2.1. The preparation of tungsten samples

Tungsten samples, of Φ20 mm×1 mm, were cut from a plate of powder metallurgy tungsten (99.95% purity). The samples were annealed for 1 h at 1273 K to eliminate the internal stress, and were polished with the electrolytic polishing method in the solution of 2% sodium hydroxide.

2.2. Irradiation tests of plasma

The plasma irradiation were performed with an equipment of electron cyclotron resonance (ECR). Fig. 1 shows the schematic diagram and a discharge picture of the ECR equipment. In the vacuum chamber, samples were placed on an electrode to control the bombardment energy, with a negative bias (50–275 V). A permanent magnet was placed under the samples to constrain the plasma converging.

Fig. 1Full-screen View

Schematic diagram of the ECR equipment (a) and photo of the ECR plasma (b).

The samples were irradiated at fluxes of 0.5×10²¹, 1.0×10²¹ and 1.5×10²¹ m⁻²·s⁻¹, with changing Ar mass flow rate to helium (the Ar volume percentage in Ar/He mixture).

2.3. Measurements of samples

Thickness loss is used to scale erosion rate of the irradiated tungsten samples, which were covered by a small piece of tungsten during the plasma irradiation. A surface profiler (SURFCOM 480 A) was employed to measure the step height which was regarded as the thickness loss. Every step was measured for 5 times at different site. The sample morphology was characterized by scanning electron microscopy (SEM) and atomic force microscope (AFM).

3.1. Influences of the negative bias on the erosion of pure helium plasma.

A series of tungsten samples were irradiated by pure He plasma of 1.0×10²⁵ m⁻² in fluence and 1×10²¹ m⁻²·s⁻¹ in flux. The sample temperature was around 650 K. The samples were biased negatively at 50–275 V. As shown in Fig. 2, the thickness loss increased almost linearly with the irradiation energy which was controlled by the bias voltage.

Fig. 2Full-screen View

Thickness loss of tungsten sample irradiated by pure He plasma, as function of the negative bias..

Thickness loss was about 0.40 μmm at 275 V. However, the surface micro-morphology did not show an apparent variation before and after the pure He plasma irradiation (Fig. 3). This indicates that pure helium plasma at 50–275 eV can hardly make an observable damage to tungsten surface.

Fig. 3Full-screen View

SEM images of tungsten sample surfaces before (a) and after (b) irradiation by pure He plasma.

This is because that the energy transfer of helium ions to tungsten was low. It can be calculated by ^[22]:

where, U is the energy transfer from an incident particle to a target atom, E is the energy of incident particle, and M₁ and M₂ are weight of the incident particle and target atom, respectively. For He and W, M₁=4 and M₂ =183.84, we have U = 0.0834 E.

Thus the transferred energy for a helium ion is only 23 eV when the negative bias is 275 V. The binding energy of tungsten is 8.90 eV, which means that a He ion of less than 106.7 eV cannot sputter a tungsten atom. A probable reason for the slight thickness loss at 50 V might be due to the energy fluctuation effect, for which a few He and W atoms could obtain an energy much higher than the average value at 650 K. Additionally, at high fluxes, the mass of interstitial atoms of helium existing in the tungsten surface layer might be another possible influencing factor.

3.2. Influences of Ar in the plasma mixture

Inert gases were considered as cooling gases to add to diverters of a fusion reactor, and argon has been added into the plasma. Polished tungsten samples were irradiated by plasma mixed with argon in atomic percentages of 0, 4.8%, 9.1%, 13.0%, 16.7%, 33.3%, 50.0%, 66.7%, 83.3%, 92.3% and 100%, at plasma flux of 1.0×10²¹ m⁻²·s⁻¹, fluence of 1.0×10²⁵/m², and bias of −200 V. The thickness losses are shown in Fig. 4. The results indicate that the Ar addition aggravated greatly the thickness loss, which increased sharply with the Ar percentage, when it was less than 20at.%, where the increasing rate of thickness loss began to lower down gradually.

Fig. 4.Full-screen View

Thickness loss vs fraction of Ar atoms in the mixed gas. (flux=1.0×10²¹/m²s, fluence=1.0×10²⁵/m², bias = −200 V)

The Ar addition effect can also be seen by SEM observations (Fig.5). In Fig.5(b), ripple-like damages can be seen on the surface irradiated by the plasma mixed with Ar at 13 at.%. For comparison, we show also morphology of the surface irradiated by plasma of pure He plasma (Fig.5a), and pure Ar plasma (Fig.5c). The latter made the most serious erosion on the surface, which was eroded to cobblestone-like morphology.

Fig. 5.Full-screen View

SEM images of tungsten surfaces irradiated by plasmas of (a) pure He, (b) He-Ar mixture (Ar, 13.0at.%) and (c) pure Ar .

AFM images taken near the steps on the surfaces irradiated by the plasma mixed with Ar at 92.3at% and 9.1at% are shown in Fig.6. The scale marks of Z axis were set as the symmetrical ranges from −1.9 μm depth to 1.9 μm height in Fig. 6(a), and from −1.2 μm depth to 1.2 μm height in Fig. 6(b). In each AFM image, the ridge-shaped topography can be seen along the step edge of the covered area, revealing that some sputtered particles could redeposited at the edge of the gap between the cover and the sample.

Fig.6.Full-screen View

AFM images of the surfaces irradiated by the plasma mixed with Ar at 92.3% (a) and 9.1%(b).

The energy transfer of Ar⁺ to tungsten can be estimated as follows. If the secondary ionization was negligible and the ratio of Ar⁺: He⁺ in plasma is assumed the same as the atomic ratio of Ar: He in the gas mixture, the thickness loss shall be in proportion to the composite transferred energy:

where, U_com is the composite transferred energy (transferred energy from composite to tungsten), which is a nominal variable; x is the fraction of Ar atoms in the mixed gas; and U_Ar and U_He are the transferred energy to tungsten atoms from an incident particle of Ar⁺ and He⁺, respectively. The mass of Ar atom is 39.95, so using Eqs. (1) and (2), we have:

The composite transferred energy increases linearly with the fraction of Ar atoms. For the incident energy of 200 eV, the relationship of composite transferred energy can be drawn as shown in Fig. 4. However, it does not fit the experiment result of the relationship of thickness loss with Ar fraction. This is due to the above assumption is too simple. Since the ionization energy of Ar is lower than He and the ratio of Ar⁺: He⁺ should not be the same as the atom ratio of Ar: He in the plasma. Further studies on influences of Ar to the thickness loss will be carried out.