BRIEF DESCRIPTION OF THE DRAWINGS
Embodiments of the invention will now be described with reference to the accompanying drawings, by way of example only, in which:
FIG. 1 is a schematic representation of an ion implanter;
FIG. 2 is a section through a beam stop according to embodiments of the present invention;
FIG. 3 is a front view of a first embodiment of a beam stop according to the present invention of FIG. 2;
FIG. 4 is a front view of a beam stop according to a second embodiment of the present invention;
FIG. 5 is a front view of a beam stop according to a third embodiment of the present invention;
FIG. 6 is a front view of a beam stop according to a fourth embodiment of the present invention;
FIG. 7 is a front view of a beam stop according to a fifth embodiment of the present invention;
FIG. 8 is a front view of a beam stop according to a sixth embodiment of the present invention;
FIG. 9 is a section through a beam stop according to further embodiments of the present invention;
FIG. 10 is a simplified section of the beam stop of FIG. 9 showing an incident high-energy ion beam;
FIG. 11 is a simplified section of the beam stop of FIG. 9 showing an incident low-energy ion beam;
FIG. 12 is a front view of a beam stop according to a seventh embodiment of the present invention; and
FIG. 13 is a front view of a beam stop according to an eighth embodiment of the present invention.
DETAILED DESCRIPTION OF THE INVENTION
In order to provide a context for the present invention, an exemplary application is shown in FIG. 1, although it will be appreciated that this is merely an example and is in no way limiting.
FIG. 1 shows an ion implanter 10 for implanting ions in a semiconductor wafer 12. The ion implanter 10 is located within a vacuum chamber 15 pumped by vacuum pump 24. In summary, ions are generated by ion source 14 to be extracted and passed through a mass analysis stage 30. Ions of a desired mass are selected by passing through a mass-resolving slit 32, and continue to strike the semiconductor wafer 12. The ion implanter 10 is operated under the management of a controller, most often a computer operating suitable software.
The ion source 14 generates an ion beam that includes a desired species. Ions generated in the ion source 14 are extracted through an exit aperture 28 using an extraction electrode assembly 26. A potential difference is applied between the ion source 14 and the extraction electrode assembly (not shown) by a power supply 21 to accelerate extracted ions to a desired beam energy suitable for transit through the ion implanter 10, the ion source 14 and mass analysis stage 30 being electrically isolated from each other by an insulator (not shown). For shallow implants, this beam energy is optimised to be high enough to avoid unacceptable blow up in the beam caused by space charge effects, and the ion beam is later decelerated prior to implantation by a deceleration lens assembly (not shown).
The mixture of extracted ions pass through the mass analysis stage 30 so that they traverse a curved path under the influence of a magnetic field. The radius of curvature traveled by any ion is determined by its mass, charge state and energy. The magnetic field is controlled so that, for a set beam energy, only those ions with a desired mass and charge state exit along a path coincident with the mass-resolving slit 32. The emergent ion beam 34 is then transported to the target, i.e. the substrate wafer 12 to be implanted held in position by a wafer holder 36. When the wafer 12 and wafer holder 36 are moved out of the path of the ion beam 34, the ion beam 34 instead continues to be incident on the beam stop 100. The beam stop 100 absorbs the heat generated by incidence of the ion beam 34 and mitigates ejection of contaminants towards the wafer 12 caused by incidence of the ion beam 34.
Either a single semiconductor wafer 12 (or other target) may be implanted at a time or many wafers 12 may be positioned on a carousel or the like that rotates to present the wafers 12 to the incident ion beam in turn.
Implantation is generally effected by scanning the ion beam 34 relative to the wafer 12 such that a required dose is achieved after many passes. Scanning may be achieved by movement of the wafer 12 using the wafer holder 36, movement of the ion beam 34 using ion optics (not shown) or a combination of the two.
The geometry in the region around the wafer 12 is indicated by axes in FIG. 1. The z axis corresponds to the direction of travel of the ion beam 34. The x axis is taken to be the horizontal, and the y axis is taken to be the vertical.
A beam stop 100 according to embodiments of the present invention is shown in section in FIG. 2. The beam stop 100 comprises a Faraday cup having an entrance aperture 102 defined by an annular collar 104 such that the aperture 102 is sufficiently wide to accept the ion beam 34. In this embodiment, the wafer 12 is scanned relative to a fixed ion beam 34. This means only small movement of the ion beam 34 is effected (effectively to correct and optimise beam shape and position) and so the aperture 102 may be correspondingly small. A larger aperture 102 would be required where the ion beam 34 is scanned during implantation. In a hybrid scanning implanter 10, one dimension of the aperture 102 is likely to be far wider than the other.
The beam stop 100 defines an internal volume 106 that extends from the entrance aperture 102 and is bounded by a side wall 108 and a back wall 110. A back plate 112 is attached to the back wall 110 in standard fashion, so as to face the entrance aperture 102. Thus, an ion beam 34 travelling as indicated by arrow A enters through the entrance aperture 102 and strikes the back plate 112. The back plate 112 is insulated from the rest of the beam stop 100 and is sized such that it covers the full extent of the ion beam 34.
The side wall 108 is optimised, or has an associated liner that is optimised, to absorb secondary electrons and other charged particles liberated from the back plate 112 by ion bombardment. The collar 104 defining the entrance aperture 102 may be similarly treated to ensure optimised absorption of ejected charged particles. A magnetic field, provided by permanent magnets or the like (not shown), may be provided across the entrance aperture 102 to suppress loss of charged particles from the beam stop 100 that might otherwise contaminate the wafer 12. The field also suppresses external electrons from entering the beam stop 100 and guides any higher-energy electrons to the side wall 108 rather than allowing them to reach the back plate 112 and provide a false reading. Thus, the back plate 112 provides an accurate measure of ions in the ion beam 34.
The back plate 112 generates a signal that is proportional to the incident ion beam current, as is well known in the art. This signal may be taken directly from the back plate 112 via an electrical connection 114. Alternatively, the signal may pass through preliminary signal processing in the beam stop 100, for example by circuit 116 shown in FIG. 2. The signal taken from the connection 114 is passed to the ion implanter controller.
FIG. 3 shows a front view of a first embodiment of a back plate 112. The back plate 112 has a circular cross-section and comprises a circular centre segment 110 surrounded by two annular segments 121 and 122, the segments 120 to 122 being separated by small gaps 130. The three segments 120 to 122 are arranged concentrically, and have equal areas in this embodiment. Each segment 120 to 122 is electrically isolated from the other two segments. A connection 118, such as those shown in FIG. 2, is provided to each segment 120 to 122 such that each segment 120 to 122 provides a signal indicative of the ion beam current incident thereon.
These signals may be used to derive information about the ion beam 34. For example, the signals may be represented as a histogram to indicate how the ion beam current decreases as a function of radial distance from the centre of the ion beam 34. As a wafer 12 is implanted by scanning along overlapping scan lines, it is important to know the radial profile of the ion beam 34. This includes both the width of the ion beam 34 and also the radial variation of the ion beam current. For example, the ion beam current may be characterised as a Gaussian. Also, knowledge of any asymmetries in the ion beam 34 is useful as they may then be corrected or compensated. Of course, total ion beam current can be found simply by summing the signals from the three segments 120 to 122.
FIG. 3 shows a circular back plate 112 with circular segments 120 to 122 that is best suited to characterising ion beams 34 with circular profiles. However, the ion beam 34 may not necessarily be circular, and the shape of the back plate 112 and its segments 120 to 122 may be altered accordingly. For example, FIG. 4 shows a back plate 112 that is identical to the back plate 112 of FIG. 2, other than that it has an elliptical shape to suit an elliptical ion beam. Thus, the back plate 112 is divided into an elliptical centre segment 120 surrounded by elliptical annular segments 121 and 122.
The number of segments 120 to 122 may be varied depending upon the amount of information required. FIG. 5 shows a circular back plate 112 divided into a circular centre segment 120 surrounded by four annular sections 121 to 124. Of course, any number of segments 121 to 124 may be chosen. Generally, the choice of segments 121 to 124 will be a compromise between the benefit of additional information and the penalty of more signal processing.
In addition to providing radial information on the ion beam profile, the back stop 100 can be used to provide centring information. This is most useful when performed with respect to the scanning directions, i.e. to provide centring information along the x and y axes in this embodiment. FIG. 6 shows an embodiment of a circular back plate 112 divided into two segments 120 and 121: a circular inner segment 120 and an annular outer segment 121. The inner segment 120 is further divided into two halves 120a and 120b by a vertical, electrically insulating gap 130. Each of the two halves 121a and 121b provides its own signal that is indicative of the ion beam current thereon.
Thus, the back plate 112 of FIG. 6 provides three signals. The two signals from the two halves 120a, 120b of the inner segment 120 may be compared to determine the centring of the ion beam 34 along the x axis. The measurements may be performed as the ion beam 34 is adjusted, i.e. use the measurements in a feedback loop to control steering of the ion beam 34. To centre the ion beam 34 on the beam stop 100, the signals from the two halves 120a and 120b should be equalised (assuming the response from each half 120a, 120b to be the same: the back plate 112 may be calibrated and differences in sensitivity compensated accordingly). Alternatively, the optimum ion beam position may not correspond to the centre of the beam stop 100, and so the ion beam 34 may be adjusted until a required offset is achieved between the segment halves 120a and 120b.
The signals from the two halves 120a and 120b of the inner segment 120 may be summed and compared to the signal from the outer segment 121 (or outer segments, where more than one are present) to allow radial profiling, as described previously.
FIG. 7 shows a variation of the embodiment of FIG. 6 where both inner and outer segments 120, 121 are divided into two halves. In addition to the centring and radial profiling already described, further analysis may be performed by comparing the radial profile of the left and right halves of the ion beam 34. Thus any asymmetries in the ion beam 34 may be detected, and also monitored during correctional reshaping of the ion beam 34 using the ion optics of the ion implanter 10.
FIG. 8 builds on the embodiment of FIG. 7 by further dividing the inner and outer segments 120, 121 with a horizontally extending gap 130. Thus the inner and outer segments 120, 121 are both divided into quarters, with each quarter providing its own signal (making eight in total). Thus, the back plate of FIG. 8 may be used to centre the ion beam 34 in both the x- and y-axis directions and also to detect asymmetries in both the x- and y-axis directions. It will be appreciated that this is achieved by summing adjacent quarters and comparing against the corresponding pair, e.g. 120a and 120c can be summed and compared with the sum of 120b and 120d to obtain x-axis centring.
FIG. 9 shows a beam stop 100 in section that broadly corresponds to that of FIG. 2. Accordingly, like reference numerals are used for like parts. In order to reduce further any problems associated with ejection of charged particles from the back plate 112, a baffle 150 is provided part way along the internal volume 106. Alternatively, two or more separate baffles 150 may be provided. The baffle 150 projects inwardly from the side wall 108 and presents a front face 152 (closest to the entrance aperture 102) and a back face 154. The back face 154 is optimised for absorbing ejected secondary electrons and other charged particles.
The baffle 150 forms a neck 156 in the internal volume 106 with a narrow bore such that many of the charged particles ejected from the back plate 112 are absorbed by the baffle 150 or the section of side wall 108a between the baffle 150 and back wall 110.
FIGS. 10 and 11 show a high-energy 160 and low-energy 162 ion beam respectively travelling through the beam stop 100. Divergence is less in the high-energy ion beam 160 where space charge effects have less time to act. Accordingly, the high-energy ion beam 160 is narrower and passes through both the entrance aperture 102 and the narrower bore 156 defined by the annular baffle 150. The blow-up of the low-energy ion beam 162 results in the edge of the ion beam 162 clipping the baffle 150. Accordingly, the baffle 150 creates a shadow downstream, although continued divergence of the ion beam 162 reduces the shadow size. As it is only the edges of the ion beam 162 that clip the baffle 150 where there is very little current, the problem of charged particle ejection is very much reduced compared to that from the back plate 112.
However, it is advantageous to measure the ion beam current incident on the baffle 150 to allow total ion beam current measurement and radial profiling. Moreover, the fact that the baffle 150 is offset along the z axis allows the ion beam trajectory and divergence to be measured. Thus the front face 152 of the baffle 150 is provided with a charge collecting plate 158 (hereinafter referred to as a baffle plate) from which a signal is provided that is indicative of the ion beam current incident thereon.
As the neck 156 formed by the baffle 150 is narrower than the entrance aperture 102, the back plate 112 may be smaller than those previously described. Although a single-piece back plate 112 may be used, it is preferred to use segmented back plates 112. FIGS. 12 and 13 show two such arrangements.
FIG. 12 is a front view of one embodiment of a vertically-divided back plate 112 with an annular baffle plate 158. The annular baffle plate 158 has a width W1, and defines an aperture 156 of width W2. The back plate 112 is circular and has a width W3. W3 is smaller than the overall width W4 of the internal volume 106, but larger than W2 to account for divergence of the ion beam 34. Again, total ion beam current may be found by summing all signals.
FIG. 13 is a front view of another embodiment of a beam stop 100 with baffles 158. The baffles 158 comprise two corresponding segments 158a and 158b (in the mathematical sense, i.e. they correspond to the area between an arc and chord of a circle, the arc formed by the side wall and the chord formed by a parallel edge). The edges 159 run vertically, as does the gap 130 that divides a circular back plate 112 into two halves 112a and 112b. As before, the back plate 112 can be used to centre the ion beam 34 in the x-axis direction. In addition, so too can the baffle plates 158a and 158b. Summing the signals from the baffle plates 158a and 158b and the back plate 112 provides the total ion beam current.
The baffles 158 of FIGS. 9 to 13 may be combined with the annular segmented back plate 112 of FIGS. 3 to 8. In particular, combining the annular baffle plate 158 of FIG. 12 with the annular segment back plate 112 is useful as it allows the trajectory and divergence of the ion beam 34 to be measured. This may be done conveniently by dividing the shadowed part of the back plate 112 (i.e. that having a width greater than W2 but less than W3) into annular segments 120, 121 and inspecting the ion beam current measured by each segment. This information allows the divergence and parallelism of the ion beam 34 to be controlled.
The person skilled in the art will appreciate that variations may be made to the above described embodiments without departing from the scope of the invention defined by the appended claims.
For example, the number, shape and size of the segments 120, etc. may be varied according to need (including the relative shape and size of the segments). Also, the division of segments 120, etc. into parts may also be varied. The shape of the beam stop may also be varied, rectangular or square and round or elliptical being particularly preferred cross-sectional shapes.
Patent Application
20080017811
