JP5614259B2 - Material action compounding analysis data management device - Google Patents

Description

  The present invention relates to a substance action combination analysis data management mechanism (material development analysis device) that supports material design based on selection of optimum material candidates.

  Conventionally, material development often includes complicated processes, and there are infinite degrees in each process and operation, so development is often difficult. In addition, since the past experimental data at the material development site is enormous, it takes a considerable amount of time just to look through all the experimental data. On the other hand, needs from the field have been diversified, and it is required to manage data including material properties and to develop the possibility to be extremely wide from those data.

  For example, in the field of semiconductor packages, the ratio of small and thin surface mount packages has increased, and stacked MCPs (Multi-Chip Packages) in which chips are stacked in multiple stages are rapidly increasing. By stacking multiple semiconductor chips in this way, the storage capacity can be increased several times, so it is frequently used in applications such as flash memory and memory for embedded software. In the manufacture of such a package, it is possible to develop a semiconductor package that efficiently satisfies the requirements by blending a wide variety of materials.

These multi-layered chip structures require a variety of die bonding film lineups such as chip / substrate, chip / chip, and wire filling, and the characteristics required for each film are also different. In addition, the manufacturing process of the semiconductor package is composed of many processes such as dicing, pick-up, die bonding, and sealing, and there are various requirements for the film in each process. In particular, at the beginning of the development of a new package, the required characteristics of the material are often unclear, and in practice, the materials are often rubbed together while trying materials having different characteristics.
Material design and development has been carried out for many years, but it has been wasteful in terms of time and labor due to failure to save and reuse various data related to compounding experiments obtained in the past. Development is often done, which hinders more efficient material design and development.

  As described above, due to inefficient data reuse and recording methods based on material design experiments, accumulated data obtained in the past cannot be used effectively, and the development speed of new materials tends to be delayed. is there. Also, only the amount of materials to be blended and the order of the processes tend to be focused on, but which material can be selected and blended from among a large number of candidates leads to the development of products that effectively meet the requirements, and semiconductor mounting In turn, the bottleneck in the development of the electronics industry is eliminated. An example was given above, but not only material design including semiconductors, but also data recording techniques and reuse in various industries such as automobiles, communications, power, clothing, food, economic decision making, social decision making, etc. Method development is becoming a bottleneck.

  In the development of new materials, referring to past data one by one is not realistic in the present age when the speed of material development is increasing. However, performing the same experiment as the past data causes extra costs such as labor costs and time.

  Therefore, if the past experimental data can be grasped efficiently, unnecessary costs can be reduced and the speed of material development can be increased. This places importance on secondary use of experimental data.

  In particular, as an example of failure, for example, when developing a new material X, when there are characteristics A, B, and C, only A is insufficient and B and C may be just right. In this case, by increasing A, all the characteristics can be set within the target value range. When it is difficult to change the characteristic A, it is often possible to improve the characteristic B and the characteristic C by using an alternative material. For example, even if data is recorded about the characteristic data and blending amount obtained in an experiment using an alternative material in the past, the purpose and policy of the new experiment are different. Often, they are unaware of the use of alternative materials without being aware of past experimental examples. In that case, there are often cases where compromises must be made in situations that are not optimal for the purpose.

  Also, Japanese Patent Application Laid-Open No. 2004-070882 discloses a method for automatically generating metadata of data provided by a user and a metadata creation program for such a problem. This is for supporting information sharing and disclosure on a portable terminal, and is not intended for material design. Japanese Patent Laid-Open No. 2006-113984 discloses a system that improves the convenience of providing information to users by evaluating the similarity of metadata regarding the contents of document data and geographical data. All of them are related to simple metadata creation without handling numerical data and characteristics such as material design and scientific law.

  By adding a function for the purpose of reusing experimental data to material development simulation that supports material design, there are materials to be used to some extent in the development site when the characteristic values to be formulated are determined. However, there may be a need to search what materials can be added to meet the target characteristics, or whether experiments have not been conducted with the same target characteristics in the past. There was no equipment for material design development by reusing the data.

JP 2004-070882 A JP 2006-113984 A

  The present invention deals with a data description method for extracting past experiment information in material design as shown in the above semiconductor device as an example, and a database for storing past experiment data and relevance and law in material composition This database and data description method consisted of two databases that store data, etc. By using these databases and the data description method, past experimental data in material development that can efficiently retrieve information required by users from a vast amount of information A substance action combination analysis data management mechanism provided with a combination substance discovery mechanism based on efficient secondary use is provided.

The present invention is as follows.
(1) One or more pieces of information selected from the characteristics, cost, scarcity, availability, environmental load, environmental compatibility, toxicity, carbon dioxide emissions, and other parameters specified for each substance. The results and features specified by the user can be linked and saved, and can be recalled. In simulation, the information can be input as conditions and the strength of the conditions by entering conditions, and fragmented. Substance action analysis data management mechanism that can find materials and materials suitable for compounding and provide the information to users by reverse simulation using a database based on data reuse .
(2) All the data and knowledge obtained from the simulation results of material design can be stored, and the simulation result data is automatically analyzed, and the information already obtained, such as its material properties, characteristics, inventory, and environmental impact The substance action combination analysis data management mechanism as described above, which allows a user to modify physical properties and knowledge, material engineering properties and knowledge, and intentional data as necessary.
(3) Consists of databases with at least two different purposes, virtually or physically, and one is a high-level scientific law, environmental load, constraints, material properties, hazards, safety, etc. A generalist database that stores information with concepts, a specialist database that stores data such as characteristics and blending amounts obtained through experiments, and a specialist database that stores user-specified characteristics and features The substance action combination analysis data management mechanism comprising a database that enables correction and deletion.
(4) Quantitatively enter each substance and action characteristic value, give the generated substance, target value and target range of action, and even if the type and quantity of the substance are unknown, the conditions from the database Data of matching materials and candidate substances can be extracted, and the generated substances, substances that satisfy the target range of action, combinations of action amounts, generated substances, and each characteristic value of action are output. Thus, the substance action combination analysis data management mechanism comprising a program group for virtually performing a combination experiment and a mechanical device for realizing the combination experiment.
(5) Based on the characteristics and characteristics stored in the specialist database, it is possible to present an experimental example to be searched at high speed based on the conditions given by the user, and further, information and composition related to fragmentary characteristics Even when possible substances are given as conditions, materials are not searched from all experimental examples, but experimental examples are automatically searched from candidates narrowed down from the conditions given by the user, and materials that are highly likely to be blended from there. The substance action combination analysis data management mechanism which can provide the user with the information sorted in the order of.
(6) In the production of a compound composed of a plurality of substances and actions and having a plurality of characteristics, extraction of data from a specialist database relating to a substance satisfying a target property value and design of actions based on the extraction are efficiently performed. The substance action combination analysis data management mechanism comprising the calculation device.
(7) For a given target characteristic value, material used, and blending amount, it is possible to exclude experimental possibilities that are less meaningful from the generalist database and access the specialist database to extract information about the remaining experimental possibilities. In addition, when there is a high possibility of being extracted or when there is no possibility, a reverse simulation function that can automatically calculate the formulation and reduce the possibility or increase the possibility, and a program that realizes it The substance action combination analysis data management mechanism comprising a group.
(8) All calculation results can be managed and saved for each keyword, and all data in actual experiments and simulations after extraction can be managed with keywords. Desired substances can be obtained by giving keywords and conditions as a search. The above-mentioned substance action combination analysis data management mechanism comprising a calculation device characterized in that the candidate and the amount of the combination can be presented, and they function automatically in the analysis of the substance action combination as well as the manual search.

  By using the substance action compounding analysis data management mechanism (material development analysis device) of the present invention, the fragmentary or complete substance type, the fragmentary or complete material type, the fragmentary or complete characteristic, the respective quantitative quantity Qualitative and qualitative data allows the system to automatically analyze candidate materials, manifesting properties, and accompanying information, search, sort, and output similar experimental examples from the past. In addition, efficient material design has become possible.

It is a block diagram which shows the substance effect | action combination analysis data management mechanism (apparatus) which concerns on this invention. It is a block diagram which shows one Example of a substance effect | action combination analysis data management mechanism at the time of storing an experiment example in a database. It is the flowchart which showed one procedure of the search of the specialist database shown in FIG.

Hereinafter, embodiments of the present invention will be described in detail.
The substance action combination analysis data management mechanism (substance action combination analysis data storage and extraction mechanism) of the present invention can also be expressed as a substance combination analysis apparatus, action combination analysis apparatus, substance combination analysis mechanism, or action combination analysis mechanism.
Therefore, in the present invention, the substance action combination analysis data management mechanism can store all the simulation result data or a part of the data in the simulation result storage device or can call it, and data that meets the conditions when calling Groups can be extracted and sorted, and by using the data, a new substance can be created by combining and combining substances with multi-element / multi-component properties. A combination of arbitrary types and amounts, and a method for finding a candidate for the characteristics of the substance desired by the producer and a satisfactory combination for each characteristic of the newly generated substance can be searched.

  In addition, the substance action combination analysis data management mechanism combines any kind and quantity from multi-element / multi-component actions when combining actions with multi-element / multi-component properties to generate new actions, For each characteristic of the action to be newly generated, a method for finding a candidate for the characteristic of the action desired by the creator and a satisfactory combination using the data stored in the simulation result storage device was formulated. The action can be searched and all or part of the result can be stored again in the simulation result storage device.

In addition, the substance action combination analysis data management mechanism digitizes and inputs the characteristic values of each substance and action, gives the generated substance, the target value and target range of action, and generates the generated substance and target range of action. The combination of the substances that satisfy the conditions, the combination of action amounts, the substances generated at that time, and the characteristic values of the action can be output, and these data can be stored in the simulation result storage device, and the combination experiment that calls these data can be performed. It consists of a group of programs to be executed virtually and a mechanical device for realizing it.
Furthermore, the substance action combination analysis data management mechanism can compare and calculate the degree of divergence between the data in the actual experiment and the data in the simulation results, and also store the divergence degree together. In addition, when there are a plurality of substance candidates, an analysis program group is provided that can compare the results for each candidate regarding the action, amount, and characteristics when the substances are combined.

  In addition, the substance action combination analysis data management mechanism can manage and store all calculation results for each keyword, and can also manage all data in actual experiments and simulations after extraction with keywords. By providing as a search, the desired substance candidates and the amount of formulation can be presented, which comprises a computing device characterized by automatically functioning in the analysis of substance action formulations as well as manual searches.

  In addition, the substance action combination analysis data management mechanism can calculate the similarity of the material with the simulation performed in the past, can present the candidate substance based on the similarity, or if the similarity is completely matched The present invention is characterized in that the calculation time can be reduced by presenting data stored in the specialist database without performing a re-experiment.

  The substance action combination analysis data management mechanism of the present invention is extremely effective when the desired target value is objective and clear, such as industrial materials, due to the characteristics of data comparison and reuse methods, Even if the desired target value is subjective, it is also effective when it is difficult to express numerical values or the like and it is somewhat unclear. Examples of such applications include cooking, perfume blending, whiskey blending, spice blending, fashion fields such as clothing, color schemes for interiors, pharmaceuticals such as herbal medicines and synthetic drugs.

  In addition, not only substances but also manufacturing conditions such as temperature, humidity, speed, etc., or color, light, sound, vibration, touch, smell, etc. For example, it may be other than a substance. In the present invention, the term “action” is used as a term representing these. It is difficult to separate the substance and action, and it is necessary to use the substance and action in combination in order to expand the application range.

  For example, it is possible to perform an analysis for enhancing the driver's satisfaction by integrating all the feelings of use of a wheelchair while driving a car, engine sounds, vibrations of a vehicle body, touch of a steering wheel, and the like. Similarly, taking automobiles as an example, it is possible to simultaneously analyze objective numerical values such as fuel efficiency and acceleration related to environmental compatibility and subjective numerical values such as sound, vibration, and touch, thereby increasing satisfaction.

Specific examples of the material include the following materials in the case of a die bond film for semiconductors. Examples of substances constituting the die bond film for semiconductor include a high molecular weight component, a thermosetting component, an inorganic filler, a curing accelerator, a catalyst, a coupling agent, and an additive.
Specific examples of the high molecular weight component include polyimide, polystyrene, polyethylene, polyester, polyamide, butadiene rubber, acrylic rubber, (meth) acrylic resin, urethane resin, polyphenylene ether resin, polyetherimide resin, phenoxy resin, and modified polyphenylene ether. Examples thereof include resins, polycarbonates, and mixtures thereof. Epoxy group-containing (meth) containing a functional monomer such as a high molecular weight component having a weight average molecular weight of 100,000 or more, for example, a functional monomer such as glycidyl acrylate or glycidyl methacrylate, and having a weight average molecular weight of 100,000 or more An acrylic copolymer etc. are mentioned. Examples of the epoxy group-containing (meth) acrylic copolymer include (meth) acrylic ester copolymer and acrylic rubber. The acrylic rubber is a rubber mainly composed of an acrylate ester and mainly composed of a copolymer such as butyl acrylate and acrylonitrile, a copolymer such as ethyl acrylate and acrylonitrile, or the like.

  In addition, as the thermosetting component, there are epoxy resin, cyanate resin, phenol resin and its curing agent. For example, when epoxy resin is used as an example in terms of high heat resistance, the epoxy resin is cured and bonded. It is not particularly limited as long as it has an action, and examples thereof include bifunctional epoxy resins such as bisphenol A type epoxy resins, and novolak type epoxy resins such as phenol novolac type epoxy resins and cresol novolac type epoxy resins. Moreover, what is generally known, such as a polyfunctional epoxy resin, a glycidyl amine type epoxy resin, a heterocyclic ring-containing epoxy resin, or an alicyclic epoxy resin, can be applied.

As the inorganic filler, aluminum hydroxide, magnesium hydroxide, calcium carbonate, magnesium carbonate, calcium silicate, magnesium silicate, calcium oxide, magnesium oxide, alumina, aluminum nitride, aluminum borate whisker, boron nitride, crystalline silica, Examples thereof include amorphous silica and antimony oxide. In order to improve thermal conductivity, alumina, aluminum nitride, boron nitride, crystalline silica, amorphous silica and the like can be mentioned. For the purpose of adjusting melt viscosity and imparting thixotropic properties, aluminum hydroxide, magnesium hydroxide, calcium carbonate, magnesium carbonate, calcium silicate, magnesium silicate, calcium oxide, magnesium oxide, alumina, crystalline silica, non-crystalline silica Examples thereof include crystalline silica. In order to improve moisture resistance, alumina, silica, aluminum hydroxide, and antimony oxide can be used.
In addition, a curing accelerator, a catalyst, an additive, a coupling agent, and the like can be given.

Further, there are various functions such as temperature, humidity, speed, etc. during film production, temperature during curing, time, film thickness, width, thickness variation, and color.
In addition to physical properties such as heat resistance, adhesive strength, elastic modulus, viscosity, thermal expansion coefficient, breaking strength, material costs, ratio of non-petroleum materials, toxicity, recycling rate, recyclability, Costs such as transportation distance related to material procurement, amount of heat used, carbon dioxide emission related to production, and environment-related characteristics can be mentioned.

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings, but the present invention is not limited to these embodiments.
FIG. 1 is a block diagram showing an embodiment of a substance action combination analysis data management mechanism according to the present invention. First, the user inputs conditions regarding materials and characteristics used from the terminal 10. Information input by the user in FIG. 1 is transmitted to the experiment condition input unit 11 via the terminal 10. The information about the material and properties transmitted here is generated by the type of material, the type of property, the material used, the property used, the property value of each material, the maximum usage of each material, the minimum usage of each material. The target value of each characteristic of the compound, the maximum target value and the minimum target value of the compound to be produced, and the substances and keywords that are unknown in compounding and the characteristics thereof.

  In FIG. 1, only the characteristics of the information such as the characteristic values and usage amounts of the materials input by the user are transmitted to the feature extraction unit 12, and information indicating the type of experiment is sent to the search target specifying unit 13 by the generalist database 19. Sent. Further, the numerical data of the experiment is transmitted from the experiment condition input unit 11 to the search target specifying unit 13. Quantitative and qualitative data obtained by the search target specifying unit 13 are transmitted to the specialist database 18 by the data search unit 14, and an experimental example in which feature data satisfies the condition and an experimental example in which numerical data satisfies the condition are sequentially stored in the memory 15. Sent to. The data of the experiment example in the memory 15 is output by the data organizing unit 16 to the user terminal 10 through the calculation result output unit 17 after setting the data of the name and amount of the substance intermittent to the experiment example.

FIG. 2 is a block diagram showing an embodiment of a substance action combination analysis data management mechanism when an experimental example is stored in a database.
When the user saves the laws, physical characteristics, theoretical characteristics, and scientific knowledge of the experiment of the target category in the database, the user transmits the data to the data management unit 22 via the knowledge input unit 21. After being converted into a format that can be stored in the database, it is stored in the generalist database 19.

  In addition, when the user inputs the substance name, characteristics, and other data related to all experiments, the data input from the terminal 10 is shaped into a format that can be stored in the database via the experiment result input unit 20, and the data The data is stored in the specialist database 18 through the shaping unit 26.

  In the specialist database 18, the user inputs not only the data indicating the substance name, the blending amount, and the characteristics, but also the characteristics of the material obtained by the blending other than the characteristics of the substance itself required for the material design by the terminal 10. Is possible. Even when the user does not input, the experiment example newly updated in the specialist database 18 at the update timing set by the user is automatically collated through the generalist database 19 and the data collating unit 24, and past experiments are performed. Are automatically classified by calculation in the condition determination unit 25. At the same time, the consistency evaluation unit 23 can evaluate whether it is consistent with the knowledge stored in the generalist database 19, and all the results are provided to the user on the terminal 10 via the data verification unit 24. Is done. If the user needs to correct the result, the user can correct the result, and the corrected experimental example is overwritten in the specialist database via the data matching unit 24.

  FIG. 3 is a flowchart showing a procedure for searching the specialist database shown in FIG. First, in the first step 30, the target property value and the substance attribute to be blended in fragments are input, and then data satisfying the attribute value input in the second step 31 is inquired and extracted from the specialist database 18. Sorting from data close to the characteristic value in step 32, data having the same plurality of characteristic values in the fourth step 33 are sorted in ascending order of the blending amount, and substances and blending amounts to be newly added in the fifth step 34 are displayed on the user terminal. To do.

10: terminal, 11: experiment condition input unit, 12: feature extraction unit, 13: search target specifying unit, 1
4: data search unit, 15: memory, 16: data organizing unit, 17: calculation result output unit, 18: specialist database, 19: generalist database, 20: experiment result input unit, 21: knowledge input unit, 22: data management Part 23: consistency evaluation part 24: data collation part 25: condition determination part 26: data shaping part 30: first step 31: second step 32: third step 33: fourth step , 34: Fifth step

Claims (7)

Experimental results on one or more information selected from the characteristics, cost, rarity, availability, environmental impact, environmental compatibility, toxicity, carbon dioxide emissions and other parameters specified for each substance the user saves to link the properties and characteristics specified, you can call the information, when the reverse simulation, the information can be input as the intensity of the condition and the condition at the input of the condition, and, fragmentary It is an input condition or information, the inverse simulation using the database based on the reuse of data to discover a substance or material which is suitable for formulation, substances acting formulation analysis data management apparatus that can provide the information to the user Because
At least a database that has two or more different purposes, either virtually or physically, one of which has high-level concepts such as scientific law, environmental impact, constraints, material properties, hazards, safety, etc. A generalist database that stores data, and a specialist database that stores data such as characteristics or blending amounts obtained through experiments. In addition, in the specialist database, the characteristics or characteristics specified by the user can be stored and modified. Or a database that allows deletion,
A data collation unit for collating data updated in the specialist database with information stored in the generalist database;
A substance action combination analysis data management device further comprising a condition determination unit for classifying data updated in the specialist database based on information stored in the generalist database . All the data and knowledge obtained from the simulation results of material design can be stored, and the simulation result data is automatically analyzed to cite the information already obtained, such as its material properties or characteristics, inventory quantity, environmental impact, etc. 2. The substance action combination analysis data management apparatus according to claim 1, wherein the physical action combination analysis data management apparatus according to claim 1, wherein the physical property and knowledge, material engineering characteristics and knowledge, and intentional data can be corrected as necessary. Each substance and action characteristic value is digitized and input, and the generated substance, action target value and target range are given, and even if the type or amount of the substance is unknown, the material that meets the conditions from the database Data and candidate substances can be extracted, and the combination of the generated substance, the substance that satisfies the target range of action, the combination of action amounts, and the generated substance and action characteristic values are output. The substance action combination analysis data management device according to claim 1 or 2 , comprising a program group for performing an experiment virtually and a mechanical device for realizing the program group. The characteristic or characteristics of the data stored in the Specialist databases, based on the conditions given by the user, can present an experimental example to be search, further information or formulation substance involved in fragmentary properties even when given as a condition, without searching all the experimental examples, it retrieves the experimental examples from among the candidates Yu over tHE was squeezed from the given conditions, sorted in order of likely compounded material therefrom The substance effect | action combination analysis data management apparatus of Claim 1 which can provide the information to a user. A computer that efficiently extracts data from a specialist database related to a substance that satisfies a target property value in the generation of a compound that has multiple characteristics and is composed of multiple substances and actions, and designs the actions based on it The substance effect | action mixing | blending analysis data management apparatus of Claim 1 which comprises. Given target characteristic value for use MATERIAL FOR or Blend amount, than generalists database, excluding the thin experimental possible significance can extract access to information to the specialist database for the remaining experiments possible, In addition, when there is a high possibility of extraction, or when there is no possibility, the inverse simulation function that can automatically calculate the formulation and narrow down the possibility or increase the possibility, and the program that realizes it substance acting formulation analysis data management apparatus according to any one of claims 1 to 5, provided with the group. All calculation results can be managed and saved for each keyword, and all data in actual experiments and simulations after extraction can also be managed with keywords. By giving keywords and conditions as searches, candidates for desired substances and The substance action combination analysis data according to claim 1, comprising a calculation device capable of presenting the amount of the combination, and the management by the keyword automatically functions not only in the manual search but also in the analysis of the substance action combination. Management device .

Images