Low-Temperature Thermal Treatment and Boron Speciation Analysis from Coals

Abstract

Despite urgent calls for decarbonization, the continued increasing demand for electricity, primarily from coals, has presented challenges in managing coal-derived wastes such as coal fly ash (CFA), which are enriched with environmentally hazardous substances like boron. This study explores a low-temperature heating process to remove boron from coal, aimed at preventing its condensation and enrichment into CFA during combustion. Initial boron concentrations in coals varied widely from 50 to 500 ppm by weight and were found to correlate with fixed carbon content (FC) through the following polynomial equation: [B]_o = 0.0929(FC)² − 14.388(FC) + 601.85; R² = 0.9173. This relationship suggests that as coal undergoes coalification, boron-containing compounds are decomposed and released, resulting in a decline in boron levels as the coal matures. Boron-removal efficiency was investigated by drying coal samples at 110 °C, 160 °C, and 210 °C under natural air convection, and nuclear magnetic resonance (NMR) spectroscopy was used to assess changes in boron speciation during heating. Our results demonstrate that boron removal ranged from 5% to 82%, with minimal improvements observed beyond 110 °C. In addition, the ¹¹B MAS-NMR spectra of the coal samples showed four peaks at isotropic chemical shift values of −1.0, 2.0, 8.0, and 14.0 ppm and suggested that the species of boron volatilized at low temperatures is the inorganic BO₄ assigned to peak no. 0 at −1.0 ppm. The association of boron with inorganic components in coal suggests potential for efficient removal, particularly in coals with higher fixed carbon content. These findings highlight the viability of low-temperature thermal treatment as a cost-effective method for boron removal, which is crucial in mitigating the risks associated with coal combustion by-products.

1. Introduction

In recent decades, improved living standards, technological industrialization, and extensive urbanization have led to an ever-increasing demand for electricity generated from various sources such as coal, oil, gas, nuclear, hydro, solar, and wind. Due to the risk of excessive global warming associated with greenhouse gas emissions, there have been insistent calls to move away from carbon-based energy sources such as coal. Despite these initiatives, the global share of electricity production by coal combustion remains consistent at about 35% of the total energy mix from the early 1980s to present [1,2,3].

In Japan, after the Great East Japan Earthquake and the ensuing Fukushima Daiichi Nuclear Accident, nuclear power stations began to be regarded as highly unsafe by the public and gained a negative impression such that energy production transitioned away from nuclear sources. This change in public perspective resulted in a considerable increase in the share of coal-fired power generation from 25% in 2011 to 33% in 2013. As suggested by market prices at that time, coal-fired power stations were considered as the least expensive marginal plant [4], resulting to the sustained increase in the consumption of coal to fuel one-third of Japan’s electricity mix. In 2018, the country imported more than 210 million tons of coal, making Japan the world’s third largest coal-importing country after China and India. About 160 million tons (76%) of coal were utilized to generate electricity in coal-fired electric power stations, while the remaining portion was used for iron and steel production [5,6]. Although Japan’s Ministry of Economy, Trade, and Industry (METI) initially aimed to decrease coal’s contribution by over 50% in 2030 to respond to the growing call for decarbonization, METI now predicts that the future energy mix will consist of approximately 21% nuclear, 23% renewable, 27% natural gas, and 26% coal, suggesting that the domestic use of coal is expected to persist in the future.

An inevitable consequence of continued coal consumption is the accumulation of industrial wastes such as coal fly ash (CFA). In 2013, more than 12.9 million tons of coal fly ash were generated in Japan [7], and under the Fundamental Law for Establishing a Sound Material-Cycle Society in Japan, recycling was rigorously carried out as a raw material for cement, landfill cover, or land reclamation [8]. However, condensation and enrichment of volatile and environmentally hazardous substances into coal fly ash during coal combustion has become a major impediment towards its effective recycling.

One of these compounds of concern is boron, which was recently enlisted as an environmentally regulated substance by the Ministry of Environment of Japan in March 2001. Although boron is an important micronutrient, it still poses risks when consumed by plants and animals at elevated levels. Using the official test method “Notification #46”, the amount of boron that may be eluted into the soil has been limited to below 1.0 mg L⁻¹ since the toxic mechanism of boron was found to start at concentrations exceeding this value [8]. Meanwhile, the World Health Organization (WHO) set a 2.4 mg L⁻¹ maximum boron content in drinking water [9]. In plants, consumption of high levels of boron was found to decrease crop yields and trigger genetic variation [10]. It was also found to slow down the rate of cell development [11,12], cause the decay of leaf tissues [13,14], and create abnormalities in fruits [15]. In humans and animals, boron intake at high concentrations and for a long duration of time was found to lead to adverse effects such as nausea, diarrhea, lethargy, dermatitis, and cardiovascular, nervous, and reproductive issues [16,17,18].

The amount of boron naturally present in coal usually ranges from 5 to 400 ppm by weight [19]; however, during combustion of coal, boron volatilizes and condenses in coal fly ash, where it becomes enriched up to more than 500 ppm by weight [20]. Consequently, recent studies have shown that the amount of boron leached from coal fly ash using the Notification #46 method could reach up to 9.1 mg L⁻¹, which is way above the environmentally acceptable limit [21]. The current status quo therefore suggests that the presence of boron in coal has become a bottleneck against safe and effective coal fly ash recycling, urging efficient and inexpensive methods of boron removal from coal.

To address this issue, it is crucial to understand the current methods that are being explored to treat boron from coal and its processing by-products, such as adsorption, chemical precipitation, advanced oxidation, and nanotechnology. Adsorption techniques make use of bio-waste-derived adsorbents [22] and carbon-based materials [23] and have shown promise in removing boron from ground water and wastewater effluents. Meanwhile, chemical precipitation methods that use layered double hydroxides (LDH) through ion exchange and surface complexation [24] and chemical oxo-precipitation (COP) using calcium-based chemicals [25] were also proposed as potent methods to treat boron-containing flue gas desulfurization wastewater. Li et al. (2023) and Liu et al. (2024) also explored more advanced technologies via electrocoagulation and electrocoagulation-flotation methods [26,27], and Ee et al. (2023) explored the use of hyperbranched nanocellulose functionalized with amino-polyol cellulose nanocrystals for rapid boron removal [28]. However, while these methods show promise in treating boron-containing effluents from coal combustion, they are typically costly and require strict control of process operations. More importantly, they only cater to post-processing waste streams, such as wastewater discharge, rather than targeting boron removal prior to coal processing itself.

Our group also previously investigated the volatilization behavior of boron from coal by heating at 200 °C to 400 °C. It was observed that the vaporization of boron occurs at 300 °C and that the boron content in coal decreases with increasing temperature [29]. Based on these prior findings, we decided to investigate the use of a low-temperature thermal treatment process to preferentially remove boron from coal, which can be utilized as a simple and low-cost pretreatment process to prevent the subsequent condensation and enrichment of boron into coal fly ash. In this study, the removal of boron from coal by drying at 110 °C, 160 °C, and 210 °C was investigated by analyzing the boron concentration and speciation in coal before and after heating.

2. Materials and Methods

Ten coal samples with origins from several countries were obtained through the Japan Coal Energy Center (JCOAL), where more than 100 kinds of coal samples were collected.

Table 1. Selected properties of the coal samples obtained from JCOAL. (FC: fixed carbon, VM: volatile matter, W: water, and A: ash).

Coal		SS007	SS028	SS030	SS039	SS045	SS054	SS066	SS068	SS082	SS087
Boron content (ppm)		103.79	60.89	503.56	170.14	234.38	61.85	130.22	55.89	95.53	143.80
Calorific value (J g−1)		27,740	28,493	26,945	24,518	19,790	29,706	30,418	33,681	33,054	29,079
Origin		AUS	AUS	JPN	CAN	JPN	AUS	IDN	VTN	AUS	IDN
Industrial analysis	FC	51.10	60.07	41.54	43.67	31.19	54.17	49.91	87.60	58.80	45.41
(wt%, JIS M 8813)	VM	31.60	26.79	37.28	35.12	28.49	32.51	42.35	6.41	33.32	43.26
	W	2.30	5.16	1.74	9.21	2.67	2.20	5.96	1.41	1.29	6.73
	A	15.00	7.98	19.44	12.00	37.65	11.12	1.78	4.55	6.58	4.61
Elemental analysis	C	70.90	74.02	64.77	64.84	48.58	74.18	78.42	89.15	78.38	70.92
(wt%, JIS M 8813)	H	4.52	4.39	5.00	4.84	4.04	4.86	5.57	3.21	5.06	4.82
	O	7.53	11.19	7.45	15.49	7.08	7.63	12.16	1.46	7.30	17.26
	N	1.14	1.72	1.05	1.38	0.86	1.58	1.68	1.04	2.10	1.65
	S	0.56	0.27	1.95	0.23	0.76	0.37	0.28	0.53	0.49	0.41

shows some of the selected physical and chemical properties of these samples as provided by JCOAL (Standard Coal Samples Database, 2014), and Figure 1 shows the actual photos of these samples. These samples were selected based on varying fixed carbon (FC), volatile matter (VM), water (W), and ash content (A), as these parameters were assumed to have some degree of influence on the amount of boron removal.

2.1. Low-Temperature Drying

To determine the feasibility of removing boron from coal via a low-temperature drying process, coal samples were subjected to a revised moisture determination method based on ASTM D3173-73 [30]. Initially, two types of drying experiments were performed to determine the influence of the presence or absence of airflow on boron removal and verify that boron does not re-deposit in coal under any of these conditions. One batch of experiments was performed under natural air convection conditions inside the muffle furnace using all 10 coal samples, while another batch was carried out in a horizontal tube furnace with a supplied airflow using 5 out of the 10 coal samples, as shown in Figure 2.

Natural air convection experiments were carried out first by drying porcelain crucibles inside a drying oven at 110 °C in air for 1 h to remove the pre-adsorbed moisture. The crucibles were then cooled in a desiccant and then weighed. Approximately 1 g of each coal sample was placed inside the crucible using a spatula and then weighed. The crucibles without covers were then placed inside the muffle furnace and heated at 110 °C, 160 °C, and 210 °C for 1 h. After heating, the crucibles were removed from the furnace, the covers were replaced immediately, and the crucibles were cooled in a desiccant and then weighed. Boron concentration and speciation in the resulting coal samples were analyzed using the methods described in the proceeding sections.

The second batch of experiments was performed using a horizontal tube furnace with supplied airflow. Using a spatula, about 1 g of coal sample was placed on a pre-weighed platinum sheet arranged on top of an MgO boat and then weighed. The MgO boat with the sample was then inserted into the constant temperature zone of the horizontal furnace. The heating temperatures were set at 110 °C, 160 °C, and 210 °C for 1 h with clean gas (79% N₂-21% O₂) flowing at a rate of 200 mL min⁻¹. After heating, the MgO boat was removed from the furnace, cooled, and then weighed. The resulting coal samples were similarly analyzed for boron concentration and ¹¹B speciation using the methods described below.

2.2. Determination of Boron Concentration

To determine boron concentration, coal samples were initially dissolved into solution using a microwave-assisted digestion method described in detail in one of our papers [29] and based on EPA Method 3052 for the digestion of organic substances [31]. First, 0.05 g of each coal sample was digested using 10 mL of an acid solution containing a mixture of 13.6 M HNO₃ and 18 M H₂SO₄ at a 5:3 volumetric ratio. Then, the coal sample and acid mixture were placed in Teflon^® vessels for the microwave digestion step using Speed Wave 4, Berghof. The heating pattern employed in this study was set at 200 °C for 20 min, 230 °C for 20 min, and then 250 °C for 20 min, with ramp-up times of 5 min each. After microwave heating, the samples were allowed to cool down naturally, and complete digestion of the coal sample was confirmed visually. The resulting solutions containing dissolved coal samples were then diluted and measured for boron concentration using inductively coupled plasma–mass spectroscopy (ICP-MS, HP4500 Agilent, Santa Clara, CA, USA).

2.3. Determination of ¹¹B Speciation by NMR

Approximately 100 mg of each of the coal samples was used for ¹¹B speciation analysis using magic angle spinning–nuclear magnetic resonance (MAS-NMR) spectroscopy. Each sample was placed into a Si₃N₄ tube with an outer diameter of 4 mm and then placed inside the bore hole of the equipment. In this study, an NMR spectrometer (JNM-ECA 600, JEOL, Tokyo, Japan)with an external magnetic field strength of 14.1 T was utilized. The maximum magic angle-spinning rate was set at 15–16 kHz, and chemical shifts in the NMR spectra were referenced to boron trifluoride etherate (BF₃·OEt₂). Data processing and line shape simulation of the resulting data were conducted using Delta v5.0.4.4 (JEOL) and DMfit [32], respectively.

3. Results

3.1. Influence of Airflow on Boron Removal

Figure 3 shows a comparison of the changes in boron content after heating at 110 °C, 160 °C, and 210 °C for 1 h under natural and supplied air flow for five coal samples (SS030, SS039, SS045, SS066, and SS087). The dashed line indicates the initial boron content, [B]_o, of the samples. It was found that the presence or absence of supplied airflow did not have a substantial effect on the amount of boron removal. At 110 °C, samples SS030 and SS045 showed slightly elevated boron concentrations, which was most likely due to difficulty in obtaining a fully representative sample of 0.05 g for the boron concentration measurements or, potentially, the uneven distribution of boron within the samples. It was also reported elsewhere that the determination of boron content in coal involves a substantial level of uncertainty due to possible volatilization of the sample during acid digestion [33,34,35,36].

It was further observed that for most samples, more boron was removed under natural airflow conditions, with convection movement proving enough to remove volatile boron from the samples at low heating temperatures. This trend could potentially be due to the cooling effect of the supplied air in the furnace due to the absence of a pre-heating step, effectively lowering the actual temperature of the coal samples. These results validate that boron did not re-deposit into coal despite the absence of supplied airflow. Moving forward, the subsequent data presented and discussed in this paper are taken from experiments under natural air convection, as this method will be more practical and economical when applied at an industrial scale.

3.2. Change in Boron Concentrations in Coals

The boron concentration in coals tend to vary widely, and in this study, the initial boron content of the coal samples was measured to range from 50 to 500 ppm by weight, as shown in

Table 2. Boron concentration in the coal samples before and after heating at 110 °C, 160 °C, and 210 °C for 1 h under natural air convection.

Sample	[B]o	110 °C			160 °C			210 °C
		[B]	ΔB	%(ΔB)	[B]	ΔB	%(ΔB)	[B]	ΔB	%(ΔB)
	ppm	ppm			ppm			ppm
SS007	103.8	30.2	73.6	70.9	35.3	68.5	66.0	37.4	66.4	64.0
SS028	60.9	37.2	23.7	38.9	35.7	25.1	41.3	40.4	20.5	33.7
SS030	503.6	479.1	24.5	4.9	455.3	48.2	9.6	456.6	47.0	9.3
SS039	170.1	95.8	74.4	43.7	87.6	82.5	48.5	84.0	86.1	50.6
SS045	234.4	184.2	50.2	21.4	168.7	65.7	28.0	146.1	88.3	37.7
SS054	61.8	31.3	30.5	49.3	32.5	29.4	47.5	26.6	35.3	57.0
SS066	130.2	110.2	20.0	15.3	86.7	43.5	33.4	105.8	24.4	18.7
SS068	55.9	18.4	37.5	67.1	13.3	42.6	76.1	12.4	43.5	77.9
SS082	95.5	70.4	25.1	26.3	53.9	41.7	43.6	78.2	17.3	18.1
SS087	143.8	120.4	23.4	16.3	110.9	32.9	22.9	111.7	32.1	22.3

. In general, no relationship between moisture (W), ash (A), or volatile matter (VM) content and initial boron concentration in the coal samples was found. It was, however, observed that the initial boron content in the samples generally decreased with fixed carbon (FC) content, as shown in Figure 4. Excluding coal sample SS030, the following polynomial relationship could be fitted with a high R² value, indicating good correlation between fixed carbon content and initial boron concentration:

[B]_o = 0.0929(FC)² − 14.388(FC) + 601.85; R² = 0.9173

(1)

These results suggest that with an increasing degree of coalification, boron-containing compounds in coals are decomposed and released, leading to a reduction in boron concentration as the coals mature. Consequently, contamination of boron from inorganic mineral sediments or from groundwater is also possible, as suggested by the significant deviation found for sample SS030. The origin of boron in coals remains debated [37,38], but the very high concentration of boron in sample SS030 despite its low fixed carbon content confirms that boron in coal could potentially come from external sources.

Table 2 also summarizes the data for the boron concentration in coal after the low-temperature thermal treatment process. It was found that heating the coal samples at 110 °C, 160 °C, and 210 °C for 1 h under natural air convection conditions removed approximately 17 to 88 ppm of boron, i.e., 5 to 82% of boron, from the original coal samples. Generally, the change in boron concentration increased with heating temperature; however, fluctuations were also observed, which can again be attributed to difficulty in obtaining a true representative sample due to the very small sample size used in boron concentration measurements or potential uneven distribution of boron throughout the sample. The data for boron concentration after heating suggest that increasing the temperature from 110 °C to 210 °C did not necessarily lead to a significant increase in boron removal from coal and that heating at 110 °C should be sufficient.

The mass percent change in boron concentration at 110 °C, 160 °C, and 210 °C for 1 h under natural air convection as a function of (a) water, (b) ash, (c) volatile matter, and (d) fixed carbon contents in the original coal sample is shown in Figure 5. It was found that the change in boron concentration with respect to initial boron content in coals tends to decrease with increasing moisture (W) and volatile matter (VM) content and increase with increasing ash (A) and fixed carbon (FC) content. Although this relationship is a little bit more scattered for moisture (W) content, these trends confirm that boron in coal may be associated with either the inorganic or organic portion of coals or possibly both. Three modes of boron occurrences in coal have previously been identified as either bound to the organic fraction, locked inside the clay minerals such as illite, or bound within the crystal lattice of tourmaline [39]. Among these, the most common occurrence was reported as that portion of boron bound to the organic component, which supports the trends observed in Figure 5.

However, significant deviation from this trend was observed for SS030 and SS045 samples, where higher ash contents of 19.44% and 37.65%, respectively, did not necessarily lead to greater change in boron concentrations. At this point, it is evident that sample SS030 is indeed an exemption. These results imply that while the initial boron concentration is related to the fixed carbon portion of most coal samples, the total amount of boron removed by heating may be related not only to fixed carbon content but to the ash content as well.

3.3. ¹¹B Speciation in Coals

Nuclear magnetic resonance (NMR) spectroscopy was also employed to examine the change in boron speciation during heating. NMR exploits the magnetic properties of atomic nuclei to determine the physical and chemical properties of atoms or molecules in which they are contained. Studies on the ¹¹B speciation in coals, however, have been found to contradict each other. Zhang et al. (2008) proposed that boron was primarily bound to the organic portion of the coal [40], whereas Burchill et al. (1990) and Kuwabara et al. (2007) resolved that boron was mainly bound to inorganic minerals and other borates in coal [41,42].

For sub-bituminous and bituminous coals (70–90% carbon content), previous NMR studies detected three peaks of four-coordinate ¹¹B [43,44]. The NMR signals in the downfield region, denoted as peak nos. 3 and 2, were located at 16.1 and 10.0 ppm, whereas one signal in the up-field region, denoted as peak no. 1, was located at 3.9 ppm. All three peaks were assigned to four-oxygen-coordinated boron atoms (BO₄ unit). Peak nos. 3 and 2 were assigned to organoboron complexes, whereas peak no. 1 was attributed to the inorganically bound boron and is potentially the reason for inconsistencies in previous findings. For blind coals (>90% carbon content), the resonance peak at the most upper field at approximately −1.0 ppm was found consistent with that of illite. Previous works have also shown two varying sets of three ¹¹B MAS-NMR peaks with isotropic chemical shift values at 3.9, 10, and 16.1 ppm [29] and at 0, 8, and 14 ppm [42,43,45].

Figure 6 shows the ¹¹B MAS-NMR spectra of all ten coal samples used in this study arranged in terms of increasing fixed carbon content from top to bottom. Four peaks were identified at isotropic chemical shift values of −1.0, 2.0, 8.0, and 14.0 ppm, which are comparable to the peaks determined by Kuwabara et al. (2007) and Takahashi et al. (2011) [42,43]. It was noted that the ¹¹B MAS-NMR peaks broadened significantly with an increase in fixed carbon content and, consequently, with a decrease in initial boron concentration, which supports our initial findings.

All peaks could be fitted with a Lorentz curve, and the quadrupole coupling constant (CQ) values were measured at ~1 MHz; hence, it was determined that all peaks could be assigned to a four-oxygen-coordinated boron atom (BO₄). While previous studies have only shown three distinct peaks of boron, the present work shows an extra speciation of boron, particularly that in sample SS039, which distinctly shows all four peaks. In this work, these peaks are labeled as peak nos. 0, 1, 2, and 3 from the up-field to down-field region, with isotropic chemical shift values of approximately −1.0, 2.0, 8.0, and 14.0 ppm, respectively.

Boron associated with peak nos. 2 and 3 was previously established to be organoboron coordinated by aromatic ligand [43]. Peak nos. 0 and 1, on the other hand, appear to be very similar in nature due to their proximity in terms of the ¹¹B MAS-NMR spectra. Correspondingly, the peak at −1.0 ppm could be fitted with a Lorentz curve with quadrupole asymmetry parameter (η) value of zero. The quadrupole coupling constant of this peak was also estimated at 1.0 to 1.3 MHz using the following equation [44]:

C_Q = (10Wv_o)^1/2

(2)

where W is the full-width at half-maximum value of the peak, and v_o is 192.4 MHz, i.e., the Lamour frequency value of the ¹¹B atom at an external magnetic field of 14.1 T. These values suggest that peak nos. 0 and 1 belong to an inorganic BO₄ unit, which confirms previous assignment of peak no. 1 to an inorganic four-coordinate boron compound such that the same can also be said about peak no. 0 [43].

3.4. Boron Species Removed by Drying

To determine the explicit species of boron removed from coal by drying, calculation of the relative peak intensity ratios was carried out based on the hypothesis that the area of an NMR resonance signal is directly proportional to the number of nuclei or simply the concentration represented by that signal [46]. The relative changes in concentration of the associated boron species can therefore be determined by simply calculating the relative ratios of the processed peak intensities using the values obtained from DMfit.

Table 3. Fitted area under ¹¹B peak nos. 3, 2, 1, and 0 and the corresponding peak ratio between peak no. 0 to peak no. 4 of as-received and heated coal samples.

Sample	Conditions	Concentration (%)				Peak Ratio
		Peak No. 3	Peak No. 2	Peak No. 1	Peak No. 0	0/3
SS030	Initial	26.46	15.33		58.20	2.20
	110 °C	27.23	19.12		53.65	1.97
	160 °C	32.18	16.30		51.52	1.60
	210 °C	30.61	19.08		50.32	1.64
SS039	Initial	31.23	26.78	15.55	26.44	0.85
	110 °C	34.26	31.44	12.57	21.72	0.63
	160 °C	30.45	26.85	26.22	16.48	0.54
	210 °C	28.71	14.32	44.09	12.88	0.45
SS045	Initial	33.40	18.00		48.60	1.46
	110 °C	31.62	23.31		45.07	1.43
	160 °C	33.88	21.09		45.03	1.33
	210 °C	29.18	16.08	23.26	31.48	1.08

summarizes the area fitting values under ¹¹B peak nos. 3, 2, 1, and 0 and the corresponding peak ratios between peak no. 0 to peak no. 4 of the as-received and heated coal samples. Meanwhile, Figure 7 shows the relationship between peak no. 0 (inorganic BO₄) to peak no. 3 (organoboron complex) integrated area ratios for SS039, SS030, and SS045, i.e., the samples that clearly demonstrated the presence of the new peak no. 0 at approximately −1.0 ppm, and the change in boron concentration, %(ΔB). It was apparent that the intensity of peak no. 0 relative to peak no. 3 increased with decreasing change in boron concentration. This relationship was found to be consistent for the three samples, with R² denoting a strong downhill linear relationship.

Only three coal samples, SS039, SS030, and SS045, were used in this analysis due to the very strong broadening and overlapping of the ¹¹B MAS-NMR peaks in the other samples, making it difficult to achieve precise peak measurements while fitting the raw spectra. Nonetheless, it was indicated that for these samples, the boron associated with peak no. 0 assigned to the inorganic BO₄ unit was preferentially removed during heating at low temperatures. For the peak like that of peak no. 0, the efficient cross-polarization between ¹¹B and ¹H nuclei suggested that peak no. 0 might contain combined water molecules [43]. It is therefore possible that most of the boron removed by drying at relatively low temperatures originated from the boron associated with peak no. 0 due to the evaporation of the combined water.

4. Discussion

For practical purposes, determination of the ¹¹B MAS-NMR peaks is rather tedious, implying the importance of developing a simple method in predicting the amount of boron that can be removed from coal by drying. Moisture (W) content in coal would be a good and straightforward parameter to measure, but after a substantial amount of storage and transport time, moisture content values could be a little ambiguous. Conveniently, it is generally agreed that as coal matures, the gradual elimination of moisture because of physical and chemical changes occurs [47]. This makes fixed carbon content a more suitable and reliable parameter in predicting the changes in boron content, as suggested by the association of the inorganic boron species to the boron portion removed by low-temperature drying.

The relationship between fixed carbon content and change in boron concentration, %(ΔB), is shown in Figure 8 for all coal samples used in this study after heating at 110 °C, 160 °C, and 210 °C. Although there is a slight scattering of data, R² suggests that there is a moderate-to-strong positive relationship between fixed carbon content and boron removal. The findings of the present work imply that more efficient boron removal is therefore achievable in coals with higher fixed carbon content, with most of the boron removed due to low-temperature heating associated with the inorganic portion of coal. Meanwhile, further studies are needed to determine whether enough boron is removed via the proposed method to prevent its elution from the resulting coal fly ash generated during combustion.

5. Conclusions

This study determined the feasibility of removing boron from coal using a low-cost, low-temperature thermal treatment process. The initial boron concentration in coals was found to vary widely, ranging from 50 to 500 ppm of boron by weight, and it obeys the following polynomial relationship:

[B]_o = 0.0929(FC)² − 14.388(FC) + 601.85; R² = 0.9173

where (FC) is fixed carbon content. This indicates that as coal undergoes coalification, boron-containing compounds break down and are released, causing a decrease in boron levels as the coal matures.

Boron removal from the coal samples was carried out by a simple drying method under natural air convection at 110 °C, 160 °C, and 210 °C. The results from the present study indicate that approximately 17 to 88 ppm of boron, i.e., 5 to 82% of boron, from the original coal samples could be removed by drying. It was also found that increasing the temperature from 110 °C to 210 °C does not necessarily lead to a significant increase in boron removal from coals, and heating at 110 °C should be sufficient.

The ¹¹B MAS-NMR spectra of the coal samples used in this study showed four peaks at isotropic chemical shift values of −1.0, 2.0, 8.0, and 14.0 ppm that were assigned to four-oxygen-coordinated boron atoms (BO₄ unit). It was further identified that the species of boron volatilized at low temperatures is the inorganic BO₄ assigned to peak no. 0 at −1.0 ppm, and more efficient boron removal is feasible in coal with higher fixed carbon content, which is of particular importance in predicting the amount of boron that can be removed for practical and industrial purposes.

Compared to other boron-removal techniques that are typically costly and require complex control of operations that are only suitable to post-combustion waste streams, the proposed low-temperature thermal treatment process can address boron removal from coal early in the processing chain. However, the findings drawn in this study are applicable only to the specific coal samples examined, requiring additional test work across a broader range of coal samples to enhance accuracy.