Abstract:
A simple method for calculating the optimum amount of HDP deposited material that needs to be removed during CMP (without introducing dishing) is described. This method derives from our observation of a linear relationship between the amount of material that needs to be removed in order to achieve full planarization and a quantity called “OD for CMP density”. The latter is defined as PA×(100−PS) where PA is the percentage of active area relative to the total wafer area and PS is the percentage of sub-areas relative to the total wafer area. The sub-areas are regions in the dielectric, above the active areas, that are etched out prior to CMP. Thus, once the materials have been characterized, the optimum CMP removal thickness is readily calculated for a wide range of different circuit implementations.
Abstract:
This description relates to a sensing product formed using a substrate with a plurality of epi-layers. At least a first epi-layer has a different composition than the composition of a second epi-layer. The sensing product optionally includes at least one radiation sensing element in the second epi-layer and optionally an interconnect structure over the second epi-layer. The sensing product is formed by removing the substrate and all epi-layers other than the second epi-layer. A light incident surface of the second epi-layer has a total thickness variation of less than about 0.15 μm.
Abstract:
A method of electroplating a metal layer on a semiconductor device includes a sequence of biasing operations that includes a first electroplating step at a first current density followed by a second immersion step at a second current density being less than the first current density, and subsequent electroplating steps of increasing current densities beginning with a third electroplating step having a third current density that is greater than the first current density. The second, low current density immersion step improves the quality of the plating process and produces a plated film that completely fills openings such as vias and trenches and avoids hollow vias and pull-back on the bottom corners of via and trench openings. The low current density second immersion step produces an electrochemical deposition process that provides low contact resistance and therefore reduces device failure.
Abstract:
A new and improved wafer support for supporting wafers in a process chamber such as an edge bead removal (EBR) chamber. The wafer support comprises multiple wafer support units each including a gripper block that engages an edge portion or bevel of the wafer. The gripper block is attached to an engaging and disengaging mechanism for selectively causing engagement of the gripper blocks with the wafer to support the wafer and disengagement of the gripper blocks from the wafer to release the wafer for removal of the wafer from the chamber. The gripper blocks contact little or none of the surface area on the patterned surface of the wafer to prevent or substantially reduce the formation of contact-induced defects on the wafer.
Abstract:
A simple method for calculating the optimum amount of HDP deposited material that needs to be removed during CMP (without introducing dishing) is described. This method derives from our observation of a linear relationship between the amount of material that needs to be removed in order to achieve full planarization and a quantity called “OD for CMP density”. The latter is defined as PA×(100−PS) where PA is the percentage of active area relative to the total wafer area and PS is the percentage of sub-areas relative to the total wafer area. The sub-areas are regions in the dielectric, above the active areas, that are etched out prior to CMP. Thus, once the materials have been characterized, the optimum CMP removal thickness is readily calculated for a wide range of different circuit implementations.
Abstract:
A new method is provided to replace tungsten plugs for wafers that trigger the WCVD backside alarm. In this new rework process, the original TiN glue layer is sputter etched back and a new (“fresh”) 100-Angstrom thick layer of TiN is deposited. The new tungsten plug is created over the top surface of the refreshed glue layer.
Abstract:
An improved method for removing residual slurry particles and metallic residues from the surface of a semiconductor substrate after chemical-mechanical polishing has been developed. The cleaning method involves sequential spray cleaning solutions of NH.sub.4 OH and H.sub.2 O, NH.sub.4 OH, H.sub.2 O.sub.2 and H.sub.2 O, HF and H.sub.2 O, and HCl, H.sub.2 O.sub.2 and H.sub.2 O. The cleaning sequence is: 1. A pre-soak in a spray solution of NH.sub.4 OH and H.sub.2 O; 2. Spray cleaning in a solution of NH.sub.4 OH, H.sub.2 O.sub.2 and H.sub.2 O; 3. Spray cleaning in a dilute solution of HF and H.sub.2 O; 4. Spray rinsing in DI-water. It is important that slurry particulates first be removed by NH.sub.4 OH, H.sub.2 O.sub.2 and H.sub.2 O, followed by spray cleaning in a dilute solution of HF and H.sub.2 O to remove metallic residues. The spray cleaning method is superior to brush cleaning methods for both oxide-CMP and tungsten-CMP and results in superior removal of slurry particles and metallic residues introduced by the CMP processes. An optional spray cleaning step using a solution of HCl, H.sub.2 O.sub.2 and H.sub.2 O results in further reduction of metallic residue contamination following oxide-CMP. Compared to traditional brush cleaning the new spray cleaning process has a 2.times. improvement in throughput, less consumption of DI water, and low risk of cross-contamination between sequentially cleaned substrates.
Abstract:
Low-k dielectric layer, semiconductor device, and method for fabricating the same. The low-k dielectric layer comprises a hardened sub-layer sandwiched by two low-k dielectric sub-layers. The hardened sub-layer is formed by a method comprising bombarding the underlying low-k dielectric sub-layer utilizing hydrogen plasma or inert gas plasma. The semiconductor device comprises the low-k dielectric layer overlying an etch stop layer overlying a substrate, and a conductive material embedded in the dielectric layer and the etch stop layer, electrically connecting to the substrate.
Abstract:
A method and system is disclosed for reducing slurry usage in a chemical mechanical polishing operation utilizing at least one polishing pad thereof. Slurry can be intermittently supplied to a chemical mechanical polishing device. The slurry is generally flushed so that a portion of said slurry is trapped in a plurality of pores of at least one polishing pad associated with said chemical mechanical polishing device, wherein only a minimum amount of said slurry necessary is utilized to perform said chemical mechanical polishing operation, thereby reducing slurry usage and maintaining a consistent level of slurry removal rate performance and a decrease in particle defects thereof. The present invention thus discloses a method and system for intermittently delivering slurry to a chemical mechanical polishing device in a manner that significantly conserves slurry usage.
Abstract:
A cobalt silicide process having a titanium-rich/titanium nitride capping layer to improve junction leakage is described. Semiconductor device structures to be silicided are formed in and on a semiconductor substrate. A cobalt layer is deposited overlying the semiconductor device structures. A titanium-rich/titanium nitride capping layer is deposited overlying the cobalt layer. Thereafter, a cobalt silicide layer is formed on the semiconductor device structures. The titanium-rich/titanium nitride capping layer and an unreacted portion of the cobalt layer are removed to complete fabrication of the integrated circuit device.