Abstract
Wood-based activated carbon microspheres (WACMs) were successfully prepared from Chinese fir. However, the weak thermal stability of the precursors resulted in melting at high activation temperatures, which led to a lower yield and poor adsorption properties. To promote the thermal stability of the precursors in the process without chemicals, this paper focused on the effects of the pre-oxidation before activation on the thermal stability and chemical structure of the precursors and the pore structure of the activated samples. The results showed that during the pre-oxidation process, the O/C ratio increased noticeably and the intensity of oxygen-containing functional groups, such as carbonyl, carboxyl, and ester groups, were enhanced, which contributed to the improvement of the thermal stability. Moreover, the etching effect of oxygen on the structure of the precursors was verified when the porous structure was enriched by activation. The activated sample pre-treated at pre-oxidation temperature 250 °C had the maximum micropore specific surface area and pore volume. As the temperature was raised to 280 °C, the mesopore specific surface area and pore volume increased, which indicated that the optimization of the pre-oxidation parameters contributed to the formation of the mesopore structure.
Download PDF
Full Article
Preparation and Characterization of Wood-based Pre-oxidized Precursors and Activated Carbon Microspheres
Junbo Shang,a,b Jian Lin,a,b,* and Guangjie Zhao a
Wood-based activated carbon microspheres (WACMs) were successfully prepared from Chinese fir. However, the weak thermal stability of the precursors resulted in melting at high activation temperatures, which led to a lower yield and poor adsorption properties. To promote the thermal stability of the precursors in the process without chemicals, this paper focused on the effects of the pre-oxidation before activation on the thermal stability and chemical structure of the precursors and the pore structure of the activated samples. The results showed that during the pre-oxidation process, the O/C ratio increased noticeably and the intensity of oxygen-containing functional groups, such as carbonyl, carboxyl, and ester groups, were enhanced, which contributed to the improvement of the thermal stability. Moreover, the etching effect of oxygen on the structure of the precursors was verified when the porous structure was enriched by activation. The activated sample pre-treated at pre-oxidation temperature 250 °C had the maximum micropore specific surface area and pore volume. As the temperature was raised to 280 °C, the mesopore specific surface area and pore volume increased, which indicated that the optimization of the pre-oxidation parameters contributed to the formation of the mesopore structure.
Keywords: Liquefied wood; Thermal stability; Pre-oxidation; Pore structure; Activated carbon microspheres
Contact information: a: Beijing Key Laboratory of Wood Science and Engineering, Beijing Forestry University, Beijing 100083, China; b: MOE Key Laboratory of Wooden Material Science and Application, Beijing Forestry University, Beijing 100083, China; *Corresponding author: linjian0702@bjfu.edu.cn
INTRODUCTION
Activated carbon microspheres are widely used in the fields of supercapacitors (Wang et al. 2014), adsorption (Martín et al. 2011; Wickramaratne and Jaroniec 2013), medicine (Ye et al. 2008), heavy metal adsorption (Chen et al. 2011), and others, because of its advantages, such as its smooth surface, high packing density, high mechanical strength, good uniformity, rolling property, low flow resistance, etc. (Shen et al. 2008). Most raw materials used for activated carbon microspheres are prepared from fossil resources, such as pitch, coal, phenolic resin, etc. Environmental concerns have prompted researchers to seek alternative raw materials for the preparation of activated carbons from biomass materials, such as wood (Liu et al. 2017), lignin (Li et al. 2016), and hemp (Guo et al. 2016).
Sawdust from wood processing is a readily available biomass resource for value-added utilization. As one of the utilization methods, wood liquefaction and its products can be used to prepare fuels, adhesives, foams, molded products and so on (Pan et al. 2009). Moreover, wood-based carbon fibers have been successfully prepared from Chinese fir from liquefied wood (Ma and Zhao 2011; Shang et al. 2015), as well as wood-based activated carbon fibers (Huang et al. 2015). For some specific applications, such as fluidized beds and water treatment devices, the advantages of spherical activated carbons are obvious (Shen et al. 2008). Moreover, further study on the structure of wood-based carbon materials has indicated that liquefied wood is more suitable for preparing porous structural carbons (Liu et al. 2015). However, the most difficult problem has been the low activation yield and small porous structure, which is a result of the low thermal stability of the precursors from liquefied wood (Lin et al. 2013).
Pre-oxidation is a requisite process for preparing carbon fibers before carbonization, with no chemicals added. The process is widely used in the preparation of pitch-based carbon fibers. During the pre-oxidation process, a series of complicated reactions, such as cyclization, cross-linking, dehydrogenation, and oxidation of the molecular chains, occur under the conditions of air or oxygen between 200 °C to 400 °C (He 2010). A partial condensation reaction has been found to occur along with removal reactions during pre-oxidation, which leads to more cross-linked reactions and a denser structure (Sun et al. 2017). Moreover, after the pre-oxidation treatment, the C1s peak of activated carbons shifts to a higher binding energy, the number of oxygen-containing functional groups increases, and more bonds appear between oxygen and the carbon matrix, which leads to high heat resistance (Liu et al. 2011). However, decomposition and crosslinking reactions both occur during the pre-oxidation treatment, mainly because the improper pre-oxidation parameters result in an etching effect of oxygen on the pore walls of the precursors (Liu et al. 2010).
To improve the thermal stability of the precursors, wood-based activated carbon microspheres were prepared with the pre-oxidation treatment in air at different pre-oxidation temperatures, which was followed with water steam activation in an inert atmosphere. The morphology, chemical composition of the precursors, and the pore structure of the activated samples were analyzed to reveal the relationship between pre-oxidation and the pore structure of the activated carbon microspheres.
EXPERIMENTAL
Materials
Chinese fir (Cunninghamia lanceolata) wood was obtained from Fujian province, China. Chemicals such as phenol, phosphoric acid, hexamethylenetetramine (HTMA), hydrochloric acid, and polyvinyl alcohol (PVA) were purchased from Lanyi Chemical Products Corp. (Beijing, China). All of the chemicals mentioned above were used directly without any further purification.
Methods
Preparation of the precursors
A mixture of 25 g of dried Chinese fir powder (20 mesh to 80 mesh) and 100 g of phenol was added to a bottle flask, and 8 g of phosphoric acid (8.0% based on the weight of the phenol) were then added. The flask was immersed in an oil bath, heated rapidly from room temperature to 160 °C, and incubated for 2.5 h. The flask was cooled to room temperature with cool water, and the resultant mixture was filtered through a glass crucible filter (G2, pore diameter of 10 μm to 15 μm) to separate the filtrate from the residue. Subsequently, the filtered residue was put into a solution that was pre-heated to 65 °C, along with 6% PVA (based on the weight of the liquefied wood), and stirred at a speed of 400 rpm for 2 h. The liquefied wood microspheres (LWCMs) were obtained by freeze drying and were then immersed in a solution of hydrochloric acid (4 M) and 5 g/L HTMA, which was incubated at 85 °C for 2 h. After the treatment, the LWCM was washed with distilled water and dried in a vacuum oven (degree of vacuum = 0.1 MPa) at 45 °C.
The wood-based pre-oxidized microspheres (LWOMs) were prepared by heating the LWCM to 190 °C, 220 °C, 250 °C, 280 °C, and 310 °C at a heating rate of 1 °C/min, and were incubated for 1 h under an air atmosphere. Then, the temperature in the furnace was gradually decreased to room temperature. The samples were marked as LWOM-Tx, where x is the pre-oxidation temperature.
Preparation of the wood-based activated carbon microspheres
The WACMC was prepared using the LWCM, which was first carbonized at 600 °C at a heating rate of 4 °C/min under N2 and incubated for 1 h. Then, the temperature was increased from 600 °C to 800 °C at a heating rate of 4 °C/min and they were incubated for 1 h under N2 and water steam.
The LWOMs were placed in an activation furnace and heated to 800 °C under N2 (flow rate = 100 mL/min) at a heating rate of 4 °C/min. Then, water vapor was introduced at a flow rate of 3.12 g/min and the LWOMs were incubated for 1 h. After the temperature decreased to room temperature, the wood-based activated carbon microspheres pre-treated with pre-oxidation (WACMOs) were obtained and marked as WACMO-Tx, where x is the pre-oxidation temperature.
Characterization
Thermo-gravimetric analysis
The weight loss behaviors and decomposition temperature of the samples were determined by using a thermo-gravimetric analysis (TGA) device (TGA STA449F3, Netzsch Corp., Selb, Germany) from room temperature to 800 °C at a heating rate of 10 °C/min in high-purity argon (flow rate = 60 mL/min).
Fourier transform infrared spectroscopy
Fourier transform infrared (FTIR) spectroscopy of the samples was performed using a FTIR spectrophotometer (Tensor 27, Bruker Corp., Karlsruhe, Germany). The samples were dried before they were mixed with water-removed KBr crystals. This mixture was put in an agate mortar and ground at a weight ratio of 1:100 (sample to KBr crystals). A few samples were pressed on a tablet press to make a transparent sheet for taking measurements. The survey resolution was 4 cm-1 with 32 scans and a scanning range of 4000 cm-1 to 500 cm-1.
Elemental analysis
Elemental analysis of the samples was performed using an elemental analyzer (FLASH EA1112, Thermo Fisher Scientific Inc., Waltham, USA). The carrier gas was helium, and the applied decomposition temperature of the carbon, hydrogen, and nitrogen was 950 °C.
Scanning electron microscopy
The samples were coated with gold and the morphologies of the samples were observed with scanning electron microscopy (SEM) (S-3400N, Hitachi Corp., Tokyo, Japan) at magnification factors of 1500x to 2000x and an accelerating voltage of 10 kV.
X-ray photoelectron spectroscopy
To determine the number and type of functional groups present on the surface of the samples, X-ray photoelectron spectroscopy (XPS) measurements were performed on a spectrophotometer (ESCALAB 250Xi, Thermo Fisher Scientific Inc.) with a monochromated Al Kα X-ray source (hv = 1486.6 eV) and a power of 420 W. The survey scans were collected from the binding energy range of 0 eV to 1350 eV. A nonlinear, least squares regression analysis program (XPSPEAK software, Version 4.1., Informer Technologies, Inc., Hong Kong, China) was used for the XPS spectral deconvolution.
Specific surface area and pore structure analysis
The N2 adsorption-desorption isotherms were measured at 77 K with a surface area and porosity analyzer (Autosorb-IQ, Quantachrome Instruments Corp., Boynton Beach, FL, USA). The samples were out-gassed at 300 °C for 3 h before taking measurements to remove any moisture or volatile compounds within the pores of the samples. The Brunauer-Emmett-Teller (BET) model was used to calculate the surface area (SBET) of the N2 adsorption isotherms over the relative pressure (P/P0) range of 0.05 to 0.3 (Okada et al. 2003). The adsorption total pore volumes at relative pressures of 0.1 and 0.995 are equal to the volume of the micropores and total volume of the micropores and mesopores (Vtot), respectively (Nabais et al. 2004; Azargohar and Dalai 2008). The micropore (Smicro) and mesopore areas (Smeso) were calculated according to the t-plot and Barrett-Joyner-Halenda methods, respectively. The pore size distribution was calculated using the density functional theory method (Lastoskie et al. 1993).
RESULTS AND DISCUSSION
Thermal Stability of the LWCM and LWOMs
The thermo-gravimetric (TG) and differential thermo-gravimetric (DTG) curves of the LWCM and LWOMs prepared at different pre-oxidation temperatures are shown in Fig. 1. Figure 1a shows that the pyrolysis weight loss rate of all of the samples during the first stage from room temperature to 100 °C was less than 5%, which was mainly because of the evaporation of water. From 100 °C to 250 °C there was a pyrolysis reaction that occurred in the LWCM, but there was no obvious weight loss during this stage, which indicated that there were few pyrolysis reactions that occurred in the LWOMs. The fastest pyrolysis weight loss reaction occurred from 300 °C to 700 °C, and the by-products, such as low molecular weight phenolic substances (CH4, CO, and CO2), combusted or evaporated as the temperature increased (Zhang and Shen 2009). The end of the pyrolysis stage occurred from 700 °C to 800 °C. During this stage, the weight loss rate decreased gradually and only a few small molecular substances continued decomposing.
Although the variation in the TG curves between the different pre-oxidation temperatures was not obvious, the promotion of pre-oxidation to reduce the pyrolysis rate was distinctive, and the minimum pyrolysis residual rate of all of the samples prepared by pre-oxidation was 48.8%, which was greater than that of the LWCM. Therefore, even there was small pyrolysis loss under the conditions of air and heat, the effect of the pre-oxidation on improving the thermal stability of the precursors was more obvious. Meanwhile, the subsequent activation process was simplified.
Fig. 1. TG (a) and DTG (b) curves of the LWCM and LWOMs prepared at different pre-oxidation temperatures
Figure 1b shows that the differences in the maximum pyrolysis reaction rate between the LWCM and LWOMs during the pyrolysis process were obvious, especially when the pyrolysis temperature was higher than 100 °C. From 100 °C to 200 °C, the DTG curve of the LWCM fluctuated more than those of the LWOMs, and the maximum pyrolysis reaction rate of the LWCM was more concentrated. Moreover, as the pre-oxidation temperature increased, the peak pyrolysis temperature shifted to a higher temperature. The maximum peak temperature appeared at 527 °C with LWOM-T310, which was higher than that of the LWCM. Therefore, the DTG curves illustrated that the molecular crosslinking degree of the LWOMs was enhanced and the thermal resistance became stronger as the pre-oxidation temperature increased.
Fig. 2. SEM images of the LWCM (a) and LWOM-T250 (b) at 500x magnification, WACMO-T250 (c) at 250x magnification, and WACMO-T250 (d) at 500x magnification
Figure 2 depicts the morphologies of the LWCM, LWOM-T250, and WACMO-T250. The chemical curing and pre-oxidation treatment had few effects on the spherical shape of the microspheres, but the chemical curing process formed a dense solidified layer on the surface of the LWCM (Ma and Zhao 2010). After the pre-oxidation treatment, the surface of the LWOM-T250 became smoother and denser than before. Because the pre-oxidation treatment resulted in a dense surface layer with few pores, further processes, such as carbonization and activation, were necessary to open the inner and outer pore connections. During the activation process, the internal pore structure was enriched and the connections between the surface and inner pores were opened. More pores were observed on the surface of the activated samples, mainly because of the etching effects of the water steam and high temperature used during activation.
Cross-linking Reaction during Pre-oxidation
Figure 3a shows the FTIR spectra of the LWCM and LWOMs prepared at different pre-oxidation temperatures. The adsorption peak from 3400 cm-1 to 3430 cm-1 was attributed to the -OH vibration frequencies of hydroxyl groups (Silverstein et al. 2005), which became wider and weaker as the pre-oxidation temperature increased. The adsorption peaks at 1352 cm-1 and 1608 cm-1, that were attributed to benzene rings, were diminished after pre-oxidation. Meanwhile, the intensity of the peaks at 1454 cm-1, which were attributed to the C-H stretching of methylene groups, and the peaks at 830 cm-1 and 756 cm-1, which were attributed to C-H vibration, showed a decreasing trend as well. This manifested when the hydrocarbon functional groups were combusted or oxidized during the pre-oxidation process. In contrast, the adsorption peak at 1703 cm-1, which was assigned to the keto, carbonyl, and ester groups, was enhanced noticeably, and the peak moved to higher frequencies as the pre-oxidation temperature increased. This indicated that a large amount of oxygen-containing functional groups was formed during the pre-oxidation process. These results were consistent with the variations in the element contents (Fig. 3b).
Fig. 3. FTIR spectra (a) and element contents (b) of the LWCM and LWOMs prepared at different pre-oxidation temperatures
The variations in the element contents for the LWCM and LWOMs prepared at different pre-oxidation temperatures are shown in Fig. 3b. Before the pre-oxidation treatment, the relative elemental carbon content was higher than those in the LWOMs, and the oxygen content and O/C ratio were lower than those in the LWOMs. As the pre-oxidation temperature increased, the carbon and hydrogen contents decreased gradually and the oxygen content increased. As a result, the O/C ratio increased noticeably from 39.7% in the LWCM to 57.7% in LWOM-T310, which indicated that the amount of oxygen-containing functional groups formed during the pre-oxidation process and the amount of aliphatic hydrocarbon functional groups containing carbon and hydrogen were both reduced. However, the oxygen content increase rate from 280 °C and 310 °C was small, which was mainly because the combustion rate of the oxygen-containing functional groups was exceeded by the formation rate under the high temperature condition (Liu et al. 2010).
Variation in the Chemical Compositions for the Different Processes
The XPS spectra of the LWCM, LWOM-T250, WACMC, and WACMO-T250 are depicted in Fig. 4a.
Fig. 4. XPS spectra of the LWCM, LWOM-T250, WACMC, WACMO-T250 (a); peak fitting of the XPS C1s peaks for WACMO-T250 (b)
Table 1 summarizes the elemental compositions on the surface of all of the samples. Carbon was the most abundant constituent in all of the samples and clearly increased after activation. The nitrogen content was the lowest mainly because there were no nitrogen-containing substances that had been introduced during the preparation process, and nitrogen escaped as by-products during activation (Mangun et al. 2001). In contrast, the oxygen content increased after the pre-oxidation treatment, mainly because oxygen combined with the LWOMs in the form of bound oxygen, which increased the amount of oxygen-containing functional groups and the oxygen-carbon ratio. Under the conditions of a high activation temperature and water steam, the thermal decomposition of oxygen-containing carbon functional groups released a large amount of carbon and oxygen compounds, which promoted the enrichment of carbon and the pyrolysis of oxygen-containing functional groups.
The C1s spectra of the LWCM, LWOM-T250, WACMC, and WACMO-T250 were similar. Therefore, only the WACMO-T250 spectrum is depicted in Fig. 4b. For all of the samples, the C1s signals showed an asymmetric tailing, which was attributed to the intrinsic asymmetry of the graphite peaks and contribution of oxygen surface complexes. Deconvolution of the four individual component peaks of the C1s spectra was performed and indicated the presence of graphitic carbon (284.6 eV to 284.8 eV); carbon present in alcohol, ether, or C=N groups (285.4 eV to 286.2 eV); carbonyl or quinine groups (286.4 eV to 287.1 eV); and carboxyl, lactone, or ester groups (288.1 eV to 288.9 eV) (Chiang et al. 2007).
Table 1. Surface Elemental Composition of the LWCM, LWOM-T250, WACMC, and WACMO-T250
Table 2 summarizes the percentages of the graphitic and functional carbon atoms in the samples. Comparison of the spectra was done between the LWCM, LWOM-T250, WACMC, and WACMO-T250. After the pre-oxidation treatment, the relative amount of graphitic carbon and C/O functional groups, such as C-O and carbonyl, were reduced, except for the carboxyl functional groups. Because of the combination with oxygen, the percentage of carboxyl functional groups increased remarkably.
Table 2. Results of the Fits of the C1s regions
After activation, the percentage of the carbonate groups both in the WACMC and WACMO-T250 increased noticeably. In contrast, the functional groups C=O and -COOH were reduced greatly, which was mainly because the increasing amounts of graphitic carbons reacted with steam water molecules to generate additional functional groups containing C-O (Lin and Zhao 2016). Furthermore, the carbonyl and carboxyl functional groups were consumed, which resulted in volatile by-products, such as CO2, CO, and H2O (Ma et al. 2017).
Pore Structure of the WACMC and WACMOs
The N2 adsorption-desorption isotherms and pore size distribution of the WACMC and WACMOs pre-treated at various pre-oxidation temperatures are shown in Fig. 5. As the pre-oxidation temperature increased from 190 °C to 310 °C, the greatest N2 adsorption capacity was 350 cm3/g, which was recorded with the WACMO-T280. The isothermal adsorption/desorption curves showed the same tendency for the different pre-oxidation temperatures. The isotherms increased in the low-pressure zone rapidly, which indicated that the samples had a typical microporous structure. As the pre-oxidation temperature increased, the isotherms of the WACMOs showed a more pronounced knee-like shape and narrower hysteresis loops than the WACMC, which illustrated that there was part of an irregular mesopore structure with a smaller size that existed in the WACMOs (Jin and Huang 2015).
Fig. 5. N2 adsorption-desorption isotherms at 77 K of the WACMC and WACMOs
The pore size distributions of the WACMC and WACMOs pre-treated at various pre-oxidation temperatures are shown in Fig. 6. Figures 6a and 6b show the micropore and mesopore areas, respectively. The presence of both micropores and mesopores was indicated by the sample curves. The same trend of the pore size distribution showed that the different temperatures had more of an effect on the mesopore size distribution than on the micropore size distribution. The pore size distribution of the micropore area was mainly concentrated in the range of 0.5 nm to 1.4 nm. However, in the mesopore region (2 nm to 30 nm), the pore size distribution was mainly concentrated from 2 nm to 10 nm. Although WACMO-T250 appeared to have a higher integral pore volume when the pore size was greater than 6 nm, the integral pore volume in the range of 2 nm to 6 nm was less than that of WACMO-T280. Consequently, it was concluded that the appropriate pre-oxidation temperature enriched the mesoporous structure of the samples.