Historically, cooling of components were realized at system level by using fans and bulky heat pipe and heat sinks. However, there is a fundamental limit with respect to how much heat can be extracted by such cooling systems. Further, such methods do not address the cooling at the source of the problem: the transistor. With the advent of novel materials and silicon fabrication techniques, and accounting for model-based design optimization, new possibilities to increase the cooling efficiency both at device and package level are emerging.
In power electronics, thermal management is key to reliability and performance of the components. E.g. for wide band gap transistors, the maximum temperature of the junction that can be reached by the device also dictates the final performance. For classical silicon technologies, despite all the efforts made to reduce the power, spreading heat uniformly still remains a difficulty due to portioned nature of the circuitry. For upcoming 3D integrated chips and devices, the thermal management remain a bottleneck due to the very poor heat conduction capabilities of stacked dice or device.
Main challenge is to enable a fast thermal dissipation of the hot spots of an IC or a power device, this can be achieved by certain materials like Aluminum nitride or Diamond, which have anisotropic heat conductance behavior and can help to dissipate the heat quickly far away from the device hot spots. Then the high amount of energy generated by the operation of the component need to be dissipated to realize maximum heat transfer from hot spot to the exterior of the system. Advanced numerical topology optimization is envisaged to improve the conceptual design (see e.g. Figure).
Figure: Optimal cooling with two different materials in a block heated on the right surface: optimal material growth (left); resulting temperature profile (right), from D. Vanderwaerde, MSc Thesis, Energy Engineering, KU Leuven (2012)
In order to support its research activities in this growing field, imec is looking for a PhD candidate to address these heat dissipation challenges. The goal of this PhD is to assess the different possibilities in integrating novel materials as heat spreader, either prior to transistor fabrication or at package level. The candidate will study theoretically and by means of computational engineering, the benefits and drawbacks of different material integration approaches. The candidate will then select the most adequate one, eventually supported by an automated optimization procedure. The experimental characterization of a prototype will also be part of this PhD. Together with the process integration and reliability teams, the candidate will come up with innovative designs and manufacturing methods using new integration technology.
Engineering, material science, physics, or related.
Type of work:
20% literature study, 50% process integration, 30% characterization.
Supervisor: Tine Baelmans
Daily advisor: Philippe Soussan and Deniz Sabuncuoglu
When you apply for this PhD project, mention the following reference code in the imec application form: ref. STS 1704-20.