Chiral Imidazolium-Functionalized Au Nanoparticles: Reversible Aggregation and Molecular Recognition.

Gold nanoparticles (AuNPs) stabilized by imidazolium salts derived from amino acids [glycine (1), rac-alanine (2), l-phenylalanine (3), and rac-methionine (4)] were prepared. The AuNPs were stabilized the most by 4, which kept the particles dispersed in water for months at pH > 5.5. These AuNPs exhibited a well-defined absorption band at 517 nm and had an average particle size of 11.21 ± 0.07 nm. The 4-AuNPs were reversibly aggregated by controlling the pH of the solution. Chiral R,R-4-AuNPs and S,S-4-AuNPs were synthesized, and the chiral environment on the nanoparticle surface was confirmed using circular dichroism; these nanoparticles exhibited a molecular recognition of chiral substrates. Furthermore, they showed potential for separating racemic mixtures when supported on a layered double hydroxide.

Introduction

Gold nanoparticles (class="Chemical">AuNPs) have been studied extensively by the scientific community becn class="Chemical">ause of their unique electronic, chemical,[1] and optical properties,[2,3] which originate from localized surface plasmon bands.[4] Special emphasis has been placed on synthesizing stable water-soluble metallic nanoparticles (MNPs)[5−7] to expand their potential applications, especially in the fields of biology,[8,9] medicine,[10,11] and green chemistry.[12−14] Molecular coatings on metal nanoparticles play an important role in providing the desired water solubility and chemical functionality and in enabling the nanoparticles to assemble, which is essential in all of the above-mentioned applications.[1]

Furthermore, over the past decade, considerable efforts have been devoted to the synthesis and characterization of opticclass="Chemical">ally active Mn class="Chemical">NPs[15] capped with chiral organic compounds.[16] Their preparation usually requires the use of protective agents such as surfactants,[17] polymers,[18] and ligands.[7] Recently, DNA,[19]l/d-tartaric acid,[20]l/d-penicillamine,[21]l/d-amino acids,[22,23]R/S-BINAP,[24] and R/S-BINAS,[25] among others, have been used to prepare chiral AuNPs, where circular dichroism (CD) spectrum is the most important tool to evaluate the chiral environment.[26]

Chirclass="Chemical">al Mn class="Chemical">NPs have the potential for application in heterogeneous enantioselective catalysis,[7,24] chiral recognition,[23] enantioseparation,[16] and nonlinear optics.[22] In particular, Guonan Chen and co-workers recently demonstrated the potential of chiral MNPs for molecular recognition and resolution of enantiomers in racemic mixtures by using a system based on AuNPs modified with functional nucleic acids to perform the enantiospecific separation of rac-tryptophan.[27] Gellman et al. studied the absorption of rac-propylene oxide on l/d-cysteine-modified AuNPs and measured the enantioselective response using optical rotation.[16] Recently, Baoxin Li and co-workers explored the potential of l-tartaric acid-capped AuNPs for chiral recognition of l-mandelic acid by hydrogen bond interactions. This system is cheap and efficient because it enabled visual chiral recognition by a change in the solution color.[20]

On the other hand, we reported the synthesis of class="Chemical">imidazolium zwitterions derived from amino acids and used them to prepare n class="Disease">N-heterocyclic carbene (NHC) gold(I) and (III) complexes.[28] These complexes exhibited a strong tendency to form AuNPs in solution. Here, we report the successful synthesis and characterization of chiral AuNPs stabilized by these imidazolium salts in aqueous media. These nanoparticles are reversibly aggregated in response to changes in the pH of the solution. Furthermore, the potential of these AuNPs for the molecular recognition of enantiomers and for resolving the racemic mixtures when supported on a layered double hydroxide (LDH) was investigated.

Results and Discussion

The class="Chemical">imidazolium zwitterions derived from the amino acids n class="Chemical">glycine (1), rac-alanine (2), l-phenylalanine (3), and rac-methionine (4) were prepared following the methodology previously reported[28] (Figure ). Additionally, S,S-4 and R,R-4 enantiomers were synthesized from S- and R-methionine, respectively.

Figure 1

Imidazolium zwitterions 1–4.

class="Chemical">Imidazolium zwitterions 1–4.

To prepare the class="Chemical">AuNPs, n class="Chemical">NaBH4 was added to HAuCl4 solution, and an imidazolium zwitterion (1, 2, 3, or 4) was subsequently added to the mixture. For 1–3, the order of addition of the ligand and NaBH4 did not affect the results; however, for 4, if the ligand was added before NaBH4, a yellow precipitate was formed immediately. To ascertain the reason for these different results, HAuCl4 and the imidazolium zwitterions (1–4) were reacted in a 1:1 ratio. Thus, when compound 1, 2, or 3 was used, the corresponding [imidazolium][AuCl4–] adduct was formed. The adduct with 2 (2-H+·AuCl4–) was fully characterized using nuclear magnetic resonance (NMR) (see Figure S1), elemental analysis, and single-crystal X-ray diffraction (XRD) studies (Figure ); the diffracted crystal contained a mixture of diastereomers R,S-2-H+·AuCl4– and S,S-2-H+·AuCl4– in a 64:36 ratio because of the use of d,l-alanine as a starting material for the synthesis of 2. This adduct 2-H+·AuCl4– crystallized in the monoclinic P21 space group. In the molecular structure, the two amino acid residues were located on different sides of the imidazolium planar ring. In the crystal structure, both carboxylic acid groups participated in the hydrogen bond interactions (Figure S2), forming rows running along the c axis with O...O distances of 2.634(14) and 2.747(11) Å for R,S-2-H+·AuCl4– and distances of 2.65(3) and 2.588(18) Å for the minor component. On the basis of these distances, the hydrogen bond could be categorized as strong (2.50–2.65 Å).[29] By contrast, the reaction with 4 resulted in the immediate formation of a yellow precipitate (4-H+·AuCl4–) that was isolated and dried under a negative pressure. This solid was unstable, and eventually, metallic gold was deposited on the vessel walls. The solid obtained with the S,S-4 enantiomer (S,S-4-H+·AuCl4–) and the S,S-4 ligand were characterized using carbon nuclear magnetic resonance (13C NMR), cross-polarization/magic-angle spinning (CP/MAS), thermogravimetric analysis (TGA), and elemental analysis. The 13C NMR CP/MAS spectrum of S,S-4 displayed doublet signals because of the lack of symmetry of the two methionine sides of the ligand (Figure a); this can be explained on the basis of solid-state 13C NMR analysis because by this spectroscopic technique, it is possible to detect two different molecular environments over the methionine carbons on both sides of the ligand molecule.

Figure 2

Single-crystal X-ray structure of R,S-2-H+·AuCl4– (45% probability ellipsoids). Selected bond distances (Å) and bond angles (°) are Au1–Cl1, 2.273(5); Au1–Cl3, 2.271(5); Au1–Cl2, 2.2687(19); Au1–Cl4, 2.2730(19); N1–C4, 1.383(12); N1–C6, 1.324(14); C5–N2, 1.361(13); N2–C6, 1.328(12); C4–C5, 1.341(15); C4–N1–C1, 126.0(12); C6–N1–C1, 124.5(11); C6–N1–C4, 109.4(11); C5–N2–C7, 123.5(10); C6–N2–C5, 108.2(9); and C6–N2–C7, 128.3(11).

Figure 3

13C CP/MAS NMR spectrum 150 MHz of (a) the S,S-4 ligand and (b) S,S-4-H+·AuCl4– and their corresponding proposed structures with the molecular formula C26H40Au4Cl10N4O8S4 based on the elemental analysis and the TGA results.

Single-crystclass="Chemical">al X-ray structure of R,n class="Chemical">S-2-H+·AuCl4– (45% probability ellipsoids). Selected bond distances (Å) and bond angles (°) are Au1–Cl1, 2.273(5); Au1–Cl3, 2.271(5); Au1–Cl2, 2.2687(19); Au1–Cl4, 2.2730(19); N1–C4, 1.383(12); N1–C6, 1.324(14); C5–N2, 1.361(13); N2–C6, 1.328(12); C4–C5, 1.341(15); C4–N1–C1, 126.0(12); C6–N1–C1, 124.5(11); C6–N1–C4, 109.4(11); C5–N2–C7, 123.5(10); C6–N2–C5, 108.2(9); and C6–N2–C7, 128.3(11).

class="Chemical">13C CP/MAS n class="Chemical">NMR spectrum 150 MHz of (a) the S,S-4 ligand and (b) S,S-4-H+·AuCl4– and their corresponding proposed structures with the molecular formula C26H40Au4Cl10N4O8S4 based on the elemental analysis and the TGA results.

Thus, the molecular asymmetry is the consequence of a fixed and stable zwitterion on one class="Chemical">methionine side chain of the molecule in comparison with the other uncharged side, as shown in Figure a, whereas broad bands were observed in the S,n class="Chemical">S-4-H+ AuCl4– spectrum (Figure b) because of the increased anisotropy caused by the AuCl4– association. For both of the compounds (Figure a,b), the chemical shift patterns were similar, except for those of the −CH2–CH2–S–CH3 fragment. For S,S-4, the 13C chemical shifts of −CH2–CH2–S–CH3 were 32.3, 30.5, and 13.74 ppm, whereas for S,S-4-H+·AuCl4–, the CH2–CH2–S signals were shifted to a single signal at 36.79 ppm and the S–CH3 signal appeared at 24.09 ppm (Figure b). The downfield chemical shift of the S,S-4-H+·AuCl4– signals was due to the coordination of the sulfur atoms to a gold atom, which is consistent with previous observations for a gold complex with a sulfur ligand.[30]

According to the studies using TGA, S,class="Chemical">S-4 decomposed at 286.94 °C (Figure S3). The S,n class="Chemical">S-4-H+·AuCl4– TGA results showed a weight loss of 15.08% at 167.37 °C and a second weight loss of 21.04% at 278.05 °C (Figure S4). On the basis of the similarity between the S,S-4 decomposition temperature and the second S,S-4-H+·AuCl4– temperature, the total amount of organic material in S,S-4-H+·AuCl4– was determined to be 36.12%. This result and the results of elemental analysis (C, 17.25; H, 2.29; N, 3.51%) were used to propose the structure shown in Figure b, which includes Au–Au interactions that are typically observed in gold complexes.[31,32] The structural differences between 2-H+·AuCl4– and S,S-4-H+·AuCl4– demonstrate the effect of the order of addition of the reagent on the 4-AuNP synthesis.

The class="Chemical">AuNPs were stabilized by 1–3 in solution for a couple of days and exhibited an absorption band at 523 class="Chemical">nm (Figure S5). n class="Chemical">Nevertheless, the AuNPs stabilization was enhanced by 4, which kept them dispersed in solution for months at pH > 5.5, and exhibited a well-defined absorption band at 517 nm. Furthermore, no gold was deposited on the vial walls and no precipitate formation indicating destabilization was observed. The transmission electron microscopy (TEM) images revealed that the 4-AuNPs were spherical with an average particle size of 11.21 ± 0.07 nm (Figure ).

Figure 5

HRTEM images (a and b) and their corresponding fast Fourier transform (FFT) (c) of the S,S-4-AuNPs at pH = 3.0. The yellow lines in d, e, and f show the distance between two nanosphere surfaces, which most likely corresponds to the distance between the two ligands interacting via hydrogen bonds.

The class="Chemical">4-AuNPs are sensible to pH changes by precipitating from solution as a blue solid at pH < 3 and redispersing to form a n class="Chemical">ruby red solution at pH > 5.5. This reversible process could be performed at least 15 times without changing the particle morphology or size, as shown by the TEM and ultraviolet–visible (UV–vis) spectroscopy results. A single sharp band was observed at 517 nm at pH > 5.5, and an extra band appeared at approximately 650 nm when the nanoparticles began to aggregate when the pH value approached 3 (Figure ). This AuNP redispersion process was previously reported for the AuNPs stabilized by carboxymethyl cellulose,[33] carboxylate-modified polyvinylpyrrolidone,[12] thiols,[34] or NHC[35] among other stabilizants. The aggregation behavior of the AuNPs synthesized with 1–3 could not be reversibly controlled by changing the pH.

Figure 4

UV–vis spectra demonstrating the reversible aggregation of the 4-AuNPs controlled by changing the pH from 8.0 to 3.0. The new peak at 650 nm indicates nanoparticle aggregation.

UV–vis spectra demonstrating the reversible aggregation of the class="Chemical">4-AuNPs controlled by changing the pH from 8.0 to 3.0. The class="Chemical">new peak at 650 class="Chemical">nm indicates class="Chemical">nanoparticle aggregation.

The characterization of the S,class="Chemical">S-4-H+·n class="Chemical">AuCl4– complex showed that 4 preferred to coordinate with the gold atoms through the sulfur atoms. Therefore, it was assumed that the sulfur atoms were bonded to the AuNP surface, leaving carboxylate groups exposed on the 4-AuNP surface. These carboxylate groups were susceptible to protonation/deprotonation when the pH was modified, enabling the formation and breaking of the hydrogen bonds between ligands in a reversible process. For pH < 5.5, aggregation is observed because hydrogen bonds are formed between the carboxylic acid moieties in both methionine side chains of the ligand. These functional groups can promote the formation of dimeric hydrogen bonds between the ligands from some other nanoparticles. Moreover, when the nanoparticles were treated under alkaline pH, redispersion of the 4-AuNPs occurred because of the formation of carboxylates in the ligands, allowing electrostatic stabilization. By contrast, for the (1–3)-AuNPs, the electrostatic interaction between the carboxylate groups of the ligands and the gold surface is disrupted in an acidic media because of the protonation of the carboxylic groups, and consequently, the (1–3)-AuNPs aggregated irreversibly.

To gain a greater insight into this controlled aggregation behavior, high-resolution transmission electron microscopy (HRTEM) anclass="Chemical">alysis of the n class="Chemical">S,S-4-AuNPs was performed at pH 3.0, 5.0, and 8.0. At pH 8.0 and 5.0, the S,S-4-AuNPs formed aggregates with nanoparticles stacked on top of each other due to sample concentration and sample dehydration. Interestingly, at pH 3.0 (Figure ), the spherical nanoparticles were arranged with an average distance of 1.43 nm between them. On the basis of the crystallographic data previously reported for 4,[28] this distance is close to the distance between two 4 ligands interacting via hydrogen bonds through their carboxylic acid groups (the distance between the sulfur atom and the carboxyl oxygen atom in 4 is 0.63 nm).[28]

HRTEM images (a and b) and their corresponding fast Fourier transform (FFT) (c) of the class="Chemical">S,S-4-AuNPs at pH = 3.0. The yellow lines in d, e, and f show the distance between two class="Chemical">nanosphere surfaces, which most likely corresponds to the distance between the two ligands interacting via n class="Chemical">hydrogen bonds.

The class="Chemical">4-AuNP precipitation/redispersion process allowed nanoparticles purification after their synthesis. Hence, after precipitating the 4-AuNP in an acidic medium, the solution was decanted and analyzed using atomic absorption spectroscopy (AAS). We found that this solution contained just 2% of the total Au present in the initial solution indicating that the AuNP yield was 98%. Additionally, the solution was evaporated, and the resultant slightly yellow hygroscopic solid (4′) was analyzed using 13C and 1H NMR (Figure S6). The 13C NMR chemical shifts (Table S1) were similar to those of 4 (e.g., 30.12, 29.11, and 14.03 ppm for CH2–CH2–S–CH3) because of an excess of 4 that remained after the synthesis. Three additional signals were observed at 24.93, 43.22, and 36.95 ppm and were similar to those reported by Deslandes et al.[36] for the CH2–CH2–S(O)–CH3 fragment of sulfoxide N-methyl-l-methionine (25.6, 50.6, and 39.2 ppm), which indicates the presence of oxidized 4 and suggests that the ligand participates directly in the formation of AuNPs. Additionally, a wet precipitated 4-AuNP sample was analyzed using TGA (Figure S7); a weight loss of 1.95% (water loss) and another weight loss of 3.67% at 289.43 °C (a similar decomposition of 4) were determined. This result allows us to determine the millimolar ratio of the ligand-Au in the 4-AuNPs, which is 1:4.

S,class="Chemical">S-4 and R,R-4 were used to synthesize n class="Chemical">AuNPs according to the methodology described previously, and the chiral environment conferred on the R,R-4-AuNP and S,S-4-AuNP metal surfaces by these ligands was analyzed using CD (Figure a). The maximum and minimum absorption bands were observed at approximately 210 nm, which is consistent with the CD curves of the R,R-4 and S,S-4 ligands (Figure b). This result confirmed the chiral nature of the nanoparticle surfaces; therefore, the ability of these nanoparticles to achieve a molecular recognition of chiral substrates was studied.

Figure 6

CD spectra of (a) R,R-4-AuNPs and S,S-4-AuNPs and (b) R,R-4 and S,S-4.

CD spectra of (a) class="Chemical">R,R-4-AuNPs and n class="Chemical">S,S-4-AuNPs and (b) R,R-4 and S,S-4.

The interactions of class="Chemical">R,R-4-AuNPs and n class="Chemical">S,S-4-AuNPs with R and S enantiomer substrates were also analyzed using CD. The following substrates were selected based on their ability to form hydrogen bonds with the ligand through carboxylic groups on the AuNP surface: alanine, phenylalanine, proline, N-acetyl alanine, N-acetyl phenylalanine, prolinol, 2-aminobutanol, sec-butylamine, methylbenzylamine, and 2-butanol. First, the CD spectra of the R and S substrates were obtained. Then, the CD spectra of the [R,R-4-AuNPs + R-substrate], [R,R-4-AuNPs + S-substrate], [S,S-4-AuNPs + R-substrate], and [S,S-4-AuNPs + S-sustrate] mixtures were acquired for each substrate. Finally, the differences between the absorption spectra of the pure enantiomers and the mixtures were analyzed. For example, in the case of prolinol, the CD ellipticity absolute values of the [R,R-4-AuNPs + R-prolinol] and [S,S-4-AuNPs + S-prolinol] mixtures increased relatively to those for R- and S-prolinol, respectively (Figure a, left). The CD ellipticity absolute values of the other two mixtures also increased relatively to those of the substrate, but the increase was smaller. Figure a (right) shows the difference between the spectra of the mixtures with the strongest interactions and the corresponding pure enantiomers, that is, the net interaction. The symmetry observed in these differences suggests that the AuNP–substrate interactions were the same in magnitude. Similar results were observed for the alanine enantiomers; the absolute intensities of the [R,R-4-AuNPs + R-alanine] and [S,S-4-AuNPs + S-alanine] CD spectra were higher than those of pure alanine and the other two mixtures (Figure b). Unlike the above substrates, for 2-aminobutanol, the CD absolute intensities of the [R,R-4-AuNPs + S-2-aminobutanol] and [S,S-4-AuNPs + R-2-aminobutanol] mixtures were lower than those of R- and S-2-aminobutanol, respectively (Figure c). Further, the maximum and minimum intensity bands of the aminobutanol mixtures underwent a red shift of 8.0 nm relative to those of the pure enantiomers (198 nm). This behavior was also observed for the other two mixtures; nonetheless, the difference in wavelength between the mixtures and the pure enantiomers was smaller. For the other substrates, the CD absorption spectra of the mixtures and the pure enantiomers were the same, suggesting that these substrates did not interact with the AuNP molecular coating. Interestingly, the three substrates that interacted with the coated nanoparticles have two functional groups that are capable of forming hydrogen bonds; they also display low steric hindrance and flexible structures; therefore, they could readily form hydrogen bonds with the AuNP carboxylic ligands.

Figure 7

CD spectra of the R,R-4-AuNP and S,S-4-AuNP mixtures with the (a) prolinol, (b) alanine, and (c) 2-aminobutanol enantiomers. The plots on the left show the spectra of the [R,R-4-AuNPs + R-substrate], [R,R-4-AuNPs + S-substrate], [S,S-4-AuNPs + R-substrate], and [S,S-4-AuNPs + S-substrate] mixtures and the pure enantiomers. The plots on the right show the difference between the spectra of the mixtures with the best interactions and the corresponding pure enantiomers.

CD spectra of the class="Chemical">R,R-4-AuNP and n class="Chemical">S,S-4-AuNP mixtures with the (a) prolinol, (b) alanine, and (c) 2-aminobutanol enantiomers. The plots on the left show the spectra of the [R,R-4-AuNPs + R-substrate], [R,R-4-AuNPs + S-substrate], [S,S-4-AuNPs + R-substrate], and [S,S-4-AuNPs + S-substrate] mixtures and the pure enantiomers. The plots on the right show the difference between the spectra of the mixtures with the best interactions and the corresponding pure enantiomers.

Becclass="Chemical">ause of the affinity between the n class="Chemical">4-AuNPs and prolinol, alanine, and 2-aminobutanol, the ability of these nanoparticles to discriminate between the enantiomers in a racemic mixture was studied. Accordingly, aliquots of racemic substrate mixtures were added to R,R-4-AuNP and S,S-4-AuNP solutions, and the resultant mixtures were analyzed using CD. For example, 10 μL aliquots of rac-2-aminobutanol solution (211 mM) were added to a vial containing 0.5 mL of S,S-4-AuNP solution (1.12 mM, based on gold atoms). The mixture was stirred for 1 minute after each addition, and then analyzed using CD. The results of the titration are shown in Figure . Then, 60 μL (1.3 × 10–2 mmol) of the rac-2-aminobutanol solution was added to the nanoparticle solution, a minimum band appeared (190–195 nm), indicating that the S,S-4-AuNPs and R-2-aminobutanol started to interact, and the solution became enriched with S-2-aminobutanol. The titration was stopped after adding 110 μL (2.1 × 10–2 mmol) because the absorption spectra did not change because of the saturation of the chiral centers on the nanoparticle surface. On the basis of these results, it was determined that the maximum interaction between the S,S-4-AuNPs and R-2-aminobutanol detected using CD spectroscopy occurred at a nanoparticle/substrate molar ratio of 42:1.

Figure 8

CD spectra of the mixtures of 0.5 mL of the S,S-4-AuNPs (1.12 mM) and 10–110 μL of rac-2-aminobutanol solution (1.1–2.1 × 10–2 mmol).

CD spectra of the mixtures of 0.5 mL of the class="Chemical">S,S-4-AuNPs (1.12 mM) and 10–110 μL of rac-n class="Chemical">2-aminobutanol solution (1.1–2.1 × 10–2 mmol).

To study the potenticlass="Chemical">al of these class="Chemical">nanoparticles to resolve the racemic mixtures, they were supported on an inorganic matrix. Aires et n class="Chemical">al.[37] reported the synthesis and stability of AuNPs supported on LDHs, and Ballarin et al.[38] reported a high affinity between an LDH and AuNPs. Hence, an LDH [Zn5Al2(OH)14CO3] was synthesized according to the coprecipitation methodology of Carbajal-Arízaga and Jiménez[39] and characterized using XRD and attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy. Then, the LDH was added to the AuNP solution, resulting in the loss of color in the solution and change in the LDH color from white to pink. This process was repeated until the solid was completely saturated, yielding a fine purple powder after drying (LDH-4-AuNPs). AAS experiments showed that 1.0 mg of gold was supported on 20 mg of dry LDH. XRD data indicated that the LDH structure presented a basal distance (related to separation of layers) equal to 7.45 Å before and after the impregnation with the AuNP solution (Figure S8); thus no intercalation of AuNP or the ligand occurred. The analysis using TEM (Figure ) revealed a nonhomogeneous AuNP distribution on the LDH surface.

Figure 9

TEM (a and b) and scanning transmission electron microscopy (STEM) images (c) for the LDH-S,S-4-AuNPs. This reveals a nonhomogeneous AuNPs distribution on the LDH surface.

TEM (a and b) and scanning transmission electron microscopy (STEM) images (c) for the class="Chemical">LDH-n class="Chemical">S,S-4-AuNPs. This reveals a nonhomogeneous AuNPs distribution on the LDH surface.

On the other hand, a simple method was not found to determine the enantiomeric excess (ee) through CD. Therefore, the ee was obtained from the cclass="Chemical">alibration curves of the pure n class="Chemical">alanine and 2-aminobutanol enantiomers. Hence for alanine, 1.4–4.3 mg/mL aqueous solutions of each enantiomer were prepared. The maximum and minimum CD ellipticity values were observed at 204 nm for R- and S-alanine (Figure S9a). The sample concentrations were linearly correlated with the ellipticity values by eqs and 2 for S- and R-alanine, respectively. Similarly, calibration curves were obtained for the 2-aminobutanol enantiomers over the concentration range of 0.5–2 μL/mL. The maximum and minimum CD ellipticity values were observed at 193 nm for R- and S-2-aminobutanol (Figure S9b), and the R and S enantiomer concentrations were linearly correlated with the ellipticity values by eqs and 4, respectively. The calibration curves were validated using mixtures of known concentrations, confirming their usefulness (Tables S2 and S3).

To resolve the racemic mixtures, two procedures were explored. In the first method (P-1), 100 mg of the chirclass="Chemical">al n class="Chemical">LDH-4-AuNPs was packed in a Pasteur pipette. Water was used as an eluant to achieve uniform packing. Then, 0.4 mL of 0.50 M racemic solution was flowed through the column, and 0.5 mL aliquots (rac-alanine and rac-2-aminobutanol in water and rac-prolinol in methanol) were collected. The collected fractions were dried, and 1 mg for alanine or 1 μL for 2-aminobutanol and prolinol was subsequently transferred into a 1 mL volumetric flask and diluted to the given volume with water. In the second method (P-2), 150 mg of the chiral LDH-4-AuNPs and 10 mg of rac-alanine or 10 μL of the other substrates were stirred in 1 mL of water for 1 h, and the sample was decanted. The decanted liquid was evaporated, and 2.0 mg for alanine or 1.0 μL for 2-aminobutanol and prolinol was dissolved in 1 mL of water for CD analysis. This procedure was repeated five times.

After performing P-1 and P-2 experiments with class="Chemical">rac-prolinol, the solutions darkened (becn class="Chemical">ause of a possible oxidation); therefore, this substrate was not studied further. For rac-alanine, fractions enriched in one of the enantiomers were not obtained after P-1; however, good results were observed after P-2. Figure a shows the results for the LDH-S,S-4-AuNPs. The spectra of the initial fraction showed an ellipticity value of approximately −7.5 mdeg, which corresponds to 17.0% ee of R-alanine, based on eq . This result agrees with the previous experiments where an affinity between the S,S-4-AuNPs and S-alanine was observed. This interaction led to the production of the enriched R-alanine fraction. By contrast, when the LDH-R,R-4-AuNPs were used, fractions enriched in S-alanine were observed, displaying an ee of 10%, based on eq (Figure b). These ee values were corroborated by using a quiral high-performance liquid chromatography (HPLC) technique according to the recommended procedure for using a Chiralcel OD-H column,[40] and they were consistent with the CD quantitative results (Figure S10, Table S4). For the resolution of rac-2-aminobutanol, the P-1 process led to better results than the P-2 method. Figure c shows the results obtained using a column packed with the LDH-R,R-4-AuNPs. The second and third fractions were enriched in R-2-aminobutanol and had an ee of 12.0%, based on eq . This result is consistent with the earlier experiments in which an affinity between the R,R-4-AuNPs and S-2-aminobutanol was observed. This interaction enabled the elution of samples enriched in R-2-aminobutanol.

Figure 10

CD spectra of (a) R-alanine-enriched fractions obtained from the interaction of rac-alanine with the LDH-S,S-4-AuNPs, (b) S-alanine-enriched fractions obtained from the interaction of rac-alanine with the LDH-R,R-4-AuNPs, and (c) R-2-aminobutanol-enriched fractions obtained from the interaction of rac-2-aminobutanol with the LDH-R,R-4-AuNPs.

CD spectra of (a) class="Chemical">R-alanine-enriched fractions obtained from the interaction of n class="Chemical">rac-alanine with the LDH-S,S-4-AuNPs, (b) S-alanine-enriched fractions obtained from the interaction of rac-alanine with the LDH-R,R-4-AuNPs, and (c) R-2-aminobutanol-enriched fractions obtained from the interaction of rac-2-aminobutanol with the LDH-R,R-4-AuNPs.

Experimental Section

General Information

class="Chemical">All chemicn class="Chemical">als were obtained from Sigma Aldrich and used without further purification. NaBH4 and HAuCl4·3H2O were kept under an N2 atmosphere to prevent hydration. All glassware and magnetic stirrer bars used to synthesize the AuNPs were washed with aqua regia (HCl/HNO3 = 3:1) to avoid unwanted nucleation during the synthesis.

The morphology and size of the class="Chemical">AuNPs were characterized by HRTEM using a FEI TECn class="Chemical">NAI F30 STWIN G2 instrument operated at 300 kV. UV–vis absorption spectra were recorded at room temperature using a Thermo Scientific Genesys 10S spectrometer. CD spectra were recorded on a Jasco J-1500 spectrometer using a 1 mm rectangular quartz cuvette. A Waters 600 pump equipped with a Rheodyme 7725i injector and a Waters 2996 photodiode array detector was used to perform chromatographic experiments with a Chiralcel OD-H column (250 mm L × 4.6 mm I.D.). 1H and 13C NMR spectra were obtained on a JEOL ECA 600 spectrometer of 14 T magnet field (600 MHz).

AAS experiments were performed using a Varian SpectrAA 220 spectrometer. For the gold anclass="Chemical">alysis, a 4 mA hollow-cathode lamp with a wavelength of 267.6 class="Chemical">nm was used. The TGA was performed using a Discovery TGA equipment from TA Instruments with a temperature gradient of 10 °C/min under a n class="Chemical">nitrogen flow.

Powder XRD patterns were obtained on a Siemens D500 system. The samples were scanned using Cu Kα source over the 2θ range from 5 to 60° at a rate of 0.02°/s using a current of 20 mA and a voltage of 30 kV.

Single crystclass="Chemical">al X-ray diffraction data of n class="Chemical">2-H+·AuCl4− were collected at room temperature on an Oxford Diffraction Gemini CCD diffractometer with graphite-monochromated Mo Kα radiation (λ = 0.71073 Å). The data were integrated, scaled, sorted, and averaged using CrysAlis software package. By using Olex2,[41] the structures were solved with the ShelXS structure solution program using Direct Methods and refined with the ShelXL[42] refinement package using Least Squares minimization. All nonhydrogen atoms were refined anisotropically. The positions of the hydrogen atoms were kept fixed with a common isotropic displacement parameter; the hydrogen atoms linked to the oxygen atoms were located on the Fourier map. A summary of the crystal data and the refinement parameters for the structural analyses is given in Table S5. The R,S and S,S diastereomers were observed in the cell in a 64:36 ratio. The [AuCl4]− anion displayed positional disorder over two positions, with the main component having an occupancy of 0.641(9). The disorder was treated with rigid bond restraints by using instructions such as SAME, SIMU, and DELU (For a detailed description of the restraints, see Müller, Peter. “Disorder” Crystal Structure Refinement. A Crystallographer’s Guide to SHELXL. Edited by P. Müller. New York: Oxford University Press, 2006, 56−96).

General Procedure for the Synthesis of 1–4

Ligands 1–4 were prepared according to our previously reported procedures.[28]class="Chemical">l- and d-methionine were used to prepare S,n class="Chemical">S-4 and R,R-4, respectively. For these enantiomers, the optical values were −18.00 ± 0.06 for R,R-4 and +18.00 ± 0.06 for S,S-4, which were obtained from a solution containing 3 g/100 mL of R,R-4 or S,S-4 in NaOH (0.1 M) at 25 °C.

General Procedure for the Synthesis of AuNPs Stabilized by 1–4

A solution of 1.0 mL of class="Chemical">HAuCl4·3H2O (6.9 mM) was added to 20 mL of n class="Chemical">water under stirring at 1000 rpm. The mixture was heated to 40 °C, and 0.8 mL of NaBH4 solution (26.4 mM in water) was added. Then, 0.8 mL of an appropriate solution of 1–4 (4.0 mM) was immediately added to the mixture. The solution was stirred for 20 min at 40 °C and then cooled to room temperature.

The purification of class="Chemical">4-AuNPs, n class="Chemical">S,S-4-AuNPs, and R,R-4-AuNPs was carried out by adding 1.0 mL of HCl (0.1 M) to the solution, which changed the color from red to deep blue. The sample was centrifuged at 2500 rpm for 2 min. The solution was decanted, and the blue solid was redispersed by adding 5 mL of water and 50 μL of 0.1 M NaOH. The resultant solution developed an intense red color.

Synthesis of 2-H+·AuCl4–

Twenty milligrams of class="Chemical">HAuCl4·3H2O (5.1 × 10–2 mmol) and 10.8 mg of 2 (5.1 × 10–2 mmol) were added to 2.0 mL of n class="Chemical">water in a vial. The yellow solution was concentrated by evaporation at room temperature to obtain crystals. Yield: ≥99%. MP: 191.0 °C. 1H NMR (D2O, 600 MHz, δ ppm): 9.07 (s, N–CH–N), 7.63 (d, CH=CH), 5.32 (m, N–CH–CH2), 1.84 (d, CH3). 13C NMR (D2O, 150 MHz, δ ppm): 172.98 (C=O), 135.84 (N–CH–N), 121.83 (CH=CH), 58.23 (CH–CH3), 16.87 (CH3). ATR-FTIR (cm–1): 3146 (m) OH, 2958 (m) CH, 1723 (s) COOH, 1552 (s) imidazole ring vibration. Elemental analysis for [C9H13N2O4][AuCl4]: calculated: C, 19.62; H, 2.20; N, 5.08%. Found: C, 19.83; H, 2.40; N, 5.76%.

Synthesis of S,S-4-H+·AuCl4–

Twenty milligrams of class="Chemical">HAuCl4·3H2O (5.1 × 10–2 mmol) was added to 2.0 mL of n class="Chemical">water in a vial. Then, 17.0 mg of S,S-4 (5.1 × 10–2 mmol) dissolved in 5.0 mL of water was subsequently added to the mixture. Immediately, a yellow precipitate was formed. The mixture was filtered and dried under vacuum. Decomposition temperature: 153 °C. 13C NMR CP/MAS (150 MHz, δ ppm): 168.68 (C=O), 136.44 (N–CH–N), 122.37 (CH=CH), 61.61 (N–CH–CH2), 35.23 (CH2–CH2–S), 24.49 (S–CH3). ATR-FTIR (cm–1): 3366 (w) OH, 3138 (w) HC=CH, 2923 (w) CH, 1724 (s) COOH, 1553 (s) imidazole ring vibration, 1169 (s) C–S–C. Elemental analysis for C26H40N4O8S4Au4Cl10: calculated: C, 17.28; H, 2.23; N, 3.10%. Found: C, 17.25; H, 2.29; N, 3.51%.

Synthesis of the Zn/Al Carbonate LDH

class="Chemical">Al(NO3)3·9H2O (10.1 g, 26.92 mmol) and Zn(NO3)2·6H2O (20.0 g, 67.31 mmol) were dissolved in 500 mL of distilled water (molar ratio of Al/Zn = 2.5) and then precipitated by adding NH4OH (14.0%) to reach a final pH of 8.0–9.0. Then, 1.42 g of Na2CO3 (13.46 mmol) was added, resulting in a pH of approximately 10. The reaction mixture was stirred for 24 h (aging time). Subsequently, the reaction was allowed to stand for 2 h. The solution was then decanted, and the solid was washed three times with water to get a neutral pH.

The obtained precipitate was transferred to a porcelain dish and dried in an oven at 70 °C for 12 h. The white solid was triturated in an agate mortar. The nominclass="Chemical">al molecular formula was cn class="Chemical">alculated to be Zn5Al2(OH)14CO3 (yield = 92%). ATR-FTIR (cm–1): 3407.3 (s) OH, 1358.5 (s) CO3=.

CD Analysis for Molecular Recognition Studies

To prepare the mixtures of the class="Chemical">S,S-4-AuNPs or n class="Chemical">R,R-4-AuNPs and the S or R substrate (alanine, prolinol, and 2-aminobutanol), the following general procedure was followed: 0.3 mL of R,R-4-AuNP solution (1.12 mM, based on gold atoms) and 0.2 mL of the R substrate (500 mM) were placed in a vial. The mixture was diluted to 3 mL with distilled water. The same quantities of R,R-4-AuNPs, S substrate, and water were added to another vial. This procedure was repeated for the S,S-4-AuNPs and R and S substrates. The pH of all of the solutions was maintained between 7.8 and 8.1 to prevent aggregation of the AuNPs. The mixtures were stirred for 15 min. For the CD analysis, 0.2 mL of each chiral substrate was diluted to 3 mL with distilled water.

AuNPs Supported on the LDH

Six milliliters of class="Chemical">R,R-4-AuNP or n class="Chemical">S,S-4-AuNP solution (1.12 mM, based on gold atoms) was added to a flask containing 1.0 g of LDH (1.8 mmol). The solution subsequently lost its red color. The mixture was centrifuged (2000 rpm for 2 min), and the liquid was decanted. The obtained solid was added to a fresh AuNP solution, and the procedure was repeated until the solution remained red, indicating that the LDH was saturated. The final solid was dried under vacuum for 12 h to give a fine purple powder. The atomic absorption results indicated that 20 mg of dry LDH absorbed 1.0 mg of gold (5.1 × 10–3 mmol). ATR-FTIR (cm–1): 3379.2 (s) OH, 1352.0 (s) CO3=.

Enantioselective Separation of rac-2-Aminobutanol (P-1 Procedure)

A smclass="Chemical">all piece of sterile cotton and 100 mg of n class="Chemical">LDH-R,R-4-AuNPs or LDH-S,S-4-AuNPs were packed in a Pasteur pipette. Using 1 mL of water as eluant, 100 μL of rac-2-aminobutanol was passed through the column. Fractions of 0.5 mL were collected and dried. Aliquots of 1.0 μL were removed from the resulting residue and diluted to a final volume of 1 mL for CD analysis.

Enantioselective Separation of rac-Alanine (P-2 Procedure)

One milliliter of distilled class="Chemical">water, 100 mg of n class="Chemical">LDH-R,R-4-AuNPs or LDH-S,S-4-AuNPs, and 10 mg of rac-alanine were added to a vial. This reaction mixture was stirred for 1 h, and then, the solid was allowed to settle. The supernatant solution (200 μL) was taken and dried. This fraction contained 2.0 mg of alanine, which was diluted up to 1 mL with distilled water and analyzed using CD. Additionally, the settled solid was washed five times with 1 mL of water and these fractions were also analyzed using CD; the alanine residue was recovered in these fractions, showing no ee.

Calibration Curves of the Alanine and 2-Aminobutanol Enantiomers

For each of the pure enantiomers (class="Chemical">alanine and n class="Chemical">2-aminobutanol), a 0.5 M solution was prepared. For alanine, five aliquots in the range of 11.2–67.2 μL were taken and diluted to 1 mL (final concentration of 1.4–4.3 mg/mL, respectively). For 2-aminobutanol, five aliquots in the range of 25–150 μL were taken and diluted up to 1 mL (final concentration of 0.5–2 mg/mL, respectively). All samples were analyzed using CD. The maximum and minimum absorption ellipticity values were linearly correlated with the sample concentrations by eqs –4. The validity of the equations was confirmed by analyzing several enantiomer mixtures with known concentrations (Tables S2 and S3).

Chiral HPLC Analysis

The protection of the class="Chemical">alanine samples with n class="Chemical">carbobenzoxy (Cbz) was necessary and was made according to the methodology reported by A. Pesic et al.[43] (who reported a yield of 98%) and corroborated by 1H-NMR before the HPLC injection. The analysis was carried out following the procedure for Chiralcel OD-H column specifications for N-Cbz-alanine.[40] The integrity of the stereocenter for the chiral N-Cbz-amino acid was maintained as determined using HPLC.

Conclusions

In conclusion, we obtained class="Chemical">AuNPs stabilized by n class="Chemical">imidazolium salts in an aqueous medium. The 4-AuNPs showed prolonged stability in water and reversible aggregation as the pH was modified, ensuring a simple and effective purification. Additionally, the synthesis of AuNPs with R,R-4 or S,S-4 gave a chiral environment to the metal surface with two stereogenic centers in each ligand, which could be used to carry out the enantioselective recognition of the substrates that could form two hydrogen bonds with the carboxylate of the ligand on the surface metal. The chiral environment at the nanoparticle surface was used to get a low enrichment of racemic mixtures of alanine and 2-aminobutanol when the AuNPs were supported on the synthesized LDH according to the CD responses. These results open the possibility of using chiral AuNP systems in the enantiomeric separation.