Patent Application
20120139088
The present invention relates to a silicon wafer (hereinafter simply referred to as a wafer) to be preferably used as a substrate for forming a semiconductor device and a method for heat-treating a silicon wafer which performs a heat treatment on a silicon wafer obtained by slicing a silicon single crystal ingot grown by a Czochralski method (hereinafter referred to as a CZ method).
With trends toward high integration of semiconductor devices and micro-miniaturization of a design rule in recent years, for a silicon wafer to be used as a substrate thereof, it is required that no void defects such as Crystal Originated Particle (COP) are present in a region near the wafer surface to be a device active region (specially from the wafer surface to a depth of 5 μm, hereinafter also referred to as a surface region).
Generally, a silicon wafer made by slicing a silicon single crystal ingot grown by a CZ method and then mirror polishing has void defects such as COP therein. The void defects such as COP in the device active region can be eliminated, for example, by subjecting the silicon wafer to heat treatment at a temperature of 1100° C. or higher in an Ar gas atmosphere for 30 minutes or longer by use of a vertical type heat-treatment apparatus or the like (for example, Patent Document 1).
Also, such a silicon wafer having no void defects can be produced, for example by controlling V/G (V indicates a pulling rate and G indicates a temperature gradient) when growing the silicon single crystal ingot in the CZ method, pulling up the silicon single crystal ingot entirely having a defect-free region and then slicing the ingot (For example, Patent Document 2).
However, in the method described in Patent Document 2, the void defects can be generated by a slight variation in V/G and it is very difficult to control it. In addition, in the case of pulling up the defect-free region by controlling V/G, generally, the pulling rate V is required to be lowered and thus productivity in growing the silicon single crystal becomes lower.
For these reasons, a method for forming a defect free layer in the wafer surface region where the device is formed is proposed (for example in Patent Document 3). The method includes growing a silicon single crystal ingot having a region where vacancy type point defects predominate (hereinafter referred to as V-rich region) so that the pulling rate V can be raised; subjecting a silicon wafer obtained by slicing the single crystal ingot to an HF treatment and thereby removing an inner-wall oxide film of a void defect present in the wafer surface region; and then subjecting the silicon wafer to a rapid heating/cooling thermal process (hereinafter also referred to as RTP).
On the other hand, a silicon wafer having highly dense BMD (Bulk Micro Defect) in a region deeper than the wafer surface region (specially, a region deeper than 5 μm from the wafer surface, hereinafter referred to as a wafer bulk region) has a gettering effect for metal impurities or the like incorporated during a device process.
In this respect, for example, a method of performing RTP in a nitrogen or inert atmosphere and controlling the cooling rate, thereby allowing excess vacancies to remain in the wafer bulk region and forming oxygen precipitate nuclei is proposed (For example, Patent Documents 4 and 5).
Further, a method of forming a thermal oxide film having a thickness that is preselected to correspond to a preselected depth of defect-free region in an oxidation atmosphere when performing RTP (for example, Patent Document 6) and a method of performing RTP by supplying an argon-based gas to the silicon wafer surface side and a nitrogen-based gas to the back surface side (for example, Patent Document 7) are known.
As an apparatus for performing such RTP, for example, such apparatus that includes a chamber for accommodating a substrate; a substrate support member arranged in the chamber and having a ring frame for supporting the edge portion of the substrate; a heating portion for heating the substrate; and a modifying gas supplying portion for supplying a modifying gas including a first gas containing an oxide atom in the molecule and a second gas as a dilution gas into a substantially closed space formed on a back surface side of a substrate surface where a semiconductor device is formed, is effective (for example, Patent Document 8).
Patent Literature 1: Japanese Patent Application Laid-Open (kokai) No. 2006-4983
Patent Literature 2: Japanese Patent Application Laid-Open (kokai) No. H8-330316
Patent Literature 3: Japanese Patent Application Laid-Open (kokai) No. 2005-123241
Patent Literature 4: Japanese Patent Application Laid-Open (kohyo) No. 2007-534579
Patent literature 5: Japanese Patent Application Laid-Open (kohyo) No. 2005-522879
Patent Literature 6: Japanese Patent Application Laid-Open (kokai) No. 2000-91259
Patent Literature 7: Japanese Patent Application Laid-Open (kokai) No. 2001-308101
Patent Literature 8: Japanese Patent Application Laid-Open (kokai) No. 2003-77851
However, according to the heat treatment method described in Patent Document 1, productivity decreases due to a heat treatment for a long time. Also, since the long-time heat treatment causes out-diffusion of oxygen from the wafer surface and great reduction of a concentration of solid-solution oxygen in the device active region, it is difficult to inhibit elongation of dislocation generated by application of damage or strain produced in the device process.
Further, according to the above-mentioned method described in Patent Document 3, although the defect free layer can be formed in the wafer surface region to be the device active region, elimination of the void defects or the like in the wafer bulk region deeper than the above-mentioned wafer surface region is not described or considered at all.
Furthermore, according to the methods of Patent Documents 4 and 5, although uniformity in formation of oxygen precipitate nuclei in the wafer bulk region can be achieved, elimination of the void defect or the like in the wafer bulk region is not described or considered at all in the above-mentioned Patent Documents 4 and 5.
As shown in
In addition, since the void defect has a shape of an octahedron, slip can easily be generated by strains concentrated on its sharp corner portion 3.
As described in Patent Documents 4 and 5, BMD may densely be formed in the wafer bulk region so as to prevent the above-mentioned void defect from becoming contamination source. However, in order to form BMD, certain conditions such as having many vacancies in the wafer bulk region and having a high oxygen concentration are needed. Thus, there are limitations on the production. In addition, in the case where the BMD is densely formed, the BMD may become a source of slip generation. Thus, it is not preferred.
Further, according to the heat treatment method described in Patent Document 6, since oxygen in the atmosphere is in-diffused into the wafer surface, concentration of solid-solution oxygen in an outermost wafer surface portion of the device active region is increased. In this outermost wafer surface portion, the void defect may remain without being eliminated because of poor solubility of the inner-wall oxide film of the void defect. In this case, mirror polishing needs to be performed again after performing RTP so as to remove the outermost wafer surface portion, and thus there is a problem of decrease in the productivity.
By contrast, since the argon-based gas is used in the heat treatment method as described in Patent Document 7, the concentration of solid-solution oxygen in the outermost wafer surface portion of the device active region is not increased. In this case, since the inner-wall oxide film of the void defect is more soluble, it is possible to eliminate the void defects in the outermost wafer surface portion. However, where interstitial Si (hereinafter referred to as i-Si) is necessary to eliminate the void defect having a dissolved inner-wall oxide film, the amount of generated i-Si is small in the argon gas compared with the oxygen atmosphere. Therefore, an area capable of eliminating the void defect by RTP which is a short-time heat treatment in the order of seconds may be limited only to the outermost wafer surface portion of the device active region. Thus, an ability of eliminating the void defect may be reduced and the void defect may remain without being eliminated in the device active region lower than the outermost surface portion (hereinafter, referred to as surface layer portion). In this case, there is a problem that the yield is decreased in the device process.
It is to be noted that in the case of using the RTP apparatus as described in Patent Document 8, the outer periphery of the backside of the silicon wafer (within about 1 to 2 mm from the outer edge of the wafer) is supported on a susceptor for example made of SiC such that the supported portion is shaped like a ring. In this case, the center part of the supported wafer is warped into a bowl shape and elastically deformed by influence of gravity and thermal expansion during RTP (
In order to suppress such introduction of slip, it is necessary to reduce the amount of deflection of the wafer and distribute the self-weight stress effectively. As one of its methods, a susceptor for supporting a whole backside surface of the wafer may be used. However, this causes a problem that contact scratches caused by the susceptor are formed on the whole backside surface of the wafer. There may be an elevational difference of about 1 μm between these contact traces and it may locally deteriorate flatness of the front surface side of the wafer where a semiconductor device is formed. Thus, it is not preferred.
The present invention is made by considering the situations as described above, and aims to provide a silicon wafer capable of preventing the void defect present in the wafer bulk region from becoming a contamination source in the device process and further preventing the void defect from becoming a source of the slip generation.
Further, the present invention aims to provide a method for heat-treating a silicon wafer capable of greatly reducing the void defect in the outermost surface portion and the surface layer portion of the device active region in the case of using the rapid heating/cooling thermal process which is the heat treatment performed in the order of seconds.
Furthermore, the present invention aims to provide a method for heat-treating a silicon wafer capable of effectively suppressing introduction of slip and reducing crystal defects such as COP in the region near the wafer surface to be the device active region, even in the case of using the susceptor which supports the outer periphery of backside of the silicon wafer in a ring shape when subjecting the silicon wafer to the rapid heating/cooling thermal process.
A silicon wafer in accordance with the present invention comprises a wafer surface region which is a defect-free region having no void defects, and a wafer bulk region which is deeper than the wafer surface region having the void defects in the shape of a polyhedron whose basic shape is an octahedron, in which a corner portion of the polyhedron is formed in a curved shape and an inner-wall oxide film of the void defect is removed.
By using such silicon wafer, it is possible to prevent the void defect present in the wafer bulk region from becoming a contamination source in a device process. Further, it is possible to prevent the void defect from becoming a source of slip generation.
The above described void defects are preferably a sphere or an ellipsoid in which all the corner portions of the above-mentioned polyhedron are formed in a curved shape.
By using such silicon wafer, it is possible to positively prevent the void defect present in the wafer bulk region from becoming a source of the slip generation.
Further, a method for heat-treating a silicon wafer in accordance with a first embodiment of the present invention is, a method for heat-treating a silicon wafer which performs a rapid heating/cooling thermal process on the silicon wafer produced by a CZ method and comprises supplying an inert gas into a first space on a front surface side of the silicon wafer where a semiconductor device is formed; supplying an oxidizing gas into a second space on a back surface side of the silicon wafer; and performing the process at a maximum achievable temperature of from 1300° C. to 1400° C.
An oxygen partial pressure in the above-mentioned oxidizing gas is preferably set to a range from 20% to 100%.
An oxide film formed on the back surface side of the silicon wafer after performing RTP preferably has a thickness of 15 nm or more.
Further, a heat-treating method in accordance with a second embodiment of the present invention is, a method for heat-treating a silicon wafer which performs a rapid heating/cooling thermal process on the silicon wafer produced by a CZ method and comprises a first step of rapidly heating the silicon wafer to a maximum achievable temperature; a second step of maintaining the maximum achievable temperature for a predetermined period of time; and a third step of rapidly cooling from the maximum achievable temperature, wherein at least the first step is performed under a condition where an inert gas is supplied into a first space on a front surface side of the silicon wafer where a semiconductor device is formed, an oxidizing gas is supplied into a second space on the back surface side of the silicon wafer, and an inner pressure of the second space is set to a negative pressure compared to an inner pressure of the first space. By applying such method, when subjecting the silicon wafer to RTP, it is possible to effectively suppress introduction of slip and reduce crystal defects such as COP in the region near the wafer surface to be a device active region even with using a susceptor which supports the outer periphery of the back surface side of the silicon wafer such that the supported portion is shaped like a ring.
The above-mentioned negative pressure is preferably −500 Pa or less. In such a negative pressure range, more significantly, it is possible to effectively suppress introduction of slip and reduce the crystal defects such as COP in the region near the wafer surface to be the device active region.
The above-mentioned first space and second space are preferably spatially connected.
By using the RTP apparatus having such a structure and performing RTP under the above-described conditions, it is possible to suppress introduction of slip more effectively.
In accordance with the present invention, it is possible to provide a silicon wafer capable of preventing the void defect present in the wafer bulk region from becoming a contamination source and further preventing the void defect from becoming a source of slip generation in the device process.
Therefore, the silicon wafer in accordance with the present invention greatly contributes to enhancing the yield in the device process.
Further, by the method for heat-treating the silicon wafer in accordance with the first embodiment of the present invention, it is possible to greatly reduce the void defects in the outermost surface portion and the surface layer portion of the device active region even in the case of performing RTP which is a heat treatment in the order of seconds.
Therefore, the silicon wafer which is subjected to the heat treatment in accordance with the first embodiment of the present invention greatly contributes to enhancing the yield in the device process.
Further, by the method for heat-treating the silicon wafer in accordance with the second embodiment of the present invention, when subjecting the silicon wafer to RTP, it is possible to effectively suppress introduction of slip and reduce crystal defects such as COP in the region near the wafer surface to be the device active region even with using the susceptor which supports the outer periphery of the back surface side of the silicon wafer such that the supported portion is shaped like a ring.
Hereinafter, the present invention will be explained in detail with reference to the accompanying drawings.
As shown in
By contrast,
The conventional wafer is obtained by slicing a silicon single crystal ingot grown by a CZ method so as to have a V-rich region and subjecting the thus-obtained silicon wafer to heat treatment so as to form a wafer surface region 1 as a defect-free layer. With respect to this conventional wafer, since oxygen in a wafer bulk region 2 is poorly out-diffused in the heat treatment, an inner-wall oxide film 6 of a void defect 5 present in the wafer bulk region 2 is not dissolved but remains.
Such inner-wall oxide film 6 has a characteristic of introducing metal impurities around the void defect 5. Therefore, in the case where there are many such void defects 5 in the wafer bulk region 2, there is a possibility that the void defect 5 itself becomes a contamination source in the device process.
For the silicon wafer in accordance with the present invention, since such inner-wall oxide film is removed, it is possible to prevent the void defect present in the wafer bulk region from becoming the contamination source in the device process.
Further, the void defect 4 as shown in
That is, since the sharp corner of the void defect 4 of the silicon wafer in accordance with the present invention is formed in a curved shape, it is possible to prevent the void defect 4 from becoming a source of slip generation.
Therefore, the silicon wafer in accordance with the present invention greatly contributes to enhancing the yield in the device process.
The above-mentioned void defect 4 is preferably a sphere or an ellipsoid in which all the corner portions of the above-mentioned polyhedron are formed in a curved shape as shown in
Such silicon wafer can positively prevent the void defect present in the wafer bulk region from becoming a source of slip generation.
For such void defect 4 present in the wafer bulk region 2, by performing grinding and polishing from the wafer surface region 1 to the wafer bulk part 2, its shape can be evaluated by using a transmission electron microscope.
Next, a method for producing a silicon wafer and a method for heat-treating a silicon wafer in accordance with the present invention will be explained.
The above-described silicon wafer in accordance with the present invention can be produced by performing RTP on a silicon wafer obtained by slicing a silicon single crystal ingot produced by a CZ method.
Further, the method for heat-treating the silicon wafer in accordance with the present invention is to perform RTP on the silicon wafer produced by slicing the silicon single crystal ingot grown by the CZ method.
The CZ method is, as well-known, a method for growing the silicon single crystal ingot by heating polysilicon filled in a quartz crucible into a silicon melt solution, causing a seed crystal to come in contact with a liquid level of the silicon melt solution, pulling up the seed crystal while rotating the seed crystal and the quartz crucible, forming a straight body portion by increasing a diameter into a desirable diameter, and then cutting it off from the silicon melt solution.
At this time, the silicon single crystal ingot is grown so as to have a V-rich region by controlling V/G (V indicates a pulling rate and G indicates a temperature gradient).
Then, after the thus-obtained silicon single crystal ingot is sliced into a wafer by using an inner diameter slicing or wire saw or the like, a silicon wafer is obtained by performing a machining operation such as beveling of the outer periphery, lapping, etching and mirror polishing.
The above-mentioned RTP can preferably be performed on the thus-obtained mirror polished wafer for example by an RTP apparatus as shown in
The RTP apparatus 10 shown in
It is to be noted that in the state that the wafer W is supported on the wafer support 30, a first space 20a surrounded by an inner wall of the reaction tube 20 and a front surface W1 side of the wafer W where a semiconductor device is formed, and a second space 20b surrounded by the inner wall of the reaction tube 20 and a back surface W2 side of the wafer W are formed.
The above-mentioned reaction tube 20 includes a gas supply port 22 for supplying a first atmosphere gas FA (solid-line arrow in the drawing) into the first space 20a on the front surface W1 side of the wafer W where the semiconductor device is formed, a gas discharge port 26 for discharging the gas from the first space 20a, a gas supplying port 24 for supplying a second atmosphere gas FB (dashed-line arrow in the drawing) into the second space 20b on the back surface W2 side of the wafer W and a gas discharge port 28 for discharging the gas from the second space 20b. The reaction tube 20 is made of, for example, quartz.
The wafer support 30 includes a susceptor 32 for directly supporting the outer periphery of the back surface W2 side of the wafer W such that the supported portion is shaped like a ring and a rotation part 34 for supporting the susceptor 32 and rotating the susceptor 32 in the diameter direction. The susceptor 32 and the rotating part 34 are made of, for example, SiC.
A heating unit 40 is disposed above the wafer support 30 and outside the reaction tube 20, and heats the wafer W from the front surface W1 side. The heating unit 40 is constituted of, for example, a plurality of halogen lamps 50.
In addition, the first space 20a and the second space 20b formed within the reaction tube 20 are spatially connected each other by a connection part 60.
In the case of performing RTP by using the RTP apparatus 10 shown in
Hereinafter, one example of RTP for obtaining the silicon wafer in accordance with the present invention by using the RTP apparatus shown in
In the heat treatment sequence shown in
Next, the maximum achievable temperature T1 is maintained for a predetermined time t (sec).
Thereafter, the light irradiation from the halogen lamp 50 is turned off and the wafer W is rapidly cooled at a predetermined temperature drop rate ΔTd (° C./sec), if desired, by using the second atmosphere gas FB.
The temperature of the wafer W in the above-described heat treatment sequence is measured, for example, by a radiation thermometer (not shown) disposed below the wafer W. Also, the temperature rise rate and the temperature drop rate are controlled using a control unit (not shown) based on the temperature measured as above by controlling the output of individual halogen lamps 50 and the flow rate of the first atmosphere gas FA or the second atmosphere gas FB or the like
As a first method for obtaining the silicon wafer in accordance with the present invention, an oxygen concentration of the silicon single crystal ingot to be grown is preferably not more than 1.1×1018 atoms/cm3. In other words, the oxygen concentration of the wafer bulk region of the silicon wafer in accordance with the present invention is preferably not more than 1.1×1018 atoms/cm3.
By such method, an inner-wall oxide film of the void defect present in the wafer bulk region 2 becomes more soluble in RTP, and interstitial-Si (hereinafter be called “i-Si”) generated in the wafer surface region 1 is diffused into the wafer bulk part 2 and enters the void defect where the inner-wall oxide film is removed. Thus, the silicon wafer shown in
The oxygen concentration as used herein means a value determined from a conversion factor of Old ASTM standards in 1970-1979 Edition, it can be evaluated by infrared spectroscopic analysis, or a secondary ion mass spectrometer (SIMS).
Further, in addition to the above-mentioned method, by increasing the predetermined time t (second) of the maximum achievable temperature T1, i-Si generated in the wafer surface region 1 is increased and many i-Si enter the void defects and fill the inside of the void defects in which the inner-wall oxide film thereof is removed furthermore. Thus, the silicon wafer as shown in
The above-mentioned oxygen concentration is preferably not less than 0.8×1018 atoms/cm3 in order to suppress slip generation in RTP and maintain the strength of the wafer.
Further, as a second method for obtaining the silicon wafer in accordance with the present invention, an inert gas is preferably used as the first atmosphere gas FA and an oxidizing gas is preferably used as the second atmosphere gas FB in the heat treatment sequence of RTP for example shown in
By this method, a large amount of i-Si is generated on the back surface W2 side and diffused to the front surface W1 side of the wafer, and dissolution of the inner-wall oxide film of the void defect present in the wafer bulk region 2 is enhanced by i-Si. Thus, the void defect in the shape of a polyhedron whose basic shape is an octahedron is shaped to approximate to a sphere which is energetically stable and i-Si fills the inside the void defect, thereby the silicon wafer as shown in
Further, by increasing the predetermined time t (second) of the maximum achievable temperature T1, many i-Si generated on the back surface W2 side of the wafer are diffused into the wafer bulk region 2, enter the void defects in which the inner-wall oxygen film thereof is removed, and fill the inside the void defects having approximate to the sphere shape furthermore. Thus, the silicon wafer as shown in
Still further, as a third method for obtaining the silicon wafer in accordance with the present invention, the oxidizing gas is preferably used as the first atmosphere gas FA and the second atmosphere gas FB in the heat treatment sequence of RTP for example shown in
By this method, a large amount of i-Si is generated on the front and back surfaces of the silicon wafer and dissolution of the inner-wall oxide film of the void defect present in the wafer bulk region 2 is enhanced by i-Si. Thus, the void defect in the shape of a polyhedron whose basic shape is an octahedron is shaped to approximate to a sphere which is energetically stable and i-Si fills the inside the void defect, thereby the silicon wafer as shown in
Further, by increasing the predetermined time t (second) of the maximum achievable temperature T1, many i-Si generated in the wafer front and back surfaces are diffused into the wafer bulk region 2, enter the void defects in which the inner-wall oxygen film is removed, and fill the inside the void defects having approximate to the sphere shape furthermore. Thus, a silicon wafer as shown in
It is to be noted that when the third method is used, void defects may remain at the outermost surface portion of the silicon wafer (for example, within 1 μm from the wafer surface). In this case, the silicon wafer in accordance with the present invention as shown in
The above-mentioned maximum achievable temperature T1 in RTP is preferably from 1300° C. to a melting point of silicon.
In the case where the above-mentioned maximum achievable temperature T1 is lower than 1300° C., it is difficult to enhance an ability of eliminating the void defects in the wafer surface region to be the device active region.
On the other hand, the maximum achievable temperature T1 exceeding the melting point of silicon is not preferred because the silicon wafer being subjected to the heat treatment is melted.
It is to be noted that upper limit of the above-mentioned maximum achievable temperature T1 is more preferably not higher than 1380° C. in terms of the durability of the RTP apparatus
Further, the inert gas to be used in RTP is preferably argon gas and the oxidizing gas is preferably oxygen.
In the case of using a nitrogen gas as the inert gas, a nitride film is formed on the wafer surface in RTP, and an etching step or the like for removing the nitride film must be newly added, which is not preferred because of increase in the number of steps. Also, a hydrogen gas is not preferred because a mixed gas of oxygen to be used as the second atmosphere gas FB and hydrogen has a danger of explosion. Furthermore, an ammonia-based gas reduces the ability of eliminating the void defects in the wafer surface region and is not preferred.
A temperature rise rate ΔTu is preferably from 10° C./sec to 150° C./sec.
The above-mentioned temperature rise rate ΔTu lower than 10° C./sec results in poor productivity. Further, there is a problem that an area where oxygen is diffused into the inside of the substrate and reached to the saturation concentration is increased, therefore an area where elimination of the inner-wall oxide film is insufficient is increased. Thus, it is not preferred. On the other hand, in the case where the temperature rise rate ΔTu exceeds 150° C./sec, there is a problem that the silicon wafer cannot withstand the excessively rapid temperature change and a slip is generated on the silicon wafer.
In addition, the temperature drop rate ΔTd is preferably from 10° C./sec to 150° C./sec. There is a problem that the above-mentioned temperature drop rate ΔTd lower than 10° C./sec results in poor productivity, whereas in the case where it exceeds 150° C./sec, the silicon wafer cannot withstand the excessively rapid temperature change and the slip is generated on the silicon wafer.
Next, a heat treatment method in accordance with a first embodiment of the present invention will be explained. RTP applied to the heat treatment method in accordance with the first embodiment of the present invention is performed on a mirror-polished silicon wafer under the conditions such that using an RTP apparatus 10 as illustrated in
More specifically, the mirror-polished wafer W is placed inside a reaction tube 20 kept at a temperature of T0 (for example, 600° C.). Subsequently, the inert gas is supplied into the first space 20a on the front surface W1 side of the wafer W where the semiconductor device is formed and the oxidizing gas is supplied into the second space 20b on the back surface W2 side of the wafer W. The temperature is rapidly heated from the temperature T0 (° C.) to the maximum achievable temperature T1 (° C.) which is from 1300° C. to 1400° C. at a predetermined temperature rise rate ΔTu (° C./sec). Thereafter, the maximum achievable temperature T1 (° C.) is maintained for a predetermined time t (sec) and rapidly cooled from the above-mentioned maximum achievable temperature T1 (° C.) to a temperature for removing the wafer W out of the reaction tube 20 (for example, temperature T0 (° C.)) at a predetermined temperature drop rate ΔTd (° C./sec).
It is to be noted that when using the RTP apparatus 10 as illustrated in
Thus, by the method for heat-treating the silicon wafer in accordance with the first embodiment of the present invention having the above-described features, void defects can greatly be reduced in the outermost surface portion and surface layer portion of the device active region even in the case of using RTP which is a heat treatment performed in the order of seconds. Therefore, the silicon wafer being subjected to the heat treatment of the present invention can greatly contribute to enhancing the yield in the device process.
Next, a mechanism for obtaining the effect of the invention will be explained.
In RTP of the present invention, the inert gas (argon in
It is to be noted that in the case where the oxidizing gas is supplied to the front surface W1 side of the wafer, the inner-wall oxide film of the void defects becomes less soluble due to an increase in concentration of solid-solution oxygen in the outermost surface portion of the device active region. Thus, the void defects may remain without being eliminated.
In RTP, when the oxidizing gas (oxygen in
Therefore, in the outermost surface portion Da, i-Si generated by the in-diffusion of oxygen cannot enter the void defects because of the inner-wall oxide film remaining in the void defects. As a result, it is considered that the void defects remain in the outermost surface portion Da (
It is to be noted that if the oxidizing gas is not supplied to the back surface W2 side of the wafer, an atmosphere where a rare gas is supplied to the front surface W1 side of the wafer does not allow a sufficient amount of i-Si generated on the front surface W1 side of the wafer. For this reason, the void defects may be eliminated only in the outermost surface portion of the device active region by RTP which is short-time heat treatment performed in the order of seconds. In the surface layer portion which is underlying layer of the outermost surface portion, the ability of eliminating the void defects is reduced and the void defects may remain without being eliminated.
In the case where the inert gas (argon in
It is to be noted that in the case where the atmosphere gas to be supplied into the first space 20a is a hydrogen gas, a mixed gas of the oxidizing gas (for example, oxygen) to be supplied to the second space 20b and the hydrogen gas poses a risk of explosion. Thus, it is not preferred.
In addition, in the case where the atmosphere gas to be supplied into the first space 20a or the second space 20b is nitrogen, a nitride film is formed on the surface of the wafer W in RTP, and an etching step or the like for removing the nitride film must be newly added, which is not preferred because of increase in the number of production steps.
Argon is preferably used as the inert gas to be supplied into the first space 20a. And, oxygen is preferably used as the oxidizing gas to be supplied into the second space 20b.
The above-mentioned maximum achievable temperature T1 is from 1300° C. to 1400° C. The ability of eliminating the void defects in the device active region can be enhanced under such temperature condition.
The above-mentioned maximum achievable temperature T1 lower than 1300° C. is not preferred because the ability of eliminating the void defects in the device active region is reduced. The above-mentioned maximum achievable temperature T1 exceeding 1400° C. is not preferred because it becomes close to the melting point of silicon and thereby causing thermal deformation of the wafer and generation of slip or the like.
More preferably, the upper limit of the above-mentioned maximum achievable temperature is not higher than 1380° C. in terms of the durability of the RTP apparatus.
The above-mentioned oxidizing gas preferably has an oxygen partial pressure of from 20% to 100%. By setting such a condition of the oxygen partial pressure, many i-Si are generated in the wafer and therefore, the ability of eliminating the void defects in the surface layer portion of the device active region can be enhanced. It is to be noted that in the case where the above-mentioned oxygen partial pressure is less than 20%, it is difficult to enhance the ability of eliminating the void defects in the surface layer portion because only a small amount of i-Si is generated in the wafer.
The above-mentioned inert gas is preferably an oxygen-free 100% argon gas and the above-mentioned oxidizing gas is preferably a 100% oxygen gas. By employing such a composition, the void defects can greatly be reduced in the outermost surface portion and the surface layer portion of the device active region.
An oxide film formed on the back surface side of the silicon wafer after the above-described RTP preferably has a thickness of not less than 15 nm.
By employing such an oxide film thickness, many i-Si are generated in the wafer and the ability of eliminating the void defects in the outermost surface portion of the device active region can be enhanced.
The holding time (t) of the above-mentioned maximum achievable temperature T1 (° C.) is preferably from 1 to 15 seconds. By employing such a configuration, the void defects can greatly be reduced in the outermost surface portion and the surface layer portion of the device active region, therefore the productivity can greatly be improved. It is to be noted that in the case where the above-mentioned holding time (t) is less than 1 second, because of the short holding time (t), there is a possibility that void defects are not reduced in the device active region. Thus, it is not preferred.
Next, a heat treatment method of a silicon wafer in accordance with a second embodiment of the present invention will be explained.
In a heat treatment sequence in accordance with the heat treatment method of the second embodiment, a mirror-polished wafer W is placed inside a reaction tube 20 in an RTP apparatus 10 kept at a temperature of T0 (for example, 600° C.) such as illustrated in
Next, temperature is rapidly heated from a temperature T0 (° C.) to a first temperature T1 (° C.) which is a maximum achievable temperature at a predetermined temperature rise rate ΔTu (° C./sec) (first step), thereafter, kept at the above-mentioned first temperature T1 (° C.) for a predetermined time t (second) (second step), and finally, rapidly cooled from the above-mentioned first temperature T1 (° C.) to a second temperature (for example, temperature T0 (° C.)) at a predetermined temperature drop rate ΔTd (° C./sec) (third step). It is to be noted that when placing the wafer W in the reaction tube 20 of the RTP apparatus 10 as illustrated in
In the heat treatment method of the silicon wafer in accordance with the second embodiment of the present invention, at least in the above-mentioned first step, an inert gas is supplied as the first atmosphere gas FA and an oxidizing gas is supplied as the second atmosphere gas FB respectively. In addition, in order to prevent the second atmosphere gas FB from flowing into the first space 20a, RTP is performed under a condition where an inner pressure of the second space 20b is set to a negative pressure (in other words, lower pressure) compared to an inner pressure of the first space 20a.
The above-mentioned negative pressure can be adjusted by adjusting a discharge pressure set point of a second discharge port 28 relative to that of a first discharge port 26. By supplying the oxidizing gas as the second atmosphere gas FB, it is possible to reduce the amount of deflection of the wafer as illustrated in
In addition, by supplying the inert gas as the first atmosphere gas FA and setting the inner pressure of the second space 20b more negative than that of the first space 20a, it is possible to prevent the above-mentioned oxidizing gas from flowing into the first space 20a and reduce crystal defects such as COP in the region near the front surface W1 side of the wafer.
It is to be noted that in the case where the inner pressure of the second space 20b is not set to a negative pressure compared to the inner pressure of the first space 20a thereby allowing the oxidizing gas to flow into the first space 20a, thick oxide film is also formed on the front surface W1 side of the wafer W. In this case, because the thick oxide films are formed on the front and back surface sides of the wafer W, it is difficult to reduce the amount of deflection of the wafer by use of difference in the amount of thermal expansion between silicon and the silicon oxide film as described above. Further, when the thick oxide film is formed on the front surface W1 side of the wafer W, dissolution of an inner-wall oxide film of COP in the region near the wafer surface is inhibited. As a result, it becomes difficult to reduce the crystal defects such as COP in the region near the wafer surface.
It is to be noted that in the case where the first atmosphere gas FA is an oxidizing gas, thick oxide film is formed on the front surface W1 side of the wafer W. Thus, as described above, it is difficult to effectively suppress introduction of slip into the wafer W. In addition, it is difficult to reduce the crystal defects such as COP in the region near the wafer surface.
Further, in the case where a gas other than the oxidizing gas is used as the second atmosphere gas FB, because oxide film is not formed on the back surface W2 side of the wafer W, it is difficult to suppress the introduction of slip effectively.
An argon gas is preferably used as the above-mentioned inert gas, and a 100% oxygen gas or a mixed gas of an oxygen gas and a helium gas is preferably used as the above-mentioned oxidizing gas.
It is preferred that the above-mentioned second space 20b is not substantially a closed space when holding the wafer W on the susceptor 32. In other words, in
If the second space 20b of the RTP apparatus is a closed space while holding the wafer W, when RTP is performed under the condition where the inner pressure of the second space 20b is set to a negative pressure compared to that of the first space 20a as described above, stress is generated in the wafer W held on the susceptor 20 in the direction of the second space 20b having a lower pressure and the wafer W is more deflected. It results in an increase in the amount of deflection of the wafer W and it becomes difficult to effectively suppress the introduction of slip.
The above-mentioned negative pressure is preferably −500 Pa or less. That is, the inner pressure of the second space 20b is preferably set to a negative pressure which is −500 Pa or less compared to that of the first space 20a. As described above, this negative pressure is adjusted by adjusting the discharge pressure set point of the second discharge port 28 relative to that of the first discharge port 26 such that the discharge pressure set point of the second discharge port 28 is −500 Pa or less compared to that of the first discharge port 26. Such negative pressure range can more remarkably and effectively suppress the introduction of slip and reduce the crystal defects such as COP in the region near the wafer surface to be the device active region.
In the case where the above-mentioned negative pressure exceeds −500 Pa, because the oxidizing gas supplied into the second space 20b flows into the first space 20a, it becomes difficult to suppress the introduction of slip into the wafer W in RTP. In addition, it becomes difficult to reduce the crystal defects such as COP in the region near the wafer surface.
The above-mentioned negative pressure is preferably from −700 Pa to −500 Pa.
In the case where the above-mentioned negative pressure is lower than −700 Pa, it becomes difficult to supply the inert gas supplied into the first space 20a in the direction of the front surface W1 side of the wafer W, and results in a decrease in the effect of reducing the crystal defects such as COP in the region near the wafer surface. Thus, it is not preferred.
The above-mentioned temperature rise rate ΔTu is, for example, from 10° C./second to 150° C./second.
The above-mentioned first temperature T1 (° C.) is preferably from 1300° C. to a melting point of silicon. If the above-mentioned first temperature T1 (° C.) is lower than 1300° C., it results in a decrease in the effect of reducing the crystal defects such as COP in the region near the wafer surface. Thus, it is not preferred. The above-mentioned first temperature T1 (° C.) exceeding the melting point of silicon is not preferred because the wafer W is melted.
In addition, the above-mentioned first temperature T1 (° C.) is more preferably from 1300° C. to 1380° C. in terms of the durability of the RTP apparatus.
The holding time t for which the above-mentioned first temperature T1 (° C.) is held is, for example, from 1 to 60 seconds.
The above-mentioned temperature drop rate ΔTd is, for example, from 10° C./second to 150° C./second.
Hereinafter, the present invention will be explained in further detail with reference to the following examples, however, the present invention is not limited thereto.
A silicon wafer (a diameter of 300 mm, a thickness of 775 μm and an oxygen concentration of 1.1×1018 atoms/cm3)obtained by slicing a silicon single crystal ingot grown by a CZ method so as to have a V-rich region was subjected to an RTP treatment by using an RTP apparatus as illustrated in
Thereafter, presence or absence of void defects in the wafer surface region (from the wafer surface to a depth of 5 μm) was evaluated by a laser light scattering method, and after removing a surface layer by grinding and polishing, void defects in a wafer bulk region (a region deeper than 5 μm from the wafer surface) were observed by using a transmission electron microscope.
As a result, it was confirmed that the wafer surface region of thus-obtained silicon wafer was a defect-free region having no void defect, and the wafer bulk region included many void defects in the shape of a polyhedron whose basic shape was an octahedron in which a corner portion of the above-mentioned polyhedron was formed in a curved shape and inner-wall oxide films of the void defects were removed.
Except for setting the holding time t of the maximum achievable temperature T1 as 15 seconds, a silicon wafer (a diameter of 300 mm, a thickness of 775 μm and an oxygen concentration of 1.1×1018 atoms/cm3) obtained by slicing a silicon single crystal ingot grown by the CZ method so as to have a V-rich region was subjected to the RTP treatment in the manner similar to that in Example 1.
Thereafter, presence or absence of void defects in the wafer surface region and void defects in the wafer bulk region were observed in the manner similar to that in Example 1.
As a result, it was confirmed that the wafer surface region of the thus-obtained silicon wafer was a defect-free region having no void defect, and the wafer bulk region included many void defects in the shape of a polyhedron whose basic shape was an octahedron. The void defects were a sphere or an ellipsoid in which all the corner portions of the above-mentioned polyhedron were formed in curved shape and inner-wall oxide films of the void defects were removed. In addition, the size of the void defects was smaller than that in Example 1.
Except for using a 100% argon gas as the second atmosphere gas FB, a silicon wafer (a diameter of 300 mm, a thickness of 775 μm and an oxygen concentration of 1.1×1018 atoms/cm3) obtained by slicing a silicon single crystal ingot grown by the CZ method so as to have a V-rich region was subjected to the RTP treatment in the manner similar to that in Example 1.
Thereafter, presence or absence of void defects in the wafer surface region and void defects in the wafer bulk region were observed in the manner similar to that in Example 1.
As a result, it was confirmed that the wafer surface region was a defect-free region having no void defect but the wafer bulk region included many void defects in which inner-wall oxide films thereof were not removed.
A silicon single crystal ingot was grown by the CZ method so as to have a region where vacancy-type point defects were predominantly present while controlling v/G (v indicates a pulling rate and G indicates a temperature gradient in the pulling axis direction within the single crystal). Thereafter, a silicon wafer was sliced from a region where the vacancy-type point defects were predominantly present and mirror-polished on both surfaces. The thus-obtained silicon wafer (a diameter of 300 mm and a thickness of 775 μm) was subjected to the RTP treatment by using the RTP apparatus 10 as illustrated in
With regard to the annealed wafer prepared under the above-described conditions, LPD (Light Point Defect) in the wafer surface where a device to be formed was measured by using Surfscan SP2 (40 nm) supplied by KLA-Tencor Corporation. In addition, LPD distribution in the depth-direction from the wafer surface was measured in a similar manner by repeatedly polishing the wafer surface where the device to be formed.
Further, with regard to the respective annealed wafers prepared under the respective conditions, LSTD (5 μm) in the wafer surface where the device to be formed was measured to calculate a decrease rate of a defect density. This measurement was carried out by using an LSTD (Laser Scattering Tomography Defect) scanner (MO-601, supplied by Raytex Corporation). Furthermore, for thickness of oxide film formed on the back surface side of the wafer, the center point of the front and back surfaces of the wafer was measured by an ellipsometric method by using a three-wavelength automatic ellipsometer Auto EL IV NIR III supplied by Rudolph Research Analytical.
The evaluation conditions of this evaluation and evaluation results of LSTD are shown in Table 1, and the measurement results of LPD in Example 3 and Comparative Examples 2 to 6 are shown in
As seen from Table 1 and
A silicon single crystal ingot was grown by the CZ method so as to have a region where vacancy-type point defects were predominantly present while controlling v/G (v indicates the pulling rate and G indicates the temperature gradient in the pulling axis direction within the single crystal). Thereafter, a silicon wafer was sliced from the region where the vacancy-type point defects were predominantly present and mirror-polished on both surfaces. The thus-obtained silicon wafer (a diameter of 300 mm and a thickness of 775 μm) was subjected to the RTP treatment by using the RTP apparatus 10 as illustrated in
Thickness of oxide film formed on the front and back surfaces of the annealed wafer obtained under the above-mentioned conditions was evaluated in the manner similar to that in Example 3. In addition, presence or absence of slip generation was evaluated by using X-ray topography (XRT 300) supplied by Rigaku Corporation. And, elimination of COP in the region near the wafer surface (a depth of from 0 to 5 μm) was evaluated. The COP was evaluated by using a LSTD scanner (MO-601 supplied by Raytex Corporation).
Examination conditions and evaluation results of Example 4 and Comparative examples 7 to 10 are shown in Table 2.
As seen from Table 2, it was confirmed that slip was not generated and the COP in the region near the surface was effectively eliminated by using the argon gas as the first atmosphere gas FA and the oxygen gas as the second atmosphere gas FB and setting the discharge pressure set point to −500 Pa or less.
Except for varying the above-mentioned discharge pressure set point (Pa), the RTP treatments were performed under the conditions similar to those in Example 4.
With regard to the annealed wafers obtained under the above-mentioned conditions, thickness of oxide film formed on the front and back surfaces, presence or absence of slip generation and elimination of COP were evaluated in the manner similar to that in Example 4.
Examination conditions and evaluation results in Examples 5 to 7 and Comparative example 11 are shown in Table 3.
As seen from Table 3, when the argon gas was used as the first atmosphere gas FA, the oxygen gas was used as the second atmosphere gas FB and the discharge pressure set point exceeded −700 Pa, although slip was not generated, the COP in the region near the surface was not eliminated but remained.
In the RTP apparatus 10 as illustrated in
Examination condition and evaluation result in Comparative example 12 are shown in Table 4.
As seen from Table 4, when the connection part 60 was sealed and the second space 20b was temporally closed in the RTP apparatus 10, although the COP in the region near the surface was eliminated, slip was generated.
The kind of gas and condition of the negative pressure as indicated in Example 4 were applied only for the first step (Example 8), only for the second step (Comparative example 13) and only for the third step (Comparative example 14) in the heat treatment sequence shown in
With regard to the annealed wafers obtained under the above-mentioned conditions, thickness of oxide film formed on the front and back surfaces, presence or absence of slip generation and elimination of COP were evaluated in the manner similar to that in Example 4.
Examination conditions and evaluation results in Example 8 and Comparative examples 13 and 14 are shown in Table 5.
As seen from Table 5, by applying the kind of gas and the condition of the negative pressure as indicated in Example 4 at least in the first step, slip generation was prevented and the COP in region near the surface was eliminated. However, in the case of using argon as the second the atmosphere gas FB in the first step and the third step (Comparative example 13) or in the first step and second step (Comparative example 14), although the COP in the region near the surface was eliminated, slip was generated.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2009-134252 | Jun 2009 | JP | national |
| 2009-153776 | Jun 2009 | JP | national |
| 2009-179319 | Jul 2009 | JP | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/JP2010/003583 | 5/28/2010 | WO | 00 | 2/16/2012 |
