Method of Manufacturing Plural Active Regions Followed by a Chain Step Formation

This invention relates to a method for manufacturing a semiconductor device, and more particularly to a manufacturing method for compressing a narrow channel effect in a semiconductor device.

Recently, with the progress of a semiconductor integrated circuit manufacturing technique, a multi-element and high element-density semiconductor device have been developed. The progress of micro processing techniques in a semiconductor device manufacturing makes it possible to manufacture a very large scale integrated circuit (VLSI) employing a 1.5 m or 1 m rule. However, the manufacturing of such VLSI gives rise to various problems due to a miniaturization of the device. One of these problems is a "narrow channel effect" by which a threshold voltage V.sub.T is increased as a channel width of a transistor is narrowed. This effect causes disadvantages because a driving voltage, which is higher than a necessary voltage, must be applied to the device; or, not enough electric current can flow in the device to drive it. This means that necessary signal charges cannot be handled in a charge coupled device (CCD).

It is, therefore, an object of this invention to provide a method for manufacturing semiconductor devices by which a narrow channel effect can be compressed.

According to this invention, a method for manufacturing a semiconductor device provides an active region formed on a semiconductor substrate and defined by a channel stopper. The method comprises: a step of forming a first oxide layer by oxidizing a surface of the semiconductor substrate; a step of forming a nitride layer on the first oxide layer; a step of forming a second oxide layer having a predetermined thickness on a predetermined portion by removing the nitride layer at which the active region is to be formed on the predetermined portion and then oxidizing; a step of removing the nitride layer; a step of implanting an impurity for forming a channel stopper by using the second oxide layer as a mask; a step of removing the second oxide layer; a step of forming an oxide gate layer of the active region by oxidizing; and a step of forming an electrode.

The features and advantages of this invention will be understood from the following detailed description of a preferred embodiment of this invention, taken in conjunction with the accompanying drawings in which:

FIGS. 1(a) to 1(d) are cross sectional views of a semiconductor device, showing the principal steps of a method of manufacturing the semiconductor device according to a conventionl method; and

FIGS. 2(a) to 2(e) are cross sectional views of a semiconductor device according to an embodiment of this invention.

First, a conventional method for manufacturing a MOS semiconductor device will be described with reference to FIGS. 1(a) to 1(d).

In manufacturing a MOS semiconductor device, a P-type substrate 1 is thermally oxidized to form a first oxide layer 2. Then, a nitride layer 3 is formed on the oxide layer 2 by means of a vapor deposition (FIG. 1(a)). A portion of the nitride layer 3, at which a field is to be formed thereafter, is removed by means of an etching technique. An impurity having the same conductivity type as the substrate, for example, a boron, is diffused into the etched portion by means of an ion implantation or a thermal diffusion (FIG. 1(b)). The field portion is heavily oxidized to form a field oxide layer 5 (FIG. 1(c)). Then, the nitride layer 3 is removed. After an active region is oxidized to form a gate oxide layer and an ion implantation for a channel dope is achieved, an electrode 6 is formed (FIG. 1(d)). In FIG. 1, reference numerals 4 and 7 identify a P.sup.+ -type channel stopper and a seminconductor region for a channel dope, respectively.

The formation of the field oxide layer 5 occurs in a series of manufacturing steps. The oxidization progresses from a boundary portion between an edge portion of the nitride layer and the field, and continues toward the active region to form a non-active region called a "bird's beak." Not only such an oxide layer, but also the boron impurity diffused as the channel stopper, are spread laterally with respect to the active region.

The "bird's beak" and the laterally spread channel stopper reduce the area of the active region. More specifically, the portion, at which the active layer is thereafter to be formed, is defined to have a width L.sub.M in FIG. 1(a). The width of the final active region is reduced to L.sub.E due to the bird's beak and to the lateral spread of the channel stopper. This causes a narrow channel effect which inevitably increases the threshold voltage in the case where the channel width L.sub.M is extremely narrowed.

FIG. 2 shows a series of steps in a method of manufacturing a MOS semiconductor device according to an embodiment of this invention.

In manufacturing a MOS semiconductor device according to this embodiment, a semiconductor substrate (a P-type substrate in this embodiment) 1 is thermally oxidized to form a first oxide layer 2. Then, a nitride layer 3 is formed on the oxide layer 2 (FIG. 2(a)). Next, a portion of the nitride layer 3 is removed at an active region which is to be formed thereafter (FIG. 2(b)). An impurity for a channel dope, for example, a phosphorus or a boron, is diffused to form a thermal oxide layer 15 having a predetermined thickness. The thickness of the thermal oxide layer 15 is set so that an impurity for the channel stopper does not progress into the active region during the implanting of the impurity for the channel stopper, in the following next step (FIG. 2(c)). Then, the nitride layer 3 is removed and the impurity for the channel stopper is implanted to form the channel stopper 12 (FIG. 2(d)). Next, the oxide layer of the upper surface is removed. Then, a thermal oxidization is again achieved to form a gate oxide layer 13. Thereafter, an electrode 14 is formed (FIG. 2(e)). In FIG. 2, reference numeral 11 identifies a channel-doped semiconductor region (an N-type silicon having a reverse conductivity as compared to the conductivity of the semiconductor substrate 1).

A feature of this embodiment is that the portion, at which the channel stopper is to be formed, is covered by the nitride layer 3. The nitride layer on the other portion, at which the active region is to be formed, is removed as shown in FIG. 2(b). Further, as shown in FIG. 2(c), when the thermal oxidization is achieved, not only is the bird's beak spread from the active region to the channel stopper, but also the channel of doped N-type silicon layer 11 is laterally spread. The impurity implanted for the channel stopper (for example, the boron) is not diffused into the portion which is covered by the relatively thick oxide layer 15 and by the bird's beak. As a result, the final active region is spread out to have the final width L.sub.A which greater than the originally defined width L.sub.M, as shown in FIG. 2(e). This means that there is no reduction of the channel width, due to the bird's beak. The narrow channel effect is reduced even when the device is miniaturized.

Another feature is obtained when this invention is applied to the embedded channel CCD. This feature is based on the fact that the impurity of the N-type silicon layer which functions as the embedded channel layer is redistributed when the relatively thick thermal oxide layer 15 is formed as shown in FIG. 2(c). More specifically, when the oxidization is performed, the boundary between the silicon and the silicon dioxide is shifted to the inner portion of the silicon. The impurity in the N-type silicon layer, for example, the phosphorus or arsenic, accumulates near the boundary. This means that the practical thickness of the N-type silicon layer can be reduced to make a junction having a shallow thickness. The impurity can be distributed closer to the boundary. This makes it possible to increase the level of the signal which may be handled by the CCD.