Towards the production of core-shell nanoparticles with fluidized bed ALD

Reducing particle size or material structure size to nanometer scale can make the material properties, such as light absorption and electronic structure, change compared to the same materials at normal scale. This gives them properties that can make them suitable for the development of highly efficient and improved micro-electronics, sensor,medicine, batteries, catalysts and third generation solar cells. There are, however some challenges that need to be overcome in the development of nanoparticle-based devices. The first is protection of nanoparticles against corrosion and oxidation. This phenomenon is increased by the large surface area available for corrosion. Furthermore, when nanoparticles or nanostructures are used in electronic devices, lowresistive electrical contacts should be made between electrodes and nanoparticles. The second challenge is, thus how to make electrical contacts without compromising the material’s nanostructure. This thesis deals with the development of a synthesis process for core-shell nanoparticles, existing of a core that is coated with a thin layer of material that is able to provide protection, or electrical contacts. The first chapter describes the electrical contact between titanium nitride (TiN), which is a metallically conductive material that is used as contact material in electronic devices, and cadmium sulfide (CdS), a II-IV semiconductor that is used in (second generation) thin film solar cells and to increase light absorption in Grätzel-type solar cells. The experiments show that indeed we can make a Ohmic contact between TiN and CdS, meaning that the contact resistance between the two materials is low and that current is not blocked by the contact. The use of thin coatings as protective layers is investigated by coating thin CdS films,which can be used as photo-catalyst in solar hydrogen production cells, with thin, inert titanium dioxide (TiO2) to protect the CdS from corrosion under influence of solar radiation. The goal of this research was to deposit a TiO2 layer that was thick enough to provide full protection against corrosion, yet thin enough to enable electrons to be transferred between the CdS electrode and the electrolyte. The TiO2 coating was deposited with Atomic Layer Deposition (ALD), a technique used to deposit extremely thin layers of material by letting two precursors (A and B) react on the surface of a substrate to form product C. The first step in this process is chemisorption of precursor A. This chemisorption reaction is self-limiting and stops whenever the complete substrate is covered with a monolayer of component A. After completion of pulse A, precursor B is fed to the reactor and reacts with component A to form component C and prepare the surface of the substrate for a new pulse of component A. By repeating this cycle coatings can be made atomic layer by atomic layer. The experiments with TiO2-coated CdS films in photoelectrochemical hydrogen production cells show that, even though the samples are coated with protective TiO2 layers, the CdS remains sensitive to photocorrosion. The photocorrosion mechanism is investigated with electrochemical measurements in which the photocurrent over time can be described with a model that strongly resembles the Johnson-Mehl-Avrami model for phase transitions in solids. Analysis of the experimental results with the model shows that the corrosion starts in small defects in the TiO2 coating and that the corrosion spreads mostly in lateral directions. The next step in the research was to deposit coatings on individual nanoparticles with a fluidized bed ALD reactor (FB-ALD) that was specially developed for this purpose. In this reactor, the nanoparticles are agitated by a constant flow of inert carrier gas. The ALD precursors, tetrakis-dimethylaminotitanium (TDMAT) and water, are added to the carrier gas and hence brought into contact with the nanoparticles and layer-by-layer form a TiO2 shell on the particles. In the design of the reactor that, is used for loose nanoparticles, care has been taken to make the reactor both safe and versatile in operation. Furthermore, the possibility of extensive monitoring of the reactor is provided. With this reactor, silica (SiO2) nanoparticles are coated with 1.6 nm TiO2 layers. The growth rate is 0.32 Å per ALD cycle and independent of precursor pulse time and exposure. Electron microscope analysis (TEM) tells us that particles have a core-shell structure in which the SiO2 core is coated by a homogenous TiO2 layer. To show that this deposition technique can also be used to deposit conductive coatings on nanoparticles, SiO2 nanoparticles have been coated with conductive TiN layers. In this case TDMAT and ammonia (NH3) were used as precursors. The growth rate of TiN showed the saturation plateau that is typical for ALD growth but depended on the amount of ALD cycles: more cycles led to a lower growth rate. This decline in growth rate can be attributed to the formation of reaction by-products that can adsorb on the particle surface and hence block the adsorption of precursor molecules. The TiN-coated nanoparticles did show good electrical conductivity, with a resistance that depended strongly on the deposition conditions. The results of this research are a step towards the use of FB-ALD in the synthesis of core-shell nanoparticles, with batteries and nanostructured third generation solar cells as the most promising applications. Future research should focus on technological challenges of the FB-ALD technique itself and, on a fundamental level, on optimization of the core-shell structure. The fundamental questions relate to the electronic structure of the nanoparticles: the behavior of nanostructured materials can be fundamentally different from behavior of "normal materials”. Fundamental studies, based on calculations and simulations of electronic structure, can determine the ideal core-shell material combination for each application. The particles can be synthesized in a fluidized bed ALD reactor. In the further development of the FB-ALD technique, safety, with respect to the processing of loose nanoparticles, will be the most important aspect to be looked at, especially when dealing with (nano-) toxic materials and materials that are easily oxidized in air. This is mostly important for loading fresh particles and in loading of processed particles. The importance of a proper loading and unloading procedure has been demonstrated with the TiN-coated particles that spontaneously ignited when they came in contact with air. Another, more practical challenge is controlling nanoparticle agglomeration and maintaining a stable, homogeneously fluidized particle bed at large scale. Several techniques are available, besides the vibrating fluidized bed described in this thesis, to break agglomeraties and create an homogeneous fluidized bed at lab scale. Scale-up of these techniques should be investigated. Furthermore, the static head (pressure drop) of large scale fluidized beds is often higher than the desired absolute operating pressure of the ALD reaction. The influence of the relatively large pressure drop over the bed on fluidization homogeneity should be thoroughly investigated. Despite the technological challenges that come with scaling up of the technique, FB-ALD is a promising technique for the production of core-shell nanoparticles. The flexibility in materials selection, both for the core and the shell, and the homogeneity and quality of the coatings will provide a large advantage over other techniques.