In Situ XRD, Raman Characterization, and Kinetic Study of CO₂ Capture by Alkali Carbonate-Doped Na₄SiO₄

Abstract

Sodium silicate, a new type of CO₂ sorbent, has a relatively low cost, but its sorption reactivity is not yet good enough. Alkali carbonate doping is commonly used as an effective means to improve the CO₂ uptake reactivity of solid sorbents. In this study, sodium orthosilicate, Na₄SiO₄, was synthesized and mixed with 5, 10, and 20 mol% of Li₂CO₃–Na₂CO₃ or Li₂CO₃–Na₂CO₃–K₂CO₃ as CO₂ sorbents. The promotion of alkali carbonates on Na₄SiO₄ in CO₂ capture was characterized using thermal analyses in an 80 vol% CO₂–20 vol% N₂ atmosphere. The phase evolution and structural transformations during CO₂ capture were characterized by in situ XRD and Raman, and the results showed that the intermediate pyrocarbonate, C₂O₅²⁻, which emerged from alkali carbonates, enhanced the CO₂ capture of Na₄SiO₄ to form Na₂CO₃ and Na₂SiO₃ from 100 °C. Isothermal analyses showed that 10 mol% of Li₂CO₃–Na₂CO₃ was the optimal additive for Na₄SiO₄ to attain better CO₂ uptake performance. The alkali carbonates were effective in reducing the activation energy for both chemisorption and bulk diffusion, improving the cycle stability of Na₄SiO₄.

1. Introduction

CO₂ emissions have become a world concern due to their huge impact on global warming [1,2]. CO₂ capture and storage (CCS) technology using different amine solvents or solid chemisorbents has been proposed as the most promising method for the removal of CO₂ from industrial flue gases [3,4]. Among CO₂ captors, Monoethanolamine (MEA) and diethanolamine (DEA) are representative amine solvents used for CO₂ capture, but high cost, energy loss, and corrosiveness plague their widespread application [5,6]. Solid sorbents such as CaO [6,7] or MgO [8,9] seem to be optional CO₂ captors due to their low cost and good reactivity. However, the dense solid product layer covering the unreacted sorbents restricts CO₂ diffusion and further reaction. Some alkali ceramics also have the potential to be CO₂ sorbents due to desirable properties, including good selectivity, sorption capacity, and proper kinetics for CO₂ capture [10]. Alkaline ceramics such as Li₂ZrO₃ [11,12], Na₂ZrO₃ [13,14], Li₃BO₃ [15,16], NaBO₂ [17,18], LiYO₂ [19], NaYO₂ [19], and Li₄SiO₄ [20,21,22] have been studied as potential CO₂ captors because they can work at high temperatures with good CO₂ uptake capacity. Among these alkali ceramics, Li₄SiO₄ has attracted more attention because of its relatively better CO₂ capture reactivity and durability. However, high raw material prices seem to be the limiting factor for Li₄SiO₄ in large-scale applications [23]. Therefore, it is desired to develop low-cost alternative materials with good CO₂ capture performance.

Some studies [23,24,25,26,27,28,29] on CO₂ capture by chemisorption have been performed using sodium silicates, which may be alternatives to Li₄SiO₄. It is worth studying whether Na₄SiO₄ has the potential to be developed as a low-cost CO₂ sorbent, as well as how to reasonably improve its CO₂ uptake performance. Rodríguez [28] found that Na₂SiO₃ can achieve CO₂ capture at 80–120 °C, but the adsorption capacity was very low due to poor kinetics. Sanna [23,26,27] prepared sodium-based silicates from fly ash and Na₂CO₃, and the main phases were Na₂SiO₃ and Na₄SiO₄, which exhibited better CO₂ sorption and desorption properties at 700 °C. This suggests that Na₄SiO₄ has better reactivity in CO₂ capture compared with Na₂SiO₃. In the previous study [24], it was found that the reaction of CO₂ with Na₄SiO₄ can be expressed as Equation (1). The CO₂ capture of Na₄SiO₄ took place from 200 °C with a minor weight gain rate until rapid chemisorption occurred above 750 °C. Although Na₄SiO₄ has a wider temperature window for CO₂ capture than Li₄SiO₄, it is necessary to take measures to enhance the CO₂ uptake of Na₄SiO₄ at different temperatures, thereby reducing energy consumption at different operating conditions.

$$\mathrm{C}{\mathrm{O}}_2(\mathrm{g})+\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4\rightleftharpoons \mathrm{N}{\mathrm{a}}_2\mathrm{S}\mathrm{i}{\mathrm{O}}_3+\mathrm{N}{\mathrm{a}}_2\mathrm{C}{\mathrm{O}}_3$$

(1)

It has been reported that CO₂ capture capacity can be improved by doping with molten salts, such as halides for CaO [6,7] and nitrates for MgO [8,9]. Alkali carbonate doping is also widely adopted as an accessible and facile method to improve the CO₂ uptake performance of moderate and high-temperature sorbents [24,25,30,31,32,33]. The facilitation of alkali carbonates in CO₂ capture is usually attributed to the eutectic melt produced from doped carbonates and products, allowing the CO₂ to diffuse in the liquid phase rather than solely in the solid phase. In addition to melt formation, carbonates can produce the pyrocarbonate C₂O₅²⁻ to speed up CO₂ diffusion. Zhang [34] found that C₂O₅²⁻ acted as an active species in molten alkali carbonates to accelerate CO₂ transport. Corradini [35] proposed that CO₂ diffusion in molten CaCO₃ took place through a Grotthuss-like mechanism via C₂O₅²⁻, which was three times quicker than solely molecular diffusion of CO₂. In the previous study [24,25], it was confirmed that alkali carbonates were thermodynamically favorable for the CO₂ uptake of Na₄SiO₄ under an air atmosphere or 80 vol% CO₂–20 vol% N₂ because alkali carbonates could bind CO₂ to form C₂O₅²⁻. This provided a new idea for alkali carbonates to promote CO₂ chemisorption of Na₄SiO₄ from 200 °C, apart from melt-accelerated CO₂ diffusion. However, when the mass ratio of Na₄SiO₄ and single carbonates was 1:0.5 [24], the proportion of Na₄SiO₄ in sorbents decreased, resulting in a lowered CO₂ adsorption capacity. Under air atmosphere, the CO₂ uptake performance of alkali Na₂CO₃-doped Na₄SiO₄ was also found to become progressively better when the content of Na₂CO₃ in the Na₄SiO₄-based sorbents was reduced from 66.7 wt% to 50 wt% and then to 33.3 wt% [25]. In light of this, the proportion of alkali carbonates in the sorbents must be appropriately reduced. Given that the contribution of carbonates includes the production of pyrocarbonates and melts, different single alkali carbonates should be reasonably combined to exert the promoting effect of carbonates corresponding to the solid or molten state at different temperatures. On the other hand, the enhanced CO₂ uptake kinetics of Na₄SiO₄ at different temperatures by pyrocarbonates and carbonate melts has not yet been discussed. Hence, it is required to fit the CO₂ uptake results using a kinetic model so that the positive roles of alkali carbonates can be assessed based on kinetic parameters. The double exponential model and Eyring’s model are the most commonly used models for analyzing CO₂ uptake kinetics, which take into account two independent processes: surface chemisorption and bulk diffusion, and fit well with experimental data [14,36,37]. Then, the fitted kinetic results are estimated by Eyring’s model to determine the activation energies of the different processes [14,38].

As a continuation of our previous work [24,25], here, Na₄SiO₄ will be synthesized and mixed with 5, 10, and 20 mol% of Li₂CO₃–Na₂CO₃ or Li₂CO₃–Na₂CO₃–K₂CO₃ as CO₂ sorbents. In situ XRD and Raman will be performed to characterize the phase evolution and structural transformations during the CO₂ uptake of alkali carbonates doped Na₄SiO₄. The CO₂ sorption properties will be investigated by isothermal analyses at different temperatures, and the CO₂ uptake kinetics will be evaluated by the double exponential model and Eyring’s model. This study will focus on the contribution of pyrocarbonates that emerged from alkali carbonates in different species and amounted to the CO₂ uptake performance of Na₄SiO₄ at different temperatures.

2. Materials and Methods

2.1. Sorbents

The chemicals in this study, such as NaOH, Na₂SiO₃·9H₂O, and alkali carbonates, were of analytical grade and bought from Aladdin Chemical Reagent Co., Ltd., Shanghai, China. Na₄SiO₄ was prepared by a wet mixing method using NaOH and Na₂SiO₃·9H₂O as the raw materials. Firstly, NaOH and Na₂SiO₃·9H₂O with a molar ratio of 2:1 were dissolved by de-ionized water in a 50 mL nickel crucible, and the solution was stirred for 1 h and dried in a vacuum drying oven at 100 °C for 24 h. Then, the powder was fired in a high-temperature furnace at 600 °C for 1 h under ambient air conditions, and blocky Na₄SiO₄ was obtained. After grinding, the Na₄SiO₄ powder was sieved and obtained with particles of 150–250 μm. Then, the binary carbonates Li₂CO₃–Na₂CO₃ and ternary carbonates Li₂CO₃–Na₂CO₃–K₂CO₃ at 28–48 μm were mixed with Na₄SiO₄ with different molar ratios of 5:95, 10:90, and 20:80; these sorbents were labeled as x(carbonates)yNa. For example, the sample named 10(LiNa)90Na refers to a mixture with a molar ratio of (Li₂CO₃–Na₂CO₃):Na₄SiO₄ = 10:90, while P–Na refers to pristine Na₄SiO₄. The doped binary carbonates or ternary carbonates were equimolarly mixed and pre-melted. The melting points of Li₂CO₃–Na₂CO₃ and Li₂CO₃–Na₂CO₃–K₂CO₃ are 500 °C and 393 °C [21], respectively.

2.2. Characterization

The X-ray powder diffraction patterns of the specimens were collected by a laboratory diffractometer (Bruker D8 ADVANCE, Bruker Inc., Karlsruhe, (Baden-Württemberg), Germany). The room temperature XRD data were collected in the range 2θ = 10–80 ° with a scan rate of 2°·min⁻¹ at ambient atmosphere. In situ XRD characterization consists of two aspects: (1) the full-pattern scanning model in steps of 0.02 ° over the range 15–65 °. CO₂ chemisorption from 50 to 850 °C in an 80 vol% CO₂–20 vol% N₂ atmosphere with a heating step of 10 °C·min⁻¹. When it was 850 °C, the temperature was held constant for 1 h, and the atmosphere switched to a 100 vol% N₂ atmosphere for desorption; (2) the characteristic peak range of 32.8–35.4 ° using the position-sensitive detector (PSD) fixed model, in which the experimental condition remained unchanged, and the heating rate was changed to 5 °C·min⁻¹ from 50 °C to 850 °C. Desorption was performed under a 100 % N₂ atmosphere and accompanied by a cooling rate of 2 °C·min⁻¹ from 850 °C to 50 °C. The gas flow rate was always 50 mL·min⁻¹. The collected diffraction data of in situ XRD was refined by DIFFRAC.TOPAS5.0 software to determine the mass percentage of different phases in the samples at different temperatures.

In situ Raman spectra of different samples during CO₂ capture were measured using a laser confocal Raman spectrometer (LabRAM HR Evolution, Horiba France SAS, Villeneuve d’Ascq (Nord-Pas-de-Calais Region), France). A carrier table attached to the measurement system was used to load samples for in situ measurements from ambient temperature to 800 °C with a heating rate of 5 °C·min⁻¹. The atmosphere composition and airflow size in the in situ Raman experiments were kept consistent with the experimental conditions described above.

The morphological features of the sorbents were characterized by scanning electron microscopy (SEM) using an FEI 450 (FEI, Hillsboro, NH, USA). To ensure visualization, a layer of platinum (thickness of 3–4 nm) was coated on the samples to enhance conductivity.

2.3. CO₂ Sorption

The CO₂ uptake performance of the sorbents was evaluated using a thermogravimetric analyzer (SDT650, TA Instrument, New Castle, DE, USA). Approximately 20 mg of the sample were loaded into a platinum crucible and heated from ambient temperature to 350 °C and held for 10 min under 100% N₂ at 50 mL·min⁻¹ to remove moisture. Then, the temperature was changed to the targeted temperatures. Dynamic analyses were conducted from 50 to 900 °C under the 80% CO₂–20% N₂ atmosphere. Isothermal experiments for pristine and doped Na₄SiO₄ were carried out in two temperature ranges, 200–300 °C and 500–600 °C for 120 min. Since the focus of this study is on the promotion of Na₄SiO₄ by alkali carbonates in CO₂ capture, 80 vol% CO₂ was selected to make the enhancement more significant and easily measured. Experiments with 20 consecutive cycles of CO₂ adsorption–desorption were performed at 600 °C with a flow of 80% CO₂ for adsorption (50 mL/min for 60 min) and 850 °C in pristine N₂ for desorption (50 mL/min for 90 min).

Since the theoretical weight gain of different samples will vary with the content of Na₄SiO₄, the conversion ${\chi }_{\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4}$ was proposed to describe the capture performance. The conversion ${\chi }_{\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4}$ was estimated according to Equation (2), in which ∆m denotes the weight change before and after adsorption, m₀ denotes the theoretical capacity, x_n is the molar fraction of compositions, and M_n is the relative molecular weight of compositions.

$$\chi =\frac{\mathsf{\Delta }m}{{{\displaystyle m}}_0}\times 100\%=\frac{\Delta m\times \Sigma (x_{\mathrm{n}}\times {\mathrm{M}}_{\mathrm{n}})}{{\mathrm{M}}_{\mathrm{C}{\mathrm{O}}_2}\times {\chi }_{\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4}}\times 100\%$$

(2)

3. Results and Discussions

3.1. Characterization of the Prepared Na₄SiO₄

In this study, Na₄SiO₄ was prepared by a wet mixing method using NaOH and Na₂SiO₃·9H₂O as raw materials, heated from ambient temperature to 600 °C at a heating rate of 2 °C·min⁻¹, and it was maintained for 60 min. Figure 1a shows that the XRD plots of the prepared Na₄SiO₄ are consistent with the standard diffraction pattern of Na₄SiO₄ (JCPDSNo.78–1432). The results of XRD characterization suggested that the desired Na₄SiO₄ was successfully prepared. No typical peaks of impurities were detected, indicating that the NaOH and Na₂SiO₃·9H₂O could react adequately under the set synthetic conditions. The morphology of the prepared Na₄SiO₄ is shown in Figure 1b, presenting the cluster-shaped particles with a dense surface, as shown in the enlarged image. Additionally, there are obvious gaps in the distribution between the particles.

3.2. In Situ XRD

To study the phase changes in CO₂ uptake of the P–Na, the in situ XRD plots of the P–Na were collected at 25–850 °C, as shown in Figure 2a. When it was 100 °C, only the Na₄SiO₄ phase was detected. When the temperature went up to 200 °C, the Na₂SiO₃ phase could be detected, which was the product of the carbonization reaction between Na₄SiO₄ and CO₂. When the temperature climbed to 400 °C, the Na₂CO₃ phase was found. The peak intensities of the Na₂SiO₃ and Na₂CO₃ phases gradually strengthened with increasing temperature, but the characteristic peak of Na₄SiO₄ was not found above 400 °C. From 400 to 800 °C, the dominant phases were Na₂SiO₃ and Na₂CO₃ during the CO₂ capture of Na₄SiO₄. In the mixture exposed to 80 vol% CO₂, the Na₂SiO₃ phase persisted after formation, and the SiO₂ phase was not found, implying that Na₂SiO₃ did not react with CO₂. This is different from the conclusion reported by Rodríguez [28] that Na₂SiO₃ can achieve CO₂ capture. When the temperature was up to 850 °C, the Na₂CO₃ phase vanished because 850 °C was approaching the melting point of Na₂CO₃, 860 °C. After that, the atmosphere was changed to 100% N₂ and kept for 60 min. Both the Na₂SiO₃ and Na₄SiO₄ phases were found, which indicated that desorption had not been completed under this condition.

To investigate the phase evolution of alkali carbonate-doped Na₄SiO₄ during CO₂ uptake, 10(LiNa)90Na and 10(LiNaK)90Na were used as representatives and characterized in the same temperature regime and atmosphere as described above, and their XRD plots are shown in Figure 2b,c. It should be mentioned that when the mass percentage of a certain phase in the mixtures is less than 5 wt%, it is not detectable in the XRD measurement. This allowed us to exclude the interference of additives to the main phases, including Na₄SiO₄, Na₂SiO₃, and Na₂CO₃, during the study of phase transition. With the addition of carbonates, the Na₂SiO₃ phase was detected at 100 °C. As the temperature increased, the Na₄SiO₄ phase disappeared at 400 °C. When the temperature gradually increased to 800 °C, Na₂SiO₃ and Na₂CO₃ were the predominant phases, and no other miscellaneous phases were found. This indicated that alkali carbonates enhanced the CO₂ chemisorption of Na₄SiO₄ and did not change the direction of the reaction from the reactant, Na₄SiO₄, to the products, Na₂SiO₃ and Na₂CO₃. In addition, the mixed carbonates did not cause the reaction between Na₂SiO₃ and CO₂. After desorption at 850 °C for one hour, Na₂SiO₃ and Na₄SiO₄ phases were found, suggesting incomplete desorption despite the addition of carbonates. As shown in Figure 3, the mass percentage of different phases in in situ XRD characterization varies with temperature from 25 to 850 °C. For samples containing 10 mol% of Li₂CO₃–Na₂CO₃ or Li₂CO₃–Na₂CO₃–K₂CO₃, when the temperature rose to 200 °C, the mass fraction of the Na₄SiO₄ phase decreased rapidly to less than 5%, while Na₂SiO₃ and Na₂CO₃ increased rapidly and accounted for the majority at higher temperatures. In contrast, the same effect was only achieved at 600 °C for P–Na.

To further explore the CO₂ capture behavior of Na₄SiO₄ promoted by alkali carbonates, the collection of typic peaks in carbonation–decarbonization was recorded by in situ XRD using a position-sensitive detector (PSD) fixed model at the temperature range of 50–850 °C. The phase changes of P–Na are shown in Figure 2d. The sorbent was heated in the 80 vol% CO₂–20 vol% N₂ atmosphere and cooled in 100% N₂ at the same flow rate, corresponding to different ramping rates of 5 °C·min⁻¹ and 2 °C·min⁻¹, respectively. Figure 2e,f represents the continuous phase changes of 10(LiNa)90Na and 10(LiNaK)90Na at the same temperature and atmosphere regime. The carbonation and decarbonization processes are discussed specifically below:

Carbonization: In Figure 2d, the Na₂SiO₃ phase appeared at 300 °C. After which, the Na₄SiO₄ phase disappeared quickly with the increasing temperature. From 300 °C to 850 °C, the Na₂SiO₃ phase appeared in abundance with Na₂CO₃. As the temperature increased, the intensities of Na₂CO₃ and Na₂SiO₃ increased slowly. In Figure 2e,f, due to the addition of carbonates, the phase of Na₄SiO₄ vanished gradually at 100 °C. Hence, it was confirmed that the alkali carbonates allowed the CO₂ capture of Na₄SiO₄ to occur at lower temperatures. Under the same experimental conditions, for samples 10(LiNa)90Na and 10(LiNaK)90Na, the intensities of the Na₂CO₃ and Na₂SiO₃ phases slightly strengthened with increasing temperature compared with those of P–Na. This suggests the promotion of doped carbonates to Na₄SiO₄ in CO₂ capture.

Decarbonization: When the temperature rose to 850 °C, the gas flow was changed to 100% N₂. In Figure 2d, Na₂CO₃ and Na₂SiO₃ phases were always present from 850 to 50 °C, and the weak intensity of Na₄SiO₄ was detected. The desorption did not proceed completely because the time spent in desorption was not long enough, and the temperature decreased gradually. In Figure 2e,f, the intensities of the Na₂SiO₃ phases weakened rapidly, and the Na₄SiO₄ phase appeared with higher intensity than that of the P–Na. When the temperature dropped from 850 to 765 °C, the Na₂CO₃ phases could be found in Figure 2e,f. This was because the product Na₂CO₃ and the mixed carbonates formed a melt at higher temperatures. As the temperature continued to decrease, the Na₂SiO₃ phase was observed. This indicates that the promotion of doped carbonates in desorption was less pronounced than in adsorption.

In situ XRD characterization of the carbonate-doped Na₄SiO₄ indicated that alkali carbonates could speed up the reaction of Na₄SiO₄ and CO₂ to form Na₂CO₃ and Na₂SiO₃, and no other impurity phases appeared. This is quite different from the results proposed by other studies on CO₂ capture by Na₂SiO₃ to produce SiO₂ [28]. During the CO₂ uptake process, the doped alkali carbonates worked as a facilitator during CO₂ uptake and did not change the reaction direction from the reactant Na₄SiO₄ to the final products Na₂SiO₃ and Na₂CO₃. For decarbonization, the weak peak intensity of the Na₄SiO₄ phase below 800 °C indicated the small contribution of carbonates to regenerating the Na₄SiO₄.

3.3. In Situ Raman

To investigate the structural transformation during CO₂ uptake by carbonate–doped Na₄SiO₄, samples of 10(LiNa)90Na and 10(LiNaK)90Na were used as representatives characterized by in situ high–temperature Raman. Figure 4 shows the in situ Raman results measured from room temperature to 800 °C in an atmosphere containing 80 vol% CO₂ in the wavenumber range of 500 to 1700 cm⁻¹. These Raman–type peaks contained two fundamental vibrational modes: υ₁–symmetric stretching vibrations and υ₃–asymmetric stretching vibrations, associated with CO₃²⁻, C₂O₅²⁻, SiO₃²⁻ and SiO₄⁴⁻ ions [34,39,40,41,42]. At ambient temperature, it was detected that typic peaks at 823 cm⁻¹ related to the υ₁ mode of Si–O in Na₄SiO₄ [41], and 1078 cm⁻¹ were related to the υ₁ mode of C–O in alkali carbonates [34]. When the temperature was above 100 °C, the peak of υ₃ mode at 609 cm⁻¹ and υ₁ mode at 962 cm⁻¹ of Si–O in Na₂SiO₃ [39,40,42] appeared, and the intensities strengthed gradually. Additionally, the peak intensity of υ₁ mode at 1078 cm⁻¹ related to C–O also exhibited a trend of gradual enhancement due to the product, Na₂CO₃. Correspondingly, the peak intensity of υ₁ at 823 cm⁻¹ related to Na₄SiO₄ showed a weakening trend. The peak at 1580 cm⁻¹, related to the υ₁ mode of the specie C₂O₅²⁻, was detected from 150 °C. These results indicate that the added alkali carbonates can exhibit significant promotion from 100 °C. Additionally, at temperatures below 300 °C, no melt was present in the mixtures. This suggests that the pyrocarbonates can be emerged from solid carbonates and replace CO₂ to react with Na₄SiO₄, expressed as Equation (3).

$${\mathrm{C}}_2{\mathrm{O}}_5^{2-}+\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4=\mathrm{N}{\mathrm{a}}_2\mathrm{S}\mathrm{i}{\mathrm{O}}_3+\mathrm{N}{\mathrm{a}}_2\mathrm{C}{\mathrm{O}}_3+\mathrm{C}{\mathrm{O}}_3^{2-}$$

(3)

When the temperature rose above 500 °C, the carbonate melt was formed. The intensities of υ₁ mode at 609 cm⁻¹ and υ₃ mode at 962 cm⁻¹ related to the Si–O in Na₂SiO₃ maintained an upward trend when the temperature increased. Meanwhile, the intensity of υ₁ at 1078 cm⁻¹ assigned to CO₃²⁻ also showed a tendency to increase. The typical peaks of C₂O₅²⁻ at 1580 cm⁻¹ and Na₄SiO₄ at 823 cm⁻¹ could also be found in the magnified view. The intensity of υ₁ mode associated with Na₄SiO₄ at 823 cm⁻¹ decreased significantly with increasing temperature, and this typic peak disappeared when it was higher than 600 °C. This revealed that the facilitating effect of alkali carbonate was manifested by the provision of pyrocarbonate and the formation of melt at higher than 500 °C, accelerating CO₂ transport and improving the CO₂ uptake performance.

3.4. Isothermal Analyses and Kinetic Evaluation

Dynamic thermal analyses of Na₄SiO₄ mixed with different amounts of alkali carbonates were performed from amient temperature to 900 °C in the 80% CO₂ –20% N₂ atmosphere, and the results were shown in Figures S1–S4. CO₂ uptake can be roughly divided into two temperature ranges, above and below 400 °C, corresponding to carbonate mixtures in solids and melts. Isothermal adsorption was performed in two temperature ranges, 200–300 °C and 500–600 °C, to clarify the facilitation of alkali carbonates in different phase states.

To explore the positive influence of solid carbonates in different quantities on the CO₂ sorption capacity of Na₄SiO₄, isothermal experiments were conducted at 200–300 °C for 120 min under an atmosphere containing 80% CO₂, and the results are shown in Figure 5a–c. The weight gain of all samples at different temperatures and the conversion of Na₄SiO₄ contained therein were listed in Table S1. The sorption capacity of all samples increased with increasing temperature. It was determined that the rate and adsorption capacity of CO₂ capture by Na₄SiO₄ were significantly enhanced at different temperatures due to the mixed alkali carbonates. At the same addition ratio, the improvement of Li₂CO₃–Na₂CO₃ seemed to be relatively close to that of Li₂CO₃–Na₂CO₃–K₂CO₃, and samples containing Li₂CO₃–Na₂CO₃ exhibited slightly better chemisorption. Among sorbents, 10(LiNa)90Na had the best CO₂ capture performance at 200–300 °C. The weight gain of 10(LiNa)90Na after sorption at 300 °C for 120 min was 13.3%, corresponding to ${\chi }_{\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4}$ = 58.6%. Under the same conditions, the weight gain of sample P–Na was 7.9%, corresponding to ${\chi }_{\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4}$ = 33.1%. However, the performance became worse when the carbonate content was doubled to 20 mol%.

To compare the contribution of molten carbonates and pyrocarbonates to the CO₂ sorption capacity of Na₄SiO₄, isothermal experiments were conducted at 500–600 °C for 120 min, and the results are shown in Figure 5d–f. In this temperature range, there was little difference in the promotion effect of binary and ternary carbonates. Overall, 5 mol% and 10 mol% of carbonates seemed to perform a similar promotion. At 500–600 °C, 10(LiNa)90Na showed the best CO₂ uptake performance among sorbents. The weight gain of 10(LiNa)90Na after sorption at 600 °C for 120 min was 16.7 %, corresponding to ${\chi }_{\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4}$= 74.9%. Under the same conditions, the weight gain of sample P–Na was 13.8 %, corresponding to ${\chi }_{\mathrm{N}{\mathrm{a}}_4\mathrm{S}\mathrm{i}{\mathrm{O}}_4}$= 57.7%. However, the CO₂ uptake performance became poor when the carbonate content was doubled to 20 mol%. It is suggested that the appropriate amount of alkali carbonate addition is critical to enhancing the CO₂ capture performance of Na₄SiO₄. Therefore, when the molar ratios of alkali carbonates in the samples are the same, the samples containing Li₂CO₃–Na₂CO₃ or Li₂CO₃–Na₂CO₃–K₂CO₃ show similar CO₂ uptake performance at the same experimental temperatures.

Taking P–Na, 10(LiNa)90Na, and 20(LiNa)80Na as representatives, the CO₂ sorption results of the samples were simulated by a double exponential model [14,27,43] expressed as Equation (4), in which y is the weight gain of CO₂ sorbents, t is the CO₂ uptake time, A₁, A₂, and A₃ are the pre-exponential factors, and k₁ and k₂ are the exponential constants for chemisorption and bulk diffusion, respectively. As shown in

Table 1. Kinetic parameters of P–Na, 10(LiNa)90Na, and 20(LiNa)80Na at 200–300 °C.

Sample	T (°C)	k1 (s−1)	k2 (s−1)	R2
P–Na	200	1.21 × 10−3	1.21 × 10−4	0.999
	250	1.93 × 10−3	1.56 × 10−4	0.999
	300	2.89 × 10−3	2.10 × 10−4	0.999
10(LiNa)90Na	200	1.97 × 10−3	2.21 × 10−4	0.999
	250	2.73 × 10−3	2.66 × 10−4	0.999
	300	3.81 × 10−3	3.10 × 10−4	0.999
20(LiNa)80Na	200	1.57 × 10−3	1.58 × 10−4	0.999
	250	2.24 × 10−3	1.96 × 10−4	0.999
	300	3.25 × 10−3	2.45 × 10−4	0.999

and

Table 2. Kinetic parameters of P–Na, 10(LiNa)90Na, and 20(LiNa)80Na at 500–600 °C.

Sample	T (°C)	k1 (s−1)	k2 (s−1)	R2
P–Na	500	2.91 × 10−3	2.01 × 10−4	0.999
	550	3.46 × 10−3	2.29 × 10−4	0.999
	600	4.23 × 10−3	2.63 × 10−4	0.999
10(LiNa)90Na	500	4.56 × 10−3	2.96 × 10−4	0.999
	550	5.13 × 10−3	3.21 × 10−4	0.999
	600	5.71 × 10−3	3.47 × 10−4	0.999
20(LiNa)80Na	500	3.34 × 10−3	2.27 × 10−4	0.999
	550	3.88 × 10−3	2.51 × 10−4	0.999
	600	4.49 × 10−3	2.76 × 10−4	0.999

, the value of the chemisorption rate (k₁) is an order of magnitude greater than that of the bulk diffusion rate (k₂) for all samples. The values of k₁ and k₂ for all samples at 500–600 °C were greater than those of samples at 200–300 °C. This indicates that the carbonates were more effective when they were in melt compared to solid form. The values of k₁ and k₂ for 10(LiNa)90Na were also greater than those of P–Na and 20(LiNa)80Na, which means that 10 mol% was a more appropriate additive amount at different temperatures.

$$y=A_1\cdot \mathrm{e}\mathrm{x}\mathrm{p}(-k_1\cdot t)+A_2\cdot \mathrm{e}\mathrm{x}\mathrm{p}(-k_2\cdot t)+A_3$$

(4)

To provide further discussion on the temperature dependence of alkali carbonates at 200–300 °C on CO₂ capture by Na₄SiO₄, the activation energies of chemisorption and bulk diffusion for P–Na, 10(LiNa)90Na, and 20(LiNa)80Na were simulated according to the Eyring’s model [14] expressed as Equation (5), in which k denotes the reaction rate constant, i.e., k₁ and k₂ above. t denotes the absolute temperature in units of K⁻¹. R is the gas constant 8.314 J·mol⁻¹·K⁻¹. k_B is the Boltzmann constant 1.380649 × 10⁻²³ J·K⁻¹. h is the Planck constant 6.62607015 × 10⁻³⁴ J·s. ΔH⁺⁺ and ΔS⁺⁺ represent the activation enthalpy and activation entropy, respectively, which can be fitted to obtain

$$\mathrm{l}\mathrm{n}(k/T)=-\mathsf{\Delta }{H}^{++}/(\mathrm{R}\mathrm{T})+ln(k_B/h)+\mathsf{\Delta }{S}^{++}/\mathrm{R}.$$

(5)

Accordingly, the fitted functional relationship of k and T is shown in Figure 6a. The mixed alkali carbonates lead to a decline in the activation energy of chemisorption and diffusion for CO₂ capture of Na₄SiO₄. In the case of chemisorption, the fitted activation energies were 15.30, 10.50, and 12.03 kJ·mol⁻¹ for P–Na, 10(LiNa)90Na, and 20(LiNa)80Na, respectively, which means that the carbonates can diminish the temperature dependence of chemisorption of Na₄SiO₄. Combined with in situ Raman characterization, it can be inferred that the doped carbonate opens up new chemical reaction pathways, as expressed in Equation (3), which greatly accelerates the CO₂ uptake reaction and diminishes the activation energy of chemisorption. In terms of the bulk diffusion, the activation energy of P–Na was 8.07 kJ·mol⁻¹, and it dropped to 3.34 and 5.41 kJ·mol⁻¹ for 10(LiNa)90Na and 20(LiNa)80Na, respectively. The significant reduction in activation energy was attributed to the formation of pyrocarbonates, which reacted with Na₄SiO₄, followed by the release of CO₃²⁻ and its recombination with CO₂. The oxo-Grothuss mechanism could improve the CO₂ uptake kinetics of Na₄SiO₄ [24]. However, more alkali carbonate content in the sorbents did not necessarily lead to better chemisorption. For 20(LiNa)80Na, the diffusion process required more activation energy than that of 10(LiNa)90Na, which indicates that too much alkali carbonates covering the sorbent surface will block the diffusion of CO₂.

The activation energies of chemisorption and diffusion for P–Na, 10(LiNa)90Na, and 20(LiNa)80Na at 500–600 °C were evaluated by Eyring’s model. As shown in Figure 6b, the addition of alkali carbonates contributed to a remarkable decrement in the activation enthalpy for chemisorption and bulk diffusion of CO₂ capture by Na₄SiO₄. The ΔH⁺⁺ values of chemisorption calculated were 14.11, 5.80, and 9.77 kJ·mol⁻¹ for P–Na, 10(LiNa)90Na, and 20(LiNa)80Na, respectively, which means that mixed alkaline carbonates result in the reduced temperature dependence of chemisorption. In the case of the diffusion process, the ΔH⁺⁺ of P–Na was 7.98 kJ·mol⁻¹, and the values of 10(LiNa)90Na and 20(LiNa)80Na dropped to 2.09 and 4.14 kJ·mol⁻¹, respectively. In contrast, the activation energies of both chemisorption and bulk diffusion were found to be further reduced at 500–600 °C than at 200–300 °C, especially for bulk diffusion. Combined with in situ Raman characterization, it can be concluded that at 500–600 °C, pyrocarbonate and molten carbonates make joint contributions to the CO₂ uptake of Na₄SiO₄. The pyrocarbonate breaks up a new CO₂ uptake pathway, and the molten carbonate reduces the resistance to CO₂ diffusion. Therefore, polycarbonates can significantly improve the kinetics of CO₂ capture by Na₄SiO₄. However, when the content of alkali carbonates was added to 20 mol%, the diffusion process required more activation energy than that of 10(LiNa)90Na, which was attributed to excessive carbonate melts, hindering CO₂ diffusion to Na₄SiO₄.

SEM images of P–Na and 10(LiNa)90Na after isothermal experiments at 300 °C and 600 °C are shown in Figure 7. For P–Na, after two hours of isothermal analysis at 300 °C, the overall morphology did not change much, and the surface showed dense particles in Figure 7a. As shown in Figure 7b, when the temperature was up to 600 °C and isothermal analysis was carried out for two hours, the large particles were refined into small particles after the reaction. However, in individual areas, only some traces of surface reaction could be observed. For 10(LiNa)90Na, after two hours of isothermal chemisorption at 300 °C, there was a distribution of fine particles on the surface of large particles, and some areas of unreacted material were exposed, as shown in Figure 7c. In Figure 7d, the reaction product of 10(LiNa)90Na was distributed in small particles and had a certain number of small sparse pores distributed on the surface. Alkali carbonates changed the morphology of Na₄SiO₄, refining the particles and forming a porous structure, thus increasing the reactivity of Na₄SiO₄.

The CO₂ capture capacities of Na₄SiO₄ phasically doped with different alkali carbonates under different experimental conditions were compared, as shown in

Table 3. Weight gain of Na₄SiO₄-based sorbents under different experimental conditions.

Sample	Dopant	Sorption Conditions	Weight Gain, %	Reference
Na2CO3-Fly ash	-	Isothermally, 700 °C, 100% CO2, 120 min	9.8	A. Sanna [27]
Na2CO3-Fly ash	5 mol% Li2CO3	Isothermally, 700 °C, 100% CO2, 120 min	10.2	A. Sanna [27]
Na2CO3-Fly ash	10 mol% Li2CO3	Isothermally, 700 °C,100% CO2, 120 min	11.4	A. Sanna [27]
Na2CO3-Fly ash	10 mol% K2CO3	Isothermally, 700 °C,100% CO2, 120 min	9.9	A. Sanna [27]
Na4SiO4	33.3 wt% Li2CO3	Dynamic thermally, 50–1000 °C, 80% CO2	14.9	Liu [24]
Na4SiO4	33.3 wt% Na2CO3	Dynamic thermally, 50–1000 °C, 80% CO2	7.7	Liu [24]
Na4SiO4	33.3 wt% K2CO3	Dynamic thermally, 50–1000 °C, 80% CO2	10.8	Liu [24]
Na4SiO4	33.3 mol% Na2CO3	Isothermally, 550 °C, Air, 190 min	3.2	Wang [25]
Na4SiO4	-	Isothermally, 300 °C, 80% CO2, 120 min	7.9	This work
Na4SiO4	-	Isothermally, 600 °C, 80% CO2, 120 min	13.8	This work
Na4SiO4	10 mol% (Li2CO3-Na2CO3)	Isothermally, 300 °C, 80% CO2, 120 min	13.3	This work
Na4SiO4	10 mol% (Li2CO3-Na2CO3)	Isothermally, 600 °C, 80% CO2, 120 min	16.7	This work

. The contribution of single alkali carbonates to CO₂ uptake varied by species and proportion for different Na₄SiO₄-based sorbents. Larger proportions of alkali carbonates in sorbents did not imply better facilitation. In this study, alkali carbonates were employed to accelerate CO₂ capture and were found to significantly improve the CO₂ uptake performance of Na₄SiO₄, especially at lower temperatures.

3.5. Cyclic Stability

Multiple absorption–desorption cycles were conducted to study the influence of doped carbonate on the long-term stability of CO₂ capture, and the results are shown in Figure 8. 10(LiNa)90Na was selected for cyclic experiments because of its better performance in isothermal analyses. As shown in Figure 8a, for P–Na, the first adsorption capacity was close to 50% of the theoretical value, but the desorption was 26% of the theoretical value. In the following 19 cycles, adsorption and desorption tended to be stable, reaching 30% of the theoretical conversion. In Figure 8b, for 10(LiNa)90Na, the capacity reached 60% of its theoretical value in the first adsorption, but desorption was not carried out completely. In the following 19 cycles, the capacity of adsorption and desorption increased gradually. The conversion finally reached 52% of the theoretical value. In general, alkali carbonates contributed greatly to the durability of CO₂ capture by Na₄SiO₄. It is suggested that molten carbonates can disperse the reaction products of Na₄SiO₄ and CO₂, reducing the aggregation of products. The gradual improvement of conversion is due to the redistribution of products and regeneration products in multiple cycles so that unreacted Na₄SiO₄ can contact and react with CO₂.

4. Conclusions

In this study, it was confirmed that the CO₂ uptake performance of Na₄SiO₄ can be effectively enhanced when mixed with Li₂CO₃–Na₂CO₃ or Li₂CO₃–Na₂CO₃–K₂CO₃. Compared with P–Na, alkali carbonate-doped Na₄SiO₄ achieved CO₂ capture from 100 °C. This is due to the fact that the alkali carbonates reacted with CO₂ at different temperatures to form pyrocarbonates, C₂O₅²⁻, which in turn reacted with Na₄SiO₄ to form Na₂SiO₃, as shown in Figure 9. When the temperature is higher than the eutectic point of alkali carbonates, the molten alkali carbonates become a carrier to accelerate CO₂ diffusion. When the temperature is above 800 °C, the spent Na₄SiO₄ will be regenerated due to the reaction between the products of Na₂SiO₃ and Na₂CO₃, whereafter CO₂ is released. In the CO₂ desorption, the molten alkali carbonates seem to provide liquid channels for CO₂ diffusion through the solid product layers.

When the proportion of alkali carbonates in the sorbents was the same, there was little difference in the promotion between binary and ternary alkali carbonates. The CO₂ uptake capacity of carbonate-doped Na₄SiO₄ increased dramatically as the proportion of carbonates was increased from 5 to 10 mol%. Overall, 10 mol% of Li₂CO₃–Na₂CO₃ was found to be the optimum addition for CO₂ capture of Na₄SiO₄ and exhibited the best CO₂ uptake capacity at different temperatures. At 200–300 °C, the activation energy values of the chemisorption and the bulk diffusion were 12.03 kJ·mol⁻¹ and 3.34 kJ·mol⁻¹, respectively, corresponding to 15.30 kJ·mol⁻¹ and 8.07 kJ·mol⁻¹ for pristine Na₄SiO₄. This was attributed to the fact that alkali carbonates could bind CO₂ to produce pyrocarbonates, C₂O₅²⁻. The alkali pyrocarbonates were formed at 100 °C and reacted with Na₄SiO₄, followed by the release of CO₃²⁻ and its recombination with CO₂. The Grothuss-like mechanism improved the CO₂ uptake kinetics of Na₄SiO₄ when doped carbonates were present in the solid state. At 500–600 °C, the enhancement of carbonates was more pronounced, as both pyrocarbonates and melts worked simultaneously. The activation energy values of chemisorption and bulk diffusion for 10(LiNa)90Na were 7.98 kJ·mol⁻¹ and 2.09 kJ·mol⁻¹, respectively, which were lower than those of 14.11 kJ·mol⁻¹ and 5.80 kJ·mol⁻¹ for pristine Na₄SiO₄. In addition, alkali carbonates significantly improved the cycle stability of CO₂ captured by Na₄SiO₄. After 20 cycles, the conversion of Na₄SiO₄ mixed with 10 mol% Li₂CO₃–Na₂CO₃ exceeded 52% of the theoretical value, which was much higher than 26% for the pristine Na₄SiO₄.

In conclusion, the CO₂ capture performance of Na₄SiO₄ was improved by alkali carbonates in solid and melt states, which were both beneficial for the CO₂ uptake kinetics of Na₄SiO₄. Future efforts will be focused on CO₂ desorption performance and cyclic stability to develop economical Na₄SiO₄-based CO₂ sorbents.