[labnetwork] Anhydrous HCl best practices

Thu Jun 12 12:56:05 EDT 2014

Tom Britton,

Thank you for your insight.

The Marvell NanoLab is about to commission an HCl feed to an epi system. 
We are planning to add a purifier in series with the pigtail from the 
cylinder to the panel. When one installs such a purifier is it normal 
practice to plumb a shunt around it to use during 
trickle-purging/cylinder changes? Although a new term to me, we have 
made it a practice to reverse-feed N2 through pigtails, to the DISS, 
during cylinder changes, with good benefit.

Also, if one install a reactive HCl purifier does this give an 
opportunity to step-down a level in gas purity for HCl, assuming it 
meets a < 1ppm H2O content?

Bob Hamilton

-- 
Robert Hamilton
University of California at Berkeley
Marvell NanoLab
Equipment Eng. Mgr.
Room 520 Sutardja Dai Hall
Berkeley, CA 94720-1754
bob at eecs.berkeley.edu
Phone: 510-809-8600
Mobile: 510-325-7557 (Emergencies only)
e-mail preferred

On 6/11/2014 9:20 AM, Tom Britton wrote:
>
> Hi John,
>
> I'm glad that you found the source of the leak. Any time a system is 
> compromised it causes huge issues with down time and cost of repair. 
> As you are probably aware, your gas line is most likely trashed to the 
> tool and needs to be replaced to the tool. Don't just replace the 
> panel and run the gas again, as corrosion spreads once it starts. Your 
> gas cabinet vendor should be able to get you a new panel, hopefully 
> quickly, so you're not down for an excessive time.
>
> In reading what you wrote about leak checking after install, this 
> shouldn't be necessary. The major Fabs don't do this as the downtime 
> and cost would be enormous. The initial installation by your 
> mechanical contactor includes a helium leak check through the cabinet 
> to the tool, and if you're performing a cylinder leak check after 
> every bottle change, you should be fine from a leak check perspective. 
> Instead focus on keeping the gas dry. It is well documented that HCL, 
> and most other corrosives, in the gaseous form, can run for years in 
> stainless steel lines with no issues, but as soon as there is the 
> slightest trace of moisture, game over. It's my guess, that when you 
> send this panel back to the gas cabinet supplier, you're probably 
> going to hear that the valve corroded through where you found the leak.
>
> The major points for using any corrosive gas are:
>
> 1.Helium leak check the connection after every cylinder change prior 
> to the introduction of process gas. We advise using a "He/Inert Gas" 
> blend for your purge gas, as you can use the purge gas to helium leak 
> test the connection.
>
> 2.Always use DISS fittings for both the process gas and the purge gas. 
> Not only do they provide a higher level of leak integrity and purity 
> over a CGA, but one is less likely to introduce a lower purity purge 
> gas by mistake than if the connection has a standard 580 CGA on it.
>
> 3.Purge Gas Purifier - We highly recommend a purge gas purifier for 
> all mineral acid gases. Granted, your cylinder costs $1200, but it's 
> cheap insurance and it will last a very long time.
>
> 4.Trickle Purge Valve - check to make you have a Trickle Purge Valve 
> in your purge line and that it's open during bottle changes. This does 
> a good job of keeping atmosphere from getting into the gas line during 
> the physical changing of the bottle.
>
> 5.Cycle Purges - 30 is not enough. Set your purge cycles at 60 for 
> both pre and post purge. The pre-purge is for the safety of the 
> operators, and does a better job of scrubbing the inside of the tubing 
> and components of the HCl, so any atmospheric moisture that gets past 
> the trickle purge doesn't react. The post purge is to dry the inside 
> of the gas line of the tube and components before the HCl is 
> reintroduced into the system. Any less and there is a chance that 
> residual moisture will remain, and once the HCl gas hits it, the 
> corrosion will start and it's a matter of time before you see the effects.
>
> I was talking with Troy on this and he ran his HCl at ON Semi for over 
> 10 years using the above practice, and some of their systems sat idle 
> for quite some time between quartz cleaning steps on seldom used 
> tools. I also talk with a lot of gas systems managers in my travels 
> and this is a fairly common way that the production Fabs keep their 
> corrosive systems running and the problems at bay.
>
> Hope this helps. See you on Sunday.
>
> Tom
>
> Tom Britton
>
> Director of Sales
>
> /Critical Systems, Inc./
>
> Direct: 208-890-1417
>
> Office: 877-572-5515
>
> /www.CriticalSystemsInc.com/ <http://www.criticalsystemsinc.com/>/__/
>
> /__/
>
> logo for email signature png
>
> -----Original Message-----
> From: labnetwork-bounces at mtl.mit.edu 
> [mailto:labnetwork-bounces at mtl.mit.edu] On Behalf Of John Shott
> Sent: Tuesday, June 10, 2014 2:20 PM
> To: Weaver, John R; labnetwork at mtl.mit.edu
> Subject: Re: [labnetwork] Anhydrous HCl after it escapes ...
>
> Labnetwork Community:
>
> First, I'd like to thank Chung Wing Leong, Dennis Grimard, Craig 
> Cheney, Tom Britton and John Weaver for taking the time to respond to 
> my query.
>
> I appreciate your collective expertise and input and am taking a 
> number of things you suggested under careful consideration as we move 
> forward.
>
> In the meantime, we have learned a bit more about this leak and I 
> thought that I would share with you a bit more about what we have learned:
>
> First, the leak seems to have occurred at the joint between the valve 
> body and the tube stub ... not, as we had originally feared and 
> suspected at one of the VCR connection points.  Nor did it occur at 
> one of the orbital welds between the tube stub on the valve and a VCR 
> gland.
>
> To me it is unusual to see a leak at that point in a valve ...
>
> particularly one that has not been in service very long.
>
> Second, this failure occurred downstream (on the tool side) of the 
> pneumatically-controlled, low-pressure outlet valve on the panel.
>
> Certainly on our panel, and in most auto-purge panels of this type, 
> this valve is typically closed during bottle changes and similar leak 
> checking activities. To be candid, this has been a bit of a blind spot 
> for me: while I told you that we leak check the panel each time that 
> we change a cylinder (which is true), now that I think about it more 
> carefully, we don't have a mechanism for leak checking anything on the 
> downstream side of this valve on a regular basis.
>
> In fact what we normally leak check is that portion of the panel that 
> is upstream of the pneumatically-controlled, low-pressure outlet valve.
>
> Since this is a manifolded cabinet that has two bottles feeding a 
> single output line, there are actually a number of potential failure 
> points that we do not regularly check that includes the output side of 
> each pneumatically-controlled output valve, the manual quarter-turn 
> valve in series with and downstream from the pneumatically-controlled 
> output valve and the fittings associated with the tee in the 
> manifold.  As this cabinet has not been service for an extended period 
> of time, those downstream fittings have likely been checked only 
> twice:  once by the cabinet manufacturer at their facility and once by 
> us during the cabinet install.  While our cabinet was helium leak 
> checked and certified by a third-party to be less than 10^-9 SCCM 
> leakage, is there a chance that we missed something then?  I'm 
> certainly wondering ...
>
> My guess is that I am not alone in having thought that we leak check 
> the entire panel and all fittings each time we change a gas cylinder ...
>
> whereas there are likely several fittings and connection points that 
> are not actually checked each time.  I will certainly think about this 
> in a new light ...
>
> Thanks,
>
> John
>
> On 6/9/2014 5:43 AM, Weaver, John R wrote:
>
> > John -
>
> > I've had experiences (certainly plural :)) with both external and 
> internal leaks. For an internal leak, where moisture has gained access 
> to the piping, I have always replaced the piping components up to the 
> point where the leak was blocked - a valve, etc. I use valves with 
> Hastelloy wetted surfaces (sometimes called Hastelloy trim), which is 
> not impervious to HCl corrosion but certainly holds up better.
>
> > For external leaks, the key is to neutralize the surfaces as quickly 
> as possible. In one instance we had to replace many components because 
> we didn't neutralize the surfaces more quickly. A mild base works well 
> for neutralization - something with a pH in the 9 range. Then a 
> thorough rinse to remove any salts is necessary.
>
> > We run all of the gas cabinet exhausts for chloride gas cabinets 
> through our plastic exhaust system, so we don't have to worry about 
> exhaust. If you use a metal exhaust on the gas cabinet where the leak 
> occurred, I'd carefully inspect the exhaust for a month or two after 
> the incident.
>
> > I hope this helps.
>
> > John
>
> >
>
> > John R. Weaver
>
> > Facility Manager
>
> > Birck Nanotechnology Center
>
> > Purdue University
>
> > jrweaver at purdue.edu <mailto:jrweaver at purdue.edu>
>
> > nano.purdue.edu
>
> > ________________________________________
>
> > From: labnetwork-bounces at mtl.mit.edu 
> <mailto:labnetwork-bounces at mtl.mit.edu> [labnetwork-bounces at mtl.mit.edu]
>
> > on behalf of John Shott [shott at stanford.edu]
>
> > Sent: Thursday, June 05, 2014 3:08 PM
>
> > To: labnetwork at mtl.mit.edu <mailto:labnetwork at mtl.mit.edu>
>
> > Subject: [labnetwork] Anhydrous HCl after it escapes ...
>
> >
>
> > Labnetwork Community:
>
> >
>
> > Sooner or later, it seems, anhydrous hydrogen chloride (or a number of
>
> > equally corrosive materials) is going to escape.  At least that is our
>
> > experience at the Stanford Nanofabrication Facility.  Despite our best
>
> > efforts to use quality components, do proper leak checking, etc., it
>
> > seems as if we end up with a leak either in a gas cabinet or in a tool
>
> > near a mass flow controller.  At that point, anything near the site of
>
> > the original leak has been covered with now moisture-laden hydrogen
>
> > chloride ... which, I believe, is far more corrosive than the original
>
> > anhydrous material.
>
> >
>
> > For those of you who have encountered similar situations, how do you
>
> > recover or what to you replace?  Do you have effective means of
>
> > neutralizing those metal surfaces? Do your replace VCR gaskets with
>
> > thicker-than-normal or grooved "super gaskets"?  Do you replace the
>
> > entire assembly?  Do you leak check, put back in service, and pray?
>
> >
>
> > Thanks for sharing your experience and insights,
>
> >
>
> > John
>
> >
>
> >
>
> > _______________________________________________
>
> > labnetwork mailing list
>
> > labnetwork at mtl.mit.edu <mailto:labnetwork at mtl.mit.edu>
>
> > https://www-mtl.mit.edu/mailman/listinfo.cgi/labnetwork
>
> _______________________________________________
>
> labnetwork mailing list
>
> labnetwork at mtl.mit.edu <mailto:labnetwork at mtl.mit.edu>
>
> https://www-mtl.mit.edu/mailman/listinfo.cgi/labnetwork
>
>
>
> _______________________________________________
> labnetwork mailing list
> labnetwork at mtl.mit.edu
> https://www-mtl.mit.edu/mailman/listinfo.cgi/labnetwork

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20140612/c6cb26a7/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 14859 bytes
Desc: not available
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20140612/c6cb26a7/attachment.png>