Prediction device

Abstract

A low viscosity is achieved in a low thermal expansion resin composition containing silica particles. [Solution] A prediction device (300) predicts the respective blending ratios of first to fourth silica particles, each with a different median diameter, in an inorganic filler that satisfy the required characteristics of a resin composition comprising a resin and an inorganic filler, comprising: a data acquisition unit (33) that acquires resin composition required characteristics data showing the required characteristics; and a recommendation data derivation unit (34) that inputs the resin composition required characteristics data into a prediction model to derive recommended inorganic filler blending data showing the blending ratios of each silica particle that satisfy the resin composition required characteristics data.

Description

[Technical Field]

[0001] One aspect of the present invention relates to a resin composition, a semiconductor encapsulant, and a prediction device. [Background technology]

[0002] With the rapid progress in miniaturization, thinning, and high-density mounting of electronic devices, the gap between the element and the substrate is becoming narrower, and semiconductor encapsulants are required to have even lower thermal expansion and higher moldability.

[0003] Silica particles are incorporated as a filler in resin compositions used as semiconductor encapsulants. Incorporating silica particles into a resin composition can reduce its thermal expansion coefficient. However, incorporating a large amount of silica particles increases the viscosity of the resin composition, reducing its moldability.

[0004] As an example of a resin composition containing silica particles, Patent Document 1 describes an epoxy resin composition containing 260 parts by mass or more of filler per 100 parts by mass of solids. Patent Document 1 also describes that by using first silica particles, which are spherical silica particles with an average particle diameter of 0.1 μm or more and less than 1.0 μm, second silica particles, which are spherical silica particles with an average particle diameter of 1.0 μm or more and 5.0 μm or less, and third silica particles, which are spherical silica nanoparticles with an average particle diameter of 20 nm or more and 200 nm or less, as fillers, an epoxy resin composition that is easy to mold can be obtained even if a large amount of filler is included. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Japanese Patent Publication No. 2014-88509 [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] However, the epoxy resin composition described in Patent Document 1 is formulated by combining multiple silica particles of a specific particle size to obtain a composition with desired properties, and it is difficult to obtain a composition with similar properties when using silica particles of other particle sizes. Since there are a vast number of options for the particle size of silica particles, their combinations, and their blending ratios, it is not easy to select and blend silica particles to obtain a resin composition with desired properties based on the description in Patent Document 1. In particular, there is a trade-off between achieving low thermal expansion by incorporating a high amount of silica particles and achieving a low viscosity resin composition, and it is not easy to achieve both.

[0007] One aspect of the present invention aims to achieve low viscosity in a low thermal expansion resin composition containing silica particles. [Means for solving the problem]

[0008] To solve the above problems, a resin composition according to one aspect of the present invention is a resin composition comprising a resin and an inorganic filler, wherein the inorganic filler comprises: First silica particles having a median diameter of 2 μm or more and 5 μm or less, Second silica particles, which are wet silica with a median diameter of 0.2 μm or more and less than 0.5 μm, Third silica particles, which are wet silica with a median diameter of 0.1 μm or more and less than 0.2 μm, It includes.

[0009] A semiconductor encapsulant according to one aspect of the present invention includes a resin composition according to one aspect of the present invention.

[0010] A prediction device according to one aspect of the present invention satisfies the required characteristics of a resin composition comprising a resin and an inorganic filler, by determining the following proportions in the inorganic filler: (I) the proportion of first silica particles having a median diameter of 2 μm or more and 5 μm or less, (II) the proportion of second silica particles which are wet silica having a median diameter of 0.2 μm or more and less than 0.5 μm, (III) the proportion of third silica particles which are wet silica having a median diameter of 0.1 μm or more and less than 0.2 μm, and (IV) the proportion of fourth silica particles which are wet silica having a median diameter of 0.5 μm or more and less than 1.5 μm. A prediction device for predicting the above, comprising: a data acquisition unit that acquires resin composition required characteristic data showing the required characteristics; and a recommendation data derivation unit that inputs the resin composition required characteristic data into a prediction model to derive recommended inorganic filler formulation data showing (I) the blending ratio of the first silica particles, (II) the blending ratio of the second silica particles, (III) the blending ratio of the third silica particles, and (IV) the blending ratio of the fourth silica particles in the inorganic filler that satisfy the resin composition required characteristic data. [Effects of the Invention]

[0011] According to one aspect of the present invention, a low viscosity can be achieved in a low thermal expansion resin composition containing silica particles. [Brief explanation of the drawing]

[0012] [Figure 1] This is a block diagram showing the configuration of the main parts of a resin composition manufacturing system equipped with a prediction device according to one aspect of the present invention. [Figure 2] This is a diagram illustrating the overview of the first prediction model. [Figure 3] This diagram illustrates the overview of the second prediction model. [Figure 4] This diagram illustrates the overview of the recommended data derivation section. [Figure 5] This figure shows an example of inorganic filler formulation input data and second input data. [Figure 6] This figure shows an example of the derivation of explanatory variables in the first algorithm execution unit. [Figure 7]It is a diagram showing an example of generating a first prediction model in the second algorithm execution unit. [Figure 8] It is a diagram showing an example of input and output in the first prediction model. [Figure 9] It is a diagram showing an example of an operation in the second prediction model generated by the third algorithm execution unit.

Mode for Carrying Out the Invention

[0013] As a result of intensive studies to further achieve a low viscosity in a low thermal expansion resin composition containing silica particles, the present inventor has found that by inputting characteristic data required for the resin composition (resin composition required characteristic data), a resin composition satisfying both low thermal expansion and low viscosity can be realized by using a prediction model that derives recommended inorganic filler blending data representing the blending ratio of each silica particle in the inorganic filler that satisfies the resin composition required characteristic data.

[0014] Hereinafter, first, the configuration of a [resin composition] that satisfies both low thermal expansion and low viscosity will be described. Then, a [resin composition manufacturing system] that generates a prediction model used to realize the resin composition and predicts the blending ratio of the inorganic filler using the generated prediction model will be described.

[0015] 〔Resin Composition〕 The resin composition according to one aspect of the present invention includes a resin and an inorganic filler. The inorganic filler includes first silica particles having a median diameter of 2 μm or more and 5 μm or less, second silica particles which are wet silica having a median diameter of 0.2 μm or more and less than 0.5 μm, and third silica particles which are wet silica having a median diameter of 0.1 μm or more and less than 0.2 μm. Further, the resin composition may further include fourth silica particles which are wet silica having a median diameter of 0.5 μm or more and less than 1.5 μm as the inorganic filler.

[0016] A resin composition according to one aspect of the present invention contains silica particles as an inorganic filler and therefore exhibits low thermal expansion. Resin compositions used as semiconductor encapsulants are preferably low in viscosity in addition to having low thermal expansion in order to improve moldability. A resin composition according to one aspect of the present invention has low viscosity, and its viscosity is preferably 250 Pa·s or less, and more preferably 200 Pa·s or less, at room temperature (20-30°C). Semiconductor encapsulants containing a resin composition according to one aspect of the present invention are also included in the scope of the present invention.

[0017] The resin composition may further contain a curing agent, elastomer, ion trapping agent, pigment, dye, defoaming agent, stress relaxant, pH adjuster, accelerator, surfactant, coupling agent, etc.

[0018] The resin composition is obtained by mixing a resin with an inorganic filler. The inorganic filler mixed with the resin may be a dispersion in a solvent.

[0019] (resin) Examples of resins included in the resin composition include epoxy resins, acrylic resins, polyurethane resins, and, as one example, epoxy resin. Examples of epoxy resins that can be used include bisphenol-type epoxy resins such as bisphenol A-type epoxy resin, bisphenol F-type epoxy resin, and bisphenol S-type epoxy resin; novolac-type epoxy resins such as phenol novolac-type epoxy resin and cresol novolac-type epoxy resin; biphenyl-type epoxy resin, xylylene-type epoxy resin, phenol aralkyl-type epoxy resin, biphenyl aralkyl-type epoxy resin, biphenyl dimethylene-type epoxy resin, trisphenolmethane novolac-type epoxy resin, glycidyl ethers, trifunctional or tetrafunctional glycidylamines, tetramethylbiphenyl-type epoxy resin and other arylalkylene-type epoxy resins; naphthalene-skeleton-modified cresol novolac-type epoxy resin, methoxynaphthalene-modified cresol novolac-type epoxy resin, methoxynaphthalene-dimethylene-type epoxy resin and other naphthalene-skeleton-modified epoxy resins; anthracene-type epoxy resin, dicyclopentadiene-type epoxy resin, norbornene-type epoxy resin, fluorene-type epoxy resin, and flame-retardant epoxy resins obtained by halogenating the above epoxy resins. These epoxy resins can be used individually or in combination of two or more types with different weight-average molecular weights.

[0020] The resin may be contained in the entire resin composition in an amount of 10% by mass or more and 40% by mass or less, more preferably 10% by mass or more and 35% by mass or less, and even more preferably 15% by mass or more and 30% by mass or less.

[0021] (Inorganic filler) The inorganic filler includes first silica particles, second silica particles, and third silica particles. The inorganic filler may also further contain fourth silica particles. The median diameter of the first to fourth silica particles is the particle size D at which the cumulative frequency in the particle size distribution determined by laser diffraction and scattering is 50%. 50This is intended. The first to fourth silica particles may be spherical. Spherical silica particles can be manufactured by the manufacturing methods of each silica particle described later.

[0022] The proportion of inorganic filler in the resin composition is 60% by mass or more and 95% by mass or less of the total resin composition, preferably 65% ​​by mass or more and 90% by mass or less, and more preferably 70% by mass or more and 85% by mass or less.

[0023] <First Silica Particle> The first silica particles are silica particles with a median diameter of 2 μm or more and 5 μm or less. The method for producing the first silica particles is not particularly limited, but they can be produced by classifying crude silica powder obtained by a specific method using a specific classification apparatus.

[0024] Methods for producing the crude silica powder used to produce the first silica particles include, for example, a melting method, a wet method, a dry method, etc., but one example is the melting method. As the melting method, a known melting method can be used, and for example, the method described in Japanese Patent Application Publication No. 2015-86120 can be used.

[0025] In the melting method, fumed silica, the raw material, is supplied to a flame produced by burning a gaseous or liquid fuel mixed with a combustion-supporting gas such as oxygen or air in a burner, using an ejector, screw feeder, and fluidized bed. The supplied fumed silica melts to produce molten spherical silica, which is then guided to a cyclone for solid-gas separation to obtain crude silica powder. The silica particles in the crude silica powder obtained in this way may be spherical.

[0026] As the raw material, fumed silica can be a silicon compound, particularly a silicon halide, generally a silicon chloride, usually obtained by a known method of burning purified silicon tetrachloride in an oxyhydrogen flame.

[0027] It is preferable that the fumed silica is hydrophobized. By using hydrophobized fumed silica, dispersibility in the flame is improved, resulting in a spherical silica crude powder with higher sphericity and uniform particle size. In the spherical silica crude powder thus produced, the amount of fine silica particles with extremely large specific surface area is reduced. The hydrophobization of fumed silica is not particularly limited and can be carried out by conventionally known hydrophobization treatments, and any known treatment agent, such as a silylation agent, can be used without any limitations.

[0028] As an example of a method for classifying the crude silica powder obtained as described above, one method is to use a mechanism that classifies the powder by balancing the centrifugal force acting on the powder due to the rotation of the classifying rotor and the drag force acting on the powder due to the airflow passing in the axial direction of the classifying rotor. In such a method, a dispersed airflow of crude silica powder is supplied to a powder classifier in which the upper surface of the classifying rotor is coated with resin, and coarse particles larger than 45 μm contained in the crude silica powder are scattered and separated from the classifying rotor. Alternatively, the classifying rotor may be disc-shaped with an internal cavity that communicates from the peripheral edge to below the axial center, and separation may be performed by selectively drawing in fine particles smaller than 1 μm into the internal cavity by suction from the internal cavity side.

[0029] <Second Silica Particles> The second silica particles are wet-processed silica with a median diameter of 0.2 μm or more and less than 0.5 μm. The second silica particles can be produced by known wet processes. An example of a wet process is the sol-gel method described in the republished patent WO2018 / 096876. The sol-gel method involves hydrolyzing and polycondensing a silicon alkoxide in a reaction medium consisting of water containing a catalyst and an organic solvent to produce a silica sol, then gelling it, removing the resulting solid, and drying it to produce silica powder. The silica powder produced by this method may be further calcined as needed.

[0030] In the sol-gel method described above, water, a polar solvent other than water (organic solvent), and a basic catalyst are charged into a reaction vessel, and silicon alkoxide (or an organic solvent solution of silicon alkoxide) and an aqueous solution of the basic catalyst are simultaneously added and reacted. By appropriately setting the addition time of the silicon alkoxide and the basic catalyst, and the reaction temperature, silica particles with a desired particle size distribution can be produced. Alternatively, the dispersion of silica particles produced by the sol-gel method may be filtered by wet filtration to remove coarse particles.

[0031] Examples of silicon alkoxides include methyltrimethoxysilane, methyltriethoxysilane, tetramethoxysilane, tetraethoxysilane, tetraisopropoxysilane, and tetrabutoxysilane. Examples of basic catalysts include amine compounds and alkali metal hydroxides. Examples of organic solvents that are polar solvents other than water include alcohols such as methanol, ethanol, isopropyl alcohol, and butanol; ethers such as tetrahydrofuran and dioxane; and amide compounds such as dimethylformamide, dimethylacetamide, and N-methylpyrrolidone.

[0032] <Third Silica Particle> The third silica particle is wet-processed silica with a median diameter of 0.1 μm or more and less than 0.2 μm. The third silica particle can be manufactured by a known wet-process method, similar to the second silica particle.

[0033] <Fourth Silica Particle> The fourth silica particle is wet-processed silica with a median diameter of 0.5 μm or more and less than 1.5 μm. The fourth silica particle can be manufactured by a known wet-process method, similar to the second silica particle.

[0034] <Other silica particles> The inorganic filler may also contain other silica particles besides the first to fourth silica particles. For example, the inorganic filler may contain silica particles with a median diameter of 1.5 μm or more and less than 2 μm. Alternatively, the inorganic filler may contain silica particles with a median diameter exceeding 5 μm.

[0035] (Surface treatment) The second silica particles, third silica particles, and fourth silica particles may be treated with a surface treatment agent. Known silane coupling agents, siloxanes, silazanes, etc., can be used without particular limitation as surface treatment agents. For example, silane coupling agents include methyltrimethoxysilane, dimethyldimethoxysilane, phenyltrimethoxysilane, hexyltrimethoxysilane, octyltrimethoxysilane, decyltrimethoxysilane, vinyltrimethoxysilane, 3-glycidoxypropyltrimethoxysilane, 3-mercaptopropyltrimethoxysilane, 3-methacryloxypropyltrimethoxysilane, 3-acryloxypropyltrimethoxysilane, 3-aminopropyltrimethoxysilane, etc. For siloxanes, disiloxane, hexamethyldisiloxane, hexamethylcyclotrisiloxane, octamethylcyclotetrasiloxane, decamethylcyclopentasiloxane, and polysiloxanes such as polydimethylsiloxane can be used. For silazanes, hexamethyldisilazane can be used. The surface treatment agent is preferably a silane coupling agent.

[0036] The method for treating the second, third, and fourth silica particles with a surface treatment agent is not particularly limited, but one method is to stir-mix the silica particles and the surface treatment agent in a mixing device and then heat-treat them. An example of a method for treating silica particles with a surface treatment agent is the method described in the republished patent WO2019 / 044929 and the Japanese Patent Application Publication No. 2014-201454.

[0037] Surface treatment with a surface treatment agent is preferably performed after obtaining a dispersion of second, third, and fourth silica particles by the sol-gel method, and before wet filtration. This removes any aggregates or residual surface treatment agent that may occur during the surface treatment. The method of adding the surface treatment agent to the mixing device is not particularly limited, but if the surface treatment agent is a low-viscosity liquid at room temperature and atmospheric pressure, it can be added directly. If the surface treatment agent is a high-viscosity liquid or solid at room temperature and atmospheric pressure, it should be added to a suitable organic solvent to form a solution or dispersion before being added to the mixing device. The amount of surface treatment agent used to treat the silica particles should be set appropriately according to the application and other factors.

[0038] (Composition ratio) The proportion a of the first silica particles in the inorganic filler is 50% by mass or more and 80% by mass or less, preferably 55% by mass or more and 80% by mass or less, and more preferably 60% by mass or more and 75% by mass or less, based on the total amount of inorganic filler.

[0039] The proportion c of the second silica particles in the inorganic filler is 5% by mass or more and 20% by mass or less, preferably 5% by mass or more and 15% by mass or less, relative to the total amount of inorganic filler.

[0040] The proportion d of the third silica particles in the inorganic filler is 5% by mass or more and 30% by mass or less, preferably 10% by mass or more and 25% by mass or less, relative to the total amount of inorganic filler.

[0041] The proportion b of the fourth silica particles in the inorganic filler is 0% by mass or more and 15% by mass or less, and preferably 0% by mass or more and 10% by mass or less, relative to the total amount of inorganic filler.

[0042] The proportion of each silica particle in the inorganic filler should be selected to satisfy the required properties of the resin composition. For example, the proportion of each silica particle in the inorganic filler should be selected to satisfy the required viscosity of the resin composition. Preferably, the proportion of each silica particle in the inorganic filler should be selected so that the viscosity of the resin composition is 250 Ps·s or less.

[0043] The proportion of each silica particle in an inorganic filler such that the viscosity of the resin composition is 250 Ps·s or less is given by the following formula (1), where (i) the proportion of the second silica particle is c, (ii) the proportion of the third silica particle is d, and (iii) the proportion of the fourth silica particle is b. -0.078×b -0.105×c -0.011×d -0.0159×b 2 -0.0217 × c 2 -0.0130 × d 2 -0.0393 × b × c - 0.0257 × c × d + 0.0218 × d × b + 25.498 ≥ 0 ····(1) It may satisfy the following. Examples 1 to 5 described later are resin compositions in which each silica particle is blended in a blending ratio that satisfies formula (1).

[0044] Furthermore, if (i) the proportion of second silica particles in the inorganic filler is c, (ii) the proportion of third silica particles is d, and (iii) the proportion of fourth silica particles is b, then the following equation (2) -0.205×b + 0.125×c + 0.244×d - 0.0151×b 2 -0.0102 × c 2 -0.0079×d 2 -0.0154 × b × c - 0.0104 × c × d + 0.0252 × d × b + 0.964 ≥ 0 ····(2) By using a blending ratio that satisfies the above, a resin composition with a viscosity of 250 Ps·s or less can be obtained with a higher probability than with a blending ratio that satisfies formula (1). Examples 3 to 5 described later are resin compositions in which each silica particle is blended in a blending ratio that satisfies formula (2).

[0045] The blending ratio of each silica particle in the inorganic filler may be calculated based on formula (1) or formula (2) above, and then the blending ratio may be adjusted according to the type and content of resin contained in the resin composition.

[0046] [Predictive device for predicting the mixing ratio of each silica particle in an inorganic filler] A prediction device according to one aspect of the present invention satisfies the required characteristics of a resin composition comprising a resin and an inorganic filler by determining the following proportions in the inorganic filler: (I) the proportion of first silica particles having a median diameter of 2 μm or more and 5 μm or less, (II) the proportion of second silica particles, which are wet silica, having a median diameter of 0.2 μm or more and less than 0.5 μm, (III) the proportion of third silica particles, which are wet silica, having a median diameter of 0.1 μm or more and less than 0.2 μm, and (IV) the proportion of fourth silica particles, which are wet silica, having a median diameter of 0.5 μm or more and less than 1.5 μm. A prediction device for predicting each of the blending ratios, comprising: a data acquisition unit that acquires resin composition required characteristic data showing the required characteristics; and a recommendation data derivation unit that inputs the resin composition required characteristic data into a prediction model to derive recommended inorganic filler blending data showing the blending ratio of (I) first silica particles, (II) second silica particles, (III) third silica particles, and (IV) fourth silica particles in the inorganic filler that satisfy the resin composition required characteristic data.

[0047] (Resin composition manufacturing system 1) A resin composition manufacturing system 1 equipped with a prediction device according to one aspect of the present invention is described below. For the sake of simplicity, explanations of matters similar to those in the prior art will be omitted as appropriate. Please note that each configuration and numerical value described herein is merely an example unless otherwise specified.

[0048] Figure 1 is a block diagram showing the configuration of the main parts of the resin composition manufacturing system 1. The resin composition manufacturing system 1 comprises a model generation device 100 and a prediction device 300. As described later, the model generation device 100 generates a prediction model for predicting the conditions for satisfying the required characteristics of a resin composition comprising one or more inorganic fillers and one or more resins. Specifically, the model generation device 100 generates a second prediction model (MODEL2), which will be described later, as the prediction model. The prediction device 300 then uses MODEL2 to predict the conditions for satisfying the required characteristics.

[0049] The prediction device 300 comprises a third input data acquisition unit (data acquisition unit) 33, a recommended data derivation unit 34, and an output unit 35. The prediction device 300 derives recommended inorganic filler blending data, which shows the blending ratios of first to fourth silica particles in the inorganic filler that satisfy the resin composition requirement characteristics, which represent the required characteristics of the resin composition. Here, the resin composition requirement characteristics are, for example, the viscosity of the resin composition. That is, the prediction device 300 may derive the blending ratios of first to fourth silica particles in the inorganic filler to obtain a resin composition with a desired viscosity. In the prediction device 300, if a desired low viscosity is set as the resin composition requirement characteristic data, the prediction model derives the blending ratios that achieve this. Then, by blending the first to fourth silica particles with the derived blending ratios to produce the resin composition, an even lower viscosity can be achieved in a low thermal expansion resin composition containing silica particles. Details of the prediction device 300 will be described later.

[0050] In the following description, one or more silica particles will be collectively referred to as inorganic filler A, and one or more resins will be collectively referred to as resin B. A resin composition containing inorganic filler A and resin B will be referred to as resin composition C.

[0051] It should be noted that the "conditions for satisfying required characteristics" in this specification are not limited to "conditions that completely satisfy the required characteristics of resin composition C." The "conditions for satisfying required characteristics" in this specification also include, for example, "required characteristics that are closest to the complete required characteristics of resin composition C" and "required characteristics that are somewhat close to the complete required characteristics of resin composition C." Therefore, the resin composition manufacturing system 1 only needs to be able to predict conditions that generally satisfy the complete required characteristics of resin C as the conditions for satisfying required characteristics. For this reason, for example, as described later, the resin composition manufacturing system 1 may predict the conditions for satisfying required characteristics based on a probabilistic mathematical model.

[0052] The processing of the resin composition manufacturing system 1 is broadly divided into a learning phase (processing in the model generation device 100) and a prediction phase (processing in the prediction device 300). The prediction phase is also called the inference phase. First, the learning phase will be explained.

[0053] (Model generation device 100) The model generation device 100 generates a prediction model used in the prediction device 300. The model generation device 100 comprises a first input data acquisition unit 11, a second input data acquisition unit 12, a first machine learning unit 21, and a second machine learning unit 22.

[0054] The first input data acquisition unit 11 acquires the first input data 110. The first input data 110 is input data that includes inorganic filler formulation input data 111. The inorganic filler formulation input data 111 is data relating to the blending ratio of inorganic filler A in a resin composition C in which multiple types of inorganic filler A are blended into resin B. As an example, the inorganic filler formulation input data 111 is data showing the blending ratio of inorganic filler A.

[0055] The first input data acquisition unit 11 may be any data acquisition interface. For example, the first input data acquisition unit 11 may be an input unit that accepts user input operations. In this case, the first input data acquisition unit 11 acquires the first input data 110 entered by the user. As another example, the first input data acquisition unit 11 may acquire the first input data 110 that has been pre-stored in a storage unit (not shown) within the model generation device 100. As yet another example, the first input data acquisition unit 11 may communicate with an external device (not shown) of the model generation device 100 and acquire the first input data 110 from the external device. Examples of external devices include a storage server and a measuring device. These descriptions of the first input data acquisition unit 11 also apply to the second input data acquisition unit 12 and the third input data acquisition unit 33, which will be described later.

[0056] The second input data acquisition unit 12 acquires the second input data 120. Specifically, the second input data acquisition unit 12 acquires resin composition characteristic input data as the second input data 120. The resin composition characteristic input data represents the characteristics of resin composition C. From this, it can be said that the second input data 120 is data that corresponds to the first input data 110.

[0057] The first machine learning unit 21 acquires the first input data 110 from the first input data acquisition unit 11 and the second input data 120 from the second input data acquisition unit 12. Based on the first input data 110 and the second input data 120, the first machine learning unit 21 generates a first prediction model (hereinafter referred to as MODEL1).

[0058] Figure 2 is a diagram illustrating the overview of MODEL 1. As shown in Figure 2, MODEL 1 is a model (mathematical model) for predicting unknown resin composition property data 1200 from arbitrary inorganic filler formulation data 1110. In this specification, a dataset containing arbitrary inorganic filler formulation data 1110 is referred to as an arbitrary input dataset 1100.

[0059] In the example in Figure 2, the arbitrary input dataset 1100 is an example of an explanatory variable (X). The unknown resin composition property data 1200 is an example of an objective variable (y). Explanatory variables are also called independent variables. In contrast, the objective variable is also called the dependent variable or the variable being explained. In the example in Figure 2, MODEL1 can be represented as a function f that shows the relationship y=f(X). Thus, MODEL1 is a model for solving forward problems (a model for deriving y from X).

[0060] In this specification, the types of data included in any input dataset 1100 are assumed to be the same as the types of training data used to generate MODEL1. That is, the data structure of any input dataset 1100 is assumed to be the same as the data structure of the training data used to generate MODEL1. Therefore, for example, the data structure of any input dataset 1100 is the same as the data structure of the first input data 110.

[0061] The second machine learning unit 22 obtains MODEL1 from the first machine learning unit 21. Based on MODEL1, the second machine learning unit 22 generates a second prediction model (hereinafter referred to as MODEL2).

[0062] Figure 3 is a diagram illustrating the overview of MODEL2. As shown in Figure 3, MODEL2 is a model for predicting predicted inorganic filler formulation data 2310 that satisfies arbitrary resin composition characteristic data 2200. In this specification, "satisfies arbitrary resin composition characteristic data 2200" means "satisfies the required characteristics of resin composition C indicated by the arbitrary resin composition characteristic data 2200." In this specification, the dataset including the predicted inorganic filler formulation data 10 is referred to as the predicted dataset 2300.

[0063] In the example in Figure 3, the arbitrary resin composition property data 2200 is an example of the dependent variable (y). The predicted dataset 2300 is an example of the independent variable (X). In the example in Figure 3, MODEL2 can be represented as a function g showing the relationship X = g(y). Note that g ≈ f -1 Therefore, function g is the approximate inverse function of function f. Thus, MODEL2 is a model for solving inverse problems (a model for deriving X from y). As described above, MODEL2 is a model that is the counterpart to MODEL1.

[0064] In this specification, the type of arbitrary resin composition property data 2200 is assumed to be the same as the type of training data used to generate MODEL2. That is, the data structure of arbitrary resin composition property data 2200 is assumed to be the same as the data structure of the training data used to generate MODEL2. Therefore, for example, the data structure of arbitrary resin composition property data 2200 is the same as the data structure of the second input data 120.

[0065] In the example shown in Figure 1, the first machine learning unit 21 includes a first algorithm execution unit 211 and a second algorithm execution unit 212. The first algorithm execution unit 211 acquires first input data 110 from the first input data acquisition unit 11. The first algorithm execution unit 211 executes a first algorithm that derives explanatory variables (X) corresponding to data indicating the properties of resin composition C obtained from the second input data 120. An example of the first algorithm will be described later.

[0066] The second algorithm execution unit 212 obtains explanatory variables (X) from the first algorithm execution unit 211 and also obtains second input data 120 from the second input data acquisition unit 12. The second algorithm execution unit 212 executes a second algorithm that generates MODEL1 based on X and the second input data 120.

[0067] The second algorithm may be any algorithm capable of generating MODEL1 based on X and the second input data 120. In other words, the second algorithm may be any algorithm capable of generating a model for solving a forward problem. As an example, the second algorithm may be: • Gaussian process regression, Support Vector Machines Linear regression, Decision tree, • Random Forest, • Neural networks, and · Gradient boosting wood, It is at least one of the following.

[0068] In the example shown in Figure 1, the second machine learning unit 22 includes a third algorithm execution unit 223. The third algorithm execution unit 223 executes a third algorithm that generates MODEL2 based on MODEL1.

[0069] The third algorithm may be any algorithm capable of generating MODEL2 based on MODEL1. In other words, the third algorithm may be any algorithm capable of generating a model for solving the inverse problem. As an example, the third algorithm may be: • Genetic algorithms, • Steepest descent method, Grid search, and • Bayesian optimization, It is at least one of the following.

[0070] In one embodiment of the present invention, the resin composition characteristic input data (second input data 120) may be data representing any characteristic of the resin composition C. For example, the resin composition characteristic input data may be data representing the resin composition C, Viscosity, fluidity, moldability, adhesion, transparency, color, strength, water absorption, coefficient of linear expansion, elastic modulus, yield stress, tensile strength, fracture toughness, electrical conductivity, dielectric constant, dielectric loss tangent, thermal conductivity, and stability. At least one of these is shown as a property of resin composition C. The resin composition property input data may also be the viscosity of resin composition C. The same explanation regarding the resin composition property input data applies to each data corresponding to said resin composition property input data.

[0071] (Prediction device 300) Next, the prediction phase will be described. The prediction device 300 includes a third input data acquisition unit (data acquisition unit) 33, a recommended data derivation unit 34, and an output unit 35.

[0072] The third input data acquisition unit 33 acquires the third input data 330. Specifically, the third input data acquisition unit 33 acquires the resin composition required characteristics data as the third input data 330. The resin composition required characteristics data represents the required characteristics of resin composition C.

[0073] The recommended data derivation unit 34 acquires resin composition requirement characteristic data (third input data 330) from the third input data acquisition unit 33, and also acquires MODEL2 from the model generation device 100 (more specifically, the second machine learning unit 22).

[0074] Figure 4 is a diagram illustrating the overview of the recommended data derivation unit 34. As shown in Figure 4, the recommended data derivation unit 34 derives recommended data 340 by inputting the resin composition required characteristic data into MODEL2. The recommended data 340 includes recommended inorganic filler formulation data 341.

[0075] The recommended inorganic filler formulation data 341 shows the formulation of inorganic filler A that satisfies the resin composition required characteristics data. In this specification, "satisfies the resin composition required characteristics data" means "satisfies the required characteristics of resin composition C as indicated by the resin composition required characteristics data." A specific example of the process for deriving the recommended data 340 by the recommended data derivation unit 34 will be described later.

[0076] As stated above, the "conditions for satisfying required characteristics" in this specification are not limited to "conditions that completely satisfy the required characteristics of resin composition C." Therefore, naturally, the recommended data 340 in this specification are not limited to "data that completely satisfies the resin composition required characteristics data." The recommended data 340 in this specification also includes "data that generally satisfies the resin composition required characteristics data."

[0077] Therefore, the recommended inorganic filler formulation data 341 in this specification only needs to show a formulation of inorganic filler A that generally satisfies the resin composition required characteristics data. These explanations regarding the recommended inorganic filler formulation data 341 also apply to the predicted inorganic filler formulation data 2310 described above.

[0078] The output unit 35 obtains the recommended data 340 from the recommended data derivation unit 34. Then, the output unit 35 outputs the recommended data 340. The output unit 35 may be any output interface. For example, the output unit 35 may be a display (display device). In this case, the output unit 35 can visually present the recommended data 340 to the user by displaying the recommended data 340. Thus, the output unit 35 may output the recommended data 340 in a visual manner. As another example, the output unit 35 may transfer the recommended data 340 to a storage unit within the model generation device 100. As yet another example, the output unit 35 may transfer the recommended data 340 to an external device of the model generation device 100.

[0079] In one embodiment of the present invention, the resin composition required characteristic data (third input data 330) may be data indicating any required characteristics of resin composition C. As is clear from the above description of the resin composition characteristic input data, as an example, the resin composition required characteristic data is of resin composition C, Viscosity, fluidity, moldability, adhesion, transparency, color, strength, water absorption, coefficient of linear expansion, elastic modulus, yield stress, tensile strength, fracture toughness, electrical conductivity, dielectric constant, dielectric loss tangent, thermal conductivity, and stability. At least one of these is indicated as a required characteristic of resin composition C. As an example, the required characteristic data for resin composition C is the viscosity of resin composition C.

[0080] (An example of processing in resin composition manufacturing system 1) The following describes an example of processing in resin composition manufacturing system 1. In the following example, we will describe the case where there are five types of inorganic filler A and one type of resin B. In the following example, both the properties of resin B and the mixing ratio of resin B are set to constant fixed values ​​(fixed conditions).

[0081] Therefore, in the following example, the properties of resin B and the blending ratio of resin B will not be considered as explanatory variables. Consequently, the resin property input data 112 and resin blending input data 114 will not be mentioned in the following explanation. For this reason, the blending ratio of inorganic filler A will be abbreviated as simply "blending ratio" below. In the following example, the viscosity (unit: Pa·s) of resin composition C will be used as an example of the properties of resin composition C. The viscosity of resin composition C will also be abbreviated as simply "viscosity".

[0082] (Examples of first input data 110 and second input data 120) Figure 5 shows an example of inorganic filler formulation input data 111 and second input data 120. In the example in Figure 5, the inorganic filler formulation input data 111 is data that shows the mixing ratio of five types of inorganic fillers in weight % (wt%). In the following explanation, the five types of inorganic fillers will be referred to as inorganic filler 0 to inorganic filler 4, respectively. The mixing ratio of inorganic filler i will be referred to as x0i, where i is an integer satisfying 0 ≤ i ≤ 4. For example, x01 represents the mixing ratio of inorganic filler 1.

[0083] As will be obvious to those skilled in the art, x00+x01+x02+x03+x04=100 The following relationship holds. Therefore, x00 is x00 = 100 - x01 - x02 - x03 - x04 As shown, it is uniquely determined according to the pre-set x01 to x04. Therefore, in the inorganic filler formulation input data 111 in the example of Figure 5, only x01 to x04 are set in order to reduce the number of dimensions of the explanatory variables.

[0084] In the inorganic filler formulation input data 111, multiple formulation ratio patterns (patterns of combinations of x01 to x04) are set. The id in Figure 5 is the identification number of the formulation ratio pattern. For example, id=1 indicates the first formulation ratio pattern (hereinafter also referred to as the first formulation ratio pattern). In the example in Figure 5, the first formulation ratio pattern is "x00=60, x01=10, x02=30, x03=0, x04=0". In the following, for example, id=1 will be abbreviated as id1 as appropriate.

[0085] In the example in Figure 5, the second input data 120 shows the viscosity (y01) of resin composition C corresponding to each blending ratio pattern. Specifically, the measured viscosity value for each blending ratio pattern is recorded in y01 in the second input data 120. Prior to the computer simulation in this example, the inventors of the present application (hereinafter simply referred to as "the inventors") manufactured resin composition C according to the first blending ratio pattern described above. The inventors then measured the viscosity of resin composition C and obtained a measured value of 455. Therefore, in the second input data 120 in the example in Figure 5, the value y01 = 455 is set for id1.

[0086] In the example in Figure 5, a dataset showing the correspondence between each blending ratio pattern and y01 is created for each ID. In the example in Figure 5, the dataset corresponding to the j-th ID (idj) is called DATASET_idj. For example, DATASET_id1 is a dataset showing the correspondence between the first blending ratio pattern and y01. Hereafter, the j-th blending ratio pattern will also be referred to as the j-th blending ratio pattern.

[0087] (Example of derivation of explanatory variables in the first algorithm execution unit 211) Figure 6 shows an example of the derivation of explanatory variables in the first algorithm execution unit 211. The first algorithm execution unit 211 derives explanatory variables (X) corresponding to data indicating the properties of resin composition C obtained from the first input data 110. In this example, the first algorithm execution unit 211 derives explanatory variables corresponding to data indicating the viscosity of resin composition C obtained from inorganic filler formulation input data 111. Thus, in this example, the explanatory variable X is derived to explain the target variable, viscosity.

[0088] Specifically, the first algorithm execution unit 211 derives X based on the inorganic filler formulation input data 111 by executing the first algorithm. In the example in Figure 6, the first machine learning unit 21 performs weighted average calculation and principal component analysis as the first algorithm.

[0089] In the example shown in Figure 6, the first machine learning unit 21 derives a particle size distribution causal vector (a vector with components xx01 to xx05 in Figure 6) as an explanatory variable by performing principal component analysis on the inorganic filler formulation input data 111. Alternatively, the first machine learning unit 21 may derive the explanatory variables by performing a weighted average on the inorganic filler formulation input data 111, for example, based on the inorganic filler's characteristic data, and then performing principal component analysis on the weighted averaged inorganic filler formulation input data. The inorganic filler's characteristic data could, for example, be the particle size distribution of the inorganic filler. The weight values ​​in the weighted average calculation may be set by a known method, for example, based on the inorganic filler's characteristic data. The dimensions reduced by principal component analysis may be set arbitrarily. The first machine learning unit 21 calculates a particle size distribution causal vector for each ID. Therefore, for example, as shown in Figure 7, the explanatory variables derived by the first machine learning unit 21 include the first particle size distribution causal vector (the particle size distribution causal vector corresponding to ID 1) described below.

[0090] In this specification, the particle size distribution origin vector corresponding to the j-th mixing ratio pattern (in other words, the particle size distribution origin vector corresponding to idj) is referred to as the j-th particle size distribution origin vector. The j-th particle size distribution origin vector is calculated from the particle size distribution of inorganic fillers 0 to 4 in resin composition C when the j-th mixing ratio pattern is applied. Therefore, for example, the first particle size distribution origin vector is calculated from the particle size distribution of inorganic fillers 0 to 4 in resin composition C when the first mixing ratio pattern described above is applied.

[0091] (Example of MODEL1 generation in the second algorithm execution unit 212) Figure 7 shows an example of MODEL1 generation in the second algorithm execution unit 212. The second algorithm execution unit 212 generates MODEL1 based on (i) the explanatory variable (X) derived by the first algorithm execution unit 211 and (ii) the second input data 120. Specifically, the second algorithm execution unit 212 generates MODEL1 based on X and the second input data 120 by executing the second algorithm.

[0092] In the example shown in Figure 7, the second algorithm execution unit 212 executes a neural network as the second algorithm. Specifically, for each ID, the second algorithm execution unit 212 obtains the target variable corresponding to X from the second input data 120. For example, the second algorithm execution unit 212 obtains the viscosity (y01) shown in DATASET_id1 as the ground truth data for the target variable (y) corresponding to X in ID1. Then, the second algorithm execution unit 212 derives a function f that satisfies the relationship between X and y in each ID by executing a neural network using this ground truth data. In this way, the second algorithm execution unit 212 generates MODEL1 as a function f that shows the relationship y=f(X).

[0093] The following example illustrates the case where MODEL1 is generated by the second algorithm execution unit 212 performing ensemble learning using the bagging method. Therefore, MODEL1 in this example includes multiple neural networks (weak learners) with different hyperparameters. Thus, MODEL1 in this example is generated as a strong learner by integrating multiple weak learners.

[0094] (Examples of input / output in MODEL1) Figure 8 shows an example of input and output in MODEL1. In Figure 8, the first granularity distribution causal vector (the granularity distribution causal vector corresponding to id1) is exemplified as X. As shown in Figure 8, by inputting X into MODEL1, a histogram (y_Hist) showing the distribution of y can be obtained. Specifically, by inputting X into each of the multiple weak learners in MODEL1, y_Hist is obtained by integrating the multiple y outputs from these multiple weak learners.

[0095] In this example, prior to the generation of MODEL1 using the second algorithm described above, preprocessing of the ground truth data is performed. Specifically, in this example, a logarithmic transformation of the ground truth data is performed prior to the generation of MODEL1. Therefore, MODEL1 is generated as a model that outputs log(y). For this reason, the horizontal axis of y_Hist in the example in Figure 8 is log(y). However, for simplicity, in this specification, we will explain MODEL1 as a model that outputs y. Furthermore, we will explain y_Hist as a histogram showing the distribution of y.

[0096] In this example, MODEL1 determines predetermined data based on y_Hist as the final predicted value (predicted value as a strong learner) and outputs this final predicted value. Specifically, in this example, MODEL1 outputs the mean value (μ) of y_Hist as the final predicted value (y). μ is also called the expected value.

[0097] According to MODEL1 generated as a reinforcement learner, as shown in FIG. 2 above, by obtaining any input data set 1100 as an explanatory variable (X), unknown resin composition characteristic data 1200 can be output as an objective variable (y). For example, by inputting the above-described first blending ratio pattern into MODEL1 as X, a predicted value of viscosity corresponding to the first blending ratio pattern can be output as y.

[0098] Note that MODEL1 may output the variance (σ 2 ) of y_Hist together with μ. This variance is an example of an index of the uncertainty of the predicted value of MODEL1. Alternatively, MODEL1 may output the standard deviation (σ) of y_Hist together with μ. This standard deviation is another example of an index of the uncertainty of the predicted value of MODEL1.

[0099] (Calculation example in MODEL2 generated by the third algorithm execution unit 223) FIG. 9 is a diagram showing a calculation example in MODEL2 generated by the third algorithm execution unit 223. The third algorithm execution unit 223 generates MODEL2 based on MODEL1 by executing the third algorithm. In other words, the third algorithm execution unit 223 determines the above-described function g (approximate inverse function of the function f) based on the function f predetermined by the second algorithm execution unit 212 (see also FIG. 3 above).

[0100] In the example of FIG. 9, the third algorithm execution unit 223 generates MODEL2 by using a grid search as the third algorithm. Inside the generated MODEL2, the following two-stage (first stage and second stage) calculations are executed.

[0101] First, in the first stage, MODEL2 uses MODEL1 to calculate a predicted viscosity (μ) for each of the multiple possible blending ratio patterns (combinations of x01 to x04). Specifically, MODEL2 inputs the explanatory variables (X) corresponding to each blending ratio pattern into MODEL1, causing MODEL1 to output μ as the target variable. In this example, MODEL1 further outputs σ in addition to μ for each blending ratio pattern.

[0102] Next, in the second stage, MODEL2 outputs a blending ratio pattern (X) that maximizes the probability of obtaining the arbitrary viscosity data input to MODEL2, based on the μ and σ calculated in the first stage (see also Figure 3 above).

[0103] In the example in Figure 9, MODEL2 predicts the prediction dataset 2300 that has the highest probability of obtaining arbitrary resin composition property data 2200. In the example in Figure 9, the case where arbitrary resin composition property data 2200 is data indicating a viscosity of "μ = 290 to 310" is illustrated.

[0104] The above numerical range of μ = 290 to 310 is an example of a numerical range set assuming that resin composition C is used as an adhesive. For example, if the viscosity of the adhesive is too high, it is difficult to mold the adhesive into the desired shape for use. On the other hand, if the viscosity of the adhesive is too low, it is prone to dripping. Therefore, it is considered that there is a suitable numerical range for the viscosity of the adhesive. The above numerical range of μ = 290 to 310 is an example of such a suitable numerical range.

[0105] In the example in Figure 9, MODEL2 selects the blending ratio pattern with the highest probability of "μ falling within the target range of 290 to 310" (hereinafter referred to as "probability relative to the target range") as the optimal blending ratio pattern. Then, MODEL2 outputs this optimal blending ratio pattern as the prediction result (i.e., prediction dataset 2300). In the example in Figure 9, MODEL2 calculates the probability relative to the target range for all possible combinations of multiple blending ratio patterns. Specifically, in the example in Figure 9, MODEL2 calculates the probability relative to the target range under the assumption that μ follows a normal distribution.

[0106] In the example in Figure 9, a dataset showing the correspondence between "each blending ratio pattern," "μ and σ," and "probability relative to the target range" is created for each ID. In the example in Figure 9, the dataset corresponding to the j-th ID (idj) is called DATASET2_idj.

[0107] In the example in Figure 9, MODEL2 searches for the dataset with the highest probability for all j relative to the target range (hereinafter referred to as the maximum probability dataset). In the example in Figure 9, the search by MODEL2 found that DATASET2_id1 is the maximum probability dataset. The first blending pattern in the example in Figure 9 is the blending pattern "x00=65, x01=5, x02=20, x03=0, x04=10".

[0108] MODEL2 selects the dataset with the highest probability as the optimal dataset (DATASET2_OPT). Then, MODEL2 determines the blending ratio pattern corresponding to the optimal dataset as the optimal blending ratio pattern. In the example in Figure 9, MODEL2 selects DATASET2_id1 as the optimal dataset. Then, the third algorithm execution unit 223 determines the above-mentioned first blending ratio pattern corresponding to DATASET2_id1 as the optimal blending ratio pattern. Thus, in the example in Figure 9, the first blending ratio pattern is output as the prediction result.

[0109] As described above, the third algorithm execution unit 223 generates MODEL2 (a model that predicts a predictive dataset 2300 that satisfies arbitrary resin composition characteristic data 2200) using the third algorithm (e.g., grid search). However, as will be apparent to those skilled in the art, the method for determining the optimal blending ratio pattern is not limited to the above example.

[0110] For example, MODEL2 may search for a blending ratio pattern that corresponds to the dataset (hereinafter referred to as the "most recent predicted value dataset") that yields the closest predicted value (μ) to 300 (the median of the viscosity range mentioned above) among all the blending ratio patterns in the example in Figure 9. Then, MODEL2 may determine the blending ratio pattern corresponding to the most recent predicted value dataset as the optimal blending ratio pattern.

[0111] (Example of recommended data derivation in the recommended data derivation unit 34) As described above, by using the generated MODEL2, the recommended data derivation unit 34 can derive recommended data 340 based on the third input data 330 (resin composition required characteristic data) (see also Figure 4 above).

[0112] For example, the third input data 330 may be data indicating viscosity, such as "μ = 350 to 370". In this case, the recommended data derivation unit 34 can derive recommended data 340 corresponding to the third input data 330 by inputting the third input data 330 into MODEL2. In this example, the recommended data 340 is, for example, an optimal blending ratio pattern corresponding to a viscosity of μ = 350 to 370.

[0113] (Specific examples of recommended data derived in the recommended data derivation unit 34) As described above, by using the generated MODEL2, the recommended data derivation unit 34 can derive the following as recommended inorganic filler formulation data that satisfies the required characteristics for resin composition C: (I) the blending ratio of first silica particles, (II) the blending ratio of second silica particles, (III) the blending ratio of third silica particles, and (IV) the blending ratio of fourth silica particles in inorganic filler A.

[0114] As an example, the recommended data derivation unit 34 derives inorganic filler formulation data that has a probability of satisfying the required characteristics of resin composition C of 25% or more as a necessary condition for satisfying the required characteristics. Also, as an example, the recommended data derivation unit 34 derives inorganic filler formulation data that has a probability of satisfying the required characteristics of resin composition C of 90% or more as a sufficient condition for satisfying the required characteristics.

[0115] For example, the third input data 330 is data indicating a viscosity of "μ = 250 Pa·s or less". In this case, the recommended data derivation unit 34 can derive recommended data 340 corresponding to the third input data 330 by inputting the third input data 330 into MODEL2. In this example, the recommended data 340 is, for example, data indicating the blending ratio of an inorganic filler that can achieve a viscosity of μ = 250 Pa·s or less with a predetermined probability.

[0116] In this example, the prediction device 300 first acquires data indicating a viscosity of "μ = 250 Pa·s or less" as the third input data 330 using the data acquisition unit 33. Then, the recommended data derivation unit 34 inputs the data indicating a viscosity of "μ = 250 Pa·s or less" into MODEL2 to derive recommended inorganic filler formulation data for resin composition C to satisfy a viscosity of "μ = 250 Pa·s or less".

[0117] The numerical range μ = 250 Pa·s or less is an example of a numerical range set assuming that resin composition C is used as a semiconductor encapsulant. For example, if the viscosity of the semiconductor encapsulant is too high, it is difficult to shape the semiconductor encapsulant into the desired form for use. Therefore, it is considered that there is a suitable numerical range for the viscosity of the semiconductor encapsulant. The above numerical range μ = 250 Pa·s or less is an example of such a suitable numerical range.

[0118] When data indicating a viscosity range of μ = 250 Pa·s or less is used as the third input data 330, the necessary conditions for satisfying the required characteristics such that the viscosity of resin composition C becomes lower than μ, as derived in the recommended data derivation unit 34, are given by the following equation (1), where (II) the blending ratio of second silica particles in inorganic filler A is c, (III) the blending ratio of third silica particles is d, and (IV) the blending ratio of fourth silica particles is b. -0.078×b -0.105×c -0.011×d -0.0159×b 2 -0.0217 × c 2 -0.0130 × d 2 -0.0393 × b × c - 0.0257 × c × d + 0.0218 × d × b + 25.498 ≥ 0 ····(1) This is data represented by [this formula].

[0119] When data indicating a viscosity range of μ = 250 Pa·s or less is used as the third input data 330, the sufficient condition for satisfying the required characteristic, such that the viscosity of resin composition C becomes lower than μ, as derived in the recommended data derivation unit 34, is given by the following equation (2), where (II) the blending ratio of second silica particles in inorganic filler A is c, (III) the blending ratio of third silica particles is d, and (IV) the blending ratio of fourth silica particles is b. -0.205×b + 0.125×c + 0.244×d - 0.0151×b 2 -0.0102 × c 2 -0.0079×d 2 -0.0154 × b × c - 0.0104 × c × d + 0.0252 × d × b + 0.964 ≥ 0 ····(2) This is data represented by [this formula].

[0120] (effect) According to the prediction device 300, the conditions for satisfying the required characteristics of the resin composition can be predicted using MODEL2, which is pre-generated in the model generation device 100. Specifically, the prediction device 300 can derive recommended data 340 by inputting the resin composition required characteristics data into MODEL2. As is clear from the above explanations, the recommended data 340 is the prediction result for the resin composition required characteristics data.

[0121] The prediction device 300 can derive inorganic filler formulation data that has the highest probability of satisfying the required characteristics of resin composition C. Furthermore, the prediction device 300 can also derive inorganic filler formulation data that satisfies the required characteristics of resin composition C with a desired probability, such as inorganic filler formulation data with a probability of 25% or more (necessary condition for satisfying required characteristics) or inorganic filler formulation data with a probability of 90% or more (sufficient condition for satisfying required characteristics). [Examples]

[0122] A resin composition was prepared using spherical silica particles A (corresponding to the first silica particle), spherical silica particles B (corresponding to the fourth silica particle), spherical silica particles C (corresponding to the second silica particle), and spherical silica particles D (corresponding to the third silica particle) as shown in Table 1. [Table 1]

[0123] Spherical silica particles (A) were Silica Excelica UF-320 (manufactured by Tokuyama), spherical silica particles (B) were Silica Sunseal SP-10P (manufactured by Tokuyama), spherical silica particles (C) were Silica Sunseal SP-03P (manufactured by Tokuyama), and spherical silica particles (D) were Silica Sunseal SP-01P (manufactured by Tokuyama). The BET specific surface area, volume-based particle size distribution, and surface carbon content of each spherical silica particle were measured as follows.

[0124] (BET specific surface area) Using the SA-1000 specific surface area measuring device manufactured by Shibata Rikagaku Co., Ltd., the BET specific surface area S(m²) was determined by the nitrogen adsorption BET single-point method. 2 The measurement ( / g) was taken.

[0125] (Volume-based particle size distribution by laser diffraction and scattering method) Approximately 0.1 g of spherical silica particles were weighed into a 50 mL glass bottle using an electronic balance, approximately 40 mL of ethanol was added, and the particles were dispersed using an ultrasonic homogenizer (BRANSON, Sonifier 250) at 40 W for 10 minutes. The volume-based 50% diameter (D50) (μm) and coefficient of variation (CV) of the spherical silica particles were then measured using a laser diffraction scattering particle size distribution analyzer (Beckman Coulter, LS 13 320).

[0126] (Surface carbon content measurement) The carbon content (mass%) of spherical silica particles was measured using the combustion oxidation method (EMIA-511, Horiba, Ltd.). Specifically, spherical silica particle samples were heated to 1350°C in an oxygen atmosphere, and the resulting carbon content was calculated per unit mass. The spherical silica particles used for measurement were pre-treated by heating to 80°C and removing adsorbed moisture from the air by reducing the pressure in the system before being subjected to the carbon content measurement.

[0127] 24.8 g of inorganic filler containing spherical silica particles (A) to (D) in the proportions of Examples 1 to 5 and Comparative Examples 1 to 2 described in Table 2 was added to 7 g of bisphenol A+F type epoxy resin (manufactured by Nippon Steel Chemical & Material, ZX-1059) and mixed by hand. In Table 2, the amount of inorganic filler indicates the proportion (wt%) of the inorganic filler to the total resin composition, and a to d indicate the respective proportions (%) of spherical silica particles (A) to (D) to the total inorganic filler.

[0128] The hand-kneaded resin composition was pre-mixed using a rotation-and-revolution type mixer (THINKY, Awatori Rentaro AR-500) (mixing: 1000 rpm, 8 minutes; degassing: 2000 rpm, 2 minutes). After pre-mixing, the resin composition was stored in a 25°C constant temperature water bath and then mixed using a three-roll mixer (IMEX, BR-150HCV, roll diameter φ63.5). The mixing conditions were a mixing temperature of 25°C, a roll distance of 20 μm, and 5 mixing cycles. The resulting resin composition was degassed under reduced pressure for 30 minutes using a vacuum pump (Sato Vacuum, TSW-150).

[0129] The viscosity of the aforementioned compounded resin composition was measured using a rheometer (Thermo Fisher Scientific, HAAKE MARS40). The measurement temperature was 25°C, and the sensor used was a C35 / 1 (cone plate type, 35 mm diameter, 1° angle, titanium material). [Table 2] Examples 1 and 2 satisfy formula (1) described above. Examples 3 to 5 satisfy formulas (1) and (2) described above. Comparative Examples 1 and 2 do not satisfy formulas (1) and (2) described above. It has been shown that by satisfying at least one of formulas (1) and (2) described above, the following resin compositions with a viscosity of 250 Pa·s can be produced.

[0130] [Examples of implementation using software] The function of the resin composition manufacturing system 1 (hereinafter referred to as "the system") can be realized by a program that causes a computer to function as the system, and by a program that causes a computer to function as each control block of the system (in particular, each part included in the model generation device 100 and the prediction device 300).

[0131] In this case, the system includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., memory) as hardware for executing the program. By executing the program using this control device and storage device, the functions described in each of the embodiments are realized.

[0132] The above program may be recorded on one or more computer-readable recording media, not temporary ones. These recording media may or may not be provided by the above device. In the latter case, the program may be supplied to the above device via any wired or wireless transmission medium.

[0133] Furthermore, some or all of the functions of each of the above control blocks can also be realized by logic circuits. For example, an integrated circuit in which logic circuits functioning as each of the above control blocks are formed is also included in the scope of the present invention. In addition, it is also possible to realize the functions of each of the above control blocks by, for example, a quantum computer.

[0134] Furthermore, each process described in the above embodiments may be performed by AI (Artificial Intelligence). In this case, the AI ​​may operate on the control device described above, or it may operate on other devices (for example, an edge computer or a cloud server).

[0135] [Additional Notes] The present invention is not limited to the embodiments described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Industrial applicability]

[0136] One aspect of the present invention can be used, for example, in the field of electronic equipment manufacturing. [Explanation of Symbols]

[0137] 1. Resin composition manufacturing system 100 Model Generators 300 Prediction Devices 33. Third Input Data Acquisition Unit (Data Acquisition Unit) 34 Recommended Data Derivation Section

Claims

1. A predictive device for predicting the following proportions of a resin composition comprising a resin and an inorganic filler, which satisfy the required characteristics of the inorganic filler: (I) the proportion of first silica particles having a median diameter of 2 μm or more and 5 μm or less in the inorganic filler; (II) the proportion of second silica particles, which are wet silica, having a median diameter of 0.2 μm or more and less than 0.5 μm; (III) the proportion of third silica particles, which are wet silica, having a median diameter of 0.1 μm or more and less than 0.2 μm; and (IV) the proportion of fourth silica particles, which are wet silica, having a median diameter of 0.5 μm or more and less than 1.5 μm. A data acquisition unit that acquires resin composition requirement characteristic data showing the aforementioned requirement characteristics, A prediction device comprising: a recommendation data derivation unit that inputs the resin composition requirement characteristic data into a prediction model to derive recommended inorganic filler formulation data that satisfies the resin composition requirement characteristic data, showing the respective (I) the blending ratio of the first silica particles, (II) the blending ratio of the second silica particles, (III) the blending ratio of the third silica particles, and (IV) the blending ratio of the fourth silica particles in the inorganic filler.

2. The recommended data derivation unit, assuming that (II) the blending ratio of the second silica particles in the inorganic filler is c (mass%), (III) the blending ratio of the third silica particles is d (mass%), and (IV) the blending ratio of the fourth silica particles is b (mass%), then the following formula (1) -0.078×b-0.105×c-0.011×d-0.0159×b ２ -0.0217×c ２ -0.0130×d ２ -0.0393×b×c-0.0257×c×d+0.0218×d×b+25.498≧0・・・・(1) The prediction device according to claim 1, which derives the above as recommended inorganic filler formulation data.

3. The recommended data derivation unit, assuming that (II) the blending ratio of the second silica particles in the inorganic filler is c (mass%), (III) the blending ratio of the third silica particles is d (mass%), and (IV) the blending ratio of the fourth silica particles is b (mass%), then the following formula (2) -0.205×b+0.125×c+0.244×d-0.0151×b ２ -0.0102×c ２ -0.0079×d ２ -0.0154×b×c-0.0104×c×d+0.0252×d×b+0.964≧0・・・・(2) The prediction device according to claim 1, which derives the above as recommended inorganic filler formulation data.

Images