Abstract: Techniques are provided for optimizing a program operation in a memory device to compensate for program speed variations due to block oxide thinning. In one approach, during a program operation, a program voltage which indicates program speed is acquired from sub-blocks with the highest and lowest program speeds. An initial program voltage for intermediate sub-blocks can be determined based on the acquired program voltages and the positions of the intermediate sub-blocks. The technique can accommodate a loss of one or both acquired program voltages if the programming is interrupted. In another approach, a program voltage which indicates program speed is acquired from one sub-block, and for a later-programmed sub-block, an appropriate offset is located from a table and summed with the acquired program voltage to determine an optimum initial program voltage.

Abstract: Method for performing an erase program operation. Various methods include: erasing a block of cells by: applying a program pulse to a block of memory elements in the three-dimensional memory that programs the block of memory elements to a level below an erase verify level, where the three-dimensional memory comprises memory elements stacked vertically; performing a verify step to verify voltage levels of a group of memory elements; determining that a memory element of the group is outside of a threshold window defined between the erase verify level and a compact erase threshold amount; and applying a second program pulse to the memory element. Where erasing the block of memory elements creates an erased block, where a width of the voltage distribution of the erased memory elements in the erased block is the same as or below a width of a voltage distribution associated with programmed memory elements.

Abstract: Techniques are provided for optimizing a program operation in a memory device to compensate for program speed variations due to block oxide thinning. In one approach, during a program operation, a program voltage which indicates program speed is acquired from sub-blocks with the highest and lowest program speeds. An initial program voltage for intermediate sub-blocks can be determined based on the acquired program voltages and the positions of the intermediate sub-blocks. The technique can accommodate a loss of one or both acquired program voltages if the programming is interrupted. In another approach, a program voltage which indicates program speed is acquired from one sub-block, and for a later-programmed sub-block, an appropriate offset is located from a table and summed with the acquired program voltage to determine an optimum initial program voltage.

Abstract: A three-dimensional memory device includes an alternating stack of insulating layers and electrically conductive layers located over a substrate, and memory pillar structures extending through the alternating stack. Each of the memory pillar structures includes a respective memory film and a respective vertical semiconductor channel. Dielectric cores contact an inner sidewall of a respective one of the vertical semiconductor channels. A drain-select-level isolation structure laterally extends along a first horizontal direction and contacts straight sidewalls of the dielectric cores at a respective two-dimensional flat interface. The memory pillar structures may be formed on-pitch as a two-dimensional periodic array, and themay drain-select-level isolation structure may cut through upper portions of the memory pillar structures to minimize areas occupied by the drain-select-level isolation structure.

Abstract: Techniques are provided for optimizing an erase operation in a memory device to compensate for erase speed variations due to blocking oxide thinning In an erase operation for a block, the channels of NAND strings in different sub-blocks can be charged up by different amounts. One approach adjusts the control gate voltage of a first select gate transistor in a NAND string. This adjusts the amount of holes generated in the channel due to gate-induced drain leakage. Another approach adjusts the control gate voltage of additional select gate transistors in the NAND string to adjust the conductivity of the adjacent channel regions. Another approach applies different bit line voltages to different rows of NAND strings in each sub-block.

Abstract: An approach to use silicided bit line contacts that do not short to the underlying substrate in memory devices. The approach provides for silicide formation in the bit line contact area, using a process that benefits from being self-aligned to the oxide-nitride-oxide (ONO) nitride edges. A further benefit of the approach is that the bit line contact implant and rapid temperature anneal process can be eliminated. This approach is applicable to embedded flash, integrating high density devices and advanced logic processes.

Abstract: Techniques are provided for optimizing a program operation in a memory device to compensate for program speed variations due to block oxide thinning. In one approach, during a program operation, a program voltage which indicates program speed is acquired from sub-blocks with the highest and lowest program speeds. An initial program voltage for intermediate sub-blocks can be determined based on the acquired program voltages and the positions of the intermediate sub-blocks. The technique can accommodate a loss of one or both acquired program voltages if the programming is interrupted. In another approach, a program voltage which indicates program speed is acquired from one sub-block, and for a later-programmed sub-block, an appropriate offset is located from a table and summed with the acquired program voltage to determine an optimum initial program voltage.

Abstract: Techniques are described for programming memory cells with reduced widening of the threshold voltage distributions. Bit line voltages are adjusted during verify tests for memory cells assigned to the upper data state in a pair of adjacent data states which are concurrently verified. An elevated bit line voltage is applied and then stepped up in successive program loops. A lower, fixed bit line voltage is used for verifying the lower data state in the pair of adjacent data states. In one option, the step size increases progressively over the program loops. In another option, the minimum level of the elevated bit line voltage is lower for higher data states. In another option, the minimum level of the elevated bit line voltage is set as a function of data states, program-erase cycles and/or temperature.

Abstract: Techniques are provided for optimizing an erase operation in a memory device to compensate for erase speed variations due to blocking oxide thinning In an erase operation for a block, the channels of NAND strings in different sub-blocks can be charged up by different amounts. One approach adjusts the control gate voltage of a first select gate transistor in a NAND string. This adjusts the amount of holes generated in the channel due to gate-induced drain leakage. Another approach adjusts the control gate voltage of additional select gate transistors in the NAND string to adjust the conductivity of the adjacent channel regions. Another approach applies different bit line voltages to different rows of NAND strings in each sub-block.

Abstract: Techniques are provided to reduce neighbor word line interference and charge loss in a multi-pass program operation. In one implementation, the first pass of a multi-pass program operation uses one or more program pulses without performing associated verify tests. The memory cells may be programmed to different intermediate threshold voltage (Vth) distributions in the first program pass. Different bit line voltages can be used to obtain the different intermediate Vth distributions when the single program pulse is applied. In other cases, multiple program pulses are applied without performing verify tests. The intermediate Vth distributions can be provided for the memory cells assigned to the higher data states but not the lower data states, or for memory cells assigned to both the higher and lower data states.

Abstract: An alternating stack of insulating layers and sacrificial material layers is formed over a substrate, and memory stack structures are formed through the alternating stack. A backside trench is formed through the alternating stack, and backside recesses are formed by removing the sacrificial material layers. An undoped aluminum oxide backside blocking dielectric layer is formed in the backside recesses and on sidewalls the backside trench. A portion of the undoped aluminum oxide backside blocking dielectric layer located at an upper end of the backside trench is converted into a carbon-doped aluminum oxide layer. An electrically conductive material is deposited in the backside recesses and at peripheral regions of the backside trench. The electrically conductive material at the peripheral regions of the backside trench is removed by an etch process, with the carbon-doped aluminum oxide layer providing etch resistivity during the etch process.

Abstract: Techniques are described for programming memory cells with reduced widening of the threshold voltage distributions. Bit line voltages are adjusted during verify tests for memory cells assigned to the upper data state in a pair of adjacent data states which are concurrently verified. An elevated bit line voltage is applied and then stepped up in successive program loops. A lower, fixed bit line voltage is used for verifying the lower data state in the pair of adjacent data states. In one option, the step size increases progressively over the program loops. In another option, the minimum level of the elevated bit line voltage is lower for higher data states. In another option, the minimum level of the elevated bit line voltage is set as a function of data states, program-erase cycles and/or temperature.

Abstract: Techniques are provided to reduce neighbor word line interference and charge loss in a multi-pass program operation. In one implementation, the first pass of a multi-pass program operation uses one or more program pulses without performing associated verify tests. The memory cells may be programmed to different intermediate threshold voltage (Vth) distributions in the first program pass. Different bit line voltages can be used to obtain the different intermediate Vth distributions when the single program pulse is applied. In other cases, multiple program pulses are applied without performing verify tests. The intermediate Vth distributions can be provided for the memory cells assigned to the higher data states but not the lower data states, or for memory cells assigned to both the higher and lower data states.

Abstract: Program disturb is suppressed during a program recovery phase of a program operation in a memory device. The duration of the recovery phase can be increased when the risk of program disturb is greater due to factors such as temperature, the position of the selected word line, the number of program-erase cycles and the program pulse magnitude. In other approaches, the risk of program disturb is reduced by providing an early ramp down of the voltages of the drain-side word line relative to a ramp down of the voltages of the source-side word lines, providing an early ramp down of the bit line voltage of the inhibited NAND strings relative to the ramp down of the select gate voltage or setting a lower recovery voltage for the source-side word lines relative to the recovery voltage of the drain-side word lines.

Abstract: Techniques are provided for optimizing an erase operation in a memory device to compensate for erase speed variations due to blocking oxide thinning In an erase operation for a block, the channels of NAND strings in different sub-blocks can be charged up by different amounts. One approach adjusts the control gate voltage of a first select gate transistor in a NAND string. This adjusts the amount of holes generated in the channel due to gate-induced drain leakage. Another approach adjusts the control gate voltage of additional select gate transistors in the NAND string to adjust the conductivity of the adjacent channel regions. Another approach applies different bit line voltages to different rows of NAND strings in each sub-block.

Abstract: A three-dimensional memory device may include an alternating stack of insulating layers and spacer material layers formed over a carrier substrate. The spacer material layers are formed as, or are subsequently replaced with, electrically conductive layers. Memory stack structures are formed through the alternating stack. Each memory stack structure includes a respective vertical semiconductor channel and a respective memory film. Drain regions and bit lines can be formed over the memory stack structures to provide a memory die. The memory die can be bonded to a logic die containing peripheral circuitry for supporting operations of memory cells within the memory die. A distal end of each of the vertical semiconductor channels is physically exposed by removing the carrier substrate. A source layer is formed directly on the distal end each of the vertical semiconductor channels. A bonding pad can be formed on the source layer.

Abstract: An alternating stack of insulating layers and sacrificial material layers is formed over a substrate, and memory stack structures are formed through the alternating stack. A backside trench is formed through the alternating stack, and backside recesses are formed by removing the sacrificial material layers. An undoped aluminum oxide backside blocking dielectric layer is formed in the backside recesses and on sidewalls the backside trench. A portion of the undoped aluminum oxide backside blocking dielectric layer located at an upper end of the backside trench is converted into a carbon-doped aluminum oxide layer. An electrically conductive material is deposited in the backside recesses and at peripheral regions of the backside trench. The electrically conductive material at the peripheral regions of the backside trench is removed by an etch process, with the carbon-doped aluminum oxide layer providing etch resistivity during the etch process.

Abstract: Techniques are provided for optimizing an erase operation in a memory device to compensate for erase speed variations due to blocking oxide thinning In an erase operation for a block, the channels of NAND strings in different sub-blocks can be charged up by different amounts. One approach adjusts the control gate voltage of a first select gate transistor in a NAND string. This adjusts the amount of holes generated in the channel due to gate-induced drain leakage. Another approach adjusts the control gate voltage of additional select gate transistors in the NAND string to adjust the conductivity of the adjacent channel regions. Another approach applies different bit line voltages to different rows of NAND strings in each sub-block.

Abstract: A three-dimensional memory device includes an alternating stack of insulating layers and electrically conductive layers located over a substrate, and memory pillar structures extending through the alternating stack. Each of the memory pillar structures includes a respective memory film and a respective vertical semiconductor channel Dielectric cores contact an inner sidewall of a respective one of the vertical semiconductor channels. A drain-select-level isolation structure laterally extends along a first horizontal direction and contacts straight sidewalls of the dielectric cores at a respective two-dimensional flat interface. The memory pillar structures may be formed on-pitch as a two-dimensional periodic array, and themay drain-select-level isolation structure may cut through upper portions of the memory pillar structures to minimize areas occupied by the drain-select-level isolation structure.

Abstract: Techniques are provided for optimizing a program operation in a memory device to compensate for program speed variations due to block oxide thinning. In one approach, during a program operation, a program voltage which indicates program speed is acquired from sub-blocks with the highest and lowest program speeds. An initial program voltage for intermediate sub-blocks can be determined based on the acquired program voltages and the positions of the intermediate sub-blocks. The technique can accommodate a loss of one or both acquired program voltages if the programming is interrupted. In another approach, a program voltage which indicates program speed is acquired from one sub-block, and for a later-programmed sub-block, an appropriate offset is located from a table and summed with the acquired program voltage to determine an optimum initial program voltage.