Access Official Technical PDF

Share Summary:LinkedInEmail

Monitoring a biomolecular reaction through only one type of signal can leave important changes undetected. A reaction may alter a sample’s dielectric properties while also changing its fluorescence behavior, and each measurement method has its own sensitivities and limitations. A hybrid CMOS biosensor developed in 28 nm technology brings these complementary measurements together on a compact platform, combining microwave permittivity sensing with single-photon optical detection.

The design illustrates a lab-on-CMOS approach: integrating different sensing modalities and their supporting electronics on one chip. Its practical significance lies not in replacing laboratory analysis across the board, but in showing how parallel, quantitative measurements can provide a richer view of reactions over time.

Why combine microwave and optical sensing?

Biomolecular sensing typically connects a bioreceptor, which interacts with a target, to a biotransducer that turns the interaction into a measurable signal. The signal might involve light, electrical properties, or another physical response. Because no single transduction method captures every aspect of a complex reaction, combining modalities can help distinguish changes that would be difficult to interpret from one measurement alone.

Conventional laboratory instruments can be large, while CMOS integration offers a route to placing multiple transducers on a small sensing platform. The reported system combines a microwave sensor for changes in complex permittivity with an optical array for photon counting and fluorescence-lifetime measurement. These modalities operate in parallel rather than relying on a single measurement channel.

How the microwave permittivity sensor works

The microwave sensing unit uses a superharmonically coupled quadrature voltage-controlled oscillator, or QVCO. When the material under test changes, its complex permittivity affects the oscillator’s electrical loading and shifts its natural frequency. The sensing architecture converts that shift into a phase-related output, avoiding the need for a power-intensive millimeter-wave frequency divider.

The sensor operates across a reported range of 37.5 to 45.1 GHz. This high-frequency region was selected in part to reduce artifacts associated with biological water’s Debye relaxation during frequency-dependent measurements. The approach is intended to probe intrinsic biomolecular properties in an ionic medium, rather than relying on a detection label to produce the signal.

Stability without a replica reference

Slow biochemical reactions can produce small changes over periods of minutes or hours, making drift and electronic noise important design concerns. Instead of depending on a separate replica sensor or a specific label for compensation, the readout uses a tank-perturbation method. The two oscillator paths are deliberately given a small differential bias, and their output response is measured as that bias is swept.

A slope is extracted from the resulting measurements. The sweep and regression process helps suppress noise, while slope extraction also provides a correlated-sampling effect that reduces low-frequency readout noise. The reported long-term stability is 5.4 ppm over eight hours. This is a stability result for the sensor readout, not a universal measure of detection accuracy for every sample or reaction.

SPADs add photon counting and lifetime information

The optical sensing unit is an array of single-photon avalanche diodes, or SPADs. A SPAD detects individual photons by operating a semiconductor junction above its avalanche breakdown voltage. When a photon triggers an avalanche, the resulting electrical pulse can be counted; with suitable timing, photon arrival information can also be used to characterize fluorescence lifetime.

Fluorescence lifetime measurement can add information beyond the amount of emitted light. In the reported platform, the SPAD array supports photon counting and lifetime measurements using pulsed optical excitation. These optical measurements complement the microwave channel, which responds to changes in the sample’s dielectric properties. Together, they provide distinct observations of the same reaction rather than two versions of one signal.

Demonstrating the dual-modality approach

The system was used to study enzyme denaturation under thermal and chemical stress. The dual-modality measurements tracked luciferase denaturation associated with both forms of stress, including chemical stress from ethylene glycol. The purpose of the experiments was to demonstrate the value of combining the microwave and optical sensing channels in a biomolecular measurement.

For engineers and system integrators, the example highlights several design considerations for an integrated sensing platform:

  • Complementary transducers: Permittivity and fluorescence measurements capture different physical responses from the sample.
  • Long-duration readout: Drift management matters when reactions evolve over hours rather than producing an immediate signal.
  • Integration trade-offs: Microwave oscillator circuitry and optical photon-detection circuitry have different implementation and measurement requirements, even when placed on the same CMOS platform.
  • Application-specific validation: The reported enzyme experiments demonstrate a sensing approach; they do not establish performance for every analyte, sample preparation, or operating environment.

Practical implications for sensing-system design

This CMOS biosensor is a research demonstration, not evidence that a particular industrial process can be monitored without additional development. Its broader engineering lesson is that integrating multiple transducers can make it possible to observe a reaction through more than one physical signal while keeping the sensing electronics compact. For applications such as biochemical process development or reaction characterization, that can support more informative measurements—provided the sensor is validated for the target chemistry, sample conditions, and required operating lifetime.

The combination of microwave permittivity sensing and SPAD-based optical detection also shows why readout strategy is as important as sensor selection. Small, gradual changes require attention to noise, drift, and measurement methodology, not just sensitivity at a single instant.

Conclusion

By pairing a QVCO-based microwave sensor with a SPAD optical array, the 28 nm lab-on-CMOS platform measures complementary properties of biomolecular reactions. Its reported 5.4 ppm eight-hour stability and enzyme-denaturation experiments demonstrate the potential of multi-parameter sensing without replica references or detection labels in the microwave channel. The practical takeaway for technical decision-makers is to treat integrated, multimodal sensing as a way to gain richer reaction data—and to assess its suitability through application-specific validation.

Access Official Technical PDF

About The Author