Printed spark ablation nanoparticle films for microelectronic applications

This thesis is about the application and characterisation of spark ablation generated nanoparticles in microelectronics. It opens in chapter 1 with a general motivation of the need for advanced materials and for nanotechnology in particular. It then describes what nanoparticles are and why they are promising materials. Some of their advantages are a high chemical reactivity, a high specific surface area, and the display of quantum effects at that scale. The chapter ends by presenting the research questions and giving the thesis structure.

Chapter 2 provides a technological background for the rest of the thesis based. It describes several applications of nanoparticles within microelectronics not researched during this PhD project: as die-attach materials, chemical sensors, or catalysts. It continues with the description and discussion of several competing nanoparticle synthesis methods and goes in-depth on the theory of spark ablation generation. It describes the effects of various parameters that govern the mass generation rate, particle size, and composition. This theory is important to be able to interpret the results in the other chapters. Impaction deposition is then described in this chapter since it is the method of printing all samples in this thesis. It explains how this method prints dots or lines of nanoparticles, that they have a Gaussian cross-section profile, and how specific deposition parameters affect the deposit. Lastly, the chapter gives a detailed description of the spark ablation synthesis and deposition equipment with which all the experiments in this thesis are performed. The generator, components, gasses, pressures, and materials are all described with diagrams and specifications. Typical synthesis and deposition settings for the generation of Au nanoparticle deposits are given (1 kV, 5 mA, 1.5 L min.¡1 Ar or N2 and 1mmnozzle distance).

The first chapter with results, chapter 3, presents a method to measure the mass deposition rate of the nanoparticle printer. Measuring the mass of microgram scale deposits is challenging due to the high sensitivity required for an accurate measurement. Balances are sensitive to changes in pressure, temperature or humidity that can already give too big errors. One solution already applied in thin film deposition methods is the use of quartz crystal microbalances (QCMs). Their resonance frequency is dependent on their mass, and thus, we can use the frequency shift during deposition to measure a mass change. The Sauerbrey equation that is used for that conversion must be valid, so a special method was developed to comply to all of its conditions. A concentric circular pattern of Au nanoparticles was printed on 10 MHz QCMs to measure the mass deposition rate. It was found that the deposition rate scales linearly with the generation current of the spark, as expected from theory, but also showing the losses in the system are either constant or scale linearly too. The film density was surprisingly constant for all tested synthesis and deposition settings, at 15.95 g cm¡3, or a porosity of £p Æ 0.18. The density was compared to models presented in literature, and it is proposed that the impaction energy likely compacts the porous structure during deposition until this density is reached. The QCM method can be applied for process monitoring using commercially available equipment and open-source software.

The first applications of printed conducting nanoparticle films are discussed in chapter 4. It describes the conductive properties of such films and the effect of annealing on their conductivity. It was found that an untreated Au film conducts 22 times worse than bulk Au. Several applications are then discussed. Here it was demonstrated that printed Au nanoparticle lines can be applied as interconnect materials as an alternative to wire bonding. Next, a method was presented to miniaturize the deposits even further by using lithography and lift-off. This reached a line width at the minimum of the lithography equipment available, at 1.2 ¹m, without significantly changing the nanostructure.

Chapter 5 deals with the application of spark ablation generated nanoparticles as thermoelectric materials. It describes in detail the synthesis and characterization of Bi2Te3 nanoparticles and their thermoelectric properties. The main finding was that the thermal conductivity was drastically lower than bulk Bi2Te3 and comparable to the state of the art for Bi2Te3 nanostructured materials, reaching a minimum of 0.2Wm¡1 K¡1 at room temperature. Unfortunately, the electrical conductivity was reduced by at least a factor 1000, easily undoing any efficiency gains from reduced thermal conductivity. Suggestions are given to possibly improve this trade-off. Additionally, this chapter shows how quickly nanostructured materials like the ones in this thesis oxidize after synthesis. From the moment the sample is printed, it gains mass and loses conductivity, so this must be counteracted if a non-noble metal is to be applied.

The final chapter before the conclusions, chapter 6, showcases another application of printed nanoparticles: as UV-sensing material. It shows the results obtained using ZnO nanoparticles to create a UV sensor that is insensitive to visible light. The nanoparticles were deposited over electrodes to fabricate a resistor that has two orders of magnitude electrical resistance reduction when exposed to 265 nm UV light. The response was slow, with 79 seconds to reach 90% of the maximum response and 82 seconds to get back to 10% again. This is attributed to the adsorption and desorption of oxygen under the influence of UV light and can be prevented by packaging the sensor. The contact behaviour between the metal electrodes and ZnO nanoparticles proved to be too unpredictable to reliably create a Schottky diode, which would have had a higher response. This dissertation ends with a list of the conclusions, the answers to the research questions, and finally, some suggestions for future work.