Influences of CO2 Bubbling Types on Preparation of Calcite Nanoparticles by Carbonation Process

This study investigates the comparison of influences of CO2 bubbling into the calcium hydroxide (Ca(OH)2) slurry through a microbubble generator (MBG) and an ordinary CO2 generator (OCG) on the preparation of calcite nanoparticles by a carbonation method. Each obtained precipitate was characterized using XRD, SEM and particle size analyses. During the carbonation process at each CO2 flow rates, it was determined that the MBG generates tiny bubbles whereas an increase in CO2 flow rates led to an increase bubble size when the OCG was used. The flow rate of CO2 was not an important parameter with using the MBG as calcite nanoparticles were prepared (<125 nm) at each CO2 flow rates. The necessary time for the complete reaction decreases with an increase in the CO2 flow rates through the MBG in comparison to the OCG. To produce calcite nanoparticles with a high production recovery in shorter times, the MBG should be adopted to the carbonation reactor.


Introduction
Calcium carbonate (CaCO 3 ) is one of the most abundant materials and widely used as a filling material in various industries (paper, paint, plastic, rubber and so on) depending on its purity, opacity, whiteness degree, particle shape, rheology, specific surface area, particle size distribution, and water or oil absorption properties.However, the use of natural CaCO 3 particles is limited in the industrial applications as the controlling of those properties was difficult.Recrystallization processes (biomimetic method and carbonation method) have been conducted to prepare synthetic CaCO 3 particles with desired qualities [1].
The carbonation method is favorable in the industry because of the availability of raw material and low cost in comparison to the biomimetic method.CaCO 3 particles are prepared by absorbing CO 2 into a Ca(OH) 2 solution.However, the production recovery is not high yield due to low absorption of CO 2 .The solubility of CO 2 is increased with the use of microbubble generator (MBG) that produces innumerable bubbles in micrometer-sizes [2].
The adaptation of those apparatus to the carbonation reactor leads to produce calcite nanoparticles with uniform morphology.The particle size of PCC was found to be 50 -100 nm using the CO 2 bubbling orificer, whereas that of PCC in the absence of orificer was obtained in micron sizes [20].It is obvious that MBG should be conducted to obtain calcite nanoparticles [5][6][7].In addition, the particle size of PCC increased within increasing CO 2 flow rates.This paper reports the determination of influences of CO 2 flow rates using the MBG apparatus on the production of calcite nanoparticles by a carbonation method.The morphology, particle size and specific surface area of each precipitate were evaluated by SEM, XRD and particle size analyses.The slurry pH was measured as an indicator for the complete reaction.Control experiments at the same conditions were performed using the OCG for the comparison.

Experimental 2.1 Preparation of calcite nanoparticles
In this study, 250 mL of 0.4 M Ca(OH) 2 slurry prepared with an analytical grade Ca(OH) 2 (Merck, purity ≥ 99%) and pure water was used for each experiment.Plexiglas reactor designed for producing calcite nanoparticles is shown in Fig. 1.CO 2 (purity: 99.9%, supplied from Cangas, Turkey) with various flow rates (0.3, 1, 3, 5, 8 and 10 L/min) was bubbled into the Ca(OH) 2 slurry through the MBG that was placed at the bottom of the reactor.The MBG has a porous structure and its length and diameter was 3 and 1 cm, respectively.With the use of this apparatus, tiny bubbles having same sizes are produced at each CO 2 flow rates.The dispersion of bubbles was better provided compared to the OCG.The duration time of each bubble generated through the MBG in the slurry was higher.The bubble size was not influenced by the CO 2 flow rates.However, the increase of CO 2 flow rates led to an increase the bubble size when the OCG was used.
A mechanical stirrer stirred the solution at 650 rpm during the carbonation.The reaction temperature was kept 293 K by a water batch.The carbonation time was variable depending on the CO 2 flow rates.The solution pH was continuously measured by pH meter (WTW 3110) during the carbonation reaction.When the solution pH was 7.30±0.10,the CO 2 flow rate was switched off and the reaction was complete.Non-reacted CO 2 was accumulated above the acidified water in a lab scale gasometer for the re-use.After conducting each carbonation, the suspension was filtered using filter paper (Whatman, No.1).
The precipitate was dried at 105 o C for 3 h in an oven in preparation for the characterization test.X-ray diffractometer (XRD, Rigaku) equipped with Cu Kα radiation with the 2θ range of 15-85 o , with a step size of 0.02 was performed.The diffraction data was identified using a PDXL software.The crystallographic properties of PCC were using a scanning electron microscope (SEM, Philips XL 30S FEG and Zeiss EVO LS-10).The average particle size of each precipitate was examined using a Malvern Master sizer (Hydro 2000 MU).
The reactor was cleaned by HCl (5%) solution to remove carbonate particles sticking on the reactor wall and the MBG or OCG, and then rinsed with water.Each experiment was repeated twice.Furthermore, control experiments were performed with an ordinary CO 2 bubbles generator (OCG) for the comparison.

Results and Discussion
The pH value was an important indicator in the carbonation test.The decrease of pH indicates that OH -ions are consumed and each Ca 2+ ion is carbonated due to adding CO 2 gas.However, if the pH was ~ 9, it is evident that the carbonation did not complete and the desired product is prepared because of the presence of OH -ions in the slurry [21].Precipitated calcium carbonate (PCC) production mechanism can be explained in the following equation ( 1) -( 5

→
Thereafter the reaction steps given below occurred, the PCC particles were formed in various morphological phases depending on the reaction conditions.It is thought that the dissolution of PCC particles is trigged with an decrease in the slurry pH.However, Han et al. [23] revealed that the PCC particles dispersed in the carbonate solution were not dissolved due to the high HCO 3 -concentration, when the slurry pH was about 7.50.Therefore, the carbonation reaction was terminated when the slurry pH decreased to the desired value.
The PCC particles are quite stable in the carbonate solution [23].The behavior of solution pH during carbonation was explained by many researchers in the previous studies [25][26][27].Wei et al. [25] explained that the behavior of slurry pH during the carbonation process divides into the two stages which are as follows: (1) constant rate of absorption stage and (2) falling rate stage.
In the first stage, more than 75% of the Ca(OH) 2 is converted to PCC particles with the injection of CO 2 into the slurry and the formation of PCC particles is completed in the falling rate stage.However, Bang et al. [26] identified the hatched region area (HR) on the pH graph where steep changes in pH occurred with an injection of CO 2 .They claimed that Ca 2+ ion was formed as PCC particles before the HR region.The slurry pH (5) did not change during the carbonation stage as the OH -ion was supplied by the dissolution of Ca(OH) 2 in the slurry until none remains [27].The decreasing trends of the slurry pH during the carbonation were similar with those of previous studies.In this paper, the behavior of the slurry pH was evaluated based on the approach of Bang et al. [26].The changes of pH in the slurry during the carbonation using the OCG and MBG are shown in Fig. 2 and Fig. 3, respectively.The absorption ratio of CO 2 was lower than that of the dissolution rate of Ca(OH) 2 as the slurry pH did not decrease at the beginning stage of each carbonation reaction as shown in Fig. 2 and Fig. 3. Thereafter, the slurry pH decreased rapidly in time depending on the CO 2 flow rates.This indicates that the Ca(OH) 2 has been totally carbonated.
The HR area in the Fig. 3 was observed in shorter times within increasing CO 2 flow rate and the use of MBG comparing to the use of OCG.When the reaction was in the HR area, the dissolution rate of CO 2 is larger than its consumption rate due to absence of Ca 2+ ion in the solution and the accumulation of H + ion leads to the decrease of solution that became acidic [6,27].These findings are in line with those of previous studies [23,25,28].
Each prepared precipitate was of calcite phase of CaCO 3 with a high purity according to the XRD examination (not given in this paper).The use of MBG did not influence on the morphology of the product.No peaks representing the aragonite or vaterite phase of CaCO 3 were observed.Fig. 4 shows the carbonation time depending on the CO 2 flow rates through the MBG and OCG.The amount of CO 2 that is necessary for the complete carbonation reaction was the same as the concentration of Ca(OH) 2 was constant for each experiment.At 0.3 L/min of CO 2 flow rate, the longest reaction time was obtained.However, the increase of CO 2 flow rate led to an accelerated carbonation and the growth time of each precipitate decreased.However, the change in the reaction time was minimal at ≥3 L/min of CO 2 flow rate, some of CO 2 bubbled into the solution left the carbonation reactor without reacting Ca 2+ ions.This may be due to the constant absorption ratio of CO 2 at experimental conditions conducted [23,29,30].
When the obtained results obtained from both apparatus were compared each other, the MBG has strong effects on the carbonation reaction [19,25].The precipitate was obtained in shorter times with the use of MBG compared to the OCG.The surface tension of bubbles increases due to generating tiny bubbles through the MBG that remain longer in the liquid phase and contact with the Ca(OH) 2 in extended times [31,32].These properties led to the increase solubility of CO 2 in the slurry, and the reaction became faster [19,33].At 0.3 L/min, the reaction time was found to be 35 and 27 min for the MBG and OCG.The percentage of reacted CO 2 was found to be 27% for the MBG and 21% for the OCG at the fixed CO 2 flow rate of 0.3 L/ min.respectively (Fig. 5).However, larger bubbles having low surface tension were produced and escaped easily from the slurry due to buoyancy when the OCG was used [26,27].This explains the difference of reaction time between both the MBG and OCG.An increase of up to 5 L/min of CO 2 flow rate lead to the decrease of reaction time.However, there was no change determined in carbonation time at higher CO 2 flow rates.The amount of reacted CO 2 decreased and it left from the reactor without reacting with Ca(OH) 2 and entered the gas holder due to the increase of CO 2 flow rates.
Fig. 6 shows the SEM images of each precipitate obtained through the MBG.It is clear that the morphology and particle size of precipitate was not influenced by the CO 2 flow rates since the MBG generates tiny bubbles at each flow rates.The precipitate was mono-dispersed, uniform and cubic shapes (Fig. 6

. (a)-(b)-(c)-(d)-(e)-(f))
. The particle size analyis results were consistent with those of findings observed from the SEM examination.The d 90 of precipitate was found to be 61.43,76.45, 86.50, 97.23, 100.42 and 122.40 nm at 0.3, 1, 3, 5, 8 and 10 L/min, respectively.These findings are consistent with those of study [34].The number of CO 2 bubbles has a no influence on the d 90 value as the size of CO 2 bubbles was almost the same size at each CO 2 flow rate.The particle size is relatively uniform and its distribution is narrow.This was due to the dispersion of CO 2 tiny bubbles that led to produce calcite nanoparticles [35].
However, the controlling of CO 2 bubble size through the OCG was impossible at higher CO 2 flow rates.The increase of CO 2 flow rate affected the morphology of precipitate (Fig. 7).This finding was in good agreement with those of previous study [20].Except for the precipitate prepared at 0.3 L/min (Fig. 7.a), the particles of PCC tend to be agglomerated and poly-dispersed.It was difficult to obtain precipitates of uniform particle size.The smaller crystals appeared on the surface of large crystals of precipitate depending on the CO 2 flow rates as shown in Fig. 7.(b)-(c)-(d)-(e)-(f).Then, the precipiate with large and small rhombohedral shapes was obtained with increasing CO 2 flow rates.This was confirmed by the d 90 of each precipitate prepared using the OCG.The d 90 of precipitate was found to be 102.56nm at 0.3 L/min of CO 2 flow rate, whereas the d 90 of precipitate prepared at 10 L/min of CO 2 flow rate was 2.01 µm.It was thought that this was related to the bubble size of CO 2 , and bigger CO 2 bubbles led to an obtain PCC in micron sizes in this paper.Researchers have suggested that the BET surface area is related to the particles diameter (D), or radius (R) and the skeletal density [28,36].For that reason, it was assumed that all of the particles were of spherical shapes and non-porous and determined the specific surface area of each product using the following equation.The calculated results are given in Table 1.The precipitates with higher surface areas were obtained by means of the MGB.Where, s g is the specific surface area of precipitate (g/m 2 ), p g is the density of CaCO 3 (2.711g/cm 3 ), d is the particle size of precipitate (µm) (6)

Conclusion
In this paper, the carbonation reaction was facilitated when CO 2 gas was bubbled into the slurry through the MBG.t led to an increase not only the percentage of reacted CO 2 but the carbonation recovery.The SEM images clearly show the influences of both apparatus on the morphology and particle size of precipitates.Calcite nanoparticles were prepared at each CO 2 flow rates by means of the MBG.There was no need to use any chemical to produce calcite nanoparticles (below <125 nm) with the narrow size and uniform morphology.However, an increase in the CO 2 flow rate affected the morphology and particle size of each precipitate when using the OCG.If the CO 2 flow rate was fast, controlling the diffusion and nucleation of those particles would be difficult, and small&large crystal particles were prepared.The MBG can overcome the difficulty of those problems.
The reaction was complete in shorter times through the MBG that generates tiny bubbles at higher flow rates.This finding was supported with the percentage of reacted CO 2 .The MBG apparatus has superior advantages over the OCG due to the decrease of reaction time by 20 -45 % depending on the flow rates.

Fig. 2 Fig. 3
Fig. 2 Change in the slurry pH during carbonation using an OCG

Fig. 4
Fig. 4 Change of reaction time depending on CO 2 apparatus

Fig. 5
Fig. 5 Amount of reacted CO 2 at various flow rates

Table 1
Particle size and calculated surface area of each precipitate