A journey across the periodic table: What the chips of the future are made of
The digital revolution is taking hold of ever broader segments of society, but without modern semiconductors in sufficient quantities it could quickly come to a standstill. In order to develop ever faster and better chips for this purpose, which in their hunger for electricity do not immediately destroy the urgently needed energy revolution, new materials beyond classic silicon are essential.
The digital revolution is taking hold of ever broader segments of society, but without modern semiconductors in sufficient quantities it could quickly come to a standstill. In order to develop ever faster and better chips for this purpose, which in their hunger for electricity do not immediately destroy the urgently needed energy revolution, new materials beyond classic silicon are essential.
They block leakage currents, i.e. unwanted power losses, ensure energy-efficient computing power for mobile use, among other things, and pave the way for green energy from wind and solar power plants to enter the grids. "Only about 25 years ago, the number of substances that were important for microelectronics was very manageable," says nanoelectronics professor Thomas Mikolajick of TU Dresden, outlining this trend. "Since then, that number has really exploded, and semiconductor technology has expanded to include a large part of the periodic table." Among them are some elements and compounds on which the industry has particularly high hopes for the near future – here’s a little "tour d’horizon":
Metal oxide semiconductors such as indium tin oxide (ITO) and indium gallium zinc oxide (IGZO).
They are seen as a beacon of hope for the three-dimensional logic and memory circuits of the next generation but one. As transparent interconnect materials, they have already been in use for years in solar cell and display manufacturing, for example. However, these metal oxides are also suitable for the 3D CMOS concept. According to expert estimates, this will become the focus of the semiconductor industry in about five to ten years, when it is about to move into the sub-nanometer world. Then ITO and IGZO could provide the key to three-dimensional stacked transistor levels. 3D architectures already exist for individual transistors such as the FinFET or for memory in USB sticks, but not for complex logic circuits. The technologies are being researched primarily at Imec in Belgium, but also at the TU Dresden.
Advantage
can be processed at low temperatures – important for layer build-up!
Disadvantage
Unlike silicon, these metal oxides are not single crystals, their charge carrier mobility is lower – i.e. it may be more difficult to achieve a high switching speed.
Applications
Highly integrated electronics for autonomous cars, smartphones, data centers, control electronics in factories etc.
Magnetic stacks with materials such as cobalt-iron-boron or iridium-platinum compounds, magnesium oxide, etc.
They should enable cheaper and more powerful memory cells.
Advantage
fast storage capability
low voltage
Disadvantage
deposition of the materials is complex and expensive
industry has not yet gained much experience with it
Applications
Memory cells embedded in logic circuits or other complex systems.
Tantalum oxides and amorphous hafnium oxide
They are suitable for memristors. This is electronics that "remembers" previous states, i.e. "collects experience" to a certain extent. Panasonic and TSMC already produce such memories.
Advantage
value for money
Disadvantage
Lasts only a relatively low number of switching cycles
industry has not yet gained much experience with it
Applications
currently in demand primarily as a low-cost alternative to flash memory cells
perspective as hardware for certain artificial intelligence (AI) computational tasks
Crystalline hafnium oxide
… in its orthorombic version (crystal with three perpendicular axes) is a ferroelectric material and a hopeful candidate for fast memories and AI hardware. The breakthrough was achieved primarily in Dresden: once at Qimonda and then at the TU’s NaMLab, Fraunhofer CNT and the "NaMLab" spin-off "FMC".
Advantages
as memory in performance and efficiency a quantum leap compared to flash
easily integrated into existing CMOS processes
Disadvantages
tolerates only a few million switching cycles
relatively high voltage compared to MRAM (about 3 volts)
Applications
very fast memory with very low power consumption
perspective: neuromorphic computing for artificial intelligence
Silicon carbide and gallium nitride
Both are semiconductors with a high band gap between the outermost electron bands. As a result, they can withstand higher voltages and stronger currents than ordinary silicon components and offer less resistance to current. In Saxony, the X-FAB plant in Dresden, among others, produces gallium nitride semiconductors.
Advantages
high reverse voltage up to several thousand volts
little resistance when switched on
tolerates strong currents
Disadvantages
more complex technology than silicon
doping is demanding
crystals are difficult to grow – therefore, in the case of gallium nitride (GaN), layer build-up on silicon wafers has been the most common method up to now
Applications
latest generation power electronics, for example rectifiers and inverters in solar and wind power plants, transformers and other power equipment, power supply equipment for consumer electronics, electric rail vehicles
GaN is also suitable for high-frequency technology in "High Electron Mobility" transistors (HEMT)
Gallium Arsenide
Gallium arsenide has been used for high-frequency electronics for years. Its use as a material for power electronics is still quite new. It is conceivable that it could be used to cover a niche between silicon in the lower power spectrum and silicon carbide or gallium nitride on the other side. Saxony is considered an important location for gallium arsenide semiconductor technology. The pioneers here include Freiberg Compound Materials (FCM) and "3-5 Power Electronics" from Dresden.
Advantages
larger band gap than silicon
higher charge carrier mobility with silicon – thus high switching speed possible
Disadvantage
more complex technology than silicon
Applications
in addition to the classics such as LEDs and high-frequency technology, power electronics diodes have also come into focus, for example for electric car charging stations
Gallium oxide
Gallium oxide could enable the next generation of power electronics after silicon carbide and gallium nitride. However, this technology is still in the research phase.
Advantage
would tolerate even higher voltages and stronger currents, the reason is that the band gap between the valence and conduction electron bands is 4800 electron volts (eV), which is even larger than that of silicon (1100 eV), gallium arsenide (1400 eV) and even gallium nitride (3000 eV)
Disadvantage
Not yet a ready-to-use technology
Application
next generation power electronics
Conclusion
High-tech companies and research institutions in Saxony are working on several of these promising material developments. One particular focus is on power electronics, among other things. Lighthouses in this segment include, for example, a gallium nitride research center in Freiberg and Freiberger Compound Materials (FMC), as well as TU Dresden’s NaMLab, Infineon’s power electronics production, X-FAB’s gallium nitride line in Dresden and others.
This article was first published as part of our NEXT magazine "In the spotlight: Microelectronics".
To the complete edition of the magazine
Related articles:
Computing trends: quantum computers, neuromorphic computers and tap-proof telephony
Germany‘s microelectronics landscape – Europe‘s semiconductor heart beats here
European Chips Act: Good, but good enough?