Zinc-Decorated and Nitrogen-Functionalized Hierarchical Porous Carbons for Carbon Dioxide Capture

Abstract

This study developed a highly facile method to synthesize Zn-decorated and nitrogen-doped hierarchical porous carbons for carbon dioxide (CO₂) adsorption. Zeolitic imidazolate framework-8 (ZIF-8) was used as the raw material, which was subjected to a thermal treatment to obtain ZIF-8-derived carbons (ZDCs) in order to develop nanocarbons with a stable framework structure, a high CO₂ adsorption capacity, and high selectivity under normal pressure. The crystallinity evolution of the samples changed from the typical ZIF-8 structure to having features of graphite carbons upon heating. The average particle sizes of the products were between 34 and 105 nm, and the specific surface areas ranged from 618 to 1862 m²/g. The nitrogen and zinc contents gradually decreased with increasing carbonization temperatures, but the changes in the distributions of the functional groups were different. The interactions between CO₂ and the ZDCs were significantly enhanced, resulting in a higher isosteric heat of adsorption. The ZIF-8 carbonized at 1123 K exhibited the highest CO₂ uptake, i.e., 3.57 mmol/g at 298 K and 101.3 kPa, while higher CO₂ uptakes at 15 kPa occurred on the ZIF-8 carbonized at 923 and 1023 K due to their high isosteric heat of adsorption of CO₂. The higher adsorption selectivity of Z8-650 for CO₂ over N₂ may be due to its higher V_<0.7nm/V_mi ratio and nitrogen and zinc contents. Consequently, the micropore area ratio and surface functional groups primarily determined the CO₂ adsorption capacity at 15 kPa. In addition, an appropriate metal Zn to Zn²⁺ ratio may have a positive effect on CO₂ adsorption. On the other hand, the ultramicropore volume ratio, micropore volume ratio, micropore area, and SSA played more significant roles at 101.3 kPa of pressure.

1. Introduction

One of the effective ways to mitigate climate change is to reduce carbon dioxide (CO₂) emissions. During the transition period to reach a low-carbon society, CO₂ capture may be the most effective strategy to reduce atmospheric carbon dioxide levels. Among them, CO₂ adsorption is known as an available technique to achieve this goal. The synthesis of porous carbon materials using zeolitic imidazolate framework (ZIF) polyhedrons as the raw materials has attracted much attention recently for green energy and environmental applications [1,2,3]. In particular, porous carbons characterized with micropores [4,5] and heteroatom surface functional groups [6,7] have been recommended for CO₂ capture.

As one of the most well-studied metal organic frameworks, ZIFs have received considerable research interest because of their unique features (permanent porosity, high thermal, and chemical stability) and potential applications in the fields of adsorption/separation and electrochemistry [1,8]. Among all the ZIFs, ZIF-8, which possesses a sodalite (SOD) topology that is composed of four- and six-member ring ZnN₄ clusters, is of particular interest due to its high thermal and chemical stability [1,9], characteristic porosity, and large specific surface area (1630–1700 m²/g) [10]. A structural feature of ZIF-8 is that its internal cavities have a diameter of 1.16 nm and are connected through small pores with a diameter of 0.34 nm. It has been reported that the kinetic diameter of nitrogen gas is 0.36 nm [11] and that of CO₂ gas is 0.33 nm [12]. Therefore, it is expected that molecules with different kinetic diameters can be separated using ZIF-8 due to its small apertures [13].

ZIF-8 polyhedrons can be readily manufactured using a variety of synthesis processes in water or organic solvents, such as conventional solvothermal [14], hydrothermal [15], colloidal precipitation [16,17], mechanochemical [18], sonochemical [18], microwave-assisted [18], dry-gel [18], microfluidic [18], electrochemical [19], or soft template [20,21] methods. The particle sizes of the samples prepared by sonochemical and dry-gel routes are significantly smaller than those prepared with other methods [18]. A higher ZIF-8 yield can be obtained using dry-gel and mechanochemical synthesis [18]. The hydrothermal or solvothermal processes are generally conducted with a reaction temperature of room temperature up to 200 °C and a reaction duration of hours to days [22]. The colloidal precipitation route is a simple and widely used approach [16] for the large-scale production of nano- to micro-size ZIF-8 crystals with controlled particle sizes and has been widely used in the literature.

It has been reported that the yield, particle size, crystallinity, and porosity of ZIF-8 are dependent on the molar ratio of imidazole to Zn [23]. For instance, the particle size of ZIF-8 decreased with an increasing molar ratio of imidazole to Zn [24]. Cubic-shaped ZIF-8 is synthesized at low imidazole to Zn molar ratios, whereas at higher imidazole to Zn ratios, rhombic dodecahedron morphologies, or even truncated rhombic dodecahedron morphologies can be obtained [23]. In addition the ratio of imidazole to Zn, the solvent type [25] and the Zn precursor [26] also affect the crystallinity of ZIF-8. In addition to the above-mentioned parameters, the synthesis time also controls the structural evolution of ZIF-8. The first ten minutes are the nucleation stage of ZIF-8. Next, comes the growth phase, during which, it reaches optimal crystallinity after approximately 30–60 min. After that, the process enters a stable phase, and the structure remains unchanged [27].

Although ZIF-8 is considered thermally stable below 723 K in air and below approximately 823 K in N₂ [28], its CO₂ adsorption capacity relative to its specific surface area is not remarkable. Pyrolysis is an efficient and simple strategy to transform ZIF-8 polyhedrons into nitrogen-enriched carbons [29]. Hence, ZIF-8-derived carbons (ZDCs) have been investigated for use in adsorption [2] and electrochemistry applications [30]. Although the high-temperature pyrolysis of ZIF-8 can improve its porosity, it is accompanied by significant decreases in yield and nitrogen and zinc contents [2]. The pore size of ZDCs is determined by the size of the ZIF-8 polyhedrons [30]. It has been reported that a hierarchical pore structure is beneficial to high exposure of active sites and rapid mass transfer [31]. Porous nanocarbons obtained through the direct carbonization of ZIF-8 crystals have been reported to have large adsorption uptakes and fast sensor responses for toluene vapors [16].

The nitrogen functional groups within imidazole ligands play a significant role in the application of ZIF-8 [32]. It was found that as the treatment temperature increased, the nitrogen content of the samples gradually decreased [30]. Compared with the bond dissociation enthalpy of C=C and C=N bonds [33,34], the breakage of C–N bonds occurs more easily. On the other hand, the methyl groups at the terminal end of the imidazole ligand are exposed. Therefore, once ZIF-8 is subject to heat treatment, the cleavage of methyl group from the imidazole ligand and the decomposition of C–N bond would probably take place first [32]. Nevertheless, the overall framework structure of carbonized ZIF-8 can be retained [33]. A systematic study [2] found that the decomposition of imidazole ligands and the evaporation of metal zinc were highly dependent on the carbonization temperature, reaction time, as well as heating rate. Once the processing temperature exceeds 1073 K, the formation of pores is accompanied by the loss of Zn and N atoms and a decrease in yield.

The interactions between CO₂ molecules and ZIF-8 or ZDCs have significant effects on their adsorption of CO₂. The pyrrolic N was found to be highly related to the CO₂ adsorption capacity and the interaction mechanisms between N functional groups and the CO₂ molecules, which could involve acid–base and hydrogen-bonding interactions [6]. ZIF-8 carbonized at 1273 K exhibits an outstanding adsorption capacity for methylene blue compared to those carbonized at 873 and 1473 K as well as ZIF-8 due to the change in the surface charge and pore size distribution [35]. The presence of N dopant on a graphite surface can enhance the chemisorption of O adsorbates due to an enhanced charge-transfer mechanism [36]. Studies have found that under atmospheric pressure, the adsorption of CO₂ on ZIF polyhedrons was restricted [1] and was only possible upon dehydration of the atmospheric gas [37]. Although several studies have tried to use ZIF-8 at high pressures [38,39,40], this approach is not economically viable.

In this study, ZIF-8 polyhedrons were prepared using a simple and rapid colloidal chemistry route. Subsequently, a facile thermal treatment was utilized to prepare ZDCs. The thermodynamic parameters and isosteric heat of adsorption of CO_2, and the adsorption selectivity of the ZDCs for CO₂ over N₂ were evaluated. Moreover, the major factors influencing the CO₂ adsorption capacity under 101.3 kPa were identified. The N₂ adsorption isotherm at 77 K combined with the CO₂ adsorption isotherms at 273 K were used to determine the pore size distributions (PSDs) of the samples. The results showed that this benefits to confirm the existence of ultramicropores on the nanocarbons. The rhombic dodecahedron morphologies were well retained after ZIF-8 was subjected to the heat treatment. This study demonstrated that Zn-decorated and nitrogen-doped hierarchical porous carbons prepared via a facile synthesis route exhibited good CO₂ adsorption capacities under 101.3 kPa and at room temperature. In the literature, there is little discussion on the role of Zn and the changes in Zn functional groups after the thermal treatment of ZIF-8, but this was investigated in this study. Scheme 1 shows a schematic diagram of the research methodology in this study.

2. Materials and Methods

2.1. Chemicals

Methanol (≥99.8%) was obtained from Mallinckrodt Chemicals (Phillipsburg, NJ, USA); zinc nitrate hexahydrate (Zn(NO₃)₂·6H₂O, ≥99%) was provided by J.T. Baker (Radnor, PA, USA); 2-methylimidazole (2-MeIM, 99%) was purchased from Thermo Scientific Chemicals (Fair Lawn, NJ, USA); and cetyltrimethylammonium bromide (CTAB, 98%) was provided by Alfa Aesar (Sisli-Istanbul, Turkey).

2.2. Synthesis and Carbonization of ZIF-8 Polyhedrons

ZIF-8 polyhedrons were prepared using a simple and rapid colloidal chemistry route [17,41]. Using the typical process, Zn(NO₃)₂·6H₂O (0.1 M) and CTAB were dissolved in methanol such that CTAB/Zn(NO₃)₂·6H₂O = 0.025 mol/mol (solution A), and 2-MeIM (0.4 M) was also dissolved in methanol (solution B). Solution A was quickly added to solution B under constant stirring. After vigorously stirring for 30 min, the mixed solution was left to stand at room temperature for 24 h [27]. The products were collected and washed with large amounts of methanol to completely remove any excess 2-MeIM and then dried at 353 K. The thermal annealing of the ZIF-8 polyhedrons was conducted by placing 1 g of the ZIF-8 sample in a ceramic boat in a horizontal tubular furnace. The samples were heat-treated at the target temperature (heating rate of 5 °C/min) for 6 h in a N₂ atmosphere. The samples heat treated at 923, 1023, 1123, and 1223 K were denoted as Z8-650, Z8-750, Z8-850, and Z8-950, respectively.

2.3. Characterizations

The morphology and interior structure of the products were examined using field emission scanning electron microscopy (FESEM) and transmission electron microscopy (TEM) using a Hitachi S-4800 microscope (Tokyo, Japan) and Hitachi H-7100 microscope (Tokyo, Japan), respectively. Ultra-high resolution scanning transmission electron microscopy (STEM) was used to observe the chemical elements on the samples using a Hitachi SU8220 STEM-TE detection device combined with a QUANTAX Annular XFlash^® QUAD FQ5060 energy dispersive X-ray spectrometer (Tokyo, Japan). The N₂ adsorption–desorption isotherms were measured at 77 K using an ASAP 2020 accelerated surface area and porosimetry system (Micromeritics, Norcross, GA, USA) to investigate the surface areas and porosities of the products. The CO₂ adsorption isotherms measured at 273 K were also collected using ASAP 2020 to determine the pore size distribution. A powder diffractometer (Rigaku TTRAX III; Tokyo, Japan) equipped with Cu-Kα radiation (λ = 0.15418 nm, 30 kV and 20 mA) was utilized to record the X-ray diffraction (XRD) patterns from 5 to 90° at a rate of 4°/min. The elemental compositions of the products were analyzed using an elemental analyzer (Elementar vario EL cube; Langenselbold, Germany). X-ray photoelectron spectroscopy (XPS) was utilized to identify the elements on the surface as well as the chemical states on the samples. A spectrophotometer (PHI 5000 VersaProbe II, ULVAC-PHI; Kanagawa, Japan) was used to collect the XPS spectra.

2.4. CO₂ Adsorption Experiments

A Micromeritics ASAP 2020 system was used to acquire the CO₂ adsorption isotherms. The adsorption temperatures were set as 298, 313, and 328 K, which was controlled using an ISO controller (Micromeritics). Based on our previous experience [42], the Freundlich equation [43] (Equation (1)) was used for adsorption data fitting. It is an empirical model that assumes heterogeneous adsorption due to the diversity of adsorption active sites, where q_e (mmol/g) is the equilibrium adsorption capacity, K_F [(mmol/g)(1/kPa)^1/n] is the Freundlich adsorption coefficient, P (kPa) is the gas pressure, and n is a constant indicating the isotherm curvature.

$$q_e=K_F{P}^{1/n},$$

(1)

The evaluation of the thermodynamic parameters for adsorption (mainly standard enthalpy change ($∆$H), standard entropy change ($∆$S), and standard Gibbs free energy change ($∆$G)) provide valuable information for understanding the spontaneity of the adsorption mechanisms. The sign of $∆$H determines whether the adsorption process is exothermic (negative) or endothermic (positive). $∆$S is an indication of the orderliness during the adsorption process. If the adsorption process becomes less random, $∆$S is negative. If it is more random, the value is positive. $∆$G is a critical parameter for determining the spontaneity of the adsorption process. If $∆$G < 0, the adsorption process is defined as a spontaneous process. If $∆$G > 0, the process is not spontaneous and not favorable [44,45]. The values of $∆$H and $∆$S were evaluated using Equation (2), where R (=8.314 J/mol/K) is the universal ideal gas constant and T (K) is the adsorption temperature. Specifically, $∆$H and $∆$S were obtained from the slope and the intercept of the regression line from the ln $K_F$ versus 1/T plot. Then, $∆$G was calculated using Equation (3).

$$\ln K_F={\displaystyle \frac{∆S}{R}}-{\displaystyle \frac{∆H}{RT}}$$

(2)

$$∆G=∆H-T∆S$$

(3)

The above thermodynamic properties were evaluated at zero surface coverage [44]. When the adsorbent’s surface is covered with a certain amount of the adsorbate, the isosteric heat of adsorption (Q_st) can be used to measure the energy heterogeneity at the gas–solid interfaces [46]. It can also provide information about the interaction between the adsorbate molecules as well as the affinity between gas molecules and the lattice atoms of the adsorbent [44]. The value of Q_st (kJ/mol) was determined using the Clausius–Clapeyron equation (Equation (4)). The adsorption selectivity for CO₂ over N₂ was evaluated using Equation (5) [47], where q and p denote the adsorption amount and the partial pressure.

$$-{\displaystyle \frac{Q_{st}}{R}}={\displaystyle \frac{d\ln P}{d\frac{1}{T}}}$$

(4)

$$Adsorption selectivity={\displaystyle \frac{q_{C{O}_2}/q_{N_2}}{p_{C{O}_2}/p_{N_2}}}$$

(5)

3. Results and Discussion

3.1. Characterization of ZIF-8 and Its Derived Carbons

In order to understand the morphology and microstructure of ZIF-8 and its derived carbons, FESEM and TEM were performed. Figure 1 presents the FESEM micrographs of the samples. The ZIF-8 crystals showed a rhombic dodecahedron morphology [48,49] with diameters of approximately 50–150 nm. The heat treatment did not change the morphology of ZIF-8. However, it was found that the smallest particles sizes occurred in the Z8-750 sample and the size of the particles gradually increased as the temperature increased. The average particle sizes were 105 ± 16 nm (ZIF-8), 37 ± 5 nm (Z8-650), 34 ± 5 nm (Z8-750), 67 ± 11 nm (Z8-850), and 70 ± 10 nm (Z8-950), which were smaller than those reported by Xiao et al. [29] but similar to the results of Ma et al. [6]. Similar results were observed in the TEM micrographs (Figure 2). The rhombic dodecahedron morphology was found to be maintained well after carbonization. The obvious decrease in particle size in Z8-650 and Z8-750 could be ascribed to the cleavage of methyl groups on the imidazole ligand and the decomposition of C–N bonds. However, the increase in particle size in Z8-850 and Z8-950 was attributed to the reduction of Zn²⁺ into metal Zn with C, which occupied part of the space, as well as the release of Zn vapor, leading to volume expansion and pore structure development. This transition even induced significant shrinkage from the outside to the inside (Figure 2e) [30], which may be due to the release of a large amount of Zn vapor, leading to the formation of internal cavities. The STEM-EDX mapping of ZIF-8, Z8-750, and Z8-950 clearly showed that the elements were distributed evenly, but the Zn content decreased with the increase in treatment temperature (Figure 2f–h).

The porous properties were further confirmed using measurements of the N₂ adsorption–desorption isotherms. Figure 3a displays the N₂ adsorption–desorption isotherms of the samples, revealing typical reversible type I isotherms. Nevertheless, the isotherms of Z8-750 and Z8-850 still displayed a hysteresis loop, indicating the existence of capillary condensation. In addition, the isotherms near P/P₀ = 1 produced a sharp rise, suggesting the existence of interparticle pores between particles. It is interesting that the nitrogen adsorption isotherms of ZIF-8 showed a two-stage jump at P/P₀ = 0.005 and 0.02 (Figure 3b), similar to those reported in the literature [50]. The first jump could be attributed to the gate-opening effect, which is known as the gas-induced rotation of the imidazolate linkers under pressure; the second jump resulted from the reorganization of the nitrogen molecules due to strong electrostatic interactions with the organic framework. Tian et al. [9] found that both the rotation of the imidazole ring and the bending of the methyl group cause the gate opening of ZIF-8. It is reported that the molar ratio of 2-MeIM to Zn²⁺ does not have a significant impact on the gate opening of ZIF-8 [24].

Using the N₂ adsorption isotherms at 77 K and CO₂ adsorption isotherms at 273 K, the PSDs were determined based on the non-local density functional theory (NLDFT) model (i.e., HS-2D-NLDFT). The CO₂ adsorption isotherms of all the samples at 273 K are shown in Figure 3c. The PSD data with/without CO₂ are compared in Figure 3d,e. It can be seen that for samples without ultramicropores (pore size < 0.7 nm), such as ZIF-8, there was no difference. However, if the samples were filled with ultramicropores, incorporating the CO₂ data into the N₂ isotherms was beneficial for determining the distributions in the low pore width range, such as in the isotherms of Z8-850. The PSDs for all the samples are displayed in Figure 3f,g, where another two peaks at about 0.4 and 0.6 nm were revealed by incorporating the CO₂ data.

The ZIF-8 polyhedrons can be transformed into Zn- and N-doped porous nanocarbons through direct carbonization. The existence of ultramicro-, micro-, meso-, and macropores was identified in all the carbonized samples (Figure 3h,i). Z8-650 and Z8-950 possessed relatively higher mesopore volumes, while Z8-750 and Z8-850 exhibited more uniform pore volume distributions. This hierarchical porous structure facilitates gas transport, thereby accelerating subsequent gas adsorption [32]. The surface and pore features of the samples are detailed in

Table 1. Surface and pore characteristics of the samples determined from N₂ and CO₂ adsorption isotherms at 77 K and 273 K, respectively.

Parameter	ZIF-8	Z8-650	Z8-750	Z8-850	Z8-950
SSA α (m2/g)	1862	618	963	1309	1369
Smi β (m2/g)	1750	485	792	1041	973
Vt γ (cm3/g)	1.2186	1.0422	0.8066	1.0153	1.8033
Vma φ (cm3/g)	0.1504	0.0867	0.1325	0.1171	0.4259
Vme η (cm3/g)	0.4213	0.7008	0.2923	0.3810	0.8418
Vmi η (cm3/g)	0.6470	0.2546	0.3818	0.5173	0.5355
V<0.7nm η (cm3/g)	0	0.1772	0.2207	0.2585	0.2795
Smi/SSA	0.94	0.78	0.82	0.79	0.71
Vmi/Vt	0.53	0.24	0.47	0.51	0.30
V<0.7nm/Vt	0	0.17	0.27	0.25	0.15
V<0.7nm/Vmi	0	0.70	0.58	0.50	0.52

^α SSA (specific surface area) was determined using the BET method. ^β S_mi was determined using the t-plot method. ^γ V_t represents the single-point total pore volume at P/P_o ≈ 0.99. ^η V_me, V_mi and V_<0.7nm were determined using the non-local density functional theory (NLDFT) model (HS-2D-NLDFT). ^φ V_ma was obtained by subtraction.

. The specific surface area (SSA) of ZIF-8 (1862 m²/g) was higher than that of the other ZDCs (Figure 3j), while ZIF-8 did not have ultramicropores, which are known as one of the key properties for CO₂ adsorption. Z8-750 and Z8-850 exhibited higher ultramicropore volume ratios (Figure 3k). Moreover, Z8-850 and Z8-750 also had higher S_mi/SSA and V_mi/V_t ratios (Figure 3l), which are indicative of microporosity.

The XRD pattern of the as-prepared ZIF-8 crystals is shown in Figure 4 and was in good agreement with literature reports. With increasing temperatures, the ZIF-8 polyhedrons began to decompose [30]. The XRD patterns of the ZDCs had weak and broad peaks at 2θ = 26.5°, 35.4°, and 43.3°, which correspond to the C(002), ZnO(002), and Zn(101)/C(101) peaks [51,52,53], respectively. In the XRD pattern of Z8-650, no Zn or ZnO peaks were detected. The signal may have been covered by the detection noise. The ZnO (002) peak gradually increased with increasing carbonization temperature. This might be the reason why the particle sizes of carbonized ZIF-8 became larger as the carbonization temperature increased. The elemental compositions of the samples are shown in Figure 5a. Z8-650 had the lowest carbon content, which could be attributed to the decomposition of free methyl groups [2]. As the carbonization temperature increased above 650 °C, the decomposition of imidazolate rings could result in mass loss [2]. For Z8-850, a significant conversion of the Zn centers to crystalline ZnO was expected due to the increase in crystalline ZnO, which is consistent with the XRD results.

To gain a deeper understanding of the products, the elemental composition and valence states present on the samples were investigated using XPS measurements. C 1s, N 1s, O 1s, and Zn 2p peaks were identified in the XPS survey scan spectra. Their atomic ratios are shown in Figure 5b; the changes in the elemental compositions were similar to those observed in the elemental analysis. The ZIF-8 crystals had abundant amounts of N 1s (28.06 at. %) and Zn 2p (9.94 at. %). Since the boiling point of metal Zn is about 907 °C [54] and nitrogen-containing functional groups are unstable upon heating, with increasing temperature, the atomic ratios of N 1s and Zn 2p gradually decreased. Despite this, Zn still remained after ZIF-8 was heat treated at temperatures greater than 900 °C, which could be ascribed to N–Zn interactions [2].

The XPS C 1s spectrum of ZIF-8 (Figure 6a) could be decomposed into three peaks located at 285, 286.1, and 287.6 eV, which were assigned to C-C/C=C, C-N, and C=O bonds, respectively. A decrease in the peak area of C-C/C=C (285 eV) was observed after ZIF-8 was subjected to heat treatment, suggesting the decomposition of free methyl groups. For all ZDCs, the XPS C 1s regions (Figure 6b–e) could be deconvoluted into five peaks: C-C/C=C (285 eV), C-N (286.1 eV), C=N/C-OH (286.6 eV), C=O (287.6 eV), and COOH (290.5 eV) [55,56,57,58,59]. Z8-750 had a higher percent of C=O, while Z8-850 had a relatively more uniform distribution of the functional groups (Figure 6f).

Due to the nitrogen atoms in the organic linker, ZIF-8 showed a symmetrical XPS N 1s peak with a binding energy of approximately 398.4 eV (Figure 7a) [33]. This indicates that the major N groups in ZIF-8 were N=C. We further explored the N functional groups in the ZDCs and found that the left tail of the N 1s peak extended towards higher binding energies. This implies that the N=C bonds decomposed as the carbonization temperature increased. The XPS N 1s peaks of the ZDCs could be divided into, at most, four nitrogen-containing functional groups, including N=C (398.4 eV), O=C-NH (400.5 eV), quaternary or protonated N (401.2 eV), and oxidized species (404.6 eV), as shown in Figure 7b–e [52]. Z8-850 exhibited the highest percent of O=C-NH (pyrrolic nitrogen), followed by Z8-750 and Z8-650, while only a little amount was detected in Z8-950 (Figure 7f).

Curve fitting of the XPS O 1s spectra (Figure 8) showed the existence of up to three surface oxides [56,57,59], including C=O/O=C-N, C-OH, and COOH groups, which were located at 531.5, 532.7, and 533.7 eV, respectively. Similar to the N 1s peak results, ZIF-8 also exhibited a symmetrical XPS O 1s peak, and the C=O/O=C-N groups were the only type of oxygen-containing functional groups. The oxygen content on ZIF-8 increased upon heating. Two other surface oxides, C-OH and -COOH, were observed on all the ZDCs.

There was only one valence state of Zn (Zn²⁺) on ZIF-8, which was confirmed by the XPS Zn 2p peaks at 1022 eV (Zn²⁺ 2p3/2) and 1045 eV (Zn²⁺ 2p1/2) in Figure 9a [25]. As ZIF-8 was carbonized, some Zn²⁺ was converted to metal Zn (Zn⁰) (Figure 9b–e). Although the atomic ratios of Zn decreased after being subjected to the heat treatment, the ratio of metal Zn to Zn²⁺ was almost unchanged in the ZDCs, except for Z8-950 (Figure 9f). This finding suggests that the Zn chemical state is unchanged after thermal treatment between 650 and 850 °C [32], and only the total content decreased. Once the temperature is higher than 950 °C, Zn²⁺ might be reduced to metal Zn with C and may lead to the release of Zn vapor [2].

3.2. CO₂ Adsorption

Figure 10a–c show the CO₂ adsorption isotherms under atmospheric pressure at different temperatures. The CO₂ adsorption performance increased with increasing CO₂ pressure, but decreased with increasing adsorption temperature, confirming that it is an exothermic reaction. The CO₂ adsorption capacities at 298 K and 101.3 kPa (Figure 10d) were ranked as follows: Z8-850 (3.57 mmol/g) > Z8-750 (3.40 mmol/g) > Z8-950 (3.25 mmol/g) > Z8-650 (2.83 mmol/g) > ZIF-8 (0.88 mmol/g). This indicated that although ZIF-8 had the largest SSA, it had the smallest adsorption capacity. Considering a typical untreated flue gas composition, the adsorption performance at 298 K and 15 kPa were recorded (Figure 10e), and the CO₂ uptake performance exhibited a different order: Z8-750 (1.26 mmol/g) > Z8-650 (1.25 mmol/g) > Z8-850 (1.13 mmol/g) > Z8-950 (0.93 mmol/g) > ZIF-8 (0.09 mmol/g). The particle sizes of the samples seems to be related to the amount of CO₂ adsorption at 15 kPa. This means that the CO₂ adsorption capacities of the samples are the result of a compromise between surface structure and surface chemistry. The Freundlich equation works well for curve fitting of the CO₂ adsorption data (Figure 10 and

Table 2. Fitted parameters for the Freundlich equation for CO₂ adsorption isotherms and the thermodynamic properties at zero loading of CO₂ adsorption on the samples.

Sample	Temperature (K)	KF (mmol/g/kPa1/n)	n	$∆$G (kJ/mol)	$∆$H (kJ/mol)	$∆$S (kJ/mol/K)
ZIF-8	298	0.0035	0.84	−5.64	−44.27	−0.130
	313	0.0020	0.80	−3.69
	328	0.0006	0.69	−1.75
Z8-650	298	0.2202	1.79	−16.79	−31.03	−0.048
	313	0.2294	1.96	−16.08
	328	0.1153	1.62	−15.36
Z8-750	298	0.3085	1.93	−16.49	−29.20	−0.043
	313	0.1982	1.82	−15.85
	328	0.1000	1.51	−15.21
Z8-850	298	0.2393	1.66	−15.66	−26.88	−0.038
	313	0.1458	1.54	−15.09
	328	0.0935	1.46	−14.53
Z8-950	298	0.1580	1.53	−14.79	−21.95	−0.024
	313	0.0978	1.42	−14.43
	328	0.0639	1.36	−14.07

). The values of $K_F$ were inversely proportional to the adsorption temperature. The values of n ranged from 0.69 to 1.96, suggesting favorable adsorption on all the ZDCs except for ZIF-8.

Adsorption thermodynamic properties can provide helpful information for understanding the mechanism of the spontaneous adsorption process. Adsorption is a spontaneous process if there is a decrease in the total free energy of the system. Moreover, the more spontaneous the CO₂ adsorption process is, the easier it is for CO₂ to be preferentially adsorbed. The three thermodynamic parameters of CO₂ adsorption on the samples are summarized in Table 2. The $∆$G values of CO₂ adsorption were negative, which indicated that the adsorption process occurs favorably and spontaneously. Despite this, the absolute value of $∆$G for ZIF-8 was the lowest compared with the ZDC samples at the same temperature. Except for ZIF-8, the largest absolute value of $∆$G was observed with Z8-650 at the same temperature, and the absolute value of $∆$G decreased with an increase in the carbonization temperature of the samples, implying that an increase in carbonization temperature is not conducive to CO₂ adsorption. In addition, the absolute value of $∆$G decreased when the adsorption temperature increased, implying that increasing the adsorption temperature is not beneficial to the adsorption process [44].

The values of $∆$H for CO₂ adsorption for all the samples were negative, which confirmed that the adsorption process is exothermic. The magnitude of $∆$H also provides insights into the adsorption mechanism. The absolute values were 21.95 to 31.03 kJ/mol for the ZDCs and 44.27 kJ/mol for ZIF-8, implying that the adsorption of CO₂ on the ZDC samples involved physisorption, while combined (both physical and chemical) CO₂ adsorption occurred on ZIF-8 [45]. The negative values of $∆$S suggested a decrease in the disorder and randomness of the adsorption system during the CO₂ adsorption process. This could be attributed to the decrease in the degrees of freedom of CO₂ molecule movement because the CO₂ molecules were adsorbed on the surface of the adsorbents, making the system more ordered.

As shown in Figure 10f, the Q_st values decreased logarithmically with the CO₂ loading from 0.01 to 1.2 mmol/g, indicating that the adsorption active sites on the material surface were energetically heterogeneous for CO₂ capture [60]. The Qst values on specific CO₂ loadings were ranked as follows: Z8-650 > Z8-750 > Z8-850 > Z8-950 > ZIF-8. This indicates that the interactions between CO₂ and the ZDCs were enhanced compared with ZIF-8 and that the N=C and C=O/O=C-N groups, which are abundant in ZIF-8, might have an insignificant role in CO₂ adsorption. However, the presence of nitrogen and zinc species on carbonized ZIF-8 actually promoted the interactions between CO₂ and the adsorbents. This may be due to the dissociation of ZIF-8 during pyrolysis, which produced more active sites, exposing nitrogen and zinc atoms, and converting them into more active functional groups. Although physical adsorption might be predominant adsorption behavior in this study, under low CO₂ loading, some chemosorption may have occurred on Z8-650 and Z8-750.

In order to evaluate the potential applications of the ZDCs for CO₂ capture from flue gas, the adsorption selectivity for CO₂ over N₂ was investigated. Since the molecular sizes of CO₂ and N₂ are very close, it is relatively difficult to separate them based on kinetics. Therefore, the ability of adsorbents to separate gas molecules that are similar in size is highly dependent on their properties. Z8-650 and Z8-850 were utilized as examples to investigate the adsorption selectivity at 298, 313, and 328 K (Figure 11a). The adsorption selectivity decreased with increasing partial pressure of CO₂ or adsorption temperature. The adsorption selectivity of Z8-650 was higher than that of Z8-850, which can be ascribed to the higher Qst of Z8-650. It is noteworthy that the adsorption selectivity for CO₂ over N₂ of Z8-650 at 298, 313, and 328 K all intersected at a CO₂ partial pressure of 0.5 atm. This implied that at a partial pressure of CO₂ over 0.5 atm, if the interactions between CO₂ and the adsorbents are stronger, the adsorption selectivity for CO₂ over N₂ is higher at higher adsorption temperatures. Simulating the flue gas conditions under atmospheric pressure (15% CO₂ to 85% N₂ mixture) at 298, 313, and 328 K, the adsorption selectivity for CO₂ over N₂ was found to be 25.9, 20.0, and 19.2 for Z8-650 and 14.5, 12.2, and 11.7 for Z8-850, respectively (Figure 11b). The higher adsorption selectivity of Z8-650 for CO₂ over N₂ may be due to its higher V_<0.7nm/V_mi ratio and nitrogen and zinc contents.

Under atmospheric pressure, all the carbonized ZIF-8 samples had superior CO₂ adsorption capacities compared to the as-prepared ZIF-8 polyhedrons. Under a lower pressure (e.g., 15 kPa), Z8-750 displayed the highest CO₂ adsorption capacity, while the CO₂ adsorption on Z8-650 was also unneglectable. These findings demonstrate that the specific surface area and total pore volume of nanomaterials may not be the most crucial factors for CO₂ adsorption. All the interactions between hierarchical porous structures and surface heteroatom functional groups determine the CO₂ adsorption performance. Figure 12 shows the relationships between CO₂ adsorption capacity and some key parameters of the ZDC samples. At 15 kPa (Figure 12a–c), besides the S_mi/SSA ratio, C-C/C=C (%), C=O/O=C-N (%), and O=C-NH (at. %) also enhanced CO₂ adsorption at 298 K. At 313 K, the main parameters were the V_<0.7nm/V_t ratio, C=O (at. %), and COOH (%), while at 328 K, the S_mi/SSA ratio, O=C-NH (at. %), C-C/C=C (%), C=O (%), and C=O/O=C-N (%) were the primary factors. At 101.3 kPa (Figure 12d–f), CO₂ adsorption at 298 K was highly correlated with the V_mi/V_t and V_<0.7nm to V_t ratios, as well as the S_mi and SSA. At 313 K, besides the aforementioned parameters (V_mi/V_t ratio, S_mi and SSA), V_<0.7nm also affect the adsorption of CO₂. However, the V_<0.7nm to V_mi ratio exhibited an inverse relationship, which implied that a higher volume ratio of the pores (between 0.7 and 2 nm) could facilitate CO₂ adsorption at 313 K. The same effect of the V_<0.7nm/V_mi ratio was observed at 328 K; the S_mi and the V_mi/V_t ratio also contributed to adsorption at 328 K. At 15 kPa, the micropore area ratio and surface functional groups primarily determined the CO₂ adsorption capacity. In addition, an appropriate metal Zn to Zn²⁺ ratio may have had a positive effect on CO₂ adsorption. On the other hand, the ultramicropore volume ratio, micropore volume ratio, micropore area, and SSA played more significant roles at 101.3 kPa.

Table 3. Comparison of CO₂ adsorption performance of the materials developed in this study with various ZIF-8-related products from the literature.

Material	SBET (m2/g)	CO2 Conc.	Temp. (K)	CO2 Capacity (mmol/g)	Reference
Z8-850	1309	15 kPa 101.3 kPa	298	1.13 3.57	This study
Z8-750	963	15 kPa 101.3 kPa	298	1.26 3.40	This study
ZIF-8 (773 K, 24 h)	942	0.15 bar 1 bar	298	0.4 1.79	[33]
ZIF-8	1016	1 bar	298	0.60	[61]
ZIF-7-8s	283	1 bar	293	1.44	[62]
ZIF-8	1169	1 bar	298	1.06	[63]
NH2-ZIF-8	886	1 bar	298	1.94	[63]
NH2-ZIF-8/bacterial cellulose foams	455	1 bar	298	1.63	[63]
ZIF-100	595	1 atm	298	0.6	[64]
Deep eutectic solvent-ZIF-8	1147	1 bar	293	3.03	[65]
KOH-activated ZIF-8	2491	1 bar	298	1.72	[66]

summarizes the CO₂ adsorption capacities of ZIF-8-related nanomaterials from the literature and shows that the Z8-850 and Z8-750 materials developed in this study are promising CO₂ adsorbents.

4. Conclusions

The ZIF-8 polyhedrons prepared in this study showed a uniform rhombic dodecahedron morphology, which was maintained after carbonization at temperatures of up to 1223 K. The decomposition of ZIF-8 upon high-temperature pyrolysis might generate significant shrinkage from outside to inside. Using N₂ adsorption isotherms at 77 K and CO₂ adsorption isotherms at 273 K, the pore size distributions of the adsorbents with ultramicropores were determined using the NLDFT method. The ZIF-8 polyhedrons had the largest SSA; however, all the ZDCs had superior CO₂ adsorption performances compared to ZIF-8. It was expected that the dissociation of ZIF-8 during pyrolysis would generate more active sites, exposing nitrogen and zinc atoms, and convert them to more active functional groups. The adsorption process occurs favorably and spontaneously, is exothermic and tends to be orderly. Physical adsorption was the predominant adsorption behavior in this study, but under a low CO₂ loading, some chemosorption might occur on the adsorbents with a high isosteric heat of adsorption. At 15 kPa, Z8-750 exhibited a better CO₂ adsorption performance, while at 101.3 kPa, Z8-850 showed the highest CO₂ adsorption capacity. Strong interactions between CO₂ and the adsorbents resulted in a higher adsorption selectivity for CO₂ over N₂ at higher adsorption temperatures. The CO₂ adsorption capacities of the adsorbents are a compromise between surface structures and surface chemistry.