Wikipedia talk:WikiProject Elements/Archive 35

Reclassifying the nonmetals: Index of discussions
This table replaces and expands the previous list. YBG (talk) 04:27, 26 April 2018 (UTC) It was restored after the bot archived it. As the index of multiple discussion threads, it should be the last one to be archived. YBG (talk) 10:21, 27 May 2018 (UTC) ——— YBG (talk) 01:02, 27 June 2018 (UTC)

Continent of stability[?]
Here. Suggests that "elements" with amu > ~300 may represent a new form of (stable) matter. Sandbh (talk) 23:03, 16 June 2018 (UTC)
 * Let me coin right here right now: "we all, already, here, live in a universe of stability". DePiep (talk) 02:06, 23 June 2018 (UTC)

Yes, I should have included a link to island of stability to make the contrasting reference to a continent of stability clearer. Sandbh (talk) 11:02, 23 June 2018 (UTC)

Transfermium Wars
Would this be a good addition to Transfermium Wars? Even if not, it should be good fun to add more details to the template. YBG (talk) 09:20, 6 June 2018 (UTC)
 * Well, it's not really a war, so not really, but I appreciate the humor behind the template :) If anything, we could use a template like Infobox historical event.--R8R (talk) 17:49, 6 June 2018 (UTC)
 * Maybe we could find a place for it next April per WP:Rules for Fools? YBG (talk) 23:09, 6 June 2018 (UTC)
 * LOL. ^_^ I guess 102 and 103 should technically be in it too, as should I guess the GSI after it got dragged into the controversy. But there is such a thing as being too clever, and adding these may make it less amusing for everyone else for the sake of sticklers. Double sharp (talk) 23:59, 6 June 2018 (UTC)
 * WP:Rules for Fools says
 * All jokes and pranks must be kept out of the "article" and "help" namespace. Jokes that affect articles, including files, categories and templates that are used in the article namespace, will be treated as vandalism.
 * But perhaps we could develop this here and then on the day, deploy it to WP:ELEM, WP:CHEM and some other spots where the audience might appreciate it. YBG (talk) 02:40, 7 June 2018 (UTC)
 * This link is about April 1st Fools Day. Methinks Cold Wars and especially Radioactive Wars are more important ;-). - DePiep (talk) 07:30, 7 June 2018 (UTC)
 * Infobox military conflict also has parameters units1, casualties1, strength1. - DePiep (talk) 07:30, 7 June 2018 (UTC)
 * Re April 1, my thought was that this would make a good April Fool's joke. Another place we could drop this template would be at WP:MILHIST. Re parameters, I've added placeholders for some of the missing parameters. Any ideas what should go in them? YBG (talk) 07:47, 7 June 2018 (UTC)


 * Found no fun in Infobox court case. Doesn't look like Roe v. Wade. You might try yourself. -DePiep (talk) 22:00, 27 June 2018 (UTC)

Single criterion definition
The background to this post is that Double sharp and I have some history discussing what is a metal.

I was recently reminded that Johnson (2007, p. 15) defines a metal as any element having an electrical conductivity at or below room temperature of at least 3 x 105 S m-1 along any direction in a single crystal of any known form of the element.

The resulting set of elements counted as metals matches the set of elements that we show on our periodic table as being metals. Johnson counts B as a nonmetal; and Si, Ge, As, Sb, Te, and At as metalloids but does not explain the difference between a metalloid and a nonmetal. That said, he is consistent with the literature in categorising Si etc as metalloids.

For a 1-criterion definition I thought his definition was reasonable.

Can anybody see any issues with it apart from maybe the arbitrariness of the conductivity value? --- Sandbh (talk) 06:04, 30 April 2018 (UTC)


 * Johnson D (ed) 2007, Metals and chemical change, RSC Publishing, Cambridge


 * Based on the data here, Johnson's definition applied literally would list phosphorus, arsenic, and antimony as metals. (I would be prepared to agree for Sb but not for P or As.) OTOH, nonmetal list disagreeing values for As and Sb that are indeed below 3×105 S/m. I still remain rather uncomfortable with this definition as it only considers physical properties without considering chemical properties; the latter are why I would be willing to call antimony a metal but not arsenic, for example. Double sharp (talk) 06:41, 30 April 2018 (UTC)
 * That is exactly why I went right to the root, chemical and physical properties of elements are both defined by their quantum natures. Plasmic Physics (talk) 06:59, 30 April 2018 (UTC) [see Density of states etc section, below]

P, As, Sb. These three all have amorphous forms (red P, black As, black Sb) which presumably fail Johnson's single crystal requirement, and the elements concerned would therefore not be counted as metals. Also, yellow As is a non-conductor. I realise the definition only considers physical properties however the correlation in this case between the physical and the chemical is close enough, in the absence of a single chemical criterion. Sandbh (talk) 11:42, 30 April 2018 (UTC)


 * As I read the definition above, it means that if any one form meets the definition, the element is counted as a metal. But in the 11:42, 30 April 2018 post above, appears to be treating the definition as requiring that every form meet the definition for the element to be counted as a metal. 14:16, 30 April 2018 (UTC)
 * Put another way, when the definition says ... of any known form of the element, does that mean "of every known form of the element" or does it mean "of at least one known form of the element". It seems to me that the English could be read in either way. What does Johnson mean? YBG (talk) 02:22, 1 May 2018 (UTC)
 * That would make every element a metal due to pressure induced metalisation. I would recommend only applying it to the standard form of the element. Plasmic Physics (talk) 05:27, 1 May 2018 (UTC)

What did Johnson mean? From having corresponded with Johnson, his requirement for an element to be regarded as a metal was that all crystalline forms of the element had to exceed the minimum conductivity requirement in every direction. My reference to P, As, and Sb was kind of wrong. Their amorphous forms may lack long range order but Johnson responded: "does not the term amorphous mean only that the crystals are too small to be exploited in X-ray studies?" Arguably then, P, As, and Sb are still not metals. Sandbh (talk) 10:31, 1 May 2018 (UTC)
 * Thank you, that helps clarify things. So then any element having an electrical conductivity at or below room temperature of at least 3 x 105 S m-1 along any direction in a single crystal of any known form of the element means "any element having an electrical conductivity at or below room temperature of at least 3 x 105 S m-1 along any every direction in a single crystal of any every known crystaline form of the element". This gives rise to two questions: (1) Does my rephrasing represents an accurate reflection of Johnson's intent? (2) Does at or below room temperature mean "at every temperature at or below room temperature" or "at some temperature at or below room temperature"? Thank you for your patience. My questions probably expose my lack of big-picture context in the physical properties of metals; no doubt, most of what I see as ambiguities are clarified in an obvious way for those with greater implicit understanding. YBG (talk) 11:57, 1 May 2018 (UTC)
 * Yes to both questions. Sandbh (talk) 00:44, 2 May 2018 (UTC)

Status of Bi and Sb

 * I've argued for calling bismuth a metalloid before on your talk page User talk:Sandbh. ^_^ For all that I wouldn't support it on WP (because it's not a majority classification), I am quite in agreement with Droog Andrey saying "you are right that Sb and Bi are not true metals, but at least they are much closer to metals than to non-metals" above. Double sharp (talk) 15:55, 1 May 2018 (UTC)

My impression is that the literature views Sb as the most metallic of the metalloids whereas Bi is viewed as the weakest of the metals.

As far as I can tell this is largely based on the nature of the oxides going down group 15, N acidic; P acidic; As acidic or amphoteric; Sb amphoteric; Bi basic. A similar trend is seen in comparing the sulfides of As, Sb, and Bi. Certainly there are more nuanced aspects of Bi that give one pause for thought (e.g. its semimetal band structure) but these appear to be seen to be outweighed, overall, by its chemically basic nature. Along these lines, in that thread on my talk page, the bulk of my observations and quotes from the literature addressed more headline properties (acid/base character; overall chemical behaviour; the aforementioned sulfides; and oxides) whereas your observations tended to address nuanced aspects of these observations. Your comments weakened the case for treating Bi as a metal but not to the point, IMO, of moving it into full-blown metalloid territory---e.g. I think the ionisation energy of Bi is too low, and Bi does not have any bulk semiconducting allotropes unlike the other metalloids.

I kind of agree with what Droog Andrey said although I have trouble reconciling his view about Sb being closer to the metals than the non-metals, given most Sb chemistry is non-metallic e.g. ‘Although the cationic tendencies of Sb(III) suggest metallic behavior, most of the chemistry of antimony is characteristic of a non-metal…’ (Jolly WL 1966, The chemistry of the non-metals, Prentice-Hall, New Jersey, p. 107). Mind you the band structure of antimony does place it closer to the metals than the nonmetals but that is just one factor. OTOH, graphite, as a semi-metal in the physics-based sense, might be said to be closer to the metals than nonmetals whereas, for example, the chemistry of carbon is closer to the nonmetals. Sandbh (talk) 00:34, 2 May 2018 (UTC)

Density of states etc
Perhaps we should determine metallicity by the overlap of the density of state distributions for the valence and conduction band with respect to the Fermi level. i.e. If the overlap is between 67 and 100% (metallic), 33 and 67% (semi-metallic), 0 and 33% (non-metallic)? Presto - non-arbitrary! Plasmic Physics (talk) 06:36, 30 April 2018 (UTC)


 * This is problematic since it would result in Bi being treated as a nonmetal. Sandbh (talk) 11:42, 30 April 2018 (UTC)

Not band overlap, but density of states overlap. Even if it this results in its treatment as a semi-metal or non-metal, it doesn't seem unreasonable to me as it has some properties that conforms to those descriptions. Plasmic Physics (talk) 05:27, 1 May 2018 (UTC)


 * Is the semimetal article wrong? It says, "Metals have a partially filled conduction band. A semimetal is a material with a very small overlap between the bottom of the conduction band and the top of the valence band. A semimetal thus has no band gap and a negligible density of states at the Fermi level. A metal, by contrast, has an appreciable density of states at the Fermi level because the conduction band is partially filled.[1]" These words appear to match with your suggested definition. However I'd equate arguing Bi to be categorised as non-metal, which is what your suggested definition would result in, as a bridge too far. The best we could get to would be to categorise it as metalloid however very few authors do that---about 1 in 20 according to list of metalloid lists. And the prospect of group 15 going N: non-metal, P: non-metal, As: metalloid, Sb: metalloid, Bi metalloid, does not fill me with confidence. Sandbh (talk) 11:01, 1 May 2018 (UTC)

Solid state physics
In solid state physics, a metal is a material with a Fermi surface. That definition is mainly used in my table. Fermi surfaces for most of the elements are shown there; see also [10.1002/pssb.2221170122] for Sb, Bi and [10.1016/0031-9163(66)90852-3] for Sn. Droog Andrey (talk) 07:51, 4 May 2018 (UTC)
 * I understand that arsenic has a Fermi surface as does graphite? Sandbh (talk) 12:08, 4 May 2018 (UTC)
 * Well, Droog Andrey says mainly, so clearly he must have some provisions (probably chemical ones) that exclude C and As. Going purely by physical properties is problematic because the gradation metal–semimetal–semiconductor–insulator is not always followed exactly when moving right on the periodic table: germanium is a semiconductor while arsenic is a semimetal, and bismuth is a semimetal while polonium and probably astatine are true metals. Here are two papers about the Fermi surfaces of the metals: 10.1080/00018736900101387 (s- and p-blocks), 10.1080/00018737100101211 (d- and f-blocks). Double sharp (talk) 16:07, 4 May 2018 (UTC)
 * Yes, they have, but both C and As become semiconductors when amorphized, that's the reason to exclude them from metals. Droog Andrey (talk) 11:15, 6 May 2018 (UTC)
 * Antimony becomes a semiconductor when amorphized (= black antimony). Presumably then you should show antimony as a nonmetal, too? Sandbh (talk) 13:40, 6 May 2018 (UTC)
 * yes, antimony is the first candidate to exclude from the list of metals. But black antimony lacks stability; also, the most of russian sources treat Sb as a metal. So it is still here. Droog Andrey (talk) 19:17, 8 May 2018 (UTC)
 * [ How do most Russian sources treat germanium? Sandbh (talk) 12:14, 13 May 2018 (UTC)]
 * Ge is classified as a metal only in schoolbooks (just to draw a simple stairline). Serious literature doesn't count it as a metal; various terms (semimetal, semiconductor, ...) are strongly supported by applications of germanium (microelectronics, lenses for infrared optics, etc.) Droog Andrey (talk) 19:33, 13 May 2018 (UTC)
 * I wonder how much of this problem is because physical properties end up with Ge and As "swapped"; As is a semimetal, while Ge is only a semiconductor. Even chemically, As is also more comfortable in the +3 than the +5 state, whereas Ge prefers +4 to +2, which raises the chemical metallicity of As (which is still not that great, since all the basic properties of As2O3 can be rationalised away as alcoholic) and lowers that of Ge (because the state where it has a fighting chance to show some metallic properties is easily oxidised to the state where it doesn't). Double sharp (talk) 00:01, 14 May 2018 (UTC)
 * That is quite interesting. Sources differ as to the stability of black antimony. Our own article on antimony says that the black form transforms to grey antimony above 100 °C (citations included). OTOH Wiberg's Inorganic Chemistry says the transformation starts even at 0° C. This article talks about using black antimony in a semiconducting bolometer so presumably its instability is not an issue, although I don't know what temperature the bolometer is supposed to be maintained at. In this vein it is weird that white phosphorus is treated as the standard form of P even though it is metastable. I guess most Russian sources treat Sb as a metal since Mendeleev did too? If so, I'd find that odd given antimony's mostly nonmetallic chemistry. Sandbh (talk) 03:11, 9 May 2018 (UTC)
 * I'd say the chemistry of antimony is right on the edge between metallic and nonmetallic. I'd even say that there's some kind of parallel between Sb and Re (both are just before the middle of the block). Yes, white phosphorus is metastable as well as diamond, for example, but they are both much more stable than black antimony. Droog Andrey (talk) 10:17, 9 May 2018 (UTC)
 * I guess I should go look at the 4d and 5d metals again, since they are the ones who make cation formation a problematic criterion. ^_^ Double sharp (talk) 00:05, 14 May 2018 (UTC)
 * Aren't arsenic and graphite's Fermi surfaces anisotropic though? Anisotropy gives rise to only partial metallicity. Plasmic Physics (talk) 14:06, 5 May 2018 (UTC)
 * I don't know about graphite but Sb and Bi have anisotropic Fermi surfaces too. It looks to me that the Fermi surface criterion is not that helpful, at least not consistently so. Sandbh (talk) 00:44, 6 May 2018 (UTC)
 * Why not just specify that a metal is a material with an isotropic Fermi surface, albeit with a list of exceptions? I don't see why bismuth is considered to be fully metallic, instead of semi-metallic, other than the plain reason of popular convention. Plasmic Physics (talk) 02:13, 6 May 2018 (UTC)
 * I like your approach of giving the general rule, and then listing the exceptions and explaining why the exceptions are still counted as metals. Bi is considered to be metallic due to its basic oxide. Sandbh (talk) 13:50, 6 May 2018 (UTC)
 * That seems like a completely arbitrary, or is it 'coincidental', property. Plasmic Physics (talk) 14:30, 6 May 2018 (UTC)
 * Well, physicists do not have a monopoly on the term metal. It seems to me that chemistry has a right to have a seat at the table negotiating the partition of the periodic table between metals and nonmetals, with a buffer zone between the two; and after all, we have metalloid as a term that specifically focuses on chemistry, rather than the physics-based term of semimetal. I am unfortunately facing a lack of time right now; I'll gather some thoughts about this soon. Double sharp (talk) 15:28, 6 May 2018 (UTC)

Two criteria definition

 * heading added YBG (talk) 06:26, 7 May 2018 (UTC)

It is proposed to define a metal with a single-criterion definition and then list the exceptions, along with a reason for in(ex)cluding them despite the definition. My question: is this appropriate? My answer: it depends on the definition of "definition". If by definition we mean a description that defines membership in a class, then this is not acceptable, in fact, it is patent nonsense. But if by definition we mean a general description, a first-approximation rule-of-thumb, then this is entirely appropriate. In this second sense, it would be perfectly appropriate to say that birds fly and mammals bear live young, and go on to discuss platapi, echidna, and bats. But this is properly a description not a definition. We are saying what is generally true and go on to discuss the exceptions to the general rule. Thus, we can say "metals are substances that generally ____" (descriptive) but if there are exceptions we probably should not say "metals are substances that _____" (definitive). Along these lines, it would be ideal if we could have two criteria, C1 and C2, and say that "metals are generally substances which satisfy both C1 and C2" and "nonmetals are generally substances that satisfy neither C1 nor C2", provided ALL elements that satisfy both C1 and C2 are indisputably metals, and all elements that violate both C1 and C2 are indisputably nonmetals. We could then go on to discuss the smallish set of elements that satisfy one of the criteria but violate the other one, and explain why each of those anomalous elements should be considered a metal or a nonmetal. YBG (talk) 23:36, 6 May 2018 (UTC)


 * I'm juggling a few things at the moment, so here are some quickish comments.
 * YBG, I tend to agree. Looking at historical conceptions of metals it may be possible to generally define a metal as any element that is relatively malleable (in its most stable form in ambient conditions). The exceptions are those elements that nevertheless have a high strength to weight ratio e.g. Be, Cr, or high density (> Fe) e.g. Os, Bi; or Hg which is a liquid but malleable when solid. I know this conception lacks any chemical properties however I am trying for a general definition that a person in the street would understand ("cation formation" and "basic oxide" don't mean anything to these people). I suspect that boron may trip me up if it has a high strength to weight ratio in which case I would need to add "good 3D electrical conductivity" as a mandatory requirement for any exceptional element, in addition to high strength or density.
 * I need to double-check all the brittle metals to make sure they are all picked up by this general definition. I may still get tripped up.
 * I appreciate the fact that the expression "relatively malleable" lacks precision, but submit that this is not an issue at a general level of abstraction.
 * The brittleness of zinc was troubling me for a while until I found a reliable source noting this is not the case wrt highly pure zinc. It would be good to find some more sources confirming this.
 * Incidentally, I lifted a cast iron bowl yesterday and was reminded how truly dense "mere" iron is. Iron is flipping heavy. The discovery of metals that floated on water must have astonished the people of the day. Sandbh (talk) 02:47, 7 May 2018 (UTC)


 * Did my C1/C2 discussion make sense? What would be the C1 and C2 be that would break the elements into four categories as follows
 * {| class=wikitable

! C1 !! C2 !! Description ! Yes !! Yes ! No !! No ! Yes !! No ! No !! Yes
 * style="background-color:;" | Elements that satisfy both C1 and C2 are, without exception, metals - and this group constitutes the bulk of the metals
 * style="background-color:;" | Elements that violate both C1 and C2 are, without exception, nonmetals - and this group constitutes the bulk of the nonmetals
 * The relatively few elements that satisfy C1 and violate C2 are classified as either metal or nonmetal based on other factors.
 * The relatively few elements that violate C1 and satisfy C2 are classified as either metal or nonmetal based on other factors.
 * }
 * Here's another way of looking at it:
 * {| class=wikitable

! Criteria ! C2 satisfied ! C2 violated ! C1 satisfied ! C1 violated
 * style="background-color:;" | Elements that satisfy both C1 and C2 are, without exception, metals - and this group constitutes the bulk of the metals
 * The relatively few elements that satisfy C1 and violate C2 are classified as either metal or nonmetal based on other factors.
 * The relatively few elements that violate C1 and satisfy C2 are classified as either metal or nonmetal based on other factors.
 * style="background-color:;" | Elements that violate both C1 and C2 are, without exception, nonmetals - and this group constitutes the bulk of the nonmetals
 * }
 * Or perhaps this way, with criteria M and N
 * {| class=wikitable

! Criteria ! Criterion N violated ! Criterion N satisfied ! Criterion M satisfied ! Criterion M violated
 * style="background-color:;" | (+M, -N) Elements that satisfy criterion M and violate criterion N are, without exception, metals - and this group constitutes the bulk of the metals.
 * (+M, +N) The relatively few elements that satisfy both criteria, M and N, are classified as either metal or nonmetal based on other factors.
 * (-M, -N) The relatively few elements that violate both criteria, M and N, are classified as either metal or nonmetal based on other factors.
 * style="background-color:;" | (-M, +N) Elements that violate criterion M and satisfy criterion N are, without exception, nonmetals - and this group constitutes the bulk of the nonmetals.
 * }
 * Am I making any sense here? YBG (talk) 04:17, 7 May 2018 (UTC)
 * Hmm. If we split the elements this way, are there still going to be metalloids as a category? I'm not opposed to a pure metal–nonmetal distinction, but first we have to decide if that is the way we are going to go. Double sharp (talk) 05:06, 7 May 2018 (UTC)
 * Re metalloids and two-criteria definitions. Several ways to go here.
 * Define "metalloid" to include every element that falls into one of the two ambiguous cells (+M, -N) or (-M, +N) and those elements only. Highly unlikely this would work out, but if it does, it would be great
 * Change the text in cells (+M, +N) and (-M, -N) to read "... are classified as either metal, metalloid, or nonmetal based on other factors". This seems the best way forward if metalloid is a full-fledged supercategory parallel to metal and nonmetal.
 * Consider metalloid to be a subcategory of nonmetal and leave the wording on the chart unchanged.  This, I believe, is the direction that Sandbh's thinking is heading, and it seems like a good idea to me if (and this is a big if) it receives broad support.
 * Here is yet another way to think about my criteria M and N
 * M is a property shared by most (but perhaps not all) metals, but not held by any of the nonmetals or metalloids.
 * N is a property shared by most (but perhaps not all) nonmetals, but not held by any of the metals.
 * So are there properties M and N that satisfy these conditions? YBG (talk) 06:26, 7 May 2018 (UTC)
 * I reckon M could be “has a basic oxide” and N could be “a packing efficiency of less than 35%”. Sandbh (talk) 12:06, 13 May 2018 (UTC)

Street level, generic, and specific definitions

 * I must confess that I don't think that we need to make the definition understandable to a person on the street. For one thing, I suspect most of the metallic elements are not known to the person in the street anyway, and those whose names are known are rarely encountered as the metal. I think that the average person on the street would be very surprised by pure sodium or even pure gold. But more importantly, the concepts that make us consider elements as more or less metallic mostly do not mean anything to a person on the street, and I don't see why we should exclude most of them as factors because of that. That would give us a very shallow set of criteria that focuses on the metals most laypeople have encountered as elements, which often are not very typical. Double sharp (talk) 05:37, 7 May 2018 (UTC)
 * It would be useful to start with the most generic/historical "definition" of a metal, followed by the discipline specific definitions that emerged more recently in history and which have been more technical. I think that is the nature of the beastie. Something like:
 * A metal is any element, alloy, or compound that is a good conductor of electricity and (a) is relatively easy to hammer into thin sheets by hammering or rolling; or (b) has a high strength or density.


 * In astronomy, a metal is any chemical element with an atomic number greater than that of helium.


 * In chemistry, a metal is any chemical element that forms a simple cation in aqueous solution, or which meets the general definition given above.


 * In physics, a metal is any substance in which the valence and conduction bands overlap or touch.


 * Note that As and Sb don't meet the definition of a metal in chemistry but that they do in physics. Sandbh (talk) 12:11, 7 May 2018 (UTC)


 * Sb simple cations are way more stable in aqueous solutions than, say, tungsten cations. Droog Andrey (talk) 19:28, 8 May 2018 (UTC)


 * There is no simple Sb cation. If there was, then I would probably classify Sb as a metal. The stable form of Sb(III) in aqueous solution is an incomplete hydrocomplex [Sb(H2O)4(OH)2]+. W does not form simple cations in aqueous solution, as far as I know. Sandbh (talk) 08:39, 9 May 2018 (UTC)


 * pH decides here. Yes, [Sb(OH2)6]3+ is significantly deprotonated even at pH = 0, but at least it behaves not so far from [Fe(OH2)6]3+, for example. Look at solutions of antimony perchlorate. As for tungsten, its simple cations are indeed exotic and questionable (because of cluster formation), but that doesn't make tungsten non-metallic :) The same holds for protactinium: its (+5) cations are hydrolysed way deeper than Sb(+3), but Pa is a true metal. Droog Andrey (talk) 10:31, 9 May 2018 (UTC)


 * OK, so you agree Sb does not form a simple cation in aqueous solution. I remember reading one or more obscure papers on antimony perchlorate but I could never find anything therein worth citing to support the notion of antimony as a metal. At least W and Pa are strong or have a high density, and these are attributes traditionally associated with metals. I am beginning to doubt that there is an inclusive definition of a metal based solely on chemical properties. Sandbh (talk) 11:08, 9 May 2018 (UTC)


 * I don't agree that aqueous Sb3+ doesn't exist, but I agree that it is highly hydrolyzed. I'd never use high density itself as a metallicity mark; yes, metallic bond has dense packing, but density does also depend on atomic weight and atomic radius. As for chemical definitions: some are proposed below. Droog Andrey (talk) 11:27, 9 May 2018 (UTC)
 * I could very well believe that Sb(ClO4)3 at least notionally contains Sb3+. The homologous Bi(ClO4)3·5H2O is well-known (see p. 620 of the Pergamon volume on As, Sb, and Bi) and not only dissolves in water to give a colourless solution containing bismuth oxycations, but also does not readily precipitate basic salts like bismuth nitrate and sulfate do (antimony sulfate does the same as its bismuth homologue). So if we are looking for a real salt of antimony, perchlorate seems to be a very good candidate, as do other salts of weakly complexing anions which are the conjugate bases of strong acids (e.g. permanganate). The notes in Metalloid certainly note the existence of antimony cations and oxocations like [Sb(H2O)4(OH)2]+ and SbO2+ in very acidic solution, so I could also very well believe in Sb3+ and Bi3+ even if they are going to be highly hydrolysed at any normal pH.
 * What's your opinion on possible Ge2+ and As3+ cations?
 * Appealing to strength or high density as one of the criteria puts us into a difficult situation with metals like sodium, and additionally does not seem to be that useful a criterion since there is not much that is categorically different about the essentially cationic chemistry of Na or the electrical conductivity of Na metal from that of the denser and stronger Fe, for instance. Double sharp (talk) 14:53, 9 May 2018 (UTC)
 * P.S. I find it strange to exclude that Sb cation as being an incomplete hydrocomplex. Apart from really weakly polarising cations like Na+, surely most metal aqua cations will undergo incomplete hydrolysis and form incomplete hydrocomplexes; I already learnt that in high school for the explanation of why salts like MgCl2 and AlCl3 dissolve in water to form acidic solutions, as you can think of it as the aqua cation protonating a water molecule. The Bi3+ aqua cation also only exists in very strongly acidic solution and readily hydrolyses to oxo-hydroxo complexes even with strong acid in excess at 1 M. Double sharp (talk) 15:27, 9 May 2018 (UTC)
 * As3+ is about as exotic as B3+. Speculations about it go mostly from water-soluble chloride complexes. Ge2+ seems to be less questionable, but still doubtful because of too low solubility of Ge(OH)2 in concentrated HClO4. Droog Andrey (talk) 21:43, 9 May 2018 (UTC)
 * OK, thanks for this explanation. Are oxocations like AsO+ any more likely for arsenic? Double sharp (talk) 23:42, 9 May 2018 (UTC)
 * AFAIK, [10.1179/cmq.2004.43.4.439] is the only source that mentions it. But, again, I'd prefer perchloric acid media :) Droog Andrey (talk) 04:24, 10 May 2018 (UTC)
 * OK. Sorry to be asking all these questions, but I still have one more. ^_^ What's your opinion of the oxoacid salts of Ge? Greenwood and Earnshaw mention Ge(OAc)4 and the unstable Ge(SO4)2; Metalloid mentions among others also Ge(ClO4)4. Double sharp (talk) 16:39, 12 May 2018 (UTC)
 * They seem to be covalent compounds like boron perchorate. Droog Andrey (talk) 20:44, 12 May 2018 (UTC)
 * OK, but can't the same criticism be applied to those Sb salts? Sb2(SO4)3 has a structure of corner-sharing SO4 tetrahedra and SbO3 forming an infinite ladder (see antimony sulfate) and is like a mixed oxide; its main claim to salthood seems to be that it forms when you react Sb with sulfuric acid (but then reacting Sb with nitric acid does not produce the nitrate). As Axiosaurus noted at Talk:Metalloid/Archive 1, Sb(ClO4)3 reacts with LiClO4 to form LiSb(ClO4)4 or Li2Sb(ClO4)5, suggesting that these "salts" have significant covalent character. Double sharp (talk) 07:41, 13 May 2018 (UTC)
 * Yes, that is. Sb(III) and Ge(II) aquacations are nearly equally unstable. But antimony's larger charge put it under heavier conditions, so that competition hardly seems fair. You see, Th(IV) is hydrolyzed much deeper than Al(III), but Al is much closer to non-metals than Th. Droog Andrey (talk) 10:56, 13 May 2018 (UTC)
 * Well, yes, but where do you draw the line then? Of course all +4 cations are going to be more heavily hydrolysed than +3 cations, and for +5 cations even Pa5+ is significantly hydrolysed already. So what is it that distinguishes Sb(OH)2+ from Te(OH)3+, since obviously Te4+ would be under heavier conditions than Sb3+ due to its higher charge? (I would say that it's because solvated Sb3+ is calculated to persist under favourable conditions without immediately deprotonating a water molecule, while Te4+ can't do that, but that ends up allowing in Ge2+ as well.) Double sharp (talk) 23:54, 13 May 2018 (UTC)
 * P.S. And the Ge "salts" I mentioned are all GeIV and not GeII, so it seems to me that SbIII has an easier time than GeIV. After all, Ge4+ is calculated to hydrolyse immediately (like Te4+), while Sb3+ is calculated to be persistent. Double sharp (talk) 07:25, 14 May 2018 (UTC)
 * I think that admissible degree of hydrolysis for metal cations should depend on their charge. For M+ the deadline lies probably near At; for M2+ near Sn; for M3+ near Sb. Larger charged cations are deeply hydrolysed even for active metals like Th, so the concept itself becomes useless. Droog Andrey (talk) 16:23, 14 May 2018 (UTC)
 * The chemistry definition needs another chemical property criterion so that we can get rid of the default back to the general definition bit. I'm a little perplexed that I cannot recall a purely chemistry based definition of a metal. Sandbh (talk) 00:15, 8 May 2018 (UTC)

Neighbour-based definition

 * heading added Sandbh (talk) 00:04, 17 May 2018 (UTC)


 * I propose the following definitions:
 * Chemical element is called a metal if it forms metallic chemical bonds with the most of its neighbours in the periodic table.
 * Chemical element is called a reactive non-metal if forms covalent or ionic chemical bonds with the most of its neighbours in the periodic table.
 * Chemical element is called a noble gas if it doesn't form any chemical bonds with the most of its neighbours in the periodic table.
 * Droog Andrey (talk) 10:45, 9 May 2018 (UTC)
 * We can also add the element itself to its eight neighbours for tie-break. Droog Andrey (talk) 21:43, 9 May 2018 (UTC)
 * How would you determine if an element forms a metallic bond with (a) one of its neighbours; (b) itself? Sandbh (talk) 11:17, 10 May 2018 (UTC)
 * There are 9 elements in most cases: 8 neighbours and an element itself. If more than 4 form a metallic bonded compounds with our element, then it is a metal. Droog Andrey (talk) 17:07, 10 May 2018 (UTC)
 * And how does one define neighbor? Although I don't think it would make any difference in the context of this proposed definition, there are a number of edge cases. H, B, Al, Y, and Og all cry out "And who is my neighbor?" YBG (talk) 13:09, 10 May 2018 (UTC)
 * In my opinion, we should ignore gaps when determining what neighbours are, so both strontium and zirconium may be considered neighbours of yttrium. I'd also consider that the noble gases neighbour both the alkali metals and the halogens, so that both iodine and caesium are neighbours of xenon. Of course, elements in the first and seventh periods will still have less than eight neighbours, but I don't think that's a problem. Double sharp (talk) 13:33, 10 May 2018 (UTC)
 * I think we shouldn't stick together opposite sides of the table. Otherwise Li, for example, will fall into uncertainty with 3 metallic (Be, Na, Mg) and 3 non-bonding neighbours (He, Ne, Ar). So, the neighbours of H are He, Li, Be; the neighbours of B are Be, Mg, Al, Si, C; the neighbours of Al are Be, B, C, Mg, Si, Zn, Ga, Ge; the neighbours of Li are H, Be, Na, Mg; and so on. Droog Andrey (talk) 17:07, 10 May 2018 (UTC)
 * Wouldn't Li itself break the tie between non-bonding and metallic-bonding? YBG (talk) 04:26, 11 May 2018 (UTC)

I propose the following definitions for the eight immediate neighbors (where by "immediately left (right)" I mean "having one less (more) proton"): This guarantees eight neighbors for all except H (3), He/B/C/N/O (5), Li/Be/F (6), Ne/Rn (7), Fr (6), Ra-Ts (5), Og (4). I'm being a big OCD about having diagonals defined whenever possible and always unambiguous. I don't think it really matters because I'm quite certain that whenever a diagonal element is potentially ambiguous or missing, I don't think it would make any difference in the definition. YBG (talk) 04:26, 11 May 2018 (UTC)
 * 1) W: element immediately left of the element (exists for all except H)
 * 2) E: element immediately right of the element (exists for all except Og)
 * 3) N: element immediately above the element (exists for all except eight: H, He, Be, B, C, N, O, F)
 * 4) S: element immediately below the element (currently ambiguous for Y; exists for all up thru Rn)
 * 5) NW: element immediately left of N or immediately above W (identical if both exist) (exists for all but eight: H, He, Li, B, C, N, O, F)
 * 6) NE: element immediately right of N or immediately above E (identical if both exist) (exists for all but seven: H, He, Be, B, C, N, O)
 * 7) SW: element immediately left of S (exists for all up thru Rn) (not always the same as below W)
 * 8) SE: element immediately right of S (exists for all up thru At) (not always the same as below E)
 * I think there are many more elements with less than eight neighbours; the 3d and 4f rows certainly have no elements above them. Of course this is just quibbling, since the definition itself is sound. ^_-☆ Double sharp (talk) 07:32, 11 May 2018 (UTC)
 * Well, to avoid overloading the definitions with tons of formal details, let's just take an element itself and its two closest neighbours in the sense of atomic number.
 * So we will always have three elements, and uncertainly could only appear if we encountered "no bonding", "metallic bonding" and "covalent/ionic bonding" in the same set. But that never happens (thanks to metallic At for that). Droog Andrey (talk) 12:41, 11 May 2018 (UTC)

p-block chemistry

 * heading added Sandbh (talk) 23:58, 16 May 2018 (UTC)

I wonder if we might not get another hint by looking at whether and how readily the p-block elements dissolve in acid or alkali (aqueous or molten). Double sharp (talk) 01:33, 15 May 2018 (UTC)

Summary of the reactions for most of the group 13 through 16 elements (mostly from Greenwood and Earnshaw's book and Ullmann's encyclopaedia):

Group 13

 * Boron is not attacked even by molten NaOH at 500 °C. Boron is not attacked by non-oxidising acids but will dissolve in a 2:1 mixture of hot concentrated H2SO4 and HNO3.
 * Aluminium is blocked from reacting much with dilute acids by its oxide layer, but destroying this layer allows the reaction to proceed. It also dissolves in hot concentrated HCl or aqueous NaOH or KOH at room temperature. Gallium is similar, but indium does not dissolve in aqueous alkali.

Group 14

 * Carbon as graphite is attacked by hot concentrated nitric acid to form mellitic acid (preserving the hexagonal structure).
 * Silicon will not react with aqueous acids, but will be oxidised by a mixture of concentrated nitric and hydrofluoric acids (because HF will etch away the SiO2 layer formed by oxidation, allowing the HNO3 to continue doing its work). It reacts violently with dilute aqueous alkali.
 * Germanium will react with hot concentrated sulfuric or nitric acids, but will not react with dilute acids or alkalis without the presence of an oxidising agent. It will react violently with molten alkali.
 * Tin will react with dilute nitric acid as well as hot concentrated hydrochloric and sulfuric acids, forming SnII compounds. OTOH, hot aqueous alkali yields SnIV compounds.
 * Lead will react slowly with aqueous hydrochloric acid and quickly with nitric acid.

Group 15

 * White phosphorus (but not red) will readily react with aqueous alkali, disproportionating to phosphine and hypophosphite. It reduces sulfuric acid to SO2 and nitric acid to nitrogen oxides.
 * Arsenic is not noticeably attacked by non-oxidising acids. Reaction with dilute nitric acid gives arsenious acid, concentrated nitric acid gives arsenic acid, and hot concentrated sulfuric acid gives arsenic trioxide. Arsenic will not readily react with aqueous alkali but will react with molten NaOH, or boiling NaOH in the presence of air.
 * Antimony is less reactive than arsenic. Dilute acids will not react with it, but concentrated oxidising acids (including aqua regia) will attack it, forming the oxoacid salt Sb2(SO4)3 upon reaction with sulfuric acid. It will only react with molten NaOH or KOH at red heat.
 * Bismuth will react with acids to form oxoacid salts.

Group 16

 * Sulfur is oxidised by nitric acid (or aqua regia, or indeed HCl in the presence of an oxidising agent) to form sulfuric acid, and concentrated sulfuric acid at 200 °C to give sulfur dioxide. It will react with NaOH (products vary with conditions).
 * Selenium is not attacked by dilute HCl and other nonoxidising acids, but will react with HNO3 to form selenous acid and H2SO4 to form the polymeric selenium cation Se82+. It dissolves in strong alkali, disproportionating into selenide and selenite.
 * Tellurium dissolves somewhat in dilute HCl in the presence of air and readily in concentrated sulfuric or nitric acid. It will also react with caustic alkali to form tellurites, but its solubility depends on temperature.
 * Polonium dissolves readily in dilute HCl. It is only slightly soluble in alkali.

Analysis and conclusion
I think it's reasonable to say that antimony is the closest to metals of the bunch, being the only one that actually forms "salts" upon reaction with acid (the others seem to form oxides, oxoanions, or oxoacids if I am reading this right, although I'd like to see more detail about germanium). Perhaps the next in line would be germanium and tellurium. Given this, I think it's quite reasonable to bring Sb over to the metallic side and leave all the rest as nonmetals. Double sharp (talk) 15:39, 16 May 2018 (UTC)

Amphoteric line

 * Oh, and we get a similar result (of course) from just considering the "amphoteric line" (with Al and Sn as high-school examples) and seeing where its rightmost edge is – the last elements where the most stable oxidation state is still amphoteric. We can draw this line at HI, BeII, AlIII, GaIII, SbIII, AtIII, and probably OgIV, and a nice metalloid line comes to the right of these elements. (Well, except for putting hydrogen on the metallic side – but hydrogen has some really very metallic properties. Does it form a cation in aqueous solution? Well, would we consider H3O+(H2O)6 to be an aqua cation? It looks like one, but it doesn't really make sense to consider H3O+ to be H(H2O)+! At least I'd say it's a wannabe metal. ^_^) Double sharp (talk) 15:51, 16 May 2018 (UTC)
 * GeO2 and TeO2 are amphoteric. In the literature, arsenic trioxide is usually regarded as being amphoteric. Only a few sources describe it as instead being weakly acidic. Sandbh (talk) 23:43, 16 May 2018 (UTC)
 * Yes, but they have significantly more acidic properties than basic properties. I should probably have specified that the limit ought to be when they are about equal, because if we're going for a binary division, the limit ought to be between predominating basic and predominating acidic properties, not between having basic and not having basic properties (which moves the line far too much to the right). Double sharp (talk) 00:25, 17 May 2018 (UTC)
 * I agree that on this basis, the metalloids are B, Si, Ge, As, and Te. I don't have enough information about At and Og to make a call, noting At has been predicted to be a true metal. I guess Sb is nevertheless commonly regarded as a metalloid in the literature on the basis of its lack of appreciable cationic chemistry and its covalent bonding structure. Sandbh (talk) 12:33, 17 May 2018 (UTC)
 * I don't think Sb has a lack of appreciable cationic chemistry; the simple Sb3+ cation may not show up very often, but Sb(OH)2+ (hydrated SbO+) certainly does: see the antimony Pourbaix diagram. And if we excluded Sb on these grounds we should have to exclude Bi as well. But I am increasingly thinking that pointing to one criterion and one criterion alone is folly and that we must simply look at what the general lay of the land is like in the descriptive chemistry of each element.
 * As for astatine, one tantalising note is that the structure of the At aqua cation may well be At(H2O)+ with only one solvating water molecule (see the Gmelin volume on At, p. 220); as such it can be considered protonated hypoastatous acid, which indicates amphoteric properties for AtOH. And given that oxides and hydroxides are more acidic in higher oxidation states, AtO(OH) should at least be amphoteric as well, and indeed it can be protonated to form AtO+ in acidic solutions and deprotonated to form AtO(OH)2− in alkaline solutions. Oganesson indeed has scant predicted information; the best we can do at the moment is extrapolate from Rn (since solvated I+ exists in pyridine solution and At+ exists in aqueous solution, we may extrapolate that since Rn2+ exists in halogen fluoride solution, Og2+ may exist in aqueous solution), and making analogies to Sn on the grounds that significant ionicity is predicted for not only OgF2 and OgF4. If Ts3+ is already expected to follow Au3+, then it should be amphoteric, and I would have a hard time then believing that Og2+ with a lower charge could be acidic; it would remind me of the cases of Ga3+ and Ge2+. But I agree that we lack enough information, even predicted information, to decide if the metalloid line should be drawn before or after oganesson. Double sharp (talk) 15:34, 17 May 2018 (UTC)
 * I may nevertheless have to look up protonated hypoiodous acid, given that HIO also has a high enough pKa that amphotericity is reasonable; it also seems to analogously exist. Concurrently I shall also have to look up all the scraps of information I can find about the solvation of the aqueous At+ cation, including whether it is At(H2O)+ or At(H2O)2+. Double sharp (talk) 04:25, 18 May 2018 (UTC)

I am uncomfortable with calling As2O3 an amphoteric oxide, despite its reaction with hydrochloric acid, because in water it dissolves to form arsenious acid – which, I will grant you, is still a weak acid that does not dissociate completely, but with a pKa of 9.29 is of similar strength to boric acid. As such I think the reaction is not well-classed as an acid–base reaction and I would agree with those authors who treat it as analogous to nucleophilic substitution of alcohols: As(OH)3 + 3 HCl → AsCl3 + 3 H2O. The same can be said of GeO2, which dissolves in water to form germanic acid: the reaction with hydrochloric acid is Ge(OH)4 + 4 HCl → GeCl4 + 4 H2O. I'm on my phone now, but looking up a few pKa's online here suggests that this argument may take care of tellurium as well. Double sharp (talk) 02:21, 17 May 2018 (UTC)

acidity decreases nearly this way: TeO2 >> GeO2 ≈ B2O3 > As2O3, and the latter is definitely on the acidic side. Of course, we could even call N2O5 amphoteric (because of nitronium salts), but the term "amphoteric" would become useless then. Droog Andrey (talk) 09:27, 17 May 2018 (UTC)
 * We can refer to the pKa chart I linked above to demonstrate this as well: SiO(OH)2 9.91, As(OH)3 9.29, B(OH)3 9.24, GeO(OH)2 9.01. We then get a big gap between that and CO(OH)2 at 6.35 and TeO(OH)2 at 6.27, and then an even bigger gap going down to SeO(OH)2 at 2.62 and PO(OH)3 at 2.16, before dropping below zero for the textbook strong acids made from the strongest nonmetals. Double sharp (talk) 14:01, 17 May 2018 (UTC)
 * Incidentally, the pKa of Sb(OH)3 is 11.0. It really is a feeble acid, and indeed at you will already get Sb(OH)2+ in acidic solution and Sb(OH)4− in alkaline solution without needing to go past pH 0 or 14; that's much better than arsenic, for which all cationic forms are uncertain. Double sharp (talk) 23:59, 17 May 2018 (UTC)

Iodine and astatine cations
Hmm. Protonated hypoiodous acid seems to exist (10.1039/JR9510002734). Its structure is H2O+I and it is the active iodinating agent in a large number of iodination reactions. At about 10−12 N it is actually the predominant iodine species in aqueous solution, which casts some light on my earlier comment that the tracer chemistry of iodine differs greatly from its bulk chemistry and approaches that of astatine (whose chemistry is all tracer chemistry). However, I wouldn't call this a normal aqua cation. Such cations are essentially bound by coulombic interactions and almost always don't meet the octet or 18-electron rule; you simply have the metal attracting ligands to itself to disperse its high positive charge, as noted at 18-electron rule. Protonated hypoiodous acid, on the other hand, takes on this structure precisely to meet the octet rule, unlike what a bare "I+" cation would do: if you draw its structure, you find that you have to assign the positive charge to the oxygen to explain its three bonds. So while H2O+Cl, H2O+Br, and H2O+I (protonated hypochlorous, hypobromous, and hypoiodous acids respectively) are definitely much more stable than Cl+, Br+, and I+ would be (with only H2O+I being stable enough to be the major iodine species at low concentrations), I would not consider them to be true solvated cations; they are essentially halogen analogues of hydronium, H3O+. Of course, being cations in water, they will also electrostatically attract other nearby water molecules, but that "H2O" in the formula is not a solvating water molecule but a part of the cation.

The question now is whether "At+" is really a solvated cation like At(H2O)2+ or whether it is simply protonated hypoastatous acid. Electromigration studies mentioned in the Gmelin volume on astatine suggest the latter, with H2O+At being about as strong an acid as HOI itself. The volume goes on to suggest that AtO+ (the astatosyl cation) may in fact be protonated astatous acid, H2AtO2+. Deprotonating this would give astatous acid, AtO(OH), but astatite does not seem to exist and the predominant anionic At(III) species is instead AtO(OH)2−, formed through a secondary hydrolysis reaction:
 * AtO(OH) + H2O AtO(OH)2− + H+

It thus seems that cationic AtI and AtIII species behave midway between being true aqua cations, like those of its horizontal neighbours bismuth and polonium, and "pseudo" aqua cations, like that of its lighter congener iodine. What is clear is that "hypoastatous acid" and "astatous acid" are certainly amphoteric compounds. If At(H2O)+ has a similar pKa to HOI (10.64), that would be less than Ag+ at ~12.0, the lowest of the common monovalent cations (alkali metals, Ag, and Tl).

I suppose it does make a good limit to how far a unipositive cation can be hydrolysed before we stop considering it as metallic, and I guess astatine's metallic band structure and oxoacid salt At+IO3− is probably enough to sway me (though not enough to stop making me think about it again). Double sharp (talk) 16:45, 18 May 2018 (UTC)
 * It seems that At+ (aq) is more stable to hydrolysis than predicted based on the (H2O)nI+ and (H2O)nAt+ structures (10.1524/ract.1989.47.23.105). Astatine is certainly a two-faced element in its chemistry! ^_^ I suppose we could simply use its predicted metallic band structure to break the tie. Double sharp (talk) 04:50, 19 May 2018 (UTC)

Neighbours
Look at chemical bonds formed by element E with itself and its two closest neighbours in the sense of atomic number.
 * If metallic phases prevail, E is a metal.
 * If covalent or ionic phases prevail, E is a reactive non-metal.
 * If bond absence prevails, E is a noble gas.
 * Droog Andrey (talk) 12:53, 11 May 2018 (UTC)
 * Following that definition, As is a non-metal because of bandgaps in germanium arsenides [10.1002/pssb.201552598] and selenides of arsenic. On the other hand, Sb appears to be a metal because Sn-Sb alloys are metallic along with elementary Sb. Droog Andrey (talk) 13:24, 11 May 2018 (UTC)
 * I'm curious: what do you think is the precise role of cation and oxoacid salt formation when determining what a metal is? Tungsten, IIRC, fails both criteria, but no one doubts that it is a metal; so there should be something metallic about its chemical as well as its physical properties. What would you consider that to be? Double sharp (talk) 16:18, 11 May 2018 (UTC)
 * We can consider other elements instead of oxygen. E.g. some transition metals form cluster cations with chlorine as well as Pb/Bi do with oxygen. Droog Andrey (talk) 20:51, 12 May 2018 (UTC)
 * Reasonable, since these class B metals are surely not going to be happy with O-donor ligands. ^_^ I'll soon write up an exposé on the behaviour of the transition metals in aqueous solution. Double sharp (talk) 07:33, 14 May 2018 (UTC)
 * When you say elementary Sb is metallic what properties are you using to determine that elementary Sb is metallic? Sandbh (talk) 05:52, 12 May 2018 (UTC)
 * I imagine it is because elementary Sb has no band gap. Double sharp (talk) 06:50, 12 May 2018 (UTC)
 * That's right :) Droog Andrey (talk) 20:51, 12 May 2018 (UTC)
 * Any comments on this? Droog Andrey (talk) 00:50, 23 June 2018 (UTC)
 * Well, I like it a lot, and I certainly don't see any problems with it. ^_^ Double sharp (talk) 15:37, 26 June 2018 (UTC)

Germanium and antimony cations
I wrote a bit on those two at WT:CHEM, as there seems to be some computational evidence that they at least can form real aqua cations (they are the ones to the left of the traditional metalloid line); I'll copy and paste my post here.

I get the feeling that part of the reason why it is difficult to find Ge2+ in aqueous solution is because that oxidation state is unstable; Ge much prefers to be tetravalent. Kinetic effects permit GeII compounds to persist for a while, though. Naturally, at more commonly encountered pH values, Ge2+ and Sb3+ are probably existing as various hydrolytic and polymeric species; but since the same is true for Sn2+ and Bi3+, and because calculations expect that the former pair should still be able to exist (unlike things like As3+, Ge4+, Sn4+, and Pb4+, which immediately react with a solvating water molecule), I suspect that what you need to find Ge2+ and Sb3+ cations is simply extremely acidic media, and for the latter conditions that slow down the oxidation to the +4 state. (Polonium has definite cationic chemistry in the +2 state, but this is also readily oxidised. I'd be interested to see if At3+ might be possible; certainly At+ and AtO+ are, but +3 is a more stable state for astatine than +1.)

I wonder if something similar might be at work with polonium, which readily self-oxidises from PoII to PoIV. The higher oxidation state behaves more like tellurium (viz. Te(OH)3+ and Po(OH)22+ as the cationic species obtained in the IV state); it does not seem possible to get the fully unhydrolysed species Te4+ and Po4+, since even Sn4+ cannot be obtained (I think the first time you might be able to get a real cation in this group in the IV state is with Lv4+). This is possibly another expression of the principle that elements are more electropositive in lower oxidation states. Even tin seems a lot more electropositive in the II state than the IV state; since the higher state is favoured, I suspect that if it weren't for the metallic β allotrope we'd all be a lot more willing to consider Sn a metalloid. Double sharp (talk) 08:00, 13 May 2018 (UTC)


 * I'm not sure about the sources for aqueous Sb3+, I studied that cation several years ago while developing a solubility table (Sb3+ was eventually excluded from it). Actually, to prevent deprotonation of hydrated cation, one has to rise the concentration of both hydrated cations and OH3+, but that leads to high concentration of anions and complexation (up to Sb(ClO4)4- mentioned earlier). And there's no strict critetia where to draw the line "cation really exists". Well, one may take an acetonitrile solution of, say, fluorocarborane acid together with its Sb(III) salt, add some water and look for OH3+ spectroscopically. If no lines are observed, then Sb3+ isn't compatible with water indeeed. The same could be done for Ge2+ and other questionable cations, thanks to weak coordination of CHB11F11-. Droog Andrey (talk) 11:39, 13 May 2018 (UTC)
 * The metallic nature of Sn and Sb is supported by their metallic alloys with other metals. On the other hand, most of germanium and arsenic alloys lack metallicity. Droog Andrey (talk) 11:45, 13 May 2018 (UTC)
 * Hmm, might this have anything to do with the one at the back of your beautiful table? ^_^
 * yes, that's the very table I mentioned :) Droog Andrey (talk) 19:36, 13 May 2018 (UTC)
 * Cool! I guess we could consider this method an extension of the old one involving perchloric acid media, for when even the weak complexing ability of ClO4− proves too much. ^_-☆ Double sharp (talk) 06:22, 14 May 2018 (UTC)
 * Well, I guess we can draw a rather unimpressive line by saying that a cation should persist for long enough to be observed by some experiments like your suggested ones. The ref I gave says "In principle, they can be classified as ‘stable’ and ‘instable’ ions, where ‘stability’ means that they do not undergo hydrolysis within the simulation period and are also mostly reported as observable in experiment." Sb3+ ends up on the "stable" side while As3+ ends up on the "instable" side (they only consider ions with a +3 or +4 charge). 10.1021/ic4031156 gives a more detailed calculation of what exactly happens to As(H2O)33+, by the way (very quick hydrolysis to arsenious acid, of course). Double sharp (talk) 14:41, 13 May 2018 (UTC)

Antimony as a metalloid
I've been wondering why, from a literature perspective, antimony came to be included among the elements commonly recognised as metalloids.

I suspect there are various "memes" involved. A meme is an idea, behaviour, or style that spreads from person to person within a culture. Here they are, in rough historical order:

0. Pliny the Elder made a distinction between "male" and "female" forms of antimony; the male form was probably the sulfide, while the female form, which is superior, heavier, and less friable, has been suspected to be native antimony.

1. Bastardry. Arsenic, antimony, and bismuth were historically called bastard metals or semimetals on account of their brittle nature. As well, metals were supposed to be fusible. The fact that arsenic sublimed rather than melted further sullied its reputation.

2. Allotropy. Antimony, like arsenic, was known in "metallic" and non-metallic forms. Tin escaped this meme because it was malleable. An equivalent non-metallic allotrope of bismuth was not known.

3. Mendeleev described tellurium as forming a transition between metals and nonmetals. Curiously, he referred to As and Sb as metals, and to Bi as a perfect [sic] metal. That got the hares running as to which other elements could be regarded as forming a transition between metals and nonmetals.

4.  Semiconductivity. Johan Koenigsberger classified solid materials as metals, insulators and "variable conductors" in 1914 although his student Josef Weiss already introduced the term Halbleiter (semiconductor in modern meaning) in his PhD thesis of 1910. The subsequent development of semiconductor physics sparked a renewed interest in Ge and Si, and to a lesser extent, B, as halfway elements. As well, the elements to either side of Sb namely Sn and Te existed in semiconducting forms (noting that grey tin behaves like a semiconductor but is actually a semimetal) so it was expected that Sb would also exist in a semiconducting form, which it did (Moss 1952, p. 173).

5. Metalloid line. Deming's 1923 periodic table made it easier to make out a notional dividing line between metals and nonmetals, naturally focusing attention on the elements to either side namely Be and B; Al and Si; Ge and As; Sb and Te; and Po and At. Note the absence of Bi.

6. Amphoterism. The amphoteric character of:
 * Ge and As, lying as they do between Ga (a metal) and Se (usually considered to be a nonmetal; and
 * Sb and Te, lying as they do between Sn (a metal) and I (a nonmetal),

came to be associated with a transition in metallic character, from metallic to nonmetallic.

The situation in period 6 was less clear. The sequence of elements involved is Pb Bi, Po, and At. Lead is a metal. Astatine was popularly thought to be a halogen, and therefore a nonmetal (although the folks who first synthesised it thought it was a metal). On this basis it could’ve been thought that Bi and Po would be amphoteric. However Bi was regarded as basic, and only Po showed some amphoteric character, which may have resulted in some authors regarding it as a metalloid.

7. Pauling published his influential book General chemistry (1947) in which he referred to B, Si, Ge, As, Sb, Te, and Po as being metalloids. (He erroneously referred to Sb as being a semiconductor.)

8. Rochow published The metalloids (1966) and recognised B, Si, Ge, As, Sb, and Te as such.

9. Group 15. The progression in metallic character going down group 15 tended to reinforce regarding Bi as a metal, but not Sb. For example:


 * "Antimony…is more nonmetallic than metallic…bismuth…more nearly approaches a metal in physical and chemical properties." (Norris & Young 1938, p. 529)
 * "The trisulphides of arsenic and antimony are acidic, forming salts with yellow ammonium sulphide and alkali, while that of bismuth is typical of a metal." (Moody 1969, pp. 267, 321)
 * "All the elements react readily with halogens but are unaffected by non-oxidising acids. Nitric acid gives, respectively, phosphoric acid, arsenic acid, antimony trioxide, and bismuth nitrate, which well illustrates the increasing metallic character as the group is descended." (Cotton & Wilkinson 1976, p. 288)
 * "The paucity of [stereochemical] information about Bi is due to the more metallic character of this element, which does not form many of the simple covalent molecules formed by As and Sb." (Wells 1984, p. 878)
 * "Bismuth(III) oxide occurs naturally as bismite and is formed when Bi combines with O2 on heating. In contrast to earlier members of group 15, molecular species are not observed for Bi2O3 and the structure is more like that of a typical metal oxide." (Housecroft & Sharpe 2008, p. 474)

Conclusion: I guess Sb came to be regarded as a metalloid mainly due to its brittle comportment; existence of a non-metallic semiconducting allotrope; proximity to the metalloid line; perceived amphoterism; and apparent lack of genuine salts. In contrast, Bi came with only one of these features.

References Cotton FA & Wilkinson G 1976, Basic inorganic chemistry, Wiley, New York Housecroft CE & Sharpe AG 2008, Inorganic chemistry, 3rd ed., Pearson, Harlow Moody B 1969, Comparative inorganic chemistry, 2nd ed., Edward Arnold, London Moss TS 1952, Photoconductivity in the elements, Butterworths Scientific Publications, London Norris JF & Young RC 1938, A textbook of inorganic chemistry for colleges, 2nd ed., McGraw-Hill, New York Wells AF 1984, Structural inorganic chemistry, 5th ed., Oxford University, Oxford.

-- Sandbh (talk) 10:20, 24 May 2018 (UTC)
 * I agree that many of these features must have played a part in it. One of the strongest may well be that Sb is right next to the metalloid line as it is usually drawn, and indeed I will spend a little more time now discussing the position of that line. I suspect this is also why Po is classified quite commonly as a metalloid, even though an objective look at its chemical and physical properties show quite little to support such a classification. (Aluminium may have escaped this fate by being very common; the hardness of the Al3+ cation may also have had something to do with it, which would also explain why beryllium is so rarely considered a metalloid.) Note that Ga, In, Sn, and Pb are also amphoteric in their common oxidation states, so I think people were drawing the metalloid line to the right of the amphoteric line.
 * Another thing that may have influenced some authors in considering Po and At to be metalloids or nonmetals is that Bi is less metallic than would be expected from its position; Po and At have truly metallic band structures, and Po2+ at least is more basic than Bi3+, but this would not have been common knowledge as high radioactivity makes the end of period 6 a chemical backwater (which was even more true then than now). So off they went to the metalloids, with astatine being considered to follow iodine as a nonmetal. (Polonium at least had tellurium as a congener to warn against this.) This "default" treatment was also enough for germanium to often get classified as a poorly conducting metal in the past. So it seems that the default position was either:
 * P-block elements to the left of the diagonal staircase are considered metals; those to the right are considered nonmetals. (A binary classification; this is probably still pretty common.)
 * P-block elements next to the diagonal staircase are considered metalloids, unless they happen to be in groups 17 or 18 (default nonmetals), or are the ubiquitous aluminium.
 * Both of these positions suffer, of course, from starting with the desired conclusion and then trying to group together the remaining elements, excluding astatine by sweeping it under the rug and aluminium by "I know it when I see it". To some extent, then, I suspect that examples exemplifying the behaviour of elements along the metalloid line may have been influenced unconsciously to give the "right" answer. So for example, while indeed Sb does not form a nitrate when dissolved in nitric acid (although the oxide it forms is definitely more impressive than arsenious acid), it does form a sulfate when dissolved in sulfuric acid. By picking the right properties it is quite possible to nudge many of these ambiguous elements one way or another; the fact that the range of elements that have been called metalloids ranges from nitrogen through zinc to radon should have made this expected, as has the fact that even selenium (!) has been considered a metal and tin (slightly less !) a nonmetal. So while the lessons of history should be learned well, we should try to reexamine the question without its biases but with its depth. Double sharp (talk) 11:41, 24 May 2018 (UTC)

Thank you. I'd like to respond further but am a bit pressed for time just now.

I was just looking in Bailar's Comprehensive Inorganic Chemistry, and in Wiberg, to clarify the acid-base character of Sb2O3. Apparently its acid character is more pronounced than its weak basic character. (I was under the impression it was the other way round!) In contrast, Bi2O3 is weakly but essentially basic, with any acidic character being bought out only in highly acidic media. I presume the bias of Sb2O3 towards acidity was another historical consideration. Sandbh (talk) 03:54, 25 May 2018 (UTC)
 * Yes, that may have contributed to it – though I'd note that while Sb2O3 may be more acidic than basic, B2O3, SiO2, GeO2, As2O3, and TeO2 are much more acidic than basic. Note that bulk aluminium will react at room temperature with aqueous NaOH but not aqueous HCl (unless hot and concentrated) because of the oxide layer. ^_^ Double sharp (talk) 16:46, 25 May 2018 (UTC)
 * Al does react with aqueous HCl, but aqueous H2SO4 is a problem indeed. Droog Andrey (talk) 17:31, 25 May 2018 (UTC)
 * That will teach me not to write this late in my time zone. ^_^ Yes, it should have been H2SO4. Double sharp (talk) 04:13, 26 May 2018 (UTC)

One more observation, which fits under meme 9 above, is that bismuth is metallic enough to form what is generally regarded as an ionic fluoride BiF3. In contrast, I'm not aware of any ionic antimony compounds. Sandbh (talk) 11:44, 14 June 2018 (UTC)

Antimony(III) perchlorate

 * Antimony(III) perchlorate Plasmic Physics (talk) 12:24, 14 June 2018 (UTC)
 * Thank you Plasmic Physics.
 * Yes, I thought about that but couldn't find anything suggesting it was an ionic compound. Do you have a reference? The best I could find was in Inorganic Substances Handbook (1996, p. 44), where Lidin says, "Main form of existence of BiIII is in the crystal lattice of salts e.g. [Bi6(OH)6+12](ClO4)6; the Bi3+ ion exists only in the crystal lattice of BiF3". Given bismuth is apparently not ionic in its crystalline perchlorate I doubt antimony would be ionic in its perchlorate.
 * I overlooked a 1995 reference to the existence of a simple [Bi(H2O)9]3+ ion in crystalline [Bi(H2O)9](SO3CF3)3.
 * On a related obscure and extraordinary note, Greenwood & Earnshaw (1998, Chemistry of the elements, p. 564) mention that in 1971, Bi10Hf3Cl18, of all things, was shown to be (Bi+)(Bi95+)(HfCl62−)3 thereby identifying Bi+ in the solid state. Sandbh (talk) 03:52, 15 June 2018 (UTC)
 * To be honest, it was more of a suggestion than an example, and I was refering to the anhydrate. Plasmic Physics (talk) 06:05, 15 June 2018 (UTC)

This item in JChemEd says that perchlorates have a high degree of ionic character, and that the perchlorate ion was estimated by Sanderson to have an EN of 4.244, which is quite impressive. I see that SbF3 (noting F has an EN of 4.0 on the Sanderson scale), has a polymeric structure. Sandbh (talk) 03:44, 18 June 2018 (UTC)
 * Yes, but perchlorate is a non-coordinating anion, and fluoride is not. Plasmic Physics (talk) 05:42, 18 June 2018 (UTC)

I looked up this article about anhydrous antimony perchlorate. Upon heating the perchlorate starts decomposing at 210 °C. It is totally hydrolysed by water with the formation of hydrated antimony trioxide. The authors note that the perchlorates of Cs and Rb are ionic but do not say anything about the status of the antimony "salt". Solubility rule tables that I have seen show perchlorates as being generally soluble in water, noting the degree of solubility will vary. Sb(ClO4)3 seems to be an exception, which might be expected given Sb struggles (IMO) to be regarded as a metal. Bismuth perchlorate doesn’t give any insoluble hydrolysis products. Sandbh (talk) 11:06, 18 June 2018 (UTC)

Position of the metalloid line
Getting back to 's earlier observations about the position of the metalloid line…

I’m not sure there was much "starting with the desired conclusion", apart from polonium and astatine.

As far as I can tell the line emerged out of the observation that metals had a metallic appearance whereas nonmetals didn’t (or so it was thought). See Hinrichs (1869) for example.

Then, in 1891, Walker published a periodic 'tabulation' with a diagonal straight line drawn between the metals and the nonmetals. His line was equivalent to the modern zigzag line but for counting eka-tellurium and eka-iodine as metals.

The same table appeared in his book Introduction to physical chemistry (1899, p. 44), which became a set text in many British universities. In it, Walker wrote:

"The distinction between metals and non-metals is not a very sharp one, for the properties of the one class gradually merge into those of the other. There is, however, a classification which is generally adopted in practice, and this classification is in conformity with the periodic table. The elements enclosed in the dotted triangle at the lower left-hand corner of Table I. are the non-metals as usually understood. All the other elements tabulated are metals."

In 1906, in his highly influential textbook Introduction to general inorganic chemistry, Alexander Smith published a periodic table with a zigzag line separating the nonmetals from the rest of elements (p. 408).

Like Walker before him, Smith’s line replicated today’s regular zigzag line with the difference that, after running under tellurium it ran under iodine. As well, he had beryllium and magnesium showing as part of group 12 so that the line separated beryllium and boron.

While germanium ended up on the metallic side of the ledger Smith astutely said that classifying it as such was of questionable propriety (p. 410).

From there it was only a question of continuing to build on the earlier commentary that the dividing line was not sharp, and that the elements to either side could show more or less metallic or nonmetallic character.


 * Beryllium and aluminium retained their identities as metals I guess because of their remarkable mechanical properties.


 * The status of boron and silicon as metalloids came to be cemented with the rise of the semiconductor industry. For its time, I find it remarkable that the Riesenfeld Periodic Table (1928), here, showed B and Si as semimetals/metalloids. On boron as a metalloid, Greenwood’s comments about the analogy between boron and metals are also remarkable.


 * Arsenic was regarded as a metalloid due to it existing in metallic and nonmetallic (black, yellow, and amorphous) forms. There was also its brittle nature, and unique volatility. And the fact that its trioxide was generally regarded as being amphoteric, despite reservations by a few authors, wouldn’t have hurt.


 * Selenium mostly escaped the metalloid label due its acidic oxide. For more on selenium’s category, see here.


 * Tin was regarded as a metal due to its malleability and ductility.


 * Antimony came to be regarded as a metalloid due to the already discussed memes.


 * Tellurium was regarded by Mendeleev as an intermediate element.


 * Bismuth, IMO, very largely escaped the metalloid label due to (a) the reasons mentioned in meme #9; (b) the "definitely basic" (Greenwood and Earnshaw, p. 575) nature of its oxide; and (c) the fact that this oxide represented a starting point for much bismuth chemistry. The historical confusion of bismuth with lead and tin probably also played a part here.

The identification of B, Si, Ge, As, Sb, and Te as metalloids looks more like a survival of the fittest outcome. Any biased thinking along the way has presumably by now been filtered out by the collective wisdom of the masses. I do appreciate, however, that what we take to be filters sometimes turn out to have been blinkers (e.g. the impossibility of noble gas compounds, or being able to isolate fluorine chemically). [Having said that, I find it odd curious that while you’ve previously argued for treating bismuth as a metalloid, you’re now suggesting antimony might be better regarded as a metal.]

The exception would’ve been polonium and astatine, Not enough critical thinking was done about the properties of polonium for the reasons you identified. Not enough was widely known about astatine so it was either defaulted as a halogen and therefore a nonmetal, or less often presumed to be a metalloid on the basis of its position next to the metalloid line, and by extrapolating from iodine. It is ironic now that astatine was historically first categorised as a metal and that it may well turn out to be one. Sandbh (talk) 12:09, 17 June 2018 (UTC)


 * I've mot not got much time right now, but this information is fascinating. To the extent it is not already in Dividing line between metals and nonmetals, I hope you'll take the time to insert it. YBG (talk)
 * Thank you very much YBG. I hope to be able to do so. Sandbh (talk) 00:08, 18 June 2018 (UTC)
 * Well, I admit that it does look rather contradictory at first glance for me to have argued in these ways for Sb and Bi. The difference is because I no longer think that it's desirable to retain metalloids as an intermediate category. So when I was arguing for calling Bi a metalloid, I meant that it wasn't a true metal, a position which I still agree with. Nevertheless, it behaves more like a metal than a nonmetal, and on that basis I would be prepared to call bismuth a metal if we are using a strictly binary division (never mind that there are a number of elements in the middle; they all seem to be closer to one side than the other). My arguing for regarding antimony as a metal ought to be interpreted in the same context, which I should certainly have elaborated on earlier; I'm sorry for creating confusion!
 * I will look at the rest of your wonderfully detailed post later and try to post a reply that does justice to it as soon as I can. ^_^ Double sharp (talk) 10:33, 18 June 2018 (UTC)

Double sharp (talk) 15:44, 10 July 2018 (UTC)

Layout advice/plans for "Compounds"
For the elements pages, my understanding is that we typically have: I suggest that we aim to standardize the layout of the Compounds section. The plan:
 * a subsection on "properties" where the behavior of the element (zero-oxidation state) is described.
 * a section on "Compounds".
 * short paragraph with an overview (most important cmpds from commerce (usually tonnage) perspective)
 * option A: subsections organized according to compound classes (oxides/sulfide, halides, organometallic, ...), describing iconic or important representatives.
 * option B: subsections organized according to principal oxidation states (Mo(VI), Mo(V), Mo(IV), ...), describing iconic or important representatives.

Unsolved problem: a lot of important compounds are often described in the refining section. Advice/comments are welcome! --Smokefoot (talk) 16:31, 7 July 2018 (UTC)
 * I can understand having oxidation states described in the element's article, but why have compounds in there? -DePiep (talk) 00:32, 8 July 2018 (UTC)
 * Oh, I hadn't thought of that. You mean omit an overview of the main compounds derived from the element? Just list the main oxidation states?  We could also remove anything from the article that does not involve the element per se.  So iron would only involve the metal and its alloys.  --Smokefoot (talk) 01:58, 8 July 2018 (UTC)
 * Why so cynical, ? You first propose a section compounds, containing either by class or by oxidation satate i.e., a systematic completeness. Now here you write "an overview of the main compounds", which is a different approach with different results. So let me rephrase my point: what approach, setup and completeness would we adopt for describing compounds in there? - DePiep (talk) 10:05, 8 July 2018 (UTC)
 * I think it is absolutely necessary to include compounds in the article. You can hardly describe the element's reactions without starting to talk about them, and they are an integral part of the element's chemistry. I find Smokefoot's suggestions all very reasonable. I don't see a problem with discussing important compounds both in Compounds (or Chemistry) and in Production; the former should focus on their chemistry, and the latter should focus on their place in the production of the element. (We would have to mention the useful ones again in Applications, after all.) Double sharp (talk) 04:41, 8 July 2018 (UTC)
 * Sounds good. Would mean the options A, B are not used btw (not all compounds in one dedicated section, but mentioned in appropriate sections per relevance. Can we apply this to all element articles? Side question: should/would some compounds end up on the infobox? - DePiep (talk) 10:11, 8 July 2018 (UTC)
 * That is not what I meant. I meant that compounds should be in both a dedicated section and mentioned in other appropriate sections as well; the dedicated section discusses chemistry, while the other sections discuss their uses, production, toxicities, and so on (Applications, Production, Precautions respectively). This is because an article on an element ought to discuss its chemistry, and it is impossible to discuss its chemistry without its compounds. That is why all our actually good element articles have a section called "Compounds", "Chemistry and compounds", or perhaps "Chemistry"; if you don't have such a section, it can't be a complete and well-organised article (recognising that it's a necessary but not a sufficient condition). Obviously, we should not be mentioning all compounds, but only the important ones that you cannot understand the element's chemistry without: we're not going to cover all of organic chemistry in the carbon article, for example. But if "compounds" as a section would be taken out of our element articles they would not even deserve B-class in my opinion, as it would be an obvious omission. The place to focus on the element alone is the infobox, not the article (so I would keep compounds out of there). Double sharp (talk) 12:16, 8 July 2018 (UTC)
 * Thanks DePiep and Double sharp for the input. I will try to review the "Compounds" section for more metals and try to ensure that each has some sort of overview.  I am leaning toward a brief intro laying out oxidation states followed by subsections on oxides/nitrides/sulfide then halides and then organometallics.  Such approach probably is more generally useful.  But the main thing is that we provide some sort of overview of the main compounds. --Smokefoot (talk) 02:02, 9 July 2018 (UTC)
 * - If molybdenum is how you would change the articles - I would suggest to have some subsection which describes the chemical properties of the element itself. You removed that it reacted with O2, and afaics the only mention of the element is now under "physical properties" which states "t does not visibly react with oxygen or water at room temperature," (does or doesn't it react with water...?). I.e. some information like does it react with O2, H2O, halogens, etc. Christian75 (talk) 11:29, 10 July 2018 (UTC)
 * My counter-suggestion, respectfully offered, is to restrict the chemistry in the "Properties" section to air (O2, CO2, N2) and water. One exception might be tarnishing of silver by sulfur compounds. My perspective is that the "Properties" section would give insight to those handling the bulk metal.  It seems preferable to avoid overlap between "Properties" and the "Compounds" sections.  But you do raise a good point that reactions of the metal with simple reagents should be mentioned somewhere because they are very "elementary" (sorry).  Virtually every element reacts with F2, Cl2, sulfur, and those simple reactions could be included. Another thing that could be included in "Properties" is resilience to common acids (HCl, H2SO4).  But that behavior is probably too sensitive to describe well.  It might be worth seeing what the German Wiki is up to because that one is very good.


 * PS: About Mo and water: no reaction. Like all metals, this non-reactivity is the result of passivated surfaces. Transition metals do not react with water near STP. In fact, I think only the only elements that react with water near RT are alkali metals and the halogens, but other editors might know exceptions.--Smokefoot (talk) 13:01, 10 July 2018 (UTC)
 * I looked at the German wiki. The have a section named "Chemische Eigenschaften" (chemical properties) and generally a lot more about the chemical itself. (Mo is not a very good article, but e.g. Wasserstoff (Hydrogen) has a lot of chemistry in one section. Some of the content is in our hydrogen, but you have to search for it. Christian75 (talk) 17:10, 13 July 2018 (UTC)
 * The alkaline earth metals from Ca downward react with water (slowly for Ca, but it gets quicker going down the group). The lanthanides and actinides will also react. (Sc and Y will also react according to Greenwood and Earnshaw, but the reaction is very slow at RT unless the metals are finely divided.) But indeed, the transition metals proper (i.e. excluding the rare earths) will not react with water near STP; the group 12 metals also do not. Mentioning of the reaction of the elements with common reagents under the properties of the element rather than its compounds is absolutely fine by me; after all, it passes the benchmark of "if Greenwood and Earnshaw do it, it's a good idea". ^_^ Double sharp (talk) 15:42, 10 July 2018 (UTC)
 * From Double sharp's "That is not what I meant. I meant that ..." comment above, I learn & understand what our good element articles have (that is: already have). But somehow Smokefoot removes the word 'Chemistry' from the TOC in Mo (a GA=i.e., good article), and much more changes that by effect impose Smokefoot's own proposal while ignoring Double sharp's setup description. I object to Smokefoot's edits, as being controversial at least but actually not supported by consensus. While some detailes may be an improvement, the larger changes are to be reverted. - DePiep (talk) 22:19, 10 July 2018 (UTC)
 * I actually find Smokefoot's edits perfectly reasonable and support them. While I would prefer to have chemistry discussed alongside compounds, there does not seem to have been any discussion of chemical reactions of Mo metal in this section beforehand, so the title is justified for now even though this is an omission here (which I expect will be remedied soon). As for the Mo article: though it is marked with a green plus, I do not believe that it should be a GA in its current state due to its deficiencies in chemical coverage (B would be more justified IMHO). This is why I referred to "actually good" articles rather than "good" articles. By the adjective "actually" I meant to imply something different than GA, as an article can pass GA and still not be good in anything but name if it is incomplete like this; I think a lot of our transition metal articles are like this (e.g. scandium). Double sharp (talk) 05:58, 11 July 2018 (UTC)

It can be useful to include a chemical properties section before discussing compounds, in order to give an overall impression of the chemical nature of the element in question. For example, Mo is one of the few metals that does not form a simple cation and does not have a basic oxide; unusually for a metal it has a relatively high EN of 2.16, and a positive standard electrode potential. Cotton et al. (Advanced inorganic chemistry, 6th ed.) note that the chemistry of Mo is among the most complex of he transition elements (p. 922). The astatine article has a nice Chemical properties section, showing the utility of such a construct. You seem to be heading towards such a structure anyway i.e. Physical properties, Chemical Properties (reactions with air etc, and common acids?), Compounds. Sandbh (talk) 08:07, 11 July 2018 (UTC)
 * Helpful comments. I do like the idea of a preface. Giving the range of oxidation states and mentioning something about important compounds might be one way to give some insights into the element. I am unfamiliar with the meaning of a "simple cation", sounds suspect or archaic. The astatine article might be unrepresentative since its chemistry is so unimportant, not much context since no one uses it. Related to the metals, I have never (or v rarely) heard any inorganicker refer electronegativity except in the case of Au. But these perceptions do not mean that I am correct and might reflect my blindspots.  But we can agree on some brief overview statement.  Cheers,--Smokefoot (talk) 13:14, 11 July 2018 (UTC)
 * Simple ion is another word for a monoatomic ion Christian75 (talk) 17:04, 13 July 2018 (UTC)

Thank you. What I said was partly wrong; I must have written it in my sleep. Molybdenum does form a simple cation (aq), and the standard electrode potential of molybdenum (–0.20) is negative, not positive. Regarding the meaning of "simple cation", and further to Christian75's comment above, Parish (The metallic elements 1977, pp. 113, 133) discusses the aqueous chemistry of the 4d- and 5d- metals (excluding group 3) and notes that only in a few cases are "simple" aquated cations known. In a similar vein, and in writing about the hydrolysis of cations in aqueous solution, Smith (Inorganic substances: A Prelude to the study of descriptive inorganic chemistry 2000, p. 173) refers to simple cations as "Mn+(aq)".

I was surprised about your observation that inorganic chemists don't, or very rarely, refer to electronegativity (but for Au). In scanning the molybdenum article I saw that the Physical properties section made mention of the electronegativity of molybdenum as well as some other chemical properties. For clarity, I've extracted all of these chemical properties and placed them in their own section. And I added a citation-supported sentence about molybdenum's disinclination to form a cation in aqueous solution.

Perhaps the important thing is whether mentioning electronegativity in the preface would add any value, or reader interest. For example, it could be added that "among the non-noble metals, only tungsten has a higher electronegativity."

I feel that flourishes such as this, together with molybdenum's refractoriness; its reluctant cationic nature; and seemingly complex chemistry, have the potential to make the article more engaging. Sandbh (talk) 06:11, 14 July 2018 (UTC)


 * Note: Smokefoot has continued this thread on my talk page, here. Sandbh (talk) 08:17, 15 July 2018 (UTC)
 * WTF ? Talk here openly at wp:elements or don't talk at all. Realy, any serious WP editor would not go private on a public issue. . -DePiep (talk) 22:54, 21 July 2018 (UTC)
 * It just seemed that the conversation was settled (or going nowhere) so I didnt want to bore other editors with my responses to sandbh on his specific points The gesture was one of politeness.  I dont understand DePiep's tone, is he in charge of something?  In any case, I paste in the main points  below.


 * I support the idea of a mini-preface when discussing compounds. The preface could give the range of oxidation states and mention something about important compounds.
 * The concept of "simple cation" is archaic, and seems to be mishandled by editors here. Not included in any class I ever got or gave.
 * The astatine article is unrepresentative since its chemistry is unimportant, my focus is on compounds of non-radioactive elements.
 * Related to the metals, I have almost never heard any inorganicker refer to electronegativity. Ligands so dominate the properties of complexes that the concept is useless.  But that is a chemistry perspective.  --Smokefoot (talk) 22:19, 22 July 2018 (UTC)


 * Mo/W electronegativity is higher only on Pauling scale because of strong multiple homoatomic bonds between their atoms. In chemical sense, their electronegativity is not very high. Droog Andrey (talk) 08:12, 19 July 2018 (UTC)

Big discussion now archived
... with the final index located in Archive 35 YBG (talk) 05:08, 23 July 2018 (UTC)

Metal article update
I've been doing some work on the Metal article.

I remember seeing it several times in the past. It was hard going. I felt disappointed given metals form such a core part of chemistry, physics, and materials science. The article itself is viewed about 720,000 times a year. Then again it must have been tough to write given it covers so many disciplines.

Anyway I've had a go and am reasonably happy with the way it looks now, apart from the History section, which needs a lot more work. Many citations need to be added, too.

Could you please let me know how the metal article looks to you now? Sandbh (talk) 05:24, 11 August 2018 (UTC)

Isotopes of X
Hello, I have opened a mass nomination of the 84 "Isotopes of X" categories composed solely of the main article and redirects to the main. You are all invited to participate at this CfD. – Laundry Pizza 03  ( d c&#x0304; ) 20:27, 15 August 2018 (UTC)

The limits of nuclear mass and charge: new paper by Witold Nazarewicz
Interesting... (10.1038/s41567-018-0163-3) Double sharp (talk) 16:15, 24 August 2018 (UTC)

Discovery era in the periodic table


PS: Along the way I happened to be looking at the Periodic_table_(discovery_periods) template, here. The background color legend to this template is a masterful piece of work. Kudos to User:DePiep for that.
 * Thanks :-), but the colors were there already before I first touched it in 2012: ;-).
 * Still on my mind: instead of time periods of discovery, why not classify them by discovery technique (X-rays, Optical spectrometer, cyclotron). -DePiep (talk) 15:22, 11 August 2018 (UTC)
 * that is an excellent idea! YBG (talk) 16:28, 11 August 2018 (UTC)
 * It is about Periodic table (discovery periods) -DePiep (talk) 22:24, 11 August 2018 (UTC)
 * Right, but maybe have a Periodic table (discovery methods or something? Close correlation, but as you point out, actually different concepts. YBG (talk) 00:44, 12 August 2018 (UTC)


 * (expanding on method of discovery:)…specify the old known metals e.g. along: "native form", "processed chemically/physically" (melting), "after 1809", ... - DePiep (talk) 08:02, 13 August 2018 (UTC)
 * Could you clarify what you mean? Sandbh (talk) 01:32, 14 August 2018 (UTC)
 * (my comment was in wrong position, refactored/repositioned). I meant to say: if we illustrate the discovery of elements not by era but by method of discovering, maybe we could be more specific about the metals' discovery. II was reading Metal. - DePiep (talk) 06:07, 14 August 2018 (UTC)
 * Ah, I see, That would be worth exploring. Metal stills need a lot of work. Sandbh (talk) 07:28, 14 August 2018 (UTC)
 * Yeah, I only wanted to privide the link. Interestingly, the period colors could be sequential by time, (say like these colors). That would illustrate, roughly, higher atomic numbers being discovered later, and outlayers for noble gases. - DePiep (talk) 10:20, 12 August 2018 (UTC)
 * That is good. There are seven discovery periods. Could we use the seven colours of the spectrum? I know using indigo is always a challenge. Perhaps the colour frequencies could be matched to the year of discovery? Sandbh (talk) 00:47, 13 August 2018 (UTC)

I could have been clearer. The colours were already there. What I was impressed about was the minimalist layout of the legend, compared to what was there before. It is unusual and refreshing to see a table like that without borders around each cell. Sandbh (talk) 00:38, 13 August 2018 (UTC)
 * See Periodic_table_(discovery_periods)/sandbox. Not ideal, hard to distinguish (cannot link an element to a legend color without guessing). - DePiep (talk) 06:01, 13 August 2018 (UTC)


 * Might this help(?): End of the rainbow? New map scale is more readable by people who are color blind.
 * "Quote : Nuñez and…Renslow tackled this problem with their new scale, which they call cividis, by using just two colors with a clear brightness hierarchy: blue and yellow. Just as with a gray scale, people perceive the brightest yellows as peaks and the darkest blues as lows. But viewers can perceive a greater level of detail with colors instead of shades of gray. Unquote"
 * There is a link in the linked page to the parent article, which is open access. Sandbh (talk) 07:53, 13 August 2018 (UTC)
 * Very useful, different approach (in a sequential scale!) than [colorbrewer2.org]. BTW, on the other issue (re more meningful legend classes): already I see some discovery-techniques in today's legend. OTOH, the "2000" border is arbitrary, and ~meaningless imo. Worth exploring? DePiep (talk) 08:02, 13 August 2018 (UTC)
 * The borders are worth exploring. They look OK to me. The 2000 border is arbitrary but then again the pre-2000, post-2000 divide has a reasonably strong recognition in modern culture. People talk about something being "so last century" for example. Sandbh (talk) 01:32, 14 August 2018 (UTC)
 * Wow, only published this month :-) . First thoughts. re cividis (new term), also viridis is mentioned. wrt the periodic table: the cividis blue-yellow scale is very good wrt color blindness (as opposed to the rainbow) -- by design. In the PT, we have discrete categories not a continuous scale. 9+1 categories may be a bit much to recognise (interpreting: reading from element cell to legend, or vice versa). However, since the categories are trendlike over a period, we might use three colorsnot two (CB-problematic colors not neighbouring). Also: the introduction correctly & nicely describes some problems with the rainbow scale (apart from CB): the rainbow has no intuitive order to the human interpretor (physically is has of course). Later more. -DePiep (talk) 08:34, 13 August 2018 (UTC)
 * No doubt there will be a less arbitrary border at some point, when the era of target–projectile reactions as we know them is halted as we bump into the proton drip line. That will nonetheless likely be closer to 2050 than to 2000! ^_^ Double sharp (talk) 16:17, 24 August 2018 (UTC)

Nihonium FA
My compliments go to all who made Nh an FA. -DePiep (talk) 21:56, 9 September 2018 (UTC)
 * Thank you, and many thanks to R8R for all your help too! Double sharp (talk) 07:07, 11 September 2018 (UTC)

End of the rainbow?
A recent Scientific American article End of the rainbow: new map scale is more readable by people who are color blind provides some interesting reading. One point in particular stuck out to me: that even for people with full color vision, the rainbow presentation of information can be misleading. Definitely food for thought. YBG (talk) 06:19, 11 September 2018 (UTC)
 * Interestingly, was pointed to by Sandbh in above  (before the ink had dried). -DePiep (talk) 06:50, 11 September 2018 (UTC)
 * Circular rainbow in the sky of Pattaya.jpg
 * This is way too funny now that I've figured out what actually transpired. I read that post, e-mailed the link to myself at work, printed out a copy to read and share with others. Some time later, I saw the article in the library. Then today the article surfaced again in the archaeological dig that is my desk, so I found the link (but not in my e-mail in box) and then sent the newly "discovered" link to a couple of color-blind folks I have worked with, and then came home thought to myself, I must share this with my WP friends. Can anyone say citogenesis? YBG (talk) 07:10, 11 September 2018 (UTC)
 * So, your rainbow went full circle. No Au for you then. -DePiep (talk) 15:17, 11 September 2018 (UTC)
 * Gotta be satisfied a different yellow element: the sulfurous odor of egg-on-face.   YBG (talk) 17:00, 11 September 2018 (UTC)

Infobox elements: add PyMOL color?
To consider: we can add the PyMOL color used for atoms in PyMOL molecule visualisation. Colors per element are in WP:CHEM/Elements_coloring_scheme.

If I understand this well, PyMOL colors are most commonly used in visualisation (1, 2). PyMOL PyMOL wiki does not list the default colors.

It could be listed under "Miscellanea", showing in a shadowy sphere + the RGB triplet number. . -DePiep (talk) 16:51, 11 September 2018 (UTC)


 * Is this the same as CPK coloring? YBG (talk) 17:04, 11 September 2018 (UTC)
 * Maybe CDK is more used. Is it complete for first ~100 elements? Cl and F still double use of green? Reading more about PyMOL, looks like its et is less stable, and very open to user choices.
 * Having a element-to-color link is step one, the reverse route is even more interesting: hovering over a colored ball-and-stick model showing atom inforation tooltip-like. -DePiep (talk) 17:21, 11 September 2018 (UTC)
 * I have the Jmol ones drawn up to element 109 (meitnerium) at User:Double sharp/Jmol (edit the page to see the precise hex values). Double sharp (talk) 23:59, 11 September 2018 (UTC)
 * shall we move that table to mainspace Periodic table (CPK colors) (keeping your credits ;-) ), and use it in article CPK coloring? Is there a stable and preferable color set? -DePiep (talk) 12:51, 22 September 2018 (UTC)

Infobox periodic table group: redesign

 * Page views over 90d: |Alkali_metal|Alkaline_earth_metal|Group_3_element|Group_4_element|Group_5_element|Group_6_element|Group_7_element|Group_8_element|Group_9_element Groups 1-9 + article, |Group_11_element|Group_12_element|Boron_group|Carbon_group|Pnictogen|Chalcogen|Halogen|Noble_gas Groups 10-18

I propose to redesign , both in template code and in information presented.


 * 1. Change into using Infobox. Today it is an elaborate wikitable construct. handles rendering much better, including layout, handling images, serving mobile applications. And of course it is serving WP:INFOBOX guidelines well. I plan to maintain the individual templates like  (no in-article template filling), as we usefully do with element infoboxes.


 * 2. Group info in top. In top there should be the group properties: meaningful image, group names, maybe point out meaning of "noble gas" in a few words, position in the periodic table. This is where the meaning of "group" is made. Could be using a few subheaders. Not all groups have all such info, though likely group 3 has something to note here. The template has flexible options in this (like a free text "description" parameter).


 * 3. Member info ("member" = group element). Apart from main info (name, symbol, Z), there should be group-related info only here. Showing the electron configuration shoud illustrate the pattern (Janet's!). Is atomic weight relevant? Images must be small (or abesent). Also, I think it could better to write-out colors's meaning right away, instead of using legend blocks at the bottom. Everything to show the patterns.

See /testcases for groups 1, 3, 18.

Comments? -DePiep (talk) 11:06, 5 September 2018 (UTC)


 * Lovely!! At first glance, I misunderstood and thought you were proposing a revamping of the table at - which would also be a good thing to do. YBG (talk) 14:41, 5 September 2018 (UTC)

Introducing the Member table
One of the problems to solve is what member info to present, and how to present that info. The what: group-relevant info like element name, period, Z, atomic weight, valence!, elconfig, category, images?, ...

I concluded that an infobox is not the best way to present this member information. Problem is that the two-column form does not invite to present the info comparatively. A simple and better way is by table. Like this:

The infobox can keep a basic list possibly images. -DePiep (talk) 15:33, 22 September 2018 (UTC)

Talk:Unbibium
FYI there is an RFC at Talk:Unbibium regarding whether a page at that title should now exist. shoy (reactions) 19:18, 24 September 2018 (UTC)

So... what now?
Hey everyone at Project:Elements! Just wanted to check in to see what everybody is thinking for the next big project that we'll be working on? In the meantime, I've been waiting to hear back from more people on Unbibium before I get working, but is there anything that anybody is working to get done? Just want to hear some input about the stuff going on... If it were up to me and I wanted to write a FA article, I'd start on chromium or beryllium. Then again, I don't have the experience to perform such an undertaking. Ah well. Back to raising c-classes to b-classes. UtopianPoyzin (talk) 03:33, 28 September 2018 (UTC)
 * Well, History of aluminium is currently at FAC. I have been short of free time to spend on WP lately, but when it does come I plan to do some more good old 2016-style GA spamming. ^_^ Double sharp (talk) 04:21, 28 September 2018 (UTC)
 * I'd welcome any contributions to the Metal article. It's sobering to think that but for metals we'd still be in the Stone Age. Sandbh (talk) 04:29, 29 September 2018 (UTC)
 * I already mentioned that History of the periodic table is an ideal title to see as an FA on main page next March (150 y). Also, appart from element articles, there are element related articles that could use an upgrade. -DePiep (talk) 06:38, 29 September 2018 (UTC)

Hi, ! We're not a particularly large project and there haven't been WikiProject-wide projects; this will probably not change anytime soon. At the present moment, you could help a lot by providing a review at the FAC page for History of aluminium; I would be really thankful for that. After the review is over, there'll still be with his very ambitious project of metal; this project will clearly require a lot of labor and he will certainly appreciate help. I will be waiting for to improve the remaining sections of aluminium, then we'll hold a GAN and probably a peer review to bring the article to the FA quality level.

If you want to try to go for FA with either Be or Cr, go for it! Personally, I essentially joined the project when I decided to bring fluorine to the FA quality; at the time I had no major Wikipedia experience and even my English was fairly weak. After lots of writing, trying this and that, observing other editors' writing, finding various information, and so on, I've developed my own writing style; I learned to enjoy writing, learned not to be afraid of ambitious projects, learned to find joy in making technical information (which our encyclopedia articles are by definition) accessible to common people who could actually learn something and feel proud for that. I will be happy if there is one more editor who writes high-quality element articles (not necessarily styled like mine) and I'll be glad to help you become one with an article of your choice. We all started somewhere, so don't be afraid to ask for help. Also, you can write me an email and I'll send you a few books that I find really useful when writing articles.--R8R (talk) 16:00, 29 September 2018 (UTC)

mass Chembox Hazards addition
has added, (hazars section) to some 40 element pages. For starters, I have moved them to an appropriate section (Precausions, Safety etc.). I don't know if these additions are FA-grade. There may occur a difference between text and hazard box (see Chromium: text says "see Chromium toxicity", hazard box says "No hazards"),. Also, all content in the box is code (phrase numbers, firediamond codes), but no description. -DePiep (talk) 17:36, 2 October 2018 (UTC)
 * I was on the fence about putting the box in when no hazards are listed, I will remove those(I did wonder if that chromium one would be confusing). But I find it odd that a hazard infobox would be a problem in an article about a dangerous element, as it is standard to include the GHS warning codes, fire diamond, ect in chemical articles. Tornado chaser (talk) 21:27, 2 October 2018 (UTC)

. Tornado chaser (talk) 21:30, 2 October 2018 (UTC)
 * I agree with that these boxes might be helpful to keep articles about hazardous elements in line with those of other hazardous substances. But  is correct to raise the issue about whether the quality of information and quality of presentation is suitable in featured articles. YBG (talk) 21:58, 2 October 2018 (UTC)
 * Looks OK now, in the section & empty ones removed. My worries are about aligning prose & hazard box (which might refer to different situations, like chromium(VI) etc.). Best would be if text and box agree nicely, as a mutual support, when relevant. I will remove the "at STP" footnote, as hazards refer to real life not laboratory situations. Will use NFPA_ref when applicable. -DePiep (talk) 09:02, 3 October 2018 (UTC)
 * Some elements might have other Chembox Hazards parameters applicable. (See parameter list, testcases demo). -DePiep (talk) 09:16, 3 October 2018 (UTC)

Single-use templates
There is a discussion over at that may be of interest to editors in this project. YBG (talk) 23:01, 15 November 2018 (UTC)

Dubnium at FAC
We're having one more FAC, and I hope it'll go well! I'll be very glad if you can find some time to comment the article. R8R (talk) 07:37, 27 February 2018‎ (UTC)
 * FAC Success, April 2018. Not on Main Page yet. -DePiep (talk) 22:31, 5 December 2018 (UTC)

infobox element: "a metal"
It crossed my mind that we could/should add to the infobox, in top: a "metal/metalloid/non-metal". -DePiep (talk) 22:17, 19 October 2018 (UTC)
 * We could add m/moid/nonm to the category in top of infobox: "non reactive" -> a non-metal". DePiep (talk) 21:30, 3 November 2018 (UTC)
 * No, mostly redundant and otherwise needless details (syaing Ln's are a metal is not top info).
 * ❌ Withdrawn. -DePiep (talk) 11:36, 16 November 2018 (UTC)

Isotopes articles first sentence
Over at MOS talk:FIRSTSENTENCE, I have asked for advice on whether and how to bold & link words like "carbon" and "isotopes" in Isotopes of carbon. -DePiep (talk) 14:29, 20 October 2018 (UTC)
 * Answered, that easy. I'll change the articles shortly. -DePiep (talk) 16:05, 20 October 2018 (UTC)
 * ✅. Unbolded, e.g. sulfur. -DePiep (talk) 23:07, 26 October 2018 (UTC)

Planned move: from (large cells) into (detailed cell)
I want to move Periodic table (large cells) --> Periodic table (detailed cells) (preliminary talk).

My reasoning: "large cells" does not count, not even as a DAB specifier. Point is, re Periodic table, that it can contain more info. -DePiep (talk) 23:46, 20 October 2018 (UTC)
 * ✅ - DePiep (talk) 20:03, 27 October 2018 (UTC)

IUPAC IUPAP JWP catchy name
is preparing nihonium for TFA. This from my talkpage:


 * Today's featured article/November 17, 2018

Hi DePiep, can you find an article for me that mentions that IUPAC subcommision and talks about its role in element naming? - Dank (push to talk) 22:51, 25 October 2018 (UTC)
 * If I understand your question: nihonium has links IUPAC and International Union of Pure and Applied Physics IUPAP. IUPAC/IUPAP Joint Working Party is a bit stubby indeed. Alse there are ref #23, ref #47. hth. -DePiep (talk) 05:49, 26 October 2018 (UTC)
 * Actually it is not a subcommission, but a cooperation ("JWP") between two independent organisations IUPAC IUPAP (Chemistry & Physics) - DePiep (talk) 06:03, 26 October 2018 (UTC)
 * Pinging User:Double sharp. I can understand the objection about the stubby article, although sometimes we do use links just to define terms. What's your preference? - Dank (push to talk) 14:14, 26 October 2018 (UTC) I mean, do you like the link, and do you like the current text, that says "subcommission"? That wording doesn't seem to be supported. - Dank (push to talk) 14:47, 26 October 2018 (UTC)
 * It is complicated, especially in this case because while IUPAC and IUPAP were working together, IUPAC went ahead and published the acceptance of the discoveries without wating for IUPAP. Perhaps it should instead say "In 2015, the International Union of Pure and Applied Chemistry (IUPAC) announced recognition of the element and assigned the naming rights to Riken"? Although that has the problem of not mentioning that IUPAP ought to play a role in these things, that may be a bit too much to explain in a blurb. Double sharp (talk) 06:40, 28 October 2018 (UTC)
 * It's up to you guys, but I think I agree, that's too much for a blurb. - Dank (push to talk) 13:07, 28 October 2018 (UTC)

What would be a nice short description/link for the JWP? -DePiep (talk) 15:41, 26 October 2018 (UTC)

Nihonium TFA
Yesterday, 2018-11-17, nihonium was TFA. Day page visits went from ~1,000 to 47,000. . Nice and great. - DePiep (talk) 21:16, 18 November 2018 (UTC)

Style errors
It seems that every single one of the articles about isotopes of elements violates the manual of style. Links should not be placed in the boldface reiteration of the title in the opening sentence of a lead ... If the article's title is absent from the first sentence, do not apply the bold style to related text that does appear. And yet is all of them that I've looked at, the name of the element is in boldface with a blue link, and sometimes the word "isotope" is also put in boldface. I have fixed some of them; people should fix the rest. 51.6.138.66 (talk) 20:38, 26 October 2018 (UTC)
 * Yes. See earlier Wikipedia_talk:Manual_of_Style/Lead_section, and MOS:BEATLESINUS. Will happen. - DePiep (talk) 21:27, 26 October 2018 (UTC)
 * ✅ - DePiep (talk) 18:29, 27 October 2018 (UTC)

Fermi to be TFA
Enrico Fermi is to be Today's Featured Article (TFA) on December 10, 2018. - Let's check the blurb & artice (both linked), I suggest. -DePiep (talk) 21:33, 5 December 2018 (UTC)

2012, 2014, 2016 and so ... 2018
Here it is again.


 * -DePiep (talk) 02:00, 19 November 2018 (UTC)
 * I'll be sure to expect it for 2020. ^_^ Perhaps we could add a note on the template pages linking to all the previous discussions to alert people who are confused by our setup? I know the talk pages already link to the previous TfD's, but perhaps having a noincluded warning on the template might help further. Double sharp (talk) 03:46, 19 November 2018 (UTC)
 * Hi. I am back from read-only into read-write mode. -DePiep (talk) 22:27, 3 December 2018 (UTC)

User Eudialytos
User Eudialytos is "helping" by putting a lot of unformated references in the articles about elements. Either we reformate all the insertions or we delete them? Here the Contributions of Eudialytos.--Stone (talk) 20:07, 6 October 2018 (UTC)

Critical point
I propose that we put the critical temperature and pressure in the infobox for elements. Eric Kvaalen (talk) 08:45, 17 December 2018 (UTC)
 * We've added this to Chembox (see ammonia). I'd like to read if we think it is relevant enough, say a defining property. -DePiep (talk) 10:32, 17 December 2018 (UTC)
 * We already have a field for the critical point. See the infobox for hydrogen, for example.--R8R (talk) 14:46, 17 December 2018 (UTC)
 * Lol. My mistake. -DePiep (talk) 16:22, 17 December 2018 (UTC)

Chromium
I started a discussion on a near word for word copy of parts of our chromium article on the talk page. Please have a look. --Stone (talk) 11:43, 25 November 2018 (UTC)
 * re . Discussion is here (I assume); results invisible. Or did I miss something? (I'm serious) -DePiep (talk) 21:00, 11 December 2018 (UTC)
 * See the talk. Is solved now (error on their side, not WP side). -DePiep (talk) 02:26, 15 December 2018 (UTC)

Old stuff. To be cleaned?

 * - DePiep (talk) 00:46, 28 December 2018 (UTC)
 * ✅, and merged SimpleNuclide into SimpleNuclide2 (they are the same now) . -DePiep (talk) 14:01, 28 December 2018 (UTC)
 * Same for ComplexNuclide and ComplexNuclide2 (no more errors for ). -DePiep (talk) 14:14, 28 December 2018 (UTC)
 * The merge was correctly reverted by  . Cause of the error: while the usage-category was empty when I merged, there still were 62 uses in mainspace not categorised apparently. I will continue cleaning up to merge.
 * merge can continue, does not have this issue. -DePiep (talk) 22:00, 28 December 2018 (UTC)
 * The 62 transclusions were in all namespaces. I fixed the two in mainspace, and three others in talk namespaces. I see you got some user sandboxes. So, I flipped my reverts back. – wbm1058 (talk) 22:42, 28 December 2018 (UTC)
 * After checking OK now: merged SimpleNuclide into SimpleNuclide2 (they are the same now). -DePiep (talk) 22:45, 28 December 2018 (UTC)

Infobox element: oxidation states data set
I have centralised all oxidation state (OS) data used in the s. Apart from some textual cleanups and ref name refinement, no changes were made (i.e., infobox shows the same OS information as before). Data set is in Infobox element/symbol-to-oxidation-state; see also there for a nice overview (all elements listed). -DePiep (talk) 18:25, 11 November 2018 (UTC)


 * This is the oxidation states data list: . This is where you can edit the content (individual element text; replacing old parameter input oxidation state).
 * Repeated references (eg Haire) best be exact copies in that page. However, a ref error message wrt this, usually only appears on that data page, not in article. -DePiep (talk) 12:59, 13 November 2018 (UTC)

Check List of oxidation states
We can now easily check the infobox data with List of oxidation states of the elements. I have created page /tablecheck to compare the two lists. Please discuss here. -DePiep (talk) 18:25, 11 November 2018 (UTC)
 * 0 values: the List clearly mentions zero valence when it appears in a compound. Sounds like  a good rule to me. Must we add those to the infobox? -DePiep (talk) 18:27, 11 November 2018 (UTC)
 * The /tablecheck has many sources. Could be helpful. Expect the heavies (E100+) to be outdated (empty). -DePiep (talk) 20:35, 12 November 2018 (UTC)

Format proposals
The overview in Template:Infobox element/symbol-to-oxidation-state shows that currently various formats are in use. The issues are:
 * Add or remove the "+" for positive values (consistently everywhere)
 * Order from low to high or from high to low? (negatives to the left or to the right?)

Here is a 2015 discussion about this: Archive_20#oxidation_states:_"+1"_notation.


 * I propose to: 1. add a "+" for all positive values (more explicit), and 2. order by lowest first (natural order). -DePiep (talk) 18:32, 11 November 2018 (UTC)
 * I agree with DePiep; that will clear ambiguity. ComplexRational (talk) 01:58, 12 November 2018 (UTC)
 * I also will agree with this. Adding the "+" to the positive oxidation states is something that I've wanted to do for a while (to further contrast the "-1" charges from the "+1" charges instead of leaving it as "1"), but decided to leave it as it is in the past. The high-to-low/low-to-high fluctuation argument is something that I was never really bothered by, but if I really had to choose what should be standard, I'd pick low-to-high for logicality of order. UtopianPoyzin (talk) 14:49, 12 November 2018 (UTC)
 * Order direction: less important; same order everywhere: hugely important. I have prepared these proposed standards in the sandbox, line-by-line comparable with current (random) formatting. See /testcases. -DePiep (talk) 17:37, 12 November 2018 (UTC)
 * I agree with adding the + sign, and with consistent ordering; ambivalent about the direction, but agree that it should be the same throughout. I am wondering if there might be some benefit to using color or some other technique to highlight the most common states - or to de-emphasize the unusual states. My concern is that the consistent order might appear to emphasize a less common state by placing it first. But even if that were not the case, being able to emphasize more common states and/or de-emphasize less common states would be helpful. YBG (talk) 17:47, 12 November 2018 (UTC)
 * The main states are bolded right now, but that is not enough I understand? The data row is already crowded, especially when there are many states, refs and/or 'predicted'-note values (heavies). Adding more graphical effects could hinder not help. I have this thought (to be developed later): we can add to the top infobox section the simple data row "Main oxidation states: +2, +3, +6" (for Fe). The full, detailed & sourced list stays where it is. As you can see, these are already in the testcases ;-). Also available: the Mendelevian value (Roman group numbers, but not A/B). - DePiep (talk) 18:06, 12 November 2018 (UTC)
 * It might be nice to also change the font color or background color. But its not a super big deal, bold is great, just that sometimes it doesn't stand out as much as I would like. But I wouldn't want to make it more busy than it already is, so if a solution doesn't jump out right away, feel free to ignore this comment. YBG (talk) 19:27, 12 November 2018 (UTC)
 * Is it possible to create two infobox parameters - main oxidation states and other/less common oxidation states? If so, it may prevent the assumption of an obscure oxidation state as the main one when reading a left-to-right list, and the use of the word main may provide additional emphasis rather than a simple bolding in a list that could indeed crowd a row in infoboxes. ComplexRational (talk) 19:56, 12 November 2018 (UTC)
 * That may have some promise. It would cost an extra line in the infobox, but it might make the long lists appear less busy. It would require adding to the data template. We should, however, consider the edge cases: what would we want to display if there were only two oxidation states, a main one and an obscure one. In such a case, would we want to expend two lines, or would another technique work - like bolding the main one and parenthesizing the obscure one.  Overall, having two one solution for elements with just a few states and a different solution for elements with lots of states might be a good thing. YBG (talk) 20:13, 12 November 2018 (UTC)
 * (ec) re ComplexRational These two 'parameters' already exist Face-smile.svg. In the /testcases you see an extra column "main", eg Fe has "+2, +3, +6". And the values are read from two lists ("main" and "complete"), no need to enter them! (technical follow later) When this new list is a stable & cleaned up in a few days, I will show new options here (eg extra data row "main: ..." indeed). Meanwhile, one can improve the data present; the data needs a thorough check and impovement (being put out in the open recently). This content quality for OSs I pointed out above in.
 * So, by calendar: First I will clean up & order the current data (on my own, a few days). After that, we will discuss improvements suggested here. Also, I will ask for quality controls & improvement (more links like for C OS=0!). If you want to engage: start to discover & look & check the /testcases. -DePiep (talk) 20:33, 12 November 2018 (UTC)
 * ✅ these two. See overview: use "+" in value "+2", and order low-to-high. Other changes (wrt prediction formatting, main indicators) are under discussion and not prejudiced. -DePiep (talk) 12:52, 13 November 2018 (UTC)

Formatting question - predictions

 * , : OK, so here is a formatting question at hand: How to format a predicted oxstat? Today, these situations and forms appear (&lt;ref>'s omitted):
 * B : 3, 2, 1, −1, −5
 * Rf: 4, (3), (2)
 * Og: −1, 0, +1, +2, +4,

That is: B has no predictions, Rf has some not all, and Og has predictions only. Some prediction are "main" oxstats. Technically, the way it is written is OK (the info is there). My question is: I would like a more simple and clear way, reusable. Especially the diff between Rf and Og is too big. I can think of:
 * 1) Use brackets per predicted oxstat always: (−1), (0), ... and/or
 * 2) Use sharp commenting: always add "  values are predictions"

I would be very pleased with some consistent, reusable and reader-easy clarifying form (reader does not need to think or research or look for legend). That's all I ask. -DePiep (talk) 21:19, 12 November 2018 (UTC)


 * Side note: Because this edit included U without a signature, the pings were not sent. As I understand it, for a ping to be sent, the user reference and the signature must be added in the same edit. Took me a long time to get that figured out. YBG (talk) 21:37, 12 November 2018 (UTC) Yes. OK. DePiep.


 * (ec) Thought of this, re refs & the examples:
 * 3. When predicted, the refs should be at the end (they cover all) not after the last "+5" value. OTOH, when a ref pertains to an explicit value, it should be "+3&lt;ref> 4, ..." of course. But an oxide type comment ("basic") goes after generic refs. -DePiep (talk) 21:42, 12 November 2018 (UTC)
 * 4. Can be built in the infobox: when a prediction is present, the lefthand text can say like: "ox stats,  = predicted". -DePiep (talk) 21:50, 12 November 2018 (UTC)


 * I'd try something like this in the infobox (oxidation states parameter) with all the predicted oxidation states in a parenthesized set (I think this is slightly cleaner than having (a), (b), (c), ... or the legend at the end):
 * main: +3 (predicted: +4) others: +2 (predicted: +6)
 * If there are no predicted oxidation states, there simply will be nothing in parentheses, and should there only be one in a category (e.g. for the alkali metals), it is still clear which is dominant without looking for boldface or notes. In cases like F where there only is one oxidation state, we can do away with the main and others labels. I hope this format can clearly distinguishes everything; if not, feel free to suggest alterations. ComplexRational (talk) 22:22, 12 November 2018 (UTC)
 * Looks very nice. May I suggest: 1. only italicise the word predicted (or prediction) + its brackets, nothing else. Is what we do elsewhere in the infobox (Nh), and prevents suggesting that these values mean something different from upright (roman font) values - they are not. 2. Also, we can add the oxide type to the main set. And 3. maintain bolding if it still is a crowded view (then do so with all elements); 4. rm brackets oxide type?
 * main: +3 (predicted: +4) others: +1, +2 (predicted: +6)
 * -DePiep (talk) 10:26, 13 November 2018 (UTC)
 * Time to call in, who works extensively with these predictions. -DePiep (talk) 10:32, 13 November 2018 (UTC)
 * I would prefer to have the oxidation states going in numerical order all the time from lowest to highest. The main ones would be bolded, and the predicted ones would be parenthesised even if everything is predicted. So the three examples you give would look like:
 * B: −5, −1, +1, +2, +3
 * Rf: (+2), (+3), +4
 * Og: (−1), (0), (+1), (+2), (+4)
 * To be very clear, I'd also like a note to appear saying that everything in parentheses is only predicted. Double sharp (talk) 10:41, 13 November 2018 (UTC)
 * Earlier proposals now done. You are all invited to edit individual entries ("parameter" values), see link in section top. When this talk about main & predictins has fleshed out, I will take care (in the sandbox). -DePiep (talk) 13:02, 13 November 2018 (UTC)

(arbitrary break)

 * Note on the bolding the main values looks good in regular text, see e.g. the examples in this subthread. However, in an infobox the font is reduced by 85% 88%. That severly reduces the visibility of the marking (as pointed to before). Maybe we can stress them typographically more (enlarge? double bold?), OTOH I am not happy with adding graphics (like colors) for this, because that's too much circus, and we're supposed to explain any such graphic effect. -DePiep (talk) 08:00, 14 November 2018 (UTC)


 * Here a mockup infobo x with three options (if we kee pthe bold-in-order arrangemetn): current, 100% (b) and 112% (c) font size. (Don't forget to check the effect on mobile). -DePiep (talk) 20:08, 14 November 2018 (UTC)
 * I'd say the second (100%) as it provides enough emphasis without making the others too difficult to read, though for some reason, even 100% seems larger on mobile relative to the normal text than on a computer screen. ComplexRational (talk) 00:33, 15 November 2018 (UTC)
 * I second on the second. UtopianPoyzin (talk) 04:59, 15 November 2018 (UTC)
 * On mobile, but using the desktop view, the slight enlargement on (b) is hardly noticeable, but the bigger one on (c) seems too noticeable. I favour just (a) as we basically have now (just using bolding), with (b) as a second choice. Double sharp (talk) 05:12, 15 November 2018 (UTC)
 * With me, on desktop the 100% version is significantly better (it actuually does stress those). Cold it be that you are too familiar, iow you doen't need the stressing that much? I'm with YBG, who stated that (on desktop) they do not jump out too much.
 * Mobile browsers do different things with font size (like, ignore settings), but on my mobile (iPhone) is see: well stressed in live version @88%, same @ 100%, so no harm there when going to 100%.
 * I find 100% best for both screens. -DePiep (talk) 07:02, 15 November 2018 (UTC)
 * I've added an option (d) that uses separate lines to distinguish common and less common states. With 7 less common states, two of which have individual refs, it seems just barely passable. I would envision common, less common, and predicted states to each be on separate lines.  If all were predicted, they would be on the first line, followed by a comment saying "(all predicted)" or the like. Thoughts? YBG (talk) 07:49, 15 November 2018 (UTC)
 * I still prefer having the common ones together with the not-so-common ones. In an article, I would normally not split discussion of compounds by common and less common oxidation states, but by classes (e.g. oxides, halides, hydrides, organometallics, etc.). Indeed, the oxidation state in question would have to be very singular and exotic, and go against almost all the rest of the chemistry of the element (which would mean something as exotic as HgIV, if we were sure that existed; the alkalides might come close), for it to be a good idea to do that. Even in the places where I would normally split by oxidation states (coordination compounds), I would do it in numerical order instead just like Greenwood & Earnshaw, as the magnitude of the oxidation state has greater predictive value on the chemistry than its commonness. For instance, FeVI is common like FeII and FeIII, but it acts more like FeIV and FeV (see high-valent iron). Double sharp (talk) 15:48, 15 November 2018 (UTC)
 * Good idea to add (d), indeed we have multiple choices to make. I have added Sg as it tests example (d) more extremely; Hs is even more extreme. re YBG, I would keep the -notation, and keep "predicted" in the LH label consistently. This is structured not free text data. To be solved with option d: where to write the oxide type? (today does not occur i.c.w. preditions). And where a ref that covers all values?
 * re Double sharp: OK if the article describes them by orther ordering like halides, but the infobox is a summary without that detail and so that detail does not need to be shown at all, not even indirectly as a secret ordering principle. Minor advantage of the multi-row notation is that the label (lefthand text) is cclarifying. (Although infobox size should be kept minimal, as this all is in the lede, I'd accept more oxidation state rows and remove lesser stuff like vapor pressure at any time). I would go with single-line order, as it is higly and simply structured and so very clear to the reader even at first sight. More so if we use bold&big font (100%). -DePiep (talk) 17:06, 15 November 2018 (UTC)


 * ✅ in live infoboxes: all predicted values have brackets (see meitnerium) . (late sign: ): DePiep, 16 November 2018

Oxidation state changes: Roundup
See also the Overview list


 * Standardised ~along the lines Double sharp described, for all (order low-to-high, use "+", just bold main OSs, always parenthesise predicted ones and add a note for this).


 * Kept: "(predicted)" when all are, "(parenthesized: predicted)" when some are.


 * Not applied: split into major/minor data entries, split order by major/minor, enlarge or color major OSs.


 * Todo: align the list with article List of OS (tablecheck compare page).


 * Todo: write "parenthesiSed" when engvar=en-GB.
 * I am under impression the word parenthesis and its derivatives are less commonly used in British English than just brackets. Can anyone confirm/contradict this?--R8R (talk) 10:04, 11 December 2018 (UTC)


 * Indeed: . So: en-GB, en-OED to write "bracket/ed" here. Being en-GB/OED, we can omit the "round" here I pose. -DePiep (talk) 10:29, 11 December 2018 (UTC)
 * I'll add my confirmation that in British English, ( and ) are brackets. If you want to be precise, they are round brackets, but I think you can drop the "round" if there are no other types of brackets around. "Parentheses" will probably be understood as the American usage is widespread, but it is definitely less common. Double sharp (talk) 10:34, 11 December 2018 (UTC)


 * OK then. An Engvar setting applies to the whole article, so when set to British the context (other spelling variants) are British too. -DePiep (talk) 10:39, 11 December 2018 (UTC)
 * ✅, see documentation. Articles affected: 0 having en-GB/OED in this ;-). -DePiep (talk) 13:26, 11 December 2018 (UTC)
 * How come? Phosphorus uses BrE.--R8R (talk) 13:31, 11 December 2018 (UTC)
 * And also thorium. And these two are just the ones I know, maybe there are also a few other ones. Why is your count 0?--R8R (talk) 13:34, 11 December 2018 (UTC)
 * Only those elements are affected that have mixed OS's (known and predicted). See Overview list: only E104 Rf and above. But none of these mixed lists are in en-GB/OED. (Fl is, but has an all-predicted list of OS, and so does not have the "bracket" explanation; as described in bullet 2 "Kept" in my OP). HTH. -DePiep (talk) 13:39, 11 December 2018 (UTC)
 * Ah, I see. Do note, however, that not all oxidation state lists are extensive: radon's, for instance, does not mention the future possibility of +4 and +8.--R8R (talk) 13:55, 11 December 2018 (UTC)
 * In that Rn case: an editor adds these predictions to the data list; will parenthesize them with round brackets; and will correctly add parenthesized in there to describe the meaning of those brackets. From here everything is easy & automated: would radon have set en-GB (quod non), the infobox template will show text as expected in en-GB. The point of this whole OS data set is, that it is centralised and standardised. Good for checking consistency over the elements (done for elements, todo for OS list as mentioned), and reusability (in other articles). Same for engvar variants in element infoboxes: centralised, maintained, checked. -DePiep (talk) 14:12, 11 December 2018 (UTC)
 * The engvar effects in our element infoboxes are listed here. -DePiep (talk) 14:14, 11 December 2018 (UTC)
 * Test: setting seaborgium into en-GB ; effect = as intended 'brackets'. (However. Test originally failed because was not prepared for engvar}} ... now is. Would be a problem if such an article would switch engvar). -DePiep (talk) 14:42, 11 December 2018 (UTC)


 * Todo: check for link-able oxidation states (single article).

-DePiep (talk) 09:49, 11 December 2018 (UTC)

Opinion: I am still not happy with the reduced size (infobox=font-size:88%). With this size, the bold (major) OSs do not stand out enough on desktop screens. This 'hides' this important info. I prefer enlargement (example (b)=font-size:100%), not coloring. , you think we can pursue this once more without being too -whatstheword-? -DePiep (talk) 09:49, 11 December 2018 (UTC)
 * Not sure what you mean by "whatstheword", or why you directed it to me in particular. No matter, it prompted me to think about it.
 * I am sympathetic about the concern about bold is not distinctive enough, but I have not actually checked it myself. As far as parens go, I'd prefer this:
 * All OS predicted: (predicted)
 * No OS predicted:
 * Some of each:  values are predictions
 * This nicely avoids the ENGVAR issue, but IMO more importantly, it is briefer. YBG (talk) 17:23, 11 December 2018 (UTC)
 * re: I pinged you because you mentioned the bad visibility of bold values (less distinctive) in these OS lists. -whatstheword- is like pushy, demanding, boring, flogging a dead horse. IOW: question may be dead by now.
 * re predictions: I see no need to add boldness or attention grabbing as you do here (bolding a word that is in brackets is contradicting, even). Also, writing italics (predicted) is gently used elsewhere too. Latter option " ..." puts the note outside of brackets, while I prefer brackets for all (consistency). Preventing engvar is no aim, we can & do handle it. - DePiep (talk) 17:33, 11 December 2018 (UTC)
 * At this moment I see no open issues. I plan to archive this thread in a few days (some opposition, see below -- DePiep). Editors can reopen any point of course. -DePiep (talk) 16:25, 14 December 2018 (UTC)
 * If you have no concerns about bold being visible, that is good enough for me. Regarding my notes above, I used bold to distinguish my suggested wording from the label. I agree that it should not be bold in the infobox; italic is definitely preferred.
 * I see no problem with having the note outside the brackets. Saying " values are predictions" is comparable to saying "ƒ fissile" by my way of thinking. If this is not acceptable, what about "(values) are predictions"? YBG (talk) 16:52, 14 December 2018 (UTC)
 * re : I'm still not sure if the bolding is clear enough to mark the main ox states, while being a very important thing we want to signal. Problem is that is we ask our local WT:ELEMENTS editors, they do not see the issue (everyone here already knows these important ones). We editors here should take a fresh look as if being an interested first reader. For this reader, is it clear enough? (I don't think so). So I will start a (non-assaulting ;-) ) campaign once more here.
 * re you remark: " values are predictions": I disagree for reasons mentioned (inconsistent notation pattern). If you want to change that default text, go ahead and start a campaign. -DePiep (talk) 20:10, 15 December 2018 (UTC)

One more call: brackets & text improvements?

 * Looking at that page (standard texts), I see how my proposed text doesn't fit in with the others. This isn't such a big deal since the different comment options never appear together, but it still is a bit untidy. Overall I think "(values) are predictions" would be a bit better than " values are predictions".
 * I also see the root cause of the problem: we are using for two completely different things: (1) to mark predicted values, and (2) to surround the comment text. This works OK provided we don't use the  in both ways in the same context, as I was suggesting.
 * So I now suggest marking the comment in a different way that doesn't require . Perhaps an em dash between the list and the comment? That, with the italics should be enough to distinguish between the list and comment.
 * This would be the very best solution if there is an absolute requirement to have all comments formatted and punctuated identically, which I think is nice but not absolutely necessary as the comments never appear with each other.
 * YBG (talk) 20:51, 15 December 2018 (UTC)
 * My reply (!vote): Seeing the complicated texts in the ox state value (that is, having multiple textual parts like refs, brackets, comments and the actual numbers), I prefer each and all suffixed comments to be in outer brackets (as they are today). That is to add regularity in formatting, which IMO helps readability. -DePiep (talk) 20:00, 29 December 2018 (UTC)


 * By the way, I hereby accept the consensus to include the entire note in parens.
 * Some side comments that have no bearing on the decision
 * What about these rewordings? No strong preference here, just offering another option. (Note: I've italicized the comments and the enclosing parens, I think it looks ever so slightly better, but my only strong feeling is that it be done consistently). By the way, these two suggestions are independent and each stands on its own.
 * (parenthesized: predicted) → (predictions parenthesized)
 * (predictions) → (all predicted) or (all values predictions)
 * Also, I see that parenthetical comments are used both relative to predicted values and to describe the oxide status. Is there any chance of needing to use both types of parenthesized comments?
 * Is the symbol-to-ox-state to be used only (or primarily) for infoboxes? If so, it would be good for the testcases to be in the same font size that they will appear in infoboxes.
 * In the testcases table, there is no need for the "wl" (wikilink) column. Why not wikilink the "Symbol" column - or the "Name" column (or both even)?
 * Thanks again for everyone's participation in this discussion. One of the things I appreciate most about this project is that we are willing to have a slow conversation in order to reach a consensus. It takes a while, but the end product is well worth it. YBG (talk) 00:56, 30 December 2018 (UTC) ( post was split into two sections from single post DePiep (talk) 08:31, 30 December 2018 (UTC)) Link showing split: special:diff/875973438
 * I don't see "!votes" weighing in to change this pattern. Must say: prefab texts are not in concrete either. Especially those for one or two element can be refined more easisy.
 * Concluding: no change from here. -DePiep (talk) 00:39, 5 January 2019 (UTC)

One more call: infobox marking the main ox states is good?
The issue: the bold numbers (main oxidation state values) do not stand out. The value listing also has parts like -bracketed values, x[12] reference indicators, and (comments): textual elements that undo the marking of these major values.

The proposal: enlarge the bold values (from 88% infobox size back to 100%). Shown in 2nd half of the demo. The goal is the same as for the bolding: point out major values. And as with bolding, there is no explanatory clarification present.

Please take a look with 'fresh reader's eyes', forget what you (experienced element editor) already expect to see.

Notes: This issue only exists in desktop view. Mobile view does not have this issue (mobile view does not change font-size in infoboxes). And: the demo shows O, Fe and Sg top cover most list situations.

-DePiep (talk) 19:09, 29 December 2018 (UTC)


 * I support enlarging the bold values; it is much clearer to the reader this way, especially when browser zoom is less than 90%. ComplexRational (talk) 23:03, 29 December 2018 (UTC)
 * I support enlarging bold values also, though deficiency of merely bolding isn't as great in the examples here as I'd feared - in fact it seems better here than it does on the testcases page. By the way, I hereby accept the consensus to include the entire note in parens.
 * Some side comments that have no bearing on the decision
 * What does it mean for some predicted values to be 'main' and others to not be 'main'?
 * Is the symbol-to-ox-state to be used only (or primarily) for infoboxes? If so, it would be good for the testcases to be in the same font size that they will appear in infoboxes.
 * In the testcases table, there is no need for the "wl" (wikilink) column. Why not wikilink the "Symbol" column - or the "Name" column (or both even)?
 * Thanks again for everyone's participation in this discussion. One of the things I appreciate most about this project is that we are willing to have a slow conversation in order to reach a consensus. It takes a while, but the end product is well worth it. YBG (talk) 00:56, 30 December 2018 (UTC) (YBG post was split into two sections from single post DePiep (talk) 08:31, 30 December 2018 (UTC)) Link showing split: special:diff/875973438
 * re
 * 1. When do we state "main" (=show bold & bigger): I don't know exactly. There might be border cases. This page shows differences between data list and the listing article.
 * 2. The subpage (data list) /symbol-to-ox-state is primarly aimed at infobox (must be OK in there). Also allows reuse in other pages (same as with, for example, atomic weight). Font size now reduced to 88% (infobox size) for the infobox data column.
 * 3. Removed one redundant column. The names etc. are gained automatically, and changing this is on the todo list. Not problematic now.
 * I add: typographically, this is a compromise solution. One should not mix font sizes and bolding easily. Also, the meaning of this effects is not mentioned (no explanation "bold = main" available). However, the overall picture (the info we want to show in the infobox, and how to) is complicated and requires extra options. -DePiep (talk) 15:10, 3 January 2019 (UTC)
 * Prepared in sandbox, see full test table. To go live shortly. -DePiep (talk) 15:10, 3 January 2019 (UTC)
 * ✅. Looks good to me. -DePiep (talk) 00:46, 5 January 2019 (UTC)

ref dump

 * (demo references are here)