Impact of Amino Acids on the Isomerization of the Aluminum

Article pubs.acs.org/IC

Impact of Amino Acids on the Isomerization of the Aluminum Tridecamer Al13 Olivier Deschaume,*,† Eric Breynaert,‡ Sambhu Radhakrishnan,‡ Stef Kerkhofs,‡ Mohamed Haouas,§ Ségolène Adam de Beaumais,∥ Valeria Manzin,∥ Jean-Baptiste Galey,∥ Laure Ramos-Stanbury,⊥ Francis Taulelle,‡ Johan A. Martens,‡ and Carmen Bartic† †

Soft-Matter Physics and Biophysics Section, Department of Physics and Astronomy, KU Leuven, Celestijnenlaan 200 D - box 2416, B-3001 Heverlee, Belgium ‡ Centre for Surface Chemistry and Catalysis, KU Leuven, Celestijnenlaan 200 F - box 2461, B-3001 Heverlee, Belgium § Lavoisier Institute of Versailles, University of Versailles Saint-Quentin en Yvelines, UMR CNRS 8180, 45 Avenue des Etats-Unis, 78035 Versailles, France ∥ L’Oréal Recherche & Innovation, 1 avenue Eugène Schueller, 93600 Aulnay-sous-Bois, France ⊥ L’Oréal Recherche & Innovation, 88 rue Paul Hochart, 94550 Chevilly-Larue, France S Supporting Information *

ABSTRACT: The stability of the Keggin polycation ε-Al13 is monitored by 27Al NMR and ferron colorimetric assay upon heating aluminum aqueous solutions containing different amino acids with overall positive, negative, or no charge at pH 4.2. A focus on the effect of the amino acids on the isomerization process from ε- to δAl13 is made, compared and discussed as a function of the type of organic additive. Amino acids such as glycine and β-alanine, with only one functional group interacting relatively strongly with aluminum polycations, accelerate isomerization in a concentration-dependent manner. The effect of this class of amino acids is also found increasing with the pKa of their carboxylic acid moiety, from a low impact from proline up to more than a 15-fold increased rate from the stronger binders such as glycine or β-alanine. Amino acids with relatively low C-terminal pKa, but bearing additional potential binding moieties such as free alcohol (hydroxyl group) moiety of serine or the amide of glutamine, speed the isomerization comparatively and even more than glycine or β-alanine, glutamine leading to the fastest rates observed so far. With aspartic and glutamic acids, changes in aluminum speciation are faster and significant even at room temperature but rather related to the reorganization toward slow reacting complexed oligomers than to the Al13 isomerization process. The linear relation between the apparent rate constant of isomerization and the additive concentration points to a first-order process with respect to the additives. Most likely, the dominant process is an accelerated εAl13 dissociation, increasing the probability of δ isomer formation.

■

formation of μ2-OH bridges between the aluminum ions of the rotated trimer and the aluminum ions in the rest of the structure. Such conformational change also leads to a more regular tetrahedral coordination of the central aluminum.5 εAl13 forms within seconds to minutes upon partial neutralization of aluminum chloride at 70 to 90 °C. Isomerization of the ε isomer to the δ isomer, also found in Al30, takes 1−2 d to complete at 90 °C.6−9 γ-Al13 was recently shown to prevail after 6 d of heating at 140−145 °C.10 Although ε, δ, and γ isomers have been observed in solution NMR spectra for decades, elucidation of their structure was only progressively achieved through improvement of synthesis procedures, enabling their crystallization and further characterization in the solid state.

INTRODUCTION Aluminum polycations play a key role in several industries and processes with high societal impact, such as water treatment, catalysis, and cosmetics, only to name a few. The polycation [Al13O4(OH)24(H2O)12]7+, hereafter denoted Al13, occurs in different isomers named according to the Baker−Figgis nomenclature initially established for tungsten polyanions.1 At pH values ranging between 3 and 5, Keggin ε-Al13, also called cagelike2 ε-Al13, is the predominant species formed at ambient temperature under many industrially and environmentally relevant conditions3 and was also the first isomer isolated from solution as sulfate or selenate crystal.4 In ε-Al13, aluminum trimers (Al3(OH)6(H2O)3) are linked together through μ1-OH bridges and to a central Al(O)4 site via μ4-O bridges. The δ, γ, β, and α isomers, respectively, contain 1, 2, 3, and 4 trimer units rotated by an angle of 60° with respect to their position in the ε structure (Figure 1). This reorientation of trimers implies the © 2017 American Chemical Society

Received: July 26, 2017 Published: September 26, 2017 12401

DOI: 10.1021/acs.inorgchem.7b01699 Inorg. Chem. 2017, 56, 12401−12409

Article

Inorganic Chemistry

Figure 1. Scheme representing structural differences between ε and δ isomers of Al13, as seen from the side (left) and top (right) of the molecule, the rotated trimer being chosen as top side. The atoms below the dotted line in the side view are not represented in the top view as well as the hydrogen atoms for a clearer visualization.

Reversely, the α isomer has been observed for a long time in the solid state, associated with silicon and fluoride in the mineral zunyite,5 but its presence in solution remains elusive. A recent report hints at its first observation in solution in the presence of D2O, that appears to slow processes related to isomerization. Large amounts of species attributed to “capped” δ isomer were detected in addition to a small and sharp peak at 72 ppm attributed to α isomer.11 In addition to the Keggin family and its derivatives, also aluminum polycations with exclusively six-coordinated aluminum atoms are known. Polycations such as Al8 and the flat-Al13 are formed at aluminum concentrations higher than 0.4 M, and, in particular, in a pH regime lower than 2. As these species only contain Al3+ under octahedral coordination, they only display signals in the 0 to 20 ppm region of 27Al NMR spectra, making them more difficult to characterize than cagelike molecules.5 Despite the issues with identification and quantification of such species, this class of aluminum polycations can now be produced in high yields12,13 and shows exciting potential for new applications, for example, when used as solution-based precursors to nanometer-thick aluminum oxide layers for electronic or optical device manufacture.14,15 Methods to stabilize the ε isomer in time or to speed generation of cagelike isomers more stable than ε-Al13 are desirable in view of their potential to prolong shelf life for Alcontaining cosmetic and water treatment products without noticeable differences in properties and efficacy. To retain their properties, aluminum species must resist the breakdown that may occur upon dilution and/or interaction with strong aluminum binders including proteins, biological membranes, and surfaces.16,17 Especially in applications where Al polycations are used as flocculants, cross-linkers or molecular spacers, the structural integrity of the molecule is essential to maintain the desired size-to-charge ratio. Different ions and molecules, inorganic, organic, and combinations thereof have been shown to interact with aluminum polycations and influence the aluminum speciation. Aluminum polycations have been shown to complex heavy metal cations such as Ni2+, Cu2+, and Pb2+.18 Sn2+ was found to accelerate the ε to δ isomerization of Al13,19 while Ga3+ stabilizes the ε isomer by replacing the central aluminum ion of Al13.19,20 Also Ca2+, glycine, and their combination influence the equilibria of aluminum polycation solutions and have been used by industry for approximately two decades to optimize the speciation in aluminum and aluminum−zirconium salt solutions to reach higher antiperspirant efficacy.21,22 Studies on aluminum complexation by amino acids have largely focused on monomeric or small size complexes, which are predominant in very dilute conditions.23−25 As a hard cation

(high charge-to-radius ratio), Al3+ preferentially interacts with hard anions such as oxygen donors. Among amino acid functional groups, aluminum therefore interacts especially with carboxylate and alcohol functions (together with phosphate for phosphorylated amino acids). With most amino acids, monodentate complexes are formed by binding through the carboxylate group, and bidentate complexes are formed when the amine function is also involved. The presence of the amine group leads to stability constants lower than the ones observed with carboxylic acids. Owing to their additional carboxylate moiety, glutamic and especially aspartic acids display higher complexation constants and delay the formation of the insoluble hydroxide phase in titration experiments. With an even larger number of carboxylate moieties, citric acid has been shown to inhibit the formation of large aluminum-containing polycations and hence also to prevent or retard the subsequent formation of particles and precipitates.26 Glycine has been used as a buffer to compensate the large pH gradients produced when titrating aluminum chloride with an alkali. It was recently proposed that glycine enables a stepwise control over isomerization, especially in the presence of divalent cations, and favors high yields for δ- and γ-Al13 isomer productions.10 A larger number of carboxylate-bearing molecules are patented for similar purposes, but little experimental data are available. With diffusion-ordered spectrocopy (DOSY) NMR, glycine was shown to bind through its carboxylate moiety and to favor isomerization from ε- to δ-Al13.11 In this study, the impact of a wide range of amino acids on the ε- to δ-Al13 isomerization is evaluated. Aluminum speciation is monitored in situ and ex situ by means of liquid-state 27Al NMR spectroscopy and ferron kinetic assay. Ferron assay is used as a fast and inexpensive screening tool, nevertheless accurate and sensitive for identifying molecules that impact aluminum speciation.

■

EXPERIMENTAL SECTION

Materials. Glycine, L-alanine, β-alanine, L-serine, D,L-aspartic acid, acid monosodium salt hydrate, L-glutamine, sodium chloride, aluminum chloride hexahydrate, aluminum atomic absorption standard solution (1 g/L aluminum in HCl), Amberlite IRA67 (15−60 mesh in free base form), and HCl 37% were purchased from SigmaAldrich (Overijse, Belgium). L-Proline, L-lysine, L-histidine, and hydroxylamine hydrochloride were purchased from Acros organics (Geel, Belgium). Calcium chloride dihydrate was purchased from Merck (Fontenay-sous-Bois, France), NaOH pellets from VWR (Leuven, Belgium), Ferron (8-Hydroxy-7-iodo-5-quinolinesulfonic acid) from Fluka (Overijse, Belgium), and acetic acid glacial from Riedel-de-Haën (Seelze, Germany). Ultrapure (UP) water was prepared using an Arium VF system. L-glutamic

12402


Article

Inorganic Chemistry

Figure 2. (a) Tetrahedral aluminum region of 27Al NMR spectra obtained while heating ε-Al13 solutions with or without glycine. Experimental spectra (black) together with the result of multipeak fitting (red) are shown for 10 and 60 min heating. (*) Peak attributed to the aluminate external reference, Al(OH)4− (80 ppm). (+) Peak attributed to aluminum in the tetrahedral cores of Al30 (70 ppm). (x) Peak attributed to aluminum in the tetrahedral core of ε-Al13 (63 ppm). (b, c) Percentage of the total aluminum concentration found in ε-Al13 (b) and Al30 (c) as a function of heating time at 90 °C, starting from an ε-Al13 solution. (d) Ferron kinetic assay determination of the aluminum fraction remaining as ε-Al13 after heating a solution of the isomer for different times in a 90 °C water bath, with or without glycine. Preparation of Additive-Free ε-Al13 by Ion Exchange. For the preparation of the initial ε-Al13 solutions, anion exchange resin Amberlite IRA67 (15−60 mesh) under free base form was used as previously described.9 Briefly, the resin was suspended in ultrapure water under continuous stirring before adding a 1 N aluminum chloride solution prepared from aluminum chloride hexahydrate to obtain the desired aluminum concentration (0.4 M and hydrolysis ratio OH/Al = 2.46).8 The resulting suspension was stirred for 3 h, while pH and conductivity were measured periodically, and in parallel, aliquots were sampled to monitor the aluminum speciation. The resin was finally separated by filtration. This protocol leads to solutions containing a small proportion of aluminum monomer as detected by NMR and ferron assay. A small amount of monomer is needed for the Al30 to form.27 Moreover, solutions containing lower levels of monomer for a total [Al3+] of 0.4 M can often be found to be overtitrated (OH/Al > 2.46), which often generates traces of aluminum hydroxide colloids, or out of equilibrium (e.g., when metathesis from sulfate crystals is used). In our case, we used the ε-Al13 solutions at least two weeks after preparation and aging at room temperature to ensure that the ε-Al13 formation was complete and that the concentration of the polycation reached a maximum. Conditions of ε-Al13 Isomerization in the Presence of Additives. Organic additives were dissolved in ultrapure water 1 d prior to use. In most experiments, the resulting solutions were adjusted to pH 4.2 using small amounts of NaOH or HCl, close to the optimal pH range for the stability of ε-Al13, and close to the pH of these solutions straight after preparation. This adjustment mostly avoids the potential change in pH produced by mixing ε-Al13 with a more acidic

or basic pH solution, which may lead to the dissolution of some polycations, or to their local aggregation/precipitation, respectively. The effect of Na+ and additional Cl− ions was checked by performing the isomerization at fixed glycine concentration and variable sodium chloride (NaCl) addition, or without adjusting the pH of glycine, and led to negligible variations in the results. Indeed, an effect of sodium could be presumed, as it was found to strongly interact with δ-Al13 in the solid state, and it was therefore suspected to play a role in the εAl13 to δ-Al13 isomerization process.6 For the isomerization experiments in the presence of additives, the aluminum polycation solutions were mixed with the additive solutions in glass tubes before homogenization with a vortex mixer and heating or transfer to NMR tubes for in situ analysis and heating. For heating treatments of ex situ measurement, the samples were heated at 90 °C for 24 or 48 h in glass tubes or at 70 °C in plastic Eppendorf tubes. The aluminum concentration used for isomerization experiments was fixed at 0.2 M, and Al/additive ratios were obtained by varying the amount of added water and the concentration/volume of additive solution. 27 Al NMR Measurements and Data Treatment. Samples were measured at 25 and/or 90 °C in 10 mm quartz NMR tubes fitted with coaxial quartz inserts filled with sodium aluminate in D2O both as reference and lock substance. A 500 MHz Bruker spectrometer was used for acquisition, equipped with an Al-free Bruker modified probehead. The 27Al spectra were recorded at a 27Al frequency of 130.32 MHz with a π/12 rad pulse length of 2.1 μs, a recycle delay of 0.1 s, acquisition time of 0.4 s, and acquisition of 1024 pulse transients. Particular attention was paid to quantitative measurement con12403


Article

Inorganic Chemistry

Figure 3. (a) Residual aluminum concentration present in ε-Al13, after heating with different concentrations of glycine, and first-order kinetic fits. (b) Calculated apparent rates of isomerization as a function of the amino acid concentration. Full symbols are for a series of concentrations tested at the same time, while hollow symbols are for additional fully independent repetitions (different time, ε-Al13 batch, additive stock solutions, etc.). The lines are linear fits of the data.

■

ditions.28,29 Free induction decay (FID) traces were treated using Bruker topspin software, and full spectral fitting was performed with Lorentzian functions using Igor Pro software. Ferron Measurements and Data Treatment. A single-beam UV−visible spectrophotometer (Ultrospec 2100 pro, GE Healthcare, Diegem, Belgium) was used throughout with plastic or quartz cuvettes (1 cm path length). An aliquot (0.1 mL) of the sample was diluted with 4.9 mL of UP water, and the solution was mixed in a closed test tube using a vortex stirrer. An aliquot (0.1 mL) of the resulting solution was then mixed with 2.4 mL of UP water and 2.5 mL of Ferron reagent (2 × 10−3 mol/L) also containing acetate buffer (0.2 mol/L) at pH 5.2. The resulting solution was also vortexed before being transferred into the UV−vis measurement cuvette. The absorbance of the solution was then recorded at 370 nm, with one reading every 3 s for a total duration of 30 min. A calibration curve was obtained using an aluminum atomic absorption standard solution (1 g/L aluminum in HCl), the absorbance of the solutions being measured 20 min after mixing the diluted standard with Ferron. Blank ferron measurements were performed with the amino acids only and with solutions containing mostly aluminum monomers and the amino acids (prepared from 1 M aluminum chloride solution without neutralization). These experiments demonstrated no significant impact of the amino acids on ferron measurements. The resulting aluminum species dissolution data was finally treated with a Matlab exponential data fitting algorithm to determine inorganic solution speciation.9,30 The fitting algorithm gives the best results for monomers and fast-reacting species up to ε-Al13. For slowly reacting species such as aluminum hydroxide particles and Al30, the fit must be parametrized with the total aluminum concentration present and a set of dissolution rate constants based on previous studies.9 Al30 and aluminum hydroxide particles are the most difficult to quantify by both solution-state 27Al NMR (due to broad spectral features and “invisible Al species”, requiring heating of the sample or magic-angle spinning (MAS) 27Al NMR to detect most of the aluminum nuclei present) and ferron assay (due to their small rates of dissolution), especially when present at low concentrations in solution. Dynamic Light Scattering. Dynamic light scattering (DLS) measurements were performed with a 90plus (Brookhaven, Holtsville, NY; 35 mW, 640 nm wavelength laser) with a detection angle of 90° and a Zetasizer Nano ZS (Malvern instruments, Malvern, United Kingdom; 4 mW maximum, 633 nm wavelength laser) at a measurement angle of 175°. For the two instruments, the manufacturer states a measurement range between 0.3 nm and 10 μm diameter. Owing to the small dimensions of the aluminum polycations, and to obtain satisfactory data for fitting, correlograms were averaged for a minimum of 1 min per run on the Zetasizer and 5 min on the 90plus, with at least four repetitions for each sample. Data treatment was performed using a software that was provided by the manufacturers, and correlograms were additionally fitted with Clementine, a toolbox developed to model decay kinetics using either maximum entropy or least-squares fitting.31,32

RESULTS AND DISCUSSION

Effect of Glycine on the Isomerization Kinetics: In Situ 27 Al NMR and Ferron Measurements. With 27Al NMR, Al13 isomerization is demonstrated and quantified by fitting the 63 ppm peak attributed to the tetrahedrally coordinated aluminum core of ε-Al13, and fitting of the 70 ppm peak attributed to the tetrahedrally coordinated aluminum cores of Al30. The spectra acquired in situ while heating ε-Al13 solutions with or without glycine were subjected to peak fitting for the quantification of aluminum nuclei present in the different coordination environments. Multipeak fitting quantification confirmed that the isomerization of ε-Al13 into δ-Al13 present in Al30 is significantly accelerated by glycine. This effect is reflected by both the acceleration of ε-Al13 consumption and that of Al30 production (Figure 2a−c). An additional peak was clearly visible at 5 ppm, and the area of the peak attributed to the aluminum monomer near 0 ppm decreased in the presence of glycine (Figure S3). This last observation can be explained by the formation of small species, probably mononuclear Al-glycine complexes, since they were absent for solutions where no monomers are initially present. This observation also demonstrates that the interaction of glycine with the polycations does not lead to a shift of the solution equilibrium toward small oligomeric aluminum species. This feature will also be discussed in relation to the comparison of the impact of different amino acids on ε-Al13 isomerization. From the ferron kinetic assay results, the kinetics of ε-Al13 consumption was also clearly accelerated in the presence of glycine. At a 1:1 Al/glycine ratio, the fraction of total aluminum present as ε-Al13 decreased from 90 to ∼20% and to ∼70%, respectively, in the presence and absence of glycine after 3 h of heating (Figure 2d). Impact of the Amino Acid Concentration on ε-Al13 to δ-Al13 Isomerization Kinetics. Aluminum speciation was analyzed by means of ferron assay after heating ε-Al13 solutions at 70 °C in the presence of different concentrations of glycine or β-alanine. At this stage, we hypothesized that the interaction of amino acids with ε-Al13, and their impact on isomerization, would be amplified if the carboxylic acid of the organic additives displayed a higher pKa as compared to glycine. With a pKa of 3.63 (2.35 for glycine) and a high solubility, β-alanine was a possible good candidate for testing this assumption. In this series of experiments, the total aluminum concentration was kept at 0.2 M, and the concentration of organic additives varied between 0.02 and 0.8 M for glycine, and between 0.02 and 0.2 M for β-alanine. 12404


Article

Inorganic Chemistry

Figure 4. Members of the different amino acid families tested, including charges displayed under the pH conditions corresponding to the domain of stability of ε-Al13 and Al30.

In both cases, the amount of ε-Al13 consumed after 48 h of heating increased with the concentration of both amino acids. The presence of low levels of the amino acids (0.02 M) already led to a distinct acceleration as compared to blank samples containing only aluminum polycations. From previous work using both 27Al NMR and ferron analysis, the change in aluminum concentration contained in εAl13 and Al30 during isomerization can be fitted with apparent first-order kinetics (see fitting used in Figures S1 & S2).19 Using the values of ε-Al13 concentrations for t0 and ttest (ferron measurement time after heating), and a zero concentration for infinite time, we estimated the value of the apparent rate constant for Al13 isomerization kiso (with respect to aluminum concentration present as ε-Al13 and the first-order kinetics hypothesis). The value of kiso is more practical to interpret and compare between experiments than residual concentrations. Furthermore, it gives a better mechanistic insight, especially when slightly different initial ε-Al13 concentrations are present due to the slow isomerization at room temperature, or when very little ε-Al13 remains in solution after heating. The apparent rate constants of isomerization increase linearly with the organic additive concentration (Figure 3). Such a trend is characteristic of a first-order reaction with respect to the additive. Similarly to the values of the apparent rate constant of ε-Al13 isomerization, the slopes of the linear trends increase in the order glycine < β-alanine. From an initial value of 1.76 × 10−4 min−1 in the absence of additive, the apparent rate of isomerization reaches 2.72 × 10−3 min−1 for 0.8 M glycine; that is, isomerization is speeded more than 15-fold. For an equal amino acid concentration, the isomerization is always faster in the presence of β-alanine. For example, for 0.2 M amino acid, the apparent rate of isomerization is ca. 7.63 × 10−4 min−1 with glycine and 1.36 × 10−3 min−1 with β-alanine. Ferron Assay Quantification of ε-Al13 to Screen Molecules for Their Impact on Aluminum Speciation. To screen a larger number of amino acids and experimental variables for their impact upon the isomerization process of ε-

to δ-Al13, a condition was selected where most water-soluble amino acids exhibit a comparable solubility and exert an observable impact on isomerization that is [Al3+] = 0.2 M, [amino acid] = 0.05 M, theating= 48 h, and T = 70 °C. ε-Al13 exhibits characteristic spectroscopic features in 27Al NMR, relatively fast and specific dissolution rate in ferron kinetic assay, and a fast parameter-dependent reorganization into a range of aluminum polycations and colloids. Consequently, εAl13 is a convenient tracer molecule evidencing the impact of the additives and other parameters on aluminum speciation. For this screening, the ferron assay was used to determine the apparent rates of ε-Al13 consumption, and 27Al NMR was used to check the final solution speciation following heat treatment. The different amino acids tested in this work are presented in Figure 4, showing the species in their dominant protonation state at pH 4.2. When ε-Al13 solutions were heated in the presence of a range of amino acids, the overall positively charged species lysine and histidine exhibit the lowest impact on the ε-Al13 consumption (Figure 5). The noncharged amino acids exhibit an increasing effect by alanine, proline, serine, β-alanine, and glutamine. The impact increases even further for the overall negatively charged aspartic and glutamic acids. This behavior reveals a strong influence of electrostatic interactions. Within the series of neutrally charged zwitterions, and in the absence of complementary interacting moieties, the ε-Al13 consumption appears to increase with the pKa of the carboxylic acid group (see Table S1). The two dicarboxylate amino acids, namely, aspartic and glutamic acids, not only exert the highest impact on ε-Al13 consumption, these have been demonstrated to delay the precipitation of aluminum hydroxide in aluminum complexation studies.23 Among oxygen donors strongly interacting with aluminum, phenolic ligands are some of the few that have been specifically studied with respect to their interaction with ε-Al13. This type of ligand displayed an initial rate-determining binding 12405


Article

Inorganic Chemistry

region may be attributed to partially dissociated or condensed cagelike species with a severely disrupted core. Occurrence of at least four distinct features in the aluminum octahedral region of the spectrum (Figure 6a) points toward rearrangement into larger oligomers rather than into monomeric or dimeric aspartic acid complexes only. This is confirmed by dynamic light scattering (DLS) showing species with sizes similar to or slightly larger than Al13 and Al30 (Figure 6b). In addition, the scattering intensities confirm the concentration of these species is of the same order of magnitude as that of the original Keggin species (Figures 6b and S4). In situ NMR of the breakdown of Keggin Al13 at room temperature in the presence of aspartic acid indicates a decrease of the area of the 63 ppm peak (Al13 tetrahedral core) together with an increase of the monomers signal at 0 ppm, related to hexaquo aluminum ions (Figure 7). Moreover, the decrease in intensity observed at 10−12 ppm (signal resulting from octahedrally coordinated nuclei in the shell of Al13) is correlated with an increase of another broad feature at 4−5 ppm. This would indicate a slow modification of solution equilibrium and the production of new oligomeric species as opposed to a simple complexation at the surface of Al13. The in situ room-temperature NMR experiment was complemented by a Ferron assay after 30 min, 24 h, and 48 h aging of ε-Al13 samples at room temperature in the presence aspartic acid. Whereas after 30 min, the ε-Al13 dissolution curves obtained in the presence of glycine, β-alanine, proline, serine, and glutamine remain nearly identical to that of the polycation alone, the curves obtained in the presence of aspartic acid show significant differences (Figure S5). In the latter case, the concentration of monomer increased markedly after 30 min, while the kinetics of dissolution slowed. No background process was observed when mixing the ferron reagent with aspartic acid alone. Ferron assay after 24 and 48 h further confirms the distinct difference between the impact of aspartic and glutamic acid on the Al13 speciation as compared to most other amino acids. In presence of amino acids other than aspartic and glutamic acid, the monomers concentration observed after 24 and 48 h remains close to the 30 min values, and dissolution rates observed are slightly lower than in absence of the amino acids. Aspartic and glutamic acid, however, promote a higher

Figure 5. Apparent rates of ε-Al13 consumption in the presence of different amino acids. The rates are plotted in logarithmic scale to better visualize the efficacy of all additives. Three groups of amino acids can be differentiated based on the observed rates, as emphasized in the color scheme used for the bars filling.

step to the polycation, followed by a rapid breakdown of εAl13.33 In the presence of aspartic and glutamic acids, a low rate of reaction with the ferron reagent was observed after the heat treatment. Different hypotheses can be raised to explain this observation: (a) all Al13 is isomerized much faster from the ε to the δ form, complexed or not by the amino acids, (b) the complexation of ε-Al13 prevents its dissolution in the ferron reagent, (c) Al13 units dissociate and reassociate into a different set of monomeric and polymeric (complexed) species with low breakdown rates in the ferron reagent. As Al13 Keggin ions still prevail in the case of the two first options, detection of the typical 27Al NMR signal of the Al13 tetrahedral core at 63 ppm would allow to discard option (c). 27 Al NMR of ε-Al13 solutions heated in the presence of aspartic acid shows the disappearance of most of the signal from the aluminum tetrahedral region, whereas it remains in the presence of glycine or β-alanine, immediately revealing the breakdown of Keggin species by aspartic acid (option (c)). The very broad feature remaining in the aluminum tetrahedral

Figure 6. 27Al NMR spectra of ε-Al13 samples treated for 48 h in the presence of different amino acids (a) and DLS intensity-based distributions obtained for samples obtained under similar conditions (b). 12406


Article

Inorganic Chemistry

Figure 7. Changes observed in selected regions of 27Al NMR spectra of an ε-Al13 sample during in situ aging at 25 °C for 78 h in the presence of aspartic acid. Spectral regions corresponding to (a) ε-Al13 tetrahedral core, (b) aluminum in octahedral coordination in monomers, polycations, and complexes with aspartic acid, and (c) monomeric aluminum. For (b), an additional data smoothing (10-point adjacent averaging) is used to facilitate the visualization of the trends. The arrows indicate the increase/decrease in peak amplitude over time.

needed to form the different isomers. The rate of Al13 dissociation is increased by heating or interaction with ligands such as O donors, leading to the release of monomeric and oligomeric species such as aluminum trimers or dimers, which then reassemble to form one of the Al13 isomers (Figure 8).

monomer concentration and induce the formation of oligomers with significantly lower dissolution kinetics as compared to Al13. The mechanism of aspartic acid interaction with ε-Al13, and the resulting change in speciation, does not appear to require much activation energy as compared to the isomerization process maintained in the presence of the other additives. Therefore, it is most likely that aspartic and glutamic acids trigger the breakdown of ε-Al13 into monomers and/or small oligomers that further reorganize into larger, more stable molecules, not containing the tetrahedral cores characteristic of Al13 and Al30. Here, the system was mostly studied at an Al/ additive ratio of 4:1, but the type and properties of the molecules produced are likely to vary as a function of this ratio. This topic would therefore require further investigation. Despite the lower pKa of their carboxylic acid group, serine enhanced isomerization with an efficacy comparable to that of glycine, and glutamine was the amino acid that induced the highest isomerization rate (at [additive] = 0.05 M). These high rates of isomerization probably originate from a soft, multiple interaction of the amino acids with aluminum species, where carboxylate acts in synergy with the alcohol moiety for serine and with the amide moiety (through, e.g., hydrogen bonding) for glutamine. Current data obtained in the presence of amino acids forming increasingly stable complexes with aluminum are better interpreted based on Al13 formation/dissociation equilibria than by rotations of the trimeric units leaving the rest of the structure unaltered. A rearrangement mechanism going through dissociation and reforming of Al13 can, for example, also explain the replacement of the central aluminum ion of cage species by other metal ions, for example, when ε-Al13 is heated in the presence of Ga3+. This interpretation is also supported by a recent density functional theory study, by Ohlin et al., demonstrating that the energies required for isomerization via trimer rotation are prohibitively high when calculated in vacuo, suggesting a dissolution−precipitation mechanism or a lowering of the energy barrier through coordination and/or solvation.34 Following the hypothesis of a dissociation−formation equilibrium, the rate of isomerization is then mostly limited by the rate of dissociation of Al13, and by the activation energy

Figure 8. Impact of amino acids on equilibria of ε-Al13 dissolution− restructuration into other isomers.

The dissociation rate grows as the pKa of the donor increases or when additional groups interact with the ε-Al13 shell. However, above a certain interaction threshold, dissociation becomes faster than Al13 formation, and/or the stability of monomeric and oligomeric complexation products disables the formation of Al13 cages. This is the case observed with aspartic and glutamic acids, as demonstrated by the changes in 27Al NMR spectra and dissolution rates of the species.

■

CONCLUSIONS AND PERSPECTIVES The present work demonstrated the link between the structure of amino acids and their impact upon aluminum speciation in relatively concentrated systems (0.2 M Al) and mildly acidic pHs (pH 4.2). In contrast with many previous studies, we focused on large aluminum polycations, such as Al13, that are the predominant species encountered under such conditions and in actual applications. In particular, we investigated how the amino acids affect their stability and conversion mechanisms. Ferron assay was used as a monitoring and screening tool, while 27 Al NMR was applied to selected in situ experiments and measurements. When studying the ε- to δ-Al13 isomerization process, NMR and ferron quantification generally agree within a few percent, making ferron an excellent first-line screening tool in simple and well-controlled cases. NMR spectroscopy 12407


Inorganic Chemistry

■

remains nonetheless an invaluable technique to fully unravel the complexity of aluminum speciation.5,35 Most amino acids speed the ε- to δ-Al13 isomerization, this effect being found to increase with the pKa of their C-terminal carboxylic acid or in the presence of soft complementary interacting groups such as alcohol or amide moiety. Hence, amino acids with a carboxylic acid group presenting a relatively large pKa such as glycine or β-alanine, or multiple softer complexing moieties such as serine and glutamine, display the highest amplification, leading to an increase of more than 1 order of magnitude in the apparent rates of isomerization, in the presence of high amino acid concentrations (∼1 M). Aspartic and glutamic acids produce an even faster reorganization of the polycations into a different class of aluminum oligomeric species, mostly displaying signals in the aluminum octahedral region of the 27Al NMR spectrum. Such species may be analogous to tri-, tetra-, or even higher nuclearity motifs (such as the flat Al13−Heidi complex) observed with many O donor organic ligands.36 For glycine and β-alanine, the apparent rate constant of Al13 isomerization kiso increases linearly with the amino acid concentration, demonstrating a first-order reaction with respect to the organic additive. Further work should help to understand in which part of the Al13 isomerization mechanisms amino acids are playing a role. The exact mechanisms for the formation, isomerization, and reorganization of aluminum polycations, for example, into insoluble aluminum hydroxides are still controversial in the literature. However, the isomerization results obtained here for amino acids forming increasing stability complexes with aluminum are best explained based on a partial dissociation of Al13, which is speeded by the soft interaction with most of the amino acids and followed by a reformation of one of the isomers of the tridecameric structure. In the case of aspartic and glutamic acids, the dissociation is enhanced one step too far, or the stability of the resulting complexes prohibits the formation of new Al13 cages. This work presents an excellent example for the production and stabilization of well-defined aluminum polycations taking advantage of soft interactions and mild reaction conditions, and this approach can be extended to other inorganic systems. Inspiring ourselves from the enhancement observed with serine and glutamine, more specific (macro)molecules could be designed to softly direct inorganic systems toward the production of specific isomers and species. Such aspects will be explored and discussed in future works.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Olivier Deschaume: 0000-0001-6222-0947 Eric Breynaert: 0000-0003-3499-0455 Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS O.D., C.B., and J.A.M. acknowledge the financial support provided by L’Oréal Paris. The authors acknowledge the Belgian Federal Government for support of the IAP project 07/ 05 Functional Supramolecular Systems. J.A.M. acknowledges the Flemish Government for long-term structural funding (Methusalem). This work was supported by the Flemish FWO via project funding (G.0885.14N) and the Hercules Foundation through an investment in NMR equipment (AKUL/13/21).

■

REFERENCES

(1) Baker, L. C. W.; Figgis, J. S. New Fundamental Type of Inorganic Complex: Hybrid between Heteropoly and Conventional Coordination Complexes. Possibilities for Geometrical Isomerisms in 11-, 12-, 17-, and 18-Heteropoly Derivatives. J. Am. Chem. Soc. 1970, 92, 3794− 3797. (2) Bi, S.; Wang, C.; Cao, Q.; Zhang, C. Studies on the Mechanism of Hydrolysis and Polymerization of Aluminum Salts in Aqueous Solution: Correlations between the “Core-Links” Model and “CageLike” Keggin-Al13 Model. Coord. Chem. Rev. 2004, 248, 441−455. (3) Furrer, G.; Phillips, B. L.; Ulrich, K.-U.; Pöthig, R.; Casey, W. H. The Origin of Aluminum Flocs in Polluted Streams. Science 2002, 297, 2245−2247. (4) Johansson, G.; Lundgren, G.; Sillén, L. G.; Söderquist, R. On the Crystal Structure of a Basic Aluminium Sulfate and the Corresponding Selenate. Acta Chem. Scand. 1960, 14, 769−771. (5) Casey, W. H. Large Aqueous Aluminum Hydroxide Molecules. Chem. Rev. 2006, 106, 1−16. (6) Rowsell, J.; Nazar, L. F. Speciation and Thermal Transformation in Alumina Sols: Structures of the Polyhydroxyoxoaluminum Cluster [Al30O8(OH)56(H2O)26]18+ and Its δ-Keggin Moieté. J. Am. Chem. Soc. 2000, 122, 3777−3778. (7) Allouche, L.; Gérardin, C.; Loiseau, T.; Férey, G.; Taulelle, F. Al30: A Giant Aluminum Polycation. Angew. Chem., Int. Ed. 2000, 39, 511−514. (8) Shafran, K.; Deschaume, O.; Perry, C. C. High-Temperature Speciation Studies of Al-Ion Hydrolysis. Adv. Eng. Mater. 2004, 6, 836−839. (9) Shafran, K. L.; Deschaume, O.; Perry, C. C. The Static Anion Exchange Method for Generation of High Purity Aluminium Polyoxocations and Monodisperse Aluminium Hydroxide Nanoparticles. J. Mater. Chem. 2005, 15, 3415−3423. (10) Smart, S. E.; Vaughn, J.; Pappas, I.; Pan, L. Controlled StepWise Isomerization of the Keggin-Type Al13 and Determination of the γ-Al13 Structure. Chem. Commun. 2013, 49, 11352−11354. (11) Oliveri, A. F.; Colla, C. A.; Perkins, C. K.; Akhavantabib, N.; Callahan, J. R.; Pilgrim, C. D.; Smart, S. E.; Cheong, P. H.-Y.; Pan, L.; Casey, W. H. Isomerization of Keggin Al13 Ions Followed by Diffusion Rates. Chem. - Eur. J. 2016, 22, 18682−18685. (12) Wang, W.; Wentz, K. M.; Hayes, S. E.; Johnson, D. W.; Keszler, D. A. Synthesis of the Hydroxide Cluster [Al13(μ3-OH)6(μ-

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.7b01699. Aluminum speciation as determined using ferron assay during the course of ε-Al13 preparation and isomerization, 27Al NMR spectra in the aluminum octahedral region during heating with or without glycine, scattering intensities of ε-Al13 heated in the presence of different amino acids, and raw ferron assay curves of ε-Al13 aged in the presence of amino acids at room temperature. Summary of the properties of amino acids used (PDF) 12408


Article

Inorganic Chemistry OH)18(H2O)24]15+ from an Aqueous Solution. Inorg. Chem. 2011, 50, 4683−4685. (13) Wang, W.; Liu, W.; Chang, I.-Y.; Wills, L. A.; Zakharov, L. N.; Boettcher, S. W.; Cheong, P. H.-Y.; Fang, C.; Keszler, D. A. Electrolytic Synthesis of Aqueous Aluminum Nanoclusters and in Situ Characterization by Femtosecond Raman Spectroscopy and Computations. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 18397−18401. (14) Smith, S. W.; Wang, W.; Keszler, D. A.; Conley, J. F. Solution Based Prompt Inorganic Condensation and Atomic Layer Deposition of Al2O3 Films: A Side-by-Side Comparison. J. Vac. Sci. Technol., A 2014, 32, 041501. (15) Perkins, C. K.; Mansergh, R. H.; Ramos, J. C.; Nanayakkara, C. E.; Park, D.-H.; Goberna-Ferrón, S.; Fullmer, L. B.; Arens, J. T.; Gutierrez-Higgins, M. T.; Jones, Y. R.; Lopez, J. I.; Rowe, T. M.; Whitehurst, D. M.; Nyman, M.; Chabal, Y. J.; Keszler, D. A. LowIndex, Smooth Al2O3 Films by Aqueous Solution Process. Opt. Mater. Express 2017, 7, 273. (16) Deschaume, O.; Shafran, K.; Perry, C. Interactions of Bovine Serum Albumin with Aluminum Polyoxocations and Aluminum Hydroxide. Langmuir 2006, 22, 10078−10088. (17) Deschaume, O.; Fournier, A.; Shafran, K. L.; Perry, C. C. Interactions of Aluminium Hydrolytic Species with Biomolecules. New J. Chem. 2008, 32, 1346−1353. (18) Lothenbach, B.; Furrer, G.; Schulin, R. Immobilization of Heavy Metals by Polynuclear Aluminium and Montmorillonite Compounds. Environ. Sci. Technol. 1997, 31, 1452−1462. (19) Parker, W. O.; Millini, R.; Kiricsi, I. Metal Substitution in Keggin-Type Tridecameric Aluminum−Oxo−Hydroxy Clusters. Inorg. Chem. 1997, 36, 571−575. (20) Thomas, B.; Görz, H.; Schönherr, S. Zum NMR-Spektroskopischen Nachweis von Dodecaaluminogallium-Ionen. Z. Chem. 1987, 27, 183−183. (21) Shen, Y.-F. Enhanced Antiperspirant Salts Stabilized with Calcium and Concentrated Aqueous Solutions of Such Salts. U.S. Patent 6,245,325, 2001. (22) Pan, L. Antiperspirant Active Compositions Having SEC Chromatogram Exhibiting High SEC Peak 4 Intensity. U.S. Patent 8,257,689, 2012. (23) Kiss, T.; Sóvágó, I.; Tóth, I.; Lakatos, A.; Bertani, R.; Tapparo, A.; Bombi, G.; Martin, R. B. Complexation of aluminium(III) with Several Bi- and Tri-Dentate Amino Acids. J. Chem. Soc., Dalton Trans. 1997, 44, 1967−1972. (24) Kiss, T.; Hollósi, M. The Interaction of Aluminium with Peptides and Proteins. In Aluminium and Alzheimer’s Disease: The Science that Describes the Link; Elsevier, 2001; p 361. (25) Kiss, T. From Coordination Chemistry to Biological Chemistry of Aluminium. J. Inorg. Biochem. 2013, 128, 156−163. (26) Dabbs, D. M.; Ramachandran, U.; Lu, S.; Liu, J.; Wang, L.-Q.; Aksay, I. A. Inhibition of Aluminum Oxyhydroxide Precipitation with Citric Acid. Langmuir 2005, 21, 11690−11695. (27) Allouche, L.; Taulelle, F. Conversion of Al13 Keggin ε into Al30: A Reaction Controlled by Aluminum Monomers. Inorg. Chem. Commun. 2003, 6, 1167−1170. (28) Gérardin, C.; Haouas, M.; Lorentz, C.; Taulelle, F. NMR Quantification in Hydrothermal in Situ Syntheses. Magn. Reson. Chem. 2000, 38, 429−435. (29) Maki, H.; Sakata, G.; Mizuhata, M. Quantitative NMR of Quadrupolar Nucleus as a Novel Analytical Method: Hydrolysis Behaviour Analysis of Aluminum Ion. Analyst 2017, 142, 1790−1799. (30) Belton, D. J.; Deschaume, O.; Patwardhan, S. V.; Perry, C. C. A Solution Study of Silica Condensation and Speciation with Relevance to in Vitro Investigations of Biosilicification. J. Phys. Chem. B 2010, 114, 9947−9955. (31) Livesey, A. K.; Brochon, J. C. Analyzing the Distribution of Decay Constants in Pulse-Fluorimetry Using the Maximum Entropy Method. Biophys. J. 1987, 52, 693−706. (32) Ablonczy, Z.; Lukács, A.; Papp, E. Application of the Maximum Entropy Method to Absorption Kinetic Rate Processes. Biophys. Chem. 2003, 104, 249−258.

(33) Forde, S.; Hynes, M. J. Kinetics and Mechanism of the Reactions of the Al13 Keggin Oligomer, [AlO4Al12(OH)24(H2O)12]7+, with a Series of Phenolic Ligands. New J. Chem. 2002, 26, 1029−1034. (34) Ohlin, C. A.; Rustad, J. R.; Casey, W. H.; Sun, S.; Jiao, H. The Energetics of Isomerisation in Keggin-Series Aluminate Cations. Dalton Trans. 2014, 43, 14533−14536. (35) Haouas, M.; Taulelle, F.; Martineau, C. Recent Advances in Application of 27Al NMR Spectroscopy to Materials Science. Prog. Nucl. Magn. Reson. Spectrosc. 2016, 94−95, 11−36. (36) Loiseau, T.; Volkringer, C.; Haouas, M.; Taulelle, F.; Férey, G. Crystal Chemistry of Aluminium Carboxylates: From Molecular Species towards Porous Infinite Three-Dimensional Networks. C. R. Chim. 2015, 18, 1350−1369.

12409


Impact of Amino Acids on the Isomerization of the Aluminum

Recommend Documents