There are golden prospects in toxic mining wastewater, and Wits chemistry Master’s student Taskeen Hasrod is setting machine learning loose to dig them out.
Hasrod was chosen as FameLab SA 2023 winner of the national leg of the International FameLab science communication competition on September 21, 2023. As the national winner, she will represent South Africa in the International FameLab 2023 competition on Friday, November 24, 2023.
Her research is focused on reclaiming valuable by-products from the treatment of Acid Mine Drainage, which itself is a toxic result of the mining that built South Africa into a regional economic powerhouse. “One man’s trash is another man’s treasure,” she says, and she believes her research has the potential to unlock a circular economy.
Hasrod says she fell in love with chemistry in Grade 7, especially the idea that “you could take really big things and break them down into fundamentals which governed everything. I was fascinated that one proton could be the difference between gold or mercury, and that this tiny particle played such a big role in things.”
That fascination clearly still drives her, as she explores how to turn poison into prosperity.
She explains that “in gold mining all the rock waste is disposed in mine dumps, those yellowish mountains that we see from the highway. All that rock that was taken from underground and has now been exposed to air and rain which leaches the toxins from the rock into the underground water table. That’s your acid mine drainage and we see it coming up through the unused mine shafts predominantly and appearing as vibrant red ponds.”
She says it’s only as recently as 2010 that this water has been treated to remove the hazard to vegetation, wildlife, and people who live close to these toxic ponds.
She says “Specifically we are looking at the sulphur containing compounds left by this treatment process, and that’s then where the idea of a circular economy comes in. You’re getting clean water out of the treatment, but from what’s cleaned you can extract by-products that have commercial value and that you can sell. Since the by-products come from a waste feedstock, it’s very cheap but still has commercial value.”
This pharmaceutical sulphur, once extracted, can be used as an antiseptic disinfectant and in anti-acne treatment in cosmetics.
But the process is not straightforward. “The trick is this,” says Hasrod. “If you do it the traditional way by taking your wastewater samples to the lab and conduct experiments to figure out how to extract the sulphur, the experiments are extremely time-consuming and expensive. Avoiding this by going the computational way allows us to get results faster, and cheaper. Using machine learning, we analyse the data that we’ve obtained from the treatment plants. This analysis allowed us to develop an app that predicts the sulphate level we usually would have gotten by doing experiments. All you have to do is just determine two other values, punch them into our app, click a button, and you get the sulphate value you would have gotten from the experiment.”
And who developed the app? “Me”, she says brightly, making it all sound easy.
She says her project was to develop the app and provide sulphate data as a pre-cursor for future simulation work. “The data enables us to simulate experiments based on statistical modelling to determine which experiment is most likely to work, and only then go to the lab. After doing the experiment, product yields and commercial value can then be established. The process that we’re looking at is one that happens in nature. We’ve seen that it does work, we just have to figure out a way to do it in a lab.”
She says that one of the difficulties of being a scientist is explaining the importance of one’s work. This is why she joined the FameLab which was being advertised at Wits to improve her scientific communication skills.
“FameLab has taught me that good scientific communication comes from avoiding jargon. In your university studies you’re being taught new words and what they mean to get you up to speed in your field. So there’s lot of jargon that you’ve been taught which is necessary for you to understand the subject and to communicate with your lecturers. But because there is so much jargon in science, when you try to communicate with someone who is not a scientist, and you don’t know how to remove the jargon and simplify it, then you lose the meaning and impact of your work.”
“Then people don’t understand your work and how it relates to them or the broader picture. What happens is that people think science is not for them. And that’s not true. Science is not just for the scientists it’s for everyone. In my case, we’re not just doing a purely academic project since it’s a real-world problem that impacts a lot of people”.
“Now when you remove the jargon people can understand your work and say ‘Oh, that makes sense’. FameLab teaches you how to take your 200-page Master’s dissertation, or a high-level journal article, and condense it into three minutes. It sounds impossible, but eventually you get the hang of it.”
It’s important to get this communication across to the broader public, says Hasrod, because “we do a lot of good research in South Africa and Africa in general. And we really need to showcase to the world what we do.”
“It can take anywhere from three months to three years just to publish a single paper of maybe 20 pages and not everyone has access to read it. So competitions like FameLab where you condense all of that down to the really important bits, allows a lot more people to hear it and see what its impact is.”
Asked what advice she would give to young researchers, Hasrod says: “Discipline, resilience, patience”.
She elaborates that “Research is such that things don’t work most of the time. Classically, in science when something works we give it to engineering to commercialise. I’m in science, and if something’s not working you have to give yourself faith that eventually it will. You’ve got to wake up every day, even if you’re failing, go back in, give it your best, try again and eventually you will see a result. It might not be a good result because sometimes you learn more from when things don’t work. It takes a very long time and research is not fast. That’s one of the reasons why I came to machine learning, to try and make research faster.”
Regarding her winning the South African leg of the competition she says, “I was very excited to win and I was really, really happy. It was something I worked hard towards and it was an even sweeter victory when I shared the news with my supervisor and my research group and they were so excited. That’s what made me realise the impact of the work. It showed me that our research is capable of getting to that international level.”
“Now I’m excited about Friday, and I’ll give it my best”.
Mr Robert Inglis, the director of Jive Media Africa, says: “For science to make an impact in the world, it needs to get out of the lab and into people’s lives. The FameLab process has helped Taskeen to hone her skills for researching public audiences.” Jive Media Africa runs the FameLab programme in South Africa, in partnership with SAASTA (the South African Agency for Science and Technology Advancement, a business unit of the NRF).