Wikidata talk:WikiProject Chemistry/Archive/2015

Active discussions


IP changed a chemical formula: [1]
Regards --Chris.urs-o (talk) 06:39, 24 January 2015 (UTC)

Chemical element in french and english

Hi, I started a discussion on enwiki WikiProject about this. See en:Wikipedia talk:WikiProject Chemistry#Chemical element : french and english definition. TomT0m (talk) 13:24, 7 June 2015 (UTC)

Chemical mixtures

@Snipre: Should chemical mixtures be tagged with material used (P186) or with has part (P527)?--Kopiersperre (talk) 16:04, 12 June 2015 (UTC)

@Kopiersperre: I would say has part (P527). It's totally the kind of relationships it's supposed to express, while material used (P186) seems to be liked not to the parts themselves but what the parts are made of : a wooden ship is made of wood, and has a hull (a wooden hull :) ). But has part would also work for
⟨ wooden ship ⟩ has part (P527)   ⟨ wooden hull ⟩
. TomT0m (talk) 16:33, 12 June 2015 (UTC)
@TomT0m: But has part (P527) is already used for the elemental composition (as seen in ethanol (Q153)). I would like to state Aerozine 50 (Q16831) consists of 1,1-dimethylhydrazine (Q161296) (50 %) and hydrazine (Q58447) (50 %).--Kopiersperre (talk) 16:51, 12 June 2015 (UTC)
@Kopiersperre: A misread ? I say has part (P527) is good :) TomT0m (talk) 16:59, 12 June 2015 (UTC)
@Kopiersperre, TomT0m: has part (P527) is a good start but we need to specify the amount or percentage. Perhaps a specific property can help to avoid to use has part (P527) for everything. But we have to think about a qualifier in any case to specify the percentage. Snipre (talk) 19:33, 12 June 2015 (UTC)
@Kopiersperre, Snipre: : Can we convert the percentage into raw numbers ? there is already a qualifier for this. quantity (P1114)  . I guess percentage qualifier is needed in the cases of classes of variable size molecules with constant ratios of components anyway, if there is some of interests ( I'm not a chemist :) ) ... TomT0m (talk) 09:05, 13 June 2015 (UTC)

I found this thesis who describes (page 13) the "has part" relationship in ChEBI exactly as this. Although I couldn't find in a quick search a real example of this in ChEBI itself The example given in ChEBI tutorial (page 9, search potassium) gives an example of an ionic bond if I'm correct ... TomT0m (talk) 09:34, 13 June 2015 (UTC)

property SMILES: canonical and isomeric

Currently we have only one property for the SMILES notation: canonical SMILES (P233). But SMILES notation can have several notations: one with the isomer description and one without. Do we need to have two properties or to add a qualifier to distinguish both notations ? Snipre (talk) 11:24, 17 June 2015 (UTC)

State & phase

Should state of matter (Q11430) and phase (Q104837) be merged? The former has "phase" as an alias, and they both point to commons:Category:States of aggregation, which suggest so. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:48, 18 June 2015 (UTC)

@Pigsonthewing: Phase is not equal to state of matter. State of matter is limited to solid, liquid, gas and plasma (I know there are some others more exotic) and phase is just "An entity of a material system which is uniform in chemical composition and physical state" according to IUPAC. You can have 2 liquid phases in a bottle (think about mixture of water and oil). Both are liquid but as there is a separation into 2 phases this can't considered as one entity. So phase used as state of matter is not correct and you can't merge both. Snipre (talk) 15:50, 18 June 2015 (UTC)
Fair enough. In that case, how should we disentangle the two items? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:03, 18 June 2015 (UTC)
@Pigsonthewing: By using clear definitions:
State of matter: one distinct form of the matter
Phase: An entity of a material system which is uniform in chemical composition and physical state. Snipre (talk) 08:35, 19 June 2015 (UTC)
But note the issues in my first post in this section: aliases, and a shared commons category. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:55, 19 June 2015 (UTC)
For the alias, we have to delete phase as alias of state, and the commons category we have to choose according to the documents we have in the category. Snipre (talk) 12:30, 19 June 2015 (UTC)


Aren't diamond (Q5283) and graphite (Q5309) instance of (P31) chemical substance (Q79529)?

No, they are
⟨ diamond ⟩ subclass of (P279)   ⟨ chemical substance (Q79529)     ⟩
. Each diamond is a substance and there is many diamond. author  TomT0m / talk page 11:19, 18 August 2015 (UTC)
@TomT0m: It seems to me an instance of (P31) of chemical substance (Q79529). See the registry at NIST WebBook. It can be a class of gemstone (Q83437), but they are chemically equivalent. Almondega (talk) 12:22, 18 August 2015 (UTC)
@Almondega: see In chebi is a is the equivalent of subclass of, if my understanding is correct. author  TomT0m / talk page 12:31, 18 August 2015 (UTC)
@TomT0m: diamond (Q5283) figures also in The Merk Index 13th Ed. (Q20819290) at inventory number (P217) 3007. But there aren't chemical substance (Q79529) that are diamond (Q5283) instances. I honestly do not see how ChEBI ontology corroborate with your thinking. Almondega (talk) 12:54, 18 August 2015 (UTC)
I do honestly do not understand what you mean. Are you aware of basic classification principles explained in Help:Classification ? author  TomT0m / talk page 13:10, 18 August 2015 (UTC)
@TomT0m: Ok, but why exactly diamond (Q5283) would be a subclass of (P279) chemical substance (Q79529)? Almondega (talk) 13:33, 18 August 2015 (UTC)
@Almondega: Because we have
⟨ Dresden Green Diamond (Q706710)     ⟩ instance of (P31)   ⟨ diamond ⟩
Each diamond is a pure substance. Like each bottle of Helium we use for parties with childrens for balloons. One bottle and one diamonds are tokens in the type/token distinction principle, real objects of the real world. Diamond is a class of real world object, a type in the type/token distintion sense, with a lot of instances. Similarly, pure substance is a class of real world object, the class of all real world objects who are chemically pure. Then the set of all diamonds is a subset of all pure chemical substances, and this is precisely what
⟨ Diamond ⟩ subclass of (P279)   ⟨ pure substance ⟩
 means. If we want to say that "diamond" is an instance of something, then the something is a metaclass. This could be labelled "chemical substance type" for example. author  TomT0m / talk page 14:15, 18 August 2015 (UTC)
@TomT0m: It is not yet clear to me. By this reasoning silicon dioxide (Q116269) should be a chemical subclass since describes several instances of that material in the world. Almondega (talk) 14:53, 18 August 2015 (UTC)
@Almondega: There is an ambiguity that we can retrieve all other chemistry, about substances and the molecules (or assimilated) they are made of, included into basic definition like chemical elements. For example if you search "silicon dioxyde" in chebi you'll find : then you'll find ( more precisely) that it's a molecular entity. A type of molecules, said in other words. Then there may be some substances made of "silicon dioxyde molecule" ... we could call them "pure silicon dioxyde substances". "silicon dioxyde molecule" would be a subclass of molecules, and "pure silicon dioxyde substance" would be a subclass of chemical, with
⟨ pure silicon dioxyde substance ⟩ has part (P527)   ⟨ "silicon dioxyde molecule" ⟩
As I said, this distinction is often ambiguous in everyday chemistry as the same word is often used for the two concepts, for example a "chemical element" may be a substance or a type of atoms depending on the definition you make (frwiki and others made the second, enwiki made the first, so we have a beautiful iw conflict to manage ...). author  TomT0m / talk page 15:21, 18 August 2015 (UTC)
@TomT0m:I'm increasingly confused :( Almondega (talk) 15:33, 18 August 2015 (UTC)
It's just that your question raised another ambiguity :) but essentially yes you were right, silicon dioxyde substance is a subclass of chemical according to the type/token distinction principle. author  TomT0m / talk page 15:44, 18 August 2015 (UTC)

Properties for IDs in commercial vendors?

ZINC database (which is itself not commercial, a property proposed at Wikidata:Property_proposal/Natural_science) lists hundreds of commercial vendor identities, for example. Should these be Wikidata property?--GZWDer (talk) 11:50, 2 September 2015 (UTC)

Why not? There are no free sources of safety data sheet (Q222067). At least Sigma-Aldrich (Q680841) should be supported, as Template:Sigma-Aldrich (Q10663387) is used e.g. over 4000 times in German wikipedia.--Kopiersperre (talk) 19:12, 2 September 2015 (UTC)
Not a good thing. Sold chemicals are not pure and commercial product should be represented as specific items describing technical, puriss or purum grades. Instead of using commercial MSDS data which are often influenced by the composition and the precence of impurities we should focus on authorities databases: they can provide all the data of a MSDS with a better reliability. Snipre (talk) 07:32, 6 September 2015 (UTC)

Qualifiers for CAS number

In my opinion CAS Registry Number (P231) should allow racemic mixture (Q467717), (+) and (-) as qualifiers. For pharmaceuticals, it would be useful to have also the CAS of the hydrochloride (Q420060).--Kopiersperre (talk) 16:35, 13 September 2015 (UTC)

@Kopiersperre: No, the rule for mixture is to create a new item:
  • one item for the (+) molecule
  • one item for the (-) molecule
  • one item for the racemate
  • one item for the hydrochloride form. Stop mixing data from different molecules in the same item. Snipre (talk) 18:40, 1 October 2015 (UTC)

simple substance (Q2512777)

Hi guys, am I totally wrong or is this the real equivalent for "chemical element" in english @Emw, Snipre:. author  TomT0m / talk page 18:53, 15 September 2015 (UTC)

Wikidata Lists

The template Template:Wikidata list is a real oppertunity in my opinion to get more people interested in Wikidata. A lot of topics that are left to the category-system of Wikipedia can be easily turned into a much more practical list: User:Tobias1984/List of uranium compounds (compare to Category:Uranium compounds). And many languages of Wikipedia don't even have such a detailed category-system and would profit even more from structured information with minimal maintainance. --Tobias1984 (talk) 15:19, 16 September 2015 (UTC)

Chemical hazard information


You may have gotten several pings about it through the {{Ping project}} template, but I have requested the creation of several properties regarding chemical hazards (and chemical information in general). This is in the context of the planned importation of the contents of the Pocket Guide to Chemical Hazards, a publication of the National Institute for Occupational Safety and Health, the U.S. government research agency dedicated to health and safety in the workplace.

The properties I proposed are ionization energy, lower flammable limit, upper flammable limit, molecular mass, minimum explosive concentration, relative density to air, OSHA liquid class, chemical exposure route, symptoms from overexposure, organ affected by chemical exposure, IDLH, NIOSH recommended exposure limit, OSHA permissible exposure limit, type of exposure limit, first aid response to exposure, and personal protective measure against chemical exposure. These proposals are listed on Wikidata:Property proposal/Natural science.

Snipre recommended that I reach out to this WikiProject to discuss how we might, in general, want to approach including chemical safety information on Wikidata. For most of these properties I proposed them based on what information there was in the Pocket Guide. While the Pocket Guide is an authoritative source of information, it is not my intention to make the it the only source for this information, but rather, as a convenient source for mass-importation. That considered, are there any thoughts on how best to proceed? James Hare (NIOSH) (talk) 16:27, 1 October 2015 (UTC)

My remark was about your proposals which were "limited" to the Pocket Guide to Chemical Hazards. In WD we should find a way to be able to add data from others databases and if this is possible, to merge properties when it is possible. Snipre (talk) 18:44, 1 October 2015 (UTC)
Referring to the proposed properties for recommended exposure limits and permissible exposure limits? I proposed them as "NIOSH recommended exposure limits" and "OSHA permissible exposure limits" recognizing that they are specifically the recommendations/regulations of the United States and other agencies (in Europe, South Korea, etc.) may have different recommendations or laws. I am supportive of a generalized property. For exposure laws, we could have a general permissible exposure level field with qualifiers for the affected jurisdiction. For example, where the U.S. agency OSHA sets an exposure limit, we would qualify that limit by saying it's effective in the United States (and whichever jurisdictions have an identical rule). For recommended (non-binding) exposure limits a different approach may be necessary, since in theory any reasonably credible organization can make a recommendation, whether a government agency like NIOSH or a reputable non-governmental research organization. Is there a "recommended by" property or something like that we could use? James Hare (NIOSH) (talk) 19:21, 1 October 2015 (UTC)
Specifically, I would recommend a property "legally permissible exposure limit" with the qualifier applies to jurisdiction (P1001) and then a property "recommended exposure limit" with the qualifier being either publisher (P123) or maintained by (P126). If people agree, I will withdraw the current proposals and propose those instead. James Hare (NIOSH) (talk) 16:54, 2 October 2015 (UTC)

If no one objects to my proposal above, I will propose it next week and withdraw my current proposals for NIOSH/OSHA specific properties. James Hare (NIOSH) (talk) 16:33, 16 October 2015 (UTC)

I have now submitted proposals to Wikidata:Property proposal/Natural science. Comments welcome. James Hare (NIOSH) (talk) 20:27, 21 October 2015 (UTC)

trade name / INN

Hi, how do you suggest solve this conflit: barnidipine (Q925327), Mepirodipine (Q6817818)? they must be two different items? Because the first is about the substance and the second one is a commercial product of that substance. --Almondega (talk) 17:33, 19 October 2015 (UTC)

Some WP create article for trademark so we can't avoid the item creation for this kind of thing. But we have to be clear: a trademark can't have a CAS number or a PubChem ID. Only item with "instance of: chemical compound" can have these properties. This is the work of people creating items for trademark to link the trademark item with the active substance item. Most of the time sold products are formulation so this is a mistake to mix formulation of active substance with the active substance. Snipre (talk) 16:11, 20 October 2015 (UTC)

Overlap in category items

Category:Terpenes and terpenoids (Q6470754), Category:Terpenes (Q13553578) and Category:Terpenoids (Q10259435) seem to overlap. The problem seems to be in the first item that contains e.g. en:Category:Terpenes and terpenoids. Any idea on how this issue may be solved? --Leyo 23:13, 26 October 2015 (UTC)

Category:terpenes, subclass category:terpenes and terpenoids
Category:terpenoids, subclass category:terpenes and terpenoids
Regards --Chris.urs-o (talk) 06:01, 27 October 2015 (UTC)
I do not really understand your proposal, i.e. concerning how it would solve the issue. --Leyo 13:50, 27 October 2015 (UTC)
If category:terpenes and category:terpenoids are a subcategory of category:terpenes and terpenoids the problem is defused for now. --Chris.urs-o (talk) 14:56, 27 October 2015 (UTC)
I still don't see what exactly you intend to do. Just do it if you are sure. --Leyo 22:17, 27 October 2015 (UTC)


Are these items the same thing: flammable object (Q21474084)/combustible matter (Q42501)? Should be merged? --Almondega (talk) 13:28, 15 November 2015 (UTC)

Almondega No. An object can be defined in term of localization. A combustible matter not. So as flammable object can have some different properties we can't merge. An flammable object is an object composed of a combustible matter. Snipre (talk) 09:50, 26 November 2015 (UTC)
+1. See also flammability (Q910035). --Leyo 11:11, 26 November 2015 (UTC)
@Snipre: Isis trucks containing gaz that were destroyed this last month were localized ... author  TomT0m / talk page 11:13, 26 November 2015 (UTC)
??? Snipre (talk) 17:21, 26 November 2015 (UTC)
Gaz may be somewhat liquid, but a certain part of this is certainly localized. author  TomT0m / talk page 17:29, 26 November 2015 (UTC)

Wikimania 2016

Only this week left for comments: Wikidata:Wikimania 2016 (Thank you for translating this message). --Tobias1984 (talk) 11:59, 25 November 2015 (UTC)

Data donation

Please see Wikidata:Bot_requests#Data_from_English_Wikipedia_chemboxes.2C_drugboxes_and_relevant_infoboxes. --- Jura 11:47, 4 December 2015 (UTC)

Return to the project page "WikiProject Chemistry/Archive/2015".