Micro total analysis systems, commonly known as lab-on-a-chip, are becoming increasingly popular in various areas, particularly at the interface of nanotechnology and biotechnology. These devices have been fabricated from glass or polymers. Among the polymers which have been used, polydimetylsiloxane (PDMS) has emerged as a favorite. Back in 2006, the Li group at Vanderbilt used a PDMS microfluidic device to conduct a fluorous solid phase extraction(FSPE) by packing a microchannel with 5 µm FluoroFlash silica gel. A recent paper (Langmuir, 2007, DOI: 10.1021/la702038t) from J. Hugh Horton’s group at Queen’s University has gone one better. They describe the manufacture and use of a fluorous modified PDMS for the FSPE of a fluorous tagged peptide.
PDMS has a number of interesting properties, depending on how large or small n is. It is considered a viscoelastic so that at high temperatures or at long flow times it will act as a viscous liquid while at low temperatures or short flow times as a rubbery substance. PDMS is available in a multitude of viscosities and is used in everything from cosmetics, contact lenses, caulking agents, and, yes, Silly Putty. The rubbery characteristics of PDMS make Silly Putty, well, silly. The flow characteristics, low temperature curing, and other factors also make PDMS an excellent material for the fabrication of microfluidic devices. One can make a glass mold of the desired device then pour some PDMS along with a curing agent to form microfluidic devices cheaply and reproducibly.
There are, however, some disadvantages. Unmodified PDMS is highly hydrophobic and is difficult to wet. The lack of ionizable surface sites means that microfluidic chips cannot support strong electroosmotic flow (EOF), which is the flow of ions in an solvent when an electric field is applied. The most familiar example of this would be capillary electrophoresis. One of the methods used to overcome this problem is to oxidize the PDMS surface by air plasma oxidation which forms silanols. This works well and results in higher EOF rates. The result, however, is temporary, due to the viscoelastic properties of PDMS. Within 24h a significant decrease in EOF is observed through the migration of unmodified PDMS to the surface. The driving force for this is presumably the lowering of surface free energy by replacing the hydrophilic silanols with the hydrophobic silanes. In order to counteract this phenomena, derivatization of oxidized PDMS with triethoxysilanes has resulted in PDMS devices capable of supporting higher EOF rates over days rather than hours.
In this latest report, Horton used a 1H,1H,2H,2H-perfluorooctyl triethoxysilane as his derivatizing agent resulting in a modified PDMS with higher EOF rates while maintaining those high rates for at least a week. They surmise that the fluorophilic layer has lower surface free energy than the unmodified substrate thereby removing the driving force for diffusion. Interestingly, the EOF rates were nearly as high as the oxidized PDMS, yet they could not detect free silanols using chemical force titrations. Finally, they physisorbed a fluorous tagged peptide, produced through elimination of a phosphoserine residue and addition with a fluorous thiol, onto the fluorous modified PDMS surface, then recovered the peptide with a MeOH wash. While the experiment was only a proof-of-principle for a FSPE on such a surface, the potential to apply it to phosphopeptide enrichment from biological samples can be seen. Overall, a very interesting paper combining physical chemistry, materials, and fluorous separations.