Analysis of a technical-grade w/o-microemulsion and its application for the precipitation of calcium carbonate nanoparticles

Abstract

For a directed synthesis of inorganic nanoparticles in water-in-oil (w/o)-microemulsions the detailed investigation of the microemulsion system is of particular importance. This study illustrates that the phase behavior of the ternary as well as of the quaternary mixture (with reactants) has to be analyzed to identify microemulsion regions suitable for particle synthesis. Besides detailed phase prism investigations the microemulsion, consisting of water/reactant, cyclohexane and Marlipal O13/40, was further characterized. Dynamic light scattering studies have been conducted for determining the droplet size and viscosity measurements have been undertaken to clarify the internal structure. A time scale analysis of the involved processes during particle formation (e.g., droplet exchange, reaction, nucleation, growth and mixing) points out, that the utilization of w/o-microemulsions for precipitation reactions is an effective way to overcome mixing problems, which usually are present in bulk precipitation. This is demonstrated for the precipitation of CaCO₃ from two microemulsions containing dissolved Na₂CO₃ and CaCl₂, respectively. The influence of microemulsion composition, initial reactant concentration and holding time on the final particle size and shape was investigated. The resulting CaCO₃ nanoparticles have been characterized by means of a transmission electron microscope. Depending on the applied reaction conditions it is shown that different particle sizes and morphologies can be obtained.

Introduction

The synthesis and processing of ultrafine solid particles at the nanometer scale (characteristic length 1–100 nm) is of interest for a great number of existing and upcoming technological applications. Beside their use as active compounds in heterogeneous catalysts they are processed, for instance, as semiconductors, for enhanced magnetic and optical devices or for the production of advanced bulk materials [1], [2].

Because the special physical and chemical properties of nanoparticles are strongly size dependent, the major challenge for a synthesis process is to control particle size, shape and particle size distribution (PSD).

Nowadays, a variety of existing methods provide the know-how of defined nanoparticle synthesis of very different materials at the laboratory scale. And since the demand of tailored nanomaterials for industrial applications is permanently increasing, the research on scaling-up approaches of developed and proved laboratory methods are becoming more and more important.

A successful technique for the controlled preparation of very small and narrowly distributed nanoparticles is the use of w/o-microemulsions. This has been evidenced by a number of groups during the last decade for different nanomaterials and a variety of microemulsion systems [3], [4], [5], [6], [7]. However, the usability of the suggested microemulsion components (costs, availability), a key issue with respect towards a technical scale realization, was neglected in most scientific investigations.

In Fig. 1 a possible flow sheet of an industrial process for the production of nanoparticles in w/o-microemulsion droplets is shown. For an effective realization of such a process the detailed investigation and understanding of the particle formation dynamics in the droplets of the chosen w/o-microemulsion system is of particular importance.

These ternary mixtures, usually consisting of water, a non-polar solvent (oil phase) and a surfactant, are thermodynamically stable, optically clear and isotropic one-phase solutions. In case of w/o-microemulsions (also named reverse micelles) nanometer sized water droplets are stabilized by a surfactant monolayer and segregated by a continuous oil phase. These three component systems exhibit a very complex phase behavior often represented in a ternary phase diagram (Gibbs triangle). As these phase diagrams are temperature dependent it is advantageous to include the temperature as ordinate and display all in one diagram at constant pressure as suggested by Kahlweit et al. [8]. In such a phase prism a multitude of structures can be found: e.g., two- and three-phase regions, liquid crystalline phases as well as microemulsion phases (one-phase regions). A systematic investigation of these one-phase regions, by studying extensively the phase behavior, is of particular importance if chemical reactions are carried out in a microemulsion. Another aspect that should be taken into account is the influence of added reactants on the extent and stability of the microemulsion phase [3], [9].

Moreover the internal nanostructure of the microemulsion region is a point to be investigated. Besides the preferred structure of isolated spheroidal water domains, larger droplet cluster or bicontinuous structures can be found and detected by viscosity or conductivity measurements in the one-phase region [10], [11]. To establish a set of information required for a directed particle synthesis in microemulsion droplets, dynamic light scattering measurements provide details about droplet dimensions under different conditions (e.g., temperature, surfactant content). This is of particular meaning if particles with specified properties should be produced, because the droplet diameter, adjustable in the nanometer range, affects the final particle size and shape as demonstrated by a number of groups [5], [12], [13].

Boutonnet et al. [14] were among the first using the advantages of segregated water droplets in microemulsions, regarded here as nanoreactors, for the production of nanomaterials. Since that time microemulsions have attracted a lot of advertence and a number of theories on the particle formation inside the droplets have been proposed [15], [16].

Besides the limiting effect of the droplet size on particle growth and especially agglomeration the dynamic behavior of the droplet population plays a decisive role. Random collisions of the droplets, caused by the Brownian motion, lead to the formation of transient dimers (fusion–fission mechanism) and subsequently to the exchange of reactants.

Thus, not only the size of the confined water domains but also this droplet exchange process, influenced by the elasticity of the surfactant film and other parameters, is crucial for particle formation and must be considered [17], [18]. In fact, the ratio of the relative rates of all involved sub-processes, e.g. droplet exchange, reaction, nucleation and growth (as shown in Fig. 2) are determining the overall process and hence the final particle size and shape [15], [16], [19].

In case of very fast reaction kinetics (e.g. ionic reactions) the mixing process has to be considered as well. For common precipitation processes in the bulk phase micromixing is usually regarded as the most dominating process [20], [21]. If complete micromixing is not achieved before reaction and nucleation start, one ends up with coarse particles and a broad particle size distribution due to inhomogeneities of the spatial concentration distribution. The following time scale analysis for a precipitation process in a stirred tank points out that the droplet exchange rate can be used to slow down and potentially control the usually very fast kinetics for reaction and nucleation.

In the present study the precipitation of CaCO₃ from aqueous Na₂CO₃ and CaCl₂ is investigated. The reaction between dissociated Ca²⁺ and CO₃²⁻ ions is assumed to be instantaneous, i.e. the time constant τ_r for the diffusion-controlled reaction could be estimated by applying the Smoluchowski equation [22]. The subsequent nucleation of CaCO₃ is considered to be primary and therefore the corresponding rate can be calculated according to the classical nucleation theory [23]. This homogeneous nucleation rate B_hom grows significantly with increasing supersaturation. For the typical set of supersaturation values 100 and 1000, the corresponding nucleation rates of 6.7 × 10¹⁹ and 2.3 × 10²⁸ m⁻³ s⁻¹ can be calculated for the specific case of calcium carbonate precipitation.

The characteristic time for nucleation τ_n is considered to be inversely proportional to the nucleation rate and can be defined as:

$${\tau }_{\text{n}}=\overline{N}/B_{\text{hom}}$$

where $\overline{N}$ is the average concentration of particles in the bulk phase as introduced by Baldyga et al. [24]. If one assumes a diffusion controlled droplet exchange, i.e. every collision leads to the complete exchange of the droplet content, the according second-order rate constant k_D,ex can be derived from the Smoluchowski equation [22]. The typical time constant τ_D,ex for this bimolecular droplet exchange process can now be estimated as:

$${\tau }_{\text{D},\text{ex}}=1/(k_{\text{D},\text{ex}}{c}_{\text{drop}})$$

with c_drop as the molar droplet concentration of the microemulsion [25], [26]. Almost all experimentally determined exchange rate constants are much lower, indicating that the droplet exchange process is not diffusion controlled.

As mentioned above, micromixing is the fastest process in the mixing sequence and is therefore determining the product quality (e.g. PSD) in bulk precipitation. Following Baldyga et al. [24], the characteristic micromixing time can be expressed by the time constant for mixing by engulfment τ_E:

$${\tau }_{\text{E}}=17.3{(\nu /\epsilon )}^{1/2}$$

In this equation, ɛ is the local turbulent energy dissipation rate and ν the kinematic viscosity of the liquid.

The particle growth of already formed nuclei is the slowest process during the particle formation. A characteristic time constant τ_g can be derived from the linear growth rate G according to the following equation:

$${\tau }_{\text{g}}=c_{\text{c}}M/(\rho GA)$$

where c_c is the concentration of precipitating substance, ρ the particle density, A the particle surface area per unit volume of suspension and M the molar mass of the precipitated solid.

In a stirred tank the term macromixing describes the process of mixing the reactor input with the entire content of the reactor. The characteristic time constant for macromixing τ_M is defined as the time needed for reaching a final concentration field [27] and can be estimated from the following equation:

$${\tau }_{\text{M}}\approx 3.4{(D_{\text{T}}/d)}^2/N$$

where D_T is the reactor diameter, d the diameter of the stirrer and N the stirrer speed. Because experiments show that the final particle size is not affected significantly by increasing the stirrer speed over a certain level it is assumed that for particle precipitation in microemulsions macromixing is the dominating mixing process [19], [28].

The typical time and length scales of the regarded sub-processes are shown in Fig. 3. In particular, for very fast kinetics (e.g. reaction, nucleation) the obviously slower droplet exchange might be an efficient way to slow down and control these processes.

Keeping in mind that the effective exchange rate is two to four orders of magnitude smaller than the calculated diffusion exchange rate [16], [25] this slow down effect is much more pronounced. Thus, the precipitation of nanoparticles from two initial microemulsions containing the reactants should be feasible without paying special attention to the micromixing conditions.

In the presented work a non-ionic, technical-grade w/o-microemulsion system was extensively investigated to evaluate the suitability as medium for precipitation reactions. As a representative system the formation of calcium carbonate in the water droplets was studied with the objective to synthesize nanoparticles of a certain size and shape.