CO2 Capture by Na2TeO4: Structure of Na2–xHxTeO4 and Kinetic

Article Cite This: Inorg. Chem. 2019, 58, 8866−8876

pubs.acs.org/IC

CO2 Capture by Na2TeO4: Structure of Na2−xHxTeO4 and Kinetic Aspects Cyrille Galven,† Thierry Pagnier,‡ Jens Dittmer,† Françoise Le Berre,*,† and Marie-Pierre Crosnier-Lopez† †

Downloaded via KEAN UNIV on July 19, 2019 at 10:44:11 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Institut des Molécules et Matériaux du Mans (IMMM), UMR CNRS 6283, Le Mans Université, Avenue Olivier Messiaen, 72085 Le Mans Cedex 9, France ‡ Université Grenoble Alpes, Université Savoie Mont Blanc, CNRS, Grenoble INP, LEPMI, 38000 Grenoble, France S Supporting Information *

ABSTRACT: β-Na2TeO4 is able to trap CO2 in a humid atmosphere due to a partial Na+/H+ exchange and the formation of NaHCO3. The RT powder X-ray diffraction pattern of the resulting Na2−xHxTeO4 shows broad and narrow hkl lines preventing the structural study. We show by the DIFFaX program that Na+/H+ exchange is topotactic since the structure, as in the mother form, consists of [TeO4]n2n− chains of TeO6 octahedra. We also show that the broadening of some hkl lines is due to stacking faults which result from the weakness of H−O···H bonds connecting the [TeO4]n2n− chains. Upon heating, a progressive structural organization takes place which has been followed by powder X-ray diffraction, Raman, and NMR spectroscopies. Around 300 °C, a well organized structure can be described from powder X-ray diffraction refinements in the monoclinic P21/n space group while ab initio computations allowed location of the hydrogen atoms with satisfactory H−O bonds. In addition, we present the CO2 sorption/desorption by Na2TeO4 and compare its performance to that of Na2SiO3. Finally, the existence of a Na2−xLixTeO4 solid solution (0 ≤ x ≤ 0.9) is evidenced, and we show that the presence of lithium in the structure leads to the disappearance of the structural transition observed for β-Na2TeO4 and to a progressive decrease of the CO2 capture ability.

■

Na2ZrO3,1 or Na2SiO3,2 which are known for their high absorption capacity. In the case of Na2ZrO3, the CO2 absorption follows the reaction

INTRODUCTION

With CO2 being today considered the main contributor to climate warming, its capture is becoming essential. Among the various technologies developed, none can be considered the final solution for limiting the atmospheric CO2 concentration increase, implying that research in this area remains a challenge. The use of metal oxides is one of the ways that has been studied for about 20 years. The principle is to combine CO2 with the oxide to form carbonate (absorption step) and to regenerate the oxide by heating (desorption step). The oxide can then be considered as a CO2 sponge. However, to be a promising candidate for CO2 capture, some criteria must be met: low energy cost for the capture and the regeneration processes, high CO2 absorption capacity, small volume change during the CO2 sorption/desorption cycles, good kinetics, reversibility, and durability. Other parameters must also be taken into consideration such as working temperatures and oxide cost. Among the metal oxides considered as interesting CO2 absorbents, some contain alkali ions, e.g., Li2ZrO3,1 Li4SiO4,1 © 2019 American Chemical Society

Na 2ZrO3(s) + CO2(g) ↔ Na 2CO3(s) + ZrO2(s)

According to this reaction, the maximal theoretical CO2 absorption rate for this sodium zirconate is then equal to 0.2376 g CO2/g adsorbent, corresponding to a weight increase of 23.75%. Different ways to improve the CO2 sorption have been studied. In some cases, the presence of water allows the absorption of 2 times more CO2 than the quantity absorbed under dry conditions, due to the formation of NaHCO3 instead of Na2CO3.2,3 Other studies show that the grain’s size4 and the grain’s morphology5,6 influence the absorption kinetics. Whatever the case, for such compounds, the CO2 Received: May 2, 2019 Published: June 18, 2019 8866

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry

mL of water. A nitric acid solution (2 M) was slowly added until the pH remained acidic and constant. After one night, the product was filtered, washed with distilled water, and dried in an oven at 60 °C. The CO2 sorption was realized by introducing about 0.5 g of Na2TeO4 inside a platinum crucible placed in a reactor (125 mL) containing 10 drops of water. A CO2 pressure of 10 bar was then fixed before heating at 140 °C for one night. For the Na2−xLixTeO4 compositions, we needed to find a synthesis method which allowed us to prepare all the compounds. We chose a coprecipitation method, based on the one used by Hottentot and Loopstra10 for CaTeO4 and SrTeO4. To prepare 1 g of the tellurates, stoichiometric amounts of Li2CO3 and NaHCO3 (Alfa Aesar, 99%) were dissolved in 30 mL of water with H6TeO6 (Sigma-Aldrich, >99%). The solution was progressively heated up to 85 °C under constant stirring until complete evaporation. The resulting solid was then mixed in an agate mortar before being pressed into a pellet and heated in an alumina crucible for 12 h in the air at 400 and 500 °C. A third heating was needed to obtain pure samples with the temperature depending on the x value (Li2TeO4, 620 °C; Na2TeO4, 650 °C). All the powders were placed immediately in a desiccator to prevent any unwanted spontaneous Na+/H+ exchange, taking into account our first observations described in ref 8. Powder X-ray Diffraction. The completeness of the reactions, the crystallinity, and the purity of all the samples were checked by PXRD patterns using a PANalytical Empyrean diffractometer equipped with a Pixel 1D detector (Cu Kα radiation). In the case of the structural study of the high temperature form of NaHTeO, the PXRD data have been collected at room temperature from a sample previously heated a 310 °C and naturally cooled, the transition being not reversible. For this experiment, a well sprinkled sample on grease was carefully prepared to promote a random orientation of the crystallites according to the strong preferential orientation observed on Na2TeO4.8 Patterns (angular range, 2θ = 10.00−150.00°, step scan increment, 2θ = 0.007°; counting time = 2 h 20 min) have been recorded and added to obtain a single pattern. For the CO2 desorption study, the same diffractometer was used, equipped with an Anton Paar HTK12 furnace (XRK900). The data (angular range, 2θ = 14.00−55.00°; step scan increment, 2θ = 0.014°; counting time = 14 min) were collected under dry air flow every 10 °C between 50 and 500 °C. For each temperature, six successive patterns were recorded in order to check the stability of the sample upon heating during the data collection and summed to obtain a single pattern for each temperature. All the refinements were performed with the Rietveld method,11 using the Fullprof profile refinement software12 with a pseudo-Voigt function applied to describe the diffraction line profiles. The background points were determined manually before being refined. The cell parameters evolution of the Na2−xLixTeO4 solid solution (0 ≤ x ≤ 0.9) was obtained by full pattern matching refinements of PXRD patterns (angular range, 2θ = 10.00−120.00°; step scan increment, 2θ = 0.014°; counting time = 2 h 20 min). DIFFax Program. DIFFaX9 is a program that calculates the PXRD pattern for crystal structures containing planar defects such as stacking faults. This is then a particularly suitable method when the pattern has broad and narrow lines preventing usual structural refinement using the Fullprof program.12 With DIFFaX, crystal structure must be described from layers connected together via stacking operations occurring with some probability. For further details, see Supporting Information (Table S1). Thermal Analysis. Thermogravimetric analysis (TGA) coupled with differential thermal analysis (DTA) was made using a TA Instruments SDT Q600 with a heating rate of 10 °C min−1 under synthetic air flow (100 mL min−1). For the CO2 sorption study, an isothermal analysis was performed on a Netzsch STA 449 F3 thermobalance equipped with a water vapor furnace. About 50 mg of the sample were deposited on an alumina flat holder. The temperature was initially increased to 60 °C at 1 °C min−1 and kept at this temperature for 90 min under a dry CO2 flow of 60 mL min−1. The sample was then submitted to humid CO2 at a relative humidity (RH) close to 50% for 14 h.

sorption depends on the alkali diffusion inside the material. Indeed, the sorption occurs first on the grain surface leading to the formation of an external shell of Na2CO3 around a ZrO2 core in the case of Na2ZrO3, for example. This shell becomes thicker and thicker as the sorption continues making the diffusion of the sodium from the core to the grain surface more and more difficult. This diffusion can then be facilitated not only by a temperature increase but also by an opened structure of the compound. This has been observed with Li2ZrO3 and Na2ZrO3:7 Na2ZrO3, which has a layered monoclinic structure faciliting the sodium diffusion, presents a better CO2 sorption than Li2ZrO3, which presents a dense monoclinic structure where lithium diffusion is more difficult. In this context, we have recently studied8 the tellurate βNa2TeO4 (named Na2TeO4 hereafter) which has an opened 1D structure with the Na+ ions located between [TeO4]n2n− chains, thus making Na2TeO4 a good candidate for CO2 capture. We showed that the CO2 capture follows the reaction Na 2TeO4 + x H 2O + xCO2 ↔ Na 2 − xHxTeO4 + x NaHCO3

This mechanism is different from that observed for Na2ZrO3 since it corresponds to a Na+/H+ exchange and to the carbonation of the sodium released to form NaHCO3. Unfortunately, the features of the powder X-ray diffraction pattern of the Na+/H+ exchanged Na2−xHxTeO4 (noted NaHTeO hereafter) made the structural study difficult due to the coexistence of broad and narrow hkl lines. We proposed an orthorhombic cell with parameters close to those of the mother form, especially the cell parameter parallel to the chains, which remains preserved. This indicates that the [TeO4]n2n− chains are most probably still present in the exchanged form NaHTeO. In order to deepen this structural study of NaHTeO, we first performed a DIFFaX9 simulation of the powder X-ray diffraction (PXRD) pattern by introducing stacking faults. This work is the subject of the first part of this paper. The second part is devoted to the structural organization of NaHTeO upon heating. For this, we used ab initio computations, X-ray diffraction, Raman, and NMR experiments. In a third part, the CO2 sorption/desorption mechanism of Na2TeO4 is presented. The sorption has been studied with thermogravimetric analysis performed under humid CO2 flow and with Raman experiments while the desorption process was followed by thermal X-ray diffraction. The CO2 capture ability is also compared with that of Na2SiO3, well-known to be highly CO2 reactive in the same temperature range.2 Finally, this work is completed by the Na+/Li+ substitution study in Na2TeO4 and its influence on the phase transition from orthorhombic to monoclinic seen for Na2TeO48 and on the CO2 capture ability.

■

EXPERIMENTAL SECTION

Synthesis. Na2TeO4 and Na2SiO3 compounds were prepared as powders using the conventional solid-state reaction described respectively in refs 8 and 2 from stoichiometric amounts of dried Na2CO3 (Alfa Aesar, 99.99%), TeO2 (Acros Organics, 99.8%), or SiO2 (Prolabo). Powders were ground, pressed into pellets, and heated in the air at 650 °C for 12 h (Na2TeO4) or 800 °C for 7 h (Na2SiO3). To perform the structural study, NaHTeO was synthesized from Na+/H+ exchange performed by introducing 12 g of Na2TeO4 in 80 8867

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry To generate this RH, the gas was first saturated with moisture by bubbling through a sintered glass disc into a water tank thermostated at 50 °C. The saturated gas was then heated to 60 °C and sent to the TGA cell by means of a heated transfer line. Raman Spectroscopy. Raman spectra were measured with a Renishaw InVia Raman spectrometer. Low wavenumber spectra (down to 5 cm−1) were measured using Super Notch filters, while dielectric filters were used for spectra measured down to 150 cm−1. Measurements were performed in backscattering mode through a ×50 objective having a long (8 mm) working distance. The detector was a Peltier-cooled CCD. Room temperature measurements were performed on as-received powders dispersed on a glassy microscope slide. For atmospheric aging, the powder was simply left in the room with no special protection. The average CO2 concentration was supposed to be close to 410 ppm, the current value in Earth’s atmosphere. For temperature measurements, a Linkam hot stage was used. The powder was placed in an alumina crucible, and the true temperature of the sample was measured using the ∼520 cm−1 band of a small piece of silicon wafer as described in ref 13. Ab Initio Calculations. Ab initio calculations were performed with the ABINIT software14 in the conventional cell (space group P21/c). A Perdew−Burke−Ernzerhof GGA functional15 was used for the exchange-correlation calculations. FHI pseudopotentials16 were obtained from the ABINIT Web site. The energy cutoff was held at 45 Ha (1224 eV), and the k-points grid was that proposed by the software, containing six independent k-points. Raman vibrational frequencies and normal modes were calculated using generalized forces as described in the Supporting Information (band calculation file) of ref 8. A correcting factor of 1.077 was applied to the calculated frequencies to take into account the underestimation inherent to the technique. This factor was the one chosen to minimize the distance between observed and calculated Ag modes in Na2TeO4.8 Solid-State NMR. Solid-state NMR experiments were conducted on a Bruker Avance III WB 300 MHz (7.0 T), equipped with a 1.3 mm H/X channel fast MAS probe. 23Na and 125Te spectra have been acquired as single pulse spectra (23Na with 1H decoupling), while for 1 H the DEPTH sequence was applied for background suppression. The MAS frequency was 60 kHz for 23Na and 1H and 45 kHz for 125 Te. For 23Na, 512 scans were accumulated with a recycle delay of 30 s, for 1H 16 scans with 120 s, and for 125Te 64 scans with 600 s. In addition, a 23Na 3Q-MAS experiment was done for the mother phase on a 4 mm H/X channel WVT probe. The MAS frequency was 10 kHz; the recycle delay 5 s. A total of 120 increments with 12 scans per increment were acquired. The 1H spectra were referenced to TMS, 23 Na to a 1 M solution of NaCl via solid NaCl as secondary reference (7.3 ppm), 125Te to (CH3)2Te via Te(OH)6 as secondary reference (692.2 ppm for the downfield peak). The NMR parameters were extracted from the 23Na spectra by means of the software dmfit.17

Figure 1. Na2TeO4 PXRD patterns: simulated by DIFFaX (a) and experimental (b). Insert: projection of Na2TeO4 structure along the c axis showing the two layers used for DIFFaX simulations.

We can remark that it can be described as a stacking of two different layers in the (a,c) plane respectively with z coordinates close to 0 and 0.5. In these conditions, the largest cell parameter of Na2TeO4, (b = 12.22603(5) Å), becomes the new c stacking direction in the DIFFaX formalism, while a and b parameters become respectively equal to 5.18998(2) and 5.75845(2) Å. Each layer thickness is then equal to c/2 (12.22603/2 Å), and the atomic coordinates for Te and O are then calculated from those refined in the Pbcn space group in order to preserve the Te−O distances.8 At this step, the stacking vector (x, y, z) and the associated stacking transition probability p must be determined: a probability p = 1 means that there is a 100% chance to find the corresponding stacking vector x, y, z. With our two layer description and the fact that layer 1 is necessarily followed by a layer 2 and vice versa, the stacking probability is 1.0, while the stacking vector is 0.0 0.0 1.0, meaning that the stacking is perfectly ordered. The simulation was performed using a pseudo-Voigt function, an unknown Laue symmetry, as advised by the authors of DIFFaX,9 while calculations were realized with a large number of layers (recursive and infinite parameters). Under these conditions, the simulated and the experimental patterns present good agreement, as one can see in Figure 1. This result validates our layered description of the Na2TeO4 structure and will allow us to tackle the simulation of the NaHTeO form. For a better understanding and in order to compare NaHTeO to Na2TeO4, NMR experiments were undertaken on Na2TeO4. A 1H experiment verifies that the phase is free of any hydrogen (Figure 2a). The 23Na spectrum of Na2TeO4 shows a certain complexity which indicates that there is more than one site (Figure 2b). A separation was done by means of a 3Q-MAS experiment (Figure S1), showing that there are two sites, as expected from structural data. Their cross sections are shown in Figure 2b. Their quadrupolar coupling of ωQ = 2π × 1.13 MHz (η = 0.14; 28.6 ppm isotropic chemical shift) and 2π × 1.19 MHz (η = 0.17; 1.1 ppm isotropic chemical shift) is relatively large,

■

RESULTS AND DISCUSSION Na2TeO4: Simulation of the PXRD Pattern by DIFFaX and NMR Characterization. In order to validate our choice of the layers for the PXRD pattern calculation with DIFFaX, we started this study by simulating the PXRD pattern of the mother phase Na2TeO4. Although we know that sodium atoms have a certain weight in the PXRD pattern, we chose not to consider them in this simulation since our goal was to prove that the respective positions of the infinite chains are responsible for the widening of some hkl lines. While in the Ruddlesden−Popper H2SrTa2O7 case,18 the choice of the (a,b) plane and consequently of the c stacking direction are imposed by the perovskite layers, this is not the case of Na2TeO4, which has a 1D structure. The first thing to do is to properly choose the c crystallographic axis that defines the stacking direction in DIFFaX. Figure 1 shows the Na2TeO4 structure projected on its (a,b) plane. 8868

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry

Figure 2. 1H (a), 23Na (b), and 125Te (c) spectra of the mother phase Na2TeO4 and of NaHTeO before and after heating to 310 °C. In b, in brown and olive, the two cross sections of the 3Q-MAS spectrum of Na2TeO4; in dark blue, their sum.

is observed (Figure 3) when two successive layers are randomly shifted by a stacking vector ± x 0.0 1.0 type with a small x value (±0.03) and an associated equal probability factor (0.5).

reflecting a relatively high degree of deviation from the octahedral symmetry: if we measure the distortion seen from the Na site by the largest deviation of the Na−O distances of the average, d = |rNa−O − ⟨rNa−O⟩|max, we find for site Na1 in Na2TeO4 d = 0.23 Å and for site Na2 d = 0.19 Å. 125Te is a nucleus that is not very often studied. It has moderate sensitivity, and we observe long T1 relaxation (in the order of some minutes), but the advantageous spin S = 1/2 in conjunction with the relatively high structural order makes the lines narrow and thus the signal easily observable. The mother phase Na2TeO4 yields one sharp signal at 736.6 ppm, relatively close to Te(OH)6 (692.6 and 685.5 ppm, Figure 2c). NaHTeO: Chemical Formulation and Simulation of the PXRD Pattern by DIFFaX. The chemical formulation of the exchanged NaHTeO was determined by a TGA experiment (Figure S2). Upon heating, NaHTeO exhibits a weight loss (Δm = 5.4%) in one step occurring in the interval 200− 500 °C. After the heating, the resulting powder corresponds to a mixture of Na2Te2O7 (Powder Diffraction File no. 04−013− 8318) and small amounts of TeO3 (Powder Diffraction File no. 01−083−4177) due to the loss of H2O. The calculated formulation of the starting compound corresponds then to Na0.8H1.2TeO4. We know from ref 8 that NaHTeO4 is obtained via a Na+/ + H topotactic exchange with a preservation of the [TeO4]n2n− chains. We decided then to use the same layers for the DIFFaX study of NaHTeO4 at RT with the cell parameters we proposed in ref 8: 4.829, 5.200, and 12.822 Å. According to the DIFFaX formalism, the long axis defines the stacking direction, this implying that the layer thickness is then equal to 12.822/2 Å. The atomic coordinates for Te and O calculated in this new cell, with the preservation of the Te−O distances as for Na2TeO4, are summarized in Table S1. A careful examination of the experimental PXRD pattern reveals that the first and the third diffraction lines are narrow (2θ ≈ 13.8 and 21.9°). As they correspond, respectively, to 002 and to 012 indexations, we can conclude that, most probably, no stacking fault occurs along the b and c directions. Consequently, we started simulations with the hypothesis of a stacking fault only along the a axis and considering only four layers to have a simple model: two identical layers 1 and two identical layers 2 (as for the Na2TeO4, a layer 1 must be necessarily followed by a layer 2 and vice versa). Several tests were performed which revealed that the second diffraction line near 2θ = 19.5° is strongly widened if an important disorder is applied, while the first and the third lines remain narrow, as expected. The best agreement

Figure 3. NaHTeO PXRD pattern at RT: simulated by DIFFaX and experimental.

This study confirms that, as in the mother phase Na2TeO4, the NaHTeO structure is constituted of TeO6 octahedra sharing edges to build infinite chains [TeO4]n2n−. Due to the Na+/H+ substitution, some Na−O bonds are replaced by weaker H−O···H bonds between the chains thus favoring stacking disorder corresponding to a random shift of two successive layers along the [100] direction (4.829 Å). This disorder is clearly responsible for the broad hkl lines of the PXRD pattern, as shown by DIFFaX simulations. Structural Organization of NaHTeO with Temperature: X-ray Diffraction, Raman, and NMR Techniques. 8869

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry Figure 4 shows the PXRD patterns evolution when heating NaHTeO up to 550 °C.

Figure 5. Observed, calculated, and difference PXRD patterns of NaHTeO previously heated at 310 °C. Vertical bars are related to the Bragg reflections positions in the P21/n SG. The ∗ shows the three small diffraction lines not indexed.

these patterns to perform correct refinement (Figure S3). We decided then to achieve the structural determination without taking into account these three additional lines considering they belong to an unknown product. Direct methods were then applied to the observed intensities as extracted from the X-ray data by the program FULLPROF (Le Bail method) using SHELXS7619 (TREF option) and allowed to locate one Te site: −0.01, 0.23, 0.56 (4e site). The structure was completed with subsequent Fourier difference calculations: one 4e site for Na and four 4e sites for O. With this atomic description and one isotropic thermal motion attributed for each kind of atom, the refinement converged quickly to correct reliability factors (Table 1, Figure 5). In this table, we assume a complete occupancy of the sodium site: indeed, if the occupation factor is refined, the site progressively fills. We performed ab initio computations starting from these experimental values in order not only to check the stability of this structure but also to propose H+ sites. For this, the structure of NaHTeO was calculated starting from the PXRD data refinement results. The converged data are gathered in Table 1 (italic values). We can notice the good agreement between these two data sets and see that ab initio computations give coherent calculated H positions. In addition and in order to increase our knowledge about this reaction, we have attempted ab initio calculations exchanging Na1 or Na2 atoms of the orthorhombic Na2TeO4 by H. Starting from the orthorhombic structure, the cell parameters were modified according to the PXRD results: a was decreased by 16%, b increased by 5%, and c left unchanged. The structure was then optimized with three steps summarized in Table 2. The H substitution for Na2 was impossible: there was no convergence of the calculation. In the case of Na1, after the second optimization step (creation of dissymmetric O−H−O bonds), it appeared that an asymmetric unit could be obtained in a monoclinic space group with the following symmetry operations: identity, π rotation about the c axis, inversion, and reflection in relation to the (a,b) plane. This encouraged us to test the P21/n space group for further optimization. The final cell parameters and atomic positions obtained in this space

Figure 4. PXRD patterns of NaHTeO at RT, 310, and 550 °C showing Na2Te2O7 (■) and TeO3 (●) after the complete dehydration process and the gradual splitting of the broad lines under heating (insert).

We observe first that the resulting product of the complete dehydration corresponds mainly to the Na2Te2O7 phase with a small amount of TeO3. This is in agreement with the orange color of the residue, the color of Na2Te2O7. Before the complete decomposition of NaHTeO, we also observe that the broad lines are progressively narrowing, leading to a split-peak nature as shown for the line at 2θ = 19.5° (Figure 4, zoom) and to a well-defined PXRD pattern near 310 °C. As this transformation is not reversible, and considering the strong preferential orientation observed in the case of the mother form Na2TeO4,8 we have recorded a new PXRD pattern at room temperature with a carefully sprinkled sample previously heated at 310 °C. With this new set of data, no orthorhombic cell could explain the peaks splitting. However, a conventional monoclinic cell (unique axis b, a = 4.8408 (1), b = 5.1974 (1), c = 12.8539 (2) Å, and β = 90.8459 (7)°) indexes all the lines. For instance, the two hkl lines appearing around 2θ = 19.5° correspond in this monoclinic cell to −101 and 101. We can remark that these cell parameters are close to those proposed for the orthorhombic cell of the NaHTeO form at RT in ref 8 and used for DIFFaX simulations. Moreover, with this cell, the parameter parallel to the chains in Na2TeO4 is preserved. The systematic absences correspond to the P21/n space group (no. 14). We choose this non-standard space group since it allows the preservation of the axis order used in DIFFaX study. However, three small lines remain not indexed in this new monoclinic cell (Figure 5) and could not be assigned to another phase. Nevertheless, a PXRD pattern of NaHTeO can be obtained below 310 °C without the three additional diffraction lines, but the split of the broad lines is not sufficiently well-defined on 8870

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry

Table 1. Organized Structure Data of NaHTeO (Previously Heated at 310°C; Results of the ab Initio Calculations Given in Parentheses) space group P21/n (no. 14) a = 4.8408(1), b = 5.1974(1), c = 12.8539(2) Å, and β = 90.8459(7)° (a = 4.8888, b = 5.2415, c = 12.9132 Å, and β = 89.218°) Rp = 11.1, Rwp = 12.4, Rexp = 2.68, χ2 = 21.4, Bragg R-factor = 4.89 η = 0.621(6); half width parametersa: u = 0.053(2), v = 0.003(2), w = 0.0071(2) atom

site

x

y

z

Biso

BV

Te Na O1 O2 O3 O4 H

4e 4e 4e 4e 4e 4e 4e

−0.0092(2) (0.0108) 0.980(1) (0.0197) 0.266(2) (0.2537) 0.253(2) (0.2530) 0.709(2) (0.7082) 0.801(2) (0.7969) (0.5617)

0.2416(3) (0.2514) 0.273(2) (0.2360) −0.153(1) (−0.1350) 0.377(1) (0.3432) 0.069(1) (0.0656) 0.074(1) (0.0771) (0.2202)

0.5639(1) (0.5634) 0.8201(2) (0.8240) 0.3314(5) (0.3334) 0.6684(5) (0.6642) 0.9464(7) (0.9494) 0.4380(6) (0.4537) (0.6574)

0.87(2) 0.43(8) 1.31(7) 1.31(7) 1.31(7) 1.31(7)

5.1 (5.5) 1.4 (1.2) 1.6 (1.9) 1.2 (2.0) 1.7 (1.9) 1.9 (1.9) (0.9)

Selected interatomic distances (Å). Te−O (Å): Te−O1, 1.903(7); Te−O2, 1.966(8); Te−O3, 1.962(7); Te−O3, 2.100(9); Te−O4, 2.044(8); Te−O4, 1.928(7) (mean distance: ⟨Te−O⟩ = 1.98 Å). Na−O (Å): Na−O1, 2.351(8); Na−O1, 2.24(1); Na−O2, 2.43(1); Na−O2, 2.43(1); Na− O3, 2.35(1); Na−O4, 2.30(1) (mean distance: ⟨Na−O⟩ = 2.35 Å). H−O (Å): H−O1, 1.02; H−O2, 1.64 (from ab initio). aCaglioti, G.; Paoletti, A.; Ricci, F. P. Choice of collimators for a crystal spectrometer for neutron diffraction. Nucl. Instrum. 1958, 3, 223−228.

parallel to the b axis (Figure 6) and connected by sodium cations in a very distorted octahedral coordination. This structure is therefore clearly related to that of the mother form Na2TeO4, meaning that the Na+/H+ exchange is topotactic. The comparison of the two structures however shows that only one of the two sodium sites (Na2) is still occupied in NaHTeO, the Na1 site being completely empty as highlighted from ab initio computations. From the point of view of distances and bond valence sums (Table 1), we can consider that the values are in agreement with the sum of the ionic radii (1.96 and 2.42 Å respectively for Te−O and Na− O).20 The mean Na−O distance is relatively small (2.35 Å), leading consequently to a bond valence sum greater than the expected value (1.4 instead of 1). The hydrogen bonds link two chains along the a axis (Figure 6) with classical H−O (1.02 Å) and O−H···O distance (1.64 Å). It is interesting to note that hydrogen bonds are aligned along the a axis (≈ 4.84 Å), the axis along which the stacking faults occur as we saw in the DIFFaX part. The Raman spectra of NaHTeO have been recorded between RT and 340 °C (Figure S4). Above 800 cm−1, large features of low intensity are observed, most probably due to H

Table 2. Optimization Steps When Substituting H for Na1 or Na2 in the Na2TeO4 Structure step definition structure optimization with all symmetries removed dissymmetrization of O−H−O bonds (creation of strong O−H bonds and hydrogen bonds) using space group 14, unique axis c, cell choice 2 (P1121/n)

result for Na1−H exchange converged E = −293.0798 Ha converged E = −293.0996 Ha converged E = −293.0997 Ha

result for Na2−H exchange did not converge

group were very close to those previously obtained (Table 1). We can therefore conclude that the Na+/H+ exchange reaction preferably empties the Na1 site and that the introduction of hydrogen implies a spontaneous transformation into a monoclinic cell. The structure of NaHTeO can be described as built from infinite chains of TeO6 octahedra sharing edges and running

Figure 6. Projections of the organized NaHTeO structure showing the [TeO4]n2n− chains running along the b axis. 8871

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry

Figure 7. Raman spectra of NaHTeO (1) before,and (2) after heat treatment at 340 °C and of Na2TeO4 aged in air at room temperature for several months (3). The vertical lines indicate the positions of the Ag bands (red, long) and Bg bands (blue, short) as calculated ab initio. The star indicates a contaminant that is neither Na2CO3 nor NaHCO3, probably some dust. Inset: fwhm of the Raman bands before (full circles) and after (open squares) heat treatment at 340 °C.

Figure 8. Thermogravimetric analyses under humid CO2 flow (50% RH) at 60 °C: Na2TeO4 prepared by solid state chemistry (▲), by coprecipitation (×), and Na2SiO3 by solid state reaction (⧫).

atom vibrations. Below 800 cm−1, we did not observe any change in the Raman spectra, except a small red shift and a band enlargement during this heat treatment. This is a good indication that no structural change has occurred. After cooling down to 25 °C, the Raman bands have become narrower, without any change in the position (Figure 7). This is in agreement with the PXRD results which showed that, upon heating, a long-range ordering occurred. The counting of the Raman-active modes can then be performed. Forty-two Raman active modes are expected (21 Ag and 21 Bg), but only 20 modes are clearly distinguished in the experimental Raman spectra, not taking into account those implying predominantly H atoms that lie above 800 cm−1. The calculated Raman band positions are also shown in Figure 7. The experimental spectrum is fairly well reproduced. In numerous cases, Ag and Bg modes overlap, which explains that

the number of observed bands is much lower than the expected one. Nevertheless, these experiments confirm the structure obtained from PXRD refinement and ab initio computations. The partial substitution of sodium by hydrogen is seen in the 1 H NMR spectrum by a signal at 11.6 ppm (Figure 2a). A second, smaller signal at 6.3 ppm is attributed to lattice water. The structural change upon Na+/H+ exchange is reflected in the 23Na and 125Te NMR spectra (Figure 2b,c). The two 23Na sites are replaced by a single one with a chemical shift of 8.9 ppm and a smaller quadrupolar coupling (ωQ = 2π × 821 kHz, η = 0.27) due to a smaller octahedral distortion (reduced to d = 0.09 Å). The signal resembles that shown by Hayashi21 for a sample denoted as Na2TeO4·2H2O in which, however, probably also Na+/H+ exchange has taken place. The 125Te signal moves to 702.7 ppm, much closer to those of Te(OH)6 (692.2 and 685.5 ppm), a shift that is significant but, on the 8872

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry

Figure 9. Raman spectra obtained during aging in the laboratory air of powdered Na2TeO4.

Figure 10. PXRD pattern evolution of a fully transformed Na2TeO4 sample under heating up to 610 °C showing the progressive recovering of the mother form Na2TeO4 (∗) and the formation of Na2Te2O7 (●) between 410 and 510 °C. Inset: progressive disappearance of Na2Te2O7 on the PXRD pattern if the temperature is maintained at 450 °C.

when the CO2 flow is charged with moisture, the mass of the three samples increases significantly. We also observe that the kinetics are different for the three samples: Na 2 TeO 4 synthesized by coprecipitation reacts faster than the two others, but after 10 h, Na2SiO3 becomes the most effective. For Na2SiO3, the PXRD pattern performed after the TGA revealed the presence of Na2SiO3, NaHCO3, and SiO2. This means that the CO2 sorption is not finished, as evidenced by the weight sample which still increases at the end of the experiment. In the case of Na2TeO4, the following reaction takes place:

other hand, small within the overall chemical shift range of 125 Te. After heating to 310 °C and cooling down again, the 23Na chemical shift and quadrupolar coupling do not change, but the signal narrows, showing a higher structural order. A similar behavior is seen in the 125Te signal, where however the high spectral resolution in addition allows the observation of a small shift to 702.2 ppm. Also, the 1H signals narrow, thus the increase in the structural order also concerns the substituting hydrogen and the lattice water. The latter in addition shifts significantly. CO2 Sorption/Desorption Study on Na2TeO4. The CO2 sorption was followed by isothermal experiments at 60 °C under humid CO2 flow (50% RH). Three samples were analyzed under these conditions: two Na2TeO4 samples prepared by different synthesis methods (solid state reactionSSRand coprecipitationCP) and one Na2SiO3 sample (SSR) used as a reference since it is well-known to be highly CO2 reactive in the same temperature range (30−60 °C).2 The experimental results are presented in Figure 8. We observe first that, as long as the three samples remain under dry CO2, no significant mass variation is visible, while

Na 2TeO4 + x H 2O + xCO2 → x NaHCO3 + Na 2 − xHxTeO4

since these two products (NaHCO3 and Na2−xHxTeO4) are present in the PXRD patterns recorded after the CO2 sorption. Concerning the SSR sample, we observed in addition the mother phase Na2TeO4, indicating thus that the capture is not completed. This demonstrates that for Na2TeO4 the decrease of the grain size (Figure S5) amplifies significantly the absorption kinetics (Figure 8) as observed in the literature for Na2ZrO34 and Na2SiO3.2 8873

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry

Figure 11. Solid solution Na2−xLixTeO4 (0 ≤ x ≤ 0.9): evolution versus the lithium content (x) of the cell parameters (Å).

Na+/H+ Exchange in the Laboratory Atmosphere. Raman spectra of Na2TeO4 left in the laboratory atmosphere were measured over several months. Some of the spectra are shown in Figure 9. The spectra clearly show a progressive substitution of the Na2TeO4 spectrum by the NaHTeO one with the final spectrum very close to that of NaHTeO obtained by ion exchange. There is no shift of the bands, this indicating there is no solid solution Na2−xHxTeO4. Furthermore, each intermediate spectrum can be considered as the sum of the Na2TeO4 and NaHTeO spectra, the weight of each evolving with time. Figure S6 shows the fractions determined by hand of each end spectrum in the intermediate spectra. This gives a semiquantitative idea of the kinetics of the Na+/H+ exchange reaction. A complete transformation is observed on the spectra in less than 150 days. We can remark that this exchange reaction is surprisingly particularly slow. Indeed, previous experiments performed with PXRD on a crushed sample left in the laboratory atmosphere revealed that the exchange becomes visible after 4 days (Figure S7). We think that this kinetic difference is due to the humidity rate of the laboratory atmosphere: the Raman spectra were taken in winter when the laboratory atmosphere was very dry (due to lab heating), while the PXRD patterns were recorded in September. This emphasizes the importance of minimum moisture for the reaction to take place. To follow the CO2 desorption, a thermal powder X-ray diffraction experiment was undertaken between room temperature and 610 °C on transformed Na2TeO4 after the CO2 sorption test (Figure 10). From 100 °C, NaHCO3 is transformed into Na2CO3. Then, diffraction lines corresponding to Na2TeO4 begin to appear at 130 °C and grow regularly until 610 °C. Surprisingly, when the temperature reaches 410 °C, additional diffraction lines appear with intensity that gradually increases before they disappear at 510 °C. These lines do not belong to either Na2TeO4 or NaHTeO but to a third tellurate, Na2Te2O7. At 450 °C, the complete disappearance of NaHTeO is observed, and all the diffraction lines can be indexed with a full pattern matching refinement with the two phases Na2TeO4 and Na2Te2O7. Their

coexistence demonstrates the difficulty of sodium to diffuse in the grains: the sodium has to circulate throughout the surface to reach the core. The progressive sodium diffusion is confirmed by the continuous disappearance of the Na2Te2O7 form on the PXRD pattern if the temperature is maintained at 450 °C, meaning that the sodium diffusion is the factor limiting the rapid formation of Na2TeO4 (Figure 10, zoom). At 610 °C, only Na2TeO4 in monoclinic form8 is obtained, showing then the entire reversibility of the CO2 capture. Na 2−x Li x TeO 4 Compositions: Synthesis and CO 2 Chemisorption Study. The tellurate Na2TeO4 has a 1D structure with Na+ cations located between [TeO4]n2n− chains. With the smaller cation Li+, the structural arrangement is quite different since Li2TeO4 adopts a structure called inverse spinel.22 This one is built from MO6 them with half of the octahedra occupied by Te6+ and the other half by Li+. These octahedra form infinite helical chains by sharing edges. The remaining Li+ cations are in a tetrahedral environment, LiO4. In 2017, Pfeiffer et al.7 studied the Li2−xNaxZrO3 composition and showed that the sample presenting the best CO2 absorption was LiNaZrO3. This formulation presents at 600 °C an efficiency of 75.3% corresponding to 0.196 g of CO2 absorbed 1 g of sorbent according to the reaction: 1 1 Li 2CO3 + Na 2CO3 + ZrO2 2 2 We then checked the CO2 sorption ability of Na2−xLixTeO4 samples. For this, we needed to find a synthesis method allowing the preparation of all compositions. We have chosen the coprecipitation method since the solid state one used for Na2TeO48 did not allow us to obtain Li2TeO4, whatever the temperature applied. Well crystallized and pure samples of both Na2TeO4 and Li2TeO4 were then synthesized by this method as well as the Na2−xLixTeO4 compositions. The PXRD patterns (Figure S8) presented only the diffraction lines of Na2TeO4 for x ≤ 0.9, while from x = 1.0, unknown additional lines appeared. This indicates the existence of a partial solid solution Na2−xLixTeO4 for 0 ≤ x ≤ 0.9. This was confirmed by the PXRD data refinements in the Pbcn space group of Na2TeO4. Figure 11 shows the cell parameters evolution versus lithium content for 0 ≤ x ≤ 0.9 (Table S2 for the values). LiNaZrO3 + CO2 →

8874

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry

Figure 12. Solid solution Na2−xLixTeO4 (0 < x < 0.9) : total mass loss of the TG analyses after the ROR test under humid CO2 (see text).

■

In good agreement with the respective ionic radii of Li+ and Na+ cations (0.76 and 1.02 Å in an octahedral coordination),20 the three cell parameters decrease linearly with x according to Vegard’s law. It is nevertheless interesting to note that the parameter parallel to the [TeO4]n2n− chains is the least affected by this substitution since it depends mainly on the connection mode of the octahedra in the [TeO4]n2n− chains. Concerning the CO2 capture of the Na2−xLixTeO4 series, we used the same procedure described in ref 8 for Na2TeO4. The PXRD patterns performed on the resulting powders after the test revealed the presence of NaHCO3 and of a partially exchanged form. This confirms that the Na+/H+ exchange and the CO2 capture have occurred. However, we also observed that the proportion of the exchanged form strongly decreased with x, meaning that the Na2−xLixTeO4 series becomes less and less H2O/CO2 reactive as the lithium content increases. It was confirmed by the TG analyses, which showed that the total mass loss decreased with x (Figure 12). Concerning Li2TeO4, no reaction with CO2 was observed. Moreover, if Li2TeO4 is placed in water, no pH increase occurs, contrary to what is observed for Na2TeO4. This means that there is no spontaneous Li+/H+ exchange in Li2TeO4. It seems in addition that the lithium in Na2−xLixTeO4 does not contribute to the CO2 capture since we did not observe Li2CO3 in PXRD patterns after the tests. The partial substitution of sodium by lithium interferes then with the CO2 capture because the sodium content decreases as x increases and due to the cell contraction resulting from the Na+/Li+ substitution, which prevents the alkali diffusion. With the Na2TeO4 form presenting a reversible phase transition from an orthorhombic to a monoclinic structure near 420 °C,8 we performed a DSC analysis on some members of the Na2−xLixTeO4 series to see the influence of the substitution on this structural transformation. Unfortunately, no peak is observed on the DSC curves, meaning that the structure seems “blocked” as soon as lithium partially substitutes sodium.

CONCLUSION In a humid atmosphere, Na2TeO4 is able to trap CO2 via a two step mechanism: Na+/H+ exchange, leading to NaHTeO, followed by a carbonation of the released sodium as NaHCO3. In order to characterize NaHTeO, we synthesized it by soft chemistry via a Na+/H+ exchange in acidic solution from Na2TeO4. The resulting compound presents a disordered structure leading to a broadening of some hkl lines on the PXRD pattern. This structure was approached from PXRD experiments and DIFFaX simulations. We showed that, as in the mother form, it was constituted by TeO6 octahedra sharing edges to build infinite chains [TeO4]n2n−. We also explained that the broadening of hkl lines resulted from a stacking disorder due to the elasticity of the hydrogen bonds. After heating at 310 °C, the NMR experiments revealed that the disorder disappears and that the TeO6 octahedra are more regular. We then obtained a well organized compound, and its structure was described in a monoclinic cell (P21/n, a = 4.8408(1), b = 5.1975(1), c = 12.8540(2) Å, and β = 90.8458(7)°) from PXRD refinements. Ab initio computations allowed us to give a coherent H position. They also revealed that the Na+/H+ exchange involved a monoclinic transformation and preferably emptied the Na1 site of the mother phase. In addition to this structural part, kinetic aspects of the spontaneous reversible CO2 absorption/desorption by Na2TeO4 were studied. The absorption is due to a strong instability of Na2TeO4 toward water, which results in a Na+/ H+ exchange and the formation of NaHTeO, while the sodium released is combined with CO2 under NaHCO3. The exchange occurs first over the grain surface; this explains why the CO2 capture kinetics is strongly accelerated as the grain size decreased. We also showed that the CO2 capture rate for Na2TeO4 is lower than that of Na2SiO3. The aging in ambient atmosphere of a Na2TeO4 sample was also followed in situ by Raman and PXRD experiments. It is strongly dependent on the humidity rate in the laboratory. Under transformation, the Raman spectrum as well as the PXRD patterns correspond to the sum of the NaTeO4 and NaHTeO spectra, confirming thus the absence of a Na2−xHxTeO4 solid solution. To complete this study, we also evidenced the existence of the orthorhombic 8875

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876

Article

Inorganic Chemistry Na2−xLixTeO4 solid solution for 0 ≤ x ≤ 0.9. As soon as sodium is substituted by lithium, the structural transition seen near 420 °C for Na2TeO4 is no longer observed and the CO2 capture ability strongly decreases, probably due to the cell contraction further limiting the alkali diffusion. To end our work, we performed the same tests of CO2 capture in a humid atmosphere on two other tellurates: we observed that Na2Te2O7, which has a 3D structure, did not absorb CO2, while Na4TeO5 (0D structure) reacts. This tends to confirm the importance of the structure dimensionality in this CO2 capture mechanism.

■

(7) Pfeiffer, H.; Vázquez, C.; Lara, V. H.; Bosch, P. Thermal Behavior and CO2 Absorption of Li2−xNaxZrO3 Solid Solutions. Chem. Mater. 2007, 19 (4), 922−926. (8) Galven, C.; Pagnier, T.; Rosman, N.; Le Berre, F.; CrosnierLopez, M. P. β-Na2TeO4: Phase Transition from an Orthorhombic to a Monoclinic Form. Reversible CO2 Capture. Inorg. Chem. 2018, 57 (12), 7334−7345. (9) Treacy, M. M. J.; Newsam, J. M.; Deem, M. W. A Gereral Recursion Method for Calculating Diffracted Intenisties from Crytsals Containing Planar Stacking Faults. Proc. R. Soc. London, Ser. A 1991, 433, 499−520. (10) Hottentot, D.; Loopstra, B. O. Structures of Calcium Tellurate, CaTeO4, and Strontium Tellurate, SrTeO4. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1979, B35, 728−729. (11) Rietveld, H. M. A Profile Refinement Method for Nuclear and Magnetic Structures. J. Appl. Crystallogr. 1969, 2 (2), 65−71. (12) Rodriguez-Carvajal, J. FULLPROF; The FullProf Team, 2014. (13) Helwig, A.; Spannhake, J.; Müller, G.; Rosman, N.; Pagnier, T. Temperature Characterization of Silicon Substrates for Gas Sensors by Raman Spectroscopy. Sens. Actuators, B 2007, 126 (1), 240−244. (14) Gonze, X.; Beuken, J. M.; Caracas, R.; Detraux, F.; Fuchs, M.; Rignanese, G. M.; Sindic, L.; Verstraete, M.; Zerah, G.; Jollet, F.; et al. First-Principles Computation of Material Properties: The ABINIT Software Project. Comput. Mater. Sci. 2002, 25, 478−492. (15) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple John. Phys. Rev. Lett. 1996, 77 (18), 3865−3868. (16) Fuchs, M.; Scheffler, M. Ab Initio Pseudopotentials for Electronic Structure Calculations of Poly-Atomic Systems Using Density-Functional Theory. Comput. Phys. Commun. 1999, 119 (1), 67−98. (17) Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calvé, S.; Alonso, B.; Durand, J. O.; Bujoli, B.; Gan, Z.; Hoatson, G. Modelling One- and Two-Dimensional Solid-State NMR Spectra. Magn. Reson. Chem. 2002, 40 (1), 70−76. (18) Crosnier-Lopez, M. P.; Fourquet, J. L. Stacking Faults in Protonated Layered Perovskite Phases: DIFFaX Simulation Studies on H2SrTa2O7. Solid State Sci. 2005, 7 (5), 530−538. (19) Sheldrick, G. M. SHELX76 Program for Crystal Structure Determination; University of Cambridge: England, 1976. (20) Shannon, R. D. Revised Effective Ionic Radii and Systematic Studies of Interatomic Distances in Halides and Chalcogenides. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1976, 32, 751−767. (21) Hayashi, S. Magic-Angle Spinning Nuclear Magnetic Resonance of Half-Integer Quadrupole Nuclei: Effect of Spin-Locking Efficiency on Powder Lineshapes. Solid State Nucl. Magn. Reson. 1994, 3, 93−101. (22) Daniel, F.; Moret, J.; Philippot, E.; Maurin, M. Etude Structurale de Li2TeO4. Coordination Du Tellure VI et Du Lithium Par Les Atomes d’oxygène. J. Solid State Chem. 1977, 22 (2), 113− 119.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.9b01282.

■

3Q-MAS experiment of Na2TeO4; TGA curve of NaHTeO; DIFFaX formalism and atomic parameters of the layers used for the DIFFaX simulations; PXRD pattern of NaHTeO (RT) previously heated at 295 °C; Raman spectra of as-prepared NaHTeO4 during the heat-treatment; SEM image of Na2TeO4 showing the grain size depending on the synthesis method; fractions of Na2TeO4 and NaHTeO extracted from Raman spectra; PXRD pattern of Na2TeO4 maintained in ambient air for different times; PXRD patterns of Na2TeO4, Na1.1Li0.9TeO4, and Li2TeO4 prepared by coprecipitation; cell parameters (Å) and volume (Å3) of Na2−xLixTeO4 (0 ≤ x ≤ 0.9) obtained from PXRD pattern refinement (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Françoise Le Berre: 0000-0002-0593-8984 Marie-Pierre Crosnier-Lopez: 0000-0002-8676-7063 Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Kumar, S.; Saxena, S. K. A Comparative Study of CO2 Sorption Properties for Different Oxides. Mater. Renew Sustain Energy 2014, 3, 30. (2) Rodríguez-Mosqueda, R.; Pfeiffer, H. High CO2 capture in Sodium Metasilicate (Na2SiO3) at Low Temperatures (30−60°C) through the CO2-H2O Chemisorption Process. J. Phys. Chem. C 2013, 117 (26), 13452−13461. (3) Santillán-Reyes, G. G.; Pfeiffer, H. Analysis of the CO2 Capture in Sodium Zirconate (Na2ZrO3). Effect of the Water Vapor Addition. Int. J. Greenhouse Gas Control 2011, 5 (6), 1624−1629. (4) Ji, G.; Memon, M. Z.; Zhuo, H.; Zhao, M. Experimental Study on CO2 Capture Mechanisms Using Na2ZrO3 Sorbents Synthesized by Soft Chemistry Method. Chem. Eng. J. 2017, 313, 646−654. (5) Nambo, A.; He, J.; Nguyen, T. Q.; Atla, V.; Druffel, T.; Sunkara, M. Ultrafast Carbon Dioxide Sorption Kinetics Using Lithium Silicate Nanowires. Nano Lett. 2017, 17 (6), 3327−3333. (6) Akram, M. Z.; Atla, V.; Nambo, A.; Ajayi, B. P.; Jasinski, J. B.; He, J.; Gong, J. R.; Sunkara, M. Low-Temperature and Fast Kinetics for CO2 Sorption Using Li6WO6 Nanowires. Nano Lett. 2018, 18 (8), 4891−4899. 8876

DOI: 10.1021/acs.inorgchem.9b01282 Inorg. Chem. 2019, 58, 8866−8876