Abstract: A method for fabricating semiconductors is provided in which a conformal layer is formed superjacent at least two conductive layers. The conformal layer has a thickness of at least 50 .ANG.. A barrier layer is then formed superjacent the conformal layer to prevent subsequent layers from diffusing into active regions. The barrier layer is preferably Si.sub.3 N.sub.4. A glass layer is then formed superjacent the barrier layer. The glass layer has a thickness of at least 1 k.ANG.. The glass layer is heated to a temperature of at least 800.degree. C. for at least 15 minutes while introducing H.sub.2 and O.sub.2 at a high temperature to cause vaporization, thereby causing the glass layer to reflow. Next, the glass layer is exposed to a gas and radiant energy for 5 to 60 seconds, thereby making said glass layer planar. The radiant energy generates a temperature within the range of 700.degree. C. to 1250.degree. C. Further, the gas is at least one of N.sub.2, NH.sub.3, O.sub.2, N.sub.2 O, Ar, Ar-H.sub.2, H.sub.

Abstract: The present invention develops a process for forming hemi-spherical grained silicon storage capacitor plates by the steps of: forming a silicon layer over a pair of neighboring parallel conductive lines, the silicon layer making contact to an underlying conductive region; patterning the silicon layer to form individual silicon capacitor plates; exposing the silicon capacitor plates to a fluorine based gas mixture during an high vacuum annealing period, thereby transforming the silicon capacitor plates into the semi-spherical grained silicon capacitor plates; conductively doping the hemispherical grained silicon capacitor plates; forming a capacitor dielectric layer adjacent and coextensive the semispherical grained silicon capacitor plates; and forming a second conductive silicon layer superjacent and coextensive the capacitor dielectric layer.

Abstract: This invention is a process for fabricating ultra-large-scale integration CMOS circuits using a single polysilicon gate layer for both N-channel and P-channel devices, a single mask step for defining the gates of both N-channel and P-channel devices, the fabrication of one set of disposable spacers for N-channel implants, and the fabrication of another set of disposable spacers for P-channel source/drain implants. The set of spacers used for P-channel implants also comprises material deposited to fabricate the spacers for the N-channel implants. The process is adaptable to LDD structures for both N-channel and P-channel devices or for only N-channel devices. The process is also compatible with anti-punchthrough implants for both types of devices.

Abstract: A low-pressure chemical vapor deposition process is disclosed for creating high-density, highly-conformal titanium nitride films which have very low bulk resistivity, and which provide excellent step coverage. The process utilizes a metal-organic compound, tetrakis-dialkylamido-titanium Ti(NR.sub.2).sub.4, as the primary precursor, in combination with a halogen gas selected from the group consisting of fluorine, chlorine and bromine. The wafer is heated to a temperature within a range of 200.degree.-600.degree. C. The halogen gas is admitted to the depositiion chamber before the introduction of the primary precursor compound or simultaneously with the primary precursor compound.

Abstract: There is a precharge circuitry that uses little real estate and can be deactivated once a word line driver is activated. Specifically, a high signal created by the selected driver is fed back to the precharge circuit to deactivate it when activating a chosen word line. Thus, alleviating the resulting effect between the low signal to activate the selected driver and the precharge high voltage current both using the same node coupled to the word line drivers.

Abstract: A semiconductor processing method of providing and planarizing an insulating layer on a semiconductor wafer includes the following sequential steps: a) providing a conformal layer of insulating material to a first thickness over a semiconductor wafer having non-planar topography; b) providing a CMP polishing protective layer over the conformal layer to a second thickness, the protective layer being of different composition than the conformal layer; and c) chemical-mechanical polishing the protective layer and conformal layer in a single CMP step using a single CMP slurry and under conditions which in combination with the slurry remove the conformal layer material at a faster rate than the protective layer material, the protective layer upon outward exposure of conformal layer material in high topographical areas restricting material removal from low topographical areas during such chemical-mechanical polishing.

Abstract: The invention is a method directed to the use of a nonvolatile precursor, either a solid precursor or a liquid precursor, suitable for chemical vapor deposition (CVD), including liquid source CVD (LSCVD), of a semiconductor film. Using the method of the invention the nonvolatile precursor is dissolved in a solvent. The choice of solvent is typically an inorganic compound that has a moderate to high vapor pressure at room temperature and that can be liquified by a combination of pressure and cooling. The solution thus formed is then transported at an elevated pressure and/or a reduced temperature to the CVD chamber. In CVD the solution evaporates at a higher temperature and a lower pressure upon entry to the CVD chamber, and the nonvolatile precursor, in its gaseous state, along with a gas reactant, produces a product which is deposited as a thin film on a semiconductor wafer.

Abstract: A process useful for isolating active areas of semiconductor devices, comprising the steps of: forming pad oxide, polysilicon, and nitride layers superjacent a substrate; patterning and etching the layers and the substrate, thereby forming a recess in the substrate; forming a nitride spacer within the recess, then growing a field oxide region therein; removing the layers thereby causing an indentation to form within the periphery of the field oxide region; disposing polysilicon within the indentation; etching the polysilicon to a level even with the field oxide region; and oxidizing the polysilicon.

Abstract: A video RAM having isolated array sections for providing write function that will not affect other array sections. The whole VRAM memory array does not have to be completely read before writing new pixel information to particular array section. At least two separate VRAM activities can be performed simultaneously to different parts of the array. Specifically, to write to one particular section of an array and to and for refreshing other parts of the VRAM. The overall read and write sequences can be shorter. When a particular pixel or memory cell has to be modified or update, only an associated SAM to the particular cell will be activated. This SAM will now only affect the column lines associated with that section of the array containing the activated SAM.

Abstract: A charge pump circuit and method for increasing a value of a supply potential. The charge pump circuit features a first stage circuit for generating an intermediate pumped potential greater than an input supply potential. The intermediate pumped potential becomes a supply potential for a portion of a second stage circuit. The second stage circuit generates a pumped output potential greater than the intermediate pumped potential. Both the first and second stage circuits have at least two capacitors, a small pump capacitor and a large pump capacitor. The first stage circuit of the invention supplies the increased intermediate pumped potential to those nodes which are used to charge the small pump capacitor of the second stage circuit. The input supply potential supplies the potential to those nodes which are used to charge the large pump capacitors of both stages and the small pump capacitor of the first stage circuit.

Abstract: Self refresh timing in a low-power dynamic memory system is governed by an oscillator having a voltage dependent resistor and a process dependent capacitor. The resistor increases in resistance to compensate for increased applied power supply voltage and variation in substrate bias voltage. The total capacitance needed for a given oscillator center frequency is made up of a plurality of capacitors having the same physical characteristics as the capacitor used in the dynamic memory cell. Power supply voltage, substrate bias voltage, and the physical characteristics of the cell capacitor affect the cell's data retention time. By compensating the oscillator for these effects, refreshing is optimally accomplished within the data retention time. A system having compensated refresh timing according to the invention is more appropriate for low power applications due to the resulting decreased power consumption.

Abstract: The invention is a storage cell capacitor and a method for forming the storage cell capacitor having a storage node electrode comprising a barrier layer interposed between a conductive plug and an oxidation resistant layer. A thick insulative layer protects the sidewalls of the barrier layer during the deposition and anneal of a dielectric layer having a high dielectric constant. The method comprises forming the conductive plug in a thick layer of insulative material such as oxide or oxide/nitride. The conductive plug is recessed from a planarized top surface of the thick insulative layer. The barrier layer is then formed in the recess. The process is continued with a formation of a second insulative layer, a potion of which is removed to form an opening exposing a portion of the barrier layer. An oxidation resistant conductive layer is deposited in the recess and forms at least a portion the storage node electrode of the capacitor.

Abstract: A non-volatile static read/write memory is formed by a bistable memory cell which is programmable to operate statically in one of two alternative output states, corresponding to binary output voltages. The static memory cell is formed by a pair of MOSFET inverters having cross-coupled inputs and outputs. A ferroelectric storage element fabricated from a material such as lead zirconate titanate or barium strontium titanate is positioned to be electrically polarized in one of two alternative orientations determined by the memory cell's output voltage. The ferroelectric storage element maintains its electric polarization upon power-down of the memory cell. Upon subsequent power-up of the memory cell, the ferroelectric storage element biases the memory cell toward one of the memory cell's two output states in accordance with the electric polarization of the ferroelectric storage element, and in accordance with the memory cell's output state at power-down.

Abstract: An array of SONOS memory cells includes: a) a pair of spaced, adjacent SONOS gates atop a silicon substrate within an array area; b) a trench between the gates, the trench having opposing downwardly elongated sidewalls and a base, the sidewalls being doped with a conductivity enhancing impurity of a first conductivity type to define separated source/drain diffusion regions in between and adjacent the respective gates of the pair, the trench being filled with an effectively electrically insulating material; c) a word line commonly interconnecting the adjacent SONOS gates of the pair; and d) separate bit lines separately electrically engaging the separated diffusion regions of the pair. LDD regions are also included. A method of producing such a construction is disclosed.

Abstract: A cleaner, selective etch process wherein NF.sub.3 ions and nitrogen ions are employed to bombard a patterned nitride layer disposed superjacent an oxide layer, thereby creating substantially vertical sidewalls, especially useful when etching submicron features.

Abstract: There is a post etching test apparatus and method to be able to only test just a few die on the wafer. Uniquely, the remainder of the die on the wafer can be salvaged, if the test identifies proper tolerances for the etching process over the entire wafer surface. If the tests show negative, the etch process can be re calibrated and the wafer can be reprocessed and tested again. Salvage of the majority of the die on the wafer under test is possible by using a fine point resist removal plate. Specifically, oxygen is forced over certain die on the wafer to remove the resist mask by using a plate barrier with only a few holes in it. The holes are located a key positions around the wafer, and restrain the oxygen laminar flow to effect only the wafers directly below these holes.

Abstract: A process for forming a thin film transistor having a lightly doped drain which is self-aligned with the transistor channel. A transistor gate is formed over a first dielectric layer, and a second dielectric layer is formed over the transistor gate. A layer of polycrystalline silicon (poly) is formed over said second dielectric layer, and the poly layer can be optionally doped with a P-type or N-type dopant to adjust the threshold voltage of the transistor. Next, an implant masking layer is formed over the gate, and has an etch mask thereupon. The exposed implant masking layer is removed, and in one embodiment the etch mask is undercut during the same etch to remove portions of the implant masking layer from under the etch mask. The exposed poly is doped with a P-type dopant. The etch mask is removed and the exposed poly is again doped with a P-type dopant to form the lightly doped drain using the implant mask to self-align the lightly doped drain with the channel region.

Abstract: A programmable logic device (PLD) with an output macrocell circuit is disclosed. Specifically, there is a macrocell having an exclusive logic signal feedback line and an exclusive external input signal line both feeding into the input of the PLD. Exactly, this PLD can disable the I/O pad and still have an internal feedback to its logic circuitry.

Abstract: A band gap voltage reference circuit operates between a positive supply voltage and ground. The inputs to a difference amplifier of the band gap reference circuit are biased above the voltage drop of the base-emitter junctions of the band gap reference. The bias voltage is then subtracted from the difference amplifier output by a second difference amplifier. In addition, a bootstrap circuit assures a nonzero output from the first difference amplifier. Other embodiments wherein the band gap reference circuit is more generally a summing circuit are disclosed.

Abstract: The present invention develops a bond pad interconnect in an integrated circuit device, by forming an aluminum pad; bonding a metal layer (such as copper (Cu), nickel (Ni), tungsten (W), gold (Au), silver (Ag) or platinum (Pt)) or a metal alloy (such as titanium nitride) to the aluminum bond pad by chemical vapor deposition or by electroless deposition; and adhering a conductive epoxy film to the metal layer, thereby forming a low resistive bond pad interconnect.