JPH0528904B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method for evaluating silicon after epitaxial growth, and in particular, the present invention relates to a method for evaluating silicon after epitaxial growth. , the misalignment between the underlying pattern and the epitaxial layer surface pattern, that is,
The present invention relates to a method of measuring the amount of pattern shift.

[Conventional technology]

Conventionally, this type of method for measuring the amount of pattern shift involves a process of forming a buried layer on a silicon wafer, a process of performing epitaxial growth on the silicon wafer on which the buried layer is formed, and a process of depositing the epitaxially grown wafer on a diamond point or the like. The process involves cutting the chips into chips several millimeters square with a cutting tool, attaching the chips to a polishing jig using wax, and performing angle polishing. After angle polishing, the chips are removed from the polishing jig and the wax is removed. The process of cleaning and stain etching to color the embedded layer to make it visible, the process of washing and drying to remove the stain etch solution, the surface pattern, and the process of coloring by stain etching. The process consisted of measuring the amount of pattern shift using a counter attached to a microscope.

[Problem that the invention seeks to solve]

The conventional method for measuring the amount of pattern shift using the angle polishing method described above takes a long time of 3 to 5 hours to obtain measurement results, and the measurement results are not available during the mass production line of the epitaxial growth process. There was a drawback that we could not get feedback quickly.

Furthermore, in order to adjust the pattern shift amount using the stainless etching method, the buried pattern in the base is diffused in the horizontal direction as well as in the vertical direction due to the heat treatment during epitaxial growth, resulting in an error due to the horizontal diffusion. , has the disadvantage that it is unsuitable for determining an accurate pattern shift amount.

The present invention was made to solve this drawback, and it shortens the time it takes to obtain measurement results, feeds back the measurement results until the next epitaxial growth, and further improves the The present invention provides a method for measuring the amount of pattern shift that eliminates errors.

[Means for solving problems]

The method for measuring the amount of pattern shift of the present invention includes a first step of forming a buried layer on a silicon wafer, a second step of performing epitaxial growth on a part of the silicon wafer on which the buried layer is formed, and a second step of epitaxial growth. A third step is to measure the distance between elements across a step between a place where epitaxial growth has occurred and a place where no epitaxial growth has occurred, and a fourth step is to measure the distance between elements at a place where epitaxial growth has not been made.
This method is characterized in that the amount of pattern shift is determined from the distances at the two locations measured in the third and fourth steps.

〔Example〕

Next, the present invention will be explained in detail with reference to the drawings.

FIG. 1 is a principle diagram showing a part of a cross section of a silicon wafer for explaining one embodiment of the present invention.

First, by the conventional method,
A silicon wafer having a silicon substrate 1 and a embedding surface 4 formed thereon was prepared. The level difference of embedded pattern 4 at this time is 200 Å.
It was hot.

Next, the silicon wafer on which the buried pattern was formed was placed on a susceptor of a silicon epitaxial growth apparatus, and a quartz plate was placed in the center of the silicon wafer to perform epitaxial growth. At this time, a 15mm x 40mm thick piece was placed in the center of the 100mmφ silicon wafer.
A quartz plate with a size of 0.5 mm was placed. Placing the quartz plate in the center of the silicon wafer has the effect of preventing epitaxial growth in the part of the silicon wafer where the quartz plate is placed, and forms an un-epitaxially grown surface 4 in a part of the silicon wafer. A buried pattern before epitaxial growth is left on the surface of this non-epitaxially grown surface 4.

Next, the epitaxial layer thicknesses of the epitaxial growth layer 2 and the non-epitaxially grown surface 4 were measured, and it was confirmed that no epitaxial growth was made on the non-epitaxially grown surface 4 on which the quartz plate was placed. Furthermore, it was confirmed that an epitaxial displacement layer having a desired thickness of 20 μm was formed in the epitaxial growth layer 2 in the area other than where the quartz plate was placed.

Next, the distance 6 from the pattern in the epitaxial growth surface 3 to the pattern in the non-epitaxially grown surface 4 is precisely measured, and the distance 6 in the non-epitaxially grown surface 4 with the same pattern spacing as the pattern measured above is further measured. 7 was measured, and the pattern shift amount 5 was determined from the difference between the distances 6 and 7 measured at the two locations. As a result, it was confirmed that the pattern shift amount was 15 μm. The time required for this measurement was 15 minutes, and it was confirmed that the measurement time was significantly shortened compared to the conventional measurement time of 3 to 5 hours. Furthermore, it was confirmed that the pattern shift amount 5 measured at this time matched the pattern shift amount obtained by the conventional stainless etching method.

Next, it was confirmed that the distance 7 between patterns in areas where epitaxial growth was not performed matched the distance of the mask pattern used when forming the buried pattern. This proves that no error occurs due to lateral diffusion of the buried layer due to the heat treatment during epitaxial growth.

〔Effect of the invention〕

As explained above, the present invention has the advantage of being able to significantly shorten the time it takes to obtain measurement results, allowing quick feedback of measurement results, and also being able to prevent errors due to lateral diffusion of the buried layer. bring.

[Brief explanation of drawings]

FIG. 1 is a principle diagram of an embodiment of the present invention. 1... Silicon substrate, 2... Epitaxial growth layer, 3... Epitaxial growth surface, 4... Buried formation surface (non-epitaxially grown surface), 5... Pattern shift amount, 6... Distance between patterns across steps, 7 …
Distance between patterns where no epi growth occurs.

Claims (1)

[Claims] 1 A first step of forming a buried layer on a silicon wafer, and a second step of performing epitaxial growth on a part of the silicon wafer on which the buried layer has been formed.
a third step of measuring the distance between elements across a step between a place where epitaxial growth has occurred and a place where no epitaxial growth has occurred; and a third step of measuring the distance between elements at a place where epitaxial growth has not occurred. a fourth step, and determining the pattern shift amount from the difference in distance between the two locations measured in the third and fourth steps.