Effects of organic additives and glass on the properties of AlN/glass tape-casting slurries and green tapes

Abstract

Low-temperature co-fired ceramics (LTCC) technology offers significant benefits over both thick film ceramics and HTCC (high temperature co-fired ceramics) due to high-density, high-radio-frequency, fast digital applications requiring hermetically sealed packaging and good thermal management [1]. LTCC tape has been used in the last 20 years for high-reliability, avionics, and automotive applications, as well as in MCMS for communication and computer applications [2]. In recent years, many LTCC systems were made up of Al2O3 and glass [3]. However, due to the low thermal conductivity of Al2O3, the LTCC systems of Al2O3 and glass have a poor thermal conductivity. AlN, whose thermal conductivity is 10 times greater than that of Al2O3, can be substituted for Al2O3 to improve the thermal conductivity of LTCC [4]. As a convenient and low-cost processing, tape casting has been widely used to produce ceramic substrate [5]. However, LTCC is a multicomponent and complicated system. Each component has a substantial effect on the properties of the slurry and green sheet. The present work was conducted on a new type of LTCC AlN/glass to study the effects of organic additives, including dispersant, binder, and plasticizer, on the properties of the slurries. The effects of glass content on the properties of slurry and green sheet was also investigated. AlN powder and borosilicate glass powder were used as raw materials. The commercial AlN powder (Shenhai Nitride Co. Ltd., China) was synthesized by the self-propagation high-temperature synthesis (SHS) method, with nominal particle size of 0.5 μm. Borosilicate glass powder was milled to 1.7 μm. Borosilicate glass used in this study was found to react with water and, therefore, azeotropic mixtures of 66:34 vol% 2-butanone ethanols were used as solvents. Triethyl phosphate (TEP) and polyvinyl butyral (PVB) were introduced as dispersant and binder, respectively. The plasticizer was a 50:50 wt% mixture of polyethylene glycol (PEG) and phthalate (PHT). The mixed powders with dispersant were first ball-milled for 24 hr in solvent with AlN ball. Binder and plasticizers were added to the slurries and milled for another 24 hr. Then the tape was cast on a glass surface through the action of a blade that leveled the slurry. The tapes were dried in a solvent atmosphere. A rotary viscometer was used (Model NDJ-7, Shanghai Balance Instrument Plant) to measure the apparent viscosity of the suspensions at 350 s−1. Sheardependent behavior was tested with a rheometer (Model SR5, Rheometric Scientific, Inc., Piscataway, NJ). The suspensions, containing 16 ml of solvent and 4 g of the AlN/Glass powder, were ultrasonically mixed with different amounts of dispersant for 10 min. Then the suspensions were settled in 25 ml graded tubes. The density and porosity of the green sheets were measured by mercury porosimetry (Model PoreSizer 9320, Micromeritics Instrument Corp., Norcross, GA). Fig. 1 shows the viscosity of three suspensions (67 wt%) with different dispersants as a function of dispersant content. For the 1.5 wt% dispersant content, the viscosity of the three suspensions simultaneously reached a minimum though the particle sizes were different for AlN and borosilicate glass. This indicates that the adsorption of dispersant onto ceramic particles was not only related to the particle size but also to the chemical nature of the ceramic powder. Fig. 2 describes the sediment volume of three suspensions versus the concentration of dispersant after resting for 4 weeks. As is shown, the sediment volume decreased as the concentration was increased. After a minimum volume was achieved, further increase in the concentration of dispersants produced a slight increase in sediment volume. For the 1.5 wt% dispersant content, the sedimentation volumes of the three suspensions reached a minimum. This indicates 1.5 wt% dispersant content was optimum for three suspensions. These results agreed well with that of the viscosity tests, though the solid loadings of the suspensions were