GB2446005A

GB2446005A - Apparatus and method for removal of selected particles from a charged particle beam

Info

Publication number: GB2446005A
Application number: GB0701265A
Authority: GB
Inventors: Superion Limited
Original assignee: Individual
Current assignee: Individual
Priority date: 2007-01-23
Filing date: 2007-01-23
Publication date: 2008-07-30
Anticipated expiration: 2027-01-23
Also published as: GB0701265D0; GB2446005B

Abstract

An apparatus for removing neutral particles from a high energy charged beam 501 is provided. It deflects the beam magnetically in two or more deflection regions 551, 552 and all beams above a particular energy, determined by the geometry of the deflecting magnets, are returned to a required beam direction. In the range of energies for which neutral particle filtering is required, the un-deflected energetic neutral particle beam 505 is obstructed by an obstructing structure, e.g. beam stop 561, which can be removed 562 for the transmission of higher energy, un-decelerated ion beams, which also have the direction and position of the beam entering the apparatus.

Description

Aooaratus and Method relatina to reval of selected narticles from a

charred narticle beam The present invention relates to apparatus and methods for acting on charged particle beams. The invention relates in particular, but not exclusively, to charged particle beam acceleration and deceleration systems using magnetic and/or electrostatic fields to achieve decelerated charged particle beams which are not substantially coutaminfited with energetic neutral particles that have not been decelerated.

At the aid of this specification there is set out a list of references which will be zefeired to in this specification to assist understanding of the invention, the contents being incorporated herein by reference.

Charged particles are extracted from ion sources [1] and magnetically analysed to achieve mass separation [2] in order to produce a high punty, directed beam of charged particles (usually positive ions) which can be used to implant into various substrates, of which semiconductor wafers, solar cells and flat panel displays are important commercial examples.

Existing technology predominantly uses a system of analysis which will be referred to as conventional mass separator' optics [3].

This type of charged particle beam mass analysis will now be described briefly with reference to Figures Ia and lb of the present specification.

Figure la shows the dispersion plane (the plane in which there is dispersion of the charged particle beam into many directions, according to their (mass) x{energy} product and charge state, of a conventional mass separator. The charged particle beam can be extracted from the charged particle source as a circular beam, but where a high beam current is required it is usual to extract from a long slot (long in this context being typically a 10:1 aspect ratio or more).

The charged particle source 120 produces charged particles which are extracted from the charged particle source aperture 100 (circular or long slot) using electrically biased extraction electrodes 121 and 122 to fbrm an charged particle beam 101 (with an energy determined by the extraction voltage between the source and electrode 122) which typically diverges from the charged particle source extraction region. The electrode 121 typically extracts a positive ion beam and the second electrode 122 decelerates the beam sufficiently to provide an accelerating field for electrons in order to pievent the removal of electrons from the beam. These negatively charged thermal electrons are needed to neutralise the positive space charge of the ion beam.

The charged particle beam is then passed between the poles of an analysing magnet 150 as also a sideview ofthe beam in Fig Ib, the inthiscasebeinga ci ribbon beam 101. This magnet has two fimctions, one being to achieve mass dispersion and the other being to focus the beam so that mass analysis can be achieved at the resolving slit 160. It is necessaly to focus through a resolving slit so that slightly lower charged particle masses (deflected through a larger angle) or slightly higher masses (deflected through a smaller angle) are not transmitted.

Referring again to Figure la, the divergence in the dispersion plane of the beam at 103, which is normally small (typically a halfangle of 1-3 ), may be acceptable for charged paiiicie implantation if it is not than an optical element 170 in the dispersion plane can be used to create a parallel beam before arrival at the target 190. This optical element may also be used to accelerate or decelerate the charged particle beam to its required energy. The energy of charged particle beam passing through the analysing magnet is determined by a nwnber of considerations. A high extraction energy is desirable when high beam currents are required but this can lead to an expensive and large magnet This magnet is usually an electromagnet, and the power required and the size of the magnet will increase as the maximum charged particle (mass}x(energy} product requirement for the charged particle beam increases. A low extraction energy is favoured when the required charged particle beam energy is low, but this can lead to a decrease in the available beam current to unacceptable values. Typically the pre-analysis energy is set to a value around 30-50kV and then to accelerate the beam to the required energy after analysis. For low energy beams, a substantial deceleration is applied to the beam by the electrostatic field between electrodes 121 and 122, giving an analysis voltage typically in the range 5-3OkeV.

There are other methods of mass analysis (and some applications may not require analysis at all) but the above technique is the most widely used and will be used as the environment for the description of the present invention.

This invention addresses a problem that occurs when charged particle beams are decelerated. In Fig. Ia, the converging beam 102 leaving the analysing magnet 150 is directed towards the resolving slit 160 and the diverging beam 103 towards the target 190. Collisions with neutral gas atoms in this region of a typical commercial charged particle implanter used for semiconductor ion implantation, where the vacuum may be of the order of 10 ton, can cause neutralisaiion of a fraction of this charged particle beam. This might typically result in neutralisation m the range 0.1% to 1% of the beam. If the optical element 170 were to be a deceleration lens, then these energGtic neutral particles in the beam 103 would not be decelerated. The target will be bombarded by particles with two different energies -low energy decelerated charged particles and higher energy undecelerated neutrals in the beam 104. For some implants into silicon as part of the manufacture of semiconductor devices, a neutral content of 0.1% can be a serious problem, particularly when the deceleration is substantial (e.g. a reduction by a factor of 5 or more). Substantial reductions in energy are desirable because it allows high beam cun'ents to be produced as a result of a high extraction energy and the convergent focusing power of the deceleration lens increases with deceleration ratio (initial energy/final energy). This lens action is necessary to overcome space charge blow-up of the beam, caused by the mutual repulsion of the charged particles in the beam, which becomes a serious problem for decelemtion to low energies If a charged particle beani is to be efficiently scanned over a target, either by scanning the beam or mechanically scanning the target, or both, then it is important that the beam cross-section size should not be too large. Mother consideration is the range of implant angles produced by space charge expansion of the beam.

Fi&2 shows the conventional way of removing neutrals from the beam arriving at the targvt or arriving at some other optical process such as acceleration, deceleration or beam scanning In Fig2a, the beam 20 IA, represented for schematic simplicity as a single trajectory representing the general direction of propagation of the beam, is deflected by a pair of electrostatically biased electrodes 252, resulting in a deflected beam trajectory 203k The magnitude of the deflection is determined by the magnitude of the electric field between the electrodes (the polarity of this field and the charged particle charge polarity determining the deflection direction), the geometry of the electrodes (simple parallel plates shown in this Figure) and the energy and charge state of the beam. This deflection is not dependent on the mass of the beam. For reasons of schematic simplicity, the deflection is shown as a sharp change in direction at a plane 251. In reality, there is a curved deflection trajectory through the electric field region.

Neutral particles are not deflected, shown as trajectory 205k In Fig.2b, the beani trajectory 2OlB is deflected by a magnetic field between the poles of a deflection magnet pole gap between poles 254, the field direction being perpendicular to the deflection plane, the field polarity and charged particle charge polarity determining the deflection direction. The deflection is shown schematically as a sharp change of direction at plane 253 producing a deflected trajectoly 203B, the magnitude and direction of the deflection being determined by the strength and polanty of the magnetic field and the {mass}x{energy} product of the charged particles and the charged particle charge state. The neutral trajectoiy 2058 is not deflected.

Fig2c shows the general case of deflection by an electric or magnetic field, the unfilled square 255 representing the deflection region.

The most common use of a charged particle beam deflection as a neutral particle filter in prioraltis in ion hnplanters which usescRnningofthebean partofthe scanmng process required to give a unifonn dose over the surface of the target The neutrals must be removed from the scan region because the neutrals will not be scanned. Fig3a shows a beam trajectoty 301A being scanned by a variable electric or magnetic field at region 355A from trajectory 306 to 307, the beam scanning across the target 390 at a constant speed, assuming the currenttobeconstant Theneutralbeam3O5Aisnotdeflecte(Jandaregionofdusewjllbe produced on the target 390 by this neutral beam. A neutral filter is needed, and this can be of the type shown in Figic, situated before the scanner (i.e. closer to the charged particle source) so that the beam 301A in Fig3a is nominally free of neutrals. The problem with this is that neutrals can be generated between the filter and the scanner. A better approach is shown in Fig3b, where the filter and the scanner are combined into a single component 3558. There is now a constant deflecting field giving the filter deflection and a superimposed variable field giving the scanning from trajectory 308 to trajectozy 309. The neutral beam 305B does not reach the target 39!.

The scanning and deflecting systems in Fi&3 can be either electrostatic or magnetic.

Electrostatic deflection and scanning is favoured for low current (low space charge) charged particle beams where the space charge forces in the beam are small. Perveance [4] is a measure of the space charge forces in the beam and its value is dependent on beam current, beam ener and charged particle mass and charge state. Electrostatic fields remove thermal electrons (of the order of 1eV enerr) from the beam which neutralise the space charge of a positively charged charged particle beam (the normal charge state for commercial charged particle imp1ntation). If the space charge forces are significant; then the beam will blow up (increase in cross-section size) as it passes through the electric field. The advantages of electrostatic fields are low cost; simplicity and high scan fiequencies.

When space charge forces are large, due to high beam cunent and/or large charged particle mass, then magnetic deflection and scanning is favoured. The magnet usually needs to be an electromagnet because the deflection angle for the neuiral filter determines the geoinetiy of the system beamline, but the need for a range of beam energies and charged particle masses means that a variable magnetic field is required. Scanning magnets clearly must be variable and therefore must be electromagnets.

Fig.4 shows the application of neutral filters to the deceleration of charged particle beams. Fig4a shows the deceleration electrodes 471,472 and 473 placed before the deflection neutral filter 455A (charged particle source side). Three electrodes are necessary to produce acceleration followed by deceleration for positive ion beams. This ensures that electrons are not removed from the beam on either side of the deceleration gap. If deceleration is required, then the deceleration voltage is the dominant voltage. The initial beam trajectory 401A passes through the apertures in the deceleration electrodes 471,472 and 473 and then passes through the deflection system. This would seem to be the logical arrangement because the deflected beam with trajectory 403A will be completely free of undecelerated neutrals which are not deflected with trajectory 40546 The disadvantage of this approach to neutral particle filtering is that the deflection field is acting on the decelerated beam, and space charge may be a serious problem. In the case of electrostatic deflection, all the thennal electrons will be removed from the beam and this is only viable at low beam currents. Magnetic deflection is far superior for high perveance charged particle beams, but magnetic deflection runs into problems at veiy low beam energies due to magnetic field induced instabilities in the beam plasma, resulting in the loss of space charge neutralising thermal electrons from the beam.

Fig.4b shows the alternative arrangement where the deceleration is carried out after the filter. The charged particle beam trajectory 4018 passes through the filter field at 455B and the deflected beam 403B is subsequently decelerated by deceleration electrodes 474,475 and 476.

The advantage of this approach is that the deflecting field is applied to the high energy beam (i.e. before deceleration) and space charge related problems are significantly less severe, particularly for high deceleration ratios. The disadvantage is that some neutrals will be generated between the filter and the deceleration region, and in the deceleration region. It is therefore important that the filter and deceleration regions should be as close to each other as possible and the background gas pressure between these regions and in the deceleration region should be as low as possible.

The neutral filters described above all rely on a single deflection to separate the neutrals from the charged pailicle beam. This is generally satisfactoty for electrostatic deflection but is not so attractive for magnetic deflection. The main problem for magnetic deflection is that the deflection angle is built into the structure of the charged particle beamline and is there for the whole range of {mass)x{energy} product charged particle beams required. The filter is mainly required for decelerated low energy beams with vezy low (mass)x(energy} products but also has to deflect higher energy high mass charged particle beams, where the filter is an irrelevant component The filter magnet has to be variable, and therefore has to be an electromagnet The present invention addresses the need for a neutral filter that fimctions as required for low energy beams but does not have to deflect the high energy, high mass beams in the same manner. This is relevant to both electrostatic and magnetic deflection, but is particulaiy severe for magnetic deflection, where electromagnets can become large, power consuming and costly when they have to deflect the high energy, high mass charged particle beams that do not require the neutral filtering function. Electrostatic deflection is intrinsically variable, but with magnetic deflection there is an attractive alternative technique using fixed field permanent magnets. High performance rare earth magnets make very small high performance deflection mignets, but the filtering geometry must allow the use of a fixed field. The compact dimensions of a rare earth magnet filter, which can be installed in the vacuum system, are particularly important when the filter needs to be retrofitted into an existing implanter design.

According to the present invention in a first main aspect there is provided apparatus for removing uncharged energetic particles from a charged particle beam which has an initial general direction of propagation and an initial position compnsing at least a first and a second region of charged particle deflection for deflecting in a said first region the charged particle beam from its initial general direction of propagation to a second general direction of propagation followed by said second deflection region, a beam obstructing component for collecting uncharged energetic particles in the charged particle beam which are not deflected by the first charged particle deflection region the arrangement being such that in operation the charged particle beam is deflected by the first deflection region to avoid wholly or partly the beam obstructing component and is then deflected by the second deflection region to a required direction of propagation and/or to a required position.

For example, the required position could be a position through which the initial beam entering the apparatus would have passed in the absence of deflection and obstruction so that a target or subsequent deceleration region can be placed in this position. The geometiy of the range of charged particle parameter beams direct all the beams through a region through which the initial beam entering the apparatus would have passed in the absence of deflection and obstruction. Alternatively the beam leaving the apparatus can be parallel to but displaced from the input beam. Again the deflection region can be arranged so that the beam leaving the apparatus with a constant deflection field is parallel to the initial beam entering the apparatus for a range of charged particle parameters According to the present invention in a second main aspect there is provided apparatus, in which the first region of deflection is followed by at least two more regions of deflection, and in which the charged particle beam direction and position oni gthe apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction.

This allows the apparatus to be fitted to an existing beamline (retrofitted) because the output beam follows the same trajectory as the input beam.

According to the present invention in a third main aspect there is provided apparatus in which the final deflection to a required direction of propagation and/or to a required position is substantially maintained for a range of charged particle parameters comprising mass and/or energy and/or charge state, with or without changes to the strength of the deflecting fields.

These charactenstics with a constant deflecting field illustrates the advantage over a simple single deflection for decelerated ion beams. The deflection fields do not have to track with the energy of the beam. When the required beani energy is high and deceleration is not required, then it is not necessaiy to increase the deflection fields to cope with this higher energy, because these higher energy beams leaving the apparatus also pass through the same region through which the initial beam would have passed in the absence of deflection and obstruction.

A low energy decelerated input beam is deflected around the beam blocking structure, but high energy beams which do not require deceleration can pass through the apparatus with smaller deflections and with the blocking structure removed.

The characteristic of the output geomeny being maintained over a range of charged particle parametei and/or deflection field strengths is important because this apparatus can be used to focus the beam, particularly when the deflecting field is magnetic. Changing the energy of the input beam or the strength of the deflecting fields can be used to change the focusing characteristics without prejudicing the uncharged envigvtic particle blocking characteristics or the required position and/or direction of the output beam.

This apparatus can not only be used to obstruct uncharged energetic particles but can also be used to selectively transmit charged particles with particular parameters. The need for high energy charged particle beams to pass through the apparatus with deflections too small to avoid the beam obstructing component requires the obstruction clement to be movable. For the case of the removal of uncharged energetic particles, the obstruction element can be a simple plate that blocks the path of the undeflected beam, which is moved from the low deflection path of high energy beams when the removal of energetic uncharged particles is not requireci When particular parameter charged particle beams are to be selected for transmission through the apparatus, this component will be moveable, and will have an aperture through which certain trajectories can pass, and the position of the aperture is moved to allow transmission of the required parameter charged particle beam. The apparatus in this case is acting as an energy filter (strictly a {mass}x{energy} analyser).

The deflecting fields can be either magnetic or electrostatic or both. The electrostatic fields are easily vanable. Magnetic fields can either be produced by electromagnets, which are vanable (by changing the current through the coils), or by pennanent magnets, which are intrinsically constant field or both. The deflection field produced by permanent mfignets can be made variable by having a variable geometry magnetic cmuit e.g by mechanically varying the pole gap. All these deflection field options can have electncal or mechanical vanability in order to change the characteristics of the output beam e.g in order to change the focusing of the deflected beam.

Electrostatic fields create a space charge problem with low energy and/or high current ion beams. The electrostatic field removes electrons from the beam which are neutralising the positive space charge of the positive ion beam. When associated with deceleration, electrostatic deflection is only likely to be appropriate when the removal of energetic uncharged neutral particles from the beam takes place before deceleration, so that only the higher energy beam before deceleration is exposed to the electrostatic field. Electron repelling electrostatic fields may be needed at the entry region of the deflection system to limit the size of the beam region exposed to electron removing fields There is also a possible embodiment in which electrostatic deceleration and deflection are combined in a single deflecting and decelerating region at the last deflection region of the apparatus.

This description of the invention will now concentrate on the important example of the use of magnetic fields in the a neutral particle filter (the apparatus being referred to as a neutral particle filter for the rest of this description). Magnetic fields do not actively extract electrons from the beam pJsmi Low energy thermal electrons trapped in the beam by the positive space charge potential of the positive ion beam are necessary for the neutralisation of this positive space charge potential and their presence is essential for the trsnsmission of high perveance (low energy and high current) beams. Deflecting magnetic fields restrain the electrons but neither remove or contain them. Low energy electrons can escape from the beam along the magnetic field lines. At moderately low energies, high current beams can be deflected by magnetic fields without any deleterious effects, but at vely low energies plasma instabilities can occur which allow electrons to escape from the beam. For this reason, the unconventional placing of a neutral particle filter before the decelemtion stage has its merits. The magnetic field is applied to the high energy side of the deceleration gap, and magnetic field induced loss of space charge neutralisation will not be a problem. Unlike the conventional arrangement of deceleration followed by neutral particle filtering (post-deceleration filtering), pre-deceleration filtering will not substantially remove all the neutrals from the beam entering the deceleration region because neutrals can be funned in the region at the exit from the filter and in the region between the filter and the deceleration region. It also will not remove neutrals formed in the deceleration region.

But in the type of beamline shown in Fig.1, it will remove all the neutrals formed in regions 102 and 103 between the analysing magnet and the entrance to the filter placed just before the deceleration region at position 170. All the neutrals entering the filter are substantially removed but a small number of neutrals formed in the filter, between the filter and the deceleration region and in the deceleration region can reach the target. Nevertheless the undecelerated neutral content of the beam leaving the deceleration region is likely to be an order of magnitude lower than without the filter and this filter can be used at very low energies where the post-deceleration filter cannot be used. The word substantial is used above because certain processes such as beam scattering caused by collisions with gas molecules in the vacuum system may allow a very small proportion of neutral particles to be transmitted.

It is also possible to place the neutral particle filter between two deceleration regions.

The first gap will decelerate the beam 103 in Fig 1 from the mass analysis energy down to the minimum beam energy that can be exposed to a magnetic field without loss of space charge neuiralisation. If this energy is at or below the required energy then the near 100% efficient post-deceleration filter can be used, without the need to use the second deceleration region. For very low energy (high perveance) beams the second deceleration region will also be used. This arrangement would appear to make it necessary for the magnetic fields in the filter to be variable, because of the wide range of beam energies which may need to be transmitted through the filter. One of the advantages of the pre-deceleration filter is that the input beam energy (from the mass analysing magnet 150 in Fig!) can be fixed at a constant {mass}x{energy} product and constant field permanent magnets used in the filter. It is therefbre possible to use the double deceleration system with a permanent magnet filter by using the first deceleration region to lower the input energy from the magnetic analysis energy down to the preset requirements of the fixed field filter and then flurther decelerating down to the final required energy. This does mean that there is a range of beam energy down to the (fixed) input requirements of the permanent msgnet filter (for total removal of neutrals from the input beam) where the beam has to be used without deceleration, unless deceleration followed by acceleration is employed. The firnction of the first deceleration region with a permanent magnet filter is not necessarily to decelerate the beam down to a particular low energy. It is also there to provide a range of input energies into the filter so that the focusing properties of the filter in a plane at iight angles to the deflection plane (both planes containing the initial beam central axial trajectory) can be varied without the need for a variable filter field. The reason for the importance of the permanent magnet filter having a constant output direction or position (or both) characteristic for a range of input energies is because of the use of the filter as a focusing element, where the focusing action is varied by changing the input energy.

The action on the beam considered so fur as been dominated by deceleration. The filter can also be used in association with acceleration, when an neutral particle filter or energy filter is required after post-analysis acceleration to give a monoenergetic output beam. There are other possible applications as a straight through mass analysis system with the input beam to the apparatus coming directly from the beam extracted from the ion source and as a pre-ifiter for a straight through magnetic scanning system (i.e. with no neutral filter built into the magnetic -system).

According to the present invention in a fourth main aspect there is provided apparatus in which the defleclions are all in a single deflection plane and in which in at least one of the deflection regions there is magnetic deflection in the pole gap of a magnetic dipole which has no focusing action in the deflection plane.

The reason for the importance of this geometry has been described in a previous patent by this inventor [5] with regard toa technique formass analysis of wide beams using a quadrupole magnet with both magnetic deflection elements having parallel entry and exit pole edge geometry and with the input beam usually at right angles to this pole edge direction. This parallel edged, identical geometry but opposite polarity pole gapswith no focusing occurring in this plane (e.g. a parallel beam in gives a parallel beam out). These characteristics are maintained for a wide range of beam (mass)x{energy) products and a wide range of magnetic field strengtkc. The plane at right angles to the deflection plane containing the beam direction has powerful convergent focusing which is used for mass analysis by utilising the parameter dependent focusing of the beam in this previous invention. The fundamental difference between this invention and the previous invention [5] is that in this invention the obstruction of particles that are not to be transmitted takes place in the deflection plane, whereas in the previous [5] invention it takes place in a plane at right angles to the deflection plane.

This corresponds to the two deflection version of this invention with a output beam parallel to the input beam. lithe second parallel edge pole is increased in width or the field strength is increased with the original pole geometly, then the result is a beam leaving the apparatus which crosses the path of the undeflected and unobstructed initial beam, and beams with a range of charged paaiclc parameter values cross the initial beam trajectory at substantially the same position. Changing the field strength in the apparatus also results in a beam trajectories which pass through this same position. The output beam parallel to the initial input beam case is useful when, for example, the end station target scanning system is not prejudiced by a displacement of the beam by an amount a little larger than the width of the beam in the deflection plane as it passes through the neutral particle filter. Mternatively with the geometiy which gives an output beam crossing the path of the initial beam, a target, or a deceleration system, can be placed at the position where the range of charged particle parameter beams converge onto the lrajectoiy of the input beam passing through the neutral particle filter without deflection and obstruction.

According to the present invention in a fifth main aspect there is provided apparatus in which the defleclions are all in a single deflection plane and in which, in all of the deflection regions, there is magnetic deflection in the pole gap of a magnetic dipole which has no focusing action in the deflection plane.

When there are three deflection regions with alternate polarity, we have the ideal output geometry where the beam leaving the neutral particle filter has substantially the same output optics as the undeflected and unobstructed initial beam for a range of charged particle parameters or deflection field strengths. This allows the focusing action in the plane at aight angles to the deflection plane containing the beam direction to be adjusted without changing the output beam direction and position in the deflection plane.

An alternative similar geomeUy is to have apparatus in which there are four magnetic deflection elements having a parallel edge pole geometry (a non-focusing geometry), the deflections being equal in magnitude and the first and forth having one deflection direction and the second and third an opposite deflection direction, leading to a beam between the second and third magnetic regions with a geneml direction of propagation parallel to the initial general direction of propagation, and leading to an output beam in which the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that of the initial charged particle beam entering the apparatus and passing through without deflection and obstruction. The beam displacement between the second and third deflection regions relative the the initial beam will be referred to as the fllter displacement'. This geometiy is essentially two quadrupole deflection systems with opposing magnetic polarity.

According to the present invention in a seventh main aspect there is provided apparatus in which there is a beam crossover substantially at the region of maximum displacement from the charged particle beam position which the beam would have on passing through the apparatus without deflection and obstruction and a inoveable beam obstructing aperture plate placed with the aperture substantially at this crossover position arranged such that uncharged particles can be removed and a particular charged particle parameter beam substantially selected for trmnmiission through the apparatus.

In this aspect, the apparatus of the invention is acting as an energy filter (actually a {mass}x{energy) filter).

According to the present invention in a seventh main aspect there is provided apparatus in which there are at least two magnetic deflections in a deflection plane; deflection regions, in which an input or output general direction of propagation coincides with the initial general direction of propagation without deflection and obstruction, having a geometry which introduces convergent or divergent focusing into this deflection plane, intermediate regions having no focusing action in the deflection plane, arranged such that the charged particle beam general direction of propagation leaving the apparatus is substantially parallel to the initial charged particle beam general direction of propagation, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the

deflecting field strength.

The geometry which introduces focusing into the deflection plane can be achieved by changing the parallel pole edge geomeuy of the first and/cr last deflection elements to an angled pole edge geometry or changrng the parallel pole face geometry to an angled pole face geometry, creating an inhomogeneous magnetic field or both. These changes can be made variable by mechanical movement. The inboinogeneous field is easily modified by changing the tilt angle. Changing the edge angle has been commonly achieved in electromagnets by having a rotatable pole edge [6).

This is possible with permanent magnets, but it would involve the manufacture of complex permanent magnet geometries or the use of iron or steel pole pieces on the permanent magnet assemblies.

According to the present invention in a eighth main aspect there is provided a preferred embodiment of the apparatus in which there are three or four magnetic deflections in a deflection plane, the first and/or last deflection region having a geomeuy which introduces convergent or divergent focusing into this deflection plane, the intermediate region or regions having no focusing action in the deflection plane, the first and last regions deflecting in the same direction the intennediate region or regions having the opposite direction, arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters

deflecting field strength.

The importance of the focusing deflections being the first and/or the last is because the beam position in these regions is fixed by the coincident input and output beam general directions of propagation, allowing the geometric changes to be mfide without substantially changing the trajectoty of the beam through the apparatus (with all non-focusing deflections).

According to the present invention in a ninth main aspect there is provided apparatus in which there are at least two magnetic deflections in a deflection plane, at least one deflection region having a geometry which gives convergent or divergent focusing of the beam in this deflection plane, arranged such that that the charged particle beam general direction of propagation leaving the apparatus is substantially to a required direction of propagation and/or to a required position.

This covers the general case where any deflection region can have any focusing geometry in the deflection plane, the direction and/or position of the beam leaving the apparatus being determined in the deflection plane by the use of inhomogeneous magnetic fields or by a variable geometly and/or magnetic field strength. The focusing in the plane at right angles to the deflection plane, parallel to the initial beam general direction of propagation is controlled by deflection plane geometric parameters, beam parameters and field strengths in the various deflection regions.

According to the present invention in a tenth main aspect there is provided apparatus in which an a n isapplied to th bea efo paclesinflhave passed through the embodiment of the apparatus of this invention.

The action can include deceleration, acceleration, beam scanning or beam formation.

Deceleration is the important action on the beam for this invention, but acceleration can also be useflul for controlling the focusing in fixed field permanent magnet neutral particle filters, particularly when more than one neutral particle filter are used.

According to the present invention in a eleventh main aspect there is provided apparatus where the charged particle beam passes through at least one more apparatus embodiment with different deflection plane orientations in order to improve control of beam focus.

An important special case is an apparatus with one more apparatus embodiment with the deflection plane orientation rotated 900 about the initial general direction of propagation of the charged particle beam entering the first apparatus. Non-fodnRing deflection planes can be conveniently used, the focusing action being in the orthogonal plane. Deceleration or acceleration action on the beam either before the first apparatus embodiment, between the two apparatus embodiments or after the second apparatus embodiment or any combination of any two or any three deceleration or acceleration positions is possible. Deceleration of the beam before the first embodiment is useflul, for example, when the first embodiment is a permanent magnet system, the second deceleration region most usefiul when the second embodiment has an electromagnetic component and the third deceleration region for deceleration to veuy low energies without exposure to magnetic fields. Alternatively, the second embodiment can be a permanent magnet filter, particularly if the region between the two embodiments is both an acceleration and a deceleration region so that the focusing of the second apparatus can be independently controlled by controlling the input beam energy. Another alternative is to control the lbcus of the first embodiment by the energy of the mass analysed beam, and conirol the second embodiment focus by acceleration/deceleration or by a electmmagnetic component in the second embodiment; followed by a final deceleration after the second embodiment; if necessaly.

This anangement allows the resolving slit for the mass analysis to be placed before or in the second embodiment. Deceleration or acceleration would not generally be acceptable before the resolving slit.

Embodiments of the invention will now be descnbed by way of example with reference to the accompanying drawings some of which illustrate known apparatus, and in which:-Figs. la and lb show mass separator optics as applied to ion implanters Fig.2 shows the prior art technique for removing energetic neutral particles from an ion beam Fig.3 shows pnor art filtering of energetic neutral particles from a scanned ion beam Fig.4 shows prior art for removing envlic neutral particles from a decelerated ion beam Fig.5 shows the simplest embodiment of the invention with a parallel ion beam Fig6 shows an alternative simple ibm of the invention Figs.7 show a preferred form of the invention with a parallel ion beam Fig.8 shoes an alternative preferred fonn of the invention Figs 9 shows the invention with a converging or diverging ion beams Fig. 10 shows convergent focusing in the deflection plane Figs 11 show various convergent focusing alternatives Fig. 12 shows the invention with vanous deceleration options Figs 13 show modelled beam trajectories for a parallel beam Figs. 14 show schematic convergent focusing in the deflection plane Figs 15 show modelled beam trajectones for a divergent beam through the apparatus of this invention Figs. 16 show schematic convergent focusing in the deflection plane Fig. 17 shows an electromagnetic embodiment of the invention Fig. 18 shows a pennanent magnet and electromagnetic embodiment of the invention Fig. 19 show angled pole faces in a cross-section of the magnetic dipole Figs 20 show a double embodiment of the invention giving focusing in two orthogonal planes Fig.21 shows a variable field permanent magnet neutral particle filter where the field is changed in strength (from that of Fig.2 la) by both changing the pole gap dimension and by introducing into the quadrupole field regions moveable shunts which divert magnetic field from the pole gap regions.

Fig.22 shows a similar arrangement to F1g21 but with pole face shunts which provide a high permeability path from the permanent magnet pole face to the moveable shunt.

Fig.23 shows a thnilfir arrangement to Fig.22 but the shunts are wedge shapet Ashasbeensetoutabove,prjorajljsshownjnFigsl to4.

Although the invention is not restricted to positive ion beams but can be applied to any type of charged particle, of positive or negative polarity, the invention will now be described for the particular important application to positive ion beams. Although the invention is not restricted to magnetic deflection, and, in particular, not restiicted to fixed field permanent magnet deflection, but can be applied to electrostatic and vanous types of magnetic deflection, the invention will now be described for the particularly important embodiment using permanent magnets.

Figs. 5a, Sb and Sc show the simpIest but not the preferred, embodiment of the invention, with two defleclions in a single deflection plane. Fig5a looks along this plane in a direction at right angles to the input beam along the direction of elongation of the poles, this direction of elongation being parallel to the beam entzy pole edges 551N and 552N, and the beam exit pole edges 551X and 552X of the magnet dipoles 551 and 552. Fig5b is a view of the deflection plane containing the beam and its two deflections corresponding to Fig5a., where the output beam is parallel to the input beam. Fig5c shows a view of a deflection plane, which does not correspond to Fig. 5a where the second deflection is larger so that the output beam is deflected towards the input beam trajectoiy passing through the apparatus without deflection or obstruction.

Fig.5a shows the beam diverging from position 500 and with a central axial trajectory 501 representing the general direction of propagation of the beam, and the size of the beam and the focus condition of the beam is shown by the outer diverging trajectories 501C and SOlD. The beam passes between parallel pole faces 551F with parallel pole edges 55 iN (the beam entry edges) and 551X (the beam exit edges) in the dipole magnet 551. There is convergent focusing of the beam in the region between pole edges 551X [7] due to the inclined trajectory of the beam relative to the exit pole edge 551X seen in Fig5b. The beam then passes through the second geometrically identical magnet dipole 552, but with the opposite magnetic polarity, with parallel pole faces 552F and parallel pole edges 552E (beam entry edge) where the beam is fwther convergently focused, and then the beam leaves the magnetic dipole at exit edge 552X The magnetic steel plates 550 complete the magnet circuits passing across the pole gaps between the geometrically identical dipoles 551 and 552, and the total magnetic flux is the same across both pole gaps and no extra magnetic circuit is required (but this would not be true if, for example, dipole 551 was shorter in the direction of elongation because of the small region occupied by the various beams as shown in Fig.Sb).

Fig.5b shows a parallel beani starting at position 500, for example corresponding to the beam optics at the resolving slit 160 in Figs. la and lb. The outer trajectories detennining the size of the beam in this plane are shown as parallel trajectories 501A and 501B. The beam is shown schematically deflected in circular paths 502, 502A and 5028 by the uniform field between the parallel pole faces of magnet dipole 551, passing through the nominally free field free region between the two magnetic dipoles as straight trajectories 503, 503A and 503B. Any neutral particles formed in this region will be undeflected by the second dipole as shown as neutral particle trajectories 506, 506A and 506B. If the {mass)x{energy} product of the beam is increased then the beam trajectories will follow the lower deflection path 504, 504A and 504B through the field free region. Energetic neutral particles in the ently beam 501, 501A and 501B are undeflected as shown as 505, 505A and 505B. The two {mass)x{energy} product beams are deflected in the second magnet dipole as paths 507, 507A and 5078, and 508, 508A and 508B; leaving the magnet army as parallel beams, parallel to the input beam shown as 509, 509A and 509B, and 510, 510A and 5lOB. These beams are parallel to the initial beam entering the apparatus for a range of {mass}x{energy) products limited only in the low {mass}x{energy} product direction by the size of the magnet 552 in the direction of elongation and in the direction of displacement of the parallel beam Leaving the apparatus. The distance between the parallel trajectories 505 and 509 (or 510) will be refened to as the filter displacement', and the apparatus embodiment will be re1ed to as a iieutral particle flute?. Neutral particle beams (and veiy high energy beams) are obstructed by the plate 561, which can be removed from the beam to position 562 when the neutral particle filter flmction is not required (e.g undecelerated beams). The aperture plate 563 is shown selecting the transmission of trajectories 510, 501A and 501B but obstructing the path of trajectories 509, 509A and 509B, thus acting as an energy filter.

For brevity, the trajectories of type 501, 501A and 5OlB, will from now on be referred to as trajectories 501; trajectoiy 501 will refer to the single trajectoiy 501.

In Fig.5c, the situation is similar to Fig.5b, but the width of the permanent magnet 552 is increased to increase the magnitude of the second deflection. The low and high {mass}x{energy} product beams are further deflected as trajectories 507 and trajectoiies 508. In Fig5c the width of the magnet 552 is twice that ofFig5b leading to substantially twice the deflection angle (only exactly double for vely small deflection angles). It can be seen in Fig. Sc that trajectories 509 and trajectories 510 and the trajectories 505 of the undeflected input beam all substantially converge on a target (or deceleration region) at 590 for a range of {mass}x{energy} product beams limited only in the low {mass}x{energy} product direction by the size of the magnet 552 in the direction of elongation and the direction of the filter displacement A plate 561 is shown blocking the neutral particle beam trajectories 505; an aperture plate could also be used (as in Fig. 5b) to select a particular {mass}x{energy} product beam, the accuracy of this selection increasing as the width of the aperture decreases. For accurate selection, the beam should focus to a crossover at this position (as shown in Fig.9c).

Fig.6 shows an alternative magnet array to that of Fig.5c where the wider second magnet is replaced by two magnets 652A and 652B which have the same geometry as the first magnet dipole 651 but with opposite magnet polarity. The deflections in magnet dipoles 651 and 652A are equal and opposite giving trajectories 611 parallel to the initial beam trajectories 601 but displaced by a distance called the filter displacement. The third magnet dipole 652B deflects the beam through the same angle as 652A to give trajectories 609 which converge with the undeflected trajectories 605 substantially at 690. The third magnet dipole 652B can be of any width between zero (beam output parallel to the input beam) and a width such that the output beam converges with the undeflected trajectory at a position required by the presence of a target or an optical element such as a deceleration region. It should be noted that neutral particles created in trajectories 609 reach a target at 690, whereas neutral particle trajectories 606 and trajectories 613 can be obstructed.

Figs7a (view along the deflection plane) and 7b (view of the deflection plane) show a preferred embodiment where the beam leaving the apparatus merges into the trajectory of the initial beam passing through the apparatus without deflection and obstruction Fig.7a shows a beam diverging from a region 700, for example, corresponding to the position 160 in Figs.la and lb. Thelrajectones 701 areconvergently focused at fringe field regions at the pole edges 75lX 752N, 752X and 753N because of; an in magnitude determined by, the deflected angles (relative to a direction at right angles to the directions of elongation of the poles) of trajectories 703 and trajectories 709 in the plane ofFig7b. The larger the filter displacement in Fig. 7b, the stronger the convergent focusing seen in Fig7a.

Fig.7c shows the trajectories for two beams with different {mass}x{ener) products, 2].

having the substantially the same output optics in the deflection plane but which will have different convergent focusing powers in the plane of Fig.7a. This allows this anbodhnent to act bothasaneutral need for the filter displacement to be slightly larger than the beam size at the position of the beam stop 761 and the deflected trajectories 707 to be within the magnetic field of magnet dipole 752.

The combined central axial trajectory 701,702,707,709,711,713 experiences three deflections, two positive deflections (anticlockwise, as viewed in Fig7c) in magnet dipoles 751 and 753, and a single negative deflection (clockwise) with a deflection angle magnitude equal to the sum of the two positive deflection magnitudes. The symmetry of the magnet dipole layout shown in Fi&7c is convenient but not essential. There may be applications where an unsymmetrical layout is advantageous. A larger {nlass}x(energy} product beam central axial trajectory is shown passing through magnet dipole 752 as trajectories 708 and leaving this embodiment substantiRlly as output trajectories 713, this final trajectory being independent of the charged particle {mass}x{energy} product value within a range limited at the low {mass}x{energy} product end of this range by the size of magnet dipole 752.

The neutral particle trajectories: trajectories 705 are obstructed by plate 761, trajectories 706 are deflected clear of the final output trajectory and trajectories 714 are obstructed by aperture plate 781, trajectories 715 are obstructed by aperture plates 781 and 782. Neutral particles formed on the inside of the beam as it is deflected through magnet 752 with trajectory 716B can be transmitted to a target at 790. This can be avoided by increasing the filter displacement but in most circumstances the amount of neutrals generated in this part of the beam (which can be transmitted through apertures in plates 781 and 782) is insignificant compared with those generated in the output trajectories 713. For this reason any post-filter deceleration region must be placed as close to the beam output region as possible so that output trajectories 713 between the filter and the deceleration region are as short as possible.

If the deceleration region is before this embodiment of a neutral particle filter, then substantially all the undecelerated neutrals are removed by plate 761 and flrrther neutral generation in and after the neutral particle filter does not result in energy contamination of the output beam.

F1g8 shows a four deflection system where trajectories 816 and trajectories 817 are parallel to each other and parallel to the input beam and are displaced from the input beam trajectoiy by an amount referred to as the filter displacement'. This central region can iii certain circumstances be used (for example) as a deceleration region but the geometty of magnets 852B and853 wouldneedtobemdjfledto comp efortheIowerbeamener Asshownin Fig 8, all the deflecticins are equal in magnitude, but this symmetiy is not essential.

Fi& bows a beam focu ingon a slitin plate the ennyedgeof the firatmagnet dipole 951. The beam is not focused in the deflection plane and the output beam is substantially the same as the initial beam passing through the apparatus without deflection and obstruction.

The trajectoiy 916 of a neutral particle formed from a small region on the inside of the bend at 907B can be transmitted to a target at 971 but this is insignificant compared with neutral formation between the neutral particle filter and any subsequent action on the beam. The plate 960 can be the resolving slit 160 in Figs.la and lb. Figs.9b, 9c, 9d and 9e show alterathve positions for a resolving slit at positions 961,962, 963 and 964. Fig9c shows a geoniefly ideal for use as an energy filter with a beam cross-over at the position of maximum displacement from the straight through trajectoiy. Mother feature of this arrangement is the small obstruction component size and therefore the small deflection necessary to achieve the neutral particle filtering In all cases the output beams have the same optics as the straight through beam. The ability to remove neutrals formed on the inside of the bend at the second magnet dipole is particularly good for geometries 9c, 9d and 9e, the trajectories shown as 916c, 916d and 916e.

Fig. 10 shows the substitution of the parallel edged magnet dipole 1053 by a sector dipole 1055 with a sloping exit edge 1054. This sector geometiy introduces convergent focusing into the deflection plane (as with magnet 150 in Fig. la) and reduces the power of the convergent focusing in the plane at right angles to the deflection plane containing the central axial trajectoiy of the input beam. The position of sloping face 1054 is chosen so that the trajectoiy of the central axial trajectoiy from 1001 to 1013 is unaffected by the pole geornetiy change.

Figs. I la to 1 if show various ways of introducing focusing into the deflection plane by modiF,'ing the first and/or the last magnet dipole. A single central trajectoiy, representing the general direction of propagation, is shown for two {znass)x{energy} products. It can be seen that in Fig I Ia (as Fig 10 but with deflection plane focusing due to the sector magnet geomeüy in both the first and last deflection regions), the output trajectoty, as a function of deflection made,isunaffctdbythegjflFjgs Ilband llcthechangetotheentiy edge geometiy of the third magnet dipole does create a deviation from the ideal geomeny of 1153, resulting in a small {mass}x{energy) dependence to the direction of the output beam (changestotheeatedgeonthefirstmagnetwouId5jJ) Figs.lld and lieshow the first magnet dipole and the first and the third magnet dipoles having parallel pole edges but non-parallel pole faces, creating an inhomogeneous magnetic field. The arrows indicate the direction of increasing pole gap and decreasing magnetic field necessaiy to give convergent focusing. A small {mass}x{energy} dependence to the output beam direction is produced. These small {mass}x{energy} dependent direction abeimtions can be reduced by introducing a small amount of inhomogeneity to the central magnet dipole.

It is desirable to introduce convergent focusing into the deflection plane because it can be used to compensate for space charge blow-up of the beam. This convergent focusing in the deflection plane reduces the powcrfiul convgent focus in the plane at right angles containing the input beam axis but this powerful focusing can be adjusted by quite small changes in the filter displacement deflection region. This sector shape for all three magnetic dipoles produces powerful focusing in the deflection plane and weak focusing in the plane at iight angles due to the small entay and exit angles between the beam direction and the noimal tothe pole edge [7]. The required beam direction leaving the apparatus can be obtained by having an electromagnetic component to the field in dipole I 152F to give deflection adjustment. Alternatively, as shown in Fig.! if, the dipole 1152F can have an inhomogeneous field, the field decreasing and the pole gap increasing in the direction of the arrow, to compensate for the increasing pole width in the direction of the arrow.

Figs 12 shows the possible arrangements of deceleration regions relative the neutral particle filter. Fig 12a shows the ideal arrangement when the ener (more correctly, perveance) of the output beam is high enough (i.e. puveance low enough) to allow the decelerated beam to be eqiosed to deflecting magnetic fields. In principle, all the undecelerated neutrals produced in or before the deceleration region of electrodes 1271, 1272 and 1273 are removed from the beam leaving the neutral filter. Fig. 12b, with the deceleration region after the neutral filter, is apprcpnate for veiy low energy, high perveance beams which have space charge problems when exposed to deflecting magnetic fields. The deceleration region(eleclrodes 1274, 1275 and 1276) must be placed close to the filter to minimise the number of neutrals formed between the filter and the deceleration region. A good vacuum will also reduce the neutral generation in this region. The advantage of this arrangement is that it allows the use of a permanent magnet filter, because the input energy can be preset to a convenient value for the filter and the filter optics is independent of the final decelerated energy.

Fig. l2c shows deceleration regions before and after the neutral filter. If the filter is a permanent magnet filter, then the first deceleration region can decelerate a relatively high, constant energy input beam down to the energy required by the filter, and have energy adjustment readily available for selection of the focusing power of the filter (chnning the energy of the input beam may inconveniently involve change of energy change of analysing mfignet field strength). If the filter is an electromagnetic or a vanable field permanent magnet type, then the energy of the beam passing through the filter can be varied over a wide range. A major advantage of this two stage deceleration is that the high energy neutrals generated before the first deceleration are completely blocked by the filter, and that the small number of neutrals generated between the exit region of the filter and the deceleration region are of a substantially lower energy and therefore have less adverse effect on the low energy implant profile. The magitude of this first deceleration can be substantial because the powerfiul focusing available in the plane of Fig. l4b can compensate lbr space charge beam blow-up of the ribbon beam thickness and the magnetic and electrostatic focusing in the ribbon plane of Fag. 14a can compensate for the less powerfiul space charge effects in this plane. In order to avoid beam blow-up in the target region, the region between the final deceleration to very low energies and the target must be free of substantial magnetic fields.

Fig. 12d shows the possibility of decelerating in the central region of the filter. The change in energy will make the filter optics asymmetric in a symmetrical fixed field sextupole filter. Any combination of deceleration before, in and after the filter is possible.

Figs. 13 shows scale drawings for lOkeV singly charged boron ion trajectories modelled by software [8]. In Fig. 13b the magnets 1351 and 1353 are neodymium-iron-boron permanent magnets having a cross-section 40mm x 20mm and magnet 1352 80mm x 20mm and the sextupole magnet cross-section dimensions are 240mm x 120mm. A resolving slit 1360 is placed at the exit edge of the first magnet in order to minimise the powerful convergent focusing in this plane (the beam passes through the centre of the first lens position at the first pole gap exit region without any significant focusing). The output beam focuses to a crossover at 1340. The beam shown in Fig,13a is a parallel beam, 40mm wide. The large dimension of the beam in this plane requires a large deflection to avoid the neutral obstruction component 1361 and leads to powerful focusing in the plane of Fig.13b (which can be reduced by increasing the length of the sextupole magnet and decreasing the deflection angLe). A larger filter displacement would prevent neutral trajectoiy 1316 from being transmitted to the target Fig.14 is schematic and shows a 40mm parallel input beam in the deflection plane in Fig. l4a with a convergent output due to an inhomogeneous magnet 1453 (the pole gap between pole faces 1453F increases in the direction of the arrow in 1453. The deceleration electrodes after the filter can be of circular symmeliy but curved rectangular slot deceleration electrodes 1474,1475 and 1476 are shown in Fig. 14a after the filter to give further convergent tbcusing in the nbbon plane. Fig.14b shows a resolving slit as the first electrode 1471 of a filter input deceleration region, electrodes 1472 and 1473 providing electron suppression and deceleration respectively (shown as straight electrodes, giving no fbcusing in the ribbon plane), before the beam enters the neutral filter and the output beam 1413 leaving as a narrow converging beam through the deceleration electrodes 1474, 1475 and 1476. In Fig. 14b, the beam is shown schematically as passing through the deceleration regions without focusing The advantage of decelerating a thin, but wide, ribbon beam in a narrow line lens is that the deceleration region can be small, thus reducing the distance over which there is no space charge neutralisation, and powcrfld (due to the nall aperture) and with a space charge blow-up profile which is independent of nl,bon beam thickness [9J. The curvature of the line lens determines the magnitude of the convergent focusing in the ribbon plane. This is shown in Fig 14b for the deceleration region after the filter, but this can be applied to either or both of these deceleration regions. An assembly of straight and/or curved electrodes (or electrodes of difibrent curvature) as illustrated in Figs.l4c and l4d can be used to give variable focusing, the distribution of decelerating field between the electrodes detennining the focusing power. Two or more gaps are needed for deceleration, one for electron suppression and one or more for deceleration. This technique can also be applied to acceleration. The electrodes will be concave (as viewed from the filter input) for convergent deceleration lbcusing and convex for acceleration convergent focusing. A powful lens action could be achieved by a large accelerating electron suppression field between convex electrodes Ibilowed by a decelerating field between concave electrodes.

This would be particularly appropriate fix the first deceleration gap, where all neutrals can be removed from the beam by the filter.

The pole gaps in Fig. 14b can be made mechanically variable so that the fields can be varied to suit the required {mass}x{energy} product beam passing through the filter.

Figs. 15 shows computer modelled [8] trajectories for a diverging I OkeV singJy charged boron beam in the deflection plane of a neutral filter placed after the resolving slit 1560 in the mass resolving plane of Fag. la, where the resolving slit 160 is represented by 1560 in Figi St The permanent magnets 1551 and 1553 in Fig 14b have cross-section dimensions 20mm x 20mm and magnet 1552 has dimensions 40mm x 20mm and the sextupole magnet cross-section is 240mm x 120mm and the pole gaps are 50mm. The magnets can be smAller than the wide ribbon beam deflection plane of the magnets of Figs. 14 because the beam width at the position of maximum filter displacement is smaller and a smaller deflection angle can be used. Small deflections flavour the lmnnnission of trajectory 1516, but Ipractice this is insignificant compared with the neutral generation in the exit region between the filter and a deceleration region. The exit beam 1516 in Fig 15a is substantially the same as the trajectoiy 1505, the trajectory of the undeflected and unobstructed initial beam.

Fig I 6a shows the deflection plane beam profile being focused (schematically) to a parallel beam by an inhomogeneous magnet dipole 1653 and reduced beam convergence in the plane shown in Fig. 16b (compared with Figs 15). A large beam in both dimensions is shown passing through large aperture deceleration region electrodes 1674,1675 and 1676 (without focusing). When a beam is circular or square in cross-section, the space charge situation favours a large beam size [10]. Deceleration can also be carried out at the resolving slit position 1651 and this would determine the energy of the beam entering the permanent magnet filter.

Fig.17 shows an electromagnet neutral filter with an iron or steel magnetic circuit 1750 which is 240mm x 150mm and the length of the magnet in the beam direction including the coils is 280mm. The trajectory 1713 shown is for a 5keV singly charged boron beani [8] and a coil cunent densityof4Amm in coils 1731, 1732 and 1733, a 20mm pole gap and the filter displacement is 45mm, sufficient for a 40mm parallel nbbon beam in the plane perpendic War to that ofFigl7..

Fig 18 shows a combined permanent magnet/electromagnet neutral filter. The permanent magnet components 1851P, 1852P and 1853P provide the field necessaly to achieve the required filter displacement, and the electromagnet coils 1831 together with magnetic iron circuit 1850 provide sufficient vaiiable field to achieve the required range of focusing capability. In this layout, the permanent magnet and electromagnet components are in series; a parallel approach is also possible.

of elongation and can be advantageously be angled in order to inaximise beani clearance through the pole gaps and to modiij focusing properties and flinge fields.

Fig2Oa shows a double filter/focus system viewed in the resolving plane of' the analysing magnet l5Oin Fig.la. andFig20bthepIaeatrjghtangles(b isrepresentedbyasingfe central trajectoiy for simplicity). This gives control of focusing in two planes, either by controlling the energy of the beam through pamanent magnet sextupole magnets, and/or by controlling the fields in the sextupole by an electromagnetic component There can be any combination of one, two or three deceleration regions. Either or both of these sextupoles can act as neutral particle filters.

Fig.21 shows the cross-section geometly of a sextupole neutral particle filter with pennanent magnet mataial 210PM typically the high field strength but low permeability neodymium-iron-boron material, with an iron return circuit 2IOFE. The field 21 IF in the pole gap 210 can be decreased to a lower strength 2l2Fby increasing the pole gap dimension as shown at212. The fieldcanbefizrtherred (ass bythe lowerfie1d1inedensity)fiomtjat in the pole gaps of Fig2la by moving the shunts from position 2l3A to position 213B, such that a significant fraction of the magnetic flux is diverted from the pole gap to a path through the high permeability shunt at position 23 lB. A single example of a pole gap and a movable shunt is described above, but this applies to the other pole gap and shunt positions in a similar way. An ently shunt separated from the filter magnetic circuit is shown at 215 and an exit shunt at 216.

These limit the fringe fields in the ion beam enhly and e,ut positions (important when there are deceleration gaps close to these regions).

Fig.22 shows high permeability pole face shunts at 224 (and on the other pennanent magnet pole faces) which improve the efficiency of the movable shunts 223 by providing a high permeability path from the pole face to the movable shunt. Fig 22b shows the movable shunts 223 in intermediate positions and Fig.22c the movable shunts 223 in high diverted field positions. The entiy and exit shunts 225 and 226 respectively are connected to the sextupole of the field across the neighbouring pole gap, thus reducing the field penetration into neighbouring regions (through which the ion beam passes).

Fig.23 shows the use of wedge shaped pole face shunts The thin end of the wedge allows the permanent magnet pole gap to be as small as possible, giving the advantages of high field across the pole gap and small flinge field penetration into neighbouring regions The thick end of the wedge gives the necessazy pole face shunt cross-section to divert the field efficiently from the pole faces to the movable shunts (shown at 233A for high pole gap fields and 223B for high

diverted field and lower pole gap fields).

REFERENCES

[1] The Physics and Technology of Ion Sources, Ed. Ian G Brown, John Wiley & Sons (1989), Chapter 3, Ion Extraction, R Keller [2] J H Freeman, Proc. Roy. Soc. A.311. 123-130 (1969) [3] J H Freeman, Proc. lilt. Mass Spectrosc. Conf., Kyoto, Japan (1969) [4] The Physics and Technology of Ion Sources, Ed. Ian G Brown, John Wiley & Sons (1989), Chapter 3, Ion Extraction, R Keller, p.27 [5] D Ait.ken, US Patent 6,498,348 B2, Dec.24, 2002.

(6] J H Freeman, UKAEA Research Group Report, AERE-R 6254, H M Stalioneiy Office (1970) [7] H A Enge, Rev. Sci. Instr., 11,278, (1964)

[8] PC-OPERA, Vector Fields Ltd., Oxford, England

[9] P. DahI, Introduction to Electron and Jon Optics, Academic Press, p.116(1973) [10] P. DahI, Jnlroduction to Elecirozi and Ion Optics, Academic Press, p.111 (1973) CLATh

Claims

1. Apparatus for removing uncharged cncgv1ic particles from a charged

particle beam which has an initial general direction of propagation and an initial position compnsing at least a first and a second region of charged particle deflection for deflecting in a said first region the charged particle beam from its initial general direction of propagation to a second general direction of propagation followed by said second deflection region, a beam obstructing component for collecting uncharged energetic particles in the charged particle beam which are not deflected by the first charged particle deflection region the arrangement being such that in operation the charged particle beam is deflected by the first deflection region to avoid wholly or partly the beam obstructing component and is then deflected by the second deflection region to a required direction of propagation and/or to a required position.

2. Apparatus according to Oaim 1, in which the arrangement is such that the charged particle beam leaving the apparatus is substantially coincident with or substantially parallel to its onginal direction

3. Apparatus according to Claim!, in which the arrangement is such that the deflected charged particle beam after deflection passes substantially through a region through which the initial charged particle beam would have passed in the absence of deflection and obstruction.

4. Apparatus according to any preceding claim, in which the first region of deflection is followed by at least two more regions of deflection.

5. Apparatus according to Claim 4, in which the charged particle beam direction and position on leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction.

6. Apparatus according to Claim 2 in which the charged particle beam direction is substantially the same for a range of charged particle parameters comprising mass and/or ener' and/or charge state; with or without changes to the strength of any of the deflecting fields.

7. Apparatus according to Claim 3 in which the region through which the beam passes is substantially the same for a range of charged particle parameters comprising mass and/or energy and/or charge state; with or without changes to the strength of any of the deflecting fields.

8. Apparatus according to aaun 5 in which the charged particle beam direction and position on leaving the apparatus is substantially the same for a range of charged particle parameters comprising mass and/or energy and/or charge state; with or without changes to the

strength of any of the deflecting fields.

9. Apparatus according to any preceding claim in which the deflections are in a single deflection plane.

10. Apparatus according to Claim 1,2,3,6,7 cr9 in which there are two deflection regions and the deflections are in opposite directions and the magnitude of the second deflection is equal to or larger than that of the first deflection.

11. Apparatus accordingtoanyofclajpjs I to4,or6,7or9in which therearethree deflection regions, the second and third deflections being in the opposite direction to the first deflection.

12. Apparatus according to any of Claims 4,5,8 or 9 in which there are three deflection regions, deflecting in alternate directions such that the magnitude of the deflection angle in the second region is equal to the sum of the magnitudes of the deflection angles in the first and third deflection regions.

13. Apparatus according to Claim 12 in which the magnitude of the first and third deflection angles are the same.

14.

deflection regions, the first and second deflections being equal in magnitude and opposite in directions, the third deflection direction being the same as the second deflection direction, the third and fourth defleclions being equal in magnitude and opposite in direction.

15. Apparatus according to Claim 14 in which all deflection angle magnitudes are the same.

16. A method for removing uncharged energetic particles from a charged particle beam having an initial general direction of propagation and an initial position compnsing providing at least a first and a second region of charged particle deflection for deflecting in a said first region the charged particle beam flom its initial general direction of propagation to a second general direction of propagation followed by said second deflection region, providing a beam obstructing component for collecting uncharged energetic particles in the charged particle beam which are not deflected by the first charged particle deflection region arranging that in operation the charged particle beam is deflected by the first deflection region, wholly or partly avoiding a beani obstructing component and then deflecting in a second deflection region to a required direction of propagation andfor to a required position.

17. Amethocj according to Claim 16; arranging for the beam leaving the apparatus to have the substantially the same direction and/or passing through the same position for a range of charged particle parameters compnsing mass and/or energy and/or charge state; with or without changes to the strength of any of the

deflecting fields.

18. A method according to Claim 16 or 17; arranging for the beam leaving the apparatus to be substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without changing the strength of the deflecting fields.

19. Apparatus according to Claim 9 in which there is no focusing of the beam in the deflection plane.

20. Apparatus according to Claim 9 in which there is convergent or divergent focusing of the beam in the deflection plane.

21. Apparatus according to any preceding claim in which the beam obstructing component is movable arranged such that high mass and/or high energy beams can pass through the apparatus with small deflections and without obstruction, when the removal of uncharged particles is not

22. Apparatus according to any of Claims I to 20 in which the beam obstructing component is a movable aperture plate arranged such that uncharged particles can be removed and a particular charged particle parameter beam substantially selected for transmission through the aperture and through the apparatus.

23. Apparatus according to any preceding claim in which electrostatic fields are used to deflect the charged particle beam.

24. Apparatus according to any preceding claim in which magnetic and electrostatic fields are used to deflect the charged particle beani.

25. Apparatus according to any preceding claim in which magnetic fields are used to deflect the charged particle beam.

26. Apparatus according to aaim 24 or 25 in which invariable magnetic fields are produced by permanent magnets.

27. Apparatus according to aaim 24 ci 25 in which variable magnetic fields are produced by pennanent magnets with a variable geometry magnetic circuit.

28.

electromagnets.

29. Apparatus according to any of Claims 24 to 28 in which the magnetic fields are produced by permanent magnets and electromagnets.

29A. Apparatus according to Claims 27,28 or 29 in which the magnetic field value is selected as required for a particular mass energy product ion beam.

30. Apparatus according toy of Claims 24 to 29 in which the deflections are all in a single deflection plane and in which, in at least one of the deflection regions, there is magnetic deflection in the pole gap of a magnetic dipole which has no focusing action in the deflection plane.

31. Apparatus according to Claim 30 in which, in at least one of the deflection regions there is mfignetic deflection in the pole gap of a magnetic dipole; in which the magnetic field direction and magnitude at any position in a cross-section of the magnetic dipole pole gap in a first plane, is constant along a direction of elongation perpendicular to this said first plane.

32. Apparatus according to Claim 30 or 31 in which, in at least one of the deflection regions, there is a magnetic deflection in the pole gap of a magnetic dipole; comprising the means of generation and distribution of a magnetic field which passes through a pole face, across a pole gap and then through the other pole face of the dipole; in which the geometric and magnetic paramets in a cross-section in said first plane, which substantially detennine the field distribution and magnitude across the pole gap, are constant along a direction of elongation perpendicular to this said first plane.

33. Apparatus according to any of Claims 30 to 32 in which, in the said at least one magnetic deflection region, there is a second plane at iight angles to the said first plane in which the geometric parameters in a cross-section of the magnetic material of the magnetic dipole are constant along a direction perpendicular to this said second plane over a range in which this cross-section contains the pole faces of the magnetic dipole, this direction being parallel to the deflection plane.

34. Apparatus according to any of Claims 30 to 33 in which the initial general direction of propagation of the beam is parallel to the said first plane.

35. A method according to any of Claims 16 to 18, arranging for deflections to be in a single deflection plane and in at least one of the deflection regions, deflecting magnetically in the pole gap of a magnetic dipole, arranging that there is no focusing action in the deflection plane.

36. Apparattzs according toy of Claims 30 to 34 in which the at least one deflection region compnses all deflection regions.

37. Apparatus according to any of Claims 25 to 36 in which the deflections are all in a single deflection plane and in which, in all of the deflection regions, there is magnetic deflection in the pole gap of a magnetic dipole which has no focusing action in the deflection plane.

3& Apparatus according to Claim 37 in which the deflections are all in a single deflection plane and in which, in all of the deflection regions, there is magnetic deflection in the pole gap of a magnetic dipole; in which the magnetic field direction and magnitude at any position in a cross-section of the magnetic dipole pole gap in a first plane, is constant along a direction of elongation perpendicular to this said first plane.

39. Apparatusaccordingtocjajms3l or38inwhicbthedeflectk,nareallinasingle deflection plane and in which, in all of the deflection regions, there is a magnetic deflection in the pole gap of a magnetic dipole; comprising the means of generation and distribution of a magnetic field which passes through a pole face, across a pole gap and then through the other pole face of the dipole; in which the geometric and magnetic parameters in a cross-section in said first plane, which substantially determine the field distribution and magnitude across the pole gap, are constant along a direction of elongation perpendicular to this said first plane and parallel to the deflection plane.

40. Apparatus according to any of Claims 37 to 39 in which, in all said magnetic deflection regions, there is a second plane at iight angles to the said first plane in which the geometric parameters in a crass-section of the magnetic material of the magnetic dipole are constant along a direction perpendicular to this said second plane over a range in which this cross-section contains the pole faces of the magnetic dipole, this direction being parallel to the deflection plane.

41. Apparatus according to any of Claims 37 to 40 in which the initial general direction of propagation of the beam is parallel to the said first plane.

42. Apparatus according toy of Claims 37 to 41 in which there are at least three magnetic deflection regions, and in which the geometric and magnetic parameters in the first and last regions are substantially the same.

43. Apparatus according to any of Claims 37 to 42 in which there are two magnetic dipole deflection regions, the deflections being equal in magnitude and opposite in direction, so that the charged particle beam general direction of propagation leaving the apparatus is substantially parallel to the initial charged particle beam general direction of propagation, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or

without a change to the deflecting field strength.

44. Apparatus according to any of Claims 37 to 42 in which there are at least two magnetic deflection regions, the deflections in at least a second region being opposite in direction to the first region, arranged such that the charged particle beam leaving the apparatus passes substantially through a region through which the initial charged particle beam would have passed in the absence of deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the

deflecting field strength.

45. Apparatus according to any of Claims 37 to 42 in which there are three magnetic deflection regions, deflecting in alternate directions such that the magnitude of the deflection angle in the second region is equal to the sum of the magnitudes of the deflection angles in first and third regions, arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a

change to the deflecting field strength.

46. Apparatus according to any of Claims 37 to 42 in which there are four magnetic deflection, the first and second deflections being equal in magnitude and opposite in directions, the third deflection being the same direction as the second, the third and fourth deflections being equal in magnitude and opposite in directions arranged such that the charged particle beam general direction of piopagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the deflecting field strength.

47. Apparatus according to Claim 46 in which the magnet deflection regions arc geomeincally symmetrical about a plane between the second and third magnetic elements and at right angles to the deflection plane and parallel to the direction of elongation of the poles, and all four deflections are equal in magnitude, so that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, fbr a range of charged particle parameters comprising mass and/or energy and/or charge state and with or

without a change to the deflecting field strength.

48. Apparatus according to aaims 46 or 47, in which the second and third deflection regions are brought into contact to create a single deflection region, so that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the deflecting field strength.

49. Apparatus according to any of Claims 37 to 48 in which there is no focusing of the beam in the deflection plane and there is convergent focusing in a plane at right angles to the deflection plane and parallel to the initial beam general direction of propagation, the strength of this focusing being determined by the charged particle parameters comprising mass and/or energy and/or charge state; and the geometric and magnetic parameters of the embodiment of the apparatus.

50. Apparatus according to any preceding Claims in which the initial beam entering the apparatus is a parallel or a diverging beam in this plane.

51. Apparatus according to any preceding asims in which the initial beam entering the apparatus is a converging beam in this plane.

52. Apparatus according to Claims 22 and 51 in which there is a beani crossover substantially at the region of maximum displacement from the charged partide beam position which the beam would have on passing through the apparatus without deflection and obstruction and a moveable beam obstructing aperture plate placed with the aperture substantially at this crossover position arranged such that uncharged particles can be removed and a particular charged particle parameter beam substantially selected for transmission through the apparatus.

53. A method according to Claim 35 of deflecting magnetically in the pole gaps of the magnetic dipoles in all the deflection regions arranging that there is no focusing action in the deflection plane.

54. Apparatus according to any of Claims 30 to 34 in which there are at least two magnetic deflections in a deflection plane; and there is at least one deflection region, in which an input or output general direction of propagation coincides with the initial general direction of propagation without deflection and obstruction, having a geometty which produces convergent or divergent focusing in this deflection plane, at least one other region having no focusing action in the deflection plane, arranged such that the charged particle beam general direction of propagation leaving the apparatus is substantially parallel to the initial charged parlide beam general direction of propagation and/or passes substantially through a region through which the initial charged particle beam would have passed in the absence of deflection and obstruction, for a range of charged particle parameters coinpzising mass and/or energy and/or charge state and with or without a change to the deflecting field strength.

55.

magnetic deflections in a deflection plane, the first deflection iegion having a geometiy which produces convergent or divergent focusing in this deflection plane the second region having no focusing action in the deflection plane, the deflections being equal in magnitude and opposite in direction, arranged such that the charged particle beam general direction of propagation leaving the apparatus is substantially parallel to the initial charged particle beam general direction of propagation, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the deflecting field strength

56. Apparatus according to any of Claims 30 to 34 or Claim 54 in which there are at least two magnetic deflections in a deflection plane, the first deflection region having a geometiy which produces convergent or divergent focusing into this deflection plane, and at least a second region having no focusing action in the deflection plane, arranged such that the charged particle beam leaving the apparatus passes substantially through a region through which the initial charged particle beam would have passed in the absence of deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the deflecting field strength.

57. Minwhichthereare1g three magnetic defleclions in a deflection plane, the first and/or last deflection region having a geornetiy which introduces convergent or divergent focusing into this deflection plane, the intermediate region or regions having no focusing action in the deflection plane, the first and last regions deflecting in the same direction, the intermediate region or regions having the opposite direction, arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to

the deflecting field strength.

58. Apparatus according to any of Claims 30 to 34 or Claim 54 or 57 in which there are three magnetic deflections in a deflection plane, the first and/or third deflection region having a geometry which produces convergent or divergent focusing into this deflection plane, the second region having no focusing action in the deflection plane, these regions deflecting in alternate directions such that the magnitude of the deflection angle in the second element is equal to the sum of the magnitudes of the deflection angles in the first and third regions, arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the deflecting field strength.

59. Apparatus according to aaim 58 in which the first and third det]ections are equal in magnitude and direction, the second deflection angle is twice the magnitude of the first or third deflection angles and opposite in direction, arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the deflecting field strength.

60. 54or57 inwhicbthereare four magnetic deflections in a deflection plane, the first and/or fourth deflection region having a geometiy which produces convergent or divergent focusing into this deflection plane, the second and third deflection regions having no focusing action in the deflection plane, the first and second deflections being equal in magnitude and opposite in directions, the third deflection being in the same direction as the second, the third and fourth deflections being equal in magnitude and opposite in directions arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, lbr a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the deflecting field strength.

61. Apparatus according to Claim 60 in which all four deflections are equal in magnitude, arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to

the deflecting field strength.

62. Apparatus according to any of Claims 54 to 61 in which there is convergent and/or divergent focusing in the deflection plane and there is convergent and/or divergent focusing in a plane at right angles to the deflection plane parallel to the initial beam general direction of propagation, the focusing in both planes being determined by the charged particle parameters comprising mass and/or energy and/or charge state, and the geometric and magnetic parameters of the embodiment of the apparatus.

63. Apparatus according to any of Claims 54 to 62 in which the geomeuy of the first and/or last magnetic deflection regions comprises angled beam entry and beam exit pole edges on each pole face.

64. Apparatus according toy of Claims 54 to 62 where the geometry of the first and/or last magnetic deflection regions comprises angled pole faces which produce an inhomogeneous magnetic field, this field being produced by the pole gap increasing or decreasing substantially in the direction of elongation of the non-focusing magnetic dipoles.

65. Apparatus according to any of Claims 54 to 62 where the geometry of the first and/or last magnetic deflection regions comprises angled beam entry and beam exit pole edges on each pole face, and angled pole faces, which produce an inhomogeneous magnetic field, this field being produced by the pole gap increasing or decreasing substantially in the direction of elongation of the non-focusing magnetic dipoles.

66. Apparatus according to Claims 63,64 or 65 where the pole geometries are mechanically variable.

67. A method according to Claim 53 of deflecting magnetically in the pole gaps of the magnetic dipoles in all the deflection regions, arranging that there is focusing action in the deflection plane

68. ApparatusaccordingtoanyofClaimslto22 and 24to29inwhichthereareatJeao magnetic deflections in a deflection plane, at least one deflection region having a geometry which gives convergent or divergent focusing of the beam in this deflection plane, arranged such that that the charged particle beam general direction of propagation leaving the apparatus is substantially to a required direction of propagation and/or to a required position.

69. Apparatus according to Qaim 68 in which there are two magnetic deflections in a deflection plane, either or both deflection regions having a geometry which gives convergent or divergent focusing of the beam in this deflection plane, the deflections being equal in magnitude and opposite in direction, so that that the charged particle beam general direction of propagation leaving the apparatus is substantially parallel to the initial charged particle beam general direction of propagation.

70. Apparatus according to Claim 68 in which there are at least two magnetic deflections in a deflection plane, at least one of the deflection regions having a geomeliy which gives convergent or divergent focusing in this deflection plane, arranged such that the charged particle beam leaving the apparatus passes substantially through a region through which the initial charged particle beam would have passed in the absence of deflection and obstruction.

71. Apparatus according to Claim 68 in which there are at least three magnetic deflections in a deflection plane, at least one of the deflection regions having a geometry which gives convergent or divergent focusing in this deflection plane, arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beam would have on passing through the apparatus without deflection and obstruction.

72.

inhomogeneous magnetic field.

73.

one in which an input or output general direction of propagation coincides with the initial general direction of propagation without deflection and obstruction, has a sector geometry magnetic dipole pole face geometry in the deflection plane, and also has a non-parallel pole gap producing an inhoinogeneous field, arranged such that the charged particle beam general direction of' propagation leaving the apparatus is substantially parallel to the Initial charged particle beam general direction of propagation and/or passes substantially through a region through which the Initial charged particle beam would have passed in the absence of deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy andlor charge state and with or without a change to the deflecting field strength.

74. Apparatus according to Claims 68 to 73 arranged such that the charged particle beam general direction of propagation and position leaving the apparatus is substantially the same as that which the beamwould have on passing through the apparatus without deflection and obstruction, for a range of charged particle parameters comprising mass and/or energy and/or charge state and with or without a change to the deflecting field strength.

75. Apparatus according to any of Claims 68 to 74 in which at least one region has a variable

geometiy and/or magnetic field strength.

76. Apparatus according to any of Claims 68 to 75, in which changes to the focusing in the deflection plane is used to determine the focusing in the plane at right angles which is parallel to the initial general direction of propagation.

77. A method according to Claim 67, arranging that there is focusing action in the deflection plane, controlling this focusing and the direction and position of the beam leaving the apparatus by inhomogeneous magnetic fields and/or variable geometiy and/or a variable field strength in at least one of the deflection regions.

78. Apparatus according to any preceding Claim in which an action is applied or a number of actions are applied to the beam before and/or after the charged particles in the beam have passed through the embodiment of the apparatus of this invention.

79. Apparatus according to any of Claim 78 where at least one of the actions, applied to the beam is deceleration.

79k ApparatusaccordingtoClaim79inwbichdecelerationisappliedtothebeambeforeand after the charged particles in the beam have passed through the embodiment of the apparatus of this invention, the apparatus removing substantially all the neutral particles generated before the first deceleration and deflecting and convergently focusing the beam in the region between the two deceleration regions.

79B. Apparatus according to Claims 27,29, 29A and 79A in which the magnetic field between the poles of the permanent magnet embodiment of tbis invention are made variable by changing the geometly of the magnet array.

79C. Apparatus according to Claim 79B in which the change of geometly consists of a change to the geometty of the pole gaps of the deflecting dipoles.

80. Apparatus according to any of Claim 78 where at least one of the actions, applied to the beam is acceleration.

81. Apparatus according to any of Claim 78 where at least one of the actions, applied to the beam is beam scannin&

82. Apparatus according to any of Claim 78 where the action, or one of the actions, applied to the beam is the formation of the beam by extraction from a charged particle source.

83. A method according to any pieceding Claim, of acting on a charged particle beam before and/or after an embodiment of this invention, removing uncharged particles and/or trnmitting selected charged particles before and/or after the actions on the beam.

84. Apparatus according to any preceding aaim where the charged particle beam passes through at least one more apparatus embodiment with different deflection plane orientations in order to control beam focus in a number of planes and remove uncharged energetic particles flDm the beam.

85. Apparatus according to Claim 84 where the charged particle beam passes through one more apparatus embodiment with the deflection plane orientation rotated 900 relative to the first apparatus embodiment about the initial general direction of propagation of the charged particle beam in order to control beam focus in two orthogonal planes and remove uncharged energetic particles from the beam.

86. Apparatus according to aaim 85 with deceleration or acceleration action on the beam either before the first apparatus embodiment, between the two apparatus embodiments or after the second apparatus embodiment or any combination of any two or any three deceleration or acceleration positions.

87. A method of providing at least one more apparatus embodiment, with different deflection plane orientations, controlling beam focus in a number of planes and removing uncharged energetic particles from the beam.

88. Apparatus according toy of Claims 79, 79A, 80 or 83 to 87 in which the acceleration or deceleration regions consist of a number of electrodes, larger than the minimum required, some or all with different aperture geometries such that the focusing power of the acceleration or deceleration region can be determined by the distribution of accelerating and/or decelerating

field between these electrodes

89. Apparatus according to Claim 88 in which the apertures are elongated slots and the geometric variable is the curvature of the electrode along the long dimension of the slot.

90. Apparatus according to Claim 27,29 and 29A in which the variable geometry magnetic circuit consists of a variable pole gap dimension such that the deflecting field strength across the pole gap can be set to a required value.

91. Apparatus according to Claim 27,29 and 29A in which the vaaiable geometry magnetic circuit consists of variable position shunts which provide a high permeability path for the magnetic field between adjacent poles of opposite polarity on each side of the pole gaps providing the deflecting fields such that the field across the pole gaps of these adjacent sets of poles can be set to a required value.

9lA. Apparatus according to Claim 27,29 and 29A in which the variable geometay magnetic circuit consists of both variable position shunts and variable pole gap dimensions.

92. ApparatusaccordingtoClaim9l or9lAinwhichhighpermeabilitypolefaceshtmtson the permanent magnet pole faces associated with pole gaps providing the deflection magnetic fields ale used to provide a high permeability path substantially parallel to the pole faces such that the field diverted though the variable position shunts can be increased for any particular position of the variable position shunts.

93. Apparatus acconuingto Claim 92 in which two of the faces of at least one pole face shunt are parallel to the associated permanent magnet pole face.

94. Apparatus according to Claim 92 in which the shunt face in contact with the permanent magnet pole face of least one pole face shunt is at an angle to the shunt face adjacent to the pole gap such the thickness of the shunt increases with increasing proximity to the vaiiable position shunts.