JP5583070B2 - Heat treatment method for silicon wafer - Google Patents

Description

  The present invention relates to a heat treatment method for a silicon wafer in which heat treatment is performed on a silicon wafer (hereinafter also simply referred to as a wafer).

  Silicon wafers used as semiconductor device forming substrates reduce void defects such as COP (Crystal Originated Particles) and LSTD (Laser Scattering Tomography Defects) in the vicinity of the surface of the wafer that will be the device active region (hereinafter referred to as the surface layer portion). There is a need for efforts to make them defect-free.

  In recent years, as a method of manufacturing such a silicon wafer with high productivity, at least a surface on which a semiconductor device is formed is mirror-polished (hereinafter, the mirror-polished surface is also referred to as a polished surface). A technique for performing a thermal treatment (Rapid Thermal Process: hereinafter also simply referred to as RTP) is known.

  As such a technique, Patent Document 1 discloses that an oxygen-containing gas atmosphere (inert gas atmosphere as referred to in the present invention) that is mainly argon or helium at a temperature exceeding about 1175 ° C. under an oxygen partial pressure of less than about 5000 ppma, A heat treatment method is disclosed in which the wafer is heated for less than 60 seconds.

  However, since the method described in Patent Document 1 performs RTP in an inert gas atmosphere such as argon or helium, it is possible to greatly reduce void defects in the surface layer portion of the wafer. When RTP is performed at a high temperature exceeding 1175 ° C. in a reducing gas atmosphere such as an atmosphere or hydrogen, polishing after RTP is caused by the influence of a natural oxide film formed on the polishing surface of the wafer before RTP. There is a problem that the surface roughness of the surface deteriorates.

  In order to solve such a problem, Patent Document 2 discloses a mixed gas of argon containing 100% hydrogen or 10% or more of hydrogen using an RTP apparatus after removing a natural oxide film on the wafer surface by hydrofluoric acid treatment. Disclosed is a heat treatment method that can reduce the microroughness of the wafer surface and remove void defects existing on the wafer surface by heat treatment in an atmosphere.

JP-T-2001-509319 JP 2000-91342 A

In such a method described in Patent Document 2, since hydrogen is terminated at silicon atoms on the wafer surface by hydrofluoric acid treatment, a natural oxide film is hardly formed on the surface. Therefore, even if the RTP is performed, deterioration of the surface roughness on the wafer surface can be suppressed.
However, in order to eliminate void defects existing in the surface layer portion of the wafer by RTP, a high-temperature heat treatment of at least 1000 ° C. is required in the inert gas atmosphere or the reducing gas atmosphere. Below, the bond of the hydrogen atom terminated by the silicon atom is easily broken, and the silicon atom is easily exposed on the wafer surface. The exposed silicon atoms are unstable and are likely to react with other atoms.

  Therefore, for example, when other reactive gas (such as nitrogen) is present in the atmosphere, it reacts and bonds with the exposed silicon atoms, and the bond is repeatedly etched by the atmosphere. Therefore, there is a problem that the surface shape of the wafer changes and the surface roughness deteriorates.

Further, when the atmosphere contains a small amount of oxygen, the exposed silicon atoms and oxygen react to form an oxide film in an island shape on the wafer surface, and this oxide film is etched by the atmosphere. However, there is a problem that concave pits are formed in the etched portion.
The above problem becomes more conspicuous as the heat treatment temperature in RTP becomes higher. On the other hand, the higher the heat treatment temperature, the higher the void defect extinction power of the surface layer portion of the wafer. .

  The present invention has been made in view of the above-described circumstances, and even when RTP is performed at a high temperature at which void defect annihilation power is high, deterioration of surface roughness can be suppressed. An object of the present invention is to provide a method for heat-treating a silicon wafer that can also suppress the occurrence of the above.

The method for heat-treating a silicon wafer according to the present invention includes a step of cleaning the surface of a silicon wafer on which at least a surface on which a semiconductor device is formed is mirror-polished with a hydrogen fluoride-based solution, and an ammonia-based cleaning of the cleaned silicon wafer. A step of performing a first rapid heating / cooling heat treatment in which a temperature is rapidly raised and maintained in a first temperature range of 900 ° C. or more and 1250 ° C. or less in a gas atmosphere, and then the temperature is rapidly lowered; The silicon wafer is rapidly heated and maintained in a second temperature range of 1300 ° C. to 1400 ° C. in a rare gas (excluding nitrogen gas, meaning a rare gas, the same applies hereinafter) atmosphere, and then rapidly cooled. And a step of performing a second rapid heating / cooling heat treatment.

The silicon wafer heat treatment method according to the present invention includes a step of cleaning the surface of a silicon wafer on which at least a surface on which a semiconductor device is formed is mirror-polished with a hydrogen fluoride solution, and the cleaned silicon wafer, In the ammonia gas atmosphere, the temperature is rapidly raised to a first temperature range of 900 ° C. or more and 1250 ° C. or less, the ammonia gas atmosphere is switched to a rare gas atmosphere in the first temperature range, and further, 1300 ° C. or more and 1400 ° C. And a step of performing a rapid temperature raising and lowering heat treatment in which the temperature is rapidly raised and held in a second temperature range of not more than 0 ° C. and then rapidly lowered.

  According to the present invention, even when RTP is performed at a high temperature with high void defect annihilation power, the deterioration of surface roughness can be suppressed, and furthermore, the generation of concave pits can be suppressed. A heat treatment method is provided.

It is a cross-sectional conceptual diagram which shows an example of the RTP apparatus applied to the heat processing method of the silicon wafer concerning this invention. It is a conceptual diagram which shows an example of the heat processing sequence of 1st RTP applied to the heat processing method of the silicon wafer concerning 1st Embodiment. It is a conceptual diagram which shows an example of the heat processing sequence of 2nd RTP applied to the heat processing method of the silicon wafer concerning 1st Embodiment. It is a conceptual diagram which shows an example of the heat processing sequence in RTP applied to the heat processing method of the silicon wafer concerning 2nd Embodiment. It is a conceptual diagram which shows another example of the heat processing sequence in RTP applied to the heat processing method of the silicon wafer concerning 2nd Embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
(First embodiment)
The silicon wafer heat treatment method according to the present embodiment includes a step of cleaning the surface (polished surface) of a silicon wafer on which at least a surface on which a semiconductor device is formed is mirror-polished with a hydrogen fluoride-based solution, and the cleaned silicon After the wafer was rapidly heated and held in a first temperature range of 900 ° C. or higher and 1250 ° C. or lower in an ammonia-based gas atmosphere, a first RTP for rapidly lowering the temperature was performed, and the first RTP was performed. Performing a second RTP that rapidly lowers the temperature of the silicon wafer after rapidly heating and holding the silicon wafer in a second temperature range of 1300 ° C. to 1400 ° C. in an inert gas atmosphere.

As described above, by further performing RTP on the silicon wafer cleaned with the hydrogen fluoride-based solution in an ammonia-based gas atmosphere, nitrogen can also be terminated at the silicon atoms on the polished surface. Therefore, the bonding force between the silicon atoms on the polished surface and hydrogen can be further increased.
Therefore, even at a high temperature of 1300 ° C. or higher and 1400 ° C. or lower, the bond is hardly broken and a stable state can be obtained.
Therefore, even if other reactive gases (such as nitrogen) are present in the atmosphere, the bonding between the silicon atoms and the reactive gas can be suppressed, so the deterioration of the surface roughness of the polished surface of the wafer is suppressed. can do. In addition, even when a small amount of oxygen is contained in the atmosphere, the reaction between oxygen and silicon atoms can be suppressed, so that generation of concave pits can also be suppressed.

The mirror-polished silicon wafer is manufactured by cutting out from a silicon single crystal ingot grown mainly by the Czochralski method (hereinafter referred to as CZ method).
The silicon single crystal ingot is grown by the CZ method by a known method.
Specifically, a silicon single crystal ingot heats a silicon raw material filled in a quartz crucible to form a silicon melt, contacts the seed crystal with the liquid surface of the silicon melt, and rotates the seed crystal and the quartz crucible. The seed crystal can be pulled up to grow the neck portion, the crown portion, and the straight body portion on the seed crystal, and then separated from the silicon melt.
Next, the grown silicon single crystal ingot is cut out by a known method and processed into a silicon wafer in which at least the surface on which the semiconductor device is formed is mirror-polished.
Specifically, a straight body portion of a silicon single crystal ingot is cut into a wafer shape with an inner peripheral blade or a wire saw, and processing such as chamfering, lapping, etching, and mirror polishing of the outer peripheral portion is performed.

  The step of cleaning with the hydrogen fluoride-based solution is performed by a known method (a method of immersing at least a polished surface of a wafer in the solution, a method of spin-coating the solution on the polished surface while rotating the wafer, or the like). be able to. Preferably, the method of immersing the polished surface of the wafer is suitable in terms of productivity and cost.

The cleaning with the hydrogen fluoride-based solution is preferably performed by bringing the polished surface of the wafer into contact with the solution for 1 minute to 10 minutes.
By setting it as such washing | cleaning time, the silicon atom of the said grinding | polishing surface can be terminated with hydrogen efficiently, suppressing the fall of productivity.

The hydrogen fluoride-based solution mainly includes a hydrofluoric acid solution (HF) and a buffered HF solution (NH ₄ F + HF).

FIG. 1 is a conceptual cross-sectional view showing an example of an RTP apparatus applied to a silicon wafer heat treatment method according to the present invention.
An RTP apparatus 10 shown in FIG. 1 contains a reaction chamber 20 for accommodating a wafer W and performing a heat treatment, a wafer holding unit 30 provided in the reaction chamber 20 for holding the wafer W, and heating for heating the wafer W. Unit 40. In a state where the wafer W is held by the wafer holder 30, the first space 20 a that is a space surrounded by the inner wall of the reaction chamber 20 and the surface (device formation surface) W 1 side of the wafer W, and the inner wall of the reaction chamber 20 And a second space 20b that is a space surrounded by the back surface W2 side of the wafer W facing the front surface W1 side.

The reaction chamber 20 has a supply port 22 for supplying atmospheric gas F _A (solid arrow) into the first space 20a and the second space 20b, and the supplied atmospheric gas F _A from the first space 20a and the second space 20b. And a discharge port 26 for discharging. The reaction chamber 20 is made of, for example, quartz.

  The wafer holding unit 30 includes a susceptor 32 that holds the outer peripheral portion of the back surface W2 of the wafer W in a ring shape, and a rotating body 34 that holds the susceptor 32 and rotates the susceptor 32 about the center of the wafer W. The susceptor 32 and the rotating body 34 are made of, for example, SiC.

  The heating unit 40 is disposed outside the reaction chamber 20 above the front surface W1 and below the back surface W2 of the wafer W held by the wafer holding unit 30, and heats the wafer W from both sides. The heating unit 40 is composed of, for example, a plurality of halogen lamps 50.

When RTP is performed using the RTP apparatus 10 shown in FIG. 1, the wafer W is introduced into the reaction chamber 20 from a wafer introduction port (not shown) provided in the reaction chamber 20, and the outer periphery of the back surface W <b> 2 of the wafer W is introduced. part was held on the susceptor 32 of the wafer holder 30 in a ring shape, it supplies the atmospheric gas F _a, while rotating the wafer W, carried out by heating the wafer W by the heating unit 40.

The first temperature range in which the temperature is rapidly raised and maintained in the ammonia-based gas atmosphere is preferably 900 ° C. or higher and 1250 ° C. or lower.
When the first temperature range is less than 900 ° C., it is difficult to terminate nitrogen on the silicon atoms on the polished surface, so that deterioration of the surface roughness of the polished surface is suppressed in the subsequent second RTP. Difficult to do. When the first temperature range exceeds 1250 ° C., the polished surface of the wafer is etched by the ammonia-based gas, and the surface roughness is deteriorated.
The ammonia-based gas includes ammonia gas (NH ₃ ) and hydrazine (H ₂ NNH ₂ ).

The RTP in the ammonia-based gas atmosphere is charged into the first temperature range by introducing the cleaned silicon wafer into the reaction chamber 20 of the RTP apparatus 10 as shown in FIG. 1 held in a temperature range of 400 ° C. or less. It is preferable to rapidly raise the temperature.
By introducing in such a temperature range, in the first RTP, it is possible to suppress the occurrence of slip due to a rapid temperature change at the time of introduction into the reaction chamber 20 while suppressing a decrease in productivity. .

FIG. 2 is a conceptual diagram showing an example of a first RTP heat treatment sequence applied to the silicon wafer heat treatment method according to the present embodiment.
As shown in FIG. 2, the heat treatment sequence used for the first RTP is such that at least a semiconductor device is placed in the reaction chamber 20 of the RTP apparatus 10 shown in FIG. 1 held at a temperature T0 (for example, 200 ° C.). The surface W1 side to be formed is mirror-polished and a wafer W cleaned with a hydrogen fluoride-based solution is placed, and ammonia-based gas is supplied into the first space 20a and the second space 20b.

Next, the temperature is rapidly increased from a temperature T0 (° C.) to a first temperature range of 900 ° C. to 1250 ° C. (temperature T1 (° C.)) at a temperature increase rate ΔTu1 (° C./sec). Then, for example, the temperature is rapidly lowered at a temperature drop rate ΔTd1 (° C./second) to a temperature T0 (° C.).
The temperatures T0 and T1 were measured by a radiation thermometer (not shown) installed below the wafer holder 30 when the wafer W was installed in the reaction chamber 20 of the RTP apparatus 10 as shown in FIG. It is the surface temperature of the wafer W (the average temperature when a plurality of radiation thermometers are arranged in the radial direction of the wafer W).

The second RTP is preferably performed in an inert gas atmosphere.
By performing RTP in such an atmosphere, the annihilation power of void defects can be increased.
When the second RTP is performed in an ammonia gas atmosphere, the polished surface of the wafer is etched by the ammonia gas, and the surface roughness is deteriorated.
Further, when the second RTP is performed in a reducing gas atmosphere such as hydrogen, oxygen easily diffuses out from the surface W1, so that the oxygen concentration in the surface layer portion of the wafer serving as the device active region decreases. Slip is likely to occur in the heat treatment when forming the semiconductor device later.
Argon gas (Ar) is preferably used as the inert gas.

It is preferable that the second RTP is rapidly heated and held in a second temperature range of 1300 ° C. or higher and 1400 ° C. or lower.
When the second temperature range is less than 1300 ° C., the void defect extinction power is reduced. When the second temperature range exceeds 1400 ° C., the silicon wafer is close to the melting point of silicon, so that the silicon wafer may be softened or melted.
The second temperature range is more preferably 1300 ° C. or higher and 1380 ° C. or lower from the viewpoint of device life as an RTP device used for performing the RTP.

In the second RTP, a silicon wafer subjected to the first RTP is put into a reaction chamber 20 of an RTP apparatus 10 as shown in FIG. It is preferable to rapidly raise the temperature to a temperature range.
By introducing in such a temperature range, in the second RTP, it is possible to suppress the occurrence of slip due to a rapid temperature change at the time of introduction into the reaction chamber 20 while suppressing a decrease in productivity. .

FIG. 3 is a conceptual diagram showing an example of a second RTP heat treatment sequence applied to the silicon wafer heat treatment method according to the present embodiment.
As shown in FIG. 3, the heat treatment sequence used for the second RTP is such that at least a semiconductor device is placed in the reaction chamber 20 of the RTP apparatus 10 shown in FIG. 1 held at a temperature T0 (for example, 200 ° C.). The surface W1 side to be formed is mirror-polished, the wafer W subjected to the first RTP is placed, and an inert gas is supplied into the first space 20a and the second space 20b.

Next, the temperature is rapidly increased from the temperature T0 (° C.) to a second temperature range of 1300 ° C. to 1400 ° C. (temperature T2 (° C.)) at a temperature increase rate ΔTu2 (° C./second), and then the temperature T2 ( After the temperature is kept constant for a predetermined time t2 (seconds), the temperature is rapidly lowered to a temperature T0 (° C) at a temperature drop rate ΔTd2 (° C / second), for example.
The temperature T2 is measured by a radiation thermometer (not shown) installed below the wafer holder 30 when the wafer W is installed in the reaction chamber 20 of the RTP apparatus 10 as shown in FIG. Surface temperature (average temperature when a plurality of radiation thermometers are arranged in the radial direction of the wafer W).

The temperature increase rates ΔTu1 and ΔTu2 in the first and second RTP are preferably 10 ° C./second or more and 150 ° C./second or less.
By setting it as such a temperature increase rate, in the said RTP, generation | occurrence | production of the contact trace and slip by the rapid temperature change at the time of rapid temperature increase can be suppressed, suppressing the fall of productivity.

The temperature drop rates ΔTd1 and ΔTd2 in the first and second RTP are preferably 10 ° C./second or more and 150 ° C./second or less.
By setting it as such a temperature fall rate, generation | occurrence | production of the contact trace and slip by the rapid temperature change at the time of rapid temperature fall can be suppressed in said RTP, suppressing the fall of productivity.

The holding time t1 in the temperature range (temperature T1 (° C.)) of 900 ° C. or more and 1250 ° C. or less in the first RTP is preferably 1 second or more and 10 seconds or less.
By setting such a holding time t1, it is possible to efficiently terminate the silicon atoms on the polished surface with nitrogen while suppressing a decrease in productivity.
The holding time t2 in the temperature range (temperature T2 (° C.)) of 1300 ° C. to 1400 ° C. in the second RTP is preferably 1 second to 30 seconds.
By setting the holding time t2 as described above, it is possible to efficiently eliminate void defects while suppressing a decrease in productivity.

(Second Embodiment)
The silicon wafer heat treatment method according to the present embodiment includes a step of cleaning at least the surface of a silicon wafer on which a surface on which a semiconductor device is formed is mirror-polished with a hydrogen fluoride-based solution; In the system gas atmosphere, the temperature is rapidly increased to a first temperature range of 900 ° C. or more and 1250 ° C. or less, and the ammonia-based gas atmosphere is switched to an inert gas atmosphere in the first temperature range. And a step of performing a rapid heating / cooling heat treatment in which the temperature is rapidly raised and held in a second temperature range of not higher than ° C. and then rapidly lowered.
In other words, the silicon wafer heat treatment method according to this embodiment is different in that the first RTP and the second RTP in the first embodiment are performed by one continuous RTP.
Since others are the same as that of 1st Embodiment, description is abbreviate | omitted.

  Since the silicon wafer heat treatment method according to the present embodiment has the above-described configuration, in addition to the effects of the first embodiment described above, the number of steps for performing the first RTP is reduced by one. Therefore, productivity can be improved and costs can be reduced.

FIG. 4 is a conceptual diagram showing an example of a heat treatment sequence in RTP applied to the silicon wafer heat treatment method according to the present embodiment.
In the heat treatment sequence used for the RTP, as shown in FIG. 4, at least a semiconductor device is formed in the reaction chamber 20 of the RTP apparatus 10 shown in FIG. 1 held at a temperature T0 (for example, 200 ° C.). A wafer W whose surface W1 side is mirror-polished and further cleaned with a hydrogen fluoride-based solution is installed, and ammonia-based gas is supplied into the first space 20a and the second space 20b.
Next, nitrogen is terminated by rapidly raising the temperature from a temperature T0 (° C.) to a first temperature range of 900 ° C. to 1250 ° C. (temperature T1 (° C.)) at a temperature increase rate ΔTu2 (° C./sec). . Subsequently, the ammonia-based gas atmosphere is switched to an inert gas atmosphere in the first temperature range (temperature T1 (° C.)) and supplied into the first space 20a and the second space 20b.
Next, from the first temperature range (temperature T1 (° C.)) to a second temperature range (temperature T2 (° C.)) of 1300 ° C. or more and 1400 ° C. or less, the temperature rise rate ΔTu2 (° C./second) is further increased. The temperature is raised and held constant for a predetermined time t2 (seconds) in the second temperature range (temperature T2 (° C)), and then rapidly, for example, at a temperature drop rate ΔTd2 (° C / second) until the temperature T0 (° C). Lower the temperature.

FIG. 5 is a conceptual diagram showing another example of the heat treatment sequence in RTP applied to the silicon wafer heat treatment method according to this embodiment.
As shown in FIG. 5, the switching from the ammonia-based gas atmosphere to the inert gas atmosphere is preferably performed in a state of being kept constant in the first temperature range (temperature T1 (° C.)).
That is, at least the surface W1 side on which the semiconductor device is formed in the reaction chamber 20 of the RTP apparatus 10 as shown in FIG. 1 held at a temperature T0 (for example, 200 ° C.) is mirror-polished, and further, a hydrogen fluoride solution The wafer W cleaned by the above is installed, and ammonia gas is supplied into the first space 20a and the second space 20b.

  Next, the temperature is rapidly raised from a temperature T0 (° C.) to a first temperature range of 900 ° C. to 1250 ° C. (temperature T1 (° C.)) at a temperature increase rate ΔTu2a (° C./sec), and the first temperature Nitrogen is terminated by holding within a range (temperature T1 (° C.)) for a predetermined time (t1a (seconds)). Subsequently, the ammonia-based gas atmosphere is continuously switched to an inert gas atmosphere in the first temperature range (temperature T1 (° C.)), and is kept constant for a predetermined time (t1b (seconds)). The temperature is further rapidly increased to a second temperature range (temperature T2 (° C.)) of 1300 ° C. to 1400 ° C. at a temperature increase rate ΔTu2b (° C./second), and the second temperature range (temperature T2 (° C.)). Then, the temperature is kept constant for a predetermined time t2 (seconds), and then, for example, the temperature is rapidly lowered to a temperature T0 (° C) at a temperature drop rate ΔTd2 (° C / second).

  When the silicon wafer heat treatment method according to the present embodiment is performed in a heat treatment sequence as shown in FIG. 5, although productivity is slightly reduced, more nitrogen may be terminated than in the heat treatment sequence as shown in FIG. 4. Furthermore, it becomes easy to completely discharge the ammonia-based gas from the reaction chamber 20 when the gas is switched. Therefore, since the risk of the wafer polishing surface being exposed to the ammonia-based gas at a high temperature exceeding 1250 ° C. is reduced, deterioration of the surface roughness of the polishing surface can be suppressed.

The holding time (t1a (second)) for maintaining the first temperature range (temperature T1 (° C.)) in the ammonia-based gas atmosphere is 1 second or more and 5 seconds or less, and after the switching, the inert gas In the atmosphere, the holding time (t1b (second)) for holding the first temperature range (temperature T1 (° C.)) is preferably 1 second or more and 5 seconds or less.
By setting it as such holding time, nitrogen can be terminated efficiently, suppressing the fall of productivity.

The temperature increase rates ΔTu2a and ΔTu2b in the RTP are preferably 10 ° C./second or more and 150 ° C./second or less.
By setting it as such a temperature increase rate, in the said RTP, while suppressing the fall of productivity, generation | occurrence | production of the contact trace and slip by the rapid temperature change at the time of rapid temperature increase can be suppressed.
Note that the temperature increase rates ΔTu2a and ΔTu2b may be the same temperature increase rate or different temperature increase rates as long as they are 10 ° C./second or more and 150 ° C./second or less.

  In the first embodiment and the second embodiment, from the viewpoint of efficiently terminating nitrogen on the polishing surface, as shown in the first embodiment, the first RTP and the second RTP are combined. It is preferable to perform them individually. By setting it as such a structure, the surface roughness after 2nd RTP can be improved.

EXAMPLES Hereinafter, although this invention is demonstrated further more concretely based on an Example, this invention is not limitedly interpreted by the following Example.
(Test 1)
Controlling v / G (v: pulling speed, G: temperature gradient in the pulling axis direction in the single crystal) by the CZ method to produce a silicon single crystal ingot having a region in which vacant point defects exist, A silicon wafer (diameter 300 mm, surface orientation {100}, thickness 775 μm, oxygen concentration 1.2 to 1.3 × 10 ¹⁸ atoms / cm ³ ) obtained by cutting out from the region and having both surfaces mirror-polished The entire wafer was immersed in a 5% hydrofluoric acid solution, washed for a certain time (5 minutes), then washed with pure water and dried.
Next, the first RTP in which the temperature T1 (° C.) is changed in the heat treatment sequence as shown in FIG. 2 is applied to the cleaned sample using the RTP apparatus 10 as shown in FIG. went.
Specifically, the cleaned wafer is put into a reaction chamber maintained at 200 ° C., ammonia gas (NH ₃ ) is supplied as an atmosphere, a temperature increase rate (ΔTu1) is 75 ° C./second, and a temperature T1 ( ° C) was changed to 800 ° C, 950 ° C, 1100 ° C, 1250 ° C, 1300 ° C, and each temperature was rapidly increased, and each temperature T1 (° C) was held for 10 seconds, and then the temperature decrease rate (ΔTd1) was 90 ° C / second The temperature was rapidly lowered to 200 ° C. (first RTP).

Next, for each sample subjected to the first RTP, the temperature T2 (° C.) is further changed by the heat treatment sequence as shown in FIG. 3 using the RTP apparatus 10 as shown in FIG. The second RTP was performed to produce an annealed wafer.
Specifically, the wafer subjected to the first RTP is put into a reaction chamber held at 200 ° C., and an argon gas (Ar) is supplied as an atmosphere, and a temperature increase rate (ΔTu 2) is 75 ° C./second. The temperature T2 (° C.) was changed rapidly to 1250 ° C., 1300 ° C., and 1350 ° C., respectively. The temperature T2 (° C.) was maintained for 15 seconds, and then the temperature decrease rate (ΔTd2) was 90 ° C./sec. (2nd RTP). The argon gas (Ar) is known to contain a very small amount of oxygen of 0.1 ppm or less as measured by an oxygen concentration meter installed in the reaction chamber.

RMS (measurement range: 3 μm × 3 μm) of the surface roughness of the obtained annealed wafer on the semiconductor device formation surface was evaluated using an AFM (Atomic Force Microscope).
In addition, the occurrence of concave pits on the semiconductor device formation surface was evaluated from the AFM image.
Furthermore, regarding the defect density in the surface layer part from the wafer surface to a depth of 5 μm, an LSTD scanner (Laser Scattering Topology Defect)
(Scanner) at a wavelength of 680 nm.
Further, as a reference example, the surface roughness on the semiconductor device forming surface of the wafer before cleaning was similarly evaluated by RMS (measurement range: 3 μm × 3 μm) using AFM.
Table 1 shows test conditions and evaluation results in this test.

As shown in Table 1, when the temperature T1 is 800 ° C. or 1300 ° C. (Comparative Examples 1 to 3, Comparative Examples 7 to 9), a tendency that the surface roughness after the second RTP deteriorates is recognized. Further, when the temperature T1 is 800 ° C. (Comparative Examples 1 to 3), generation of concave pits is observed. Furthermore, when temperature T2 is 1250 degreeC (comparative examples 1, 4, 5, 6, and 7), the tendency for a defect density to become high is recognized.
Therefore, by setting the temperature T1 to 900 ° C. or more and 1250 ° C. or less and further setting the temperature T2 to 1300 ° C. or more, deterioration of the surface roughness can be suppressed, concave pits are not generated, and defects It can be seen that the density is low.

(Test 2)
With respect to each of the cleaned samples, the temperature T1 (° C.) and the temperature T2 (° C.) are set to the test 1 using the RTP apparatus 10 as shown in FIG. An annealed wafer was produced by performing RTP under the same conditions. Since others are the same as those in Test 1, description thereof is omitted.
Specifically, the cleaned wafer is put into a reaction chamber maintained at 200 ° C., and ammonia gas (NH ₃ ) is supplied as an atmosphere, and the temperature T1 is 800 ° C., 950 ° C., 1100 ° C., 1250 ° C., The temperature is changed to 1300 ° C., and the temperature is rapidly increased to a temperature T1 at a temperature increase rate (ΔTu2) of 75 ° C./second. Thereafter, the atmosphere is changed from ammonia gas (NH ₃ ) to argon gas (Ar) at temperature T1. The temperature T2 was further changed to 1250 ° C., 1300 ° C., and 1350 ° C., and each temperature was rapidly increased to a temperature T 2 at a temperature increase rate (ΔTu 2) of 75 ° C./second, and held at the temperature T 2 for 15 seconds. The temperature was rapidly lowered to 500 ° C. at a temperature drop rate (ΔTd 2) of 90 ° C./second. The argon gas (Ar) is known to contain a very small amount of oxygen of 0.1 ppm or less as measured by an oxygen concentration meter installed in the reaction chamber.
The surface roughness on the semiconductor device formation surface of the obtained annealed wafer, the occurrence of concave pits, and the defect density were measured in the same manner as in Test 1.
Table 2 shows the evaluation results in Test 2.

  As shown in Table 2, in Test 2, the surface roughness tends to be slightly rough (RMS is about 0.01 to 0.03 nm) compared to Test 1, but Test 2 also has Test 1 A similar tendency is observed. That is, when temperature T1 is 800 degreeC or 1300 degreeC (Comparative Examples 10-12, Comparative Examples 16-18), the tendency for the surface roughness after 2nd RTP to get worse is recognized. Further, when the temperature T1 is 800 ° C. (Comparative Examples 10 to 12), generation of concave pits is observed. Furthermore, when the temperature T2 is 1250 ° C. (Comparative Examples 10, 13, 14, 15, 16), a tendency that the defect density tends to increase is recognized.

  Therefore, in Test 2, as in Test 1, the temperature T1 is set to 900 ° C. or more and 1250 ° C. or less, and further, the temperature T2 is set to 1300 ° C. or more. Pits are not generated and the defect density is low.

10 RTP apparatus 20 Reaction chamber 30 Wafer holding part 40 Heating part

Claims (2)

Cleaning at least the surface of the silicon wafer on which the surface on which the semiconductor device is formed is mirror-polished with a hydrogen fluoride-based solution;
A step of performing a first rapid heating / cooling heat treatment in which the cleaned silicon wafer is rapidly heated and held in a first temperature range of 900 ° C. or higher and 1250 ° C. or lower in an ammonia-based gas atmosphere;
The second rapid heating / cooling heat that rapidly cools after the temperature of the silicon wafer subjected to the first rapid heating / cooling heat treatment is rapidly raised and maintained in a second temperature range of 1300 ° C. to 1400 ° C. in a rare gas atmosphere. And a step of performing a treatment. A method for heat treating a silicon wafer. Cleaning at least the surface of the silicon wafer on which the surface on which the semiconductor device is formed is mirror-polished with a hydrogen fluoride-based solution;
The cleaned silicon wafer is rapidly heated to a first temperature range of 900 ° C. or higher and 1250 ° C. or lower in an ammonia-based gas atmosphere, and the ammonia-based gas atmosphere is switched to a rare gas atmosphere in the first temperature range. And a step of performing a rapid temperature raising and lowering heat treatment in which the temperature is rapidly raised and held in a second temperature range of 1300 ° C. or higher and 1400 ° C. or lower and then rapidly lowered, and then a silicon wafer heat treatment method.

Images