Vibrational and First-Principles Analyses of Aragonite, Calcite, and Vaterite CaCO3 Prepared from 13CO2 Isotope Authors: Takayoshi Fujimura, Kunio Matsui, Ippei Maruyama Published: 2025 ABSTRACT To clarify the isotopic fractionation of CaCO3, Ca12CO3 and Ca13CO3 were synthesized from 12CO2 and 13CO2, respectively, and analyzed using infrared absorption spectroscopy. During the synthesis, calcite, aragonite, and vaterite were successfully separated. Many different vaterite structures have been proposed, and the structure of the synthesized vaterite was identified using X-ray diffraction. The infrared absorption data for calcite, aragonite, and vaterite composed of 100% 13C were obtained for the first time, providing precise data on the isotopic shift of the reference vibration of the CO3 2– ion in CaCO3. Vibrational analyses using first-principles calculations reproduced the experimentally measured isotope shifts to a certain degree. The reduced isotopic partition function ratios for each vaterite structure, which were not experimentally determined, were theoretically obtained via vibrational analysis of each vaterite structure. Introduction Global warming has become an overarching problem in recent years, where the main probable cause is the emission of greenhouse gases into the atmosphere. Greenhouse gases include carbon dioxide, methane, nitrous oxide, and chlorofluorocarbons, of which carbon dioxide accounts for more than half. In the construction industry, extensive efforts are being made to reduce CO2 emissions and achieve carbon neutrality. The carbon dioxide emitted during cement production via the calcination of limestone through combustion using fossil fuels accounts for 8% of the carbon dioxide emitted worldwide. , Thus, various approaches have been proposed for reducing the use of conventional Portland cement as a cementitious material. The alternatives include blending industrial byproducts or natural SiO2- and Al2O3-bearing resources into cement, as well as utilizing carbonated concrete waste by converting the hydration products into aluminosilicate gels and calcium carbonates for reuse as cementitious components. − Additionally, strategies for storing CO2 in concrete in the form of inorganic carbonates by incorporating these products into concrete formulations are being explored. − Isotopic tracing is considered a promising analytical method for providing a deeper understanding of the reaction pathways and carbonation mechanisms involving the target CO2 gases. Although several calcium carbonate phases, such as calcite, aragonite, vaterite, amorphous calcium carbonate, and calcium aluminate carbonate hydrates, have been detected during the carbonation of cementitious materials, the isotopic fractionation among these polymorphs remains poorly understood. This study aims to contribute to the mechanistic understanding of carbonation in cementitious systems through investigations employing carbon-13 (13C) isotopic labeling. Three types of carbon isotopes are found in nature. The most common isotope is 12C, which is a stable isotope that accounts for more than 98.9% of the natural isotopes. The next most common isotope is the 13C stable isotope, which accounts for 1.06%. The remaining radioactive 14C isotope accounts for approximately 10–10%. The atomic ratio of the heavy isotopes (13C and 14C) in a material can be characterized by comparison relative to that of a standard material. During the carbonation of concrete, the carbon-isotope ratio of the generated CaCO3 is generally different from that of gas-phase carbon dioxide. CaCO3 exists in polymorphic forms such as calcite, aragonite, and vaterite, where vaterite has various structures. Even if the composition is the same, the isotope ratio differs if the structures are different. The differentiation process is known as isotope fractionation. In a previous study, 14C was used to identify the carbon source to confirm the direct capture of atmospheric CO2 by cement paste. The normal absorption frequency measured by infrared (IR) spectroscopy was used to calculate the reduced isotopic partition function ratio. For calcite and aragonite, the normal frequencies and reduced isotopic partition-function ratios have been determined via both experimental and theoretical calculations. − However, for vaterite, which has many possible structures, there are no reports of the reduced isotopic partition function ratios for the individual structures, and the degree to which the isotope partition coefficients differ for the various vaterite structures has not been established. In addition, it is not known which positions in the vaterite structure are amenable or resistant to substitution with isotopic carbon, and the reduced isotopic partition function ratio of isotopic carbon at these positions has not yet been clarified. Furthermore, the synthesis of calcite, aragonite, or vaterite using gas-phase carbon dioxide composed of 100% 13C carbon and measurement of the vibrational absorption frequency have not been attempted. In this study, we first aim to separately synthesize calcite, aragonite, and vaterite using 100% 13C carbon dioxide to obtain highly accurate infrared absorption data for 13C calcite, aragonite, and vaterite. Powder X-ray diffraction (XRD) is used to determine whether the synthesized product has the desired structure, and the Fourier-transform infrared (FTIR) absorption spectra of the synthesized materials are acquired. In FTIR measurements, characteristic absorption curves that reflect the structure of the materials can be obtained, and the normal vibrational frequencies can be estimated from the absorption peaks. Moreover, because it is impossible to clarify the normal vibrational frequency and reduced isotopic partition function ratio of each individual structure of vaterite because the data cannot be experimentally determined for the variety of structures, we attempted to clarify this through vibrational analysis using first-principles calculations. Experimental Section Experimental Procedure Generally, for the industrial synthesis of calcium carbonate globally, CO2 gas is bubbled into a Ca­(OH)2 slurry. The polymorphs and particle shape of the resulting calcium carbonate strongly depend on the reaction temperature, concentration of the Ca­(OH)2 slurry and CO2 gas, flow rate, and other factors, making these factors difficult to control. In the present study, the Ca­(OH)2 slurry–CO2 gas bubbling method was adopted for the synthesis of all three types of calcium carbonate, with the objective of preparing Ca13CO3 using 13CO2 gas as the carbon source. The apparatus used for the synthesis is shown in Figure . This apparatus was designed with reference − to be suitable for smaller-scale synthesis. Furthermore, to avoid 12CO2 contamination in the air and enhance the utilization efficiency of 13CO2, it was constructed as a fully closed system. A 500 mL polypropylene reaction vessel with a sealed structure was employed, equipped with a gas inlet line fitted with a nozzle at the tip and a gas outlet line for the unreacted gas. The flow rate of the introduced CO2 gas was controlled using a digital flowmeter (ALICAT SCIENTIFIC, M series), and the total amount of CO2 gas was determined. To monitor the reaction progress, a pH sensor and an electrical conductivity sensor (DKK-TOA Corp., MM-43K) were simultaneously inserted into the reaction vessel. For small-scale reactions involving less than 100 mL of slurry, a double-layer configuration was used by placing a smaller polypropylene container inside a 500 mL vessel. Experimental setup. A gas collection bag was attached to the outlet line to capture the unreacted CO2 gas. When the CO2 gas flow rate was set below 50 mL/min, the gas was introduced using a gas bag prefilled with CO2 at atmospheric pressure and a tube pump, prioritizing the accuracy of the gas flow quantification. The reaction progress was monitored by analyzing the pH and electrical conductivity, which are important metrics for tracing the progress of the reaction from the Ca­(OH)2 slurry and CO2 gas because both parameters provide more detailed information than pH only. − This method is considered essential, especially when determining the end-point of the reaction. The electrical conductivity of amorphous calcium carbonate (ACC), which was initially formed, was used for in-depth monitoring of the crystallization of calcite. − Using the following literature conditions as the starting point, several preliminary syntheses were conducted to determine the detailed experimental conditions, such as the starting concentration of the Ca­(OH)2 slurry and CO2 gas flow rate, taking into consideration the reaction efficiency (=(CO2 absorption volume calculated based on 100% carbonation of Ca)/(CO2 gas volume from integrated flow rate)) and the purity of the desired crystal phase, as estimated from XRD. Prior to the experiments using 13CO2, the following experiments were conducted using 12CO2 gas to maximize the reaction efficiency of 13CO2. The selected reaction conditions of 12CO2-gas synthesis experiments were directly adapted to the 13CO2-gas synthesis experiments. The following raw materials were used in the synthesis: Special-grade Ca­(OH)2 (FUJIFILM Wako Pure Chemical Corp.) was used to prepare the Ca­(OH)2 slurry. CaO obtained by calcining special-grade CaCO3 (FUJIFILM Wako Pure Chemical Corp.) at 1000 °C was used for the synthesis of vaterite. The calcination program for CaCO3 using an electric furnace was as follows: The sample was heated from a preheating temperature (40 °C) to 1000 °C over a period of 2 h, held at 1000 °C for 4 h, and then cooled down to 200 °C. Following this, the sample was cooled to ambient temperature in a glass desiccator containing silica-gel and a CO2 adsorbent. XRD measurements conducted prior to the synthesis experiment indicated that the calcined sample contained 0.43 mass% Ca12CO3. The 12C contamination in the vaterite synthesis was considered negligible. 12CO2 gas (99%; Tokyo Koatsu Yamazaki Corp.) and 13CO2 gas (99%; Shoko Science Corp.) were used for the synthesis. Synthesis of Each CaCO3 Polymorph The conditions for the synthesis of each CaCO3 polymorph are listed in Table . The detailed procedure is described below. 1Experimental Conditionscrystalline phasecalcitearagonitevateritestarting slurryCa(OH)2–H2OCa(OH)2–H2OCaO–CH3OH–H2Oliquid weight (g)50 (H2O)50 (H2O)5.3 (H2O) + 50 (CH3OH)CaO/liquid (mass%, as CaO)2.03.02.7 molar ratio = 11additivenoneNaOH Na/Ca molar ratio = 1.86nonereaction temp. (°C)255042CO2 gas rate (mL/min)251010reaction progress 12CO2 13CO2 12CO2 13CO2 12CO2 13CO2 total reaction time (min)57.251.087.375.582.380.5final pH of solution7.107.1011.9011.907.887.41final conductivity (S/m)0.1130.11012.4710.380.01890.0178 Experimental Conditions Calcite CO2 gas was bubbled through the Ca­(OH)2–H2O slurry. After a predetermined time, the slurry was suction-filtered through a filter paper (ADVANTEC, filter paper No. 2). The filter cakes on the filter paper were washed with H2O, followed by washing with acetone, whereby the excess water was replaced with acetone. The filter cakes were washed with H2O on a rote, followed by similar washing with acetone; thus, the excess water was replaced with acetone. Aragonite Aragonite is more easily formed at higher temperatures than calcite. However, the method using NaOH was employed herein, where NaOH was added to a Ca­(OH)2–H2O slurry to obtain a NaOH/CaO molar ratio of 1.86. The filter cake obtained after the reaction was suspended in 300 mL H2O, stirred, and filtered again. This was repeated three times. After the second time, the electrical conductivity of the suspension decreased to 6.1 S/m. The filter cakes were then washed with H2O and acetone on a filter paper. Filtration was performed in the same manner as for calcite, and the final filter cakes were vacuum-dried at ambient temperature to obtain a powder. Vaterite In the Ca­(OH)2–CH3OH–H2O–CO2 reaction system, vaterite of various particle sizes can be synthesized by changing the reaction temperature. , This method was used as the basis for vaterite synthesis. CaO powder (2.7 mass%), prepared by calcining CaCO3 at 1000 °C, was suspended in a slurry (5.3 g H2O + 50 g CH3OH), and the reaction was performed by bubbling CO2 gas into the hydrated Ca­(OH)2–H2O slurry. The H2O/CaO molar ratio was determined as 11. The vaterite particles were so fine that a centrifuge was used for solid–liquid separation. The suspension and centrifugation processes were repeated three times, and acetone was used for washing. Characterization Method XRD measurements were performed using an Aeris device (Malvern Panalytical, Germany) with reflection geometry, a Cu–Kα X-ray source, tube voltage of 40 kV, tube current of 15 mA, scanning range (2θ) of 5–70°, step width of 0.00346°, and single-test duration of 310 min. At the incident beam path, a 0.04 rad Soller slit, a 1/4° divergence slit, and a 13 mm beam mask were used. For the diffracted beam path, a 7 mm antiscatter slit and a 0.04 rad Soller slit were set before the 1D detector (Malvern Panalytical, Germany). Typically, 900 mg of powdered sample was poured into a Ø16 mm sample holder, and back-side loading was then applied to create a flat measuring surface. Rietveld analysis was performed to determine the phase-composition of the synthesized CaCO3 powders using TOPAS software (version 6.0, Bruker, Germany) based on the external standard method. , In this study, the vibrational absorption spectra of the synthesized CaCO3 powder were analyzed. ATR-FTIR spectroscopy was conducted using a Bruker ALPHA II instrument. FTIR measurements were performed as soon as possible on samples that had been vacuum-dried and stored in sealed polypropylene (PP) bottles at 25 °C in air. The spectra were acquired in the wavelength range of 400–4000 cm–1 at a resolution of 2 cm–1, where 128 scans were averaged. Thermogravimetric analysis (TGA) was performed using a NETZSCH STA 449F5 instrument. Approximately 20 mg of each sample was heated from room temperature to 950 °C at a rate of 10 °C/min under N2 flow at a rate of 50 mL/min. Field-emission scanning electron microscopy (FE-SEM) observations were performed with a JEOL JSM-IT800SHL instrument at an acceleration voltage of 5 kV. Before observation, the sample was coated with osmium to enhance its electrical conductivity. Calculation Method Vibrational analysis of calcite, aragonite, and vaterite was performed using first-principles calculations. For vaterite, the eight structures discussed by Demichelis et al. were analyzed. The formation energy and crystal structure of these structures were discussed in that study. The reduced isotopic partition function ratio β, which is an indicator of isotopic fractionation, was calculated from eq using the ratio of the isotopic partition functions of each carbonate according to the simplified model proposed by Bigeleisen and Mayer. β=Q′Q=∏i(exp(−ui′/2)1−exp(−ui′)1−exp(−ui)exp(−ui/2))gi 1 Here, u i = hcω i /kT, u i ′ = hcω i ′/kT, where ω i and ω i ′ are the vibrational frequencies of the materials with 12C and with 13C, respectively. h,c,k,T are Planck’s constant, the speed of light, Boltzmann’s constant, and temperature, respectively. g i is the degeneracy of the corresponding vibrational modes. The initial lattice constants of calcite, aragonite, and vaterite were obtained from previous studies. ,, The initial lattice constants were used for optimization of the lattice constants. The lattice constants were optimized by varying the volume of the unit cell and fitting the total energy obtained by first-principles calculations as a function of the volume to the Murnaghan equation of state. The code in the HiLAPW package was used to calculate the Murnaghan equation of state. In the optimization calculations, the internal coordinates were not relaxed in order to maintain the symmetry of each structure. The equilibrium lattice constants obtained from the optimization calculation were used to create a unit cell for the vibrational analysis. In the vibrational analysis, the forces acting between the atoms were calculated by displacing each atom of one CO3 2– ion in the unit cell. The displacement was 0.05 Bohr radius. Subsequently, force constant and dynamics matrices were created, and the frequency of the normal vibration of the CO3 2– ion was obtained by solving an eigenvalue problem. However, in the vibration analysis, the normal frequency was calculated based on the Γ point. The β value for each site within the structure was determined using eq from the normal frequency of each CO3 –2 site. Sites with larger β values are more likely to contain 13C isotope. The first-principles calculation program PHASE/0 was used, which implements the plane-wave pseudopotential method based on density functional theory (DFT) within the Perdew–Burke–Ernzerhof-type generalized gradient approximation (GGA–PBE). − An ultrasoft pseudopotential was used. The cutoff energy was 25 Ry for the wave function and 225 Ry for the charge density. The k-point sampling was set to 4 × 4 × 4 mesh for calcite and aragonite using the Monkhorst–Pack method. For the vaterite C1_2 structure, which has the largest number of atoms to calculate, only the Γ point was used for k-point sampling. For the other vaterite structures, the k-point sampling was set to 2 × 2 × 2 mesh or more. Results and Discussion XRD, TG, FE-SEM and FTIR The XRD patterns of the CaCO3 polymorphs are shown in Figure , and the content of each crystalline phase quantified by Rietveld analysis is shown in Table . Note that Rwp (Weighted Profile R-factor) here is a dimensionless number representing the fit quality in Rietveld analysis. Since all values are around 10, this indicates that the fitting was good. Furthermore, it is considered that the targeted crystalline phase was obtained. Experimental XRD patterns and fitting data for synthetic CaCO3 powder, (a) 12CO2-calcite, (b) 13CO2-calcite, (c) 12CO2-aragonite, (d) 13CO2-aragonite, (e) 12CO2-vaterite, (f) 13CO2-vaterite. 2XRD/Rietveld Analysis 12CO2–Calcite 13CO2–Calcitecalcite (mass %)85.1479.25amorphous (mass %)14.8620.75 R wp 9.48210.895 12CO2–Aragonite 13CO2–Aragonitearagonite (mass %)87.8384.56calcite (mass %)4.427.53amorphous (mass %)7.757.91 R wp 9.9968.962 12CO2–Vaterite 13CO2–Vateritevaterite C1_2 (mass %)84.2883.11vaterite P1121 (mass %)10.9011.19calcite (mass %)0.774.28amorphous (mass %)3.321.42 R wp 10.3610.067 XRD/Rietveld Analysis Similar to the results of the syntheses using 12C and 13C, small amounts of calcite contamination were observed during the synthesis of aragonite and vaterite. The amount of calcite and aragonite synthesized using 12C differed from that using 13C. Kinetic control of the CO2 bubbling method used in this study is difficult because the state of the bubbles is difficult to control and depends on the degree of agitation, slurry viscosity, and other conditions. Therefore, this is not the subject of discussion here. Furthermore, the amorphous components listed in Table were not strictly quantified; rather, the values were calculated from the difference in the peak fittings. The lattice constants of the synthesized calcite, aragonite, and vaterite structures obtained by XRD analysis are summarized in Table . The error range shown here for the lattice constant represents computational errors in the Rietveld analysis. Actual measurements include uncertainties from sample mounting, sample inhomogeneity, and repeatability errors. Therefore, such precision is unlikely to be achieved. However, the sharpness of the diffraction peaks suggests an accuracy of four significant figures is achievable. The synthesized vaterite was formed from two structures. The main structure of the synthesized vaterite is shown in Figure (c). As identified by Saeki et al. the structured is 6-layered with monoclinic symmetry, and according to the notation in previous studies, it is denoted as C1. This is the second-lowest-energy metastable structure among the C1 structures. Herein, this structure is denoted as the Vaterite C1_2 structure. The remaining components, shown in Figure (d), are structures with P1121 space-group symmetry. 3Lattice Constants of Each Structure Obtained from XRD lattice parametersstructureisotope a (Å) b (Å) c (Å)α (deg)β (deg)γ (deg)calcite 12C5.00009 ± 0.000055.00009 ± 0.0000517.0538 ± 0.000290°90°120° 13C5.00047 ± 0.000065.00047 ± 0.0000617.0499 ± 0.000290°90°120°aragonite 12C4.96461 ± 0.000067.97217 ± 0.000095.74921 ± 0.0000690°90°90° 13C4.96404 ± 0.000057.97157 ± 0.000095.74897 ± 0.0000690°90°90°vaterite C1_2 12C12.4243 ± 0.00067.1356 ± 0.000325.7821 ± 0.000990.237° ± 0.007°99.6° ± 0.005°90° 13C12.4249 ± 0.00067.1352 ± 0.000325.7826 ± 0.000990.237° ± 0.007°99.6° ± 0.005°90°vaterite P1121 12C7.1396 ± 0.00077.1152 ± 0.000625.437 ± 0.00290°90°60.278° ± 0.007° 13C7.1398 ± 0.00077.1167 ± 0.000625.436 ± 0.00290°90°60.281° ± 0.007° Lattice Constants of Each Structure Obtained from XRD (a) Calcite, (b) aragonite, (c) Vaterite C1_2 structure, (d) Vaterite P1121 structure. Blue balls are Ca atoms, brown balls are C atoms, and red balls are O atoms. There was almost no difference in the lattice constants of CaCO3 synthesized with 12C and 13C. The difference was less than 0.05%. The TG data for the CaCO3 polymorphs are presented in Figure . For all samples, weight loss was observed from approximately 600 °C, and the weight loss was almost complete at 800 °C. These results indicate that the product was almost purely single-phase CaCO3. Furthermore, no differences were observed for the samples prepared with 12C versus 13C. For synthesized calcite, decomposition corresponding to a small amount of raw material was observed in the range of 200–600 °C. This may be due to the presence of a small amount of amorphous CaCO3, according to the results of the Rietveld analysis (Table ). SEM analysis of calcite (Figure ) showed agglomerated particles consisting of sharp crystals in the form of strips, whereas the SEM images of aragonite show columnar crystals, and vaterite comprised almost homogeneous spherical particles. These are typical particle shapes of CaCO3 polymorphs. However, there was no significant difference in the particle morphology for the samples prepared with 12C versus 13C. The SEM images in Figure show agglomerated particles consisting of sharp crystals in the form of strips or plates for calcite, aggregates of columnar crystals for aragonite, and almost spherical particles with a uniform diameter for vaterite. These are typical particle shapes of CaCO3 polymorphs. However, there was no significant difference in the particle morphology for the products obtained using 12C and 13C. TG curves of synthesized CaCO3 powder: (a) 12CO2-calcite and 13CO2-calcite, (b) 12CO2-aragonite and 13CO2-aragonite, (c) 12CO2-vaterite and 13CO2-vaterite. FE-SEM images of synthesized CaCO3 powder: (a) 12CO2-calcite, (b) 13CO2-calcite, (c) 12CO2-aragonite, (d) 13CO2-aragonite, (e) 12CO2-vaterite, (f) 13CO2-vaterite. The Fourier-transform infrared (FTIR) data for the synthesized calcite, aragonite, and vaterite are shown in Figure . For all structures, the absorption curve shifted to the lower wavenumber side for the products synthesized with 13C versus 12C. Table shows the normal vibrational frequencies of the CO3 2– ion in each structure, estimated from the peaks in the absorption curves. 4Frequency of Normal Vibration of CO3 2– Ion in CaCO3 Estimated from FTIR Absorption Curve structureω1 ω1 ′ ω2 ω2 ′ ω3 ω3 ′ ω4 ω4 ′ calcite1074107487184514061363712710aragonite1083108385382614731420713710vaterite1088108787584814101369745742degeneracy1122aω i and ω i ′ are the normal vibrations of 12CO3 2– and 13CO3 2–, respectively. The subscript i indicates the mode of vibration. i = 1 is symmetric stretching vibration, i = 2 is the out-of-plane bending vibration, i = 3 is the asymmetric stretching vibration, i = 4 is the in-plane bending vibration. Frequency of Normal Vibration of CO3 2– Ion in CaCO3 Estimated from FTIR Absorption Curve ω i and ω i ′ are the normal vibrations of 12CO3 2– and 13CO3 2–, respectively. The subscript i indicates the mode of vibration. i = 1 is symmetric stretching vibration, i = 2 is the out-of-plane bending vibration, i = 3 is the asymmetric stretching vibration, i = 4 is the in-plane bending vibration. FTIR absorption curves of (a) calcite, (b) aragonite, and (c) vaterite. Black lines are absorption curves of the materials synthesized from 12C and red lines from 13C isotope. The absorption intensity is normalized with the strongest absorption intensity set as 1.0. ω i = (i = 1,2,3,4) is the normal vibration of 12CO3 2–. The CO3 2– ion as a single molecule has 12 degrees of freedom in three-dimensional space. This is because the number of degrees of freedom corresponds to the number of atoms constituting a molecule multiplied by the spatial dimension. Among these degrees of freedom, three are due to translational symmetry and three are due to rotational symmetry. Therefore, the remaining six vibrational modes were considered. For the CO3 2– ion, which has ideal symmetry, the two vibrational modes are doubly degenerate, resulting in four observed vibrational modes. Moreover, when considering infrared spectroscopy analysis of the energy of the CO3 2– ion in the crystal structure, the translational and rotational components of the energy are much smaller than the vibrational component of the energy, and therefore, generally do not need to be considered. A schematic of the normal vibration mode is shown in Figure . The normal vibration mode ω1 is a symmetric stretching vibration; ω2 is an out-of-plane bending vibration; ω3 is an asymmetric stretching vibration; and ω4 is an in-plane bending vibration. The primed frequencies ω i ′ refer to 13C isotopes, and the vibrational mode is the same as ω i . Types of normal vibrations: (a) Symmetric stretching vibration, (b) out-of-plane bending vibration, (c) asymmetric stretching vibration, and (d) in-plane bending vibration. Black circles are carbon atoms, and white circles are oxygen atoms. Plus and minus signs indicate motion toward the paper and away from the paper, respectively. DFT and Vibration Analysis Data The equilibrium lattice parameters of calcite, aragonite, and vaterite obtained using DFT are summarized in Table . Comparison of the calculated equilibrium lattice constants with the experimental results in Table indicates a 1.3% overestimation for both calcite and aragonite, and within 1.0% overestimation for both Vaterite C1_2 and Vaterite P1121. Calculations using the GGA functional tended to overestimate the lattice constants, where an error within 2% was considered acceptable. 5Lattice Constants from First-Principles Calculation lattice parameters structure a (Å) b (Å) c (Å)α (deg)β (deg)γ (deg)atomscalcite5.05305.053017.276890.00°90.00°120.00°30aragonite5.02418.07215.815390.00°90.00°90.00°20vaterite C1_212.49137.182526.018490.43°99.88°90.29°180vaterite P1121 7.18787.177525.632390.00°90.00°60.34°90vaterite C1_112.48737.178826.012790.46°99.78°90.24°180vaterite C212.37767.27489.405690.00°115.16°90.00°60vaterite Cc12.43087.22879.485590.00°115.48°90.00°60vaterite P65 7.18877.188525.682290.00°90.00°120.00°90vaterite P31217.19617.196125.613190.00°90.00°120.00°90vaterite P32217.20047.200425.592290.00°90.00°120.00°90aAtoms is the number of atoms in the unit cell used in the calculation. Lattice Constants from First-Principles Calculation Atoms is the number of atoms in the unit cell used in the calculation. The theoretically calculated normal vibrational frequencies ω i and ω i ′ of one CO3 2– ion in the crystal structure of CaCO3 are summarized in Table . The values of ω3 and ω4 are the average values of the two vibrational frequencies for the same vibrational mode, obtained by calculation. 6Vibrational Frequency of Carbonate Minerals Obtained from First-Principles Calculation and Degeneracy of Vibrational Modea structureω1 ω1 ′ ω2 ω2 ′ ω3 ω3 ′ ω4 ω4 ′ βcalcite10341034811786132212866586561.2905aragonite10381038809784138713486526491.3020vaterite C1_210151014820795136013216436411.2988vaterite P1121 10111011816790136113226406381.2991vaterite C1_110101010818792136113236416391.2993vaterite C210071006812787135713196416391.2976vaterite Cc10071007818792134613086416391.2965vaterite P65 10121012818792136613276416391.3001vaterite P312110071006811785136313246396371.2987vaterite P322110071007817791135613176386361.2981degeneracy1122 aThe unit of frequency is cm–1. β indicates the value at 298.15 K. Vibrational Frequency of Carbonate Minerals Obtained from First-Principles Calculation and Degeneracy of Vibrational Modea The unit of frequency is cm–1. β indicates the value at 298.15 K. The notation for each vaterite structure follows that used in previous studies. This notation is thought to primarily follow the Hermann–Mauguin notation for space groups. The vibrational frequencies listed in Table are the normal vibrational frequencies of the CO3 2– ion containing the carbon atom at the substitution position where the β value is the largest. For example, in Vaterite C1_2, the location of CO3 2– can be divided into 18 different sites based on the atomic arrangement around the carbon atom. The normal vibrational frequencies of these 18 geometrically different CO3 2– are calculated individually, and the normal vibrational frequency of the site with the largest β value is listed because the carbon at this site is most likely to be converted to 13C during isotopic fractionation between 12C and 13C. Upon substituting 12C with the heavier isotope 13C, a common change in the vibrational frequency was observed for all structures. The marginal change in ω1 and ω4 with isotopic substitution of the carbon atom and the largest change observed for ω3 can be explained by the atomic arrangement and shape of the vibrational mode of the CO3 2– ion. The CO3 2– ion forms an equilateral triangle as an ideal shape. A carbon atom is at the center of gravity of the equilateral triangle, and an oxygen atom is located at each vertex. In the symmetric stretching vibration of ω1, the carbon atom at the center of gravity is not displaced due to symmetry. Thus, even if the carbon is substituted with a heavier isotope, the normal frequency does not change. ω4 is an in-plane bending vibration, and for bending vibrations, the C–O bond length does not change, but the O–C–O bond angle changes. As the bond angle changes, the carbon atom undergoes a small displacement and vibrates. Thus, the frequency slightly changes with isotopic substitution of the carbon atom. On the other hand, ω2 is also a bending vibration, but the displacement direction of the vibrating atom is not within the plane of the CO3 2– ion, and the carbon atom and oxygen atom are displaced in opposite directions perpendicular to the plane. Because the carbon is displaced perpendicular to the plane, it is thought that the change in ω2 occurs when the C–O bond length is slightly extended. ω3 is an asymmetric stretching vibration, in which the carbon atom at the center of gravity is displaced in-plane, and the change in the C–O bond length is also large. Thus, is thought that this results in the largest change in frequency. Comparing the experimental FTIR data in Table with the calculated vibration analysis data in Table shows that the frequencies ω i and ω i ′ do not match between the experimental and calculated values. However, the frequency-shift of ω2, which shows the sharpest peak in the absorption curve, to the ω2 ′ frequency upon isotopic substitution of 12C with 13C is in good agreement with the experimental and calculated values for calcite, aragonite, and vaterite. Therefore, the vibrational analysis in this study, in which 12C was replaced with the 13C isotope, reproduce the experimental results to a certain extent. In previous studies, the calculated and experimental frequencies did not match, and several corrections were proposed. , It has also been argued that this discrepancy is caused by the difference between the experimental and calculated lattice constants. Vibrational analysis was also performed by fixing the lattice constants obtained in the experiments, but this did not produce a satisfactory match with the experimental values. A method that starts from this nonequilibrium state may not satisfy the calculation conditions because vibrational analysis using first-principles calculations requires displacing atoms from the equilibrium state to obtain accurate results. The obtained β ratio for each mineral was used to calculate the fractionation factor between the minerals. , The β differs for the various structures of vaterite, resulting in isotopic fractionation between the vaterite structures. Although the largest β value for vaterite was obtained for the P65 structure, taking into account the experimental results, the β value of the vaterite C1_2 structure was used as the β value of the vaterite structure. The smallest β value obtained specifically for vaterite C1_2 is β = 1.2921. Thus, the fractionation coefficient between the site that was most easily replaced by 13C and the site that was least easily replaced in vaterite C1_2 was 0.52%. This is larger than the fractionation coefficient between aragonite and vaterite obtained from the calculation, demonstrating the importance of performing calculations for each site within a structure, as done in this study. Conclusions The synthesis of CaCO3 using 12CO2 and 13CO2 successfully produced calcite, aragonite, and vaterite. X-ray diffraction (XRD) analysis of the synthesized vaterite showed that the main component was the C1_2 structure, and the minor component had a P1121 structure. The infrared absorption data for calcite, aragonite, and vaterite almost composed of 13C were obtained for the first time. This enabled precise analysis of the isotope shift from 12C to 13C for calcite, aragonite, and vaterite. Vibrational analysis using DFT reproduced the experimentally measured isotope shifts to a certain degree. Vibrational analysis of CO3 2– sites with different symmetries in each vaterite structure was used to theoretically determine the reduced isotopic partition function ratio for each structure, which has not been previously reported. For the experimentally synthesized vaterite C1_2 structure, which is the main component of vaterite, the maximum partition coefficient between the sites was found to be larger than that between aragonite and vaterite. This indicates the importance of considering the reduced isotopic partitioning function ratio of each site in the structure when discussing isotopic fractionation. As a future development, it is possible to clarify the structures that emerge during carbonation via carbon dioxide absorption in actual concrete through FTIR measurements by making appropriate corrections to the results of the first-principles calculations in this study. Supplementary Material