Development of Self-Sensing Adhesive Materials for Predictive Maintenance in AI Hardware
Structural Adhesives
University of Leicester | Chemical, Earth and Environmental Sciences
Giving adhesives a way to signal damage before failure
The processors used in artificial intelligence and data centres generate considerable heat and operate under repeated thermal and mechanical stresses. Adhesives help hold components together and transfer heat, but their properties can gradually deteriorate. Because conventional adhesives do not report their own condition, damage may go unnoticed until a component overheats or equipment fails.
Morgan’s project will develop self-sensing adhesives containing special molecules known as mechanophores. These molecules respond to mechanical force by changing their chemical structure, producing a quantifiable signal such as a change in colour or light. The research will identify mechanophores that can withstand the high temperatures and stresses found in advanced electronics, before incorporating them into realistic adhesive formulations, including those containing thermally conductive fillers. Through mechanical testing, the project will investigate how these signals relate to early signs of damage and changes in adhesive performance under operating conditions.
The long-term goal and impact is significant: Developing the next generation of adhesives for advanced electronics applications that could help prevent hardware failures, extend product lifetimes and reduce maintenance costs across sectors.