Semiconductor Fabrication Basics - Thin Film Processes, Doping, Photolithography, etc.

[@johndavid360]

You should make a series called ‘sand to nand’.

[@KonradTheWizzard]

Search for "Sand to Silicon" - there are multiple videos on this. Most of them more polished and less in-depth of course...

[@Meyoutoo2222]

My god man. I learned about you from a Wired article today and had to check out your channel. I am absolutely blown away to say the least. I'm a software developer by trade and have never even imagined creating my own ICs. I've dabbled with premade discreet transistors, but nothing like this even seemed possible in a DIY lab. My hat's off to you, subscribed and will be watching EVERYTHING you post!

[@mjvnoth]

After 7 years this video is priceless 🖤🖤🖤🖤 finally

[@mike3891973]

I worked for Emulsitone in the early 90’s in Whippany NJ and made these dopants. I worked with phosphorus pentoxide, boron, arsenic pentoxide and many others. Our biggest seller at that time was EMS 1146 which was used as a laser scribing solution. It was a water based solution with very little health hazards. The film he speaks of is a liquid glass I used to make called silicafilm (Si02), which was a mixture of silicon tetrachloride and acedic anhydryde. After the Si02 Crystal was purified you would mix up to 30% of it with ethyl alcohol and behold, liquid glass. This was all possible due to a brilliant man named Milton Genser who I miss. Ion implanting was new to diffusion sourcing when I left the company which only makes me wonder what new methods have been developed since then.

[@antoinelavelle8629]

Yo thank you young man I'm studying microelectronics and Nano manufacturing Georgia Institute of technology ( Georgia Tech) And with all due respect because I respect my instructors they have very strong Asian accents and I can't understand a word they're saying in English so I want to thank you because I can relate to the words that you're saying and how you're explaining it without trying to decipher a dialect or accident I really appreciate that I would like to learn more from you I am in a veteran service oriented program that's getting veterans in the workforce to learn how to produce a manufacture semiconductors.

[@ayushbanerjee1187]

Yo representing GT also. I am also an ECE

[@Gideonrex1]

Thanks, man! I’m a chemical engineer graduate hoping to get in the semiconductor industry. Hopefully these videos will help me nail any interviews...once I get one eventually lol

[@jayakrishna023]

Same here! Let me know how was search ? I hope you got into the job you are aiming for.

[@justaman553]

@@jayakrishna023 Good luck guys. I was a mechanical engineer at school and ended up working in the semiconductor industry . I can't stretch enough how amazing and interesting it is to be a semiconductor engineer. let me know if I can help. people often forget how hard or complicated it is to find a job right after college. I went through that so I know. let me know. peace

[@BJB104]

You ever get into the industry?

[@executorarktanis2323]

​@@justaman553 i need help man

[@justaman553]

@@executorarktanis2323 how can I help?

[@somekindofdevil]

These videos are priceless. Thank you.

[@lucasc5622]

These comments are priceless. Thank you.

[@DefaultFlame]

Just found out about your channel. Homemade ICs and a garage scanning electron microscope are the most amazing things I have ever heard of.

[@ArtisanTony]

0:52 LOL, I am a carpenter Jim, not a rocket scientist :) but I appreciate you sharing this knowledge with those who can absorb it :)

[@fllev4121]

My favorite KZbin Channel! Kudos!

[@gulfarn863]

Right now i am trying to learn as much as possible about cpu architecture and design and then i might dive into this and try to make a mikroprocessor ic. I know it will be challening and that there is a great chance i will fail but thanks to people like you it won't be impossible. Thank you!

[@gulfarn863]

@@hey_1232 Its going great! I built Bean Eater's graphics card. Took a part of Onur Mhutlus KZbin lectures on computer architecture and programmed a mips core in verilog. I tried to put it onto an FPGA but it didn't work. Next year I am going to uni to study electrical engineering and eventuelly I will probably major in embedded systems. How are you doing?

[@justtrying1975]

I watch lots of educational videos on you tube, this one is very good with real technical content, simple and well explained. Thanks for doing this.

[@SixOhFive]

Awesome stuff man, love the cleanliness and precision of your work and lab.

[@2morow858]

Hi Mr Sam! could you please make a video about crystal orientation?

[@MichaelOfRohan]

I dont know you from adam. I dont know how old you are, or how you got here. I do know your name, that one was given. But I am so damned proud of what youve succeeded in doing here. I dream of microcontroller fab stations getting popular, like how 3d printing has gotten.

[@lesliefanny]

Hello!! thanks a lot for this awesome video! I just have one question! in the minute 7:10 how do you get rid of the hidrophiyc layer that you have created before exactly? is this when you apply the mask for the litography? thanks!!

[@Crykir]

i supose he heats the surface until no water is present, should be about 100c to 200c.

[@zinckensteel]

Would it be possible (if far from ideal) to get a functioning transistor structure by a shorter process, perhaps skipping the field oxide, then etching in HF, followed by pairs of tiny dots of phosphorus dopant solution, as close as you can get them without significant overlap of the diffusion regions formed after baking/cleaning/driving in, followed by a single stage of oxide growth for the gate, and then your normal process of making electrical connections? Apologies if the needed information is presented somewhere in this video - I didn't watch the whole thing. Looking forward to seeing your creations.

[@SamZeloof]

So you're saying to do doping without a diffusion mask? It's possible. You would have to be very careful with the wafer before pre deposition because the dots of phosphorus solution could slide around. I believe you could make a device like a jfet this way. Interesting.

[@zinckensteel]

And a jfet doesn't need gate oxide, right? That simplifies the oven requirements, I imagine. It looks like there might be (at least) two ways to make a jfet here, depending on how deeply you can diffuse dopants, and the minimum size of substrate you're able to handle. Maybe try a ring-shaped application of dopant for the gate, then make drain connection in the center of the ring, and the source connection around the outer perimeter. Or diffuse from both sides of a thin chip, to get a pinched region of p-type between two n-type wells. Now the question is, do you connect the n wells together and use them as a gate, or do we use the substrate as a base and the n wells as emitter and collector? :-)

[@SamZeloof]

Right, you could fabricate a JFET without any oxide growth I guess. I think using the substrate as a base and n wells as collector and emitter may actually work quite well. I'll give it a try maybe this weekend, I got a new tube furnace last week that I want to try out.

[@gibsonhan1323]

You are a legend my friend.

[@luketorpedo]

I'm a little confused by the metal deposition stage to link the individual elements on the wafer together. Can multiple transistors be made on the same wafer with this methodology only, as what is to prevent any connection on the P type area interacting to any connection on the N type and vice versa? Or is the intention to create individual discrete devices on their own wafers and link by wires to other wafers, in a hybrid integrated circuit kind of way? Very interesting stuff, Just looking forward to it taking that leap to a complex multi transistor function!

[@KonradTheWizzard]

The whole point of ICs is to make many transistors on the same wafer. You prevent unintentional linking of the base by separating transistors with pure silicon zones, by making sure the distance between them has enough resistance to not matter or by putting n-type and p-type material in a pattern that ensures those form diodes that are blocking in the operating range of the transistor. Or you simply put transistors together for which a connected base does not matter. I'm pretty sure an actual IC engineer could give you more methods and better descriptions. The wires that you imagine are the metallization layers he talks about and are deposited directly onto the wafer. (Base=substrate = either the base of a BJT or the hidden part of the MOSFET that contains the opposing pole to the gate to form an electric field (the F in MOSFET). In the MOSFET symbol it is the little thingy with the arrow that inexplicably is missing from schematics that show chip-level MOSFETs.)

[@MrConnor128]

Absolutely amazing video and explanation. Thank you!!

[@excitedbox5705]

what about streaming oxygen into the furnace for the oxidation process? I am thinking if you get an o2 concentrator and use the lowest setting to stream maybe 20-30% oxygen over the wafer. You obviously don't want it to catch fire but the increased oxygen should help.

[@killerjoke56]

why are the oxides over the n+ doped regions less dense? Also, I'm not really understanding why the oxide grows faster for the N+ doped regions.

[@TheLindarella2u]

Dude, what a good avoidance of the word "Piranha" for removal of negative resist! That stuff is Nasty!

[@KonradTheWizzard]

HF is worse. Granted Piranha has higher etching power, but you know when you got some on yor skin - it hurts and you instinctively get it washed off. HF (esp. the slightly diluted version) gets onto your skin, does not hurt initially and you only notice when it is too late to save that limb when you are in howling pain because it is eating your bones.

[@omsingharjit]

After reading lots of Wikki as my intrest i found ZnO is very Magic Si Cond matrial as it's very Easy to make and work with . To make n and p type simi con , Piezoelectric , Electroluminescent , led Lasers , sensors etc Can you make related video ?

[@BeeBaux]

Thanks Sam for putting it

[@manumohan1986]

Great work bro. Precise explanation really helped. Thank you.

[@rock3tcatU233]

How does the resist in an e-beam lithography process not "gas out" in a vacuum?

[@SamZeloof]

It has been pre baked on a hot plate to drive off all volatile solvents

[@johnhodgson4216]

This is Amazing!

[@emiliocolon3245]

greetings from Mexico, could you tell me what was the bibliography you used to make the Integrated Circuit

[@ObstTube]

Thank you very much! Awesome vid.

[@omsingharjit]

Is it Difficult to build basic Si Mi conductor Laser diode For you ?

[@AyranGabriel]

Pls, how i can traslate this video for people can watch in my language?

[@monad_tcp]

What happened to the video 7 ? can you upload it to censor-free site, like x-videos ? or was it just a bad upload?

[@_azaad_]

Which video?

[@behnamkiani3736]

God bless you,with your video my dreams come true

[@codesurge6838]

U r a legend!

[@tousifahmed871]

Can you do a video on your lab setup, cost, etc.?

[@mmughal]

Hi before i start this, do can you please link the other two in the description below

[@nazigermany1945]

In Vietnam, we can buy Hydrofluroic acid (HF 55%) without any license. as easy as buying a Coca-Cola can

[@responsiblestudent7938]

Same here in Pakistan

[@gvl610]

Where to buy please?

[@nazigermany1945]

@@gvl610 bạn cũng đam mê tìm hiểu về bán dẫn à ?

[@nazigermany1945]

@@gvl610 tôi chưa thấy một đồng hương nào trước giờ !

[@gvl610]

@@nazigermany1945 vn cần dạy thứ này

[@sandeepchaturvedi3232]

Did you try your hand with GaAs MESFET or HBT

[@jdcd001]

Where to buy all your equipments?

[@John-vl6hg]

I wonder why you would pump in H20 and not pure Oxygen for growing the oxide?

[@John-vl6hg]

Also why nitrogen rich for doping?

[@possible-realities]

@@John-vl6hg The important thing is probably to make it oxygen poor, to avoid unnecessary oxidation during the diffusion step. I would guess that nitrogen gas is the easiest way to do that.

[@possible-realities]

He said so in the video, because hydroxyl ions diffuse faster into the silicon than O2 does.

[@slabaugh5]

I install the semiconductor tools in cleanrooms, but I’m trying to break through and climb through the company. Are there any courses you, or anyone reading this, recommend to get me a better understanding (complete beginner) into mechatronics/robotics? My goal is to completely know and understand every step and part of the semiconductor tools.

[@davidrave563]

well I imagine you would need to look into taking some electrical engineering courses, or some kind of technician course to be able to troubleshoot these devices. What exactly do mean by climbing through the company? your coworkers and supervisors might be able to point you in the right direction and type of training you would need for a better position. what kind of tools do you work with if you dont mind me asking? if youre interested learning DIY robotics info, YT has many great communities for this, my favs are Dronebot Workshop and GreatScott!, they work with devices with picos and arduino controllers, but I imagine this is at a lower level than what your interested in.

[@KatwereJames]

Thanks for the love it, no one does this for sure.

[@MichaelWaisJr]

Katwere James I’m doing it.

[@ReagueOfRegends]

40:00 very similar to quadrupole mass spectrometers

[@SciHeartJourney]

I want to make a heavy duty power FET!

[@houssemouali6720]

Amazing !! god bless you!

[@TheMadisonHang]

Do you not work for an electric company? Or ehat have you 😮

[@ethanmye-rs]

What's your background? HS, college, post grad etc?

[@SamZeloof]

HS

[@ethanmye-rs]

Sam Zeloof me too actually. I'm jealous -- having access to all that equipment must be nice!

[@zinckensteel]

I was more in love with arcs, sparks, plasma, and vacuum tubes back in HS, but nonetheless the tinkering was an excellent foundation for eventual professional work. You have my admiration; you're operating at a level well above most folks.

[@SamZeloof]

Thanks, I appreciate it! Arcs and plasmas are still awesome...

[@SciHeartJourney]

Does it matter?

[@SciHeartJourney]

Yay, NOR gate! All logic can be made from NOR gates!

[@ScienceDiscoverer]

NAND2TETRIS!

[@shridharhegde5760]

Thank you very much.

[@williamtarabalka4920]

Your into some coll stuff.

[@estuardorecargador2495]

Please update the wiki🙄

[@mmomsboy]

When working in a factory environment with these processes which chemicals seem to easily cause a blood cancer?

[@anandbhushan7709]

I submitted a request on your blog. Looking forward to get your reply.

[@TO-jd8rg]

I cannot find a single vid about LIVA or TIVA.

[@magnuswootton6181]

I need the easiest possible way to make a diode. (not just one, i need thousands of them.)

[@TheMadisonHang]

Chemical engineering ftw

[@lukehill6272]

Hopefully someone with the expertise and knowledge can help me here I cannot find my answers anywhere... I have been tasked to produce a power point on LED manufacturing and its requirement for vacuum. id be forever grateful for any the answers to any of these questions I've watched endless videos on KZbin. Thanks in advance. The questions are: 1)A wide variety of systems used in LED manufacture require a certain level of vacuum pressure, what is the necessity for vacuum in this environment? 2) what are the various methods of achieving low and high vacuum pressure (pumps) 3) what measurement systems are used to indicate the various ranges (gauges)

[@KonradTheWizzard]

It's probably too late for your assignment, but anyway: 1) usually for contamination during the process. CVD, PVD and plasma etching need vacuum to keep the plasma clean and at the exact right pressure to get an ideal reactivity - not too fast or too slow. The air would contaminate the plasma and add reactions that you don't want (e.g. oxydation from the O2). Implantation needs the ion stream to be clean and unobstructed - air would collide with the beam and make it unusable by dispersing and contaminating it. There are some measurement processes that need vacuum because air would change the results. There are a few more, but those are the main vacuum processes. 2) Low: turbo pumps. High: cryo pumps. Look it up on Wikipedia - cryo pumps are fascinating... ;-) 3) no idea. I just look at the screen and act suitably impressed. Sorry, I just do automation. ;-)

[@stevenriofrio7963]

Please subtitles in spanish :(

[@treemanzoneskullyajan711]

I is da daddy when it comes to doping at home init