The adsorption properties of microporous activated carbon prepared from pistachio nut shell for low-concentration VOCs under low-medium temperatures

The control of low-concentration VOCs in coal-fired flue gas is one of the research hotspots at present. In this work, K2CO3 and K2CO3-KCl were employed to activate the agricultural wastes (pistachio nut shell) to prepare activated carbon (AC), named PSAC-1 and PSAC-2, respectively. By testing the adsorption performance of the prepared AC and commercial activated carbon (CAC) for the five target VOCs, it was observed that the adsorption capacity of PSAC-2 was the best compared to the other two. Particularly, the adsorption capacity of PSAC-2 (225 mg·g-1) for phenol was 3.8 times that of CAC (59 mg·g-1). In addition, the pseudo-first-order model, pseudo-second-order model, and Elovich model all fitted the adsorption process well, which indicated that both physical adsorption and chemical adsorption existed simultaneously, in which physical adsorption played a dominant role and chemical adsorption played a minor role. Weber-Morris kinetic model was used to illustrate the rate-controlling mechanism; the results confirmed that the stage of external membrane mass transfer was the control stage of adsorption rate. The results of this study can provide some references for the commercial production of biomass-derived AC and the removal of VOCs in coal-fired flue gas.


Introduction
With the continuous development of China's economy and the improvement of urbanization, VOC emissions would continue to grow. During "The Fourteenth Five-Year Plan" period, in order to further improve air quality, the prevention and control of VOCs were one of the "protagonists." Research had shown that VOCs in southern China mainly had the following sources, paint solvent usage, paint solvent usage + liquefied petroleum gas (LPG) usage, biomass burning, coal burning + industry combustion source, gasoline vehicle exhaust gas, diesel exhaust, industrial sources, the fuel evaporation (gasoline), contributing 11%, 22%, 13%, 17%, 12%, 8%, 11%, and 6%, respectively . Compared with other sources, coal and industrial combustion sources accounted for a large proportion of VOCs, reaching 12%. Coal-fired VOCs discharged into the atmospheric environment can not only affect human body, animals, and plants because of its pathogenicity and carcinogenicity but can also be considered as the precursor of ozone, dioxin, and aerosol in the atmosphere, and they do great harm to the environment (Liu et al. 2020). There were many researches on the control of VOCs, but most of them focused on low-temperature and high-concentration VOCs, which were not applicable for the VOCs in coal-fired flue gas with large gas capacity, medium and high temperature, low concentration, and complex composition (Cheng et al. 2018).
The control methods of VOCs could be divided into recovery (such as adsorption, absorption, condensation) and destruction (such as combustion, photocatalytic oxidation) methods. Among these technologies, adsorption was widely regarded as an effective strategy to remove VOCs due to its simple operation and low cost (Qu et al. 2009). Owing to its porous structure and surface properties, activated carbon (AC) was the most common adsorbent used in adsorption. However, the high cost of preparation of commercial activated carbon limited its large-scale industrial application (Cardoso et al. 2008). The preparation of high-performance AC from cheap agricultural wastes such as rice husk (Liu et al. 2012a), bamboo (Deng et al. 2015), coconut shell (Sarswata and Mohan 2016), date palm root (Hadoun et al. 2013), and wood sawdust (Foo and Hameed 2011) had attracted many researchers' attention.
Pistachio nut shells were agricultural wastes in recent years, with a composition of 4.0% moisture, 73.4% volatility, 21.6% fixed carbon, and 1.0% ash (Foroushani et al. 2016). Pistachio nut factories produced nearly 1 million tons of pistachios for consumption each year (Marett et al. 2017). The shells accounted for less than half of the total nut mass, estimating that between 400,000 and 500,000 tons of pistachio nut shells are produced globally every year (Açıkalın et al. 2012). Numerous scholars at home and abroad had carried out researches on the preparation of AC from pistachio nut shell. Yang and Lua (2003, Lua et al. 2004, Lua and Yang 2004a, c, d, 2005, Yang and Lua 2006 prepared AC from pistachio nut shells during 2003-2009, by controlling the activation method, the activation temperature, the types of activator, pyrolysis environment, the pore structure, and chemical properties of AC prepared under different experimental conditions were studied. In addition, Kaghazchi et al. (2010) and Dolas et al. (2011) made researches on how to prepare high specific surface area pistachio nut shell AC; Kamandari et al. (2013Kamandari et al. ( , 2015, Foroushani et al. (2016), Bazan-Wozniak et al. (2017), and Nasernejad (2017, 2018) had researched the effects of the preparation process and process parameters on the characteristics of pistachio nut shell AC; the application of pistachio nut shell AC had been researched by Vijayalakshmi et al. (2010). Thus, pistachio nut shell is a kind of potential biomass carbon precursor.
This was demonstrated in a number of studies that carbonization and activation were the key factors that affect the adsorption capacity of biochar. K 2 CO 3 was a commonly used activator for its low cost and easy recovery, though it had never been used in the production of pistachio nut shell biochar. K 2 CO 3 was employed to prepare pistachio nut shell biochar, and then biochar was applied to control VOCs in coalfired flue gas, which not only realized the reuse of agricultural waste, but also had great significance for the purification of environmental pollution.
Herein, this work investigated the effect of different activators on the pore structure of ACs prepared from pistachio nut shell. The adsorption experiments of ACs on various VOCs were conducted to evaluate the adsorption capacity of ACs and the influence of adsorbate characteristics on the adsorption performance. The pseudo-first-order kinetic model, pseudo-second-order kinetic model, Elovich kinetic model, and the Weber-Morris kinetic model were used to analyze the adsorption behavior and mechanism of VOCs on ACs. The results of this study can provide some references for the commercial production of biomass-derived AC and the removal of VOCs in coal-fired flue gas.

Raw material
Pistachio nut shell was selected as a precursor of biomassderived AC. K 2 CO 3 was selected as the activator and KCl as the combined activator to investigate the adsorption effect of AC prepared by combined activation. Furthermore, commercial coconut shell AC, denoted as CAC, was selected for comparison. The chemical reagents were all analytical grade.
Benzene, toluene, p-xylene, p-dichlorobenzene, and phenol, five typical VOCs commonly found in coal-fired flue gases (Fernández-Martıńez et al. 2001;Cheng et al. 2018), were chosen as adsorbates. Their main properties are shown in Table 1. Molecular dimension was first calculated by using Materials Studio software to draw organic molecules; then, DMol3 method was used to optimize the molecular structure. Measure the bond length and bond angle of the VOCs through software, and the length and width of molecules were calculated by combining the van der Waals radius of each atom (Yang et al. 2016).

Preparation of pistachio shell ACs
The pistachio nut shell was washed by ultrapure water, dried, and crushed; the shell and activator were mixed with 100 mL ultrapure water and dried at 80°C for 24 h. Subsequently, the dried shell was placed in a tube furnace under nitrogen atmosphere and raised to 700°C at a heating rate of 10°C/min, held at 700°C for 2 h. Washing the sample to the neutral of filtrate with diluted hydrochloric acid and ultrapure water, the product was ground into powder after being dried and denoted as PSAC-n (n = 1, 2, 3, 4) The use of activator and the nomenclature of product are shown in Table 2.

Characterization
The specific surface area, pore volume, and pore size distribution of AC samples were measured by automatic three-station specific surface and pore distribution instrument (Microtra BELSORP-Max, Japan). Please refer to Text S1 (supplementary data) for the specific calculation method of the characterization employed in this work.
The surface morphology of the sample was observed by field emission scanning electron microscope (FE-SEM) (ULTRA PLUS-43-13, Germany). The X-ray photoelectron spectroscopy (XPS) (Thermo ESCLAB 250xi, USA) was used to determine the content of various elements on the surface of AC samples and the types and distribution of various functional groups.

Adsorption process analysis
The selected target pollutants' adsorption kinetics behavior onto adsorbents was fitted via the pseudo-first-order model, pseudo-second-order model, and Elovich model (Shafiei et al. 2018). Weber and Morris proposed a kinetic model for the intraparticle diffusion to determine rate control during adsorption (Jiang et al. 2019). The analysis involved in the model is detailed in text S2(supplementary data).

VOCs adsorption test
The experimental facility is shown in Fig. 1, which is mainly composed of three parts, VOCs generator, a fixed-bed, and VOC concentration detection device. VOCs are generated through two paths: N 2 carrier gas and N 2 balance gas. The flow rate of N 2 carrier gas was controlled by mass flow meter and then entered the thermostat to take out the volatile organic vapors and mixed with N 2 balance gas in the static mixer. Moreover, the concentration of VOC gas was controlled by adjusting the inner radius of VOC evaporate bottle mouth, the temperature of the thermostat, and the amount of balance and carrier gas.
The concentration of VOC gas was measured online by a portable total hydrocarbon analyzer equipped with a flame ionization detector (POLARIS FID 300, Italy). After passing through the static mixer, the gas was divided into two paths to measure the inlet and outlet concentration. The saturated adsorption capacity of the adsorbent could be calculated by the breakthrough curve formula as follows: where q e (mg·g -1 ) represented the saturated adsorption capacity of VOCs per unit mass of adsorbent, Q (L· min -1 ) was the total gas flow rate, C 0 (mg·m -3 ) was the inlet VOCs concentration, m (g) was the adsorbent mass, t e (min) was the adsorption equilibrium time, C t (mg·m -3 ) was the outlet VOCs concentration at any time t, and t (min) was the adsorption time.

Results and discussion
Material characterization results

Pore development and surface microstructure
The nitrogen adsorption-desorption isotherms and pore diameter distribution of ACs are depicted in Fig. 2, as well as the specific surface area, pore volume, and average pore size parameters of different ACs could be seen in Table 3. The N 2 adsorption-desorption isotherms of PSAC-1 and PSAC-2 increased rapidly in the relatively low-pressure region, whereafter became horizontal (Fig. 2a), and they were type I physical adsorption isotherm according to the IUPAC classification (Ma et al. 2020), suggesting that they were microporous materials (microporosity: 94.41% for PSAC-1, 92.41% for PSAC-2). The N 2 adsorption-desorption isotherms of CAC belonged to type I and type IV mixture physical adsorption isotherm (Gao et al. 2013); in Fig. 2a, microporous adsorption occurred in the range of 0~0.4 relative pressure, followed by hysteresis loop due to capillary condensation, indicating CAC contained mesoporous structure (microporosity: 57.55% for CAC). PSAC-3 and PSAC-4 showed almost no adsorption of nitrogen (Fig. 2a), and PSAC-3 had only a small amount of mesoporous structure (Fig. 2b), and the pore size distribution of PSAC-4 could not be calculated according to the isotherm data of PSAC-4 ( Fig. 2a). In Fig. 2b, CAC, PSAC-1, and PSAC-2 had rich pore structure. Compared with CAC, PSAC-1 and PSAC-2 had narrower micropore structures, which were centrally distributed in 0.5~0.7 nm. The mesoporous structure of CAC was more abundant than other two adsorption materials. The specific surface area of the four adsorbents except PSAC-4 was arranged as PSAC-3 (3.24 m 2 g -1 ) < CAC (686.76 m 2 g -1 ) < PSAC-1 (758.25 m 2 g -1 ) < PSAC-2 (1012.2 m 2 g -1 ).
The disadvantage of the 77K N 2 adsorption method was that the diffusion kinetic energy of N 2 molecules was very low at low temperatures without sufficient diffusion into the narrow microporous pores, so the measured adsorption data were not realistic. The 273K CO 2 adsorption method could make up for the shortcomings of the 77K N 2 adsorption method to characterize the microporous pore structure.
In Fig. 3, the characteristic curves of the three ACs were all linear, manifesting that CO 2 was mainly used to fill the micropores. The pore volume of the micropores (Table. 4) was much larger than that measured by nitrogen adsorption (Table. 3). The average pore size of CAC was mesoporous, and PSAC-1 and PSAC-2 were microporous, it was consistent with N 2 test results, but the average pore diameter of ACs was significantly smaller than that measured by N 2 adsorption. Besides, the characteristic curves of PSAC-1 and PSAC-2 deviated under low pressure since they contained very small and narrow micropores, and CO 2 molecules could not be sufficiently diffused under low pressure. Figure 4 showed surface microstructure of ACs the under the field emission scanning electron microscope, it can be seen that PSAC-1 (Fig. 4b) and PSAC-2 ( Fig. 4c/d) contained rich pore structure, and the pore development was better than CAC (Fig. 4a). There were many tiny pores on the wall surface of PSAC-1 and PSAC-2; combined with its high porosity, the results inferred that this should be the main area of adsorption. Additionally, the electron microscope images of PSAC-2 (Fig. 4c) depicted the phenomenon of pore sleeve pore. In contrast, the surfaces of PSAC-3 (Fig. S1a) and PSAC-4 (Fig. S1b) were smooth without obvious pore structure.

Surface chemistry
XPS was adopted to analyze the chemical composition and surface functional groups of ACs. The corresponding results are shown in Fig. 5 and Table 5. The XPS spectra of the ACs had three peaks, two peaks of C1s (284.7 eV) and O1s (532.4 eV) were more prominent, and the peaks of N1s (400.5 eV) were relatively weak. The C1s peak could be segmented into five peaks, which were located at 284.6 ± 0.2 eV, 285.0 ± 0.2 eV, 286.5 ± 0.2 eV, 288.5 ± 0.2 eV, and 290.4 ± 0.2 eV, respectively. The five peaks corresponded to graphite carbon, C-O (connected to hydroxyl group and C in ether), C=O (C in carbonyl group), COOH (C in carboxyl group), and π-π* (Meng et al. 2019;Liu et al. 2012b;Puziy et al. 2008), indicating that ACs contained some oxygen-containing functional groups. Compared with CAC, biochar had relatively lower content of π-π*; relatively high content of graphite carbon, carboxyl group, and ester group; and little difference between ether, hydroxyl group, and carbon group. The relative contents of the three elements (C, N, O) of PSAC-1 and PSAC-2 were similar, but the functional group contents were  different. After mixed activation by adding KCl, the graphite carbon was reduced; however, the ether, hydroxyl, carbon group, carboxyl group, ester group, and π-π* were increased at the same time.

Mechanism analysis of activator influence on carbonization activation
Through the characterization and analysis of two materials (PSAC-3, PSAC-4), it was found that the specific surface area of PSAC-3 was only units digit; the specific surface area of PSAC-4 was too small to be calculated. Furthermore, the two carbon materials had no obvious pore structure under field emission scanning electron microscope. Hence, activation of biomass without activator or only with KCl did not have good activation effect. K 2 CO 3 played a primary role in the activation stage. First point, K 2 CO 3 reacted directly with the carbon in the biomass to increase the activation rate, as shown in the following chemical reactions Eqs. (2)-(3) .
Next point, under the gasification condition of C-H 2 O (g), K 2 CO 3 can form surface complexing salt CO -K + . Due to the electron-donating effect of potassium, the carbon chain broken and disconnected, and then complexing salt was formed again, as well as the catalysis process of ring-opening -chainbreaking -ring-opening was repeated. In this way, at the beginning of activation, there was a condition for continuous gasification reaction near the active point containing potassium salt or potassium oxide; then, gasification was preferred to produce macropores by successive reactions with this point and adjacent carbon.
When K 2 CO 3 -KCl was used for combined activation, the presence of KCl enhanced the electron-donating effect of potassium and promoted the above ring-opening -chainbreaking -ring-opening reaction process, whereafter generated the pores and expanded tiny pores. It was also conducive to the gas passes through more biomass sites, generated pores. Consequently, the specific surface area of PSAC-2 was higher than PSAC-1, but the microporosity was lower than PSAC-1, Fig. 4 The field emission scanning electron microscope images of ACs  Figure 6 showed the dynamic adsorption process of CAC, PSAC-1, and PSAC-2 at 40°C for 5 target pollutes (benzene, toluene, p-xylene, p-dichlorobenzene, and phenol) with initial concentration of 110 mg·m -3 and their corresponding saturation adsorption capacities. When the outlet concentration reached to 95% of the inlet concentration, it was considered that the adsorption reached saturation, which was the saturation point. The dynamic adsorption process was influenced by many factors. In addition to external conditions such as adsorption temperature and initial concentration of adsorbates, the properties of adsorbates and adsorbents also affected the adsorption process.

Influence of adsorbent characteristics on adsorption process
Combined with the data of ACs in Table 3 and the adsorption capacity of each ACs (Fig. 6), it was be observed that large specific surface area is favorable for adsorption on account of more adsorption sites could be provided for VOCs (Kim and Ahn 2012). Li et al. (2020) investigated that the adsorption  5 The XPS spectra of survey (a), the C1s peak and the fitted curves for CAC (b), the C1s peak and the fitted curves for PSAC-1 (c), the C1s peak and the fitted curves for  capacity was linearly correlated with the specific surface area and pore volume of the adsorbent, with linear correlation coefficients of 0.9782 and 0.8969, respectively. The pore size distribution of the ACs used in this paper varied considerably; thus, the results were inconsistent with it. In the process of adsorption, macropores and mesoporous usually played the role of channels, and micropores especially the narrow micropores were the main place for adsorption. Lillo-Ródenas et al. (2005) investigated that the narrow micropore content better reflected the adsorption capacity of ACs for low-concentration benzene compared to the micropore content. The adsorption capacity of ACs for low-concentration VOCs was largely dependent on the narrow micropores. PSAC-1 and PSAC-2 were rich in narrow microporous structure with significantly higher microporosity than CAC. Accordingly, although the pore volume of CAC was larger than that of PSAC-1, the specific surface area and VOC adsorption of PSAC-1 were larger than that of CAC.

Influence of adsorbate characteristics on adsorption process
Three adsorbates with the same intramolecular element composition, benzene (80.1°C), toluene (110.6°C), and paraxylene (138.5°C), were selected for discussion here. Apart from the adsorption of PSAC-1 to p-xylene, the adsorption capacity increased with the increase of the boiling point of the adsorbates. The main reason was that VOC adsorption on porous adsorbents was similar to the process of vapor-liquid phase transformation and liquid-like condensation; in other words, the adsorbates changed from gas phase to liquid phase in the adsorbent pores (Chiang et al. 2001;Guo et al. 2012). VOCs with high boiling point were preferentially adsorbed as they were easier to be converted into liquid and had stronger intermolecular forces with adsorbents. In Fig. 6d, CAC and PSAC-2 had no obvious increase in pxylene adsorption, since the molecular size of the adsorbent also affected the adsorption process. In general, when the molecular size of VOCs was larger than the pore size of AC, spatial site resistance occurred and adsorption did not happen. When the molecular size of VOCs was equal to the pore size of AC, AC had a strong ability to trap VOCs and it was not easy to desorption. When the molecular size of VOCs was smaller than the pore size of AC, capillary condensation took place in the pores of AC, making the adsorption capacity of biochar larger. When the molecular size of VOCs was much smaller than the pore size of AC, the adsorbed molecules were easily desorbed, leading to a decrease in adsorption capacity . CAC, PSAC-1, and PSAC-2 all contained narrow micropores, and PSAC-1 showed the highest content of narrow micropores and the widest distribution, with the Fig. 6 Breakthrough curves of VOC adsorption by CAC (a), breakthrough curves of VOCs adsorption by PSAC-1 (b), breakthrough curves of VOCs adsorption by PSAC-2 (c), and the corresponding adsorption capacities (d).
narrowest pore size up to 0.6 nm or less. In Table 1, the molecular size of p-xylene was larger than that of benzene and toluene. Hence, it might be attributed to the fact that the molecular size of p-xylene was larger than the pore size of ACs causing steric hindrance, resulting in a decrease or insignificant increase in its adsorption.
It can be observed from Fig. 6 that the adsorption capacity of p-dichlorobenzene was larger than that of benzene, toluene, and p-xylene with smaller molecular size, because the boiling point of p-dichlorobenzene was nearly 40°C higher than benzene, toluene, and p-xylene, and the molecular weight difference exceeded 40 g·mol -1 , which made the conversion of pdichlorobenzene in the adsorbent pores much easier and the liquid phase condensation adsorption increase. Such a rule can also explain that the adsorption capacity of toluene and pxylene was higher than that of benzene. Besides, the abundant oxygen-containing functional groups on the surface of AC could increase the dipole-dipole interaction between AC and p-dichlorobenzene, promoting adsorption (Almazán-Almazán et al. 2007).
It was evident from Fig. 6 that the adsorption capacity of ACs for dichlorobenzene was greater than that of phenol with higher boiling point and smaller molecular size. Owing to the presence of hydroxyl groups in the phenol molecule, with higher polarity than the other four VOCs, and the predominance of non-polar C-C on the surface of ACs, the adsorption capacity of ACs for polar molecules phenol was much weaker than that of non-polar molecules.

Influence of temperature on adsorption
Temperature was another important external factor affecting adsorption, the increase of temperature would increase chemical adsorption and affect the adsorption process of AC with high micropore content. The related analysis is detailed in text S3 (supplementary data).

Kinetics model fitting
The adsorption processes of two VOCs (toluene and pdichlorobenzene) with large differences in molecular weight and boiling point on micro mesoporous AC (CAC) and microporous AC (PSAC-2) were selected for kinetic model fitting, respectively (Fig. 7). According to the calculation results (Table 6), pseudo-first-order kinetic model fitting results were better, R 2 (0.9894~0.9965) was greater than that of the other two models, and the fitted equilibrium adsorption amount was more in line with the experimental value. Based Fig. 7 Nonlinear fitting curves of three kinetic models. Fitting curves of toluene adsorption process by CAC (a). Fitting curves of toluene adsorption process by PSAC-1 (b). Fitting curves of p-dichlorobenzene adsorption process by CAC (c). Fitting curves of pdichlorobenzene adsorption process by  on the assumption of the pseudo-first-order kinetic model (Drenkova-Tuhtan et al. 2017), the adsorption rate was determined by the diffusion step in the adsorption process.
In Table 6, the rate constant (k 1 ) of PSAC-2 was lower than that of CAC, which was caused by the richer mesoporous content of CAC and the faster diffusion rate of VOCs in it. So, a certain number of mesopores might benefit to accelerate the adsorption rate. The rate constant (k 1 ) of p-dichlorobenzene was lower than that of toluene, due to the higher molecular weight and boiling point of p-dichlorobenzene, resulting in a lower molecular diffusion rate.
Pseudo-second-order kinetic model was also good fit for adsorption of VOCs with R 2 value reaching 0.9826~0.9892; the Elovich model could reach R 2 values of 0.9447-0.9831 except for the fit of p-dichlorobenzene in the adsorption process of PSAC-2, which indicated that the adsorption process was also affected by certain chemical adsorption ; it was related to the part of oxygen-containing functional groups on the surface of ACs.

Rate-controlling mechanism
To further determine the adsorption rate-controlling mechanism, it was simulated by using the Weber-Morris kinetic model to obtain the adsorption rate constants (Table 7) for three stages (Jiang et al. 2019). In Fig. 8, none of the straight line crosses through the origin, illustrating that intraparticle diffusion was not the only mechanism controlling the adsorption rate. Combining with the above adsorption kinetic analysis, this phenomenon might be caused by the presence of chemical adsorption in the adsorption process, which accelerated the adsorption rate.
The VOC adsorption process can be explained as follows according to the above analysis. (A) This stage had a flat curve (Fig. 8) and small values of k A (Table 7); VOC molecules diffused from the gas phase to the surface of AC, which was mainly affected by the outer surface area of AC and initial concentration of VOCs; this stage controlled the adsorption process rate. (B) The diffusion rate at this stage was faster than the last stage (Fig. 8, Table 7); VOC molecules were adsorbed on the surface of ACs and diffused into the internal structure of ACs through the pores of ACs. Hence, the diffusion rate depends on the pore diameter distribution of ACs. (C) With the increases of VOC molecules entering the AC, the free channels of VOC molecules became narrower, the diffusion process was blocked, in Fig. 8 the curve of this stage flattened again, the adsorption process slowed down, and finally the adsorption reached equilibrium. The adsorption mechanism of VOCs with low concentrations on PSAC-2 was proposed based on the above analysis. The adsorption mechanism is shown in Fig.9.

Conclusions
In this work, PSAC-2 prepared by K 2 CO 3 -KCl activation from pistachio shells showed the best pore structure and VOCs adsorption capacity. The adsorption capacities of five typical VOCs on CAC, PSAC-1, and PSAC-2 were in the range of 36 mg·g -1~9 9 mg·g -1 , 71 mg·g -1~1 63 mg·g -1 , and 130 mg·g -1~3 23 mg·g -1 , respectively. The properties of adsorbates and adsorbents affected the adsorption process; the larger specific surface area and micropore volume of ACs, the higher boiling point, molecular weight, and the weaker the molecular polarity of VOCs were conducive to adsorption. Both physical and chemical adsorptions existed in the adsorption process, with physical adsorption dominant. VOC adsorption involved three stages, and the external membrane mass transfer was the adsorption rate-limiting step due to the lower VOC concentration. Availability of data and materials All data generated or analyzed during this study are included in this published article and its supplementary information files.

Declarations
Ethical approval Not applicable.
Consent to participate Not applicable. Fig. 8 Intraparticle diffusion kinetic plots for toluene and pdichlorobenzene adsorption. Fitting of toluene adsorption processes at CAC and PSAC-1, respectively (a). Fitting of pdichlorobenzene adsorption processes at CAC and PSAC-1, respectively (b) Fig. 9 Mechanism of VOC adsorption on PSAC-2