How to make glass microfluidic device? mask design and photolithography

During my PhD at the University of Wyoming, I may have made ~400-500 microfluidic devices all from glass substrate. My PhD laboratory uses borosilicate based substrate/cover plate to make the chips for variety of applications such as immunoassay development, separation experiments, fuel cell units. Glass is one of the oldest materials in microfluidic field and has certain advantages (see below) over other materials that are being used in this field of research.

1. Inertness to many chemicals
2. Optical transparency
3. Low fluorescence
4. High resistance to mechanical stress
5. Well established surface modification procedures

Below are the steps taken for the fabrication of glass microfluidic device: starting from designing the channel network using CADopia software to bonding the glass plates. I will take you through each of these steps.

1. Mask design: We create a channel design using CAPopia drawing software. I had never used any drawing softwares before and found this one easy to use. Once the design is complete, we send it out to a printing company (Fineline Imaging, Colorado) to get the mask. 

Mask with channel design

2. Photolithographic procedure: Once the photo-mask arrives from the printing company, we perform basic photo-lithography procedure to transfer the channel pattern design onto the glass substrate.  We buy borosilicate glass substrate (Telic, CA, USA) which is 4” x 4” in dimension and 1.65 mm thick. One side of this glass substrate is coated with a layer of chromium and photoresist on top of each other. Both chromium and photoresist layers are about 100-200 nm in thickness. The photoresist coated on the substrate is polymer based positive photoresist which is sensitive to UV radiation. Therefore the box containing the substrate plates must be opened only in dark room.

In photolithography room turn on the UV light source before at least 15 min you plan to shine it to the substrate plate. We use a custom made box where one can place photo-mask on top of substrate plate. This assembly is then gently pressed by putting a glass plate on top of it. The whole thing is inside a box in which a shutter can be opened when ready to expose to UV radiation.

Typically we expose the substrate to the UV radiation for ~ 30-45sec. During this step the polymer of photoresist breaks down and is removed by soaking the substrate in a photo-developing solution (MF-319, Rohm and Haas) for ~ 5 min. The substrate is then washed with deionized (DI) water (DirectQ Water Purification System, Millipore) and dried by blowing N2 gas. After this washing step, one can clearly see the channel pattern on the glass substrate. 

click HERE for second part of this post

Microfluidic Approaches to Improving Enzyme-Linked Immunosorbent Assays

This was the title of my PhD dissertation work. I successfully defended it on Tuesday. Now it feels really great!!! Just to give you some idea about my dissertation work, below is the general description of it. Details of the research work can be found in the dissertation which will be available online soon and in upcoming publications as journal articles. 

Enzyme-linked immunosorbent assay (ELISA) is arguably one of the most practiced techniques used for reliable quantitation of biological analytes in complex sample matrices. The advantages of ELISA arise mainly from the specific interaction between an antigen and its corresponding antibody, and the signal amplification due to an enzyme reaction that produces multitude of detectable species per binding event of the target molecule to the assay surface. The conventional methods of performing ELISA are predominantly based on polystyrene microtiter plates that require relatively large amounts (~100 µL) of expensive and/or precious samples or reagents. This volume may not be large itself but it becomes an issue when one has to deal with expensive reagents and limited amount of sample. For example, analysis of multiple biomarkers from the content of a single cell. There is a need to miniaturize these ELISA methods. Moreover, the limited ability of this assay format to measure biomarker concentrations (e.g., antigens/antibodies) circulating in bodily fluids restricts its use in the detection of several fatal diseases (e.g., cancers) at an early stage. Therefore, there is a need to improve the minimum analyte concentration detectable (limit of detection) by these techniques. In addition to this insufficient assay sensitivity, the microtiter plate version of ELISA requires that the enzyme substrate undergoes a change in its spectral signature during the enzyme reaction necessitating that its fluorophore be acted upon by the enzyme-label. This constraint, in turn, limits the number of possible enzyme-substrate couples available for use in the ELISA method. 

In my dissertation work, these major limitations of conventional ELISA methods have been addressed to broaden their utility in biomedical applications. The sample/reagent volume requirement of microtiter plate based ELISA has been reduced by simultaneously miniaturizing this assay and enhancing its multiplexing capabilities using the microfluidic platform. In an effort to improve the limit of detection, we have developed a pre-concentration ELISA method based on field amplified stacking of enzyme reaction product molecules in the ELISA micro-channel using two buffers of different ionic strengths.

In addition, we have demonstrated an ELISA method for the first time in which the enzyme substrate does not need to undergo a change in its spectral properties to allow its distinction from the enzyme reaction product in the assay chamber. We have accomplished this goal by synthesizing a rhodamine B based substrate molecule for the alkaline phosphatase enzyme-label, and then applying it to measuring the concentration of human TNF-α in a microfluidic device.

Lastly, a microfluidic method for simultaneous pre-concentration of cationic and anionic chemical species is described in my dissertation. This method is again based on the field amplified sample stacking process, and has been realized by injecting a relatively large plug of a sample prepared in a low-conductivity buffer into a microfluidic channel.