2.4.3 Mineral carbonation
Mineral carbonation is one of technologies utilizing CO2, used to form carbonated materials by the reaction between CO2 and Ca or Mg-bound compounds such as wollastonite (CaSiO3), olivine (Mg2SiO4), and serpentine (Mg3Si2O5(OH)4) (Maroto et al. 2005). This technology can be also considered as an accelerated carbonation in terms of making the reaction shorter using high purity CO2. One of advantages of this technology compared to other CO2 storage technologies is that CO2 is stably stored in final products such as CaCO3 and MgCO3.
2.5 Carbon dioxide capture sorbents
A variety of materials have been reported (Feng et al. 2007) to be able to absorb CO2. These can be classified as follows:
Microporous and Mesoporous Materials. These include carbon-based sorbents such as activated carbon, carbon fiber, and carbon molecular sieves, zeolites, and chemically or physically modified mesoporous materials such as MCM- 41.
Metal Oxides. Many metal oxides exhibit carbonation and calcination reaction. However, a majority of these metal carbonates are thermally stable and calcined only at higher temperatures. The calcination temperatures of some metal carbonates (CaCO3 ~750
°C, MgCO3 ~385 °C, ZnCO3 ~340 °C, CuCO3 ~290 °C and MnCO3
~440 °C) are within the temperature range of interest (200-800 °C)
(Butt et al. 1996). Various metal oxides such as those listed in Table 2-1 have strong affinity to CO2.
Table 2-1 CO2 capacity of metal oxides (Feng et al. 2007)
Metal Oxide CO2 capacity (g of CO2/g of oxide)
Hydrotalcite-like Compounds. These belong to a large class of anionic and basic clays, also known as layered double hydroxides (LDH). They are composed of positively charged brucite-like (Mg(OH)2) layers with trivalent cations substituting for divalent cations at the centers of octahedral sites of hydroxide sheet whose vertex contain hydroxide ions, and each -OH group is shared by three octahedral cations and points to the interlayer regions (Feng et al. 2007).
2.6 Absorption system to capture CO2
Reaction-based processes can be used for separating CO2 from flue gas. This process is based on the carbonation reaction in which gaseous CO2 reacts with a solid metal oxide (represented by MO) to yield the metal carbonate (MCO3) (Gupta & Fan 2002).
MO + CO2 MCO3
Once the metal oxide has reached its ultimate conversion, it can be thermally regenerated to the metal oxide and CO2 by heating the metal carbonate beyond the calcination temperature. The calcination reaction can be represented by
MCO3 MO + CO2
There are two types of adsorption systems determined by IEA (International Energy Agency) that have been widely studied (Gomes and Yee 2002; Takamura et al. 2001; Ishibashi et al. 1996; Diagne et al. 1995) PSA (Pressure Swing Adsorption) and TSA (Temperature Swing Adsorption).
In the PSA process, CO2 gases are captured at higher pressure and released at lower pressure, this technology requires a vacuum unit for regeneration. Similarly, TSA process adsorbed CO2 at lower temperature and desorbed by heating. One important economical factor in these technologies is strongly related to the ultimate adsorption capacity of the solvent/sorbent (quantified by kg CO2/ton solvent or sorbent) and kinetics of the adsorption process. Higher equilibrium capacities allow lower sorbent/solvent requirement and handling; lower regeneration costs and therefore lower capital and operating cost.
2.7 Carbonation and calcination cycles in the CaO-CO2 absorption process
Calcined lime (main component, CaO) can be used to capture CO2 in the exhaust gas or in the reactor during the utilization of fossil fuels (Wang et al. 2009). That is, calcium oxide (CaO) absorbs CO2 to yield calcium carbonate (CaCO3), and the CaCO3 is then thermally decomposed to CaO, releasing nearly pure CO2 for sequestration. In fact, the heat for decomposing CaCO3 can be supplied by combusting fossil fuels, such as coal and natural gas, in a calciner. These two reactions are showed as follow:
CaO(s) +CO2 (g) →CaCO3(s) ∆H0 298=−178 kJ/mol CaCO3(s) →CaO(s) + CO2 (g) ∆H0 298=178 kJ/mol
Specifying the conditions for a carbonation step must strike a balance between high temperatures which favor the speed of reaction, and low temperatures which favor the equilibrium conversion. Barker (1973) found that the carbonation reaction took place in two stages; an initial rapid rate was followed by a slower approach to a conversion plateau.
Both the limestone and dolomite react with CO2 in these two stages.
The initial carbonation stage is kinetically controlled with a fast reaction rate between CaO and CO2. At this stage, the outlet CO2 concentrations are not higher than 3% for a stable absorption period. After the stable absorption period, CO2 concentrations abruptly increased and then increased very slowly, indicating that the CO2 capture capacity of the CaO was nearly exhausted. Owing that large part of active CaO was converted to CaCO3 and the carbonation stage moved to the second stage controlled by the diffusion in the product layer (Fang et al 2009b). During
the latter and slower stage of the reaction, it was found that CO2
concentration does not affect the rate of the reaction if it is maintained much higher than the corresponding equilibrium concentration for that temperature.
Abanades and Alvarez (2003) reported that the maximum carbonation conversion decreased during the carbonation/calcination cycles due to the loss in the porosity associated with the small pores and the increase in the porosity associated with the large pores. The formation of a larger volume molar product (CaCO3) compared to CaO led to the plugging of these pores thereby causing a loss in the active surface area.
Figure 2-6 shows the concept of pore filling and plugging at the pore-mouths of these sorbent particles by CaCO3 product layer, preventing the access of CO2 to un-reacted CaO at the pore interiors.
Figure 2-6 Non-reacted and reacted particle of CaCO3 (Hassanzadeh and Abbasian2010)
The calcination reaction is favored by higher temperatures. The reactions proceed only if the partial pressure of CO2 in the gas above the solid surface is less than the decomposition pressure of the CaCO3. The latter pressure is determined by equilibrium thermodynamic considerations (Stanmore and Gilot 2005). A typical expression for equilibrium decomposition pressure Peq (Silcox et al. 1989) is:
Peq 4.137 10 e T atm
Figure 2-7 plots three of the expressions listed in the literature; the agreement is good except at lower temperatures. (Silcox et al. 1989; Hu and Scaroni 1996; Hartman and Tmka 2003).
Figure 2-7 Decomposition pressure of carbon dioxide over calcium carbonate (Stanmore and Gilot 2005)
2.8 Sorbent deactivation
After the reaction of CaO with CO2, the product CaCO3 must undergo calcination to regenerate CaO for repeated use (Fang et al. 2009).
In the calcination process, some pores are produced inside the CaO particle but, at the same time, CaO sintering occurs because of the high calcination temperature, which reduces the surface area and porosity with an increasing residence time. These two factors are very important for the reaction of CaO with CO2, but sintering sharply reduces surface area and porosity, which strongly affects the CO2 capture capacity and the reaction rates of CaO with CO2. As the number of carbonation/calcination sorbents already approach their lowest ultimate conversion, degradation is found to be more severe under more highly sintering calcination conditions at higher temperatures, higher residence times, higher partial pressures of CO2 and H2O. Abanades and Alvarez (2003) indicated that for limestones, it appears that the reaction ceases when the product has built up to a depth of 50 nm, averaged across the total surface area.
Thus, the loss of surface area by sintering is a further contributing factor in the fall in sorbent capacity. However, there is still controversy surrounding the exact mechanism of deactivation. Also Abanades et al.
(2007) mentioned that the surface texture of cycled limestone commonly features shrinkage of smaller pores, usually accompanied by growing macro pores. These trends are typical of solid-state sintering in an intermediate stage (Randall and German 1996) in which vacancies (or voids) generated by temperature-and-ion-sensitive lattice defects direct void volume from smaller to larger pores, whereas mass moves in the opposite direction.
2.9 Cost assessments
Freund (2003) estimated the costs for CO2 transportation $1–$3/ t /100 km and sequestration ($4–$8)/t CO2, these costs are small compared to the cost for CO2 capture, estimated at $35– $55/t CO2 capture (Abanades and Alvarez 2003). Singh (2003) indicated that the high cost of CO2 capture is due to the considerable amount of energy required in the separation process. Therefore, reducing the cost of CO2 capture is absolutely necessary to make CCS more economically attractive.
However, the cost of sorbent must also be considered. Costs of limestone and dolomite are $26.7/ t and $30/ t, respectively (U.S. GPO 2005).
Manovic and Anthony (2008) cited that synthesized sorbent with high performance and low cost would be highly beneficial for the CO2 capture process using the carbonation/calcination cycle.
2.10 Magnesium carbonate
The reversible chemical reaction for CO2 removal involving magnesium oxide is:
MgO + CO2 ↔ MgCO3 ∆H≈-96 kJ/mol
Some reports suggest that high CO2 absorption is possible only by chemisorptions of carbon dioxide molecules with metal oxide at higher temperatures (Song et al. 1998). However Bhagiyalakshmi et al. (2010) found that the mesoporous MgO is highly basic with well-ordered pores to hold high CO2 at lower and higher temperatures. The large surface active sites of mesoporous MgO initially holds the CO2 molecules with smaller affinity and are trapped into the pores by chemical reaction of MgO and CO2 to form MgCO3. It has been shown (Hassanzadeh and
Abbasian 2010) that the behavior of these materials is related to their lattice structure.
When the solid is porous in nature, the gaseous reactant diffuses into the interior pores of the particles and reacts with the active solid species at the surface of the pores. This can be described by the grain model, in which the solid particle is treated as an assemblage of numerous smaller grains. Surrounding these grains are macro-pores through which the gas has to diffuse to reach the grains. The reaction occurs at the surface of each grain, according to the un-reacted shrinking core model. As the reaction proceeds, the difference in the molar volumes of the solid product and the molar volume of the reactant results in an increase in the grain size. This decreases the pore volume between the grains, and also decreases the diffusion rate of the gaseous reactant through the sorbent particles.
At temperatures below the equilibrium temperature for carbonation of magnesium oxide (i.e. between 300 and 450 °C), the reactivity of the sorbent improves with increasing temperature, while at a higher temperature (i.e., 500 °C), because of the significant increase in the rate of reverse (i.e., regeneration) reaction, a decline in the sorbent reactivity is observed (Hassanzadeh and Abbasian 2010).
2.11 CaO sorbents mixed with other metals
It is believed that the capacity decay in CO2 absorption process is mainly owing to the sintering of CaO and CaCO3 (formed during carbonation) in the regeneration process, or the physical aggregation of the crystals leading to increased particle size, or reduced surface area of the produced CaO for the carbonation reaction in the next cycle. In order
to address this loss-in-capacity, many methods have been tested with varying degrees of success (Liu et al. 2010).
It has been reported (Li et al. 2009) that the method of incorporating the inert materials as MgO has a critical effect on the long term stability of the CaO-based absorbent. The investigation of Albrecht et al. (2008a) also showed that the rate of decline in CaO activity can be reduced by incorporating finely dispersed MgO in the sorbent. The absorption capacity of a limestone-based sorbent with 20 wt % MgO was 45% greater than that of a similar material without MgO by the end of the test, and the rate of decline in absorption capacity was very small.
In the study of Liu et al. (2010) a mixture of CaO and MgO was formed, in which during the regeneration of the CaO sorbent at 900 °C, the MgO particles, with a sintering temperature of 1289 °C, act as a physical barrier to prevent the sintering and aggregation of the CaCO3
nano particles, which typically sinter at 527 °C. Therefore the high CO2
capture capacity of the sorbent is maintained over a multitude of carbonation- regeneration cycles.
Chrissafis and Paraskevopoulos (2005) indicated that the degrading performance of CaO in cycles was mainly due to sintering, and by mixing high melting point compounds, such as Al2O3 or MgO, sintering may be inhibited to some extent. Li et al. (2009) reported a type of stable MgO-doped CaO absorbent, produced by mechanical mixing of small MgO particles with Ca(CH3COO)2, followed by high temperature calcination, gives as high as 53 wt % CO2 capacity after 50 carbonation/decarbonation cycles. But without MgO addition the absorption capacity decrease to 26wt% for the 50th cycle under the same conditions. Some studies related to the enhancement of the sorbent for carbonation/calcination cycles can be found in Table 2-2.
Table 2-2 Performance of some Adsorbents for CO2 Capture (Li et al. 2009) Materials System Conditions Absorption
Capacity Reference
CHAPTER THREE EXPERIMENTAL METHOD
3.1 Procedure and method
There are several parts involved in the operation of the experimental study: sorbent preparation, characterization of the samples, CO2 sorption/desorption condition and systems procedures. Figure 3-1 shows the flow chart of this work.
3.2 Sorbent preparation technique
The sea water precipitated CaCO3 and MgCO3 sorbents were obtained from Power Research Institute, Taiwan (PRIT). A method permitting a selected production of calcium carbonate and magnesium carbonate precipitates was developed (Lan and Hong 2005) for the fixation of carbon dioxide by chemical precipitation it is described briefly as follows:
50 ml of 4 M sodium hydroxide solution and 20 ml of 2 M sodium carbonate solution were added to 1.8 L of sea water in a 2L flask. This mixture was magnetically stirred for about 30 min then allowed to settle for 4 hr. After that, three fourths of the supernatant sea water was considered to be provided by directly injecting carbon dioxide into the resulting sea water a flow rate of 1L/min for 48 hours. Sufficient amount of carbon dioxide was considered to provide by directly injecting carbon dioxide into the resulting sea water at flow rate of 1.0 L per minute for 48hrs. In this instance, calcium carbonate precipitates could be
exclusively separated from the suspended sea water by 0.45um membrane filtration. Magnesium carbonates precipitates, were produced by complete drying of the filtrate sea water, rinsed with de ionized water and dried.
Figure 3-1 Flow chart of the experimental study
Calcium and Magnesium Carbonate mixture
The mixture of calcium and magnesium carbonate for the production of an anti-sintering sorbent were prepared by the following procedure based on the study of Chen (2009).
Appropriate amounts of calcium carbonate and magnesium carbonate were mixed with distilled water and methanol. The purpose of adding methanol was to dissolve the magnesium carbonate for better mixing.
The above mixture was vigorously stirred for 1 hour at 75oC, and then dried at 110oC overnight. After drying, the sample was ground into a fine powder.
Figure 3-2 Calcium and Magnesium carbonate mixture procedure.
3.3 Characterization
Characterization of CaCO3 and MgCO3 were done by XRD, SEM, ICP-MS and BET surface area measurements. These tests were performed in order to study the chemical and physical composition, and the morphology of the samples.
X-ray Powder Diffractometer (XRPD)
CaCO3 and MgCO3 were characterized by X-ray diffraction (XRD) patterns on a D/MAX-RB diffract meter using Cu KR radiation (λ) 1.5406 Å. The operating conditions are at an emission voltage of 30 kV, and an emission current of 20 mA. XRPD patterns were obtained for crystalline phase detection between 10 and 70°
(2θ).
BET surface area
BET surface area and pore size distribution measurements were performed using nitrogen adsorption and desorption isotherms on a Micromeritics ASAP 2000 volumetric adsorption analyzer. The CaCO3 and MgCO3 sorbents were degassed at 350 °C at a pressure of 10-6m bar for at least 8 hr. in the degassing port of the apparatus before the actual measurements. The pore size distribution measurements were obtained using the BJH method.
Scanning Electron Microscope (SEM)
The morphology of the products was observed on a Hitachi S-4700 scanning electron microscope (SEM) with 15 keV of acceleration. The samples were prepared by placing the sorbents on double-sided carbon tape mounted on the sample holder.
3.4 Equipment and chemicals component
1. CO2 analyzer (AGM4000-1-2-2) Zhi Shang instrument, Taiwan.
Reaction time: T90/45, Lower Detection Limit: 1%. Sample Flow Rate: 120cc/min, Linearity Error < 2%.
2. Furnace: Thermolyne, 1400, USA
3. Magnetic stirrer: Cimares2, Thermolyne, Lowa, USA 4. Sieve: 16-30 mesh, Zhong Xin, Taiwan
5. Nitrogen (N2) gas cylinder: 99%, Taiwan chiah lung company 6. Carbon dioxide gas cylinder: 20% CO2/N2, Taiwan chiah lung
company
7. Air gas cylinder: Taiwan chiah lung company
8. CaCO3 and MgCO3 :Power Research Institute , Taiwan Power Company
9. Methanol: You He company 10. Limestone: Hualien, Taiwan
3.5 Experimental apparatus TGA system
A thermo gravimetric analysis (TGA) (TG 209 F1, NETZSCH, Germany) was used to obtain the CO2 absorbed amount.
The TGA consists of an electronic balance, vertical furnace, reactor tube, a carrier gas system, and computerized data acquisition system. The crucible is made of Al2O3. A schematic diagram of the TGA system for cyclic CO2 dry absorption is shown in Figure. 3-3
Figure 3-3 Schematic of TGA system for absorption
The two chemical reactions occurring during the cyclic test are:
Carbonation: CaO(s) + CO2(g) CaCO3 (s) Calcination: CaCO3(s) CaO(s) + CO2 (g)
A small amount of the sample (25-30mg) was placed in the crucible; the gas flow rate was 20 ml/min for a typical run the temperature was brought at 850°C for calcination at rate of 40oC/min.
After complete calcination, the temperature was decreased at a rate of 60
°C/min to the carbonation temperature. When approaching the carbonation temperature the valve was switched to allow the gas mixture of CO2 with an inlet concentration of 20 vol% (N2 balance) to flow over the calcined absorbent.
The conditions for TGA experiments are described in Table 3-1.
Sorbent
Table 3-1 Typical test condition for the cyclic test of carbon dioxide capture using TGA
Experimental parameters Conditions Weight of the sample 25~30 mg
Absorption cycle conditions 650-730℃, 20% CO2/N2, 10min
Desorption cycle conditions
850℃, pure N2, 1min Heating rate 40℃/min, pure N2
Cooling rate -60℃/min, pure N2
Flow rate 20ccm
In order to find the optimal condition for the CO2 capture, different parameters were also tested. Table 3-2 shows the changes in carbonation, calcination temperature, the effect of the cooling and heating rate.
Table 3-2 Change in parameters for finding the optimal conditions for the carbon dioxide capture using TGA
TGA Changes in parameters
Effect on carbonation temperature ( ) 650, 700, 730, 760 Effect on desorption temperature ( ) 800, 825, 850
Effect of Cooling rate (℃/min) 20, 60
Effect of heating rate (℃/min) 40, 50, 60 Absorption time (min) 10, 30
Desorption time (min) 1
3.6 Packed column test system
Packed column tests were also performed to evaluate the cyclic CO2 capture efficiency and its corresponding CO2 absorbed amount as well. The CO2 capture efficiency was calculated using the following equation:
CO2 capture efficiency = 1 - C1/ C0,
Where C1 is the CO2 outlet concentration and C0 is the CO2 inlet concentration.
A schematic diagram of the packed column system for carbonation/calcination cyclic CO2 is shown in Figure 3-4. For avoiding the pressure drop, all the sorbents were pelletized, crushed, and then sieved into grains using No.16 and 30 meshes (corresponding to 0.59
~1.19 mm in grain size). 8.2±0.3g of sorbent was packed in a quartz tube with an inner diameter of 1.5 cm.
Figure 3-4 Schematic of packed column system for absorption
Before absorption, the sorbents were pretreated under N2 flow at 850oC for 1 hour, and then the temperature was decreased to 700oC for carbonation with CO2 inlet concentration of 15 vol. % (balanced with 6 % O2 and 79 % N2). The carbonation time was 10 min. After carbonation, the column temperature was increased to 850oC and kept for 30 min
N2
under N2 flow to release CO2, and then the temperature was decreased to 700oC again for the cyclic test.
The total gas flow rate was 500 ml/min (at 1 atm, 25°C) for both carbonation and calcination processes, which corresponded to an empty-bed gas residence time of around 0.36 s at absorption temperature of 700oC. The inlet and outlet concentrations of CO2 were continually
The total gas flow rate was 500 ml/min (at 1 atm, 25°C) for both carbonation and calcination processes, which corresponded to an empty-bed gas residence time of around 0.36 s at absorption temperature of 700oC. The inlet and outlet concentrations of CO2 were continually