JP2672743B2 - Evaluation method of contaminant impurities - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an evaluation method for evaluating the type and amount of contaminant impurities generated in the manufacturing process of semiconductor devices.

[0002]

2. Description of the Related Art Semiconductor devices such as MOS LSIs and bipolar ICs are manufactured through hundreds of manufacturing processes such as a cleaning process, an oxidation process, and a photolithography process. In such a manufacturing process, when heavy metals such as iron, nickel and chromium are mixed in, the performance of the semiconductor device is deteriorated, so that heavy metal contamination is removed with great care. However, with the progress of higher integration and higher densities, new manufacturing processes have to be introduced, and the required cleanliness has also become greater than before, so the need to understand heavy metal contamination is greater than ever before. It is getting bigger.

Under such circumstances, Auger analysis (AE) is a method for detecting the degree of contamination by heavy metals.
There are various physical analysis methods such as S), SIMS, and total reflection X-ray fluorescence analysis. Such an analysis method has an advantage that the type of contaminant impurities can be determined and the amount of the impurities can be obtained. However, the former two have the disadvantages that it is necessary to cut the semiconductor substrate into a sample of a desired size, and that the sample needs to be put into an evaluation device and evacuated, and that it takes time. . Furthermore, when cutting the sample,
It is possible that other contaminating impurities are mixed in and the measurement data is confused. The total reflection fluorescence X-ray analysis has an advantage that it can be evaluated without cutting the semiconductor substrate, but has a disadvantage that it takes time to measure the in-plane distribution of the wafer (semiconductor substrate) in detail. In addition, physical analysis generally has insufficient detection sensitivity, and remains unsatisfactory as compared with electrical measurement.

On the other hand, there is a method of measuring the lifetime of a semiconductor substrate and detecting the presence or absence of contaminant impurities based on the measured value.
This utilizes the phenomenon that the carrier concentration is reduced and the carrier lifetime is reduced when contaminant impurities are mixed in the semiconductor substrate, and has advantages such as simplicity and high sensitivity. In addition, in this method, the lifetime can be measured by a non-contact method, and since it is not necessary to draw a vacuum, it is possible to measure in a short time without adding any processing to the semiconductor substrate to be measured. There are advantages.

In order to improve the yield, it is necessary to detect the type of contaminant impurities, the amount of contaminant impurities, the position of the contaminant source, etc. in order to eliminate the contaminant source. However, the conventional lifetime measurement method has a problem in that it is not possible to detect the type of contaminant impurities, the amount of contaminant impurities, and the position of the contaminant source that contributed to the decrease in lifetime.

[0006]

As described above, in the conventional method, there is a method of easily measuring the presence or absence of contaminant impurities in a short time without contact, but the type of impurities cannot be specified. was there.

The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a method capable of easily detecting the type of impurities in addition to the presence or absence of impurities.

[0008]

Therefore, in the present invention, an n-type semiconductor substrate and a p-type semiconductor substrate are prepared, a process for evaluating the contamination occurrence state is added to these, and the n-type semiconductor substrate after the addition is provided. And the p-type semiconductor substrate are respectively detected with a carrier lifetime, and the detected value is compared with the correlation of the lifetime data of the impurity with respect to the n-type semiconductor substrate and the p-type semiconductor substrate which have been measured in advance. The type and / or amount of impurities are detected from the result.

[0009]

In the present invention, when contaminant impurities are mixed in the semiconductor substrate, the influence on the tie-time differs depending on whether the carriers of the semiconductor substrate are electrons or holes, that is, the conductivity type. According to this, it becomes possible to easily detect the type of contaminant impurities in a non-contact manner.

That is, the carrier time of the semiconductor substrate of each conductivity type when it is previously contaminated with various heavy metals is measured and stored as a reference value. The lifetime is measured for each conductivity type, and each value is compared with the reference value to identify the type of impurity.

If the impurity concentration is changed to measure the reference value of the lifetime and the impurity concentration is quantitatively measured, not only the type of contaminant impurities but also the impurity concentration can be identified. . Further, since the measurement on each wafer surface can be performed in a short time in a non-contact manner, the impurity distribution can be easily measured.

[0012]

Embodiments of the present invention will be described below in detail with reference to the drawings.

First, p-type impurities and n having the same impurity concentration are used.
A p-type silicon substrate and an n-type silicon substrate, into which type impurities have been injected, respectively, are prepared, and a first heating furnace is used to
Heat treatment was performed at 00 ° C to 1150 ° C for 1 hour. After that, the lifetime was measured using the photoconductive decay method. This method is a method of irradiating a sample with light for a certain period of time, and measuring the time from when the carriers are excited to the conduction band to when they return to the ground state. Here, the measurement was performed at 5 points and the average value was plotted on the XY coordinates. The results are shown by black circles in FIG. For comparison, the lifetimes of the p-type silicon substrate and the n-type silicon substrate without contamination (before heating) are indicated by triangle marks in FIG. Here, both are 400 μsec.

Here, the heat treatment temperature is 1000 to 1050.
In the range of ° C, the pollution degree is low, and the amount of contaminant impurities is small.
The lifetime is reduced only for the silicon substrate.
This change is continuous.

When the temperature exceeds 1100 ° C., the amount of contaminant impurities increases, and the p-type and n-type lifetimes decrease.

As a result of total reflection X-ray fluorescence analysis on these silicon substrates, iron (Fe) of each concentration was detected corresponding to the lifetime value. The concentration values in FIG. 1 are based on this X-ray analysis.

Next, using the second heating furnace, 1000
The heat treatment was performed at 1 ° C to 1150 ° C for 1 hour. After that, the lifetime was measured using the photoconductivity decay method.
It is plotted on the Y coordinate. The results are shown by white circles in FIG.

As is apparent from this figure, the lifetime is reduced for both p-type and n-type silicon substrates.

As a result of total reflection X-ray fluorescence analysis on these silicon substrates, copper (Cu) at each concentration was detected corresponding to the lifetime value. The concentration values in FIG. 1 are based on this X-ray analysis.

Since the black circle and the white circle draw different trajectories in this way, if the contaminant impurities are unknown, the lifetime can be measured and plotted on this drawing to easily distinguish between iron and copper.

The results of the heat treatments in the first and second heating furnaces are plotted in FIG. 1 on the XY coordinates as reference values. That is, the lifetime is measured and the correlation is obtained with respect to the type and the concentration of the impurity, while the type and the concentration of the impurity are detected by another method in advance.

After the preparation is carried out in this way, the p-type and n-type silicon substrates are heat-treated in the third heating furnace which is the object of contamination measurement. The lifetimes of the p-type and n-type silicon substrates were measured by the photoconductivity decay method and plotted on FIG. 1 to find that they were at the positions indicated by A in FIG. This point A is on the iron standard curve and is 1 × 10 ¹¹ cm ^-2 to 1 × 1
It lies between 0 ¹⁰ cm ^-2 . Therefore, it can be seen that in the third heating furnace, 1 × 10 ¹¹ cm ^-2 to 1 × 10 ¹⁰ cm ^-2 is mixed as iron as a contaminant impurity.

For confirmation, a total reflection X-ray fluorescence analysis was carried out. As a result, iron (Fe) was detected at 2 × 10 ¹⁰ cm ^-2 .

As described above, according to the method of the present invention, the type and amount of the contaminant impurities can be measured very easily and in a short time without contact and nondestruction. With this method, in-plane measurement is 5 minutes for 2 minutes,
Even 500-point measurement can be performed in 1 hour, and can be easily performed in an extremely short time. Therefore, the pollution factor (source)
Can be extracted, and pollution reduction measures can be easily taken.

Furthermore, since detailed data of the impurity distribution on the wafer surface can be obtained in a short time, the position and direction of the contamination source can be detected with high accuracy.

In the above-mentioned embodiment, the reference data was measured only for two kinds of contaminated impurities, but if a large number of data are obtained and stored for possible impurities, various kinds of data will be obtained. It can be measured for impurities.

Although the measured values are plotted on the XY coordinates in the above embodiment, they may be stored as a database and calculated by a computer.

Further, the amount of contamination may be detected by using the average value of the measured values at each position on the wafer surface.
The distribution of contaminant impurities at each position may be obtained.

[0029]

As described above, according to the present invention, by measuring the carrier lifetime for both the p-type substrate and the n-type substrate, non-contact, non-destructive, easy and short-time operation is possible. It is possible to detect the type of contaminant impurities.

[Brief description of the drawings]

FIG. 1 is a diagram showing a result of measuring a lifetime by a method according to an embodiment of the present invention.

[Explanation of symbols]

 A detection point

Claims (1)

(57) [Claims] 1. A step of mixing a contaminant impurity to be evaluated into both a p-type semiconductor substrate and an n-type semiconductor substrate, measuring a carrier lifetime, and storing each measured value as a reference value, The p-type semiconductor substrate and the n-type semiconductor substrate are processed in the same manner in the manufacturing process in which the contamination generation status is to be evaluated, and the carrier lifetime is detected for both the processed n-type semiconductor substrate and p-type semiconductor substrate. Then, the method includes a step of comparing the detected value with the previously measured reference value, and detecting the type of contaminant impurities based on the correlation between the p-type and n-type lifetimes from the comparison result. Evaluation method of contaminant impurities.