Metallization: Making Conductive Traces on Silicon Chips. 

Actual technical aspects aside, I was much more impressed and entertained by the critical thinking and problem solving involved in this one. It's very satisfying to see you go through all the stages of hypothesis, experimentation and adaptation

As someone who has coated telescope mirrors with precipitated silver, once you finish them, they are never touched with anything other than running water.

The allure of cheap materials is their cost, but the low cost always comes with terrible side effects. You need a proper magnetron if you're gonna do this. Trying to do it without one is like a carpenter that doesn't own a hammer. And if you've got a magnetron, you have a plasma cleaner if you just fill it with oxygen and light it with RF. You can't make chips without a variety of vacuum systems. They're as essential as the furnace. Don't misunderstand, your efforts thus far are awesome. But you're going to hit a wall really really fast IMHO. Keep going, looking forward to seeing how far you make it through this.

I'm commenting before I finish watching the video just because. I make soap and I can very much attest that dissolving NaOH in water produces a *lot* of heat and can even boil the water if you add it too quickly. The "go slow and be careful" is very good advice for anyone not familiar with it.

One thing you could try is electroless copper plating, the process they use to create the initial copper layer in plated throughholes in PCBs (in PCB manufacture they then build up that layer to final thickness through electroplating). It's reasonably quick (from what I can find about 15-20 minutes for a 1 micron layer), works on glass, and DIY-kits for it are available online (although not necessarily cheap).

every time you upload I get excited to see how much progress you've made, can't wait until the day you make your first chip, even if it's something as simple as an And-Gate or something else.

Thermal evaporation wasn't really used in IC fabrication past the 70's to early 80's, as it isn't really that much cheaper or easier than magnetron sputtering, but it does have many unique disadvantages. If you do intend on acquiring dedicated equipment for metal thin film deposition, I would *strongly* suggest going with a magnetron sputtering setup, even when choosing between a DIY bell-jar-based magnetron system and a commercial thermal evaporator. As for the current process, getting an adherent film on a homogeneous, smooth and flat SiO2 surface is great and all, but I fear insurmountable issues lay ahead if you try to use this process to contact exposed silicon through holes in the SiO2 insulation.

I think you can avoid the problem of silver coalescing into small blobs by first alloying the deposited silver with a bit of copper. You can deposit some copper very quickly (you don't need much) with a copper sulphate solution and a DC voltage. As soon as you have some deposition, take it out and bake the chip. Over time, the silver and copper will alloy into each other (fairly quickly as well, since copper in particular is amazing at solid state diffusion), and you'll end up with a layer of some grade of sterling silver instead of pure silver. The sterling silver should allow you to anneal it without it coalescing into blobs and losing all electrical continuity.

I Love videos like this! Creative solutions for normally complex or difficult tasks that normally require extremely specialized & EXPENSIVE machines to accomplish. Said machines are almost always so far outside the realm obtainable to all but laboratories or people with large free capital. It was a great video where you get brought along for the ride through the trials and tribulations. Unnecessary content was removed and only the most fundamental parts were kept which means 100’s upon hundreds of hours of footage had to be edited and broken into distinct stages. Before reviewing and finalizing transitions and adding the audio commentary. If anyone can’t appreciate the skills needed to edit let alone the immense time invested to get a result, then i urge them to try to record just a single one of their own multi day projects to see what actually goes into creating something that keeps people attention.

Just an idea but check out plating plastic with metals videos. Now that you have basic conductive layer using the mirror method, you could use some light electroplating to increase durability significantly on your deposited layer. This would likely create a thicker more continuous layer without require anything to major in material or equipment. The electroplating rub on method that is often used with gold plating and other expensive metals should allow relatively controlled deposited layer that is actually adhered to itself and should be relatively quick and easy to do. I hope this is helpful in improving you design. Can't wait to see what your next version is like, awesome work

This is my favorite Project on RU-vid right now. Thank you so much for your amazing work ❤❤

It's fascinating to watch you solve problem after problem with improvisation, this is a great series. It looks like the acetic acid is working well, but if you're so inclined, you can make a "poor man's" nitric acid by mixing sulfuric acid and potassium or sodium nitrate. Potassium nitrate is commonly available as Spectracide "stump killer."

This made my Thursday. Thank you! It's very motivating to see you persevere through all of the challenges of this project.

This is the kind of amazing content that is so often sorely missing from this site! Thank you so much for the highly detailed info

Loving these videos and your progress, also a huge thank you for sharing the issues you faced and how you solve them rather than skipping straight to the "here's one i prepared earlier" style.

8:45 never underestimate the strength of silver fulminate, even small amounts you can barely see

finnaly i wasted 2 years of my life trying to do exactly this, never finished. thank you

BABE! WAKE UP! NEW PROJECTS IN FLIGHT VIDEO JUST DROPPED!! Man, I look forward to your videos! I'm so invested in seeing you succeed ;)

Honestly I’m amazed you can get Riston to stick to a polished wafer, that is no easy task! In the past I’ve used spray-on negative pcb resists on silicon wafers with good adhesion but they can be a pain to remove, required warm baths of 2M hydroxide. Good luck with the evaporator, in my experience, it’s cheaper and faster to use coiled tungsten baskets as opposed to boats. The boats require 100s of amps whereas a tungsten basket gets just as hot with maybe 50-60amps. They work really well for evaporating chromium because it doesn’t melt it just sublimes at the right temperature, if ur using gold or aluminum then sometimes the boats are your only option. Also as a hobbyist I would definitely shy away from the expensive deposition monitors and just use a dremel to break open a 1-10Mhz crystal oscillator and use a USB oscilloscope to measure the changing in frequency with a small bjt driven amplifier.

Honestly I have just been living this series.im excited to see what your next project so going to be once you finish this one

Loving the progress in these videos!

Im really enjoying this series as I just found it a few days ago. One suggestion which could be of use is to maybe create your own spincoater. This could potentially help with your silver deposition if spun at slow enough speeds using a dropper.

This is what engineering is about! Though I do suspect that you'll end up having trouble down the line with using silver. So probably better to slowly start working on that thermal evaporator or sputtering magnetron sooner rather than later :P

Very cool, step by step finding a way. As always a pleasure to watch, thx for sharing.

I am having one of those very rare events where the algorithm truly picks something and someone worthy to introduce. Great info, amazing work, subbed right away. Thank you for the hardwork.

As a previous commenter noted, the critical thinking and problem solving are very interesting/entertaining but I also enjoy your speaking style, your clear enunciation and intonation. Thanks for sharing your work; I know it takes a lot of effort to research, plan, execute and record, then edit and post these videos. Very nice!

Amazing video! What a showing of skill and determination

I really like your style of narration. good work!

well done, enjoyed the whole process!

Wow! This is amazing progress. I really want to see you make a functional chip. Maybe something simple like a few not gates, or other 7400 logic. I can imagine a future where instead of making just printed PCBs, maker spaces might be building their own self contained ASIC fabs for their projects. This is freaking amazing!! Edit: p.s. This is the first channel I've ever hit the little bell on. I don't want to miss an update!!

This is such an awesome project, can't wait to see more!

Dude keep these coming so good !!!!!!!!

amazing work.

Fantastic! Thank you very much for another incredible video.

This is awesome. You are quickly becoming my favorite channel. For the home lab a tiny vacuum chamber should be plenty and I think if we could make a few custom Mason jar lids with a vacuum adapter and various tools it should allow anyone to do thermal deposition, CVD, PVD, and ALD for under a few hundred $. It would be big enough to fit 2" wafers and old DLP projector optics allow projection lithography to be done very cheaply.

Very impressive!

Hell yea, finally. Love the series

WOW. Just WOW I'm amazed with how much you are accomplishing. God I love living in the future!

❤ for your dedication

I've been following you for a few months - Excellent approach and progress. Consider UV Resin as a hard/permanent protective coating!

Absolute legend right there.

congrats man :D I would love to have one of your throw-away failed little Si chiplets, would be neat to have a little piece of what may become *eventually* the go-to method for home labs making their own custom integrated circuits... you could be onto something huge - and all the very best to you my man! :D

wow congrats! what a journey

So enthused to see another video up

Like always a great video. I have to try this at some point.

This is some methodical and metriculous work that brought you to this success, well done and congratulations! Back in the day, we used to selectively create small silver particles directly where it was exposed to light, without extra photo resist (a.k.a. common silver photography). I am wondering if such process could create a useful conductive layer, though I suspect that the particles will be far apart and not connected.

Wow. You are good chemist.

Great work. You are getting close to make your first semiconductor device😊

Dude... This is awesome

thanks for sharing.

Awesome video !!!

Absolutely amazing! I was already considering trying to figure out my own ways of doing these things without a giant lab like Sam Zeelof, and (if i ever do try anything like this lol) this is amazing information, especially since I am quite useless at chemistry...

Fascinating!

Would be neat to see a working circuit using this method. So neat and way cleaner then most chip manufacturing processes. If you give it a coat of UV Epoxy that would effectively make it a durable product that is also quick to apply. Love this.

Thank you for share your efforts and experiences. Pleas make a video on transistor making at home by common available chemicals.

I really liked your conductive paint, so I looked up online, yes! There is a thing called doped silicon ink. You can just "print your own transistor" with doped silicon on paper(?) with a suitable printer.

I've been wondering about silver for hobby IC labs for a while, great to see it in a video. A future problem you might run into with silver interconnect is electromobility; low melting point metals tend to have atoms pushed out of place over time by electrons colliding with them as current flows, altering electrical characteristics, I think it's just a problem with smaller wires though, as aluminum interconnect is widely used in the thicker global interconnect layers and also has a low melting point.

very cool! It might be interesting to try electroless copper plating on the silver traces to increase their robustness instead of the copper paint

As someone has already suggested, I would try to electroplate the silver layer with copper and then bake it as you did. Copper layer might prevent silver from beading, and it might adhere better to the silicone(oxide) substrate. You should be able to make the plated layer even thicker then original silver layer, so conductivity will be better too. Might not work with just silver on silver layer, but you can give it a try. Electroplating should be relatively easy and you can try different metals too, but I believe copper is the easiest to achieve and could work well with silver and heat.

Very cool, keep it going

That's awesome!

Can you use sputtering for depositing a thin layer of metal?

To solve the silver beading problem try annealing the layer with an additional Si or glass chip laying flat on top of the target chip. The surface tension should spread out the Ag layer (the method has been used for growing thin films of metallic single crystals from metallic solutions)

You are insane :) insanely good

I could never be patient and methodical like that. I'd probably lose my mind way before this point.

In commercial mirror making, I think they use wet deposition of tin or chromium as an adhesion layer, or occasionally some metal oxides. Tin (II) Chloride mixed with HCl, if I remember right, but it's annoyingly hard to find any open access articles for it right now. I kinda suspect it will be really hard to avoid using vacuum processes for silver deposition of high enough quality and adhesion for this use case though.

Would it be possible to use the silver anneal beading effect to create patterned discontinuities?

Forgot to mention this to anyone testing electroplating the mirror finish deposit on silicon, would likely require the use of alternative electroplating solution as the ph of tradition electroplating solution would likely be too acid. A simple alternative could be using DES, deep eutectic solvent, which is essential mixing two solid chemicals that make a liquid chemical. Often this process can be done with just mixing the chemical together, might require mild heating. The ph of is often closer to neutral than water based plating solutions. There are a large variety of DES that could be used for this but I would want to use something biodegradable like choline chloride with ethylene glycol, choline chloride with urea or even just urea or choline with metal salt that is deposited, ie silver chloride. DES can be formed using metal halogens as being one of the chemicals. Good luck to anyone trying.

Very cool! Cheers...

Very cool!

Do you think if you physically etched the chip face would the deposited silver be adhered better... using either ceramic, carbide, garnet, or even cut glass as a media? I have a small glass etching "airbrush" made by PASCHE which can do small fine work at lower pressures.

I was a little sad to see the beaker die an energetic death, but then I realized if it's that easy then it's easy enough for me to try.

Thanks!

This is really cool! I was wondering whether it would be possible to use the initial silver plating to then use electroplating for a thicker layer of e.g. copper that would survive the annealing process and better adhere to the silicon surface?

Maybe spray-on photoresist like Positiv 20 could be used to avoid fiddling with the teflon tape. And maybe some thin layer of electroplated copper could improve the durability and conductivity of the silver. It won't improve adhesion though. Pre-etching the SiO2 layer a little bit might help with that by modifying the surface.

Genius!

Is it possible to do microfluidics with the etching technique ?

i may be completely wrong here but would it not be possible to use something like fluorinated tin oxide for the connections or is that process way too complex/incompatible with the rest of the project?

I think electroplating the silver with some other metal *before* annealing would prevent it from beading up. the top layer of metal would stay solid and squish the silver down.

18:40 Nice, you made flexible electronics.

Love all your effort and I am trying to guess how to keep the silver layer stable for the moment the semiconductor heats up.

i wonder if pasive electroplating could improve the sructural strength of the base metal , there are cold tining solutions avalable for normal electronics but it is ostly ment for copper but if it work it could stick together loose silver bits into single metal foil made out of tin and silver. i am not sure if it would stick to silicon by itself and even aplied to silver it would still relay mostly on silver to silicon bonds

Very impressive! At this point would you consider vacuum deposition again?

You can use paint of photo resist

You could try to plate the silver layer with copper or another plating compound

Bro, your glucose concentration is actually 9.09% not 10% not sure if 1% would trow off your results. Thank you for sharing this video it really inspires me to get back in the lab.

OK so rather than a plasma cleaner, UV ozone is a good final cleaning step and dead simple. It will make things more reproducible. Just get a UV lamp with the right envelope so the 185 nm light gets through and produces ozone. Expose the surface to be cleaned to the output of the UV lamp at relatively close range. At ambient temperatures it trivially eats few-nm layers of organics and makes everything pass the water break test with flying colors every time, the water just zips across the surface at first contact, so it's good after wet cleaning. If the surface is heated, you can even eat through um thick layers of organics. Logistical issues: run the lamp in an enclosure, but UV ozone will slowly erode anything organic. A simple setup is to have the electrodes outside of an enclosure, and the silica envelope of the lamp goes through the enclosure. Watch out for some supposedly single ended lamps that aren't really single ended, they just have a PTFE coated wire going to the other end. Also disposing of the ozone can be a pain, but it is decomposed by 254 nm light, so if you get a second non-ozone-producing lamp you can use that to destroy ozone. Putting the setup near a fume extracting device is a good idea regardless. This may be a bit much, but it's a lot cheaper and easier than a plasma cleaner and gives as good or better results for many applications.

8:23 "It just works" -Skyrim man

@ProjectsInFlight What if you put the post-etch post-acetone deposited silver layer into an electroless nickel plating solution for a few minutes to firm it up a little bit (physically and electrically.) I would be interested to know the answer to this one. --Josh S.

I think with silver you're going to run into issues pretty fast as you try to make more complex designs and add more processes onto it. There's some thermal evaporator build series on RU-vid that could give you a good starting point.

Mr Projects, would you like an ACIDIC SILVERING process i stumbled upon? Maybe it won't destroy your photoresist and you could try the lift-off method. You need your silver nitrate, citric acid, disodium EDTA and ascorbic acid, yes, vitamin C. Parts are sensitised with SnCl2 and AgNO3. I've used it to plate things like flowers, plastic dice, and glass of course. You can ask here for deets. You can also check out Brashear's method for silvering telescope mirrors for the SnCl2 thing and other ideas in general, they end up with thicker layers that can be polished using the alkaline silver method.

amazing video bro 😍😍😍😍😍😍😍😍😍😍😍❤❤❤❤❤❤

17:31 I thought you were gonna say..sometimes the best solution..is found in the kitchen

It's not the nitric acid that allows dissolving silver metal but rather the nitrate ion. You can actually add a small amount of potassium nitrate to another acid and it should work just as well as aqua regia. Don't remember where I saw this, but it was another chem youtuber

sick

You could do tests with other metals. I know you can do copper with that one blue crystal. Copper sulfide? Idk. As for cleaning, is there a reason a laser engraver or cleaner wouldnt work? And what materials could you deposit to make transistors?

What about quickly dipping the wafer in molten aluminum and then etch away the excess thickness with HCL or HCL-H2O2 until reach a desire thickness? Another Idea would be to laminate the wafer with aluminum foil and the put in the furnace for sometime until melts.

Hi man, weren't you able to finish the semiconductor device? I was expecting the remaining part of the process I think you do an amazing job here, best regards!

this is a weird combination between surprisingly simple and complex. Stradling the line between, I could do this and I really don't want to do this

What about taking advantage of the conductivity of the poorly sticking silver layer just as a precursor for electroplating? If you have all of the traces temporarily shorted together in the design you could nickel electroplate over the silver for example. You could keep building it up to a fairly robust layer. Hand Tool Rescue uses the method to nickel coat steel tool handles. I have no personal experience or data, it just popped into my head. I also wonder if you could attach super thin metal foil in one corner before you silver plate so that you have a pad to contact electrodes to.