Joint Research Activities

WP20 – JRA1 – Ion Sources and Beams

Timeframe within RADIATE: Months: 1-40
Partner involved: UNIVIE, HZDR, ATOMKI-HAS, CNRS, ETH Zürich, Ionplus, KU Leuven, OP, RBI, UNI BWM

Improving the beam characteristics and broadening the diversity of ion species available to users is central for all applications of ion beams: increasing beam intensity and brightness (reducing spread in space, direction, and energy) for faster, more accurate modification or characterization (T20.1 to T20.3); decreasing beam spot size on target down to the nm range for high resolution imaging (T20.1); achieving these improvements over a wider energy range (from MeV to few eV – T20.1 to T20.4) and over a wider range of ion species, in particular towards heavier elements (T20.1 and T20.2), thereby extending the fields of application for the existing techniques (all tasks) as well as for new ones (e.g. ultra-low energy ion implantation of 2D materials, T19.4) currently under development.

Since ion beam optics is a crucial, common component of these developments, an Ion Optics Pool of expertise will be established within RADIATE (administrated by ATOMKI and UW). This pool will centralize and further develop the expertise (currently scattered among RADIATE partners) on ion optics hardware and on state-of-the-art software for ion optics simulations (COMSOL, WinTRAX and SIMION), as well as on related fields of relevance (electronics, vacuum, molecular gas dynamics, mechanical design, etc.) (D20.1).

This task will ensure that European users are provided with state-of-the art ion beams constantly being upgraded, following the pace of international progress in this field. Smaller spot size is generally beneficial but can more easily be achieved for lighter ions (protons) and low intensities. For intense proton beams, typically used for PIXE imaging with micrometer resolution, the spot size will be reduced to the level of 300 nm and become routinely available for TA (D20.3). Improved ion optics for stronger demagnification of the beam, and careful elimination of parasitic effects like electromagnetic stray fields, mechanical vibrations and thermal drifts, are the key ingredients. Existing setups (RBI, JSI, Atomki, RBI, CNRS) will be optimized by identification and elimination of weaknesses.

The design of new instrumentation will be supported by the Ion Optics Pool. To satisfy the increasing user demand for heavy ion beams with small spot sizes, the developments will be pushed towards spot sizes around 100 nm and below 1000 nm for low and high-intensity beams, respectively (D20.4). The antagonism between size and intensity is imposed by the lower beam brightness of heavy ions. Brightness improvements of the Caesium Sputter Ion Source, which is the workhorse for heavy ion production (T20.2), will contribute to this goal. Where intensity losses are acceptable, collimation will be explored as a means to further decrease the beam diameter. CNRS together with industry partner Orsay Physics will use their expertise in producing ultra-thin capillaries to replace the standard apertures of the existing “PELIICAEN” platform. This will not only vastly reduce beam size, but also extend the life time of the aperture.

The development of apertures with long lifetimes and customer-tailored size will considerably expand the portfolio of techniques offered in TA (D20.5) but will simultaneously have a major impact on the product range of Orsay Physics.

In this task, a commercially available (Ionplus) Caesium Sputter Source for negative ions will be further developed to deliver higher intensity, brightness, and negative ionization yield for a wide variety of elements, from Be to the actinides.
This will allow for higher intensity in smaller beam spots for heavy-ion beam analysis and modification and will improve the sensitivity of AMS. An experimental program will be launched to characterize negative ion beams from various ion sources of the consortium partners. Based on this data, ETHZ will develop simulation tools to model generation of ion beams under realistic operation conditions. UW will upgrade its ion source, in order to increase the throughput of TA measurements, and has a scientific interest in clarifying the underlying physical processes in the ion source (i.e. surface versus plasma ionization), which is a prerequisite for simulations. Studies will include methods to enhance yields by optical methods (i.e. laser ablation and optically stimulated ionization). Based on designs exhibiting promising improvements, such as higher primary sputter beams, prototypes will be constructed at Ionplus (D20.2), which will also contribute engineering capacity to modify components from existing instruments to boost performance. The general aim of Ionplus is to provide more intense heavy ion sources, as a commercial product for the general ion beam community.

The detection of natural 41Ca for AMS studies, e.g., in geomorphology, will be enabled at the HZDR DREAMS facility by improving the separation from interfering 41K which so far has only been possible at the largest available accelerators, as, e.g., the Munich Tandem accelerator. The advantages will be lower instrumental effort and higher overall yield. The successful extension of the capability of DREAMS will be demonstrated for titanium- and iron-rich mineral fractions, for which the production of 41Ca by fast neutrons is well understood. In parallel, based on the strong expertise in sample preparation at HZDR and UBW, an efficient routine process to prepare CaF2 sputter samples for the production of CaH3- beams will be established, so that 41Ca AMS can be made available for TNA (D20.6). ETHZ as well will carry out performance tests with CaH2 using smaller accelerators.

Ion cooling technology has tremendous untapped potential as a solution for various current limitations of ion beam techniques (e.g. to increase sensitivity and applicability of AMS and the reliability of ultra-low energy ion implantation).

In an ion cooler, beams are electrostatically decelerated and cooled in a buffer gas. Although the technology is well established at radioactive ion beam facilities (dedicated to experimental nuclear physics), broader application in ion beam analysis (IBA) and modification of materials is still in its infancy, and no commercial provider exists. An ion cooler has recently been developed at UW for AMS applications, based on Ion-Laser Interaction Mass Spectrometry.

The device is used to decelerate negative ions, which greatly increases the interaction time with a laser beam (up to few milliseconds) required for isobar separation. By significantly increasing sensitivity, this approach has the potential to more than double the number of ultra-trace isotopes available to AMS users, boosting numerous new application fields and attracting new user com-munities. The exploratory prototype shows good results, and so far, was commissioned for 36Cl, a well-established AMS isotope. Methods for new isotopes (135Cs and 93Zr) will be developed in the framework of this task, becoming available for TNA within the project duration (D20.7). In this task, KUL will explore the use of ion cooling for ultra-low energy (ULE) ion implantation (down to few eV), to enable a completely new field of ion beam modification of 2-dimensional materials (e.g. synthesis and doping of graphene and MoS2). Although commercial ion implanters provide reliable implantation conditions for conventional materials down to hundreds of eV, recent research has shown that controllable modification of 2D materials require even lower energies (for effective implantation within the first atomic monolayers, with minimal ion beam damage). Existing ULE implantation facilities (HZDR and KUL) are based on ion sources with low energy spread and electrostatic deceleration on target.

However, this approach cannot deliver the required beam specifications (energy spread below 1 eV, uniform and reproducible implantation in wafer scale) in the ultra-low energy extreme. By directly decreasing the energy spread and the beam emittance, ion cooling has the potential to revolutionize ULE ion implantation of 2D materials, enabling a vast new user community in a field of extreme strategic relevance in the European context. In this task, an ion cooler for ULE ion implantation will be designed (MS20.1), developed and commissioned at KUL (D20.8). The pool of ion optics expertise described above is fundamental for this task, which will also benefit from the abovementioned experience of UW in ion cooling, and of ETHZ on gas dynamics in a similar pressure regime (developed in the context of ion stripping).

In a broader context, by decreasing the energy spread and emittance, ion cooling dramatically increases beam brilliance, and will therefore be considered as an alternative to the developments described in T20.2, for applications where the
gain in brilliance justifies the added complexity of ion cooling.

WP21 – JRA2 – Detectors and Electronics

Timeframe within RADIATE: Months: 1-36
Partners involved: RBI, ATOMKI-HAS, CAEN, CNRS, IMEC, INFN, Ionoptika Ltd, IST, JSI, JYU, SURREY

The development of efficient and reliable detection techniques with high resolution, sensitivity, throughput and durability are key elements to successful and wider application of IBA techniques. The increasing complexity and sensitivity of detection hardware, as well as the fast-growing amount of information that signals carry, require the simultaneous advancement of associated electronics and data acquisition hardware. Particular attention will be given to lower ion energies, both for applications in characterization and single ion implantation. The technical developments performed in WP20 will result in a significant extension of the portfolio of TA services available to users within the duration of RADIATE. Furthermore, the close collaboration of IONOPTIKA and CAEN (industrial partners) with the academic partners who are at the forefront of their specific domains, will allow for an effective distribution of the resulting new products and capabilities to the wider ion beam community and user groups.

Deterministic single-ion implantation has become a Holy Grail in the field of quantum science and technology. However, the capability to position atoms of any element into any material, with a high spatial resolution, and counted individually, will require major advances in reliable detection of single ion impacts. Whereas >99% efficiency of hit detection may be achieved for high-energy heavy ions (e.g. through emission of secondary electrons), the detection of impacts of low energy and light ions is far more demanding. In the low energy ion range (<20 keV), where the high spatial resolution (<20 nm), can be obtained using Focused Ion Beam technology, efficiency of single ion detection will be improved and tested using new secondary ion detection systems (SUR, IONOPTIKA) and using highly charged ions (HZDR), becoming available for TA (D20.1).

In the range of ion energies above 100 keV, different secondary products (ions, electrons, photons) and low noise detectors, will be explored by ATOMKI and RBI. For the development of future quantum technologies, ions may be also detected using detectors embedded in the target substrate. For these applications, low noise IBIC systems will be developed and tested for substrates of importance for quantum technology applications, as diamond, Ga2O3, GaN, SiC, etc. (RBI, IST).

The outcome of the developments and tests carried out in this task, in terms of efficiency for single ion hit detection under different experimental conditions (ion species, ion energy, detection system) will be compiled in a report (D21.6).

Recent developments of energy-dispersive CCD x-ray cameras and transition edge sensor (TES) x-ray microcalorimeters have opened a number of avenues for new or improved instruments that exploit the characteristic xrays emitted by materials under ion beam bombardment.

The position sensitivity of CCD cameras will be exploited together with crystal diffraction elements to achieve high resolution PIXE (JSI, RBI) allowing dis-crimination of closely overlapping x-ray emission peaks and sensitivity to the chemical state of elements. A compact CCD camera-based invacuum
spectrometer will be developed for PIXE microprobe (RBI). In the soft x-ray regime, the energy resolution will be extended further by coupling the CCD with curved focusing crystals to reach sub-natural linewidth resolution and
perform chemical state analysis of bulk materials (JSI).

Polycapillary x-ray optics will be combined with flat crystal diffraction (JSI for in-air microPIXE (D21.7). The use of micropore optics coupled directly to a CCD detector will be explored for imaging soft x-ray emission (Soft-PIXE) from a uniformly irradiated field of view (CNRS).

An alternative approach to obtain high-resolution PIXE spectra is provided by TES microcalorimeters, in which the temperature increase in a small absorbing element is measured by a TES. Typically, arrays of such microcalorimeters are used, to achieve high overall count rates whilst maintaining very high energy resolution. Multi-pixel TES spectrometers will be developed for high resolution PIXE both in air (INFN, JYU) and in vacuum (JYU), and for high resolution PIXE imaging with polycapillary optics (JYU). The potential of high-resolution PIXE detectors for measurements in channeling mode (e.g. for studies on lattice location of dopant atoms in electronic materials) will also be explored (KUL).

Progress in particle detectors for IBA applications is driven by the need to separate close molecular masses (by MeV SIMS) and distinguish neighbouring elements in depth-resolved analysis (by RBS and ERD). Small fragmentation, high desorption efficiency and minimal sample pre-treatment have recently made MeV SIMS very attractive to new users over a wide range of applications.

In this task, its analytical capabilities will be further explored, in particular taking advantage of the possibility of combining it with other IBA techniques. Existing MeV SIMS systems will be upgraded to provide higher mass resolution. Reflectron mass analyser geometry will be implemented at RBI (D20.2). A feasibility study about coupling novel advanced orbitrap mass spectrometer with MeV ion excitation to perform MeV-SIMS will be carried out at JSI. High mass res-olution nano-electrospray ionization system (DAPNE) will be coupled to PIXE and secondary electron imaging, increasing at the same time positioning accuracy during the sampling of extremely low sample volumes (SUR). High depth resolution RBS will be developed using a magnetic spectrometer combined with a Si strip detector (IMEC, and made available for TA at KUL – D21.3) and time of flight spectrometer (JYU) systems.

Furthermore, the applicability of TOF ERD systems to be used with lower energy heavy ion beams will be explored at JYU and IMEC, in order to open this multi-elemental depth profiling technique to smaller accelerator laboratories.

This task is devoted to further developing the capabilities of ion beam analysis by combining multiple detectors and multiple techniques in the same measurement systems, also taking advantage of recent developments in detector and data acquisition technology.

Helium ion microscopy (HIM) has now reached the status of standard imaging technique, employing secondary electrons emitted by ~40 keV He ions. In this task, we will explore the combination of spatially resolved ion backscattering spectrometry (BS) with SIMS (HZDR) and Auger electron spectroscopy (JYU, D21.5) in He and Ne microscopes. The BS and SIMS measurements will greatly benefit from the development of pulsed gas-field ion source with pulse length below 5 ns (HZDR).

The efficiency and sensitivity of more widely used IBA techniques will also be enhanced by several developments.

Cost efficient multi-detector systems and electronics for ultra-high sensitivity Rutherford backscattering (IMEC, KUL, CNRS) and forward scattering (RBI, IST) will be developed and made available for TA (D21.4). For nuclear microprobe analysis of biological tissues, the efficiency will be improved by implementing annular segmented SDD covering solid angle of 1 strd (JSI). Options to use one SDD segment for detection of the backscattering (EBS) ions will be  experimentally explored. The development of integrated photon and charged particle detector chip (INFN, IST) will enable more efficient simultaneous PIXE and RBS measurements. A common development enabling larger throughput and multi-detector setups with common data acquisition will involve the use of digital pulse processing (all partners, including CAEN). This will also be promoted by the JRA activities on multiparameter data treatment in WP22.

WP22 – JRA3 – Software and Data Handling

Timeframe within RADIATE: Months: 1-36
Partners involved: UNI BWM, HZDR, CNRS, IMEC, IST, JYU, KU Leuven, SURREY

Next to hardware (WP20 and WP21), ion beam techniques are strongly based on software: to simulate beam-target interactions, enabling controlled modification of materials; to perform data analysis for quantitative characterization; and for automated control of ion beam hardware.

This work package deals with software developments to address the increasing demands of user communities. These new software demands originate either directly from new fields of research (ion beam modification of nanostructures and 2D materials – T22.1 – and characterization of nanostructures – T22.2) or indirectly, i.e. to cope with hardware developments which, in turn, were motivated by new user demands (e.g. multi-parameter filtering and analysis and deep learning algorithms to handle large amounts of data – T22.3 and T22.4).

These software developments are oriented towards broad dissemination across user communities and ion beam facilities beyond RADIATE. Therefore, although this work package deals with the research component of these activities, it is
directly linked with networking activities of WP3, WP6 and WP7.

Ion beam bombardment is one of the main tools to modify the properties of materials by controllable introduction of impurities and defects. At the moment there are two wide-spread approaches to simulate the non-equilibrium processes of ion impacts onto the targets.

The most common tools are TRIM and the high-fluence extension TRIDYN based on the binary collision approximation (BCA), which are computationally very efficient to predict ion ranges and estimate the number of defects produced even on a desktop computer. However, 2D and 3D shaped nanomaterials are not considered. Therefore, TRI3DYN has been developed as a 3D extension of TRIDYN, which, however, is presently only accessible by experts.

The second approach to model the non-equilibrium processes is to use molecular dynamics (MD) simulations, which include also detailed analysis of energy transfer and multi particle dynamics in a solid or fluid after an ion impact.
Compared to BCA simulations, MD simulations are computationally rather expensive, but allow obtaining a detailed understanding of materials modification by ion bombardment on an atomic level. Several MD codes are available for ion impact simulations, e.g. LAMMPS and PARCAS. However, simulations of irradiation effects in novel systems, including the modification of 2D materials with ultra-low energy ion implantation (T19.4) require further modification of these codes.

Within this task, the two programs will be further developed to obtain robust, generally applicable codes that can be used by non-experts. Together with the codes, common input and output data structures (WP6), user manuals and tutorials will be developed and made available via the Ion Beam Web Portal (WP3) (D22.1 and D22.3).

In this way, extended tools will be offered to users to tailor their ion implantation/irradiation parameters (e.g. energy and ion species) for the desired material modification to be carried out in the TA facilities. In particular, we will modify the two temperature molecular dynamics model aimed at describing the impacts of high-energy ions into solids to account for the reduced dimensionality of two-dimensional materials and correspondingly develop the dedicated software based on the LAMMPS code. Moreover, we aim to combine the two approaches and develop an algorithm/simulation setup, which would make it possible to assess the amount of damage produced in 2D materials on Si/SiO2 and other substrates. For light ions, e.g., using a He ion microscope, the majority of the damage in the 2D material results from the backscattered ions and atoms sputtered from the substrate, not by direct impacts. The process cannot be simulated by MD alone, as it becomes computationally too demanding. Besides, interatomic potentials are not always available to describe the interactions between the atoms in the target and substrate. TRIM-type software will therefore be used to assess the number of backscattered ions and sputtered atoms, and the results will be fed into MD calculations with the substrate modelled as an external repulsive potential.

Ion beam analysis techniques such as Rutherford backscattering spectrometry (RBS) and elastic recoil detection (ERD) analysis are typically used for the characterization of thin films which are homogeneously grown in two dimensions.

Thus, depth profiles are calculated from the energy spectra of detected ions assuming that no considerations on lateral fluctuation of thicknesses or 3D topology are required. However, these assumptions are not generally valid, especially when corrugated thin films, nanostructures or 3D nanomaterials in general are analysed. In order to disentangle depth profiles and 3D topologies, RBS and ERD must be measured for various angles of ion beam incidence or exit.

However, extracting 3D information from those data requires advanced simulation and fitting software, which will be developed within this tasks by IMEC and SUR. Software will be developed for the simulation of energy spectra from 3D structured materials for given detector geometries. In a second step of 3D tomography, elemental distributions will be extracted
from multiple energy spectra (D22.4).

One of the current key developments in ion beam analysis is the use of pixelated and multi detector arrangements in order to enhance detection efficiency, analysing speed, reduce radiation damage and to enhance depth resolution, sensitivity and accuracy of the elemental profiles (WP21). Examples are focal plane detector arrangements in magnetic spectrographs for high resolution RBS and ERD and pixelated particle or X-ray detectors at ion microprobes. In addition, different kinds of ion beam analysis methods, such as secondary ionization mass spectroscopy (SIMS), can be performed simultaneously or sequentially where the combination of information of the different techniques allow a substantial improvement in the interpretation of ion beam analyses. In parallel to the hardware developments it is therefore necessary to develop adequate data acquisition, filtering and analysis tools. Thus, new data structures (WP5) and software packages to perform adequate data acquisition, optimized data filtering and data analysis will be developed in this task:

• Software for routine, one or more level filtering of multi-parameter data (UBW, JYU, RBI, IMEC) e.g. filtering and analysis of the multi-dispersive data (position and energy) from a magnetic spectrometer
• Algorithms to combine and analyze the spectra from a large set of detectors (CNRS, IMEC, INFN) (D22.2)
• Software to make use of multi techniques analysis, such as IBA and SIMS, in a common data evaluation (SUR, IST, RBI) (D22.5)

Nowadays, deep learning algorithms find widespread use in various fields where modelling algorithms are unsuitable (i.e. are not efficient or even cannot solve a given complex problem). In the domain of ion beam analysis and ion beam modification of materials, we have identified two fields where these artificial intelligence approaches will particularly enhance the performance of ion beam methods.

The first field of application is to solve the inverse problem for Rutherford backscattering spectrometry (RBS) where one has to solve the ambiguity of elemental and depth information mixed together in a single energy spectrum of backscattered particles. These data may be supported by multiple technique approach where additional information is obtained from complementary analysis techniques (KUL, SUR). By training the system with the large set of already measured and well-known samples one can expect the evaluation of new measured energy spectra by the deep learning algorithms. This will allow an easier analysis of energy spectra even by users that are not as experienced in data evaluation. Using advanced learning methods and network architectures, as well as by simultaneously combining spectra from multiple detectors (WP21, T22.3), this approach will be implemented for more challenging measurements, e.g. where signals overlap, where the spectra are very noisy due to a low count rate, or where the samples exhibit extreme roughness.

Finally, we aim at implementing such networks in elastic backscattering spectrometry (EBS), where the contribution of non-elastic scattering events makes the operator-based spectrum analysis very tedious and timeconsuming (KUL, JYU, IST, SUR) (D22.6).

A second field of application is the targeted irradiation of cellular substructures. An up to now unsolved problem is the reliable recognition of substructures of a cell, e.g. cell nuclei, nucleoli, mitochondria or other organelles from phase contrast images or from staining these structures and imaging them under the epifluorescence microscope. The main challenge is to detect these substructures with high reliability (high detection success rate with low reporting of false structures). Since image quality and structure visibility vary from cell to cell and from image to image, direct modelling and analysis algorithms are not suitable. On the other end, the abundance and complexity of the data make it an ideal model application for deep learning algorithms. (UBW, SUR) (D22.7)