Remote Ion Source Plasma Electron Gun

Technical Field

The invention relates to large area electron guns and more particularly to a secondary electron emis¬ sion gun associated with a gas plasma.

Background Art

Cold cathode, secondary electron emission guns were first developed in the early 1970's for ionizing high power lasers. In French patent 72 38 368, D. Pigache describes an electron gun in which an ion source powered by filaments and magnetic fields emits an ion beam which bombards a cold cathode, emitting secondary electrons. These electrons then travel back through the ion source and exit into air through a thin metal window. The ion and electron paths are coaxial, but counter- flowing due to different polarities.

In U.S. Pat. No. 3,970,892 G. Wakalopulos de¬ scribes an electron gun in which a gas plasma is ionized in a manner permitting ions to be extracted from the plasma boundary to bombard a metal cathode from which the secondary electrons are emitted. The electrons flow counter to the ions and are allowed to escape through a window in a housing for the plasma and the secondary emitter.

In U.S. Pat. No. 4,025,818 R. Giguere et al. disclose a similar wide area electron gun except that the hollow cathode forming the secondary emission surface in the first mentioned patent is replaced by a wire, thereby allowing for a much more compact design.

In U.S. Pat. No. 4,642,522, Harvey et al. dis¬ close the addition of an auxiliary grid for better con¬ trol in switching an electron beam on and off.

In U.S. Pat. No. 4,645,978, Harvey et al. dis¬ close a radial design for an ion plasma electron gun.

The radial design is useful in switching large amounts of electric power.

In U.S. Pat. No. 4,694,222, Wakalopulos dis¬ closes an ion plasma electron gun which features grooves in the cathode to increase secondary electron yield.

The prior art relating to ion plasma electron guns may be summarized in a general way by observing that usually two adjacent chambers are employed in a single housing. These chambers are separated by a grid and are evacuated and backfilled with helium to a pressure of 10 to 30 millitorr. In one chamber, a plasma is established using a low voltage power supply. A high voltage nega¬ tive supply at 100 to 300 kilovolts is connected to a cold cathode in the second chamber. The negative field of the cold cathode attracts and accelerates ions from the boundary of the plasma. The accelerated ions bombard the cold cathode releasing 10 to 15 secondary electrons per ion. The electrons generally travel back through a grid separating the two chambers and through the plasma. A windpw is provided so that the electrons can escape the plasma chamber and exit into air. The ions and electrons are traveling in counter-flowing paths, with the electron distribution being directly proportional to the ion dis¬ tribution. The geometry of the plasma chamber, its cur- rent density, the gas and gas pressure determine the shape and distribution of the plasma. In turn, the shape of the plasma determines the general shape of the ion and electron beams.

The grid which separates the plasma chamber from the high voltage chamber must be transparent to the electron beam and is therefore typically 80 to 90% open in area. This transparency makes the operating pressure in both chambers nearly equal, which tends to cause high voltage breakdown or arcing in the high voltage region.

In order to achieve improved electron beam uni¬ formity and electron current densities required for com¬ mercial electron beam processing applications, i.e. 100 - 500 micro-amps per centimeter squared, the plasma chamber

has to be operated at high pressure, i.e. 1 - 30 milli- torr. This pressure causes the anode-cathode spacing in the high voltage chamber to decrease in order to minimize Paschen breakdown, i.e. arcing due to high gas pressure or large anode-cathode spacing. The reduced spacing re¬ quirements increase the electric field stress of the electrodes, causing a higher probability of vacuum break¬ down, i.e. arcing in the vacuum due to close electrode spacing. The arcing process is undesirable because it causes current surges in the power supply and results in operational down time.

An object of the present invention was to de¬ vise a large area electron gun which has a compact geome¬ try yet which was not subject to Paschen or vacuum break- down. Another object was to devise a large area electron gun which had better beam control and efficiency, reli¬ ability and operational range.

Disclosure of the Invention

The above objects have been achieved with the realization that in an ion plasma electron gun, the ion source could be removed from the path of the electrons so that deleterious counter-flowing streams of ions and electrons, which characterize the prior art, no longer exist. Instead, an ion source is isolated in an auxilia¬ ry housing removed from a main housing for the high vol¬ tage chamber, the two being separated by a narrow aper¬ ture. Now, a pressure differential may be maintained be¬ tween the two housings so that better efficiencies are achieved. The separation of the plasma region from the electron beam formation region allows both the plasma and the electron beam to be separately shaped and controlled for optimal density, pattern and uniformity. For exam¬ ple, magnetic fields could be used to confine the plasma in one housing, yet not affect the electron beam which might be controlled electrostatically in another housing.

A preferred design involves a main housing with a central high voltage chamber at low pressure and

peripheral or side plasma housings feeding energetic ions into the main housing by gas flow through a narrow aper¬ ture and toward an elongated metal target in the main housing. Now, an electron beam formed from secondary 5 electron emission from the target need not penetrate the plasma nor the ion extraction grid. This allows fine mesh grids to be used for ion beam shaping, turning and fo¬ cusing. The high energy electron beam will no longer destroy wire control grids since it is not coaxial with 10 the ion beam. Other advantages of the invention will be seen below.

Brief Description of the Drawings

Fig. 1 is a side plan view of a remote source - _j _ ₅ electron gun in accord with the present invention.

Fig. 2 is a detail of a spark plate ignition source for an ion chamber of Fig. 1.

Fig. 3 is a first embodiment of a secondary emission electrode structure used in the apparatus of

20 ^{Fig* λ} -

Fig. 4 is a second embodiment of a secondary emission electrode structure used in the apparatus of

Fig. 1.

Fig. 5 is a cross sectional view of an ion gun ₂ configuration taken along lines 5-5 in Fig. 5A.

Fig. 5A is an isometric view of an elongated ion gun of the present invention.

Best Mode for Carrying Out the Invention

30 With reference to Fig. 1, a main housing 12 has a gas impermeable wall 14, seen in cross section. The wall is cylindrical, having a length of several feet, but could be shorter and could be spherical or perhaps rec¬ tangular or an asymmetric shape. A high voltage elec¬ trode 16 penetrates wall 14 and is supported within insu¬

35 lating sheath 18 which itself is supported by support block 20. Wall 14 is grounded by means of electrical ground 15. High voltage electrode 16 is connected to the

high voltage power supply 22, capable of supplying several thousand volts for short intervals, but usually supplying a few hundred volts. Electrode 16 is connected to secondary electron emitter 24 using a cathode cable 5 connector 26. The emitter 24 is supported within a cathode shield 28 by means of metal blocks 30.

A vacuum pump 32 communicates with main housing 12 via connecting pipe 34. Vacuum pump 32 has the cap¬ ability of pumping main housing 12 down to less than 0.1 0 millitorr, which is a preferred condition. Pressure in the main chamber should not exceed 1.0 millitorr He.

A beam shield 36 is spaced from cathode shield 28 by ion entrance slits 38 and 40. Beam shield 36 has an opening distal to the secondary electron emitter 24 ₅ which is a cathode shadow grid 42. This grid is a wire mesh used for shaping an emergent electron beam which is shaped to flow ^' toward a thin foil, forming beam window 44. The thin foil maintains the vacuum within main housing 12, yet allows penetration of an electron beam. _Q Beam window 44 is held in place by foil backup grid 46. Outside of the main chamber, cylindrical auxiliary chambers 52 and 54 are adjacently disposed. Each of the auxiliary chambers is connected to the main chamber by means of a connecting passageway 56. The ₅ auxiliary chamber typically has the same longitudinal ex¬ tent as the main chamber. A gas supply 58 feeds the auxiliary chambers through a connecting pipe 60, opening into the auxiliary chamber. Helium is the preferred gas, introduced and maintained at a pressure in the range of 0 10-20 millitorr. Each chamber has an electrode 62 con¬ nected to a plasma power supply 64 capable of forming an ionized plasma from the gas delivered from gas supply 58. Typically, plasma power supply 64 consists of a current regulated positive polarity, regulated d.c. power source. The voltage needed to form a low temperature ionized 5 plasma is usually greater than 5 kV for plasma ignition with a total current of 10 to 50 milliamps per "linear inch of plasma. Once the plasma has been ignited,

voltage in the supply drops to several hundred volts. The operation of a low temperature plasma source is de¬ scribed in U.S. Pat. No. 3,156,842 to McClure. Briefly, if electrode 62 is formed into a thin wire, electrons are caused to orbit about the wire in long paths. The ener¬ getic electrons ionize the gas and maintain a discharge process. Positive ions are accelerated towards the walls of the auxiliary chambers 52 and 54 where they liberate secondary electrons. A control and focus power supply 66 maintains voltages on control electrodes 68 surrounding passageway 56. It is well known that cold cathode plasma discharge characteristics change with time. Oxide coatings and other insulating impurities greatly increase the secondary electron emission and this facilitates plasma ignition and maintenance. However, after a long operating time, the continuous ion bombardment generally removes all impurities from the inside of the plasma chamber walls. The result of such atomically clean surfaces is reduced electron emission and thus a higher current is necessary for plasma maintenance and higher starting voltages are required to ignite the plasma. Voltages as high as 20 V may be used or a hot filament electron source has been successful.

To overcome this problem without the use of a hot filament, a spark ignition system is used. A spark plug 51 is installed on the side or end of the plasma ion source. It is connected to the plasma power supply by a pulse generator 53, an automotive capacitor ignition circuit 55, and a spark coil 57. The spark plug is fired every time this plasma is switched on. This will facilitate ignition and make it independent of operation time. The ions and electrons produced by the spark easily ignite the plasma. The location of the spark source is important in plasma ignition. Generally, it is more efficient to locate the spark plug at an end of the plasma, near the termination of wire 62, where it can inject axial electrons into the plasma chamber.

To eliminate the sensitivity to spark location, a wide area spark source is used. These wide area plasma sources emit electrons over a wide linear dimension and thus help in uniform plasma ignition. The use of ceramics to facilitate surface discharges also aid in the generation of wide area electron sources. Many ignition techniques are possible due to the remote location of the plasma source. The absence of high energy electrons fa¬ cilitate the placement of insulators in the plasma region. Finally, the spark source can be pulsed con¬ tinuously from 100 to 300 Hertz to also help in main¬ taining the discharge. This mode of operation requires less plasma current since the spark source provides free electrons to keep the discharge going. The spark source may be either a spark plug, which is a point spark source, or may be a wide area spark source. In the situation where a spark plug is used, spark plug 51 is mounted near the termination of wire 62. The endwise injection of electrons encourages the formation of spiral electron orbits about wire 62. As the electrons traverse the wire in a helical path, coaxial with the wire, gas atoms in the chamber are ionized. The spark plug could be located elsewhere in the auxiliary chamber, but the formation of helical electron trajectories about the wire would be more difficult to establish.

In Fig. 2, a wide area spark source is shown which would be mounted along the length of the auxiliary chamber, parallel with wire 62. The extended spark source would be fed from a spark coil, similar to the spark plug source. A series of metal plates 61, spaced apart by insulative gap 69 would form a continuous first electrode at high potential fed by wire 65. A series of second spaced apart electrodes 63 would be maintained at ground potential by wire 67. The material of gap 69 may be alumina or similar ceramic material. The theory of operation is similar to a spark plug wherein a high vol¬ tage arcs across gap 69 from the high voltage plates to

ground potential. Electrons formed along the length of the wide area source migrate toward the high voltage wire and begin orbiting the wire after collisions with gas atoms between the outer wall of the chamber and the central wire.

Returning to Fig. 1, once a plasma is formed in the auxiliary chambers, ions are extracted from the peri¬ phery of the plasma by the electrodes 68 and travel through the passageway 56 into the main chamber. The ions are focused both by the electrodes and by the strong high voltage field in the main chamber. Ions are directed towards the cathode shield 28 which is main¬ tained at a high negative potential because of contact with secondary electron emitter 24. The ions pass through elongated ion entrance slits 38 and 40 because of alignment of the passageways 56 with the secondary elec¬ tron emitter 24. The emitter is typically molybdenum metal, but other materials could also be used. Once ions strike the secondary electron emitter, electrons are energetically released from the emitter surface and move towards cathode shadow grid 42 and thence toward beam window 44. Ion trajectories inside of the beam shield can be modified by allowing more or less electric field penetration through the cathode shadow grid 42. The secondary electron yield of molybdenum bom¬ barded by 200 kV helium ions is approximately 10 to 15 electrons per incident ion at 0° from normal. At 30°, this doubles and at 80 to 90°, it is a factor of 3 to 4. The efficiency is thus enhanced by bombarding the target at steep angles. This may be done in a manner discussed below with reference to Fig. 3. In accord with the present invention, the main ion beams from the auxiliary chambers are transverse to the electron beam formed from electrons emitted from the secondary emitter. In Fig. 1, there is an approximate right angle relationship between the ion beam coming from sides of the main chamber and the electron beam which is emitted downwardly from the main chamber. The secondary electrons leave the target

surfaσe with 10 to 50 volts of energy and then follow field lines inside of beam shield 36. It is important to adjust the distance from the secondary emitter 24 to the cathode shadow grid 42. This distance, along with the grid transparency and the geometry of the ion passageway determine the field inside of beam shield 36. The field must be stronger in the vicinity of the cathode shadow grid 42 to make the electrons travel in that direction. If the ion aperture field is stronger, the electrons will loop back to the ion source. Although, all electrons leaving the cathode surface initially travel in paths normal to the surface.

Electrons which leave the surface of the secon¬ dary electron emitter 24 are then accelerated towards the cathode shadow grid 42 where they attain their maximum speed. The cathode shadow grid 42 is aligned with the foil backup grid 46 in order to minimize electron inter¬ ception by the foil backup structure. The electron beam thus has a shadow of the cathode grid and exits into air outside of the main chamber through the thin beam window 44 without hitting the foil backup grid 46. The elec¬ trons are then directed to a deposition surface where they may induce chemical change, such as curing of poly¬ meric material or any other desired use. The electron beam may be made uniform across beam window 44 for wide processing applications, namely in the situation where main housing 12 is a cylinder.

With reference to Figs. 3 and 4, ion and elec¬ tron beam trajectories may be seen. In Fig. 2, ionized plasmas exist in auxiliary chambers 52. Ion beams are formed therein and pass through passageways into main housing 12 where electric fields guide the ion beams 72 towards secondary electron emitter 24 after the beams enter the aperture defined between the cathode shield 28 and the beam shield 36. In both Figs. 2 and 3 it is seen that the ion beam 72 is at approximate right angles to the electron beam 74. In Fig. 2, the ion beam is at less than a right angle to the electron beam, while in Fig. 4

it is at slightly more than a right angle. Usually, the ion beam is within plus or minus 30° to the plane of the secondary electron emitter 24, and preferably within plus or minus eight degrees. Actually, the secondary electron emitter need not be a plane, but may be segmented in a discontinuous manner, as explained below.

In Fig. 3, the ion beam emerging from the auxiliary chamber on the right controls the right portion of the electron beam 74 passing through the right side of the beam window 44. Similarly, the ion beam on the left controls the left portion of the electron beam 74. The distribution of ions within each ion beam can be matched or staggered so that at the secondary emitter the valley of one beam covers the peak of its neighbor and vice versa. This geometry allows for uniform electron beams covering a wide area.

Besides the angular variation of the ion beam, Fig. 4 illustrates that the secondary emitter may be formed by a plurality of spaced apart parallel ribs 76. In this manner, the top surface of the ribs is almost normal to the incident ion beams, thereby promoting higher secondary emission efficiency. Emitted electrons travel through the ribs toward cathode shadow grid 42 with a higher electron flux than in the embodiment of Fig. 3. Moreover, the location of the ion beam 72 above the plane of the ribs 76 has an advantage where access into the main housing 12 is difficult.

While electrostatic focusing was discussed for forming the ion and electron beams, one might substitute magnetic focusing electrodes for the electrostatic elec¬ trodes. In the event that main housing 12 is spherical, the auxiliary housing 52 may be made toroidal. Where the main housing 12 is cylindrical, auxiliary housings 52 are also cylindrical. Pressure in auxiliary housings 52 is always higher than in main housing 12 so that the pres¬ sure differential encourages ion flow from the auxiliary housing into the main housing. Even though the main

force on the beams is electrostatic or magnetic, the pressure differential also encourages beam formation.

Figs. 5 and 5A show an arrangement of auxiliary chambers 102 on one side of main chamber 114 and other auxiliary chambers 104 on the opposite side of the chamber. Auxiliary chambers 102 are offset from chambers 104 such that ion beams 106 overlap with ion beams 108. At the center of the main chamber 114 the overlapping beams form a generally uniform plasma. An advantage of the configuration of Fig. 5 is that a very long electron source may be constructed, without the need for long, continuous ion sources. Instead, a plurality of offset, relatively small size, ion sources may be disposed on each side of the central chamber 114. The width of each auxiliary source should be sufficient to produce a generally uniform plasma at the center of the main chamber 114.