Facile Preparation of Carbon Nitride-ZnO Hybrid Adsorbent for CO 2 Capture: The Signiﬁcant Role of Amine Source to Metal Oxide Ratio

: The presence of CO 2 in gaseous fuel and feedstock stream of chemical reaction was always considered undesirable. High CO 2 content will decrease quality and heating value of gaseous fuel, such as biohydrogen, which needs a practical approach to remove it. Thus, this work aims to introduce the ﬁrst C 3 N 4 -metal oxide hybrid for the CO 2 cleaning application from a mixture of CO 2 -H 2 gas. The samples were tested for their chemical and physical properties, using ﬁeld emission scanning electron microscopy (FESEM), transmission electron microscopy (TEM), physical adsorption analysis (BET), fourier-transform infrared (FTIR), x-ray diffraction (XRD), and x-ray photoelectron spectroscopy (XPS). The CO 2 capacity test was carried out by means of a breakthrough test at 1 atm and 25 ◦ C using air as a desorption system. Among the samples, amine/metal oxide mass ratio of 2:1 (CNHP500-2(2-1)) showed the best performance of 26.9 wt. % (6.11 mmol/g), with a stable capacity over 6 consecutive cycles. The hybrid sample also showed 3 times better performance than the raw C 3 N 4 . In addition, it was observed that the hydrothermal C 3 N 4 synthesis method demonstrated improved chemical properties and adsorption performance than the conventional dry pyrolysis method. In summary, the performance of hybrid samples depends on the different interactive factors of surface area, pore size and distribution, basicity, concentration of amine precursors, ratio of amines precursors to metal oxide, and framework stability.


Introduction
Proton-exchange membrane fuel cell (PEMFC) technology is considered a promising energy source to meet the growing demand for energy supply [1]. Apart from the use of renewable fuel, its waste, which consists only of water, is environmentally friendly. However, it requires high-purity hydrogen gas fuel with more than 99.97% quality, as stated by the International Organization for Standardization (ISO FDIS14687-2) [2]. Traditional H 2 production from fossil fuels has been able to supply the desired fuels but with a persistent fossil depletion and release of carbon dioxide (CO 2 ) into the environment during its production process. As a result, biohydrogen gas was seen as a greener method of producing hydrogen (H 2 ). To date, raw biohydrogen gas typically consists of H 2 -CO 2 mixture with a CO 2 ratio of 40-50 %. Therefore, the collected gas must first be purified by removing CO 2 by using either solid, liquid, or membrane adsorbents [3][4][5].
Scientists have attempted for years to absorb CO 2 using various solid adsorbents, such as activated carbon, silica, zeolite, and metal-organic frameworks [6][7][8]. Common improvement works are carried out by impregnation of amine additives, such as dimethylamine and polyethylamine (PEI), on their materials, assuming that the adsorbent basicity thus increases the affinity of CO 2 gas to the adsorbent. However, issues, such as low CO 2 adsorption performance, low sample basicity, and sample degradation over repetitive cycles, are still under discussion. This study thus aims to use a material that is readily amine-rich, such as carbon nitride (C 3 N 4 ), as an adsorbent support. C 3 N 4 is a graphene-like organic material with an orderly 2D structure and a high nitrogen content. Apart from that, C 3 N 4 possesses unique properties such as moderate band gap, nitrogen richness, and high thermal and chemical stability, making it a good photocatalyst, electrocatalyst, photoelectrocatalyst, as well as gas adsorbent for energy storage [9][10][11][12]. The materials can also be prepared by using low-cost precursors, such as urea, melamine, and dicyanamide [13].
To date, few studies have been performed to capture CO 2 on C 3 N 4 . However, they showed weak CO 2 adsorption performance due to their low porosity when they exist in bulk structures. Few researchers have tried to improve the performance of C 3 N 4 by enhancing its porosity through a hard template method, apart from impregnating it with an amine-rich source, such as dimethylamine and PEI. However, the performance will eventually decrease due to PEI volatilization at temperatures above 100 • C and sample degradation over continuous cycles [14,15]. In some situations, a high-pressure operation may lead to favorable performance but with a high operating cost [16,17]. It is important to note that several preparation steps are needed for the common hard template process. Thus, preparing a C 3 N 4 sample with a simpler route but still with a high yield and capturing performance is necessary.
On the other hand, few researchers have investigated the role of various metal oxides in the capture of CO 2 , such as lithium oxide, calcium oxide, and magnesium oxide [18,19]. However, apart from the chemical sorption problems, these materials are costly. Zinc oxide is considered to be a favorable candidate for its low cost and easy regeneration of oxides. This is due to the formation of slightly exothermic and stable physisorption states between the CO 2 and polar ZnO surfaces by a linear adsorption mode [20]. In addition, small binding energies (0.06-0.1 eV) between zinc oxide and CO 2 indicate that the CO 2 adsorption and desorption process can easily occur instead of activating the reaction [20,21]. Past research also showed a positive effect of temperature and pressure on the CO 2 uptake of ZnO adsorbent [22]. To date, the hybridization of ZnO onto C 3 N 4 has not been recorded for CO 2 capture studies. Hence, by a simple hydrothermal and pyrolysis process, we report the production of high-yield carbon nitride and zinc oxide hybrids. Firstly, we concentrated on raw C 3 N 4 and C 3 N 4 -ZnO hybrid development via traditional dry-pyrolysis process. Then, we construct a more stable hybrid sample through the double preparation route with different ratio of amine to metal oxide. In short, we aim to manufacture a high-yield hybrid with high CO 2 capture performance.

Results and Discussion
Initial morphology test was performed by using field emission scanning electron microscopy (FESEM) analysis of sample CNHP500-2(2-1) as shown in Figure 1a,b. The hybrid sample displayed an exfoliated and agglomerated structure of C 3 N 4 with a coralreef-like framework. Sample measurement was recorded as 2 µm. In the meantime, detailed structure analysis through transmission electron microscopy (TEM) shows a dispersion of crystalline lattice clusters in a dimension of 10 nm (Figure 1c). These lattice structures with parallel array framework may indicate good hybridization of ZnO clusters on modified C 3 N 4 , as seen in Figure 1d. Brunauer-Emmett-Teller (BET) technique was conducted to further assess the sample surface area and pore morphology. Figure 2 shows the nitrogen adsorption isotherm and their corresponding pore size distribution curves for sample CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1). Sample CNHP500-2(3-1) displayed the highest surface area of 36.3 m 2 /g, with average pore diameter of 18.1 nm. Meanwhile, samples CNHP500-2(4-1) and CNHP500-2(2-1) displayed a surface area of 18.2 and 12.3 m 2 /g with pore size of 17.4 and 42.5 nm, respectively (Table 1). Although all samples displayed the same multilayer adsorption isotherm with mesoporous saturation point (Type V), each had different pore distribution curves. The pore numbers of sample CNHP500-2(4-1) is higher than sample CNHP500-2(3-1), with similar pore size uniformity. Meanwhile, sample CNHP500-2(2-1) displayed the lowest pore number counts and least uniform pore size distribution. In summary, the optimal amine to metal oxide ratio with the maximum surface area and uniform pore distribution is 3:1.  Brunauer-Emmett-Teller (BET) technique was conducted to further assess the sample surface area and pore morphology. Figure 2 shows the nitrogen adsorption isotherm and their corresponding pore size distribution curves for sample CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1). Sample CNHP500-2(3-1) displayed the highest surface area of 36.3 m 2 /g, with average pore diameter of 18.1 nm. Meanwhile, samples CNHP500-2(4-1) and CNHP500-2(2-1) displayed a surface area of 18.2 and 12.3 m 2 /g with pore size of 17.4 and 42.5 nm, respectively (Table 1). Although all samples displayed the same multilayer adsorption isotherm with mesoporous saturation point (Type V), each had different pore distribution curves. The pore numbers of sample CNHP500-2(4-1) is higher than sample CNHP500-2(3-1), with similar pore size uniformity. Meanwhile, sample CNHP500-2(2-1) displayed the lowest pore number counts and least uniform pore size distribution. In summary, the optimal amine to metal oxide ratio with the maximum surface area and uniform pore distribution is 3:1.
The samples were then characterized by using Fourier transformation infrared (FTIR) spectroscopy ( Figure 3a). As shown by CN500 and CNH500 spectrum, the strong peak at 806 cm −1 can be attributed to the triazine unit vibration of C 3 N 4 ; the peak at 1240, 1327, 1412, and 1465 cm −1 can be assigned to aromatic C-N heterocyclic stretching mode [23]; meanwhile, the peak at 1568 and 1635 cm −1 represents the stretching mode of C-N and C=N [24]. The broad band between 3000-3674 cm -1 can be correlated to the terminal amine group (-NH 2 or =NH) and hydroxyl (O-H) vibration [13]. For CNHP500-1, the spectrum resembles the CN500 band but in a weakened intensity, revealing overlapping bands of C 3 N 4 and ZnO. Both CNHP500-1 and CNHP500-2(2-1) had a weak peak around 2200 cm −1 , suggesting that modification with zinc acetate at 2:1 ratio promotes nitrile bond C≡N formation, as detailed by Deng and co-workers [14]. Since CNHP500-2 preparation was modified through a hydrothermal synthesis, the typical spectrum of C 3 N 4 was again reserved as seen in Figure 3b. This was due to the universal solvent properties of water that dissolve all polar reactants during preparation steps. All CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1) samples exhibited the same functional group properties, with peak intensity decreases in the order of decreasing amine to metal oxide ratio. It may also be noted that an amine to metal oxide ratio higher than 2:1 does not promote nitrile formation. The samples were then characterized by using Fourier transformation infrared (FTIR) spectroscopy ( Figure 3a). As shown by CN500 and CNH500 spectrum, the strong peak at 806 cm −1 can be attributed to the triazine unit vibration of C3N4; the peak at 1240, 1327, 1412, and 1465 cm −1 can be assigned to aromatic C-N heterocyclic stretching mode [23]; meanwhile, the peak at 1568 and 1635 cm −1 represents the stretching mode of C-N and C=N [24]. The broad band between 3000-3674 cm -1 can be correlated to the terminal amine group (-NH2 or =NH) and hydroxyl (O-H) vibration [13]. For CNHP500-1, the spectrum resembles the CN500 band but in a weakened intensity, revealing overlapping bands of C3N4 and ZnO. Both CNHP500-1 and CNHP500-2(2-1) had a weak peak around 2200 cm −1 , suggesting that modification with zinc acetate at 2:1 ratio promotes nitrile bond C≡N formation, as detailed by Deng and co-workers [14]. Since CNHP500-2 preparation was modified through a hydrothermal synthesis, the typical spectrum of C3N4 was again reserved as seen in Figure 3(b). This was due to the universal solvent properties of water that dissolve all polar reactants during preparation steps. All CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1) samples exhibited the same functional group properties, with peak intensity decreases in the order of decreasing amine to metal oxide ratio. It may also be noted that an amine to metal oxide ratio higher than 2:1 does not promote nitrile formation.

Sample
BET Surface Area (m 2 /g) Pore Size (nm) Powder X-ray diffraction (XRD) analysis was done to study the sample crystalline properties (Figure 4). The XRD patterns of sample CN500, CNH500, CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1) showed a sharp peak at 27.4 • , 27.2 • , 27.7 • , 27.7 • , and 27.2 • , respectively. It can be assigned as the turbo static ordering of carbon and nitrogen atoms (002) in CN-graphene like layers with a uniform distribution of N atoms. While Niu and co-workers suggested a reduction of gallery distance when a peak shift towards higher angle, a shift to the lower angle suggests an improvement in the gallery distance between the basic sheets of g-C 3 N 4 nanosheets due to ZnO incorporation, as shown by sample CNH500 and CNHP500-2(2-1) [25]. A slightly broad peak of CNHP500-2(2-1) suggested a semi-crystalline nature that can be related to the fabrication of amorphous carbon in the sample. For CNHP500-1, a wide and diffusing peak indicate an assembly of a randomly arranged amorphous structure. This can be ascribed to the incorporation of metal oxide caused the interference of crystal growth of C 3 N 4 [26]. Meanwhile, an additional peak at 13.3 • of all samples indicates the in-plane structural packing motif indexed as (100) phase of g-C 3 N 4 [27]. Additionally, the low-intensity diffraction peaks related to ZnO were detected at 32, 34, and 36 • . However, the peaks were not visible in sample CNHP500-2(3-1), suggesting smaller crystallite size or highly dispersed of ZnO particles on the C 3 N 4 mesoporous surface.
Powder X-ray diffraction (XRD) analysis was done to study the sample crystalline properties (Figure 4). The XRD patterns of sample CN500, CNH500, CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1) showed a sharp peak at 27.4°, 27.2°, 27.7°, 27.7°, and 27.2°, respectively. It can be assigned as the turbo static ordering of carbon and nitrogen atoms (002) in CN-graphene like layers with a uniform distribution of N atoms. While Niu and co-workers suggested a reduction of gallery distance when a peak shift towards higher angle, a shift to the lower angle suggests an improvement in the gallery distance between the basic sheets of g-C3N4 nanosheets due to ZnO incorporation, as shown by sample CNH500 and CNHP500-2(2-1) [25]. A slightly broad peak of CNHP500-2(2-1) suggested a semi-crystalline nature that can be related to the fabrication of amorphous carbon in the sample. For CNHP500-1, a wide and diffusing peak indicate an assembly of a randomly arranged amorphous structure. This can be ascribed to the incorporation of metal oxide caused the interference of crystal growth of C3N4 [26]. Meanwhile, an additional peak at 13.3° of all samples indicates the in-plane structural packing motif indexed as (100) phase of g-C3N4 [27]. Additionally, the low-intensity diffraction peaks related to ZnO were detected at 32, 34, and 36°. However, the peaks were not visible in sample CNHP500-2(3-1), suggesting smaller crystallite size or highly dispersed of ZnO particles on the C3N4 mesoporous surface.  The X-ray photoelectron spectroscopy (XPS) analysis was further carried out to analyse the elemental composition of sample. From Figure 5, it can be seen that both CN500 and CNH500 displayed a typical spectrum of carbon nitride, with peaks of C 1s and N 1s at 285 and 396 eV, respectively. Meanwhile, the survey spectra of CNHP500-1 and CNHP500-2(2-1) showed obvious signals from elements C, N, O, and Zn, indicating the The X-ray photoelectron spectroscopy (XPS) analysis was further carried out to analyse the elemental composition of sample. From Figure 5, it can be seen that both CN500 and CNH500 displayed a typical spectrum of carbon nitride, with peaks of C 1s and N 1s at 285 and 396 eV, respectively. Meanwhile, the survey spectra of CNHP500-1 and CNHP500-2(2-1) showed obvious signals from elements C, N, O, and Zn, indicating the successful formation of C 3 N 4 -ZnO hybrid. It proves that the second synthesis method is better than the first method given the differences in intensity. The C 1s deconvolution spectrum of CNHP500-2(2-1) represents the C-C and C-N interaction at the binding energy of 284.5 and 288.2 eV [28]. Meanwhile, the peak of O 1s at 529.7 eV signifies the Zn-O interaction [29]. For N 1s, the peak at 397 eV indicates the formation of pyridinic-N structure [13] as the best basic site for CO 2 interaction. The pyridinic-N is considered the most favorable site for CO 2 binding owing to its high Lewis basic characteristic [30]. The formation of Zn 2+ , with peaks at 1021.7 and 1044.7 eV, denotes the peak of Zn 2p 3/2 and Zn 2p 1/2 [31]. The elemental composition of CNHP500-2(2-1) sample (  For the final analysis, the adsorbent samples were further tested for CO 2 adsorptiondesorption activity at low temperature and pressure. Figure 6 represents the adsorption curves (a), desorption curves (b), and CO 2 capture capacity (c) for all samples. In the meantime, Figure 6d represents the adsorption capacity of sample CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1) for three consecutive cycles. Based on adsorption and desorption curves, the samples CNHP500-2(3-1) and CNHP500-2(4-1) exhibited a longer time for CO 2 adsorption compared to other samples, but steeper curves of desorption were shown, suggesting a more complex and time-consuming desorption process. On the other hand, CNHP500-2(2-1) and CNHP500-1 samples demonstrated a rapid and easy desorption of CO 2 , which is beneficial for efficient adsorption cycle.
In terms of capture performance, sample CNHP500-2(2-1) displayed the highest adsorption capacity of 26.9 wt. %. The sample displayed 3-fold higher CO 2 adsorption capacity when compared to CNHP500-1, although both samples were prepared with similar amine to metal oxide ratio. This shows the importance of water as a universal solvent during hydrothermal synthesis in contrast to the conventional dry-pyrolysis method. Meanwhile, the performance of hybrid samples with different amine to metal oxide ratio displayed an increasing adsorption capacity in the order of CNHP500-2(4-1) < CNHP500-2(3-1) < CNHP500-2(2-1). This can be attributed to the increasing numbers of metal oxide per active site in the sample. Theoretically, C 3 N 4 linkages are good for CO 2 capture because they provide the required basic site for acidic CO 2 adsorption activities. Recent analysis of density functional theory (DFT) between CO 2 and C 3 N 4 materials shows that the weak van der Waals bond form by abundantly accessible pyridinic-N structure of C 3 N 4 has been favorable for CO 2 interaction [32]. On the other hand, metal oxide ZnO can further enhance the intensity of accessible active site by forming low binding strength interaction, as explained by Tang and co-workers [20]. That explained why the hybrid samples performed better than a pure C 3 N 4 (CN500). As for the three-cycle adsorption test in Figure 6d, CNHP500-2(2-1) sample showed a decrease in capture capacity after the first to second cycles and then slight increase towards the third cycle. Furthermore, CHNP500-2(3-1) sample also demonstrated similar performance in the second and third cycles. However, considering the desorption isotherms in Figure 6b, CNHP500-2(2-1) sample demonstrated far more rapid and easier desorption process for CO 2 compared to CNHP500-2(3-1), suggesting better adsorbent characteristics. Further raising the ratio to 4:1 would not be advantageous for the capture performance, as the low CO 2 capture capacities of CNHP500-2(4-1) sample were shown in all three consecutive cycles. This can be explained by the reduced amount of ZnO on the adsorbent surface as the ratio increases, thus lessening the promotional effects of ZnO. For the final analysis, the adsorbent samples were further tested for CO2 adsorptiondesorption activity at low temperature and pressure. Figure 6 represents the adsorption curves (a), desorption curves (b), and CO2 capture capacity (c) for all samples. In the meantime, Figure 6(d) represents the adsorption capacity of sample CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1) for three consecutive cycles. Based on adsorption and desorption curves, the samples CNHP500-2(3-1) and CNHP500-2(4-1) exhibited a longer time for CO2 adsorption compared to other samples, but steeper curves of desorption were shown, suggesting a more complex and time-consuming desorption process. On the other hand, CNHP500-2(2-1) and CNHP500-1 samples demonstrated a rapid and easy desorption of CO2, which is beneficial for efficient adsorption cycle. The effect of melamine and urea combination in C3N4 synthesis can be seen in sample CNH500. Compared to a pure melamine-based C3N4 (CN500), the sample displayed a better adsorption capacity, as in Figure 6(c). Similar to a study by Zheng and co-workers, the incorporation of urea created a disturbance during C3N4 formation, thus providing a higher sheet gap and a better surface area [33][34][35][36]. To note, all hybrid samples in this study applied the combination of melamine and urea as the amine source to boost the sheet gap and surface contact area. Figure 6. (a) Adsorption and (b) desorption isotherm of sample CN500, CNH500, CNHP500-1, CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1). (c) CO 2 capture capacity of different samples at 50 ml/min and room condition (1 atm, 25 • C). (d) CO 2 capture capacity of sample CNHP500-2(4-1), CNHP500-2(3-1), and CNHP500-2(2-1) for 3 continuous cycles.
The effect of melamine and urea combination in C 3 N 4 synthesis can be seen in sample CNH500. Compared to a pure melamine-based C 3 N 4 (CN500), the sample displayed a better adsorption capacity, as in Figure 6c. Similar to a study by Zheng and co-workers, the incorporation of urea created a disturbance during C 3 N 4 formation, thus providing a higher sheet gap and a better surface area [33][34][35][36]. To note, all hybrid samples in this study applied the combination of melamine and urea as the amine source to boost the sheet gap and surface contact area.
To further comprehend the stability of the best sample, CNHP500-2(2-1) was tested for adsorption-desorption activity over six continuous cycles. Based on Figure 7a, the sample showed a rapid overall cycle with a complete cycle measured as 20 minutes each, indicating a quick adsorption-desorption process. Figure 7b displayed a sharp and consistent adsorption isotherm of sample except at the first cycle. In terms of desorption activity, the result from Figure 7c clearly showed an excellent profile consistency for over six cycles. Hence, fast desorption activity confirms the absence of major clogging issue. Based on Figure 7d, the sample displayed high and consistent CO 2 adsorption capacity, which indicated a low material degradation rate. Meanwhile, a slight decline in CO 2 capacity for the second cycle afterwards can be explained by several factors, such as the pore heterogeneity or arrangement and minor chemical adsorption. In terms of moisture consequence on CO 2 capture performance, the durability of the sample was proven through the usage of air as desorbing gas. Although there is a little percentage of moisture in the air, the sample still displayed a stable CO 2 capacity.
Based on Figure 7d, the sample displayed high and consistent CO2 adsorption capacity, which indicated a low material degradation rate. Meanwhile, a slight decline in CO2 capacity for the second cycle afterwards can be explained by several factors, such as the pore heterogeneity or arrangement and minor chemical adsorption. In terms of moisture consequence on CO2 capture performance, the durability of the sample was proven through the usage of air as desorbing gas. Although there is a little percentage of moisture in the air, the sample still displayed a stable CO2 capacity.

Chemical and Reagents
Reagent grade melamine powder was purchased from Sigma-Aldrich, Saint Louis, MO, USA. Meanwhile, urea and zinc acetate powder were purchased from R&M Chemicals (London, UK). CO2-H2 gas mixture with vol. % of 50:50 was supplied by Linde Malaysia Sdn Bhd, Selangor, Malaysia. All materials were used without further treatment.

Chemical and Reagents
Reagent grade melamine powder was purchased from Sigma-Aldrich, Saint Louis, MO, USA. Meanwhile, urea and zinc acetate powder were purchased from R&M Chemicals (London, UK). CO 2 -H 2 gas mixture with vol. % of 50:50 was supplied by Linde Malaysia Sdn Bhd, Selangor, Malaysia. All materials were used without further treatment.

Synthesis of Adsorbents
A carbon nitride hybrid with porogen sample, namely CNHP500-1, was prepared via a simple dry in-situ pyrolysis method (1st method). Initially, amine sources (urea and melamine) and zinc acetate were placed into a closed-lid crucible and underwent pyrolysis at 500 • C for 1 h. The sample was left to cool at room temperature for 24 h before being ground into fine powder in a mortar. Meanwhile, the sample hybrid CNHP500-2(x-y), where x-y represent the amine to metal oxide mass ratio, was prepared through a facile hydrothermal and pyrolysis method (2nd method). Initially, appropriate amounts of the amine sources (urea and melamine) and zinc acetate were added with water and stirred for 24 h. The resulting solution was sonicated and dried for 24 h. The dried sample was then pyrolyzed at 500 • C for 1 h under closed crucible. The sample was left to cool for 24 h before being ground into powder. The ratios of amine to metal oxide of 2:1, 3:1, and 4:1 were denoted as CNHP500-2(2-1), CNHP500-2(3-1), and CNHP500-2(4-1), respectively. For comparison, bulk C 3 N 4 namely CN500 was prepared through the conventional drypyrolysis of melamine powder at 500 • C for 1 h. Another bulk C 3 N 4 sample was prepared through direct pyrolysis of melamine-urea mixture in 1:1 mass ratio at 500 • C for 1 h and was labelled as CNH500. Table 3 summarizes the samples synthesized in this study. Hydrothermal and dry pyrolysis Melamine and urea 2:1 CNHP500-2  Hydrothermal and dry pyrolysis Melamine and urea 3:1 CNHP500-2(4-1) Hydrothermal and dry pyrolysis Melamine and urea 4:1

Characterization
To view the sample morphology, field emission scanning electron microscopy (FESEM) and transmission electron microscopy (TEM) was carried out using SEM ZEISS Supra 55 VP (Oberkochen, Germany) and TEM Thermo Fischer (Model Talos 120C) (OR, USA), respectively. Textural analysis was conducted using Micromeritics ASAP 2020 (Norcross, GA, USA) with Brunauer-Emmett-Teller (BET) model was used to determine the specific surface area. Bulk Scientific M, Perkin Elmer (Waltham, MA, USA) Fourier-transform infrared (FTIR) spectrophotometer was used to analyze the functional groups. FTIR ATR sampling technique was applied while the sample used, as it is in powder form. Sample crystallinity study was conducted via X-ray diffraction (XRD) by Bruker AXS D8 Advance diffractometer (Bremen, Germany) operating at 40 kV and 40 mA using Ni filtered Cu Kα source (λ = 0.15406 nm), and XRD patterns were measured between 2θ = 5 -80 • at a scan rate 0.25 • s −1 . Meanwhile, X-ray photoelectron spectroscopy (XPS) analysis was employed on Kratos/Shimadzu Axis Ultra DLD (Manchester, UK), with Al Kα (1486 eV) monochromatic X-ray as the excitation source.

CO 2 Capture Test
The CO 2 capture test was done through a breakthrough experiment, as in previous study [37]. In summary, the sample was packed into a column before being connected to gasflowmeter-sample-CO 2 analyzer system. A CO 2 analyzer from Quantek Instrument Model 906 (Grafton, MA, USA) was used with HOBO software (HOBOware Pro 3.7.18 version, Onset, Cape Cod, Massachusetts, USA, 2018) as data logger. All tests were conducted on a fixed flowrate of 50 mL/min at room temperature and pressure.

Conclusions
To conclude, the C 3 N 4 -ZnO organic-inorganic hybrid, namely CNHP500-2(2-1), was successfully prepared via a facile hydrothermal and pyrolysis route. The simplistic fabrication of this sample, with low cost with high yield, high CO 2 capture capacity, and high sample stability, qualify it as a prospective adsorbent for future CO 2 removal application. To highlight, past C 3 N 4 studies for CO 2 capture application favored a high surface area sample via hard template synthesis to improve the mass transport of gas. Through this study, a low surface area C 3 N 4 with optimal ratio of metal oxide hybridisation, a disturbed sheet of N heterogens, and favorable bond strength between adsorbent and adsorbate was able to display a comparable CO 2 capture performance. Thus, the synthesis of abundantly graphitic-N C 3 N 4 with ZnO hybridization is an interesting and promising study that may be conducted in the near future. Above all, we hope this work can open a new path to the development of a versatile adsorbent for CO 2 capture soon.