Evaluation of two fibrous clay minerals (sepiolite and palygorskite) for CO2 Capture

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering journal homepage: www.elsevier.com/locate/jece

Evaluation of two fibrous clay minerals (sepiolite and palygorskite) for CO2 Capture

T

J.A. Ceciliaa, E. Vilarrasa-Garcíab, C.L. Cavalcante Jr.b, D.C.S. Azevedob, F. Francoa, ⁎ E. Rodríguez-Castellóna, a

Universidad de Málaga, Departamento de Química Inorgánica, Cristalografía y Mineralogía, Facultad de Ciencias, 29071 Málaga, Spain Universidade Federal do Ceará, Departamento de Engenharia Química, GPSA-Grupo de Pesquisa em Separaçoes por Adsorçao, Campus do Pici, 60455-760 Fortaleza, Brazil

b

A R T I C LE I N FO

A B S T R A C T

Keywords: Sepiolite Palygorskite CO2 adsorption Microwave treatment

Two fibrous clay minerals (sepiolite and palygorskite) have been employed as adsorbents to assess their adsorption capacity of pure CO2 in a volumetric setup. The raw sepiolite reached a CO2 uptake of 1.48 mmol g−1 due to the presence of nanocavities acting as a molecular sieve. Both sepiolite and palygorskite were modified by microwave-assisted acid treatment, which led to an increase in specific surface area and pore volume due to Mg2+ leaching, particularly in the case of sepiolite. However, the partial digestion of these fibrous structures does not improve CO2 adsorption uptake due to the progressive increase of the size of nanocavities. In a next step, both fibrous clay minerals were functionalized with amine species by different procedures (grafting with (3-Aminopropyl) triethoxysilane (APTES), impregnation with polyethyleneimine (PEI) and double functionalization by grafting with APTES and then impregnation with PEI. In all cases, it can be observed that the incorporation of amine species favors the chemical interaction between the amines species and the CO2 molecules, although it also produces obstruction of the nanochannels so the adsorption takes place mainly on the outer surface of the fibers. Finally, the incorporation of amine species by double functionalization led to the highest CO2 adsorbed concentration of 2.07 mmol g−1 at 760 mm Hg and 65 °C due to a larger proportion of available amines sites as well as the use of higher adsorption temperature, which favored the diffusion of CO2 molecules within the adsorbent.

1. Introduction

can also have an important impact on the migrations, particularly when less developed countries are affected [3]. Taking into account the consequences of global warming, stringent environmental regulations have been put forward to limit the emissions of anthropogenic CO2. The proposed strategy relies on the premise that reducing the current energy demand by increasing energy efficiency and productivity and promoting the transition to a low-carbon sustainable economy is the best way to avoid CO2 emissions and boost economic growth. Even though more sustainable processes with net zero-carbon balance are under development, such as fuel cells or batteries, CO2 release is still unavoidable due to steam reforming (hydrogen production) or combustion (power generation) of carbon, natural gas, or biomass [4]. Considering these premises, CO2 capture and storage (CCS) is the most realistic strategy to minimize CO2 emissions at short- and mid terms, especially from stationary sources. Previous research has reported that the most costly step in the CCS process is attributed to CO2 capture, which account for 50–90% of the total cost

In recent years, the consequences of global warming by the greenhouse effect have become more evident. The Intergovernmental Panel on Climate Change (IPCC) has predicted that the average temperature of the Earth will have increased 4 °C in the year 2100 as a consequence of the emission of anthropogenic CO2, which is mainly attributed to the combustion of the fossil fuels [1]. This inflicts harsh consequences on the planet. Among others, the acidification of the oceans seriously affects marine organisms by the reducing the concentration of carbonate ions, which causes the dissolution of shells of macro invertebrates and coral-reef [2]. Global warming also gives rise to more unstable rain regimes, torrential in many cases, followed by long periods of drought. This causes a great impact on agriculture and ranching and serious problems with availability of potable water, leading to the consumption from aquifers with poor quality, which has become a significant source of diseases. In addition, the higher frequency of extreme weather events

⁎

Corresponding author. E-mail address: [email protected] (E. Rodríguez-Castellón).

https://doi.org/10.1016/j.jece.2018.07.001 Received 28 April 2018; Received in revised form 19 June 2018; Accepted 1 July 2018 Available online 02 July 2018 2213-3437/ © 2018 Elsevier Ltd. All rights reserved.

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Scheme 1. Elemental structures of sepiolite and palygorskite.

properties as sieves to retain CO2. As mentioned previously, the most expensive step of the CCS process is the CO2 capture. Most of the solids reported in the literature have a high potential in adsorption processes; however their large-scale synthesis would prohibitively raise the cost of CCS. In order to cheapen the costs, low-cost adsorbents with high availability and reduced energy demand for their regeneration have been developed for CO2 capture. With this respect, natural zeolites have been pointed out as alternative adsorbents in the purification and separation of gases [39]. Clay minerals are another abundant component on the terrestrial surface with a wide range of applications in the field of adsorption and catalysis [40]. Several authors have investigated clay minerals as CO2 adsorbents by using, for example, kaolinite [41], montmorillonite [42,43] or sepiolite [45,46] as starting materials. The aim of this work is to evaluate CO2 adsorption capacity of fibrous clay minerals sepiolite and palygorskite. Both minerals are relatively abundant on the surface of the Earth, which can potentially reduce the cost of the CCS process. Both fibrous clay minerals contain ribbons with 2:1 type layer structure, where each ribbon is linked to the next inverted SiO4 tetrahedral sheet by Si-O bonds. Thus, tetrahedral apices point in opposite direction in adjacent ribbons. The width of the apical oxygen atom strip consists of a tetrahedral ring in palygorskites and 1.5 ring in sepiolites. The ideal chemical composition of sepiolite is Si12O30Mg8(OH)4(H2O)8 where the trioctahedral positions are occupied by Mg2+ and small proportions of Al3+ and Fe3+. However, the ideal chemical composition of palygorskite is Si8O20(Al2Mg2) (OH)2(OH)4(H2O)4, which has a dioctahedral character, where the octahedral positions are occupied by Mg2+ and Al3+ generating voids in the octahedral sheets [47,48]. These inversions of the tetrahedral position lead to the formation of channels with dimensions of 0.37 nm × 1.06 nm in the case of sepiolite and 0.37 nm × 0.64 nm for palygorskite, where the charge deficiency is counterbalanced by the presence of protons and small number of exchangeable cations and zeolitic water [47,49] (Scheme 1). The formation of these microcavities may be an interesting feature of these materials for the adsorption of small gas molecules such as CO2. On the other hand, the physicochemical properties of these materials have been modified by microwave-assisted acid treatment to increase their surface specific area and pore volume according to the methodology proposed by Franco et al. in previous research [50]. The microwave activation is an interesting method to modify the magnesian clay minerals since it requires shorter activation times than traditional acid

[4–6]. Therefore, a great deal of effort is focused on the development of efficient technologies for CO2 capture. In the last decades, several technologies have been proposed for this goal. Among them, membrane selective permeation, cryogenic distillation, absorption in liquid amines and adsorption onto porous materials have been proposed [4,7], although each of these technologies has some intrinsic limitations. Membrane selective permeation has been used in gas separation with interesting results for highly concentrated CO2 streams, but the efficiency of this process decreases for low CO2 concentrations. Cryogenic distillation is also an effective technology to separate components by compression, cooling and expansion steps; however, it requires high energy consumption, so it tends not to be feasible on a small scale. The most mature technology used in CO2 capture is absorption in liquid amines or chilled ammonia. The process itself is relatively inexpensive in comparison to other technologies and it displays high yields of CO2 capture [8]. On the other hand, it has a number of drawbacks related to the volatility of the amines, which causes the equipment corrosion and the progressive loss of capture efficiency. In addition, amine regeneration requires high temperature, which increases the cost of the process [9]. The use of solid adsorbents has shown to be an alternative in CO2 capture. Several solids with basic properties, such as CaO [10,11], MgO [12,13] or mixed oxides coming from hydrotalcites [14,15] have been highly used due to their low-cost, high availability and high capacity to capture CO2. However, the interaction between these oxides and CO2 molecules is very strong so that high temperatures are required to regenerate them and subsequently there is a high energy penalty involved. Other solids act as sieves to retain CO2. With this respect, the literature brings an increasing number of reports of metal organic frameworks (MOFs), formed by organic-inorganic structures with high specific surface area and high microporosity, which may provide high capacity and selectivity to adsorb CO2 [16,17]. Recently, new hybrid materials such as graphene-organic frameworks (GOFs) [18] or gyroid mesoporous materials [19] have shown high potential to retain CO2. The design and synthesis of porous silica has evolved a lot over the last 25 years since Mobil scientific discovered a family with regular array of uniform mesopores, denoted as M41S [20]. Thus, porous silica with different textural parameters, such as SBA-15 [21–25], MCM-41 [26,27], MCM-48 [28], HMS [29,30], KIT-6 [31] or mesocellular silica foams [23,25,32,33] have been investigated for CO2 adsorption processes. The possibility of tailoring pore size, as in the case of zeolites [34–36] or activated carbons [37,38], renders these materials suitable 4574

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

a double functionalization method, following the methodology proposed by Sanz et al. [22]. In a typical synthesis, 0.3 g of sample were dried at 110 °C overnight in He flow at 110 °C. Then, the sample was functionalized by grafting with APTES, according to the methodology previously described [21]. Then, the sample was filtered, washed with toluene and dried overnight. Later, a larger amount of amines species were incorporated by impregnation of PEI following the methodology reported by Vilarrasa-García et al. [25], and then, the samples were dried in air at 110 °C overnight. The samples were labeled Sep for sepiolite and Pal for palygorskite. The modification of the raw materials by microwave-assisted acid treatment was indicated with the term 8, indicating the time, in minutes, of the acid treatment (i.e. Sep 8 or Pal 8). Finally, the incorporation of amines species was labeled as 20 A, 40 P and 20 A40 P, where 20 A is the concentration of amine grafting solution vol.(%) in the case of the APTES, 40 P is the wt.(%) percent for PEI, while 20 A40 P indicated the amount of APTES and PEI in the double functionalization.

activation. In addition, CO2 uptakes in the materials obtained by this activation method have not been reported in literature yet. These fibrous clay minerals were functionalized with amine species by grafting 3-aminopropyltriethosilane (APTES) and/or impregnating polyethyleneimine (PEI) since it is generally accepted that loaded amine species favor a chemical interaction with the CO2 molecules. This interaction is also favored at temperatures above room temperature (e.g. 75 °C) due to enhanced diffusion of CO2 molecules through the adsorbent and thus greater access to adsorption sites. The present research offers a comparative study of CO2 adsorption in fibrous clay minerals submitted to microwave assisted acid treatment and amine functionalization, which has not been reported in the literature yet. In addition, this manuscript also evaluates the efficiency of the incorporation of amines species on CO2 adsorption. Finally, note that the present research is carried out with pure CO2, which we understand to be a necessary preliminary step before simulating real emissions from an industrial facility [51–53]. 2. Experimental 2.1. Materials

2.4. Characterization techniques

The clay minerals used in the present study were sepiolite, under the commercial name Pangel S9 and palygorskite, denoted as Minclear, both supplied by Tolsa S.A. (Spain). Nitric acid (HNO3, VWR 68%) was employed for the partial digestion of the fibrous materials by acid treatment. Both raw materials and those obtained after the microwave acid treatment were functionalized with amine groups by grafting with 3-aminopropyltriethoxysilane (APTES) (Aldrich, 98%) using toluene (Aldrich, 99.5%) as solvent or by impregnating branched polyethylenimine (PEI) (average Mn ≈600, Aldrich) dissolved in methanol (Aldrich, 99.9%). The gases required in the characterization and adsorption experiments were He (Air Liquide, 99.99%,), N2 (Air Liquide, 99.9999%) and CO2 (AirLiquide, 99.998%).

The chemical analysis of the starting materials and the materials obtained by microwave-assisted acid treatment was performed by means of the MagiX X-ray fluorescence (XRF) spectrometer of PANanytical. X-ray diffraction patterns (XRD) for the samples were collected on an X'Pert Pro MPD automated diffractometer (PANalytical B.V.) equipped with a Ge (111) primary monochromator (strictly monochromatic Cu Kα1 radiation with a wavelength of λ = 1.5418 Å) and an X'Celerator detector. The overall measurement time and step size were set according to Franco et al. [50]. The morphology of the materials under study was examined by scanning electron microscopy (SEM) using a JEOL SM-6490 LV. The morphology of the adsorbents was also studied by transmission electron microscopy (TEM), by using a FEI Talos F200X equipment, which combines outstanding high-resolution S/TEM and TEM imaging with energy dispersive x-ray spectroscopy (EDS) signal detection, and 3D chemical characterization with compositional mapping. DRIFT spectra were collected on a Harrick HVC-DRP cell fitted to a Varian 3100 FT-IR spectrophotometer. The interferograms consisted of 200 scans, and the spectra were collected using a KBr spectrum as a background with a resolution of 4 cm−1. N2 adsorption-desorption isotherms at -196 °C were measured to determine the textural properties of the materials, by using an automatic ASAP 2020 system (Micromeritics). Prior the measurements, the samples were outgassed overnight at 110 °C and 10−4 mbar. The specific surface area was estimated using the BET equation considering a N2 cross-section of 16.2 Å2 [55]. The microporosity of the samples was determined using de Boer’s t-plot method [56]. The pore size distribution (PSD) was estimated from the desorption branch of the isotherm using Non-local Density Functional Theory (NLDFT) [57]. The total pore volume was calculated from adsorbed N2 at relative pressure of P/P0 = 0.996. The nitrogen content of samples that underwent the functionalization step was determined by elemental chemical analysis using a LECO CHNS 932 analyzer.

2.2. Modification of the starting materials by acid treatment The textural properties of the raw materials were modified by microwave-assisted acid treatment, as reported in previous research [50,54]. Briefly, 5 g raw material were treated with 50 ml HNO3 (0.2 M) solution for 8 min in an open glass reactor EMS20100OX (Electrolux, Stockholm, Sweden), operating at 800 W and 2.45 GHz. The acid treatment was carried out by applying microwave irradiation discontinuously so as to avoid higher temperatures than 100 °C. Thus, the irradiation was applied for 1 min and then switched off for 5 min to allow for cooling. Then, the samples were centrifuged and washed with water to remove excess nitrate from the medium, as determined by the Griess test. Finally, the samples were dried at 60 °C overnight. 2.3. Functionalization with amine groups The functionalization with amine groups by grafting of the fibrous phyllosilicates was carried out according to the methodology proposed by Hiyoshi et al. [21]. For each functionalization by grafting, 0.3 g sample were dried overnight at 110 °C under He flow. Then, the dried material was functionalized in a three-neck flask using 15 ml APTES solution in toluene (20% vol.) under reflux for 24 h at 110 °C. Finally, the suspension was filtered, washed with toluene and dried under air flow at 110 °C overnight. The incorporation of amine groups by impregnation with PEI was carried out following the procedure described by [25]. In each impregnation, 0.3 g of samples were dried at 110 °C overnight and then added to a PEI solution (0.12 g in 3 ml methanol) so as to reach 40 wt.% PEI in all adsorbents. Then, the sample was stirred for 1 h and dried in air at 110 °C overnight. The amine species were also incorporated into the clay minerals by

2.5. CO2 adsorption measurements The CO2 adsorption capacity of the raw and synthesized materials was determined from their CO2 isotherms, measured using a Micromeritics ASAP 2020 Analyzer (i.e., volumetrically) between 25 and 65 °C. Prior to the measurements, samples were outgassed at 110 °C and 10−4 mbar until complete outgassing was achieved. 4575

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

carbonate species (Cal and Dol) disappear after the acid treatment. It is also noteworthy that the baseline of the samples modified by microwave-assisted acid treatment is higher in the range 2θ = 15–26° as compared to the starting materials, which could be related to a possible generation of amorphous silica triggered by the acid treatment, mainly in the case of the sepiolite. As a matter of fact, Franco et al. have established that Mg2+ species located on the octahedral layers tend to leach during microwave-assisted acid treatment [50]. This process is accompanied by the release of amorphous silica, as indicated by the increase of the baseline between 2θ = 15–26°. In fact, previous studies have reported that Mg2+ species are more easily leached than Al3+ in smectites [54]. This might as well be true for sepiolites/palygorskites, in which case the higher proportion of Al3+ (esp. in palygorskites) renders this fibrous phyllosilicate a higher resistance to the acid treatment, as confirmed from the chemical analysis (Table 1).

2.6. Equilibrium models The CO2 adsorption isotherms were fitted to the SIPS model (Eq. 1)

q = qm ×

(b × P )1/ m 1 + (b × P )1/ m

(1)

Where q is the adsorbed concentration (mmols of adsorbed gas per unit mass of adsorbent) at a given T and P; qm is the maximum adsorbed concentration in mmol g−1 [58]; b is a parameter related to the adsorbate-adsorbent affinity; and m is a parameter related to the heterogeneity of the adsorbent. It may be observed that if m = 1, the SIPS model will be reduced to the Langmuir model. Hence, an adsorbent is more heterogeneous when it shows m values higher than 1 [59]. Dual-site Sips model was used to fit the experimental data of the functionalized materials. This model mathematically accounts for two types of adsorption sites, as shown in Eq. (2).

3.3. SEM images

1

1

q = qm1 ×

(b1 × P ) m1 1

1 + (b1 × P ) m1

+ qm2 ×

(b2 × P ) m2 1

(2)

1 + (b2 × P ) m2

The morphology of the raw sepiolite and palygorskite was studied by SEM (Fig. 2). The anisotropic morphology of raw sepiolite is shown in Fig. 2A.1, with particles sizes lower than 3 μm in all cases. The magnification of this sample (Fig. 2A.2) reveals that these particles are formed by an agglomerate of fibrous structures. After the acid treatment (Fig. 2B.1), the sample seems to be more disaggregated with a slightly lower fiber length, as indicated in its magnification (Fig. 2B.2 and Supplementary Information Fig. 1A). The micrograph of raw palygorskite also displays heterogeneous particles with variable sizes (Fig. 2C.1). In addition, the presence of cubic morphologies is noteworthy, which could be ascribed to the presence of impurities, such as MgO. After the acid treatment, the cubic morphologies disappear, confirming the results reported in the XRD data (Fig. 1B). The magnification of both palygorskite samples hardly reveals any changes upon acid treatment (Supplementary Information Fig. 1B). In both cases, a denser structure may be observed (Fig. 2C.2 and Fig. 2D.2) in comparison to the sepiolite, where the fibrous structure is less visible.

where q is the adsorbed concentration of gas (mmols of adsorbed gas per unit mass of adsorbent) in equilibrium with the gas phase pressure P (mmHg). qmi (also mmol g−1) is the maximum adsorbed concentration in site i with strength characterized by the magnitude of constant bi (mmHg−1). The fitting of experimental data with this model may supply information about the relative density of chemisorption and physisorption sites. In this work we have assigned site 2 as the site with higher value of parameter b and thereby the site where adsorption occurs with higher adsorbent-adsorbate affinity. 3. Characterization of the adsorbents 3.1. Elemental analysis The chemical composition of the sepiolite and palygorskite and the materials obtained by microwave-assisted acid treatment are shown in Table 1. The chemical analysis suggests that the acid treatment causes partial solution of the Mg2+ species, which are located on the octahedral sheets. However, the Al3+ species are more resistant to acid treatment, as previously reported by Franco et al. for dioctahedral and trioctahedral smectites [54].

3.4. TEM micrographs The morphology of the fibrous phyllosicates was also analyzed by TEM (Fig. 3). In all cases, it is clear that both sepiolite and palygorskite display a fibrous morphology with variable length but with similar thickness. In addition, one may appreciate that the fibers of the acidattacked sepiolite are shorter and with irregular edges (Fig. 3A), which is in agreement with the decrease in crystallinity observed by XRD (Fig. 1), while palygorskite suffers fewer modifications (Fig. 3B). According to XRD data (Fig. 1), sepiolite displays less impurities (Fig. 3A–B), while palygorskite shows the presence of fibers together with cubic structures and agglomerates ascribed to smectites (Fig. 3C). After the acid treatment, the cubic structures disappear; however, the structure of the smectite resists the acid treatment (Fig. 3D).

3.2. X-ray diffraction X-ray diffraction patterns of the starting materials are compiled in Fig. 1. The diffractograms reveal that sepiolite does not show other diffraction peaks to be attributed to other crystalline phases, whereas palygorskite exhibits several impurities in minor proportions, such as quartz (Qtz), calcite (Cal) and dolomite (Dol). Moreover, the presence of small bands at low 2θ angles (4.7–7.8º) in the case of palygorskite, suggests the existence of another phyllosilicate, probably a smectite with variable hydration sphere in its interlayer spacing. After microwave-assisted acid treatment, the diffractogram of the sepiolite shows that all diffraction lines are maintained, although the intensity of these peaks decreases after the acid treatment. With regard to the palygorskite, the intensity of the diffraction peaks attributed to clay minerals and Qtz are maintained, while the diffraction peaks assigned to

3.5. FT-IR spectra From the FTIR spectrum of the raw sepiolite in the range 40002800 cm−1 (Fig. 4), two bands located about 3689 and 3626 cm−1 may

Table 1 Chemical analysis of sepiolite and palygorskite and their respective materials after acid treatment determined by XRF. Sample

SiO2 (%)

MgO (%)

Al2O3 (%)

K2O (%)

TiO2 (%)

Fe2O3 (%)

CaO (%)

MnO (%)

P2O5 (%)

Sep Sep 8 Pal Pal 8

56.41 64.23 64.01 69.97

22.393 16.390 15.430 12.918

1.56 1.35 12.370 11.147

0.219 0.215 0.918 0.871

0.072 0.073 0.470 0.458

0.436 0.409 3.900 3.641

0.824 < 0.60 1.81 0.90

< 0.05 < 0.05 < 0.05 < 0.05

< 0.045 < 0.045 < 0.045 < 0.045

4576

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Fig. 1. X-ray diffractograms of sepiolite (A) and palygorskite (B) before and after the microwave-assisted acid treatment (8 min).

be assigned to the presence of −OH stretching vibrations coordinated with the magnesium species [60]. The band located at 3572 cm-1 is ascribed to stretching vibration modes of H2O coordinated with the Mg2+ species [60]. The broad bands with maxima located at 3416 and 3237 cm−1 are assigned to zeolitic water present in the channels formed by the inversion of the tetrahedral layer and physisorbed water on the surface of the layers, respectively. It is well-known that sepiolite only displays Mg2+ in the octahedral layer whereas palygorskites can contain Mg2+, Al3+ and Fe3+. These small chemical variations cause changes in the FT-IR spectra of both starting materials. Frost et al. reported that palygorskite displays four bands located at 3616 and 3554 cm−1, which have been ascribed to the −OH stretching vibrations of the Mg2+ and Al3+/Fe3+ cations, while the bands located at 3400 and 3255 cm−1 are attributed to the zeolitic and physisorbed water, respectively, in the palygorskite structure [60]. The FTIR spectra of the raw sepiolite and palygorskite between 1800 and 600 cm−1 (Fig. 4) shows an asymmetric band with a maximum located at about 1660 cm−1, which is attributed to the overlapping of the HeOeH bending modes of the zeolitic (1660 cm−1) and adsorbed water (1640 cm−1), respectively [61]. The set of bands located between 1350 and 850 cm−1 are ascribed to the overlapping of the SieOeSi longitudinal asymmetric stretching and SieOeSi transverse asymmetric stretching [60], while the band with the maximum at about 785 cm−1 is assigned to SieOeSi symmetric stretching vibrations and the band located at 470 cm−1 are assigned to the SieOeSi bending vibrations [62]. The band at about 655 cm−1 is assigned to Mg3OH bending vibrations with different coordination [62], while the band located at about 430 cm−1 is assigned to SieOeMg bending vibrations modes [63]. In the case of palygoskite, the presence of a shoulder at about 510 cm-1 can also be observed, which is ascribed to SieOeAl bending vibration modes [62,63]. The FTIR spectra of the samples after the microwave-assisted acid treatment reveal that eOH stretching vibrations coordinated with the magnesium species of the pristine sepiolite decrease upon acid treatment, while the FTIR spectrum of palygorskite hardly presents any change. Likewise, the band located at about 430 cm−1, assigned to the SieOeMg bending mode, also diminishes after the acid treatment, which suggests the Mg2+ leaching coming from the octahedral layer with the acid treatment. This fact is accompanied by an appearance of a small band located about 3740 cm−1, which is detected for both sepiolite and palygorskite, being attributed to the formation of silanol groups. The FTIR spectrum of the acid treated sepiolite also shows a broader Si-O-Si stretching band, which reinforces the formation of amorphous silica after the acid treatment [50]. The FTIR spectrum of palygorskite after the acid treatment is similar to that shown for the starting material, confirming that palygorskite is more resistant to the acid treatment than sepiolite. This also rules out the leaching of Mg2+,Al3+ or Fe3+ coming from the octahedral layer of palygorskite, which is in agreement with the findings previously reported in the XRD data (Fig. 1).

3.6. Textural parameters The textural parameters of the starting and modified materials were evaluated by N2 adsorption-desorption at -196 °C (Figs. 5A–B). According to the IUPAC classification, the isotherms of all samples can be considered as Type II [64], which is typical of physisorption of most gases on macroporous solids, as indicated by the increase in adsorbed N2 at higher relative pressures. Even though these materials are thought to be macroporous judging from the isotherm shape, a pronounced increase of adsorbed N2 is also observed at low relative pressures, so their microporosity is also significant. In all cases, the hysteresis loops are quite narrow (type H3). This type of hysteresis is usually assigned to non-rigid aggregated particles, as clay minerals, or macroporous structures that are not completely filled with condensate [64]. The specific surface area was estimated from the BET equation [55] (Table 2), reaching 182 m2 g−1 for sepiolite and 93 m2 g−1 for palygorskite. It is also worth noting that sepiolite exhibits higher microporosity and pore volume than palygorskite. A distinct increase of SBET following acid treatment is observed for sepiolite, as shown in Table 2 and Fig. 5A, reaching a value of 326 m2 g−1, by the partial solution of the Mg2+ species of the octahedral layer [50], which causes an increase of the microporosity. In the case of palygorskite (Fig. 5B), the microwave-assisted acid treatment produces a less pronounced increase of the SBET (from 93 to 122 m2 g−1). This indicates that the acid treatment hardly affects the textural properties of the palygorskite, which is in agreement with the findings from XRD (Fig. 1) and FT-IR (Fig. 4) data. The pore size distribution was determined using the NLDFT method [57] (Fig. 6). As suggested from the N2-isotherm profiles, the pore size distribution confirms a certain degree of microporosity, with pore sizes lower than 2 nm. These may be attributed to the channels formed by the inversion of the tetrahedral layer in the sepiolite and palygorskite. In addition, an important contribution due to the presence of macroporosity may be assigned to voids between fibers. The acid treatment causes a slight increase of the micropore size range in sepiolite, as suggested in its isotherm profile (Fig. 5A) probably due to the enlargement of its cavities by Mg2+ leaching. As expected, pore size distribution of palygorskite hardly undergoes any change upon acid treatment. 4. CO2 adsorption 4.1. CO2 adsorption in fibrous clay minerals (without amine species) CO2 adsorption isotherms at 25 °C are shown in Fig. 7 for pristine and microwave-assisted acid treated sepiolite and palygorskite. Note that both fibrous materials present impurities such as Qtz, Cal or Dol, although the CO2 adsorption capacity of these structures is negligible as compared to that of the fibrous clays. It is noteworthy that none of the adsorption isotherms exhibits a linear behavior with increasing 4577

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Fig. 2. SEM images of sepiolite (A), sepiolite after the acid treatment (B), palygorskite (C) and palygorskite after the acid treatment (D). Magnification: (1) X500, 50 μm and (2) X5000, 5 μm.

pressures. This is typical of adsorbents with small pore sizes, such as MOFs, activated carbons or zeolites [4], due to the high quadrupole moment of CO2, which favors its binding polar groups or ions on the solid surface [7]. Raw sepiolite exhibits an excellent CO2 adsorption capacity, reaching 1.48 mmol g−1 at 25 °C and 760 mm Hg. On the other hand, raw palygorskite only reaches a CO2 uptake of 0.41 mmol g−1 under the same conditions. These differences can be attributed to the larger channels of sepiolite, as indicated in Scheme 1, as well as the textural properties of each material. Pristine sepiolite has higher specific surface area and micropore volume, which favors higher CO2

uptakes in comparison to pristine palygorskite, which has lower microporosity. Both sepioliote and palygorskite have been evaluated for the geological storage of CO2 [65]. It is claimed that carbonic acid (CO2+H2O) is formed, which causes the partial dissolution of Mg2+ within the sepiolite thus increasing the microposity, similarly to what happens in the microwave-assisted acid treatment. However, palygorskite is more stable to the leaching, maintaining its SBET values. On the other hand, in this work, the absence of water discards the formation of carbonic acid and subsequent leaching of the Mg2+ species. The modification of the starting materials by microwave-assisted 4578

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Fig. 3. TEM micrographs of sepiolite (A), sepiolite after the acid treatment (B), palygorskite (C) and palygorskite after the acid treatment (D). Scale: (1) 1 μm and (2) 200 nm.

estimated b parameter. This fact may be attributed to the formation of cavities with appropriate sizes upon acid treatment, which enhance adsorbate-adsorbent interactions. It has been reported in literature that the functionalization of porous materials with amine species improves the chemical interactions between a given adsorbent and the CO2 molecules via zwitterion intermediate to form ammonium carbamate species [64]. Thus, both raw sepiolite and palygorskite, as well as their acid-treated counterparts, were functionalized by APTES grafting, by PEI impregnation and by double functionalization (grafting + impregnation with APTES-PEI).

acid treatment (Fig. 7) has a strong impact on CO2 adsorption capacity for both fibrous phyllosilicates. Acid-treated sepiolite suffers a strong decrease in CO2 adsorption capacity, going down to 0.93 mmol g−1 at 760 mm Hg and 25 °C (Fig. 7A), which may be ascribed to the loss of magnesium species and the subsequent slight widening of the channels, as takes place in zeolites subject to acid treatment. The increase observed in SBET upon acid treatment (Table 2) is accompanied by an increase in micropore size (Fig. 6A), which impoverishes its behavior as molecular sieve to retain CO2 molecules due to weaker physical interaction between the CO2 molecules and the wider nanochannels of acidtreated sepiolite. In the case of palygorskite, CO2 adsorption capacity improves after the acid treatment, attaining a value of 0.98 mmol g−1 (Fig. 7B) probably due to an increase of the microporosity and the micropore volume, approaching the uptake found for the raw sepiolite. All isotherms were well fitted with the Sips model (Table 3). In all cases, the m parameter is higher for the raw materials probably due to the stronger confinement effects in the channels of the pristine fibrous clays. This parameter is likely higher in the case of the palygorskite due to a higher content of impurities, as indicated by the XRD data (Fig. 1). The acid treatment partially dissolves these impurities leading to lower m value by the formation of more homogeneous systems. The parameters obtained from the SIPS equation reinforce the hypothesis that the acid treatment causes a decrease in CO2 adsorption capacity in sepiolite due to a lower adsorbent-adsorbate interaction, as indicated the b value. In the case of the palygorskite, the opposite trend can be observed, i.e. the acid treatment leads to an improvement in the CO2 adsorption capacity related to a stronger adsorption interaction between the palygorskite and the CO2 molecules, also confirmed by the

4.2. CO2 adsorption in fibrous clay minerals functionalized by APTES grafting The FTIR spectra of the fibrous phyllosilicates after the functionalization with amine species (Fig. 8) show the presence of new bands in comparison with those data reported for the pristine samples in Fig. 4. Thus, two bands located at about 1570 and 1480 cm−1 are noticeable, assigned to the CeH asymmetric and symmetric bending vibrations [44,46], which are more pronounced for the samples impregnated with amine-rich polymers (PEI), while the band located at about 1310 cm−1 is attributed to CeN stretching vibration modes [44]. The bands centered at 2950 and 2840 cm-1 are assigned to the asymmetrical and symmetrical CeH stretching vibration modes [24]. On the other hand, the FTIR spectra should display two bands about 3370 and 3270 cm−1, which are assigned to symmetrical and asymmetrical NeH stretching [67]; however, the presence of water may have masked the presence of these bands. The high hydration of the samples functionalized with

Fig. 4. FTIR spectra of sepiolite (A) and palygorskite (B) before and after the microwave-assisted acid treatment (8 min). 4579

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Fig. 5. N2 adsorption-desorption isotherms at -196 °C of sepiolite (A) and palygorskite (B) before and after the microwave-assisted acid treatment (8 min).

low SBET and pore volume, which increases the chances of blocking of the cavities upon functionalization. However, the adsorbents still maintain a relevant porosity after the grafting step so these materials still act as molecular sieves to a certain extent, which evidences that CO2 adsorption takes place on both physical and chemical sites. Sepiolite maintains its crystallinity after the grafting process. In the case of palygorskite, the ordering of the smectite fraction disappears, as indicated by the loss of its d100 diffraction, while palygorskite also maintains its crystallinity (Fig. 10A). APTES-grafted sepiolite (Sep 20 A) (Fig. 9A) exhibits lower CO2 adsorption capacity (0.99 mmol g−1 at 760 mm Hg and 25 °C) than the original clay. The small pore size of the sepiolite, about 1.2 nm, hinders the access of APTES molecules into the channels so the grafting reaction takes place mainly on the surface of the fibrous material. The grafting with APTES may even block the access to the micropores thus decreasing the available physical adsorption sites in comparison to the starting materials. Although palygorskite displays narrower channel size than sepiolite, the grafting with APTES improves the CO2 adsorption capacity of palygorskite (Fig. 9B), reaching 0.74 mmol g−1 at 760 mm Hg and 25 °C, higher than the uptake of the starting material. Considering the narrow pores of palygorskite, the grafting probably occurs on the outer surface of the fibrous structure. The microwave-assisted acid treatment causes an increase of the CO2 adsorption capacity in both cases, attaining uptakes of 1.30 mmol CO2 g−1 for Sep 8–20 A and 1.01 mmol CO2 g−1 for Pal 8–20 Aat 760 mm Hg and 25 °C. The improvement is associated to the partial digestion of the Mg2+ located in the octahedral sheet [46,50], which causes an increase of the pore volume, as indicated in Table 2. This larger pore volume favors the access of the APTES molecules to higher population of available silanol groups as isuggested from the data in Table 4, where N content increases for both acid-treated clays. The presence of N species is directly related to adsorbate-adsorbent interactions that are stronger than those observed in the adsorbents without functionalization, as indicated the b2 value estimated by the Dual site Sips model (Fig. 9 and Table 4). In addition, the presence of these amine

Table 2 Textural properties obtained from N2 adsorption-desorption isotherms at -196 °C for the raw materials and the samples after the acid treatment. Sample

SBET (m2 g−1)

Sext (m2 g−1)

t-plot (m2 g−1)

VP (cm3 g−1)

VMICROP (cm3 g−1)

Sep Sep 8 Pal Pal 8

182 326 93 122

134 224 81 94

48 102 12 28

0.608 0.775 0.397 0.463

0.021 0.045 0.005 0.011

Sep-20 A Sep 8-20 A Pal-20 A Pal 8-20 A

48 69 37 39

38 52 34 35

10 17 3 4

0.184 0.207 0.141 0.151

0.004 0.007 – –

Sep-40 P Sep 8-40 P Pal-40 P Pal 8-40 P

29 31 18 15

29 31 18 15

– – – –

0.064 0.071 0.053 0.059

– – – –

Sep-20 A40 P Sep 8-20 A40 P Pal-20 A40 P Pal 8-20 A40 P

13 15 7 10

13 15 7 10

– – – –

0.055 0.062 0.045 0.047

– – – –

amine groups, mainly those impregnated with PEI or APTES-PEI, can also be observed due to the increase of the HeOeH bending vibration, located at 1640 cm-1 [44]. The CO2 adsorption isotherms of the samples functionalized by grafting are compiled in Fig. 9. CO2 adsorption is much higher in the low pressure range in comparison to the samples without functionalization. This suggests the appearance of an additional binding mechanism in CO2 adsorption due to the incorporation of amine species by grafting on the surface of the samples. The functionalization with APTES causes a decrease in SBET and mainly in the microporosity. This behavior is slightly more pronounced in the case of the palygorskitebased materials (Table 4) because the starting materials already have

Fig. 6. Pore size distribution of sepiolite (A) and palygorskite (B) before and after the microwave-assisted acid treatment (8 min). 4580

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Fig. 7. CO2 adsorption isotherms at 25 °C of sepiolite (A) and palygorskite (B) before and after the microwave-assisted acid treatment (8 min).

attributed to the high polymer content causing the blockage of the micropores by the stacking of this amine-rich polymer on the surface of the adsorbent (Table 2), although the crystalline structure of the clay mineral is maintained after the PEI impregnation (Fig. 10B), following a similar trend for ATES grafting. Thus, it seems that CO2 adsorption is exclusively due to a chemical interaction between the amine species that cover the fibrous structure with CO2 molecules. Sepiolite impregnated with PEI (Sep-40 P) (Fig. 11A) also lower CO2 uptake at higher pressure than that observed for the raw sepiolite, reaching 1.02 mmol g−1. On the other hand, PEI-impregnated palygorskite (Pal.40 P) (Table 5 and Fig. 11B) has improved uptake in comparison to the respective raw material, reaching 0.73 mmol g−1 at 760 mm Hg and 25 °C. The difference in the CO2 adsorption capacity of both materials is attributed to their textural properties (Table 2). The higher SBET value, and mainly external surface, of the sepiolite minimizes the stacking of the polymer favoring its dispersion on the surface of the fibrous structure, which implies higher proportion of available chemical sites to interact with the CO2 molecules in comparison to the raw palygorskite. The acid treatment does not improve CO2 adsorption capacity in sepiolite, reaching of only 0.76 mmol g−1at 760 mmHg and 25 °C, even though the partial digestion of the octahedral layer causes an increase of the surface area and pore volume. Note that this process also provokes the liberation of amorphous silica (Fig. 1) coming from the tetrahedral sheets. All of this contributes to the loss of functional groups that would stabilize the amine-rich polymer by electrostatic interactions, mainly hydrogen bonds, on the surface of the adsorbent. The absence of these groups is likely to provoke the stacking of the polymer thus diminishing the available amine sites. In the case of palygorskite, the acid treatment hardly affects its structure and physicochemical properties, as was reported in the characterization data. This fact leads to similar CO2 isotherms before and after acid treatment (Pal-40 P and Pal 8–40 P). The CO2 isotherms were fitted with the Dualsite Sips model (Table 5). The increase in CO2 uptake at low absolute pressure suggests that adsorption occurs by chemical interaction between the amine groups and the CO2 molecules. As previously indicated, chemical adsorption sites interact stronger with the CO2 molecules than the physical adsorption sites, which is in agreement with the much higher value of the b2 parameter with respect to b1 in all samples impregnated with PEI. The high proportion of amine species stacked on the surface of the adsorbent leads to a more homogeneous system in comparison to the samples functionalized with APTES, since grafting only occurs where silanol groups are present, whereas PEI covers the whole surface. The reaction mechanism of CO2 with amine groups is through zwitterion formation, according to the following reactions (Scheme 2): The formation of the zwitterion is followed by deprotonation with a base to produce a carbamate. CO2 adsorption efficiency, defined as the ratio between moles of adsorbed CO2 and moles of N in the material, is a useful metrics to quantify the effectiveness of amine-based CO2 adsorbents. In the mechanism of zwitterion formation, under anhydrous conditions, a second amine typically acts as the base to produce an

Table 3 SIPS parameters of the CO2 adsorption at 25 °C for the raw materials and the samples after the acid treatment. Sample

q760 (mm Hg)

qm (mmol g−1)

b (mm Hg−1)

m

Sep Sep 8 Pal Pal 8

1.48 0.93 0.41 0.98

2.50 1.93 0.92 1.22

0.002 0.001 0.001 0.006

1.24 1.10 1.46 1.26

species on the surface of the adsorbents leads to CO2 adsorption by chemical interaction. This generates more heterogeneous adsorption systems, as indicated by the increasing values of parameter m (Table 4). It is generally accepted that the interaction of CO2 with amine species takes place by the formation of carbamate species, through zwitterion intermediate, in the absence of H2O. The theoretical efficiency of this type of CO2 adsorption follows the molar ratio CO2/ N = 0.5. The coexistence of physical and chemical sites in the samples functionalized by grafting makes it difficult to assess the efficiency of the N species. If q760 or qm were considered, the obtained data would be CO2/N > 0.5, because the contribution of physical adsorption sites adds to that of chemisorptions sites. Considering that the N species are only involved in the chemical adsorption and that this adsorption takes place at low absolute pressure, it has been assumed that the adsorption at q50 (50 mm Hg) is only attributed to the chemical sites (Table 4). From these data, it may be observed that the CO2/N efficiency is in the range (0.27-0.41). These values are lower than 0.5, which may be explained as follows. The primary amine formed in the grafting process may react with another APTES molecule or neighboring silanol groups leading to secondary or ternary amines [21], which are less efficient in capturing CO2 than the primary amine. As shown in Table 3, sepiolitebased materials display higher amine capture efficiency in comparison to palygorskite materials. Likewise, the acid treatment also causes an improvement in the CO2/N efficiency. In summary, the APTES grafting does not lead to a significant improvement of CO2 adsorption capacity in either fibrous clay due to the small nanopores of both materials that are unlikely to host the amine effectively. In the case of mesoporous silicas [29,68], which have a larger pore diameter, the incorporation of APTES by grafting does increase the CO2 adsorption due to the coexistence of physical and chemical sites. In contrast, the incorporation of APTES in sepiolite and palygorskite produces pore blockage, so the physical adsorption sites decay too much to reach appreciable CO2 adsorption values. 4.3. CO2 adsorption in fibrous clay minerals functionalized by PEI impregnation The adsorption isotherms at 25 °C for both fibrous phyllosilicates functionalized by impregnation with an amine-rich polymer (PEI) are compiled in Fig. 11. All isotherms exhibit a rectangular profile. This is 4581

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Fig. 8. FTIR spectra of sepiolite (A), sepiolite modified with acid treatment (B), palygorskite (C) and palygorskite modified with acid treatment (D) functionalized with APTES (A), PEI (P) and APTES-PEI (A–P).

Fig. 9. CO2 adsorption-desorption isotherms at 25 °C of sepiolite and sepiolite treated with acid treatment functionalized by grafting with APTES (A) and palygorskite and palygorskite treated with acid treatment functionalized by grafting with APTES (B). Adsorption (filled marks) and desorption (non-filled marks). 4582

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Table 4 SIPS parameters of the CO2 adsorption at 25 °C for the samples functionalized by grafting with APTES. Sample

SBET (m2 g−1)

N (%)

q50 (mmol g−1)

q760 (mmol g−1)

qm1 (mmol g−1)

b1*1E4 (mm Hg−1)

m1

qm2 (mmol g−1)

b2*1E3 (mm Hg−1)

m2

molCO2/molN (q50)

Sep-20 A Sep 8-20 A Pal-20 A Pal 8-20 A

48 69 37 39

2.29 3.07 2.15 2.46

0.58 0.89 0.42 0.64

0.99 1.30 0.74 1.01

2.21 1.04 0.11 0.78

1.09 0.40 1.02 2.38

1.22 1.02 1.00 1.01

1.17 1.82 1.26 1.52

16.51 13.71 2.34 4.82

4.68 3.01 3.11 3.75

0.35 0.41 0.27 0.36

Fig. 10. Comparative of the x-ray diffraction before the functionalization and after the functionalization by grafting with APTES and by impregnation with PEI for sepiolite (A) and palygorskite (B).

Fig. 11. CO2 adsorption-desorption isotherms at 25 °C of sepiolite and sepiolite treated with acid treatment functionalized by impregnation with PEI (A) and palygorskite and palygorskite treated with acid treatment functionalized by impregnation with PEI (B). Adsorption (filled marks) and desorption (non-filled marks). Table 5 SIPS parameters of the CO2 adsorption at 25 °C for the samples functionalized by impregnation with PEI. Sample

SBET (m2 g−1)

N (%)

q50 (mmol g−1)

q760 (mmol g−1)

qm1 (mmol g−1)

b1*1E4 (mm Hg−1)

m1

qm2 (mmol g−1)

b2 (mm Hg−1)

m2

molCO2/molN (q50)

Sep-40 P Sep 8-40 P Pal-40 P Pal 8-40 P

29 31 18 15

10.33 11.11 10.67 11.32

0.91 0.64 0.68 0.74

1.02 0.76 0.73 0.78

0.24 1.01 0.17 0.20

9.03 0.36 7.41 5.81

1.05 1.05 1.01 1.01

0.94 0.63 0.68 0.73

20.49 11.30 19.86 24.71

2.11 2.15 1.02 1.03

0.12 0.08 0.09 0.09

subject to double functionalization show a rectangular profile, which again suggests chemical CO2 adsorption in all cases. Nonetheless the CO2 adsorption capacity is higher than those obtained with PEI impregnation only, which suggests an increase of available amine sites or a higher proportion of primary amines (more efficient for CO2 capture) coming from the grafting with APTES. Thus, the raw sepiolite treated with the double functionalization reaches a CO2 uptake of 1.41 mmol g−1 at 760 mmHg and 25 °C, while the raw palygorskite attains 1.04 mmol g-1 under the same conditions. Acid treatment does not improve the adsorption capacity of either sepiolite or palygorskite, leading to CO2 uptakes of 1.10 mmol g−1 and 0.79 mmol g−1 for Sep 8 20 A40 P and Pal 8 20 A40 P, respectively, at 760 mmHg and 25 °C. The single functionalization by grafting with APTES led to higher CO2

ammonium carbamate, giving a theoretical maximum amine efficiency of 0.5. The obtained values (0.08-0.12) for CO2/N efficiency (Table 5) are lower than those obtained for the samples functionalized with APTES probably because the stacking of the amine-rich polymer hinders the access of CO2 molecules to these N species. 4.4. CO2 adsorption in fibrous clay minerals functionalized with APTES + PEI The isotherms at 25 °C for adsorbents with incorporation of amine species by double functionalization, i.e. by APTES grafting followed by impregnation with PEI, are shown in Fig. 12. As observed for the adsorbents impregnated with PEI in Fig. 11, the isotherms of the samples 4583

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Scheme 2. Mechanism of the CO2 adsorption via zwitterion and carbamate. Fig. 12. CO2 adsorption-desorption isotherms at 25 °C of sepiolite and sepiolite treated with acid treatment functionalized by double functionalization APTES-PEI (A) and palygorskite and palygorskite treated with acid treatment functionalized by double functionalization APTES-PEI (B). Adsorption (filled marks) and desorption (non-filled marks).

Table 6 SIPS parameters of the CO2 adsorption at 25 °C for the samples modified by double functionalization grafting with APTES and later impregnation PEI. Sample

SBET (m2 g−1)

N (%)

q50 (mmol g−1)

q760 (mmol g−1)

qm1 (mmol g−1)

b1*1E5 (mm Hg−1)

m1

qm2 (mmol g−1)

b2 (mm Hg−1)

m2

molCO2/molN (q50)

Sep-20 A40 P Sep 8-20 A40 P Pal-20 A40 P Pal 8-20 A40 P

13 15 7 10

12.92 12.73 13.87 12.97

1.34 1.30 1.00 0.71

1.41 1.10 1.04 0.79

0.11 0.77 0.21 0.21

10.31 0.11 1.12 0.01

1.01 1.12 1.01 1.00

1.37 1.13 1.01 0.75

29.50 34.51 33.21 26.64

1.40 3.31 1.06 1.02

0.15 0.14 0.10 0.08

uptakes for fibrous phyllosilicates that underwent acid treatment; however, their CO2 adsorption capacity was not improved when they were functionalized with PEI probably due to the loss of functional groups (mainly in the case of sepiolite) that would interact with the basket molecule (PEI) and favor its dispersion. In the case of the double functionalization, it seems that the loss of the functional groups plays a similar role to that observed in the single impregnation, worsening CO2 adsorption capacity. The CO2 adsorption isotherms obtained for the samples subject to double functionalization were also fitted to the Dualsite Sips model (Table 6). These isotherms follow a similar trend as that observed for the samples functionalized by single impregnation with PEI. The almost exclusive presence of chemical adsorption sites leads to a strong adsorbate-adsorbent interaction, as indicated by the b2 values, which are in the same range of those samples only impregnated with PEI. In addition, the homogeneity of the adsorption increases since all the clay surface is covered with the amine-rich polymer so there are no preferential sites for the CO2 adsorption. Finally, the CO2/N efficiency also displays similar values to those obtained for the samples impregnated with PEI. These data are not in agreement with other authors [23], which detected an improvement in the efficiency using the double functionalization in mesoporous silicas. The adsorbents with the highest CO2 adsorption uptakes, i.e. Sep,

Sep 8–20 A, Sep-40 P and Sep-20 A40 P, were then chosen to carry out experiments to assess the influence of the temperature on CO2 adsorption (Fig. 13). In the case of the raw sepiolite (Fig. 13A), CO2 uptake readily decreases with increasing temperature, from 1.48 mmol CO2 g−1 at 25 °C to 0.85 mmol CO2 g−1 at 65 °C. This decrease is attributed to the adsorption being ruled by physisorption, which is disfavored by temperature [64]. In fact, the adsorbate-adsorbent interaction (b parameter) decreases with increasing temperatures (see Table 7). For the sample functionalized by grafting showing the highest adsorption capacity (Sep 8–20 A), CO2 uptake hardly suffers any change when the temperature is modified, attaining values in the range 1.24–1.31 mmol CO2 g−1 (Fig. 13B). Previous reports have pointed out that APTES grafting leads to the coexistence of chemical and physical sites in the CO2 adsorption [44,66]. As previously indicated, increasing temperatures decrease the binding efficiency of physical sites. At the same time, increasing temperatures improve the efficiency of chemical sites, because the interactions between the –NH2 species located on the adsorbent surface and the CO2 molecules are favored. These effects on chemical and physical sites with respect to temperature compensate one another, leading to similar CO2 adsorption capacities in all cases. For the sample functionalized by PEI impregnation (Fig. 13C), CO2 adsorption is thought to be exclusively attributed to chemisorption so 4584

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

Fig. 13. CO2 adsorption-desorption isotherms at 25 °C, 45 °C and 65 °C of raw sepiolite (A), sepiolite modified with acid treatment (8 min) and functionalized with APTES (B), raw sepiolite functionalized with PEI (C) and raw sepiolite functionalized with APTES-PEI (D). Adsorption (filled marks) and desorption (nonfilled marks).

Table 7 SIPS parameters of the CO2 adsorption at 25 °C–65 °C for Sep, Sep 8–20 A, Sep-40 P and Sep-20 A40 P. Sample

Temp (ºC)

q50 (mmol g−1)

q760 (mmol g−1)

qm1 (mmol g−1)

b1*1E4 (mm Hg−1)

m1

qm2 (mmol g−1)

b2 (mm Hg−1)

m2

molCO2/molN (q50)

Sep Sep Sep Sep 8-20 A Sep 8-20 A Sep 8-20 A Sep-40 P Sep-40 P Sep-40 P Sep-20 A40 P Sep-20 A40 P Sep-20 A40 P

25 45 65 25 45 65 25 45 65 25 45 65

0.35 0.16 0.07 0.89 0.90 0.88 0.91 0.98 1.24 1.34 1.59 1.93

1.48 1.13 0.85 1.30 1.31 1.24 1.02 1.07 1.43 1.41 1.65 2.07

2.50 2.49 2.49 1.04 1.02 0.99 0.24 0.23 0.22 0.11 0.08 0.06

20.01 10.05 5.99 0.40 0.29 0.11 9.03 7.01 6.95 1.03 0.79 0.61

1.24 1.08 1.08 1.02 1.05 1.00 1.05 1.21 1.19 1.01 1.00 1.00

– – – 1.82 1.72 1.71 0.94 1.01 1.32 1.37 1.68 2.09

– – – 13.71E-3 17.01E-3 17.99E-3 20.49 22.03 22.05 29.50 41.21 53.34

– – – 3.01 2.99 2.72 2.11 2.15 2.10 1.40 2.58 2.99

– – – 0.41 0.41 0.41 0.12 0.13 0.17 0.15 0.17 0.21

5. Conclusions

that CO2 uptake at 760 mmHg improves with increasing temperatures from 1.02 mmol g−1 at 25 °C to 1.43 mmol g−1 at 65 °C. Previous publications have pointed out that increasing temperature enhances the diffusion of CO2 molecules within the adsorbent due to a rearrangement of the amine-rich polymer, increasing the proportion of available amine sites on the surface of the adsorbent [20]. This leads to an increase in the adsorbate-adsorbent interaction, as shown by the b value, and an enhancement in CO2/N efficiency from 0.12 to 0.17 (Table 7). The influence of the temperature on the adsorption isotherms for the sample subject to double functionalization (Fig. 13D) follows the same trend as the sample only impregnated with PEI (Fig. 13C), where the rearrangement of PEI upon a temperature rise diminishes its stacking on the surface and favors the diffusion and chemical adsorption of CO2 molecules. The improvement is clear for the double functionalization since the highest uptake at 760 mmHg is obtained (2.07 mmol CO2 g−1 at 65 °C) with an increase in CO2/N efficiency from 0.15 to 0.21 and a stronger adsorbate-adsorbent interaction (Table 7).

Inexpensive fibrous phyllosilicates (sepiolite and palygorskite) have been used as starting materials to yield potential adsorbents for CO2 capture. In a first step, the textural properties of both materials were modified by microwave-assisted acid treatment using a short activation time (8 min). The characterization of these materials reveals that sepiolite is more susceptible to the acid treatment. Mg2+ coming from the octahedral layer is leached and amorphous silica is formed, increasing the SBET and pore volume, whereas palygorskite is more resilient to acid treatment. The evaluation of the CO2 adsorption on these fibrous materials show that raw sepiolite has high potential for CO2 capture attaining uptakes of 1.48 mmol g−1 at 760 mm Hg and 25 °C, while palygorskite shows worse performance for CO2 adsorption due to its lower microporosity. This indicates that a low-cost material such as sepiolite may be used for pressure swing adsorption (PSA) processes due to the presence of nanocavities that act as inexpensive molecular sieve. The incorporation of amine species by grafting with APTES or by impregnation with PEI causes a decrease of the CO2 adsorption, although the amount of chemisorbed CO2 increases. Acid treatment only 4585

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

improves the functionalization by grafting due to a slight increase of the pore size due to the partial or total blockage of the nanochannels of the fibrous clay minerals. The raw materials led to higher CO2 adsorption values when the adsorbents were impregnated with larger proportions of PEI. The decrease in surface area and pore volume suggests that CO2 adsorption must take place on the outer surface. The double functionalization (APTES-PEI) led to the highest adsorption capacity due to the higher amount of available amine sites, which favors the chemical interaction with the CO2 molecules. Increasing temperatures improved the adsorption capacity at 760 mmHg, reaching a maximum uptake of 2.07 mmol g−1 for Sep-20 A40 P, thanks to a rearrangement of the amine-rich polymer favoring the diffusion of the CO2 molecules in the adsorbent, enhancing the CO2/N efficiency. This value is obtained at low absolute pressures so Sep-20 A40 P sample has high potential to be used in vacuum swing adsorption (VSA) processes. The CO2 adsorption capacity of these fibrous clay minerals, shown in the present research, takes place using pure CO2. These adsorption conditions can be considered as unreal since the flue gas is a mixture CO2/air or the extraction of natural gas contains CO2/CH4. However, the adsorption of pure CO2 displays a high potential to the selection of adsorbents with potential to be implanted at larger scale.

[15] A. Garcia-Gallastegui, D. Iruretagoyena, V. Gouvea, M. Mokhtar, A.M. Asiri, S.N. Basahel, S.A. Al-Thabaiti, A.O. Alyoubi, D. Chadwick, M.S.P. Shaffer, Graphene oxide as support for layered double hydroxides: enhancing the CO2 adsorption capacity, Chem. Mater. 24 (2012) 4531–4539. [16] J.R. Li, Y. Ma, M.C. McCarthy, J. Sculley, J. Yu, H.K. Jeong, P.B. Balbuena, H.C. Zhou, Carbon dioxide capture-related gas adsorption and separation in metalorganic frameworks, Coord. Chem. Rev. 255 (2011) 1791–1823. [17] K. Sumida, D.L. Rogow, J.A. Mason, T.M. McDonald, E.D. Bloch, Z.R. Herm, T.H. Bae, J.R. Long, Carbon dioxide capture in metal-organic frameworks, Chem. Rev. 112 (2012) 724–781. [18] E. Haque, M. Islam, E. Pourazadi, S. Sarkar, A.T. Harris, A.I. Minett, E. Yanmaz, S.M. Alshehri, Y. Ide, K.C.W. Wu, Y.V. Kaneti, Y. Yamauchi, S.A. Hossain, Boronfunctionalized graphene oxide-organic frameworks for highly efficient CO2 capture, Chem. Asian J. 12 (2017) 283–288. [19] W.C. Chu, B.P. Bastakoti, Y.V. Kaneti, J.G. Li, H.R. Alamri, Z.A. Alothman, Y. Yamauchi, S.W. Kuo, Tailored design of bicontinuous gyroidme soporous carbon and nitrogen-doped carbon from poly(ethylene oxide-b-caprolactone) diblock copolymers, Chem. Eur. J. 23 (2017) 13734–13741. [20] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Smith, C.T.W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, A new family of mesoporous molecular sieves prepared with liquid crystal templates, J. Am. Chem. Soc. 114 (1992) 10834–10843. [21] N. Hiyoshi, Y. Yogo, T. Yashima, Adsorption characteristics of carbon dioxide on organically functionalized SBA-15, Micropor. Mesopor. Mater. 84 (2005) 357–365. [22] R. Sanz, G. Calleja, A. Arencibia, E.S. Sanz-Pérez, CO2 adsorption on branched polyethyleneimine-impregnated mesoporous silica SBA-15, Appl. Surf. Sci. 256 (2010) 5323–5328. [23] E.S. Sanz-Pérez, M. Olivares-Marín, A. Arencibia, R. Sanz, G. Calleja, M.M. MarotoValer, CO2adsorption performance of amino-functionalized SBA-15 under postcombustion conditions, Int. J. Greenh. GasControl 17 (2013) 366–375. [24] E. Vilarrasa-García, J.A. Cecilia, E.M. Ortigosa-Moya, C.L. Cavalcante Jr, D.C.S. Azevedo, E. Rodríguez-Castellón, “Low-cost” pore expanded SBA-15 functionalized with amine groups applied to CO2 adsorption, Materials 8 (2015) 2495–2513. [25] E. Vilarrasa-García, E.M. Ortigosa-Moya, J.A. Cecilia, C.L. Cavalcante Jr, J. Jiménez-Jiménez, D.C.S. Azevedo, E. Rodríguez-Castellón, CO2 adsorption on amine modified mesoporous silicas: effect of the progressive disorder of the honey comb arrangement, Micropor. Mesopor. Mater. 209 (2015) 172–183. [26] X. Xu, C. Song, J.M. Andresen, B.G. Miller, A.W. Scaroni, Novel polyethy leniminemodified mesoporous molecular sieve of MCM-41 type as high-capacity adsorbent for CO2 capture, Energy Fuels 16 (2002) 1463–1469. [27] M.R. Mello, D. Phanon, G.Q. Silveira, P.L. Llewellyn, C.H. Ronconi, Amine-modified MCM-41 mesoporous silica for carbon dioxide capture, Micropor. Mesopor. Mater. 143 (2011) 174–179. [28] S.C.G. Santos, A.M. Garrido Pedrosa, M.J.B. Souza, J.A. Cecilia, E. RodríguezCastellón, Carbon dioxide adsorption on micro-mesoporous composite materials of ZSM-12/MCM-48 type: the role of the contents of zeolite and functionalized amine, Mater. Res. Bull. 70 (2015) 663–672. [29] J.A. Cecilia, E. Vilarrasa-García, C. García-Sancho, R.M.A. Saboya, D.C.S. Azevedo, C.L. Cavalcante Jr, E. Rodríguez-Castellón, Functionalization of hollow silica microspheres by impregnation or grafted of amine groups for the CO2 capture, Int. J. Greenh. Gas Control 52 (2016) 344–356. [30] E.S. Sanz-Pérez, A. Arencibia, G. Calleja, R. Sanz, Tuning the textural properties of HMS mesoporous silica. Functionalization towards CO2 adsorption, Micropor. Mesopor. Mater. 260 (2018) 235–244. [31] R. Kishor, A.K. Gholar, Polyethy lenimine functionalized as-synthesized KIT-6 adsorbent for highly CO2/N2selectiveseparation, Energy Fuels 30 (2016) 9635–9644. [32] X. Yan, L. Zhang, Y. Zhang, K. Qiao, Z. Yan, S. Komarnemi, Amine-modified mesocellular silica foams for CO2 capture, Chem. Eng. J. 168 (2011) 918–924. [33] X. Feng, G. Hu, X. Hu, G. Xie, Y. Xie, J. Lu, M. Luo, Tetraethylenepentaminemodified siliceous mesocellular foam (MCF) for CO2 capture, Ing. Eng. Chem. Res. 52 (2013) 4221–4228. [34] G. Li, P. Xiao, P. Webley, J. Zhang, R. Singh, M. Marshall, Capture of CO2 from high humidity flue gas by vacuum swing adsorption with zeolite 13X, Adsorption 14 (2008) 415–422. [35] J. Merel, M. Clausse, F. Meurier, Experimental investigation on CO2 post-combustion capture by indirect thermal swing adsorption using 13X and 5A zeolites, Ind. Eng. Chem. Res. 47 (2008) 209–215. [36] F. Su, C. Lu, S.C. Kuo, W. Zeng, Adsorption of CO2onamine-functionalized Y-type zeolites, Energy Fuels 24 (2010) 1441–1448. [37] M.G. Plaza, C. Pevida, A. Arenillas, F. Rubiera, J.J. Pis, CO2 capture by adsorption with nitrogen enriched carbons, Fuel 86 (2007) 2204–2212. [38] C. Pevida, M.G. Plaza, B. Arias, J. Fermoso, F. Rubiera, J.J. Pis, Surface modification of activated carbons for CO2 capture, Appl. Surf. Sci. 254 (2008) 7165–7172. [39] M.W. Ackley, S.U. Rege, H. Saxena, Application of natural zeolites in the purification and separation of gases, Micropor. Mesopor. Mater. 61 (2003) 25–42. [40] A. Vaccari, Clays and catalysis: a promising future, Appl. Clay Sci. 14 (1999) 161–198. [41] Y.H. Chen, D.L. Lu, Amine modification on kaolinites to enhance CO2 adsorption, J. Colloid Interface Sci. 436 (2014) 47–51. [42] M.L. Pinto, L. Magra, J.M. Guil, J. Pires, J. Rocha, Adsorption and activation of CO2 by amine-modified nanoporous materials studied by solid-state NMR and 13CO2 adsorption, Chem. Mater. 23 (2011) 1387–1395. [43] C. Chen, D.W. Park, W.S. Ahn, Surface modification of a low cost bentonite for postcombustion CO2 capture, Appl. Surf. Sci. 283 (2013) 699–704. [44] E. Vilarrasa-García, J.A. Cecilia, D.C.S. Azevedo, C.L. Cavalcante Jr, E. Rodríguez-

Acknowledgements We thank project CTQ2015-68951-C3-3-R (Ministerio de Economía y Competitividad, Spain) and FEDER funds, project P12-RNM-1565 (Junta de Andalucía, Spain) and European project 295156 FP7PEOPLE-2011-IRSES for financial support. E.Vilarrasa-Garcia thanks CAPES/PNPD (Brazilian Ministry of Education) for post-doctoral fellowship. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jece.2018.07.001. References [1] Intergovernmental Panel on Climate Change, Climate Change: Mitigation of Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2014. [2] O. Hoegh-Guldberg, J.F. Bruno, The impact of climate change on the world’s marine ecosystems, Science 328 (2010) 1523–1528. [3] Intergovernmental Panel on Climate Change, Impacts, Adaptation, and Vulnerability. Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2014. [4] M. Pera-Titus, Porous inorganic membranes for CO2 capture: present and prospects, Chem. Rev. 114 (2014) 1413–1492. [5] A. Brunetti, F. Scura, G. Barbieri, E. Drioli, Membrane technologies for CO2 separation, J. Membr. Sci. 359 (2010) 115–125. [6] R. Venemana, H. Kamphuis, D.W.F. Brilman, Post-combustion CO2 capture using supported amine sorbents: a process integration study, Energy Procedia 37 (2013) 2100–2108. [7] S. Choi, J.H. Drese, C.W. Jones, Adsorbent materials for carbon dioxide capture from large anthropogenic point sources, ChemSusChem 2 (2009) 796–854. [8] P.V. Danckwerts, Reaction of CO2 with ethanol amines, Chem. Eng. Sci. 34 (1979) 443–446. [9] E. Rinker, S.S. Ashour, O.C. Sandall, Absorption of carbon dioxide into blends of diet hanol amine and methyl diet hanol amine, Ind. Eng. Chem. Res. 39 (2000) 4346–4356. [10] J.C. Abanades, E.J. Anthony, J. Wang, J.E. Oakley, Fluidized bed combustion systems integrating CO2 capture with CaO, Environ. Sci. Technol. 39 (2005) 2861–2866. [11] G.S. Grasa, J.C. Abanades, CO2 capture capacity of CaO in long series of carbonation/calcination cycles, Ind. Eng. Chem. Res. 45 (2006) 8846–8851. [12] E. Abbasi, A. Hassanzabeh, S. Zarghami, H. Arastoopour, J. Abbasian, Regenerable MgO-based sorbent for high temperature CO2 removal from syngas: 3. CO2 capture and sorbent enhanced water gas shift reaction, Fuel 137 (2014) 260–268. [13] A.T. Vu, Y. Park, P.R. Jeon, C.H. Lee, Mesoporous MgO sorbent promoted with KNO3 for CO2 capture at intermediate temperatures, Chem. Eng. J. 258 (2014) 254–264. [14] A. Hanif, S. Dasgupta, S. Divekar, A. Arya, M.O. Garg, A. Nanoti, A study on high temperature CO2 capture by improved hydrotalcite sorbents, Chem. Eng. J. 236 (2014) 91–99.

4586

Journal of Environmental Chemical Engineering 6 (2018) 4573–4587

J.A. Cecilia et al.

[45]

[46]

[47] [48] [49] [50]

[51] [52] [53] [54]

[55] [56]

[57] J. Landers, G.Y. Gor, A.V. Neimark, Density functional theory methods for characterization of porous materials, Colloid Surface A 437 (2013) 3–32. [58] I. Langmuir, The adsorption of gases on plane surfaces of glass, mica and platinum, J. Am. Chem. Soc. 40 (1918) 1361–1403. [59] D.D. Do, Adsorption Analysis: Equilibria and Kinetics, Imperial College Press, 1998. [60] R.L. Frost, O.B. Locos, H. Ruan, J.T. Kloprogge, Near-infrared and mid-infrared spectroscopic study of sepiolites and palygorskites, Vib. Spectrosc. 27 (2001) 1–13. [61] V.C. Farmer, The Infrared Spectra of Minerals, Mineralogical Society of London, 1974. [62] J. Madejová, FTIR techniques in clay mineral studies, Vibr. Spectrosc. 31 (2003) 1–10. [63] M.A. Vicente-Rodríguez, M. Suárez, M.A. Bañares-Muñoz, J.D. López-Suárez, Comparative FT-IR study of the removal of octahedral cations and structural modifications during acid treatment of several silicates, Spectrochim. Acta A 52 (1996) 1685–1694. [64] M. Thommes, K. Kaneko, A.V. Neimark, J.P. Oliver, F. Rodríguez-Reinoso, J. Rouquerol, K.S.W. Sing, Physisorption of gases, with special reference to the evaluation of surface area and pore size distribution (IUPAC technical report), Pure Appl. Chem. 87 (2015) 1051–1069. [65] E. Galán, P. Aparicio, A. Miras, Sepiolite and palygorskite as sealing materials for geological storage of carbon dioxide, Dev. Clay Sci. 3 (2011) 375–392. [66] A. Sayari, Y. Belmabkhout, R. Serna-Guerrero, Flue gas treatment via CO2 adsorption, Chem. Eng. J. 171 (2011) 760–774. [67] X. Wang, V. Schwartz, J.C. Clark, X. Ma, S.H. Overbury, X. Xu, C. Song, Infrared study of CO2 sorption over “molecular basket” sorbent consisting of polyethylenimine-modified mesoporous molecular sieve, J. Phys. Chem. C 113 (2009) 7260–7268. [68] E. Vilarrasa-García, J.A. Cecilia, M. Bastos-Neto, C.L. Cavalcante Jr, D.C.S. Azevedo, E. Rodríguez-Castellón, CO2/CH4 adsorption separation process using pore expanded mesoporous silicas functionalizated by APTES grafting, Adsorption 21 (2015) 565–575.

Castellón, Evaluation of porous clay heterostructures modified with amine species as adsorbent for the CO2 capture, Micropor. Mesopor. Mater. 249 (2017) 25–33. M. Irani, M. Fan, H. Ismail, A. Tuwati, B. Dutcher, A.G. Russell, Modified nanosepiolite as an inexpensive support of tetraethylenepentamine for CO2 sorption, Nano Energy 11 (2015) 235–246. E. Vilarrasa-García, J.A. Cecilia, M. Bastos-Neto, C.L. Cavalcante Jr, D.C.S. Azevedo, E. Rodríguez-Castellón, Microwave-assisted nitric acid treatment of sepiolite and functionalization with polyethylenimine applied to CO2 capture and CO2/N2 separation, Appl. Surf. Sci. 410 (2017) 315–325. M. Súarez, E. García-Romero, Advances in the crystal chemistry of sepiolite and palygorskite, Dev. Clay Sci. 3 (2011) 33–65. S.W. Bailey, Structures of Layer Silicates. Crystal Structures of Clay Minerals and Their X-Ray Identification, Mineralogical Society, London, 1984. B.F. Jones, E. Galán, Sepiolite and paly gorskite, Rev. Mineral. 19 (1988) 631–674. F. Franco, M. Pozo, J.A. Cecilia, M. Benítez-Guerrero, E. Pozo, J.A. Martín Rubí, Microwave assisted acid treatment of sepiolite: the role of composition and “crystallinity”, Appl. Clay Sci. 102 (2014) 15–27. A.J. Ward, K. Arola, E.T. Brewster, C.M. Mehta, D.J. Batstone, Nutrient recovery from wastewater through pilot scale electrodialysis, Water Res. 135 (2018) 57–65. M.Y. Menetrez, An overview of algae biofuel production and potential environmental impact, Environ. Sci. Technol. 46 (2012) 7073–7085. R. White, R. Ryan, Long-term cultivation of algae in open-raceway ponds: lessons from the field, Ind. Biotechnol. 4 (2015) 212–220. F. Franco, M. Pozo, J.A. Cecilia, M. Benítez-Guerrero, M. Lorente, Effectiveness of microwave assisted acid treatment on dioctahedral and trioctahedral smectites. The influence of octahedral composition, Appl. Clay Sci. 120 (2016) 70–80. S. Brunauer, P.H. Emmett, E. Teller, Adsorption of gases in multi molecular layers, J. Am. Chem. Soc. 60 (1938) 309–319. J.H. de Boer, B.C. Lippens, B.G. Linsen, J.C.P. Broekhoff, A. vandenHeuvel, T.J. Osinga, The t-curve of multi molecular N2-adsorption, J. Colloid Interface Sci. 21 (1966) 405–414.

4587