Incorporation of monoethanolamine (MEA), diethanolamine (DEA) and methyldiethanolamine (MDEA) in mesoporous silica: An alternative to CO2 capture

Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

Contents lists available at ScienceDirect

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

Incorporation of monoethanolamine (MEA), diethanolamine (DEA) and methyldiethanolamine (MDEA) in mesoporous silica: An alternative to CO2 capture Simone G. de Ávilaa,* , Marco A. Loglia , Luís Carlos C. Silvab , Márcia C.A. Fantinic , Jivaldo R. Matosa a

Institute of Chemistry, University of São Paulo, 05508-000, São Paulo, SP, Brazil Institute of Chemistry, University of Brasília, 70910-900, Brasília, DF, Brazil c Institute of Physics, University of São Paulo, 05508-090, São Paulo, SP, Brazil b

A R T I C L E I N F O

Article history: Received 3 May 2016 Received in revised form 1 September 2016 Accepted 13 October 2016 Available online 14 October 2016 Keywords: CO2 capture Ethanolamine Adsorption SBA-15

A B S T R A C T

Monoethanolamine (MEA), diethanolamine (DEA) and methyldiethanolamine (MDEA) are substances used in the CO2 capture industrial process. The inconvenience of their use is that these species have low thermal stabilities, are volatile, and in the presence of air undergo oxidation processes to form corrosive products. This work proposed the incorporation of MEA, DEA and MDEA in a mesoporous silica substrate as an alternative for CO2 adsorption. The mesoporous materials have a high surface area and a large pore size. The high material porosity enabled the incorporation of a large quantity of ethanolamine. The presence of MEA inside the silica pores promoted the acceleration of the decomposition process. The materials prepared using a mixture of mesoporous silica and DEA and MDEA had an optimal CO2 capture efficiency than the isolated ethanolamines. This increased CO2 capture efficiency was most significant in the sample prepared using MDEA. The increase of CO2 efficiency capture by the incorporated material was 30% higher compared with free MDEA. DEA and MDEA incorporated in mesoporous silica had a higher CO2 capture efficiency in the second and third adsorption cycles than free DEA and MDEA. The study showed that the incorporation of DEA and/or MDEA in mesoporous silica may increase CO2 capture efficiency. ã 2016 Elsevier Ltd. All rights reserved.

1. Introduction The increase of CO2 concentration is the principal reason for global climate change. Aiming to reduce the atmospheric CO2 concentration, some materials capable of CO2 absorption/adsorption have been developed. Liquid ethanolamine solutions, principally monoethanolamine (MEA), diethanolamine (DEA) and methyldiethanolamine (MDEA), are utilized to capture CO2 in industrial processes. The CO2 capture occurs by passing gaseous fluid through the ethanolamine solution, producing carbamates and/or bicarbonates [1].

* Corresponding author at: Institute of Chemistry, University of São Paulo, Av. Professor Lineu Prestes, 748, Butantã São Paulo, SP, Brazil, CEP: 05508-000, Brazil. E-mail address: [email protected] (S.G. de Ávila). http://dx.doi.org/10.1016/j.jece.2016.10.015 2213-3437/ã 2016 Elsevier Ltd. All rights reserved.

While ethanolamines have a high capacity for CO2 capture, they are volatile substances that are easily oxidized. The oxidation of ethanolamines can produce carboxylic acids, ammonia and other corrosive substances. The regeneration of the solutions involves high energy consumption and there is the inconvenience of the work using liquid substances [1–4]. The presence of oxygen in a gaseous fluid can promote the degradation of ethanolamines, producing ammonia and a mixture of organic compounds such as organic acids and aldehydes, especially acetic acid and formic acid. This is an irreversible process, so it is necessary to replace the starting material [5]. The efficiency of CO2 capture is inversely related to the thermal stability. Steric impediment inhibits the reaction between the ethanolamine and CO2. Thus, although MEA is less stable than the other ethanolamines, it has the greater CO2 capture efficiency. The thermal decomposition of DEA involves an intramolecular reaction

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

that produces MEA and ethylene oxide. The presence of the methyl radical increases the thermal stability of MDEA, due to steric impediment [6]. Due to the low thermal stability of ethanolamines, it is necessary to develop solids that are CO2 adsorbent in order to obtain a stable material that is efficient at CO2 capture. Porous solid materials can be used as adsorbent surfaces, catalysts, catalytic supports, pharmaceutical excipients, and chromatographic support, among others. These applications are possible because these materials have a large pore size, high surface area and ordered pore distribution. According to the IUPAC definition, porous materials are divided into three classes based on the pore diameter [7–10]: microporous (<2 nm), mesoporous (2–50 nm) and macroporous (>50 nm). The incorporation of ethanolamines in microporous materials has already been studied. The presence of Al in the zeolite structure increases the polarity of the material, then; the interaction between the material structure and water molecules is more intense in this kind of solids. However, because of their highly hydrophilic character, the flue gas needs extensive drying prior to CO2 capture. The other inconvenient about the incorporation of ethanolamine in microporous materials is the limitation of pore sizes (<2 nm) [1,11]. In addition to low cost, ideally an adsorption medium for CO2 removal and recovery at ambient temperature and high pressure should combine high CO2 uptake, complete regeneration under mild condition, high thermal stability, favourable adsorption– desorption kinetics and selective CO2 adsorption. Current research activity on CO2 capture processes focused on oxides and mixed oxides, high surface area porous materials such zeolitas, carbon, metal organic frameworks (MOFs), organo- silica and surfacemodifield silica [11–17]. It is however intriguing that despite the significant growth in the area of periodic mesoporous materials since their discovery there are only few studies devoted to CO2 adsorption on plain mesoporous silica such as SBA-15. Surface grafting method makes the interaction between supported materials and functional groups much stronger and stable so that the adsorbents can be operated at relatively high temperature. A series of mesoporous materials grafted with aminosilanes were prepared to enhance CO2 adsorption performance [18]. The synthesis of grafting materials is more sophisticate and involve high cost. The amine impregnated in mesoporous materials can be a more interesting alternative because it is not necessary the amine functionalization in the silica structure. The large pore sizes of the mesoporous material allow the incorporation of a large quantity of ethanolamine. Moreover, the encapsulation of these substances inside the silica pores can increase their thermal and chemical stability. The incorporation of MEA, DEA and/or MDEA in mesoporous materials may provide an interesting and new alternative to CO2 capture, as demonstrated in this work. The pore size, pore volume and surface area of supported materials were observed to play an important role in the CO2 capture process. Due to the high pore volume and the large pore size, it is possible to incorporate a large quantity of ethanolamine in the solid materials, which may in turn increase the thermal stability of ethanolamines. This study proposed the incorporation of MEA, DEA and MDEA in a SBA-15 silica, which has a 2D configuration and a hexagonal pore structure that is synthesized using the triblock copolymer Pluronic1 P123 as a template [19]. SBA-15 presents a unique structure with its hexagonally ordered mesoporous structure connected by irregular micropores [20]. A thermal analytic study of the materials produced by the incorporation of ethanolamine in the SBA-15 silica was conducted and the efficiency of these materials for the CO2 capture process was tested.

4515

2. Materials and methods 2.1. Adsorbent synthesis 2.1.1. SBA-15 synthesis SBA-15 silica was prepared based on established techniques [19] with a minor modification; it was synthesized with triblock poly(ethylene oxide)–poly(propylene oxide)–poly(ethylene oxide) (EO20–PO70–EO20, Pluronic P123; BASF) copolymer as a structure directing agent (template) and tetraethyl orthosilicate (TEOS; Aldrich, USA) as a silica source. TEOS was added dropwise into the homogeneous solution. The mixture was maintained in a Teflon bottle and it was subjected to hydrothermal treatment at 100
C for 48 h. The solid product (SBA-15) was filtered, washed with distilled water and air-dried at room temperature. SBA-15 was heated in a tube furnace under air by slowly increasing the temperature at 1
C min1 from room temperature to a target temperature of 550
C under a nitrogen atmosphere and calcined at the target temperature for 6 h under an air atmosphere. The template, Pluronic P123, has surfactant properties. The following stoichiometric molar ratios for the SBA-15 synthesis were used: 1.0 TEOS: 0.0167 Pluronic P123: 5.82 HCl: 190 H2O. Two major routes for the formation of silica- based ordered mesoporous materials have been described in the literature. According to the first route, the surfactant molecules organize independently of the inorganic species into a liquid crystal phase; the silicate ions condense and polymerase around the pre-formed structure. The second one is the cooperative templating mechanism and, in this case, the building blocks are the micelles; so the cooperative templating mechanism occurs at low surfactant concentrations. In such a mechanism, the interactions between the surfactant and the inorganic precursor are responsible for the mesoporous materials formation. The synthesis can be achieved through either an electrostatic pathway, based on a supramolecular assembly of charged surfactants with charged inorganic precursors or a neutral pathway (S0 H+  X I +), in which hydrogen bonding is responsible for the cohesiveness between the surfactant and the inorganic precursor [21–23]. Therefore, the interaction between Pluronic P123 and the acid specie are important during the mesoporous silica formation. The ionic force of the solution influences the mesoporous structure, so the acid concentration and the anionic species are important to form the material structure [24]. 2.1.2. Ethanolamine incorporation in SBA-15 silica The MEA and DEA studied samples were donated by Oxiteno S. A., and the MDEA sample was donated by Dow of Brazil S.A. Mixtures of SBA-15/MEA, SBA-15/DEA and SBA-15/MDEA were prepared with different mass ratios, starting at a 1:1 mass proportion until the addition of a large amount of ethanolamine was possible while conserving the material in a solid state. Mixtures made using MEA were made between 1:1 and 1:3 mass proportions; mixtures made using DEA were made between 1:1 and 1:4 mass proportions; mixtures with MDEA were made only at 1:1 and 1:2 mass proportions. The ethanolamine incorporation was performed by adding MEA, DEA or MDEA to a SBA-15 and acetone suspension. The amount of acetone used was sufficient to dissolve the ethanolamine. The mixture was prepared by magnetic stirring until the formation of a pasty material, and then dried in a desiccator under reduced pressure. 2.2. Characterization The morphology of the samples was studied with a scanning electron microscope (SEM) (Quanta 600 FEI) and the pore

4516

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

distribution images were studied with a transmission electron microscope (TEM) (LEO 906e). Small angle x-ray scattering (SAXS) characterization was performed using a Bruker AXS Nanostar small angle X-ray scattering camera. This camera has a microfocus Genix 3D system (source + focusing mirrors), two scatterless slits sets for collimation and a 2D Vantec-2000 detector. The X-ray wavelength was l = 0.15418 nm (Cu Ka radiation), the cross section of the beam was 0.8  0.8 mm2 and sample to detector distance were 66.7 cm. The range of scattering vectors q (q = (4p/l)sinu, where 2u is the scattering angle) was 0.01 Å1 to 0.35 Å1. The samples, in powder form, were placed between two mica sheets and the measurements were performed at room temperature for exposure times of 600 s. The diffraction pattern was indexed according to the p6 mm symmetry group, corresponding to a hexagonal bidimensional structure of pores in the ordered portions of the sample. The diffraction peaks were isolated from the background through the subtraction of a straight-line background under the (100) peak and another for the (110) and (200) peaks. Nitrogen adsorption isotherm measurements were executed using a Micromeritics ASAP 2010. The samples were degassed for 2 h at 100
C, with the data acquisition being performed at 196
C, with relative pressure ranging from 10–6 to 0.995. Specific surface areas were calculated with the Brunauer–Emmett–Teller (BET) equation and the pore size distribution was obtained using the Barrett–Joyner–Halenda (BJH) method. 2.3. Adsorbent thermal stability 2.3.1. Initial study Thermogravimetric (TG) tests were performed using a thermogravimetric analyzer (Shimadzu, model TGA-51), in a synthetic air atmosphere (50 mL min1) over a temperature range of 25–900
C and a heating rate (b) of 10
C min1. A platinum (Pt) crucible containing 15 mg of the studied samples was used. The equipment was calibrated using a CaC2O4H2O (Sigma Aldrich) analytical standard. The Fourier transform infrared spectra (FTIR) technique was used to understand the chemical changes during the different stages of material decomposition. The FTIR spectra were taken with a FTIR spectrometer (PerkinElmer, Spectrum One) in the range of 4000–400 cm1 to monitor the template extraction. 2.3.2. Kinetic study of thermal decomposition (SBA-15/DEA) 2.3.2.1. Ozawa method. The Arrhenius equation (Eq. (1)) shows the relationship between the rate constant and the temperature of reaction. kðTÞ ¼ AeðEa =RTÞ

ð1Þ

where k is the rate constant, A is a frequency factor, Ea is the activation energy (J mol1), R is the gas constant (8.314 J mol1 K1) and T is temperature (K). The kinetic study of DEA incorporated in SBA-15 silica (1:1) was done considering the Thermogravimetric Non-Isothermal Method (Ozawa Method) [25]. In the Ozawa method, the decomposed fraction (a) during the thermogravimetric analysis is defined by Eq. (2):

a¼

ðm0  ma Þ ðm0  mf Þ

ð2Þ

where, m0 is the initial mass, ma is the resulting mass considering a certain decomposed mass fraction and mf is the final mass (residual mass). The decomposition rate (da/dt) can be described

according to Eq. (3): da ¼ kðTÞf ðaÞ dt

ð3Þ

where k(T) is the Arrhenius rate constant as defined in Eq. (1) and f (a) is a function related to the reaction mechanism. The heating rate (b) is defined as the relation between the temperature variation and the time: dT ¼b dt

ð4Þ

From Eqs. (1) and (3), it follows that: da ¼ AeðEa =RTÞ dT

ð5Þ

Using Eqs. (4) and (5) results in Eq. (6):
 da a ðEa =RTÞ ¼ e dT b

ð6Þ

Further, by integrating Eq. (6), gðaÞ ¼

ZT
 A ðEa =RTÞ dT e T0

b

ð7Þ

This integral does not present an analytical solution. Its resolution is performed by approximate methods. Some resolution methods propose a limit of integration change between the limits   of 0 and T, introducing a functionpðEa R TÞ, valid only forEa R T > 20 [26]. ZT

eðEa =RTÞ dT ¼



 Ea Ea p R RT

ð8Þ

0

Where

 Ea Ea  2:315  0:4567 log p RT RT

ð9Þ

The combination of Eqs. (7)–(9) and, assuming that g(a) is constant to Ta, Ozawa proposed a kinetic treatment [25]:

 AEa Ea  loggðaÞ  0:4567 ð10Þ log b ¼ 2:315 þ log R TRa 2.3.2.2. Experimental condition. The kinetic study of SBA-15/DEA sample thermal decomposition was performed using a TGA-51 (Shimadzu) thermogravimetric analyser in a synthetic N2 atmosphere (50 mL min1). The heating rate (b) was 5.0, 7.5, 10.0, 15.0 and 20.0
C min1 in the temperature range 25–600
C. A Pt crucible containing 15 mg of the studied samples was used for the non-isothermal method (Ozawa Method). 2.4. CO2 capture efficiency CO2 capture tests were performed with MEA, DEA and DMEA in both the free state and incorporated in mesoporous material. 2.4.1. Free ethanolamines These experiments were done using test tubes containing 3.0 g of ethanolamine samples. The tubes were wrapped in aluminium foil and sealed with plastic film, through which a tube was inserted to introduce gas from a CO2 cylinder. The gas flow was 20 mL min1. The CO2 capture was determined by mass measurements every 20 min even when mass variation was not observed. It is important to explain that the CO2 adsorption experiments were done using pure CO2, because the air may cause a matrix

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

4517

Fig. 1. SEM image of SBA-15 with 5000 magnification (a) and TEM image of SBA-15 (b).

Fig. 2. N2 isothermal adsorption and pore size distribution of SBA-15 sample.

modification. The pure CO2 in the fluid allows the comparison of CO2 capture efficiency of ethanolamine and the control of the experimental condition. 2.4.2. Ethanolamine incorporated in SBA-15 A sample port was made using a glass pipette (10 mL capacity), which was cut and filled with adsorbent material and its ends were filled with glass wool. Before the material introduction in the sample port, the sample was heated at 50
C (SBA-15/MEA) and 80
C (SBA-15/DEA and SBA-18/MDEA) until constant mass was obtained. This process was done to eliminate some substances that could be adsorbed in the silica structure. The sample port was connected to the CO2 gas cylinder by silicone hose. The gas flow was 20 mL min1. After the first adsorption stage the samples were heated and a new adsorption process was initiated. 2.4.3. Characterization FTIR was utilized to characterize the samples before and after the CO2 capture process. The FTIR spectra were taken with a FTIR spectrometer (PerkinElmer, Spectrum One) in the range of 4000– 400 cm1 to monitor the chemical changes in the CO2 capture process.

3. Results 3.1. SBA-15 characterization Fig. 1 shows the SEM and TEM images of SBA-15 synthesized by direct calcination. The materials presented rod shapes in an ordered form. The TEM image shows that the materials have an ordered pore distribution, which is organized in symmetrical form and has a hexagonal structure in a 2D symmetry. The nitrogen isothermal adsorption measurements (Fig. 2) show an adsorption profile characteristic of type IV mesoporous materials, according to the IUPAC classification. In Fig. 2, a hysteresis phenomenon is observed, which occurs because the nitrogen condenses in the mesopores and, during desorption, capillary evaporation occurs [27]. In other words, the pores are large enough to allow condensation but small enough to permit capillary action. The pore distribution sizes can be seen in Fig. 2. The specific surface area, pore volume and pore diameter, respectively, of the material is 476 m2 g1 (BET), 0.757 cm3 g1 and 9.5 nm. Fig. 3 shows the SAXS results. There are five indexed peaks related to (100), (110), (200), (210) and (300) diffraction planes,

4518

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

Fig. 3. X-ray diffraction pattern obtained by SAXS of the SBA-15 sample. Table 1 SAXS parameters and wall thickness of SBA-15 (direct calcination). hkl

2u

q/Å1

d/nm

a/nm

t/nm

100 110 200 210 300

0.90 1.56 1.80 2.36 2.68

6.40 11.1 12.8 16.8 19.0

9.82 5.66 4.91 3.74 3.30

11.3 11.3 11.3 11.3 11.3

1.8 1.8 1.8 1.8 1.8

characteristics of ordered hexagonal materials. Table 1 presents the calculated parameters [28–32] such as lattice parameter, wall thickness and pore diameter (obtained by nitrogen isothermal adsorption). All SAXS peaks provided the same wall thickness (1.8 nm), indicating that the material is regularly structured. 3.2. Thermal stability of the adsorbed material 3.2.1. Initial study The TG curves of SBA-15, MEA, and SBA-15/MEA 1:1 (m/m) are illustrated in Fig. 4(a). The SBA-15 curve displayed two stages of weight loss: Dm = 9.8% between 25 and 200
C and Dm = 4.3% between 200 and 900
C. The first event is attributed to adsorbed water elimination. The second event occurs gradually and corresponds to water elimination as a consequence of the dehydration reaction of silanol groups to form siloxane groups. The TG curve of MEA incorporated in SBA-15 also displayed two stage of weight loss. The first stage (25–200
C) had the same

profile as the MEA TG curve and corresponds to a mass variation of 47.8% and is attributed to elimination of MEA present on the SBA15 surface and some water molecules that can remain in the material structure. The second event (Dm = 10.3%) occurs between 200 and 600
C and corresponds to removal of MEA that exists in the porous material. These results show that a part of MEA remains in the pores of SBA-15 and can increase the thermal stability of ethanolamine. Considering the MEA TG curve, it is possible to observe the complete weight loss of this ethanolamine at 167
C. Therefore, the quantity of MEA inside the pores of SBA-15 (10.3%) had its thermal stability increased. Fig. 4(b) illustrates the TG curves of free DEA and DEA incorporated in the SBA-15 matrix (1:1). Two decomposition stages were observed in the free DEA TG curve; in the incorporated material, three stages were observed. The first event (25 and 120
C Dm = 13.1%) is more evident in the TG curves of incorporated material than the TG curve of free DEA. This mass loss is related to the intramolecular reaction between DEA molecules, producing MEA and ethylene oxide [6], and is also attributed to adsorbed water elimination. The second event (Dm = 34.3%) is due to volatilization of DEA that remains on the silica surface. It occurs in the same temperature interval (120–250
C) that free DEA volatilizes/decomposes. The third event of weight loss (Dm = 8.9%, between 250 and 600
C) is related to the DEA decomposition in the silica pores. So, the presence of DEA in the silica pores increased the thermal stability. The thermal decomposition of free DEA is initiated at 25
C and it is finished at 246
C and, the amount of DEA that remains in the silica pores suffers the thermal decomposition until 600
C. The DEA samples were incorporated in silica matrices in order to compare the capacity of DEA incorporation in SBA-15 silica, considering the CO2 capture process. Gel silica and silica obtained by the thermal treatment of rice husks were used, into both of which DEA was incorporated at a 1:1 mass ratio. The incorporation procedure was performed by adopting the same procedure as for DEA incorporation into SBA-15 silica. The silica sample produced by rice husks was donated and its generation was part of a master’s work [28]. The material produced by gel silica presented a pasty texture, due to the reduced porosity of this material. Fig. 5 shows the overlap of TG curves for DEA incorporated in different kinds of silica and SEM images of each silica type. It is possible to observe three weight loss events. The first event (25–120
C) is a consequence of water elimination on the surface of the silica and the DEA intramolecular reaction [6]. The second event is attributed the decomposition/volatilization of DEA present on the surface of silica and the last event is due to elimination of DEA from the silica pores.

Fig. 4. TG curves obtained under a dynamic air atmosphere (50 mL min1), between 25 and 900
C, utilizing Pt0 crucibles and 15 mg of MEA, SBA-15 and SBA-15/MEA (1:1) samples (a) and 15 mg of DEA, SBA-15 and SBA-15/DEA (1:1) samples (b).

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

4519

Fig. 5. TG curves obtained under a dynamic air atmosphere (50 mL min1), between 25 and 900
C, utilizing Pt0 crucibles and 15 mg of DEA incorporated in different kind of silica and SEM images of all silica kind sample. Table 2 Data obtained by TG curves of DEA incorporated in different silica matrix. Sample (1:1 m/m)

%Dm1 (25–102
C)

%Dm2 (102–287
C)

%Dm3 (287–600
C)

m (g)

SBA-15/DEA Rice husk/DEA Gel silica/DEA

10.4 14.7 25.8

44.9 40.7 29.6

13.6 4.2 7.0

0.437 0.104 0.189

Both materials, those produced by gel silica and those produced from the silica synthesized by rice husks, presented a higher quantity of DEA on the surface than the SBA-15 material, indicating that SBA-15 has a larger porosity than the other analysed silica. The SEM images of each silica type (Fig. 5, inset) confirm that they have different morphologies. The quantities of DEA encapsulated in the pores of each silica type were calculated from the TG curves and are

DEA

pores)/m (g) silica

shown in Table 2. The quantity of DEA in the SBA-15 pores is 2.3 times higher than the DEA in the gel silica pores and 4.2 times higher than DEA in the rice husk-derived silica pores; therefore, SBA-15 has a larger porosity than the other studied silica. Fig. 6 shows the TG curves of various mass fractions of DEA incorporated in SBA-15 and the resulting data are shown in Table 3. The profile of TG curves depends on the incorporated DEA quantity.

Fig. 6. TG curves obtained under a dynamic air atmosphere (50 mL min1), between 25 and 900
C, utilizing Pt0 crucibles and 15 mg of DEA incorporated in SBA-15 in different mass proportion.

4520

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

Table 3 Data obtained by TG curves of DEA incorporated in SBA-15 in different mass proportion. SBA-15/DEA (m/m)

Dm1 (%) Dm2 (%) Dm3 (%) DEA (mol) pores/g SBA-15

1:1 1:2 1:3 1:4

13.4 16.3 9.5 10.9

34.3 50.6 60.7 64.1

8.9 5.3 5.6 4.9

0.00195 0.00182 0.00220 0.00232

Using the data of TG curves, it was possible to calculate the quantity of DEA in the SBA-15 pores (Table 3). The quantity of DEA in the pores is independent of the mass fraction of DEA incorporated because, in all samples, the quantity of DEA in the pores is similar, suggesting that the materials present homogenous pore size and pore quantity. The mean quantity of DEA in the SBA15 pores is 0.00207  0.00023 mol DEA per gram of SBA-15. FTIR experiments were performed on SBA-15, DEA and the combination of DEA and SBA-15 (Fig. 7), even during the formation of intermediate products of each weight loss event observed in the TG curves of DEA incorporated in SBA-15 silica. The spectra show absorption bands resulting from fundamental vibrations of the silica structure (460, 810, 960, 1080, 1600 and 3400 cm1). The band localised at approximately 3400 cm1 is characteristic of the silanol (SiOH) group stretch, while the smaller band (1600 cm1) is a consequence of the OH angular deformation of water molecules adsorbed on the silica surface. The band present at 1080 cm1 is characteristic of the Si O Si group stretch and the band at 460 cm1 is due to the angular deformation of this group. The band at 950 cm1 is due to the angular deformation (off-plan) of Si-OH groups. All bands are present in the SBA-15 sample spectrum and in all intermediate products.

The intermediate products 1 and 2 presented a white colour and in both spectra the absorption bands characteristic of DEA groups (CH2 asymmetric stretch and NH vibration) were observed, indicating that this species is present on the surface and in the pores of SBA-15. A SBA-15 plus DEA sample was synthesized with a sufficient DEA amount to fill the silica pores. The TG/DTG curves (Fig. 8) of this material show the stage of water elimination and the stage of the DEA decomposition in the pores. An event due to the presence of DEA on the surface was not observed. These results allowed the conclusion that the incorporation of DEA in SBA-15 is initiated by the filling of the silica pores. The first molecules of DEA adhere to the internal silica surface, filling the pores completely and, finally, the material excess is adsorbed on the silica surface. In this case, since there is not an excess of DEA molecules observed on the surface of the silica, the DEA must be protected in the interior of the pores and the thermal decomposition initiates at 300
C. The MDEA TG curves (Fig. 9 and Table 4) have the same profile as the materials prepared with MEA and DEA. Three mass loss events were observed; the first is attributed to the water molecules adsorbed on the material surface, while the following two events arise from the MDEA quantity on the surface and in the silica pores, respectively. The quantity of MDEA in the SBA-15 pores is less than the quantity of DEA that fills the silica pores (Table 4) due to the larger molecular volume, caused by the stearic hindrance of the methyl group, in MDEA compared to DEA. 3.2.2. Kinetic study of DEA thermal decomposition in SBA-15 matrix A kinetic study of DEA thermal decomposition was conducted when this substance was incorporated in SBA-15. The Ozawa Method was used to analyse the kinetics of thermal decomposition

Fig. 7. FTIR spectra of SBA-15, DEA, DEA incorporated in SBA-15 silica showing the formation of intermediate thermal decomposition products.

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

4521

Fig. 8. TG curves obtained under a dynamic air atmosphere (50 mL min1), between 25 and 600
C, utilizing Pt0 crucibles and 15 mg of DEA incorporated in SBA-15.

The Ozawa Method was used to calculate the decomposition time of DEA in the silica material, considering different mass rates decomposed (a) for free DEA, DEA on the silica surface, and DEA in the silica pores (Table 6). The similar decomposition times of free DEA and DEA on silica surface suggests that the presence of DEA on the material surface does not influence the thermal stability of this ethanolamine. In contrast, the time necessary for thermal decomposition of DEA in the material pores is significantly longer, suggesting that DEA incorporation in mesoporous silica increases the thermal stability of DEA. 3.3. Characterization of adsorbed material The SAXS results (Fig. 10) of the mixture of SBA-15 and ethanolamines do not present differences compared with the SAXS results of pure SBA-15, indicating that the ethanolamine incorporation did not promote an alteration of the silica structure. All samples have five indexed peaks related with (100), (110), (200), (210) and (300) reflections, showing that all the materials present an ordered hexagonal structure. The SAXS data and lattice parameters of SBA-15 and DEA or MDEA with SBA-15 are similar. It is possible to observe a small variation in the intensity of the (100), (110) and (200) peaks of the materials that have incorporated MEA, besides a small increase of lattice parameters. These variations could be related to small structural changes of the materials incorporated with MEA. Thus, while MEA incorporation may cause some modification in the structural parameters, these variations do not cause a drastic change of the ordered hexagonal mesoporous structure of the material.

Fig. 9. TG curves obtained under a dynamic air atmosphere (50 mL min1), between 25 and 900
C, utilizing Pt0 crucibles and 15 mg of MDEA, SBA-15 and SBA-15/MDEA (1:1) samples.

of free DEA and DEA incorporated in SBA-15 at a 1:1 mass ratio. The kinetic parameters (Table 5), such as activation energy and the reaction order, of the thermal decomposition of free DEA and DEA present on the silica surface are similar. However, the DEA in the SBA-15 pores has a complex thermal decomposition. In this case, the reaction order is two, showing the great influence of the molecular confinement in the porous material.

Table 4 Data obtained by TG curves of SBA-15/MDEA. SBA-15/MDEA (m:m)

Dm1 (%) (25–100
C)

Dm2 (%) (100–210
C)

Dm3 (%) (210–600
C)

MDEA (mol) in the pores/g SBA-15

1:1 1:2

37 25

26 47

5 4

0.00128 0.00116

Table 5 Kinetic parameters – Ozawa Method (T = 40
C). Sample/atmosphere DEA – N2 SBA-15/DEA 1:1 – N2 (surface) SBA-15/DEA 1:1 – N2 (pores)

Ea (kJ mol1) 61  1 62  1 64  1

Frequency factor (min 6

2.062  10 1.992  106 3.933  105

1

)

Order

a

K (min1)

0 0 2

0.18–0.90 0.18–0.90 0.20–0.90

1.377  104 1.475  104 3.621 104

4522

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

Table 6 Decomposition/volatilization time of free DEA, in the SBA-15 surface and in the SBA-15 pores. Surface

Free

Pores

a

Time (days)

a

Time (days)

a

Time (days)

0.21 0.41 0.62 0.82

3 6 13 17

0.25 0.35 0.65 0.90

5 7 14 19

0.22 0.32 0.68 0.84

26 67 157 437

Table 8 Efficiency of CO2 capture utilizing different ethanoamines incorporated in SBA-15 in the 1:2 mass proportion (SBA-15: ethanolamine). Sample

Efficiency (%) 1a Adsorption

Efficiency (%) 2a Adsorption

Efficiency (%) 3a Adsorption

MEA/SBA-15 – Sample 1 MEA/SBA-15 – Sample 2 DEA/SBA-15 – Sample 1 DEA/SBA-15 – Sample 2 MDEA/SBA-15 – Sample 1 MDEA/SBA-15 – Sample 2

17.65 18.01 58.98 59.05 98.60 99.10

13.70 14.03 57.26 58.78 87.40 90.56

13.17 13.80 57.15 58.55 85.65 88.78

3.4. CO2 capture efficiency

Fig. 10. X-ray diffraction pattern obtained by SAXS for pure SBA-15 and SBA-15 incorporated with MEA, DEA and MDEA 1:2 (m/m).

Table 7 Efficiency of CO2 capture by ethanolamines. Ethanolamine

Efficiency (%) 1a Adsorption

Efficiency (%) 2a Adsorption

Efficiency (%) 3a Adsorption

MEA – Sample 1 MEA – Sample 2 DEA – Sample 1 DEA – Sample 2 MDEA – Sample 1 MDEA – Sample 2

73.90 77.05 52.43 53.67 67.38 68.01

– – 48.11 48.12 66.20 67.12

– – 47.71 47.23 66.13 67.03

3.4.1. Free ethanolamine The results of ethanolamine CO2 capture are presented in Table 7. Among the studied ethanolamines, MEA presented the best efficiency in the first CO2 capture process. However, from the second adsorption process, a negative efficiency was observed for MEA, which is attributed to a low amount of initial material, due to the volatilization or decomposition processes. MDEA presented the second best efficiency for CO2 capture and the efficiency did not decrease when the adsorption process was repeated. The data show that MEA, while giving the best efficiency in the CO2 capture process, it is the most unstable species, while the MDEA presented the larger stability of the studied ethanolamines. The stability of MDEA is related to the steric impediment in MDEA molecules. MEA, being a primary amine, presents less steric hindrance, which eases the reaction with CO2; however, it is the most volatile species and is less stable. The efficiency decrease in CO2 capture of the DEA sample compared with MDEA is related to a low chemical stability due to the intramolecular reaction between DEA molecules [6]. The FTIR spectra (Fig. 11) of the ethanolamines before and after the CO2 capture process show the enlargement of the band related to the angular deformation of the OH group (3400 cm1), suggesting that during the capture process, protonation of ionic oxygen atoms on carboxylate groups occurs, producing a carboxylic acid. This phenomenon can occur more readily in the presence of water. The other evidence of CO2 capture by ethanolamines is the band localized between 1730 and 1760 cm1 caused by axial deformation of the C¼O group, which is most evident in the CO2 capture by MEA and MDEA. 3.4.2. Ethanolamine incorporated in SBA-15 The CO2 capture results by ethanolamine incorporated in SBA15 shows that the MEA efficiency decreases when it is incorporated in mesoporous silica (Table 8). It is possible that the SBA-15 catalysed the MEA decomposition. A colour change to yellow was observed when the mixture of MEA and SBA-15 was prepared,

Fig. 11. FTIR spectra of ethanolanimes before and after the CO2 capture process.

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

4523

Fig. 12. FTIR spectra of ethanolamines incorporated in SBA-15 silica before and after the CO2 capture process.

suggesting that some reaction occurred during the incorporation process. However, the efficiency change was not observed after the second adsorption process, suggesting the stability of the MEA present in the material pores. An increase of DEA CO2 capture efficiency occurred when this substance was incorporated in SBA-15. The efficiency of CO2 capture was constant after the first absorption cycle, suggesting that SBA-15 promoted the DEA stabilization. Likewise, the CO2 capture efficiency significantly increased, by about 30%, when MDEA was incorporated into SBA-15 in comparison to free MDEA. The FTIR spectra of the incorporated material (Fig. 12) shows the enlargement of the band corresponding to the OH group axial stretch (3400 cm1) upon CO2 capture as a consequence of the carboxylic acid formation in the presence of water. It is possible to observe a small increase of the band localized at 1700 cm1 in the FTIR of SBA-15, which is a consequence of the vibration of adsorbed water molecules; however, the C¼O chemical linkage also has an absorption band in this spectral region due to the axial deformation of C¼O, evidencing CO2 capture. The CO2 capture tests performed with pure SBA-15 showed no interaction with CO2; thus, the capture process was due solely to the ethanolamine incorporated. 4. Conclusion The presence of mesopores in the SBA-15 silica enabled the incorporation of a larger quantity of ethanolamines than the others types of studied silica, including gel silica and silica produced from rice husks. Although the SBA-15 surface has the capacity to incorporate a large ethanolamine quantity, no significant contribution related to increase of ethanolamine stability was observed when the ethanolamine was on the silica surface. The kinetics study of DEA thermal decomposition in the SBA-15 shows that the activation energy and reaction order are the same for free DEA and DEA on the silica surface. However, the increase of activation energy for the decomposition of DEA in the SBA-15 pores shows that the presence of the ethanolamine in the porous silica can contribute to thermal stabilization of the ethanolamines. Considering the efficiency of CO2 capture by the adsorbents studied in this work, the confinement of MEA in the porous silica caused an acceleration of the thermal decomposition of this material. On the other hand, DEA and MDEA had an increase in efficiency of CO2 capture when they were incorporated in the SBA15 and the increase was most evident in the MDEA sample. The increase in the thermal stability of MDEA and DEA was evidenced by the stability of CO2 captured in the second and third adsorption cycles. MDEA was the ethanolamine that presented the best optimization of the CO2 capture process when incorporated in mesoporous silica compared with DEA and MEA. The use of

mixtures of SBA-15 and MDEA may be a good alternative for the CO2 capture process because it increases the MDEA thermal stability and, consequently, increases CO2 adsorption efficiency. Acknowledgements The authors thank CNPq and CAPES for financial support, the Oxiteno and Dow companies for the donation of the samples, Antonio Carlos F. Silveira (IF-USP) for the SAXS experiments and Dr. Fábio Toniolo (UFRJ) for the isothermal adsorption measures. Márcia C. A. Fantini is researcher of CNPq – Brazil. References [1] A. Sayari, Y. Belmabkhout, R. Serna-Guerrero, Flue gas treatment via CO2 adsorption, Chem. Eng. J. 171 (2011) 760–774. [2] F. Chang, K.J. Chao, H.H. Cheng, C.S. Tan, Adsorption of CO2 into amine-grafted mesoporous silica, Sep. Purif. Technol. 70 (2009) 87–95. [3] G.D. Pirngruber, S. Cassiano- Gaspar, S. Louret, A. Chaumonnot, B. Delfort, Amines immobilized on a solid support for post combustion CO2 capture – a preliminary analysis of the performance in a VSA or TSA process based on the adsorption isotherms and kinetics data, Energy Procedia 1 (2009) 1335–1342. [4] S. Shimizu, C. Song, M. Strano, Role of specific amine surface configurations for grafted surfaces: implication for nanostructured CO2 adsorbents, Langmuir 27 (2011) 2861–2872. [5] B. Epp, H. Fahlemkamp, M. Vogt, Degradation of solution of monoethanolamine: diglycolamine and potassium glycinate in view of tailend CO2 adsorption, Energy Procedia 4 (2011) 75–80. [6] S.G. Ávila, M.A. Logli, J.R. Matos, Kinetic study of the thermal decomposition of monoethanolamine (MEA) diethanolamine (DEA), triethanolamine (TEA) and methyethanolamine (MDEA), Int. J. Greenh. Gas Control 42 (2015) 666–671. [7] E. Borodina, S.I. Karpov, V.F. Selemenev, W. Schwieger, S. Maracke, M. Fröba, F. Röbner, Surface and texture proprieties of mesoporous silica materials modified by silicon-organic compounds containing quaternary amino groups for their application in base catalyzed reaction, Microporous Mesoporous Mater. 203 (2015) 224–231. [8] U. Ciesla, F. Schüth, Ordered mesoporous silica and zincornias: control on length scale between nanometer and micrometer, Microporous Mesoporous Mater. 27 (1999) 131–149. [9] F. Hoffmann, M. Cornelius, J. Morell, M. Fröba, Silica based -mesoporous organic-inorganic hybrid materials, Angew. Chem. Int. Ed. 45 (2006) 3216– 3251. [10] H. Zhao, J. Hu, J. Wang, L. Zhou, H. Liu, CO2 capture by the amine -modified mesoporous materials, Acta Phys. Chim. Sin. 23 (2006) 801–806. [11] Y. Belmabkhout, R. Serna-Guerrero, A. Sayari, Adsorption of CO2 from dry gases on MCM-41 silica at ambient temperature and high pressure 1: pure CO2 adsorption, Chem. Eng. J. 64 (2009) 3721–3728. [12] X.P. Wang, J.J. Yu, J. Cheng, Z.P. Hao, Z.P. Xu, High-temperature adsorption of carbon dioxide on mixed oxides derived from hydrotalcite-like compounds, Environ. Sci. Technol. 42 (2) (2008) 614–618. [13] X. Xu, C. Song, B.G. Miller, A.W. Scaloni, Influence of moisture on CO2 separation from gas mixture by a nanoporous adsorbent based on polyethylenimine-modified molecular sieve MCM-41, Ind. Eng. Chem. Res. 44 (21) (2005) 8113–8119. [14] H. Yang, Z. Xu, M. Fan, R. Gupta, R.B. Slimane, A.E. Bland, I. Wright, Progress in carbon dioxide separation and capture: a review, J. Environ. Sci. 20 (2008) 14– 27. [15] A.A. Olajire, CO2 capture and separation technologies for end-of-pipe applications: a review, Energy 35 (2010) 2610–2628. [16] M.B. Yue, L.B. Sun, Y. Cao, Y. Wang, Z.J. Wang, J.H. Zhu, Efficient CO2 capturer derived from As-Synthesized MCM-41 modified with amine, Chem. Eur. J. 14 (2008) 3442–3451.

4524

S.G. de Ávila et al. / Journal of Environmental Chemical Engineering 4 (2016) 4514–4524

[17] D.R. Radu, C.Y. Lai, K. Jeftinija, E.W. Rowe, S. Jeftinija, V.S.Y. Lin, A Polyamidoamine dendrimer-capped mesoporous silica nanosphere-based gene transfection reagent, J. Am. Chem. Soc. 126 (41) (2004) 13216–13217. [18] Y. Jing, L. Wei, Y. Wang, Y. Yu, Synthesis, characterization and CO2 capture of mesoporous SBA-15 adsorbents functionalized with melamine-based and acrylate-based amine dendrimers, Microporous Mesoporous Mater. 183 (2014) 124–133. [19] J.R. Matos, L.P. Mercuri, M. Kruk, M. Jaroniec, Toward the synthesis of extralarge-pore MCM-41 analogues, Chem. Mater. 13 (2001) 1726–1731. [20] P.F. Fluvio, M. Jaroniec, Optimization synthesis time of SBA-15 materials, Stud. Surf. Sci. Catal. 156 (2005) 75–82. [21] V.L. Zholobenko, A.Y. Khodakov, M. Impéroc-Clerc, D. Durand, I. Grillo, Initial stages of SBA-15 synthesis: an overview, Adv. Colloid Interface 142 (2008) 67– 74. [22] J.L. Blin, M. Impéroc-Clerc, Mechanism of self-assembly in the synthesis of silica mesoporous materials: in situ studies by X-ray and neutron scattering, Chem. Soc. Rev. 42 (2013) 4071–4082. [23] Y. Yang, S. Karmakar, J. Zhang, M. Yu, N. Mitter, C. Yu, Synthesis of SBA-15 rods with small sizes for enhanced cellular uptake, J. Mater. Chem. B 2 (2014) 4929– 4934. [24] S.G. Ávila, L.C.C. Silva, J.R. Matos, Optimisation of SBA-15 properties using Soxhlet solvent extraction for template removal, Microporous Mesoporous Mater. 234 (2016) 277–286. [25] T. Ozawa, Kinetic analysis of derivative curves in thermal analysis, Therm. Anal. 2 (1970) 301–324.

[26] C.E. Doyle, Approximations to the equation of thermogravimetric data, Nature 207 (1965) 290–291. [27] K.S.W. Sing, Reporting physisorption data for gas/solid systems: with special reference to the determination of surface area and porosity, Pure Appl. Chem. 54 (1982) 2201–2218. [28] J.F. Oliveira, Obtenção de sílicas mesoporosas altamente ordenadas a partir de cinzas de casca de arroz e de bagaço de cana-de-açúcar, Institute of Chemisty— University of Sao Paulo, 2014, pp. 154 (Master dissertation). [29] M. Ignat, L. Sacarescu, P. Cool, V. Harabagiu, Glycerol- derived mesoporous carbon: N2-sorption and SAXS data evaluation, Mater. Today Proc. 2 (2015) 3836–3845. [30] S. Pikus, E.B. Seler, M. Jaroniec, L.A. Solovyov, M. Kozak, Studies of intrawall porosity in the hexagonally ordered mesostructures of SBA-15 by small angle X-ray scattering and nitrogen adsorption, Appl. Surf. Sci. 256 (2010) 5311– 5315. [31] R.A. Jesus, A.S. Rabelo, R.T. Figueiredo, L.C.C. Silva, I.C. Cosentino, M.C.A. Fantini, G.L.B. Araújo, A.A.S. Araújo, M.E. Mesquita, Synthesis and application of the MCM-41 and SBA-15 as matrices for in vitro efavirenz release study, J. Drug Deliv. Sci. Technol. 31 (2016) 153–159. [32] F. Mariano-Neto, J.R. Matos, L.C.C. Silva, L.V. Carvalho, K. Scaramuzzi, O.A. Sant’anna, C.P. Oliveira, M.C.A. Fantini, Physical properties of ordered mesoporous SBA-15 silica as immunological adjuvant, J. Phys. D Appl. Phys. 47 (2014) 1–10.