FTIR Analysis of the Functional Group Composition of Coal Tar Residue Extracts and Extractive Residues

Abstract

The extracts and extractive residues were obtained by single-stage extraction with acetone, CS₂ and acetone/carbon disulphide solvent (IACDSM) under mild conditions, and they were then characterised by infrared and segmented peak fitting analysis. The results showed that acetone and CS₂ solvents could dissolve more aromatic structures containing fatty side chains, while the mixed solvents could extract more phenolic substances. The acetone extracts were mainly asymmetric -CH₂ stretching vibrations, while the remaining two extracts were dominated by aliphatic -CH. A comparison of the IR results of the tar residue and the three solvent extracts showed that the main structure of the tar residue was not altered throughout the process, but the aromatic structure was altered more by the extraction. Compared to the original sample, the proportion of tetrasubstituted or trisubstituted benzene rings in the three solvent extracts increased, while the proportion of pentasubstituted benzene rings produced a significant decrease. Comparison of the IR structural parameters showed that the aromatic ring condensation was higher in the as-received CS₂ extracts and acetone extracts, with the acetone extracts having longer straight aliphatic chains and fewer branched chains. The analysis of the composition of the extraction results provides a reference basis for the separation and purification technology of coal tar residue, enabling the further development of China’s coal chemical industry.

1. Introduction

The coal coking production process produces a large amount of residues, among which coal tar residue is classified as hazardous waste due to the large amount of carcinogenic polycyclic aromatic hydrocarbons (PAHs) [1,2], which make it a wasted available resource [3]. Although there are many comprehensive utilization methods for tar residue, the lack of in-depth research on the nature of tar residue and the high cost of treatment have prevented the effective treatment and high value-added utilization of tar residue [4].

Coal tar residue still contains more organic aromatic compounds and fixed carbon. The solvent extraction method uses the “similar solubility” principle of the organic components in coal tar residue [5,6] and the extraction solvent to achieve the separation and recovery of organic components in coal tar residue [7,8] and improve its utilization rate.

The extracts are generally determined by gas chromatography–mass spectrometry (GC/MS) [9], Fourier transform infrared spectrometry (FTIR) [10], and pyrolysis gas chromatography-mass spectrometry (PY-GC/MS) [11] techniques to determine their compositional structures. Liang et al. [12] investigated the effect of deashing on the structure of Yimin lignite using FTIR and concluded that the deashing process had a greater effect on the aromatic structure and oxygen-containing functional groups, mainly due to the three effects of substitution, hydrolysis, and ion exchange that occurred during the deashing process. Liang et al. [13] conducted a qualitative and semi-quantitative study of the functional groups of long-flame and anthracite coals based on the FTIR and split-peak fitting techniques. The results showed that the organic matter content of low-degradation coals was less than that of high-degradation coals, and the organic matter maturity of high-degradation coals was higher than that of low-degradation coals. As the degree of coal metamorphism increases, the degree of coal aromatization becomes higher and the degree of coal aromatic ring condensation becomes stronger and more prone to condensation. Wang et al. [14] used acid-washed and Hefeng sub-bituminous coals as research objects, and analysed the molecular structures of the extracted products with the help of FTIR and segmented peak fitting methods. The results showed that the hydroxyl hydrogen bonding structure of all five extraction levels was dominated by self-conjugated hydroxyl hydrogen bonds; CS₂ dissolved a larger proportion of aromatic structures containing fatty side chains in the five extraction levels. The functional groups contained in the extracts were the same at all levels. Zhao et al. [15] analyzed the structural characteristics of different chemical components of Yimin lignite using Fourier transform infrared spectroscopy. The results show that humic acid and asphaltene have fewer aromatic rings and are mostly in the form of monocyclic rings, while residual coal has more aromatic rings and is mostly in the form of naphthalene rings; and that humic acid and asphaltene contain a large number of aliphatic hydrocarbons, which are shorter in length and more branched.

In this paper, single-stage extraction experiments with various solvents were carried out on tar residues using ultrasonic-assisted extraction at room temperature. The extracts and extractive residues were analysed by FTIR detection and segmented peak fitting techniques in order to obtain the structural characteristics of the soluble organic components in the tar residues.

2. Materials and Methods

2.1. Sample Preparation and Reagents

Coal tar residue came from the Yulin Lanchar plant. The samples were dried under vacuum conditions at 40 °C for 2 h, then taken out and placed in a desiccator at room temperature and sealed in bags. The elemental and industrial analyses of the samples are shown in

Table 1. Proximate and ultimate analyses of the coal tar residue samples.

Samples	Industrial Analysis (Mass Fraction)/%				Elemental Analysis (Mass Fraction)/%
	Mad	Ad	Vdaf	FCad	Cdaf	Hdaf	Ndaf
Coal tar slag	5.33	13.43	17.58	67.55	77.86	1.911	2.125

Note: M_ad is air dry basis moisture, A_d is dry basis ash, V_daf is dry ash-free basis volatile fraction and FC_ad air dry basis fixed carbon.

. The acetone and CS₂ solvents used in the experiments were commercially available analytically pure reagents, which were refined by rotary evaporator before use.

2.2. Extraction Experiments

Weigh 5 g of coal tar residue into a filter paper cartridge in a beaker, add 40 mL of each of the three organic solvents, acetone, CS₂ and CS₂/acetone (1:1, V:V), place in an ultrasonic cleaner and shake for 20 min; ultrasonic power: 100%, extraction temperature: 20 °C. After extraction, remove and filter the filter paper cartridges, concentrate the extract using a rotary evaporator, transfer to a sample bottle to dry, and weigh naturally.

2.3. FTIR Determination and Split-Peak Fitting of Samples

The structure and functional groups of coal tar residue extracts and extractive residues were analysed using a Nicolet iN10 & iZ10 Fourier infrared spectrometer from Thermo Fisher Scientific, Waltham, MA, USA. The IR measurement conditions were: 100 um, 2 min scan, resolution 0.4 cm⁻¹, scan range 400–4000 cm⁻¹, and test specimens were produced using the KBr compression method.

Due to the complexity of the structure of coal tar residue, its IR spectra cannot directly give specific information on the structure of the functional groups corresponding to the overlapping peaks and shoulder peaks, so this paper uses Origin 2019b to fit the measured IR spectra to the peaks, referring to the literature [16,17] for the partitioning of the aromatic structures in the 700–900 cm⁻¹ region, the distribution of oxygen-containing functional groups and C=C bonds of benzene rings in the 1000–1800 cm⁻¹ region, aliphatic hydrogens in the 2800–3000 cm⁻¹ region, and hydroxyl groups in the 3000–3600 cm⁻¹ region. The distribution of hydroxyl functional groups in the 700–3600 cm⁻¹ region was quantified in order to obtain more adequate information on the structure of tar residues.

3. Results and Discussion

3.1. Infrared Spectroscopic Characterisation of Extracts

Figure 1 shows the FTIR spectra of the three solvent single-stage extracts of coal tar residue. The absorption peaks around 3440 cm⁻¹ for the three solvent extracts are the characteristic absorption peaks of joined hydroxyl groups or the absorption peaks of stretching vibrations of intramolecular hydrogen bonds. The peaks of some intensity at 3043 cm⁻¹ for the three solvent extracts are the absorption peaks of C-H stretching vibrations on the benzene ring. Increasing aromatic ring condensation and higher aromatic ring substitution reduces the C-H structure and thus makes the absorption peak weaker. The peak near 1594 cm⁻¹ is the absorption peak of the C=C backbone stretching vibration of the aromatic ring. The multiple absorption peaks between 2850–2920 cm⁻¹ represent the aliphatic hydrogen absorption peaks of the -CH₂- structure, while the absorption peak near 1435 cm⁻¹ is the characteristic aliphatic hydrogen peak of the -CH₃- structure, both of which indicate a high level of aliphatic structure and aromatic ring oligomerisation in the three extracts. 910–650 cm⁻¹ is the characteristic absorption peak of the aromatic compound. The absorption peaks at 910–850 cm⁻¹ are usually the out-of-plane bending vibration absorption peaks of a single (i.e., isolated) hydrogen atom by a substituted benzene ring; at 815–790 cm⁻¹ they are caused by the out-of-plane bending vibration of two neighbouring hydrogen atoms by a substituted benzene ring; at 750 cm⁻¹ they are caused by the out-of-plane bending vibration of three neighbouring hydrogen atoms by a substituted benzene ring; and at 710 cm⁻¹ they are the out-of-plane bending vibration absorption peaks of aromatic ring skeletons. The absorption peak at 710 cm⁻¹ is an out-of-plane bending vibration of the aromatic ring backbone. All of these indicate that the extract contains a high concentration of aromatic compounds [14].

3.2. Infrared Spectral Split Peak Fitting of Extracts

3.2.1. Results of Peak Splitting Fits for Extracts in the 700–900 cm⁻¹ Interval

Figure 2 shows the fitted IR peak splitting curves for the three extracts in the 700–900 cm⁻¹ interval. The peaks of the infrared spectra in this region are the characteristic absorption peaks of aromatic compounds. The position of the absorption peaks in this wave number range can be used to determine the substitution of hydrogen atoms on the aromatic ring. The CS₂ extracts were predominantly disubstituted, tetrasubstituted, and pentasubstituted benzene rings. The acetone extracts and the acetone/CS₂ mixture extracts were dominated by benzene ring trisubstitution. The fitted data results in

Table 2. Infrared spectra of three solvent extracts 700–900 cm⁻¹ split peak fitting ratio of each absorption peak area.

Peak #	Position σ (cm−1)	Assignment	Area Percentage (%)
			ACE	CS2	ACE/CS2
1	900−860	Five adjacent H deformation	9.78	30.27	38.49
2	860−810	Four adjacent H deformations	29.47	23.30	14.82
3	810−750	Three adjacent H deformations	31.28	16.30	27.05
4	750−720	Two adjacent H deformations	27.37	30.13	19.64
5	710	One adjacent H deformations	2.10	0	0

show that acetone and CS₂ can dissolve more aromatic structures containing aliphatic side chains.

3.2.2. Extracts 1000–1800 cm⁻¹ Interval Peak Splitting Fit Results

Figure 3 shows the IR peak fitting curves of the three solvent extracts in the 1000–1800 cm⁻¹ interval. Combining the fitted parameters of the spectra shows that the acetone extracts are dominated by asymmetric vibrational absorption peaks in the C-H plane and absorption peaks associated with the C-O-C band, which is probably the ether group of dibenzofuran. The CS₂ extracts are dominated by asymmetric vibrational absorption peaks in the C-H plane and asymmetric =C-O-C absorption peaks in aromatic ethers. The mixed extracts are dominated by the C-O stretching vibrations of the phenols. From the results of the fitted data in

Table 3. Infrared spectra of three solvent extracts 1000–1800 cm⁻¹ split peak fitting ratio of each absorption peak area [14].

Peak #	Position σ (cm−1)	Assignment	Area Percentage (%)
			ACE	CS2	ACE/CS2
1	1600	Aromatic C=C	8.41	9.68	8.35
2	1470	CH3-,CH2-	4.36	0	15.95
3	1440	Asymmetric -CH3, -CH2	22.80	8.69	7.38
4	1380	Symmetric -CH3, -CH2	6.10	28.27	13.73
5	1320	Ar-O-C	18.56	24.65	9.05
6	1150	C-O-C	17.47	5.99	33.85
7	1110	S=O	0.68	5.32	1.23
8	1050	Si-O-C	21.62	17.39	10.46

, it is clear that the acetone/CS₂ solvent mixture extracts more phenolics.

3.2.3. Results of Peak Splitting Fits for Extracts in the 2800–3000 cm⁻¹ Interval

Figure 4 shows the fitted IR peak splitting curves for the three solvent extracts in the 2800–3000 cm⁻¹ interval, and the corresponding fitted data results are shown in

Table 4. Infrared spectra of three solvent extracts 2800–3000 cm⁻¹ split peak fitting ratio of each absorption peak area.

Peak #	Position σ (cm−1)	Assignment	Area Percentage (%)
			ACE	CS2	ACE/CS2
1	2956	Aliphatic -CH3	12.95	8.36	7.70
2	2919	Asymmetric aliphatic -CH2	38.90	22.16	6.87
3	2881	Aliphatic -CH	5.71	63.51	53.16
4	2850	Symmetric aliphatic -CH2	42.44	5.97	32.27

. The peaks in this interval are attributed to the stretching vibration of the aliphatic hydrocarbon CHx, and this dictates that six to eight peaks are appropriate for the deconvolution of this section of the spectrum. From the fitted parameters it is clear that both extracts are dominated by aliphatic -CH except for the acetone extract, which is dominated by the symmetric -CH₂ stretching vibration.

3.2.4. Results of Peak Splitting Fits for Extracts in the 3000–3600 cm⁻¹ Interval

Figure 5 shows the fitted IR peak splitting curves for the three extracts in the 3000–3600 cm⁻¹ interval, and the corresponding fitted data results are shown in

Table 5. Infrared spectra of the three extracts 3000–3600 cm⁻¹ split peak fitting ratio of each absorption peak area [15].

Peak #	Position σ (cm−1)	Assignment	Area Percentage (%)
			ACE	CS2	ACE/CS2
1	3600−3500	OH-π	8.31	5.83	8.64
2	3500−3350	self-associated OH	61.62	67.84	52.23
3	3300−3150	ring hydroxy	23.89	23.52	33.04
4	3150–3000	OH-N	6.18	2.81	6.09

. There are six main forms of hydroxyl functional groups present in coal, namely hydroxyl-N hydrogen bonding near 3150 cm⁻¹, ring-conjugated hydroxyl hydrogen bonding near 3200 cm⁻¹, hydroxyl-ether oxygen hydrogen bonding near 3300 cm⁻¹, hydroxyl self-conjugated hydrogen bonding near 3410 cm⁻¹, hydroxyl-π hydrogen bonding near 3516 cm⁻¹, and the free hydroxyl group near 3611 cm⁻¹ [18,19]. The fitted parameters of the spectra show that all three extracts are dominated by hydroxyl self-conjugated hydrogen bonds and contain a large number of ring-conjugated hydroxyl hydrogen bonds.

3.2.5. Analysis of the Structural Parameters of the Three Extracts

The structural parameters obtained from the split-peak fitting results were used for the semi-quantitative analysis of the different solvent tar residue extracts. The structural parameters were calculated according to Equations (1) and (2) and the results are shown in

Table 6. Structural parameters of the three extracts.

Extract Sample	DOC	A(CH2)/A(CH3)
ACE	0.91	3.00
CS2	1.07	2.65
ACE/CS2	0.75	0.89

.

DOC = A_900−700/A₁₆₀₀,

(1)

where DOC indicates the degree of aromatic ring condensation of the extract.

A(CH₂)/A(CH₃) = A₂₉₂₀/A₂₉₅₀,

(2)

CH₂/CH₃ characterises the length of the fatty side chains and the degree of branching in the samples. As can be seen from Table 6, the CS₂ extract had a larger DOC, indicating a higher degree of aromatic ring condensation and the presence of more aromatic hydrocarbons with a large number of aromatic rings, while the acetone/CS₂ mixed solvent extract had a smaller DOC value. Comparing the A(CH₂)/A(CH₃) data for the three extracts shows that the acetone extract has the largest CH₂/CH₃ index, indicating that the extract has the longest aliphatic side chain and the fewest branched chains.

3.3. Comparative Infrared Spectroscopy of Coal Tar Residues and Extractives

The IR spectra of the coal tar residue and its three solvent extracts are given in Figure 6. It can be seen that the position of the characteristic peaks did not change before and after extraction, but only the peak intensity changed, which indicates that the ultrasonic-assisted extraction method does not destroy the main structure of the tar residue itself. The peak intensity of the hydroxyl functional group at 3440 cm⁻¹ was somewhat reduced compared to the original sample, indicating that some of the hydroxyl-containing compounds were extracted by the solvent. The 910–650 cm⁻¹ of the absorption peaks were classified as characteristic absorption peaks of aromatic compounds. The intensity of the absorption peak at 745 cm⁻¹ is somewhat reduced compared to the original sample, indicating a reduction of aromatic compounds in the extractives.

3.4. Infrared Spectral Split Peak Fitting of Coal Tar Residue and Extractives

3.4.1. Results of the Peak Splitting Fit for the 700–900 cm⁻¹ Interval

Figure 7 shows the IR peak splitting curves of the coal tar residue and its three solvent extractables in the 700–900 cm⁻¹ region, and the corresponding fitted data are shown in

Table 7. Infrared spectra of four samples 700–900 cm⁻¹ split peak fitting ratio of each absorption peak area.

Peak #	Position σ (cm−1)	Assignment	Area Percentage (%)
			E	ACE	CS2	ACE/CS2
1	900−860	Five adjacent H deformation	44.68	49.25	39.49	42.30
2	860−810	Four adjacent H deformations	15.30	0	0	0
3	810−750	Three adjacent H deformations	17.47	46.89	30.56	29.69
4	750−720	Two adjacent H deformations	22.55	3.86	23.42	25.26
5	710	One adjacent H deformations	0	0	6.53	2.06

. The acetone/CS₂ mixed solvent residue was dominated by benzene ring pentasubstitution and benzene ring tetrasubstitution, while the rest of the two solvent residues were dominated by benzene ring pentasubstitution and benzene ring trisubstitution. The above data show that the aromatic structure of the tar residue after extraction changed significantly, with the proportion of tetrasubstituted or trisubstituted benzene rings increasing and the proportion of pentasubstituted benzene rings decreasing in the three solvent extracts compared to the original samples. This suggests that the aromatic ring compounds may have undergone substitution reactions during the extraction process, resulting in structural changes.

3.4.2. Results of the Peak Splitting Fit for the 1000–1800 cm⁻¹ Interval

Figure 8 shows the IR peak splitting curves of coal tar residue and its three extractives in the 1000–1800 cm⁻¹ region, and the corresponding fitted data results are shown in

Table 8. Infrared spectra of four samples 1000–1800 cm⁻¹ split peak fitting ratio of each absorption peak area.

Peak #	Position σ (cm−1)	Assignment	Area Percentage (%)
			E	ACE	CS2	ACE/CS2
1	1600	Aromatic C=C	11.03	14.63	14.30	14.85
2	1440	Asymmetric -CH3, -CH2	31.37	4.87	27.18	36.16
3	1380	Symmetric -CH3, -CH2	0	33.57	0	3.82
4	1320	Ar-O-C	11.15	13.66	22.59	14.43
5	1150	C-O-C	37.39	13.97	18.90	4.96
7	1110	S=O	0	2.16	2.58	19.91
8	1020	Si-O-C	9.06	17.14	14.45	5.87

. The asymmetric vibrational absorption peaks in the C-H plane of the group were dominant in the CS₂ residue, and the C-N absorption peak was dominant in the CS₂ residue. 1158 cm⁻¹ was the C-O stretching vibration of the phenolic hydroxyl group, and the percentage of the phenolic hydroxyl group was higher in all three residues than in the tar residue. The percentage of this group in the CS₂ extracts was the most variable, at 18.90%.

3.4.3. Results of the Peak Splitting Fit for the 2800–3000 cm⁻¹ Interval

Figure 9 shows the IR peak splitting curves for coal tar residue and its three extractives in the 2800–3000 cm⁻¹ region, and the corresponding fitted data results are shown in

Table 9. Infrared spectra of four samples 2800–3000 cm⁻¹ split peak fitting ratio of each absorption peak area.

Peak #	Position σ (cm−1)	Assignment	Area Percentage (%)
			E	ACE	CS2	ACE/CS2
1	2955	Aliphatic -CH3	9.97	9.74	13.48	20.53
2	2920	Asymmetric aliphatic -CH2	25.36	49.15	47.94	34.72
3	2886	Aliphatic -CH	57.70	10.45	17.33	24.87
4	2853	Symmetric aliphatic -CH2	6.97	30.67	21.25	19.88

. The acetone residue was dominated by symmetric and asymmetric -CH₂ stretching vibrations, while the CS₂ and acetone/CS₂ mixed solvent residues were dominated by asymmetric -CH₂ stretching vibrations.

3.4.4. Results of the Split Peak Fit at 3000–3600 cm⁻¹

Figure 10 shows the IR peak fitting curves of the coal tar residue and its three extractives in the 3000–3600 cm⁻¹ region. The hydroxyl group is the main functional group that forms hydrogen bonds, and hydrogen bonds are a major non-covalent bond in the structure of coal macromolecules. As can be seen in

Table 10. Infrared spectra of four samples 3000–3600 cm⁻¹ split peak fitting ratio of each absorption peak area.

Peak #	Position σ (cm−1)	Assignment	Area Percentage (%)
			E	ACE	CS2	ACE/CS2
1	3600−3500	OH -π	24.55	10.89	8.38	9.25
2	3500−3350	Self-associated OH	52.16	54.01	62.08	63.00
3	3300−3150	Ring hydroxy	20.84	32.16	28.98	27.75
4	3150–3000	OH-N	2.44	2.95	0.56	0

, the three solvent extracts were dominated by hydroxyl self-conjugated hydrogen bonds and ring-conjugated hydrogen bonds near 3200 cm⁻¹.

3.4.5. Analysis of Structural Parameters of Tar Residues and Extractives

In addition to the two structural parameters DOC and CH₂/CH₃, two other structural parameters were used to further characterise the structure of the tar pomace and extractive residues, and the results of the calculated data are shown in

Table 11. Structural parameters of coal tar residue and its three extractive residues.

Sample	DOC	A(CH2)/A(CH3)	AR	A
E	0.87	2.54	2.11	0.33
ACE	1.80	5.05	1.09	0.21
CS2	0.42	3.56	1.28	0.25
ACE/CS2	0.40	1.69	0.84	0.32

.

The aromaticity (AR) is the aromatic C-H bending vibration/fatty C-H, which can be used to determine the degree of aromatisation:

AR = A_700–900/A_2800–3000,

(3)

The A-factor, commonly used to assess the hydrocarbon potential and maturity of cheese roots, is higher for greater A, and lower for greater maturity.

A = A_2800–3000/(A_1490–1600 + A_2800–3000),

(4)

Table 11 shows that the aromaticity of the three solvent extracts was less than that of the original tar residue, indicating that the degree of aromatisation was lower than that of the tar residue. The highest A value for tar residue was 0.33, proving that it was the least mature, and the higher CH₂/CH₃ value indicates the length and branching of the fat side chains in the sample. A comparison of A(CH₂)/A(CH₃) showed that the acetone extracts had the longest fat chains and the least branched chains. A larger DOC for the acetone residue indicates a higher degree of aromatic ring condensation and the presence of more aromatic hydrocarbons with a large number of aromatic rings, whereas the DOC value for the acetone/CS₂ mixture residue is smaller.

4. Conclusions

(1) The tar residue contains a large number of hydroxyl hydrogen bonds and polycyclic aromatic hydrocarbons, and the main substitution mode of the aromatic ring is benzene ring pentasubstitution, therefore, ultrasonic extraction has a large impact on the asymmetric=C-O-C absorption peak and aromatic structure of the aromatic ether in the tar residue.

(2) By comparing the extracts, extract residues, and the original samples after the ultrasonic extraction, it was found that all three extracts contained a large amount of PAHs, CS₂ solvents could dissolve more aromatic structures containing aliphatic side chains, and the acetone/CS₂ mixture could extract more phenols. Among the aliphatic structures, both extracts were dominated by aliphatic -CH, except for the acetone extract, which was mainly a symmetric -CH₂ stretching vibration.

(3) A comparison of the IR results of the original tar residue and the five solvent extracts showed that the main structure of the tar residue did not change throughout the process, but the extraction changed the aromatic structure considerably. Compared to the original sample, the proportion of tetrasubstituted or trisubstituted benzene rings in the three solvent extracts increased, while the proportion of pentasubstituted benzene rings produced a significant decrease.

(4) The analysis of the extracted separated material from coal tar residue was carried out by using infrared fractionation fitting, thus demonstrating the effect of extraction on PAHs and laying the foundation for the subsequent selection of a suitable extractant for the extraction process.