SPME Method for Chemical Analysis of Heavy Organic Trace Compounds in Synthesis Gas

INOM

EXAMENSARBETE TEKNIK, GRUNDNIVÅ, 15 HP

STOCKHOLM SVERIGE 2020,

SPME Method for Chemical

Analysis of Heavy Organic Trace Compounds in Synthesis Gas

An Experimental Manual for a Proof of Concept Study

SIMON EDIN, FELISE ELEMIA FREIRE,

CHANG HO LEE

SPME Method for Chemical Analysis of Heavy Organic Trace Compounds in

Synthesis Gas

An Experimental Manual for a Proof of Concept Study

Examiner:

Klas Engvall

Supervisors:

Efthymios Kantarelis Lars Pettersson

Authors:

Simon Edin Felise Elemia Freire

Chang Ho Lee

Sammanfattning

I nuläget finns det ingen kommersiell metod för att snabbt extrahera och analysera spår av tjär- komponenter i gasströmmar. Tidigare har solid phase microextraction (SPME) med polydimetylsiloxan (PDMS) som fast fas undersökts som en möjlig kandidat då den ej kräver lösningsmedel och kan enkelt återanvändas. Detta projekt hade som mål att bevisa att SPME kan anpassas tillräckligt känsligt för att analysera spår av tjära i syngas med en koncentration på mindre än 0,1 mg/Nm ​

. På grund av komplikationer som uppstod i samband med Covid-19 pandemin var det inte möjligt att utföra den praktiska delen av projektet. Istället så har en design tagits fram för ett koncept som beskriver hur man kan genomföra den praktiska delen.

Designen beskriver en två-kammare lösning som kan användas för att ta prover från syngas som kommer direkt från en förgasare. Proverna tas vid temperaturer om 60 °C och 125 °C för att uppnå maximal känslighet. En uppsättning kommersiellt tillgängliga sorbentrör används för att kontrollera resultaten från SPME.

Nyckelord: ​SPME, PDMS, tjära, biomassaförgasning, syngas, experimentell design.

Summary Abstract

There currently exists no commercialized method for rapid sampling and analysis of trace tar in gas streams. Solid phase microextraction (SPME) with a polydimethylsiloxane (PDMS) solid phase has been previously investigated as a possible candidate due to its solvent-free nature and reusability. This project set out to deliver a proof of concept study to test whether SPME can be sufficiently tuned to analyse trace tar content in syngas below the concentration of 0.1 mg/Nm ​

. Due to complications that arose from the Covid-19 pandemic, it was unfeasible to carry out the practical elements of the project. Instead a concept design for carrying out such a study has been successfully developed. This design envisions a two-chamber setup able to sample syngas directly from a gasifier at 60 °C and 125 °C respectively and is illustrated in the text. It utilizes commercially available solvent tubes to cross-check and verify the SPME results.

Key Words: ​SPME, PDMS, tar, biomass gasification, syngas, experimental manual

Sammanfattning 2

Summary Abstract 3

Terminology and Abbreviations 5

Acknowledgement 6

Introduction 7

Purpose and Scope 9

Background Information 11

Biomass Gasification 11

Tar 12

SPME and Adsorption 14

Sorbent Tubes 17

Method 19

Theoretical 19

Empirical 19

Results 21

Materials Required 21

Before Assembly 21

Assembly 22

Before Initial Sampling 23

Experimental Procedure 23

Before Analysis 24

Analysis 24

Illustrations 26

Discussion 29

Conclusion 31

References 32

Terminology and Abbreviations

SPME Solid Phase Microextraction

SPA Solid Phase Adsorption

PDMS Polydimethylsiloxane

PA Polyacrylate

DCM Dichloromethane

GC-FID Gas Chromatography - Flame Ionisation Detector

Acknowledgement

The authors wish to extend their deepest gratitude to their supervisors ​Efthymios Kantarelis

and Lars Petterson, lab supervisor Iraj Bavarian, as well as their examinator Klas Engvall for their

guidance, their support, and their patience throughout the project. They invested much time in

personal meetings, in the lab watching over the authors, and spent numerous hours over video

chat giving invaluable advice and feedback.

Introduction

The world of today uses ever more energy, derived mostly from non-sustainable resources. In 2016, a mere fourteen percent of all global energy consumption came from renewable energy sources (World Bioenergy Association, 2019). Considering the environmental impact of a growing number of sophisticated lifestyles, being able to utilize more sustainable energy sources is essential to mankind’s survival. One such alternative that has shown great growth in the past years and has significant potential is synthesis gas produced by biomass gasification (International Energy Agency, 2020).

The market for biogas is relatively young but has grown exponentially during the past years, in part because it has seen substantial support from governments, such as EU regulations aimed at cutting greenhouse gas emissions. Much of this energy is today used for heating and electricity (Scarlat, Dallemand, and Fahl, 2018). However, there are far more applications possible such as the production of syngas, which is made up mainly of CO ​

, CO, H ​

, and CH ​

(Brage et al. 1997), the latter of which can be used as engine fuel and thus be utilized for transportation.

Synthesis gas can be used to produce methanol and synthetic fuel, the latter of which is used as fuel in combustion engines in transport systems. Syngas itself also has a wide variety of applications, for example, the gas can be upgraded to concentrated hydrogen gas for use in fuel cells. In comparison to the biogas market, syngas from biomass has not been commercialized to the same extent due to a by-product that causes problems in its utilization.

The use of syngas both to power stoves but also cars and fuel cells is not without its drawbacks.

During the heating of the raw materials, which are commonly hard biomass as in wood and

wood shavings, the lignin inside will break apart and form smaller aromatic compounds (WBA,

syngas. Unless kept at high temperature during the whole journey from production to end consumption the tar will condense somewhere along the way. Condensed tar can block pipes, cake filters, and coke catalysts causing expensive and unnecessary maintenance and replacements(Engvall, Liliedahl, and Dahlquist, 2013).

The maximum levels of tar allowed in syngas varies for the various applications, but this study will follow the tar limits set out by Ahmadi Svensson (2013) which is below 0.1 mg/Nm ​

. Currently there are two methods in place that can come close to this limit and both of them have significant drawbacks. The first is cold trapping, which requires large samples and ample amounts of time as well as the use of dichloromethane (DCM) for eluting the analyte (Brage et al. 1997). The second method is solid phase adsorption (SPA), which does not require large sample sizes but also uses DCM and whose detection limit varies from 0.5 to 3.0 mg/Nm ​

and thus is not accurate enough to confidently determine syngas purity (Woolcock et al. 2013). It is therefore of high interest to develop a method to monitor syngas that is unintrusive to the analyte, of high accuracy, high repeatability, and high speed while also being easy to use. A promising alternative is solid phase microextraction (SPME).

Purpose and Scope

Due to the extraordinary circumstances of the Covid-19 Virus Pandemic the purpose of this study had to be reevaluated from its original design. The original purpose was to investigate a viable method, SPME, for measuring tar concentration in syngas. As it is no longer possible to carry out laboratory experiments at this time this thesis will instead endeavor to develop a thorough theoretical guideline that can be used as a manual for carrying out the experiments once the crisis has passed.

There are several reasons why SPME is a promising candidate. The extracted volume from the sample is microscopic and thus negligible in comparison to the overall sample (Ouyang and Jiang, 2017). It is entirely independent of solvents, eliminating the use of eluents such as DCM or other potentially harmful chemicals, making it a quick and simple screening tool (Prosen and Zupančič-Kralj, 1999). Additionally the absence of eluents increases the efficiency and accuracy of the following chromatographic separation making it possible to analyze tiny concentrations of tar rapidly, with no sample preparation required.

Being able to quickly and confidently determine syngas purity will enable the optimization of biomass gasification processes for tar-free syngas production, allowing it to become more accessible and reliable as an alternative energy source. This will contribute to increase the share of renewables in global energy consumption and help the world reach its emission goals.

This study therefore poses the following research question:

How can heavy organic trace compounds (tar) at a concentration below 0.1 mg/Nm ​

reliably be detected using SPME?

For the sake of this study, when measuring tar concentration the focus will be on the following twelve compounds also investigated by Ahmadi, Svensson, and Engvall (2013), presented below in order of increasing molecular weight.

1. Benzene 2. Toluene 3. Indane 4. Indene 5. Naphthalene 6. Acenaphthylene 7. Acenaphthene 8. Fluorene 9. Phenanthrene 10. Anthracene 11. Fluoranthene 12. Pyrene

This study will not investigate other compounds that could be an issue in syngas application, save the twelve listed above. It was to also focus exclusively on birch wood as a raw material.

Other sources of syngas may cause variations in the composition, which will not be taken into account. Lastly, the method will be validated with sorbent tubes, which will then be eluted with DCM for the GC-FID analysis.

Background Information

This chapter will present the theoretical background necessary for the project. It will begin with

an overview of biomass gasification and its techniques. This is followed by a review of tar and

its origins in syngas. It will then conclude with a detailed presentation of SPME, its method and

underlying concepts.

Biomass Gasification

The core of biomass gasification lies in the pyrolysis of the biomass. This is when the raw materials are processed with gasification agents, which is usually air and steam, under high temperatures ranging from 500 ​ °C ​ to 1 000 ​ °C ​ depending on gasifier design and catalysts.

Temperature will have a significant impact both on the molar fraction of the different components in syngas as well as tar concentration. It has been observed that a temperature between 700 ​ °C to 800 ​ °C will produce this highest fraction of CH ​

. Higher temperatures will favor the production of H ​

and CO at the expense of CH ​

and CO ​

as thermal cracking of the carbon compounds will occur (Fagbemi, Khezami, and Capart, 2001).

Looking at the process design, Ahmadi Svensson (2013) lists four different types of gasifiers.

These are updraft, downdraft, fluidized bed, and entrained flow. An updraft gasifier adds the fuel from the top and the gasification agent from the bottom, allowing the streams to move in opposite directions. This is generally marked by a lower gas quality. The downdraft gasifier also feeds the fuel from the top, but in comparison the gasification agent is fed through the middle and the final gas is tapped out through the bottom. The gas quality in this case is described as decent. The fluidized bed gasifier on the other hand provides a high degree of mixing between the fuel and the gasification agent allowing for a high reaction yield and a good gas quality.

Finally, the entrained flow gasifier is marked by high operating temperatures exceeding 1000

°C, which allows it to produce a high-energy, almost tar-free product.

The result of the gasification is syngas of varying composition and purity. The most significant products are, as mentioned earlier, CO, CO ​

, H ​

, and CH ​

with impurities consisting of mainly tar and sulfides. Wood has been shown to have some of the highest tar yield, up to 35 mass%

(Fagbemi, Khezami, and Capart, 2001) ​ . This again underlines the importance for having a

reliable detection method.

Tar

There are a few variations in language but Yung, Jablonski, and Magrini-Bair (2009) define tar neatly as “the condensable fraction of organic gasification products, largely consisting of aromatic compounds, such as benzene, toluene, and naphthalene.” Below you will find Table 1, listing all the tar compounds under investigation in this study, as well as some of their physical properties that are of interest. Moreover, their individual responses during GC-FID analysis and the temperature at which these responses occurred are also presented.

Figure 1 - Present are all of the twelve organic compounds to be studied in this report in order of molecular weight, from lowest (benzene) to highest (pyrene). All pictures of the structural formulae were taken from ChemSpider.

Table 1 - Names, molar masses, and boiling points of the tar compounds under investigation (PubChem, 2020) as well as the temperature for their strongest recorded response using SPME as sampling method and GC-FID for analysis in addition to approximate analysis responses (Ahmadi, Elm Svensson, and Engvall, 2013).

Name

Molar Mass

[g/mol] Boiling/Dew Point [°C]

T [°C] for Strongest Response

Approximate response

[μV*s]

Benzene 78.1 80.0 20 300

Toluene 92.1 110.6 20 1 000

Indane 118.2 177.9 20 1 550

Indene 116.2 182.0 20 2 250

Naphthalene 128.2 217.9 60 7 250

Acenaphthylene 152.2 280.0 100 5 000

Acenaphthene 154.2 279.0 100 4 500

Fluorene 166.2 295.0 100 6 500

Phenanthrene 178.2 340.0 125 16 000

Anthracene 178.2 339.9 125 7 500

Fluoranthene 202.3 384.0 125 5 000

Pyrene 202.3 404.0 125 5 000

It can be seen in Table 1 that both the response intensity as well as the difference in temperature for a maximal response varies significantly between the compounds, up to a difference of 105 °C. This does provide a challenge as the measurement quantities this study is focusing on will be tiny and therefore the sampling temperature needs to be optimized in order to maximize responses. Given the observed spread, it is recommendable to use more than one sampling temperature in order to see responses for all twelve components.

SPME and Adsorption

Solid phase microextraction is a sampling method that relies on adsorption in order to extract the analyte particles from the sample. The SPME device resembles a syringe, inside in which a silicone fiber covered with a polymer matrix is hidden. This matrix is exposed directly into the analyte, in this case the gas stream, for a set period of time. The analyte particles will attach to the polymer on the needle of the SPME-device via weak intermolecular interactions. The particles will be competing for binding sites until saturation. However, even after saturation is achieved, particles of higher affinity will replace particles of lower affinity. It is therefore recommendable to not expose the polymer too long so as not to eskew the relative concentrations in the sample (Górecki, Yu, and Pawliszyn, 1999). Following the sampling process, the polymer matrix is transferred to a GC-FID, where the analyte is desorbed from the matrix by a carrier gas and transferred through the column. One major advantage of SPME is that the polymer matrix will be reconditioned as the analyte is desorbed into the carrier gas at high temperatures of 300 °C and is therefore immediately reusable (Ahmadi Svensson, 2013).

The method cuts down on time spent on equipment and sample preparation while also eliminating the use of solvents for washing.

Since the SPME method is based on adsorption, temperature is a key factor in the extraction

process. This is clearly illustrated in the method for modelling adsorption, the basis for which is

the Antoine equation for modelling vapor pressure. The equation relates the logarithm of the

vapor pressure (P) to a set of three compound specific variables (A,B, and C) at a certain

temperature (T). The model is presented as equation 1 below.

l og P = A −

(eq 1)

The relation between the amount of analyte adsorbed onto the matrix and the total amount of analyte in the sample is labeled as the partition coefficient ​K. ​It is based on the Nernst distribution law, as defined by Martos and Pawlizyn (1997) and presented in equation 2 below.

K =

(eq 2)

air

C_fiber

Wherein C

is the concentration of analyte adsorbed onto the polymer matrix and C

is the concentration of analyte in the surrounding gas. Martos and Pawliszyn (1997) move on to prove that an estimation of K can be found by the below presented model.

og K l = a ( )

+ b (eq 3)

The constants ​a​ and ​b​ are defined as

a =

(eq 4)

b = l og(

)

i *

−

(eq 5)

Wherein H Δ

is the heat of vaporization of the modelled compound, R is the gas constant of the Ideal Gas Law using appropriate units, p

is the analyte vapor pressure at the temperature

, and is the activity coefficient. Using this model, the amount of analyte in the sample at

T

γ

a given temperature can be calculated for any of the identified compounds.

Ahmadi Svensson (2013) applies the Langmuir adsorption model in combination with the above

equations in order to gain more insight into the adsorption process itself, as well as the relation

of tar compound in the sample. The Langmuir adsorption model stipulates a set of assumptions:

firstly that there will only be surface adsorption in a monolayer manner. Secondly, all adsorption sites are identical and allow bonding with only one species. Thirdly, these sites do not influence each other and their individual bonding ability. The actual model is presented as equation 6 below.

=

(eq 6)

max

+

max

Wherein C is the concentration in the sample, Γ is the amount adsorbed onto the polymer matrix, Γ

is the maximum amount that can be adsorbed as C increases, and finally L is the Langmuir equilibrium constant.

However in another paper published around the same time Ahmadi, Svensson, and Engvall (2013) put the Langmuir models ability to the test against the competing Freundlich model, which takes into account adsorption in multilayers, and found that the latter exhibits a much higher degree of conformity to the data. The R ​

value of the Freundlich parameters is at least 0.995 compared with the Langmuir model’s highest R ​

of 0.632 . Therefore the Freundlich model they used will be presented below in equation 7.

x /m = Lp

(eq 7)

Here, x stands for the mass of the tar compounds in the sample that are adsorbed, while m is

the mass of the polymer matrix, p is the vapor pressure of the tar compounds at equilibrium,

while L and n are adsorbate and adsorbent specific constants at a given temperature. Having

estimated a value for ​K ​one can combine equation 7 with equation 2 in order to get a relation

between the concentration adsorbed to the SPME fiber and the actual concentration of tar in

the sampe.

Aside from the temperature, another major factor affecting the adsorption process is the choice of polymer matrix. Different matrices will be more prone to adsorb different compounds depending on their individual compositions. The two main matrices in use are poly- dimethylsiloxane (PDMS) and polyacrylate (PA), each with their own adsorption coefficient. For PDMS the adsorption coefficient is relatively close to that of a nonpolar organic solvent. This implies that PDMS allows fast adsorption of nonpolar compounds onto the matrix (Martos and Pawliszyn, 1997). This study aims to examine the adsorption of aromatic compounds, and will therefore make use of the PDMS polymer matrix in order to achieve maximum adsorption.

When all of the factors are optimized it should be possible to detect quantities below 0,1 mg/Nm ​

(Ahmadi, Svensson, and Engvall, 2013).

Sorbent Tubes

It is important to be able to monitor whether or not the SPME is taking correct readings in order to make any conclusions. The idea is therefore to use sorbent tubes, a commercially available and certified gas analyser, suitable to sample tar compounds from gas. These tubes must at all times stay in a vertical position for best results to be obtained as instructed by the supplier ​Scantec Nordic​. The samples collected in the tubes will have to be desorbed using DCM. The ID numbers of the tubes intended for use in this experiment are detailed in Table 2.

Additionally the tubes will require carefully monitored airflow. The supplier lists the upper

airflow limit for the heavy compound sorbent tube as 5 litres per minute. Volume flow in the

lighter compound sorbent tube should not exceed 1 litre per minute as instructed by the

manufacturer (SKC FAQ, 2020). The readings from the sorbent tubes will be used to determine

whether the results from the SPME samples are accurate and reliable.

Method

This chapter will describe the process by which the instructions for carrying out the experiments have been designed. It is subdivided into two sections, theoretical and empirical.

The theoretical section will give an overview on what design decisions are based on the literature. The empirical section will elaborate on the parts of the design that have been deduced from carrying out laboratory work.

Theoretical

The most important consideration to be answered by the theory is the temperature at which the sampling should be carried out. As stated earlier, the temperature is one of the most crucial factors to consider when optimizing the adsorption and therefore the limit of detection. In short, a high temperature will prevent condensation as the gas is able to hold more of the individual tar compounds, but will also make it more likely for the compounds to desorb from the polymer matrix. Regarding the responses in Table 1 and the findings by Ahmadi, Elm Svensson, and Engvall (2013) it is estimated that, for a two-chamber analysis, the optimal temperatures should be 60 °C and 125 °C. The same paper provided a compelling argument that the sample time should be 10 minutes. This should allow for enough time for the analytes to adsorb to the polymer matrix in sufficient quantity while also being short enough to be operationally practical.

Next the choice of polymer matrix must be made. This choice is also done with the intent of

maximizing the sensitivity of the analysis. As presented earlier, PDMS is the more nonpolar of

the commonly available matrices. Since the analytes are aromatic hydrocarbons, PDMS is the

matrix chosen to be used in the experiment.

Empirical

The setup is designed based on 6 mm piping. Should other pipe sizes be used, a need might arise for more materials such as insulation and heating tubes. The number of fittings required is based on the assumption that two fittings are required per individual pipe.

The biomass gasifier that was intended to be used for the pilot-scale test run which this project aimed to investigate is a 5 kW atmospheric bubbling fluidized bed gasifier, with a hot gas ceramic filter and a catalytic reactor operating at 800 ​ °C. Excluding the ceramic filter to remove particulate matter, there are no cleaning processes to remove impurities. In the operating temperature of the gasifier, phenols are known to decompose and no longer contribute to tar composition (Engvall, 2020, personal communications).

For the initial setup, everything was planned to be constructed inside a fume hood. A key factor to consider here is the accessibility of the SPME access points as well as the sorbent tubes. The sorbent tubes as well as the SPME syringes need to be attached to and detached from the setup at the start and finish of every experiment. For this purpose the setup was designed with the SPME access points at the top of the center gas chambers and sorbent tubes as easily detachable units on the sides. The gas feed was placed at the bottom of the setup so as not to obstruct access.

The analysis part of the instructions is kept to a bare minimum as this will be a standard GC-FID analysis and not at all unique to this experiment. The use of DCM as an eluent for the sorbent tube sample is a simple choice as it is a nonpolar solvent that shall work well in the gas chromatograph with the nonpolar tar compounds and will not interfere with the actual analysis.

Flowmeters, also known as rotameters, are installed at the entrance and exit to the system in

designed to sample at a specific flow rate. For this setup the flow rate will be capped at 1 litre/minute because this is the maximum gas flow the more fragile tubes can withstand.

The instructions presented in the next chapter comprise the steps that were planned to be

carried out before the coronavirus pandemic of 2020 forced a change of plans.

Results

This chapter gives the instructions on how to carry out the experiments in order to verify whether or not tar compounds can be detected in quantities below 0.1 mg/Nm ​

. It will start with listing the materials required in Table 2. This is followed by detailed, step-by-step instructions on how to set up the equipment and carry out the experiments. The chapter is then concluded with three illustrations intended to enhance the text’s accessibility.

Warning ​: Always ensure that the assembly area is well ventilated in case of a gas leak.

Materials Required

The materials and the amounts required described in Table 2 shall suffice for carrying out the experiment. In areas where it says ​circumstantial, ​the amount required is highly dependent on experimental conditions, such as size of the fume hood and distance to the gas source, and can therefore vary significantly. Hence no estimated quantity is given. Other amounts such as fittings, tools, and piping sizes may vary depending on the individual design adaptation.

Two types of instructions will be shown in parentheses below the subheadings. The first is: (In recommended order) which designates the order that the authors determined is the most efficient to carry out a set of instructions. The second is: (In order) which designates the order that the authors determined is crucial to follow in order to carry out the experiment successfully.

Before Assembly

❖ Read through the complete instructions carefully.

❖ Make sure all materials listed in Table 2 are accessible and available in required quantities.

❖ Measure, cut, and bend the pipes into required length and shape. The shorter the better

❖ Measure heating tube effect. Knowing this will allow better control over the temperature later.

❖ It is important to have a designated sampler to carry out the actual SPME sampling.

Having a designated sampler will reduce the systematic error that will occur from having different samplers.

Assembly

(in recommended order)

❖ Attach heating cables to each of the sampling chambers. The cables should cover the majority of the chambers’ surface to allow for maximum heat transfer.

❖ Encase the chambers with the heating cables in insulation material. This will help stabilize the temperature in the chambers.

❖ Assemble the chambers as seen in Figure 2 using the appropriate piping. Feed gas (compressed air or nitrogen) through the chambers and check for leaks.

➢ If leaks are found tighten the identified fittings and test again.

➢ If no leaks are found proceed with assembly.

❖ Insert thermocouples at all points marked in Figure 2. This will allow close monitoring of temperature changes in the gas.

❖ Insert rotameter at points marked in illustration. This will give information over the gas volume entering the chambers.

❖ Surround the wide tubes with heating cables and encase them with insulation material.

These will stabilize the temperature in the sorbent tubes.

❖ Attach sorbent tubes to each other and then attach the heavy-compound tube to the sample chamber gas outlet, using the supports if necessary. This will prevent the light-compound tube being clogged with heavy compounds.

❖ Place the heated and insulated wide tube around the sorbent tubes. This will prevent condensation in the sorbent tubes.

❖ Connect the parts of the setup as seen in Figure 2.

❖ Attach heating tubes to exposed pipes. Ideally, encase pipes in insulation. This shall keep the temperature in the tubes high and prevent condensation.

❖ Repeatedly check for leaks during the setup to avoid unnecessary deconstruction.

Before Initial Sampling

(in order)

❖ Make sure the sorbent valves are set to vent the gas as in Figure 4.

❖ Feed the system with gas (compressed air from a pump or nitrogen) via the setup inlet and check for leaks.

➢ If leaks are found, tighten the identified fittings and test again.

➢ If no leaks are found, proceed with the initial run.

❖ Make sure all thermocouples, heating tubes, and insulation are functioning properly.

❖ Monitor the gas temperature. The temperature should match the ones indicated in Figure 3.

➢ If the temperature is too high - decrease the power to the heating tubes.

➢ If the temperature is too low - increase the power to the heating tubes and check the insulation.

➢ If the temperature remains unstable - adjust the insulation.

❖ Attach the gas inlet of the setup to the outlet of the gas source (ex. gasifier).

Experimental Procedure

(in order)

❖ Feed a stable stream of gas from the gas source into the sampling setup. The stream shall not exceed 1 litre/minute.

❖ Wait until the temperature of the gas has stabilized in the entire setup at the levels indicated in Figure 3.

❖ The following steps should be executed as simultaneously as possible:

➢ Insert one SPME syringe into each of the sampling chambers and expose the

➢ Open the valves to vent the gas through the sorbent tubes

➢ Start recording time.

❖ Wait until the end of the sampling time (10 minutes).

❖ At the end of the sampling time:

➢ Retract the SPME matrices into the syringe and extract the syringe.

➢ Close the sorbent valves.

➢ Detach the sorbent tubes.

❖ Store sorbent tubes and SPME-syringes as cold as possible to prevent evaporation of the analyte.

Before Analysis

❖ Elute the sorbent tubes using DCM (Bavarian, 2020, personal communication). Now they can be stored at low temperatures (-15 °C) for an extended period of time. Discard the used tubes appropriately.

❖ Mix a standard containing DCM and known concentrations (recommended 0.1 mg/Nm ​

) of all twelve compounds under investigation.

Analysis

(in recommended order)

❖ Run the standard using a GC-FID. Run analysis and record results.

❖ Insert SPME-syringe into the appropriate access point on the GC-FID and expose the fiber. Run analysis and record results. The syringe will now be ready for the next trial.

❖ Insert eluted sample from sorbent tubes into appropriate access points on the GC-FID.

Run analysis and record results.

❖ The resulting samples should be analyzed using available software.

Table 2 - The materials and the amounts required are derived as described in the Empirical section of the Method.

Materials Amount Required

Sampling Chambers 2

Piping Circumstantial

Fitting 2/pipe

Pipe Cutters 1

Isolation Materials Circumstantial

Heating Cables 5 (min)

4-way pipe splits 2

Valves 4 On-Off + 1 Regulator

Thermocouples (3.0 mm) 3

Thermocouples (0.5 mm) 2

Thermocouple Access Points 1/Thermocouple

Thermocouple reader Depends on type of reader

Flowmeters 3

Wide Tube 2

Support Circumstantial

Gas Source 1

Wrench 2 (min)

SPME needles 2

Sorbent Tube Anasorb CSC ID 226-01 (Light Compounds)

2/trial

Rubber fittings matching sorbent Tube (Light Compounds)

4

Sorbent Tube Puff ID 226-124 (Heavy Compounds)

2/trial

Rubber fittings matching sorbent Tube (Heavy Compounds)

4

Illustrations

Figure 2 - The experimental setup as envisioned. The gas enters via the center pipe, is distributed equally between the two chambers, and then either vented out of the system or proceeds through the sorbent tubes, after which it then leaves the system. The sorbent tubes shall always maintain a vertical position for which they might require additional support. Their makeup is shown in greater detail in Figure 4.

Figure 3 - A temperature profile overlay for the system depicted in Figure 2. It is important that the gas in the individual sampling chambers hold the same temperature as the temperature of the gas passing through the sorbent tubes in connection with them to ensure that the samples are taken at the same conditions and thus analysis is comparable.

Figure 4 - The layout of the sorbent tubes. The tubes are connected via an adaptor piping as they will have different sizes. The Heavy Compound Tube shall be mounted closest to the sample chamber to prevent the heavy compounds from clogging the Light Compound Tube. The gas then flows from the Light Compound Tube to the flowmeter. Both tubes are inserted into the heated and insulated mantle to control the temperature. The tubes shall always

maintain a vertical position to ensure that the gas spreads equally through the adsorbent and prevent channeling.

Discussion

This chapter will discuss the manual presented under the Results heading. It will evaluate its strengths and weaknesses, as well as suggest possible adaptations to the method.

A strength of this manual is that much of the initial work required for the setup has already been tested by the authors. When the laboratory then became inaccessible due to the Covid-19 pandemic, a majority of the steps had already been empirically determined. Therein though lies also a weakness as this particular setup has never been tested in practice in full. However, the design is based on several previous studies, which have been successful, and therefore the authors are confident that the setup is realistically feasible.

Another positive feature of this setup is its rather small size. This allows the user to construct it within a fume hood, therefore minimizing the risk of a gas exposure. However, while the setup might not be excessively wide, it sports some height, which could pose a challenge when assembling inside the fume hood. The authors recommend using a stool to stand on.

It should be noted that maintaining the ideal temperature profile of the system as uniform as shown in Figure 3 might pose a significant challenge. This is because the metal piping has a high thermal conductivity and the gas temperature will drop quickly once it leaves the gasifier.

Proper insulation and heating equipment will therefore be of utmost importance. Finding the ideal practical management of the temperature will depend on the quality of insulation used, as well as the effect of the heating tubes and the accuracy of the thermocouples.

Elaborating on the thermocouples it is important to keep an eye on the effect they have on the

gas flow. If the thermocouples are too large relative to the pipe diameter, they will obstruct the

flow and provide potential points for condensation. A test run has not been done to confirm if

the sizes presented in Table 2 are small enough to have no significant impact on the gas stream.

Keeping on the topic of temperature, the heating of the sorbent tubes will most likely provide an additional challenge. The current system is designed so the internal sorbent tubes can quickly be removed from the heating mantle. The optimal size of the mantle as well its insulation could not be empirically determined by the authors before laboratory access ceased.

Other foreseeable challenges include the actual detachment of sorbent tubes from the sampling chambers between experiments as well as maintaining a consistent heat profile in the mantle. This is further complicated since taking the actual temperature inside the tubes will be impossible without compromising the adsorbents. Simultaneously, it is important to keep the temperature stable throughout the system to allow proper comparison between the observed results.

A limitation that arises with using SPME is that it cannot have an internal standard. The standard cannot be put through the same extraction process as the sample gas. This means that an external standard is required in order to quantify results from the samples, which will carry over systematic errors inherent to an external standard.

A beneficial improvement would be to add a third sampling chamber to the setup. This would allow the experiment to be carried out at three different temperatures and would allow for better fine-tuning. The current setup forces the temperature to at least 60 °C to prevent condensation of the medium weight compounds which sacrifices accuracy of the analysis of the light weight compounds. Introducing a third chamber would allow sampling at 20 °C, 66 °C, and 125 °C, optimizing the response of all compounds, therefore lowering the limit of detection thus making the method more sensitive. This was not included in the presented setup as the laboratory conditions only provided two sampling chambers.

An issue that will occur in the lower temperature sampling chambers, particularly 20 °C and

60 °C is, as mentioned before, condensation. It is not possible to prevent small tar particles

from depositing inside the pipes or the sampling chambers when the gas temperature drops.

This will impact the gas flow and the heat capacity of the system, thus impacting the required energy supply. If the system is to be used for repeated analysis, regular maintenance will be required. It is therefore worth noting the amount of condensables in the lower temperature sampling chambers during the practical experiments. A suggested cleaning method would be to flood the system with hot gas, evaporating the condensed tar and carrying it away. As presented in Table 1 the highest boiling point lies at 404 °C. However, whether the system can withstand such heat will be determined by the materials used in construction and is therefore difficult to predict.

A speculative improvement to the sampling process would be if one could determine a correlation between the concentration of the lighter and heavier compounds in the syngas.

Although unlikely, if such a connection could be determined, then one would only have to sample a small subset of the compounds. The setup would no longer require several sample chambers with different temperatures, but could be highly optimized to one temperature.

Finding said correlation would also be an extremely useful insight into the kinetics and reaction mechanisms of tar formation during biomass gasification.

Conclusion

This study aimed to create a set of instructions by which a proof of concept study to investigate

whether SPME can be sufficiently tuned to analyse samples with a tar concentration of less

than 0.1 mg/Nm ​

could be carried out. This set of instructions has been successfully designed

and now awaits execution. The authors are confident that the setup will work based on the

empirical and theoretical research carried. Slight adaptations depending on the setup

circumstances will be necessary.

References

Ahmadi Svensson, M. (2013). ​Sampling and analysis of tars by means of photo ionization detection and solid phase micro extraction (Doctoral dissertation, KTH Royal Institute of Technology).

Ahmadi, M., Elm Svensson, E., and Engvall, K. (2013). Application of solid-phase microextraction (SPME) as a tar sampling method. Energy and Fuels, 27(7), 3853-3860.

Brage, C., Yu, Q., Chen, G., and Sjöström, K. (1997). Use of amino phase adsorbent for biomass tar sampling and separation. ​Fuel​, ​76​(2), 137-142

Engvall, K., Liliedahl, T., and Dahlquist, E. (2013). Biomass and black liquor gasification. In Technologies for Converting Biomass to Useful Energy (pp. 217-258). CRC Press.

Fagbemi, L., Khezami, L., and Capart, R. (2001). Pyrolysis products from different biomasses:

application to the thermal cracking of tar. ​Applied Energy,​ 69(4), 293-306.

Górecki, T., Yu, X., and Pawliszyn, J. (1999). Theory of analyte extraction by selected porous polymer SPME fibres. ​Analyst​, ​124​(5), 643-649.

IEA (2020), Outlook for biogas and biomethane: Prospects for organic growth, IEA, Paris [online]

https://www.iea.org/reports/outlook-for-biogas-and-biomethane-prospects-for-organic-growt h ​ [Accessed] May 5th 2020

Lundgren, J., Waldheim L., Marklund M., Landälv I. & Gebart R. (2016). Biomassaförgasning och

pyrolys - Nycklar till fossilfrihet. Strategisk Innovationsagenda.

Martos, P. A., and Pawliszyn, J. (1997). Calibration of solid phase microextraction for air analyses based on physical chemical properties of the coating. ​Analytical Chemistry​, 69(2), 206-215.

Ouyang, G., and Jiang, R. (2017). Solid Phase Microextraction Recent Developments and Applications (1st ed. 2017.. ed.).

Prosen, H., and Zupančič-Kralj, L. (1999). Solid-phase microextraction. TrAC Trends in Analytical Chemistry, 18(4), 272-282.

PubChem (2020) ​Explore Chemistry​, US National Library of Medicine, [online]

https://pubchem.ncbi.nlm.nih.gov/ ​ [Accessed] February 10th 2020

Scarlat, N., Dallemand, J. F., and Fahl, F. (2018). Biogas: Developments and perspectives in Europe. ​Renewable Energy​, ​129​, 457-472.

SKC FAQ ​ ​(2020) ​What is the highest flowrate that can be used for sorbent tube sampling? ​SKC, [online] ​ https://www.skcinc.com/catalog/infotext.php?action=showEntry&data=303&id=4000 [Accessed] May 8th 2020

Woolcock, P. J., Koziel, J. A., Cai, L., Johnston, P. A., and Brown, R. C. (2013). Analysis of trace contaminants in hot gas streams using time-weighted average solid-phase microextraction:

Proof of concept. Journal of Chromatography A, 1281, 1-8.

Yung, M. M., Jablonski, W. S., and Magrini-Bair, K. A. (2009). Review of catalytic conditioning of biomass-derived syngas. ​Energy and Fuels​, ​23​(4), 1874-1887.

World Bioenergy Association (2019) WBA Global Bioenergy Statistics 2018: Summary Report,

WBA, Stockholm. [online]

https://worldbioenergy.org/uploads/181017%20WBA%20GBS%202018_Summary_hq.pdf [Accessed] May 5th 2020

Image Index

Chemspider. (2020). Royal Society of Chemistry. [online] ​ http://www.chemspider.com/

[Accessed May 4th 2020].

TRITA TRITA-CBH-GRU-2020:127