Hollow waveguide optimization for fluorescence based detection
ABSTRACT Previously, we created antiresonant reflecting optical waveguides (ARROWs) with hollow cores that guide light through gas and liquid media. We have demonstrated that these ARROWs can be used in sensing applications with single particle sensitivity using fluorescence correlation spectroscopy. To increase sensitivity for single molecule sensing, we have improved our initial designs and fabrication methods to decrease ARROW background fluorescence and improve transitions between solid and hollow waveguides. Photoluminescence of ARROW layers creates background fluorescence that masks the desired fluorescence signals. To improve sensitivity, we have optimized the PECVD ARROW layers to minimize the photoluminescence of each layer. Sensing applications require that hollow waveguides interface with solid waveguides on the substrate to direct light into and out of test media. Our previous ARROW designs required light at these interfaces to pass through the anti-resonant layers. Although in theory, high transmission through ARROW layers can be achieved, in practice, passing through these layers has limited transmission efficiencies. A new design coats the top and sides of the hollow core with only silicon dioxide, allowing light at interfaces to pass directly from silicon dioxide into the hollow core. This new design exhibits good mode confinement in the hollow core.
[show abstract] [hide abstract]
ABSTRACT: A design criterion that permits truly omnidirectional reflectivity for all polarizations of incident light over a wide selectable range of frequencies was used in fabricating an all-dielectric omnidirectional reflector consisting of multilayer films. The reflector was simply constructed as a stack of nine alternating micrometer-thick layers of polystyrene and tellurium and demonstrates omnidirectional reflection over the wavelength range from 10 to 15 micrometers. Because the omnidirectionality criterion is general, it can be used to design omnidirectional reflectors in many frequency ranges of interest. Potential uses depend on the geometry of the system. For example, coating of an enclosure will result in an optical cavity. A hollow tube will produce a low-loss, broadband waveguide, whereas a planar film could be used as an efficient radiative heat barrier or collector in thermoelectric devices.Science 12/1998; 282(5394):1679-82. · 31.20 Impact Factor
[show abstract] [hide abstract]
ABSTRACT: The basic theory of laser propagation in hollow waveguides is considered in the context of laser-plasma physics. The physical model of waves reflecting between the guide walls is used to show that there is a discrete series of modes, and to give the mode dispersion relation and losses in terms of a given reflectivity. The mathematical connection between this model and the solution of Maxwell's equations for lossless propagation in a cylinder is given. Thus the solutions for low loss propagation for any given reflectivity can be obtained, provided it is close to 1. Results are given using Fresnel reflectivity for perfect dielectric and finite conductivity waveguides. The relationship of the breakdown intensity in dielectric waveguides to known breakdown intensities is also derived. The practical implications for the guiding of intense laser pulses and the limitations of the model are discussed. The theory is shown to explain, at least qualitatively, a number of previous experimental results.Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics 12/2000; 62(5 Pt B):7168-80.
[show abstract] [hide abstract]
ABSTRACT: We present a fully planar integrated optofluidic platform that permits single particle detection, manipulation and analysis on a chip. Liquid-core optical waveguides guide both light and fluids in the same volume. They are integrated with fluidic reservoirs and solid-core optical waveguides to define sub-picoliter excitation volumes and collect the optical signal, resulting in fully planar beam geometries. Single fluorescently labeled liposomes are used to demonstrate the capabilities of the optofluidic chip. Liposome motion is controlled electrically, and fluorescence correlation spectroscopy (FCS) is used to determine concentration and dynamic properties such as diffusion coefficient and velocity. This demonstration of fully planar particle analysis on a semiconductor chip may lead to a new class of planar optofluidics-based instruments.Lab on a Chip 10/2007; 7(9):1171-5. · 5.67 Impact Factor
Hollow waveguide optimization for fluorescence based detection
Evan J. Lunt*a, Brian S. Phillipsa, Cory J. Jonesa, Aaron R. Hawkinsa,
Philip Measorb, Sergei Kuehnb, Holger Schmidtb
aECE Department, Brigham Young University, 459 Clyde Building, Provo, UT 84602;
bSchool of Engineering, University of CA Santa Cruz, 1156 High Street, Santa Cruz, CA 95064
Previously, we created antiresonant reflecting optical waveguides (ARROWs) with hollow cores that guide light through
gas and liquid media. We have demonstrated that these ARROWs can be used in sensing applications with single
particle sensitivity using fluorescence correlation spectroscopy. To increase sensitivity for single molecule sensing, we
have improved our initial designs and fabrication methods to decrease ARROW background fluorescence and improve
transitions between solid and hollow waveguides. Photoluminescence of ARROW layers creates background
fluorescence that masks the desired fluorescence signals. To improve sensitivity, we have optimized the PECVD
ARROW layers to minimize the photoluminescence of each layer. Sensing applications require that hollow waveguides
interface with solid waveguides on the substrate to direct light into and out of test media. Our previous ARROW designs
required light at these interfaces to pass through the anti-resonant layers. Although in theory, high transmission through
ARROW layers can be achieved, in practice, passing through these layers has limited transmission efficiencies. A new
design coats the top and sides of the hollow core with only silicon dioxide, allowing light at interfaces to pass directly
from silicon dioxide into the hollow core. This new design exhibits good mode confinement in the hollow core.
Keywords: Integrated optics, sensors, ARROW waveguides, microfluidics, hollow waveguides, photoluminescence.
Hollow waveguides have received a great deal of attention in the optics world over the past decade. Excitement over
hollow waveguides arises from two of their features. First, light guiding is achieved without total internal reflection,
which is the mechanism employed in traditional optical waveguides. Second, since total internal reflection is not
necessary, light can be guided in very low index media, like water (n=1.33) or air (n=1). This allows for high optical
intensity light to interact with gaseous or liquid media over long distances. Optical sensors for these media using hollow
waveguides have been generating considerable interest.
During their development, hollow waveguides have taken various forms, including photonic crystals1, Bragg waveguides
and omniguides2, Fresnel reflecting hollow channels3, and antiresonant reflecting optical waveguides (ARROWs)4.
Hollow waveguides with the lowest loss have been produced with photonic crystal fibers, but fibers are not always the
ideal form for creating optical sensors. Planar, on-chip sensors offer many packaging and scalability advantages,
including methods for introducing air or liquid samples into hollow waveguides. ARROW-based waveguides are a
natural choice for on-chip sensors. We have pursued ARROWs fabricated through surface micromachining that can be
made into integrated networks for both liquid and gas sensing.
We have previously demonstrated that ARROWs designed to be filled with liquids can be used for very sensitive
measurements in biology and chemistry such as to identify single fluorescent particles through fluorescence correlation
spectroscopy5. In this case, a sensor platform was constructed that took advantage of a surface microfabrication process
to produce both solid and hollow core waveguides. These waveguides intersected each other and were used to carry light
through the medium of interest as well as on and off the chip to light sources and detectors (see Figure 8).
However, in order to increase ARROW sensitivity for single molecule detection, they can be improved in a number of
ways. These include design and fabrication improvements. This paper describes specific advances to decrease ARROW
background fluorescence and improve transitions between solid and hollow waveguides. We will first provide a brief
description of the ARROW principle and our standard sacrificial-etching based fabrication technique.
*firstname.lastname@example.org; phone 1 801 422-5414; fax 1 801 422-0201
Advanced Fabrication Technologies for Micro/Nano Optics and Photonics,
edited by Thomas J. Suleski, Winston V. Schoenfeld, Jian Jim Wang, Proc. of SPIE
Vol. 6883, 68830H, (2008) · 0277-786X/08/$18 · doi: 10.1117/12.763448
Proc. of SPIE Vol. 6883 68830H-1
2008 SPIE Digital Library -- Subscriber Archive Copy
2.1 ARROW principle
ARROWs allow light to propagate in lower index media by utilizing Fabry-Perot reflectors in the direction transverse to
light propagation. Fabry-Perot reflectors constructed from dielectric layers act like mirrors, exhibiting high reflectivity
when operating at anti-resonance6. The finite transmission of the dielectric cladding layers makes the ARROW by
definition a leaky waveguide. However, the loss decreases when more antiresonant dielectric layers are added. For our
design, each additional layer decreases the loss by approximately three times7. The ARROW is also effectively a single
mode waveguide since higher order modes are suppressed by high optical losses. To achieve the anti-resonant, highly
reflective condition, the dielectric layers must be deposited to a fairly precise thickness determined by equation (1) 6.
1 ) 1
In the equation, nj and nc are the index of refraction of the jth layer and the core, respectively, dc is the thickness of the
core, λ is the wavelength, and M is an integer representing the anti-resonance order. These correspond with Figure 1,
which illustrates a cross sectional view of an ARROW.
Fig. 1. Cross section illustration of an ARROW waveguide.
2.2 ARROW fabrication
We have chosen to pursue a true bottom-up fabrication procedure that relies on two well-known microelectronics based
processes. The first is the chemical removal of a material applied to a substrate, and the second is chemical vapor
deposition (CVD) of silicon based thin films. The heart of our process is the creation of hollow tubes by surrounding a
sacrificial core with silicon dioxide or silicon nitride and then removing the sacrificial layer with acid etching. This
process is depicted in Figure 2.
Fig. 2. Fabrication steps used to create hollow ARROW waveguides based on removal of a sacrificial core.
Proc. of SPIE Vol. 6883 68830H-2
680 700720740760 780
First, a silicon substrate is coated with alternating layers of silicon dioxide and nitride using plasma enhanced chemical
vapor deposition (PECVD). These serve as the Fabry-Perot reflectors of the ARROW. This process takes place at
approximately 250°C. A thin layer of sacrificial material is then deposited and defined into a thin line using
photolithography and etching. A variety of sacrificial materials may be used, including photosensitive polymers and
metals. Overcoat layers of PECVD oxide and nitride are then grown which cover the sacrificial material and form the
roof of the ARROW. The conformal nature of this process is important to ensure that the sacrificial layer is completely
enclosed. The final step of the procedure is to expose the sacrificial material to an acid etch from either end of the
channel. Upon completion of the etching, we are left with a hollow tube with walls composed of either silicon dioxide or
A number of sacrificial layers have been investigated in the context of our outlined fabrication process: aluminum8, SU-
8 (a photosensitive epoxy)9, and photoresist10. Aluminum is most quickly removed using a nitric and hydrochloric acid
etching solution, while SU-8 and photoresist are removed using a sulfuric acid and hydrogen peroxide solution. The
different sacrificial materials result in different hollow core cross sections, providing flexibility when designing
waveguides. Typical cross sections range from 10 to 50 µm in width and around 3 to 6 µm in height.
3. DECREASING FLUORESCENCE
3.1 Background Fluorescence
Hollow ARROW waveguides are particularly useful for fluorescence sensing applications, including fluorescence
correlation spectroscopy4. In sensing applications, the desired signal must be distinguishable from the fluorescence
background for reliable detection. Silicon, silicon nitride, and silicon dioxide have long been known to
photoluminesce11,12. For our integrated ARROW platforms, we investigated the relative photoluminescence (PL)
contributions of PECVD silicon nitride and silicon dioxide for layers 150 nm thick on silicon wafers (Figure 3). For our
PL tests, normal excitation was provided by a HeNe laser (λ=632.817 nm), and an objective (50x, 0.5 NA) was used to
couple light into a LabRAM HR spectrometer (Horiba Jobin Yvon).
Fig. 3. Detected PL for PECVD silicon nitride and silicon oxide layers.
Figure 3 shows that, for layers of equal thicknesses, the PL of PECVD silicon nitride is significantly higher than that of
silicon dioxide in the measured wavelength span. Although our ARROW waveguide design calls for a greater total
thickness of oxide than of nitride, the much higher nitride PL per unit thickness means the nitride layers are the main
sources of background fluorescence in ARROW sensors.
Proc. of SPIE Vol. 6883 68830H-3
660680 700720 740 760780
3.2 PECVD recipe optimization
Our hollow ARROW waveguides use several layers of silicon nitride. Silicon nitride is also commonly used in many
other integrated optics applications, including in Bragg mirror stacks. In many applications, excessive PL of nitride
layers can be detrimental to the overall performance.
The characteristics of PECVD silicon nitride layers vary greatly according to growth parameters. In order to decrease the
overall background fluorescence for our integrated ARROW platforms, we investigated how the PL of nitride varies
with the refractive index of the film and the substrate temperature during deposition. As many of our fluorescence
sensing applications utilize a detection window in the wavelength range of 660-780 nm, we have focused on the PL in
this range. We controlled the index of our nitride films by altering the ratio of gas flows (SiH4:NH3). Increasing this ratio
increases the ratio of silicon to nitrogen in the deposited film and increases the refractive index of the film. Figure 4
shows the relative PL intensity of nitride layers with different refractive indexes, deposited at 250°C.
Fig. 4. Detected PL for PECVD silicon nitride with different refractive indexes at 250°C.
Figure 4 shows that the PL intensity for nitride tends to decrease with decreasing index. Others have reported that the PL
of silicon nitride originates from defects and interface effects in the amorphous silicon nitride films13. Since higher-index
nitride films are more silicon-rich than lower-index films, it is likely that the defects and PL centers in the film are
increased by increasing the relative number of silicon atoms.
To compare the PL intensity for nitride at different temperatures, we deposited the same thickness of films (300 nm) at
250°C and 300°C. Figure 5 shows the resulting PL spectra.
Fig. 5. Temperature dependent PL in PECVD nitride.
Proc. of SPIE Vol. 6883 68830H-4
As shown in Figure 5, the PL intensity collected for the same index is lower for 300°C than for 250°C. Others have also
reported that increasing the deposition temperature does decrease the PL for PECVD films12. Therefore, our results
showed that PECVD nitride PL decreases with increasing deposition temperature, as expected.
3.3 PL from solid-core waveguides
Our ARROW test platforms use solid-core waveguides to couple light into and out of the hollow ARROW waveguides
(see Figure 8). In addition to investigating the PL contributions from the ARROW layers themselves, we also
investigated the relative PL contributions from two different types of solid waveguides (see Figure 6).
Fig. 6. a) Rib waveguide on ARROW layer stack. b) SU-8 waveguide on top of an ARROW layer stack.
The first type was rib waveguides etched in the top thick oxide ARROW layer (Figure 6a). This type of waveguide is the
type that we have previously used in integrated ARROW designs4. The second type was waveguides of patterned SU-8
on top of the ARROW layer stack (Figure 6b). SU-8 was chosen as a waveguide because of its low optical loss. In fact,
using the structure shown in Figure 6b, we have created waveguides with losses as low as 0.035 cm-1 for rectangular
waveguides 5x12 um. SU-8 has low loss because the sidewalls in patterned SU-8 are very smooth, and SU-8 does not
have many scattering centers. The relative PL spectra from these two types of solid waveguides are shown in Figure 7.
Fig. 7. Detected PL for waveguides made from PECVD oxide and SU-8.
As shown in Figure 7, the oxide waveguides have much lower PL than those made from SU-8. Although SU-8 has very
low optical loss, its high PL precludes it use as solid waveguides in our ARROW fluorescence based sensing platforms.
Proc. of SPIE Vol. 6883 68830H-5