Three-dimensional ion micro-tomography

https://doi.org/10.1016/S0168-583X(01)00618-8Get rights and content

Abstract

The technique of ion micro-tomography (IMT) provides three-dimensional distribution information about a sample's mass density and elemental composition. The required data are obtained by doing a scanning transmission ion microscopy (STIM) tomography experiment followed by a particle-induced X-ray emission (PIXE) tomography experiment. The experiment times have been vastly reduced now that data are collected with MicroDAS, the new fast data acquisition system. Moreover, the experiment is easier to perform because sample manipulation is automated via computer control. To obtain comparable spatial resolutions between the STIM and PIXE data, the PIXE tomography experiment is performed by implementing a large solid angle between the sample and X-ray detector. To correct for the inherent three-dimensional nature of such an experimental setup, a specially developed tomographic reconstruction technique is used to combine the STIM and PIXE tomography data sets to create an accurate quantitative tomogram of the sample. The efficacy of the entire IMT process is tested with a characterised “standard” sample. The calculated data agree well with the quantitative and structural information known about the sample. To interpret the three-dimensional distribution information, a special volume rendering program is used to visualise various aspects of the tomogram. Each aspect is colour coded to facilitate the easy visualisation of multiple complex three-dimensional structures.

Introduction

The motivation for this work was to devise a technique that generates a three-dimensional distribution map of the mass density and elemental composition of micro-samples. These data can then be analysed and visualised with appropriate computer software. The mass density data are obtained with scanning transmission ion microscopy (STIM) experiments and the elemental concentration data are obtained with particle-induced X-ray emission (PIXE) experiments. However, such experiments only provide “depth-averaged” quantitative data. To resolve depth information, the technique of computer-assisted tomography (CAT) is used to extend the STIM and PIXE techniques. This involves collecting multiple two-dimensional STIM and PIXE scans (projections) at different angles around the sample.

For the IMT experiment to be practical and the reconstructed data to be accurate, two main attributes must be optimised. The first attribute is the data acquisition time, which must be minimised. This is achieved by using a newly designed data acquisition system (DAS) [1] that can accurately collect data at very large acquisition rates. The second attribute is an accurate three-dimensional representation of the sample, to be generated by using an appropriate reconstruction algorithm. This is achieved with the the discrete image space reconstruction algorithm (DISRA) [2], [3] that was developed to compensate for the many non-linear processes that occur during STIM and PIXE experiments.

Many techniques have been developed in an attempt to compensate for the non-linear processes that occur during STIM and PIXE experiments [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]. The main improvement over these other techniques is the ability to compensate for the effects of X-ray attenuation for a geometry in which the X-ray detector subtends a very large three-dimensional cone with the sample, as shown in Fig. 1. The purpose for such a geometry is to allow the acquisition rate of X-ray data to be sustained as the spatial resolution of the PIXE data is reduced to be comparable with the spatial resolution of the STIM data.

To demonstrate this, an experiment was done on a characterised “standard” sample. The X-ray detector was placed 12 mm from the sample at a scattering angle of 90°. With an area of 34.8mm2, the geometric efficiency was 0.018. At this position, the spatial resolution of the PIXE data was 0.5μm and was acquired at an average rate of 5.6k counts per second. This is compared to the spatial resolution of 0.25μm for the STIM data, which was acquired at an average rate of 9.9k counts per second.

Section snippets

Automation of experiment

There are many features required by a DAS to optimise the quality and throughput of collecting data for a tomography experiment.

The features of the STIM experiment are: data rates above 40k events per second, a beam blanking signal, data triggering for the scan dwell, a raster scan and a signal to block the ion beam when data are not being collected.

The high data rate allows for reduced experiment times by increasing the current of the STIM beam. A beam blanking signal deflects the beam off the

Reconstruction technique

The technique used to reconstruct STIM and PIXE tomography data is the DISRA [2], [3]. DISRA is our extension of the image space reconstruction algorithm (ISRA) [16]. The main feature of DISRA is that the density of inhomogeneous samples can be reconstructed from experiments in which complex physical phenomena occur when a probe interacts with the sample. This is accomplished whilst also retaining accurate density information at every voxel.

Fig. 2 shows a diagram of the steps required for the

Test using a “standard” sample

A test sample was made by selecting some powder chemicals and cutting one “ladder” from a 2000 mesh copper grid. These were embedded in Araldite. The powder chemicals are lead fluoride (PbF2), vanadium (V) oxide (V2O5), copper (II) oxide (CuO), lead (II) chloride (PbCl2) and potassium chloride (KCl). There are a few reasons why these chemicals were chosen. The X-ray peaks do not overlap so they are easy to identify and their yield can be accurately measured. Also, a few aspects of the

Conclusion

The experimental setup for IMT has been optimised to collect accurate STIM and PIXE data with large data acquisition rates. Even though these experiments still take several hours to complete, the total automation of these experiments does help alleviate the physical and mental burden of the user.

The three-dimensional distributions of the mass density and elemental concentration of micro-samples can be determined accurately. However, the inability of PIXE to detect low-Z elements has a

References (16)

  • A. Sakellariou et al.

    Nucl. Instr. and Meth. B

    (2001)
  • A. Sakellariou et al.

    Nucl. Instr. and Meth. B

    (1997)
  • A.E. Pontau et al.

    Nucl. Instr. and Meth. B

    (1989)
  • A.E. Pontau et al.

    Nucl. Instr. and Meth. B

    (1990)
  • A.E. Pontau et al.

    Nucl. Instr. and Meth. B

    (1991)
  • G. Bench et al.

    Nucl. Instr. and Meth. B

    (1991)
  • G. Bench et al.

    Nucl. Instr. and Meth. B

    (1992)
  • R.M.S. Schofield et al.

    Nucl. Instr. and Meth. B

    (1992)
There are more references available in the full text version of this article.

Cited by (26)

  • Accuracy of three-dimensional proton imaging of an inertial confinement fusion target assessed by Geant4 simulation

    2023, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
  • Application of silicon drift detector in PIXE tomography system of Fudan University

    2022, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
  • Reconstruction of hydrogen distribution in Inertial Confinement Fusion target shell by tomography with proton-proton scattering coincidence measurement

    2021, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
    Citation Excerpt :

    With tabulated stopping power information of ions in the scanned matter, the density information could be derived [5]. The second ion beam method integrated with CT is particle-induced X-ray emission (PIXE) [3,6]. PIXE-T doesn’t detect the energy loss of ions, instead, the emitted X-ray photons along the ion beam path during one scan is detected.

  • A dual-PIXE tomography setup for reconstruction of Germanium in ICF target

    2017, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
    Citation Excerpt :

    During target fabrication process, quantitative doping of Ge should be accurately controlled. PIXE has many advantages, like non-destructive analysis, high sensitivity for trace elements and high resolution of micro-scale [5–7]. It has already been proven useful to measure Ge planar distribution in ICF target in our laboratory [8].

  • Performance of a gas flow ionization detector filled with He-iso-C<inf>4</inf>H<inf>10</inf> mixtures for STIM-T

    2015, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms
View all citing articles on Scopus
View full text