We started our studies by further investigating and developing the device architecture from which we obtained our preliminary results. We quickly discovered that part of the basic physics eluded us, prompting us to launch a new round of fundamental research to elucidate the operating mechanisms. It turned out that we were wrong in the sign of one carrier temperature gradient at the basis of the photon-electricity conversion: the flow of electrical current from this gradient was opposite to what we initially thought. We were able to understand the physical reasons as to why this sign was opposite to what we expected. A practical consequence of this finding is that the flow of electrical current stemming from this gradient was actually opposite to that stemming from another gradient present in our devices: the carrier density gradient. In other words, the two gradients in these devices were actually countering one another, lowering the effective output electrical power.
This realization naturally prompted us to rethink our device architecture, so as to force the two gradients to act cooperatively rather than destructively. The new architecture changes by the geometry of the metasurfaces. We also invented a way to make the layer of CQDs even more efficient. All in all, the improvements allowed us to raise the open-circuit voltage to 20 µA/cm2 for a 40-nm thick device, representing a 100-fold enhancement compared to the original design presented in the proposal.
Even with this 100-fold enhancement, the efficiency of the device remained below 1% twelve months after the starting of the project, prompting us to focus on the development of photodetectors rather than solar cells during the last six months of the project. We identified a practical application for which our devices were uniquely fitted: photodetectors that can be integrated into places that are usually not compatible with conventional detectors.
The outcomes of the project are a better photovoltaic architecture (but still not satisfying in terms of measured efficiency) as well as a demonstrator of a near-infrared photodetector with unique integration capabilities.