Reducing the supply voltage is one of the most pressing issues for the further scaling of CMOS logic transistors since their energy efficiency is directly linked to the supply voltage. However, the carrier distribution in a semiconductor limits the steepness of the on/off transition in a MOSFET to 60 mV/dec at room temperature, the so-called Boltzmann limit. Therefore, the supply voltage cannot be reduced without significantly increasing the off-state leakage of the device. In the year 2008, Salahuddin and Datta proposed to use a negative capacitance created in a ferroelectric layer to get an internal voltage amplification and by this increase the on/off switching steepness below the Boltzmann limit.
This proposal on the one side created a new research direction and on the other side led to an intense scientific dispute. While first simulation studies using simplified models were very useful to find the expected fingerprint of the device that was called negative capacitance FET after the pioneering papers, experimentalists immediately went on to put ferroelectrics on field-effect transistors, creating a situation that could not be easily understood by simple models anymore. As a result, many papers claiming alternative explanations for the observed results appeared. Other researchers go much further and doubt that the theoretical basis of negative capacitance in ferroelectrics is on solid ground at all. In many cases, it is nearly impossible to confirm either the optimistic or the pessimistic view on negative capacitance since real world devices are much more complex than the simplified models used in many simulations. It is clear that a more nuanced understanding of negative capacitance is needed.
When we received the opportunity to write a comment on this topic for Nature Electronics, we tried to take a step back and look at the broader picture and the possible next steps. We observed that there is a significant divide in negative capacitance research: On the one hand, some groups have focused on the fundamental understanding of diverse negative capacitance effects mainly in ferroelectric model systems. On the other hand, the majority of research has focused on experimentally realizing practical devices, so far with limited success. In our article, we emphasize that this slow progress seems to be a direct result of the general focus on the application without having a thorough understanding of the negative capacitance effect in practically relevant (HfO2 based) ferroelectric materials.
We felt that the community needs guidance on which problems to focus on moving forward. As our own research is situated at the boundary between the more fundamental material science/physics and the more practical device engineering side of the topic, we felt that we have a unique perspective in this regard. Our goal was to encourage these rather separate communities to come together and bridge this gap in our understanding of negative capacitance. Only once we have established a basic understanding of the effect in HfO2 based materials, we might be able to overcome the Boltzmann limit in electronics and might also enable applications beyond negative capacitance FETs.