The recent eruption of COVID-19 has demonstrated, first, that changing climate and relentless globalization increase risks from novel infectious agents, and second, we need test methods that will rapidly and accurately detect these microorganisms.
For several years, a Korean research team has been developing such a sensitive flexible test platform. Now, they have improved the speed, sensitivity, and robustness, moving toward clinical application.1
The sensor is built around silicon microring resonators (SMRs) with surfaces modified to bind specific target organism DNA. As molecules bind to the SMR, the effective refractive index shifts, modifying the resonant wavelength. By measuring changes in the optical power of a wavelength-scanned laser source, the resonant shift can be quantified. The latest innovation uses ball lenses to optimize the input and output coupling, thereby improving the device’s performance.
Single-step measurement
The ideal clinical test provides accurate results quickly, in an easy-to-use process. Requirements for rapidity, accuracy, and simplicity often push a design in opposite directions, so development requires not only rigor, but innovation. This research team, composed of members from Yonsei University and three other Korean organizations, built on the foundation of sensitive SMR detection.
SMRs are silicon waveguides formed in small, circular loops. The geometry and index of refraction determine the resonant frequency. When a ring resonator is adjacent to a linear waveguide, the evanescent field that extends outside of the linear waveguide will transfer energy at the ring’s resonant wavelength. With a broadband source, that energy transfer shows up as a dip in the power spectral density transmitted through the linear waveguide. In this implementation, the Korean researchers used a device with four SMRs: one they left buried in the silicon oxide to monitor temperature effects, while they etched the other three free of the oxide and implemented an innovative biochemical detection strategy.
They bound forward DNA primer to the surface of the three sensing SMRs and immersed the sensors in a solution containing additional primers, enzymes, and the test sample. The sensor is maintained at 38°C, and if the sample DNA matches the primers, that DNA will replicate in place on the primers bound to the SMRs. A large fraction of the energy in the resonant modes of the SMRs is in evanescent waves, so there is a large interaction between light in the waveguide and molecules bound to the surface.
The researchers verified a shift in resonance in clinical samples from patients with Coxiella burnetii infections—a difficult-to-diagnose infection that can lead to Q-fever, a serious human illness. The resonant wavelength shift was clearly distinct from the signature of uninfected patients, in a procedure that took 30 minutes.2
Sensitivity and speed were achieved, but the sensor required careful alignment of the input beam, which limited its clinical utility. To reduce misalignment sensitivity, the researchers added ball lenses to optical fibers that couple light to and from the sensor. This had two additional consequences. First, it increased the depth of focus of the input light in the linear waveguide. That is, the beam remained narrower for a longer distance, maximizing the coupling to the SMRs. Second, the coupling into and out of the sensor was improved as well, leading to a higher signal-to-noise—equivalently, a higher detection sensitivity.
The design now met requirements for sensitivity, speed, and robustness; the next step was clinical validation.
Real-world testing
The updated configuration of the sensor sends light from a swept wavelength tunable laser centered at 1550 nm into a polarization-maintaining fiber, then into a custom-fabricated, 300-µm diameter ball lens. The ball lens directs the beam into the SMR sensor waveguide, where it propagates in proximity with the three functionalized resonant rings and the single non-functionalized ring. The transmitted light is gathered by a matching ball lens/fiber assembly, then detected by a photodiode. The two ball lens assemblies are automatically aligned to produce maximum throughput (see figure).
In operation, the sensing region is immersed in the enzyme/primer/analyte solution as the wavelength sweeps through a desired range, monitoring any shift in SMR resonance. In the presence of the target analyte, more molecules are bound to the surface, creating greater resonant shifts. The researchers validated the sensor performance by measuring clinical samples from 18 different patients suffering from fever. After 20 minutes of incubation, nine with Coxiella burnetii infections were clearly distinguished from nine whose fever stemmed from other causes.
The sensor can be modified to detect other infectious agents simply by changing the primer design. Professor Yong Shin of Yonsei University, a member of the research team, says they are now testing larger cohorts of clinical samples and on the road to “commercializing the optical sensor in clinical applications.”
REFERENCES
1. B. Park et al., Biosensors, 11, 125 (2021); doi:10.3390/bios11040125.
2. B. Koo et al., J. Biophotonics, 11, 1–9 (2018); doi:10.1002/jbio.201700167.
