Ruthenium clusters in lead-borosilicate glass in thick film resistors
- PDF / 3,111,157 Bytes
- 13 Pages / 576 x 792 pts Page_size
- 65 Downloads / 185 Views
An interparticle glass matrix in ruthenium dioxide-based thick film resistors has been studied intensively by means of analytical and high resolution transmission electron microscopy. The ruthenium dioxide phase interacts with lead-borosilicate glass at high temperature by dissolving ruthenium ions and incorporating a small number of lead and aluminum ions on the surface. Ruthenium ions diffuse through the glass network at least over a distance of 1 fim during firing, but are accommodated in the glass structure by an amount only less than 7 at. % at room temperature. High resolution electron microscopy reveals numerous ruthenium-pyrochlore crystallites in high-lead glasses, but hardly any Ru-based clusters/crystallites in low-lead glasses, where lead-rich glass clusters due to glass immiscibility and reduced lead metal clusters are more commonly observed instead of ruthenium clusters. Lead oxide is prone to reduction both in high- and low-lead glasses upon irradiating with a high-energy incident electron beam. Comparison with gold-based resistor and estimation of average dispersion length of ruthenium clusters, 2 to 4 nm, prefer the model of electron hopping via ruthenium clusters/crystallites as a dominant conduction mechanism in thick film resistors.
I. INTRODUCTION A mixture of fine and conductive oxide particles and insulative glass frits, screen-printed and fired typically on alumina substrates, is called thick film resistor (TFR). It provides inexpensive, reliable, and high-capacity microresistors in hybrid IC's and as chip resistors. In the thick film technology, it is very important to clarify the electrical conduction mechanism, because many of the output electrical properties, such as temperature coefficient of resistivity (TCR), current noise, resistance to electrical overloadings, etc., depend significantly on the structure property of the electrical conduction paths. As transmission electron microscopy (TEM)1-9 of the TFR substructure reveals that conductive RuO 2 particles and their agglomerates are always surrounded by glassy matrix, it seems unlikely that the conductive particles would touch each other to form chains of overall conduction paths connecting one edge to the other of TFR, which was the postulation in the early work.2 Instead, locally formed particle-touching paths are in most cases interrupted by thin glassy regions, and the system for the conduction may be considered essentially as a metalinsulator-metal (MIM) system.3 Prudenziati3 considers that models of tunneling,4 hopping,5 and narrow conduction bands6 are possibilities for the essential conduction mechanism in TFR. Pike and Seager4 showed a mechanism of tunneling conduction between conductive particles with resonant centers scattered in the glass matrix. Charge accumulation on each conductive particle was assumed to account for 1866 http://journals.cambridge.org
J. Mater. Res., Vol. 9, No. 7, Jul 1994
Downloaded: 17 Mar 2015
the negative change of TCR. Forlani and Prudenziati5 have explained this from the viewpoint of the hopp
Data Loading...
