The Hard Problem No One Is Talking About

How to Design Biosensors in the Absence of a Clean Pre-Binding State to Reference

Until now, the central theme of my posts and articles has focused on understanding and optimising electrochemical biosensing platforms. Primarily, this has been achieved through the lens of interpreting electrochemical data to gain deeper insights into the physical platforms we are developing. This could involve identifying bottlenecks in the signal transduction pathway, defects and heterogeneity in films or monolayers, rate-limiting steps, and so on. While this has been intentional (this is Electrochemical Insights after all), I have been avoiding an issue which has come up again and again in my own work.

For those who know me personally, it's no secret that I'm not your typical academic. While I do hold a PhD, my interests have always leaned more toward industry, where projects are grounded in solving real-world problems with immediate impact. That’s not to say I think one is more valuable than the other. I believe academic research is just as important, and I collaborate with academic partners on nearly every project I work on. The breakthroughs that come from fundamental research are often what make applied innovation possible. I simply find myself more energised by the fast-paced, applied nature of industrial work.

With that said, there is a critical issue in the field of electrochemical biosensing that receives surprisingly little attention during the R&D phase, both in academia and industry. It is the reality that, in real-world settings, the target analyte we aim to detect is never alone. It is always surrounded by background noise, and often, already present in unknown concentrations. This creates an even deeper challenge — the absence of a clean pre-binding state to reference. In most field applications, there is no opportunity to first record a baseline signal in a target-free condition before the analyte arrives. The sensor is exposed to the sample as it is, meaning the very first signal you measure may already include contributions from binding events you can't track or separate.