Turning Skyscrapers into Power Plants Through Quantum Dots


With the need for renewable energy sources, new buildings are working towards being self-sufficient in energy. Many research projects are investigating nanocrystal structures, highly luminescent quantum dots, and LED and PV devices within the buildings and construction sector. The vision is that as part of the construction design and build, regular windows will be turned into semi-transparent solar panels.


One of the approaches involves the development and optimization of light-emitting and photovoltaic devices with the aid of machine learning (ML), and the support of multiple online materials databases positioned around the globe. The project, called AI4QD, united three partners: (1) The Italian Institute of Technology (IIT); (2) Glass to Power (GTP), an Italian SME focused on the development of building integrated photovoltaic windows based on semiconductor colloidal nanocrystals. (3) MaterialsZone (us).

With the help of the Materials Informatics Platform and scientific knowledge of all partners, the project applied machine learning (ML) algorithms and online notebook management tools, to the research on colloidal nanomaterials and their applications through a ‘reverse tailoring’ approach.

MaterialsZone has exploited large databases and suggested intelligent protocols for the synthesis of novel and improved crystals that are suitable for applications in the field of optoelectronics, such as highly luminescent quantum dots and LEDs.


Targeting the synthesis of novel materials via a ‘reverse tailoring’ approach, the partners used artificial intelligence (AI) for optimizing existing synthesis and fabrication protocols. Already employed at an industrial level, in order to obtain higher performance of the materials and devices, in particular for luminescent nanocrystals for solar concentration.

advanced materials research, as performed at the IIT is now FAIR (Findable, Accessible, Interoperable and Re-usable), and facilitates the interaction of a large set of parameters (material choice, synthesis conditions, process type, the composition of the device) that are tuned based on empirical knowledge, and intuition of the scientists. Here, the ML and optimization algorithms of MaterialsZone provides a completely new approach towards the development of new materials, and the design of efficient LEDs and solar harvesting.


Through experimental studies and theoretical predictions, the project was successfully identifying key parameters that affect the light concentration capacity of the nanocrystals. Luminescent solar concentrators (LSCs) that absorb and concentrate sunlight are mainly made from semiconductor nanoparticles that can obtain high-performance values while keeping the production costs as low as possible.

In conclusion, the AI4QD project is an ongoing process of optimizing the device performance as well as the yield (cost and material efficiency) of the synthesis of the colloidal materials that constitute a major activity towards to EU’s sustainable living goals.