At present a lot of companies over the world produce different kinds of ion sources. These are the duoplasmotrons, Kauffmann ion sources, Hall accelerators, Penning sources, saddle-field ion sources, radio frequency ion sources, electron resonance ion sources and others. Part of them is used mainly for research and special purposes. The Kauffman ion sources are more often used for the optical fabrication and thin film deposition. At present a number of companies offer radio frequency sources. Here we will not talk about these ion sources because you can find a lot of information on them without problems. I would like to discuss design and parameters of Hall accelerators which are one of the best in my opinion for technological applications. These sources have been developed in former Soviet Union in 60’s and at first were used as ion propulsion thrusters for satellites and other space systems. I think that Dr. Aleksey I.Morosov’s group had run the first research and development of these kind of thrusters in Kurchatov’s nuclear institute, later this work were continued in other enterprises and R&D institutes. In the end of 70’s – beginning of 80’s the first attempts to use these devices as a tool for optical surface treatment have been made. During the 80’s very good results have been collected which confirmed the advantages of these ion sources for the technological purposes. After the well known events in the former USSR a lot of works has been stopped however at present there are a lot of small companies and private experts who continue successfully work in this area.
The "Hall accelerator" is the Western name of this kind of ion sources. In Russia they are called as "The Plasma Accelerator with Close Drift of Electrons" (PACDE) and are divided by two classes: Plasma Accelerator with Extended Zone of Acceleration (PAEZA) and Accelerator with Anodic Layer (AAL). I like these names because they better describe the physical aspects of their design and work.
The physical and electrical configuration of this source is shown in Fig.1. The main element of its design is the dielectric accelerating channel consists of two coaxial cylinders with flange on the one side. The anode is placed inside the channel. It looks like hole metallic tore: the working gas is supplied inside through the gas supplying pipe and exits into the accelerating channel through a lot of holes with small diameters. So the anode is often called by too complicated name: anode-gas-distributor.
Fig.1
The accelerating channel with anode is placed in magnetic system consisting of metallic core and solenoid. The cathode-compensator and the ignition electrode are placed on the top of magnetic system near the output of the accelerating channel.
The flow of the accelerated low temperature gas plasma is creating in the accelerating channel by the way of the working gas atoms (Ar, Kr, Xe) or molecules (O2, N2) ionization.
The accelerator works as follows: After the anode voltage +UA and solenoid current IS have been supplied to anode and magnetic system accordingly, the area of crossed electrical field E and magnetic field H is created at the top area of accelerating channel . The vector of the electrical field is directed along the accelerating channel and the vector of the magnetic field is directed along the radius of cylinders.
Under the influence of the magnetic field the cloud of electrons emitted by the heated cathode is moving inside accelerating channel to the anode along the helical (spiral) trajectories. By other words, electrons drift in crossed electrical and magnetic fields diffusing in the same time to anode. On their way they collide with the atoms or molecules of the working gas and ionize them. The conductivity of accelerating channel for electrons strongly decreases and the electric field of high strength is created, increasing the kinetic energy of created ions. These ions are accelerated in electrical field and go out of accelerating channel like annular divergent flow.
The space charge of the ions in accelerating channel is compensated with the electron cloud drifting in channel. The electrons which are emitted from cathode compensate the space charge of the created ion beam. Part of these electrons compensate the lost of electrons in accelerating channel. So as a result we obtain the flow of more or less compensated plasma, i.e. the plasma beam.
The ignition electrode makes easy the ignition of discharge under the low anode voltage and low consumption of working gas.
The general parameters of this plasma source are as following:
| Anode voltage, V | +150 ... +600 |
| Discharge current, A | Depends on dimensions of source and gas consumption, usually 0.6 ... 1.0 A, but may reach up to a few amperes |
| Ionization of gas | Up to 90% |
| Ion current density, mA/cm2 | Up to 20, depends on dimensions of source and gas consumption |
| Gas consumption, cm3/s | 0.2 ... 2.0 and more depends on vacuum system |
| Ion energy, eV | 90 ... 400 |
| Working gases | All gases, may be limited by material and design of cathode |
| The half angle of beam divergence, grad | 10 ... 15 |
The scheme of this source is shown on Fig.2. It looks and works very like PAEZA but it does not has dielectric accelerating channel and the plasma is created in a thin layer of crossed electrical field of anode and magnetic field of magnetic system close to surface of anode. This source can work in two modes: without a cathode-compensator (so-called "ion mode") and with a cathode-compensator (so-called "plasma mode"). The source without cathode often is called "ion source with cold cathode". Let look how does it work.
Fig.2
Increasing the anode voltage we increase the electric field strength in narrow gap between the anode and nearest grounded metal surface (in our case it is a surface of magnetic core poles). If the electric field strength is big enough any casual electron get enough energy to ionize the atom or molecule of working gas by collision. In this collision the ion and second electron are created and process is repeated. The electrons move in crossed electric and magnetic fields along the helical (spiral) trajectories, that decrease the conductivity of gap and increase the probability of electron – atom collisions, i.e. concentration of created ions. In general, we obtain the independent glow discharge in crossed fields. The created ions are accelerated in strong electric field and leave system, so we get annular divergent flow of ions.
Typical parameters of this source are the following: the discharge ignition voltage is about +600 ... +800 V, the anode voltage range is usually up to +5000 ... +7000 V, the discharge current depends on the source dimensions and gas consumption and usually is in the range of 0,1 – 1 A, the ion current density is about dozen microamperes per square centimeter. The ion source works with any gases.
For sure, there are not enough electrons for full compensation of the ion beam space discharge. However the beam is usually partially compensated due to processes of charge exchange with molecules of gas in chamber, secondary electron emission from the chamber details and the target due to influence of energetic ion flow, and some other processes. The electrical potential of beam u is the index of its degree of compensation. In the fully compensated ion beams (plasma beams) u = +5 ... +10 V (PAEZA), in the ion sources with anodic layer and with cold cathode u may be about a several tenth volts and more.
If we will add the hot cathode or other electron emitter, the ion source start to work in plasma mode with the parameters which are close to parameters of PAEZA source described above, i.e. low anode voltage, high currents, low energy of the ions and well compensated beam.
Two PAEZA sources you can see in Fig.3. These sources I used to use for the ion beam milling and ion assisted deposition.
Fig.3
The AAL sourse is shown on Fig.4. This source has been used for the deep substrates cleaning and also for the ion assisted deposition.
Fig.4
These two kinds of ion sources are very simple and cheap in fabrication and may be used for solving of many technological problems. They may be installed as on the flange of vacuum chamber, as in the volume of chamber with possibilities of moving during the technological run. Well designed sources very often don’t need water cooling. PAEZA sources provide very good technological sputtering rates of dielectrics – up to a few nanometers per second and save a good quality of sputtered surface. These sources work very good together with the magnetron sputtering systems. In the last case you usually don’t need the cathode because the plasma of magnetron discharge is a good source of electrons. Therefore the ion source with anodic layer and with cold cathode works with magnetron in intensive plasma mode. The PAEZA sources for technological applications are fabricated usually with circular beam output with diameter from 35 mm to 70 mm. However the sources with anodic layer are fabricated both with circular and linear output. Last ones may reach a few meters in length. Both solenoid and permanent magnets may be used like a source of the magnetic field. You can use any electron emitters: heated cathode made from tungsten wire or LaB6 rod (last one is more expensive but works well in oxygen and needs two times less heating current), or gas discharge emitter of any design. The gas discharge emitter has biggest lifetime but it is more expensive device. Sometimes it is easier to use simple piece of tungsten wire and change it one time per week. But it is up to you...
These ion sources practically don’t need any maintenance or repair during the months of work and don’t need any tuning or alignment. You need only to clean them approximately each six months depend on your technology and perharps to change cathodes if you use heated ones. The maintenance works may be done by personal with no special skills and qualification, doesn’t need special tools and takes usually not more than couple of hours with full sources disassempling.
The experts often say that the lack of these sources is the wide spectra of the energy of ions. It is correct, but I don’t think that it is very important for a main applications in optical technologies, i.e. ion beam milling and ion assisted deposition. Beside these sources need some more gas flow than, for instance, Kauffmann ion sources with the same dimensions. The reason is that they have opened discharge chamber and gas is going out of the anode straight in the vacuum chamber. Let me discuss this problem in detail because it is very important.
The balance between the pumping speed and the leak rate determines the ultimate pressure in the vacuum chamber of the given volume. (Actually you must take into account and outgassing rates which determine the ultimate vacuum in the case of leakage and gas supply absence.) In regular vacuum plants for optical coating deposition without gas supply you can easy reach the ultimate pressure about 10-3... 10-4 Pa with the more or less well working pumping system. May be it will take 20 min, may be it will take 60 min, but early or later you will have pumped it down and start to deposit your coating. But when you install the ion source in your chamber, you can get the problem. The matter of fact is that from the point of view of vacuum technique the ion source is a big leak! The working pressure in your chamber will be determined now by balance between the gas flow of ion source and the pumping speed. But the gas consumption (not pressure in chamber!) determines all parameters of ion source. By the way it is the main difference between the ion sources and other plasma devices, magnetron, for example. The work of magnetron sputtering system depends on pressure in the vacuum chamber (usually 10-1 – 1 Pa), so a magnetron will not work in outer space, but ion sources will (remind ion propulsion thrusts). You have read in the Manual that your ion source provides the parameters you need with working pressure 10-2 – 3x10-2 Pa and gas consumption 0.5 – 1.0 cm3/s (and very often you won't find in the Manual any requirements for vacuum system). You think: ”OK, I’ll use it for the ion assisted deposition, my e-gun works well in this pressure.” And when you start up this ion source you see that it starts to work only when the pressure reaches 7x10-2 Pa, when you get a good discharge in e-gun. You measure the gas consumption – it is like in the Manual. It just means that your pumping speed is too low to provide good for you and for your e-gun pressure in the vacuum chamber. So be very careful! It may be when you try to install too big ion source or when your work with the old or self-made equipment. Usually you have no problem working with the modern vacuum equipment when the chamber of the volume about 0.5 ... 0.7 m3 is pumping down by the high vacuum pump with the pumping speed about 4000 – 8000 l/s.
Let me say a couple words about the power supplies. These are no problems usually with the solenoid power supply (20 ... 30 W, 0 ... 6 A DC), with the cathode power supply (about 100 W, 0 ... 50 A AC for tungsten cathode Ø0,6 ... 0,8 mm x 40 mm), and with the ignition power supply (you can use any pulse system with sufficient pulse voltage, including car ignition coil). The anode (discharge) power supply must provide DC voltage controlled in the range of +(50 V ... 600 V) and current up to 1 ... 2 A for PAEZA or DC voltage controlled in the range +(100 V ... 5000 V) and current up to 0,5 A for AAL. It is very good (but not necessary; we'll see it later) if the anode power supply provides stabilized voltage and current. But the main problem is that in any ion source an arc discharge may spring up sometimes. Usually it takes place between anode or its leads and nearest grounded parts due to bad or dirty insulation, or a gas leak. The big arc current destroys anode which is the most complicated and expensive part of an ion source. So your anode power supply has to be arc protected without fail. It may be different but relatively fast systems - from simplest electro-magnetic current relay to complicated electronic units. Anyway they have to drop anode voltage if the discharge current starts to increase dramatically and to stop the developing of the arc discharge.
At present you can find on a market a big number of power supplies providing well controlled stabilized voltages and currents in the wide power range with very good arc suppression. One problem is that the prices of these power supplies reach up to a few tens thousand dollars depending on their power. By the way a number of Russian companies produce at present the ion sources power supplies which have the same quality and parameters as power supplies produced by leading brand name companies but are a few times cheaper. You can find a wide range of Russian power supplies and other vacuum equipment (coating chambers, controllers, ion sources, magnetrons) at VECOR - Vacuum Equipment, Coatings, and Optics from Russia web site. In my practice I worked with a number of power supplies from simplest unstabilized single phase bridge rectifier with LC filter and current limiting relay to semiconductor frequency converters with electronic arc suppression and voltage - current stabilization. And I didn't see any difference in technological results. So you may use simplest power supplies especially if you just start to use this technology. Later you'll understand if you really need of perfect and expensive device. It depends on your goals, technologies, production rate and volume, employees qualification etc.
And at last I would like to say a little about gas control. We noted before that the main parameter of the ion source is not the pressure in a vacuum chamber but the gas consumption. Of course you can use any regular leak valve to control gas flow but it will be better to use the gas flow controller with digital indication of gas flow. You can find these devices in catalogues of practically all leading vacuum equipment manufacturers, including VECOR.