Because the conventional ways did not provide efficient solutions for the peeling between the conductive layer and the mask layer of a gate structure, the inventors studied and analyzed the possible reasons for the peeling.
After studying embodiments wherein the conductive layer comprised tungsten silicide and the mask layer comprised silicon nitride, it was found that when the conductive layer and the mask layer were stacked, there were different stresses between the conductive layer and the mask layer, resulting in different degrees of peeling thereby. An actual statistical result showed that the peeling increased if the stress of the junction was 0 million Pascal. As the stress decreased to −100 million Pascal, the peeling decreased as well. If the stress decreased further to −200 million Pascal, there was an obvious decrease in peeling. That is, the peeling increases as the stress between the conductive layer and the mask layer increases. This relationship shows that the observed peeling may result from the stress generated from the material difference between the conductive layer and the mask layer.
Based on the aforementioned observations, the present invention provides a method for modifying the surface of the conductive layer to reduce the stress between the conductive layer and the mask layer, so as to eliminate the peeling between the conductive layer and the mask layer and enhance the yield of the gate manufacturing process.
A method for manufacturing a gate is described below for illustrating the present invention in detail.
Then, in step 107, a conductive layer 207 is formed on the polysilicon layer 205. The conductive layer 207 is preferably a metallic layer. The conductive layer 207 can be, but not limited to, a tungsten metallic layer or a tungsten-containing layer such as a tungsten silicide layer. Taking the tungsten silicide layer as an example, the conductive layer 207 can be deposited from the reaction product of tungsten hexafluoride and silane by LPCVD.
Subsequently, in step 109, the surface of the conductive layer 207 is modified. In an embodiment, the treatment is proceeded by bombarding the surface with a nitrogen-containing plasma gas as shown in
In step 111, a mask layer 209 is deposited. In an embodiment, the mask layer 209 is a dielectric layer such as, but not limited to, a silicon nitride layer. In step 113, the dielectric layer 203, the polysilicon layer 205, the conductive layer 207, and the mask layer are patterned to form a gate stack structure 213. In step 115, an insulation layer is deposited on the gate stack structure 213. Finally, in step 117, the insulation layer is anisotropically etched to form a spacer 215 at a side of the gate stack structure 213. A gate structure as shown in
The following table shows a comparison between the numbers of defects and defect-containing dies of a first wafer that was not treated with the method of the present invention and those of a second wafer that was treated with the method of the present invention. The number of defects indicates the defects observed in the wafer. The number of defect-containing dies indicates the number of minimum operation unit areas with defects. As shown in the table, the first wafer had 270 defects and 170 defect-containing dies, whereas, the second wafer only had 60 defects and 21 defect-containing dies. Thus, the second wafer had fewer defects and defect-containing dies. Therefore, the peeling can be significantly eliminated by using the stress-reducing method of the present invention.
In conclusion, the present invention treats the tungsten silicide layer that is to be in contact with a silicon nitride mask layer. Accordingly, as a result of the reduced stress between the conductive layer and the mask layer, the peeling is prevented. In other embodiments where the conductive layer of the gate is metal, such as a tungsten layer, the aforementioned plasma process can also reduce the stress between the conductive layer and the mask layer. Therefore, with the gate manufacturing method of the present invention, the stress between the conductive layer and the mask layer would be decreased, the peeling between the conductive layer and the mask layer would be avoided, and the production yield of the semiconductor devices would be improved.
The sequence of the aforementioned steps for the manufacture of a gate is simply for illustration. Any gate manufacturing method being conceivable for people skilled in this art can adopt the method of the present invention to reduce the stress between the conductive layer and the mask layer to eliminate the peeling between the two layers.
The above disclosure is related to the detailed technical contents and inventive features thereof. People skilled in this field may proceed with a variety of modifications and replacements based on the disclosures and suggestions of the invention as described without departing from the characteristics thereof. Nevertheless, although such modifications and replacements are not fully disclosed in the above descriptions, they have substantially been covered in the following claims as appended.
| Number | Date | Country | Kind |
|---|---|---|---|
| 095135792 | Sep 2006 | TW | national |