- Email: [email protected]
Scanning transmission ion microscopy computed tomography (STIM-CT) for inertial confinement fusion (ICF) targets
Fusion Engineering and Design 88 (2013) 188–194
Contents lists available at SciVerse ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Scanning transmission ion microscopy computed tomography (STIM-CT) for inertial confinement fusion (ICF) targets Y.Q. Li a , C. Habchi b,c , X. Liu d , Y.Y. Liu d , Y. Zheng a , X.Y. Li a , H. Shen a,∗ a
Institute of Modern Physics, Applied Ion Beam Physics Laboratory, Fudan University, Shanghai 200433, China University of Bordeaux, CENBG, UMR 5797, F-33170 Gradignan, France CNRS, IN2P3, CENBG, UMR 5797, F-33170 Gradignan, France d Research Center of Laser Fusion, CAEP, Mianyang 621900, China b c
h i g h l i g h t s I I I I I
ICF target quality requires surface finishes on the order of submicron-scale. In STIM inner and outer wall profile can be mapped. In STIM the thickness and nonconcentricity of shell-wall in ICF targets can be measured. STIM-CT is a powerful method for obtaining three-dimensional density maps within ICF targets. STIM-CT can obtain internal structure with identifying non-uniformities in the ICF targets.
a r t i c l e
i n f o
Article history: Received 17 September 2012 Received in revised form 17 January 2013 Accepted 22 January 2013 Available online 22 March 2013 Keywords: ICF target STIM-CT Three-dimensional
a b s t r a c t ICF target quality control in the laser fusion program is vital to ensure that the energy deposition from the lasers results in uniform compression and minimization of Rayleigh–Taylor instabilities, which requires surface finishes on the order of submicron-scale. During target fabrication process the surface finish and the dimensions of the hohlraum need be well controlled. Density variations and nonspherical or nonconcentric shells might be produced. Scanning transmission ion microscopy computed tomography (STIM-CT) is able to reconstruct the three-dimensional quantitative structure of ICF targets a few tens of micrometers in size. Compared to other types of probe techniques, the main advantage of STIMCT is that quantitative information about mass density and sphericity can be obtained directly and non-destructively, utilizing specific reconstruction codes. We present a case of ICF target (composed of polyvinyl alcohol) characterization by STIM-CT in order to demonstrate the STIM-CT potential impact in assessing target fabrication processes. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In inertial confinement fusion (ICF), targets for direct-drive experiments consist of a spherical capsule made of glass or plastic that contains the D or DT gas used in these experiments [1]. Target structure as well as the intensity and shape of the driving pulse is vital to ensure that the bombarding laser energy yields high compression of the fuel and ignition [2]. To minimize Rayleigh–Taylor instabilities, current direct drive inertial confinement fusion (ICF) targets require surface finishes. Density variations and nonspherical or nonconcentric shells produced during the fabrication process can cause hydrodynamic instabilities. Non-uniformities of targets are the main source of perturbations which cause a departure from
∗ Corresponding author. Tel.: +86 02155664131. E-mail address: [email protected] (H. Shen). 0920-3796/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fusengdes.2013.01.080
one-dimensional performance [3]. Polyvinyl alcohol (PVA) is one of the materials used in ICF capsules and has higher tensile strength and is less permeable to hydrogen isotopes than polystyrene (PS). However, PVA is easily degraded by beta radiation, which leads to a reduction in nominal retention capability [4]. The surface finish should be appropriate for the current ICF experiments. Using an interfacial polycondensation reaction (IPCR) to produce the membrane leads to an improved surface finish [5,6]. To diagnose a single ICF target capsule, the ideal analysis method would be non-destructive, accurate, sensitive and have the ability to measure the areal density, total concentration and homogeneity. There are a number of methods for diagnosing target capsule. X-ray fluorescence (XRF) can measure elements with Z ≥ 11 with elaborate procedures to allow precision quantification of dopants in shells; Scanning transmission electron microscopy/energy dispersive X-ray diffraction (STEM/EDXRD) can be used to analysis structure and composition of solid specimen with 5–500 nm thick;
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
auger electron spectroscopy (AES) with ion milling is able to characterize the composition of all elemental except H and He. It can similarly characterize the sample in depth. The surface of the shell and the cross section of a fractured wall were observed using a scanning electron microscope (SEM) or an atomic force microscope (AFM) using a knife to broke the shell in order to produce a cross section of the wall [7]. Other techniques, such as mass spectroscopy (MS), thermo gravimetric analysis (TGA), and atomic absorption (AA), are discussed in Ref. [8] as well as in numerous other publications. Furthermore, there are several conventional methods to describe the density or thickness of materials. The method of X-ray phase contrast imaging or X-ray phase-contrast computed tomography (CT) can provide substantially increased contrast over conventional absorption-based imaging [9] and is suitable for weak-absorption materials and to diminish the total absorption dose [10]. The positron emission tomography (PET) or positron emission tomography computed tomography (PET-CT) is suitable for mapping samples with millimeter-level spatial resolution. By comparison, scanning transmission ion microscopy (STIM) has the potential for producing structural images of sample. STIM relies on measuring the energy loss of a beam of highly focused MeV ions as it passes through a sample. Because the transmitted protons in general maintain a straight path as they pass through sample, then a high quality structural image of a relatively thick specimen can be formed [11]. Combining the modern computed tomography (CT), STIM allows for nondestructive quantitative characterization of targets without slicing sample itself [12–14]. The technique of CT is used to determine the 3D distribution of a physical property in a specimen from a set of projections taken at different orientations [15]. STIM-CT is a powerful non-destructive tool for obtaining three-dimensional density maps and identifying non-uniformities within the targets [16] with a high probing efficiency (nearly 100%) and low ion beam current (∼fA). Features such as thickness variations or shell delamination in ICF targets are distinguished with micron and submicron-scale spatial resolution due to the small beam size [17]. The depth resolution is several tenths nanometers influenced by the stopping power of incident ion, the energy resolution of the STIM detector and the energy straggling. STIMCT could be explored how accurately we can resolve geometric defects. It also measured a target with density variations caused by latex sphere clusters similar in density and composition to the hydrocarbon matrix. The physical quantity measured in scanning transmission ion micro-tomography (STIM) is the energy loss of the ions. This primary data can then be fed into a STIM-CT 3D reconstruction process [18], assuming that sample is homogeneous in composition. Because of this energy-loss mechanism, we can obtain quantitative total density measurements with the advantage that low atomic number (Z) constituents are not masked by their high Z counterparts. In STIM, a single projection, which can be used to identify the shape or sphericity of the ICF target, does not allow visualisation of the internal structure of the sample. The internal structure becomes accessible using the tomography technique. The intention of this paper is to describe the system and present the examples of STIMCT characterization of ICF targets for three-dimensional density, homogeneity, and concentration in order to demonstrate the capabilities of the technique as a means of assessing future ICF target production methods for the first time.
2. Experimental setup and conditions The experimental setup has been designed especially for 3Dtomography analysis of ICF targets. A schematic diagram of a STIM-CT experiment is shown in Fig. 1. A windowless Si-PIN diode
189
Fig. 1. Schematic diagram of a STIM-CT experiment. For the experiment described in this paper, the spot size of the incident beam at the specimen surface was adjusted to 2 m. Energies of protons are measured with a Si-PIN dector.
(Hamamatsu 1223-01) with a resolution of 25 keV at full width at half-maximum (FWHM) is placed behind the sample holder (about 6 mm away from the sample) and measures the residual energy of the protons after traversing the sample. The detector suffers from damage during the experiment since the protons are implanted into the crystal. Thus, as soon as the deviation of the measured energy for protons in vacuum decreases, the detector is whirled to expose a fresh region [18]. A sample manipulator with a steel needle is utilized as rotation axis. The manipulator is attached to a PC magnetic rotary drive unit, which is based on an x–y–z target manipulator (x, y: ±12.5 mm, z: 0–50 mm, step: 5 m). The manipulator can be used to correct the specimen position in case it moves out of the field of view during specimen rotation at the beginning data acquisition [19]. During the experiments the projection angle of the tomography axis is performed by a computer controlled precision step motor. The step motor is capable of rotation in minimum step of 0.05◦ . The control and data acquisition system of the microprobe facility is running on a PC under MS Windows 2000. The detectors of STIM is connected to OM1000, which convert the analog signals to digital signals. The computer controls the scanning size (range 2 mm × 2 mm) with the Oxford Microbeams Ltd. Data acquisition (OMDAQ) system. The data acquisition and the 3D-scan analysis are fully automated. A single shell target composed of polyvinyl alcohol (PVA) was prepared for the first experiment. These microspheres were chosen as a model for STIM-CT study due to their regular spherical shape, certified relatively good sphericity, concentricity characteristics and low mass density of 1.19–1.31 g/cm3 . The composition, at least for the main chemical elements, is assumed uniform within all the volume. Although the size of these microspheres does not reach sub-micron level, they are suitable for a first test in STIM-CT. The sample was mounted on top of steel needle using cyanoacrylate adhesive or super glue and then was simply dried in air about 12 h. The sample was aligned along the vertical axis in air and in vacuum. A 3D-STIM-tomography experiment consists of recording a number of 2D-STIM images of the sample, called projections, under different incident angles from 0–180◦ (the third dimension) [20]. From these projections the 3D density distribution can be reconstructed. In our experiment, a 3 MeV H+ beam was used as a probe. The current rate was adjusted to 1000–3000 Hz to avoid any damage of the sample during the analysis. In these conditions, STIM-CT is considered as a non-destructive technique. The transmitted ions are collected in detector, mechanically whirled at 0◦ on the incoming beam axis during the acquisition [21]. For the analysis the sample was scanned over an area of 750 m × 750 m with a total of 128 horizontal slices, obtained
190
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
from a 128 × 128 pixels. The projections were recorded by horizontally scanning the beam over the sample from top to bottom [20]. With the purpose of avoiding sample damage, several rapid scans were performed for every projection. The scan speed was here selected to be 10 s/pixel. The rotation was performed by a computer controlled precision step motor. A total of 100 projections were recorded over 180◦ angular range with steps of 1.8◦ , with using 50 scans/projection. The total duration of the analysis was typically about 5 h.
3. Three dimensional (3D) data reconstruction and image display software For data treatment the TomoRebuild 2 data reduction program was used. The program, developed at CENBG, is a code used to process STIM-CT data before reconstruction of 2D or 3D tomography images. It builds the projections, corrects the sinograms, calculates the rotation centers and calibrates the data in the Data Treatment step. In the TomoRebuild step, it calculates the stopping power for each pixel of the original file, calculates each slice of the resulting sinogram with a reconstruction algorithm. Intermediate results can be easily accessed to be checked and/or modified at different steps of the reconstruction. Automatic correction procedures can be implemented to improve noisy data. Three reconstruction techniques are available in TomoRebuild (filtered backproection, MLEM and OSEM). For this experiment, we used the filtered backprojection algorithm, which is the easiest and also the least time-consuming reconstruction technique for transmission tomography data [22]. The reconstruction of ideal data is straightforward. However, with experimental data, the reconstructed image may be drastically deteriorated due to noise and misalignment problems in the raw data [19]. A detailed description of how the TomoRebuild software operates can be found in Ref. [17] for the previous version. A new version is now available and a description can be found in Ref. [23]. The graphic software AMIRA® [24] was used to display the image of 3D structure of the sample by associating a colour code to density values. It allows you to visualize scientific data sets from various application areas, e.g. medicine, biology, chemistry, physics, or engineering. 3D objects can be represented as grids suitable for numerical simulations, notably as triangular surface and
Fig. 2. STIM spectra of target, fixed with general glue on the top of the glass capillary. Peak marked with outer was the incident energy. The lower energy (marked with inner) was caused by energy losses of the protons that pass through target.
volumetric tetrahedral grids. Amira provides methods to generate such grids from voxel data representing an image volume, and it includes a general purpose interactive 3D viewer. Meanwhile several cutting planes can be selected to vision internal structure at any orientation. Automatic calculations can be carried out on selected materials to determine thickness, volume and average density of sample. Using the method quantitative information about the target is easily approachable (Fig. 2). 4. Results and discussion In STIM, inner and outer wall profile (Fig. 3) can be mapped according to sorting energy area on the STIM spectra (Fig. 2). From the rest, the outer wall center is as same as the inner wall center. Then nonconcentricity between the inner and outer wall is zero. Meanwhile, two-dimensional projection (Fig. 4) of the microcomposite sample shows the median energy loss of the proton beam measured at every position of the 750 m × 750 m scan. At this resolution, two regions, with rather uniform density, could be observed: the shell-wall (around 25 m thick) and the hollow interior. The shape or sphericity of the ICF target is identified from the only projection. Whereas a single projection does not allow
Fig. 3. STIM images of target. (a) Inner wall profile and (b) outer wall profile.
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
191
Fig. 4. Two-dimensional projection of the target shows the median energy loss of the proton beam measured at every position of an area of 750 m × 750 m with a total of 128 horizontal slices, obtained from a 128 × 128 pixels. The linear gray scale ranges from zero (red) to the highest value (black). The thickness variations are visible.
visualisation of the internal structure of the sample. The internal structure becomes accessible using the tomography technique. Three-dimensional structure of the target was reconstructed (Fig. 5). Two isosurfaces are displayed. In Fig. 6, the density closed to zero is characteristic of the limit between the sample and surrounding vacuum. The highest densities show the wall of target (white). Their spherical shape is visible in the slices, selected along the target as indicated in Fig. 5(b). The mass density ranged from zero (black) to the highest areal density (bright white) was used to represent the data. The reconstruction was performed assuming a uniform theoretical composition of polyvinyl alcohol for all the target. Because of sample composition, the reconstructed values of the mass density given hereafter are only indicative, and should not be considered as a precise determination [21]. The local concentrations lead to a better calculation of the stopping power of the sample. In return, the knowledge of composition and mass density leads to a more precise calculation of proton energy loss [25]. However many deviations from the precise determination occur. The experimental conditions affect getting an approximation closer to the ideal projection process. One point to be considered is the energy dependence of the stopping power [18]. The energy loss of an ion passing through a specimen can be described by the linear stopping power, which relates to the incident beam energy. The energy loss is converted to the projected areal mass density along this trajectory prior to the reconstruction of the stopping power [26]. Another point to be considered is the beam broadening within the specimen (lateral straggling) [27]. It is important that the interaction volumes for each ion trajectory do not overlap otherwise the spatial resolution of the reconstructed specimen is degraded. For a given specimen, the incident energy and the specimen thickness control the broadening of the proton beam [18]. A further effect caused by the statistical nature of the multiple interactions between ions and the electrons of the material is energy straggling. A beam of defined energy will have a Gaussian energy distribution after interacting with matter [28,29]. From the experimental conditions, the error was about 10%.
Fig. 5. The direction of X-axis was the beam direction. (a) 3D-reconstruction of the reference sample. (b) Position of the three selected slices at three orientations within the reconstructed volume.
In Fig. 6(a), the non circular shape observed is surrounded by streaks which could lead to shape artefacts in the reconstructed circle. Such artefacts would be inherent to the filtered backprojection process. Also it could come from a slightly wrong determination of the position of the rotation axis. We have tried to reconstruct the slice with a slightly different position of rotation center and filtered backprojection. However the shapes were same as Fig. 6(a). From reconstructed ion microtomography slice of target, it can be inferred that a processing defect caused the inner and outside surface to become nonconcentric. In other words, this target was used to help identify the key processing steps during fabrication which led to the formation of such flaws. For illustrative purposes, we have imposed a density threshold over the full reconstruction volume in order to only highlight the distribution of density and target slice [3]. Fig. 7 shows the distribution of different density. It can be directly known that the precise density at every position. Fig. 8 presents the possibility to stack the reconstructed slices and perform three-dimensional tomography. An artificial cut has been made along the symmetry axis to
192
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
Fig. 6. Reconstructed ion microtomography slice of target from three orientations. (a) Axial orientation, the round of the out surface was irregular having a nonconcentric outer surface. (b) Coronal; the shell of the coronal orientation is symmetry. (c) Sagittal orientation having a nonconcentric inner surface.
Fig. 7. The distribution of different density from (a) to (c) the density threshold is growing.
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
193
induced X-ray emission (PIXE) tomography will be combined for elemental analysis of micrometer-sized samples. Acknowledgement The authors wish to thank accelerator group for good beam condition during experiments. References
Fig. 8. Three-dimensional image of target. An artificial cut has been made along the symmetry axis for illustration of the internal structure as well as the target wall surface. (a) front of artificial cut and (b) side of artificial cut.
illustrate the shape of the object as well as its internal structure. The inside surface was finish and the dimensions of the hohlraum are of symmetry. Therefore it is illustrated that the surface finish and the dimensions of the mandrel be well controlled in the target fabrication [1]. However the outside surface was coarse. 5. Conclusions STIM-CT on the beam line at the FUDAN nuclear microprobe laboratory can be applied to analyse the 3D structure of inertial confinement fusion targets, providing submicron-scale characterization. The method can apply not only to PVA material for ICF capsules, but also to other materials being used for capsules such as beryllium or high-density carbon (diamond). The projection data reveal the need for careful corrections of misalignment of the rotation centers obtained for every slice. For more accurate quantitative STIM tomography in further experiments, the rotation axis alignment will be performed with higher precision. Also a more accurate quantitative mass density will be determined. Particle
[1] K.R. Schultz, J.L. Kaae, W.J. Miller, D.A. Steinman, R.B. Stephens, Status of inertial fusion target fabrication in the USA, Fusion Engineering and Design 44 (1999) 441–448. [2] U. Kubo, H. Tsubakihara, Development of a coating technique for inertial confinement fusion plastic targets, Journal of Vacuum Science and Technology A 4 (1986) 1134–1137. [3] A.J. Antolak, A.E. Pontau, D.H. Morse, Ion microtomography and particleinduced X-ray emission analysis of direct drive inertial confinement fusion targets, Journal of Vacuum Science and Technology A 10 (1992) 1164–1169. [4] L.A. Scott, R.G. Schneggenburger, P.R. Anderson, Deuterium permeation properties of beta-irradiated and unirradiated poly(vinyl alcohol) and polystyrene shells, Journal of Vacuum Science and Technology A 4 (1986) 1155. [5] U. Kubo, H. Tsubakihara, Development of a coating technique for inertial confinement fusion plastic targets, Journal of Vacuum Science and Technology A 4 (1986) 1134. [6] M. Takagi, M. Ishihara, T. Norimatsu, T. Yamanaka, Y. Izawa, S. Nakai, Development of foam shell with plastic ablator for cryogenic laser fusion target, Journal of Vacuum Science and Technology A 11 (1993) 2837. [7] U. Kubo, H. Nakano, Hyo-gun Kim, Fabrication of cross-linked polymer shells for inertial confinement fusion experiments, Journal of Vacuum Science and Technology A 15 (3) (1997) 683–685. [8] C. Brundle, C. Evans, S. Wilson, Encyclopedia of Materials Characterization, Butterworth-Heinemann, Manning, 1992. [9] P.P. Zhu, K. Zhang, Z.L. Wang, Y.J. Liu, X.S. Liu, Z.Y. Wu, S.A. McDonald, F. Marone, M. Stampanoni, Low-dose, simple, and fast grating-based X-ray phase-contrast imaging, Proceedings of the National Academy of Sciences of the United States of America 107 (2010) 13576–13581. [10] J.Y. Zhang, J. Gao, F. Jiang, J. Chen, B. Chen, H. Ming, X-ray phase-contrast computed tomography, Chinese Journal of Quantum Electronics 22 (2005) 579–584. [11] Ren Minqin, J.A. van Kan, A.A. Bettiol, Lim Daina, Chan Yee Gek, Bay Boon Huat, H.J. Whitlow, T. Osipowicz, F. Watt, Nano-imaging of single cells using STIM, Nuclear Instruments and Methods in Physics Research Section B 260 (2007) 124–129. [12] C. Michelet, Ph. Moretto, 3D mapping of individual cells using a proton microbeam, Nuclear Instruments and Methods in Physics Research Section B 150 (1999) 173. [13] Eric Stindel, K. Jayaram, Udupa, E. Bruce, D. Hirsch, Odhner, Christine Couture, Computerized Medical Imaging and Graphics 23 (1999) 75. [14] R. Wu, H. Shen, Y. Mi, M.D. Sun, M.J. Yang, A computer program for scanning transmission ion microscopy simulation, Nuclear Instruments and Methods in Physics Research Section B 231 (2005) 149–155. [15] A.C. Kak, Principles of Computerized Tomographic Imaging, IEEE Press, New York, 1987. [16] T. Andrea, M. Rothermel, T. Butz, T. Reinert, The improved STIM tomography set-up at LIPSION: three-dimensional reconstruction of biological samples, Nuclear Instruments and Methods in Physics Research Section B 267 (2009) 2098–2102. [17] C. Michelet-Habchi, S. Incerti, P. Aguer, Ph. Barberet, E. Gontier, K. Grente, Ph. Moretto, D.T. Nguyen, A. Pouthier, T. Pouthier, F. Rebillat, R.W. Smith, TomoRebuild: a new data reduction software package for scanning transmission ion microscopy tomography, Nuclear Instruments and Methods in Physics Research Section B 231 (2005) 142–148. [18] M. Schwertner, A. Sakellariou, T. Reinert, T. Butz, Scanning transmission ion micro-tomography (STIM-T) of biological specimens, Ultramicroscopy 106 (2006) 574–581. [19] M. Wegdén, M. Elfman, P. Kristiansson, N. Arteaga Marrero, V. Auzelyte, K.G. Malmqvist, C. Nilsson, J. Pallon, Development of a microtomography system at the Lund sub-micron beamline, Nuclear Instruments and Methods in Physics Research Section B 249 (2006) 756–759. [20] T. Reinert, A. Sakellariou, M. Schwertner, J. Vogt, T. Butz, Scanning transmission ion microscopy tomography at the Leipzig nanoprobe LIPSION, Nuclear Instruments and Methods in Physics Research Section B 190 (2002) 266–270. [21] N. Gordillo, C. Habchi, L. Daudin, A. Sakellariou, F. Delalée, Ph. Barberet, S. Incerti, H. Seznec, Ph. Moretto, Technical developments for computed tomography on the CENBG nanobeam line, Nuclear Instruments and Methods in Physics Research Section B 269 (2011) 2206–2209. [22] G. Bench, K.A. Nugent, M. Cholewa, A. Saint, G.J.F. Legge, Submicron STIM tomography reconstruction techniques, Nuclear Instruments and Methods in Physics Research Section B 54 (1991) 390. [23] C. Habchi, N. Gordillo, S. Bourret, Ph. Barberet, C. Jovet, Ph. Moretto, H. Seznec, Beyond filtered backprojection: a reconstruction software package for ion beam microtomography data, Nuclear Instruments and Methods in Physics Research Section B 295 (2013) 42–49.
194
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
[24] AMIRA web site. Available from: http://www.tgs.com [25] C. Habchi, D.T. Nguyen, Ph. Barberet, S. Incerti, Ph. Moretto, A. Sakellariou, H. Seznec, Reconstruction of 3D ion beam micro-tomography data for applications in cell biology, Nuclear Instruments and Methods in Physics Research Section B 267 (2009) 2107–2112. [26] A. Pontau, A.J. Antolak, D.H. Morse, A.A. Ver Berkmoes, J.M. Brase, D.W. Heikkinen, H.E. Martz, I.D. Procter, Ion Microbeam tomography, Nuclear Instruments and Methods in Physics Research Section B 40/41 (1989) 646–650.
[27] A.J. Antolak, An empirical treatment of nonrelativistic proton beam spreading, Nuclear Instruments and Methods in Physics Research Section B 30 (1988) 182–184. [28] C. Michelet, Ph. Moretto, G. Laurent, W.J. Przybylowicz, V.M. Prozesky, C.A. Pineda, Ph. Barbaret, F. Lohste, J. Kennedy, Measurement of lateral straggling using a microbeam, Nuclear Instruments and Methods in Physics Research Section B 181 (2001) 157–163. [29] M. Breese, D.N. Jamieson, P. King, Materials Analysis using Nuclear Microprobes, Wiley Press, New York (USA), 1996, 102 pp.
Contents lists available at SciVerse ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Scanning transmission ion microscopy computed tomography (STIM-CT) for inertial confinement fusion (ICF) targets Y.Q. Li a , C. Habchi b,c , X. Liu d , Y.Y. Liu d , Y. Zheng a , X.Y. Li a , H. Shen a,∗ a
Institute of Modern Physics, Applied Ion Beam Physics Laboratory, Fudan University, Shanghai 200433, China University of Bordeaux, CENBG, UMR 5797, F-33170 Gradignan, France CNRS, IN2P3, CENBG, UMR 5797, F-33170 Gradignan, France d Research Center of Laser Fusion, CAEP, Mianyang 621900, China b c
h i g h l i g h t s I I I I I
ICF target quality requires surface finishes on the order of submicron-scale. In STIM inner and outer wall profile can be mapped. In STIM the thickness and nonconcentricity of shell-wall in ICF targets can be measured. STIM-CT is a powerful method for obtaining three-dimensional density maps within ICF targets. STIM-CT can obtain internal structure with identifying non-uniformities in the ICF targets.
a r t i c l e
i n f o
Article history: Received 17 September 2012 Received in revised form 17 January 2013 Accepted 22 January 2013 Available online 22 March 2013 Keywords: ICF target STIM-CT Three-dimensional
a b s t r a c t ICF target quality control in the laser fusion program is vital to ensure that the energy deposition from the lasers results in uniform compression and minimization of Rayleigh–Taylor instabilities, which requires surface finishes on the order of submicron-scale. During target fabrication process the surface finish and the dimensions of the hohlraum need be well controlled. Density variations and nonspherical or nonconcentric shells might be produced. Scanning transmission ion microscopy computed tomography (STIM-CT) is able to reconstruct the three-dimensional quantitative structure of ICF targets a few tens of micrometers in size. Compared to other types of probe techniques, the main advantage of STIMCT is that quantitative information about mass density and sphericity can be obtained directly and non-destructively, utilizing specific reconstruction codes. We present a case of ICF target (composed of polyvinyl alcohol) characterization by STIM-CT in order to demonstrate the STIM-CT potential impact in assessing target fabrication processes. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In inertial confinement fusion (ICF), targets for direct-drive experiments consist of a spherical capsule made of glass or plastic that contains the D or DT gas used in these experiments [1]. Target structure as well as the intensity and shape of the driving pulse is vital to ensure that the bombarding laser energy yields high compression of the fuel and ignition [2]. To minimize Rayleigh–Taylor instabilities, current direct drive inertial confinement fusion (ICF) targets require surface finishes. Density variations and nonspherical or nonconcentric shells produced during the fabrication process can cause hydrodynamic instabilities. Non-uniformities of targets are the main source of perturbations which cause a departure from
∗ Corresponding author. Tel.: +86 02155664131. E-mail address: [email protected] (H. Shen). 0920-3796/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fusengdes.2013.01.080
one-dimensional performance [3]. Polyvinyl alcohol (PVA) is one of the materials used in ICF capsules and has higher tensile strength and is less permeable to hydrogen isotopes than polystyrene (PS). However, PVA is easily degraded by beta radiation, which leads to a reduction in nominal retention capability [4]. The surface finish should be appropriate for the current ICF experiments. Using an interfacial polycondensation reaction (IPCR) to produce the membrane leads to an improved surface finish [5,6]. To diagnose a single ICF target capsule, the ideal analysis method would be non-destructive, accurate, sensitive and have the ability to measure the areal density, total concentration and homogeneity. There are a number of methods for diagnosing target capsule. X-ray fluorescence (XRF) can measure elements with Z ≥ 11 with elaborate procedures to allow precision quantification of dopants in shells; Scanning transmission electron microscopy/energy dispersive X-ray diffraction (STEM/EDXRD) can be used to analysis structure and composition of solid specimen with 5–500 nm thick;
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
auger electron spectroscopy (AES) with ion milling is able to characterize the composition of all elemental except H and He. It can similarly characterize the sample in depth. The surface of the shell and the cross section of a fractured wall were observed using a scanning electron microscope (SEM) or an atomic force microscope (AFM) using a knife to broke the shell in order to produce a cross section of the wall [7]. Other techniques, such as mass spectroscopy (MS), thermo gravimetric analysis (TGA), and atomic absorption (AA), are discussed in Ref. [8] as well as in numerous other publications. Furthermore, there are several conventional methods to describe the density or thickness of materials. The method of X-ray phase contrast imaging or X-ray phase-contrast computed tomography (CT) can provide substantially increased contrast over conventional absorption-based imaging [9] and is suitable for weak-absorption materials and to diminish the total absorption dose [10]. The positron emission tomography (PET) or positron emission tomography computed tomography (PET-CT) is suitable for mapping samples with millimeter-level spatial resolution. By comparison, scanning transmission ion microscopy (STIM) has the potential for producing structural images of sample. STIM relies on measuring the energy loss of a beam of highly focused MeV ions as it passes through a sample. Because the transmitted protons in general maintain a straight path as they pass through sample, then a high quality structural image of a relatively thick specimen can be formed [11]. Combining the modern computed tomography (CT), STIM allows for nondestructive quantitative characterization of targets without slicing sample itself [12–14]. The technique of CT is used to determine the 3D distribution of a physical property in a specimen from a set of projections taken at different orientations [15]. STIM-CT is a powerful non-destructive tool for obtaining three-dimensional density maps and identifying non-uniformities within the targets [16] with a high probing efficiency (nearly 100%) and low ion beam current (∼fA). Features such as thickness variations or shell delamination in ICF targets are distinguished with micron and submicron-scale spatial resolution due to the small beam size [17]. The depth resolution is several tenths nanometers influenced by the stopping power of incident ion, the energy resolution of the STIM detector and the energy straggling. STIMCT could be explored how accurately we can resolve geometric defects. It also measured a target with density variations caused by latex sphere clusters similar in density and composition to the hydrocarbon matrix. The physical quantity measured in scanning transmission ion micro-tomography (STIM) is the energy loss of the ions. This primary data can then be fed into a STIM-CT 3D reconstruction process [18], assuming that sample is homogeneous in composition. Because of this energy-loss mechanism, we can obtain quantitative total density measurements with the advantage that low atomic number (Z) constituents are not masked by their high Z counterparts. In STIM, a single projection, which can be used to identify the shape or sphericity of the ICF target, does not allow visualisation of the internal structure of the sample. The internal structure becomes accessible using the tomography technique. The intention of this paper is to describe the system and present the examples of STIMCT characterization of ICF targets for three-dimensional density, homogeneity, and concentration in order to demonstrate the capabilities of the technique as a means of assessing future ICF target production methods for the first time.
2. Experimental setup and conditions The experimental setup has been designed especially for 3Dtomography analysis of ICF targets. A schematic diagram of a STIM-CT experiment is shown in Fig. 1. A windowless Si-PIN diode
189
Fig. 1. Schematic diagram of a STIM-CT experiment. For the experiment described in this paper, the spot size of the incident beam at the specimen surface was adjusted to 2 m. Energies of protons are measured with a Si-PIN dector.
(Hamamatsu 1223-01) with a resolution of 25 keV at full width at half-maximum (FWHM) is placed behind the sample holder (about 6 mm away from the sample) and measures the residual energy of the protons after traversing the sample. The detector suffers from damage during the experiment since the protons are implanted into the crystal. Thus, as soon as the deviation of the measured energy for protons in vacuum decreases, the detector is whirled to expose a fresh region [18]. A sample manipulator with a steel needle is utilized as rotation axis. The manipulator is attached to a PC magnetic rotary drive unit, which is based on an x–y–z target manipulator (x, y: ±12.5 mm, z: 0–50 mm, step: 5 m). The manipulator can be used to correct the specimen position in case it moves out of the field of view during specimen rotation at the beginning data acquisition [19]. During the experiments the projection angle of the tomography axis is performed by a computer controlled precision step motor. The step motor is capable of rotation in minimum step of 0.05◦ . The control and data acquisition system of the microprobe facility is running on a PC under MS Windows 2000. The detectors of STIM is connected to OM1000, which convert the analog signals to digital signals. The computer controls the scanning size (range 2 mm × 2 mm) with the Oxford Microbeams Ltd. Data acquisition (OMDAQ) system. The data acquisition and the 3D-scan analysis are fully automated. A single shell target composed of polyvinyl alcohol (PVA) was prepared for the first experiment. These microspheres were chosen as a model for STIM-CT study due to their regular spherical shape, certified relatively good sphericity, concentricity characteristics and low mass density of 1.19–1.31 g/cm3 . The composition, at least for the main chemical elements, is assumed uniform within all the volume. Although the size of these microspheres does not reach sub-micron level, they are suitable for a first test in STIM-CT. The sample was mounted on top of steel needle using cyanoacrylate adhesive or super glue and then was simply dried in air about 12 h. The sample was aligned along the vertical axis in air and in vacuum. A 3D-STIM-tomography experiment consists of recording a number of 2D-STIM images of the sample, called projections, under different incident angles from 0–180◦ (the third dimension) [20]. From these projections the 3D density distribution can be reconstructed. In our experiment, a 3 MeV H+ beam was used as a probe. The current rate was adjusted to 1000–3000 Hz to avoid any damage of the sample during the analysis. In these conditions, STIM-CT is considered as a non-destructive technique. The transmitted ions are collected in detector, mechanically whirled at 0◦ on the incoming beam axis during the acquisition [21]. For the analysis the sample was scanned over an area of 750 m × 750 m with a total of 128 horizontal slices, obtained
190
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
from a 128 × 128 pixels. The projections were recorded by horizontally scanning the beam over the sample from top to bottom [20]. With the purpose of avoiding sample damage, several rapid scans were performed for every projection. The scan speed was here selected to be 10 s/pixel. The rotation was performed by a computer controlled precision step motor. A total of 100 projections were recorded over 180◦ angular range with steps of 1.8◦ , with using 50 scans/projection. The total duration of the analysis was typically about 5 h.
3. Three dimensional (3D) data reconstruction and image display software For data treatment the TomoRebuild 2 data reduction program was used. The program, developed at CENBG, is a code used to process STIM-CT data before reconstruction of 2D or 3D tomography images. It builds the projections, corrects the sinograms, calculates the rotation centers and calibrates the data in the Data Treatment step. In the TomoRebuild step, it calculates the stopping power for each pixel of the original file, calculates each slice of the resulting sinogram with a reconstruction algorithm. Intermediate results can be easily accessed to be checked and/or modified at different steps of the reconstruction. Automatic correction procedures can be implemented to improve noisy data. Three reconstruction techniques are available in TomoRebuild (filtered backproection, MLEM and OSEM). For this experiment, we used the filtered backprojection algorithm, which is the easiest and also the least time-consuming reconstruction technique for transmission tomography data [22]. The reconstruction of ideal data is straightforward. However, with experimental data, the reconstructed image may be drastically deteriorated due to noise and misalignment problems in the raw data [19]. A detailed description of how the TomoRebuild software operates can be found in Ref. [17] for the previous version. A new version is now available and a description can be found in Ref. [23]. The graphic software AMIRA® [24] was used to display the image of 3D structure of the sample by associating a colour code to density values. It allows you to visualize scientific data sets from various application areas, e.g. medicine, biology, chemistry, physics, or engineering. 3D objects can be represented as grids suitable for numerical simulations, notably as triangular surface and
Fig. 2. STIM spectra of target, fixed with general glue on the top of the glass capillary. Peak marked with outer was the incident energy. The lower energy (marked with inner) was caused by energy losses of the protons that pass through target.
volumetric tetrahedral grids. Amira provides methods to generate such grids from voxel data representing an image volume, and it includes a general purpose interactive 3D viewer. Meanwhile several cutting planes can be selected to vision internal structure at any orientation. Automatic calculations can be carried out on selected materials to determine thickness, volume and average density of sample. Using the method quantitative information about the target is easily approachable (Fig. 2). 4. Results and discussion In STIM, inner and outer wall profile (Fig. 3) can be mapped according to sorting energy area on the STIM spectra (Fig. 2). From the rest, the outer wall center is as same as the inner wall center. Then nonconcentricity between the inner and outer wall is zero. Meanwhile, two-dimensional projection (Fig. 4) of the microcomposite sample shows the median energy loss of the proton beam measured at every position of the 750 m × 750 m scan. At this resolution, two regions, with rather uniform density, could be observed: the shell-wall (around 25 m thick) and the hollow interior. The shape or sphericity of the ICF target is identified from the only projection. Whereas a single projection does not allow
Fig. 3. STIM images of target. (a) Inner wall profile and (b) outer wall profile.
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
191
Fig. 4. Two-dimensional projection of the target shows the median energy loss of the proton beam measured at every position of an area of 750 m × 750 m with a total of 128 horizontal slices, obtained from a 128 × 128 pixels. The linear gray scale ranges from zero (red) to the highest value (black). The thickness variations are visible.
visualisation of the internal structure of the sample. The internal structure becomes accessible using the tomography technique. Three-dimensional structure of the target was reconstructed (Fig. 5). Two isosurfaces are displayed. In Fig. 6, the density closed to zero is characteristic of the limit between the sample and surrounding vacuum. The highest densities show the wall of target (white). Their spherical shape is visible in the slices, selected along the target as indicated in Fig. 5(b). The mass density ranged from zero (black) to the highest areal density (bright white) was used to represent the data. The reconstruction was performed assuming a uniform theoretical composition of polyvinyl alcohol for all the target. Because of sample composition, the reconstructed values of the mass density given hereafter are only indicative, and should not be considered as a precise determination [21]. The local concentrations lead to a better calculation of the stopping power of the sample. In return, the knowledge of composition and mass density leads to a more precise calculation of proton energy loss [25]. However many deviations from the precise determination occur. The experimental conditions affect getting an approximation closer to the ideal projection process. One point to be considered is the energy dependence of the stopping power [18]. The energy loss of an ion passing through a specimen can be described by the linear stopping power, which relates to the incident beam energy. The energy loss is converted to the projected areal mass density along this trajectory prior to the reconstruction of the stopping power [26]. Another point to be considered is the beam broadening within the specimen (lateral straggling) [27]. It is important that the interaction volumes for each ion trajectory do not overlap otherwise the spatial resolution of the reconstructed specimen is degraded. For a given specimen, the incident energy and the specimen thickness control the broadening of the proton beam [18]. A further effect caused by the statistical nature of the multiple interactions between ions and the electrons of the material is energy straggling. A beam of defined energy will have a Gaussian energy distribution after interacting with matter [28,29]. From the experimental conditions, the error was about 10%.
Fig. 5. The direction of X-axis was the beam direction. (a) 3D-reconstruction of the reference sample. (b) Position of the three selected slices at three orientations within the reconstructed volume.
In Fig. 6(a), the non circular shape observed is surrounded by streaks which could lead to shape artefacts in the reconstructed circle. Such artefacts would be inherent to the filtered backprojection process. Also it could come from a slightly wrong determination of the position of the rotation axis. We have tried to reconstruct the slice with a slightly different position of rotation center and filtered backprojection. However the shapes were same as Fig. 6(a). From reconstructed ion microtomography slice of target, it can be inferred that a processing defect caused the inner and outside surface to become nonconcentric. In other words, this target was used to help identify the key processing steps during fabrication which led to the formation of such flaws. For illustrative purposes, we have imposed a density threshold over the full reconstruction volume in order to only highlight the distribution of density and target slice [3]. Fig. 7 shows the distribution of different density. It can be directly known that the precise density at every position. Fig. 8 presents the possibility to stack the reconstructed slices and perform three-dimensional tomography. An artificial cut has been made along the symmetry axis to
192
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
Fig. 6. Reconstructed ion microtomography slice of target from three orientations. (a) Axial orientation, the round of the out surface was irregular having a nonconcentric outer surface. (b) Coronal; the shell of the coronal orientation is symmetry. (c) Sagittal orientation having a nonconcentric inner surface.
Fig. 7. The distribution of different density from (a) to (c) the density threshold is growing.
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
193
induced X-ray emission (PIXE) tomography will be combined for elemental analysis of micrometer-sized samples. Acknowledgement The authors wish to thank accelerator group for good beam condition during experiments. References
Fig. 8. Three-dimensional image of target. An artificial cut has been made along the symmetry axis for illustration of the internal structure as well as the target wall surface. (a) front of artificial cut and (b) side of artificial cut.
illustrate the shape of the object as well as its internal structure. The inside surface was finish and the dimensions of the hohlraum are of symmetry. Therefore it is illustrated that the surface finish and the dimensions of the mandrel be well controlled in the target fabrication [1]. However the outside surface was coarse. 5. Conclusions STIM-CT on the beam line at the FUDAN nuclear microprobe laboratory can be applied to analyse the 3D structure of inertial confinement fusion targets, providing submicron-scale characterization. The method can apply not only to PVA material for ICF capsules, but also to other materials being used for capsules such as beryllium or high-density carbon (diamond). The projection data reveal the need for careful corrections of misalignment of the rotation centers obtained for every slice. For more accurate quantitative STIM tomography in further experiments, the rotation axis alignment will be performed with higher precision. Also a more accurate quantitative mass density will be determined. Particle
[1] K.R. Schultz, J.L. Kaae, W.J. Miller, D.A. Steinman, R.B. Stephens, Status of inertial fusion target fabrication in the USA, Fusion Engineering and Design 44 (1999) 441–448. [2] U. Kubo, H. Tsubakihara, Development of a coating technique for inertial confinement fusion plastic targets, Journal of Vacuum Science and Technology A 4 (1986) 1134–1137. [3] A.J. Antolak, A.E. Pontau, D.H. Morse, Ion microtomography and particleinduced X-ray emission analysis of direct drive inertial confinement fusion targets, Journal of Vacuum Science and Technology A 10 (1992) 1164–1169. [4] L.A. Scott, R.G. Schneggenburger, P.R. Anderson, Deuterium permeation properties of beta-irradiated and unirradiated poly(vinyl alcohol) and polystyrene shells, Journal of Vacuum Science and Technology A 4 (1986) 1155. [5] U. Kubo, H. Tsubakihara, Development of a coating technique for inertial confinement fusion plastic targets, Journal of Vacuum Science and Technology A 4 (1986) 1134. [6] M. Takagi, M. Ishihara, T. Norimatsu, T. Yamanaka, Y. Izawa, S. Nakai, Development of foam shell with plastic ablator for cryogenic laser fusion target, Journal of Vacuum Science and Technology A 11 (1993) 2837. [7] U. Kubo, H. Nakano, Hyo-gun Kim, Fabrication of cross-linked polymer shells for inertial confinement fusion experiments, Journal of Vacuum Science and Technology A 15 (3) (1997) 683–685. [8] C. Brundle, C. Evans, S. Wilson, Encyclopedia of Materials Characterization, Butterworth-Heinemann, Manning, 1992. [9] P.P. Zhu, K. Zhang, Z.L. Wang, Y.J. Liu, X.S. Liu, Z.Y. Wu, S.A. McDonald, F. Marone, M. Stampanoni, Low-dose, simple, and fast grating-based X-ray phase-contrast imaging, Proceedings of the National Academy of Sciences of the United States of America 107 (2010) 13576–13581. [10] J.Y. Zhang, J. Gao, F. Jiang, J. Chen, B. Chen, H. Ming, X-ray phase-contrast computed tomography, Chinese Journal of Quantum Electronics 22 (2005) 579–584. [11] Ren Minqin, J.A. van Kan, A.A. Bettiol, Lim Daina, Chan Yee Gek, Bay Boon Huat, H.J. Whitlow, T. Osipowicz, F. Watt, Nano-imaging of single cells using STIM, Nuclear Instruments and Methods in Physics Research Section B 260 (2007) 124–129. [12] C. Michelet, Ph. Moretto, 3D mapping of individual cells using a proton microbeam, Nuclear Instruments and Methods in Physics Research Section B 150 (1999) 173. [13] Eric Stindel, K. Jayaram, Udupa, E. Bruce, D. Hirsch, Odhner, Christine Couture, Computerized Medical Imaging and Graphics 23 (1999) 75. [14] R. Wu, H. Shen, Y. Mi, M.D. Sun, M.J. Yang, A computer program for scanning transmission ion microscopy simulation, Nuclear Instruments and Methods in Physics Research Section B 231 (2005) 149–155. [15] A.C. Kak, Principles of Computerized Tomographic Imaging, IEEE Press, New York, 1987. [16] T. Andrea, M. Rothermel, T. Butz, T. Reinert, The improved STIM tomography set-up at LIPSION: three-dimensional reconstruction of biological samples, Nuclear Instruments and Methods in Physics Research Section B 267 (2009) 2098–2102. [17] C. Michelet-Habchi, S. Incerti, P. Aguer, Ph. Barberet, E. Gontier, K. Grente, Ph. Moretto, D.T. Nguyen, A. Pouthier, T. Pouthier, F. Rebillat, R.W. Smith, TomoRebuild: a new data reduction software package for scanning transmission ion microscopy tomography, Nuclear Instruments and Methods in Physics Research Section B 231 (2005) 142–148. [18] M. Schwertner, A. Sakellariou, T. Reinert, T. Butz, Scanning transmission ion micro-tomography (STIM-T) of biological specimens, Ultramicroscopy 106 (2006) 574–581. [19] M. Wegdén, M. Elfman, P. Kristiansson, N. Arteaga Marrero, V. Auzelyte, K.G. Malmqvist, C. Nilsson, J. Pallon, Development of a microtomography system at the Lund sub-micron beamline, Nuclear Instruments and Methods in Physics Research Section B 249 (2006) 756–759. [20] T. Reinert, A. Sakellariou, M. Schwertner, J. Vogt, T. Butz, Scanning transmission ion microscopy tomography at the Leipzig nanoprobe LIPSION, Nuclear Instruments and Methods in Physics Research Section B 190 (2002) 266–270. [21] N. Gordillo, C. Habchi, L. Daudin, A. Sakellariou, F. Delalée, Ph. Barberet, S. Incerti, H. Seznec, Ph. Moretto, Technical developments for computed tomography on the CENBG nanobeam line, Nuclear Instruments and Methods in Physics Research Section B 269 (2011) 2206–2209. [22] G. Bench, K.A. Nugent, M. Cholewa, A. Saint, G.J.F. Legge, Submicron STIM tomography reconstruction techniques, Nuclear Instruments and Methods in Physics Research Section B 54 (1991) 390. [23] C. Habchi, N. Gordillo, S. Bourret, Ph. Barberet, C. Jovet, Ph. Moretto, H. Seznec, Beyond filtered backprojection: a reconstruction software package for ion beam microtomography data, Nuclear Instruments and Methods in Physics Research Section B 295 (2013) 42–49.
194
Y.Q. Li et al. / Fusion Engineering and Design 88 (2013) 188–194
[24] AMIRA web site. Available from: http://www.tgs.com [25] C. Habchi, D.T. Nguyen, Ph. Barberet, S. Incerti, Ph. Moretto, A. Sakellariou, H. Seznec, Reconstruction of 3D ion beam micro-tomography data for applications in cell biology, Nuclear Instruments and Methods in Physics Research Section B 267 (2009) 2107–2112. [26] A. Pontau, A.J. Antolak, D.H. Morse, A.A. Ver Berkmoes, J.M. Brase, D.W. Heikkinen, H.E. Martz, I.D. Procter, Ion Microbeam tomography, Nuclear Instruments and Methods in Physics Research Section B 40/41 (1989) 646–650.
[27] A.J. Antolak, An empirical treatment of nonrelativistic proton beam spreading, Nuclear Instruments and Methods in Physics Research Section B 30 (1988) 182–184. [28] C. Michelet, Ph. Moretto, G. Laurent, W.J. Przybylowicz, V.M. Prozesky, C.A. Pineda, Ph. Barbaret, F. Lohste, J. Kennedy, Measurement of lateral straggling using a microbeam, Nuclear Instruments and Methods in Physics Research Section B 181 (2001) 157–163. [29] M. Breese, D.N. Jamieson, P. King, Materials Analysis using Nuclear Microprobes, Wiley Press, New York (USA), 1996, 102 pp.