3.1 FTIR spectra of apricot shell cellulose and ASLC
The yield of ASLC higher than that of apricot shell cellulose, as listed in Table 1. In reviewing the literature, the characteristic absorption peaks of cellulose β-glycosidic bond at ~895 cm-1, ~1056 cm-1, and ~1162 cm-1.27 The characteristic absorption peaks of benzene ring skeleton of lignin at ~1430 cm-1, ~1508 cm-1, and ~1615 cm-1.28, 29 Fig. 1 (a) indicates that there are absorption peaks at wavelengths of 895 cm-1, 1056 cm-1, and 1162 cm-1, suggesting the presence of cellulose β-glycosidic bond in apricot shell cellulose. The stretching vibration of the -CH3/-CH2 at ~1461 cm-1. The absorption peaks of syringyl at ~1322 cm-1 and ~1112 cm-1. Additionally, the absorption peak at wavelength 1270 cm-1 or 1215 cm-1 indicates the presence of guaiacol.28, 29 As a result, -CH3/-CH2, benzene ring skeleton, syringyl, and guaiacol were existed in apricot shell lignin, as shown in Fig. 1 (b). Notably, Fig. 1 (b) equally indicates that apricot shell lignin has a minor amount of cellulose (apricot shell lignin was doped with cellulose), which was due to the limitation of the experiment. Thus, apricot shell lignin was doped with cellulose was denoted as ASLC.
3.2 Thermogravimetric analysis
To further investigate the pyrolysis characteristics of apricot shell cellulose and ASLC, the weight loss (TG) and the rate of weight loss (DTG) upon the heating rate of 10 ℃/min were analyzed, as exhibited in Fig. 2. Fig. 2 (a) shows that the pyrolysis processes of the apricot shell, apricot shell cellulose, and ASLC were composed of dry dehydration stage, pyrolysis stage, and pyrolysis-carbonization stages. The first stage was a slow weight loss stage, while the second and third stages were rapid devolatilization stages. In dry dehydration stage, apricot shell cellulose possesses the largest weight loss, followed by apricot shell and ASLC. In the two hind stages, apricot shell, apricot shell cellulose, and ASLC lose a lot of weight and formed significant weight-loss peaks. Different from the apricot shell and ASLC, apricot shell cellulose contains only one significant weight-loss peak. Similar results were also found for cellulose model components.30 Besides, the two significant weight-loss peaks formed during the pyrolysis of ASLC were similar to that of the lignin model components.30, 31 Notably, for the apricot shell, apricot shell cellulose, and ASLC, the pyrolysis-carbonization processes were completed at ~ 550 ℃. Consequently, in preparation for activated carbon, considering the property and cost, the activation temperature was 550 ℃. Furthermore, the pyrolysis residues of the apricot shell, apricot shell cellulose, and ASLC were 24.49 mass%, 16.96 mass%, and 33.13 mass%, respectively. Compared with the yield of the apricot shell and apricot shell cellulose pyrolysis residues, the yield of ASLC pyrolysis residue was higher. Accordingly, in the following experiments, the ASLC was selected if not mentioned as otherwise.
3.3 Characterization of AAC
3.3.1 Textural properties of AAC
H3PO4 and KOH are generally used as activators for the preparation of activated carbon.32, 33 As we all know, N2 adsorption and desorption isotherms at low relative pressure indicate micropores are filled; in other words, there are a host of micropores. At moderate relative pressure, isotherms with a hysteresis loop demonstrate the presence of mesopores. The steep rise of isotherms at high relative pressure exhibits capillary condensation in the macropores; that is to say that the existence of macropores. Fig. 3 indicates that the AAC has hierarchical structures of porosity. At low relative pressure, the N2 adsorption/desorption isotherm of AAC-K-2 is the steepest, followed by AAC-K-1, AAC-P-3, and AAC-P-4. Consequently, the order of the micropore specific pore volume (Vmicro) of AAC was determined: AAC-K-2 > AAC-K-1 > AAC-P-3 > AAC-P-4. Inversely, the N2 adsorption/desorption isotherm of AAC-K-2 was the gentlest upon moderate relative pressure, followed by AAC-K-1, AAC-P-3, and AAC-P-4, which means AAC-K-2 has the lowest mesopore specific pore volume (Vmeso) and AAC-P-4 has the highest Vmeso. On the other hand, the N2 adsorption and desorption isotherms of AAC-K-2, AAC-K-1, AAC-P-3, and AAC-P-4 display different degrees of an upward trend at high relative pressure, which indicates that AAC-K-2, AAC-K-1, AAC-P-3, and AAC-P-4 have macroporous structures. These results indicated the hierarchical micro/meso/macroporous nature of AAC. Moreover, the textural properties of AAC can be obtained by analyzing the adsorption and desorption isotherms of N2, as shown in Table 2.
Compared to AAC-K-2 (2136.56 m2/g), the specific surface areas of AAC-K-1, AAC-P-3, AAC-P-4 decreased to 1669.94 m2/g, 1475.16 m2/g, and 1370.35 m2/g, respectively. Montane et al.32 used H3PO4 as an activator to synthesize lignin-based activated carbon with a specific surface area of ~1000 m2/g. Zhang et al.34 indicated the specific surface area of KOH-assisted synthesis of activated carbon was 1672 m2/g. In other words, the specific surface area of lignin-based activated carbon lower than that of AAC-P-3 and AAC-P-4, and the specific surface area of KOH-assisted synthesis of activated carbon lower than that of AAC-K-2. Moreover, the Vmicro/Vmeso values of AACs were over 45.00%, especially AAC-K-2, its Vmicro/Vmeso value was as high as 375.27%, these results indicate that AACs with well-developed micropores. Compared with the AAC-P-3 and AAC-P-4, the average pore sizes of AAC-K-1 and AAC-K-2 were narrower. These results were consistent with the conclusion that the micropore specific pore volumes of AAC-K-1 and AAC-K-2 were higher. In this work, the pore-wall thickness was calculated by the following formula: Pore-wall thickness = 2/(ρSBET), where ρ = 2.2 g/cm3 represents the density of carbon wall equated to the density of graphite.35, 36 The order of the pore-wall thickness of AAC was established: AAC-P-4 > AAC-P-3 > AAC-K-1 > AAC-K-2, as listed in Table 2.
To further understand the pore structures of AAC, the pore size distribution is presented in Fig. 4. Unlike AAC-P-3 and AAC-P-4, AAC-K-1 and AAC-K-2 have intensely abundant micropores, especially micropores with a pore size of ~ 0.75 nm. Furthermore, the micropores of AAC-P-3 and AAC-P-4 were primary concentrated in the range of 1-2 nm, and the micropores below 1nm were few. For mesopores within the range of 2-5 nm, AAC-K-1 and AAC-K-2 have only a minor amount of distribution, while AAC-P-3 and AAC-P-4 possess a majority of distribution. Interestingly, the previous study has shown that porous materials rich in 2-5 nm are more beneficial to the adsorption of dioxin;37 from this, we can infer that AAC equally possesses a promising application in removing dioxins. On the other hand, the activated carbon prepared by H3PO4 activation was rich in mesopores, while the activated carbon prepared by KOH activation was almost all micropores, with fewer mesopores. Consequently, in actual production, different activation strategies can be selected according to activated carbon requirements. According to literature, porous materials possess high specific surface area and proper pore size distribution hold a promising application as electrode materials for capacitors.38, 39 Sudhan et al.40 prepared a rice straw-derived activated carbon that possesses a specific surface area of 1007 m2/g and employed it as an electrode material for supercapacitors. Notably, the specific surface areas of AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 are higher than 1007 m2/g, especially AAC-K-1 and AAC-K-2. As a result, AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 also have a certain value in the application of electrode materials for capacitors.
3.3.2 Micro-morphology of AAC
To observe the natural hierarchical porous structure of AAC more intuitively, SEM tests using AAC-P-3 and AAC-K-2 as examples were performed. Fig. 5(a) indicates that the surface of the AAC-P-3 is similar to the honeycomb, which means AAC-P-3 has an enriched pore structure. It is worth noting that the 3D architecture with hierarchical structures is presented in Fig. 5(b). Moreover, Fig. 5 (b) shows that the porosity is composed of irregularly oriented narrow pores. These micro-morphologies are consistent with and also more intuitively than the results of N2 adsorption and desorption isotherms.
3.4 Electrochemical performance of AAC
Fig. 6 (a) demonstrates GCD curves of AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 at the current density was 0.5 A/g. As we all know, if the GCD curve is approximately an isosceles triangle, it indicates an ideal double-layer capacitor behavior.41 AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 have excellent double-layer capacitor behavior. The specific capacitances of AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 were 169.14 F/g, 159.98 F/g, 188.50. F/g, and 236.00 F/g, respectively. Consequently, considering the specific capacitance and different preparation processes, AAC-P-3 and AAC-K-2 were choose for the CV test.
The capacitance behavior shown in these CV curves has belike quasi-rectangular shapes accompany by some humps, resulting from the combined effects of electrical double-layer capacitance and pseudocapacitance.42, 43 The CV curves of AAC-P-3 and AAC-K-2 upon various scan rates are shown in Fig. 6 (b) and (c), respectively. CV curve is similar to a rectangle, which means the electrolyte can diffuse into the micropores within a sufficient time. In other words, the charge and discharge reversibility are better. Fig. 6 (b) and (c) indicate the AAC-P-3 and AAC-K-2 possess a better charge/discharge reversibility at a low scan rate (5-20 mv/s). Additionally, the resistance of micropores is higher than that of mesopores and macropores. Which results in the electrolyte ions cannot quickly enter the micropores upon a high scan rate. Hence, with an increase in scan rate (>20 mv/s), the degree of CV curve deviation from the rectangle gradually increased. Nonetheless, none of them exhibited redox peaks, demonstrating that the AAC-P-3 and AAC-K-2 have excellent electrochemical stability.
On the other hand, the specific capacitances of AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 are listed in Table 3. Meanwhile, the plot of specific capacitance variation of AAC is given in Fig. 7. In brief, the specific capacitances of AAC-P-3 and AAC-P-4 were comparable to that of the reported commercially available specific capacitances of supercapacitors. However, the specific capacitances of AAC-K-1 and AAC-K-2 were over the specific capacitances of commercially available supercapacitors.44 With an increase in the current density, the specific capacitances of AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 gradually decreased, which were similar to the results of previous studies.45-49 The order of the specific capacitances of AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 was established at the same current density: AAC-K-2 > AAC-K-1 > AAC-P-3 > AAC-P-4. Notably, at the current density was 0.50 A/g, the specific capacitance of AAC-P-3 was higher than that of rice straw-derived activated carbon and jute stick-derived activated carbon.40, 50 Moreover, the capacitance retention rates of AACs at 10.00 A/g exceed 67.39%.
EIS was employed to further understand the electrochemical performances of AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2, as shown in Fig. 6 (d). As we all know, the low-frequency region and high-frequency region indicate the charge diffusion-limited process and the resistance of charge transfer at the interface between the electrolyte and active materials, respectively.51, 52 Compared with the Rct values of AAC-P-3, AAC-P-4, and AAC-K-1 (~ 1.9 Ω), the Rct value of AAC-K-2 was higher (~ 2.25Ω). In addition, the high slope in the low-frequency region shows that AAC-P-3, AAC-P-4, AAC-K-1, and AAC-K-2 have pleasurable capacitance performance.
3.5 Textural properties/specific capacitance relationship
For porous materials, the textural property directly affects the performance of porous materials.53, 54 For instance, micropores provide a large surface area for ion adsorption, mesopores serve as transport channels into the micropores, and macropores serve as ion-buffering reservoirs, which can reduce the diffusion distance of ions.55 Furthermore, some studies have exhibited the specific surface area affects specific capacitance.56-58 It is worth noting that the specific surface area is only one of the textural properties. As a result, it is necessary to systematically understand the influence of textural properties (e.g. specific surface area, specific pore volume, pore-wall thickness) on specific capacitance.
In this work, linear fitting was used to evaluate the relationship between textural properties and specific capacitance. Thus, the data in Table 2 was associated with the data in Table 3, as shown in Fig. 8. The average values of the correlation coefficients (AR2) in Fig. 8 (a), (b), (c), (d), (e), (f), (g) were calculated respectively, which were 0.983, 0.312, 0.956, 0.808, 0.978, 0.837, 0.975. Among these, the AR2 (0.983) between specific surface area and specific capacitance was highest, followed by the AR2 (0.978) between Vmicro/Vmeso and specific capacitance, the AR2 (0.975) between pore-wall thickness and specific capacitance; from this, we can infer that the specific surface area plays a key role in specific capacitance. A similar phenomenon has been observed in previous studies.56-58 It is worth noting that 0.983, 0.978, 0.975, 0.956 are very close and are all higher than 0.95. Consequently, the specific capacitances of the AACs depend not only the specific surface area, but also on the Vmicro/Vmeso, pore-wall thickness, and Vmicro. On the other hand, the AR2 (0.312) between Vt and specific capacitance was the lowest; that is to say that Vt and the specific capacitance were not linearly positively correlated. Based on the AR2, we can infer that the specific surface area, Vmicro/Vmeso, pore-wall thickness, and Vmicro can be used as a universal predictor of specific capacitance upon employed KOH solution as the electrolyte.