MARSH, Roy, Vincent (55 Freshfield Bank, Forest Row, East Sussex RH18 5HW, GB)
SIMMONS, Jonathon, Yancey (1430 Cherrydale Drive, San Jose, California, 95125, US)
BOYD, Wendell, Glen, Jr. (15203 Monticello Way, Morgan Hill, California, 95037, US)
MARSH, Roy, Vincent (55 Freshfield Bank, Forest Row, East Sussex RH18 5HW, GB)
SIMMONS, Jonathon, Yancey (1430 Cherrydale Drive, San Jose, California, 95125, US)
CLAIMS
1. A method of loading a substrate on an electrostatic chuck of a substrate holder, comprising: placing a substrate onto the chuck; supplying a first voltage to an electrode in the chuck thereby to establish and to maintain an electrostatic force due to attraction of the substrate to the chuck; and, subsequently, supplying a second voltage to the electrode greater than the first voltage thereby to establish and to maintain an increased electrostatic force.
2. The method of claim 1, comprising supplying the first voltage when the substrate is at rest .
3. The method of claim 2, comprising moving the substrate only when the second voltage is supplied to the electrode.
4. The method of claim 3, wherein the first voltage is sufficient to hold the substrate securely in place when the substrate is static but is not sufficient to hold the substrate securely in place when the substrate is scanned.
5. The method of any preceding claim, comprising gradually increasing the voltage supplied to the electrode to reach the first voltage.
6. The method of any preceding claim, further comprising supplying a coolant gas to the chuck thereby to cool the substrate, and then supplying the second voltage to the electrode .
7. The method of claim 6, comprising gradually increasing the voltage supplied to the electrode to reach the first voltage and gradually increasing the pressure of the coolant gas supplied to the chuck.
8. The method of claim 7, comprising increasing the voltage to reach the first voltage and then increasing the pressure of the coolant gas.
9. The method of claim 7, comprising increasing the voltage and the pressure of the coolant gas concurrently.
10. The method of any preceding claim, comprising gradually increasing the voltage supplied to the electrode to reach the second voltage.
11. A method of scanning a substrate through an ion beam in an ion implanter, comprising: placing a substrate onto an electrostatic chuck of a substrate scanner; supplying a first voltage to an electrode in the chuck thereby to establish and maintain an electrostatic force due to attraction of the substrate to the chuck; supplying a coolant gas to the chuck thereby to cool the substrate; supplying a second voltage to the electrode greater than the first voltage thereby to establish and maintain an increased electrostatic force; scanning the substrate through the ion beam; and decreasing the voltage supplied to the electrode.
12. The method of claim 11, further comprising gradually decreasing the supply of coolant gas to the chuck while decreasing the voltage supplied to the chuck.
13. The method of claim 12, comprising ensuring a supply of coolant gas to the chuck when the voltage supplied to the chuck reaches zero volts.
14. The method of any preceding claim, wherein supplying the first voltage to the electrode in the chuck further comprises supplying a complementary first voltage, being of the same magnitude but opposite polarity, to a further electrode in the chuck; and wherein supplying the second voltage to the electrode further comprises supplying a complementary second voltage, being of the same magnitude but opposite polarity, to the further electrode.
15. A method of unloading a substrate from an electrostatic chuck holding the substrate in place by virtue of electrostatic attraction arising from a voltage placed on an electrode of the chuck, the chuck also being provided with a coolant gas that provides a pressure acting to force the substrate from the chuck, the method comprising decreasing the voltage placed on the electrode to zero volts while decreasing the supply of coolant gas such that the supply of coolant gas is maintained when the voltage reaches zero.
16. The method of claim 15, comprising ensuring a supply of coolant gas at zero volts that is sufficient to stop the substrate sticking to the chuck.
17. The method of claim 15 or claim 16, wherein the substrate is held in place by virtue of complementary voltages placed on a pair of electrodes of the chuck, the method comprising decreasing the magnitude of the complementary voltages placed on the electrodes to zero volts .
18. The method of any preceding claim, further comprising using a feedback loop to control the voltage supplied to the electrode.
19. The method of claim 18, wherein using the feedback loop comprises measuring directly or indirectly the attractive force experienced by the substrate.
20. The method of claim 19, wherein measuring the attractive force experienced by the substrate comprises using one of: an optical method, a load cell, and a contact resistance method.
21. Apparatus comprising a controller arranged to implement the method of any preceding claim.
22. A computer programmed to implement the method of any of claims 1 to 20.
23. A computer program that, when executed, implements the method of any of claims 1 to 20.
24. A computer readable medium carrying thereon the computer program of claim 23. |
SECURING A SUBSTRATE TO AN ELECTROSTATIC CHUCK
Field of the Invention The present invention relates to securing a substrate to an electrostatic chuck to minimise damage to the substrate. In particular, the present invention relates to securing a substrate to an electrostatic chuck provided as part of a substrate scanner in an ion implanter.
Background of the Invention
Although the present invention is not limited to the field of ion implanters, this field corresponds to a contemplated application and provides a useful context for understanding the invention. Hence there follows a description of ion implanters.
Ion implanters are well known and generally conform to a common design as follows ' . An ion source produces a mixed beam of ions from a precursor gas or the like. Only ions of a particular species are usually required for implantation in a substrate, for example a particular dopant for implantation in a semiconductor wafer. The required ions are selected from the mixed ion beam using a mass-analysing magnet in association with a mass-resolving slit. Hence, an ion beam containing almost exclusively the required ion species emerges from the mass-resolving slit to be transported to a process chamber where the ion beam is incident on a substrate held in place in the ion beam path by a substrate holder. Often, the cross-sectional profile of the ion beam is smaller than the substrate to be implanted. For example, the ion beam may be a ribbon beam smaller than the substrate
in one axial direction or a spot beam smaller than the substrate in both axial directions. In order to ensure ion implantation across the whole of the substrate, the ion beam and substrate are moved relative to one another such that the ion beam scans the entire substrate surface. This may be achieved by (a) deflecting the ion beam to scan across the substrate that is held in a fixed position, (b) mechanically moving the substrate whilst keeping the ion beam path fixed or (c) a combination of deflecting the ion beam and moving the substrate. For a spot beam, relative motion is generally effected such that the ion beam traces a raster pattern on the substrate. To ensure good throughput of substrates through the ion implanter, the substrates are subjected to strong acceleration and deceleration forces as the substrate changes direction at the end of each scan line and as the substrate is stepped between scan lines.
Our US Patent No. 6,956,223 describes an ion implanter of the general design described above. A single wafer is held in a moveable substrate holder. While some steering of the ion beam is possible, the implanter is operated such that the ion beam follows a fixed path during implantation. Instead, the wafer holder is moved along two orthogonal axes to cause the ion beam to scan over the wafer following a raster pattern. The substrate holder may be provided with an electrostatic chuck to which the substrate is secured. Typically, the electrostatic chuck will comprise an electrode embedded in an insulator. A substrate is placed on the chuck and a voltage is applied to the electrode. This induces an accumulation of charge in the substrate such that the substrate is secured to the chuck by electrostatic attraction. In an alternative arrangement, two electrodes
are provided in the chuck and biased with opposite polarities. This bipolar arrangement has the advantage of not placing a net bias on the substrate and allows an overall reduction in the voltages applied. US 5,606,485 provides further details of electrostatic chucks.
The voltage applied to the electrode (s) in the chuck must be enough for the electrostatic attraction to secure the substrate firmly in position, particularly when being scanned through the ion beam where shifts in position will be at the expense of uniformity in the implant.
In addition, gas cooling of the substrate may be employed that is typically effected by providing the chuck with channels that open to the face of the chuck that supports the substrate. Gas is circulated through these channels at pressure to provide cooling to the substrate. Hence, the electrostatic attraction must also be enough to overcome the pressure of the gas that will try to force the substrate away from the chuck.
The large force necessary to secure a substrate firmly in place on an electrostatic chuck when faced with the large accelerations experienced during scanning can damage the substrate. In particular, this is seen for semiconductor wafer processing where defects and other damage is sometimes seen in the silicon on the back face of silicon wafers. Clearly, such structural damage has adverse effects on the performance of the high-value semiconductor devices formed on the wafers .
Summary of the Invention Against this background, and from a first aspect, the present invention resides in a method of loading a substrate on an electrostatic chuck of a substrate holder, comprising:
placing a substrate onto the chuck; supplying a first voltage to an electrode in the chuck to establish and to maintain an electrostatic force due to attraction of the substrate to the chuck; and, subsequently, supplying a second voltage to the electrode greater than the first voltage to establish and to maintain an increased electrostatic force.
Preferably, the method comprises preferentially supplying the first voltage when the substrate is at rest. Then the substrate may be moved only when the second voltage is supplied to the electrode. So, the first voltage may be supplied predominantly when the substrate is at rest, only being increased to the second voltage when the substrate is to be moved. Advantageously, the first voltage may be sufficient to hold the substrate securely in place when the substrate is static but may not be sufficient to hold the substrate securely in place when the substrate is scanned.
Hence, damage to the substrate may be minimised by applying a decreased voltage whenever the substrate is not being moved. Put another way, an increased voltage is only applied when the substrate is being moved. Thus, the present invention can also be viewed as a method of securing a substrate to an electrostatic chuck of a substrate holder comprising (a) generally, when the substrate is at rest, providing a voltage to an electrode of the chuck sufficient to secure the substrate in place with the substrate at rest but not sufficient to hold the substrate in place were the substrate to be moved; (b) immediately prior to moving the substrate, providing an increased voltage to the electrode sufficient to secure the substrate in place when the substrate is moved; (c) maintaining the increased voltage while the substrate is being moved; and (d) immediately
after the substrate has been moved, providing a voltage to the electrode sufficient to secure the substrate in place with the substrate at rest but not sufficient to hold the substrate in place were the substrate to be moved. Optionally, the method further comprises supplying a coolant gas to the chuck thereby to cool the substrate, and then supplying the second voltage to the electrode.
The present invention also resides in a method of scanning a substrate through an ion beam in an ion implanter, comprising: placing a substrate onto an electrostatic chuck of the substrate scanner; supplying a first voltage to an electrode in the chuck to establish and to maintain an electrostatic force due to attraction of the substrate to the chuck; supplying a coolant gas to the chuck thereby to cool the substrate; supplying a second voltage to the electrode greater than the first voltage to establish and to maintain an increased electrostatic force; scanning the substrate through the ion beam; and decreasing the voltage supplied to the electrode. As before, this method may follow the principle that the second voltage is applied whenever the substrate is being moved, and the first voltage is applied whenever the substrate is at rest save for the short periods immediately before and after movement when the voltage is being increased and decreased respectively.
Optionally, the method further comprises gradually decreasing the supply of coolant gas to the chuck while decreasing the voltage supplied to the chuck to ensure a supply of coolant gas to the chuck when the voltage supplied to the chuck reaches zero volts. The supply of coolant gas may be decreased in proportion to the decrease in voltage, or vice versa. Advantageously, by keeping a small supply of
gas to the chuck, the consequent pressure of the gas on the chuck can be enough to overcome a tendency for the substrate to stick to the chuck, even in the absence of a voltage on the electrode, as a result of charge accumulation on the wafer.
The present invention also resides in a method of unloading a substrate from an electrostatic chuck holding the substrate in place by virtue of electrostatic attraction arising from a voltage placed on an electrode of the chuck, the chuck also being provided with a coolant gas that provides a pressure acting to force the substrate from the chuck, the method comprising decreasing the voltage placed on the electrode to zero volts while decreasing the supply of coolant gas such that the supply of coolant gas is maintained when the voltage reaches zero. Preferably, the method comprises ensuring a supply of coolant gas at zero volts that is sufficient to stop the substrate sticking to the chuck. The exact pressure required may be determined as a matter of trial and error. Any of the above methods may further comprise using a feedback loop to control the voltage supplied to the electrode. In this way, the voltage supplied to the electrode may be checked and regulated according to any indication that is not at the desired level. For example, the feedback loop may comprise measuring directly or indirectly the attractive force experienced by the substrate. This is useful as it allows fine tuning of the voltage supplied to the electrode such that the desired force felt by the wafer that keeps it' secured to the chuck may be optimised. The attractive force experienced by the substrate may be measured by determining the deflection of
the substrate or the force exerted by the substrate on a part of the chuck.
This may be done in many different ways. For example a load cell may be used to measure the force applied by the substrate. Alternatively, a contact resistance method may be used to determine the force applied by the substrate. In a contemplated embodiment, a circuit may be formed using a pair of electrodes within the chuck, the substrate completing the circuit, such that the contact resistance can then be measured. The contact resistance may be determined by monitoring the current flowing through the circuit. Alternatively, the deflection of the substrate may be monitored such as by using optical techniques . Preferably, a laser source is used to illuminate the surface of the substrate with a laser beam and the reflected light beam is then detected. Various techniques may be used to determine the deflection of the substrate, for example using interference between the incoming and outgoing beam (whether that be laser light or otherwise) , or by looking at the changes in reflection intensity due to the angle of incidence of the beam changing with respect to the substrate surface as it deflects and bends.
The invention may also be used with bipolar electrostatic chucks having a pair of electrodes. In this instance, complementary voltages may be applied to the electrodes, i.e. voltages of equal magnitude but opposite polarity.
Brief Description of the Drawings In order that the present invention may be better understood, a preferred embodiment will now be described with reference to the accompanying drawings, in which:
Figure 1 is a schematic representation of an electrostatic chuck that may be used with the present invention;
Figure 2 is a schematic representation of an alternative form of electrostatic chuck that may be used with the present invention;
Figure 3 is a schematic representation of an ion implanter that may be used with the present invention;
Figure 4 is a block diagram illustrating a method of implanting a wafer according to an embodiment of the present invention;
Figure 5 is a graph showing the pressure on the substrate as electrode voltage and gas pressure are varied during loading of a wafer on an electrostatic chuck; Figure 6 shows the electrostatic chuck of Figure 2 further modified to include a load cell for measuring the attractive force on the wafer;
Figure 7 shows the electrostatic chuck of Figure 2 further modified to include an optical system for measuring the deflection of the wafer and, thereby, the attractive force on the wafer;
Figure 8 shows the electrostatic chuck of Figure 2 further modified to include a circuit for measuring contact resistance and, thereby, the attractive force on the wafer; Figure 9 corresponds to Figure 4, but shows a modification of the method to include a step of monitoring the attractive force on the wafer and compensating therefor;
Figure 10 corresponds to Figure 4, but shows a modification of the method to include further steps using force measurements to optimise the voltage used to secure the wafer to the electrostatic chuck;
Figure 11 corresponds to Figure 4, but shows a modification of the method to include simultaneous increase in chucking voltage and gas pressure; and
Figure 12 corresponds to Figure 5 and shows the pressure on the substrate during loading of a wafer on an electrostatic chuck according to the method shown in Figure 11.
Detailed Description of the Invention An electrostatic chuck 1 with which the present invention may be used is shown in Figure 1. The chuck 1 is shown holding a silicon wafer 12 in place and may be mounted to any suitable substrate scanner, such as a cantilevered arm. The chuck 1 comprises an insulating body, for example formed from Aluminium Nitride (AlN) . The chuck 1 contains a pair of electrodes 2 (only one electrode is visible in the section of Figure 1) positioned so as to be adjacent to the wafer 12. The electrodes 2 are connected to a power supply unit 3 by a cable 4 that provides positive and negative biases to the electrodes 2. The consequent charge accumulation in the adjacent back face of the wafer 12 results in an electrostatic attraction that urges the wafer 12 against the chuck 1. Supplying suitable voltages to the electrode 2 will see the wafer- 12 held firmly in position.
The chuck 1 is also provided with a gas coolant system comprising a closed loop flowing to and from a pressurised gas source 5. The. gas source 5 chills the gas as it circulates around the closed loop. The closed loop takes chilled gas to the chuck 1, circulates the chilled gas around the chuck 1 thereby to cool the wafer 12, and takes the warmed gas back to the gas source 5 to be chilled once
more. As can be seen from Figure 1, the closed loop comprises conduits 6 to take gas to and from the chuck I 7 channels 7 provided in the chuck 1 for circulating the gas around the chuck 1, and outlets 8 that allow the chilled gas to contact the wafer 12 and hence remove heat through conduction.
An alternative form of electrostatic chuck 1 is shown in Figure 2. The chuck 1 is of a broadly similar design to that of Figure 1, and like reference numerals are used to refer to like parts. The electrostatic chuck 1 of Figure 2 supports the wafer 12 on an array of mesas 9. These mesas 9 define valleys 10 therebetween. The cooling gas is passed through the conduits 7 that open to the valleys 10 at their outlets 8. However, in alternative arrangements, the outlets 8 are provided at the top of the mesas 9 (both in addition and as an alternative to outlets 8 placed in the valleys 10) .
As mentioned above, the present invention may find application in an ion implanter, although it is to be understood that the invention is not limited to such use. The following description of an ion implanter is not intended to be limiting, but will provide a useful context to aid in the understanding of the present invention.
Figure 3 shows a conventional ion implanter 11 for implanting ions in semiconductor wafers 12. The ion implanter 11 comprises a vacuum chamber 15 evacuated by pump 24. Ions are generated by an ion source 14 to be extracted and follow an ion path 34 that passes, in this embodiment, through a mass analysis stage 30. Ions of a desired mass are selected to pass through a mass-resolving slit 32 and then to strike the wafer 12. The ion source 14 generally
- ii -
comprises an arc chamber 16 containing a plasma for generating the desired ions.
Ions from within the arc chamber 16 are extracted using a negatively-biased (relative to ground) extraction electrode 26. The mixture of extracted ions are then passed through the mass analysis stage 30 so that they pass around a curved path under the influence of a magnetic field. The radius of curvature travelled by any ion is determined by its mass, charge state and energy, and the magnetic field is controlled so that, for a set beam energy, only those ions with a desired mass to charge ratio and energy exit along a path coincident with the mass-resolving slit 32. The emergent ion beam is then transported to the process chamber 40 where the target is located, i.e. the wafer 12 to be implanted or a beam stop 38 when there is no wafer 12 in the target position.
The semiconductor wafer 12 is mounted on an electrostatic chuck 1 of the wafer holder 36, wafers 12 being successively transferred to and from the wafer holder 36, for example through a load lock (not shown) . The chuck
1 may correspond to the one shown in Figure 1 or the one shown in Figure 2.
The ion implanter 11 operates under the management of a controller, such as a suitably programmed computer 50. The controller 50 controls scanning of the wafer 12 through the ion beam 34 to effect desired scanning patterns. These scanning patterns may comprise raster scans, including interlaced patterns, as is well known in the art.
Figure 4 presents a method of scanning a wafer 12 through an ion beam 34 using the ion implanter of Figure 2, including loading the wafer 12 on an electrostatic chuck 1, in accordance with a first embodiment of the present invention.
This method may be implemented using the ion implanter 11 of Figure 3 , including the chuck 1 of Figure lor Figure 2. In particular, the method may be implemented by the controller 50. The method comprises a two-step process for loading the wafer 12 onto the chuck 1, as opposed to the conventional one- step process. To illustrate the advantages of the present invention, Figure 5 shows the pressures felt by the wafer 12 during both one-step and two-step processes. A single vertical marker is used in Figure 5 to denote the one-step process, whereas a double vertical marker is used to denote the two-step process.
At 52, the controller 50 issues instructions such that the wafer holder 36 is rotated to present the chuck 1 horizontally for accepting a wafer 12. At 54, the controller 50 controls a wafer-handling robot (not shown in Figure 3) to load a wafer 12 onto the chuck 1. Once released, the wafer 12 is held on the chuck 1 by gravity. Guides or indicia may be provided on the chuck 1 to facilitate correct placement of the wafer 12 on the chuck 1. In the timeline of Figure 5, the start point shown as t 0 corresponds to completion of step 54, i.e. with the wafer 12 placed on the chuck 1.
At 56, the chucking voltage applied to electrodes 2 is ramped up to a low level between times t 0 and t x , as indicated at 76 in Figure 5. This is the first part of a two-step process according to a first embodiment of the present invention. Figure 5 also shows the one-step process currently practiced. As can be seen, step 56 of the present invention sees the chucking voltage on the chuck's electrodes 2 set to a lower level than for the one-step process. This leads to a reduced pressure being felt by the wafer 12 at time ti relative to the one-step process.
At 58, the gas pressure of the coolant gas is increased. As can be seen at 78 in Figure 5 between times ti and t 2 , the increase in gas pressure is similar for both the one-step and two-step processes. The chucking voltage placed on the electrodes 2 is kept constant during this phase. Hence, in both processes, the pressure felt by the wafer 12 decreases as the gas pressure acts to force the wafer 12 away from the chuck 1. For the two-step process, the reduced chucking voltage applied at 56 must be sufficient to keep the wafer' 12 securely in place as the gas pressure is increased at 58.
At 60, the chucking voltage applied to the electrodes 2 is increased for the two-step process as shown at 80 in Figure 5 between t 2 and t 3 . During this period, the chucking voltage is kept constant in the one-step process. In fact, the chucking voltage is increased in the two-step process to match that used in the one-step process at time t 3 . Of course, the pressure felt by the wafer 12 in the two-step process increases as the chucking voltage is increased. This increase is designed to allow the wafer 12 to be moved, i.e. the pressure felt by the wafer 12 between to and t 3 is sufficient to keep the wafer 12 in place provided the wafer 12 is not moved, whereas at t 3 and beyond the pressure is sufficient to keep the wafer 12 in place even when moved through the ion beam 34.
As can be seen from Figure 5, although the pressure felt by the wafer 12 is the same at t 3 whether the one-step or two-step process is followed, the pressures felt by the wafer 12 is lower at all times from t 0 through to t 3 for the two-step process. Hence, the wafer 12 is less likely to be damaged in this two-step loading process.
With the loading process complete, the chuck 1 is rotated at 62 so as to bring the wafer 12 to vertical ready for scanning. Scanning is then effected at 64 by moving the wafer 12 through the ion beam 34 to complete an implant according to the desired scan pattern. Once an implant is complete, the chuck 1 is rotated to the horizontal once more, as indicated at 66, so as to be ready for unloading the wafer 12. In order to minimise the pressure experienced by the wafer 12, the chucking voltage on electrode 2 is ramped straight down to zero at 68. In addition, the gas pressure is ramped down concurrently at 70. These two steps are co-ordinated such that drop in gas pressure shadows the drop in chucking voltage with the result that the slight pressure exerted by the gas on the wafer 12 stops the wafer 12 sticking to the chuck 1 (this effect is otherwise common because of residual charge accumulation such as from contaminants) .
At 72, the wafer 12 is removed from the chuck 1 such that the chuck 1 is ready for another wafer 12 to be loaded. The chucking voltage required to produce a desired attractive force on the wafer 12 may be readily determined. For example, values may be determined by a manufacturer and the ion implanter 11 supplied with these values stored in memory for access by the controller 50. However, manufacturing tolerances in making the chuck 1 mean that predetermined voltage values are unlikely to be optimal for any particular chuck 1. The variations that have the greatest effect are seen in the heights of the mesas 9, the depth of the electrode 2 within the chuck 1 , and the resistivity of the AlN used to form the chuck 1. As a result, typical voltage values used are on the high side to ensure very few wafers 12 slip or are dropped during
scanning. This means that the chucking voltages used are higher than necessary for the majority of chucks 1, providing the potential for more wafers 12 to be damaged.
A further consideration is that the required chucking voltage can vary over the lifetime of the chuck 1 due to changes in the condition of the chuck surface.
The chuck 1 described above may be modified to address these issues by including a monitoring system to determine the attractive force felt by the wafer 12. Thus, a feedback loop may be used to provide a check that the chucking voltage applied to the electrodes 2 in the chuck 1 is producing the desired attractive force. Moreover, the feedback loop allows the chucking voltage to be adjusted until the desired attractive force is achieved. The wafer 12 may be monitored in many different ways to ascertain the attractive force, and three exemplary embodiments are shown in Figures 6 to 8.
Figure 6 shows a further embodiment of an electrostatic chuck 1 for use with the present invention. The chuck 1 is broadly similar to the chuck 1 of Figure 2, and like reference numerals refer to like parts. The chuck 1 is modified to include one or more load cells 90 positioned within the valley 10 between mesas 9 to measure the deflection of the wafer 12. As will be appreciated, the attractive force felt by the wafer 12 causes it to deflect in the regions overlapping the valleys 10. Preferably, the load cells 90 are positioned centrally within the mesas 9, where wafer deflection is greatest. The height of the load cells 90 is such that they terminate at or just below the plane occupied by the back face of the wafer 12. Deflection of the wafer 12 causes a force to be applied to the load cells 90. The load cells 90 may be used to measure this
force and, where correct calibration is employed, to deduce the attractive force felt by the wafer 12.
The signals provided by the load cells 90 are provided to the controller 50 via connections 92 that pass through and then out of the chuck 1. The controller 50 reads these signals, assesses their values and controls the power supply unit 3 in response thereto (via connection 94) . For example, if the controller 50 determines that the signals have too high a value, the controller 50 may cause the power supply unit 3 to provide a lesser chucking voltage to the electrodes 2. This determination may be performed in numerous ways, e.g. direct comparison of the signal values to expected values, or after conversion to other quantities such as the attractive force felt by the wafer 12. A single load cell 10 may be used, or multiple load cells 10 may be used to allow redundancy such that only a single value is used, checking to ensure consistency between values or averaging where multiple values are used.
The embodiment of Figure 7 operates on the same principle of measuring deflection of the wafer 12 into the valleys 10 between the mesas 9. However, an optical method is used in place of the load cell method of Figure 6. The optical method of Figure 7 comprises a light source 100, such as a laser, that operates under the control of the controller 50. The light source 100 sends a beam of light 102 to the surface of the wafer 12 above a valley 10 where the deflection is greatest. The reflected light beam 102 is received by a photodetector 104 or similar whose signal is provided to the controller 50. The controller 50 may deduce the deflection of the wafer 12 from the received signal, and hence the attractive force felt by the wafer 12. The deflection may be deduced
in any standard way, e.g. by using interference between the incoming and outgoing light beams 102. The controller 50 is connected to the power supply unit 3 via connection 106, such that the controller may control the chucking voltage placed on the electrodes 2, as has already been described in respect of Figure 7.
Figure 8 shows a further embodiment of the electrostatic chuck 1 that may be used with the present invention. The attractive force felt by the wafer 12 is determined by measuring the electrical contact resistance between the wafer 12 and mesas 9. To this end, a circuit is provided that comprises a constant voltage source 110, an electrical current measuring device 112, a connection to the controller 114 and a pair of electrodes 116 that are located within mesas 9. The wafer 12 completes the circuit. As the attractive force on the wafer 12 varies, the electrical contact resistance between the wafer 12 and the electrodes 116 varies, and so the current measured by the current measuring device 112 varies. Accordingly, the controller 50 can deduce the attractive force and control the power supply unit 3 via connection 118, as has been described previously.
The chucking voltage provided to the electrodes 116 must be small enough not to interfere with the chucking voltage provided to the electrodes 2, but large enough to provide a measurable current. A voltage of 10-50V has been found to give good results.
The method of Figure 4 may be adapted to include feedback control of the chucking voltage in many different ways. Figure 9 shows one such adaptation where a continual or near-continual monitoring method is employed, as shown figuratively at 120. The monitoring method may be any of those described above .
According to this method, as soon as a chucking voltage is provided to the chuck 1 at step 56, the feedback system begins operation. The feedback system sees the controller 50 check the chucking voltage at frequent intervals to ensure that it compares well to the expected value, and to adjust operation of the power supply unit 3 if it does not. This monitoring continues throughout the two-step chucking process (steps 56-60), during scanning at steps 62 and 64, and during the ramp-down of the chucking voltage at step 70. The feedback system only ends its operation when the chucking voltage has reached zero.
A further embodiment is shown in Figure 10. When compared with Figure 4, it will be seen that two further steps are introduced at 57 and 61. At step 56, the chucking voltage is first ramped up to a low level. Once this is complete, a feedback system (such as any of those described above) checks the attractive force felt by the wafer 12 at step 57. If the force measured does not compare favourably to the desired force, the controller 50 adjusts the power supply unit 3, iteratively until the attractive force measured corresponds to the desired value. Once the desired force is measured, the method continues to step 58, as per Figure 4.
At step 60, the chucking voltage is ramped up to a high level. Once this is complete, step 61 begins. Step 61 effectively corresponds to step 57 in that the attractive force is measured and, if required, the power supply unit is adjusted iteratively until the correct force is obtained. Further steps like 57 and 61 may be added to the method. Moreover, a method that uses features from both
Figures 9 and 10 may be used, i.e. there may be periods of continual monitoring like that of Figure 9, while other
periods may be subject to one-off checks like those of Figure 10.
As will be clear to the skilled person, variations may be made to the embodiments described above without departing from the scope of the invention as defined by the appended claims .
For example, the exact implementation to cause electrostatic attraction between the wafer 12 and chuck 1 is not critical. The present invention is equally well suited to work with single electrode chucks and bipolar electrode chucks .
Figure 4 shows a process where the voltage placed on the electrodes 2 are increased between t 0 and ti before the gas pressure is increased between t x and t 2 . However, the voltage and the gas pressure may be increased concurrently, for example so that they increase proportionately. Thus, an end point will be reached where the coolant gas is supplied and the lower voltages are applied to the electrodes 2 such that the wafer 12 is held in position securely provided the wafer 12 is not moved. Then, at any time, the voltages applied to the electrodes 2 may be increased prior to the wafer 12 being moved.
Alternatively, the gas pressure may be increased at the same time as the chucking voltage is increased at step 60. Such a method is shown in Figures 11 and 12. As can be seen, the gas pressure is increased at step 58 at the same time as the chucking voltage is increased at step 60. Moreover, Figure 12 shows that this is done such that the attractive force felt by the wafer 12 remains essentially constant. In more detail, the period between to and ti sees the voltage placed on the chuck 1 increase to a low level, in accordance with step 56. Between ti and t 2 , the chucking
voltage placed is increased, as shown by line 76 and, at the same time, the gas pressure is increased as shown by line 78. As a result, the attractive force felt by the wafer 12 increases between t 0 and t x , and then remains constant between ti and t 2 .
Conveniently, a monitoring system like those shown in Figures 6 to 8. may be employed to ensure that the attractive force felt by the wafer 1 remains essentially constant between ti and t 2 . As shown in Figure 11 at 120, a monitoring system is used continuously or near-continuously to ensure that the attractive force felt by the wafer 1 corresponds to the desired level, as has been described previously with respect most particularly to Figure 9. This continuous monitoring may be used to ensure that the gas pressure and chucking voltage rise in tandem, as required to maintain a constant force on the wafer 1.
As a further modification, the de-chucking procedure may be varied. For example, the chucking voltage may be decreased first without a reduction in gas pressure, i.e. step 70 may be started before step 68. The gas pressure may be decreased only when the chucking voltage has reached zero. Alternatively, the wafer 12 may be monitored to determine when it "pops off" the chuck 1 during the decrease in chucking voltage, and the gas pressure reduced once the pop off has been seen. The pop off may be detected by, for example, by measuring the capacitance between the wafer 12 and chuck 1 or by any of the monitoring systems of Figures 6 to 8.
The inclusion of a coolant gas supply is not necessary. Where gas cooling is not used, the same principle of applying a lesser voltage at times when the. wafer 12 is not being moved may be used. Cooling of the chuck 1 may be
effected in other ways. For example, liquid cooling may be used. This liquid cooling may see a recirculating chilled water supply routed through the chuck 1 using pipes. Alternatively, the water may be piped through a supporting structure where the supporting structure is in thermal contact with the chuck 1. Either way, the chuck 1 will be kept cool and, in turn, the wafer 12 may be kept cool through heat conduction. Where the wafer 12 is supported on mesas 9, the gas supply still helps in that gas fills the valleys 10 between the mesas 9 and helps heat conduction from the wafer 12 to the chuck 1.