Poking Atoms with a VERY Sharp Stick

hey everyone today we're talking about an atomic force microscope probably the coolest piece of equipment that i've ever had the chance to use but before we get into the details about how an afm works just take a moment to appreciate these scans because they are so so cool this is the surface of a silicon wafer that has been partially melted with a laser and so you can see some melt ripples these are some tungsten disulfide micro platelets it's a van der waals material sort of like graphite here's the surface of a precision ground gauge block this is part of an integrated circuit from 1979. and these may or may not be nanodiamonds the jury's still out on this one how cool is that i just uh i can't even i've looked at these images a dozen different times while i was rendering them for this video and preparing and every time i look at them it just brings a smile to my face so i just wanted to tease them up front because it seemed like a crime to wait till the very end of the video to show those because just because so those images were taken on an atomic force microscope and that's the topic of the video today afm which is a member of the scanning probe microscope family where you physically take a probe and scan it across the surface of your substrate and that's what gives you the topographic information so this differs from the confocal that we saw in a prior video in that we're physically touching the surface rather than sending down a stream of photons and then recording those positions so this has a lot of advantages because when you're physically touching the surface you can use a very very small probe to do that contact so it's not uncommon for atomic force microscopes to have probes on the order of 20 to 100 nanometers which is considerably smaller than even ultraviolet or deep ultraviolet light so today i've got not one but two different machines to show you i have the actual atomic force microscope which is this and it has literal nanometer precision and then i have this thing that i built and it doesn't have nanometer precision at all i built this macro afm as a demonstration model to show how the principles of atomic force microscopy work because the real afm has the cantilever and probe on the end of this pcb and you can't really see the cantilever at all even under a high magnification microscope it's really hard to see the little cantilever doing its thing so rather than trying to just explain things by waving my hands in the air i built this macro afm to give us a much larger model to investigate how the cantilever works how it interacts with the surface and what kind of software is used to control a machine like this this macro afm is actually based on a paper from an mit graduate course i believe they build a much more advanced version of this for the course i took the concept of that paper and distilled it down to essentially the most bare bone system possible because i'm not an mit graduate student and this is about all i can get away with the macro afm that i built is cobbled together out of an old 3d printer that i had sitting around so it has probably 150 millimeters cubic volume and about 0.1 millimeter step size that it can accurately hit so 100 microns xyz resolution in contrast the real atomic force microscope has a 20 by 20 micron in x and y and 10 microns in z and get this the x y resolution is 0.5 nanometers while the z resolution is about one nanometer right i mean i just that kind of breaks your brain to think about that all right so let's dive into how this macro afm works and we'll use that to gain an understanding of afm more generally and then we'll go look at the n-gage unit and do some scans on the table well not scans of the table scans while it's sitting on the you know what i mean so you can see here this is my diy macro afm it operates essentially the same as a what's known as a tapping mode atomic force microscope although in reality it's much closer to a prophylometer than an afm just because of the the forces involved here versus an actual afm the cantilever in my device is just a piece of brass shim stock that's been cut to size it has a little tip folded on the end and a magnet glued on top and the cantilever is driven by this voice coil essentially it's just magnet wire that's been wound around a 3d printed mandrel and this is excited at near the resonant frequency of the cantilever which if i remember correctly is about 26 hertz and so the way a tapping mode afm works is it senses the displacement of this cantilever and uses the amplitude of the oscillations to determine when it's in contact with the surface so to demonstrate this principle you can put something in the way of the cantilever so my finger here is now impacting it and the amplitude of the cantilever is changed and if we look at the chart of the data being collected you can see that as i adjust my finger up and down the amplitude that's being collected changes so a tapping mode afm what it does is it monitors this amplitude and tries to get back to a predefined set point if it sees that the amplitude is growing too small then it moves up because that essentially means you're too close to the surface and if the amplitude gets too large it moves down and this up and down movement is how it actually tracks the height of the sample that's under test there's also non-contact afm where the cantilever actually doesn't make contact with the substrate at all and it relies on the attractive and repulsive forces when you're very close to a surface and so van der waals forces and electrostatic forces all these kind of longer distance forces will actually push on the probe and make it oscillate just above the surface there's also contact mode afm where you actually just drag the tip along the surface that's not used very often because it wears out the tip really quickly now there are a lot of different ways that you can sense the displacement of the probe traditionally this is done with a small mirror that's glued onto the end of the tip and you bounce a laser off that mirrored into a photo detector mine uses back emf of the voice coil so what that means essentially is that half the cycle the arduino is driving the voice coil down and then the other half when it's coming back up it's sensing the current that's being induced inside this coil by the magnet and we can use that to determine how much amplitude essentially is in the system the n-gage unit has a piezo-resistive sensor built into the probe itself and so as the cantilever is flexing there's a little resistive device that is giving a similar measurement as this back emf or the laser diode cool so let's see this in action so if i go to the ui i can get everything lined up so now we're roughly in position we can have the afm approach so what that's doing what it's doing now is it found the set point essentially sixty percent of the oscillation and that's where it's going to try to get to so it moves into position you can hear it starting to tap on it a little bit and the chart you can see we get under the set point so we're kind of like roughly in position so now we can start the scan and we'll move over to the corner of the scan and start rastering across the sample and if we watch the chart you can see that the chart is always trying to get back to the set point region as defined as by those little two bars and that's plus or minus four percent of the set point so the way it works essentially is that it'll move to the next xy position check to see if it's within tolerance of the set point if it's not it will move up or down depending on where the amplitude is in relation to that set point and when it finally gets back with intolerance it will record that z location and move on to the next spot now there are some interesting things to note about a scanning probe technique so this tip directly impacts what you can actually scan so if we stop it for a second you can look and see that i tried to make the tip as sharp and tall as possible and the reason for that is the wider the tip or the more of an angle on the tip the more likely you are to hit side features and so in this lattice for example as it goes into the hole if it's a really wide tip you're likely to hit the edges of the hole rather than the bottom of the hole and that'll skew your readings so the geometry of the tip itself is actually very important another thing to note is that because this is a cantilever that's fixed at one end and kind of flexing it's impacting the surface at an angle going through an arc and this has impacts on what kind of geometries you can scan and how different geometries are resolved because you're coming in at an angle depending on which direction you're going and the orientation of a geometry and whether has an overhang or not will impact the type of data that you can gather from it so there's just kind of like unique things about a cantilever system like this now this is obviously very exaggerated because this is i don't know maybe traveling five or six millimeters in its arc whereas a cantilever that's a few microns it's a very different travel so different considerations when you get smaller but it's the same idea and there are different ways that afn manufacturers deal with this by pre-angling the tip and having kind of compensation control inside the software and while i built this macro afm just as a demonstration to show kind of the cantilever and how it interacts with the substrate the scans actually come out pretty good so here's a scan of that lattice you can see that it basically looks like the lattice there's the hole in the middle all the features line up the deepest part is about four millimeters the individual kind of crevices between the lattice is about two millimeters and yeah it's it's a shockingly good system for you know as janky as it looks and is all right so the machine itself if you're interested is an n gauge atomic force microscope from a company called icspibe and it caught my eye because traditional atomic force microscopes are not this small right this is a very small unit whereas traditional ones are typically the size of kind of a small table and the reason for that is twofold first you need very precise positioning of the probe right if you're working in a 20 by 20 micron volume and you've got 0.5 nanometer resolution you need to stick that probe you know within 0.5 nanometers of where you tell it to go so that's typically done with a piezo actuator to move the sample around underneath the probe and then traditional afms also usually have a laser that bounces off a mirror that's on the cantilever and all the optics associated with it and all of that adds mass right and so the more mass you have in the system the lower the resonant frequency and this is important because low frequency vibrations are things like someone walking on the floor above you or a car going by outside and these low frequency vibrations are really hard to damp out usually and so you need a big isolation system to protect the device from those vibrations so typically you'll have a machine that's i don't know like a lunch box two lunchbox size like not terribly large but then it's sitting on a big 300 pound granite slab and that's sitting on top of a vibration isolation table and the whole thing's usually shrouded in some kind of shield so they can be quite large just to house this really tiny little probe so when i was scrolling through vendors and i saw this it really caught my eye because i don't have space in my lab for a big afm but this sits on my bench now you might be wondering how can this unit be so small if all the other afms are so large like what about this allows it to be small and compact without a big vibration isolation table so the answer goes back to these probes that i showed earlier that are on the end of these little pcbs instead of the big optical setup bouncing a laser off the probe and using piezo actuators to move things around all of the hard work has been baked into a mems device that's attached to the end of this pcb so the mems device actually does the translation of the probe in x and y and z and there's a little piezo resistive sensor that measures the displacement of the probe on the mems chip itself and so essentially it's all been miniaturized down onto the chip which means it's super super lightweight now and the rest of the device can also be lightweight so long story short i shot him an email and we got chatting and we came to an agreement where i'll be filming some product videos for their website in an exchange they sent me this unit to play with give them feedback on and showcase it on the channel for my various projects so it's not really a sponsored video and that i'm actually doing work for them in exchange but it's a little unorthodox payment so i want to just be fully transparent about that up front okay so here is the n-gage atomic force microscope you've got a thumb screw here which mounts the chip so the chip goes there and then there's a stage with two micrometer thumb screws to kind of give you some adjustment there's a small stepper motor on the side which raises and lowers the stage so to use this you pull one of the chips out of the case and mount it to the thumb screw which is relatively straightforward you do want to be careful because the tip is very fragile so you don't want to drop it like so once it's in place put the thumb screw back on tighten that down and now we're ready to go alright so at this point we need to enter the software and do a frequency sweep so each cantilever has a specific frequency that it resonates at and it's a little different just based on manufacturing tolerances so the software will sweep through a range of frequencies to find the one that it best resonates at and so in this case this particular cantilever is around 9 000 hertz which is obviously quite a bit higher than the 26 hertz that mine did from here we are ready to approach and start scanning a sample so today we'll be scanning a tungsten carbide insert this just goes to one of my milling tools and you place it underneath the probe carefully right about there normally this would be sitting on my desk where i have an inspection microscope above it which makes alignment quite a bit easier but for a large object like this you can do it pretty easily by eye and then you tell it to approach this will move the stage up into contact with the probe the first time it did this i about had a heart attack because it looks like it's going to crash but it knows when to stop and you'll see on the amplitude chart the amplitude starts getting smaller until it locks in and right now we are in contact with this sample so it's actually tapping on the surface at this point for our scan we're going to take a 256 by 256 image which is kind of just a quick scan since we're filming this live and 20 micron by 20 micron area and spending 500 microseconds at each point so we go ahead and hit scan and we'll start collecting data now i'm not overly familiar with what tungsten carbide looks like under an afm i've only scanned it a few times it is a sintered material so you kind of expect it to be kind of glompy with big particles and that's generally what i've seen when i've scanned different parts of this particular insert yes if you look at the data coming in you can see that let's see we've got 2.3 microns between the lowest point and the highest point so the lowest of the moments over here and then the high points are at the top of these kind of particles down here the line profile is showing you the z height of the most recently scanned line and then there's two lines here it shows you the the forward scan and the reverse scan and ideally you want those to match up because that lets you know that the pi controller is working correctly but yeah that's that's pretty much it i mean i'm basically entirely new to afm and you know you can get a scan out of this pretty quickly which is remarkable in my opinion at 20 by 20 microns which just is what breaks your brain when you when you think about it okay so that is what the back side of a tungsten carbide insert looks like so let's do another sample so this is a piece of glass that i initially deposited a thin layer of aluminum on and then ablated away a pattern and i noticed while looking at it under a microscope there's kind of this cool diffraction color to it which made me think that it wasn't fully ablated in the regions and in fact when you look at it under the afm you can tell that much of the metal didn't get ablated or was like recast back onto the surface do another 20 micron scan yes you can see the dark areas here this is the glass which is pretty flat even though it's just a cheap slide it has a surface roughness of you know 50 100 nanometers but then you have these blobs and this is actually the recast aluminum and i can't quite tell i think it's actually material that was ablated away and then was essentially ejected from the surface and then re-lands as molten material and balls up because if you look at a section that has the thin film right next to the recast layer the thin film is quite a bit more smooth and flatter so i think this is really kind of ejecta that lands back on the surface which is pretty cool and it's relatively spherical in nature which is also pretty cool you can see the highest point it's probably about there and there is 1400 nanometers so 1.4 microns [Music] it is just pretty cool and this also explains why there's kind of that diffraction coloring effect happening because the distances between these two points like if we load it up in gwyddion we can see it a little better but it's i don't know if i had to guess that's maybe seven or eight microns which is kind of on the range of being able to get some of that diffraction coloring effect all right so here is a gauge block which should be pretty flat and smooth but you can see some grinding marks so you can see there's a lot of texture here from the grinding but the absolute height the difference is only 400 nanometers from highest to lowest so it's actually pretty flat other than that little spot there's 600 and that's probably just some dust or dirt on the surface but the actual grinding marks are pretty shallow you know that might actually be on the gauge block it kind of looks like it's part of this that little spot and i might i wonder if that's like a particle from the grinding process either the actual grinding material itself or maybe just some metal that got pushed up while it was being ground that's cool so i think you can tell from the excitement of my voice during this whole video that i am just pumped to have this machine in my shop i think it's going to open up a lot of really cool projects that i really wouldn't have been able to do otherwise so the nano diamond image that you saw for example i did that a couple months ago and i was pretty sure i replicated the paper because it was a pretty easy protocol but then i was looking at the glass slide and realized i just had no way to actually quantify if i had made nanodiamonds or not because it's a thin layer of literally nano diamonds how are you going to quantify that with things you have around the shop so it's projects like that that i just had hit pause on because i didn't have a way to actually measure it which are now completely unlocked because of this machine so i think there's gonna be a lot of really cool stuff coming down the pipeline now that i've got this capability and i'm excited to see what comes from it big thanks to the folks at icspy for making all of this possible they've been super great the machine is obviously wonderful and i just really appreciate the flexibility that they showed in getting everything up and running i'm not quite sure the timeline but probably within the next month or two you'll start to see some of the videos that i'm filming for them show up on their website so check it out should be pretty cool if you feel like supporting the projects on this channel financially i did just start a patreon i'm not really sure what the rewards will look like it's a work in progress probably some combination of behind the scenes footage stuff like the green screen editing for the blade runner intro or afm images that haven't made it on the channel just because i've been scanning literally everything in my shop and stuff like that but honestly at this point subscriptions probably mean more to me and the success of this channel than financial support so if you're not subscribed and you liked today's episode please consider subscribing i'd really appreciate it okay cool i think that's all i got for you today thanks for watching i'll see you next time

Channel: Breaking Taps

Views: 394,567

Keywords: afm, atomic force microscope, ngauge, icspi, macro-afm, microscopy, scanning probe microscopy, spm, atomic force microscopy, mems, probe, atomic force microscope how it works, atomic force microscopy images, atomic force microscopy explained, atomic force microscopy principle, tapping mode, non-contact mode, scanning probe techniques, scanning probe microscope magnification, gwyddion, sem, scanning electron microscope, micrograph monday, afm microscope, scanning electron microscopy

Id: 2Kv6KwADn7Q

Length: 22min 39sec (1359 seconds)

Published: Sat Jun 05 2021