About C4F6 and C5F8 gas detector 2006-11-27

Sent: Monday, November 27, 2006 7:53 PM

To: SEMI EHS Grapevine

Subject: About C4F6 and C5F8 gas detector

SEMI Environmental Health & Safety Grapevine

 

Hi,

As we know, C4F6 and C5F8 are hazardous gas. Gas detectors are installed at work place. Who could tell me the TLV value of C4F6 and C5F8? What's the alarm value?

Peter

 

Peter,

One of the manufacturers of C4F6 dramatically reduced the OEL from 5 ppm to 0.5 ppm based on recent testing.

Eugene Ngai

 

 

Any Rule of Thumb on monitoring for gas at bottom of cylinder cabinet if the specific gravity is > than? 2008-12-12

 

Sent: Friday, December 12, 2008 11:24 AM

To: SEMI EHS Grapevine

Subject: Any Rule of Thumb on monitoring for gas at bottom of cylinder cabinet if the specific gravity is > than?

SEMI Environmental Health & Safety Grapevine

 

EHS Professionals:

I recognize that all gas cylinder cabinets must be designed with adequate air flows to ensure << PEL, ceiling or 50% IDLH in event of leak, and need to have a negative pressure BUT…. Is there any rule of thumb for when to locate a toxic gas monitor at bottom of gas cabinet versus the top??? We are using C4F6 and it has a specific gravity of 5.6 (1.0=air) so there would be some motivation to sense for a leak at the bottom if the gas molecules (during leak) fall outside of the exhaust stream boundary layer, right?

 

Is there any rule of thumb? That if SG >? should locate at bottom (in addition to top?)

Mark

 

Mark,

A properly designed and operated gas cabinet should capture and exhaust any gas leakage regardless of their specific gravity. A gas sensor just before the duct should pick up any leakage. I cannot envision any advantage to putting a gas sensor at the bottom. This implies that the gas leakage which would be at the top of the cabinet would diffuse downward against the ventilation air flow to the bottom. The only time this would happen is if the air flow is off. Locating a gas sensor at the bottom also exposes them to impact damage when moving cylinders in and out.

Eugene Ngai

 

Arsine: Toxic Gas Monitoring 2008-03-21

 

Sent: Friday, March 21, 2008 8:33 AM

To: SEMI EHS Grapevine

Subject: Arsine: Toxic Gas Monitoring

SEMI Environmental Health & Safety Grapevine

 

How low does your site monitor for arsine, and what alarm level settings are you using?

Bart

 

 

Bart,

As a manufacturer of Arsine we use a considerable number of gas sensors at our production and warehouse facilties. The recent change to the TLV caused us to reexamine our protocols and detection methods We benchtested our current electrochemical and papertape detectors with low levels of Arsine and found that the papertape readily detected at the 2 ppb level while the electrochemical was >10ppb. The problem is that it maxs out at 25 ppb Our facilities will select the monitoring based on activity

 

1.     Areas that are routinely occupied and process Arsine will use papertape with a warning at TLV and alarm at 4 X TLV

2.     Areas that are not routinely occupied (Storage, transportation) will use electrochemical, warning at PEL and Alarm at 4X PEL. Note these locations are also periodically IH tested

 

Eugene Ngai

 

 

Bulk Silane Installations Gas Detection 2006-12-21

 

To "SEMI EHS Grapevine"

Subject: Bulk Silane Installations

SEMI Environmental Health & Safety Grapevine

 

For those of you that work at facilities that have outdoor bulk silane systems, I have the following questions regarding toxic gas monitoring (TGM) at the bulk system itself.

 

1. Do you employ the use of hydride type monitoring to detect releases?

2. If so, what type of technology is used – electrochemical cells, tape?

3. Upon high level activation does the TGM perform auto-shutdown of the system or does it just provide notification to a central monitoring station?

4. If you have TGM, is it reliable or have you had a history of false shutdowns? With no specific code requirement (that I am aware of) for gas monitoring in outside, unconfined bulk installations (meeting all the other safety requirements specified by FM and GCA - flame detection, deluge system, etc) the decision seems to be left up to facilities to determine whether or not TGM is required and if so whether or not to perform auto shutdown. Thanks for your time in advance.

 

Regards,

Peter

 

Peter,

There has always been a debate over this topic and no agreement as to the right answer Hydride detectors in some outdoor areas are prone to false alarms and under some weather conditions it's doubtful that the leak will drift by the sensor before it's carried away by the wind, as a result the codes do not mandate gas detection for bulk installations which are outdoors. In recent years we have detected minor leaks which were not visually evident, not on fire, fuming or slowly growing SiO2. These were detected by Hydride detectors. In our release testing in May, we were able to release Silane without ignition times under s variety of conditions. In these cases only a hydride detector would have detected a leak. Of course the dilemma in these cases would be "now what do we do?" As described in Dr Tamanini's study, shutting off the flow was the ignition source. That was how we blew up the gas cabinet

At the SESHA Conference this April, I will be speaking about Silane, TMAl and ClF3 Safety and ER.

 

Eugene Ngai

 

 

Sent: Wednesday, December 20, 2006 11:29 AM

To: SEMI EHS Grapevine

Subject: Re: Bulk Silane Installations

SEMI Environmental Health & Safety Grapevine

 

We utilize hydride tape detection at all silane VMBs and tool gas jungles. Activation of these gas det points will shutdown the source cylinders. At the bulk system itself, UV/IR detection is utilized that will shut down the cylinders and open a deluge valve. An alarm signal is sent to 24/7 security. Since the bulk system is not within an exhausted enclosure, no TGM is used.

Brian

 

 

C5F8/C4F6 detection 2009-05-02

 

Steve

This is the real beauty of the grapevine. There is open and honest communications on a topic.

I the last 24 hours I have learned much more about FTIR and its capabilities.

Eugene

 

Sent: Saturday, May 02, 2009 3:48 PM

To: SEMI EHS Grapevine

Subject: Re: C5F8/C4F6 detection

SEMI Environmental Health & Safety Grapevine

 

All,

Normally I prefer not to name names as I believe the rules for the Grapevine dictate, but I believe clarity to Jay's comment below is required, and in this one (never before) situation I disagree with Eugene.

 

An IR-148 is a narrow band IR detector. In other words it relies upon very few points of IR absorption to determine the presence of target chemistries. Obviously a system that does not utilize the entire IR spectrum is going to be limited in capability, and present an unfair bias for years of a method use that have established laws determining molecular identification through IR absorption. Simply stated, if you are going to decide to use an IR method, don't automatically limit your ability by using a device that uses only a small portion of the spectrum. An IR method is the only method that will not erroneously identify a cross interferant with a target chemistry, and is best for use with fluorinated compounds.

 

Electro-chemical and colorimetric detection methods provide an inexpensive solution, relatively speaking, to IR methods. Considering this, it is sometimes preferable to use these methods. However, utilizing a method that relies upon chemical reaction will always present potentials for cross interferences.  Colorimetric and electro-chemical methods that detect target chemistries that must be "cracked" in order to react are especially sensitive to cross inferences since so many materials are bound by fluorine. Although it is possible to physically filter these cross interferants, in most cases doing so also removes or slows the target chemistry as well.

 

Ok, now that I’ve upset our sales force, and alienated myself from the best Grapevine source we have through Eugene, let me know if I can provide any further clarifications for anyone.

Best Regards,

Steve

 

 

 

Date: Sat, 2 May 2009 10:53:04 -0400

To: SEMI EHS Grapevine<sehs@ehs.semi.org>

Subject: FW: C5F8/C4F6 detection

SEMI Environmental Health & Safety Grapevine

 

Gas Detection C5F8/C4F8

 

Date: May 1, 2009

 

Since most of the fluorocarbon (CxFy) molecules are not chemically reactive, the more specific gas detection methods using an electrochemical cell is not available.

 

The 2 primary gas detection methods currently used are Infrared or Pyrolysis followed by Electrochemical detection. Both have significant interference from other fluorocarbon and chlorofluorocarbon molecules as well as other gases that might be present in the Fab. (SF6, CO2. Refrigerant Gases). Isopropyl alcohol (IPA), has been reported to be a significant interferent, there is a filter available to remove it before the sensor but it must be changed every few months. Galden fluids which are perfluoropolyethers, containing carbon, fluorine and oxygen and are used in heat transfer systems have also been a problem.

 

Infrared

Two infrared methods can be used, FT-IR (Fourier Transform IR) or ND-IR (Non dispersive IR). The detector will measure the absorbance of light at various wavelengths in a long path length cell (>6 m). Typically a ND IR will focus on one wavelength characteristic of the gas, unfortunately most similar gases will also have an absorption at this wavelength. The FT-IR may be a little more selective since it will fingerprint the molecule and use a variety of wavelengths. This however will still be fooled by similar molecules

 

Pyrolysis

The sensitivity of detection is a function of the efficiency of cracking the molecule and forming a fluoride gas which will be detected by a mineral acid electrochemical cell or impregnated paper. These are more sensitive but still have a variety of interferences.

 

As reported by R. Matsuoka, et al “New Semiconductor Gas C5F8”, The Earth Technologies Forum The Conference on Climate Change and Ozone Protection ,Sept. 27-29, 1999 the destruction of the fluorocarbons at 600oC is

 

Molecule	Destruction
C4F8	5%
C3F8	20%
C2F6	40%
CF4	75%
CHF3	95%
C5F8	100%
NF3	100%

 

Pyrolyzers are made of stainless steel or quartz. Corrosion from the fluoride byproducts can be of concern for system maintenance.

 

Interferences

Any fluorocarbon or chlorofluorocarbon will interfere since they will be pyrolyzed to the corresponding halogen which will be detected by the mineral acid detector. While the mineral acids, HF, HCl, BF3, WF6, etc will not be pyrolyzed they will immediately react with the acid detector. IPA will be pyrolyzed but not detected.

Eugene Y. Ngai

 

Sent: Friday, May 01, 2009 10:40 PM

To: 'SEMI EHS Grapevine'

Subject: RE: C5F8/C4F6 detection

 

To All

Discussion on this topic goes back to 2004. Since I’m home recuperating, I took the time to recap the answers and added some of my own. This reinforces Jay’s comments

Eugene

 

Sent: Friday, May 01, 2009 2:39 PM

To: SEMI EHS Grapevine

Subject: RE: C5F8/C4F6 detection

SEMI Environmental Health & Safety Grapevine

 

We have been using these and similar gases for several years.

 

We have used “IR-148” systems for detection.

 

Interferences with other fluorinated materials is significant and we have been unable to isolate the interferences.

 

We have learned (in part through experience) that paper tape technology used for NF3 detection (pyrolizing the sample, then detecting with mineral acid tape) detects these gases with less interfence than the IR detectors.

 

However, there is no calibration for this detection system for these fluorinated gases.

Jay

 

 

Sent: Friday, May 01, 2009 10:40 AM

To: SEMI EHS Grapevine

Subject: Re: C5F8/C4F6 detection

SEMI Environmental Health & Safety Grapevine

 

Jen,

Your best source of detection will be an IR based product. It is the only method that can detect these specific targets (easily) without cross sensitivities. If you any questions regarding this please feel free to contact me.

Regards,

Steve

 

Date: Fri, 1 May 2009 11:28:42 -0400

To: SEMI EHS Grapevine<sehs@ehs.semi.org>

Subject: C5F8/C4F6 detection

SEMI Environmental Health & Safety Grapevine

 

Good morning all,

We currently are having cross sensitivity issues with the C5F8/C4F8 detectors that we are using and are looking to change them out. Before doing that I would like toknow what detectors otherpeople are using to detect C5F8 and C4F6 that are not cross sensitive to fluorinert or other inertfluorinated gases.

 

Thanks for your time on this matter.

Jennifer

 

Explosion-Proof Refrigerator/Freezer 2004-09-27

 

Please note that Explosion Proof does not mean that it will contain an explosion. It means that the electrical contacts/wiring will not ignite a flammable atmosphere either by the use of special enclosures or by current limiting devices

Eugene Ngai

 

To: SEHS Grapevine

Subject: RE: Explosion-Proof Refrigerator/Freezer

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

As long as it is accessible by emergency responders (ERT, Security, Fire Dept) I wouldn't see a problem. I believe I've seen units commercially available that come with lockable doors.

Steve

 

 

Sent: Monday, September 27, 2004 1:00 PM

To: SEHS Grapevine

Subject: Explosion-Proof Refrigerator/Freezer

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

All,

Does anyone see any safety reason to not put a lock on an explosion-proof refrigerator/freezer? The unit is used for resist and other materials storage, and because we are a user facility we like to safeguard these items. It is of my opinion that the unit is designed to contain an explosion so there is no reason a lock would hinder its use or design safety wise.

 

Any and all comments are welcome.

Thanks,

Joe

 

 

Nitrogen Trifluoride Monitoring 2007-11-01

 

I would like to clear up the confusion on the NFPA health rating.

The reference study

 

Vernot, E.H., Haun, C.C., MacEwen, J.D., and Egan, G.F., “Acute Inhalation Toxicology and Proposed

Emergency Exposure Limits of Nitrogen Trifluoride,” Toxicol. Appl. Pharmacol., Vol. 26, 1973 pp 1-13

 

has determined the LC50 of NF3 to be 6700 ppm

 

As such it is classified as a 1

 

Hope this helps. I am also forwarding this to Matheson to review

Eugene Ngai

 

Sent: Wednesday, October 31, 2007 4:19 PM

To: SEMI EHS Grapevine

Subject: RE: Nitrogen Trifluoride Monitoring

SEMI Environmental Health & Safety Grapevine

 

Doug,

Thank you for the response. We used MSDS information for the hazard rating. Matheson gives it an NFPA rating of 2-0-0. The Praxair MSDS and the Air Products MSDS both give an NFPA rating of 1-0-0.

Jamie

 

 

Sent: Wednesday, October 31, 2007 11:39 AM

To: SEMI EHS Grapevine

Subject: RE: Nitrogen Trifluoride Monitoring

SEMI Environmental Health & Safety Grapevine

 

Jamie,

What source are you using to determine that the rating went from a 3 to a 2? HMEX? Other? We're still monitoring and auto evacuating.

Thanks,

Doug

 

Sent: Wednesday, October 31, 2007 10:29 AM

To: SEMI EHS Grapevine

Subject: Nitrogen Trifluoride Monitoring

SEMI Environmental Health & Safety Grapevine

 

I’d like to know what current practices are for continuous toxic gas monitoring for NF3 (nitrogen trifluoride). When the NFPA health hazard rating went from 3 to 2 several years ago we continued to monitor but do not automatically evacuation. We have a tech alarm and conduct an investigation to determine the source of the alarm.

 

What are other companies doing?

Jamie

 

personal chlorine single gas monitors 2005-07-27

 

We evaluated a number of portable gas monitors for ER approx 7 years ago and standardized on the GMD Autostep Papertape monitor as being better suited for our ER use. They worked with us to make some improvements It is a papertape monitor which has a number of benefits versus the TLD-1 which we also use

 

Selector switch rather than keys for 3 gases (Hydride, Cl2 and HCl)

One hand operation

Demand Sampling Button

Cassette rather than a Reel

Display is easier to see during use rather than twisting your arm to position it

 

For us in a leak scenario, a papertape monitor is better suited to ER  because they can be rezeroed quicker by advancing to fresh section of tape. This is why the demand sampling is critical. It doesn't advance until you tell it to, otherwise you can very quickly run out of tape. The gel and liquid type sensors will also get saturated and can take a long time to rezero, sometimes the cell has to be replaced.

 

Our requirements are much higher than the gas enduser since we are coming in for a known leaker which will subject the instruments to much higher levels

Eugene Ngai

 

 

Sent: Wednesday, July 27, 2005 8:27 PM

To: SEHS Grapevine

Subject: RE: personal chlorine single gas monitors

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

Our ERT detection concerns in the various areas of the Fab are hydrides, acid gases, and CL2. We are currently evaluating the handheld Miniwarn Drager unit which has the capability to detect 4 gases at once and seems to work well for our ERT scenarios. The CL2 takes a while to stabilize, but once stable, the ERT like the convenience of the unit and the multigas capability.

With Regards,

Mike

 

 

Sent: Wednesday, July 27, 2005 11:55 AM

To: SEHS Grapevine

Subject: RE: personal chlorine single gas monitors

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

We use SPMs (by Zellweger) for much of our emergency response. As you are probably aware, they utilize  Chemcassette technology. The units are definitely not considered personal devices but they are portable. Emergency responders will typically bring them into the area where they are responding.

 

 

Sent: Wednesday, July 27, 2005 12:12 PM

To: SEHS Grapevine

Subject: personal chlorine single gas monitors

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

I am looking at alternative ways to monitor for chlorine gas for use in emergency responses. We currently use the TMX412 with a chlorine sensor. The problem I have been having with this unit is during the calibration. I received some information on the BW Gas Alert from a local vendor. I would really like some input from this group on what you are using for personal chlorine monitoring and if you are happy with its performance.

Thank you,

Kim

 

Question about non-Silane Pyrophoric Gas Detection... 2009-10-02

 

The combustion of metal hydride gases such as Arsine, Phosphine, Silane, Hydrogen Selenide is typically quantitative, there will be no traces of the gas to detect and they form solid byproducts which are also not detected. I have been in ER’s where the hydride gas detector alarms briefly before the hydride gas ignites. It doesn’t alarm again until the fire goes out.

 

A thermal detector relies on radiant and/or convective heat transfer. It can take minutes to detect. A small fire which is the most likely will probably not be detected. UV/IR detection is millisecs but you must be aware of shadows and sensitivity. To prevent spurious alarms most users have a time delay of 5 secs. At the typical UV/IR sensitivity setting, it can detect a 10” flame at 15’. A 10” flame is 70 lpm of silane! Most users have them in gas cabinets where considerably smaller flames will be detected, the challenge is positioning it properly in that tight space.

 

The best practice is to have gas detection and UV/IR for the pyrophorics like Phosphine, Diborane and Silane. Thermal detection is a backup. I have been able to release other pyrophorics at high pressure without ignition in experiments with Prof Chen

Eugene Ngai

Chemically Speaking LLC

 

 

Sent: Friday, October 02, 2009 10:08 AM

To: SEMI EHS Grapevine

Subject: Question about non-Silane Pyrophoric Gas Detection...

SEMI Environmental Health & Safety Grapevine

 

A little guidance needed….

With all the emphasis lately on Silane gas and Silane Safety Awareness, one of the key issues was recommendation of using UV/IR sensing for flame detection (versus use of indirect thermal sensing).

 

Although current fire codes “allow” for the exception to use thermal sensing to detect a flame it has been rationalized that the cooling effect due to gas cabinets/gas box airflow can result in a Silane flame NOT being detected if the thermal switch is significantly far away from flame/fitting. Because of the rapid burn of Silane, the combustion by products of Silane cannot be detected with a gas sensor. BUT….I do not believe this is the case with other pyrophorics, such as Diborane, Phosphine or Germaine.

 

1.     Can the same gas detection (not flame detection) be used to detect a leak of say Phosphine (both unignited < LEL and ignited) …due to the detection of its products of combustion from the burn???

2.     What is the chemical/semiconductor industry feedback on NOT using thermalswitches in areas of high cooling’, and NOT installing UV/IR either… but instead relying only on gas detection in these special cases?.

 

Mark

 

TGM-NF3 2006-02-09

 

Sent: Tuesday, January 24, 2006 4:34 PM

To: SEHS Grapevine

Subject: TGM-NF3

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

I am soliciting responses to the questions below concerning TGM levels with respect to NF3. While NF3 has a relatively low TLV (10 ppm) and very poor odor warning properties, there is no STEL published (to my knowledge) and its IDLH is 1000ppm. From a code perspective, it's not an HPM, or a highly toxic

or toxic chemical. With that being said it is an oxidizer.

1.     What TGM levels are you using for ambient detection? Are you using typical TGM levels such as warning at 􁪽 TLV and high alarm at TLV (with subsequent gas shutdown, evacuation, etc)?

2.     What about in a gas cabinet exhaust, are you using the same levels as above, with shutdown of the cabinet at TLV or do you have different alarm levels?

3.     If you don't use the above but yet have TGM, what has been your rational for treating it differently (raising those levels)?

Regards,

Peter

 

 

Sent: Friday, January 27, 2006 6:09 AM

To: 'SEHS Grapevine'

Subject: RE: TGM-NF3

 

Peter

Sorry for the late reply, I have been traveling in Asia and have been rushing to finish all my meetings, etc before the lunar New Year Attached please find the AP Safetygram for NF3. As summarized in the

Safetygram, the ACGIH Time Weighted Average (TWA) of 10 ppm (29 mg/m3 for NF3 was based on the 19-week subchronic inhalation study that was conducted in rats. In that study, rats that were exposed to 100 ppm of NF3 for 7 hours per day, five days per week for 19 weeks developed pathological changes in

their livers and kidneys. The ACGIH incorporated an additional safety factor of 10 to account for longer exposure periods when setting the TLV.

 

The LC50 of 6700 ppm is above the 5000 ppm level for it to be classified as a Toxic Gas under the US Transportation regulations and is well above the 2000 ppm level for the US Fire Codes to classify as a Toxic gas. Some Asian countries however have an additional classification for toxicity if the TLV is <200 ppm. Nitrogen Trifluoride and Silane would be classified as Toxic in those countries. Also note the  emergency Exposure limits (EEL) of 750 ppm for 30 minutes and 375 ppm for 60 minutes.

 

As a precaution AP recommends monitoring for users at the same levels as a HPM gas since it is a odorless gas. Warning would be at 50% of TLV and alarm at 100%. Please note that since this is not an HPM, some customers do not use gas cabinets nor monitor these gases. In some of our processing areas which may contain hundreds of mechanical fittings and large flanges we allow an increase of the alarm level to 4X the TLV to reduce the likelihood of false alarms shutting down the process. In normal operation it is 0 ppm. While NF3 is classified as an oxidizer it exhibits this property only under

high temperature/pressure conditions similar to Nitrous Oxide. Release testing that were conducted in the last few years comparing ClF3, 20% F2, O2, N2O & NF3 showed NF3 to be less of an oxidizer than pure Oxygen. Testing was done at atmospheric pressure onto burning clothing, wooden splints, candles

and cigarettes.

 

I hope this answers your questions

Eugene Ngai