Method for processing a substrate
AssigneeTokyo Electron Limited
InventorsEric Chih-Fang Liu, Sophie Thibaut, Nicholas Joy, Christopher Cole
Life of this patent
Estimating the term…
Abstract
A method for processing a substrate includes forming a first hardmask layer over a target layer and a second hardmask layer over the first hardmask layer, and patterning the second hardmask layer. The method further includes forming a tone inversion mask between portions of the patterned second hardmask layer, removing the patterned second hardmask layer, patterning the first hardmask layer using the tone inversion mask as an etching mask, and etching the target layer using the patterned first hardmask layer as another etching mask.
Claims20
A method for processing a substrate, the method comprising:
- forming a first hardmask layer over a target layer and a second hardmask layer over the first hardmask layer;
- forming a first mandrel and a second mandrel over the second hardmask layer;
- forming a spacer layer over the first mandrel and the second mandrel;
- forming a non-mandrel block between the first mandrel and the second mandrel;
- forming spacers around sidewalls of the first mandrel and the second mandrel from a first portion of the spacer layer;
- forming a bottom spacer from a second portion of the spacer layer under the non-mandrel block;
- removing the first mandrel and the second mandrel;
- patterning the second hardmask layer by transferring a pattern of the spacers and the bottom spacer to the second hardmask layer;
- forming a tone inversion mask between portions of the patterned second hardmask layer;
- removing the patterned second hardmask layer;
- patterning the first hardmask layer using the tone inversion mask as an etching mask;
- and
- etching the target layer using the patterned first hardmask layer as another etching mask.
A method for processing a substrate, the method comprising:
- forming a mandrel over a target layer;
- forming a first cut mandrel and a second cut mandrel by bisecting the mandrel with a mandrel cut process;
- forming a spacer layer over the first cut mandrel and the second cut mandrel, the spacer layer filling a cut between the first cut mandrel and the second cut mandrel;
- etching the spacer layer, wherein remaining portions of the spacer layer form spacers;
- exposing a first hardmask layer by removing the first cut mandrel and the second cut mandrel, the first hardmask layer being over the target layer;
- transferring a first pattern of the spacers to the first hardmask layer;
- forming a tone inversion mask between portions of the patterned first hardmask layer;
- and
- transferring a second pattern of the tone inversion mask to the target layer with a subtractive etching process, wherein a tone of the second pattern is the reverse of a tone of the first pattern.
A method for processing a substrate, the method comprising:
- forming a first hardmask layer over a conductive layer and a second hardmask layer over the first hardmask layer;
- forming a first mandrel and a second mandrel over the second hardmask layer;
- forming a spacer layer over the first mandrel and the second mandrel;
- forming a non-mandrel block between the first mandrel and the second mandrel;
- performing an anisotropic etching technique on the spacer layer, wherein a first remaining portion of the spacer layer forms spacers around sidewalls of the first mandrel and the second mandrel, and wherein a second remaining portion of the spacer layer under the non-mandrel block forms a bottom spacer;
- exposing the second hardmask layer by removing the first mandrel and the second mandrel;
- transferring a pattern of the spacers and the bottom spacer to the second hardmask layer;
- forming a tone inversion mask between portions of the patterned second hardmask layer;
- removing the patterned second hardmask layer;
- patterning the first hardmask layer using the tone inversion mask as an etching mask;
- and
- performing a subtractive etch on the conductive layer using the patterned first hardmask layer as another etching mask.
Dependencies are inferred from the claim text.