JP2010153889A - Method for manufacturing semiconductor device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a flash lamp annealing method without causing any damage on an Si wafer. <P>SOLUTION: The method for manufacturing a semiconductor device includes: a process for arranging a flash lamp light source 5 having a plurality of flash lamps, above the Si wafer 1 having a single crystal Si region; a process including a first heat treatment and a second heat treatment of heating the Si wafer by heat radiated from the light source, wherein the Si wafer is heated in each of the first and second heat treatments so as to form such a light intensity distribution on the Si wafer by the light that the intensity has a maximum value in a direction differing from that of a crystal orientation of the single crystal Si region, and an array direction of the plurality of flash lamps in the first heat treatment is different from that in the second heat treatment. The crystal orientation 9 of the single crystal Si region is a cleavage plane orientation of the single crystal Si region. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a semiconductor device and a method for manufacturing the same, and more particularly, to a method for manufacturing a semiconductor device in which damage caused in a substrate due to light heating is reduced.

  Improvement of LSI performance has been achieved by increasing the degree of integration, that is, by miniaturizing elements constituting the LSI. As device dimensions are reduced, parasitic resistance and short channel effects increase. Therefore, the formation of a low-resistance and shallow pn junction has become increasingly important.

  A shallow impurity diffusion region can be formed by optimizing ion implantation with low acceleration energy and subsequent annealing. On the other hand, in order to reduce the resistance of the impurity diffusion region, it is necessary to perform annealing for activating impurity ions at a high temperature.

  As impurity ions to be ion-implanted, boron (B) ions, phosphorus (P) ions, or arsenic (As) ions are used. These impurity ions have a large diffusion coefficient in silicon (Si). Therefore, in RTA (Rapid Thermal Anneal) using a halogen lamp, inward diffusion and outward diffusion of impurity ions occur, and it becomes increasingly difficult to form a shallow impurity diffusion region.

  The inward diffusion and outward diffusion can be suppressed by lowering the annealing temperature. However, when the annealing temperature is lowered, the activation rate of impurity ions is greatly reduced. Therefore, it is difficult to form a shallow impurity diffusion region even if a technique of lowering the annealing temperature is employed.

  As described above, it is difficult to form a low-resistance and shallow (20 nm or less) impurity diffusion region by the RTA process using a conventional halogen lamp.

  Therefore, in recent years, a flash lamp annealing method using a flash lamp in which a rare gas such as xenon (Xe) is enclosed has been studied as a method for instantaneously supplying energy necessary for activating impurity ions. .

  The 1/2 pulse width of the flash lamp is about 10 milliseconds. For this reason, when the flash lamp annealing method is used, the time during which the wafer surface is kept at a high temperature is extremely short, so that impurity ions implanted into the wafer surface hardly diffuse. Therefore, the impurity ions can be activated with almost no change in the distribution of the impurity ions implanted on the wafer surface.

  However, the conventional flash lamp annealing method has the following problems.

In order to obtain a sufficient annealing effect, a large power strength of 20 J / cm ² or more is required. In addition, a rapid temperature rise occurs on the wafer surface. As a result, a temperature difference is generated between the front side and the back side of the wafer, and thermal stress increases inside the wafer. Such an increase in thermal stress causes damage (thermal damage) such as slip and crack (breakage) in the wafer. Such thermal damage to the wafer causes a decrease in production yield.

  Thermal damage to the wafer can be avoided by lowering the temperature at which the wafer is preheated or the irradiation energy density of the flash lamp before the flash lamp is turned on. However, in that case, sufficient activation of impurity ions cannot be expected.

  As described above, in the conventional flash lamp annealing method, it is difficult to sufficiently activate the impurity ions implanted into the wafer surface without causing damage such as slip and crack (breakage) in the wafer. Therefore, at present, it is difficult to form a shallow impurity diffusion region without damaging the wafer.

  The present invention has been made in view of the above circumstances, and its object is to easily damage the substrate when the substrate is heated by light and the impurity ions in the substrate are activated. An object of the present invention is to provide a method of manufacturing a semiconductor device that can be prevented.

  A method of manufacturing a semiconductor device according to an aspect of the present invention includes a step of disposing a flash lamp light source including a plurality of flash lamps above a Si wafer including a single crystal Si region, and the Si emitted by the light emitted from the light source. A process including a first heat treatment and a second heat treatment for heating the wafer, wherein each of the first heat treatment and the second heat treatment is formed on the Si wafer by the light. The Si wafer is heated such that the light intensity distribution is such that a distribution having a maximum intensity in a direction different from the crystal orientation of the single crystal Si region, and the first heat treatment is performed in the first heat treatment. The step of arranging the plurality of flash lamps and the step of changing the arrangement direction of the plurality of flash lamps during the second heat treatment; The crystal orientation of the region, wherein said a surface orientation of the cleavage plane of the single-crystal Si region.

  According to the present invention, when a substrate is heated by light to activate impurity ions in the substrate, it is possible to realize a method for manufacturing a semiconductor device that can easily prevent damage generated in the substrate. Become.

Sectional drawing which shows the manufacturing process of the semiconductor device which concerns on the 1st Embodiment of this invention. The top view which shows the relationship between several flash lamps and the crystal orientation of Si wafer at the time of the heating process of 1st Embodiment The top view which shows the relationship between the crystal orientation of several flash lamps and Si wafer at the time of the conventional heating process The figure which shows the process window of 1st Embodiment regarding board | substrate preheating temperature and irradiation energy density The figure which shows the process window of the comparative example regarding the substrate preheating temperature and irradiation energy density Sectional drawing which shows the manufacturing process of the semiconductor device which concerns on the 2nd Embodiment of this invention. The top view which shows the relationship between a lamp arrangement | positioning direction, a crystal orientation, and a line pattern arrangement | sequence direction at the time of the heating process of 2nd Embodiment. Sectional drawing which shows the manufacturing process of the MOS transistor which concerns on the 3rd Embodiment of this invention. Sectional drawing which shows the manufacturing process of the semiconductor device which concerns on the 4th Embodiment of this invention. The figure which shows the relationship between the uniformity of the intensity | strength of the light in the surface of Si wafer, and irradiation distance L in 4th Embodiment. The figure which shows the relationship between ripple width / average light intensity and irradiation distance in 4th Embodiment The figure which shows the light emission waveform of the flash lamp of this embodiment and a comparative example The figure which shows the temperature distribution of the thickness direction of Si wafer in each time after the flash lamp irradiation in this embodiment The figure which shows the temperature distribution of the thickness direction of Si wafer in each time after the flash lamp irradiation in a comparative example The figure for demonstrating the stress which generate | occur | produces in the thickness direction of Si wafer in each time after the flash lamp irradiation in this embodiment The figure for demonstrating the stress which generate | occur | produces in the thickness direction of Si wafer in each time after flash lamp irradiation in a comparison The figure which shows the light emission waveform of another comparative example (2nd comparative example). Diagram showing brittle fracture curve of Si wafer The figure which shows the process window of 5th Embodiment regarding board | substrate preheating temperature and irradiation energy density The figure which shows the process window of the comparative example regarding the substrate preheating temperature and irradiation energy density The figure which shows the relationship between the gate leakage current of the MOS capacitor comprised by the impurity diffusion area | region of 7th Embodiment, and pulse width The figure which shows the relationship between the junction leakage current of the pn junction comprised by the impurity diffusion area | region of 7th Embodiment, and pulse width The figure which shows the relationship between the yield of a device comprised by the impurity diffusion area | region of 7th Embodiment, and pulse width

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a cross-sectional view showing a manufacturing process of a semiconductor device according to the first embodiment of the present invention.

  First, as shown in FIG. 1A, a mask 2 for ion implantation is arranged on a single crystal Si wafer (Si substrate) 1, and impurity ions 3 are placed on the mask 2 by a well-known ion implantation apparatus (not shown). Into the Si wafer 1.

  Here, the Si wafer 1 may be a normal bulk Si wafer or an SOI wafer. Further, a wafer (substrate) whose main component is a semiconductor material other than silicon such as silicon germanium may be used. The mask 2 may be a normal mask or a stencil mask. As the impurity ions 3, for example, boron (B) ions are used as p-type impurities, and phosphorus (P) ions or arsenic (As) ions are used as n-type impurities.

  Next, as shown in FIG. 1B, the flash lamp light source 5 is placed in a state where the Si wafer 1 is placed on the hot plate 4 and the Si wafer 1 is heated from the back side (preheated state). The Si wafer 1 is heated from the surface side by the light 6 emitted from (by flash lamp annealing).

  By the heating process, the impurity ions 3 are activated, and a shallow impurity diffusion region 7 of 20 nm or less is formed on the surface of the Si wafer 1 as shown in FIG.

  The heating temperature (substrate preheating temperature) of the Si wafer 1 by the hot plate 4 is set to 400 ° C., for example. The heating temperature is not limited to 400 ° C., and may be in the range of 300 to 600 ° C. If it is the said temperature range, the impurity diffusion area | region 7 which has a desired density | concentration profile can be formed easily. The substrate preheating temperature is generally set lower than the heating temperature of the Si wafer 1 by the flash lamp light source 5.

  Here, the hot plate 4 (resistance heating means) is used to heat the Si wafer 1 from the back surface (for preheating), but heating means (light heating) such as a halogen lamp which is one of infrared lamps. Means) may be used.

The flash lamp light source 5 includes a plurality of flash lamps filled with a rare gas such as Xe gas. The energy of the light 6 emitted from the flash lamp light source 5 is, for example, 35 J / cm ² . The energy of the light 6 is not limited to 35 J / cm ² and may be 60 J / cm ² or less.

  FIG. 2 is a plan view showing the relationship between the plurality of flash lamps 8 constituting the flash lamp light source 5 and the crystal orientation 9 of the Si wafer 1 during the heating process. The crystal plane of the Si wafer 1 is, for example, (100), and the crystal orientation 9 is the plane orientation of the cleavage plane of the Si wafer 1, for example, <011>.

  As can be seen from FIG. 2, in this embodiment, the positional relationship between the flash lamp light source 5 and the Si wafer 1 is set so that the arrangement direction (lamp arrangement direction) 10 of the plurality of flash lamps 8 and the crystal orientation 9 are different. Is done. In the present embodiment, the longitudinal direction of the plurality of flash lamps 8 is also set to a direction different from the crystal orientation 9, but this is not always necessary.

  On the other hand, in the conventional flash lamp annealing method (comparative example), as shown in FIG. 3, the positional relationship between the flash lamp light source 5 and the Si wafer 1 is that the lamp array direction 10 (longitudinal direction of the flash lamp 8) and the crystal orientation. 9 is set to be parallel.

  FIG. 4 shows a process window of the present embodiment regarding the heating temperature (substrate preheating temperature) of the Si wafer 1 by the hot plate 4 and the irradiation energy density on the Si wafer 1 by the flash lamp light source 5. FIG. 5 shows a comparative process window with respect to substrate preheating temperature and irradiation energy density.

  The higher the substrate preheating temperature, the lower the irradiation energy density required for activating the impurities, but at the same time, the irradiation energy density at which thermal damage (slip, crack) occurs in the Si wafer 1 also decreases.

  Here, when the present embodiment and the comparative example are compared with respect to the irradiation energy density at which thermal damage occurs to the Si wafer 1, from FIG. 4 and FIG. 5, the present embodiment is better at any substrate preheating temperature. You can see that it is higher. That is, it became clear that the process window of this embodiment is wider than that of the comparative example.

  In general, in the case of lamp heating, the light intensity tends to be high immediately below the lamp, and the light intensity tends to be low between the lamps. For this reason, a temperature difference occurs between the lamp and the lamp. Due to such a temperature difference, thermal stress is generated in the Si wafer 1.

  In the case of the comparative example, it is considered that the thermal stress is generated along the crystal orientation where the substrate strength is weak. This is considered to be the reason why thermal damage is likely to occur in the Si wafer 1 in the comparative example.

  On the other hand, in the case of the positional relationship between the Si wafer 1 and the flash lamp light source 5 of the present embodiment, the light intensity distribution formed on the Si wafer 1 by the light 6 has the maximum intensity in a direction different from the crystal orientation of the Si wafer 1. A light intensity distribution is formed.

  As a result, the direction in which the thermal stress is generated deviates from the crystal orientation of the Si wafer 1. Thereby, in this embodiment, it is thought that board | substrate intensity | strength was securable and the tolerance of Si wafer with respect to thermal damages, such as a slip and a crack, improved.

  As described above, according to the present embodiment, when flash lamp annealing is performed, the thermal damage generated in the Si wafer 1 is caused by setting the Si wafer 1 and the flash lamp light source 5 in a predetermined positional relationship. Can be easily prevented. That is, the shallow impurity diffusion region 7 can be easily formed in the Si wafer 1 without causing thermal damage.

(Second Embodiment)
FIG. 6 is a cross-sectional view showing a manufacturing process of a semiconductor device according to the second embodiment of the present invention. Note that parts corresponding to those in the previous figures are denoted by the same reference numerals as in the previous figures, and detailed description thereof is omitted (the same applies to the third and subsequent embodiments).

  First, as shown in FIG. 6A, a film to be processed 11 to be a pattern is formed on the Si wafer 1. The processed film 11 is, for example, an insulating film such as a silicon oxide film, a semiconductor film such as a polycrystalline silicon film, a metal film such as an aluminum film, or a resist such as a photoresist.

  Next, as shown in FIG. 6B, the film 11 to be processed is processed by known photolithography and etching to form a pattern 11p including a plurality of line patterns.

  At this time, the pattern 11p is formed so that the arrangement direction of the plurality of line patterns (line pattern arrangement direction) and the crystal orientation 9 are different. The pattern 11p is, for example, an insulator pattern such as a gate insulating film, a semiconductor pattern such as a polycrystalline silicon gate electrode, a metal pattern such as an aluminum wiring, or a resist pattern.

  Next, as shown in FIG. 6C, the Si wafer 1 is heated by the flash lamp light source 5 and the hot plate 4 as in the first embodiment.

  FIG. 7 is a plan view showing the relationship among the lamp array direction 10, the crystal orientation 9, and the line pattern array direction 12 during the heating process. As can be seen from FIG. 7, in this embodiment, the lamp array direction 10 is set to be different from the crystal orientation 9 and the line pattern array direction 12.

  Since the light absorption coefficient of the Si wafer 1 and the light absorption coefficient of the pattern 11p are different, a temperature difference is generated between the Si wafer 1 and the pattern 11p. As a result, thermal stress is applied to the pattern 11p.

  However, it has been clarified that when the lamp array direction 10 is different from the line pattern array direction 12 and the crystal orientation 9 as in the present embodiment, a temperature difference hardly occurs.

  Therefore, according to the present embodiment, it is possible to prevent thermal damage such as slips and cracks from occurring in the Si wafer 1 during the heating process of the Si wafer 1 including the pattern 11p.

  The lamp array direction 10 and the line pattern array direction 12 may be the same, but different ones can be expected to have a higher thermal damage reduction effect.

(Third embodiment)
FIG. 8 is a cross-sectional view showing a manufacturing process of the MOS transistor according to the third embodiment of the present invention.

  First, in FIG. 8A, a gate insulating film 21 and a gate electrode 22 are formed on a Si wafer 1 by a known method.

  Next, as shown in FIG. 8B, impurity ions 3 are implanted into the surface of the Si wafer 1 using the gate electrode 22 as a mask.

  Next, as shown in FIG. 8C, the Si wafer 1 is heated (first heat treatment) in the same manner as in the process of FIG. 1B of the first embodiment. As a result, the impurity ions 3 are activated, and extension regions (first impurity diffusion regions) 24 are formed on the surface of the Si wafer 1 in a self-aligned manner.

  At this time, similarly to the first embodiment, since the lamp array direction 10 and the crystal orientation 9 are set in different directions, thermal damage such as slips and cracks does not occur in the Si wafer 1, and the process The window also opens.

  Next, as shown in FIG. 8D, a gate sidewall insulating film (spacer) 25 is formed by a well-known method, and then impurity ions (not shown) are formed using the gate sidewall insulating film 25 and the gate electrode 22 as a mask. Is injected into the surface of the Si wafer 1.

  Next, as shown in FIG. 8E, the Si wafer 1 is heated (second heat treatment) as in the step of FIG. 1B of the first embodiment. As a result, the impurity ions are activated, and source / drain regions (second impurity diffusion regions) 26 are formed on the surface of the Si wafer 1.

  At this time, as in the first embodiment, since the lamp array direction 10 and the crystal orientation 9 are set to different directions, damage such as slips and cracks does not occur in the Si wafer 1, and the process window Also spread.

  Further, during the first heat treatment, the Si wafer 1 or the flash lamp light source 5 is rotated, and the lamp arrangement direction 10 during the second heat treatment is different from the arrangement direction of the flash lamps 8 during the first heat treatment. Set to direction. Thereby, the direction of the thermal stress generated in the Si wafer 1 due to the irradiation of the flash lamp during the second heat treatment, and the heat generated in the Si wafer 1 due to the irradiation of the flash lamp during the first heat treatment. The directions of stress are different from each other.

  As a result, the load due to the thermal stress accumulated in the Si wafer 1 is effectively reduced. Thereby, even when performing the heat treatment twice, the resistance against thermal damage such as slips and cracks of the Si wafer 1 can be easily improved.

  Even when the heat treatment is performed three times or more, the above effect can be obtained by setting the lamp arrangement directions 10 in the respective heat treatments to be different from each other.

  In all the heat treatments, the lamp array directions 10 do not necessarily have to be different from each other. For example, when heat treatment is performed three times or more, the first heat treatment and the last heat treatment can be the same. This is because the first heat treatment and the last heat treatment are separated in time, and the influence of the accumulation of these heat treatments is small.

  The multiple-time heat treatment method of the present embodiment is also effective for the formation process (anneal process) of impurity diffusion regions other than the impurity diffusion regions (extension region 24, source / drain region 26) of the MOS transistor. Furthermore, it is also effective for thermal processes other than the impurity ion annealing process.

  When forming a plurality of MOS transistors, it is preferable that the arrangement direction of the plurality of gate electrodes 22 is set to a direction different from the crystal orientation 9 and the lamp arrangement direction 10 (see the second embodiment).

(Fourth embodiment)
FIG. 9 is a cross-sectional view showing a manufacturing process of a semiconductor device according to the fourth embodiment of the present invention.

  First, as shown in FIG. 9A, an ion implantation mask 2 is arranged on the Si wafer 1, and impurity ions 3 are implanted into the Si wafer 1 from the mask 2 by an ion implantation apparatus (not shown).

  Next, as shown in FIG. 9B, the Si wafer 1 is placed on the hot plate 4 and the light 6 emitted from the flash lamp light source 5 in a state where the Si wafer 1 is heated from the back side. The Si wafer 1 is heated from the surface side. By the heating process, the impurity ions 3 are activated and an impurity diffusion region is formed.

  Here, the distance (irradiation distance) L between the Si wafer 1 and the flash lamp light source 5 is set to a value in the range of 23 to 46 mm. The energy of light generated from a plurality of flash lamps constituting the flash lamp light source 5 is substantially at the same level.

  The reason why 23 mm ≦ L ≦ 46 mm is set is as follows.

  FIG. 10 is a diagram illustrating the relationship between the intensity uniformity σ of the light 6 in the plane of the Si wafer 1 and the irradiation distance L. FIG. 11 is a diagram showing the relationship between the ripple width / average light intensity and the irradiation distance L. The ripple width is a difference (D1−D2) between the light intensity D1 on the surface of the Si wafer 1 immediately below the flash lamp and the light intensity D2 on the surface of the Si wafer 1 immediately below between two adjacent flash lamps. The average light intensity is the average intensity of the light 6 on the surface of the Si wafer 1. 10 and 11 show the results when the size of the Si wafer 1 is 8 inches, but similar results were obtained with other sizes.

FIG. 10 reveals that the value of the light intensity in-plane uniformity σ increases when the irradiation distance L is shorter than 23 mm and increases even when the irradiation distance L is longer than 46 mm. Furthermore, FIG. 10 also shows that the value of the light intensity in-plane uniformity σ is 1% or less in the range of 23 mm ≦ L ≦ 46 mm. On the other hand, the value of σ = 1% or less is a sufficient value in terms of specifications. From FIG. 11, the ripple width / average light intensity is sufficiently low at the irradiation distance L = about 30 mm and around 69 mm. However, when the irradiation distance L is around 69 mm, the value of the light intensity in-plane uniformity σ increases as shown in FIG.

  Therefore, by setting the irradiation distance L in the range of 23 mm or more and 46 mm or less, the dependency of the irradiation distance on the light intensity in-plane uniformity σ and the ripple width / average light intensity can be reduced simultaneously. Thereby, the influence of the thermal stress due to the light intensity in-plane uniformity σ and the ripple width / average light intensity is sufficiently reduced, and the resistance of the Si wafer to thermal damage such as slips and cracks is improved.

  Further, when the average yield of a logic circuit composed of MOSFETs using the impurity diffusion region 7 of this embodiment as a source / drain region was examined, the irradiation distance L was set to a range of 23 mm or more and 46 mm or less. % Could be achieved.

  In addition to the above method, as a method for simultaneously reducing the dependency of the light intensity in-plane uniformity σ and the ripple width / average light intensity on the irradiation distance, a light diffusing plate is provided between the Si wafer 1 and the flash lamp light source 5. Alternatively, there is a method of providing a light intensity filter. However, these methods have the following problems.

  The energy of the light 6 emitted from the flash lamp light source 5 is attenuated by the light diffusing plate or the light intensity filter before reaching the Si wafer 1. Therefore, it is necessary to supply the flash lamp light source 5 with electric power (high voltage) larger than that of the present embodiment by an amount corresponding to the amount of energy attenuation caused by the light diffusion plate or the light intensity filter. This deteriorates the pressure resistance of the plurality of flash lamps constituting the flash lamp light source 5 and shortens the life of the flash lamp light source 5.

  On the other hand, according to the present embodiment, since the energy of the light 6 is hardly attenuated, the Si wafer 1 can be efficiently heated. Thereby, it is not necessary to apply a large electric power (high voltage) to the flash lamp light source 5, and it is possible to prevent the flash lamp light source 5 from being shortened.

  Note that the method of the present embodiment and the method of the second embodiment may be combined, or the method of the present embodiment and the method of the third embodiment may be combined.

  In the first to fourth embodiments, an example in which a flash lamp is used as a light source for heat treatment has been shown. However, other lamps such as a halogen lamp can be used, and other than a lamp such as a laser. It is also possible to use a light source.

  For example, when a laser that emits a linear beam is used, the lamp arrangement direction in the first to fourth embodiments is read as the longitudinal direction of the laser beam, so that the same as in the first to fourth embodiments. The same effects as those of the first to fourth embodiments can be obtained.

(Fifth embodiment)
The semiconductor device manufacturing method of this embodiment is different from the conventional method in that the 1/2 pulse width of the flash lamp light source 5 is set to a predetermined value (1 ms) or less.

  First, as shown in FIG. 1A, an ion implantation mask 2 is arranged on the Si wafer 1, and impurity ions 3 are implanted into the Si wafer 1 from above the mask 2 by an ion implantation apparatus (not shown). The

  Next, as shown in FIG. 1B, the light emitted from the flash lamp light source 5 in a state where the Si wafer 1 is placed on the hot plate 4 and the Si wafer 1 is heated from the back side. 6, the Si wafer 1 is heated from the surface side.

  Here, the heating temperature of the Si wafer 1 by the hot plate 4 is 500 ° C. Instead of the hot plate 4, other heating means such as a halogen lamp which is one of infrared lamps may be used. Further, the heating temperature of the Si wafer may be other than 500 ° C.

The energy of the light 6 emitted from the flash lamp light source 5 is, for example, 20 to 40 J / cm ² . The 1/2 pulse width of the flash lamp light source 5 is set to 1 msec or less.

  The flash lamp light source 5 of the present embodiment is originally developed for research, and can be set to a short ½ pulse width of 1 msec or less, which is impossible with a conventional flash lamp light source. . In order to shorten the 1/2 pulse width, the capacitance of the capacitor connected to the circuit of the flash lamp light source may be reduced. For example, if a capacitor of several hundred μF is connected in parallel to the flash lamp, a ½ pulse width of 1 ms or less can be achieved. Actually, 0.7 ms can be achieved at 400 μF.

  FIG. 12 shows a light emission waveform of the light 6 emitted from the flash lamp light source 5 of the present embodiment. In this embodiment, as shown in FIG. 12, the 1/2 pulse width is set to 0.3 milliseconds. FIG. 12 also shows a light emission waveform when the ½ pulse width is set to 3.0 milliseconds as a comparative example.

  By the above heating process, the impurity ions 3 are activated, and a shallow impurity diffusion region 7 of 20 nm or less is formed on the surface of the Si wafer 1 as shown in FIG.

  FIGS. 13 and 14 respectively show the Si wafer 1 at each time after flash lamp irradiation in the present embodiment (1/2 pulse width = 0.3 msec) and the comparative example (1/2 pulse width = 3 msec). It is a figure which shows the temperature distribution of thickness direction.

  In the case of the present embodiment, the surface temperature of the Si wafer 1 rapidly increases after flash lamp irradiation, and reaches a maximum of 1100 ° C. after about 0.3 msec. At this time, the back surface temperature of the Si wafer 1 is governed by the substrate preheating temperature by the hot plate 4. As a result, a temperature difference close to 600 ° C. occurs between the front surface side and the back surface side of the Si wafer 1.

  On the other hand, in the case of the comparative example, the surface temperature of the Si wafer 1 reaches a maximum of 1100 ° C. after about 3 milliseconds after the flash lamp irradiation. The 1/2 pulse width of the comparative example is longer than the 1/2 pulse width of this embodiment. Therefore, the temperature distribution in the wafer thickness direction of the comparative example becomes gentler than that of the present embodiment.

  FIG. 15 and FIG. 16 are diagrams for explaining the stress generated in the thickness direction of the Si wafer 1 at each time after the flash lamp irradiation in this embodiment and the comparative example, respectively, and FIG. FIG. 15B is a diagram schematically showing the stress in the cross section of the Si wafer, in which the stress on the surface of the Si wafer is 0 mm in depth. The substrate preheating temperature in this embodiment and the comparative example is both 500 ° C.

  In each of FIG. 15A and FIG. 16A, the area defined by the waveform indicating tensile stress and the straight line of stress = 0 is the waveform indicating compressive stress regardless of the time after irradiation time. And the area defined by the straight line.

  In both this embodiment and the comparative example, stress grows in the Si wafer 1 after the flash lamp irradiation. The stress grows as a compressive stress in the front surface portion of the Si wafer 1 and grows as a tensile stress from the inside of the Si wafer 1 to the back surface portion.

  In the case of the present embodiment, the stress generated in about 0.3 to 0.5 milliseconds after flash lamp irradiation is maximized, and the maximum tensile stress value is approximately 40 MPa. On the other hand, in the case of the comparative example, the stress generated in about 3 to 5 milliseconds after the flash lamp irradiation is maximized.

  Further, the depth of change from compressive stress to tensile stress is greater in the comparative example than in the present embodiment. In other words, in the comparative example, the balance between the compressive stress and the tensile stress is maintained in a region closer to the back side of the Si wafer 1 than in the present embodiment. Therefore, although the difference between the surface temperature and the back surface temperature of the Si wafer is larger in the present example than in the comparative example, the absolute value of the tensile stress is larger in the comparative example than in the present embodiment. In the case of the comparative example, the tensile stress grows to a value that reaches about 120 MPa at the maximum.

  FIG. 17 is a diagram illustrating a light emission waveform of another comparative example (second comparative example). The second comparative example has a shorter rise time of the light emission pulse than the comparative example (first comparative example) shown in FIG. 12, but the 1/2 pulse width is the same at 3 milliseconds.

  When the stress distribution and the maximum stress amount in the second comparative example were examined, the same results as in the first comparative example were obtained. From this result, it has been clarified that it is necessary to shorten the ½ pulse width in order to move the depth at which the tensile stress is generated to the wafer surface side.

  FIG. 18 is a diagram showing a brittle fracture curve of a Si wafer. FIG. 18 shows that the strength against tensile stress of the Si wafer decreases as the heating temperature of the Si wafer increases.

  Furthermore, at the substrate preheating temperature of 500 ° C. in FIG. 15 (this embodiment) and FIG. 16 (comparative example), the maximum tensile stress value at which the Si wafer does not break (slip or crack) is about 100 MPa. I understand. Since the tensile stress of the present embodiment is 10 to 30 MPa and the tensile stress value of the comparative example is 100 to 120, the Si wafer is less likely to break in the present embodiment than in the comparative example. That is, the process window of this embodiment is wider than that of the comparative example.

  FIG. 19 and FIG. 20 show process windows of the present embodiment and a comparative example with respect to the substrate preheating temperature and the irradiation energy density, respectively.

  The higher the substrate preheating temperature, the lower the irradiation energy density required for activating the impurities, but at the same time the irradiation energy density at which slip or cracks occur in the Si wafer 1 also decreases.

  Here, when the present embodiment and the comparative example are compared with respect to the irradiation energy density at which slips or cracks occur in the Si wafer 1, from any of the substrate preheating temperatures shown in FIG. 19 and FIG. You can see that it is higher. Therefore, it is considered that the process window becomes wider as the substrate preheating temperature becomes lower and the 1/2 pulse width becomes shorter.

  As described above, according to the present embodiment, the tensile stress value can be reduced by setting the 1/2 pulse width of the flash lamp light source 5 short. As a result, the thermal stress resistance of the Si wafer 1 is improved, so that impurity ions implanted into the Si wafer 1 can be activated without causing thermal damage such as slips or cracks in the Si wafer 1. become. That is, according to the present embodiment, it is possible to realize a process of an impurity diffusion region having a wide process window.

  FIG. 21 is a diagram showing the relationship between the gate leakage current and the ½ pulse width of the MOSFET using the impurity diffusion region 7 as the source / drain region. In the figure, for the sake of simplicity, the horizontal axis represents not the ½ pulse width but the pulse width (the same applies to other figures). From FIG. 21, when the 1/2 pulse width becomes 1 msec or less, the gate leak current rapidly decreases (below spec), and when the 1/2 pulse width becomes 0.5 msec or less, the gate leak current is almost constant and low. It can be seen that the value stabilizes. Similar results were obtained in the MOS capacitor using the impurity diffusion region 7 of the present embodiment.

  FIG. 22 is a diagram showing the relationship between the junction leakage current of the pn junction constituted by the impurity diffusion region 7 and the ½ pulse width. From FIG. 22, when the 1/2 pulse width becomes 1 msec or less, the junction leakage current sharply decreases (below spec), and when the 1/2 pulse width becomes 0.5 msec or less, the junction leakage current is almost constant and low. It can be seen that the value stabilizes.

  FIG. 23 is a diagram showing the relationship between the yield of the logic circuit composed of the MOSFET and the 1/2 pulse width. From FIG. 23, it can be seen that when the 1/2 pulse width becomes 1 ms or less, the yield increases rapidly, and when the 1/2 pulse width becomes 0.5 ms or less, the yield stabilizes at a substantially constant high value.

  From the above results, it can be seen that the device characteristics can be improved and the yield can be improved by setting the ½ pulse width to 1 ms or less, more preferably 0.5 ms or less. This is considered to be because the tensile stress value that causes thermal damage is sufficiently reduced by the above-described mechanism by setting the 1/2 pulse width to 1 ms or less.

  The present invention is not limited to the above embodiment. For example, an embodiment in which the above embodiments are appropriately combined is also possible. For example, an embodiment in which the fifth embodiment is applied to the first to fourth embodiments is also possible.

  Furthermore, the above embodiments include inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if some constituent requirements are deleted from all the constituent requirements shown in the embodiment, if the problem described in the column of the problem to be solved by the invention can be solved, the configuration in which this constituent requirement is deleted Can be extracted as an invention.

  In addition, various modifications can be made without departing from the scope of the present invention.

  DESCRIPTION OF SYMBOLS 1 ... Si wafer (Si substrate), 2 ... Mask, 3 ... Impurity ion, 4 ... Hot plate, 5 ... Flash lamp light source, 6 ... Light, 7 ... Impurity diffusion region, 8 ... Flash lamp, 9 ... Crystal orientation, 10 ... Lamp array direction, 11 ... Film to be processed, 11p ... Pattern, 12 ... Line pattern array direction, 21 ... Gate insulating film, 22 ... Gate electrode, 24 ... Extension region, 25 ... Gate sidewall insulating film, 26 ... Source / drain region.

Claims (2)

Disposing a flash lamp light source including a plurality of flash lamps above a Si wafer including a single crystal Si region;
A step including a first heat treatment and a second heat treatment for heating the Si wafer by light emitted from the light source, wherein each of the first heat treatment and the second heat treatment includes: Heating the Si wafer such that a light intensity distribution formed on the Si wafer by light forms a distribution having a maximum intensity in a direction different from the crystal orientation of the single-crystal Si region; and The step of arranging the plurality of flash lamps during the first heat treatment and the step of arranging the plurality of flash lamps during the second heat treatment are different from each other. The method for manufacturing a semiconductor device, wherein the crystal orientation is a plane orientation of a cleavage plane of the single crystal Si region.   By rotating the plurality of flash lamps, the arrangement direction of the plurality of flash lamps during the first heat treatment is different from the arrangement direction of the plurality of flash lamps during the second heat treatment. The method of manufacturing a semiconductor device according to claim 1, wherein:

Images