#homecmos IRC log for Tuesday, 2011-07-19

lekernelbart416, i've read that sputtering outside of a vacuum gives bad result... the deposition is not uniform, is brittle, oxidizes, etc.09:10
bart416lekernel, yeah but it's possible09:11
lekernelprobably not for microelectronics :)09:11
bart416SolidRaven> The thing is, you could theoretically sputter coat at atmospheric pressure09:11
lekerneleven for depositing aluminum on potato chips bags they use vacuum09:12
Action: bart416 once again points at theoretically :P09:21
bart416You'd just need a strong electric field to force the ions in the right direction09:24
Laurencebive spluttered paper in air11:32
Laurencebwith graphite, using a large cap11:32
Laurenceband 1KV  XD11:32
Laurencebpropelling pencil lead11:36
Laurencebare there processes for two component aluminium deposition?12:30
Laurencebanyone here?13:21
azonenbergJust got up but heading to work shortly13:22
Laurencebare there processes for two component aluminium deposition?13:22
Laurencebim sure ive seen two component alu deposition using sprays13:23
Laurencebbtw, have you seen jeri ellsworths work with diy mosfets?13:23
azonenbergLaurenceb: Yes, that was actually the inspiration for this work13:25
azonenbergI wanted to do a process shrink on her basic design13:25
Laurencebiirc i saw a video on low volume cd production13:27
Laurencebthey had a two component alu deposition spray thing13:27
Laurencebwish my chemistry was better :(13:27
azonenbergvery interesting but i'm not familiar with the process13:28
azonenbergIt'd have to be something like sol-gel13:28
Laurencebi remember we did chrome deposition in liquid state in 6th form13:28
azonenbergBut the problem is that Al oxidizes sooooo easily13:28
azonenbergYou'd almost certainly have to do it under inert gas13:28
azonenbergThough possibly not low pressure13:29
Laurencebim trying to work out if you could do a liquid state process, like jeris but without the lame comductive epoxy13:29
azonenbergI'm going to be playing with an evaporator on campus shortly13:29
azonenbergPrepping samples tonight/tmrw and running the experiment thursday13:29
Laurencebnice, its not very complex but vacuum complicates things13:29
Laurencebi like jeris oven+nitrogen, very basic13:29
Laurencebi also like the idea of liquid microjets, but i know its a little mad13:30
azonenbergI'd probably use argon since i'd use it for sputtering13:30
azonenbergEven if it costs a bit more, cheaper than two tanks13:30
azonenbergBut in any case i'll be playing with lift-off and normal etching on ~100nm Cu films13:30
azonenbergIf it's even remotely successful i plan to build one of my own over the coming months13:31
Laurencebtheres got to be a use for that13:32
azonenbergDeposition of particles for nanostructures? Hmm13:33
azonenbergThe thing is, you're depositing single particles13:33
azonenbergyou'd have a giant polycrystal with no cohesion between grains13:33
Laurencebyeah but if you used it for dopants13:33
azonenbergYou'd need to heat it to near the melting point to densify the film (still polycrystalline but at least they stick together)13:34
Laurenceband alu maybe13:34
azonenbergDoping is trivial13:34
azonenbergSpin coat doped SiO2 and pattern with HF13:34
azonenbergThen bake and let it diffuse13:35
Laurencebi dont understand the spin coating agent13:36
Laurencebits dissolved in alcohol?13:36
azonenbergI dont know the exact chemistry for spin coating SiO2 but its a silicate compound of some sort13:36
Laurencebi see13:36
azonenbergthat upon drying and gently heating (200C ish) decomposes to SiO213:37
Laurencebso how is it converted to SiO2?13:37
Laurencebah i see13:37
azonenbergSo basically you'd pattern the doped film with photoresist and HF13:37
azonenbergThen bake to drive it in13:37
azonenbergYou could also backfill the entire wafer with SiO2 to prevent further oxidation13:37
azonenbergand do dopant drive-in in the air13:37
Laurencebso you could print the dopant13:38
azonenbergTheoretically, yes13:38
Laurenceband you can print HF13:38
azonenbergBut in that case it'd be just dopant and not oxide too13:38
azonenbergYou dont need to13:38
azonenbergCoat the entire wafer13:38
azonenbergJust print the masking layer13:38
Laurenceboh course13:38
azonenbergThe only reason you'd print is if its easier than the alternate method13:38
Laurencebhmm i need to understand latlor cones13:39
azonenbergAnd for somethign like a wet etch, it isnt13:39
azonenbergIn any case i'll be running some experiments tonight involving diluting photoresist and doing tantalum oxide liftoff using thinner layers13:41
azonenbergBut for now, off to work for me13:41
Laurenceb_azonenberg: i was looking at microjets20:55
Laurenceb_looks very hard20:55
azonenberg_workThey're interesting but i have no plans to pursue them21:00
Laurenceb_there are a few research groups working on the same idea21:00
Laurenceb_so its not totally mad, but getting consistent drops that are small enough is very hard21:00
Laurenceb_arg stupid google21:02
azonenberg_worklol where's the actual paper?21:02
Laurenceb_i was getting down to 350nm or less with a nebuliser21:03
Laurenceb_- page 1121:03
azonenberg_workThis is you?21:03
Laurenceb_i was using NaCl in water and an ultrasonic nebuliser21:05
azonenberg_workStill reading21:06
azonenberg_workRPI's electronics club did a high altitude balloon launch two summers ago but i wasnt involved21:07
azonenberg_workAnd it was a proof-of-concept that just took photos and didnt carry any other sensors21:07
kristianpaul(cheap autonomous sonde) highaltitude balloons have interestinhg aplications21:08
azonenberg_workLaurenceb_: It's probably bad that i understand a lot of your math dealing with the aerodynamics of the particles21:10
azonenberg_workDespite having a degree in comp sci :P21:10
Laurenceb_im sure theres errors in there... but i got a first for it so :P21:11
azonenberg_workYour mean free path looks a little long, i recall 58nm, but i think the delta is due to STP vs room temperature21:12
azonenberg_workAnd very interesting on the thermal precipitator21:12
azonenberg_workHow often is that used?21:13
Laurenceb_its quite common aiui21:13
azonenberg_workHmm... the electrostatic seems interesting too21:14
Laurenceb_i was origionally thinking the particle size range was interesting, but maybe electrostatic coating could have uses21:14
azonenberg_workAnd lol @ impactor21:16
azonenberg_workProbably not the most effective method to use at low velocity...21:30
azonenberg_workWhat kind of SEM were you using? I dont recognize the pin design21:38
azonenberg_workLooks too short for a Jeol and too tall for a Zeiss21:38
Laurenceb_it was modified on a lathe21:42
Laurenceb_to reduce the diameter so it would fit in the precipitator21:42
azonenbergI see21:59
azonenbergLaurenceb_: So are you at Oxford?22:04
Laurenceb_i got a job in the real world after graduating22:04
azonenbergGraduated already? Ah22:05
Laurenceb_2008 - i had a bit of a 'crisis' after a while22:05
azonenbergI see - with me it was "in for a penny in for a pound"22:06
Laurenceb_decided to do a phd22:06
azonenbergSame here22:06
azonenbergAs of now i plan to go through with it22:06
azonenbergYou bailed and got a real job instead?22:06
Laurenceb_yes, but didnt enjoy it22:06
Laurenceb_i worked for ~1year at surrey satellite technology on galileo22:07
azonenbergI've been a de facto grad student for a while lol22:07
azonenbergMy grades started slipping last year as an undergrad because i was so into my research i spent less time on class work22:07
Laurenceb_i didnt like work in a largish company as there was no room for creativity22:08
Laurenceb_im now working at nottingham university doing a PhD on biomedical sensing22:09
Laurenceb_at sstl i was stuck in a team working on vhdl for decoder ics22:11
Action: azonenberg prefers verilog22:11
azonenbergI'm actually ordering a whole load of Xilinx CPLDs (5 to 10 units each of four types, ranging from 36 to 144 macrocells) to fool around on22:12
azonenbergFPGAs will come later, they pretty much need massively multilayer boards22:13
azonenbergdue to the multiple power rails etc22:13
azonenbergI've done some FPGA designs but only run them on simulations22:13
Laurenceb_anyway i need some sleep, 11.15pm here22:13
azonenbergCPLDs look cheaper to actually USE22:13
azonenbergIts only 18:15 here, and ttyl22:13
--- Wed Jul 20 201100:00

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