In situ characterisation of hydrocarbon layers in TEXTOR by laser induced ablation and laser induced breakdown spectroscopy

https://doi.org/10.1016/j.jnucmat.2010.11.055Get rights and content

Abstract

Laser based methods in combination with spectroscopy are proposed for in situ characterisation of the first wall in ITER and are tested presently and applied in TEXTOR. A ruby laser is used to heat up a defined area at a surface of a test limiter, which then leads to ablation. The released particles are observed spectroscopically either at the surface or in the edge of the tokamak plasma.

The absolute calibration affords the knowledge of the ablation process, especially the ablated debris compounds as a function of laser energy. Beyond that, the interaction of the released material with the edge plasma and the atomic and spectroscopic data of the species must be known, e.g. breakup of clusters and molecules, ionisation and excitation of atoms.

In laboratory experiments the ablation process has been investigated in detail. Carbon test samples with known layer properties are exposed by laser radiation in laboratory and in TEXTOR with comparable power densities.

Introduction

Laser based methods in combination with spectroscopy have been proposed as in situ diagnostic for characterisation of the first wall [1], [2]. In contrast to laser induced desorption spectroscopy (LIDS) [3], which by moderate heating only desorbs volatile gases, mainly hydrogen, without destroying the layer and bulk material, laser induced ablation spectroscopy (LIAS) and laser induced breakdown spectroscopy (LIBS) are in principle destructive methods. The adsorbed laser energy leads to an evaporation of all wall materials, which are detected by spectroscopic methods in the laser produced plasma or in the edge plasma of a tokamak [4].

At JET the wall material ablated by the edge lidar laser was observed spectroscopically [5] and elements expected at the exposed wall element were identified. A quantitative evaluation was suffering from the unknown conversion factors and high fluctuation levels of the signals preventing a reliable evaluation of the measured data. Similar measurements were performed in a TEXTOR pilot experiment. A hydrocarbon layer of 140 nm thickness was deposited ex situ on a tungsten bulk limiter in a dc assisted glow discharge (PADOS [6]) and characterised by surface analysis methods. The limiter was then positioned in TEXTOR just 1 cm outside the last closed flux surface ra = 47 cm and exposed to an intense ruby laser for 10 ns during the discharge. The layer was removed completely at a laser energy fluence of 0.25 J/cm2 in a single exposure (Fig. 1a), without melting the bulk tungsten material. The observed spectra during ablation from a side view spectrometer is shown in Fig. 1b. These experiments showed the principal possibility of the technique to detect the amount and composition of deposits in fusion devices by laser induced layer ablation in connection with local spectroscopy. However, the TEXTOR experiment showed also a much reduced ablation threshold of the C layer on W if compared with pure C and difficulties to obtain self-consistence with absolute values of the spectroscopic data.

Laser induced ablation was also applied ex situ on carbon tiles of the TEXTOR belt limiter (ALT II), in connection with LIBS spectroscopy of the laser plasma. With a laser fluence of 20 J/cm2 the measurements were performed in the net deposition zone where thick carbon layers deposited from long term TEXTOR campaign (5–10 μm in this case) which typically contain also boron from boronisation process of few %. The data showed in comparison to measurement on carbon bulk material a surprisingly fast removal of the layer in only about three laser shots, as can be seen in the LIBS spectra on the development of boron and hydrogen isotopes in consecutive laser shots in Fig. 2.

As a consequence of these results, more detailed investigations of the ablation mechanism have been started in lab experiments, concentrating on the analysis of the absolute number of released particles and their angle and species distribution in the dependence on laser energy fluence and material properties. These parameters determine largely the LIAS line intensities in connection with the ratio of ionisation and excitation rate coefficient for each transition. The LIBS plasma, which always proceeds the LIAS emission in the edge plasma, has different ionisation and excitation rate coefficients, is expanding rapidly and changes its parameter correspondingly. The plasma decays within 1 μs and neutral particles are formed by recombination.

This paper focuses on research of the ablation processes itself with special emphasis on the species and angle distribution of ablated carbon material in dependence on the laser energy fluence.

Section snippets

Experimental set-up

The investigation of the ablation process in the laboratory was performed in an UHV chamber as shown in Fig. 3. The base pressure is below 10−6 mbar, the samples are mounted on a target holder movable in xy direction and rotatable around z-axis. The laser light from a Q-switched ruby laser with 1 J maximum energy and 15 ns pulse duration is focussed onto the target under 45° to the surface normal direction, which is oriented into a cross beam quadrupole mass analyser to measure the composition

Results and discussion

The carbon atom density across the collector plate is presented in Fig. 5. The maximum is slightly shifted from the normal axis of the laser exposed area. The measured curve can be converted into polar coordinates and fits reasonably with a superposition of two cosine functions with different order of magnitude corresponding to I = A(cos α)n + Bcos α. The first term describes a jet of ablated atoms and the second the contribution of evaporated thermal particles during heat up and cooling phase [8].

Summary and conclusion

To prepare the application of laser induced ablation and breakdown spectroscopy to characterise the composition and amount of deposited wall material in situ in TEXTOR, basic measurements on the ablation process of carbon and carbon deposits on W have been performed with respect to the species composition and angular and energy distribution. With a ruby laser with ns pulse duration the absorbed energy fluence must be above 8 (10) J/cm2 in order to receive reproducible results on fine grain

References (10)

  • C.H. Skinner et al.

    J. Nucl. Mater.

    (2003)
  • C. Grisolia et al.

    J. Nucl. Mater.

    (2007)
  • B. Schweer et al.

    J. Nucl. Mater.

    (2009)
  • D.D.R. Summers et al.

    J. Nucl. Mater.

    (2001)
  • J. Winter et al.

    Nucl. Instrum. Methods

    (1987)
There are more references available in the full text version of this article.

Cited by (32)

  • Study of spectral intensity of the laser ablated tungsten plasma and ablation mass at various laser spot sizes and laser fluence in vacuum environment

    2023, Spectrochimica Acta - Part B Atomic Spectroscopy
    Citation Excerpt :

    One laser-based diagnostics named laser-induced breakdown spectroscopy (LIBS) is an ideal candidate for in situ, real-time, online diagnosing the fuel retention and impurity deposition processes on the first wall and divertor tiles in tokamaks [3,4]. Since Razdobarin and Federici firstly proposed in 2001 that using the LIBS technique to measure the dust on the plasma-exposed surfaces and the grooves of PFCs in the next generation of nuclear fusion devices (e.g., ITER) [5], a few investigations have been reported to diagnose the surface characterization on PFCs in WEST, KSTAR, FTU, JET, TEXTOR and TORE SUPRA Tokamaks using LIBS [6–9]. During the last decade, our group has carried out the systematically LIBS diagnostic experiments for study of the PWI processes in Experimental Advanced Superconducting Tokamak (EAST), such as the deuterium retention, and depth profiles of the lithium‑deuterium (LiD) co-deposition layer, other impurity elemental compositions like carbon (C), molybdenum (Mo), iron (Fe), silicon (Si), sodium (Na), calcium (Ca), et al. [10–12].

View all citing articles on Scopus
View full text