Physics and Technology

Physics forms the basis for many quantitative energy estimates. For instance, it allows to compute the power reaching the earth from the sun. From this the potential for most regenerative sources (solar, wind, water) can be estimated. It can also serve as a driver for technological innovations. I only highlight a few aspects of my particular interest.


In terms of their electrical conductance materials can be classified as conductors (metals), semi-conductors, and insulator. For instance, copper is a good conductor and hence is used frequently in power cables. Doped silicon is one of the semi-conductors, which have many applications in electronics, e.g., for gates in integrated circuits, but also photovoltaic cells or light emitting diodes (LEDs). A different class of materials, so-called superconductors, was discovered in 1911. These materials can conduct electricity without any resistance. This phenomenon is useful for many applications, including power transport and the generation of strong magnetic fields. However, many of these materials need to be cooled down to very low temperatures (liquid nitrogen temperature 77K, liquid helium 4K), which inhibits their technological use. Superconductivity is a very active field of research both in physics and engineering.


Thermoelectrics are materials which can convert thermal energy into electrical energy. How effectively they manage to do that is described by the figure of merit ZT. ZT is proportional to the square of the Seebeck coefficient S, linearly proportional to the the electrical conductivity and the temperature T, and inversely proportional to the thermal conductivity. The Seebeck coefficient of thermopower is a measure of the thermoelectric voltage in response to a temperature difference. The quest is therefore to find material with high Seebeck coefficient and electrical conductivity but low thermal conductivity. Some materials have been shown to have a figure of merit ZT>1, which could make them interesting for applications. Since waste heat is abundant, thermoelectrics have enormous potential to change the energy landscape. It has, however, remained very difficult to find materials with the right properties.

Strongly correlated materials

One fundamental concept in statistics and machine learning is that of correlations, e.g., between features, or between features and a target. Interestingly, in condensed matter physics there is a class of materials, which is called strongly correlated. These are materials where the behavior of the electrons is far from independent. These material show a very diverse electric and magnetic behavior and often complicated phase diagrams. Interestingly, one class of superconductors, so-called high temperature superconductors, based on copper-oxide planes are strongly correlated materials with many exotic properties.