From cmfaulkn at snf.stanford.edu Wed May 5 13:35:57 2004 From: cmfaulkn at snf.stanford.edu (Carl Faulkner) Date: Wed, 5 May 2004 13:35:57 -0700 (PDT) Subject: free 3-6p Message-ID: From zlrao at stanford.edu Thu May 6 18:32:07 2004 From: zlrao at stanford.edu (Zhilong Rao) Date: Thu, 6 May 2004 18:32:07 -0700 Subject: reservation on 3-6pm Saturday removed References: <1083878933.409aae1586436@webmail.stanford.edu> Message-ID: <002701c433d3$1be7ab20$39aa0c80@Rao> From chquay at MtHolyoke.edu Fri May 7 16:03:30 2004 From: chquay at MtHolyoke.edu (Charis Quay Huei Li) Date: Fri, 7 May 2004 19:03:30 -0400 (EDT) Subject: quick question Message-ID: is there copper and/or iron in the innotec? i.e. are there pellets in the plastic drawer? thanks! i need to know before my evaporation tomorrow....charis. From cmfaulkn at snf.stanford.edu Mon May 10 10:43:29 2004 From: cmfaulkn at snf.stanford.edu (Carl Faulkner) Date: Mon, 10 May 2004 10:43:29 -0700 (PDT) Subject: cancelled 7a-10a tomorrow Message-ID: From smalekos at unr.nevada.edu Mon May 10 14:51:29 2004 From: smalekos at unr.nevada.edu (Steven Malekos) Date: Mon, 10 May 2004 14:51:29 -0700 Subject: Need Innotec time Message-ID: <000001c436d8$f3134a20$5224c586@me.unr.edu> Hello All, I am a remote user from the University of Nevada, Reno and I lost my highly anticipated Innotec reservation last week due to the chiller problem. If anyone cancels their reservation this week, could you please let me know (preferably Wednesday or Thursday). Thank you, Steven 775-232-3687 -------------- next part -------------- An HTML attachment was scrubbed... URL: From sbank at stanford.edu Mon May 10 19:38:25 2004 From: sbank at stanford.edu (Seth Bank) Date: Mon, 10 May 2004 19:38:25 -0700 Subject: Al Deposition Message-ID: <5.2.1.1.2.20040510193413.02e78620@sbank.pobox.stanford.edu> Hi All, Is anyone planning to deposit Al in the near future who wouldn't mind if I loaded a couple samples with them? We just need to deposit enough Al on a couple Si wafers to give a mirror finish, thickness isn't critical - anything ~1000Ang or more is great. Thanks in advance, Seth Seth Bank sbank at stanford.edu Doctoral Candidate Ph: 650-725-8313 James S. Harris Group Fax: 650-723-4659 Electrical Engineering Stanford University CIS-X Rm. 126X Via Ortega Stanford, CA 94305-4075 From cmfaulkn at snf.stanford.edu Tue May 11 13:59:11 2004 From: cmfaulkn at snf.stanford.edu (Carl Faulkner) Date: Tue, 11 May 2004 13:59:11 -0700 Subject: res canceled for tomorrow 7a-10a Message-ID: <5.2.1.1.0.20040511135835.00b43898@snf.stanford.edu> From jperez at snf.stanford.edu Thu May 13 15:19:01 2004 From: jperez at snf.stanford.edu (Jeannie Perez) Date: Thu, 13 May 2004 15:19:01 -0700 Subject: We have Au in the Innotec! Message-ID: <40A3F455.23965B51@snf.stanford.edu> Please, make sure you are always checking under the SHUTTER for any peeling before deposition. Thanks, Jeannie From mwiemer at stanford.edu Sun May 16 14:27:46 2004 From: mwiemer at stanford.edu (Michael West Wiemer) Date: Sun, 16 May 2004 14:27:46 -0700 (PDT) Subject: reservation removed Message-ID: sunday 1:30-7 Drytek4 is not working so my samples are not ready. From jhemanth at stanford.edu Mon May 17 12:27:24 2004 From: jhemanth at stanford.edu (Hemanth Jagannathan) Date: Mon, 17 May 2004 12:27:24 -0700 Subject: Reservation on tuesday night - cancelled Message-ID: <001401c43c44$fcbb1be0$c66340ab@JAGANNATHANPC> Reservation on Tuesday from 10 pm to 3pm removed. -------------- next part -------------- An HTML attachment was scrubbed... URL: From grupp at snowmass.Stanford.EDU Mon May 17 17:42:31 2004 From: grupp at snowmass.Stanford.EDU (Cosmo) Date: Mon, 17 May 2004 17:42:31 -0700 (PDT) Subject: Innotec: logjam Message-ID: Hi everyone, With the innotec being so maxed out, i am wondering if people have ideas on how to make the system more efficient so we can all get more stuff thru it. THanks, DAn --------------------------------------------------------------------------- Dr. Daniel Grupp, Visiting Scholar Center for Integrated Systems Stanford University Stanford, CA 94305 (650) 724-6911 FAX: 723-4659 --------------------------------------------------------------------------- From mwiemer at stanford.edu Mon May 17 18:29:23 2004 From: mwiemer at stanford.edu (Michael West Wiemer) Date: Mon, 17 May 2004 18:29:23 -0700 (PDT) Subject: res removed Tues 7pm-10:30pm Message-ID: From ben.jian at arrayedfiberoptics.com Tue May 18 19:57:29 2004 From: ben.jian at arrayedfiberoptics.com (ben.jian) Date: Tue, 18 May 2004 19:57:29 Subject: machine available now to 4 PM Message-ID: <20040518195730.277.qmail@server266.com> Wafers got delayed by the fire alarm. -------------- next part -------------- An HTML attachment was scrubbed... URL: From pethe at stanford.edu Tue May 18 15:48:43 2004 From: pethe at stanford.edu (Abhijit Pethe) Date: Tue, 18 May 2004 15:48:43 -0700 Subject: removed res thurs 2300hrs and fri 2100hrs Message-ID: <002701c43d2a$455314d0$296440ab@flash> Wafers not ready -------------- next part -------------- An HTML attachment was scrubbed... URL: From pethe at stanford.edu Thu May 20 17:32:16 2004 From: pethe at stanford.edu (Abhijit Pethe) Date: Thu, 20 May 2004 17:32:16 -0700 Subject: removing res 7:30pm Message-ID: <001101c43ecb$11a13fc0$296440ab@flash> Sorry for the late notice. Wafers are not ready. Abhijit Abhijit Jayant Pethe PhD Candidate Department of Electrical Engineering Stanford University, CA Phone :(O)(650)-725-3608; (C) (650)-387-6435 e-mail: pethe at stanford.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: From cmfaulkn at snf.stanford.edu Fri May 21 16:35:01 2004 From: cmfaulkn at snf.stanford.edu (Carl Faulkner) Date: Fri, 21 May 2004 16:35:01 -0700 (PDT) Subject: Won't be using my 6:30pm res Message-ID: I won't be back from Berkeley in time to use it. Since I can't cancel from here - ignore my reservation. Carl From pethe at stanford.edu Fri May 21 18:31:59 2004 From: pethe at stanford.edu (Abhijit Pethe) Date: Fri, 21 May 2004 18:31:59 -0700 (PDT) Subject: Cancel Reservation tonite at 9:00pm. Wafers not ready Message-ID: From zlrao at stanford.edu Fri May 21 19:06:55 2004 From: zlrao at stanford.edu (Zhilong Rao) Date: Fri, 21 May 2004 19:06:55 -0700 Subject: reservation on 14:30-18:00 saturday cancelled Message-ID: <000a01c43fa1$75404f90$39aa0c80@Rao> An HTML attachment was scrubbed... URL: From grupp at snowmass.Stanford.EDU Mon May 24 16:15:02 2004 From: grupp at snowmass.Stanford.EDU (Dan Grupp) Date: Mon, 24 May 2004 16:15:02 -0700 (PDT) Subject: INNOTEC: Usage question Message-ID: Hello Everyone, I think it would be informative for all of us and to help spped things up, to know how people are using their time. Right now, the system pumps to 2E-7 T in under 2 hrs. What are people doing with their time in a 4+ hour reservation, which almost all of them are? Is this time really being used (like by doing multiple runs)? Further, it pumps into the sixes in under 45 minutes. This is often good enough for most applications. If you are unsure what presswure you need, we can figure it out, and maybe save you and others lots of time. Often, you may be ready to go once the gate valve opens. I'd like everyone to talk about this online. It worked really well for solving the Raith's problems. Regards, DAn --------------------------------------------------------------------------- Dr. Daniel Grupp, Visiting Scholar Center for Integrated Systems Stanford University Stanford, CA 94305 (650) 724-6911 FAX: 723-4659 --------------------------------------------------------------------------- From pxnguyen at mail.arc.nasa.gov Mon May 24 17:44:24 2004 From: pxnguyen at mail.arc.nasa.gov (Pho Nguyen) Date: Mon, 24 May 2004 17:44:24 -0700 Subject: INNOTEC: Usage question In-Reply-To: Message-ID: <000001c441f1$6d5cab50$4a55e88f@lime> Dan, Going to E-6 T is what I did after wasting too much time for E-7. Also, I think J. Perez told me (please correct me if I'm wrong) that SNF usually charges users for the entire 4-hr block, instead of 1 or 2 hours, since this is a standardized time allocation for the tool. As such, members may not have the incentive to complete their runs quickly. Pho Nguyen Research Scientist 650-604-0480 -----Original Message----- From: Dan Grupp [mailto:grupp at snowmass.Stanford.EDU] Sent: Monday, May 24, 2004 4:15 PM To: innotec at snf.stanford.edu Subject: INNOTEC: Usage question Hello Everyone, I think it would be informative for all of us and to help spped things up, to know how people are using their time. Right now, the system pumps to 2E-7 T in under 2 hrs. What are people doing with their time in a 4+ hour reservation, which almost all of them are? Is this time really being used (like by doing multiple runs)? Further, it pumps into the sixes in under 45 minutes. This is often good enough for most applications. If you are unsure what presswure you need, we can figure it out, and maybe save you and others lots of time. Often, you may be ready to go once the gate valve opens. I'd like everyone to talk about this online. It worked really well for solving the Raith's problems. Regards, DAn ------------------------------------------------------------------------ --- Dr. Daniel Grupp, Visiting Scholar Center for Integrated Systems Stanford University Stanford, CA 94305 (650) 724-6911 FAX: 723-4659 ------------------------------------------------------------------------ --- From wistey at snowmass.Stanford.EDU Mon May 24 19:50:30 2004 From: wistey at snowmass.Stanford.EDU (Mark Wistey) Date: Mon, 24 May 2004 19:50:30 -0700 (PDT) Subject: INNOTEC: Usage question In-Reply-To: Message-ID: I should preface my comments by saying that I work with UHV most of the time, so I might be overly paranoid. So I'll play the devil's advocate here... On Mon, 24 May 2004, Dan Grupp wrote: > Right now, the system pumps to 2E-7 T in under 2 hrs. I'm told by people using evaporators elsewhere that this is unusually long, and the base pressure should also be better than that. As a vacuum guy, that tells me whatever's in the chamber is something with a high vapor pressure, like oxygen or water. (Is there a mass spectrometer on the machine? Or could we add one, at least temporarily?) I think the 1e-7 rule of thumb is based on the mean free path at that pressure. If you're depositing something that reacts with oxygen (Ti, Ni), and the background gas is mostly oxygen/water, and your process is sensitive to the oxide, then you'll need the mean free path to be longer than the distance from the source to the wafer. > Often, you may be ready to go once the gate valve opens. Is it wise to stick large amounts of junk on the walls? Sometimes what's on the walls of a chamber comes back to haunt you, especially when it's got high surface area like the frequent peeling in the Innotec. Short term tricks can become long term problems. - Mark From chquay at MtHolyoke.edu Mon May 24 23:08:42 2004 From: chquay at MtHolyoke.edu (Charis Quay Huei Li) Date: Tue, 25 May 2004 02:08:42 -0400 (EDT) Subject: INNOTEC: Usage question In-Reply-To: Message-ID: On Mon, 24 May 2004, Mark Wistey wrote: > On Mon, 24 May 2004, Dan Grupp wrote: > > Right now, the system pumps to 2E-7 T in under 2 hrs. > > I'm told by people using evaporators elsewhere that this is unusually > long, and the base pressure should also be better than that. As a > vacuum guy, that tells me whatever's in the chamber is something with > a high vapor pressure, like oxygen or water. (Is there a mass > spectrometer on the machine? Or could we add one, at least > temporarily?) This is true. Another evaporator I use takes about 30 minutes to get to that pressure. The cryopump seems to be regenerated quite frequently on the Innotec, so probably the problem is something else as you say. > Short term tricks can become long term problems. Amen. Treat the root cause, not the symptoms. OK. Enough platitudes for today. Charis. From shott at snf.stanford.edu Tue May 25 08:08:57 2004 From: shott at snf.stanford.edu (John Shott) Date: Tue, 25 May 2004 08:08:57 -0700 Subject: INNOTEC: Usage question References: Message-ID: <035501c4426a$33b9d5c0$286540ab@jds> Innotec community: Let me first agree with Mark's concern about outgassing being a concern in this system ... but disagree mildly with the claim that a 1E-7 rule of thumb is based primarily on mean free path considerations. If I remember correctly, kinetic theory of gases predicts that the mean free path is inversely proportional to pressure and that the mean free path of air is about 5 meters at 1E-5 Torr and 50 meters at 1E-6 Torr. When the distance from source to substrate is less than a meter, as it is in the innotec, I'd think that mean free path arguments would indicate that 1E-6 Torr is more than adequate. The more relevant thing to consider, in my opinion, is the ratio between the number of metal ions reaching the surface during evaporation each second relative to the number of residual gas atoms reaching the surface each second ... this will determine, to first order, the fraction of non-metallic species incorporated into the metal film. At 1E-6 Torr, gas molecules are hitting the surface at a rate of about 3.8E14 molecules per cm2 per second. In other words, it would take about 2.2 seconds to form a monolayer assuming a sticking coeffiicient of 1. At 1E-7 Torr the molecular incident rate falls by 10 and the monolayer formation time increases by 10. How fast is the metal coming down? For aluminum (mol wt. = 27 and density = 2.7 gm/cm2 so the math is easy ... 6E22 atoms per cc), if the deposition rate is 10 nm/sec, I calculate that the incident flux of aluminum atoms would be about 6E16 molecules per cm2 per second. So, if the vacuum is 1E-6 Torr, slightly more than 0.5% of the incident species would be from the background gas whereas at 1E-7 Torr this number goes down by a factor of 10. If these numbers are correct, is 1E-6 Torr good enough? I'm not sure, but I think that the real question is when does gas incorporation in the evaporated films cause some sort of problems? I'd guess that bulk resistivity isn't affected by modest incorporation of background gas ... but that surface reflectivity and contact resistance may be more sensitive to evaporation at elevated pressures. Here I think that we likely need some discussion of an appropriate test structure followed by some experimentation. I'd also guess that we collectively know a number of people with significant evaporation experience that may be able to provide some useful insight ... maybe they can help to short-circuit this study. Related to pump down times, we should make sure that we are comparing similar machines ... in other words, non-load locked machines that have a similar sized bell jar that is used for extensive evaporation. The evaporated metal on the walls of the chamber act like a big gas sponge that is loaded up with air and water vapor each time the bell jar is opened ... and that after pumping for a while, we are seeing a base pressure that is dominated by steady outgassing of these somewhat porous films. If our pumpdown times are long in comparison to other similar machines, then we need to take a look at a problem ... but I would argue that machines that are load locked, or regularly baked out, or not used for extensive evaporations should be eliminated from our comparison study. For a high-demand tool like the Innotec, we should ask ourselves how to minimize our cycle times, while still maintaining acceptable film quality, and I thank Dan for starting this discussion and appreciate everyone inputs and consideration of this issue. Thanks, John From jperez at snf.stanford.edu Tue May 25 10:59:00 2004 From: jperez at snf.stanford.edu (Jeannie Perez) Date: Tue, 25 May 2004 10:59:00 -0700 Subject: Finished early 5/25/04 at 11 AM Message-ID: <40B38964.1030806@snf.stanford.edu> From wistey at snowmass.stanford.edu Tue May 25 17:08:21 2004 From: wistey at snowmass.stanford.edu (Mark Wistey) Date: Tue, 25 May 2004 17:08:21 -0700 Subject: INNOTEC: Usage question Message-ID: I fully agree with John that it's good that Dan has raised the question as to why we wait for pumpdown. If there's one less I've learned from working in a hands-on research group, it's that you should always question the reasons for the advice that's handed down from previous generations, because the reasons might be wrong. (In fact, my thesis came from that discovery!) But there's also the flip side, which is that previous generations may have had very good reasons for setting up certain equipment guidelines, so the guidelines might be right even if the reasons have been lost with time. Anyway, enough rambling preamble... Thanks to John for correcting my numbers on the argument of mean free vs. surface coverage. He's right that surface adsorption is probably worse than collisions en route. I'd just offer a couple of minor additions... If background contaminants are a problem, then mean free path could still be an issue at 1E-6. To use his numbers, assuming a mean free path of 50 meters, then on average, 1-2% of the metal will collide with something on its way to the wafer (a little less than a meter). So mean free path could still be an issue for critical layers. The other catch is that not everyone is fortunate enough to be depositing bulk aluminum, or at such high rates. :-) For typical deposition rates of Ni, Ge, or Ti, or even gold, the deposition rates can be 10-40x slower, with correspondingly higher fractions of contaminants. At least in theory. John wrote... >we likely need some discussion of an appropriate test structure >followed by some experimentation. Experiment always trumps theory. :-) And yet I wonder whether it would tell us that much for the general user. I would guess most people can probably live with sort-of-adequate films. I just get nervous when users start on the path from impatience to carelessness to abuse. (But then again, my daily work is on a much more expensive and fragile machine, so I'm overly cautious.) Maybe it would be useful to know roughly what the background gases are, so we could tell whether they're bad (water, oxygen) or neutral (He, N2?). >Related to pump down times, we should make sure that we are comparing >similar machines ... Yup, that's what I was comparing. I think the other evaporators (two other university fabs) probably get cleaned around the bell jar more often, which reduces the "sponge." - Mark >>>>> -- - - - - - - - - - - - - - - - - - - - - - - - - + - - - - - - - - - - - - Mark Wistey, wistey at snow.stanford.edu (Internet)| the love of the Lord is 575H Stanford Ave, Palo Alto CA 94305 (BikeNet) | the beginning of wisdom From grupp at snowmass.Stanford.EDU Tue May 25 18:35:13 2004 From: grupp at snowmass.Stanford.EDU (Dan Grupp) Date: Tue, 25 May 2004 18:35:13 -0700 (PDT) Subject: INNOTEC: Usage question In-Reply-To: Message-ID: Note that for most pumpdowns, when the shutter is opened, the pressure shoots up due to the walls being directly heated. This extra pressure is mostly water. So base pressure isn't the whole story! Also, cryo's pump oxygen better than nitrogen, so there's not much oxygen in the chamber at 1e-7 base pressure. Great discussion of the fundamentals! Good education for everyone! To recap the two rules of thumb discussed: 1)1E-6 T is a monolayer/ second 2)5/Pressure in mT = mean free path in cm (or John's 5 m at 1E-5 T) THanks, Dan On Tue, 25 May 2004, Mark Wistey wrote: > I fully agree with John that it's good that Dan has raised the > question as to why we wait for pumpdown. If there's one less I've > learned from working in a hands-on research group, it's that you > should always question the reasons for the advice that's handed down > from previous generations, because the reasons might be wrong. (In > fact, my thesis came from that discovery!) But there's also the flip > side, which is that previous generations may have had very good > reasons for setting up certain equipment guidelines, so the > guidelines might be right even if the reasons have been lost with > time. Anyway, enough rambling preamble... > > Thanks to John for correcting my numbers on the argument of mean free > vs. surface coverage. He's right that surface adsorption is probably > worse than collisions en route. I'd just offer a couple of minor > additions... > > If background contaminants are a problem, then mean free path could > still be an issue at 1E-6. To use his numbers, assuming a mean free > path of 50 meters, then on average, 1-2% of the metal will collide > with something on its way to the wafer (a little less than a meter). > So mean free path could still be an issue for critical layers. > > The other catch is that not everyone is fortunate enough to be > depositing bulk aluminum, or at such high rates. :-) For typical > deposition rates of Ni, Ge, or Ti, or even gold, the deposition rates > can be 10-40x slower, with correspondingly higher fractions of > contaminants. At least in theory. > > John wrote... > >we likely need some discussion of an appropriate test structure > >followed by some experimentation. > > Experiment always trumps theory. :-) And yet I wonder whether it > would tell us that much for the general user. I would guess most > people can probably live with sort-of-adequate films. I just get > nervous when users start on the path from impatience to carelessness > to abuse. (But then again, my daily work is on a much more expensive > and fragile machine, so I'm overly cautious.) Maybe it would be > useful to know roughly what the background gases are, so we could > tell whether they're bad (water, oxygen) or neutral (He, N2?). > > >Related to pump down times, we should make sure that we are comparing > >similar machines ... > > Yup, that's what I was comparing. I think the other evaporators (two > other university fabs) probably get cleaned around the bell jar more > often, which reduces the "sponge." > - Mark > >>>>> > -- > > - - - - - - - - - - - - - - - - - - - - - - - - + - - - - - - - - - - - - > Mark Wistey, wistey at snow.stanford.edu (Internet)| the love of the Lord is > 575H Stanford Ave, Palo Alto CA 94305 (BikeNet) | the beginning of wisdom > --------------------------------------------------------------------------- Dr. Daniel Grupp, Visiting Scholar Center for Integrated Systems Stanford University Stanford, CA 94305 (650) 724-6911 FAX: 723-4659 --------------------------------------------------------------------------- From mwiemer at stanford.edu Thu May 27 16:35:44 2004 From: mwiemer at stanford.edu (Michael Wiemer) Date: Thu, 27 May 2004 16:35:44 -0700 Subject: reservation removed Message-ID: <004d01c44443$56350770$766140ab@longmorn> on Sunday afternoon, Monday & Tuesday early AM. From pethe at stanford.edu Thu May 27 19:29:10 2004 From: pethe at stanford.edu (Abhijit Pethe) Date: Thu, 27 May 2004 19:29:10 -0700 (PDT) Subject: Removed Res Fri 05/28 1830 hrs Message-ID: wafers not ready From curlwang at stanford.edu Fri May 28 18:27:20 2004 From: curlwang at stanford.edu (Ke Wang) Date: Fri, 28 May 2004 18:27:20 -0700 (PDT) Subject: innotec free tonight Message-ID: tried hard but still can't catch it... sorry for very late notice! _____________________________________________ Ke Wang PHD Candidate Department of Applied Physics, Stanford University CISX B113-14 Stanford, CA 94305-4070 Phone: (650)723-8040