A New High-Throughput Focused Mev Ion-Beam Analysis Setup

instruments

Article A New High-Throughput Focused MeV Ion-Beam Analysis Setup

Sören Möller 1,* , Daniel Höschen 1, Sina Kurth 1, Gerwin Esser 1, Albert Hiller 1, Christian Scholtysik 2, Christian Dellen 1 and Christian Linsmeier 1

1 Institut für Energie- und Klimaforschung, Forschungszentrum Jülich GmbH, 52425 Jülich, Germany; [email protected] (D.H.); [email protected] (S.K.); [email protected] (G.E.); [email protected] (A.H.); [email protected] (C.D.); [email protected] (C.L.) 2 Peter Grünberg Institut–9, Forschungszentrum Jülich GmbH, 52425 Jülich, Germany; [email protected] * Correspondence: [email protected]

Abstract: The analysis of material composition by ion-beam analysis (IBA) is becoming a standard method, similar to electron microscopy. A pool of IBA methods exists, from which the combination of particle-induced-X-ray emission (PIXE), particle induced gamma-ray analysis (PIGE), nuclear-reaction-analysis (NRA), and Rutherford-backscattering-spectrometry (RBS) provides the most complete analysis over the whole periodic table in a single measurement. Yet, for a highly resolved and accurate IBA analysis, a sophisticated technical setup is required integrating the detectors, beam optics, and sample arrangement. A new end-station developed and installed in Forschungszen- trum Jülich provides these capabilities in combination with high sample throughput and result accuracy. Mechanical tolerances limit the device accuracy to 3% for RBS. Continuous pumping enables 5 × 10−8 mbar base pressure with vibration amplitudes < 0.1 µm. The beam optics achieves a demagniﬁcation of 24–34, suitable for µ-beam analysis. An in-vacuum manipulator enables scanning 50 × 50 mm2 sample areas with 10 nm accuracy. The setup features the above-mentioned IBA detectors, enabling a broad range of analysis applications such as the operando analysis of batteries or the post-mortem analysis of plasma-exposed samples with up to 3000 discrete points per day. Citation: Möller, S.; Höschen, D.; Custom apertures and energy resolutions down to 11 keV enable separation of Fe and Cr in RBS. Kurth, S.; Esser, G.; Hiller, A.; This work presents the technical solutions together with the quantiﬁcation of these challenges and Scholtysik, C.; Dellen, C.; Linsmeier, C. A New High-Throughput Focused their success in the form of a technical reference. MeV Ion-Beam Analysis Setup. Instruments 2021, 5, 10. https:// Keywords: ion-beam analysis; Rutherford-backscattering spectrometry; nuclear reaction analysis; doi.org/10.3390/instruments5010010 particle induced x-ray spectroscopy; material analysis

Received: 13 January 2021 Accepted: 25 February 2021 Published: 28 February 2021 1. Introduction Ion beam analysis (IBA) is a versatile group of methods using a broad range of pro- Publisher’s Note: MDPI stays neutral jectiles with kinetic energies per ion in the order of 1 to 10 MeV and ion currents in the with regard to jurisdictional claims in range of pA to µA to probe surface near ~10 µm concentrations of a sample composition. published maps and institutional affil- The method is, in most cases, non-destructive and reference free. It yields depth pro- iations. files of element and isotope amounts and concentrations when combined with computer based interpretation. Furthermore, a combination with scanning methods enables µm resolved compositional tomography when using µm sized ion-beam diameters (so-called “microbeam” variants). For these reasons, the applications of IBA span from historical Copyright: © 2021 by the authors. paintings over thin-film technology up to material development, with each application Licensee MDPI, Basel, Switzerland. having its special requirements on the analysis setups. This article is an open access article Usually, the measurements are conducted in vacuum due to the detrimental effects distributed under the terms and of matter on ion beams. The measurements rely on the detection of the products of the conditions of the Creative Commons ion-matter interaction. These can be secondary electrons, gamma rays, x-rays, recoils, scat- Attribution (CC BY) license (https:// tered projectiles, and nuclear reaction products. Many different techniques are available to creativecommons.org/licenses/by/ exploit the information hidden in these products [1]. The combination of the detection of 4.0/).

Instruments 2021, 5, 10. https://doi.org/10.3390/instruments5010010 https://www.mdpi.com/journal/instruments Instruments 2021, 5, 10 2 of 13

several products in parallel proved to be useful for determination of unambiguous results, since for example, backscattering from a heavy element present deep in the sample or from a light element present at the sample surface may result in the same spectra [2]. A combination with a particularly low amount of ambiguity and a straightforward integration is nuclear-reaction-analysis (NRA), Rutherford-backscattering-spectrometry (RBS), and particle-induced-X-ray emission (PIXE), also referred to as Total-IBA [3]. This combination can provide ppm detection limits and sub-% accuracy for light (NRA), intermediate (PIXE), and heavy (PIXE+RBS) elements in a self-consistent way, which are particular advantages compared to electron beam based analysis. These advantages in combination with the iso- topic sensitivity of RBS, NRA, and PIGE are particularly interesting for materials research, for example for nuclear fusion reactors [4–6]. For this set of methods, the ion-beam and the end-station properties deﬁne the analysis properties. A high intensity ion beam with a small beam diameter is required to obtain decent counting statistics (=result accuracy) and lateral spatial resolution (=beam spot size). This can be achieved with beam optics that ﬂexibly focus the beam onto the samples with the minimal spot sizes depending on the initial beam emittance from the accelerator and the properties of the beam optics. Numerous devices exist, regularly achieving spot sizes in the order of 1 µm by combining focusing with modern ion sources and accelerators, leading to high-brightness ion-beams [7,8]. This reference work introduces a new setup combining a magnetic quadrupole focusing system with a custom vacuum chamber and sample positioning system, which potentially realizes µm spot sizes with a special emphasis on high measurement throughput. The technical issues arising from sample positioning on the µm scale in a non-magnetic vacuum environment with high sample throughput is discussed and solutions are presented. Unfor- tunately, the available 1.7 MV Cockcroft–Walton tandem accelerator with a Duoplasmatron source (both dating 1983) limited the achievable spot-sizes of the presented device due to the 1–2 orders of magnitude larger energy-normalized emittance (>20 π mm mrad (MeV)1/2) compared to modern systems. This limits the minimal spot-sizes correspondingly.

2. Materials and Methods The new setup combines a quadrupole triplet focusing magnet, an UHV chamber, a vibration isolated table, and the detector setup connected to a beam-line and a tandem ion accelerator, as shown in Figure1. Generally, the relative alignment accuracy between a setup components and the sample limit the base accuracy and result reproducibility of an IBA setup. The counting performance improves with smaller dimensions and distances, since e.g., a smaller sample to detector distance increases the detector solid angle at a given detector size. This induces technical layout conﬂicts, since smaller dimensions limit sample dimensions and increase relative alignment inaccuracy with a given manufacturing tolerance. A technical setup providing precision in the percent range in combination with the possibility of frequent sample exchange, high sample throughput, and the requirements Instruments 2021, 5, x 3 of 13 on precise angular alignment and low ion doses required the development of a new analysis setup. This new setup is called µNRA and will be described in the following text.

Figure 1. Sketch of the µNRA setup with the required modules and the target (green star) attached Figureto a tandem 1. Sketch accelerator of the µ. NRABlue objects setup with depict the dipole required magnets. modules Orange andthe squares target depict (green ion star) beam attached fo- to a tandemcusing devices. accelerator. Blue objects depict dipole magnets. Orange squares depict ion beam focusing devices.

3. Vacuum Chamber and Beamline Figure 2 shows the design of chamber and its supporting systems. The manufacturing results in 0.1 mm tolerances at the points where an impact on measurement angles and distances is relevant, e.g., the ion-beam impact angle. The beamline is made from CF100 316 stainless steel vacuum parts and has a length from sample to switching magnet of about 5 m. The beamline is equipped with the identical turbo- and fore-pump as the µNRA chamber and reaches a base pressure of 2*10−9 mbar. Two Faraday cups and one beam profile monitor are installed for beam diagnostics, which can be moved out of the beam path during the sample analysis. An iron-core steering magnet powered by two remote-controlled power supplies (0.2 mA steps) after the first (object) aperture provides additional flexibility for the beam control. A second aperture is installed right before the focusing magnets (see Figure 2). Together these apertures collimate the beam in terms of dimension and divergence. Both apertures feature a fixed 5 mm diameter hole followed by four motor controlled blocks representing the four sides of the adjustable rectangular aperture. All parts are equipped with tantalum shields for radiation levels <10 µSv/h in the laboratory when operating with H, D, and He ions. Heavier ions cannot be accelerated to relevant energies.

Aperture Ion beam injection Magnets Pumping Camera observation system Camera view line

Samples

Detectors Sample installation door

Manipulator

(a) (b) Figure 2. CAD drawings of the µNRA setup (dotted square in Figure 1). (a) The µNRA vacuum chamber (flanges in yellow) together with the three focusing magnets and an aperture on a vibration-damping table. (b) Detailed crosscut of the vacuum chamber with sample manipulator, observation, charged particle detectors, and holder.

Instruments 2021, 5, x 3 of 13

Instruments 2021, 5, 10 Figure 1. Sketch of the µNRA setup with the required modules and the target (green star) attached3 of 13 to a tandem accelerator. Blue objects depict dipole magnets. Orange squares depict ion beam focusing devices.

3. Vacuum3. Vacuum Chamber Chamber and Beamline and Beamline FigureFigure 2 shows2 shows the design the design of chamber of chamber and and its supporting its supporting systems. systems. The The manufactur- manufacturing ing resultsresults in in 0.1 0.1 mm mm tolerances tolerances at at the the points points where where an an impact impact on on measurement measurement angles angles and and distancesdistances is is relevant, e.g., thethe ion-beamion-beam impact impact angle. angle. The The beamline beamline is is made made from from CF100 CF100316 316 stainless stainless steel steel vacuumvacuum partsparts and has a a length length from from sample sample to to switching switching magnet magnet of of aboutabout 5 m. 5 m.The The beamline beamline is equipped is equipped with with the identical the identical turbo turbo-- and andfore- fore-pumppump as the as the −9 µNRAµNRA chamber chamber and reaches and reaches a base a pressure base pressure of 2*10 of−92 mbar.× 10 Twombar. Faraday Two cups Faraday and cupsone and beamone profile beam monitor profile are monitor installed are installedfor beam fordiagnostics beam diagnostics,, which can which be moved can be out moved of the out of beamthe path beam during path the during sample the sampleanalysis analysis.. An iron An-core iron-core steering steering magnet magnet powered powered by two by two remoteremote-controlled-controlled power power supplies supplies (0.2 mA (0.2 steps) mA steps) after afterthe first the (object) first (object) aperture aperture provides provides additionaladditional flexibility flexibility for the for beam the beam control. control. A secon A secondd aperture aperture is installed is installed right rightbefore before the the focusingfocusing magnets magnets (see Figure (see Figure 2). Together2). Together these these apertures apertures collimate collimate the beam the beam in terms in terms of of dimension and divergence. Both apertures feature a fixed 5 mm diameter hole followed dimension and divergence. Both apertures feature a fixed 5 mm diameter hole followed by four motor controlled blocks representing the four sides of the adjustable rectangular by four motor controlled blocks representing the four sides of the adjustable rectangular aperture. All parts are equipped with tantalum shields for radiation levels <10 µSv/h in aperture. All parts are equipped with tantalum shields for radiation levels <10 µSv/h in the laboratory when operating with H, D, and He ions. Heavier ions cannot be accelerated the laboratory when operating with H, D, and He ions. Heavier ions cannot be accelerated to relevant energies. to relevant energies.

Aperture Ion beam injection Magnets Pumping Camera observation system Camera view line

Samples

Detectors Sample installation door

Manipulator

(a) (b)

Figure 2. CAD drawingsFigure of the 2. µCADNRA drawings setup (dotted of the square µNRA in setup Figure (dotted1). ( a) ThesquareµNRA in Figure vacuum 1). chamber(a) The µNRA (flanges vacuum in yellow) together with the threechamber focusing (flanges magnets in and yellow) an aperture together on with a vibration-damping the three focusing table. magnets (b) Detailed and an crosscut aperture of on the a vacuumvibra- chamber with sampletion manipulator,-damping observation, table. (b) Detailed charged crosscut particle of detectors, the vacuum and chamber holder. with sample manipulator, observation, charged particle detectors, and holder. The sample chamber is manufactured from 316 stainless steel and mechanically pol- ished from the inside. It features 18 CF flanges for feedthroughs, detectors, sample observation, and sample access together with an individually designed short flange for flexible connection of the beam tube going through the focusing magnets to the analysis chamber. Two self-assembled thermocouples of type K with 500 V isolation voltage and 1.5 K ac-

curacy, electrical wirings for in-vacuum connections for example of lithium batteries as depicted in Figure3, and a multi-pin feedthrough equipped with 0.2 mm Kapton isolated wires are installed here. The multi-purpose wiring enables for example charging and discharging of batteries during IBA measurement via an external cell tester. During the assembly of the beamline and chamber, special attention is paid on avoiding stray electro-magnetic fields in the beamline and the detector region in order to provide compact ion beams and reproducible scattering angles. Magnetic field measurements revealed the vacuum gauges (low stray field, Pfeiffer Vacuum PKR 360), the (electro- pneumatic) valve actuators, and the magnetic-bearing turbo-molecular pumps as critical sources of stray fields with strength in the mT range. To mitigate the stray field influence Instruments 2021, 5, x 4 of 13

Instruments 2021, 5, 10 4 of 13 The sample chamber is manufactured from 316 stainless steel and mechanically pol- ished from the inside. It features 18 CF flanges for feedthroughs, detectors, sample observation, and sample access together with an individually designed short flange for flexible onconnection the ion beam of the and beam charged tube going products, through these the devices focusing are magnets mounted to onthe extension analysis chamber. flanges providingTwo self-assembled sufficient distance thermocouples for the field of type amplitude K with to500 reduce V isolation below voltage 10 µT at and the 1.5 beamline K accu- centerracy, electrical and in the wirings measurement for in-vacuum chamber, connections respectively. for example The fields of lithium are measured batteries using as de- a hand-heldpicted in Figure magnetic 3, and compass a multi with-pin an feedthrough effective handling equipped resolution with 0.2 of 1mmµT Kapton (0.01 µT isolated digital resolution).wires are installed The field here. dropped The multi below-purpose the detection wiring enables limit at for a distance example of charging 450 mm and for thedis- turbo-molecularcharging of batteries pump during and 200 IBA mm measurement for the vacuum via an gauges external and cell actuators, tester. respectively.

FigureFigure 3. 3.Camera Camera sample sample image image of of an an in-situ in-situ contact contact of of an an all-solid all-solid state state lithium lithium thin-film thin-film battery battery (red (red circle) using a W contact needle (silver part on the left) attached to an isolated 2 pole DC feed- circle) using a W contact needle (silver part on the left) attached to an isolated 2 pole DC feedthrough through (not shown) with an outside cell tester. The cell to source wiring accounts to 1 Ω DC re- (not shown) with an outside cell tester. The cell to source wiring accounts to 1 Ω DC resistance. sistance. The beamline is mounted and aligned with the help of a leveling camera. All flanges are centeredDuring the to ± assembly1 mm (camera of the beamline accuracy) and around chamber, the idealspecial track, attention including is paid beamline, on avoid- magnets,ing stray apertures,electro-magnetic and the fields sample in the chamber beamline on itsand table the (Figuredetector2). region After in centering, order to thepro- chambervide compact is further ion beams corrected and reproducible for rotational scattering alignment angles. with theMagnetic help of field adjusting measurements screws andrevealed a laser the beam vacuum injected gauges 5 m away (low at stray the backside field, Pfeiffer of the Vacuumswitching PKR magnet 360), to the a precision (electro- ofpneumatic) 10−4 degree valve against actuators, the plane and ofthe vertical magnetic axis-bearing and beam turbo and-molecular furthermore pumps against as crit theical perpendicularsources of stray plane fields by with a level strength of 0.01 in◦ precisionthe mT range. rating. To After mitigate the beamlinethe stray field and chamberinfluence assembly,on the ion the beam sample and manipulatorcharged products, is installed these intodevices the chamber.are mounted The on degree extension of precision flanges achievedproviding in sufficient mounting distance allows for for fine the adjustingfield amplitude all parts to withreduce their below individual 10 µT at fine the adjusting beamline screwscenter inand the in next the stage.measurement chamber, respectively. The fields are measured using a handThermal-held magnetic drifts of compass the structures with an relevant effective for handling relative resolution beam spot of movements 1 µT (0.01 µT between digital theresolution). accelerator, The apertures, field dropped and samplebelow the are detection assessed bylimit ambient at a distance measurements of 450 mm over for one the weekturbo in-molecular summer. pump The room and temperature200 mm for the variations vacuum aregauges found and to actuators, be in the orderrespectively. of up to 1 K/hThe over beamline an operational is mounted day. Theand steepest aligned gradientwith the typicallyhelp of a observedleveling camera. lies between All flanges 10:00 toare 12:00. centered The to overall ±1 mm temperature (camera accuracy) variations around between the ideal 8:00 track and, 20:00, including the usualbeamline, working mag- hours,nets, apertures, stayed within and 2the K. sample The maximum chamber assumed on its table temperature (Figure 2). difference After centering is 1 K, as, the all metalcham- componentsber is further provide corrected decent for thermalrotational conductivity alignment forwith equilibration. the help of adjusting The measurement screws and also a revealedlaser beam changes injected of relative5 m away humidity at the backside in the range of the of switching 30–70%. Tomagnet avoid to humidity a precision related of 10- swelling,4 degree against only metals the plane were usedof vertical in the axis whole and support beam and structures. furthermore against the perpen- dicularFor plane ion beam by a measurementslevel of 0.01° precision requiring rating. long After acquisition the beamline times, aand relative chamber positional assem- driftbly, between the sample last aperturemanipulator and sampleis installed (which into are the both chamber. on the same The table) degree is ofconsidered precision asachieved the relevant in mounting quantity, allows since for the fine ion beamadjusting focusing all parts projects with their the aperture individual opening fine adjust- onto theing analyzedscrews in spotthe next on thestage. sample. Distance changes are not relevant, as they will be negligibleThermal compared drifts of to the the structures focal length relevant discussed for relative in the beam following spot movements sections. The between main concernthe accelera are verticaltor, apertures, displacements and sample induced are by assessed thermal by expansion ambient of measurements the support structures. over one Allweek materials in summer. below The the room table temperature surface are selected variations to beare identical. found to Thebe in part the aboveorder theof up table to 1 surface is assumed to induce the same displacement to all parts mounted on it with a relevant height of the beamline above the table of 150 mm. With the thermal expansion coefficient of aluminum (2.3 × 10−5/K) used for construction X-profiles minus the one of

Instruments 2021, 5, 10 5 of 13

stainless steel 316 (1.6 × 10−5/K), a maximum displacement of 1 µm is calculated as the upper limit neglecting temperature conduction.

4. Pumping System and Vibrations The pumping system of the vacuum chamber is constructed to provide a low base pressure (5 ± 2 × 10−8 mbar) for stable sample surfaces and ion energy losses <0.1 keV together with quick sample exchange and pump-down times in the order of 10 minutes. In addition, the pumping system has to avoid inducing magnetic stray fields and vibrations into the measurement chamber. The reduction of these quantities has an important role for positioning accuracy, charged particle analysis, and the measurement of small electrical signals in the detector system (vibration-induced voltages). Therefore, the vacuum system consists of three parts, subdivided by gate valves and elastic elements. These parts are the fore-vacuum pump, the turbo-pump, and the µNRA chamber. Every part is equipped with a vacuum gauge and a venting valve for safe access and easy maintainability. The pumping system is installed at a CF100 side port of the sample chamber. An edge-welded bellow separates beam-line and µNRA chamber. The rotary vane fore-pump is found primarily responsible for the vibrations. The turbo- molecular pump (Pfeiffer 300M) works on fully contactless magnetic bearings, strongly mitigating its external vibration amplitude. Rigid vacuum lines can transmit those vibrations to the samples and devices in the analysis chamber. For reducing this transmission, low stiffness vacuum components are installed with the goal of reducing the vibration amplitude well below 1 µm. To quantify the success of this approach, the vibrations amplitude is measured by a high-accuracy LK-H052 laser distance meter from KEYENCE. The displacement sensor works with a repeatability of 0.025 µm, a spot diameter of 50 µm, and a measurement-frequency of 50 kHz (maximum detectable vibration frequency of 25 kHz), enough to cover the first few harmonics of all possible mechanical frequencies. The laser is located on a stand next to the µNRA table. Due to the impact of the artificial lightning on the measurement, the measurements are conducted with lights off. The laser is positioned on the different analysis points. For each point, the distance is recorded for 60 s, yielding 3 million points. In steady-state, electronic noise and vibrations are assumed to feature different spectral distributions with vibrations following certain rotational frequencies such as 50 Hz. This assumption allows neglecting the white noise in the Fourier spectrum. For the evaluation, the data are corrected for a baseline-shift and the vibration amplitude is taken as the 3σ (99.73%) value of the real spectrum. First, vibrations of the concrete floor and the supports are excluded by Laser measurements on several static objects and walls. All measurements resulted in signals at the electronic noise limit. The amplitude of oscillation of the vacuum fore-pump in operation is measured to 45 µm for a 1-phase and 30 µm for a 3-phase motor of the same pump type. With its significantly lower vibration levels, the 3-phase pump is selected for all further measurements. The turbo-pump is selected as second measurement point, since it features a vibration amplitude above the detection limits. Fitting different components between fore-pump and turbo-pump enables further optimization of the vibration amplitude. Table1 lists the measurement results. Softer flexible components reduce the vibration transmission to a level of 1.6 µm, a factor 20 below the source level (fore-pump). The vibration amplitude at the samples will be significantly smaller, since another flexible element and the rigid support table of the vacuum chamber further mitigate the vibrations. At the sample position, no signal is detectable anymore in this situation equivalent to a vibration amplitude <0.1 µm. Instruments 2021, 5, 10 6 of 13

Table 1. Amplitude of oscillations at the turbo-pump with different damping components in between turbo-pump and fore-pump. Plastic clamps and edge-welded bellows reduce the transmission of the vibration to an acceptable level. The base amplitude of the fore-pump is 30 µm. The electronic noise background is 0.1 µm. Plastic connections reduce the vibration transmission and allow for a galvanic separation.

Connecting Tubing and Clamping Vibration Amplitude [µm] Al KF clamp Flexible hose - 3.18 Plastic KF clamp Flexible hose - 3.03 Al KF clamp Flex. hose + Edge welded bellow Al clamp 2.06 Plastic KF clamp Flex. hose + Edge welded bellow Plastic clamp 1.63 Al KF clamp Flex. hose + Edge welded bellow Plastic clamp 1.60

5. Optical Observation of Samples A FLEA3-FL3-U3-88S2C color camera with a 8.8 MPixel 1/2.5” detector is combined with a 0.28× tele-centric lens from Techspec with a working distance of 180 mm, providing a field of view of about 20 mm, a depth of sharpness of ±7 mm, and a tele-centricity of <0.03◦ for sample observation. The camera observes the samples through a vacuum window and a planar mirror positioned above the ion-beam tube in the µNRA chamber, see camera view line in Figure2b. The mirror position above the ion beam axis requires a rotation of the mirror normal of 30◦ against the sample normal (instead of the ideal 45◦). Due to the camera tele-centricity preserving all distances over the whole field of view, even in a rotated situation and for an offset between ion-beam and camera observation spot, samples are always observed virtually head-on. This eases beam position and size/area determination on scintillators exposed to the beam, but results in a thin sharpness line due to the mirror angle. No error will be connected to storing the beam outline on the camera image and navigating on non-scintillating samples. A position verification by the means of the camera system is required after each sample exchange, as mechanical tolerances and changes in the beam optics limit the accuracy of the relative positioning of samples and ion beam. A far-field microscope could have offered a better spatial resolution than the tele-centric lens, but the chosen solution provides the larger field of view, higher depth of field, tele-centricity, and also significantly lower costs. Imaging by beam scanning is also possible, but slow compared to the optical alignment. Upon installation, adjusting screws allow tilt, rotational, and distance adjustments by observing samples on the adjusted manipulator, see Section7 for further details. The tele-centric camera setup is first calibrated in a test stand with the help of a standardized vacuum compatible test pattern also used for the later beam resolution tests (Figure4). The best spatial resolution was found by varying the angle of incidence, lens aperture, the test patterns location in the camera field of view, the test patterns lighting, the lens aperture opening, and the distance between lens front and test pattern. The distance between two line centers on the test pattern compared to the number of pixels between two lines, yields the pixel width calibration, as shown in Figure4. Ten different groups are investigated with variations from 5.521 µm/pixel to 5.861 µm/pixel. On average, a pixel corresponds to 5.71 ± 0.12 µm, compared to nominal calibration of 5.5 µm/Pixel according to the datasheets. A change of the test patterns location in the camera field of view had no influence on this calibration, proving the lens’ tele-centricity. The optical spatial resolution is checked with regard to the angle of incidence, lens aperture, the test pattern location in the camera field of view, the test patterns lighting and the distance between camera front and test pattern (working distance). Similar to the spatial calibration, no influence of the observation angle on the spatial resolution was present, besides the effect of varying working distance over the field of view in case of non-perpendicular observation. The working distance is varied between 177 mm and 185 mm for perpendicular observation, with a nominal working distance of 180 mm. The spatial resolution is determined by checking which elements of the test pattern the camera are still separable by eye. The best spatial resolution is achieved at full aperture with Instruments 2021, 5, 10 7 of 13 Instruments 2021, 5, x 7 of 13

alens maximum aperture of opening, 19.7 + 0and− 2.1the µdistancem at a 180between mm distancelens front between and test camerapattern.lens The frontdistance and testbetween pattern. two Onlyline centers negligible on the resolution test pattern degradation compared occurs to the betweennumber aof 180pixels and betwe 183 mmen distancetwo lines, (Figure yields4 the). Thepixel manufacturer’s width calibration, statement as shown of in a Figure spatial 4. resolution Ten different of 10 groupsµm at reducedare investigated aperture withopening variations was not from achieved, 5.521 probably µm/pixel because to 5.861 the µm/pixel. lighting wasOn average, insufﬁcient a atpixel minimal corresponds lens aperture to 5.71 and± 0.12 camera µm, compared noise becomes to nominal dominating calibration at low of 5.5 light µ m/Pixel conditions ac- (seecording Figure to 5thea). datasheets. Unfortunately, A change the application of the test patterns in an ion-beam location analysis in the camera chamber field prohibits of view morehad no intense influence lighting on this due calibration, to a possible proving damage the oflens’ the tele light-centricity. sensitive radiation detectors.

FigureFigure 4.4. 1”1" × 3"3” Positive, Positive, USAF USAF 1951 1951 Wheel Wheel Target Target with with Chromium Chromium lines lines on on scintillating scintillating float ﬂoat glass. Eachglass. block Each containsblock contains a set of a triple set of lines triple with lines decreasing with decreasing size. The size. line The distances line distances and line and width line of each width of each triplet are known, thus the smallest resolvable line triplet yields the camera or ion triplet are known, thus the smallest resolvable line triplet yields the camera or ion beam resolution. beam resolution. The right image shows a pair of lines with 22 pixels center to center distance. TheAccording right image to the showsspecifications, a pair ofthere lines are with eight 22 pairs pixels of lines center per to mm center on this distance. triplet. According Therefore, toa the Instruments 2021, 5, x 8 of 13 speciﬁcations,pair of lines has there a distance are eight of 0.1 pairs25 mm of lines and the per pixels mm on are this calibrated triplet. to Therefore, a spatial width a pair of 5.68 lines µm. has a distance of 0.125 mm and the pixels are calibrated to a spatial width of 5.68 µm. The optical spatial resolution is checked with regard to the angle of incidence, lens aperture, the test pattern location in the camera field of view, the test patterns lighting and the distance between camera front and test pattern90 (working distance).Open Similar aperture to f/6the spatial calibration, no influence of the observation angle on the spatial resolution was present, 80 Half aperture f/14 besides the effect of varying working distance over the field of view in case of non-per- 70 pendicular observation. The working distance is varied between 177Minimum mm and aperture 185 mm f/22 for perpendicular observation, with a nominal working60 distance of 180 mm. The spatial resolution is determined by checking which elements50 of the test pattern the camera are still separable by eye. The best spatial resolution is achieved40 at full aperture with a maximum of 19.7 +0 −2.1 µm at a 180 mm distance between camera30 lens front and test pattern. Only negligible resolution degradation occurs between[ΜM] RESOLUTION a 180 and 183 mm distance (Figure 4). 20 The manufacturer’s statement of a spatial resolution of 10 µm at reduced aperture opening 10 was not achieved, probably because the lighting was insufficient at minimal lens aperture 176 178 180 182 184 186 and camera noise becomes dominating at low light conditions (see Figure 5a). Unfortu- DISTANCE CAMERA - TEST PATTERN [MM] nately, the application in an ion-beam analysis chamber prohibits more intense lighting due to(a )a possible damage of the light sensitive radiation(b) detectors.

FigureFigure 5. 5.( (aa)) DetailDetail ofof the USAF 1951 test test pattern pattern with with f/6 f/6 aperture. aperture. In Inthe the best best case, case, the the lines lines of the of thefifth ﬁfth element element of the of the fifths ﬁfths group group (box) (box) can canbe separated, be separated, giving giving a spatial a spatial resolution resolution of 19.7 of 19.7µm. µ(bm.) (bResolution) Resolution with with respect respect to toaperture aperture and and working working distance. distance.

6.6. Magnetic Magnetic Focusing Focusing System System AA micro-beam micro-beam has has to to provide provide a a maximum maximum ion ion current current at at a a minimum minimum beam beam diameter diameter forfor high high lateral lateral resolution resolution ion-beam ion-beam analysis. analysis. For For this this purpose, purpose, beam beam focusing focusing and and beam beam aperturesapertures are are required. required. The The Quadrupole Quadrupole triplet triplet consists consists of of three three OM-56 OM-56 type type magnets magnets from from thethe manufacturer manufacturer Oxford Oxford Microbeams Microbeams Ltd. Ltd. (Oxfordshire, with 10 mm UK) bore with and 10 100 mm mm bore length and per 100 mmmag- lengthnet. The per triplet magnet. is set The at triplet123 mm is distance set at 123 between mm distance the last between magnets the iron last core magnets and the iron sample core and(see the Figure sample 6). The (see magnets Figure6 have). The a maximum magnets have field a at maximum the pole tip ﬁeld 0.4 T. at They the pole can be tip aligned 0.4 T. Theywith can1 µm be alignedprecision. with The 1 µbeamm precision. current Theis specified beam current to 0.6 isnA speciﬁed for a spot to 0.6size nA of for1 µm a spot at a sizedistance of 1 µ ofm 1 at70 a mm distance between of 170 last mm magnet between and last target magnet by the and manufacturer target by the of manufacturer the magnets. ofPractically, the magnets. detection Practically, limits detectionand accuracy limits for and sub accuracy-µm depth for resolution sub-µm depth requires resolution ion currents >0.1 nA in order to achieve sufficient counts within measurement times in the order of hours which can be provided by the magnetic focusing system. Since the micro-focus in this setup is necessarily connected to a convergent beam, an angular spread of the impacting ions of up to 1.2° is present for the maximum specified demagnification and a 5 mm aperture, see Figure 6. This geometrically induces ranges of probing depth and energy-loss of the projectiles in the target. Typically, smaller apertures and less demagnification is required for the investigated samples resulting in values of ≤0.3° for most cases. This value and the resulting loss of energy resolution is negligible compared to the geometrical straggling induced by the finite detector’s width. Two high-stability power supplies with up to 3 V and 100 A with <10 mA ripple and <5 ppm drift power the magnets. Connecting the power supplies in different polarity and grouping allows producing vertically elongated, horizontally elongated, or symmetric ion-beam spots. In general, it has to be considered that, due to the axial length of each magnet of 100 mm, the maximum demagnification on the sample position is achieved with quite different current at each magnet. To compensate for this, the central magnet is powered in opposite polarity to the two outer magnets and the final magnet is powered individually. The magnets can be spatially aligned on the scale of ±1-2000 µm vertically and horizontally and <5° rotation around the beam axis with respect to the initial adjustment of their baseplate by adjusting individual micrometer screws at each magnet base. Before operation, all magnets are centered on the beam axis by individually connecting each magnet in both polarizations and reducing their horizontal and vertical steering and the rhomboid focus effect to zero with the help of these micrometer screws. The beam spot on the sample is observed using the sample observation camera and a scintillating LiAlO2 single crystal target irradiated by a 3 MeV proton beam of about 1 nA.

Instruments 2021, 5, x 9 of 13

For the final step of setting up the focusing magnet currents, different groups of the USAF 1951 test image (Figure 5) are targeted with the sample manipulator discussed in the next section. The magnet currents were increased and the beam fluorescence size is compared to the known line width of the test image groups. The other beam parameters remain constant. Smaller line groups are targeted until no light can be observed next to the metallized part. This procedure is repeated for horizontal and vertical line groups, leading to an accuracy in the beam size determination of about 5% due to the finite inter group differences. The procedure allows determining the beam size, even for beam spots smaller than the optical resolution of the camera and for high current density beams. The minimum spot size is determined to 50x35 µm² with an object aperture of 1200x1200 µm². Instruments 2021, 5, 10 Correspondingly, a demagnification of 24 and 34 is achieved, respectively. Smaller8 spot of 13 sizes can be achieved by narrowing the beam apertures resulting in smaller beam currents depending on the emittance and current density of the connected accelerator. The beam- opticalrequires aberrations, ion currents the >0.1 sample nA in to order magnet to achievedistance sufﬁcient, and the countsaccelerator within beam measurement brightness limittimes the in theminimum order of achievable hours which beam can spot be provided size. by the magnetic focusing system.

Figure 6. SketchSketch of of the the focusing focusing setup. setup. The The ion ion beam beam originates originates from from the right the right side. side. The object The object ap- ertureaperture defines deﬁnes its itssize. size. Three Three quadrupole quadrupole magnets magnets focus focus the the beam beam onto onto the the movable movable sample. sample.

7. SampleSince theStage micro-focus in this setup is necessarily connected to a convergent beam, an angular spread of the impacting ions of up to 1.2◦ is present for the maximum specified demagnificationA remote controlled and a 5 mm in-vacuum aperture, manipulator see Figure6 .(see This Figure geometrically 2) allows induces for mapping ranges the of probingsamples. depth This setup and energy-loss provides several of the advantages projectiles in over the ion target.-beam Typically, scanning, smaller as a spatial apertures cal- ibrationand less is demagnification not required, the is requiredion beam for can the always investigated stay in samplesthe idealresulting beam axis, in valuesa smaller of magnet≤0.3◦ for-sample most cases.distance This (no value scanning and the magnet) resulting and loss a larger of energy scanning resolution area are is negligibleachieved. Thecompared manipulator to the geometricalfeatures 3 linear straggling movement induced and by 1 thevertical finite rotational detector’s axis, width. all driven by piezoTwo-electric high-stability engines. The power rotational supplies axis with enables up to 3extended V and 100 sample A with selection <10 mA and ripple an andad- <5justment ppm drift of the power projectile the magnets. impact angle, Connecting e.g., for the channeling power supplies features in with different single polarity crystals and or gracinggrouping incidence allows producing analysis. The vertically manipulator elongated, offers horizontally manipulation elongated, ranges orof symmetric±25 mm in ion-lin- earbeam axes spots. with In a general, positioning it has accuracy to be considered of 10 nm that, and duea position to the axialresolution length of of 1 each nm magnetand an infiniteof 100 mm, rotational the maximum axis. The demagnification manipulator employs on the sample non-magnetic position ma is achievedterials and with engines quite withdifferent a dynamic current at range each in magnet. movement To compensate of about 10 for6 and this, movement the central magnetspeeds ofis powered nm/s to 10in mm/s.opposite polarity to the two outer magnets and the final magnet is powered individually. The magnetsA significant can be handling spatially advantage aligned on theis the scale overloading of ±1–2000 toleranceµm vertically of the and piezo horizontally engines, preventingand <5◦ rotation damage around to the theengine beam when axis too with much respect force tois induced the initial dur adjustmenting mounting. of their The manipulatorbaseplate by adjustingis calibrated individual and referenced micrometer by screwsits internal at each positioning magnet base. sensors, Before which operation, were furtherall magnets verified are centeredby the tele on-centric the beam camera. axis by The individually manipulator connecting can handle each weights magnet of in 300 both g polarizationsto 600 g, limited and by reducing the capabilities their horizontal of the motor and (± vertical150 g engine steering force), and including the rhomboid the sample focus holder.effect to This zero weight with the tolerance help of theserange micrometer was selected screws. for the The typical beam samples, spot on thebut sampleit can be is tailoredobserved to using other therequirements sample observation by changing camera the counter and a scintillating-acting spring LiAlO of the2 singlemanipulator. crystal Ontarget the irradiated manipulator, by a different 3 MeV proton types beamof sampl of aboute holders 1 nA. can be installed on a self-centring adapterFor to the fit final either step many of setting small samples up the focusing or fewer magnet and larger currents, samples. different A so-called groups carrou- of the selUSAF holder 1951 offers test image40 positions (Figure for5) 10 are × targeted12 × 5 mm³ with PSI the-2 [9 sample] standard manipulator samples. The discussed rotational in the next section. The magnet currents were increased and the beam fluorescence size is compared to the known line width of the test image groups. The other beam parameters remain constant. Smaller line groups are targeted until no light can be observed next to the metallized part. This procedure is repeated for horizontal and vertical line groups, leading to an accuracy in the beam size determination of about 5% due to the finite inter group differences. The procedure allows determining the beam size, even for beam spots smaller than the optical resolution of the camera and for high current density beams. The minimum spot size is determined to 50 × 35 µm2 with an object aperture of 1200 × 1200 µm2. Correspondingly, a demagnification of 24 and 34 is achieved, respectively. Smaller spot sizes can be achieved by narrowing the beam apertures resulting in smaller beam currents depending on the emittance and current density of the connected accelerator. The beam- optical aberrations, the sample to magnet distance, and the accelerator beam brightness limit the minimum achievable beam spot size. Instruments 2021, 5, 10 9 of 13

7. Sample Stage A remote controlled in-vacuum manipulator (see Figure2) allows for mapping the samples. This setup provides several advantages over ion-beam scanning, as a spatial calibration is not required, the ion beam can always stay in the ideal beam axis, a smaller magnet-sample distance (no scanning magnet) and a larger scanning area are achieved. The manipulator features 3 linear movement and 1 vertical rotational axis, all driven by piezo-electric engines. The rotational axis enables extended sample selection and an adjustment of the projectile impact angle, e.g., for channeling features with single crystals or gracing incidence analysis. The manipulator offers manipulation ranges of ±25 mm in linear axes with a positioning accuracy of 10 nm and a position resolution of 1 nm and an infinite rotational axis. The manipulator employs non-magnetic materials and engines with a dynamic range in movement of about 106 and movement speeds of 1 nm/s to 10 mm/s. A significant handling advantage is the overloading tolerance of the piezo engines, preventing damage to the engine when too much force is induced during mounting. The manipulator is calibrated and referenced by its internal positioning sensors, which were further verified by the tele-centric camera. The manipulator can handle weights of 300 g to 600 g, limited by the capabilities of the motor (±150 g engine force), including the sample holder. This weight tolerance range was selected for the typical samples, but it can be tailored to other requirements by changing the counter-acting spring of the manipulator. On the manipulator, different types of sample holders can be installed on a self-centring adapter to fit either many small samples or fewer and larger samples. A so-called carrousel holder offers 40 positions for 10 × 12 × 5 mm3 PSI-2 [9] standard samples. The rotational axis enables selection of the 8 sample slots in this holder type. The samples are fixed in a profile avoiding contact of the front surface. A flat plate holder features two slit-like slots of 50 mm height and 5–10 mm width for clamping of samples between the plate and a backside spring-supported bar. This method conserves the sample to detector geometry with varying sample thickness. Alternatively, the sample thickness has to be known and corrected for by displacing the holder equally. The rotational and tilt angles of the manipulator are aligned using screws for the tilt and a correction factor for the rotational axis. Their offset is determined by laser reflection from a mirror-finished sample mounted on the sample holder. For this purpose, a laser beam is injected into the back of the switching magnet. The laser beam travels through the 5 m long ion-beam line through a 2 mm aperture onto the sample chamber and is then reflected back from a mirror mounted on the sample holder. The manipulator tilt and rotation angles are corrected until the laser reflects back through the 2 mm entrance hole of the ruler at 100 mm distance. This leads to an angular accuracy of the alignment of ±0.2◦. Therefore, the ion-beam impact angle has the same accuracy, but the detector has an additional uncertainty stemming from the relative alignment uncertainty of the ruler and the detector holder of ±1 mm. These tolerances in the ion-beam impact angle can lead to deviations of the Rutherford cross-sections in the order of 3%. The deviations of other cross-sections might be smaller and a general statement for the result uncertainty related to geometrical uncertainties is impossible. RBS generally features the lowest uncertainties among the four methods used here, due to the known and high cross-sections. Combining their results yields a total uncertainty larger than the largest individual uncertainty due to the rules of error propagation. Therefore, we state a minimal uncertainty of 3% for the results of the presented setup.

8. Ion Beam Analysis Detector Setup The setup contains in total four detectors as depicted in Figure7. For charged particle detection for nuclear reaction analysis (NRA) and backscattering spectrometry (RBS and EBS), two silicon detectors are installed at reaction angles of 150 ± 2◦ for separating different particle species. The detector surface has a distance to the analysis/beam spot of 32.5 ± 0.5 mm. Light induced noise in the detectors is avoided by a full coverage of all windows during measurement. A high-resolution detector with 11 keV FWHM at 1.6 µs Instruments 2021, 5, 10 10 of 13

rise-time (or 16 keV at 0.6 µs), 50 mm2 size, and 100 µm nominal active thickness detects the full energy of protons up to about 3.3 MeV and α’s up to about 13 MeV. Figure8 shows the excellent resolution of this detector via the separation of Fe and Cr signals in a steel. The actual thickness of this detector is around 300 µm as full energy measurements of 6 MeV protons from NRA reactions suggest. The second detector has a 1500 µm nominal active thickness, 150 mm2 size, and 22 keV FWHM resolution at 1.4 µs rise-time. This detector is equipped with a stopping foil of Polyimide of 25 µm thickness coated with 30 nm Al to block incident light and α-particles below about 5 MeV. Both detectors can be equipped with 2 mm thick steel apertures in the shape of a segment of the 150◦ circle to adapt count rates and limit geometrical straggling. The effective detector solid angle ratio between both detectors and different aperture options are experimentally calibrated using simultaneous acquisition of the 7Li(p, α)4He reaction signal of a thin lithium containing ﬁlm on a silicon substrate to an accuracy of 1% (>10,000 counts) for all cases with e.g., 3.52 ± 0.04 times reduced solid angle for a 1.5 mm aperture on the high-resolution detector and an open InstrumentInstruments 2021s 2021, 5, ,x 5 , x thick detector. The connected RC pre-ampliﬁers limit the usable count rates up11 to 11of a few13of 13

10,000/s with signiﬁcant degradation of energy resolution at higher values.

FigureFigureFigure 7. 7.Sketch 7.Sketch Sketch of of theof the the detector detector detector and and and chamber chamber chamber arrangement. arrangement. arrangement.

simulated simulated 50005000 Cr Cr Fe Fe W 40004000 W experimental experimental

30003000

counts counts 20002000

10001000

0 0 20002000 21002100 22002200 23002300 24002400 25002500 26002600 27002700 28002800 EnergyEnergy [keV] [keV]

3 FigureFigureFigure 8. 8.RBS 8.RBS RBS detector detector detector signal signalsignal using usingusing 2960 29602960 keV keVkeV 3He 3He Heions ionsions on ononHiperFer HiperFerHiperFer 17Cr 17Cr17Cr steel steel with with <2% <2% dead dead dead-time-time-time togethertogethertogether with with with SimNRA SimNRA SimNRA evaluation evaluation. evaluation. The The. The setup setup setup allows allows allows for for fora aseparation a separation separation of of theof the the Fe Fe andFe and and Cr Cr Crsignals signals signals with with with a a clearcleara clear edge edge edge at at ~2350 ~2350at ~2350 keV keV, keV, yielding , yielding a matching a matching Cr Crconcentration concentration in thein the st steel.eel. steel.

ForFor particle particle induced induced x-ray x-ray analysis analysis (PIXE), (PIXE), a KETEK a KETEK AXAS AXAS-D -siliconD silicon-drift-drift detector detector of of 450450 µm µm active active thickness thickness and and an an8 µm 8 µm thick thick Beryllium Beryllium foil foil counts counts the the emitted emitted X- rays.X-rays. The The detectordetector allows allows for for count count rates rates up up to 200to 200,000/s,000/s for for <10% <10% deadtime deadtime and and <135 <135 eV eV resolution. resolution. TheThe detector detector normal normal points points at theat the sample sample surface surface with with an anangle angle of 62.5of 62.5° from° from a port a port in thein the upperupper plane. plane. Its Itsdistance distance to tothe the sample sample can can be beadjusted adjusted from from 30 30to to200 200 mm mm to tooptimize optimize countcount rates. rates. The The detector detector is calibratedis calibrated by byK andK and L linesL lines from from high high purity purity W, W, Mo Mo and and Y Y targets.targets. Its ItsH factorH factor (energy (energy dependent dependent sensitivity) sensitivity) is determinedis determined by bythe the same same samples samples in in combinationcombination with with normalization normalization by bythe the respective respective RBS RBS signals. signals. A HPGeA HPGe detector detector with with 76.2 76.2 mm mm crystal crystal diameter diameter is installedis installed at theat the backside backside flange flange of of thethe chamber chamber as asseen seen by bythe the ion ion beam beam for for particle particle induced induced gamma gamma-ray-ray analysis analysis (PIGE). (PIGE). In In thisthis geometry, geometry, the the photons photons will will be bedetected detected only only after after passing passing the the sample sample and and the the samp sample le holder,holder, but but for for the the relevant relevant photon photon energies energies and and the the holder holder construction, construction, only only a minora minor absorptionabsorption can can be beexpected. expected. For For the the efficiency efficiency calibration, calibration, known known lithium, lithium, aluminum aluminum, and, and boronboron samples samples are are exposed exposed together together with with RBS RBS and and NRA NRA measurements measurements for for cross cross-ca-libra-calibration.tion. The The system system achieves achieves an anoverall overall detection detection efficiency efficiency in thein the order order of 1%of 1% (478 (478 keV keV pho- pho- tons).tons).

Instruments 2021, 5, 10 11 of 13

For particle induced x-ray analysis (PIXE), a KETEK AXAS-D silicon-drift detector of 450 µm active thickness and an 8 µm thick Beryllium foil counts the emitted X-rays. The detector allows for count rates up to 200,000/s for <10% deadtime and <135 eV resolution. The detector normal points at the sample surface with an angle of 62.5◦ from a port in the upper plane. Its distance to the sample can be adjusted from 30 to 200 mm to optimize count rates. The detector is calibrated by K and L lines from high purity W, Mo and Y targets. Its H factor (energy dependent sensitivity) is determined by the same samples in combination with normalization by the respective RBS signals. A HPGe detector with 76.2 mm crystal diameter is installed at the backside flange of the chamber as seen by the ion beam for particle induced gamma-ray analysis (PIGE). In this geometry, the photons will be detected only after passing the sample and the sample holder, but for the relevant photon energies and the holder construction, only a minor absorption can be expected. For the efficiency calibration, known lithium, aluminum, and boron samples are exposed together with RBS and NRA measurements for cross-calibration. The system achieves an overall detection efficiency in the order of 1% (478 keV photons). A quantitative analysis of the sample composition by ion beam analysis requires determining the Particle*Sr value (=incident ion dose * detector solid angle). This value is the product of detector solid angle and incident ion charge/dose. The ion current is measured using a Keithley 6487 pico-ampere meter and integrated by software to a precision <0.1%. The actual relevant (systematic) uncertainty in the collected charge arises from the emission of secondary electrons (SE) by the ion beam impact. For suppressing these SE and measuring only the ion current, an electric field driving the SE back onto the sample is required. In the current setup, the sample holder is isolated against the surrounding structures and the (metallic) sample holder itself is biased to −10 to +500 V against the ground using the voltage sources of the pico-ampere meter. When adding a conductive tape to the sample front and operating with

Table 2. Measured sample current for two different ion beam settings on different materials. In both cases, the current increases with negative and decreases with positive biasing of the sample holder. At around +80 V bias, the secondary electrons seem to be completely suppressed in both cases.

Bias [V] −10 −4 −2 −1 0 +1 +3 +10 +30 +50 +60 +80 +100 1.4 MeV 4He+ [nA] 13.1 12.2 11.4 10.4 9.1 7.7 5.8 3.1 2.4 2.28 2.22 2.22 2.22 2.96 MeV 3He+ [nA] 9.1 8.4 7.2 3.2 2.6 1.65 1.5 1.4 1.4

9. Conclusions A new setup for ion-beam analysis on the µm scale is developed, tested, and eval- uated. The setup combines state-of-the-art industrial products to a reliable device with easy handling, high precision, and high throughput in the order of 40 samples or about 3000 points on one sample per day, as already demonstrated in [4]. The setup consists of four independent parts, the vacuum system, the magnetic focusing, the sample observation and positioning, and the detectors for ion-beam analysis. Special attention is paid on main- taining a vacuum below 10−7 mbar without inducing stray magnetic ﬁelds and vibrations into the sample chamber. In the sample chamber, a combination of a ﬁxed beam axis with an Instruments 2021, 5, 10 12 of 13

nm precision piezo-electric manipulator and a high-resolution tele-centric camera proved to be reliable and practical for sample positioning. Four detectors for charged particles, γ-rays, and X-rays with special apertures and stopping foils enable a Total-IBA analysis with high depth resolution, sub-% accuracy, and ppm detection limits using 0.01 to 50 nA ion beam currents. The good energy resolution in combination with the low geometrical straggling enable a surface near separation of Fe and Cr in steels using RBS. Generally, the RBS and NRA detector count-rates limit the performance of the presented setup due to excessive dead-time or loss of energy resolution in the connected amplifier electronics. The installation and adjustment of the device posed a particular challenge, since sub-mm and sub-degree precision is required in the alignment of the different parts. The chosen compact arrangement improves the analytical performance, but also increases the requirements on accuracy. Several steps using levels, leveling cameras, lasers, and adjusting screws with two stages of accuracy on several spots yielded success in alignment. The setup precision in angles and distances is sufficient to deliver 3% accuracy in the ion beam measurements. The final step required alignment of the magnets on the µm scale with direct ion beam observation and allowed to obtain the specified demagnification of about 24. The use of a test pattern on (scintillating) float glass allowed the determination of beam position and size. The connected tandem accelerator from 1984 limits the achievable beam spot sizes to a few 10 µm, but a modern accelerator would allow for beam spots in the order of 1 µm due to increased beam brightness. For the future, several projects on depth analysis with side cuts, grain resolved studies, ppm sensitivity elemental analysis, and in-situ experiments of battery cells and a commercial product based on this design are planned. As an experimental setup, the technical details and implemented devices will change and improve over time, but the principle choice in the design space will remain.

10. Patents The patent WO002019219103A1 relates to this publication: Sören Möller 1,*, Daniel Höschen 1, Sina Kurth 1, Gerwin Esser 1, Albert Hiller 1, Christian Scholtysik 2, Christian Dellen 1, and Christian Linsmeier 1.

Author Contributions: Conceptualization, S.M., D.H., and C.L.; methodology, S.M.; software, S.M.; investigation, S.M., S.K., and C.D.; resources, C.S., G.E., A.H., and C.D.; data curation, S.M.; writing—original draft preparation, S.M.; writing—review and editing, S.M.; visualization, S.M.; supervision, C.L.; funding acquisition, C.L. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Data Availability Statement: Data sharing is not applicable. Conﬂicts of Interest: The authors declare no conﬂict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References 1. Möller, S. Accelerator Technology: Applications in Science, Medicine, and Industry; Springer International Publishing: New York, NY, USA, 2020. 2. Wang, Y.; Nastasi, M. Handbook of Modern Ion Beam Materials Analysis; Materials Research Society Warrendale: Warrendale, PA, USA, 2009. 3. Jeynes, C.; Bailey, M.J.; Bright, N.J.; Christopher, M.E.; Grime, G.W.; Jones, B.N.; Palitsin, V.V.; Webb, R.P. “Total IBA”—Where are we? Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. At. 2012, 271, 107–118. [CrossRef] 4. Möller, S.; Ding, R.; Xie, H.; Gao, B.F.; Wang, B.G.; Peng, J.; Liu, S.C.; Gao, W.; Kirschner, A.; Breuer, U.; et al. 13C tracer deposition in EAST D and He plasmas investigated by high-throughput deuteron nuclear reaction analysis mapping. Nucl. Mater. Energy 2020, 100805. [CrossRef] 5. Möller, S.; Krug, R.; Rayaprolu, R.; Kuhn, B.; Joußen, E.; Kreter, A. Deuterium retention in tungsten and reduced activation steels after 3 MeV proton irradiation. Nucl. Mater. Energy 2020, 23, 100742. [CrossRef] Instruments 2021, 5, 10 13 of 13

6. Möller, S.; Kuhn, B.; Rayaprolu, R.; Heuer, S.; Rasinski, M.; Kreter, A. HiperFer, a reduced activation ferritic steel tested for nuclear fusion applications. Nucl. Mater. Energy 2018, 17, 9–14. [CrossRef] 7. Simiˇciˇc,J.; Pelicon, P.; Budnar, M.; Šmit, Ž. The performance of the Ljubljana ion microprobe. Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. At. 2002, 190, 283–286. [CrossRef] 8. Malmqvist, K.G.; Hyltén, G.; Hult, M.; Håkansson, K.; Knox, J.M.; Larsson, N.O.; Nilsson, C.; Pallon, J.; Schoﬁeld, R.; Swietlicki, E.; et al. Dedicated accelerator and microprobe line. NIMB 1993, 77, 3–7. [CrossRef] 9. Kreter, A.; Brandt, C.; Huber, A.; Kraus, S.; Möller, S.; Reinhart, M.; Schweer, B.; Sergienko, G.; Unterberg, B. Linear Plasma Device PSI-2 for Plasma-Material Interaction Studies. Fusion Sci. Technol. 2015, 68, 8–14. [CrossRef]