Dec 1, 2011 - of the two plasmas and the ion beam current dependence on these potentials are .... ion current density at the IOS inlet was kept constant.
Eur. Phys. J. D 65, 475–479 (2011)
DOI: 10.1140/epjd/e2011-20402-y
Influence of ion-beam plasma on ion extraction efficiency in a single-grid ion source S.V. Dudin and D.V. Rafalskyi
Eur. Phys. J. D 65, 475–479 (2011) DOI: 10.1140/epjd/e2011-20402-y
THE EUROPEAN PHYSICAL JOURNAL D
Regular Article
Influence of ion-beam plasma on ion extraction efficiency in a single-grid ion source S.V. Dudin1 and D.V. Rafalskyi2,a 1 2
Department of Physics and Technology, V.N. Karazin Kharkiv National University, Ukraine Scientific Center of Physical Technologies of Ukrainian Academy of Science and Ministry of Education and Science of Ukraine, Kharkiv, Ukraine Received 6 July 2011 / Received in final form 23 September 2011 c EDP Sciences, Societ` Published online 1st December 2011 – a Italiana di Fisica, Springer-Verlag 2011 Abstract. The influence of ion-beam plasma on ion extraction efficiency in a single-grid ICP ion source is researched. The single-grid ion source is considered as a system with two plasmas, ion-beam plasma and the source plasma, separated by an extraction grid. Results of experimental measurements of the potentials of the two plasmas and the ion beam current dependence on these potentials are presented. It is shown that the ion extraction efficiency depends equally on both the acceleration potential and on the potential of the ion-beam plasma. The obtained results demonstrate the key role of the ion-beam plasma in the ion source operation, which is important in technological applications and space thrusters.
1 Introduction At the present time, broad-beam low-energy ion sources (IS) are widely used in different technologies particularly in rare gas and reactive ion beam processing, micromachining and nanostructure fabrication [1–3]. Usually, in such devices gas discharge plasma is used as a source of charged particles, while ion acceleration is provided by a multigrid ion-optical system (IOS) consisting of thin perforated electrodes (grids) spaced by several millimeters, with high voltages applied between the grids. Considering the significant transversal sizes of the IOS and the essential thermal load, the grids’ manufacturing, adjustment and the maintenance of geometry during the operation become rather challenging, so the IOS is one of the most demanding and expensive parts of the IS. Besides, using the three-grid IOS requires application of an additional high-voltage power supply (even at low ion energies) with the necessity of careful voltage adjustment. Moreover, some specific problems like insulator coating by conductive films and intergrid breakdowns are probable. At the same time, single-grid ion sources are known [4–7] to be free of the inherent drawbacks of multigrid sources. In addition, the single-grid source with RF grid biasing is capable of the simultaneous generation of coincident ion and electron flows [5,6] in contrast to more common two- or three-grid sources. Application of such sources is usually expedient in reactive ion-beam etching with low ion energy (10 keV) focused ion beams [8]. This research was mostly devoted to the investigation of the influence of IBP on the beam divergence. Another direction of IBP research was the study of the ion-beam plasma effect on the processed dielectric surface potential in technological systems with low energy broad ion beams [11]. However, in the gridded ion sources the IBP may also influence the ion extraction efficiency. Since the IBP potential is always positive, the accelerated ion is decelerated in the outlet of the grid aperture and its trajectory can be significantly changed. Usually this effect is neglected and the value of the IBP potential is rarely measured in experiments. In the present paper we will show that the ion-beam plasma has a strong influence on the ion extraction efficiency in single-grid ion source.
476
The European Physical Journal D
Fig. 1. Measured dependences of the target current Itgt on the acceleration potential ϕacc and the target potential ϕtgt (solid lines). The ion energy distribution function (IEDF) f (Ei ) (Ei is the ion energy) for the ϕacc = 100 V is shown by dotted line. This graph is a compilation of previous paper results [5] and recently measured data (see below in Chapt. 3). Dependences Itgt (ϕtgt ) and f (Ei ) were measured at same acceleration potential (ϕacc = 100 V); dependence Itgt (ϕacc ) was measured at zero target potential.
2 Experimental setup The described experiments were conducted using a singlegrid ion source with inductively coupled plasma (ICP). The scheme of experimental setup is shown in Figure 2. The discharge vessel of the ICP source is made of 40 mm diameter and 80 mm length quartz tube 1 and two metallic flanges. In the outlet of the source the grounded 0.18 mm thick metal grid 5 with holes of 0.3 mm diameter is placed. The active area of the grid is 3 cm2 and the optical transparency is 0.19. RF power (13.56 MHz) in the range 0–200 W is applied to the 4-turn inductor 2 through the matching box. The current density is measured by a planar probe 6 biased at –25 V. For measurement of the ICP and IBP potentials hot emissive probes 4, 7 located in the source and in the beam transport space are used. A hollow cylindrical electrode 11 allowing control of the ionbeam plasma potential is placed inside the beam transport space. A DC acceleration bias voltage ϕacc (0–200 V) is applied between the grounded grid 5 and the copper cylindrical electrode 3 with longitudinal cuts placed inside the ICP tube. All potentials here are measured versus the grounded beam transportation chamber. The target accepting the ion beam is made as a 50 mm diameter singlegrid retarding-field energy analyzer (EA) 8–10 placed at 50 mm distance from the source grid. The EA grid holes diameter is 0.1 mm, thickness 0.12 mm and transparency 0.2. There are possibilities to bias the grid 8 independently or to connect it to the EA shield 9 and collector 10 for measurement of the full target current. In all the described experiments, the argon pressure in the beam transportation chamber was 3 × 10−4 torr and in the ICP vessel was 5 × 10−3 torr. Such a pressure drop appeared due to pumping by a turbo pump with 700 L/s throughput.
Fig. 2. (Color online) Experimental setup.
3 Experimental results The main parameter of interest of this paper is the ion extraction efficiency that means here the ratio between the ion current densities at the IOS outlet and inlet. It should be emphasised that in all the described experiments the ion current density at the IOS inlet was kept constant (5 mA/cm2 ) by RF power adjustment. As we mentioned in the Introduction, the extracted ion current has a strong dependence on the acceleration potential and, moreover, this dependence is not monotonic (see Fig. 1). In order to explain this behaviour of the extracted current we have measured potentials of both ICP (ϕICP ) and IBP (ϕIBP ) versus the acceleration potential ϕacc (see Fig. 3a). It is clearly seen from the figure that when ϕICP is below some critical potential ϕcr , the potentials of both plasmas are closely connected, while at higher electrode potentials ϕIBP saturates. This means that the accelerating potential difference between the plasmas exists and the directed ion beam is formed only when ϕICP is significantly higher then ϕIBP . Otherwise, the ion velocities in the transport space are close to the Bohm velocity, and ions are accelerated only near to the target. The value of ϕcr can be estimated using Child’s Law [12] for the case when sheath thickness is equal to the grid
S.V. Dudin and D.V. Rafalskyi: Influence of ion-beam plasma on ion extraction efficiency in a single-grid ion source 477
Fig. 3. (a) Dependence of ICP (ϕICP ) and IBP (ϕIBP ) potentials and target current Itgt on the acceleration potential ϕacc . (b) Dependences of the target current Itgt on the target potential ϕtgt measured at different acceleration potentials ϕacc .
Fig. 4. Dependences of ion-beam plasma potential ϕIBP on: (a) acceleration potential ϕacc measured at different constant target potentials ϕtgt ; (b) potential of additional electrode ϕE in the beam transport space at different constant acceleration potentials ϕacc .
aperture radius RA : 3/2 4 2e0 φcr ε0 RA ≈ , 9 mi ji
(1)
where ε0 is the permittivity of free space, e0 is the electron charge, mi is the ion mass, and ji is the ion current density. The estimation (1) in our case gives a value of about 30 V for ϕcr , which is close to the experimental result (see Fig. 3a). Thus at the ICP potentials lower than the critical potential the plasma penetrates to the extracting holes (“plasma regime”). At higher ICP potentials the sheath thickness increase leads to the plasma expulsion out of the grid holes, and ions are accelerated through the grid apertures (“beam regime”). However, when ϕICP is not much higher than ϕcr the sheath boundary has a defocusing configuration and the majority of the accelerated ions are intercepted by the grid. With the acceleration potential increase the sheath boundary becomes nearly flat and the extracted ion current tends to saturation (Fig. 3a). Thus the potential difference between ICP and the grid is one of the main parameters controlling the extraction efficiency. As it was suggested in the introduction, another factor affecting the ion extraction efficiency is the IBP potential.
Figure 3b shows that at any acceleration potential the increase of target potential leads to significant target current reduction. However, these dependences don’t make clear the physical reason for the mentioned effect. The fact is that the IBP potential is perturbed by the target potential change, which can be seen from Figure 4a. In order to separate the influence of the IBP and target potentials on the ion extraction efficiency we introduced an additional electrode into the beam transportation chamber (see Fig. 2). It allows us to change the ion-beam plasma potential whilst keeping the target potential constant. The possibility of IBP potential control is demonstrated by the curves shown in Figure 4b. Using the described technique we have measured the dependence of the EA collector ion saturation current on the ion-beam plasma potential at different acceleration potentials, with the EA grid potential set to 0 V. It can be seen from Figure 5 that in all cases, the increase of the IBP potential leads to a fast decrease in the extracted ion current down to zero at potentials close to the ICP potential. Thus, the dependence of the target current on the target potential can be explained by the strong dependence of the extraction efficiency on the IBP potential.
478
The European Physical Journal D
Fig. 5. Dependences of the EA collector ion saturation current Icoll on the ion-beam plasma potential ϕIBP at different acceleration potentials ϕacc .
4 Discussion As it was shown above, the ion extraction efficiency in a single-grid ICP ion source is strongly affected both by the potentials of IBP (Fig. 5) and ICP (Fig. 3a). In order to compare the contributions of each factor the current dependencies shown in Figure 5 have been plotted versus the potential difference between ICP and IBP (see Fig. 6). It is clearly seen from the figure that all 7 dependences are in close coincidence. Therefore, we can conclude that in the described single-grid system, the efficiency of ion extraction is defined just by the potential difference between ICP and IBP. This conclusion is also confirmed by the comparison of the mentioned curves with the dependence of the target ion saturation current (Fig. 3a) replotted versus the potential difference ϕICP − ϕIBP (Fig. 6). Despite the fact that this dependence has been obtained at constant IBP potential, one can see the similar shape of the curve. Hence, ion extraction efficiency depends equally on both the ICP potential and on the potential of the ion-beam plasma in the beam transport space. The relationship between the potential difference and the extracted current may seem obvious. Indeed, in the “plasma regime” when both plasmas (ICP and IBP) are connected and have the same potential, the extracted current is defined by the potential difference between the plasma source and its external environment. However, in the beam regime the “environment potential” is screened by the grid. This statement is based on the experimentally proven fact that the ICP in the beam regime doesn’t “feel” the potential change in the IBP. The measured values of the ICP potential and ion current extracted from ICP to the plane probe placed near to the grid show no dependence on the IBP potential change in the whole range of parameters, while the ion current extracted from the source changes dramatically. Therefore, formation of the ion flow from ICP to the grid apertures in the beam regime is controlled mainly by the potential difference between the ICP and the grid that is natural for any ion source. From this point of view, the suppression of the ion current to the target is expected
Fig. 6. Dependences of the EA collector ion saturation current Icoll (solid lines) and of the target current Itgt (dashed line) on the potential difference between ICP and IBP ϕICP − ϕIBP .
at the target potentials corresponding to the ion energy. However, our experimental results demonstrate that the “environment potential” affects the extracted current despite its non-interference in the source volume bounded by the grid. In order to understand the physical sense of the discovered dependence of the ion extraction efficiency on the potential difference between the plasmas let us consider the ion motion between ICP and the target. The potential distribution along the system axis is qualitatively shown in Figure 2. The change of velocity of an ion created in ICP occurs mainly in the three regions: (i) in the space charge sheath between ICP and the extraction grid (Ei ≈ e0 ϕICP ); (ii) in the space charge sheath between the extraction grid and the ion-beam plasma (Ei ≈ e0 (ϕICP − ϕIBP )); (iii) in the space charge sheath between the ion-beam plasma and the target (Ei ≈ e0 ϕICP ). In Figure 7 the qualitative picture of ion motion between ICP and IBP is shown for cases of high and low potential difference ϕICP − ϕIBP at constant acceleration voltage. Due to the potential dip in extracting cells any ion moving off the axis is affected by the radial electric field. Therefore, at the IBP edge it has a noticeable energy of radial motion. The potential difference between the plasmas ϕICP − ϕIBP defines the full kinetic energy of the ion on the IBP edge. If at some point in the ion trajectory the full ion energy is equal to the radial motion energy then the ion will be reflected from IBP. The lower the ϕICP − ϕIBP value, the more ions will be reflected (see Fig. 7). Naturally, the question about ion motion in a real system is quite complicated and the proposed explanation of the discovered phenomenon is just an attempt at a qualitative description. Since the electric field distribution in the grid apertures is non-homogeneous, the accurate description of ion motion is impossible without 2D or 3D numerical simulations. The description is further complicated by the necessity of taking into account the ions created in IBP. However, numerical modeling of the described system is out of the scope of the present paper.
S.V. Dudin and D.V. Rafalskyi: Influence of ion-beam plasma on ion extraction efficiency in a single-grid ion source 479
that the extraction efficiency depends equally on both the acceleration potential and on the potential of the ion-beam plasma in the ion beam transport space. The obtained results demonstrate the key role of ionbeam plasma in the ion source operation that is important in technological and space thruster applications [9].
References Fig. 7. Ion motion between ICP and IBP for the cases of high (dashed lines) and low (solid lines) potential difference ϕICP − ϕIBP at constant acceleration voltage ϕICP = 100 V.
The obtained results allows us to give some practical recommendations. To achieve the maximum ion extraction efficiency of a single-grid system the potential of ion-beam plasma should be reduced as much as possible. For this purpose all conductive parts in the beam transport space should be grounded and insulators have to be removed or shielded. To achieve the highest efficiency of the singlegrid source the ion-beam plasma potential can be further reduced using additional electron sources in the ion transport space, even for conducting materials processing [13].
5 Summary In summary, in the present paper, the influence of ionbeam plasma on ion extraction efficiency in a single-grid ICP ion source is researched. Results of the experimental measurements of potentials of the two plasmas separated by an extraction grid and the dependence of the ion beam current on these potentials are presented. It is shown
1. I.G. Brown, The Physics and Technology of Ion Sources: Second, Revised and Extended Edition (Wiley-VCH, New York, 2004) 2. G. Rius, J. Llobet, M.J. Esplandiu, L. Sole, X. Borrise, F. Perez-Murano, Microelectron. Eng. 86, 892 (2009) 3. E. Chason, S.T. Picraux, J.M. Poate, J.O. Borland, M.I. Current, T. Diaz de la Rubia, D.J. Eaglesham, O.W. Holland, M.E. Law, C.W. Magee, J.W. Mayer, J. Melngailis, A.F. Tasch, Appl. Phys. Rev. 81, 6513 (1997) 4. H.R. Kaufman, J. Vac. Sci. Technol. A 4, 764 (1985). 5. S.V. Dudin, D.V. Rafalskyi, Europhys Lett. 88, 55002 (2009) 6. S.V. Dudin, D.V. Rafalskyi, A.V. Zykov, Rev. Sci. Instrum. 81, 083302 (2010) 7. D.V. Rafalskyi, S.V. Dudin, Appl. Phys. Lett. 97, 051501 (2010) 8. A.J.T. Holmes, Phys. Rev. A 19, 389 (1979) 9. C. Charles, J. Phys. D 42, 163001 (2009) 10. D.M. Goebel, I. Katz, Fundamentals of Electric Propulsion (Hoboken, USA, Wiley, 2008) 11. S.V. Dudin, A.V. Zykov, V.I. Farenik, Rev. Sci. Instrum. 65, 1451 (1994) 12. I. Langmuir, Phys. Rev. 2, 450 (1913) 13. S.V. Dudin, Proceedings of the 21st EPS Conference on Controlled Fusion and Plasma Physics (Montpellier, France, 1994), p. 477
