Plasma chamber target for reducing defect of workpiece in dielectric sputtering
Dielectric targets with controlled particle sizes and RF/DC cathode configurations address defects in sputtering processes by minimizing arcing and delamination, enhancing substrate quality and production efficiency.
Patent Information
- Application Number
- JP2024118527
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2020-07-01
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-06
AI Technical Summary
Sputtering processes in semiconductor manufacturing cause defects due to unwanted deposits on chamber shields leading to arcing and delamination, which affect substrate quality.
The use of dielectric targets with controlled particle sizes and configurations, combined with RF and DC cathodes, reduces defects by minimizing arcing and delamination during dielectric sputtering.
Significantly reduces defects on workpieces by enhancing target grain size uniformity and chamber design, resulting in faster production with fewer defects and improved substrate quality.
Smart Images

Figure 2024156735000001_ABST
Abstract
Description
[Technical field]
[0001] SUMMARY OF THE DISCLOSURE Embodiments of the present disclosure generally relate to substrate processing chambers used in semiconductor manufacturing systems. [Background technology]
[0002] Sputtering, also known as physical vapor deposition (PVD), is a method of forming features in integrated circuits that is typically performed in a process chamber. Sputtering deposits a layer of material, such as a dielectric material, onto a workpiece (e.g., substrate / substrate). A source material, such as a target, is bombarded with ions that are highly accelerated by an electric field. The ion bombardment ejects material from the target, which collects or deposits on the substrate (e.g., workpiece). During deposition, the ejected particles may also deposit on other surfaces, such as shields or other interior surfaces of the process chamber.
[0003] Unwanted coating on the shield can cause defects in the substrate being processed or can contribute to defects during subsequent substrate processing. Defects can occur, for example, when unwanted deposition occurs on the shield and the unwanted deposition accumulates charge causing arcing or spalling of dielectric material that has collected on the shield.
[0004] Thus, the inventors have provided improved target embodiments for reducing defects during dielectric sputtering. Summary of the Invention
[0005] Presented herein are methods and apparatus for plasma chamber targets for reducing workpiece defects during dielectric sputtering. For example, in some embodiments, the apparatus can include a dielectric sputter deposition target, which can include a dielectric compound having a predetermined average grain size in the range of about 65 μm to 500 μm, and the dielectric compound is at least one of magnesium oxide or aluminum oxide.
[0006] In at least some embodiments, the processing chamber includes a chamber body defining an interior volume, a substrate support supporting a substrate within the interior volume, and a plurality of targets to be sputtered onto the substrate including at least one dielectric target and at least one metallic target, the dielectric target including a dielectric compound having a predetermined average grain size in a range of about 65 μm to 500 μm.
[0007] In at least some embodiments, a method of performing physical vapor deposition in a process chamber can include selecting a first target in the process chamber and depositing a dielectric compound having a predetermined average grain size in a range of about 65 μm to 500 μm on a workpiece via the first target, and selecting a second target in the process chamber and depositing a metal over the dielectric compound on the workpiece via the second target.
[0008] Other, further embodiments of the present disclosure are described below.
[0009] Embodiments of the present disclosure, briefly summarized above and discussed in more detail below, can be understood by reference to illustrative embodiments of the disclosure, which are depicted in the accompanying drawings. The accompanying drawings, however, depict only typical embodiments of the present disclosure, and therefore should not be considered as limiting the scope of the present disclosure, as the present disclosure may admit of other equally effective embodiments. [Brief description of the drawings]
[0010] [Figure 1] 1 is a schematic diagram of a multi-cathode processing chamber according to an exemplary embodiment described herein. [Diagram 2] 2 illustrates a spatial arrangement of the chamber of FIG. 1 according to an exemplary embodiment described herein. [Diagram 3] 1 is a magnified image of a target having a particular grain size. [Figure 4]1 is a magnified image of a target having a grain size according to an example embodiment described herein. [Diagram 5] FIG. 1 is a flow diagram of a method for reducing defects on a workpiece according to an example embodiment described herein.
[0011] For ease of understanding, the same reference numerals have been used, where possible, to designate like elements common to each of the figures. The figures are not to scale and may be simplified for clarity. Elements and features of one embodiment may be beneficially incorporated in other embodiments even if not specifically described otherwise. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0012] Magnetic Random Access Memory (MRAM) requires a barrier layer to be constructed as part of the memory device. The barrier layer must be of high purity and have a low number of defects to operate properly. Magnesium oxide (MgO) and / or aluminum oxide (Al2O3) are dielectric materials that can be used as the barrier layer. However, the use of RF power to deposit MgO on the substrate surface inherently leads to poor defect performance.
[0013] Presented herein are embodiments of sputter deposition targets including a dielectric compound having a predetermined average grain size for reducing defects in a workpiece (e.g., a substrate) when manufacturing MRAM. As used herein, grain size refers to the average grain size throughout the target of the compound, and the grain size itself has a wide range of distribution. In some embodiments, each grain is within 20% of the predetermined grain size, while in other embodiments, each grain is within 30% of the average grain size (e.g., the predetermined grain size). According to some embodiments, at least 80% of the grains in the target, or in other embodiments at least 90%, have a grain size within 20% or 30% of the preferred grain size. The preferred grain size ranges from about 20 μm, 40 μm, and 50 μm to 500 μm for pure single crystals. In another embodiment, at least 80% of the grains in the target have a grain size of at least 20 μm.
[0014] During physical vapor deposition of a dielectric film on a substrate (e.g., a substrate) in a process chamber, dielectric powder can also form on the shield in the process chamber because the shield is closer to the target than the workpiece. When RF power is then applied to the process chamber, arcing can occur between the shield and the substrate due to the high concentration of positive ions on the shield, e.g., sheath voltage. In other words, the dielectric film accumulates on the shield, and because the dielectric film is not conductive, a charge builds up on the shield. This charge on the shield does not easily dissipate, which can lead to arcing. Furthermore, after multiple subsequent substrates are processed in the process chamber, more dielectric film can build up on the shield, and poor adhesion between the film and the shield can cause the film to flake off or peel off and deposit on the substrate. The inventors discovered that modifying the target grain size and the target and shield configuration reduced workpiece defects by preventing arcing and by preventing the dielectric film from peeling off from the shield.
[0015] The inventors have observed that when targets containing dielectric compounds (e.g., MgO, Al2O3) are used in the process chamber, the majority of defects on the workpiece are dielectric particle defects. For example, the inventors have observed that when a metal target (e.g., tantalum) is exposed in a shielded process chamber, the number of Ta defects is small. However, when a dielectric target is exposed in the chamber, a significant number of defects are found on the workpiece in the form of dielectric particles. The origin of the defects was verified by experiments such as energy dispersive X-ray spectroscopy (EDX), although other methods can be used. The shielding remained unchanged, and therefore the defects were confirmed to originate from the dielectric target.
[0016] In some embodiments, a multiple cathode PVD chamber (e.g., process chamber 100) includes multiple cathodes 106 with corresponding multiple targets (at least one dielectric target 110 and at least one metal target 112), (e.g., six cathodes in a 3RF×3DC alternating configuration) attached (e.g., via upper adapter assembly 142) to a chamber body 140 that defines an interior volume. Other RF / DC cathode configurations, such as 1:1, 2:2, 4:4, 5:5, etc., can also be used. The numbers indicate the ratio of RF-powered cathodes to DC-powered cathodes. In some embodiments, the RF and DC cathodes alternate in the upper adapter assembly 142. In other embodiments, an RF cathode can be adjacent to another RF cathode, as well as for a DC cathode. In still other embodiments, the ratio of RF and DC cathodes can be unequal, such as 1:2, 2:1, 1:3, 3:1, 2:3, etc. When multiple RF cathodes are used, the operating frequencies can be offset to reduce any interference during the deposition process. For example, in a three RF cathode embodiment, the first RF cathode can operate at a frequency of 13.56 MHz, the second RF cathode operates at a frequency of 13.66 MHz (+100 kHz), and the third RF cathode operates at a frequency of 13.46 MHz (-100 kHz). The offset can be chosen based on crosstalk prevention for a given number of cathodes.
[0017] An RF cathode is typically used with a dielectric target 110 for depositing the dielectric film on the substrate. A DC cathode is typically used with a metal target 112 for attachment after the dielectric film is deposited on the substrate. This attachment reduces the chance of particle formation and defects in the deposited film. Having a process chamber with RF and DC cathodes allows for faster production of substrates since attachment and dielectric deposition can occur in one chamber. In addition, having multiple cathodes of the same type allows for higher attachment and deposition rates. Higher deposition rates mean that the substrates spend less time in the chamber to achieve a given film thickness. Less time in the chamber or residence time results in fewer defects in the substrate.
[0018] In some embodiments, the metal target 112 may be formed of a metal, such as, for example, tantalum, aluminum, titanium, molybdenum, tungsten, and / or magnesium. The dielectric target 110 may be formed of a metal oxide, such as, for example, titanium oxide, titanium magnesium oxide, and / or tantalum magnesium oxide. However, other metals and / or metal oxides may be used instead.
[0019] The dielectric target 110 includes a dielectric compound having a predetermined particle size, where the particles are generally uniform throughout the dielectric target 110 with low size variation, e.g., 20-30%. The predetermined particle size varies in various embodiments. In some embodiments, the particle size is selected from a range of about 20 μm to 500 μm. In some embodiments, at least 80%, or in some embodiments at least 90%, of the particles in the target have a particle size equal to a preferred particle size (e.g., 20 μm, 40 μm, 50 μm, and up to about 500 μm or even greater).
[0020] FIG. 3 shows an exemplary target with a grain size of 8 μm, and the substrate contains 218 defects with a grain size greater than 40 nm. However, as shown in FIG. 4, increasing the grain size of the target to about 30 μm reduces the number of grains (defects), for example to 96. In some embodiments, the grain size is increased to about 40 μm, where the defects are estimated to be about 50-60 in grain count, a significant reduction from 218 defects. In other embodiments, targets have grain sizes of about 80 μm, 120 μm, 400 μm, 500 μm, and even more. While individual results for all of the grain sizes contemplated in these embodiments are not shown, FIGS. 3 and 4 illustrate the reduction in grain defects on a substrate that can be obtained by decreasing the target grain size, for example, of the dielectric target 110. While some embodiments described herein are directed to the grain sizes shown above, other grain sizes of about 20 μm or greater can also be used to reduce the number of defects on a substrate.
[0021] The number of grain defects decreases significantly as the grain size increases from 8 μm to at least 20 μm. This is because the surface area to volume ratio increases significantly with larger grain sizes, and the surface area to volume ratio is greater for smaller grains. As the grain size increases, the grain boundary area decreases, as can be seen in FIG. 4. The larger grain size results in a smaller grain boundary area and fewer defects. In other embodiments, the finish of the dielectric target 110 is modified. For example, the target can have a sputtering surface that ranges in surface roughness from about 6 μm surface roughness to a polished surface with a mirror finish (nearly no surface roughness). The inventors have discovered that the closer to a mirror finish that is achieved, the fewer grains are detected on the workpiece. In some embodiments, the dielectric target 110 has a density of at least 99.7% of pure single crystal MgO, or in some examples, at least 99.98% of pure single crystal MgO, to reduce defects in the workpiece. As used herein, target density refers to the ratio of the target to the theoretical pure single crystal density that can be theoretically obtained.
[0022] The processing chamber 100 also includes a substrate support 130 for supporting a substrate 132. The processing chamber 100 includes an opening (not shown) (e.g., a slit valve) through which an end effector (not shown) can extend to position the substrate 132 on lift pins (not shown) for lowering the substrate 132 onto a support surface 131 of the substrate support 130. In the embodiment shown in FIG. 1, the dielectric target 110 and the metal target 112 are disposed substantially parallel to the support surface 131. The substrate support 130 includes a bias source 136 that is coupled to a bias electrode 138 disposed on the substrate support 130 via a matching network 134. The upper adapter assembly 142 is coupled to an upper portion of a chamber body 140 of the processing chamber 100 and is grounded. Each cathode 106 can have a DC power source 108 or an RF power source 102 and an associated magnetron. In the case of an RF power source 102, the RF power source 102 is coupled to the cathode 106 via an RF matching network 104.
[0023] The shield 121 is rotatably coupled to the upper adapter assembly 142 and shared by each cathode 106. In some embodiments, the shield 121 includes a shield body 122 and a shield top 120. In other embodiments, the shield 121 has an aspect where both the shield body 122 and the shield top 120 are integrated into a single piece. In still other embodiments, the shield 121 can be more than two pieces. Depending on the number of targets that need to be sputtered simultaneously, the shield 121 can have one or more holes to expose the corresponding target or targets. The shield 121 advantageously limits or eliminates cross-contamination between multiple targets, including the dielectric target 110 and the metal target 112. The shield 121 is rotatably coupled to the upper adapter assembly 142 via a shaft 123. The shaft 123 is attached to the shield 121 via a coupler 119. In addition, because the shield 121 is rotatable, areas of the shield 121 that are not normally subject to attachment can move to allow the areas to be attached, thereby significantly reducing flaking of the build-up deposits and particle formation. The processing chamber 100 can also include at least one shunt disposed on the top surface of the shield 121 for containing and shielding at least another one of the multiple targets that is not to be sputtered.
[0024] An actuator 116 is coupled to shaft 123 opposite shield 121. Actuator 116 is configured to rotate shield 121, as indicated by arrow 144, and to move shield 121 vertically up and down along a central axis 146 of process chamber 100, as indicated by arrow 145. During processing, shield 121 is raised to an upper position. The raised position of shield 121 exposes targets used during processing and shields targets not used during processing. The raised position also grounds the shield for RF processing.
[0025] In some embodiments, the processing chamber 100 further includes a process gas supply 128 for supplying a process gas to the interior volume 125 of the processing chamber 100. The processing chamber 100 may also include an exhaust pump 124 fluidly coupled to the interior volume 125 for exhausting the process gas from the processing chamber 100. In some embodiments, for example, the process gas supply 128 may supply oxygen to the interior volume 125 after the metal target 112 has been sputtered.
[0026] FIG. 2 is a diagram of a spatial configuration of the processing chamber 100 of FIG. 1 according to an exemplary embodiment described herein.
[0027] According to some embodiments, the dielectric target 110 and the metal target 112 (along with any other targets) are positioned away from the edge of the shield 121, typically about 0.5 inches to 2.0 inches away from the edge of the shield 121. When the targets are positioned at the edge of the shield, the film deposited there tends to be more stressed than other areas, causing particles to flake off and deposit on the workpiece. When the targets are positioned towards the center of the shield 121, the film is less concentrated in a small area, which significantly reduces the stress on the film, resulting in less flake off and particle defect deposition on the workpiece.
[0028] FIG. 5 illustrates a flow diagram of a method 500 for depositing a film on a substrate according to an example embodiment described herein.
[0029] The method 500 begins at 502 and proceeds to 504. At 504, a first target is selected in the process chamber 100 and a dielectric compound (e.g., MgO, Al2O3) is deposited onto a workpiece (e.g., a semiconductor substrate) via the first dielectric target. For example, in at least some embodiments, the dielectric compound can be a single crystal target having a diameter of about 2 inches to about 6 inches. The predetermined particle size varies in various embodiments. For example, in some embodiments, the predetermined particle size is selected from a range of about 20 μm to about 500 μm. In some embodiments, the predetermined particle size can be about 65 μm to about 500 μm. In some embodiments, at least 80%, or in some embodiments at least 90%, of the particles in the target have a particle size equal to a preferred particle size (e.g., 20 μm, 40 μm, 50 μm, 400 μm, 500 μm, or even greater).
[0030] A second target is then selected in the process chamber 100 and a second metal is deposited through the second target onto the dielectric material on the workpiece at 506. The second target is typically a metal target, and the metal may be, for example, tantalum.
[0031] The processing chamber 100 is then ready for the next or subsequent workpiece and the method 500 ends at 508 .
[0032] While the forgoing is directed to embodiments of the present disclosure, other and further embodiments of the present disclosure may be devised without departing from the basic scope thereof.
Claims
1. 1. A dielectric sputter deposition target consisting essentially of a dielectric compound having a predetermined average grain size in the range of about 65 μm to 500 μm, A dielectric sputter deposition target, wherein the dielectric compound is at least one of magnesium oxide or aluminum oxide.
2. 2. The dielectric sputter deposition target of claim 1, wherein at least 80% of the grains of the dielectric compound are within 20% of the predetermined average grain size or within 30% of the predetermined average grain size.
3. The dielectric sputter deposition target of claim 1 , wherein the dielectric compound is single crystalline.
4. The dielectric sputter deposition target of any of claims 1 to 3, wherein said dielectric compound is magnesium oxide having a predetermined average grain size of about 400 μm.
5. 10. The dielectric sputter deposition target of claim 1, wherein the dielectric compound has a density of at least 99.7%.
6. The dielectric sputter deposition target of any of claims 1 to 3 or 5, wherein the dielectric compound has a density of at least 99.98%.
7. a chamber body defining an interior volume; a substrate support for supporting a substrate within the interior volume; and a plurality of targets to be sputtered onto the substrate, the targets including at least one dielectric target and at least one metal target, the dielectric target including a dielectric compound having a predetermined average grain size in the range of about 65 μm to 500 μm.
8. The processing chamber of claim 7 , wherein the dielectric compound is single crystalline.
9. 8. The processing chamber of claim 7, wherein the dielectric compound is magnesium oxide having a predetermined average grain size of about 400 μm.
10. The processing chamber of claim 7 , further comprising a plurality of cathodes coupled to the chamber body and corresponding to the plurality of targets.
11. 8. The processing chamber of claim 7, further comprising a shield coupled to the chamber body and having at least one hole for exposing at least one of the multiple targets to be sputtered, the multiple targets being positioned at least 0.5 inches from an edge of the shield.
12. The processing chamber of any of claims 7 to 11, wherein the shield is rotatably coupled to an upper portion of the chamber body.
13. the shield further comprising:
8. The processing chamber of claim 7, comprising at least one shunt disposed on a top surface of the shield for receiving and shielding at least another one of the plurality of targets that is not to be sputtered.
14. The processing chamber of any of claims 7 to 11 or 13, wherein the shield is configured to rotate about and move linearly along a central axis of the processing chamber.
15. 1. A method of performing physical vapor deposition in a process chamber, comprising: selecting a first target in the processing chamber and depositing a dielectric compound having a predetermined average grain size in a range of about 65 μm to 500 μm onto a workpiece through the first target; selecting a second target in the processing chamber and depositing a metal onto the dielectric compound on the workpiece through the second target.
16. The method of claim 15 , wherein the first target is a dielectric target and the second target is a metallic target.
17. 17. The method of claim 15 or 16, wherein the dielectric compound is monocrystalline magnesium oxide having a predetermined average grain size of about 400 μm.
18. 16. The method of claim 15, wherein the metal is tantalum.
19. The method of claim 15, wherein the first target is positioned about 0.5 inches to 2.0 inches from an edge of a shield of the processing chamber.
20. 20. The method of claim 15, 16, 18, or 19, wherein the first target has a mirror finish.
