GB1587852A - Electron beam apparatus - Google Patents

Description

(54) ELECTRON BEAM APPARATUS (71) We, INTERNATIONAL BUSINESS MACHINES CORPORATION, a Corporation organized and existing under the laws of the State of New York in the United States of America, of Armonk, New York 10504, United States of America do hereby declare the invention for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particuarly described in and by the following statement: The present invention relates to electron beam apparatus.

Electron beam columns have been adapted for use in systems for the microfabrication of largescale integrated semiconductor circuits. For example, U.S. Patent 3,644,700 describes an electron beam column adapted to form or "write" selected patterns on semiconductor wafers. Such columns have particular utility in the writing of such patterns on photoresists, i.e., exposing selected areas on photoresists which are then developed to form the photo-resist masks extensively used in a wide variety of operations during integrated circuit fabrication. The typical electron beam columns utilized in connection with such integrated circuit microfabrication applications generally include an electron beam source, condenser lenses, alignment stages, demagnification lens stages, a projection lens, a deflection unit, and a target area, arranged in well-known fashion.Typical electron beam columns and components thereof are further described in United States Patent 3,949,228 and United States Patent 3,984,678. Typical optical systems and components for such columns are further described in U.S. Patent 3,930,181 and in the publication "New Imaging and Deflection Concept for Probe-Forming Microfabrication Systems", H.C. Pfeiffer, J. Vac.

Sci. Technol ., November/December 1975, Vol. 12, No. 6, pp. 1170 - 1173.

The advantages of the square shaped electron beam over the more traditional Gaussian round beam has been set forth in detail in the above Pfeiffer article in the J. Vac. Sci.

Technol, as well as in the above mentioned U.S. patents 3,644,700 and 3,949,228. As set forth in these publications resolution and current density in electron optic systems are determined by the electron optical configuration and are effectively independent of the size of the target image. The Pfeiffer article indicates that a beam of relatively uniform intensity having twentyfive times the area of the usual Gaussian spot at approximately the same edge dose gradient can be obtained by projecting a square shaped beam onto a target.

This result will be summarized with respect to the comparison of the round beam and square shaped beam in Figure 2 of the accompanying drawings. The comparison illustrates the advantage of square beam over a Gussian round beam with identical resolution. In Figure 2, the shape and size of a Gaussian beam spot 25 is compared to a square shaped beam spot 26 and in the graph beneath each beam spot layout the intensity distribution, i.e, intensity is plotted with respect to spot area. As set forth in the Pfeiffer article, in a conventional roundbeam system, the intensity distribution half-width (d) equals the spatial resolution.The resolution of a Gaussian round beam is determined by the superposition of all n aberration disks 6i plus the demagnified Gaussian source, which typically equals the quadrature sum of the aberrations for optimum current density:

n d6ussin = ( i aberrations2 +' 5 Gaussian2) 1/2 i=1 To produce a true copy of the pattern, the half-width of the spot has to be at least five times smaller than the smallest element of the pattern.For the square spot, resolution is determined by the edge slope of the intensity distribution, caused only by the superposition of all n aberration disks:

n dsquare = ( ss 6 i aberrations2) 1/2 The Gaussian disk of the source does not contribute to the edge shape. The size of the square spot is independent of resolution and can be chosen to match the smallest segment of the pattern. This entire pattern segment is exposed at once, thereby speeding up the exposure rate by a factor of twentyfive over a comparable round-beam system.

It may be seen from Figure 2 that using the square shaped beam, rectilinear area 27 (i.e., an area defined by straight lines) may be totally exposed by six stepped exposures ( 1 through 6) with square beam 26 while the same rectilinear area 27' would require in the order of one hundred and forty stepped exposures with the round Gaussian beam 25.

While present shaped or square beam systems provide significant electron beam capability in the integrated circuit fabrication field, it is foreseeable that in future technologies wherein portions of the patterns to be exposed may have dimensions below two microns, the application of the shaped electron beam in forming such patterns may be limited. In such dense integrated circuits having aperture widths and/or line widths with smallest dimensions below two microns, the "blooming effect" produced by double and greater multiple exposures to which some pattern areas may be subject when using shaped aperture apparatus may be beyond the dimensional tolerances of the integrated circuits.

The problem of double exposure will be elaborated on in greater detail with respect to Figures 3 and 4. However, it may be seen in its simplest form with respect to area 27 which is exposed by the square shaped beam in Figure 2. Utilizing a beam providing a square shaped spot 26, area 27 is exposed by six stepped exposures. Since the selected area to be exposed does not have dimensions which are integral multiples of spot 26, a portion 28 (crosshatched) will be double exposed. This double exposure results in the known "blooming effect" wherein the double exposed areas, for example, in the photoresist develop at a faster rate than the normal singly exposed areas during development. This produces undercutting and edge irregularity in the remaining photoresist defining the exposed area.With the present integrated circuits having line widths and apertures with smallest pattern dimensions of at least two microns, this "blooming effect" is within dimensional tolerances and presents no problems. However, with the denser, more advanced integrated circuits having lateral dimensions below two microns, the "blooming effect" may produce dimensional irregularities beyond the lateral tolerances.

In addition, in advanced integrated circuit technology, it would be highly desirable if the time required for electron beam exposure of selected patterns could be reduced thereby increasing the integrated circuit fabrication throughput.

Accordingly, it is an object of the present invention to provide a method and apparatus for exposing target to an electron beam in a selected pattern wherein the areas subject to multiple exposure may be reduced or eliminated, and wherein the speed of the exposure operation may be increased, compared to the prior art.

According to the present invention there is provided an electron beam apparatus having a source of electrons, a target area toward which the electrons are directed, and electron beam shaping means along the path from the source to the target area, the shaping means comprising a first beam shaping member having a first aperture therein, a second beam shaping member having a second aperture therein, means for focusing the image of the first aperture in the plane of the second aperture whereby the cross-sectional shape of the beam is defined by the overlap between the image and the second aperture, and means for deflecting the image of the first aperture laterally with respect to the second aperture in order to vary the degree of overlap so as to vary the dimensions of the beam cross-section, the image of the source being focussed in a plane coincident with the virtual centre of the deflection.

Embodiments of the invention will now be described with reference to the accompanying drawings, in which: Figure 1 is a schematic diagram of an electron beam apparatus in accordance with the present invention; Figure 2 shows the arrangement and number of exposures required to expose a given area with, respectively, a round beam, a square beam, and a variable shaped beam, the figure also showing the current density for the various beams:: Figure 3 is a diagrammatic plan view of a portion of an rectilinear target pattern showing the exposure steps required in order to fully expose the portion utilizing the prior art square beam of fixed dimensions; Figure 4 is a diagrammatic plan view of the same target pattern portion modified to show the electron beam exposure steps required when exposing with a variable shaped beam; Figure 5 is a schematic view of a prior art square shaped electron beam apparatus; Figure 6 is a schematic view of a variable shaped electron beam apparatus according to the present invention syowing linked-beam trace when there is no deflection; Figure 7 is a view similar to Figure 6 showing the modification of the linkedbeam trace when there is deflection of the image of the first aperture with respect to the second aperture;; Figure 8 is a schematic view of a deflection apparatus which may be used to laterally deflect the image of the first aperture during beam spot shaping; and Figure 9 is a view of the same apparatus as Figure 8 showing the operational modifications in moving the virtual centre of deflection into coincidence with the plane of the focused source image.

With reference to Figure 1, an apparatus for shaping the beam spot will be described in general. An electron source 10 directs a beam of electrons 11 along an axis 12 in an electron beam column toward a target which is not shown. The beam is shaped into a variable rectangular spot by first passing the beam through square aperture 14 in shaping member 13. Condenser lens 15 focuses the image 14' in the plane of square aperture 16 of shaping member 17 and focuses the image of the source at a position 18 in a plane coincident with the center of deflection provided by deflection means 19 which move the focused image 14' of the first aperture 14 laterally with respect to aperture 16.In the embodiment of Figure 1, deflecting means 19 are conventional electrostatic deflection plates with plates 20 and 20' acting to deflect image 14' in the X direction while plates 21, 21' act to deflect image 14' in the Y direction. For optimum results, image 14' has the same dimensions and shape as aperture 16. The final beam spot shape is determined by that portion 22 of image 14' which is not blocked by the plate of shaping member 17 and passes through aperture 16 as shaped composite image 23. While the operation has been illustrated with respect to a deflection in the X direction, it will, of course, be understood that with various combinations of deflections of image 14' in the X and Y direction a wide variety of rectangular shapes may be achieved for composite spot 23.

The variable spot shaping apparatus shown in general in Figure 1 may be used in combination with standard electron beam columns such as those described in the above mentioned Pfeiffer article in the J. Vac. Sci. Technol or in U.S. Patent 3,644,700. When used in such columns, the composite spot 23 may then be passed through demagnification lenses and projection lenses to project the image of spot 23 onto a target. Also, the beam may be deflected in the conventional manner for scanning purposes with respect to such targets. Standard demagnification, projection and deflection equipment such as that described in U.S. Patent 3,644,700 and the Pfeiffer article may be used for this purpose.

One important aspect of the apparatus of Figure 1 which will be discussed in greater detail subsequently with respect to Figures 6 - 9 is the optical separation of the image of source 10 from the image of aperture 14. The image 14' of the aperture 14 is focused in the plane of aperture 16 by condenser lens 15 while the same lens focuses the image of source 10 in a plane near the center of deflection produced by deflection means 19. There will subsequently be described with respect to Figures 6 - 9 how the plane of the image 18 of source 10 is brought into coincidence with the center of deflection. In any event, by such optical separation of the focused images, the illumination, i.e., current density of the beam spot at the target, which is dependent on the position of the source image remains constant.

This constant illumination is ensured because the source image is not deflected and remains essentially aligned with the electron beam column axis. This alignment also provides for maximum edge resolution of the beam spot at the target after conventional demagnification and projection stages.

While the aperture shaping technique of the present embodiment is described herein with respect to conventional electron beam apparatus using raster scanning of the beam with respect to the target, it should be clear that the technique is equally applicable to apparatus employing other modes of scanning, e.g., vector scanning.

With a variable shaped beam produced as above, the multiple exposure and consequent "blooming problems" are clearly avoidable. This may be simply understood again with reference to Figure 2. The overlapped region 28 resulting from the use of the square shaped beam is avoided when the variable shaped beam is used to fill in the area 27" which is equivalent to area 27. The exposure is achieved in two steps: exposure by a first shaped beam 29 followed by an exposure second shaped beam 30 which abuts the area exposed by beam 29. As may be seen from the accompanying profile of current density, there is no overlap which could lead to increased exposure of "blooming", i.e., total current density at interface 31 between the two exposed areas does not exceed the current density or intensity of illumination of the beam spot which remains constant irrespective of the shape of the beam.

It should be noted from the beam profiles in Figure 2 that the resolution of the variable shaped beam remains substantially constant. Resolution is described in the above mentioned Pfeiffer article in J. Vac. Sci. Technol. as the half-width of the intensity distribution (d) of a round or Gaussian beam having the selected intensity, i.e., current density. The Pfeiffer article further indicates that the embodiment is described herein with respect to conventional square shaped beam will have the same resolution as the round beam provided that the edge slope of the square shaped beam equals the half-width (d) of the round beam.We have further found with respect to the variable shaped beam that so long as the smallest dimension of the pattern being subjected to stepped exposure by the variable shaped beam is at least five times the resolution (d) of the round shaped beam of the selected intensity, then the pattern will be subject to the same resolution as the one exposed by a plurality of round shaped beams. For example, with reference to Figure 2, although beam spot 30 has one dimension of only 4d, the total dimension of the pattern is 14d. Thus, it will be subject to substantially the same resolution as pattern 27' which is exposed to the round beam.

The advantages of the variable shaped beam both with respect to the elimination of multiple exposure and increased speeds for exposure, i.e., increased throughput, will be more readily understood with reference to Figures 3 and 4. Section 32 in Figure 3 is the layout of an orthogonal rectilinear pattern to be exposed by an electron beam column. The pattern is to be formed in a photoresist, for example, and is to define metallization lines. If pattern element 32 were to be exposed by a square beam in accordance with conventional practice, this square beam would be limited to a spot 33 having a width, W, which is no greater than the minimum dimension of the entire pattern being exposed. With such a square beam spot 33, it would require twenty-seven stepped exposures to completely expose portion 32 of the pattern.Because of the geometric limitations, areas 34 which are crosshatched would be doubly exposed and at least one region 35 would be quadruply exposed. Such multiple-exposed areas would be subject to "blooming effects".

On the other hand, with reference to Figure 4, if the portion of the rectilinear pattern 42 which is equivalent in shape and dimensions to pattern portion 32 in Figure 3 is exposed to a variable shaped beam, then only eight exposure steps are necessary to expose the total pattern portion. Five different beam shapes would be involved. The first beam shape would be used in exposing areas 43, 43A; a second shape for areas 43B and 43C; a third shaped spot to expose area 44; a fourth shaped spot to expose area 45, and a fifth shaped spot used to expose area 46. The total exposure of pattern portion 42 is accomplished without any overlap; each of the stepped exposed regions abuts the adjacent region. There is no need for any exposed region to overlap the adjacent exposed region.In addition, the total exposure for the same pattern portion is accomplished by the variable shaped beam apparatus in one-fourth as many stepped exposures as with the fixed square shaped aperture. Thus, the total time for exposure of substrate patterns in integrated circuit fabrication would be substantially reduced, thereby greatly enhancing thoughput.

The electron beam shaping apparatus will be more clearly understood, particularly with respect to the application of principle of optical separation during the beam shaping operation, when reference is made to Figures 5, 6 and 7. Figure 5 which is a diagrammatic view of a conventional electron beam column as described in the above mentioned Pfeiffer article in the J. Vac. Sci. Technol. comprises an electron source 50, a condenser lens 51, a square spot shaping aperture 52, blanking plates 53 which operate in the manner described in United States Patent 3,644,700 a demagnification system comprising a first demagnification lens 54 and a second demagnification lens 55. The apparatus further includes a circular aperture plate 56 for defining the axial portions of the beam which have the maximum intensity in the manner described in U.S.Patent 3,644,700 as well as a standard projection lens 57 having a central deflection yoke 58 and standard dynamic elements 49 for correcting field curvature and axial and deflection astigmatism. This projection lens and yoke are a standard structure as described in detail in the Pfeiffer article. They may also have the structure described in U.S. Patent 3,930,181 or 3,984,687.

The column directs a beam at a target 59, e.g., photoresist covered semiconductor wafer on which a rectilinear pattern is to be exposed.

The linked-beam trace shown in Figure 5 is the same as that shown in the above mentioned Pfeiffer article. The design of the column in Figure 5 optimizes the beam current density, i.e., intensity distribution, simultaneously with the square shaping of the spot by implementing the linked-beam imaging concept of A. Koehler, Z. Wiss. Mikroskopie, 10, 433 (1893), further described in H.C. Pfeiffer and K. H. Loeffler, Proceedings of the 7th International Conference on Electron Microscopy, Grenoble (1970), p. 63.As the drawn linked-beam trace shows, condenser lens 51 images the source 50 into the entrance pupil of the first demagnification lens 54 of the demagnification section, to thereby provide the most efficient and uniform illumination for the beam, i.e., constant current density. (The trace of the source beam is indicated by the widely spaced cross lines while the projection of the image of square aperture 52 as indicated by the narrow spaced cross lines.) Projection lens 57 generates the electron beam spot 60 by projecting the image of aperture 52 demagnified by demagnification lenses 54 and 55 onto the target wafer 59.

In the prior art structure of Figure 5, the condenser lens 51 and the projection lens 57 are the critical elements of the respective square aperture image and source beam tracing; demagnification lenses 54, 55 establish the link between the condenser lens 51 and the projection lens 57. The square aperture 61 which is being imaged to form the beam spot is formed in a thin metal plate 52. In the apparatus shown, the image of square aperture 61 is demagnified in two steps through lenses 54, 55. For example, with an aperture 61 in the order of 400 microns square, the demagnification system will reduce to a final beam spot about 2.5 microns square. While the image of aperture 61 is being thus demagnified, the first demagnification lens 54 simultaneously creates a magnified image of the source in the plane of circular aperture 63 centered about the electron beam column axis 62.The second demagnification lens 55 images the round aperture 63 at the centre 64 of projection lens and defines the semiangle of convergence. Thus, uniform beam current density is provided since circular aperture 63 admits only the central or axial portion of the source beam trace which minimizes aberrations. For a given round aperture size, the second demagnification lens determines the final beam convergence angle and consequently the required brightness, e.g., about 3 x 105 A/cm2-sterad is necessary to achieve a target current of 3 microamps.

The final or projection lens provides the required working distance deflection yoke 58 in order that the beam may be deflected over the target field to be exposed which is in the order of 5 millimeters square. The beam is deflected by a conventional deflection yoke.

During the writing of patterns by a stepped electron beam, the intensity of the beam spot 60 may be modulated by electrostatic beam blanking plates 53 which operate essentially in the manner described in Patent 3,644,700.

Referring now to Figure 6, the operation of a beam shaping apparatus according to the present invention will be described by showing the trace of the linked-beam concept by which optimum optical separation of the source beam from the shaping aperture image is achieved. As will be seen, such separation is highly desirable in that it permits a uniform current density irrespective of the shape of the spot. In the column of Figure 6, the lower portion of the apparatus has substantially the same structure and operational characteristics as the apparatus described in Figure 5. First and second demagnification lenses 64 and 65 are respectively equivalent to demagnification lenses 54 and 55 in Figure 5. Round aperture plate 66 is substantially the same as round aperture plate 56 in Figure 5.Projection lens 67 performs the same function as lens 57 in Figure 5, deflection yoke 68 performs the equivalent function of deflection yoke 58 in Figure 5, and dynamic correction elements 98 perform the same functions as elements 49 in Figure 5. Target 69 is a photoresist wafer on which a rectilinear pattern is to be formed. Also, blanking plates 63 in the column of Figure 6 performs the conventional blanking function as described with respect to plates 53 in Figure 5.

In performing the beam shaping function of the present embodiment which was previously described in general with respect to Figure 1, source 70 directs a beam of electrons along the axis of the electron beam column 71. Following the imaging concept used in Figure 5 and in the Pfeiffer article in the J. Vac. Sci. Technol., the trace of the source imaging is indicated by the widely spaced cross lines while the trace of the first aperture image and of the composite image formed by the first and second apertures is shown by the narrow spaced cross lines. The beam is shaped into a variable rectangular spot by first passing the beam through a square aperture 72 formed in plate 73. Condenser lens 74 which may suitably be a magnetic lens of conventional design in the electron beam art, performs two functions.It focuses an image of aperture 72 in the plane of square aperture 75 formed in plate 76. In addition, condenser lens 74 focuses the image 77 of source 70 at a point along the axis and in the center of the image deflection means provided by electrostatic plates 78 and 78'. This pair of plates has the capability of deflecting the focused image 79 of the first square aperture 72 with respect to second square aperture 75 during the beam shaping operation. There is, of course, a second pair of plates not shown in Figure 6 but shown in Figure 1 which act to deflect the beam laterally in the other orthogonal direction during the shaping operation. The deflection of the image 79 of the first aperture with respect to the second aperture 75 is shown in Figure 7.For the optimum operation of the electron beam column of the present embodiment the focused image 77 of source 70 must be at the virtual center of deflection of the deflection means provided by electrostatic plates 78 and 78' as well as the corresponding pair of plates for deflection in the other orthognonal direction. The focal length of lens 74 is determined primarily to focus the aperture image 79- in the plane of aperture 75. Therefore, the attendant focusing of source image 77 will not necessarily occur at the center of deflection. While it may be possible to move deflection plates 78 and 78' along the axis of the column to position source image 77 at the center of deflection, this is not considered to be very practical.There will subsequently be described with respect to Figure 8 and 9 a suitable expedient for moving the center of deflection of an electrostatic deflection system into coincidence with the plane of the focused image 77 of the source without physically moving plates 78 and 78'.

The significance of having the source image 77 at the center of deflection becomes apparent when one considers the action of condenser lens 80 which may be any standard magnetic condenser lens within which the second aperture plate 76 is disposed. Condenser lens 80 acts to image source image 77 into the entrance pupil of the first demagnification lens 64 in much the same manner that condenser lens 51 in Figure 5 acted to image the source itself, 50, into the entrance pupil of first demagnification lens 54.When a voltage differential is applied between electrostatic plates 78 and 78' in order to deflect the image 79 of the first aperture with respect to the second aperture 75 as shown in Figure 7 (Figure 7 is a schematic of the linked-beam trace of the column of Figure 6 showing the trace where the deflection apparatus operates to deflect), source image 77 is not deflected; it remains stationary because it is at the center of deflection. Consequently, irrespective of the deflection of aperture image 79 in the X or Y direction, focused source image 77 remains stationary, and the image of source image 77 projected by condenser lens 80 into the entrance pupil of first demagnification lens 64 will remain constant in position at the axis of column 71.

The demagnification system and projection system of the column in Figure 6 act on the composite image determined by laterally deflected image 79 and aperture 75 in essentially the same manner that the demagnification and projection system of Figure 5 acted upon the image of square aperture 61. Thus, in the columns of Figures 6 and 7, the composite image is demagnified in two steps through demagnification lenses 64 and 65. (In Figure 6, the composite image is identical with the second aperture 75). While the composite image is being thus demagnified, first demagnification lens 64 simultaneously creates a magnified image of the source in the plane of circular aperture 81. This image of the source is, of course, dependent on the position of source image 77.Since source image 77 remains stationary irrespective of the deflection in forming the composite aperture image, focused image 82 of the source remains centered about the column axis at aperture 81. Thus, substantially uniform current density is provided by circular aperture 81 which admits only the central or axial portion of the Gaussian source being traced and minimizes aberrations generated in the final lens.

The second demagnification lens 65 and projection lens 67 operate in the same fashion as lenses 55 and 57 in Figure 5 in accomplishing this. Likewise, deflection yoke 68 provides for the deflection of the composite beam spot 83 across the target field in the manner previously described with respect to yoke 58 of Figure 5. In addition, since the image of the source projected upon the entrance pupil of demagnification lens 64, will be centered about the axis irrespective of the deflection, only the central portions of the lenses in the demagnification system and of the projection lens will be primarily utilized. Thus, degradation of beam spot edge resolution which would result if the source image projected upon demagnification and projection lenses were off center is avoided.It should be noted that this latter effect is significant in systems which do not utilize a round beam forming aperture like aperture 81.

In this connection, it should be noted that in electron beam columns, it is possible to eliminate physical apertures like round aperture 81 which restricts the beam diameter in the demagnification and projection stages of the column by using properly scaled images of the source itself. However, the scaled image approach has its shortcomings in that there is non-uniform current density distribution of the source which leads to higher aberrations for the same total beam current or brightness.

Referring now to Figures 8 and 9, apparatus associated with the deflection means is provided for moving the center of deflection into coincidence with the plane in which the source image is focused. The apparaus will be described with respect to one pair of electrostatic deflection plates 88 and 88'. However, it should be understood that the same adjustment may be made with the pair of electrostatic deflection plates which move the aperture laterally in the other direction. The movement of the center of deflection into coincidence with the plane in which the image of the source is focused is customarily carried out before the apparatus is actually in operation. It may be conveniently accomplished during the calibration period that is described in U.S. Patent 3,644,700 for an electron beam column. Once the center of deflection is adjusted into coincidence with the focused source image, the further modification should usually be unnecessary during the operation of the electron beam column irrespective of the number or nature of lateral deflections carried out to change the beam spot shape during the normal operation.

With reference first to Figure 8, the voltage differential between plates 88 and 88' is provided by the conventional push-pull circuit through which deflection is achieved by applying a signal to amplifier 84, the output of which is applied to plate 88' and to amplifier 85. The level and the sense (positive or negative) of the signal will determine the extent of deflection. The output of amplifier 85 is, in turn, applied to plate 88 to provide a push-pull circuit in which the voltage level of plate 88 swings negative when plate 88' is positive and vice versa. This portion of the structure represents a conventional arrangement for developing the voltage differential between a pair of electrostatic plates. In addition, the present structure comprises a pair of auxiliary plates 89 and 89'.Auxiliary plates 89 and 89' are each connected to the outputs of amplifiers 84 and 85 respectively through the variable resistors 90 and 90'. Thus, let us consider the initial condition, shown in Figure 8 when a differential voltage is applied between primary plates 88 and 88'. Since the contacts 91 and 91' respectively between plates 89 and 89' and balanced resistors 90 and 90' are centered with respect to such resistors, plates 89 and 89' will be at the identical voltage level (half-way between voltage levels of plates 88 and 88'), and the deflection of the path of the first aperture image diagrammatically represented by line 92 will be deflected as shown.

Also, as the beam trace 93 of the source image indicates, the source image will be focused at point 94. The virtual center of deflection 95 of the apparatus is determined by the intercept of the extension 96 of the beam path as deflected and axis 97 which was the original path of beam 92 before deflection.

Because it is desired that the center of deflection be moved into coincidence with the focused beam at 94, the apparatus will be adjusted as indicated in Figure 9 to bring such coincidence about. Since the center of deflection 95 produced by voltage drop between primary plates 88 and 88' is above focused image 94 of the source, the center of deflection 95 is moved down by applying a potential difference between auxiliary plates 89 and 89' in the same sense as the potential difference between primary plates 88 and 88', i.e., if primary plate 88' has a positive voltage with respect to plate 88, then plate 89' is made positive with respect to plate 89. This is accomplished by moving variable resistor contact 91' as indicated in Figure 9 so that the portion of variable resistor 90' between plates 89' and 88' is reduced to move the voltage level of plate 89' toward that of plate 88'.Similarly, variable resistor contact 91 is moved to the position indicated to reduce the portion of variable resistor 90 between plate 89 and plate 88 whereby voltage level on plate 89 approaches that of plate 88. This shifts the path 92 of the aperture image as shown whereby extension 96 crosses axis 97 as shown to produce a virtual center of deflection 95 in coincidence with focus source image 94.

Conversely, if it is desired to move the center of deflection 95 upward, then variable resistor contact 91' is moved to increase the resistance between auxiliary plate 89' and primary plate 88' and to thus diminish the resistance between auxiliary plate 89' and primary plate 88. At the same time, variable resistor contact 91 may be moved to increase the resistance between auxiliary plate 89 and primary plate 88, and to thus diminish the resistance between auxiliary plate 89 and primary plate 88'. As a result, a voltage drop is produced between auxiliary plates 89 and 89' which is in the opposite sense to the voltage drop between the primary plates. This has the effect of bucking or opposing the deflection action of the primary plates and thereby moving the center of deflection upward.

In suitable operating conditions. for the deflection apparatus of Figure 8 and 9, the voltage drop which may be applied between auxiliary plates 89 and 89' may be in the order of about ten per cent of the voltage drop across the primary plates, e.g., when the voltage swing between the primary plates is in the order of about twenty volts, the voltage swing between the auxiliary plates would be in the order of about two volts.

While the description in Figures 8 and 9 was directed to moving the center of deflection with respect to means for deflecting the aperture image in one lateral direction, similar center of deflection adjusting apparatus may be used in connection with the electrostatic deflection plates which deflect the beam in the other lateral direction.

As a practical matter, the movement of the center of deflection into coincidence with the focused image of the source is accomplished during electron beam calibration by first measuring the current density of the shaped beam spot at the target using any standard measurement technique under conditions where there is no lateral deflection of the first aperture image with respect to the second beam shaping aperture. Then, if the center of deflection is coincident with the source image, the current density will remain constant irrespective of the lateral deflection of the first aperture image in the X and Y directions.

Accordingly, after the initial reading as to current density, the first aperture image is deflected in the X and/or Y direction and the auxiliary plates are "tuned" by moving variable resistor contacts 90 and 91' until the constant current density at the initial level is achieved. This indicates that the center of deflection of the deflection means is in coincidence with the source image. Once this coincidence is achieved by the initial "tuning", the current should remain constant thereafter. No additional changes should be necessary during the operation of the electron beam column when the beam shape is changed from step to step. Thus, the apparatus provides for rapid change in beam aperture size and shape during the electron beam column operation to effectively expose rectilinear regions in a target without any multiple exposure or exposure overlap and with increased throughput.

Although electrostatic deflection apparatus has been disclosed for the preferred embodiment as the means for deflecting the image of the first aperture and for moving the center of deflection, it will be clear that other deflection apparatus such as magnetic deflection apparatus may be used for the same purpose.

WHAT WE CLAIM IS: 1. An electron beam apparatus having a source of electrons, a target area toward which the electrons are directed, and electron beam shaping means along the path from the source to the target area, the shaping means comprising a first beam shaping member having a first aperture therein, a second beam shaping member having a second aperture therein, means for focusing the image of the first aperture in the plane of the second aperture whereby the cross-sectional shape of the beam is defined by the overlap between the image and the second aperture, and means for deflecting the image of the first aperture lateral with respect to the second aperture in order to vary the degree of overlap so as to vary the dimensions of the beam cross-section, the image of the source being focussed in a plane coincident with the virtual centre of the deflection.

2. The apparatus of claim 1, wherein the first and second apertures have a common rectangular shape.

3. The apparatus of claim 2, wherein the focused image of the first aperture has the same dimensions as the second aperture.

4. The apparatus of claim 1, 2 or 3, wherein the deflecting means are electrostatic deflection means positioned between the first and second apertures.

5. A method of exposing a plurality of portions of a target successively to an electron beam comprising locating the target in the target area of an electron beam apparatus as claimed in any preceding claim, and directing the electron beam produced by such apparatus at each of the said portions of the target in turn, the degree of overlap being different in respect of at least some of the said portions so that the exposed portions are not .al.l of the same dimensions.

6. An electron beam apparatus as claimed in claim 1, substantially as described with reference to the accompanying drawings.

**WARNING** end of DESC field may overlap start of CLMS **.

Claims (6)

**WARNING** start of CLMS field may overlap end of DESC **. coincidence with the source image. Once this coincidence is achieved by the initial "tuning", the current should remain constant thereafter. No additional changes should be necessary during the operation of the electron beam column when the beam shape is changed from step to step. Thus, the apparatus provides for rapid change in beam aperture size and shape during the electron beam column operation to effectively expose rectilinear regions in a target without any multiple exposure or exposure overlap and with increased throughput. Although electrostatic deflection apparatus has been disclosed for the preferred embodiment as the means for deflecting the image of the first aperture and for moving the center of deflection, it will be clear that other deflection apparatus such as magnetic deflection apparatus may be used for the same purpose. WHAT WE CLAIM IS:

1. An electron beam apparatus having a source of electrons, a target area toward which the electrons are directed, and electron beam shaping means along the path from the source to the target area, the shaping means comprising a first beam shaping member having a first aperture therein, a second beam shaping member having a second aperture therein, means for focusing the image of the first aperture in the plane of the second aperture whereby the cross-sectional shape of the beam is defined by the overlap between the image and the second aperture, and means for deflecting the image of the first aperture lateral with respect to the second aperture in order to vary the degree of overlap so as to vary the dimensions of the beam cross-section, the image of the source being focussed in a plane coincident with the virtual centre of the deflection.

2. The apparatus of claim 1, wherein the first and second apertures have a common rectangular shape.

3. The apparatus of claim 2, wherein the focused image of the first aperture has the same dimensions as the second aperture.

4. The apparatus of claim 1, 2 or 3, wherein the deflecting means are electrostatic deflection means positioned between the first and second apertures.

5. A method of exposing a plurality of portions of a target successively to an electron beam comprising locating the target in the target area of an electron beam apparatus as claimed in any preceding claim, and directing the electron beam produced by such apparatus at each of the said portions of the target in turn, the degree of overlap being different in respect of at least some of the said portions so that the exposed portions are not .al.l of the same dimensions.

6. An electron beam apparatus as claimed in claim 1, substantially as described with reference to the accompanying drawings.