Oxygen-enriched surface modification for improving the dispersion of iron oxide on a porous carbon surface and its application as carbon molecular sieves (CMS) for CO₂/CH₄ separation

Abstract

The separation of CO₂/CH₄ can be enhanced by impregnating porous carbon with iron oxide. Dispersion of iron oxide is one of the critical factors which supports the separation process performance. Iron oxide dispersion can be enhanced by enriching the oxygen functional groups on the carbon surface. This study investigates three distinct oxidation processes: oxidation with a 10% H₂O₂ solution, ozonation with distilled water, and ozonation with a 10% H₂O₂ solution. The research steps included the following: (i) oxidation, (ii) impregnation of iron oxide followed by calcination, (iii) material characterization, and (iv) material performance analysis. Materials were characterized using N₂ sorption analysis, X-ray diffraction analysis (XRD), scanning electron microscopy-energy dispersive X-ray spectroscopy analysis (SEM-EDX), and Fourier transform infrared analysis (FT-IR). Iron oxide was well dispersed on the carbon surface, as evidenced by the elemental mapping of materials. In addition, the oxygen functional groups increased significantly in the range of 28.6–79.7% following the oxidation process, as indicated by the elemental component using SEM-EDX analysis. The impregnation of iron oxide on oxidized carbon ozonated with distilled water (COA–Fe) obtained a maximum CO₂ uptake capacity of 3.0 mmol g⁻¹ and CO₂/CH₄ selectivity increased by up to 190% at a temperature of 30 °C and pressure of 1 atm. Furthermore, the enhancement of CO₂/CH₄ separation up to 1.45 times was the best performance achieved by COA–Fe. Thus, improving iron oxide dispersion on oxidized carbon surfaces has a potential application in CO₂/CH₄ separation.

---

1 Introduction

Biogas is a renewable energy source that can be substituted for fossil fuels and natural gas. A typical biogas mixture contains 50–70% CH₄ and 30–49% CO₂.^1,2 Carbon dioxide removal from CO₂/CH₄ gas mixtures is critical since it results in a more energy-dense product, due to the high calorific value of methane. In contrast, carbon dioxide has no heating value.³ A number of technologies are currently available to remove carbon dioxide, including absorption, membrane separation, and cryogenic separation.^4,5 However, these technologies are energy-intensive and costly. Adsorption-based separation is considered promising for CO₂/CH₄ separation because it produces high-purity methane (>98 percent vol), is relatively inexpensive, easy to operate, and energy-efficient.⁶ Adsorption-based separation occurs due to the difference in affinity and diffusivity of CO₂ and CH₄.⁷ Several types of molecular sieves, including zeolite molecular sieve (ZMS),^8,9 carbon molecular sieve (CMS),^10–13 and metal–organic framework (MOF),^14,15 are widely used for adsorption-based separation. Molecular sieves made from carbon are relatively stable long-term due to the low heat of adsorption and can be synthesized from a variety of materials, including coal,^16,17 biomass,^10,18 and polymer.^12,19 As a consequence, it is simple to regenerate.

CMS material derived from palm kernel shells (PKS) was investigated for CO₂/CH₄ separation. This material achieved separation ratios of up to 2.¹⁰ However, the adsorption capacity of CO₂ is quite limited, which will require further investigation. Metal oxide impregnation may enhance carbon dioxide uptake.²⁰ In contrast, metal oxide agglomerations are common due to the fact that carbon is non-polar and hydrophobic, whereas metals are polar and hydrophilic. To enhance dispersion, reduced metal oxide agglomeration is employed. In recent years, several authors have discussed the use of highly dispersed iron oxide for a wide variety of applications.^21–23

The addition of oxygen groups to the surface chemistry of carbon can increase its hydrophilic character.²⁴ As a consequence, it becomes wettable, and metal oxide dispersion is enhanced. Oxygen-enriched surfaces have been widely used in a variety of modification processes, including gaseous or aqueous oxidation, ozonation, and gamma irradiation, etc. Aqueous oxidation enhanced carboxylic acid functional groups significantly, whereas gaseous oxidation enhanced carbonyl and hydroxyl functional groups significantly. While ozonation increases the number of acid surface groups, it alters the surface area and porosity.²⁵

In this study, porous carbon impregnated with iron oxide was utilized to fabricate a molecular sieve for CO₂/CH₄ separation. Palm kernel shell biomass was a precursor to porous carbon. Three different oxidation processes were evaluated in order to increase the hydrophilicity of carbon: oxidation with 10% H₂O₂, ozonation with distilled water, and ozonation with 10% H₂O₂ solution. In the CMS preparation, the oxidation processes are followed by iron oxide impregnation and calcination. An investigation of the effect of three different oxidation processes on the dispersion of iron oxide impregnated porous carbon on separation performance is conducted, which has not been previously described. In this study, the characterization of materials is extended to ensure the feasibility of impregnation, adsorption isotherms, and breakthrough separation for the purpose of separation.

2 Experimental

2.1. Materials

Porous carbon from palm kernel shell (20–25 mesh) was obtained from PT Home System Indonesia. Hydrogen peroxide with a purity of 50% from PT Indonesia Inti Pratama. Iron(III) nitrate nonahydrate for analysis EMSURE® ACS, ISO, Reag. Ph Eur from Merck. Nitrogen with a purity of 99.95% was used as inert gas in the calcination process. Methanol for analysis EMSURE® ACS, ISO, Reag. Ph Eur from Merck as a solvent for metal impregnation. CO₂ and CH₄ with a purity of 99.9% were obtained from PT. Aneka Gas Industri Indonesia as adsorbate. A mixed gas CH₄/CO₂ (55/45% v/v) as a biogas representative.

2.2. Preparation of oxidized carbon

Oxidized carbon was obtained by oxidizing porous carbon with H₂O₂ (10% v/v), ozonating it with distilled water, and ozonating it with an H₂O₂ (10% v/v) solution. To oxidize a porous carbon with a 10% H₂O₂ solution, a 15 g porous carbon was mixed with 150 mL 10% H₂O₂ solution at room temperature for two hours, followed by filtering and heating to 60 °C for 24 hours. As part of the ozonation process, a 1 : 10 (w/v) mixture of porous carbon and distilled water was placed in the ozone contact reactor (IONTECH QLA-3G ozone generator). Ozone gas was fed into the system at a flow rate of 3000 mg h⁻¹ and a power output of 60 watts for 120 minutes. The ozone–H₂O₂ reaction was initiated by substituting distilled water for a 10% H₂O₂ solution.

2.3. Impregnation of iron oxide

The iron oxide was dispersed on porous carbon by an incipient wetness method, followed by calcination. At first, pristine 15 g carbon was degassed at a temperature of 150 °C for 2 hours in order to clean the pores. Carbon pores were then slowly filled with a salt solution of 5.4 g iron(III) nitrate nonahydrate in 12 mL methanol. The sample was heated under nitrogen flow for 6 hours at 500 °C during the calcination process. Iron as iron oxide was targeted at 5% by weight. The procedure was also applied to oxidized carbon. A schematic of the CMS preparation is presented in Fig. 1.

Fig. 1 Schematic procedure of iron oxide impregnation onto porous carbon as CMS.

2.4. CMS characterization

Fourier transforms infrared (FTIR) analysis was carried out for the functional group's characterization of the material by using Nicolet Avatar 360 IR. In this study, samples were analyzed at a wavelength of 400–4000 cm⁻¹. The morphology of the CMSs was characterized by SEM-EDX instrument using JEOL JSM-6510 LA at a voltage of 10 kV. The structure of crystalline materials was characterized by XRD using Bruker D2 Phaser. The diffractogram was obtained by using Cu-K_α radiation (λ = 1.5406 Å) in the range 10° < 2θ < 90°.

2.5. Adsorption isotherm measurement

The adsorption capacity of CO₂ and CH₄ was measured by the volumetric method from 0 to 1.2 atm at a temperature of 30 °C. An ultrahigh vacuum adsorption apparatus rig was constructed using Swagelok® VCR valves and fittings. First, CMS samples were degassed for 6 hours at 150 degrees celsius until a static pressure of at least 0.01 torr was reached. These results are presented as an adsorption isotherm curve. The schematic diagram and procedure for measuring adsorption isotherms have been published in the literature.²⁶

2.6. Separation performance with a breakthrough analysis

The composition of mixed gases (CH₄ and CO₂) was determined using a portable gas analyzer (Biogas Analyzer Gas Board 3200plus, Hubei Cubic-Ruiyi Instrument Co., Ltd.). The material was first placed in the column with fixed bed dimensions (D = 9.5 mm and L = 300 mm), followed by a 200 mL min⁻¹ nitrogen flush until no gas content of CH₄, CO₂, or O₂ could be detected on the gas detector. The mixed gas mixture, comprised of CH₄ and CO₂, was then introduced into the system at a flow rate of 50 mL min⁻¹ at room temperature and 1.2 bar pressure. The separation cycle was continued until the gas composition returned to the initial concentrations of the gas inlet.¹³

3 Results and discussion

3.1. CMS characteristics

Functional groups and surface morphologies. Fig. 2(a) illustrates the FT-IR spectra of CMSs. The spectral characteristics of pristine and oxidized carbon correspond to wavelengths of 3440 cm⁻¹, 2350 cm⁻¹, 1580 cm⁻¹, 1130 cm⁻¹. The peak of 3440 cm⁻¹ represents the presence of a hydroxyl group (O–H stretching).²⁷ Meanwhile, the peak of 1580 cm⁻¹ and 1130 cm⁻¹ denote the presence of a carboxyl group (C=O stretching)^27–30 and C–O group (stretching), respectively.²⁹ It is remarkable that there is an increase in the intensity of the O–H stretching, C=O stretching, and C–H stretching when compared to pristine carbon. As a result, the oxygen groups on all oxidized carbons increase.

Fig. 2 FT-IR spectra of (A) oxidized carbon, and (B) oxidized carbon impregnated with iron oxide.

As shown in Fig. 2(b), the FTIR spectrum and absorption bands of iron oxide impregnated porous carbon. There is an addition at the peak of 590 cm⁻¹ in comparison to the pristine and oxidized carbon prior to iron oxide impregnation. This peak is thought to be caused by Fe–O bonds. A peak at 576.3 cm⁻¹ was also observed in a composite of activated carbon and iron oxide (Fe₃O₄).³¹

SEM images of each material reveal the oxidized carbon's morphological structure (CH, COA, and COH). Fig. 3 illustrates the results of SEM images of C (a), CH (b), COA (c), and COH. SEM images reveal an increase in the size of oxidized carbon pores.²⁵ This increase has been modified by either oxidation treatment using 10% H₂O₂ solution, ozonation using distilled water, or ozonation using 10% H₂O₂ solution. According to the SEM images, the pore size of the carbon can become irregular as a result of the ozonation process, whether using distilled water (COA) or a 10% H₂O₂ solution. In comparison to carbon, which is oxidized with a 10% H₂O₂ solution (CH). Despite the increased size of the CH pore cavities, the carbon structure remains regular.

Fig. 3 SEM images of (a) C, (b) CH, (c) COA, and (d) COH.

The composition of C and O was determined via SEM-EDX analysis. Table 1 summarizes the percentage increase in oxygen groups for CH, COA, and CH. It can be concluded that each treatment increases the oxygen groups in the following order: COA > COH > CH.

Table 1 Percentage increase in oxygen groups for CH, COA and CH

Oxidation method	Sample	Element (%)		Addition of oxygen group (%)
		C	O
Pristine carbon	C	93.06	6.94	—
Oxidation using 10% H2O2 solution	CH	91.07	8.93	28.67
Ozonation using distilled water	COA	87.77	12.33	79.74
Ozonation using 10% H2O2 solution	COH	89.95	10.05	38.43

---

The SEM micrographs of C–Fe (a), CH–Fe (b), COA–Fe (c), and COH–Fe (d), as well as the elemental mapping of iron oxide on C–Fe (e), CH–Fe (f), COA–Fe (g), and COH–Fe (h), are shown in Fig. 4. Based on Fig. 4(a–d), it can be seen that there are some small aggregates of iron oxide visible from the lateral view of the CMS. These aggregates are brighter in colour and are supported on the darker surface of the porous carbon.

Fig. 4 SEM images of C–Fe (a), CH–Fe (b), COA–Fe (c), COH–Fe (d) and their elemental mapping of Fe on C–Fe (e), CH–Fe (f), COA–Fe (g), COH–Fe (h).

As shown in Fig. 4(a) and its elemental mapping in Fig. 4(e), an agglomeration of iron oxide exists on the pristine carbon surface. This demonstrates that the iron salt solution is incapable of spreading uniformly across the pristine carbon surface, resulting in the formation of iron oxide agglomerates. This is because the iron salt solution is unable to wet the surface of the pristine carbon due to its hydrophobic nature. While iron oxide is widely dispersed on the surface of oxidized carbon, as illustrated in Fig. 4(b–d) and as indicated by the elemental mapping in Fig. 4(f–h). This indicates that the addition of oxygen functional groups increases the wettability of carbon while decreasing its hydrophobicity. The composition of iron oxide impregnated porous carbon was determined via SEM-EDX analysis. The results are presented in Table 2.

Table 2 Elements detected on CMSs by EDX analysis

Sample	Element (%)
	C	O	Fe
C–Fe	83.47	10.57	5.96
CH–Fe	80.97	13.78	5.25
COA–Fe	73.23	20.96	5.81
COH–Fe	79.71	14.79	5.50

---

Crystallinity of CMSs. The crystal structure of iron oxide on porous carbon can be determined using X-ray diffraction patterns. Fig. 5 illustrates the XRD patterns of porous carbon and iron oxide impregnated carbon. Amorphous carbon is indicated by the diffraction peaks of pristine and oxidized carbon in the planes of (002), (101), and (110) indicate the amorphous carbon. Meanwhile, for all oxidized carbon, the diffraction peaks at 2θ value of 30.5°, 35.9°, 43.5°, 54.0°, 57.2°, and 63° indicate the presence of Fe₃O₄ in the cubic spinel structure of (220), (311), (400), (422), (511), and (440) confirmed.³² Thus, the iron oxides prepared are magnetite. From the diffraction peak obtained, no other phase of iron oxide is detected in the sample.

Fig. 5 XRD pattern of CMSs.

An estimation of the magnetite Fe₃O₄ crystallite size was performed using Debye–Scherrer's equation. The mean crystallite size calculated from the main diffraction peak (311) around 35.9° was listed in Table 3.

Table 3 Crystallite size of Fe₃O₄ for all oxidized carbon from Scherrer's equation calculation

Sample	Crystallite size, nm
C–Fe	18.34
CH–Fe	9.65
COA–Fe	16.93
COH–Fe	14.57

---

Nitrogen sorption analysis. The textural properties were determined using N₂ adsorption–desorption analysis. Fig. 6 illustrates the results. According to the IUPAC classification, both pristine and oxidized carbon exhibit a type I isotherm, which is characteristic of microporous materials. Pristine carbon has a large amount of nitrogen adsorbed (ca. 225 cm³ g⁻¹ STP at 1P/P₀). The volume of nitrogen adsorbed by oxidized carbon decreases. Interestingly, the isotherm of adsorption–desorption for oxidized carbon is not closed. This is most likely due to the presence of pores in the shape of an ink bottle following the oxidation process.¹³ The pore size distribution was evaluated using the Horvath–Kawazoe (HK) model. Pristine carbon-centered at 0.75 nm. These findings establish unequivocally that pristine carbon exists in a microporous region. Due to the formation of cavities during oxidation, the center of oxidized carbon shifts to a larger size of approximately 0.8–1 nm.²⁵

Fig. 6 Nitrogen adsorption–desorption isotherm for oxidized carbon (a) and iron oxide impregnated carbon (b) (inset: pore size distribution). Unit y-axis is in Standard Temperature and Pressure (STP).

Fig. 6(b) shows a nitrogen sorption isotherm for oxidized carbon impregnated with iron oxide. As can be seen from the isotherm, the CMS also has a type I isotherm. Interestingly, COA–Fe has a narrow and sharp pore size distribution compared to other materials, which is advantageous for CMS applications.

Table 4 summarizes the specific surface area (SSA), micropore surface area (S_mic), total pore volume (V), micropore volume (V_mic), and mean pore diameter (D_avg). Pristine carbon has a high surface area of 708 m² g⁻¹, but its specific surface area decreases by ca. 15–25% after oxidation and iron oxide loading. The reduction is most likely the result of cavities formed during the oxidation process³³ and the pore being occupied during the impregnation process.^13,21,34 All materials exhibit a predominance of microporous structure, both in terms of surface area and pore volume.

Table 4 Pore textural properties evaluated by nitrogen sorption

Parameter	C	CH	COA	COH	CH–Fe	COA–Fe	COH–Fe
SSA, m^2 g^−1	708	602	561	528	649	612	622
Smic, m^2 g^−1	651	560	520	491	603	573	573
% Smic	92.0	93.0	92.7	93.0	93.0	93.6	92.1
V, cm^3 g^−1	0.33	0.28	0.27	0.26	0.31	0.29	0.30
Vmic, cm^3 g^−1	0.25	0.22	0.21	0.20	0.23	0.22	0.22
% Vmic	75.8	78.6	77.8	76.9	74.2	75.9	73.3
Davg, nm	1.88	1.90	1.95	1.96	1.89	1.86	1.95

---

3.2. CO₂ and CH₄ adsorption isotherm

The CO₂ and CH₄ adsorption isotherm curves in. Fig. 7(a) were determined at a temperature of 30 °C and pressure of up to 1 atm. As shown in Fig. 7(a), the CO₂ uptake capacity was significantly greater than the CH₄ uptake capacity. It was consistent with the other outcomes.¹³ Furthermore, impregnation with iron oxide can increase CO₂ uptake capacity while decreasing CH₄ uptake capacity. The uptake capacity of CH₄ on the impregnated carbon is lower than that of pristine carbon. The increased capacity for CO₂ uptake is most likely due to the iron oxide, as the active site has a higher affinity for CO₂ than the pristine carbon surface. However, impregnation of carbon with iron oxide resulted in a decrease in the affinity of CH₄. It will improve CO₂/CH₄ selectivity, which is beneficial for CO₂/CH₄ separation. The pristine carbon has a CO₂ uptake capacity of 1.85 mmol g⁻¹ at a pressure of 1 bar, whereas the CH–Fe, COA–Fe, and COH–Fe have capacities of 2.94 mmol g⁻¹, 3.00 mmol g⁻¹, and 2.94 mmol g⁻¹, respectively. Meanwhile, the CH₄ uptake capacities of pristine carbon, CH–Fe, COA–Fe, and COH–Fe are respectively 1.08 mmol g⁻¹, 0.85 mmol g⁻¹, 0.92 mmol g⁻¹, and 0.89 mmol g⁻¹. In comparison to other studies in the literature (Table 5), the CO₂ uptake capacity of oxidized carbon impregnated with iron oxide is attractive and can compete with other impregnation of metal oxide.

Fig. 7 Adsorption of CO₂ and CH₄ on pristine carbon, CH–Fe, COA–Fe and COH–Fe (a) and selectivity of CO₂/CH₄ (b) at temperature of 30 °C.

Table 5 Comparison of adsorption capacity with other investigations at various temperatures and the pressure of 1 bar

Sorbent type	Metal/metal oxide	Uptake capacity of CO2, mmol g^−1	Temp., °C	Ref.
Oxidized AC (CH)	Fe3O4	2.95	30	This work
Oxidized AC (COA)	Fe3O4	3.01	30	This work
Oxidized AC (COH)	Fe3O4	2.94	30	This work
Mesoporous carbon	NiO	2.00	30	37
AC	MgO	2.72	0	38
AC	Cu/Zn	2.25	30	39
AC	CuO	0.30	25	40
Unmodified AC	NiO	3.02	30	41

---

The selectivity of CO₂/CH₄ is a critical factor in the separation of CO₂/CH₄. Fig. 7(b) illustrates the selectivity of CO₂/CH₄. The data indicate that as pressure increases, the selectivity value of pristine carbon decreases, which is consistent with previous literature.^13,35 On the pressure range investigated, the selectivity of CO₂/CH₄ for pristine carbon was ca. 3.75–1.7. Interestingly, for iron oxide-impregnated carbon, the selectivity increased as the pressure was increased. This is most likely because CO₂ chemisorption occurs at active sites such as iron oxide. Iron oxide's active sites may have a strong affinity for CO₂. As a result, the intermolecular interaction between CO₂ and iron oxide is much stronger than the interaction between CH₄ and iron oxide, resulting in a significant increase in the uptake capacity of CO₂ as the gas's pressure increases. As a result, CO₂ has a greater selectivity than CH₄.³⁶ Meanwhile, physisorption occurs as a result of CO₂ adsorption onto pristine carbon. When iron oxide impregnated oxidized carbon was compared to pristine carbon, the selectivity of CO₂/CH₄ increased up to 190%.

3.3. Performance of CO₂/CH₄ separation

The breakthrough analysis was conducted to determine the CO₂/CH₄ separation performance of the materials. A 45 : 55% mixture of CO₂/CH₄ was flowed into a packed bed column containing the material, and the gas composition at the outlet was monitored over time. The CO₂ breakthrough curve is depicted in Fig. 8. The breakthrough curves plot the outlet concentration of gas species at a certain time (C_t)/the initial inlet concentration (C₀) versus time. As illustrated in Fig. 8(a), there was no difference in the time required to achieve CO₂ breakthrough between pristine and oxidized carbon. Thus, oxidation enhances the oxygen functional group on the carbon surface but does not transform the time of the breakthrough. Based on the data, iron oxide impregnated-oxidized carbon has an increase in CO₂ breakthrough time (see Fig. 8(b)). Pristine carbon achieved a CO₂ breakthrough time of 400 s, while iron oxide impregnated-pristine carbon increased slightly (425 s). However, iron oxide impregnated-oxidized carbon resulted in a 25–45% increase. The breakthrough time for CO₂ was defined as when the concentration of CO₂ reaches 5%, whereas the requirement for biomethane is at least 95%. CH–Fe, COA–Fe, and CH–Fe had CO₂ breakthrough times of 500 s, 570 s, and 530 s, respectively. The CO₂ breakthrough curve may indicate that iron oxide impregnated oxidized carbon prepared via ozonation with distilled water has a superior iron oxide dispersion. The order of the CO₂ breakthrough time is CH–Fe < COH–Fe < COA–Fe. The results of this study indicate that the addition of oxygen groups to carbon surfaces can increase iron oxide dispersion on the surface, which can enhance CO₂/CH₄ separation.

Fig. 8 Carbon dioxide-breakthrough curves of mixed gas CO₂/CH₄ for (a) C, CH, COA and COH; (b) C–Fe, CH–Fe, COA–Fe and COH–Fe.

Fig. 9 displays the breakthrough curves of CO₂ and CH₄ for the materials. As shown in Fig. 9, no CH₄ or CO₂ gas is present initially until approximately 250 seconds, after which CH₄ appears with a purity of >98 percent until a certain time, and finally, CO₂ emerges until the final concentration in the outlet equals the inlet. A similar CH₄ breakthrough time was obtained for all materials, but with a different CO₂ breakthrough time. A delayed flow of CH₄ at the outlet could indicate a slower diffusion of gas through the bed, caused by the interaction of carbon surface with gas species. Additionally, a larger curve between the CH₄ and CO₂ signals indicated that CO₂ and CH₄ had a better separation performance. COA–Fe generates the largest curve.

Fig. 9 Breakthrough curves of mixed gas CO₂/CH₄ (♦C, ▲CH–Fe, ●COA–Fe, ■COH–Fe for CO₂ and ◊C, ΔCH–Fe, □COA–Fe, ○COA–Fe for CH₄).

Generally, the separation process resembles that of chromatographic separation. The chromatography separation occurs when the components with a faster rate of movement are separated first. In the separation of CO₂/CH₄, molecule types with higher diffusivity values or those that diffuse more rapidly are separated first. The illustration mechanism of separation is depicted in Fig. 10. In a column, as shown in Fig. 10(a) and (b), CH₄ diffuses more rapidly than CO₂.¹³ Therefore, CH₄ and CO₂ can be separated. CO₂ diffuses more slowly over iron oxide impregnated carbon in Fig. 10(b) than it does over pristine carbon (Fig. 10(a)). Iron oxide exhibits a stronger affinity for CO₂. This can result in increased interaction between CO₂ and the iron oxide active surface, resulting in a slower rate of CO₂ diffusion in the CMS column. As a result, the process of separation lengthens.

Fig. 10 Separation mechanism using (a) pristine carbon and (b) oxidized carbon impregnated with iron oxide.

3.4. Performance comparison on CH–Fe₂O₃ and CH–Fe₃O₄

The type of iron oxide formed in this study is Fe₃O₄. CH–Fe₃O₄ was mentioned previously in conjunction with CH–Fe, which can improve CO₂/CH₄ separation performance by up to 125%. Its performance will be compared to that of other types of iron oxide. Hematite (Fe₂O₃) was chosen as a representative of other iron oxides due to its high affinity for CO₂.²⁰ Fig. 11 compares the CO₂ breakthrough for CH, CH–Fe₃O₄, and CH–Fe₂O₃. The CO₂ breakthrough curves for all samples are ideal, indicating that mass transfer occurs instantaneously.¹³ The order of breakthrough time is CH–Fe₂O₃ (260 s) < CH (400 s) CH–Fe₃O₄ (500 s).

Fig. 11 Carbon dioxide-breakthrough curves of mixed gas CO₂/CH₄ for CH, CH–Fe₃O₄, and CH–Fe₂O₃.

4 Conclusions

The feasibility of preparing iron oxide impregnated porous carbon was investigated. In this study, increased dispersion of iron oxide on the carbon surface can be achieved by adding oxygen via the three oxidation processes used. The impregnation of iron oxide on oxidized carbon can be characterized as CMS for CO₂/CH₄ separation. CMS, which was obtained by aqueous ozonation with distilled water, followed by impregnation with iron oxide and labeled COA–Fe, exhibited superiority. The COA–Fe can provide an enhancement of CO₂ uptake capacity up to ca. 1.7 times at 30 °C and 1 atm, while the enhancement of CO₂/CH₄ separation up to ca. 45% compared to pristine carbon. Furthermore, the iron oxide of Fe₃O₄ is more favorable for CO₂/CH₄ separation than that of Fe₂O₃.

Author contributions

Nur Indah Fajar Mukti: conceptualization, investigation, methodology, formal analysis, funding acquisition, writing-original draft. Teguh Ariyanto: supervision, funding acquisition, writing-review and editing. Wahyudi Budi Sediawan: supervision, writing-review and editing. Imam Prasetyo: conceptualization, supervision, funding acquisition, writing-review and editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was supported by Indonesia Endowment Fund for Education (LPDP), Indonesian Ministry of Finance through Dissertation Research Grant (Beasiswa BUDI-DN). The authors wish thanks to PT Home System for the gift of porous carbon produced from palm kernel shell.

References

1. I. U. Khan, M. H. D. Othman, H. Hashim, T. Matsuura, A. F. Ismail, M. Rezaei-DashtArzhandi and I. W. Azelee, Energy Convers. Manage., 2017, 150, 277–294.
2. I. Angelidaki, L. Treu, P. Tsapekos, G. Luo, S. Campanaro, H. Wenzel and P. G. Kougias, Biotechnol. Adv., 2018, 36, 452–466.
3. M. T. Kallo and M. J. Lennox, Langmuir, 2020, 36, 13591–13600.
4. O. W. Awe, Y. Zhao, A. Nzihou, D. P. Minh and N. Lyczko, Waste Biomass Valorization, 2017, 8, 267–283.
5. X. Y. Chen, H. Vinh-Thang, A. A. Ramirez, D. Rodrigue and S. Kaliaguine, RSC Adv., 2015, 5, 24399–24448.
6. I. Durán, N. Álvarez-Gutiérrez, F. Rubiera and C. Pevida, Chem. Eng. J., 2018, 353, 197–207.
7. I. Prasetyo, R. Rochmadi and E. Wahyono, Reaktor, 2010, 13, 24–30.
8. N. K. Jensen, T. E. Rufford, G. Watson, D. K. Zhang, K. I. Chan and E. F. May, J. Chem. Eng. Data, 2012, 57, 106–113.
9. M. Younas, M. Sohail, L. L. Kong, M. J. K. Bashir and S. Sethupathi, Int. J. Environ. Sci. Technol., 2016, 13, 1839–1860.
10. I. Prasetyo, N. I. F. Mukti, R. B. Cahyono, A. Prasetya and T. Ariyanto, Waste Biomass Valorization, 2020, 11, 5599–5606.
11. I. Prasetyo and D. D. Do, Chem. Eng. Sci., 1998, 53, 3459–3467.
12. I. Prasetyo, R. Rochmadi, E. Wahyono and T. Ariyanto, Eng. J., 2017, 21, 83–94.
13. T. Ariyanto, K. Masruroh, G. Yunita, S. Pambayun, N. Indah, F. Mukti, R. B. Cahyono, A. Prasetya and I. Prasetyo, ACS Omega, 2021, 6, 19194–19201.
14. W. Lou, J. Yang, L. Li and J. Li, J. Solid State Chem., 2014, 213, 224–228.
15. D. Britt, H. Furukawa, B. Wang, T. G. Glover and O. M. Yaghi, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 20637–20640.
16. J. Alcañiz-Monge, J. P. Marco-Lozar and D. Lozano-Castelló, Fuel Process. Technol., 2012, 95, 67–72.
17. Z. Yang, D. Wang, Z. Meng and Y. Li, Sep. Purif. Technol., 2019, 218, 130–137.
18. A. R. Mohamed, M. Mohammadi and G. N. Darzi, Renew. Sustain. Energy Rev., 2010, 14, 1591–1599.
19. S. J. Kim, Y. Kwon, D. Kim, H. Park, Y. H. Cho, S. E. Nam and Y. I. Park, Membranes, 2021, 11(482), 1–34.
20. A. Hakim, T. S. Marliza, N. M. Abu Tahari, R. W. N. W. Isahak, R. M. Yusop, W. M. Mohamed Hisham and A. M. Yarmo, Ind. Eng. Chem. Res., 2016, 55, 7888–7897.
21. B. G. H. Briton, L. Duclaux, Y. Richardson, K. B. Yao, L. Reinert and Y. Soneda, Appl. Water Sci., 2019, 9, 1–14.
22. J. G. Mahy, L. Tasseroul, A. Zubiaur, J. Geens, M. Brisbois, M. Herlitschke, R. Hermann, B. Heinrichs and S. D. Lambert, Microporous Mesoporous Mater., 2014, 197, 164–173.
23. I. Mazilu, C. Ciotonea, A. Chirieac, B. Dragoi, C. Catrinescu, A. Ungureanu, S. Petit, S. Royer and E. Dumitriu, Microporous Mesoporous Mater., 2017, 241, 326–337.
24. A. Barroso-Bogeat, M. Alexandre-Franco, C. Fernández-González and V. Gómez-Serrano, Arab. J. Chem., 2019, 12, 3963–3976.
25. A. Annisa, I. Prasetyo, D. Swantomo and T. Ariyanto, 4th Int. Semin. Chem., 2021, 2349, p. 020007.
26. I. Prasetyo, N. I. F. Mukti, M. Fahrurrozi and T. Ariyanto, ASEAN J. Chem. Eng., 2018, 18, 09–16.
27. B. Lesiak, N. Rangam, P. Jiricek, I. Gordeev, J. Tóth, L. Kövér, M. Mohai and P. Borowicz, Front. Chem., 2019, 7(642), 1–16.
28. V. Tucureanu, M. Alina and M. A. Andrei, Crit. Rev. Anal. Chem., 2016, 46, 502–520.
29. W. Yang, Z. Du, Z. Ma, G. Wang, H. Bai and G. Shao, RSC Adv., 2016, 6, 3942–3950.
30. J. L. Figueiredo, M. F. R. Pereira, M. M. A. Freitas and J. J. M. Órfão, Carbon, 1999, 37, 1379–1389.
31. P. L. Hariani, M. Faizal, R. Ridwan, M. Marsi and D. Setiabudidaya, Sustain. Environ. Res., 2018, 28, 158–164.
32. S. Kamilah, C. Soh, A. Azzura, A. Rahman and M. Shamsuddin, Malaysian J. Anal. Sci., 2018, 22, 768–774.
33. G. Lota, P. Krawczyk, K. Lota and A. Sierczy, J. Solid State Electrochem., 2016, 20, 2857–2864.
34. S. H. Khalil, M. K. Aroua and W. M. A. W. Daud, Chem. Eng. J., 2012, 183, 15–20.
35. W. Liang, Z. Liu, J. Peng, X. Zhou, X. Wang and Z. Li, Energy Fuels, 2019, 33, 493–502.
36. M. Bokare, S. Bano, P. S. Antony, S. Pal and R. Biniwale, Adsorption, 2020, 26, 51–59.
37. J. A. Schott, Z. Wu, S. Dai, M. Li, K. Huang and J. A. Schott, Microporous Mesoporous Mater., 2017, 249, 34–41.
38. S. Shahkarami, A. K. Dalai and J. Soltan, Ind. Eng. Chem. Res., 2016, 55, 5955–5964.
39. S. Hosseini, I. Bayesti, E. Marahel and F. Eghbali, J. Taiwan Inst. Chem. Eng., 2015, 1–9.
40. B. Kim, K. Cho and S. Park, J. Colloid Interface Sci., 2010, 342, 575–578.
41. D. Jang and S. Park, Fuel, 2012, 102, 439–444.

---