Medical Device Link.

Originally Published IVD Technology June 2002

Processing Technologies

Developing microfluidics and bioMEMS

Jay N. Sasserath and David Fries

A processing technology offers alternatives for designing and manufacturing microcomponents.

Simple microfluidic channels that were formed in riston 8080 negative photoresist. Microfluidics can be rapidly prototyped using maskless photolithography, allowing for faster development of new devices.

MEMS are also used for biomedical applications and are called either bioMEMS or microfluidics, which use liquid flow as part of their normal operation.¹ Examples of microfluidics include miniature biochemical reaction chambers, lab on a chip, and microelectrode arrays.² Since these components do not have the same properties as other MEMS, they cannot be made using the same processes. Manufacturers have therefore had to use other fabrication methods and techniques. Maskless photolithography is one such technique for prototyping and developing bioMEMS and microfluidics.

Key Differences

A microfluidic channel that was produced from a master image using an embossing technique. The master image for this component was fabricated using maskless photolithography.

In order to understand the challenges that biomedical manufacturers encounter, it is necessary to compare the processes that are used to produce bioMEMS and microfluidics with those methods utilized in a silicon product fabrication line.³ The comparisons highlight the differences in the physical makeup of the components, as well as economic considerations.

Substrates and Materials. While silicon product fabricators use primarily standard, well-defined silicon wafers, bioMEMS and microfluidics makers are often forced to work with nonsilicon materials.⁴ The reasons for this restriction include the incompatibility of silicon with many biological fluids, and the manufacturers' need to start with low-cost materials. Examples of substrates that are used in bioMEMS and microfluidics include glass slides, polypropylene and other polymeric materials, and rigid plastics that have topography.

Size Features. While manufacturing processes for silicon products strive for higher densities and small, sub-micron geometries, bioMEMS and microfluidics often require thicker materials and larger features. Since many biomaterials, such as red blood cells, that are used in bioMEMS and microfluidics tend to be large, the channels and other parts within these components also need to be large. In addition, the materials from which these components must be constructed are easier to process if they are larger and thicker. The result is that many of the essential pieces of process equipment used in producing silicon products are not needed for bioMEMS and microfluidics fabrication.

Figure 1. A schematic of the SF-100 maskless exposure system. (Click to enlarge)

Rapid Prototyping. Because the processes involved in making metal oxide semiconductors and other silicon-based components are well known and documented, a number of models exist for designers to use. Such models enable manufacturers to design new silicon components rapidly, which means that the fabrication of a new component may take up to a few weeks. BioMEMS and microfluidics designers also have many models at their disposal. However, due to the wide variety of materials and designs, the models cannot accurately predict all cases and designs. They must rely on empirical test results to finalize component design and fabrication. Hence, manufacturers of bioMEMS and microfluidics encounter multiweek fabrication cycles, which could slow down component development to unacceptable levels.

Cost Sensitivity. Because production volumes are typically smaller than comparable silicon product runs, bioMEMS and microfluidics makers are more cost-sensitive than silicon product makers.⁵ While electronic companies order silicon components in lots of several thousands, bioMEMS and microfluidics are usually ordered in lots of 50–100. This situation presents an economic challenge to bioMEMS and microfluidic fabricators that need to produce low-cost components in small quantities.

Standard Photolithography

Using maskless photolithography for fabricating bio-MEMS and microfluidics enables companies to create prototypes and test new ideas, while realizing reduced capital costs and a faster time to market with new products.

Figure 2. A photo of the SF-100 maskless exposure system.

Photolithography is the key process for transferring images to an electronic or microfluidic component and is used in manufacturing semiconductors and other components. Standard photolithography processes use photomasks, glass plates that are selectively patterned with chrome or another opaque material. These masks are a critical part of the image-transfer process. The standard photolithography process involves four major steps: coating, exposure, development, and hardbake.

Photoresist Coating. To create the photopatternable surface, a substrate is coated with photoresist, a liquid polymeric material onto which the image will be transferred or projected during the photolithography process. To ensure even coating over the entire substrate surface, photoresist is deposited onto the substrate surface while the substrate is spinning at speeds between 1000 and 5000 rpm. An alternative is to employ dry-film photoresists that can be laminated on the substrate.

Exposure. Once the substrate has been coated with photoresist, it is exposed using a system designed for the exposure process. In standard processes, the system shines light through a photomask that has on it a master image of the component. By shining light through the photomask and onto the substrate, individual areas of the photoresist are selectively exposed to light, which causes a chemical change in the photoresist.

Figure 3. A thin-film glucose monitor component. (Click to enlarge)

Development. After the exposure phase, the substrate is typically immersed in an aqueous developer solution that dissolves the areas of the photoresist that were exposed to light. Once this step is completed, the photoresist is patterned with the master image that was on the photomask.

Hardbake. The substrate is then baked in an oven or on a hot plate at temperatures between 100° and 120°C. Hardbaking is needed to remove any liquids that may have been absorbed by the substrate and to cross-link the remaining photoresist. Cross-linking the polymer increases the mechanical and chemical stability of the material, allowing it to be used in further processing of the substrate.

Photolithography Systems

Examples of photolithography systems that require photomasks for image transfer include the following.

Contact and Proximity Printers. In these systems, typical processes require that the mask be held very close to the substrate being exposed or that the photomask actually come in contact with the substrate during exposure. The key benefits of these systems include a well-understood, proven technology.

Monochromatic Steppers. These systems use a single wavelength of light to obtain high resolution. During the photolithography process, a small area, typically 1.0 x 2.0 cm, is exposed on a substrate. Although costly, these systems are effective when small, submicron features are required for component fabrication.

Maskless Photolithography

With standard photolithography, both the cost and the time involved in component processing are heavily influenced by the availability and cost of the photomask that is used to transfer the pattern. One solution would be to employ a maskless technology that could yield a large number of designs depending on the component. One such maskless exposure system is the SF-100 by Intelligent Micro Patterning (St. Petersburg, FL).

Figure 4. A photoresist-patterned polycarbonate plastic material. (Click to enlarge)

The SF-100 is a photo imaging system that uses any full-scale image created on a Windows-based computer and reduces it to a size as small as 5 µm, while maintaining all relative proportions and resolutions. The SF-100 is also used in the exposure step of a typical photolithography process, in which a standard Windows-based PC interfaces directly with the system, providing system control and image storage for the exposure process (see Figures 1 and 2).

In the SF-100, any Windows-based software can be used to create the desired design for a microchip, MEMS, or microfluidics. Once the design is transferred to a Smart Filter assembly that allows an optical image to be generated from an inputted electronic signal, light is introduced into the system using a polychromatic white light source.

Complex microfluidic channels that were formed in positive photo polymeric materials.

A directly coupled optical delivery system ensures efficient transfer of the image to the Smart Filter subassembly, which incorporates all of the optical and electronic components necessary to transfer an image onto the substrate. The projected image is free of distortion and is uniform throughout the exposure area. The light emanating from the Smart Filter then shines directly onto the surface of the substrate.

Since the area of the image is typically only a few square centimeters, a step-and-repeat motion can be used to expose the entire surface of the substrate. To move the substrate precisely, an x-y-z stage is built into the base unit with piezoelectronic motors that provide step increments of 0.25 µm.

Multiple layers are often required for more-complex components, in which many functions are integrated to provide greater device performance. By using a high-resolution microscope above the substrate, the user can control the image-to-substrate alignment, providing the capability of fabricating multilayer devices.

To avoid substrate exposure while image-to-substrate alignment is adjusted, a removable UV filter can be placed between the light source and the substrate. If the filter were absent, the photoresist would be exposed during the alignment process, resulting in a thinner photoresist. This effect would cause defect and repeatability problems at subsequent etch and deposition steps.

Alternative Systems

Maskless photolithography offers a number of advantages and disadvantages when compared with standard photoresist exposure systems that require photomasks (see Table I).

Other maskless photolithography systems are currently available on the commercial market. Examples of such other systems include laser writing systems that use a laser to write an image directly onto a photoimagable material; electron-beam writing systems perform the same function with an electron beam. In both of these systems, the images are typically written pixel by pixel, and special photoresist materials must be used for the processing.

Product Results

A number of advanced components have been designed, fabricated, and tested using maskless photolithography, including those described below.

Figure 5. A polymer lattice used for biomaterials growth. (Click to enlarge)

A thin-film glucose monitor component is an example of how the maskless photolithography technology can be applied (see Figure 3). The ability to process thin-film plastic materials is critical for many bioMEMS and microfluidics. The low cost of plastic materials and the ease of processing makes them ideal for many biotech applications where cost is critical. Also, the chemicals in many plastics are very compatible with biofluids and biomaterials. The photolithography process described earlier has been used in many such applications. For example, a 50 x 50 mm plastic square was patterned with standard photoresist materials (see Figure 4). To fabricate a more complex structure, further processing could be required, and other steps, such as plating or etching, could be performed. In its present form, this component has the characteristics that are sought in a disposable diagnostic device. The substrate is made of polycarbonate, which can be injection molded to high tolerance. The patterned layer can be coated with laminate, followed by a photopatterning to create the analytical sequences.

Another component is a polymeric matrix, which is a three-dimensional lattice that is produced by exposing multiple layers of photosensitive materials sequentially (see Figure 5).⁶ Through the repeated exposure of individual layers, a three-dimensional structure can be produced. Such a structure has utility as a cell-capturing matrix or as a scaffolding structure for tissue engineering. Using this multiple-layer exposure technique, combined with the maskless photolithography process, biomaterials can be grown in specific shapes and sizes, which may be useful in future tissue transplant or regeneration procedures.

Figure 6. A 750-µm-diam stainless-steel rod patterned around the circumference with photoresist.

Other advanced products are also available using the Smart Filter technology (see Figure 6). For example, a 750-µm-diam cylinder has been patterned with photoresist materials around its entire circumference and has been used in medical devices. The examples above demonstrate that maskless photolithography offers medical diagnostic device designers the ability to fabricate mechanical components and integrate them into mechanical devices.

Conclusion

Maskless photolithography provides rapid prototyping capabilities for fabricators of bioMEMS and microfluidics. Although similar in principle to silicon-based commercial techniques, a number of critical changes have been implemented to ensure the success of this technology.

Maskless photolithography ensures that the next generation of bioMEMS and microfluidics can be developed quickly and at a reasonably low cost.

References

1. M Mehregany, "Chemical and Biomedical Sensors," in MEMS Technology Tutorial, ed. S Walsh (Mountain View, CA: Semiconductor Equipment and Materials International, 1997), 70–86.

2. N Maluf, An Introduction to Microelectromechanical Systems Engineering, (Norwood, MA: Artech House, 2000), 163–176.

3. Viva La Revolución—Advances in Semiconductor Processing Technology, (San Francisco: Thomas Weisel Partners, 1999), 67–153.

4. A Heuberger and R Hintsche, "Electrical Biochips: A New Class of MST-Devices" (presented at International MEMS/MST Industry Forum, Munich, April 23, 2001).

5. R Grace, "Microelectromechanical Systems (MEMS): U.S. Market Overview" (presented at Sensors Expo, Detroit, October 23, 1997).

6. D Fries, "Direct Writing of Tissue Support Scaffolding Using Microoptoelectromechanical Systems" (presented at BioTechnologies for Spinal Surgery; Halle, Germany, April 11–13, 2002).

Jay N. Sasserath, PhD, is chief executive officer and David Fries is chief technical officer at Intelligent Micro Patterning, LLC (St. Petersburg, FL). The authors can be reached via jays@intelligentmp.com and dfries@intelligentmp.com, respectively.

Copyright ©2002 IVD Technology