Soldering silicon

I tried once to replace a wire on a solar cell that had broken off. AFIK, solder won't stick to silicon directly, only to the deposited metal traces. These are very easy to damage and then nothing will stick. (humble opinions deferred to those with far more experience) Oppie

Yup, and you have to worry about the quality of the contact. There'll be oxide on top of the silicon, and you have to get through that before you can make an electrical connection.

Even if you do, the connection will be a Schottky barrier unless you've done something special.

Cheers

Phil Hobbs

You could check out US3132419

If you need to flux silicon, it'll take something that etches quartz to do the job; you will have to get the metallization onto the silicon in a vacuum to get it done right. Hydrofluoric acid is the usual flux and cleaner, you do NOT want to do this at home.

Aluminum on HF-etched silicon makes an ohmic contact, and can be welded to (for instance) gold bondwires. Soldering is not usually done directly to thin aluminization, because the solder could dissolve the metallized layer.

Many people regard aluminium as unsolderable but for those in the know there are aluminium solders with special fluxes, I worked for a company that made ultrasonic equipment one of their products was a heated cast-iron solder tub with an ultrasonic transducer mounted through a hole in the side (asbestos gasket!) this item would easily tin small aluminium assemblies without messy flux.

Silicon is a metaloid, having a metallic lustre but being brittle like glass, so other than having metallisation process to create solderable lands I seriously doubt its solderable.

Maybe a hydroflouric acid etch followed by an ultrasonic tinning bath would prove once and for all whether bare silicon can be soldered.

There are a couple of metals that make ohmic contacts to silicon, because the height of the Schottky barrier is actually negative. However, iirc they aren't the common ones like aluminum, tungsten, titanium, copper, or gold. Apart from the fancy stuff, making an ohmic contact to silicon requires either (a) alloying (the old method) or doping the daylights out of the Si under the contact.

Cheers

Phil Hobbs

do

It seems to me that any place where contact was expected to be made would be heavily doped already.

I've never tried to solder silicon, but I have unsoldered it.

I once took a defunct 8751 in the ceramic package, unsoldered the gold-plated quartz window holder, and unsoldered the raw chip from the substrate to make a tie-tac. :-)

The corners of the chip were dangerously sharp.

So somebody, somehow, soldered the silicon chip to the ceramic substrate, albeit I have no idea what kind of prep had been done to do the trick. For all I know, they had gold-plated the ceramic - gold solders real good, if you're quick about it. ;-) (the gold dissolves in the solder.)

Cheers! Rich

Methinks that a gold-aluminum generates what was called the purple plague, not to be confused with spirit duplicators.

..and then there is the cadmium loaded solder that works VERY well; good for sniffing either before or after HCN..

I agree, but there wasn't much context from the OP. It's pretty hard to solder to a chip pad by hand, and reworking a dead chip with a soldering iron is one of those recreational impossibilities that Douglas Adams used to talk about. (RIP).

Cheers

Phil Hobbs

Nice. Die attach is often done by sputtering metal on to the bottom of the wafer and soldering to that, iirc with AgSn solder.

I have a refrigerator magnet with three generations of IBM memory ICs embedded in it: 4 Mb, 16 Mb, and 64 Mb. I worked on process control gizmos for those ones--their code names were Gazelle, Luna, and Oberon. (1 Mb was Antelope, and 256 Mb was Titan. Then the names got boring again.)

Cheers

Phil Hobbs

there

Do you have to make ohmic contacts?

Mark L. Fergerson

Hi,

Here's an interesting solar panel soldering technology:

from that page: "The core of Day4 Energy's technological advantage is the patented Day4 Electrode that is used to inter-connect photovoltaic (PV) cells. This technology represents a fundamental change in the way that both PV cells and PV modules are built.

The Day4 Electrode is comprised of a polymer film embedded with a number of specially coated copper wires. These wires are coated with a proprietary low-temperature melting point alloy and are designed to establish a low-resistance electrical contact with the surface of the PV cell. The Day4 Electrode is a direct replacement of the conventional PV cell soldering process that is currently widely used in the industry."

cheers, Jamie

I remember the Tandberg Hybrid CTV which had an improvised thermal trip on the line scan valve anode cap.

It resembled a safety pin soldered to a terminal post thermally coupled to the anode cap - there was a little packet of low temperature cadmium solder for remaking the thermal cutout.

I would think that the solder must have a higher MP than "plain" AgSn, as i use SMT chips in a HV shunt regulator that i guarantee operational to 204C and there is never a problem. Tho one time for grins, i took a unit to 220C and somewhere along the way from 200C it died and never recovered.

Do you have to make ohmic contacts?

Mark L. Fergerson

Yes, must conduct...

Having read all these great responses I was thinking about a mechanical fixing, something akin to a staple. Perhaps then copper plate the whole thing to keep the pressed interface clean.

Neil

Like Phil Hobbs said, your making a Cat Whisker diode.

conductive expoxy ?

Jamie