Fig 4.3 Depth of implantation: a) schematic portraying the range of ion implantation; b) dependence of ion implantation range R p and standard deviation ∆ R p on energy of implantation of nitrogen ions into iron; c) Gaussian implantation profile for nitrogen ion implantation in iron; d) distribution of concentration of ions implanted in an amorphous body and distribution of defects caused by them; e) defect formation in substrate C by an incident ion (several thousand atom translocations in the lattice) and the formation of sputter cascades A and B (several hundred vacancies and several hundred interstitial atoms); - implantation profile; - defects profile; - implanted atoms diffusion; - sputtered atoms effect is insignificant and then there may occur an additional maximum of concentration of implanted ions in the vicinity of R max The curve which determines the distribution of implanted ions at different depths in the host material is known as the implantation profile Maximum concentration of implanted ions, especially those of light elements, is not at the surface of the host material but at a distance of some © 1999 by CRC Press LLC tenths of a micrometer from it, due to backscatter and the non-elastic character of interaction of the ions with electrons of the host material Implanted ions, colliding with atoms of the host material, causes their displacement which, in turn, causes the formation of radiation defects Since the energy of the ion (several dozen keV) is several thousand times greater than the energy of atom bonds in the lattice (in metals it is approximately 25 eV), one ion is capable of knocking even several thousand atoms out of their nodular positions In this way, the implanted ion generates along its path a strongly defected zone, known as a cascade, of several to several tens nanometers, which propagates laterally to the direction of the ion movement, due to the secondary interaction of atoms knocked out from their positions The number of defects exceeds the number of implanted ions by a factor of to and usually is so big that there is a defect saturation [14] Because the probability of defect formation depends on the cross-section of nuclear deceleration, the distribution profile of atom displacement of the implanted material (radiation defects) is similar to the profile of implantation, with the defect maximum, however, always occurring closer to the surface [5, 15] (Fig 4.3d) When the ion dose exceeds 10 14 ions per cm 2, separate disturbed zones superimpose and the enhanced density of point defects may cause the formation of amorphous zones, dislocations and microporosity, through the coagulation of point defects The boundary value of the dose necessary to amorphize the substrate decreases with a rise in the mass of ions and with a drop in substrate temperature The formation of microblisters may occur during implantation of metals by ions of inert gases [1] The implanted ions may be located at dislocation boundaries, take up substitution positions or form inclusions of a new phase [14] Many defects formed during the implantation suffer atrophy already at room temperature, while the distribution of the remaining ones is similar to that of an alloying additive The number and range of atom displacement in the implanted material depend strongly on the ion dose, i.e., on the number of implanted ions per unit surface [3, 6, 9] In the majority of cases of practical utilization of ion implantation, it is important to obtain in the surface layer of a concentration of the alloying additive from several to several tens per cent The corresponding doses should be contained within the range of 10 16 to 10 18 ions per cm Such doses are delivered within tens and hundreds of seconds [1] For small doses, the implantation profile corresponds well to the theoretical Gaussian distribution (see Fig 4.3c and Fig 4.4 - curve 1’) An increase of the ion dose causes an increase in the number and range of atom displacements which are connected with significant defects in the crystalline lattice Deepest penetrating are those ions, the incidence direction of which is in agreement with the direction of “empty channels” in the lattice (stemming from the spatial distribution of nodes) and with the crystallographic axis of the implanted material This is the tunneling effect, mentioned earlier Partially decelerated ions also migrate athermically © 1999 by CRC Press LLC Fig 4.5 Dependence of profile for nitrogen ions, implanted at 75 keV energy into pure iron on ion dose: - dose: 3·10 17 N + /cm 2; - dose: 6·10 17 N + /cm ; - dose: 1·1018 N+/cm2 (From Iwaki, M [20] With permission.) Fig 4.6 Dependence of profile and depth of implantation on atomic number of ions implanted into gold, and on ion energy (From Deicher, M et al [21] With permission from Elsevier Science.) Usually, deviations are observed at the surface in the direction of increased concentration, which is the result of ion etching of the implanted material, or under the surface of the material (for x > 2∆Rp), where ions may occur due to random tunneling (Fig 4.6) Besides the above, ion mixing of materials in the deceleration zone, as well as possible diffusion processes, causes deviations from the Gaussian distribution curve [14] The depth of penetration of ions into the solid is relatively small - it exceeds µm only exceptionally and drops rapidly with a rise of ion mass (Fig 4.6) For ions of the same elements it rises with a rise of the accelerating voltage (Fig 4.7a) and ion energy (Fig 4.7b) In order to obtain an implantation profile without a sharp maximum peak it is possible to raise the energy of the ions during the implantation process The final profile will be an algebraic sum of implantation profiles obtained at different energy levels (Fig 4.7b) Another method is to increase the ion dose [8] © 1999 by CRC Press LLC Since implantation involves the introduction of ions to the host material but basically without arise in volume, ion implantation is accompanied by the formation of compressive stresses and a local rise in the surface temperature of the implanted material The bombarding ion may bring about, in the cascade zone, local temperature of approximately 1000ϒC in a time shorter than 10-11 s Heating up of the implanted material depends predominantly on energy and dose of ions and is described by the density of power supplied At power density of 10 kW/m the surface of the material heats up to approximately 100ϒC, while at 100 kW/m2, it heats up to 350 to 500ϒC within several minutes At the maximum achievable power density of 6000 kW/m 2, the material may melt or even vaporize [8, 9, 24, 25] Usually, the implantation process is conducted in such a way as not to allow the temperature of the implanted material to exceed 200ϒC, thus minimizing or eliminating changes of properties and deformation of the implanted material [14] 4.3.2 Pulse ion beam implantation On account of the non-stationary character of interaction of the pulse ion beam with the solid, due to the sufficiently short ion pulse (ns, µs) of very high energy, the beam causes melting of the thin surface layer of the solid (which is not observed with the application of the continuous beam) and the introduction of a foreign component (beam ions) into the molten liquid In the case of metals and alloys, modification of surface properties takes place as the result of the coexistence of three processes [26]: – thermal processes: melting, recrystallization, rapid cooling (at rates of 107 to 1011 K/s); – stress processes, caused by the propagation of stresses formed by the shock wave (ablative or dilatation) connected with extremely rapid vaporization of a portion of the molten material; – physico-chemical processes, connected with the supply of a foreign component to the molten zone in the form of ions or material coating the substrate, which are sputtered out by the ions of the beam, combined with thermal processes; these are, naturally, alloying processes The relative participation of these processes depends on the parameters of ion beams, their mean energy and thermal properties of the material subjected to implantation From the scientific point of view, these processes have not been thoroughly researched to date 4.4 Ion beam implantation equipment To carry out ion beam implantation, special ion accelerators are used, called ion beam implanters [27] Ion implanters, in the most general sense, may be divided into two groups: – with continuous ion beam, traditionally called simply ion implanters, © 1999 by CRC Press LLC – with pulse ion beam, which are versions of high voltage ion diodes or ionotrons The first group has been used for ion implantation since the advent of implantation technology, while the second group is in existence since the 1980s and used only for laboratory purposes [9] 4.4.1 Continuous ion beam implanters The main component systems of the implanter are (Fig 4.8): the ion source, separator, focusing and accelerating system, deflecting system vacuum system and the working (implantation) chamber [28−35] Ion source This is the most significant functional element of the implanter, serving to produce the initially formed beam of positive ions of a given type It comprises a discharge chamber where ionization takes place of gas, vapour or gaseous compounds of solids, and an extractor, used for extracting ions from the ionization zone, their initial formation into a beam and directing it to the focusing-accelerating system Properties of the source are determined by technological possibilities and the effectiveness of the implanter [28−35] Depending on the mechanism of ionization of the ion-forming substance, ion sources are divided into three groups [33−37]: With discharge in the gas phase, often referred to as plasma implanters: – with extraction of ions from the discharge plasma: spark (very seldom used in implanters), with capillary arc discharge (also seldom used); – with low voltage arc discharge in magnetic field or without a magnetic field; the best known and used in implanters are the Bernas, Nielsen and Freeman sources The Bernas source operates with an arc discharge at several tens to several hundred volts and a current of several tens amperes in a gas under pressure of approximately 1.33 Pa It operates in a magnetic field generated by an electromagnet with controlled induction up to several hundred Gauss, perpendicular to the direction of ion extraction The field serves to limit the escape of ions from the central portion of the discharge chamber and to enhance the effectiveness of ionization by extending the path of moving electrons (Fig 4.9a) Discharge in the source takes place between the hot cathode and the anode (anti-cathode) In order to avoid condensation of the ionized material, the discharge chamber may be heated by a resistance heater and screened by foil, e.g., tungsten or molybdenum At the same time it must be intensively water cooled The ionized material can be a pure element or a chemical compound It is supplied to the discharge chamber in the form of vapour under appropriate pressure in quantities ranging from fractions of a milligram to several hundred milligrams, depending on the dose of implanted ions and on the type of chemical substance used for the formation of ions of the given element Vapours of the implanted elements or chemical compounds are formed in separate heating chambers and supplied in a controlled manner to the discharge cham- © 1999 by CRC Press LLC the elements which, in turn, allows high energy of ions without the need to raise the extraction voltage [28−35] In the Nielsen source (Fig 4.9b) the cathode, powered by direct current in the form of a coil, supplies electrons to the discharge zone and the anode is a graphite cylinder The source serves to produce ions of both gaseous and solid elements The above-described ion sources belong to the most versatile and allow the production of the majority of ions within the range of atomic masses from to 240 Their characteristic feature is the hot cathode and the heater which enables vaporization of solid materials Such sources are used mostly in laboratory implantators [30] In industrial implantators there are usually applied specialized sources of ions of one element, often with a cold cathode which, although featuring long service life, allow the obtaining of relatively small currents of the ion beam and can operate with gases only [31] An example of an ionization source, especially for the generation of strong fluxes of nitrogen ions, is the J.H Freeman slot source (Fig 4.9c) in which the cathode, made of tungsten wire, is located very close to the extraction opening in the shape of a slot This, in conjunction with the strong cathode glow current (up to approximately 100 A) and the effect of the magnetic field, perpendicular to the direction of ion extraction, allows the obtaining of maximum plasma density opposite the extraction slot [31]; – with electron oscillation (so-called F.M Penning sources), also referred to as sources with cold or hot cathode The best known are G Sidenius sources and their different modifications Fig 4.10 shows the principle of operation of such a source The ion emitter here is plasma of low pressure discharge, ignited in the electrode system, in which the cathode has the shape of a hollow cylinder (either a wire coil [Fig 4.10b] or solid [Fig 4.10c]), forming a niche The glow of the hot hollow cathode is from direct current and the cathode is heated to a temperature at which a high electron emission current is obtained (Fig 4.10b) When appropriate pressure conditions are met, arc discharge is activated in the discharge chamber and plasma is formed, screened from the hot cathode by a bipolar layer of spatial electrical charge Practically the entire used interelectrode voltage is situated in this layer, and in this zone electrons, emitted by the cathode, achieve sufficient energy for the ionization of the gas In the discharge chamber there may or may not exist a magnetic field In the second case, the magnetic field created by the flow of current through the cathode may be compensated by an external field In the absence of a magnetic field, electrons move perpendicularly to the axis of the source and after passing through plasma they are decelerated by the electric field in the layer situated opposite the spot where they are emitted by the cathode The electrons are slowed down, their direction is reversed and they are accelerated in the direction of the discharge plasma; next, decelerated at the cathode, and thus the cycle repeats itself Electrons oscillate inside the hollow cathode Maximum ionization occurs near the © 1999 by CRC Press LLC – very high frequency (seldom used in implanters); – duoplasmotrons, i.e., sources with extraction of ions from plasma which diffused from the discharge chamber to the zone of beam formation Rather seldom used in implanters; characterized by high values of ion currents Thermoemissive - with thermal emission of ions from the surfaces of solids (Fig 4.11) This source operates on the principle of utilization of ionization of atoms colliding with the surfaces of metals The hollow tungsten cylinder with the extraction orifice of 0.2 mm diameter, together with rhenium foil is heated to approximately 2500 to 3000°C by electrons emitted by glowing external cathodes Atoms of the substance designated for implantation pass from the vaporizer to the inside of the hollow cylinder where, after making contact with the surface of the rhenium foil they undergo thermoionization Such sources allow the obtaining of ions not only of single atom elements but also of molecules [31] Fig 4.11 Schematic of thermoemissive ion source: - tungsten cylinder; - rhenium foil; - cathode I; - cathode II; - evaporator; - extraction aperture (From M˙czka, D., et al [31] With permission.) Field - with surface ionization in which the difference between the work of exit of electrons from the metal and the ionization potential of elements for ionization is utilized With bombardment by accelerated electrons (not used in implanters) Extraction systems These systems are built of one, two or three flat, cylindrical or conico-cylindrical electrodes, the first of which serves to extract ions and the remaining two to the formation of the beam (Fig 4.12) To the electrodes, voltages are applied of several tens kV Control of the extraction system involves either a change of value of the applied voltage, or change of location of electrodes relative to ion source [8, 9, 28, 30−37] In the case of single electrode extraction systems, the extraction and acceleration of ions occur with the aid of a conico-cylindrical electrode, placed near the extraction orifice with a negative potential relative to © 1999 by CRC Press LLC Fig 4.13 Principle of action of sector magnetic field on a divergent beam of monoenergy ions: m1, m - ion masses; Φ m - angle of divergence of magnetic field; U - potential accelerating ions; rm1, r m2 - curvature radii of ion paths with masses m1, and m 2; B vector of induction of magnetic field Mass separators Separators are used for precise selection of ions in the beam, based on an analysis of mass of the beam ions Separators allow through only ions with a given e/m (charge to mass) ratio, utilizing the effect of the homogenous magnetic field, limited by two planes (so-called magnetic lens) This field affects the beam of single energy ions but of different masses in two different ways: it focuses the divergent beam of ions of same masses but the sites of focusing depend on masses (Fig 4.13) in accordance with the equation [31]: (4.4) where: r m - radius of curvature of ion path; υ - ion velocity; B - magnetic induction Thanks to this, the beam passed through the separator is free from contaminations in the form of ions or particles of a mass other than desired The source of such contaminations may be incompletely purified gases or vapours, desorbed particles of gas from the ion source electrodes, atoms sputtered out of the electrode surface and ions from vapours of an ionized compound, supplying the implanted ions For example, if the substrate of Cr ions is CrCl, the Cl ions are separated and arrested in the separator, while only Cr ions are allowed to pass through, similarly to Na ions, obtained from the NaCO compound [22] Most often used separators are electromagnetic, as well as separators operating on the principle of crossed electrical and magnetic fields [16] Scanning systems To scan the treated material with an ion beam in order to ensure the required homogeneity of the implantation process, the following systems are used: System for deflecting the ion beam in the x-z axis (scanner): a) mechanical - in the form of a rotating and possibly laterally moving shield with openings, modulating the ion beam mechanically [39]; © 1999 by CRC Press LLC Fig 4.14 Schematic of scanning system (collector) of Polish-built UNIMAS-79 implanter: - ion beam; - probe; - implanted material (From M˙czka, D et al [31] With permission.) Fig 4.15 Simplified schematic of ion implanter (without separator and beam deflector) with fixed beam and movable stage (in z-x plane): - discharge chamber with ion source and extracting electrode; - three-electrode beam focusing system; - movable stage; - treated load; - work chamber b) electrical - in the form of two generators of saw-tooth shaped voltage, producing voltage with an amplitude dependent on maximum ion energy and on size of implanted surface, with a frequency of deflection voltage from to 100 kHz (Fig 4.14) System for mechanical feed in the x-z plane of the stage with the implanted material and a fixed ion beam (Fig 4.15) with possible masking off of beam System for mechanical deflection of the ion source together with the extraction and acceleration system, usually in one direction, with masking off of ion beam © 1999 by CRC Press LLC The last two systems usually find application in industrial implanters for the implantation of gases and metals [33-35] Vacuum systems Systems of vacuum pumps, valves and vacuum gauges serve to obtain vacuum in the zone of extraction and acceleration within the working chamber High vacuum is of special significance in systems with low working voltages, while soft vacuum in the single optical system may cause losses of beam current even up to 90% and deteriorate resolution [8, 9, 28, 30] Work chamber This chamber is used for placing of the treated load It should be equipped with mechanical systems for fixing and moving (in two directions) or rotating the load (of special importance for complex shapes) in such a way that the treated elements not mutually screen off the beam incidence line (Fig 4.16) The chamber is, moreover, equipped with loading/unloading systems which are usually cooled The cooling is either forced by a water jacket or water canals, or it may be natural, through radiators which give off heat Laboratory implanters are equipped with more systems than industrial implanters, e.g., heaters, goniometer, refrigerators, etc [40] Fig 4.16 Examples of providing movement to load by means of movable z - x stage (From Podgórski, A et al [38] With permission.) © 1999 by CRC Press LLC Table 4.1 Technical data relating to some implanters (Data from [8, 17, 33, 34, 42] and various other sources) Brand (manufacturer) Ion energy [keV] Beam current intensity [mA] Accelerating voltage [kV] Mass separator Load temperature [ C] Type of implanted ions Maximum zone of implantation [mm] Load mass [kg] PIMENTO (ARE, Harwell, U.K.) 10−100 100 NO - nitrogen 50−80 - First industrial implanter Working chamber 20 times smaller than the high current unit 60 10−25 - NO - nitrogen 1000 20 1500 Used for implantation of large size components Diameter of working chamber: 2.5 m; length: 2.5 m Pressure: 1.3 10 -5 Pa NO