THE USE OF A HIGH LIMESTONE CONTENT MINING WASTE AS A SORBENT FOR CO2 CAPTURE

In this work, a high limestone content waste was evaluated as a potential material for CO2 capture. The influence of calcination conditions on the CO2 capture capacity was evaluated using 5 cycles of calcinationhydration-carbonation reactions. A Central Composite Design of Experiments was set using calcination temperatures and time as variables. The response evaluated was the CO2 capture measured by thermogravimetric analysis. The results indicate that both calcination temperature and time influence the CO2 capture capacities in the initial cycles but, after a large number of cycles, the effect becomes less relevant. The optimum calcination temperature did not change significantly between cycles about 893 °C in the first and 850 °C in the fourth cycle. However, the optimum calcination time decreased from 40.1 min in the first to 22.5 min in the fourth cycle. The maximum CO2 capture capacity declines over the reaction cycles due to the sorbent sintering, which becomes more noticeable. Moreover, the waste used in this work is suitable for separating CO2 from flue gas, achieving more than 0.2 g/g of capture capacity after five cycles.


INTRODUCTION
The continuous increase of atmospheric carbon dioxide levels since the Industrial Revolution, mostly due to fossil fuel combustion (Yu et al., 2012), has encouraged extensive research to develop technologies to reduce this gas emission (Phalak et al., 2012).CO2 capture technologies are being implemented and could become a necessary option to achieve the necessary reduction (Zeman, 2007).This process consists of three steps: separation of CO2 from flue gas -the most expensive step-, transport and sequestration.
Among the promising technologies to separate CO2 by post-combustion capture processes, carbonation-calcination loops use an extremely cheap and abundant regenerable sorbent (Abanades et al., 2004;López-Periago et al., 2013): calcium carbonate.This process is based on the cycles of the gas-solid reaction (carbonation) of calcium oxide with carbon dioxide, forming calcium carbonate (CaCO3) (Barker, 1973;Santos et al., 2012) and the reverse reaction (calcination) to regenerate the CaO, releasing a concentrated stream of CO2 (Phalak et al., 2012) which can be stored or used for other purposes such as dry ice, refrigeration equipment, carbonated beverages and fire extinguishing equipment (Kikkinides and Yang, 1993).
Despite low costs, CaO-based sorbents show a fast deactivation and a decrease in their CO2 capture capacity over reaction cycles because of sintering phenomena, crystal growth and pore blocking during the Brazilian Journal of Chemical Engineering calcination step (Abanades and Alvarez, 2003;Symonds et al., 2009;Grasa and Abanades, 2006;González et al., 2008;Manovic et al., 2009;Kuramoto, 2003).Previous studies show that high calcination temperatures accelerate this deactivation process (Grasa and Abanades, 2006;González et al., 2008).In order to overcome the sintering and enhance the sorbent reactivity, a range of methods has been proposed, such as doping of CaO-based sorbents (Al-Jeboori et al., 2012) to produce synthetic sorbents, thermal preactivation of fresh sorbent (Valverde et al., 2013) and hydrating (Yu et al., 2012;Phalak et al., 2012;Blamey et al., 2010).A number of authors have shown that including an intermediate hydration of CaO improves the sorbent morphology, forming porous agglomerates (Kuramoto, 2003) and enhancing pore size, likely a shift from smaller to larger pores, which avoids the pore pluggage (Yu et al., 2012;Phalak et al., 2012).
Previous studies investigating carbonation of Ca(OH)2 and carbonation of CaO show that Ca(OH)2 reaches a higher conversion than CaO (Nikulshina et al., 2007;Kuramoto et al., 2003), whereas the carbonation of CaO is initially chemically controlled and then becomes controlled by diffusion, while the carbonation of Ca(OH)2 at high temperatures (up to 400 °C) is all chemically controlled (Nikulshina et al., 2007).
Usually the studies that use calcium carbonate as a potential sorbent for CO2 capture processes are conducted with freshly mined materials.Although this type of material is cheap and abundant, the environmental impact of a mining process is always high.Therefore, it would be more environmentally friendly to use a residue as a CO2 sorbent.Consequently, this work concerns the evaluation of a high limestone content waste from Vale Fertilizantes produced in an apatite concentration process for separating CO2 through calcination-hydration-carbonation reactions under different calcination conditions to determine the optimum calcination temperature and time to provide the best performance.

EXPERIMENTAL Reaction Cycles
A set of 5 cycles (calcination−hydration−carbonation) was performed for each calcination condition, keeping the hydration and carbonation conditions constant.These conditions were fixed because there are studies in the literature which pointed out that the calcination step is usually the most critical one in terms of sintering the sorbent materials (Ramkumar et al., 2010).
The reactions developed in this work are provided below: The samples were calcined in a muffle furnace under 30 mL/min of synthetic air flow.Calcination temperatures and time ranges of 723-977 °C and 11-49 min, respectively, were set by central composite design of the experiments, with five replicates in the central point (850 °C and 30 min) and an alpha orthogonality factor of 1.267.Table 1 shows an experimental matrix set by the central composite design.Thereby, for each cycle, 13 runs were performed, for different calcination temperatures and times.The hydration and carbonation reactions were carried out under fixed conditions in a laboratory-scale fixed-bed reactor with internal diameter of 0.012 m and length of 0.317 m.The solid samples, after each calcination, were supported in this reactor using quartz wool, forming a packed bed and the reactor was placed in a heated zone provided by an electrical tube furnace.Liquid water was pumped at 25 °C, heated to 150 °C and then fed to the reactor as steam.Hydration was carried out at 500 °C for 30 min (Phalak et al., 2012) under 90% steam and 10% argon.Carbonation was performed at 650 °C and 30 min (Phalak et al., 2012) under 10% CO2 (balanced with an argon stream).
Figure 1 shows schematically the experimental setup used to perform the cycle of reactions.

Characterization
The crystalline phases of the waste and the extent of calcination and carbonation were tested using Xray diffraction analysis, on a Rigaku Miniflex diffractometer with CuKα radiation at 30 kV and 30 mA.The analyses were performed varying 2θ from 5 to 90 °, with steps of 0.02 ° and a time of 2 seconds per step.Average crystallite size after first and fifth calcinations was estimated with the Scherrer equation: where D is the mean size of the crystallite; κ is the Scherrer constant, which depends on the shape factor, with a typical value of about 0.9; λ is the X-ray wavelength; β(2θ) is the line broadening at half the maximum intensity, in radians; and θ is the Bragg angle.
The BET surface area was determined from N2 adsorption isotherms, at 77 K, using a Quantachrome adsorption analyser model Quantsorb Jr.

CO2 Capture Capacity
A thermogravimetric analysis (TGA) was used to evaluate the extent of CO2 capture in the cycles.After each carbonation step, about 5 mg of sample was tested in the TGA.The analyses were carried out under N2 flow and the same conditions used in the calcination step.Thus, if the sample was calcined at 750 °C for 15 min, in the TGA test the sample was heated from 25 to 750 °C under N2 and held isothermally at 750 °C for 15 min.

Reactivity of Calcium Oxide
The reactivity of the oxide formed was experimentally studied following the procedure described by ASTM -C -110-76.This methodology is based on the temperature increase due to the heat released during the hydration reaction of the oxide.After each calcination, around 2.5 g of sample and 10 g of water were introduced into an adiabatic receiver and shaken by a magnetic stirrer.The increase of the temperature was monitored using a thermometer placed inside the receiver and the temperature variation reached after 10 min was recorded.

RESULTS
A diffractogram of the waste is shown in Figure 2. As can be seen, the waste is composed mostly of calcium carbonate.Calcite (CaCO3, rhombohedral) is the predominant phase.Limestone with a high concentration of calcite has received the most focus for CO2 capture because it allows the highest capture of CO2 per unit mass (Blamey et al. 2010).The waste has a specific surface area of 3 m 2 /g, as determined by BET. Figure 3 shows XRD diffraction patterns of samples after the first and the second carbonation cycles and Figure 4 after the first and the second calcination.Both samples were calcined at 950 °C for 45 min and carbonated at 650 °C for 30 min.It can be seen that the carbonations and calcinations are reversible in practice since, after the first and second carbonations, there is only calcium carbonate and after calcinations there is mostly calcium oxide.Kuramoto et al. (2003) also reported that the sorbent regenerates, maintaining the CO2 capture capacities of this type of mineral over calcination-hydration-carbonation cycles.The authors compared the results with and without intermediate hydration of CaO, suggesting that the hydration treatment provides the durability of the sorbents in repetitive CO2 sorption.
Figure 4 shows the diffractograms of a sample after the first and fifth calcination cycles.The increase of intensity of the major peaks is evidence that the sample is sintering as it is submitted to calcinationhydration-carbonation cycles.After the first and fifth calcination the average crystallite sizes for the CaCO3 phase obtained using the Scherrer equation were (0.298 mm) and (0.340 mm) respectively.This shows that CaCO3 particles are sintering during the cycles, which makes their diameter larger and decreases their surface area.Besides the CaCO3 phase, one can also notice some small peaks of MgO due to calcined dolomite and some peaks of Ca(OH)2, probably due to the contact of the oxide with air moisture.The extent of CO2 captured during the cycles, for different calcination conditions, is shown in Figure 5.
As can be seen, samples calcined at temperatures up to 800 °C performed better in the first cycle.It can be noted that the CO2 capture capacity drops from the second to fifth cycle for all samples.However, as the number of cycles increases, the CO2 capture capacity seems to stabilize.Similar results were obtained by Rhida et al. (2012) working with natural limestones and pellets prepared from acetified limestone, with kaolin binder or with Al(OH)3 binder.Therefore, the calcination temperature seems to have a pronounced effect on CO2 capture in the first cycle, since samples calcined at higher temperatures performed much better.However the effect diminishes with increasing number of cycles, and all samples achieved similar CO2 uptake in the fifth cycle.
In Figure 5 it can be seen that samples calcined at temperatures up to 800 °C achieved almost 0.4 g/g of CO2 capture capacity in the first cycle and about 0.22 g/g in the fifth cycle.In a study using natural limestones, Rhida et al. (2012) obtained about 0.5 g/g to 0.24 g/g, in the first and fifth cycle, respectively.This result suggests that the waste used in this work is suitable for separating CO2 from flue gas.Another interesting effect can be noted in Figure 5: the samples calcined at lower temperatures showed an increase in the CO2 uptake from the first to second cycle.This probably happened because the use of lower temperatures is not sufficient to decompose completely CaCO3 into CaO in the first cycle.Figure 6 shows XRD data for a sample calcined at 800 ºC for 5 minutes (Fig. 6A) and for 30 minutes (Fig. 6B).The presence of a small CaCO3 diffraction peak (2θ around 30 o ) in Fig. 6B indicates that the use of a temperature of 800 ºC is not sufficient to calcine completely the sample, even after 30 minutes.Therefore, the use of temperatures below 800 ºC, such as 723 ºC and 750 ºC (shown in Figure 5) are possibly insufficient to convert all the CaCO3 into CaO, even if longer calcination times are used.After the second calcination-hydration-carbonation cycle, probably all CaCO3 was converted to CaO.
Figure 7 shows the response surface for each cycle.In the first and second cycles, the CO2 capture capacity increased, as the calcination temperature increased, until reached the best region and then began to decline.This is expected because, for low calcination temperature and time, there is probably still unreacted carbonate in the first cycles and, as the calcination temperature and time increase, the sorbent become less reactive due to the sintering processes.However, from the third to fifth cycles the ability to capture CO2 for all conditions tends to the same value.For a larger number of cycles, it seems reasonable to assume that the decay in CO2 capture capacity would be more dependent on the amount of cycles than on the calcination conditions, within this experimental range.
The maximum capacity declined over the cycles.For instance, in the first cycle, the maximum CO2 capture capacity was around 0.42 gram of CO2 per gram of waste and in the fifth cycle, it was only about 0.23.Previous studies investigating CO2 capture with the same type of materials also show the decay in the maximum capture capacity along the cycles (Grasa and Abanades, 2006;González et al., 2008).Usually this decline in the CO2 capture capacities is attributed to the sintering phenomena (Abanades and Alvarez, 2003;Symonds et al., 2009) which becomes more noticeable with each cycle and is enhanced by increasing calcination temperatures (Borgwardt, 1989).
The optimum conditions for each reaction cycle are shown in Table 2.For the fifth cycle, the optimum conditions were out of the experimental region.The CO2 capture capacities for the first to fifth cycles can be described by Equations ( 5) to (9), respectively.Brazilian Journal of Chemical Engineering Vol. 33, No. 03, pp. 599 -606, July -September, 2016 Results obtained in reactivity tests of samples calcined under five different conditions are shown in Figure 8.It can be observed that samples calcined at temperatures up to 800 °C are more reactive, but as the number of cycles increases, the reactivity of all the samples seems to converge to similar values.The results suggest that reactivity of the oxide, as reflected in the CO2 capture capacities, depends more strongly on the number of cycles than on calcination conditions within the range tested.

CONCLUSIONS
In this paper, we have experimentally investigated the CO2 capture capacity of a high limestone content waste via five calcination-hydration-carbonation reaction cycles under different calcination conditions.The parametric optimization shows that the increase of the calcination temperature and time increases the quality of quicklime and CO2 capacity capture for the initial cycles, until the best region of operation is reached and then it decreases.Furthermore, for the first cycle, those samples treated at higher calcination temperatures showed better ability to capture CO2.However, over several cycles, the ability to capture CO2 for all conditions tended to similar values and the difference between the best and the worst condition tested decreased.Therefore, it appears that the calcination temperature and time influence CO2 uptake in the initial cycles, but as the number of cycles increases, the effect becomes less significant.The loss of CO2 capture capacity can probably be attributed to the sintering, which is faster for sorbents calcined at higher temperatures.Although there is a loss in the CO2 capture capacity of CaO over the cycles for all samples, which is common for this kind of material, the waste seems to be suitable for CO2 capture.

Figure 1 :
Figure 1: Schematic drawing of the experimental system used for the calcination-hydration-carbonation reaction cycles.

Figure 3 :
Figure 3: Diffractograms of the carbonation product after the first and second cycles.C -Calcite (JCPDS 01-072-1651).Calcination carried out at 950 °C for 45 min, hydration at 500 °C for 30 min and carbonation at 650 °C for 30 min.

Figure 5 :
Figure 5: CO2 Capture Capacity vs. the number of cycles under different calcination conditions.

Figure 8 :
Figure 8: Sorbent reactivity vs. the number of cycles for different calcination conditions.