Table of contents

The chemical erosion properties of DIII-D tokamak graphite tiles are examined using in-situ hydrocarbon visible spectroscopy, along with knowledge of hydrocarbon dissociation and excitation rates. We previously inferred a long-term (eight years) reduction in divertor chemical sputtering of carbon from divertor surfaces. Ex-situ analysis of sample DIII-D tiles from both the divertor and main wall is also performed. Boron is found to be present on all plasma-facing surfaces with 5–34% surface coverage. Periodic application of 100 nm thick boronization films provides the source of boron, while plasma erosion and transport determine self-consistently the surface composition of boron and carbon. Chemical erosion yield (C/D⁺) is uniformly ∼1% in regions where the edge plasma is attached (i.e. electron temperature T_e > 10 eV), consistent with the yield expected from boron-doped graphite. The long-term four-fold reduction in chemical erosion yield at the outer divertor is consistent with boron-doping effects. Inferred yields in detached (T_e < 2 eV) plasmas are much lower (∼0.01%) than the expected yield of boron-doped graphite. However, experimental evidence shows that spectroscopy may be influenced by the dissociation and prompt redeposition of hydrocarbons. DIII-D divertor tile samples are exposed in the PISCES-B linear plasma device. Results indicate that the high net erosion rate in the PISCES device greatly modifies the surface and erosion properties of the plasma-facing surfaces, making quantitative yield assessments problematic.

The injection of hydrocarbons into the JET Gas Box divertor in ohmic and L-mode discharges has been used to calibrate the intrinsic hydrocarbon erosion yields and to determine the variation of the photon efficiency with the edge ion flux. Methane erosion yields of 5% for CD₄ and 3% for CH₄ were measured, with no sign of any flux dependence. The erosion yield of ethane/ethene was lower but increased with the edge ion flux (increasing edge plasma density, decreasing edge electron temperature) such that it dominated the molecular carbon source at high density. The methane and ethane photon efficiencies (the number of molecular dissociations per emitted photon) were both not constant, but were found to increase with increasing density (increasing edge ion flux, decreasing edge electron temperature).

Experimental results of graphite erosion under high fluxes by using a high flux irradiation device in Osaka University are reviewed. Results of chemical sputtering experiments by 5 keV D₃⁺ beams with fluxes from 0.5 × 10²¹ to 3.9 × 10²¹ D/m² s showed an increase in peak temperatures from 900 K to 980 K with flux for isotropic graphite and CFC. These peak temperatures are higher than those of low flux results (750 K-850 K, ≤ 10²⁰ m^-2 s^-1). The peak yields did not show clear flux dependence. Results of radiation-enhanced sublimation (RES) experiments by 5 keV Ar⁺ beams with fluxes from 3.5 × 10¹⁹ to 1.5 × 10²¹ Ar/m² s showed a significant decrease in the RES yields with flux for isotropic graphite and pyrolytic graphite. The flux dependence of the RES yields (total yields minus physical sputtering yields) was about ϕ^-0.26 (ϕ: flux), while the flux dependence in fluxes from ∼10¹⁷ to ∼10¹⁹ Ar/m² s was about ϕ^-0.07. Comparison between models and experiments for chemical sputtering and RES are also shown.

Investigations of chemical erosion of carbon (EK98) were carried out in hydrogen and deuterium low pressure discharges. Atomic hydrogen fluxes to the surface were obtained from the radiation of the Balmer lines (Γ_H = Γ_D = 2.6 × 10²¹ m^-2 s^-1). From the diffusion flow onto a surface, the reflection coefficients of graphite for H and D atoms were determined to be 40% and 65%, respectively. Erosion yields were derived from weight loss measurements and emission spectroscopy of the CH and C₂ bands at ion energies between 5–40 eV in the substrate temperature range of 300–1100 K. Spectroscopic results indicate the influence of higher hydrocarbons. An isotope effect of 2.5–3 was measured, which vanishes at room temperature and approx. 5 eV ion energy. Here, an erosion yield of 0.1% remains. The measurements were compared with model calculations and were in satisfying agreement.

Former measurements at PSI-1 have been completed by new investigations of the chemical erosion yield for a D⁺ flux density of 1.8 · D⁺/m² · s in the upgraded plasma generator PSI-2. Si doped CFC samples (NS 31) were exposed to a deuterium plasma. By increasing the neutral gas pressure the electron temperature could be kept below 10 eV. Changing the heat contact between the sample and the cooling components of the water cooled target holder the surface temperature was varied between 300 and 900 °C. The measured erosion yield of about 0.5% at a target temperature of 500 °C confirms with the decreasing flux dependence.

Over the past few years, the thermo-oxidation of codeposited films from ASDEX-U, TEXTOR, JET, TFTR and DIII-D has been studied, with widely varying erosion yields observed. In this paper, we present an overview of the work done to date on the thermo-oxidation of tokamak codeposits, and make comparisons with the oxidation of carbon substrate materials, and laboratory-produced carbon films.

Graphite erosion leading to the formation of carbon clusters and dust particles under high power deposition has been studied using an electron beam facility and in various types of controlled fusion devices with carbon walls. It is shown that brittle destruction of graphite takes place in both systems and it results in the production of debris of a sub-millimeter to millimeter size. In electron beam experiments fine grain graphites and carbon fiber composite have been subjected to intense transient heat loads with deposited energy densities of up to several tens of MJ m^-2 in order to simulate plasma disruptions (5 ms pulse duration) or vertical displacement events (100–1000 ms pulse duration). In a reversed field pinch the brittle destruction has been caused by a pronounced mode activity (and related power loads to the wall, over 1 GW m^-2) at the very edge of the plasma during wall locking phase of discharges. Particles or dust generated during the thermal load tests and in fusion devices have been collected and analyzed using optical and various electron microscopy methods.

The aim of this work is to present recent results of photon efficiency (D/XB) measurements and chemical erosion yields of hydrocarbons in the region of the outer divertor of ASDEX Upgrade.

In order to study the influence of plasma edge parameters on the photon efficiencies and therefore on the evaluation of the erosion yield, different hydrocarbons were injected in a localized region of the outer divertor. The emission of radiative species was studied via recently installed new lines of sight which are centred on the gas injector.

The photon efficiency D/XB (CH or CD bands) of methane is determined first, a the calibrated gas flux injected during an experiment. These values are compared with a second technique based on evaluation of the photon efficiencies from the spectroscopic measurements of photon flux densities (Hγ or Dγ) which, in their turn, are compared with ion flux densities derived from Langmuir probe measurements. The difference observed between these two methods is related to the fraction of hydrogen that is released as neutral after molecular break-up. The measured fraction is discussed in terms of possible molecular break-up reactions. In a similar way by introducing CH₄ or CD₄ a dependence of D/XB on the hydrogen mass is observed. The results show an increase of the photon efficiency by a factor of 1.3 to 1.8, in case of CD₄ compared to CH₄, resulting in an increase of the isotope-effect on the erosion yield derived from spectroscopic measurements. Erosion yields of 1–4% are obtained for the low ion flux density range (Γ_i < 10²² m² s^-1)

LHD (Large Helical Device) has been operated during 2 years after initiating the first plasma on March of 1998. During the first year (1st and 2nd experimental campaigns) the discharges were mainly carried out with a low magnetic field of B_t = 1.5 T. Before the 3rd experimental campaign begin in 1999, carbon divertor plates were installed in the vacuum chamber at the end of the divertor legs. These were connected to the ergodic layer for the purpose of heat removal and separation from the vacuum wall (SS316L). After the installation metal impurities decreased and high influxes of carbon and oxygen were measured. During accumulated number of discharges the metal impurities increased and returned finally to the former level. The re-deposition of metals on the carbon plates was also observed. However, discharges were successful and the impurity level did not lower critically the plasma performance. The behaviour of impurities, Z_eff (effective charge) and density limit is discussed and results from spectroscopic measurements are presented.

JET has conducted experiments specifically designed to tackle the issue of CD₄ redeposition at the divertor target plates. This paper reports the results obtained by these experiments, quantifies the percentage of prompt CD₄ redeposition and evaluates the importance of chemical erosion at the divertor in an ITER relevant regime such as the steady state ELMy H-mode.

This paper presents ERO-TEXTOR simulation calculations of chemical erosion of a carbon limiter exposed to the TEXTOR Scrape-Off-Layer. The spatial distribution of CD-emission near a testlimiter is calculated in dependence on the plasma parameters and on the sticking probability for hydrocarbons hitting the limiter surface. The comparison of D/XB values with experimental data on TEXTOR-94 indicates that eroded hydrocarbons returning to the limiter surface have an extremely low effective sticking. This can be understood in terms of a high re-erosion of the freshly deposited a-C:D layer. The comparison of the radial distribution of the CD- and CII-emission between simulation and experiment shows a good agreement.

In order to study the impurity production, recycling and power deposition a Si doped CFC test limiter (NS31) was used in TEXTOR-94. The release of impurities (C, Si, O, Cr, CD radicals) was measured spectroscopically. A reduced methane production was found in the Si doped graphite when compared to a pure graphite limiter. A smaller decrease of the carbon fluxes could also be observed. The limiter contained about 1%–1.5% of Si, but a relative Si flux (Si/D) from the Si doped CFC surface between 0.12% and 0.4% has been measured. A chemical erosion of Si due to formation of SiD_x has not been observed. Silicon evaporated from the surface at temperatures above 1500 °C. This led to an increase of Si concentration and total radiation losses from the plasma. Surface analysis shows the formation of microcracks and holes on the plasma exposed limiter surface. The released Si was deposited in the vicinity of the tangency point of the limiter. Whereas a Si depletion was observed in the area of highest power loading with values reaching in and in-between fibres values of 0.03% and 0.02% respectively.

In the chemical interaction of hydrogen plasmas with carbon-based plasma-facing components in fusion research important issues are still unresolved: the extrapolation of measured erosion yields to high particle fluxes as expected for a fusion reactor divertor and the transport and deposition of hydrocarbon molecules on the divertor plate and walls. At the occasion of the Carbon Workshop in a discussion session the present status of knowledge was defined, suggestions for improved measurements and data evaluation discussed and the most pressing data needs reviewed.

In response to continuing concerns over the use of carbon containing materials in the DT phase of a next-step fusion device, a range of techniques for removing accumulated dust, flakes and films from the vessel as well as diagnosing the total inventory without vessel intervention have been evaluated. These techniques are described and the results of initial proof-of-principle experiments are discussed.

Erosion of the divertor and first-wall plasma-facing components, tritium uptake in the re-deposited films, and direct implantation in the armour material surfaces surrounding the plasma, represent crucial physical issues that affect the design of future fusion devices. In this paper we present the derivation, and discuss the results, of current predictions of tritium inventory in ITER-FEAT due to co-deposition and implantation and their attendant uncertainties. The current armour materials proposed for ITER-FEAT are beryllium on the first-wall, carbon-fibre-composites on the divertor plate near the separatrix strike points, to withstand the high thermal loads expected during off-normal events, e.g., disruptions, and tungsten elsewhere in the divertor. Tritium co-deposition with chemically eroded carbon in the divertor, and possibly with some Be eroded from the first-wall, is expected to represent the dominant mechanism of in-vessel tritium retention in ITER-FEAT. This demands efficient in-situ methods of mitigation and retrieval to avoid frequent outages due to the reaching of precautionary operating limits set by safety considerations (e.g., ∼350 g of in-vessel co-deposited tritium) and for fuel economy reasons. Priority areas where further R&D work is required to narrow the remaining uncertainties are also briefly discussed.

A numerical simulation model for brittle destruction of carbon based materials (CBMs) based on crack generation by thermal shocks was developed. Crack generation in heated graphite samples depends on the grain anisotropy, the grain size, the failure stress distribution of the bonds connecting adjacent grains and on the temperature and the temperature gradient in the sample. A realistic value for the mean failure stress was obtained from a numerical simulation of destructive compression tests and from a comparison of calculated and measured onset of brittle destruction for volumetric heating. Graphite predamaging due to crack propagation into the depth of the sample or due to volumetric heating might occur, resulting in an enhancement of brittle destruction under cyclic heat loads for typical ITER-FEAT off-normal events. Results of a consistent analysis of brittle destruction and of target screening by dust particles and of enhancement of brittle destruction under cyclic heat loads are presented.

For operation within a wide range of plasma and magnetic parameters an open divertor structure of the Wendelstein 7-X stellarator (optimised HELIAS configuration with superconducting coils, main parameters: R = 5.5 m, a = 0.55 m, B = 3 T, rotational transform variable between 5/6 and 5/4) has been developed. The most critical components of the divertor are ten 3D-shaped target surfaces, located along the "helical edge" of the five-fold rotational symmetry magnetic configuration of the machine. The plasma outflow concentrates on the target areas, designed to withstand a maximum stationary heat flux of 10 MW/m². To prevent excessive radiation losses all plasma facing components will be manufactured using low-Z materials, e.g. CFC, B₄C. The paper presents the optimised design of the target plates, which takes into account the latest results of the physics, the requirements of the heating and diagnostic systems, the manufacturing costs and the mounting operations in order to build a technical reliable system.

Actively cooled plasma facing components armoured with carbon fibre composite flat tiles have been developed for Tore Supra Tokamak since 1994. After fabrication of prototypes successively tested in a high heat flux test bed the special bond between the carbon tiles and the hardened copper of the heat sink was particularly studied. Since 1997, 660 elements are in fabrication. The infrared acceptance test underlines the importance of the heat treatment during the manufacture and revealed some unexplained cracked tiles.

In future fusion devices, like in the stellarator Wendelstein 7-X, the target plates of the divertor will be exposed to heat loads up to power densities of 10 MW/m² for 1000 s. For this purpose actively cooled target elements with an internal coolant flow return, made of 2-D CFC armor tiles brazed onto a two tube cooling structure were developed and manufactured at the Forschungszentrum Jülich. Individual bent- and coolant flow reversal elements were used to achieve a high flexibility in the shape of the target elements. A special brazing technology, using a thin layer of plasma-arc deposited titanium was used for the bonding of the cooling structure to the plasma facing armor (PFA). FEM-simulations of the thermal and mechanical behavior show that a detachment of about 25% of the bonded area between the copper tubes and the PFA can be tolerated, without exceeding the critical heat flux at 15 MW/m² or a surface temperature of 1400 °C at 10 MW/m² by using twisted tape inserts with a twist ratio of 2 at a cooling water velocity of 10 m/s. Thermal cycling tests in an electron beam facility up to a power density level 10.5 MW/m² show a very good behavior of parts of the target elements, which confirms the performance under fusion relevant conditions. Even defected parts in the bonding interface of the target elements, known from ultrasonic inspections before, show no change in the thermal performance under cycling, which confirms also the structural integrity of partly defected regions.

After a brief description of the rationale of the material and geometry selection for each carbon armoured plasma facing components, this paper describes the European development of the two basic geometries, namely the monoblock and the flat tile. An overview of the non-destructive inspection techniques specifically developed for these components is also presented.

Non-destructive ultrasonic testing has been applied for the characterization of joints by means of a polymer transducer. One of the advantages of the polymer transducer is flexibility in its shape and the possibility to install multiple transducers in one probe, which can reduce the time for inspection. As a first step, the size effect of the transducer on the resolution and sensitivity was examined to detect the joint flaw. Transducers with circumferential angles of 5°, 10° and 30° were tested. For this test a small divertor element with a drilled hole was prepared, which simulates a joint defect. The transducers with angles of 30 could not characterize the size of the artificial joint flaw. On the contrary, the size of the artificial defect was successfully detected with an accuracy of 90% by means of the transducers with angles of 5 and 10°. From the viewpoint of the sensitivity of the detection of the joint flaw, the transducer with the angle of 10° is appropriate because it could detect the largest intensity of the reflected signal caused by the same artificial defect of the joint interface.

This paper presents thermal cycle experiments of CFC/DSCu mock-ups to investigate the thermal performance and the durability of the joining interface between CFC and DSCu after neutron-irradiation. Test specimens have a duplex structure of a 1D or 2D CFC armor and dispersion strengthened copper (DSCu) heat sink. The 5 mm thick CFC armors were brazed onto the DSCu heat sink with 1 mm thick OFHC-Cu interlayer by a Cu-Ti-Ag braze. The dimensions of the specimens are 23 mm^l × 15 mm^w × 22 mm^h. These specimens were irradiated with neutron fluences of 0.3–0.5 dpa for CFC at an irradiation temperature of 280–320 °C. After the neutron irradiation, thermal cycle experiments were carried out under a heat load of 5 MW/m² for 10 s to simulate the ITER steady state thermal conditions. As a result, all specimens withstood 1000 thermal cycles without failure. The surface temperature of 1D-CFC at 0.43 dpa reaches 650 °C while the surface temperature of non-irradiated 1D-CFC specimen is 250 °C. The surface temperature of 2D-CFC at 0.3 dpa reaches 800 °C. Higher surface temperature of the irradiated specimens can be attributable to a degradation of thermal conductivity of the CFC armors. The surface temperature of the 2D-CFC specimen irradiated at 0.45 dpa decreases from 1200 °C to 1000 °C during the thermal cycles, which could be caused by an annealing effect.

For the actively cooled first wall panels of the W7-X stellarator under construction at Greifswald, Germany, boron carbide (B₄C) coatings with a thickness up to 500 µm are being developed as plasma facing surface on the stainless steel component. Different coating technologies (vacuum plasma spray (VPS), atmospherical plasma spray (APS)) are investigated. Characterisation of the coatings comprises metallographic and SEM investigations, impurity determination and measurement of the thermal conductivity, characterisation of the sputtering behaviour and the behaviour under plasma exposure. First in-pile-tests in the tokamak TEXTOR-94 at the Research Centre Juelich have been performed. 10 poloidal limiter elements and 5 test limiters made of copper with a 170 µm thick B₄C VPS coating were exposed to plasma discharges with ohmic and additional heating. By varying the limiter positions and plasma density, different loading conditions were imposed. After exposure, the test limiters were removed for further analyses. First results showed the formation of small craters in the coating, probably due to arcing. Although melting occurred within a limited zone around the craters, the coatings did not detach from the substrate or show severe cracking outside the molten area.

Recent progress in R&D of SiC fibers and reinforced SiC matrix (SiC/SiC) composites in Japan, especially focusing on the activities of CREST-ACE program, is presented. Firstly, the present status of high performance SiC fiber development, such as Hi-Nicalon Type-S and Tyrano-SA, is provided. The high performance SiC matrix production by reaction sintering (RS) method improved in both strength and thermal conductivity are accomplished. The efforts to make appropriate fiber-matrix interfacial microstructure by CVI and PIP methods have been successful, resulting in the production of high strength and high fracture toughness SiC/SiC composites. Several joining processes using PIP, RS and mechanical fastener for composites are introduced. Dimensional stability under radiation damage has been studied by neutron and charged particle irradiation. The SiC/SiC composites prepared with Type-S SiC fiber with a stoichiometric composition did not exhibit mechanical property degradation. Based on the development of SiC composites, test module concepts to verify the advanced fluid systems including SiC/SiC/Be/He coolant blanket are presented.

Improved carbon-based plasma facing materials are developed by doping graphite with different carbides and optimizing the microstructure in order to refine their thermomechanical properties and to reduce both processes of chemical erosion. As starting material a mixture of mesophase carbon powder with mean particle size of 0.6 µm, and carbide powders (TiC, V₈C₇, WC, ZrC, SiC and B₄C) with particle sizes around 1 µm have been used. The carbides were added to the carbon powder up to a metal concentration of about 5 at.%. Most dopants showed a catalytic effect on graphitization, V₈C₇ being the most effective one. The materials obtained exhibit a high density, a low open porosity and good mechanical properties. Chemical erosion was reduced with doping, typically by a factor of the order of two.

The formation of carbon-based mixed materials is unavoidable on the plasma-facing components (e.g. first wall and divertor) of fusion devices when carbon is used together with other materials. On the surfaces of these components very different conditions with respect to particle and energy impact occur. To predict the mixed material formation under these conditions the precise knowledge of the fundamental mechanisms governing these interactions is essential. In this paper we present the results of carbon interaction with titanium and silicon, as model substances for metallic and covalent carbides, during thermal treatment. To perform basic studies of the reactions of carbon with different elements, thin carbon films are produced by electron beam evaporation on the different substrates under UHV conditions. All measurements for chemical analysis are performed using x-ray photoelectron spectroscopy (XPS). We discuss first the properties of the deposited carbon films. The carbon films are characterized on inert gold surfaces and are compared to bulk graphite. Annealing of the carbon films up to 970 K leads to a transition from a disordered carbon network into a graphitic structure. Preparation of carbon films at room temperature on titanium or silicon leads to a limited carbide formation at the carbon/substrate interface. Carbon deposited in excess of several monolayers is present in elementary form. Annealing of the samples leads to complete carbidization consuming the available carbon in both cases. Titanium reacts to TiC and additional substoichiometric carbide, silicon forms SiC with exact stoichiometry.

The 9th International Workshop on Carbon Materials was held on the 18th and 19th of September 2000 in Hohenkammer Castle nearby Munich, Germany. The symposium was jointly organised by the Max-Planck-Institute für Plasmaphysik (Garching) and Forschungszentrum Jülich, Germany. Previous symposia of this series took place at the Forschungszentrum Jülich, Germany (seven meetings) and once at the Royal Institute of Technology (KTH) in Stockholm, Sweden. Participation of experts from overseas is always essential to keep high scientific standards. Therefore, the Carbon Workshops have been organised as satellite ones to a large conference held in Europe. This time, it was in connection with the 21st Symposium on Fusion Technology (SOFT) held a week earlier in Madrid, Spain.

These workshops have been perceived as a forum for open discussion aiming at the evaluation of the behaviour and properties of carbon as a material for plasma facing components in controlled fusion devices. Unfortunately, the list of relevant materials for the first wall in a fusion reactor is short and carbon is one of only few choices; the others are beryllium and tungsten. A very important and attractive feature of carbon is, that its properties can be tailored to meet designed requirements regarding, for instance, thermal conductivity and mechanical strength. In this sense, "carbon" is a common name for a great variety of graphite-based materials: from highly purified isotropic graphite to multidirectional fibre composites and plasma sprayed layers, including those doped with boron, silicon, titanium or their carbides.

It has been a tradition of this workshop that each time three specific subjects have been selected for detailed discussions. The topics presented in Hohenkammer were:

• Erosion and Deposition

• Dust Formation and Tritium Retention

• New Carbon-based Materials

Basic scientific and technical aspects have always been in the scope of the meeting. Therefore, among 64 registered participants there were experts from institutes of plasma physics, material engineering laboratories and industrial companies. The scientific programme consisted of 17 invited talks, two guided summary & general discussion sessions and presentation of over 20 posters. The programme covered a wide range of topics spanning from fundamental processes underlying the reactions between hydrogen atoms and CH₃ radicals with carbonaceous surfaces to the development of plasma facing components for a next-step device. There were talks on the evaluation of spectroscopic data from the edge and divertor plasmas and, on the "wash-and-flush" concepts to remove dust from the vacuum vessel of fusion devices. Much attention was focused on the influence of hydrogen flux density (so-called "flux dependence") on the erosion rate of carbon. This volume contains a selection of works presented.

The presentations documenting that significant progress have been made in all the treated topics since the 8th Carbon Workshop. However, it has also been obvious that many open questions remain. In our view, the most important are:

• the flux dependence of carbon erosion and the evaluation of corresponding data;

• erosion and deposition of hydrogenated carbon layers because of their huge contribution to the total fuel inventory in plasma facing components;

• basic erosion processes (the energy gap for hydrogen erosion between thermal energies and about 10 eV);

• the influence of material mixing on the performance of plasma facing components.

Impressive progress has been achieved in the development and manufacturing of actively cooled plasma facing components. We are looking forward to see the results of the first large-scale application, in Tore Supra, which will be known in two years. It appears that a safe prediction for a material choice for a next-step device cannot be made on the basis of available data and much more efforts have to be undertaken to address open questions in the field. We will hopefully be able to present a number of answers to those open questions on the occasion of the 10th Carbon Workshop. Finally, the symposium was also a great opportunity to express deep gratitude and to warmly say "good bye" to some of the protagonists and pioneers in the field. Gottfried Veider, who acted as a member of the programme committee of this workshop, retired in the mid of 2000. Egon Vietzke, who devoted most of his scientific career to the basic studies and deep understanding of hydrogen - carbon interactions, will retire at the end of this year. We wish them good, good health and many exciting activities beyond carbon and fusion.

We would like to thank our sponsors: Max-Planck-Institute für Plasmaphysik and Forschungszentrum Jülich. Their support helped to keep the registration fee low and to make the meeting effective, fruitful and pleasant in the beautiful scenery of Hohenkammer. We thank our secretary, Mrs. Katharin Marx, for her efforts in the organisation of the symposium and Hans Meier for the creation and maintenance of the workshop web-page. We thank all the participants, authors and referees of the articles for their co-operation.

A list of the members of the Programme Committee