Patent Application
20070261964
This application relates to reactors, systems, and methods for electroplating microfeature workpieces having a plurality of microdevices integrated in and/or on the workpieces. Particular aspects of the present invention are directed toward electroplating workpieces using a low-profile reactor.
Microfeature devices, such as semiconductor devices, imagers, displays, and micromechanical components, are generally fabricated on and/or in microfeature workpieces using a number of machines that deposit and/or etch materials from the workpieces. Many current microfeature devices have interconnects and other very small sub-micron sized features (e.g., 45-110 nanometers) formed by depositing materials into small trenches or holes. One particularly useful process for depositing materials into small trenches and/or vias is electroplating. Typical electroplating techniques include plating processes that deposit copper, solder, permalloy, gold, silver, platinum, electrophoretic resist, and other materials onto the workpieces.
One challenge of plating metals into narrow, deep trenches or holes is that it is very difficult to completely fill the very small features without creating voids or other non-uniformities in the metal. For example, an ultra thin seed layer must be used to deposit a metal into a trench having a dimension of 45 nm to 250 nm so that the trench has sufficient vacant space for the subsequently plated metal. Ultra thin seed layers, however, are problematic because they may not cover the workpiece uniformly. For example, ultra thin seed layers may have voids or other non-uniform physical properties that can result in non-uniformities in the plated material. This is particularly common on the side walls of such small features. These challenges can be overcome by plating onto the seed layers or plating a seed layer directly on a barrier layer to provide competent seed layers that are well-suited for bulk plating metals into trenches or holes with small critical dimensions. One particularly useful method is to plate a material using a processing solution with a low conductivity. Suitable electrochemical processes for forming competent seed layers are disclosed in U.S. Pat. No. 6,197,181, which is herein incorporated by reference.
One problem with low conductivity baths, however, is that they require more power to plate the material onto the workpieces compared to acidic plating solutions with high conductivities. As a result, conventional reactors require larger power supplies to operate with the low conductivity baths used for plating onto seed layers or plating directly onto barrier layers. This increases the operating costs of conventional reactors because they consume more power, and it may be inconvenient because the facility may need to be renovated to provide high-voltage power outlets. Another challenge of using low conductivity baths is that they are generally more sensitive to shielding and geometries of the system. As such, non-uniformities in the electrical field near the anode are more likely to cause non-uniformities on the workpiece compared to highly conductive baths.
Plating onto seed layers or plating materials directly onto barrier layers in high pH processing solutions presents additional challenges. For example, inert anodes are generally required in high pH processing solutions because the high pH passivates consumable copper anodes. This generally produces copper hydroxide particles and/or flakes that may create defects on the wafer. However, one drawback of using inert anodes is that the deposited material exhibits a significant increase in resistivity over a relatively small number of plating cycles (e.g., short bath life). The resistivity of the deposited layers can be maintained in a desired range by frequently changing the processing solution, but this increases the operating costs of electrochemically processing seed layers and/or depositing materials directly onto barrier layers.
As a result, there is a need for a reactor that reduces power consumption, enables the use of inert anodes in low conductivity processing solutions, and enables use of high pH processing solutions.
The present invention provides electroplating reactors for processing seed layers, depositing materials directly onto barrier layers, and/or depositing materials onto microfeature workpieces in other applications. The reactors enable the use of an inert anode and a low conductivity processing solutions so that the reactors of the invention are not plagued by the flaking problems associated with using consumable anodes in high pH solutions. The reactors achieve this result by significantly reducing or eliminating oxidation of certain constituents of the processing solution to prolong its operating life. This reduces the operating costs of using the reactor because the tool does not need to be frequently shut down to change out the processing solution. A further benefit of reactors in accordance with the invention is that they are designed to operate within conventional power consumption ranges and still plate a uniform layer of material on a seed layer or directly on a barrier layer. The reactors also enable the use of high pH processing solutions in the same system. As a result, reactors in accordance with the invention are expected to further reduce the operating costs of electroplating seed layers, depositing materials directly onto barrier layers using low conductivity processing solutions, and/or using high pH processing solutions.
One aspect of reactors in accordance with the invention is that they have a first chamber configured to direct a first processing solution to a processing zone, a second chamber configured to contain a second processing solution different than the first processing solution, and an ion exchange membrane between the first chamber and the second chamber. The reactors also include (a) a support member in the first chamber that contacts the ion exchange membrane across the surface of the membrane, and (b) a counter electrode in the second chamber. The ion exchange membrane enables a low conductivity catholyte to be used in the first chamber and an inert counter electrode in the second chamber. More specifically, the ion exchange membrane prevents nascent oxygen that evolves from the inert counter electrode from reaching the catholyte. This reduces oxidation of constituents of the catholyte, which results in good quality deposits over a long bath life. The ion exchange membrane is also quite useful in applications that use a low conductivity, high pH catholyte. Therefore, reactors of the invention provide a platform for using inert counter electrodes for precision plating processes.
The support member provides several advantages for using a low-profile reactor. The support member contacts the ion exchange membrane to impart the desired contour or profile to the membrane. This prevents the membrane from having a shape that disturbs the electrical field or traps bubbles. The support member also directs a flow of the first processing solution over the ion exchange membrane and toward the processing zone, and the support member configures the electrical field in the first chamber. The support member enables the reactors to reduce the distance between the processing zone and the electrode in the second chamber so that less power is required to establish a sufficient electrical field compared to conventional reactors. Reactors in accordance with the invention accordingly enable the use of low conductivity, high pH processing solutions with conventional power supplies that are already widely used in plating tools.
Several other features of the reactors in accordance with the invention enhance the uniformity of the layers deposited onto the workpieces. For example, the electrode in the second chamber can be shaped to further enhance the uniformity of the electrical field at the processing zone. Therefore, several embodiments of reactors in accordance with the invention also enable the uniform deposition of materials onto workpieces using low conductivity, high pH processing solutions.
The reactor 100 further includes a workpiece holder 140 having a plurality of electrical contacts 142 for applying an electrical potential to a workpiece W mounted to the workpiece holder 140. The workpiece holder 140 can be a movable head configured to position the workpiece W in the processing zone 112 of the first chamber 110, and the workpiece holder 140 can further be configured to rotate the workpiece W in the processing zone 112. Suitable workpiece holders are disclosed in U.S. Pat. Nos. 6,080,291; 6,309,520; 6,527,925; 6,773,560; and U.S. patent application Ser. No. 10/497,460; all of which are incorporated herein by reference. In applications that use low conductivity baths, it is expected that “dry” contact assemblies that seal against the perimeter of the workpiece are efficacious. Suitable dry contact assemblies are disclosed in U.S. Pat. Nos. 6,773,560 and 6,309,520. Such dry contact assemblies are also useful for plating onto high resistance layers (e.g., seed layer enhancement or direct-on-barrier plating).
The reactor 100 further includes (a) a support member 150 in the first chamber 110 that contacts a first surface 132 of the membrane 130, and (b) a counter electrode 160 in the second chamber 120. The support member 150 spaces the ion exchange membrane 130 apart from the workpiece processing zone 112 by a controlled distance. This feature provides better control of the electrical field at the processing zone 112 because the distance between the ion exchange membrane 130 and the workpiece processing zone 112 affects the electrical field at the processing zone 112. The support member 150 generally contacts the first surface 132 of the ion exchange membrane 130 such that the distance between the first surface 132 and the processing zone 112 is substantially the same across the first chamber 110. Another feature of the support member 150 is that it also shapes the ion exchange membrane 130 so that bubbles do not collect along a second side 134 of the membrane 130. The support member 150 can also be configured to impart a slanted or non-planar contour to the ion exchange membrane 130 for increasing/decreasing the strength of the electrical field in local regions of the processing zone 112 or providing additional bubble control.
The support member 150 is further configured to direct a flow F1 of the first processing solution laterally across the first surface 132 of the ion exchange membrane 130 and vertically to the processing zone 112. The support member 150 accordingly controls the flow F1 of the first processing solution in the first chamber 110 to provide the desired mass-transfer characteristics in the processing zone 112. The support member 150 also shapes the electrical field in the first chamber 110. The upper surface of the support member 150, for example, can be contoured to shape the electrical field within the reactor or otherwise control the electrical field of the workpiece. These features of the support member 150 enable the reactor 100 to provide good electrical and mass-transfer properties at the processing zone 112 with a relatively short distance D1 between the processing zone 112 and the counter electrode 160. As a result, the reactor 100 is particularly well suited for use with low conductivity processing solutions because less power is required to establish the desired electrical field at the processing zone 112 over the relatively short distance D1 compared to conventional reactors in which the space in between the workpiece W and the counter electrode is significantly larger.
The counter electrode 160 is spaced apart from the second surface 134 of the ion exchange membrane 130 by a gap distance D2 such that a flow F2 of the second processing solution moves radially outward across the second surface 134 of the ion exchange membrane 130 at a relatively high velocity. The flow F2 of the second processing solution accordingly sweeps oxygen bubbles and/or particles from the ion exchange membrane 130. The short distance D2 also limits the volume of the second processing solution in the second chamber 120 so that the second processing solution can be quickly withdrawn from the second chamber 120. The reactor 100 further includes a flow restrictor 170 around the counter electrode 160. The flow restrictor 170 is a porous material that creates a back pressure in the second chamber 120 to provide a uniform flow between the counter electrode 160 and the second surface 134 of the ion exchange membrane 130. As a result, the electrical field can be consistently maintained because the flow restrictor 170 mitigates velocity gradients in the second processing solution where bubbles and/or particles can collect. The configuration of the counter electrode 160 and the flow restrictor 170 also maintains a pressure in the second chamber during plating that presses the ion exchange membrane 130 against the support member 150 to impart the desired contour to the ion exchange membrane 130.
The reactor 100 operates by positioning the workpiece W in the processing zone 112, directing the flow F1 of the first processing solution through the first chamber 110, and directing the flow F2 of the second processing solution through the second chamber 120. As the first and second processing solutions flow through the reactor 100, an electrical potential is applied to the workpiece W via the electrical contacts 142 and the counter electrode 160 to establish an electrical field in the first and second chambers 110 and 120. When the counter electrode 160 is inert, the ion exchange membrane 130 prevents bubbles and particles from entering the first chamber 110. As such, the bubbles that evolve from the counter electrode 160 are carried out of the second chamber 120 by the flow F2 of the second processing solution.
The reactor 100 enables the efficient use of inert electrodes and/or low conductivity solutions for several reasons. First, the ion exchange membrane 130 mitigates oxidation of constituents in the first processing solution(e.g., chelating agents or other organic/inorganic species). Second, the total distance D1 between the workpiece W and the counter electrode 160 is significantly less than conventional reactors. This reduces the power necessary to establish the requisite electrical field through low conductivity processing solutions, which reduces the operating costs of the reactor 100 compared to conventional reactors. Third, the support member 150 controls the shape of the ion exchange membrane 130 and the flow F1 of the first processing solution in the first chamber to provide the desired electrical field and mass-transfer characteristics in the processing zone 112. The support member 150 enables the total distance D1 between the processing zone 112 and the counter electrode 160 to be relatively short without unduly sacrificing the desired electrical and mass-transfer properties at the processing zone 112.
The reactor 100 illustrated in
Furthermore, gold plating or plating other precious metal or other metals that depend on use of an inert anode are plating processes that are candidates for membrane reactors employing an anion exchange membrane because the anodic process can be separated from the catholyte by the ion exchange membrane. Properties of gold deposits (such as roughness) will be fairly constant over the bath life. The bath stability and bath life of these plating processes can be greatly improved by isolating the plating species from the rest of the bath solutions. The anodic oxidation reaction in the anolyte of these plating baths does not influence the chelating agents or other organic agents for precious metal plating in the catholyte. The reactor can be operated in a continuous fashion indefinitely with no anode maintenance as the metal salt is added as a solution to the catholyte side and the metal ions are plated. The counter ions move across the membrane and pass into solution in the anolyte—where hydroxyl ions are lost due to the inert anodic reaction of oxygen production.
Furthermore, there are advantages in using an anion exchange membrane reactor plating for Sn (and other metals with multiple valence states). Any metal with multiple valence states can be plated from most of their stable states. Since the charge required to deposit any metal is directly proportional to the electrons required for their reduction, metals in their valence states closest to their neutral states consume less energy for reduction to metal. However most of the metals in their closest state with their valence are inherently unstable, and therefore production worthy plating is not feasible. With the reactor of this invention inherently unstable plating solutions can still be applied for a production worthy process because the separation of the anodic and cathodic process results in lesser oxidation of the inherently unstable metal species.
Most Sn—Ag—Cu alloy plating solutions prefer Sn(II) as the species for Sn plating. For such alloy plating systems, the control of Sn, Ag and Cu ions needs to be tight and the use of Sn, Ag, Cu as an anode is ruled out. These anode systems could not only cause stability issues by plating/reacting with the anodes, but also pose a problem for uniform metal replenishment. The use of inert anodes, however, causes nascent oxygen which not only oxidizes any organics in the plating bath, but also oxidizes Sn(II) to a stable Sn((V) ion. To prevent problems from occurring in such baths, the use of an inert anode anionic membrane reactor in accordance with this invention can greatly simplify replenishment schemes, and does not oxidize any organic agents or any metal species. This separation between catholyte and anolyte is ideal to solve such problems encountered in plating baths. The baths are relatively more stable, and the bath life can be extended.
Furthermore, for systems that use chelated baths, high pH, or other systems where inert anodes are used, the reactor of this invention can be very beneficial. First and foremost the advantages of using the membrane reactor of this invention for a high pH Cu bath with a chelating agent are: (a) the conductivity and pH of the catholyte is fairly constant; (b) the properties of deposit (eg. resistivity) are constant with bath age; (c) the chelating agent is not oxidized at the inert anode; (d) bubbles from the inert anode reaction are removed and therefore do not cause problems to the cathode; and (e) changes in the conductivity and pH of the anolyte are isolated from the catholyte and therefore the impact of such changes on the catholyte are minimized.
During a plating cycle, the second processing chamber 120 is filled with the anolyte and an electrical potential is applied to the workpiece W and the counter electrode 160. As the copper ions are removed from the catholyte and plated onto the workpiece, sulfate ions SO42− accumulate in the catholyte near the first surface 132 of the ion exchange membrane 130. Additionally, the positively charged counter electrode 160 causes hydroxyl ions OH− at the inert counter electrode 160 to liberate oxygen and produce water. This creates a charge gradient that causes the negatively charged sulfate ions to move from the first surface 132 to the second surface 134 of the ion exchange membrane 130. The transfer of negatively charged sulfate ions from the catholyte to the anolyte during a plating cycle maintains the charge and mass balance of the system. Additionally, to maintain the concentration of boric acid in the catholyte during plating, the concentration of boric acid in the anolyte is significantly greater than that of the catholyte to prevent the boric acid in the catholyte from moving to the anolyte during plating. It should be noted that during an idle state the anolyte level is dropped to separate the anolyte from the ion exchange membrane to prevent boric acid from crossing the ion exchange membrane 130. The hydroxyl ions could also move from the catholyte to the anolyte during plating, but they are inhibited from moving by maintaining the at least approximately the same pH level in both solutions.
The operation of the reactor 100 in the example set forth in
Another feature of the operation of the reactor 100 in the example illustrated in
The first chamber 510 illustrated in
The support member 550 further includes a plurality of openings 556 that direct the flow F1 of the first processing solution upwardly toward a processing zone near the rim 514 of the inner vessel 512. The openings 556 are generally arranged in a pattern to provide a generally uniform flow of the first processing solution in a direction normal to a plane at the rim 514 of the inner vessel 512, but the openings 556 can be arranged to provide a non-uniform flow toward the workpiece. For example, the openings 556 can have a uniform size and distribution across the support member 550, or the support member 550 can have a first region with first openings and a second region with second openings that are configured to provide different flow rates and/or different field components of the electric field. The first and second openings can have different sizes and/or be arranged in different patterns. The openings 556 also act as virtual electrodes that define the shape of the electrical field in the lower portion of the inner vessel 512. To avoid creating high magnitude points in the electrical field at the processing zone, the spacing between the openings 556 can be correlated with the distance to the processing zone. For example, the distance to the processing zone can be at least the distance between the openings 556. In operation, the workpiece holder (not shown in
In an alternative embodiment, the support member in the first chamber 510 is a porous material that provides a highly uniform distribution of the electrical field and the flow of the first processing solution. The porous material can be a porous ceramic, a porous plastic, or any other suitable porous material. Although porous support members may provide a highly controlled distribution of the processing solution in the electrical field, they may also be a source of particles that can contaminate the workpiece. In still additional embodiments of the support member, the upper surface of the support member 550 can be contoured to shape the electrical field in the first chamber 510. For example, the center of the support member 550 can have a thickness that is less than the thickness at the perimeter to provide a higher current at the center of the wafer relative to the perimeter. This may enable more uniform plating on high resistance seed layers or allow a desired non-uniform plating shape (e.g., a dome or dish) for subsequent chemical-mechanical planarization. Additional embodiments of the support member 550 can include a non-uniform pattern of openings 556 to provide different flow rates and current densities in the first chamber 510. For example, the density or size of the openings can be greater at a center region of the support member 550 than at a perimeter region to increase the deposition rate at the center of the region relative to the perimeter region.
Referring back to
In operation, the second processing solution level can be lowered during an idle state to separate the second processing solution from the ion exchange membrane so that ions do not pass through the ion exchange membrane. For example, when the reactor 500 is operated using an anion membrane as described above with reference to
Referring to
The term “microfeature workpiece” is used throughout to include substrates upon which and/or in which microelectronic devices, micromechanical devices, data storage elements, micro-optics, and other features are fabricated. For example, microfeature workpieces can be semiconductor wafers, glass substrates, dielectric substrates, or many other types of substrates. Microfeature workpieces generally have at least several features with critical dimensions less than or equal to one micron, and in many applications the critical dimensions of the smaller features on microfeature workpieces are less than 0.25 micron or even less than 100 nanometers. Furthermore, the terms “electrochemical processing” and “electrochemical deposition” include both electroplating processes and electroless plating processes. Where the context permits, singular or plural terms may also include the plural or singular term, respectively. Moreover, unless the word “or” is expressly limited to mean only a single item exclusive from other items in reference to a list of at least two items, then the use of “or” in such a list is to be interpreted as including (a) any single item in the list, (b) all of the items in the list, or (c) any combination of the items in the list. Additionally, the term “comprising” is used throughout to mean including at least the recited feature(s) such that any greater number of the same features and/or types of other features and components are not precluded.
From the foregoing, it will be appreciated that specific embodiments of the invention have been described herein for purposes of illustration, but that various modifications may be made without deviating from the spirit and scope of the invention. For example, the reactors and systems disclosed herein are not limited for use with low conductivity, high pH solutions. In many applications, the processing solution can have only a low conductivity, only a high pH, or neither a low conductivity nor a high pH. Accordingly, the invention is not limited except as by the appended claims.
