Selective electrochemical deposition methods using pyrophosphate copper plating baths containing ammonium salts, citrate salts and/or selenium oxide

Abstract

An electrochemical fabrication process and apparatus are provided that can form three-dimensional multi-layer structures using pyrophosphate copper plating solutions that contain citrate salts, selenium oxide, and/or excess ammonium salts. In some embodiments the citrate salts are provided in concentrations that yield improved anode dissolution, reduced formation of pinholes on the surface of deposits, reduced likelihood of shorting between anode and cathode during deposition processes, and reduced plating voltage throughout the period of deposition. A preferred citrate salt is ammonium citrate in concentrations ranging from somewhat more that about 10 g/L for 10 mA/cm2 current density to as high as 200 g/L or more for a current density as high as 40 mA/cm. In some embodiments deposits having enhanced ductility and/or reduced tendency to crack are provided.

Description

RELATED APPLICATIONS

[0001] This application claims benefit of U.S. Provisional Patent Application No. 60/379,176, filed on May 7, 2002 which is hereby incorporated herein by reference as if set forth in full.

FIELD OF THE INVENTION

[0002] Embodiments of the invention relate to the field of electrochemical fabrication where single layer or multi-layer structures may be formed from a plurality of adhered layers, and more particularly where selective electrodepositions of material are formed using copper pyrophosphate plating baths and masks of the contact type or of the adhered type wherein material exchange around deposition regions is limited. Some embodiments relate to plating baths containing citrate salts, ammonium salts, and/or selenium oxide.

BACKGROUND

[0003] A technique for forming three-dimensional structures (e.g. parts, components, devices, and the like) from a plurality of adhered layers was invented by Adam L. Cohen and is known as Electrochemical Fabrication. It is being commercially pursued by MEMGen® Corporation of Burbank, Calif. under the name EFAB™. This technique was described in U.S. Pat. No. 6,027,630, issued on Feb. 22, 2000. This electrochemical deposition technique allows the selective deposition of a material using a unique masking technique that involves the use of a mask that includes patterned conformable material on a support structure that is independent of the substrate onto which plating will occur. When desiring to perform an electrodeposition using the mask, the conformable portion of the mask is brought into contact with a substrate while in the presence of a plating solution such that the contact of the conformable portion of the mask to the substrate inhibits deposition at selected locations. For convenience, these masks might be generically called conformable contact masks; the masking technique may be generically called a conformable contact mask plating process. More specifically, in the terminology of MEMGen® Corporation of Burbank, Calif. such masks have come to be known as INSTANT MASKS™ and the process known as INSTANT MASKING™ or INSTANT MASK™ plating. Selective depositions using conformable contact mask plating may be used to form single layers of material or may be used to form multi-layer structures. The teachings of the '630 patent are hereby incorporated herein by reference as if set forth in full herein. Since the filing of the patent application that led to the above noted patent, various papers about conformable contact mask plating (i.e. INSTANT MASKING) and electrochemical fabrication have been published:

[0004] 1. A. Cohen, G. Zhang, F. Tseng, F. Mansfeld, U. Frodis and P. Will, “EFAB: Batch production of functional, fully-dense metal parts with micro-scale features”, Proc. 9th Solid Freeform Fabrication, The University of Texas at Austin, p161, August 1998.

[0005] 2. A. Cohen, G. Zhang, F. Tseng, F. Mansfeld, U. Frodis and P. Will, “EFAB: Rapid, Low-Cost Desktop Micromachining of High Aspect Ratio True 3-D MEMS”, Proc. 12th IEEE Micro Electro Mechanical Systems Workshop, IEEE, p244, January 1999.

[0006] 3. A. Cohen, “3-D Micromachining by Electrochemical Fabrication”, Micromachine Devices, March 1999.

[0007] 4. G. Zhang, A. Cohen, U. Frodis, F. Tseng, F. Mansfeld, and P. Will, “EFAB: Rapid Desktop Manufacturing of True 3-D Microstructures”, Proc. 2nd International Conference on Integrated MicroNanotechnology for Space Applications, The Aerospace Co., April 1999.

[0008] 5. F. Tseng, U. Frodis, G. Zhang, A. Cohen, F. Mansfeld, and P. Will, “EFAB: High Aspect Ratio, Arbitrary 3-D Metal Microstructures using a Low-Cost Automated Batch Process”, 3rd International Workshop on High Aspect Ratio MicroStructure Technology (HARMST'99), June 1999.

[0009] 6. A. Cohen, U. Frodis, F. Tseng, G. Zhang, F. Mansfeld, and P. Will, “EFAB: Low-Cost, Automated Electrochemical Batch Fabrication of Arbitrary 3-D Microstructures”, Micromachining and Microfabrication Process Technology, SPIE 1999 Symposium on Micromachining and Microfabrication, September 1999.

[0010] 7. F. Tseng, G. Zhang, U. Frodis, A. Cohen, F. Mansfeld, and P. Will, “EFAB: High Aspect Ratio, Arbitrary 3-D Metal Microstructures using a Low-Cost Automated Batch Process”, MEMS Symposium, ASME 1999 International Mechanical Engineering Congress and Exposition, November, 1999.

[0011] 8. A. Cohen, “Electrochemical Fabrication (EFABTM)”, Chapter 19 of The MEMS Handbook, edited by Mohamed Gad-EI-Hak, CRC Press, 2002.

[0012] 9. “Microfabrication—Rapid Prototyping's Killer Application”, pages 1-5 of the Rapid Prototyping Report, CAD/CAM Publishing, Inc., June 1999.

[0013] The disclosures of these nine publications are hereby incorporated herein by reference as if set forth in full herein.

[0014] The electrochemical deposition process may be carried out in a number of different ways as set forth in the above patent and publications. In one form, this process involves the execution of three separate operations during the formation of each layer of the structure that is to be formed:

[0015] 1. Selectively depositing at least one material by electrodeposition upon one or more desired regions of a substrate.

[0016] 2. Then, blanket depositing at least one additional material by electrodeposition so that the additional deposit covers both the regions that were previously selectively deposited onto, and the regions of the substrate that did not receive any previously applied selective depositions.

[0017] 3. Finally, planarizing the materials deposited during the first and second operations to produce a smoothed surface of a first layer of desired thickness having at least one region containing the at least one material and at least one region containing at least the one additional material.

[0018] After formation of the first layer, one or more additional layers may be formed adjacent to the immediately preceding layer and adhered to the smoothed surface of that preceding layer. These additional layers are formed by repeating the first through third operations one or more times wherein the formation of each subsequent layer treats the previously formed layers and the initial substrate as a new and thickening substrate.

[0019] Once the formation of all layers has been completed, at least a portion of at least one of the materials deposited is generally removed by an etching process to expose or release the three-dimensional structure that was intended to be formed.

[0020] The preferred method of performing the selective electrodeposition involved in the first operation is by conformable contact mask plating. In this type of plating, one or more conformable contact (CC) masks are first formed. The CC masks include a support structure onto which a patterned conformable dielectric material is adhered or formed. The conformable material for each mask is shaped in accordance with a particular cross-section of material to be plated. At least one CC mask is needed for each unique cross-sectional pattern that is to be plated.

[0021] The support for a CC mask is typically a plate-like structure formed of a metal that is to be selectively electroplated and from which material to be plated will be dissolved. In this typical approach, the support will act as an anode in an electroplating process. In an alternative approach, the support may instead be a porous or otherwise perforated material through which deposition material will pass during an electroplating operation on its way from a distal anode to a deposition surface. In either approach, it is possible for CC masks to share a common support, i.e. the patterns of conformable dielectric material for plating multiple layers of material may be located in different areas of a single support structure. When a single support structure contains multiple plating patterns, the entire structure is referred to as the CC mask while the individual plating masks may be referred to as “submasks”. In the present application such a distinction will be made only when relevant to a specific point being made.

[0022] In preparation for performing the selective deposition of the first operation, the conformable portion of the CC mask is placed in registration with and pressed against a selected portion of the substrate (or onto a previously formed layer or onto a previously deposited portion of a layer) on which deposition is to occur. The pressing together of the CC mask and substrate occur in such a way that all openings, in the conformable portions of the CC mask contain plating solution. The conformable material of the CC mask that contacts the substrate acts as a barrier to electrodeposition while the openings in the CC mask that are filled with electroplating solution act as pathways for transferring material from an anode (e.g. the CC mask support) to the non-contacted portions of the substrate (which act as a cathode during the plating operation) when an appropriate potential and/or current are supplied.

[0023] An example of a CC mask and CC mask plating are shown in FIGS. 1(a)-1(c). FIG. 1(a) shows a side view of a CC mask 8 consisting of a conformable or deformable (e.g. elastomeric) insulator 10 patterned on an anode 12. The anode has two functions. FIG. 1(a) also depicts a substrate 6 separated from mask 8. One is as a supporting material for the patterned insulator 10 to maintain its integrity and alignment since the pattern may be topologically complex (e.g., involving isolated “islands” of insulator material). The other function is as an anode for the electroplating operation. CC mask plating selectively deposits material 22 onto a substrate 6 by simply pressing the insulator against the substrate then electrodepositing material through apertures 26a and 26b in the insulator as shown in FIG. 1(b). After deposition, the CC mask is separated, preferably non-destructively, from the substrate 6 as shown in FIG. 1(c). The CC mask plating process is distinct from a “through-mask” plating process in that in a through-mask plating process the separation of the masking material from the substrate would occur destructively. As with through-mask plating, CC mask plating deposits material selectively and simultaneously over the entire layer. The plated region may consist of one or more isolated plating regions where these isolated plating regions may belong to a single structure that is being formed or may belong to multiple structures that are being formed simultaneously. In CC mask plating as individual masks are not intentionally destroyed in the removal process, they may be usable in multiple plating operations.

[0024] Another example of a CC mask and CC mask plating is shown in FIGS. 1(d)-1(f). FIG. 1(d) shows an anode 12′ separated from a mask 8′ that comprises a patterned conformable material 10′ and a support structure 20. FIG. 1(d) also depicts substrate 6 separated from the mask 8′. FIG. 1(e) illustrates the mask 8′ being brought into contact with the substrate 6. FIG. 1(f) illustrates the deposit 22′ that results from conducting a current from the anode 12′ to the substrate 6. FIG. 1(g) illustrates the deposit 22′ on substrate 6 after separation from mask 8′. In this example, an appropriate electrolyte is located between the substrate 6 and the anode 12′ and a current of ions coming from one or both of the solution and the anode are conducted through the opening in the mask to the substrate where material is deposited. This type of mask may be referred to as an anodeless INSTANT MASK™ (AIM) or as an anodeless conformable contact (ACC) mask.

[0025] Unlike through-mask plating, CC mask plating allows CC masks to be formed completely separate from the fabrication of the substrate on which plating is to occur (e.g. separate from a three-dimensional (3D) structure that is being formed). CC masks may be formed in a variety of ways, for example, a photolithographic process may be used. All masks can be generated simultaneously, prior to structure fabrication rather than during it. This separation makes possible a simple, low-cost, automated, self-contained, and internally-clean “desktop factory” that can be installed almost anywhere to fabricate 3D structures, leaving any required clean room processes, such as photolithography to be performed by service bureaus or the like.

[0026] An example of the electrochemical fabrication process discussed above is illustrated in FIGS. 2(a)-2(f). These figures show that the process involves deposition of a first material 2 which is a sacrificial material and a second material 4 which is a structural material. The CC mask 8, in this example, includes a patterned conformable material (e.g. an elastomeric dielectric material) 10 and a support 12 which is made from deposition material 2. The conformal portion of the CC mask is pressed against substrate 6 with a plating solution 14 located within the openings 16 in the conformable material 10. An electric current, from power supply 18, is then passed through the plating solution 14 via (a) support 12 which doubles as an anode and (b) substrate 6 which doubles as a cathode. FIG. 2(a), illustrates that the passing of current causes material 2 within the plating solution and material 2 from the anode 12 to be selectively transferred to and plated on the cathode 6. After electroplating the first deposition material 2 onto the substrate 6 using CC mask 8, the CC mask 8 is removed as shown in FIG. 2(b). FIG. 2(c) depicts the second deposition material 4 as having been blanket-deposited (i.e. non-selectively deposited) over the previously deposited first deposition material 2 as well as over the other portions of the substrate 6. The blanket deposition occurs by electroplating from an anode (not shown), composed of the second material, through an appropriate plating solution (not shown), and to the cathode/substrate 6. The entire two-material layer is then planarized to achieve precise thickness and flatness as shown in FIG. 2(d). After repetition of this process for all layers, the multi-layer structure 20 formed of the second material 4 (i.e. structural material) is embedded in first material 2 (i.e. sacrificial material) as shown in FIG. 2(e). The embedded structure is etched to yield the desired device, i.e. structure 20, as shown in FIG. 2(f).

[0027] Various components of an exemplary manual electrochemical fabrication system 32 are shown in FIGS. 3(a)-3(c). The system 32 consists of several subsystems 34, 36, 38, and 40. The substrate holding subsystem 34 is depicted in the upper portions of each of FIGS. 3(a) to 3(c) and includes several components: (1) a carrier 48, (2) a metal substrate 6 onto which the layers are deposited, and (3) a linear slide 42 capable of moving the substrate 6 up and down relative to the carrier 48 in response to drive force from actuator 44. Subsystem 34 also includes an indicator 46 for measuring differences in vertical position of the substrate which may be used in setting or determining layer thicknesses and/or deposition thicknesses. The subsystem 34 further includes feet 68 for carrier 48 which can be precisely mounted on subsystem 36.

[0028] The CC mask subsystem 36 shown in the lower portion of FIG. 3(a) includes several components: (1) a CC mask 8 that is actually made up of a number of CC masks (i.e. submasks) that share a common support/anode 12, (2) precision X-stage 54, (3) precision Y-stage 56, (4) frame 72 on which the feet 68 of subsystem 34 can mount, and (5) a tank 58 for containing the electrolyte 16. Subsystems 34 and 36 also include appropriate electrical connections (not shown) for connecting to an appropriate power source for driving the CC masking process.

[0029] The blanket deposition subsystem 38 is shown in the lower portion of FIG. 3(b) and includes several components: (1) an anode 62, (2) an electrolyte tank 64 for holding plating solution 66, and (3) frame 74 on which the feet 68 of subsystem 34 may sit. Subsystem 38 also includes appropriate electrical connections (not shown) for connecting the anode to an appropriate power supply for driving the blanket deposition process.

[0030] The planarization subsystem 40 is shown in the lower portion of FIG. 3(c) and includes a lapping plate 52 and associated motion and control systems (not shown) for planarizing the depositions.

[0031] Another method for forming microstructures from electroplated metals (i.e. using electrochemical fabrication techniques) is taught in U.S. Pat. No. 5,190,637 to Henry Guckel, entitled “Formation of Microstructures by Multiple Level Deep X-ray Lithography with Sacrificial Metal layers. This patent teaches the formation of metal structure utilizing mask exposures. A first layer of a primary metal is electroplated onto an exposed plating base to fill a void in a photoresist, the photoresist is then removed and a secondary metal is electroplated over the first layer and over the plating base. The exposed surface of the secondary metal is then machined down to a height which exposes the first metal to produce a flat uniform surface extending across the both the primary and secondary metals. Formation of a second layer may then begin by applying a photoresist layer over the first layer and then repeating the process used to produce the first layer. The process is then repeated until the entire structure is formed and the secondary metal is removed by etching. The photoresist is formed over the plating base or previous layer by casting and the voids in the photoresist are formed by exposure of the photoresist through a patterned mask via X-rays or UV radiation.

[0032] When desiring to electroplate copper, various types of plating baths are known in the electrodeposition arts. One such standard bath type involves use of pyrophosphate copper. Atotech USA, Inc. of Somerset, N.J. supplies materials and recommendations for pyrophosphate copper plating baths under the name UNICHROME®. The UNICHROME pyrophosphate copper plating process is taught by Atotech in their technical information sheet revision 4/93. A range of components and optimum values of the UNICHROME plating bath are supplied in those teachings and are set forth in the following table (Table 1). 1 TABLE 1 UNICHROME ® Plating Solution Table Component or Condition Range Optimum Copper as metal 18.8 to 30.0 g/L (2.5 to 4.0 oz/gal 22.5 g/L (3 oz/gal) P2O7:Cu ratio 7.4 to 8.0:1 7.7:1 Ammonia as NH3 or 0.375 to 2.25 g/L (0.05 to 0.3 oz/gal) 1.20 g/L (0.16 oz/gal) Ammonium as 1.41 to 8.67 mL/L (0.18 to 1.11 fl. oz/gal) 5 mL/L NH4OH pH (electrometric) 8.0-8.5 8.3 Temperature 46° C. to 57° C. (115° F.-135° C.) 52° C.(125° F.) Cathode Current 2.2 to 3.2 A/dm2 (20-30 A/ft2) 2.7 A/dm2 (25 A/ft2) Density Anode Current 2.7 to 5.4 A/dm2 (25-50 A/ft2) 3.8 A/dm2 (35 A/ft2) Density

[0033] The primary components for Atotech's “optimum” bath are Atotech components C-11-Xb (35 mL/L) and C-10-Xb (333 mL/L). C-10-Xb supplies copper pyrophosphate while the C-11-Xb provides some form of pyrophosphate without copper. The “optimum” solution additionally includes 5 mL/L of NH4OH with the remainder being water. The solution also probably contains additives (included within the C-11-Xb and/or the C-10-Xb components) which are proprietary to Atotech and which are unknown to the present inventor.

[0034] General recommendations for pyrophosphate baths can be found in various sources. From the article “Copper Plating” in the ASM Handbook Vol. 5 (Surface Engineering, Ed. C. M. Cotell et al, ASM International, 1994), by L. M. Weisenberger and B. J. Durkin, and from “Electroplating: Fundamentals of Surface Finishing”, McGraw-Hill Book Co., 1978, by F. A. Lowenheim, generalized copper pyrophosphate plating solution components and operating conditions are believed to be as indicated in the following table (Table 2): 2 TABLE 2 Typical Components or Conditions Copper metal (comes from copper 19-30 g/L pyrophosphate Potassium pyrophosphate 235-405 g/L Copper pyrophosphate 52-84 g/L Ammonium hydroxide 3.75-11.0 mL/L Potassium nitrate 3.0-12.0 g/L Weight ratio (pyrophosphate:copper) 7:1-8:1 Additives As recommended Operating conditions Temperature, ° C. 43-60° C. Current density, mA/cm2 10-70 Anode and cathode efficiency, % ≈100 pH 8.0-8.7 Anode OFHC copper Anode/cathode area ratio 1:1-2:1 Agitation Air

[0035] From “Modern Electroplating”, 4th Edition, edited by Mordechay Schlesinger and Milan Paunovic, John Wiley & Sons, Inc., 2000, Part D, pages 122-133, the optimum range of constituents for a copper pyrophosphate plating solution are given in the following table (Table 3): 3 TABLE 3 Typical Components Copper, Cu2+ 22-38 g/L Pyrophosphate, (P2O7)4− 150-250 g/L Nitrate, NO3− 5-10 g/L Ammonia, NH3 1-3 g/L Orthophosphate, (HPO4)2− <113 g/L Additives As required Operating conditions Weight ratio (pyrophosphate:copper) 7.0:1 to 8.0:1 pH 8.0-8.8 Temperature 50-60° C. Cathode current density 1-8 A/dm2 Anode and cathode efficiency, % ≈100 Anode/cathode area ratio 1:1-2:1 Agitation - m3(min)−1(m−2 of surface) 1-1.5

[0036] From these three tables it can be seen that there isn't a single “optimal” bath formulation and set of operating conditions for all plating situations and further there is even some variability in the range of formulation parameters and use parameters for what may be considered “standard-type” plating situations. The references from which these tables have been extracted are hereby incorporated herein by reference as if set forth in full herein.

[0037] The '630 patent as well as the other CC masking and electrochemical fabrication publications noted above describe copper as a preferred material for deposition during the selective deposition process.

[0038] Even though pyrophosphate copper is used in normal plating operations and its use in CC mask plating and in electrochemical fabrication has been briefly described in the above references, it optimization for CC mask plating processes and electrochemical fabrication processes and particularly for micro-CC mask plating processes or micro-electrochemical fabrication processes have not been taught.

[0039] Though citrate salt containing plating solutions have been proposed in the literature, Applicants are unaware of any commercial use of such formulations. More particularly, Applicants are unaware of any formulations that have combined citrate solutions and copper pyrophosphate solutions and more particularly Applicants are unaware of such a combination being used to address difficulties in an electrochemical fabrication process that is used to construct three-dimensional structures from a plurality of adhered layers that contain feature sizes in the range of microns to hundreds of microns.

[0040] When using a plating bath formulated according to the Atotech's “optimal” teaching when practicing CC mask plating using CC masks with thin (e.g. less than about 100 microns) patterns of conformable material, problems with anode dissolution have occurred and depositions having pits extending to up to half the deposition thickness have occurred.

[0041] The '630 patent and each of the above noted publications that is explicitly directed to the CC mask plating and electrochemical fabrication processes, whether taken alone or in combination, are silent concerning such problems.

[0042] A need remains in the field of electroplating with CC masks, and particularly when conformable mask material is thin, and in electrochemical fabrication for improved processes that can allow more versatile structural formation (e.g. formation of regions with smaller lateral features), more rapid formation (e.g. by minimizing the amount of gross deposition thickness that is in excess of the desired layer thickness), more reliable formation (e.g. a reduction in the rate of shorting), enhanced process latitude associated with successful formation of structures, and/or more cost effective formation (e.g. less material waste, less waste of machine time, longer effective CC mask life) regardless of whether such characteristics will be explicitly required or experienced during any particular formation process.

SUMMARY OF THE INVENTION

[0043] It is an object of certain aspects of the invention to provide a CC mask plating or electrochemical fabrication process that allows for improved versatility when using copper as a selective deposition material.

[0044] It is an object of certain aspects of the invention to provide a CC mask plating or electrochemical fabrication process that allows for improved speed when using copper as a selective deposition material.

[0045] It is an object of certain aspects of the invention to provide a CC mask plating or electrochemical fabrication process that allows for enhanced reliability of structure formation when using copper as a selective deposition material.

[0046] It is an object of certain aspects of the invention to provide a CC mask plating or electrochemical fabrication process that has enhanced process latitude associated with the successful formation of structures when using copper as a selective deposition material.

[0047] It is an object of certain aspects of the invention to provide a CC mask plating or electrochemical fabrication process that is capable of being more cost effective when using copper as a selective deposition material.

[0048] Other objects and advantages of various aspects of the invention will be apparent to those of skill in the art upon review of the teachings herein. The various aspects of the invention, set forth explicitly herein or otherwise ascertained from the teachings herein, may address any one of the above objects alone or in combination, or alternatively may not address any of the objects set forth above but instead address some other object ascertained from the teachings herein. It is not intended that all of these objects be addressed by any single aspect of the invention even though that may be the case with regard to some aspects.

[0049] In a first aspect of the invention an electrochemical fabrication process for producing a three-dimensional structure from a plurality of adhered layers, includes: (A) selectively depositing at least a portion of a layer onto a substrate, wherein the substrate may comprise previously deposited material; and (B) forming a plurality of layers such that each successive layer is formed adjacent to and adhered to a previously deposited layer, wherein said forming comprises repeating operation (A) a plurality of times; wherein at least a plurality of the selective depositing operations comprise (1) contacting the substrate and a pattern mask; (2) in presence of a plating solution, conducting an electric current through the at least one opening in the selected mask between an anode and the substrate such that a selected deposition material is deposited onto the substrate to form at least a portion of a layer; and (3) removing the mask from the substrate; and wherein the selected deposition material is copper; and wherein the plating solution comprises pyrophosphate copper in combination with at least one of the following (a) an ammonium ion concentration greater than about 1.5 g/l, (b) ammonium ions and citrate ions, or (c) ammonium ions and tartrate ions.

[0050] In a variation of the first aspect, the process additionally includes supplying a plurality of preformed masks, wherein each mask includes a patterned dielectric material that includes at least one opening through which deposition can take place during the formation of at least a portion of a layer, and wherein each mask comprises a support structure that supports the patterned dielectric material, and wherein the contacting of the substrate and a patterned mask comprises contacting a selected one of the preformed masks to the substrate.

[0051] In a second aspect of the invention a conformable contact masking process for producing a structure, includes (A) supplying at least one preformed mask that comprises a patterned dielectric material that includes at least one opening through which deposition can take place during the formation of at least a portion of a layer, and wherein the at least one mask comprises a support structure that supports the patterned dielectric material; and (B) selectively depositing at least a portion of a layer onto a substrate, comprising (1) contacting the substrate and the dielectric material of the preformed mask; (2) in presence of a plating solution, conducting an electric current through the at least one opening in the selected mask between an anode and the substrate such that the selected deposition material is deposited onto the substrate to form at least a portion of a layer; and (3) separating the selected preformed mask from the substrate; wherein the selected deposition material is copper; and wherein the plating solution comprises pyrophosphate copper in combination with at least one of the following (a) an ammonium ion concentration greater than about 1.5 g/l, (b) ammonium ions and citrate ions, or (c) ammonium ions and tartrate ions.

[0052] In specific variations of the first two aspects of the invention, the dielectric comprises a conformable material.

[0053] In other variations a source of citrate ions comprises a citrate salt. In further variations the citrate salt comprises at least one of (1) potassium citrate, (2) sodium citrate, (3) potassium sodium citrate, or (4) ammonium citrate, (5) ammonium hydrogen citrate.

[0054] In other variations one of the following conditions is met (1) the citrate salt is supplied in a concentration that is effective in holding the plating voltage at a reduced level substantially during the entire duration of a plating operation; (2) the citrate salt is supplied in a concentration that is effective in substantially reducing the occurrence of shorting during plating operations that achieve the desired thickness of deposition; (3) the citrate salt is supplied in a concentration that is effective in achieving a uniformity of deposit that has a minimum thickness which is no less than 50% of a maximum thickness.

[0055] In other variations a source of tartrate ions comprises a tartrate salt. In further variations the tartrate salt comprises at least one of (1) potassium tartrate, (2) sodium tartrate, (3) potassium sodium tartrate, or (4) ammonium tartrate, or (5) ammonium hydrogen tartrate.

[0056] In still other variations the selenium oxide is supplied in a concentration that is effective in reducing cracking to an acceptable level. In further variations the reduction in cracking substantially eliminates cracking that contacts patterned features. In even further variations the reduction in cracking substantially eliminates cracking.

[0057] In additional variations the concentration of ammonium ions is sufficient to allow plating to occur at a current density of about 20 mA/cm2 or greater in achieving a desired plating thickness. In even further variations the concentration of ammonium ions is sufficient to allow plating to occur at a current density of about 30 mA/cm2 or greater in achieving a desired plating thickness.

[0058] In additional variations the concentration of ammonium ions, in combination with any citrate or tartrate ions, is sufficient to allow a desired plating thickness to be achieved without the voltage increasing by a factor of two or more prior to a desired plating thickness or plating time to be reached.

[0059] In additional variations the concentration of ammonium ions, in combination with any citrate or tartrate ions, is sufficient to allow a desired plating thickness or plating time to be reached without the occurrence of shorting on at least an average of 50% of attempts to form depositions of desired thickness during the formation of a given structure.

[0060] In even additional variations the concentration of ammonium ions, in combination with any citrate or tartrate ions, is sufficient to allow a desired plating thickness or plating time to be reached without the occurrence of shorting on at least 80% of attempts to form depositions of desired thickness during the formation of a given structure.

[0061] Further aspects of the invention will be understood by those of skill in the art upon reviewing the teachings herein. Other aspects of the invention may involve combinations of the above noted aspects of the invention and/or addition of various features of one or more embodiments. Other aspects of the invention may involve apparatus that can be used in implementing one or more of the above method aspects of the invention. These other aspects of the invention may provide various combinations of the aspects presented above as well as provide other configurations, structures, functional relationships, and processes that have not been specifically set forth above.

BRIEF DESCRIPTION OF THE DRAWINGS

[0062] FIGS. 1(a)-1(c) schematically depict side views of various stages of a CC mask plating process, while FIGS. 1(d)-(g) schematically depict a side views of various stages of a CC mask plating process using a different type of CC mask.

[0063] FIGS. 2(a)-2(f) schematically depict side views of various stages of an electrochemical fabrication process as applied to the formation of a particular structure where a sacrificial material is selectively deposited while a structural material is blanket deposited.

[0064] FIGS. 3(a)-3(c) schematically depict side views of various example subassemblies that may be used in manually implementing the electrochemical fabrication method depicted in FIGS. 2(a)-2(f).

[0065] FIGS. 4(a)-4(i) schematically depict the formation of a first layer of a structure using adhered mask plating where the blanket deposition of a second material overlays both the openings between deposition locations of a first material and the first material itself.

[0066] FIG. 5 provides and image of an anode surface after CC mask plating with a plating solution lacking ammonium hydrogen citrate.

[0067] FIG. 6 depicts a deposition having many pinholes that result from use of a plating solution lacking ammonium hydrogen citrate.

[0068] FIG. 7 depicts a deposition resulting from a plating solution containing ammonium hydrogen citrate.

[0069] FIG. 8 provides an image of the anode surface after CC mask plating when the plating solution contained ammonium hydrogen citrate.

[0070] FIG. 9 provides a plot of plating bath voltage versus time for 10 different plating baths some having ammonium hydrogen citrate and some lacking ammonium hydrogen citrate.

[0071] FIG. 10 provides a plot of voltage versus time for four plating baths having different concentrations of ammonium hydrogen citrate and having CC mask conformable material thicknesses less than those used for generating the curves of FIG. 9.

[0072] FIG. 11 provides a picture of a structure formed by electroplating which includes a number of circular features and where the deposit has formed numerous cracks that extend through the structure many of which extend between features of the structure.

[0073] FIG. 12 provides a plot of plating voltage versus plating time for three modified versions of a base plating bath (CU1000).

[0074] FIG. 13 provides a plot of plating voltage versus plating time for a primary plating bath (CU1000) and for three variations of that plating bath where the modified baths show an increase in voltage during the plating operation prior to a plating time reaching a desired 3600 seconds.

[0075] FIG. 14 provides a plot of plating voltage versus plating time for three further modified versions of the CU1000 plating bath, one of which shows an increase in voltage prior to reaching a desired plating time.

[0076] FIG. 15 provides a plot of plating voltage versus plating time for a number of pe plating baths that have been modified to include different concentrations of selenium oxide.

[0077] FIG. 16 provides a plot of plating voltage versus plating time up until shorting occurs between anode and cathode.

DETAILED DESCRIPTION

[0078] FIGS. 1(a)-1(g), 2(a)-2(f), and 3(a)-3(c) illustrate various features of one form of electrochemical fabrication that are known. Other electrochemical fabrication techniques are set forth in the '630 patent referenced above, in the various previously incorporated publications, in various other patents and patent applications incorporated herein by reference, still others may be derived from combinations of various approaches described in these publications, patents, and applications, or are otherwise known or ascertainable by those of skill in the art from the teachings set forth herein. All of these techniques may be combined with those of the various embodiments of various aspects of the invention to yield enhanced embodiments. Still other embodiments be may derived from combinations of the various embodiments explicitly set forth herein.

[0079] FIGS. 4(a)-4(i) illustrate various stages in the formation of a single layer of a multi-layer fabrication process where a second metal is deposited on a first metal as well as in openings in the first metal where its deposition forms part of the layer. In FIG. 4(a), a side view of a substrate 82 is shown, onto which patternable photoresist 84 is cast as shown in FIG. 4(b). In FIG. 4(c), a pattern of resist is shown that results from the curing, exposing, and developing of the resist. The patterning of the photoresist 84 results in openings or apertures 92(a)-92(c) extending from a surface 86 of the photoresist through the thickness of the photoresist to surface 88 of the substrate 82. In FIG. 4(d), a metal 94 (e.g. nickel) is shown as having been electroplated into the openings 92(a)-92(c). In FIG. 4(e), the photoresist has been removed (i.e. chemically stripped) from the substrate to expose regions of the substrate 82 which are not covered with the first metal 94. In FIG. 4(f), a second metal 96 (e.g., silver) is shown as having been blanket electroplated over the entire exposed portions of the substrate 82 (which is conductive) and over the first metal 94 (which is also conductive). FIG. 4(g) depicts the completed first layer of the structure which has resulted from the planarization of the first and second metals down to a height that exposes the first metal and sets a thickness for the first layer. In FIG. 4(h) the result of repeating the process steps shown in FIGS. 4(b)-4(g) several times to form a multi-layer structure are shown where each layer consists of two materials. For most applications, one of these materials is removed as shown in FIG. 4(i) to yield a desired 3-D structure 98 (e.g. component or device).

[0080] Though the embodiments discussed herein are primarily focused on conformable contact masks and masking operations, the various embodiments, alternatives, and techniques disclosed herein may have application to proximity masks and masking operations (i.e. operations that use masks that at least partially selectively shield a substrate by their proximity to the substrate even if contact is not made), non-conformable masks and masking operations (i.e. masks and operations based on masks whose contact surfaces are not significantly conformable), and adhered masks and masking operations (masks and operations that use masks that are adhered to a substrate onto which selective deposition or etching is to occur as opposed to only being contacted to it).

[0081] A basic standard plating configuration (i.e. non-CC mask plating configuration) includes an anode and a cathode which are immersed in a plating bath. The distance between the anode and cathode is at least 1 mm. A power source provides a pre-set current passing through the plating cell so that the anode metal usually dissolves into the plating bath and the metal ions in the plating bath are reduced at the cathode to become a metallic deposit. Depending on various parameters, including the composition of the plating bath, the plating bath is usually operated at a constant temperature some wherein the range of between 20-60° C. The plating bath is agitated mechanically or by compressed air to ensure that fresh plating solution is delivered to the cathode and that the products of the electrochemical reactions are removed from the electrodes into the bulk solution.

[0082] Through-mask plating is a selective plating process since the substrate (cathode) is patterned by a thin non-conductive material (e.g. a patterned photoresist). Otherwise, its plating configuration is the same as that of standard plating process as outlined above. As such, through-mask plating, for the purposes herein, may be considered a selective form of standard plating.

[0083] CC mask plating is different from normal and through-mask plating in several aspects. In one form of CC mask plating, the plating bath is trapped in a closed volume defined by the substrate, the side walls of the conformable material, and the anode. Examples of such closed volumes 26a and 26b are depicted in FIG. 1(b). Another form of CC mask plating may involve the use of a porous perforated support and a distal anode. FIGS. 1(d)-1(g) illustrate this form of CC mask plating. In this alternative form of CC mask plating, the barrier presented by the support portion of the CC mask, though allowing at least some ion exchange, may present a sufficient hindrance to the exchange of some components of the plating solution that the solution in the deposition region may still be considered to be substantially isolated from the bulk solution. This trapping results in little or no mass exchange between the volume of solution in the plating region and the bulk solution and as such no or little fresh solution with proper additives can be supplied into the microspace and no or little reaction products can be removed.

[0084] A preferred form of CC mask plating involves closed volumes where at least one of the dimensions of at least one of the plating volumes is on the order of tens of microns (e.g. 20 to 100 &mgr;m) or less. As such, this form of CC mask plating may be considered to be a microbath plating process (i.e. micro-CC mask plating).

[0085] In micro-CC mask plating, the preferred separation between the anode and cathode is presently between about 20 &mgr;m and about 100 &mgr;m, and more preferably between about 40 and 80 &mgr;m. As such, regardless of the size of the area being deposited, these preferred embodiments may be considered to be micro-CC mask plating processes. Of course thinner separation distances (e.g. 10 &mgr;m or less) and thicker separation distances (e.g. 300 &mgr;m or more) are possible. Due to this close spacing between anode and cathode, deposition processes at the cathode and dissolution processes at the anode, unlike standard plating, are highly interacting. This is believed to also be true for proximity mask plating and adhered mask plating where an anode is placed in proximity.

[0086] Agitating the plating bath, as is common with standard plating processes, though possible, is not necessarily desirable in electrochemical fabrication due potentially to the high interaction between anode and cathode processes and due to the believed enhanced risk of shorting when agitation is used. Shorting refers to a portion of the deposition height bridging the space between the cathode and the anode prior to the lapse of the desired deposition time, in which case the current is directed almost solely through deposited conductive material as opposed to flowing primarily through the plating bath as intended such that the continuing of deposition is inhibited.

[0087] Using a pyrophosphate bath at high temperature (i.e. above around 43° C. to 45° C.), though recommended in the standard plating processes, is not desirable in the current form of micro-CC mask plating due to the higher rate of attack at the interface between the CC mask support and the conformable material and the associated shortening of CC mask life. The preferred temperature range is between about room temperature and about 35° C.

[0088] CC mask plating has its own characteristics and the conventional wisdom associated with standard plating processes may be more of a hindrance than a help in developing commercially viable CC mask plating processes and systems. The following table (Table 4) provides a detailed comparison of various aspects of the two forms of standard plating (i.e. non-selective and through-mask plating) and micro-CC mask plating. 4 TABLE 4 Standard Plating Through-mask Micro-CC mask Characteristic Normal plating plating plating Patterned cathode No Yes Yes Separation between Macroscale Macroscale Microscale, anode and cathode <˜80 &mgr;m Closed microbath No No Yes plating Bath agitation Useful & Easy Useful & Easy Possible but problematic Bath heating Useful & Easy Useful & Easy Possible but problematic Fresh bath supplied Yes Yes No to electrodes Products removed Yes Yes No from electrodes Cathode:anode area Variable Variable 1:1 ratio Plating time Not limited Not limited Limited Interaction between Insignificant Insignificant Significant cathode and anode

[0089] Experiments were performed to test the effectiveness of copper pyrophosphate plating solutions for use in CC mask plating and in electrochemical fabrication. Experiments were performed with a pyrophosphate copper solution that was formulated according to the Atotech's “optimum” recommendations for the UNICHROME Pyrophosphate Copper Plating Process. The bath was formulated using (1) 35 mL/L of C-11-Xb, (2) 333 mL/L of C-10-Xb, and (3) 5 mL/L of NH4OH, with the remainder of the solution being H2O. It is believed that the Atotech supplied components contained various additives but it is not known precisely what these additives might be. It is possible that one of the additives is a nitrate such as potassium nitrate as it is generally considered a standard component in pyrophosphate baths. To minimize contamination risk, the solution was purified by being pumped through a 1.0 &mgr;m filter device (POLYCAP™ HD from Whatman) before being used for CC mask plating. The resulting solution matched Atotech's “optimum” formulation recommendations set forth previously.

[0090] The function of each of the primary components of the pyrophosphate copper plating bath may be briefly summarized as follows. Copper pyrophosphate, Cu2P2O7, dissolves in a potassium pyrophosphate (K4P2O7) solution forming the complex ion Cu(P2O7)26− which is the source of copper ions in the plating bath. The bath pH is important since at pH values above 11, Cu(OH)2 precipitates, while either CuH2P2O7 or Cu2P2O7 precipitate at a pH below 7. Between pH values of 7 and 11, the bath is relatively stable, but undergoes slow hydrolysis as defined by:

P2O74−+H2O→2PO43−+2H+ (or 2HPO42−)

[0091] The concentration limit of orthophosphate (PO43−) is about 100 g/L. Above this limit, the conductivity and bright-plating range suffer. Improper operation such as low pH, high P2O7:Cu ratio, or temperature above 60° C. will also build up orthophosphate. Since there is no way to remove orthophosphate from the bath, the bath will eventually need to be discarded. Nitrate ions in the bath can permit a higher limiting current density since they act as hydrogen acceptors to reduce cathode polarization. A small amount of ammonia in the bath produces more uniform and lustrous deposits and improves anode dissolution. Organic additives within controlled, limited concentrations refine the grain structure, impart leveling characteristics to the plating bath, and act as brighteners. Although some baths are operated without additives, most commercial pyrophosphate baths employ proprietary materials.

[0092] The electrochemical deposition reactions at the cathode and the dissolution reaction at the anode can be described as follows:

[0093] At the cathode:

Cu(P2O7)26−+e−→Cu(P2O7)27−

Cu(P2O7)27−+e−=Cu+2P2O74−

[0094] At the anode:

Cu−2e−→Cu2+ (via a multistep mechanism)

Cu2++2P2O74−→Cu(P2O7)26−

[0095] When the above noted “optimum” plating solution was used in the CC mask plating process, an unexpected result occurred. As indicated in the background section of this application, when using a plating bath formulated according to the Atotech's “optimal” teachings when practicing CC mask plating using CC masks with thin (e.g. less than about 100 microns) patterns of conformable material, problems with anode dissolution have occurred and depositions having pits extending to up to half the deposition thickness have occurred. FIG. 5 depicts an example of a copper anode after about 60 minutes of CC mask plating at about 10 mA/cm2 of current density. A brown passive film has formed on the anode and the anode dissolution is not uniform. When plating with this “optimal” solution, typical plating voltages (i.e. the voltage between the anode and cathode) have been observed to be between about 1.4-1.7 volts. The deposits resulting form this plating solution have also been observed to have many pinhole-like imperfections. An example of such imperfections is shown in FIG. 6. This figure depicts a copper deposit on a nickel substrate. A number of black spots 102 are visible in the Figure along with a plethora of small pin-like indentations or dimples 104. The black spots represent nodules (i.e. bumps) that are believed to be the result of contamination while the dimples are an imperfection in the deposition itself. It has also been observed that when plating with this solution a significant rate of premature shorting occurs (i.e. the deposit has a non-uniform deposition height with a protrusion extending to the anode that shorts the flow of current prior to the deposit as a whole reaching its desired height).

[0096] Solutions to these problems have been found. Applicants have observed that if ammonium hydrogen citrate is added to the plating solution these problems are greatly reduced. It is believed that the ammonium hydrogen citrate (a) acts as a secondary complexing agent for the copper (i.e. a substance that helps keep copper in solution) to stabilize the bath, (b) acts to help dissolve copper from the anode more uniformly, (3) allows an increase in the plating current density, and (d) increases deposit brightness.

[0097] Applicants' additional observations and experimentation have shown that ammonium ions play a stronger roll in solving these problems than the citrate does but they also indicate that citrate does play a roll as well. Applicants' additional observations and experiments have linked the presence of ammonium ions to microcracking and apparent brittleness of deposits and the presence of citrate as a reducer of these effects. Applicants' further observations and experiments have shown that use of selenium oxide (SeO2) without or without the presence of citrate has even a stronger influence on the reduction of microcracking and brittleness.

[0098] It has been observed that the quantity of ammonium hydrogen citrate can be adjusted to optimize the effectiveness of the plating solution. Tests have indicated when plating when using a current density of 10 mA/cm2 with the conformable mask material having a thickness of around 75 microns, a 15 g/L concentration of ammonium hydrogen citrate is effective in yielding good deposits with uniform anode dissolution and with greatly reduced risk of shorting. FIG. 7 provides an image of a copper deposit made with the modified plating bath (i.e. a plating bath having the Atotech “optimal” formulation but additionally including 15 g/L of ammonium hydrogen citrate (i.e. (NH4)2HC6H5O7)). FIG. 7 depicts a smooth and flat deposit along with a few black spots 102 which are believed to be contaminates or the result of contaminates. FIG. 8 provides an image of an anode surface after 60 minutes of plating using a current density of 10 mA/cm2 and using the modified solution. FIG. 8 is shown at the same level of magnification as that used in FIG. 5. FIG. 8 illustrates the improved condition of the anode as a result of plating using the modified solution. The following table (Table 5) provides results from testing of several solutions containing varying amounts of ammonium hydrogen citrate under different operating conditions: 5 TABLE 5 Concentration of Ammonium Thickness of Hydrogen Conformable Current Density of Current Density of Current Density of Citrate Material 10 mA/cm2 15 mA/cm2 20 mA/cm2 15 g/L 75 &mgr;m Good Deposit (e.g. Acceptable Deposit Acceptable deposit an initial coating (e.g. an initial but significant depth of 12-13 &mgr;m coating depth of 14- pitting Worse than probably yields a 15 &mgr;m probably results at 15 good 8 &mgr;m layer) yields a good 6-7 mA/cm2 No shorts &mgr;m layer) Significant level of Uniform Dissolution Some shorts shorting of Anode Some pitting of the anode but grain is still visible) 20 g/L Better than above but still suffers from excessive amounts of shorting 30 g/L Worked well (similar to 15 g/L at 10 mA/cm2 50 g/L Worked well also

[0099] FIG. 9 depicts a plot of plating voltage versus time for two identically controlled experiments using plating baths having the Atotech “optimal” formulation and eight identically controlled experiments using a plating bath having the Atotech formulation along with 15 g/L of ammonium hydrogen citrate. The two upper curves 122 and 124 are derived from the two identically controlled experiments using the Atotech “optimal” formulation. As can be seen, even though these experiments were identically controlled, differences obviously existed and it may be concluded that the use of the Atotech optimal formulation is accompanied by a very small process latitude. On the other hand, the curve at the bottom 126 actually represents the data from eight identical experiments using the formulation modified with the ammonium hydrogen citrate. The use of the ammonium hydrogen citrate not only improved the process latitude of the operation, it also lowered significantly the plating voltage of the process. The plating voltage was reduced from about a range of about 1.4 to 1.7 volts to a range of about 0.5 to 0.7 volts. It is believed that this reduction in voltage is a useful indicator of the effectiveness of the combination of (1) the plating current density, (2) the gap between the anode and the cathode (i.e. the thickness of the conformable material), and (3) the quantity of ammonium hydrogen citrate present so that improved anode dissolution is achieved (e.g. more uniformity), improved deposition (e.g. reduction in or complete elimination of dimples), and reduction in premature shorting (i.e. shorting prior to the deposition as a whole reaching its desired thickness).

[0100] Further experimentation showed an interesting effect. The separation between the anode and cathode had an affect on the effectiveness of the added ammonium hydrogen citrate. Even though it was demonstrated that 15 g/L of ammonium hydrogen citrate was effective in yielding improved plating when the separation was about 70 &mgr;m, when the separation was reduced to 27 &mgr;m the ammonium hydrogen citrate seemed to loose its effectiveness after about 5-6 minutes after which a jump in voltage occurred. In the small trapped volumes that exists in some embodiments of CC mask plating, it is possible that the ammonium hydrogen citrate is consumed. When the concentration of ammonium hydrogen citrate was increased to about 50 g/L, the effectiveness of the ammonium hydrogen citrate remained until shorting occurred. These results are illustrated in FIG. 10 where curves 132 and 134 represent experiments with a concentration of 15 g/L and curves 136 and 138 represent experiments performed with a concentration of 50 g/L. As can be seen in curves 132 and 134 a jump in voltage occurred prior to the eventual shorting that terminated each plating operation (which is indicated by the drop in voltage at the end of each curve).

[0101] From Applicants' experimentation the following general rules may be stated: (1) the higher the ammonium hydrogen citrate concentration the higher the current density that can be handled, (2) higher ammonium hydrogen citrate concentrations do not cause plating problems when lower current densities are used, (3) the thinner the gap between the anode and the cathode, the higher the concentration of ammonium hydrogen citrate that is needed. For the experiments performed based on the Atotech “optimal” formulation and anode to cathode spacings of about 70 &mgr;m, it was determined that

[0102] 10 mA/cm2, required more than about 10 g/L to get a good deposit (e.g. about a 15 &mgr;m coating thickness that can yield a layer thickness of about 8 &mgr;m with no shorting);

[0103] 20 mA/cm2, 30 g/L worked well;

[0104] 30 mA/cm2, 100 g/L worked well; and

[0105] 40 mA/cm2, 200 g/L worked well.

[0106] In addition to ammonium hydrogen citrate being an effective additive to improve the results of the plating operations, it is believed that potassium citrate, sodium citrate, or potassium sodium citrate, as well as variations thereof will also help reduce plating problems with or without use of a suitable ammonium hydroxide. It is also believed that ammonium citrate would be as effective as ammonium hydrogen citrate in addressing the above noted problems. Furthermore, due to its similar structure it is believed that tartrate salts will also be effective additives. These salts include ammonium tartrate (C4H4O6(NH4)2), sodium tartrate, potassium tartrate, and sodium potassium tartrate, as well as variations thereof.

[0107] Experiments with the modified solutions showed an extension of the useful pH range under which plating could occur. Experiments showed that acceptable results could be obtained with pH as low as 7.9 and as high as 9.3. It is possible that the effective range is even broader as experimentation showed no failure.

[0108] As noted above, the addition of various concentrations of ammonium hydrogen citrate have been added to Atotech's “optimal” formulation (designated herein as Cu1000) to get uniform copper deposits from IM plating. Copper anodes can be electrochemically dissolved in baths modified in this manner. A particular plating bath which contains Cu1000 plus 15 g/l ammonium hydrogen citrate for purposes of description herein has been labeled Cu1001.

[0109] Though deposits formed using Cu1001 provide drastic improvements over Cu1000 deposits with regard to some problems (thus making the modified solution useful for some embodiments), it has been found that CC mask plated deposits from the Cu1001 bath were not without problems. Under SEM examination micro-cracks in deposits have been observed. In fact the evacuation of the SEM chamber has proved to be a reliable diagnostic for not only observing cracks but also for producing them. An example of a deposit representing a solid layer with a number of holes extending there through is shown in FIG. 11. In this Figure a number of very small cracks 150 (i.e. micro-cracks) can be observed. It has been observed that in standard non-selective plating operations the use the Cu1001 plating solution has not caused any cracks (at least that have been observed). As such, it is believed that the use of a contact mask plating process (e.g. CC mask plating process influences deposit properties in some unexpected way. It is also believed that this result holds true for proximity mask plating operations and even adhered mask plating operations where a relatively trapped plating volume would exist (e.g. when the anode is pressed against or located in proximity to the masking material).

[0110] Without being limited to a particular theory of operation, it is believed that the cracking phenomena related to deposit properties such as the brittleness of the deposits that result from the Cu 1001 plating bath. Alternatively, the cracking phenomena may be related deposit properties, such as brittleness and stress, that make deposits very sensitive to their environment. The cracking of the deposits in the SEM chamber, may result from a vacuum interaction or from a temperature change which results from the from evacuation process. These cracks may be a problem in themselves or they may cause a problem by leaving small gaps that a second material can deposit into.

[0111] Applicant's experiments have shown that with increased concentration of ammonium ions, deposition uniformity improves and shorting reduction occurs, however this comes at the apparent cost of increased brittleness of deposited materials. It has been found that increased concentration of citrate ions seems to lower the brittleness and an increased concentration of selenium oxide (with or without the citrate) even lowers it further. Increased concentrations of both of these materials seems to reduce or eliminate the presence of cracking.

[0112] Experiments were performed using a plating current density of 10 mA/cm2 and pyrophosphate plating baths of having a pH of about 8.3.

[0113] To determine the influence that the different parts of ammonium hydrogen citrate have on plating, different groups of experiments were performed. In a first group of experiments, two chemicals, ammonium hydrogen phosphate, (NH4)2HPO4, and ammonium sulfate, (NH4)2SO4 were substituted into CU 1000 in place of the 15 g/l ammonium hydrogen citrate (which formed CU 1001) to test the influence that the presence of ammonium has on deposit properties. These two materials were chosen as it was believed that the presence of HPO42− or SO42− would not have a influence copper plating. In separate baths, 8.8 g/l of ammonium hydrogen phosphate and ammonium sulfate were added to Cu1000. The 8.8 g/l of these chemicals was chosen so that the amount of ammonium ions present would be about the same as were present in the Cu1001 bath. Both baths acted as the CU1001 bath does for copper CC mask plating (i.e. plating voltages were similar to those for Cu 1001; uniform deposits were obtained, and cracks were observed). FIG. 12 provides a plot of plating voltage versus time for each of the Ammonium hydrogen phosphate modified plating bath, the ammonium sulfate modified plating bath, and the ammonium hydrogen citrate modified plating bath. As can be seen from the plot, the voltage characteristics were very similar over the course of a one hour long deposition time.

[0114] In a second group of experiments, two different concentrations of potassium citrate were added to CU1000 (20 g/l and 40 g/l). The first of these concentrations was selected as it results in about the same amount of citrate ions as does 15 g/l of ammonium hydrogen citrate. This particular additive was selected as it is believed that potassium ions (K+) do not have a significant influence on copper plating operations. FIG. 13 provides a plot of plating voltage versus time over the course of a one hour plating operation for plating baths modified according to the two concentrations noted above. As can be seen for both concentrations, the plating voltage started off low but then after some time jumped up to higher potential (went down in the Figure). The plating solution modified with the lower concentration sustained the low voltage for 22 minutes before the voltage jumped to 1.7-1.8 volts, the higher concentration yielded a low voltage for about 45 minutes indicating that increased concentrations of citrate were useful in improving plating properties. Even at this higher concentration a desired one hour plating time was not achievable at low voltage. The deposits showed pitting (which is believed to be a result of the increased voltage) but they did not show any cracking. From these results it may be concluded that the presence of potassium citrate does help improve the plating process although by itself (at up to these concentration levels), it didn't produce completely acceptable results (under the assumption that one hour of plating is required at a current of 10 mA/cm2 to obtain the desired deposition thickness) In some embodiments, where a thinner deposition is acceptable, one or both of these baths may prove acceptable.

[0115] From the above experiments and observations, it can be concluded that ammonium ions are useful additives in obtaining uniform deposits with low shorting failures; however they may also play a significant role in causing deposition to crack (e.g. in an SEM chamber). In summary any chemicals that supply ammonium ions to a copper pyrophosphate bath will aid in achieving more uniform deposits, lower plating voltage, and or reduced risk of shorting as long as the negative ions in those chemicals do not interfere with the copper plating process. Some examples of useable ammonium salts include ammonium citrate, ammonium chloride, ammonium dihydrogen phosphate, and ammonium hydroxide, and the like.

[0116] Unfortunately, in at least in some embodiments, the benefits of more uniform deposition, lower plating voltage, and reduce risk of shorting, may be offset by a tendency for cracking. One way to reduce the tendency to crack is to lower the concentration of ammonium ions but this will tend to reduce the benefits noted above unless some form of compensation is added. This compensation may take the form of boosting citrate ion concentration by use of non-ammonium containing citrate salts (potassium citrate, for example). This compensation approach was tested using two modified plating baths: (1) CU1000+20 g/l potassium citrate+5 g/l ammonium hydrogen citrate and (2) CU1000+20 g/l potassium citrate+2.5 g/l ammonium hydrogen citrate. The second modified plating bath showed a jump in plating voltage after about 35 minutes but the first modified plating bath worked well as its plating voltage was low and constant for a full 60 minutes. The results of two tests with the first modified plating bath and one with the second are shown in the plot of plating voltage versus plating time of FIG. 14. The first modified plating bath resulted in a factor of three reduction in added ammonium concentration (from ammonium hydrogen citrate) when compared to plating bath CU1001. The samples from this first modified bath were put in the SEM chamber and it was found that the numbers of cracks in the deposits were dramatically reduced and the crack found did not link features or extend from feature corners but instead were located rather innocuously located in the bulk deposits between pits or nodules.

[0117] Selenium (IV) oxide (i.e. selenium oxide, SeO2), has been used as a stress reducer or a brightener for various plating baths. Various concentrations of this chemical were tested in a modified CU1000 plating bath (CU1000+20 g/l potassium citrate+5 g/l ammonium hydrogen citrate). The test results indicated that some cracks were observed (in the SEM chamber) when a concentration of 0.02 g/l were used but that no cracks existed when concentrations were 0.1, 0.5, 1, 1.5, 2 and 5 g/l. With increasing concentration of SeO2 in the plating baths, the plating voltages shifted to lower values and became more stable. This is illustrated with the plating voltage plot versus time of FIG. 15. A further test was performed to see how long plating could continue before shorting resulted (when using an initial anode to substrate separation of about 70-75 &mgr;m). Plating was able to continue for 160 minutes and achieved a nominal thickness of 33 &mgr;m. Previous results from CU1001 yielded a time of 120 minutes but after 90 minutes plating voltage jumped and became unstable. A plot of these results are shown in FIG. 16.

[0118] Further tests were performed with Cu1001 using 0.5 and 1 g/l SeO2. In both cases, no cracks were observed in the deposits. Also a test was performed with a plating solution consisting of Cu1000+8.8 g/l ammonium sulfate+1 g/l SeO2 and no cracks were observed in the deposits.

[0119] To speed build operations, it is sometimes desirable to use higher plating current densities (higher than 10 mA/cm2) and as higher current densities are achievable using baths have higher concentrations of ammonium ions, experiments were performed to test the ability to use such baths while still avoiding the occurrence of cracking. In one experiment, a plating bath consisting of Cu1000+50 g/l ammonium hydrogen citrate+1 g/l SeO2 was used while in another experiment a plating bath consisting of Cu 1000+30 g/l ammonium sulfate+1 g/l SeO2 was used. In these experiments, a current of 20 mA/cm2 for 30 minutes was used. In both cases, no shorting resulted and no cracks were observed.

[0120] Tests were also performed to determine whether the inclusion of SeO2 in the baths resulted in more ductile deposits. Disk shaped copper deposits were made onto stainless steel substrates as they tend to resist deposit adhesion. An attempt was made to peel off the deposited disks produced using three different baths: (1) Cu 1001 without SeO2, (2) Cu 1001 with 1 g/l SeO2, and (3) a bath containing 20 g/l potassium citrate+5 g/l ammonium hydrogen citrate+1 g/l SeO2. For the deposits produced from Cu 1001 without SeO2, the deposits were so brittle that it was very difficult to peel them off the stainless steel substrates. Only one of five disks was successfully removed. The bath containing Cu 1001 with 1 g/l SeO2 and the bath containing 20 g/l potassium citrate+5 g/l ammonium hydrogen citrate+1 g/l SeO2 produced deposits that could easily be removed. These deposits could bend several times without any damage. From these observations and experiments, it is concluded that SeO2 can make ductile deposits. Without limiting ourselves to a particular theory of operation, it is believed that this change in deposit property resulted in the elimination of the cracking of the deposits.

[0121] As noted above, recommended plating baths may have concentrations of ammonium ions up to about 1.47 g/l In some embodiments, ammonium ions may be present in plating baths at concentration levels greater than what is normally considered acceptable (i.e. greater than 11 ml/l) with or without the presence of citrate ions. In other embodiments the concentration of ammonium ions may be within a range that is considered acceptable when combined with citrate ions that help improve deposition quality. In still other embodiments, selenium oxide may be added to a plating bath containing ammonium ions, with or without the presence of citrate ions, in a concentration that is effective to reduce the tendency to crack to an acceptable level.

[0122] According to Tables 1 and 2 above, the maximum concentration of ammonium hydroxide in standard copper pyrophosphate plating baths is about 11 ml/l which converts to about 1.47 g/l (0.134 g/ml*11 ml/l) of NH4+ ions. In some embodiments the concentration of ammonium ions from all sources will result in a net concentration below this amount while in other embodiments it will be greater. For example in Cu1001 (where 15 g/l of ammonium hydrogen citrate is used), the total NH4+ is about 3.1 g/l (slightly more than twice the recommended maximum amount), in the bath containing of Cu1000+20 g/l potassium citrate+5 g/l ammonium hydrogen citrate, the total NH4+ is about 1.5 g/l or slightly more than the recommended maximum but bath also contains a larger concentration of citrate that did the Cu1001 bath. For the bath of Cu 1000+8.8 g/l of either ammonium hydrogen phosphate or ammonium sulfate, the total concentration of NH4+ ions is about 3.1 g/l.

[0123] In summary, for effective micro-mask plating baths with NH4+ concentrations greater than the recommend maximum for standard pyrophosphate plating baths and/or the baths have concentrations of citrate ions that when combined with the concentration of ammonium ions result in relatively uniform deposits at low voltages can be obtained. Also if higher current densities are desired (i.e. for faster build up during plating), even higher concentrations of NH4+ ions are preferred bath (e.g. a current of 20 mA/cm2 using a bath containing about 30 g/l ammonium hydrogen citrate or more. Brittleness of deposits or cracking of deposits can be reduced by favoring citrate ions over ammonium ions and/or by inclusion of an effective concentration of selenium oxide in the solution. Applicants' additional observations and experimentation have shown that ammonium ions play a stronger roll in solving these problems than the citrate does but they also indicate that citrate does play a roll as well. Applicants' additional observations and experiments have linked the presence of ammonium ions to micro-cracking and apparent brittleness of deposits and the presence of citrate as a reducer of these effects. Applicants' further observations and experiments have shown that use of selenium oxide (SeO2) without or without the presence of citrate has even a stronger influence on the reduction of microcracking and brittleness.

[0124] It believed that based on the teachings herein those of ordinary skill in the art will be able to perform experiments to test varying concentrations of citrate salts (e.g. ammonium hydrogen citrate, ammonium citrate, sodium citrate, and/or potassium sodium citrate), varying concentrations of ammonium salts, and/or with various concentrations of selenium oxide and with different contact or adhered mask thicknesses, and desired plating depths, desired current densities, and the like. The effectiveness of varying solutions will be readily ascertainable from which benefit to plating operations can be enabled. Such experimental analysis might involve, among other things, (1) the visual examination of deposits, (2) the visual examination of anodes after deposition, (3) the monitoring of plating voltage during plating operations, (4) brittleness testing, (5) stress testing, and/or (6) ductility testing, and the like.

[0125] Various other embodiments of the present invention exist. Some of these embodiments may be based on a combination of the teachings herein with various teachings incorporated herein by reference. Some embodiments may not use any blanket deposition process and/or they may not use a planarization process. Some embodiments may involve the selective deposition of a plurality of different materials on a single layer or on different layers. Some embodiments may use blanket depositions processes that are not electrodeposition processes. Some embodiments may use selective deposition processes on some layers that are not conformable contact masking processes and are not even electrodeposition processes. Some embodiments may use nickel as a structural material while other embodiments may use different materials such as gold, silver, or any other electrodepositable materials that can be separated from the copper and/or some other sacrificial material. Some embodiments may use copper as the structural material with or without a sacrificial material. Some embodiments may remove a sacrificial material while other embodiments may not. In some embodiments the anode may be different from the conformable contact mask support and the support may be a porous structure or other perforated structure. Some embodiments may use multiple conformable contact masks with different patterns so as to deposit different selective patterns of material on different layers and/or on different portions of a single layer. In some embodiments, the depth of deposition will be enhanced by pulling the conformable contact mask away from the substrate as deposition is occurring in a manner that allows the seal between the conformable portion of the CC mask and the substrate to shift from the face of the conformal material to the inside edges of the conformable material.

[0126] In view of the teachings herein, many further embodiments, alternatives in design and uses of the instant invention will be apparent to those of skill in the art. As such, it is not intended that the invention be limited to the particular illustrative embodiments, alternatives, and uses described above but instead that it be solely limited by the claims presented hereafter.

Claims

1. An electrochemical fabrication process for producing a three-dimensional structure from a plurality of adhered layers, the process comprises:

(A) selectively depositing at least a portion of a layer onto a substrate, wherein the substrate may comprise previously deposited material; and

(B) forming a plurality of layers such that each successive layer is formed adjacent to and adhered to a previously deposited layer, wherein said forming comprises repeating operation (A) a plurality of times;

wherein at least a plurality of the selective depositing operations comprise

(1) contacting the substrate and a pattern mask;

(2) in presence of a plating solution, conducting an electric current through the at least one opening in the selected mask between an anode and the substrate such that a selected deposition material is deposited onto the substrate to form at least a portion of a layer; and

(3) removing the mask from the substrate; and

wherein the selected deposition material is copper; and wherein the plating solution comprises pyrophosphate copper in combination with at least one of the following (a) an ammonium ion concentration greater than about 1.5 g/l, (b) ammonium ions and citrate ions, or (c) ammonium ions and tartrate ions.

2. The process of claim 1 additionally comprising supplying a plurality of preformed masks, wherein each mask comprises a patterned dielectric material that includes at least one opening through which deposition can take place during the formation of at least a portion of a layer, and wherein each mask comprises a support structure that supports the patterned dielectric material, and wherein the contacting of the substrate and a patterned mask comprises contacting a selected one of the preformed masks to the substrate.

3. The process of claim 2 wherein the dielectric comprises a conformable material.

4. The process of claim 1 wherein a source of citrate ions comprises a citrate salt.

5. The process of claim 4 wherein the citrate salt comprises at least one of (1) potassium citrate, (2) sodium citrate, (3) potassium sodium citrate, or (4) ammonium citrate, (5) ammonium hydrogen citrate.

6. The process of claim 4 wherein the citrate salt is supplied in a concentration that is effective in holding the plating voltage at a reduced level substantially during the entire duration of a plating operation.

7. The process of claim 4 wherein the citrate salt is supplied in a concentration that is effective in substantially reducing the occurrence of shorting during plating operations that achieve the desired thickness of deposition.

8. The process of claim 7 wherein the citrate salt is supplied in a concentration that is effective in achieving a uniformity of deposit that has a minimum thickness which is no less than 50% of a maximum thickness.

9. The process of claim 1 wherein a source of tartrate ions comprises a tartrate salt.

10. The process of claim 4 wherein the tartrate salt comprises at least one of (1) potassium tartrate, (2) sodium tartrate, (3) potassium sodium tartrate, or (4) ammonium tartrate, or (5) ammonium hydrogen tartrate.

11. The process of claim 1 wherein the selenium oxide is supplied in a concentration that is effective in reducing cracking to an acceptable level.

12. The process of claim 11 wherein the reduction in cracking substantially eliminates cracking that contacts patterned features.

13. The process of claim 11 wherein the reduction in cracking substantially eliminates cracking.

14. The process of claim 1 wherein the concentration of ammonium ions is sufficient to allow plating to occur at a current density of about 20 mA/cm2 or greater in achieving a desired plating thickness.

15. The process of claim 1 wherein the concentration of ammonium ions is sufficient to allow plating to occur at a current density of about 30 mA/cm2 or greater in achieving a desired plating thickness.

16. The process of claim 1 wherein the concentration of ammonium ions, in combination with any citrate or tartrate ions, is sufficient to allow a desired plating thickness to be achieved without the voltage increasing by a factor of two or more prior to a desired plating thickness or plating time to be reached.

17. The process of claim 1 wherein the concentration of ammonium ions, in combination with any citrate or tartrate ions, is sufficient to allow a desired plating thickness or plating time to be reached without the occurrence of shorting on at least an average of 50% of attempts to form depositions of desired thickness during the formation of a given structure.

18. The process of claim 1 wherein the concentration of ammonium ions, in combination with any citrate or tartrate ions, is sufficient to allow a desired plating thickness or plating time to be reached without the occurrence of shorting on at least 80% of attempts to form depositions of desired thickness during the formation of a given structure.

19. The process of claim 1 wherein the formation of each of a number of layers comprise at least one blanket deposition as well as the selective deposition wherein for a given layer the selectively deposited material is different from a material deposited by blanket deposition.

20. The process of claim 1 wherein the plurality of selective depositions comprise the deposition of a plurality of different materials.

21. The process of claim 1 wherein at least a portion of one layer is formed by a non-electroplating deposition process.

22. The process of claim 1 wherein a plurality of depositions occur during the formation of each of a number of layers wherein at least one of the depositions on each of the number of layers deposits copper and at least one of the other depositions on the number of layers deposits nickel.

23. The process of claim 1 wherein the selective depositing for each of a number of layers comprises at least two selective depositions.

24. The process of claim 1 wherein a number of the plurality of layers are each formed by depositing at least one structural material using at least one deposition and by depositing at least one sacrificial material by using at least one other deposition.

25. The process of claim 24 wherein at least a portion of the at least one sacrificial material is removed after formation of a plurality of layers to reveal a three-dimensional structure comprised of at least one structural material.

26. The process of claim 2 wherein, for each of a plurality of masks, the support for the conformable material for a mask comprises the anode involved in the deposition associated with the use of the mask.

27. The process of claim 3 wherein, for each of a plurality of masks, the support for the conformable material of a mask comprises a porous medium which does not act as the anode during the deposition associated with the use of the mask.

28. The process of claim 2 where a thickness of the dielectric material of at least one selected mask is less than 100 &mgr;m and is more preferably less than 50 &mgr;m.

29. The process of claim 1 wherein the formation of at least a plurality of layers additionally comprises removing a portion of the deposited material from the substrate such that a desired surface level is obtained.

30. A conformable contact masking process for producing a structure, wherein the process comprises:

(A) supplying at least one preformed mask that comprises a patterned dielectric material that includes at least one opening through which deposition can take place during the formation of at least a portion of a layer, and wherein the at least one mask comprises a support structure that supports the patterned dielectric material; and

(B) selectively depositing at least a portion of a layer onto a substrate, comprising

i) contacting the substrate and the dielectric material of the preformed mask;

ii) in presence of a plating solution, conducting an electric current through the at least one opening in the selected mask between an anode and the substrate such that the selected deposition material is deposited onto the substrate to form at least a portion of a layer; and

iii) separating the selected preformed mask from the substrate;

wherein the selected deposition material is copper; and wherein the plating solution comprises pyrophosphate copper in combination with at least one of the following (a) an ammonium ion concentration greater than about 1.5 g/l, (b) ammonium ions and citrate ions, or (c) ammonium ions and tartrate ions.