A stacked microfluidic device for improving experiment throughput

Jiandong Wuab, Xun Wubc and Francis Linabcd*
a Department of Biosystems Engineering, University of Manitoba, Canada
b Department of Physics and Astronomy, University of Manitoba, Canada
c Department of Immunology, University of Manitoba, Canada
d Department of Biological Science, University of Manitoba, Canada
Email: flin[at]physics.manitoba.ca

Why is this useful?

Owing to the advantages in miniaturization and cellular microenvironmental control, microfluidic devices have been increasingly applied to cell biology research [1]. Particularly, microfluidic devices can precisely configure chemical concentration gradients and flexibly manipulate the gradient conditions in space and in time [2, 3]. Various microfluidic gradient-generating devices have been used for studying cell migration and chemotaxis [2, 3]. These studies rely on live cell microscopy and usually only one experiment can be performed at a time. Previously, a double gradient device was demonstrated for parallel cell migration experiment with a motorized stage to image cells in different gradient channels [4]. However, the full XYZ motorized stage is expensive and thus often times limits the practical use of high-throughput microfluidic devices.

To overcome this limitation, here we report a stacked microfluidic device that allows parallel live cell imaging experiments on a single chip with only a Z motorized stage. This device is fabricated with multiple stacked layers of PDMS devices, and the cell imaging channels in each layer are aligned so they all fit into a single microscope viewing field. Thus, by only adjusting the vertical focus using a Z motorized stage, multiple cell channels can be imaged repeatedly over time. If a full XYZ motorized stage is available, the throughput of the stacked device can be further increased along horizontal dimensions. Making a stacked device is straightforward and this strategy can be useful for improving experiment throughput, especially in a limited microscopy facility.

What do I need?

  • Two identical PDMS chips and a glass coverslide
  • An oxygen plasma cleaner
  • A fluorescent microscope equipped with a programmable motorized Z stage or a full XYZ motorized stage, and a CCD camera

What do I do?

  1. Fabricate your SU-8 mask using standard photolithography. We used a simple ‘Y’ shape design with a 350µm wide main channel.
  2. Make two PDMS replicas of the same design from your SU-8 masks using a standard soft-lithography method.
  3. Bond the two PDMS chips using O2 plasma treatment. The channels should all face down and the main channels in the two chips should be vertically aligned. To avoid overlap of the inlets, we align the two layers so that their inlets are separated by some distance, while the main channels of the two layers are still within the same viewing field in the microscope. We used a 10X objective in our experiment, and this strategy works well with a 350µm channel. However, this method may not work if higher magnification is required, and it may also cause shifted gradients in the two layers. As an alternative strategy, we can align the two layers perfectly along the vertical direction, but punch inlet holes for the two layers with different distance relative to the ‘Y’ junction so the inlets of the two layers can be separated (this will require punching inlet holes for the top layer before plasma bonding). Two masters with different inlet designs can also be used to separate inlets of the stacked device.
  4. Punch the holes for making fluidic inlets and outlets.
  5. Bond the double-layer PDMS chips to a glass coverslide using O2 plasma or air plasma treatment to complete the simple stacked device (Figs. 1A & 1B).
  6. Image the channels and cells using a microscope equipped with a Z motorized stage.
  7. We used food coloring to show the channels in the two layers of the stacked device (Fig. 1B). Furthermore, we show that we can generate chemical gradients in each layer of the stacked device by mixing buffer and FITC-Dextran as well as imaging cells loaded to the main channel of each layer only with a Z motorized stage (Fig. 2).
  8. Finally, using a different design that consists of two gradient channels in each layer and a full XYZ motorized stage, we demonstrate that four individual channels can be imaged in the stacked double-layer device. Again, food coloring is used to show the channels in the upper layer and the bottom layer (Fig. 1C).

Figure 1
Fig. 1. Illustration of the stacked microfluidic device. (A) The schematic drawing of the stacked double-layer device that consists of 2 identical ‘Y’ shape channel; (B) A real picture of the stacked device of the same design. Food coloring is used to show the channels in the 2 layers. (C) A real picture of the stacked device of the design that consists of 2 gradient-generating channels in each layer. Again, food coloring is used to show the channels in the 2 layers.

Figure 2
Fig. 2. Gradient generation and cell images in the double-layer stacked device. (A) Gradient of FITC-Dextran 10kDa generated in the bottom layer channel using the ‘Y’ shape design, and images of Jurkat cells in the same channel. (B) Gradient of FITC-Dextran 10kDa generated in the upper layer channel using the ‘Y’ shape design, and images of Jurkat cells in the same channel. Only a Z motorized stage is used for the imaging.

What else should I know?

Here we demonstrate the simple double-layer stacked device. More layers of PDMS can be stacked to further increase the throughput. In addition, in the current demonstration, the channel in the bottom layer is formed between PDMS and a coverslide while the channel in the top layer is formed between PDMS. If needed, the double-layer chip can be first bonded to another piece of PDMS before bonding to the glass substrate. This way, both layers of channels are in PDMS for consistency.


[1] G. B. Salieb-Beugelaar et al., Latest developments in microfluidic cell biology and analysis systems. Anal. Chem., 2010, 82, 4848-4864.
[2] S. Kim, H. J. Kim, and N. L. Jeon, Biological applications of microfluidic gradient devices. Integr. Biol., 2010, 2, 584-603.
[3] J. Li and F. Lin, Microfluidic devices for studying chemotaxis and electrotaxis. Trends Cell Biol., 2011, 21, 489-497.
[4] W. Saadi et al., A parallel-gradient microfluidic chamber for quantitative analysis of breast cancer cell chemotaxis. Biomed. Microdevices, 2006, 8, 109-118.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)