Three‐dimensional Ion Distribution in a Filtered Vacuum Arc Discharge

Three-dimensional measurements of the ion flux along the filter of a magnetically filtered d-c vacuum arc are presented. The device includes a metallic plasma-generating chamber with cooper electrodes coupled to a substrate chamber through a quarter-torus magnetic filter. The filtering magnetic field was high enough to magnetize the electrons but not the ions. The ion current distribution was studied using a multi-element Cu probes, placed at three different positions along the filter. The ion saturation current of each probe was measured by biasing the probe at –70V with respect the grounded anode. Preliminary results of the three dimensional ion flux distribution and the floating potential of the plasma as functions of the bias filter voltage and magnetic field intensity are reported.


INTRODUCTION
high to magnetize the plasma electrons.
In a previous work, the magnetic filter was The cathodic vacuum arc is widely used to characterized along its axis, by using only a single probe [7]. In this work we present preliminary produce coatings [1]. The deposited material comes results of a three dimensional study of the ion from highly ionized plasma ejected from cathode spots. The kinetic energies of the ions are in the distribution and the plasma floating potential, measured as functions of the bias filter potential and range 15-120 eV, depending on the cathode material magnetic field intensity performed with a and on the charge-state of the ion [2], and with a total ion current amounting to 8-10 % of the total magnetically filtered . discharge current [3]. The presence of microdroplets of melted cathode material in the coatings is a EXPERIMENTAL SETUP disadvantage in vacuum arc technology, since for some applications this macroparticles increase the The investigations were carried out in a d-c filtered porosity and roughness of the coating. The most vacuum arc system, which is shown schematically in popular system used to separate the metallic plasma Fig.1 (DCF2 device at INFIP). There is a plasma from the microdroplets, are based on a focusing generation chamber that includes a water-cooled magnetic field that magnetize the electrons and guide copper cylindrical cathode (60 mm in diameter) the metallic plasma through the filter to the substrate surrounded by a floating shield, an annular waterbut leave unchanged the microdroplet flux [4].
cooler cooper anode (80 mm in diameter), and a However, part of the plasma flux is lost in the filter; tungsten striker which is brought into contact with so many efforts have been devoted to the the cathode surface and later removed to trigger the optimization of filters (by applying a positive bias discharge. The anode was grounded. At the exit of voltage of about 10-20 V to the filter with respect to this chamber is connected a magnetic quarter torus the plasma potential). Straight [5] and curved [6] filter (500 mm length, 100 mm inner diameter) made filters have been studied. Presently, the one most of corrugated stainless steel, including an external often employed in practice is the so-called "quarter coil that produces the filtering magnetic field. The torus filter" (developed by Aksenov). It consists in a bending angle of the torus is 90q. At the exit of the circular non-magnetic metallic tube with a bending torus a deposition stainless steel vacuum chamber angle of 90 o , and with a toroidal magnetic field (cross shaped) is connected. The plasma generation generated by an external coil. The magnetic field chamber, magnetic filter and deposition chamber are electrically isolated among them. By employing an independent d-c power source, the magnetic filter can be biased with respect to the plasma at a positive V BF potential. Two vacuum systems (composed of mechanical and diffusion pumps) pump separately the plasma generation and deposition chambers to a base pressure of less than 10 -4 mbar. The arc was operated in a continuous mode with an arc current of 100 A. The magnetic field generating coil was fed by an independent d-c variable current source, so that the filter magnetic field B F values (measured by a Hall probe) varied in the range 0-110 G. The maximum field corresponded to a coil current of 70 A, and was obtained at the knee of the torus.
The ion current distribution was studied using a multi-element probe, placed at three different positions along the filter, (at the entrance (A), in the middle knee (B) and at the exit (C)). Three individual spherical Cu probes (2 mm diameter), electrically insulated among them, were located on a PVC probe holder. The three dimensional measurements were done by rotating the holder and moving it along the filter.
In the Fig.2 a scheme of the probe sensed positions on a transversal section of the filter are indicated. It should be noted that at position B the lower points (6-7, 14-17) could not be sensed because of geometrical limitations of the multiprobe diagnostic.
The ion saturation current I ION of each probe was measured by biasing the probes at -70V with respect the grounded anode and its values were registered by measuring the induced voltage drop on a resistor connected in series with the biasing power source.

Saturation Ion Current
In the Fig. 3 I ION vs. B F with floating filter potential for positions A3, A7 and A9 (entrance) is presented. Within experimental uncertainties it can be seen that I ION is almost independent of the position and this result was found also for the other position investigated (not shown for clarity). Also it was found that I ION was independent of B F . (I ION a 2.5 mA)  In the Fig. 4    Typically, the current collected at the lower positions increased from 0.02 ± 0.02 mA (B F #0 G) to 0.18 ± 0.02 mA (B F #110 G).
By comparing Fig. 3 to 5, the ion loses in the filter are apparent. The efficiency of the filter, I ION (C)/ I ION (A), increases with B F from 0.8 % to 7% for a floating filter.
In Fig. 6 I ION as a function of V BF , for the maximum intensity of B F (B FMAX ) at the knee of the QT filter, is shown for the positions B3, B8 and B9. I ION is higher at the central position B3 and the beginning increases with V BF reaching a broad maximum for V BF a 15 V. In this case I ON increases from 0.4 ± 0.2 mA (V BF #0 V) to 2.2 ± 0.2 mA (V BF #15 V). Also I ION strongly decreases in the radial direction and there is no a defined behavior with V BF for the other positions shown in the figure. In Fig.7 I ION as a function of V BF , for B FMAX at the exit of the QT filter, is shown for the positions C3, C7 and C9. In can be seen a similar behavior of I ION as that shown in Fig. 5 (that is the ion flux concentrated at the upper positions of the filter), but with absolute values increased due to the bias effect. For example, I ION increases at the upper probe positions from 0.2 ± 0.2 mA (V BF #0 V) to 0.8 ± 0.2 mA (V BF #20 V).
By comparing the intensity of I ION , the ion loses in the filter are again, apparent: the efficiency of the filter, I ION (C)/ I ION (A), increases with V BF from 7 % to 14 % for B FMAX .

Floating Potentials
In Fig. 8 V FF   The dependence of the V FP with the filter magnetic field B F was studied along the filter, using the probes showed in Fig. 2. For B F =0 G and nonbiased filter, V FP decreased along the filter from -6 ± 4 Vat position A to -14 ± 4 V at position B and C , but the local value of V FP increases at the outer points influenced by the filter potential at position C.
The higher values at the entrance seems to be influenced by the anode potential (V anode a 0 V) and in a weakly way by V FF = -10 ± 2 V.
For B FMAX (V FF = 1 ± 1 V) , V FP takes higher values with respect to the previous situation, reaching -8 ± 4 Vat the entrance, to -12 ± 4 V at the QT exit.
On the other hand, when the QT filter is biased, there is no considerably change in the V FP values at position A (V FP = -8 ± 4 V) for different values of V BF . At position B and C also no significantly changes in V FP with V BF were found. The only point to note is that at the filter exit, the V FP values at the upper position of the filter (V FP =-10 ± 4 V) were somewhat higher that those found at the lower position (V FP =-4 ± 4 V).

CONCLUSIONS
The measurements of I ION indicate that there is an electron confinement due to B F , and the filter efficiency increase with its value. The biasing of the filter improves the efficiency of the system, reaching 14 % for the optimized parameters of DCF2.
The obtained data for the ion current and plasma floating potential will be coupled in the near future to a hydrodynamic model for the filter.