Improving Heating Uniformity on AZO Conductive Glass: What Went Wrong?

[paulzhen]

Hi!

I tried to heat the AZO conductive glass with a current source = 0.5A.

However, I found the voltage near the contact points was almost 3 times higher than that on the glass surface (please see the picture for details). It means the heating effect mainly happened near the contact points, very slow heating at the middle of the glass.

I check and did my best to keep good contacts between all components, and I have tried several different ways to make contacts including stick the exposed conducting wire directly onto the cooper tape or the surface of the glass, but the results are pretty much the same (heating still mainly happened near the contacts)!

Please someone explain to me what went wrong, and what would you recommend if I want to achieve an uniform and fast heating effect on AZO glass.

Thanks so much!

 

[Dr. Courtney]

Image not available. Post is not clear.

 

[paulzhen]

Image not available. Post is not clear.

edited, pic can be seen now?

 

[Hyo X]

It means your contact is not good. Specifically not an Ohmic contact, which is not surprising contacting a metal to a glass.

Try these to improve your contact quality:
You could try evaporative deposition with a shadow mask,
or you could try using colloidal conductor, like colloidal silver.
You could try liquid metal like galinstan available online.

General tips: shrink all the dimensions of your system. Make the glass smaller. If you are trying to heat up the HPC on the glass, cover all of the glass with HPC uniformly.

Do a control where you use an external heater to heat up the glass and see if it has the effect you want to see without "Joule heating".

 

[paulzhen]

It means your contact is not good. Specifically not an Ohmic contact, which is not surprising contacting a metal to a glass.

Try these to improve your contact quality:
You could try evaporative deposition with a shadow mask,
or you could try using colloidal conductor, like colloidal silver.
You could try liquid metal like galinstan available online.

General tips: shrink all the dimensions of your system. Make the glass smaller. If you are trying to heat up the HPC on the glass, cover all of the glass with HPC uniformly.

Do a control where you use an external heater to heat up the glass and see if it has the effect you want to see without "Joule heating".

Thanks so much!

This is what I have done today:

I sputtered aluminum onto the AZO glass, however...the results are similar, the heat was mainly happened near the contact points, and the voltage distribution is also similar as before... I will try liquid metal later, but why the situation did not show any noticeable improvement?

Please help!

 

[Jeff Rosenbury]

Why is the aluminum black? Did you use an inert gas for the sputtering?

Perhaps you could (in an inert gas) etch the glass before sputtering? There may be a boundary layer problem.

 

[paulzhen]

Why is the aluminum black? Did you use an inert gas for the sputtering?

Perhaps you could (in an inert gas) etch the glass before sputtering? There may be a boundary layer problem.

The black spots are formed because of that, I tape a aluminium foil as a shield for the center area when doing sputtering, and the edge of the foil did not perfectly taped.

 

[Hyo X]

The problem with your new resistor geometry is that current is always going to take the path of least resistance, AKA the shortest path. In this case, the current will only travel through the small gaps between the aluminum pads. Your other geometry was better, with identical linear contact pads on opposite sides.

How thick is your alumnium? aim for at least 200 nm thickness.

Can you measure the total current with an Ammeter? total amount of heating is going to be proportional to current flow regardless of where the voltage is dropped. If you don't have a short then there will be heating somewhere.

 

[paulzhen]

The problem with your new resistor geometry is that current is always going to take the path of least resistance, AKA the shortest path. In this case, the current will only travel through the small gaps between the aluminum pads. Your other geometry was better, with identical linear contact pads on opposite sides.

How thick is your alumnium? aim for at least 200 nm thickness.

Can you measure the total current with an Ammeter? total amount of heating is going to be proportional to current flow regardless of where the voltage is dropped. If you don't have a short then there will be heating somewhere.

I tried silver paste today, similar results. I doubt that if this is the right way to heat up a conductive thin film, it is really frustrated. I mean it should be a straight forward experiment, right?
Do you know any workable alternative to heat up the film electrically? Or maybe heating the HPC material directly? Please advise! A million thanks!

 

[jim hardy]

Just how uniform is the thickness of your conductive film?

I have never "sputtered" , only spray painted. (though at my meager skill level you might call it spattering )
Does that sputtering process assure same thickness everywhere?
If i do not sweep my spraygun across the work i get a thick application where the nozzle points, thin at edges.

That mistake would make a coating that heats only around the edges.

just a thought...

Try placing one voltmeter lead at the center of your glass and observe voltage gradient across the whole workpiece.

 

[paulzhen]

Just how uniform is the thickness of your conductive film?

I have never "sputtered" , only spray painted. (though at my meager skill level you might call it spattering )
Does that sputtering process assure same thickness everywhere?
If i do not sweep my spraygun across the work i get a thick application where the nozzle points, thin at edges.

That mistake would make a coating that heats only around the edges.

just a thought...

Try placing one voltmeter lead at the center of your glass and observe voltage gradient across the whole workpiece.

Yes the sputtering process should produce a uniform aluminium layer (for my case, the thickness was around 150nm).

I still believe this should be a straight forward entry-level experiment, but I have worked on this for a week, tried a lot of different ways, but none of them came out as I expected...

 

[Jeff Rosenbury]

Another thought. If your goal is even heating, you could use peltier coolers to offset the heat near the connection giving even heating. It's not elegant, but...

 

[paulzhen]

Another thought. If your goal is even heating, you could use peltier coolers to offset the heat near the connection giving even heating. It's not elegant, but...

But anyone can tell what exactly is the problem with the over heating near the connections?

 

[jim hardy]

When one cannot say what his problem is
he must resort to eliminating what it is not.

Seems to me heating would be in proportion to both current density and resistivity.

Would a map of simple ohmmeter readings on the workpiece prove that resistivity is uniform? A simple jig to hold two probes at constant distance would seem a natural.first step.. toward producing that map

Would a map of voltage gradient readings localize areas of high current density? That was your first troubleshooting step, can you improve on it ?
Why would current density be higher near the contacts?
The same simple test prod jig would enable readings.of volts-per-unit-length

You'll have to keep poking at it until it gives up a secret or two. We can't do that from way out here.

 

[jim hardy]

If you make a test prod holder with 1cm spacing you can measure voltage gradient in millivolts per cm.
By rotating the holder you can infer a direction and draw little arrows to scale.
A result like this says you have good contact along your aluminum strips -

gradient starts perpendicular to alumiinum strip, equipotential lines would be perpendicular to gradients parallel to aluminum strips.

a result like this says you have contact only where the alligator clips press your aluminum down against the glass.

gradient radiates outward from point of contact, equipotential lines would be quarter circles segments centered on point of contact.

just a thought experiment.
You are obviously far more advanced than i.

 

[Jeff Rosenbury]

Jim, I was assuming the metal to glass contact was more resistive than the glass overall. This would cause the I²R losses to concentrate there.

An oxide layer or some such could cause this. I think ZnO is a semiconductor. Perhaps there's some sort of diode depletion region causing an increased resistance.

Also it's remotely possible the current's movement into the depth of the glass might add some. (Movement in the third dimension could show up as extra heating in the area. It sounds a little suspect to me, but something is causing the heating.)

 

[davenn]

post deleted

 

[jim hardy]

I could be way off.
I never heard of AZO before, a quick search led me to think it's a conductive film applied to glass substrate.
http://materion.com/ResourceCenter/...s/Oxides/AZOTransparentConductiveCoating.aspx

Description and Applications
Films of Aluminum-doped Zinc Oxide, AZO, are transparent and electrically conductive. ...

The full range of sheet resistance, from < 50 Ω/sq to M Ω/sq, can be obtained with AZO by varying deposition thickness and parameters. ...

Optical and Electrical Properties
AZO is deposited by sputtering from targets composed of 2-4% Al metal incorporated in ZnO. Electrical conductance, measured as bulk resistivity or as sheet resistance, is related to deposition properties and thickness. Figure 1 presents the sheet resistances achieved for several DC magnetron depositions made with the Materion 2%/98% target on ambient-temperature glass substrates.

So i was suggestng simple checks first for uniform film conductivity then for successful contact between aluminum strips and the film.

If it's the actual glass that's conducting, well, i missed the boat (again) .

Perhaps there's some sort of diode depletion region causing an increased resistance.

Hmmmm maybe he's made a giant FET that's pinching off ?

 

[Jeff Rosenbury]

I could be way off.
I never heard of AZO before, a quick search led me to think it's a conductive film applied to glass substrate.

Good job. My search just showed some crystals of the stuff. I was assuming it was all doped ZnO, but you are likely correct. A thin film would be more reasonable.

I doubt you missed the boat. They rarely leave without the captain.