HCl removal performance of Mg-stabilized carbide slag from carbonation/calcination cycles for CO₂ capture

Abstract

Mg-stabilized carbide slag (MSCS) was fabricated with carbide slag, magnesium nitrate hydrate and a by-product of biodiesel from transesterification by combustion, and was used as a CO₂ sorbent in calcium looping cycles. The cycled MSCS containing a ratio of CaO to MgO of 80 : 20 from the calcium looping cycles (i.e. carbonation/calcination cycles) for CO₂ capture was subsequently used as an HCl sorbent. The HCl capture performance of the cycled MSCS which had experienced repetitive CO₂ capture cycles using calcium looping was investigated in a triple fixed-bed reactor. The reaction products of the cycled MSCS after HCl absorption are CaClOH, CaO, and MgO. MgO is an inert support. The cycled MSCS reaches the highest HCl capture capacity at 750 °C. The number of CO₂ capture cycles increases the effect of chlorination temperature on the HCl capture capacity of the cycled MSCS. The HCl capture capacity of the cycled MSCS drops slightly with the number of CO₂ capture cycles. The HCl capture capacity of the MSCS which has undergone 10 CO₂ capture cycles can retain 0.21 g HCl/g sorbent, which is about 1.7 times as high as that of the carbide slag. The presence of CO₂ leads to a reduction in the HCl capture capacity of the cycled MSCS. MSCS can maintain a more stable microstructure due to the presence of MgO during the repetitive CO₂ capture cycles. The cycled MSCS from the CO₂ capture cycles exhibits a more porous structure than the cycled carbide slag, especially in the pore size range of 2–10 nm in diameter, which benefits HCl capture. Therefore, the cycled MSCS from CO₂ capture cycles using calcium looping appears promising to remove HCl.

---

1 Introduction

Mankind faces an increasing energy crisis and environmental pollution. China relies on fossil resources such as coal for the bulk of it’s energy, and emissions like SO₂, NO_x and CO₂ are pumped into the atmosphere. Biomass and refuse derived fuel (RDF) are good substitute fuels, but their flue gas often contains a significant quantity of HCl. HCl concentration can reach 1900 ppm during combustion and gasification processes of biomass,¹ and the HCl concentration in flue gas from RDF-fired boilers can reach levels even higher than 2500 ppm.^2,3 HCl has a negative impact on the corrosion of the boiler heating surface and results in environmental pollution by polychlorinated dibenzo-dioxins (PCDDs) and furans (PCDFs).^4–6 Thus, in order to take advantage of fuels containing chlorine such as biomass and RDF as substitutes for coal, it is essential to remove HCl efficiently during the application process of such fuels.

Methods for HCl removal are commonly classified as “wet” and “dry”.⁷ In the dry method, calcium-based sorbents such as limestone and dolomite have often been used as HCl sorbents. Although the optimum reaction temperature for a chlorination reaction between CaO derived from various calcium-based sorbents and HCl may not yet be in full agreement, lots of references have reported that CaO has a high HCl capture capacity in a temperature range of 600–800 °C.⁸ Limestone can be used to capture HCl by direct chlorination (as is shown in eqn (1)) or indirect chlorination (chlorination after the decomposition of limestone as is shown in eqn (2) and (3)) at various temperatures.⁵ The indirect chlorination of limestone shows a higher HCl capture capacity than direct chlorination. CaO from different precursors other than limestone such as calcium hydroxide, dolomite, hydrogarnet, calcium magnesium acetate and calcium propionate exhibited a higher HCl capture capacity than that of limestone.^4,5,9

CaCO₃ + 2HCl → CaCl₂ + H₂O + CO₂ (1)

CaCO₃ → CaO + CO₂ (2)

CaO + 2HCl → CaCl₂ + H₂O (3)

Calcium-based sorbents are not only used to remove HCl, but are also used to capture CO₂. CO₂ capture and separation are regarded as potential methods to continue using fossil fuel-fired or biomass-fired power plants for electricity production, as ways to capture the main CO₂ emission sources.^10,11 Calcium looping using carbonation/calcination cycles of CaO is considered to be one of the most promising CO₂ capture technologies for coal-fired power plants and hydrogen production.¹² Calcium looping involves a carbonation reaction of CaO in a carbonator (650–700 °C) and a subsequent calcination reaction of the formed CaCO₃ to regenerate CaO in a calciner (>850 °C), as is shown in eqn (4). Fuel is burnt during the oxy-combustion process in the calciner to produce the necessary heat for the regeneration of CaO. The almost pure CO₂ stream from the calciner can be sequestrated or reused. Calcium looping can be applied in both pre-combustion CO₂ capture and post-combustion CO₂ capture,^12–14 which has attracted a great deal of attention recently, owing to a number of advantages: the relatively small efficiency penalty, the potential use in large-scale circulating fluidized beds; it is an excellent opportunity for integration with cement manufacture.¹⁵

However, the CO₂ capture capacity of the natural calcium-based sorbent decreases sharply with the number of cycles due to the sintering of CaO at high temperatures.^16,17 Improving the CO₂ capture capacity and cyclic stability of the calcium-based sorbent has become a research focus.^18–23 The criterion for the design of the synthetic sorbent for CO₂ capture is to increase the active surface area, stability of the pore structure, and mechanical stability of the sorbent, which enhances the CO₂ capture capacity of the sorbent during calcium looping.²⁴ Dispersing CaO into an inert carrier to weaken the sintering and aggregation of CaO particles has been frequently studied. The presented works have proved the effectiveness of adding various inert carriers such as Al₂O₃,^25–27 ZrO₂,²⁸ SiO₂,^13,29 Y₂O₃,³⁰ MnO₂,³¹ CaTiO₃,³² Nd₂O₃ (ref. 33) and calcium aluminate cement.^34–39

Dolomite mainly composed of CaO and MgO shows a higher CO₂ capture capacity than limestone during calcium looping cycles, because MgO as the inert carrier can hinder CaO sintering at high temperatures and maintain a stable pore structure during the cycles.^40,41 MgO is a good support at high temperatures.⁴² MgO does not react with CO₂ at calcium looping conditions. Some researchers have proposed MgO as an inert carrier to fabricate a synthetic CaO/MgO sorbent during calcium looping cycles. The CO₂ capture performance of synthetic CO₂ sorbents strongly depends on the synthesis routes and precursors¹⁹ and the cost should be considered for industrial applications. Broda et al.⁴³ pointed out that MgO in the synthetic sorbent prevented the sintering of CaO particles and they used calcium acetate hydrate, alcohols, alkanes or toluene as raw materials through a “one-pot” re-crystallization method. Luo et al.⁴⁴ adopted a wet mixing method to prepare CaO/MgO sorbents with calcium carbonate, calcium nitrate tetrahydrate, citric acid monohydrate, magnesium carbonate and magnesium nitrate hexahydrate. Liu et al.^45,46 synthesized CaO/MgO sorbents through a wet mixing method and spray-drying method with calcium D-gluconate monohydrate and magnesium salts of D-gluconic acid, and the sorbents could achieve a CO₂ capture capacity of 0.56 g CO₂/g sorbent after 24 cycles and 0.46 g CO₂/g sorbent after 44 cycles, respectively. Our group fabricated a novel synthetic CaO/MgO sorbent, i.e. Mg-stabilized carbide slag (MSCS), with carbide slag, magnesium nitrate hydrate and a by-product of biodiesel obtained from the transesterification process for biodiesel production.⁴⁷ MSCS showed a high CO₂ capture capacity, which could retain a CO₂ uptake of 0.42 g/g sorbent after 20 cycles.

Carbide slag as an industrial waste is a by-product from the hydrolysis process of calcium carbide (CaC₂) for the industrial production of ethylene as a main material of polyvinyl chloride.⁴⁸ Xie et al.⁴⁹ proposed a new HCl removal route where HCl can be removed by the cycled carbide slag from CO₂ capture cycles using calcium looping. They used carbide slag and aluminum nitrate to fabricate CaO/Ca₃Al₂O₆ sorbent and found that the cycled synthetic sorbent experienced 20 and 50 carbonation/calcination cycles for CO₂ capture that were 2.3 and 2.6 times greater than those of the cycled carbide slag after the number of same cycles.⁵⁰ HCl removal by sorbents from calcium looping that can realize the sequential capture of CO₂ and HCl in the flue gas of fuels containing chlorine from fired power plants is promising.

Although our previous work⁴⁷ has proved that the novel synthetic CO₂ sorbent, i.e. MSCS prepared from carbide slag, magnesium nitrate hydrate and a by-product of biodiesel, possesses a high CO₂ capture capacity and good stability during calcium looping cycles, the HCl capture performance of the cycled MSCS from CO₂ capture cycles has not been reported and is necessary to be investigated.

In this work, the HCl capture performance of the cycled MSCS from CO₂ capture cycles using calcium looping was studied. The effects of the chlorination temperature, the number of CO₂ capture cycles, the HCl concentration and the presence of CO₂ in the chlorination atmosphere on HCl removal by the cycled MSCS from calcium looping cycles were discussed. In addition, the HCl capture performance of the cycled MSCS was compared to that of the cycled carbide slag from CO₂ capture cycles.

2 Experimental

2.1 Sample preparation

Carbide slag was sampled from a chlor-alkali plant in Shandong Province, China. The by-product of biodiesel (>90 wt% glycerol content) was obtained from the transesterification process of peanut oil reacted with methanol.⁵¹ The chemical components of the carbide slag analyzed by X-ray fluorescence (XRF) are shown in Table 1. The carbide slag with a particle size of less than 0.125 mm after being sieved was chosen for the following experiments.

Table 1 Chemical components of carbide slag (wt%)

CaO	MgO	SiO2	Fe2O3	Al2O3	SrO	TiO2	Others	LOI
69.52	0.02	2.34	0.17	1.52	0.03	0.03	0.57	25.80

---

The preparation process for MSCS is presented as follows: 80 mL of the by-product of biodiesel and 50 mL of deionized water were added into a beaker and continuously stirred for 10 min. Then 5 g of carbide slag and 5.617 g of magnesium nitrate hydrate (analytical grade Mg(NO₃)₂·6H₂O, > 99 wt%) were added to the by-product of biodiesel and stirred at 80 °C for 1 h to make them completely dissolve. The mass ratio of CaO derived from the carbide slag to MgO derived from magnesium nitrate hydrate was 80 : 20. Lastly, the solution was put into a muffle furnace for combustion under air at 850 °C for 1 h.

The synthetic sorbent, i.e. MSCS, was ground and sieved to a size of < 0.125 mm.

2.2 Triple fixed-bed reactor (TFBR)

The triple fixed-bed reactor (TFBR) mainly includes a calciner, a carbonator and a chlorinator, which is operated under atmospheric pressure, as is shown in Fig. 1. A gas flow rate of 2 L min⁻¹ was controlled by the reducing valves and mass flow meters. The carbonator was operated at 700 °C for 20 min in 20% CO₂/80% N₂, and the calciner was controlled at 850 °C for 10 min in pure N₂ (99.999%). The chlorinator was operated at 700–800 °C for 90 min in a gas mixture consisting of HCl (0.10–0.20%) and N₂ (balance). 10% CO₂ was added into the chlorination atmosphere in order to investigate the effect of the presence of CO₂ on the HCl capture by the cycled MSCS. All of the gases were mixed in a mixer and the gas mixture was then sent to the TFBR.

Fig. 1 Schematic of the TFBR.

500 mg of sample was evenly put into a boat where the thickness of the sample was less than 2 mm, and was firstly placed in the carbonator. After finishing carbonation, the sample boat was sent to the calciner for complete calcination. Then the 1st CO₂ capture cycle, i.e. carbonation/calcination cycle, was finished. The calcined sample from the calciner was put into the carbonator again for the next CO₂ capture cycle. The sample after 0 cycles denotes the initial sorbent which has not experienced carbonation/calcination cycles. The sorbent samples which had experienced 0, 1, 5, 10 and 20 CO₂ capture cycles in the TFBR were sent into the chlorinator for HCl absorption. In the experimental process, the mass change of the sample was measured by an electronic balance (Mettler Toledo-XS105DU) with a resolution of 0.1 mg.

The HCl capture capacity can be calculated according to the mass change of the sample. The HCl capture capacity of the sample is defined as the HCl absorption amount per unit mass of the sample, which is presented as follows:

where N denotes the number of cycles; Y_HCl,N represents the HCl capture capacity of the cycled sample experiencing N CO₂ capture cycles, g (HCl)/g (sorbent); t denotes the reaction time, min; m_0,N represents the mass of the sample after the Nth calcination, g; m_ch1,N(t) represents the mass of the chlorinated sample experiencing N CO₂ capture cycles at t, g.

It should be noted that each group of experiments has been repeated three times, and the accidental errors were all less than 2.8%. In order to show the results clearly, four figures were chosen to draw error bars.

2.3 Phase and microstructure analysis

The phase components of the cycled samples after 90 min of chlorination were analyzed by a D/Max-III A X-ray diffractometer (XRD). The surface morphologies of the cycled samples after different conditions were observed by a SUPRATM 55 field emission scan electron microscope (SEM). An Oxford INCA sight X energy dispersive spectrometer (EDS) was used to detect elements on the surfaces of the samples. The pore structure parameters of the samples were measured by a Micromeritics ASAP 2020-M nitrogen absorption analyzer. BET specific surface areas and pore size distributions of the samples were calculated according to the Brunauer–Emmett–Teller (BET) method and Barrett–Joyner–Halenda (BJH) model, respectively.

3 Results and discussion

3.1 XRD analysis of chlorinated MSCS experiencing CO₂ capture cycles

During the preparation process of MSCS, the mass ratio of CaO to MgO was chosen as 80 : 20, because our previous work proved that the obtained MSCS possessed the highest CO₂ capture capacity during multiple cycles.⁴⁷ Fig. 2 shows the XRD spectra of the chlorinated MSCS and the chlorinated carbide slag experiencing various CO₂ capture cycles. The main products of the cycled carbide slag after 90 min of chlorination are CaO and CaClOH, as is shown in Fig. 2(a), which agrees with the results reported by Xie et al.⁴⁹ The main components of MSCS are CaO and MgO,⁴⁷ and the mass ratio of CaO to MgO is 80 : 20. The reaction products of MSCS after chlorination are CaO, MgO and CaClOH, as is presented in Fig. 2(b). This means that MgO as a support is inert and can not absorb HCl. CaCl₂ is not generated during the 90 min chlorination of MSCS. Chin et al.⁹ and Partanen et al.⁵² both thought that CaClOH should be the first product generated when HCl was captured by CaO, but CaClOH ultimately converted to CaCl₂ after a long reaction time (e.g. 800 min). However, HCl capture by the cycled MSCS from CO₂ capture cycles may be applied in the fluidized bed reactors and the residence time of the sorbent in the reactor is short. Thus, the possible chlorination product of the cycled MCSC should be CaClOH. The chlorination reaction of MSCS after 0 and 10 cycles can be described as follow.

CaO·MgO + HCl → CaClOH·MgO (6)

Fig. 2 XRD spectra of chlorinated MSCS (a) and chlorinated carbide slag (b) from various CO₂ capture cycles (chlorination at 750 °C in 0.20% HCl/N₂ for balance, chlorination time: 90 min).

3.2 Effect of the number of CO₂ capture cycles on HCl capture by cycled MSCS

The effect of the number of CO₂ capture cycles on the HCl capture performance of the cycled MSCS is depicted in Fig. 3. The HCl capture capacity and chlorination rate of the cycled MSCS are both obviously higher than those of the cycled carbide slag from the same CO₂ capture cycles. It is found that the HCl capture capacity of the cycled MSCS linearly increases with chlorination time. As the number of CO₂ capture cycles increases from 0 to 1, the HCl capture capacities of MSCS after 30 min and 90 min drop by about 30% and 11%, respectively. However, with increasing the cycle number from 1 to 20, the HCl capture capacities of the cycled MSCS only experience little change. Therefore, MSCS which has not experienced CO₂ capture cycles achieves a higher HCl capture capacity, compared to the cycled MSCS during 20 cycles. The number of CO₂ capture cycles exhibits little effect on Y_HCl,N of the cycled carbide slag during 20 cycles. After 30 min of chlorination, Y_HCl,0 and Y_HCl,10 of MSCS are about 2.5 and 2.2 times greater than those of the carbide slag, respectively.

Fig. 3 Effect of the number of CO₂ capture cycles on Y_HCl,N of cycled MSCS from various CO₂ capture cycles (chlorination at 750 °C in 0.20% HCl/N₂ for balance).

Fig. 4 presents a comparison between MSCS and the carbide slag, which have both experienced N CO₂ capture cycles, of Y_HCl,N for 90 min of chlorination. Both the cycled MSCS and the cycled carbide slag exhibit relatively stable HCl capture capacity with an increasing number of CO₂ capture cycles. The cycled MSCS achieves a much higher Y_HCl,N than the cycled carbide slag. Y_HCl,5 and Y_HCl,20 of MSCS are about 1.7 and 1.8 times higher than those of the carbide slag, respectively. This may be attributed to the support of MgO in MSCS. MSCS not only exhibits a higher cyclic CO₂ capture capacity than the carbide slag,⁴⁷ but shows a higher HCl capture capacity after CO₂ capture cycles. Therefore, MSCS is a promising synthetic sorbent for sequential CO₂/HCl capture.

Fig. 4 Comparison between MSCS and carbide slag from various CO₂ capture cycles in Y_HCl,N for 90 min chlorination (chlorination at 750 °C in 0.20% HCl/N₂ for balance).

3.3 Effect of chlorination temperature on HCl capture by cycled MSCS

The effect of chlorination temperature in the range of 700–800 °C on the HCl capture capacity of the cycled MSCS under 90 min of chlorination is shown in Fig. 5. The HCl capture capacity of the cycled MSCS increases with increasing temperature from 700 to 750 °C, while it decays when further increasing the temperature to 800 °C. Therefore, the optimum chlorination temperature of the cycled MSCS from CO₂ capture cycles is 750 °C. When the reaction temperature is low, a high temperature is beneficial for HCl absorption by CaO according to the chemical equilibrium between the gas and solid. However, when the reaction temperature is above 750 °C, the sintering of the sorbent is intensified which leads to the blockage and collapse of CaO in MSCS. That is not beneficial for HCl absorption by MSCS. Daoudi and Walters⁵³ also found adverse effects of high-temperature sintering on HCl absorption by CaO. On the other hand, Weinell et al.⁵⁴ found that at temperatures exceeding 750 °C, a liquid phase of chlorination product saturated with CaO was formed during the absorption of HCl by CaO, which is not beneficial for HCl absorption. Thus, the feasible chlorination temperature of MSCS should not exceed 750 °C according to the two possible reasons above mentioned.

Fig. 5 Effect of chlorination temperature on Y_HCl,N of cycled MSCS from various CO₂ capture cycles (chlorination for 90 min in 0.20% HCl/N₂ for balance).

The chlorination temperature in the range of 700–800 °C has little effect on the cycled MSCS during the first 5 cycles. Y_HCl,10 at 700 °C and 800 °C is about 5% and 17% lower than that at 750 °C, respectively. However, as soon as the number of CO₂ capture cycles exceeds 5, the chlorination temperature exhibits a great impact on the HCl capture capacity of the cycled MSCS. Y_HCl,20 at 700 °C and 800 °C is about 34% and 40% lower than that at 750 °C, respectively. This indicates that the number of CO₂ capture cycles intensifies the effect of chlorination temperature on the HCl capture capacity of the cycled MSCS.

3.4 Effect of CO₂ on HCl capture by cycled MSCS

When CO₂ is present in the chlorination atmosphere, CO₂ and HCl simultaneously react with calcium-based sorbents. Thus, carbonation has an impact on the chlorination of the sorbent. The effect of 10% CO₂ in the chlorination atmosphere on the HCl removal capacity of the cycled MSCS is depicted in Fig. 6. In the presence of CO₂, Y_HCl,0 of the MSCS is almost the same as Y_HCl,10, and both are lower than those in the absence of CO₂, respectively. The volume fraction of CO₂ is 50 times higher than that of HCl in the chlorination atmosphere, so the cycled MSCS may react with CO₂ preferentially to form CaCO₃. Subsequently, the formed CaCO₃ reacts with HCl. However, the pore structure of CaCO₃ is significantly less developed than that of CaO, so the HCl capture capacity of CaCO₃ is lower. Therefore, CO₂ shows an adverse impact on HCl capture by the cycled MSCS. Duo et al.⁵⁵ also found that the direct chlorination property of CaCO₃ is lower than that of CaO. Y_HCl,0 and Y_HCl,10 of MSCS in the presence of 10% CO₂ is reduced by about 35% and 20% due to the presence of CO₂, respectively. This means that the negative impact of CO₂ on HCl capture by the cycled MSCS declines with an increasing number of CO₂ capture cycles. This phenomenon can be explained as follows. In fact, the CO₂ capture capacity of the calcium-based sorbent always decays with the number of CO₂ capture cycles due to sintering. For example, our previous work⁴⁷ has shown that the CO₂ capture capacity of MSCS decreases by about 10% when increasing the cycle number from 1 to 10 in the absence of HCl. Therefore, when HCl is removed by the cycled MSCS, less CaCO₃ is formed in the presence of CO₂ with an increasing number of cycles. Accordingly, the negative impact of CO₂ on the HCl removal of the cycled MSCS also becomes lower with an increasing number of cycles.

Fig. 6 Effect of CO₂ on Y_HCl,N of cycled MSCS from various CO₂ capture cycles (chlorination at 750 °C in 0.20% HCl/N₂ for balance).

3.5 Effect of HCl volume fraction on HCl capture by cycled MSCS

Fig. 7 demonstrates the effect of the HCl volume fraction on the HCl capture capacity of MSCS and carbide slag for 90 min of chlorination which has experienced various CO₂ capture cycles.

Fig. 7 Effect of HCl volume fraction on Y_HCl,N of MSCS and carbide slag from various CO₂ capture cycles (chlorination at 750 °C for 90 min).

This shows that the HCl capture capacities of the cycled MSCS and the cycled carbide slag both almost increase linearly with HCl volume fraction. As the HCl volume fraction rises from 0.1% to 0.2%, Y_HCl,1 and Y_HCl,5 of MSCS increase by 150% and 170%, respectively. Y_HCl,1 of the carbide slag under 0.1%, 0.15% and 0.2% HCl are 78%, 45% and 9% higher than Y_HCl,5 under the corresponding HCl volume fraction, respectively. This reveals that the cycle number has a great effect on Y_HCl,N of the carbide slag under a low HCl volume fraction, but shows little effect under a high HCl volume fraction (e.g. 0.2% HCl). It is interesting to find that Y_HCl,1 and Y_HCl,5 of MSCS are almost equal, regardless of a high or low HCl volume fraction. This is possible because MSCS maintains a better pore structure with the aid of the support of MgO during the CO₂ capture cycles, while the carbide slag suffered severe sintering and requires a high HCl volume fraction to offset the adverse effect of the change of pore structure. For the same HCl volume fraction and cycle number, MSCS exhibits higher Y_HCl,N than the carbide slag. Y_HCl,5 of MSCS under 0.15% and 0.2% HCl are approximately 2.1 and 1.7 times higher than those of the carbide slag, respectively.

3.6 Microstructure analysis

The SEM-EDS mapping of MSCS is illustrated in Fig. 8. The elements of Ca, O and Mg are evenly distributed on the surface of MSCS, which means that MgO is well mixed with CaO. The good dispersion of CaO and MgO in MSCS plays an important role in MgO as the support for CaO, and helps MSCS maintain a high sintering resistance during the CO₂ capture cycles and a high HCl capture capacity. The EDS spectrogram of the MSCS in Fig. 8 is illustrated in Fig. 9. The mass ratio of Ca to Mg in Fig. 8 is approximately equal to 4.3 : 1, which agrees with that in MSCS.

Fig. 8 SEM-EDS mapping of MSCS (mass ratio of CaO : MgO = 80 : 20).

Fig. 9 EDS spectrogram of MSCS in Fig. 8.

The BET specific surface areas of MSCS and the carbide slag experiencing various CO₂ capture cycles are shown in Table 2. The cycled MSCS displays a much larger specific surface area than the cycled carbide slag for the same CO₂ capture cycles.

Table 2 BET surface areas of MSCS and carbide slag after various CO₂ capture cycles

Sample	Number of CO2 capture cycles	BET surface area (m^2 g^−1)
Carbide slag	0	10.3
Carbide slag	10	4.2
MSCS	0	28.9
MSCS	10	20.8

---

The specific surface areas of MSCS after 0 and 10 CO₂ capture cycles are 2.8 and 4.9 times higher than those of the carbide slag after the corresponding cycles, respectively. As the number of CO₂ capture cycles rises from 0 to 10, the specific surface areas of MSCS and the carbide slag decrease by 28% and 59%, respectively. This suggests that MSCS can maintain a more stable microstructure due to the presence of MgO during repetitive CO₂ capture cycles. During the combustion step in the preparation of MSCS, the by-product of biodiesel is rapidly burnt and the release of the gaseous combustion product as a foaming agent impacts the sorbent, which leads to the formation of the porous structure. Thus, the cycled MSCS from CO₂ capture cycles shows a higher surface area. A larger specific surface area is beneficial for HCl capture by calcium-based sorbents.

The pore size distributions of MSCS and carbide slag after various CO₂ capture cycles are plotted in Fig. 10. After the same cycles, the pore volume of MSCS is evidently larger than that of the carbide slag in the entire measured pore size range, especially the volume of the pores in the range of 2–10 nm. Xie et al.⁵⁰ thought that the pores distributed in the 2–10 nm range were the most important areas for HCl absorption. The volume of pores in the range of 2–10 nm (V_{2–10 nm}) and in the range of 10–150 nm range (V_{10–150 nm}) of the two sorbents after 0 and 10 cycles are presented in Table 3. V_{2–10 nm} of MSCS after 0 and 10 cycles are 4.5 and 5.7 times greater than that of the carbide slag after the same cycles, respectively.

Fig. 10 Pore size distributions of MSCS after various CO₂ capture cycles (calcination at 850 °C for 10 min in pure N₂ and carbonation at 700 °C for 20 min in 20% CO₂/80% N₂).

Table 3 Volume of pores of MSCS in the 2–10 nm and 10–150 nm range after various CO₂ capture cycles^a

Samples	Number of CO2 capture cycles	V2–10 nm (cm^3 g^−1)	V10–150 nm (cm^3 g^−1)
a V2–10 nm denotes volume of pores in size range of 2–10 nm in diameter. V10–150 nm represents volume of pores in size range of 10–150 nm in diameter.
Carbide slag	0	0.0035	0.0305
Carbide slag	10	0.0021	0.0185
MSCS	0	0.0159	0.1052
MSCS	10	0.0119	0.0661

---

Furthermore, V_{10–150 nm} of MSCS after 0 and 10 cycles are 3.4 and 3.6 times greater than those of the carbide slag, respectively. The pores in the range of 10–150 nm in diameter are beneficial for HCl diffusion and absorption. The cycled MSCS from various CO₂ capture cycles shows a higher HCl capture capacity, because it possesses better pore size distribution characteristics than the cycled carbide slag.

4 Conclusions

The cycled MSCS fabricated with industrial waste, i.e. carbide slag, magnesium nitrate hydrate and a by-product of biodiesel from transesterification by combustion, which had experienced carbonation/calcination cycles for CO₂ capture was proposed to remove HCl. The reaction products of MSCS after HCl absorption are CaClOH, CaO and MgO. MgO in the MSCS is an inert support. MSCS which has not experienced CO₂ capture cycles achieves a higher HCl capture capacity. The number of CO₂ capture cycles exhibits little effect on the HCl capture capacity of the cycled carbide slag during 20 cycles. The cycled MSCS achieves a much higher HCl capture capacity than the cycled carbide slag for the same CO₂ capture cycles. This may be attributed to the support of MgO in MSCS. Chlorination temperature in the range of 700–800 °C has little effect on the cycled MSCS during the first 5 cycles. However, as soon as the number of CO₂ capture cycles exceeds 5, the chlorination temperature exhibits a great impact on the HCl capture capacity of the cycled MSCS. The optimum chlorination temperature of the cycled MSCS from CO₂ capture cycles is 750 °C. CO₂ shows an adverse impact on HCl capture by the cycled MSCS. The negative impact of CO₂ on the HCl capture by the cycled MSCS declines with an increasing number of cycles. The good dispersion of CaO and MgO in MSCS plays an important role in MgO as the support for CaO, which helps MSCS maintain a high sintering resistance during the CO₂ capture cycles and great HCl capture capacity. The cycled MSCS experiencing CO₂ capture cycles possesses a higher surface area and better pore size distribution characteristics than the cycled carbide slag, so it exhibits a higher HCl capture capacity. It is proved that the highly active sorbent for CO₂ capture from repetitive CO₂ capture cycles can also capture HCl efficiently. Thus, MSCS is a promising synthetic sorbent for sequential CO₂/HCl capture.

Acknowledgements

Financial supports from the National Natural Science Foundation of China (51376003) and the Fundamental Research Funds of Shandong University, China (2014JC049) are gratefully appreciated. We thank Dr Hui Li for providing the by-product of biodiesel for this work.

References

1. D. Porbatzki, M. Stemmler and M. Müller, Biomass Bioenergy, 2011, 35(4), 397.
2. C. S. Chyang, Y. L. Han, L. W. Wu, H. P. Wan, H. T. Lee and Y. H. Chang, Waste Manage., 2010, 30(7), 1334.
3. M. Y. Wey, K. Y. Liu, W. J. Yu, C. L. Lin and F. Y. Chang, Waste Manage., 2008, 28(2), 406.
4. J. M. Yin and Z. S. Wu, Chin. J. Chem. Eng., 2009, 17(5), 849.
5. B. Shemwell, Y. A. Levendis and G. A. Simons, Chemosphere, 2001, 42(5–7), 785.
6. S. Fujita, N. Ogawa, T. Yamasaki, T. Fukuda, S. Sataka and K. Suzuki, Chem. Eng. J., 2004, 102(1), 99.
7. N. Verdone and P. D. Filippis, Chem. Eng. Sci., 2006, 61(22), 7487.
8. C. S. Chyang, Y. L. Han and Z. C. Zhong, Energy Fuels, 2009, 23, 3948.
9. T. Chin, R. Yan, D. T. Liang and J. H. Tay, Ind. Eng. Chem. Res., 2005, 44(10), 3742.
10. S. D. Kenarsari, D. Yang, G. Jiang, S. Zhang, J. Wang and A. G. Russell, RSC Adv., 2013, 3(45), 22739.
11. W. Z. Bi, T. Chen, R. Zhao, Z. Wang and J. H. Wu, RSC Adv., 2015, 5(44), 34913.
12. E. J. Anthony, Greenhouse Gases: Sci. Technol., 2011, 1(1), 36.
13. H. C. Chen, C. S. Zhao and Q. Q. Ren, J. Environ. Manage., 2012, 93(1), 235.
14. V. Manovic and E. J. Anthony, Int. J. Environ. Res. Public Health, 2010, 7(8), 3129.
15. C. C. Dean, J. Blamey, N. H. Florin, M. J. Al-Jeboori and P. S. Fennell, Chem. Eng. Res. Des., 2012, 43(7), 836.
16. J. C. Abanades, E. J. Anthony, D. Y. Lu, C. Salvador and D. Alvarez, AIChE J., 2004, 50(7), 1614.
17. V. Manovic, J. P. Charland, J. Blamey, P. S. Fennell, D. Y. Lu and E. J. Anthony, Fuel, 2009, 88(10), 1893.
18. H. Ping and S. Wu, RSC Adv., 2015, 5(80), 65052.
19. W. Liu, H. An, C. Qin, J. Yin, G. Wang and B. Feng, Energy Fuels, 2012, 26(5), 2751.
20. J. M. Valverde, J. Mater. Chem. A, 2012, 1(3), 447.
21. J. Blamey, E. J. Anthony, J. Wang and P. S. Fennell, Prog. Energy Combust. Sci., 2010, 36(2), 260.
22. J. C. Abanades, B. Arias, A. Lyngfelt, T. Mattisson, D. E. Wiley and H. Li, Int. J. Greenhouse Gas Control, 2015, 40, 126.
23. A. Islam, S. H. Teo, E. S. Chan and T. Y. Yun, RSC Adv., 2014, 4(110), 65127.
24. J. Sun, W. Q. Liu, Y. C. Hu, M. K. Li, X. W. Yang, Y. Zhang and M. H. Xu, Energy Fuels, 2015, 29(10), 6636.
25. Z. Li, N. S. Cai and Y. Huang, Ind. Eng. Chem. Res., 2006, 45(6), 1911.
26. M. Broda and C. R. Müller, Adv. Mater., 2012, 24(22), 3059.
27. L. Barelli, G. Bidini, A. D. Michele, F. Gallorini, C. Petrillo and F. Sacchetti, Appl. Energy, 2014, 127(7), 81.
28. A. Antzara, E. Heracleous and A. A. Lemonidou, Appl. Energy, 2015, 156(7), 331.
29. J. M. Valverde, A. Perejon and L. A. Perezmaqueda, Environ. Sci. Technol., 2012, 46(11), 6401.
30. X. Zhang, Z. Li, Y. Peng, W. Su, X. Sun and J. Li, Chem. Eng. J., 2014, 243(1), 297.
31. R. Y. Sun, Y. J. Li, H. L. Liu, S. M. Wu and C. M. Lu, Appl. Energy, 2012, 89(1), 368.
32. S. F. Wu and Y. Q. Zhu, Ind. Eng. Chem. Res., 2010, 49(6), 2701.
33. Y. C. Hu, W. Q. Liu, J. Sun, M. K. Li, X. W. Yang, Y. Zhang and M. H. Xu, Chem. Eng. J., 2015, 273(8), 333.
34. V. Manovic and E. J. Anthony, Ind. Eng. Chem. Res., 2009, 48(19), 8906.
35. V. Manovic, P. S. Fennell, J. A. Moham and E. J. Anthony, Ind. Eng. Chem. Res., 2013, 52(23), 7677.
36. V. Manovic, Y. Wu and E. J. Anthony, Environ. Sci. Technol., 2012, 46(22), 12720.
37. V. Manovic and E. J. Anthony, Fuel, 2011, 90(1), 233.
38. H. C. Chen, C. S. Zhao, Y. M. Yang and P. P. Zhang, Appl. Energy, 2012, 91(1), 334.
39. H. C. Chen, C. S. Zhao and W. W. Yu, Appl. Energy, 2013, 112(4), 67.
40. A. Silabna, M. Narcida and D. P. Harrison, J. Exp. Biol., 1996, 146(1), 149.
41. J. M. Valverde, P. E. Sanchez-Jimenez and L. A. Perez-Maqueda, Appl. Energy, 2015, 138(C), 202.
42. Y. Q. Qiao, J. Y. Wang, L. Huang, Q. W. Zheng, D. O'Hare and Q. Wang, RSC Adv., 2015, 5(101), 82777.
43. B. Marcin, M. K. Agnieszka and R. M. Christoph, Adv. Funct. Mater., 2014, 24(36), 5753.
44. C. Luo, Y. Zheng, Y. Q. Xu, N. Ding, Q. W. Shen and C. G. Zheng, Chem. Eng. J., 2015, 267(5), 111.
45. W. Q. Liu, B. Feng, Y. Q. Wu, G. X. Wang, J. Barry and C. J. Da, Environ. Sci. Technol., 2010, 44(8), 3093.
46. W. Q. Liu, J. J. Yin, C. L. Qin, B. Feng and M. H. Xu, Environ. Sci. Technol., 2012, 46(20), 11267.
47. X. T. Ma, Y. J. Li, L. Shi, Z. R. He and Z. Y. Wang, Appl. Energy, 2016, 168(4), 85.
48. Y. J. Li, R. Y. Sun, C. T. Liu, H. L. Liu and C. M. Lu, Int. J. Greenhouse Gas Control, 2012, 9(3), 117.
49. X. Xie, Y. J. Li, W. J. Wang and L. Shi, Appl. Energy, 2014, 135(12), 391.
50. X. Xie, Y. J. Li, C. T. Liu and W. J. Wang, Fuel Process. Technol., 2015, 138(10), 500.
51. H. Li, S. L. Niu, C. M. Lu, M. Q. Liu and M. J. Huo, Sci. China: Technol. Sci., 2014, 57(2), 438.
52. J. Partanen, P. Backman, R. Backman and M. Hupa, Fuel, 2005, 84, 1674.
53. M. Daoudi and J. Walters, Chem. Eng. J., 1991, 47(1), 11.
54. C. E. Weinell, P. I. Jensen and K. Dam-Johansen, Ind. Eng. Chem. Res., 1992, 31(1), 164.
55. W. Duo, J. P. K. Sevill, N. F. Kirkby and R. Clift, Chem. Eng. Sci., 1994, 49(24), 4429.

---