JP4124412B2 - Wafer metal contamination prevention device and metal contamination measurement method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an apparatus for preventing metal contamination of a semiconductor wafer and a method for measuring the amount of metal contamination.
[0002]
[Prior art]
The quality of semiconductor wafers deteriorates when they are contaminated with metal. Therefore, in each process of manufacturing a semiconductor wafer, a clean environment is secured in order to avoid metal contamination of the semiconductor wafer.
[0003]
However, since abrasives, wires, tools, etc. are used when processing semiconductor wafers, the metal that constitutes these abrasives may enter the inside of the semiconductor wafer, that is, in the bulk, and the semiconductor wafer may be contaminated by the metal. . In particular, since copper (Cu) has a high diffusion rate in silicon, it easily enters the bulk of a silicon wafer.
[0004]
Therefore, a sampling inspection is performed at a semiconductor wafer manufacturing site where a semiconductor wafer is extracted and the amount of metal contamination of the semiconductor wafer is measured to quantitatively determine in which process the metal contamination occurs.
[0005]
One method for measuring the amount of metal contamination inside (bulk) a semiconductor wafer is a hot plate method. Although there is “SIMS” as another method for quantitatively measuring the amount of metal contamination inside (bulk) the semiconductor wafer, it is not suitable for practical use in terms of insufficient sensitivity.
[0006]
Known documents relating to the hot plate method include, for example, JP-A-9-64133, JP-A-9-64052, JP-A-9-260325, and JP-A-2000-290100.
[0007]
[Problems to be solved by the invention]
However, when the hot plate method is used, there is a possibility that the semiconductor wafer to be measured may be contaminated by the copper contained in the hot plate itself, thereby reducing the amount of copper existing in the bulk of the semiconductor wafer. It became clear by the present inventors that it cannot be measured accurately.
[0008]
First, a procedure for quantitatively measuring the amount of contamination of a wafer contaminated with copper by the hot plate method will be described with reference to FIG.
[0009]
As shown in FIG. 2A, a wafer 1 contaminated with an unknown amount of Cu (hereinafter referred to as Cu-contaminated wafer 1) is placed on a hot plate 3 via a clean wafer 2 as shown in FIG. To place. In this case, the Cu-contaminated wafer 1 is placed so that the front surface 1a is upward and the back surface 1b is downward. Here, the surface 1 a is a surface on the mirror surface side of the wafer 1. A “clean wafer” is a wafer whose metal contamination is below the measurement limit and is prepared in advance for measurement. The clean wafer 2 is interposed as the “underlay” of the Cu-contaminated wafer 1 because the Cu-contaminated wafer 1 itself prevents the copper (Cu) from entering the Cu-contaminated wafer 1 from the hot plate 3 by diffusion. This is for measuring the amount of Cu contamination contained.
[0010]
In this state, the hot plate 3 as a heating device is operated to heat-treat (anneal) the Cu-contaminated wafer 1 to be measured at 300 ° C. to 500 ° C. for 2 hours in a clean room of class 1000 or higher.
[0011]
The reason why the heat treatment is performed in a clean room of class 1000 or higher is to prevent the Cu-contaminated wafer 1 from being contaminated by the atmosphere. In particular, the Cu-contaminated wafer 1 itself is prevented from being contaminated from the atmosphere of copper. This is because only the amount of Cu contamination contained is measured.
[0012]
When the heat treatment is started, copper existing in the bulk of the Cu-contaminated wafer 1 starts to move and diffuses due to heat, and is deposited on the surface 1 a of the Cu-contaminated wafer 1. For this reason, the copper in the bulk is concentrated on the surface 1a after the heat treatment.
[0013]
Next, as shown in FIG. 2C, the Cu-contaminated wafer 1 immediately after the heat treatment is placed on the aluminum (Al) plate 4 and rapidly cooled. For example, the Cu-contaminated wafer 1 heated to 500 ° C. is rapidly cooled to 50 ° C. in about 20 seconds.
[0014]
Next, as shown in FIG. 2D, the surface analysis of the Cu-contaminated wafer 1 is performed using a total reflection fluorescent X-ray analysis method (TXRF) which is one method of the surface analysis method. That is, by using the total reflection fluorescent X-ray analyzer TXRF, X-rays are incident on the surface 1a of the Cu-contaminated wafer 1 at a shallow angle and totally reflected on the surface 1a, whereby the amount of copper concentrated on the surface 1a (copper) ).
[0015]
An example of the measurement result is shown in FIG. FIG. 3 shows the measurement results when the Cu-contaminated wafer 1 is heat-treated at 300 ° C. for 4 hours.
[0016]
FIG. 3A shows the distribution of the amount of Cu contamination (copper concentration) on the surface 1a of the Cu-contaminated wafer 1 after the heat treatment.
[0017]
FIG. 3B shows the Cu contamination amount distribution after the heat treatment of the clean wafer surface 2 a used as an underlay for the Cu contaminated wafer 1.
[0018]
In these FIGS. 3 (a) and 3 (b), the size of the circles in each part in the wafer surface corresponds to the amount of Cu contamination in each part in the surface. 3A and 3B, the position of each part in the wafer plane is specified by the X and Y coordinate positions (X, Y) with the center of the wafer as the origin (0, 0) of the XY coordinate system. The
[0019]
As shown in FIG. 3A, after the heat treatment, the Cu contaminated wafer 1 is almost uniformly 10 in the plane. ¹³ atoms / cm ² An amount of Cu contamination on the order of was detected. These detected coppers existed in the bulk of the Cu-contaminated wafer 1 before the heat treatment, and were diffused by heat and deposited on the surface 1a. When the surface 1a was measured using TXRF before heat treatment, copper was detected (1 × 10 ¹⁰ atoms / cm ² )
[0020]
Also, the clean wafer 2 used as an underlay also had copper below the detection limit before the heat treatment, but after the heat treatment, Cu contamination was confirmed on the surface 2a as shown in FIG. 3 (c). The Cu contamination of the clean wafer 2 is due to copper diffused from the hot plate 3.
[0021]
A confirmation experiment was performed to confirm whether the Cu contamination of the clean wafer 2 was caused by copper diffused from the hot plate 1 or the copper diffused from the Cu contaminated wafer 1.
[0022]
FIG. 4A shows a mounting example of a wafer used in this confirmation experiment.
[0023]
As shown in FIG. 4A, in the confirmation experiment, an underlying clean wafer 2 (N-type silicon wafer) is placed on the hot plate 3, and the Cu contamination shown in FIG. Instead of the wafer 1, a clean wafer 5 (N-type silicon wafer) free from Cu contamination is placed.
[0024]
In the confirmation experiment, heat treatment was performed at 300 ° C. for 4 hours. The measurement results are shown in FIGS. 4 (b) and (c). FIG. 4B shows the Cu contamination amount distribution on the clean wafer surfaces 2a and 5a before the heat treatment, and the Cu contamination was below the measurement limit. 4C shows the Cu contamination amount distribution on the surface 5a of the upper clean wafer 5 after the heat treatment, and FIG. 4D shows the Cu contamination amount distribution after the heat treatment on the surface 2a of the underlying clean wafer 2. FIG. Yes.
[0025]
As shown in FIGS. 4B, 4C, and 4D, after the heat treatment, only the surface 2a of the underlying clean wafer 2 is contaminated with Cu, so that the underlay in FIG. It was confirmed that the contamination of the clean wafer 2 was not caused by the copper diffused from the Cu-contaminated wafer 1 but by the copper diffused from the hot plate 3.
[0026]
In FIG. 2B, it has also been confirmed by a confirmation experiment that the copper detected on the surface 1 a of the Cu-contaminated wafer 1 to be measured is not copper diffused from the hot plate 3.
[0027]
FIG. 5A shows a mounting example of a wafer used in this confirmation experiment.
[0028]
As shown in FIG. 5A, in the confirmation experiment, clean wafers 6 and 2 (N ⁻ Two types of silicon wafers) are stacked, and Cu contaminated wafer 1 (P ⁺⁺ Mold type silicon wafer).
[0029]
In the confirmation experiment, heat treatment was performed at 500 ° C. for 4 hours. The measurement results are shown in FIGS. 5 (b), (c) and (d).
[0030]
FIG. 5B shows the distribution of Cu contamination on the surface 1a of the Cu-contaminated wafer 1 after the heat treatment, and FIG. 5C shows the distribution of Cu contamination after the heat treatment on the surface 2a of the upper clean wafer 2 as an underlay. FIG. 5D shows the Cu contamination amount distribution after the heat treatment of the surface 6a of the lower clean wafer 6 serving as an underlay.
[0031]
As shown in FIGS. 5B, 5C, and 5D, although a large amount of Cu contamination is confirmed on the surface 1a of the Cu-contaminated wafer 1 to be measured, the underlying clean wafers 6 and 2 Since the amount of copper detected on the surfaces 6a and 2a was very small, it was confirmed that the copper detected on the surface 1a of the Cu-contaminated wafer 1 to be measured was not copper diffused from the hot plate 3.
[0032]
From the above confirmation experiment, the copper contained in the hot plate 3 enters the bulk of the clean wafer 2 underlying the Cu-contaminated wafer 1 to be measured and contaminates the clean wafer 2 during the heat treatment. For this reason, there is a concern that the Cu-contaminated wafer 1 to be measured is contaminated through the underlying clean wafer 2.
[0033]
For this reason, the methods shown in Reference Examples 1 and 2 below are carried out.
[0034]
(Reference Example 1)
In FIG. 2B, the underlying clean wafer 2 is exchanged together with the Cu-contaminated wafer 1 after completion of measurement every time heat treatment is performed, that is, every time the amount of Cu contamination of the Cu-contaminated wafer 1 is measured.
[0035]
(Reference Example 2)
As shown in FIG. 6, two clean wafers 6 and 2 are placed on the hot plate 3, and the Cu-contaminated wafer 1 to be measured is placed thereon.
[0036]
For the clean wafers 2 and 6 that are the underlay, each time the heat treatment is performed, that is, every time the Cu contamination amount of the Cu-contaminated wafer 1 is measured, the Cu contamination amount is measured. After ozone cleaning, reuse it as an underlay. If Cu contamination is confirmed in the clean wafers 2 and 6, the clean wafers 2 and 6 are exchanged together with the Cu contaminated wafer 1 after the measurement is completed. However, it is desirable to replace the upper clean wafer 2 together with the Cu-contaminated wafer 1 for each measurement.
[0037]
However, in any of the reference examples 1 and 2, the underlying clean wafer needs to be replaced with a new one in a complicated manner, resulting in a problem that the economy is not good.
[0038]
In the reference example 2, as compared with the reference example 1, it is possible to more effectively prevent the contaminated Cu-contaminated wafer 1 from being contaminated through the underlying clean wafer.
[0039]
However, in both Reference Example 1 and Reference Example 2, contamination of the Cu-contaminated wafer 1 due to copper diffused from the hot plate 3 cannot be completely stopped with a clean wafer alone. In particular, if the clean wafer is neglected and the clean wafer is left untreated, the Cu-contaminated wafer 1 to be measured may be contaminated by copper diffused from the hot plate 3.
[0040]
The present invention has been made in view of such a situation, and when measuring the amount of metal contamination on the wafer surface using the hot plate method, it is more complete that the wafer to be measured is contaminated by the metal diffused from the hot plate. The problem to be solved is to make it possible to prevent it. Further, when a clean wafer is used for the underlay, it is an object of the present invention to lengthen the replacement cycle and improve the economy.
[0044]
[Means for solving the problems and effects]
The first invention is
In measuring the amount of metal contamination on the surface of the silicon wafer heated by the hot plate using the hot plate method, in the metal contamination prevention apparatus for the wafer for preventing metal contamination from the hot plate to the wafer,
A silicon wafer on which a silicon oxide film is formed is interposed between the hot plate and the silicon wafer to be measured.
It is characterized by.
[0045]
According to the first invention, as shown in FIG. 1B, the silicon wafer 10 on which the silicon oxide film 10c is formed is interposed between the hot plate 3 and the silicon wafer 1 to be measured. The silicon oxide film 10c has a contamination metal diffusion rate lower than the contamination metal diffusion rate in the silicon wafer 1 to be measured. For this reason, when measuring the amount of metal contamination on the surface 1a of the silicon wafer 1 using the hot plate method, it is possible to more completely prevent the silicon wafer 1 to be measured from being contaminated by the metal diffused from the hot plate 3. it can.
[0046]
The second invention is
In measuring the amount of metal contamination on the surface of the silicon wafer heated by the hot plate using the hot plate method, in the metal contamination prevention apparatus for the wafer for preventing metal contamination from the hot plate to the wafer,
A silicon wafer having a silicon oxide film formed thereon, a clean wafer, and a silicon wafer to be measured were sequentially stacked on the hot plate.
It is characterized by.
[0047]
According to the second invention, as shown in FIG. 1B, the silicon wafer 10 on which the silicon oxide film 10c is formed is interposed between the hot plate 3 and the silicon wafer 1 to be measured. The silicon oxide film 10c has a contamination metal diffusion rate lower than the contamination metal diffusion rate in the silicon wafer 1 to be measured, and effectively prevents the contamination metal on the hot plate 3 from entering the silicon wafer 1 to be measured. Can do. Furthermore, since the silicon wafer 2 is interposed, it is possible to more effectively prevent the contaminated metal of the hot plate 3 from entering the silicon wafer 1 to be measured.
[0048]
The third invention is the first invention, the second invention,
The contaminant metal measured on the wafer surface is copper
It is characterized by.
[0052]
The fourth invention is
In the method for measuring the amount of metal contamination of a wafer for measuring the amount of metal contamination on the surface of a silicon wafer heated by a hot plate using the hot plate method,
A step of interposing a silicon wafer formed with a silicon oxide film between the hot plate and the silicon wafer to be measured;
A process of measuring the amount of metal contamination on the surface of the silicon wafer to be measured using the hot plate method;
It is characterized by including.
[0053]
In the fourth invention, the apparatus invention of the first invention is replaced with the method invention.
[0054]
The fifth invention
In the method for measuring the amount of metal contamination of a wafer for measuring the amount of metal contamination on the surface of a silicon wafer heated by a hot plate using the hot plate method,
A step of sequentially stacking a silicon wafer on which a silicon oxide film is formed, a clean wafer, and a silicon wafer to be measured on the hot plate;
A process of measuring the amount of metal contamination on the surface of the silicon wafer to be measured using the hot plate method;
It is characterized by including.
[0055]
In the fifth invention, the apparatus invention of the second invention is replaced with a method invention.
[0056]
The sixth invention
In the method for measuring the amount of metal contamination of a wafer for measuring the amount of metal contamination on the surface of a silicon wafer heated by a hot plate using the hot plate method,
A step of sequentially stacking a silicon wafer on which a silicon oxide film is formed, a clean wafer, and a silicon wafer to be measured on the hot plate;
Measuring the amount of metal contamination on the surface of the silicon wafer to be measured using the hot plate method;
A step of replacing the clean wafer together with the silicon wafer to be measured each time the amount of metal contamination of the clean wafer reaches a limit value;
It is characterized by including.
[0057]
According to the sixth invention, as shown in FIG. 1B, every time the amount of metal contamination of the clean wafer 2 reaches a limit value, the clean wafer 2 is exchanged together with the silicon wafer 1 to be measured. Here, the diffusion of metal from the hot plate 3 to the clean wafer 2 is delayed by the oxide film-formed wafer 10. For this reason, compared with the structure of FIG. 6, the exchange cycle of the clean wafer 2 can be lengthened, and economical efficiency can be improved.
[0058]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings. In the present embodiment, a silicon wafer is assumed as the semiconductor wafer, and copper is assumed as the contaminating metal. The embodiments described below can be applied to semiconductor wafers other than silicon, and can also be applied to contaminated metals other than copper.
[0059]
(Example 1)
FIG. 1 is a diagram for explaining Example 1, and shows a procedure for quantitatively measuring the amount of contamination of a wafer contaminated with copper by a hot plate method.
[0060]
As shown in FIG. 1A, a wafer 1 contaminated with an unknown amount of Cu, that is, a Cu-contaminated wafer 1 is prepared as a measurement target. Further, as shown in FIG. 1B, a clean wafer 2 and an oxide film-formed wafer 10 are prepared.
[0061]
The oxide film-formed wafer 10 is a silicon wafer in which a silicon oxide film (SiO2) 10c is formed with a thickness of, for example, 5000 mm on the entire surface, that is, on both surfaces 10a and 10b. The silicon oxide film 10c has a copper diffusion coefficient smaller than that of the silicon wafer. In general, the diffusion coefficient of copper is 10 in silicon. ^-6 -10 ^-5 cm ² / sec (temperature condition of 300 ° C. to 500 ° C.) and 10 in silicon oxide (SiO 2). ^-15 -10 ^-11 cm ² / sec (temperature condition of 300 ° C. to 500 ° C.).
[0062]
Furthermore, the interface between the silicon wafer and the silicon oxide film has an effect of gettering copper, and the diffusion of copper can be stopped at the interface.
[0063]
Therefore, if the oxide film-formed wafer 10 is used instead of the clean wafer 6 on the lower side of FIG. 6, the diffusion of copper can be delayed.
[0064]
Therefore, as shown in FIG. 1B, the oxide film-formed wafer 10 and the clean wafer 2 are sequentially stacked on the hot plate 3, and the Cu-contaminated wafer 1 to be measured is further mounted thereon.
[0065]
In this case, the Cu-contaminated wafer 1 is mounted such that the front surface 1a which is the mirror surface side is upward and the back surface 1b is downward. Similarly, the oxide film-formed wafer 10 which is the lower underlay is also placed so that the front surface 10a on the mirror surface side is upward and the back surface 10b is downward. Similarly, the clean wafer 2 serving as the upper underlay is also placed so that the front surface 2a on the mirror surface side is upward and the back surface 2b is downward.
[0066]
In this state, the hot plate 3 as a heating device is operated to heat-treat (anneal) the Cu-contaminated wafer 1 to be measured at 300 ° C. to 500 ° C. for 2 hours in a clean room of class 1000 or higher.
[0067]
When the heat treatment is started, copper existing in the bulk of the Cu-contaminated wafer 1 starts to move and diffuses due to heat, and is deposited on the surface 1 a of the Cu-contaminated wafer 1. For this reason, the copper in the bulk is concentrated on the surface 1a after the heat treatment.
[0068]
Next, as shown in FIG. 1C, the Cu-contaminated wafer 1 immediately after the heat treatment is placed on the aluminum (Al) plate 4 and rapidly cooled. For example, the Cu-contaminated wafer 1 heated to 500 ° C. is rapidly cooled to 50 ° C. in about 20 seconds.
[0069]
Next, as shown in FIG. 1 (d), the surface analysis of the Cu-contaminated wafer 1 is performed using a total reflection fluorescent X-ray analysis method (TXRF) which is one method of the surface analysis method. That is, by using the total reflection fluorescent X-ray analyzer TXRF, X-rays are incident on the surface 1a of the Cu-contaminated wafer 1 at a shallow angle and totally reflected on the surface 1a, whereby the amount of copper concentrated on the surface 1a (copper) ).
[0070]
As will be described later, in the oxide film-formed wafer 10, even after 100 hours of heat treatment, copper contamination could not be confirmed on the surface 10a.
[0071]
Therefore, if the heat treatment time for one time is 2 hours, even if 50 Cu-contaminated wafers 1 are continuously measured, that is, even if heat treatment is continuously performed for 100 hours, Cu due to copper diffused from the hot plate 3 is used. Contamination of the contaminated wafer 1 can be prevented. For this reason, the amount of Cu contamination of the Cu-contaminated wafer 1 can be accurately measured.
[0072]
The exchange of the oxide film-formed wafer 10 is performed every time 50 Cu-contaminated wafers 1 are measured, that is, after 100 hours of heat treatment.
[0073]
For the clean wafer 2 that is the underlay, whenever the heat treatment is performed, that is, each time the Cu contamination amount of the Cu-contaminated wafer 1 is measured, the Cu contamination amount on the surface 2a is measured, and the Cu contamination amount is less than the measurement limit. For example, after ozone cleaning, reuse it as an underlay. If Cu contamination is confirmed on the clean wafer 2, it is replaced with a new clean wafer 2 together with the Cu-contaminated wafer 1 after the measurement is completed.
[0074]
Therefore, according to the first embodiment, the replacement cycle of the underlying clean wafer 2 and oxide film-formed wafer 10 can be lengthened as compared with the reference example 2 in FIG. improves.
[0075]
Next, the Cu diffusion delay effect of the oxide film-formed wafer 10 will be described.
FIG. 7 shows an experimental procedure for confirming the Cu diffusion delay effect of the oxide film-formed wafer 10.
[0076]
As shown in FIG. 7A, the Cu contamination amount is 7.45 × 10 using the pipette 20 on the back surface 10 b of the oxide film-formed wafer 10. ¹³ atoms / cm ² It is contaminated by the copper standard solution 30.
[0077]
Therefore, as shown in FIG. 7B, on the hot plate 3, an oxide film-formed wafer 11, a clean wafer 12, an oxide film-formed wafer 10 contaminated with copper (hereinafter referred to as a Cu-contaminated oxide film-formed wafer 10), and a measurement sample. Wafers 1 are sequentially stacked and placed.
[0078]
The measurement wafer 1 is used here because the Cu contamination on the surface 10a of the Cu-contaminated oxide film-formed wafer 10 cannot be measured by the TXRF measurement method, and by measuring the surface 1a of the measurement wafer 1 This is because the Cu contamination on the surface 10a of the Cu-contaminated oxide film-formed wafer 10 is indirectly measured. As the measurement wafer 1, a clean wafer whose Cu contamination amount is below the measurement limit is used.
[0079]
Therefore, the Cu-contaminated oxide film-formed wafer 10 is placed so that the front surface 10a is on the upper side and the contaminated back surface 10b is on the lower side. With the surface 1a facing downward, the measurement surface 1a is placed in contact with the surface 10a of the Cu-contaminated oxide film-formed wafer 10.
[0080]
The clean wafer 12 is placed so that the front surface 12a is upward and the back surface 12b is downward. Similarly, the oxide film-formed wafer 11 is also placed so that the front surface 11a is upward and the back surface 11b is downward.
[0081]
The reason why the clean wafer 12 and the oxide film-formed wafer 11 are provided under the Cu-contaminated oxide film-formed wafer 10 is to eliminate the influence of contamination by copper diffused from the hot plate 3.
[0082]
In this state, the hot plate 3 serving as a heating device is operated to heat-treat (anneal) the Cu-contaminated oxide film-formed wafer 10 and the measuring wafer 1 in a clean room of class 1000 or higher at 500 ° C. for 2 hours.
[0083]
Next, the measuring wafer 1 immediately after the heat treatment is placed on an aluminum (Al) plate and rapidly cooled. The measuring wafer 1 heated to 500 ° C. is rapidly cooled to 50 ° C. in about 20 seconds.
[0084]
Next, as shown in FIG. 7C, the surface analysis of the measurement wafer 1 is performed using a total reflection X-ray fluorescence analysis (TXRF). That is, the amount of Cu contamination on the surface 10a of the Cu-contaminated oxide film-formed wafer 10 is indirectly measured.
[0085]
FIG. 8 shows the measurement results of the surface 10a for each sample in which the type of dopant and the thickness of the oxide film of the Cu-contaminated oxide film-formed wafer 10 are changed.
[0086]
As shown in FIG. 8A, a sample having a P-type oxide film thickness of 5000 mm, a sample having an N-type oxide film thickness of 5000 mm, a sample having a P-type oxide film thickness of 1000 mm, and an N-type oxide film. About the thing which contaminated the back surface 10b of the sample whose film thickness is 1000cm with a copper standard solution, the amount of Cu contamination of those surface 10a was measured.
[0087]
The measurement results for each sample are shown in FIG. The X and Y coordinate positions (X, Y) in FIG. 8B correspond to the X and Y coordinate positions (X, Y) of the wafer surface shown in FIGS. 3A and 3B. In FIG. 8B, “ND” means that the amount of Cu contamination is below the measurement limit.
[0088]
As shown in FIGS. 8A and 8B, in either case of P-type or N-type, or in any case where the thickness is 1000 mm or 5000 mm, the heat treatment for about 2 hours is performed in the oxide film-formed wafer 10. It was confirmed that the copper diffusion was slow, not reaching the surface 10a.
[0089]
In the experiment, heat treatment was further continued for 100 hours. Even in this case, copper contamination could not be confirmed on the surface 10a of the oxide film-formed wafer 10.
[0090]
An experiment was also conducted on the difference in the Cu diffusion rate in the Cu-contaminated wafer 1 due to the difference between the P type and the N type.
FIG. 9 shows the procedure of an experiment for examining the difference in the Cu diffusion rate in the Cu-contaminated wafer 1 due to the difference between the P-type and the N-type.
[0091]
As shown in FIG. 9A, the back surface 1b of the P-type and N-type Cu-contaminated wafer 1 is used with a pipette 20, and the amount of Cu-contamination is 7.45 × 10. ¹³ atoms / cm ² It is contaminated by the copper standard solution 30.
[0092]
Therefore, as shown in FIG. 9B, the oxide film-formed wafer 11, the clean wafer 12, and the Cu-contaminated wafer 1 to be measured are sequentially stacked on the hot plate 3.
[0093]
In this state, the hot plate 3 as a heating device is operated to heat-treat (anneal) the Cu-contaminated wafer 1 at 500 ° C. for 2 hours in a clean room of class 1000 or higher.
[0094]
Next, the Cu-contaminated wafer 1 immediately after the heat treatment is placed on an aluminum (Al) plate and rapidly cooled. The measuring wafer 1 heated to 500 ° C. is rapidly cooled to 50 ° C. in about 20 seconds.
[0095]
Next, as shown in FIG. 9 (c), the surface of the Cu-contaminated wafer 1 is analyzed using total reflection X-ray fluorescence analysis (TXRF) to measure the amount of Cu contamination on the surface 1 a of the Cu-contaminated wafer 1.
[0096]
FIG. 10 shows the measurement results of the surface 1a for each sample in which the dopant type of the Cu-contaminated wafer 1 is changed.
[0097]
As shown in FIG. 10A, the amount of Cu contamination on the front surface 1a of the P-type sample and the N-type sample of which the back surface 1b was contaminated with the copper standard solution was measured.
[0098]
The measurement results for each sample are shown in FIG.
[0099]
As shown in FIG. 10, it was found that the diffusion of copper is delayed more in the N type than in the P type.
[0100]
In the first embodiment described above, the clean wafer 2 is used for the underlay, but this may be omitted.
[0101]
(Example 2)
That is, in Example 2, the clean wafer 2 is omitted in FIG. 1B, the oxide film-formed wafer 10 is placed on the hot plate 3, and the Cu-contaminated wafer 1 to be measured is placed thereon. .
[0102]
In this case, the Cu-contaminated wafer 1 is mounted such that the front surface 1a which is the mirror surface side is upward and the back surface 1b is downward. Similarly, the underlying oxide film-formed wafer 10 is also placed with the front surface 10a on the mirror surface side upward and the back surface 10b downward.
[0103]
In this state, the hot plate 3 as a heating device is operated to heat-treat (anneal) the Cu-contaminated wafer 1 to be measured at 300 ° C. to 500 ° C. for 2 hours in a clean room of class 1000 or higher.
[0104]
When the heat treatment is started, copper existing in the bulk of the Cu-contaminated wafer 1 starts to move and diffuses due to heat, and is deposited on the surface 1 a of the Cu-contaminated wafer 1. For this reason, the copper in the bulk is concentrated on the surface 1a after the heat treatment.
[0105]
Next, as shown in FIG. 1C, the Cu-contaminated wafer 1 immediately after the heat treatment is placed on the aluminum (Al) plate 4 and rapidly cooled. For example, the Cu-contaminated wafer 1 heated to 500 ° C. is rapidly cooled to 50 ° C. in about 20 seconds.
[0106]
Next, as shown in FIG. 1 (d), the total amount of copper concentrated on the surface 1a by performing surface analysis of the Cu-contaminated wafer 1 using total reflection X-ray fluorescence analysis (TXRF). Measure.
[0107]
According to the second embodiment, as in the first embodiment, even if 50 Cu-contaminated wafers 1 are continuously measured, that is, even if the heat treatment is continuously performed for 100 hours, they diffuse from the hot plate 3. Contamination of the Cu-contaminated wafer 1 with copper can be prevented. For this reason, the amount of Cu contamination of the Cu-contaminated wafer 1 can be accurately measured.
[0108]
The exchange of the oxide film-formed wafer 10 is performed every time 50 Cu-contaminated wafers 1 are measured, that is, after 100 hours of heat treatment.
[0109]
For this reason, according to the second embodiment, the replacement cycle of the underlying oxide film-formed wafer 10 can be lengthened as compared with the reference example 1 in FIG. To do.
[0110]
Various modifications can be made to the first and second embodiments described above.
[0111]
In the embodiment, the silicon oxide film 10c is formed on both surfaces 10a and 10b of the oxide film-formed wafer 10, but a silicon oxide film may be formed only on one of the surfaces.
[0112]
In the embodiment, the oxide film-formed wafer 10 is interposed between the hot plate 3 and the Cu-contaminated wafer 1 to be measured. However, the wafer 10 is not necessarily formed with an oxide film, and at least silicon. It suffices if an oxide film layer is interposed. For example, instead of the oxide film-formed wafer 10, only a silicon oxide film may be interposed.
[0113]
In the embodiment, the diffusion of copper is delayed by the silicon oxide film. However, if the diffusion of copper can be delayed more than that in the silicon, this is used for the hot plate 3 and the Cu-contaminated wafer 1 to be measured. It is also possible to intervene between the two. For example, a silicon carbide (SiC) layer may be interposed between the hot plate 3 and the Cu-contaminated wafer 1 to be measured, and a nitride is interposed between the hot plate 3 and the Cu-contaminated wafer 1 to be measured. (Si ₃ N ₄ ) Layer may be interposed.
[Brief description of the drawings]
FIGS. 1A, 1B, 1C, and 1D are diagrams for explaining the present embodiment and for explaining a procedure for measuring a copper contamination amount. FIG.
FIGS. 2A, 2B, 2C, and 2D are diagrams for explaining a reference example and for explaining a procedure of a process for measuring the amount of copper contamination.
FIGS. 3A, 3B, and 3C are diagrams showing measurement results of the contamination amount of the wafer used in FIG. 2B.
4 (a), (b), (c), and (b) are diagrams illustrating an experiment for confirming a copper contamination path in the configuration shown in FIG. 2 (b).
5 (a), (b), (c), and (b) are diagrams illustrating an experiment for confirming a copper contamination path in the configuration of FIG. 2 (b).
FIG. 6 is a diagram illustrating a mounting example of a wafer of a reference example.
FIGS. 7A, 7B, and 7C are diagrams used for explaining a delay effect of copper diffusion in an oxide film-formed wafer. FIG.
8A and 8B are diagrams showing the results of measuring the amount of copper contamination on the surface of the oxide film-formed wafer in the configuration of FIG. 7B.
FIGS. 9A, 9B, and 9C are diagrams illustrating an experiment for investigating a difference in diffusion rate of copper depending on a difference between P-type and N-type.
10 (a) and 10 (b) are diagrams showing the results of measuring the amount of copper contamination on the surface of P-type and N-type wafers in the configuration of FIG. 9 (b).
[Explanation of symbols]
1 Cu contaminated wafer
2 Clean wafer
3 Hot plate
10 Oxide film forming wafer

Claims (7)

In measuring the amount of metal contamination on the surface of the silicon wafer heated by the hot plate using the hot plate method, in the metal contamination prevention apparatus for the wafer for preventing metal contamination from the hot plate to the wafer,
An apparatus for preventing metal contamination of a wafer, wherein a silicon wafer on which a silicon oxide film is formed is interposed between the hot plate and a silicon wafer to be measured. In measuring the amount of metal contamination on the surface of the silicon wafer heated by the hot plate using the hot plate method, in the metal contamination prevention apparatus for the wafer for preventing metal contamination from the hot plate to the wafer,
An apparatus for preventing metal contamination of a wafer, wherein a silicon wafer on which a silicon oxide film is formed, a clean wafer, and a silicon wafer to be measured are sequentially stacked on the hot plate. The contamination metal measuring apparatus for a wafer according to claim 1, wherein the contamination metal measured on the wafer surface is copper. In the method for measuring the amount of metal contamination of a wafer for measuring the amount of metal contamination on the surface of a silicon wafer heated by a hot plate using the hot plate method,
A step of interposing a silicon wafer formed with a silicon oxide film between the hot plate and the silicon wafer to be measured;
Measuring the amount of metal contamination on the surface of a silicon wafer to be measured using a hot plate method. In the method for measuring the amount of metal contamination of a wafer for measuring the amount of metal contamination on the surface of a silicon wafer heated by a hot plate using the hot plate method,
A step of sequentially stacking a silicon wafer on which a silicon oxide film is formed, a clean wafer, and a silicon wafer to be measured on the hot plate;
Measuring the amount of metal contamination on the surface of a silicon wafer to be measured using a hot plate method. In the method for measuring the amount of metal contamination of a wafer for measuring the amount of metal contamination on the surface of a silicon wafer heated by a hot plate using the hot plate method,
A step of sequentially stacking a silicon wafer on which a silicon oxide film is formed, a clean wafer, and a silicon wafer to be measured on the hot plate;
Measuring the amount of metal contamination on the surface of the silicon wafer to be measured using the hot plate method;
And a step of replacing the clean wafer with the silicon wafer to be measured each time the metal contamination amount of the clean wafer reaches a limit value. The method for measuring a metal contamination amount of a wafer according to any one of claims 4, 5, and 6, wherein the contamination metal measured on the wafer surface is copper.

Images