Determination of BCl3 and PCl3 in chlorosilane samples by GC

[onepiece]

Hi,

Will it be possible to determine BCl3 and PCl3 (ppm to ppb) in pure chlorosilane (Trichlorosilane/Silicon Tetrachloride) or chlorosilane mixture by Gas Chromatography?

Sample may be either gaseous or liquid form.

Kindly suggest the following:
1. Sample introduction technique
2. Sample pre-treatment
3. Chromatographic conditions (injector, oven, detector, etc.)
4. Carrier Gas and make up gas (flow rates...)
5. Column (length, diameter, MOC, packing material, etc.)
6. Detector (type, gases, etc.)
7. Vents/Exhaust gases
8. Calibration Standards (proposed vendor/supplier)
9. Standards Dilution
10. Challenges

If it is not feasible with GC, then what method or analytical technique do you suggest? Please share to me your ideas.

Hope someone out there can help me on this...

Any comments are highly appreciated.

Thanks in advance.

[aidnai]

Boron/Phos contamination in chlorosilanes is typically done with some wet chemistry followed by ICP/MS or ICP/AES. It is not easy and the instrumentation is fairly expensive (around $500k). Another method is epitaxial deposition followed by resisitivity. Some even go to the extent of doing a mini CVD to poly silicon, then melting and pulling a single crystal, then wafering and measuring resistivity on that.

I have looked into doing this with IR, either standard FTIR (full spectrum) or tunable diode laser type (limited spectrum) but we didn't have the money to proceed (and when we were looking several years ago the technology wasn't readily available or mature).

Good luck,
Aidan

[onepiece]

Thanks aidan,

Actually, I had found one literature. (Please see the link below)
http://pubs.acs.org/doi/abs/10.1021/ac00250a063

But, the thing is, it was published long time ago. I am also not sure, how it is validated.
I would like to check if someone using the same or almost the same methodology.
Or is there a SEMI method available for detecting BCl3/PCl3 by Gas chromatography.

What driving me to go by GC instead of ICP-MS, because aside of being much cheaper and it is faster and much less contamination. Especially if samples are introduced to GC by direct injection using a switching valve.

Anyway, thank you again. I am glad that you share your insights.
It seems that you have great experience with polysilicon industry.

To be honest with you, this is something new for me.

Regards,
Onepiece

[aidnai]

That's an interesting looking article. I haven't heard of that technique being used. Unfortunately the pay-wall keeps me from seeing a lot of the interesting details (I am curious about POCl3 vs PCl3 -- usually the phosphorus impurities are referred to as PCl3). Of course, Boron is not measured with this method, so I'm not sure how helpful it is...

Doing syringe injections of TCS/STC is problematic. Chlorosilanes break down in moist air and produce HCl and silicon oxides. Quite a mess. I've done it using argon shield purging (installed a bag around the injection port, filled it with argon and used it like a glove box to inject from a headspace vial). But the syringe gets destroyed quickly by the silica deposits. In the article they mention using 'mixed etchant' (HF, HNO3, and CH3OOH) which is a standard mix for etching silicon -- but the HF is also bad for the syringe, and certainly unpleasant to work with. These issues could be avoided with a gas sampling valve, but that presents problems as well. If at any time moisture or oxygen are present in the GSV, it may become clogged, the volume injected may change, or it may become mechanically damaged due to the formation of solids internally.

Another challenge you will certainly need to address -- the chlorosilane waste and vapor need to be sent through a water scrubber.

Again, good luck. If you make headway, I'd love to hear about it.