JP3290352B2 - Semiconductor substrate crystal defect evaluation method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method for non-destructively evaluating a crystal defect of a semiconductor substrate.

[0002]

2. Description of the Related Art When a semiconductor substrate is subjected to ion implantation or the like, crystal defects occur. Conventionally, evaluation of such a crystal defect is performed by using Rutherford Back Scattering (hereinafter referred to as RBS).
Method, or transmission electron microscopy (Transmissi
on electron microscopy (hereinafter, referred to as TEM).

However, ion implantation is performed on a semiconductor substrate,
When evaluating the residual defects existing after the crystallinity is recovered by performing a heat treatment thereafter, the RBS method has a problem that the sensitivity is low. In observation using a TEM, the TEM
However, because of the high magnification, the field of view is so narrow that it is not easy to observe, it takes at least several days to prepare samples, it is impossible to evaluate many samples, and the semiconductor substrate must be destroyed. was there.

In recent years, Thermal Wave (hereinafter referred to as TW)
Method or photothermal expansion displacement (PhotoAcoustic Displaceme
Non-destructive methods for evaluating crystal defects, such as the nt (hereinafter, referred to as PAD) method, have been proposed. In both cases, the modulated excitation light is narrowed down to the order of μm to irradiate the surface of the semiconductor substrate.
The Ne laser is irradiated with the same diameter as the excitation light, the reflected light is detected, and the defect is evaluated.

The TW method is described in A. Rosencwaig et al., Appl.
hys. Lett. 46, 1013 (1985), which uses an Ar ion laser beam output from an Ar laser as a pumping beam, as shown in FIG.
The sample is modulated to MHz and irradiates the sample as excitation light. Then, the sample is irradiated with a He-Ne laser having a diameter of 1 μm, which is the same as the excitation light, to determine the reflectance R of the detection light. This reflectance R has a different rate of change ΔR / R when the excitation light is scanned on the sample between a portion having a crystal defect and a portion having no crystal defect. Therefore, TW determined by the rate of change ΔR / R
Find the value. For the measurement of this TW value, see WLSmith
et al., Appl. Lett. 47, 584 (1985).

[0006] On the other hand, the PAD method is a method proposed in Horiuchi, Crystal Processing and Evaluation Technical Committee, and the 62nd meeting of the Study Group (Japan Society for the Promotion of Science, 1993). FIG.
As shown in (2), the sample surface is irradiated with excitation light from an LED having a wavelength of 813 nm and modulated at 87 kHz. This causes the sample surface to thermally expand, causing displacement. This photothermal displacement (Pa value) is measured by a phase change of the He-Ne laser detection light.

Conventionally, for example, T. Hara et al., I
n semiconductor Silicon, editedby HRHuff et al.
(The Electrochem. Soc. Proc. 90-7, Pennington, NJ,
(1990) pp.972-982)
The results of evaluating residual defects by the method a have also been reported. However, all of these reports merely compare the magnitude of the TW value or the Pa value, and have been qualitatively evaluated.

[0008]

As described above, conventionally, it has not been possible to quantitatively evaluate crystal defects of a semiconductor substrate without destruction.

In view of the above circumstances, an object of the present invention is to provide a semiconductor substrate crystal defect evaluation method capable of non-destructively, quantitatively and quickly evaluating crystal defects of a semiconductor substrate.

[0010]

According to the method for evaluating crystal defects of a semiconductor substrate of the present invention, the surface of a semiconductor substrate having a crystal defect is irradiated with modulated excitation light by performing a predetermined process, and detection light is applied to the same portion. A step of measuring at least one of a TW value obtained by a change in reflectance at the time of irradiation or a Pa value of a semiconductor substrate surface obtained by irradiating modulated excitation light and irradiating detection light to the same place. Observing a surface portion of the semiconductor substrate and measuring at least one of a density of crystal defects and a total peripheral length obtained by summing peripheral lengths of the crystal defects;
Value and the density of the crystal defects, the TW value and the total peripheral length of the crystal defects, the Pa value and the density of the crystal defects, or at least one of the Pa value and the total peripheral length of the crystal defects. Obtaining a predetermined relationship established between the combinations; measuring at least one of a TW value and a Pa value with respect to another semiconductor substrate different from the semiconductor substrate; Determining a density or a total perimeter of crystal defects using the value or the Pa value and the predetermined relationship.

Here, the measurement of the TW value or the Pa value is performed by measuring one of the surfaces of the semiconductor substrate and the other semiconductor substrate.
Irradiating a point or a plurality of points with excitation light and detection light;
The scanning may be performed by scanning excitation light and detection light on one or more lines.

In addition, the semiconductor substrate and the other semiconductor substrate are ones containing crystal defects by implanting impurity ions and performing an annealing process (in many cases, the defect types take the form of dislocation loops). You may.

The density or the total peripheral length of the crystal defects may be measured by observing a sample prepared from the surface of the semiconductor substrate and the other semiconductor substrate using a transmission electron microscope. Alternatively, the density of the crystal defects is measured by selectively etching the semiconductor substrate and the other semiconductor substrate to generate etch pits in the crystal defects, and observing the surface using a scanning electron microscope. You may.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

First, FIG. 1 shows a procedure of a method for evaluating a crystal defect of a semiconductor substrate according to the present embodiment. Step 101
Then, impurity ions are implanted into a semiconductor substrate, activation annealing is performed, and a sample having crystal defects is formed.

In step 102, the sample is
The TW value is measured by the TW method described above, or PAD
The Pa value is measured by the method.

In step 103, the surface of the same sample is observed using a TEM, the number of crystal defects existing per unit area is counted to determine the defect density, or the total length obtained by summing the peripheral length of each crystal defect is calculated. Find the perimeter. Here, the total perimeter can be determined by 2πrN using the number N of crystal defects existing per unit area and the average radius r of the crystal defects.

In step 104, the correlation between the TW value or Pa value and the defect density, or the correlation between the TW value or Pa value and the total perimeter is determined.

Step 105: TW of unknown sample
The value or the Pa value is measured.

In step 106, the correlation between the TW value or Pa value of the unknown sample and the TW value or Pa value obtained in step 104 and the defect density, or the correlation between the TW value or Pa value and the total peripheral length. Based on the relationship, the defect density or the total peripheral length of the sample is obtained, and the crystal defect is evaluated.

By the above steps 101 to 106, the defect density or the total peripheral length of the unknown sample is obtained, and the crystal defects can be quantitatively evaluated.

Next, using a silicon substrate, arsenic (A) generally used for forming a MOS transistor is used.
An example in which crystal defects remaining after performing s ⁺ ) ion implantation and activation annealing will be described with reference to FIG.

In step 111, the plane orientation (10
0), an n-type CZ silicon substrate was prepared.

In step 112, a silicon oxide film having a thickness of 25 nm was formed on the silicon substrate by a thermal oxidation method.

In step 113, arsenic ions are implanted on the surface of the silicon substrate at an acceleration voltage of 60 keV and a dose of 1 *.
The injection was performed at 10 ^{15 to} 3 * 10 ¹⁶ / cm ² .

In step 114, the silicon substrate was annealed at 800, 850, and 900 degrees Celsius for 10 minutes in a diffusion furnace in a nitrogen (N ² ) atmosphere.

In step 115, the silicon oxide film was removed.

In step 116, the TW value of the silicon substrate was measured by the TW method. For the measurement of the TW value, a Thermo-Probe TP-200 was used, contour mapping measurement was performed on the entire surface of the silicon substrate, and a good value within a standard deviation of 3% was obtained, and an average value was obtained.

In step 117, a cross section and a plane of the sample were observed using a TEM. FIG. 3 shows a distribution of crystal defects obtained by observing a cross section of the sample by a cross-sectional TEM method. Small dislocation loops having a diameter of about 24 nm were distributed in a state localized at almost the same depth (about 73 nm) from the surface. FIGS. 4A to 4C show changes in the sample surface when the dose amount and the annealing temperature are changed. FIG. 4A shows a dose amount of 1 * 10 ¹⁵ / cm ² and an annealing temperature of 80 degrees Celsius.
FIG. 4B shows a dose of 1 * 10 ¹⁵ / cm ² and an annealing temperature of 900 ° C. FIG. 4C shows a dose of 1 *
The planar TEM observation results near the respective sample surfaces when the annealing temperature is 10 ¹⁶ / cm ² and the annealing temperature is 850 degrees Celsius are shown. Thus, it can be seen that the density of the residual defects changes when the ion implantation conditions and the annealing conditions change.

The defect density of the sample used in this embodiment is as high as 10 to 270 defects / μm ² , and the defect has a uniform density in a measurement area with a beam diameter of 1 μm by one TW method. Can be considered as distributed.

In step 118, the density, size, and total peripheral length of the dislocation loop are obtained. By observing the sample surface as shown in FIG. 4, the number of defects per unit area present on the silicon substrate, the size of each defect (mainly the diameter), and the periphery calculated from the diameter of each defect Find the total length.

In step 119, the relationship between the TW value, the dislocation loop density, and the total perimeter is determined. FIG. 5 shows the relationship between the TW value measured in step 116 and the dislocation loop density determined in step 118. FIG.
The relation of the total perimeter of the dislocation loop to the value is shown. As is clear from the comparison between FIG. 5 and FIG. 6, it is the total peripheral length of the dislocation loop that shows a clear correlation with the TW value. There is a proportional relationship between the increase in the TW value and the increase in the total perimeter, and can be approximated by a straight line using the least squares method.

Next, a description will be given of the evaluation of the defect of the unknown sample using the relationship between the TW value obtained in the above steps 111 to 119 and the density of the dislocation loop or the total peripheral length.

As step 120, the TW of the unknown sample
The value was measured. FIG. 7 shows that the dose is 0, 1, 3, 7, 1
It shows the TW value when it is changed as 0 * 10 ¹⁴ / cm ² . It can be seen from FIG. 7 that the TW value is almost proportional to the dose. Here, processing conditions other than the dose amount are as follows:
The same as in steps 111 to 115 was used.

Further, as shown in FIG. 9A, ion implantation is performed under the same conditions as in an actual product.
The measurement was performed on a region 201 of 00 μm * 300 μm and a plurality of regions 202 having a width of 20 μm arranged at intervals of 20 μm shown in FIG. FIG. 9 shows the scanning for performing the TW measurement.
The operation was performed in the directions of the arrows shown in (a) and (b).

FIG. 8 shows the change in the TW value when the heat treatment temperature was changed while the ion implantation conditions were constant (arsenic ions were implanted at an acceleration voltage of 60 keV and a dose of 1 * 10 ¹⁵ / cm ² ). Show. When the temperature is 500 degrees Celsius or less, it has become amorphous by ion implantation, and the TW value shows a high value. When the temperature exceeds 600 degrees Celsius, the amorphous phase is crystallized by solid phase growth, and when the annealing temperature is increased, crystal defects are reduced.

As in step 121, the relationship between the TW value of the unknown sample shown in FIG. 7 or FIG. 8 and the density of the dislocation loop or the total peripheral length with respect to the TW value shown in FIG. 5 or FIG. , The dislocation loop density or the total perimeter of the unknown sample can be determined.

In the above embodiment, the TW value was measured.
When the a value is measured, the above-described steps 111 to 12 are also performed.
1 can be applied.

According to the evaluation method of the present embodiment, the TW value or Pa value of a sample containing a defect prepared in advance is measured,
Defect density or total perimeter is determined by measuring the density or total perimeter of the defect by EM observation, and measuring the TW value or Pa value for the unknown sample to be evaluated. It can be easily obtained. Therefore, according to the present embodiment, it is possible to quantitatively evaluate crystal defects by non-destruction, and it is also possible to monitor changes in crystal defects of the same substrate during processing.

In a semiconductor manufacturing process, it is known that a crystal defect generated by implanting impurity ions remains without being eliminated by a subsequent low-temperature process, thereby causing a leak at a shallow junction. . Therefore, by using the crystal defect evaluation method according to the present invention,
Optimal process conditions for reducing residual defects can be determined. As described above, the present invention can contribute to improvement in the yield and reliability of the device.

The above-described embodiment is merely an example, and does not limit the present invention. For example, in the embodiment, T
Although the sample surface is observed using EM, it is not limited to this. For example, the density of defects or the total perimeter may be determined by performing selective etching on a sample in which only crystal defects selectively generate etch pits and observing the surface with a scanning electron microscope.

In the embodiment, as shown in FIG. 9, the TW value is measured by scanning the excitation light and the detection light on one line of the surface of the semiconductor substrate, and the mapping measurement (multiple points in the wafer surface) is performed. In some cases, the measurement may be performed by irradiating one or more points, or by scanning a plurality of lines.

Further, the embodiment is applied to the case of evaluating a crystal defect after implanting impurity ions. However, the evaluation method of the present invention can be applied to those in which crystal defects have occurred due to processes other than ion implantation. For example, the present invention can be applied to evaluation of a crushed layer generated by machining such as slicing, grinding, polishing, and honing, or evaluation of a high-density dislocation loop, and evaluation of stacking faults and oxygen precipitation. it can.

[0044]

As described above, according to the method for evaluating a semiconductor substrate of the present invention, the TW value or Pa value of a sample prepared in advance is measured, the surface is observed with a microscope, and the density of defects or the total area is measured. It is possible to determine the defect density or the total perimeter by using the relationship determined by measuring the TW value or the Pa value of the sample to be evaluated by measuring the length and determining the relationship between the two. Crystal defects can be evaluated quantitatively.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a method of evaluating a crystal defect of a semiconductor substrate according to an embodiment of the present invention for each process.

FIG. 2 is a flowchart showing an evaluation method in each example in which a crystal defect of a semiconductor substrate is evaluated according to the embodiment.

FIG. 3 is a longitudinal sectional view showing a distribution of crystal defects in a sample section observed by TEM in the example.

FIG. 4 is a TEM photograph showing the distribution of crystal defects on the sample surface observed by TEM in the same example.

FIG. 5 is a graph showing the relationship between the TW value and the density of dislocation loops in the example.

FIG. 6 is a graph showing a relationship between a TW value and a total peripheral length of a dislocation loop in the example.

FIG. 7 is a graph showing a relationship between a dose and a TW value in the example.

FIG. 8 is a graph showing a relationship between an annealing temperature and a TW value in the example.

FIG. 9 is an explanatory view showing an impurity ion implanted region in the embodiment.

FIG. 10 is a block diagram showing a schematic configuration of an apparatus used for evaluating a crystal defect of a semiconductor substrate by a TW method.

FIG. 11 is a block diagram showing a schematic configuration of an apparatus used for evaluating a crystal defect of a semiconductor substrate by a PAD method.

[Explanation of symbols]

 201, 202 area

──────────────────────────────────────────────────続 き Continuation of front page (56) References JP-A-62-119439 (JP, A) JP-A-63-244752 (JP, A) JP-A-4-348050 (JP, A) JP-A-6-119 338548 (JP, A) JP-A-8-21801 (JP, A) JP-B-46-27639 (JP, B1) (58) Fields investigated (Int. Cl. ⁷ , DB name) H01L 21/66 H01L 21 / 265

Claims (5)

(57) [Claims] 1. A method according to claim 1, wherein a predetermined processing is performed to irradiate a modulated excitation light onto a surface of the semiconductor substrate on which a crystal defect exists, and a thermal wave value obtained by a change in reflectance when the same location is irradiated with detection light. Hereinafter, at least one of TW value) and photothermal expansion displacement (hereinafter, referred to as Pa value) of the semiconductor substrate surface obtained by irradiating modulated excitation light and irradiating the same portion with detection light is measured. Observing a surface crystal portion on or near the surface of the semiconductor substrate, and measuring at least one of a density of crystal defects, or a total peripheral length obtained by summing the peripheral lengths of the crystal defects, A TW value and a density of the crystal defect, a TW value and a total peripheral length of the crystal defect, a Pa value and a density of the crystal defect,
Or a step of obtaining a predetermined relationship that is established between at least one of the combination of the Pa value and the total peripheral length of the crystal defects; and TW for another semiconductor substrate different from the semiconductor substrate.
Measuring at least one of the value or the Pa value; and using the TW value or the Pa value of the other semiconductor substrate and the predetermined relationship to determine a density or a total peripheral length of crystal defects. A method for evaluating crystal defects of a semiconductor wafer, comprising: 2. The method of measuring the TW value or the Pa value, comprising irradiating one or more points on the surface of the semiconductor substrate and the other semiconductor substrate with excitation light and detection light, or using one or more lines. 2. The method according to claim 1, wherein the scanning is performed by scanning the line with excitation light and detection light. 3. The semiconductor substrate according to claim 1, wherein said semiconductor substrate and said another semiconductor substrate contain crystal defects by implanting impurity ions and performing an annealing process.
The method for evaluating crystal defects of a semiconductor substrate according to the above. 4. The method according to claim 1, wherein the measurement of the density or the total peripheral length of the crystal defects is performed by observing the surfaces of the semiconductor substrate and the other semiconductor substrate using a transmission electron microscope. 4. The method for evaluating a crystal defect of a semiconductor substrate according to any one of claims 1 to 3. 5. A method of measuring the density of crystal defects, comprising selectively etching the semiconductor substrate and the other semiconductor substrate to generate etch pits in the crystal defects, and observing the surface using a scanning electron microscope. The method according to claim 1, wherein the method is performed.