Since the establishment of the first CD Laboratory at today's Institute for Microelectronics in 1989, there have been tremendous developments: Today, microelectronic components range from the smallest, highly sensitive sensors in medical technology to inverters for photovoltaic systems, which must withstand very high voltages. To date, eight CD Laboratories at the Institute for Microelectronics have contributed to this development.
The conquest of the third dimension
Then, as now, transistors have been the key components that convert applied current into logical "zeros" and "ones", forming the basis for bits and bytes. Early electronic circuits can be imagined as villages: transistors as houses, connected by roads, i.e. lines that must not be overloaded ("burn out"). Over time, microelectronics conquered the third dimension: skyscrapers and subways were integrated and villages grew into vast metropolises like Manhattan or Tokyo. But how are all these transistors optimally built, arranged, wired, and switched? The CD Laboratory for Integrated Devices, which was launched in 1989, was already working on computer-based mathematical models and simulation software for silicon enriched (doped) with phosphorus, boron, or arsenic. The behavior of these materials under applied voltage was modeled using a computer program (FORTRAN at the time). The software tools developed in this way proved to be highly successful and sought after, leading to the founding of a spin-off company in 2008 by members of the institute: Global TCAD Solutions. Today, the company has over 40 employees and offices in Taiwan and India and is now a commercial partner of a CD Laboratory.
Subsequently, the simulation tools were expanded to address an increasing range of challenges: power components, wiring, memory, and sensors - covering an ever-increasing number of materials explored and used.
The combination does the trick
The chemical bonds between atoms in electronic components can be described quantum-mechanically using the Schrödinger equation. However, even the thinnest layers used in today’s advanced devices still contain many thousands of atoms – making this equation impossible to solve for such large systems. A multiscale approach is therefore used: for a few atoms, their movements can still be calculated quantum-mechanically. At the next level, these calculations are abstracted, incorporating measurement results as well. This process continues step by step until an entire device, and ultimately even complex circuits, can be simulated. This combination of calculation and measurement has proven to be very successful in describing the behavior of components, including performance, reliability, and aging, with sufficient accuracy.
No modeling without measuring
Other CD Laboratories are working on the foundations for very powerful and fast electronic measuring devices. These are essential for detecting and localizing randomly occurring individual defects in the atomic structure: If an atom is in the wrong place or an electron "gets lost", this leads to reduced current flow at the same applied voltage. Highly refined voltmeters and ammeters allow us to observe individual electrons and compare the data with theoretical predictions, making it possible for these defects to be identified and localized.
The path to the future
The major trends in microelectronics - miniaturization, differentiation for various applications, and the use of different elements - will continue to be the focus of research in the future. AI and machine learning will also play an increasingly important role. At the same time, urgent new challenges are emerging, particularly for sustainability in production. Issues such as water and energy consumption, the release of environmental toxins, and recyclability remain critical concerns. Currently, recyclability is often hindered by the wide variety of elements used. While considerable progress has been made, the rapid growth of production has outpaced these advancements. Research is now actively addressing these challenges.