Laser Microfabrication: Paving the Way Toward the Optofluidic Lab-on-a-Chip

---

By Matthew Wronski, Peter R. Herman, Phd, Jianzhao Li, PhD, James Dou, Nigel Munce and Lothar Lilge, PhD

What is Optofluidics?
The advent of micro total analysis systems (m-TAS) is revolutionizing routine analytical tasks that can now be carried out much faster than previously possible, requiring substantially reduced reagent and analyte volumes. Such systems often employ microfluidic channels for analyte transport and manipulation. High-throughput stand-alone diagnostic microchips with applications in remote sensing, environmental monitoring or homeland security require scalable and fail-safe detection schemes. While electrochemical, conductivity and amperometric detection methods have demonstrated reasonable performance, optical methods such as laser-induced fluorescence or interferometry offer unmatched sensitivities and superior resolution.

The disadvantage with optical detection is the large size of traditional bulky confocal microscope and spectrophotometer instruments. Optofluidics is now an emerging technology that combines microfluidics with integrated optics and, as such, provides largely miniaturized alignment-free analytical devices capable of single-molecule detection on a compact module. Last year, the DARPA Center for Optofluidic Integration at the California Institute of Technology (Pasadena, CA) set out to explore the potential of this novel technology with applications ranging from adaptive information processing to bio-sensing.1

Why Laser Microfabrication?
As with any nascent technology, fabrication techniques that allow rapid device prototyping are essential. Having a final product by the end of the day is a crucial outcome of a quick fabrication process. Furthermore, rapid prototyping enables the fabrication of highly customized optofluidic devices tailored for particular analytical modalities such as flow cytommetry or capillary electrophoresis.

Laser microfabrication is a particularly viable means of prototyping complex optofluidic microchips. The process typically involves a focused laser beam and computer-controlled motion stages that either deflect the beam or move the sample. By adjusting the laser energy, optical focus and scan velocities, a wide range of biophotonic components may be written on or inside the substrate. Additionally, by tuning the process parameters, individual components may be specifically tailored for custom applications.

Although alternative fabrication processes such as soft lithography are available, laser processing offers a number of inherent advantages. A variety of materials may be processed — such as glass, polymer or silicon — and sub-micron feature resolutions are commonly supported. Given that no masks are used in such a direct-write approach, components may be readily added and trimmed at any step in the process. Thus, a diverse range of m-TAS applications may be flexibly targeted with a single process. Particularly exciting is the novel capability today to write optical devices inside the bulk of transparent materials that interface with microfluidic and other micro-laboratory components.

The Laser Nanofabrication Facility (LNF) at the University of Toronto (Toronto, ON) houses several laser processing stations. The focus is mostly set on processing transparent media that are challenging to structure because of the low absorption of standard laser systems. High-resolution micromachining of high-purity fused silica — a material that is commonly used in both chromatography and photonic circuits — is possible with a fluorine excimer laser that operates at a 157-nm wavelength, shown in Figure 1. The high-energy photons associated with the deep-UV radiation are efficiently absorbed in glasses and polymers to enable high-resolution crack-free nanostructuring. Furthermore, the F2 laser enables surface or refractive index modification inside glass.

The LNF also provides two novel burst-mode ultrafast laser systems that produce high repetition rate (several MHz) pulses of 40 to 400 femtosecond (10 to 15 seconds) duration. Although the ultrafast lasers operate in the near-infrared region, the extremely high intensities (more than 1012 W/cm2) produced only in the focal volume drive highly localized non-linear absorption and thus provide a different means of nanostructuring transparent materials. Glass and polymers are also highly transparent at near-infrared wavelengths, allowing these femtosecond lasers to focus deep inside the material, thus enabling truly 3-D optofluidic integration. The ultrafast laser processing setup is shown in Figure 2.

Integration is Key
Deep ultraviolet and femtosecond laser processing at the LNF has enabled the fabrication of a wide range of components with direct applications in optofluidics. These components include channels, wells, vias, lenses, waveguides and gratings, as well as electrodes.

By modifying the F2-laser energy, scan speed and beam shape, microchannels with different sidewall profiles may be machined. Certain applications such as electrophoresis or single-cell transport benefit from a circular channel profile, while a rectangular geometry is better suited when channels are combined with micro-optical components. Other fluidic components include high-aspect ratio microvias for interconnecting multilayered channels and microwell arrays that facilitate single-cell trapping and analysis. Surface channels, such as the ones on the right in Figure 4 or on the left in Figure 3a may be effectively sealed with elastomeric materials such as PDMS. Moreover, buried microchannels or tunnels have also been demonstrated by combining ultrafast laser processing and chemical etching techniques.2,3

A variety of telecom-quality photonic components may be integrated alongside the fluidics. Precise F2-laser surface nanostructuring provides an excellent means of shaping integrated lenses — such as the one shown in Figure 3d, which may substantially improve fluorescence collection efficiencies from micro-analytical chambers — while high-periodicity gratings offer much potential for on-chip spectral analysis. Buried optical waveguides are another key component, and are used for guiding optical signals around the diagnostic microchips. Intercepting waveguides and microfluidic channels enables highly sensitive fluorescence or interferometric detection directly on the biochip or m-TAS. Figure 3a shows such a device in which an optical fibre was permanently affixed to the waveguide by means of a UV-curable epoxy. Bonding fibres to optofluidic chips enables sturdy alignment-free connections between on-chip waveguides and external laser sources or detectors. Waveguides formed in bulk glasses may be used to route light around bends (Fig. 3b) or to split a single laser beam for fluorescence excitation and probing of multiple channels or wells. The capability of fabricating channels and waveguides at different depths below the chip surface opens the door to massively parallel high-throughput detection for far-reaching applications such as drug screening or protein analysis.

Integration of a number of these components on a single biophotonic chip for single-cell proteomic studies is the goal behind a collaborative effort between the LNF and the Ontario Cancer Institute at the Princess Margaret Hospital in Toronto, Ont. Commercially available glass and polymer microfluidic chips have been custom-processed using the F2 laser system to enable parallel single-cell electrophoretic separation. Figure 4 shows a polymer chip obtained from Micronit Microfluidics bv (Enschede, the Netherlands) in which microchannels several microns wide were written with the F2 laser. Efficient extraction of bioanalytes from single acute myeloid leukemia cells was accomplished with precisely laser-etched injector structures shown on the right of Figure 4. The future effort is aimed at accurately laser-patterning electrode structures on the chip surface for dielectrophoretic trapping and manipulation of single cells.

The New Frontier
Integrated optofluidics technologies are poised to revolutionize the fields of analytical chemistry and biology and target a vast range of environmental, medical, pharmaceutical and defence-related application areas. Recent progress in microfluidics has spurred commercial growth of fluidic-based laboratory products by companies such as Agilent Technologies Inc. (Palo Alto, CA) and Micralyne Inc. (Edmonton, AB). Furthermore, CAD packages offered by companies such as Coventor Inc. (Cary, NC) and IntelliSense Software Corp. (Woburn, MA) enable microfluidic system design. Integrated optics, which has traditionally fuelled the telecommunications industry is now finding a niche in the new and exciting world of m-TAS.

Being a viable rapid prototyping technology, laser microfabrication offers a great deal of processing latitude. Deep ultraviolet and ultrafast lasers are particularly well suited for high-resolution structuring of 3-D fluidic and photonic components in a wide range of transparent media. The future biochip will likely include fluidic sample transport, optical probing and electronic processing. The laser-machining process presents a great deal of flexibility for fabricating and integrating such devices, thus advancing the forefronts of modern diagnostic instrumentation.


References:
1) Optofluidics: What is Optofluidics? 2005. DARPA Center for Optofluidic Integration. Accessed August 8, 2005. (http://www.optofluidics.caltech.edu/optofluidics)

2) Bellouard, Y., A. Said, M. Dugan, and P. Bado. “Fabrication of high-aspect ratio, micro-fluidic channels and tunnels using femtosecond laser pulses and chemical etching.” Optics Express 12.10 (2004): 2120.

3) Fisette, B. and M. Meunier. “Femtosecond laser three-dimensional microstructuring inside photosensitive glasses.” Proc. of the SPIE 5578 (2004): 677.


Matthew Wronski completed an M.Sc. degree in electrical and computer engineering at the University of Toronto (U of T) (Toronto, ON) in the area of laser microfabrication of optofluidic biochips. He is currently pursuing a PhD in medical biophysics at U of T, and is stationed at the Sunnybrook and Women’s Research Institute (Toronto, ON). Peter R. Herman, PhD is a professor in U of T’s department of electrical and computer engineering. Jianzhao Li, PhD is a post-doctoral fellow in Herman’s group. James Dou is completing his M.ASc. in electrical and computer engineering at U of T, and is in Herman’s group. Nigel Munce is a doctoral student in U of T’s department of medical biophysics. Lothar Lilge, PhD is a scientist at the Ontario Cancer Institute (Toronto, ON), and an assistant professor at U of T’s department of medical biophysics.