaszubska M. and Wzorek M.. 2018. Development of Measurement Techniques for Siloxanes in Landfill Gas. International Journal of Thermal and Environmental Engineering. 16(2), 91-96.

Narros, A., Del Peso, M.I, Mele, G., Vinot, M., Fernandez, E., Rodriquez, M.E. 2009. Determination of siloxanes in landfill gas by adsorption on Tenax tubes and TD-GC-MS. CISA Publisher (Italy).

Nyamukamba, P., Mukumba, P., Chikukwa, E.S., Makaka, G. 2020. Biogas Upgrading Approaches with Special Focus on Siloxane Removal - A Review. Energies 2020. 13.

Pichota G.. 2021. Siloxanes in Biogas: Approaches of Sampling Procedure and

GC-MS Method Determination. Molecules. 26, 1953.

Roels, A. 2010. Biogas in Vlaanderen: een SWOT-analyse. Universiteit Gent Faculteit Economie en Bedrijfkunde (Gent, Belgium).

Watson, N., Morris, P., Grosshans, P. Monitoring Siloxanes in biogas using thermal desorption tube sampling. Markes International.

Kim N., Chun S., Cha D.K., Kim C.. 2013. Determination of Siloxanes in Biogas by Solid-phase Adsorption on Activated Carbon. Bulletin of the Korean Chemical Society. 34 (8), 2353-2357.