Wikipedia talk:WikiProject Chemicals/Archive 2017

Missing lower case redirects
For some reason, en.wikipedia prefers article names of chemicals to be in upper case, e.g. 1,1-Dibromoethane (not 1,1-dibromoethane). As in this example, the lower case name, i.e. how the name would appear in the text, is a red link. There are many more such red links that need to become redirects, e.g. these and those (some are false-positives). Is there an easy way to create the missing redirects? --Leyo 15:32, 10 January 2017 (UTC)
 * It might be an interesting feature-request to make the following expansion of long-standing wikilink handling: "treat first alphabetic character as capital even if other non-alphabetic characters are before it ". Currently foo is treated as foo, so this would extend the magic to 1foo being 1foo . Take it to VPT (do they have a perennial-proposals archive)? DMacks (talk) 17:03, 10 January 2017 (UTC)
 * My intention was that the missing redirects are being created to eliminate red links. The feature request is probably a good idea, but may take way to long to get implemented (if at all). --Leyo 23:30, 10 January 2017 (UTC)

WikiJournal of Medicine promotion
T.Shafee(Evo &#38; Evo)talk 10:33, 19 January 2017 (UTC)

Inconsistency in article names of dyes
There are dye article names in upper case such as these and in lower case such as those Shouldn't we have a consistency in the naming? Please have a look at Category:Azo dyes and other articles in (subcategories of) Category:Dyes. --Leyo 15:11, 17 January 2017 (UTC)
 * Texas Red
 * Disperse Yellow 26
 * Disperse Orange 1
 * Sudan Red G
 * Oil Yellow DE
 * Brilliant cresyl blue
 * Titan yellow
 * Indian yellow
 * Acid orange 7
 * Azo violet
 * I guess that the upper case spellings should be used since these are proper names.


 * Dyes and pigments are a problem in Wikipedia, at least for those that are not used in biomedical area. Unlike what most of us edit, they are actually used for something in the real world.  Few editors have knowledge or interests in this area.  One start would be to construct a work list of the most pervasive dyes and pigments.  --Smokefoot (talk) 18:27, 17 January 2017 (UTC)
 * Well, there are List of dyes and List of inorganic pigments. --Leyo 19:25, 17 January 2017 (UTC)
 * I am surprised that nobody really cares about this mess. :-( --Leyo 23:18, 27 January 2017 (UTC)
 * I don't know what the answer is, other than the usual "check what refs say" regarding whether they are proper nouns or not. I just checked one set, and here is Sigma-Aldrich using "Sudan R ed" in a non-title context. That ref cites this, a European Food Safety Authority press release, that likewise capitalizes it and several other classes of dyes. Compare that to this article that does not capitalize brilliant cresyl blue. DMacks (talk) 23:35, 27 January 2017 (UTC)

Depicting hydrates vs anhydrous salts, slightly nerdy MOS concept
User:Riventree posed a question at Talk:Iron(II) sulfate: "I understand that this compound occurs as a hydrate, but since the data box chemical formula is FeSO4, perhaps the images should match this formula?"
 * This simple comment implies a couple of issues that we might try to converge on and then get into an MOS.
 * For Ferrous sulfate the ChemBox could just show the structure of anhydrous FeSO4, not the anhydrous material of relevance to almost no one since it is a rare entity. Using that logic, I had swapped out the image of anhydrous CuSO4 from the ChemBox of copper(II) sulfate for the far more common pentahydrate.  The structure of anhydrous CuSO4 is buried deeper into that article.  Again, anhydrous CuSO4 is a rarity.
 * Notice too that the two images of Ferrous sulfate in the ChemBox dont match. One is a portion of the heptahydrate lattice, and the ChemDraw is an image of the two ions that comprise the salt in aqueous solution. In this way, the ChemBox shows the structures of the most prevalent species that readers might encounter.
 * Overall I vote for showing Xray structures for the most common salt and ChemDraws for the component ions. Now we could take this approach to an extreme and show for sodium chloride both the lattice and [Na(H2O)6]+,Cl-.  However inorganic chemists dont typically show aquo complexes of monovalent ions.  Views from others are welcome of course.--Smokefoot (talk) 12:40, 4 February 2017 (UTC)
 * I understand it can be solved/improved with available options. If there is a problem with covering this, please let me know. -DePiep (talk) 08:52, 7 February 2017 (UTC)
 * I agree with Smokefoot on this - the first image should be the most common thing covered. Don't forget the NaCl·2H2O structure, but for such an important substance some of these hydrates or salt solutions could deserve their own article eg Brine. 21:34, 7 February 2017 Graeme Bartlett


 * See if I get this. The article title says "Iron(II) sulfate" unspecified/unaquatic. The infobox shows the aquatic form, because that is the common one. All fine so far.
 * Let me introduce Wikidata (WD). WD wants to be very tough on identifications & classifications (each compound form is distinct; while we still can mix list and detail them in an article). WD has item conected with this article, and item  (via ) for this more common aqua form. So the situation is: the infobox  shows a different molecule that the article title is.
 * Sure the "most common form in top [while article title says something different]" could be a MOS. The WD-enwiki interaction must cover this. Just to keep in mind, I have no solution yet. Pease continue. -DePiep (talk) 22:15, 7 February 2017 (UTC)
 * Compared to the number of chemboxes this issues should be rare (?). Could we create a hidden parameter in chembox WD item name to specify the WD item name? Biggest issue will be making sure that all the info is for the same version of the compound. I agree that we should focus on the most important/common form but we should always explain that in the lead. (BTW, what do we do about the periodates, chemboxes for those compounds all contain data for 2 different compounds) --Project Osprey (talk) 23:45, 7 February 2017 (UTC)
 * Yes to every sentence, good descriptions. (and a "?, I postpone" for the periodates). Glad I'm getting this issue. Re WD item name. I plan to add QID to exactly for this (special) situation: article title =/= main compound. I think, WD-WP issues should not determine how we at enWP organise an article (not any WD editor is objecting).
 * TL;DR: we can solve this in, and IMO it still fits the WD chemicals classification (or 'chemicals data model'). -DePiep (talk) 21:57, 8 February 2017 (UTC)
 * I wanted to thank those above for helpful advice. DePiep and Osprey, I hope that WD people/code can be trained to be more open minded or at least to not get too pushy with content adders. BTW, if you see common salts that are poorl described, drop me a note. I am going to get the structure of Graeme's NaCl·2H2O.  --Smokefoot (talk) 23:14, 8 February 2017 (UTC)
 * Currently we have CASNo, CASNo1, SMILES, SMILES1 etc, could we have QID QID1 and then link them all together so that a chembox can have several sets of data but WD can still understand what all of it is? As a second step you could set CASNo_Comment CASNo1_Comment etc etc to show the QID name as default which would straighten out a lot of our own chemboxes. That would probably require you to set each QID in two parts: a full IUPAC name for WD and a shorthand for the page (e.g. R-enantiomer, hydrochloride) so that we don't have wrapping problems. Would that be too ambitious?--Project Osprey (talk) 09:51, 9 February 2017 (UTC)
 * 1. Yes QID=, QID1=, QID2=, ... are already in there :-) :-).
 * 2. Just finished in the sandboxes: add like index2_label just once will add that prefix to each identifier2 output (no need to use all CASNo2_Comment, SMILES2_Comment, ...). This old proposal is about to go live.
 * 3. However, these changes do not interact with Wikidata yet. For two reasons: first I wanted to prepare the infobox setup (until now, there was no requirement that all index2 identifiers are about the same compound, example in case: iron(II) sulfate needs so). Second, building and testing this for a dozen identifiers better be done in steps.
 * 4. For now, I am not planning to show Wikidata data automatically (e.g. a name taken from QID2). That is because here at WT:CHEMS is a extensive talk about the quality of data at WD. Even PubChem CID + name imported from the PubChem database (!) could have errors. I will add some features to help with data cleaning, starting with tracking categories next week or so ('PubChem CID local input different from Wikidata').
 * TL;DR: Wikidata links like QID, QID1, QID2, .. will be added shortly, passive only. Params index_label, index1_label, index2_label will be added (one label for all index-n identifiers, replaces all id_Comment). No automated Wikidata showing at first, but tracking categories will follow. -DePiep (talk) 11:05, 9 February 2017 (UTC)


 * Ahead of me as always :-) . There's still the issue of the data already in chemboxes, this will need to be checked and organised to match the new numerical system. The only option I can see is manual checking. Obviously tedious, but more importantly impossible without a work-list. We would need a new hidden maintenance category category:chemboxes with 2 or more identities (or somesuch) which would list all pages with 2 or more CAS, ChemSpiderID and SMILES etc. But then we'd need a way of marking off pages that have been checked. Simplest option would be a hidden note () but I think long term a better option would be index data checked which you manually date-stamp, with those pages then being moved into a sub-cat.--Project Osprey (talk) 11:47, 9 February 2017 (UTC)
 * The proposal is on: Wikipedia_talk:Chemical_infobox. Please continue there when appropriate.
 * Quick reply here: See (500+ P). About checking & signing ok: the proof is in the edit! Once you add (R), this shows with each #2 identifier. Now it would be very bad editing not checking these actual id's per compound. But that can happen anyway. Future checking step: check against Wikidata per indexed compound. For this I want to use the QIDn, obviously. But this step is a long time ahead, because this Wikidata-check must be done for all compounds (17k infoboxes Chem and Drug). For this topic (chemical infobox & MOS etc): enough to state that we must aim to be correct is addressing multiple compounds. -DePiep (talk) 15:54, 9 February 2017 (UTC)
 * Seems good for future data comparison. Good job. Snipre (talk) 17:03, 9 February 2017 (UTC)

Addressing granularity in Bonnie and Clyde issues
In the above discussion, a common problem seems to be that a single (en)WP page can only be linked to a single WD page, and vice versa (per language), whereas the granularity of the concepts treated in the respective pages may differ. Such problems arise in all sorts of contexts, and so there is a WikiProject over at WD aimed at scouting out potential solutions to such issues, which are often referred to as Bonnie and Clyde issues (where enwp has one article, but WD has three corresponding items &mdash; one for the pair, and one for each of the two people). -- Daniel Mietchen (talk) 10:20, 3 February 2017 (UTC)
 * Your link 'problems' already describes the solution (within the Wikidata concepts and without enforcing a localWP to change say article titles). It nicely opens saying it is 'a problem for editors new to WD'.
 * As for chemicals on enwiki, allows chemical article 'Bonny and Clyde' (substance having isomers) to specify (add) 'Bonny' and 'Clyde' through indexes like PubChem/PubChem1/PubChem2/.... Coming weeks I plan to refine this rude form, mostly to match the WD concepts.
 * On the WD side, I understand the bot masters are refining issues (errors like: having two Clyde items, or Bonny and Clyde not yet split into its parts). See d:here, note the original number of violations was 659 (first section), being reduced (cleaned) now showing many single-item(QID) rows.
 * This issue I call 'data model issues'. We need to get the data model right, for each chemical, on in both WD and WP, in general: identifiers are unique, their classification being correct (whether Xor class or and/or class).
 * A totally different issue is: 'PubChem CID is wrong in WD, in WP or in both'. -DePiep (talk) 10:56, 3 February 2017 (UTC)
 * +1 The relation solution is the correct solution: WD items should be linked together and WP contributors choose the most similar concept in WD to link WP article to WD. Then they can easily access to the data of all items linked to the main item using its relations in WD. Snipre (talk) 17:57, 3 February 2017 (UTC)
 * Wikidata is incredible flexible and WP infoboxes can be programmed (with Lua) in a way so they can flexibly react to what WD is providing. E.g. if WD has several isoforms on a compound, the infobox can render this accordingly, same goes for salt forms (using isoform and has part properties). We did that with the ProteinBoxes already, pulling together gene and protein information from linked WD items. In addition, for more control, infobox parameters can still be used. Sebotic (talk) 01:36, 7 February 2017 (UTC)
 * LOL. Even without Lua, and even being crippled with this current wikitable format, we can attack the Wikidata limits hands down. We will also inject  parallels, we so eagerly kept up. Sebotic (are you a bot?), tone down from your WD throne and cooperate. -DePiep (talk) 02:33, 7 February 2017 (UTC)
 * can you clarify what you meant by "attack the Wikidata limits hands down"? Was your point that some WP pages would draw from so many WD items that it would hit some WD performance limit?  Something else?  (And I can attest that Sebotic is not a bot.) Best, Andrew Su (talk) 06:06, 7 February 2017 (UTC)
 * , I said so for LOL only to keep the spirits positive (but a bit too rough maybe? ). Sure is an old wikitable setup, with gives extra problems in options & ease, but for now I think we can still make a good WD-support (covering the data types, add tracking categories). The "attack" is about confronting tough WD-WP data checks. I think we do not require a Lua form yet (of course the templates should go there later on; however that better be kept independent of the WD connection). And  will ride along. I know about 'expensive' calls, but I do not foresee limits at the moment and I will only bother when we hit that wall. Expensive calls are allowed after all. I said "tone down" to stress that we can go a long way before we would reach any limit with the current infoboxes here. My post was a cheer up, to say we are not about to disappoint WD people :-). -DePiep (talk) 08:43, 7 February 2017 (UTC)
 * , super, thanks for the clarifications. All your points make sense to me.  I think Sebotic's point was that it is possible to handle the edge cases being discussed using Lua, not that we needed to switch to that immediately.  And I appreciate your desire to stay positive -- we are all on the same team here, but as we all know it is too easy to let personality friction (real or perceived) get in the way of that sometimes.  Best, Andrew Su (talk) 18:40, 7 February 2017 (UTC)
 * OK. And such frictions are a risk indeed. Now since English is not my first language I'd better stay out of irony & subtleties. Or iron the subtleties out. -DePiep (talk) 20:29, 7 February 2017 (UTC)
 * please accept me toning down. As I described, it was enthousiasm foremost, but in wrong words. (now allow me: so you were actually given a 'bot' name at birth? ;-) ) -DePiep (talk) 22:26, 7 February 2017 (UTC)
 * Sure, my nick predates any Wiki stuff ^^. Anyway, how would you like to proceed. The comments I made regarding Lua were just to clarify that we have a lot of flexibility on how to get data from WD, so as long as the items (e.g. isomers/salts) are linked in a consistent fashion on the WD side, they can be rendered in an infobox. What would certainly be helpful if users would cleanup the issues I posted and we reboot template drafts? Sebotic (talk) 10:00, 10 February 2017 (UTC)
 * TL;DR. How to proceed: (1) Bonnie and Clyde is about chemical classification. Postpone for now, first start: (2) At enWP categorise PubChem CID WD-WP situations. Similar to this by SoCalChemBot. Check, crosscheck PETSCAN/Venn, empty some categories, edit WD? WP?, etc. Open question: how to edit from here, so that it helps WD chemicals, and (your) WD-side? You can wait responding until I make practical proposals (adding categories in ); no need to detail answers in theory now. [end of TL;DR]
 * (1) this Bonnie & Clyde issue is about 'chemical classification' of course. I start getting the WD setup, its aim, and future infobox option. (2) The other topic is individual data points, like getting PubChem CID right. Discussion is above, so far resulting in this SoCalChemBot WD overview. This is my first goal, starting with PubChem CID: creating these lists here at enWP as categories (done smart, use PETSCAN to Venn, try catching non-templated chemicals, ...). This should be useful in improving the chemicals, both in WD and in WP, by checking. A questionis, how such edits invited from enWP can make acceptable improvements. (eg, what to do when a bad PubChem CID is found?). Also, some categories should be emptied. Unsolved IMO: how to state that a WD claim is OK? (extremis: a PubChem CID is sourced in the PubChen DB of course, so how can someone claim that the source is wrong? How do you solve that at WD?). -DePiep (talk) 14:12, 11 February 2017 (UTC)
 * So on Wikidate for chemicals should we using "has part" or "part of" for the situations like optical isomers, different hydrates (eg iron sulfate vs Szomolnokite, Rozenite, Siderotil, Ferrohexahydrite, Melanterite)? How about the relation between compound classes and their members: eg between aluminium halides and aluminium trichloride, or sulfate and potassium sulfate? Graeme Bartlett (talk) 11:55, 8 February 2017 (UTC)
 * We should use specific properties to link compounds like salt withe their acidic form, hydrates with the "dried" form.
 * For example NaCl is a salt of HC, where is a salt of is a property. Currently we don't discussed about the kind of properties and the ontology we want to use in WD. For the large classes we can use instance of/subclass of but I would prefer to use functional group decomposition of the chemical to describe chemicals instead of traditionnaly chemical families. The possibilities to use WD by machines will be lead to many application linking properties of chemicals with their structure. Snipre (talk) 20:18, 8 February 2017 (UTC)
 * For that purpose, I successfully proposed the 'has role' property a while ago. For chemical classes, we could use as many ChEBI classes as possible. Sebotic (talk) 10:04, 10 February 2017 (UTC)

Imidazolone vs Imidazolinone, advice sought
Help is welcome in figuring out the difference between imidazolone and imidazolinone. Maybe the same thing. They are oxo derivatives of dihydroimidazoles. My feeling is that we need an article guiding readers on these heterocycles.
 * imidazolone. Example: Imidazol-4-one-5-propionic acid
 * Imidazolinone. Example: Imazaquin

While we are at it and to invite corrections/suggestions: an mini-disambiguation article might also be started on imidazolidinones. 2-imidazolidinone are cyclic ureas. 4-imidazolidinone are used in organocatalysis.--Smokefoot (talk) 17:57, 11 February 2017 (UTC)


 * The two heterocycles differ by oxidation state. Also they are incompletely defined as the carbonyl group could be at position-2 or -4. The parent heterocycle (without the carbonyl group) is imidazole and adding the carbonyl back gives:
 * imidazole – aromatic → imidazolone, could be 2H-imidazol-2-one or 4H-imidazol-4-one
 * imidazoline – one double bond reduced → imidazolinone, could be 1H-imidazol-2(5H)-one or 1H-imidazol-4(5H)-one
 * imidazolidine – two double bonds reduced → imidazolidinone, could be imidazolidin-2-one or imidazolidin-4-one
 * Boghog (talk) 20:08, 11 February 2017 (UTC)


 * Names seem to originate from the Hantzsch–Widman nomenclature. My interpretation would be that either is allowed. Imidazolone is based on the suffix -ole, for any unsaturated 5 membered ring. Imidazolinone seems to be based on the suffix "-oline", which is a special suffix reserved for 5-membered nitrogen heterocycles containing one double bond, when there can be more than one non-cumulative double bond (see here). If it ends up with a page you may as well explain all that somewhere.--Project Osprey (talk) 10:11, 13 February 2017 (UTC)

Comment on wikidata transclusion from the User:ProteinBoxBot team
I've copied my reply to this discussion at WP:MED because I think it's relevant here, and hopefully will prevent confusion or misunderstanding down the road...

Hello folks, speaking for the ProteinBoxBot team here. Let me provide a bit of context and history. Since ~2007, our team has been keeping the infoboxes on pages for human genes/proteins up to date. For many years, we have been making those edits using our User:ProteinBoxBot account. Our bot synced data from many source databases, and populated infoboxes here on Wikipedia using GNF Protein box. When Wikidata transclusion became possible, we made plans to convert our bot so that gene/protein data would be written to wikidata, and the relevant content would be queried from there. This was not a trivial change. We spent months (or more) tweaking the data model on Wikidata so that our imported data would properly represented. Once we were confident in the result (and only then), we reimplemented our gene/protein template as Infobox gene -- same presentation, but using data drawn from wikidata. This conversion was done in stages (completed ~ May 2016), and everything was announced and coordinated with WP:MCB.

In addition to adding gene/protein data to Wikidata, we are also interested in adding content related to human diseases and drugs. (This data has value for us even outside of its potential use on WP.) We have been importing data from commonly used sources in the biomedical research community. This process is still ongoing, and as for gene/protein data, it's also not a trivial process. The data modeling and the bot coding are both continually getting better, and we always appreciate input from both the WD and WP community on how to improve things. However, we have never made any claims about being fully confident enough for them to appear in templates like Infobox medical condition or drugbox. If we ever did want to propose using wikidata as a resource, we would certain to coordinate with WP:MED and/or WP:CHEMS. Bottom line, we fully support the important oversight role that the wikiprojects play, and fully support applying a very critical eye to transcluding wikidata content. We take no offense if the verdict is that wikidata content is not mature enough yet, but we also ask not to be "blamed" anytime people outside of our team use wikidata content without consulting us or the relevant wikiproject. Cheers, Andrew Su (talk) 18:26, 24 January 2017 (UTC)
 * Does this work include Chembox, which intentionally overlaps Infobox drug (esp wrt identifiers and chemical/physical data)? -DePiep (talk) 18:37, 24 January 2017 (UTC)
 * Hi the chemical data that we are loading into wikidata certainly overlaps the fields that are contained in Chembox as well.  But again, we haven't taken an active role in changing any template to use Wikidata content (with the exception of infobox gene).  (I'm pretty sure you know this from discussions over at WT:CHEM, but just to be sure!)  If your question is whether I think caution should be exercised when transcluding content from wikidata here on WP and chembox, I think the answer is generally yes.  If you let us know which fields you are interested in transcluding, we can comment on how stable things are (data model, data sources, and update cycle).  Sound reasonable?  Cheers, Andrew Su (talk) 00:38, 26 January 2017 (UTC)
 * Where is this Project being talked? For now, it feels like enwiki template development is being frozen into 'don't touch it, just wait'. You'll get this. (I have a serious data model for both and  by chemical compound. How to check with this bot team?). -DePiep (talk) 01:25, 26 January 2017 (UTC)
 * We do most of our Wikidata discussions at WD:MB, and you are welcome to check in there about any issue in general or any data you want to transclude in specific. (Or even participate in our data loading/modeling, of course.)  I think that would be a good first step before considering transcluding into well-established WP infoboxes.  I hear and understand your frustration with the cautionary message.  My concern of course is that we not move so quickly that the WD data is not properly vetted by WP standards, which I think would lead to long-term distrust.  This is how I'm thinking about it, but of course would welcome your thoughts as well.  Cheers, Andrew Su (talk) 06:51, 26 January 2017 (UTC)
 * I remain to have serious concerns about this. I have since we were talking about using WikiData for identifiers earlier run into other data that is being compared to WikiData.  For one template (I would call it data 'of general public interest'), that concerns about 3000 datapoints, of which more than half are categorised as 'not the same as on WikiData'.  That would suggest to me that at about 750 of those are wrong there, but that then also goes for the smaller half (the ones that are the same) .. it would mean that about half of them are plainly wrong.  And this is data 'of general interest', not some specific database identifier in a corner of science.  I would not expect the former to go wrong this much, on the other hand, database ids can be compared from official database files (we can download the whole of drugbank), which may give less problems.
 * Before we start transcluding, I think it is good to see numbers for 'how many we have here and they don't have there', 'how many they have there that we don't have here', 'how many have a different here and there', and 'how many are the same here and there'. The second category should not contain more than a couple of % of the total number .. otherwise transcluding from WikiData will make our situation significantly worse than what we have now (I agree, from the previous example, we may have 1500 wrong here as well - though for some identifiers I trust our data more as we actively looked at it, and if it is here, people will actually repair it because they don't have to figure out where to do it there - displaying becomes increasingly ugly if we have to include edit links to external wikis).  --Dirk Beetstra T  C 03:55, 26 January 2017 (UTC)
 * Great point. User:Sebotic started exactly such an analysis on PubChem IDs (I think discussed somewhere in the archives).  We got distracted by other projects, but I think he will have it complete soon and will report results at WT:CHEM.  Cheers, Andrew Su (talk) 06:51, 26 January 2017 (UTC)
 * I saw the PubChem one as well, there are 36 where Wikipedia is different from WikiData, meaning that at least one of the two is wrong - there are 36 mistakes together (out of 195 total transclusions).  There are more cases  where they are both the same, or where WikiData does not have the data, but by extension, some of those must be wrong on both ends (and if they were copied from one place to another .. ).  I have other of those statistics on my radar where the number of cases with different data is 20.600 (and where other, missing data is automatically pulled from WikiData ... and thát may be a BLP concern).  --Dirk Beetstra T  C 07:46, 26 January 2017 (UTC)
 * As Beetstra says. Last December I had in the sandbox for a setup that would categorize two pilot checks: A. articles with/out a QID (+link to WD item page inviting to edit). B. comparing & analyse (pilot datapoint) CAS Registry Number local input vs. Wikidata value. I stopped it because I get from Andrew Su's posts that we'd better wait nutil the bots have made the WD datra perfect or so. In short: I still have no clue on how the (respected) Bot and editors are supposed to improve this WD data & usage. (I think I am in line with Beetstra's remark; tell me if I mistake). -DePiep (talk) 10:11, 26 January 2017 (UTC)
 * - how much coding would it take to categorise CASNo on en.wikipedia into 'en=value;wd=empty','en=empty;wd=value','en=wd','en!=wd' (the rest of the transclusions of the chembox are then obviously empty) for the 'main CAS number' in the chemboxes? --Dirk Beetstra T  C 10:56, 26 January 2017 (UTC)
 * And you are well in line with me :-) .. I want to add that I get a feeling that there is a huge push from WD along the lines of 'we took a lot of data from you guys, and we worked hard on it, it is now time that you use it back ... --Dirk Beetstra T  C 10:58, 26 January 2017 (UTC)
 * Not much coding, I was working on that in sandboxes. I also wanted to get these categories right from the start (like cat naming pattern, useful subdivisions). Then there is this topic we better discuss too: the data model for a Chemical page (~= having or ). In short, article could be: chemical substance (e.g. a group), chemical compound (what we think of first), with multiple chem compounds (by indexes CASNo1), vaccine/mab (drugs only). One cover-all parameter for this (to improve specific categising)? Because: when we have an article without CAS number, the maintenance task depends:  CAS number is missing, not assigned, or doesn't exist because it's a chemical group etc. Also, I'd like to categorise articles by "having a QID" or not (no QID=not in WD!).
 * Those maintenance categories are a huge invitation to edit WD (or the en:article) of course. It's just, I keep getting mixed signals from on whether we should start this route here on enwiki. Sounds like we would interfere with the bot process. A simple answer possible, Andrew Su?
 * re 2nd part: Good! -DePiep (talk) 14:45, 26 January 2017 (UTC)
 * IOW Beetstra, I'm ready to dive into the sandboxes again, and make formal talking points/proposals for this. And indeed, categories as you mention. -DePiep (talk) 14:48, 26 January 2017 (UTC)
 * Thanks to you both. I actually wasn't aware of those categories that highlight differences between WP and WD -- good idea.  Once Sebotic updates his analysis script and results, we'll report here and figure out together the best path moving forward.  Regarding 's comment about editing WD, my recommendation would be to go ahead and change WP if WD is right.  If WP is right and WD is wrong, go ahead and change WD if you like, but most important would be to pass along those examples to our team.  Ultimately we want to make sure our upload and synchronization bots on WD are perfect relative to the sources, so we need to examine the edge cases to figure out where things are going wrong. Cheers, Andrew Su (talk) 19:20, 26 January 2017 (UTC)
 * This I can work with. Expect proposals as mentioned by me, in Februari. One detail: when changing WD it might be difficult to 'pass along those examples to our team' for sure, because that would be a manual WD edit and so require discipline (or a smart history filter?). Will de talked again later on. -DePiep (talk) 19:37, 26 January 2017 (UTC)
 * Contribution by the WP community need to be well referenced (database it came from + database ID, as it is in the WD ref guidelines), then, our bots will keep these contributions. Sebotic (talk) 20:02, 26 January 2017 (UTC)
 * As Andrew said, I will re-run/complete the analysis in the coming days. Just to give you a little context, I have been importing ~140K chemical compounds into Wikidata in recent months and I have also been strongly improving and expanding the existing ~16K compounds in WD, discussion here. This all happened with automatic imports from primary resources like PubChem, ChEBI, FDA UNII, Guide to Pharmacology, etc. So the data on these chem compound WD items, touched either by SoCalChemBot or ProteinBoxBot, should be in very good shape, As far as I could see, if there is a disagreement between WD and WP on these items, WD is usually right. There is however, one issue: I identified ~1.2K compounds which are a mess because people put alls kinds of IDs on it, making it inconsistent to a degree where these problems cannot be programmatically resolved and need human intervention. I posted this list, so WD users could fix it (fixed 250 by hand, painful), but it got moved to archives before being fully resolved. The major problem I see is that people constantly get stereochemistry and salt forms wrong. We can resolve that with Wikidata and pull the salt forms and different stereoisomere info into WP infoboxes, if we want to. For that purpose, I already created a stereoisomer property for that purpose. Happy to discuss, will come back asap and post my WD to WP comparison. Sebotic (talk) 20:02, 26 January 2017 (UTC)
 * This is all good news. I think here we could start out with the "show local WP value over WD" (for pilot CAS number), and categorise. The &Delta; articles ask for attention. Let's see what the numbers are. -DePiep (talk) 20:27, 26 January 2017 (UTC)

About WD and WP data quality (1)

 * My two cents. The major problem is not to know who is right or wrong but to define correctly the compounds. I just take one example from : Vernolic acid. CAS number from WP article: 503-07-1, CAS number from associated WD item: 32381-42-3. Who is wrong ? No one. Value in WP article is for cis-(+)-vernolic acid and value in WD is for cis-(-)-vernolic acid. From my opinion, the problem is coming from WP articles which often mix data for mixture of stereoisomer and for the different stereoisomers. WP articles should be more acurate and provide data for all compounds (stereoisomers and mixture) or to choose only one (mixture or one stereoisomer). Why do we have in Vernolic acid article always the identifiers for both stereoisomers (ChemSpider, PubChem, InChI,...) and only one value in the case of CAS number ?
 * Looking at that example, I can already say that all statistical analysis are just bullshit because based on messy data. WD needs to be improved but WD is better organized with stronger policies than in WP.
 * Thanks for the work done about data importations in WD but please before any new large data import perform a preliminary check of the data from the different databases: put all data one table for each chemical (see below an eample) and perfom some matching analysis in order to detect potential problems and to import in WD only data having a high matching rank.


 * What is missing is a powerful comparison tool providing some confidence grade about faithful value: if most of the databases have the same value, then the value has a high pobability to be correct and can be imported, if there are several different values then the confidence is low and the data should not be imported. Snipre (talk) 21:03, 26 January 2017 (UTC)
 * Correct that none of the articles WD/EN was wrong. But en:Vernolic acid was missing one of its two enantiomers CAS number (I have just added it), and CAS for the racemate too (added too). WD has one article for each enantiomer (but for this compound only for the (-)-isomer - the two others are N/A). We can only link to one compound at WD so the editor (robot) has to make a choice. When a chemical compound have two notable enantiomers we describe them in the same article. All optical active "chemistry"-chemical compounds - both enantiomers are notable. For drugs, most optical active compounds - only one of the two enantiomers are notable, and the other shouldn't be mentioned in the infobox. Christian75 (talk) 21:48, 26 January 2017 (UTC)
 * I think this can be solved by linking to the racemate WD item if all enantiomeres are relevant and should appear in the chem infobox or just to the specifiic enatiomer if only a specific one is relevant/notable, e.g. for a drug. Sebotic (talk) 22:14, 26 January 2017 (UTC)
 * Below I wrote about this (sort of ec). Also an idea to get all three values from WD. -DePiep (talk) 22:26, 26 January 2017 (UTC)
 * I object to the suggestion to leave out compounds because we could only retrieve one value from WD (which is not true btw). Such a perceived limitation of WD should not determine a thing like infobox content. -DePiep (talk) 22:39, 26 January 2017 (UTC)


 * What is this??? Above, : Bottom line, we fully support the important oversight role that the wikiprojects play, and fully support applying a very critical eye to transcluding wikidata content. But four months ago: this. Why was this not discussed/announced at Template talk:Chembox? Why did no one here even mention you WD people were already working with this while I was just proposing it? -DePiep (talk) 21:26, 26 January 2017 (UTC)
 * That edit was not made by a member of our team. (For reference, the list of our team members can be found here.)  Does that address your questions?  If not, can you restate your working with this question -- I'm not understanding...  Cheers, Andrew Su (talk) 21:36, 26 January 2017 (UTC)
 * ''(Struck my post). I made a mistake. My excuses to for wrongfooting you. I totally misunderstood the situation. -DePiep (talk) 21:42, 26 January 2017 (UTC)
 * No problem, lots of moving parts here. Your comment does prompt me to update User:ProteinBoxBot here on WP, just to explicitly identify our team members (which I will do momentarily)... Cheers, Andrew Su (talk) 21:45, 26 January 2017 (UTC)
 * (ec) Let me restart. Above, mentions . This is populated by PubChem, a standalone template used inline, not in/through the infobox .  has 195 transclusions only. (My mistake was: I misread it for, which is the subtemplate that handles PubChem ID values in  8500&times; ;-) ).
 * More on the Vernolic acid example Snipre mentions. It has PubChem=6449780 in the body text, and 5281128, 6449780 in infobox.
 * Also, Infobox mentions all three CAS numbers well-noted, using parameters CASNo, CASNo2, CASNo3 etc.:
 * 503-07-1 (+)-isomer; 32381-42-3 (−)-isomer; 58800-58-7 (Racemate)
 * So, it looks like in the infobox (not the bodytext) the issue is handled completely and correctly. (is my reply to Snipre). Of course, the article body text should be written correct manually.
 * Next issue: if we want to automate WD reading, in Vernolic acid we must ask for three CAS number values. My thoughts were to use article title/item for default Property reading (as CASNo). Then manually add QID2, QID3} for the other two compounds. Voila, the infobox can read & check & use WD. There are some 500 articles with multiple compounds. -DePiep (talk) 22:17, 26 January 2017 (UTC)
 * Note: Christian75 has completed the three CAS numbers in a moment ago, so is not what Snipre first saw. See Christian75's post above. -DePiep (talk) 22:26, 26 January 2017 (UTC)
 * As a general comment, I think we will eventually want to roll things out in phases, and I think that we should start with the absolutely simplest cases we can find -- single compounds, unambiguous and concordant mappings between resources, etc. Even with those cases, we're going to find edge cases and problems.  Once we get those simple compounds dealt with, then we can move on to increasingly complex examples.  As mentioned above, for the gene/protein infoboxes, that meant that we created a brand new template with the exact same styling and content as the original.  We then prototyped on only one infobox that queried wikidata.  Once that looked good, then we converted 10 more, then 20, then 50, then 100, etc.  At each step, we did lots of manual sanity checking.  Anyway, I think we should do something similar here, and in that case I think we can tackle vernolic acid a bit later in the process.  My two cents... Cheers, Andrew Su (talk) 22:54, 26 January 2017 (UTC)
 * Now this is across the distinction of "getting WD right" and "changing the templates". What's wrong with getting the en:article datamodel right first? And would you oppose an other route of introduction (already described a few times)? I get the impression that you've not digested the proposals Beetstra mentioned. -DePiep (talk) 07:48, 27 January 2017 (UTC)
 * In WD, the structure for this case is the next one:
 * mixture of stereoisomers (cis vernolic acid): one WD item ([Item]) defines as subclass of chemical compound
 * (-)-isomer (cis-(-)-vernolic acid): one WD item ([Item]) defines as instance of cis vernolic acid
 * (+)-isomer (cis-(+)-vernolic acid): one WD item ([Item]) defines as instance of cis vernolic acid
 * My proposition: link WP article with the isomers mixture item and then use lua to retrieve data from isomer items using the instance of relation. Lua language allows you to retrieve data from different items. You just have to know the QID of the different items. Snipre (talk) 08:12, 27 January 2017 (UTC)
 * Is what I described. Would we require to have automated adding "this is an isomer of"? -DePiep (talk) 08:17, 27 January 2017 (UTC)
 * We can create new relations(properties in WD) but this creates always more work to keep everything in the good format. The best is to infer the relation by using the SMILES properties: if A and B have the same canonical SMILES, and B has a value for isomeric SMILES, then B is an isomer of A. Snipre (talk) 10:56, 27 January 2017 (UTC)
 * OK. Not an issue in near future. -DePiep (talk) 11:12, 27 January 2017 (UTC)
 * I have some problems with your category :
 * sodium oxide: Puchem CID in WP: 73971. Puchem CID in WD: 73971
 * Hexazine: Puchem CID in WP: 11966278. Puchem CID in WD: 11966278
 * Isoleucine: Puchem CID in WP: 791. Puchem CID in WD: 6306. But WD item is referring to L-isoleucine and WP article is referring to DL-isoleucine with a chembox called L-Isoleucine, a picture representing L-Isoleucine, the chembl ID for zwitterion form of L-Isoleucine, the Chemspider ID for the L-Isoleucine. Little confusing and perhaps a cleaning is necessary. That's your stuff, you are free to mix data as you want but please don't try to do comparison between WD and WP as long as the data are not comparable. Snipre (talk) 16:57, 27 January 2017 (UTC)
 * re : problems with your category: Please read my posts. It is not "my" category at all. -DePiep (talk) 23:40, 27 January 2017 (UTC)
 * Sorry for the your category but this doesn't change the content of my comment: after analyzing 4 chemicals, two WP articles showed wrong or incomplete data in the chembox. Does it means that 50% of WP articles are wrong ? No and I hope that anyone who will perform an assessment of the data quality in WD will keep in mind the fact that his dataset can be the source of the errors. CU next time when you will finish your data comparison. Snipre (talk) 00:44, 28 January 2017 (UTC)
 * TL;DR: 1. Yes. 2. That category is populated erroneously. Useless. 3. We should categorise in infobox.
 * 1. You're right, and I saw that pitfall too. I am sandboxing this check for CAS registry number first. Do you want it for PubChem CID asap too? -DePiep (talk) 12:35, 28 January 2017 (UTC)
 * 2. Forget that category! Do not use it, don't waste time. It is populated by the inline template PubChem, not by infobox . Inline has a systematical error (so should be changed). That inline template always checks against property P662 of the article/item (default QID), even when it is about a different substance. In your example sodium oxide (PubChem CID=73971 ✅):
 * &rarr; (false positive, wrong categorisation).
 * 3. Proposed is: let and  do the categorising (would be in ~18k articles), be smart when 'article/item is not a chemical compound' (e.g. is a mixture, is a chemical group, ...), allow for multiple substances in an article (by using indexes like PubChem2 and QID2). Then the categories are useful (PubChem values to be checked). -DePiep (talk) 12:35, 28 January 2017 (UTC)


 * I have improved the  made. No content change (title added). -DePiep (talk) 08:47, 2 February 2017 (UTC)
 * Question: in the table, how are WD and WP adding to the quality level? They are not independent sources right (example in case: where WD=xxx, where does that come from?). OTOH, somehow WD or WP could be adding to the quality when they have been edited carefully (individual research). -DePiep (talk) 08:51, 2 February 2017 (UTC)
 * Another quality measurement would be like: "Overall, PubChem DB has less errors than CHEMBL DB. Therefor the Data Quality value of PubChem DB should be weighed heavier." (The table would have an extra column with some weighing factor for that DB). -DePiep (talk) 08:55, 2 February 2017 (UTC)
 * I don't think we should use this system as quality measurement but only to check if WD item and WD article are correctly matched based on their set of identifiers. If we assume that both sets of identifiers have some errors, we have to use some confidence interval to define if the link is correct or not. Then later once we have set up on both sides (WD and WP) the definition of each item/article (racemate, cis, trans, R, S,...) then we can use this system to analyze data from a larger number of databases and check what is the most correct value for each identifier. Snipre (talk) 17:40, 3 February 2017 (UTC)


 * Wait wait. Why cannot we claim that the 'own data' in an external database is correct? Is the PubChem CID in a PubChem DB not a 99.99% guarantee, but not the CHEMBL ID in PubChem DB. -DePiep (talk) 11:55, 3 February 2017 (UTC)

About WD and WP data quality (2)
I fully agree that these statistics as such do not say a lot - it would need to be done properly. But if I see, consistently through multiple identifiers, 10-20% discrepancy I will not believe that the fault is on one side completely. For these identifiers you may be more correct as they come out of downloadable databases, but that still needs interpretation on both sides. D-, L-, and rac have different pubchems and CAS, sure, but on both sides you need a human editor to check whether the assignment is correct. And the discrepancy rate is similar, or even higher than on identifiers that do not come out of downloadable databases.. I have done lots of these on en.wikipedia as part of the CheMoBot effort, and WD faces the same problem as here.

What I suggest is to categorise, make a representative subset of the ones which have discrepancies, and count individual errors. That gives a better idea. If the error rate is really low on WD side then transclusion can be implemented, otherwise you transclude wrong data that (especially new or one-off) editors have troubles to repair. Until that has been shown, I oppose the transclusion (of any WD data). --Dirk Beetstra T C 08:41, 27 January 2017 (UTC)
 * Beetstra This is already ongoing on WD side by using the constraint reports which provides lists of multiple uses of identifiers which should be unique. Each time I curate one constraint violation I clean the item in order to split data in the correct item (but I don't check if the interwikis are correctly matched with the item). The problem is when the reference databases are wrong. I contact them in order to spot the problem and with some databases I receive a feedback but for some of them no answer. I miss a reference database providing the CAS number and the correct structure including cis/trans and chiral configuration (InChI, SMILES, IcChIKey) : the best would be •SciFinder but I only have Reaxys.
 * I don't think data quality in WD side is really better but WD offers more facilities to work on data: check unique use of one identifier, check of correctness of CAS numbers using check digit to prevent typing error, data extraction for data analysis and comparison with external dataset,... Both WP and WD data need tools and processes to improve data quality and protect data against vandalism. Snipre (talk) 10:47, 27 January 2017 (UTC)
 * That is close to the point that I want to make. I have done the work here on Wikipedia, and see the conflicts throughout.  Multiple CAS numbers for seemingly the same species, opposite enantiomers seeing other identifiers from different sites (and impossible to verify easily from original sites), etc. etc.  Now we have WD, and I expect WD to have the same problem.  Then there is the additional problem regarding linking up the correct Wikipedia item with the correct WD item.  I fully appreciate the work there, I think it is a good thing to do, but the data has to be of extremely high quality before we should even consider to use it here.
 * Now, we can, for all items where we do not have the identifier here, blindly take over the value from WD. All fine.  But then we get, in some cases, thousands of translusions of wrong data here.  Editors here may see where the error lies, but when editing locally do not see the value that they want to repair.  So every value then needs an 'edit on WD link' (ugh .. that is ugly, already for the few cases that do that, let alone on 12 or so identifiers in the chembox!!!)?  Or we need to have in every article a remark 'if you don't see it here, try to find it on WD' .. what do we tell a local editor who wants to repair it.
 * At the time of the above poll, 36 out of 195 were mismatched. That has several reasons, but several seem to be due to .. mismatching compounds here and there.  Still, that is 18%.  Lets say that only half of them are wrong on WikiData, so that is 9% is wrong.  On 20.000 chemboxes that is 1800 pages that would transclude a wrong number.
 * So I guess the question is .. if we take a representative subset of WD data, how many are wrong there? Is that number, at this time, sufficiently small that it will only happen in a few cases that the editor runs into problems?  Is that number large and do we choose to have that ugly (and rather unclear) 'edit this on WD' link on every page?  Or do we wait with transclusion until the number is small enough?  And yes, it makes a difference whether it is wrong there or wrong here - if it is wrong here it is easy to repair here, but if it is wrong here because it is wrong there, it cannot be repaired here.  --Dirk Beetstra T  C 13:00, 27 January 2017 (UTC)
 * I think we agree together about the fact that use of WD data in WP is not a priority and this is not my main objective. I just point the fact that most differences between WD and WP are the result of 1) bad interwikis (WP articles are not linked to correct items) and of 2) WP articles mixing data about different isomers in the same article. Again I don't want to say you doing bad stuff, but when I read that you want to do data comparison, I say: What are you comparing ? If you are not comparison exactly the same type of data, better don't do any comparison because you will have differences and at the end the conclusion will be wrong.
 * My proposition is to work on correcting interwikis based on a clear policy from WP:en side:
 * - for chemboxes mixing data for different isomers and racemate, provide a clear way to identify to which compound the data is relevant for
 * - for article about one specific isomer, don't add any data about other isomers or racemate in the chembox. Add the data in the text or a table to clearly indicate the specificity of the article.
 * So for the case of one compound with one chiral atom or a double bond we should have:
 * - always 3 values for each data entry of the chembox (racemate and both isomers). PS: for me not applicable or unknown is a value.
 * - or always one value for each entry (one isomer only)
 * As long as you don't have a strong data structure, as long as you have sometimes one value, sometimes 2 and sometimes 3, don't do any comparison because you have not guarantee to compare comparable data. Snipre (talk) 16:34, 27 January 2017 (UTC)


 * It is however part of the initial post of this thread .. "so that gene/protein data would be written to wikidata, and the relevant content would be queried from there .... we reimplemented our gene/protein template as -- same presentation, but using data drawn from wikidata".  I agree with you, User:Snipre that we should be working on connecting the compounds correctly, know what is here and what is there, but even after that, I first want to know how many datapoints are wrong.  We have thousands of 'simple' compounds in en.wikipedia, with methane and carbon dioxide we do not have any problems with misalignment of the CAS, there can only be one (yes, these two high-profile ones are probably correct).  A right-handed Base Ball Bear does not have a different Twitter than a left-handed Base Ball Bear (on this article, the problem is something different and totally unrelated to the issue, Rob Brantly is a case).  Still we have more than 1600 cases where BLP's have twitters are different between en.wikipedia and WD (many due to other issues, though).
 * My main problem in my CheMoBot work were the enantiomers. We can link correctly whatever we want, different databases have sometimes different assignments.  En.wikipedia is, almost by definition, worse.  Having the data in two places hence is going to make it even worse.
 * Again, I would suggest first to categorise and get some numbers, in the meantime we can work on improving correcting the links (which should change ratios in the suggested categories) .. and improve in data. Still, as now we do not know how many (real) errors there are (let alone how many we wrongly assign at the moment) we should not even remotely consider to start transluding them (even if WD has a value and en.wikipedia does not have a value).  I understand that people are eager to transclude from WD (as I said, they have the data, they want something to be done with it), but I have seen resistance already.  You don't want this to become another fiasco as how WMF had problems with 'improvements' they push(ed) through (mediaviewer to name one).  I would really like to use the data, but if we get a major RfC about suggestions to transclude WD data (and I have heard talk about a major WD RfC, not sure of the topic), I would currently !vote strongly against transclusion.  And then it may become difficult to turn the community around.  --Dirk Beetstra T  C 03:55, 29 January 2017 (UTC)
 * About the practical side, seen from . Thanks you both. I think I follow, and can help in you both's concerns without enforcing anything. I will finalise the sandboxes into full proposals, they are sleeping since December (after this week I have time to discuss and edit). Mainly: add tracking categories re WP-WD diff, smartly diffused, maybe add parameter datatype to add data info for non-single-compound articles/items, address model issues like isomers, all hidden, initially do not change shown values, add this for both and  that's 10.6k + 7.4k = 18k&times; (so far it is not relevant for a compound which infobox we choose to use). Glimpses as of December are here and here (pls do not discuss now).
 * This will give the rough numbers and maintenance jobs. This is an extra tool to check data quality between the sites, without pushing any WD data into readers' sight. Meanwhile, WD-people & -bots can continue to do their research undisturbed.
 * Just a quick question: so far I am working on CAS registry number as a pilot. Would it be better to start with PubChem CID? -DePiep (talk) 08:42, 29 January 2017 (UTC)
 * That is an interesting question .. I would suggest PubChem/DrugBank, etc. over CAS, as CAS does not have a downloadable database which could be uploaded into WD - everything there needs to be done by hand (taking into account their usage policies). A lot of CAS is hence also indirect (identify drugbank to compound, take the CAS that drugbank mentions).  Maybe someone on WD side can tell us which identifier they think is 'most correct'.  --Dirk Beetstra T  C 08:55, 29 January 2017 (UTC)
 * From my experience the best identifier is InChIKey: it is independent of any database and it is easily generated from the chemical structure. The advantages are we can identify racemate and stereoismers or cis/trans compounds and we can associate racemate to their stereoisomers.
 * 2-butanol: BTANRVKWQNVYAZ-UHFFFAOYSA-N
 * (S)-2-butanol: BTANRVKWQNVYAZ-BYPYZUCNSA-N
 * (R)-2-butanol: BTANRVKWQNVYAZ-SCSAIBSYSA-N
 * Not defined stereoisomers or cis/trans compounds are recognized by the second serie of letters (UHFFFAOYSA). Most of databases generate InChIKey so this a good candidate as matching criterion.
 * Compounds having the same first serie of letters but divergent second serie can be defined as steroisomers.
 * PubChem is good but no so rigourous as many contributors to this database don't respect stereoisomers differences especially when naming the compounds or associating PubChem compound with external identifiers. Snipre (talk) 17:29, 3 February 2017 (UTC)

I now performed the comparison of PubChem CIDs in WD and WP, you can have a look at it here. Also, I randomly selected 10 compounds where CIDs do not match, in every case WD did a better job compared to the CID in the WP infobox. WP had outdated CIDs, wrong salt form, wrong composition, wrong stereoisomer etc. To me, currently WD does a better job, but certainly, still some cleanup required. We could proceed with transcluding PubChem CIDS for a start. Sebotic (talk) 00:46, 1 February 2017 (UTC)
 * That was also generally my starting point. Matching those with external (primary) databases should go correct to a high level (primary database is 'PubChem database for PubChem', 'ChemSpider API for ChemSpider', NOT 'ChemSpider API for CAS').  Our templates could be written at first like that: match local InCHI with WD InChI and take the other WD data according to that link.  That could then also be our starting point here: make the maintenance categories I suggest for InChI, making the link as perfect as possible (low discrepancy percentage, random checks to have less than an agreed percentage of real errors (which would mean repairs on each side where needed, I am sure we have wrongly assigned InChI's, they are not human understandable and vandalism prone).  From there we also make maintenance categories for other data based on the en.wiki InChI - WD InChI link.  And then take it from there.  (Since the CAS db, e.g.  is very incomplete and does not have listed InChIs those are not easy to properly check, we'll take such cases last).  --Dirk Beetstra T  C 05:24, 4 February 2017 (UTC)
 * A brilliant investigation! Thanks! I get the 'broader checks are to be done on the WD-side' (leaving me jobless here at enwiki ;-) ). Will possibly ask later about: 'Most errors are due to mix-ups of different stereoisomers or salt forms', or other manual edits that could be usefully done here at WP. Also interested in how would like to proceed. -DePiep (talk) 08:55, 1 February 2017 (UTC)
 * I would expect a rather good job on WikiData, but a hitset of 0 does not exactly show how many errors there really are on WD (and this would indeed suggest that we could transclude PubChem). 6% mismatch is on the low end for the whole set, a better job than on other (non chem) identifiers - are there typical reasons (wrong template calls), or are it real errors?  --Dirk Beetstra T  C 10:39, 1 February 2017 (UTC)
 * Thats not the picture I get when checking the salts. E.g. I checked the indium salts. E.g. InF3 pubchem wrong. Some of them they are linking to what looks like the same compound but have different pubchem number, e.g. Indium(III) selenide WD: pubchem, WD: pubchem (the first 2D image looks insane, but its probably autogenerated).  Is possible to get a list of chembox articles only? Christian75 (talk) 11:31, 1 February 2017 (UTC)
 * That is more like what I got when I was working this myself. It is darn difficult get a right link in some cases.  What my main concern is, that if you are collecting PubChems for compounds, and list all of them in a db, then 'the right one is obviously there'.  That is the case for WD.  Whether that, on WD, is always correct is unlikely to me, just as that it is unlikely to me that Wikipedia has it always right.  Mixing two databases with each a certain error rate will result in a larger overall error.  That is in part defined by the 6% mismatch (which is, from my experience, rather small, I mismatches generally larger than 10%).
 * The problem that we always had here on en.wikipedia was that you have to have a first identifier done to very, very high precision. After that, you can match the rest rather easy (except for the mismatches between different databases - they all have their own error rate).  I am somewhat impressed by '0 out of 10' .. but that can still be 5 out of 100, with a real percentage of 10% (if your number 11 is wrong, you go from 0% to 9% errors).  If you extrapolate from that to the next identifier, you will see that the second one will have higher error rates .. etc. etc.  --Dirk Beetstra T  C 11:41, 1 February 2017 (UTC)

About WD and WP data quality (3)
I updated our analysis/comparison of Wikidata and Wikipedia PubChem CID}. Now, each item/article which needs curation is listed. Best, Sebotic (talk) 01:28, 7 February 2017 (UTC)
 * Thanks, . Now who is to edit what? -DePiep (talk) 22:42, 11 February 2017 (UTC)
 * I take care about group C chemicals. Someone from WP:en should take care of group D. For group A and B, we can see after. Snipre (talk) 09:51, 13 February 2017 (UTC)
 * OK, will work towards that. Do you think the two infoboxes could use a tracking parameter like chemclass or wdIssue manually, to help eventually cure the Groups? -DePiep (talk) 10:25, 13 February 2017 (UTC)

2-Isopropyl-5-methyl-1-(2,6-dihydroxy-4-nonylphenyl)cyclohex-1-ene
Does this compound have a CAS RN? --Leyo 01:13, 7 January 2017 (UTC)
 * Not that I can find. Doing a similarity search gave an isomer with Me's added to alpha and beta carbons of the tail, and two fewer carbons in length.  It has a paper and a patent, referring to cannabinoids. The propyl-tailed derivative gets 130 hits, again cannabinoid literature.--Smokefoot (talk) 01:25, 7 January 2017 (UTC)
 * Thanks. Did you do the search using SciFinder? --Leyo 20:01, 7 January 2017 (UTC)
 * I checked SciFinder; no exact hits, and the one Smokefoot notes--an isomeric nonyl sidechain--is the closest by similarity to it. No library near me has a copy of The Total Synthesis of Natural Products. (cited ref 1). I won't have access to the TetLett article (by the same author, cited ref 2) until Monday to see what it says, but I did a SciFinder search for all compounds in that article. It found several structures having that same isononyl sidechain but no other major alkyl-sidechain analog. I wonder if someone somewhere dropped a leading "isomer of" or "iso" on the "nonyl" name? DMacks (talk) 20:39, 7 January 2017 (UTC)
 * Meodipt created that article and should thus be in the position to answer that question. --Leyo 19:28, 17 January 2017 (UTC)
 * I will check in my copy of that article from Total Synthesis of Natural Products to confirm when I get a chance but pretty sure that structure matches what is in there. However it could well be a mistake in the review article, as dimethylheptyl would make more sense than n-nonyl in terms of substituents here. Don't think I ever double checked it against the original Tetrahedron Letters paper as I didn't have access either. Meodipt (talk) 02:11, 20 January 2017 (UTC)
 * Ok I have checked the original review article and the mistake was mine - Razdan drew the core structure but did not draw out the side chain, it just says C9H19 as the R substituent, without specifying what this is, and I mistakenly assumed this meant n-nonyl. However elsewhere in the article he explains that unless otherwise specified, C9H19 refers to 1,2-dimethylheptyl. So you guys are correct, this n-nonyl derivative of cannabidiol does not exist (though both the 1,2-dimethylheptyl and 1,1-dimethylheptyl isomers are known), and the page should be deleted. However too many people have subsequently edited the page to just add a db-author tag to the top so I'm not sure what the best process to follow is here. As it happens there has been quite a bit more research into the 1,1-dimethylheptyl derivative of cannabidiol, which is known as CBD-DMH, and I may eventually get around to making a page for that one, which could also mention the 1,2-dimethylheptyl isomer. This n-nonyl isomer though does not appear to have been made, and probably doesn't share the same activity profile as the dimethylheptyl versions either, so best to just delete this page and start afresh I think. Meodipt (talk) 21:11, 27 January 2017 (UTC)
 * Thank you for your investigation. I fixed your DR: Articles for deletion/2-Isopropyl-5-methyl-1-(2,6-dihydroxy-4-nonylphenyl)cyclohex-1-ene --Leyo 23:14, 27 January 2017 (UTC)


 * BTW: This unsourced addition of a CAS RN by a new user might be checked. --Leyo 20:25, 7 January 2017 (UTC)
 * Structure found by SciFinder for that CAS looks like the same. DMacks (talk) 20:41, 7 January 2017 (UTC)
 * Thanks. --Leyo 23:58, 10 January 2017 (UTC)


 * I has translate this article some years ago in the deWP. Same question as here in the deWP. As a result I take a look into the tetrahedron letters paper cited and there was given as a structure for the questionable sidechain: C9H19 = CH(CH3)CH(CH3)-n-C5H11 - the 1,2-dimethylheptyl-derivativ. In my opinion the title is wrong but the article isn't it. --Codc (talk) 13:27, 29 January 2017 (UTC)


 * On a similar note, I noticed when checking information for CBD-DMH that the page for Abnormal cannabidiol was also incorrect, it showed the cis isomer when it should be trans like natural cannabidiol. I have changed it to show the correct enantiomer and added the CAS number that is listed by suppliers of this compound like Tocris and Caymanchem, however I note that PubChem shows the incorrect cis enantiomer (which I presume is where I would have copied the mistake through from when I made the page back in 2010). They also list a different CAS number which I would guess may be for the cis enantiomer, but I am pretty certain this is not the enantiomer generally used in research. Someone with Scifinder access might have to check to make sure though. Meodipt (talk) 19:55, 1 February 2017 (UTC)

What about File:IPMDHNPC.png? Should it get deleted as incorrect? --Leyo 23:21, 12 February 2017 (UTC)


 * I suppose so, yes that should be deleted also. Meodipt (talk) 10:36, 13 February 2017 (UTC)

Wikipedia talk:WikiProject Chemistry
There is currently and RFC on what do do with the shortcuts used for the chemistry-related projects. Please comment. Headbomb {talk / contribs / physics / books} 16:11, 14 February 2017 (UTC)

Chemical compound
I'm surprised this hasn't been mentioned on this talk page, but Chemical compound has been under some contention recently regarding its accuracy and overall content. Your assistance in improving this article is requested. Primefac (talk) 13:10, 22 February 2017 (UTC)

Is nitrous oxide "volatile"?
Can an expert in chemistry please check the concern I have posted regarding the article on the Texas City disaster, here? Thank you very much. 129.97.18.167 (talk) 02:06, 1 March 2017 (UTC)

RFC of interest to this project.
See Talk:Chemistry. Please contribute if you have an opinion. Thanks. -- Jayron 32 02:52, 4 March 2017 (UTC)

WP:CHEMISTRY/WP:CHEMICALS shorcut updated
Note that per this RFC, the shortcuts to WP:CHEMISTRY/WP:CHEMICALS have been updated.


 * WP:CHEM now refer to WP:CHEMISTRY
 * WP:CHEMS now refer to WP:CHEMICALS
 * WP:CHM is deprecated

Old discussions have had their shortcuts updated already. If I have made a mistake during an update, feel free to revert. Headbomb {talk / contribs / physics / books} 15:57, 23 March 2017 (UTC)

Ferric carboxymaltose
Long article, but chembox/drugbox is missing. --Leyo 12:27, 6 April 2017 (UTC)
 * Uggh! You might want to flag this over at WP:Med as well, the article is going to need a lot of editing. --Project Osprey (talk) 12:34, 6 April 2017 (UTC)
 * Was redirected to Iron supplement. --Leyo 11:41, 7 April 2017 (UTC)

Polymethylsiloxane polyhydrate
Is it possible and reasonable to add a general chemical structure? --Leyo 09:39, 11 April 2017 (UTC)
 * That article seems suspicious. Looks like an advert to me.  Probably yet another WP:COI contribution. --Smokefoot (talk) 17:10, 11 April 2017 (UTC)
 * Well, maybe. Anyway, is there an answer to my question? --Leyo 23:03, 15 April 2017 (UTC)
 * It is total crap. If you look at the references they are all in Russian or have a page number that doesn't exist.  A search on the term comes up with potentially quack remedy sites.  I think it needs to go.JSR (talk) 23:31, 15 April 2017 (UTC)
 * Well, would you like to file a deletion request? --Leyo 19:34, 16 April 2017 (UTC)
 * That will be done.JSR (talk) 19:36, 16 April 2017 (UTC)
 * See Special:Diff/775736336/776623731. --Leyo 21:26, 21 April 2017 (UTC)
 * Noticed that. Nothing substantial though.  Links to web site of manufacturer and a few studies which are small.JSR (talk) 21:49, 21 April 2017 (UTC)

Popular pages report
We – Community Tech – are happy to announce that the Popular pages bot is back up-and-running (after a one year hiatus)! You're receiving this message because your WikiProject or task force is signed up to receive the popular pages report. Every month, will post at /Popular pages with a list of the most-viewed pages over the previous month that are within the scope of. We've made some enhancements to the original report. Here's what's new:
 * The pageview data includes both desktop and mobile data.
 * The report will include a link to the pageviews tool for each article, to dig deeper into any surprises or anomalies.
 * The report will include the total pageviews for the entire project (including redirects).

We're grateful to for his original, and we wish his bot a happy robot retirement. Just as before, we hope the popular pages reports will aid you in understanding the reach of, and what articles may be deserving of more attention. If you have any questions or concerns please contact us at m:User talk:Community Tech bot.

Warm regards, the Community Tech Team 17:16, 17 May 2017 (UTC)

Should Enzymes be in WP:Chemicals?
A question of scope: Should enzymes be in this wikiproject? The issue was raised briefly in 2007, but noting conclusive was said. The project now has a random scattering of a few hundred enzymes (approximate list). My opinion is that, since they're already in WP:MCB, they're not necessary to put in WP:CHEMS, especially since they're typically several kilodaltons. Peptides (e.g. Dactinomycin would be more appropriate). What're people's opinions? T.Shafee(Evo &#38; Evo)talk 12:23, 18 May 2017 (UTC)
 * I also see no problems having them in both. There are alwas questions of where to draw a line.  --Dirk Beetstra T  C 12:49, 18 May 2017 (UTC)
 * I would leave them out and tag them only for MCB. The general article, protein deserves a chemistry proj tag though. Graeme Bartlett (talk) 13:24, 18 May 2017 (UTC)

In general, I would suggest drawing the line at approx 4 kiloDaltons. That would include most short peptides (both ribosomally synthesised peptides and non-ribosomal peptides), but exclude most proteins (inc. enzymes). There is inevitably a grey area, but I think a 500 kDa enzyme isn't usefully included in WP:CHEM's scope. T.Shafee(Evo &#38; Evo)talk 03:10, 31 May 2017 (UTC)

Expert help requested at Gel point
Hello Chemistry experts - would someone - perhaps a polymers expert? - please take a look at Gel point. It's horrible (not a WP policy phrase), and has been horrible for some time. I think it started to go wrong here. To my eyes the article doesn't actually explain what "gel point" is. Would it be reasonable to roll back the article to the Revision as of 21:31, 17 March 2013 ? Shhhnotsoloud (talk) 11:46, 28 May 2017 (UTC)
 * Its not much but at least my revision is probably more understandable.--Smokefoot (talk) 16:56, 31 May 2017 (UTC)

Thiazine
Is this article about the group of substances (according to text) or about the individual compound (according to chembox)? --Leyo 09:13, 28 June 2017 (UTC)
 * Either, to my eyes Category:Thiazines doesn't contain any real thiazines and I've not manages to find others via structural searching. So that page is all we have. --Project Osprey (talk) 10:31, 28 June 2017 (UTC)
 * I've changed the infobox: "Thiazine" the group is the top name (if group is the right word?), then index1_label and ChemspiderID1 etc. are the [1,4]Thiazine compound. Other [x,y] compounds can be added. -DePiep (talk) 12:05, 28 June 2017 (UTC)


 * When this individualk item is solved, this generalisation: are there more sunch examples where could & should serve a situation: group + its compounds? If so, is this the advisable approach? -DePiep (talk) 12:05, 28 June 2017 (UTC)

Arsenic compounds' categorization
Because of their current difficulty to navigate, I'm going to be reorganizing the arsenic compounds' categories over the next couple of days. I am open to suggestions and criticism. While I am working on them the categories will likely be a big mess. OrganoMetallurgy (talk) 18:45, 1 July 2017 (UTC)

Diorganophosphites
Related to OrganoMetallurgy plea above, organophosphorus is an area where some advice would be welcome. Overall organophosphorus chemistry (formally phosphites are not organophosphorus since there is no P-C bond, I guess) is complicated even for semi-experts, so it must be nearly impossible for outsiders. One idea would to make organophosphorus compound a graphical disambiguation page with every reasonable permutation of H-OR-O--R-Cl-anion-cation on phosphorus pictured with links.--Smokefoot (talk) 02:03, 10 July 2017 (UTC)
 * Maybe diethylphosphite, which I call a diorganophosphite, could be retitled diethylphosphonate. But phosphonate describes conjugate bases of organo-P)O)(OH)2, phosphonic acids with P-C bonds.
 * The article phosphite ester is weird since I at least had never heard that term before Wikipedia. It now describes two very different classes of compounds (P(III) vs P(V)).
 * Phosphite is about the conjugate base of phosphorous acid, which are fairly exotic salts. Every organometallic chemist considers "phosphite" to be ligand of the type P(OR)3 (R = organic).  We might devise a specialized label for HPO32- salts, and then let phosphite be about the ligands and their P(V) relatives.
 * I was also getting confused about some of the new articles that you were writing. In many cases the official name also contains "hydrogen". Note also that O=P-H on a P(V) atom probably very easily converts to H-O-P on a P(III) atom, so some of these possibilities would be equivalent. We will have to checkout what IUPAC says, and then confuse ourselves further by Chemical Abstracts naming and see what we find out. Graeme Bartlett (talk) 06:29, 10 July 2017 (UTC)


 * I agree that a WP:SETINDEX graphical table is a good way to help our readers.DMacks (talk) 13:05, 10 July 2017 (UTC)


 * Also related to this is Tetraethyl pyrophosphate (or diphosphites). There are also halo derivatives, one is named diethyl phosphorochloridate (but this is a phosphate). Dichloro methyl phosphite with two chlorines and a methoxy on the P. Dimethyl chlorophosphite also exists with one chlorine connection and two methoxy.  For similar things in the phosphate series, there is a =O groups as well.
 * Pubchem likes to use the term "xxxxphosphite for the ion, and then adds "hydrogen" to have "xxxx hydrogen phosphite" for the neutral molecule. Their images show the POH tautomer, even if that is not the real thing! (I called P(V) and P(III) based on how many bonds to P, rather than saying H is + and O is -). Graeme Bartlett (talk) 11:21, 11 July 2017 (UTC)

HP(O)(OR)2
First about these HP(O)(OR)2 compounds, I can find no comment on their existence as the (RO)2POH tautomer. Apparently that pattern also applies to phosphorous acid: no trihydroxyphosphine. The hydroxy tautomer only appears when there are C-P bonds, and even then the POH tautomer is usually minor, if observed at all. These statements are a mixture of opinion and fact. More reading may change the situation. About nomenclature, here is CAS's list of names for HP(O)(OEt)2: My feeling is that organic chemists, which seem to dominate the area, call this compound "diethylphosphite" or "diethyl phosphite". Aldrich uses the latter. Aldrich switches to phosphonate when the connectivity is RP(O)(OR')2, i.e. methylation of diethylphosphite gives methyldiethylphosphonate.--Smokefoot (talk) 12:46, 10 July 2017 (UTC)
 * Phosphonic acid, diethyl ester
 * Ethyl phosphite, (EtO)2(HO)P (6CI) [ I think this is referring to the P(III) tautomer]
 * Diethoxyphosphine oxide
 * Diethyl acid phosphite
 * Diethyl hydrogen phosphite
 * Diethyl phosphite
 * Diethyl phosphonate
 * Hydrogen diethyl phosphite
 * Yes, as an organic chemist who's used it a little in research, I've always called it diethyl phosphite. It's understood with phosphorus that diethyl phosphite has the P=O present, even though triethyl phosphite doesn't!  Those names are in very common use, in my personal experience (in the UK and US).  Walkerma (talk) 13:02, 10 July 2017 (UTC)

How is it that we're defining the tautomers as PIII and PV ? It has admittedly been a very long time since I did it by hand but to my eyes both forms are PIII. The naming of a lot of these things is a mess, I believe that historically chemists worked out the molecular formulas fairly early but couldn't agree on structures - and thus names - for some time (here's a nice example from 1954 of them still trying to organise things). Regarding phosphite ester: It was intentionally split off from phosphite because having all the possible oraganic and inorganic things in the same page was just too confusing. We did something similar with Sulfate and Organosulfate. Perhaps we should consider a name change from phosphite ester → Organophosphite? --Project Osprey (talk) 19:35, 10 July 2017 (UTC)
 * , I recently edite the diethylphosphite article and added oxidation states, as:
 * (C2H5O)2PV(O)H ⇌ (C2H5O)2PIII(OH)
 * I think it is clear that the latter form is phosphorous(III). I rationalised the former as:  Each ethoxy group is formally a 1&minus; anion and the oxide is 2&minus;.  The phosphorous must be a cationic centre, to which a proton would not bind, so the hydride needs to be treated as a 1&minus; anion...  thus the P centre is surrounded by 5&minus; charge and so must be assigned as phosphorous(V).
 * The same reasoning has H3PO3, which looking at this formula should be phosphorous(III), as having a phosphorous(III) form P(OH)3 and a (dominant) phosphorous(V) form (HO)2P(O)H, with the two acidic hydrogens in the +1 state and the non-ionisable hydrogen being a formal hydride in a P&mdash;H covalent bond. However, our article states in the section Phosphorous acid that the solid-state form is the dominant form but has phosphorous in a +3 oxidation state.
 * I came to post here to ask everyone whether I am missing something / reasoning incorrectly, or are our articles wrong in asserting these tautomerisations involve only phosphorous(III) centres.
 * I agree that some renaming appears necessary, but also rewriting and reorganising. Our Organophosphorus template, for example, list the phosphite esters as having only P&mdash;O bonds, but diethylphosphite not only has a P&mdash;H bond, but the article itself declares it to be "the site of high reactivity".  EdChem (talk) 01:10, 11 July 2017 (UTC)


 * I'm not sure I agree with your assessment, the diethylphosphite article describes P-H as being deprotonatable, which would infer acidic character making the H 1+ and thus the tautomers PIII in both cases. As an added argument; conservation of charge is a physical law, even in the somewhat made-up world of oxidation state. If you're going from III to V then you need an oxidising agent and your equilibrium reaction needs to be balanced in some way, electrons on the right-hand-side perhaps. You can't just change states. We could perhaps have a page on dialkyl hydrogen phosphite esters, providing we can find enough to say, but I think they should still be mentioned at phosphite ester as they are notionally related. We do similar things on other pages, ester mentions orthoesters for example. --Project Osprey (talk) 10:37, 11 July 2017 (UTC)
 * 0) slight pedantic (apologies) comments...
 * 1) The relevant articles that need renaming are phosphite for the obscure anions H2PO3- and HPO32- (maybe phosphite anion) and phosphite ester for compounds "everyone" calls phosphites. IMHO.
 * 2) Diethylphosphite ((EtO)2P(O)H) is a common name, re-affirmed by user:Walkerma and Aldrich and titles of articles in eEROS and Org Syn... So we cannot toy with those article titles probably.
 * 3) The oxidation state formalism and the idea of redox tautomerism can seem a little weird. Assigning (RO)-, O2-, and H- defines the ligand set making phosphorus P5+. This is what EdChem indicates above. H3PO3 is mainly the P(V) tautomer, the P(OH)3 tautomer has not been detected but only inferred.  Ditto for the alkoxy derivatives.  I recall ΔG(298) > 10 kcal/mol, which translates to a monstrous Keq.  H3PO3 is a deceptive presentation of the formula, but (HO)2P(O)H is too cumbersome.  Somewhat llke many transition metal "hydrides", the (EtO)2P(O)H center is deprotonatable, but requires very strong base.  We probably dont want to get into the protic vs hydridic labeling of P-H center to decide oxidation state.  That battle was fought long ago and is semi-useless.--Smokefoot (talk) 13:20, 11 July 2017 (UTC)
 * Yeah, I think I'm going to have to admit that I don't understand this properly - I also really don't want to have conversation about things being amphoteric. --Project Osprey (talk) 22:09, 11 July 2017 (UTC)
 * You are not losing it, the assignments are semi-arbitrary. The phosphite anion (HPO32-), which I am insisting is P(V), is isoelectronic with bisulfite (HSO3-), which most would prefer to describe S(IV).  The difference is that the bisulfite is reasonably acidic. --Smokefoot (talk) 00:58, 12 July 2017 (UTC)

Phosphite
OK, I just moved "phosphite" to "phosphite anion", which should alleviate much confusion. The article had fairly extensive (hand-wringing?) discussion nomenclature, which I compressed. It also had images of the PO33- anions, which are really only of theoretical interest. I will make chemdraws of the H2PO3- (Cs symmetry) and HPO32- (C3v symmetry). I hope that these changes are not controversial, but if they cause concern, leave a note here or on the Talk pages. --Smokefoot (talk) 14:26, 11 July 2017 (UTC)

My FRD-903 and GenX disruption
I just created stub articles 2,3,3,3-tetrafluoro-2-(heptafluoropropoxy)propanoic acid (FRD-903) and GenX (chemical process). Then I got a robotic message that I should "incubate the article with minimal disruption". Anyone else feels invited to edit? -DePiep (talk) 23:28, 22 July 2017 (UTC)
 * The aritcle has no references for verification, but looks like a good potential topic. I thought it best to draftify until sources were added so that somebody didn't tag it for deletion and it can be worked on with minimal disruption from new page patrol. Once it has sources it can easily be moved back to main article space. —  InsertCleverPhraseHere  00:06, 23 July 2017 (UTC)
 * Yes, that's why the was added. There wasn't even time to link to the QID (iw).  And arguing "I deleted it to prevent deletion" is not that convincing. "minimal disruption from page patrol": just leave us alone I'd say. -DePiep (talk) 02:03, 23 July 2017 (UTC)
 * "The aritcle has no references for verification" &mdash; is has/had two external links to identifiers. -DePiep (talk) 02:05, 23 July 2017 (UTC)
 * I moved it back to 2,3,3,3-Tetrafluoro-2-(heptafluoropropoxy)propanoic acid, and added iw link to nlwiki. Graeme Bartlett (talk) 13:13, 23 July 2017 (UTC)
 * I agree that it clearly meets WP:V to exist., you and other new page patrollers should well be aware that just because there are no footnotes in a a new article, there could be embedded external links in an article or infobox. DMacks (talk) 13:54, 23 July 2017 (UTC)
 * Apologies for stepping on toes, and for missing the external links. Obviously draftification wasn't the right choice in this case, and it is hard to get it right every time.
 * In any case there are some things I'd like to know while I am here: Are chemical listings with no further information like this common? These external links seem like routine listings that exist for nearly every chemical in existence, how does that make the topic notable? As far as I am aware chemicals must still meet WP:GNG, that means significant coverage is required. Routine listings with no further expounding on the applications, properties, or usefulness of the chemical don't seem like significant coverage to me. And as far as I know, Notability_(natural_sciences) never became policy, so existence alone isn't enough to justify an article. I only ask because this topic (before the recent ref was added) seems clearly not notable from what I can see in the guideline pages. Is there some obscure policy page or common practice related to chemistry articles that I am missing or haven't seen? —  InsertCleverPhraseHere  14:18, 23 July 2017 (UTC)
 * Yes, I felt something on my toes because it is quite frustrating when I want to edit a new article and it's gone.
 * Now about notability etc: First, using just external links (identifiers) to base the existence of a substance to me looks safe enough for a stub level, more so when an iw (wikidata QID) is available. From there the article can expand. (There must be a below-stub-level quality, but I've not experienced that). And IMO this chemical substance is notable because, in combination with its GenX (chemical process) (new too), it is produced in the RL chemical industry, while being (potantially) toxic and biologically harmful. -DePiep (talk) 15:00, 23 July 2017 (UTC)
 * Notability is certainly not met just by existing and we're not in the business of creating stubs for every compound. But some are fairly important in their own fairly small niche, or are important as an "intermediate" in a larger process. Any intermediate in an industrial process is a commercially important chemical in its own right, one could editorially decide that it is best covered in a unified article about the process rather than stand-alone. In this case, the article mentions the importance of this chemical in a process that someone thinks is notable enough to have its own article. Without FRD-903, you can't do GenX, and that article further has a specific cite discussing this chemical. DMacks (talk) 15:15, 23 July 2017 (UTC)
 * @DePiep Just so you know, stubs do not enjoy any special protections from the General Notability Guideline, so I wouldn't bet on identifiers being enough to hang notability on, even for a stub. The other stuff you mention is what makes it notable, but none of that stuff was in the original article. I would suggest having a couple of solid detailed in-depth sources in your new articles before publication if you want to be assured that NPP will keep clear of your little piggies in the future. A lot of new page patrollers patrol from the front of the list, to catch the most egregious violations early, so they are likely to land rather quickly on them. Sorry again for getting in your way; one of the reasons I usually check the author of the page before draftifying, but of course the one time I didn't check it is one of the most prolific editors on the wiki. Just my luck.
 * @DMacks I get all that. When I came across the article the GenX article was a redlink, so it wasn't clear what exactly that even was. A ref like the one that Graeme Bartlett added would have been good to have in the starting article to demonstrate notability. —  InsertCleverPhraseHere  15:31, 23 July 2017 (UTC)
 * As I wrote, those earliest links provided are not basing notability, but are a defining source. Being solid?: no need to doubt CAS or PubChem in this, and surely not afterwards only to make a point. Next, in the very first version notability was present, the same notability described above (GenX process). So far, you have not based your accusation of "disruption". -DePiep (talk) 15:56, 23 July 2017 (UTC)
 * As someone mentioned earlier. A CAS entry does not provide notability (it only states that someone wrote about it in an article - the compound does not even have to exist). CAS contains more than 130 million organic and inorganic substances and 67 million sequences. Notability was not present. Notability is made with external sources which cover the topic in depth. In the past we have had a lot of chemical articles with no content and no notability. As far as I can see Graeme Bartlett reference discuss the ammonium salt of the acid. So maybe the salt is more notable than the acid? Christian75 (talk) 17:14, 23 July 2017 (UTC)
 * re As someone mentioned earlier: yes, I was the one who did that. A CAS Number is a reliable source. As is a PubChem identifier. I also wrote (3rd time by now) that that does not claim notability, but it does claim WP:RS. Notability of FRD-903 is based on the GenX link. -DePiep (talk) 21:52, 23 July 2017 (UTC)
 * In the case of this page it was very easy to do an online search and see that there were many writings about the topic, and so that it was notable. For new page patrol it only needs a claim of importance, which it did have, so NPP should not have moved it, perhaps tagged it with refimprove or no references tag. The ref I added is not the best (reliable), but it proves that it is an important topic. Graeme Bartlett (talk) 23:26, 23 July 2017 (UTC)
 * Thanks us all. By now, we have two serious new articles, well scrutinised. Let's drop the detail-quarrelings. I'm fine with this. -DePiep (talk) 23:33, 23 July 2017 (UTC)

Chemspider ID verification question for Nitenpyram
Hi, Since the Chemspider ID was first added to the Nitenpyram page (on 12 August 2013), it has been flagged as an unverified change (with a red cross). I checked the Chemspider website and it is the correct Chemspider ID. I couldn't see anything in the instructions here: WikiProject Chemicals/Chembox validation for what to do to verify that a change is correct. Please could someone help me out? Thanks DferDaisy (talk) 23:54, 23 July 2017 (UTC)

Purple pH indicator
Dear Chemistry team, I've just started an article on Metacresol purple, I've got a medical background and I am struggling to compare/contrast three similar chemicals With:
 * Metacresol purple
 * m-Cresol, and
 * Bromocresol purple

Could you add a little chemistry finesse to the article. Many thanks, DemandAmbition (talk) 15:39, 30 July 2017 (UTC)

RfC: Colour group 12 as post-transition metals
I am seeking comments on a proposal to color code the group 12 elements as post-transition metals in the Wikipedia periodic table, rather than transition metals as they are currently color coded.

The RfC can be found here. Sandbh (talk) 23:06, 9 August 2017 (UTC)

History of WikiProject Chemicals


Following Wikimania 2017 in Montreal I and some others went to nearby SUNY Potsdam for a Wikipedia event hosted by WikiProject Chemicals' own Martin Walker, user:Walkerma. He told stories about Wikipedia around 2005 including the one in this video. He mentioned, , , and , so I thought WikiProject Chemicals would be interested. I looked around elsewhere to see what other media existed around WP:Chemicals and found these.



There was also a WikiProject Report in The Signpost in 2012. It is challenging to recognize history when it happens but I think that WikiProject Chemicals might have been a significant and successful attempt at sorting essential information in a way that made it radically more available than anything which had come before. Congratulations to participants here for all that you have accomplished.  Blue Rasberry  (talk)  14:24, 18 August 2017 (UTC)
 * Great, I'll have a look at this, thanks Martin! --Dirk Beetstra T  C 15:47, 18 August 2017 (UTC)


 * Good to hear Martin's recollections about work we did 10 years ago. I rarely edit now but it's heartening to see things are still going well. --Rifleman 82 (talk) 02:28, 24 August 2017 (UTC)

WP:CHEMICALS: disconnect -talk from Chemical infobox-talk
At the moment, talkpage of Chembox is Wikipedia talk:Chemical infobox (it's Redirected). I propose to give Template:Chembox its own talkpage Template talk:Chembox (content in there). Discussion here. -DePiep (talk) 07:42, 30 August 2017 (UTC)

Help with Belladonnine
There are apparently two chemicals that have this name (α-Belladonnine and β-Belladonnine), that are stereoisomers of each other according to chemspider and others. Are either of these actually notable enough to merit an article? Does the simple name generally refer to only one of them? Are they together a notable set of chemicals (or are ambiguously discussed in the literature enough) to have one article for the simple name and then separate chembox for each isomer? DMacks (talk) 18:55, 2 September 2017 (UTC)
 * Better start to write about the stereoisomers in Belladonnine using dedicated sections and once you have enough material then it will easy to justify the creation of new articles. Snipre (talk) 23:18, 3 September 2017 (UTC)

Right structure for Solvant brown 41
I have difficulties to get the right structure for Solvant brown 41 or Bismarck brown Y: is this chemical a mixtures of compounds having different E,Z double bonds configuration or this name is used only for the E,E-molecule ?


 * E,E-configuration : ChemIDplus, Colour Index International database
 * mixture of E,E-, E,Z-, Z,Z-configurations: ECHA, Reaxys

Or more simpler, can someone give me the real structure of CAS 1052-38-6 : E,E-configuration or undefined configuration ? Thanks Snipre (talk) 23:29, 3 September 2017 (UTC)
 * Here is the low-down. Bottom line is that your picture is fine but this stuff is slightly artsy in its prep.

--Smokefoot (talk) 20:23, 4 September 2017 (UTC)
 * Diarylazo compounds are almost always E, so the E vs Z question is a non-issue. That is just my guess.
 * Quote from Ullmann's encyclopedia (Azo Dyes): "Vesuvin and Bismarck Brown products are obtained when m-phenylenediamine, 2,4- and 2,6-diaminotoluene, or mixtures thereof ... consist of mixtures of monoazo, disazo, and polyazo dyes, with the disazo species probably predominating...."
 * CAS search doesnt show anything particularly interesting. 90 hits, the literature <1959 is not shown (probably they did not know what it was).
 * It was introduced in 1863, so is probably the first azo dye (good catch!).
 * Here is a confusing and partially mistaken quote from Ullmann's (Phenylenediamines and Toluenediamines): "m-phenylenediamine is reacted [SIC] with excess nitrous acid, both amino groups are diazotized and couple with unreacted m-phenylenediamine to give 5,5�-[(4-methyl-m-phenylene)bisazo]bis(toluene-2,4-diamine), Bismarck Brown, one of the first azo dyes..."

RfC: Replace categories of poly/diatomic nonmetal with less active/active nonmetal
I am seeking comments on a proposal to change the name and composition of two of the colour categories appearing on our periodic table, as follows:

The RfC can be found here. Sandbh (talk) 23:49, 9 September 2017 (UTC)

Fresh eyes welcome at borane, edits by user:Plasmic Physics
From my perspective, our article borane has problems, consisting of sections that seem to be OR and subject to WP:UNDUE, often supported by narrow referencing. The problematic editing, IMHO, is by user:Plasmic Physics, who has often been controversial in the Wikipedia chemistry area, and with whom I disagree on how hydrides should be described. For example, never has the "amphotericity" of BH3 entered into consideration in my work and teaching with boron hydrides. --Smokefoot (talk) 13:54, 11 September 2017 (UTC)

National Nanotechnology Day
National Nanotechnology Day is coming up on October 9 (because nano is 10^-9, of course). DNA nanotechnology is planned for TFA that day, and it would be great to have some DYKs as well. A lot of nanoparticles by composition articles are missing or short. Shockingly, there's no Gold nanoparticle article—it redirects to Colloidal gold, which is an overlapping but different subject. Similarly, there are a lot of chemical reactions involving nanomaterials that could be created or expanded; some sections of Fullerene chemistry, Carbon nanotube chemistry, and Selective chemistry of single-walled nanotubes could be split out into new articles. I'm happy to collaborate on nanotechnology articles; please ping me if you're interested. John P. Sadowski (NIOSH) (talk) 04:42, 29 August 2017 (UTC)
 * Wikipedia periodically gets a rush of edits to nano this and that, but these edits are almost always very narrow and probably self-citations. So we have a problem with overviews.  So a priority list from you or an expert would be welcome, and perhaps we can organize or improve some key articles.  I take it that gold nanoparticle is your #1 recommendation. --Smokefoot (talk) 12:48, 29 August 2017 (UTC)
 * I made a nanotechnology vital articles list a while ago (Edit: I made this list as a volunteer before I started working for NIOSH) . Besides individual nanoparticle types there are a lot of lithography and microscopy methods and biographies (e.g. winners of the Feynman Prize in Nanotechnology) that are good DYK targets.  I'm happy to give an expedited GA review who wants to work on more-developed articles.  I've been doing some cleanup to Nanomaterials and Nanosensor myself lately and would welcome help there as well.  John P. Sadowski (NIOSH) (talk) 19:13, 29 August 2017 (UTC)
 * Very helpful list, except the list of people, which just invites trouble/egos/coi. One approach is to start with technology. Which nanotechnologies are most successful commercially?  Maybe microelectronics. Even then, I am unsure if mass fabrication is being conducted on the nanometer scale yet.--Smokefoot (talk) 19:22, 29 August 2017 (UTC)
 * Well, standard semiconductor fabrication has been using sub-100 nm features for well over a decade now; currently the 10 nanometer process is being used. I can try to find sources about which nanomaterials are being used commercially.  John P. Sadowski (NIOSH) (talk) 02:23, 30 August 2017 (UTC)

I've looked up some sources to to get ideas for the most commonly used nanomaterials. A lot of them don't exist and would make good targets for DYKs: John P. Sadowski (NIOSH) (talk) 00:20, 7 September 2017 (UTC)
 * Titanium dioxide nanoparticle, silicon dioxide nanoparticle, and zinc oxide nanoparticle are the most produced nanomaterials worldwide.
 * Silver nanoparticle is the most popular advertised nanomaterial in the Nanotechnology Consumer Products Index.
 * Nanomaterials widely considered important for risk assessment are gold nanoparticle, silicon dioxide nanoparticle, silver nanoparticle, titanium dioxide nanoparticle, cerium oxide nanoparticle, fullerene, iron nanoparticle, iron oxide nanoparticle, and carbon nanotube.
 * Some others used in consumer products: aluminum oxide nanoparticle, aluminum nanoparticle, zirconium nanoparticle, zirconium oxide nanoparticle, hydroxyapatite nanoparticle, ceramic nanoparticle, platinum nanoparticle, manganese oxide nanoparticle, titanium nanoparticle, carbon black

Update: Titanium dioxide nanoparticle is now up. Comments and improvements are welcome! John P. Sadowski (NIOSH) (talk) 04:19, 16 September 2017 (UTC)
 * This article is very good. It gives readers a feel for the material, which is widely used.  I will say that the composition of the article in terms of weight and referencing illustrates the #1 challenge to writing chemistry-related articles: the sources available on-line are heavily focused on environmental health and safety issues.  Not only is such content free and abundant, it is easily digested by nontechnical editors and those without access to pay-wall protected sources.  So far as I can see, almost all secondary articles are focused on EHS issues.  More detailed suggestions can be put on the Talk page. --Smokefoot (talk) 14:34, 16 September 2017 (UTC)

Dimethyldiborane
Dimethyldiborane has two infoboxes, each with a melting point. The second has recently been changed and is now showing an error because it includes "?". The reason for the recent change is that the melting point was −150.2 K—a highly unlikely value! That seems to have been added in August 2015 and hasn't been fixed since, I think. Can someone please investigate. Johnuniq (talk) 11:17, 20 September 2017 (UTC)
 * Yes, clearly −150.2 K is not correct. I have removed it.  If someone can find a reliable source (reporting a non-impossible value) that would be great.  For now it can be blank.  -- Ed (Edgar181) 12:43, 20 September 2017 (UTC)
 * Looking a little closer, there is a referenced value for melting point in the article text, −150.2 °C. So it looks like it was just a typo, C not K.  -- Ed (Edgar181) 13:05, 20 September 2017 (UTC)
 * The article should probably be split into the 1,1- and 1,2- isomers, even though each is pretty obscure.--Smokefoot (talk) 14:29, 20 September 2017 (UTC)
 * I split out the 1,1-Dimethyldiborane just now. I (or someone) need to go back to the dimethyldiborane, move it to 1,2 and the cull the 1,1-material.  I removed a paragraph on complicated redistribution reactions as being overspecialized.  Others might disagree, so go ahead and act on your instincts.  I was also planning to remove statements and references to PhD thesis. It is possible that a number of mistakes remain or were made by me, as I was acting fast. --Smokefoot (talk) 17:35, 20 September 2017 (UTC)
 * The redistribution reactions are an important feature of the x-methyldiboranes, as they occur spontaneously at room temperature, and there will normally be a mixture in what started as a pure substance. Graeme Bartlett (talk) 09:09, 22 September 2017 (UTC)
 * OK, I will dig more into this. It appears that H. C. Brown, who was a prominent organic chemist, used (MeBH2)2 without observing formation of Me2BH-derivatives but I could be wrong.  In any case we might continue this specialized conversation on the [[Talk:1,2-Dimethyldiborane.  Thanks for the heads up.--Smokefoot (talk) 16:15, 23 September 2017 (UTC)

Trifluoromethyl sulphur pentafluoride
I wonder why there is no box. — Preceding unsigned comment added by 213.55.176.150 (talk) 18:48, 4 September 2017‎ (UTC)
 * The absence of that box is a message from the spirit world that your services are needed. Here it is, paste it in and start filling it out! --Smokefoot (talk) 19:45, 4 September 2017 (UTC)


 * The talkpage has the chemspider link, which will help get some details. I moved the article to trifluoromethylsulfur pentafluoride as a more MOS-compliant name (I guess the spacing could go either way—I chose what original journal-pub used—but the "ph"→"f" is WP:SULF). And I just added a lead ref for the chemical itself that might have some additional properties. The article said it was first identified in the atmosphere, rather than a long-known compound be detected there recently. DMacks (talk) 20:00, 4 September 2017 (UTC)
 * For some reason, the article wasn't in any chemistry-related maintenance categories (i.e. subcategories of Category:WikiProject Chemicals). There may be more similar cases then. How to find them? --Leyo 22:29, 4 September 2017 (UTC)
 * WP:PETSCAN is our friend. Like:
 * "Having & not in Category:WikiProject Chemicals"
 * "Having & not in Category:WikiProject Chemicals"
 * -DePiep (talk) 23:55, 4 September 2017 (UTC)
 * This would not have worked out at all in the relevant version. --Leyo 08:34, 5 September 2017 (UTC)
 * That's more difficult, right? ;-) --Leyo 11:33, 10 September 2017 (UTC)
 * You're right. For those:
 * WP:PETSCAN:
 * "Not having & in Category:WikiProject Chemicals"
 * "Not having & in Category:WikiProject Chemicals"
 * Of course, an article being negative on all three lists, will not be in view. Unless there is another marker (having Chem?). I'm low in time now, cannot write a petscan. -DePiep (talk) 11:38, 10 September 2017 (UTC)
 * Forgot: plus there is the problem of WikiProject Chemicals being on the talkpage. Have not figured out how to fix that in petscan. -DePiep (talk) 11:41, 10 September 2017 (UTC)
 * Are you a bit familiar with petscan? Very useful, given the wiki category system. -DePiep (talk) 19:23, 10 September 2017 (UTC)
 * I am familiar with PetScan. To answer your question above, there is the option Use talk pages instead.
 * Unfortunately, your suggestions would not have worked for the relevant version (and not for similar cases). --Leyo 21:26, 10 September 2017 (UTC)
 * Did not mean to be pedantic, it's just that I did not see a response on that point. And about that version: out of luck situation. -DePiep (talk) 21:58, 10 September 2017 (UTC)

Is the structural formula shown by ChemSpider correct? --Leyo 21:32, 28 September 2017 (UTC)
 * Yes, it's correct. OrganoMetallurgy (talk) 00:39, 29 September 2017 (UTC)

Split cyclooctene
If there is no objection, could I get some administrative help splitting cyclooctene into Trans-cyclooctene and Cis-cyclooctene? The two compounds are pretty different and readers would be interested in one or the other, but usually not both. --Smokefoot (talk) 13:40, 28 September 2017 (UTC)
 * ✅ OK I have split. Please check.  The chembox will need some work to get properties unique as I suspect they differ. Graeme Bartlett (talk) 23:39, 28 September 2017 (UTC)
 * Many thanks. I have started editing the individual article.  The "languages" links should probably be to the more common cis isomer (maybe the disambiguation article?).  I will do that in the next day or so.--Smokefoot (talk) 14:34, 29 September 2017 (UTC)

name change
Proposal: Chembox &rarr; Infobox chemical. See Template_talk. -DePiep (talk) 11:21, 1 October 2017 (UTC)

Should we have separate articles on a WHO Essential medicines?
This was posted at WT:MED. To gain wider input I am posting it here.


 * Discussion at Talk:Potassium permanganate (medical use) Sizeofint (talk) 04:33, 9 October 2017 (UTC)

Tricky COI issue
Amphipols has just received a minor update and bunch of new references courtesy of User:‎Jean-Luc Popot. The author of these papers? You guessed it: J. l. Popot. So there's a plain COI, however Google Scholar implies that Popot is actually a leader in this area and the refs he's added are mostly to well cited secondary reviews (although admittedly his own ones). I can't decide whether to assume good faith or revert. --Project Osprey (talk) 11:57, 11 October 2017 (UTC)


 * I think we can assume good faith; from his edit summary it seems he just found the article and wanted to help since he's an expert in it. And WP:SELFCITE says:
 * Using material you have written or published is allowed within reason, but only if it is relevant, conforms to the content policies, including WP:SELFPUB, and is not excessive. Citations should be in the third person and should not place undue emphasis on your work. When in doubt, defer to the community's opinion: propose the edit on the article's talk page and allow others to review it.
 * His edits look good to me; so we should probably just tell him about the COI and that he should be somewhat careful. Galobtter (talk) 13:14, 11 October 2017 (UTC)
 * AGF is a joke in cases of scientific self-citation in my opinion. It written by and for editors who have not weathered scientific publishing. Here's my approach to these situations:
 * rm non-secondary refs contributed by COI editor, to put the editor on notice.
 * check to see if the COI author ever cites anyone other than themselves.
 * leave a note with the COI editor reminding them of WP:COI, the tone of the note being determined by preceding two bullet points.
 * Unfortunately, the response is often this with COI editors: "Moi? I didnt know about COI... I love Wikipedia... I will definitely contribute more broadly...." silence. --Smokefoot (talk) 13:32, 11 October 2017 (UTC)

Molecular Summary Table
Dear all,

first of all I apologize for my bad English. I will improve it. I´m still new here at wikipedia and my instruction consists in publishing scientifical articles. Most of them contain chemical infoboxes (similar like drugboxes, enzymes)calles "molecular summary table". I tried everything to publish the following table but it doesn´t work.

It´s about the following infobox:

https://de.wikipedia.org/wiki/Wikipedia:Redaktion_Chemie#/media/File:Wikipedia_Support_2.jpg

The molecular summary table is almost published on a german wikipedia support page. I need urgent help in converting this table into wiki syntax. I would be very grateful for support!

Thanks and best regards

--Gunnar Römer (talk) 14:01, 31 October 2017 (UTC)
 * I'm not sure exactly what you want to do. If you want to create infobox, the article (BI 99179) has to be created first (only one paper seems be published on it so not sure of notability), and then a infobox can be filled there from the information in your image. Also it'd be better if you linked to the source of the information in your image. Galobtter (talk) 14:39, 31 October 2017 (UTC)
 * It is possible you should not be working on this wikipedia article at all. According to many of your contributions so far (for example, commons:Special:Contributions/Gunnar Römer) you are doing this as part of paid work for a company potentially aimed at promoting one of their chemicals or results. Please read our conflict of interest guideline before continuing. DMacks (talk) 14:47, 31 October 2017 (UTC)


 * On the English Wikipedia, we use the Chembox template for creating those summary tables. I see on the German Wikipedia, you were given information about the equivalent process they have for them (de:Wikipedia:Redaktion Chemie). DMacks (talk) 14:43, 31 October 2017 (UTC)

Dear DMaks, thank you for the information. Please excuse the inconvenience. I´ll make sure that all guidelines will be respected during my work. It´s true, that I publish these articles on behalf. I assure you that the articles don´t contain any kind of advertising. The only aim is making these molecules more famous in scientifics. I work exclusively as a freelancer. Best regards. --Gunnar Römer (talk) 12:54, 1 November 2017 (UTC)
 * 'the only aim is making these molecules more famous in scientifics' .. ?? That sounds like promoting/advertising to me, and hence not really in line with Wikipedia's goals.  By the way, for starters, please take wmf:Terms of use at heart.  But even with that, WP:COI should stop you at the level of merely suggesting that the article be created (or maybe that you create it in Draft space, and then leave it others to decide whether it is moved, and leaving it alone when it is in mainspace).  --Dirk Beetstra T  C 13:19, 1 November 2017 (UTC)

Indoxyl sulfate and Indole-3-carboxaldehyde
I'm seeking feedback on what to name the articles on these compounds. Please see WT:MCB for further details.  Seppi  333  (Insert 2¢) 22:29, 10 November 2017 (UTC)


 * Indican seems to come up a lot for the former. Regarding the latter, which compound are you refering to 1H-indole-3-carbaldehyde or indole-3-carbaldehyde - is the nitrogen atom hydrogenated or not? Plasmic Physics (talk) 01:09, 11 November 2017 (UTC)
 * Yes, but, both would fall under the same page scope anyway, would they not? Anyway, "indican" (i.e., the phytochemical) isn't really the same thing as indoxyl sulfate, which is a metabolite of indican in humans (indoxyl is too).  Seppi  333  (Insert 2¢) 02:58, 11 November 2017 (UTC)


 * They would fall under the same scope, but one is clearly superior. For example, even though methyl radical and methane fall under the same scope, it doesn't very much sense to describe methane in the context of the methyl radical if you only had one article. You suggestion of indoxyl sulfate is fine. Plasmic Physics (talk) 06:39, 11 November 2017 (UTC)

Can someone please add a structure diagram for indoxyl sulfate?  Seppi  333  (Insert 2¢) 05:15, 16 November 2017 (UTC)
 * User:Edgar181 did it. DMacks (talk) 18:22, 16 November 2017 (UTC)
 * Thanks!  Seppi  333  (Insert 2¢) 01:46, 18 November 2017 (UTC)

Can someone please add a structure diagram for indole-3-carboxaldehyde?  Seppi  333  (Insert 2¢) 01:46, 18 November 2017 (UTC)

Disambiguation links on pages tagged by this wikiproject
Wikipedia has many thousands of wikilinks which point to disambiguation pages. It would be useful to readers if these links directed them to the specific pages of interest, rather than making them search through a list. Members of WikiProject Disambiguation have been working on this and the total number is now below 20,000 for the first time. Some of these links require specialist knowledge of the topics concerned and therefore it would be great if you could help in your area of expertise.

A list of the relevant links on pages which fall within the remit of this wikiproject can be found at http://69.142.160.183/~dispenser/cgi-bin/topic_points.py?banner=WikiProject_Chemicals

Please take a few minutes to help make these more useful to our readers.&mdash; Rod talk 14:04, 3 December 2017 (UTC)

3-Methylhistidine
This article was created today and needs a structure diagram. If someone is willing to create, upload, and add it to the chembox, I'd appreciate it.  Seppi  333  (Insert 2¢) 21:41, 25 December 2017 (UTC)