Graphene algorithm replaces quality control guesswork

Engineers at Monash University have developed a world-first machine-learning algorithm which can assess the quality of graphene in under 14 minutes, potentially bringing quality control costs of the $1000 per gram material down significantly.
Graphene, derived from graphite, possesses a range of properties that make it highly valued for applications in energy storage, water purification and sensing. It was first discovered in 2004 and is created by exfoliating graphite down into a single layer structure.
Originally published in global journal, Advanced Science, the engineers took hundreds of images from a quantitative polarised optical microscope, then used the machine learning algorithm to identify data clusters, after which they applied image analysis to quantify the proportions of each cluster.
Previously, the quality of graphene could only be tested in products after they were manufactured, further compounding a lack of consistency between batches, laboratories, and manufacturers. Now, the algorithm allows manufacturers to detect the quality and properties of graphene before using it.
Professor Mainak Majumder from the Depart of Mechanical and Aerospace Engineering at Monash University led the breakthrough study alongside the Australian Research Council’s Hub on Graphene Enabled Industry Transformation.
Majumder believes there was a definite gap in the market for such a technology as manufacturers would like more assurance in the quality of the graphene they are purchasing, especially considering the price is so steep.
“Graphene possesses an extraordinary capacity for electric and thermal conductivity. It is widely used in the production of membranes for water purification, energy storage and in smart technology, such as weight loading sensors on traffic bridges” Majumder said.
“Therefore, manufacturers need to be assured that they’re sourcing the highest quality graphene on the market. Our technology can detect the properties in graphene in under 14 minutes for a single dataset of 1936 x 1216 resolution. This will save manufacturers vital time and money, and establish a competitive advantage in a growing market”.
Dr Shibani believes their work would especially benefit industries wanting to deliver high quality, high functioning graphene to their customers.
“There are a number of ASX listed companies attempting to enter this billion-dollar market, and this technology could accelerate this interest,” said Shibani.
Director of the Australian Research Council’s Hub on Graphene Enabled Industry Transformation, Professor Dusan Losic said the outcomes from the research efforts were outstanding.
“These outstanding outcomes from our ARC Research Hub will make a significant impact on the emerging multibillion-dollar graphene industry, giving graphene manufacturers and end-users new and a simple quality control tool to define the quality of their produced graphene materials which is currently missing.”
