Physiochemical Properties of Biochar and Activated Carbon from Biomass Residue: Influence of Process Conditions to Adsorbent Properties.

This study evaluates the influence of hydrothermal carbonization (HTC) or slow pyrolysis (SP) process conditions on the physicochemical properties of precursor biochars and activated carbon (AC). The AC is achieved through a direct or a two-step method with subsequent chemical activation using KOH. A theory is developed on the biochar propensity to be chemically activated based on the lignocellulosic structure composition. X-ray photoelectron spectroscopy elemental analysis shows that the O/C ratio decreases after chemical activation for HTC biochar but remains the same for SP biochar. X-ray powder diffraction indicates that the SP biochar and all ACs have broad amorphous carbon peaks, whereas corn stover and the HTC biochar have distinct cellulosic crystalline peaks. Vanillin adsorbent experiments were performed on various ACs with up to 98% reduction shown. The best adsorbent for vanillin was the AC produced directly from corn stover, followed by AC HTC and then AC SP.

Introduction

In a circular bioeconomy, maximizing the use of lignocellulosic biomass waste is paramount for the full utilization of energy, products, and chemical commodities with minim<span class="Chemical">al environmental harm.[1] For example, corn stover is one of the most produced agricultural residues in the United States and is one of the primary feedstocks for cellulosic ethanol.[2] Corn stover, consisting of stalks, leaves, and cobs, was removed from approximately 6.3% of corn operations in the United States, suggesting an ample supply with a minimal demand.[3] In addition to bioethanol, corn stover can be used for other purposes, including fibers, hydrocarbons, and animal feed. The full utilization of lignocellulosic biomass by turning corn stover into value-added products will help progress the circular bioeconomy, increase the agricultural sector’s profitability, and decrease the dependence on non-renewable resources.

One beneficial and low-cost v<span class="Chemical">alue-added product that can be produced from corn stover is activated carbon (AC).[4,5] AC is a high surface area, porous structure made from various carbon sources using either a direct one-step or a two-step process requiring an initial carbon precursor before activation. AC has several uses,[6−8] especially as an adsorbent for wastewater treatment. Wastewater treatment plants utilize AC to remove pollutants, such as dyes, pharmaceuticals, heavy metals, and organic contaminants. The removal of industrial phenolic waste such as vanillin is especially important because it has adverse environmental effects.[9−11] The adsorbent capacity is often one of the key metrics used to evaluate the effectiveness of ACs. Converting agricultural residues into AC for wastewater treatment can lower the costs and create a more sustainable pathway to clean drinking water.

AC properties, such as surf<span class="Chemical">ace functional groups, pore size, and surface area, can be modified to fit the desired criterion or application. The AC properties can often be tailored by changing the reaction parameters for the preparation of biochar precursors and activation methods. One way to form biochar precursors is the use of hydrothermal carbonization (HTC). HTC is a green process that uses mild temperatures, water as a solvent, and an inert environment. HTC can produce three fractions: solid (char), liquid (bio-oil), and gas. This process is performed at a point where water is subcritical, which is useful in breaking down the polymeric backbone of the biomass. HTC uses temperatures between 180 and 250 °C[12] and has several advantages compared to pyrolysis, including lower energy inputs, no need of drying the feedstock, reduced ash content, and higher solid yields.[13] Pyrolysis, an alternative to HTC, is a standard thermal method to convert biomass into biochar. It can use low or high temperatures with little to no oxygen. There are three major categories of pyrolysis based on the duration and associated temperature ramp: flash, slow, and fast pyrolysis.[14] For slow pyrolysis (SP), the ramp rate is on the order of minutes or hours, ranging between 10 °C/min and 10 °C/h. The temperature range is lower than that of the flash pyrolysis, less than 500 °C.[15] To optimize biochar production, one should focus on utilizing low temperatures and moderate ramp rates like that of SP.

AC can be made from biochar precursors using either physic<span class="Chemical">al or chemical activation. Physical activation requires two separate steps: first with pyrolysis or thermal treatment and then exposure to an oxidizing gas such as steam or CO2.[16] Chemical activation can be done either by a direct one-step process or a two-step process. The direct one-step process involves impregnating or mixing biomass with a chemical activating agent under thermal treatment.[17] In the one-step method, the carbonization and activation step are performed simultaneously. The two-step process is done by an initial carbonization step to form biochar, followed by thermal activation.[18] Several types of chemicals can be used as activating agents, including K2CO3, NaOH, ZnCl2, and KOH.[18−20] In all cases, the activating agent is used with an ideal ratio to biomass to ensure complete activation and formation of pores and improved surface area. Several groups have studied the activation mechanism using KOH. Huang et al.[21] found that KOH reacts with carbon around 530 °C to produce K2O and K2CO3, which react to create metallic K and a graphite-like microcrystalline structure. Otowa et al.[22] also found the formation of K2O by dehydration and K2CO3 by a carbonate reaction. Metallic K is formed and intercalated in the carbon matrix at high temperatures, resulting in atomic layers of carbon being widened and forming pores.

The method used for biochar precursor preparation can ultimately affect the AC properties (surf<span class="Chemical">ace area, porosity, and prevalent surface functional groups). This study performs an in-depth investigation on how biochar precursors derived from the direct method, HTC, and SP can affect the physiochemical properties of ACs. Vanillin adsorption experiments are used to probe the adsorption capabilities of AC. Vanillin was chosen to represent a phenolic pollutant in water. We utilized chemical activation with KOH due to its superior dehydrating abilities and propensity to achieve the highest surface areas in AC. There are many studies on using HTC[23−25] or SP[15,26,27] of biomass residues to produce biochar; however, only a few compare the two methods for corn stover.[28] This article analyzes the biochar precursors, ACs, and probe the AC’s physiochemical and adsorptive properties.

Results and Discussion

Physiochemical Properties of Biochar

Hydrothermal Carbonization

We studied the influence of dwell temperature at 200, 220, and 240 °C on the formed biochar after hydrothermal treatment. Figure a shows the X-ray powder diffr<span class="Chemical">action (XRD) pattern of HTC-treated corn stover with a 2 h dwell time at different temperatures. The patterns are compared with that of untreated corn stover. The XRD pattern of corn stover has prominent cellulosic peaks at ∼16 2θ (101) and ∼22 2θ (220). The cellulosic peaks of the formed biochars decrease in intensity and become broad as the temperature increases. This broadening coincides with the decrease in the crystallite sizes of the cellulose and hemicellulose. It indicates that the HTC process promotes the partial breakdown of the cellulosic and hemicellulosic components of corn stover. Interestingly, the turbostratic carbon (t-carbon) peak at ∼26 2θ increases as the temperature increases, showing the potential growth of graphene layers. T-carbon is a unique class of carbon having structural ordering in between that of amorphous carbon phase and crystalline graphite phase.[29]

Figure 1

(a) XRD spectra of the HTC biochar prepared at different temperatures and a dwell time of 2 h. (b) Surface areas of the HTC-formed biochar plotted as a function of temperature.

(a) XRD spectra of the HTC biochar prepared at different temperatures and a dwell time of 2 h. (b) Surf<span class="Chemical">ace areas of the HTC-formed biochar plotted as a function of temperature.

Figure b shows the change in the surface area as a function of dwell temperature and time. The biochar formed at 200 °C for 1 h produced the lowest surf<span class="Chemical">ace area at 1.0 ± 0.12 m2/g. As the temperature increased to 220 °C, the surface area increased to 3.5 ± 0.38 m2/g. This change may coincide with cellulose chains hydrolyzing at temperatures > 220 °C, as corroborated with our XRD data.[30] At 240 °C, the surface area decreased to 2.9 ± 0.34 m2/g. The same trend was observed with the 4 h dwell time at 200, 220, and 240 °C with surface areas of 3.3 ± 0.36, 4.5 ± 0.50, and 2.8 ± 0.41 m2/g, respectively. Based on our XRD and surface area analysis, we believe that the surface area increases as the biomass constituents are broken down, with the hemicellulose degrading first and the cellulose second. We would expect this degradation order as hemicellulose is much less resistant to hydrolysis than cellulose.[31] Any further thermal degradation of biomass may occur through hydrolysis, isomerization, dehydration, and fragmentation.[32,33] The surface areas for the 2 h dwell time at dwell temperatures 200, 220, and 240 °C were 3.0 ± 0.17, 2.6 ± 0.20, and 6.9 ± 1.3 m2/g, respectively. We note that the highest achieved surface area for the HTC experiments was found using a 2 h dwell time and 240 °C dwell temperature. Figure S1a shows the average pore size as a function of temperature and dwell time. The average pore size increased when the temperature increased from 200 to 220 °C in all the cases. The increase in the pore size could coincide with the widening of pores caused by cellulose chain hydrolyzing. When the temperature was further increased to 240 °C, in the 1 h run, there was a slight increase in the average pore size, but in the 2 and 4 h runs, the average pore size decreased.

The color of the biochar varied with temperature and dwell time (Figure S2). The biochar color varied from light brown to a black fine powder consistency. The milled corn stover had the appearance of sawdust prior to the hydrotherm<span class="Chemical">al treatment. The color change of the biomass is most likely due to the Maillard reaction and the degradation of the sugars contained in the biomass.[34] The morphological changes of corn stover for select biochar samples were studied using scanning electron microscopy (SEM). Figure b,c shows the evolution of corn stover to HTC biochar as a function of temperature and duration. As the temperature and time increase, the surface becomes rougher, indicating structural degradation from the hydrothermal process. Figure a shows the milled corn stover before HTC. The corn stover structure is rigid and fibrous. Figure b shows the biochar formed after the HTC process at 220 °C and 1 h dwell time. The surface changes are apparent with the formation of pits that are possible starting locations for pore formation. As the temperature and time are increased, as shown in Figure c,d, to 240 °C for 4 h dwell time, the most dramatic changes are observed, with little to no resemblance to the original corn stover. Hydrothermal degradation occurs due to a myriad of simultaneous reactions including hydrolysis,[12,25] dehydration,[31] and decarboxylation.[35] At higher temperatures, other reaction mechanisms are dominant such as condensation polymerization.[35,36] The temperatures of this study are sufficient for the promotion of these simultaneous reactions that become prevalent at 240 °C and promote the degradation of corn stover. The effect of temperature on the properties and morphology of HTC biochar is complex and multifaceted. Despite a greater degree of degradation at higher temperatures, our findings indicate no clear trend in the surface area and pore size.

Figure 2

SEM images of the HTC biochar at different temperatures and dwell time. (a) Corn stover milled 1 mm (mag: 1.43k×). (b) HTC 220 °C 1 h (mag: 1.43k×). (c) HTC 220 °C 2 h (mag: 1.15k×). (d) HTC 240 °C 4 h (mag: 1.17k×).

SEM images of the HTC biochar at different temperatures and dwell time. (a) Corn stover milled 1 mm (mag: 1.43k×). (b) <span class="Chemical">HTC 220 °C 1 h (mag: 1.43k×). (c) HTC 220 °C 2 h (mag: 1.15k×). (d) HTC 240 °C 4 h (mag: 1.17k×).

We varied the ratio of water to biomass from 8:1 to 5:1 or 10:1 at 220 °C to determine if the amount of <span class="Chemical">water influenced the biochar characteristics. Table S1 shows the results for varying the ratio. The 5:1 ratio had the lowest surface area at 2.2 ± 0.24 m2/g, the 8:1 ratio had a higher surface area at 2.6 ± 0.41 m2/g, and the 10:1 ratio had the highest surface area of 3.6 ± 0.40 m2/g. Interestingly, the 8:1 experiment had the lowest solid biochar yield, while the highest biochar yield was achieved from the 5:1 ratio. Considering the solid loss after the experiments, we surmise that the difference in solid retention is a higher production of liquid and gaseous products for the 8:1 ratio compared to the other ratios. Nevertheless, the higher water-to-biomass ratio improves the surface area and pore structure.

We wanted to determine if the reactor’s volume influenced the biochar’s properties as the amount to charge the re<span class="Chemical">actor would be of interest for scaling up this process. We maintained the 8:1 ratio for water to biomass and studied HTC on 1, 5, and 10 g of biomass. Table S1 shows the comparison of the three different runs as a function of added water volume. The run with 1 g of corn stover achieved the highest surface area at 5.6 ± 0.62 m2/g. However, the biochar yield was very low due to some of the biochar residue adhering to the reactor’s walls. The run for the 10 g sample indicated a higher biochar yield but a lower surface area, 4.1 ± 0.45 m2/g. We surmise the different mass to reactor volume allowed for different heat transfer rates, which in turn promoted different biochars. Thus, the reactor size has significance when optimizing the formation of biochar.

Slow Pyrolysis

Figure a shows the XRD patterns for the biochar derived from SP at 300, 500, and 700 °C. The XRD spectra for the biochar formed at 300 °C show cellulosic peaks comparable to the peaks for corn stover. This indicates that at 300 °C, there was a minimal change in the structur<span class="Chemical">al characteristics of the biochar. The biochars formed at 500 and 700 °C are quite different, with the formation of broad peaks between 15 and 30 2θ, resembling amorphous carbon peaks. We surmise that temperatures above 500 °C were substantial enough to break down the lignocellulosic structure for the complete breakdown of cellulose and hemicellulose. Interestingly, all the HTC biochars have cellulosic peaks, while those of SP biochars at elevated temperatures above 500 °C were apparently amorphous. This is an indication of different reactions occurring in the different carbonization methods. The biochars from SP and HTC are both precursors to AC, and comparing the biochar formed from the two different processes is essential. This breakdown in the structure corresponds to a higher surface area for SP biochar compared to that formed from HTC.

Figure 3

(a) XRD spectra for SP of corn stover at different temperatures. (b) Surface area of SP of corn stover for 1 h over a range of 300–700 °C.

The change of the surface area as a function of the SP temperature is shown in Figure b. The trend shows that as the temperature increases, the solid residue’s surf<span class="Chemical">ace area increases as well. From 300 to 500 °C, the surface area went from 1.5 ± 0.03 to 5.0 ± 1.5 m2/g, but once the temperature reaches 550 °C, the surface area increases to 111 ± 23 m2/g. An SP 240 °C experiment was conducted to compare with the HTC 240 °C biochar. Compared to HTC, the SP biochar’s surface area formed at 240 °C is 1.2 ± 0.2 m2/g versus the HTC biochar’s surface area of 6.9 ± 1.3 m2/g. We surmise that HTC, using subcritical water, may be more effective with breaking the down biomass structure than SP at low temperatures.

The decrease in the pore size and surface area after 550 °C could be due to the limited re<span class="Chemical">activity of the lignin-rich biochar. The decomposition of hemicellulose and cellulose usually occurs between 200 and 450 °C. Particularly, cellulose decomposition reactions dominate between 300 and 450 °C.[37] The thermal decomposition of corn stover is limited at low temperatures, and the large concentration of hemicellulose and cellulose limits the surface area and pore morphology. This can be seen when comparing HTC and SP at 240 °C, which highlights the impact of subcritical water in breaking down the biomass. Fragmentation becomes a dominating reaction at higher temperatures and is at its maximum around 600 °C.[37] The lignin polymeric structure has a higher kinetic threshold for decomposition that dominates at temperatures above 500 °C.[37] It is possible that the increase in surface area at 550 °C is primarily due to the decomposition of cellulose and hemicellulose forming high surface area biochar. As the temperature increases, the solid degrades into aromatic species and other hydrocarbons that escape into the gas phase where a portion can be condensed into the oil phase. There is a small spike from 600 to 650 °C, which could be due to polymerization and formation of some other products on the solid.[37] Generally, the surface area of the SP carbon samples decreases at higher temperatures, which may be attributed to the formation of large pores. Initially, the pore size of the samples is mesoporous, then decreases and increases again at elevated temperatures. Table S1 shows the morphological changes of carbon as a function of a change of duration at 550 °C. It appears the longer the dwell time, the higher the surface area. However, the increase in the surface area from 1 to 8 h is less than 5%. The limited change after 1 h confirms that the reaction reaches steady state in 1 h. The average pore size is plotted in Figure S1b as a function of temperature. Interestingly, the average pore size plot is an inverse of the surface area plot between 500 and 700 °C, Figure S1b. As the average pore size decreases, the total surface area increases, which would suggest that there is an increased number of smaller pores that causes the increases in the available surface area. At higher temperatures, we observe a decrease in surface area. We speculate that the decrease can be attributed to either pore collapse or an increase in the size of the pores, as can be seen for temperatures above 650 °C.

The visual appearance of the biochar from the SP remained consistently bl<span class="Chemical">ack for all dwell times and temperatures tested. The SP did not show much color change over the range as the lowest temperature, 300 °C, would be past the Maillard reaction’s upper limits. In Figure a–d, the biochar SEM images for select samples from SP can be seen in which the structure changed as the temperature increased. Notably, SP biochar samples still had the rigid structure of the fibrous corn stover structure. Nevertheless, Brunauer–Emmett–Teller (BET) analysis confirms that the structure does become more porous as the temperature increases. HTC and SP produced distinct biochar from each other, where the degradation of cellulose and hemicellulose played an important role. HTC led to lower surface areas and larger average pore sizes, while SP led to higher surface areas and smaller average pore sizes. The subcritical water of HTC and higher temperatures in SP leads to different reactions occurring and some were more prominent than others.

Figure 4

SEM images of SP biochar at different temperatures. (a) SP 400 °C 1 h (mag: 931×). (b) SP 500 °C 1 h (mag: 1.23k×). (c) SP 600 °C 1 h (mag: 934k×). (d) SP 600 °C 1 h (mag: 1.28k×).

AC from Biochar Derived from HTC and SP

The biochars from the HTC and SP methods were chemic<span class="Chemical">ally activated to produce AC for the adsorption of phenolic compounds. The properties of the ACs were characterized using X-ray photoelectron spectroscopy (XPS), XRD, Fourier transform infrared spectroscopy (FTIR), and SEM and compared to that of direct chemical activation of the corn stover. The results of XRD of select ACs show the formation of amorphous carbon (Figure ). The AC XRD patterns have two broad peaks from ∼20 to 30 and ∼40 to 50 2θ. These represent amorphous carbon peaks as one would generally find broad peaks ranging from 10 to 30 and 35 to 50 2θ for amorphous carbon composed of aromatic carbon sheets oriented in a considerably random fashion.[38] The XRD pattern for the direct activation of corn stover has a peak at 29.5°, which corresponds to silicate minerals in the sample. XPS was conducted to perform an elemental analysis of the biochar materials prior to and after activation with KOH. We determined that the O/C content of the HTC biochars and AC direct materials decreased after activation. However, the O/C content of the SP biochar remained the same. This confirms that the SP biochar is not as amenable as the HTC biochar for subsequent chemical activation by KOH.

Figure 5

XRD spectra of AC prepared from corn stover directly and SP and HTC biochars.

XRD spectra of AC prepared from corn stover directly and SP and <span class="Chemical">HTC biochars.

Tables and S2 contain the surface areas of the <span class="Chemical">AC samples using biochar precursors from the direct, HTC, and SP methods. Surprisingly, the highest surface areas were achieved with the direct and HTC biochar precursors. For example, the highest surface area from the activation of the HTC biochar sample prepared at 240 °C for 2 h was 1167 ± 164 m2/g. We activated milled corn stover with no prior treatment as a control and observed a surface area of 956 ± 39 m2/g. This surface area is around what can be found on many commercial ACs. The highest surface area from the AC SP series of experiments was from the biochar formed from the 300 °C 1 h run at 1008 ± 94 m2/g. The lowest achieved surface area was from 550 °C SP biochar at 376 ± 107 m2/g. Despite having higher starting surface areas compared to HTC biochars, the ACs formed from SP biochars have lower surface areas than those created from HTC biochars. Thus, the increased biochar surface area is inversely proportional to the AC surface area; this trend can be seen in Table S2. Figure S3 shows the N2 adsorption/desorption isotherms of the produced ACs. AC direct (Figure S3a) and AC HTC (Figure S3b) had the highest porosity, suggesting saturation of micropores and mesopores in the structure. The isotherm for AC SP (Figure S3c) indicates a lower adsorptive capacity with a microporous structure. The average pore size of all the ACs as seen in Table S2 was between 4 and 9 nm, indicating that there are mesopores in each sample. There were minimal differences in the AC direct, AC HTC, and AC SP average pore size. The most noticeable difference between the samples was surface area primarily contributed by the micropore region. AC direct and AC HTC had surface areas with a higher micropore area.

Table 1

Surface Area and X-ray Photoelectron Spectroscopy Elemental Analysis of the Biochar and the Associated ACs

		relative atomic concentration (%)
sample	surface area (m2/g)	O	N	C	S	O/C
CS	1.2 ± 0.13	24	1	75	0	0.32
SP500	5.0 ± 1.5	18	1	81	0	0.22
HTC240	6.9 ± 1.3	22	1	77	0	0.29
AC direct	956 ± 39	17	0	82	0	0.21
AC SP500	646 ± 19	18	1	81	0	0.22
AC HTC240	1167 ± 164	13	1	86	0	0.15

Figure a–d shows the SEM images of the AC samples for select samples. The <span class="Chemical">HTC AC images in Figure a–d have a stark difference in appearance and morphology than the SP biochar AC images in Figure c,d. The HTC biochar AC samples have significant surface changes with a clear breakdown of the rigid structure and courser appearance. This is especially apparent when compared to the original corn stover shown in Figure a. The SP AC images from the SP biochar still have a rigid structure with less visible change compared to HTC ACs. The differences observed from SEM in the AC derived from either the HTC or SP biochar could indicate the differences in the surface area and pore sizes.

Figure 6

SEM images of AC. (a) AC HTC 200 °C 1 h (mag: 1.14k×). (b) AC HTC 240 °C 2 h (mag: 1.14k×). (c) AC SP 400 °C 1 h (mag: 1.0k×). (d) AC SP 550 °C 1 h (mag: 1.0k×).

SEM images of AC. (a) <span class="Chemical">AC HTC 200 °C 1 h (mag: 1.14k×). (b) AC HTC 240 °C 2 h (mag: 1.14k×). (c) AC SP 400 °C 1 h (mag: 1.0k×). (d) AC SP 550 °C 1 h (mag: 1.0k×).

We surmise that the difference in the maximum surface area <span class="Chemical">achieved and the porosity for HTC biochar or SP biochar-derived AC is due to cellulose and lignin concentration. For instance, a study by Tiryaki et al. created AC from tomato leaves that had 10.9% cellulose and 24.8% lignin that produced a surface area of 305 m2/g.[39] In comparison, carbon with a higher cellulose content with a ratio of 26.2% cellulose and 36.5% lignin had a surface area of 839 m2/g. A study performed by Zhang et al. determined that as the lignin concentration increased, the surface area decreased.[40] It was theorized that the polysaccharide structure of cellulose allowed the formation of a mesoporous structure. The complex polymeric aromatic structure of lignin though contributed to the layered and microporous structure. We conclude that the SP biochars at high temperatures above 500 °C lack an appreciable amount of cellulose, corroborated by XRD, and thus are less prone to mesoporous structure formation. The hydroxyl groups in the cellulose and hemicellulose structure are reactive to the KOH chemical activation, reducing micropores. In contrast, lignin’s aromatic backbone is more predisposed to produce macropores and carbon sheets. Thus, the HTC samples rich in cellulose are more prone to micropore formation.

Adsorbent Properties of ACs for Vanillin

Characterization of Functional Groups

The adsorption characteristics of the <span class="Chemical">AC HTC, AC SP, and AC direct materials depend on the surface functional groups. FTIR analysis was used to analyze the surface functional groups of the formed biochars and the associated ACs. Figure a shows the comparative structures of biochar formed from HTC and SP. Both spectra show a broad band between 3150 and 3400 cm–1 attributed to the O–H stretching of the hydroxyl groups. The area between 3000 and 2800 cm–1 is attributed to the C–H stretching. The peak intensity for HTC 240 °C is greater, showing that there are more of these functional groups formed after hydrotreatment. The peaks between 1700 and 1650 cm–1 and at 1050 cm–1 can be assigned to the C–O stretching of the carboxyl groups. The peaks between 1100 and 1000 cm–1 refer to the C–OH and C–O stretch. These results indicate that the surface of the biochars is mostly oxygen-containing groups, including hydroxyl (−OH) and carboxylic (−COOH) functional groups. The ACs from all biochars formed from the direct, HTC, and SP had peaks indicative of C–O, C–H, and C–OH bonds as seen in Figure b. The lower intensities infer that the surface of the carbon has predominant hydroxyl and C–O groups, however, at a lower concentration than the biochar precursors. This is also corroborated by the O/C ratio measured by XPS in Table , showing a lower O/C ratio than the biochars. The additional peaks and stronger intensity for the HTC corn stover over the other samples show that it has more oxygen-containing groups, which could be attributed to the higher concentration of cellulose and hemicellulose.

Figure 7

FTIR analysis of the (a) biochar from either the HTC or SP and (b) AC.

FTIR analysis of the (a) biochar from either the <span class="Chemical">HTC or SP and (b) AC.

Adsorption Performance

The physical adsorption of <span class="Chemical">vanillin molecules depends on the functional groups on the surface of the AC. We compared the vanillin adsorption performance of the AC with respect to their biochar precursor characteristics and synthesis. Batch adsorption experiments of the ACs were carried out, and the results are presented in Figures and 9. The results show that the samples perform almost the same for concentrations between 50 and 200 mg/L (Figures and 9a). For example, 98% of vanillin can be removed within 60 min by using the AC prepared from the corn stover direct method and the HTC biochar. Comparatively, it takes the AC prepared from the SP biochar greater than 120 min to adsorb 98% of vanillin.

Figure 8

Vanillin adsorbate capacity normalized to surface area of the ACs as a function of (a) vanillin concentration, (b) dosage, and (c) time.

Figure 9

Vanillin removal as a percentage of ACs as a function of (a) time and (b) mass.

Vanillin <span class="Chemical">adsorbate capacity normalized to surface area of the ACs as a function of (a) vanillin concentration, (b) dosage, and (c) time.

Vanillin remov<span class="Chemical">al as a percentage of ACs as a function of (a) time and (b) mass.

The surface function<span class="Chemical">al groups, pore volume, and pore size of AC positively influence the adsorption rate and amount of vanillin adsorbed onto the AC surface. Therefore, the effect of the mass of the AC to the reaction medium was used to determine the effects of the removal of vanillin. The solution volume and concentration were kept constant at 50 mL of 50 mg/L and were used at room temperature with a 60 min contact time. As shown in Figures and 9b, at 20 mg loading, AC direct had the best removal of vanillin at ∼14% more removed than AC HTC and even greater for AC SP. While AC direct had slightly less surface area, it adsorbed more than AC HTC by a noticeable amount, which may have implications that surface area is not the only parameter that affects the adsorption of vanillin. Parameters like surface functional groups and pore structure may affect the adsorption of vanillin. Each sample adsorbed more vanillin when the loading increased from 20 to 35 mg, with AC direct and AC HTC having removed similar amounts. With further increased amounts, AC HTC and AC direct remained the same, and AC SP continued to increase in adsorption capabilities. At the 50 mg loading, over 98% is removed for AC direct and AC HTC, while only 58% is removed for AC SP. The surface and morphology of the AC SP are drastically different from the other two ACs. This can be seen in the normalized adsorption capacity, where even per m2 the adsorption of vanillin onto AC SP is less than that of the others. This is an indication that more than the surface area is involved in the removal of vanillin, which we speculate could be due to the number of adsorption sites, the pore structure, and/or the surface functional groups.

Surface function<span class="Chemical">al groups can affect the adsorptive properties of AC. The presence of dissolved oxygen on AC can increase the adsorptive capacity of phenolic compounds through oxidative coupling reactions.[41] From the FTIR data, the HTC biochar had more oxygen-containing surface functional groups than the SP biochar. However, when activated, distinguishing which has the most oxygen-containing surface functional groups becomes difficult. The results from the AC Direct and AC HTC are very similar, indicating similar surface functional groups, which we speculate is a reason for the higher adsorption of vanillin compared to AC SP.

The contact time between the <span class="Chemical">AC and vanillin is a critical parameter for the adsorption process; thus, contact time optimization was investigated. The effect of contact time on the adsorption of vanillin by the prepared ACs was examined at room temperature by using 50 mg of AC and 50 mL of 50 mg/L vanillin solution. The adsorption experiment was carried out for up to 2 h to determine the adequate adsorption time, and the result is presented in Figure c. The amount of the absorbed vanillin for AC HTC and AC direct increased from 0 to 30 min and plateaued afterward. The adsorption on SP, however, increased rapidly from 60 to 120 min. For AC HTC and AC direct, the results show that the adsorption sites were saturated after 30 min. The AC SP continued to adsorb as the time increased, indicating that the maximum adsorption had not been achieved. This shows a possibility that the adsorption rate for SP is slower when compared to that of the AC HTC and AC direct.

Conclusions

This paper compared the properties of biochar precursors and ACs that were chemic<span class="Chemical">ally activated either from corn stover using the direct method, HTC, or SP. XRD and BET analysis showed that HTC and SP biochar differed in their lignocellulosic composition after the reaction at elevated temperatures. The role of water in HTC plays an integral part in the decomposition of corn stover as it allows hydrolysis and other reactions to break down the biomass more efficiently than SP. SEM analysis showed that the HTC and SP biochars formed more pores as the temperature increased. FTIR spectroscopy analysis showed that HTC biochars had a higher density of oxygen-based surface functional groups than SP biochars. Additionally, the formed AC from corn stover and HTC biochar precursors was apparently more oxygen-rich.

The AC formed from the studied biochar precursors was then probed for the adsorption of <span class="Chemical">vanillin. The lignocellulosic composition of the biochar is influential on the surface area, pore size, pore structure, and available surface functional groups for the formed ACs. In terms of their adsorptive abilities, the AC direct and AC HTC had significantly better adsorption of vanillin than the AC SP. When normalized to the surface area, AC direct and AC HTC had improved performance compared to AC SP as a function of duration and the total AC amount used, which implies that other properties such as pore structure and surface functional groups are important. The AC direct and AC HTC performed better than AC SP, indicating the importance of the biochar pretreatment method to AC properties. Overall, AC was produced from a highly relevant agricultural residue, corn stover. The results suggest that the AC generated directly from corn stover had properties comparable to the absorbents made from HTC and SP biochars. The production of AC as adsorbents for phenolic compounds may not warrant the extra thermochemical step of biochar precursor synthesis from HTC and SP.

Materials and Methods

Hydrothermal Carbonization

A 300 mL Series 4561 Bench Parr Reactor was used for the <span class="Chemical">HTC reactions. Corn stover was milled to 1 mm. Deionized water and corn stover were added 8:1 by mass, unless noted otherwise, to the reactor and purged with nitrogen for 10 min. The reactor was heated and held at the desired dwell temperature for 1, 2, or 4 h. Once finished, the reactor was submerged in ice water to stop the reaction. The liquid and solid phases were separated using vacuum filtration. The solid phase was rinsed with 300 mL of DI water to remove most of the bio-oils, leaving the solid biochar behind. While a portion of the liquid phase is not water-soluble and may still be left in the porous solid structure, we believe that this is a negligible amount. The biochar was dried overnight in an oven at 105 °C before chemical activation with KOH.

Slow Pyrolysis

A Thermo Scientific Type 1315M Benchtop Muffle Furnace inside a <span class="Chemical">nitrogen glovebox was used for SP. Corn stover, milled to 1 mm, was placed inside a crucible and purged in a nitrogen environment. The sample was then placed in the muffle furnace and heated to the desired temperature at a ramp rate of 10 °C/min. The sample was held for 1, 4, or 8 h at the desired temperature and then allowed to cool to room temperature.

Chemical Activation of Biochar

The same muffle furnace setup for the SP experiments was used for the chemic<span class="Chemical">al and thermal activation of carbon. The biochar, either from HTC or SP, was combined with KOH in a 2:1 ratio of biochar to KOH by mass. DI water was added to the mixture and stirred for 1 h to ensure it was homogeneous. The mixture was then dried in an oven at 105 °C. The sample was transferred into the nitrogen environment muffle furnace, heated to 300 °C for 2 h at a ramp rate of 10 °C/min to remove moisture, and then further heated to 800 °C for 3 h with a ramp rate of 10 °C/min. Once cooled, the sample was washed with a 0.1 M HCl solution to neutralize any remaining KOH. The sample was vacuumed-filtered and washed with DI water until the filtrate was pH neutral. The sample, now AC, was dried in an oven overnight at 105 °C before characterization. For the direct chemical activation method, corn stover replaced the biochar in equal amounts and the rest of the procedure for chemical activation remained the same.

Surface Area Analysis

Surface an<span class="Chemical">alysis was conducted using the Micromeritics ASAP 2020 and ASAP 2020 Plus physisorption instruments to perform BET measurements. The sample was loaded and degassed for 4–11 h until an outgassing rate of less than 5 μmHg/min was achieved to ensure moisture and volatile contaminants were removed before analysis. N2 physisorption and five-point BET analysis were used to measure the surface area, pore volume, and pore size. The BET was calibrated with a silica–alumina reference material with a standard error of 2.5%. Replicates of most of the biochar and AC were performed to determine the intrinsic errors in the surface areas and pore sizes with 95% confidence levels.

Scanning Electron Microscopy

A TESCAN Vega3 SBH SEM was used to capture images of the various corn stover, biochar, and AC samples. A Thermo Fisher Scientific NNS450 was <span class="Chemical">also used to capture images of AC samples. Before imaging, the samples were placed under vacuum, purged with argon, and then sputter-coated with Au for 10 s to improve the clarity of the images. The images were taken between 900 and 1700 times magnification with a voltage of 5 kV.

XRD and XPS

A PANalytic<span class="Chemical">al Empyrean Series 2 XRD instrument was utilized to evaluate the carbon structures. The emission source was Cu Kα (1.54056 A wavelength) with a Ni beta filter. A zero-diffraction plate was employed to minimize the background peaks. XPS characterization was carried out using a Kratos AXIS ULTRA XPS system equipped with an Al X-ray source and a 165 mm mean radius electron energy hemispherical analyzer. Neutralizing was applied during the measurements to compensate for sample charging.

Fourier Transform Infrared Spectroscopy

Surface function<span class="Chemical">al groups of char and the ACs were investigated using an FTIR Spectrometer (Nicolet iS10, Thermo Scientific) equipped with a Diffuse Reflectance Infrared Fourier Transform Spectroscopy accessory (Praying Mantis, Harrick) and a High Temperature Reaction Chamber (HVC, Harrick). Gathered spectra was an average of 64 scans with 8 cm–1 resolution between the range of 650–4000 cm–1. A general procedure would be diluting a small amount of the sample with KBr. The ratio of sample to KBr was about 1:100 by mass. The mixture was ground into a fine powder with a pestle and mortar and loaded into the chamber. The sample was held at 100 °C under helium flow for 50 min before a spectrum was taken. A background spectrum consisting of only ground KBr was collected under the same heating conditions before FTIR experiments were done that day.

Batch Adsorption Study of Vanillin

The batch experiments of the vanillin adsorption studies using the <span class="Chemical">AC from SP, HTC, and the direct method were conducted at room temperature in a 150 mL beaker. For each run, 20–50 mg of the adsorbent was placed in a beaker containing 50 mL of a vanillin solution, which had a range of concentration between 50 and 200 mg/L. The suspension was stirred for a desired time, between 30 and 120 min, using a magnetic agitator. After agitation, the suspensions were gravity-filtered. The concentration of the filtrate was determined by using an Agilent Cary 60 UV–visible spectrophotometer. The absorbance wavelength was measured between 200 and 500 nm at a rate of 60 nm/min and a 0.50 nm interval.

The adsorbate cap<span class="Chemical">acity, normalized to the surface area, was calculated using the equation belowwhere C is the concentration of the adsorbate at time t in mg/L, C0 is the initial concentration of the adsorbate in mg/L, m is the mass of the AC in mg, V is the volume of the adsorbate solution in L, and the surface area of the adsorbent is in m2.