Demo and teardown of an X-ray fluorescence gun (measures chemical composition)

today on applied science i'd like to show you this x-ray fluorescence gun not only does this thing look like a sci-fi prop it actually works like one too you just aim it at something and pull the trigger and it even makes this fanciful kind of ray gun noise and what's happening is it's emitting x-rays and then examining the reflection from the metal there and after 10 seconds it's captured enough of these reflected x-rays and it gives us a complete chemical breakdown of what it was just looking at down to a fraction of a percent and then it uses all this information to figure out what the metal alloy is with very high accuracy so you can see here this is stainless steel 304 and it even measures things down to hundredths of a percent level so i'm molebum 0.03 percent even so in today's video i'm going to talk about how this thing works uh some different tricks you can do with it i actually learned quite a few things about x-ray fluorescence that i didn't know before and then i have an identical model to this gun that doesn't work and so we're going to do a tear down on this one and a failure analysis and you'll see there's a really unusual power supply design and then i have a funny story about trying to buy one of these on ebay and being rejected by the seller once he found out that i was going to show it on youtube so let's see how these things work let's talk about what this machine actually emits it's true that you wouldn't want to turn this on and you know take a nap next to it because it does emit x-rays but to put this in perspective this is about 10 or 100 000 times less powerful than a typical medical imaging x-ray tube so i've covered up this proximity sensor on the end here so that this thing will emit x-rays even when it's not butted up against an object and i'm going to aim it at my mighty ohm geiger counter there so you can see it does trigger the geiger counter but it's actually not as intense as even some radiation check samples that you can get this software is configured to emit seven micro amps at 40 kilovolts so again that's you know 10 or 100 000 times less than a medical imaging tube i'm going to try to set this up in a dark box here with a fluorescent screen so you can get a sense of how big the x-ray spot is okay so x-rays are emitted by this thing but then what happens how does it work well if we aim these x-rays at some material that's made of atoms here's the nucleus with neutrons and protons and we've got electron shells going around the old-fashioned way of naming these shells is k l m but this is the same as one two three it's just that the x-ray field for some reason uses this older notation and then within each of these shells there's the spd orbitals and stuff that's not going to matter for today's discussion so we've got this you know neutrally charged atom with the same number of electrons distributed through these shells as there are protons in the nucleus and everything's fine but what happens is when we come in with an x-ray remember photons don't have mass but they do have momentum and so when one of these photons just by chance hits one of these electrons in just the right way it actually knocks the electron clean out of the atom and this leaves it in a very unstable state so immediately one of the other electrons from somewhere else in the shells will drop down to this lower level and when this happens it's going from a you know kind of a weakly bound state to a very tightly bound state and when this transition happens it emits another x-ray in return so it's a cool process where we get you know x-rays in and then x-rays out interacting with the matter now the really clever part is that the amount of attraction that this inner electron feels is proportional to the amount of charge in the nucleus so you can see how this thing was able to identify chemical elements we have a photon coming in that could be a lot of different energies knocks the electron out this other electron comes down to here and then the amount of x-ray energy that's coming back is proportional to the number of protons in here in other words the atomic number of that element so we know exactly what element it is you can see there might be a few variations here the electron could come from this outer m shell or it could come from the l shell and so there might be two distinct energies that come out and this is you'll see this later that the peaks are known as k alpha or k gamma and that's just basically uh different energies coming out from different electrons replacing this knocked out electron but the point is that you get this fingerprint of x-ray energies based on what atom it is another cool factor is that this inner electron shell is very heavily influenced by the nucleus but it is not influenced by other chemical bonding that's going on very far outside the atom so this atom could be in a chemical compound it could be in an alloy it could be anywhere and this technique really only focuses on atomic numbers so you have this really like direct line path toward figuring out what chemical you're dealing with with this x-ray fluorescence technique one of the burning questions that i had when researching this is why when you excite something with ultraviolet light the color emitted is not dependent on the atomic number so if we look at the periodic table here i just said that if we do this x-ray fluorescence technique the elements that are very low in atomic number emit very little x-ray energy and as we move down the chart every one of these elements will have more and more x-ray fluorescence energy and that's true because of this relationship with the nucleus now when we excite something with ultraviolet light or with electrons we're really only exciting this outer electron and this might have an energy value to knock one of these other outer electrons out could be you know five electron volts or something not very much and when we do that this outer electron doesn't really feel very much of the attraction from the nucleus because it's shielded by all these other shells of electrons in here so this five electron volt you know needed to get one of these outer electrons out is very not influenced by the atomic number in fact almost not at all so when we look at the chart at first it seems almost like well if this was red you know you'd expect green light here and then blue light down here but it ends up not working that way in fact the atomic number has almost nothing to do with the energy needed to kick one of these outer electrons out if one of these outer electron volts only requires five electron volts that places that you know in the blue light or the ultraviolet range one of these inner electrons requires you know tens of thousands of electron volts which puts it into the x-ray range and so the reason that we need x-rays to do this this you know molecular probing sort of thing to figure out what the atomic number is is just because these inner electrons are so tightly bound it requires these tens of thousands of electron volts of energy to knock one out you can see that there's actually another limitation here if the returning energy is so low that it is absorbed by everything that's between our detector and the thing that's emitting it then we can't effectively detect it and this ends up being a really big practical problem so if we're inputting energy at 40 kilovolts right because i said that the x-ray tube in here operates at 40 kilovolts so let's just say it's a 40 kilo electron volt photon going in if our return is only 500 electron volts let's say it's a very very weak x-ray it's possible that even the air and the the you know construction materials of the detector itself would absorb that photon and we can't detect it so this this gun is you know not a particularly new model and it's only able to detect things that are heavier than somewhere around titanium so if you're looking at a sample of aluminum the gun actually does not detect the aluminum at all but it can sometimes tell you what alloy you're looking at based on the trace element analysis but it actually doesn't see aluminum in fact anything lighter than about titanium is essentially no signal this could be good i mean if you ignore oxygen that's fine because you may not care about oxygen in your sample you're mostly interested in the base metals but keep in mind that the heavier elements are easier to detect because they emit much more strong strong signals and those strong signals penetrate all the foils like for example the window on this thing is made of metal it has this you know shiny metal window so the return x-rays have to make it back through that window and that means they have to be more intense than at least a couple kilo electron volts say one of the biggest commercial uses for these guns is that metal scrap yards because they're so good at identifying alloys so if you want to melt down a bunch of scrap you have to know what you're putting into the furnace and these are really good at figuring that out but i found a couple other interesting applications these are some glass samples that i made in a previous video and if you recall that the way that we color the glass is by putting metal ions in there so if you forget a recipe or you want to know what color is being produced by what ion you can actually aim the gun at one of these glass samples and find out in just 10 seconds the reason that it takes 10 seconds is because it's accumulating all of these return photons and so remember i mentioned that you're putting in this high-energy photon and x-ray photon and you're getting one back if you're getting too many back at the same time we can't measure the energy easily of each one so actually instead of just turning up the power on the x-ray tube that wouldn't necessarily make our job much easier what we need to do is basically set the power level low enough where we get back photons that are essentially separated in time so that we can measure the energy of each one of these photons that's coming back so over the course of the 10 second measurement window we might get you know tens of thousands of these photons back that we're going to measure and instead if we got you know tens of millions of these photons back they would be overlapping with each other and be too hard to pick out what's going on so if we look at the screen we can see that chromium 0.2 percent and that's right chromium is what makes it green if we did the same test for the blue one we'd find out it's copper for that similarly i had this naturally occurring piece of calcite and this has a really cool phosphorescent quality to it where if you aim a ultraviolet laser pointer at this or actually a 405 nanometer laser pointer it has this really cool glowing effect and if you search around on the web you'll find out that calcite by itself is not phosphorescent there's actually an impurity in this naturally occurring crystal that causes it to have this property so sure enough i took the gun and analyzed this calcite sample and as we see it's actually manganese that makes it it gives it this property of phosphorescence tiny amounts i mentioned that the gun also works on compounds and it even works on powders in a plastic container so the plastic is thin enough and absorbs so little x-ray energy that we get a good signal even through this plastic container and um it's a white powder we don't know what it is so we'll wait for the machine to give us some information here let's take a look and as we can see yttrium 96.8 percent and as it turns out yeah it's citrium oxide in here and remember the oxygen doesn't show up in our analysis because it's too light so even though the oxygen atoms are in there and they are emitting an x-ray signal that signal is either getting blocked either by the plastic itself or the detector in this gun is just not sensitive enough to detect that weak x-ray signal from oxygen okay let's see what's inside these magic devices that lets them work i've already taken this one apart and taken all the screws out just so it's easy to pull apart this first thing sticking out is just a heat pipe the sensor that does all this photon counting in the x-ray range is probably needing to be cooled there's no external cooler in here but i suspect there might be a peltier junction inside the detector which is going to heat up pump the heat out into this outer metal thing so then there's this funny heat pipe just to you know get the heat out of there and it actually transfers it up to this metal plate behind the computer and then there's a snout here that just protects the stuff that's underneath it i don't think that filter does much but keep dirt out and remember this is for the proximity sensor so there's just an optical sensor in there and unless there's something opaque pushed up against the front of this it won't activate the x-ray tube and then we can see that there's definitely two things going on here this is the sensor and this is the x-ray emitter side and we can pull that off and interestingly inside here there's a little piece of metal and i suspect that is silver pure silver metal that little square of metal that's in there and what happens is the x-rays come in from the tube at various different energy levels and when they hit the silver the silver emits a very specific energy level so that the x-rays that end up coming out of that little aperture there are mostly caused by the silver the line basically created by silver you can see the detector is actually very small in fact it's self-contained entirely within this you know sort of typical looking semiconductor package of course it has a window on the front and that window i'm sure is very delicate it's probably super thin metal to let in as many x-rays as possible but there's nothing sort of hiding inside here that it really is a very small thing and then on the other side this is the x-ray tube and it also just unplugs if i can get a handle on it here and there we are so the x-ray tube is a traditional construction thing i should point out that this pack here actually here let's open it up it has this very funny expansion bellows here because this whole chamber was filled with oil and so i had to use a syringe and suck out all that oil which is messy and gets everywhere but the reason that it's filled with oil is to prevent arcing so remember this is a 40 kilovolt power supply and it's it's hard to contain 40 kilovolts and you the oil also makes a convenient heat sink so that all the stuff in there is cooled properly so the x-ray tube slides through here and has this o-ring to seal the oil in and plugs into there so let's zoom in on this power supply which is kind of the most interesting bit of this whole device this power supply is built using a bunch of diode and capacitor voltage multipliers and so each one of these boards is basically wired up in series and we start off at essentially ground potential here and by the time we're over to this side it should be about negative 40 kilovolts what's weird here is how it gets power to each one of these voltage multiplier ladders there's a toroid on each one of these boards and that's sort of the supply of the low voltage the challenge is that if you've got this voltage multiplier up here let's say this is going from 30 to 40 kilovolts how do you get your your ground voltage in there i mean you need a transformer of some kind and so this weird white thing is actually just a one turn piece of copper with a big insulator on it and it links together all of those toroids that are on the uh on these voltage multiplier boards kind of a weird design and so there's one toroid on the bottom that energizes this single turn of copper wire and then the single turn basically distributes power into all the little toroids that are on these boards as it works its way up in voltage kind of a weird topology i've never really seen anything quite like that before this thing has to deliver like i say 40 kilovolts at around 10 microamps and so it's not a huge amount of current seems kind of like a weird way to build it but whatever another strange design principle is that it actually is a negative 40 kilovolt supply so typically if you want to do an x-ray tube the way that you make this work is you have to put a filament voltage on the bottom and you would make this zero volts and then you would put plus 40 kilovolts over here the anode and the reason you'd want to do that is because you need to control the voltage going into the filament to control how much emission current you get so the way that you operate an x-ray tube is to just put 40 kilovolts on it and then measure how much current is going into the anode and then raise the temperature of the filament until you get the emission current that you want and typically you would have like an op-amp or a servo control loop you know monitoring this the reason that you'd want the anode to be at 40 kilovolts and the bottom to be at zero volts is so that you can attach your filament control circuit you know near ground i mean this is going to be six volts to control the filament and it'd be convenient to have all this stuff be at ground potential but the designers of this device had a problem the front of the tube has to be physically very close to the snout of this device right like the front of the tube is right up in here crammed up against all this other metal so the requirement it seemed like they decided the requirement was that the front had to be at zero volts which means the back of the tube has to be at negative 40 kilovolts which means that all the filament control circuitry also has to be at negative 40 kilovolts this is going to be a really challenging power supply so then the designers put another one of these single turn transfer capacitor or transfer transformers in here to control the filament remotely without having any galvanic connection with the rest of the circuit so it's it's quite a complicated confusing thing i'm not sure is entirely necessary but it is very interesting to look at it has all kinds of interesting stuff i didn't there's a couple other circuit boards outside the outside the high voltage that does probably some of the filament monitoring so there's potentially a way that this thing measures the amount of current going through the tube it's probably this little purple wire that goes all the way back so basically sense this wire and when it gets up to seven microamps cut off the filament control voltage and if it drops below seven microamps give it some more filament current and that's how it regulates itself um i i did actually find out what caused this whole thing to fail so it it doesn't emit x-rays like the you pull the trigger and the red the red light comes on so the circuit thinks it's emitting x-rays but hold it up to the geiger counter no x-rays come out so that pretty much narrows it down to either a bad power supply or a bad tube and upon closer inspection there's actually a nick in the glass envelope here and when i first took this thing apart and if you can still see it there's actually a few drops of oil in the tube yeah you can see it right there so it normally is supposed to be vacuum inside the x-ray tube and a little bit of a crack there probably due to some mechanical energy input into it sucked in the oil and that was the end of that x-ray tube so i kind of thought about you know maybe it might be somehow possible to fix that probably not and so i think what i'm going to do is take the detector and use it in another project and since that's a pretty cool little device there and i have a feeling that detector might come in quite interesting projects so we'll close with a funny story about me being rejected on ebay years ago there was a publication that showed if you unrolled tape like this in a vacuum it produces x-rays and i did a couple videos on this and my results weren't super conclusive but but it is a real effect the triboelectric charging right at the junction where the adhesive is getting ripped apart will transfer some charge and then as the tape gets unrolled it's basically like a van de graaff generator where you're physically separating the charge and then eventually there's a breakdown the charge slams into you know equalizes itself out on the order of 50 kilovolts can be you know built up this way and you end up with x-rays because now you've got a 50 kilovolt charge with some electrons flying off and making the x-rays for you so the the researchers that found this principle started a company to produce xrf guns based on this principle and as we saw the the power supply is pretty complicated so if you avoid the x-ray tube and the power supply yeah you might be able to build the guns for for less money so this company existed um and i i don't think they went out of business but they're kind of ceasing operations or something but you can still find the guns on ebay and i did and i thought it'd be cool to do a teardown of this commercially produced tape unrolling x-ray x-ray xrf gun so i found one for sale and contacted the seller and said well you know i'm not interested in the gun i just want the cartridge the the the tape unrolling doesn't work forever so the gun has a cartridge that you put into it that has the tape mechanism inside i wanted to buy just the cartridge because that's where the interesting bits are and the seller asked what i was going to do with it if i had one of these guns and needed a replacement cartridge or whatever and i said no no i run a youtube channel i want to do a tear down and show how it works it'd be pretty neat and he absolutely flipped out and said no no no you can't do that i he was actually involved with the company and wants to get it going again and was you know completely scared that i was gonna give away all their secrets or something online now you know mind you this thing is patented and you can download the patent and read exactly how it works and it's it's true that i mean the thing i was curious about was how do you not run out of tape in there is it like continuous loop or what but as it turns out uh there was a little bit of a pivot and instead of actually unrolling tape inside the gun it uses like a silicone pad that is stuck down to a metal target and then it's pulled away suddenly so it's basically like constantly adhesive and then ripping it off so it's just a linear movement just a solenoid or something inside there but i thought it'd be pretty cool maybe if someone has one of these cartridges or kind of wants to buy one or something we can tear it down and see see how it works anyway i hope you found that interesting see you next time bye

Channel: Applied Science

Views: 326,914

Keywords:

Id: KdfHVcU8U7U

Length: 22min 47sec (1367 seconds)

Published: Sun Aug 30 2020