Electrical Conductivity of Aluminium Nitride

[will123123]

My company is making use of aluminium nitride plates (of approx. size 6mm x 20mm by 0.6mm thick) for a thermally conductive, but electrically insulating interface for a high voltage device. In controlled conditions these parts work well, but in other environments we have observed varable electrical resistance. Also we have observed that if a person blows gently on to the plate during tests the resistance falls significantly (by at least 80%) and then recovers slowly over a period of around 10 seconds. I assume that humidity is causing these changes, but has anyone got any experience with this phenomenon. Is anyone able to provide information on the electrical resistance of AlN with respect to humidity? Through a web search I have found a limited amount of information that suggests resistance can drop in humid conditions if the material is porous and I am investigating further, but any assistance would be appreciated.

 

[Ceramicguy]

I doubt the bulk material is the problem. Surface resistivity is nearly always lower than the bulk property, sometimes significantly lower especially in high resistivity materials. I classify AlN as a high resistivity dielectric with a bulk resistivity above 10E12 ohm cm. Humidity will usually worsen the situation. The surface is not native AlN but aluminum hydroxide that is probably contaminated with other residues picked up during processing. I would suggest a surface analysis to see what is there. Usual suspects include cleaning reagents and plating bath residues. You can try firing the ceramic in air at 700 C for 1/2 hour to see if there is any improvement. A test of as-received ceramic vs ceramic that has been through your manufacturing process might be enlightening.

Bruce

http://www.accuratus.com

 

[will123123]

Ceramicguy,

Could you advise on suitable surface analysis tools, this is the first time I have dealt with ceramics. I usually work with metals and composites and is those cases, to examine the surface I would usually use SEM and FTIR respectively.


Regards

Carl W

 

[Ceramicguy]

Try to find an SEM with EDX probe. If contamination is present at high levels the FTIR might yield some useful results.

Bruce

Bruce

http://www.accuratus.com

 

[IRstuff]

How high a voltage?

TTFN
faq731-376

 

[IRstuff]

So, according to

http://www.ntktech.com/AlN/ALN for web.pdf,

AlNi is good to 14 kV/mm, which means that your voltage has to be much less than 8.4kV to have no leakage.

There are other high-voltage design techniques that must be followed to prevent such things.

TTFN
faq731-376