ºÚÁϸ£ÀûÍø

In recent years, scientists have developed new generations of biosensors capable of detecting individual molecules separately. Rather than a smooth measurement curve, these sensors produce a signal made up of discrete events, precisely at the moment molecules bound or released. This offers an unprecedented level of detail, but also fundamentally changes the nature of the measurement.

Chris Vu's research focused on what this shift toward single-molecule detection means for the functioning and limits of continuous biosensors. As a model system, he used Biosensing by Particle Motion: a setup in which microscopically small particles are coupled to a surface through molecular bonds. By tracking the motion of the particles over time, individual binding and release events could be directly observed.

What the research demonstrated

By tracking individual particles over many hours, Vu showed that no two sensor elements behaved in exactly the same way. Small differences in molecular interactions accumulated over time and led to clearly measurable variations between sensor elements. Using mathematical models and computer simulations, these observations were traced back to fundamental processes at the molecular scale, such as the number of functional binding molecules on the surface of sensor elements and how these molecules gradually change over time.

Beyond describing these effects, the research posed a more fundamental question: where did the limits of continuous measurements based on individual molecules lie? How quickly could changes be detected, and how accurate could a measurement in principle be when every signal was built up from separate molecular events? By systematically investigating different scaling laws, the research showed that random processes at the molecular scale imposed unavoidable limitations, but also that thoughtful design choices could bring sensors closer to those limits.

A surprising finding

One of the most striking outcomes of the research challenges the current consensus in the field. It is widely assumed that strong molecular interactions are needed to detect low concentrations of molecules. However, the models developed in this work showed that weak binders can achieve detection limits comparable to those of strong binders. The added benefit is that sensors based on weak interactions can produce continuous measurements, which is especially important for tracking dynamic processes in real time.

What this means for society

The insights from this thesis are relevant to anyone developing or applying biosensors, from medical diagnostics to fundamental scientific research. The findings offer:

• A better understanding of variations between sensor elements and how to account for them
• Guidance for designing more accurate and faster biosensors
• A principled framework for what is and is not measurable at the single-molecule level

In short, what at first appeared to be noise or instability turned out to contain valuable information. By embracing the complexity of single-molecule behavior and making use of it, this work contributed to a more principled and powerful understanding of biosensing at the smallest scale.