JP2007263960A - Quantitative evaluation device and method of atomic vacancy existing in silicon wafer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a quantitative evaluation device of atomic vacancy existing in a wafer, in which the atomic vacancy concentration in a silicon wafer can be evaluated quantitatively without performing acceleration processing, for example, raising the concentration by forming a thin film oscillator subjected to rationalization on the surface of a silicon sample. <P>SOLUTION: The quantitative evaluation device of atomic vacancy comprises a magnetic force generating means 2 for applying an external magnetic field to a silicon sample 5 obtained by cutting out a predetermined part from a silicon wafer; a temperature control means 3 which can cool/control the silicon sample 5 to a temperature zone of 50K or below: and an ultrasonic oscillation/detection means 4 for oscillating an ultrasonic pulse on the surface of the silicon sample 5 to allow the pulse to propagate through the silicon sample 5 and detecting a variation in sound velocity of the propagating ultrasonic pulse. The quantitative evaluation device is characterized in that a thin film oscillator 8 having physical properties capable of following up expansion of the silicon sample 5 in the temperature zone and having C-axis arranged in a predetermined direction is formed directly on the surface of the silicon sample 5. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention can identify the types of atomic vacancies present in a silicon crystal wafer produced by the Czochralski method (CZ method) or the float zone method (FZ method) used in the semiconductor industry, and The present invention relates to an apparatus and method for quantitative evaluation of atomic vacancies existing in a silicon wafer, in which the pore concentration can be directly and quantitatively evaluated without indirect methods such as measuring the total void volume. is there.

  Silicon crystals are considered to be the purest and ideal crystals that humans have. However, since there is an entropy term of free energy, at the melting point of 1412 ° C. where the crystal is grown, there is always a disorder of the crystal due to intrinsic point defects (atomic vacancies and interstitial silicon).

  There has been no means for measuring the abundance of isolated atomic vacancies in a silicon crystal. However, for example, by performing a heat treatment such as a silicon crystal growth process or a silicon device manufacturing heating process, oxygen precipitates formed by the reaction of interstitial oxygen atoms and atomic vacancies that are supersaturated in the CZ crystal are formed. By making it manifest, the presence of atomic vacancies was qualitatively determined. In addition, excessive atomic vacancies introduced during crystal solidification aggregate in the cooling process during crystal growth, and grow into voids that are secondary defects having a size of about 100 nm. The concentration of atomic vacancies introduced during crystal growth was estimated by measuring the volume by infrared tomography. However, such a method only indirectly measures the abundance of atomic vacancies.

For this reason, one of the inventors of the present invention has proposed a method in Patent Document 1 that can quantitatively measure the concentration of atomic vacancies in a silicon crystal wafer without performing acceleration processing. The method described in Patent Document 1 applies an external magnetic field to a crystal sample, passes ultrasonic waves through the crystal sample while cooling, changes in the ultrasonic sound velocity or ultrasonic absorption in the crystal sample, and the cooling temperature of the crystal sample. Inherent point defect concentration can be obtained based on the steep drop amount of the curve showing the relationship, and in order to oscillate and receive ultrasonic pulses on the silicon wafer as the test material, A vibrator made of, for example, LiNbO ₃ is attached via an adhesive.
JP-A-7-174742

However, when the inventors conducted further detailed studies after that, when the silicon wafer was cooled to an extremely low temperature of 50K or less, the vibrator might partially peel from the silicon wafer surface due to the cooling. When the vibrator is peeled off, there is a problem that it is impossible to accurately detect the change in the speed of the ultrasonic pulse propagated through the silicon sample.
The reason why the vibrator peels off from the silicon wafer surface is that when the vibrator is cooled to an extremely low temperature of 50K or less, the vibrator contracts, while the silicon wafer expands. Therefore, there is a large difference in thermal expansion between the vibrator and the silicon wafer. It is considered that peeling occurs due to this.

  An object of the present invention is to produce a thin film vibrator with optimization on the surface of a silicon sample, and to manufacture it by the Czochralski method (CZ method) or the float zone method (FZ method) used in the semiconductor industry. A silicon wafer that can identify the type of atomic vacancies present in a silicon crystal wafer and quantitatively evaluate the atomic vacancy concentration without performing acceleration processing such as increasing the concentration. An object of the present invention is to provide a quantitative evaluation apparatus and method for atomic vacancies existing therein.

In order to achieve the above object, the gist of the present invention is as follows.
(1) Magnetic force generating means for applying an external magnetic field to a silicon sample obtained by cutting a predetermined part from a silicon wafer, temperature control means for cooling and controlling the silicon sample to a temperature range of 50K or less, and the surface of the silicon sample In contrast, it has an ultrasonic oscillation / detection means that oscillates an ultrasonic pulse, propagates the oscillated ultrasonic pulse through the silicon sample, and detects a change in sound velocity of the propagated ultrasonic pulse. It exists in a silicon wafer characterized by directly forming a thin film vibrator having physical properties capable of following the expansion of a silicon sample accompanying a temperature drop in the temperature range and having the C axis substantially aligned in a predetermined direction. Quantitative evaluation device for atomic vacancies.

(2) The ultrasonic wave oscillating / detecting means includes a reference wave pulse signal obtained by directly measuring the oscillated ultrasonic pulse, and a sample passing wave pulse signal measured after the ultrasonic pulse is propagated through the silicon sample. The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to (1), comprising means for detecting a phase difference.

(3) The apparatus for quantitative evaluation of atomic vacancies present in a silicon wafer according to (1) or (2), wherein the thin film vibrator is made of zinc oxide (ZnO) or aluminum nitride (AlN).

(4) The apparatus for quantitative evaluation of atomic vacancies in the silicon wafer according to (1), (2) or (3), wherein the thin film vibrator is formed on the silicon wafer by physical vapor deposition.

(5) The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to any one of the above (1) to (4), which has a gold thin film between a thin film vibrator and a silicon crystal.

(6) The thin-film transducer has a C-axis inclined at an angle of 5 to 60 ° with respect to the surface of the silicon sample. Of the longitudinal wave component and the transverse wave component in the ultrasonic wave detected by propagating through the silicon sample, The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to any one of (1) to (5), wherein at least a shear wave component is measured.

(7) The apparatus for quantitative evaluation of atomic vacancies present in a silicon wafer according to any one of (1) to (6) above, wherein the thickness of the thin film vibrator is in the range of 0.5 to 200 μm.

(8) The apparatus for quantitative evaluation of atomic vacancies present in a silicon wafer according to any one of (1) to (7) above, wherein the resonance frequency of the thin film vibrator is in the range of 10 MHz to 10 GHz.

(9) The apparatus for quantitative evaluation of atomic vacancies in the silicon wafer according to any one of (1) to (8), wherein the magnetic force generating means is in the range of 0 to 20 Tesla.

(10) The apparatus for quantitative evaluation of atomic vacancies present in a silicon wafer according to any one of (1) to (9), wherein the temperature control means includes a dilution refrigerator that can be cooled to an extremely low temperature of 5 mK.

(11) The apparatus for quantitatively evaluating atomic vacancies existing in a silicon wafer according to any one of the above (1) to (10), wherein the ultrasonic wave generation / detection means uses an ultrasonic pulse having a pulse width of 10 μs or less. .

(12) The ultrasonic wave generation / detection means includes means for performing zero detection by changing the oscillation frequency so that a phase difference caused by a change in sound speed due to a temperature or a magnetic field is constant. The quantitative evaluation apparatus of the atomic vacancy which exists in the silicon wafer of any one of these.

(13) Atomic vacancies present in the silicon wafer according to any one of (1) to (12), wherein a phase difference can be measured simultaneously with a plurality of silicon samples and a plurality of points of one silicon sample as measurement targets. Hole quantitative evaluation device.

(14) With a temperature drop of the silicon sample in the temperature range while cooling in a temperature range of 25 K or less with an external magnetic field applied to the silicon sample cut out from the silicon wafer as required. An ultrasonic pulse is oscillated in a silicon sample directly formed on the surface of a thin film vibrator having physical properties that can follow expansion and the C axis is aligned in a predetermined direction, and the oscillated ultrasonic pulse is passed through the silicon sample. Propagation, detecting the change in the sound velocity of the propagated ultrasonic pulse, calculating the decrease in elastic constant accompanying the decrease in cooling temperature from this change in sound velocity, and present in the silicon wafer from the decrease in elastic constant calculated A method for quantitative evaluation of atomic vacancies present in a silicon wafer, characterized by quantitative evaluation of the type and concentration of atomic vacancies.

  According to the present invention, an optimized thin film vibrator is formed on the surface of a silicon sample, and is manufactured by the Czochralski method (CZ method) or the float zone method (FZ method) used in the semiconductor industry. The type of isolated vacancies present in a silicon crystal wafer can be identified, and the concentration of vacancies can be directly and quantitatively evaluated without performing acceleration processing such as increasing the concentration. be able to.

Next, a quantitative evaluation apparatus for atomic vacancies existing in a silicon wafer according to the present invention will be described with reference to the drawings.
FIG. 1 is a schematic diagram of a quantitative evaluation apparatus of the present invention.

  The quantitative evaluation apparatus 1 shown in the figure is mainly composed of a magnetic force generation means 2, a temperature control means 3, and an ultrasonic oscillation / detection means 4.

  The magnetic force generating means 2 is disposed so as to surround a position where the silicon sample 5 is set in order to apply an external magnetic field to the silicon sample 5 obtained by cutting a predetermined part from the silicon wafer. An example of the magnetic force generating means 2 is a superconducting magnet. Further, in the present invention, in order to detect a change in the sound velocity of the ultrasonic pulse propagated through the silicon sample 5 with an external magnetic field applied to the silicon sample 5 as necessary, the magnetic force generating means 2 includes 0-20. It is preferably controllable within the range of Tesla, more preferably within the range of 0 to 6 Tesla (see FIG. 9). For example, the type of isolated atomic vacancies in a silicon crystal wafer can be specified by applying an external magnetic field, which will be described later.

  The temperature control means 3 is configured to be able to cool and control the silicon sample 5 in a temperature range of 50K or less. FIG. 1 shows a case where a dilution refrigerator is used as the temperature control means 3. This dilution refrigerator is cooled and controlled to a cryogenic temperature of, for example, 4.2 K on the upper side of the apparatus and 5 mK on the lower side of the apparatus by appropriately circulating a mixed solution of helium 3 and helium 4 in the mixing chamber 6. can do. FIG. 1 shows a configuration in which the sample holder portion 7 on which the silicon sample 5 is set is immersed in a mixed solution of helium 3 and helium 4 in the mixing chamber 6 and directly cooled, but only in this configuration. Is not limited. For example, the member that forms the cooled mixing chamber 6 can be made of a material having high thermal conductivity, and the silicon sample 5 can be indirectly cooled using the heat conduction from the member that forms the mixing chamber 6. In the case of such a configuration, it is particularly advantageous in that the temperature range for cooling can be expanded to the high temperature side.

  The ultrasonic wave oscillating / detecting means 4 oscillates an ultrasonic pulse on the surface of the silicon sample 5, propagates the oscillated ultrasonic pulse through the silicon sample 5, and detects a change in sound velocity of the propagated ultrasonic pulse. To place.

  FIG. 2 is an enlarged view of the sample holder portion 7 in which the silicon sample 5 is set, which constitutes the quantitative evaluation apparatus 1 of FIG.

  In the present invention, prior to setting the silicon sample 5, the surface of the silicon sample 5 has physical properties that can follow the expansion of the silicon sample 5 in a temperature range of 50K or less, and the C axis is aligned in a predetermined direction. The thin film vibrator 8 is formed directly or via a gold thin film. By adopting this configuration, even if the silicon wafer is cooled to an extremely low temperature of 50K or less, the thin film vibrator can follow the expansion of the silicon wafer. As a result of the accurate detection of changes in the speed of ultrasonic pulses that have propagated through the sample, the type and concentration of isolated atomic vacancies in the silicon crystal wafer must be accelerated to increase the concentration. And can be directly and stably quantitatively evaluated.

  Further, as shown in FIG. 3, the ultrasonic oscillation / detection means 4 measures a reference wave pulse signal obtained by directly measuring the oscillated ultrasonic pulse, and after the ultrasonic pulse is propagated through the silicon sample. A means for detecting a phase difference from the sample passing wave pulse is preferable.

  The thin film vibrator 8 is preferably made of zinc oxide (ZnO) or aluminum nitride (AlN).

  The thin film vibrator 8 is formed on a silicon wafer by physical vapor deposition such as sputtering, for example. The silicon wafer and zinc oxide (ZnO) are closely bonded at an atomic level, and zinc oxide having excellent adhesion. As a result of forming (ZnO) on the silicon wafer, it is preferable in that the physical properties can follow the expansion of the silicon sample 5 in a temperature range of 50K or less.

  In addition, it is more preferable to have a gold vapor deposition film between the thin film vibrator 8 and the silicon sample 5 in terms of preventing peeling during cooling and increasing the conductivity.

  The thin film vibrator 8 has a C axis inclined at an angle of 5 to 60 ° with respect to the surface of the silicon sample, and at least of the longitudinal wave component and the transverse wave component in the ultrasonic wave detected by propagating through the silicon sample. It is preferable to measure the shear wave component from the viewpoint of increasing the shear component and improving the resolution. When the angle is less than 5 °, the generation of the longitudinal wave component contained in the ultrasonic wave is mostly generated, and the generation efficiency of the transverse wave component is remarkably reduced. When the angle exceeds 60 °, the longitudinal wave ultrasonic wave and the transverse wave ultrasonic wave are generated. This is because the generation efficiency of both is significantly reduced.

The angle of the C-axis is more preferably in the range of 40 to 50 ° from the viewpoint of improving the generation efficiency of both longitudinal and transverse ultrasonic waves in a balanced manner. FIG. 14 shows a sample material in which two FZ samples cut out from the same sample were formed as ZnO with a C-axis inclined at an angle of 40 ° and 80 ° with respect to the sample surface as a vibrator. It is the result of performing ultrasonic measurement on each sample material at a resonance frequency of 400 MHz.

From the results of FIG. 14, when the C-axis angle is 40 °, the change in the elastic constant in the cryogenic region can be measured with high accuracy, whereas when the C-axis angle is 80 °. Since the generation of longitudinal and transverse ultrasonic waves is small, it can be seen that the change in the elastic constant in the extremely low temperature region cannot be measured.

  As a manufacturing method of the thin film vibrator 8 in which the C axis is inclined at a predetermined angle, for example, a method in which a silicon sample is arranged obliquely with respect to a ZnO target can be mentioned.

  The thickness of the thin film vibrator 8 is preferable in that ultrasonic waves that can be measured are in the range of 0.5 to 200 μm. This is because if the thickness exceeds 200 μm, the measurement accuracy tends to decrease, and if the thickness is less than 0.5 μm, high-frequency electrical measurement tends to be difficult.

  The resonance frequency of the thin film vibrator 8 is preferably in the range of 10 MHz to 10 GHz from the viewpoint that ultrasonic measurement can be applied. This is because, when the resonance frequency is higher than 10 GHz, electrical measurement at a high frequency tends to be difficult, and when the thickness is less than 10 MHz, measurement accuracy tends to decrease.

  It is preferable that the ultrasonic wave generation / detection means 4 uses an ultrasonic pulse having a pulse width of 10 μs or less because the sound speed of a silicon sample having a thickness of 10 mm or less can be measured. This is because when the pulse width is wider than 10 μs, it is difficult to distinguish between adjacent pulses. FIG. 10 shows an example when the pulse width is 0.2 μs in the upper diagram and the pulse width is 12 μs in the lower diagram.

  It is more preferable that the ultrasonic wave generation / detection means 4 includes means for performing zero detection by changing the oscillation frequency so that the phase difference caused by the change in sound speed due to temperature or magnetic field becomes constant.

  Moreover, it is preferable that the quantitative evaluation apparatus 1 of the present invention can simultaneously measure the phase difference of a plurality of silicon samples and a plurality of points of one silicon sample. FIG. 11 shows a case where gold (Au) / zinc oxide (ZnO) / gold (Au) is vapor-deposited directly on the wafer in order to simultaneously measure the phase at a plurality of points (four places in FIG. 11) of one silicon wafer. An example when a child is formed is shown.

  FIG. 4 shows a prototype of a non-doped CZ silicon ingot having a diameter of 6 inches and schematically showing a longitudinal section thereof. As is apparent from FIG. 4, it was recognized that an intrinsic point defect region (Pv region and Pi region) extending about 3 cm exists in the central portion.

Then, next, silicon samples (A) and (B) were cut into 4 mm × 4 mm × 7 mm from the Pv region and Pi region, which are intrinsic point defect regions, respectively, and installed in the quantitative evaluation apparatus shown in FIG. 1 and FIG. By the quantitative evaluation method of the present invention, the change in the elastic constant with respect to the cooling temperature when it was cooled to 30 to 20 mK was measured. The measurement results are shown in FIG. Note that the sound velocity v used for obtaining the elastic constant was calculated from the following equation using the phase difference φ _n of the ultrasonic pulse shown in FIG.
Formula: φ _n = 2π (2n-1) lf / v
Here, (2n-1) l is the propagation length of the nth echo, and f is the ultrasonic frequency.

  From the results shown in FIG. 5, in the sample (A) in the Pv region which has been considered to be rich in the frozen atomic vacancy region, the elastic constant is remarkably proportional to the reciprocal of the temperature in the cryogenic region from 20K to 10mK. It can be seen that the temperature decreases, in other words, softens at a low temperature. On the other hand, such a decrease in elastic constant was not observed in the Pi region sample (B) which was considered to be silicon-rich between the lattices.

  FIG. 15 shows an example of the result of measuring the change in elastic constant with respect to the temperature of a sample in which ZnO is formed as an oscillator on an FZ silicon crystal. In FIG. 15, a dilution refrigerator was used as the temperature control means, and measurement was performed up to an extremely low temperature of 20 mK. From the results shown in FIG. 15, the phenomenon of softening at a low temperature was recognized in the FZ silicon crystal as well as the above-described CZ silicon crystal.

In addition, the magnetic field dependence was also investigated using a B-doped FZ silicon single crystal and a B-free FZ silicon single crystal. FIG. 9 shows an example of a change result of the elastic constant when a magnetic field of 0 to 16 Tesla is applied. The upper diagram shows the case where B is not added, and the lower diagram shows the case where B is added. From the results of FIG. 9, low temperature softening in the B-added FZ silicon single crystal occurs when a magnetic field of about 4 Tesla or less is applied and disappears by applying a magnetic field larger than that, but low-temperature softening in the undoped FZ silicon single crystal. It was found that crystallization does not occur over the entire range of the magnetic field. This indicates that the bond between the charge state of the vacancies and the strain is the origin of softening. The atomic vacancies of the undoped FZ silicon single crystal are in a non-magnetic charge state in which four electrons are captured, and the B-doped FZ silicon single crystal is in a magnetically charged state in which three electrons are captured. The molecular orbitals of atomic vacancies are split into singlets and triplets, and the Jahn-Teller effect due to the combination of triplet electric quadrupoles and strains reduces C ₄₄ and (C ₁₁ -C ₁₂ ) / 2 softening. It is thought that it is awake. In the additive-free FZ silicon single crystal, antiferromagnetic quadrupole interaction exists between atomic vacancies, T _d symmetry around atomic vacancies is maintained even at a minimum temperature of 20 mK, and the triplet is degenerate. The electric quadrupole fluctuation appears to exist.

  From these results, low-temperature softening of the elastic constant due to an odd number (3 or 5) of trapped electrons has a magnetic field dependence, whereas an atom that has captured even (4) electrons. In the present invention, the kind of atomic vacancies can be determined based on the presence or absence of magnetic field dependence by utilizing the knowledge that softening of the elastic constant due to vacancies has no magnetic field dependence.

Next, an example of a method for quantitative evaluation of atomic vacancies existing in a silicon wafer according to the present invention will be described below.
In the quantitative evaluation method of the present invention, an external magnetic field is applied to a silicon sample obtained by cutting out a predetermined part from a silicon wafer as necessary, and the silicon sample is cooled in the temperature range while being cooled in a temperature range of 25 K or less. An ultrasonic pulse is generated by oscillating an ultrasonic pulse to a silicon sample having a physical property that can follow expansion and having a C-axis aligned in a predetermined direction, directly on the surface or via a gold thin film. The pulse propagates through the silicon sample, detects the change in sound velocity of the propagated ultrasonic pulse, calculates the elastic constant associated with the decrease in cooling temperature from this change in sound velocity, and calculates the amount of decrease in the calculated elastic constant in the silicon wafer. It is possible to quantitatively evaluate the type and concentration of atomic vacancies present in

  The above description is merely an example of the embodiment of the present invention, and various modifications can be made within the scope of the claims.

  The atomic vacancy concentration present in the silicon wafer was quantitatively evaluated using the quantitative evaluation apparatus for atomic vacancies present in the silicon wafer of the present invention, and will be described below.

(Example)
With a 6 inch diameter, 45 kg charge, total length 800mm undoped CZ ingot, the region present in the CZ ingot (void region, R-OSF region, _{P V} region, _{P i} region) and, using a Cu decoration method 6, after specifying the boundary line of each region, the samples (Y-1 and Y-6 to Y-10) are cut into 4 mm × 4 mm × 7 mm size at the six positions shown in FIG. After forming a vibrator made of ZnO having a C axis with a film thickness of 10 μm and an inclination of 40 ° with respect to the sample surface directly on both surfaces of each sample, each sample was quantified as shown in FIGS. The elastic constant change (ΔC ₄₄ / C ₄₄ ) with respect to the cooling temperature was measured when set to an evaluation apparatus and cooled to 30-20 mK by the quantitative evaluation method of the present invention. The measurement results are shown in FIG. Note that the vertical axis (ΔC ₄₄ / C ₄₄ ) in FIG. 7 is a relative value, and is shifted so that the values of the samples do not overlap.

(Comparative example)
Since the sample was prototyped in the same manner as in the above example except that a vibrator made of LiNbO ₃ was attached to the surface of the Y-7 sample via an adhesive, the same measurement as in the example was performed. It was. The measurement results are shown in FIG.

As is apparent from the results shown in FIG. 7, in the measurement in the examples, the samples Y-6 to Y8 in the Pv region, which has been considered to be rich in frozen atomic vacancy regions, are poles ranging from 10K to 20mK. while the elastic constants are significantly reduced in proportion to the inverse of the temperature in the low temperature region, the sample of other areas, including P _i region, change of elastic constant at the cryogenic region was observed.

  On the other hand, in the measurement in the comparative example, as is clear from the results shown in FIG. 13, the change in the elastic constant observed in the cryogenic region from 10K to 20mK in Example (Y-7) was not recognized. It was. The change in elastic constant can be confirmed at a temperature of about 4K in FIG. 13. This change is due to the occurrence of adhesion failure (adhesion peeling) in the cryogenic region. It is thought that accurate measurement was not possible.

An example in which the relationship between the elastic constant and temperature in a sample formed on the surface using AlN as a vibrator instead of ZnO is plotted is shown in the lower diagram of FIG. The upper diagram of FIG. 12 shows data obtained by plotting the relationship between the elastic constant and temperature of a sample formed on the surface of Zn-8 as a vibrator on the Y-8 sample used in FIG. 7 for comparison. From the results of FIG. 12, it can be seen that the same result as that obtained when ZnO was used as the vibrator was obtained when AlN was used as the vibrator.

Next, FIG. 8 shows the result of calculating the atomic vacancy concentration for the samples Y-1 and Y-6 to Y-10. The atomic vacancy concentration on the vertical axis in FIG. 8 is shown as a relative value when the sample Y-7 is 1.0, and the atomic vacancy concentration of the sample Y-7 is 2.46 × 10 ¹⁵ / cm ^3. Met. The decrease in elastic constant follows C = C ₀ (TT _c ) / (T-Θ). The difference Δ = T _c −Θ between the characteristic temperature Tc and Θ obtained in the experiment is proportional to the atomic vacancy concentration N. The absolute value of the atomic vacancy concentration N can be experimentally determined by the relational expression N = Δ · C ₀ / δ ² using Δ obtained in the experiment. Here, δ is the magnitude of energy change (deformation energy) of the electronic state of the atomic vacancies with respect to externally applied strain.

From the results of FIG. 8, the sample Y-6~Y-8 cut out from the P _V region, when compared with the sample Y-9 and Y-10 cut out from the P _i region, high atomic vacancy concentration of the former, the latter The atomic vacancy concentration is low. Further, the sample Y-1 cut out from the void region does not exist as atomic vacancies but exists as voids, so the atomic vacancy concentration is low. Furthermore, when compared between the sample Y-6~Y-8 cut out from the P _V region, and a sample Y-6 positioned on R-OSF region side, atomic air in the sample Y-8 positioned in the P _i region side it can also be seen that the direction of atomic vacancy concentration in the sample Y-7 than the hole concentration was cut from the center position of the P _V region is highest.

According to the present invention, an optimized thin film vibrator is formed on the surface of a silicon sample, and is manufactured by the Czochralski method (CZ method) or the float zone method (FZ method) used in the semiconductor industry. The type and concentration of isolated atomic vacancies in a silicon crystal wafer can be directly and quantitatively evaluated without performing an acceleration process such as increasing the concentration.
In particular, in the semiconductor industry, the demand for silicon wafers using perfect crystals that do not have secondary defects such as voids is rapidly increasing, but in the prior art, the concentration of atomic vacancies in the wafer is directly measured. Although there was a problem that it was difficult to observe and quantitatively evaluate and there was a case where the defect rate of the manufactured silicon device was large, by using the atomic vacancy quantitative evaluation apparatus of the present invention, Quantitative evaluation of the type and concentration of vacancies is possible, which can be said to have a significant impact on the semiconductor industry.

1 is a schematic view of an apparatus for quantitatively evaluating atomic vacancies according to the present invention. FIG. It is an enlarged view when the sample holder part 7 which set the silicon | silicone sample 5 which comprises the quantitative evaluation apparatus 1 of FIG. 1 is extracted. It is a flowchart for demonstrating the method to detect a phase difference using an ultrasonic pulse. An example of the longitudinal section of a non dope CZ silicon ingot is shown typically. It is a figure when the change of the elastic constant with respect to cooling temperature is measured when it cools to 30-20 mK by the quantitative evaluation method of this invention. A longitudinal section of a non-doped CZ silicon ingot used in Example, each region existing (void region, R-OSF region, P _V region, P _i region) of the boundary line of each region using a Cu decoration method Is shown in a specific state. When the samples (Y-1 and Y-6 to Y-10) were cooled to 30 to 20 mK at the six positions shown in FIG. 6, the change in elastic constant with respect to the cooling temperature (ΔC _{L [111]} / C _{L [111]} is a diagram when measuring. It is a figure which shows a result when calculating an atomic vacancy density | concentration about the sample Y-1 and Y-6 to Y-10 used in FIG. It is the figure which plotted an example of the change result of the elastic constant when applying a magnetic field using the BZ addition FZ silicon single crystal (upper figure) and the FZ silicon single crystal which doped B (lower figure). The example of the pulse signal given to the vibrator is shown, and the upper diagram shows the case where the pulse width is 0.2 μs, and the lower diagram shows the case where the pulse width is 12 μs. Schematic showing an example of forming a vibrator directly on a wafer by vapor-depositing gold (Au) / zinc oxide (ZnO) / gold (Au) in order to simultaneously measure the phase at four locations on one silicon wafer FIG. It is the figure which plotted the relation between the elastic constant and the temperature in the sample which formed the vibrator on the surface of two CZ silicon crystals cut out from the same sample, and the upper figure is the measurement result in the sample which formed ZnO on the surface using the vibrator The figure below shows the measurement results of a sample formed on the surface using AlN as a vibrator. It is the figure which plotted the measurement result of the comparative example. Two FZ silicon crystal samples cut out from the same sample were used as specimens with ZnO formed by tilting the C-axis at angles of 40 ° and 80 ° with respect to the sample surface as vibrators. It is the figure which plotted the measurement result when performing ultrasonic measurement to the sample material at the resonant frequency of 400 MHz. It is the figure which plotted the result of having measured the change of the elastic constant with respect to temperature about the sample which formed ZnO in the FZ silicon crystal as a vibrator.

Explanation of symbols

1 Quantitative evaluation device for atomic vacancies 2 Magnetic force generation means 3 Temperature control means (or dilution refrigerator)
4 Ultrasonic oscillation / detection means 5 Silicon sample 6 Mixing chamber 7 Sample holder 8 Thin film vibrator

Claims (14)

A magnetic force generating means for applying an external magnetic field to the silicon sample;
Temperature control means capable of cooling and controlling the silicon sample to a temperature range of 50K or less;
An ultrasonic oscillation / detection means for oscillating an ultrasonic pulse to the surface of a silicon sample, propagating the oscillated ultrasonic pulse through the silicon sample, and detecting a change in sound velocity of the propagated ultrasonic pulse;
A thin film vibrator having physical properties capable of following expansion of a silicon sample accompanying a temperature drop in the temperature range and having a C axis substantially aligned in a predetermined direction is directly formed on the surface of the silicon sample. A quantitative evaluation system for atomic vacancies in silicon wafers.   The ultrasonic wave oscillating / detecting means calculates a phase difference between a reference wave pulse signal obtained by directly measuring an oscillated ultrasonic pulse and a sample passing wave pulse signal measured after the ultrasonic pulse is propagated through the silicon sample. 2. The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to claim 1, further comprising means for detecting.   3. The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to claim 1, wherein the thin film vibrator is made of zinc oxide (ZnO) or aluminum nitride (AlN).   4. A quantitative evaluation apparatus for atomic vacancies existing in a silicon wafer according to claim 1, wherein the thin film vibrator is formed on the silicon wafer by physical vapor deposition.   The quantitative evaluation apparatus for atomic vacancies existing in a silicon wafer according to any one of claims 1 to 4, further comprising a gold thin film between the thin film vibrator and the silicon crystal.   The thin film vibrator has a C axis inclined at an angle of 5 to 60 ° with respect to the surface of the silicon sample, and at least the transverse wave component of the longitudinal wave component and the transverse wave component in the ultrasonic wave detected by propagating through the silicon sample. The apparatus for quantitative evaluation of atomic vacancies present in a silicon wafer according to any one of claims 1 to 5, wherein components are measured.   The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to any one of claims 1 to 6, wherein the thickness of the thin film vibrator is in a range of 0.5 to 200 µm.   The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to any one of claims 1 to 7, wherein a resonance frequency of the thin film vibrator is in a range of 10 MHz to 10 GHz.   The apparatus for quantitative evaluation of atomic vacancies present in a silicon wafer according to any one of claims 1 to 8, wherein the magnetic force generating means is in a range of 0 to 20 Tesla.   The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to any one of claims 1 to 9, wherein the temperature control means has a dilution refrigerator that can be cooled to a cryogenic temperature of up to 5 mK.   The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to claim 1, wherein the ultrasonic wave generation / detection means uses an ultrasonic pulse having a pulse width of 10 μs or less.   12. The ultrasonic wave generation / detection means includes means for performing zero detection by changing an oscillation frequency so that a phase difference caused by a change in sound speed due to a temperature or a magnetic field becomes constant. For quantitative evaluation of atomic vacancies in silicon wafers.   The apparatus for quantitative evaluation of atomic vacancies existing in a silicon wafer according to any one of claims 1 to 12, wherein a phase difference can be measured simultaneously with a plurality of silicon samples and a plurality of points of one silicon sample as measurement targets.   A silicon sample cut out from a silicon wafer, with an external magnetic field applied as necessary, while cooling in a temperature range of 25K or less, following the expansion of the silicon sample due to the temperature drop in the temperature range An ultrasonic pulse is oscillated to a silicon sample having a physical property that can be obtained and a C-axis aligned in a predetermined direction directly on the surface, and the oscillated ultrasonic pulse is propagated through the silicon sample. The change in the velocity of the ultrasonic pulse is detected, and from this change in the velocity of sound, the amount of decrease in the elastic constant accompanying the decrease in the cooling temperature is calculated. From the amount of decrease in the calculated elastic constant, the type of atomic vacancies present in the silicon wafer Quantitative evaluation method of atomic vacancies existing in silicon wafer, characterized in that quantitative evaluation of concentration is performed.