Ask Me Help Desk - Need expert on nitrogen gas purification

I would very much like to speak with an expert on the purification of nitrogen gas for use in drying semiconductor material. I believe that my medical grade and Ultra High Purity nitrogen gas is contaminated and need to know if it can be purified further.
Thanks,
Bill

It's only "contaminated" if the impurities bother you. The UHP stuff, I think comes with an assay of impurities. I do believe there is a semiconductor grade available as well.

Quote:

Originally Posted by KeepItSimpleStupid

It's only "contaminated" if the impurities bother you. The UHP stuff, I think comes with an assay of impurities. I do believe there is a semiconductor grade available as well.

It bothers one of our customers so by default... Whatever the contaminant is it sticks to the wafer surface and cannot be removed except by repolishing. Observing the output of the nitrogen gun on a wafer under the microscope you can visibly see "stuff" coming out of the gun. Our gas supplier seems concerned but this is the best they can offer.

Bill

You need to use a better gun. One with an integral filter. This is a bit more complex: Cleanroom Ion Air Guns, Ionizing Blow-Off Guns, Ionizer Air Flow Controllers and Gun Filter Replacements - Inn Innotech Products

With a 0.02 micron filter and neutralization of static charges.

Quote:

Originally Posted by KeepItSimpleStupid

You need to use a better gun. One with an integral filter. This is a bit more complex: Cleanroom Ion Air Guns, Ionizing Blow-Off Guns, Ionizer Air Flow Controllers and Gun Filter Replacements - Inn Innotech Products

with a 0.02 micron filter and neutralization of static charges.

I do appreciate your responses. We use that same Ionizing gun you linked to with a .1 mircon filter tip. Upstream there is a .01 micron inline filter before the regulator. Everything except the Ionizing gun was supplied new from our gas supplier. We cannot understand why anything is getting past the filters. I noticed no improvement when the Ionizing gun was installed. The contaminant levels vary a great deal from bottle to bottle and visually look identical whether we are using standard grade, medical grade or UHP grade.

We've been chasing this problem for a year with the biggest improvement coming from the switch to medical grade/UHP. Removing filters, installing filters does not seem to make a noticeable difference.

Our supplier has never mentioned semiconductor grade gas before so we are making an inquiry today.

Bill

I think you have things way backwards. The air flows from the tank to the regulator (if attached directly), then some supply tubing (What's the material?), Possibly a large filter, then gun, then the smallest filter.

I had to find it. This was the gun we used. http://www.mykrolis.com/publications...-2594-1006.pdf

Note, it filters to 0.003 um

Quote:

Originally Posted by KeepItSimpleStupid

I think you have things way backwards. The air flows from the tank to the regulator (if attached directly), then some supply tubing (What's the material?), Possibly a large filter, then gun, then the smallest filter.

I had to find it. This was the gun we used. http://www.mykrolis.com/publications...-2594-1006.pdf

Note, it filters to 0.003 um

Agreed about being backwards. Initially we did not have the ionizing gun. The system went>>tank>>regulator>>.01um filter then a straight gun. Later we replaced the straight gun with the ionizing gun and it came with .1um filters.

We have ordered the .003um gun you mentioned and are working on obtaining semiconductor grade nitrogen and will start fresh. Do they make semiconductor grade regulators? We thought that the problem may have been our new medical grade regulator but with the bottle to bottle variability of contaminants we assumed that the regulator was OK.

Thanks,
Bill

What kind of line did you use? How long?

Moisture?

A High Purity regulator would have been a better choice, I think.
Gas Handling Equipment: Gas Delivery Pressure Regulators

The 9460 is designed for semiconductor applications. VCR fittings etc. Cross purge assemblies are used with vacuum pumps and purge lines in semiconductor systems, so moisture and contaminants cannot enter the lines. The vacuum pump is there to remove gasses that spontaneously ignite with air such as Silane and in inert gas is for purging. Tied diaphrams are required when a vacuum system is connected to a regulator.

Keep me posted as to the outcome.

Quote:

Originally Posted by KeepItSimpleStupid

What kind of line did you use? How long?

Moisture?

Good morning. From the bottle to regulator the line is stainless braided PTFE. Between regulators it is stainless/brass and after the regulator the line is polypropylene. There is a urethane line T'eed off between the regulators to supply a toluene airbrush. The .001 filter is in a stainless housing. From the regulator to airbrush there is about 10' of line. Ditto on the ionizing airgun. The PTFE line is about four feet.

Moisture... we have nothing in place to deal with that as we did not suspect it as a problem. Our gas was not being supplied with an assay certificate but once we requested them it turns out that the dirtiest bottles had much higher moisture levels than the cleaner bottles.

Can you suggest line materials for the system or even a layout for the whole system? I would guess that replacing the whole system will be necessary if we switch to semiconductor grade gas.

Thanks,
Bill

Hi again,
I was reading more about the support option mentioned in your replies and I click on your UserName then the support link as suggested but got the following:

"Return to MerchantPayments By PayPal

We cannot process this transaction because there is a problem with the PayPal email address supplied by the seller. Please contact the seller to resolve the problem. If this payment is for an eBay listing, you can contact the seller via the "Ask Seller a Question" link on the listing page. When you have the correct email address, payment can be made at www.paypal.com.

PayPal protects your privacy and security.

For more information, read our User Agreement and Privacy Policy."

I will certainly suggest the support option to management for your assistance with our problem but the link will need to be working.\

Cheers,
Bill

I tried the Support option and it worked fine for me. I'd suggest trying again and also trying it at home just to see if for some reason there was a glitch or paypal was down.

First order comments:

Can you give me a link to the Stainless Braided PTFE" line that you used?

Can you send me a picture of the setup if it doesn't violate confidentiality? I doubt you would want to post it publicly. You can send it to the address in the Support link once we figure out why it doesn't work.

Do you have a pic of the contaminants under the microscope?

What solvents are you drying off? The solvents themselves can leave residues behind. The best rinse I know of is a dip in Methanol, trichlorletane followed by a dip in liquid Freon and dry.

How are you cleaning your substrates?

What solvents are present just before the air brush dry?

What is the regulator delivery pressure?

Do you have any evidence that it's related to substrate cleanliness?

I really doubt that you will have to change to a Semiconductor gas, but I have these first order recommendations:

1. The toluene air brush should have a separate system.
2. The gun should have a separate system.
3. The air gun system should have lines as short as possible
4. The gun should be "cleaned" before use. e.g. Burst it away from the device before using it on the substrate or make sure you aim it away from the substrate before using it to dry the substrate.

If you prefer to continue via email because of confidentiality that's fine.

Quote:

Originally Posted by KeepItSimpleStupid

I tried the Support option and it worked fine for me. I'd suggest trying again and also trying it at home just to see if for some reason there was a glitch or paypal was down.

First time I clicked on your name>>view public profile>> support and it falied (still does). If i just click your name then support it works.

First order comments:

Can you give me a link to the the Stainless Braided PTFE" line that you used?

I don't have a link. It was supplied by our gas supplier and is made by Superior Products in Cleveland. I cannot find a part# but it is 1/4" ID PTFE core.

Can you send me a picture of the setup if it doesn't violate confidentiality? I doubt you would want to post it publicly. You can send it to the address in the Support link once we figure out why it doesn't work.

Pictures would have to be discussed with management. The system is bottle>>regulator dropping from bottle pressure to 150psi>>.001 filter>>T w/urethane line going to air brush, other side of T goes to second regulator we use to vary pressure from 120to 20psi>>polypropylene line going to Innotech air gun you linked to.

Do you have a pic of the contaminants under the microscope?

Yes but I think a description would be better as the detail in the images is washed out due to brightness variations. The primary contaminant is silver metallic in appearance. I once used a clean scalpel while observing under the microscope to try and remove some of the material. I was suprised by how firmly it was stuck and also that it appeared top be fibrous in nature. It almost looked like a dried up spitball. Some of this debris comes out at a high enough velocity to actually cause microfractures at the site of impact. If it encounters the wafer at a low angle of incidence it will leave scratches with minimal amounts of deposited metallic looking debris. The other contaminant we see looks like dried droplets with a color to the that ranges between gun metal blue and maroon. These droplets are water soluble and come off with a rinse in DI water. The metallic debris will not come off without re-polishing. If you still want the pictures of debris let me know and I'll post them.

What solvents are you drying off? The solvents themselves can leave residues behind. The best rinse I know of is a dip in Methanol, trichlorletane followed by a dip in liquid Freon and dry.

Primarily toluene. Lately because of the droplets I've mentioned we are doing the final clean with DI water rinse, 2-Propanol, toluene then nitrogen dry. All of our solvents are CMOS grade. Our wafers are Indium Antimonide (InSb) and Gallium Antimonide (GaSb)

How are you cleaning your substrates?

See above.

What solvents are present just before the air brush dry?

Toluene.

What is the regulator delivery pressure?

After the polish process we rinse with DI water then dry with nitrogen @ 100-120psi. During final cleaning the delivery pressure depends on wafer diameter but no more than 40psi.

Do you have any evidence that it's related to substrate cleanliness?

We polish the wafers on a SS head and inspect prior to demounting. During this inspection the majority of wafers are pristine. We re-polish if a defect is found. Once they come out of the demount vessel they are air brushed with toluene the dried with nitrogen at pressures of 40psi or less. Then flatness is verified, a final microscope inspection then final cleaning described above. The more often we use the nitrogen on the wafers the dirtier they become.

I really doubt that you will have to change to a Semiconductor gas, but I have these first order recommendations:

1. The toluene air brush should have a separate system.
2. The gun should have a separate system.
3. The air gun system should have lines as short as possible
4. The gun should be "cleaned" before use. e.g. Burst it away from the device before using it on the substrate or make sure you aim it away from the substrate before using it to dry the substrate.

We always give a quick blast away from wafers before drying. It's part of the training. We also start high above the wafer then quickly bring the gun down to the wafer for drying.

Bill

All that looks reasonably good. I think you may have two problems:

1. Particulates which I think the new gun will remove and keeping the delivery pressure low.
2. A liquid contaminant. Possibly water. Could be from the source or could be from the environment (high humidity).

Comments:
Do you notice any variation of the spots with season. e.g. better in winter?

Would you happen to know the size of the particulates?

Are you using Grade 6 or better Dry Nitrogen?

By Liquid Freon, I mean Freon TF.

Toulene is not compatible with HDPE which your not using. Not sure about PolyPropolene.

Chemical/Material Compatibility Chart for container materials of construction

While you at it, why don't you contact this company and discuss their moisture filters for your application. It might not be suitable because of volume which may not be significant.

Gas Chromatography Super Clean™ Gas Purifiers

Here is another link and they have a bit more info suggesting that it would be applicable for your application:

http://www.vici.com/matsen/purifiers.pdf

GC use is probably one of the most sensitive aparatuses for moisure.

I wonder if you can compute the dew point for 2 cases by getting the following temperatures and RH values:

RH: Room
Temp: Wafer

Temp: Air stream (case #1)
Temp: Ambient (case #2)

What I'm getting at is if the moisture is from the tank, then use a moisture eliminator that's used for gas chromatagraphs if they will work. If it's due to humidity in the air, dehumidify the air.

Ppp

Quote:

Originally Posted by KeepItSimpleStupid

All that looks reasonably good. I think you may have two problems:

1. Particulates which I think the new gun will remove and keeping the delivery pressure low.
2. A liquid contaminant. Possibly water. Could be from the source or could be from the environment (high humidity).

We are thinking about installing a .001 filter after the second regulator but our concern was that it might drop the pressure too much. We need the higher pressure to help remove the water after the polishing process as surface tension with the water is high. If the contaminant was liquid why does it appear fibrous when scraping it loose with a scalpel? There is a liquid component to the contamination that appears to clear up with water rinse. I believe we have two nitrogen related contaminants.....one that is fibrous (metallic looking particles) and the liquid variety which causes blue/maroon colored droplets.

Comments:
Do you notice any variation of the spots with season. e.g. better in winter?

No. I don't see a difference in contaminant levels from season to season or during high/low humidity. I do see a cloudiness in the toluene dispensed from the airbrush during periods of high humidity but this has not translated to increased contaminant levels on the wafers.

As a test we decided to try some UHP Argon over the last few days. The contaminant levels are off the chart. I switched the first bottle after using less than 100psi. Put another bottle of argon on line>>>same story, this time confirmed by another inspector. I just swapped to a new bottle of UHP nitrogen and so far no contaminants. Night & day difference.

Would you happen to know the size of the particulates?

We have not measured them directly because size has not so far been a concern. The fact that they are there at all is the concern. I can calibrate the software and get a measurement if needed.

Are you using Grade 6 or better Dry Nitrogen?

We are told that the UHP is Grade 5.

By Liquid Freon, I mean Freon TF.

Bill

ppp

The particle size would give you a handle as to whether the filter in the gun I suggested should work. That's why I'm curious about the size.

Changing the gas from UHP Nitrogen to UHP Argon should have made zero difference except for cost.

I looked here:
Instrument Grade Pure Gases, UHP Calibration Gas Cylinders, Products, Grades, Specifications

And the contaminant levels seem to be about the same.

That test of Argon vs Nitrogen I was thinking about. The results MAY indicate a vendor quality problem. Nitrogen is cheap enough, what about ordering some from various vendors?

You know that you can order a high pressure liquid nitrogen cylinder and tap the gas side to get Nitogen gas, but I don't know how pure it would be.

Nitrogen generator link with cylinder comparison. One of their units is capable of 7N+ purity.

http://www.sigmaaldrich.com/Graphics.../4600/4545.pdf

I'm now wondering if they have an evaluation unit you can use/rent/free for a length of time.