Control of Carbon Dioxide Sequestration by Mechanical Activation of Red Mud

Mineral carbonation is a potentially attractive sequestration technology for the permanent safe disposal and immobilization of CO2. In this technology, CO2 is chemically reacted with calcium, sodium, and magnesium containing materials to form thermodynamically stable and environmentally harmless minerals, usually carbonates. In our research, mechanical activation of red mud was carried out in order to enhance its reactivity by means of mechanochemical reactions (surface activation), and its sequestration behaviour was investigated using carbon dioxide gas at 25 °C temperature and at high pressure (5 bar) in an autoclave. The reacted red mud was characterized by Fourier-transformed infrared spectrometer, scanning electron microscopy, X-ray diffraction, laser particle size analyzer, BET specific surface area measurement, and pH measurement. It was found that mechanical activation improved the CO2 sequestration ability by 1.7 wt% of red mud, as demonstrated by the above investigations. The pH of red mud slurry can be lowered by reacting it with carbon dioxide. During our measurements, the pH of the suspension decreased from 10 to 6.81. Furthermore, the carbonation process can be successfully used to decrease the amount of harmful PM10 (particles with a diameter of 10 μm or less) and PM2.5 (particles with a diameter of 2.5 μm or less) fraction. The proportion of 10 μm particles can be reduced by 40% and that of 2.5 μm by 20%.


Introduction
Increasing CO 2 emission which accelerate global warming, is one of the major global environmental issues. In order to reduce CO 2 footprint there are several techniques developed for its sequestration and capturing. Geological sequestration, disposal into the ocean, and biological fixation are widely investigated. Mineral carbonation is one of the most promising method, with ever increasing experiments and applications of CO 2 fixation in basalts or waste materials. Mineral carbonation is a reaction of CO 2 with non-carbonate minerals: or wastes to form geologically and thermodynamically stable forms of carbonates [1][2][3] of the natural rocks, olivine-rich mafic and ultramafic rocks can be used for CO 2 sequestration [4].
Nowadays, the most commonly used materials for CO 2 sequestration are soils [5], steel slags [6,7], fly ash [8], amine solutions, zeolites, porous membranes, serpentine, olivine [9][10][11], cementitious materials [12], construction and demolition waste [13,14], metal-organic frameworks, deep geological strata [15] and red mud [3,16]. An optimal field among the possible techniques is the application of industrial wastes as sequestration media. Red mud as the waste of alumina production from bauxite, can be used as a material to capture and storage of anthropogenic CO 2 , mainly as near-source capture for fossil fuel power plants [17], cement factories, foundries, and metal-working plants.
The disposal of such a large quantity of alkaline sludge is expensive due also to its hazardous nature posing several environmental risks: the high pH (11)(12)(13), high Na concentration (~ 50 g/kg), and very fine particle size (90% under then 10 µm). The elevated pH makes red mud a caustic, highly corrosive material, possible source of severe accidents. Pond-base exfiltration into the basement may lead to Na-contamination of groundwater resources and soils. The fine grain size leads to atmospheric contamination in the PM10 value by dusting on the drying surfaces. Related to physical properties arises also the major barrier in the reutilization of red mud: the high bound water content which rises the costs if it has to be evaporated prior to beneficiation.

3
The carbonic acid formed in the CO 2 bearing fluid neutralizes the bases and precipitates as sodium, calcium, and magnesium carbonates. For the aqueous carbonation process, irreversible hydration of calcium oxide is followed by simultaneous dissolution of Ca(OH) 2 and dissociation of aqueous CO 2 and precedes the carbonation reaction. As the Ca 2+ ions are converted to CaCO 3 and precipitated, more Ca(OH) 2 dissolves to equalize the Ca 2+ concentration. In the aqueous carbonation of silicate bound Ca and Na, the first step is CO 2 dissolution in the aqueous phase producing H + ions. In the resulting slightly acidic environment Ca (Mg) and Na leaches from the silicate matrix, and the system proceeds to carbonate precipitation. The rate and extent of cation leaching were found to be inversely related to particle size and pH and increased with higher temperature, pressure, and specific surface area. The extent and rate of carbonation depends on the diffusivity and reactivity of CO 2 controlled by porosity and physical conditions (partial pressure, relative humidity, temperature, and pressure). According to Huntzinger et al. [61], Huijgen et al. [62], Uibu et al. [63], and Kuusik et al. [64], the main mechanism affecting the rate and extent of carbonation are transportation-controlled mechanisms such as CO 2 and Ca 2+ ions diffusion to/from reaction sites, boundary layer effects (diffusion across precipitate coatings on particle surface) dissolution of Ca(OH) 2 at the particle surface and pore blockage/coating. The main products of slurry forced carbonation are Na 2 CO 3 and NaHCO 3 . Soluble Na-carbonates provide less permanent CO 2 storage than solid Ca-Mg-carbonates because of their solubility. As dissolved NaOH is enriched in surface layers of the ponds [65] precipitation of Na-carbonates through drying is favored, but rehydration will remove these solid phases.
Several studies proved that mechanical activation of various minerals and secondary raw materials improved the CO 2 sequestration capacity [66,66,67,67,68] by increasing material fineness and generating amorphous layers on grain surfaces, crystallite size reduction, and micro-strain accumulation resulted in enhanced reactivity of solids. One of the major components of red mud is cancrinite [69], a tectosilicate which contains additional anions, most frequently CO 3 − , OH − , Cl − and SO 4 [70]. If ratio of OH − increases precipitation of hydroxyl cancrinite is possible. Since cancrinite structure has anion exchange capacity and is able to crystallize in low temperature hydrothermal conditions in highly alkaline environment, it could be expected to be involved in carbonation reactions.
(4) 3Ca(OH) 2 ⋅ 2Al(OH) 3 (s) + 3CO 2 (aq) ↔ 3CaCO 3 (s) + 2Al(OH) 3 (s) + 3H 2 O (5) Na 6 AlSiO 4 6 ⋅ 2NaOH + 2CO 2 (aq) ↔ Na 6 AlSiO 4 6 + 2NaHCO 3 Based on the above introduction experiments were conducted on different types of waste materials and commercially available primary materials for CO 2 sequestration. However, till date no such literature is available on the utilization of mechanically activated red mud for CO 2 sequestration. Grinding work was against the binding energy of crystalline materials. The internal energy of a unit mass dispersed system is the sum of the binding energy and the surface energy of the system. Grinding as a mechanical activation can be characterized by a change in the binding energy [71]. The hypothesis was to enhance the reactivity (carbonation) of red mud prior to CO 2 sequestration by surface activation. Therefore, our research aim was to investigate the effect of mechanical activation of red mud on the CO 2 sequestration behaviour and find correlation between the grinding parameters and the microstructural properties of the resulted carbonate bearing material.

Materials
Red mud used in this study was acquisitioned from Almásfüzitő red mud repository. Initial moisture content of 38.16 wt% and particle density of 2.69 g/cm 3 are specific. Grain size distribution analysis shows 0.4-13 µm with the median size d 50 = 2.37 µm and 31,400 cm 2 /cm 3 geometric and 28.44 m 2 /g BET actual specific surface area. Hematite (21%), cancrinite (17.7%), and goethite (28.5%) are the major phases identified in the red mud by XRD measurement. Several accessory minerals were also identified (gibbsite, calcite, hydrogarnets), and the amount of amorphous phase was 13.4 wt%. For the sequestration tests research grade (99.5%) Biogon C CO 2 gas was used without further purification.
Sample notation scheme was set up as "AV-I.5.60" in which "AV_I" means Almásfüzitő red mud, "5" is grinding time in minutes, and "60" is the carbonation reaction time in minutes.

Mechanical Activation
The raw red mud was dried in a laboratory oven at 105 °C till constant weight. Mechanically activation was done by grinding in a Retsch planetary mill for 5, 15, 30, 60 min. Each milling stage was produced on a new batch of raw red mud. The mill speed was set to 128 rpm, and the grinding ball diameter was 40 mm. The ground material was later investigated with laser particle size analyzer and FTIR measurements.

CO 2 Sequestration Experiment
The mineral carbonation reactions were performed in a Parr high-pressure-and temperature-controlled autoclave that with the active volume of 0.3 dm 3 . Suspension was prepared by mixing red mud with distilled water at 0.3 L/S ratio (3.33 g/L) to obtain a sufficient volume of slurry for carbonation reaction and the back-pressure regulator controlled the system pressure.
Experiments were carried out at 5 bar pressure with continuous mixing. To ensure the sufficient amount of CO 2 , the red mud slurry was set to fill up 70% volume of the reaction vessel. The applied process parameters were 25 °C temperature, and 1, 2, 3, and 4 h reaction times, with 245 1/min revolution number of stirrer.
Pressure was applied to increase CO 2 dissolution in the fluid expecting enhanced carbonation reactions. Slight decrease of pressure with time was observed, due to the consumption of CO 2 by carbonation. In order to ensure the CO 2 necessary for the reaction we have refilled the amount of gas required to achieve the initial pressure every 15 min. The variation of pH was measured before and after reaction with a digital pH meter. The high pH before reaction is caused by the NaOH in the Bayer-process liquor. The carbonated samples weres dried in a laboratory oven to carry out analytical measurements (FTIR, SEM, XRD). The particle size distribution of the reacted samples was measured after reaction immediately in wet state by laser particle size analyzer.

Particle Size Analysis
The particle size distribution of the raw and ground red mud was measured by HORIBA LA-950V2 laser particle size analyzer in wet mode by using distilled water as dispersing media. This measurement was carried out after the carbonation reaction also. For better dispersion of fine particles, before the measurement, sodium pyrophosphate was added, and it was measured under ultrasonic stirring for one minute. Particle size distribution was calculated based on the Mie-theory with the red mud refractive index.

Mineralogical Characterization by X-ray Diffraction
The mineralogy of the red mud sample before and after CO 2 sequestration was determined by X-ray powder diffraction (XRD) with a Bruker D8 Advance diffractometer using Cu K-alpha radiation (40 kV, 40 mA) in parallel beam geometry obtained with Göbel mirror, in the 2°-70° (2θ) range with a 0.007° (2θ) step interval and a 24 s step counting time. The crystalline phases were identified by Bruker DiffracPlus software package in its EVA module ICDD PDF-2 (2005) database for search/matching of phases. The quantitative evaluation was carried out by Rietveld refinement in TOPAS4 software, where the amorphous content was determined by the amorphous hump method.

Fourier Transformed Infrared Spectroscopy
Stretching and bending vibrations of chemical bonds in the samples before and after CO 2 sequestration was detected by a JASCO FT-IR 4200 type Fourier Transformed Infrared Spectrometer in reflection mode with a diamond ATR. The infrared transmission spectra of samples were recorded for 400-4000 cm −1 with the resolution of 4 cm −1 . The observed spectra are plotted in the function of wavenumber and transmittance. The interpretation of observed FTIR bands was done based on the handbook by Chukanov and Chervonny [72].

Scanning Electron Microscopy
Scanning electron microscopy (SEM) was used to take images of the samples for visual observation of particle size, surface morphology, and formation of new phases. The back-scattered images (BSE) were taken with a Phenom ProX Desktop SEM instrument with ×6900 magnification under 15 kV voltage. Additionally, energy dispersive X-ray spectrometry (EDX) was used to study the chemical composition of the samples.

BET Specific Surface Area Measurement
For the measurement of the total specific surface area of samples a Micromeritics Gemini VII apparatus was used. Each sample was degassed before the measurement. The gas used for adsorption was nitrogen. The specific surface area was calculated by a 5-point BET method.

Particle Size Distribution and Specific Surface Area
The aim of mechanical activation by grinding was to increase and activate the particle surfaces on which the carbonation reactions take place. The increased SSA and amorphization of particle surfaces is beneficial to enhance carbonation reactions. Based on the particle size distribution (Fig. 1a, b) the amount of the coarse fraction changed slightly in each case except for 60 min grinding where a significant increase can be observed. The sample milled for 5 min had the finest particle sizes with narrow distribution. On the frequency distribution curve (Fig. 1b) the amount of fraction between 1.7 and 4.5 µm started to increase after 5 min grinding, then further grinding shifted the distribution to higher size fraction with parallel decrease of specific surface area due to the aggregation of particles. The results show that the highest BET specific surface area can be reached by grinding after 15 min, the specific surface area of the initial raw material increased from 2828.44 up to 2929.96 m 2 /g (Fig. 1c). The grinding for 55, 30, and 60 min decreased the specific surface area down to 28.42 m 2 /g 26, 26.45 m 2 /g, and 2525.75 m 2 /g respectively due to aggregation/ agglomeration. These results are not quite well correlated with the particle size distribution of the ground red mud. The particle size decreases in the first 15 min, the geometric surface area and particle sizes indicate the tendency of the powder to form agglomerates due to prolonged milling. However, these agglomerates contain meso-and macro pores, which are of great importance in surface reactions. The pores in relatively weaker (soft) agglomerates remain accessible for nitrogen, thus BET results describe more closely the actual SSA.

Structural Characterization of the Red Mud by FTIR
The FTIR spectra of the mechanically activated red mud samples (by grinding) is shown in Fig. 2, showing that band positions and symmetry of the raw and ground material were not changed. However, differences were observed in their intensities, the decreasing transmittance values indicating the aggregation of fine particles, the same process being observed by PSA (Fig. 1a, b). The H 2 O and OH stretching region of 3300-3000 cm −1 shows the presence of OH (in silicates and goethite) and strongly bound water in mesopores [73]. The peak at 2360 cm −1 could be attributed to the H-O-H group of weakly bound water [74], which disappears upon drying after CO 2 sequestration. By milling, this absorption region and the 1640 cm −1 is decreasing in intensity, supporting the presence of water which was not removed by 105 °C drying and is mostly contained in goethite [75]. However, H 2 O in cancrinite is also producing bands in the 1630-1650 cm −1 region [76].
Part of the strongly bond water and also OH groups is supposed to be contained in the amorphous phase. The presence of carbonate groups is confirmed by strong asymmetric C-O stretching vibration in the 1420-1450 cm −1 region, with double maximum centered at 1435 cm −1 [74].

Particle Size Distribution
Particle size measurements were carried out to observe how particle sizes from different grinding times varies with the same reaction time, depending on the SSA. The variation in particle size is influenced mainly by two circumstances: (1) coarsening due to cementation of particles by carbonate coating, (2) particle decrease by digestion and dissolution upon longer reaction times. Digestion is a part of the process which produces new reaction surfaces. Therefore, changes in particle size distribution are one of the indicators of surface reactions.
During the first 1 h of reaction time the particle size distribution curve shows an increase which may be the result of carbonate cementation of particles (Fig. 3a). This carbonate layer seams to cover the particles equally as proved by the shift of the curve to coarser sizes.
In the second experimental set with 2 h of reaction time (Fig. 3b) the particle size increased depending on the grinding time. The highest coarsening was reached with 15 min grinding. This may be the result of micro cracks (highest BET specific surface area) of the particles due to grinding, in which the dissolved CO 2 can react with the silicates present in the red mud. The particle size of the 30 min milled material did not change significantly that may be the result of the clogging of the previously formed cracks that reduced the reaction surface. In the experiment with 3 h reaction time no trend was found in the variation of the particle size distribution (Fig. 3c) due to the multiple variation in dissolution-reprecipitation of formed carbonate bearing materials. The variation of median particle size (d 50 ) at various grinding times is also responding to the CO 2 fixation. It is noted that the most significant difference was found at 60 min milled sample, where d 50 increased from 3.33 to 7.34 µm (120 min reaction time) and decreased again to 4.01 µm (180 min reaction time). The lowest variation in particle size was found at the 30 min mechanically activated red mud after sequestration which has one of the lowest BET surface area and second coarsest median particle size. The maximum value of the median particle size was reached at 120 min residence time of CO 2 sequestration, except of raw red mud where continuous increase was detected.
Concentration of particulate matter (PM) in air is an important issue from environmental point of view in Europe as well as worldwide (Air quality in Europe, 2017). Two important factors are the PM 10 and PM 2.5 which means particulate matter below 10 µm and 2.5 µm particle size respectively. Since red mud is a very fine material its dusting is a risky characteristic, therefore it is important to see how the CO 2 sequestration change the particle size of red mud particles (Fig. 4).
The initial material was nearly 100% below 10 µm particle size. After 60 min. CO 2 sequestration residence time this size fraction decreased to 85%. In the first 60 min. the AV-I. 5. does not contain the size fraction above 10 µm that may be the result of large SSA that was covered equally with the thin carbonated layer. The agglomerated particles with coarser size produce bigger particles after carbonation as proved the AV-I.60 sample. The 60 min reaction time is further and significantly reducing the finer then 10 µm fraction, which is the result of enhanced carbonate formation. This was the reaction time where the highest CO 2 sequestration values were obtained from this point of view. In the 180 min residence time there is slight decrease in particle size, which is probably due to the dissolution of carbonated particles. For PM 2,5 size fraction (< 2.5 µm) for the first 60 min, the curves are like that of the PM 10 size fraction (< 10 µm) except for the initial deviation from grinding. The 120 min carbonation residence time reduced most significantly the proportion of particles below 2.5 µm, but the material of the AV-I.30 sample has already begun to dissolve.  Based on the above results, precipitation from the solution during the first two hours was dominated, followed by dissolution resulting in the reduction of particle size.

Structural Characterization of the CO 2 Sequestrated Material by X-Ray Diffraction
Changes in the RM composition ( Fig. 5 and Table 1) the main components of red mud are hematite, gibbsite, and cancrinite. However, the sample does not contain detectable amounts of titanium minerals (anatase, rutile), which several researchers have described [26,78]. This is presumably due to how the mud is classified by density after it is placed in storage. Due to its density, the titanium phase is placed at the bottom of the pitfall [79]. An increase of amorphous content is the result of the mechanical activation by grinding, which is related mostly to the hematite content decreasing continuously with grinding time. It is important to note that slight change in amorphous content might be the indicator of surface activation which is one of the states of mechanochemical reactions [80]. Regarding the CO 2 reacted samples the peaks of cancrinite developed an asymmetry (shoulder) indicating a change in cancrinite composition. This observation is in contrast with Yadav et al. [3] who observed that cancrinite decomposes into nepheline, calcite and water under the action of CO 2 reaction. However, their experiments did not involve mechanical activation, only chemical reactions. The composition of the applied RM was also different, containing chantalite, CaAl 2 SiO 4 (OH) 4 . Since the chantalite major peaks in XRD are overlapping with goethite (~ 21°2Th), calcite (~ 29.5°2Th) and hematite (49.5°2Th) as comparing with our results, we cannot identify chantalite in our samples, nor decide if identification by [3] was done correctly. On the other hand, minerals identified in our samples are well in the type and range for most common red mud components, according to Reddy et al. [81].
In order to solve the measured patterns, a cancrinite structure related to carbonate bearing variety was introduced in the Rietveld refinement, applying C to OH substitution and fitting the unit cell parameters. In this way, an increasing amount of carbonated cancrinite was detected with increasing reaction time. Hibschite, a hydroxyl bearing garnet, was also decreasing in the reacted samples, due to the reaction with CO 2 and Na from amorphous material, and probably resulting in carbonated cancrinite also. An increase in gibbsite concentration can be observed similarly to the work of Ramesh et al. (2010). This increase in concentration can be related to the carbonate conversion of hibschite. This transformation is described by [82] with another member of the hydrogrossular group with the katoite. It improved the dissolution of hibschite with mechanical activation, which is consistent with that described by [83].
The sequestration of CO 2 in solid phase must lead to increase in weight of the samples. On the other hand, CO 2 will be liberated upon high temperature calcination from any mineral structure. Based on these facts, in order to control the CO 2 capture of samples, we conducted loss on ignition (L.O.I.) tests at 950 °C, with 15.8 °C/min heating rate and 90 min residence time.
The recorded L.O.I. values (Fig. 6b) correlated well with the CO 2 capture and subsequent weight loss, except for the material milled to 60 min, in this case it increased by 1.6 wt% after a reaction time of 60 min, then decreased to 1.4 wt% after 120 min and reacted to − 3.2 wt% after 180 min. A similar dropping trend is observed for the 15 and 30 min milled material at 180 min reaction time, indicating that agglomeration of particles starts at 15 min grinding and is inhibiting CO 2 sequestration. Basically, it can be established that increased L.O.I. value compared to the initial data due to the CO 2 sequestration on mechanically activated red mud proved the carbonation reactions. The best results were Table 1 The results of the quantitative evaluation (wt%, error ± 1 relative%) of XRD on raw red mud and after mechanical activation and after carbon dioxide sequestration Chemical composition of the phases is shown in Fig. 5 Phase name Grinding-time (min)-reaction time (min) obtained with a grinding and reaction time of 60 min, in which case the sample contained 1.6 wt% carbonate. In this way, 16 g/kg of CO 2 can be isolated, which is 12 g more than in the case of ophiolite [84] and 6 g more then mechanicated, and 6 g more than can be sequestration on mechanical activated olivine-basalt [4].

Structural Characterization by FTIR Spectroscopy
The FTIR investigation of the carbonated red mud samples helps to understand their CO 2 sequestration ability. The positions of the absorption bands are nearly similar. The changes in transmittance of CO 3 2− group bands can be observed in the carbonated sample. Figure 7a-c show the FT-IR spectra of raw red mud and neutralized red mud (NRM). The 875 cm −1 band belongs to calcite, while the 833 cm −1 band of hibschite is shifted to ~ 810 cm −1 , due to Ca-Fe substitution. The strong bands in the 900-1200 cm −1 region belong mainly to the cancrinite structure, especially important is the 1100 cm −1 band, showing the presence of OH − and CO 3 −2 cancrinite structure types. The change in intensity and position of this band supports the observations by XRD, the development of carbonate bearing cancrinite. The band at 2160 cm −1 confirmed that the fresh red mud captured CO 2 from the atmosphere, however through molecular absorption and not carbonation. The 1420-1450 cm −1 bands indicate the stretching vibrations of C=O, show the presence of carbonate groups, 1427 cm −1 for calcite and 1400 for carbonate cancrinite. This is due to chemisorbed CO 2 and carbonate formation in fresh red mud, and since calcite amount is relatively stable according to the XRD, the variation of band position and intensity of this region is also related to cancrinite carbonation. Chukanov et al. [76] reported overlapping bands at 1396, 1444 and 1489 cm −1 for a hydroxycancrinite-cancrinite transitional mineral phase. Values of 1381, 1397, 1434, 1480 and 1512 cm −1 are reported by Ventura et al. [85]. Combined with the data from Chukanov and Chervonnyi [72] one observes that carbonate vibrations in cancrinite structure are complexe and unique for each sample, although a common region can be delimited.
After 60 min reaction time (Fig. 7a) the mechanically activated samples show slightly more intense carbonate bands at ~ 1435 cm −1 indicating carbonate fixation. The 5 min milled sample shows the most intense band, which may be the result of higher specific surface area in which the carbonate layer formed. A further slight decrease occurs as a function of increased grinding time.
The 120 min CO 2 residence time decreased the transmittance of AV-I.5 and AV-I.30 sample indicating decrease in carbonation (Fig. 7b) in accordance with the previously presented particle size analyses. The dissolution begins after this 120 min reaction time.
The 180 min reaction time (Fig. 7c) shown the most intense carbonate peak in the case of 15 min grinding which is well correlated with the L.O.I. values (Fig. 6b). The 5 and 30 min. milled samples had similar transmittance supporting the role of particle size distribution, in carbonation, through available activated surface. The lowest amount of carbonates remained in the samples ground for 60 min and in the raw red mud sample. Based on the measurements, according to the carbonate band characteristics, the 15 min mechanically activated samples gave the best results, which required a 180 min reaction time.
Based on the measurements, it was observed that the samples reached a maximum point in the amount of captured CO 2 followed by a decrease instead of stabilization which shown fluctuation. The evolution of OH and H 2 O content of red mud phases is also responding to mechanical activation and carbonation, but due to high goethite amount the fine details related to hibschite and cancrinite phases remain hindered. However, in the case of 180 min residence time, for the AV-I-60-180 sample there is a significant overall drop in transmittance, which is more significant for goethite water and Fe-O-Fe bands, as compared to the raw sample. The phenomenon is linked to the strong agglomeration by milling, but also amorphization process observed for this sample by XRD.

pH Measurements
The red mud suspension rapidly reacts with CO 2 in the form of carbonic acid and neutralizes the excess base in the form of NaOH, Na 2 CO 3 , Al(OH) 4 . The red mud suspension rapidly reacts with CO 2 in the form of carbonic acid and neutralizes the excess base in the form of NaOH, Na 2 CO 3 , Al(OH) 4 . The pH of the suspension decreased from 10 down to 6.81 the initial alkaline to a slightly acidic nature due to the weak acid in solution dissolved in the water to be absorbed, which results in the presence of the red mud coated with carbonates began to dissolve. At the end of cycle which indicated the changes of the high pH towards slightly acidic (Fig. 8). The equilibrium pH was achieved between 6.1 and 7.1 due to HCO 3 − ions. This may be the result of the solved CO 2 in water that caused the solution of some minerals of the red mud.
When the pH decreased there were some Na bearing minerals which dissolved easily such as cancrinite. The highest decrease of pH was experienced in the first reaction hour, when there was much free reaction partner for CO 2 sequestration. The pH reached the value of 6.2 meaning that it became slightly acidic. In this way it is expected that CO 2 sequestration using red mud will solve the environmental problem of red mud storage (high alkalinity). This was a result similar to that of Changming et al. [86], obtained in their measurements on red mud. Figure 9a and b shows the morphological structure of red mud before and after carbonation. SEM images of the microstructure of red mud show the aggregates of fine particles about 3…5 µm size which is in correspondence with the laser particle size analysis results. It can be observed that, the original red mud was fine grained, so the individual particles cannot be seen, only in aggregates due to the drying after wet sequestration. The images indicate that significant difference in the grain size or morphology cannot be observed after CO 2 sequestration, that may be the result of surface encrusting. The EDX analysis of the red mud sample shows the presence of particular elements as shown in  Fig. 9a (39.89 and 41.79 wt%), which means increasing oxygen content after CO 2 sequestration. Additionally, carbon content increased as well after red mud-CO 2 reactions indicating carbonate formation which is in correspondence with the results of X-ray diffraction and FTIR measurements.

Conclusions
Previously, researchers' work has provided studies on the mechanical activation of red mud and CO 2 sequestration of red mud separately. However, the two synergetic process together have not yet been studied to a sufficient extent. The following conclusions can be drawn based on the results of CO 2 sequestration behavior of mechanically activated red mud: -Mechanical activation can influence the CO 2 sequestration ability of red mud, but process parameters have to be optimized. The highest amount of CO 2 captured on the 15 min mechanically activated red mud sample with 180 min reaction time. -Cancrinite captured CO 2 in its crystal structure as carbonate which is the main mineral phase responsible for  carbonation of red mud proved by XRD. It can be stated that in the first 120 min the precipitation was the main process, which turned into dissolution after 180 min reaction time. -The amount of captured CO 2 has a maximum point on the FTIR curves. The intensities of the carbonate band on the FTIR spectra varied due to the process of precipitation and dissolution. -SEM EDX results indicated increased oxygen and carbon content after CO 2 sequestration. -The pH decreased from 10 to 6 at the end of neutralization. Utilization of red mud for sequestration of CO 2 and its neutralization will reduce its environmental problem and storage area. -The carbon-dioxide capture has multiple benefits. The residual sodium reacts with carbon dioxide while reduces the pH of the red mud. The neutralized reaction product is environmentally friendly, it is suitable for construction materials, beside the positive effect of coarsening of the carbonated particles reducing PM 10 and PM 2.5 .
Funding Open access funding provided by University of Miskolc.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.