JP2020038896A - Junction structure and method for manufacturing the same - Google Patents

Abstract

To provide a junction structure capable of preventing the characteristic change and breakage of a second adherend by preventing a portion of the second adherend facing a first adherend from forming uneven stress to provide excellent operation reliability to the second adherend, and a method for manufacturing the junction structure.SOLUTION: A junction structure 1 comprises a first adherend 11, a metal porous body 13 formed on the first adherend and a second adherend 12 arranged on the metal porous body. The coefficient of variation of stress in a portion of the second adherend facing the first adherend is within 10%.SELECTED DRAWING: Figure 1

Description

  The present invention provides a first adherend (for example, an electronic component such as a semiconductor package or a semiconductor chip), a metal porous body formed on the first adherend, and a metal porous body formed on the first porous body. The present invention relates to a joint structure including a second adherend (for example, another electronic component having a metal substrate, an electrode, and the like) disposed therein, and a method for manufacturing the joint structure.

  Conventionally, as a material for bonding electronic parts, a dispersion solution of metal fine particles (a paste of metal fine particles) obtained by dispersing metal fine particles (for example, copper fine particles) in a nano-size particle size range in an organic solvent has been used. . As a method of bonding an electronic component to another electronic component or a metal substrate, for example, a dispersion solution of metal fine particles is applied to the bonding surface between the electronic component and another electronic component or the metal substrate, and the metal fine particles are heated. Sintering to form a joint between the electronic component and another electronic component or a metal substrate to join the electronic component and the metal substrate.

  Since the joining portion is a sintered body of metal fine particles, there is an advantage that the thermal conductivity is high, the mechanical strength is high even at a high temperature, and the joining reliability is excellent. However, since the sintered body of the metal fine particles is harder than the solder, the stress applied to the electronic component as the adherend tends to increase due to the difference in the coefficient of thermal expansion between the joint and the electronic component as the adherend. It is in. When the stress applied to the electronic component increases, the characteristics of the electronic component change, and the reliability of the operation of the electronic component may decrease.

  Therefore, the internal stress of the silicon wafer is reduced by a bonding method in which the thermal expansion coefficient of the porous sintered body forming the bonding portion and the substrate, which is a silicon wafer on which a large number of devices are integrated, does not exceed three times. It has been proposed to prevent destruction of electronic components (Patent Document 1).

  However, in some cases, even if the coefficient of thermal expansion between the joint and the electronic component is controlled as in Patent Literature 1, the characteristics and operation reliability of the electronic component cannot be sufficiently obtained. In particular, a power device can be considered as small devices connected in parallel. Therefore, when a stress from a joint applied to some small devices increases, a characteristic difference occurs between the devices. For example, when a device having a threshold voltage lower than a set value is generated, the device is turned on first. Then, a current flows intensively into the device which is turned on first, so that the temperature of the device rises, and as a result, the device may be destroyed. Therefore, in order to prevent destruction of the device, conventionally, the gate resistance has been made uniform by using a structure such as a gate runner. The characteristics of a semiconductor device are sensitive to pressure, and the characteristics of the semiconductor device may change under the influence of stress from a joint. In particular, if the stress exerted from the junction to the device is not uniform (the stress is strong in some regions of the device and the stress is weak in other regions), the characteristics of the device will be non-uniform depending on the site, In addition, the device may be destroyed for the above reasons.

JP 2015-23186 A

  In view of the above circumstances, the present invention prevents the non-uniformity of the stress in the portion of the second adherend facing the first adherend, thereby preventing a change in the characteristics of the second adherend. It is an object of the present invention to provide a joint structure in which destruction is prevented and which can impart excellent operational reliability to the second adherend, and a method for manufacturing the joint structure.

The gist configuration of the present invention is as follows.
[1] A bonding structure including a first adherend, a metal porous body formed on the first adherend, and a second adherend arranged on the porous metal body Body
A joint structure in which a coefficient of variation of stress in a portion of the second adherend facing the first adherend is within 10%.
[2] The joined structure according to [1], wherein the second adherend is an electronic member.
[3] The bonding structure according to [2], wherein the electronic member is a silicon chip.
[4] The bonding structure according to any one of [1] to [3], wherein the first adherend is a metal substrate or another electronic member.
[5] The bonding according to any one of [1] to [4], wherein the metal porous body is a sintered body of metal fine particles, and the average primary particle diameter of the metal fine particles is 2 nm or more and 500 nm or less. Structure.
[6] The bonding structure according to [5], wherein the metal fine particles include copper.
[7] A bonding structure including a first adherend, a metal porous body formed on the first adherend, and a second adherend disposed on the metal porous body. A method of manufacturing a body,
A step of applying a paste containing metal fine particles between a first member to be the first adherend and a second member to be the second adherend, to form a composite;
Heating the composite at a heating rate of 100 ° C./min or more and less than 5000 ° C./min to sinter a paste containing metal fine particles to form a metal porous body;
A method for manufacturing a joined structure including:
[8] The method for manufacturing a joined structure according to [7], wherein the heating rate is 4000 ° C./min or less.
[9] The bonding according to [7] or [8], wherein the paste containing the metal fine particles includes metal fine particles having an average primary particle diameter of 2 nm or more and 500 nm or less, and an organic solvent containing a divalent or higher alcohol. The method of manufacturing the structure.
[10] The method for manufacturing a joint structure according to any one of [7] to [9], wherein the metal fine particles include copper.
[11] The dihydric or higher alcohol is glycerin, 1,1,1-tris (hydroxymethyl) ethane, 1,2,4-butanetriol, 1,2,3-hexanetriol, 1,2,6- Production of the joined structure according to [9] or [10], including at least one trihydric alcohol selected from the group consisting of hexanetriol and 2-ethyl-2-hydroxymethyl-1,3-propanediol. Method.
[12] The method according to [9] or [10], wherein the dihydric or higher alcohol contains glycerin.

  According to the aspect of the present invention, the second adherend has a joint structure in which the variation coefficient of the stress in the portion of the second adherend facing the first adherend is within 10%. However, since the non-uniformity of the stress in the portion facing the first adherend is prevented, the characteristic change and breakage of the second adherend are prevented, and the second adherend is excellent. Operational reliability can be provided.

  According to the aspect of the present invention, the paste containing the metal fine particles applied between the first member to be the first adherend and the second member to be the second adherend is heated at a heating rate. Heating at a temperature of 100 ° C./min or more and less than 5000 ° C./min to sinter the paste containing the metal fine particles to form a porous metal body at a joint between the first adherend and the second adherend; Thereby, the bonding can be performed in a state in which the stress on the second adherend is not made nonuniform. Therefore, when joining the second adherend to the first adherend, it is possible to prevent the characteristic change and destruction of the second adherend, and to ensure excellent operation reliability of the second adherend. Can be provided.

FIG. 2 is an explanatory cross-sectional view of the joint structure according to the embodiment of the present invention.

  Next, the joint structure of the present invention will be described in detail. As shown in FIG. 1, the joint structure 1 of the present invention includes a first adherend 11, a metal porous body 13 formed on the first adherend 11, And a second adherend 12 disposed on the surface of the second adherend 12 facing the first adherend 11 (facing the joint portion 22 with the porous metal body 13). The coefficient of variation of stress on the (surface) is within 10%. In the joint structure 1 of the present invention, the metal porous body 13 is disposed between the first adherend 11 and the second adherend 12, and the metal porous body 13 is The joint between the body 11 and the second adherend 12 is formed. Further, the second adherend 12 has a stress from the metal porous body 13 at a portion facing the first adherend 11, that is, at a surface facing the joining portion 22 in contact with the metal porous body 13. The coefficient of variation of (F) is reduced to within 10%. Therefore, in the second adherend 12, the coefficient of variation of the stress (F) from the metal porous body 13 at the joint portion 22 is made uniform.

  The porous metal body is formed by heat-treating a paste (dispersion solution of the fine metal particles (P)) containing the fine metal particles (P), which is a joining material, to form a sintered body of the fine metal powder. The paste containing the metal fine particles (P) contains the metal fine particles (P1) and the organic solvent (S) containing a dihydric or higher alcohol. The paste containing the metal fine particles (P) is a dispersion in which the metal fine particles (P1) are dispersed in an organic solvent (S) containing a dihydric or higher alcohol as a dispersion medium.

Metal fine particles (P1)
The metal fine particles (P) of the paste containing the metal fine particles (P) include the metal fine particles (P1). The metal species of the metal fine particles (P1) is not particularly limited, and examples thereof include noble metals such as copper (Cu), gold (Au), silver (Ag), platinum (Pt), and palladium (Pd). . Among these, copper (Cu) is preferable from the viewpoint of the balance between the conductivity of the metal porous body 13 and the bonding reliability of the metal porous body 13 and the production cost. These metal species may be used alone or in combination of two or more.

  The average primary particle diameter of the metal fine particles (P1) is not particularly limited, but the lower limit thereof is to prevent an increase in manufacturing cost due to the use of advanced particle size control technology, and the metal fine particles (P1) are made of metal fine particles (P1). From the viewpoint of preventing the dispersion state from becoming non-uniform due to aggregation in the paste containing P) and preventing a decrease in storage stability, 2 nm is preferable, and 10 nm is particularly preferable. On the other hand, the upper limit of the average primary particle diameter of the metal fine particles (P1) is such that the metal surface (P1) maintains a high surface area and decomposes a dihydric or higher alcohol contained in the organic solvent (S). From the viewpoint that the catalytic effect of the fine particles (P1) is more reliably exhibited and the bonding reliability is further improved, 500 nm is preferable, and 100 nm is particularly preferable.

  Here, the “primary particles” mean individual primary particles constituting the secondary particles. The “average primary particle diameter” of the metal fine particles (P1) refers to the longest diameter of each of the primary particles of a plurality of metal fine particles (P1) selected as a measurement target, and the measurement is performed. It means the average value calculated by dividing the value by the number. Specifically, for example, using a scanning electron microscope (SEM) (manufactured by Hitachi High-Technologies Corporation, product name: SU8020), observation is performed under conditions of an acceleration voltage of 3 kV and a magnification of 200,000, and becomes a measurement target. An SEM image of the metal fine particles (P1) is obtained. From the acquired SEM images, arbitrarily select 20 primary particles of metal fine particles (P1). The longest diameter of the selected primary particles of the metal fine particles (P1) is measured, and the average of the measured values is calculated to determine the average primary particle diameter.

Other metal fine particles (P2)
The paste containing the metal fine particles (P) may further contain other metal fine particles (P2) in addition to the metal fine particles (P1). The other metal fine particles (P2) have a larger average primary particle diameter than the metal fine particles (P1). By blending the other metal fine particles (P2), the nano-sized metal fine particles (P1) can be easily interposed between the other metal fine particles (P2). As a result, the free flow of the nano-sized metal fine particles (P1) can be effectively restricted, and the metal fine particles (P1) can be stably mixed in a moderately dispersed state in the paste of the metal fine particles (P). it can.

  The average primary particle diameter of the other metal fine particles (P2) is not particularly limited as long as the average primary particle diameter is larger than the average primary particle diameter of the metal fine particles (P1). More than 0.5 μm is preferable from the viewpoint of more effective restriction. On the other hand, the upper limit of the average primary particle diameter of the other metal fine particles (P2) is such that streak-like coating defects occur during printing of the paste containing the metal fine particles (P), thereby impairing bonding reliability. Is preferably 50 μm from the viewpoint of preventing

  With respect to the other metal fine particles (P2), “average primary particle diameter” means a primary particle diameter (D50) having a cumulative volume percentage of 50% by volume as measured by a laser diffraction / scattering method.

  The metal species of the other metal fine particles (P2) is not particularly limited. For example, gold (Au), silver (Ag), platinum (Pt), palladium (Pd), titanium (Ti), iron (Fe), cobalt (Co), nickel (Ni), copper (Cu), tantalum (Ta), tungsten (W), aluminum (Al), zinc (Zn), indium (In), tin (Sn), lead (Pb), bismuth (Bi) and the like. Of these, copper (Cu) is preferred. These metal species may be used alone or in combination of two or more.

Polymer dispersant (D)
The surface of the metal fine particles (P1) (for example, the average primary particle diameter is in the range of 2 to 500 nm) may be coated with the polymer dispersant (D). By coating the surface of the metal fine particles (P1) with the polymer dispersant (D), aggregation of the metal fine particles (P1) in the paste of the metal fine particles (P) can be prevented, and the metal fine particles (P1) Dispersibility can be improved. The entire surface of the metal fine particles (P1) may be coated with the polymer dispersant (D), or a partial region of the surface may be coated with the polymer dispersant (D).

  The polymer dispersant (D) is not particularly limited as long as it is a compound that can improve the dispersibility of the metal fine particles (P1) by coating the surface of the metal fine particles (P1). -CON =). Examples of the compound having an amide group (-CON =) include, for example, N-methylacetamide, N-methylformamide, N-methylpropanamide, formamide, N, N-dimethylacetamide, 1,3-dimethyl-2-imidazolide Non, N, N-dimethylformamide, 1-methyl-2-pyrrolidone, hexamethylphosphoric triamide, 2-pyrrolidone, alkyl-2-pyrrolidone, ε-caprolactam, acetamido, polyvinylpyrrolidone, polyacrylamide, and N-vinyl -2-pyrrolidone and the like. Of these, polyvinylpyrrolidone is preferred. These compounds may be used alone or in combination of two or more.

  Mass ratio (D / P1) between the polymer dispersant (D) and the metal fine particles (P1) to be coated with the polymer dispersant (D) (that is, the polymer dispersant on the surface of the metal fine particles (P1)) The coating amount of (D) is not particularly limited, but, for example, the lower limit is preferably 0.1, and particularly preferably 0.5, from the viewpoint of reliably improving the dispersibility of the metal fine particles (P1). On the other hand, the upper limit of the mass ratio (D / P1) is set when the first adherend 11 and / or the second adherend 22 is subjected to the heat treatment by applying the paste of the metal fine particles (P) to the first adherend 11 and / or the second adherend 22. Is preferably 5.0 from the viewpoint of preventing the polymer dispersant (D) from remaining in the metal porous body 13, thereby reliably preventing a decrease in mechanical strength of the metal porous body 13, and particularly preferably 3.0. preferable.

  The method of measuring the coating amount of the polymer dispersant (D) on the surface of the metal fine particles (P1) is not particularly limited, and can be measured using, for example, a carbon / sulfur concentration analyzer. Specifically, first, metal fine particles (P1) coated with a polymer dispersant (D) to be measured are introduced into a reaction vessel of a carbon / sulfur concentration analyzer, and oxygen gas is introduced into the reaction vessel. To bring the carbon component contained in the metal fine particles (P1) coated with the polymer dispersant (D) to be measured in a state of carbon dioxide. The concentration of the extracted carbon dioxide is measured by a carbon dioxide detector. The amount of carbon is calculated from the measured concentration of carbon dioxide, and the mass of the coated polymer dispersant (D) is calculated from the calculated amount of carbon. In this way, the coating amount of the polymer dispersant (D) actually coated on the surface of the metal fine particles (P1) can be determined.

Organic solvent (S)
When the atmosphere during sintering does not contain a reducing gas such as hydrogen gas, the organic solvent (S) needs to have a reducing ability to the organic solvent. Therefore, when the reducing gas is not contained in the atmosphere during sintering, the term solvent (S) contains dihydric or higher alcohols (S1 to S3). The dihydric or higher alcohol (S1 to S3) has a reducing property for the metal fine particles (P). Therefore, when the organic solvent (S) contains a dihydric or higher alcohol (S1 to S3), the paste of the metal fine particles (P) can be applied to either the first adherend or the second adherend, or In the step of applying to each and sintering by heat treatment (sintering step), sintering of the metal fine particles (P) is promoted to form the metal porous body 13 having excellent mechanical properties and conductivity. Can be. “Dihydric or higher alcohol” means an alcohol having two or more hydroxyl groups (—OH) in one molecule.

  As the dihydric or higher alcohol (S1 to S3), dihydric alcohol (S1) having two hydroxyl groups (-OH) in one molecule, and three hydroxyl groups (-OH) in one molecule Trihydric alcohols (S2) include polyhydric alcohols (S3) having four or more hydroxyl groups (-OH) in one molecule.

Dihydric alcohol (S1)
Although it does not specifically limit as a dihydric alcohol (S1), For example, ethylene glycol, diethylene glycol, 1,2-propanediol, 1,3-propanediol, 1,2-butanediol, 1,3-butanediol, , 4-butanediol, 2,3-butanediol, 2-butene-1,4-diol, pentanediol, hexanediol, 3-methyl-1,5-pentanediol, octanediol, glyceraldehyde, dihydroxyacetone, etc. No. Of these, ethylene glycol and diethylene glycol are preferred. These may be used alone or in combination of two or more.

Trihydric alcohol (S2)
Although it does not specifically limit as a trihydric alcohol (S2), For example, glycerin (glycerol), 1,1,1- tris (hydroxymethyl) ethane, 1,2,4-butanetriol, 1,2,3-hexane Triol, 1,2,6-hexanetriol, 2-ethyl-2-hydroxymethyl-1,3-propanediol, threose, erythrulose, erythrose and the like. Of these, glycerin (glycerol) is preferable because it promotes sintering of the metal fine particles (P1) and is excellent in decomposition in the sintering step. These may be used alone or in combination of two or more.

Polyhydric alcohol (S3)
Examples of the polyhydric alcohol include, but are not particularly limited to, for example, polyglycerin, threitol, erythritol, pentaerythritol, hexaerythritol, xylitol, ribitol, arabitol, hexitol, mannitol, sorbitol, dulcitol, arabinose, Ribose, ribulose, xylose, xylulose, lyxose, glucose, fructose, mannose, idose, sorbose, growth, talose, tagatose, galactose, allose, altrose, lactose, isomaltose, glucoheptose, heptose, maltotriose, lactulose, trehalose, etc. Is mentioned. Of these, polyglycerin is preferred. These may be used alone or in combination of two or more.

  Further, any one of the dihydric alcohol (S1), the trihydric alcohol (S2), and the polyhydric alcohol (S3) may be used alone, or two or more kinds may be used in combination. On the other hand, when the atmosphere at the time of sintering contains a reducing gas, it is not necessary to give the organic solvent a reducing ability, so that the organic solvent used is not particularly limited.

  As the first adherend 11, for example, a substrate made of a metal such as copper or a copper alloy, a coated substrate obtained by coating a substrate made of a material such as alumina, silicon nitride or ceramic with a copper foil, or a metal made of a metal such as copper or a copper alloy is used. Electronic members such as a lead frame, a semiconductor package, and a semiconductor chip can be used. In addition, as the electronic member, a silicon chip or the like can be given.

  Examples of the second adherend 12 include an electronic member such as a semiconductor package, a semiconductor chip such as a silicon chip, a SiC chip, a GaN chip, and a diamond chip.

  The stress (F) applied from the metal porous body 13 at the joint portion 22 of the second adherend 12 that is in contact with the metal porous body 13 that is the joint with the first adherend 11 is: It is different in the surface of the joint portion 22. In the joint structure 1, the coefficient of variation of the stress (F) from the metal porous body 13 on the surface in contact with the metal porous body 13 and facing the joint portion 22 of the second adherend 12 is within 10%. It is. That is, in the second adherend 12, the stress (F) from the porous metal body 13 in the surface of the joint portion 22 is uniform. Therefore, in the joint structure 1, variation and destruction of the characteristic change of the second adherend 12 are prevented, and excellent operational reliability is imparted to the second adherend 12.

  The coefficient of variation of the stress (F) is not particularly limited as long as it is within 10%, but is preferably as low as possible, that is, as close to 0%, for example, more preferably within 9%, particularly preferably within 8%.

In this specification, the “coefficient of variation of stress (F)” means a value calculated by the following equation (1).
Coefficient of variation (%) of stress (F) = standard deviation of stress (F) at 10 or more measurement points ÷ average value of stress (F) at 10 or more measurement points × 100 (1)
The stress (F) in the above equation (1) is a value measured on the surface of the second adherend 12 facing the joint portion 22. The stress (F) applied from the porous metal body 13 to the surface of the second adherend 12 facing the bonding portion 22 is predetermined at room temperature (20 ± 5 ° C., the same applies hereinafter). Is measured by Raman spectroscopy using the measurement wavelength of

  Next, a method for manufacturing the joint structure 1 according to the embodiment of the present invention will be described in detail. The bonding structure 1 has a paste of metal fine particles (P) as a bonding material disposed between the first adherend 11 and the second adherend 12, and a paste of the metal fine particles (P). Can be produced by heat treatment. More specifically, for example, (1) a step of applying a paste of metal fine particles (P), if necessary, (2) a drying step of drying the applied paste of metal fine particles (P), and (3) a metal step Through the step of sintering the paste of the fine particles (P), the joint structure 1 can be manufactured.

(1) Coating Step In the coating step, a paste of metal fine particles (P), which is a bonding material, is applied to a first member to be the first adherend 11 and a second member to be the second adherend 12. Or between the first member to be the first adherend 11 and the second member to be the second adherend 12 by applying a predetermined method to each of them. This is a step of applying a bonding material to form a composite of the first member and the second member.

  The paste of the metal fine particles (P), which is a bonding material, includes metal fine particles (P) including metal fine particles (P1) having a specific average primary particle diameter (for example, an average primary particle diameter of 2 nm or more and 500 nm or less), It can be prepared by dispersing in an organic solvent (S) containing a dihydric or higher alcohol. The method for dispersing the metal fine particles (P) in the organic solvent (S) is not particularly limited. For example, mixing, kneading, or kneading is performed at room temperature using a known mixer, kneader, kneader, or the like. Method.

  The mass ratio (P / S) between the metal fine particles (P) and the organic solvent (S) contained in the paste of the metal fine particles (P) is not particularly limited. By preventing the viscosity from decreasing, shape stability when applying the paste of the metal fine particles (P) is maintained (for example, generation of sagging is prevented), and application of a complicated shape such as pattern formation is performed. 10/90 is preferable, and 20/80 is particularly preferable because it can cope with the above. On the other hand, the upper limit of the mass ratio (P / S) is preferably 85/25, particularly preferably 75/35 from the viewpoint of imparting excellent coatability to the paste of the metal fine particles (P). Therefore, when the mass ratio (P / S) is in the range of 10/90 to 85/25, the first member to be the first adherend 11 and the second member to be the second adherend 12 The joining material can be suitably arranged between the members. The range of the mass ratio (P / S) can be appropriately set according to a method of applying a paste of metal fine particles (P) described later.

  The method of applying the paste of the metal fine particles (P) is not particularly limited, and any known application method can be used. Examples of the application method include various methods such as a squeegee method, screen printing, mask printing, inkjet printing, dispenser printing, spray coating, bar coating, knife coating, spin coating, applicator, blade coating, and roll coating.

(2) Drying Step If necessary, a drying step may be performed on the composite of the first member and the second member obtained in the coating step. In the drying step, the fine metal particles (P) applied between the first member to be the first adherend 11 and the second member to be the second adherend 12 in the above (1) coating step This is a step of pre-drying the paste under a predetermined heating condition such as a predetermined temperature and a predetermined heating time. That is, before the sintering step, the content of the organic solvent (S) contained in the paste of the metal fine particles (P) is reduced in advance in a drying step.

  Drying conditions such as preliminary drying performed in the drying step include, for example, a method of depositing the metal fine particles (P) in the paste of the metal fine particles (P) under an atmosphere of an inert gas (for example, nitrogen gas, argon gas, or the like) or an atmospheric gas. Drying is performed so that the content is in the range of 87 to 93% by mass in the total amount of 100% by mass of the joining material. Specifically, for example, holding at a drying temperature of about 100 ° C. to 120 ° C. for about 30 to 60 minutes may be mentioned.

(3) Sintering Step In the sintering step, the metal fine particles (P) applied between the first member to be the first adherend 11 and the second member to be the second adherend 12 This is a step of sintering the paste of (1) by heat treatment. By sintering the paste of the metal fine particles (P), the metal porous body 13 is formed from the paste of the metal fine particles (P). The formed metal porous body 13 serves as a joining portion for joining the first adherend 11 and the second adherend 12.

  The temperature rising rate in the sintering step is 100 ° C./min or more and less than 5000 ° C./min. When the temperature rise rate is 100 ° C./min or more, the second adherend 12 is joined to the first adherend 11 in a state where unevenness of stress on the second adherend 12 is prevented. it can. Specifically, the coefficient of variation of the stress (F) from the porous metal body 13 is within 10% in the surface of the second adherend 12 that is in contact with the porous metal body 13 and faces the joint portion 22. In this state, the second adherend 12 can be joined to the first adherend 11. Therefore, when joining the second adherend 12 to the first adherend 11, it is possible to prevent variation and destruction of the characteristic change of the second adherend 12, and the second adherend 12 is excellent. Operational reliability can be provided.

  The lower limit of the heating rate in the sintering step is not particularly limited as long as it is 100 ° C./min. From the viewpoint of making the stress (F) from the metal porous body 13 more uniform, the second adherend 12 It is preferable that the temperature increase rate is higher as long as the heat damage can be prevented. Specifically, for example, the lower limit of the heating rate is preferably 500 ° C / min, more preferably 2000 ° C / min, and particularly preferably 3000 ° C / min. On the other hand, the heating rate is less than 5000 ° C./min from the viewpoint of preventing thermal damage to the second adherend 12. For example, the upper limit of the heating rate is the heat of the second adherend 12. 4000 ° C./min is preferable, and 3800 ° C./min is particularly preferable from the viewpoint of reliably preventing damage.

  The sintering by the heat treatment performed in the sintering step includes a method of sintering by heating under no pressure or under a predetermined pressure. As a sintering condition, for example, when heating under pressure, a first member to be a first adherend and a second member to be a second adherend are used by using a heat press machine having a built-in heater. After holding the entire member with a uniform pressure balance, degassing (reducing the pressure) and holding the member at a heating temperature (sintering temperature) of 190 to 400 ° C. for about 1 to 120 minutes. The firing atmosphere in the sintering step is not particularly limited, and may be, for example, an air atmosphere, a reducing gas atmosphere, an inert gas atmosphere, or a mixed atmosphere thereof. The reducing gas includes, for example, hydrogen gas, and the inert gas includes, for example, nitrogen gas. Each atmosphere gas may be used alone or in combination of two or more.

  Next, examples of the present invention will be described, but the present invention is not limited to these examples unless it exceeds the gist.

(1) Average primary particle size of metal fine particles (P1) Observed using a scanning electron microscope (SEM) (manufactured by Hitachi High-Technologies Corporation, product name: SU8020) under conditions of an acceleration voltage of 3 kV and a magnification of 200,000. Then, an SEM image of the metal fine particles (P1) to be measured was obtained. From the acquired SEM images, arbitrarily select 20 metal fine particles (P1), measure the longest diameter of the primary particles of the selected metal fine particles (P1), and calculate the average of the measured values. To determine the average primary particle size.

(2) Evaluation of stress on second adherend A silicon chip and a SiC chip were used as the second adherend. The stress on each chip was determined by measuring the inside of each chip surface (the surface facing the bonding site) by Raman spectroscopy. The measurement wavelength of the Raman spectroscopy was 633 nm for the silicon chip and 532 nm for the SiC chip. The measurement temperature was room temperature (23 ° C.). The stress measurement sites were 11 or more points on the surface of each chip except for the 1 mm end of each chip, and the standard deviation and average value were calculated. Those with a coefficient of variation of stress (standard deviation ÷ average value × 100) of 10% or less were evaluated as acceptable because the uniformity of stress was high in the chip surface and did not affect the device characteristics.

(3) Evaluation of Transistor Characteristics The characteristics of a transistor as a chip as a second adherend were evaluated. For evaluation of transistor characteristics, a lateral MOS transistor was manufactured. The transistor was formed on a (001) plane p-type silicon substrate of 1 to 10 Ω / cm. The gate film thickness was 50 nm, the gate length was 4 μm, and the gate width was 500 μm. In the source and drain regions, P was ion-implanted to form n + regions. Further, in order to secure conduction with the source and the drain, the ion implantation region, the gate oxide film and the gate electrode were overlapped. The gate electrode was made of n + polysilicon. For the SiC transistor, a 4H type (0001) surface semi-insulating type substrate was prepared. A p-type of 1 × 10 ¹⁷ cm ⁻³ was epitaxially grown on the Si surface on the substrate at a thickness of 10 μm. Al was used as a dopant. Thereafter, the source and drain regions were ion-implanted with P at a dose of 5 × 10 ¹⁹ cm ⁻³ and heat-treated at 1700 ° C. At the time of ion implantation, SiO ₂ was appropriately formed on the surface so that the concentration of P became maximum on the surface. The gate thickness of the transistor was 20 nm, the gate length was 1 μm, and the gate width was 1000 μm. The gate oxide film was formed by wet thermal oxidation at 1300 ° C. On both the Si and SiC substrates, the substrate thickness was reduced to 230 μm on the back surface (the surface on which no transistor was formed), and after hydrofluoric acid treatment, Ti / Ni / Au = 100/450/200 nm was vacuum deposited. The measurement conditions were a source-drain voltage of 0.1 V and a gate voltage of 10 V. As an evaluation of transistor characteristics, a source-drain current was measured. The measurement site of the source-drain current was 11 points or more in the plane except for the 1 mm end portion of the transistor, and its standard deviation and average value were calculated. If the variation rate of the source-drain current (standard deviation ÷ average value × 100) is within 15%, the uniformity of the source-drain current in the transistor surface is high and the transistor characteristics are not affected, and the result is evaluated as pass. did.

(Separation / recovery process of metal fine particles (P))
Metal fine particles (P1) were added to an aqueous solution in which 10 g of polyvinylpyrrolidone (PVP, number average molecular weight: about 3,500) as a polymer dispersant (D) was added to 1979.3 ml of distilled water and dissolved. 29.268 g of cupric hydroxide (Cu (OH) ₂ , particle size range: 0.1 to 100 μm, manufactured by Kanto Chemical Co., Ltd.) was added and stirred to prepare an aqueous cupric hydroxide solution. The liquid temperature of the cupric hydroxide aqueous solution obtained as described above was adjusted to 20 ° C., and 150 mmol of sodium borohydride (NaBH ₄ manufactured by Kanto Chemical Co., Ltd.) and water were stirred while being stirred in a nitrogen gas atmosphere. 20.73 ml of an aqueous solution containing 480 mmol of sodium oxide (manufactured by Kanto Kagaku Co., Ltd., NaOH) was added dropwise, and a reduction reaction was carried out with stirring for 45 minutes to obtain a metal coated with the polymer dispersant (D) in the mixed solution. Fine particles (P1) were generated. Note that the reduction reaction was terminated when the generation of hydrogen gas from the reaction system was completed.

To the mixed solution containing the metal fine particles (P1) coated with the polymer dispersant (D) formed as described above, 66 ml of chloroform (CHCl ₃ manufactured by Kanto Chemical Co., Ltd.) is added as an aggregation promoter. The mixture was stirred for several minutes and allowed to stand for several minutes to precipitate the metal fine particles (P1) coated with the polymer dispersant (D). The mixed solution containing the solid content of the metal fine particles (P1) coated with the polymer dispersant (D) precipitated as described above is supplied to a centrifugal separator, and solid-liquid separation is performed. The solid content containing the metal fine particles (P1) coated with (D) was recovered.

  Next, methanol is added to the collected solid content including the fine metal particles (P1) coated with the polymer dispersant (D), and the mixture is stirred and supplied to a centrifuge to perform solid-liquid separation. The metal fine particles (P1) coated with the molecular dispersant (D) were washed with methanol. The methanol washing was repeated several times to separate and collect the fine metal particles (P1) coated with the polymer dispersant (D). Next, vacuum drying was performed for 7 hours to obtain dried metal fine particles (P1). The degree of vacuum during vacuum drying was -80 kPa or less, with the atmospheric pressure being 0 kPa.

  A part of the dried metal fine particles (P1) coated with the polymer dispersant (D) obtained through the above separation / recovery step is collected, and the metal fine particles coated with the polymer dispersant (D) are collected. The primary particle diameter of (P1) was measured by the method of (1) above, and the average primary particle diameter was calculated to be 50 nm.

(Step of obtaining paste of metal fine particles (P))
Paste 1: for sintering under pressure in a nitrogen gas atmosphere 60 g of metal fine particles (P1) coated with the polymer dispersant (D) obtained as described above, and an organic solvent (S ) As a trihydric alcohol (manufactured by Kanto Chemical Co., Ltd.) (glycerol) (C ₃ H ₅ (OH) ₃ , boiling point: 290 ° C.) (40 g) is placed in a container, and a centrifugal kneader (manufactured by Shinky Corporation) ARE-250), and kneading was repeated four times under the condition of 2000 rpm × 3 minutes to obtain a paste-like dispersion solution of metal fine particles (P) (paste 1).
Paste 2: for pressurized or non-pressurized sintering in hydrogen gas atmosphere 60 g of metal fine particles (P1) coated with polymer dispersant (D) obtained as described above and α-terpineol (Kanto Chemical Co., Ltd.) 40 g), put in a container, put into a centrifugal kneader (product name: ARE-250, manufactured by Sinky Co., Ltd.), and kneaded four times under the condition of 2000 rpm × 3 minutes. A dispersion solution (paste 2) of the paste-like metal fine particles (P) was obtained.

(Preparation of chip as second adherend)
As a silicon chip, a (001) oriented silicon substrate having a thickness of 230 μm and a diameter of 15 cm was prepared. One surface of the silicon substrate was polished and the other surface was etched. As a SiC chip, a 4H—SiC (0001) oriented Si surface substrate (230 μm thick) was prepared. One surface of the SiC substrate was mirror-finished, and the other surface was etched. The mirror-finished surface of the SiC chip was defined as the Si surface.

  Immediately after removing the oxide film on the substrate surface by dilute hydrofluoric acid treatment, the substrate was housed in a vacuum evaporation apparatus. With this apparatus, Ti / Ni / Au was deposited on the polished surface of the silicon substrate and the mirror-finished surface of the SiC substrate, respectively, to form films. Thereafter, a dicing tape was attached to the deposition surface, and diced into 10 mm squares to form chips. Next, the dicing tape was irradiated with UV light under a nitrogen atmosphere. By the irradiation of the UV light, the adhesive force of the dicing tape was lost, and the chip was picked up to obtain the chip.

(Step of applying paste of metal fine particles (P))
A copper substrate (13 mm square, 1 mm thick Cl020 material, tempered 1 / 2H material) was prepared as a first adherend. A stainless steel metal mask (200 mm square metal mask having a square opening of 8 mm square, 100 μm thickness) is placed on one surface of the copper substrate, and the paste obtained in the step of obtaining a paste of the metal fine particles (P) is obtained. The dispersion solution of the metal fine particles (P) was applied by a squeegee method.

(Drying process)
-In the case of pressure sintering, place on a hot plate with the coated surface of the copper substrate coated with the metal fine particle (P) paste facing upward, and hold at a drying temperature of 110 ° C for 30 to 60 minutes in an air atmosphere. The organic solvent (S) until the content of the fine metal particles (P) in the paste of the fine metal particles (P) becomes about 90% by mass based on 100% by mass of the total amount of the paste of the fine metal particles (P). Was preliminarily dried to evaporate.
-In the case of pressureless sintering No drying process was performed.

(Sintering process)
Pressure sintering in nitrogen gas atmosphere (Examples 1-3, 10-12, Comparative Examples 1, 2, 7, 8)
The chip (10 mm square, 230 μm thick, structure of the back electrode of Ti / Ni / Au = 100 nm / 450 nm / 200 nm). Next, the copper substrate on which the chips were arranged was set in a press device, and a pressure of 10 MPa was applied to the entire copper substrate on which the chips were arranged with a uniform pressure balance. Thereafter, nitrogen gas was flowed into the press device to reduce the oxygen concentration to 500 ppm or less. Thereafter, heating was performed at a heating rate shown in Tables 1 and 2 below to sinter the paste of the metal fine particles (P) and bond the chip to the copper substrate via the metal porous body. I got The temperature was raised to 300 ° C., and the temperature was kept at that temperature (300 ° C.) for 5 minutes. Then, the heater of the press device was turned off, and it cooled by natural cooling.

Pressure sintering in a hydrogen gas atmosphere (Examples 4 to 6, 13 to 15, Comparative Examples 3, 4, 9, and 10)
The chip (10 mm square, 230 μm thick, the structure of the back electrode is Ti / Ni / Au = 100 nm / 450 nm / 200 nm). Next, the copper substrate on which the chips were arranged was set in a press device, and a pressure of 10 MPa was applied to the entire copper substrate on which the chips were arranged with a uniform pressure balance. Then, the inside of the press was evacuated and replaced with hydrogen gas. Thereafter, heating was performed at a heating rate shown in Tables 1 and 2 below to sinter the paste of the metal fine particles (P) and bond the chip to the copper substrate via the metal porous body. I got The temperature was raised to 300 ° C., and the temperature was kept at that temperature (300 ° C.) for 5 minutes. Then, the heater of the press device was turned off, and it cooled by natural cooling.

When pressureless sintering is performed in a hydrogen gas atmosphere (Examples 7 to 9, 16 to 18, Comparative Examples 5, 6, 11, and 12)
A chip (10 mm square, 230 μm thick, back electrode structure: Ti / Ni / Au = 100 nm / 450 nm / 200 nm). Next, the copper substrate on which the chips were arranged was set in a press device. No pressure was applied because no pressure was applied. Then, the inside of the press was evacuated and replaced with hydrogen gas. Thereafter, heating was performed at a heating rate shown in Tables 1 and 2 below to sinter the paste of the metal fine particles (P) and bond the chip to the copper substrate via the metal porous body. I got The temperature was raised to 300 ° C., and the temperature was maintained at that temperature (300 ° C.) for 120 minutes. Then, the heater of the press device was turned off, and it cooled by natural cooling.

  Tables 1 and 2 show the results of the stress evaluation on the chip as the second adherend and the results of the transistor characteristic evaluation.

  As shown in Tables 1 and 2 above, in both the silicon chip and the SiC chip, at a heating rate of 100 ° C./min to 3600 ° C./min, the fluctuation rate of the stress is suppressed within 10% and the source of the transistor is reduced. The fluctuation rate of the drain current was reduced to 14% or less. Therefore, by setting the temperature rising rate in the sintering step to 100 ° C./min or more, the bonding can be performed in a state where the stress applied to the chip is not made uniform. It has been found that can be prevented. Further, it has been found that by setting the temperature rising rate in the sintering step to 100 ° C./min or more, excellent operational reliability can be imparted to the chip.

  On the other hand, at a rate of temperature rise of 10 ° C./min, the variation rate of the stress increases to 15 to 20% for the silicon chip and 17 to 20% for the SiC chip, and the source-drain current of the transistor increases. The fluctuation rates have increased to 20 to 26% and 35 to 42%, respectively. Therefore, it has been found that when the heating rate in the sintering step is 10 ° C./min, uneven stress on the chip cannot be prevented, and a change in the characteristics of the chip cannot be prevented. At a temperature rise rate of 5000 ° C./minute, cracks occurred in the silicon chip due to a rapid temperature rise in pressure sintering, and cracks occurred in the SiC chip, and bonding was not possible. Further, in pressureless sintering at a temperature rising rate of 5000 ° C./min, chips could not be joined due to evaporation of the organic solvent.

DESCRIPTION OF SYMBOLS 1 Joining structure 11 1st adherend 12 2nd adherend

Claims (12)

A joint structure comprising a first adherend, a porous metal body formed on the first adherend, and a second adherend disposed on the porous metal body. hand,
A joint structure in which a coefficient of variation of stress in a portion of the second adherend facing the first adherend is within 10%.   The joined structure according to claim 1, wherein the second adherend is an electronic member.   The joint structure according to claim 2, wherein the electronic member is a silicon chip.   4. The joint structure according to claim 1, wherein the first adherend is a metal substrate or another electronic member. 5.   The joint structure according to any one of claims 1 to 4, wherein the metal porous body is a sintered body of metal fine particles, and the metal particles have an average primary particle diameter of 2 nm or more and 500 nm or less.   The joint structure according to claim 5, wherein the metal fine particles include copper. Manufacturing of a bonded structure including a first adherend, a porous metal body formed on the first adherend, and a second adherend disposed on the porous metal body The method
Applying a paste containing metal fine particles between a first member to be the first adherend and a second member to be the second adherend, to form a composite;
Heating the composite at a heating rate of 100 ° C./min or more and less than 5000 ° C./min to sinter a paste containing metal fine particles to form a metal porous body;
A method for manufacturing a joined structure including:   The method for manufacturing a bonded structure according to claim 7, wherein the heating rate is 4000 ° C / minute or less.   9. The method for manufacturing a joined structure according to claim 7, wherein the paste containing the metal fine particles includes metal fine particles having an average primary particle diameter of 2 nm or more and 500 nm or less, and an organic solvent containing a divalent or higher alcohol. 10. .   The method for manufacturing a bonded structure according to claim 7, wherein the metal fine particles include copper.   The dihydric or higher alcohol is glycerin, 1,1,1-tris (hydroxymethyl) ethane, 1,2,4-butanetriol, 1,2,3-hexanetriol, 1,2,6-hexanetriol and The method for producing a joined structure according to claim 9 or 10, wherein the method comprises at least one trihydric alcohol selected from the group consisting of 2-ethyl-2-hydroxymethyl-1,3-propanediol.   The method for producing a bonded structure according to claim 9 or 10, wherein the dihydric or higher alcohol includes glycerin.

Images