Edited by: Rajib Mukherjee, Texas A&M University, United States
Reviewed by: Adrian Loy, Monash University, Australia
This article was submitted to Sustainable Chemical Process Design, a section of the journal Frontiers in Sustainability
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Carbon-based nanomaterials have exceptional physicochemical properties like high surface area and active sites, suitable for the adsorption of inorganic and organic compounds. Currently, these materials are being tested for environmental applications, e.g., detecting emerging pollutants in drinking and surface water. This work reviews different methods to prepare graphene (G) or graphene-based materials (GBM) using biomass or its constituents. This text brings together the methods used to revalue biomass by converting it into graphene materials from thermal treatments and their application in the adsorption of glyphosate from an aqueous solution. Computational studies were also added to evaluate the information about interactions between the herbicide and graphene layers.
Graphene (G) and graphene-based materials (GBMs) have been exceptional compounds due to their superior mechanical, electronic, thermal, and optical properties. These characteristics make them suitable materials for adsorption, photocatalysis, energy storage, gas sensing, and medical applications (Aïssa et al.,
To achieve a higher cost-effective process, higher G yield, and reduce the environmental impact, G synthesis using monomers and biopolymers (Li et al.,
Graphene is a hexagonal monolayer formed by sp2 hybridized carbon atoms, but some defects can appear during the synthesis. They are classified as intrinsic when the crystalline order is modified without the presence of foreign atoms (vacancies, 5 or 7 carbon rings) or extrinsic if G has impurities in the lattice (N, O, and Si). These atoms can extend to a more carbonaceous network, and the reduced dimensionality of G is related to a lower number of possible defect types (Banhart et al.,
Lignocellulosic biomass comprises five primary components: cellulose, hemicellulose, lignin (between the three constitute 85–90%), volatiles, and ash. Cellulose is the most abundant biopolymer on the earth and is formed by glucose. Hemicellulose and lignin are heteropolymers constituted by polysaccharides and polyalcohols, respectively (Ding et al.,
Two methods have been used for the synthesis of G from biomass. The first increases the amount of fixed carbon in the biomass by subjecting it to pyrolysis or torrefaction. Later, the product produced by this process is combined with metallic salts, and a heat treatment is carried out at high temperatures. In the other method, monomers of the macromolecules constituting the biomass are used; they are united through chemical reactions catalyzed by temperature, ultrasonic or ultraviolet radiations, and organic linkers. The G material obtained by the two methods is expected to have similar physicochemical properties. However, some interactions between char and the volatile compounds released during the direct pyrolysis of biomass can affect the result (Ding et al.,
In part, G materials can be obtained from both methods because many chemical reactions can be carried out by increasing the temperature or adding catalysts, e.g., the Diels-Alder reaction. Kong et al. indicated that pyrolysis of biomass produced highly porous G, and hence processes at high temperatures (graphitization and activation) are used to prepare high surface area G compounds. However, the materials obtained can have disordered layers and defects. They are classified as graphene-like materials because, despite the imperfections, they show the physicochemical properties of G (Kong et al.,
Many techniques can be applied to prepare graphene-like materials. Some include the salt-based method, template-based confinement, chemical blowing, coupling with hydrothermal treatment, and post-exfoliation (Kulyk et al.,
This method uses salts or mixtures that allow, among other things, to increase the surface area and porosity during the activation. The most common compounds contain potassium (KOH, K2CO3, and K2O) because this cation can intercalate in the carbonaceous structure of the precursor, and the lattice increases during thermal treatment. Furthermore, potassium salts can participate in redox reactions, increasing the degassing (release of H2O and CO2) and, therefore, the porosity of the G-like material (Nanaji et al.,
Materials with controlled physicochemical characteristics can be produced using different potassium salts. The impregnation of soybean shell with KOH generates graphene materials with a surface area of 1,152 m2 g−1; this result can be compared with materials obtained when the impregnation salt is potassium formate, whose surface area increases to 1,816 m2 g−1, and the mesopore volume also increases. Conversely, the layer thickness decreases (near to 3.8 nm), indicating that the G material comprises fewer layers, making it suitable for electronic and optical applications (Liu et al.,
Sucrose is a glycosyl glycoside present in biomass, and it is formed by the combination of one molecule of glucose and one molecule of fructose. This compound was mixed with KOH in a ratio of 1:0.3 (sucrose: KOH) (Dorontić et al.,
Glucose was also used to produce G-like structures using potassium salts. The LiCl/KCl (molten salt: melting point of 353°C) mixture was investigated to prepare G materials composed of few-layer G. This salt is mixed with glucose in a ratio of 1:100 (Glucose: salt) and thermal treatment at 500–800°C. The carbon nanostructures are generated from glucose polymerization in a solvent reaction medium formed by the ionic species (salts) (Liu et al.,
Experimentally, this finding is corroborated using X-ray diffraction (XRD) studies; the results showed a change in the peak position (2θ) from 18 to 23°, indicating a decrease in the C-C bond distance due to the release of water when the temperature increases. Raman spectroscopy also showed a rise in the IG/ID ratio and the peak formation at 2,600 cm−1 associated with the 2D peak and sp2 carbon atoms (Liu et al.,
On the other hand, at temperatures close to 300°C, caramelization reactions (Maillard reaction) prevent the salt from melting, forming a single phase where the product does not have the characteristics of G or its derivatives. Above this temperature (e.g., 700°C), two phases are observed by liquefaction of the salt. Moreover, the structure undergoes rearrangements, increasing the graphitic material by crystallization. In this regard, at low temperatures (<300°C), the surface area decreases from 630 (at 700°C) to 30 m2 g−1 (Li et al.,
Temperature is not the only factor that affects the surface area. Modifications in the methodology or precursor can also influence this parameter. A summary is shown in the following table.
Summary of conditions used to prepare G and GBM and their surface area.
Gingko shells (Hong et al., |
Biomass, NaCl and Urea (0.5:1.4:4 g) | 1,133 |
Cellulose (Perondi et al., |
Cellulose, FeCl3 and ZnCl2 (30 g: 3M; ZnCl2 ratio 1:3 biomass:salt) Pyrolysis temperature: 700°C | 1,229 |
Cellulose (Perondi et al., |
Cellulose, FeCl3 and ZnCl2 (30 g: 3M; ZnCl2 ratio 1:6 biomass:salt) Pyrolysis temperature: 700°C | 1,227 |
Coconut shell (Sun et al., |
Coconut shell, ZnCl2, FeCl3 (3g:9g:50mL of 3M solution) Activation-graphitization conditions: N2 atmosphere by heating the sample at 900°C for 1 h | 1,874 |
Sorghum stalk (Khalil et al., |
Stalk:ZnCl2 (15 g:200 mL of 10% solution) Pyrolysis: 500°C for 1h | 1,817 |
Lignin (Ma et al., |
Lignin:ZnCl2 (1:2) Pyrolysis: 450°C for 2h | 1,769 |
Among the salt processes to prepare graphene materials, we found a method called simultaneous graphitization-activation (SAG). This process allows the synthesis of two-dimensional (2D) carbon-based materials with intermediate properties between graphene and porous materials (graphene-like nanosheets, PGNS). In addition to the advantages already mentioned, these materials present a reduction in environmental pollution compared to synthesizing graphene oxide (GO) through the Hummers method (Ambika and Srilekha,
The SAG method involves the incorporation of the graphitization (FeCl3) and activation (ZnCl2) agents simultaneously in the carbonaceous matrix (biomass). The chemical reactions are catalyzed by the temperature (>850°C, 1 h), and an inert atmosphere is necessary when the pyrolysis is carried out. Graphitization is catalyzed by Fe3+ ions of the carbon source, forming a carburized phase, and ZnCl2 is used during activation to generate porous structures (Hou et al.,
This method uses templates on which the carbon source (biomass or its constituents) is deposited to form graphene sheets after heat treatment. Ruiz-Hitzky et al. used sucrose as a carbon source and sepiolite (a layered silicate) as inorganic support. They melted the sucrose on the support to form humin (polymeric material resulting from water loss and volatile organic species). The polymer intercalation in the silicate is assisted by microwave irradiation generating clay-humin nanocomposites. Then, this structure will form graphene-like-clay nanocomposites by thermal treatment at 750°C in an inert atmosphere. The XRD results obtained by the authors showed that the graphene-like-clay nanocomposites have an interlayer distance carbon-clay of 0.4 nm (Ruiz-Hitzky et al.,
Electrochemical impedance spectroscopy (EIS) measurements corroborated the conversion of molten caramel to graphene. It is observed that the intercalation of the nanocomposite containing humin in the silicate increases the electrical conductivity from 10−12 S cm−1 (in the material generated after thermal treatment) to 10−1 S cm−1 (Ruiz-Hitzky et al.,
Likewise, the presence of graphene is corroborated by 13C NMR analysis, where a single band at 125 ppm is observed, indicating the formation of C=C bonds from aromatic groups. Additionally, no bands are associated with oxygenated groups, suggesting that GO was not formed (Ruiz-Hitzky et al.,
The material previously described was used in electronic applications due to the decrease in the composite surface area to 1 m2 g−1 by the humin incrustation in the clay porosity. The authors suggest using other laminated clays with higher porosity (Ruiz-Hitzky et al.,
Tang et al. used carbohydrates (glucose, fructose, and sucrose) and PEG 20000 as carbon sources and soft template, respectively. The authors mixed 8.9 wt% glucose solutions and 0.2 wt% PEG solutions in glass bottles with a tightened cover; the bottle was heated in a conventional microwave oven at 595 W for 1, 3, 5, 7, or 9 min (microwave-assisted hydrothermal conditions). According to the authors, PEG 20000 forms micelles, in which sugars are incorporated using microwaves. Then, the mixture is cooled to room temperature. A color change from transparent to light yellow corroborated the presence of GQD in the solution. Electron energy loss spectroscopy evaluated the structure and chemical composition of GQD. The results showed oxygenated surface groups (539 eV), but the main signals were found at 284 and 291 eV, peaks related to sp2 C=C bonds (Tang et al.,
In addition to clays, zeolites, and other laminar materials, compounds such as boric acid can be used as a template. Ling et al. used boric acid dissolved in distilled water and mixed it with gelatin (other proteins from biomass could be used) in a 10:1 ratio. The solution was stirred at 80°C until the water had evaporated entirely. The obtained material consists of boric acid nanoplates coated by a gelatin layer (thin thickness of 100–200 nm). Finally, the composite was treated at 900°C for 1 h (ramp rate: 5°C min−1) under an inert atmosphere. The sample was washed using reflux and centrifugation with deionized water for 1.5 h, followed by freeze-drying (Ling et al.,
Among the graphene preparation systems that include conventional thermal treatments, several can be mentioned and are described in the specialized literature. Several techniques have been established for graphene synthesis. However, mechanical cleavage (exfoliation) (Novoselov et al.,
Systems that can be classified as advanced systems for graphene synthesis are widely reported and are experimentally complex in terms of the setups and the experience that the experimenter must have. An interesting experiment has recently been published and is reported to be a very good example of this type of synthesis. The arc discharge system is constructed with an anode and a cathode graphite rod, which are attached to the mechanical feed channel; the anode was thus adjusted to the separation of the cathode. Both electrodes were allowed to hang inside a horizontal quartz tube, and the quartz tube was surrounded by a tungsten heating coil. They then placed inside the quartz tube and against the inner wall of the quartz tube a 0.2 2 cm, 50 mm thick Cu foil with a purity of 99.95%. The quartz tube with the vacuum arc electrodes was enclosed in a high vacuum chamber which was evacuated at a base pressure of 10−7 torr. The center of the quartz furnace was heated to various processing temperatures from 300 to 400°C. Next, a gaseous mixture of H2 (10%) with Ar was introduced into the chamber for 120 min, and the furnace was heated to 600°C for 40 min and then held for 20 min under an ambient pressure of 50 torr. The H2 contained at 600°C is essential for graphene growth. This pre-deposition process removed the native oxide and resurfaced the Cu surface to aid carbon diffusion and segregation. Next, the authors quenched H2 and provided 40 V DC and 100 A electrical power through the vacuum gap of the graphite electrodes to cause a vacuum arc for 25 s. This generated a very high temperature of 2,400°C, vaporized the graphite, and injected carbon into the Cu surface. It also generated a very high flux of photons with a broad spectrum. By photon radiation, the Cu surface can be heated to its melting point within 1 s. After striking an arc under vacuum, the chamber was cooled down to the ambient temperature.
When synthesizing graphites, graphite oxides, and families of these compounds, it is important to perform an excellent characterization. One of the techniques to be used in this sense is to take Fourier-transform infrared (FTIR) spectra of the samples with a good number of scans using a suitable instrument; they are usually scanned between 4,000 and 600 cm−1. Another important technique in graphene characterization is solid-state 13C nuclear magnetic resonance (NMR) spectra taken at a slew rate of 62.5 kHz (4.0 μs, 115 W). The X-ray photoelectron spectroscopy (XPS) analysis spectroscopy should be performed using an instrument equipped with a monochromatic Al Kα excitation source operating at 1,486.7 eV. X-ray diffraction patterns (PXRD) should be acquired using an instrument equipped with a CuKα1/2 radiation source (λKα = 0.15406 nm) and a Lynxeye type detector. Samples should be scanned in the 2θ range of 10–90° 2θ using a step size of 0.026 °. Raman and solid-state photoluminescence (PL) spectroscopy spectra can be obtained by having equipment with a dual-frequency Lexel SHG argon ion laser. Silica should be used as a substrate. Raman measurements were performed with a 514.5 nm laser (green), and PL measurements were performed with an excitation wavelength of 244 nm. Total scattering data were collected at the Brockhouse high energy beamline at the Canadian Light Source using a wavelength of λ = 0.2081° and a PerkinElmer XRD1621 area detector XRD1621 positioned 160 mm downstream of the sample (Mokoloko et al.,
Finally, an NMR is taken in the specific case of hydroxyapatite, which is reviewed in the literature. In this way, the solid state allows to analyze the evolution of the functional groups in the rGO-HAP
The application of G and GBM on glyphosate (GLY) adsorption has been studied in some investigations, including experimental and computational research. In this section, we will show and discuss the most relevant results of some investigations related to this application, compare them, and analyze the interactions and mechanisms involved in this type of process. We clarify that the materials presented here are GBM nanocomposites that can be used to improve graphene physicochemical characteristics and its performance in environmental applications.
Currently, there is an environmental and health concern over the use of herbicides. These are chemical agents that act in different metabolic pathways of undesirable plants, herbs, and weeds; due to the constant increase in worldwide food demand, massive use of these chemicals in agriculture has become necessary. Among others, glyphosate is one of the most controversial broad-spectrum herbicides that are currently used, although about 20 countries have banned it. The health and environmental risk of GLY is related to the induction of antibiotic resistance; toxicity in organisms like amphibians, fishes, and reptiles; endocrine disruption; metabolic alteration; and potential teratogenic, tumorigenic, and carcinogenic effects (Meftaul et al.,
Adsorption is the process that takes place when a solid is in contact with a fluid (gas or liquid). It consists of the adhesion of fluid molecules (adsorptive) on the solid surface (adsorbent), thus increasing the fluid density in the vicinity of an interface (adsorbate) (Rouquerol et al.,
where
As mentioned, adsorption is an interface process, so it is important to understand the nature of the adsorbate–adsorbent interactions. To do so, it is necessary to analyze the physicochemical characteristics of GLY and GBM. First, GLY, also named as N-(phosphonomethyl) glycine (IUPAC), is a relatively small organic molecule that contains a carboxyl moiety (–CO2H), amino moiety (–NH), and phosphonic acid moiety [–PO(OH2)], which can undergo protonation and deprotonation reactions according to the pKa values. Thus, when GLY is dissolved in water, it can be found as a cationic, anionic, or zwitterionic form, depending on the degree of protonation of each functional group, i.e., depending on the pH of the solution. In addition, GLY has an amphoteric nature due to the presence of a basic group (amine) and two acidic groups (carboxylic and phosphonate) in its structure.
On the other hand, as mentioned earlier, G is a single-atom-thick material with a honeycomb lattice of sp2 hybridized carbon atoms, and therefore, it presents local aromaticity with two π-electrons distributed over every hexagonal ring (
Novel GBM have been synthesized for application in glyphosate adsorption, including metal-doped graphene, magnetic reduced graphene oxide (rGO), MOF-G composites, and graphene aerogels, among others. As shown previously, the materials obtained
Magnetic composites draw the attention of investigators because they present several advantages like ease of separating from solution (using a magnet), the presence of metallic centers that create new active adsorption sites, and an increase in particle size. In 2016, researchers of Université Laval reported the first study related to GLY adsorption onto magnetic GBM (Ueda Yamaguchi et al.,
A complementary investigation was done in 2019 by researchers of the same university (Marin et al.,
Other examples of magnetic GBM composites are as follows. First, a nanocomposite of magnetite and rGO, labeled as Fe3O4/rGO, was synthesized through a coprecipitation method (Li et al.,
All these composites have in common the use of iron oxides, besides GO and rGO. However, they were synthesized by different methods, different reactants, and different conditions that led to the production of different products. Indeed, these are composites of manganese ferrite (MnFe2O4), magnetite (Fe3O4), and iron oxide (III) (α-γ-Fe2O3). In general, they differ with respect to the oxidation state of iron, its crystal structure, and morphology, which attributes to the differences in their physicochemical properties. For example, in MnFe2O4, the oxidation states of iron and manganese are 3+ and 2+, respectively, with a spinel structure (cubic closed packed). Fe3O4 is also a ferrite, and in fact is the most common in nature. In this case, iron is presented as iron (II) and iron (III), in a spinel structure too. Finally, the oxidation state of iron in hematite and maghemite is 3+, but their crystal structures are hexagonal and cubic, respectively.
The adsorption results obtained by these composites are presented in
Simulated Langmuir isotherms using the parameters KL and
As can be seen in
To understand the reason behind these differences, it is necessary to compare some physicochemical parameters, such as surface functional groups, surface area, and porosity. The FTIR spectra of all composites showed the typical surface groups of GO and rGO, like hydroxyl (~3,400 cm−1), carbonyl (~1,650 cm−1), aromatic carbon (~1,150 cm−1), and epoxy group (~1,050 cm−1). In addition, the vibrations due to Fe–O bond appear as two bands at 450 and 565 cm−1 for all composites, due to the presence of the magnetic iron oxide. Another important characteristic related to the surface functional groups is the surface charge, which is measured as the pH of the point of zero charge (pHPZC) or isoelectric point (pHIEP). This parameter describes the surface charge depending on the pH of the work solution. When pH > pHPZC/pHIEP, the surface has a negative net charge; in contrast, when pH < pHPZC/pHIEP, the surface net charge is positive. In this case, all composites have an acidic pHPZC/pHIEP. It is worth noting that all researchers assured that adsorption experiments were carried out in a solution with a pH < pHPZC/pHIEP, and hence the surface net charge of the composites was positive. In addition, considering the pKa values of GLY and the investigated pH values in each investigation (3.0–6.0), GLY has a net negative charge under these conditions. Therefore, electrostatic attraction forces between the positively charged surface of composites and negatively charged GLY are responsible for the adsorption.
With regard to textural parameters, the BET surface area trend is MnFe2O4-rGO >> α-γ-Fe2O3-GO = MnFe2O4-rGO/VAC, and unfortunately, no textural parameters were determined for Fe3O4/rGO. BET area of MnFe2O4-rGO (305 m2 g−1) is ~15 times higher than the other two composites (20 m2 g−1). The average pore diameter of MnFe2O4-rGO, α-γ-Fe2O3-GO, and MnFe2O4-rGO/VAC is 3.4, 18, and 3.8 nm, respectively, since all of them are classified as mesoporous materials. Interestingly, α-γ-Fe2O3-GO does not have the highest surface area, although it has the highest adsorption capacity. So, it could be attributed to a better interaction between GLY molecules and surface functional groups of α-γ-Fe2O3-GO.
As shown previously, surface functional groups determined by FTIR were almost the same in all the composites. However, it is important to highlight that there are some differences between α-γ-Fe2O3-GO and MnFe2O4-rGO adsorption active sites, specifically in the type and the number of accessible adsorption sites. As proposed by the authors, electrostatic interaction is the principal mechanism of adsorption in these systems, and metallic centers are very important in this mechanism. They are Lewis acids, since they can act as electron-pair acceptors, while oxygenated charged groups of GLY are Lewis bases, because they are electron-pair donors. In addition, according to Pearson's acid–base theory, acids and bases could also be classified as hard and soft, depending on their size, charge, and polarizability. In addition, this theory states that soft acids prefer to react with soft bases and vice versa. In this case, oxygenated groups of GLY are hard bases, so they react preferably with hard acids. α-γ-Fe2O3-GO contains just Fe3+ metallic centers, while MnFe2O4-rGO contains Mn2+ and Fe3+ metallic centers, both being hard acids. Nevertheless, as Mn2+ radius is bigger than Fe3+ and they are isoelectronic, Mn2+ is softer than Fe3+.
Another important difference is related to their crystal structures and particle morphology, and their relationship with the surface chemistry of exposed crystal facets on the iron minerals. Probably, the crystal structure of manganese ferrite impedes the optimal interaction between GLY and its metallic centers, while the crystal structures of hematite and maghemite possibly favor GLY adsorption because of the distribution of surface functional groups on a specific crystal face, which are considered as facet–dependent properties.
Other types of non-magnetic GBM have been synthesized, such as MOF composite, aerogel, and polymeric membrane. A composite of GO and a Zr-based MOF (known as UiO-67) were synthesized using a solvothermal methodology (Yang et al.,
Langmuir's maximum adsorption capacity of GLY (
To compare the
Surface functional groups of CM-C/GA are the same as GO, including an amide group that comes from the carboxymethyl chitosan structure. In the case of UiO-67/GO, it presents Zr–OH groups that favor the formation of complexes with phosphonic and carboxylic groups of GLY. The textural properties of UiO-67/GO were not characterized; however, as MOFs got a tunable pore structure with different types of cavities, it is well-documented that they develop high BET-specific surface areas (>1,000 m2 g−1) (Kaur et al.,
In this case, the exceptional performance of CM-C/GA could be attributed principally to its high specific surface area and its porous network structure. According to the authors, the adsorption and removal of GLY using CM-C/GA are done
Even though there are some proposed adsorption mechanisms of GLY over GBM, there is still a lack of understanding of this mechanism. In addition, experimental studies on GLY–GBM interactions are limited, and therefore computational investigations have been conducted to explore these types of interactions at an atomic scale. Density functional theory (DFT) research has been performed to study the adsorption of GLY over GBM. In general terms, optimization of the electronic structures of isolated GLY and GBM are done, thus obtaining an energetic quantity for each one of the structures, i.e.,
An investigation was performed with pyridine-like nitrogen-doped graphene decorated with platinum and copper clusters (PNG/Pt-Cu) system (Gulati and Kakkar,
Another computational investigation was done to explore the adsorption mechanism of GLY on pristine G and G doped with Ti, V, Cr, B, Ca, N, Cu, O, Pt, and Pd (Li et al.,
The thermal techniques presented here allow to obtain GBM avoiding the use of other toxic techniques, such as Hummer's method. For this reason, these techniques result in greener synthesis methods in comparison to the traditional ones.
As to GLY adsorption using GBM, we showed that efficient and optimal adsorption systems have been designed, such as the G aerogel (CM-C/GA) and MOF-GO composite (UiO-67/GO). However, it is still necessary to investigate the application of these materials in dynamic adsorption, because this process mode is more used on a big scale. Also, it would be interesting to make an economic analysis comparing other carbonaceous materials with the same performance, because it is well-known that G is in principle an expensive material. In addition, as we propose here, graphene-derived materials obtained
Finally, computational investigations are a very important tool to understand the mechanism involved throughout the adsorption process. This favors the design of optimal adsorption systems because more focused experimental investigations could be carried out, reducing time and economic costs. Nevertheless, it is required to keep investigating the mechanisms to produce other GBM, such as GO, since this is the most common GBM used in the experimental investigations reported in the literature.
DH-B, VB, and LG collected the information and wrote the manuscript. DH-B made the figures. JM-P and PR-E revised and corrected the manuscript. All authors contributed to the article and approved the submitted version.
This work was funded by Science Faculty from Universidad de los Andes, with the project INV-509 2021-128-2257. An additional support of “Publica tus Nuevos Conocimientos y Expón tu Nuevas Creaciones”, de la Vicerrectoría de investigaciones de la Universidad de los Andes (Bogotá, Colombia).
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Authors want to thank ‘Fondo de Apoyo Financiero para Doctorado' from Universidad de los Andes and Science Ministery from Colombia (MINCIENCIAS). The authors also thank the Research and Postgraduate Committee-Faculty of Sciences of the Universidad de los Andes, Colombia. On the other hand, the authors thank the framework agreement between Universidad Nacional de Colombia and Universidad de los Andes (Bogotá, Colombia) under which this work was carried out. Juan Carlos Moreno-Piraján also thank for an award from the Facultad de Ciencias of Universidad de los Andes, number INV-2021-128-2257.