Semiconductor ICF Target Processing

Abstract

A method of manufacturing a semiconductor ICF target is described. On an n-type silicon wafer a plurality of hard mask layers are etched to a desired via pattern. Then isotropically etching hemispherical cavities, lithographically patterning the hard mask layers, conformally depositing ablator/drive material(s) and shell layer material(s), inserting hollow silicon dioxide fuel spheres in the hemisphere cavities, thermally bonding a mating wafer with matching hemisphere cavities and etching in ethylene diamine-pryrocatechol-water mixture to selectively remove n-type silicon and liberate the spherical targets.

Claims

1. A process for manufacturing a plurality of individual spherical devices from a wafer of semiconductor material, comprising: creating a first wafer by: depositing one or more hard mask layers upon a single-crystal n-type silicon substrate; applying a uniform coat of photoresist pattern on top of the one or more hard mask layers; etching through the one or more hard mask layers with vias; forming hemispherical cavities using isotropic dry or wet etch of the n-type silicon substrate having one or more hard mask layers; using a Chemical Mechanical Planarization (CMP) process to polish and remove one or more of the hard mask layers; and incorporating a p-type dopant on a surface of the hemispherical cavities; placing hollow silicon dioxide fuel spheres within the hemispherical cavity; and filling the hollow silicon dioxide fuel spheres with a fusion fuel mixture.

2. The process of claim 1, further comprising: repeating the steps of claim 1 to create a second wafer, wherein the second wafer has the same pattern of hemispherical cavities as the first wafer; mating the first and second wafers; bonding the first and second wafers together using thermal and/or adhesive techniques; etching to selectively remove what remains from the single-crystal n-type silicon substrate to liberate a spherical device; and repeating these steps to produce a plurality of spherical devices.

3. The process of claim 2, wherein said single-crystal n-type silicon substrate is customized to a thickness ranging from approximately 500 μm to 5 cm.

4. The process of claim 3, wherein the plurality of hard mask layers are composed of one of the following materials: silicon dioxide, silicon nitride and chromium.

5. The process of claim 4, wherein the plurality of hard mask layers are each customized to a thickness ranging from approximately 200 Å to 100,000 Å.

6. The process of claim 5, further comprising: isotropically etching the hemispherical cavities with a sulfur hexafluoride dry etch.

7. The process of claim 6, wherein the diameter of said hemispherical cavities are formed in the range of approximately 100 μm to 5 cm.

8. The process of claim 7, further comprising: using thermally diffuse boron atoms in the p-type doping of the surface of the hemispherical cavity.

9. The process of claim 8, wherein the p-type doping of the surface of the hemispherical cavity is performed to a depth of approximately 1-200 μm for the thermally diffuse boron atoms.

10. The process of claim 9, further comprising; conformally depositing a shell layer on top of an ablator/drive layer on top of a surface of the hemispherical cavities; and removing any excess from the deposition of shell layer and ablator/drive layer by a planarization polish.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0012] FIG. 1 shows a silicon wafer of n-type conductivity with a plurality of deposited hard mask films.

[0013] FIG. 2 shows a photoresist coating and etched vias.

[0014] FIG. 3 shows isotropic etched hemispherical cavities.

[0015] FIG. 4 shows the lithographic patterning of the nitride CMP stop layer.

[0016] FIG. 5 shows hollow silicon oxide fuel spheres, shell, and ablator/drive region within a single crystal n-type substrate.

[0017] FIG. 6 shows a bonded wafer wherein a second wafer is turned upside down and mated with the original wafer where their flat surfaces are thermally bonded together.

[0018] FIG. 7 shows the hollow silicon oxide fuel spheres liberated from the single crystal n-type silicon layer.

[0019]

TABLE-US-00001 REFERNCE NUMERALS 100 Silicon Wafer of N-type Conductivity 102 Single-Crystal N-type Silicon Substrate 104 SiO.sub.2 Layer 106 Si.sub.3N.sub.4 Layer 108 Cr Layer 202 Photoresist Pattern 402 Boron P.sup.+ 502 Ablator/drive region 504 Shell 506 Hollow SiO.sub.2 Fuel Sphere 600 Bonded Wafer

DETAILED DESCRIPTION

[0020] FIGS. 1-6 show a cross-sectional view of a semiconductor ICF Target manufacturing process by the method according to the invention. FIG. 1 shows a silicon wafer of n-type conductivity 100 with a customizable thickness (ranging from 500 μm to 5 cm) for the single-crystal n-type silicon substrate 102. A plurality of hard mask layers may be deposited on top of the single-crystal n-type silicon substrate 102 such as the silicon dioxide layer 104, silicon nitride layer 106 and followed by a chromium layer 108. The thickness of these layers could range from 200 Å to 100,000 Å.

[0021] As the wafer is rotated at a high speed, a uniform coat of photoresist pattern 202 may then be applied to the surface of the wafer as seen in FIG. 2. Vias may be patterned and etched through one or more adjacent layers until the etching reaches the single-crystal n-type silicon substrate 102. These vias could range in diameter from 1 μm to 1 mm.

[0022] The photoresist layer may then be chemically stripped by a solvent to reveal a via pattern transferred from the mask. The via hard mask pattern can be used to provide a plurality of hemispherical cavity structures as shown in FIG. 3. This revealed hemispherical cavity pattern is formed by isotropic etching using a sulfur hexafluoride (SF.sub.6) dry etch or a silicon wet etch or other suitable isotropic etches. The diameter of these cavities could range from 100 μm to 5 cm. After the etching is complete, the photoresist may then be removed, and the hemispherical cavities become visible. The chromium (Cr) hard mask layer 108 could then be stripped away resulting in FIG. 3. Optionally, one could form a self-aligned flange layer by a controlled undercut of the silicon nitride layer 106 after the isotropic etching of the single-crystal n-type silicon substrate 102. This would allow for the fuel sphere (506, FIG. 5) to have an air-gap between it and the silicon shell (504, FIG. 5) if so desired.

[0023] Optionally a thermal oxidation and strip could be applied to smooth out the surface roughness of the single-crystal n-type silicon substrate 102 on the hemispherical cavity surface. In FIG. 4, a lithographic patterning step may be performed to etch selective regions of the nitride top layer to form the pattern as shown. Then the exposed portion of the silicon surface may be doped with p-type boron to a depth of approximately 1-200 μm using thermally diffuse boron atoms (approximately 5×10.sup.19 cm.sup.−3 peak doping concentration). In addition, a variety of other materials could be substituted for boron p+ for the freeing etch stop, including but not limited to silicon dioxide. Next the silicon dioxide layer 104 and silicon nitride layer 106 may be chemically stripped away.

[0024] Forming an ablator/drive region 502 and shell layers 504 above the p-type boron 402 is shown in FIG. 5. This is done by a conformal deposition of each layer followed by a planarization polish to remove the excess of both layers from the surface. An ablator/drive region in an ICF target is a region that when penetrated by a shock wave, a portion of that ablator layer may be ablated. It can range in thickness from 1 um to 3 cm. Similar thicknesses are possible for the shell layer. While there are many well known methods to polish and remove unwanted layers, using a Chemical Mechanical Planarization (CMP) process is one preferred approach to polish and remove unwanted layers from above the silicon nitride top layer, using the nitride layer as a polish stop 106. Then the silicon nitride layer 106 is stripped away. Optionally a patterned and etched fuel gas diffusion trench can be formed starting from the hemispherical cavity and going through the boron p+ region. A CMP process may be performed again to allow for a flat, smooth mating surface. Hollow silicon dioxide fuel spheres 506 are then placed within the hemispherical cavity.

[0025] Alternatively, the ablator/drive and shell material regions could be formed using the silicon dioxide and/or silicon nitride hard masks. Either evaporating a metal flash or depositing a titanium nitride (TiN) layer is then performed to act as a seed layer for electroplating. Excess seed metal is removed from above the silicon nitride layer using CMP. With the hemispherical cavities and silicon electrically conductive and the surface an insulator, electroplating can now be performed biasing the substrate in order to selectively deposit the ablator/drive and shell material regions on the hemispherical cavities.

[0026] Next, hollow silicon dioxide fuel spheres 506, located in the hemispherical cavities, may be filled with deuterium-tritium (DT) fuel or any one of a variety of other fusion fuel mixtures. A second wafer processed as above, without fuel spheres, is turned upside down and mated with the original wafer and their flat surfaces are thermally bonded together resulting in a bonded wafer 600, as seen in FIG. 6. Optionally, an adhesive layer (not shown) can be applied to one or both of the wafer surfaces to promote adhesion of the two wafers and their hemispheres. The two wafers must be accurately aligned to each other to form spheres from the matching hemispheres, which can be done by forming alignment marks on each and microscope viewed mechanical alignment, as known in the art.

[0027] FIG. 7 shows what remains after the single-crystal n-type silicon substrate 104 of the mated wafers is selectively removed by etching in an ethylene diamine-pyroccatechol-water mixture. The plurality of target spheres are liberated from the silicon substrates. These target spheres are composed of a hollow silicon dioxide fuel sphere 506 filled with a fusion fuel mixture, surrounded by a shell 504, ablator/drive region 502, and p-type boron doped silicon 402.

[0028] There are many advantages for this invention such as parallel processing for thousands of targets on a single wafer, target manufacture at reduced cost, using mature and available semiconductor high volume processing tools and processes, utilizing precision lithography pattern definition, accurate film thickness deposition, and flexibility with ablator/drive region and shell material choices.

[0029] Additionally, the set of embodiments discussed in this application is intended to be exemplary only, and not an exhaustive list of all possible variants of the invention. Certain features discussed as part of separate embodiments may be combined into a single embodiment. Additionally, embodiments may make use of various features known in the art but not specified explicitly in this application.

[0030] It should be noted that embodiments can be scaled-up and scaled-down in size, and relative proportions of components within embodiments can be changed as well. The range of values of any parameter (e.g. size, thickness, density, mass, composition, etc.) of any component of an embodiment of this invention, or of entire embodiments, spanned by the exemplary embodiments in this application should not be construed as a limit on the maximum or minimum value of that parameter for other embodiments, unless specifically described as such.