Date of Award
Master of Science (MS)
School of Materials Science and Engineering
Kyle S Brinkman
Advances in integrated circuit (IC) fabrication has led to microelectronic devices with sub-micrometer size features. These small features have been used to enhance the functionality of devices including faster processing. At the same time, this feature miniaturization has also caused unintended reliability concerns from corrosion. While the susceptibility of materials to degradation has not changed, the small length scales have meant that a shorter time of or amount of corrosion can lead to device failure. Prior work in understanding corrosion in microelectronic devices has included classifying failure mechanisms in microelectronic devices. These works indicated that the primary factors contributing to corrosion events included moisture in the operating environment, ionic contaminants diffusing into the devices, small distances between metallic lines that created large electric fields, and electric contact between dissimilar materials.
While there has been substantial research on bulk material corrosion mechanisms, more work is needed to understand the corrosion of films and features. The intention of this study was to understand the electrochemical performance of Ti and TiNx films. This material system was selected due to its wide appearance in microelectronic devices as adhesion layer and diffusion barrier layer for the metallization. Ti/TiN layered systems also appear in other coating applications such as wear-resistant coatings, where the architectural arrangement of layers has been of interest to researchers. These layered configurations could lead to both galvanic and interfacial corrosion.
The aims of the research were to (1) understand how the pH of electrolyte influences the electrochemical behavior of Ti and TiNx thin films independently; (2) evaluate the stability of electrochemical performance of the Ti and TiNx films during 105 days of exposure to each electrolyte; (3) study effect of galvanic coupling on electrochemical performance of Ti and TiNx; and (4) determine if macro-scale corrosion tests can show a considerable difference between the behavior of Ti/TiN multilayered and electrically coupled samples.
Monolithic Ti and TiN, bilayered Ti/TiN, and nanolaminated Ti/TiN with 100 nm thick layers in all cases were deposited onto (100) Si wafers using sputtering deposition. Electrolytes were 3 wt. % chloride ions with different pH values, ranging from acidic to basic. While all wafers in a process chamber were considered a single sample, we will report the tests of the replicates made from each sample. Four (4) sample replicates were kept in each electrolyte and tested independently.
Results showed more negative potentials for monolithic and electrically coupled systems in the basic electrolyte. The range of OCP values in the basic cell was -0.2 V to -0.8 V, while it was -0.4 V to +0.4 V in the neutral and acidic cells. Some instabilities were observed in basic electrolytes in terms of fluctuations in OCP curves. We expected to see some differences between the behavior of monolithic Ti and TiN because of having oxide layers of different stoichiometry and composition, and the difference between the two was pronounced in the acidic electrolyte with an order of magnitude higher polarization resistance for TiN. Electrically coupling Ti and TiN affected the behavior of Ti in acidic cell. Macro-scale corrosion tests such as OCP and LPR did not show a considerable difference between the electrochemical performance of bilayered Ti/TiN and electrically coupled Ti/TiN. However, Mott-Schottky analysis showed different flat-band potentials (-0.08 V for bilayer and -2.34 for electrically coupled system). EIS technique showed that the peaks in the low frequency range of Bode phase diagrams appeared at different frequencies for bilayer and electrically coupled films.
Najaf Tomaraei, Golnaz, "Understanding the Electrochemical Behavior of Monolithic, Bilayer, and Electrically Coupled Nano-Scaled Films in Different Environments" (2017). All Theses. 3239.