[Oseania]

I stumbled upon a method where metallic coatings can be transferred to PDMS substrates. The idea is that the metal (Au/Ag) is deposited on glass surface which is then soaked in 3-MPTMS ethanol solution for 3 hours. A separate PDMS sheet is treated with corona and the glass and the PDMS sheet are then pushed against each other. According to the authors the metal surface is transferred to PDMS and the adhesion should be really good.

I was thinking about to test this method but I was hoping if someone could explain what actually happens here. How 3-MPTMS affects the glass so that the metal can be transferred? How the good adhesion between the PDMS and metal comes about?

More information about the method: http://www.rsc.org/binaries/loc/2003/volume1/155-307.pdf

[Arkcon]

Polydimethylsiloxanes are a common reagent used in soft-lithography:  https://en.wikipedia.org/wiki/Polydimethylsiloxane#Soft_lithography  So, just stamping the traces onto a glass substrate is a well known process.  Once you've made a photoresist, by the traditional stamp or your more innovative process, the procedure continues much like lithography. http://www.nanotech-now.com/columns/?article=332

[Oseania]

Thank you Arkcon for your reply.

I started wondering if the above mentioned method (3-MPTMS) could work if the electrodes are deposited on Si wafer instead of glass?

[Enthalpy]

The linked paper uses Au for one electrode, and I fear that Au will stick too much on Si because it makes a known eutectic. Au botches the electric properties of Si (rather mobile dopant in Si, introduces deep levels with short carrier lifetime) so the wafer will be made unusable. If processing at any warm temperature, one normally deposits W or Pt beween Si and Au to avoid it; expect this barrier layer to be transferred together with gold to the polysiloxane. But trying is easy.

If you like the planeity of Si and other good properties, why not grow or deposit an SiO₂ layer on Si? Chances are that the adhesion of metal to silica is similar to glass. (Not completely sure, because for instance at metal-to-glass bonding, ion conduction in glass is paramount).

Besides silicon, you can also purchase silica and sapphire wafers which are meant for semiconductor processes: size, planeity, purity and so on.

----------

What happens there... The authors must explain it in their previous paper, where they describe the 3-MPTMS process. My poorly substantiated bet is that it creates a single layer of a silicon thing on the metal surface, maybe trimethoxysilane with Si-metal bonds, and this monolayer reacts properly with the siloxane to achieve a decent adhesion. 3-MPTMS in solution has the advantage of reaching all points of the metal surface for as long as needed until the first metal-to-siliconthing succeeds; the second step, siliconthing-to-siliconthingy adhesion or reaction is the easier part.

The activation and oxidation of the polysiloxane by corona must first clean the polysiloxane, create pending bonds, and also increase the proportion of oxygen atoms in the polysiloxane, which would (maybe perhaps) react with the monolayer's methyls to achieve true chemical bonds.

Whatever happen in details, replacing at the polysiloxane the highly antiadhesive methyl surface by a highly adhesive silica-like surface looks logical.

Unclear is: how good an insulator the corona-treated surface of the polysiloxane is, since it's so brutally modified. In case the sensor needs a better insulation than obtained, it could be worth trying to:
- Remove the surperficial, degraded polysiloxane, say by a plasma etch, leaving the metal as much as possible
- Reduce again the polysiloxane to methylated compounds. How, I don't know.

----------

To my understanding, 3-MPTMS acts on the top of the metal to let it adhere to the strongly treated polysiloxane. 3-MPTMS doesn't act essentially on the glass, so whether this works with silicon seems unrelated with 3-MPTMS.