Adsorption of CO2 on Activated Carbons Prepared by Chemical Activation with Cupric Nitrate.

Activated carbons were prepared from a lignocellulosic material, African palm shells (Elaeis guineensis), by chemical impregnation of the precursor with solutions of 1-7% w/v Cu(NO3)2 at five different concentrations. These were carbonized in a carbon dioxide atmosphere at 1073 K to obtain different carbons. Their textural properties were characterized by nitrogen and carbon dioxide adsorption isotherms in order to evaluate the pore-size distribution. The immersion enthalpies of the activated carbons in benzene, dichloromethane, and water were determined. The CO2 adsorption capacities of the materials at 273 K under low-pressure conditions were also determined. Chemical characterization was performed by mass spectrometry, Fourier transform infrared spectroscopy, and temperature-programmed reduction. With this method of preparation under the concentrations described, activated micro-mesoporous carbons were obtained, with the formation of highly mesoporous solids that favored the process of diffusion of molecules of CO2 into the material. Here, we show that activated carbons were obtained with different textural characteristics: surface Brunauer-Emmett-Teller areas varied between 473 and 1361 m2 g-1 and micropore volume between 0.18 and 0.51 cm3 g-1. The activated carbon with the highest values of textural parameters was ACCu5-1073. Micro-mesoporous solids were obtained with the methodology used. This is important as it may help the entry of CO2 molecules into the pores. The adsorption of CO2 in the materials prepared presented values between 103 and 217 mg CO2 g-1; the values of volume of narrow microporosity obtained were between 0.16 and 0.45 cm3 g-1. The solid with the greatest capacity for adsorption of CO2 and volume of narrow microporosity was ACCu3-1073. The use of these solids is of importance for future practical and industrial applications. The adsorption kinetic of CO2 in the activated carbons prepared with metallic salt of copper is in good accordance with the intraparticle diffusion model, for which diffusion is the rate-limiting step. The adsorption of CO2 in the prepared activated carbons is favorable from the energy and kinetic point of view, as these accompanied by the presence of wide micro-mesoporosity favor the entry of CO2 into the micropores.

Introduction

In recent years, awareness of climate change has been generated in the international community. It promoted the study of technologies to reduce emissions of greenhouse gases such as CO2, CH4, N2O, and O3, among others, which are responsible for the atmospheric contamination level and climate change. The CO2 is generated by the growing energy demands, which makes it difficult to cut back. In the short term, one of the most viable options to reduce its emissions is capture and storage, as agreed by the Intergovernmental Panel on Climate Change (IPCC), which has the ambitious objective of reducing CO2 emissions by 50% by 2050.[1]

For this reason, in the last decades, new solid materials have been developed, aiming at the capture of harmful gases. Within these materials are porous solids in all forms, which have been implemented due to their great versatility. CO2 is produced by different types of anthropogenic activities, reaching up to 400 ppm concentration in the environment.[2]

There are different technologies for the capture of carbon dioxide released during combustion processes; these techniques include absorption, cryogenics, cycles of carbonation–calcination, the use of membranes, and adsorption. The last method has advantages over the others as they have higher selectivity and lower energy consumption and associated costs at production. The adsorbents used must show a porous structure and favorable surface chemistry. Given that CO2 is an acidic gas, the surface of the solid carbonaceous material must be appropriate to increase selectivity toward this contaminant.[3,4] For adsorption of CO2, the solids of interest that are studied at present are activated carbons, zeolites, clays, metal–organic frameworks, and zeolite-like metal–organic frameworks.[5]

Activated carbons are porous solids with great versatility in the process of adsorption of CO2, as a result of both their textural properties and surface chemistry. In this research, we chose as a precursor material for the preparation of activated carbons African palm shells (Elaeis guineensis), a solid residue resulting from processing of the fruit. The African palm shell is a lignocellulosic precursor suitable for the preparation of activated carbon because it has a carbon content of around 60% and is considered an economic and profitable raw material.

The present research shows the differences that occur in the physicochemical characteristics of activated carbons that were prepared by impregnation with Cu(NO3)2 solutions of different concentrations. These salts catalyze the gasification reaction of lignocellulosic material, resulting in the formation of micro–mesoporous solids with a varied pore size. The activation of the lignocellulosic precursor with this salt raises the CO2 adsorption capacity, which is closely related to the porous structure obtained in the materials. This is attributed to the change in the decomposition path of the precursor material because it prevents the formation of tars, which can block the pore structure, promoting the formation of open pores.[6] This allows greater accessibility and reduces the time required for the adsorption process.

Results and Discussion

The preparation yield of the materials was 20.6–29.3%, which is in accordance with the activation yields reported, of 5–40% for the initial dry biomass. To evaluate the effect of the metallic catalyst, it was necessary to identify the factors that can affect the catalytic gasification, which are the structure, the nature of its surface, its surface characteristics, and the gaseous environment.[7,8] The last has a strong influence on the gasification reaction of the precursor material favored by the impregnation by Cu(NO3)2. CO and H2 can inhibit the gasification process and thus affect final yields;[9] the material performance percentage was calculated as the mass loss of precursor for each experimental condition. Figure presents the performance percentage as a function of the activating agent concentration. It can be seen that the values decrease with the increasing concentration of activating agent. The weight loss produced during the impregnation and carbonization processes occurred at low concentrations of salt because the small impregnated amounts could be distributed uniformly in the precursor surface, giving rise to activated carbon. The catalytic action of these agents increases with concentration, and therefore there is a greater amount of material to be removed from the carbonaceous matrix.

Figure 1

Percentage yield based on the concentration of the activating agent.

Reactions that takes place in the surface

Adsorption Isotherms of N2 at 77 K

Figure presents the adsorption isotherms of N2 for the activated carbons with different Cu(NO3)2 salt concentrations at 1073 K. Solid micro–mesoporous structures were obtained, represented by type II isotherms with H4 hysteresis cycles, according to the latest classification of IUPAC.[10] The isotherms present an initial stage of monolayer-multilayer adsorption on the walls of the micropores and mesopores, followed by a change in the nitrogen adsorption tendency; this change is related to the capillary condensation of the adsorbate in the mesopores. It is evidenced by the rise in the textural parameters with the increase in the impregnated salt concentration up to 5%, for instance, the increasing intensity of the attractive forces in these pores and the potential of adsorption.[11] At a concentration of 7%, these parameters were reduced, as evidenced by a decay in nitrogen adsorption capacity in the carbons. At 1 and 3% concentrations, these isotherms are similar, which may be due to the activation at these concentrations having the effect of increasing the textural characteristics compared to the material obtained at 2% concentration of Cu(NO3)2.

Figure 2

Nitrogen adsorption isotherms for the series ACCu-1073.

Table presents the textural characteristics of the prepared activated carbons; the Brunauer–Emmett–Teller (BET) area ranges between 473 and 1361 m2 g–1. The micropore volume and the characteristic energy of adsorption for this series were determined using the Dubinin–Radushkevich model with values of 0.18–0.51 cm3 g–1 and 15.56–19.86 kJ mol–1, respectively. The results show that there is a correlation with the surface area values, as the solids that showed the greatest micropore volume had the highest values of BET area. These results are comparable with those reported in other investigations where activated carbons were prepared from lignocellulosic precursors similarly impregnated with metallic salts through gasification reactions.[12−16]

Table 1

Textural Parameters for Activated Carbons Obtained from the N2 Adsorption Isotherm at 77 K

	adsorption of N2 at 77 K
sample	SBETa (m2 g–1)	VO (cm3 g–1)	Vmeso (cm3 g–1)	V0.99 (cm3 g–1)	EO (kJ mol–1)
ACCu1-1073	679 (C+)	0.26	0.04	0.30	19.8
ACCu2-1073	473 (C+)	0.18	0.05	0.23	15.6
ACCu3-1073	638 (C+)	0.24	0.06	0.30	17.5
ACCu5-1073	1361 (C+)	0.51	0.09	0.60	19.9
ACCu7-1073	758 (C+)	0.33	0.09	0.42	16.4

Corrected value in optimal range p/p0[10]

The differences in the textural characteristics between the solids obtained show that the change in the concentration of activating agent influences the process that generates porosity in the precursor. During the impregnation, the chemical reagent is introduced into the particles of the precursor and produces some hydrolysis reactions that are seen in a loss of weight, in the exit of volatile material, in the weakening of the structure, and in the increase of the elasticity, and also the chemical agent produces the swelling of the particles. At higher impregnation with cupric nitrate, hydrolysis and swelling are accentuated and the agent cannot be distributed uniformly within the particles; although the total pore volume increases, the distribution of pore sizes is more heterogeneous, which may be due to the decrease in textural parameters in the ACCu7-1073 sample with respect to the solid ACCu5-1073, the meso–macroporosity being more important;[17] with the change in concentration, a variation in the textural parameters of the materials is observed, showing a maximum concentration of Cu(NO3)2 7% w/v reducing the surface area and the volume of micropores. The activated carbon with the highest values of surface area and micropore volume corresponds to the sample ACCu5-1073, with values of 1361 m2 g–1 and 0.51 cm3 g–1, respectively.

Figure presents the relation between the surface area and the micropore volume for the carbons according to the concentration of the activating agent; with the increase in the concentration of the activating agent, a variation between the SBET and VO, due to the increase in the Cu(NO3)2 concentration, causes a greater removal of carbon atoms from the matrix, which generates a higher porosity. It is evident that at an impregnating concentration of 5%, these relationships are favored by obtaining the greatest textural parameters.

Figure 3

Development of the micropore volume and surface area depending on activating agent concentration.

We calculated the pore-size distributions (PSDs) of the carbons. Figure shows this distribution for the sample ACCu5-1073, which results in the development of micro–mesoporosity with pores up to 3 nm, using the experimental data obtained from adsorption of N2 at 77 K. For this determination, the density functional theory (DFT) model was used for this sample, obtaining a pore volume of 0.55 cm3 g–1 and a half pore width 0.34 nm, showing the a better fit of the quenched solid DFT (QSDFT) model; this presents advantages for the determination of pore distributions in geometric and chemically carbonaceous materials disordered because it takes into account the effects of roughness and the heterogeneity of the surface as it has been shown in other investigations;[18,19] it is correlated with the textural parameters because this solid is the one with the highest area and pore volume, presenting the combined development of narrow micropores and wide micropores in its structure. The material has the presence of pores >2 nm, which is useful in order to help in faster diffusion of CO2 into the prepared adsorbents.

Figure 4

PSD using the DFT model for activated carbon ACCu5-1073.

Figure presents the fit of the QSDFT and non-linear DFT (NLDFT) models to the ACCu1-1073 and ACCu5-1073 adsorption data. The materials have rough surfaces and are heterogeneous, with a kernel of mixed pores for all samples. It can be seen that, with the average error adjustments for the different models shown in Table , the results show a better fit of the experimental data with the QSDFT model, which describes a system of combined (slit and cylindrical) pores,[20] with an average error percentage of 0.034–0.157% compared to 0.150–0.274% calculated for the NLDFT model.

Figure 5

Experimental isotherm of N2 at 77 K in semilogarithmic scale for samples (a) ACCu1-1073 and (b) ACCu5-1073, next to the isotherm as computed by QSDFT, NLDFT, and for combined (slit/cylindrical) pores.

Table 2

Comparison of the Average Error (%) of Modeled vs Experimental Isotherms for Different Forms of Pores (Cylindrical, Slit, and Combined) for NLDFT (Homogeneous Surface) and QSDFT (Rough or Heterogeneous Surface) Models

NLDFT
sample	slit pores [%]	cylindrical pores [%]	slit/cylindrical pores [%]
ACCu1-1073	0.614	0.374	0.150
ACCu2-1073	0.761	0.416	0.245
ACCu3-1073	0.643	0.723	0.157
ACCu5-1073	0.642	0.192	0.151
ACCu7-1073	0.424	0.370	0.274

FTIR Spectroscopy

Figure presents the Fourier transform infrared (FTIR) spectra for the activated carbons obtained and also the spectrum obtained from the precursor material without any treatment; by comparing this with the prepared solids spectra, the qualitative changes that occur in the material by activating with the different solutions of Cu(NO3)2 can be seen, from which three bands of interest can be distinguished.

Figure 6

FTIR spectra of the precursor and ACCu series at 1073 K.

First, a band of intensity between 900 and 1450 cm–1. In this region, it is difficult to assign the bands with certainty, as there is overlap of C–O stretching of different surface groups. In this area, assignments have been made to the C–O vibration of esters (1100–1250 cm–1), carboxylic acids and cyclic anhydrides (1180–1300 cm–1), lactones (1160–1370 cm–1), ethers (942–1300 cm–1), cyclic ethers (1140 cm–1), and phenolic groups (1180–1220 cm–1).[21−24] There is a second peak around 1580 cm–1 that can be attributed to the polyaromatic C=C vibration of sp2 hybridized carbons; such vibration is increased by the adsorbed surface oxygen. The third peak located between 3100 and 3700 cm–1 is characteristic of the stretching vibration of OH hydroxyl groups of carboxylic acids and phenolic compounds. These results confirm that through the chemical impregnation with Cu(NO3)2 at different concentrations, changes occur in the surface chemistry of the materials with respect to the precursor material, as shown in Figure These changes are indicated by a decrease in the vibrational characteristics of the oxygenated groups of these materials; the content of surface groups is related to the nature of the activating agent and the concentration used in the preparation of the carbons.

Temperature-Programmed Reduction

Figure compares the different temperature-programmed reduction (TPR) profiles of each of the prepared carbons with the reference mass (CuO). It is found that there is an interaction of copper with the carbon surface; such interaction makes the maximum signal for the reduction move toward higher temperature. In addition to the above, it is noted that the interaction of copper with carbon presents two signs of reduction, at low and high temperature, which indicates a weak interaction probably with the species that are especially on the surface, and another, strong interaction, with species that may be trapped within the porous network of such materials. It is important to note that, as the burden of Cu is increased, the signal at low temperature moves lower, suggesting that the greater burden of Cu on the surface makes it easier to reduce. However, at high temperatures, there is no correlation with the burden of Cu. On the other hand, the consumption of hydrogen is similar for all solids. Table indicates that the presence of Cu after the purification stage of the carbons is the same. This allows us to conclude that, regardless of the burden of Cu used, the remaining amount of Cu is the same, but there is a major change in the way it distributed on the new materials. This is directly related to the adsorptive capacity.

Figure 7

Analysis of the TPR of each of the prepared carbons at 1073 K.

Table 3

Consumption of Hydrogen and Presence of Cu in the Activated Carbons Obtained

solid	temperature reduction (±6 K)	total consumption of mmol H2 g–1 (±0.3)	amount of Cu present (mmol H2 g–1)
ACCu1-1073	754	1.7	0.8
	927
ACCu2-1073	740	1.6	0.8
	901
ACCu3-1073	734	1.7	0.8
	933
ACCu5-1073	736	1.8	0.9
	956
ACCu7-1073	712	1.8	0.9
	944

The following reaction takes place

Determination of the Enthalpies of Immersion

Table presents the results of the enthalpies of immersion in water, benzene, and dichloromethane, along with the hydrophobic factor, which was calculated as the ratio between the enthalpy of immersion in benzene and the enthalpy of immersion in water.[25] The enthalpies of immersion in water, benzene, and dichloromethane are exothermic in nature, which is consistent with the surface process that takes place between the solid and the liquid. The immersion of enthalpies in water are between −28.8 and 49.1 J g–1, in benzene between −117.0 and −55.5 J g–1, and in dichloromethane between −117.3 and −50.6 J g–1. These results are comparable with those from other investigations that employed similar methodologies and the use of lignocellulosic precursors from agro-industrial waste in the production of activated carbons.[26−33]

Table 4

Immersion Enthalpies in Water, Benzene, and Dichloromethane as Well as the Hydrophobic Factor of the Activated Carbons Designed

sample	–ΔHimm H20 (J g–1)a	–ΔHimm C6H6 (J g–1)a	–ΔHimm CH2Cl2 (J g–1)a	hydrophobic factor (−ΔHimm C6H6/−ΔHimm H2O)
ACCu1-1073	28.8 ± 0.5	58.0 ± 0.8	80.6 ± 0.2	2.01
ACCu2-1073	33.1 ± 1.0	55.5 ± 0.6	50.6 ± 0.5	1.68
ACCu3-1073	34.0 ± 0.4	67.3 ± 0.7	73.9 ± 1.0	1.98
ACCu5-1073	49.1 ± 1.1	117.0 ± 1.2	117.3 ± 0.5	2.38
ACCu7-1073	40.0 ± 0.1	65.0 ± 1.0	69.3 ± 1.0	1.63

The standard deviations of the immersion calorimetry are among ±0.1–1.2 J g–1.

Hydrophobic factor values were between 1.63 and 2.38, showing that the activated carbons obtained with the larger BET areas also have a greater hydrophobic factor; these two parameters favor the adsorption of nonpolar molecules. This type of material tends to have a hydrophobic character, which is indicative of the adsorption of nonpolar-type adsorbates as is the case with CO2 into this type of material.[34]

Figure presents the calorimetric curves obtained by the immersion in benzene and dichloromethane of the carbons prepared, and it is noted that the magnitude of the peak in each sample is proportional to the enthalpy of immersion obtained in this analysis; the enthalpy with greater calculated value corresponds to ACCu5-1073 in both benzene and dichloromethane. There is a correlation between the enthalpies of immersion of the solids in benzene and dichloromethane.

Figure 8

Comparison between the obtained calorimetry curves from immersion in benzene and dichloromethane.

Figure shows the relationship between the enthalpies of immersion of materials in benzene on the basis of the surface area obtained; there is an increase in the enthalpy of immersion in benzene that is consistent with the increase of the BET area of the solids, demonstrating that the increase in the BET area is proportional to the rise in enthalpy; this behavior is adjusted as described and to that obtained in similar investigations that have prepared carbonaceous materials that have been characterized by benzene immersion calorimetry.[25,27,30,32,34] This correlation shows that the materials prepared under these conditions have an adequate interaction with benzene. It is possible to evaluate the adsorption of molecules with similar characteristics.

Figure 9

Relationship between immersion enthalpies in benzene and the BET surface areas of the materials.

Figure shows the relationship between the enthalpies of immersion in benzene as a function of the hydrophobic factor; it is noted that the enthalpies of immersion tend to increase the hydrophobicity of the materials. This behavior has been described by other researchers who believe that the hydrophobic character of the carbons decreases as the oxygen within the functional groups present on the surface of the solid increases.[35,36] This favors the adsorption of nonpolar compounds, bearing in mind that in this research the adsorbate molecule is CO2.

Figure 10

Relationship between the enthalpy of immersion in benzene and the hydrophobic factor.

Adsorption Isotherms of CO2 at 273 K

Figure presents the adsorption isotherms of CO2 at 273 K and 1 bar of pressure on the carbons. These isotherms are of type I and with the increase in the concentration of the solutions of Cu(NO3)2 to 3%, the largest volume of adsorbed gas is obtained; this solid shows the highest production of narrow microporosity with respect to the 5% activated sample because at this concentration, the production of (Vn) is gradually reduced, as this concentration of the agent used generates a greater amount of wide micropores and reduces the narrow micropores, causing a decrease in CO2 adsorption on this solid. This result correlates with the textural parameters and PSDs of this sample because it does not have the greatest surface area; the distribution of narrow pores of this sample is higher, as evidenced in Table , and these are greater than the other materials. This is evidence of a decline in the capacity to adsorb CO2 at 7% activating agent concentration, a fact that can be explained by the reduction of the surface area and pore volume.

Figure 11

CO2 adsorption isotherms at 273 K for the ACCu1073 series.

There is a tendency of the isotherms at low pressures to be close to a line and to subsequently decrease the adsorption to find values of volume that is held constant, which can be related to a mechanism of adsorption of CO2 in the coverage of the surface is associated with a range of pore sizes 2 times the molecular dimension of CO2 (0.33 nm);[37,38] for this research, the isotherms presented show a trend in accordance with those described in earlier studies.

Table reports the amounts of CO2 adsorbed at 273 K and 1 bar. It is noted that the solids have an adsorption capacity of between 103 and 217 mg CO2 g–1. The sample, ACCu3-1073 presents the highest adsorption capacity with a value of 217 mg CO2 g–1, with the greatest volume of narrow micropores (0.45 cm3 g–1), which shows that the fact that solid surface chemistry plays an important role in the adsorption of CO2 is not the only parameter that influences the adsorption, and therefore it is required that the adsorbent possesses appropriate textural features that ensures the ingress of CO2 into the porous solid for subsequent adsorption in the pores.[38,39] Researchers report that CO2 molecules better interact with the walls of the pores smaller than 2 nm in which their adsorption can be the result of both a micropore filling and a surface coverage mechanism.[40]

Table 5

Amount of Adsorbed CO2 and Narrow Pore Volume at 273 K and 1 bar

sample	volume adsorbed (mg CO2 g–1)	Vn (cm3 g–1)
ACCu1-1073	179	0.32
ACCu2-1073	103	0.16
ACCu3-1073	217	0.45
ACCu5-1073	201	0.36
ACCu7-1073	134	0.24

Figure shows the correlation between the amount of adsorbed CO2 and the narrow micropore volume determined from the experimental data of adsorption of CO2 and applying the DR model, you get a narrow wide microporosity, which correlates with the adsorbed volume of CO2 and favors the adsorption process. Activation with solutions of this type of metallic salt leads to the formation of porosity in solids and also generates a proportion of wide mesopores. This behavior has been reported by other authors such as Wahby et al., who prepared carbons with a particular pore diameter and high values of surface area, obtaining solids with a micropore-size distribution strait.[41]

Figure 12

Adsorption of CO2 vs Vn.

Table reports the kinetic data of CO2 adsorption on the mentioned activated carbons. Kinetic data were interpreted in light of three different models; the best adjustment to the intraparticular diffusion model is the rate-limiting step. In all samples, K1 is less than K2, which indicates that in the second stage, the adsorption is faster and the amounts of CO2 adsorbed are highest. In all materials, the correlation between the capacity and the speed of the CO2 adsorption process was observed; the kinetics of CO2 adsorption for this material is enhanced by its micro–mesoporous nature, in which the access routes to the micropores are larger. The solids that presented the highest kinetics correspond to ACCu3 and ACCu5 and these correlate with the determined textural parameters; in Figure the three models used for the ACCu3-1073 sample are compared; it was used to identify the diffusion mechanism. According to this theorywhich can be linearized as follows

Table 6

Adsorption Constants of Kinetic Models Applied to Materials

	pseudo first order		pseudo second order		intraparticular diffusion
sample	K1 (min–1)	R2	K2 (g mmol–1 min–1)	R2	K1 (mmol g–1 min–1)	R2	K2 (mmol g–1 min–1)	R2
ACCu1-1073	0.003	0.447	0.001	0.857	0.424	0.723	21.36	0.964
ACCu2-1073	0.003	0.479	0.002	0.849	0.232	0.767	16.39	0.966
ACCu3-1073	0.008	0.475	0.001	0.642	0.315	0.663	32.64	0.978
ACCu5-1073	0.005	0.508	0.002	0.873	0.419	0.672	32.52	0.974
ACCu7-1073	0.005	0.524	0.008	0.321	0.297	0.703	16.34	0.955

Figure 13

Adjustment of experimental data of adsorption of CO2 at 273 K to the kinetic models of solid ACCu3-1073.

Conclusions

Activated carbons were obtained with different textural characteristics: BET surface areas varied between 473 and 1361 m2 g–1, and micropore volume between 0.18 and 0.51 cm3 g–1. The activated carbon with the highest values of textural parameters was ACCu5-1073. Micro–mesoporous solids were obtained with the methodology used. This is important, as it may help the entry of CO2 molecules into the pores.

The adsorption of CO2 in the materials prepared presented values between 103 and 217 mg CO2 g–1; the values of volume of narrow microporosity obtained were between 0.16 and 0.45 cm3 g–1. The solid with the greatest capacity for adsorption of CO2 and volume of narrow microporosity was ACCu3-1073.

The enthalpies of immersion in water, benzene, and dichloromethane were exothermic, obtaining values in water between −28.8 and 47.5 J g–1, in benzene between −117.0 and −55.5 J g–1, and in dichloromethane between −117.3 and −50.6 J g–1. The highest values calculated in these apolar solvents were for the sample ACCu5-1073.

The adsorption kinetic of CO2 in the activated carbons prepared with metallic salt of copper is in good accordance with the intraparticle diffusion model, for which diffusion is the rate-limiting step.

The adsorption of CO2 in the prepared activated carbons is favorable from the energy and kinetic point of view; these accompanied by the presence of a wide micro–mesoporosity favor the entry of CO2 into the micropores.

Material and Methods

The precursor material was crushed in a knife mill and sieved to obtain a 4 mm particle size. For the impregnation of the precursor, African palm shell particles were mixed with Cu(NO3)2 solutions of concentrations 1, 2, 3, 5, and 7% w/v for 48 h at 358 K. These conditions were selected according to our previous investigations.[42−45] The carbonization was carried out in a horizontal Carbolite oven, heating to 1073 K at a heating ramp of 5 K min–1 and a N2 flow of 110 mL min–1. Subsequently, a change to CO2 atmosphere was made, maintaining this temperature for 6 h, then the temperature was reduced to 673 K for 2 h in a N2 atmosphere at the same flow to remove the excess CO2. The solids were first washed with HCl 0.01 M at 323 K and then with distilled water at the same temperature to constant pH to remove the excess activating agent present. Finally, the material was dried in an oven at 383 K for 12 h and stored in hermetically sealed plastic containers in a nitrogen atmosphere.[46] The identification of each activated carbon sample is made according to the following convention: AC (granular activated carbon) followed by the letter Cu (according to the salt used in the activation), the number 1, 2, 3, 5, or 7 representing the concentration (w/v) of the activating agent, and, finally, 1073 corresponding to the activation temperature (K).

Adsorption Isotherms of N2 at 77 K and CO2 at 273 K

The textural characterization of activated carbons was performed by physical adsorption of N2 at 77 K. The activated carbons were degassed at 523 K for 24 h. The apparent surface areas were calculated on the basis of the BET model,[47] and the micropore volume, Vo (N2) through the implementation of the Dubinin–Radushkevich model. The total pore volume, Vt was calculated on the basis of the volume and adsorbed at a relative pressure of 0.99.[48] The PSD was determined using the DFT model, taking into account the functional QSDFT and NLDFT.

Studies of CO2 adsorption at low pressure were made in a volumetric adsorption instrument at 273 K and pressure of 1 bar, reaching a pressure of 1.35 × 10–6 bar in the system. For these determinations, the automatic adsorption equipment, Autosorb iQ2 analyzer (Quantachrome Instruments) was used. The kinetic test was performed with the same volumetric instrument by measuring the CO2 adsorbed as a function of time, until the equilibrium value was reached.

The FTIR adsorption spectrum was determined for each carbon. A sample (0.100 g) of each was crushed and mixed with potassium bromide. This powder mixture was compressed in a mechanical die press to form a translucent disc and analyzed in a diffuse reflectance cell. The Thermo Nicolet 6700 FTIR equipment was used.[49]

Determination of Immersion Enthalpy

The enthalpies of immersion of the porous solids were determined in three solvents: water, benzene, and dichloromethane, in a heat conduction microcalorimeter, which measured the flow of heat through thermopiles of semiconductor materials installed in an aluminum block into which was inserted a stainless-steel cell of 10 mL capacity where the solvent was placed.[50] A sample of the solid (0.100 g) was weighed in a glass ampoule which was placed in the calorimetric cell, and the electric potential was measured for approximately 40 min to get a stable baseline. After that, the glass bulb breakage takes place and the generated thermal effect is recorded while the potential readings continue for 15 min more. Finally, the device is calibrated electrically.[51]

Temperature-Programmed Reduction

The TPR was carried out in a 300 Chembet (Quantachrome) instrument, using a thermal conductivity detector and a quartz reactor. Hydrogen (99.995% purity) was used as a reducing gas with argon (99.998% purity) as the purge gas and drag. The sample was previously degassed at 323 K for 1 h in an argon flow. The analysis was carried out with a heating rate of 10 K min–1, with a mixture of 10% v/v H2/Ar (3.1 mmol H2 cm–3), and a flow velocity of 0.27 mL(STP) s–1.[52]