Evaporation of gold circuit

[Erikve]

Dear all,

I have a quite easy plan: W want to evaporate a gold circuit on a Brass plate. Between this copper and gold I need to make an isolating layer to prevent short-currents. i thought to make this layer of Silicon-oxide. Is the possible to evaporate and what source can I use this for best?

PS. I know that between the gold and SiO2 should be a small layer of Titanium.

 

[MATLABdude]

This is not a trivial matter, unfortunately. The whole reason you need an adhesion layer is to prevent lattice and/or temperature coefficient mismatch from causing your films to lift-off of whatever you're trying to deposit it onto. Will amorphous SiO2 stick to bronze? How about if you used a chrome or titanium adhesion layer? For thin layers, the following discussion may help:
http://www.finishing.com/368/90.shtml

How thick of gold are we talking about here? Could you use a flex PCB instead?

EDIT: You can evaporate SiO2 (pretty much stochiometricaly) or sputter, or CVD it. With varying levels of built-in stress.

 

[Erikve]

Okay, we are talking about brass (copper+small amount of zinc), I cannot replace this component. In principle the isolating layer could be everything, if you have better idea, please tell me!
Titanium I normaly use as adhesion layer of the gold, so maybe this can be used indeed also on the copper.

 

[Mapes]

Dear all,

I have a quite easy plan: W want to evaporate a gold circuit on a Brass plate. Between this copper and gold I need to make an isolating layer to prevent short-currents. i thought to make this layer of Silicon-oxide. Is the possible to evaporate and what source can I use this for best?

PS. I know that between the gold and SiO2 should be a small layer of Titanium.

I've personally found that evaporated silicon dioxide (really SiO_2-x; it's not at all stoichiometric) is pinhole-filled and a poor insulator (base evaporation pressure of 10^-7 torr). PECVD oxide is better, but for you it depends on the brass having micron-level smoothness and no defects, or the oxide won't coat it suitably. How about an low-outgassing resin or other polymer coating that might be sprayed on to the brass and then evaporated upon?

The whole reason you need an adhesion layer is to prevent lattice and/or temperature coefficient mismatch from causing your films to lift-off of whatever you're trying to deposit it onto.

An adhesion layer is needed because gold doesn't form a chemical bond with copper oxide or zinc oxide.

 

[MATLABdude]

I've personally found that evaporated silicon dioxide (really SiO_2-x; it's not at all stoichiometric) is pinhole-filled and a poor insulator (base evaporation pressure of 10^-7 torr). PECVD oxide is better, but for you it depends on the brass having micron-level smoothness and no defects, or the oxide won't coat it suitably. How about an low-outgassing resin or other polymer coating that might be sprayed on to the brass and then evaporated upon?

An adhesion layer is needed because gold doesn't form a chemical bond with copper oxide or zinc oxide.

I think it's possible to get near stoichiometric with enough process tuning / e-beam evaporation; no idea of how porous / non-insulating the resulting film would be. To the OP: if someone else at your university / microfab has this figured out, great! If not, you probably want to go with something else. Again, a flex PCB may be ideal, assuming it suits your purposes. If you can put tapped holes in your bronze thing, that'd be even better.

I suppose that (good) bonding with lattice / thermal coefficient mismatch would cause warping / fracturing of the materials in question, rather than having a deposited film pop off.

 

[f95toli]

I think it's possible to get near stoichiometric with enough process tuning / e-beam evaporation; no idea of how porous / non-insulating the resulting film would be. To the OP: if someone else at your university / microfab has this figured out, great! .
I suppose that (good) bonding with lattice / thermal coefficient mismatch would cause warping / fracturing of the materials in question, rather than having a deposited film pop off.

E-beam evaporated SiO2 is essentially an amorphous mess with lots of defects and pin-holes (and SiO2 regardless of evaporation method is never a good dielectric, there are lots of two-level fluctuators etc).

Whether or not SiO2 (or anything else) would be good enough really depends on the requirements. What kind of resistance do you need?

Also, are there any requirements when it comes to dielectric constant/loss tangent?

 

[Erikve]

This gold circuit will serve as a calibration coil on a capacitive moving plate. So in principle I want as less as possible loss between coil and plate, the same for dielectric loss, since this influence the capacitive value. Best would be if I have totally isolated surface between copper plate and golden coil.
is there not something more easy chemical coating where I do not have problems with pin holes in the surface?