CO2 Adsorption on Aluminosilicate Single-Walled Nanotubes of

Sep 6, 2012 - ABSTRACT: Adsorption of CO2 at subatmospheric pressure at temperatures about ambient has been studied on three materials: (i) imogolite ...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JPCC

CO2 Adsorption on Aluminosilicate Single-Walled Nanotubes of Imogolite Type Cristina Zanzottera,† Marco Armandi,‡ Serena Esposito,§ Edoardo Garrone,† and Barbara Bonelli*,† †

Department of Applied Science and Technology, and INSTM Unit Torino-Politecnico, Corso Duca degli Abruzzi 24, Politecnico di Torino, I-10129 Turin, Italy ‡ Centre for Human Space and Robotics @ Polito, Istituto Italiano di Tecnologia, Corso Trento 21, 10129 Turin, Italy § Department of Civil and Mechanical Engineering, Università degli Studi di Cassino e del Lazio Meridionale, Via G. Di Biasio 43, 03043 Cassino (FR), Italy ABSTRACT: Adsorption of CO2 at subatmospheric pressure at temperatures about ambient has been studied on three materials: (i) imogolite (IMO, chemical formula (OH)3Al2O3SiOH)) a hydrated alumino-silicate occurring as nanotubes (NTs) with bridged AlOHAl groups at the outer surface and Si−OH groups at the inner surface; (ii) an imogolite-like material (Me-IMO, chemical formula (OH)3Al2O3SiCH3) with Si-CH3 groups replacing Si−OH at NTs inner surface; (iii) a material (Me-IMO-NH2) obtained by grafting 3-aminopropylsilane at the outer surface of MeIMO. All materials, being in the form of NTs, exhibit rather high specific surface area values (355−665 m2 g−1) and are accessible to CO2 molecules. Infrared spectroscopy shows that carbon dioxide may interact in a variety of ways. At the inner surface of IMO, linear molecular species are reversibly formed by interaction with silanols, whereas at the outer surface carbonate-like species are given rise with partial reversible character. With Me-IMO, no interaction takes place at the inner surface: linear species are formed in the intertube nanopores as well as carbonate species as in the case of IMO. Finally, with Me-IMO-NH2, all species present in Me-IMO are found, as well as reversible carbamate species arising from the reaction with amino groups. Optical isotherms concerning molecular adsorption have Langmuir character, whereas those for the reversible formation of carbonates/carbamates are of Henry-type. Volumetric isotherms are interpreted as due to two independent families of adsorption sites, respectively Langmuir and Henry: comparison between optical isotherms (measured at ca. 33 °C) and volumetric isotherms (measured at 0 °C) allows a semiquantitative estimate of the adsorption enthalpy for molecular species, corresponding to ca. −20 kJ mol−1, for linear species reversibly formed by interaction with inner silanols in IMO, and to a relatively high adsorption enthalpy for molecular species formed in the larger intertube nanopores of Me-IMO (ca. −32 kJ mol−1). silanols are as acidic as the species of amorphous silica;7 concerning the properties of the outer surface, an IR spectroscopic and catalytic study proved that it has an amphoteric character, with weak acidic properties ascribed to Al(OH)Al groups and basic sites due to nucleophilic oxygen atoms in AlOAl bridges.7 Functionalization of both inner and outer surface is pursued, due to the possible applications of IMO-derived materials, e.g. as humidity sensor,9−1 nanofiller in new composites with increased mechanical and optical properties,12−15 inorganic support for biomolecules,16−18 catalyst,7,19,20 gas adsorbent,2,3,21 and ions adsorbent.22−26 Recently, a variety of IMO has been obtained and thoroughly characterized, hereafter referred to as Me-IMO, with a fully

1. INTRODUCTION Imogolite, (OH)3Al2O3SiOH (IMO), is a naturally occurring aluminosilicate1 with a peculiar structure, consisting of singlewalled nanotubes (NTs) some micrometers long, with an inner diameter of ca. 1 nm and an outer diameter of ca. 2 nm. The outer surface of IMO consists of a gibbsite-like layer, where OH species are bridged over two Al adjacent ions, whereas the inner surface is lined with silanols (Scheme 1). IMO NTs form intertwined bundles, where a pseudo regular arrangement is observed, close to hexagonal packing. This gives rise to three kinds of pores (Scheme 2): (i) those proper to NTs, ca. 1 nm wide, which become accessible to probe molecules after removal of naturally present water (pores A); (ii) smaller intertubes micropores, ca. 0.3 nm wide, not accessible even to water molecules (pores B); and (iii) larger slit-mesopores among bundles (pores C).2,3 Besides occurring ubiquitously as a minor component of soils, IMO can be synthesized via a sol−gel process.4−8 Inner © 2012 American Chemical Society

Received: June 22, 2012 Revised: September 4, 2012 Published: September 6, 2012 20417

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425

The Journal of Physical Chemistry C

Article

Scheme 1. Sections of a Natural Imogolite Nanotube (IMO, Chemical Formula: (OH)3Al2O3SiOH) with an Inner Diameter of about 1.0 nm; Imogolite-Like Materials, Namely Me-IMO (Chemical Formula (OH)3Al2O3SiCH3, with Methyl Groups Replacing Inner SiOHs in IMO); and MeIMO-NH2, Obtained from Me-IMO by Grafting with 3-APS

Scheme 3. Geometrical Evaluation of the Diameter of B Pores in IMO (a) and Me-IMO (b)

hereafter referred to as Me-IMO-NH2.29 The inner surface of the latter features only Si-CH3 groups: the outer surface is complex, due to the presence of residual aluminols, one-, two-, and three-legged species, some of them doubly bound to the surface (Scheme 1). In the present paper, the accessibility of both inner and outer surfaces of IMO, Me-IMO, and Me-IMO-NH2 NTs is studied by means of CO2 adsorption as followed by both volumetric measurements (at 0 °C) and IR spectroscopy at room temperature (actually about 33 °C, as estimated by taking into account the heating effect of the incident IR beam) in the subatmospheric pressure range (0.000−0.030 bar). CO2 was chosen as probe molecule because (i) adsorption occurs at temperatures near the ambient, unlike CO adsorption, for instance, for which often low temperatures are required; (ii) it probes both acidic sites on which it adsorbs linearly (e.g., uncoordinated cations or hydroxyls)30−34 and basic sites involved in the formation of either anionic carbonate or hydrogen−carbonate species;35−38 and (iii) in the presence of −NH2 groups, as with Me-IMO-NH2, formation of carbamates occurs.39 Indeed, CO2 appears to be a probe of election for the materials studied in this paper. Few experimental data are actually available in the literature about CO2 adsorption of imogolite NTs: Ackerman et al.2 reported long ago isotherms obtained at higher pressures (up to 1 bar CO2 equilibrium pressure), in order to study the possible storage properties of NTs. In the present paper, instead, CO2 is adsorbed at lower pressures, to use the molecule as a probe for the surface structures involved. Recent Monte Carlo simulations of carbon dioxide adsorption on isolated imogolite NTs40 have become available, which showed the prominent role of flexible hydroxyls in CO2 adsorption.

Scheme 2. Sketch of the Three Kind of Pores Occurring in Imogolite Bundles: Pores A ca. 1 nm Wide, Proper to NTs; (ii) Pores B, Smaller Inter-Tubes Micropores, ca. 0.3 nm Wide; and (iii) Pores C, Larger Slit-Mesopores among Bundles

2. EXPERIMENTAL SECTION 2.1. Materials. All reagents were from Sigma-Aldrich (ACS grade). IMO was synthesized as described in refs 4−7: briefly, at 20 °C, TEOS (Tetra-ethoxysilane) and Al(s-butoxide)3 were added to a 75 mM aqueous solution of HClO4 in the molar ratios Si/Al/HClO4 = 1.x:2:1. A slight excess of Si precursor is necessary to prevent the formation of aluminum hydroxide. The solution was stirred for 18 h, diluted to 20 mM, put in autoclave at 100 °C for 4 days, dialyzed for 4 days against deionized water, and then dried at 50 °C. Me-IMO was synthesized using Al-sec-butoxide, as source of aluminum, and TEMS (triethoxymethylsilane), as source of Si, with molar ratio Al/Si = 2:1. The obtained sol was dialyzed against deionized water for 4 days.27,28 For the synthesis of Me-IMO-NH2, Me-IMO was dried at 150 °C under vacuum for 4 h to remove moisture before reaction with 3-aminopropylsilane (3-APS). To a stirred

methylated inner surface, corresponding to the chemical formula (OH)3Al2O3SiCH3 (Scheme 1).27,28 Me-IMO has NTs with a larger inner diameter of ca. 2 nm, an outer diameter of ca. 3 nm, and consequently larger intertube pores (Scheme 3) that could become accessible to small molecules, like CO2. Still recently, the outer surface of Me-IMO has been functionalized by means of 3-APS, so obtaining a new material, 20418

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425

The Journal of Physical Chemistry C

Article

structure of all materials, in no case was diffusional hindrance to attainment of equilibrium observed.

suspension of Me-IMO in anhydrous toluene (ca. 60 mL) was added 3-APS in the molar ratios Me-IMO/3-APS = 1:0.3. The resulting mixture was then refluxed at 100 °C for 12 h under nitrogen atmosphere. The slurry was cooled to room temperature and washed with toluene. The product was then filtered and dried at room temperature. 29 Before CO 2 adsorption measurements the Me-IMO-NH2 sample was washed with a diluted NaOH aqueous solution so to neutralize −NH3+ species usually formed on aminated aluminosilicates,41,42 which prevent reaction with carbon dioxide to carbamates. All samples underwent outgassing prior to measurements, at (i) room temperature (r.t.), to desorb weakly held water, mainly from the hydrophobic inner surface of Me-IMO and Me-IMO-NH2, and (ii) 300 °C, to free inner NTs from water in the case of IMO: thermal treatment at such temperature also causes dehydration of the outer surfaces. As it concerns MeIMO-NH2, an outgassing temperature of 150 °C has also been considered, representative of a partially dehydrated surface still carrying intact aminopropyl moieties. 2.2. Methods. XRD patterns of powder samples were obtained by a X’Pert Phillips diffractometer using Cu Ka radiation in the 2.5−16° 2θ range (step width = 0.02°). N2 isotherms at −196 °C were measured to determine samples BET SSA (Brunauer−Emmett−Teller specific surface area) and PSDs (pore size distributions). The latter were calculated by using the NL-DFT (non-local density functional theory) method and applying a N2-silica kernel to isotherm desorption branch. Corresponding results are reported in Table 1.

3. RESULTS AND DISCUSSION 3.1. Samples Textural Characterization. Figure 1 reports powder XRD patterns of the three materials, which show

Figure 1. Low angles powder XRD patterns of Me-IMO (1), MeIMO-NH2 (2), and IMO (3).

basically the same diffraction peaks, although with different relative intensities and slightly different positions. According to the literature, the packing array of IMO and related materials is either hexagonal or tetragonal,43 depending on the water content, because NTs cross-section may change from elliptic to circular on varying the amount of water.8 With Me-IMO (curve 1) all peaks are more intense than with Me-IMO-NH2 (curve 2) and, especially, IMO (curve 3), because of a more ordered packing of NTs in the first material.27,29 The peak due to the (100) reflection is at 2θ = 3.37° in both Me-IMO and Me-IMO-NH2,29 the corresponding cell parameter (calculated as a = 2d(100)/√3 and equal to the center-to-center distance between two aligned NTs within a hexagonal packing) being 2.98 nm for Me-IMO and 3.02 nm for Me-IMO-NH2. With IMO, all peaks are shifted to higher angles, the cell parameter a being smaller, i.e., 2.68 nm (Table 1). Peaks at 2θ = 5.10, 8.25, and 12.2° (Me-IMO and Me-IMONH2) are assigned to diffraction of (110), (001), and (211) planes.29,44,45 Figure 2a reports the N2 isotherms at −196 °C, from which the textural parameters reported in Table 1 were calculated: IMO shows a type I isotherm, proper for a microporous material, whereas Me-IMO and Me-IMO-NH2 exhibit type IV isotherms, with limited hysteresis loops, indicating the presence of mesopores. The steep increase at low p/p0 values indicates the occurrence of some micropores, as well, especially with MeIMO, whereas the extent of adsorption within micropores partially decreased after grafting. The latter finding is confirmed by the BET SSA and micropores volume values reported in Table 1. PSDs reported in Figure 2b confirm the presence of IMO proper nanopores with 0.98 nm diameter, whereas larger pores occur in imogolite-like materials.27,29 After grafting, the relative abundance of mesopores changes in imogolite-like materials, since a decrease of the pores population at ca. 2 nm occurs along with some formation of larger mesopores.29

Table 1. Textural Parameter of the Three Materials As Obtained by XRD Powder Diffraction and N2 Isotherms at −196°C sample IMO Me-IMO MeIMONH2 MeIMONH2c

cell parameter (nm)a

BET SSA (m2 g−1)

total pore volume (cm3 g−1)

microporous volumeb (cm3 g−1)

2.68 2.98 3.02

352 665 303

0.22 0.39 0.32

0.10 0.09 0.004

2.99

518

0.35

0.004

a

Calculated as a = 2d100/√3 by assuming a hexagonal packing array. As calculated by applying the αs method. cOut-gassed at 150 °C to prevent degradation of the organic linker.

b

The same instrument used for N2 isotherms (Quantachrome Autosorb) was used to measure adsorption of CO2 (99.998% purity, Rivoira) in the 0.000−0.030 bar equilibrium pressure range at 0 °C by using an ice−water bath to keep the temperature constant during the experiment. IR spectra of self-supporting wafers were collected at 2 cm−1 resolution on an FT-IR spectrophotometer (Equinox 55, Bruker) equipped with a MCT (mercury−cadmium telluride) cryodetector. Powders were pressed into self-supporting wafers (with an optical density of about 10 mg cm−2); prior to CO2 adsorption, samples were outgassed at either r.t. or 300 °C. CO2 was dosed at r.t. by increasing the equilibrium pressure in the 0.000−0.030 bar range: samples were then evacuated at r.t. to check reversibility of adsorption. Notwithstanding the NT 20419

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425

The Journal of Physical Chemistry C

Article

Figure 3. CO2 adsorption isotherms at 0 °C (0.000−0.030 bar) on IMO (circles); Me-IMO (squares) and Me-IMO-NH2 (triangles) outgassed at r.t. (section a) and dehydrated at 300 °C (section b).

Figure 2. N2 adsorption/desorption isotherms at −196 °C (section a) and PSDs (section b) of IMO dehydrated at 300 °C (circles); MeIMO dehydrated at 300 °C (squares); and Me-IMO-NH2 dehydrated at 150 °C (triangles), the latter sample having been dehydrated at a lower temperature in order to avoid thermal degradation of the organic moieties grafted at the outer surface of NTs. Solid and open symbols refer to adsorption and desorption branches, respectively.

(curves 1) and 300 °C (curves 2). That of IMO outgassed at r.t. (curve 1 in Figure 4a) is dominated by the signal of molecular water (intense band at 1650 cm−1), and the broad adsorption in the OH stretching range (3800−2800 cm−1) is due to extended H-bonding among either water molecules or water/surface hydroxyls. These comprise both inner silanols and outer bridged AlOHAl groups, the former being more hydrophilic than the latter.7,8 Outgassing at 300 °C brings about the thorough removal of water: only observed are the features of surface hydroxyls (broad absorption band between 3750 and 3000 cm−1) and of carbonate species (weak bands at about 1600 and 1480 cm−1) in dehydrated IMO (curve 2 in Figure 4a). As reported in the literature,7 the outer surface of IMO has amphoteric properties, and features some basic sites, onto which atmospheric CO2 may react by forming carbonates/ bicarbonates. With respect to IMO, the IR spectrum of Me-IMO outgassed at r.t. (curve 1 in Figure 4b) shows additional bands at 1270, 2917, and 2973 cm−1 due to vibrations of the inner methyl groups. Although Me-IMO is basically hydrophobic,27 outgassing at r.t. does not bring about complete removal of water, as a band at 1650 cm−1 is still observed. Water is probably located in intertubes spaces (pores B in Scheme 2): depletion of molecular water is instead achieved by dehydration at 300 °C. The IR spectra of Me-IMO-NH2 (Figure 4c) show additional signals due to amino-propyl chains, namely the bands at 3304 and 3043 cm−1 (asymmetric and symmetric stretch of -NH2

Also, grafting causes a dramatic decrease in the micropore volume of Me-IMO-NH2, by a factor more than 20, which can be explained by the blocking of type B pores because of 3-APS grafting at the mouth of such micropores. 3.2. Volumetric Isotherms at 0 °C. Figure 3 reports data concerning the samples outgassed at room temperature (section a) and 300 °C (section b). As a whole, CO2 adsorption on the hydrated samples gives very similar results (Figure 3a), whereas samples behavior changes after dehydration. In particular, the uptake of Me-IMO increases, whereas the opposite is observed with Me-IMO-NH2. As shown later, the increase is probably related to the availability of intertube pores as adsorption centers. The latter phenomenon is instead probably related to a decrease in SSA caused by thermal treatment, as shown by the comparison between Me-IMO-NH2 outgassed at 300 and 80 °C (Table 1). In turn, such a decrease is probably related to the onset of thermal decomposition of the outer organic linker, as previously reported.29 Further elaborations of the data in Figure 3, on the basis of related IR spectra, are reported below. 3.3. IR Spectra of the Bare Samples. Figure 4 reports IR spectra of the three samples outgassed at room temperature 20420

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425

The Journal of Physical Chemistry C

Article

Figure 4. FT-IR spectra of self-supporting wafers of IMO (a), MeIMO (b), and Me-IMO-NH2 (c) outgassed at r.t. (curves 1) and 300 °C (curves 2).

groups) and the bands at 2878 and 2823 cm−1 (asymmetric and symmetric stretching of CH2 groups). 3.4. CO2 Adsorption at r.t. as Followed by IR Spectroscopy. Figure 5 shows selected IR spectra obtained after dosing CO2 on the samples previously outgassed at room temperature; in this and in the following figure, difference spectra are reported with respect to the spectra of bare samples reported in Figure 4. To allow comparison, spectra were normalized to samples unit specific weight. Concerning IMO outgassed at r.t. (Figure 5a), a sharp band forms at 2342 cm−1, markedly pressure dependent, due to the (IR active) asymmetric stretch of CO2 molecules: its absorption coefficient being very high, the related band is very intense, going out of scale even for moderate CO2 equilibrium pressures: at 2277 cm−1 (asterisk) the band appears due to 13 CO2 molecules, the population of which is ca. 1%.30,31 At lower wavenumbers, less intense bands are seen at ca. 1650, 1555, 1460, and 1410 cm−1, due to monodentate carbonates (1555 and 1410 cm−1) and bicarbonates (1650 and 1460 cm−1) formed at the outer surface of NTs. The latter bands are partly irreversible at r.t. (dotted curve) and fully removed only by outgassing at 100 °C. Also with Me-IMO (Figure 5b) an intense band at about 2341 cm−1 is seen, together with less intense absorption due to carbonate-like species. The latter finding is expected, since the outer surface of the two materials is the same. The frequency of the 2341 cm−1 band, being very close to both that of the free gas (2343 cm−1) and to that of CO2 physisorbed on silicalite, an all-silica microporous material (2339 cm−1),30,31 seems to indicate that it is related to a weak interaction. Adsorption, however, at the inner surface of Me-IMO NTs, only featuring methyl groups, is not expected: notwithstanding the similarity

Figure 5. FT-IR difference spectra as obtained after dosing CO2 (equilibrium pressure range: 0.000−0.030 bar) on IMO (a), Me-IMO (b), and Me-IMO-NH2 (c) outgassed at r.t. Difference spectra were obtained after subtraction of spectra of the bare samples reported in Figure 4. Spectra were normalized to samples unit weight to allow comparison.

in vibrational frequency, the species observed with IMO and Me-IMO are probably different. Indeed, an estimate of the corresponding standard adsorption enthalpy made below shows that this is definitely larger with Me-IMO. We favor as adsorption location the intertube cavities shown (pores B in Scheme 2), which are about three times larger in Me-IMO (Scheme 3), and can provide a confined environment to the carbon dioxide molecule, and relatively large van der Waals 20421

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425

The Journal of Physical Chemistry C

Article

interactions between the molecule and the walls. A piece of evidence in favor of a symmetric environment for the molecular species responsible for the 2341 cm−1 is the absence of absorption at 1381 cm−1, the region of the ν1 mode (IR inactive symmetric stretch vibration), which becomes instead partially active in the case of the molecular species interacting with silanols in IMO outgassed at 300 °C, as discussed below (Figure 6b). Measurements of ammonia adsorption on the same system (to be published) also indicate that intertube nanopores in Me-IMO become accessible to adsorption after outgassing at 300 °C. Finally, as discussed below, the comparison of CO2 adsorption in molecular form at the surface of both Me-IMO and Me-IMO-NH2 affords conclusive evidence on the location of molecular species in these systems in B-type pores. The spectral features related to CO2 adsorption on Me-IMONH2 outgassed at r.t. (Figure 5c) are basically the same of MeIMO: the band of physisorbed CO2 is seen at 2339 cm−1, whereas in the lower wavenumbers region, a more complex bands envelope is observed, due to the (expected) formation of carbamates species, as shown by the typical band at 1554 cm−1 (arrow) not observed with the other two samples, and superposed to that of carbonates Interestingly, after prolonged outgassing at room temperature, CO2 adsorption was not completely reversible at room temperature with Me-IMO-NH2 (dotted curve) due to the presence of amino functionalities. Figures 6 and 7 report IR data concerning CO2 adsorption on samples dehydrated at 300 °C. As to IMO (Figure 6a), basically the same linear species is observed as with the sample after dehydration at r.t., though at a lower intensity. The inset to Figure 6a shows the spectrum taken at the highest CO2 pressure in the OH stretching range. Notwithstanding the inherently poor quality of the spectrum, a negative peak at 3742 cm−1 is observed, due to isolated inner SiOHs species interacting with CO2 molecules. With the same sample outgassed at r.t., a spectrum like the one reported in the inset to Figure 6a was not observed because of the large amount of physisorbed water. Simultaneously, a broad absorption forms centered at about 3200 cm−1, due to OH species perturbed by the interaction with CO2 molecules: this is evidence that CO2 molecules interact with silanols inside NTs and probably, as discussed below, also with water hydroxyls. Direct evidence about the interaction between silanols and CO2 molecules is provided by Figure 6c, which shows that proportionality exists between the decrease in intensity of the 3742 cm−1 band and the intensity of the 2341 cm−1 band. The environment of the adsorbed molecule is probably asymmetrical, as documented by the activation of the IR-forbidden ν1 mode, which is in agreement with the formation of H-bonds with terminal oxygen atoms (Figure 6b). Spectra of CO2 adsorbed on dehydrated Me-IMO are less intense than those recorded with the sample outgassed at r.t., especially as far as the carbonate range is concerned (1800− 1300 cm−1, Figure 7a). This finding may be due to the fact that at 300 °C some dehydroxylation of the outer aluminum hydroxide already takes place;7,27 therefore, removing some external OH acting as adsorbing sites. The reason for the decrease in intensity of the molecular species is instead different. Ueno and Bennet46 when studying the adsorption of carbon dioxide on amorphous silica, found that interaction occurs between the linear CO2 molecule and a pair of silanols at a suitable distance and that H-bonding with one hydroxyl only is not enough to stabilize the interaction. Clearly, in the r.t.

Figure 6. FT-IR difference spectra as obtained after dosing CO2 (equilibrium pressure range: 0.000−0.030 bar) on IMO outgassed at 300 °C. Panel a: 2400−2250 cm−1 range (CO2 asymmetric stretch region, ν3); inset to section a reports the spectrum taken at the highest CO2 equilibrium pressure (bold curve in the Figure) in the OH stretch region (3700−2650 cm−1). Panel b: 1800−1300 cm-1 range; the downward arrow indicates the CO2 symmetric stretch, ν1, which becomes IR active when CO2 molecules are linearly adsorbed. Panel c: linear correlation between intensities of the 3742 cm−1 band and that of adsorbed CO2 (ν3 mode).

dehydrated sample still carrying a substantial amount of water, slightly acidic hydroxyls at a distance suitable for interaction may more easily occur than on the dehydrated sample, showing 20422

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425

The Journal of Physical Chemistry C

Article

Figure 7. FT-IR difference spectra as obtained after dosing CO2 (equilibrium pressure range: 0.000−0.030 bar) on Me-IMO (a) and Me-IMO-NH2 (b) outgassed at 300 °C. Difference spectra were obtained after subtraction of spectra of the bare samples reported in Figure 4. Spectra were normalized to samples unit weight to allow comparison.

a regular array with high density7 (9.1 OH nm−2, about twice as much that at the surface of hydrated amorphous silicas47). Similar features are seen with Me-IMO (Figure 7a), in that much weaker bands due to chemisorbed CO2 forming carbonate-like species. As to physisorbed CO2, the decrease in its population caused by the higher temperature of dehydration has probably the same explanation as in the case of IMO, i.e., the assistance by water molecules in double interactions. In the present case, however, the pores involved are those of type B and the hydroxyls those bridged to two Al atoms. With Me-IMO-NH2 (Figure 7b) formation of both carbonates and carbamates occurs. Also, a marked decrease in the population of the linear CO2 species with respect to MeIMO is noted, which is commented on below. 3.5. Semiquantitative Aspects of CO2 Adsorption. Taking advantage of the fact that signals due to molecular species and carbonate/carbamate species do not overlap, it is possible to draw optical isotherms concerning each type of surface species. A few examples of these isotherms are reported in Figure 8. In all cases data concerning molecular species are described rather satisfactorily by Langmuir isotherms: A CO2 = AL KLPCO2 /(1 + KLPCO2)

Figure 8. Optical isotherms as obtained by plotting the intensity of the ν3 CO2 bands in Figures 6 and 7 as a function of equilibrium pressures for IMO (a), Me-IMO (b), and Me-IMO-NH2 (c) outgassed at 300 °C. The curve-fits were obtained by applying the Langmuir adsorption model. Results of the curve-fits are reported in Table 3.

where AL and KL correspond to the maximum intensity and to the Langmuir equilibrium constant, respectively. As to anionic species, in some cases a weak reversible interaction is seen, described by a simple Henry law A CO2 = C HPCO2

(2)

where CH is the product AHKH, in which AH is the maximum intensity and KH is the Henry equilibrium constant. In other cases, a fraction nearly irreversible held is seen. All related parameters are reported in Table 2.

(1) 20423

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425

The Journal of Physical Chemistry C

Article

Table 2. Data Concerning CO2 Adsorption at 0 °C on Samples Outgassed at 300 °C as Derived by Applying a Langmuir + Henry Model to Volumetric Isothermsa

sample IMO MeIMO MeIMONH2 a

CO2 uptake at 0.030 bar (cm3g−1 STP)

VL (cm3 g−1)

KL (bar−1)

CH (cm3 g−1 bar−1)

Table 3. Data Concerning CO2 Adsorption at r.t. on Samples Outgassed at 300 °C, as Derived by Applying a Langmuir Model to Optical Isotherms in Figure 8, Obtained by Reporting the Intensity of the ν3 Band As a Function of CO2 Equilibrium Pressure

adj. Rsquare

sample

Imax (cm−1)

KL (bar−1)

adj. R-square

IMO Me-IMO Me-IMO-NH2

1.39233 2.41071 0.326

69.8 26.1 27.8

0.9996 0.9998 0.9995

32.9 41.6

20.7 23.8

165.7 105.3

478.7 743.1

0.99723 0.99969

24.3

12.8

188.1

438.6

0.9998

hydroxyls actually engaged in the interaction and the rest free because of steric hindrance. Also, it is noted that the population of the linear species measured from optical isotherms (AL) for Me-IMO-NH2 is smaller than that of Me-IMO by a factor of ca. 8, in fair agreement with the decrease in microporosity. We take this as strong evidence supporting the idea that with Me-IMO and Me-IMO-NH2 systems molecular adsorption takes place in Btype pores.

The irreversible fraction has been systematically subtracted.

Leaving aside those cases where IR data show the presence of an irreversibly held fraction, the volumetric isotherms have been interpreted by assuming the simultaneous presence of a Langmuir and a Henry contributions VCO2 = VLKLPCO2 /(1 + KLPCO2) + C HPCO2

(3)

4. CONCLUSIONS CO2 molecules may interact with the surface of imogolite-like systems in a variety of manners. With basic sites at the outer surface of NTs, CO2 forms anionic species: either bicarbonates or carbonates by reaction with basic OH groups or basic oxygen atoms, respectively, or carbamates, when −NH2 groups are present. As it concerns the structurally important feature of NTs cavities, interaction occurs in a molecular form with IMO only, because of the possible hydrogen bond with inner silanols and/or water molecules. No interaction is instead possible with the methyl-lined inner surface of Me-IMO (or Me-IMO-NH2). Here, CO2 is hosted with relatively high energy in the intertube pores, wide enough to allow the molecule to enter and narrow enough to provide a sort of confinement effect. Again, a double interaction is likely to take place, with assistance by either aluminols and/or water molecules.

where KL and KH have the same meaning as before, VL corresponds to the maximum adsorbed volume, and CH is the product VHKH, in which VH is the maximum adsorbed volume. Values of VL, KL, and CH are gathered in Table 2. It is worth noting that the equilibrium constant estimated from optical isotherms is different for IMO and the two other systems, whereas the values for Me-IMO and Me-IMO-NH2 nicely coincides, so showing that two different species are present, respectively. From the values of KL coming from IR spectroscopy (taken at a temperature that we roughly estimate to be 33 °C, as a consequence of the heating by the IR beam) and from volumetry at 0 °C, use of van’t Hoff formula ln[K (T2)/K (T)] = ΔH 0/R[1/T2 − 1/T] 1 1

(4)



allows to make a semiquantitative estimate of the standard enthalpy of interaction in the two cases. It results in ΔH0 values of −20 and −32 kJ mol−1, respectively. Notwithstanding the lack of more adsorption isotherms taken at different temperatures, the first value is compatible, at least on a semiquantitative basis, with a weak H-bond interaction, and indeed close to the value reported by Ueno and Bennet;46 the latter value, closer to the values measured for the CO2 interaction with protonic zeolites,34 is in agreement with a van der Waals interaction with narrow pores. Data in Table 2 allow some considerations concerning the relative populations of the sites responsible for molecular adsorption. As to IMO dehydrated at 300 °C, the overall specific surface has two contributions, one from the inner SiOH lined surface (pores A) and the other from the Al(OH)Al covered external surface (pores C): the surface of intertubes B pores should not contribute. The relative extent of the two contributions can be evaluated through the microporous surface area of IMO, 237 m2 g−1 according the αs method, i.e., about 2/3 of the overall (BET) SSA: most of IMO specific surface is indeed provided by the inner surface. By using (i) the specific surface reported in Table 1, (ii) the structural density of OH’s at the inner surface, and (iii) the amount adsorbed at full coverage (Table 3), the ratio between adsorbed molecules and surface OH is ca. 26%. This value is in good agreement with the proposed model of double interaction, which leads to a 52% of

AUTHOR INFORMATION

Corresponding Author

*Tel.: +39 011 0904719. Fax: +39 011 0904699. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



REFERENCES

(1) Yoshinaga, N.; Aomine, A. Soil Sci. Plant Nutr. 1962, 8 (3), 22− 29. (2) Ackerman, W. C.; Smith, D. M.; Huling, J. C.; Kim, Y.; Bailey, J. K.; Brinker, C. J. Langmuir 1993, 9, 1051−1057. (3) Wilson, M. A.; Lee, G. S. H.; Taylor, R. C. Clays Clay Miner. 2002, 50 (3), 348−351. (4) Farmer, V. C.; Fraser, A. R. In Proceedings of the International Clay Conference; Mortland, M. M., Farmer, V. C., Eds.; Elsevier: Amsterdam, 1978; pp 547−554. (5) Wada, S. I.; Eto, A.; Wada, K. In Proceedings of the International Clay Conference; Mortland, M. M., Farmer, V. C., Eds.; Elsevier: Amsterdam, 1978; p 348. (6) Farmer, V. C.; Adams, M. J.; Fraser, A. R.; Palmieri, F. Clay Miner. 1983, 18, 459−472. (7) Bonelli, B.; Bottero, I.; Ballarini, N.; Passeri, S.; Cavani, F.; Garrone, E. J. Catal. 2009, 264, 15−30. (8) Creton, B.; Bougeard, D.; Smirnov, K. S.; Guilment, J.; Poncelet, O. Phys. Chem. Chem. Phys. 2008, 10, 4879−4888. 20424

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425

The Journal of Physical Chemistry C

Article

(45) Mukherjee, S.; Bartlow, V. M.; Nair, S. Chem. Mater. 2005, 17, 4900−4909. (46) Ueno, A.; Bennett, C. O. J. Catal. 1978, 54, 31−44. (47) Iller, R. K. The Chemistry of Silica: Solubility, Polymerization, Colloid and Surface Properties, and Biochemistry of Silica; John Wiley and Sons: New York, 1979.

(9) Zang, J.; Nair, S.; Sholl, D. S. J. Chem. Phys. 2011, 134, 184103. (10) Oh, J.; Chang, S.; Jang, J.; Roh, S.; Park, J.; Lee, J.; Sohn, D.; Yi, W.; Jung, Y.; Kim, S. J. J, Mater. Sci: Mater Electron 2007, 18, 893−897. (11) Smirnov, K. S.; Bougeard, D. J. Phys. Condens. Matter 2010, 22, 284115. (12) Calvert, P. Nature 1992, 357, 365−366. (13) Shikinakaa, K.; Koizumia, Y.; Osada, Y.; Shigehara, K. Polym. Adv. Technol. 2011, 22, 1212−1215. (14) Yamamoto, K.; Otsuka, H.; Wada, S. I.; Sohnd, D.; Takahara, A. Soft Matter 2005, 1, 372−377. (15) Jiravanichanun, N.; Yamamoto, K.; Irie, A.; Otsuka, H.; Takahara, A. Synth. Met. 2009, 159, 885−888. (16) Inoue, N.; Otsuka, H.; Wada, S. I.; Takahara, A. Chem. Lett. 2006, 35, 194−195. (17) On Yah, W.; Yamamoto, K.; Jiravanichanun, N.; Otsuka, H.; Takahara, A. Materials 2010, 3, 1709−1745. (18) Ishikawa, K.; Akasaka, T.; Abe, S.; Yawaka, Y.; Suzuki, M.; Watari, F. Bioceramics Develop. App. 2011, 1, D110133. (19) Imamura, S.; Hayashi, Y.; Kajiwara, K.; Hoshino, H.; Kaito, C. Ind. Eng. Chem. Res. 1993, 32, 600−603. (20) Imamura, S.; Kokubu, T.; Yamashita, T.; Okamoto, Y.; Kajiwara, K.; Kanai, H. J. Catal. 1996, 160, 137−139. (21) Pohl, P. I.; Faulon, J.-L.; Smith, D. M. Langmuir 1996, 12, 4463−4468. (22) Parfitt, R. L.; Thomas, A. D.; Atkinson, V. R.; Smart, S. C. Clays Clay Miner. 1974, 22, 455−456. (23) Clark, C. J.; McBride, M. B. Clays Clay Miner. 1984, 32 (4), 291−299. (24) Su, C.; Harsh, J. B.; Bertsch, P. M. Clays Clay Miner. 1992, 40, 280−286. (25) Harsh, J. B.; Traina, S. J.; Boyle, J.; Tang, Y. Clays Clay Miner. 1992, 40, 700−706. (26) Arai, Y.; McBeath, M.; Bargar, J. R.; Joye, J.; Davis, J. A. Geochim. Cosmochim. Acta 2006, 70, 2492−2509. (27) Bottero, I.; Bonelli, B.; Ashbrook, S.; Wright, P.; Zhou, W.; Tagliabue, M.; Armandi, M.; Garrone, E. Phys. Chem. Chem. Phys. 2011, 13, 744−750. (28) Bonelli, B.; Bottero I.; Garrone, E. Italian patent no. 0001380065. (29) Zanzottera, C.; Vicente, A.; Celasco, E.; Fernandez, C.; Garrone, E.; Bonelli, B. J. Phys. Chem. C 2012, 116, 7499−7506. (30) Bonelli, B.; Onida, B.; Fubini, B.; Areán, C. O.; Garrone, E. Langmuir 2000, 16, 4976−4983. (31) Bonelli, B.; Civalleri, B.; Fubini, B.; Ugliengo, P.; Areán, C. O.; Garrone, E. J. Phys. Chem. B. 2000, 104, 10978−10988. (32) Garrone, E.; Bonelli, B.; Lamberti, C.; Civalleri, B.; Rocchia, M.; Roy, P.; Areán, C. O. J. Chem. Phys. 2002, 117, 10274−10282. (33) Bonelli, B.; Ribeiro, M. F.; Antunes, A. P.; Valange, S.; Gabelica, Z.; Garrone, E. Microporous Mesoporous Mater. 2002, 54, 305−317. (34) Armandi, M.; Garrone, E.; Areán, C. O.; Bonelli, B. ChemPhysChem 2009, 10, 3316−3319. (35) Lavalley, J. C. Catal. Today 1996, 27, 377−401. (36) Jacobs, P. A.; Cauweleart, F. M.; Vansant, E. F. J. Chem. Soc. Faraday Trans. 1973, 69, 2130−2139. (37) Morterra, C.; Magnacca, G. Catal. Today 1996, 27, 497−532. (38) Morterra, C.; Orio, L. Mater. Chem. Phys. 1990, 24, 247−268. (39) Chang, A. C. C.; Chuang, S. S. C.; Gray, McM.; Soong, Y. Energy Fuels 2003, 17, 468−473. (40) Zang, J.; Chempath, S.; Konduri, S.; Nair, S.; Sholl, D. J. Phys. Chem. Lett. 2010, 1, 1235−1240. (41) Hiyoshi, N.; Yogo, K.; Yashima, T. Microporous Mesoporous Mater. 2005, 84, 357−365. (42) Zengh, F.; Tran, D. N.; Busche, B. J.; Fryxell, G. E.; Shane Addleman, R.; Zemanian, T. S.; Aardahl, C. L. Ind. Eng. Chem. Res. 2005, 44 (9), 3099−3105. (43) Kang, D. Y.; Zang, J.; Wright, E. R.; McCanna, A. L.; Jones, C. W.; Nair, A. ACS Nano 2010, 4, 4897−4907. (44) Cradwick, P. D. G.; Farmer, V. C.; Russell, J. D.; Wada, K.; Yoshinaga, N. Nat. Phys. Sci. 1972, 240, 187−189. 20425

dx.doi.org/10.1021/jp3061637 | J. Phys. Chem. C 2012, 116, 20417−20425