Method of fractionating oxidic nanoparticles by crossflow membrane filtration

Method for classifying oxide nanoparticles by cross-flow membrane filtration

Method of fractionating oxidic nanoparticles by crossflow membrane filtration

The present invention relates to a method of classifying oxide nanoparticle dispersions by membrane filtration. The present invention also relates to the oxide nanoparticle dispersion obtained by the method of the present invention.

In recent years, academia and industry have seen a steady increase in nano-sized particles (in other words particles with a diameter of less than 1 micron), because the properties of nano-particles make it suitable for applications such as electronics, optics, and chemical products. Has great potential. Of particular interest in this article are particles with diameters in the range below 100 nanometers. In this range, effects commonly called "nano-effects" usually occur, such as quantum effects, which can be attributed to factors including the influence of the surface area of large particles. Furthermore, with these particles, the light scattering will be reduced to the observable increase in transparency in the "nanocomposite" where the particles are embedded in a matrix (often a polymer or coating material) to enhance their properties. Degree.

However, regarding the application of nano-particles in composite materials, it is important that the spherical particles do not cohesive first, and secondly do not exist in a narrow size distribution. Even only a small portion of the coarser particles or viscous polymer may adversely affect the properties of the composite. This is especially true for transparency. Nanoparticles are often modified to adapt to a specific matrix. The purpose of this modification is to produce a better dispersion and prevent the formation of viscose.

There are various methods for the synthesis of nanoparticles. In addition to gas phase synthesis, it is possible to operate in solution, and in this case a template is sometimes used. Another option is to grind coarser particles. The feature of this method is that it is more cost-effective than synthesis from molecular precursors.

Regardless of whether particles are synthesized by molecular precursors or obtained by grinding, the resulting product always has a size distribution. Although in the case of particles in the micrometer range, coarser particles can be separated by sedimentation, centrifugation, or screen filtration, these methods have limited practicality in the case of nanoparticles. If the nanoparticle is in the dispersion, the coarser particles may also be separated by sedimentation or centrifugation. However, in this case, the extreme surface area and time requirements and batch operation morphology are very unfavorable, making these methods practical. Almost trivial. Likewise, techniques such as particle size exclusion chromatography (SEC) or gel electrophoresis are not suitable for larger operations.

In many applications, such as the manufacture of UV-resistant polymer composites or fluorescent materials (Journal of Nanoscience and Nanotechnology, 2006, 6, 409-413), the nanoparticles of interest are those composed of metal oxides. Therefore, for industrial implementation, it is useful to have a continuous and easy-to-implement method that can classify oxide nanoparticles. There have been many proposals suggesting the use of membrane filtration in this regard. It should be noted here that the separation of the film is affected by the specific interaction between the particles and the film.

Generally, in order to separate particles from the suspension according to specified criteria (such as particle size), filtration techniques are usually used. In this case, a batch operation of dead-end filtration or a dynamic operation of cross-flow filtration is generally used. In the case of end-point filtration, the entire amount to be filtered is passed directly through the filter medium, and in this case the deposited particles can generally accumulate into a filter cake, which therefore collectively determines the result of the filtration. The result of the filtration here is then not only determined by the nature of the filter medium, but in particular by the filter cake that has been formed and changed during operation. Therefore, cake-forming filtration cannot be used to classify particle dispersions. Only the depth filter operating according to the endpoint method can classify within certain limits by allowing the particles to be separated to penetrate the structure of the filter medium and separating according to the adhesion of the particles to the inner surface of the filter medium. The limitation of this method is that it can only handle very thin dispersions and the classification effect has a high inherent inaccuracy; as a result, a considerable amount of the target product will remain adhered to the filter medium and thus be lost.

In the case of cross-flow filtration, the medium to be filtered is passed through the filter medium in shear. The cut-off limit is determined by the pore size of the filter medium. Important applications are in microfiltration, ultrafiltration and nanofiltration.

Cross-flow filtration attempts to avoid the shortcomings of the end-point filtration method by making the flow that hits the filter medium sheared compared to traditional filtration in this case. The feed stream is divided into a filter stream passing through the filter medium and a parallel stream on the filter medium. In membrane technology, the flow through the membrane is called permeate. The substance remaining on the film is called the retentate. According to the result of this flow method, the retained components are transported back to the retained stream from the surface of the filter medium. This will therefore counteract the formation of sediments and filter cake layers on the filter medium.

Advanced Materials 2005, 17(5), 532-535 explains how to use membrane cross-flow filtration technology to classify metal nanoparticles. For this project, a special film containing nano-level channels was manufactured. The experiment was conducted on a very small laboratory scale, and no instructions were provided for transfer to an industrial scale. It does not mention any essential description about the pore size or composition of the membrane required to classify oxide nanoparticles. Furthermore, the observations among them suggest that it is impossible to classify nanoparticles using traditional thin films.

Journal of Membrane Science 2006,284,361-372 describes the cross-flow membrane filtration of silica nanoparticle dispersions. However, the details of the possibility of classification are not provided here; instead, the effect of the formation of a dynamic filter cake layer by nanoparticles is studied, which would make classification impossible.

Bokela, Karlsruhe (Germany) sells a mesh filter system (Dynofilter) that can separate coarse particles as small as 10 microns from the particle dispersion by dynamic mesh filtration. As far as the possibility of using thin films in this system is concerned, there is no way to know.

Langmuir 1997, 13, 1820-1826 describes the study of membrane filtration of polymer particles with constant surface charges. Among them, the retention of nanoparticles is improved by introducing surface charges. As far as the classification effect is concerned, there is no explanation.

Anal. Chem. 2006,78,8105-8112 outlines how to separate organic colloids from aqueous solutions by cross-flow ultrafiltration. Among them, the film retains more than 99% of the nanoparticles.

US 2004/0067485 A1 describes the use of protein as a template to synthesize nano-scale semiconductors based on zinc and cadmium in combination with the elements S, Se and Te. It is pointed out that the obtained complexes such as zinc sulfide/iron-deficiency ferritin may be classified by end-point membrane filtration, and the membrane pore size used is much higher than the particle size. It lists a large number of optional membrane materials, but there are no examples of filtration methods. With this method, it is impossible to perform nanoparticle classification and filtration on an industrial scale because of the formation of filter cakes. Therefore, the requested method can only be performed on a laboratory scale, in which the required filter material is likely to be replaced frequently.

WO 2006/116798 A1 describes the production of radioactive nanoparticles based on metal Tc. These particles are subjected to an end-point membrane filtration process. The membrane used in this procedure is hydrophilic. Therefore, the endpoint method is also used here, which can cause the particles to be almost completely deposited on the film even under low-level nanoparticle cohesion. The proposed nanoparticle classification method can therefore only be carried out on a laboratory scale at a very low nanoparticle concentration.

In the prior art, there is no known thin film method that can be used on an industrial scale and can classify fine parts from particle dispersions mainly containing nano particles and coarse particles. All the prior art methods form a filter cake layer on the surface of the film, so although separation is possible, it is impossible to classify the nanoparticles by self-dispersion.

Therefore, the object of the present invention is to provide a method for classifying the dispersion of metal oxide nanoparticles, which can overcome the shortcomings of the prior art (especially the accumulation of the filter cake layer on the film), so that the nanoparticles can be classified , Even if it is implemented on an industrial scale.

This objective is achieved by a method of classifying oxide nanoparticle dispersions. The method is characterized in that at least one step of the method is a membrane cross-flow filtration step, in which the dispersion is made in the Flow over the film.

Surprisingly, it was found that the method of the present invention can prevent the accumulation of a filter cake layer on the film and can perform the classification of the oxide nanoparticle dispersion.

The method of the present invention is carried out using a filter unit operating on the principle of cross-rotation (CR). These are filter machines that generate shear force and flow in parallel with the filter medium by moving the inside. This may also include the filter media itself (e.g. rotating membrane). A possible version of this filter machine is a cross-rotating filter. In contrast to cross-flow filtration, cross-rotation filtration uses an additional rotor to decouple the flow on the membrane from the feed and pressure build-up. In a typical CR filter, the filter medium and the rotor are stacked on each other in a sandwich formation. There is a rotating shaft driving the rotor in the middle of the stack, and the rotating shaft can reach an average cross flow velocity much higher than 6 m/s-that is, a spatially average cross flow velocity across the entire filter medium. The supplied medium is guided into the layer through the plate stack. In the plurality of sections of the plate stack, the medium is concentrated in stages. By decoupling the feed (by the pump) from the flow on the membrane (by the rotor), a high specific filtrate flux with low system pressure can be achieved. These systems with rotating internals are now used in applications that require the separation of very small particles or viscous materials and that require very high concentrations.

Microfiltration, ultrafiltration and nanofiltration are also included in pressure-driven membrane processes. The structure called a thin film is generally a two-dimensional structure that forms a barrier between two fluid phases and allows selective exchange of materials between the two sides. Therefore, the membrane is a filter medium that has a limited critical value or retains particles of a specific size in the presence of a driving force (pressure). The nature of the driving force, as well as the pressure and flow conditions at the membrane that are related to the nature of the membrane, determine the separation result. The function of dividing the synthetic film into different categories is carried out according to the structure and aggregation state of the synthetic film and its electrochemical characteristics. In the meaning of this application, the film has a diameter of at most 10 microns. If it has a larger pore size, its structure is called a screen.

As the film, any commercially available film can be used, such as those made of polyethylene, polytetrafluoroethylene, polytetrafluoroethylene, or cellulose. It is preferably a symmetrical membrane, that is, the membrane has a fixed pore size over its entire cross-section. The film used according to the present invention has a pore size of at most 10 microns, preferably between 0.01 and 5 microns, more preferably between 0.1 and 1 microns.

The average cross-flow velocity generated in the method of the present invention is between 5 and 25 m/sec in total, preferably at least 8 m/sec, and more preferably at least 10 m/sec. These conditions are achieved in the form of the relative speed between the medium to be filtered and the membrane, through the rotation of the contents (such as an agitator). The average cross flow velocity is determined by measuring the velocity on the entire filter medium and then taking the average value. In this context, it is particularly advantageous to decouple this cross flow velocity from the pressure build-up, so as to generate the above-mentioned high cross flow velocity at low permeable membrane pressures (<1 bar). Under these conditions, the formation of the filter cake layer can be almost completely prevented. In addition, the formation of the filter cake layer can be offset by regular backwashing with permeate or liquid (such as water) or a liquid/gas mixture (such as water and compressed air).

The nanoparticle used in the meaning of the method of the present invention can be any metal oxide particle. The metal oxide can be prepared, for example, by flame pyrolysis, precipitation procedures, or solvogel operations. The metal oxide can also be a mixed oxide derived from two or more different metals. Preferably, it is titanium oxide, cerium oxide, aluminum oxide, silicon dioxide, zirconium dioxide, zinc oxide, indium tin oxide, antimony tin oxide, barium titanate, or mixed oxides containing these components. Furthermore, two or more different metal oxides in the form of a mixture may also be present in the dispersion. The preparation of the metal oxide nanoparticle dispersion by grinding in the liquid phase can be carried out, for example, in the manner described in DE 10204470 A1. The liquid jets are sprayed with each other under high pressure, causing them to collide and, therefore, the principle of reducing the size of the materials contained in them and dispersing them is also called wet jet milling. A concomitant situation in its production is that the dispersion has an asymmetric particle size distribution. Other techniques for making dispersions include, for example, the use of jet mills, stirred ball mills, ultrasonic dispersion equipment, rotor-stator machines, Ultra-Turrax, planetary kneaders/mixers, or high-pressure homogenizers, alone or in combination. The dispersion formed by such techniques can also have a symmetrical particle size distribution, and is also suitable for the method of the present invention.

The dispersion used for the purpose of the method can, for example, be obtained directly from the metal oxide by synthesis, or obtained by redispersing the metal oxide in advance to dry the solid part of the dispersion. Before or during the dispersion operation, for example, adding a dispersing aid to adjust the pH or perform chemical surface modification of the metal oxide (for example, by using a reactive silane compound or a compound electrostatically bound to the particle to modify ). This also applies to the case of redispersing metal oxides. The liquid component used in the dispersion can be water, water-based liquid, organic liquid or ionic liquid, or a mixture of two or more representatives of one of the groups, or different groups A mixture of its representatives. These liquids can also be solutions.

In addition, in the liquid component of the dispersion, there may be dissolved or sufficiently finely dispersed components, examples of which are stabilizers, inhibitors, anti-aging agents, biocides, dyes, antistatic agents, and salts , Surface active substances or corrosion control agents.

In a specific example of the method of the present invention, the membrane cross-flow filtration step is connected with the upstream grinding process so that the separated coarse material (retentate) is returned to the grinding operation, and the fine part (permeate 1) is removed from the operation . Then, if appropriate, through a separation and/or concentration step, such as ultrafiltration, to increase the solid content of the permeate 1, while the permeate 2 obtained in the concentration process is substantially free of nanoparticles and is hereinafter It is called the clean part, or the concentrated liquid corresponding to the permeate 2 is returned to the operation, that is, the previous filtration step, and is used as "washwater". However, it is also possible to add "washing water" from the outside. In a preferred embodiment, the retentate in the filtration step is cleaned by membrane percolation. In this case, the new "washing water" is continuously supplied to the retentate until the old "washing water" has been completely replaced through the film, and the residual nanoparticles present in the retentate enter the permeate through the film until. Another possibility is to meter it back to the grinding operation. In a preferred embodiment, the entire operation is performed continuously.

Figure 1 shows a basic flow chart of a possible operation method.

In order to more substantially separate the fine fraction from the dispersion, it may be necessary to use additional "washing water", which is used together with the dispersion medium if appropriate.

In addition to the use of the term water or "washing water", the term liquid may be used more generally, because in addition to water and water-based liquids, liquid organic compounds or ionic liquids may also be appropriately used. The liquids can also be a mixture of two or more compounds of the same substance, or a mixture of individuals of the liquid group, which can not only form the substrate of the nanoparticle dispersion, but also can be used as "washing water".

The following examples illustrate the present invention in more detail, but should not be regarded as limiting the protection scope of the present invention.

Example

The following tests are carried out with a 30% dispersion of titanium dioxide in water, which can be prepared, for example, by the method described in DE 10204470 A1. The dispersion used has a D of about 0.09 microns₅₀Value and D of about 0.2 microns₉₀Value; that is, 50% of the particles have a diameter less than or equal to 0.9 microns, and 90% have a diameter less than or equal to 0.2 microns.

Use the following commercially available films from Millpore:

Test 1: PTFE film with 1 micron pore size

Test 2: PTFE film with 0.45 micron pore size

Test 3: PTFE film with 0.2 micron pore size

The agitator speed in each case was 1264/min. This corresponds to an average cross flow velocity of about 9 m/s.

The particle size distribution is measured before classification (measurement series using square measuring points) and after (measurement series using circular measuring points) (Figure 2).

In addition, scanning electron micrographs of the dispersion of Experiment 3 were taken at two different resolutions (UHR detector, high voltage, 10 and 20 kV) (Figure 3).

From the results of experiments 1 to 3, it can be clearly seen that through the method of the present invention, the particle size distribution of the dispersion used is shifted to a smaller particle size, so that after the classification is completed, the dispersion does not have 100 nanometers or more. Large diameter particles. This finding is also supported by the scanning electron micrograph in a purely qualitative manner.

Figure 1 shows a basic flow chart of a possible operation method.

Figure 2 depicts the particle size distribution measured before (measurement series using square measuring points) and after (measurement series using circular measuring points).

Figure 3 shows the scanning electron micrographs of the dispersion of Test 3 taken at two different resolutions (UHR detector, high voltage, 10 and 20 kV).