www.inseto.co.uk POWER MODULES 17 www.power-mag.com Issue 2 2023 Power Electronics Europe The Pressure is On John Govier, Sales Director of Inseto, summarises how far sintering has come and hints at what’s in store. The EV sector is constantly striving for higher power densities (in terms of W/m3) and in particular with regards to the electronics for controlling power, both within vehicles and within charging stations. Reducing the volume of a component reduces the quantity of materials to build it, thus reducing production costs. Power modules, as used in inverters for example, are required to switch high voltages at high frequencies into loads that draw hundreds of Amperes. These requirements have made silicon carbide (SiC) the semiconductor material-ofchoice, and EV power modules typically contain several SiC-based MOSFETs or IGBTs. In addition to the requirements to switch more power, there is also a need for deep power cycling and providing high reliability operation in relatively harsh environments. All of these factors place challenges on device packaging - certainly if industry is to get the most from all that SiC has to offer. Although SiC devices boast low power losses these are relative to very high (achievable) power densities. Essentially, SiC-based high power switching structures run hot, and the heat must be dissipated. Moreover, SiC-based die can run far hotter than the melting point of even specialist solders. Sintering – Materials & Processes As readers will be aware, pressure sintering is an alternative to soldering but let’s take a detailed look at the materials and how they affect/govern the sintering process. A sinter paste comprises monometallic particles (typically less than 1µm in size), an organic compound (with an evaporation temperature of about 150˚C) and possibly an oxygen reduction agent. Sinter paste OEMs have their own recipes but common to all is that the monometallic particles account for about 90% of the volume. The material of choice is currently silver. However, all paste OEMs are conducting R&D into copper. The cost of copper is about 5% that of silver, but its use presents many challenges because of how readily the metal oxidises, preventing the formation of good mechanical bonds – unless the entire process is performed in an inert atmosphere. The most popular method of applying the sinter paste to the substrate (which is typically direct bonded copper, DBC) is using a stencil, where the holes are usually 5 to 10% larger than to die to be placed. The applied paste will typically be between 100 to 120µm thick, though some device manufacturers are experimenting with 80µm. The next stage is pre-drying. This tends to be in an oven with a nitrogen atmosphere. The substrate is heated to between 120 and 130˚C for 30 minutes. This removes any moisture present and results in the thickness of the paste reducing by about 20% while still remaining tacky, as the organic compound is still present. After drying, the substrate is moved to a pick and place machine. Here, many OEMs are using hot-head tools (at about 100˚C) as it improves the bond quality. Note: the underside of the SiC die is already metalized with silver. Next, it’s to the sinter press. If the substrate is large, it is advisable for the press to have a pre-heating stage. This reduces the risk of thermal stress (warping). Also, if the substrate is a large Figure 1 – Above, a cross-section of a BLT. Figure 2 – Equipment used in the fabrication of a power electronics component/module requiring die to be sintered to the substrate. Not shown, but a recent industry development is the use of a foil with a similar composition to a paste. Printing and pre-drying can be skipped, but the foil is expensive and requires special handling. The technique is more for R&D work than volume production.
RkJQdWJsaXNoZXIy MjQ0NzM=