Influence of Amino Acids on the Mobility of Iodide in Hydrocalumite

Abstract

In the cement system, hydrocalumite is a candidate adsorbent for low-level ¹²⁹I anionic species. However, the stability of hydrocalumite after immobilizing I⁻ is unclear when they are exposed to pedosphere characterized by organic substances derived from living organisms. In the present work, five amino acids were selected as simplified models of natural organic substances under alkaline conditions. L-cysteine (H₂Cys) and L-aspartic acid (H₂Asp) accelerated the release of I⁻ from I-hydrocalumite through ion-exchange. Ion-exchange of Cys²⁻ with I⁻ in I-hydrocalumite was faster than Asp²⁻, and the interlayer spacing (d₀₀₃) of Cys-hydrocalumite was smaller than that of Asp-hydrocalumite. DFT simulations not only supported the above results but also predicted that there was a positive correlation between the formation energies and interlayer spacings of amino acids intercalated hydrocalumite, depending on the configurations. Moreover, in the DFT predictions, the interaction between amino acids and metallic hydroxide layers was responsible for the formation of hydrogen bonds and Ca-O chemical bonds between the -COO⁻ groups and [Ca₂Al(OH)₆]⁺. The other three amino acids did not show intercalation through ion-exchange. The stability of I-hydrocalumite is influenced differently by coexisting amino acids, depending on the ionic sizes, charge numbers, and hydrophilicity, which cause the second contamination.

1. Introduction

In the process of cement hydration, tricalcium aluminate, Ca₃Al₂O₆ (C₃A), is one of the most reactive substances, and several hydrated phases, such as SO₄-ettringite (3CaO·Al₂O₃·3CaSO₄·32H₂O) and SO₄-hydrocalumite (3CaO·Al₂O₃·CaSO₄·12H₂O), can be formed [1,2,3]. Among them, hydrocalumite is one of the layered double hydroxides (LDHs) that is well known to immobilize anionic species due to its positively charged surfaces and large ion-exchange capacities [4,5,6]. The low-level radioactive anionic species can be immobilized in a cement matrix, which could be stored in a relatively shallow underground zone under anoxic conditions.

¹²⁹I possesses a half-life of 1.57 × 10⁷ years and is generally produced by nuclear power plants, which can be extremely hazardous to the pedosphere and human health in the long term [7,8,9]. Under anoxic repository conditions, ¹²⁹I shows a stable iodide (I⁻) form which cannot be directly precipitated like iodate (IO₃⁻) [10]. Therefore, it is highly mobile in the pedosphere. Hydrocalumite plays an important role in I⁻ immobilization through surface adsorption and incorporation into structures, suggesting that hydrocalumite can be a good candidate for the stabilization of I⁻ in Portland cement systems [11,12]. Aimoz et al. [13] found that SO₄-hydrocalumite can take up I⁻. However, the stability of hydrocalumite after immobilizing I⁻ is of concern for a deeper understanding of the effectiveness of cement stabilization/solidification (S/S).

The stability of I⁻ in the hydrocalumite interlayer is always seriously affected by its coexistence with other anions, which have higher affinities with hydrocalumite. In contaminated soil or water systems, there are a variety of coexisting inorganics and organic ions, which possibly come into contact with cementitious materials after immobilizing target pollutants [14,15,16,17]. Generally, the effects of major inorganic ions, such as Cl⁻, OH⁻, CO₃²⁻, and SO₄²⁻, in environments on the stability of hydrocalumite have been explored. Aimoz et al. [13] explored the retention behavior of I⁻ in the hydrocalumite phase, and found that Cl⁻ and CO₃²⁻ showed a competitive relationship with I⁻ to intercalate into the hydrocalumite interlayer and reduced the stability of I-hydrocalumite and SO₄-hydrocalumite. Nedyalkova et al. [18] also confirmed that the capability of hydrocalumite bearing I⁻ became weak due to the formation of monocarbonate as the reaction time went on. The selectivity of anions in hydrocalumite is important to immobilize anionic pollutants in the environment. In environmental mineralogy, the effect of organic substances cannot be ignored because they are abundant in the pedosphere. However, the effects of organic substances on the stability of I-hydrocalumite have not yet been explored. Amino acids are monomers of proteins that exist in free single molecules and degradation products of soil, small animals, plants, and microorganisms. Natural amino acids contain amine (-NH₂) and carboxyl (-COOH) groups and can be classified into hydrophobic, neutral, and hydrophilic groups depending on their specific side chains (R group). Some amino acids can be charged to −2 due to their two carboxyl groups and the existence of a thiol group. Therefore, to some extent, the chemical properties of amino acids can represent complicated natural organic substances as models [19].

In the current work, I-hydrocalumite was prepared through co-precipitation. Five amino acids were selected to explore the stability of I-hydrocalumite, considering the molecular sizes, charge numbers, functional groups, and hydrophilicity. Based on Ca and Al oxide hydration under alkaline conditions, the released behavior of I⁻, adsorption behavior of amino acids, and solid residue phase changes were elaborated. Density functional theory (DFT) simulations were applied to determine the possible configurations and support the experimental results and the possible release mechanisms of I⁻ in the presence of amino acids.

2. Materials and Methods

2.1. Reagents

All reagents used in the present work were purchased from Wako Chemicals (Osaka, Japan). Inorganic reagents including CaCO₃ (98%), Al₂O₃ (99%), NaOH (97%), and CaI₂·nH₂O (99.5%) were of analytical grade.

There are 20 types of amino acids that exist in nature. Depending on their specific side chain groups, they can be classified as hydrophilic, hydrophobic, and amphipathic. L-phenylalanine (HPhe, C₉H₁₁NO₂, 99%) and L-tryptophan (HTrp, C₁₁H₁₂N₂O₂, 99%) have aromatic groups that make them more hydrophobic than L-cysteine (H₂Cys, C₃H₇NO₂S, 99%), L-aspartic acid (H₂Asp, C₄H₇NO₄, 99%), and glycine (HGly, C₂H₅NO₂, 99%), which have hydrophilic side chain groups. Each amino acid has different acid dissociation constants (pK_a); depending on the relationship between the solution pH and pK_a, the different negative charge numbers for each amino acid can be produced. Under the present experimental conditions, where pH was higher than 12.0, HGly, HTrp, and HPhe had one negative charge but H₂Cys and H₂Asp had two negative charges. Hence, different behaviors could be expected. H₂Cys, H₂Asp, HGly, HTrp, and HPhe of a special grade were used. For high performance liquid chromatography (HPLC) analyses, phenyl isothiocyanate (PITC, 98%) of a special grade, acetonitrile, methanol, and n-hexane of HPLC grade were used. All the solutions were prepared using decarbonized Milli-Q water (integral water purification system, Millipore (Burlington, MA, USA)) that was boiled and purged with N₂ gas for 2 h.

2.2. Synthesis of I-Hydrocalumite

Tricalcium aluminate, Ca₃Al₂O₆ (C₃A), was prepared by the calcination of a mixture of 30.027 g of CaCO₃ and 10.196 g of Al₂O₃ at a molar ratio of 3:1 for 72 h at 1300 °C in a furnace [11,20]. Thereafter, the obtained solid was obtained for the following experiments. I-hydrocalumite, Ca₄Al₂(OH)₁₂I₂·4H₂O, was prepared by co-precipitation using 0.02 mol C₃A and 0.02 mol CaI₂·nH₂O, stirring at 600 rpm in a glovebox (95% N₂, 5% H₂, COY, M-160, Grass Lake, MI, USA) to avoid carbon dioxide (CO₂) contamination. After 72 h, the suspension was centrifuged to separate the solid residues, which were rinsed with decarbonized water, freeze-dried for 24 h, and then stored for further use.

2.3. Reaction with Amino Acids

A total of 50 mg of synthesized I-hydrocalumite were suspended in 50 mL of solutions, with and without 1.7 mM amino acids. The initial pH of each solution was adjusted to pH 12.0 using 3 M NaOH. A pH of 12.0 was selected to simulate the alkalinity conditions of a real cement system. The suspensions were shaken at 100 rpm in polyethylene plastic bottles for different time intervals, from 0.5 to 24 h, followed by collection of leaching solutions using disposable syringes and solid residues by filtration using 0.2 μm membrane filters to freeze-drying for 12 h. The solubility of I-hydrocalumite in ultrapure water was evaluated separately without adding amino acids. After reacting with amino acids, the soluble fraction of I⁻ was evaluated.

2.4. Chemical Analysis and Solid Characterization

Inductively coupled plasma optical emission spectrometry (ICP-OES, Optima 8300, Perkin Elmer, Waltham, MA, USA) was adopted to determine the concentrations of Ca and Al. The concentration of I⁻ was determined by ion chromatography (IC, ICS-2100, Thermo Scientific, Yokohama, Japan). High-performance liquid chromatography (HPLC, LC-NetⅡ/ADC, Jasco, Hachioji, Japan) was used for determination of the concentrations of amino acids. The solid residues were characterized using X-ray diffraction (Ultima IV XRD Rigaku, Akishima, Japan) using Cu Kα radiation (40 kV, 40 mA, λ = 1.54184 Å) with a Ni filter at a scanning speed of 2°/min and scanning step of 0.02°. The morphology of the hydrocalumite samples, before and after reaction with different amino acids, was observed using a field emission scanning electron microscope (FlexSEM 1000, Hitachi, Tokyo, Japan) at 20 kV. The soluble fraction was calculated based on concentrations of Ca²⁺, Al(OH)₄⁻, and I⁻, following Equation (1):

Dissolved fraction (%) = C_t/C₀ × 100

(1)

where the C_t (mmol) is the released amount of ions in the solution at time t (h) and C₀ (mmol) is the amount of ions in pristine I-hydrocalumite solid.

2.5. DFT Simulation

Density functional theory (DFT) simulations were applied to predict the intercalation of amino acids, as well as the possible interaction between amino acids and hydrocalumite. Due to the expanded peaks in the XRD patterns that only appeared in the presence of Asp²⁻ and Cys²⁻, the intercalation behavior of Asp²⁻ and Cys²⁻ may have occurred. Thus, Asp²⁻ and Cys²⁻ were selected as the model amino acids for the DFT simulation.

The projector augmented-wave (PAW) method [21,22] implemented in VASP code [23,24,25] was employed for the precise description of electronic structures and also due to its low computational costs. The PBEsol-type of exchange-correlation functional within the generalized gradient approximation (GGA) [26,27,28], combined with D3 van der Waals correction [29,30], was used with the plane-wave cutoff energy of 550 eV. Dispersion forces belonging to van der Waals forces are types of force acting between atoms and molecules that are normally electrically symmetric [31]. Hydrogen bonds are primarily the electrostatic forces of attraction between a hydrogen (H) atom, which is covalently bound to a more electronegative atom or group, and another electronegative atom with a single pair of electrons [32]. A weak interaction includes electrostatic and dispersion forces. Thus, in the DFT simulation, although H-bonded interactions are dominant, the contribution of dispersion forces is not negligible in computing accurate interaction energies. The states as valence electrons were: 1s for H; 2s and 2p for C, N, and O; 3s and 3p for Al and S; 3p and 4s for Ca.

Initial structural models were prepared as shown in Figure S1, reported in the Supplementary Materials. Firstly, the structural model of positively charged hydroxide layers for Ca₂Al-LDH was constructed for the computer simulations. The hydroxide layers of Ca₂Al-LDH used in the simulations were set using atomic coordinates extracted from the previously reported crystal structure of hydrocalumite with the composition of [Ca₈Al₄(OH)₂₄(CO₃)(Cl)₂·10H₂O] [33]. By removing the intercalated carbonate (CO₃²⁻) and chloride (Cl⁻) ions, an “empty” LDH model was created. Since the composition of the LDH layer was [Ca₂Al(OH)₆]⁺, the number of Al atoms corresponded to the positive charge of the empty LDH model. Secondly, specific Asp²⁻ and Cys²⁻ anions and water molecules were placed in the interlayer spaces of the empty LDH layers at random positions and rotation angles using the pymatgen python library [34] so as to neutralize the positive charge. Thus, more than ten simulated unit cells consisting of the positively charged hydroxide layers of [Ca₂Al(OH)₆]⁺ and desired intercalated anions were constructed. The relaxation of lattice constants and internal coordinates was performed until the residual stress and force decreased to 4 MPa and 1 meV/Å, respectively. The pymatgen python library was used to extract the structural information such as interlayer spacing and the number of chemical bonds from the optimized structures. The VESTA code was used to visualize the optimized structures [35].

3. Results and Discussion

3.1. Characterizations of Pristine I-Hydrocalumite

Figure 1a,b presents the XRD patterns of the calcined tricalcium aluminate (C₃A, 3CaO·Al₂O_3, PDF#32-0149) and I-hydrocalumite (3CaO∙Al₂O₃∙CaI₂∙10H₂O, PDF#042-1474). The solid phase of C₃A is consistent with a previous report [36]. After being co-precipitated with CaI₂·nH₂O for 72 h, I-hydrocalumite was formed in a pure phase with a d₀₀₃ value of 8.75 Å at a diffraction angle of 10.10° (2θ), which is consistent with previous reports [13,18]. The molar ratio of Ca/Al is 1.87 is shown in

Table 1. Elemental compositions of I-hydrocalumite.

LDHs	Chemical Formula	Ca/mmol⋅g−1	Al/mmol⋅g−1	I/mmol⋅g−1	Ca/Al
I-hydrocalumite	Ca6.0⋅Al3.2⋅(OH)18⋅(I)2.19‧6H2O	6.0	3.2	2.19	1.87

. Therefore, considering the above results, the formation of I-hydrocalumite is expressed as the following reaction [37] (Equation (2)):

3CaO·Al₂O₃ + CaI₂ + 10H₂O = Ca₄Al₂(OH)₁₂I₂·4H₂O

(2)

3.2. Effects of Amino Acids on Releasing of I⁻ from I-Hydrocalumite in Alkaline Solutions

The changes in I⁻, Ca²⁺, Al(OH)₄⁻, pH, and amino acids at different suspension time intervals are shown in Figure 2. Without any amino acids, after I-hydrocalumite was suspended in an alkaline solution for 0.5 h, and the leaching amount of I⁻ reached 50.0% of the total amount in pristine I-hydrocalumite. As the reaction time continued, the concentration of the released I⁻ reached 2.00 mM (91.0%) and the dissolution equilibrium was almost achieved after 6 h. The concentrations of the released Ca (Figure 2b) and Al (Figure 2c) reached 5.37 and 2.78 mM after 24 h, respectively. In addition, the ratios of released Ca/Al were always within 1.85–1.95 which were close to the Ca/Al ratio in pristine I-hydrocalumite. This indicates that the metallic hydroxide layer structure of I-hydrocalumite easily enabled simple dissolution, even though it was under strongly alkaline conditions (Equation (3)).

Ca₄Al₂(OH)₁₂I₂·4H₂O = 4Ca²⁺ +2Al(OH)₄⁻ + 2I⁻ + 4OH⁻ + 4H₂O

(3)

To verify the dissolution equilibrium, the reactions in this work were continued until the 24 h mark. The XRD patterns of solid residues collected from blank and amino acids tests at different reaction times are presented in Figure 3. Figure 3a shows the results of the blank test, where the phase of I-hydrocalumite was maintained for 3 h, and, after that, it was gradually phase-transferred into the metastable phase of calcium aluminum oxide hydrate (4CaO·Al₂O₃·xH₂O, PDF#02-0077) and 3CaO·Al₂O₃·xH₂O, PDF#02-0083). After 6 h, the phase of 4CaO·Al₂O₃·xH₂O completely disappeared and 3CaO·Al₂O₃·xH₂O became the end-product, but the Ca²⁺ and Al(OH)₄⁻ concentrations were kept constant, which indicated that ion-exchange of I⁻ with OH⁻ also happened in the simple dissolution process.

Amino acids of H₂Cys, H₂Asp, HGly, HTrp, and HPhe at a concentration of 1.7 mM were used to explore the environmental impact of amino acids on the stability of I-hydrocalumite. As shown in Figure 2d, the equilibrated pH in all series was higher than 12.0, so the -COOH, -NH₃⁺, and -SH (thiol) functional groups in all amino acids were dissociated because of the pKa values of five amino acids. This means H₂Cys, H₂Asp, HGly, HTrp, and HPhe existed in the form Cys²⁻, Asp²⁻, Gly⁻, Trp⁻, and Phe⁻ under the present pH. Based on the bond length and bond angle of each amino acid molecule, the sizes including length (Å) and width (Å) were estimated [38,39]: glycine (3.9/3.5 Å), L-cysteine (6.4/3.7 Å), L-aspartic acid (6.5/3.5 Å), L-phenylalanine (9.7/4.7 Å), and L-tryptophan (10.9/5.6 Å), so we can clearly distinguish that L-phenylalanine and L-tryptophan had larger molecular sizes. Combing the van der Waals volumes [40] of glycine (48 ų), L-cysteine (86 ų), L-aspartic acid (91 ų), L-phenylalanine (135 ų), and L-tryptophan (163 ų), the glycine had the smallest molecular size and L-cysteine was smaller than L-aspartic acid.

In the presence of H₂Cys, at 0.5 h and 1 h, the leaching fraction of I⁻ corresponded to 74.7% (1.637 mM) and 91.9% (2.016 mM), respectively (Figure 2a). After 24 h, the released concentration was 2.128 mM, corresponding to 97.1% of the original content of I⁻ in I-hydrocalumite. Therefore, the release rate of I⁻ was great in the presence of H₂Cys, compared with the blank test at all times; finally, the leaching concentration was larger than 2.003 mM in the blank. From Figure 2e, the decreasing changes in the H₂Cys concentration suggest a possible ion-exchange mechanism of Cys²⁻, as in Equation (4):

Ca₄Al₂(OH)₁₂I₂·4H₂O + Cys²⁻ = Ca₄Al₂(OH)₁₂(Cys)·4H₂O + 2I⁻

(4)

After 0.5 h, the concentration of Cys²⁻ reduced to 0.75 mM, which meant that 0.95 mM Cys²⁻ was immobilized in the solid. According to the charge balance, Cys²⁻ provided 1.9 meq/L negative charges, which were enough to substitute with 1.637 mM I⁻ from I-hydrocalumite. In addition, the XRD patterns verified the expectation. In Figure 3b, two reflections at d₀₀₃ 10.306 Å (8.58°, 2θ) and 11.694 Å (7.56°, 2θ) appeared after 0.5 h and should be assigned to Cys-hydrocalumite. After 3 h, the d₀₀₃ peak intensity of Cys-hydrocalumite at 8.54° (2θ) decreased and almost disappeared after 24 h, resulting in 3CaO·Al₂O₃·xH₂O, which is the main phase. The Ca²⁺ and Al(OH)₄⁻ concentrations until 6 h, in Figure 2b,c, showed higher values than the blank test, which might suggest the ion-exchanging process destabilized the metallic hydroxide layer of I-hydrocalumite. After 6 h, the decrease in Ca²⁺ and Al(OH)₄⁻ concentrations were ascribed to the co-precipitation with OH⁻ to form 3CaO·Al₂O₃·xH₂O.

H₂Asp showed a slightly different trend from H₂Cys, while both amino acids were charged to −2 under the present pH conditions. In the presence of H₂Asp, the release rate of I⁻ was slower than that of the H₂Cys series but larger than the blank test, as shown in Figure 2a. At the suspension times of 0.5 h and 1 h, the released concentrations of I⁻ were 1.258 mM (57.4%) and 1.426 mM (65.1%), respectively, which were much lower than in the H₂Cys series. After 24 h, 2.087 mM I⁻ was released, corresponding to 95.2% of the original amount of I⁻ in I-hydrocalumite. The changes in Asp²⁻ concentrations in Figure 2e show that adsorbed amounts of Asp²⁻ were much smaller than Cys²⁻. After 0.5 h, the concentration of Asp²⁻ decreased from 1.70 mM to 1.42 mM, so 0.56 meq/L of negative charge were involved in ion-exchange with I⁻. In Figure 3c, a phase of Asp-hydrocalumite appeared within 0.5 h with a d₀₀₃ value of 11.305 Å (7.82°, 2θ) and was maintained until 6 h with a small intensity. Therefore, this solid phase changing process also verified the ion-exchange of Asp²⁻ with I⁻, as in Equation (5):

Ca₄Al₂(OH)₁₂I₂·4H₂O + Asp²⁻ = Ca₄Al₂(OH)₁₂(Asp)·4H₂O + 2I⁻

(5)

In Figure 2b,c, the released Ca²⁺ and Al(OH)₄⁻ concentrations were always lower than in the blank test, except in the presence of H₂Cys, demonstrating that leaching of I⁻ was not caused by only simple dissolution. Thus, the residual 0.698 mM I⁻ at 0.5 h was ascribed to ion-exchange with OH⁻ in I-hydrocalumite to form 3CaO·Al₂O₃·xH₂O.

As observed in Figure 3b,c, the d₀₀₃ diffraction peaks of Cys/Asp-hydrocalumite finally disappeared after 24 h, but the released fraction of Ca and Al concentrations were kept mostly constant. This means that Cys/Asp-hydrocalumite phases were unstable and changed into a 3CaO·Al₂O₃·xH₂O phase through ion-exchange of Cys²⁻/Asp²⁻ with OH⁻, as in Equation (6):

Ca₄Al₂(OH)₁₂(Amino acid)·4H₂O + yOH⁻ = 3CaO·Al₂O₃·xH₂O + (Amino acid)²⁻ + Ca²⁺

(6)

The different performances of H₂Cys and H₂Asp on the release rate of I⁻ might be caused by their different molecular sizes and the existence of a thiol group (-SH) in H₂Cys. Cys²⁻ possesses a slightly smaller molecule than Asp²⁻ under the same alkaline conditions, which made it easier to enter the interlayer of hydrocalumite.

However, in the HGly and HTrp series, Figure 2a shows a similar amount of released I⁻ in the blank test within the initial 0.5 h. During the period of 0.5 h to 6 h, the leaching amount of I⁻ was lower than in the blank. Furthermore, the released amounts of Ca²⁺ (Figure 2b) and Al(OH)₄⁻ (Figure 2c) were also lower than in the blank. This means that the surface adsorption of Gly⁻ and Trp⁻ by electrostatic interaction on hydrocalumite might happen more easily than their intercalation into the [Ca₂Al(OH)₆]⁺ metallic interlayer. Figure 2e presents the changes in amino acids concentrations. The adsorption of Gly⁻ and Trp⁻ on solid phases were always very low over the reaction times, which is consistent with the prediction that only surface adsorption of them happened, rather than intercalation. This phenomenon relates to their properties. HGly has the smallest molecular size among these amino acids, and is dissociated to Gly⁻ at the presented pH. Therefore, the ion-exchange of Gly⁻ with I⁻ appears to be difficult. For HTrp, which possesses a relatively large molecular size, one negative charge, and an aromatic group that made it more difficult to go into the hydrocalumite interlayer by substituting with I⁻. After adding 1.7 mM HGly, the phase of I-hydrocalumite was maintained within 6 h but disappeared and transformed into 3CaO·Al₂O₃·xH₂O as the main product after 24 h (Figure 3d). It seems that HGly acted to stabilize I-hydrocalumite a bit longer, compared with no co-existing amino acids. It seems that no d₀₀₃ diffraction peak of Gly-hydrocalumite appeared in small diffraction angles, so the intercalation of HGly did not significantly occur. When HTrp was co-existing with I-hydrocalumite (Figure 3f), the intermediate phase of calcium aluminum oxide hydrate of 2CaO·Al₂O₃·xH₂O (PDF#12-0008), 4CaO·Al₂O₃·xH₂O, and 3CaO·Al₂O₃·xH₂O appeared after 3 h. However, the XRD peaks assigned to 2CaO·Al₂O₃·xH₂O seemed to overlap with I-hydrocalumite from 3 h to 6 h.

HPhe showed slightly stronger suppressing effects to release I⁻ than HGly and HTrp (Figure 2a). HPhe and HTrp belong to aromatic amino acids that have similar molecular sizes and structures, but within 1 h, the concentration of HPhe decreased to 1.04 mM (Figure 2e), indicating that 0.66 mM Phe⁻ was kept in the solid phase. Combining the solution data with the corresponding XRD patterns, the surface adsorption of Phe⁻ can be determined. In Figure 3e, 3CaO·Al₂O₃·xH₂O became the main phase after 24 h but I-hydrocalumite in the solid residues was still maintained. It is obvious that I⁻ was still preserved in hydrocalumite rather than released into the solution. This phenomenon is consistent with the solution results in Figure 2a, indicating the more significantly inhibitory effects of HPhe. Between 1 h and 3 h, the concentration of Phe⁻ in the solid phase decreased, and after that it became constant. However, after 3 h, the main phase was transformed from I-hydrocalumite to 3CaO·Al₂O₃·xH₂O (Figure 3e), suggesting that HPhe might be likely to be retained on the surface of I-hydrocalumite and 3CaO·Al₂O₃·xH₂O. Based on the above results, HPhe and HTrp showed different effects on the leaching of I⁻ from I-hydrocalumite. HPhe and HTrp have one -COOH group and a side chain with an aromatic group to give it a hydrophobic property. Initially, the ion-exchange and surface adsorption of negatively charged Phe⁻ and Trp⁻ are able to occur on I-hydrocalumite. The adsorbed Phe⁻ and Trp⁻ may modify the surface properties of I-hydrocalumite from hydrophilic to hydrophobic to some extent [41,42]. Under a high concentration of 1.7 mM of HPhe and HTrp, more hydrophobic molecules were adsorbed onto the surface rather than stabilized in the interlayer. However, HPhe has a stronger hydrophobic property when Phe⁻ is adsorbed onto the surface by electrostatic interaction; OH⁻ in the solution had difficulty in contacting I-hydrocalumite, thus, it might prevent OH⁻ ion-exchange with I⁻.

To ensure the decomposition of I-hydrocalumite, the particle morphologies of solid residues, before and after reaction with different amino acids for 24 h, were observed using SEM and are shown in Figure 4. Figure 4c shows the morphology of I-hydrocalumite solid residue after being suspended in an alkaline solution in the absence of amino acids. A hexagonal structure, irregular structures, and a collapsed layer structure were observed, which are characteristic of typical layered double hydroxides [43]. After suspension in different amino acids solutions, in the H₂Cys (Figure 4d), H₂Asp (Figure 4e), HGly (Figure 4f), and HTrp (Figure 4h) series, there were not only collapsed layer structures, but also aggregation and the frame of a hexagonal structure, suggesting the dissolution of I-hydrocalumite. For HPhe (Figure 4g), the layered structure and clear hexagonal-like morphology still existed, which was consistent with the XRD pattern at 24 h (Figure 3e).

Therefore, based on the above results, HGly, HTrp, and HPhe suppressed the dissolution rate of I-hydrocalumite, and HPhe also suppressed the released amount of I⁻, while H₂Cys and H₂Asp enhanced the release of I⁻ from I-hydrocalumite to form other solid phases, but at different rates. Furthermore, a small amount of I-hydrocalumite still existed according to the SEM images, where a hexagonal morphology was maintained.

3.3. Behavior of Amino Acids in I-Hydrocalumite

To explore the different effects between Cys²⁻ and Asp²⁻ on the interaction with I-hydrocalumite, the total formation energies (DE) were calculated following Equations (7) and (8):

[Ca₂Al(OH)₆]⁺ + 0.5(Asp²⁻/Cys²⁻) → Ca₂Al(OH)₆·0.5(Asp/Cys)

(7)

DE = E(Ca₂Al(OH)₆·0.5(Asp/Cys)) − [E([Ca₂Al(OH)₆]⁺) + E0.5(Asp²⁻/Cys²⁻)]

(8)

where E(Ca₂Al(OH)₆·0.5(Asp/Cys)), E([Ca₂Al(OH)₆]⁺), and E0.5(Asp²⁻/Cys²⁻) were the calculated energies from DFT.

Considering different configurations of Cys-hydrocalumite and Asp-hydrocalumite, total formation energies were plotted against their interlayer spacings (Figure 5). There was a positive correlation, which indicated that the total formation energies of the system also increased with the increase in interlayer distances. Higher formation energy means a more unstable system. Here, we selected the most stable and most unstable systems, which possessed the lowest and highest total energies, respectively. In Figure 6a, the most stable configuration Ca₄Al₂(OH)₁₂·Asp (0 kJ/mol) showed an interlayer distance of 7.34 Å with one Ca-O bond and seven hydrogen bonds in one unit cell. In Figure 6b, the most unstable system Ca₄Al₂(OH)₁₂·Asp (71.2 kJ/mol) showed a layer spacing of 8.57 Å with one Ca-O bond and six hydrogen bonds between the metal hydroxide layers and Asp²⁻ in one unit cell. For H₂Cys, the most stable system of Ca₄Al₂(OH)₁₂·Cys with a formation energy of 0 kJ/mol showed a 7.22 Å interlayer spacing including one Ca-O and three hydrogen bonds between metal hydroxide layers and Cys²⁻ (Figure 6d). The most unstable system, Ca₄Al₂(OH)₁₂·Cys, showed 8.92 Å of interlayer distance in 85.6 kJ/mol (Figure 6e). From these results, the simulations predicted a much lower interlayer spacing than the observed XRD results (Figure 3b,c). As can be confirmed, the water molecules can cause the expansion of the interlayer spacings in experimental results [44,45]. Thus, we considered a new system, which included one water molecule in one unit cell of hydrocalumite including Asp²⁻, as shown in Figure 6c. In the system of Ca₄Al₂(OH)₁₂Asp·4H₂O, the corresponding interlayer spacing was 7.99 Å under the lowest formation energy of 0 kJ/mol, which indicated that the presence of water molecules can affect the expansion of the interlayer distances. Therefore, in a real system, the most stable state is possibly overlapped by the d₀₀₃ peak in XRD, assigned to 3CaO·Al₂O₃·CaI₂·10H₂O in 11.6°, 2θ. The XRD diffraction peaks at 8.54° (2θ), 7.61° (2θ), and 7.92° (2θ) in Figure 3b,c might be caused by the different water molecule numbers included in each system, affecting the configurations of Asp²⁻ and Cys²⁻ in the metallic hydroxide layer. Another factor might be related to the difference in the basal layer charge densities, which has been confirmed through the positive correlation relationship between them [46].

4. Conclusions

The interaction of five selected amino acids with iodide intercalated hydrocalumite was investigated by solid and solution chemistry integrated with DFT simulations. I-hydrocalumite was very fragile and caused dissolution, even at pH 12.0. After being suspended in amino acids solutions, the main factors that caused the release of I⁻ were divided into ion-exchange and simple dissolution. H₂Cys and H₂Asp accelerated the releasing rate of I⁻ through ion-exchange between Cys²⁻/Asp²⁻ and I⁻. Using XRD, expanded d₀₀₃ spacings were observed, which was different from the other three amino acids. However, the effects of Cys²⁻ and Asp²⁻ on the ion-exchange with I⁻ in hydrocalumite and the stability of I-hydrocalumite were different. Asp²⁻ maintained a structure of I-hydrocalumite for a longer time than Cys²⁻, and Cys-intercalated hydrocalumite showed one smaller d₀₀₃ spacing than Asp-intercalated hydrocalumite in the XRD patterns. This was also verified by DFT simulations. DFT simulations also confirmed that the interlayer spacings might be strongly influenced by the presence of water molecules and water numbers. There are interaction forces between amino acids and hydrocalumite in which the hydrogen and Ca-O bonds contributed to their configurations, but the Asp/Cys-hydrocalumite system was still unstable. In the presence of HGly, HTrp, and HPhe, the simple dissolution and ion-exchange with OH⁻ contributed to I-hydrocalumite decomposition. Moreover, HPhe showed slight inhibitory effects on the release of I⁻ by surface adsorption. Thus, some organic matter that is negatively charged with a large molecule and less affinity than I⁻ might make hydrocalumite stable. Hence, for iodide isotope stabilization in cementitious materials, I-hydrocalumite may be formed, but its stabilization is strongly influenced by the geochemical environment.