When Diana Virgovicova was 14 she travelled to India with her mother. She saw rivers so polluted they were black.
“That’s when the idea was born,” she says. “I decided I wanted to spend my life cleaning dirty water.”
When she returned home to Slovakia from her trip, she contacted a chemistry professor, who told her about how quantum chemistry could be used to model new molecules before they are synthesized in the lab, such as ones that could help purify water. She spent the spent the next three years experimenting with quantum chemistry software and at just 17 found a new type of molecule that had the potential to purify water when exposed to sunlight.
Now a student at the University of Toronto, Virgovicova has also founded a company, Xatoms, that uses quantum chemistry and artificial intelligence to continue to discover and refine new materials to clean water of specific contaminants like pesticides, herbicides, viruses and bacteria.
The materials are in a powder form, added directly to the water. They are photocatalytic, so when exposed to sunlight they create highly reactive radicals which react with and eliminate the specific organic pollutant they have been designed to target, while also absorbing volatile chemicals that cause unpleasant odours or tastes. The powder can then be filtered out, leaving clean water behind. The company also plans to develop contained filters using the same materials.
Dennis Salahub, a computational chemist at the University of Calgary, says quantum chemistry has been used for decades to calculate the properties of molecules, or predict the properties of new molecules. It allows chemists to understand things like what the electrons are doing in a given molecule, to predict the energy levels of those atoms, or how the material will interact with light.
“It’s a screening process to save time in the lab,” he says. “You can do the calculations much faster than synthesizing and testing new molecules.”
Over the past decade or so, artificial intelligence and machine learning have revolutionized quantum chemistry, says Salahub. Chemists can now use machine learning to trawl through huge databases of chemical properties to help guide the choice of molecule to be subjected to quantum chemistry calculations, drastically reducing the time required to come up with new materials.
“The use of machine learning tools is well-advanced,” he says. “It’s highly likely Xatoms will be able to make good use of them.”
Virgovicova says Xatoms’ materials are perfect for use in the developing world because they are easy to use, low cost, and don’t require electricity. They have launched a pilot project in Kenya, where they are working with a local company and communities to find ways to provide affordable access to the materials. Another pilot in South Africa, in which they will test the technology in a flowing river, is expected to launch later this fall.
The company is part of U of T Scarborough’s The Bridge start-up accelerator program, and is supported by the 776 Climate Fellowship and Compute for Climate, a fellowship from by the International Research Centre on Artificial Intelligence, backed by Amazon Web Services and UNESCO. Computer for Climate aims to help founders capitalize on their momentum, and get their “groundbreaking” companies off the ground, says Kathryn Van Nuys, head of startup business development at AWS.
Virgovicova says this support, for the company’s R&D as well as business development, will help her focus on the company and begin helping people, especially women and girls, around the world who need access to clean water.