Highly Stable Layered Double Hydroxide Colloids: A Direct Aqueous

Article pubs.acs.org/Langmuir

Highly Stable Layered Double Hydroxide Colloids: A Direct Aqueous Synthesis Route from Hybrid Polyion Complex Micelles Géraldine Layrac,† Mathias Destarac,‡ Corine Gérardin,*,† and Didier Tichit*,† †

Institut Charles Gerhardt, UMR 5253 CNRS/ENSCM/UM2/UM1, Matériaux Avancés pour la Catalyse et la Santé (MACS), Ecole Nationale Supérieure de Chimie, 8, rue de l’Ecole Normale, 34296, Montpellier Cedex 5, France ‡ Laboratoire Interactions Moléculaires et Réactivité Chimique et Photochimique, UMR 5623 CNRS-UPS Toulouse, Université Paul Sabatier Toulouse, 118 route de Narbonne, 31062 Toulouse Cedex 9, France S Supporting Information *

ABSTRACT: Aqueous suspensions of highly stable Mg/Al layered double hydroxide (LDH) nanoparticles were obtained via a direct and fully colloidal route using asymmetric poly(acrylic acid)-b-poly(acrylamide) (PAA-b-PAM) double hydrophilic block copolymers (DHBCs) as growth and stabilizing agents. We showed that hybrid polyion complex (HPIC) micelles constituted of almost only Al3+ were first formed when mixing solutions of Mg2+ and Al3+ cations and PAA3000-b-PAM10000 due to the preferential complexation of the trivalent cations. Then mineralization performed by progressive hydroxylation with NaOH transformed the simple DHBC/Al3+ HPIC micelles into DHBC/aluminum hydroxide colloids, in which Mg2+ ions were progressively introduced upon further hydroxylation leading to the Mg−Al LDH phase. The whole process of LDH formation occurred then within the confined environment of the aqueous complex colloids. The hydrodynamic diameter of the DHBC/LDH colloids could be controlled: it decreased from 530 nm down to 60 nm when the metal complexing ratio R (R = AA/(Mg + Al)) increased from 0.27 to 1. This was accompanied by a decrease of the average size of individual LDH particles as R increased (for example from 35 nm at R = 0.27 down to 17 nm at R = 0.33), together with a progressive favored intercalation of polyacrylate rather than chloride ions in the interlayer space of the LDH phase. The DHBC/ LDH colloids have interesting properties for biomedical applications, that is, high colloidal stability as a function of time, stability in phosphate buffered saline solution, as well as the required size distribution for sterilization by filtration. Therefore, they could be used as colloidal drug delivery systems, especially for hydrosoluble negatively charged drugs.

■

INTRODUCTION The preparation of highly stable aqueous colloidal suspensions of layered double hydroxide (LDH) particles on a nanometer length scale would dramatically increase the applications of these materials in several areas such as drug delivery, catalysis, and reinforcement of polymer materials. LDHs have the general formula [MII1−xMIIIx(OH)2]x+An−x/n·nH2O, where MII and MIII stand for a divalent and a trivalent cation octahedrally coordinated with hydroxyl groups sharing edges to form brucite-like layers. These edge-sharing octahedral units form infinite layers with the hydroxyl ions sitting perpendicular to the plane of the layers. In LDHs, a fraction of the divalent cations in the brucite lattice is isomorphically substituted by trivalent cations such that the layers acquire a positive charge, which is balanced by intercalation of anions between the layers (represented as An− in the general formula). Water molecules hydrogen-bonded to the anions are also found in the interlayer space. When the charge compensating anion in the LDH is a biologically active species, bio-LDH hybrids can be prepared, and the formation of stable suspensions of such nanoparticles as gene or drug delivery systems is highly desired due to the © 2014 American Chemical Society

unique properties of low cytotoxicity and high biocompatibility of the LDHs.1−4 Bio-LDH hybrids have been obtained by intercalation of DNA, ATP, or nucleosides, for instance.2 Used as nanofillers, LDH nanoparticle suspensions would improve properties in polymer reinforcement.5 If LDH suspensions were used as catalysts, the improved accessibility of the active sites would give rise to higher activities in many reactions.6−8 Small particle sizes associated with a low aggregation degree and a high colloidal stability of LDH aqueous suspensions are the required characteristics for further developing abovementioned applications. But LDH nanoparticles stable in aqueous medium are difficult to obtain by conventional preparation methods. Indeed, the long precipitation times and the growth of the first nuclei throughout the precipitation process of LDH generally lead to wide distributions of particle sizes.9 For example, controlled supersaturation conditions lead to particle sizes in the range 80−350 nm,10 which could be further narrowed down to 60−80 nm by separating nucleation Received: June 3, 2014 Revised: June 30, 2014 Published: August 3, 2014 9663

dx.doi.org/10.1021/la502159x | Langmuir 2014, 30, 9663−9671

Langmuir

Article

was obtained by simply varying the amount of copolymer per metal ion.26,29 The present paper deals with the synthesis of stable colloidal suspensions in water of Mg/Al LDH nanoparticles using asymmetric poly(acrylic acid)-b-poly(acrylamide) (PAA-bPAM) DHBC. The main objective was to overcome the main drawbacks previously pointed out when using the different coprecipitation methods and to obtain well-dispersed LDH nanoparticles of controlled size directly as aqueous colloids. The influence of the degrees of metal hydroxylation and of metal complexation on the properties of the intermediate micelles and of the LDH colloids were studied by dynamic light scattering (DLS), transmission electron spectroscopy (TEM), energy-dispersive X-ray (EDX) spectrometry, and X-ray diffraction (XRD). The formation of the LDH-based particles, the identification of the intermediate HPIC micelles, and their subsequent transformation by mineralization were particularly investigated.

and aging processes, but unfortunately particles remained highly aggregated.11 pH adjustment during the precipitation by addition of anion-exchange resin in the solution allowed obtaining hexagonal platy LDH particles with average size of 50 nm. However, the resin must be then removed by filtration12,13. Moreover, in the case of unaged primary crystals, a glue effect due to the presence of small amounts of amorphous phase is responsible for the aggregation of the particles, and aging treatments (necessary to improve the crystallinity) lead to large rod-shape aggregates due to face-to-face association of primary crystals.14 An interesting study reports LDH particles with an average size of about 10 nm synthesized in the presence of tripodal ligands but the strong interactions between the ligands and the layers probably hinder their removal.15 Postsynthesis hydrothermal treatments allowed particles to be obtained as monodisperse platelets (∼60 nm), however, arranged in liquid crystalline phases.14 Many attempts have been made to prepare nanocrystallites of individual LDH platelets in stable suspensions. Remarkably, O’Hare et al. reported that the “water pools” formed in reverse microemulsion systems (aqueous solutions of salts, oil, and surfactants) acted as nanoreactors allowing control of the LDH platelets size and aspect ratio by varying the molar ratio of water to surfactant. Addition of triblock copolymers at different stages of the synthesis allowed changing growth orientation and morphology (beltlike or rodlike structures) of LDH particles.16 They showed a control of the morphology due to particle formation in a confined environment, and obtained platelet dimensions exceeding 200 nm. Improvement of the reverse microemulsion approach was obtained using oleylamine, leading to LDH nanoparticles with sizes in the range 40−80 nm.17 Stable colloidal suspensions of single LDH layers were also obtained by delamination using either organophilic anions/ alcohols or amino acid/polar solvent combinations.18,19 Delamination in water was scarcely performed, except when lactate anions,20 short-chain aliphatic carboxylates, or organic sulfonates21,22 were previously intercalated. Also, Sasaki et al. have successfully delaminated Co/Al LDH intercalated by nitrate anions using formamide solvent.23 Li et al. obtained LDH colloids sterically stabilized by polyethylene glycol (PEG) chains terminated by sulfate or carboxylate functions.24 The surrounding polymer shell around the preformed LDH particles led to hydrodynamic radii of dispersed PEG/LDH nanocomposites in the range 200−500 nm. Unfortunately, the previously cited routes imply either the use of organic solvents or fail in controlling the particle sizes below 200 nm in water and in maintaining a good degree of dispersion. This makes them unsuitable for the direct preparation of stable aqueous suspensions of LDH nanoparticles. The above-mentioned drawbacks were recently overcome in the case of the preparation of metal (hydr)oxides, for example, of Al3+, La3+, Cu2+, Ni2+, Zn2+, and Ca2+ or metal sulfides25 by performing nanoparticle syntheses in the presence of double hydrophilic block copolymers (DHBCs) constituted of a metalcomplexing block (e.g., poly(acrylic acid) (PAA)) and a stabilizing block (e.g., polyacrylamide or poly(ethylene oxide)).26 With this strategy, hybrid polyion complex (HPIC) micelles formed upon complexation of multivalent metal cations by the ionizable complexing block of DHBC polymers and stabilized by their neutral blocks27 act as nanoreactors for subsequent mineralization reactions.28 Highly stable colloids of metal hydrous oxides were thus prepared. Moreover, remarkable control of the particle sizes and of the stability

■

MATERIALS AND METHODS

Sample Preparation. The DHBC used was a poly(acrylic acid)-bpoly(acrylamide) (PAA-b-PAM) where the PAA metal-binding block is smaller than the PAM neutral block. The number-averaged molecular weights (Mn) of the PAA and PAM blocks are 3000 and 10 000 g/mol, respectively, and are indicated close to each block name; the polymer is then named as follows: PAA3000-b-PAM10000. This DHBC was prepared according to a well-established sequential polymerization of acrylic acid and acrylamide performed in a hydroalcoholic medium in the presence of a xanthate RAFT/ MADIX transfer agent.22 MgCl2·6H2O and AlCl3·6H2O (Sigma-Aldrich) were used as sources of metal cations. Deionized water (Purelab, Elga, France) was always used for the preparation of the solutions. The synthesis of colloidal suspensions of hybrid LDH nanoparticles proceeds in two steps, that is, micellization and mineralization steps. Micelle Formation. An aqueous solution of MgCl2·6H2O and AlCl3·6H2O (molar ratio Mg/Al = 2) was prepared. Separately, an aqueous solution of PAA3000-b-PAM10000 at a concentration of 6.4 10−4 mol/L in acrylate function was prepared, its pH was adjusted at 5.5, which corresponds to a degree of dissociation of the acrylic acid groups of 50%. The two solutions were then mixed. The amount of PAA3000-b-PAM10000 added to the cation solution depends on the complexation ratio R, which is defined as the molar ratio of acrylate groups (AA) per total amount of cation (R = AA/(Mg + Al)); R was varied from 0 to 1. The Mg and Al concentrations in the micelle suspensions are maintained constant at [Mg2+] = 10−2 mol/L and [Al3+] = 5.10−3 mol/L whatever the R values. The solutions were stirred for 10 min before the mineralization step. Mineralization by Progressive Hydroxylation of a Mg−Al Cation Solution in the Absence and in the Presence of DHBC Copolymer (R = 0.8). Without DHBC. A volume of 5 mL of cation solution (Mg/Al = 2; [Mg2+] = 10−1 mol/L) was titrated with NaOH (0.4 M). The hydroxylation degree h is defined as the molar ratio of OH ions per cation (h = OH/(Mg + Al)). h was varied from 0 to 2.6. With DHBC. A volume of 5 mL of cation solution (Mg/Al = 2; [Mg2+] = 10−2 mol/L) containing the DHBC polymer (R = 0.8) was titrated with NaOH (0.4 M). h was varied from 0 to 2.6. Mineralization by Coprecipitation at Constant pH = 10 of Mg− Al Cation Solutions with Varying Amounts of DHBC Copolymer. LDH formation was induced by cohydroxylation of Mg2+ and Al3+ at a constant pH of 10 upon addition of increasing amounts of NaOH (0.2 mol/L) under air atmosphere. A 5 mL solution of cations (Mg/Al = 2; [Mg2+] = 10−2 mol/L) containing the DHBC polymer in varying amounts (R varied from 0 to 1) was added at a feeding rate of 0.4 mL/ min in a 5 mL aqueous solution, whose pH was maintained at 10 by addition of NaOH (0.2 mol/L). The colloidal suspensions were then stored at ambient temperature in closed flasks. 9664


Langmuir

Article

step. Formation of the HPIC micelles was first investigated with mixtures of Mg2+ and Al3+ cations (molar ratio Mg/Al = 2) and PAA3000-b-PAM10000 copolymers at increasing complexation degree R varied from 0 to 1 (R = AA/(Mg + Al)). Then, the mineralization step was specifically investigated on a mixture of cations and DHBC polymer prepared at R = 0.8, which was hydroxylated by addition of increasing amounts of NaOH (h was varied from 0 to 2.6). Finally, the formation of LDH colloids at varying R ratios was studied by mineralization at constant pH = 10 of DHBC/Mg−Al solutions. Micellization Step. HPIC micelles were characterized first by DLS measurements; the hydrodynamic diameter (Dh) and the polydispersity index (PDI) of the scattering objects and the scattered intensity I (normalized by the copolymer amount) are reported in Table 1. For comparison, DLS measurements were

Preparation of a Reference PAA Homopolymer/LDH Sample by Coprecipitation at Constant pH = 10. A 50 mL aqueous solution containing 2.925 g of MgCl2·6H2O and 1.737 g of AlCl3·6H2O were added drop by drop to a sodium polyacrylate solution (5000 g/mol, Sigma-Aldrich). The PAA amount corresponds to three times the LDH anionic exchange capacity. The pH is maintained at 10 by addition of a NaOH solution (2 mol/L). The product is put under reflux at 80 °C for a night, and then it is washed three times with distilled water. Dynamic Light Scattering. The dynamic light scattering (DLS) measurements were performed at 25 °C using a Malvern Autosizer 4800 spectrogoniometer, with a 50 mW laser source operating at 532 nm. The scattering measurements were done at a scattering angle of 90°. The measured normalized time correlation function was analyzed by using the CONTIN algorithm. The values of the hydrodynamic radii were calculated from the decay times using Stokes−Einstein equation: Dh = kBT/6πηD, where kB is the Boltzmann constant, T is the absolute temperature, η is the solvent viscosity, and D is the diffusion coefficient. The Dh values are intensity-averaged hydrodynamic diameters. Sterilizing Filtration. Suspensions were filtered through 0.22 μm cellulose mixed ester (CME) microfilters. They were quantitatively analyzed by DLS before and after filtration. X-ray Diffraction. Powder X-ray diffraction (XRD) patterns were recorded using a Bruker D8 Advance diffractometer and the monochromatic Cu Kα1 radiation (λα = 1.54184 Å, 40 kV and 50 mA). They were recorded with 0.02° (2θ) steps over the 2−70° 2θ angular range with 0.2 s counting time per step. The patterns were performed on powders obtained by evaporation of the colloidal suspensions on glass slides at 40 °C during 2 h. Elemental Analysis. Mg and Al elemental analysis of the samples was performed by EDX microprobe using a Cambridge Stereoscan 260 apparatus. Transmission Electron Microscopy. A volume of 100 μL of the colloid suspensions was diluted into 900 μL of water at pH 10. A drop was then placed on a carbon−copper grid. After a few minutes of evaporation in air at ambient temperature, the remaining solution was sucked with a filter paper to obtain a few particles on the grid. The sample was then characterized using a JEOL 1200 EXII microscope operated at 80 kV. The particle size distribution was determined using the measure It software.

Table 1. Hydrodynamic Diameter and Scattered Intensity of HPIC Micelles Formed with Multivalent Cations and PAA3000-b-PAM10000 at Different Complexation Ratios R samples 2+

DHBC/Mg DHBC/Al3+ DHBC/(Mg2+, Al3+)

DHBC alone

R

Dh (nm)

I/[DHBC]

PDI

2 3 0.53 0.66 0.80 1.00 2.33

8 28 16 21 20 21 24 6

1403 10612 4290 7820 5490 4370 11502 574

0.27 0.17 0.14 0.23 0.27 0.24 0.26 0.45

also performed on mixtures of PAA3000-b-PAM10000 with either Mg2+ or Al3+ cations alone using in each case the required amounts of negatively charged DHBC to compensate the cationic charges (R = AA/Mg = 2; R = AA/Al = 3; R = AA/ (Mg + Al) = 2.33). A rather constant Dh value was found over the whole range of R values between 0.53 and 2.33 for the mixture of Mg2+ and Al3+ with PAA3000-b-PAM10000 showing that the size of the formed micelles did not depend on the complexation degree R30 (Table 1). Precisely, Dh equals 24 nm at the stoichiometric ratio R = AA/(Mg + Al) of 2.33. This size was slightly smaller than the one obtained with simple micelles (28 nm) of PAA3000-b-PAM10000 and Al3+ cations only (Table 1). It is noteworthy that micelles did not form when PAA3000-bPAM10000 polymers were mixed with Mg2+ only. In this case a Dh value of 8 nm was found, which is quite close to the hydrodynamic diameter of copolymers alone, 6 nm, at pH = 5.5. The values of the scattered intensities confirm that no micelles formed when Mg2+ are the only cations (I/CDHBC(Al− Mg) = 11 502 kcounts, I/CDHBC(Al) = 10 612 kcounts/s, and I/ CDHBC(Mg) = 1403 kcounts/s, while I/CDHBC alone = 574 kcounts/s). In the case of Mg−Al suspensions, the scattered intensity normalized by the polymer concentration is about 20 times higher than that of the reference copolymer solution, while it is about 18.5 times higher in the case of the Al micelles, and 2.4 times higher than that of the reference (I/CDHBC alone) in the case of Mg solutions. It can be suggested that micelles of similar aggregation number formed in the case of Mg−Al and Al solutions, while mixtures of DHBC and Mg cations may have led to only dimers or trimers but not true micelles, as shown by the low values of Dh and I/CDHBC. Speciation of the PAA-b-PAM DHBC/Mg−Al micelle suspensions was determined after performing precipitation of the micelles in dioxane, which is known as a poor solvent for

■

RESULTS AND DISCUSSION The new synthesis route which is proposed here for the preparation of LDH colloids was inspired by a strategy that we previously developed in the case of simple metal hydroxides using DHBC polymers.18 The synthesis of stable colloidal suspensions in water of Mg/Al LDH nanoparticles using asymmetric PAA-b-PAM DHBC has never been performed and it appeared to be a challenging task for the two following reasons. First, the formation of intermediate mixed HPIC micelles containing both divalent and trivalent cations was never reported, and competition between Mg2+ and Al3+ for complexation by PAA could be expected. It is indeed wellknown that polyacrylate polymers preferentially complex trivalent rather than divalent ions,23 especially Mg2+. Second, LDHs are anion exchangers and a preferential intercalation of the negatively charged PAA blocks of DHBC between LDH layers might occur, since multicharged polymers can act as charge-compensating anionic species.26 PAA intercalation could inhibit the role of the DHBC as growth control and colloid stabilizing agent. Therefore, the synthesis of DHBC/LDH stable colloids was challenging and thus, it required a full understanding of, first, the formation of HPIC micelles in the presence of both Mg2+ and Al3+ ions, and second of the micelle transformation upon hydroxylation during the mineralization 9665


Langmuir

Article

PAM blocks.31 EDX analysis of the obtained precipitates was performed. Results are reported in Figure S1 in the Supporting Information. They show that Mg/Al molar ratios lay in the range 0.01−0.13 for all R values between 0.53 and 2.6, and are thus lower by far than the nominal value of 2. This reveals that, as a first approximation, the micelles prepared with a solution of Mg2+ and Al3+ cations did not contain any Mg2+, regardless of the value of R, in the range 0.53−2.6. These results confirm the higher selectivity toward trivalent cations in the competitive complexation of Mg2+ and Al3+ by polyacrylate-based DHBC. They also confirm the inability of Mg2+ cations to form an insoluble complex phase with PAA blocks, and thus, the impossibility of forming Mg-based micelles. Mg2+ ions remained in solution, certainly as small soluble complexes containing not more than 2 or 3 Mg cations. In the presence of both Mg2+ and Al3+ cations, instead of mixed micelles, simple DHBC/Al3+ micelles were formed, in a first approximation. Mineralization by Progressive Hydroxylation of a Mg−Al Cation Solution in the Presence of DHBC Copolymer (R = 0.8). The progressive hydroxylation of a DHBC/Mg−Al suspension was examined in the case of R = 0.8. The titration curve of the mixture of cations and DHBC was determined by using NaOH as the titrating base (Figure 1).

and complexed metal cations are hydroxylated on a wider pH range. Moreover, the end point of the first plateau at h = 1 corresponded to the amount of OH− necessary to compensate the positive charge provided by the Al3+ cations and hydroxylation of Al3+ into Al(OH)3 (Al3+ represent 1/3 of the cations in the mixture (Mg + Al) at Mg/Al = 2). This end point was shifted down to h = 0.76 and the pH increase is less sharp in the presence of DHBC. Such behaviors accounted for a competition between hydroxylation of the Al3+ ions to form Al(OH)3 and complexation of Al3+ ions with PAA3000 blocks. The amount of complexed Al3+ with PAA blocks could thus be estimated at 24% of the initial amount. The Mg/Al ratios within the colloids formed in the presence of DHBC at increasing hydroxylation ratio were determined after precipitation of the colloids in dioxane and analysis by EDX. At low hydroxylation ratio (below h=0.7), Mg/Al ratios are close to zero. The Mg/Al molar ratio start increasing at the end of the first plateau in the titration curve, just before the formation of the LDH phase (Figure 1);32 Mg/Al equals 0.17 at h=0.8 and 0.39 at h=1.07. This shows that Mg2+ ions enter the inorganic condensed phase only after hydroxylation of Al3+ ions, as it is also well-known in the absence of DHBC in the classical precipitation of Mg−Al LDH by progressive pH increase. Once Mg2+ ions started to enter the colloids, the formation of the LDH phase could begin. The second higher pH plateau occurred at pH 9.80 and the end point at h = 2.33 corresponded to the compensation of the positive charge provided by Mg2+ and Al3+ (2/3 Mg2+ + 1/3 Al3+ for Mg/Al = 2). A final Mg/Al molar ratio of 1.74 was obtained in the particles formed in the presence of DHBC at h = 2.5, it is lower than the one obtained in the absence of DHBC, which is close to 2. The smaller value of Mg/Al obtained in the presence of DHBC can account for the limited incorporation of Mg2+ ions in the LDH phase under formation. In the presence of DHBC, a fraction of Al3+ ions remains complexed by acrylate functions when hydroxylation occurs; acrylate-complexed Al3+ ions are thus not totally available for cohydroxylation with Mg2+ ions, and this subsequently inhibits the full incorporation of Mg2+ ions in the colloids. That phenomenon results in the formation of colloids with a Mg/Al molar ratio smaller than the nominal value of 2. It must be emphasized that, in the presence of DHBC, the titration curves and the variation of the Mg/Al molar ratio showed that Mg2+ ions were progressively introduced in the colloids in the course of the hydroxylation process, but only after complete hydroxylation of Al3+ ions. XRD powder patterns of the dried colloidal suspensions in the case of R = 0.8 at h = 0.80, 1.40, and 2.33 precipitated in dioxane are depicted in Figure S3 in the Supporting Information. In all patterns, the very broad signal in the 12− 32° (2θ) range was due to the presence of the copolymer corona. The sample prepared at h = 2.33 exhibits a weak shoulder centered at 7° (2θ) corresponding to the (003) reflection of a LDH phase. The LDH phase was on the contrary absent in the other samples obtained at h = 0.80 and 1.40, that is, below the pH of the precipitation plateau. This confirms that LDH phase formation in the DHBC-based colloids occurs only when the pH reaches the plateau of precipitation. These results showed that the formation of DHBC/Al3+ micelles instead of mixed Mg−Al-DHBC micelles does not impede the formation of the Mg−Al LDH phase. Moreover, the nanoparticles were prepared in a fully aqueous colloidal route. Initially, DHBC/Al3+ based micelles were formed while Mg2+ remained free (or as small soluble complexes) in solution due

Figure 1. Titration curve (empty symbols) and Mg/Al molar ratio (full symbols) of the PAA3000-b-PAM10000/Mg−Al mixture (R=AA/ (Mg + Al)=0.8; Mg/Al = 2) (○, ●) as the hydroxylation degree h increases.

The procedure corresponded to the coprecipitation of the LDH phase by increasing the pH. The titration curve can be compared to that of a reference Mg−Al solution with no DHBC (Figure S2, Supporting Information), which exhibits two distinct plateaus.32,33 The first plateau corresponds to the formation of aluminum hydroxide and the second one to the formation of Mg/Al LDH.32,33 The last sharp pH increase (above pH 10.5), following the second plateau, was due to the addition of free hydroxide ions in solution. It is noteworthy that the initial pH decreased from 4 in the absence of DHBC down to 3.2, in the presence of DHBC; this is the result of the complexation reaction of the metal cations by PAA blocks.27 Indeed, complexation of cations by some acrylic acid functions leads to acrylate−Mn+ bonds, which induces the release of H+ ions in solution and leads to a decrease in pH. Moreover, in the presence of DHBC, a less distinct plateau was observed at lower pH, compared to the case without DHBC: hydrolysis of AA-complexed metal cations occurs on a broader pH range due to a more distributed speciation of metal ions in solution; free 9666


Langmuir

Article

to the preferential complexation of the trivalent ions by the PAA blocks. Then, as the hydroxylation degree of the metal cations increased, aluminum hydroxide formed in the DHBCstabilized colloids. At higher h values, Mg2+ were introduced in the colloids due to the formation of hydroxo bridges between Al3+ and Mg2+ ions, and that lead to the formation of the Mg− Al LDH phase directly in the DHBC-protected colloids. This led us to suggest a general scheme summarizing the behavior of the DHBC/Mg−Al mixtures during the micellization and mineralization steps (Scheme 1). Scheme 1. Schematic Representation of the Formation of the DHBC/Mg−Al LDH Colloids in Water by Progressive Hydroxylation Figure 3. XRD patterns of Mg−Al/Cl LDH (a); Mg−Al LDH colloids at R = 0.13 (b); R = 0.33 (c); R = 0.66 (d) and Mg−Al/PAA LDH reference sample (e) (*NaCl).

turbid as R increased and were clear above R = 1. Below the flocculation threshold, R1, the supernatants were recovered and were analyzed by DLS, together with the stable colloids obtained above R1. The scattered intensity increases when R increases from R = 0 to 0.27 due to the increase of the fraction of colloidal species stable in suspension. In that domain, the fraction of macroscopic precipitate decreases while the fraction of colloids increases as R increases up to R1. Above R = R1, the scattered intensity together with the hydrodynamic diameter decrease as R increases (Dh decreases from 530 to 60 nm as R increased from 0.27 to 1) (Figure 4); this accounts for the

Mineralization by Coprecipitation of DHBC/Mg−Al Cation Solution in the Presence of Copolymer (R varies from 0 to1). The mineralization of DHBC/Mg−Al systems was then investigated on micelle suspensions prepared at different R values ranging from 0 to 1 with the aim to study the influence of the complexation degree on the formed LDH colloids. First, great changes in the macroscopic aspect of the solutions were noted (Figure 2). In the absence of DHBC (R

Figure 2. Macroscopic aspect of the solutions obtained at different R values.

= 0), macroscopic precipitation of Mg−Al LDH phase occurred. This was confirmed by the characteristic diffraction peaks observed in the powder XRD pattern of the dried precipitate, particularly at 11.52 and 23.20° (2θ) respectively assigned to the (003) and (006) reflections on the basal planes (Figure 3a). The d003 basal spacing of ca. 0.77 nm was consistent with the intercalation of Cl− as charge compensating anions brought by the precursor Mg2+ and Al3+ salts (Figure 3a). One can note the presence at 31.8° (2θ) of the most intense characteristic peak of the excess NaCl phase formed after drying of the suspension and which was not eliminated, since no washing was performed on this sample. In the presence of DHBC, mineralization of the micelle precursors led to different behaviors depending on R. At low DHBC amounts, the large fraction of noncomplexed cations led to the formation of a macroscopic precipitate upon base addition. Above R = 0.27, precipitates were no longer present, while stable and opalescent colloidal suspensions were observed. This value of R = 0.27 corresponded to the flocculation threshold, called R1. Suspensions became less

Figure 4. Scattered intensity (□) and hydrodynamic diameter (●) as a function of the complexation ratio R of colloidal PAA3000-b-PAM10000/ Mg−Al LDH particles.

increase of the overall surface area of the colloids as the copolymer amount increases. The amount of copolymer, that is, the degree of complexation, determines the size of the mineralized colloids, since the copolymer stabilizes the external surface of the LDH particle. Figure 4 also presents the size of the corresponding micellar precursors as a function of R. The increase in the size of the colloids upon mineralization of the micelles was due to enlargement of the micelle core and to coalescence of micelles upon cohydroxylation reactions between Mg and Al ions.26 It was of utmost importance to confirm the formation of Mg−Al LDH phases when coprecipitation of both Mg2+ and Al3+ cations with the alkaline solution is performed in the micellar suspensions, and, also to characterize the morphology of the particles. For this purpose a drop of the colloidal 9667


Langmuir

Article

Figure 5. TEM images of DHBC/LDH colloids at R = 0.27 (a, b) and R = 0.33 (d, e); and particle size distributions at R = 0.27 (c) and R = 0.33 (f).

progressive favored intercalation of PAA blocks of DHBC rather than of Cl− ions. The LDH phase was also characterized by TEM. In all cases, typical morphology of LDH particles was observed with well dispersed elongated platelets containing few stacked layers. Figure 5 shows the images of the particles after drying of the colloidal suspensions at R = 0.27 and 0.33. The size of 145 LDH particles was measured, allowing us to determine the evolution of size distribution as the degree of complexation increased. Size distribution in the range 25−45 nm with a mean size of 35 nm at R = 0.27 becomes narrower, in the range 10− 25 nm with a mean size of 17 nm at R = 0.33. At R = 0.66, such measurements were not performed as the density of particles was too low. The evolution observed between R = 0.27 and 0.33 shows that the degree of complexation controls the size of individual LDH particles formed in the colloids. This confirms that not only the mean hydrodynamic diameter of the colloids diminishes as R increases, but also the mean LDH phase particle size. One must emphasize the low aggregation degree of the particles in the colloidal DHBC/LDH suspensions whatever the value of R, in contrast to LDHs obtained by classical coprecipitation at constant pH in the absence of polymer (Figure S4, Supporting Information). In addition, the appearance of the TEM images implies that the stable colloids of polymer-stabilized LDH nanoparticles are constituted of an aggregate of a few LDH particles sterically stabilized by a polymer corona. Deeper analysis of TEM images confirmed the evolution of the different LDH basal spacings in the colloids at different R values (Figure 6). Almost 100 elongated dark domains, apparently corresponding to cross sections of brucitelike layers, were measured in samples at R = 0.20 and R = 0.33. In the former case, the mean thickness of these domains is of about 1.30 nm. The XRD pattern has previously revealed a d003 value of 0.77 nm corresponding to Cl− intercalation at R = 0.13 (Figure 5). Therefore, the thickness of the dark domain observed on the image accounts for two superimposed layers intercalated with Cl− anions. This indeed corresponds to 0.77 + 0.48 = 1.25 nm, in close agreement with the obtained value of 1.30 nm. At R = 0.33, two populations in the range 1−1.4 and 1.4−1.9 nm were distinguished, accounting for domains containing two brucite-like layers intercalated with Cl− ions, and PAA anionic species, respectively. Their amount reaches 40% in the former case and 60% in the latter case. Both the

suspensions was deposited on a glass slide or on a carbon− copper grid and evaporated to perform X-ray diffraction and TEM analyses, respectively. LDH phases were checked by powder XRD analysis of the evaporated suspensions. For R= 0.13, when flocculation still occurs, the XRD pattern of the dried suspension exhibited the characteristic (003) and (006) reflections of the LDH phase with d 003 = 0.77 nm corresponding to Mg−Al LDH intercalated with Cl− (Figure 3b) as already reported in DHBC-free conditions. The LDH phase formed in the colloids was similar to the precipitated one at R = 0. In spite of the preferential orientation of the particles due to the deposition mode, the XRD pattern showed a less crystallized structure at R = 0.13 than at R = 0 with a crystallite coherent size of 1.70 nm (compared with 2.90 nm at R = 0) along the c direction, as calculated by the Scherrer equation. This reveals that particle growth was slightly inhibited. Characteristic LDH patterns were also observed at higher R values. It is noteworthy that, at R = 0.33, basal spacings determined from the positions of the two most intense diffraction peaks, that is, 1.26 and 0.67 nm, do not exactly match (003) and (006) harmonics (Figure 5c). The second harmonic of the (003) peak at 1.26 nm and the (003) peak due to Cl− intercalation probably overlapped in the very broad second peak. Besides, the crystallite coherent size along the c direction decreases to 1.40 nm. At R = 0.66, two broad peaks corresponding to basal spacings of 1.31 and 0.65 nm well accounted for (003) and (006) harmonics (Figure 3d). The d003 value obviously corresponded to the intercalation of anionic species of larger size than Cl−. The most probable ones were the negatively charged PAA blocks of the DHBC. To check this hypothesis, a reference PAA/LDH sample was prepared by coprecipitation at constant pH = 10 in the presence of a sodium polyacrylate homopolymer. The powder XRD pattern (Figure 3e) has the same profile as that of the dried colloidal suspensions at R values higher than 0.33 (Figure 3c, d). The d003 value of 1.16 nm was almost similar to those obtained at R = 0.33 and R = 0.66 (d003 = 1.26 and 1.31 nm, respectively), however with a slight decrease probably accounting for a different arrangement of the PAA species in the interlayer space;34−36 therefore, besides the role of DHBC as stabilizers of the particles, the small polyacrylate chains act as charge-compensating anions of the LDH phase. Moreover, the evolution of the XRD patterns as R increases suggested a 9668


Langmuir

Article

was studied on samples prepared by coprecipitation at constant pH = 10, for varying R ratios. It was shown that at R = 0.27, colloidal stability was maintained for 1 week, but a precipitate formed after 2 weeks. At R equal or larger than 0.33, colloids were stable as a function of time (the hydrodynamic diameter did not vary), for more than 3 weeks. It was then shown that colloids whose initial hydrodynamic diameter was smaller than 220 nm were stable for more than 1 month. Second, for biomedical applications, it could be necessary to establish a robust method of sterilization of the colloidal suspensions if they were used as drug delivery systems. An easy way of sterilizing solutions is by filtration. The present suspensions underwent a sterilizing filtration test (over 0.22 μm filters), and their properties were compared before and after filtration. The scattered intensities together with the hydrodynamic diameters were measured for samples with initial hydrodynamic diameters from 80 to 170 nm (Figure S5, Supporting Information). Figure 7 shows that neither the hydrodynamic diameters nor

Figure 6. Basal spacings measured on TEM images at R = 0.2 (a,b), R = 0.33 (c−e), and R = 0.66 (f,g).

observed distances and the relative amounts are in agreement with the XRD pattern (Figure 3c) showing higher intensity of the (003) peak assigned to the layers intercalated with PAA. It is possible that in the previous cases two individual brucitelike layers could not be distinguished in each dark domain due to the too low contrast between the layers and the interlayer space and because of insufficient resolution. At R = 0.66, as previously emphasized, a low density of dark domains was observed on the image; however, it was possible to measure the thickness of 45 of them. The obtained value of 0.52 nm is consistent with that of a brucite-like sheet (0.48 nm). Moreover, a basal spacing of ∼1.24 nm can be determined which is close to that obtained by XRD (d003 = 1.31 nm). Besides, this value agrees well with that obtained for the PAALDH reference sample. That confirms that the polyacrylate chains are the main charge compensating anionic species of the Mg−Al LDH at this value of R. Individual brucite-like sheets are more easily distinguished than in samples at lower R values. Mg/Al molar ratios of dried LDH particles were determined by EDX analysis as a function of R. Mg/Al ratio equals 2.08 at R = 0, and regularly decreases from 1.78 at R = 0.13 down to 1.73 at R = 0.66. This is consistent with our previous result (Mg/Al = 1.74) on the sample prepared at R = 0.8 by a progressive hydroxylation procedure. The results confirm that, as R increases, the amount of Al atoms complexed by acrylate functions increases and impedes the formation of a Mg−Al LDH hydroxide framework with a stoichiometric ratio of Mg/ Al = 2. The incomplete hydroxylation of Al atoms (due to complexation) inhibits the introduction of Mg atoms up to Mg/Al = 2. The higher the complexing ratio R, the lower the Mg/Al ratio of the colloidal LDH due to surface complexation (of Al) by DHBC polymers. Despite their ability to intercalate between LDH layers, DHBC polymers were shown to be able to control the size of the LDH nanoparticles. Properties of DHBC/LDH Particle Colloidal Suspensions. First, the stability of the colloids as a function of time

Figure 7. Comparison of the scattered intensity (■) and the hydrodynamic diameter (○) of the colloidal particles before and after sterilizing filtration over 0.22 μm filters.

the scattered intensities were changed upon filtration: it means that the size distributions did not change and that all the particles went through the filter. So, suspensions prepared with R higher than 0.33 passed the sterilization test successfully. Finally, the stability of all the suspensions (R > 0.3) was tested in a phosphate buffered saline solution (PBS), and we observed that neither the colloidal stability nor the hydrodynamic size was affected by the PBS medium. This is related to the high steric stabilization brought by the DHBC with long neutral PAM chains whose solubility in water is high. Finally, the values of the zeta potential of the colloids vary between −5 and −7 mV when R increases from 0.5 to 1, in agreement with sterically stabilized particles protected by anionic-neutral DHBC polymers. All these properties are highly favorable for biomedical applications of DHBC-protected LDH particles directly as colloids. The ability of the LDH phase to host negatively-charged drugs can then be explored in a further extension of the present study.

■

CONCLUSIONS A simple and direct method for the preparation of aqueous colloidal suspensions of polymer-stabilized Mg−Al LDH particles has been developed. It consists of using doublehydrophilic block copolymers as growth control agents and for controlling the colloidal stability directly in water. This was achieved by first mixing Mg2+ and Al3+ cations with the PAAcontaining DHBC polymers and then performing metal cation hydroxylation in the HPIC micelles. When mixing the cations with the DHBC, competition, highly favorable to Al3+, occurs 9669


Langmuir

Article

between Mg2+ and Al3+ for complexation by the negatively charged PAA blocks. Therefore, DHBC/Al3+ HPIC micelles were exclusively formed, while Mg2+ cations remained as soluble entities in solution. Then hydroxylation occurs and leads first to the formation of aluminum hydroxide from the AlHPIC micelles; then further hydroxylation allows introduction of Mg2+ in the colloid core and then precipitation of the LDH phase in the colloids. Such mechanism was confirmed by several main experimental features. The micellar core thus acts as a confined nanoreactor in which precipitation of the LDH phase takes place. The complexation degree is the main parameter governing the stability of the suspensions, the size of the LDH particles but also their anionic composition. Above a critical complexation degree R = 0.27, the precipitate disappeared and stable colloidal suspensions were obtained with a hydrodynamic diameter of the colloids decreasing from 530 to 60 nm as R increased from 0.27 to 1. It is noteworthy that, as the degree of complexation increases, Cl− anions were progressively replaced by the negatively charged PAA blocks of DHBC acting as the compensating anionic species of the LDH phase. Therefore, the DHBC not only acts like a growth control agent and stabilizer of the LDH phase, but its negatively charged PAA block is also the main compensating anion of the structure. The new route presented here allowed the direct synthesis in aqueous medium of stable suspensions of LDH particles which are less aggregated and of lower mean size than those obtained by the classical precipitation methods. Properties such as sizes of the colloids, LDH particle sizes, and distribution could also probably be adjusted using other types of complexing DHBCs. Moreover, the procedure could be adapted to the preparation of LDHs of other compositions, intercalated with different guest entities providing multifunctional hybrid materials. Biomedical applications are now explored, since LDH nanoparticles appear as very promising colloidal drug delivery systems.

■

(4) Kwak, S.-Y.; Kriven, W. M.; Wallig, M. A.; Choy, J.-H. Inorganic delivery vector for intravenous injection. Biomaterials 2004, 25, 5995− 6001. (5) Leroux, F.; Taviot-Gueho, C. Fine tuning between organic and inorganic host structure: New trends in layered double hydroxide hybrid assemblies. J. Mater. Chem. 2005, 15, 3628−3642. (6) Abello, S.; Medina, F.; Tichit, D.; Perez-Ramirez, J.; Cesteros, Y.; Salagre, P.; Sueiras, J. E. Nanoplatelet-based reconstructed hydrotalcites: Towards more efficient solid base catalysts in aldol condensations. Chem. Commun. 2005, 11, 1453−1455. (7) Liu, S.; Jiang, X.; Zhuo, G. Heck reaction catalyzed by colloids of delaminated Pd-containing layered double hydroxide. J. Mol. Catal. A: Chem. 2008, 290, 72−78. (8) Mastalir, Á .; Király, Z. Pd nanoparticles in hydrotalcite: Mild and highly selective catalysts for alkyne semihydrogenation. J. Catal. 2003, 220, 372−381. (9) Kannan, S. Influence of synthesis methodology and post treatments on structural and textural variations in MgAlCO3 hydrotalcite. J. Mater. Sci. 2004, 39, 6591−6596. (10) Oh, J. M.; Hwang, S. H.; Choy, J. H. The effect of synthetic conditions on tailoring the size of hydrotalcite particles. Solid State Ionics 2002, 151, 285−291. (11) Zhao, Y.; Li, F.; Zhang, R.; Evans, D. G.; Duan, X. Preparation of layered double-hydroxide nanomaterials with a uniform crystallite size using a new method involving separate nucleation and aging steps. Chem. Mater. 2002, 14, 4286−4291. (12) Nitoh, K.; Ayral, A.; Ogawa, M. Preparation of well-defined nanometer-sized layered double hydroxides by novel pH adjustment method using ion-exchange resin. Chem. Lett. 2010, 39, 1018−1019. (13) Naito, S.; Nitoh, K.; Ayral, A.; Ogawa, M. Preparation of finite particles of nitrate forms of layered double hydroxides by pH adjustment with anion exchange resin. Ind. Eng. Chem. Res. 2012, 51, 14414−14418. (14) Liu, S.; Zhang, J.; Wang, N.; Liu, W.; Zhang, C.; Sun, D. Liquidcrystalline phases of colloidal dispersions of layered double hydroxides. Chem. Mater. 2003, 15, 3240−3241. (15) Kuroda, Y.; Miyamoto, Y.; Hibino, M.; Yamaguchi, K.; Mizuno, N. Tripodal ligand-stabilized layered double hydroxide nanoparticles with highly exchangeable CO32−. Chem. Mater. 2013, 25, 2291−2296. (16) Hu, G.; O’Hare, D. Unique layered double hydroxide morphologies using reverse microemulsion synthesis. J. Am. Chem. Soc. 2005, 127, 17808−17813. (17) Wang, C. J.; O’Hare, D. Synthesis of layered double hydroxide nanoparticles in a novel microemulsion. J. Mater. Chem. 2012, 22, 21125−21130. (18) Adachi-Pagano, M.; Forano, C.; Besse, J.-P. Delamination of layered double hydroxides by use of surfactants. Chem. Commun. 2000, 91−92. (19) Hibino, T.; Jones, W. New approach to the delamination of layered double hydroxides. J. Mater. Chem. 2001, 11, 1321−1323. (20) Hibino, T.; Kobayashi, M. Delamination of layered double hydroxides in water. J. Mater. Chem. 2005, 15, 653−656. (21) Iyi, N.; Ebina, Y.; Sasaki, T. Water-swellable MgAl-LDH (layered double hydroxide) hybrids: Synthesis, characterization, and film preparation. Langmuir 2008, 24, 5591−5598. (22) Iyi, N.; Ebina, Y.; Sasaki, T. Synthesis and characterization of water-swellable LDH (layered double hydroxide) hybrids containing sulfonate-type intercalant. J. Mater. Chem. 2011, 21, 8085−8095. (23) Liu, Z. P.; Ma, R. Z.; Osada, M.; Iyi, N.; Ebina, Y.; Takada, K.; Sasaki, T. Synthesis, anion exchange, and delamination of Co-Al layered double hydroxide: Assembly of the exfoliated nanosheet/ polyanion composite films and magneto-optical studies. J. Am. Chem. Soc. 2006, 128, 4872−4880. (24) Li, D.; Xu, X.; Xu, J.; Hou, W. Poly(ethylene glycol) haired layered double hydroxides as biocompatible nanovehicles: Morphology and dispersity study. Colloids Surf., A 2011, 384, 585−591. (25) Tarasov, K.; Houssein, D.; Destarac, M.; Marcotte, N.; Gerardin, C.; Tichit, D. Stable aqueous colloids of ZnS quantum dots prepared

ASSOCIATED CONTENT

S Supporting Information *

Mg/Al molar ratio in HPIC micelles at different complexation ratios (R); titration curve of Mg/Al solution without DHBC; XRD powder patterns of the dried micelle suspensions precipitated in dioxane obtained with varying R; TEM image of Mg-Al LDH particles obtained by coprecipitation at constant pH = 10 in the absence of copolymer; scattered intensity and hydrodynamic diameter of colloidal suspensions before and after sterilizing filtration on 0.22 μm filters. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Faraji, A. H.; Wipf, P. Nanoparticles in cellular drug delivery. Bioorg. Med. Chem. 2009, 17, 2950−62. (2) Choy, J. H.; Kwak, S. Y.; Jeong, Y. J.; Park, J. S. Inorganic layered double hydroxides as nonviral vectors. Angew. Chem., Int. Ed. 2000, 39, 4042−4045. (3) Tyner, K. M.; Roberson, M. S.; Berghorn, K. A.; Li, L.; Gilmour, R. F.; Batt, C. A.; Giannelis, E. P. Intercalation, delivery, and expression of the gene encoding green fluorescence protein utilizing nanobiohybrids. J. Controlled Release 2004, 100, 399−409. 9670


Langmuir

Article

using double hydrophilic block copolymers. New J. Chem. 2013, 37, 508−514. (26) Gérardin, C.; Sanson, N.; Bouyer, F.; Fajula, F.; Putaux, J.-L.; Joanicot, M.; Chopin, T. Highly Stable Metal Hydrous Oxide Colloids by Inorganic Polycondensation in Suspension. Angew. Chem., Int. Ed. 2003, 42, 3681−3685. (27) Sanson, N.; Bouyer, F.; Destarac, M.; In, M.; Gerardin, C. Hybrid polyion complex micelles formed from double hydrophilic block copolymers and multivalent metal ions: Size control and nanostructure. Langmuir 2012, 28, 3773−3782. (28) Bouyer, F.; Sanson, N.; Destarac, M.; Gerardin, C. Hydrophilic block copolymer-directed growth of lanthanum hydroxide nanoparticles. New J. Chem. 2006, 30, 399−408. (29) Antonietti, M.; Breulmann, M.; Göltner, C. G.; Cölfen, H.; Wong, K. K. W.; Walsh, D.; Mann, S. Inorganic/organic mesostructures with complex architectures: Precipitation of calcium phosphate in the presence of double-hydrophilic block copolymers. Chem.Eur. J. 1998, 4, 2493−2500. (30) Bouyer, F.; Gérardin, C.; Fajula, F.; Putaux, J. L.; Chopin, T. Role of double-hydrophilic block copolymers in the synthesis of lanthanum-based nanoparticles. Colloids Surf., A 2003, 217, 179−184. (31) Sanson, N.; Putaux, J.-L.; Destarac, M.; Gérardin, C.; Fajula, F. Hybrid organic-inorganic colloids with a core-corona structure: A transmission electron microscopy investigation. Macromol. Symp. 2005, 226, 279−288. (32) Boclair, J. W.; Braterman, P. S. Layered double hydroxide stability. 1. Relative stabilities of layered double hydroxides and their simple counterparts. Chem. Mater. 1999, 11, 298−302. (33) Radha, A. V.; Kamath, P. Aging of trivalent metal hydroxide/ oxide gels in divalent metal salt solutions: Mechanism of formation of layered double hydroxides (LDHs). Bull. Mater. Sci. 2003, 26, 661− 666. (34) Tanaka, M.; Park, I. Y.; Kuroda, K.; Kato, C. Formation of hydrotalcite−acrylate intercalation compounds and their heat-treated products. Bull. Chem. Soc. Jpn. 1989, 62, 3442−3445. (35) Vaysse, C.; Guerlou-Demourgues, L.; Duguet, E.; Delmas, C. Acrylate Intercalation and in situ polymerization in iron-, cobalt-, or manganese-substituted nickel hydroxides. Inorg. Chem. 2003, 42, 4559−4567. (36) Oriakhi, C. O.; Farr, I. V.; Lerner, M. M. Incorporation of poly(acrylic acid), poly(vinylsulfonate) and poly(styrenesulfonate) within layered double hydroxides. J. Mater. Chem. 1996, 6, 103−107.

9671


Highly Stable Layered Double Hydroxide Colloids: A Direct Aqueous

Recommend Documents