1. Introduction
Global warming is still a controversial topic worldwide. Its origin is mainly attributed to (1) Planet Earth’s natural cycles, which cause changes in the global temperature, and (2) human activities (anthropogenic global warming), which cause the deterioration of nature and a greenhouse effect. This has allowed a better compression of both research lines and how they can be connected. Anthropogenic global warming has gained importance in recent years because of scientific research and reports showing the relationship between the increase in gas emissions and the increase in global temperature [
1,
2,
3,
4,
5]. There are different greenhouse gases, but global warming is mainly related to large amounts of carbon dioxide emissions [
6,
7,
8,
9,
10,
11], of which about 78% is generated by industry through combustion processes [
8,
12,
13]. Industrial emissions (flue gas) are composed of 15%–20% CO
2, 5%–9% O
2, traces of other gases, and the N
2 balance [
14,
15]. However, current methods are not good enough to control CO
2 emissions [
16,
17]. The most frequently used method at the industrial level is absorption, but this has significant problems related to solvent regeneration, corrosion, and a high energy consumption, among others [
18,
19,
20,
21]. Considering this, many institutions around the world are promoting the exploration of other options for stabilizing and controlling CO
2 emissions in future years [
6,
7,
8,
22]. The International Energy Agency presents different options that could stabilize the emissions by 2035, such as enhancing the energy efficiency (50%), alternative energies (16%), biofuels (4%), nuclear energy (9%), and the carbon capture and storage process (CCS) (22%).
Therefore, CSS has been promoted and evaluated around the world in the last decades [
13,
21,
23,
24]. This method consists of three main stages: (1) CO
2 separation from flue gas, (2) CO
2 transport to the geological store, and (3) CO
2 injection into deep geological storage sites (more than 300 m [
21]). In this case, supercritical CO
2 is stored through the free space in the geological medium. Nevertheless, industrial massification of the CCS process still has different technical and economic challenges [
13,
21], mainly in the first and second stages (separation), which can consume about 70%–80% of the total CCS project costs [
25,
26,
27].
Accordingly, Rodriguez et al. [
24] presented an enhanced CCS process (e-CCS) based on nanotechnology to avoid separation, in which transport is avoided as CO
2 is separated and stored underground in-situ in shallow geological reservoirs (<300 m). There are many geological media that could be used for the e-CCS process. Natural geological media, such as gas/oil reservoirs, could be perfect for CO
2 storage due to their physical conditions, such as nature seals, which could prevent possible operational problems in the future. In sandstone reservoirs, the adsorption capacity is less than 0.0013 mmol g
−1, which can be improved in terms of the selectivity and uptake by surface modifications of the rock [
24]. In this case, CO
2 remains at gas conditions and, through the injection of carbon nanospheres, selective adsorption controls the capture and storage processes [
24]. On the other hand, the spherical structure and nanometric size allow the conservation of the porous structure of the geological media, avoiding the clogging of pores, which prevents operational problems due to formation damage and the associated loss of injectivity [
28,
29]. One great advantage of these nanospheres is that these materials are synthesized with specific chemical-physical characteristics for CO
2 adsorption process [
30], but they are not selective for carbon dioxide, as they show a high adsorption capacity towards other molecules in the flue gas stream, such as nitrogen.
It is worth mentioning that carbon nanomaterials have many advantages related to their high surface area, tunable porous texture, and easy surface functionalization [
20,
31,
32,
33]. Additionally, these materials can be obtained from carbon biomass precursors, resulting in a low cost [
34,
35].
Although there are reports of other materials with high adsorption capacities (for example, Mg-MOF74, with a CO
2 adsorption capacity of up to 10 mmol g
−1 at 25 °C and 0.01 MPa [
36]), their stability and physical properties may limit their application at the reservoir level. In the same way, 2D and 3D nanometric structures (nanofibers, nanosheets, and nanotubes) have been reported for CO
2 adsorption, where the capacity at a high pressure can be greater than 4–5 mmol g
−1. Nevertheless, these materials have limitations for geological application [
20,
32,
37,
38,
39]. Hence, due to their nanometric size and shape, nanospheres are the best option for e-CCS. However, the literature is very limited for these spherical nanomaterials applied to the CO
2 capture process evaluated at atmospheric pressure (average CO
2 adsorption capacity is between 2 and 4 mmol g
−1) and high pressure (average CO
2 adsorption capacity is greater than 3–5 mmol g
−1) [
20,
24,
37,
40,
41,
42].
In this sense, agriculture is an integral part of the economy of many countries. In Colombia, agricultural activities are developed in many rural areas that have been vulnerable to violent conflict [
43,
44]. These areas are significant producers of cane, rice, corn, and fruits, among other products, but several production chains have problems related to waste management. Organic waste or biomass, in some cases, represents social, health, and environmental problems [
43,
44,
45]. Nevertheless, these organic residues have the potential to be used as carbon precursors for micro/nanomaterial production, especially those that have a high carbohydrate content [
20,
30,
35]. Therefore, the International Sugar Organization estimates that about 110 countries produce sugar, and production from cane represents nearly 80% of global sugar production. In general, 100 tons of sugar cane will yield 10–11 tons of sugar and 3–4 tons of molasses [
46], which makes it an excellent candidate for the synthesis of carbon-based nanomaterials.
To the best or our knowledge, in the scientific literature, there is no research related to the use of carbon nanospheres synthesized from biomass residue with molecular sieve-type behavior for improving the natural selectivity for CO2 uptake.
Therefore, the main objective of this study was to synthesize and characterize carbon sphere molecular nano-sieves (CSMNS) from cane molasses. Nanofluids were prepared using the obtained CSMNS at different loads for impregnating the rock surface at 10% and 20%, in order to obtain a considerable change in the adsorption capacity of sandstone. Then, the materials were evaluated in the CO2/N2 adsorption processes at different pressures to mimic the e-CCS process conditions. As the main novelty, carbon nanospheres obtained from cane molasses behave like molecular sieves, which allows a natural selectivity towards CO2 in mixtures of CO2/N2 under different pressure and temperature conditions of an e-CCS process. The material could be considered a molecular sieve, due the fact that it only allows the entry of CO2 molecules within its porous structure, which is evidence of a high adsorption capacity, while it does not allow the entry of N2 molecules, thus exhibiting a poor adsorption capacity.
2. Materials and Methods
The carbon sphere molecular nano-sieves (CSMNS) were synthesized by means of a hydrothermal method using latex nanospheres as a template to obtain “hollow” spheres. The synthesized nanostructures were carbon micro/nanostructures obtained from cane molasses as a carbon precursor (CN.RON), and carbon nanostructures obtained from resorcinol/formaldehyde as a carbon precursor (CN.POL), which was synthesized to conduct a comparison with the adsorption performance of CN.RON. The synthesis from cane molasses was carried out by employing the H
2O/molasses mass ratios of 1800:1 (CN.RON1) and 3600:1 (CN.RON2). The materials were characterized to select the best material, considering important parameters in relation to the e-CCS process, such as the smallest size, adsorption capacity, and lower technical and economic cost [
24]. The best material was impregnated over Ottawa sandstone, which was used as a porous medium. The CO
2 adsorption was evaluated at atmospheric pressure and 0 °C, as well as at high pressure and temperatures of 25 and 50 °C (reservoir conditions).
2.1. Materials and Reagents
Cane molasses mainly consist of water, carbohydrates, protein, and fibers. The composition of the molasses employed for the nanomaterial synthesis is shown in
Table 1.
Furthermore, for synthesis, cleaning, drying, and carbonization, the following chemicals were used, all from Sigma–Aldrich (USA): Styrene (97%), acetoacetoxyethyl methacrylate (97%), sodium peroxydisulfate (97%), formaldehyde (37%), resorcinol (≥99%), Pluronic F127, acetone (99.9%), and ethanol (99.5%). N2 (high purity, grade 5.0) was used for carbonization.
2.2. Synthesis of Nanomaterials
The materials developed were latex spheres as templates for carbon spheres from resorcinol/formaldehyde (CN.POL) and carbon spheres from cane molasses (CN.RON).
2.2.1. Synthesis of Latex Sphere Templates
Latex was selected as the template considering that it is composed of carbon-based molecules, which are not completely removed from the nucleus after carbonization at 500 °C. This results in a different porous internal structure, which can be modified with other compounds to enhance CO2 capture and storage.
The process for obtaining the latex template was based on the method proposed by Agrawal et al. [
47]. For this, water (85 g), styrene (9.5 g), and acetoacetoxyethyl methacrylate (0.5 g) were mixed and charged into a double-wall glass reactor, using a mechanical stirrer, temperature controller, and nitrogen inlet. First, N
2 was bubbled through the reaction media for 30 min. After that, the reaction temperature was increased to 70 °C, and aqueous sodium peroxydisulfate (SP) solution (mass relation SP/water of 0.03:1) was added to reaction media to start the polymerization. The reaction was carried out at 70 °C for 24 h. The polystyrene spheres were dried under dynamic vacuum.
2.2.2. Synthesis of CN.RON
The process was adapted from White et al. [
35], using the latex template and defining particular concentrations of reagents and carbon precursor. The initial water:carbon precursor mass ratio was 1800:1 and the latex/carbon precursor mass ratio was 1:10. The latex nanoparticles were placed in water with 0.05 mL of Span 80 to disperse hydrophobic particles in the aqueous medium, at 25 °C for 4 h and 200 rpm. After that, the carbon precursor was added to the system, which was followed by stirring for 30 min. This solution was put in a hydrothermal reactor (Techinstro, Nagpur, India) with a capacity of 200 mL, at 180 °C for 24 h. After the reaction, the carbon gels were filtered and washed with excess deionized H
2O.
Later, gels are washed with acetone for 3 days, to permit the exchange of water molecules and preserve the porous structure during carbonization. Finally, the obtained polymer was dried at 120 °C for 12 h and carbonized under N2 flow at 60 mL min−1 and 500 °C (1 °C min−1) for 6 h, using a tubular furnace (Thermo Fisher Scientific, Waltham, USA). CN.RON1 was obtained. After that, the same procedure was carried out at a lower molasses concentration. The cane molasses was diluted at a mass ratio (H2O/carbon precursor) of 3600:1 to obtain smaller particle sizes. In this case, CN.RON2 was obtained.
2.2.3. Synthesis of CN.POL
The general procedure was adapted from Fang et al. [
48]; the carbon precursor and some operating conditions were changed to obtain a nanometric size and high surface area. The original method uses phenolic resol as the carbon precursor without a template, but the material yield is less than 10%. For this reason, resorcinol was used as a carbon precursor. Pluronic F127 was employed as a soft template, in order to obtain a porous structure.
Initially, a solution of resorcinol/formaldehyde (R/F) (molar ratio of 1:2), Pluronic F127 (concentration of less than 10−7 mol L−1), and deionized water (molar relation water/resorcinol ratio of 5556:1) was stirred at 200 rpm, 25 °C, for 18 h. Parallel to this, the latex nanoparticles were placed in water with 0.05 mL of Span 80 to disperse hydrophobic particles in the aqueous medium, at 25 °C for 4 h and 200 rpm. After that, the carbon precursor was added to the system, which was followed by stirring for 30 min. The solution was placed in a hydrothermal reactor (Techinstro, Nagpur, India) with a capacity of 200 mL, for 24 h at 130 °C. The obtained polymer was cleaned with acetone for 3 days to remove the water inside the porous structure. After that, the material was carbonized in N2 at 700 °C (1.5 °C min−1) for 4 h, using a tubular furnace (Thermo Fisher Scientific, Waltham, USA).
2.3. Impregnation of Sandstones
The sandstone was modified to enhance the rock surface with the nanoparticles and increase the chemical-physical condition and molecular interactions [
24,
49]. According to Rodriguez et al. [
24], the minimum nanoparticle percentages needed to enhance the CO
2 adsorption capacity are 10% and 20%. Considering this, in this work, Ottawa sandstone (SS) was impregnated with CN.RON2 at mass fractions of 10% and 20% by the immersion and soaking method [
49]. The nanofluid was developed by dispersing the nanoparticles in deionized water, which was sonicated at 40 °C for 4 h. The SS was introduced into the nanofluid at 60 °C for 24 h. This method mimicked the reservoir conditions. The modifying SS was dried at 110 °C for 12 h.
2.4. Characterization of Materials
Chemical-physical properties can be evaluated by different techniques. For e-CCS, it is essential to obtain nanomaterials with the smallest size, a spherical structure, and a high surface area. The details of characterization are presented below.
2.4.1. Size and Structure
Scanning electron microscopy (SEM) was used to obtain the dry particle size, size distribution, and morphology of CN.RON and SS. Transmission electron microscopy (TEM) was used to analyze the “hollow” structure. The analysis was carried out using a JSM−7100 emission electron microscope (JEOL, Nieuw-Vennep, The Netherlands), a GEMINI-LEO1530 VP FE-SEM emission electron microscope (Carl Zeiss, Cambridge, UK), and a Tecnai F20 Super Twin TMP transmission electron microscope (FEI, Hillsboro, USA).
A NanoPlus-3 zeta/nanoparticle analyzer (Micromeritics, Norcross, USA) at 25 °C was used to obtain the mean particle size of nanoparticles dispersed in fluid (dynamic light scattering (DLS)), where the particles were hydrated and interacted with each other [
24]. In this case, the nanomaterials were dispersed in water or ethanol (10 mg L
−1) and sonicated for 6 h before analysis.
2.4.2. Porous Structure
Materials were characterized by N
2 and CO
2 adsorption at −196 and 0 °C, respectively, using Autosorb adsorption equipment (Quantachrome Instruments, Anton Paar Quanta Tech, Boynton Beach, USA). The total adsorbed volume (
V0.95) is the physisorbed N
2 volume at a relative pressure of
P/
P0 = 0.95. The surface area (
ABET) was obtained by the Brunauer–Emmett–Teller (BET) model [
50]. The micropore volume (
Vmic) and average pore size (
L0) were obtained by the Dubinin–Radushkevich model [
51,
52]. The mesopore volume (
Vmeso) was obtained by the Barrett–Joyner–Halenda (BJH) model [
53,
54].
2.4.3. Chemical Composition
The chemical characterization was carried out by carbon, hydrogen, oxygen, and nitrogen (CHON) analysis for nano/micromaterials using a CE 440 elemental analyzer (Exeter Analytical Inc, Chelmsford, USA). Moreover, Fourier transform infrared spectroscopy (FTIR) for carbon materials and sandstone was performed using potassium bromide at a KBr/material ratio of 30:1 (% w/w) and at 25 °C. For this, an IRAffinity-1S FTIR spectrometer was used (Shimadzu Scientific Instruments, Columbia, USA). Thermogravimetric analysis (HP TGA 750, TA Instruments, New Castle, USA) was used to analyze the impregnation percentages of nanoparticles on sandstone. For this, the samples were burned at 800 °C (10 °C min−1) under an air atmosphere.
2.4.4. Dispersion of Nanoparticles in Aqueous Media
A NanoPlus-3 zeta/nanoparticle analyzer (Micromeritics, Norcross, GA, USA) was used to analyze the surface charge of particles and their dispersion stability at 25 °C (electrophoretic light scattering (ELS)). Nanoparticle/water suspensions were prepared with a pH adjusted to between 2 to 12 by adding solutions of 0.1 mol L
−1 HCl or 0.01 mol L
−1 NaOH, and then subjected to analysis. The zeta potential was calculated using the Smoluchowski equation, derived from the calculation of the Doppler effect [
55,
56,
57].
2.5. Adsorption Tests
The HP TGA 750 thermogravimetric analyzer (TA Instruments, New Castle, USA) was used to evaluate the performance adsorption of CN.RON2, sandstone, and impregnated sandstone (with a mass fraction of 10% and 20% of CN.RON2) at 25 and 50 °C and high pressure from 0.03 to 2.5 MPa for CO
2 and N
2. Initially, the material was cleaned by vacuum to remove gases and humidity. For this, the thermogravimetric analyzer had a vacuum pump coupled with an oven, in which the sample was placed. The moisture and adsorbed gases were removed from the surface and immediately afterwards, the adsorption process was carried out. The system was operated at 0.0025 MPa and 120 °C for 12 h. This equipment used magnetic levitation top-loading balance, which generated a uniform electromagnetic field. The sample weight was proportional to the current required to maintain the balance position. Additionally, this system allowed a highly sensitive analysis. The contribution of the buoyancy effect was manually subtracted from the data using blank tests carried out in the same conditions, but with an empty sample holder [
24]. The amount of each material was around 15 mg for nanoparticles, and 40 mg for sandstone and impregnated sandstone, in order to produce enough total surface area for adsorption.
The isotherms were fitted with the Sips and Toth models, which take into account multilayer adsorption. The models are presented in
Table 2 [
24,
34,
58,
59].
KS and
KT represent the adsorption equilibrium constants for the Sips and Toth models, respectively, and the
n and
t parameters indicate the heterogeneity of the system for the Sips and Toth models, respectively.
Nads (mmol g
−1) is the adsorbed amount,
Nm (mmol g
−1) is the adsorption capacity at equilibrium, and
P (kPa) is the equilibrium pressure. The heterogeneity may originate from the solid structure, the solid energy properties, or the adsorbate [
34].