Search

Focus of this work is the design and characterization of a versatile coating system based on magnetron sputter deposition. This technology consists of a rotary target (primary target) that will be sputtered at one position and also can be coated at a different position with a secondary material by another sputter process. This simultaneous operation and the serial order of two sputter processes is the serial co-sputter process. The flexibility of this deposition system is proven by the possibility to combine different materials and to vary the stoichiometry continuously. The serial co-sputter process allows the fabrication of alloys which are not available as a target for manufacturing reasons. In addition the two sputter processes can operate in separated atmospheres. A highly elaborated gas separation allows the operation of the primary sputter process in a reactive gas atmosphere whereas the secondary process can be driven in a non-reactive atmosphere. Compared to conventional co-sputtering the gas separation enables a stable operation of the secondary sputter process even if reactive gas is added to the primary sputter process. To develop an understanding of the process dynamics of serial co-sputtering the rate enhancement of metal oxides by co-doping with heavy atoms has been investigated first. Sputtering is a highly inefficient process because most of the electric power will be lost in collision cascades in the target by target heating. If heavy elements are added to the target material the collision cascades can be reflected back towards the target surface by a more efficient momentum transfer and therefore increase the sputtering rate. Especially for metal oxides the sputtering process is accompanied by an order of magnitude lower sputtering rate due to a higher surface binding energy of the oxide. The addition of heavy atoms can be achieved by serial co-sputtering. In the secondary sputter process the heavy element will be sputter deposited onto the rotary target. When entering the erosion zone of the primary sputter process the heavy atoms will be partially sputtered away and partially recoil implanted beneath the target surface. The later will contribute to the sputter yield amplification effect described above. In this work the sputter yield amplification effect has been investigated for the metal oxides Al2O3 and TiO2 by co-doping of a aluminum and titanium rotary target with the heavy element tungsten (Z=74) and bismuth (Z=83). The primary process variables are the O2-gas flow which determines the working point of the primary sputtering process, the rotation speed of the rotary target and the sputtering power of the secondary process (dopant concentration). The investigations show that already for low W-dopant concentrations < 2 at.% in sputter deposited layers a significant rate enhancement is possible for both metal oxides (70% for Al2O3:W and 160% for TiO2:W). For co-doping with the heavier element Bi no rate enhancement is observed. This result is surprising at first and is in contrast to earlier simulations. Our experiments show, however, that not only a high mass contrast of the collision partners is required for an improved momentum transfer but also a higher surface binding energy of the dopant to the target surface is necessary. A low surface binding energy leads to a high sputter yield of Bi, thus Bi is quickly sputtered away before it can contribute efficiently to the sputter yield amplification effect. This work shows the successful design, production and characterization of a serial co-sputter setup and the development of a first understanding of process dynamics in serial co-sputtering. Furthermore the first proof of the sputter yield amplification effect for metal oxides in a magnetron sputter deposition process is provided.

In dieser Arbeit wurden Polymer-Metall-Nanokomposite durch gleichzeitiges Sputtern von Silber und Polymeren aus zwei unabhängigen Magnetron-Sputterquellen hergestellt. Durch die vernetzte Struktur und die dielektrischen Eigenschaften eignet sich PTFE zur Verwendung als Matrixmaterial. Der Metallgehalt der Kompositfilme wurde über die Messung von EDX-Intensitäten ermittelt. Die Größe der Nanopartikel wurde mit Hilfe von TEM und XRD bestimmt und liegt im Bereich von 10 nm. Die Kompositmaterialien zeigen an der Perkolationsgrenze einen schlagartigen Übergang der Eigenschaften vom polymer- zum metallähnlichen Verhalten. Der spezifische Widerstand fällt in einem Bereich zwischen 35 und 40 % Silbergehalt um mindestens neun Größenordnungen ab. Die optischen Eigenschaften ändern sich mit dem Metallgehalt. Die durch Plasmonenresonanz verursachte Absorption, die bei kleinen Füllgraden bei ca. 400 nm ihr Maximum hat und eine intensive gelbe Farbe des Kompositmaterials zur Folge hat, verschiebt sich mit zunehmendem Silbergehalt zu längeren Wellenlängen. Die Unterschiede im optischen Verhalten der Nanokomposite und der reinen Polymere lässt sich bei der Herstellung von Bragg-Reflektoren ausnutzen. Leichte Änderungen im Metallgehalt können bei den Polymer-Silber-Nanokompositen im Bereich der Perkolationsschwelle also eine drastische Änderung in den Eigenschaften bewirken, wodurch Anwendungen als Sensoren denkbar sind.