#homecmos IRC log for Saturday, 2011-06-25

B0101hi01:57
azonenbergHopefully going to get some more litho work done this weekend05:53
azonenbergThe hope is to refine the liftoff process for the hardmask a little more05:54
azonenbergIn the meantime i'm starting a wiki page on the process06:14
Harold_parkernice man06:15
Harold_parkerhow'd you get on with those probes last night?06:15
azonenbergI didnt do any more work since testing them on the pic a little06:15
azonenbergit was enough to verify i could hit bond pads and other big stuff06:15
azonenbergi dont need to do anythimg smaller until i get my process shrunk :p06:16
Harold_parkerahh nice :)06:20
Harold_parkerexcellent06:20
Harold_parkeryeh for sure man06:20
azonenberghttp://www.youtube.com/watch?v=Jx77NLsQfg8&feature=related06:55
azonenbergIs it bad that i recognized this SEM as a recent model Zeiss before even seeing the logo?06:55
azonenbergBut it seems strange to see someone using one of these things who isnt dressed in a bunny suit :p06:55
CIA-67homecmos r92 | wiki/Si2DLithoProcess.wiki | Created wiki page through web user interface.07:23
azonenbergEarly draft, any thoughts? http://code.google.com/p/homecmos/wiki/Si2DLithoProcess07:25
bart416Pictures!07:29
azonenbergSo I want to start exploring alternate hardmasks19:21
azonenbergTa2O5 is still worth looking into for when vacuum isnt available but i think i can get better results with evaporated Ni19:21
azonenbergalso patterned by liftoff19:21
azonenbergUnfortunately i dont have a working evaporator yet19:22
berndjwhat are you hoping to achieve with this hardmask stuff?19:45
berndjand, does it relate to that PR "peeling" problem you were having?19:46
azonenbergberndj: Yes19:47
azonenbergA hardmask of some sort is necessary because photoresist is attacked by KOH and will disintegrate in seconds19:47
azonenbergSo you need a material that you can pattern using photoresist (by either lift-off or etching) and then use that material as a mask for the KOH etch19:48
azonenbergNormally people use silicon nitride, iirc it's patterned by an HF wet etch19:48
azonenbergBut my resist is not very good against HF19:48
azonenbergSo I tried using lift-off, which works but i am having issues with so far19:49
berndjKOH is to get the oxide off?  or to actualy "dig into" the silicon?19:49
azonenbergKOh is for etching the silicon19:49
azonenbergTo significant depth (for MEMS applications)19:49
azonenbergas in several microns to a few hundred microns19:49
berndjwas about to ask if this was for MEMS19:49
azonenbergYes, thats the main application19:49
berndjdidn't think you'd absolutely need that for making just circuits19:49
azonenbergNo, you typically dont19:50
berndjfor accelerometer type devices: how do you get "under" the probe mass?19:50
azonenbergThere are a couple of methods depending on the process19:50
azonenbergI was going to use bulk micromachining and actually have it free hanging19:51
azonenbergetch from the back up until the center of the die is only a few microns thick19:51
azonenbergthen do another etch from the top down with my target pattern19:51
azonenbergAnother is based on surface micromachining and involves depositing a sacrificial spacer followed by polysilicon19:51
azonenbergyou make the moving parts out of poly-si then dissolve the spacer19:51
azonenbergpresumably at some point in between you'd put support structures through the spacer layer to hold your device up19:52
berndjisn't MEMS an order of magnitude harder to DIY than just "normal" devices (just circuits)?19:53
azonenbergNot for something simple like a comb drive19:53
azonenbergbecause CMOS is so sensitive to trace metal contamination19:53
berndji'm only assuming so because it's a decade or two or three further along in history19:53
azonenbergmems dont really care19:53
azonenbergAnd making really complex mems usually needs RIE19:53
azonenbergBut for simpler stuff, not necessary19:54
berndjdoes NaOH also work to etch Si?19:54
azonenbergYes19:54
berndjworks too well?19:55
azonenbergBut KOH is apparently more repeatable19:55
azonenbergIn mixed signal CMOS+MEMS devices they usually use TMAH which is purely organic and has no metal ions19:55
berndjTMAH?  geez, another james bond movie death-chemical!19:59
azonenbergberndj: Lol20:01
azonenbergWhat do you mean?20:01
azonenbergIt can be toxic but i'm not aware of it having been used in film :p20:01
berndjjust looking at its wikipedia article, that makes it sound like another one of those death-on-contact chemicals like HF20:01
azonenbergIt can definitely be nasty, i'm not planning to use it without more research20:02
azonenbergWhich is one of the reasons i am sticking with mems for the time being20:02
azonenbergsince i definitely dont need it20:02
azonenbergi want to explore alternative photoresist developers, preliminary research suggests straight household ammonia might work20:03
berndjif you can tell me a household-chemicals means of making PR (for pcb type resolution) i'll give you a wet sloppy kiss20:04
azonenbergI'm not aware of any, chemistry isnt my main strong point lol20:05
azonenbergi'm into this mostly for the engineering of devices20:05
azonenbergthe process development is a necessary step to make that happen20:05
berndjmaybe i'll just dick around sometime with indoor varnish + UV, see if anything looks workable20:06
berndjmy outdoor timber suggests there's some UV action, but i don't want to sit with multi-year exposures!20:06
azonenberglol :p20:07
--- Sun Jun 26 201100:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!