CO₂ Sorption and Regeneration Properties of K₂CO₃/Al₂O₃-Based Sorbent at High Pressure and Moderate Temperature

Abstract

In this study, the CO₂ sorption mechanisms and regeneration properties of alumina-based sorbent using K₂CO₃ loading under high-pressure and moderate temperature conditions were examined. To investigate the mechanism of CO₂ sorption, a zirconium-based sorbent was compared with an alumina-based sorbent. The CO₂ capture capacities of the PAI10, 20, 30, and 40 were 32.3, 63.0, 95.4, and 124.5 mg CO₂/g sorbent, respectively. To investigate the CO₂ sorption mechanism of an alumina-based sorbent, we performed XRD, TG/DTG, and FTIR analyses after CO₂ sorption in the presence of 10 vol% CO₂ and H₂O each at 200 °C and 20 atm. For PAI10–40 sorbents, KHCO₃ and KAl(CO₃)(OH)₂ phases were observed by TG/DTG and FTIR analysis. For PAI-x sorbents, it was confirmed that the captured CO₂ is desorbed completely at a temperature below 400 °C at 20 atm.

1. Introduction

Fossil fuels, coal, petroleum, and natural gas account for about 86 percent of the world’s primary energy. Greenhouse gas emissions are associated with human activities [1]; in particular, carbon dioxide (CO₂) is a major greenhouse gas emitted into the atmosphere by coal-fired power plants. Improved power generation technology can help prevent global warming and climate change by reducing carbon dioxide emissions from coal-fired power plants. Currently, several technological processes for CO₂ capture have been widely investigated: pre-combustion, post-combustion, and oxyfuel technology. In the case of pre-combustion process, CO₂ capture occurs before combustion from the synthesis gas of an integrated gasification combined cycle (IGCC). The total pressure in the pre-combustion process is 20–70 atm, and the CO₂ content is between 15 and 60%. In addition, the separation of CO₂ is significantly easier because of the high CO₂ partial pressure when compared with the post-combustion process [2,3,4]. The capture of CO₂ using dry CO₂ sorbent is one of the most efficient techniques in an IGCC plant. The ideal dry solid CO₂ sorbent for applying the pre-combustion process must have a high CO₂ capture capacity, a fast sorption rate, high selectivity for carbon dioxide, a stable CO₂ capture capacity for multiple cycles, and excellent physical properties, due to the high pressure and moderate temperature [5,6]. Many researchers use inorganic materials such as magnesium oxide, aluminum oxide, hydrotalcite, calcium oxide, and lithium-containing material to develop a dry CO₂ sorbent, applicable in the pre-combustion process in IGCC plants [7,8,9,10,11,12,13,14,15,16,17]. K₂CO₃-promoted alumina-based CO₂ sorbents have several advantages over these materials, including a lower cost, high physical strength, and a relatively high CO₂ capture capacity [18,19,20,21,22,23,24,25,26,27,28,29,30]. As a result, many researchers have been studying the interaction between carbon dioxide and alumina-based CO₂ sorbents loaded with K₂CO₃ at relatively low temperatures of below 150 °C and low atmospheric pressure.

To understand the interaction between CO₂ and K₂CO₃ loaded alumina-based CO₂ sorbent under high-pressure and moderate temperature conditions, we investigated the CO₂ capture capacity and sorption mechanism of a K₂CO₃ loading at 200 °C and 20 atm.

This article reports the CO₂ capture capacities and absorption mechanism of PAI10–40 sorbents prepared by the impregnation method. X-ray diffraction (XRD), thermogravimetric analysis/differential thermal gravimetry (TGA/DTG), and temperature-programmed desorption (TPD) experiments were also carried out to investigate the CO₂ sorption mechanism, physical change, and regeneration properties of an alumina-based sorbent, respectively after CO₂ sorption at 20 atm.

2. Materials and Methods

2.1. Preparation of the CO₂ Sorbent and Analysis

The aluminum oxide (Sigma Aldrich) and zirconium oxide (Sigma Aldrich) sorbent impregnated with K₂CO₃ (Sigma Aldrich) in this study was prepared by the impregnation method. The preparation procedure was as follows: Al₂O₃ and ZrO₂ supports were added to aqueous solutions of K₂CO₃ from 10 to 40 weight % in de-ionized water. Then, the solutions were mixed with a magnetic stirrer for 24 h at room temperature. After stirring, the mixed samples were dried in a vacuum rotary evaporator [5,19,20,21,22]. These samples were calcined in an air furnace for 5 h at 500 °C. The calcination temperature was raised to 500 °C at a rate of 5 °C/min and then ground and sieved to produce a particle size range of 150–250 µm in diameter. These CO₂ sorbents are denoted as PAI-x and PZIx, where P, A, Z, I, x represent potassium carbonate, aluminum oxide, zirconium oxide, the impregnation method, and the potassium carbonate loading amount, respectively.

The changes in the crystalline phase before and after CO₂ sorption were examined by XRD (Philips X’PERT) using Cu Kα radiation at the Korea Basic Science Institute. TGA and DTG (SDT Q600, TA instrument) were conducted to investigate the thermal characteristics after CO₂ sorption. Nitrogen gas was used as the carrier gas, and the CO₂ sorbent was heated from 30 °C to 500 °C at a ramping rate of 5 °C/min. The N₂ adsorption-desorption isotherms at −196 °C were recorded using a Micrometrics ASAP 2020 instrument to evaluate textural properties such as the BET surface area, pore volume, and pore size. The changes in the properties of the sorbents after CO₂ sorption were investigated by Fourier transform infrared (FTIR, PerkinElmer Inc., Waltham, MA, USA). The TPD experiment was conducted in a nitrogen atmosphere, and the sample was heated from 100 °C to 450 °C at a heating rate of 2 °C/min.

2.2. Apparatus and Procedure

The CO₂ capture capacity of the aluminum and zirconium oxide sorbents were conducted using a stainless steel fix-bed reactor with an outside diameter of 1 inch. The CO₂ sorbent was dried for 12 h at 350 °C in a nitrogen atmosphere to remove moisture and impurities from the sorbent surface. The CO₂ capture capacity of the CO₂ sorbent was investigated by packing 10 g of the sorbent into a stainless steel reactor and maintaining the CO₂ sorption pressure at 20 atm with a back pressure regulator. An analog pressure gauges and two digital pressure gauges were used to check pressures during CO₂ sorption. The composition of feed gases was 10 vol% CO₂, 10 vol% H₂O and the remainder was N₂. The GC column used in the analysis was a 1/8 inch stainless tube packed with Porapak Q. The outlet gas was automatically analyzed every 2 min with a thermal conductivity detector (Agilent, Santa Clara, CA, USA, 6890) equipped with an autosampler (Valco, Houston, TX, USA). The CO₂ capture capacity was calculated from the CO₂ breakthrough curve, which indicates the amount of CO₂ absorbed until the output concentration of CO₂ reaches 10 vol%, which is the same as the inlet concentration. The CO₂ capture capacity is determined by the amount of CO₂ absorbed per 1 g of sorbent (mg CO₂/g sorbent). The CO₂ capture capacity was calculated according to Equation (1) as follows:

$${\mathrm{CO}}_2 \mathrm{capture} \mathrm{capacity} = (\frac{P\times {\mathrm{V}}_{{\mathrm{CO}}_2}}{R\times T}\times M_{{\mathrm{CO}}_2}\times t, milligram)/(\mathit{gram}\mathit{per}\mathit{sorbent})$$

(1)

3. Results

3.1. CO₂ Capture Capacity of a CO₂ Sorbent

Table 1. Textural properties of ɣ-Al₂O₃, PZI30, and potassium-based alumina sorbents.

Sample	Surface Area (m2/g)	Pore Volume (cm3/g)	Pore Size (nm)
ɣ-Al2O3	157.65	0.31	4.91
ZrO2	8.04	0.03	3.83
PZI30	3.92	0.02	3.82
PAI10	134.98	0.22	4.32
PAI20	69.67	0.17	4.32
PAI30	44.26	0.17	4.32
PAI40	27.37	0.09	4.31

provides information on the surface area, pore volume, and pore size of ɣ-Al₂O₃, ZrO₂, PZI30 sorbent, and PAI-x sorbents. The bare ZrO₂ had a surface area of 8.05 m²/g, a pore volume of 0.02 cm³/g, and a pore size of 3.83 nm. The ɣ-Al₂O₃ had a surface area of 157.65 m²/g, a pore volume of 0.31 cm³/g, and a pore size of 4.91 nm. Compared with the ZrO₂ support material, the surface area and pore volume of the PZI30 sorbent decreased; it might be expected that the very low surface area and pore volume of PZI30 sorbent results in poor CO₂ capture capacity. When K₂CO₃ is loaded on the support materials, surface area decreases, whereas a significant change in pore size is not observed regardless of the K₂CO₃ loading amount. When the K₂CO₃ loading is 40 wt% (PAI40 sorbent), the surface area is reduced to about 17.3% compared with the Al₂O₃ support material, and the pore volume is also significantly reduced. The decrease in surface area and pore volume is caused by filling the pores or blocking the pore openings with K₂CO₃.

Figure 1 presents the breakthrough curves (a) and the CO₂ capture capacity (b) per 1 g of sorbent for PAI10 to 40 and PZI30 sorbents in the presence of 10 vol% CO₂, 10 vol% H₂O, N₂ balance at 200 °C and 20 atm. Figure 1b shows the CO₂ capture capacities of PAI10 to 40 and PZI30 sorbents calculated from the CO₂ sorption breakthrough curves shown in Figure 1a. Under high-pressure conditions (20 atm), PZI30 sorbent was used to investigate the reaction mechanism of alumina-based sorbent. PZI30 and PAI10–40 sorbents had CO₂ capture capacities of 19.7 and 32.3, 63.0, 95.4, and 124.5 mg CO₂/g sorbent, respectively. CO₂ capture capacities of PAI-x sorbents increased with increasing K₂CO₃ loading. Compared with the PZI30 sorbent, the PAI30 sorbent showed a higher CO₂ capture capacity.

Figure 2 shows the amount of CO₂ sorption per 1 g K₂CO₃ as a function of the K₂CO₃ loading. The CO₂ sorption per 1 g K₂CO₃ was calculated from the CO₂ capture capacities of PZI30 and PAI10–40 sorbents, as shown in Figure 1. The CO₂ capture capacities of PAI10–40 sorbents were maintained at approximately 311.2–323.1 mg CO₂/g K₂CO₃. These calculated values corresponded to approximately 98–101% of the K₂CO₃ theoretical value of 318.4 mg CO₂/g K₂CO₃. However, the CO₂ capture capacity of PZI30 sorbent was approximately 65.7 mg CO₂/g K₂CO₃, which is equivalent to approximately 26% of the K₂CO₃ theoretical value. The differences in the CO₂ capture capacities of PZI30 and PAI10–40 sorbents mean that the CO₂ sorbents can be absorbed through different reaction mechanisms.

3.2. Physical Characteristic Analysis after CO₂ Sorption

The results of the XRD analysis of PZI30 and PAI10–40 sorbents conducted to identify the structural changes before and after CO₂ sorption are shown in Figure 3. The XRD patterns of PZI30 in the fresh state show the sharp peaks of ZrO₂ (JCPDS No. 83-0940) and small peaks of K₂CO₃ (JCPDS No. 16-0820). For PAI10–40 sorbents before CO₂ sorption, only K₂CO₃, and ɣ-Al₂O₃ (JCPDS No. 10-0425) phases were observed. In the case of PAI10–40 sorbents, the peak intensity of K₂CO₃ increased with increasing K₂CO₃ loading. For PZI30 sorbent, the KHCO₃ phase could not be detected after CO₂ sorption in the presence of water vapor at 200 °C and 20 atm, as shown in Figure 3II. After CO₂ sorption, on the other hand, the XRD patterns of PAI10–40 sorbents showed two phases: Al₂O₃, and KAl(CO₃)(OH)₂ (JCPDS No. 15-3303). XRD patterns of PAI10–40 sorbents showed no KHCO₃ phase after CO₂ sorption regardless of the K₂CO₃ loading. Based on the XRD results, the KAl(CO₃)(OH)₂ phase formed from K₂CO₃ and ɣ-Al₂O₃ over PAIx sorbents during CO₂ sorption in the presence of 10 vol% CO₂ and 10 vol% H₂O at 200 °C and 20 atm.

FTIR analysis of PZI30, PAI10–40 sorbents, and pure KHCO₃ reagent (Sigma-Aldrich, CAS No. 298-14-6) was conducted to confirm the formation of the KHCO₃ structure after CO₂ sorption in the range of 400 to 4000 cm⁻¹. These results are shown in Figure 4. In all PZI30 and PAI10-40 sorbents, the broad peaks at a wavenumber of 3400 cm⁻¹ (the expected bond is O-H) are attributed to H₂O on the surface. In the case of PAI-x and PZI30 sorbents after CO₂ sorption, the weak peaks at wavenumber 662 cm⁻¹ (the expected bond is O₁CO₂), and 840 cm⁻¹ (the expected bond is CO₃) are attributed to carbonate. In addition, the strong peaks at wavenumber 1007 cm⁻¹ (the expected bond is C-O+C∙∙∙O), 1402 cm⁻¹ (the expected bond is O-H∙∙∙O), and 1630 cm⁻¹ (the expected bond is C=O) are attributed to bicarbonate species [17,24,27,31]. These results are in agreement with pure KHCO₃ analysis. The weak peaks at 1100 cm⁻¹ of PAI10-40 sorbents are due to Al-O stretching and Al-O-H bending vibration, and the broad peaks at 1500 cm⁻¹ of PAI10-40 sorbents are attributed to the H-O-H bend of absorbed H₂O [32]. The FTIR results show that for the K₂CO₃ loaded alumina-based CO₂ sorbent, the KHCO₃ structure was formed during CO₂ sorption in the presence of 10 vol% CO₂ and 10 vol% H₂O at 200 °C and 20 atm. We suggest that the KHCO₃ structure is formed during CO₂ sorption in the presence of water vapor at 200 °C and 20 atm [24,25,26]. According to the XRD and FTIR results, the alumina-based CO₂ sorbent could form the KHCO₃ phase and KAl(CO₃)(OH)₂ phase after CO₂ sorption in the presence of water vapor at high pressure and 200 °C.

The results of TGA/DTG of PZI30 and PAI10–40 sorbents conducted to investigate the thermal properties after CO₂ sorption are shown in Figure 5. PZI30 and PAI10–40 sorbents showed weight loss at temperatures between 30 °C and 100 °C. This is due to the desorption of adsorbed water on the sorbent surface. The weight loss of PZI30 sorbent after CO₂ sorption at 20 atm was 5% at temperatures from 140 °C to 200 °C. The DTG peak of PZI30 sorbent showed one peak at the same temperature range as the weight loss range of TGA. Figure 3II shows that the PZI30 sorbent did not change structure after CO₂ sorption at 20 atm, but the TGA/DTG confirmed that the PZI30 sorbent showed weight loss at temperatures ranging from 140 °C to 200 °C, which is attributed to CO₂ desorption from the KHCO₃ phase. At temperatures ranging from 140 °C to 350 °C, the weight loss of the PAI10–40 sorbents was 5.73%, 7.28%, 10.63%, and 11.61%, respectively. The DTG peak of PAI10 and 20 sorbents showed one peak between 200 °C and 350 °C due to the decomposition of the KAl(CO₃)(OH)₂ structure. The DTG peak of PAI30 and 40 sorbents showed two peaks at 140 °C to 200 °C and 200 °C to 350 °C, respectively. In the case of 140 °C to 200 °C, it was confirmed that the temperature range of the DTG peak was the same as for the PZI30 sorbent. In the XRD analysis for PAI30 and 40 sorbents, no KHCO₃ structures were found after CO₂ sorption, but we confirmed a weight loss in the temperature range from 140 °C to 200 °C. The TGA/DTG results confirmed that for the PAI10–40 sorbents, there are two CO₂ sorption mechanisms involving KHCO₃ and KAl(CO₃)(OH)₂ structures formed with increasing K₂CO₃ loading under water vapor at high pressure [28].

From these analysis results, it was concluded that unlike for the PZI30 sorbent, the CO₂ absorption mechanism of the PAI-x sorbent was due to the formation of KHCO₃ and KAl(CO₃)(OH)₂ structures in the presence of water vapor in the conditions of 200 °C and 20 atm.

3.3. TPD at 20 atm of PZI30 and PAI10–40 Sorbents

TPD experiments of PAI10–40 sorbents were conducted to examine the regeneration properties under 20 atm, as shown in Figure 6. In addition, a test of PZI30 sorbent was conducted to identify the decomposition temperature of the KHCO₃ structure at 20 atm. The CO₂ desorption of PAI-x sorbents started at 150 °C, and the desorbed CO₂ peak was observed in the temperature range from 150 °C to 400 °C. The TPD results of PAI10 and 20 sorbents showed one peak between 250 °C and 350 °C, and the TPD results of PAI30 and 40 sorbents showed two peaks between 150 °C and 400 °C. In the case of low K₂CO₃ loading such as PAI10 and 20, the CO₂ peak was produced by the decomposition of the KAl(CO₃)(OH)₂ structure. In the case of high K₂CO₃ loadings such as PAI30 and 40 sorbents, the CO₂ peak was produced by the decomposition of the KHCO₃ and KAl(CO₃)(OH)₂ structures. Particularly, the KHCO₃ decomposition of PAI30 and 40 sorbents appeared at the same temperature range as the PZI30 sorbent, and the CO₂ peak caused by the decomposition of KAl(CO₃)(OH)₂ increased with increasing K₂CO₃ loading. The CO₂ peak caused by KHCO₃ decomposition also increased due to increased K₂CO₃ loading. Based on the TPD results, the alumina-based CO₂ sorbent was confirmed to be capable of regeneration at temperatures below 400 °C at 20 atm. This study concluded that the potassium-loaded alumina-based sorbent had a higher CO₂ capture capacity than the zirconium-based sorbent due to the high surface area of alumina and the KHCO₃ and KAl(CO₃)(OH)₂ conversion. This means that these sorbents could be used as a support and additive to prepare a sorbent with the potential for CO₂ capture and regeneration under high and moderate temperature conditions.

4. Conclusions

This study examined the CO₂ absorption mechanisms and regeneration properties of alumina-based sorbents according to K₂CO₃ loading in the presence of 10 vol% CO₂, 10 vol% H₂O, N₂ balance at 200 °C and 20 atm. For the alumina-based sorbent loaded with K₂CO₃, such as PAI10–40, the CO₂ capture capacities were found to be 32.3, 63.0, 95.4, 124.5 mg CO₂/g sorbent, respectively, which was about 98% to 101% of the K₂CO₃ theoretical value. However, the CO₂ capture capacity of PZI30 sorbent was 19.7 mg CO₂/g sorbent, corresponding to 26% of the K₂CO₃ theoretical value. The higher CO₂ capture capacity of PAI-x sorbent relative to the PZI30 sorbent was confirmed by an XRD, FTIR analysis, and TPD experiment. This was attributed to KHCO₃ and KAl(CO₃)(OH)₂ structures. As for the regeneration properties measured in the TPD experiment, the PAI10 and 20 sorbents showed one peak due to the decomposition of KAl(CO₃)(OH)₂ at temperatures of 250 °C and 350 °C. The PAI30 and 40 sorbents showed two peaks due to the decomposition of KHCO₃ and KAl(CO₃)(OH)₂ at temperatures of 150 °C and 400 °C. In the case of PAI-x sorbent, the KHCO₃ structure generated during CO₂ sorption was confirmed to have the same decomposition temperature as the KHCO₃ structure generated after CO₂ sorption of PZI30 sorbent. This study demonstrated that the regeneration of an alumina-based sorbent is possible at temperatures lower than 400 °C. This low regeneration energy is advantageous because the CO₂ capture cost can be reduced under high pressure and moderate temperature conditions.