Applying integrated optics to single bioparticle counting
Optical detection plays a major role in chemical and biological analysis. By measuring optical probes attached to biomolecules or the intrinsic optical properties of a sample, scientists can determine concentrations of specific species during disease analysis, genetic profiling, and basic scientific discovery. Classically, this has required either use of bulky table-top instruments for high-sensitivity measurements or production of many photons so that low-sensitivity cameras or visual readouts are adequate.
Minimizing the size of optical readout systems would enable analysis using handheld devices. Providing light generation, routing, and detection on a single photonic chip promises low costs, high speeds, and compact packaging. One of the most natural applications for a photonics chip is bio-fluid characterization, but visible light guiding in fluids has presented a challenge. Most materials used to contain a fluid, such as glass or plastics, have higher refractive indices than the fluid, which makes classical waveguiding through total internal reflection impossible. To address the challenge of fluidic waveguiding, we applied the principles of optofluidics, a new field that seeks to marry micro-scale fluid and light manipulation,1, 2 to develop an optofluidic chip that can detect bioparticles in a fluid with unprecedented sensitivity.
Our optofluidic biosensors combine integrated optics with microfluidics on silicon substrates. The integrated optics component includes both solid ridge waveguides etched into a glass surface and hollow waveguides capable of guiding light when filled with a fluid. Fluid-filled waveguides rely on the anti-resonant reflective optical waveguide (ARROW) principle, which uses precise film thicknesses to reflect optical modes through interference.3 The chips' microfluidic component relies on micron-scale channels that push test fluids from one micro-reservoir to another, making possible particle analysis in a fluid stream.
The most critical element of our chip is the orthogonal intersection of a hollow-core waveguide and a solid-core waveguide. Bioanalysis is performed when fluorescently tagged molecules within the liquid sample are excited as they flow past the orthogonal solid core. This intersection isolates a femtoliter-sized sample volume. The resulting fluorescence signal is confined in the liquid core and guided into another solid-core waveguide that directs it off the chip. At the edge of the chip, the signal is collected by a photodiode and can be quantified and analyzed.
Construction of our chips relies on standard silicon-based microfabrication techniques. Figure 1 illustrates the basic process flow. We begin by depositing ARROW cladding layers on silicon. These cladding layers consist of alternating films of dielectric materials, typically silicon dioxide (SiO2) and tantalum pentoxide. To create hollow core channels, we first photolithographically define a line of photoresist, which is then coated with SiO2 using chemical vapor deposition to form the walls of the channel. Subsequently, the photoresist is removed with acid, leaving a hollow core. We can then define solid core waveguides in the SiO2 coating by additional photolithography and plasma etching steps. We have developed several generations of devices, which have continued to improve on optical throughput by elevating hollow cores on a self-aligned pedestal,4 overgrowing top cladding layers,5 and using materials with low photoluminescence.6
We have used our optimized optofluidic chips to analyze a number of single bioparticles.7 For example, we performed a target and control test with H1N1 and H3N2 influenza virus samples at clinically relevant concentrations (104 particles/milliliter). Ribonucleic acid (RNA) was extracted from both samples and mixed with a fluorescent molecule tag designed to specifically bind to H3N2 RNA. In this experimental setup, a fluorescent signal is produced only if a tag successfully binds to the target RNA. Our measurements indicated a positive response to the H3N2 sample (565 spikes in optical intensity in 200s, corresponding to the detection of 565 individual H3N2 RNA molecules). By comparison, the H1N1 sample produced only a handful of optical peaks, implying that the H3N2 probe and H1N1 targets did not bind and no strong fluorescence signals were created. This result suggests that our optofluidic platform can be used for amplification-free detection of clinically relevant virus samples and other bioparticles.
Our optofluidic platform represents a nearly ideal application of integrated optics, microfluidics, and biophotonics. With our optimized chip, we can analyze single bioparticles in clinically relevant solutions and provide an alternative to gold standard techniques like polymerase chain reaction in a compact, point-of-care package. We are continuing to improve the sensitivity of our biosensors and to test their performance with a variety of clinical samples.
This work is supported by the National Institute of Allergy and Infectious Diseases, part of the National Institutes of Health, and the National Science Foundation under grant numbers 1R21AI100229, ECCS-1101801, ECCS-1101902, CBET-1159423, and CBET-1159453. We would also like to thank the W. M. Keck Center for Nanoscale Optofluidics at the University of California, Santa Cruz; the Integrated Microfabrication Laboratory at Brigham Young University; Philip Measor at Liquilume Diagnostics; and Charles Chiu at the University of California, San Francisco.
Aaron Hawkins is a professor of electrical and computer engineering. He received his BS from the California Institute of Technology and his MS and PhD from the University of California, Santa Barbara.
Lynnell Zempoaltecatl is an MS student in electrical engineering. She received her BS from Brigham Young University.
Holger Schmidt is a professor of electrical engineering. He received his diploma in physics from the University of Stuttgart and his MS and PhD from the University of California, Santa Barbara.