#homecmos IRC log for Sunday, 2012-11-04

--- Sun Nov 4 201200:00
nmz787anyone need FIB work done?11:13
nmz787anyone around?22:14
azonenbergyep22:20
azonenbergand no, i have no immediate short-term need for a FIB22:20
nmz787have you used one?22:22
azonenbergNo, I've done a decent amount of SEM work and am currently learning TEM though22:22
azonenbergAnd I have things in mind that a FIB would be useful for but don't have the need just yet22:23
azonenbergDo you own one or just have access to it?22:23
nmz787hmm, i'm thinking of using this FIB for making lithography masks for microfluidics22:23
azonenbergI think that's massive overkill22:23
nmz787but I'm looking for a way to reduce the jagged edges since its raster CNC22:23
azonenbergfluidic masks don't need to be that small22:23
azonenbergso using a fib is a waste of time and money22:23
nmz787yeah but I think its cheap, and quick too22:24
azonenbergcheap?? quick??22:24
azonenbergno way22:24
azonenbergnot any fib i've ever heard of22:24
azonenbergif you want cheap, quick masks at micron-scale features check out a place like laserlab.com22:24
nmz787the guy who is gonna let me use his said free/cheap, and that he engraved the entire bible in about 1 sq inch in an afternoon22:24
nmz787yeah I think you may have told me about laser lab22:25
nmz787but even that will still have jagged edges I think22:25
nmz787so still would be interested in a way to blur that step22:25
azonenbergYes, but they're dirt cheap - like $50 for a mask the size of a large sheet of paper at 3.125um lambda22:25
azonenbergDo you need truly rounded edges or are diagonals OK?22:25
azonenberg45deg bends etc22:25
nmz787right but for my fluidics a jagged edge can mean a local pressure drop which can them make salts fall out of solution and a seed crystal develop which will change the volume and restriction22:26
azonenbergAlso, look at the details of the etch processing etc after the mask22:26
nmz787well the edges between pixels is what i'm worried about22:26
azonenbergYour pixels will be rounded, not perfectly square22:26
nmz787unless the beam is twice the dpi22:26
azonenbergChoose a mask resolution smaller than your etch resolution22:27
nmz787i mean the beam is twice the diameter of center to center pixel spacing22:27
azonenbergthat's what it comes down to22:27
azonenbergthen you'll end up low-pass filtering off all of the high frequency noise22:27
nmz787so what do you mean my etch res?22:28
azonenbergWell, the sum total of your photoresist, developing, and etch process will have some limiting resolution22:28
nmz787i was planning on doin contact lithography with photoresin, then curing, then pouring PDMS over that and peeling back when cured22:28
azonenbergSo what is the smallest feature you can actually produce doing this?22:28
nmz787maybe i could etch the cured resist22:29
nmz787then pour PDMS22:29
nmz787well the FIB can do like 20nm or something like that22:29
azonenbergYeah22:29
azonenbergBut you wont get 20nm with contact litho22:29
azonenberglol22:29
nmz787maybe smaller than that22:29
nmz787nah22:29
nmz787i only need maybe 1 micron22:29
azonenbergSo set your mask resolution to be maybe 500nm then22:29
nmz787but I'd like the edge between pixels to be smooth as possible22:29
nmz787ok so the FIB makes dots at 500nm spacing22:30
azonenbergI am almost certain that if you do contact litho and have 500nm noise that it will blur out unless you deliberately try to get super sharp focus22:30
azonenbergand for your needs, you dont want that22:30
nmz787what controls the blur22:30
nmz787i guess that's what i'm looking for22:30
azonenbergLots of factors, things like the photoresin you're using22:30
nmz787just diffraction?22:30
azonenbergHow long you develop22:30
azonenbergDiffraction22:30
azonenbergDistance from mask to wafer/substrate22:31
azonenbergThickness of your resist22:31
nmz787would there be more blur with longer or shorter develop time?22:31
azonenbergIf you develop longer, there will be a tendency to attack very thin features22:31
nmz787hmm, ok22:32
nmz787well that's enough for me to think about for now22:32
nmz787let me know if you have any FIB ideas, I was thinking fresnel lens mold masters, so I could make a magnifying lens22:33
nmz787just for fun, but maybe it could be some $ making gimmick22:33
nmz787fresnel lens for your microbible22:34
azonenbergLol22:34
azonenbergIf you can do small chrome-on-glass masks that would be nice, but I suspect that the fib will cost more to operate than you anticipate22:34
nmz787i have no idea what the operating costs are22:34
nmz787the guy says he charges $375/hr to customers22:35
nmz787but if I was playing around on my own he could do way cheaper22:35
azonenbergThat sounds about right, maybe a little on the high side22:35
nmz787like 10X cheaper22:35
azonenbergI can get time on the one at RPI for $188.50/hr as an outside industrial user22:35
azonenbergor, if i was on a sponsored research project, my grant would be billed at $45/hr22:35
nmz787yeah I just moved to portland OR, and have been looking for industry open labs22:35
azonenbergthat's a flat rate for the cleanroom, includes use of any tool there22:35
nmz787really want metrology access22:36
azonenbergWe have a really nice FESEM in the cleanroom here (zeiss supra 55)22:36
nmz787i used a really sweet white light interferometer at RIT that just shot out 3D profiles22:36
azonenbergThere's also a bunch of scopes in the matsci department that are outside the cleanroom that are much cheaper to get time on22:36
azonenberg$45/hr for the jeol JSM-840 which is a dinosaur but good enough for a lot of stuff22:37
azonenberg65/hr for the FESEM22:37
azonenbergand they have a bunch of TEMs which i'm just learning how to use for a class but don't anticipate actually using in any of my work22:37
nmz787you might look into temwindows.com and their parent site simpore.com22:39
nmz787i was thinking about using them for protein sieves22:39
nmz787when working with DNA and protein reactors22:39
azonenberginteresting22:40
nmz787but if I get going with this FIB, I might be able to make small enough pores that would work, I'm not sure22:40
azonenbergwell for the class i'm only using premade tem samples22:40
nmz787they say they can control the pore size well22:40
azonenbergand i dont anticipate using it myself outside of class22:40
nmz787within like 10 or 15 nm of your specified diameter22:40
azonenbergand FIBing pores sounds very expensive22:44
nmz787bbl22:59
Syncyeah that seems to be inefficient22:59
AdrianGwhat is the meaning of this23:28
--- Mon Nov 5 201200:00

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