From the Swiss Federal Institute of Technology in Zurich, an innovative method for the rapid and cost-effective selection of metal catalysts for large-scale conversion of hydrogen and carbon dioxide into green methanol.
In an article by the World Economic Forum on energy transition, the writers sort of throw out a challenge, wondering where green methanol has disappeared to in the race to find sustainable alternatives to fossil fuels.
Let’s not forget, as the International Council on Clean Transportation (ICCT) puts it, methanol – which is pretty much «a close cousin of ethanol in liquid form» – comes in four flavors. There’s “grey methanol“, made from natural gas; “blue methanol“, also from natural gas but with a bit of carbon capture and storage action; “biomethanol“, from biomass (which also gets to wear the “green” or “renewable” methanol hat), and last but not least, e-methanol, made by using renewable energy to grab hydrogen and carbon dioxide (CO2) out of the air, which is the type of green methanol we’re talking about here [source: “A Step Forward For “Green” Methanol And Its Potential To Deliver Deep GHG Reductions In Maritime Shipping” – International Council on Clean Transportation].
The Methanol Institute reckons green methanol can slash carbon dioxide emissions by a whopping 60-95%. But circling back to the WEF’s big question, «its production is still on the low side, with less than 0.2 million tonnes cranked out globally each year, against the 98 million tonnes of the regular methanol from fossil fuels».
Even though the green methanol market is expected to grow, with various countries (China leading the pack) seeing its potential, its rollout, analysts point out, «is being slowed down by costs, which are still way higher than those for methanol made from natural gas and coal».
TAKEAWAYS
Green methanol: unpacking the high costs of production
The International Renewable Energy Agency (IRENA) breaks it down for us: producing e-methanol can set you back between 800 and 1,600 bucks per tonne. These hefty price tags come from how complex and time-consuming its production is, needing something called “electrolysers.” These gadgets are all about splitting hydrogen from oxygen. Then, this hydrogen gets turned into green methanol in a reactor, through a cool process with carbon dioxide [source: “Innovation Outlook: Renewable Methanol” – International Renewable Energy Agency].
In the science world, a catalytic process is a big deal because it’s all about using catalysts to make reactions go faster. Especially metal catalysts, they’re the secret sauce to getting more product out of the reaction.
When we’re talking about making chemicals and fuels, finding the right metal catalysts is like looking for a needle in a haystack. It’s crucial but involves sifting through loads of options, with thousands or even millions of combinations. And when it comes to green methanol production, it’s even trickier. Scientists say «the metal combinations are endless. And the chemical space we’re digging through for these metal catalysts is about as big as 1020 possibilities» note some researchers from the Department of Chemistry and Applied Biosciences at the Swiss Federal Institute of Technology in Zurich. They’re part of this tech group SwissCat+ (housed at the École Polytechnique Fédérale de Lausanne) and put their heads together for the study “Accelerated exploration of heterogeneous CO2 hydrogenation catalysts by Bayesian-optimized high-throughput and automated experimentation,” which hit the Chem Catalysis mag in February 2024.
Their mission? To make the hunt for the right metal catalysts for green methanol synthesis a bit less like finding a needle in a cosmic haystack. They’re aiming to cut down the overwhelming number of possible good, scalable, and cheap options to pick from.
Getting smart with catalytic composition calculations
Digging into the nitty-gritty, the Swiss Federal Institute of Technology in Zurich crew has come up with a wayto digitally crunch numbers for the metal catalyst mixes needed for green methanol. Let’s walk through it slowly.
For starters, to be considered big-league and wallet-friendly, the catalyst mixes need to be made from metals that don’t cost an arm and a leg. That’s why the research gang went for iron, copper, and cobalt, plus a few supporting materials like four metal oxides. Mixing all these up, they say, «we’re still looking at twenty million possible mixes», the authors share.
That’s when they whipped up a machine learning algorithm that’s pretty savvy, using something called “Bayesian Optimization.” This clever trick helped them sift through the three chosen metals to find the top mix for making metal catalysts.
Bayesian Optimization is all about smart guessing, helping to find «the likeliest best settings for any given system or model». It’s a fresh take compared to old-school optimization methods, zooming in on those bits of data that are more likely to give us the good stuff [source: “Bayesian Optimization” – ScienceDirect].
Hunting down the best catalyst for the production of green methanol
When they first put the machine learning algorithm through its paces, it picked out twenty-four metallic catalyst mixes that seemed to tick all the boxes based on what it had been taught.
But the team at the Swiss Federal Institute of Technology in Zurich lets us in on a little secret: «this initial pick was more about giving the AI’s decision-making process a whirl than actually nailing the perfect catalyst for green methanol right out of the gate». So, at first, “finding the best catalyst for green methanol production wasn’t the top priority.” There was another angle:
«…right now, most of what we know about making fuel comes from the oil industry’s playbook. When we shift focus to chemical reactions for the green energy scene, we’re kinda flying blind with a serious lack of solid data for training AI algorithms. Yet, these algorithms need that kind of info to really get cracking in the huge realm of chemical possibilities»
So, this initial round was a test to see how well the machine learning algorithm, armed with Bayesian Optimization, could handle the training data thrown at it by the Swiss lab.
Looking at the twenty-four metallic catalytic mixes the AI picked and tested in just a week, they were all about copper and zirconium. The authors point out:
«Any scientist would have probably landed on this mix after digging through research, but here’s an AI, with no special knowledge but a good handle on Bayesian Optimization and the initial chemical space and objectives it was given, coming up with it on its own»
Moving to the second round of tests, starting with those first picks, the Bayesian Optimization method dialed down the copper (too much of it can mess with the catalyst’s mojo) and threw in a bit of zinc and cerium. These additions are «top picks among copper-based ternary catalysts», aimed at boosting performance without bumping up costs.
By the time they hit the third and final round, the Bayesian optimizer had the formula for green methanol synthesis down pat, suggesting a switch to indium-based catalysts from copper, all backed up by zirconium oxide. These elements were all in the mix from the start but got a closer look and were tweaked across three stages, all wrapped up in five weeks.
Glimpses of Futures
Diving into the world of techniques and methods to quickly and affordably fine-tune metallic catalysts for the big-time conversion of hydrogen and CO2 into green methanol is basically setting the stage for a big shift towards a renewable fuel that’s seen as a game-changer, especially for cutting down emissions in land, sea, and air transport. This is a big deal in the Laurelin project’s playbook – yep, that’s the one backed by EU funds – focused on cooking up innovative green methanol production solutions, kicked off in May 2021 and set to wrap up by April 2025.
The work by the Swiss Federal Institute of Technology in Zurich, which has cracked what usually took years in just a few weeks, is blazing a trail for a fresh take on that age-old problem of speeding up the research and development of high-performance, low-cost, and practical metallic catalysts.
Those indium and zirconium oxide catalysts that got the thumbs up from AI using Bayesian Optimization? They’re now being put through their paces in the SwissCat+ labs. But beyond their main gig of helping turn hydrogen and CO2 into green methanol, the team points out that these catalysts could down the line play a role in other chemical reactions in wider chemical playgrounds, where, say, the catalyst’s evolution might be steered by different rules and goals.
Leaning on the STEPS matrix, let’s try to scope out future scenes by looking into how the evolution of crunching numbers for metallic catalyst mixes in green methanol production might shake things up socially, technologically, economically, politically, and in terms of sustainability.
S – SOCIAL: on the social front, the positive vibes from evolving the method the Zurich crew talked about could mainly make shifting to green methanol as a sustainable fuel for road transport smoother. Today, this sector’s responsible for chucking more than 22% of greenhouse gases into the air over Europe, as per the Annual European Union Greenhouse Gas Inventory 1990-2021 and Inventory Report 2023 by the European Environment Agency. This move is in sync with the EU’s climate targets and, more specifically, with the tentative deal struck on 9 February 2024 by the EU Council and Parliament for slicing emissions from heavy-duty vehicles – think trucks, buses, and trailers – by 2030, 2035, and 2040.
T – TECHNOLOGICAL: down the road, the tech behind digitalising the recipe for metallic catalyst mixes in green methanol production – right now a combo of machine learning and Bayesian Optimization – will need to nail down and systematically sort out the algorithms’ training data, which has got to be spot-on, reliable, and leaning more on the know-how of the sustainable energy scene instead of borrowing from the oil industry’s notes.
E – ECONOMIC: economically speaking, if we fast forward to a future where the rush towards green methanol picks up speed – thanks to quickly optimised metallic catalysts ready for large-scale action in flipping hydrogen and CO2 into green methanol – it’ll mean pouring investments across the whole value chain. We’re talking about pumping money into developing the tech, setting up industrial and storage infrastructures, and getting things off the ground. These moves will likely link up with business strategies aimed at job creation in the e-methanol niche and bringing in professionals with the right skill set.
P – POLITICAL: looking ahead, the idea of green methanol taking over from fossil fuels, especially in the transport world, could trigger a political rush to sort out a generally agreed-upon sticky point in energy production: crafting policy tools to lock in «fair tax treatment and a guaranteed minimum price in the long term for the production of e-methanol» [source: “Renewable Methanol: An Enabler for Carbon Neutrality in the Chemical & Liquid Fuel Sectors” – International Renewable Energy Agency].
S – SUSTAINABILITY: picturing a future where the global shift to green methanol – with a little help from AI techniques – is done and dusted means, as the International Renewable Energy Agency (IRENA) puts it, «having broadened the use of e-methanol as a fuel base, hitting the Net Zero emissions target across all industrial sectors, not just in transport». That’s the ultimate sustainability goal that the EU’s green policy and, on a wider scale, the United Nations’ 2030 Agenda (Goal 13 – “Take urgent action to combat climate change and its impacts”) are shooting for.
