Holding device

Abstract

To suppress damage to a joining unit by improving cooling performance in a holding device even when heat input to a plate-like unit for holding an object becomes large.SOLUTION: A holding device for holding an object includes: a plate-like unit formed in a plate-like shape; a base unit that supports the plate-like unit, is mainly composed of a material having a thermal expansion coefficient different from the thermal expansion coefficient of a material mainly forming the plate-like unit, has a cooling function, and is formed in a plate-like shape; and a joining unit that is arranged between the plate-like unit and the base unit, includes an adhesive, and joins the plate-like unit and the base unit to each other. The thermal resistance of the joining unit is 5.0×10-4 (m2K / W) or less, and the strain amount at the time of maximum shearing stress of the joining unit is 0.5 mm or more.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a retention device. [Background technology]

[0002] Conventionally, as a holding device for holding an object, for example, a wafer or the like in manufacturing semiconductors has been used. Electrostatic chucks for holding objects are known. Generally, electrostatic chucks have a structure in which an object is placed on the electrostatic chuck. a ceramic portion in which a refrigerant flow path is formed, a base portion in which the ceramic portion and the base portion are connected For example, Patent Document 1 describes a material for the adhesive layer (joint). Silicone resin, silicone resin with aluminum oxide or aluminum nitride, etc. It is described that an adhesive made of a composite resin containing a conductive filler is used. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2008-300491 A Summary of the Invention [Problem to be solved by the invention]

[0004] According to the technology described in the above Patent Document 1, silicone resin is used as the material for the joint. This reduces the stress between the ceramic part and the base part even under relatively high temperature conditions. However, for example, by exposing the electrostatic chuck to a higher power plasma, When the heat input to the ceramic part is large, such as when used in a ceramic There was a possibility that the temperature of the ceramic part would rise excessively, resulting in a decrease in the processing accuracy of the wafer. There was a possibility that the joints would be damaged due to stress generated in the joints by excessive temperature rise. Therefore, even if the heat input to the ceramic part is large, the There has been a demand for technology that can improve cooling performance and prevent problems such as damage to joints. [Means for solving the problem]

[0005] The present disclosure can be realized in the following forms. (1) According to one aspect of the present disclosure, a holding device is provided. The holding device is formed in a plate shape. and a plate-shaped portion that supports the plate-shaped portion and has a thermal expansion coefficient different from that of a material that mainly forms the plate-shaped portion. A base portion formed in a plate shape and mainly made of a material having a thermal expansion coefficient and having a cooling function; the plate-shaped portion and the base portion, the plate-shaped portion containing an adhesive and connecting the plate-shaped portion and the base portion; and a junction portion for joining the thermal resistance of the junction portion is 5.0×10 -4 (m 2 K / W) or less and the strain of the joint at maximum shear stress is 0.5 mm or more. According to the holding device of this embodiment, even when the heat input to the plate-shaped portion is large, the holding It is possible to improve the cooling performance of the device and suppress the inconvenience caused by the stress generated at the joint. This can be done. (2) In the holding device of the above embodiment, a tensile test is performed on the joint, The shear stress when the deformation length due to shear force is 0.5 mm is 0.1 MPa or more. The pressure may be 3.5 MPa or less. With this configuration, the pressure generated at the joint is Since it is possible to suppress shear stress, it is effective in suppressing inconvenience caused by stress generated at the joint. This can further improve the results. (3) In the holding device of the above aspect, the joint portion contains the adhesive and an inorganic filler. The structure may be a composite structure. In this structure, the bonding may be performed by using an adhesive. The flexibility of the part is ensured to increase the amount of strain at maximum shear stress, and the inorganic filler is used This ensures the heat conductivity of the joint and reduces the thermal resistance of the joint. (4) In the holding device of the above embodiment, the radius of the inscribed circle when the inorganic filler is viewed in cross section If the radius of the circumscribed circle is R1 and the radius of the circumscribed circle is R2, the average value of R2 / R1 is 1.10 or more. With such a configuration, more contact points between the inorganic fillers can be formed. This makes it easier for a thermal conduction path to form within the joint, reducing the thermal resistance of the joint. do. (5) In the holding device of the above embodiment, the rigidity modulus of the joint is 0.03 MPa or more, 2.0 MPa or less. With such a configuration, the flexibility of the joint is ensured. Therefore, by improving the stress relaxation performance at the joint, when a shear force is applied to the joint, Damage to the joint can be suppressed, and the shape retention of the joint can be improved. (6) In the holding device of the above embodiment, the joint portion is a resin layer containing the adhesive, The adhesive layer is formed so as to extend in the surface direction of the joint and is made of an inorganic material. In this configuration, the inorganic sheet may be By providing this, the thermal resistance and the amount of strain at maximum shear stress can be made uniform within the joint surface. At the same time, the temperature distribution within the surface of the joint can be made uniform. (7) In the holding device of the above embodiment, the plate-shaped portion is mainly made of ceramic, The device includes an adhesion electrode for holding an object, but does not include a heater electrode for heating the object. With this configuration, the object and the plate-shaped portion are heated by the heater electrode. Even when the plate is not heated, the heat input to the plate is large and the temperature difference between the plate and the base is large. In the use condition where the size tends to become large, the cooling performance of the holding device is improved and the joint This can significantly reduce the inconvenience caused by the stress generated. The present disclosure may be realized in various forms other than those described above. For example, a semiconductor manufacturing device including a holding device The present invention can be realized in the form of a manufacturing device, a manufacturing method for a holding device, a method for forming a joint, etc. [Brief description of the drawings]

[0006] [Figure 1] 1 is a perspective view illustrating a schematic external appearance of an electrostatic chuck according to a first embodiment. [Diagram 2] FIG. 1 is a cross-sectional view illustrating a schematic configuration of an electrostatic chuck. [Diagram 3] FIG. 2 is an explanatory diagram showing how to calculate “R2 / R1” for inorganic filler particles. [Figure 4] FIG. 4 is an explanatory diagram showing how heat transfer paths are formed between particles of an inorganic filler. [Diagram 5] FIG. 4 is an explanatory diagram showing the relationship between the content ratio of the inorganic filler and the thermal conductivity of a joint. [Figure 6] FIG. 4 is an explanatory diagram regarding stress occurring at a joint. [Figure 7] FIG. 11 is a cross-sectional view illustrating a schematic configuration of an electrostatic chuck according to a second embodiment. [Figure 8] FIG. 4 is an explanatory diagram showing the measurement results and evaluation results of each sample. [Figure 9] FIG. 4 is an explanatory diagram showing the measurement results and evaluation results of each sample. [Figure 10] FIG. 4 is an explanatory diagram showing the measurement results and evaluation results of each sample. [Figure 11] FIG. 4 is an explanatory diagram showing the measurement results and evaluation results of each sample. [Figure 12]FIG. 4 is an explanatory diagram illustrating a method for calculating maximum shear stress and strain amount. [Figure 13] FIG. 1 is an explanatory diagram showing an example of actual measured values ​​(sample S25) in a tensile test. [Figure 14] FIG. 1 is an explanatory diagram showing an example of actual measured values ​​(sample S25) in a tensile test. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0007] A. First embodiment: (A-1) Structure of electrostatic chuck: FIG. 1 is a perspective view showing a schematic external appearance of an electrostatic chuck 10 according to a first embodiment. 2 is a cross-sectional view showing a schematic configuration of an electrostatic chuck 10. In FIG. 1 and 2, as well as FIGS. 6 and 7, which will be described later, show the method. In order to specify the direction, mutually orthogonal XYZ axes are shown. In this specification, the Z axis indicates the vertical direction, and the X axis indicates the The X-axis and Y-axis indicate the horizontal direction. Note that the above figures are only schematic representations of the layout of each part. The dimensions of each part are not shown exactly.

[0008] The electrostatic chuck 10 is a device that attracts and holds an object by electrostatic attraction. It is used to fix a wafer W, which is an object, in a vacuum chamber of a conductor manufacturing device. The electrostatic chuck 10 includes a ceramic portion 20, a base portion 30, and a bonding portion 40. These are, in the -Z axis direction (vertically downward), the ceramic portion 20, the joint portion 40, the base The electrostatic chuck 10 in this embodiment is referred to as a “holding device.” Also called.

[0009] The ceramic portion 20 is a substantially circular plate-like member made of ceramic (e.g., aluminum oxide). In the present specification, the specific component is "The main component" or "the material that mainly forms the product" refers to the content of the specific component. The ceramic portion 20 has a diameter of, for example, 50 mm. The thickness should be about 200mm to 500mm, and is usually about 200mm to 350mm. The thickness of the ceramic portion 20 may be, for example, about 1 mm to 10 mm. It is also called the "plate portion."

[0010] As shown in FIG. 2, an attraction electrode 22 is disposed inside the ceramic portion 20. The electrode 22 is made of a conductive material such as tungsten or molybdenum. When a voltage is applied to the chucking electrode 22 from a power source (not shown), an electrostatic attraction force is generated. The wafer W is attracted and fixed to the mounting surface 24 of the ceramic portion 20 by this electrostatic attraction. The electrode 22 may be of a bipolar type or a unipolar type. Inside is a resistor made of conductive material (e.g., tungsten, molybdenum, etc.). A heating element (not shown) is configured to heat the wafer W fixed by suction to the mounting surface 24. A heater electrode may also be provided.

[0011] The base portion 30 is a plate-like member that includes a metal and is formed into a substantially circular shape. For example, aluminum, magnesium, molybdenum, titanium, tungsten, and nickel The alloy may contain at least one of the following metals: molybdenum, titanium, tungsten Among the metals mentioned above, stainless steel has a relatively small thermal expansion coefficient, so In the case where the base portion 30 is made of a single metal, the base portion 30 and the ceramic portion 20. "Thermal expansion coefficient" refers to "linear expansion coefficient." In addition, magnesium has a relatively small Young's modulus, Therefore, when the base portion 30 is made of magnesium, the thermal stress generated in the base portion 30 is large. Aluminum is also desirable because it has a relatively high thermal conductivity and can reduce the stress caused by heat. Therefore, when the base portion 30 is made of aluminum, In this case, the cooling efficiency of the ceramic part 20 and the wafer W by the base part 30 can be improved. This is desirable because it is possible to reduce the manufacturing cost of the electrostatic chuck 10. From the viewpoint of increasing the cooling efficiency and reducing the manufacturing cost, the metal content of the base portion 30 is It is desirable that the ratio is high, and it is desirable that the base portion 30 contains metal as a main component. For example, , and contains 90% or more by mass of aluminum, which is highly versatile (e.g., A6061, A5 It is preferable to use an aluminum alloy such as 052 for the base portion 30. The diameter of the base portion 30 may be, for example, 10 mm or less. For example, it may be about 220 mm to 550 mm, and typically it is 220 mm to 350 mm. The thickness of the base portion 30 may be, for example, about 20 mm to 40 mm.

[0012] A plurality of refrigerant flow paths 32 are formed inside the base portion 30 along the XY plane. By passing a coolant such as a fluorine-based inert liquid, water, or liquid nitrogen through the coolant flow passage 32, The base portion 30 is cooled. Then, the base portion 30 and the ceramic portion 40 are bonded to each other via the bonding portion 40. The ceramic portion 20 is cooled by heat transfer between the ceramic portion 20 and the mounting surface 24 of the ceramic portion 20. The held wafer W is cooled. In this way, the temperature of the wafer W is controlled. In addition to a configuration in which the refrigerant flow path 32 is provided inside the base portion 30, a configuration in which the refrigerant flow path 32 is provided from the outside of the base portion 30 The base portion 30 may have a cooling function by cooling the base portion 30 .

[0013] The joint portion 40 is disposed between the ceramic portion 20 and the base portion 30. The joint 40 is made of a resin material and contains an adhesive. The joint 40 may further include a paste for forming the joint 40, and the properties of the joint 40. It may contain various fillers (inorganic fillers) to adjust the properties. The joint 40 may be made of a compound containing an adhesive and an inorganic filler. If the joint 40 satisfies the properties described below, the joint 40 does not contain an inorganic filler. From the viewpoint of reducing the thermal resistance of the joint 40, the thickness of the joint 40 may be, for example, It is good to set it to 1.00mm or less, 0.45mm or less is preferable, and 0.40mm or less is more preferable. The thickness of the joint 40 is preferably 0.35 mm or less, and more preferably 0.35 mm or less. From the viewpoint of ensuring the flexibility and strength of the material, it is sufficient to set the thickness to 0.05 mm or more.

[0014] The electrostatic chuck 10 further includes a plurality of gas supply paths 50. 50 is provided penetrating the ceramic portion 20, the joint portion 40, and the base portion 30 in the Z direction. The gas outlet 52 is formed in the mounting surface 24 and opens at the gas outlet 52 (see FIG. 1). The gas supply line 50 supplies an inert gas, such as helium gas, from a gas supply device (not shown). The inert gas is supplied from the gas outlet 52 to the space between the mounting surface 24 and the wafer W. This increases the heat transfer between the ceramic portion 20 and the wafer W, The controllability of the temperature distribution of the W is further improved. The gas supply path 50 may not be provided in the electric chuck 10 .

[0015] (A-2) Joint configuration: The structure of the joint 40 will be described below. As described above, the joint 40 is made of resin. and further comprising an inorganic filler. The joint 40 of the electrostatic chuck 10 of the present embodiment has a thermal resistance of 5.0×10 -4 (m 2 K / W) or less, and the strain at maximum shear stress is 0.5 mm or more. The strain under large shear stress is a numerical value that is an index of the flexibility and stress relaxation performance of the joint 40. The amount of strain at maximum shear stress will be explained in detail later.

[0016] The adhesive that constitutes the joint 40 may be, for example, a silicone resin, an acrylic resin, or the like. In particular, silicone resin has relatively high heat resistance and Among these resins, silicone resin has a relatively high elastic modulus. Since the temperature is low, the function of mitigating thermal stress occurring at the joint 40 is high, and the heat resistance temperature is relatively high. High, therefore desirable.

[0017] Inorganic fillers include ceramics, metal oxides, metals, or other inorganic compounds. The material may be in the form of particles or powder, and may be made of various inorganic materials, including inorganic compounds. Inorganic fillers include, for example, aluminum nitride (AlN) and aluminum oxide. (Alumina: Al2O3), Zirconium oxide (Zirconia: ZrO2), Yttria oxide Yttrium (Yttrium oxide: Y2O3), Yttrium fluoride (YF3), Silicon carbide (SiC) , silicon nitride (Si3N4), silicon dioxide (silica: SiO2), iron oxide, barium sulfate The inorganic filler may be composed of the above-mentioned inorganic materials such as calcium carbonate, calcium carbonate, etc. The material generally has a higher thermal conductivity than the resin used as the adhesive, so an inorganic filler is used in the joint 40. By adding the above, the thermal conductivity at the joint 40 can be increased. From the viewpoint that the conductivity is relatively high and the thermal resistance of the joint 40 can be easily suppressed, the inorganic filler The material constituting the film is preferably aluminum nitride, aluminum oxide, or silicon carbide. , aluminum nitride and aluminum oxide are particularly preferred.

[0018] The bonding portion 40 may further include a catalyst for promoting the curing reaction, or a bonding agent for promoting curing and adhesion. Silane coupling agents for imparting heat, crosslinking agents, and reaction agents for adjusting the curing speed of adhesives The joint 40 may contain a reaction inhibitor, a viscosity adjuster, or the like. Various conventional catalysts can be used, such as platinum catalysts, rhodium catalysts, and titanium catalysts. Among them, platinum catalyst, which has high reactivity, is used. The silane coupling agent contained in the bonding portion 40 is not particularly limited, and examples thereof include , as organic reactive groups, vinyl groups, epoxy groups, methacryl groups, amino groups, mercapto groups, Among the conventionally known silane coupling agents, those having either an isocyanate group or Alternatively, instead of the silane coupling agent, a titanate coupling agent may be used. Alternatively, a coupling agent based on ethoxylated or aluminate may be used. The crosslinking agent is an organohydride having at least three hydrosilyl groups in one molecule. More specifically, for example, polymethyl halides polysiloxane can be used. Hydrogensiloxane and poly(dimethylsiloxane-methylhydrogensiloxane) At least one of the above can be used.

[0019] As the reaction inhibitor contained in the joint 40, various reaction inhibitors known in the art can be used. For example, 1,3,5,7-tetramethyl-1,3,5,7-tetravinylcyclopentasiloxane Lasiloxane, 1,1,3,3-tetramethyl-1,3-divinyldisiloxane, Tria The viscosity modifier contained in the joint 40 may be, for example, butyl isocyanurate. Various viscosity modifiers known in the art can be used, such as fumed silica, fumed silica, etc. Silica, colloidal silica, fumed alumina, colloidal alumina, etc. The above-mentioned catalysts, silane coupling agents, crosslinking agents, reaction inhibitors, and The type and amount of viscosity modifier, etc., depends on, for example, the type of resin constituting the joint 40. An appropriate selection may be made depending on the situation.

[0020] As described above, in this embodiment, the thermal resistance of the joint 40 is 5.0×10 -4 (m 2 K / W) or less. The thermal resistance of the junction 40 is 4.5×10 -4 (m 2 K / W) or less It is desirable to do so at 4.0×10 -4 (m 2It is more preferable to keep it below K / W. 3.5×10 -4 (m 2 It is more desirable to set the thermal conductivity λ and Here, the thermal resistance R refers to a value at room temperature (22° C.). The thermal resistance of the junction 40 is R (m 2 K / W), the thickness of the joint 40 is t (m), and the thermal conductivity of the joint 40 is λ (W / mK). Then, the thermal resistance R of the joint 40 is calculated by the following formula (1).

[0021] R(m 2 K / W) = t(m) ÷ λ(W / mK) … (1)

[0022] By setting the thermal resistance of the joint 40 to the above value, the thermal resistance of the base portion 30 via the joint 40 Heat transfer between the ceramic portion 20, i.e., heat transfer from the ceramic portion 20 to the base portion 30 This makes it easier for the electrostatic chuck 10 to cool, thereby improving the cooling efficiency of the electrostatic chuck 10.

[0023] According to formula (1), the thermal resistance of the joint 40 is reduced by reducing the thickness of the joint 40. For example, by making the thickness of the joint 40 less than 0.5 mm, 0 thermal resistance to 5.0×10 -4 (m 2 However, it is easy to keep the connection By increasing the thermal conductivity of the joint 40 to, for example, 1.0 W / mK or more, the thickness of the joint 40 Even if the thickness is increased to 0.5 mm or more, the thermal resistance of the junction 40 is 5.0×10 -4 (m 2 K / W) In order to increase the thermal conductivity of the joint 40, for example, In order to achieve this, inorganic fillers made of materials with higher thermal conductivity are used, or the inclusion of inorganic fillers is It would be better to increase the proportion of such companies.

[0024] The thermal resistance of the joint 40 depends on, for example, the material of the inorganic filler contained in the joint 40, The content of inorganic filler in the joint 40 and the type of resin contained in the joint 40 vary. The use of inorganic fillers made of materials with higher thermal conductivity allows By increasing the content of inorganic filler, the thermal conductivity is improved. By using a resin with a high thermal conductivity, the thermal conductivity of the joint 40 is increased, and the thermal resistance of the joint 40 is reduced. It can be made smaller.

[0025] Furthermore, the thermal resistance of the joint 40 depends on the shape of the inorganic filler particles contained in the joint 40. In order to reduce the thermal resistance of the joint 40, the inorganic filler is removed. When the radius of the inscribed circle is R1 and the radius of the circumscribed circle is R2, the ratio R2 / R It is desirable that the average value of "1" is 1.05 or more, and even better that it is 1.10 or more. The closer "R2 / R1" is to 1.0, the closer the particle shape is to a perfect sphere (hereafter, the term "perfect sphere" is used). This indicates that the ball has a high degree of sphericity.

[0026] The method for identifying the inscribed and circumscribed circles and the method for calculating "R2 / R1" are as follows: That is, in an image of a cross section of the joint 40 observed with a scanning electron microscope (SEM), Focus on one particle being observed. The particle is enclosed and in contact with three vertices of the particle. Draw a circle and call it the circumscribing circle. Also, draw a circle that is contained within the particle and touches the particle's periphery at three points. Draw a circle so that the circle is inscribed, and call this the inscribed circle. If multiple circles can be drawn, choose the circle with the smallest radius. The circles are the inscribed and circumscribed circles. Identify the inscribed and circumscribed circles for 10 particles in the same way. Then, measure the radius R1 of the inscribed circle and the radius R2 of the circumscribed circle, and calculate "R2 / R1" from the average value. " is calculated.

[0027] FIG. 3 is a diagram showing how to determine "R2 / R1" when a particle of the inorganic filler 42 is viewed in cross section. FIG. 4 is an explanatory diagram showing how a heat transfer path is formed between inorganic fillers 42. In FIG. 3, the inscribed circle of the inorganic filler 42 when viewed in cross section is indicated by a dashed line, and the circumscribed circle is indicated by a dashed line. The center of the inscribed circle is indicated as O1, and the center of the circumscribed circle is indicated as O2. In FIG. 4, a thermal conduction path is formed at the location where the inorganic fillers 42 contact each other. The state is represented by a double-headed arrow. The value of "1" is equal to or greater than the above-mentioned lower limit, and the shape of the inorganic filler 42 is different from a perfect sphere and has a surface roughness. As the surface becomes uneven, the number of contact points between the inorganic fillers 42 increases, as shown in FIG. This makes it easier for a heat conduction path to be formed within the joint 40, and reduces the thermal resistance of the joint 40.

[0028] FIG. 5 shows the relationship between the inorganic filler particles at the joint 40 and the particle shape of two different inorganic fillers. FIG. 5 is an explanatory diagram showing an example of the relationship between the filler content and the thermal conductivity of the joint 40. The inorganic filler is aluminum nitride (AlN) and aluminum oxide (Alumina Among the inorganic fillers used here, aluminum nitride is more effective. The value of "R2 / R1" is larger than that of aluminum oxide. The "R2 / R1" of minium is 1.15, and the "R2 / R1" of aluminum oxide is 1. The number is 05.

[0029] As shown in FIG. 5, even if the content ratio of the inorganic filler in the joint 40 is the same, Aluminum nitride with a larger value of "R2 / R1" has a higher thermal conductivity value of the joint 40. As shown in Figure 5, the higher the inorganic filler content, the greater the aluminum nitride The difference in thermal conductivity between aluminum and aluminum oxide relative to the content ratio is significant. This is because the larger the content of inorganic filler, the greater the gap between inorganic filler particles. The distance becomes smaller, making it easier for the inorganic fillers to come into contact, and the number of contact points between the inorganic fillers increases, improving heat transfer. This is because the effect on the conductivity of the inorganic filler becomes more apparent. When the inorganic filler is close to a perfect sphere, even a slight decrease in the content ratio is sufficient. If the surface of the inorganic filler is uneven, the contact points between the inorganic filler particles are easily reduced. This is thought to be because the effect of reducing the contact points between inorganic fillers is relatively small even if the ratio is reduced. In this way, when the content of the inorganic filler is relatively small (for example, 20 vol.%), When the following conditions are met, the effect of the shape of the inorganic filler on the thermal conductivity is small, and Due to the difference in thermal conductivity of the material itself, there is a slight difference in the thermal conductivity of the joint 40. When the content of inorganic filler is relatively high and the inorganic fillers are likely to come into contact with each other, Since the shape of the filler has a large effect on the thermal conductivity, the thermal conductivity of the joint 40 is greatly affected. It is believed that there will be a difference. In the inorganic filler used in Figure 5, aluminum nitride has a larger "R2 / R1" value than aluminum oxide, but the "R2 / R1" value is Regardless of the inorganic filler material, various changes are possible by adjusting the manufacturing method, etc. do.

[0030] The inorganic filler may be formed from a homogeneous material, or may be formed from a material that is a particle body of the inorganic filler. A coating layer made of a material different from the material constituting the coating layer may be provided on the surface. By appropriately selecting the filler, the water resistance of the inorganic filler can be improved, and the inorganic filler and the resin can be bonded to each other. It is possible to improve the compatibility and increase the flexibility of the joint 40. When using aluminum nitride powder as a filler, the aluminum nitride that constitutes the inorganic filler The surface of the aluminum particles is covered with silicon dioxide (SiO2), aluminum oxide (Al2O3), Alternatively, a coating layer containing aluminum phosphate (AlPO4) may be provided.

[0031] The lower limit of the thermal resistance of the joint 40 is, for example, 0.6×10 -4 (m 2 K / W) In order to suppress the thermal resistance of the joint 40, the inorganic filler can be used as described above. A method of appropriately selecting the constituent materials and increasing the content of inorganic filler is considered. However, if the content of the inorganic filler is excessively increased, the flexibility of the joint 40 decreases. As a result, it becomes difficult to sufficiently secure the magnitude of the strain at the maximum shear stress of the joint 40, which will be described later. In order to suppress the thermal resistance of the joint 40, the joint 40 should be made thin. However, if the joint 40 is made too thin, the strength of the joint 40 may be deteriorated. As the strength of the joint 40 decreases, it becomes difficult to ensure the flexibility of the joint 40, and the strain amount at the maximum shear stress It may be difficult to set the thermal resistance of the junction 40 in the desired range. , the above 0.6 × 10 -4 (m 2 It is desirable to set the temperature at or above 100 K / W.

[0032] As described above, in this embodiment, the strain amount at the maximum shear stress of the joint 40 is The strain at the maximum shear stress of the joint 40 is 0.6 mm or more. It is preferable that the thickness is 1.0 mm or more, and more preferable that the thickness is 1.1 mm or more. It is more desirable that the strain at the maximum shear stress of the joint 40 is generally 5 The strain at maximum shear stress is determined by the flexibility and stress relaxation of the joint 40. This is an index value that indicates the ability of the joint 40 to withstand shear force. When the shear stress is maximum, i.e., when the maximum shear stress occurs at the joint 40 The maximum shear stress is the magnitude of the strain (displacement in the shear direction) that occurs at the joint 40. The greater the strain under stress, the greater the flexibility of the joint 40. The specific measurement method using a tensile tester to measure the amount of strain will be explained in detail later. If the strain amount at the maximum shear stress of the joint 40 is set to the above value, the flexibility of the joint 40 is improved. Therefore, it is possible to sufficiently ensure the durability and stress relaxation performance.

[0033] 6 is an explanatory diagram of the stress generated at the bonding portion 40. In the present embodiment, the ceramic portion 20 is heated by exposure to plasma or the like. The base portion 20 is heated and expands. The base portion 30 is provided with a refrigerant flow path 32, so that the cooling FIG. 6 shows how the ceramic portion 20 is heated and expands, and how the base The state in which the joint 30 is cooled and contracts is shown by the white arrow. In the bonding process, a pulling force acts on the ceramic portion 20 and the base portion 30. The force with which the ceramic portion 20 is pulled by the base portion 30 is expressed as follows: 6, the amount of deformation in the X-axis direction occurring at the joint 40 is shown by the arrow. The amount of distortion is shown as δ.

[0034] In addition, in the electrostatic chuck 10, the base portion 30 is generally larger than the ceramic portion 20. It has a higher thermal expansion rate than quartz and expands and contracts significantly due to temperature changes. In other words, the degree of thermal expansion of the base portion 30 is greater than the degree of thermal expansion of the ceramic portion 20. In this way, the expansion and contraction between the ceramic portion 20 and the base portion 30 may occur. Since the degree of compression is different, a shear force in the X-axis direction is applied to the joint 40, and the shear Therefore, as described above, the strain amount at the maximum shear stress of the joint 40 is large. By increasing the size, even if a large shear force is applied to the joint 40, This reduces the shear stress occurring at the joint 40, and suppresses damage to the joint 40.

[0035] The amount of strain at the maximum shear stress of the joint 40 is determined by the adhesive (resin) contained in the joint 40. For example, the maximum shear stress can be changed depending on the type of resin that constitutes the joint 40. In addition to being able to change the amount of strain when force is applied, even if the same type of resin is used, the resin is a polymer material. By controlling the distance between the crosslinking points in the resin, the strain amount at the maximum shear stress of the joint 40 can be Specifically, the content of the functional group that serves as the crosslinking point in the resin (functional group equivalent) By reducing the crosslinking point distance, the flexibility of the resin is improved, and the maximum shear strength is reduced. The amount of strain under stress can be increased. In addition, the content of reactive functional groups in the resin is Even in the same case, the distance between crosslinks may vary depending on the curing conditions such as the resin curing temperature and curing time. That is, when the curing temperature is increased or the curing time is extended, In this case, the resin hardens further, the crosslink density increases, and the distance between crosslinks becomes shorter. become.

[0036] In addition, the amount of strain at the maximum shear stress of the joint 40 increases as the thickness of the joint 40 increases. So it can be made larger.

[0037] In addition, the greater the content of the inorganic filler in the joint 40, the smaller the maximum stress of the joint 40. The strain under tensile stress tends to be smaller. This is because the content of inorganic filler is high. The degree to which the inorganic filler binds the surrounding resin (resin composition) increases, and the joint 4 This is thought to be because the flexibility of the joint 40 decreases, making it difficult for the joint 40 to distort.

[0038] According to the electrostatic chuck 10 of the present embodiment configured as described above, the electrostatic chuck 10 The thermal resistance of the junction 40 is 5.0×10 -4 (m 2 K / W) or less, and the maximum shear stress Since the strain under force is 0.5 mm or more, the heat input to the ceramic portion 20 is larger. Even if the temperature of the electrostatic chuck 10 is too high, the cooling performance of the electrostatic chuck 10 can be improved and the bonding portion 40 can be prevented from being damaged. This can suppress the inconvenience caused by the stress generated. The smaller the thermal resistance of the portion 40, the higher the cooling efficiency of the electrostatic chuck 10. Therefore, by setting the thermal resistance of the joint 40 to the above value, for example, excess This prevents the deterioration of wafer processing accuracy caused by temperature rise. At the same time, the shear stress occurring at the joint 40 can be suppressed. The greater the strain at maximum shear stress, the greater the stress relaxation performance of the joint 40. By setting the strain amount at the maximum shear stress of the joint 40 to 0.5 mm or more, Even if shear stress occurs in the joint 40, damage to the joint 40 caused by the shear stress is prevented. From the above, the thermal resistance and the strain amount at the maximum shear stress of the joint 40 can be suppressed. By setting the above values, the performance of the entire electrostatic chuck 10 can be improved.

[0039] Such effects are particularly prevalent when the electrostatic chuck 10 is exposed to higher power plasmas. This is particularly noticeable when the heat input to the ceramic part is large. Even if a heater electrode for heating the block portion 20 is not provided, a high plasma power can be obtained. When the electrostatic chuck 10 is used in combination with the ceramic portion 20, the mounting surface of the ceramic portion 20 and the base portion 30 Since the temperature difference between the two is likely to be large, the effect of this embodiment is remarkable.

[0040] By improving the stress relaxation performance at the joint 40, when a shear force is applied to the joint 40, In order to suppress damage to the joint 40, when a shear force is applied to the joint 40, It is desirable that the shear stress generated at the joint 40 is small. In order to prevent this, a tensile test was performed on the joint 40, and the length of deformation due to shear force was 0 The shear stress generated at the joint 40 when the shear strain is 0.5 mm (hereinafter referred to as "shear strain 0. The shear stress at 5 mm should be 3.5 MPa or less. It is more preferable that the ratio is 2.0 or less, and even more preferable that the ratio is 2.0 or less. A specific test was performed using a tensile testing machine to measure the shear stress at a breaking strain of 0.5 mm. The measurement method will be explained in detail later.

[0041] In addition, the "shear stress when the shear strain is 0.5 mm" must be 0.1 MPa or more. For example, the electrostatic chuck 10 may be used as a vacuum chuck. In this case, the ceramic in the electrostatic chuck 10 is The mix section 20 side may be exposed to a vacuum and the base section 30 side may be exposed to the atmosphere. If the "shear stress when the shear strain is 0.5 mm" is 0.1 MPa or more, then the above Even in the above-described manner of use, it is easy to ensure sufficient strength at the joint 40. In addition, if the usage is different from the above, for example, "When the shear strain is 0.5 mm, The "shear stress" may be less than 0.1 MPa.

[0042] By improving the stress relaxation performance at the joint 40, when a shear force is applied to the joint 40, In order to prevent damage to the joint 40, as described above, the flexibility of the joint 40 is ensured. As another index showing the flexibility of the joint 40, the rigidity modulus of the joint 40 is mentioned. The rigidity of the joint 40 is preferably, for example, 2.5 MPa or less. It is preferable that the pressure is 2.0 MPa or less, and more preferable that the pressure is 1.5 MPa or less. However, if the rigidity of the joint 40 is too small, the shape of the joint 40 may not be maintained. The electrostatic chuck 10 may become unstable when a wafer is attached to or detached from the electrostatic chuck 10 or when the electrostatic chuck 10 is transported. When the ceramic portion 20 is inserted, it is possible that the ceramic portion 20 may move to an undesirable extent. From the viewpoint of suppressing such inconvenience, the rigidity modulus of the joint 40 may be set to 0.03 MPa or more. The method for determining the rigidity modulus will be explained in detail later.

[0043] As described above, the joint 40 needs to have a shear strain of 0.5 mm or more. In order to achieve this, it is desirable for the adhesive used in the joint 40 to have a tensile elongation of 40% or more. It is more preferable that the ratio is 50% or more, and even more preferable that the ratio is 60% or more. The tensile elongation of the adhesive is preferably as large as possible, but it is usually 500% or less. This refers to the condition in which a hardened adhesive sample breaks when subjected to a tensile test using a tensile tester. The tensile elongation is the length at which the sample is stretched divided by the initial length of the sample. Only for adhesives with a quantity of more than 100, conditions for joining were examined and the flexibility of the joint 40 was evaluated. The measurements are the maximum shear stress, the strain at the maximum shear stress, and the shear strain 0.5 By measuring the shear stress and rigidity at 1 mm, the efficiency of adhesive selection can be improved. Cut.

[0044] As described above, since the joint portion 40 has sufficient flexibility, the ceramic of the electrostatic chuck 10 As described above, the flatness of the mounting surface 24 of the mixer 20 can be improved. In the case of the joint 10, even if a shear force is applied to the joint 40 due to a temperature change, By increasing the flexibility of the joint 40, the shear stress occurring at the joint 40 can be suppressed. In both cases, the force with which the joint portion 40 pulls the ceramic portion 20 is suppressed, and deformation of the ceramic portion 20 is suppressed. That is, the decrease in flatness of the ceramic portion 20 is suppressed. As defined in B0621 "Definition and display of geometric deviations", "The geometry of a planar feature It indicates the "degree of deviation from the objectively correct plane."

[0045] B. Second embodiment: FIG. 7 is a cross-sectional view showing a schematic configuration of an electrostatic chuck 110 according to the second embodiment. The electrostatic chuck 110 of the embodiment includes a multi-layer joint 140 instead of the single-layer joint 40. The electrostatic chuck 1 according to the second embodiment has the same configuration as that of the first embodiment except for the above. 10, the same reference numerals are used to designate parts common to the electrostatic chuck 10 of the first embodiment. 7, the configuration inside the base portion 30, such as the coolant flow path 32, is omitted.

[0046] The bonding portion 140 includes a first resin layer 142, an inorganic sheet 144, and a second resin layer 146. In the bonding portion 140, the first resin layer 14 2, the inorganic sheet 144, and the second resin layer 146 are laminated in this order. In this embodiment, the first resin layer 142 and the second resin layer 146 are collectively referred to as a "resin layer." In the embodiment, the inorganic sheet 144 is arranged in the resin layer in the plane direction (X-axis direction) of the joint portion 140. The heat of the joint 140 having the three-layer structure as described above is The resistance and the strain amount at maximum shear stress are in the same range as those of the joint 40 of the first embodiment. That is, the thermal resistance of the entire joint 140 is 5.0×10 -4 (m 2 K / W) or less, and the strain at maximum shear stress is 0.5 mm or more.

[0047] The first resin layer 142 and the second resin layer 146 include an adhesive made of resin, As the adhesive, the adhesive already described can be used, which is the same as that used for the joint 40 in the first embodiment. In particular, it is desirable to use a silicone resin having a relatively low elastic modulus and a relatively high heat resistance. In addition, the first resin layer 142 and the second resin layer 146 may further include the adhesive layer 144 of the first embodiment. The joint 40 may contain the inorganic filler described above. In particular, aluminum nitride, aluminum oxide, and silicon carbide are desirable due to their relatively high thermal conductivity. The first resin layer 142 and the second resin layer 146 are preferably formed by using a material selected from the group consisting of an adhesive, an inorganic filler, and a combination of the adhesive and an inorganic filler. When an inorganic filler is contained, the content of the inorganic filler may be different from each other.

[0048] The inorganic sheet 144 is made of an inorganic material, and is, for example, Similar ceramics as previously mentioned (e.g., aluminum oxide, aluminum nitride, etc.) are used as the main components. However, the inorganic sheet 144 and the ceramic portion 20 may be formed as different components. The inorganic sheet 144 may be formed using various kinds of ceramics. In addition to or instead of ceramics, inorganic materials other than ceramics, such as carbon fibers, Alternatively, the inorganic sheet 144 may be made of, for example, hysteresis loop. The data may be configured as

[0049] In the second embodiment, the joint 140 as a whole has a maximum shear stress strain amount and a thermal resistance The thicknesses of the first resin layer 142 and the second resin layer 146 are set so that the value of falls within the above-mentioned numerical range. , the thickness of the inorganic sheet 144, the thickness of the resin constituting the first resin layer 142 and the second resin layer 146, The type, the content of the inorganic filler in the first resin layer 142 and the second resin layer 146 (the inorganic filler content For example, the maximum shear stress of the joint 140 may be set as follows: The amount of strain when a force is applied increases as the first resin layer 142 and the second resin layer 146 are made thicker. In addition, the strain amount at maximum shear stress of the first resin layer 142 and the second resin layer 146 is increased. The thermal resistance of the joint 140 increases as the temperature of the first resin layer 142 and the inorganic sheet 1 increases. The smaller the thermal resistance of each of the first resin layer 44 and the second resin layer 146 is, the smaller the capacitance becomes.

[0050] Here, the thermal resistances of the first resin layer 142, the inorganic sheet 144, and the second resin layer 146 are The resistance is calculated by the following formula (2). In formula (2), R n represents the thermal resistance of each layer , t n represents the thickness of each layer, and λ n represents the thermal conductivity of each layer. However, in equation (2), The first resin layer 142 has n=1, the inorganic sheet 144 has n=2, and the second resin layer 146 has n=1. For n=3,

[0051] R n (m 2 K / W)=t n (m)÷λ n (W / mK) …(2)

[0052] The thermal resistance R of the entire joint 140 is calculated by the following formula (3). In the above, N represents the total number of resin layers and inorganic sheets, and the joint 1 of the second embodiment In 40, N=3.

[0053] R(m 2 K / W)=R1+R2+…+R N …(3)

[0054] With this configuration, the thermal resistance of the entire joint 140 and the strain amount at maximum shear stress are By setting the value of in the above-mentioned numerical range, the same effects as those of the first embodiment can be obtained. Since the joint portion 140 includes the inorganic sheet 144 having a higher thermal conductivity than the resin layer, the first resin The total thickness of the layer 142 and the second resin layer 146 is set to be greater than the thickness of the joint 40 in the first embodiment. This is because, when the inorganic sheet 144 is provided, the thickness of the inorganic sheet 14 can be reduced. 4 also has thermal resistance, so the thermal resistance of the entire joint 140 must be kept within the above-mentioned numerical range, and the resin The thermal resistance of the first resin layer 142 and the second resin layer 146 is higher than that of the same thermal resistance obtained by the layer alone. As a result, the resistance of the first resin layer 142 and the second resin layer 146 can be reduced. This is because the total thickness of the joint 140 can be made thinner. The surface area of ​​the adhesive layer exposed at the bonding surface (the surface of the adhesive layer) is reduced, and the plasma resistance of the bonding portion 140 is improved. Furthermore, according to the second embodiment, the bonding portion 140 includes an inorganic sheet 144. This makes it possible to make the thermal resistance and the amount of strain at maximum shear stress uniform within the surface of the joint 140. At the same time, the temperature distribution within the surface of the joint portion 140 can be made uniform.

[0055] In the second embodiment, the bonding portion 140 includes one layer of the inorganic sheet 144. However, a plurality of inorganic sheets may be provided. In this case, between each of the inorganic sheets, It is sufficient to dispose a resin layer similar to the first resin layer 142 and the second resin layer 146. Even if the ceramic portion 20 is provided with a plurality of inorganic sheets, the ceramic portion 20 in contact with the ceramic portion 20 at the joint portion is The layer including the surface in contact with the base portion 30 and the layer including the surface in contact with the base portion 30 may be a resin layer. In the case where the inorganic sheets are provided, at least one of the inorganic sheets is different from the other inorganic sheets. Even when a plurality of inorganic sheets are provided, the bonding If the thermal resistance and strain at maximum shear stress for the entire part are within the ranges already mentioned, In the case where the joint portion includes a plurality of inorganic sheets, each inorganic sheet may be formed on the joint portion. The thermal resistance of the sheet and the resin layer can be calculated by the above formula (2). The resistance can be calculated using the above formula (3).

[0056] C. Other embodiments: The present disclosure is applicable to holding devices other than electrostatic chucks that use electrostatic attraction to hold a wafer W. That is, a plate-shaped portion, a base portion, and a joint portion for joining the plate-shaped portion and the base portion may be provided. Other holding devices for holding an object on the surface of a ceramic part, such as CVD, P The present invention can also be applied to heater devices for vacuum devices such as VD and PLD, and vacuum chucks.

[0057] In the above-described embodiments, the plate-shaped portion having the mounting surface is made of ceramic as a main component. Although the ceramic portion 20 is used, the plate-shaped portion may be made mainly of a material other than ceramic. The base portion 30 may have a thermal expansion coefficient different from the thermal conductivity of the material that mainly constitutes the plate-shaped portion. The plate-shaped portion, the base portion, and the plate-shaped portion may be made of a material having a certain elongation ratio. and a joint portion that joins the base portion to the holder. By applying the disclosed configuration, the cooling performance of the holding device can be improved to prevent the generation of heat at the joint. The same effect of suppressing stress and suppressing damage to the joint even if shear stress occurs at the joint is obtained. EXAMPLES

[0058] The following describes the holding device of the present disclosure based on an embodiment. As samples corresponding to various joints with different strains at maximum shear stress, samples Sheet-shaped samples S1 to S33 were prepared. Samples in the form of electrostatic chucks having joints with the same composition as those in samples S1 to S33 were fabricated. Below, the electrostatic chuck-type sample was also tested using a sheet-shaped sample with the same composition as the joint. The samples are called by the same sample number as the sample.

[0059] Figures 8 to 11 show the thermal resistance of the joints of each sample, the amount of strain at maximum shear stress, thickness, and thermal Along with the conductivity value, the cooling performance of the ceramic part through the joint and the susceptibility of the joint to damage are also measured. 10 and 11 are explanatory diagrams showing the results of evaluation of the fraying (peeling). Tensile tests were then conducted on the joints of samples S24 to S33 to determine the deformation caused by shear force. Shear stress when the length becomes 0.5 mm (shear stress when the shear strain is 0.5 mm) The values ​​of the flatness and the flatness of the ceramic part 20 of the electrostatic chuck are also shown. FIG. 10 also shows the rigidity modulus of samples S24 to S29. In addition, in Fig. 11, the inorganic filler "R2 / R1" in the joints of samples S30 and S31 is The value of " is also shown.

[0060] <Preparation of each sample> The electrostatic chucks of the samples S1 to S32 are all the electrostatic chucks of the first embodiment shown in FIG. The electrostatic chuck of sample S33 has a configuration similar to that of the electrostatic chuck 10. The electrostatic chuck 110 has a similar configuration to that of the electrostatic chuck 110 of the second embodiment. This paper describes the method for preparing the sheet-shaped sample and the electrostatic chuck-shaped sample, which are common to all the samples. Then, the materials and preparation conditions of each sample will be explained.

[0061] [Preparation of sheet-like samples] The sheet-shaped sample was prepared as follows: A resin material, which is an adhesive before hardening, was (100 parts by weight) and inorganic filler, platinum catalyst (platinum content 0.003 parts by weight) The constituent materials include a silane coupling agent (2 parts by weight), and a crosslinking agent (3 parts by weight). The mixture is stirred under vacuum using a vacuum degassing mixer, and then kneaded using a three-roll mill. A paste-like resin composition (adhesive paste) was prepared. , and coated onto polyethylene terephthalate (PET) film using a doctor blade. Next, the PET film with the adhesive paste spread on it was cut, and then the cut The adhesive paste with the PET film is dried in a dryer at an appropriately set temperature and time. Each sample was semi-cured or cured by heating under conditions between 100 and 200°C. The adhesive sheet was then prepared.

[0062] In each sample, silicone resin was used as the adhesive. The adhesive material (resin material) used was vinyl-terminated polydimethylsiloxane. In order to obtain the above, dimethyldichloro was used as a raw material for the bifunctional D unit having two organic substituents. We used silane ((CH3)2SiCl2). In addition, a monofunctional silane with three organic substituents was used. As raw materials for the M unit, trimethylchlorosilane ((CH3)3SiCl) and hexamethion We used rudisiloxane ((CH3)3SiOSi(CH3)3). In addition, In order to introduce the group, dimethylvinylchlorosilane ( (CH3)2(CH2=CH)SiCl) is used to introduce reactive functional groups into the molecular chain. To this end, CH3(CH2=CH)SiCl2, the raw material for the functionalized D unit, and , (CH3)2(CH2=CH)SiCl was used. Samples S1 to S25, S30 to S Samples S26 to S29 are made of the same resin material, and the raw materials of the D units are The resin material used was one in which the mixing ratio of the raw materials of the M units was different from that of the above sample S1 and the like. By varying the mixing ratio of D unit raw material and M unit raw material, the average molecular weight of the resin material can be reduced. In this way, the content (density) of reactive functional groups in the resin material was increased. The average molecular weight of the resin materials S1 to S25 and S30 to S33 is 63,000. The average molecular weight of the resin material of sample S26 is 33,000, and the average molecular weight of the resin material of sample S27 is 33,000. The average molecular weight of the resin material in sample S28 is 18,000. The average molecular weight of the resin material of sample S29 is 15,000, and the reactivity is in this order. High functional group content.

[0063] [Preparation of electrostatic chuck-type sample] The sample in the form of the electrostatic chuck 10 was prepared by applying the semi-cured adhesive paste described above to a ceramic substrate. The adhesive paste is then placed between the backing portion 20 and the base portion 30 and allowed to harden. The sample in the form of the electrostatic chuck 110 was prepared by the semi-cured adhesive paste described above. An inorganic sheet 144 mainly composed of aluminum oxide is sandwiched between two sheets of paper. Place the adhesive paste between the ceramic part 20 and the base part 30 and then allow the adhesive paste to harden. The ceramic portion 20 is made of aluminum oxide. The ceramic part (thermal expansion coefficient 7 ppm / K) is used, and the base part 30 is made of aluminum. The base part (thermal expansion coefficient: 23 ppm / K) was made of ceramic part 20 The diameter of the joint was 350 mm.

[0064] [Samples S1-S5] Samples S1 to S5 all use the same resin material as described above, and contain inorganic fillers. Aluminum nitride (average particle size: 10 μm) was used. The curing temperature was 150° C. in all cases. In addition, when the resin was cured in samples S1 to S5, The time period was 10 hours for sample S1 and 10 hours for sample S2 so that the strain at maximum shear stress was the same. 11 hours for S2, 15 hours for S3, 16 hours for S4, and Samples S1 to S5 have different amounts (contents) of inorganic filler added. Specifically, the inorganic filler content is 32% by volume in sample S1 and 32% by volume in sample S2. Sample S2 was 28.5% by volume, sample S3 was 21% by volume, and sample S4 was 20% by volume. % by volume, and 11.5% by volume for sample S5. The higher the ratio, the greater the thermal conductivity of the joint and the smaller the thermal resistance.

[0065] [Samples S6-S8] Samples S6 to S8 all use the same resin material as described above, but the curing temperature is By changing the temperature, the distance between the crosslinking points in the resin is changed. Sample S6 was cured at 155°C, sample S7 at 165°C, and sample S8 at 170°C. The curing time was 10 hours in both cases. The higher the curing temperature, the higher the crosslink density and the greater the distance between crosslink points. As shown in Fig. 8, the higher the curing temperature, the smaller the maximum shear stress. The amount of strain was small when the sample was filled with inorganic filler. The aluminum nitride contained in the samples S1 to S5 is the same as that in the samples S6 to S7. In all of the samples S1 and S8, the content was 32% by volume.

[0066] [Sample S9~S20] Samples S9 to S20 all use the same resin material as described above, but the curing temperature By changing the thickness, the distance between the crosslinking points in the resin is changed. The curing temperature of samples S9, S12, S15, and S18 was 155°C, and that of samples S10, S13, and S1 6. The curing temperature of S19 is 165℃, and the curing temperatures of samples S11, S14, S17, and S20 The curing temperature was set to 170°C. As shown in Figures 8 and 9, the higher the curing temperature, the greater the maximum shear stress. The amount of strain during curing was small. The curing time was 11 hours for samples S9 to S11, and Samples S12 to S14 were 15 hours, samples S15 to S17 were 16 hours, and samples S18 to The S20 was 28 hours.

[0067] In addition, samples S9 to S20 contain the same nitride as samples S1 to S5 as an inorganic filler. The amount of inorganic filler added was 28.5% by volume for samples S9 to S11. ,Samples S12 to S14 are 21 vol%,Samples S15 to S17 are 21 vol%,Samples As shown in FIG. 8 and FIG. 9, the inorganic filler S18 to S20 were 11.5% by volume. The greater the amount of added, the greater the thermal conductivity of the joint and the smaller the thermal resistance.

[0068] [Sample S21~S23] Sample S21 is the same as sample S1. Samples S22 and S23 are Only the thickness was different from sample S21 (S1). The thickness of the joint was different from that of sample S21. (S1) is 0.3 mm, sample S22 is 0.4 mm, and sample S23 is 0.4 mm. As shown in Fig. 9, the thicker the joint, the greater the strain at maximum shear stress. The quantity became larger.

[0069] [Sample S24~S29] Sample S24 is the same as sample S1. Samples S25 to S29 are inorganic fillers. The sample S24 (S1) contains aluminum nitride as an inorganic filler, but the inorganic filler The amount of addition was set to 44.4% by volume, which is higher than that of sample S24 (S1). As shown in FIG. 10, in samples S25 to S29, compared to sample S24 (S1), The thermal conductivity of the joint increased and the thermal resistance decreased.

[0070] In addition, samples S25 to S29 differ in the manner of the adhesive that constitutes the joint. As mentioned above, sample S25 uses the same resin material as sample S24 (S1). The curing conditions are 140℃ and 10 hours, and the curing temperature is low. Therefore, the crosslink density is reduced. As shown in Figure 10, the distance between the crosslinking points is thought to have increased. ) the "shear stress at a shear strain of 0.5 mm" was larger. Samples S26 to S29 were cured under the same conditions as sample S25, but as mentioned above, the resin The density of functional groups that serve as crosslinking points for the silicone resin is changed, resulting in The distance between the points is different. As shown in Figure 10, the higher the density of reactive functional groups, In other words, in the order of samples S25 to S29, the shear stress at a shear strain of 0.5 mm was " and rigidity modulus values ​​increased.

[0071] [Sample S30~S32] Samples S30 to S32 all use the same resin material as described above, but the inorganic filler is used. The types of inorganic fillers are different. Sample S30 is aluminum nitride. S31 uses aluminum oxide, and S32 uses boron nitride. was used.

[0072] The inorganic filler contained in sample S30 is the same as that of aluminum nitride shown in FIG. The inorganic filler contained in sample S31 is the same as the aluminum oxide shown in FIG. In samples S30 and S31, the average particle size of the inorganic filler was 10 μm. The content of the inorganic filler in the bonding portion 40 was 44% by volume in all cases.

[0073] In sample S32, the average particle size of the plate-shaped boron nitride, which is an inorganic filler, is 2 μm. When boron nitride is used as an inorganic filler, it is preferable to use aluminum nitride or aluminum oxide. The viscosity of the adhesive paste tends to be higher than when using miniature. It was difficult to increase the content of inorganic filler, which is a component of the composite material that can form a joint. The inorganic filler content was reduced to the range, and the joint of sample S32 was filled with inorganic filler. The content of boron nitride was set to 16% by volume.

[0074] [Sample S33] As described above, the electrostatic chuck of the sample S33 is the electrostatic chuck of the second embodiment shown in FIG. In sample S33, the first resin layer 142 and the second resin layer 143 are the same in structure as the first resin layer 142 and the second resin layer 143. The second resin layer 146 is made of the same resin material as that of the sample S25, and the curing time and the like are set. The content ratio of the organic filler and the inorganic filler was the same as that of sample S25. The first resin layer 142 and the second resin layer 146 each have a thickness of 0.15 mm. The conductivity is 1.44 W / mK and the thermal resistance is 1.04 × 10 -4 m 2 K / W. In the sample S33, the inorganic sheet 144 is made of aluminum oxide (alumina: Al2 O3), with a thickness of 1.0 mm, thermal conductivity of 30 W / mK, and thermal resistance of Resistance is 0.33 x 10 -4 m 2 The name was K / W.

[0075] <Measurement of thermal conductivity> The thermal conductivity of the sheet sample was measured using a known thermal conductivity meter (Kyoto Electronics Manufacturing Co., Ltd. The measurements were performed using a fast thermal conductivity meter QTM-710. When measuring the thermal conductivity of a joint built into an electrostatic chuck, The ceramic part of the chuck is removed with a surface grinder, etc. to expose the joint, and then a knife, etc. is used to The bond can be stripped using a thermal conductivity tester and the thermal conductivity can be measured.

[0076] <Calculation of thermal resistance> The thermal resistance R is calculated as follows using the measured thermal conductivity of each sample as described above. In equation (1), t is the thickness of the joint, λ is the thermal Conductivity, i.e., the thermal conductivity of each sample, measured as described above.

[0077] R(m 2 K / W) = t(m) ÷ λ(W / mK) … (1)

[0078] <Measurement of maximum shear stress> The maximum shear stress and the strain at maximum shear stress described later were measured using a known tensile testing machine ( The measurements were performed by tensile testing using an autograph (AGS-5kNX) manufactured by Shimadzu Corporation.

[0079] FIG. 12 is an explanatory diagram showing a schematic diagram of a method for calculating the maximum shear stress and the amount of strain. 12(A) shows the front view of the tensile test, and Figs. 12(B) and 12(C) show 12(A) and 12(B) show the state at the start of the test. FIG. 12(C) shows the state after the start of the test. 0 is a semi-cured adhesive sheet of each sample, 25 mm wide x 100 mm long x 1 mm thick. 25 mm x 12.5 mm, from the edge of the two aluminum plates 72 to a position 12.5 mm away The two aluminum plates 72 are attached to the respective portions and pulled in opposite directions. After pasting them together in the correct orientation, the semi-cured adhesive sheet is cured to create the product. The thickness t of the test piece 70 at the start of the test was set to the thickness of the joint shown in Figs. Then, the two aluminum plates 72 are moved relative to each other so that a shear force acts on the test piece. Here, a tensile tester was used to tensile test one of the aluminum plates in one direction parallel to the adhesive surface. The load and the strain δ were measured as a function of the distance traveled while moving the specimen at a tensile speed of 2 mm / min. (See Fig. 12(C)). In Fig. 12(B), the relative movement of the two aluminum plates is shown. The direction of the load is indicated by a white arrow. The bond area of ​​the test piece before the load was transferred (25 mm x 12. The shear stress was calculated by dividing the thickness by the thickness of the aluminum plate. The relative movement is continued until the test piece 70 breaks, and the shear stress at the time when the shear stress becomes maximum is measured. The force was defined as the maximum shear stress (unit: MPa).

[0080] <Measurement of strain at maximum shear stress> The strain at maximum shear stress (unit: mm) was measured in the tensile test shown in Figure 12. The strain amount when the shear stress reached its maximum was defined as δ.

[0081] FIG. 13 shows an example of the measured values ​​in a tensile test, showing the strain δ and FIG. 13 is an explanatory diagram showing the relationship with shear stress. In FIG. 13, the horizontal axis shows the amount of strain (mm), The vertical axis shows the shear stress (MPa). For each sample, as shown in Figure 13, Based on the results of the tensile test, the shear stress is determined to be the maximum shear stress τ max The amount of strain when δa was calculated.

[0082] <Measurement of shear stress when shear strain is 0.5 mm> The shear stress at a shear strain of 0.5 mm is shown in the tensile test shown in Figure 12. This refers to the shear stress when the amount δ is 0.5 mm. As shown in the figure, based on the results of the tensile test, the shear stress when the strain amount δ becomes 0.5 mm is force τ 0.5 asked for.

[0083] <Rigidity> The rigidity modulus G is calculated by dividing the shear stress τ by the shear strain γ as shown in the following formula (4): The shear strain γ is calculated by dividing the strain amount δ by the sample before the test as shown in the following formula (5). This is the value divided by the sample thickness t0.

[0084] G(Mpa)=τ(MPa)÷γ(-) … (4) γ(-)=δ(mm)÷t0(mm) … (5)

[0085] Figure 14, like Figure 13, shows the results of sample S25 as an example of actual measurements from a tensile test. FIG. 14 is an explanatory diagram showing the strain instead of the strain amount (mm) on the horizontal axis. The strain shown on the horizontal axis of Fig. 14 is multiplied by 1 to obtain the shear strain γ in the above formula (5). The modulus of rigidity G is expressed as a percentage by multiplying it by 00. It is a value calculated from the gradient of the lines of force. As an example of the rigidity modulus G for each sample, The calculation results for sample S25 are shown below, using values ​​when the strain amount δ is 0.5 mm. The thickness t0 of sample S25 before the test was 0.3 mm (see Figure 10), so the amount of strain was Shear strain γ when δ is 0.5 mm 0.5 From equation (5), we get "0.5÷0.3≒1. 667 (166.7%)". And the shear stress when the strain δ is 0.5 mm is τ 0.5 is τ 0.5 = 0.90 (see Fig. 10), the rigidity modulus G is calculated from equation (4). "0.90 ÷ 1.667 ≒ 0.54".

[0086] In addition, the maximum shear stress of the joint already built into the electrostatic chuck When measuring the strain under force, the shear stress when the shear strain is 0.5 mm, and the rigidity For example, the following procedure is used to remove the joints: The specimen is cut out together with the adherend (ceramic part and base part). The shape of the test piece to be cut out is the same as that of the tensile test. Two adherends that can be held by the fixture of the testing machine and are joined are shown in Figure 12. The shape of the material must be such that it can be pulled in opposite directions. The area and thickness of the joint of the cut-out test piece are measured. A tensile test was carried out in the same manner as described above, and the maximum shear stress, strain at maximum shear stress, and The shear stress and rigidity modulus at a shear strain of 0.5 mm are measured.

[0087] <Evaluation of cooling performance of ceramic parts> For each of the electrostatic chuck samples S1 to S33, cooling of the ceramic part through the joint Specifically, the initial temperature of the ceramic part of each sample was set to 150°C. The ceramic part was heated to 150°C. The base part was cooled to -10°C. A coolant was supplied to cool the base part. After the ceramic part reached 150°C, Heating of the laminated portion was stopped, and the time until the ceramic portion was cooled to 20°C was measured. If it takes more than 30 seconds for the ceramic part to cool to 20°C, the cooling performance is poor ( When the cooling time was 25 seconds or less, the cooling performance was evaluated as excellent (◎). If the time it takes for the ceramic part to cool to 20°C is longer than 25 seconds but less than 30 seconds was rated as having good cooling performance (○).

[0088] <Evaluation of peeling of joints> A peeling test was performed on each of the electrostatic chuck samples S1 to S33 to check for peeling of the joint. Specifically, each sample was placed in a commercially available heat cycle tester and heated to a maximum temperature of 1 100 cycles of thermal cycle testing at 50℃ and a minimum temperature of 0℃ were performed, and then the ceramic was returned to room temperature. The presence or absence of peeling at the joint between the mix part and the base part was evaluated. The presence or absence of defects was determined using a known ultrasonic flaw detector. Ultrasonic waves were irradiated from the ceramic part side, A reflected echo (flaw echo) was detected from a depth corresponding to the joint interface between the ceramic part and the joint. If so, it was determined that there was peeling at the bonding interface between the ceramic part and the bonding part. Ultrasonic waves were irradiated from the base part side, and when a reflected echo (defect echo) was detected from the depth corresponding to the bonding interface between the base and the bonding part, it was determined that there was peeling at the bonding interface between the base part and the bonding part. For at least one of the above two interfaces where peeling was observed, peeling was evaluated as "present", and for those where no peeling was observed by ultrasonic flaw detection, peeling was evaluated as "absent".

[0089] <Calculation of R2 / R1> For each sample in the form of the electrostatic chuck 10, in order to calculate the value of "R2 / R1", an arbitrary location was cut out in the X-Y direction shown in FIG. 1 with a size of 50 mm × 50 mm and penetrated in the Z direction, and the part of the bonding part 40 was taken out from the obtained test piece with a cutter. After mirror-polishing the cross-section of the cut-out bonding part, it was observed using a scanning electron microscope (SEM). For 10 particles of inorganic fillers observed in a field of view of 2 00 μm × 200 μm, their shapes were observed. Using the image analysis software WinROOF, the radius R1 of the inscribed circle and the radius R2 of the circumscribed circle were measured to obtain the value of "R2 / R1" (see FIG. 3), and the value of "R2 / R1" for each sample was calculated as the average value of 10 particles.

[0090] <Flatness> The definition of flatness is as described in JIS B0621 "Definition and Representation of Geometric Deviations", which is the magnitude of the deviation from the geometrically correct plane of a planar body. When the use surface is sandwiched between two geometrically correct parallel planes, the dimension of the interval where the interval between the two parallel planes is minimized was defined as the flatness. Specifically for the measurement method of flatness, a known three-dimensional measuring machine can be used for measurement. Here, the Keyence-made one-shot 3D shape measuring machine "VR series" was used. For the above two parallel planes ​​​​​​ When the dimension of the smallest gap is less than 50 μm, the flatness is very good ( When the thickness was 200 μm or more, the flatness was evaluated as low (×). When the thickness was less than 200 μm, the flatness was evaluated as good (◯).

[0091] As shown in Figs. 8 to 11, the thermal resistance of the junction is 5.0 × 10 -4 (m 2 K / W) or less Sometimes, the ceramic part is evaluated as having good cooling performance (○), and the thermal resistance of the joint is Resistance is 3.5×10-4(m 2 K / W) or less, the cooling performance is excellent (◎) It was also confirmed that the strain at the maximum shear stress of the joint was 0.5m m or more, no peeling occurs at the joint, and distortion at the joint is suppressed and the joint It was confirmed that the effect of suppressing damage to the joint caused by the strain of the joint was sufficient. If we set the strain at maximum shear stress to 2, it is especially desirable for the strain to be 1.0 mm or more. It is possible.

[0092] The indicators for ensuring the flexibility of the joints in order to improve the stress relaxation performance at the joints are as follows: In addition to the strain at maximum shear stress, the shear stress at 0.5 mm shear strain and the joint As mentioned above, in each sample shown in Fig. 10, the rigidity of the joint is The curing temperature of the resin, the content of inorganic filler, and the density of reactive functional groups in the resin are varied. As shown in Figure 10, the values ​​of the above indicators are different. The shear stress at a breaking strain of 0.5 mm must be 3.5 MPa or less, or the rigidity of the joint must be It was confirmed that the flatness of the ceramic part was improved by setting the pressure to 2.0 MPa or less. In particular, the "shear stress at a shear strain of 0.5 mm" is set to 2.0 MPa or less, or By setting the rigidity of the joint at 1.2MPa or less, the flatness of the ceramic part is improved. It has been confirmed that this will be the case.

[0093] In addition, as can be seen by comparing sample S30 and sample S31 in FIG. If the content of the chromium is the same, the R2 / R1 value is 1.10 or more and the sphericity is lower. The sample S30, which has a lower thermal conductivity, has a lower thermal resistance at the junction. However, as shown in Fig. 11 as sample S32, the inorganic filler When using boron nitride, which is a plate-like particle, the sphericity of the inorganic filler is considered to be lower. However, as mentioned above, the viscosity of the adhesive paste increases and the inorganic filler content increases. As a result, the thermal resistance is also relatively high at 5.0×10 -4 ( m 2 K / W).

[0094] Although data is not shown in FIG. 11, in addition to samples S30 to S32, carbon nanotubes We also prepared samples using tubes as inorganic fillers. This is because the thermal conductivity of the joint is known to be very large. The thermal conductivity value was further reduced (0.27 W / mK). When using inorganic filler for the tube, the inorganic filler is in the form of fibers, so the adhesive paste It is believed that the amount of inorganic filler that can be mixed into the resin is limited, which is why the thermal conductivity did not increase. In other words, fibrous inorganic fillers have a higher mechanical strength than granular or spherical inorganic fillers. This is thought to be because the amount of the material that could be blended was limited due to the low fluidity of the material itself.

[0095] The present disclosure is not limited to the above-described embodiments, and may be modified without departing from the spirit and scope of the present disclosure. For example, the above-mentioned embodiments described in the Summary of the Invention can be realized in various configurations. The technical features in the embodiments corresponding to the technical features solve some or all of the above problems. In order to achieve some or all of the above effects, the following may be replaced or recombined as appropriate. In addition, the technical features are not essential in this specification. If it is not explained in the text, it may be deleted as appropriate. [Explanation of symbols]

[0096] 10,110...Electrostatic chuck 20…Ceramic section 22...Adsorption electrode 24...Placement surface 30…Base section 32... Coolant flow path 40,140...Joint part 42…Inorganic filler 50…Gas supply line 52…Gas outlet 70...Test piece 72…Aluminum plate 142...First resin layer 144...Inorganic sheet 146…Second resin layer

Claims

1. A holding device for holding an object, A plate-shaped portion formed in a plate shape; a base portion formed in a plate shape, which supports the plate-shaped portion, is mainly made of a material having a thermal expansion coefficient different from that of a material mainly forming the plate-shaped portion, has a cooling function, and a joining portion disposed between the plate-shaped portion and the base portion, the joining portion including an adhesive and joining the plate-shaped portion and the base portion; Equipped with The thermal resistance of the junction is 5.0×10 －４ (m ２ K / W) or less, The strain amount at the maximum shear stress of the joint is 0.5 mm or more, The thermal conductivity of the joint is 0.8 W / mK or more, The adhesive is characterized in that it is an adhesive material obtained by crosslinking vinyl-terminated polydimethylsiloxane consisting of bifunctional D units having two organic substituents and monofunctional M units having three organic substituents. holding device.

2. 2. The holding device according to claim 1, The joint is characterized in that a tensile test is performed on the joint, and the shear stress when the deformation length due to shear force becomes 0.5 mm is 0.1 MPa or more and 3.5 MPa or less. holding device.

3. 3. A holding device according to claim 1 or 2, The bonding portion is formed of a composite material containing the adhesive and an inorganic filler. holding device.

4. 4. The holding device according to claim 3, The inorganic filler is characterized in that, when the radius of an inscribed circle in a cross-sectional view of the inorganic filler is R1 and the radius of a circumscribed circle is R2, the average value of R2 / R1 is 1.10 or more. holding device.

5. A holding device according to any one of claims 1 to 4, The rigidity modulus of the joint is 0.03 MPa or more and 2.0 MPa or less. holding device.

6. A holding device according to any one of claims 1 to 5, The bonding portion is characterized by comprising: a resin layer containing the adhesive; and an inorganic sheet formed in the resin layer so as to extend in a surface direction of the bonding portion and made of an inorganic material. holding device.

7. A holding device according to any one of claims 1 to 6, The plate-shaped portion is mainly made of ceramic, includes an attraction electrode for holding the object, and does not include a heater electrode for heating the object. holding device.

Images