ARTICLE pubs.acs.org/Langmuir
Tuning High Aqueous Phase Uptake in Nonionic Water-in-Oil Microemulsions for the Synthesis of MnZn Ferrite Nanoparticles: Phase Behavior, Characterization, and Nanoparticle Synthesis Carolina Aubery,† Conxita Solans,† and Margarita Sanchez-Dominguez*,†,‡ †
Instituto de Química Avanzada de Catalu~ na, Consejo Superior de Investigaciones Científicas (IQAC-CSIC), CIBER en Biotecnología, Biomateriales y Nanomedicina (CIBER BBN), Jordi Girona 18-26, 08034 Barcelona, Spain ‡ Centro de Investigacion en Materiales Avanzados, S. C. (CIMAV), Unidad Monterrey, GENES-Group of Embedded Nanomaterials for Energy Scavenging, Alianza Norte 202, Parque de Investigacion e Innovacion Tecnologica, 66600 Apodaca, Nuevo Leon, Mexico
bS Supporting Information ABSTRACT: In this work, the formation of water-in-oil (w/o) microemulsions with high aqueous phase uptake in a nonionic surfactant system is investigated as potential media for the synthesis of MnZn ferrite nanoparticles. A comprehensive study based on the phase behavior of systems containing precursor salts, on one hand, and precipitating agent, on the other hand, was carried out to identify key regions on (a) pseudoternary phase diagrams at constant temperature (50 °C), and (b) pseudobinary phase diagrams at constant surfactant (S):oil(O) weight ratio (S:O) as a function of temperature. The internal structure and dynamics of microemulsions were studied systematically by conductivity and selfdiffusion coefficient determinations (FT PGSE 1H NMR). It was found that nonpercolated w/o microemulsions could be obtained by appropriate tuning of composition variables and temperature, with aqueous phase concentrations as high as 36 wt % for precursor salts and 25 wt % for precipitating agent systems. Three compositions with three different dynamic behaviors (nonpercolated and percolated w/o, as well as bicontinuous microemulsions) were selected for the synthesis of MnZn ferrites, resulting in nanoparticles with different characteristics. Spinel structure and superparamagnetic behavior were obtained. This study sets firm basis for a systematic study of MnZn ferrite nanoparticle synthesis via different scenarios of microemulsion dynamics, which will contribute to a better understanding on the relationship of the characteristics of the obtained materials with the properties of the reaction media.
1. INTRODUCTION Inorganic nanoparticles have been intensively investigated in recent years for their distinctive electrical, optic, and magnetic properties. Their functionality is closely related to their structural characteristics, such as crystalline structure, particle size distribution, and shape, which allow their application in several fields including ferrofluid technology, chemical catalysis, contrast agents for magnetic resonance imaging (MRI), transistors, and magnetically guided site-specific drug delivery.1,2 Several wet chemistry methods can be listed for the synthesis of inorganic nanoparticles.3 Conventional methods include coprecipitation, solgel, and impregnation processes; however, the very small and controlled particle sizes required in certain applications are not achieved with these techniques, despite modern refining that is nowadays present in such methods.3 A method with a growing interest is the microemulsion reaction method (MRM),4 which has been found to be a suitable option to deal with those drawbacks. A microemulsion is a system of water, oil, and an amphiphile, which is a single optically transparent and thermodynamically stable liquid dispersion.5 Depending on the system components, composition r 2011 American Chemical Society
variables, temperature, electrolyte concentration, etc., water-inoil (w/o), oil-in-water (o/w), and bicontinuous microemulsions can be formed. Numerous works have been carried out for the formation of inorganic nanoparticles in w/o microemulsions.68 In most of the studies reporting the synthesis of inorganic nanoparticles by the MRM, the surfactant systems used are of the ionic type, which may have some drawbacks as, for example, functional species could be absorbed at the particle surface and interfere with its growth; possible contamination of ceramic nanoparticles with the counterions; and reduction in aqueous phase uptake of ionic microemulsions in the presence of precursor salts due to screening effects.914 Nonionic surfactants of the nonylphenol ethoxylated type (e.g., Triton X-100) have been used as well,1517 but these are not biodegradable. Fewer studies employ nonionic aliphatic poly(oxyethylene) alkyl ether surfactants (ethoxylated fatty alcohols), despite the fact that these are more environmentally Received: February 17, 2011 Revised: October 29, 2011 Published: October 31, 2011 14005
dx.doi.org/10.1021/la203125x | Langmuir 2011, 27, 14005–14013
Langmuir friendly and relatively common at commercial level. Nonionic surfactants have a great capacity of hydration by their ethoxylated (EO) units;18 hence, an appropriate selection of surfactant, oil, and precursor salts/precipitating agent concentration, in combination with the rich structural behavior that such a system displays as a function of temperature, can lead to a highly optimized system in terms of aqueous phase uptake and hence reactants loading. The dynamic behavior of microemulsions depends on the surfactant composition as well as on the concentration of the dispersed phase; that is, when the aqueous phase is increased in a w/o microemulsion, the discrete phase domains can increase in size, and the flexibility and percolation of the droplets are also progressively changing;1921 this can directly influence the reaction kinetics.22 Although some conclusions have been reported about several variables affecting the synthesis of nanoparticles by the MRM, that is, reagent and aqueous phase concentrations, the type of surfactant, oil, addition of cosurfactant, and mainly microemulsion droplet size,2326 the dynamic aspects of microemulsions have not been sufficiently taken into account. The present investigation is focused on the phase behavior and characterization of a novel microemulsion system, based on a commercial aliphatic poly (oxyethylene) alkyl ether nonionic surfactant (Synperonic 13/6.5). The system is extensively studied in terms of microemulsion type and dynamic behavior, by means of phase diagrams, conductivity, FT PGSE 1H NMR, and dynamic light scattering (DLS). The interest in the system under study lies on the fact that it has been designed to solubilize a high concentration of aqueous phase containing the necessary reagents for the synthesis of MnZn ferrite nanoparticles. A high aqueous phase solubilization allows for a broad range of compositions for the synthetic studies. This gives rise to ecological and economical advantages, due to the concomitant increase in synthesis capacity by microemulsion volume and the decrease in oil content. Furthermore, conditions have been sought so that w/o microemulsions of both precursor salts and precipitating agent can be obtained around 50 °C, which has been determined as an optimal reaction temperature for the direct formation of MnZn ferrite nanoparticles with a crystalline spinel structure,13,27 without the need for calcination steps. Finally, a deep knowledge in the dynamics of the microemulsions formed by this system will allow the study of different scenarios for the nanoparticle synthesis, contributing to a better understanding on the relationship of nanoparticle characteristics with the properties of the system used as reaction media.
ARTICLE
homogenized (vortex), and the vials were flame-sealed to avoid evaporation. The samples were left to equilibrate at 50 °C, for phase diagrams at constant temperature. For phase diagrams as a function of temperature, the samples were left to equilibrate at each temperature, and observations were made once equilibrium was reached. The aqueous phase consisted of Milli-Q water, aqueous solution of the precursor salts (PS), and aqueous solution of the precipitating agent TMAH (PA) 0.5 M. The PS and their concentrations were FeSO4 3 7H2O 0.25 M, MnSO4 3 1H2O 0.0625 M, and ZnSO4 3 7H2O 0.0625 M, dissolved in H2SO4 0.5 M (to avoid oxidation of Fe2+). Conductivity Measurements. Conductivity experiments were carried out using a Crison conductimeter GLP31. The conductivity cell was a model 52-92 (Crison) with Pt electrodes and a cell constant of 1 cm1. The temperature was controlled to (0.05 °C by a CAT temperature sensor, model 55-31 (Crison). Different aqueous solution concentrations at several surfactant:oil (S:O) wt ratios were measured. The aqueous solutions were the PS and PA, with the same concentrations as described above. Measurements as a function of temperature (typically from 10 to 60 °C) were carried out. Self-Diffusion Coefficients. The diffusion coefficients of the microemulsion components (for PA microemulsions only) were measured with the Fourier transform pulsed-gradient spinecho (FT PGSE) technique monitoring the 1H NMR spectrum. The self-diffusion coefficient is obtained from the attenuation of the spin echo under the influence of pulsed magnetic field gradients. Most of the measurements were developed on a Varian Inova 500 MHz spectrometer, while the diffusion coefficient of pure isooctane was measured on a Bruker DMX 500 MHz spectrometer. The gradient compensated stimulated echo spin lock (DgcsteSL) was the sequence selected for diffusion experiments, adjusting the gradient strength (G = 13 G/cm), the pulse interval (Δ = 50300 ms), and, eventually, the gradient pulse length (δ = 24 ms). The decay of the echo intensity is given by: A ¼ A0 exp½ DðγGδÞ2 ðΔ δ=3Þ where A is the echo amplitude in the presence of the gradient pulse, A0 is the echo intensity in the absence of the gradient pulse, D is the self-diffusion coefficient, and γ is the proton gyromagnetic ratio (=2.67515255 108 s1 T1).28 Measurements were performed at 50 ( 0.1 °C using 5 mm NMR capillary tubes, replacing the water of the microemulsions by D2O; nevertheless, a certain amount of H2O was present in the aqueous phase, arising from crystallization water of TMAH. All reported values are the average of three independent measurements. The self-diffusion coefficient D obtained by this function was related to the hydrodynamic radius (RH) by the StokesEinstein equation: RH ¼
kB T 6πμD
2. EXPERIMENTAL SECTION Materials. Synperonic 13/6.5 (ethoxylated isodecanol with an
average of 6.5 EO units, cloud point ∼65 °C, HLB ∼12.5) was a gift from CRODA. Isooctane (Suprasolv, for gas chromatography) and sulfuric acid (9597%, p.a.) were purchased from Merck. Tetramethylammonium hydroxide pentahydrate (TMAH, 98%) was purchased from Alfa Aesar. Iron(II) sulfate heptahydrate (puriss. p.a. ACS 99.5% min), manganese(II) sulfate monohydrate (puriss. p.a. ACS 99% min), zinc sulfate heptahydrate (puriss. p.a. ACS 99.5% min), and hydrogen peroxide (purum p.a. 30% w/w min) were purchased from Fluka. Deuterium oxide (99.9 atom %D) was purchased from Aldrich. All materials were used as received. Deionized water (Milli-Q) had a resistivity of 18.2 MΩ cm1. Phase Behavior. The samples for phase behavior were prepared by weighing the appropriate amount of Synperonic 13/6.5, isooctane, and aqueous phase in glass vials with a long neck. The samples were
where kB is the Boltzmann constant, T is the absolute temperature, and μ is the viscosity of the continuous phase. DLS. The hydrodynamic diameter of microemulsion droplets was estimated by dynamic light scattering measurements (Malvern Instruments, 4700). The equipment consists of a monochromatic laser light source (λ = 488 nm), directed to the cuvette containing the sample, placed into a thermostatted bath. The detector amplitude, the beam intensity, and the dispersion angle were adjusted to obtain a signal of ∼80 K counts s1. According to this, the dispersion angle was set to 90° for PS microemulsions and 150° for PA microemulsions. After detection (photomultiplier), the light was analyzed by a correlator. The samples were measured at 50 °C. The data were analyzed by the non-negative least-squares (NNLS) algorithm, except for the sample corresponding to the precipitating agent system and 10 wt % aqueous solution concentration, which was analyzed with CONTIN. The viscosity of the continuous phase (isooctane) at 50 °C was μ = 0.365 cP,29 and the isooctane 14006
dx.doi.org/10.1021/la203125x |Langmuir 2011, 27, 14005–14013
Langmuir
ARTICLE
refractive index at 50 °C was experimentally determined, η = 1.3860 (Optilab, rEX).
Synthesis of MnZn Ferrites by the Microemulsion Reaction Method. Nonpercolated, percolated, and bicontinuous microemulsions were selected for the synthesis of nanoparticles (S:O ratio 25:75, 5 wt % and 24 wt % aqueous solution concentration and S:O ratio 85:15, 50 wt % aqueous solution concentration, respectively). The syntheses were carried out by adding the PA microemulsion to the PS microemulsion. Before mixing, both microemulsions were equilibrated at 50 °C, and the mixing was also carried out at this temperature. The mixing ratio was adjusted to reach pH 12.5. According to this, nonpercolated and percolated microemulsions were mixed at 5:1 PA to PS microemulsion wt ratio, and at 6:1 microemulsion wt ratio for bicontinuous microemulsions. The reaction mixture was kept stirring at 50 °C during 1 h. A small amount of concentrated H2O2 (30% w/w) was then added, and the reaction mixture was left stirring at 50 °C during 30 min. After the synthesis was finished, the product was separated from the microemulsion by addition of absolute ethanol:water 1:1 v/v, followed by several cycles of centrifugation and washings. Finally, the product was dried overnight at 60 °C. In another experiment, a one-microemulsion strategy was used: a PS microemulsion with S:O ratio 25:75 and 12 wt % aqueous phase was prepared, and the PA was added directly as a solution until pH 12.5 was reached. Both solutions were previously equilibrated at 50 °C. After this step, the rest of the process was the same as when the two-microemulsion strategy was used. Characterization of MnZn Ferrites. The products were characterized by transmission electron microscopy by using either a HRTEM 200 kV JEOL 2100 LAB6 (objective polar piece with a resolution of 2.3 Å) or a HRTEM 300 kV Philips CM30 (Super-Twin objective polar piece, 2.0 Å resolution). The samples for TEM were prepared by the addition of one drop of final microemulsion reaction mixture (before washings) to 1 mL of isopropanol, and dispersing by ultrasound during 1 min. One drop of this solution was immediately deposited onto a holey-Formvar carbon TEM copper grid. The dried materials were also characterized by X-ray diffractometry (XRD Siemens D-500) and magnetometry (Magnetometer SQUID Quantum Design MPMS XL). The particle size was estimated from the broadening of the spinel XRD peaks (dXRD).
3. RESULTS AND DISCUSSION 3.1. Phase Behavior. The ideal system for the synthesis of MnZn ferrite nanoparticles should form microemulsions around 50 °C, because it is the minimum temperature that favors direct formation of the spinel crystalline structure, eliminating the need for a calcination step.13,27 On the basis of exploratory phase behavior studies, the system selected was aqueous phase/ Synperonic 13/6.5/isooctane. Three different aqueous components were used: Milli-Q water (as a reference), PS (FeSO4, MnSO4, and ZnSO4 in H2SO4 0.5 M), and PA (TMAH, whose role is to increase the pH during the reaction) aqueous solutions. Figure 1a shows the phase behavior of the system with Milli-Q water. The surfactant is not soluble in water, but the water is dissolved in the surfactant up to 23 wt %. The formation of monophasic liquid microemulsions (L region) extends along the surfactant oil axis from a S:O ratio of 8:92 incorporating a maximum of 24 wt % water. Addition of the reactants had a great influence on the phase behavior, as shown in Figure 1b and c for the systems with PS and PA, respectively. In both systems, the microemulsion zone is shifted toward higher water concentrations, increasing the minimum amount of aqueous solution needed to form one phase microemulsions; this was more pronounced for the PS system.
Figure 1. Phase diagrams at 50 °C of the aqueous solution/Synperonic 13/6.5/isooctane systems, showing single phase microemulsion regions (L) and multiphasic regions (M, M0 ). Aqueous solution: (a) water, (b) precursor salts, and (c) precipitating agent (TMAH).
It was also observed that in both systems there are two maxima in aqueous phase solubilization. For the PS system, one of the maxima (up to 45 wt % aqueous phase solubilized) is produced at S:O ratio 25:75, while for the PA system, the first maximum (∼26 wt % aqueous phase solubilized) is produced at S:O ratio 20:80. For both systems, the second solubility maximum is produced at around S:O ratio 85:15. From Figure 1b and c, it is observed that some overlapping is produced in the microemulsion regions of PS and PA systems, which allows for the selection of appropriate microemulsions to carry out synthesis (i.e., at S:O ratio 25:75; the synthesis of nanoparticles could be undertaken from 10 to 22 wt % aqueous solution). 14007
dx.doi.org/10.1021/la203125x |Langmuir 2011, 27, 14005–14013
Langmuir
ARTICLE
Figure 2. Phase behavior as a function of the temperature and aqueous solution concentration, of samples of the aqueous solution/Synperonic 13/6.5/isooctane systems with a S:O ratio 25:75, showing the microemulsion regions. Aqueous solution: (a) precursor salts and (b) precipitating agent. The lower (b) and upper (O) microemulsion formation limit and the percolation temperature, Tp (2), are indicated. The region below the Tp line denotes percolated microemulsions.
The differences in phase behavior between the system with water and those with PS and PA solutions can be attributed to dehydration of EO surfactant groups caused by the reactants in the aqueous phase (“salting out effect”). As a consequence, a decrease in repulsion between headgroups occurs and a shift in the hydrophiliclipophilic balance is produced, leading to phase separation at low aqueous phase concentrations (M0 region). In fact, the M0 region appears to consist of a microemulsion phase plus a very small amount of excess surfactant (not aqueous phase). Changes in the phase inversion temperature (PIT) for selected microemulsions described in the Supporting Information (Figure S1) provide evidence of the influence of reactants addition on the hydrophiliclipophilic balance properties of this system. As compared to classic w/o microemulsion systems, it may appear at first sight that certain ionic surfactant systems, such as water/Aerosol OT/isooctane, are superior, because for that system the phase diagram indicates a maximum aqueous phase solubilization of around 65 wt % at S:O ratio near 35:65.30 Nevertheless, when similar reagents are incorporated into this system, such as in the work by Yener and Giesche,9 a composition at S:O ratio 40:60 and aqueous solution concentration of only 25 wt % were outside of the microemulsion region, despite having a maximum water uptake of nearly 50 wt % at this S:O ratio in the absence of reactants. Although the full phase diagram with incorporated reagents was not reported, this behavior is an indication of an important shrinkage on the microemulsion region area upon addition of reactants. Another example is the system water/CTAB/hexanol; it has a maximum of 50 wt % water uptake at S:O 40:60 in the absence of reagents, which is shifted toward higher surfactant concentration upon addition of similar reagents (S:O ratio 45:55 for PS and 50:50 for PA).31 In the present work, the maximum aqueous phase uptake was at much lower S:O ratios (25:75 and 20:80, incorporating 45 and 26 wt % aqueous phase for PS and PA, respectively); at these S:O ratios, the water/CTAB/hexanol system incorporated less than 20 wt % for PS and less than 15 wt % for PA.
Figure 3. Conductivity k (b) and d[log(k)]/dW (O) as a function of aqueous solution concentration W for microemulsions with a S:O ratio of 25:75 at 50 °C. Aqueous phase: (a) precursor salts and (b) precipitating agent. M0 denotes the composition range in which microemulsions with a slight excess of surfactant phase were obtained, while L denotes the composition range for single-phase microemulsions.
To estimate the optimum water solubilization at temperatures close to 50 °C, and hence widen the range of microemulsion compositions available for synthesis, a phase behavior study as a function of temperature and aqueous solution concentration was carried out at S:O ratio 25:75. The results showing the microemulsion regions are displayed in Figure 2. For the PS system (Figure 2a), the solubility maximum is located around 50 °C, while for the PA system (Figure 2b), it is located at around 46 °C, explaining why in the pseudoternary phase diagrams at 50 °C of Figure 1, the microemulsion region for the PS system is much wider than that of the PA system. 3.2. Microemulsion Characterization by Conductivity. 3.2.1. Conductivity at Constant Temperature (50 °C). The phase diagrams at 50 °C hint the possible presence of different types of microemulsion structures (Figure 1b and c). In particular, the solubility maxima found along S:O ratios 25:75 and 85:15 (PS system), and S:O ratios 20:80 and 85:15 (PA system), are indicative of structural transitions taking place within the microemulsion region. Conductivity of microemulsions was studied at different S:O ratios, increasing the amount of aqueous solution at constant temperature (50 °C) to establish their dynamic behavior, hence the presence or absence of percolation induced by increase in aqueous solution concentration. Figure 3a shows example data for the PS system at a S:O ratio 25:75. It should be noted that samples below the lower solubilization limit (11 wt %) were also included; the small amount of excess surfactant was removed prior to measurement. The conductivity (k) remains very low up to 9 wt % aqueous solution concentration (
