Method of making pillar-type structure on semiconductor substrate

The present invention relates to a method in making a pillar-type structure on a semiconductor substrate, and more particularly to a method in making pillar-type capacitor node on a silicon substrate.

In U.S. Pat. No. 5,482,885 Water Lur et al. had provided a method to form a capacitor in a DRAM cell by depositing a conductive polysilicon electrode layer on the substrate. Oxide lines are then formed on the polysilicon layer. Using the oxide lines as a mask to etch the polysilicon layer, pillar-type capacitor node is formed in the plolysilicon electrode layer.

The present application however provides a different way to form a pillar-type capacitor node on the silicon substrate.

It is therefore an object of the present invention to provide a method in making a pillar-type structure on semiconductor substrate, by depositing sequentially a conductive polysilicon electrode layer, a nitride layer and a silicon layer on the substrate, and then forming required oxide pillars in the silicon layer to act as a mask for etching the conductive polysilicon electrode layer.

FIG. 1 shows that a conductive polysilicon electrode layer 3 , a nitride layer 4 and a first silicon layer 5 are formed on the substrate 1 .

FIG. 2 shows that the first silicon layer 5 is etched being divided into a plurality of regions to be spaced apart by a distance of 5 X, and each region has a width of 3 X.

FIG. 3 shows that a doped SiO 2 layer 6 of thickness X is deposited.

FIG. 4 shows that etchingback process is conducted to form spacers 6 abutting the first silicon layer 5 .

FIG. 5 shows that the remained first silicon layer 5 is removed.

FIG. 6 shows that a second silicon layer 7 of thickness X is deposited.

FIG. 7 shows that etchingback process is conducted to form spacers 7 abutting the spacers 6 .

FIG. 8 shows that a doped SiO 2 layer 8 is deposited.

FIG. 9 shows that etchingback process is conducted on the SiO 2 layer 8 to form spacer 9 and stud 10 .

FIG. 10 shows than the spacer 7 is removed so that the distance between the remained oxide pillars is X.

FIG. 11 shows that an etching process is conducted on the Si 3 N 4 layer 4 and the polysilicon layer 3 below the spacer 7 .

FIG. 12 shows that the remained spacer 6 , spacer 9 , stud 10 and the Si 3 N 4 layer 4 are removed.

Referring to FIG. 1, a semiconductor silicon substrate 1 is provided first to form a diffusion region 11 (e.g. a source region) therein, then a CVD process is conducted to form a dielectric layer 2 on the semiconductor silicon substrate 1 , followed by a photolithography process and an anisotropic plasma etching process to remove part of the dielectric layer 2 , so as to form a recess to expose the diffusion region 11 of the silicon substrate 1 .

Next, a CVD process is conducted again to deposit a polysilicon layer 3 of thickness 3000˜8000 Å to cover the dielectric layer 2 and fill up the recess, then a CVD process is conducted for form a Si 3 N 4 layer 4 of thickness 300˜800 Å on the polysilicon layer 3 .

A first silicon layer 5 of thickness 2500˜5000 Å is then deposited on the Si 3 N 4 layer 4 . The first silicon layer 5 can be a doped polysilicon layer, an undoped polysilicon layer or an amorphous silicon layer, but preferably using doped polysilicon.

Referring to FIG. 2, which just shows the Si 3 N 4 layer 4 and the first silicon layer 5 to simplify the drawing, but the silicon substrate 1 , the dielectric layer 2 and the polysilicon layer 3 are still exist below the Si 3 N 4 layer 4 . As shown in FIG. 2, a photolithography process and an etching process are employed to etch the first silicon layer 5 to expose the Si 3 N 4 layer 4 intermittently such that the first silicon layer 5 is divided into a plurality of regions to be spaced apart by a distance of 5 X, and each region has a width of 3 X (X for example is between 100˜500 Å).

Referring to FIG. 3, a doped SiO 2 layer 6 of thickness X (100˜500 Å) is deposited as shown, and then etched back by fluorine plasma to form spacers 6 abutting the first silicon layer 5 , as shown in FIG. 4 .

Referring to FIG. 5, the remained first silicon layer 5 is removed by isotropic etching in SF 6 gas plasma, so that the distance between the spacers 6 is 3 X.

Referring to FIG. 6, a second silicon layer 7 of thickness X (100˜500 Å) is deposited and then etched back to form spacers 7 abutting the spacers 6 , as shown in FIG. 7 . The distance between the spacers 7 is also X. The second silicon layer 5 can be a doped polysilicon layer, an undoped polysilicon layer or an amorphous silicon layer, but preferably using doped polysilicon.

Referring to FIG. 8, a doped SiO 2 layer 8 is deposited to cover everything above the Si 3 N 4 layer 4 and fill up the spaces between spacer 7 .

Referring to FIG. 9, by etchingback the doped SiO 2 layer 8 to form spacers 9 of thickness X (100˜500 Å) abutting the the spacers 7 and form studs 10 between the the spacers 7 , as shown in the figure.

Referring to FIG. 10, the spacers 7 are removed by isotropic etching in SF 6 gas plasma so that the distance between the remained oxide pillars is X (100˜500 Å).

Referring to FIG. 11, anisotropic etching the Si 3 N 4 layer 4 and the polysilicon layer 3 below the spacers 7 to a depth without exposing the oxide layer 2 on said semiconductor silicon substrate 1 .

Referring to FIG. 12, removing the spacers 6 , the spacers 9 , the studs 10 and the Si 3 N 4 layer 4 that remained to expose the polysilicon layer 3 , a pillar-type capacitor node is formed on the silicon substrate 1 .

The scope of the present invention depends only upon the following Claims, and is not limited by the above embodiment.