Fluorescence microscopy is one of the most commonly used analytical tools in biology, chemistry, and medicine. Several sophisticated fluorescence techniques can also achieve single molecule sensitivity, but they require bulky setups with several optical elements.1 Fully integrated designs accordingly generate a high level of interest due to their inherent advantages, such as compactness, robustness, and inexpensive fabrication.2 Recently, integrated waveguides for liquids and gases have become available.3 However, the excitation of molecules and the detection of their fluorescence with a planar optofluidic chip requires light to be guided through liquids. Furthermore, wavelength filtering of the detected light is required to reject the much stronger excitation light.
One device that shows great promise for this application is the antiresonant reflecting optical waveguides (ARROW).4 It is based on surrounding the waveguide core with several dielectric layers that act as Fabry-Pérot reflectors. Migrating this concept from its original solid-state implementation to fluid cores has allowed us to overcome the technical problems associated with waveguiding in liquid environments.
The geometry of an ARROW with a non-solid core is shown in Figure 1(a) for a low-refractive index core (nc) surrounded by high-index cladding layers (n1,2) on both sides. Such structures can be fabricated using plasma-enhanced deposition of the confinement layers (typically silicon dioxide and nitride) with intermediate deposition of a sacrificial core layer. After all dielectrics are deposited, the sacrificial layer is removed in a selective etch, leaving a hollow core behind.5
Figure 1. (a) Side view of a hollow-core antiresonant reflecting optical waveguide. Light propagates with wave vector β along the z direction in the liquid core. (b) Transmission spectrum of a 3mm-long and 12μm-wide air-core antiresonant reflecting optical waveguide (blue) and theoretical trace (red).
Waveguiding in ARROWs is achieved by selecting each of the cladding layers such that the transverse wave vector (kT) is antiresonant. Thus, the light in this direction is highly reflected at the interfaces and it remains confined to the core. The resonance condition is highly wavelength-dependent, which provides design flexibility for selecting the wavelengths to be transmitted through the device: see Figure 1(b).6
Fluorescence resonance energy transfer (FRET) is a technique that detects energy transfer between fluorescent molecules. It is commonly used to obtain information such as the distance between molecules, e.g. between fluorescent dyes.7 Integrated detection with built-in pump filtering would represent a major step forward in an application designed for on-chip single molecule detection of FRET signals: see Figure 2(a). Designing an ARROW to filter out the excitation beam (λexc) while transmitting both donor and acceptor fluorescence (λem1 and λem2) would also improve the signal-to-noise ratio. Figure 2(b) shows an ARROW device in which both excitation and detection paths run in the plane of the chip. This geometry not only enables spatial filtering, but also reduces the excitation volume, thus allowing single molecule sensitivity and parallelization.2
Figure 2. (a) Calculated transmission for a 15μm×4.5μm×1cm liquid-core fluorescence resonance energy transfer antiresonant reflecting optical waveguide (ARROW) filter. The design is for the dye pair Alexa 546 (donor) and Alexa 647 (acceptor). (b) Scanning electron microscope image of intersecting hollow- and solid-core waveguides. Inset: observed mode image.
Liquid-core ARROW waveguides are now considered most valuable in building planar optofluidic devices for single molecule detection. Their main advantages are the ability to guide both high- intensity light and liquids through the same micrometer-scale channels and the capacity to transfer light between solid- and liquid-core channels. Future developments will focus on incorporating these waveguides into larger optofluidic systems with the ultimate goal of achieving complete optical and fluid integration on the same chip.
We are grateful to Prof. Vahid Sandoghdar of the Eidgen¨ossische Technische Hochschule (ETH) Z¨ urich for his support setting up the broadband excitation scheme. Support for this research was provided by the National Academies Keck Futures Initiative under grant NAKFINano14, and by the National Institutes of Health (NIH) under grant R01EB006097 and the National Science Foundation (NSF) under grant ECS-0528730.
Solid State Physics/Nanometer Consortium
Philip Measor, Dongliang Yin, Holger Schmidt
School of Engineering, University of California/Santa Cruz
Santa Cruz, CA
Evan J. Lunt, Aaron R. Hawkins
Electrical and Computer Engineering, Brigham Young University
5. J. P. Barber, E. Lunt, Z. George, D. Yin, H. Schmidt, A. R. Hawkins, Integrated hollow waveguides with arch-shaped cores, IEEE Photonics Technol. Lett. 18, no. 1, pp. 28-30, 2006.doi:10.1109/LPT.2005.859990
6. U. Hakanson, P. Measor, D. Yin, E. Lunt, A. R. Hawkins, V. Sandoghdar, H. Schmidt, Tailoring the transmission of liquid-core waveguides for wavelength filtering on a chip, Proc. SPIE 6477, pp. 647715, 2007.doi:10.1117/12.703156