# Bonding resistance wire

1. Jan 20, 2010

Did anyone here ever successfully wire bond Chromel, Manganin, Konstantan or some other such material with low ductility? What is the key to success here? Pad temperature? Pressure? Power? Is gold still the best pad material? I heard also the needle makes a difference?

2. Jan 20, 2010

### f95toli

"Bond" as in using e.g. Manganin as a bond wire?
I doubt that is possible, the only materials that are "easy" to bond are -as far as I know- gold and Al-Si

I have colllegues that sometimes "glue" other types of wire using silver paste, but that is quite messy,

3. Jan 20, 2010

A friend told me that his wire bonder manual listing different wire materials, had the remark "hard to do" written next to Konstantan.

4. Jan 20, 2010

### f95toli

Are you sure he didn't mean materials you can bond TO?
Bonding is a bit tricky in that only certain combinations of material work well together, it is e.g. easy to bond to a gold pad using Al wire; but it is impossible to bond TO an aluminium pad using Au wire (due to the fact that the Au can't break the oxide).

Also, the shape of the wedge/ball makes a huge difference which is why there are special wedges/balls for Al and Au (you can't bond Al using an Au wedge). Hence, not only would you have to find the right wire you would also need the right type of wedge...

5. Jan 20, 2010

### Mike_In_Plano

I was in the instrumentation business for years and we had to bond high nickel wires on a regular basis. One way is to epoxy your copper wire nearby such that your stripped end of the copper is held in place. Wrap your high nickle wire about the end of the copper wire, and melt the two together using a carbon probe.

The probe is nothing speacial, in fact it can be quite large. However, the power source has to right. You want a current limited power supply with a fairly high open circuit voltage ( maybe 60 volts? ).

I don't recall the rest, that was out of my area, but I suspect you could work it out with a lab supply and few hours.

6. Jan 21, 2010

### davidrit

K&S has some information on studies they did for copper bonding. I don't think they cover the alloys you mention but the process considerations would be similar:

http://www.kns.com/KNS07/Templates/showpage.asp?DBID=1&LNGID=1&TMID=111&FID=549&PID=976 [Broken]

You normally want a uniform intermetallic formation between the wire and pad for good bond strength. I'm guessing that may be more difficult with alloys.

I think you said it backwards. Thermosonic gold wire bonded to aluminum pads is common. I think the heat and ultrasonic energy is sufficient to break the oxide.

Last edited by a moderator: May 4, 2017
7. Jan 22, 2010

### f95toli

No, but I do think we are referring to somewhat different things. I don't know of anyone who actually uses heat while bonding "research" samples; simply because if it hot enough to modify the surface chemistry it might also change the properties of the sample (remember that we are talking about an uncontrolled atmosphere here, at least for normal manual bonders). The heater on the sample stage is generally only used to keep the sample a few degrees about RT to avoid having moisture condensing on it.

Anyway, the main point here is that I was talking about bonding using ONLY the ultrasonic (since I sort of assumed that is what Deadbeef was referring to), in which case bonding gold to Al is if not impossible so at least extremely difficult, it is at least not possible on the manual K&S bonders I use most of the time.

It is interesting to not that Al wire works on almost everything (I use it on Al, Nb, Au, Sn etc) except copper. I am not sure why, but whenever I have to bond to copper I have to use Au wire instead (which of course doesn't work on Al, Nb, Sn....).

8. Jan 25, 2010

The wires as well as the substrate have thicknesses of about $$30\mu m$$ Soldering causes problems because of the surface tension which produces .3mm thick balls even on the .1mm soldering iron tips. Fixating the wires is hard on those dimensions too. Normal spot welders don't work with wires this thin and would also break the substrate.