- If I search for something basic like iron, I get tons of alloys with Fe but no info on the iron itself. Maybe you can work on the search, so this kind of info appears first?
Please, use the terms unary, binary, ternary, quaternary, etc. to restrict the chemical elements number, e.g. see the following query: Materials Platform for Data Science
- So far, I have downloaded via API data relating to physical properties of interest connected with the composition. Still, I have not yet understood if it is possible to connect the same properties with the crystalline structure instead of the composition. Unfortunately, my knowledge of chemistry is quite limited, and I am trying to understand a few things to make the most out of the data.
Sure, in most situations you’ll need both the crystal structures and the physical properties, inter-linked. This linking is achieved via the distinct phases concept. On the MPDS the crystalline structures (S-entries) are tightly connected to the physical properties (P-entries) and to the phase diagrams (C-entries). The standard unit of the MPDS data is an entry (e.g. S377634, P600028, C100027). Another dimension of the MPDS data is the distinct phases; they link the S-, P-, and C-entries together logically, even if they’re reported in the different publications. Each distinct phase has its own unique combination of (a) chemical formula, (b) crystalline space group, (c) Pearson symbol. Only these three together can define the distinct phase uniquely. In addition, each distinct phase has its permanent integer identifier called phase_id. We can find out what’s there for some phase_id e.g. 109070 like this: Materials Platform for Data Science.
So how to get the data via API in an inter-connected manner? First, please request a property of your interest. See the full hierarchy here: Materials Platform for Data Science. Please always save the phase_id’s inside an intermediate data container, then request the corresponding crystal structures (or the other data) for exactly these phase_id’s using the phases API request parameter (see Table 2 in MPDS API docs). This way you’ll be able to group the entries via the phases they belong. See more examples here: mpds-api/kickoff at gh-pages · mpds-io/mpds-api · GitHub (namely, Statistical relationship of physical property and crystalline structure).
- I’m using python (so the mpds_client). I have my API key and everything, but I’m unable to make any requests. I receive
{"error":"Sorry, your subscription plan does not include full access to these data."} when making a request. So I guess I’m having a difficult time telling what is/isn’t publicly available.
See Materials Platform for Data Science license (§3. Open MPDS data).
- I had some first experiences with MPDS, and I have two questions concerning the best search modus. In the first step, I was looking for oxide materials with a melting temperature above 1800 °C which should have similar coefficients of thermal expansion. My search terms were: “oxide”, “binary”, “temperature for congruent melting” combined with “range search”. Would you do it the same way? Are there possibilities for comparing the coefficients of thermal expansion within the results?
Try something along these lines in the MPDS GUI:
https://mpds.io/#inquiry/classes=oxide&numeric=linear thermal expansion coefficient,>%2C0.0001%3Blinear%20thermal%20expansion%20coefficient%2C<%2C0.0003%3Btemperature%20for%20congruent%20melting%2C>%2C2073
Or try the following MPDS API data-miner: mpds-api/kickoff/miner_twofold_props.py at a95c271a1780b5da7fcccd862d9d728da8ba3710 · mpds-io/mpds-api · GitHub
- Can I use digitized Phase Diagram data to get specific points like eutectics?
Surely yes! See more examples on the phase diagrams data-mining here: mpds-api/kickoff at gh-pages · mpds-io/mpds-api · GitHub (namely, Extracting the liquidus lines for a given binary system and Retrieval of binary systems producing no compounds).
- Do you have data about materials manufactured by additive manufacturing?
In connection to the possible complex alloy systems (e.g. Al-Cu-Mg-… etc.) relevant for the 3d-printing, the MPDS provides you a lot. Try searching some combinations of the elements of your interest, there will many relevant hits. You might be also interested in the mechanical properties (see the entire hierarchy of the solid state properties: Materials Platform for Data Science) however please note that the MPDS is a general-purpose data provider. So we don’t really focus on the metallurgy, microstructure, plasticity, etc. per se. One may get such data either from the phase diagrams available, or doing the simulations based on our starting data. I think, no one really knows exactly today which combination of elements works better for each particular 3d-printing technology, and the strength of the MPDS is it just provides you everything that was published in the last 130 years, so you are free to play with these data. Of course, here the MPDS API is also relevant, so if you have some experience with Python, you might try these examples we created to tease your imagination: mpds-api/kickoff at gh-pages · mpds-io/mpds-api · GitHub