The Study of Cu(II) Adsorption onto Synthetically Modified Geopolymers

Abstract

The study of Cu(II) from aqueous solutions using the adsorption process on synthetically modified geopolymers was performed under static conditions. Three geopolymers (based on metaphase of Serbian clay, metaphase of German clay and metaphase of German clay plus 10% of carbon cloth) were used. The geopolymers were made by condensing a mixture of metaphases and alkali activator solution at a fixed ratio at room temperature and then at a temperature of 60 °C in a dry oven. Then, the geopolymer samples were pre-crashed to a fixed-radius size. Their properties were characterized by X-ray diffractometry (XRD), Diffuse Reflectance Infrared Fourier Transform (DRIFT) analysis and Scanning Electron Microscopy (SEM) with Energy Dispersive Spectroscopy (EDS). Adsorption experiments were carried out under batch process as a function of the dose, concentration of metal, and contact time. The uptake of Cu(II) was rapid, and it increased with increasing metal concentration. The sorption percentage decreased with increasing concentration of Cu(II). The equilibrium adsorption capacity of geopolymers was measured and extrapolated using more isotherms. The data fit very well the linear Langmuir isotherm model. The pseudo-second-order kinetic model can well describe the adsorption behavior of Cu(II) ions with geopolymers samples. These results show that used geopolymers hold great potential to remove Cu(II) from industrial wastewater.

1. Introduction

The essence of environmental sustainability is the minimization of environmental burdens, which can also include contamination of water with heavy metals (a consequence of the development of industrial production, battery production, sheet metal production, and mining activities, etc.) [1,2,3].

Copper is an essential element for all plants and animals. All copper compounds, unless known otherwise, should be considered toxic [4,5]. The lethal dose of copper sulfate for humans is approximately 7 to 10 g. Copper is the third most abundant trace element in the body after zinc and iron. The suggested safe concentration in drinking water for humans varies depending on the source but tends to stabilize between 1.5 and 2 mg L⁻¹. Dietary reference intake: the tolerable upper limit for adult consumption of dietary copper from all sources is 10 mg/day [6,7].

An important part of copper toxicity comes from its ability to accept and donate electrons when its oxidation state changes. This catalyses the formation of highly reactive radical ions, such as the hydroxyl radical, in a manner similar to the Fenton reaction [8]. This catalytic activity is exploited by enzymes with which copper is normally associated and is therefore toxic only when separated and unmediated. This increase in unmediated free radicals is called oxidative stress and is the subject of active research in various diseases where copper may play an important but lesser role than in acute toxicity [9,10].

There are several procedures to reduce the concentration of Cu(II) in water. Among the most widely used are chemical precipitation [11,12], membrane filtration [13], ion exchange [14], solvent extraction [15], photocatalysis [16], reverse osmosis [17], electrochemical treatment [18], chemical coagulation [19], and adsorption [20,21]. Adsorption on natural or synthetic materials is one of the cheap, feasible, and effective methods for the removal of toxic metal ions from water [22,23].

Geopolymers are a favorable class of materials for the elimination of harmful substances from industrial and household liquid waste. It is commonly made from an uncomplicated reaction between alkali activator and reactive aluminosilicate. Commonly used as an alkali activator, it is a mix of sodium hydroxide and sodium silicate (NaOH/Na₂SiO₃) or KOH. Aluminosilicate and alkali activator through a reaction of geopolymerization creates a material responsible for pure fabrication [24]. Concerning the application of geopolymers, their properties can be controlled with the consumption of as little energy as possible through a process that is wholly environmentally friendly. Geopolymer manufacture is one of the effective methods for creating a rapidly inorganic ionic conductor [25]. Furthermore, the thermodynamic properties of the mixture alkali activator (density, viscosity, speed of sound) are conducive to easier arrangement of engineering processes [26]. Afterwards, geopolymers have a number of advantages: heat and acid/alkaline resistance, immense strength, and beneficial solidification of toxic waste [27]. After 1979, the term “geopolymer”, which categorizes a new class of alumino-silicate materials, was introduced by Davidovits [28]. Geopolymers are inorganic, solid, and stable at elevated temperatures; they are inflammable materials that are cured at low temperatures and with low CO₂ emissions, compared to Portland cement [29,30]. Si–O–Al structures with interchangeable Si and Al tetrahedral associate with each other by sharing oxygen atoms; they are building blocks of geopolymer samples [31].

A special three-dimensional network structure with determined pore sizes and pathways that allow certain heavy metals to pass through makes geopolymer an effective adsorbent. More than one decade ago, Wang et al. [32] investigated the adsorption capacity of fly ash-based geopolymer for Cu(II) in an aqueous solution. The value of adsorption capacity was 92 mg g⁻¹, which is a significantly higher value than those of fly ash. Zhang et al. [33] immobilized Cr(VI), Cd(II), and Pb(II) by using a fly ash-based geopolymer binder. They concluded that a chemical connection occurs in geopolymer gels. Yousef et al. [34] confirmed that geopolymers based on natural zeolitic tuff has a high adsorption capacity in terms of methylene blue and Cu(II) ions.

The use of geopolymers for the removal of Cu(II) from water has been investigated. A zeolite-based geopolymer was used for Cu(II) removal by Siyal et al. This material had a Cu(II) adsorption capacity per gram of geopolymer of 7.8 mg g⁻¹ [35]. Andrejkovicva et al. investigated the adsorption capacity of metakaolin-based geopolymers. The adsorption capacity of these materials, depending on their quantitative composition, ranged from 28.38 to 44.73 mg g⁻¹ of adsorbed Cu(II) [36]. Qiaoqiao et al. Al prepared NaOH activated geopolymer. They were able to prepare a material with a huge adsorption surface area. This material was prepared synthetically, and its theoretical adsorption capacity of Cu(II) was 335.43 mg g⁻¹ [37]. Ge et al. used a methacoline-based porous geopolymer to remove Cu(II) from water. The adsorption capacity of the material was 50.06 mg g⁻¹ [38]. Wang et al. used coal gangue and red mud based geopolymers for adsorption of Cu(II) from aqueous solutions. The theoretical adsorption capacity of these geopolymers ranged from 18.8 to 60.0 mg g⁻¹ [11]. Cheng et al. used a metakaolin-based geopolymer to adsorb toxic metal ions including Cu(II). The theoretical maximum Cu(II) adsorption capacity of their material was 48.78 mg g⁻¹ [39].

The adsorption of Cu(II) onto a bio-modified geopolymer from a multi-metal system of Cu(II), Fe(II) and Zn(II) was studied by Mama et al. The conclusion of their study was that Cu(II) adsorbed the worst among these ions. The calculated maximum adsorption capacities increased in the order of 35.01 mg g⁻¹—Cu(II), 45.18 mg g⁻¹ Fe(II) and 44.63 mg g⁻¹ Zn(II) [40].

The subject of this research was to prepare three clay-based geopolymers, characterize their structure, and compare their adsorption efficiencies with respect to Cu(II) adsorption. The effect of pH, amount of adsorbent, initial concentration of Cu(II), and contact time was also studied.

2. Materials and Methods

2.1. The Preparation of Adsorbent Samples

Geopolymers were synthesised by reaction of solid precursor and an alkali activator. A solid precursor was obtained using three different system: (a) metaphase of kaolinite of Serbian origin (MS), (b) metaphase of German clay (MG) and (c) mixing MG and 10 mas % of carbon cloth (MGC). The percent of carbon cloth is related to metaphase—MG. The carbon cloth was manufactured in ‘‘Vinča” Institute of Nuclear Sciences (Serbia) [41]. Origin of carbon cloth was viscose rayon cloth (from the Viskoza Factory Loznica, Yugoslavia). The cloth was carbonized in a nitrogen flow and activated in a carbon dioxide flow at 850 °C for 1 h. Methaphases were prepared by calcining kaolinite (Al₂Si₂O₇·2H₂O) of Serbian origin (Rudovci, Serbi) and clay that originated in Germany at 850 °C for 3 h. The alkaline activator solution was made by mixing a solution of Na₂SiO₃ (technical grade) and a solution of NaOH (analytical grade) with a concentration of 12 mol dm^–3 (12 M). The solution of Na₂SiO₃ supplied by a manufacturer from Serbia, “DEM Company”, Belgrade, (the chemical composition of Na₂SiO₃ was comprised of Na₂O = 14.7%, SiO₂ = 29.4%, and water 55.9%). Different metaphases and alkaline activators were mixed for 15 min. The ratio of mass of solid precursor and mass of liquid phase of alkali activator was 1.5. The obtained pastes were poured out into molds and shaken for 1 h to eliminate the bubbles. The samples stayed at room temperature for 1 day, and after that for 2 days in a sample-drying oven at 60 °C. Prior to using them as adsorbents, the geopolymer samples were crushed and sieved (hole diameter-355 µm). These samples of geopolymers were marked as GPS (geopolymer based on metakaolin of Serbian clay), GPG (geopolymer based on metaphase of German clay) and GPGC (geopolymer based on metaphase of German clay and mixed with carbon cloth).

Due to a microstructure evaluation and semi-quantitative elemental analysis of the surface, geopolymer samples were characterized by X-ray diffractometry (XRD), Diffuse Reflectance Infrared Fourier Transform (DRIFT) analysis, and Scanning Electron Microscopy (SEM) with Energy Dispersive Spectroscopy (EDS).

2.2. Cu(II) Syntetic Sollutions

For all the experimental aqueous solutions of Cu(II), analytical reagent grade CuSO₄ · 5 H₂O (Chemapol, Praha) was used. The equivalent mass of CuSO₄ · 5 H₂O (p. a.) was dissolved in deionized water. Deionized water was prepared via reverse osmosis Demiwa, Watek, Czech Republic. For the preparation of standard solution for AAS analysis was used a certificate referenced material—Cu(II) with a concentration of 1001 ± 4 mg L⁻¹ (Fluka Analytical Copper Standard for AAS).

2.3. Adsorption Experiments

For the adsorption experiments the 50 mL PVC tubes were used. In the tube 0.025 g of an adsorbent sample was mixed with 25 mL of copper solution. The tubes with the adsorbent sample and with aqueous solution of Cu(II) were shaken at a temperature of 25 °C in an orbital shaker ES-20/60 (bioSan, Riga, Latvia) at 200 rpm. After the required mixing time the suspension was centrifugated by a rotation centrifuge machine at 5000 rpm. The adsorption experiments were performed in duplicate using two independent samples. For determination of the copper concentration in the solution before and after adsorption experiments, atomic absorption spectrometry (AAS) was used. The experimental error bar limit between duplicate samples was lower than ±5%.

For the calculation of the percentage of Cu(II) adsorption (Ads. %), this equation was used (1):

$$Ads. \%=\frac{c_0{ - c}_t}{c_0}\times 100$$

(1)

where c₀ is an initial Cu(II) concentration (mg L⁻¹) and c_t is the Cu(II) concentration left in aqueous solution at time t (mg L⁻¹).

The amount of the adsorbed Cu(II) per mass unit of the adsorbent in time q_t (mg g⁻¹) was calculated by Equation (2):

$$q_t=\frac{\left(c_0{ - c}_t\right)V}{m}$$

(2)

where V is the volume of the Cu(II) aqueous solution (L), and m is the mass of adsorbent sample (g).

2.4. Atomic Absorption Spectrometry

The concentrations of Cu(II) before and after adsorption experiments were determined by using the atomic absorption spectrometer AVANTA Σ (GBC Scientific, Melbourne, Australia) with flame atomization. For the flame-fuel a mixture of acetylene and air was used. The experimental data were processed using a GBC Avanta v. 2.0 software (GBC Scientific, Melbourne, Australia). The working wavelengths were 217.90 nm for Cu(II) concentrations up to 60 mg L⁻¹ and 222.60 nm for a Cu(II) concentration limit lower than 300 mg L⁻¹. For periodically checking the instrument response standard metal ion solutions were used.

2.5. XRD Analysis

In order to determine the phase composition of the synthesized samples, XRD analysis was applied. Samples were recorded at room temperature using an Ultima IV Rigaku diffractometer; an X-ray copper tube was used with a generator voltage (40.0 kV) and a generator current (40.0 mA). The selected angular range of the recording was 5–80° 2θ in a continuous scan mode with a scanning step size of 0.02° and at a scan rate of 5°/min. The samples were prepared on mono crystalline Si-sample holders, and an ultra-high-speed detector D/TeX was used. The PDXL2 (Ver. 2.8.4.0) software was used to evaluate the phase identification and microstructure properties of the material. All obtained powders were identified using the ICDD data base [26]. For phase identification, selected PDF card numbers were used:

• PDF: 00-024-1047, Paragonite NaAl₂(AlSi₃)O₁₀(OH)₂
• PDF: 00-039-1380, Faujasite-NaNa₂Al₂Si₄O₁₂ · 8 H₂O
• PDF: 01-078-1252, Quartz SiO₂
• PDF: 01-089-0769, Zeolite X
• PDF: 00-043-0685, Illite KAl₂(Si₃Al)O₁₀(OH)₂
• PDF: 00-003-0052, Kaolinite Al₂O₃ · 2 SiO₂ · 2 H₂O

2.6. Diffuse Reflectance Infra Red Fourier Transform Spectroscopy (DRIFTS)

The samples for the DRIFT technique were mixed with KBr. By application of this technique the need for pelleting of samples was avoided. Contrary to the FTIR, the DRIFT technique can outdo the spectra in increased resolution and lower interference from water bands [42], but DRIFT spectra can be exposed to dissimilarity in the spectra due to sample particle size differences. That problem recurs throughout the pellet production. The DRIFT spectra in this research were obtained using the Perkin-Elmer FTIR spectrometer, Spectrum Two. The spectra of samples were scanned at 4 cm⁻¹ resolution and collected in the mid-IR region from 4000 to 400 cm⁻¹.

2.7. SEM Analysis

SEM analysis was used to determine the morphological characteristics of the samples, and EDS was used to determine the semi-quantitative chemical composition. Samples were Au coated prior to analysis and analyzed using the Scanning electron microscope model: JEOL JSM 6390 LV under a low vacuum with a backscattered electron detector and electron acceleration voltage of 25 kV. The EDS spectra were collected by scanning a certain area of the surface.

3. Results and Discussion

3.1. Characterization of the Adsorbents Samples

3.1.1. X-ray Diffraction

As can be seen from the identified phases of the powder diffractogram (Figure 1), the clay mineral phases represented in both clays were: kaolinite and illite, while quartz was the secondary mineral that accompanied these clay samples [43]. In the sample of Serbian clay (D), slightly more peaks belonging to the illite phase were identified, indicating a probably slightly higher content of illite in this sample compared to the sample of German clay (G). The kaolinite clay used for this research has already been described [26], and depending on the sampling locality of the clay itself, the results pointed to the illite phase, which occurs in different proportions in the sample itself [26,44].

Kaolinite and illite are clay minerals with a layered structure. From a chemical point of view, these are hydrated silicates of K, Na, and Mg, and from a chemical and structural point of view, these minerals have a significant content of OH groups. Clay minerals from a structural point of view can be represented as different orientations and interconnections among the octahedral and tetrahedral layers in the structure [45]. Figure 2 illustrates the crystal structure of kaolinite and illite.

The quartz represents the main crystalline residue in the amorphous geopolymer matrix. During the activation process, quartz (Figure 3) contributes with the share of silica in addition to the activation solution; however, it is not completely degraded. For this reason, it can be said that the geopolymer structure is semi-crystalline, where the intensity of the main quartz peak decreases and the baseline increases, i.e., the amorphous part on the diffractogram [46].

XRD results of geopolymer samples are shown on Figure 4. Based on the X-ray patterns of samples, the presence of a mixture of different zeolitic phases, obtained after alkaline activation of kaolinite and thermal treatment is noticed.

During the addition of Si, Al, and Na in excess during the formation of geopolymer, the formation of most likely synthetic faujasite (Zeolite with X type of structure) occurs in the geopolymer matrix. These characteristics of zeolite formation from a matrix rich with Si, Al, and Na have already been described by other authors [47]. The sample GPS consists mainly of three phases—faujasite, quartz, and paragonite. Faujasite has the chemical formulae Na₂A_l2 Si₄O₁₂ · 8 H₂O and represents sodium aluminum silicate hydrate. Paragonite represents Na-rich mica mineral, by weathering its converts with its analogue illite [48]. Paragonite is sodium dialuminum phyllo-decaoxodihydroxoalumotrisilicate and with the zeolite group of minerals can also be found in the geopolymer matrix.

All identified phases represent sodium aluminosilicates. Samples GPGC and GPG are very similar in terms of their phase composition where, in addition to quartz, which is the common phase for all three samples. The zeolite X phase was also identified; this phase, together with zeolite Y, belongs to the family of aluminosilicate molecular sieves with a faujasite-type structure (FAU). It is characterized by the formula (Ca, Mg, Na₂). Identified types of zeolites can be synthesized using kaolinite clay with an activation solution, such as the geopolymerization process with included thermal treatment [44,49].

In the amorphous structure of geopolymers, it appears that the negative charge is not localized, and it is more or less uniformly distributed in the framework. The Si⁴⁺ and Al³⁺ cations in the framework of aluminosilicate geopolymeric gels are tetrahedrally coordinated and linked by oxygen bridges. The negative charge on the AlO 4–group is charge-balanced by alkali cations (typically Na⁺ and/or K⁺). Geopolymers can also be viewed as amorphous analogs of zeolites, capable of cation exchange and possessing catalytic properties [27]. Based on the XRD analysis, we found the presence of zeolite in the GPG and GPGC samples, which additionally influences these two adsorbents to show better adsorption properties. DRIFT analysis confirmed the Al-Si-O amorphous structure, which is also responsible for the adsorption of the tested copper ions. Carbon cloth can adsorb small but signification quantities of Cu(II) ions from aqueous solutions. It is well known that the most common functional groups on the carbon surface are carboxyl, lactonic, carbonyl, and phenolic.

Figure 5 illustrates the structure of sodium rich zeolite Fujasite and Paragonite.

3.1.2. DRIFT Analysis

DRIFT is a fast and non-destructive way of evaluating clay minerals and their products [50]. DRIFT analysis results for GPS, GPG and GPGC are shown in Figure 6a–c. In the process of geopolymerization, dissolution of the raw aluminosilicate material took place. Silica and alumina were released and reacted to form a geopolymer gel which is subject to condensation to form a geopolymer. The basic steps of geopolymerisation involve dissolution of solid alumino-silicate oxides in MOH solution (M: alkali metal), diffusion or transportation of dissolved Al and Si complexes from the particle surface to the inter-particle space, formation of a gel phase resulting from the polymerization between added silicate solution and Al and Si complexes, and finally, hardening of the gel phase. The following reaction scheme was proposed by Xu and Van Deventer (2000) [51] for the polycondensation taking place during geopolymerisation of minerals:

Al–Si material (s) +MOH (aq) + Na₂SiO₃ (s, aq)

Al–Si material (s) + [Mz (AlO₂) × (SiO₂)y · n (MOH) · m(H₂O)](gel)

Al–Si material (s)[Ma ((AlO₂)a (SiO₂)_b) n MOH · m(H₂O)]

DRIFT results of geopolymer samples are not so easy to explain. The movement of some peaks, as well as the formation of new ones and the disappearance of some derived from the starting compounds point to the growth of newly formed geopolymers [52]. For geopolymer samples, a broad band around 3000–3450 cm⁻¹ and a band around 1650 cm⁻¹ usually appear. In our case the region 3000–3450 cm⁻¹ is covered by noise, but there are indications of the existence of a broad band in this region in the GPS and GPG samples. These two bands can be attributed to stretching vibrations of –OH and H–O–H groups, respectively [53]. The band at 2352 to 2371 cm⁻¹, which correlates with to the –CH stretching vibration with a strong hydrogen bond, is most pronounced in geopolymer GPS while it is almost non-existent in geopolymer GPG [54]. The band at about 1450 cm⁻¹ corresponds to the stretching vibrations of (CO₃)^2– groups in carbonates [55]. A powerful peak at closely 1000 cm⁻¹ of DRIFT spectra is joined with Si–O–Si asymmetric stretching vibrations, which is one of the fingerprints of geopolymerization [56].

The displacement of vibrational bands matching Si–O–X (X is Si, Al or O) bond stretching to lower wave numbers point to the elongation of the bonds, as well as the shrinkage of the bond angle [57]. In our case, a shift of the vibrational bands corresponding to Si–O–X stretching was observed. In the MK sample, the Si–O–Si band appears at 1063 cm⁻¹ while in the activated samples it moves to lower wave numbers [32]. In the analyzed geopolymer samples, these bands were indicated at around 1022 cm⁻¹ and 1024 cm⁻¹. This movement is due to the creation of Al–Si gel, which indicates the condensation of Si–O as well as SiO₄ and AlO₄ tetrahedra in the new network, geopolymeric network [55].

In the part of the spectrum corresponding to wave numbers at 1220–1000 cm⁻¹ there are two wide bands: Si–O and Si (Al)–O stretching vibrations [56,57]. The typical Al^IV absorption detected in metakaolin at 806 cm⁻¹ was superseded by some smaller bands on the 600 to 800 cm⁻¹ scale after geopolymerization. The band at about 460 to 600 cm⁻¹ is due to the bending of Si–O vibrations, Si–O– (Si, Al) bonds. The presence of quartz was confirmed by vibrational bands from 790 cm⁻¹ to 804 cm⁻¹; this was also observed by XRD analysis. In the DRIFT spectrum of geopolymer samples between 900 cm⁻¹ and 1300 cm⁻¹ there is a clear band associated with Si–O–M (M = Si or Al) asymmetric vibrations [58]. The peaks mentioned above are characteristic peaks that indicate that a geopolymeric structure has been created. The band at 700 cm⁻¹ refers to the bending vibration of Si–O–Al, which is an indication that there is a reaction between the original materials and the activator and the formation of an alkali-activated Si–O–Al structure [59]. In addition to the above, bands at 850–1030 cm⁻¹ were also detected, which can be attributed to asymmetric vibrations of Si–O tetrahedron stretching, while the absorption peak at 795 cm⁻¹ is directly correlated to the Si–O–Si symmetric stretching [60]. The band at 466 cm⁻¹ is attributed to the deformation vibration (O–Si–O), while the band between 600 and 800 cm⁻¹ can be attributed to the asymmetric stretching vibration of Al–O.

3.1.3. SEM Analysis

The morphology of synthesized geopolymers was investigated by SEM, as shown in Figure 7. The geopolymer samples (GPS, GPG, GPGC) were analyzed in a powder state. Amorphous, semicrystalline, and crystalline particles widened surrounded by the aluminosilicate gel of all the geopolymer samples. All of the crystalline, semicrystalline, and amorphous phases were mainly comprised of Si, Al, Na, Ca compounds, as claimed by results obtained from the EDS analysis. The presence of carbon in the GPS and GPG sample, according to the EDS analysis, suggests that certain of the compounds could be carbonates. Formation of Na₂CO₃ in the geopolymer samples is due to the existence of Na in the MS and MG samples and in the alkali-activators. Besides Na₂CO₃, calcite (CaCO₃) was created in the geopolymer samples, GPS and GPG. Carbon in GPGC also originated from carbon cloth.

The structure of the geopolymer matrix observed by SEM microanalysis depends on the molarity of the NaOH solution and the Si/Al ratio in the geopolymer samples. Matrix structures of investigated geopolymer samples consist of crystalline phases of quartz and zeolite, fujasite and paragonite. Fujasite and paragonite are found especially in the geopolymer made from Serbian clay (GPS) (Figure 7a). The geopolymer synthetized from Germany clay and carbon cloth consists predominantly of crystalline phase zeolite (Figure 7c). The gel structure was formed mainly from Si, Al, Na; it represents the major product, besides the unreacted metaphase particles, among all the samples studied from this time period. The microstructures of all geopolymer samples were inhomogeneous. Figure 7a shows semicrystalline, unevenly formed, individually arranged particles of GPS samples. All observed particles were of various sizes and have porous surface structures (Figure 7a). The geopolymers synthesized from metakaolin of clay of Serbian origin are the most porous, which can be explained by the ratio Si/Al (2.10) [43].

Table 1. EDS results of surface analysis of GPS, GPG and GPGC.

Elements	wt, %
Sample:	GPS	GPG	GPGC
C	17.59	18.33	19.80
O	45.75	45.13	44.53
Na	8.07	5.08	4.95
Al	8.20	9.78	9.59
Si	17.20	18.18	17.82
K	1.90	1.70	1.68
Ca	0.65	0.56	0.50
Ti	0.1	0.56	0.42
Fe	0.54	0.68	0.70
Total:	100.00	100.00	100.00
Si/Al	2.10	1.86	1.86
Na/Al	0.98	0.52	0.52

shows the elemental composition of all geopolymer samples as well as the ratio of Si/Al to Na/Al.

The existence of pores in the matrix of a geopolymer sample (Figure 7a’) shows the possibility of good adsorption characteristics of this geopolymer. The micrograph in Figure 7b is completely different from the one in Figure 7a. Unequally shaped low particles grouped together in agglomerates and singly arranged particles on the surface of great geopolymer plates are presented. In Figure 7b, groups of much larger particles can be seen, as well as smaller ones between them, which practically join them. The appearance of smaller particles on their surface is characteristic of this geopolymer. The geopolymer plates on the surface of the grate are obviously small particles of undissolved MG. Figure 7b’ shows the presence of a platy unreacted metaphase, as well as a dense amorphous, porous structure that is suitable for adsorption.

By adding carbon cloth to the metaphase of German clay, by alkaline activation, the microstructure of GPGC compared to GPG was changed (Figure 7c,c’), although the Si/Al and Na/Si ratios did not change. The structure of GPGC is more homogeneous, with a lot of small particles distributed throughout the surface of the amorphous gel structure. Figure 7c’ shows prismatic shapes which are most likely the result of the presence of crystalline phases of zeolite. Micro porosity is lower compared to the previous two samples.

Figure 7c shows a homogenous and less crystalline microstructure in contrast with the previous structure of two geopolymers (GPS and GPG). The amorphous phase formation is an outcome of the creation of alkali-activated products, C-(A)-S-H gel and N-(A)-S-H gel. Zeolite X is identified in the XRD analysis next to the sodium aluminate phases. The presence of zeolite phases leads to an interface between these zeolites and the alkali-activated matrix. An increase in the interface between phases gives space for essential weakness of the alkali-activated matrix, leading to low strength.

3.2. Adsorption Experiments

3.2.1. Effect of Adsorbent Dose

When the surface of a geopolymer is dry and subjected to fluids, the geopolymer adsorbs the fluid of the capillary structure [61]. From Figure 8 it was observed that the adsorption capacity decreased with increasing adsorbent dose. The amount of the adsorbed Cu(II) per mass unit of the adsorbent sample at equilibrium q_e (mg g⁻¹) was calculated according to Equation (3):

$$q_e=\frac{\left(c_0{ - c}_e\right)V}{m}$$

(3)

where c₀ and c_e (mg L⁻¹) were initial and equilibrium Cu(II) ions concentrations, respectively, m is the mass of the adsorbent sample (g) and V is the volume of metal solution (L). Regarding the maximum adsorbent surface occupancy at GPS—the sample was at 0.015 g of the sample mass. At the highest doses of adsorbent, the adsorbent surface was fully occupied by Cu(II) ions, and the amount of metal on the adsorbent surface could not have been higher. Different adsorbent sample composition accounted for the different shapes of the plot q_e vs. m by the GPS sample. The composition of the individual samples was described in Section 3.1. The most effective adsorbent was the GPGC sample, which is consistent with the theory of increasing the specific surface area by adding a carbon [62].

3.2.2. Effect of pH

The pH of the aqueous solution is a significant variable which controls the adsorption of the Cu(II) at the geopolymer–water interfaces. The theoretical calculated precipitation forms of Cu(II) in water solution are shown in Figure 9.

The visualization of precipitated forms was calculated with the software Visual MINTEQ v. 3.1 and the database minteqa.dat. From Figure 9 it is clear that the precipitated forms of Cu(II) begin at pH 6.5 approximately. The pH value in which the precipitated form was initiated depends significantly on the initial concentration of Cu(II) in the water solution, the temperature, and the atmospheric pressure. The data were interpreted as a function of q_e or a percentage of a selected form of Cu(II) cation and pH value. The experimental results of pH value before and after the adsorption process showed that the pH values of approximately 5.4 after adsorption of Cu(II) by geopolymers samples agree well with the theoretical calculation of speciation processed as summarized in Figure 9. The precipitation process of Cu(II) ions in insoluble hydroxides of Cu start at pH values higher than pH = 6 [62,63,64,65].

3.2.3. Effect of Initial Concentration Cu(II)

The influence of the initial concentration of Cu(II) on the adsorption capacity of geo-polymer samples was investigated in the interval of initial concentration of Cu(II) from 25 to 200 mg L⁻¹ (Figure 10).

The dose of the geopolymers samples and volume of Cu(II) solutions was constant in all of the experimental tubes (0.025 g of geopolymers per 25 mL of the Cu(II) ions solution). Starting at a low concentration, the adsorption capacity increased significantly with increases in the initial Cu(II) concentration. When the initial Cu(II) concentration exceeded 100 mg/L, the adsorption capacity still increased gradually.

Higher equilibrium adsorption capacities (q_e) were found for GPGC and GPG adsorbents (72.86 mg g⁻¹, 70.29 mg g⁻¹, respectively); GPS adsorbent showed lower efficiency with an equilibrium capacity of 58.30 mg g⁻¹. The adsorption capacity of geopolymers can be expressed in percentages of adsorption. Figure 11 illustrates the dependence of adsorption % on the initial concentration of Cu(II) in solutions for geopolymer samples. The highest experimental adsorption capacity at a concentration of Cu(II) 25 mg L⁻¹ had a geopolymer GPGC (96.92%); the lowest geopolymer GPS (85.93%).

3.3. Adsorption Isotherms

The experimental adsorption data were fitted to linear forms of adsorption isotherms models. The data were calculated from selected adsorption isotherms models, using Equations (4)–(7) in

Table 2. Equations of linear forms of used adsorption isotherms.

Isotherm Model	Equation
Langmuir	$\frac{c_e}{q_e}=\frac{1}{b q_m}+\frac{1}{q_m}{c}_e$	(4)
Freundlich	${\log q}_e{=\log K}_F+\frac{1}{n}{\mathrm{logc}}_e$	(5)
Redlich-Peterson	$\ln \frac{c_e}{q_e}{=\beta \ln c}_e - \ln A$	(6)
Sips	${\beta }_s \ln c_e=–\ln \left(\frac{K_s}{q_e}\right)+\ln \left(a_s\right)$	(7)

at the experimental temperature of 25 °C.

The experimental data were plotted for isotherm models (Figure 12, Figure 13, Figure 14 and Figure 15). By substituting them into the formulas, the parameters of the isotherms were calculated, which are listed in

Table 3. Calculated parameters of adsorption isotherms models.

Parameters	GPS	GPG	GPGC
Langmuir	qm (mg g−1) = 58.48 ± 0.06	qm (mg g−1) = 71.43 ± 0.02	qm (mg g−1) = 72.46 ± 0.07
	b (L mg−1) = 0.3353 ± 0.0003	b (L mg−1) = 0.4389 ± 0.0007	b (L mg−1) = 0.5391 ± 0.0005
	R2 = 0.9998	R2 = 0.9998	R2 = 0.9991
Freundlich	1/n = 0.1592 ± 0.0079	1/n = 0.2109 ± 0.0099	1/n = 0.1773 ± 0.0087
	KF = 26.6502 ± 0.0033	KF = 28.0543 ± 0.0039	KF = 33.8870 ± 0.0042
	R2 = 0.8887	R2 = 0.8157	R2 = 0.8202
Redlich-Peterson	β = 0.3053 ± 0.0012	β = 0.3004 ± 0.0047	β = 0.2851 ± 0.0021
	aR = 0.0378 ± 0.0002	aR = 0.0359 ± 0.0002	aR = 0.0300 ± 0.0003
	KR = 0.2563 ± 0.0010	KR = 0.2366 ± 0.0037	KR = 0.2345 ± 0.0024
	R2 = 0.9959	R2 = 0.9844	R2 = 0.9899
Sips	qmax (mg g−1) = 45.2489 ± 2.2624	qmax (mg g−1) = 27.0270 ± 1.3514	qmax (mg g−1) = 57.8035 ± 2.8902
	Ks (L mg−1) = 1.1510 ± 0.0576	Ks (L mg−1) = 2.7407 ± 0.1370	Ks (L mg−1) = 1.2815 ± 0.0641
	nS = 6.7267 ± 0.3363	nS = 5.1988 ± 0.2599	nS = 7.6029 ± 0.3801
	R2 = 0.9766	R2 = 0.9709	R2 = 0.9675

, where q_e was the mass of Cu(II) ions adsorbed until the time unit per mass unit of geopolymer was in equilibrium. q_m was the maximum adsorption capacity of the monolayer of used samples of adsorbent material (mg g⁻¹); c_e was the concentration of Cu(II) ions in the solution (mg L⁻¹); b was the Langmuir constant related to the affinity of adsorbate-adsorbent (L mg⁻¹) in equilibrium; 1/n, resp. n was an empirical parameter related to the intensity of adsorption (where for values in the range of 0.1 < 1/n < 1, adsorption is favorable); K_F was the surface adsorption equilibrium constant (mg g⁻¹); 𝐴 was the Redlich-Peterson isotherm constant (L g⁻¹); 𝛽 was the exponent that lay between 0 and 1; 𝐾_𝑠 was the Sips isotherm model constant (L g⁻¹); 𝛽𝑠 was the Sips isotherm exponent; 𝑎_𝑠 was the Sips isotherm model constant (L g⁻¹). The alternative isotherms fit better for heterogeneous surfaces, while the Langmuir model fits better to experimental data of monolayer adsorption onto surfaces of geopolymers samples [66,67,68,69]. From the values is significant, that the data fitted the best to Langmuir isotherms (R²; GPS = 0.9998; GPG = 0.9998 and for GPGC = 0.9991).

The values of adsorption capacities for the adsorption of Cu(II) by geo-polymers used in the literature with geopolymers of the present study are summarized in

Table 4. Adsorption capacities of different geopolymers for adsorption of Cu(II).

Used Adsorbent Material	qm (mg g−1)	Author/’s
Sodium silicate based geopolymer	48.78	[39]
Green geopolymer/alginate hybrid spheres (GAS)	62.50	[70]
Geopolymer based on fly ash and iron ore tailing	69.11	[71]
Germany clay based geopolymer	71.43	In this study
Metakaolin-based geopolymer (MKG)	86.60	[72]
Geopolymer based on Serbian kaolinite	58.48	In this study
Beihai in Guangxi Province, China —based kaolin porous geopolymer	52.63	[38]
Cetyl-trimethylammonium bromide (CTAB)-based geopolymer	40.00	[73]
Unmodified kaolinite	10.80	[74]
Organic carbon based geopolymer prepared from coffee residue	31.20	[75]
Geopolymer based on Germany clay with addition of a carbon	72.46	In this study

. Although direct comparison is difficult, owing to the differences in experimental conditions, it was found that the adsorption capacities of geopolymers used in this work were comparable to geopolymers presented in Table 4.

The auxiliary hypothetical yield from Langmuir adsorption isotherm is the value of the dimensionless separation factor calculate R_L, which characterizes the nature of adsorption within the Langmuir isotherm. This R_L value was calculated from Equation (8) and their values appear in

Table 5. The values of R_L for bentonites in concentration range 25–200 mg L⁻¹ Cu(II) at 25 °C.

Adsorbent	Concentration (mg L−1)
	25	50	100	150	200
GPS	0.1035 ± 0.0052	0.0560 ± 0.0028	0.0296 ± 0.0015	0.0205 ± 0.0010	0.0155 ± 0.0008
GPG	0.0810 ± 0.0041	0.0433 ± 0.0022	0.0228 ± 0.0011	0.0157 ± 0.0008	0.0119 ± 0.0006
GPGC	0.0670 ± 0.0034	0.0356 ± 0.0018	0.0186 ± 0.0009	0.0128 ± 0.0006	0.0097 ± 0.0005

.

$$R_L=\frac{1}{{1+b · c}_0}$$

(8)

where b is the equilibrium constant calculated from Langmuir isotherm model (L mg⁻¹) and c₀ is the initial Cu(II) concentration (mg L⁻¹). The results of R_L factor reveal whether the adsorption process is favorable, by the condition: (0 < R_L < 1), unfavorable—when the values of R_L factor are higher than 1. The separation factor can be a theoretical tool for balancing a linear 1 < R_L or irreversible (R_L = 0) adsorption process. The calculated R_L values at a concentration range from 25 to 200 mg L⁻¹ of Cu(II) for used samples of geopolymers are shown in Table 5. The R_L values indicate that the Langmuir model of adsorption isotherms is favorable—that is indicated by 0 < R_L < 1.

3.4. Kinetic Studies

The adsorption characteristics of Cu(II) by geopolymers depending on the reaction time (5–120 min) are shown in Figure 16.

The adsorption process was significantly fastest up to 30 min of contact time. Contact time longer than 120 min did not significantly influence the amount of adsorbed Cu(II) ions.

To explore the adsorption behavior, pseudo-first-order kinetic (Equation (9)) and pseudo-second-order (Equation (10)) kinetic models were utilized to examine the experimental data, which are described as follows:

$$\ln \left(q_e-q_t\right)=\ln q_e-k_{1}t$$

(9)

$$\frac{t}{q_t}=\frac{1}{k_2·q_e^2}+\frac{1}{q_e}t$$

(10)

where q_e and q_t are the adsorbed amounts of Cu(II) at equilibrium and at the time t (mg g⁻¹) respectively; t is the contact time (min); k₁ is the pseudo-first-order adsorption rate constant (L min⁻¹); k₂ is the pseudo-second-order adsorption rate constant (g min⁻¹ mg⁻¹). The calculated values obtained from the linear plots in (q_e − q_t) vs. time (t) did not agree with the experimental q_e values, and the obtained R² values are depicted in

Table 6. Comparison of the calculated and experimental q_e values and the adsorption rate constants for pseudo-first and pseudo-second-order reaction kinetics of the Cu(II) adsorption on geo-polymers at the temperature 25 °C.

Pseudo-First Order
Sample	Experimental qe (mg g−1)	k1 (L min−1)	Calculated qe (mg g−1)	R2
GPS	46.80	0.0071 ± 0.0004	18.81 ± 0.9405	0.1360
GPG	47.09	0.0069 ± 0.0003	25.54 ± 1.2770	0.1627
GPGC	50.52	0.0061 ± 0.0003	19.19 ± 0.9595	0.2039
Pseudo-Second Order
Sample	Experimental qe (mg g−1)	k2 (g mg−1 min−1)	Calculated qe (mg g−1)	R2
GPS	46.80	0.0026 ± 0.0002	42.19 ± 0.31	0.9965
GPG	47.09	0.0031 ± 0.0005	48.54 ± 0.17	0.9936
GPGC	50.52	0.0017 ± 0.0001	50.00 ± 0.32	0.9925

. The corresponding linear results were fitting for the pseudo-second-order presented in Figure 17 and Table 6.

The data in Table 6 show, significantly, that the correlation coefficients R² (0.9925–0.9965) for pseudo-second-order models are higher than the R² (0.1360–0.2039) values for pseudo-first order for adsorption Cu(II) onto used geopolymer samples. Comparison of experimental and theoretical data shows relative agreement. These results show that in terms of mechanism, chemisorption has a major influence—it controls the rate of the entire sorption process. Adsorption follows the pseudo-second order of the adsorption kinetic model.

4. Conclusions

Sorption performance of three synthetically modified geopolymers was studied for the adsorption of Cu(II) from aqueous solutions. All geopolymer samples were characterized by XRD, DRIFT, and SEM/EDS analysis. XRD analysis showed that all geopolymer samples represented sodium aluminosilicates. A strong band at 1000 cm⁻¹ is associated with Si–O–Si asymmetric stretching vibrations, which is the one of the fingerprints of geo-polymerization. Vibrational bands in the region from 790 cm⁻¹ to 804 cm⁻¹ confirmed the presence of quartz, which was also observed by XRD. The appearance of carbon in GPS and GPG showed that the some of the compounds could be carbonates. In the amorphous phase, crystalline and semicrystalline particles are mainly Si, Na, Ca, Al compounds. By the SEM analysis, the presence of the platy unreacted metaphases, as well as a dense, amorphous, porous structure that is suitable for adsorption were observed. The adsorption capacity of used geopolymers followed the trend: GPGC > GPG > GPS. The adsorption capacity increased with increases from the initial concentration. Isotherm studies indicated that the Langmuir isotherm model fitted the experimental data better than the other models. The adsorption equilibrium was described well by the Langmuir isotherm model with maximum adsorption capacities of 72.86, 70.29 and 58.30 mg g⁻¹ for Cu(II) on GPGC, GPG and GPS, respectively. The values of Ads. % were in accordance with this statement (96.92, 93.67 and 85.93% for GPGC, GPG and GPS, resp.) for the initial concentration of Cu(II) c₀ = 50 mg L⁻¹. The experimental and theoretical calculated data from the kinetics study are best fitted to pseudo-second order of adsorption process. To help summarize the presented study, the fact that all of the geopolymers used can be used for effective immobilization—adsorption of Cu(II) from aqueous solutions. The adsorption process of Cu(II) was favorable for all investigated geopolymer samples.