Recording and get into it so today we’ve got the the second part oh and now I have no navigation if that’s up here maybe that wasn’t the best choice um so uh today we have the second part of the atom probe uh of the atom probe lecture and today what we’re going to
Look into is atom probe instrumentation so um typically uh when you go to conferences and you see people presenting adom probe results um very often they uh there is it’s very very much application driven so people don’t necessarily uh concern themselves very much with the limits of the instrumentation the detection system
And so on um but I think it’s very important to understand the instrumentation before we go to the the field of operation physics part so that they know what data we actually get and what the implications are and what we know and what we don’t know because obviously there’s things we uh because
Of the limitations of instrumentation we can’t possibly know and there is things we things we do know um and so today what we’re going to look into is overview of an eron probe what does the instrumentation actually look like um and last week we already talked about our boundary conditions are
That we are sort of have 20 to 80 uh 20 to 80 Kelvin temperature and we have a pressure of around 10 the- 10 to 10 the- 11 milars uh here in elong we even one order of magnitude below that but this is sort of uh the general uh um uh the
General vacuum level we work at um and uh this sort of sets the the the general stage of what the instrument looks like and you’ve already seen the atom probes downstairs and most what you most of what you see is actually the vacuum system and the vacuum and the
Cooling system that really dominates the the appearance of the instrumentation uh and so we’re going to look into UV generation and practice uh so what is important about it and why is ultra high vacuum a bit different than regular vacuum because in we’re talking another about um sort of the pressures we have
But residual contamination is really the important bit then we’re going to talk about detector technology um and very important for us are micr Channel plates and Del lay line detectors because this is essentially what all modern inst atom prop instrumentation uses for detection uh and then we’re going to look into how
Voltage pulsing works and how laser pulsing works and here voltage pulsing is a little bit more we’re going to look a little bit more into that because it evolves involves a little bit more engineering there are more boundary conditions uh and Laser policing is uh a little bit more straightforward from
From the engineering side maybe I’ll think about that differently in about two weeks when we get our first laser that we’ll build ourselves but we’ll see and already have okay then we’ll have to make sure that Mr cine can hear us and we have oh is it working now can you hear us
Now okay good so yeah laser pulsing uh is a challenge we still got ahead of us but uh during the course of this semester you’ll experience life uh how a new laser system is is built okay so this is the the general layout of a local electroid atom probe
And this is really the dominating type of atom prob that’s out there uh there’s been a bunch of uh a bunch of different Acquisitions and merges over time uh before um Kamika which is the current or IMO as it was called at the time was the the dominant cell of adom probes
There were a couple of different ones and at the end of the lecture I’ll go and get a uh adom probe stage so essentially the the the part where the the everything happens from an older adom probe and one from Oxford nanoscience that I have lying around in
My office so that they can see what what things looked like before and this is an adom prop that’s been used for 15 years so definitely works well and is also the one um that the reflect draw so high mass spectral resolution was first implemented in that we’ll talk about at
The end of the lecture so this is one of the early ones 2004 and the reason why I took that one is because it’s the simplest sketch that’s possible and the part that we have is essentially a three-stage vacuum system so we have a load lock chamber where the pressure goes from one
Bar down to uh about down to about 10 the- 6 10- 7 10- 8 mbar yeah so you put your specimen in there and then it goes from ambient pressure to round about this pressure then we have a buffer chamber it can then move your specimens into the
Buffer chamber where it reside at the pressure of about of about 10 to the- 9 Ms again that’s instrument specific on the leaps is about 10^ the- 9 so our home build wants it about 10us 10 uh the main purpose of having this is to facilitate a continuous movement of
Specimens so you want to utilize your instrument continuously so to in order to have a continuous movement of specimens you have a buffer chamber so they can move some out and some in without always having to go to air because this going to from 1 bar to 10us
8 will take at least about 3 hours least about 3 hours but realistically 12 plus hours until we’re at a bottom pressure and what happens here is something you’re going to we’re going to talk about in the uh later in the lecture is a disor of water vapor
Yeah so if you open something to Atmosphere a monol layer of water vapor will form on the surface and this monol layer will slowly evaporate if we heat up to high temperature it will evaporate faster but at room temperature it can take take a little while and then we get
To 10 – 8 m here 10- 9 M uh and in the analysis chamber we have 10 to Theus 11 mbar roughly and so we have different pumps here we’ve got turbo pumps which you can see here this is a turbo pump here you don’t see it underneath is the
Same turbo pump oh no sorry it’s not underneath it’s back here the same turbo pump and here we’ve got an iron gter pump we’ll also talk about that one so there a couple of different ways to set up a vacuum system to achieve that here’s an iron G pump uh it’s big and
Heavy but very powerful um and uh it can pump any kind of contamination you have in our systems we use non-evaporative G pumps we’ll talk about those as well they have the upside that they’re much smaller for the same amount of pumping power um but they they um they pump
Relatively selectively so you need to keep your system a little bit cleaner in order to use them but it makes everything a lot lighter here we’ve got the detector the detector is fled on here and the specimen sit somewhere in here and the iron fly fly towards here
And then we’ve got a detector here with the electronic so here we’ve got the uh the various attachments for uh the delay L detector high voltage power supply and so on and we’ll talk about that later on but this is sort of the basic layout that pretty much most systems will
Follow oh and one thing I forgot here we’ve got the cryostat so here we’ve got a gford MC Mayan cryostat we’ll talk a little bit about what those cry stats are as well so essentially it’s a closed cycle helium system helium piston system that allows us to cool our specimens
Down to to the temperature we want them to so the bottom out temperature of these systems is typically around around 4 to 10 Kelvin or 2 to 10 Kelvin but due to um losses in trans transmission of the cooling power we don’t get down to those levels we get down to about 20
Kelvin which for ad prob is more than good enough and it was a study not long ago where people Tred to go even lower and they found out that nothing interesting really happens below 20 Kelvin so yeah that’s said okay uh so this is the basic layout
And we’re going to talk about the individual components starting with the vacuum system because it’s the all encompassing system that we see here then we’ll talk about the cooling system how we can cool how we can measure the temperature how we can measure the pressure and what we can what what we
Know and we don’t know about it what the vacuum chamber is about yeah because you see the vacuum chamber is also not like a regular vacuum chamber you know from yourm any low vacuum systems it’s a ultra high vacuum system uh meaning everything is completely made from metal
So we can heat it up to high temperatures uh somewhere between 120 and 150° C so that we get fast Distortion of water w paper and we can get a very clean system and we call that baking and we’ll get to that and lastly we talk about the detection system
Pulsing system and so on okay so let’s start with the vacuum system and the vacuum system we have in vacuum systems we can have different vacuum levels and we can talk we we call them rough vacuum medium vacuum High vacuum and ultra high vacuum and these are more or less arbitrary limits
They’re more or less arbitary limits but they have some some some system to them as a rough vacuum means typically we can use mechanical pumps to establish that vacuum we can usually establish that vacuum through through pumps so one hectopascal is around 1 milar so uh 10
To the Z milars would be 10 to the two pascals so everything here is in SI unit in pascals but most people will talk about milar or to and to and milar are sort of kind of equivalent 730 toys one one bar but we’re talking about orders
Of magnitudes that are very similar so just remember that this is 10 to the zero uh milars and this is something you can easily you can reach with mechanical pumps this is why it’s called a rough vacuum because we’re using roughing pumps for that uh then we have a med medium vacuum
That goes down to around 10 the minus 3 and this is something we can reach with oil sealed pumps oil seal pumps then we have high vacuum up to around 10 to Theus 7 Ms or 10 to Theus 5 hectopascals uh we can reach that one with turbo
Pumps uh and we can reach that without baking the system so without heating up the system which means we can use rubber we can use other polymers pretty easily polymers that only can withand room temperature uh and then we have Ultra vacuum systems and ultra high vacuum
Systems means we have some um uh we have usually iron pumps or aborption pumps no longer pumps that get rid of the gases but pumps that trap the gases so so some kind of sorption pumps plus a bay cable system so the system is more or less sealed
From the outside because we can’t uh close from the outside because we can’t use any regular seals that would allow for any any motion yeah just for example if you think about uh the semm then you have some detectors where there’s an O-ring and something is uh some Rod is
In the O-ring and you can move it in and out the vacuum will get a little bit worse for a short amount of time but apart from that it seals pretty well this is no longer possible at these vacuum levels yeah if you want to have any motion inside typically what we need
Is some magnetically coupled devices so we have some magnets that move something um that move something inside by coupling it through the through some wall and we have sorption pumps bakeable and a bakeable system and typically if we look at what we do field iron microscopy and atom probe tomography so
Field iron microscopy sort of lives in the base pressure already ultra high vacuum I think it would be possible to make a field iron microscope based on non ultra high vacuum components uh but it would be a bit less practical and atom probe tomography uh or adom probe definitely need ultra high vacuum
Systems and this is why we’re going to look at ultra high vacuum systems but we’re not going to forget about rough medium and high vacuum be because in order to establish an ultra high vacuum we need to go through rough medium and high vacuum to get to the ultra high vacuum
And uh we do not necessarily have pumps that can go we we we neither have pumps nor do we have measurement devices that work through all orders of magnitude there’s no single pump that can get you from rough vacuum to ultra high vacuum there’s no single measurement device
That can allow you to measure from rough vacuum to ultra high vacuum and this is why we still have to look at a rough medium and high vacuum when we get to the ultra high vacuum yeah and but we’re not alone yeah so we have particle accelerator people that use the same
Techniques uh we especially have people doing surface studies and there’s quite a few of them here in elong especially at chemistry and chemical engineering there’s quite a few people that have ultra high vacuum systems for surface studies because in the end what we do in atom probe tomography is also surface studies we
Remove atoms from surfaces um and of course the big bias in that area is semiconductor processing so very often if you want to understand what equipment you have available and why you can’t get your hands on certain components at certain amounts of time uh it’s important to look at semiconductor
Processing because there you have equipment that also goes down into the uh alra High vacuum um into the alra high vacuum range um yeah the increase in the temperature in ultra high vacuum is just because of the kind of pumps that we use
In there or uh if we want to go if the PS which are currently used for high vacuum are Advanced and future for example we can use them for ultra high vacuum then we don’t we will not have the increase in the temperature or if no
No they increase the increase in the so in in ultra high vacuum equipment we need to go to elevated temperatures to disorb all the water vapor and other contaminants um that we introduced from opening the system so if you build the system originally or you have to do some
Maintenance you need to open the system to atmosphere and this will introduce water vapor and potentially also other contaminant like hydrocarbons so how to get rid of them uh has no impact or the amount or the type of pumping doesn’t really have an impact on the speed of
Disor because this is just a disspation from the chamber balls into the inside of the chamber and this is why we’re heating up that’s independent of the type of pumping that we’re using uh and by the way there’s people that are going even further but this is sort
Of the best vacuum systems that are that are available for the cleanest vacuum systems available that are open systems open systems meaning that I can introduce a specimen there’s people that go down to 10 to the minus 17 but there people for example building atomic clocks they
Build a clock you close it up you clean it as much as you can and then you never open it ever again yeah but for us of course we need to introduce specimen so we don’t have that luxury okay so um for us the important bit is now uh how what contaminants can
We have uh which contaminants do give us problems and how to get rid of them and this is what we talked about right now this is the the whole heating issue yeah so uh and the important thing for us is the vapor pressure yeah so what vapor
Pressure here again in pescal so 10us 8 means 10 to minus uh 10 ms so this is at the end roughly where we want to end up at versus temperature um and you can already see that um things like helium uh hydrogen neon and so on they all have equilibrium
Pressures um that are so high at these temperatures that they will just immediately disorb anyway so if you pump something down and you have some residual helium no resid helium will adhere to your to your surface and similar things for oxygen um methane argon Co CO2 but for hydrogen if you look at
Hydrogen uh sorry if you look at water vapor at 300 Kelvin where we’re at then we already have a a much much much much low equilibrium water pressure water uh vapor pressure but this is of course for bulk water we now thinking about water absorbed on surfaces so things look a
Little bit different but this sort of gives you a rough tendency which cases you don’t have to worry about helium definitely you don’t have to worry about helium hydrogen neon at least as is as introduced originally we’ll talk about hydrogen separately because this is a this is the last contaminant to go
Usually the reason for that is that we us usually use stainless steel Chambers and the hydrogen comes out of the stainless steel chamber but essentially the further up we go from here the worse our contaminants get and you can imagine what’s probably worse than H2O I mean these are all surface sorbing
Species so the larger the molecule probably uh the less it will it will want to dis want to dissolve because we’re mostly talking about fous forces so H Jo is bad but if about hydrocarbons if you have any oils for example then it gets even way worse
Because they they even want to they want to they want to evaporate even less okay and so the question for us is how long does it take to get rid of them yeah so always remember um at a pro pressure is about 10us 12 pascals yeah
And so how long would it uh would it take for molecules with a certain disor energy to disorb um um to disorb from the surface and so you can see that uh we have Maxima here so essentially we have got two parts we’ve got two parts and this part here
Um is uh this part here is kinetically limited yeah this is kinetically limited means um we get dis sorption of the molecules and it will take a certain amount of time this is this is the the time to reach a certain pressure If I have one monol layer absorbed on a
Surface whereas here on this side here here we’re energetically limited we’re energetically limited so essentially if the sorption energy is higher than a certain uh than a certain energy then um the the the the molecules just sit stable on the surface so if you have very heavy contaminants then what will happen is
That they might not even disorb just because the disor energy would be too high but now if you’re saying okay then I can just leave some oil in my ultra high vacuum chamber and it will be fine uh no it won’t because you will always have some Cosmic radiation you will always
Have some uh some processes by which these uh these molecules Decay into smaller fragments and they will then start disorb um and uh when we’re downstairs in the lab later in the leure you will see that uh Cosmic radiation is everywhere and we see in our detection systems uh quite a lot and
Um yeah and so we will we would have a pretty large amount of molecules disorb just because of uh because of Decay um issues uh and so what you can see is that based on the pressures we’re after yeah so 10 to the minus uh 10us 8 pascals was I the wrong way
Around eight so roughly this we with with our system would be roughly somewhere here in between somewhere here in between we can see that molecules at a probably about 120 130 KJ per mole dsorption energy would probably be the worst and here we’re roughly at a oneyear dsorption time so you can see
One monol layer for one monol layer to disorb would take about a couple of years couple of months to a couple of years was very very strongly depend on what molecule we have and what room temperature we have you can even notice that um if you look at the logs of our
Instrumentation yeah if we have a sunny day outside and the sun is shining directly on the chamber and it gets one or two degrees warmer you can see that in the vacuum levels just because the residual disspation will just speed up a little bit um okay uh and here we just have
Some typical disor energies um this is what I could find in literature typically you don’t really find the ones that are most important for us but you can see hydrogen from Platinum uh this is something that sticks pretty well hydrogen from ruthenium doesn’t stick pretty well but typical disor energies can be around
That that sort of energy level so what happens now if we go to elevated temperature here’ got 573 Kelvin uh which would would be a typical bakeout temperature yeah um and there you can see that um um that even for the pressures that we’re looking at the
Distortion times are more in the in the minutes rather than the hours or years what that means is that you don’t need to bake your system I know people that build Ultra vacuum systems that are confident enough that they don’t need to do any maintenance they just close them
Up leave them for a couple of months and then they’re at base pressure as well yeah um but uh then you need to have the patience and you don’t need to need the uh and you don’t have to need the instrument but but what we do is go to
These temperatures and then our Distortion times are relatively short still uh baking it typically takes 24 to 48 hours the reason for that is that very often we have thermally insulated Parts in the system and it just takes a little while until they warm up if you
Just want to uh if you just want to uh bake if you just have a chamber and you want to P the surface as soon as everything’s warm everything will disorb in a couple of minutes and then the contamination is gone okay so standard vacuum Chambers uh typically we have stainless Steels
Three4 and 316 so for everyone that’s not a u that’s not a uh steel person it’s essentially just regular ortic stainless steel that is stuff your your Cutler is made from uh nothing nothing too fancy uh it’s because it’s got good weldability it’s easy to polish and has good
Corrosion resistance very often if you want to Polish electr Polish system so produce even uh uh even better surface qualities we go for 36 in L because it’s it’s better to electr polish and typically we also use electr slag refined grades and they can you can only
Get them in 360 anyway I think that’s the even the bigger reason but certainly uh almost all components are available in 304 uh and some are available in 316 and 360 na this is the standard vacuum uh system stuff um and uh you will get some out
Guessing from that yeah so here we’ve got out guessing of different uh out guessing of different um materials over time and you can see that if you look at different um different materials if you got something like nepr like a polymer you get a lot of it I have no idea what Aral
Diet is um but here you’ve got you’ve got brass um brass very often we have the problem that the zinc or the tin in the brass or the bronze can be a problem during the baking it’s very often that’s why we don’t use that one but you can
See stainless steel here starts out strong yeah but then eventually will go and taper off and aluminium you know starts out worse and then eventually becomes very good the reason for that is that aluminium uh doesn’t the aluminium itself has a very low hydrogen diffusion rate and as such we don’t get hydrogen
Diffusing out of the chamber but it also has a porous oxide slightly porous oxide aluminium oxides a lot of dirt typically hides a lot of sored material usually hides in that oxide as so initially to get rid of all the dirt that’s in the oxide maybe you can maybe you can’t but then
It gets very good um and then with with stainless heal eventually the hydrogen is going to be the last residual to go uh of course baking is a little bit more limited with aluminium compared to stainless steel but it’s a different story what’s missing here is titanium
And titanium is our friend so we are building more and more systems here in airong and from titanium and um the reason for the that is hydrogen permeability yeah so here we’ve got measurements of accumulation of hydrogen in metal in the vacuum PP is result of permeation from the atmosphere at
25° and you can see okay U it will go very very quickly with uh um with low carbon steel and relatively slow with copper which is FCC here we’ve got BCC materials and um if we look at the per ation constant versus temperature for hydrogen you can see okay this is nickel
A low carbon steel um do we have a stainless steel no but probably somewhere in the middle you can see aluminium very low penetration um yeah uh and just to see the this orders of magnetes in between the reason why I put that slide on is because the
General wisdom was and to an extent is that some of the hydrogen that goes into the chain but diffuses through the chain into the vacuum um this is not something that we can really uh confirm because our um our experience especially with uh with Ultra Clean vacuum systems is it
Doesn’t doesn’t necessarily happen and probably the contamination mostly comes from the dissolved hydrogen in the metals themselves okay so which Alternatives do we have uh we have we can use aluminium Chambers or we can use titanium Chambers and the good thing about aluminium Chambers is that they have a very high
Thermal conductivity compared to stainless steel so it’s uh it’s much um much quicker to bake the system because everything will just thermally equilibri very quickly and uh we have a uh seven orders of magnitude less hydrogen permeability uh which gives us uh a really good Clean vacuum um but the
Problem is and the reason why we don’t usually use it all that much is because temperatures are limited because aluminium it’s precipitation hardened aluminium UH 60 UH 60 61 very often and so the baking temperature is limited and what’s even more important is aluminium is relatively soft and if
You have ever handled uh Ultra vacuum systems you’ll know that if you uh if you hit a flange the wrong way yeah it will get a Nick in it and then the chamber is done and this is really really Critical with aluminium CF components because it’s very easy to
Ruin your chamber and this is one of the reasons why we we don’t use them uh as the alternative we have titanium yeah um and titanium has a relatively low thermal conductivity so baking is somewhat slower uh we don’t get any hydrogen from the uh from the
Chamber we also get no carbon I don’t know if that’s really if it’s really relevant but that’s people that try to sell that very often bring it in as well um it’s lighter than steel which may or may not be relevant uh sometimes the chamber still turn out heavier than the
Steel Chambers just because you need because you need thicker wall thicknesses because they’re typically welded tubes because you can’t get them as uh as as finished tubes you have to make the tubes Yourself by welding and then you need thick material in order for it not to
Warp uh and yeah very well suited for extreme Harum capabilities but it’s expensive but I would cross it out it’s usually not a major expense so our titanium Chambers typically didn’t contribute to the expense of building the system much at all so i’ I’d really put it in relatively expensive but it’s
Not the main contributed to the instrument costs okay so this was the chamber material uh but of course we also need other materials and the materials we typically need is materials that either thermally insulate or ther conduct or electrically insulate or electrically conduct because this is the typical um
Um the typical requirements that you have and combinations thereof yeah sometimes you need an electrical insulator that’s a thermal conductor sometimes you need a thermal insulator that’s an electrical conductor and so on okay um and for that we can use Ceramics and we also need to be able to
Make the parts for our system from that and typically what we can use is machinable Ceramics machinable Ceramics are Ceramics like make yeah Mako is um I think alumina alumina pilia some combination of that so it’s essentially it’s a it’s a ceramic that you can
Machine so that you can Mill turn and so on uh the problem with that is just that maker becomes conductive after long irradiation um which is not important for us but if you ever build uhv systems for for example synchrotron and so on where really irradiate stuff for a long
Time then this can become a problem the the the uh the thing about Ceramics though is almost all of them um are Rel all almost all of them are pretty okay to use but uh usually we use the ones that we can match in thermal in thermal expansion coefficient to the
Metals that we use because while while you B your system obviously gets warm and cold and expands and contracts so you need to have things that are thermally compatible with each other uh and finally if we want to have uh electrical insulation and thermal insulation we also can’t get around
Polymers really uh if you need both of that uh and polymers you can use in ultra high vacuum but it needs to be polymers that have a very low outcasting rate and there’s a couple of those and essentially uh they are either poly image based yeah Pi based uh or PK based
Uh or or some per fluorinated polymers like PTF PF um or viton which is PE flated Rubber and they are also fine to use but they tend to have temperature limitations yeah so typically PTF typically we use Pon which is a flurocarbon um rubber as
Seals where we can uh we use teflon we use PTFE um uh PTFE also in areas where typically we need something with low friction but the problem with PTFE is that it’s easily hardened and breaks down under IR radiation and it also tends to emit small amounts of
Fline a good example of that is there Ultra vacuum transmission electron microscopes that use uh PTF a lot inside and they always seem to have a fluorine peak in their uh in their microanalysis data just because it get some out guessing uh what we use a lot is either
Poly ether ether Ketone or poly imid poly imid you might know as kton this is typically this uh um these sort of yellowish brownish yellowish films um and they are both suitable uh PK is good for mobility and machineability uh and it’s very widely used in uhv components
Um and it’s a relatively expensive polymer but for the amounts that we need it’s not super expensive uh poly imate uh is also very well suitable but poly imate is extremely expensive so we’ve got little rods of poly imate downstairs about this size that cost about ,000
Each just the material so if you see dark brown polymers lying around in the lab put them where they’re locked away because we don’t want to lose them okay and I think we’ll jump over that it’s just the data so uh lastly you know for us um 3D printing is something
That we also like the question is can I use 3D printed Parts well probably not from acrylic so if you use our printers downstairs um like our resin printers probably not because you can see the amount of residual um residual gases that you get is into the 10^ Theus 7 10-
6 10us 7 m range um but if you use uh and if you have even if you have a 3D printed nylon which you can also get similar levels of vacuum but if you have metallic 3D prints here the example silver for is silver uh then you have
Right a little issue and the residual is the residual that you always have which is hydrogen which comes from the chamber um we haven’t yet used a lot of uh 3D printed Parts in our systems but that might change okay so the question is then how
Do I build a system from that uh and typically we build systems so using so-called conflat flanes and this stands for con flat confl flanes uh and what they are is essentially you’ve got very sharp edges we’ve got a flange with very sharp edges that are pushed
Into some kind of soft metal seal soft metal seal and confl flanes are just one version of it uh if you ever get to take apart one of our ultra high vacuum valves they have a different shape yeah still same principle but a little bit of a different shape and typically the soft
Metal is copper that we use soft metal that we use is copper but you can also get them in India for example and you can imagine all kinds of soft soft metals that you can use for that uh so typically for larger systems you can get indium wires for
Example for for sealing um and the sealing action is a metal seal it’s a metal to metal seal the reason for the metal to metal seal is because um we can get it to high temperatures typically up to up to 4 50° C so the flanges themselves it
Doesn’t mean that everything else in the chamber will be able to go to the temperature but the flanges themselves can go to 450° C and they have uh a limited number of uh of standardized sizes uh and they are uh if you work with the systems a lot you will know
Them on top of your head pretty quickly which is 16 40 63 100 160 200 um and and then the larger Series start but this is sort of the the most common ones and you’ll see them around a lot and typically you kind of have to design
Your system around these sizes and they actually stand for the clear board diameter so this is since I’ve I’ve worked with a lot of them I notice this is a cf40 yeah and this means that internally we have a diameter of larger than 40 mm at least 40 mm
Uh and you can already see okay there’s a big jump between 16 and 40 then a smaller one between 40 and 63 and then uh about an equal one between 63 and 100 um but this is essentially will get all of the components only in these sizes sometimes you will have to deviate
From that so currently we’re building a system for a for alra vum system for uh Neutron scattering um that is uh o building is building that system uh where we need a rectangular window and then you have to design it yourself essentially and then everything becomes custom made everything takes much longer
To make much longer to order uh and you get no guarantee that it works now with these you can just buy them and a small flange will cost very little money uh even though um most of the ultra high vacuum systems do cost a lot of money when you design a system you
Have to you have to consider two more things and the first thing this is just an escape from one part out of a vacuum system is that if usually you will need some screws somewhere to attach stuff to other things and then you always need to make sure that there’s no dead volumes
So just imagine you screw a screw into here and then the volume behind the screw uh would be trapped and the air would just slowly come out over time and you’ll have something called a virtual leak yeah so you’ll try to find a leaky if a flange is leaking and you don’t
Find a leaking flange eventually it turns out you had air trapped somewhere in a hole because always remember we working at 10 Theus 10 m let’s say 10 Theus 9 and it’s easier to calculate um which is 12 orders of magnitude removed from ambient pressure so if you have a
Volume of let’s say 1 cubic cm right then it’s 10 to the 9 cubic cm at this pressure so very large volume so with be that in mind that small trapped volumes at these low pressures suddenly become very large volumes yeah so typically what you can
Do is drill Escape holes or cut an escape slot things like that uh the other thing is something I already talked about which is thermal expansion very important if you for example have a window which is some kind of glass or now that we building uhv uh sorry uh tpv
Laser systems it needs to be Sapphire or maybe diamond and this will have a different exponent of thermal expansion than the flange itself and you need to somehow combine them and typically what’s done is that you have some kind of uh sign some kind of
Um uh sheet metal piece that sort of can take some of the expansion but of course it’s always good if you can match the thermal expansion typically those windows we have L is called kova kova is called kova because it’s got the same form expansion as glass for example as Flint glass they
Always need to find uh if you if you m if you really Mount things on top of each other you need to make sure that they have sort of similar thermal expansion coefficients okay the question is now we have a chamber how do we get that
Chamber to ultra high vacuum now how do we get the chamber to ultra high vacuum um and this is of course by having a pumping system and a pumping system for an ultra high vacuum system typically consists of um uh of three pieces which is or four pieces really which is the
Work chamber and which is the chamber where we do our experiments then we have a some kind of ultra high vacuum pump and again the ultra vacuum pump typically has no gas Escape Route it will be some kind of sorption pump this is kind of not not 100% the case anymore as
Turbo pumps which we look into are now getting into regimes where you can get to ultra high vacuum with the turbo pump which removes gas but typically that’s not the case and so what we need on top of that is some kind of roughing pump that gets us from ambient uh from um
From around one bar down to wherever the ultra vacuum pump can start maybe 10 to the 0 10 to Theus 1 10us 2 Ms depending on what it is and we need some kind of vacuum gauge sometimes the pumps connect as gauges or the gauges connect as pumps
But typically we need both and we’ll we’ll see L later why that why that is the case why they can act this gauges or act this pumps and so we’ll look into roughing pumps Tu molecular pumps and ion G pumps or G pumps in general not necessarily ion G pumps aborption pumps
So roughing pumps roughing pumps there for us there’s four important uh four important types um and uh we can group them in two different groups uh which is dry pumps and oil lubricated pumps now uh the simplest dry pump is a membrane pump it’s very simple it’s essentially a
Membrane here it’s also called a diaphrag pump uh and the membrane is moved by an eccentric motion up and down uh and then just uh has a suction and an exhaust Port uh with vales uh to uh to avoid back flow now you can see very simple
Pump uh all you have it’s it’s clean uh and all you have to do is every once in a while change the membrane uh because the membrane uh is sort of a rubber membrane and probably after uh one or two or 3 years typically those membranes
Fail then you have to replace it now but it’s relatively simple operation they’re dry they’re very they’re relatively silent and they don’t cost a lot of money and they get you down to typically 1 to 2 Ms and modern tobo pums can actually deal with quite a good amount of back
Pressure so they typically these days good enough to back uh to pumps and we use those a lot for example in the leap system but also in other Ultra vacuum systems that we operate uh the other one that’s that’s very uh very popular are rotary vein pumps uh rotary vein pumps
Pum in rotary vein pumps uh are pumps where you have an sort of a Centric motion as as well where you have two slots here so two uh two baffles here that move in and out um and through that through the rotary motion uh you get a
Compression motion of the uh um of the gas that you suction in and they they are good they oil lubricated which means they can deal with with contamination pretty well which in Ultra vacuum systems is not relevant because you’re not typically pumping out dirt from your chamber they’re relatively affordable
You can get them on eBay for or on on on AliExpress for like 150 bucks of course not the professional ones that we buy a little bit more expensive but you can buy them for very little money and they will get you down to about 10 to the
Minus 3 Ms and the limitation thereby is the outgasing from the oil the outgasing from the oil um and this means we don’t want to get the oil anywhere but the pump and this is why we usually try to avoid them because if something goes wrong say you’ve got uh you’ve got for
Example a um um electricity outage what can happen is that they can get some backstreaming of the oil into your system and if you get a pump oil into an ultra high V vacuum system it’s going to be very very difficult to remove remve it again but we still use them with some
Precautions typically with some buffer tank in between or so that the that the oil can stream back and then they they just chuck chuck away for a long time very low maintenance uh and very good and relatively good vacuum uh scroll pumps are also a type of pump that we
Use very often and essentially they are two Scrolls that move in a fashion like this with respect to each other so one is static and the other one moves in an Ecentric motion and through that uh we have a trapped volume we have a trapped volume and that trapped volume moves
From the outside to the inside or the inside to the outside depending on what uh on how the thing is set up um and um this is what facilitates the pumping motion typically those craw tips are made from PTFE so they sort of self-lubricating um and uh this makes them uh you can
Also get them lubricated but typically they dry and they can get you down to about 5 * 10 to Theus 3 Ms they have one downside though and that is that the the Scrolls they rub on each other yeah and if you don’t have any gas flow continuously going through then you have
The problem that any wear any any wear material gets accumulated and that accelerates the wear because then you have dust inside and then they can they can wear Rel quickly so they’re great for things like load locks and so on where you pump down pretty regularly
Because then all the dust gets clean out and they good to use again but very popular they in a lot of our vacuum systems also the the regular vacuum systems for example all of our semms have scroll pumps as pre pumps as well and then finally we have Roots
Pumps Roots pumps are essentially pumps with pum in German so they there pumps where you have these two Pistons rotary Pistons that with very tight tolerance move with respect to each other uh and they essentially seal against the outside through very very narrow tolerances uh and produce a pumping
Action through that and typically they don’t just have one stage but multiple stages and the good thing about them is they’re dry they’re very clean and they can get to a very good vacuum uh and they’re very but they’re very expensive expensive and one thing is they’re very
Very loud yeah so in the beginning we had a couple of them running but by popular demand of all the PhD students we don’t run any Roots pumps anymore because you literally need to sit in the in the room with uh with uh with hearing protection on this is why we typically
Don’t use them but we have some of them still kicking around okay so typically typical membrane pump so this would be the one that’s used in our commercial atom probe uh it’s a say F mvt BL whatever and what it is is a dual stage diap frame pump so
There’s two di frame stages and that brings us down to a bottom out pressure um a bottom out pressure of about of about 1 M that’s very typical for these pumps but you can go and look the the bottom out pressure out up for for the
Individual pumps um that you want to use but this is typically what you can expect but for modern turbo pumps that’s usually pretty okay which brings us to Turbo pumps the tub molecular pumps are pumps that be used in a lot of systems here in in um at The Institute um not
Just Ultra vacuum systems but also High vacuum systems like for example our vacuum furnaces used to pumps uh our electron microscopes use turbo pumps and so on a very popular pumps and essentially what they are is they are sorption sorption and desorption kinetic pumps so this is the
Uh this is the simplest form invented about over 100 years ago and essentially what we have is a spinning cylinder that spins very fast and what you get is um you get a molecule coming in from this side absorbing on the cylinder and then eventually desorbing again and if the
Cylinder spins fast enough then you will have a net motion of molecules on the secondary side you can already imagine why they called tub molecular pumps yeah because it’s a molecular sorption desorption process and this means that they need to spin very very very fast but because it’s aborption and
Desorption process there’s always a probability for desorption into the inlet yeah and there the result of that is that these are only differential pressure pumps yeah so uh the out the out that essentially will have a um a pressure ratio a pressure ratio of in intake to Output pressure yeah so a
Pressure ratio between Inlet and Outlet pressure that is essentially exponent of some GE geometry Factor uh the blade velocity and so you can imagine the faster the blades spin um the better the the uh the ratio is the faster the blade spin the better the ratio
Is uh of course this is not what modern uh what modern uh pumps look like modern pumps look like turbines uh but don’t confuse them with an actual turbine because an actual turbine is a gas flow device so you will have some gas flow directing in this case you’re literally
Just giving the molecules a sort of an average speed that points more towards the outlet than the inlet uh and typically we have a RoR and a sta uh and so uh if we look um from the perspective of a molecule a molecule from the incident side effectively sees
A at the velocity uh sees a uh an area or an angle area of this size for the adsorption whereas it sees an angular so if the blade spinning it sees a larger angular size when it’s dissing because when it’s dissing it has the same speed as the blade when it’s
Adsorbing it has a speed relative to the overall World frame and this is just the this is just the the principle so always remember it’s a differential pump and the blade speed of the blade essentially gives us the pumping the pumping speed yeah so this is a uh a current model and
This is a f High pace and I’ve just picked that one because we use a lot of them and not all of them uh all of the pumps that we use are fer ones there’s also some that we use from Edwards and some from leska and G no what companies
Um but they all work in relatively similar ways and the most important thing for us is that we have to have some kind of bearing that supports the fast spinning of the blades uh and typically a 80 L pers second turbo pump which is a turbo pump
That fits on a CF 633 flange here so will always find some very distinct sizes with the pumps as well because they’re typically match to the flange sizes this would be one for a in this case we haven’t talked about KF flanes so this is ISO k63 flung
It’s essentially a rubber sealed version of the cf63 flange of the confl flange typically that they have the same sizes but you can get them in either rubber sealed or metal sealed and this is the rubber sealed version um and then it has about 80 L per
Second um of uh um 80 L per second of um um of pumping speed uh and what we typically have is we have one permanent magnetic uh one permanent magnetic bearing um and One ceramic bearing so these this uh this roers that typically half suspended half suspended because a pump
Like this runs at 1500 Hertz 1500 Hertz or 990,000 RPM this is obviously pretty fast so you need a bearing to match that fast motion so this is this brings us to the biggest downside of Turbo pumps and that is that they are vibration sensitive
Yeah so turbo pump if a if if you have a turbo pump on a device that sits on the table if I do this what will happen is uh you will hear very bad noises from the turbo pump and if you’re unlucky the turbo pump might even blow up just
Because the um the rotor will start touching the this STA and then will just be will just be slowed down very quickly uh and there’s one example of an atom probe at colleagues in France uh where someone has put some where it was an air suspended table and because it was laser
Atom prop and someone has put something too too hard on the table and exactly that happened and Tu pump blew up and took the essentially everything in the chamber with it yeah so be careful with tu pump systems don’t move it around when the TU pump’s running yeah because
We have some systems where you can do that could do that for example o nuggles new system is currently on Wheels and the to pump’s running so if you go there and you shake it that that would be very bad so don’t do that um and so here the maintenance is
Bottom bearing typically needs to be changed every couple of years but apart from that they’re pretty low maintenance and they’re very clean and very easy to operate and they can uh together with a backing pump get you from ambient pressure down to where is the bottom out pressure bottom out pressure bottom out
Pressure Uh bottom out pressure doesn’t say it anywhere but it should be oh no bottom out pressure so you can see we’re at less than 5 * 10us 10 mbar so higher uh higher um um ultra high vacuum level but still vacuum level what you can also
See is that um it will have different compression ratios for different elements so heavier elements are easier pumped than lighter elements so it’s harder to get rid of hydrogen for example compared to uh argon or nitrogen or something like that very important to remember um and uh remember it’s a
Differential pump so what we do for example in one of our systems is back a turbo pump with another turbo pump then you can go to even lower pressures than that yeah and this is what happens if you have a tu pump failure um and if you
Want to see a failed tur pump you can go into H verna’s office he has one from back in the day when uh our um old semm had some issues with uh gas coming in too early where you rented it and eventually the TU pump blew up as well
Not as not as violent yeah so it didn’t take the entire sem into its grave with it but you can see if something looks like this yeah and um it broke at one 1,500 HZ 90,000 RPM you can imagine Parts flying around the chamber being a problem okay so this was uh mechanical
Pumps uh the other thing that we very often have are sublimation G pumps in this and in this case the principle is chemisorption so you have some uh some chemically active surface some reactive surface and then you have a disor time that’s dependent on the uh on the dsorption energy aborption
Energy um and then obviously uh the temperature but you can have aborption energies that are so high that essentially you’ve trapped the gas atoms or the contaminant atoms and uh typical sorption energy of oxygen on metal surfaces because oxygen is obviously something that we pretty
Regularly would like to pump uh is in a range of 12 to 70 KJ per mole yeah so relatively low if you compare to what I’ve showed you from the bakeout uh stuff as this is something where the oxygen would sort of float around in the chamber and not stay absorbed on the
Surfaces but if you have oxygen on blank titanium so titanium without an oxide on top of it we’re more talking about, K per Mo which means if the oxygen bonds to the titanium it will never ever leave it will just stay there and this is the
Principle of chemor of Gap pumps and be using that chemisorption um of course if you have something that’s not chemically reactive like an inert gas atom these pumps will do absolutely nothing yeah so if you have something like helium or Argon then these pumps will do absolutely
Nothing um and if you have an air leak for example always remember air is about 1% argon so if you have an air leak you might get rid of the nitrogen you might get rid of the oxygen but you will not get rid of the Argon through these
Pumps um okay um yeah the times probably not that important um typical aborption pumps that we have um are that titanium sublimation pumps now what we do here is we have a titanium filament and we just have titanium suating from the um from some hot titanium wire and then re
Subliming onto some cold surface cold surface can be room temperature or can be liquid nitrogen cooled or can be water cooled depending on whether you just want to intermittently do that or continuously operate it for us typically we when we work in Ultra vacuum we do uh
A run of titanium sublimation and then that’s good for a couple of days or a couple of weeks uh but if you’re in semiconductor processing equipment you might have them running constantly and then you need some water cooling uh in order for continuous sublimation to happen um but these are these are very
Practical very easy to run pumps uh we have one of them on our uh on our field line microscope and that we uh that we used in because we wanted to add on prob tomography for field microscopy wasn’t all that important U but uh this obviously means you supplementing some
Titanium uh and uh you you’re put inputting heat into your system which you may or may not want so an alternative are so-called non-evaporative G pumps so this is titanium sublimation pump and non-evaporative G pump and what a non-evaporative g pump is is the same thought they produce a highly reactive
Surface of typically some uh Titanium or Titanium zium or zonum aluminium alloid is a couple of different ones out there and that then sorbs just as the titanium would Sor here but in this case it’s an alloy where the oxide that swallows its own oxide it’s a reason at
A reasonably low temperature titanium does it as well by the way but by titanium you need to go up between 450 and 600° C before titanium takes takes up its own uh takes up its own its own oxide which is not really doable for a Ultra vacuum system to really go to
These temperatures now with neck pumps it’s typically more like 250 to 450° so it just heat that up typically buy them with heating integrated and then this thing will just go and swap um the good thing about them is that they’re relatively easy to operate they don’t
Need any power while they’re running you once activated they will just keep keep sorbing for a very long time unless you have a leak or something uh but if you put in too much contamination they will eventually uh well turn into all oxide which means dust yeah I’ve taken apart an old atom
Probe that had an non revertive catap pump that by the end was essentially just dust yeah we took it apart and uh I first thought okay we can still use that one but then when you touch the uh the the the pellets the the neck pellets it was just crumbling to dust
Um and they’re relatively expensive but we have quite a few of them in operation in our atom probes and also in our titanium in our um also in our uh thermal um in uh in our thermal disspation system so uh very good because it’s yeah the very robust pumps
Unless you you give them too much contamination then they become a bit of a problem another pump uh that we use and we use that one a lot is a so-called ion pump ion G pump igp ion G pump and that’s the main pump for example in the
Leap system but that’s also the main pump typically in transmission electron microscopes including our transmission electron microscope uh it it’s also the main pump in electron microscope columns so all the field emission electron microscopes all also have ion G pumps as pumps uh and the way that they work is essentially
That you have um that you have an anode and that they have anodes and cathodes in two directions and that they produce a magnetic field that ionizes any gas atom that that that comes in and then the gas atoms get accelerated towards the uh towards the cathode and be
Thereby two two things first of all they might get buried but also they can sputter the surface of the cathode which is typically titanium and by that action they produce fresh titanium surface and then you have chemisorption again but because the uh the ions also can get buried you also have some
Pumping action for noble gas atoms yeah because essentially just shoot the noble gas atom into the material and it just gets stuck there it’s not great for noble gases but it has some pumping action for noble gases um and uh basic principle so we have a a um
We have an anode cylinder and a cathode plate in this case is a very typical uh way of doing that and through the uh through a large magnetic field uh you will have any ion produced in there flying around in a cork screw fashion which just increases the
Residence time for it to get ionized by some collisions uh and then eventually it will hit it will hit the surface downsides are big magnets uh big magnets mean very heavy pumps uh they’re big and heavy but they’re very robust to operate uh for us in in atom prop tomography
They also have one additional downside and that is some of the ions typically can escape and can get sucked into our detector so we need to have some good shielding in order for the ions not to get sucked into our detector this is by the way the pump
That we have on in the leap system Sy and one thing you can see is that it weighs about 100 kilos a very heavy pump uh which means if you design the system the only place you can put the pump is essentially at the bottom just because
It’s very heavy can’t put it on the side top or anything okay so noble gas noble gas pumping uh typically you can see here helium pumping if you have got a regular one is about 10% of the pumping speed compared to something like a nitrogen or
Water Vapor but there are some um some uh versions where you get you get improved um they get improved pumping speeds for noble gases if that’s something that you need but it can also help for instruments like a field iron microscope because in a field iron microscope you typically want to pump
Everything but the noble gas because that’s what you use for Imaging so there this is even helpful okay which brings us to uh to uh to the cryocoolers so the cryocoolers that we use are typically GED mcon cryocoolers so there’s essentially three different types of cryocoolers um and we probably won’t go
Through all of the different ones but I think you all know what a sterling process looks like yeah so Sterling process essentially means that you have um some kind of displacer and regenerator yeah and you have um you have Heat going in here and through compression of in our case helium
Because the temperatures we’re at only helium and hydrogen are even anything but solid everything else is already solid yeah so we’re using helium great so we’re using helium uh and the helium gets cooled on the primary side on the secondary side um uh the heat is taken out uh the problem with Sterling
Machines is that they relatively that they’re relatively expensive to build yeah because they’re custom piston systems so we are typically using Gord McAn systems where um the compression bit so here’s the compressors the compression bit and the displacer bit are separated so you can use just a regular type compressor regular gas
Compressor to compress the helium um and then just have a cold head uh that does the displacing uh motion and and um this is typically the sound that you hear when you go into an atom probe lab is the the sound of the displacer going back and
Forth yeah um yeah um so this is these are all closed cycle helium refrigerators you can also have open cycle Refrigeration essentially just use liquid helium but this is typically only economically feasible if you have some helium recuperation which we do not have in this building but for example the
People over at physics they they have that in their buildings so they can just go and get some liquid helium from the helium liquifier which the university owns one a relatively big one actually so we can make our own liquid helium um and then you can cool uh which means you
Don’t have this awful sound and you don’t have the vibrations that are caused by this go and I think we’re going to go very quickly through the cryos to just quickly go through the uh quickly go through the pressure gauges yeah so we also need pressure measurements and this is something
That’s going to be very important for you if you operate any vacuum equipment here uh be it an atom probe or an sem or something else to understand what the measurement readings are and essentially uh we have uh hot cathodes so we’ve got hot filament and cold cathode gauges uh
Essentially at uhv we typically have ionization gauges yeah so it’s not mechanical gauge like a uh B gauge that you know B manometer it you know from I don’t know your regular vacuum equipment um that you know just we just have the the dial or for example the high pressure
Equipment yeah so our thousand bar um systems for example they use mechanical gauges but here we’ve got iron gauges so essentially we ionize the residual contaminants and measure the current that they produce and typically we can do that in two ways either by having a hot filament now where we have
Ionization at a hot surface or by having a cold cathode which is essentially uh exactly what I just explained with an iron G pump but just use as a gauge that’s why I said ion G pumps you can use as gauges or cold cathod gauges you
Can use as ion G pumps this is where iron G pumps even come from back in the 50s 60s when they were developed people have noticed that gaau with uh the chambers with with these iron gauges um tend to get the vacuum reading tends to get better over time and first they
Thought it’s it’s some kind of drift effect but eventually uh it was clear that no actually vacuum gets better because the gauge itself acts as a pump uh and the hot filament gauges the good thing about them is they’re very simple uh they have no magnetic fields
And I think next time at the next lecture I’ll just bring some of them I have some old ones lying around um but the downset is they produce heat and they have a finite life and typically we work at 20 Kelvin so we try to avoid any heat
Sources uh the good thing about cathod called cathod gauges is they give off no heat um but the magnetic field May interfere with your measurement if you have something um where magnet magism plays plays a role and the reading again depends on the gas type and this is very
Important always to remember that the gauge reading doesn’t mean you know what the vacuum is uh because you don’t know what the partial pressures are you don’t know if the vacuum is exclusively hydrogen or nitrogen or argon or something okay and for us very important
Just like the iron G pumps also we get ion emission so some of the ions get ionized get make the way out of the gauge and can get sucked into the detector as well which just means we need to be strategic where to place the gauge so typically of
Course you get a lot less emitted from them compared to an iron to iron G pumps okay we’re not going to go through the details uh of how they work here just uh if you actually want to know with a cold cathode gauge what the pressure is and there are conversion
Factors typically the reading is for air now so if the residual gas that you have is exclusively air maybe you have a leak then the reading is correct yeah but if it’s um if it’s hydrogen for example we have a strong um uh we have a strong underestimation of
The pressure uh if it’s uh if it’s all something like argon then we have an overestimation of the pressure so unless you know what the partial pressures are always bear in mind that you don’t know exactly what the pressure is yeah uh of course now is a thermal dsorption system we have one
Where we actually get partial pressure measurement but this is a whole different topic that we’re not going to cover here um at higher pressures I said this these ion gauges they only work at very high at very low pressure so typically below 10^ Theus 3 mbars so we also need
Some measurement of pressure between 10^ minus 3 and ambient pressure and that’s typically so-called Panic gauges yeah you they very often see them as Pig um and they’re very simple devices it’s just a hot wire um and uh what we measure is the uh the change in conductivity of that wire that stems
From cooling by convection so the gas molecules essentially cool the wire and we measure the change in resistivity again we have some dependency on gas type uh but it’s relatively gas agnostic and it can take us roughly from 10 to the minus 3 uh to ambient pressure if it’s a regular one
There versions with pulses we essentially let a lot of gas absorb and then you send a pulse uh you you just heat it up for a very short amount of time uh then we can get down to 10^ Theus 4 and these days there also some
On M basis that can go down to 10 Theus 5 you will find all of them in some systems that we have just bear in mind that that’s how the measurement Works in this case and this is just an overview of uh where we can where we can use them yeah
So ionization gauges typically go down to 10^ Theus 11 uh on the ox cut for example the vacuum is better than what we can measure with regularly available measurement gauges uh but there you can get some uh some customade ones very special hot iation gauges and so on but for us if we
Know from the experiment we see no residual contaminant in the experiment then I don’t know if you really care if we have if we can precisely measure what the what the pressure is yeah and then we have our Panic gauges at at higher pressures and typically what we have is
So-called wide range gauges which are a combination of the pirani gauges and and I either hot cathode or a cold cathode gauge so the Panic gauge essentially measures from ambient pressure down to some set Point typically 10 to Theus 3 and then the gauge switches over to
Either cold or hot ionization as so if you pump down the system and you see suddenly a jump in pressure yeah it’s typically the switch point of the of the wide range gauge to go from Pani to ionization gauge okay which brings us to the last important bit um for the atom probe
Which is the detector yeah and the detector and all current atom probes a micr channnel plate as preamplifier plus a delay line for for position encoding plus some intermediate Electronics plus some intermediate electronics and this is a model of a cross delay line single detector single
Iron detector uh and we can just we can go downstairs and I’ll show you what a real delay line detector looks like because we are very one of the very few people that just have one lying around that’s not that’s not used somewhere and so essentially these
Things is first the micr channel plate is a preamplifier it’s a pit here in front and it’s a glass plate we talked about it last time already and here at the back is a delay line which essentially is just round wires that are used for the position encoding I used for the position
Encoding and eventually uh that one those one wires go into some Electronics we’ll have a look into first typically in the so-called constant fraction discriminator and then at time the digital converter but these days we could also use analog Electronics or analog to digital converters they are
Becoming fast enough for also uh for usage in atom probe and there are there is one instrument in France it already runs on this uh on this schedule okay so first how do we amplify the atoms and um this is done by microgen plates um and what micr chenel plates
Are is just a lot of scintillators so cators is a photo is an electron multiplier so you have some incident electron that hits um a surface somewhere produces secondary electrons and they produce more secondary electrons and so on and so forth so essentially you get an electron
Avalanche yeah so here it’s it’s drawn with an electron but you can also use an ion or a photon to do that they just need to have sufficient energy yeah or sufficient impulse yeah what microps react to is impulse so mass times velocity um which means that heavier
Ions typically uh can be a problem but not within anything that we are concerned about so for all of the ions we are concerned about they work perfect perfectly fine it’s a couple of Manufacturers out there in adro very often you find one from hamu and I’ll
Show you later on why but you can buy them from a couple of different vendors and what they do is essentially typically give you a gain of around so the gain the amplification is typically around 10 to the power of three so about a thousand times uh and what we typically do
Therefore is is having a so-called two-stage mCP setup yeah so we have two on top of each other so that we get a gain of about 10 to the^ of six um and the the way we do that is of course we need to have some kind of Supply
Voltage some kind of Supply voltage high voltage uh that triggers that electron Avalanche and here we can see the gain depending on the supply voltage typically we uh we operate about 1.2 Kilts per plate so two around 2ish k Kilts so two to 2 and a half Kilts for the entire
Stack uh main application of those by the way is military great night vision so uh this is why in times of Crisis uh sometimes these things get characterized as dual use goods and it gets a little bit more difficult to buy them or you have to fill out stuff and saying yeah
I’m not going to ship them to North Korea or something because that’s that’s where the main application is okay so um typical setup Chevron stack Chevron stack because we have an atom prob specimen and we’ve got an ion emitted and Chevron stack means that the channels are at an angle and the angles
The angles are twisted 180° between the subsequent between the subsequent plates the reason for that is that if we have an iron and it flies into a channel and it has the exact orientation of the channel it will still hit hit a channel in the second plate so it still has a
High chance of hitting a channel in a second plate um and two two uh again because uh we have a 10 to this we want to have a high amplification we use two and not three just simply because uh eventually we’ll get into saturation and will not
Be able to extract more charge out single iron hit and this is already the case if you have irons hitting after two after two plates so you can use three but it’s not going to help you anything um yeah so this is a typical setup so for positive iron detection
Yeah so you have a you have the stack of the two of the two things here and then in this case you have a phosphor screen for Imaging and this is what we do in uh this is what we do for field iron microscopy yeah just the electron cloud
That we produce goes into a phosphos scen yeah for atom of course we want to have time and single particle time and and and spatial resolution there we need to have some way to read out when this electron cloud was produced and where it was
Produced so how do we make them and what are the limitations well there is a very distinct process to making them and this is uh also one of the reasons why we have some residual measurement background because we need to have some glasses where we have one
Channel glass and one soluble glass that we can where we can make out of the channel glass we can make a tube put a soluble core so a soluble glass inside so we can get some kind of inner and outer Tube Stack them together then they’re fused drawn out stacked further
And eventually we eventually out of an entire bundle we cut a dis yeah and this is our microgen plate and in order to be able to do that we need to have a glass that’s soluble inside a glass that in in that’s not soluble in some kind of
Solvent uh and uh the only glass combinations that we have are using potassium containing glasses uh for the channel glass and potassium has some residual uh radioactivity from potassium 40 and this is the reason why we have some background in our detection always because that potassium 40 will Decay and
Each Decay it’s an I think it’s an alpha decay I should look that up but I think it’s an alpha decay and each DEC event will trigger the detector this is typically the background that we see and after etching this is what an mCP looks like and you
Can already imagine if an iron hits here nothing will happen but if an iron hits in the channel then it will get Amplified what it also means is that we have a detection efficiency and which is typically 0.6 to 0.9 depending on the type of the microchannel plate 0.6 to 0.9
Roughly this is what one what what the micen plate looks like and what you can seees also that we have some bundle boundaries here and you will also see them show up in your atom prop data because also the detection efficiency uh varies a little bit at those bundle boundaries so you
Can they can show up in the Adon prop data if you look with very very good statistics uh I talked about efficiency yeah so standard uh so a standard microchannel plate uh has an open error ratio of about 60% so iron hits between the channels nothing happens iron hits
In the channel it gets Amplified so people a couple of years ago from hamamasu developed so-called funnel mcps where they used some iron processing to sharpen up the end at the end bits yeah uh and so in this case we’re talking about an open area ratio of 90% active
Area um these 90% are just based on measurements like this one so they put an iron current uh and just look at how much higher it is compared to regular mCP um the open error ratio directly doesn’t really have any meaning on an mCP like this yeah because theoretically
The open area ratio is 100% but we don’t see 100% amplification because some iron still get stuck somewhere here or at least don’t get Amplified um important thing to remember in this case is here ions either hit the hole or don’t hit the hole here if they’re somewhere up here they may or
May not get Amplified but of course the higher the energy is that the impact with the higher the probability for amplification so here for these we have a saturation at about 2 K for the ions yeah so at about above 2 KV uh we amplify all the ions whereas uh for the
Funel mcps we have a first plateau and then it takes a little bit until until at about 5 KV all ions get fully Amplified or at least 90% of them get Amplified so we have some energy dependence so how does the position en coding work well it’s relatively simple
I think we talked about it already last time a little bit so we have essentially y mandas we have wi mandas and the electron cloud that the mccb produces hits somewhere here yeah and it’s relatively large it might be a centimeter across yeah depending on how
Far away from okay I don’t I think we’re going to stop on time today and I hope next next time there isn’t going to be as much construction going on anymore and what will happen is that this electron cloud then will induce a pulse that just like
If you throw a stone into a pond of water it will have a uh you will have an electron wave moving towards the ends yeah and so if it hits somewhere in the middle then uh you will get signals at about the same time on each side if it
Hits somewhere towards this side of course you will get an a signal quicker at in this case X1 then X2 yeah and so essentially uh we can encode the position by looking at the time difference we don’t even need a start time yeah but if we additionally have an
External time reference we can also say from the time average we can also say when the atom hit and so this way we get time so we get x y and time so we get x y and time from the time sum and a time difference of course in a wire coil it
Only works for One Direction it only works for One Direction so we need to have at least two coils to get X and Y some detectors you only have one coil because you might only be interested in one coordinate uh for example we do some spectrometry uh sometimes people only need
One uh typically an adom prob or very often what we have is also a third one and that’s what you’re going to see downstairs so we have three so they we have one redundant delay line which is great if we have multiple ions impacting
And we need to figure out uh we need to figure out where the multiple ions so state of the art is you can do single and multi particle detection we typically have spatial resolutions in a range of 20 microns which depends obviously on how far the the holes and
The mcps are apart and temporal resolution is typically better than 100 200 PS but it’s sort of in the in that range say 100 200 PS um you typically have detectors up to 150 mm active diameter they’re typically not us in at prob for the reason that um
For these very large detectors the uh the mcps typically have have smaller holes compared to the parts in between because it’s a large glass plate and you need to be able to manipulate it without breaking it um and so typically we only use up 80 mm because that’s the largest way you
Can get MC I know call inart for example take us up to 120 mm you can certainly go up and nowadays there some mCP manufacturers that can make capillary mcps the typic new type up to 20 cm as you can make larger and larger detectors uh and for improved um
Multihit detection you have this r del lines the P was filed 2000 so they were R out 2 years ago so now you can even buy them commercially without having to buy en uh and the reason for the for the um the reason for the uh for the run and
Delay line is essentially that you have a dead area so remember we have this P of electrons uh and just imagine you have two hitting two electrons hitting close to each other and this is just double hit distance yeah so you can see if one hits here and the other one
Relative to it hits here that it’s fine but if it hits here and here then this coordinate we can’t recover and just simply because we would have two electron cars across to each other we can’t really distinguish them we distinguish them and if you have a redundant delay line we have
Additional information and then you would only have a very small soal dead area where you can’t distinguish between individual kits in this area typically is around um 3 to 5 mm on the detector the reason being that the electron cloud typically moves or the electron rate typically moves with about 1 mm per
N and the pulse WID is is in the range of 2 3 4 NS let’s say four to 5 NS and so typically if it’s uh if it’s a if it’s a distance closer than that then we can’t really distinguish it uh of course this is because of the
Electronics in between which we don’t have time to get into but if we have some smart electronics that can take those uh two PES and deconvolve them then you might make might be able to get rid of that uh just by b or might know by now you can our friends in France
Have done that okay and with that uh I think we’re going to just maybe quickly touch what a reflectron is because one of our instruments has one yeah so um before before the detector sometimes we have something called a reflectron what the reflectron is is an electrostatic mirror is an electrostatic
Mirror uh and what it does is essentially if the ions fly into it if an ion has more energy than another one then it flies deeper into the reflectron and has a longer flight path and if you do that in a smart way then you can compensate for energy differences in
Your ions and the energy differences come essentially from the fact that if you use a voltage pulse then some ions might come off at the beginning of the voltage pulse and some of a little bit later and so they might have a little bit of an energy
If you use laser puling then they typically have no energy difference because the Thal energy you input is very very very small compared to the electrostatic energy you put into the and so in a reflectron they fly in the ions fly in a little deeper if they
Have more energy and there’s a result we get time focusing and energy focusing and we get a very very high mass resolution we get a very very high mass resolution uh the downside in this case is that in order to produce the field to do that we
Need to have a mesh that’s terminating the field and obviously some ions might crash into the mesh and get lost and so typically that’s about half of the ions because the ions need to fly in and back out again and modern reflectron also have a little a few
Measures in between so they can accommodate wide view angles um and so we lose a lot of ions but we gain a lot of M solution and our leap at prob has such a refr our home doesn’t have okay and with that uh we’ll uh we’ll end for
The instrumentation and next time we’ll go into field evation phics and I’ll just take it downstairs for a minute and I’ll show you what a detector looks like in reality m