On-chip Electrophoresis Devices: Hardware

26 March 2007

• Don't use 5 minute epoxy for attaching wells to quartz or glass substrates. 2-Ton 30 min epoxy (ITW Devcon, Danvers, MA) holds well and is still removable after curing. Cut the narrow end off of a pipette tip, invert it (wide side down), and epoxy over a microchip reservoir for a quick custom fitting as shown on Fig. 2.
  
  

  Figure 2: 200 microlitre pipette tip wells attached to a glass microchip using 30 min epoxy. Epoxy was allowed to set for 24 hours before filling the wells.

  To remove the well and epoxy, use a razor blade held at a shallow angle to the chip surface. 
• Don't rely too much on PDMS for high quality electrophoretic chip separations.  This material notoriously adsorbs and absorbs molecules [9, 10], conducts ionic current (through the substrate), and is highly hydrophobic (favoring bubbles) [11].  In our experience, published surface treatments (e.g., oxygen plasma) wear off rapidly (within minutes if dry; within hours if wet).  PDMS also has very low thermal diffusivity and conductivity (compared to glass), and so it favors Joule heating at relatively low electric fields.  The latter limits achievable resolution and separation speed. 
• Do consider the buffering capacity of your wells.  The table provided for download uses the simple Henderson-Hasselbach equation [12] for a (typical) buffer which uses a very strong acid (or strong base) and which has one dominant pKa value.  The columns depict the desirable tolerance for changes in pH and the applied charge.  The rows depict various buffer molarities, starting from a (typically unacceptable) value of 100 µM to a relatively strong buffer of 100 mM.  
• Do consider mechanically secured (i.e., versus glued) fittings and reservoirs for chip-to-world interfaces when using glass microchips (as shown in Fig. 3).
  

  Figure 3: Chip-to-world interface for glass microchip. Schematic of chip holder, side (a) and top view (b), and example of acrylic holder with nuts and O-rings (c,d)

  Each chip is worth a hundred dollars or so and it may be easier to get the chip thoroughly cleaned if the reservoirs are not permanently attached.  Threaded Upchurch fittings, tapped plastic pipette tips, and red (not the black) silicone O-rings work very well.  Silicone (red) o-rings work well because they don't seem to shed microparticles that can jam up your chip (don't use Buna-N O-rings).
• Do consider Upchurch "Nanoports" to interface a chip with tubing in your Lab-on-a-Chip design (as shown on Fig. 4). They bond permanently to the chip, but are handy and come in different types and sizes depending on the application.  They also serve as reservoirs. A large number of fittings are adaptable to Nanoports [13].  
  
  

  Figure 4: Upchurch Nanoports bonded to glass chip, and interfaced to a capillary with nut and ferrule (black).

           

•           

References

9. Toepke, M.W. and D.J. Beebe, Lab. Chip., 2006, 6(12), 1484-1486.
10. Xia, Y.N. and G.M. Whitesides, Angew. Chem. Int. Ed., 1998, 37(5), 551-575.
11. Whitesides, G.M., Ostuni, E., Takayama, S., Jiang, X., Ingber, D.E., Ann. Rev. Biomed. Eng., 2001. 3, 335-373.
12. Henderson, L.J., Am. J. Physiol., 1908, 21(2), 173-179.
13. http://www.upchurch.com/Products/specsheet.asp?vSpecSheet=742&vPart=N-121H&vFrom=L Accessed 26th March 2007.