The present invention relates to substrate carriers for use in plasma processing apparatus and the like, particularly "sandwich"-type substrate carriers. Methods for manufacturing substrate carriers and methods for mounting a substrate are also disclosed.

High throughput requirements for plasma processing combined with commercially viable substrate sizes create the need for high etch and deposition rates (very high etch or deposition rate during single wafer processing or medium etch or deposition rate for batch processing). Such high rate processes tend to heat the substrate (or "wafer"), which must therefore be cooled to maintain the required process temperature. This is achieved by promoting heat transfer between the substrate and the surface on which it is supported during processing, such as a wafer table. Typically this is achieved by supplying a heat transfer fluid to the rear surface of the substrate, such as helium gas, which has the effect of raising the pressure behind the substrate. The resulting pressure is typically in the range 5 to 26 mbar. Where the pressure in the processing chamber is less than the pressure behind the substrate, the substrate must be secured into position to prevent the pressure difference blowing the wafer off the supporting surface. Various techniques for securing substrates have been disclosed and fall generally into two categories. First, electrostatic chucks (usually formed integrally with the substrate table of the processing chamber) are a proven technique for holding single conductive, semi-conductive or metal-coated substrates. Here, clamping is achieved by electrostatic attraction between the substrate and the carrier. Some examples of electrostatic chucks are given in US-B-6344105 and US-A- 2003/0219986. However, electrostatic chucks are relatively complex, expensive devices and are also less suitable for securing dielectric substrates, such as sapphire, since a very high voltage must generally be applied in order to attain a sufficient clamping force. Electrostatic chucks have been proposed for clamping multiple substrates in a batch (see for example WO-A-2007/043528), but the increased risk of any one or more of the wafer holding positions failing renders the high cost of the system unattractive.

The alternative is to secure the substrate(s) mechanically, which is typically achieved by holding the substrate(s) between lower and upper components; hence the resulting assembly is typically referred to as a "sandwich" carrier. Such constructions can be used for substrates of any material and can be configured for single wafer processing or to carry multiple substrates simultaneously. Sandwich carriers also have the advantage that different substrate sizes can be accommodated in the processing chamber straightforwardly, simply by selecting the appropriate carrier (or, depending on the design of the carrier, simply by changing its top plate), without requiring any changes to the hardware (e.g. wafer table or table clamp) forming part of the processing chamber itself.

A typical example of a sandwich carrier of the type well known in the industry is depicted in Figures 1 and 2 hereto. Figures 1 a and 1 b show, respectively, a plan view and a cross-sectional view of a sandwich carrier clamped via a clamp ring 5 to an electrode or wafer table 7 in a plasma processing chamber (not shown). A number of substrates 4 are placed between a top plate 1 and a bottom plate 2. The top plate 1 has apertures 10 each of which exposes one substrate 4 to the processing chamber, with protrusions 11 of the top plate extending over portions of the substrates to retain them in place. The bottom plate 2 has channels 2a therethrough for delivery of a heat transfer fluid to the rear of each substrate 4, and seals 3 such as O-rings adjacent the periphery of each substrate to reduce or prevent leakage of the fluid into the processing chamber. A seal 6 is also provided between the underside of the bottom plate 2 and the wafer table 7. Substrates 4 become fully secured due to bolted connections (not shown) between the top plate 1 and the bottom plate 2 and/or may rely on the clamping force applied via clamp ring 5 to the carrier while in the process chamber.

The requirement to have one or more protrusions of the top plate extending over each substrate inherently reduces the useable surface area of each substrate. However, this problem is exacerbated by the formation of an exclusion zone around the perimeter of the substrate where the process conditions are nonuniform. This is illustrated in Figure 2 for the carrier of Figure 1 , assuming standard conditions typically used for plasma etching of a sapphire wafer (high density plasma, high DC bias and low pressure). The presence of the top plate 1 adjacent the substrate affects the geometry of the plasma sheath 8 which forms above the substrate. As shown in Figure 2a, which is a cross section along line Y-Y' shown in Figure 1 a, adjacent the edge of aperture 10 in top plate 1 , the sheath 8 undergoes localised deformation. In the region marked 8a this results in a decreased density of ions, whereas an increased density of ions reaches the surface of the top plate (8b). The region of decreased ion density 8a includes an edge area E of the substrate 4. Additionally, the sheath deformation promotes ion bombardment deviating from the desired direction (perpendicular to the substrates' top surface). As a result, features etched in this region E will be tilted. The region E is referred to as the exclusion zone since in practice, features etched or deposited in this region will generally be unusable. Some examples of related art are disclosed in US-A-4400235, US-A-5421401 , US-A-6123804, US-A-5660673 and US-A-2010/0162957. Figure 2a discussed above shows the exclusion zone E at a location in between protrusions 1 1 , where the top plate 1 does not itself extend over the substrate 4. The exclusion zone E extends yet further into the interior of the substrate at the positions of the protrusions 1 1 , as illustrated in Figure 2b, which is a cross section along line Z-Z' in Figure 1a.

In typical sandwich carrier constructions, even in the intervals between protrusions 11 , the exclusion zone E can extend several millimetres onto the substrate 4, which can be a substantial proportion of the total substrate area. For example, exclusion zones of 3 mm or more are typical. It would be desirable to reduce the exclusion zone in order to increase the usable surface area of (and hence ultimately the yield of good devices from) the substrate.

The present invention provides a substrate carrier for plasma processing apparatus, the substrate carrier comprising: a base plate assembly comprising a platform on which a substrate is placed in use, the volume to be occupied in use by a substrate placed on the platform constituting a substrate zone; and

a top plate assembly on the base plate assembly, the top plate assembly comprising a top plate having an aperture therethrough surrounding the substrate zone and one or more retention members extending into the aperture, over a portion of the substrate zone, such that in use a substrate disposed in the substrate zone is retained in a fixed position between the base plate assembly and the top plate assembly;

wherein the upper surface of the top plate, facing away from the base plate assembly, is substantially planar and the one or more retention members protrude above the substantially planar upper surface of the top plate.

By arranging the one or more retention members to protrude above the substantially planar upper surface of the top plate (in other words, reducing the upper surface level of the top plate relative to that of the retention members), the degree of sheath deformation can be substantially reduced. In particular, only the retention members themselves need protrude above the height of the wafer surface and the present invention enables the surrounding portions of the top plate to be reduced in height. As such, the exclusion zone can be significantly reduced in size, or eliminated entirely, in the intervals between retention members, thereby increasing the usable surface area of the substrate. Preferably, a plurality of retention members are provided which are spaced from one another about the periphery of the aperture. In the intervals between retention members there is nothing extending over the substrate and hence here it is fully exposed to the reaction chamber. Any number of retention members could be provided, most preferably at least three. Advantageously the retention members are evenly spaced about the periphery of the aperture. The retention members are configured to abut and apply pressure to the upper surface of the substrate in use, to hold it in position.

Any degree of reduction in the height difference between the surrounding upper surface of the top plate and the substrate surface (relative to the height difference between the top of the retention members and the substrate surface) will reduce the exclusion zone. In some cases there may be an optimum height difference value which is non-zero. However in particularly preferred embodiments, in use the top surface of the substrate is level with the top surface of the carrier. For example, this has been found by the present inventors to produce very good results in typical sapphire etch processes. Hence, preferably, the upper surface of the top plate is substantially coplanar with the upper surface of the substrate zone. Any feature protruding above the level of the substrate surface (such as the retention member(s)) produces disruption to the plasma sheath and thus, preferably, the number and size of such features should be kept to a minimum. By positioning the upper surface of the top plate at the same level as the upper surface of the substrate, the geometry of the plasma sheath is substantially undisturbed resulting in substantially uniform processing conditions across a larger proportion of the substrate. By "substantially coplanar", it is preferred that the upper surface of the top plate is coplanar with the upper surface of the substrate zone to within +/- 1 mm, preferably +/- 0.75 mm, still preferably +/- 0.5 mm. Similarly it is preferred that the upper surface of the top plate is parallel to the upper surface of the substrate zone (and/or hence to the platform, which will generally define the orientation of the substrate).

In preferred embodiments where the upper surface of the top plate is substantially coplanar with the upper surface of the substrate, the undersides of the retention members (which contact the substrate in use) are substantially level with the upper surface of the top plate, in use. Again, here "substantially level" preferably means to within +/- 1 mm, advantageously +/- 0.75 mm, still preferably +/- 0.5 mm.

Typical substrates have thicknesses ranging between 0.5 and 2mm, more usually between 0.4 and 1.5 mm, and it is advantageous if the level of the upper surface of the top plate sits within one substrate thickness, more preferably half the substrate thickness above or below the upper surface of the substrate, most preferably at equal height as mentioned above. Hence it is preferred that the upper surface of the top plate is at a height of between zero and 4 millimetres, preferably between 0.25 and 3 millimetres, more preferably between 0.4 and 2 mm, still preferably between 0.4 and 1.5 mm from the upper surface of the platform depending on the substrate in use. The plasma sheath may also be deformed by apparent depressions in the substrate / top plate upper surface level such as may be formed by any gap between the top plate and substrate. Hence it is preferred that the aperture is sized such that any lateral gap between the top plate and the substrate zone is 2 millimetres or less, more preferably 1 mm or less. In this way, sheath deformation is further reduced as is the resulting exclusion zone.

The top plate assembly can be implemented in a variety of ways. In one embodiment, the retention member(s) are an integral part of the top plate and in fixed relation to one another. For example, the top plate may be moulded or machined to incorporate both the retention members and a recessed top plate upper surface as described above.

However, in particularly preferred examples, the top plate and the retention member(s) are flexibly coupled relative to one another. This reduces the risk of damage to substrates occurring during loading or unloading of the carrier since only the retention members contact the substrate, with the clamping force from the top plate being applied indirectly via the flexible coupling. Since the top plate may typically be subject to a clamping force which varies across its area, the top plate may suffer from localised deformations which, if transferred to the substrate, increase the risk of breakage due to the creation of stress concentrations. Similarly, if the clamping structure were to come into point or line contact (as opposed to planar contact) with the substrate during loading or unloading of the carrier - the risk of which is increased by deformation of the top plate - breakages may occur. By partially decoupling the clamping force applied to the substrate(s) from that applied by the top plate, which is achieved by providing a flexible coupling between the top plate and the retention members which contact the substrate in use, such risks are greatly reduced since motion or deformation of the top plate relative to the substrate is absorbed or modified (e.g. in terms of intensity and/or direction by the flexible coupling before transmittal to the substrate).

It is further preferred that, where the carrier is configured to hold a plurality of substrates (discussed further below), the respective retention member(s) for each of the substrates can also move relative to the top plate independently of the retention member(s) for the other substrates. In this way the force ultimately applied to each substrate can be varied by the flexible connections to account for non-uniformities in the pressure applied by the top plate. Still further, each of the retention members may be independently mounted so that even those members arranged to retain the same substrate provide independent adjustment.

Flexible couplings of the sorts mentioned above can be achieved in a number of ways. In some preferred embodiments, the retention member(s) are carried on the top plate via resilient connection(s). This reduces the overall part count and may allow for faster assembly since the top plate and retention members can be placed in position in one action. However, particularly where the carrier is designed to hold multiple substrates (for batch processing) this would require the substrates to be positioned with high accuracy on corresponding platforms prior to the placement of the top plate assembly, with any misalignment of the substrates giving rise to a risk of damage.

As such, in particularly preferred examples, the retention member(s) are carried by a retention body which is separate from the top plate, the retention member(s) preferably forming an integral part of the retention body. The retention body can then be positioned in alignment with the substrate before the top plate is provided to hold the retention body in position. The flexible coupling between the top plate and retention members could take the form of a flexible interconnecting component between the top plate and the retention body (e.g. a spring), but more preferably is incorporated into the retention body itself.

Thus, in a preferred embodiment, the top plate is configured to contact the retention body at one or more contact portions of the retention body which are resiliently connected to the retention member(s) such that urging of the top plate towards the base plate assembly causes elastic deformation of the resilient connection. For example, the contact portion(s) may be located on spring beam parts of the retention body which will undergo elastic deformation relative to the retention members upon contact from the top plate.

The retention body could comprise a single retention member but advantageously the retention body carries a plurality of spaced retention members, preferably equally spaced from one another. For example the retention body could have 4 retention members spaced by 90 degrees from each other, 6 retention members spaced by 60 degrees from one another, or any other number. Typically the larger the substrate, the greater the number of retention members that will be required to secure it. Should the substrate have particular areas where contact is preferred, the retention member positions can be designed for that purpose.

Where the retention body comprises a plurality of retention members, each one preferably has a resilient connection with the one or more contact portions which is substantially independent of the resilient connections to the other retention members. In this way, should the force applied to the contact portions by the top plate vary spatially, the force will be transmitted to the retention members by the various resilient connections to varying degrees, reducing the effect of the variation and minimising stress concentrations on the substrate. In some advantageous embodiments, the retention body comprises a flange disposed between the top plate and the base plate assembly such that in use the flange is urged by the top plate towards the base plate assembly (or, analogously, by the base plate assembly against the top plate), and there is a resilient connection between the flange and the retention member(s). Here, the flange provides the one or more contact portions discussed above. In a first preferred implementation, the resilient connection comprises a joint between the flange and the retention body, the flange being pre-stressed towards the top plate such that compression of the top plate against the base plate assembly causes elastic deformation of the flange about the joint. It should be noted that the flange may be an integral part of the retention body with the joint being formed by a fold line in the material of the retention body defining the flange. In another preferred implementation, the resilient connection comprises an extension spring extending between the flange and the retention member, the extension spring preferably forming an integral part of the retention body. The top plate may include recesses or protrusions arranged on its underneath surface configured to accommodate the retention body and/or to contact it at selected positions. The retention body could take any shape but preferably the retention body is shaped so as to substantially follow the periphery of the substrate zone. That is, for a circular substrate the retention body preferably has a circular cross-section (as discussed below) whereas for a square or rectangular substrate the retention body will similarly be square or rectangular in cross-section. Most typical substrates are substantially circular (although may for instance be provided with a flat edge) and hence in preferred embodiments, the retention body substantially cylindrical or frusto-conical (thus having a circular cross section), the one or more retention members extending inwards towards the interior of the cylinder or frustum and any flange(s) (if provided) extending outwards away from the interior of the cylinder or frustum. By arranging the retention body to be of the same shape as the substrate, placement of the retention body on to the substrate can assist in centering the substrate in its correct position prior to clamping with the top plate. Where there is a lateral gap between the top plate and the substrate zone, in preferred embodiment this gap is at least partially filled by the retention body. As such, in the intervals between the retention members, preferably the upper surface of the retention body is substantially level with the upper surface of the top plate. This assists in presenting a substantially flat surface to the processing chamber and hence further reduces deformation of the plasma sheath.

Preferably the resilient connection(s) between the contact point(s) and the retention member(s) share the same symmetry as that of the substrate zone, preferably rotational symmetry. Since it is the force resulting from deformation of the resilient connection(s) which is transferred to the substrate via the retention members, due to the symmetry of the spring arrangement in the retention body, the majority of force components which are not perpendicular to the substrate's surface will cancel out, drastically reducing the possibility of the retention members sliding on substrates during installation and potentially causing damage.

Since it is necessary for the retention member(s) to extend over the substrate in order to secure it, it is not possible to entirely eliminate the exclusion zone in the vicinity of each retention member. However, preferably the or each retention member has a maximum height in the direction parallel to the normal of the substrate zone of 3 millimetres or less, more preferably 2 millimetres or less. In this way the extent of the exclusion zone around the retention members can be minimised.

The lateral extent of the retention member(s) is also preferably kept to the minimum necessary to secure the substrate in place. Preferably, the or each retention member extends over the substrate zone by 3 millimetres or less, preferably 2 millimetres or less, from the top plate. In particularly preferred examples, the total area of the substrate zone covered by the one or more retention members is 2% or less of the full area of the substrate zone, preferably 0.5% or less.

The base plate could consist solely of the platform with the top plate assembly configured to fit around or on top of the platform. However, in more preferred examples the platform is a defined region forming part of a base plate which extends under the top plate outside the platform region. The platform could be level with the rest of the base plate but in particularly preferred examples, the platform is raised relative to its surroundings on the base plate, and preferably extends into the aperture provided in the top plate. This arrangement permits the use of a thicker top plate, which is more robust and capable of applying a greater and more uniform clamping force, whilst still keeping its upper surface at substantially the same level as that of the substrate (as is preferred for reasons discussed above). Advantageously, the platform is of substantially the same lateral shape as that of the substrate zone. Where the retention member(s) are carried on a retention body, preferably the platform is surrounded by one or more recesses in the base plate adapted to accommodate the retention body. This also assists in correctly locating and seating the retention body on the base plate during assembly of the carrier, which ultimately improves alignment of the substrate.

As discussed above, etching and deposition processes tend to heat the substrate during processing and so it is strongly desirable to be able to remove heat from the wafer in a spatially uniform manner in order to maintain consistent processing conditions. Therefore, in particularly preferred embodiments, the platform has one or more channels therethrough, to permit the delivery and/or removal of heat transfer fluid, such as helium gas, to and/or from the underside of a substrate disposed in the substrate zone in use. This significantly reduces the thermal resistance between the substrate and the underlying wafer table (which is typically temperature-controlled).

Similarly, where the base plate assembly extends under the top plate assembly outside the platform it is preferred that one or more channels are provided therethrough outside the platform to permit the delivery and/or removal of heat transfer fluid to and/or from the underside of the top plate assembly. This improves heat transfer between the top plate and base plate assemblies. The substrate could sit flat on the platform during processing, i.e. be supported by the platform across its whole surface. However this can lead to a nonuniform distribution of heat transfer fluid behind the wafer. As such, in some preferred examples, the platform further comprises a (preferably continuous) raised ledge disposed around the periphery of the platform whereby in use, a substrate placed on the raised ledge defines a volume between the underside of the substrate and the platform for the provision of heat transfer fluid. By providing a finite space of this sort between the substrate and the platform, the uniformity of heat transfer fluid pressure is improved and hence so is that of heat transfer itself. The heat transfer fluid could be allowed to flow out to join other gases in the processing chamber and hence in many embodiments no sealing element may be provided behind the substrate. The omission of a seal provides a number of attendant advantages, as will be discussed further below, which lead to a reduced risk of damage to the substrates during loading and unloading as well as allowing for use of a thinner top plate.

However, in other implementations it is preferable to reduce leakage of the heat transfer fluid into the chamber and hence the base plate assembly further comprises an elastomeric seal disposed in a recess extending around the periphery of the platform, whereby, in use, a substrate placed on the seal defines a sealed volume between the underside of the substrate and the platform for the provision of heat transfer fluid. This provides the further advantage that in the absence of a significant outgoing flow of fluid from the backside region of the wafer, the pressure of the fluid becomes substantially uniform throughout the region defined between the substrate, seal and platform surface. Therefore, the uniformity of heat transfer in this region is primarily determined by the actual gap between the substrate and the platform.

Whether or not a seal is provided between the platform and substrate, the base plate assembly may preferably further comprise an elastomeric seal provided on the underneath surface of the base plate assembly, whereby, in use, placement of the base plate assembly on a substrate table defines a sealed volume between the underside of the base plate assembly and the substrate table for the provision of heat transfer fluid.

In preferred examples, the elastomeric seal comprises an O-ring and may be made, for instance, of rubber. Advantageously, the elastomeric seal has a Shore A hardness of 60 or less. This is preferred since reasonably soft materials of this sort can be more readily deformed leading to a good seal between the substrate and platform. Harder materials (with higher Shore A hardness values) will require a greater force to be applied by the top plate in order to achieve the desired seal, which may not be compatible with the preferred resilient coupling described above. Further, greater force levels require a thicker and more rigid top plate, raising the level of its upper surface which is undesirable for the reasons already discussed. Preferably, the main components of the substrate carrier are formed of materials which conduct heat and electrical charge efficiently, in order that the desired thermal and electrical processing conditions can be established through the carrier from the wafer table on which it is fitted. For instance, as discussed above it is desirable to transfer heat away from the substrate to the (cooled) wafer table during processing and conductive materials will assist in this. Similarly, in many processes a bias voltage or RF bias is applied to the substrate by the wafer table, and a conductive carrier will enable transmission of the bias from the table to the substrate. Thus, preferably, the base plate assembly, top plate assembly and/or retention body (if provided) comprises an electrically and thermally conductive material, preferably aluminium or aluminium alloy. In another preferred version, the base plate is formed from conductive material while the top plate is formed from insulating material, such as alumina, causing the RF bias to be preferentially applied to the substrates. If the substrate carrier is positioned horizontally in the processing chamber, the weight of the top plate assembly may be sufficient to provide the clamping force necessary to secure the substrate in position. However, preferably a connection means will be provided to hold the top plate assembly in position against the base plate assembly. This could be provided by the existing processing chamber hardware, e.g. a wafer table clamp could be used to clamp the top plate assembly directly to the wafer table (with the base plate assembly in between). Alternatively or additionally, the substrate carrier could be provided with separate means for joining the base plate and top plate assemblies. Hence in preferred embodiments, the carrier further comprises a connector for urging the top plate assembly against the base plate assembly. The connector may comprise for example a clamp assembly or a bolted assembly configured to urge the top plate assembly against the base plate assembly. As already mentioned, the substrate carrier may be configured to hold one or more substrates. Where the carrier is adapted to carry a plurality of substrates, the base plate assembly preferably comprising a plurality of platforms, each platform for supporting one substrate in use, each platform defining a respective substrate zone, and the top plate having a plurality of apertures positioned to correspond to the respective substrate zones, at least one retention member extending from the top plate assembly into each aperture, over a portion of each respective substrate zone. As before, the retention members could be integral with the top plate or, preferably, resiliently coupled but most preferably, the retention member(s) extending into each aperture are carried on respective retention bodies, one retention body being provided for each aperture. This is advantageous since, as each substrate is placed on its respective platform, a retention body can be fitted to align the substrate correctly and hold it in place. This can be repeated individually for each substrate before applying the top plate.

The platforms could be provided as separate components, each platform being fitted individually to the top plate or placed on the wafer table for example. However, preferably, the plurality of platforms form part of a base plate connecting the platforms. Forming the platforms as an integral unit reduces the part count and simplifies assembly of the carrier.

The invention further provides a substrate carrier assembly comprising a substrate carrier as described above and at least one substrate disposed in and defining the substrate zone. Thus, the upper surface of the substrate corresponds to the upper surface of the substrate zone mentioned above. The carrier can be used for any type of substrate, but advantageously the at least one substrate is a dielectric substrate, preferably a sapphire substrate.

Also provided is a plasma processing apparatus comprising a substrate carrier or a substrate carrier assembly each as described above. The plasma processing apparatus may for example be a plasma etching tool or a plasma deposition chamber. The apparatus could be configured for example to carry out plasma etching to produce patterned sapphire substrates (PSS) which are typically used in the production of high brightness light emitting diodes (LEDs).

Typically the plasma processing apparatus will further comprise a substrate table on which the substrate carrier can be disposed and a table connector, preferably a clamp arrangement or bolt arrangement, for urging the substrate carrier against the substrate table.

The present invention further provides a method of manufacturing a substrate carrier for plasma processing apparatus, the method comprising:

selecting a size of substrate to be carried by the substrate carrier;

providing a base plate assembly comprising a platform configured for placement of a substrate of the selected size thereon in use, the volume to be occupied in use by a substrate placed on the platform constituting a substrate zone of the selected size; and

providing a top plate assembly on the base plate assembly, the top plate assembly comprising a top plate having an aperture therethrough surrounding the substrate zone and one or more retention members extending into the aperture, over a portion of the substrate zone, such that in use a substrate disposed in the substrate zone is retained in a fixed position between the base plate assembly and the top plate assembly;

wherein the upper surface of the top plate, facing away from the base plate assembly, is substantially planar and the one or more retention members protrude above the substantially planar upper surface of the top plate.

As described above, by configuring the upper surface of the top plate to sit at a level lower than the upper surface of the retention members, plasma sheath distortion is reduced leading to a decrease in the size of any exclusion zone. Again, it is preferable that a plurality of retention members are provided which are spaced from one another about the periphery of the aperture, most preferably at least three. The retention members are configured to abut and apply pressure to the upper surface of the substrate in use, to hold it in position. As before it is preferable that the upper surface of the top plate is substantially coplanar with the upper surface of the substrate zone, the location of which will correspond to the position of the upper surface of the selected substrate once in position on the platform. This can be determined from knowledge of the selected substrate thickness. Advantageously, the upper surface of the top plate is coplanar with the upper surface of the substrate zone to within +/- 1 mm, preferably +/- 0.75 mm, still preferably +/- 0.5 mm. Likewise, the undersides of the retention members (which contact the substrate in use) are preferably substantially level with the upper surface of the top plate, in use, to within +/- 1 mm, advantageously +/- 0.75 mm, still preferably +/- 0.5 mm. Preferably, the upper surface of the top plate is parallel to the upper surface of the substrate zone. Advantageously, the aperture is sized such that any gap between the top plate and the substrate zone is 2 millimetres or less. Also provided by the present invention is a method of mounting a substrate for plasma processing, comprising:

placing the substrate on a platform forming part of a base plate assembly; and

placing a top plate assembly on the base plate assembly, the top plate assembly comprising a top plate having an aperture therethrough surrounding the substrate and one or more retention members extending into the aperture, over a portion of the substrate such that the substrate is retained in a fixed position between the base plate assembly and the top plate assembly;

wherein the upper surface of the top plate, facing away from the base plate assembly, is substantially planar and the one or more retention members protrude above the substantially planar upper surface of the top plate.

By arranging the retention member(s) to protrude above the upper surface of the top plate, the same advantages as identified above are achieved. As before, it is preferable that a plurality of retention members are provided which are spaced from one another about the periphery of the aperture, most preferably at least three. The retention members are configured to abut and apply pressure to the upper surface of the substrate in use, to hold it in position. Again it is most preferable that the upper surface of the top plate is substantially coplanar with the upper surface of the substrate (in its mounted position). Advantageously, the upper surface of the top plate is coplanar with the upper surface of the substrate to within +/- 1 mm, preferably +/- 0.75 mm, still preferably +/- 0.5 mm. Likewise, the undersides of the retention members (which contact the substrate in use) are preferably substantially level with the upper surface of the top plate, in use, to within +/- 1 mm, advantageously +/- 0.75 mm, still preferably +/- 0.5 mm. It is preferred that the upper surface of the top plate is parallel to the upper surface of the substrate. Preferably, the aperture is sized such that any gap between the top plate and the substrate is 2 millimetres or less.

As discussed previously the top plate assembly can be constructed in a number of ways. In a particularly preferred embodiment, the retention member(s) are carried by a retention body which is separate from the top plate, the retention member(s) preferably forming an integral part of the retention body, and the method comprises placing the retention body on the base plate assembly, the retention member(s) extending partially over the substrate, and then placing the top plate over the retention body. This assists in the loading/unloading procedure and also ensures good alignment of the substrate in the carrier.

Advantageously, the retention body substantially follows the peripheral shape of the substrate such that placement of the retention body over the substrate results in centering of the substrate with respect to the platform.

Where a plurality of substrates are to be mounted for plasma processing, the method preferably further comprises placing each of the plurality of substrates on a respective platform of the base plate assembly prior to placing the top plate assembly over the base plate assembly. Advantageously, a corresponding plurality of retention bodies is provided and a retention body is placed on the base plate assembly, the retention member(s) extending partially over each respective substrate, before placing the top plate over the plurality of retention bodies. This further simplifies loading and unloading of the carrier and helps to avoid damage to the substrates. Examples of substrate carriers, substrate carrier assemblies and corresponding methods in accordance with the present invention will now be described and contrasted with examples of conventional systems with reference to the accompanying drawings, in which:

Figures 1 a and 1 b illustrate an exemplary substrate carrier assembly of conventional construction, in plan view and cross-sectional view respectively; Figures 2a and 2b are cross-sections through the substrate carrier assembly of Figure 1 along lines Y-Y' and Z-Z' respectively;

Figure 3 depicts a partial cross-section of a substrate carrier assembly in accordance with a first embodiment of the invention;

Figure 4 shows an exemplary base plate assembly for use in the first embodiment;

Figure 5 shows an exemplary top plate for use in the first embodiment;

Figures 6a, 6b and 6c show three examples of retention bodies for use in the first embodiment;

Figure 7 shows a schematic cross-section through line Q-Q' of Figure 3;

Figure 8 is a partial cross-section of a substrate carrier assembly in accordance with a second embodiment of the invention;

Figure 9 is a partial cross-section through a substrate carrier assembly in accordance with a third embodiment of the invention;

Figure 10 is a partial cross-section through a substrate carrier assembly in accordance with a fourth embodiment of the invention;

Figure 11 is a cross-section along the line R-R' of Figure 10;

Figure 12 is a partial cross-section of a substrate carrier assembly in accordance with a fifth embodiment of the invention;

Figure 13 is a schematic cross-section through a substrate carrier assembly in accordance with a sixth embodiment of the invention;

Figure 14 is a schematic cross-section through a substrate carrier assembly in accordance with a seventh embodiment of the invention;

Figure 15 is a plot depicting experimental results in which the non-uniformity of features etched in a substrate carried in (i) a conventional carrier and (ii) a carrier in accordance with an embodiment of the present invention was measured for different distances from the substrate edge; and Figure 16 provides exemplary SEM images of features etched in a substrate (a) using a conventional substrate carrier at (i) 1 millimetre from the substrate edge and (ii) 3 millimetres from the substrate edge; and (b) using a substrate carrier in accordance with an example of the present invention at the same distances from the substrate edge.

For ease of reference, the embodiments of the invention discussed below will primarily be explained by reference to examples of substrate carrier assemblies, which include a substrate carrier as well as at least one substrate (or "wafer") loaded into the substrate carrier, as depicted in Figure 3 for example. It will be appreciated that the substrate carrier itself will typically be provided unloaded, with one or more substrates being mounted in the carrier prior to processing. In the case of an unloaded substrate carrier, the volume which the substrate itself will occupy once mounted is the "substrate zone" and has the same dimensions as the substrate to be loaded. Thus, in the Figures, the items labelled "4", whilst identified below as substrates, can alternatively be considered to represent substrate zones in an unloaded substrate carrier.

Figure 3 shows a first embodiment of a substrate carrier assembly 20 in which a substrate 4 is retained between a base plate assembly 22 and a top plate assembly comprising at least a top plate 21 and retention members 25. In this example, the substrate carrier is designed to carry a plurality of substrates 4, for batch processing. Portions of adjacent substrates 4 are just visible in Figure 3. In practice, the substrate carrier may be adapted for the holding of any number of substrates, including a single substrate, the mechanisms via which each substrate is held being substantially identical. As such, the components of the substrate carrier will be discussed with reference to a single one of the substrates, although it will be appreciated that the same description applies to any additional locations for carrying substrates which are provided on the apparatus.

The substrate 4 sits on a platform 23 of the base plate 22, which is preferably raised relative to the surrounding parts of the base plate. The platform 23 is of substantially the same size and shape as the substrate 4 which is to be carried, typically circular or near circular. One or more channels 28 pass through the platform 23 to enable the passage of a heat transfer fluid such as helium gas to the back side of the wafer 4 as discussed further below. The wafer 4 is retained in position on the platform 23 by retention members 25, each of which extends a small distance over the periphery of the wafer 4. The underside 25c of each retention member 25 contacts the substrate 4 such that each retention member 25 exerts a downward force on the substrate 4, the force preferably being substantially perpendicular to the surface of the substrate, to hold the substrate 4 in a fixed position. Any number (one or more) of retention members 25 may be provided depending on the dimension of the wafer to be carried, but preferably a plurality of such retention members are provided, spaced from one another about the periphery of the aperture by gaps (intervals) in which the substrate is not covered right up to its edge. Typically, larger wafer diameters will require a larger number of retention members 25. Where the carrier is designed to hold multiple wafers for batch processing, preferably the force applied to each substrate by the retention members is independent of the forces applied to other substrates installed in the carrier and means for achieving this will be discussed below.

In this example, the retention members 25 are carried by a retention body 26, discussed further below. The retention body 26 is urged against base plate 22 by a top plate 21 fitted over the retention body 26. The top plate 21 surrounds the substrate 4 which is exposed to the processing chamber environment through an aperture 21 a in the top plate 21 , again of substantially the same shape as substrate 4. The upper surface 21 b of the top plate 21 is lower in height (along the direction parallel to the normal of the plate) than the retention members 25, which protrude above the surface 21 b in order to fit over the substrate 4 and thereby retain it in position. As discussed below, the resulting topography of the substrate carrier surface presented to the processing chamber has little effect on the shape of the plasma sheath, leading to a decrease in the size of any exclusion zone at the edge of the substrate 4. Thus, the substrate carrier comprises three primary components as shown disassembled in Figures 4, 5 and 6. Figure 4 is a perspective view of the base plate assembly 22 utilised in the first embodiment, and in this Figure all five wafer processing locations on the carrier are visible. Each location is defined by a platform 23 on which the wafer 4 will be placed in use. In this example, as shown best in Figure 3, each platform is raised relative to its surroundings, although this is not essential. Further, in this example each platform 23 is defined by a peripheral recess 23a which is used to seat and locate retention body 26; this too is however optional. The size and shape of each platform 23 is preferably determined by that of the substrate 4 which is to be carried and hence the diameter D _P of each platform 23 will typically be substantially the same or slightly larger (e.g. by up to 4 mm) as that of the substrate 4. It should be noted that each of the platforms 23 may be differently sized in order that different sizes of substrates can be accommodated.

The base plate assembly 22 is typically formed of a thermally and electrically conductive material such as a metal or metal alloy, preferably aluminium alloy. The use of such materials promotes good heat conduction away from the substrate 4 and good electrical conduction of any bias applied to the substrate 4 during processing.

In this example, one integral base plate 22 is adapted to carry multiple substrates. However, in other cases batch processing may be achieved utilising a base plate assembly 22 comprising a plurality of disparate platform units which may for example be mounted directly to the wafer table individually.

The top plate 21 is depicted in perspective view in Figure 5. Here, the top plate 21 comprises a substantially flat plate having apertures 21a corresponding to each of the platforms 23. Again, the size and shape of the apertures 21a is preferably determined based on that of the substrates 4 to be carried and the retention body 26 (as discussed further below), although in practice the size of each aperture will need to include minimal clearance (e.g. 1 - 2 mm) between the substrate 4 and the aperture edge. As such, the diameter D _A of each aperture will typically be slightly larger than that of the substrate 4 (and platform 23). The top plate 21 is preferably also made of a thermally and electrically conductive material, such as aluminium alloy.

The retention body 26 carrying retention members 25 can take a number of different forms. In one example, the retention body 26 could simply comprise a solid cylinder (or frustum of e.g. less than 10 degree slope) having an outward facing flange 27 as its lower end and inward facing retention numbers 25 disposed about its upper end. Placement of the top plate 21 over the retention body 26 as shown in Figure 3 brings the top plate 21 into contact with the flange 27, thereby holding the retention body 26 in place and hence securing the substrate 4.

The use of a retention body 26 which is separable from the top plates 21 in this way is preferred since ease of loading and unloading the carrier is enhanced. In particular, to load the carrier, a first substrate 4 may be placed on one of the platforms 23 illustrated in Figure 4. A retention body 26 may then be placed over the substrate 4, the cylindrical (or frusto-conical) shape of the retention body 26 being such that it will be readily located and seated about platform 23 in recess 23a. This process will cause self-centering of the substrate 4 on the platform if necessary. As such, when the top plate 21 is then placed over the base plate 22 it can be ensured that no collisions between the top plate 21 and any one or more of the loaded substrates 4 will occur. Further, each substrate will be held in position by its corresponding retention body 26 whilst the remaining substrates are loaded. Overall, the risk of damage to the substrates is reduced and speed of loading increased.

Thus, the retention body 26 may simply act as a means for individually positioning and temporarily retaining each substrate before the top plate 21 is applied. However, in more preferred embodiments, a flexible coupling is provided between retention members 25 and top plate 21 and advantageously this flexible coupling may be achieved with the design of retention body 26. By providing a flexible coupling between the top plate 21 and the retention members 25, pressure is not applied to the substrate 4 directly by the top plate 21 , but only indirectly and to a modified degree via the flexible connection. Thus, the force of any impact from the top plate 21 applied to the carrier will be partially absorbed rather than transmitted to the substrates and further any stress concentrations due to distortions in the top plate 21 will be dissipated to an extent. Moreover each retention body 26 will act independently of each other retention body provided in the carrier for other substrates.

Figures 6a, 6b and 6c show three examples of retention bodies 26 which incorporate resilient members resulting in a flexible coupling between the top plate 21 and retention members 25. Retention body 26 shown in Figure 6a comprises a substantially cylindrical body as described before with a flange 27 extending outward at the base of the cylinder and retention members 25 extending radially inward at the top end of the cylinder. The flange 27 is connected to the walls of the cylinder only at spaced intervals corresponding to each retention member 27a. Between the connection points 27a, the flange 27 is divided, resulting in two free ends at each location marked 27b. These portions of the flange thereby act as spring beams which are able to flex in the vertical direction about the points which they join to the rest of the retention body. When assembled, the top plate 21 rests on the flange 27 as shown in Figure 3. In particular, the top plate 21 rests on the flange 27 in the regions of its free ends around positions 27b. This may be achieved either by pre-stressing the free ends of the flange 27 upward and/or by providing contact points on the underside of the top plate 21 which protrude downwards. As such, when the top plate 21 is placed on retention body 26 and urged towards base plate assembly 22, the flange 27 will undergo elastic deformation and the force ultimately applied to the substrate 4 via retention members 25 is substantially uniform, without localised stress concentrations.

In the Figure 6a example, the spring functionality of the retention body is effectively achieved by embedding simple beams in the flange 27 (i.e. the free ends 27b of the flange constitute beams). In other examples, more complex configurations of beams can be formed in the cylindrical wall of the retention body and an example of this approach is depicted in Figure 6b. Here, the retention body 26' is the same overall shape as previously described, with a generally cylindrical form having a flange 27 at its base extending outwards and retention members 25 at its top end extending inwards. The cylindrical wall 26a of the retention body 26' is formed of a network of beams, together constituting a spring which can be extended and/or compressed along the axial direction of the cylinder. As such, pressure applied to the flange 27 by top plate 21 will be modified by the embedded spring before transfer onto the substrate 4 by the retention members 25.

In each of the above examples, the retention body 26 comprises a flange 27 extending radially outwards to engage with the top plate 21. However, this is not essential and, more generally, what is required is that the top plate 21 contacts the retention body at one or more contact locations which are resiliently connected to the retention members 25. A further example is shown in Figure 6c, in which retention body 26" is again substantially cylindrical, with retention members 25 extending radially inwards at its top edge. However, in this example there is no flange and instead the top plate 21 is configured to rest on the top edge of the cylindrical wall of retention body 26 at positions away from the retention members 25. The cylindrical wall is formed into a series of circumferential beams 26b, the free ends 26c of which are free to flex in the vertical direction. Once assembled, the top plate 21 rests on the retention member 26" in the region of the free ends 26c. This may be achieved either by pre-stressing the beams 26b in the upward direction and/or providing contact locations protruding downwards on the underside of the top plate 21 in appropriate corresponding positions. Whichever approach is adopted, the flexible mechanical coupling between the retention members 25 and the top plate 21 enables relative vertical movement between the retention members 25 and the substrate 4 to be reduced or virtually eliminated during loading/unloading of the carrier and throughout handling. As such, the introduction of concentrated stress regions to the substrate surface can be avoided since the risk of point or line contact between the substrate and the carrier is reduced, the underneath surfaces of retention members 25 being able to remain parallel to the substrate surface at all times. It is further preferred that the beams or other types of flexible connection between the top plate 21 and retention members 25 share symmetry with that of the allocated substrate 4. For example, in the present case the apparatus is designed to carry a circular or near circular substrate and the spring arrangements depicted in Figures 6a, 6b and 6c each follow a circular configuration. As such, due to the symmetry of the spring arrangement in the retention body 26, the majority of any applied force components which are not perpendicular to the substrate surface will cancel one another out, significantly reducing the possibility of the retention member(s) sliding on the substrate 4.

Thus, in preferred examples such as this, an independent retention body 26 is fitted over each substrate 4. Each of the retention bodies 26 has a resilient connection such as embedded beams which transfer the clamping force provided by the top plate 21 to the retention members 25. By using retention bodies which are separate from the top plate 21 , relative movement of the top and base plates 21 , 22 does not produce high stress concentrations such as point or line contacts on the substrates 4. The force holding each substrate in position against the pressure of heat transfer fluid on its rear surface is applied when the top plate 21 is urged towards the base plate assembly 22 (e.g. via connection means such as screws or bolts, or table clamp 5 shown in Figure 1 b), and each retention body distorts independently to retain its respective substrate 4, regardless of minor distortions of either the top plate 21 or bottom plate 22. In each of the above examples, the retention body (including the retention members) is preferably formed of a thermally and electrically conductive material such as aluminium alloy.

As mentioned above, once the top plate 21 is fitted to the base plate 22, in all of the embodiments, the upper surface 21 b of the top plate (facing towards the interior of the processing chamber) is recessed relative to the retention members 25 which rise above it, as shown in Figure 3. As a result, the exclusion zone E described above with respect to Figures 2a and 2b is reduced or even eliminated in the intervals between retention members 25. This is illustrated in Figure 7 which is a cross-section along line Q-Q' in Figure 3. In the particularly preferred example shown in Figure 7, the upper surface 21 b of the top plate 21 is substantially co-planar with the top surface 4a of substrate 4 installed in the carrier (and hence with the undersides of the retention members 25). In practice, the best result will generally be achieved where the upper surface 21 b of the top plate 21 is substantially parallel with the top surface 4a of the substrate 4 and any discrepancy in height between the upper surface 21 b of the top plate 21 and the top surface 4a of the substrate is less than half of the thickness t of the substrate. For example, any discrepancy in the height of the two planes is preferably one millimetre or less, more preferably 0.75 millimetres or less. This is achieved by designing the substrate carrier to take into account the thickness t of the substrate 4 which is to be carried and configuring the components such that the height h of the upper surface 21 b of the top plate 21 relative to the surface of platform 23 (when the top plate 21 is secured in position) will substantially correspond to the thickness t of the substrate. Thus, in preferred examples, since typical substrates have thicknesses ranging between 0.5 and 2 millimetres, more generally between 0.5 and 1 millimetre, it is preferred that the upper surface 21 b of the top plate is at a height h of between 0.5 and 3 millimetres, preferably between 1.5 and 2 millimetres, more preferably between 0.5 and 1.5 millimetres from the upper surface of the platform 23.

As shown in Figure 7, the result is that, in an ideal situation away from the retention members 25 an approximately planar surface formed by the combination of top plate 21 and substrate 4a (and optionally retention body 26) is presented to the process chamber. As such, the plasma sheath 8 will suffer minimal distortions in the edge region, resulting in substantially uniform processing conditions. In tests, implementations of the present invention have achieved exclusion zones extending into the substrate interior by less than 1 millimetre which is a significant improvement on previous carriers.

Following the same principle, it is noted that any feature protruding above the wafer surface 4a produces disruption in the plasma sheath 8 and as such the number and size of such features is preferably kept to a minimum. Preferably, the retention member 25 has a maximum height in the direction parallel to the normal of the substrate 4 of 3 millimetres or less, more preferably 2 millimetres or less in order to minimise the extent of the exclusion zone established around each retention member. The lateral extent of each retention member 25 is also preferably kept small. For example, each retention member 25 may extend over the substrate 4 by 3 millimetres or less, preferably 2 millimetres or less. In particularly preferred examples the total area of the substrate zone covered by the one or more retention members is 2% or less of the full area of the substrate 4, more preferably 0.5% or less. It will be noted from the Figures that in these examples, each retention member 25 projecting above the substrate 4 is of generally cuboid shape but preferably having bevelled edges rather than 90 degree corners. The upper chamfer 25a (Figure 6a) is believed to further reduce deformation of the plasma sheath since, compared with a full cuboid shape, a sharp corner is removed. The lower chamfer or rounded edge 25b is provided to reduce stresses on the substrate during assembly. The underside 25c contacts the upper surface of the substrate in use.

Since the section of the upper surface 21 b of the top plate 21 located in the intervals between retention members 25 acts as an extension of the processing surface of the substrate 4, any lateral gap between the top plate 21 and substrate 4 could also give rise to sheath deformation. As such, it is preferred that any such lateral gap is kept to a minimum. In the first embodiment described above, the aperture 21a formed in the top plate 21 is designed so as to accommodate the substrate 4 and expose the retention body 26, which allows for easy assembly since no rotational alignment between the retention body and top plate 21 is required. As shown in the cross-section of Figure 7, this results in a discontinuous surface adjacent to the periphery of the substrate due to the gaps between the top plate 21 , retention body 26 and substrate 4. Preferably, the region is made to approximate a planar surface as closely as possible with the upper surface of the retention body in the intervals between retention members 25 being approximately level with the upper surface 21 b of the top plate. It should be noted that, in order for this to be achieved in practice, the examples of retention bodies 26 in Figures 6 a and c will be modified such that the upper edge of the cylindrical surface reaches the same height relative to the retention members 25 as shown in the Figure 6 b example.

In a second embodiment of the invention, as depicted in partial cross-section in Figure 8, the sheath deformation can be further reduced by reducing the size of the aperture 21 a' such that the edges of the top plate 21 approach the substrate directly with a minimal practical clearance gap of e.g. 2 millimetres or less. In order to achieve this configuration, the top plate 21 is shaped so as to fit over the circumferential surface of the retention body 26 and is provided with cut outs 21 a" in the edge of the aperture 21 a to accommodate the retention members 25. As a result, in the intervals between retention members 25, there is very little sheath deformation and the conditions can be considered a good approximation to the ideal situation discussed above. In this example, the upper edge of the cylindrical surface of the retention body 26 will be lower to accommodate the thickness of upper plate 22, as shown in the Figures 6 a and c examples. The corresponding edges shown in the Figure 6 b example would be lowered to the same height for use in this embodiment.

It will be noted that in the first and second embodiments discussed above, unlike the conventional carrier described previously with respect to Figures 1 and 2, there is no backside seal provided behind the substrate 4. However, the present inventors have found that the omission of a seal is preferred in many situations. Whilst the omission of a seal will allow the heat transfer fluid to flow out to join the process gases in the chamber, this is acceptable in many circumstances and results in a number of benefits. Firstly, the provision of an elastomeric seal such as 3 in the conventional carrier of Figures 1 and 2 leads to the need for a relatively high force to be applied by the top plate in order to compress the elastomeric seal 3 and prevent the passage of fluid past the seal. This in turn requires a high rigidity and hence increased thickness of the top plate. By omitting the seal 3, less compression force is required and hence a thinner top plate 21 may be used. This assists in lowering the profile of the upper surface 21 b of the top plate 21 , thereby contributing to low plasma sheath deformation as discussed above. Similarly, the high compression force required for the utilisation of elastomeric seals 3 increases the risk of localised deformation of the top plate 21 during clamping of the substrates. This in turn increases the risk creating stress concentrations during loading or unloading, such as those which occur if the mechanical clamping structure has point or line contact with the substrate, which can lead to the breakage of the substrate before or after processing. However, this problem is reduced by the use of resilient connections between the top plate 21 and retention members 25 as discussed above with respect to Figure 6. Similarly, the desired sealing effect will only be achieved if the substrate 4 is positioned accurately in alignment with the elastomeric seal 3 and in conventional systems this is difficult to achieve. The misalignment may only be detected once the substrate carrier is loaded and processing begun, being identified by a sudden increase of heat transfer fluid flow needed to achieve a set pressure behind the substrates. In such cases, using a conventional system, the loaded carrier would need to be removed, unloaded and the process restarted. In the presently disclosed system, this problem too is mitigated by the use of independent retention members 26 which as described above can be used to correct the positions of each wafer before the top plate 21 is introduced.

Another problem that can be encountered during loading or unloading of the carrier where an elastomeric seal is provided is that the elastomeric seals, when compressed against the substrate, fill cavities present in the back surface of the substrate and in the surfaces of the recess provided in the platform which carries the elastomeric seal. When combined with temperature changes encountered during processing, it is common to experience significant adhesion between the seal and the adjacent bodies, which may make unloading of the substrate time consuming and presents a risk of damage to the substrate. This too is mitigated to a degree by the use of individual retention bodies since each substrate will be unloaded individually allowing the operator's full attention to be given to each substrate as they are manipulated one by one.

In light of the above, in many embodiments it is preferred not to provide any seal between the substrate 4 and platform 23. In this case, the substrate 4 may sit flat against platform 23. That is, the substrate 4 may be in contact with and supported by the platform across the whole area of the substrate 4. However, to improve the uniform provision of heat transfer gas behind the substrate 4, in may be preferred to provide a small gap between the platform surface and the rear of the substrate which allows the passage of heat transfer gas to reach all regions of the substrate. An example of such an embodiment is shown in Figure 9, and here the platform 23 includes a (continuous) raised ledge 29 extending about the whole length of its periphery. In use, the substrate 4 rests on the raised ledge 29, creating a partial seal and leaving a small planar region 29a between the platform 23 and the substrate 4 which in use can be filled by heat transfer fluid through channel 28. The remainder of the components of the substrate carrier are substantially the same as described with respect to the previous embodiments, although in this example the retention body 26 carries only three evenly spaced retention members 25 rather than six as in the previous examples.

Despite the disadvantages of using elastomeric seals discussed above, in some circumstances their benefits may outweigh these problems. In particular, since the provision of a seal reduces leakage of heat transfer fluid from behind the wafer into the process space (which is typically held at vacuum level) the absence of significant outgoing fluid flow means that the pressure of the heat transfer fluid behind the wafer is substantially spatially uniform. As such, the uniformity of heat transfer in this region is primarily determined by the actual gap dimension between the substrate 4 and the relevant surfaces of the base plate 22 on which the substrate is placed. Uniform heat transfer is desirable since this will also improve the uniformity of the resulting structures formed during the processing.

Figure 10 therefore depicts a further embodiment of the invention in which elastomeric seals 30 are provided about the periphery of the platform 23 behind each substrate 4. The elastomeric seal may for example comprise an O-ring, such as a rubber O-ring, and preferably has a Shore A hardness of 60 or less. This is preferred since reasonably soft materials of this sort can be more readily deformed leading to a good seal between the substrate and platform. Harder materials (with higher Shore A hardness values) will require a greater force to be applied by the top plate in order to achieve the desired seal, which may not be compatible with the preferred resilient coupling described above. Further, greater force levels require a thicker and more rigid top plate, raising the level of its upper surface which is undesirable for the reasons already discussed.

The remaining components of the substrate carrier are identical to those discussed previously with respect to Figures 3 to 7. By providing a seal of this sort, the apparatus can be used for processes where high leakage of the heat transfer fluid into the process chamber is unacceptable, the seal enabling a very low leakage rate to be achieved. As discussed above, the seal also achieves an increase in uniformity of the backing fluid pressure beneath the majority of the substrate surface. However, one additional issue which the inclusion of an elastomeric seal may present is that of thermal non-uniformity in the seal region. As depicted in Figure 1 1 , which is a cross-section along line R-R' shown in Figure 10, the provision of an elastomeric seal 30 on the rear side of the substrate 4 divides the back surface of each substrate 4 into three zones. The main (central) zone 9a of the substrate is in direct contact with the heat transfer fluid at the desired pressure and here uniform processing conditions can be achieved. Where the seal 30 comes into contact with the substrate 4, the presence of the heat transfer fluid is reduced and since the elastomeric material has a different heat transfer characteristic from that of the heat transfer fluid, there will be a difference in temperature in the seal region 9b. Outside the seal 30 in the outer zone 9c, there is substantially no heat transfer fluid and the substrate is exposed to the process chamber vacuum environment. Each of the zones 9a, 9b and 9c represents different heat transfer conditions (and different bias conditions, if a bias is applied through the wafer table), which can reduce uniformity adjacent the periphery of the substrate.

In order to reduce this effect, the elastomeric seal 30 is preferably located as close to the edge of the substrate 4 as possible and its lateral width is kept to a minimum. Again, omitting the elastomeric seal 30 as in previous embodiments avoids such problems since the substrate placement surface of the base plate 22 has a uniform structure such that thermal (and bias) non uniformities (other than those related to any backside pressure gradient) will not be introduced. Alternatively, depending on the selection of the seal, the adverse effects may be reduced to an acceptable level. For instance, the width of the seal is preferably kept to a minimum and the seal material may be selected to be thermally and electrically conductive. In another example, different seal cross-sections may be utilised in place of o-rings. For example, a seal with a triangular cross-section may be used to reduce the size of zone 9b: the seal may arranged such that the wafer presses against one corner of the seal, allowing for minimal contact surface area. This, preferably together with the seal being located close to the wafer's edge and with increased localised deformation of the seal, would provide good sealing with reduced thermal non-uniformity. It should be noted that an elastomeric seal of the sort described above can be incorporated into any of the embodiments disclosed herein. For example, Figure 12 depicts a further embodiment corresponding to that discussed above with respect to Figure 8, modified solely by the inclusion of an elastomeric seal 30 for the reasons given above. Whilst in the embodiments disclosed in Figures 8 and 12, assembly the carrier structure will be less straightforward, owing to the need to align retention members 25 with recesses 21 a", as discussed previously, the overall plasma sheath deformation and hence exclusion zone will be reduced by the reduced gap between top plate 21 and substrate 4 which is presented. As discussed above, it is preferable that the top plate 21 and retention members 25 are moveable relative to one another, most preferably flexibly coupled, in order to attain the benefits in terms of simplified loading procedure and reduced risk of substrate breakage discussed above. However, this is not essential and aim of reducing the exclusion zone may be achieved without also providing these additional benefits.

To illustrate this point, Figure 13 shows an embodiment in which the retention members 25 are integrally formed with the top plate 21. For example, the top plate 21 may be moulded or otherwise formed to incorporate raised protrusions extending into and forming retention members 25, or the upper surface 21 b could be recessed to below the level of retention members 25 by machining out of an originally flat plate. Implementations such as those depicted in Figure 13 may be particularly suitable for single wafer processing as opposed to batch processing to reduce the level of difficulty during loading of the substrate since here careful alignment with the top plate will be necessary.

Figure 14 shows a cross section through a further embodiment in which the retention members 25 are carried by top plate 21 but are flexibly coupled thereto in order that pressure from top plate 21 is only indirectly transferred to the substrate, retaining some of the additional benefits discussed above. In this example, the retention members 25 are carried in slots provided about the aperture 21a and held within those slots by extension springs 24. Thus, the clamping force applied by the top plate 21 will be distributed more uniformly by the resilient connections 24 between the top plate 21 and the retention members 25. The retention members 25 could all be linked, e.g. formed as part of a retention body 26 similar to those discussed above, or could be separate from one another and hence moveable entirely independent of one another. In both embodiments described with respect to Figures 13 and 14, the upper surface 21 b of the top plate 21 is again substantially coplanar with the top surface of substrate 4. Consequently, the undersides 25c of retention members 25 are also substantially coplanar with the upper surface 21 b of the top plate 21. Thus, by reducing the level of the upper surface of the top plate relative to the retention members, plasma deformation can be decreased and hence the exclusion zone reduced. To illustrate this result, Figure 15 shows experimental data wherein the plasma etch selectivity non-uniformity in the edge area of a 100mm diameter sapphire substrate has been measured after two very similar low pressure chlorinated chemistry process conditions in a ICP plasma processing tool. "Selectivity" refers to the ratio of the sapphire etch rate vs. the photoresist etch rate (i.e. [sapphire etch rate in nm/min] / [photoresist etch rate in nm/min]). The selectivity non-uniformity measurement in Figure 15 is based on test points located at distances between 1 millimetre and 5 millimetres from edges of the sapphire wafer, plus one point in the centre of the wafer. The selectivity non-uniformity value given for each position along the x-axis in the Figure is determined by comparing the selectivity ratio at each of the test points which falls on or inside the indicated distance from the wafer edge. Hence, at 1 mm from the edge of the wafer, the non-uniformity value is calculated based on etch depths of sapphire and photoresist measured across the diameter up to 1 mm from the wafer edge, whereas at 3 mm from the edge of the wafer, the non-uniformity value is based on data across the diameter up to 3mm of the wafer edge. The non-uniformity values correspond to a degree of variation in the measured selectivity ratio between the features at the relevant test sites. In this example the non-uniformity values were calculated using the formula:

+/- 100 x (max - min)/(2 x mean)

across the data set, where "max" is the maximum selectivity ratio in the data set, "min" is the minimum selectivity ratio in the data set, and "mean" is the mean average selectivity ratio in the data set.

The measurements were repeated using (i) a conventional substrate carrier such as that described above with respect to Figures 1 and 2, and (ii) using a substrate carrier in accordance with the preferred embodiment of the discussed above with respect to Figure 3. It will be seen that, using conventional apparatus (line (i)) the degree of non-uniformity was more than twice that obtained using the presently disclosed carrier (line (ii)) at 1 and 2 mm from the substrate edge. With the conventional apparatus there is a significant decrease in non-uniformity between 2 and 3 mm from the wafer edge, indicating the approximate location of the end of the exclusion zone in this region. Using the presently disclosed carrier, there is a much less pronounced change in non- uniformity at this position indicating that the process conditions are essential uniform across the 1 mm, 2mm and 3mm positions.

Figure 16 shows SEM images of etched features obtained after processing sapphire substrates with a typical low pressure chlorinated chemistry plasma etch process. In Figure 16a, the two images show features obtained using a conventional sandwich substrate carrier (such as that depicted in Figures 1 and 2) located (i) at 1 millimetre from the substrate edge and (ii) at 3 millimetres from the substrate edge. It will be seen that at both locations there is very strong deformation of the etched features (which are intended to be symmetrical) as well as a large difference in the etch rates between the two locations. For instance, in the image taken at 1 millimetre (Figure 16a(i)), a significant amount of resist is still visible, whilst this has largely been removed at the 3 millimetre location (Figure 16a(ii)). This indicates that the etch rate at 1 millimetre from the substrate edge is significantly lower than that at 3 millimetres from the substrate edge. Figure 16b shows corresponding images of features produced using a substrate carrier in accordance with the Figure 3 embodiment following the same etching process. In this case it can be seen that at the locations 1 millimetre (Figure 16b(i)) and 3 millimetres (Figure 16b(ii)) from the substrate edge, the etch rate is virtually identical and there is substantially no tilt to either feature.

The disclosed sandwich substrate carrier therefore preferably provides for:

• Good heat transfer between each wafer and the carrier plate

· Good heat transfer between the carrier plate assembly and the wafer table

• Distribution of a heat transfer gas (typically helium) to the rear surface of each wafer

• Ease of assembly (quick to load the plate with wafers, and reliable

sealing of all helium gas seals)

• Similar heat transfer and RF coupling characteristic over the entire wafer area

• Reduced exclusion zone size. The base plate assembly preferably has the following functionality:

Accommodates one or a plurality of substrates on one or more top faces (platforms) and defines their positions during processing;

(Optionally) provides means of restricting back side leak of cooling fluid from the central area of the plate to the process chamber;

- (Optionally) provides flow paths (channels) from the back side of itself to the back side of processed substrates;

(Optionally) provides conditions for limited cooling fluid leak up from behind clamped substrates to the process chamber. The top plate preferably has the following functionality:

Allows for predefined location of substrates;

Provides optimal topography of the top face of the assembled carrier by recessing its level below that of the retention members which hold the substrates in place - most preferably creating conditions where the sections of the top plate in proximity of the of substrates are substantially co-planar with the top surfaces of those substrates - to thereby reduce the exclusion zone;

Constrains retention members in order to allow them to produce clamping force;

- (Optionally) limits exposure of the retention members to plasma processing.

The retention member(s) or retention body(ies) preferably have the following functionality:

- Produce individual clamping force acting from the top of each substrate towards the base plate;

Restrict movement of the substrates during assembly and disassembly of the carrier;

Provide minimum surface contact with the processed substrates.