Platinum Metals Review
Total Page:16
File Type:pdf, Size:1020Kb
PLATINUM METALS REVIEW A quarterly survey of research on the platinum metals and of developments in their applications in industry VOL. 11 OCTOBER 1967 NO. 4 Contents Ruthenium Oxide Glaze Resistors Cobalt-Platinum Alloy Magnets The Platinum Metals in Fuel Cells Further Expansion in Platinum Production The Platinum-Molybdenum System Electron Configuration and Crystal Structure of Platinum Metal Alloys Brazing Graphite to Metals The Structure of Supported Platinum Catalysts Thc Platinum Metals in Catalysis Iridium Coatings in Ion Engines Carbonyl Halide Complexes of the Platinum Metals Performance of Platinised Titanium Anodes Abstracts New Patents Index to Volume 11 Communications should be addressed to The Editor, Platinum Metals Revimv Johnson, Matthey & Co., Limited, Hatton Garden, London, E.C.1 Ruthenium Oxide Glaze Resistors NEW SCREEN PRINTING PREPARATIONS FOR THICK FILM CIRCUITRY By G. S. Iles and Miss M. E. A. Casale, B.s~. Research Laboratories, Johnson Matthey & Co Limited many years for discrete component manu- The rapid deoelnpment of thick jlrn facture, and these are now employed in integrated circuits has created a need for thick film integrated circuit production for preparations that will provide resistor connections and interconnections. Probably Jilrns on a variety of substrates. In the the most important new requirement for decelopment of the new rangp of integrated circuits was a preparation capable rnateriuls described in this urticle of producing resistive films. This problem adimntagr has bern takpn ofthr complex has been approached by developing suspen- rnrrhanisrn of conduction through sions, usually of powdered glaze (frit) and ruthenium dioxide. powders of one or more noble metals dis- persed in an organic medium. After screening to the substrate, the preparation is fired to The past few years have witnessed mounting burn away the organic material, fuse the interest in integrated circuits and there is now glaze component and complete any other little doubt that within the next decade a reactions necessary. By varying the composi- substantial proportion of electronic equip- tion, a variety of different film resistivities can ment will be based on them. be produced but if close limits of resistance Conventional circuits are normally as- are required, they can be achieved by removal sembled from discrete components by solder- of part of the resistive film after firing. ing them on to a printed circuit board. In integrated circuits, on the other hand, the Until recently the majority of the resistive circuit elements are deposited as films on to preparations available required a temperature substrates, a number of which are often of 700°C or above, this high firing tempera- assembled together. It was first believed that ture being necessary to complete reactions vacuum deposition was the ideal technique within the preparation. This not only im- for producing these circuits, but within the posed the necessity of very close control of past year silicon integrated circuits and, to a furnace atmosphere and of the firing cycle, lesser extent, thick film circuits, have gained but also limited the choice of substrate to considerable ground. Here the elements and materials such as high-alumina ceramics their connections are applied as pastes to the capable of withstanding this firing tempera- substrate by screen printing and subsequent ture. High surface finish of the substrate is firing. While the circuits so produced are necessary for this work, and mica and most sometimes bulkier than their thin film glasses, which inherently have high surface counterparts, they have the advantage of finishes, were ruled out. simpler and well-established manufacturing Against this background the Johnson techniques, greater versatility in manufacture Matthey Research Laboratories have de- and fewer problems in making connections. veloped glaze resistor preparations based on Silver and gold preparations capable of ruthenium dioxide. The objective was an ink being screen printed have been available for incorporating a glaze based on a fully-reacted Platinum Metals Rev., 1967, 11, (4), 126-129 126 One of the new Johnson Matthey preparations based on ruthenium oxide has been used in the production of these resistor plates by silk screening and $ring. The substrate was mica, which required no sur- face treatment. One of the assembled but unencapsulated circuits is also shown in the photograph. preparation that would be far less dependent and his co-workers at Philips in 1950 (I). It on firing conditions than those hitherto was shown that introduction of suitable ions available. To be viable the material had to into the lattice structure of a variable oxide satisfy three other conditions : could, without deforming it, balance the ions (I) The metal/glaze system had to be of deviating valency already within the lattice capable of producing a wide range of and still maintain overall neutrality. For resistivities. example, Verwey obtained a composition (2) The films had to have acceptably low Li8+Ni2+(l-,8)Nia3-0 by calcining lithium temperature coefficients. carbonate with nickel oxide at IZOOTunder (3) The ruthenium had to be used as oxidising conditions, The product had the economically as possible. same structure as nickel oxide, but with a smaller unit cell, and the Ni3+ content was Conduction through Ruthenium broadly equivalent to the amount of lithium Dioxide oxide added. Ruthenium dioxide is a black, electrically This suggested that valency variations in conducting crystalline solid with the rutile ruthenium dioxide might be controlled by a structure. Unlike palladium oxide, it can be similar “doping” technique, leading to a heated in air to 110o”cwithout physical or better reproducibility from batch to batch, chemical change, and is almost completely together with a measure of control over both insoluble in a wide variety of frit and glass resistivity and temperature coefficient. compositions. It can seldom, if ever, be prepared as Control of Valency stoichiometric RuO,, and is usually partially The oxides of Group Va metals were defective in oxygen, with a corresponding selected for investigation. Pentavalent ions amount of Ru3+in place of Ru4+in the crystal would be necessary to balance the Ru3+ions lattice. Valency control within narrow limits in the lattice and maintain overall neutrality, was obviously necessary if stable resistors and M5+ions of Group Va metals have a based on ruthenium dioxide were to be radius within &IS per cent of that of the developed. RLP ion, which is about the limit for the Work on the control of deviating valencies entry of an ion of one species into the lattice in semiconducting oxides, in particular of another in significant quantity. It was nickel oxide, was reported by E. J. W. Verwey found that niobium pentoxide could be Platinum Metals Rev., 1967, 11, (4) 127 introduced into the ruthenium dioxide lattice positive influence of the silver being compen- in quantities up to 50 per cent molecular, and sated by the negative influence of the niobium that the results obeyed Vegard's Law, which pentoxide on the temperature coefficient. states in effect that the extent of the change Thus silver provided an additional means of in lattice parameter of the host oxide is controlling temperature coefficient in addition proportional to the molecular percentage of to reducing the cost of the resistor preparation. added dopant. This linearity provides a useful means of monitoring the composition The Glaze Component by X-ray diffraction before processing into Investigation of the glaze component of the a resistor preparation. resistor compositions showed that this had a Moreover, since the temperature coefficient significant effect on some electrical properties. of resistance of ruthenium dioxide is metallic Glasses of the lead borosilicate type promoted in nature and strongly positive, introduction high positive temperature coefficients, often of a non-conducting oxide might be expected exceeding 500 x IO-~/"Cand 2000 x IO+/"C to exert a negative influence on the tem- respectively with doped and undoped ru- perature coefficient. Thus control of tem- thenium dioxide. Better results were obtained perature coefficient in addition to resistivity with zinc and cadmium borosilicate glasses. might be achieved. Further work showed that resistance values Electrical Properties were largely governed by the ratio of doped At present four basic ruthenium oxide ruthenium dioxide to glass, and temperature preparations are available commercially (2), coefficients by this ratio in conjunction with covering the range from 100 to 3000 ohms/ the molecular percentage of niobium pent- sq./mil., but it is expected that seven pre- oxide in the ruthenium dioxide lattice. For parations will ultimately be produced, firing example, ruthenium dioxide glaze films in a at 600°C upwards, to cover the range 5 to wide range of resistance values were found to IOO,OOO ohms/sq. /mil. Intermediate values have positive coefficients in excess of 1000 x may of course be obtained by blending the two 10 "OC. As the molecular percentage of standard compositions nearest to the desired niobium pentoxide in the calcine was in- resistance. creased the temperature coefficient decreased, Little difficulty should be experienced in reaching a negative value of IOO x IO-~//"Ccontrolling values to within &zo per cent with 20 per cent molecular content of of nominal, with the possibility of maintaining niobium pentoxide. better than AIOper cent with good machines Since the niobium and ruthenium oxides under closely controlled conditions. are reacted by calcination before incorpora- Temperature coefficients in the range ~ 100 tion in the resistor preparation, no reaction to f~oox IO-~/"C can be expected with sheet occurs when the preparation is subsequently resistivities from 50 to 1000ohms/sq./mil. As fired on the substrate, and electrical properties resistivity increases the temperature coeffi- were not unduly affected by variations in the cient tends to become more negative, and time of firing or in the temperature and values of +50 to -250 can be expected with atmosphere in the furnace, resistivities from 1000 to 10,000 ohms/sq./mil.