CN117871614A - Method for improving longitudinal impurity distribution accuracy of extended resistance test - Google Patents

Abstract

The invention belongs to the field of silicon-based semiconductor chip extension resistance test, and discloses a method for improving the longitudinal impurity distribution accuracy of extension resistance test, which comprises the following specific steps: step one, performing angle calibration on a sample subjected to an extended resistance test by adopting a step instrument; step two, performing needle mark calibration on the sample subjected to the extended resistance test under a 50-time microscope, and measuring and calculating the number a of points separated between the needle mark starting point and the dividing line; step three, carrying out extension resistance test on the sample again, and placing a+b test points on the left side of the boundary according to the result of the step two; and fourthly, performing needle trace tracking on the retested sample under a 50-time microscope, finding out the point number c on the left side of the boundary, removing the redundant c point numbers in subsequent result analysis, and endowing the real angle value obtained in the first test with a second extended resistance test result. The invention can calibrate the angle of the actually prepared sample and calibrate the actual test point.

Description

Method for improving longitudinal impurity distribution accuracy of extended resistance test

Technical Field

The invention belongs to the field of silicon-based semiconductor chip extension resistance test, and particularly relates to a method for improving the longitudinal impurity distribution accuracy of extension resistance test.

Background

The extended resistance technology has the characteristics of high longitudinal resolution and wide measurement range, and is widely applied to incoming material detection of silicon-based semiconductor chips and doping process detection in the preparation process. The extended resistance technology can measure the longitudinal resistance change of the silicon chip and also can measure the thickness of the silicon epitaxial wafer, the width of the transition region and the like. Although the extended resistance tester can obtain accurate longitudinal impurity distribution theoretically, in fact, the extended resistance tester is not so, and in the actual testing process, the phenomenon that the extended resistance test result of the silicon epitaxial wafer is inconsistent with the thickness result of the silicon epitaxial wafer tested by the infrared interferometry is often found, and the error is more than 2um at maximum.

1. Sample preparation and testing principle

Sample preparation firstly, a silicon wafer is attached to an angle gauge with a fixed angle, and a sample of a test wafer is produced through bevel technology processing. After the junction is exposed with the bevel treatment, a series of two-point probe tests can be performed on the bevel. At each point, the vertical distance of the probe is recorded and a resistance measurement is made. The resistance value of each point measured changes with the dopant concentration. The relation between depth and resistance value and doping content of each layer is calculated by a computer. And the computer utilizes the obtained data to construct a profile of the doping concentration of the measured sample.

2. Sample testing longitudinal distribution error analysis

The longitudinal distribution error of the impurities, which is tested by the extension resistance, is mainly in the following aspects:

2.1 Angle error caused by sample preparation Process

During sample preparation, the actual grinding angle β occurs greater or less than the angle gauge angle α due to grinding imbalance, as shown in fig. 5.

When β is greater than α in the case of fig. 2A, then the actual depth is greater than the depth tested;

when β is smaller than α in the case of fig. 2B, the actual depth is now smaller than the depth tested;

2.2 errors caused by the probe to the starting Point position

Before the test starts, a series of preset points to be tested are usually placed on the surface of the tested sample, as shown in fig. 3A, which is an ideal starting point, the center of the probe is just pressed on the boundary line between the surface and the inclined plane, but usually, the actual test point of the test equipment has a certain deviation from the preset points, as shown in fig. 3B and 3C, the starting point is either on the left side of the boundary or on the right side of the boundary, and at this time, the test depth is smaller than the actual depth or the test depth is larger than the actual depth.

Therefore, it is highly desirable to find a suitable method to improve the extended resistance test longitudinal impurity profile accuracy.

Disclosure of Invention

The invention aims to solve the technical problems that: in the actual test process, the silicon epitaxial wafer expansion resistance test result is inconsistent with the silicon epitaxial wafer thickness result tested by the infrared interferometry, and the error is larger.

In order to solve the technical problems, the specific technical scheme of the invention is as follows:

a method for improving the accuracy of longitudinal impurity distribution of an extended resistance test comprises the following specific steps:

step one, performing angle calibration on a sample subjected to an extended resistance test by adopting a step instrument;

step two, performing needle mark calibration on the sample subjected to the extended resistance test under a 50-time microscope, and measuring and calculating the number a of points separated between the needle mark starting point and the dividing line;

step three, carrying out extension resistance test on the sample again, and placing a+b test points on the left side of the boundary according to the result of the step two, wherein b is a constant;

and fourthly, performing needle trace tracking on the retested sample under a 50-time microscope, finding out the point number c on the left side of the boundary, removing the redundant c point numbers in subsequent result analysis, and endowing the real angle value obtained in the first test with a second extended resistance test result.

Wherein a+b > c.

The invention has the following advantages: the angle of the actually prepared sample is calibrated, and the actual test point position is calibrated, so that larger test errors caused by two errors can be avoided to a great extent, and useful test data can be provided for chip design better.

Drawings

FIG. 1 is a schematic diagram of an extended resistance test;

FIG. 2A is a schematic diagram of the actual angle of the sample and the angle of the angle gauge;

FIG. 2B is a schematic view of the actual angle of the sample and the angle of the angle gauge;

FIG. 3A is a schematic view of the actual probe start point position;

FIG. 3B is a schematic view of the actual probe start point position;

FIG. 3C is a schematic view of the actual probe start point position;

FIG. 4 is a schematic representation of needle marks under a 50-fold microscope;

FIG. 5 is a preset point location for performing an extended resistance test after trace tracking;

FIG. 6 is a practical point location of the 2 nd extension resistance test;

fig. 7 shows test results before and after calibration using the present method.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings for a better understanding of the objects, structures and functions of the present invention.

As shown in fig. 4, white circles in the drawing represent actual needle marks, and black circles represent points at which the distance between the actual starting point and the boundary line can be placed.

The invention relates to a method for improving the accuracy of longitudinal impurity distribution in an extended resistance test, which comprises the following specific steps:

step one, performing angle calibration on a sample subjected to an extended resistance test, wherein a step instrument is adopted in the method;

step two, performing needle mark calibration on the sample subjected to the extended resistance test under a 50-time microscope, and estimating the number of points a between the needle mark starting point and the boundary line as shown in fig. 4;

step three, carrying out expansion resistance test on the sample again, and placing b test points on the left side of the boundary according to the result of the step two, as shown in fig. 5;

and fourthly, performing needle trace tracking on the sample tested in the second time under a 50-time microscope, finding out the point number c on the left side of the boundary, removing redundant c point numbers in subsequent result analysis as shown in fig. 6, and giving the real angle value obtained in the first test to the test result of the second time of the extended resistance, wherein the pair of test results before and after calibration such as shown in fig. 7 is obtained according to the method. In fig. 7, the solid line represents the test result before calibration, and the broken line represents the test result after calibration.

According to the invention, the angle of the actually prepared sample is calibrated by adopting the step instrument, and the actual test point positions are calibrated before and after the test, so that larger test errors caused by two errors can be avoided to a great extent, and useful test data is better provided for chip design.

Although embodiments of the present invention have been described in conjunction with the accompanying drawings, it will be apparent to those skilled in the art that several variations and modifications may be made without departing from the principles of the invention, which are also considered to be within the scope of the invention.

Claims (2)

1. A method for improving the accuracy of longitudinal impurity distribution of an extended resistance test is characterized by comprising the following specific steps:

step one, performing angle calibration on a sample subjected to an extended resistance test by adopting a step instrument;

step two, performing needle mark calibration on the sample subjected to the extended resistance test under a 50-time microscope, and measuring and calculating the number a of points separated between the needle mark starting point and the dividing line;

step three, carrying out extension resistance test on the sample again, and placing a+b test points on the left side of the boundary according to the result of the step two;

and fourthly, performing needle trace tracking on the retested sample under a 50-time microscope, finding out the point number c on the left side of the boundary, removing the redundant c point numbers in subsequent result analysis, and endowing the real angle value obtained in the first test with a second extended resistance test result.

2. The method for improving the accuracy of the longitudinal impurity profile of an extended resistance test according to claim 1, wherein a+b > c.