The GEM (Getting the MOST out of the Sun) Project – From computational chemistry to the generation of sustainable energy

Through high-performance computing technologies applied to the study of a family of “special” molecules, the research project funded by the IFAB shows that a sustainable revolution in energy production, storage and distribution systems is possible.

The challenge

Sustainable energy production is the true challenge of our time, which is why it is increasingly the focus of scientific research and technological development. The “GEM” (Getting the MOST out of the Sun) project, which the IFAB chose through its “Call for Projects 2022”, is based precisely on this challenge and aims to radically innovate the solar energy storage scenario. The study, conducted by researchers and professionals at the University of Bologna and E4 Computer Engineering, aims to use artificial intelligence to identify and select (from millions of potential molecules) the most promising candidates for “MOST” – Molecular Solar Thermal – applications, i.e. molecules able to store solar energy that can be used as sources of heat and energy.

The solution

The project’s starting point is azobenzene, a molecule that is able to absorb sunlight (which is why it is used as the basis for many dyes) and release it at a later time in the form of heat. What makes azobenzene and its derivatives particularly interesting is not only their ability to absorb sunlight, but also the possibility of converting it into thermal energy and of storing this energy for long periods (even for several years, in some cases) before it is released. To give a readily comprehensible example: the principle by which the molecule functions is similar to that of electric batteries, which store their chemical energy that can be converted into electrical energy as required.

The “GEM” Project uses artificial intelligence and computational chemistry tools at various levels: firstly, to create a vast database of molecules with a similar chemical structure to azobenzene; then to analyse the effectiveness and performance of these molecules without having to resort to onerous synthesis operations and laboratory testing. The outcome of the research is potentially revolutionary, as not only does it make it possible to identify among a great many potential molecules those that are most efficient for energy storage purposes but, thanks to artificial intelligence, it is even able to invent new molecules (not present in the initial database) that are even more effective.

The “GEM” Project consists of three phases:

  • Phase 1 – “First principle predictor”: by applying computational chemistry tools (appropriate software and supercomputers) a database of thousands of azobenzenes and their physical properties of relevance for their use in MOST applications is compiled.
  • Phase 2 – “Artificial intelligence predictor”: having been trained on the database compiled in Phase 1, the artificial intelligence selects the best-performing molecules in terms of the storage and subsequent release of thermal energy.
  • Phase 3 – “Molecular creator”: the artificial intelligence uses the information obtained to create new molecules (not present in the database compiled in phase 1) that offer even higher performance.


The possibility of storing solar energy and releasing it, in a controlled manner, in the form of heat is in itself a revolutionary possibility for man, science and society. More specifically, a MOST material can be used in three particular settings:

  • the creation of organic solar collectors based entirely on MOST technology, which starting from the accumulation of solar energy will provide, in a controlled manner, clean thermal energy to large and small, public and private buildings;
  • combining MOST organic solar collectors with conventional ones, to create a system able to absorb more energy with the same surface area;
  • the development of technologies able to produce electrical current from the thermal energy stored in MOST materials, thanks to thermoelectric chips;
  • the generation and management of thermal energy without combustion, with low production costs and without using heavy and rare metals.

The results that can be achieved with the GEM Project are just one of many examples of how an interdisciplinary approach between chemistry, physics and artificial intelligence and partnerships between the world of research and industry allow us to transform innovation into reality, with a positive outlook towards the future.


  • E4 Computer Engineering S.p.A.
  • University of Bologna

Sustainable Development Goals

Stay updated on the latest IFAB events and projects.Subscribe to our monthly newsletter