Pore Characteristics for Efficient CO 2 Storage in Hydrated Carbons

Development of new approaches for carbon dioxide (CO 2 ) capture is important in both scientific and technological aspects. One of the emerging methods in CO 2 capture research is based on the use of gas-hydrate crystallization in confined porous media. Pore dimensions and surface functionality of the pores play important roles in the efficiency of CO 2 capture. In this report, we summarize work on several porous carbons (PCs) that differ in pore dimensions that range from supermicropores to mesopores, as well as surfaces ranging from hydrophilic to hydrophobic. Water was imbibed into the PCs and the CO 2 uptake performance, in dry and hydrated forms, was determined at pressures of up to 54 bar to reveal the influence of pore characteristics on the efficiency of CO 2 capture and storage. The final hydrated carbon materials had H 2 O to carbon weight ratios of 1.5:1. Upon CO 2 capture, the H 2 O/CO 2 molar ratio was found to be as low as 1.8, which indicates a far greater CO 2 capture capacity in hydrated PCs than ordinarily seen in CO 2 −hydrate formations, wherein the H 2 O/CO 2 ratio is 5.72. Our mechanistic proposal for attainment of such a low H 2 O/CO 2 ratio within the PCs is based on the finding that most of the CO 2 is captured in gaseous form within micropores of diameter < 2 nm, wherein it is blocked by external CO 2 -hydrate formations generated in the larger mesopores. Therefore, in order to have efficient high pressure CO 2 capture by this mechanism, it is necessary to have PCs with a wide pore size distribution consisting of both micropores and mesopores. Furthermore, we found that hydrated microporous or supermicroporous PCs do not show any hysteretic CO 2 uptake behavior, which indicates that CO 2 –hydrates cannot be formed within micropores of diameter 1-2 nm. Alternatively, mesoporous and macroporous carbons can accommodate higher yields of CO 2 -hydrates, which potentially limits the CO 2 uptake capacity in those larger pores to a H 2 O/CO 2 ratio of 5.72. We found that high nitrogen content prevents the formation of CO 2 hydrates presumably due to their destabilization and associated increase in system entropy via stronger noncovalent interactions between the nitrogen functional groups and H 2 O or CO 2 .


INTRODUCTION
Porous materials are important for the capture, separation and conversion of greenhouse gases such as carbon dioxide (CO 2 ). [1][2][3] Among the many existing classes of porous materials, porous carbons (PCs) have received a great deal of attention as adsorbents due to their attractive physical and chemical properties and stability. 4,5 In particular, the prospect of designing PCs with well-defined porosity within the range of supermicroporosity to macroporosity is currently attracting much research effort for potential applications in catalysis, energy storage, gas separations and for environmental remediation and conservation. [6][7][8][9] Additionally, the development of PCs with precise pore dimensions within the sub-nanometer range that can rival porous materials such as zeolites or metal-organic frameworks is of interest. 10,11 Particularly important are micropores and supermicropores, which are relevant to any efforts to physically and selectively trap CO 2 via molecular sieving approaches. Such capture of CO 2 requires narrow pores that are close to the kinetic diameter of molecular CO 2 , i.e., 0.33 nm. 12,13 The molecular sieving approach is especially important for post-combustion CO 2 capture technologies that require lower CO 2 adsorption enthalpy and faster kinetics for release of the CO 2 . Alternatively, larger diameter mesoporous and macroporous carbons have been shown to be amenable to surface modifications that act to enhance CO 2 capture capacity and selectivity via increased CO 2 adsorption enthalpy. [14][15][16] Therefore, various surface modification techniques, such as heteroatom doping to increase Lewis basic sites, [17][18][19] surface oxidation to increase the polarity, 20,21 as well as noncovalent doping using polymeric amines, 15 are an important part of recent research toward development of CO 2 capture technology. Introducing traces of water within porous media has been shown to greatly enhance CO 2 capture efficiency via the formation of CO 2 gas hydrates. 22-24 4 Gas-hydrates are crystalline host-guest compounds consisting of ordered hydrogenbonded porous water clusters that contain gas molecules in a void cavity. [25][26][27][28] A dodecahedral water cluster with CO 2 inside the cavity is shown in Figure 1. The stability of gas-hydrates depends on the strength of the noncovalent interactions between the host water molecules and the guest molecules. Hence, their stability differs significantly upon changing the guest molecules, which can be CO 2 , nitrogen (N 2 ), methane (CH 4 ), hydrogen sulfide (H 2 S) or other gases. 29,30 Therefore, apart from CO 2 capture technologies that are based on the kinetic diameter of CO 2 and strong noncovalent interactions of CO 2 with amines, the gas-hydrate based CO 2 capture processes gives an added dimension for tuning the CO 2 capture capacity and selectivity of porous materials. 31 Gas-hydrates have been effectively employed for methane storage and transportation. 32 CO 2 capture technology using hydrate-based processes is increasingly being explored although the mechanism for adsorption is not well elucidated. 33 Various materials such as porous aluminium, 34 activated carbons, 35,36 silica, 37 as well as metal-organic frameworks, 38,39 have been 5 employed for hydrate-based CO 2 capture processes. In general, there are three forms of gashydrate structures, which are denoted as structure I, structure II and structure H. 40 All three structures have been identified and investigated using X-ray diffraction studies. 40 These three forms differ in the size of their crystallographic unit cell, cavity types, water content and types of gas molecules that they can host within their cavities. CO 2 mainly exists in CO 2 -hydrates in the form of structure I, which has a unit cell of a 1.2 nm cube with 46 molecules of water and two types of cavities. Structure II has larger amounts of water molecules, and therefore has larger unit cell dimensions. Structure H is the least stable, and has the smallest unit cell dimensions, with three different types of cavities that can host various size molecules such as methane and larger chain hydrocarbons. 41 In confined porous media, particularly when the pore diameters approach the sizes of the unit cell of gas-hydrates, the properties and stability of the gashydrates are expected to change and approach the limits of geometrical restrictions governed by the porous structure.

C O O
Herein, we report the CO 2 capture characteristics of hydrated PCs that differ in their pore size distribution and surface functionalities. To explore the effect of pore size distribution on CO 2 uptake, we accordingly selected PCs with diameters ranging from the microporous/supermicroporous regime to predominantly mesoporous. This work builds on our recent preliminary report on gas-hydrate-based CO 2 storage in porous carbon materials, 42 but, via a series of carefully selected samples and experiments, goes much further in more clearly elucidating the effect of the pore dimensions, elemental composition and surface functionalities.
By exploring the CO 2 uptake performance of several PCs in their dry and hydrated forms, we are able to discuss mechanistic details and the role of pore dimensions and surface functionalities on high-pressure CO 2 storage via CO 2 -hydrate formation. 6

Micro-/Mesoporous Carbon Characterization.
PCs were selected that have pore size distribution within different size ranges, namely, narrow distribution of ultra-microporosity, mixed pore size distributions within micro-and mesoporosity, and predominantly mesoporous.
Details of sample preparation procedures and associated characterization data, including analysis for morphology of the PCs have been previously reported. 9,[42][43][44][45] Herein, we report on selected and additional data on the pore structure, surface composition and elemental composition of the PCs, which are relevant to the hydrate-based CO 2 capture process. Details of pore structure, such as porosity, Brunauer-Emmett-Teller (BET) surface area, pore volume, relative content of microporosity and the pore size distribution (PSD) of the PCs were analyzed by means of nitrogen sorption (at 77 K), and the isotherms are shown in Figure 2, while the corresponding PSDs are shown in Figure 3. As seen from Figure 2, both samples G-800 and G-850-5 exhibit Type I isotherm, with the major adsorption at low relative pressures, i.e., < 0.05, which is characteristic of microporous adsorbents. 46 The isotherm of sample G-2.3-2 shows some deviation from Type I, and indicates the presence of some mesopores. The PSDs in Figure 3 reveal that sample G-2.3-2 has some pores of size >2 nm, while on the other hand, both sample G-800 and G-850-5 do not exhibit any pores of size > 2 nm. This is a significant observation in the context of the present study especially given that, despite the differences in their PSD, samples G-800, G-850-5 and G-2.3-2 have comparable surface area (Table 1).   which represents a porous carbon with a mixture of microporous and mesoporous characteristics, along with high BET surface area of 2860 m 2 g -1 and pore volume of 1.65 cm 3 g -1 . uGilT has 9 73% of micropore volume ( Table 1). The nitrogen sorption isotherm for uGilT is intermediate between type I and IV, and the sample exhibits the narrowest mesopore size distribution (centered at ca. 2.5 nm) of all the studied materials ( Figure 2 and 3).
The surface characteristics of the PCs were studied by means of X-ray photoelectron spectroscopy (XPS). Table 1  1.8% for superhydrophobic surfaces, to 15% for superhydrophilic surfaces. 47 Sample G-800 has the highest oxygen content and is therefore expected to be the most hydrophilic, with a C/O ratio of 9.5. In comparison, sample G-850-5 has C/O ratio of 37.5 (Table 2) and is expected to be hydrophobic. We highlight the O content and hydrophilicity or hydrophobicity of these two samples (G-850-5 and G-800) as they have comparable pore structure and surface area.
Deconvoluted peaks of high resolution XPS spectra (C 1s and O 1s) of the PCs are shown in  Table 2. Table 2. Elemental composition of porous carbons estimated from the C 1s and O1s peaks of high resolution XPS.  Table 2. In general, the PCs show higher content of C─O groups, except for G-850-5, which has higher proportion of C=O groups and the lowest oxygen content. It is also noteworthy that G-2.3-2 and G-2.7-2 have higher content of nitrogen functionalities ( Table 1). The N 1s spectra for both were deconvoluted into two main peaks, with binding energies at ~398.5 and ~401.1 eV, and which correspond to pyridinic N and pyrrolic N-bonding configurations, respectively ( Figures S8 and S9). The effect of each surface bonding configuration in combination with the pore structure were further evaluated with respect to the CO 2 -sorption properties of the hydrated PCs.
12 Figure 4. C 1s and O 1s X-ray photoelectron spectra of PCs.

CO 2 -Uptake Characteristics of the Hydrated Micro-/Mesoporous Carbons.
Water containing (i.e., hydrated) PCs were prepared as previously reported. 42 The hydrated PCs were prepared to a target water content of 150 wt%, and are hereinafter denoted as PC(H 2 O) (see Experimental Section for details). It is also noteworthy to mention that upon hydration of PCs BET surface area drastically decrease, revealing the enrichment of the pores by water. 42 The Comparison of the CO 2 sorption isotherm for dry uGilT to that of hydrated uGilT(H 2 O) indicates 14 that the total CO 2 uptake capacity (at pressure of 54 bar) is similar for both samples at 30 mmol It is important to note that uGilT has a micropore volume content of 73% with the remainder of pore volume presumably arising from mesopores. Using in situ IR-spectroscopy, we have previously shown that hysteretic behavior of the CO 2 -sorption isotherm is due to the formation of CO 2 -hydrates that start to form at ~ 20 bar in the confined pore spaces of uGilT(H 2 O). 42    shifted to higher pressures. The other feature of the CO 2 sorption on hydrated PCs that was 18 explored is the sample to sample variation in the hysteresis width, and its relationship with the proportion of microporosity. Figure 7 plots the relationship between micropore % and the hysteresis width (bar), as well as micropore % + 3×N% (nitrogen content) and the hysteresis width (bar). (The dashed line represents a linear relationship). Without incorporation of the N content, the relationship is rather scattered, as can be seen in Figure 7a. In Figure 7 has the smallest hysteresis width, along with micropore content of 73% and negligible N content. Thus, in order to achieve even smaller hysteresis width according to the linear relationship shown in Figure 7, we would suggest a PC with microporosity in the range of 73% -90%. This suggestion is in accord with the mechanism proposed in this work, which is based on gas phase adsorption of CO 2 in micropores that are 'blocked in' by formation of CO 2 -hydrates within the 'outer' meso-and macropores. Therefore, it is beneficial to have mixed micro-and mesopore distribution in PCs in order to achieve efficient CO 2 -capture based on the CO 2 -hydrate formation. Importantly, it is noteworthy that for conventional CO 2 -hydrates, the CO 2 ×nH 2 O ratio, n is 5.72. [25][26][27][28][29][30] However, in our case, this ratio is considerably smaller (n = 1.8 and n = 2.8, depending on temperature). 42 Such a small ratio shows that PC(H 2 O) samples adsorb much more 19 CO 2 within the pores to an extent that is structurally impractical to form for conventional CO 2hydrates. Figure 7. The diagram representing the direct comparison of the relationship between the CO 2hydrate formation hysteresis width and the micropore percentage A) before and B) after the inclusion of nitrogen content percentage.
We have previously described the pressure and temperature dependence of the hysteretic CO 2 sorption isotherm and their relevance to the CO 2 -hydrate formation process. 42 Here, we explore the relationship between pressure of the CO 2 -hydrate formation (i.e. pressure at the beginning of the formation of hysteresis loop) and temperature in the range of 258-323 K so as to generate a phase-equilibrium diagram for CO 2 -hydrate formation within uGilT(H 2 O) as shown in Figure 8. 42 The pressure-temperature range of stability for CO 2 -hydrate formation within uGilT(H 2 O) was compared with the phase-diagrams of pure CO 2 gas-liquid transition, and pure CO 2 -hydrate gas-solid and liquid-solid transitions ( Figure 8). 51,52 As shown in Figure 8

CONCLUSIONS
In summary, detailed analysis and characterization of the pore structure and surface functionalities of PCs, and their effect on CO 2 -uptake performance in the presence (hydrated) or 21 absence (dry) of water has revealed the importance of pore size and pore size distribution as well as the nitrogen content on the adsorption process. Supermicroporous carbon, with pores < 2 nm, Characterization. The XPS analyses were obtained on a PHI Quantera SXM scanning X-ray microprobe system using a 100 µm X-ray beam with take-off angle of 45° and pass energy of  24 equipped with an attenuated total reflectance system (Nicolet, Smart Golden Gate) and a MCT-A detector.