Efficient toluene adsorption/desorption on biochar derived from in situ acid-treated sugarcane bagasse

Carbon-based materials with great adsorption performance are of importance to meet the needs of industrial gas adsorption. Having massive agricultural wastes of sugarcane bagasse, China could use this waste into wealth. However, the comprehensive utilization of sugarcane bagasse as precursor for biochar that can be used as adsorbent has not been extensively explored. In this study, a series of in situ sulfuric acid–modified biochar was prepared by hydrothermal carbonization process. The prepared biochar (SBAC-7) has a combination of two main advantages that are high microporosity (micropore surface area = 1106 m2/g) and being rich in S-containing functional groups on the surface. In particular, SBAC-7 showed an excellent adsorption capacity of toluene (771.1 mg/g) at 30 °C, which is nearly 3 times as high as that of the commercial activated carbons. Meanwhile, it showed great stability and cyclic regeneration performance with five toluene adsorption-desorption test cycles. This study provides a high-performance biochar for the adsorption-desorption cycle in practical engineering applications, and would contribute to the sustainable “sugarcane production–bagasse utilization” circular economy.


Introduction
Volatile organic compounds (VOCs) are widely used in industries including petrochemicals, printing, pharmaceuticals, and painting. VOCs are typical precursors in atmospheric chemistry, contributing to the production of ozone, secondary organic aerosols, and greenhouse gases . In recent decades, various technologies for VOCs removal have been investigated, such as adsorption, membrane separation, catalytic combustion, and photocatalytic degradation (Huang et al. 2020b;Li et al. 2020;Shu et al. 2019;Wang et al. 2020c;Wang et al. 2021). Among them, the adsorption method has been considered as one of the most practical and effective technologies because of its low cost, easy operation, and high treatment effects. Activated carbon (AC) is commonly used as adsorbent of VOCs because of its developed surface area and large pore volumes. However, traditional raw material like coal for AC preparation is a non-renewable resource. And the main disadvantages of using these commercial AC materials for the VOC adsorption are high production costs and secondary pollution during the preparation process.
Biochar derived from hydrothermal carbonization of carbohydrate-rich bio-resources was an ideal material for the purification of polluted water or air (Oliveira et al. 2019). The hydrothermal carbonization process could be directly applied to biomass with high moisture without predrying. Biochar has attracted much attention because of its potential in several crucial fields, such as catalysis, energy storage, CO 2 utilization, and air purification (Wang et al. 2018a resources, sugarcane bagasse (SB) with high carbon content, natural fibrous structure, and huge amount of production is considered as an ideal precursor (Huang et al. 2020a). Sugarcane is a perennial C 4 crop cultivated in subtropical and tropical zones worldwide. The high yields of lignocellulosic SB are considered as an excellent source for substituting fossil fuel as precursors. China is ranked third in the world in sugarcane production (Huang et al. 2020a); however, the utilization of SB as precursor of biochar has not been extensively studied.
There is a growing consensus on modulating the functional group and internal textural structure, which play key factors in improving the adsorption performance of carbon-based material. The surface functional group can be modified by using various methods including acid/base treatment, chemical oxidation, or impregnation with metal elements (Jin et al. 2020;Tang et al. 2020;Wang et al. 2020a). Notably, acid modification can change the surface alkalinity and oxygen-containing functional groups, which thus enhance the VOC selectivity and adsorption capacity (Kim et al. 2006;Tham et al. 2011;Vega et al. 2013b). Pak et al. reported that the AC treated by 10 vol% sulfuric acid showed a 47% increase in toluene adsorption capacity (Pak et al. 2016). Although these methods are effective in increasing the oxygen-containing functional groups on the surface of AC, the specific surface area often decreases during the treatment process due to the block of internal textural structure. On the other hand, there are many other methods to modify activated carbon on internal textural structure, such as microwave modification, heat treatment modification, and hot steam treatment modification (Alslaibi et al. 2013). Based on the above studies, it is recognized that the adsorption capacity of biochar can be greatly enhanced by modulating the surface functional group and internal textural structure simultaneously.
Herein, we prepared a series of biochar by developing an in situ acid-treated process using SB as bio-resource. Toluene, one of the typical VOCs, was chosen as the probe molecule to evaluate the adsorptive properties of the as-prepared samples by the dynamic breakthrough experiments. The desorption property of in situ acid-treated biochar was also studied. This work is expected to expand the utilization of agricultural waste for air pollutant removal, plus a simple, low-cost, and efficient in situ modification method.

Sample preparation
The hydrothermal method was used to prepare activated carbon with deionized water or different concentrations of sulfuric acid solution (the concentration of sulfuric acid = 3, 5, 7, 9 wt%). Typically, 3.0 g of the pretreated SB and 60 mL of deionized water or sulfuric acid solution were mixed in the reaction kettle and hydrothermal carbonization was allowed to proceed for 10 h at 240°C. After cooling down naturally, the black samples were washed thoroughly with deionized water to neutral and dried at 105°C overnight. And the carbonized product was obtained. Then, the carbonized product was impregnated with KOH solution (KOH/carbonized product weight ratio was 1.0) for 12 h. After drying at 105°C for 12 h, the impregnated samples were activated at 800°C for 1 h under N 2 flow, with a heating rate of 10°C/min. After cooling down naturally, the samples were washed with HCl (10 wt%) and deionized water until pH = 7 ± 0.05. Finally, the samples were dried at 105°C for 12 h. The final products were denoted as SBAC-x (x = 0, 3, 5, 7, 9); x means the concentration of sulfuric acid solution.

Characterization
The crystalline phase of the samples was determined by X-ray diffraction (XRD, Rigaku/SmartLab SE), which was referred to the International Centre for Diffraction Data (ICDD). The morphology was detected by scanning electron microscopy (SEM, ThermoFisher/Apreo S HiVac). The specific surface area, pore volume, and pore diameter distribution were measured by N 2 adsorption-desorption isotherms at − 196°C using Micromeritics Tristar 3020. The specific surface area was calculated by using the BET method according to nitrogen adsorption data in the relative pressure (P/P 0 ) range of 0.05-0.30. Sulfur, carbon, and oxygen species in the samples were determined by X-ray photoelectron spectroscopy (XPS, AXIS SUPRA+) and Fourier transform infrared (FT-IR) spectra on a Bruker Tensor II spectrometer.

Adsorption/desorption studies
The toluene adsorption performance of SBAC-x was carried out by dynamic adsorption experiments at room temperature. The simulated exhaust gas consisted of 1000 ppm toluene, 20% O 2 , and N 2 as balance gas. Forty milligrams of sample was weighed and put into a quartz tube, with quartz wool blocked on both sides. Before the adsorption experiments, the sample was firstly degassed in 110°C under N 2 flow to remove those adsorbed impurities. After cooling down to room temperature, the simulated exhaust gas was introduced to flow through the sample at a rate of 100 mL/min at a GHSV of 150,000 mL/(g·h). The tail gas from the reaction tube was led to the gas chromatograph (GC), and the concentration of toluene was noted every 4 min. The adsorption capacity was calculated through the integrals of the breakthrough curve by using Eq. (1) where q e (mg/g) is the calculated adsorption capacity, F (mL/min) is the gas flow rate, M (g/mol) is the relative molecular mass of the adsorbate, C 0 (mg/mL) is the initial toluene concentration, C t (mg/mL) is outlet toluene concentration at time t (min), m (g) is the mass of adsorbent used in the absorption experiment, and t (min) is the adsorption time.
In the toluene desorption experiment, the temperature of the tested samples was elevated from 30 to 450°C with a heating rate of 2.5°C/min under 100 mL/min N 2 . The toluene concentration of the effluent gas was measured by GC. Carbon balance (B c ) is calculated based on the peak areas of toluene desorption (q desorption toluene ).
Adsorption kinetic and isotherm models Adsorption isotherms are considered essential while designing adsorption systems, since they provide information about the distribution of the adsorbate between the gas and solid phases at various equilibrium concentrations. To further understand the adsorption mechanism of toluene on as-prepared SBAC-x, four common models including quasi-first-order, quasi-second-order, Elovich, and Bangham kinetic models were used to fit the experimental data (Tang et al. 2016;Zhang et al. 2019). The models were described in detail as follows: (1) Pseudo-first-order model where q t and q e were the amount of toluene adsorption at time t and equilibrium (mg/g), and k 1 was the quasi-secondorder rate constant (min −1 ).
(4) Bangham model where k is the Bangham constant (min −1 ) and z is a constant.
All kinetic and isotherm models were fitted to experimental data using their nonlinear equations (Table 4). Non-linear regression analysis was performed using the OriginPro 9.0 software, since the non-linear modeling is considered the best for estimating kinetic and isotherm parameters, due to the inherent bias, diverse estimation errors, and fit distortions, which might be resulting from the linearization.

Results and discussion
Textural properties XRD was carried out to investigate the crystallinity of SBACx as shown in Fig. 1. The as-prepared biochar showed broad peaks, indicating the amorphous structure (Gao et al. 2015). The broad peak in the range of 20-30°could be assigned to the (002) plane of amorphous carbon. And the broad hump in the range of 40-50°was related to the (100) plane, which was caused by diffusion scattering of the amorphous carbon (Chen et al. 2012;Cheng et al. 2020).  Table 1. According to Fig. 2a, the sorption isotherm of SBAC-0 presented a hysteresis loop when the relative pressure P/P 0 > 0.4, which was associated with the capillary condensation of N 2 . This phenomenon revealed that the untreated SBAC-0 was rich in mesoporous channels (Wang et al. 2020b), and the surface area (S BET ) and total pore volume (V t ) of SBAC-0 was 1137 m 2 /g and 0.76 cm 3 /g, respectively.
Notably, after the in situ sulfuric acid treatment, the S BET and V t of 5,7,9) were significantly enhanced. Specifically, the S BET of SBAC-3, 5, 7 sharply increased to 2154, 2215, and 2455 m 2 /g, while the V t values increased to 1.24~1.26 cm 3 /g, respectively (Table 1). These results may be attributed to the in situ interaction between sulfuric acid and the fibers of sugarcane during the hydrothermal carbonization, which could provide more adsorption site for the adsorbate. The fibers of the sugarcane are composed of three major components (cellulose, hemicellulose, and lignin), and it was reported that the hemicellulose can be removed by acid through etching effect (Huang et al. 2020a). Thus, both the micropore surface area (1106 m 2 /g) and the mesoporous surface area (1349 m 2 /g) enlarged over the SBAC-7 sample than those of SBAC-0 under an appropriate concentration of sulfuric acid. When further increasing the sulfuric acid concentration to 9 wt%, however, the micropore surface area of SBAC-9 decreased precipitously to 196 m 2 /g, accompanied by the surge of mesoporous surface area (1926 m 2 /g). This implies that the skeleton structure of sugarcane began to collapse under 9 wt% sulfuric acid, resulting in the vanishment and blocking of micropores ). The pore size distribution of the samples has a similar tendency with S BET (Fig. 2b). It was recognized that the micropore played a decisive role in VOC adsorption, especially when the VOC concentration is low . Besides, the diffusion of the VOC molecule is well situated to benefit from the presence of mesopores. Therefore, the SBAC-7 sample treated with the optimized acid concentration is better at keeping micropores and a considerable amount of mesopores, which will own more excellent adsorption property. The morphology of the as-prepared biochar was characterized by SEM as shown in Fig. 3. As shown in Fig. 3a, b, the SBAC-0 was in the shape of a fiber block with a rough surface and few pores on the surface. After the hydrothermal carbonation in certain concentrations of sulfuric acid solution (i.e., 3~7 wt%), the biochar could maintain the vascular bundle structure (Fig. 3c-h), which proved that the cage construction of the SB precursor has a good corrosion resistance. However, when the acid concentration was further increased to 9 wt% (Fig. 3i, j), the pore diameter of SBAC-9 began to increase and the carbon skeleton was found to be dilapidated and hollowed, which finally led to a decrease in the micropore surface area (Wang et al. 2018b). This result is highly consistent with the N 2 sorption isotherms. In addition, the acid-treated SBAC-x  have a more smooth surface, which could be due to the cleaning effect of sulfuric acid solution for the surface impurities (Tang et al. 2016). Note that the etching between sulfuric acid and hemicellulose produced a lot of tiny pores on the surface, thus creating many interconnected channels perpendicular to the stems of SB. With the elevating of the sulfuric Figure 3 SEM images of asprepared activated carbon samples SBAC-0 (a, b), SBAC-3 (c, d), SBAC-5 (e, f), SBAC-7 (g, h), and SBAC-9 (i, j) acid concentration, more external pores appeared (Jain et al. 2016), which was consistent with the increase of the specific surface area of SBAC-x (Table 1).

Surface chemical properties
It had been reported that the type and number of chemical functional groups on the surface had great influence on adsorption performance (Chen et al. 2021). In order to explore the surface chemical functional group of as-prepared samples, FT-IR analysis was applied and is presented in Fig. 4. The weak absorption peaks observed in the 3917-3539-cm −1 range and 672 cm −1 were assigned to the stretching vibration of the dissociative O-H group. The spectra showed a strong absorption peak at 3435 cm −1 represented the O-H stretching vibration in carboxyl and phenol (Pezoti et al. 2016). The band located at 2362 cm −1 was attributed to the C=O stretching vibration, which was due to carbon dioxide in the air. The peak located at 1721 cm −1 was related to the C=O stretching vibration in aliphatic ketone, which appeared after the addition of sulfuric acid. The band at 1630 and 1400 cm − 1 corresponded to the antisymmetric and symmetric stretching vibrations of the -COOgroup. The bands located at 1120 cm −1 and 830 cm −1 were connected with sulfur-containing functional groups, which was symmetrical stretching vibration of O=S=O and C-O-S, respectively. And with the increase of the sulfuric acid concentration, these two peaks became more obvious. The peaks at 1581 and 1123 cm −1 were -SO 2and S=O stretching vibration, which appeared while the sulfuric acid concentration reached 7 wt%. It could be easily observed that the number of sulfur-containing functional groups increased with the increase of sulfuric acid concentration. At the same time, the types and number of oxygencontaining functional groups also increased, which might change the surface charge, hydrophilicity, polarity, and other surface chemical properties of the as-prepared biochar.
In order to further study the existing state of surface elements of activated carbons, the XPS analysis method was adopted and the results are shown in Fig. 5 and Table 2. The survey spectra of all the as-prepared activated carbons contain C 1s, O 1s, and S 2p spectra. The sulfur content of SBAC-0 was only 0.25%, which was due to the biological uptake by sugarcane growth. After sulfuric acid treatment, the surface S content increased to some extent, which indicated that the sulfur element had loaded on the surface of biochar during preparation. Meanwhile, carbon content decreased and oxygen content increased with sulfuric acid concentration increase. The XPS C 1s spectra of the samples show three peaks at the binding energies of 284.7, 286.3, and 289.3 eV, which were related to C-C, C-O, and O=C-OH, respectively (Cheng et al. 2020;Wang et al. 2019). The deconvoluted XPS S 2p signals at binding energies of 160.0, 164.3, 165.3, and 168.9 eV corresponded to functional groups such as S 2− , disulfide (C-S-S-C), sulfinyl group (C 2 S=O), and sulfone (C 2 S(=O 2 )) reported in previous studies (Grzybek et al. 2004;Ting et al. 2018). The XPS O 1s spectra, which are shown in Fig. 5c, can be disassembled into three peaks. The peak at the binding energies of 531.5 eV was ascribed to the O=C of ketone, carbonyl, and/or lactone groups. The band observed at 532.3 eV was assigned to C-O in ether and/or alcohol. And the peak located at 533.3 eV corresponded to O=C-OH (Goel et al. 2015;Guo et al. 2020). It is believed that O=C and C-O were responsible for surface basicity and O=C-OH reflected the surface acidity (Tiwari et al. 2018). The surface adsorption active sites of biochar were related to the surface functional group as mentioned above (Vega et al. 2013a). The amount and proportion of surface basicity increased with the rise of sulfuric acid concentration, which led to a rise in the pH pzc of the zero potential point on the surface, thus enhancing the non-polarity of the activated carbon. Additionally, toluene was a weak or non-polar molecule. The increased surface basicity can promote the adsorption capacity of toluene by increasing the π-π electron diffusion capability on the biochar.  Adsorption capacity of the biochar The dynamic adsorption behaviors of toluene on different samples were considered, and the breakthrough curves are presented in Fig. 6. The corresponding saturated adsorption capacity was 387.6, 641.4, 695.0, 771.7, and 711.8 mg/g from 0 to 9 wt% of sulfuric acid, respectively. Apparently, the above results agreed with the textural properties (specific surface area and pore properties) and surface functionalities (amount and proportion of surface groups). Besides, the breakthrough time was defined as the time when the outlet toluene concentration reached 1% of feed concentration, which was more commonly used in practical applications. For the SBAC-x samples, the breakthrough time was 16, 44, 60, 68, and 60 min, respectively. As is known, the adsorption capacity of commercial activated carbon is usually at 200 to Figure 5 a C 1s, b S 2p, and c O 1s XPS spectra of the as-prepared samples 300 mg/g, which is only one third of that of our best sample (SBAC-7). Table 3 compares the toluene adsorption capacity of the as-prepared SBAC-7 with those of other carbon-based adsorbents reported in the literature, which also indicate the excellent adsorption capacity of SBAC-7 in this work. As described in Fig. 2 and Table 1, the specific surface area and ratio of the micropore reach the maximum value when sulfuric acid was added at 7 wt%. Meanwhile, the carbon skeleton and surface structure were the most abundant. In addition, as shown in Fig. 5 and Table 2, the addition of sulfuric acid enhanced the surface basicity. These suggest that the pore structure and surface functional groups are the core factors to improve the adsorption capacity of biochar for toluene. Although the adsorption capacity of the adsorbent is important, the desorption capacity which determines the regeneration effect is also noteworthy. The common method of desorption is treatment in high temperature with N 2 or water vapor, which means that the lower temperature and higher desorption efficiency could reduce energy consumption as much as possible. In order to understand the regeneration process, a desorption test of SBAC-0 and SBAC-7 was compared and the result is shown in Fig. 7. On both samples, the desorption peak appeared at 90°C, which was much lower than the previously reported 110°C (Zhu et al. 2020). Moreover, the carbon balance of SBAC-7 (ca. 98.7%) was much higher than that of SBAC-0 (ca. 73.6%). This may be explained by the existence of mesopores which promoted toluene transfer and suitable adsorption strength due to the huge amount of Scontaining functional groups. Considering the stability of the SBAC-7 sample, a adsorption-desorption cycle test was performed as shown in Fig. 8. The regeneration temperature in each cycle was set at 90°C according to the desorption test. It was found that the breakthrough time of SBAC-7 did not change remarkably during the five cycles. The saturation adsorption capacity was 771.7, 759.7, 753.3, 748.8, and 742.3 mg/g, respectively. The last adsorption capacity only decreased by 3.8% compared to that of the first time. An analysis of the reactor effluent at the desorption steps during the five cycles confirmed that the toluene was desorbed completely. This is proved by the excellent carbon balances obtained during cycling (> 96%). These results imply that SBAC-7 was renewable, reusable, and recyclable during the adsorption-desorption cycle.

Adsorption kinetic model
The fitting curves of all the considered kinetic models are shown in Fig. 9, and relevant kinetic parameters are estimated in Table 4. In the four kinds of kinetic model for the synthetic SBAC-x samples, the correlation coefficients (R 2 ) of the Bangham model were > 0.99, which were the most close to 1. Meanwhile, the adsorption capacity predicted by this model was closer to the actual measured value, from which we can conclude that the Bangham model was the best model for toluene adsorption in the as-prepared samples. Therefore, the pore diffusion model could better represent the actual adsorption situation. In comparison, the correlation coefficients of the pseudo-first-order and pseudo-second-order kinetic model were less than that of Bangham model. Moreover, the experimental adsorption capacities were remarkably different from the calculated ones. All these indicated that the two models could not reasonably explain the experimental data. Additionally, the Elovich model was also considered as an unsuitable model to explain the adsorption behavior.
It revealed that toluene adsorption involved several parts: toluene adsorption at the surface and diffusion in the pores (Lei et al. 2020). Moreover, the intraparticle diffusion played The consecutive toluene adsorption-desorption cycles of SBAC-7 a major role, which could affect the adsorption rate (Gong et al. 2019). The adsorption rate was affected by the initial concentration and surface pore size. Instantaneous adsorption or external surface adsorption was the result of the superposition effect of the adsorption field. Then was the slow adsorption phase, and the intraparticle diffusion was the controlling  Figure 9 Adsorption kinetic model fit adsorption curve: a pseudo-first-order model, b pseudo-second-order model, c Elovich model, d Bangham model factor of the adsorption rate. The last stage was the final equilibrium stage. As most of the pores were filled with adsorbates, the diffusion of internal particles slowed down further and the adsorption rate decreased further (Ren et al. 2020). These adsorption steps could well explain the piecewise fitting form of the intraparticle diffusion model and prove the important role of intraparticle diffusion as one of the rate control mechanisms.

Conclusions
In short, we presented a facile in situ modification method for biochar with excellent performance. The as-prepared biochar had high specific surface areas, great pore volumes, and an abundant surface chemical group. SBAC-7 exhibited the best toluene adsorption capacity of 771.7 mg/g, which was about 3 times higher than that of commercial ACs, while retaining mild flexibility. Meanwhile, it showed considerable stability and cyclic regeneration performance with five toluene adsorption-desorption test cycles. The outstanding performance was associated with its superior physicochemical properties. On one hand, the etching effect of sulfuric acid was conducive to a higher specific surface area (2245 m 2 /g) and the formation of more micropores. Others, the surface Scontaining functional groups, surged by adding sulfuric acid, which enhanced the surface basicity and non-polarity of the biochar. These factors simultaneously promoted the adsorption and internal diffusion of the toluene molecule. This work provided a valuable guide to produce applicable adsorbents with agricultural wastes for the adsorption of toluene in practical implications.