Carbon chemical erosion in high flux and low temperature hydrogen plasma

(1)

Carbon chemical erosion in high flux and low temperature

hydrogen plasma

Citation for published version (APA):

Westerhout, J. (2010). Carbon chemical erosion in high flux and low temperature hydrogen plasma. Technische Universiteit Eindhoven. https://doi.org/10.6100/IR675452

DOI:

10.6100/IR675452

Document status and date: Published: 01/01/2010 Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

(2)

(3)

(4)

Carbon chemical erosion in

high flux and low temperature

hydrogen plasma

Proefschrift

ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de rector magnificus, prof.dr.ir. C.J. van Duijn, voor een commissie aangewezen door het College voor Promoties in het openbaar te verdedigen op donderdag 1 juli 2010 om 14.00 uur

door

Jeroen Westerhout

(5)

Dit proefschrift is goedgekeurd door de promotoren:

prof.dr. N.J. Lopes Cardozo en

prof.dr. W.J. Goedheer

Copromotor:

dr.ir. G.J. van Rooij

This work is part of the research programme of the Foundation for Funda-mental Research on Matter (FOM), which is financially supported by the Netherlands Organisation for Scientific Research (NWO).

(6)

Carbon chemical erosion in

high flux and low temperature

(7)

A catalogue record is available from theEindhoven University of Technol-ogy Library

(8)

Few can foresee whither their road will lead them,

till they come to its end. J.R.R. Tolkien – The Lord of the Rings

(9)

(10)

Chemische erosie van koolstof

onder invloed van een

hoge flux van waterstofplasma

van een lage temperatuur

Samenvatting

Hoe lang houdt de wand van een fusiereactor het vol: uren (veel te kort) of jaren?

Deze vraag is van groot belang voor de fusiereactor ITER, het grote internationale experiment in de ontwikkeling van kernfusie als een schone, veilige en onuitputtelijke energiebron. In ITER wordt de fusiebrandstof – een heet waterstofplasma – bijeen gehouden met magneetvelden. Terwijl de temperatuur in het midden van de reactor extreem hoog is (200 miljoen graden Celsius), is het plasma veel kouder waar het de wand raakt. Deze plek, de “uitlaat” voor het reactieproduct helium, wordt de “divertor” ge-noemd. Hier is de plasmatemperatuur slechts 10.000 ◦_{C. De toestroom}

(“flux”) van waterstofionen is echter heel groot, tot wel 1024 _m−2 _s−1_{, en}

de warmtebelasting is meer dan 10 MW m−2_{. Als gevolg van deze extreme}

flux van deeltjes is de bestendigheid van materialen waar het plasma de wand raakt een kritische factor voor het succes van ITER en dus voor de ontwikkeling van fusie-energie. Koolstof is momenteel voorzien als mate-riaal voor een gedeelte van de divertor waar de grootste warmtebelasting plaatsvindt. Het doel van dit proefschrift is om de chemische erosie van koolstof te meten in ITER-achtige condities.

De experimenten in deze studie zijn gedaan in de lineaire plasmagene-rator Pilot-PSI bij het FOM-Instituut voor Plasmafysica Rijnhuizen. In dit apparaat wordt een bundel waterstofplasma van ∼1 cm breed gepro-duceerd, die naar een gekoeld target wordt geleid met behulp van een mag-neetveld tot wel 1.6 Tesla (vergelijkbaar met het magmag-neetveld in een MRI-scanner). De plasmadichtheid is in te stellen tussen de 1019_{en 10}21_m−3_en

(11)

plasmatemperatuur zo laag dat er slechts sprake is van chemische erosie en geen sputtering, het proces waarbij er atomen van het oppervlak worden weggeslagen. In Pilot-PSI zijn deeltjesfluxen van meer dan 1025

waterstof-ionen m−2 _s−1 _{en vermogensdichtheden van meer dan 10 MW m}−2

mo-gelijk. Dat betekent dat in Pilot-PSI alle omstandigheden (en meer!) zijn na te bootsen die relevant zijn voor de ITER-divertor.

Om te beginnen hebben we verschillende koolstof composieten bloot-gesteld aan een systematische reeks belastingen met verschillende plas-maflux en plasmatemperatuur en hebben deze vergeleken. Daarbij bleken alle composieten hetzelfde te reageren. Daarna is fine grain grafiet gese-lecteerd als testmateriaal voor onze studies.

Er zijn twee methodes gebruikt om de erosie te meten. Ten eerste een spectroscopische meting om de “bruto” erosie te bepalen. Dit is de stroom van koolstof die van het oppervlak af komt in de vorm van koolwaterstof-moleculen. Deze methode wordt veel gebruikt in huidige fusie-experimen-ten, zij het bij hogere plasmatemperaturen dan in de ITER-divertor worden verwacht. We hebben deze methode uitgebreid naar het lagere tempera-tuurgebied dat hier bestudeerd wordt. Hier hebben we aangetoond dat in dit temperatuurgebied dissociatieve recombinatie van koolwaterstofio-nen het dominante proces is wat leidt tot de emissie van licht. Dit in tegenstelling tot het hogere temperatuurgebied waar excitatie door elek-tronenbotsingen het belangrijkste excitatiemechanisme is.

Naast de bruto erosie werd ook de netto erosie bepaald – de netto hoeveelheid materiaal die van het oppervlak verdwenen is ten gevolge van de erosie. De netto erosie kan significant verschillen van de bruto erosie, omdat een gedeelte van het geërodeerde materiaal weer kan neer-slaan op het target (“redepositie”). De netto erosie werd bepaald door het diepteprofiel van het oppervlak te meten na een volledige blootstelling. We zagen dat deze redepositie voornamelijk plaatsvindt bij een plasmatem-peratuur hoger dan 5000 ◦_{C. Dit kan worden verklaard door ionisatie en}

magnetische opsluiting van de koolwaterstofmoleculen die van het opper-vlak af komen. Bij lagere plasmatemperaturen (onder de 5000◦_{C) was er}

vrijwel geen redepositie. In die omstandigheden was er een goede overeen-stemming tussen de metingen van de bruto en de netto erosie.

(12)

netto erosie van fine grain grafiet systematisch bepaald voor een plasmaflux variërend tussen de 1 · 1023 _m−2 _s−1 _{en 4 · 10}24 _m−2 _s−1_{, bij verschillende}

plasmatemperaturen. De resultaten voor de effectiviteit van de bruto erosie (de yield) – de hoeveelheid koolstof atomen (“C”) die worden geërodeerd per inkomend waterstof ion (“H”) – zijn als volgt samen te vatten:

• De yield neemt sterk toe met de plasmatemperatuur: ∼0.6% C/H voor een temperatuur onder de 5000 ◦_{C, en ∼6% C/H voor een}

temperatuur boven de 10.000 ◦_{C, gemeten bij een plasmaflux van}

1 · 1024 _m−2 _s−1_{. Voor temperaturen boven de 10.000} ◦_{C is er geen}

verdere toename van de yield gemeten. Deze afhankelijkheid van de plasmatemperatuur kan (gedeeltelijk) veroorzaakt zijn door de op-pervlaktetemperatuur van het target. In de experimenten met een plasmatemperatuur boven de 10.000◦_{C was de}

oppervlaktetempera-tuur rond de 500 ◦_{C, waarbij de chemische erosie maximaal is. In}

experimenten met plasmatemperaturen onder de 5000 ◦_{C was het}

oppervlak kouder (200–300 ◦_C).

• Zowel bij plasmatemperaturen van 5000◦_{C als bij boven de 10.000} ◦_C

is de erosie yield omgekeerd evenredig met de plasmaflux. Dit im-pliceert dat de bruto erosiesnelheid – de hoeveelheid eroderend ma-teriaal per seconde – onafhankelijk is van de plasmaflux.

De centrale conclusie is daarom dat de bruto erosie van koolstof bij de plasmafluxen en temperatuurgebieden die verwacht worden in ITER verzadigd is: het kan niet sneller. Dit geeft aan dat de maximale erosie-snelheid wordt bepaald door de tijd die het duurt om een koolwaterstof-molecuul te vormen en los te laten komen van het oppervlak.

Met deze informatie kunnen we berekenen hoe lang een koolstofwand het vol kan houden in de divertor van ITER. Voor een plasmatemperatuur van onder de 5000 ◦_{C vinden we een erosiesnelheid van ∼6 · 10}21 _koolstof

atomen m−2 _s−1 _{(er erodeert per seconde een laag van ∼0.065 µm).}

Rede-positie speelt hier geen rol, aangezien de bruto en netto erosie gelijk zijn. Het is mogelijk wel een onderschatting, omdat de oppervlaktetemperatuur laag is. Voor plasmatemperaturen boven de 10.000 ◦_{C is de}

(13)

is een bovengrens vanwege het recirculeren van koolstof. Het is daarom aannemelijk dat de effectieve erosiesnelheid tussen deze twee waarden ligt. Dit resulteert in een verwachte levensduur van 4 tot 40 uur (wat over-eenkomt met 40 tot 400 ontladingen) voor een ITER-divertor van kool-stof. Deze voorspelling wijst erop dat een koolstof divertor mogelijk een beperkende factor is voor het bedrijf van ITER. Bij deze conclusie is uit-sluitend de erosiesnelheid in beschouwing genomen. Andere complicerende factoren zoals het (ongewenst) vasthouden van waterstof en de vorming van stofdeeltjes door een koolstofwand zijn hier niet in meegenomen.

Het feit dat de bruto erosiesnelheid verzadigd is geeft aan dat er weinig ruimte is om deze te verminderen. Wel zorgt de redepositie van koolstof er voor dat de netto erosie beduidend lager kan zijn dan de bruto erosie. De grootte van deze reductie hangt echter af van zaken als de geometrie van de divertor en transport van geërodeerd materiaal, en is op basis van resultaten in deze studie niet te voorspellen.

(14)

Carbon chemical erosion in

high flux and low temperature

hydrogen plasma

Summary

How long will the wall of a fusion reactor last: hours (i.e. much too short) or years?

This question is investigated for the conditions foreseen in the fusion reactor ITER, the world’s joint experiment in the development of nuclear fusion as a clean, safe and inexhaustible energy source. In ITER, hot hydrogen plasma is confined by magnetic fields. Whereas the temperature in the center of the reactor is extremely high (200 million◦_{C), it is much}

lower where the plasma touches the wall. The contact area is situated in the so-called divertor. Here, the plasma temperature is a mere 1 eV (ten thousand ◦_{C), but the flux density of hydrogen ions is very high, up to}

1024 _m−2 _s−1_{, and the power flux can exceed 10 MW m}−2_{. As a result of}

these extreme flux densities, the resilience of the material at the contact area is a critical factor for the success of ITER, and indeed, fusion energy. Carbon is presently the material selected for part of the ITER divertor. The aim of this thesis is to measure its chemical erosion rate in ITER-like conditions.

The experiments were carried out in the linear plasma generator Pilot-PSI at the FOM Institute for Plasma Physics Rijnhuizen. In this device a ∼1 cm diameter hydrogen plasma beam is produced, which is guided to a cooled target by a magnetic field of up to 1.6 T. The plasma density ne and temperature Te at the target can be varied in the range 1019 < ne <1021 m−3 and 0.1 < Te <4 eV. Like in ITER, Te is so low that no physical sputtering occurs, which would invalidate the experiment. Flux densities of more than 1025 _{hydrogen ions m}−2 _s−1_{, and tens of MW m}−2

can be realized. Hence, the parameter regime accessible with Pilot-PSI fully covers the conditions relevant to the ITER divertor.

(15)

density at different plasma temperatures in Pilot-PSI. After a comparison of different carbon composites, which turned out to behave similarly, we selected fine grain graphite as the test material for our studies.

Two methods were used to measure the erosion. Firstly, a spectro-scopic measurement was used to determine the “gross” erosion, i.e. the instantaneous flux of carbon that leaves the surface in the form of hydro-carbon molecules. This method, which in present fusion experiments is commonly used at higher plasma temperatures, was extended to the tem-perature range studied here (Te ∼1 eV). A different excitation process was shown to prevail: dissociative recombination of hydrocarbon ions rather than electron impact excitation. The latter is dominant at Te >3 eV.

Secondly, the “net” erosion, i.e. the net amount of material lost from an entire exposure, was determined by postmortem surface profilometry. The net erosion can differ substantially from the gross erosion because part of the eroded material can be redeposited. It was observed that redeposition occurs primarily in conditions with plasma temperatures Te >0.5 eV. This can be understood as being the result of the ionization and subsequent magnetic trapping of the hydrocarbon molecules coming off the surface. At lower plasma temperatures (Te < 0.5 eV) very little redeposition was observed. In those cases there was good correspondence between the gross erosion as measured with spectroscopy and the postmortem net erosion measurements.

Having established and calibrated the methodology, the gross and net erosion of fine grain graphite was systematically assessed, for plasma flux densities ΓH varied from 1 · 1023 m−2 s−1 to 4 · 1024 m−2 s−1, at different

plasma temperatures. The results for the gross erosion yield Ychem, i.e.

the number of eroded carbon atoms per incoming hydrogen ion, can be summarized as follows:

• The chemical erosion yield increases strongly with increasing Te: Ychem ∼0.6% C/H for Te <0.5 eV, and ∼6% C/H for Te>1.0 eV, measured at ΓH = 1 · 1024 m−2 s−1. For Te >1.0 eV no further in-crease of Ychemwas observed. The Tedependence may in part be due to the surface temperature of the target. In the experiments with Te > 1.0 eV the surface temperature was around 500 ◦C, at which

(16)

the chemical erosion is at its maximum. For Te<0.5 eV the surface was cooler (200–300 ◦_C).

• Both at Te∼0.5 eV and Te>1.0 eV, Ychem is inversely proportional

to the plasma flux density: Γ−1

H . This implies that the gross erosion

rate – the amount of material eroded per second – is independent of the flux density.

Thus, the central conclusion is that the gross erosion of carbon at the flux densities and temperature ranges expected in ITER is saturated: it cannot go any faster. A likely explanation is that the hydrogenation of the carbon at the surface is rate-limiting; the time it takes for a hydrocarbon molecule to form and leave the surface is longer than the average consecu-tive ion impact time (∼10−4_{s at a plasma flux density of 1 · 10}23 _m−2 _s−1_).

Using this information to calculate the lifetime of a carbon wall in the ITER divertor we find for Te < 0.5 eV a rate of ∼6 · 1021 carbon atoms m−2 _s−1 _{(∼65 nm s}−1_{). This measurement is not affected by redeposition}

since the gross erosion is equal to the net erosion, but it is an underesti-mation as the surface temperature is low. For Te>1.0 eV the erosion rate is ∼6 · 1022 _{carbon atoms m}−2 _s−1 _{(∼0.65 µm s}−1_{). This measurement is}

probably an upper estimate due to the recycling of carbon. It is there-fore very likely that the effective erosion rate will be in between these two limiting estimates.

This results in an estimated lifetime of 4 to 40 hours (corresponding to 40 to 400 discharges) for the ITER divertor if it is made of carbon. This implies that the carbon divertor is likely to be a limiting factor for ITER operation. This conclusion is based on a consideration of the gross erosion rate alone; other complicating factors, such as the retention of hydrogen and the formation of dust, have not been considered.

The fact that the gross erosion rate is saturated indicates that there is little scope for measures that reduce this erosion rate. The redeposition of carbon does effectively reduce the net erosion in our experiment. The ex-tent of this reduction will depend on material migration and the geometry of the divertor, and these effects have not been assessed in this thesis.

(17)

(18)

Introduction

Where would we be without energy? We (people) need energy to survive. Burning fossil fuels is a popular way to produce useable energy, but its resources are depleting and it accelerates global warming. In order to provide the coming generations with a world worth living in, we have to develop ways to create clean and sustainable energy.

There are plenty of sustainable energy sources, but only a few show real potential in solving the energy problem. Nuclear fusion is one of them. Fusion energy holds the promise to be clean, safe, inexhaustible, and available for everyone (no limitation on the availability of fuel). It is however difficult to achieve fusion on earth.

This thesis deals with one of the main challenges of fusion: the reactions in a fusion reactor occur in a hot hydrogen plasma, but this plasma touches the wall of the reactor. How long will the wall of a fusion reactor last: hours (i.e. much too short) or years?

The next generation fusion reactor: ITER

The lifetime of a fusion reactor wall is investigated for the conditions fore-seen in the fusion reactor ITER, the world’s joint experiment in the devel-opment of nuclear fusion. In a fusion reactor deuterium 2_{H (also written}

(23)

The divertor

Figure 1.1: The design of the ITER tokamak, with in the inset shown the exhaust of the machine: the divertor.

and a neutron:

2_{H +} 3_{H →} 4_{He + n + 17.6 MeV.} _(1.1)

This requires temperatures of 200,000,000 ◦_{C. At these temperatures the}

fuelling atoms are fully ionized and form a plasma. Because plasma consists of charged particles, it can be confined by magnetic fields. There are different ways to magnetically confine a fusion plasma. The easiest and most often used configuration is called a tokamak, which has a toroidally (doughnut) shaped plasma. The amount of fusion power depends on the volume of the plasma, whereas the amount of losses depend on the gradient of the temperature, so approximately on the width of the plasma. Hence: bigger is better.

(24)

Euro-pean Torus), located in the United Kingdom. JET, as well as other smaller machines, already had successful fusion plasma operation, but the power multiplication – the amount of power obtained from fusion reactions di-vided by the amount of power put into the plasma – is limited to ∼70%. In other words, JET is not big enough to produce energy from fusion.

The next generation fusion reactor is ITER, which is roughly 2 times bigger than JET. This results in a factor of 8 larger plasma volume. ITER will be big enough to have a power multiplication factor of 10, i.e. the obtained fusion power is 10 times higher than what is put into the plasma. A picture of ITER is shown in figure 1.1. ITER is still an experimental device, as it will not use the fusion power to produce electricity. After ITER, there will be the Demonstration Power Plant: DEMO. DEMO will be a full working fusion reactor that provides electricity to the grid.

A fusion reactor must have an exhaust to remove the waste products and to allow for new fuel to be injected. The exhaust of a fusion reactor is called a divertor, which is shown in the inset of figure 1.1. Here the fusion plasma is cooled down and neutralized in order to be pumped away. Interaction between the plasma and the material wall is a key issue for the success of the future fusion reactor ITER as the flux density of hydrogen ions is very high, up to 1024 _m−2 _s−1_{, and the power flux can exceed}

10 MW m−2_.

Materials choices

The selection of materials for use in a fusion reactor is limited. Not only the plasma facing components have to be chosen wisely, but also the material for the support structure and the vacuum vessel. For example: materials should be non-magnetic due to the high magnetic fields. More importantly, the materials have to be able to withstand high energy neutron damage. The neutrons released in the fusion process are not confined by the mag-netic field and can therefore penetrate through the vessel walls. This may cause a weakening of the material, or nuclear transmutation, which possi-bly leads to activated materials. It is therefore important that materials are chosen which have only short living radioactive transmutation prod-ucts.

(25)

There are three main candidate materials for the plasma facing compo-nents in ITER. The main wall will consist of beryllium (Be), as it is a light metal (atomic number Z = 4) with a relatively high melting temperature (close to 1300◦_{C). The efficiency of the fusion plasma depends on the}

ef-fective atomic number of the plasma Zeff. In ITER, Zeff should be kept

below ∼1.7 [1]. Beryllium contamination in the core plasma has therefore a relatively small effect on Zeff. The power and particle load on the divertor

will be much larger than on the main wall. The divertor will therefore con-sist of tungsten (W) and carbon (C). A disadvantage of tungsten is that it melts, albeit at a very high melting temperature (∼3400 ◦_{C). Tungsten is}

a heavy metal (atomic number Z = 74). If tungsten impurities penetrate into the core fusion plasma, due to the large number of electronic shells there are a large number of electronic transitions which result in radia-tive losses. These losses cool the plasma and decrease the efficiency of the fusion reactions dramatically. Carbon can withstand extreme heat loads and has a sublimation temperature of ∼3600◦_{C. Besides that carbon is a}

low-Z material (atomic number Z = 6) it has the advantage that it does not melt. This means that if carbon is sublimated, the surface properties after sublimation are the same as before sublimation, whereas tungsten melts and recrystallizes when cooled down again. Carbon is therefore con-sidered for ITER as wall material in the areas of strongest particle and power loads.

Carbon chemical erosion

A serious problem of carbon is that chemical processes induce erosion of the wall even at low incident particle energies. The immediate consequence is the compromise of the lifetime of the plasma facing component. Although this may be manageable in the present tokamaks, the particle fluence in ITER will be several orders of magnitude higher as both the plasma pulse duration and the particle flux will be much larger than in current machines. Furthermore, the eroded material will be deposited elsewhere as hydrogen rich amorphous layers and as such form a fuel (especially tritium) retention problem. Tritium is radioactive (with a half-life of 12.3 years), which sets a safety limit for the operation of a fusion reactor. Disintegration

(26)

of these deposited layers, for example by the impact of so-called ELMs (quasi-periodic bursts of power and particles reaching the material wall), contributes to dust formation and as such to an explosion risk.

If the wall tiles are eroded they can be replaced, but the fuel (deuterium and tritium) retained in the deposited layers can not easily be removed. Due to safety regulations, the total amount of tritium in the ITER machine may not exceed 700 g. Each fusion plasma pulse of 400 seconds will use 50 g of tritium [1]. It is therefore essential that the injected tritium is used in the fusion reaction, and not retained in hydrocarbon deposits. Since ITER is the next generation fusion reactor, this brings all new physics issues. The conditions expected in ITER are based on the performance of present day devices. A correct extrapolation is critical to predict whether ITER lasts only for one shot or whether it can operate for years before it has to be stopped for cleaning and maintenance.

The ion flux dependence of the chemical erosion of

carbon

It has been established that the chemical erosion of carbon depends on the surface temperature and the ion energy. The ion flux dependence has been investigated with many different machines. Data was selected for surface temperatures close to the maximum yield and the local plasma conditions were normalized to an ion impact energy of 30 eV. An overview of the results is presented by Roth et al. [2] and is shown in figure 1.2. It shows measurements with ion beams [3], on the linear machines PSI-1, PSI-2 [4,5], and PISCES B [6], and on the tokamaks JET [7], Tore Supra [8,9], TEXTOR [10], JT-60U [11], and ASDEX Upgrade (not shown) [12]. This flux dependence predicts a decrease in the chemical erosion yield at higher fluxes.

These machines combined cover a large flux range, but the very high flux range (∼1024 _m−2 _s−1_{) remains unexplored. In addition, the high}

flux data that mostly determine the flux dependence in this database were obtained at plasma temperatures larger than ∼10 eV, which means that physical sputtering was important in these experiments (but was corrected for). The high energy ions also damage the material lattice which increases

(27)

Ion flux (m-2_s-1₎

Chemical erosion yield (at/ion)

10-3

10-2

10-1

1019 ₁₀20 ₁₀21 ₁₀22 ₁₀23 ₁₀24

Ion beams IPP PSI-1 JT-60U Tore Supra 1999 Tore Supra 2002 TEXTOR PISCES JET 2001

Figure 1.2: The ion flux dependence of the carbon chemical erosion yield, taken from Roth et al. [2] with permission of the International Atomic Energy Agency (IAEA).

the chemical reactivity. This is not expected to be the case in the ITER divertor. ITER will rely on a divertor plasma that is partially detached from the wall to decrease the incident power to the target which requires temperatures down to 1–10 eV.

This work

The central question in this thesis is: what is the ion flux dependence of the carbon chemical erosion yield in the high flux regime (i.e. fluxes ∼1·1024_m−2_s−1_{) in the explicit absence of physical sputtering (i.e. plasma}

temperatures ∼1 eV)?

In order to address this question, we need a plasma generator capable of creating ITER-relevant plasma conditions. It also has to provide good diagnostic access and has to be suitable for well controlled carbon erosion experiments. At the FOM Institute for Plasma Physics Rijnhuizen we have developed such a machine: Pilot-PSI. It is the forerunner of the

(28)

larger machine Magnum-PSI, which will become operational in 2010. In Pilot-PSI a ∼1 cm diameter hydrogen plasma beam is produced, which is guided to a cooled target by a magnetic field of up to 1.6 T. The plasma density ne and temperature Te at the target can be varied in the range 1019_{< n}

e<1021m−3 and 0.1 < Te<4 eV. Like in ITER, Teis so low that no physical sputtering occurs, which would invalidate the experiment. Flux densities of more than 1025 _{hydrogen ions m}−2 _s−1_{, and tens of MW m}−2

can be realized. Hence, the parameter regime accessible with Pilot-PSI fully covers the conditions relevant to the ITER divertor.

Carbon targets are systematically exposed to scans of the plasma flux density at different plasma temperatures in Pilot-PSI. Two methods are used to measure the erosion. Firstly, a spectroscopic measurement is used to determine the “gross” erosion, i.e. the instantaneous flux of carbon that leaves the surface in the form of hydrocarbon molecules. This method, which in present fusion experiments is commonly used at higher plasma temperatures, is extended to the temperature range studied here. Sec-ondly, the “net” erosion, i.e. the net amount of material lost from an entire exposure, is determined by postmortem surface profilometry. The net erosion can differ substantially from the gross erosion because part of the eroded material can be redeposited.

Having established and calibrated the methodology, the gross and net erosion of fine grain graphite is systematically assessed, for plasma flux densities ΓH varied from 1 · 1023 m−2 s−1 to 4 · 1024 m−2 s−1, at different

plasma and surface temperatures.

Outline of this thesis

We first give a background on the chemical erosion of carbon in chapter2. Here we explain the physics of chemical erosion and define the important parameters that determine the erosion rate.

Chapter 3 [Phys. Scr. T128 (2007) 18] reports on the developments of the linear plasma generators at FOM Rijnhuizen. The design and the status of construction of Magnum-PSI are described. Results obtained at Pilot-PSI are presented that demonstrate ITER-relevant plasma conditions at the remote target.

(29)

Chapter 4 [Phys. Scr. T138 (2009) 014017] gives a comparison of the eroded carbon flux of different carbon composites, including the current ITER-reference material, determined with CH spectroscopy.

Spectroscopy on the molecular CH A − X band makes it possible to quantify in situ the chemical erosion of carbon wall elements in contact with hydrogen plasma. This is described in chapters 5[Appl. Phys. Lett.

95(2009) 151501] and 6 [Submitted to Nuclear Fusion].

Finally we can determine the flux dependence of the carbon chemical erosion in the high flux and low temperature regime. The effect of the plasma flux density, as well as the plasma and surface temperature on the chemical erosion rate is evaluated in chapter 7[To be submitted].

The conclusions and an outlook are given in chapter 8.

Publications

The following publications are related to this thesis:

• J. Westerhout, W.R. Koppers, W.A.J. Vijvers, R.S. Al, S. Brezin-sek, S. Brons, H.J.N. van Eck, R. Engeln, B. de Groot, R. Koch, H.J. van der Meiden, M.P. Nuijten, V. Philipps, M.J. van de Pol, P.R. Prins, U. Samm, J. Scholten, D.C. Schram, B. Schweer, P.H.M. Smeets, D.G. Whyte, E. Zoethout, A.W. Kleyn, W.J. Goedheer, N.J. Lopes Cardozo, and G.J. van Rooij

‘PSI research in the ITER divertor parameter range at the FOM PSI-lab’

Phys. Scr. T128 (2007) 18

• J. Westerhout, D. Borodin, R.S. Al, S. Brezinsek, M.H.J. ’t Hoen, A. Kirschner, S. Lisgo, H.J. van der Meiden, V. Philipps, M.J. van de Pol, A.E. Shumack, G. De Temmerman, W.A.J. Vijvers, G.M. Wright, N.J. Lopes Cardozo, J. Rapp, and G.J. van Rooij

‘Chemical erosion of different carbon composites under ITER-relevant plasma conditions’

(30)

• J. Westerhout, N.J. Lopes Cardozo, J. Rapp, and G.J. van Rooij, ‘CH spectroscopy for carbon chemical erosion analysis in high density low temperature hydrogen plasma’

Appl. Phys. Lett. 95 (2009) 151501

• J. Westerhout, D. Borodin, S. Brezinsek, N.J. Lopes Cardozo, J. Rapp, D.C. Schram, and G.J. van Rooij

‘The breakup of methane under ITER divertor hydrogen plasma con-ditions for carbon chemical erosion analysis with CH spectroscopy’ Submitted to Nuclear Fusion

• J. Westerhout, A.E. Shumack, N.J. Lopes Cardozo, J. Rapp, and G.J. van Rooij

‘Carbon chemical erosion in high flux and low temperature hydrogen plasma’

To be submitted

Other publications

• B. de Groot, R.S. Al, R. Engeln, W.J. Goedheer, O.G. Kruijt, H.J. van der Meiden, P.R. Prins, D.C. Schram, P.H.M. Smeets, V.P. Vere-miyenko, W.A.J. Vijvers, J. Westerhout, A.W. Kleyn, N.J. Lopes Cardozo, and G.J. van Rooij

‘Extreme hydrogen plasma fluxes at Pilot-PSI enter the ITER diver-tor regime’

Fusion Eng. Des. 82 (2007) 1861

• H.J. van der Meiden, R.S. Al, C.J. Barth, A.J.H. Donné, R. Engeln, W.J. Goedheer, B. de Groot, A.W. Kleyn, W.R. Koppers, N.J. Lopes Cardozo, M.J. van de Pol, P.R. Prins, D.C. Schram, A.E. Shumack, P.H.M. Smeets, W.A.J. Vijvers, J. Westerhout, G.M. Wright, and G.J. van Rooij

‘High sensitivity imaging Thomson scattering for low temperature plasma’

(31)

• W.A.J. Vijvers, C.A.J. van Gils, W.J. Goedheer, H.J. van der Mei-den, D.C. Schram, V.P. Veremiyenko, J. Westerhout, N.J. Lopes Cardozo, and G.J. van Rooij

‘Optimization of the output and efficiency of a high power cascaded arc hydrogen plasma source’

Phys. Plasmas 15 (2008) 093507

• A.E. Shumack, V.P. Veremiyenko, D.C. Schram, H.J. de Blank, W.J. Goedheer, H.J. van der Meiden, W.A.J. Vijvers, J. Westerhout, N.J. Lopes Cardozo, and G.J. van Rooij

‘Rotation of a strongly magnetized hydrogen plasma column deter-mined from an asymmetric Balmer-beta spectral line with two radi-ating distributions’

Phys. Rev. E 78 (2008) 046405

• G.M. Wright, A.W. Kleyn, E. Alves, L.C. Alves, N.P. Barradas, G.J. van Rooij, A.J. van Lange, A.E. Shumack, J. Westerhout, R.S. Al, W.A.J. Vijvers, B. de Groot, M.J. van de Pol, H.J. van der Meiden, J. Rapp, and N.J. Lopes Cardozo

‘Hydrogenic retention in tungsten exposed to ITER divertor relevant plasma flux densities’

J. Nucl. Mater. 390–391 (2009) 610

• W.A.J. Vijvers, B. de Groot, R.S. Al, M.A. van den Berg, H.J.N. van Eck, W.J. Goedheer, A.W. Kleyn, W.R. Koppers, O.G. Kruijt, N.J. Lopes Cardozo, H.J. van der Meiden, M.J. van de Pol, P.R. Prins, J. Rapp, D.C. Schram, A.E. Shumack, P.H.M. Smeets, J. Westerhout, G.M. Wright, and G.J. van Rooij

‘Multiple discharge channels in a cascaded arc to produce large di-ameter plasma beams’

Fusion Eng. Des. 84 (2009) 1933

• G.J. van Rooij, H.J. van der Meiden, M.H.J. ’t Hoen, W.R. Koppers, A.E. Shumack, W.A.J. Vijvers, J. Westerhout, G.M. Wright, and J. Rapp

(32)

‘Thomson scattering at Pilot-PSI and Magnum-PSI’

Plasma Phys. Contr. Fusion 51 (2009) 124037

• J. Rapp, G.J. van Rooij, A. Litnovsky, L. Marot, G. De Temmerman, J. Westerhout, and E. Zoethout

‘Temperature effect on hydrocarbon deposition on molybdenum mir-rors under ITER-relevant long-term plasma operation’

Phys. Scr. T138 (2009) 014067

• G. De Temmerman, R.P. Doerner, P. John, S. Lisgo, L. Marot, S. Porro, P. Petersson, D.L. Rudakov, G.J. van Rooij, J. Westerhout, and J. Wilson

‘Interactions of diamond surfaces with fusion relevant plasmas’

Phys. Scr. T138 (2009) 014013

• G.M. Wright, R.S. Al, E. Alves, L.C. Alves, N.P. Barradas, A.W. Kleyn, N.J. Lopes Cardozo, H.J. van der Meiden, V. Philipps, G.J. van Rooij, A.E. Shumack, W.A.J. Vijvers, J. Westerhout, E. Zoet-hout, and J. Rapp

‘Carbon film growth and hydrogenic retention of tungsten exposed to carbon-seeded high density deuterium plasmas’

J. Nucl. Mater. 396 (2010) 176

• D. Borodin, A. Kirschner, D. Nishijima, R. Doerner, J. Wester-hout, G.J. van Rooij, J. Rapp, A. Kreter, R. Ding, A. Galonska, and V. Philipps

‘Modelling of impurity transport in the linear plasma devices PIS-CES-B and Pilot-PSI using the Monte-Carlo code ERO’

Accepted for publication in Contributions to Plasma Physics

• G.M. Wright, J. Westerhout, R.S. Al, E. Alves, L.C. Alves, N.P. Barradas, M.A. van den Berg, D. Borodin, S. Brezinsek, S. Brons, H.J.N. van Eck, B. de Groot, A.W. Kleyn, W.R. Koppers, O.G.

(33)

Kruijt, J. Linke, N.J. Lopes Cardozo, M. Mayer, H.J. van der Mei-den, P.R. Prins, G.J. van Rooij, J. Scholten, A.E. Shumack, P.H.M. Smeets, G. De Temmerman, W.A.J. Vijvers, and J. Rapp

‘Materials research under ITER-like divertor conditions at FOM Rijn-huizen’

Accepted for publication in Journal of Nuclear Materials

Conference contributions

• J. Westerhout, G.J. van Rooij, R. Engeln, B. de Groot, A.W. Kleyn, N.J. Lopes Cardozo, P.H.M. Smeets, and W.J. Goedheer ‘Cavity ring down spectroscopy on Pilot-PSI’

Poster on the 8th _{Workshop on the Exploration of Low Temperature}

Plasma Physics, Kerkrade, The Netherlands, November 24–25 2005 • J. Westerhout, G.J. van Rooij, R. Engeln, H.J. van der Meiden,

W.A.J. Vijvers, B. de Groot, R.S. Al, A.W. Kleyn, N.J. Lopes Car-dozo, P.H.M. Smeets, and W.J. Goedheer

‘Balmer-alpha absorption spectroscopy on the Pilot-PSI plasma jet’ Oral on the 18th _{CPS/NNV Spring Symposium for Plasma Physics}

and Radiation Technology, Lunteren, The Netherlands, March 22–23 2006

• G.J. van Rooij, J. Westerhout, H.J.N. van Eck, W.R. Koppers, V.P. Veremiyenko, W.J. Goedheer, B. de Groot, P.H.M. Smeets, R. Engeln, D.C. Schram, N.J. Lopes Cardozo, and A.W. Kleyn

‘Carbon exposed to the intense hydrogen plasma jet of Pilot-PSI’ Poster on the 17th _{International Conference on Plasma Surface}

In-teractions in Controlled Fusion Devices, Hefei, China, May 22–26 2006

• J. Westerhout, W.A.J. Vijvers, R.S. Al, S. Brezinsek, R. Engeln, B. de Groot, H.J. van der Meiden, P.R. Prins, D.C. Schram, D.G. Whyte, E. Zoethout, A.W. Kleyn, W.J. Goedheer, N.J. Lopes Car-dozo, and G.J. van Rooij

(34)

‘Carbon erosion in Pilot-PSI’

Poster on the 11th _{International Workshop on Plasma Facing}

Mate-rials and Components for Fusion Applications, Greifswald, Germany, October 10–12 2006

• J. Westerhout, W.A.J. Vijvers, R.S. Al, S. Brezinsek, R. Engeln, B. de Groot, H.J. van der Meiden, M.P. Nuijten, M.J. van de Pol, P.R. Prins, D.C. Schram, D.G. Whyte, E. Zoethout, A.W. Kleyn, W.J. Goedheer, N.J. Lopes Cardozo, and G.J. van Rooij

‘First measurements on carbon erosion in Pilot-PSI’

Oral on the 9th _{Workshop on the Exploration of Low Temperature}

Plasma Physics, Kerkrade, The Netherlands, November 23–24 2006 • J. Westerhout, W.A.J. Vijvers, A.W. Kleyn, N.J. Lopes Cardozo,

and G.J. van Rooij

‘Carbon exposed to ITER-relevant conditions in the Magnum-PSI programme’

Poster on the FOM-meeting Physics@Veldhoven 2007, Veldhoven, The Netherlands, January 23–24 2007

• J. Westerhout, W.A.J. Vijvers, R.S. Al, S. Brezinsek, R. Engeln, B. de Groot, H.J. van der Meiden, R.J.E. van de Peppel, M.J. van de Pol, P.R. Prins, D.C. Schram, A.E. Shumack, D.G. Whyte, E. Zoethout, A.W. Kleyn, W.J. Goedheer, N.J. Lopes Cardozo, and G.J. van Rooij

‘Carbon chemical erosion yield experiments in the ITER flux regime and beyond’

Poster on the 19th_{CPS/NNV Spring Symposium for Plasma Physics}

and Radiation Technology, Lunteren, The Netherlands, March 7–8 2007 [awarded with the best poster prize]

• J. Westerhout, R.S. Al, S. Brezinsek, B. de Groot, A.J. van Lange, J. Leijssen, H.J. van der Meiden, M.J. van de Pol, P.R. Prins, A.E. Shumack, L.W. Veldhuizen, W.A.J. Vijvers, G.M. Wright, J. Rapp, N.J. Lopes Cardozo, and G.J. van Rooij

(35)

‘Carbon chemical erosion yield experiments in the ITER flux regime and beyond’

Oral on the 10th _{Workshop on the Exploration of Low Temperature}

Plasma Physics, Kerkrade, The Netherlands, November 15–16 2007 • J. Westerhout, R.S. Al, M.A. van den Berg, S. Brezinsek, C.A.J.

van Gils, B. de Groot, W.R. Koppers, A.J. van Lange, J. Leijssen, H.J. van der Meiden, M.J. van de Pol, P.R. Prins, A.E. Shumack, L.W. Veldhuizen, W.A.J. Vijvers, G.M. Wright, J. Rapp, N.J. Lopes Cardozo, and G.J. van Rooij

‘Temperature dependence of carbon erosion under ITER-like plasma conditions’

Oral on the 20th _{CPS/NNV Spring Symposium for Plasma Physics}

and Radiation Technology, Lunteren, The Netherlands, March 4–5 2008

• J. Westerhout, R.S. Al, S. Brezinsek, R. Engeln, B. de Groot, A.J. van Lange, J. Leijssen, H.J. van der Meiden, M.J. van de Pol, P.R. Prins, D.C. Schram, A.E. Shumack, L.W. Veldhuizen, W.A.J. Vijvers, G.M. Wright, J. Rapp, N.J. Lopes Cardozo, and G.J. van Rooij

‘Carbon chemical erosion yield at ITER-relevant plasma fluxes’ Poster on the 18th _{International Conference on Plasma Surface}

In-teractions in Controlled Fusion Devices, Toledo, Spain, May 26–30 2008

• J. Westerhout, D. Borodin, G.J. van Rooij, R.S. Al, S. Brezinsek, M.H.J. ’t Hoen, A. Kirschner, S. Lisgo, H.J. van der Meiden, V. Philipps, M.J. van de Pol, A.E. Shumack, G. De Temmerman, W.A.J. Vijvers, G.M. Wright, N.J. Lopes Cardozo, and J. Rapp

‘Chemical erosion yields of different carbon composites under ITER-relevant plasma conditions’

Poster on the 12th_{International Workshop on Plasma Facing}

Materi-als and Components for Fusion Applications, Jülich, Germany, May 11–14 2009

(36)

Chapter 2

Background: chemical

erosion of carbon exposed to

a hydrogen plasma

2.1 Introduction

When carbon is exposed to a hydrogen plasma there are a number of erosion mechanisms, which are described in detail by Jacob and Roth [13], and Roth et al. [2,14]. The two most relevant mechanisms are:

• Chemical erosion (or chemical sputtering): chemical reactions be-tween thermal species from the gas phase or low energy (thermal) ions and surface atoms can create volatile molecules/radicals at the surface. These molecules are released when the surface tempera-ture is high enough or due to ion induced desorption. High energy ions induce damage to the material lattice and as such increase the chemical reactivity. This is then called radiation enhanced chemical erosion/sputtering.

• Physical sputtering: momentum transfer due to collisions between the incoming ions to the atoms in the material set off collision cas-cades in the target. When such cascas-cades recoil and reach the target surface with an energy above the surface binding energy, an atom

(37)

can be ejected. The physical sputtering threshold of graphite for sputtering by hydrogen is ∼35 eV [15].

Physical sputtering is a very destructive erosion mechanism with ero-sion yields approaching 1, i.e. 1 carbon surface atom is removed per in-coming ion, when the inin-coming ions are heavy particles (e.g. carbon, neon or argon) [16]. The threshold for physical sputtering by heavy ions is typ-ically > 50 eV. To prevent severe damage from physical sputtering it is essential to keep the ion energy low.

The ion impact energy Eion is estimated as Eion ∼ 2Ti + 3Te ∼ 5Te [17], where Ti and Te are the ion and electron temperature, respectively. In the divertor of ITER the plasma temperature will therefore be 1–10 eV. However, even at these low plasma temperatures there is still chemical ero-sion of carbon by hydrogen via the release of CH3 and other hydrocarbons.

This is explained in the next section (section 2.2).

2.2 Carbon chemical erosion

2.2.1 Chemical erosion cycle

The chemical erosion of carbon by atomic hydrogen was first described by Horn et al. [18] and by Küppers [19], Roth et al. [14] and Mech et al. [20]. A simplified model of the chemical erosion cycle is shown in figure 2.1.

Hydrogenation of sp2 _{carbon surface atoms results in sp}3 _hydrocarbon

complexes where CH3 radicals are formed at the surface (the left hand

side of figure 2.1). After desorbing H2 and if the surface temperature is

high enough (> 100◦_{C), the CH}

3 is released and the carbon surface atoms

return to an sp2 _{configuration (the right hand side of figure}_2.1_).

A further increase of the surface temperature leads to additional pro-cesses. At temperatures > 300 ◦_{C the first hydrogenation step can be}

reversed as the incoming hydrogen atoms may recombine with adsorbed atoms. This reduces the sp3 _{concentration and therefore reduces the}

chem-ical erosion. It is also possible to desorb the CH3 radical (and go from an

sp3 _{back to an sp}2 _{configuration), but only when the temperature is above}

(38)

CH₃ T > 400 °C T_{> 300 °C} T> 100 °C erosion hydrog enat ion dehydrogenation sp2 H H spx H H H sp3H H CH₃ H H₂ spxH CH₃ H

Figure 2.1: A simplified model of the chemical erosion cycle of carbon, described by Horn et al. [18], Küppers [19], Roth et al. [14] and Mech et al. [20].

The three main parameters that determine the chemical erosion are the surface temperature, the ion energy, and the ion flux:

Surface temperature:

it has been established in literature that the chemical erosion yield increases with temperature to a maximum at Tmax, and then

de-creases at higher temperatures [3, 14, 18, 20, 21, 22, 23, 24]. This maximum is typically 300 < Tmax < 700 ◦C and depends on the

plasma conditions (i.e. the ion energy and the ion flux). Ion energy:

(39)

yield increases for increasing ion energy and Tmax shifts to higher

temperatures [3,20,22,23].

Ion flux:

the ion flux dependence is presented in chapter 1. This flux depen-dence – an empirical result based on data of several different machines – shows an approximately constant chemical erosion yield for fluxes <1021 _m−2 _s−1 _{and a decrease at higher fluxes [2].}

Experimentally it may prove to be difficult to distinguish the different dependencies, as the three parameters are usually correlated.

2.2.2 Erosion by hydrogen atoms or ions and the formation

of higher hydrocarbons

The chemical erosion cycle described in section 2.2.1 is based on erosion by atomic hydrogen, whereas most experiments involve hydrogen ions. Vietzke et al. [21,25,26] have shown that with (high energy) hydrogen ion impact, CH4 is released instead of CH3. It is suggested that when carbon

is exposed to high energy hydrogen ions (energies ranging from hundreds of eV’s to keV’s), still CH3 is formed, but deeper in the bulk. When this

CH3 diffuses towards the surface it is likely that it will recombine with a

free hydrogen atom in the material and then leave the surface as CH4 [21].

Higher hydrocarbons CxHy (with x ≥ 2) are also formed when carbon is exposed to either hydrogen atoms or ions [11,21,22,26]. The contribution from these higher hydrocarbons to the total carbon erosion can be similar to that from CH4 [22,26,27].

In this thesis we only consider methane and its breakup products, since the experimental techniques used here were not sensitive enough to detect higher hydrocarbons. Results presented in chapter 7 show a correspon-dence between a spectroscopic measurement of erosion by CHy (see sec-tion 2.4) and the net amount of material lost, which indicates that the contribution of higher hydrocarbons is limited.

(40)

2.3 Hydrocarbon chemistry

The main reaction products of chemical erosion of carbon are CH4(or CH3)

or higher hydrocarbons. We will first consider the hydrocarbon chemistry when CH4 is released into the hydrogen plasma. All reaction pathways

given below are described in detail by Janev and Reiter [28].

There are several different mechanisms that cause a breakup of methane: • electron-impact direct (I) and dissociative (DI) ionization of CHy:

e+ CHy →CH+y + 2e (2.1)

→CH+_y−k+ (H, H2) + 2e (2.2)

→CHy−k+ (either H+ or H+2) + (H, H2) + 2e (2.3)

• electron-impact dissociative excitation (DE) of CHy to neutrals: e+ CHy → e+ CHy−k+ (H, H2) (2.4)

• electron-impact dissociative excitation (DE) of CH+

y ions:

e+ CH+_y → e+ CH+_y−k+ (H, H2) (2.5)

→ e+ CHy−k+ (either H+ or H+2) + (H, H2) (2.6)

• electron capture auto-ionization dissociation (CAD) of CH+

y ions: e+ CH+_y →CH∗∗_y → e+ CH+_y−k+ (H, H₂) (2.7) • electron-impact dissociative ionization (DI) of CH+

y ions:

e+ CH+_y →2e + CH+_y−k+ (either H+ or H+₂) + (H, H2) (2.8)

• electron dissociative recombination (DR) with CH+

y ions:

e+ CH+_y →CH_y−k+ (H, H₂) (2.9) • proton-impact charge and atom exchange (CX) reactions:

H+_{+ CH}

y →H + CH+y (2.10)

→H₂+ CH+

(41)

0 . 1 1 1 0 1 0 - 1 7 1 0 - 1 6 1 0 - 1 5 1 0 - 1 4 1 0 - 1 3 1 0 - 1 2 0 . 1 1 1 0 1 0- 1 7 1 0- 1 6 1 0- 1 5 1 0- 1 4 1 0- 1 3 1 0- 1 2 D I D I D E R at e co ef fi ci en t (m 3 /s ) E l e c t r o n o r i o n t e m p e r a t u r e ( e V ) C X C H 4 + H + / e : I o n s C H ₄ + H + / e : N e u t r a l s C A D I o n s1 1 N e u t r a l s1 1 D I D E D E E l e c t r o n t e m p e r a t u r e ( e V ) D R C H 4 + + e :

Figure 2.2: The destruction processes of CH4 and CH+4 via

capture-auto-ionization dissociation (CAD), charge exchange (CX), dissociative excita-tion (DE), dissociative ionizaexcita-tion (DI), or dissociative recombinaexcita-tion (DR), leading to (hydro-)carbon ions (dotted lines) or (hydro-)carbon neutrals (solid lines) [28]. The data is taken from HYDKIN [29].

All these processes result in the formation of smaller hydrocarbons CHy (with y ≤ 3) until C + H(2) remains. Spectroscopy on the molecular CH

A − X band makes it possible to quantify the CH4 particle flux, as will be

explained in section2.4and in chapters5and6. It is therefore important to know all the possible reaction pathways and the corresponding branching ratios that lead to the formation of CH.

The rate coefficients for each of the above given reactions depend on the electron temperature (or ion temperature for charge exchange). These reaction rates can be calculated using the online HYDKIN solver [29], a reaction kinetic solver for the catabolism of hydrocarbons in hydrogen plasma. Figure 2.2 shows the rate coefficients of the different breakup mechanisms of CH4 (left hand side) and CH+4 (right hand side). The

temperature dependencies of the different mechanisms for smaller hydro-carbons are similar (data not shown). The figure shows that at low plasma temperatures (Te . 2 eV) the chemistry simplifies as charge exchange of CH4 and dissociative recombination of CH+4 are dominant.

The breakup processes of higher hydrocarbons CxHy (with x ≥ 2) are similar to those for CH4 [30]. Again, charge exchange and dissociative

(42)

recombination are the dominant processes at low plasma temperatures. This was also observed in experiments on deposition of hydrocarbon layers using acetylene (C2H2) in ∼1 eV argon plasma [31,32,33].

2.3.1 Particle mean free paths

As mentioned in section 2.2.2 we do not consider higher hydrocarbons. The following calculations therefore refer to the formation of CH from the breakup of methane.

The first process when CH4 enters the plasma is charge exchange with

H+ _[28]: H+_{+ CH} 4 →H + CH+4 (2.12) →H₂+ CH+ 3 (2.13) The CH+ 4 or CH +

3 then dissociatively recombines with an electron to

produce CH [28]: e+ CH+₄ −→25% CH + H2+ H (2.14) e+ CH+₃ −→14% CH + H2 (2.15) (e + CH+ 2 25% −→CH + H) (2.16)

The dissociative recombination of CH+

2 (given between brackets) also

re-sults in the formation of CH, but since CH+

2 can only be formed via charge

exchange of H+_{with CH}

2, it is considered as a second or third order

reac-tion pathway, i.e. CH2 can be formed in the dissociative recombination of

CH+ 4 or CH

+ 3.

The rate coefficient for charge exchange of CH4 at 1 eV is 8·10−15m3/s

[29]. Together with a density of 1 · 1020 _m−3_{, this results in a lifetime of}

1.2 µs. The particle velocity is estimated to be around 500 m/s (thermal velocity), which yields a mean free path of 0.6 mm. The rate coefficients for the dissociative recombination at 1 eV is 6 · 10−15 _m3_{/s for CH}+

4, and

4 · 10−15 _m3_{/s for CH}+

3 [29]. This yields a mean free path of 0.9 mm and

1.3 mm, respectively. So this means that the typical length scale for CH formation from charge exchange of CH4 and dissociative recombination of

CH+

4 or CH +

(43)

proportional to the plasma density and increases slightly with increasing plasma temperature (since the rate coefficient of dissociative recombination decreases with Te).

2.4 CH spectroscopy

Optical emission spectroscopy on hydrocarbon species is the standard pro-cedure in tokamaks to determine the gross chemical erosion yield in situ [34]. Only CH, CH+ _{and C}

2 can be observed in the visible spectral range.

The CH A − X Gerö band is the main representative for CHy and the C2

Swan band for C2Hy [27,34,35].

The Photon Efficiency (PE), which relates the number of photons to the number of particles, is required to quantify the chemical erosion of carbon in fusion experiments. Often the inverse of this quantity (PE−1₎

is calculated as the ratio of the dissociation rate that leads to particle destruction and the electron excitation rate weighted with the branching ratio, the so-called inverse photon efficiency “D/XB”.

The PE−1 _{values for CH emission from CH}

4have been measured [6,27,

36], and modelled with HYDKIN [29]. HYDKIN uses the data from Janev and Reiter [28] (see section 2.3) to calculate the CH density for a given CH4 particle flux, assuming a constant zero-dimensional plasma. The CH

emission is then calculated from the direct excitation of CH:

e+ CH → CH∗+ e (2.17)

Dividing the CH4 particle flux by the CH photon flux then yields PE−1.

The electron temperature and density dependence of PE−1 _{taken from}

HYDKIN is shown in figure 2.3. PE−1 _{is typically increasing from 10 to}

80, for Te from 10 to 100 eV and ne > 1018 m−3. Modelling of PE−1 for temperatures below ∼3 eV becomes uncertain as the electron excitation rate of CH decreases orders of magnitude [35,37].

Fantz et al. [35] and Brezinsek et al. [27] have shown that C2Hy also leads to CH emission. The corresponding PE−1 _{for CH emission from}

C2Hy is similar to that from CH4. The contribution from C2Hy to CH emission can be monitored by measuring the C2 emission. This was

(44)

Electron density (m-3): 1 · 1018 Electron density (m-3_): _{2 · 10}18 Electron density (m-3): 4 · 1018 Electron density (m-3_): _{1 · 10}19 Electron density (m-3): 2 · 1019 Electron density (m-3_): _{4 · 10}19 Electron density (m-3_): _{1 · 10}20 0 . 1 1 1 0 1 0 0 1 0 0 1 0 1 1 0 2 1 0 3 1 0 4 1 0 5 1 0 6 In v er se p h o to n e ff ic ie n cy

Electron temperature (eV)

Figure 2.3: The temperature dependence of the inverse photon efficiency (PE−1_{) for CH emission from CH}

4 for a range of electron densities

cal-culated with HYDKIN [29]. At temperatures below ∼3 eV, modelling of PE−1 _{becomes uncertain due to steep gradients in the underlying reaction}

cross sections [35].

not sensitive enough. It is noted that this is a different spectrometer than the one used for measuring CH emission.

The CH A − X band spans a relatively large wavelength range: 415 to 445 nm. As this range also contains emission lines or bands from other species (e.g. emission of CII at 426.7 nm and the CH+ _{band at 423.7 nm)}

it is common to use the CH A − X band head from 430.0 to 431.5 nm to quantify the CH emission. This requires a scaling factor to determine the photon flux of the full CH band, which depends on the rotational temperature of the molecule. The scaling factor increases from ∼2.2 at a rotational temperature of 300 K, to ∼3.7 at 6000 K [35]. In this thesis we apply a factor of 2.8 corresponding to a rotational temperature of ∼3000 K.

(45)

2.5 Plasma flux to the target

The derivation of the plasma flux density to a solid surface is given in detail by Stangeby [17].

The ion flux density ΓHfor a plasma in absence of this solid surface is

given as: ΓH= 1 4nic= 1 4× ni× q 8kTi/πmi, (2.18) where ni is the ion density, c is the average (thermal) particle speed, k is Boltzmann’s constant, Ti is the ion temperature and mi is the ion mass.

When the plasma contacts a solid surface, the electrons, which are lighter and therefore faster than the ions, charge up the surface negatively. This repels the incoming electrons, but accelerates the incoming ions, until an equilibrium is reached (i.e. the electron losses to the wall are equal to the ion losses). The region in front of the surface where the plasma becomes positively charged (due to the decreased number of electrons) is called the “sheath”. The sheath is only a very short distance from the surface, since an electrostatic potential on the surface is shielded out by the ions and electrons. The ions are accelerated towards the surface over a distance called the “pre-sheath”, which is several times the length of the sheath.

The flow velocity of the plasma at the entrance of the plasma sheath equals the sound velocity, where the density drops a factor of 2 with respect to the pre-sheath. The ion flux density is then defined as:

ΓH= 0.5 × ne× q

k(Te+ γTi)/mi. (2.19) The factor γ is 1 for isothermal flow, 5/3 for adiabatic flow with isotropic pressure, and 3 for 1-dimensional adiabatic flow. For simplicity we assume here that γ = 1. Since we are not able to measure the ion density ni, but we do have a direct measurement of the electron density ne, we assume quasi-neutrality and therefore ni = ne. Furthermore we also assume that Ti= Te [38].

These assumptions may lead to a systematic error of the ion flux den-sity. However, a comparison with different methods to determine the total ion flux presented in chapter7show that the assumptions are valid within the accuracy of the measurement (∼30%).

(46)

Introduction to:

Chapter 3

PSI research in the ITER

divertor parameter range at

the FOM PSI-lab

The next chapter describes two of the main experimental devices of the FOM PSI-lab used to study plasma surface interaction (PSI) under ITER-relevant conditions: Magnum-PSI and Pilot-PSI. The largest linear plasma generator is Magnum-PSI, which is expected to become operational in 2010. All experiments described in this thesis were done with its forerun-ner: Pilot-PSI. This chapter presents initial experiments on exposing fine grain graphite samples.

It has to be noted that the conclusions given in this chapter on the limited size of the plasma beam in Pilot-PSI are incorrect due to findings presented in later chapters (especially chapter 7). Furthermore, the first erosion measurements presented here are not chemical erosion but carbon sublimation. The corresponding high erosion rates are therefore not rel-evant for ITER. In the end of this chapter we present first spectroscopic measurements, which do show chemical erosion.

(47)

(48)

Chapter 3

PSI research in the ITER

divertor parameter range at

the FOM PSI-lab

*

3.1 Introduction: overview of the FOM PSI-lab

Plasma surface interaction (PSI) in the divertor of ITER and future fusion reactors beyond is a critical research area in the development of fusion power. To address the physics of PSI in the extreme parameter regime that is expected for ITER (electron density ne up to 1021 m−3 and elec-tron temperature Te in the 1 − 5 eV range [1,39, 40, 41]), FOM, in col-laboration with its Trilateral Euregio Cluster (TEC) partners and as part of the EURATOM fusion programme, is building an integrated PSI lab-oratory [42]. Figure 3.1 gives an overview of the PSI-lab. Magnum-PSI is the high flux linear plasma generator with superconducting coils; it is combined with plasma diagnostics and in situ surface diagnostics (SFG = Sum-Frequency Generation), an ex situ, in-vacuo surface analysis station, and the existing Pilot-PSI plasma generator. Pilot-PSI is its forerunner, of

*

Published as: J. Westerhout, W.R. Koppers, W.A.J. Vijvers, R.S. Al, S. Brezinsek, S. Brons, H.J.N. van Eck, R. Engeln, B. de Groot, R. Koch, H.J. van der Meiden, M.P. Nuijten, V. Philipps, M.J. van de Pol, P.R. Prins, U. Samm, J. Scholten, D.C. Schram, B. Schweer, P.H.M. Smeets, D.G. Whyte, E. Zoethout, A.W. Kleyn, W.J. Goedheer, N.J. Lopes Cardozo, and G.J. van Rooij,Phys. Scr. T128 (2007) 18

(49)

Control units, incl. SFG Target station Analysis-station

Plasma diagnostics Pilot-PSI

Surface-PSI Thin Film PSI

XPS, STM, SEM station

Magnum-PSI

Figure 3.1: Schematic of PSI-lab. The gray elements are already opera-tional.

which results are presented below to demonstrate the unequivocally high steady state plasma fluxes that will be available in Magnum-PSI (compare for example with the overview of carbon erosion experiments presented in [43]). Thin Film PSI is a custom-built advanced coater, used for re-search aimed at the development of highly resilient extreme ultra-violet optics, as well as for the preparation of samples for erosion experiments in Pilot-PSI and Magnum-PSI. It is connected under ultra high vacuum with an X-ray photoelectron spectrometer (XPS), scanning electron microscope (SEM) and a scanning tunneling microscope (STM; under construction). The Surface-PSI experiment is an ultra high vacuum plasma-surface inter-action device to study elementary plasma surface interinter-action processes at the surface at very low flux. Samples from Magnum-PSI can be analyzed in the analysis station while vacuum is maintained. Transportation to the XPS station at a distance of 50 m from Magnum-PSI can be done under controlled atmosphere.

In this chapter we report on recent developments with respect to the linear plasma generators. The design and the status of construction of Magnum-PSI are described. Recent results obtained at Pilot-PSI are pre-sented that demonstrate ITER-relevant plasma conditions at the remote target. Finally, first exposures of carbon targets to these plasma conditions are presented.

(50)

heating chamber

source chamber target chamber to pumps target analysis chamber

Figure 3.2: Total overview of the Magnum-PSI experiment with target station and target manipulator. Shown are (from left to right) the source-, heating- and target chamber with pump ducts. Next to these, the pump-ing station for the third stage is shown. On the right hand side, the target station with target and target manipulator are visible. In the target anal-ysis station, the targets can be analyzed in detail with surface analanal-ysis equipment.

3.2 Magnum-PSI: design and status of

construc-tion

The Magnum-PSI device is in the stage of detailed design of which an impression is given in figure 3.2. The cascaded arc source used in Pilot-PSI [44] is being scaled up to produce the plasma in Magnum-Pilot-PSI. Scaling studies predict power efficiencies in excess of 10%. Like in Pilot-PSI, the plasma will be additionally heated by Ohmic dissipation (of current to the target or to a ring electrode in front of the target) and radio frequency heating. Pressure control is essential for efficient plasma transport to the target as well as ITER-relevant neutral densities at the target. Three stage differential pumping based on roots pumps compatible with the large influx of neutral hydrogen will maintain pressures of ∼1 Pa in the exposure chamber as is confirmed by modelling and experiments on Pilot-PSI. The

(51)

superconducting magnet has been predesigned and will have a bore of 1.3 m and a length of 2.5 m, with 2 × 8 diagnostic ports. It will be placed on rails so that it can be moved for access to the vacuum vessel. This vessel consists of three elements (the source-, heating-, and target chamber) that can be modified if necessary. The target chamber has been designed for optimal diagnostic access. Magnum-PSI will allow targets with a width of <10 cm and a length of < 60 cm. The sample manipulator allows tilting to grazing incidence, rotation and axial translation, and will have 100 kW cooling capacity. Targets are exchanged in the target analysis chamber, where also first surface analysis can be performed. The project produced first plasma in June 2009.

3.3 Pilot-PSI: record plasma parameters at the

target

Pilot-PSI consists of a 1.2 m long, 0.4 m diameter vacuum vessel (0.1−1 Pa background pressure) placed inside five coils that produce a pulsed axial magnetic field B up to 1.6 T (continuous at 0.2 T, 10 s pulse length at 1.6 T). The plasma source is a cascaded arc [45], which exhausts into the vessel along the magnetic field axis. It consists of three tungsten cathodes in a cathode chamber, a stack of 5 electrically insulated water cooled copper plates with a 4 mm hole that form a 30 mm length discharge channel, and a copper-tungsten nozzle that also serves as anode. The source is operated on hydrogen with a typical gas flow of 2.5 standard liter per minute (slm) = 1.1 · 1021_H

2/s and a discharge current of 100 to 200 A. The target is at

0.56 m distance from the nozzle of the source. Thomson scattering (TS) is employed at either 40 mm downstream the source nozzle to characterize the source performance, or at 17 mm in front of the target to determine the exposure conditions. The TS system collects light with an array of fibers to obtain radial profiles of ne and Te over an observational chord of 25 mm.

TS results near the source confirmed a large experimental window: ne in the range 5 · 1019₋_{4 · 10}21_m−3 _{and T}

e = 0.7 − 4 eV. At moderate source operation parameters (100 A discharge current, 2.5 slm flow), the peak electron density increases linearly with B from 2 · 1020 _m−3 _{at 0.4 T to}

Carbon chemical erosion in high flux and low temperature hydrogen plasma

Carbon chemical erosion in high flux and low temperature

hydrogen plasma

Carbon chemical erosion in

high flux and low temperature

hydrogen plasma

Proefschrift

Jeroen Westerhout

Carbon chemical erosion in

high flux and low temperature

Chemische erosie van koolstof

onder invloed van een

hoge flux van waterstofplasma

van een lage temperatuur

Samenvatting

Carbon chemical erosion in

high flux and low temperature

hydrogen plasma

Summary

Contents

Chapter 1

Introduction

The next generation fusion reactor: ITER

Materials choices

Carbon chemical erosion

The ion flux dependence of the chemical erosion of

carbon

This work

Outline of this thesis

Publications

Chapter 2

Background: chemical

erosion of carbon exposed to

a hydrogen plasma

2.1

Introduction

2.2

Carbon chemical erosion

2.3

Hydrocarbon chemistry

2.4

CH spectroscopy

2.5

Plasma flux to the target

Introduction to:

Chapter 3

PSI research in the ITER

divertor parameter range at

the FOM PSI-lab

Chapter 3

PSI research in the ITER

divertor parameter range at

the FOM PSI-lab

*

3.1

Introduction: overview of the FOM PSI-lab

Magnum-PSI

3.2

Magnum-PSI: design and status of

construc-tion

3.3

Pilot-PSI: record plasma parameters at the

target