Materials Informatics- Overview of MI2I and handling of data –
MI2I Program Manager/JST
Satoshi ITOH
Feb. 29, 2016
Two approaches to material and device development
Social and urban infrastructures are based on materials science
Solid laser
−1
2∇2+w r Φ(r)=EΦ(r)
Direct problem scheme
Conventional approach
Based on the fundamental physics, e.g. DFT, we investigate material properties .
Inverse problem scheme
Materials Informatics
GMR head
Superconductivity
By getting help of the data science, we explore materials having a desirable properties.
Materials database
MatNavi
API/UI
B
A
C
X Z
Y
Materials informatics application tools
Magnetic
Thermal
Battery
Target subjectsNew subjects
Open data‐platform
National project on materials informatics “Materials Research by Information Integration” Initiative (MI2I)
Start : July 1, 2015Period: 5 yearsBudget: about 2 billion JPY
Key point
We not just prepare materials database and materials informatics tools as a useful elements, but also we propose a solution scenario for problem by "connecting the dots".
Project leader: Kiyoyuki Terakura (NIMS)Program manager: Satoshi ITOH (JST)
MI2ISociety/Industry
Construction of “all‐Japan” structure , “connecting the dots”
“all‐Japan” by inter‐regional cooperation• NIMS/JST as hub organization• Academia network・・・Hokkaido Univ.,
Tohoku Univ., Univ.Tokyo, TUT, JAIST, Nagoya Inst.Tech., Kyoto Univ., Osaka Univ., Hiroshima Univ., AIST, IMS, JFCC
• Translational research satellites東北大
統数研阪大
北陸先端大
名工大JFCC
東大
京大
東工大
産総研
北大
広大
Government/ participation of Municipalities
Network of organizations
National project on materials informatics “Materials Research by Information Integration” Initiative (MI2I)
MI2I consortium ( 4/1/2016 ~ )• Collaboration in equal partnership of participants• Sharing of information and achievements as possible
If you are interested → e‐mail : mii‐[email protected]
“all‐Japan” by inter‐sector cooperation• Industry‐government‐academia cooperation• A new scheme of multi‐client collaboration
Rough information on atomic positions obtained by XRD and HR‐TEM
Grain size
Grain boundary composition
Magnetic property
AI
Surface structure
Interface structure
Crystal structure
TEM image×AInm~μm
Magnetic property
Electronic structure
Crystal structure
Materials database
Forecast of magnetic property
Present image by TEM
Schematic example for magnetic material
See the web site,http://mits.nims.go.jp/index_en.html
“MatNavi” consists of about 20 database (polymer, inorganic materials, superconductivity, etc.) with high reliability.
“MatNavi” provides data visualization tools and simple prediction simulator of material properties.
Materials database system : MatNavi
Genome sequence
CGAGTGTTGAAGTTCGGCGG
Amino acid sequence
APRKNVRWCTISQPEWFKCRR
Protein A • Bioactivity• Isoelectric point• Molecular weight
Bio Database
DDBJ
PDB
KEGGSSDB
Omics Space Analysis
Bio Informatics
≒
Protocol
Protocol
Bio Informatics
Open data
=
DNA sequencer
Function SiC
• Composite ratio• Crystal structure• Impurity content
• Electric resistivity• Dielectric constant• Hardness
Material specification
Materials Informatics
~1024 atoms in a realspecimen
All atomic positionscannot be determined in a real material !
Manufacturing process
Materials database
All atomic species and positions
Materials science/Materials informatics
Open data
Unknown data
=
Characteristics of materials database
All of material functions Information of all atoms in a specimen
Absolutely impossible !!
Composite ratio A
Process α Character 1
Process β Character 2
Public databasePrivate database
Ideal database
Austenite Modified austenite
Ferrite phase
Ferrite and pearlite
Hot rolling process(熱間圧延プロセス)
Fe+C+Mn+ …
Grandmaster craftsman
Swordsmith (刀鍛冶)
Characteristics of materials database
Categorizing of material functions• Function A • Function B• Function C• ・・・
Materials data
Manufacturing process(It is better than no data, even if insufficient)
+Informatics
Custom‐made database
• The manufacturing process is key know‐how in materials development.
• The know‐how is dependent on person.
• Manifestation of know‐how• Quantification of know‐how• ・・・
SPring‐8
TEM
The most reliable means is specimen box ! ( Materials DB ?)
Characteristics of materials database
Materials data
Manufacturing process
Custom‐made database
The manufacturing process is key know‐how in materials development.
Sharing know‐how ?
Specimen box
+
Open data Share data Closed data
Standard dataBibliographic data
CollaborationConsortium
Private know‐howUnpublished
Unser certification Unser certification Unser certification
User communities Data Classification and Handling Policy
Concluding remarks
A new national project called MI2I on materials informatics, which is a collaboration between materials science and information engineering, has been started from July 1, 2015, in Japan.
The knowledge of manufacturing process is key information on the development of materials, which is dependent on person as know‐how.
Next‐generation materials database contains standard material properties, functions, manufacturing processes, together with know‐how and specimens.
We have to discuss on data classification and handling policy, and incentive of database developer and data providers.