Scientists use Artificial Intelligence (AI) to guide the process of designing new materials based on biomass waste to capture carbon dioxide gas.
Climate change is one of the most important issues in the world today because of its dire consequences in the future. The World Bank reports, due to climate change, more than 200 million people in the world are forced to become climate migrants. In the next three decades, they will have to leave their homes due to water scarcity, hunger, or rising sea levels.
Climate change, according to the Ministry of Environment and Forestry’s definition, is a significant change in climate, air temperature and rainfall ranging from decades to millions of years. Climate change occurs due to increasing concentrations of carbon dioxide gas and other gases in the atmosphere that cause the greenhouse gas effect. The increase in greenhouse gas concentrations is caused by various human activities such as fossil fuel emissions, changes in land use, waste and industrial activities.
Various efforts have been made to reduce the amount of carbon dioxide or CO2 emissions. One of them is carbon dioxide sequestration or carbon dioxide sequestration, which is a method used to capture and store carbon dioxide from the atmosphere. This method is also known as Carbon Capture and Storage (CCS).
Activated Carbon from Biomass Waste
If we want to reduce climate change, we must find cost-effective and sustainable ways to reduce carbon emission dioxide (CO2) industry. Unfortunately, the CCS method that is widely used today to “capture” and store carbon in industrial post-combustion sources has significant drawbacks, including high cost, potential for toxicity to the environment, or durability issues.
Waste biomass can be used to produce porous carbon, or what we know as activated carbon, which can absorb CO2 gas emitted from large emission sources, for example, power plants, cement industry, and so on.
One of the advantages of using porous carbon for CO2 absorption is that it can be produced from biomass waste, such as agricultural waste, food waste, animal waste, and forest debris. This makes the porous carbon from biomass waste (biomass waste-derived porous carbons (BWDPC) attractive not only in terms of its low cost, but also because the porous carbon from biomass waste can be an alternative to utilizing biomass waste.
However, to date, there are no general guidelines on how such high quality porous carbon should be synthesized or what the optimal operating conditions for porous carbon should be.
In a recent study, scientists used machine learning-based methods to determine the core factors that should be prioritized in making porous carbon from this biomass waste in order to achieve the best CO2 adsorption or adsorption performance which will eventually pave the way to a circular economy.
Against this backdrop, many researchers have focused on what may be our best step for the next generation CCS system: CO2 adsorption using solid porous carbon materials.
Porous carbon from waste biomass brings us closer to realizing a circular economy, but this field of study is relatively new, and there are no clear guidelines or consensus among scientists on how porous carbon from waste biomass should be synthesized or what properties and composition of materials should be used. they have to use.
Application of Machine Learning for Carbon Synthesis Strategy
Utilization of AI to increase CO2 adsorption on activated carbon or porous carbon from waste biomass.
Is the role of Artificial Intelligence (AI) needed and can help scientists answer this challenge? Recent studies published in Environmental Science and Technology said the research team from Korea University and the National University of Singapore used a machine learning-based approach that might guide the development of porous carbon synthesis strategies in the future.
Scientists have identified three main factors that influence the adsorption properties of porous carbon CO2 from waste biomass: the elemental composition of the porous solid, its textural properties, and the adsorption parameters under which it operates, such as temperature and pressure. However, how should these factors be prioritized when developing porous carbon from waste biomass? Until now, this is still a question.
To help solve this problem, the team from the two universities first conducted a literature review and selected 76 publications describing the synthesis and performance of various porous carbons from waste biomass. After the curation process, the papers provided more than 500 datapoints which were used to train and test decision tree/tree-based models.
“The main aim of our work is to explain how machine learning can be harnessed for predictive analytics and used to bring valuable insights into CO2 adsorption processes using porous carbon from waste biomass,” explains Professor Yong Sik Ok from Korea University who led the research.
The input features of these models are the three main factors, while the output is the level of CO2 adsorption. Although the models themselves essentially become a ‘black box’ after the training process, they can be used to make accurate predictions about the performance of porous carbon based solely on the main factors under consideration.
Most importantly, through feature analysis, the research team determined the relative importance of each input feature to make accurate predictions. In other words, this prediction will establish the most important core factor to achieve high CO2 adsorption rate.
The results show that the adsorption parameters contribute more than the other two main factors so that the machine learning model can make the right predictions. What is underlined here is the importance of optimizing operational conditions first. Meanwhile, the texture of the porous carbon, such as pore size and surface area, ranks second. And the elemental composition factor is in the last order.
Potential to Find New Material
It should be noted that the model predictions and the results of the feature importance analysis are supported by the existing literature and our current understanding of the mechanisms behind the CO2 capture process. The results of this study strengthen the application of data-driven strategies in the real world, not only for porous carbon from waste biomass but also for other types of materials.
“Our modeling approach is cross-deployable and can be used to investigate other types of porous carbon for CO2 adsorption, such as zeolites and metal-organic frameworks, and not just those from waste biomass,” said Professor Ok.
The team from the two universities is now planning to design a porous carbon synthesis strategy from biomass waste with a focus on optimizing the two most important key factors. In addition, they will also continue to add experimental datapoints to the database used in this research and make it open source so that the research community can also benefit.
With this research, we hope that all the efforts made will lead us to a sustainable society that can stop climate change and achieve the UN Sustainable Development Goals, one of which is Goal 13, namely Climate Action.