James,

Have you considered that you may be oxidizing away the poly?  That would
explain the rise in resistivity with further drive-in steps.

Paul



> Hello all,
>
> I'm presently having some difficulty with doping my polysilicon. My
> substrate structure is as follows (from bottom to top):
>
> 300 microns thick p-type Silicon
> 400 nm low Stress LPCVD SiN
> 150 nm undoped polysilicon.
>
> Now I would like to reduce the sheet resistance of the polysilicon. To
> do this I've used Emulsitone's Borosilicafilm spin on dopant called
> "Borosilicafilm I" -data sheet available on Emulsitone website.
>
> I had initial problems to do with removing the Borosilicafilm after the
> diffusion (all to do with my lack of oxygen ambient).
>
> I used a Jandel 4 point probe to try and measure the sheet resistance of
> my undoped polysilicon. It was essentially immeasurable implying a sheet
> resistivity of greater than 10 mega ohms /square! I was looking to get
> the sheet resistance value to between 50-250 ohm persquare. I then would
> like to dope the contact regions (I am fabricating polysilicon
> resistors) to around 10 ohm per square.
>
> I used the following protocol to dope my polysilicon:
>
> 1. Standard clean
> 2. Spin Borosilicafilm dopant
> 3. Bake 20 minutes 200oC (essentially to remove the solvent from the
> dopant film)
> 4. Diffused for 15/30 minutes at 1050oC (this is the maximum temperature
> of my furnace) in 95%/5% N2/O2
> 5. Borosilicafilm is etched off either using CHF3 dry etch or 49% HF
> (90% of my tests I use wet etching rather than dry etching)
> 6. Measure sheet resistance.
>
> After diffusion for 15 minutes the sheet resistance of my polysilicon
> reduces to around 200 ohms pr square while diffusion for 30 minutes
> results in a sheet resistance of 150 ohms per square. So far so good.
>
> However when I doped the polysilicon film for a second time (recall I'd
> like a sheet resistance of 10 ohms persquare in the contact regions),
> following the exact same protocol as above, I found the sheet resistance
> INCREASES to around 270 ohms per square after 15 minutes and 350 ohms
> per square after 30 minutes.
>
> My simple question is this: Can anyone shed on light on this phenomenon?