Introduction

With increasing CO2 emissions, global warming is accelerated, accompanied by abnormal climate changes1,2. According to the 2015 Paris Climate Conference, the global agreement for the reduction of CO2 emissions was strengthened3. Although most of the emitted CO2 has been produced from the combustion of fossil fuels, it is inevitable to use fossil fuels as major energy sources because of their availabilities and economic efficiencies, resulting in high demand4. For decreasing CO2 emissions, carbon dioxide capture and sequestration (CCS) is considered as one of the promising technologies. In the CCS technology, CO2 capture accounts for the largest portion of the total cost; thus, it is imperative to develop efficient methods for capturing CO25. Various approaches have been applied for capturing CO2, such as absorption, adsorption, and membrane separation. Among these approaches, absorption by using liquid amines has been widely used for capturing CO2 because of its large capacity and facile application. However, absorption requires enormous energy during regeneration as well as produces polluted products6. On the other hand, membrane separation is a simple, continuous operation; however, it suffers from a drawback of limited performance in case of low CO2 concentrations7. Recently, adsorption is considered to be an efficient technology for capturing CO2 because of its low energy consumption and facile regeneration, without the production of pollution or by-products8.

Zeolites and metal–organic frameworks (MOFs) are widely used commercial CO2 adsorbents, but they suffer from a drawback of degradation of the CO2 adsorption ability under humid conditions9. However, porous carbons can be used for the adsorption of CO2 even under humid conditions, attributed to their hydrophobic properties; they also exhibit advantages of large surface area with high porosity, rapid adsorption–desorption kinetics, thermal and chemical stability, as well as cost-effective preparation10.

Moreover, the characteristics of porous carbon adsorbents can be varied by the use of carbon precursors and activation methods. Activation by physical or chemical treatment is conducted for developing porosities in carbonaceous materials; in physical activation, carbon is partially gasified using oxidising gases, such as steam, CO2, and air11,12,13. On the other hand, chemical activation is conducted by chemical reaction between carbonaceous materials and chemical agents, such as KOH, NaOH, K2CO3, ZnCl2, and H3PO414,15,16,17,18. Generally, as compared to physical activation, chemical activation is favored because it creates a well-developed microporous structure of carbon19. Also, several studies have attempted to correlate CO2 adsorption capacity and textural properties, such as surface area and pore volume, for activated porous carbon20,21,22. Recent studies have suggested that CO2 adsorption capacity at ambient temperature and pressure is closely related to the narrow micropore volume of the adsorbent23,24.

Carbonaceous materials are derived from various carbon precursors, such as polymer, biomass, coal, petroleum residues, and bones25,26,27,28,29. Unur et al. have prepared nanoporous carbon derived from hazelnut shells via carbonization and KOH activation30. The obtained porous carbon exhibited a high specific surface area of 1700 m2 g−1 and pore volume of 0.79 cm3 g−1, the values of which are approximately five times greater than those of carbon without activation. Kante et al. have synthesized coffee-based activated carbon by ZnCl2 activation, with a specific surface area and total pore volume of 1121 m2 g−1 and 0.95 cm3 g−1, respectively31. The textural properties of the samples thus prepared significantly depend on the ZnCl2/C ratio.

In this study, for preparing novel CO2 adsorbents, activated microporous carbon was prepared from the carbonization of wheat flour, followed by KOH activation. Wheat contains various polysaccharides, such as starch, cellulose, glucose, and xylose, which are released from carbonaceous structures when wheat is pyrolyzed at high temperature32,33,34,35. To the best of our knowledge, no studies have been reported about the application of wheat-based porous carbon for CO2 capture. In our study, varying KOH/C ratios were applied during activation, and the effects of the KOH/C ratio on the textural properties and CO2 adsorption abilities of the prepared porous carbon were investigated. In addition, the relationship between the textural properties of wheat-based porous carbon and its CO2 adsorption capacity was investigated, and various characteristics of CO2 adsorption, such as isosteric heat of adsorption, selectivity of CO2 over N2, adsorption kinetics, and cyclic stability, were investigated.

Results and Discussion

Characteristics of KOH-activated MCCs

Figure 1 shows the morphologies of pristine wheat flour, microporous carbon compartments (MCCs) and MCC activated with a KOH/C ratio of 3 (hereafter, referred to as MCC-K3). As shown in Fig. 1a, pristine wheat flour exhibited large irregular-shaped chunks. After carbonization, wheat flour changed to MCC, which is composed of flakes forming big compartments; these compartments contribute to high surface area and pore volume (Fig. 1b). After activation, MCC did not exhibit any significant change in morphology (Fig. 1c). As shown in the TEM image (Fig. 1d), MCC-K3 exhibited worm-like micropores distributed randomly through the carbon. Figure S1 (Supplementary Information) shows the XRD spectra of the porous carbon materials thus obtained, which are regarded as amorphous carbon.

Figure 1
figure 1

SEM images of (a) pristine wheat flour, (b) MCC, and (c) MCC-K3. (d) TEM image of MCC-K3.

Figure S2 (Supplementary Information) shows the thermogravimetric analysis (TGA) of pristine wheat flour. At or less than 100 °C, pristine wheat flour exhibited a weight loss of approximately 7.6 wt%, attributed to water vapour or steam. In addition, weight started to decrease at ~200 °C, and a substantial weight decrease was observed at around 317 °C. During carbonization, a significant amount of non-carbon elements are believed to be pyrolyzed, with the simultaneous development of pores. Figure S3 (Supplementary Information) shows the Fourier transform infrared (FTIR) spectra of pristine wheat flour, MCC, and MCC-K3, confirming the decomposition of these foreign components. Pristine wheat flour exhibited peaks related to various functional groups. A broad band was observed at 3293 cm−1, attributed to hydroxyl O–H, while bands were observed at 2926 and 2857 cm−1, attributed to C–H stretching, and at 1742 cm−1, attributed to C = O stretching vibration36. Peaks were also observed at 1648 and 1536 cm−1, characteristic of stretching vibrations of C–O from amide I and C–N from amide II, respectively37, while peaks observed at 1149 and 1014 cm−1, attributed to C–O stretching vibration36,38. However, no noticeable peaks were observed in the spectra of MCC and MCC-K3, suggesting that the characteristic functional groups of non-carbon elements disappear during carbonization.

Figure 2a shows the N2 adsorption isotherms at 77 K for MCC and KOH-activated MCCs. According to the IUPAC classification, the samples exhibited type I isotherms, indicative of microporous structures39. With increasing KOH/C ratio, the amount of adsorbed N2 increased because of the development of pores. Pore size distribution was estimated using N2 adsorption data and non-local density functional theory (NLDFT), as shown in Fig. 2b. MCC mainly exhibited narrow micropores with a size of 0.5–0.8 nm; this size is known to be favorable for the adsorption of CO240. More narrow micropores with a pore size of less than 0.8 nm were newly developed as more KOH was consumed during activation up to a KOH/C ratio of 3. However, additional larger pores with a size of 0.8–2.0 nm were simultaneously observed for KOH-activated MCCs, attributed to the further reaction of small-sized pores in carbon with KOH, resulting in enlarged pores. As can be clearly observed for highly activated MCC (KOH/C = 4 or 5), when a large amount of KOH was used for activation, the number of enlarged pores became significantly greater than that of newly developed small-sized pores. Hence, the narrow micropore volumes of MCC-K4 and MCC-K5 are less than that of MCC-K3.

Figure 2
figure 2

(a) N2 adsorption isotherms at 77 K, (b) pore size distribution, and (c) cumulative pore volume of MCC and KOH-activated MCCs.

Table 1 summarizes the textural properties obtained from N2 adsorption isotherms at 77 K. MCC exhibited a high specific surface area of 648 m2 g−1 and a total pore volume of 0.299 cm3 g−1, indicating potential for applications as an adsorbent. The textural properties of MCC were further enhanced by KOH activation. With increasing KOH/C ratio during KOH activation, the textural properties gradually increased, and the highest specific surface area of 2192 m2 g−1 and total pore volume of 1.076 cm3 g−1 were obtained for MCC-K5. Also, the volume of narrow micropores having a size of less than 0.8 nm was estimated from the cumulative pore volume (Fig. 2c). The volume of the narrow micropores increased with increasing KOH/C ratio until 3, followed by the decrease in volume for MCC-K4 and MCC-K5, attributed to pore enlargement.

Table 1 Textural properties and CO2 adsorption capacities at temperatures of 0, 25, 50, and 75 °C for MCC and KOH-activated MCCs.

Figure 3 shows the relationship between the KOH/C ratio during activation and the textural properties of samples. With increasing KOH/C ratio, new micropores were developed, and the size of pores enlarged, resulting in increased specific surface area, total pore volume, and micropore volume. According to the calculated linearity (coefficient of determination, R2 > 0.98), the specific surface area, total pore volume, and micropore volume of MCC were highly correlated with the amount of KOH used during activation. However, poor correlation was observed between the KOH/C ratio and volume of narrow micropores, with a pore size of less than 0.8 nm, because excess of KOH (KOH/C = 4 or 5) widened pores, eventually decreasing the narrow micropore volume.

Figure 3
figure 3

Relationship between the KOH/C ratio and textural properties, such as (a) specific surface area, (b) total pore volume, (c) micropore volume, and (d) narrow micropore volume.

CO2 adsorption on KOH-activated MCCs

Table 1 also summarizes the CO2 adsorption capacities at 0, 25, 50, and 75 °C. MCC exhibited CO2 adsorption capacities of 3.44 mol kg−1 and 2.28 mol kg−1 at 0 °C and 25 °C, respectively, and after KOH activation, the adsorption capacities gradually increased with increasing KOH/C ratio up to 3, with the highest CO2 adsorption capacity of 5.70 mol kg−1 and 3.48 mol kg−1 at 0 °C and 25 °C, respectively, for MCC-K3. The CO2 adsorption capacity of MCC-K3 was higher or comparable to those reported recently for porous carbons, such as N-doped carbon and porous carbon activated by KOH, CO2, or steam (Table S1, Supplementary Information)41,42,43,44,45,46,47,48,49,50,51,52. When excess of KOH was used for activation, the CO2 adsorption capacity rather decreased to 3.13 and 2.56 mol kg−1 at 25 °C for MCC-K4 and MCC-K5, respectively.

Figure 4a,b show the adsorption isotherms of CO2 and N2 for MCC and MCC-K3, respectively. Adsorption isotherms fitted well with the dual-site Langmuir–Freundlich model. High CO2 adsorption capacities were obtained at low temperature, implying that the adsorption of CO2 on MCC and MCC-K3 corresponds to exothermic physisorption. As shown in Figure S4 (Supplementary Information), other KOH-activated MCC samples also exhibited similar trends. In addition, notably, MCC and MCC-K3 exhibited higher adsorption capacities for CO2 as compared to N2, attributed to the high quadrupole moment and polarizability of CO2 molecules, which in turn induce stronger interaction between the carbon structure and CO2 as compared to N253.

Figure 4
figure 4

CO2 adsorption isotherms at 0, 25, 50, and 75 °C, and N2 adsorption isotherm at 25 °C of (a) MCC and (b) MCC-K3. Symbols and dashed lines represent experimental data and fitted results, respectively. (c) Isosteric heat of adsorption and (d) IAST-predicted adsorption selectivity of CO2 over N2 at 25 °C for a CO2/N2 binary gas mixture (CO2/N2 = 15:85) for MCC-K3.

Figure 4c shows the isosteric heat of adsorption calculated from the Clausius–Clapeyron equation for MCC-K3 using CO2 isotherms at 0, 25, and 50 °C54. Under low adsorption coverage, isosteric heat was as high as 28.1 kJ mol−1, attributed to the strong interaction between CO2 molecules and the most favorable active sites on porous carbon. However, with increasing adsorption coverage, the favorable active sites were consecutively occupied by CO2, and the isosteric heat of adsorption decreased to 26.9 kJ mol−1, accompanied with weak interactions between adsorbent and adsorbate. The calculated isosteric heat of adsorption varied in the range of ordinary physisorption (<40 kJ mol−1); hence, facile desorption is expected during regeneration.

Typically, the gas emitted from the combustion of fossil fuels contains approximately 15% CO2 with mostly balanced N2. Hence, it is imperative that the adsorbent exhibits high selectivity for CO2 over N2. Selectivity can be calculated by the ideal adsorption solution theory (IAST) using isotherms of CO2 and N2 at 25 °C55. The IAST-predicted selectivity of MCC-K3 for a 15% CO2 and 85% N2 mixture was 15–23 (Fig. 4d). CO2/N2 selectivity determined by applying an initial slope calculation method was ~16 (Figure S5, Supplementary Information); these selectivity values are comparable to those of porous carbon reported recently (Table S1, Supplementary Information). The selectivity of CO2 over N2 can be further enhanced by amine modification or heteroatom doping56,57,58,59.

For better understanding the effect of textural properties on CO2 adsorption, the CO2 adsorption capacities measured at 0, 25, 50, and 75 °C were correlated with the specific surface area, total pore volume, micropore volume, and volume of narrow micropores with a pore size of less than 0.8 nm for KOH-activated MCCs. The specific surface area, total pore volume, and micropore volume were not related to CO2 adsorption capacities (Figure S6, Supplementary Information). However, the volume of the narrow micropores with a pore size of less than 0.8 nm exhibited very close correlation with CO2 adsorption capacities, having high R2 values (Fig. 5). This result is in agreement with those reported in recent studies, suggesting that the narrow micropore volume plays a significant role in determining CO2 adsorption performance60,61,62. The CO2 adsorption capacity has been reported to be primarily affected by the number of narrow micropores because the van der Waals force from the surrounding walls of narrow micropores is thought to provide favorable interaction between the carbon structure and CO2 molecules63. Hence, for the efficient adsorption of CO2, it is imperative to develop narrow micropores of less than 0.8 nm in porous carbon by controlling the KOH/C ratio.

Figure 5
figure 5

Correlation between CO2 adsorption capacities and volume of narrow micropores with a pore size of less than 0.8 nm.

As important characteristics for CO2 adsorbents, adsorption kinetics, cyclic stability, and ease of regeneration are as important as high CO2 adsorption capacity. For evaluating adsorption kinetics, the adsorption of CO2 on MCC-K3 was measured by TGA at 30, 40, and 50 °C under atmospheric pressure. After the start of CO2 gas flow, CO2 adsorption uptake reached 50% of the equilibrium capacity within 2 min and 70% within 5 min, indicating a rapid CO2 adsorption rate. In addition, for describing the CO2 adsorption rate of MCC-K3, the widely used pseudo-first-order and pseudo-second-order models were applied. The pseudo-first-order model is based on the adsorption rate, which is proportional to the number of possible adsorption sites, and it can be expressed as follows:

Here, qe and qt denote the adsorption uptake at equilibrium and at certain time t, respectively, and k1 is the pseudo-first-order adsorption rate constant. By applying the boundary conditions of qt = 0 at t = 0 and qt = qe at t = ∞, the integrated form of equation can be written as follows:

The pseudo-second-order model is based on the assumption that the adsorption rate is proportional to the square of the number of possible adsorption sites, and the adsorption rate can be expressed as follows:

Here, k2 is the pseudo-second-order adsorption rate constant. By integrating the rate equation with the same boundary conditions of qt = 0 at t = 0 and qt = qe at t = ∞, equation (3) can be rearranged as follows:

Figure 6a shows the fitted curves using the pseudo-first-order and pseudo-second-order kinetic models, and Table 2 summarizes the related parameters. As compared to the pseudo-first-order model (R2 = 0.70–0.72), the pseudo-second-order model well fitted the experimental data (R2 = 0.92–0.96) at various adsorption temperatures. The small normalized standard deviation (ΔErr%) also confirms that the pseudo-second-order model is more suitable than the pseudo-first-order model for describing experimental data.

Table 2 CO2 adsorption kinetic parameters for MCC-K3 using pseudo-first-order and pseudo-second-order models at temperatures of 30, 40, and 50 °C.
Figure 6
figure 6

(a) CO2 adsorption kinetics for MCC-K3. Symbols and dashed lines denote experimental data and model fittings, respectively. (b) Arrhenius plot of pseudo-second-order adsorption rate constants for the estimation of activation energy of CO2 adsorption on MCC-K3. (c) Cyclic stability test for MCC-K3. Solid and dashed lines represent the adsorption and desorption steps, respectively.

Basically, rate constants are dependent on temperature, and temperature dependency can be described by the Arrhenius equation as follows:

Here, A is the Arrhenius pre-exponential factor, Ea is the apparent activation energy, R is the universal gas constant and T is the absolute temperature. Figure 6b shows the linear plot of the natural logarithm of k2 against reciprocal temperature, which exhibited reasonable linearity. From the slope and intercept of the linear plot, MCC-K3 exhibited an Ea of 23.61 kJ mol−1 and A of 710.17 mol kg−1 min−1. Activation energy ranged from 5 to 40 kJ mol−1, indicating that CO2 adsorption is predominantly physisorption64. The obtained Arrhenius parameters can be used for predicting the adsorption rate constants at different operating temperatures.

For investigating the cyclic stability of MCC-K3, a repeated adsorption–desorption test was conducted by alternatingly switching the flowing gas between N2 and CO2 at a fixed temperature at 25 °C and ~1 bar, as shown in Fig. 6c. Adsorbed CO2 was easily desorbed simply by changing the gas flow from CO2 to N2, which was expected from the low heat of adsorption of MCC-K3. During 10 cycles, MCC-K3 was successfully regenerated by gas purging, with cyclic stability, retaining ~99% of the initial CO2 adsorption capacity.

Conclusions

In this study, microporous carbon compartments (MCCs) were prepared from wheat flour via carbonization and KOH activation for the capture of CO2 by adsorption. The carbonization of pristine wheat flour induced the pyrolysis of non-carbon elements, developing big rooms, which can permit gas molecules to access micropores, and the textural properties of MCCs were further improved by chemical activation with KOH. The KOH/C ratio was varied from 1 to 5, and the effects of KOH/C ratio on the textural properties and CO2 adsorption performance were investigated. With increasing KOH/C ratio up to 3, narrow micropores of less than 0.8 nm were primarily developed, while larger micropores were also produced from pore enlargement with excess KOH. An optimal KOH/C ratio was clearly identified for activation for attaining the maximum CO2 adsorption capacity of MCC: MCC-K3 exhibited the highest CO2 adsorption capacity of 5.70 and 3.48 mol kg−1 at 0 and 25 °C, respectively. The CO2 adsorption capacity significantly depended on the volume of the narrow micropores with a pore size of less than 0.8 nm rather than the surface area or pore volume of larger pores. MCC-K3 also exhibited excellent cyclic stability with facile regeneration, high selectivity for CO2 over N2, and rapid adsorption rates. As compared to the pseudo-first-order model, the pseudo-second-order kinetic model described the experimental adsorption data methodically. The outstanding overall CO2 adsorption performance indicate that KOH-activated MCCs can be promising CO2 adsorbents.

Methods

Synthesis of MCC and KOH-activated MCCs

For preparing microporous carbon compartments (MCCs), first, pristine wheat flour was placed in a horizontal cylindrical furnace with an inner diameter of 50 mm and then heated at a rate of 5 °C min−1 up to 900 °C under N2. Heating was maintained for 2 h. The MCC thus obtained was ground with an agitator mortar.

For activation, MCC was mixed with the prepared KOH solution by continuous stirring at 60 °C for 2 h. The KOH to MCC weight ratio varied in the range from 1 to 5. The mixture was then dried at 110 °C overnight, and the dried material was activated at 700 °C for 1 h at a heating rate of 5 °C min−1 under N2. The resulting material was thoroughly washed with a 10% HCl solution for removing any inorganic salt and then washed with distilled water. KOH-activated MCC is denoted as MCC-Kx, where x denotes the KOH/C ratio.

Sample characterization

The surface morphologies of KOH-activated MCCs were analyzed by scanning electron microscope (SEM, S-4300, Hitachi) and high-resolution transmission electron microscope (HRTEM, G2 F30ST, Tecnai). The specific surface area was estimated by the Brunauer, Emmett, and Teller (BET) equation based on the information obtained by N2 adsorption at 77 K using a volumetric sorption analyzer (ASAP2020, Micromeritics). X-ray diffraction patterns were recorded with an X-ray diffractometer (XRD, X’ Pert MPD, Philips) using Cu Kα radiation in the 2θ range of 5–50°. A thermogravimetric analyzer (TGA, Q50, TA instruments) was used for measuring the weight loss of the pristine wheat flour at a heating rate of 10 °C min−1 under N2.

CO2 adsorption uptake was measured by the same volumetric sorption analyzer used for N2 adsorption isotherms. Prior to adsorption tests, samples were degassed at 150 °C under a vacuum (10 μm-Hg) for 12 h, and CO2 adsorption isotherms were obtained at four different temperatures of 0, 25, 50, and 75 °C. Cyclic adsorption stability was tested by TGA with CO2 adsorption at 25 °C and atmospheric pressure and desorption under N2 purge. CO2 adsorption was carried out for 40 min and the sample was regenerated using N2 purge for 90 min after each CO2 adsorption step. Also, CO2 adsorption kinetic studies were conducted by TGA, recording the weight change under CO2 for 3 h at temperatures of 30, 40, and 50 °C.

Additional Information

How to cite this article: Hong, S.-M. et al. CO2 Capture in the Sustainable Wheat-Derived Activated Microporous Carbon Compartments. Sci. Rep. 6, 34590; doi: 10.1038/srep34590 (2016).