Microelectronics is currently undergoing major changes: the industry is working on promising new materials and chip architectures. But this also means that novel electronic materials must be tested carefully to ensure that they will function reliably for a sufficiently long time during operation.
Together with teams from IBM and the National University of Singapore, TU Wien has now shown that tests of novel insulators for semiconductor technology are almost always carried out in a way that does not provide reliable information. However, using a new practical method, suitable measurement data can indeed be used to estimate the actual expected lifetime of electronic components. In research and development, this procedure can therefore help identify the right materials and manufacturing techniques more quickly and with greater confidence.
Measuring until it breaks
The lifetime of a transistor depends crucially on which insulating material is used, and at which electrical voltages it is operated. “For thick insulating layers, you can simply apply a voltage, keep increasing it, and measure the voltage at which it breaks down,” says Prof. Tibor Grasser from the Institute for Microelectronics at TU Wien. “The voltage at which the material fails is then converted accordingly. Roughly speaking: if the insulating layer in the transistor is only one tenth as thick, one assumes that it will break down at the same field strength, and therefore withstand only one tenth of the voltage.”
For the nanometre-thin layers used in microelectronics, however, this approach is an oversimplification. “The breakdown field strength of thin layers is not simply a material constant; rather, breakdown is a random and time-dependent phenomenon. The failure of the layer therefore depends sensitively on the test conditions,” explains Tibor Grasser.
Breakdown does not occur instantaneously when a certain voltage is reached; instead, it takes a certain amount of time. It therefore matters greatly how quickly the voltage is increased during the measurement. Nor does the layer fail everywhere at once, but rather at specific weak points — locations where the material is not perfectly structured, but contains certain defects.
The temperature and geometry of the sample also have a significant influence on the result: the larger the area, the more weak points can be expected.
“Our experience clearly shows that if one estimates the robustness of materials for electronic components using this oversimplified method, the results are usually completely wrong,” Grasser emphasises.
This is precisely what makes the issue so serious for the semiconductor industry: incorrectly estimating failure rates in electronics is unacceptable and leads to enormous economic losses. This is especially true now, when the most suitable insulator for future technologies still needs to be identified. Reliable and rapid lifetime prediction is therefore crucial for material selection.
Until now, the only way to obtain better results was an extremely laborious testing procedure: different constant voltages are applied to material samples of different surface areas, and one waits until the insulators break down. These results can then be evaluated statistically. “With this method, however, it can take months to obtain usable data,” says Tibor Grasser.
Sophisticated statistical methods
But this does not mean that researchers have to continue groping in the dark. “We propose a more practical method,” says Tibor Grasser. “We ramp up the voltage in a precisely defined way until the component fails. However, we do not directly transfer the resulting data to electronic components; instead, we use a special statistical procedure. This shows that at least three different voltage ramp rates must be tested. From these, voltage acceleration can then be statistically linked to the probability of failure. In this way, we can calculate from the measurement results exactly the data that matter for industry.”
“This allows us to say: what maximum voltage can a certain number of components of a certain area withstand over a certain period of time? That tells us which maximum operating voltage may be used for these components,” says Tibor Grasser. “This makes it possible to compare different materials reliably — without having to fear discovering years later that the wrong materials were used, and that they reach the end of their lifetime sooner than predicted.”
Nature Electronics
Experimental study
Not applicable
Industrial reliability testing of transistor gate dielectrics
24-Jun-2026