 As I said earlier, we went from one really great talk to another. We've got Rob up here who's going to be talking about nanofabrication and this is going to be a wonderful talk. So please welcome him to the tour camp stage. Hi everybody. Thanks for coming. If you were at tour camp a couple of years ago, you would have seen me on the stage talking about this crazy ancient microscope that I inherited. It's a scanning electron microscope that was made in the 1980s. It was right around 1980. I was alive, but I was not that old in 1980. It had a lot of problems. I talked about all the crazy problems that I had to go through to get this thing working again. Immediately after tour camp, it ran into another really difficult problem, and so that's what I want to tell you about, which is this scope is, can I just see a show of hands? Who here has any experience at all on scanning electron microscopy? See, this is why I love tour camp. My God. It's not just people in my hacker space who have played with this scope. I mean, like there's people out here. So, hands up if you've used a thermionic. Hands up if you know what a thermionic is, maybe we'll start there. Okay. I'll get into the differences here. If you have played with a scope in school, at college, whatever, by far the most common kind of electron emitter is a thermionic emitter. It's basically a light bulb. You have a little loop of tungsten wire, you pass a current through it, and it emits photons and electrons. This scope uses a different kind of technology called an FEG. An FEG is a field emission electron gun. So, in your thermionic, this represents a little loop of tungsten just like a filament and a light bulb. That's a really common kind. Pass a current, it makes a beam of electrons. It makes a cloud of electrons. You use electromagnetics to form it into a beam and accelerate it. But the tighter you can get this little hairpin, the smaller you can make the effective spot. So, there are other kinds of technologies that will make tighter and tighter spots. Another kind of technology is a Lab 6, a Lanthanum hexaboride crystal, where you're actually using the edge of a crystal that was grown to emit electrons. Then down at the other end here, we got the FEG. The FEG is a single crystal of tungsten that has been etched to be atomically thin at the end to provide the smallest possible point and therefore the greatest possible magnification. So, like the difference in technology here is vast. Like this, I take a piece of 0.1 millimeter tungsten. I bend it in a loop. I spot weld it onto an emitter and I'm good. This is a $5,000 piece of machine tungsten. So, my challenge after TorCamp was, my emitter died. So, what am I going to do now with this thing that I've sunk a year into and well, you know, this is what the emitter module looks like when you removed it from the scope. This was a very heart-rending exercise because I knew that as soon as I cracked the copper seal that holds this thing in place, I was in for a very expensive repair if I couldn't do it myself. So, this thing, like to give you a sense of scale, it's about the size of this microphone. The tip here is, it's very smoothly polished. It's an anode. In the thermionic, it's called the Venelt. You have, let me just go back for a second. You have all of these basically work in the same kind of principle. You have something that makes electrons happen and then you have this Venelt, this surrounding piece and there's a high voltage potential between the two. This whole thing is in a vacuum and an acceleration happens between these two points so the electrons move down the column. And, you know, when I say column, the column is actually quite tall. In here, the FEG is up here at the top so there's a little tiny point. For an FEG, you need to be at ultra high vacuum so this is a 10 to the minus nine tour which is very difficult to achieve. I'm not gonna get into all of that. That was the last talk. This talk is about what the hell do you do when the thing is broken? So you have your electrons are emitted here. They're accelerated down a tube. There are very electromagnetic coils that steer and focus the beam. Your sample is down here. The electrons hit the sample, they reflect off into sensors and then that's what turns into your picture. So what I had to do first was rip this thing out so you've got, this is all porcelain. This is a razor sharp edge that is a knife edge that's gonna dig into a copper gasket. That is the only thing that's strong enough to withstand the atmospheric pressure to keep you at ultra high vacuum. The whole system has to be baked to remove the water vapor and extraneous air molecules out of the system. It's really difficult to get the vacuum back once you've cracked this. So once it was open like, well, I got nothing to lose now. Let's keep digging. Let's see what's inside of it. So if I take the top of the vent out off, so if I unscrew this thing, what you end up with is this. So this is a little loop of tungsten wire that's a 10th of a millimeter thick and on top of it is this little tiny stick here and on top of that there's a little tiny point and that point is one or two atoms of tungsten thick if it were a good emitter. This one is a broken emitter. So just to give you a sense of scale, that's how big it is. We've got eight millimeters across. We've got a 10th of a millimeter here. We've got one millimeter here and then it goes down to, well, it goes down to 500 picometers on the end of that. Yeah, so half a nanometer is what we're trying to go for. I called a couple of different places saying, hi, will you take pity on me? I really need one of these modules. They don't really make cold cathode FEGs anymore. They're kind of finicky and it's a lot easier to use the lab sixes because generally you're not going down to 600,000 X magnification on a day-to-day kind of basis and you want your grad students to be able to just turn the machine on and do the thing. So lab six will get you down to 50, maybe even 100,000 X which is more than enough for most people's application and it's not as finicky. But me, I wanna push it as far as I could possibly go. So of course I wanna manufacture one of these things. No one would talk to me. They thought I was a terrorist because I was not attached to a university. Like I don't know, it was very difficult to convince people that yes, I am a good person and I just wanna fix my microscope. So I was kind of on my own. So the first thing I had to learn how to do was spot weld tungsten wire. And this was my first attempt. I needed a spot welder, so I made a spot welder. We ripped a microwave oven transformer out of a microwave. I had some help from my shopmates. This used to be thousands of windings of copper and we chiseled the thing out of there and put in some double-ought copper. So this became a step down transformer instead of setting up to 3,000 volts. It steps down to about a volt at many, many, many amps. And we stuck a foot switch on it. It comes out here to a couple of tungsten electrodes and I figured okay, I'll clamp my thing in there and I'll hit it and we'll see what happens. So this is a magnification of those two tungsten electrodes coming together and then this is the loop of wire and this is the little tiny one millimeter piece and that's an infrared and you could see, okay, I'm spot welding tungsten, great. Well, what I was really doing though was I was making a whole lot of tungsten oxide and destroying the piece that I was trying to weld. So I couldn't even start at a tenth of a millimeter. It was way too small. Like that's like human hair-ish, right? So this is a quarter of a millimeter because I thought I would start easy, start with something I can at least see and I couldn't do it. It was terrible. It forms a very brittle weld. This is the original emitter for scale. So you can see it's twice as thick. After the weld, the tungsten would just break. I tried using shield gas because the part of the tungsten oxide was, of course it's getting very hot and the tungsten is reacting with oxygen in the air so I used argon as a shield gas and it was a mess. This was a terrible idea and after a little bit more research it turns out it's because it's completely the wrong approach because I was putting way too much power over much too long of a time on a very small piece. So what I needed was a capacitive discharge welder where you charge up a capacitor bank and then you quickly discharge it. The problem is a CBD is about a $3,500 device that I knew that I would use a couple of times. I called around, I didn't know anybody who actually had one. I looked into actually making one. I know a little bit about capacitors and high voltage. It's a sort of a sideline hobby. But I didn't really wanna design a switch that would be reliable enough within the timings that I needed. It was just too much and besides I could take advantage of eBay. So are there any drone enthusiasts in the audience? I just wanna say personally thank you for making market forces such that I could go onto eBay and find a $100 CBD welder because you need a capacitive discharge in order to weld tabs onto the battery terminals on lithium ion packs because you can't solder them because of course you'll make a lithium fire if you try to do that. So you need a CBD welder. So this thing showed up from mainland China about two weeks after I ordered it and I only had to take it apart twice. The first time I plugged it in it just tripped the breaker. Like okay something is clearly wrong in here and I took it apart and their method of assembly was really interesting. They had a PC board wedged up against of a piece of sheet metal that was just instantly shorting. Like I don't even know how it was supposed to work in the first place. Like what are you guys doing? So I ordered some parts. I repaired the CBD welder. So okay so to be fair it was $100 plus 10 bucks in another week but I got it working. So at that point this little guy has a foot switch and a tiny little anvil I just kind of made and then boom I could spot weld tungsten to tungsten. Okay great. This is a laser cut out like this is to scale that little loop of wire with the one millimeter thing on the top that you might remember. So this was my first ever attempt to spot weld and it totally worked and didn't turn into nasty oxides. So once I had that I realized that I couldn't just spot weld tungsten wire. I had to spot weld a monocrystal of tungsten. Monocrystal and tungsten is available through different scientific markets but it's extremely expensive and it's expensive because it's extremely fragile. The problem is that a monocrystal is brittle. So you can make monocrystal in tungsten and you need a monocrystal let me just take a step back. Tungsten wire you can get it on eBay all day long because people make old fashioned light bulbs out of it and the way that they make it is they center tungsten powder. So you heat up tungsten powder you force it through a dye and so you end up with this kind of like packed granule kind of structure. It's not a single crystal it's lots and lots of little tiny grains all pointed in different directions. So if you try to etch that you're gonna end up with a point that's lots of little granules and little tiny points. It's not gonna work. So I found a paper online for generating for creating your own monocrystal and tungsten in the lab. So what this is, this is a piece of Pyrex tubing that you use for like in a chem lab for making drains out of it's basically a piece of tubing with gaskets on either side. So O-rings got a piece of aluminum on the top and the bottom. I drilled a couple holes and put spark plugs inside and it's important if you're gonna do this at home if you wanna make one of these. There are two kinds of spark plugs those with the EMF limiting resistors in them and those without and we're all in the state supposed to be using the EMF limiting resistors like take those like either remove the resistors or you can find from India basically. Yeah, no resistor plugs. So do that, hook up a variac and away you go. You make effectively a great big light bulb. You hang a weight on the bottom of it. You turn the current up until it's nice and orange and then you let it sit for a couple of hours and slowly that polycrystal and tungsten will melt and relax and then there you go, you got your monocrystal. So you take it out of there and what you end up with is two very sharp, very thin pieces of very brittle tungsten that the first time you pick it up it's going to break and fall onto the ground and you're never gonna see it again. So then you make another one and then yeah, then you got your monocrystal. So after I did that, this is kind of what I ended up with. Rather than trying to make one of these from scratch this is part of my problem. This little ceramic base, this is the only one that I have. It's the working one and I didn't have it together to try to make one of these yet. That's kind of a back burner project at this point. But instead of trying to make the loop what I ended up doing was this is the original emitter and I picked off the little tiny point and then I spot welded my thing on there. So I just sort of recycled it. So once you've got that, this is a picture of a spot weld that I did before I removed the emitter. So this is with the old broken emitter. This is a piece of polycrystal and tungsten spot welded to a monocrystal and so I could inspect the weld and it was quite good. I got good wetting. So I knew I was on the right track. So then how do you make it even sharper? A tenth of a millimeter is nowhere near five orders of magnitude off, six orders of magnitude off. So how are you going to get it sharp? There were lots of people suggested different methods for making it sharp. Some people said just cut it with a sharp pair of tweezers or sharp pair of pliers. That's sharp, I mean it's sharp enough to go into your finger but that's, it makes it worse than when you started because cutters will spread the edge. So it looks kind of flat. You can, somebody suggested pulling it. Just take it and pull it really fast and while that's true, you know, you can get a pinch. You can't, I did a bunch of experiments and you can't get it reliably to a point and even then you're still off by about 10,000 times. So the method that I found that works really well is to electro-etch it in lye. So you make about a three molar solution of an AOH of lye. You, this is just a screw. We've got my little popsicle stick here and a little loop in the point. I just screw the screw down until there's, you know, about two millimeters or so of tungsten hanging down at the bottom of the solution and then you pass a current through it. So you're passing a current through the point into the solution. You got a sacrificial anode of, or sorry, sacrificial cathode of carbon inside the solution. And I found this method, fortunately, there was at least one paper that wasn't behind a paywall that this is a thesis from Anna Sophie Lucier from 2004. And she was writing about, you know, she's a grad student and they have similar problems to me. They have no budget. So they need to make tungsten points for various different kinds of machinery, not necessarily for an SEM, but I thought that the method might be applicable and it totally was. So the physical thing that's happening here is when you pass a current, this represents the piece of tungsten and then this is the top of the solution and there's a meniscus and the meniscus wicks up the sides of the tungsten. As you start to pass current through it, it will preferentially etch at the meniscus. That's where all of the action is happening. So as it continues, it forms this kind of a, like an hourglass shape and it's incredible. You can watch it as the current is passing through it. It just gets thinner and thinner and thinner. It's bubbling and you can see it getting really, really thin. And then eventually at some point it becomes so thin that it breaks and in that instant that it breaks, you have atomically thin tungsten on both of these. In fact, some people design special little catchers in the bottom in order to preserve the second point so you can make twice as many of them. I didn't think it was really that necessary. I just made more, but it's really wild to watch this thing work. So the first time I tried to do that, this is what I got. I'm off by about an order of 10 to the fifth, I think. Yeah, I mean that's clearly like a two micron point. This is the end of that monocrystal and it's nice and round and well, what did I do wrong? Well, what I did wrong is that when this drop-off happens I said at the instant that it breaks, you have a monocrystal, but then in the very next instant you're continuing the etching process. So the longer that the current is applied the rounder and rounder this point is gonna get. Which actually makes an argument for making the catcher because this one stops etching because there's no current. But luckily the timings aren't really that bad. You really just need to turn off the current within a couple of hundred nanoseconds or something. But fortunately a 555 timer will do that job for you if you make your selections of RCs carefully enough. So that's exactly what I did. This is from another paper on manufacturing sharp pointy bits of tungsten. This one was from 2000. So all this thing is, I mean I wired this thing up in an hour in my shop. I mean it's a 555 timer, a couple of resistors and a well-chosen capacitor. And what you end up with is a circuit that you can apply anywhere from five to let's say 30 volts DC and the secret sauce in why these things cost $5,000 and are a well-guarded industrial secret are specifically in the choice of voltages and timings and concentrations of sodium hydroxide. So all I had to do was iterate on that a little bit. And there you go. So this was one of my next experiments. It's much sharper. If we zoom in on that a little, we can see now we're at 10,000 X so we're much less than a micron. And keep in mind this is still with the old emitter so I couldn't get really good focus at this point because the emitter was dying as I was trying to do this. And so yeah, at 50,000 X, this is about the best I could go and I messed up meeting I'm probably at about 10 nanometers or maybe smaller which is within an order of magnitude, right? Like maybe two orders of magnitude. I'm getting really close. So here's the problem with the chicken and the egg now. Like I want to make one. I want to inspect it and make sure that it is correct. And then I want to install it and I can't. So I just went for it and I guessed, you know? Like I did it, I made a couple. They looked like I had a process. And so 48 hours later, and it's 48 hours because you remember that ultra high vacuum I mentioned before. So once you have cracked the top off of this microscope, you need to, okay, do your etch, clean it off. How do you clean atomically thin? I'm not even gonna get into it, but you know, there can't be anything on it. And then you reassemble it and then you realign it because the top of that then has a little hole in it and that hole has to be perfectly aligned with the center of the, okay. So I do all that optically and I hope that it's good enough. And then I don't exactly have a procedure for this. I'm just making it up, right? So then you reinstall it, you bolt it all down and then you heat the whole thing up. The whole microscope is wrapped in nichrome wire. So you pass a current through that and you get it to a couple hundred degrees. You bake it, I sleep over because I don't trust that the scope is not gonna burn my shop down while I'm at home, so I'm sleeping on the couch. And then the next day you cool it and because Boyle's Law, we got PVNRT, after the vacuum system is removed enough of the intervening material when it's hot as it cools, the pressure drops down because you're maintaining volume. So if you're lucky and you don't have any breaks in your copper gasket, you will get a very, very, very sharp pull down to UHV. So then you can turn the machine on and see if you get a spot. And hopefully you do, because otherwise you get to repeat all of that. And I got really lucky and it turns out that it worked. So thank you. So you'll notice that this is a very different kind of thing. Like this almost looks like math. I look at this and I think this is like abstract, you know? I mean, this is a beautiful Bezier curve, right? No, this is electro-etched machine tungsten, like someone made this. This is the original emitter. So when I had, remember I told you I recycled the little loop and I popped off the old emitter, I put it on a tab because I wanted to be able to inspect it. And it looks like it goes down into infinity, right? Like it goes down here, down here, down here, and then boop, like there's nothing. And I mean, at this point, we're at 10 micro, if this is a 10 micron line, like you're already smaller than light will let you see. I mean, like even if you had a good optical scope, you couldn't see the tip. There was no way, not with light. So if we zoom down with electrons, there's your problem. That's what all this was about. This was about a, you know, it's about a micron chunk of random gunk and those. So inside the vacuum system, even though it's as clean as you can make it, there's still random crap that floats around. You know, you're hitting an organic sample with an electron beam and bits of it fly off and eventually it finds its way up the beam and then attaches itself to the emitter and then you have to cook the emitter in order to get that crap back off of it. That's why these effigies are finicky because there's this constant interplay between keeping the tip clean and using the tip to image. So this represents random junk, bits of tungsten atoms that have migrated downward over time, you know, and it looks like a little bullfrog on the end of your thing that's supposed to be atomically thin. So now I have a new emitter. Okay, great. So I went kind of nuts this last six months taking pictures and now I have a bunch of them blown up and hanging up on the wall. So that's what I wanted to make the rest of the talk about. I guess I'm doing pretty good on time. So I'm gonna just talk about like, stuff we took pictures of from here on out. So before I go on, are there any technical questions about the manufacturer? Yes. Yes, yes, very good question. Is it possible to recycle this tip and make a pointy again? Not with the meniscus method, unfortunately, because part of the component of making that drop happen is that you have a couple of millimeters of tungsten hanging below the meniscus and there's just not enough material here. You know, the popsicle stick is too short. So it's already about a millimeter and I don't know how the hell you could, I mean, maybe you could get a half a millimeter. No, I don't think you could. I think the meniscus is probably bigger than a half a millimeter even. So I don't think that you have enough hanging down to make the pinch happen, unfortunately. I guess you could start longer. I mean, there's no reason why you couldn't start with a longer tip and then slowly etch your way down. It'd save you a weld. Yes, the question is how sensitive is the tip length? The critical dimension is that the top of the tip must be quite near to the end of the vent out. So as long as you can move the rest of that module up and down, I don't see why you couldn't have a long one and then slowly make it shorter and shorter like a candle. But I did not do that this time. I was trying to go for something as close as possible to the original because I had no idea if it would work. Yes. Is the trigger for when to stop with the 555 circuit? Yes, good question. The trigger for when to stop with the 555. So I didn't point it out earlier, but the 555 senses changes in current through the circuit. So it has its own trigger. You have a little potentiometer that you can pick. So I just chose a voltage that, I chose a position that made it close as close to the drop-off as I could see. And then I looked at a few under the microscope. And then once I saw where it was close enough, that's where I left the potentiometer set. So yeah, you hit a button to turn it on and then it turns off automatically. What does the end of the new tip look like? I wish I could tell you because unfortunately, it's not possible to image it with itself. I am now, now that it's been another year since I did this, it's time to make another one. So after tour cam now, I'm gonna go home and make another one and pull the old tip off and yeah, we'll do this again. So the curve is proportionate to the density of the lie because of the meniscus? Ah, the curve. Ah, I see what you're getting at, but to a degree, yes, the curvature is proportional to, but it's more than just the concentration of the lie. It's the chosen etching voltage is way more important. In fact, this first tip that I showed you here, you might notice that there's two curves here. There's this curve and then it gets a lot sharper and that's because I didn't know which input voltage to use. I was just guessing and this was taking way too long and I was like, okay, this is like, it was a 12 volts I think and I'm like, well, it's just, it's going and going. I'm just gonna crank it up to 15 and as soon as I did it, it was done. So I think the voltage has way more to do. Yes? Three molar, strong or moderate? Fairly strong wouldn't want to get it in your eye. What's the number? Sorry? What's the number? I had a good question. I don't know. If I, at first I chose one molar and it didn't really seem to do much and three molar seemed to be plenty active and I didn't go past that. Yes? Direct tip once you've reinstalled it. How do you direct? The tip. Yes? The tip on. Ah, yes, the question is how do you direct, how do you position it once you've installed it? The module itself, this whole thing sits in a small holder that has set screws that you can finally position. So yeah, you move the whole base. The tip is where the tip is and you hope it's straight up because you can't do anything about that. Yes? What are the methods that they used to make it with a charm hack? Like with triple darts? That's a very good question. I think it's a refinement of this process but I haven't been able to find that out. I don't really know. As far as I can tell, it's a fairly well protected industrial secrets. People generally don't make their own FG tips and when they come from the manufacturer, typically what you're supposed to do is have an arrangement with the manufacturer and they come out and service the machine. So they don't have any incentive to have an aftermarket for components. Yes? Yes, yes, yes. The question is how do you know it's a monocrystal? You can see the difference in imaging. I think I have a picture. This, no actually I don't really, well you can almost sort of see it here. You do it through inspection. I don't have another way to do it except to look at it under the microscope. But the region out here, it's kind of hard to see I know. It looks a little bit more granular but by the time you get out here, it gets a lot smoother and you can see the difference between this kind of rough even though it's blurry versus this. This is a very good monocrystal. I do not yet know if I have achieved a very good monocrystal but I know that in the different samples that I've made I have had better and worse examples of it. And you can really tell by the, just by the way that it breaks when you pull on it, the monocrystal is extremely brittle and other bits of it are not as. I don't know how to describe it otherwise. Yeah, one more. I do some of this similar speak myself so I tried to re-crystallize it. But I tried to do that under an argon atmosphere but because it was used for 42 tons of water. Wow. Because I'm trying to build the man uprobes. Same problem supply, welding it to a piece of copper wire to try to get it attached to it. Should I just give up on the argon atmosphere and go for a high vacuum? I did it under vacuum and I had no trouble. Because I stumbled on the same paper and they suggest a high vacuum but I don't have a high vacuum in the kitchen. Yeah, so for a high vacuum, I was probably only at about 10 to the minus four-ish, I would say. Rich off-the-shelf roughing pump kind of thing. Yep, yep. Oh, so you think that would actually... That's exactly what I used. Yeah, I used a roughing pump with large gaskets. So I think I was having problems with mine. So I think you had an awesome idea of hanging in it by that. Yeah, the weight makes a big difference. I think that was... Yeah, yeah, I think 10 to the minus four, minus five seems to be plenty for it. I let it glow for a couple of hours and then I crank the voltage until it broke. Yeah, most of my time. Forget the argon. You have insufficient vacuum. That's your problem. We're going to vacuum system. All right, I'm gonna move on to some of the imagery now and we'll see if there's more time for questions at the end. I wanted to start with something that I thought that you all would appreciate. This is the insides of a Yuba key, the security key. This is not the Yuba key. This just, I wanted to give you a sense of scale because the problem with imagery is you're gonna be constantly lost here. So unless you know what the scale is, you have no idea what you're looking at. This is a bond wire pad. So I see some people nodding. If you've ever tried to look at a bond wire pad, they're too small to see with your eye, right? Like you need some magnification. They're really, really tiny. This is the pad. So everything that we're gonna look at now is about 5,000 times or 10,000 times smaller than this pad, which is the thing that you can barely see. Okay, just to be clear. So this is what happens if you dissolve a Yuba key in strong acid. They don't like people like me trying to do this kind of thing. And there is a whole shielding layer that's on top of the actual circuitry. So what we're looking at is the discarded remnants of this tattered tapestry of RF and optical shielding to prevent you from looking at reverse engineering the stuff underneath. This was our first attempt at ever doing this. We didn't make a serious RE attempt. This was just like, hey, can we do it? So once you scrape that crap out of the way, you could see individual circuits. I'm not a circuit designer, so I don't know what I'm looking at, but I know it looks capacitive and really interesting. It's storing some state there, I don't know. But there's banks and banks and banks at this stuff. Let me tell you, it goes on like Tron for miles. So this is at 10,000 acts. Once you scrape that top surface away, these are like bus bars. I don't know how else to describe it, but this is like power distribution. So the actual circuits sit on top of this and interconnect different pieces, and some of them are very clearly plus and minus, but I don't know, it just looks like really cool wallpaper. It's amazing. Polysilicon. Yeah, so man-made stuff looks really geometric, and now I'm gonna give you the insect alert. So if you are at any weight, triggered by disturbing pictures of insects, now is your chance to close your eyes because here's a spider. So this guy, he's smiling at you and look it up. This is an optical picture, this is not the SEM. It was Christmas day, and I had gone and gotten a Christmas tree, and it was a cute little tree, but I noticed it was kind of dry that year. I don't know why, we just couldn't really keep it alive, but we're going to a Christmas party, and my wife, Phoebe, says, hey, what's that on your hair? And there's a little spider, I'm like, oh, that's weird. Yeah, we came home from the Christmas party and the tree had hatched, and there were spiders all over the apartment. But they were babies, I don't know that, I don't know, okay, I'm also gonna say I'm not a biologist, I'm not a taxonomist, I'm talking about a whole lot of things I don't know anything about, but I can take pictures of them. So these are little bitty baby spiders. And so this guy, this is actually a focus stack of 20 images, he's only about, I'd put him at about a millimeter and a half tall. So very, very small. And I was impressed at how liquid they are. When they're babies especially, they're like little bags of jelly covered in hairs. That's the best that I could do looking at it under an optical scope, any more than that, and you really couldn't see it. So here's your SEM of happy smiling. Spider, it's kind of dark in here, yeah, I'm sorry, I apologize for the glare. But this guy, here I'll go back one, you can see he's got like, one, two, three, four, five, six, seven, eight little eyes. And you can zoom in on each one of those guys. So this is one, two, three, four, five, six, seven, eight little eyes. This scale, this is at about 2000X or so I would say. But each individual hair has lots of little barbs on it. And like, here's a leg, check out those barbs. Those barbs are on the nanometer scale. These guys operate in a completely different physical universe than you and I are used to interacting in. Could I train them to do my, that's an excellent suggestion. I think I need a spider army of minions, possibly. So if spiders don't squeak you out enough, here's a mosquito. So mosquitoes have a very different kind of, kind of, like if spiders are like the serial killers of the insect world, like these guys are just, I don't know, they're like the berserker, you know. So here's dozens of eyes, right? With little tiny hairs in between the eyes. And the whole thing is very hairy. This is the probiscus here. These were the first couple of images that I took with the new emitter. I was still kind of getting the hang of it. So I'm gonna move on to this and I wish I had taken a better picture of what this actually is. This is another optical picture. It's another focus stack. It's like another dozen pictures or something. So it's quite small. This is a moth's head. And specifically, this is the dividing line in the middle of the head. This is one eye. So a moth has compound eyes. It has two large sort of hemisphere kind of globes of eyes. And I was amazed at what happened next with this because I knew that a compound eye has many cells and each cell interacts with the light in interesting ways because each one is like a little part of a honeycomb. Every one of these little dots is like a part of a honeycomb. So, like a radio telescope array. Keep that in mind as we continue down the rabbit hole here. Yes, it's a lot like an array of antennas. I just didn't realize how much. So if we start with this little box here, this is gonna be this. So we just zoomed in many thousand times. And here are your individual honeycomb. Got your hexagons. And the first thing that I noticed when I got down at this level is I can't focus on it. It's the weirdest thing. And like, I'm not even gonna get into the issues of focusing in FEG. We can talk offline about how this is not a camera. This is an interactive performance art display of a human being desperately trying to make an electron beam interact with a surface. It is not photography. It is something else entirely. It's a lot of fun if you ever wanna come down to the shop. But I could not focus. I could focus on this stuff. You can see this is perfectly sharp. Here's a little hair. Those are disturbing between the cells. Lots of random debris and crap. But the hexes themselves I could not focus on. And it could be because the sample is holding too much charge is generally the problem. And because the beam interferes with itself. You have a negative beam shooting onto a surface that holds a negative charge. You can't hit it. It's gonna deflect the beam. But that wasn't it. I just couldn't focus on it. So if we take this square and we're gonna go down onto that now. Here's what I saw. So I was not expecting this. I thought it would be a surface of a cell. And what it was, it has lots of little cells. Like okay, well that's interesting. I know nothing about insects. But wow, this is really weird. It's like an array of little antennas. So now we're at 10,000 X. So this line represents one micron, which you breathe 10 micron particles and don't even notice it. So these are very small. And so that's clearly much smaller than that. I wanted to point out that these pictures are actually much larger than the screen will let you see. These are 13 megapixel images each. So if we don't even zoom anymore, I'm just going to upscale here so you can see the same number of pixels. Here's what we're looking at. So here's your one micron line, 10,000 X. This stuff over here, I've seen enough stuff to tell you that this is probably grass pollen. It had a little pollen in its eye. But these little things were intriguing. So to get down further magnification, to get sharper, what we're going to do is I'm going to take this little square and we're going to go in. But one thing that you constantly do with the scanning electron microscope is you're sacrificing resolution for zoom. Because the faster you can scan, the deeper you can go because the less energy you're putting on the sample. But if you try to go slow, you're depositing overall more charge onto the sample. So when you go faster, you can go deeper, but the resolution's not going to be as good. So just forewarned, there it is. So now we're at about NTSC resolution. This isn't 13 megapixel anymore. This is all, that's all the dots. But you can see this is a 100 nanometer line where it's 65,000 X down here. And these are tiny little hairs on the tiny little cells on the tiny little head. So we could just go one more just for the fun of it. So if we take that and we go down, now we're at 200,000 X. So we're looking at hairs that are probably maybe 100 nanometers long, maybe 20 nanometers across, and each of those hairs is still smaller than the smallest eye that we know how to manufacture for a single hair on a single cell, on a single eye of a, it's a universe is what I'm trying to get through to you people. Like, I don't know how to express how bizarre the sense of scale through the universe is because like it blew my mind the first time that I had this scope and I figured out how to use it. And I start like zooming down and zooming down and zooming down and focusing and keeping the whole thing quiet and level and turning off the fans and turning off the music and waiting for the trains out back to go away because the slightest breath is going to make the screen do this, you know? And you do that for hours in the dark staring at the screen and then like, okay, I'm done. And you step outside, you look up and then there are the stars. I just broke my brain a little bit and I think I liked it. So, yes, so this was the moth eye. And if you want to see DNA, you've still got to go quite a bit deeper. DNA, we're talking about two nanometers in diameter, about a third of a nanometer between the bases, or you know, roughly on the order of the size of the probe that had to be manufactured to show you this picture. Can moths see ultraviolet? Now here's the thing, I'm not, I know nothing about insects, so I don't know how to answer your question, but I strongly suspect that there's not an optical thing going on here. We are sub, sub, like even UV. So like, they're gonna be seeing some crazy interference patterns. 10 to 400 nanometers? Yeah, so each hair is, okay, okay, yeah, yeah, yeah. Maybe they're into UV, yeah, I don't know. But there's gonna be something strange going on in their brain with individual hairs lighting up as quantum effects are happening on the surface of their eyeball. Oops, sorry, that's the wrong one. There we go, so let's, well, I have a few minutes left. Yeah, I got a few minutes left. Any guesses at what kind of creature this would be? We're back to, is it a blood cell? That's a good question, it looks kinda similar. I'll give you another, I'll give you another example. There's a, looks like it came out of Star Wars kinda down in the crack and pit thing. Sorry, fungus, good question. Is it a spore? Now we're, yeah, we're getting closer. So like I say, without scale, you have no idea what we're looking at, right? This actually turns out to be pollen. And what happens with pollen? It's like a little hard case that holds genetic material, and then you go and be really cruel to it and you stick it in a vacuum. And when that happens, it explodes sometimes. So these are, this is ordinarily pretty smooth, and then this is the chunk of it that blew off and flipped over and, you know, it's the inside of the pollen. But it probably took an hour to figure out what the hell I was looking at. Because you need to know the scale and you need to know what you think you put in there and then there's a very good chance that you're just looking at random contamination on the stub, which is only, you know, five millimeters across to begin with anyway. But it does look really neat, like there's some kind of crystalline thing going on down here. I don't even know what it is. This is probably a spore because there were mushroom spores on the same sample. And that's about the right size. I was amazed at how much mushroom spores, at least this specific species of mushroom look like white blood cells, they're red blood cells, excuse me. They're about the right shape and size. Here's some unbroken pieces of that same pollen. They just like, you know, that is no moon kind of thing. So here's another piece. So this came from a tree. I don't know what kind of tree. I just walked along and then I really collected some pollen. This is a very regular kind of geometric shape. I love the baby pollen. It looks like a little raspberry or something. I don't know. It's a hundred times smaller than the piece of pollen that contains all the genetic material for the tree. And then finally, I'm gonna end on this guy, which is another piece of the same kind of tree pollen. The stuff on the top and bottom, the way that I collected this pollen, there were, it was like a husk that had a whole bunch of pollen in it. And so the brown kind of husk material where the pollen comes out of, that's what the top and bottom is. And so there were, you know, thousands of pieces of pollen inside of, I mean, scale is relative. So what can I say, it was too small to see. And so that's little bits of husk holding it together wherever the pollen came from. So now I sound like a babbling idiot because like I say, I don't know what I'm looking at. I made a probe that can show you this stuff. It's a lot of fun to take pictures of it. And I'm based in Seattle. So if any of y'all have things you would like to see on a very, very, very deep magnification scope, please bring them by and we would be happy to take pictures. That's all I got. So any more questions before we wrap it up? How did you diagnose that your emitter was bad? How did I diagnose that the emitter was bad? There is a, you get a sense of beam stability when you are taking images. Here's what happens, there's a curve that happens. You do this thing called flashing the emitter. You hit a button. The button passes current through that loop. That gets the emitter hot. And when you do that, it ejects all of the crap that's on the emitter, the end of the emitter off. Then you let it cool and you can image with it for hours. It's rock solid and stable. And then eventually you get more crap on the end of the emitter and then you have to flash it again. And usually you can go like a day or two without having to flash it. As it starts to break down, that time gets shorter and shorter. And then eventually you cannot maintain a stable beam and it's really heartbreaking because you'll be down there, you'll be zoomed all the way in. Everything is great and you go to take a slow scan photo and you'll get varying levels of brightness as the emitter intensity is wavering. So that's how it's starting to go. And that's where I'm at now. So I'm gonna make another one. How do you get stuff in without breaking the vacuum? Ah, how do you get stuff in without breaking the vacuum? Good question. We have an airlock. And in fact I have right here. There we go. There it is. This guy actually has two airlocks. So there is a gradation of vacuum. You've got 10 to the minus nine here. We've got 10 to the minus six here separated by a physical valve and then you have 10 to the minus five-ish down here. Inside here you've got an airlock. So you put your sample on the end of this stick that has a seal. You shove that in here and it goes through a cycle that it injects some nitrogen to clear out the moisture and then evacuates all that back out and then you're at 10 to the minus four behind the seal. You stick it in there and then reverse to pull it back out again. And this guy actually has two airlocks because this was originally intended to be an inspection scope for Seagate. So they manufactured silicon. So like this guy, I'll take a whole silicon wafer in the side and you can go and inspect it. But I don't know why you would do that because a single circuit on a silicon wafer would take me months to properly analyze let alone hundreds on a wafer. Like I don't even, I'm not in silicon manufacture so I don't know how they used it. But yeah, they were into looking at lots of things. Any other questions? Oh, one more. Thoughts on desktop SEMs? We should talk offline. Completely different beast. All right, thanks everybody.