Detection and Characterization of Organic Gases during Coal Carbonization Using VUV-SPI-TOFMS

Abstract

A VUV-SPI-TOFMS system was developed for the continuous analysis of pyrolysis gases generated during coal carbonization and aromatic molecules were detected using high-resolution quantitative techniques. Although IR spectroscopy is suitable for detecting low-molecular-weight gases present in dry distillation gas, such as CH₄ and CO₂, it is not sensitive to aromatic hydrocarbon gases. On the other hand, the VUV-SPI-TOFMS technique is applicable for the quantification of low-concentration gases at the ppb concentration level, including aromatic hydrocarbons, generated during coal carbonization, which entails continuous heating from room temperature to 800°C. Benzene and toluene were predominantly detected at 540–590°C in the dry distillation gas of bituminous coal having relatively low oxygen concentration, whereas gases containing OH groups, such as phenol and cresol, are predominantly generated from sub-bituminous coal and lignite with high oxygen concentrations. Solid ¹³C NMR spectra obtained for each natural coal sample exhibited substantial proportions of oxygen bound to aromatic carbons (aromatic–O) in young coals. The temperature range at which cyclohexane was generated was found to be lower than that of aromatic molecules, indicating that coals releasing it may exhibit a structure that is more susceptible to thermal decomposition.

1. Introduction

Carbon dioxide (CO₂) emissions from the steel industry account for 14% of Japan’s total CO₂ emissions,¹⁾ making it a major source of CO₂ emissions in Japan. The chief reason for the high CO₂ emissions in the steelmaking process is the use of large amounts of coal for the reduction of iron ore. One approach toward achieving an energy-efficient process with low CO₂ emissions is to maximize the use of coke oven gas (COG), a byproduct generated in the process of producing coke from coal. In view of scientific and industrial applications, it is necessary to elucidate the coal carbonization process in detail to maximize the utilization of COG in coke facilities.

The general COG composition (vol%) is as follows: CO₂: 2–5 vol%, hydrocarbon gas: 2–4 vol%, O₂: 0.1–0.5 vol%, CO: 5–8 vol%, CH₄: 25–30 vol%, H₂: 50–60 vol%, N₂: 3–7 vol%.²⁾ The components can be quantified by collecting the gas in a sampling bag and analyzing it using a gas chromatography–photo ionization detector (GC-PID). Because COG is used as a fuel gas to operate various facilities in steel mills, fluctuations in its composition and production rate directly affect the control of calories burned. As the availability of high-quality coal resources has been decreasing in the last decade, the use of poor-quality coal as a coke raw material is increasing.³⁾ The utilization of inferior-quality coal results in changes in the properties of the chemical products derived from COG, such as tar, because the composition of COG directly affects tar synthesis. To standardize the quality of COG-derived chemical products, it is necessary to optimize the distillation temperature according to the desired properties of COG, which continuously changes in composition. Real-time monitoring of gases generated at varying dry distillation temperatures is necessary to optimize the operation method.

Based on these technical requirements, one of the authors has established online analytical techniques for the analysis of gas emissions during coal carbonization using FT-IR spectroscopy, entailing the rapid identification of major gas components, including relatively small molecules, such as CO, CO₂, and CH₄.^4,5,6) By continuous analysis of the heating process up to 1000°C, dry distillation gases such as H₂, CO, CH₂, C₂H₄, C₂H₆, and CO₂ were quantified in the coking and non-fine coking coal. It was found that, compared to coking coal, non-fine coking coal generates more CO₂ during dry distillation in the temperature range of 300–800°C, indicating that there may be differences in inorganic gas generation depending on the coking coal structure.

COG contains a small amount of aromatic molecules in addition to inorganic gases, such as CO, CO₂, and CH₄; however, these minor aromatic components are hazardous atmospheric pollutants.

Although the gases produced during the pyrolysis process of coal should take into account its secondary reactions, the detail of the composition in gases generated during coal dry distillation is derived from the chemical structure of coal. However, the relationship between the emission temperature ranges of aromatic molecules generated during the carbonization of coal species in the coking process is not fully understood and published reports on this topic are scarce. Lian et al.⁷⁾ and Odolphus et al.⁸⁾ sampled coal combustion gas using bags, stored the samples in vacuum bottles, and determined the composition of aromatic molecules using gas chromatography–mass spectrometry (GC-MS), resulting in that a mixture of molecules, such as benzene and toluene was detected. As the temperature range starting from 500°C is limited, differences in the occurrence of these molecules have not been evaluated when sampling bags were used for gas capture. Furthermore, molecules with high boiling points, such as naphthalene, adhere to the walls of the bag, preventing real-time analysis.

Herein, the continuous and fragment-free determination of several aromatic molecular species in gases produced during coal carbonization was achieved using vacuum ultraviolet single-photon ionization time-of-flight mass spectrometry (VUV-SPI-TOFMS). This technology is suitable for the effective quantification of various types of molecules, particularly aromatics, at the sub-ppb level. A VUV-SPI-TOFMS device capable of highly sensitive on-site measurement was developed by Tonokura et al.^9,10) and Zimmermann et al.¹¹⁾ In addition to the capability for continuous gas introduction into the device without pretreatment, this technology enables soft ionization using a vacuum ultraviolet (VUV) laser, allowing for the analysis of aromatic molecules in their pristine state, in contrast to previous studies where their cleavage was unavoidable.¹²⁾

The VUV-SPI-TOFMS system has been employed for the analysis of automobile exhaust gas,^13,14,15) fuel volatile components,^13,14,16) cigarette smoke,^{17,18,19,20,21)} coffee volatile components,^16,22) and nitro-aromatic compounds, which are environmentally hazardous substances.^23,24,25) Although the real-time monitoring of aromatic molecules was achieved, in all of the above-mentioned cases the analyzed gases are characterized by a uniform gas temperature, and gas introduction into the apparatus is straightforward. In contrast, pyrolysis gas introduction into the analyzer is difficult because the gas temperature is exceedingly high and the gas properties vary depending on the conditions. Continuous evaluation of dry gases at high temperatures is further hampered by the deposition of tar and consequent clogging of the inlet of the measuring device, as tar is gaseous at the time of generation. In our previous study, we analyzed coal carbonization gases at each carbonization temperature using a VUV-SPI-TOFMS device that allows the generated gas to be introduced into the device at a constant temperature.²⁶⁾ Dry distillation gases generated from four different coal species were analyzed, and the contents of aromatic molecules such as benzene, toluene, and phenol were found to differ in the content among the different types of coal.

The purpose of the present study is to improve the VUV-SPI-TOFMS system for the quantitative analysis of each generated species, especially aromatic compounds. The pyrolysis gases generated from bituminous and sub-bituminous coals with varying degrees of coal carbonization has been quantitatively analyzed employing the improved VUV-SPI-TOFMS. The pyrolysis gases are continuously introduced into the system to monitor fluctuations in the composition of aromatic molecules in the generated gases.

2. Experimental

2.1. Sample Preparation and VUV-SPI-TOFMS Equipment

Seven coal samples with varying degrees of coalification were prepared. The samples are referred to as Coal A to Coal G, corresponding to semi-anthracite for Coal A, bituminous for Coal B– Coal E, and sub-bituminous for Coal F and G. Chemical compositions and other characteristic values are listed in Table 1. The average coal particle size was ≤ 200 μm in diameter.

Table 1. Chemical composition, adhesive water, volatile matter (VM), ash, and fixed carbon (FC) contents of coal samples.

(mass%)
Coal	C	H	N	O*	S	Water	VM	Ash	FC
A	81	3.7	1.6	3.4	0.4	3.4	13.2	10.4	76.4
B	80.8	4.3	1.78	3.9	0.47	2.9	19.5	8.3	72.2
C	80.3	4.7	1.8	3.6	0.5	1.3	23.6	8.6	67.8
D	78.7	4.6	1.7	5.7	0.7	1.1	28.8	9.4	61.8
E	76.0	5.1	1.8	7.3	0.6	2.4	35.4	8.8	55.8
F	69.2	4.9	1.5	16.2	0.7	6.5	42.0	6.3	51.7
G	69.0	4.6	1.0	21.5	0.2	19.9	49.7	2.0	48.3

*  Oxygen content is obtained according to JIS M8813.

The Van Krevelen diagram showing the O/C and H/C atomic ratios of the seven coal samples is depicted in Fig. 1.²⁷⁾ Figure 2 shows schematics of the coal carbonization and VUV-SPI-TOFMS gas analysis system. Coal carbonization was performed using the following procedure. A coal specimen (20 mg) was placed on an alumina boat, maintained at a temperature of 50°C for 5 min in Ar atmosphere, and heated to 1000°C at a heating rate of 10°C/min in a tube oven. During carbonization, Ar gas flowed in the tube oven at 500 mL/min, mixing with the vaporized gas from the coal sample. In this study, a large amount of matrix gas, Ar gas, is introduced to dry distillation gas in the transfer tube. In this analysis system, since gases generated during the pyrolysis process are analyzed immediately, secondary gas decomposition reactions are suppressed in the measurement system.

Fig. 1.

Oxygen and hydrogen atomic ratios to carbon for coal samples plotted on a Van Krevelen diagram.²⁷⁾

Fig. 2.

Schematic of the VUV-SPI-TOFMS systems used for monitoring the carbonization process.

The pyrolysis gas from the tube furnace was introduced into the VUV-SPI-TOFMS system through a stainless pipe 1/8” in diameter. The stainless steel pipe was heated to 60°C, and a membrane filter was attached to the equipment end of the stainless-steel pipe as a tar trap. Although the total pyrolysis gas flow was 500 mL/min, a representative one-fifth (100 mL/min) was introduced into the VUV-SPI-TOFMS measurement system. The pyrolysis gas introduced into the VUV-SPI-TOFMS device was continuously passed through a pinhole 50 μm in diameter into the ionization chamber.

Because the hot pyrolysis gases generated in the tube furnace are rapidly cooled before reaching the pipe pinhole section, the gas inlet of the VUV-SPI-TOFMS system is heated to avoid the solidification of aromatic compounds in the pyrolysis gas. The tar trap is installed by the membrane filter as shown in Fig. 2. On the view point of the tar trap, the higher heating temperature is valuable; however, the maximum heating temperature for the gas inlet portion is limited to 60°C, because of the use of an O-ring or a similar device to maintain a vacuum at the pinhole position.

The VUV-SPI-TOFMS system comprises VUV light, an ionization chamber, and a reflectron-type time-of-flight mass spectrometer. Coherent VUV radiation at 10.5 eV (118 nm) was generated using Xe gas through frequency tripling of the third harmonic (355 nm) of an Nd³⁺:YAG laser (Big Sky Laser Series Ultra 100, Lumibird, USA, 30 mJ/pulse, 20 Hz). An einzel lens and deflectors were located just above the ion acceleration regions to control ion trajectories. The ions were passed into the field-free drift region through a microchannel plate (MCP F4655-10X, Hamamatsu Photonics K.K., Japan). The photon energy of the vacuum ultraviolet light used in the VUV-SPI-TOFMS system was 10.5 eV. Because most aromatic molecules present in the gas generated during coal drying have ionization energies (IEs) of below 10.5 eV, these molecular species are ionizable by single photons. Single-photon ionization allows for direct ionization from the ground state without the molecules having to pass through intermediate fragmented states. Thus, the developed method is characterized by single-photon ionization using the VUV laser and is applicable to the analysis of molecular species, mainly aromatic molecules, without fragmentation thereof.

2.2. Solid-state NMR and Thermogravimetry (TG)

Solid-state nuclear magnetic resonance (NMR) spectra were applied to determine the structural parameters of the coal. ¹H and ¹³C magic angle spinning (MAS) NMR spectra were acquired using an Agilent INOVA 500 spectrometer under a static magnetic field strength of 11.75 T. ¹H MAS spectra were collected applying a single-pulse excitation sequence, with a sample spinning frequency of 55 kHz to reduce spectral broadening caused by the ¹H-¹H homonuclear dipolar interaction. For the ¹³C MAS NMR spectra obtained at a MAS rate of 20 kHz, a spin-echo pulse sequence (90–180°) was utilized to avoid distortion of the spectral baseline. The accumulation numbers for the ¹H and ¹³C NMR spectra were 128 and 2000-4000, respectively. Pulse repetition rates of 5 s for ¹H and 100 s for ¹³C were sufficient to fully relax the z-magnetization. The ¹³C chemical shift was referenced to hexamethylbenzene at 17.3 ppm, which corresponds to the tetramethylsilane shift at 0 ppm.

To determine the pyrolysis characteristics of the seven coals used, simultaneous differential thermogravimetry measurements were performed using a Seiko Instruments Co., Ltd., TG / DTA6300 instrument. A 10 mg coal sample was heated from 20 to 1000°C at a heating rate of 10°C/min under an Ar atmosphere. The flow rate of the Ar gas was 200 mL/min, and the heating pattern was the same as in the VUV-SPI-TOFMS measurement. The change in the differential thermal weight was measured employing TG analysis.

3. Results

3.1. Chemical Structure of Natural Coals and Weight Change during Carbonization

Figure 3 shows the solid-state ¹³C NMR spectra of the seven coals at room temperature. A variety of signals was observed in the range of 0–240 ppm. As observed in the Van Krevelen diagram in Fig. 1, the solid-state ¹³C NMR spectral peaks indicate significant variations among the coal types.

Fig. 3.

¹³C MAS spectra for (a) Coal A, (b) Coal B, (c) Coal C, (d) Coal D, (e) Coal E, (F) Coal F and (g) Coal G. (Online version in color.)

Signals appearing in the region of 0–50 ppm were assigned to aliphatic carbons, whereas those in the region of 90–165 ppm were attributed to aromatic carbons. Signals indicative of oxygen-containing functional groups were observed at various chemical shifts: aliphatic carbons connected to oxygen at 50–90 ppm, aromatic carbons connected to oxygen at 150–165 ppm, carboxyl acids and/or esters at 165–180 ppm, and carboxyls at 185–240 ppm. The peak assignments are listed in Table 2.^28,29)

Table 2. Assignment of coal carbon structures based on ¹³C MAS spectra.

13C chemical shift (ppm)	Carbon functional group	Symbol
240-185	carbonyl	C=O
185-165	carboxyl	C(=O)O
165-150	aromatic carbon connected with oxygen	Aro-O
150-135	aromatic carbon connected with carbon	Aro-C
135-90	aromatic carbon connected with hygrogen & Internal/bridgehead carbon	Aro-H & bridgehead
90-50	aliphatic carbon connected with oxygen	Ali-O
50-22	methylene & methine carbon	CH2 & CH
22-0	methyl carbon	CH3

The relative peak intensity of aliphatic carbon peaks (0–90 ppm) decreased as the degree of coalification increased from Coals G to A. For coal samples with a lower degree of coalification, for example, Coals F and G, a wide variety of signals were observed, especially in the region of 90–240 ppm, which were assigned to oxygen-bonded carbons. To acquire further quantitative information, the following structural parameters were obtained from the ¹³C NMR spectra:   

$$f\mathrm{a}={\mathrm{S}}_{90-165}/{\text{S}}_{0-240}$$(1)

  

$$\begin{array}{l}\text{Percentage of oxygen-bonded carbons}=\\ {\text{( S}}_{\text{50}-\text{90}}+{\text{S}}_{\text{150}-\text{165}}+{\text{S}}_{\text{165}-\text{185}}+{\text{S}}_{\text{185}-\text{240}}{\text{)/S}}_{\text{0}-\text{240}}\text{.}\end{array}$$(2)

Here,

fa: aromatic carbon ratio,

S_90–165: peak integrals in the 90–165 ppm region,

S_0–240: peak integrals in the 0–240 ppm region,

S_50–90: peak integrals in the 50–90 ppm region,

S_150–165: peak integrals in the 150–165 ppm region,

S_165–185,: peak integrals in the 165–185 ppm region,

S_185–240,: peak integrals in the 185–240 ppm region.

The cluster size and number of carbon atoms in condensed aromatic rings provide useful information on the condensation of aromatic rings in coal. The cluster size was calculated from the ¹H and ¹³C NMR data and the molar concentration of hydrogen atoms.³⁰⁾ The structural parameters are presented in Table 3.

Table 3. Aromatic carbon ratio (fa), percentage of oxygen-bonded carbons (carbonyl, carboxyl, aromatic and aliphatic carbons bonded to oxygen) and cluster sizes estimated from the ¹³C and ¹H MAS spectra and chemical composition of coals.

Coal	fa	C=O, -C(=O)O, Aro-O, Ali-O (%)	Cluster size
A	0.87	3.6	20
B	0.83	3.7	18.7
C	0.80	3.6	17.1
D	0.77	7.8	14
E	0.68	7.3	11.3
F	0.64	12	11.9
G	0.57	22	8

As the degree of coalification increased, the fa value, that is, the percentage of aromatic carbons to total carbons, and the cluster size increased, but the percentage of oxygen-bonded carbons decreased.

Figure 4 shows the TG profiles of the seven coal samples. Weight loss occurred largely in the 300–800°C temperature range for all samples. However, the weight loss of Coal G started at approximately 100°C, reaching 10% at 300°C, followed by significant weight loss in the 300–800°C range. For Coal A, the weight loss at 1000°C was only 13%, while a substantial two-stepped weight loss was observed for Coal G at this temperature. Comparing the weight loss for all samples in the 300–800°C range, it can be concluded that the percentage of weight loss increased with a decreasing degree of coalification.

Fig. 4.

Weight loss curves for the seven coals at atmospheric pressure.

3.2. Identification of Molecular Species in Pyrolysis Gas

Because it is important to separate the simultaneously generated tar from the gases to be analyzed, a tar trap was installed in the section of pipe leading up to the VUV-SPI-TOFMS apparatus, whereby the pipe temperature was decreased to approximately 60°C. The tar trap system was highly effective in eliminating blockages caused by tar and ensuring the stable flow of a certain amount of pyrolysis gas into the equipment. Figure 5 shows representative mass spectra for pyrolysis gases obtained from Coal A, Coal D, and Coal G in the temperature range of 450–600°C. The 450–600°C temperature range is the softening and melting temperature range of coal, and it is known that coal softening commences at 450°C and resolidification terminates at up to 600°C.³¹⁾ Because the gas generation temperature of aromatic molecules coincides with the coal softening and melting temperature range, it is considered that the gas generation temperature and the type and amount of gas generated affect flowability. It has been reported that the flow range of many coking coals is in the 450–600°C range,³²⁾ and the mass spectra shown in Fig. 5 correspond to gas produced in such a temperature range.

Fig. 5.

VUV-SPI-TOFMS Spectra for (a) Coal A, (b) Coal D and (c) Coal G in the temperature range of 450–600°C.

The main peaks observed in the mass spectra of Coal A, Coal D, and Coal G were at mass numbers (m/z) of 34, 78, 84, 92, 106, 120, 128 and 142, and in addition to these peaks, only Coal G gave rise to mass numbers (m/z) of 94 and 108. These peaks were assigned based on the results obtained from previous studies³³⁾ to compounds that are known to be generated in pyrolysis gases from coal dry distillation or are known molecules inferred to be generated during pyrolysis based on the coal structure.^34,35,36)

The peak at m/z 34 was assigned to hydrogen sulfide, m/z 78 to benzene, m/z 84 to cyclohexane, m/z 92 to toluene, m/z 94 to phenol, m/z 106 to xylene, m/z 108 to cresol, m/z 128 to naphthalene, and m/z 142 to methylnaphthalene.^33,34,35,36) Benzoquinone at m/z 108 may be present in the pyrolysis gas, however the ionization probability of cresol is more than one order of magnitude higher than that of benzoquinone. Therefore, the peak at m/z 108 was attributed to cresol because cresol is more easily detected than benzoquinone. In addition to the assigned peaks, several other peaks were observed, and the number of peaks for Coal G was greater than that for Coal A. The m/z mass values in the range of 20–40 were identified as arising from Ar gas and the divalent Ar atom. Because the energy of the VUV light used does not exceed the ionization potential of Ar atoms in the one-photon mode, Ar atom excitation was considered to the result from ionization by two photons of VUV light or by partial UV light of 355 nm, which is used to generate VUV laser light.

The mass spectra obtained for Coal A and D were comparable in terms of major peaks; therefore, the mass spectra for Coal B and C are omitted. The mass spectra obtained for Coal E to Coal G were of the same group, in which the same major peaks were observed. Thus, the mass spectrum of Coal G is shown as a representative mass spectrum of Coal E, Coal F, and Coal G.

Figure 6 shows the temperature dependence of the concentration of molecules generated from the carbonization of these coals. The horizontal axis in Fig. 6 represents the pyrolysis temperature, and the vertical axis represents the concentration of each molecule. The signals obtained for each interval of 10°C were integrated and plotted, and the signal intensities were converted to quantitative values using calibration curves measured separately for each compound, namely, benzene, cyclohexane, toluene, phenol, xylene, naphthalene, and methylnaphthalene.

Fig. 6.

Temperature dependence of generated gas composition during the carbonization process for (a) Coal A, (b) Coal B, (c) Coal C, (d) Coal D, (e) Coal E, (f) Coal F and (g) Coal G.

The maximum benzene concentration occurred in the 580–590°C temperature range in all cases. The highest concentration of toluene was observed at a lower temperature range than that of benzene, and the temperature range for xylene concentration maxima was lower than that of toluene. Since secondary decomposition is unlikely to occur in this experimental system, this difference in the maximum temperatures generated for each molecule is due to primary decomposition rather than secondary decomposition. Compared to the C-C bond between the benzene ring and the methyl group or side chain, the C-C bond between the methyl group of toluene or xylene and the side chain may be more easily thermally decomposed because the bond energy is weaker. The difference in bond energy may have led to the difference in the temperature of maximum concentration. Cyclohexane reached the highest concentration between 460 and 480°C, which is approximately 100°C lower than the temperature range of benzene. The maximum naphthalene concentration occurred at the same temperature range as that of benzene for Coal A through Coal D, whereas its content was minimal in Coal E through Coal G. The highest concentrations of the seven molecules for which the quantitative values were calculated were toluene for Coal A, xylene for Coal B to Coal D, cresol for Coal E, and phenol for Coal F and G.

Figure 7 shows the average concentrations of each molecular species in the seven coal samples, obtained by normalizing the concentrations by unit time over the entire temperature range from 300 to 800°C based on the data shown in Fig. 6. Coals with higher aromatic carbonization rates (fa) tended to have higher benzene concentrations. Cyclohexane content tends to increase with decreasing the degree of coalification, but this trend is not observed for Coal G. Benzene, toluene, and xylene were the major compounds generated by Coal A–Coal D, while phenol and cresol were the major components of Coal E, Coal G. The concentrations of toluene, xylene, naphthalene, and methylnaphthalene, as well as benzene, show an increasing trend in coals with a higher degree of coal carbonization, but the toluene concentration was lower in Coal A than in Coals B, xylene and naphthalene were lower in Coal A than in Coal B, and methylnaphthalene in was lower in Coal A than in Coal B–Coal D. Coal A tended to have lower concentrations of xylene and naphthalene than Coal B and Coal C. The concentrations of phenol and cresol increased as the degree of coalification decreased; however, the concentration of phenol was lower in Coal G than in Coal F, and the concentration of cresol was lower in Coal G than in Coal E and Coal F.

Fig. 7.

Compound concentrations in the gas generated during the coal carbonization process of Coal A–Coal G in the temperature range of 300–800°C.

4. Discussion

4.1. Gas Generation Behavior Based on the Present Equipment Setup

The VUV-SPI-TOFMS system has been developed for the analysis of pyrolysis gas components during the dry distillation of coal. This system features fragment-free detection of molecules ionized by a single photon, whereby molecular ionization occurs without cleavage, thereby avoiding intermediate molecular states. The developed technique is highly sensitive for the quantification of aromatic species in pyrolysis gas.

One of the subjects to improve the system is to decrease the tar component into the VUV-SPI-TOFMS system.

Since aromatic compounds with one or two benzene rings remain in the gaseous state when heated to 60°C, they can be analyzed using the VUV-SPI-TOFMS system. This heating step was crucial for the removal of tar components and did not negatively affect the study. As indicated in Fig. 5, most of the tar components were removed by heating to 60°C and not detected in the VUV-SPI-TOFMS spectra. The concentration of OH-containing molecules such as phenol and cresol shows tailing at temperatures above 700°C, as shown in Fig. 6. Coal E to Coal G were dry distilled at 500–600°C, and large amounts of phenol and cresol were generated and some remained in the SUS piping and tar traps as liquids. As the dry distillation temperature increased further to 600–800°C, significantly lower amounts of phenol and cresol were generated, but the vapor gases of liquid components remaining in the SUS pipes and tar traps were introduced into the VUV-SPI-TOFMS system, resulting in tailing.

Although polycyclic aromatics, such as anthracene, are present in carbonized coal gas, their presence is not observed in the results of this study. This is because the gas was analyzed after the temperature is rapidly cooled to room temperature and reheated to 60°C in the SUS piping, and any generated aromatics with more than three benzene rings is still remain as the liquid such as tar in the tar trap and not introduced into the inlet of the VUV-SPI-TOFMS equipment. Therefore, no signal was detected for the mass number of aromatics with three or more rings, and this is a future subject in the present system.

4.2. Characteristics of Pyrolysis Gas from the Seven Coals

Considering the changes in the thermogravimetric curves shown in Fig. 4, the weight loss percentages over the entire temperature range of 25–1000°C are comparable to the volatile matter (VM) mass % values indicated in Table 1. Because the VM is determined by the weight loss percentage after annealing at 900°C for 7 min, it is reasonable that the weight loss results and VM values are comparable. There is a significant weight loss (10%) at 100°C only in the case of Coal G, which has a relatively high water content (19.9%, Table 1); thus, this weight loss is indicative of moisture adhered to the coal surface. None of the other samples exhibited weight loss under 100°C; therefore the water contents indicated in Table 1 correspond to the inner water component of these coals.

Excluding the weight loss occurring below 100°C in Coal G, which corresponds to the surface water component, the weight loss changes observed for Coal E to Coal G are similar, being ~20% in the temperature range of 400–500°C. The weight loss changes in Coal A to Coal D were lower than those of Coal E to Coal G in the same temperature range. Thus, based on the obtained weight loss changes, the specimens were divided into two groups: Coals A–Coal D and Coal E–Coal G.

The major carbonization compounds of semi-anthracite (Coal A) and bituminous (caking) coals (Coal B–Coal D), denoted as “coals with a high degree of coalification” are found to be benzene, toluene, and xylene by the ¹³C NMR measurement results. On the other hand, OH-containing aromatic compounds, such as phenol and cresol, are mainly detected during the carbonization of bituminous coal with low coalification (Coal E) and subbituminous coal (Coal F and Coal G), denoted as “coals with a low degree of coalification”.

Considering the chemical structures of the natural coals in these two groups, it is evident that the concentrations of aliphatic species, such as CH₂ and CH, differ between the groups, as shown in Fig. 3. Coals with a low degree of coalification have higher amounts of aliphatic carbons as well as oxygen functional groups. These oxygen functional groups readily form hydrogen bonds with water; consequently, coals with a low degree of coalification have higher moisture contents than those with a high degree of coalification.³⁷⁾ The mass spectrum of Coal G shown in Fig. 5(c) contains peaks corresponding to phenol and cresol, which contain OH groups.

Because the mass spectrum of Coal G is a representative mass spectrum for Coal E–Coal G with low degrees of coalification, it is clear that phenol and cresol are the main aromatic species present in these specimens. On the other hand, benzene and toluene are the major aromatic species in the samples of Coal A–Coal D, as shown in Fig. 5.

Coal E is categorized as bituminous coal, as well as Coals B–Coal D; however, the Van Krevelen diagram in Fig. 1 shows that Coal E has higher H/C and O/C ratios than Coal B–Coal D. The results obtained from the VUV-SPI-TOFMS system shown in Fig. 6 indicate that Coal E has higher concentrations of phenol and cresol, similar to Coal F and G, which are sub-bituminous coals. Consequently, Coal E, a bituminous coal, is considered to have properties similar to those of Coal F and G. Bituminous coal is the most common type of coal used in coke facilities. If the use of the same bituminous coal with properties similar to Coal E increases, the concentrations of phenol and cresol in the COG will increase, and the quality of the chemical products will change. To control the chemical product quality, the results obtained using the VUV-SPI-TOFMS system are possible to be used as a reference.

4.3. Relationship between the Natural Coal Structure and Organic Gas Species Generated in the Carbonization Process

Comparing the ¹³C NMR spectra for the natural coals before carbonization with the molecular components of the pyrolysis gases examined by the VUV-SPI-TOFMS system, the relationship between the coal structure and the aromatic molecules generated in the pyrolysis gases are discussed.

As shown in Figs. 5, 6, and 7, the major components of coals with a high degree of coalification are benzene, toluene, and xylene. These results are also supported by the ¹³C NMR data in Fig. 3; the signal intensity attributed to Aro-H/bridgehead and Aro-C, being decomposed to benzene and xylene, is higher in coals with a high degree of carbonization. For coals with a lower degree of carbonization and a higher oxygen concentration, the concentrations of oxygen-bonded carbons, such as Aro-O, Ali-O, and carboxyl groups observed in the ¹³C NMR spectra are higher, as indicated in Fig. 3, and the gases generated in the dry distillation of coal and analyzed using the VUV-SPI-TOFMS system contain more oxygen-comprising aromatic molecules, such as phenol and cresol, as shown in Figs. 5, 6, and 7.

Although Coal G has a higher oxygen content than Coal F, it has a lower content of aromatic species with OH groups, such as phenol and cresol, as indicated in Fig. 7. This is presumably because Coal G being a lower-rank sub-bituminous coal contains more carboxyl groups, resulting in higher CO and CO₂ emissions, which are detectable by FT-IR rather than VUV-SPI-TOFMS. Combining VUV-SPI-TOFMS and FT-IR instruments for analyzing aromatic molecules, and CO₂, CO, CH₄, and other aliphatic molecules, respectively, will enable a more accurate understanding of coal carbonization reactions. It is also known that lower carbonized coals such as Coal G contain more metal species³⁷⁾ with higher catalytic activity, and it is possible that phenol and cresol may be degraded by the catalytic reaction, however the percentage of catalytic effect cannot be clearly determined from the present results. Notably, the developed VUV-SPI-TOFMS system is portable, similar to the FT-IR apparatus, which is advantageous because this enables on-site gas monitoring in coke plants and other chemical plants. Based on these characteristics, this system is expected to be widely utilized in various industries that require gas management.

The Aro-O species in natural coal are decomposed into phenol and cresol containing Aro-O bonds. However, the NMR results, in which the Aro-O signal of Coal G is greater than that of Coal F, are inconsistent with the VUV-SPI-TOFMS data, which indicate higher phenol and cresol contents in Coal F than in Coal G. This contradiction may be caused by the presence of other aromatic molecules containing OH groups that are not assigned by the VUV-SPI-TOFMS measurement.

Cyclohexane which is a representative molecule of the naphthene ring is detected in all the sampled coals with varying degrees of coalification, as shown in Fig. 7. This is potentially indicative of the naphthene ring being a partial component of natural coal. The results shown in Fig. 6 indicate that cyclohexane reaches its highest concentration between 460°C and 480°C, while benzene concentration is the highest value at 580°C. This indicates that the naphthene ring is more readily pyrolyzed in coal than aromatic rings. It has been reported that in the structure of coals with a lower degree of carbonization, such as sub-bituminous and lignite coals, aliphatic hydrocarbons and naphthene rings are more abundant in the side chains than in bituminous coal.^34,35,36) Based on the results shown in Fig. 7, the concentration of cyclohexane tends to increase as the carbonization degree of coals decreases. These results indicate a correlation between the amount of cyclohexane generated and naphthene ring side chain content in the coal structure.

5. Conclusion

The following conclusions are reached by combining the measurement results acquired by the improved VUV-SPI-TOFMS instrument capable of continuous pyrolysis gas analysis with the results obtained from solid-state NMR analysis of the coal chemical structure.

(1) The tar trap was installed by the membrane filter at a position where the piping was heated to 60°C, allowing the continuous introduction of dry distillation gases without tars into the VUV-SPI-TOFMS system in a stable manner. The laser used in the VUV-SPI-TOFMS system is suitable for the efficient quantitative monitoring of the gases produced during coal pyrolysis over a wide temperature range, from room temperature to 800°C.

(2) Based on the weight loss results, the seven coals are divided into two groups by the degree of coalification. Using continuous analysis of pyrolysis gas produced from them, it is made clear that semianthracite and bituminous coals with the high degree of coalification are characterized by benzene, toluene, and xylene as the major aromatic products, whereas low-rank bituminous and sub-bituminous coals are characterized by a higher concentration of molecules with OH groups, such as phenol and cresol. Semianthracite and bituminous coals predominantly contain Aro-H and bridgehead carbons, as identified in the ¹³C NMR spectra. In contrast, the ¹³C NMR spectra of low-rank bituminous and sub-bituminous coals contain a higher percentage of Aro-O peaks.

(3) Cyclohexane which is a cyclic aliphatic compound is detected in all the coal species with varying degrees of coalification. However, its generation temperature range is lower than those of aromatic hydrocarbon gases, which potentially indicates the partial structure of coal.

(4) The lowest-carbonized sub-bituminous coals contain the highest proportion of Aro-O compared to other sub-bituminous coals, as implied in the ¹³C NMR spectra, but lower concentrations of phenol and cresol. It is considered that this is due to higher emissions of species such as CO and CO₂, and the possible presence of aromatics containing OH groups other than phenol and cresol.

References

• 1)  National Institute for Environmental Studies: Greenhouse Gas Emissions Data for Japan (2021), https://www.nies.go.jp/whatsnew/20211210/20211210.html, (accessed 2022-09-06) (in Japanese).
• 2)  Handbook of Chemistry: Pure Chemistry, 6th ed., ed. by The Chemical Society of Japan: Maruzen Publishing Co., Ltd., Tokyo, (2021), 56 (in Japanese).
• 3)   K.  Miura and  K.  Kato: Tetsu-to-Hagané, 92 (2006), 703 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.92.12_703
• 4)   M.  Nishifuji,  K.  Saito,  Y.  Fujioka,  K.  Kanehashi,  Y.  Ishiharaguchi and  M.  Ueki: J. Jpn. Inst. Energy, 83 (2004), 895 (in Japanese). https://doi.org/10.3775/jie.83.895
• 5)   M.  Nishifuji,  Y.  Fujioka and  K.  Saito: Tetsu-to-Hagané, 89 (2003), 994 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.89.9_994
• 6)   M.  Nishifuji,  K.  Kanehashi,  Y.  Fujioka,  K.  Saito,  M.  Ueki and  Y.  Ishiharaguchi: Tetsu-to-Hagané, 91 (2005), 299 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.91.2_299
• 7)   H.  Lian,  T.  Katou and  Y.  Hanai: Bull. Inst. Environ. Sci. Technol., Yokohama Natl. Univ., 8 (1982), 37 (in Japanese). https://ynu.repo.nii.ac.jp/index.php?action=pages_view_main&active_action=repository_action_common_download&item_id=5207&item_no=1&attribute_id=20&file_no=1&page_id=59&block_id=74, (accessed 2022-09-06).
• 8)   O. S. L.  Bruinsma,  R. S.  Geertsma,  P.  Bank and  J. A.  Moulijn: Fuel, 67 (1988), 327. https://doi.org/10.1016/0016-2361(88)90315-8
• 9)   N.  Kanno and  K.  Tonokura: Appl. Spectrosc., 61 (2007), 896. https://opg.optica.org/as/abstract.cfm?URI=as-61-8-896
• 10)   K.  Tonokura,  N.  Kanno,  Y.  Yamamoto and  H.  Yamada: Int. J. Mass Spectrom., 290 (2010), 9. https://doi.org/10.1016/j.ijms.2009.11.004
• 11)   T.  Adam and  R.  Zimmermann: Anal. Bioanal. Chem., 389 (2007), 1941. https://doi.org/10.1007/s00216-007-1571-x
• 12)   K.  Tonokura: Bunseki, 12 (2011), 720 (in Japanese).
• 13)   F.  Mühlberger,  J.  Wieser,  A.  Ulrich and  R.  Zimmermann: Anal. Chem., 74 (2002), 3790. https://doi.org/10.1021/ac0200825
• 14)   F.  Mühlberger,  J.  Wieser,  A.  Morozov,  A.  Ulrich and  R.  Zimmermann: Anal. Chem., 77 (2005), 2218. https://doi.org/10.1021/ac048319f
• 15)   D. J.  Butcher,  D. E.  Goeringer and  G. B.  Hurst: Anal. Chem., 71 (1999), 489. https://doi.org/10.1021/ac9807971
• 16)   F.  Mühlberger,  R.  Zimmermann and  A.  Kettrup: Anal. Chem., 73 (2001), 3590. https://doi.org/10.1021/ac010023b
• 17)   F.  Mühlberger,  K.  Hafner,  S.  Kaesdorf,  T.  Ferge and  R.  Zimmermann: Anal. Chem., 76 (2004), 6753. https://doi.org/10.1021/ac049535r
• 18)   T.  Adam,  T.  Streibel,  S.  Mitschke,  F.  Mühlberger,  R. R.  Baker and  R.  Zimmermann: J. Anal. Appl. Pyrolysis, 74 (2005), 454. https://doi.org/10.1016/j.jaap.2004.11.021
• 19)   F.  Mühlberger,  T.  Streibel,  J.  Wieser,  A.  Ulrich and  R.  Zimmermann: Anal. Chem., 77 (2005), 7408. https://doi.org/10.1021/ac051194+
• 20)   T.  Adam,  T.  Ferge,  S.  Mitschke,  T.  Streibel,  R. R.  Baker and  R.  Zimmermann: Anal. Bioanal. Chem., 381 (2005), 487.
• 21)   F.  Mühlberger,  M.  Saraji-Bozorgzad,  M.  Gonin,  K.  Fuhrer and  R.  Zimmermann: Anal. Chem., 79 (2007), 8118. https://doi.org/10.1021/ac071217f
• 22)   R.  Dorfner,  T.  Ferge,  C.  Yeretzian,  A.  Kettrup and  R.  Zimmermann: Anal. Chem., 76 (2004), 1386. https://doi.org/10.1021/ac034758n
• 23)   C.  Mullen,  A.  Irwin,  B. V.  Pond,  D. L.  Huestis,  M. J.  Coggiola and  H.  Oser: Anal. Chem., 78 (2006), 3807. https://doi.org/10.1021/ac060190h
• 24)   B. V.  Pond,  C.  Mullen,  I.  Suarez,  J.  Kessler,  K.  Briggs,  S. E.  Young,  M. J.  Coggiola,  D. R.  Crosley and  H.  Oser: Apply Phys. B, 86 (2007), 735. https://doi.org/10.1007/s00340-006-2465-x
• 25)   N.  Tsuji,  Y.  Matsuzaki and  S.  Hayashi: Bunseki Kagaku, 61 (2012), 359 (in Japanese). https://doi.org/10.2116/bunsekikagaku.61.359
• 26)   N.  Tsuji,  M.  Nishifuji and  S.  Hayashi: Hyperfine Interact., 216 (2012), 127. https://doi.org/10.1007/s10751-013-0822-9
• 27)   D. W.  Van Krevelen: Fuel, 29 (1950), 269.
• 28)   T. H.  Fletcher,  S.  Bai,  R. J.  Pugmire,  M. S.  Solum,  S.  Wood and  D. M.  Grant: Energy Fuels, 7 (1993), 734. https://doi.org/10.1021/ef00042a006
• 29)   T.  Yoshida,  Y.  Nakata,  R.  Yoshida,  S.  Ueda,  N.  Kanda and  Y.  Maekawa: Fuel, 61 (1982), 824. https://doi.org/10.1016/0016-2361(82)90309-X
• 30)   M. S.  Solum,  R. J.  Pugmire and  D. M.  Grant: Energy Fuels, 3 (1989), 187. https://doi.org/10.1021/ef00014a012
• 31)   N.  Tsubouchi,  R.  Naganuma,  Y.  Mochizuki,  H.  Hayashizaki and  T.  Shishido: Tetsu-to-Hagané, 107 (2021), 24 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2020-053
• 32)   N.  Tsubouchi,  Y.  Mochizuki,  R.  Naganuma,  K.  Kamiya,  M.  Nishio,  Y.  Ono and  K.  Uebo: Energy Fuel, 30 (2016), 2095. https://doi.org/10.1021/acs.energyfuels.5b02914
• 33)   N.  Mallya and  R. A.  Zingaro: The Chemistry of Low-Rank Coals, ACS Symp. Series, Vol. 264, American Chemical Society, Washington, D.C., (1984), 133.
• 34)   P. J. J.  Tromp and  J.  Moulijn: New Trends in Coal Science, NATO ASI Series C, Mathematical and Physical Sciences, Vol. 244, Kluwer Academic Publishers, Boston, (1988), 305.
• 35)  Aromatics and Tar Industry Handbook, 3rd ed., The Japan Aromatic Industry Association, Tokyo, (2000), 56 (in Japanese).
• 36)   J. H.  Shinn: Prepr. Pap. Am. Chem. Soc., Div. Fuel Chem., 41 (1996), 510.
• 37)   H.  Miyakoshi,  T.  Isobe and  K.  Otsuka: J. Min. Metall. Inst. Jpn., 100 (1984), 643 (in Japanese). https://doi.org/10.2473/shigentosozai1953.100.1158_643