Misplaced Pages

Revision as of 06:37, 25 December 2020

Elements Project‑class

This page is supported by WikiProject Elements, which gives a central approach to the chemical elements and their isotopes on Misplaced Pages. Please participate by editing this page, or visit the project page for more details.ElementsWikipedia:WikiProject ElementsTemplate:WikiProject Elementschemical elements Project This page does not require a rating on Misplaced Pages's content assessment scale.

Misplaced Pages:Misplaced Pages Signpost/WikiProject used Misplaced Pages:Misplaced Pages Signpost/WikiProject used

This is the talk page for discussing WikiProject Elements and anything related to its purposes and tasks.

Put new text under old text. Click here to start a new topic. New to Misplaced Pages? Welcome! Learn to edit; get help. Assume good faith Be polite and avoid personal attacks Be welcoming to newcomers Seek dispute resolution if needed

Archives: Index, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63Auto-archiving period: 2 months

FA	A	GA	B	C	Start	Stub	FL	List	Category	Disambig	Draft	File	Portal	Project	Redirect	Template	NA	???	Total
29	0	96	103	127	95	34	0	172	307	3	0	116	1	22	8,462	103	11	0	9,681

31

31 featured content items: Featured articles (30) H, He, O, F, Zn, Ge, Y, Nb, Tc, Xe, Cs, Pb, At, Fr, Th, U, Pu, Cf, Db, Hs, Nh, Ts, Og, noble gas, metalloid, periodic table, heavy metals, radiocarbon dating, history of aluminium, island of stability Featured topics (1) period 1 elements Signpost interviews: 2011, 2013 view edit Detailed list Article quality in the periodic table Other high-quality articles

Major: Interesting publications

Eric Scerri YouTube workshop

This workshop (1h 41m) was conducted remotely via Zoom. 150+ people from 16 countries registered. It was organized by the American Chemical Society in association with Yale University.

Topics discussed by Eric include:

• teaching 3d-4s orbitals
• Madelung Rule
• differentiating electrons
• Group 3
• HFSS (half-filled sub-shells).

--- Sandbh (talk) 01:27, 5 November 2020 (UTC)

Ten most influential chemists today

Here. --- Sandbh (talk) 21:49, 5 November 2020 (UTC)

FYI: Hydrous oxides, hydrated oxides, hydrous hydrated oxides, hydroxides, and hydrous hydroxide

Here’s an extract from Bray & Latimer 1949, A course in general chemistry, 3rd ed., Macmillan, New York, pp. 89–90. It’s the first time I’ve seen a concise explanation of these five terms, in one place.

”The oxides of Group I and of the heavier elements of Group II react readily with water to form hydroxides which are soluble strong bases; i.e., their solutlons contain hydroxide ion and the ions of the metals. The oxides of the nonmetallic elements are soluble in water, with but few exceptions, to form acid solutions. This difference in behaviour depends mainly upon the size and charge of the positive kernel or ion.

The oxides of the remaining metals, in general, do not react with water; but compounds (hydrous oxides, hydrated oxides, or hydroxides) are usually formed when solutions of the positive ions are made alkaline. In a few cases the precipitate is the anhydrous oxide. When the precipitates consist of particles of the oxides with an indefinite amount of absorbed water they are called hydrous oxides, e.g., SnO₂ • xH₂O. If there is a definite number of water molecules combined with the oxide, the name hydrated oxide is applied, e. g., Al₂O₃ • H₂O. However, the hydrated oxides may absorb an indefinite amount of water to form gelatinous precipitates or hydrous hydrated oxides. In some cases the precipitate has a definite hydroxide structure, especially with ions having a +2 charge, e. g., Fe(OH)₂ and Mg(OH)₂ and these may be correctly called hydroxides.

If, as is often the case, additional absorbed water is present they may be referred to as hydrous hydroxides. The character of a hydrous precipitate varies with the method of preparation and changes upon ageing and drying. For the sake of simplicity, we shall often write the hydroxide formula without implying that this is correct.”

—- Sandbh (talk) 08:33, 30 November 2020 (UTC)

@Sandbh: Yup, I like having the terms all defined together too.

Of course, size and charge of the positive ion as Bray & Latimer state is exactly Fajans' rules. It's nice indeed to see an RS offering this up as the rationalisation. ^_^ Double sharp (talk) 12:17, 30 November 2020 (UTC)

Water into peroxide, by itself

Stanford scientists find water can transform into hydrogen peroxide when condensing on cold surfaces. Sandbh (talk) 10:21, 24 November 2020 (UTC)

Subdividing the nonmetals in 1844

I stumbled on this source…

• Dupasquier A, Traité élémentaire de chimie industrielle, Charles Savy Juene, Lyon, pp. 66–67

…and was struck by how advanced it was, for its time—a quarter century before DIM published his periodic table.

Dupasquier was a doctor, pharmacist and chemist. Per Google translate, here's his nonmetal taxonomy:

"There are 13 metalloids: nitrogen, boron, bromine, carbon, chlorine, fluorine, hydrogen, iodine, oxygen, phosphorus, selenium, silicon and sulfur.

To facilitate the study of metalloids, they will be divided into four groups or sections, as in the following table:

Section	Name	Constituents
1st	Organogens	O, N, H, C
2ns	Sulphuroids	S, Se, P
3rd	Chloroides	F, C, Br, I
4th	Boroids	B, Si

Organogenic bodies, as their name indicates, are those which constitute, as essential elements, organized matter, plant or animal. Plant matter is mostly composed of oxygen, hydrogen and carbon; animal matters generally have one more principle, nitrogen. There are exceptions to this rule; nitrogenous vegetable substances are known, and some matters of animal origin, which do not contain nitrogen.

The prototype of sulfuroids is sulfur; their general characteristics are: To be solid, volatile at an average temperature between 100 degrees and red heat, very combustible, and very inflammable. They have more affinity for the oxygen gene, with which they form powerful oxoacids, than for hydrogen. They can be divided into two secondary sections: 1 ° sulfuroids forming acid combinations (hydroacids) with hydrogen: sulfur, selenium; 2 ° sulfuroid in basic combination with hydrogen: phosphorus.

Chlorine is the prototype of the third section; the four bodies which constitute it have between them a great analogy of nature and properties; their main characteristics are: To be gaseous or very volatile at ordinary temperature; to give off a strong odor similar to that of chlorine; to have a lot of affinity for hydrogen and very little for oxygen; to form strong hydroacids, gaseous or very volatile, very greedy for water, and easily decomposable oxoacids; finally to constitute by their union with the metals combinations haloids. (See salts in general.)

The fourth section, which is boron-type, contains only two bodies, both fixed, odorless, of dull brown color, both forming solid, fixed, odorless oxoacids, and other acids, gaseous and volatiles, by their combination with fluorine and with chlorine."

Comments

• The organogens + the sulfuroides = our other nonmetals
• F was not isolated until 1866
• The discovery of the noble gases occurred from 1895 onwards.

Sandbh, 00:25, 8 December 2020

A nice pair

--- Sandbh (talk) 00:25, 12 November 2020 (UTC)

References

1. (1992)
2. (2018)
3. (2016)

How chemistry and philosophy combine

https://academicinfluence.com/interviews/chemistry/eric-scerri Sandbh (talk) 22:38, 21 December 2020 (UTC)

La v Lu paper, open access (2020)

"Averaged Scale in Electronegativity Joined to Physicochemical Perturbations. Consequences of Periodicity"

Antonio Campero and Javier Alejandro Díaz Ponce
ACS Omega 2020 5 (40), 25520-25542

"Concerning the periodicity in electronegativity of La and Lu, La belongs to three tendencies: group G3, period P6, and lanthanides. In the case of Lu, it only belongs to the last two, but the other lanthanides to any of the three sequences, and Hf only belongs to period P6. Then, the orbital f functions as an orbital intern between the 6s and 5d orbitals to give continuity in the electronegativity from groups G1 and G2 to group G3 in period 6, not considering the increment of atomic number Z. This continuity is not present for orbital d between orbitals s and p for the post-transition elements. We conclude then that the natural sequence of atomic number Z in the large periodic table, as Scerri has explained,(52c) implies that La is in group G3. The tendency of the electronegativity values in absolute values and ratios of the elements near to La and Lu, with their associated overall physicochemical properties, reaffirms this conclusion. Also, the electronegativity values determined in this work are correlated to similar chemical behavior of the ions through Knight’s, isodiagonal, vertical, and horizontal periodicity."

--- Sandbh (talk) 07:21, 23 December 2020 (UTC)

Major: Interesting publications, with discussion

Philosophy: Coming to grips with the fuzzy nature of chemistry

NB: I’ve prefixed the section title with "Philosophical:" so that those of us less enamoured by such matters can skip the whole section.

In this JChemEd article from 2002 , about electronegativity scales, the author warns us that certain central notions of chemistry are broached, in chemistry textbooks, from a set of diverse and not always concordant explanations.

18 years later, a couple of philosophers analysed that 2002 article .

Here are some extracts and observations from their open access article:

1. …on certain occasions, such as for scientific and didactic purposes, in order to avoid confusion either in learning or in the development of an experimental process, the simpler option is better; that is why traditional scales are so popular even nowadays.”

2. OTOH, “Real science is home to a wide variety of scales, and associated conceptualizations, that coexist in the scientific communities...The question is not one about which scale should be chosen, but is instead about the reasoning for choosing only one.”

3. “Objective and univocal truth is not an aim of scientific practice.”

4. "The teaching of a 'mummified' chemistry, free of conceptual problems and the associated debates, does not reflect the scientific practice."

5. "At the same time, we believe that scientific monism, according to which there is only one scientific story about the world that can be told, should be avoided as far as possible as well."

6. "There is a vast philosophical literature and a scientific practice that supports this perspective…objective and univocal truth is not an aim of scientific practice. Pluralism must engage in cultivating multiple scientific systems and lines of enquiry, as science is a multi-aimed enterprise, not the search of literal truth…"

7. "But why is it better to be pluralistic? Why keep multiple systems of knowledge alive? The immediate reason for this is the sense that we are not likely to arrive at the one perfect theory or viewpoint that will satisfy all our needs...If we are not likely to find the one perfect system, it makes sense to keep multiple ones."

Suggested implications for WP:ELEM
1. Writing about e.g. the chemistry of the elements and the periodic table is hard.

2. In many case there is no one true way. At the same time, "for scientific and didactic purposes, in order to avoid confusion…in learning…the simpler option is better."

3. The question will not be about which approach should be chosen, but instead about the reasoning for choosing only one.

4. The teaching of chemistry free of conceptual problems and associated debates does not reflect scientific debate.

5. Some pragmatism and judgement will inevitably required in order to accommodate the desirability for simplicity and the fuzziness of chemistry.

Grateful for your thoughts. Sandbh (talk) 03:37, 14 November 2020 (UTC)

Sandbh, when I said that your article touched on areas that are not sufficiently explored in the literature, this is exactly the sort of material that I had in mind. It is, IMO, central to some of the problems with writing an article on the PT, starting from considering who are the readers. It goes then to a more general question as to what the PT actually is. Is it a tool for summarising information about the properties of elements? Is it a tool for students to use to learn relationships between elements (like valency and what the formula of phosphorous(III) oxide is)? Is it a convenient summary form for relative atomic masses for use in carrying out stoichiometry? Is its purpose to provide electron configurations? Is it really meant for use by scientists and so should incorporate complexities even if these are confusing to students and lay people? Should there be different tables for different groups or different applications, each tailored to specific needs? Is including complexities an aid to student learning about multiple and co-existing models or is the confusion produced a hinderance – or is it both, an exemplar of the Perry Model in action?

There are a wealth of interesting question here, ones that are well-suited to the literature and rather less well-suited to an encyclopaedia while the discussions continue unresolved. If we had an article on the Perry Model, this would be a suitable topic for it, but the question for the project (and certainly for discussion at talk:Periodic table is how to take what is in RS and give it DUE weight to present a clear and readable article that provides our readers with what they want / need to know about the periodic table.

As a chemistry educator, I regularly think about these issues. I make sure to introduce students to the idea that a model is not the same as reality, that models are tools that do not fit / work well in all circumstances. A classic example (for me) is in introducing oxidation states. This is a model with great power for analysing redox processes, but at the same time is predicated on assumptions that are fanciful. I know that dichromate oxidation of 2-butanol to butanone involves reduction as chromium goes from Cr to Cr, but I also know that this model seeing dichromate as a Cr cation surrounded by oxide anions and with no covalency in bonding is ridiculous. I value the model of oxidation state for its utility in understanding redox chemistry, whilst simultaneously looking at it as a tool with limitations and not as a reflection or representation of reality. I have the freedom when teaching to explore this concept. However, as a Wikipedian writing about oxidation state, I do not have the freedom to work in content on models and representations of reality without RS on the subject with sufficient coverage to pass DUE for inclusion. Otherwise, I am engaging in OR and SYNTHesising what I know of models and learning and the literatures of those areas and applying them to a chemistry topic in a way unsupported by the literature.

For the Perry Model and multiple models, I introduce / explore these when discussing acids and bases as students often first try to choose between Arrhenius, Lowry–Bronsted, and Lewis models as if one is true. I encourage them to see each model as useful for different circumstances, each model being more complicated than the last but the simplest model that is valid being the wisest to choose for a situation. None of them are reality or truth, each is simply a model to understand reality, and choice of model is dependent on which best fits the situation. Like ArbCom being a sledgehammer and the right tool to use to crack some disputes, in circumstances where it is the wrong tool, its use may not have the desired result.

I also explore the breakdown of models in discussing gas laws and explain that the ideal gas law is a mathematically-constructed model to approximate reality, based on assumptions, and that when those assumptions are unjustified, the model's approximations to reality become unreliable.

I'm happy to chat more about such topics on one of our user talk pages, but for the continuation of the discussion here, I ask we concentrate on how the topic you raise can be implemented (if at all) in article space in line with policy, and to which articles it is supported by RS given DUE weight to apply. EdChem (talk) 04:19, 14 November 2020 (UTC)

PS: This is also overlapping with the comments made by DGG at the ArbCom case request – Who are our readers? Students? Experts in Chemistry? Lay persons interested in Chemistry? And, what is it that they need from us in terms of what content we cover and how? There is the problem, however, that if the RS literature consensus sees the PT as (say) a tool for chemists / experts to use in their work then WP must say so, even if that group is not our typical reader... and raises consequent challenges of covering RS and giving DUE while producing understandable and appropriate encyclopaedic content. To your point, Sandbh: yes, writing about chemistry and the PT is hard but we have guidelines from what RS tells us on what to say and from considering our readers in deciding how to say it. EdChem (talk) 05:02, 14 November 2020 (UTC)

EdChem, (btw, as background, I'm a biochemist who has also worked as a chemistry librarian). Since the PT is a tool used by chemists to organize and conceptualize their work, of course we should say so and explain how they use it. But the PT is also a teaching device, central to both the teaching of general chemistry and to what the public knows about chemistry, and we need to discuss this aspect also. (Similarly, we discuss other aspects, such as the historical aspects (best handled I think in the frWP article), and its use as the prototypical organizing device in other fields, sometimes seriously, sometimes humorously). This problem of the dual goals of being understandable and being accurate has long been a problem with all encyclopedic coverage in WP and in other encyclopedias of such fields as mathematicvs and much of physics. But I don't think this is so esoteric that it need to be split into two articles on the analogy of Introduction to general relativity .This problem of the dual goals of being understandable and being accurate has long been a problem with all encyclopedic coverage in WP and in other encyclopedias of such fields as mathematics and much of physics, and is now confronting us also with biology--and the subject of a recent arb case in which I did not participate--Misplaced Pages:Arbitration/Requests/Case/Medicine. DGG ( talk ) 06:08, 14 November 2020 (UTC)

EdChem; DGG, The PT article is not “mine”. I recall adding some content leading up to the FA process , but wasn’t the instigator of it.

The PT is a model, that’s all. The important thing is to explain the context for the particular model in use. This is relatively easy. The La form remains popular. The colour categories are the relatively popular ones: AM, AEM, Ln, An, TM, metalloids, halogens and noble gases. That just leaves the metals between the TM and the metalloids; and the nonmetals between the metalloids and the halogens (plus H).

Those categories illustrate the L-R progression in metallic to nonmetallic character going across the table, and the top-down increase in metallic character seen in most notably in the p-block. This is traditionally taught by contrasting the alkali metals with the halogens.

As long as I’ve used RS, I’ve never understood what the “undue” concern is about SYNTH and DUE etc. The RS keep me on track. As long as the end result is a better encyclopaedia. I’ve seen other editors who view minor IGF infringements of SYNTH and DUE as relatively harmless. YMMV.

Whoever said perfect is the enemy of good was right on the mark. Sandbh (talk) 08:47, 14 November 2020 (UTC)

I kind of agree with everyone here. The questions EdChem raises are matters that are extremely interesting and, I daresay, peeve-inducing once you get to know about them. But the fact of the matter is that for many of those cases the situation is that the literature is not really settled and a lot of different things are taught to students. For example, just try finding a universally accepted definition of what a block is in the literature. The fact that you'll never find one doesn't stop people from using the term, however. And if that's the situation, then we should tread somewhat carefully and mostly follow the textbooks. Are they sometimes just plain wrong and contradicting themselves with specialised RS being able to prove it? Yes – but we can't say that here because that would give the issue disproportionate significance and rather defeat the point of explaining things to laymen. It might make a good commentary for JChemEd, but their goals are not ours. Does the way they write sometimes lead to inevitable hands raised at the back of the classroom? Yes – but I guess these things ought to be discussed, if at all, in footnotes. Because having tried to explain periodicity myself in a way that I would find completely accurate to literature esoterica and still understandable, it seems to me that while it's not impossible, what you end up with if you try it becomes far far far to the extremes when it comes to what most textbooks do. Which is not, I think, what an encyclopaedia should be doing. Double sharp (talk) 11:16, 14 November 2020 (UTC)

Thanks for your comments, DGG, and I do agree that issues such as those raised here and elsewhere in the disputes / discussions have parallels in other areas and pose challenges that have been solved in other contexts on WP before.

Sandbh, I agree that no one WP:OWNs any WP article. I did comment at talk:Periodic table that you and Double sharp are the originators of more than 50% of the bytes in that article, according to the history and page statistics tools, but that tells us nothing about which parts. It's an FA with strong and weak parts and I don't know who contributed what, nor do I plan to investigate. I just hope we can all work together to improve the weaker parts.

As for your comment on RS, DUE, SYNTH, etc, I am reminded of the movie Reach for the Sky about WW2 fighter pilot Douglas Bader, where he and Harry Day exchange the maxim "rules are for the obedience of fools and the guidance of wise men." EdChem (talk) 20:59, 14 November 2020 (UTC)

• Keep in mind that any model is inevitably a simplification of reality that spares the reader overwhelming complexity. A central question for each article using the PT is "Who is the audience for this article?" What we often miss at Misplaced Pages is that for every editor there are hundreds of silent readers. We need to think about what will serve the reader best. Sometimes less complexity is more valuable. We can have daughter articles that go into greater depth. We can provide links to deeper treatments of the topic. Jehochman 05:15, 15 November 2020 (UTC)

• The overall readership for periodic table is millions per editor. The pattern of readership is cyclical – a peak in September and a steady decline to August. As this aligns with the academic year, this indicates that the readership is mainly schoolchildren and students rather than professional chemists. Understanding of the periodic table is commonly tested in chemistry exams – see this syllabus, for example. This should inform our coverage of the topic. Andrew🐉(talk) 11:24, 15 November 2020 (UTC)

Book chapter: Why don’t we really teach about the periodic table?

• Rayner-Canham G 2020, Advances in teaching inorganic chemistry, vol. 2: Laboratory enrichment and faculty community, Chapter 6, pp 69–77, ACS Symposium Series, vol. 1371 doi:10.1021/bk-2020-1371.ch006

From the intro:

"We lie to students. Such lies are often justified by saying that we are “simplifying” the topic. But in the process of simplifying, are we hiding the fascinating “messiness” of the periodic table? In this chapter, I focus on two facets of the periodic table. First, that richness, complexities, and controversies, should be highlighted in our teaching. Second, that terminology, sometimes used inconsistently or ambiguously, needs unambiguous definitions to give to our students. In the first half of this discourse, four of the many issues will be addressed; others are addressed elsewhere , particularly the very divisive issue of the lower members of Group 3 ."

Rayner-Canham covers the placement of H; do we overdo group trends; do we underdo other trends; and where do the early actinoids belong? --- Sandbh (talk) 12:09, 19 November 2020 (UTC)

Sandbh I read some of it, it was fascinating. I have to admit however that I do not think his proposed solution to group 3 (sometimes Lu, sometimes a gap before Lu, sometimes a gap) will actually stop the issue from being divisive, on the grounds that the current compromise IUPAC table with a gap below Y has been around for decades and if anything the debate actually got more divisive since it came out. So I will continue to wait for IUPAC. ^_^ Double sharp (talk) 12:20, 19 November 2020 (UTC)

Double sharp: R-C does not discuss a proposed solution to group 3 in this chapter, aside from citing my FoC article. Are you referring to something else? Sandbh (talk) 09:01, 20 November 2020 (UTC)

@Sandbh: Yes, your second link . Sorry for not making it clear that I meant that one when I said I'd read it. I haven't been able to read this first one yet. ^_^ Double sharp (talk) 10:03, 20 November 2020 (UTC)

Double sharp, in this context, is the German periodic table p-block worth revisiting? Their colour category treatment of Ge, Sb, Se, Po, and At is interesting. They are for sure not telling lies to children. Sandbh (talk) 11:31, 20 November 2020 (UTC)

@Sandbh:, my concern is not about telling lies to children, but rather about following the sources. I don't see any significant sources in English doing this sort of striping, so I do not feel we should do it. The source situation may be different between English and German: I have not checked because I'm not an active editor on German Misplaced Pages and therefore I'm not involved in checking if their content is supported by their literature situation.

Again, please understand that when I support ACS/LANL/2010 for WP, I mean that I think it best reflects the coloured-tables literature to do so, not that I think the colouring is all that great personally. There are a number of things like that on WP, where I currently support the WP table showing He in group 18 and La in group 3 on the grounds of most of the literature saying that, even though if you asked me personally my viewpoint is that both are serious mistakes. And that's why my current opinion for WP is that group 3 should not be revisited until we hear from IUPAC, not because I feel that would impact the scientific situation (it might if new arguments are presented, but that's not my main point) but because I feel that would impact the source situation that matters for WP. About the same thing goes for calling Ts a halogen and Og a noble gas from my perspective: I don't like it, if I was writing outside WP I'd probably call it a serious mistake as well, but I feel it's not WP's place to say it if most sources are doing it anyway. Similarly I think of the whole metalloid-line problem as an issue of an "original sin" of not clearly stating definitions, and would favour giving the children the physical-based definition "a metallic element has to have a Fermi surface in all stable or metastable allotropes at standard conditions" which mostly matches the chemistry (and then the metal-nonmetal dividing line runs between Be and B, between Al and Si, between Ga and Ge, between Sb and Te, probably between At and Rn, and in the 7th period everything is a metal but Cn and Og probably). I feel it makes sense because "metal" seems to have started as a physical term according to Origin and use of the term metalloid which you wrote, most people are more familiar with the physical idea of a metal than the chemical one (K is by far a stronger metal chemically than Au, yet the latter is for everyone the king of metals, and the former was found so weird it was sometimes called a "metalloid"!) and the chemistry just mostly correlates because the same thing of easily mobile electrons is behind both (though the correlation is not perfect). And I'd do that if I were writing some sort of text on periodicity. But again, to a first approximation people don't do that, there's no clear consensus definition in the literature, and so I feel my personal preference here means nothing for WP no matter how persuasive I find my own argument. So, I've listed three issues, containing a mixture of agreement and disagreement with your view, where I have a personal view and don't advocate it for WP because I don't feel the current source situation supports it. The fact that you probably have a different preference for some of this is for me further proof that trying to analyse the sources ourselves for some sort of scientific correctness is not the way to go; because of differing philosophies, some arguments of mine won't be found convincing by you, the same goes the other way, and nothing can get done that way for an encyclopaedia. There we have to present the views in the literature given their weight there; each following our own philosophy is something that only works if we write a separate monograph each, which isn't what we're doing here. That's why I feel that this sort of thing should be left to monographs and papers, leaving encyclopaedias to just describe the majority situation first even if we both agree that it sucks and there's some minority situation around that is better; I feel that's consistent with WP:DUE anyway. In other words: if most of the sources are lying to children, then I feel it is our duty to be following that even if mentally I cannot help but protest. You don't have to agree, but I'd ask that you understand where I'm coming from. And please don't try to convince me that the lying to children is bad; you'd first of all be trying to convince me of something I'm already convinced by, and second of all not be addressing my actual argument and therefore probably not manage to change my mind. ^_^ Double sharp (talk) 11:40, 20 November 2020 (UTC)

P.S. A table drawn according to how I would prefer if I was writing outside WP can be found at User:Double sharp/Template Periodic table. I hope its drastic difference from everything I am proposing for WP makes my view above clear. I wish to also make clear that while I personally, speaking with the scientist/educator's hat on, think this is the best way to go; putting my encyclopaedist's hat on, I have precisely zero intention of proposing that as a general WP table; and if it appears somewhere, it will only be in something off-WP which I, or perhaps I and somebody else, write in the future. And even then I would not propose it for the WP general table, despite my personal belief in its superiority, unless most people adopt it – which, even if it does happen, will probably take a long while. I hope that clarifies what I've said above. Though maybe it would be a good and fun exercise for everyone here to draw the PT to personal inclinations once, just to see how great the differences are and why I feel that the best way forward is to unquestioningly adopt something common in the literature rather than seeking to create something better by analysing it ourselves. ^_^ Double sharp (talk) 12:15, 20 November 2020 (UTC)

Double sharp, A picture is just a form of words. No doubt you'd've heard the expression, "A picture is worth a thousand words?" The p-block in the German PT picture is probably the most representative of the entire English RS literature. How do you see this now?

On lies to children I suggest this is a question of the quality of RS. Very crudely, there are: 3. less RS; 2. RS; and 1. "top-tier" RS, with reputation. Lies to children (ltc) RS fall between tiers 3 and 2. There's no need to necessarily rely on ltc-RS, since plenty of RS and top-tier RS are available. Indeed. a theme of distinguishing RS based on e.g. publisher reputation, author reputation, age of publication, etc pervades WP:RS.

Sandbh (talk) 22:01, 20 November 2020 (UTC)

Well, Sandbh, I understand that there may be a difficulty when it comes to the difference between what exactly is the best that can be shown from sources (e.g. for the superheavies), and what individual sources say, and I confess I am not sure what I believe in here anymore. I am also unsure if I really feel strongly for any of the schemes in particular in the first place as opposed to others at the moment. For the time being, I'd say you can go ahead with your RFC, because I feel like I have to think about this and what I feel should be there in WP. To edit effectively, I think I do need to to some extent believe that what I'm doing is the best path for creating an encyclopaedia, and frankly I'm not sure exactly what is at the moment. So I'm also not sure if I should set something before the community to pass judgement on if I only half-heartedly believe in it myself. I will have to think about it. Maybe go on semi-wikibreak mode to do so, hopefully with no hard feelings attached since we resolved all mutual issues.

(That being said: if your RFC is planned to be as at User:Sandbh/Nonmetal rfc, then I suggest you consider removing the "Aim" section. Reason being that it makes the thing look not quite neutral by arguing against "reactive nonmetals". Feel free to take or ignore this advice as you see fit, of course. ^_^) Double sharp (talk) 22:14, 20 November 2020 (UTC)

Double sharp, righto. No hard feelings whatsoever. The discussion has been engrossing. I'll have a look at the aim section. Maybe rewrite it into a context section. Not sure yet. Sandbh (talk) 00:34, 21 November 2020 (UTC)

​​​​​​​​​​​​​​​​​​​

Thanks for emailing me the article, Sandbh. Since a significant part of this seems to be about what is over- or under-emphasised in the literature, I guess I will not comment about it much for WP coverage of the periodic table in that overview article, but it is interesting and could maybe be used to cite a couple of warning examples that the typical trends are not as straight as the ones typically chosen (maybe good for a footnote there). I also like his classification of Sb as a metal rather than a metalloid, but only personally, since most people still include it. ^_^ Double sharp (talk) 23:31, 22 November 2020 (UTC)

Major: Group 3

IUPAC Group 3 project update

​​ From Eric Scerri: "Before delivering a report I will be publishing an article in Chemistry International which discusses our preliminary conclusions and will be asking for comments and feedback. I'm still writing the article." "By all means share this update." --- Sandbh (talk) 09:07, 8 November 2020 (UTC)

@Sandbh: Thank you, this is useful information to keep in the back of our heads. We can start to consider it for content discussions when his article appears – which I hope happens soon. Double sharp (talk) 19:11, 9 November 2020 (UTC)

@Sandbh: If I may ask: was an approximate timeframe for the appearance of this article given? (Totally fine if it wasn't, just curious.) Double sharp (talk) 22:03, 30 November 2020 (UTC)

@Sandbh: No. I’ll ask Eric how he’s going. I could understand his reluctance to be pinned down to a specific time, however. Sandbh (talk) 00:34, 1 December 2020 (UTC)

@Sandbh: I'd definitely understand it too; I'm just curious. Double sharp (talk) 01:31, 1 December 2020 (UTC)

Group 3 discussion: formal conclusion (Dec 2020)

(Reacting to Double sharp's ping). We write for a generalist readership. Therefore, our primary goal is to describe the current consensus, as set out (for example) in modern textbooks. The La/Lu debate is only of interest to specialists, and is unresolved. Its proper place is in a section, or perhaps an article, and nowhere else; so as not to confuse readers who just about know what the PT is. Narky Blert (talk) 10:42, 20 December 2020 (UTC)

@Narky Blert: We have a section Periodic table#Group 3 and a main article Group 3 element#Composition of group 3, the latter of which discusses the situation in more detail; is that what you're thinking of?

Thanks for your opinion. IUPAC was planning to resolve it and I got some information from Sandbh here (who's in touch with the IUPAC project chair) that something is going to be published about it by them soon. That's why I decided that the issue might as well be dropped until we hear from them, since whatever happens we have to look at it then anyway. But, more recent events made me worried that some others who had opinions here might not be happy with this "temporary drop". So my asking was intended to just gauge if that was so or not. Double sharp (talk) 10:50, 20 December 2020 (UTC)

@Double sharp: I haven't read them - but yes, exactly that

I haven't followed this whole debate; but would suggest closure with some headnote or another until IUPAC decides something. There's no point in dragging this discussion out still further; and what IUPAC says may make parts of it moot. Narky Blert (talk) 11:04, 20 December 2020 (UTC)

@Narky Blert: I agree totally with you. I also like your idea of putting some headnote up to say "look, let's wait for IUPAC", because this issue seems to be amazingly effective at driving people who care about it to passionate argumentation (myself included, I've been trying to cut it out). Well, let's just see if everyone else is happy with such a closure, since I never actually asked when I requested a withdrawal there. ^_^ Double sharp (talk) 11:13, 20 December 2020 (UTC)

Yes, there should be some sort of note in the archive to indicate our consensus to let things stand at least until we can look at the IUPAC group's report and the reaction to it. But I am not sure whether it is kosher to add a headnote to an already archived thread. I think the best thing would be to start a new "==" thread, invite all participants in the old discussion to ascent to our implicit closure, and then archive it at the top of archive 52. YBG (talk) 17:06, 20 December 2020 (UTC)

Yes, as confirming the status quo. Cannot be a new discussion. -DePiep (talk) 19:49, 20 December 2020 (UTC)

I agree with Double sharp. I think we should just wait for the IUPAC. -UbnBr₂ (Talk | Contributions | Actions) 05:13, 21 December 2020 (UTC)

Proposal

As a more-or-less neutral observer, I propose that this and any similar live thread be closed with a headnote along the lines of:

This discussion is closed without WP:CONSENSUS, and should not be reopened until IUPAC makes a firm recommendation about the La/Lu question.

I don't care about the exact wording. Anything to the same effect would do. This debate has in the past spilled over into at least two dramaboards. Let's draw a line under it.

Support

• Support, as proposer. Narky Blert (talk) 20:25, 20 December 2020 (UTC)
• Support. Good enough for me. It doesn't make sense to fight now for something this controversial, when a "no consensus" result is the most likely outcome, when it seems likely that the IUPAC taskforce will decisively put the matter to rest rather shortly. Double sharp (talk) 03:26, 21 December 2020 (UTC)
• Support. Sandbh (talk) 03:46, 21 December 2020 (UTC)
• Support. YBG (talk) 10:11, 21 December 2020 (UTC)
• Support. -DePiep (talk) 10:23, 21 December 2020 (UTC)

Oppose

Comments

(signed by proposer) Narky Blert (talk) 20:25, 20 December 2020 (UTC)

IUPAC textbook survey

Released one year ago, the results are here.

The ratios for the three periodic table options are as follows:

                Sc-Y-La  Sc-Y-Lu  Sc-Y-*-**   Z
======================================================
1970s to 2010s  4.33     1.0      1.1        193
2000s to 2010s  2.7      1.0      1.0         94
2010s           2.66     1.0      1.83        33

A single decade is probably too small a sample to base any conclusion on.

Anyway that is old news.

From a block perspective, it's interesting to note Lu appears in the f-block about 85% of the time (1970s to 2010s). That is because it shows in the f-block in the traditional Sc-Y-La-Ac form, and in the *-** form. Conversely, La is in the f-block about 33.3% of the time. That is because it shows in the f-block in the *-** form and in the Sc-Y-Lu-Lr form. --- Sandbh (talk) 05:37, 23 December 2020 (UTC)

FYI: Campero and Ponce (2020)

​​Disclaimer: No version of Group 3 is better than another; each version has it uses, depending on the properties of interest.

I flagged C&P's article in the interesting publications section.

There were two items in their article that I noted:

"We conclude then that the natural sequence of atomic number Z in the large periodic table, as Scerri has explained,(52c) implies that La is in group G3."

"Also, the electronegativity values determined in this work are correlated to similar chemical behavior of the ions through Knight’s, isodiagonal, vertical, and horizontal periodicity."

In my recent article, "The location and composition of Group 3 of the periodic table" I argued for the placement of La in group 3, including on the grounds of isodiagonal and horizontal relationships.

I've posted elsewhere about the vertical trends going down (a) B-Al-Ga-In-Tl, (b) B-Al-Sc-Y-La, and (c) B-Al-Sc-Y-Lu. Based on 23 physical and chemical properties, it turns out the trendlines are the smoothest for (b) B-Al-Sc-Y-La. The next best option is either (a) or (c).

On , there is a knight's move relationship between Ca and La:

• The ionic radius of Ca is 114 pm; that of La is 117 pm (cf. Lu 100).
• The similarity in sizes means La will compete with Ca in the human body, and usually win on account of having a higher valence for roughly the same hydrated radius.
• The basicity of La₂O₃ is almost on par with CaO₂ whereas Lu₂O₃ is the least basic of the Ln oxides.
• Freshly prepared La₂O₃ added to water reacts with such vigour that it can be quenched like burnt lime (CaO) whereas Lu₂O₃ is insoluble in water.
• The electronegativity of Ca is 1.0; that of La is 1.1 (cf. Lu 1.27).

--- Sandbh (talk) 00:26, 24 December 2020 (UTC)

Major: Colouring scheme

On creating a compromise colouring

There's something Andrew D. said at the ArbCom case page that I think makes a lot of sense, too: "perfect is the enemy of good".

Now, I've stated a way above in which one can think of the current colour scheme as policy-compliant. It reflects the situation in the sources adequately in the sense that most sources show some kind of colouring, and that it is usually not too far from this one. It seems to me that as long as this variation exists in the sources one cannot do better, but one is not doing too bad.

Now, I'll also note that we've had almost this colour scheme since literally 2002. And it served us pretty well up to 2013: the categories there would correspond basically to

alkali metal | alkaline earth metal | lanthanide | actinide | transition metal | post-transition metal | metalloid | other nonmetal | halogen | noble gas

So let's ask: what's changed since then? Not very much, honestly.

• Group 3. Yeah, let's not touch that right now. It is one of the things that changed, let's note that and move on. On moratorium until we hear about the IUPAC project progress, if you will.
• Group 12 moved from transition metals to post-transition metals. Well...the 2017 RFC was basically a snow close, so no need to revisit that: let's leave well enough alone.
• Elements beyond Hs. Nothing to say here. They were coloured in at first, but then it was pointed out that there had been no chemical studies, so they were uncoloured. Cn and Fl got chemical studies that at first looked conclusive but then turned out not to be so, so they got coloured in and then uncoloured back again. That seems to be about how it should work.
• Polonium. Well, no one seems to have disputed it since I brought it up back in 2012. Most people who focus on it seem to agree it is a metal, so there it went.

Then there is the one element that caused all the trouble.

• Astatine. Problem is, this is where halogens leave the nonmetals. Like I said in 2012, I would agree with colouring Po as a post-transition metal and At as a metalloid, but what do we do with the halogens category? And thus began the quest to get a better category name than "other nonmetals", deal with astatine effectively, and try to have something similar.

I think that, looking at Sandbh's RFC with so many plausible options, as a contrast to that linked 2017 RFC that dealt with group 12, this may be a case of "perfect being the enemy of good". It seems to me that the literature generally agrees that if a PT is coloured, halogens should be coloured, but it doesn't seem to have any agreement on what to call the other nonmetals. That would certainly explain why Sandbh was able to marshal some level of support for multiple options, but I don't think there is total dominance for any one of them.

So I'd like to turn our eyes back to 2002 and think: suppose at that point, just having built a good enough PT, we had just heard that At turned out from early chemical studies and modern theoretical corroboration to be a metal. What would we do, now that it's not clear in the literature even whether halogens extend past iodine?

I think there's a relatively simple solution. Just replace "halogen" with "halogen nonmetal" as has been suggested above. Then that actually manages to please all sides. If you agree with the part of the literature that thinks a halogen has to be a nonmetal: then it's just a pleonasm, but not wrong. If you agree with the part of the literature that thinks a halogen is anything in group 17: then it's actually necessary to be right. So both parties will agree it is right, and the general idea that F-Cl-Br-I are distinct enough to deserve being a category is there. For astatine, we invoke WP:CONTEXTMATTERS and say: OK, well, it's an element most people will never encounter, so let's look at the few sources that cared about it. In fact, that's going to create exactly Sandbh's 3-category solution, so I guess he may be happy. Although he may be less happy about what I propose to call the new-old category. ^_^

Sure, it's not "perfect". But it seems to me that many of us are actually looking favourably at restoring halogens just because that's what most sources have. I worry that in Sandbh's proposed RFC, because there are so many possible options for the remaining nonmetals, none will get a clear consensus. I also worry that the fact that there are so many possible options rather strongly suggests that the literature isn't unified and that it will scotch any possible consensus (in fact, I still can't bring myself to really strongly support any option that isn't "other nonmetals"). So I'd ask: why let the perfect get in the way of the good? Let's just do a small improvement that keeps things still OK, and it'll be better than nothing. If we try to go for the gold from the beginning, we'll get nothing.

Therefore I would like to ask: could everyone accept this compromise? It doesn't matter if you think it's not the best we could do, just whether you think it could be better than what we have now. I'd rather we have one modest improvement than try for a huge and great improvement that ultimately is too radical to pass. This way, we at least get halogens back on the table like they usually are, even if we don't have the name for the other category.

I edited this from a 2011 revision; obviously it'd be updated to current styles. We can change the colour scheme later, but that's a different topic. Thank you User:DePiep and User:R8R for your work there. ^_^

In that spirit, I would like to suggest that if this is acceptable, we "lock in" the categorisation and the layout for a while. No sense in arguing about it again when there's an entire article to improve and any further improvements run the risk of getting no consensus. That's inspired by what User:Jehochman said at User:EdChem's talk page: Can we have this discussion once, come to a decision, and then not have to revisit it too often? Obviously only for the first 108 elements (we need to be able to reflect scientific advances if elements 109 and up get chemical studies of course).

If need be, the name of the category "other nonmetals" can be revisited later, and that particular thing need not be locked in. But remember, the change from group 12 to leave the transition metals was tried twice. It failed the first time because then the category name was "poor metals" and its situation in the literature was more iffy than "post-transition metals". The second time, the new category name was "post-transition metals", and it resulted in a WP:SNOW success. Basically: if generally there did seem to be agreement between Sandbh and me at least in October IIRC that splitting out the halogens was justified (and that maybe counts for something because we generally post here more stuff than anyone else about all this categorisation stuff), I would prefer that that less controversial step not be jeopardised by mixing it with something more controversial.

Do we need disclaimers? I guess on the main periodic table article we do, and for the infoboxes we probably do need them just as we do for something like {{infobox zinc}} or {{infobox polonium}}. But the literature situation is inconsistent, so I think more than that is the perfect being the enemy of the good again. So I'd support this compromise in the form of just changing all the PT images and templates to this, standard style everywhere, contract to 18-column consistently with small asterisks, etc. etc. etc.

No, this is not my personal favourite ideal colouring across the whole spectrum of colourings, but I think that probably speaks well for it being a good compromise: no one is completely satisfied, but people should feel it's better than nothing. And within the constraint of being close to the centre of what sources really give, I cannot think of any way to do much better, which makes me think that trying to better it is trying too hard.

Why do I think this might be a plausible compromise? Because such a split of the nonmetals is something Sandbh has proposed, so it is presumably not going to cause an argument; because the name "other nonmetals" is one of the options Sandbh mentions in his RFC, so it is presumably not completely unacceptable to him; because it is my preferred option from his RFC and seems the most obviously non-controversial one; because IIRC most supported some sort of restoration of halogens; and because it is based on something that lasted for years uncontroversially before new information about astatine threw a spanner in the works. But most of all; because I think that these reasons give it a chance to stop the fight completely, as something like this was around for over a decade and no fighting resulted, so I feel like tweaking it the minimal amount possible now to correctly reflect new information about astatine is a plausible way to go. But we will see; the more important thing is to have some compromise stop the fight, rather than for it to be this one specifically. Anything that can solve this issue swiftly, take our minds completely off it, and let us work together on actually improving the parts of the periodic table article without controversy would be a good thing in my eyes even if it doesn't end up being this. Double sharp (talk) 20:23, 13 November 2020 (UTC)

Could we perhaps agree to freeze this issue for a while too? It seems to me that so far, we have a couple of views on the current colouring, but it seems to have worked fine since 2018, and I don't think anyone's current idea is that radically far from it. So, I'd like to suggest we stop talking about it at least temporarily and get to work on the periodic table article itself as EdChem is suggesting to us. We may perhaps return to it when the memory of the difficult previous situation is not so close. Double sharp (talk) 22:30, 14 November 2020 (UTC)

Yes please. Our today's discussion flow is not effective. (Is there something about 2020?). -DePiep (talk) 23:33, 14 November 2020 (UTC)

Although if Sandbh wishes to go ahead with his proposed nonmetal-splitting RFC, I think it may not be too much of a problem provided we just !vote, explain ourselves, and don't argue with each other. Double sharp (talk) 00:39, 15 November 2020 (UTC)

Category colors

Here's a table comparing contrast ratios between the colors we've used, plus black:

Legend

•  C : Black (#000000)
•  #0645AD : wikilink blue font (#0645AD)
•  Fr : Alkali metals, dark version (#ff6666)
•  Li : Alkali metals, light version (#ff9d9d)
•  Be : Alkaline earth metals (#ffdead)
•  Ln : Lanthanides (#ffbfff)
•  An : Actinides (#ff99cc)
•  M : Transition metals (#ffc0c0)
•  Al : Other metals (#cccccc)
•  Si : Metalloids (#cccc99)
•  P : Other nonmetals (#a0ffa0)
•  X : Halogens (#ffff99)
•  Ng : Noble gases (#c0ffff)
•  Ubu : Superactinides, as used in Extended periodic table (#d1ddff)
•  ? : Unknown properties (#e8e8e8)

Symbols are readable against all the colors used, though the Fr-alkalis are markedly less so, but none of the colors are AA-compliant with each other. In particular, only 3 pairs exceed a contrast ratio of 2.000 (Fr-alkalis against nonmetals, halogens, or noble gases). The Li-alkali do not contrast as well as the Fr-alkalis against the other colors, especially for visually impaired readers, and are especially hard to distinguish from the actinides. Aditionally, the transitions contrast poorly against the lanthanides and the post-transitions; the post-transitions are very hard to distinguish from the metalloids; and the most similar pair was nonmetals vs. unknown. The most important colors to distinguish are the ones that appear adjacent on the periodic table, such as the transitions and the post-transitions; these pairs are bolded in the table above. The nonmetals could be darkened a bit, but we'll have to come up with a completely new color scheme for the metals. –LaundryPizza03 (dc̄) 12:54, 5 December 2020 (UTC)

@LaundryPizza03: I agree. I understand R8R and DePiep have made some progress on this front (see e.g. User:R8R/PT search for colors). Double sharp (talk) 16:06, 5 December 2020 (UTC)

First quick remarks: the {{Color contrast ratio}} applies to fontcolor vs background color, not neighbouring bg colors. So the red bg's in the table are not that meaningfull. That said, there is an other reason to strive for between-categories-color-distinction: the Reader being able to recognise category bg color from PT-cell-to-legend ('which category is this cell?' and v.v. ('which cells are metalloids?'). With this, as a third distinction, backgrounds should not look alike (confusion as noted: 'is this AM or PTM?'). -DePiep (talk) 22:06, 5 December 2020 (UTC)

I have added row #2, "wl": the wikilink blue font color (as in Main page). Its contrast with background is meaningfull. Because: wl-blue is a fontcolor, replacing black fontcolor. -DePiep (talk) 22:20, 5 December 2020 (UTC)

@LaundryPizza03: Why checking black as a bg color, and white fontcolor? All your contrast checks are black-font on color-bg, right? (Is why I added wl-blue this way). -DePiep (talk) 22:31, 5 December 2020 (UTC)

LaundryPizza03, thanks for your contribution on colours that is looking from the perspective of accessibility / readability. I agree that this is an important consideration and is one that I feel needs advice from those with expertise in the area. I know that DePiep has been thinking about these issues as well.

I can only speak for myself (though I suspect I am not alone) in agreeing with the importance of making a choice of colouring scheme that supports visually-impaired readers in principle but having little knowledge of how this is evaluated in practice. Looking at the above table, for example, I get that larger contrast ratios are desirable for readability and I infer that a goal of greater than 2 is an appropriate guideline, but I don't know how DePiep's comments fit. Is this an area where those lacking expertise (like me) should just stand back, or is there some material that I can read to help make a useful contribution? Are there contributions that would be useful from a chemical perspective? For example, is it better for metal categories to be based in a related group of colours (say, shades from purple / pink / red / orange) and then different groups for non-metals (blue / green) and metalloids (yellow)? Or, is it better for the metal colours to be unconnected (say, green, yellow, red, brown) and then blue metalloids and pink and orange non-metals (just picking different colours, many alternatives possible)? Alternatively, is there a WikiProject or group that deals with accessibility / readability and colours that could offer help? EdChem (talk) 23:56, 5 December 2020 (UTC)

The table is not exactly to the point.. W3C contrast checks are about font-vs-background colors, not neighbouring bg colors. (There is no AEM-bg vs LN-bg check needed anywhere). The table should be removed. Except for row 1 and my wl-blue addition. -DePiep (talk) 00:01, 6 December 2020 (UTC)

@EdChem: in other words: the table does not address WP:ACCESSABILITY. I also wrote in my 1st reply above. -DePiep (talk) 00:08, 6 December 2020 (UTC)

Thanks for the explanation. H:Colorblind suggests using only 5 colours (including black and white) as backgrounds... would that approach work, picking a background (blue, say) for the non-metals and use stripes (horizontal, vertical, and none), then another color for metalloids, then two colours for metals with stripes (including diagonal if needed)? I imagine it would be messier than just with differing colours, but potentially more accessible – or even link to such a PT as an alternative for the colour one for those struggling with readability? EdChem (talk) 00:41, 6 December 2020 (UTC)

re, short: we must provide ten (10) category colors. Difficult, if even possible (too manymuch restrictions). Only good news: it is a PT-pattern, so a l-to-r flow. Also we can gain by dropping font-colors for phase (red H, green Hg).

And: this table is not the optimal guide, as noted.. -DePiep (talk) 01:21, 6 December 2020 (UTC)

As for coloring the categories, EdChem, we are sailing between Scylla and Charybdis. Satisfying all contrast and CB (colorblind) needs, will give us bleak (not even pastel) colors. That are failing the "recognise a category" check.

Some self-redactions into more helpful wording. -DePiep (talk) 17:43, 6 December 2020 (UTC)

Major: Organising the nonmetals

Compromise 1: other nonmetals + halogen nonmetals

Nonmetal categories

(The anchor for the old thread name, so it still works. Double sharp (talk) 13:21, 18 November 2020 (UTC)) Similarly, there may be some grounds for just deciding on other nonmetals and halogen nonmetals and calling it a day there.

1. We had other nonmetals from 2002 to 2013, and no one complained too much.
2. That's what Britannica does at least for other nonmetals. LANL and ACS are naughty and say "nonmetals" without "other", but an "other" makes that simple and accurate (since halogens and noble gases are nonmetals).
3. Britannica, LANL, ACS, and RSC all call out "halogens". I think that strongly suggests that including halogens as a category is the common thing to do and that we should do it somehow. The only difficulty is astatine because it is ambiguous: does being a metal (agreed by RS focusing on it) exclude it from being a halogen? Some think yes, some think no, so "halogen nonmetals" solves the issue. If you think yes, it's pleonastic but not wrong, and if you think no, it becomes actually needed for it to be right. (Besides, when most people think about halogens, eight-hour-half-life astatine is not really on the radar, is it? So if only F-Cl-Br-I are mentioned, perhaps it isn't so bad.)
4. Since it appears that most sources for beginners aren't bothering with a special name for H-C-N-O-P-S-Se, perhaps we should not try either. It appears that people generally feel it's too much for beginners. Remember, we're not looking for a perfect category scheme that somehow avoids "other" categories (I have doubts you could make one in the first place although I'm ready to be convinced), but something that is more or less near the centre of what beginners' RS do considering our likely audience, and manages to stay at least not obviously self-contradictory while doing so.

In the past, judging from Misplaced Pages talk:WikiProject Elements/Archive 15#Polonium – metalloid or post-transition metal?, the way we solved these issues was just to post at the talk page, see how the response was, and if no strong objections were raised just carry out and do it without needing to RFC anything. That seems in keeping with how we treat the entire rest of the article: anyone can improve it, there's no need for anyone's changes to go through an RFC if it's not generally felt that they'd be controversial. So I try doing it this way to avoid (1) a bureaucratic logjam and (2) extending the issue when there's the whole rest of the article to work on. If people feel generally OK with it, that's good; if not, then that's also good.

@Sandbh, R8R, YBG, DePiep, EdChem, and Andrew Davidson: Double sharp (talk) 13:42, 15 November 2020 (UTC)

@Double sharp: I am rather opposed to the idea. I'm not particularly clinging onto having "reactive nonmetals," and I could live with "other nonmetals." However, I am worried about "halogen nonmetals." As far as I know, it is common to consider halogens either as group 17: fluorine, chlorine, bromine, and iodine (nobody cares about astatine one way or another), or as group 17: fluorine, chlorine, bromine, iodine, and astatine (and maybe even tennessine). Specifically excluding astatine from the set of halogens, when the element is explicitly mentioned, is not common; at least I don't expect those readers before whom we'll be wearing our educator's hats to be familiar with it. Talking about "halogens" is common; talking about "halogen nonmetals" (as opposed to merely "halogens," because we would otherwise use the shorter name) isn't.

We could do it, and it wouldn't be the end of the world: astatine is a relatively marginal element, and we can talk ourselves into thinking it's okay. But I'd prefer we didn't create such ambiguous categorizations when there is an option not to.--R8R (talk) 15:07, 15 November 2020 (UTC)

@R8R: Thanks for your comment. Well, I also don't particularly cling to "reactive nonmetals" (I want to ask: reactive relative to what?), and could live with "other nonmetals", so I see we are halfway in agreement. ^_^

As for the halogens problem: I think the issue is just what you say, that nobody cares about astatine. Indeed the only reason we are in such a bind is that the latest research suggests that At is a metal. If not for that, I think having At under "halogens" alone would satisfy everybody. So I'd ask: even if specifically excluding At is not always a thing, is it not the case that it's practically excluded just because the only halogens anyone gets to play with to a first approximation are the first four? For example Greenwood and Earnshaw spends its halogens chapter almost totally on F, Cl, Br, and I and only covers At in a brief section at the end. And when halogens were covered to me in school, while there were some of the usual questions about predicting properties of At, it really wasn't covered in any detail at all unlike the first four. Finally: it seems common to mark out the halogens in some form, and it's only At not being a nonmetal that makes this difficult. Then I would like to ask: if we agree that astatine is peripheral, then is it really the right order to have problems with At block the reintroduction of halogens? I feel that the commonness of halogens as a category for beginners means that it should come back in some form: if I understand correctly, Sandbh may feel similarly. But if we are going to have halogens back, then there seem to be only two options for At: either it goes under halogens (undermining detailed reports that it is a metal), or it goes under metals and generally doesn't make things too bad because At often gets informally excluded from halogen discussion for beginners anyway. So that's my argument: both treatments have problems, but maybe this has fewer problems.

I think that suggests that things may not be so bad, and I guess that when you say that we could do it and it wouldn't be the end of the world it means you're willing to talk. But if you're not convinced, it's totally fine. I'd also like to hear from the rest. Double sharp (talk) 15:30, 15 November 2020 (UTC)

I see. Your logic seemed somewhat strange to me at first, but I think I got it. It appears to me there is a difference in what we seek to prioritize: you seem the want to prioritize categorization of "big" elements first, whereas I'm looking for a self-consistent scheme first. That's why I am trying to get rid of inconsistencies first, and you seek to clarify which elements are big first when there is a potential inconsistency. Both of these approaches are valid, I'm just glad to have figured it out. (IIRC you displayed similar logic when it came to coloring of copernicium.) With this clarification, I believe my position is clear, and I have nothing of substance to add. I'm interested to hear what other editors think on this topic.--R8R (talk) 16:04, 15 November 2020 (UTC)

@R8R: Yes, your position is clear and you've understood my approach here. (That's generally my approach just for categorisation because I think of that as just a human-made give and take; for the record, I feel differently about layout questions, but you already know that and those issues are frozen.) We can wait and hear from the others indeed. One question, though, is raised by what you've written. You mention getting rid of inconsistencies: do you think that our current scheme with reactive nonmetals, or my proposed one with other and halogen nonmetals, has an inconsistency? And if so, could you briefly explain what inconsistency/ies you see? (Because I'm personally not seeing them.) Double sharp (talk) 16:33, 15 November 2020 (UTC)

As you know in the past, I have argued vociferously though not effectively against the trifurcation of the nonmetals, but I have now come to believe that the time and energy I spent on that was a total waste. I have a limited amount of time to spend on WP and on ELEM, and I should spend it on things most helpful to readers, and I rather think that article improvement trumps element categorization by orders of magnitude.

This brings to mind film colorization with Peter Jackson and They Shall Not Grow Old on the one hand – and Ted Turner, Orson Wells, and Citizen Kane on the other. Which should be our model as we take our digital paintbrushes to Dmitri's black-and-white masterpiece? Let us not go the way of acrimony, but rather devote our energy on things of greatest benefit to our readers. YBG (talk) 16:48, 15 November 2020 (UTC)

@YBG: So...is that a "go ahead", "don't do it", "please please please stop talking about this so we can work on the article", or something else? (Just as not-too-serious characterisations in an attempt to lighten the mood.)

I only suggested this because both Sandbh and I seemed to be in favour of trifurcating the nonmetals. And since this is one of the issues where we fought the most, I thought that since we both agreed on it we could just do it, really call it a day without people having "and then we're going to change this!!" at the back of our minds, and really focus on the article itself. That was really my worry, that if we stop talking about this one people will still have it in their back of their heads and won't focus effectively on the article like we should, so I wanted to take a leaf from that 2012 thread on polonium and try to resolve it very quickly in a few days and put a result into action and then agree not to talk about it again without any feelings of "the discussion wasn't finished!". If you think we shouldn't be talking about it at all now, that's perfectly fine too; it's just that I don't understand what you're saying. T_T Double sharp (talk) 16:56, 15 November 2020 (UTC)

I am saying that I have no objection to change or non-change. My preference is to spend the least amount of time at the crossroads and start forthwith down one path or another, provided, of course, that we don't have to retrace our steps in the near future. YBG (talk) 17:03, 15 November 2020 (UTC)

@YBG: Right, thanks for the clarification. I also want to spend the least amount of time on this as possible, which is why I'm trying to mimic the old Po thread in 2012 that managed to amazingly enough wrap itself up in exactly one week (26 June – 2 July 2012). ^_^ Double sharp (talk) 17:07, 15 November 2020 (UTC)

YBG and Double sharp, remember that the outcome doesn't have to be dichotomous. Yes, we can only represent one way on the PT shown, but that does not preclude a section later showing a different perspective linking to a more detailed discussion in a different article. There is no reason that the halogen article has to start with halogen = group 17 and not address / recognise that more limited sub-sets are used by the terms in certain situations and further down in the article. A general rule need not apply at all times. EdChem (talk) 00:41, 16 November 2020 (UTC)

@EdChem: Yup, it doesn't have to be. I was speaking only for the "general" situation we're dealing with at the periodic table article and for stuff like infobox templates. Specialised articles need not by any means do the same thing of course, they will do and discuss whatever is more relevant for their contexts. ^_^ Double sharp (talk) 00:45, 16 November 2020 (UTC)

@EdChem: Just asking you since you are a chemist: exactly how is astatine's status as a halogen treated in your experience by your colleagues? Is there any difference between views of organic and inorganic chemists? Does anybody really care much (outside I guess the few who work with it, since they probably care)? Have you ever actually gotten to work with astatine or know someone who has? Is it well-known that recent results (both theoretical and empirical) are leaning towards astatine behaving more like a metal than not? And if it is, does anyone think that that has anything to do with whether At is a halogen or not? Thanks for your answers if you decide to give them (since your statement about the group 3 thing at ArbCom was useful). I understand you would prefer to let us decide this among ourselves, but in this case I think any information you would give on this would be useful to let us do that. ^_^ Double sharp (talk) 16:43, 15 November 2020 (UTC)

Double sharp, I wasn't particularly aware that astatine was coming to be seen more as a metal than a metalloid. I've never worked with it and don't know anyone who has. My guess is that most would view At as a halogen just from being in group 17, but that more as a default that a considered analysis of the term "halogen"... though organic chemists do often limit the term halogen to those typically used. I do know some with strong views on La / Lu, though, but that tends to be inorganic / organometallic chemists who work with lanthanoids. EdChem (talk) 00:35, 16 November 2020 (UTC)

@EdChem: Ah, so it is my suspicion then: nobody cares much about any of this except those who work with the elements involved. ^_^ Thank you, this helps a lot! Double sharp (talk) 00:38, 16 November 2020 (UTC)

Compromise idea 2: other nonmetals and halogens with At as a halogen?

Here's another possibility: why don't we simply colour astatine as a halogen, with no quibbles about whether or not it is a metal? That agrees with IUPAC, first of all. It also happens to agree with Britannica, LANL, ACS, and RSC.

This is, indeed a more group-like definition of what a halogen is. But it's consistent with what pretty much everybody else is doing, including with tennessine. That makes me think that actually nobody seems to be thinking that the metallic properties in any way block At from also being a halogen, particularly when it does sometimes act like a halogen.

As R8R correctly notes, it is not common to specifically exclude astatine from the halogens. Indeed, as EdChem seems to be suggesting, the majority view among chemists is that astatine is a halogen almost as a default, and that its informal exclusion is not so much due to people having analysed At chemistry and what "halogen" means but more due to people just not caring about it. Perhaps we should follow suit if that is how the literature seems to be, rather than trying to be better than them for an element with a still relatively undeveloped chemistry.

There is a con: doing this means we have to stop with the explicit "umbrella" categories of metals and metalloids and nonmetals. That's because "halogen" no longer implies "nonmetal" (not that it seems to have universally done so in the literature in the first place; maybe for some, by no means for all). So, instead of the legend looking like this:

it would always look like this:

I think though that this is not a big deal, because the majority of categories still retain "metal" / "metalloid" / "nonmetal" in the name, and because this kind of compact navbox without the umbrella categories is already present in {{Periodic table (navbox)}} right now. Besides, this reflects actual usage: there is a certain ambivalence about whether "halogen" and "noble gas" imply the whole group or go along with the baggage of nonmetallicity as well, and not a consensus in the literature, so it seems correct to be ambiguous about it when everybody does that. It should be noted after all that Britannica, LANL, ACS, and RSC all call tennessine a halogen too – but (1) we can appeal to IUPAC's written word, and (2) most importantly we have given ourselves an out by having an "unknown chemical properties" category, which no one can deny Ts truly belongs to at the moment even if some definitions might also give it a pass as a halogen! That neatly solves how to deal also with questions about category overlaps: although most sources don't stripe, the result should be OK as long as the element in question is shown in one of the two categories it really fits in.

We thus end up with a totally consistent set of categories: IUPAC is used for the common categories (AM, AEM, Ln, An, TM, halogens, noble gases), with "does it appear in Britannica, RSC, ACS, and LANL" as a hopefully unassailable benchmark for "common" when it comes to beginners' literature. (Admittedly this does mean that group 12 could go either way. I have a thread about that, it is a separate issue.) Then for the region in between; the metalloids are charted by Sandbh's amazing work at lists of metalloids as just the common six, and then the ones to the right are called nonmetals and the ones to the left post-transition metals, which is also common. So everything is well-informed by reliable sources.

So that should reflect the literature situation, in which At is considered a halogen nearly universally, but its status as a metal or a nonmetal is up for debate with most basic treatments thinking of it as a nonmetal like its lighter congeners and most more advanced treatments saying "um, that might not quite be the case for all of its behaviour". Faced with the situation of a poorly understood chemistry and reliable sources still not quite agreeing about whether astatine is a nonmetal, a metalloid, or a metal, the "halogen" category therefore neatly sweeps under the rug the problem of just what happens here, when it is not something the average layman should be too interested in.

As YBG has already made statements about being neutral about whether or not to change, this is mostly meant for consideration by R8R and Sandbh. DePiep may be interested as well.

Illustration:

(Avoiding the plurals is also me playing with semantics to be technically correct while being simple. Because if I see a legend saying "halogens", I get the impression that "oh, these are all the halogens". Whereas if it says "halogen", then one can argue "yes, it says these five are halogens, but it doesn't say that these are all of them". Or maybe this is getting a bit too desperately twisty. ^_^)

As a final reiteration of the salient point: this table makes no claim at all about whether astatine is a metal, a metalloid, or a nonmetal. All of that can be discussed in the text where the appropriate context is listed. All this table says about astatine is the incontrovertible fact that it is a halogen. If readers focusing on astatine do not want to read the detailed text about it and assume what is not said here to classify it, then we cannot really stop them from not reading the fine print when the issue is a fine-print one. Double sharp (talk) 23:00, 16 November 2020 (UTC)

@R8R, Sandbh, and EdChem: Comments requested. Double sharp (talk) 23:04, 16 November 2020 (UTC)

Sandbh

@Double sharp:

Thank you for your very gracious reference to the lists of metalloids as being amazing.

Considerations

Proposed nonmetal RFC

• I have a good option in the form of the proposed RFC, and I'd be interested to learn of the community's view, including from our colleagues at WP:CHEM.

Metallicity trends

• Having a legend in which the top half shows metal-metalloid-nonmetal is consistent with the L-R progression in metallic character.
• To the extent that we try and follow the literature, I'm uncomfortable with the observation that, according to my (as yet unpublished) COPTIC database, 85% of 62 tables surveyed showed sundry categories (50%) or metals-metalloids-nonmetals (35%).
• In the p-block we can see the progression in nonmetallic to metallic character going down the groups. This progression wholly or partly follows a nonmetal→metalloid→metal trend.
• We know iodine has a metallic appearance under white light; that its crystalline structure resembles that of gallium; that it's a semiconductor (band gap ~1.3 eV); that it's electrical conductivity is better than that of selenium; and that its bonding has some delocalized character. So, it's not surprising astatine has been predicted to be an fcc metal, with the team making this call including the recipient of the 1981 Nobel prize in chemistry. The third author, Ashcroft is a solid state physicist; his textbook on solid-state physics, written with N. David Mermin, is a standard text in the field. We know their Sep 2013 paper has been cited 35 times without dissent.
• For Ts, we note that, "GSI (14 December 2015). "Research Program – Highlights". superheavies.de. GSI. Retrieved 9 November 2016. If this trend were followed, element 117 would likely be a rather volatile metal. Fully relativistic calculations agree with this expectation, however, they are in need of experimental confirmation." Although we have not called Ts yet, and that's fair.
• We can see the same thing happening in the noble gases, with some cationic behaviour by Rn. For Og the speculation is that, unlike the insulating noble gases, it is likely to be a reactive semiconducting metallic-looking solid, with a band gap of 1.5±0.6 eV cf radon 7.1±0.5 eV, and +2 and +4 as its most stable oxidation states.
• To some extent, I think that using "halogen" as a colour-category in a metallicity-based scheme is telling lies to children, so to speak. As opposed to halogen nonmetals which seems like it is plain as day. What is happening in group 17 is not so hard for the general reader (our primary audience) that it needs to be abstracted out of the way. It works OK with using AM and AEM since all the elements involved are metals. It runs into problems in the p-block, with the diagonal appearance of the nonmetal to metal transition. Which does continue into the halogens and the noble gases + Og (solid, reactive).

• The WP table is a metallicity table rather than a groupic table.

ACS, EB, LANL etc tables

• The ACS, EB, LANL, and RSC tables show that categorisation schemes vary, although the broad contours tend to converge, even if they are fuzzy at the boundaries. So there will always be a TM category, even though the L and R boundaries may vary.
• The RSC table is groupic, rather than metallic-nonmetallic progression based. The colour-category scheme of the EB table predates our own WP PT from 2003, and has not been updated in that regard since J J Lagowski (decreased 2014) drew it. The ACS table is more like our table and they show At and Ts as halogens. They also make dubious calls on the status of Cn to Og: they colour Cn as a TM when Cn is more likely to be an insulating nonmetal, and they colour categorise Og as a noble gas! The LANL colour categories have never been updated, since they were first posted prior to the WP PT. For ptable.com it shows astatine as metalloid.

Astatine's category history

• When we changed At from nonmetal to metalloid, there were no complaints. Ditto changing At from metalloid to metal. That's presumably consistent with most chemists not caring since so few would ever get to experiment with At.
• Regardless, astatine will remain, of course, a halogen.

"Textbook errors"

• Astatine raises a question on the reliability of RS. I presume we are all familiar with the idea of textbook errors, which arise from the authors not doing their research. JChemEd ran an extensive series on these for more than ten years.
• Nearly all RS have no idea about At. They copy whatever the other RS say, parrot-like, due to "publish or be damned" and figuring it is not worth the effort to do more research. They ignore relativistic effects, which explain why At is a metal rather than a semi-conducting metalloid, and e.g. the colour of gold, and why Hg is a liquid.
• We do however have the 35 more credible RS who have cited Hermann, Hoffmann, and Ashcroft on At being a metal, once relativistic effects are taken into account.
• That At is expected to be a metal has been around since it was discovered in 1940.
• Siekierski and Burgess (2002) in their short 195 pp book, Concise Chemistry of the Elements, refer to astatine as a metal; Turova (2011), in her 157 pp. book, Inorganic Chemistry in Tables, appears to refer to astatine as a metal (hers is an English translation from Russian, so that's why I say "appears").

Philsophy?

• Perfect is the enemy of good. I feel the general reader could understand the difference between halogen nonmetals and halogen metals, just as there are nonmetals, metalloids, and metals among the pnictogens and chalcogens.
• Oxygen is universally regarded as a chalcogen, as is sulfur. We do not have a chalcogen colour category since that (by itself) is not a metallicity category.
• The astatine gordian knot is not worth the bother.
• I feel At as a halogen does not represents a better solution.

In light of the above considerations I prefer to ascertain the communities' position via the proposed RFC. Sandbh (talk) 00:10, 18 November 2020 (UTC)

Thank you for your sourced opinion, Sandbh. Personally, I do not agree with your view of the source situation for At as a metal, but I feel that this case is better made at an RFC which I now agree is the better path forward.

I would nonetheless like to coordinate my proposal of this with yours in some way. To that end, I would like to first hear from R8R what he thinks of this proposal with At as a halogen, since his previous objection was to "halogen nonmetals".

My current proposals and thoughts are evolving slightly as we discuss these matters, so I would like to request some time before any RFC's launch, and some discussion of exactly how the nonmetal one will go forward. Depending on response from R8R and further discussion with Sandbh regarding how similar our proposals really are, we may have two separate RFC's instead of one. Double sharp (talk) 00:23, 18 November 2020 (UTC)

@Double sharp: I tend to think the non-metal RFC should go ahead. You could vote to keep two categories, and in the discussion section explain your reasoning for doing so, i.e. that you have a separate RFC that you will put, following the nonmetal RFC, should consensus not be established for three categories. That may the easiest way going forward, but let me think about it some more, including your comments and request at my talk page. Sandbh (talk) 01:52, 18 November 2020 (UTC)

Thanks for your response, Sandbh. Now I'll let you think. ^_^ Double sharp (talk) 01:53, 18 November 2020 (UTC)

@Double sharp: I think the nonmetal RFC should go ahead. Whatever happens, it could be followed at any time by an RFC focusing on the status At. That seems to me to be the cleanest way to proceed.

I support our historical decision to categorise At as a PTM. I haven’t read anything since then that merits revisiting this decision. The formation of a monocation in aqueous solution, without any inordinate difficulty, is a hallmark. Equally, At shows some non-metallic behaviour as could be expected from a PTM.

Astatine could equally have been expected to be a metalloid (i.e. a semiconductor) but for relativistic effects.

Upon reflection, our decision to go two nonmetal categories was less than optimal. I think WP:ELEM members recognised that. But nobody had a better option/s at the time. Now we do, IMHO. --- Sandbh (talk) 05:38, 18 November 2020 (UTC)

@Sandbh: Well, it's not seeming that clean to me anymore, unfortunately. T_T The problem is that the status of astatine impacts exactly what the category is going to be called: "halogen" vs "halogen nonmetals". And that was why R8R opposed Compromise 1, so it is important. As it stands, my proposals are getting further and further away from what you want, unfortunately. That's why I feel like I should try to get mine over with first, because mine is general for a large swath of the PT, and yours is about just the nonmetals; and because mine offers an alternative that is quite different from all of yours for the particular region your proposal is about. So I continue to feel like trying to lump mine and yours together in the same RFC does a disservice to both your proposal and my proposal. But maybe I will ask for some advice.

P.S. There is indeed strong evidence that some sort of cationic aquated At species exists, but its structure is not clear. The two alternatives listed seem to be (only one solvating water molecule!) and ; if it is the former then there is a real question about whether this should really be considered an "astatine aqua cation" or rather "protonated hypoastatous acid" (which is analogous to what iodine can do!) It also seems that in both cases according to my source the water has the partial positive charge and the astatine has the partial negative charge, thus raising serious questions about whether or not this is truly quite like a "metal" cation. Since the paper on condensed-phase At is only theoretical, I would say that I am not convinced that astatine should really be reported this authoritatively to be a metal. And I say this while continuing to believe in metallic astatine as a hypothesis! Double sharp (talk) 11:30, 18 November 2020 (UTC)

@Double sharp: If consensus is not achieved with the nonmetal rfc, you can follow with an rfc proposing the reintroduction of the halogen category including At. If consensus is achieved with the nonmetal rfc you can follow up with an RFC proposing the reintroduction of the halogen category. I don't expect their would be any drama trimming the "nonmetals" from halogen nonmetals. The nonmetal rfc has sufficient merit by itself. Other rfc(s) of course can be put forward; my feeling is that each deserves their time in the sun, preferably not at the same time.

I'm happy to agree to disagree with you on the status of astatine an fcc metal. Sandbh (talk) 21:57, 18 November 2020 (UTC)

@Sandbh: Well, that's kind of precisely the thing. My RFC is not going to be just about astatine anymore. It is currently planned to be about changing in entirety to the ACS/LANL/2010 scheme (see the bottom of this page), which impacts not just At, but also group 12, Po, and the superheavies. So the issue is separate but overlaps, which is what makes it difficult. That's why I've been asking if you would object strongly if mine comes first, because:

• if yours runs first, and it gains consensus, then mine is left in an awkward position because the issue is separate but overlapping: my proposal is related to the corner you're concerned with but impacts some of the rest of the table as well. So we are left without a clear consensus on the other issues. This becomes difficult unless I offer my scheme as a side one in yours, but then that drags it off your desired topic, also not good.
• if mine runs first, and it gains consensus, then it decides the whole thing, and then there's no such problem: your section (nonmetals) would also have been decided. (If that is not an outcome you like, note that you are perfectly free to oppose my proposal when it runs to preserve the status quo or even support your own split instead; if enough people are convinced to your side that I get no consensus, then you can open your own, and everything works well; and if enough people are convinced to mine that I get consensus, then it basically acts as though I oppose all your options in yours and ask for At in halogens, and enough people are convinced by that.)
• if either of us runs first and don't gain consensus, then of course there is no prejudice against the other running his.

Because of recent tensions I have been worried about how it would look to do that. However I feel that for this reason it might be simpler if I ran mine first. You are of course perfectly free to oppose mine when it runs, if it runs before yours. ^_^ I would like to ask that we consider it just as a matter of how each one can get its deserved time in the sun most effectively without overlaps. Double sharp (talk) 22:07, 18 November 2020 (UTC)

@Double sharp: ack and will take on per your reconsider request. Sandbh (talk) 23:08, 18 November 2020 (UTC)

R8R

I have to think about it some more, and I'm not ready to give an answer right now. I suspect, however, that even if we remove the metal/nonmetal distinction from the PT, it will still be thought to be somewhere there. It will still look like the categories to the left are metals and to the right are nonmetals, and astatine will still be thought of as a metal. Besides, is removing those two categories, metal and nonmetal, a good thing or is this loss at least outweighed in terms of encyclopedic value by having a new halogen category? I doubt it.

On the other hand, even if it's unprecedented for us to have a category that has both nonmetals and a metal, that's still an interesting category name, and a popular one, too. I need to think whether I can justify having it in our new element system and whether I'll be able to call that scheme consistent. I'm currently leaning no, but your blocks-only proposal also seemed too much to me at first, but I grew to it over time and even wanted to support it had you ended up having an RfC on the matter.--R8R (talk) 06:34, 18 November 2020 (UTC)

@R8R: Thanks for your opinion, looking forward to your answer. I've been discussing these things somewhat and thinking of them on the grounds of what others here (Andrew D., EdChem, Jehochman, Softlavender) have said, so a compromise 3 should also come shortly. Double sharp (talk) 11:32, 18 November 2020 (UTC)

@R8R: Please take this as superseded by #Compromise idea 3: ACS/LANL/2010. Double sharp (talk) 13:16, 18 November 2020 (UTC)

Compromise idea 3: ACS/LANL/2010

Idea planned for DS RFC

NB: Actual RFC draft is at User:Double sharp/Category RFC. The below was just sounding the idea out with R8R first and will not be the actual RFC text! Double sharp (talk) 22:12, 18 November 2020 (UTC)

(I add this to preserve the previous anchor and to make it clear where this is planned to go.) Double sharp (talk) 22:11, 18 November 2020 (UTC) Here's a third compromise idea (and probably the one I will end up proposing at an RFC): follow the tables of the ACS and LANL in their category boundaries are exactly the same there. In the cases of disagreement of category names (next to the metalloids), simply be neutral and use "other metals" and "other nonmetals" to create a complete return to the version of 2010, before any of us three still discussing this (me, R8R, Sandbh) had ever touched the template.

(Well not quite, because at that point the discovery of Ts had not yet been announced. But I guess that is a side issue.)

This is also almost exactly the ACS table (except we have "other nonmetal" instead of "nonmetal"; I feel this is justified since noble gases are generally regarded as nonmetals). It is also almost exactly the LANL table (except it has also "post-transition metals" – but there are multiple names for those metals as detailed at Post-transition_metal#Related_groupings, and "post-transition metals" raises unproductive questions about aluminium, so I feel we'd better leave well enough alone). Anyway, the category boundaries are exactly the same there.

I recognise that this is not quite the current table. Scientifically speaking, if we look at the specialised sources, it is true that some elements could be treated better. One can make arguments about polonium and astatine, and also argue about the colouring of the superheavy elements. But:

• The perfect is the enemy of the good, following Andrew D. The proposed version is evidently thought good enough by serious English-language secondary sources (American Chemical Society, Los Alamos National Laboratory)
• The changes from this to the current table seem to have generally been because of elements like astatine where the situation is not well-known, and it seems to me that caring too much about this is not an ideal situation for the readers. That's based on things Jehochman has mentioned on YBG's talk page: Another way to think of this is that any diagram or table will oversimplify whatever it is representing. Some oversimplify one way, some another. Don't let perfect be the enemy of good. Imagine you are explaining chemistry to a first year high school student. What do they need to know? Any extra complexity an be explained in the article text (who say this alternative, who says that alternative), or footnotes.
• Jehochman suggested the LANL table as a compromise on the ArbCom request page: PERIODIC TABLE OF ELEMENTS: LANL - the closest thing to a standard that I can find. This is "A Resource for Elementary, Middle School, and High School Students", which means it should be suitable for a general audience such as Misplaced Pages readers. If anybody is an advanced chemistry student, they will know where else to look to find more advanced information. ArbCom can't impose a standard, of course. This is a suggestion that might help us all find a path forward from this dispute.
• No one at the basic level we're writing for, apparently, cares that the chemical properties of the heaviest elements are unknown when it comes to colouring them in anyway. According to the 2016 IUPAC recommendations on naming new elements, Ts and Og are in fact new elements of the halogen and noble gas groups. So we already have an official precedent for doing what ACS and LANL do and calling Ts a halogen and Og a noble gas even if their chemistry probably does not match. Do I think this is not a good idea scientifically? Yes, I am quite against it as a matter of opinion, and find it silly to call an element that probably behaves like tin a noble gas. I also find it awkward to colour elements when their chemistry has not yet been investigated. But if the sources do it, I feel that is far more important.
• The group 12 issue is not really a matter of 50-50. Jensen, the source for this assertion in post-transition metals, wrote that it is 50-50 only in "advanced monographs on coordination chemistry and organometallic chemistry". Later he discusses the situation in "general chemistry and lower-level inorganic texts". Then he writes when describing periodic tables with transition and main-group elements distinguished "In all cases, the Zn group was incorrectly labeled as being a member of the d block or transition block". So, although he doesn't like it, it seems that group 12 as transition metals simply overwhelms in basic sources. The exclusion is based on advanced considerations about bonding that don't need to be gone into at this level, and anyway opens some cans of worms with mercury.
• Post-transition metals are explained above. The term isn't IUPAC-approved, there are multiple terms for this part of the periodic table (post-transition metals, poor metals, chemically weak metals, semi-metals, B subgroup metals), and therefore I feel that if "other metals" is good enough for the ACS, it ought to be good enough for us. Anyway it gets rid of tiresome questions about how aluminium can be post-transition when there are literally no transition metals before it.
• This is what we had in 2010 before any of us touched anything, which may be welcome given some statements of Softlavender at the ArbCom case. We had this scheme from 2002 in fact. Presumably, all of us feel that it is not perfect, but there is no consensus in sight for how to improve it. Since we changed it, we've been changing things back and forth every few years, but this remained stable for literally eight years. If you don't count the "uncolouring" of the heaviest elements, then literally ten years. That tells me that this is good enough and therefore I suggest that we stick to it. And since this predates any of us changing things around, I believe this sends a powerful signal that we are willing to drop the matter and stop the arguing, that has been productive at nothing except getting the community annoyed at us.
• To me, all ideas of philosophical consistency pale before (1) following the sources because they don't seem to care about it and (2) ending the issue using extremely common category names.
• Yes, even if we don't say it, it will still look like we have a division into metals-metalloids-nonmetals, and it was still look ike At, Ts, and Og are nonmetals. However, given that experimental results for At are still unclear (At displays both halogenic and metallic properties in experiments), experimental results for Ts and Og are still nonexistent, and the predicted metallicity of these three elements is an entirely theoretical matter at this point, I feel that it is not our place to intervene at this moment, and that the reader will not, in fact, be ill-served. On the contrary, the reader might actually be served well by it, because it will become clear when s/he reads the specific articles that there is some difference between sources in how they use the terms "halogen" and "noble gas" and whether or not they imply an element is a nonmetal. Remember, this is only for a general thing, and in specific cases we do not have to follow it 100%. Just when there are no pressing arguments for a different scheme, as EdChem said above when discussing Compromise 1: YBG and Double sharp, remember that the outcome doesn't have to be dichotomous. Yes, we can only represent one way on the PT shown, but that does not preclude a section later showing a different perspective linking to a more detailed discussion in a different article. There is no reason that the halogen article has to start with halogen = group 17 and not address / recognise that more limited sub-sets are used by the terms in certain situations and further down in the article. A general rule need not apply at all times.

Given his reaction to colouring At as a halogen above as Compromise 2, I think Sandbh is likely to be against this, but it would be nice to have a confirmation. That being said, this is most likely what I will put forward as an RFC anyway. Obviously, my RFC statement will not be this long, and will simply neutrally suggest this form. So I am mostly looking for R8R's opinion here before I go ahead with an RFC. I would like to ask him to take his time and consider.

Oh, and just to clarify: I still scientifically don't like this scheme very much in how it treats those problematic radioactives near the bottom, and I still think it's not perfect. But I also think that if we don't give way on some things, we will never get any compromise out. If everyone goes for "my way or the highway", this will never end. But hopefully, the gravitas of appealing to ACS and LANL (since IUPAC doesn't, unfortunately for us, have a coloured PT) will mean that the result will be accepted as a permanent status quo. And we may apply this principle to future elements too: follow the sources, and if they want to apply categories by group, we'll do that. 119 will be an alkali metal, 120 will be an alkaline earth metal, 121+ will be superactinides for a while, and hopefully by the time that immensely long series is finally left there will be some consensus on how the periodic table looks like beyond there. I think ACS/LANL can have a chance at stopping the debate, and to support it I have put thoughts of "but this is scientifically wrong, I have specialised sources!" and trying to win aside. It all depends though on the opinions of others here and that of the community when I put this up for RFC. Double sharp (talk) 13:10, 18 November 2020 (UTC)

Aside from the group 12 question, I see three questions here:

1) polonium is now a metalloid,

2) unknown chemical properties are gone, and

3) a new halogen category.

Here's what I think:

1) Why is polonium a metalloid? Is metallicity of polonium being put into question? Or is it an accidental mistake?

2) I'm inclined to say we should not recolor those elements whose chemical properties remain unknown; it's not right for us as an encyclopedia (supposedly a collection of firm knowledge) to do so. Saying that flerovium is a metal seems a speculation at this point. While from the educational point of view it doesn't really matter, we're an encyclopedia first and foremost; I'm all for accessibility and education, but they shouldn't compromise our encyclopedic goals.

3) If you follow my logic with predicted metallicity, then here's a question. Let's say we color elements 113-116 back to light gray. What do we do with Ts and Og then? If they are also gray, then we have pretty much returned to the previous proposal. If not, this will likely raise questions. Neither solution seems good to me.--R8R (talk) 17:07, 18 November 2020 (UTC)

@Double sharp:--R8R (talk) 17:08, 18 November 2020 (UTC)

@R8R: It's not a mistake, it's deliberate. Reason being that both LANL and ACS colour polonium as a metalloid, and it can be chemically justified.

I actually agree with you that it's naughty to colour some of the elements beyond 108. It's speculation indeed. The problem is that nobody seems to care. ACS does it, LANL does it, Britannica does it, RSC does it, PubChem does it (actually uses exactly this scheme). And moreover: is it really accurate to say chemical properties of flerovium are unknown? They are not wholly unknown, some experiments have been done. Only their interpretation is not sure. So "unknown chemical properties" also seem problematic for me for Cn, Nh, and Fl. So either way we have to lie a bit. Either we say Cn is "unknown chemical properties", which doesn't match most general RS, and then walk it back and say "oh, it's not unknown after all"; or we say Cn is a transition metal sort of by fiat, which matches most general RS, and then say that combined results of theory and experiment make it seem possible that actually something different happens. But the latter is cutting-edge stuff (paper published literally last year) and I don't think it should be the first thing we're telling the beginners when that's not the first thing anyone else who hasn't been influenced by us seems to be telling the beginners. We're a collection of knowledge sure, but knowledge in the sense of what RS say. It's WP:VNT after all. If most basic sources say something, even if we know it's simply not true, who are we to go beyond them for a standard simplification, especially when the correct thing is much harder to say?

Anyway, that's just my view, you can disagree. Thanks for sharing your view, it is as always thought-provoking, but I think this means neither you nor Sandbh are going to convince me otherwise for this one, so I'll start the RFC following advice from Jehochman in probably a few hours (to take my mind off it and then come back for any last-minute things that may become apparent). Double sharp (talk) 18:33, 18 November 2020 (UTC)

I think you've got a solid case on your hands. I still think that the present scheme is, in fact, the scheme that does the job and it has the benefit of not raising those little questions (which I consider the most important), go no further. However, there is, without a doubt, good merit in yours, and it's likely to obtain at least some (or quite possibly, a lot of) support. I won't be too disappointed if it is chosen by the community.--R8R (talk) 19:09, 18 November 2020 (UTC)

Right, R8R, thank you so much for your really kind words. I appreciate them a lot. ^_^

I have consulted Jehochman and EdChem on Jehochman's talk page. The former has already looked at it; the latter plans to. If all goes well, I should be following their advice for RFC drafting, and we should be able to launch soon enough. Like you, I won't be too disappointed if it is not chosen, as I'm not trying to "win" anymore. ^_^ Double sharp (talk) 20:53, 18 November 2020 (UTC)

I would also like to ask Sandbh if he is OK with such an RFC (yes or no, supporting or opposing this option) starting within a day or two probably. Double sharp (talk) 21:07, 18 November 2020 (UTC)

@Double sharp and R8R: In principle, I'm OK with such an RFC, as I am with any RFC. My only request is to proceed with our RFC's in an orderly and cooperative manner. The nonmetal rfc was proposed on 29 Oct. It is now 19 Nov, three weeks later, and has not yet launched.

The nonmetal rfc proposal prompted mention of another five RFCs. That's OK but I do not think proposing an RFC in response to an RFC and asking for the second RFC to then go before the first RFC is a "proper(?)" way to air of treat the merits of the first RFC, unless the proposer of the first RFC agrees, which I don't. So, I politely oppose starting this option before the nonmetal rfc. Sandbh (talk) 22:13, 18 November 2020 (UTC)

Thank you for your polite opposition, Sandbh. Could I politely ask you to reconsider? I feel that the problem here is that this RFC is about a much more general issue than the first one: for example, yours seems to take trifurcation as implying "halogen nonmetals" gets included, but as can be seen from R8R and I there isn't a consensus for that. At least I would favour trifurcation without "halogen nonmetals", but it is very difficult for me to !vote for that without dragging your RFC off topic into other issues, because it is not one of your options. If you do not wish to reconsider, then the best !vote I can think of that I could put on your RFC that really reflects my position would be along the lines of "oppose all options, as none of the Britannica, RSC, ACS, and LANL uses either 'reactive nonmetal' or 'halogen nonmetal', yet all options presume one of them is included. I intend an RFC on whether to institute the ACS/LANL scheme". But even that seems fairly confrontational in a way that I don't want (because it basically smacks of saying "I don't like this RFC and I'm going to start mine"), and it also drags things off topic because instituting the ACS/LANL scheme as I would really support means revisiting many issues outside the nonmetals.

Therefore, for the purpose of discussion flow, could I please ask you to reconsider so that we can go from global issues to local issues and avoid this problem? Because if my RFC proceeds first and is rejected, then I no longer have any of this problem because it would mean my preferred option was truly off the table, and I could shift my support to one of your options without any difficulty.

I would like to ask EdChem and Jehochman for their considerations as well as to the best way to proceed. Double sharp (talk) 23:03, 18 November 2020 (UTC)

@Double sharp: I'll reconsider as per your request. I have a few RL obligations today and cannot predict when I'll be able to get back to you (hopefully later today). Sandbh (talk) 23:05, 18 November 2020 (UTC)

Thank you, Sandbh. I've left a note at Jehochman's talk page noting that the discussion is ongoing here firstly. Please take your time. I agree that all of our proposed RFC's have merit; I am simply making my request out of a desire that they proceed in a way that allows the discussion to flow more easily and in a focused manner, making sure any RFC reflects the current variation of options that have reasonable support. I have no strong feelings that my proposed order is the best way to do it; I just feel that such a procedure, no matter how it is attained, would be beneficial, and ask that we work together to find one first of all. Double sharp (talk) 23:10, 18 November 2020 (UTC)

@Double sharp: Something you wrote about the silliness of our "unknown chemical properties" category got me thinking. If we changed this to "category unconfirmed" that would address your concern. Astatine could also be shown this way since its fcc metal structure only has the status of a calculated theoretical prediction, albeit with 35 citations. F–I could then be colour-categorised as halogens.

My only reservation is that a +1 monocation of astatine does exist in aqueous solution in the 1−2 pH range., 55 citations. OTOH nobody seems fussed by the reported cationic behaviour of radon.

Subject to your thoughts, I doubt this would necessitate an rfc. That is, we could change unknown chemical properties to category unknown; change At to cat. unk.; and F,Br,Cl,I to halogens. And, I suppose, the SeSPONCH nonmetals to other nonmetals, followed by the one voter, many votes rfc to change other nonmetals to one of the 22 other possibilities, of which I guess only a half-dozen or so are prospects.

--- Sandbh (talk) 11:18, 19 November 2020 (UTC)

@Sandbh: (edit conflict) I've considered what you're saying. It is a good idea, and I would definitely move my support to it provided my current idea doesn't get support. (And if so group 12 would probably be brought up later and afterwards separately.) After all, there are some discussions about At along the lines of "is it more like a halogen or more like a metal", so we're technically not wrong there either.

But, my main concern with the "unknown chemical properties" thing is not so much what you've mentioned: that was a secondary thing. It is rather that the big names (LANL, RSC, ACS, Britannica, PubChem) don't seem terribly fussed about the lack of chemical studies and colour things in anyway, and that even IUPAC seems to be referring to Ts and Og as members of the halogen and noble gas groups even though their chemistry is probably not going to be anything like that of a normal halogen and noble gas. And that to me is a somewhat convincing argument. You don't have to be convinced by it yourself, but if R8R also thinks the idea has merit and there's a solid case while disagreeing, then it seems to me that I should bring it up for community consensus. We all seem to think each other have plausible cases, arguing won't move anyone, but trying to get a community consensus will let the community decide. But, for the above reasons, I think this is more difficult if the RFC for this scheme doesn't go first. It's not important to me if it gets up or if it crashes and burns; what's important to me is that this idea and the short argument for it be considered by the community.

Therefore, I apologise, but I still have to reiterate my earlier request for an RFC for this one to go first. If it doesn't pass, then I'll indeed have no problems supporting "category unknown" with astatine also included, F-Cl-Br-I "halogens", H-C-N-O-P-S-Se "other nonmetals". Double sharp (talk) 12:16, 19 November 2020 (UTC)

@Double sharp: I'll sleep on it. Sandbh (talk) 12:23, 19 November 2020 (UTC)

@Sandbh: No problem. Have a good night! ^_^ Double sharp (talk) 12:26, 19 November 2020 (UTC)

Update: no, I can't currently bring myself to support At outside halogens if halogens is a category. That's just because almost all authors who use halogens as a category include it. If my proposal fails, then yes, I will think about it. Double sharp (talk) 21:22, 19 November 2020 (UTC)

An attempt to clarify

The following is addressed to Sandbh in a spirit of trying to resolve this issue agreeably. Double sharp (talk) 23:35, 19 November 2020 (UTC)

I don't feel that this compromise is dragging multiple issues into one. On the contrary, I see it as a single issue on how policy is interpreted to apply to this case. To that end I will therefore first explain slightly more briefly why I am supporting the ACS/LANL/2010 option. Basically: there are lots of categories – but when it comes to colouring the PT, quite a few things seem to be generally agreed, such as AM/AEM/Ln/An/TM/halogens/noble gases with what happens in the middle iffy. In particular, it seems nearly universal for beginning texts to take group 12 as part of TM and to have halogens as a category with astatine (and probably now even tennessine) as a part of them. When there is disagreement like in the region between TMs and halogens + H: it may be due to textbook errors, or it may be due to different contexts. But then I feel like while following one source is OK (since the others are usually very close), weighing up the sources ourselves for iffy elements and producing something that none of a whole bunch of common sources do (including an "unknown" category or either removing halogens or removing At and Ts from them) just trips my OR/SYNTH meter.

So it's not about dragging three issues (group 12, Po/At/superheavies, category names) into one. Rather it is about one issue: does the community agree with my interpretation of how policy applies to this issue, or not. It's very possible that my interpretation is not the community-supported one, in which case I will of course change it to conform to the community's. However, that will remain to be seen by the community. Considering the nature of this rationale, I feel that if your RFC starts first, I don't have a good option on how to !vote. If I honestly state my preference, I drag the scope beyond nonmetals alone and into the whole table, precisely because the nature of my argumentation for supporting this makes it an "all or nothing". If you are OK with such a !vote appearing on your RFC, despite it dragging the topic open that way, then there is no problem. (When I say OK with it, I mean obviously whether you think it conforms to the process to !vote that way; obviously, as is your right, you would probably not support it, but that's not the point.) But if you are not OK with such a thing, but you still go first, then I feel we end up at an impasse where I cannot honestly !vote for my preference. Forcing group 12, polonium, and the superheavies outside the scope of an RFC means that I cannot !vote honestly for my preference, because my argument simply doesn't work for changing just the nonmetals only. It's about following categorisation boundaries wholesale and so it only works if you change everything at once. Which is why I feel we need an RFC to see if there is consensus on those issues one way or the other first. If there is in favour of the logic that leads to ACS/LANL/2010; then we're done. If there isn't; then I see consensus is not for that logic, so I change my policy interpretation in accordance with it and seek something else, in which case I can quite likely !vote for one of the options on your RFC. But not yet, which is the problem.

If a way can be found in which I can make an honest !vote given my current justification, then all my objections evaporate, of course. But I can't think of a way of doing it without either not being honest about what my argument leads to or going off-topic. Whereas you could easily respond to my RFC proposing anything you want, because mine is about colouring the whole table and therefore none of what you're suggesting would ever go off-topic. You could write in any option you want, really. You could for example !vote for one of your preferred trifurcation options, maybe with your preferred option replacing "other nonmetals", or maybe with "other nonmetals" first with an eye to another RFC to see if it should be changed. All that seems to work and I would not consider it off-topic at all.

I hope we can resolve this amicably between ourselves. I note that you said at Wikipedia_talk:WikiProject_Elements#Discussion_4 the following text: That said, anyone can put an RFC at any time. (13 Nov) That makes me encouraged that we can agree on something, although I have no intention to start an RFC unilaterally if you don't agree. So: given that my previous RFC didn't even last 24 hours (started 20 Jul, withdrawn by myself 21 Jul even if it wasn't formally closed till 6 Aug, so I've barely used any RFC air time in the first place; you won't find any comments there after 21 Jul apart from the closer), and given the asymmetrical nature of the problem (I can't seem to !vote honestly on your RFC without going off-topic, but you can do so on mine), could I please ask for your agreement to allow mine to go first? If you say no this time, I'll accept it and try to somehow find another way, but for now my request stands.

(And to make it clear by repetition: if you don't agree, I have no intention to start mine unilaterally. I would want to consult EdChem for advice, but I do not wish to do anything to jeopardise the current peace.)

P.S. I agree that the way the ACS/LANL/2010 scheme colours the heaviest elements sucks scientifically, that it is stuck in a weird limbo between being a metallicity and a groupic table, and that polonium is better regarded as a metal without any qualification, so I would like to ask in good spirit that people do not try to convince me of what I am convinced by already. ^_^ I am supporting it for other reasons, which is that the tables of PubChem/ACS/LANL all seem to care about none of those issues and therefore that I am not convinced we should either. Please understand that my support of something for WP does not imply my personal support of it. For another example, I know R8R and I support Lu personally if you asked us what we thought was the better form, but I know he thinks and he has convinced me that for now WP should show La pending the outcome of the IUPAC project. But, again, that doesn't mean either of us personally support it and would show it if we were writing about periodicity outside WP: we wouldn't. Double sharp (talk) 21:22, 19 November 2020 (UTC)

@Double sharp: I'll reply at my talk page, with a ping of course, Sandbh (talk) 02:00, 20 November 2020 (UTC)

• @Double sharp and Sandbh:, this thread for archive? -DePiep (talk) 23:12, 5 December 2020 (UTC)

Conclusions to recent threads, as at 15 Nov

For those who are overwhelmed by the threads.

• Interpersonal conflict DS vs Sandbh is totally resolved and over. Not sure about the other one, but seems to be developing promisingly.
• Argument over La vs Lu placed on freezer until IUPAC says something. DS is happy to have it at La for now, recognising he has made an attempt to change it already and that it has not resulted in a consensus in favour of his preferred Lu. He plans to revisit the situation when IUPAC says something (which based on info from Sandbh will probably start soon, albeit coming in various stages until the final IUPAC decision or lack thereof is made clear), as that will impact the source situation in a way that nothing else probably will . So that argument is settled.
• Remaining categorisation questions raised civilly here and seem very close to being resolved.
• Everyone is working together fine with no need to pass everything to external parties, although occasional questions where outside expertise might be helpful are of course still being brought up.
• Plans have already been set in motion to improve periodic table the article, with EdChem's excellent proposals being kept in mind as a guide.

Hopefully to be updated happily as news to ArbCom members that everything is settled without need for their involvement. Double sharp (talk) 23:17, 15 November 2020 (UTC)

Major: Element articles, untangling

I came upon Wikiproject elements after doing some cleanup on the lithium page, and being struck by the combination of two topics in one. It seems this is quite general to many articles, at least according to a quick scan of the first three rows of the PT.

I recommend that each article be divided into two.

1. New articles entitled, say, Elemental XXX. This article will be in exactly the same form as that recommended for all other chemical compounds in the Manual of Style/Chemistry. After all, elemental forms are only unique chemical compounds in that they only contain one element. No need to treat them differently IMHO.

2. The existing article will have information specific to the chemistry of the elemental form migrated out except where it illustrates the general chemistry of the element.

Looking forward to your reaction to the concept. Please don't bother wordsmithing the two sections below, they serve only to illustrate the principle.

Sections to be migrated out of the article "Lithium"

Infobox lithium:

- Move from Infobox 'Physical Properties' and 'Other Properties'

- Copy from 'History' In 1855, larger quantities of lithium were produced through the electrolysis of lithium chloride by Robert Bunsen and Augustus Matthiessen. The discovery of this procedure led to commercial production of lithium in 1923 by the German company Metallgesellschaft AG, which performed an electrolysis of a liquid mixture of lithium chloride and potassium chloride. to History

- Copy from Chemistry Lithium reacts with water easily, but with noticeably less vigor than other alkali metals. The reaction forms hydrogen gas and lithium hydroxide in aqueous solution. Because of its reactivity with water, lithium is usually stored in a hydrocarbon sealant, often petroleum jelly. Though the heavier alkali metals can be stored in denser substances such as mineral oil, lithium is not dense enough to fully submerge itself in these liquids. In moist air, lithium rapidly tarnishes to form a black coating of lithium hydroxide (LiOH and LiOH·H2O), lithium nitride (Li3N) and lithium carbonate (Li2CO3, the result of a secondary reaction between LiOH and CO2).

When the metal burns strongly, the flame becomes a brilliant silver. Lithium will ignite and burn in oxygen when exposed to water or water vapors. Lithium is flammable, and it is potentially explosive when exposed to air and especially to water, though less so than the other alkali metals. The lithium-water reaction at normal temperatures is brisk but nonviolent because the hydrogen produced does not ignite on its own. As with all alkali metals, lithium fires are difficult to extinguish, requiring dry powder fire extinguishers (Class D type). Lithium is one of the few metals that react with nitrogen under normal conditions.

Copy from Military applications Metallic lithium and its complex hydrides, such as Li, are used as high-energy additives to rocket propellants.

Move section 'Precautions'

Example New article ("Elemental Lithium")

New article

Elemental lithium Names IUPAC name Lithium Identifiers CAS Number 7439-93-2 3D model (JSmol) Interactive image ChEBI CHEBI:30145 ChemSpider 2293625 PubChem CID 3028194 UNII 9FN79X2M3F UN number 2680 InChI InChI=1S/Li SMILES Properties Chemical formula Li Molar mass 6.94 g/mol Appearance metal Odor none Density 0.534 g/cm Melting point 180 °C (356 °F; 453 K) Boiling point 1,330 °C (2,430 °F; 1,600 K) Hazards GHS labelling: Pictograms Signal word Danger Hazard statements H260, H314 Precautionary statements P223, P231+P232, P280, P305+P351+P338, P370+P378, P422 NFPA 704 (fire diamond) 3 2 2W Chemical compound ElementalLithium is a metal with the formula Li. It is a soft, silvery-white alkali metal. Under standard conditions, it is the lightest metal and the lightest solid element. Like all alkali metals, lithium is highly reactive and flammable, and must be stored in mineral oil. When cut, it exhibits a metallic luster, but moist air corrodes it quickly to a dull silvery gray, then black tarnish. Production Lithium metal is produced by electrolytic reduction of lithium chloride Applications Elemental lithium has no known industrial uses but it is used a precursor in academic synthetic chemistry. It can appear as an unwanted layer in lithium batteries, and there is much research to develop solid state batteries which could use lithium as cathode. References "Lithium 265969". Sigma-Aldrich. Technical data for Lithium Archived 23 March 2015 at the Wayback Machine. periodictable.com External links International Chemical Safety Card 0710

(end of new article)

--Oldboltonian (talk) 23:49, 16 November 2020 (UTC)

@Oldboltonian: Although I have some sympathy to the idea, I am eventually against it. First of all, most general inorganic texts cover the elemental form's properties within the context of a general discussion of the compounds and chemistry (Greenwood & Earnshaw, Holleman & Wiberg), so this would go against a conflation that is perfectly standard in the literature. Second of all, following on from that point, the distinction between an element and a simple substance seems rather better-known in other languages than it really is in English. Third of all, articles on compounds usually discuss their chemical reactions and the results thereof (e.g. sulfuric acid which of course mentions inevitably the sulfate anions), so we can hardly get a clean cut between the simple substance and the element anyway. Last of all the fact that we have been doing things this way since 2002 when the project was founded, combined with the difficulties that I have just described, tells me we should probably leave what has worked well enough for almost two decades alone. But it would be interesting to hear the views of others. Double sharp (talk) 00:22, 17 November 2020 (UTC)

P.S. The project is also having a few tough times with conflicts. That doesn't mean you can't suggest something now, but I do recommend we all tread a little bit carefully to not ruin a peace that has just been starting to form a few days ago. My sincerest apologies that you had to encounter us at such a terrible time, and my welcome to the project in case you want to stick around. ^_^ Double sharp (talk) 00:26, 17 November 2020 (UTC)

Article distinction

• What, you think, is the essential separation ground (say, the test question: why does subtopic X belong in article Li/1 or Li/2?). Following, could the titles be like, disambiguated: , , , , . -DePiep (talk) 00:58, 17 November 2020 (UTC)
  • pending any other responses as requested by Double sharp, @DePiep: How about: information specific to the elemental form(s) of the element, in particular history, properties, uses (primarily industrial, secondarily academic), and above all hazard data. Otherwise, into the main article as it exists today. Not sure I follow your second question. There would be a lithium article, therein a section 'Compounds and Chemistry', therein a succinct description of the chemistry of its compounds including theelemental form, and a link to main article 'Elemental Lithium'? --Oldboltonian (talk) 14:06, 17 November 2020 (UTC)
    • With the reader in mind, I ask the question: what is the separation ground? If you have property X of Li, which check do you perform to decide: should go in article Li-such or in Li-so? Then my 2nd question is: This same separation ground should be understandable from the title (to me, "elemental" is not). In short: If you have two lithium articles, what's their difference? -DePiep (talk) 17:24, 17 November 2020 (UTC)
      • Good question. For elements that are metals or gases one could add the word 'metal' (I note that you already gave your opinion on the use of the word 'metal' to indicate 'elemental' or 'gas'. For the common allotropes of C and O, these articles aleady exist. But I take your points, DePiep and double sharp, let's let nature take its course here, and where there is enough meat (like the C&O allotropes) there will be a separate article). Still I'd like to see the content around the chemistry and properties of the elemental state grouped together within the articles--Oldboltonian (talk) 11:59, 21 November 2020 (UTC)

Uniform structure of articles

In spite of there being a recommended structure for articles on the project page, adherence to the proposed structure is quite patchy (see table below, analysis just the first 3 rows of the PT. Any energy for editing towards some uniformity?

--Oldboltonian (talk) 12:18, 21 November 2020 (UTC)

@Oldboltonian: The guidelines are mostly historical and nobody really seems to be taking them very seriously. They were probably good for a phase-I standardisation, but at some point we moved into a more of a phase-II article improvement of getting as many to GA and FA as we could, and then the structures naturally diverge per element. So I'd say there's not so much reason to change. Double sharp (talk) 10:58, 23 November 2020 (UTC)

If they are merely historical artifacts, perhaps we should note that? YBG (talk) 03:22, 24 November 2020 (UTC)

Sounds cool to me, YBG. I've marked it as historical and taken it off the tabs list. That being said, I do think an updated version of these guidelines just for a "standard WP layout decision + colouring style" might be a good idea if/when one is decided on via RFC. (Former is frozen till we hear from IUPAC, latter is to come.) Or maybe that is better for the MOS like WP:ALUM. Anyway, we have time before we'll have such a thing. Double sharp (talk) 21:26, 24 November 2020 (UTC)

Isotope page improvements

I propose 3 changes to isotope pages: First, the Isotopes of dysprosium page says
Dy
is the heaviest theoretically stable nuclide. It is actually
Zr
because the stable nuclides in between are either theorized to decay or spontaneous fission. Second, mark the remaining observationally stable nuclides to observationally stable (
Ir
to
Tl
). Third, adding navigation links to nearby element isotopes pages, sort of like on the element pages. -UbnBr₂ (Talk | Contributions | Actions) 04:57, 23 November 2020 (UTC)

About the navigation you mention: what do you have in mind? With elements, it is simply four (N-E-S-W in the periodic table). Would it involve decay chains?

FYI, a complete navigation table, isotope pages by periodic table, is at the bottom, like at Isotopes of dysprosium#References. This navigation aid does not show in mobile view (being a navigation aid). OTOH, if you see a strong content-based relationship in 'nearby', that could be in body text and/or the isotopes infobox.

I note that I am not happy with the N-E-S-W neighbors addition to the element infobox. For the same reason: it is more practical navigation, not a strong content relationship (many more such flaw relationships exist in the PT that do not deserve to be in an WP:infobox). -DePiep (talk) 10:07, 26 November 2020 (UTC)

Yes, I mean N-E-S-W. The links will lead to the different elements' isotope pages. -UbnBr₂ (Talk | Contributions | Actions) 20:24, 27 November 2020 (UTC)

In the element infoboxes, the NESW links make sense because of the adjacent periodic table to provide context. Far more important to providing those links would be to display a periodic table with each cell linking to the Isotopes of X page. But the idea of showing the complete decay products for each isotope seems interesting to me. YBG (talk) 23:28, 27 November 2020 (UTC)

In general, I reject adding navigation aids to the infobox. Strongly related isotopes info (-pages, -series, -relations) may be added if there is good reason, when numbers are low. How many data points, say links, would that be? Keep in mind that the infobox should contain main article info only, not all info. -DePiep (talk) 13:29, 28 November 2020 (UTC)

For the element infoboxes, the N-S-E-W relationships make sense. E and W are next door in atomic number, and getting to the next or previous element makes some sense for navigation. N and S also make sense: they are in the same group and are hence homologous, so obviously also related. For the isotope infoboxes, E and W still make sense as they are the products of a single beta decay. However, N and S don't make any sense anymore because that second dimension of the PT is based on the electronic structure, not the nuclear one.

That being said, I think a really better solution would be to make the PT bigger so that you actually had enough room to place the symbols and click on them. Yes, that would imply 18-column rather than 32-, but it really is the same thing anyway. Then we would not need any NESW links anymore because they would already be there in the PT.

Showing complete decay products may be problematic due to branched decays. Double sharp (talk) 13:43, 28 November 2020 (UTC)

Another idea on my mind was linking the decay products to their respective isotope pages, but that would be too many links. -UbnBr₂ (Talk | Contributions | Actions) 06:00, 29 November 2020 (UTC)

Use of sources at WP:ELEM

I just finished reading WP:PSTS, which discusses WP:PRIMARY, WP:SECONDARY, and WP:TERTIARY sources. IMHO, I think understanding these policies and how they apply to our domain of knowledge would be of great benefit to our project. It would help us (a) create good content, (b) following WP policy, and (c) avoiding conflict. What our project needs is to interpret these rules and specify what are (1) primary WP:ELEM sources, secondary WP:ELEM sources, and tertiary WP:ELEM sources. I am not saying I have the answers here, just saying that we need to collaboratively reach a consensus on these three question. Thoughts? YBG (talk) 04:11, 25 November 2020 (UTC)

YBG,

1. Primary sources are those in the peer-reviewed academic literature reporting OR.

2. Secondary source are those primarily relying on secondary sources. Very crudely there are three tiers:

C. less RS;

CB. lies to children (ltc) RS

B. RS; and

A. "top-tier" RS, with reputation e.g. Wiberg; G&E; C&W; Oxford University Press, Pergamon Press

Misplaced Pages articles usually rely on material from reliable secondary sources.

3. Tertiary sources are dictionaries, encyclopaedias and compendium. Many introductory undergraduate-level textbooks are regarded as tertiary sources because they sum up multiple secondary sources. Reliable tertiary sources can be helpful in providing broad summaries of topics that involve many primary and secondary sources, and may be helpful in evaluating due weight, especially when primary or secondary sources contradict each other. Some tertiary sources are more reliable than others, and within any given tertiary source, some entries may be more reliable than others.

Principles

• Deciding whether primary, secondary, or tertiary sources are appropriate in any given instance is a matter of good editorial judgment and common sense, and should be discussed on article talk pages.
• There's no need to necessarily rely on ltc RS, since plenty of RS and top-tier RS are available. Indeed, a theme of distinguishing RS based on e.g. publisher reputation, author reputation, age of publication, etc pervades WP:RS.
• Higher-level texts, for elements that are hard to study, will be more reliable.
• The age of a source does not necessarily reduce its reliability.
• Introductory undergraduate-level textbooks have a reputation for textbook errors, , ,

—- Sandbh (talk) 04:08, 29 November 2020 (UTC)

• I find it very worrying and troublesome that now it has come this far . -DePiep (talk) 09:59, 22 December 2020 (UTC)

Launch of revamped 2010 category scheme

Some discussion between Sandbh, YBG, and myself has been going on at my talk page. It seems that the three of us are all OK with the 2010 scheme if Po is changed to being showed as a metal. Furthermore, Sandbh does not seem to think an RFC is necessary, and R8R has said above that he wouldn't be too disappointed if this scheme is chosen.

Therefore, in the spirit of WP:BOLD, I have decided to launch the new colour scheme.

I understand that Sandbh would prefer for some articles to be updated first to match, such as our articles on post-transition metal and the nonmetals. However, I feel that if we ask for too many prerequisites, we are likely to never get anything done. The current articles seem more or less serviceable in a pinch (the PTM one will need a small update to not call them PTMs anymore, but that's all); new ones can come later. After all, when element 119 is discovered, we will not be completely correct either, and there's nothing saying we cannot update things slower rather than faster.

I also understand that Sandbh and I currently disagree on whether or not the status of At, Cn, Ts, and Og should be flagged out in the navboxes. But since this disagreement does not extend to the actual colouring, I do not think we should let it stand in the way of the recolouring either, since it can easily be discussed afterwards.

Everything I do can, as always, be reverted and discussed afterwards.

Double sharp (talk) 16:32, 30 November 2020 (UTC)

Mostly done at this point (may have overlooked some), for templates. Images will have to wait. Double sharp (talk) 17:16, 30 November 2020 (UTC)

Most obvious pictures on Periodic table done. That should be enough for now, also to make things easier if it turns out that people object to this and it needs to be reverted. Double sharp (talk) 19:37, 30 November 2020 (UTC)

May we expect an overview of the reasoning here? As in, arguments, lines of thinking, source handling? -DePiep (talk) 20:26, 30 November 2020 (UTC)

@DePiep: Yes, since you asked. Give me a moment to put it together, most of it is from where something almost like it was proposed above as Compromise 3. Double sharp (talk) 20:49, 30 November 2020 (UTC)

No hurry, take three moments (especially if that serves the "briefness" part of concise ;-) ). Oh and may I add: pls also describe the content nuts, and how they were cracked. -DePiep (talk) 21:03, 30 November 2020 (UTC)

@DePiep: I wrote it below. Sorry that it is not brief, but since pretty much every category from transition metals onwards in the legend was affected, it is probably unavoidable. I tried to keep each one to one paragraph, allowing myself a bit more for the halogen/noble gas one since that is basically about two categories at once. Double sharp (talk) 21:41, 30 November 2020 (UTC)

I support Double sharp’s bold edits. Sandbh (talk) 21:53, 30 November 2020 (UTC)

@Sandbh: Thank you! Double sharp (talk) 22:02, 30 November 2020 (UTC)

Comment: I am glad to see progress happening and discussion between WT:ELEM participants. I am also grateful that Double sharp has provided a detailed rationale below, promptly responding to DePiep's request. I have been watching the user talk page discussion and wondering whether to comment, and I would like to make a procedural comment. I am all for collaboration and discussion but have been concerned that a user talk page discussion might result in a local consensus that might not reflect the consensus of this WikiProject or the WP community as a whole. It appears that this concern has not been borne out in this case, but I ask that all of us bear in mind that discussions that become likely to be implemented in article space be moved to an appropriate article or project talk page. Alternatively, bring the conclusion to a suitable page and seek broader input or endorsement. The final option (and the one I prefer least) is to make a BOLD implementation coupled with a talk page discussion, accompanied by a willingness to revert / step-back and discuss if anything proves controversial. Double sharp, I thank you for making clear that there was a user talk space discussion that led to your edit and for making clear your openness to further discussion should it prove necessary, which is very much in the spirit of BRD. DePiep, I thank you for asking politely for details and indicating a willingness to wait for a response rather than going straight to a reversion. Hopefully discussion will lead to a consensus that all support. Thank you also to Sandbh and YBG for working collaboratively on areas where agreement and productive incremental improvements to article space are possible. This implementation of BOLD appears to be working, which is great, but discussions leading to article space changes are best held or ultimately endorsed outside of user talk space, and I ask that this be borne in mind going forward. Thanks. EdChem (talk) 20:29, 1 December 2020 (UTC)

TBH, EdChem, I am a worried too about the From-Userspace-Boldly-Into-100-Articles. At leasst it could have a "proposal" step on this project page, building consensus. -DePiep (talk) 18:40, 10 December 2020 (UTC)

I agree that consensus should ideally be built on article talk pages or on a suitable project page (and then linked from the article talk page / with the edit). Userspace-developed consensus can be problematic, though as with much of WP, rules are not absolute. Double sharp, Sandbh, and YBG, I think you were all involved with user space discussion prior to these edits, would you please take note of DePiep's point and try to ensure, when seeking consensus, that the discussion include the relevant article talk page(s) and/or a WiiProject page such as this one, as appropriate? BOLDly making changes is encouraged, though I strive to temper boldness with wisdom... if a change is major or (more importantly) if I know it is likely to be controversial, discussion prior to implementation is desirable, and it starts in a much nicer environment than can exist after a bold edit and revert sequence. Taken further, a BRRRRRRRR sequence can get downright chilly! I'm all for keeping this project welcoming for Goldilocks and friends... not too hot, not cold, but nice and warm and comfy.  :)

EdChem (talk) 03:00, 11 December 2020 (UTC)

Rationale

I will ignore the parts of the periodic table that were not changed.

Mostly, sources that present periodic table colourings are working more or less at an overview level, not delving deep into the chemistry of a single element. They also are usually meant as a didactic introduction to signal out groups of like elements, especially the easy-to-cover ones that will be introduced early in chemistry. Therefore, priority was given to such sources as ACS, LANL, RSC, and Britannica. Priority was also given to IUPAC (as the relevant authority) when it had something to say, which is unfortunately not always. Further down the list were textbooks and monographs (e.g. Greenwood & Earnshaw, Holleman & Wiberg as very common reputable textbooks), and finally individual research papers for the "difficult" superheavy elements that are not well-known.

Group 12 as transition metals. This is something that has been strongly argued against. However, even Jensen who was strongly against this in 2003 had to admit that nearly all introductory general chemistry textbooks were doing it, and that it was only the advanced monographs on specific topics that sometimes don't (but even there it was 50-50). So do all four of ACS, LANL, RSC, and Britannica. IUPAC presents the definition including them first in the Red Book 2005 (their latest word on the matter): For example, the elements of groups 3–12 are the d-block elements. These elements are also commonly referred to as the transition elements, though the elements of group 12 are not always included. That is also evidently their primary definition, as when they discuss the organometallic nomenclature in pages 228 through 232, group 12 is not discussed with the main group elements. (For those who don't know: that means "the elements that are not transition elements.) Greenwood & Earnshaw as well as Holleman & Wiberg both consider group 12 as transition metals, though Cotton & Wilkinson don't. Therefore it seems more justifiable to follow the more general perspective. This also matches what we had in 2010.

Other metals rather than post-transition metals. Between ACS, LANL, and Britannica (RSC classes by group in this area, which is a rarer choice) there is not agreement on which to use: there are also a bunch of other names in the literature such as poor metals, chemically weak metals, p-block metals, B subgroup metals, and so on. As such it seems more prudent to not take a side and simply use other metals, which is indisputable as a catch-all qualifier. This also matches what we had in 2010.

Metalloids. There is not agreement between ACS, LANL, and Britannica about whether there should be a metalloid category. The first two have one as {B, Si, Ge, As, Sb, Te, Po}; Britannica lacks one, splitting {Ge, Sb, Po} to metals and {B, Si, As, Te} to nonmetals. Most English-language sources do have one, but are often a bit hazy about which elements are in there. The literature survey done by Sandbh whose results are at lists of metalloids was consulted: here {B, Si, Ge, As, Sb, Te} emerge as the most common six inclusions by far, with a huge drop to the next inclusions {Po, At} that have always had a certain amount of iffiness hanging around them due to their intense radioactivity. In general, it seems that most sources that focus on polonium do treat it as a metal: in this special case of a difficult-to-study but still fairly well-characterised element (unlike most radioactives, Po is not ignored or quarantined off into its own special section in Greenwood & Earnshaw), following them seems reasonable.

Other nonmetals. ACS, LANL, and Britannica are unanimous about splitting out the noble gases and the halogens from most of the nonmetals. Both are, after all, some of the first examples of group trends an average chemistry student encounters, and make sense to highlight at this overview level (that's not true for the pnictogens and chalcogens, say). There is not agreement as to whether these should just be called "nonmetals" or "other nonmetals" between them, but since "other nonmetals" is the only accurate one anyway (everybody agrees that chlorine and argon are nonmetals), we have used it. This also matches what we had in 2010.

Halogens, with reference to At and Ts. The problem is that there are two definitions of "halogen" running about, as there seems to be for a lot of these categories. In the literature focusing on the superheavy elements, there is some feeling I detect that you have no right to call an element a "halogen", a "transition metal", or a "noble gas" before you have actually characterised it chemically, and that if the properties are badly matching enough it's possible that Ts is not a halogen (e.g. here). So that's the specialised practice. However, outside the subset of chemists focused on these issues (which is very tiny given the short half-lives of the problematic elements At, Ts, and Og), nobody cares. ACS, LANL, Britannica, and RSC are unanimous about calling At and Ts halogens. Greenwood & Earnshaw and Holleman & Wiberg have no qualms about calling At a halogen (they were published before Ts was discovered). The 2005 IUPAC Red Book is absolutely clear that the halogens include At. It was published before Ts and Og were discovered, true, but IUPAC reports in 2016 made it clear that for them Ts and Og were in the halogen and noble gas groups. Indeed, that's the very reason why they are called tennessine and oganesson with the -ine and -on suffixes of their groups, and not -ium like for other elements. So clearly for them it means the whole group.

I feel that on the above grounds using the definition in which Ts is not automatically a halogen, and Og not automatically a noble gas, and using an "unknown chemical properties" category, is catering only to the superheavy-element community and not the rest of the chemical world that does not seem to care about such issues. I'm sad about that because I think the superheavy-element community has the right idea here, but they're not usually in the business of colouring periodic tables as an overview of the elements people actually care about.

Finally, I think it is simply premature to classify astatine as a metal, metalloid, or a nonmetal. As far as is experimentally known, it sometimes reacts characteristically like its lighter buddies in the halogen group which are all clearly nonmetals (fluorine, chlorine, bromine, and iodine). But sometimes it reacts more like a metal instead. Experimental results on copernicium appeared to suggest metallicity – but later theoretical calculations pointed out that the result was also consistent with nonmetallicity. Therefore I feel that any metallicity category here is an overreach, and that it is a good thing that we have a group category to save the day. We only have predictions telling us that At is probably a metal, that Cn is probably a nonmetal (like a noble gas), that Ts is probably a metal, and that Og is probably a semiconductor (so maybe a metalloid like silicon?). We can hardly use these predictions without some element of original research for Og, and the predictions have changed before for elements like Cn, Fl, and Og: who's to say they may not change again?

This also matches what we had in 2010.

Noble gases, with reference to Og. All of the above on Ts applies here to Og. This also matches what we had in 2010.

Colouring Mt through Lv. I agree that somehow it looks problematic, colouring them with category names that include "metal" when it's not actually sure if these elements are metals. On the other hand, ACS, LANL, RSC (for Mt through Cn), and Britannica all do that anyway. Besides Cn, no one seriously expects anything different, and for Cn, predictions have changed back and forth and they may well do so again. I also feel that it is not a problem that the category names become "formal" and a bit divorced from the actual meaning when everybody is making them so anyway. No one asks how bismuth is "choking" when that's where "pnictogen" comes from (well, I suppose you could choke on a large piece); no one asks how polonium can be an "ore-former" when it's too unstable to form ores; no one wonders why Be and Mg are "alkaline earth metals" in English (I know they do in Russian and Japanese, but that's not exactly the practice the English Misplaced Pages should probably be prioritising); no one wonders why the rare earth elements are not rare. With all this behind us, what's wrong with Og as a "noble gas" that is neither noble nor a gas, if the literature's mostly happy with that? This also matches what we had in 2010.

Therefore this colour scheme seems to be one of the best solutions within the bounds of what can be justified from sources to the whole problem. It was worked out as a compromise in the first place, seeking to follow the sources but at the same time try to be as "correct" in a way that both of us could agree on (and you know we sometimes differ very strongly about what exactly is "correct"). It is not what I would like at all for an off-WP situation, but for a WP situation I don't think we have the right to do anything vastly different based on the source situation. Because of the support I got for it on my talk page, I have decided to launch it.

Note that subtleties regarding how the status of Mt is not actually experimentally known, as well as the problematic metallicity of At, Ts, and Og and possible lack thereof for Cn, are all discussed where appropriate in the infoboxes. See {{infobox meitnerium}}, {{infobox astatine}}, {{infobox tennessine}}, {{infobox oganesson}}, {{infobox copernicium}}. For the everyday reader, I submit that what we show is enough as an overview, and that there is no need to footnote elements that to a first approximation no one cares about and that no one will get to see for a long while if ever.

Finally, we note for clarification that:

• So as to avoid having to rush a discussion at the last minute when elements 119 and 120 get discovered, we will add that unless sources do something massively different, element 119 will be coloured as an alkali metal (like Fr) when discovered, and element 120 as an alkaline earth metal (like Ra) too. Future elements may be in the g block already (I think they are, some people think they aren't, you can probably guess that this is a group 3 thing) and that will require some more far-reaching decisions based on what sources do. That is really unanswerable at present because until element 121 appears, nobody will have to make that decision. And it will be difficult for other reasons because IUPAC does not immediately put newly reported elements on its periodic table, but only does so once the IUPAC-IUPAP JWP evaluates the claims and deems it to fit the discovery criteria (so we won't have their guidance). Therefore any decision on them should probably be postponed till when the eighth period begins with elements 119 and 120.
• This scheme only decides the categories. In particular, it makes no decision about the precise colours used for the categories. Those may be discussed separately. It also makes no decision about the composition of group 3. That is a subject that will be returned to hopefully when Scerri's article appears in Chemistry International outlining what has gone on at IUPAC, as it is expected soon. AFAICS, there is no point in working with incomplete information when whatever this project concludes is going to be very important to the source situation and it seems that complete information is coming. We will see according to the situation.

Double sharp (talk) 21:39, 30 November 2020 (UTC)

Discussion of changes

I have a couple of minor comments / suggestions:

1. At the top, I would like to see some greater vertical separation between "Group" and "Period" – they look squashed together to me (this is a minor aesthetic point)
2. This may have been discussed / resolved before, but I think having the atomic numbers and symbols the same size makes the symbols less prominent as Z increases. I wonder if a smaller size for the atomic number and a larger size for the symbol would look better. I also wonder about slightly larger squares, so that there is more space around the symbol – the boxes at the bottom look crowded.
3. On the points about a colouring for 'properties unknown', I do take the point that many PTs do not have such a grouping, but it does make sense pedagogically to make the point that there are things that aren't known. It emphasises that the PT is continuing to evolve and that Chemistry is an experimental science where empirical evidence always takes precedence over theoretical expectations. I wonder if these cells might be divided diagonally and have two colours, or a footnote that the properties of recently discovered elements (XXX = symbols) need to be explored experimentally to see if they display properties in line with expectations. This idea may not find favour here at WT:ELEM, in which case of course I will respect consensus, and there may be better ways / ideas to address this point, but I do feel it is worth discussing.
4. Can the colour for the alkali metals be darkened to increase the contrast to the transition metals? Perhaps consult one of the WikiProjects that deals with accessibility for readers for comment on the colour scheme on different platforms and for readers with other-than-typical vision?
5. Would the asterisks between La and Hf and adjacent to Ce look better if vertically centred, like the period numbers and the double star for Ac / Rf / Th appear?

EdChem (talk) 20:45, 1 December 2020 (UTC)

Dear EdChem, thank you for your comments.

Regarding 1, 2, and 5, I have no opinion on the PT image layout issues. I think your proposals are sound, but since I am not the best at svg editing, I will probably have to leave it to somebody else (figuring out how to change the colours by directly editing the svg in a text editor was interesting enough for me ^_^).

Regarding 4, previously the alkali metal colour was indeed the darker   . It was lightened to    in a WP:BOLD action by R8R in September: see Wikipedia_talk:WikiProject_Elements/Archive_50#Alkali_metal_color. I personally found it OK, but since you've raised concerns similar to DePiep that this may be a problem, I have reverted it back to the darker colour. In any case, there has been some talk about revamping the colours altogether, in which case both problems will eventually go away anyway.

Regarding 3 (sorry for the despairing wall of text XD) – frankly I sympathise. But the trouble is that tension between two definitions of "halogen". If "halogen" means not only "in group 17", but also that the element has to act like {F, Cl, Br, I}: then you're right, because we don't know how Ts behaves, and theoretical calculations say it's really not much like a halogen at all, we don't have the right to say it's a halogen in that case. And that's how most sources focusing on superheavy elements indeed see it. But it seems that a lot of sources, including IUPAC, seems to see "halogen" as a synonym for "in group 17". In which case the category is formal, and Ts is automatically a halogen regardless of whether it acts like {F, Cl, Br, I} or not. We can argue maybe about periodic table placement demanding the electron configuration is known – but theoretical predictions say Ts is sp anyway, so that won't change things. Pedagogically, I think this is clearly stupid when apparently the closest match to known elements among Ts is Ga (yeah, personal communication from computational chemist User:Droog Andrey, but the interpretation matches the sourced predictions): a category purporting to put gallium and chlorine together as similar elements is quite clearly chemically deranged, so I'm sure that one putting together chlorine and tennessine likewise is. But I can't see a way out of it even though I'd dearly like one.

I think we may be in the same situation as astatine: even from its discovery, its discoverers were already noting it had significant metallic character, today we know it chemically seems to at least be a funny mix between halogen and metal properties, and chemists seriously working with it explicitly question whether it is more like a halogen or more like a metal (implying that it being a metal, which is getting some serious theoretical and experimental support but is not really sure yet, would rule it out from being a halogen). Eighty years on: nobody cares, IUPAC explicitly calls At a halogen, Greenwood & Earnshaw and Holleman & Wiberg explicitly call At a halogen, everyone calls At a halogen, and questions keep getting set on high-school exams asking to predict the properties of At (now also Ts) based on the lighter halogens (and therefore basically ask you to pick the wrong answer, amusingly). Anyone who actually cares about these elements knows that they behave quite differently (At seems mixed between halogen and metal properties chemically, Ts is probably like Ga), but (correct me if I am wrong, as I would love to be wrong) I cannot imagine that the proportion of chemists whose work significantly involves astatine is in any way significant among all chemists. Well: astatine has been known for eighty years, it's been on the periodic table for much longer than tennessine has. If that wasn't enough to make people care with an eight-hour half-life, then are people going to care about tennessine with a half-life under a minute? The end result is likely going to be the stupidity of calling Ts a halogen when calculations say it acts like Ga, and calling Og a noble gas when calculations say it acts like Sn (so if you thought gallium and chlorine in the same basket is chemically deranged, we now have basically tin and argon in the same basket, which is even more chemically deranged. ^_^). But there are worse already: if the rare earth metals are not rare, and beryllium is an alkaline earth metal despite being more amphoteric than alkaline, then I don't think most chemists are going to bat an eye about oganesson being likely neither noble nor a gas. Particularly when the most stable isotope of Og has a half-life of under a millisecond and probably its greatest relevance for the average chemist is completing the seventh row and making the periodic table look nice and complete, rather than anyone actually being able to use it for anything. And especially when IUPAC changed its naming rules in 2016 to make sure they'd be tennessine and oganesson with the matching suffixes; previously they'd have ended in -ium by the old naming rules! (The article that suggested this in Nature compounded the problem by implying that halogens and noble gases cannot also be metals – but when we get to At, Ts, and Og that may stop being true.)

In the end, I think what I would like putting on the educator's hat (halogens ending at iodine with astatine ambiguous; noble gases ending at radon) is just not easy to defend on WP, because it is hard to say "properties might not match" just because not everyone agrees that properties should define what a halogen or a noble gas is. And once you've done that and coloured Ts and Og, we can hardly not colour Mt through Lv when actually more is known experimentally about the chemistry of Cn than that of Og! (Not that Cn is conclusively characterised by any means: experimental results are rather underwhelmingly consistent with both Cn the eka-Hg and Cn the noble eka-Rn. I just mean that there actually have been experiments done on Cn and that has not happened for Og yet.) I cannot currently find a way out of this, which is why I put forward this scheme as something that might be a workable compromise. To my pleasant surprise it turned out to be one. If you can come up with a good argument why we could in this case follow the specialists' definition of halogens and noble gases – I will be most happy to uncolour At (as ambiguous) as well as Mt through Og. My worry is just that doing so will just make us inconsistent with what the general reader will see everywhere else and with what most people are defining halogens and noble gases are: pedagogically I agree with you and if I could find a way to justify it to myself would do it in a heartbeat. I have tried to address this point in some text at Periodic table#Categories, also at Periodic table#Further periodic table extensions, and in templates like {{infobox tennessine}} and {{infobox oganesson}}. Double sharp (talk) 21:32, 1 December 2020 (UTC)

re #4, color for alkali metals: the dark color has issues too - also with accessability. (font contrast). So it's 6 of one, half a dozen of the other. For stability, there is no gain in flip-flopping (reverted). Improvement can be easily achieved with some considerations. -DePiep (talk) 23:38, 1 December 2020 (UTC)

PT image in lede
I've updated the image to bring it into line with recently agreed changes. I haven't read EdChem's comments yet. I've flagged Cn, At, Ts, and Og, via underlining their symbols and adding a note about what this means. Since M used the PT to make predictions, and since we mention the predictive capacity of the PT about ten times in our article, I presume these flags and its accompany note are OK. Sandbh (talk) 01:48, 2 December 2020 (UTC)

@Sandbh: Looks OK to me. ^_^ Double sharp (talk) 02:17, 2 December 2020 (UTC)

@EdChem and Double sharp: I've 1. added group and period labels, and 2. reduced the size of the atomic numbers by ½ a point. I suspect the footnote addresses #3. For #4, I can easily change the the colour of the AM back to the darker red. On #5 I believe the image addresses this. I'll experiment some more, later. I'll see if the group names could be added to the white space between the s- and p-blocks. I'll post a prototype here.

I'd also like to lighten the shade for the other metals. Sandbh (talk) 00:44, 3 December 2020 (UTC)

I like most of the changes – the labelling of groups and periods is better, as are the font sizes of atomic numbers – though maybe another 0.5 point? I do notice that:

• There appears to be a white line at the bottom of each period 2 element.
• I wonder if stars would be better than daggers to point to notes.
• I think some copyediting might be helpful. For example:
  • Single Star (or a or or ) on Group 3, Lanthanides, and Actinides: * The placement and display of lanthanides and actinides is not the same on all periodic tables due to differing scientific views on the composition of Group 3.
  • Double Star (or or or ) on Transition Metals and Group 12: **Elements Zn, Cd, and Hg are sometimes considered as "other metals" rather than "transition metals"; Element Cn displays some properties consistent with being a non-metallic noble liquid
  • Placing a triple star symbol (or or or ) on the halogen and noble gas categories, note *** The "Halogen" and "Noble Gas" descriptors are commonly applied to all members of their respective groups. There is theoretical and some experimental evidence that elements At, Ts, and Og display properties inconsistent with being nonmetals.
  • Maybe a or or (rather than quad star) on other metals noting ****Little is known of the chemical properties of Nh, Fl, Mc, and Lv, but they are predicted to be metallic elements.
• I don't think mention of band gaps or Og's claimed appearance are needed as the former is not information a generalist will comprehend and no one has made enough Og to actually see its appearance.
• I don't think the meaning of the underline is sufficiently clear – perhaps clarify or remove. Maybe also should be underlining La and Ac as the note at present implies they are transition metals?
• On colours, maybe we seek outside ideas – DePiep is correct that accessibility is an important consideration.

Just my thoughts / ideas, open to comments, alternatives, criticisms, ... EdChem (talk) 06:50, 3 December 2020 (UTC)

@EdChem and Double sharp: I reduced Z by another half-point. Removing the white line will take me more work than I have time for just now. I experimented with multiple asterisks; they take up more room. I've changed the footnote markers to be consistent with sequence suggested by the Australian Government Publishing Service Style Manual, since I have that one at hand. I have changed the footnote explanations along the lines you suggested, with some rationalisation to reduce duplication. The underlining connection is more visible know. I haven't said anything about Nh, Fl, Mc, and Lv, since we have no reason to expect that they will be anything other than metals. This is case for the elements from 99 onwards, none of which have been produced in bulk. OTOH, for Cn, At, Ts, and Og, theory suggests they will be something else than their lighter congeners. Sandbh (talk) 03:38, 5 December 2020 (UTC)

@Sandbh: 114 (Fl) may have issues, see Flerovium#Experimental chemistry. Double sharp (talk) 09:33, 5 December 2020 (UTC)

Too Bold

I wish to note that this revamp appears to be way too much using "Bold", without searching for consensus. There seems to have been some voting process on a userpage. There was no fleshed out proposal -- instead, a clarification was only provided after someone asked for it. Also, since the change involved dozens of mainspace edits which were bulldozed over us, this does not look like a true Misplaced Pages process. FWIW, any next time I meet this I might claim like "No Consensus". -DePiep (talk) 19:52, 10 December 2020 (UTC)

@DePiep, EdChem, Double sharp, R8R, and YBG: There are a few other things to consider.

• Bold edits don’t require consensus
• As far as I know, WP operates on the basis of continuous improvement
• If there is a dispute about content that is when consensus may need to be sought
• All the edits since our experiment on continuous cooperative improvement editing started have been by myself, R8R, YGB, and Double sharp
• There have been no car crashes, traffic jams or road rage incidents.

The article probably needs more work. I intend to review it, as can any editor at any time, subject to other interests and priorities. Sandbh (talk) 04:55, 11 December 2020 (UTC)

@DePiep, EdChem, Double sharp, R8R, YBG, and Sandbh:. There are a number of things that could have been better in this process. There could have been (1) discussion at WT:ELEM instead of user space (2) a summary of the changes and pages affected in the WT:ELEM announcement (3) a list of deferred changes that need to happen and (4) a pause between the announcement and the implementation. Nevertheless, there was an announcement and a stated offer to accept BRD. So in the same way we look for continuous cooperative improvement editing, let us also seek to have continuous cooperative interaction improvement. YBG (talk) 22:53, 11 December 2020 (UTC)

@YBG: Both ways are fine by me. Sandbh (talk) 23:33, 11 December 2020 (UTC)

@Sandbh: Not sure which two ways you are talking about. YBG (talk) 02:01, 12 December 2020 (UTC)

@YBG: I mean continuous cooperative improvement editing or (and) continuous cooperative interaction improvement. Choose your poison, so to speak. Sandbh (talk) 02:29, 12 December 2020 (UTC)

Actually, I am fully supportive of both and see no contradiction at all. We can, and should, do both. YBG (talk) 02:53, 12 December 2020 (UTC)

RE @Sandbh: @YBG:

"Bold edits don’t require consensus" true by itself, but then again accept straight reversal or other editors being bold. As wikipedia works, without considering critique beforehand one should expect BRD. When applied as lightly as you suggest & with forseeable effects, this turns easily into bordering or editwarring, and poor content. Note that "discussing" is not mentioned.

"WP operates on the basis of continuous improvement". Yes. How does this excuse non-talking?

"If there is a dispute about content that is when consensus may need to be sought". May? Why not "must"?

"continuous cooperative improvement editing" has not been defined or even described. I only met wp:bold edits, which is not a new approach. AFAIK, WP policies and guidelines are well-descibed and sufficient for handling development. "continuous cooperative improvement editing", whatever process it actually is, is not replacing any policy. No editor it tied to this (unspecified) idea. Also, if it involves changing the way WP:ELEM handles policies like BRD, RS, DUE, there is no base to invoke it.

"no car crashes, traffic jams or road rage incidents." -- Actually, there were and we both were involved.

All in all, I am not convinced that this suggested "new" or "different" approach to editing is that 'cooperative', nor does it work to improve the encyclopedia. Not one bullet is about 'time to talk'. -DePiep (talk) 20:33, 19 December 2020 (UTC)

Splitting the p-block

Am I going mad?

For group 13 as the traditional B-Al-Ga-In-Tl, the trend line for the sum of the 1st three ionisation energies is disturbed by the appearance of the 3d shell in Ga. Thus, the trend line has a goodness of fit value (gf) of "only" 0.54

For group 3 as B-Al-Sc-Y-La, gf is 0.95; as B-Al-Sc-Y-Lu it's 0.94

This prompted me to check the gf values for other group 3 properties as either B-Al-Sc-Y-La or B-Al-Sc-Y-Lu.

The following table includes all the values involved, as well as the values for group 3 as Sc-Y-La-Ac and Sc-Y-Lu-Lr:

                   B-La  B-Lu  B-Tl  | Sc-Ac  Sc-Lr
---------------------------------------------------
Density            0.98  0.94  0.98  | 0.99   0.99
1st IE             0.62  0.62  0.65  | 0.97   0.97
Ionic radius       0.98  0.61  0.84  | 0.99   0.61
3rd IE             0.96  0.95  0.42  | 0.98   0.83
Melting point      0.57  0.42  0.81  | 0.15   0.11
Electron affinity  0.62  0.10  0.08  | 0.85   0.99
Sum of 1st 3 IE    0.95  0.94  0.54  | 0.96   0.96
Electronegativity  0.96  0.94  0.47  | 0.58   0.99
---------------------------------------------------
Average            0.77  0.65  0.60  | 0.83   0.69

The best one is obviously B-Tl with the smallest average. Droog Andrey (talk) 16:22, 6 December 2020 (UTC)

​ Why? Sandbh (talk) 23:07, 6 December 2020 (UTC)

​ Because of secondary periodicity (odd vs. even periods). Droog Andrey (talk) 08:47, 9 December 2020 (UTC)

Secondary periodicity

​Secondary periodicity doesn't show in group 4: C−Pb = 0.69; C−Hf = 0.71. Chemically speaking, groups 1–3 are better shown as Li-Fr; Be-Ra; B-Ac. That's not going to happen; I expect group B-Al will remain over Ga. And He will stay over Ne. So an electronically consistent (EC) table is He over Be, and Lu in group 3. A chemically consistent (CC) table is He over Ne, and B-Al over Sc. Where one goes from there depends on the context. The EC table can make a concession to chemistry and move He over Ne. The CC table can make a concession to electronics and move B-Al- over Ga, in which case one has the popular form. Sandbh (talk) 00:52, 10 December 2020 (UTC)

​ Secondary periodicity does present in Group 4: Ge is more electronegative than Si, atomic radius increases more from C to Si and from Ge to Sn than from Si to Ge and from Sn to Pb, and so on.

B-Al-Sc-Y-La-Ac is neither chemically, nor electronically consistent. It is just comfortable for those who don't understand anything but straight similarities and linear trends. Droog Andrey (talk) 14:49, 12 December 2020 (UTC)

Can someone please provide a definition of secondary periodicity? The term is not used in article space. YBG (talk) 08:23, 14 December 2020 (UTC)

@YBG: Secondary periodicity is the typical difference in properties of elements in even vs odd-numbered periods. It holds in all blocks except s. Even-period elements generally have smaller atomic radius, are more electronegative, prefer lower oxidation states. Odd-period elements differ the opposite way. Not always perfect of course. Therefore, whenever you graph a periodic trend outside the s block, it will rarely show an exact straight line, but rather show a definite zigzag instead.

The term is not already in article space because apparently mostly Russian sources know this generalisation and English ones don't. Perhaps because it seems to be a Russian discovery (Biron for the p block, Chistyakov for the d block). Double sharp (talk) 09:06, 14 December 2020 (UTC)

Chemical *and* physical properties

@Droog Andrey: The trends are based on second degree polynomial curves of best fit, using the property value in question cross referenced against Z.

I've added nine physical properties, and one more chemical property, for a total of 17 properties. The combined goodness of fit values for all 17 properties are: B-Al-Sc-Y-La = 0.71; B-Al-Y-Sc-Lu = 0.615; B-Al-Ga-In-Tl = 0.685

For the relevance of chemical and physical properties please see the quote by Jensen, here.

So there's a case for keeping B-Al- over Ga, if you want to emphasis blocks i.e. p elements in the p block.

The better fit for B-Al-Sc-Y-La v B-Al-Sc-Y-Lu prompted me to check the historical record for periodic tables showing group 3 as B-Al-Sc-Y-La. There are 17 examples listed here.

What was it that these authors, including luminaries such as Rydberg, Seaborg, and Pauling, knew about B-Al-Sc-Y-La-Ac?

Deming (1947, p. 617) categorised Al as a pre-transition metal, along with the alkali and alkaline earth metals. Pauling (1988, pp. 635, 697) refers to the congeners of B being Al-Sc-Y-La. He discusses Ga-In-Tl separately in his chapter on Cu, Zn, Ga and their congeners. He writes, "The metals Zn, Cd, and Hg (group IIb) are also much different from the alkaline earth metals (group II), as are Ga and its congeners (group IIIb) from the elements of group III. Cox (2004, p. 185) refers to the pre-transition metals as those in groups 1 and 2, and Al. Further suggesting B and Al are better placed over Sc, here is a 2019 periodic table (by me) recording oxidation number trends,.

In this vein, Rayner-Canham (2020, pp. 178–181) writes:

“It was Rang in 1893 who seems to have been the first, on the basis of chemical similarity, to place boron and aluminum in Group 3.

Such an assignment seems to have been forgotten until more recent times. Greenwood and Earnshaw have discussed the way in which aluminum can be considered as belonging to Group 3 as much as to Group 13 particularly in its physical properties. Habashi has suggested that there are so many similarities between aluminum and scandium that aluminum’s place in the Periodic Table should actually be shifted to Group 3.

In terms of the electron configuration of the tripositive ions, one would indeed expect that Al (electron configuration, ) would resemble Sc (electron configuration, ) more than Ga (electron configuration, 3d). Also of note, the standard reduction potential for aluminum fits better with those of the Group 3 elements than the Group 13 elements (Table 9.2) — as does its melting point.

In terms of their comparative solution behaviour, aluminum resembles both scandium(III) and gallium(III). For each ion, the free hydrated cation exists only in acidic solution. On addition of hydroxide ion to the respective cation, the hydroxides are produced as gelatinous precipitates. Each of the hydroxides redissolve in excess base to give an anionic hydroxo-complex, …There does seem to be a triangular relationship between these three elements. However, aluminum does more closely resemble scandium rather than gallium in its chemistry. If hydrogen sulfide is bubbled through a solution of the respective cation, scandium ion gives a precipitate of scandium hydroxide, and aluminum ion gives a corresponding precipitate of aluminum hydroxide. By contrast, gallium ion gives a precipitate of gallium(III) sulfide. Also, scandium and aluminum both form carbides, while gallium does not.”

B-Al-Sc-Y-La-Ac = (sp) cores; Lu has another core type (spf); Ga, In & Tl have other core types (spd) and (fspd).”

Based on the preceding evidence, B-Al-Sc-Y-La-Ac is chemically and electronically consistent.

The popular La table represents a case where B-Al- were moved over Ga, on electronic grounds. That said, the n +10 relationship (Rayner-Canham 2020, pp. 173−194) for groups 3 and 13 is still seen. Sandbh (talk) 04:12, 13 December 2020 (UTC) Courtesy ping Double sharp

Sandbh, I am not sure why I am being pinged even as a courtesy here when the conversation is between you and Droog Andrey. I watch this page, so there's no need to ping me unless I am being directly addressed But since I am pinged, I confess I do not see how B-Al-Sc-Y-La-Ac can be seen as electronically consistent considering their electron configurations. Boron and aluminium are sp, that doesn't match the sd configuration of the other four. Neither do B nor Al have low-lying d orbitals in the way Lr with its famous anomalous configuration does, that would save the idea with chemically relevant excited states. Neither does B have an (sp) core, as the noble gas it is over is helium which lacks a filled p shell. As I see the logic presented, one could just as well argue that Ar with an (sp) cores does not belong with Ne with an s core, nor with Kr and Xe with (spd) cores, and hence split apart the noble gases. Considering that that is literally one of the most homogeneous groups in the periodic table, I find myself unconvinced. Double sharp (talk) 07:21, 13 December 2020 (UTC)

Double sharp, that was not well written be me. Sorry. B-Al-Sc-Y-La-Ac are electronically consistent in that each has an underlying noble gas core. As discussed by G&E (p. 222), for Ga and In it is NG + d. For Tl it is NG + 4f5d. "This variation has a substantial influence on the trends in the chemical properties of the group." Sandbh (talk) 05:43, 15 December 2020 (UTC)

Yes, but Sandbh, what other groups in the periodic table have a consistent noble gas core? Look at every p block group from C-Si-Ge-Sn-Pb onwards. At the third element, d is added to the core. At the fifth element, f is added on top of that. Look at every d block group from Ti-Zr-Hf onwards. At the third element, f is added to the core. Why then should the B-Al- and the Sc-Y- groups be exceptions? It simply breaks the global trend across the periodic table and breaks electronic consistency. Not my OR argument, Jensen used it for the d block. Double sharp (talk) 05:48, 15 December 2020 (UTC)

I'll let DA answer for himself. But since I agree with him, I'll give also my reason: because we expect a contraction to wreck a linear trend everywhere else in the table. Just look at the "kinked" p-block EN trends for example. Therefore a linear trend is "wrong" here and we should expect a kinked one: so a lower goodness-of-fit value means we are closer to what the trend should look like.

The "kinked" trend of Al-Ga-In is just like that for Si-Ge-Sn, P-As-Sb, S-Se-Te, Cl-Br-I. All is clear. Linear trends, as measured by best-fit lines, are a fiction that frankly only exist in groups 1 and 2 due to weird preemptive filling of s orbitals as explained by Ostrovsky. I think the more global understanding of the RS (e.g. Jensen) is better. Double sharp (talk) 02:59, 7 December 2020 (UTC)

• Question: Are these correlation coefficients of properties vs. period number? Double sharp, Sandbh, Droog Andrey? EdChem (talk) 12:15, 7 December 2020 (UTC)
  • @EdChem:, I think Sandbh is best placed to answer that question. I assumed it was either properties vs. period number or properties vs. atomic number, myself, as it'd have made sense that way. Double sharp (talk) 12:18, 7 December 2020 (UTC)
    • If they are against period number (as in the diagram above) then either I am missing something important or the analysis is statistically flawed. EdChem (talk) 12:26, 7 December 2020 (UTC)

@EdChem and Double sharp: They're properties v Z. Sandbh (talk) 00:51, 8 December 2020 (UTC)

• Sandbh, I have some concerns that I would like to discuss.
  1. Firstly, these "goodness of fit" values are correlation coefficients for you fitting a degree 2 polynomial to the data for property vs. Z, correct? If so, I wonder why is the use of a model for continuous data being used for categorical data? In no way is Z meaningful for any non-integer value, yet you are modelling with continuous curves against Z? Are the goodness of fit values statistically justified? Would other statistical measures be more appropriate?
  2. On wat basis is the choice of a degree 2 polynomial justified as the model chosen? Why is a linear model or an alternative degree polynomial rejected / discarded / whatever? Note, obviously there must be an integer N such that a polynomial of degree N or higher will have an R = 1 as the polynomial will pass through every data point... only polynomials of degree less than N are relevant.
  3. On what basis is the choice of model a polynomial rather than some other function (exponential, logarithmic, etc)?
  4. At times, in looking at your comments, I have the impression that you are presenting your view / thoughts / opinions / conclusions that come from your own analysis and then looking for historical work / publications, etc, that support that position. Is this a reasonable impression to form and does it reflect your approach? In considering material to include in article space, what are you doing to avoid confirmation bias in looking for supporting materials and to ensure that materials selected to support a position do not need to be synthesised together to provide that support?
  5. I ask because there are some strategies that are totally appropriate for the primary research literature that don't suit an encyclopaedia. For example, the changes here to our periodic table article have Al shown in group 3 and the caption of the illustration presents an argument that supports La over Lu in group 3, but neither of these is, for me, really supported by the references shown. My impression is that it more reflects what you believe / want to say than what the sources presented support... am I being fair? The introductory comment on teaching reflects exactly what you have written in your referenced article, though it is unsupported by references here or there. Now, it is reasonable to say that contrasting the alkali metals and halogens is an approach used in teaching. However, the context in which it is made is to begin a section on analogous categories that suggests expanding that approach to contrast pairs like most non-metals with transition metals or Sn and Bi with metalloids... and I don't see the literature support to justify not only including what is stated, but also for WP's voice to be used to imply what is included here. My own experience as a researcher, for whatever it might be worth, involved mostly organometallic compounds of one of the noble metals and I don't see an analogy between it and the noble gases is apt. My experience is OR for deciding what to include, of course, but speaking as an educator I am not inclined to adopt the approach that you suggest.
• I am pinging all regular and / or recent contributors to this page – YBG, DePiep, Double sharp, R8R, Droog Andrey, ComplexRational – in search of a breadth of perspectives and thoughts. Please, can we focus comments on editing rather than on editors, and bear in mind that discussion of differences of opinion / perspective can be valuable if mutually respectful, open-minded, grounded in the extant knowledge of chemistry, and collegial / professional? Thank you. EdChem (talk) 06:18, 13 December 2020 (UTC)
  • @EdChem:, I think a polynomial is difficult to justify in the first place, because of the secondary periodicity of Evgeny Biron that Droog Andrey mentions. I also do share Droog Andrey's impression that Sandbh's analysis here is not paying enough attention to patterns and general trends. But I also don't think that that's the main issue for WP. WP is intended to reflect what the literature is rather than what we feel it ought to be; if we all agree to work on reflecting the former rather than the latter, then there would be no issue. Of course, sometimes it can be difficult when the literature is not united; solving this problem of how to describe the situation for an encyclopaedia like WP is a delicate matter that requires some discussion. But it's not the same as determining which view one personally supports and considers scientifically correct and writing in favour of it in an academic article. What EdChem remarks in 4 and 5 is what makes me worried that the two have been conflated. Double sharp (talk) 06:57, 13 December 2020 (UTC)

@EdChem: Thanks for your interest.

1. Yes, the goodness of fit value is the coefficient of determination for a second order polynomial curve of Z v the value of the property in question. I don't understand your reference to continuous data. For example, if I have five data points for the values of 1st IE for B-Al-Sc-Y-La, I'm looking to see how smooth the trend line is running down the group. Greenwood & Earnshaw (p. 223) plot trend lines for the 1st, 2nd, 3rd, and sum of 1st to 3rd ionization energies for groups 3 and group 13 v Z. The smoothness of the curve for the sum of the 1st to 3rd ionization energies for B-Al-Sc-Y-La is remarkable compared to the choppy trend line for B-Al-Ga-In-Tl. I used the same approach for six charts in my article on the location and composition of group 3. As noted in that article, a comparison of ionic data by Atkins et al. concluded that Sc-Y-La is preferred over Sc-Y-Lu. Their comparison is expressed in the context that ionic radii generally increase down a group. See Fig. 4 in my article, including the 0.99 v 0.61 goodness of fit values.

2. Polynomial curves consistently gave the best goodness of fit values. There was no requirement to go to a higher degree.

3. As per above.

4. I doubt there is that much new about the periodic table that has not already been documented somewhere in RS. Yes, I can see you may have formed the impression that I form an opinion then look for evidence in the literature to support it. In this case, prompted by G&E's plot, I found some evidence supporting B-Al over Sc, and it turned out this has a long history in the literature. I had already read in Misplaced Pages that Al was more similar to Sc, than Ga, but had not thought more of it at that time.

When I edit, my practice is to provided citations for anything controversial. I don't consciously worry about confirmation bias or WP:SYNTH since so much of descriptive chemistry is just a collection of facts.

5. I posted those changes to our periodic table article since I saw how clean the article structure looked, subsequent to our trial of continuous cooperative improvement editing, and noticed the periodic trends and patterns section could be improved with some discussion about isodiagonality and oxidation number trends. In this context I don't understand Double sharp's comment about not paying enough attention to patterns and general trends. Those changes had nothing to do with my advocacy for La in group 3, in a chemistry based table. I also advocate that Lu in group 3, and He over Be, is an excellent electronic based table. More important than either of these is that the context for the choice of particular periodic table is explained. It so happens that the La table happens to be currently more popular in the corpus of chemistry.

Aluminium in group 3 is convenient way to show the diagonal relationship between Be and Al and between Al and Ti, as has been reported in the literature. The Inorganic chemist's periodic table, for example, shows Al over Sc, with a diagonal relationship. Rayner Canham's corresponding article on isodiagonality doi:10.1007/s10698-011-9108-y has a PT with Sc over Al, showing the same kinds of relationships, with a diagonal line running through Li-Mg-Al-Ti-Nb-W. I see no controversy here.

Pairing. The remainder of the caption is non-controversial; as you say, there is no controversy with the comment that, "A traditional aspect of teaching the periodic table is to contrast the alkali metals with the halogens." I discussed the pairing approach, including its didactic implications, at length in my article in the Journal of Chemical Education, on organising the metals and metalloids, as cited.

There is nothing new in pairing. It is seen at the highest level in the metals and nonmetals. Literally thousands of authors have written about the associated L-R progression in metallic to nonmetallic character going across the PT e.g., "The elements change from active metals to less active metals, to metalloids, to moderately active nonmetals, to very active nonmetals, and to a noble gas." Welcher (2001). Or consider this quote :

"The horizontal rows are called series. A series begins with an inert gas, the next member of it is strongly electropositive, the next less strongly so, the next is but weakly electropositive, and the next quite indifferent in this respect. Then follow a very weak electronegative, a more strongly electronegative, and finally an element with intense electronegative qualities. The next element is another inert gas and so begins the next series." (Lemon 1946, p. 308)

The pairing approach article (2020) has so far been downloaded 6,500+ times, and cited twice, including by Rayner-Canham, in his article, "Why don't we really teach about the periodic table?".

PS: For the post-transition metals and metalloids there is Dingle (2017, p. 101): "…with 'no-doubt' metals on the far left of the table, and 'no-doubt' non-metals on the far right…the gap between the two extremes is bridged first by the poor metals, and then by the metalloids – which, perhaps by the same token, might collectively be renamed the 'poor non-metals'." Sandbh (talk) 09:32, 14 December 2020 (UTC)

Noble gases and metals. For the analogy between the two I gave a citation from Wiberg. The synthesis in 1962 of the first noble gas compound, of approximate composition XePtF₆, involved a noble metal. For another example, xenon, in the +8 oxidation state, forms a pale yellow explosive oxide, XeO₄, while osmium forms a yellow strong oxidizing oxide OsO₄. There are parallels in the formulas of the oxyfluorides: XeO₂F₄ and OsO₂F₄, and XeO₃F₂ and OsO₃F₂ (Rayner-Canham and Overton 2010, p. 205). Noble metals and the noble gases each have biological applications:

• Brooks R. R. Noble metals and biological systems: Their role in medicine, mineral exploration, and the environment. CRC Press, Boca Raton, 1992
• Medici S.; Peana M.F.; Zoroddu M.A. Noble metals in pharmaceuticals: Applications and limitations. In: M. Rai; A. Ingle; S. Medici (eds). Biomedical Applications of Metals. Springer: Cham, Switzerland
• Winkler, D.A.; Thornton, A; Farjot G.; Katz, I. The diverse biological properties of the chemically inert noble gases. Pharmacology & Therapeutics. 2016, 160, 44-64
• Winkler D. A.; Katz, I.; Farjot, G; Warden, A. C.; Thornton, A. W. Decoding the rich biological properties of noble Gases: How well can we predict noble gas binding to diverse proteins? ChemMedChem. 2018, 13, 1931.

--- Sandbh (talk) 04:37, 14 December 2020 (UTC)

Physical evidence

​​I compared the trend lines for B-Al-Sc-Y-La; B-Al-Sc-Y-Lu; and B-Al-Ga-In-Tl, for the following properties: 1. density; 2. electrical resistivity; 3. enthalpy of atomisation; 4. enthalpy of fusion; 5. enthalpy of vaporisation; 6. standard molar entropy; 7. heat conductivity; 8. specific heat capacity; and 9. surface tension at the melting point. These properties were used by Horovitz and Sârbu C 2005, "Characterisation and classification of lanthanides by multivariate-analysis methods", Journal of Chemical Education, vol. 82 no. 3, pp. 473–483, doi:10.1021/ed082p473. They also used the Gibbs energy of formation of the chloride LnCl₃ per mol of Cl, as a characteristic of chemical reactivity. I added this to my chemical property data set and removed density and MP, which explains the updated chemical values below.

The average goodness of fit values, alongside the chemical values were:

         Physical     Chemical  Combined
B-La     0.63         0.88      0.755
B-Lu     0.49         0.81      0.65
B-Tl     0.80         0.62      0.71

*1st update to accommodate nine physical and nine chemical properties; and double checking of property values, and charts, and corrections thereto. More work still to be done here. Sandbh (talk) 05:16, 16 December 2020 (UTC)

Physically, it seems like the d electrons in Sc-Y-La etc exert sufficient influence, compared to the sp electrons in B-Al, to upset the smoothness of physical trends. So the B-Tl trend has a better fit.

Chemically, it seems like the consistency of the underling (sp) core in B-La makes for better trends lines, compared to the irregular underlying cores in Lu (spf) and Ga-Tl (spd) and (fspd) Sandbh (talk) 02:30, 10 December 2020 (UTC)

@Sandbh: The reason this argument doesn't convince me is because most groups in the PT don't have consistent underlying cores. The 3d and 4f contractions are well-known effects in the literature (cf. Greenwood & Earnshaw), and the secondary-periodicity effects that DA mentions are well-known even if not under that name (why Br is so much harder to get in the +7 state than Cl or I, for example). When you proceed Si-Ge (p), P-As (p), S-Se (p), Cl-Br (p), Ar-Kr (p), you always add an underlying d to the core, so it seems natural to ask that Al-Ga (p) do the same. (Basically Jensen's argument applied to the p instead of the d block.) Therefore, just because B-La by itself gives a better trend line locally, does not clearly dictate that it should be the standard format, because B-Tl + Sc-Lu gives trend lines that are globally more consistent with other groups. And that is likely the reason why B-Tl is standard. Of course, to rigorise this argument requires clearly defining what exactly a block is, in order to make it clear why we should expect the p trend to follow the others. (Without defining what a block is, you cannot answer why they shouldn't be split.) But since most natural definitions that have been thrown around for blocks are based on physics and electronic structure rather than chemistry, they are surely not going to favour B-La. Double sharp (talk) 02:38, 10 December 2020 (UTC)

@Double sharp: Yes, B-Al over Sc is not going to happen. Neither can I see He over Be happening. I suppose the situation is at the broad contour level, where the broad contours are the four idealised blocks. Where you go from there depends on the context, and as long as that is explained, you'll be good. For example, moving He over Ne is one chemistry-based step. Moving La-Ac into group 3 is another. Moving B-Al over Sc, old school style, is another. The current popularity of the La form has many historical vectors. Will Lu topple La? I don't know. I tend to doubt it, unless IUPAC adopts this form. Sandbh (talk) 03:11, 10 December 2020 (UTC)

​BTW check out this old version of our scandium article, position in periodic table section. Sandbh (talk) 03:19, 10 December 2020 (UTC)

@Sandbh: I tend to think that Lu will eventually topple La, the reason being the quote from Jones you like to give.

From the perspective of those primarily interested in classifying by the final chemistry observed, the Sc-Y-La system is not less than useful, so they won't bother with reexamining it. However, if Sc-Y-Lu had been standard already, it would also not be less than useful. All the elements here are fairly chemically similar, and Y-Lu is not chemically worse than Y-La. It just prioritises different chemical relationships. This rather neatly explains why the general feeling among chemists wrt this issue seems to be not so much supporting one side or the other but total indifference. So if Sc-Y-Lu become standard, chemists of this view would probably just shrug their shoulders and use it anyway, typical of the way in which Be-Mg-Ca and B-Al-Ga got acquiesced to. Because to them the issue is not of any importance.

However, from the perspective of those primarily interested in going from physics, Sc-Y-La is quite definitely less than useful, and Sc-Y-Lu is needed for consistency. For them, Sc-Y-La must be scrapped and replaced by Sc-Y-Lu.

So you have a situation where physicists and physically minded chemists will push for Lu endlessly, and less physically minded chemists won't care. Ergo, the push for will be greater than the push back, which suggests an eventual Lu victory. That seems borne out by the steady fall in the percentage of textbooks that show a La table. The fact that the IUPAC project got established after fifty years of the fight being around is, in my view, a manifestation of this: fifty years were not enough for Lu supporters to win, but they gave them enough representation and made the dispute visible to the public. It is easy to find different books and different posters with different representations of group 3. And since most people who care about this are going to be Lu supporters for the above reasons, plus Scerri's comments, I find it highly likely that Lu toppling La is something that will eventually happen, because more will be pushing that way than the other. Extrapolation of the figures in textbooks suggests that it should happen around midcentury: of course, if the IUPAC project comes to a Lu conclusion, then that would speed it up quite a bit. (The interesting question might be whether Lu topples La before element 121 is discovered. I suspect that might happen. If not, it would be interesting to see if anyone actually applies the block-start criterion for the La table that I see as flawed in order to delay the g block to start at element 125. I highly doubt it to be honest.)

Once that happens, helium becomes the only exception from the four blocks that everyone starts from. (They are "idealised" only from a chemical perspective, I guess; from a physical perspective they are quite real.) So if we look forward in time to a period when He-Ne + Sc-Y-Lu is standard, I can actually see arguments for He-Be starting to become more mainstream. They are already appearing now in significant journals from noble gas chemists (Grochala, Grandinetti), which is already more recognition than the self-published book of Henry Bent. Extrapolation naturally suggests that, especially as helium chemistry takes off, suddenly He-Ne will seem more and more like an anomaly to more chemists. But that is a far more fraught extrapolation. Double sharp (talk) 03:32, 10 December 2020 (UTC)

@Double sharp: This perfectly ties into one of the main arguments behind my support for Lu-Lr. We see with known elements that ground state electron configurations (mostly) shape the blocks, yet there is a single exception to the pattern in the case of La and Ac. Scerri himself says that we don't make an exception for Th to delay the onset of the actinides, and starting the g-block at 125 seems very imprudent. Not to mention relativistic effects and the fact that if/when 125 is discovered, it will certainly be of more interests to physicists. In other words, if every exception to the Aufbau rule or deviation due to relativistic effects were accommodated, the periodic table would be a jumbled mess, and elements would no longer be in a meaningful order of atomic number or grouped by similar chemical properties and valence electron configurations. (And this would likely confuse everyone ranging from high school chemistry students to theoretical physicists.) As I see it, splitting group 13 not only disregards the issues at the root of the group 3 debate, but justifies breaking trends and whimsically making exceptions. I know this is not a proposal for change on WP, but I very strongly oppose it on these grounds. ComplexRational (talk) 16:47, 12 December 2020 (UTC)

@ComplexRational: Thank you! Double sharp (talk) 07:25, 13 December 2020 (UTC)

What's important

@ComplexRational and Double sharp: Regardless of the form of the periodic table, the most important consideration is to explain the context for the chosen form.

• Splitting the s-block, so that He is over Ne; and splitting the p-block so that B-Al is over Sc (as was done historically, pre the quantum revolution, and even after by e.g. Pauling); and splitting the d-block, is chemically congruous.
• From there, an electronic concession can be made by placing B-Al over Sc, as is the case with the currently popular form.
• From there:
  • One can make a concession to chemistry, and show a 15-column wide f-block as per the IUPAC form; or
  • Another concession can be made to electronics by showing Lu in group 3, as is the case with the "popular" Lu form.
• Finally, one can go the whole electronic hog and move He over Be.

As Scerri says, there is no optimal form of table: doi:10.1098/rsta.2019.0300. Each of the above options have their place. None is worse or better than any other one. The only confusion that arises is caused by not explaining the context for the different forms.

The popular form is an accident of history combined with acceptance by the chemistry community whereby the noble gases where all placed into group 0 because of their seeming inertness; La made it into group 3 because it was discovered before Lu, and when it was discovered that the f-shell closed at Yb rather than Lu, nothing changed about the chemistry involved, so La stayed where it was; B-Al were moved out of group 3 and into group 13, on electronic grounds (never mind the damned chemistry). This is all easily explained, as are the four less popular variations.

Here at WP we can't do anything per se, about the fact the La form is currently the most popular. We can however write about the contextual relevance of this form, and at least the other four variations. Then we could stop "arguing", so to speak, and focus yet more on improving articles. --- Sandbh (talk)

Conclusions

Based on these eight properties:

• Group 3 is best shown as B-Al-Sc-Y-La.
• If B-Al remain in group 13, then group 3 is more cohesive as Sc-Y-La-Ac.
• Seemingly the focus on lining up groups into pure blocks detracts from the regularity of periodic trends.

Pauling, in his book, General Chemistry (1988) has group 3 as B-Al-Sc-Y-La in his PT showing EN values (p. 182). Elsewhere he refers to the congeners of B as being Al-Sc-Y-La. He discusses Ga-In-Tl separately in his chapter on Cu, Zn, Ga and their congeners. He writes, "The metals Zn, Cd, and Hg (group IIb) are also much different from the AEM (group II), as are Ga and its congeners (group IIIb) from the elements of group III." (p. 697)

Goodness! Not only do we have a split s-block, and a split d-block, now we have a split p-block! --- Sandbh (talk) 06:49, 4 December 2020 (UTC)

Sandbh It's nice to see this being discussed. ^_^

What you have shown is exactly why, as you know, I am so against La in the d block. Because as you have just demonstrated, the arguments that lead to cleaving group 3 away from group 4 also succeed marvellously at cleaving group 13 away from group 14. Of course I was also saying that in the previous discussion, so it pleases me to see that you're seeing this now. Indeed, see the quote you used at Misplaced Pages talk:WikiProject Elements/Archive 49#Split d-block from R. J. P. Williams:

So a split between groups 13 and 14 is just as present as one between groups 3 and 4. Behaviour of Al is also closer mostly to Sc than to Ga, according to Rayner-Canham. Metallicity drops clearly going Al to Si, Ga to Ge, even In to α-Sn. The +3 state of Al-Ga-In has significant ionic tendencies that are not present for the +4 state of Si-Ge-Sn. And in fact, because Cn-Nh-Fl breaks the 234 pattern in all likelihood, whereas Ra-Ac-Th does not, in fact the 234 pattern also points to B-Al-Sc-Y-La-Ac as group 3. Isodiagonality also works better with Al atop Sc, as shown in your own article on group 3.

I also point out that at Names for sets of chemical elements there is "earth metals" that unites groups 3 and 13, and that preserves left-to-right order better with B-Al-Sc: B, Al, Sc and Ga, Y and In, etc. A +3 d-block contraction is highlighted better with B-Al-Sc, because Ti is the first tripositive cation with a d electron and Ga is the last. Shchukarev's imitator trend, if we believe in Sc-Y-La, looks more consistent with a Ti-Ga d block: then we start each half with +4 (Ti, ), dip down to +2 (Mn, Zn) before restoring +3 (Fe, Ga) at the last element; just like respectively Ce/Tb (and indeed again Tb is much less good at this than Ce!), Eu/Yb, Gd/Lu.

I hope you can now see why I was always saying that Sc-Y-La alone, without something like B-Al-Sc, is a double standard. Maybe I would be less harsh about it now, but I'd still say you can't have one without the other: the same sort of arguments imply both. (I took some of your ten arguments above and applied them to B-Al-Sc!)

Well, you seem to be seeing it as "since I believe Sc-Y-La, I have to believe B-Al-Sc". And that's consistent! Only, I think you'll find some more problems once you look at C-Si-Ti-Zr-Ce-Th. That trend is linear and obeys a 345 rule, whereas C-Si-Ge-Sn-Pb-Fl shows kinks at every block insertion and fails 345 at Nh-Fl-Mc (likely 3-2-3).

But I see it a little bit differently. I say "I cannot believe B-Al-Sc". After all, the vast majority of chemists show B-Al-Ga. Contradiction. Therefore, the logic goes: "since Sc-Y-La implies B-Al-Sc, and that's wrong, Sc-Y-La must also be wrong". Because I have no confidence in an argument that fails to give the expected answer in the non-controversial cases. If it can fail for group 13, then why should I believe it will not fail for group 3 too? I am not so sure of my own chemical knowledge as to think I have the right answer here when almost no chemist sees group 13 as an unsettled question anymore. Does anyone think so other than Habashi, really? Granted, Sc-Y-La is the more common form there, so either way a significant number of chemists are wrong. But Sc-Y-Lu is gaining ground, it has already a significant minority: there's hardly any minority arguing for B-Al-Sc anymore. Even for He-Be there is at least a minority of multiple people, a significant number of whom are noble gas chemists. And I can far more readily believe that some but not all chemists are wrong for the not-so-common rare earth elements, than that practically all chemists are wrong for ultra-common aluminium; especially when a majority of chemists who closely examines the group 3 question comes out in favour of Lu.

And then everything is clear: it's simply that after a contraction, things change. B and Al are normal; then Ga has a kink in the trend, because 3d has been inserted. We should not expect it to continue the trend. We know Ga comes after the 3d contraction, because it cannot use 3d anymore: Zn is the last element that can do that. That's exactly the same as the kink that happens going from C to Si to Ge; from N to P to As; and so on. In the exact same way: Sc and Y are normal; then Lu has a kink in the trend, because 4f has been inserted. We should not expect it to continue the trend. And we know Lu comes after the 4f contraction, because it cannot use 4f anymore: Yb is the last element that can do that. Just like the kink that happens going from Ti to Zr to Hf; from V to Nb to Ta; and so on. As Rayner-Canham noted in the article you emailed me, in fact the nice linear trends you see in the alkali and alkaline earth metals are not the norm at all in the periodic table. So we shouldn't ask for a goodness of fit to a linear trend, when we're not expecting a linear trend. It'd be like taking points perfectly spaced on the parabola y = x, and saying that there's no relationship between the variables because you can't fit a line. Of course you can't, the trend is clearly quadratic in that case.

If you look only at chemistry, you'll never decide between B-Al-Sc and B-Al-Ga; they both look plausible. If you like linear trends too much, you may even go for B-Al-Sc. It's just like Sc-Y-La vs Sc-Y-Lu. That's why chemistry alone is inconclusive and you have to drill down to the electronic structure. As Scerri noted, chemical and physical properties will never provide a decisive answer, and you've got to go for something more fundamental. And that is electronic structure which has the bonus of explaining chemical and physical properties. That's why I go for strict physics and strict blocks alone. And even in your article on group 3, you say that "From a Platonic symmetry perspective or perhaps that of physics ... it can be argued that lutetium is better placed under yttrium" in the conclusion (my emphasis). That should explain to you briefly why I am a strong partisan for the Lu side: I have rejected chemistry as a criterion as inconclusive and decided on physics, following the philosophy of Scerri: 'Scerri thinks a focus on chemical or physical properties is misguided. He compares it to early botanists’ classification of flowers by their color or petal number. “You’ve got to go for something fundamental,” Scerri says, like electronic configuration. “Just to amass properties is never going to give you a definitive answer.”' And that's why the table I am convinced is theoretically correct has not a single split block: even helium takes its place in the s block above beryllium. Of course, just because I'm convinced by it does not necessarily mean it's the right answer, just that I find it the most convincing. But we'll see what the IUPAC project says.

It is alright if you are not convinced right now. After all, it would be very hypocritical of me to criticise you for not being immediately convinced, when I was vacillating about the group 3 issue for quite a while after DA talked to us in the middle of 2018, and only finally decisively rejected Sc-Y-La in late 2019 / early 2020 and never looked back. And of course there's no reason this should impact what we show on WP for now when we haven't heard from the IUPAC project yet. After all: I think He-Be is definitely the correct form, but I don't advocate it for the mainspace because it's so obviously not the majority view. So in this thread we're just having a friendly chat, bringing to the table interesting points. But it really pleases me to see that you're now applying your argument outside the bounds of group 3, both because it means I feel you're no longer artificially limiting your arguments, and also because it means you're now examining a case where the standard classification doesn't seem less than useful to any significant number of people rather than deciding it's not worth losing sleep about. That is exactly what took me on the first step to decisively rejecting Sc-Y-La: using B-Al-Sc as a test case. And the beauty of this is that this is science: if the right evidence comes in, I will reevaluate my stance. As you may know, if some evidence came in suggesting that actually Lu had some categorical 4f valence involvement and it wasn't just buried and giving incomplete shielding like it is for Hf, I'd certainly move to thinking that Sc-Y-La isn't that bad after all. It's a contest of ideas only, not personalities, and a nice diversion while we take a break from describing what is in the literature, and dream of what perhaps should be – and in my case, what I hope will soon be. Before returning to earth of course once we saunter out into the mainspace. ^_^

I'll probably not have time to reply at this length after today, so I'll let you think about it. I also feel that this is the sort of thing that works better if you think about it by yourself for a while: that's how it worked for me. And so I think it's best if this is my only long reply about this topic to not let it get out of hand. However you choose to see it, it will for you hopefully be a journey of scientific inquiry and examination. And I feel it is always good to go down such a journey away from the WP mainspace sometimes. Whatever your conclusions are, I have no doubt they will be interesting, even if it's away from the WP mainspace. Double sharp (talk) 17:44, 4 December 2020 (UTC)

Just to get it: what would the split be, like? -DePiep (talk) 00:13, 6 December 2020 (UTC)

I don't recall having seen a split p-block in any of my studies or work in chemistry, so I would need a LOT of RS to be persuaded that any such split being shown widely on WP would be justifiable on DUE grounds. Given the reference to Pauling, there may be a place for a mention somewhere. Double sharp's comments on how the idea led to a firm view on La v. Lu etc, if they reflect discussion in RS, could be of relevance for main space somewhere as well. This thread includes much OR, it appears to me, but there are parts that could be article-space relevant if sufficient sources exist. EdChem (talk) 00:26, 6 December 2020 (UTC)

@EdChem: Yes, that's why I ended off by saying that it was interesting to talk about but not for the mainspace. Only source I'm aware of that mentions analogy of situations of B-Al-Sc-Y vs B-Al-Ga-In to Sc-Y-La-Ac vs Sc-Y-Lu-Lr and uses it to argue for the latter situation is a 2004 Russian article by Droog Andrey: I think it is cited under Periodic table#Primogenic symmetry for something else already. Might be worth a tiny mention, but not a do-or-die thing for me as it is a minor thing. Double sharp (talk) 08:08, 6 December 2020 (UTC)

What it looks like

There it is, a PT with split s-, p- and d-blocks.

I cottoned on to this as a result of reading Greenwood & Earnshaw's discussion of the chemistry of Al-Ga. I recall I was prompted to do so by something DS wrote, which I've now forgotten. Anyway, G&E have a graph (p. 223) showing trends in successive ionisation energies of B-Al-Sc-Y-La v B-Al-Ga-In-Tl. They write:

"It is notable that these irregularities for the Group 13 elements do not occur for the Group 3 elements Sc, Y, and La, which shows a steady decrease in IE from B to Al, all 5 elements having the same type of underlying core (noble gas). This has a decisive influence on the comparative chemistry of the two subgroups…the steady decrease in EN in the series B > Al > Sc > Y > La > Ac is reversed in Group 13 and there is a steady increase in EN from Al to Tl." (pp. 222, 224)."

A similar pattern is seen in standard electrode potentials (p. 225).

B-Al-Sc-Y-La tables

• Rang (1893)
• Gooch & Walker (1905)
• Cuthbertson & Metcalfe (1907)
• Baur (1911)
• Rydberg (1913)
• Black & Conant (1920)
• Lewis (1923)
• Hubbard (1924)
• Deming's table (1925), which popularised the medium-long form
• Antropoff (1926)
• LeRoy's table (1927)
• Irwin (1939)
• Seaborg (1945), with Al in group 3 and in group 13
• Yost & Russell (1946)
• Coryell (1952)
• Pauling's table (1960)
• Habishi's Metallurgist's Periodic Table (1992), as mentioned by DS (Habishi leaves B in group 13).

I suspect what has happened here is that it was historically known that group 3 was better represented as B-Al-Sc-Y-La-. Then, with the advent and rise of modern electronic structure theory, B-Al- got moved to the p-block because, after all, they were p-block elements, never mind the damned chemistry. And La stayed in the d-block since it was the first element to show 5d electron, and 4f did not show until Ce. DS and I have discussed the latter phenomenon at length.

Nowadays, this represents forgotten chemistry. It underscores the fact that electron configurations are not the final arbiters of periodic trends.

@Double sharp, Droog Andrey, and EdChem: I found the smoking gun. Here's an extract from Mellor's modern inorganic chemistry (1943 impression of 1939 edition, p. 677):

"21 Relationships of the elements of Group III. The exact sub-classification to be adopted in this group has in the past occasioned some difficulty, but it is now generally agreed that B and Al are best associated with Ga, In and Tl. This is supported by the conclusions at present accepted for the electronic configurations of these elements…"

So B-Al- over Ga-In-Tl is an electronic decision albeit B-Al- over Sc-Y-La is a better fit in terms of chemical and physical periodic trends going down group 3. Fortunately, these trends can still be seen via the n + 10 relationship, as discussed by Rayner-Canham (2020), for group 3 and group 13. Sandbh (talk) 00:09, 11 December 2020 (UTC)

B-Al-Sc-Y-La-Ac as a linking or bridging group

It's odd to see how group 3 consistently represents a linking group. So B-Al link the s-block with the bulk of the p-block; Sc-Y link to the bulk of the d-block; and La-Ac link to the f-block.

On Sc-Y-La-Ac, Greenwood & Earnshaw write:

"All are rather soft-slivery-white metals, and the display the gradation in properties that might be expected for elements immediately following the strongly-electroposive AEM and preceding the transition metals proper."

B is between Be (a metal) and C (a nonmetal). G&E note B has many similarities with C (and Si). At the same time, Greenwood (2001) commented:

"The extent to which metallic elements mimic boron (in having fewer electrons than orbitals available for bonding) has been a fruitful cohering concept in the development of metalloborane chemistry…Indeed, metals have been referred to as "honorary boron atoms" or even as "flexiboron atoms". The converse of this relationship is clearly also valid…"

Al is between Mg, a metal and Si, a metalloid. It has some properties that are unusual for a metal; its crystalline structure shows some evidence of directional bonding. Aluminium bonds covalently in most compounds. The oxide Al₂O₃ is amphoteric and a conditional glass-former. Stott (1956) labels aluminium as a weak metal. It has the physical properties of a metal but some of the chemical properties of a nonmetal. Steele (1966) notes the paradoxical chemical behaviour of aluminium: "It resembles a weak metal in its amphoteric oxide and in the covalent character of many of its compounds…Yet it is a highly electropositive metal… a high negative electrode potential". Moody (1991) says, "aluminium is on the 'diagonal borderland' between metals and non-metals in the chemical sense."

--- Sandbh (talk) 06:31, 6 December 2020 (UTC)

EdChem

So, if I am following correctly:

• Habishi argued in 1994 for moving Al but not B to group 3
• Greenwood and Earnshaw comment on the similarity of metals to B (also noted by Habishi) and note irregularities in IE, while retaining B and Al in group 13 and so not actually presenting a split p-block
• Moody's comment from 1991 that Al is on the "'diagonal borderland' between metals and non-metals" sounds to me to be consistent with placing B and Al in group 13 alongside all other such elements
• And, you have no RS literature in support of a split p-block with B and Al in group 3 that was published in the last 60 years?

Arguing that much older work had such placements tells us nothing about the present RS consensus, and nor does referring to it as "forgotten chemistry." Do you have any significant body of recent RS that actually uses a split p-block? If not, then what you are discussing may be useful for our articles on the history of the PT but it doesn't offer any reason to alter the presentation of the PT as it is currently constructed by RS and used in practice. EdChem (talk) 08:18, 6 December 2020 (UTC)

@EdChem:​ Yes, you're following correctly. It's interesting to talk about and not for the mainspace other than in a historical sense. I refer to it as forgotten chemistry since modern texts do not discuss the idea of a split p-block. It's peculiar that modern periodic tables seem to be electronic block-based tables, rather than chemistry-based tables in which B-Al are over Sc. Even Seaborg, who arguably contributed to the dominance of the electronic table, showed Al in two places ie group 3 and 13. Looking at the list of tables with B-Al over Sc, there are some other relatively big hitters, too: Rydberg; Lewis; Hubbard; Deming; Antropoff. What was it that these luminaries knew about B-Al-Sc-Y-La-Ac that is deemed to be no longer relevant, and why is that the case?

I recall seeing split-p block tables from time to time but have never paid much attention to them. The most recent one I have at hand is in the corrected 1988 edition of Pauling's General Chemistry.

Deming (1947, Fundamental Chemistry, 2nd ed. p. 617) located Al with the pre-transition metals in groups 1−2. Cox (2004, Inorganic Chemistry, 2nd ed. p. 185) refers to the pre-transition metals as those in groups 1 and 2, and Al. Here's a 2019 periodic table (by me), recording oxidation number trends, further suggesting B and Al are better placed over Sc.

Rayner-Canham (2020, The periodic table: Past, present, and future, pp. 178–181) writes:

"It was Rang in 1893 who seems to have been the first, on the basis of chemical similarity, to place boron and aluminum in Group 3.

In terms of their comparative solution behavior, aluminum resembles both scandium(III) and gallium(III). For each ion, the free hydrated cation exists only in acidic solution. On addition of hydroxide ion to the respective cation, the hydroxides are produced as gelatinous precipitates. Each of the hydroxides redissolve in excess base to give an anionic hydroxo-complex, …There does seem to be a triangular relationship between these three elements. However, aluminum does more closely resemble scandium rather than gallium in its chemistry. If hydrogen sulfide is bubbled through a solution of the respective cation, scandium ion gives a precipitate of scandium hydroxide, and aluminum ion gives a corresponding precipitate of aluminum hydroxide. By contrast, gallium ion gives a precipitate of gallium(III) sulfide. Also, scandium and aluminum both form carbides, while gallium does not."

As a chemist, what is your take on the merits of B-Al-Sc-Y-La-Ac in comparison to B-Al-Ga-In-Tl? Sandbh (talk) 22:46, 6 December 2020 (UTC)

As a chemist, I think the periodic table has an unsplit p-block and B and Al are p-block elements. As a Wikipedian, I think the discussion is wandering into WP:NOTAFORUM territory. As a researcher (and engaging in a little OR), I think that the correlations (if I am not mistaken in how they are constructed) have serious problems with validity in a statistical sense. EdChem (talk) 12:22, 7 December 2020 (UTC)

How the B-Al-Sc vs. B-Al-Ga situation relates to the Sc-Y-La vs. Sc-Y-Lu situation

@EdChem: As noted above the source I am aware of that mentions this is an old article by Droog Andrey (= Andrey Kulsha). Link is here. I grant it is not formally published but seems to be something he did as a student at Belarusian State University in 2004. But he has gotten published outside as an academic, here (some computational chemistry about superacids) and here (in 1999 as a student, but already in the journal Хiмiя: праблемы выкладання 1999, 5, pp. 102–109; this is his article supporting Lu in group 3). Hopefully this is good enough as an RS for WP; if you feel it's not, I will listen to that.

Translation courtesy of Google Translate: "It should be noted that until recently, not La - Yb and Ac - No, but Ce - Lu and Th - Lr were attributed to the families of f-elements, while the d-elements of group IIIb were La and Ac instead of Lu and Lr from for the passage of the first (n - 2) f-electron to the (n - 1) d-subshell . However, a detailed study of the regularities of changes in the chemical and physical properties of elements in the series Ba - Hf and Ra - Rf , as well as a comparative study of the Sc-Y-La-Ac and Sc-Y-Lu-Lr groups and drawing analogies with the B-Al-Sc-Y and B-Al-Ga-In groups, respectively, showed that it is more expedient, more practical, and more logical to attribute the families La - Yb and Ac - No to f-elements . At present, the version of the periodic system with "shifted" families is still widespread , but it is gradually being replaced by a new form, formally more rigorous and better describing the real properties of f-elements and d-elements of IIIb subgroup."

If I used it, it would in any case only be for a sentence in the main group 3 element article, along these lines: "Kulsha pointed out that the changes going down Sc-Y-La-Ac and Sc-Y-Lu-Lr are respectively analogous to those going down B-Al-Sc-Y and B-Al-Ga-In; since B-Al-Ga-In is standard, this constitutes an argument for Sc-Y-Lu-Lr." In any case there are very many other arguments for Lu in the literature there, so even if this is considered not RS enough, it's still OK with me. Again, it's just for a descriptive sentence, in what's basically a subarticle and not the main periodic table article, not about changing the periodic table format.

I'm not arguing that we change to -Lu-Lr right now on WP anymore. Oh, yes, I still think it's the absolutely correct form and all that, but I also know that's not the point for WP. Please correct me if I'm wrong about the outside situation, but it seems the situation there among chemists and on WP is this:

• The majority of chemists don't care and if anything wonder what all the fuss is about.
• A minority thinks the issue is of great and paramount importance.
• When this minority writes articles about the matter, a crushing majority of them argues for -Lu-Lr for various reasons. (Link is to all the sources I've managed to collect, or at least have mentions of, in the literature. If you know somewhere where I can find the ones that don't have links, please write on my talk page!) There are a few who care greatly about this and who think the situation supports -La-Ac (Sandbh's view), but it is very few. So: as a general rule, those who care about this issue think -Lu-Lr is the scientifically correct, or at least a scientifically better, form.
• However, -La-Ac was the traditional form for a long time. Therefore, among those who wonder what all the fuss is about, that form predominates because it's what they learnt and they don't think there are any strong reasons to change it.
• Therefore the effect is that most textbooks continue to use the -La-Ac form.
• However, some textbook writers have been convinced by the -Lu-Lr partisans to change their periodic tables. So thanks to them, the amount of support for -La-Ac has eroded noticeably, but it still has a majority.
• The debate also seems to be responsible for the -*-** compromise of IUPAC gaining ground, but that is not as common either. See this 1988 report by IUPAC, in which Sc-Y-Lu was admitted to be the correct form according to electronic configurations (with some sources being given to justify this), but Sc-Y-* was chosen as a compromise and appeared in the next Red Book because Sc-Y-La dominated in the literature. Although inconsistently: in the next 1990 Red Book, the periodic table was shown in 18 column format as Sc-Y-*, and then on the next page in 32 column format as Sc-Y-Lu. I don't know if this was just a mistake or a deliberate attempt to please everybody; if so it clearly didn't work. The 32 column form was dropped from the 2005 Red Book anyway, so for now IUPAC shows Sc-Y-*, but does not recommend it. Strange, I know. Judging by the fact that if anything the rate of publications on this argument increased rather than decreased after 1988, it seems that this ended up being the classic sort of compromise that satisfied nobody that was promptly ignored in favour of continuing the fight.
• Therefore the average reader is likely to be familiar with all forms because it's easy to find textbooks on both sides (La-Ac: Greenwood & Earnshaw; Lu-Lr: Clayden et al.) and even popular-science books on both sides (La-Ac: The Cartoon Guide to Chemistry, Gonick & Criddle; Lu-Lr: The Periodic Table: Its Story and Its Significance, Scerri), but -La-Ac is what he or she will find most of the time when reading off WP at his or her likely level if he or she is looking up basic stuff about the PT anyway.
• Ergo, the conflict is between which sources to listen to. Should we go for the sources people are more familiar with, in which -La-Ac predominates with -Lu-Lr as just a rising upstart in the minority? Or should we go for the sources that directly take on the matter, in which -Lu-Lr crushingly predominates, except for the small problem that those sources are written at a level quite a few levels up from the basics that the periodic table is at?
• The issue has been raised a couple of times this year by me, and it is somewhat split; but the viewpoint in favour of going for the sources directly about the thing did not get a clear majority. So I have to admit there is currently no consensus for a change. Both viewpoints seem legitimate to me by the measure of "it's not against WP policy", and I do still think WP is worse off for not showing -Lu-Lr, but I think that pales before the need to respect consensus. Without that, WP can no longer remain WP. So I hold my peace.
• The issue has been going on since 1965 and apparently enough students and instructors are confused by the differences in group 3 found in the literature that IUPAC started a project to deal with it in 2015. Its chair's words here make me hopeful that -Lu-Lr will be the ruling. If that happens, the chance of getting a consensus change will be far greater, and then I will propose it again. But not before that and not if it doesn't happen. After all: at any time it is possible that they rule something that isn't -Lu-Lr, and then suddenly -Lu-Lr has even less of a position by WP policy than before. So there isn't any point in raising it again now until and unless a -Lu-Lr ruling happens soon: not much will have changed.

So, to make it clear, I'm not looking to change anything about how we present group 3 now. (I mention this here because I'm not sure I've said all this in one place yet.) It has been fairly tried and gotten no consensus, and although I still prefer -Lu-Lr as everyone knows by now, WP has to work by consensus and I respect that. Not only that, a IUPAC ruling is coming that with a bit of luck could end the entire debate: since we usually wait for IUPAC approval in any case when new elements get discovered (the only common way the PT changes at this moment), it makes sense to use that as a precedent here and wait for it too.

It is only that Sandbh's mentioning of the group 13 situation made me remember that this argument exists in the literature, and wondered if it would be permissible by WP standards to give it a one-sentence mention in the article group 3 element that discusses the arguments (of course not in the main periodic table article for which it would be overkill). That's all I'm wondering. Double sharp (talk) 11:09, 6 December 2020 (UTC)

@Double sharp and Droog Andrey: DA's paper is less than reliable.

The key sentence is:

"However, detailed study of the regularities of changes in the chemical and physical properties of elements in series Ba - Hf and Ra - Rf , as well as a comparative study of groups Sc-Y-La-Ac and Sc-Y-Lu-Lr and drawing analogies with the groups B-Al-Sc-Y and B-Al-Ga-In, respectively, showed that it is more expedient, more practical, and more logical to attribute to f-elements precisely the families La - Yb andAc - No .

No summary is provided as to "the regularities of changes in the chemical and physical properties of elements in series Ba - Hf and Ra - Rf ."

References 5−7 are:

5. Ugai Ya.A. General and inorganic chemistry. - M .: Higher school, 1998

6. Physical quantities: Handbook. - M .: Energoatomizdat, 1991.

7. Kulsha A.V. f-Elements in the periodic system of D.I.Mendeleev // Chemistry: problems laid out 1999 , No. 5, pp. 102 - 10

I expect these references would be hard to check.

Reference 3 i.e. Winter M. WebElements Periodic table. URL: http://www.webelements.com, is not a RS.

Reference 8 is Jensen. Scerri wrote that Jensen was too selective in the evidence he put forward in order to support his case.

The real-life situation seems to be as follows:

1. An electronic (physics-based) table would show (a) He over Be; and (b) group 3 as Sc-Y-Lu-Lr; and (c) group 13 as B-Al-Ga-In-Tl
2. A chemistry-based table would show (d) He over Ne; and (e) B-Al over Sc-Y-La-Ac.

What we have instead is a hybrid table with 1(c) and 2(d).

Is that a fair summary? --- Sandbh (talk) 23:42, 6 December 2020 (UTC)

I am asking EdChem, actually. I am not sure why WebElements is not an RS considering that we actually use it in chemical elements data references. Source #7 you mention is exactly where the "regularities of changes" are mentioned: it is his 1999 article that I already linked you. The selectivity of Jensen seems to be only a charge that Scerri levels: it did not seem to stop many people from citing and referring to him as the authority for the Sc-Y-Lu table, and Scerri himself also supports Sc-Y-Lu for different reasons.

I also do not think you have a fair summary of the chemistry-based situation. (The physics-based one is correct, with maybe some quibbles about He-Be since He chemistry is not too well-known yet.) The grounds that you accept B-Al-Sc for (mostly historical and chemical) are exactly those that have in the same era been used to support Be-Mg-Zn, according to RS. As Jensen noted, Be and Mg do share more in common with Zn than with Ca from a chemical perspective. Jensen explained the situation in that paper:

And indeed, if you look at modern texts, you will find that indeed those similarities exist and are widely acknowledged. Greenwood & Earnshaw on p. 113 says it: "In these properties , as in many others, the heavier alkaline earth metals resemble the alkali metals rather than Mg (which has many similarities to Zn) or Be (which is analogous to Al)." What's changed is that nobody thinks it has significance for PT placement anymore. Of course, you can't exactly have Be-Mg-Zn and B-Al-Sc in the same periodic table. It would break the order of increasing atomic number because B would have to come to the left of Be. That doesn't look good for your argument: if we insist that chemical similarity is the first criterion, we end up with both Be-Mg-Zn and B-Al-Sc and hence break the periodic law!

Jensen has clearly pointed out here that this is simply coming from the change between chemical and physical understandings of the periodic table. B-Al-Sc and Be-Mg-Zn both come from this same era (of course never in the same book because it breaks the atomic number order). And in both cases, we can still find significant relationships to the other one: Al-Ga is well-known, Mg-Ca is also well-known. Look at any chemistry textbook that describes main group elements, some of their similarities will be mentioned. Over 50 or 60 years ago, because physics and electronic periodic tables weren't the norm or widely understood yet, people were using chemistry as the sole guide. But modern periodic tables are based on physics and electronic structure, not chemistry. That's also why, for example, hydrogen is placed where it is over lithium, and not over fluorine. (Helium over neon is arguable because one can think of 1s vs 2s2p as the core, in which case electronic structure is not against He-Ne: I expect this will change as helium chemistry gets revealed, because that is what proves that those are not just core electrons. Which is fascinatingly consistent with how a significant number of He-Be partisans seem to be noble gas chemists: Grochala, Grandinetti.) As Scerri himself noted, chemical arguments are not conclusive: they are rather the equivalent of classifying flowers by colour or petal number rather than genetics. (Jensen citing Jorgensen also pointed out that the path from electronic structure to final chemical behaviour is not always that straight, especially for the d and f block elements with multiple low-lying excited states that contribute for chemistry. Seems to me just like the winding road from genotype to phenotype.) And as Jensen also noted, if you just pick and choose elements from different rows and columns arbitrarily, then thanks to the possibility of trends cancelling each other you can prove pretty much anything, which is no good. As he said, you have to get the self-consistent periodic trends from the atomic structure. That's why his latest publicly expressed stand (2017) seems to be that the most important argument for -Lu-Lr is because La and Ac have low-lying f orbitals that they use for chemistry, but Lu and Lr don't: in other words, the physical argument is coming first for him too. That is why B-Al-Ga is standard, that is why Be-Mg-Ca is standard. Elements are placed in blocks according to their electronic structure, even if some other elements may be chemically more similar. And although I still can't find a source making the point I am about to make, it boggles my mind exactly how chemical properties could still be thought of as a valid way to build up the periodic table today, when everyone is familiar with nitrogen and bismuth in the same group. You could hardly come up with two more different elements. (If you object that N and Bi share the maximum oxidation state +5, simply make it nitrogen and moscovium.) In short: what you have, talking about chemistry, is widely agreed by authorities to be a historical view: everything I said above that isn't in small print is from the RS. My impression concurs with EdChem's. Other than a few people on the side, nobody thinks these added chemical similarities actually mean the periodic table should change anymore. Be-Mg-Zn and B-Al-Sc are interesting sidelines in the history of the periodic table, connected with the trouble encountered when you expand Mendeleev's original 8 column format to the modern 18 and 32 column ones; but they are only historical sidelines. As could be expected because Be and Mg are more like Zn than Ca according to RS, and B and Al are more like Sc than Ga according to RS, so if you adopted chemical similarity as the criterion for both you'd end up with the atomic number sequence getting lost because B came before Be!)

I might also add that if one really believed in chemistry as the basis, then Th, Pa, and U would not be under Ce, Pr, and Nd, but under Hf, Ta, and W. That is where they were before WWII as well, because the valences match better that way, and the physics was not then well-understood. Nd cannot be hexavalent; U can and often is.

Of course, nobody believes that anymore: the physical fact that the f orbitals are involved here has trumped all chemical ideas here. No doubt to a traditional chemist of Mendeleev's time, the placement of U under Nd would look deranged: now it has conquered all. Chemistry has lost utterly to physics here. I suspect this paper of Scerri is of relevance here, given its header "The Actinides. A Case Where Spectroscopy Triumphs over Chemistry?" In any case, even if one tried to use chemistry here: one could not. Th, Pa, and U would be jammed in the same spaces as Rf, Db, and Sg which cannot be placed anywhere else but as eka-Hf, eka-Ta, and eka-W. Just another example of how chemistry simply doesn't work as a basis for PT placement: it forces you to break the law that elements appear in order of increasing atomic number. It has been abandoned for a very good reason.

I don't think the future is going to be particularly good for Sc-Y-La. We seem to be agreed that physics is not for it; whether chemistry is for it or not is somewhat disputed (Jensen thinks no, Restrepo thinks yes – but only for La, since data for Ac/Lr is thin on the ground), but chemistry has already been supplanted as a basis for the PT by physics long ago. So the dispute has almost become moot because according to the modern criterion of physics, Sc-Y-Lu is a shoo-in. After all, using chemistry as the basis means wrecking the atomic number order (B before Be, Al before Mg) and forcing two elements in the same place (U and Sg jammed together under W). Scerri has outright said in his recent workshop: "My money's on number two and I certainly will be arguing to IUPAC that they should adopt number two. ... The one I'm suggesting and other people on this committee have suggested is lutetium and lawrencium should be the official version." I doubt he'd say that if he didn't think the IUPAC committee he's chairing was on board with it, particularly since he's apparently writing an article summarising the results of their deliberations. So, I grant that I am an outsider to these deliberations, but I would be very, very surprised after hearing that if the results were not an IUPAC ruling in favour of -Lu-Lr. So I'd personally advise once again that you drop that idea and return to Sc-Y-Lu as you supported before 2016, before I to my regret convinced you otherwise based on wrong data (in fact, what I told you were the condensed-phase configurations of the Ln and An, that convinced you to -La-Ac, were not the correct such configurations; my bad). RS indicate that for the wider chemical community, periodic tables based on chemistry rather than electronic structure are a historical thing. We've advanced since then, and the reason why -Lu-Lr which is consistent with this idea hasn't gotten up seems to be just inertia or people unaware of the matter. It seems to be going as DA said: "Eventually Lu will replace La. That's just a matter of time needed for obsolete understandings to die." But it's up to you, of course. You can have whatever view you like, and so can I; the point is just to work together to let WP accurately display what's in the literature, and update things only with it. As I said, I'm totally OK with WP only getting updated to the -Lu-Lr format when an IUPAC ruling comes along. Double sharp (talk) 00:38, 7 December 2020 (UTC)

Double sharp, if you are going to leave aside any action on group 3 until there is movement from IUPAC, then why put a lot of time into discussion of the topic, especially in light of WP:NOTAFORUM? EdChem (talk) 12:31, 7 December 2020 (UTC)

@EdChem: Well, I did it mostly because it seemed to me that Sandbh was using partly the same sort of argumentation as he was for Sc-Y-La as he's now using for B-Al-Sc. In this case the RS situation is a lot clearer (to a first approximation no one seems to take B-Al-Sc seriously), so I thought this might serve well at explaining to Sandbh why I feel RS was not supporting his approach. Maybe it skirts to NOTAFORUM, but my judgement might be a bit compromised by a lingering guilt: before 2016 Sandbh was a Lu supporter and I convinced him away from it based on what I now with better knowledge think was wrong data and some bad arguments. Since in 2018 DA explained to us more about this and I realised that what I was saying wasn't really correct, but Sandbh wasn't convinced, I find it a bit difficult to suppress the lingering guilt and it might impact what I say here more than it really should. Anyway, it seems clear now from his reply that it didn't really work at explaining to Sandbh why I feel this use of correlation to look for a linear trend isn't justified per Jensen. So I decided after my second post today (see the bottom) to stop posting about it here. In other words: I agree with you now, and had just decided to stop. ^_^

Anyway, the point of me starting this subthread was just to ask you if you thought DA's article could serve as an RS for including the argument "B-Al-Sc vs B-Al-Ga is analogous to Sc-Y-La vs Sc-Y-Lu, and since B-Al-Ga is standard that makes Sc-Y-Lu seem more reasonable by analogy" in the ancillary article on group 3 element (not the main periodic table article): it was a source reliability for WP question. Unfortunately I guess it is now buried. I'd still like an opinion but it's not really that important; there are lots of other arguments which have already been listed in that article and are cleanly cited to RS, so having or not having this one isn't that important.

I am afraid to say I have some doubts now about whether trying to continue the collaboration to improve periodic table to real FA status is really going to be good for us as a project. That's because of recent edits by Sandbh coupled with this discussion. The two I have in mind are this and this. Please correct me if I am wrong but I have serious doubts if these two trends (category analogies and isodiagonality respectively) are that common in the introductory chemistry literature to deserve this kind of high-profile mention. They are things that Sandbh has discussed in his recent journal-published articles of 2020 (this year), and he cites himself for them, but I was not aware of them being common trends to discuss among chemists rather than just appearing in a few places (isodiagonality is common for a few specific pairs Li-Mg, Be-Al, B-Si, I know, but I'm not aware of it being commonly applied to things like Ca-Y-Ce as Sandbh gives in his illustration). Again, please correct me if I'm wrong and these actually are common to discuss in the ways the figures show, but as it stands I'm not really comfortable with these additions. I'm especially not happy with the figure for isodiagonality, which he took from his own article on group 3, has aluminium moved to over scandium (as he's discussing here) against common chemical practice, contains an uncited claim about "predominantly ionic" chemistry that to my understanding is against typical freshman knowledge and not really often used to describe chemistry (surely that always depends on the counteranion; sure, NaCl is predominantly ionic, but one can hardly say it for Na₃Bi, say; surely one can't exactly say the chemistry of Ce is "predominantly ionic" without making it clear what elements we're comparing it against, because with most elements Ce probably forms some sort of intermetallic phase anyway; and isn't the more common way to express this electropositivity?), and finally at the end throws in what seems to be a gratuitous Sc-Y-La argument from the atomic number triad (21+57)/2 = 39 that isn't exactly relevant to this topic (isodiagonality) IMHO. I cannot say I'm happy with this, but perhaps my opinion is coloured by the fact that I'm generally against Sandbh's research conclusions on periodicity, so your opinion might help as a third one. But in general I feel that (1) we just have very different visions for how the article should look, and (2) it doesn't look very good if I'm always the one who pounces and points out something I feel is problematic when it's the same editor each time, and (3) it seems likely that in the absence of a categorical statement of Sc-Y-Lu from IUPAC this and other issues will be hard to really provide a ruling on because this La vs Lu issue keeps sneaking in in the form of these sorts of periodic table figures and illustrations. I'd be fine with it if it was just the default PTs being shown as La, without any text trying to "prove" it correct that way when the debate is not of importance; but when we have a PT shown like that, and specially moved to have Al atop Sc in something else that the editor in question is championing, with a justification for Sc-Y-La presented in the middle of a long caption in a section not related to the dispute, it somehow leaves a bad taste in my mouth as if the argument was being given too much weight and it spilled outside its own section. So I suspect it might be better for me to just leave this article be and let Sandbh do as he wants with it, when there are a bunch of much less controversial elements in the middle of the table for which good reliable sources are not hard to come by and for which categorisation is completely undisputed (I admit that probably means just some nondescript transition metals...). After all, if what he's done is really problematic, it's better for someone less involved in the dispute to say it, and if it isn't, then I'd better also hold my peace; and I think writing this neutrally is going to be so so so much easier once IUPAC says something and I don't trust myself to be able to do it perfectly either till then. For instance, I guess the GAs on vanadium and palladium ought to somehow be restored by an article improvement drive. Most of the remaining not-even-GA articles are very meaty (e.g. gold, sulfur) or hard-to-research-because-historical (radium) topics, but taking a GA somewhere and FA-ing it might be fun. All that might very well be far less controversial than trying to achieve a shared vision for the periodic table article as the situation stands... Double sharp (talk) 13:04, 7 December 2020 (UTC)

Double sharp, I mentioned WP:NOTAFORUM to both you and Sandbh as there are some pretty long posts appearing and they are not clearly directed to article-space issues. It was meant as a reminder and an invitation for reflection, so please don't take it as more than was intended.

You asked if I thought that DA's article could serve as an RS for including the argument "B-Al-Sc vs B-Al-Ga is analogous to Sc-Y-La vs Sc-Y-Lu, and since B-Al-Ga is standard that makes Sc-Y-Lu seem more reasonable by analogy" in the ancillary article on group 3 element (not the main periodic table article). Leaving aside reliability, a more pertinent question is whether DA's article makes that argument, or if it has been used by other RS to make that argument... because if the argument is built from DA's article but the article does not actually make the argument directly then it would be OR or SYNTH to use it unless another RS has already advanced the argument using DA's work.

Sandbh citing his own work does not raise OR or SYNTH concerns so long as the work cited does support what it is used to support. That doesn't mean that its inclusion is appropriate based on DUE, however. You link to two additions and I can see why you have concerns based on a first look. Starting a talk page discussion and neutrally inviting comment from the Chemistry WikiProject might be a way forward. I tried to encourage work on the PT article to look at the non-controversial parts that can be improved, but it does not appear that my approach is developing along the lines that I envisaged. EdChem (talk) 13:52, 7 December 2020 (UTC)

@EdChem: Right, I understand what you intended now. ^_^

Well, DA's article says (relevant bit in translation) However, a detailed study of the regularities of changes in the chemical and physical properties of elements in the series Ba - Hf and Ra - Rf , as well as a comparative study of the Sc-Y-La-Ac and Sc-Y-Lu-Lr groups and drawing analogies with the B-Al-Sc-Y and B-Al-Ga-In groups, respectively, showed that it is more expedient, more practical, and more logical to attribute the families La - Yb and Ac - No to f-elements . I guess it's pretty clear what he means by it, but it's maybe expressed in a way that requires the leader to draw the line to the final step. Do you think it'd be SYNTH to say what I said? I'm not sure, that's why I'm asking. ^_^

Oh yes, I agree that Sandbh citing his own work is totally OK wrt OR and SYNTH. It is cited to an RS and it supports what it claims to. My concern was just about DUE, as mentioned. But I think I'll let you opine first, because as I mentioned I may have some unconscious biases. Double sharp (talk) 15:11, 7 December 2020 (UTC)

Group 2 as Be-Mg-Ca-Sr-Ba-Ra or Be-Mg-Zn-Cd-Hg

I mapped the trend-lines going down these two options using the same eight properties I used for group 3 and 13.

The average goodness-of-fit values were:

Be-Mg-Ca-Sr-Ba = 0.84

Be-Mg-Zn-Cd-Hg = 0.66

This is better than the figures for group 3 and group 13, which I did not find surprising. It is consistent with 16 of the 17 B-Al-Sc-Y-La-Ac tables. Only Hubbard shows Be-Mg-Zn-Cd-Hg. It is consistent with C & W (1999): "The Group 12 elements differ markedly from those in Group 2 in nearly all aspects except having II as their only important oxidation state." (p. 599).

I haven't double checked my figures, so maybe I got something wrong.

The quality of DA's source 7 is too low to be legible.

I agree many people cite and refer to Jensen as the authority for the Sc-Y-Lu table. The only people I know who have looked critically at Jensen rather than superficially, as I did at first, are Scerri, myself and you. And I only did so with the help of Scerri. In our joint IUPAC paper we showed how nearly all (as I recall) of Jensen's arguments don't hold water.

I feel there is too much confidence being placed in physics, as opposed to chemistry, given the nature of IUPAC. I sort of addressed this is in my article on the location and composition of Group 3:

"Nor does IUPAC prescribe which elements, for example hydrogen and helium, belong in which group. All they have done is approve collective names for like elements e.g. chalcogens (oxygen, selenium, selenium, tellurium, polonium) and noble gases (helium, neon, argon, krypton, xenon, radon) (Connelly et al. 2005, p. 61). For example, if helium is placed over beryllium then helium still belongs to the noble gases.

IUPAC does not recommend or approve any particular format of periodic table or system, nor does it mandate the composition of Groups."

An IUPAC format table leaves open the implications associated with each group 3 option.

It says the chemistry of La to Lu is the primary item of interest, with group 3 being able to be treated flexibly as Sc-Y-La-Ac or Sc-Y-Lu-Lr.

If consistency is also important, you can move B-Al over Sc-Y-La-Ac if the focus is on chemistry ; or, with Sc-Y-Lu-Lr, move He over Be, if the focus is on electronic structure .

So the options are:

                                                                          Basis
                          Group 2  Group 3          Group 13  Group 18  Chem  Elec
 Lu form (consistent)  He-Ra    Sc-Y-Lu-Lr       B-Tl      Ne-Rn
                          Elec     Elec             Elec      Elec       0     4 
---------------------------------------------------------------------------------
 Traditional Lu form   Be-Ra    Sc-Y-Lu-Lr       B-Tl      He-Rn
                          Chem     Elec             Elec      Chem       2     2
---------------------------------------------------------------------------------
 IUPAC form            Be-Ra    Sc-Y-*-**        B-Tl      He-Rn
                          Chem     ?                Elec      Chem       2½    1½ 
---------------------------------------------------------------------------------
 Traditional La form   Be-Ra    Sc-Y-La-Ac       B-Tl      He-Rn
                          Chem     Chem             Elec      Chem       3     1
---------------------------------------------------------------------------------
 La form (consistent)  Be-Ra    B-Al-Sc-Y-La-Ac  Ga-Tl     He-Rn
                          Chem     Chem             Chem      Chem       4     0
---------------------------------------------------------------------------------

I note that:

• the Lu form with He over Be is perfectly unambiguous, and highly symmetrical.
• the traditional Lu form is a fence-sitter in that it cannot make up its mind if it is Arthur or Martha;
• the IUPAC form is not really sitting on the fence i.e. it has a chemistry bias: chem. (2½) elec. (1½);
• the most common, traditional form is inconsistent—not so surprising given the nature of the subject matter i.e. chemistry; and
• the La form is unambiguous, and requires three split blocks.

Above table and notes updated. --- Sandbh (talk) 00:57, 11 December 2020 (UTC)

I know of nothing that would prompt IUPAC to make a decision on the composition of group 3. Recall that the commissioning of the IUPAC project on group 3 does not necessarily mean the recommendation of the project will be accepted by IUPAC.

Of course, I may be be mistaken, and I'm looking forward to what eventually comes out of the IUPAC Group 3 project, and its reception, by the chemistry community, and IUPAC and its world-wide affiliated bodies, as much as you are. --- Sandbh (talk) 06:17, 7 December 2020 (UTC)

@Sandbh: I think you're still asking for goodness-of-fit trend values comparing against straight lines. But the point is that the trend should not be a straight line. If you do that, then it's like staring at exact parabolas and shouting "there's no trend! there's no trend!". Of course there is, but if one insists on the linear dogma, then it becomes simply impossible to see. Just try and plot the goodness-of-fit value for the following data:

The correlation coefficient happens to be exactly zero. But there is a clear relationship between x and y here. An exact one, in fact. I think it's pretty obvious. ^_^

There's two kinds of trends in the table: Be-Mg-Ca-Sr-Ba-Ra is the "straight line" kind, Be-Mg-Zn-Cd-Hg-Cn is the "kinked" block-insertion kind. The latter is natural everywhere outside the s block, which is why for those trends a lower goodness-of-fit value is actually better: it indicates we have the right type of trend. But if you insist on those, then I suggest plotting goodness-of-fit values for C-Si-Ti-Zr-Ce-Th (all four electrons above a noble gas) vs C-Si-Ti-Zr-Hf-Rf (Earth scientist's periodic table) vs C-Si-Ge-Sn-Pb-Fl (standard). The former will probably be better. And the same for N-P-V-Nb-Pr-Pa vs N-P-V-Nb-Ta-Db vs N-P-As-Sb-Bi-Mc. Are those forgotten chemistry also? Maybe in a way, they are the A vs B secondary analogies that Jensen agreed were kind of "lost knowledge" today. But that doesn't mean that a true chemical periodic table "should" really contain C-Si-Ti or N-P-V (unless we talk about the 8 column form still common in Russia, in which case yes, C-Si-Ti is there, but so is C-Si-Ge simultaneously); apart from the Earth scientist's periodic table for C-Si-Ti, nobody thinks either of these belong in the 18 or 32 column forms. Also maybe try Ti-Zr-Hf-Th vs Ti-Zr-Hf-Rf, V-Nb-Ta-Pa vs V-Nb-Ta-Db, and Cr-Mo-W-U vs Cr-Mo-W-Sg. Thanks to the actinide contraction probably the former would look better in best-of-fit values. ;)

Indeed, commissioning a project does not mean the recommendations will be accepted. But doing so would defeat the entire purpose of it, which is that students and instructors are puzzled by the differences. Many IUPAC projects do indeed result in recommendations, even for such basic issues as those that appear on the PT. Such has happened quite recently for the project on naming new elements and the project on discovery criteria for new elements. Well, they've since both progressed to recommendations and have become the new rules. The first was already in effect for why Ts and Og have -ine and -on suffixes; the second is already also in effect, it just hasn't been applied yet because no serious claims for elements 119 and up have come. Don't see any reason why this one will not progress similarly when Scerri already did write in Chemistry International back in 2012 about how IUPAC should be recommending one form of group 3 and indeed exactly the -Lu-Lr one. That's a IUPAC magazine, so IUPAC knows what his stand is. I also suspect IUPAC must have learned its lesson from the 1988 -*-** compromise which pleased nobody. After that debacle, they probably know that if they don't make a ruling, the issue will just end up being kicked down the can for another few decades before flaring up again just like it did this time (1988 compromise, 2015 opening of IUPAC project). Besides, predicted chemical properties meant absolutely nothing to IUPAC when Ts and Og were recommended to be named with -ine and -on; the symmetry of keeping the whole column the same won the day. So in that case chemistry was shot down in favour of symmetry which came from physics (the predicted electron configurations match F-Cl-Br-I-At and Ne-Ar-Kr-Xe-Rn, even if chemical behaviour probably totally doesn't). So if IUPAC was fine going for physics rather than chemistry then, it doesn't seem that implausible that they'd do it again for -Lu-Lr. Again, I'm not saying I know this is how it will work out (I don't); just saying that the situation and what Scerri has said makes it seem to me that we should not be too surprised if the end result is a IUPAC recommendation of -Lu-Lr. Of course that's not the only thing that might happen; I just think it seems likelier than I had imagined previously.

I am indeed in the camp that says that Lu is the only correct form and that the La form is an old historical misunderstanding, as you know. That's a stronger view than your support of La, I see, which seems to account for the possibility of different forms being more or less valid in different contexts. (This I see as secondary relationships, like Jensen.) So let me just state this honestly: I've been trying to fix the consequence of my faulty understanding in 2016. Then I had a faulty understanding, and used what I now know were faulty data and faulty arguments to influence you in User talk:Double sharp/Archive 11, probably you would not be supporting Lu now. If I knew then what I knew now, you'd never have been influenced by me to La, and probably you'd be still supporting Lu unless somebody else made La arguments to you that convinced you. You may recall that I totally withdrew my support for that old IUPAC submission we worked together on and now think that actually Jensen's arguments do hold a great deal of water after all. For me to do that when it went that far says something, probably. If you don't think that's enough for you to reevaluate your position, and you want to continue supporting La, that's your choice and I respect it. So if this post of mine doesn't convince you, then that's alright and my stance returns to "let's wait for IUPAC". Scerri's the chair of that project, he understands quite a bit about these issues, and I am sure he will come up with something interesting and that is worthy of following on WP as a IUPAC recommendation. And then we'll see. Till then, there's no point discussing it further. But you may still go ahead if you like and post interesting things you see here like your B-Al-Sc support, of course. I just probably won't reply, both because it'll go in circles, and because you probably already know what I think and what I see the RS as saying. Basically, my viewpoint is settled enough and tested to my personal satisfaction, and I don't think further discussion of this point is going to change our disagreement: only a IUPAC ruling might. To that end also, I think that if you're not convinced by what I've pointed out here (plotting goodness of fit will probably also support C-Si-Ti which G&E also show), then there's no point in me discussing the PT article until IUPAC comes around, as our visions for it are going to differ too much for a very meaningful collaboration. Maybe I'll just find an element article that is far away from all PT placement questions and also PT categorisation questions. That probably means a nondescript transition metal. Most are GA already, but some could use further improvement pretty easily. So maybe I'll channel whatever free time I have (not too much from tomorrow onwards, but we'll see) into that and wait for IUPAC to say something about group 3 before touching the PT article. There is no deadline after all.

And in any case, speaking personally: I'm already happy that you see my He-Be + Sc-Y-Lu personal view as a "perfectly unambiguous" one. I recognise it's pretty much on the opposite end of the spectrum to your preferred one, so it really pleases me that you see something in it. Double sharp (talk) 09:20, 7 December 2020 (UTC)

Basis for goodness of fit

Double sharp: There are no straight lines, as understand it. The goodness-of-fit trend values are for polynomial lines.

Oh yes, He-Be + Sc-Y-Lu is wonderful in its clarity! As far as IUPAC is concerned, they may well endorse it, on these grounds! Consider what Jensen wrote in JChemEd (2008):

"Above all, such demands should be tempered by the sobering recollection that IUPAC is the organization that brought us density in units of kg/m, 4πε₀ in the denominator of Coulomb's law, and the finger-count labels 1–18 in the periodic table."

If IUPAC introduced the dumbed down 1−18 group numbering system, it would seem to be within the realm of possibility that they could recommended at least Sc-Y-Lu on "simplicity" grounds, and just put up with He over Ne.

It's ironic that Jensen is on the IUPAC Group 3 project and will presumably support Lu on, inter alia, symmetry grounds even though he disparagingly referred to the 1−18 numbering system, and has rallied against the abuse of symmetry considerations when it comes to organising the periodic table! We know he won't support Lu on the grounds of triads per Scerri, since Jensen has disparaged the relevance of triads. Since Scerri referred to Jensen's argument as being too selection, I don't know on what basis Jensen would support Lu. Lavelle is also on the project team, and his support for La and opposition to Lu is well known. And Restrepo most recently supported La.

On students and instructors being puzzled by the differences, that is the fault of IUAPC in not providing sufficient guidance in e.g. the Red Book, along the lines I suggested.

On group 4 v 14, I looked at this, and the average goodness of fit values are about the same. This being so, I'd leave group 14 as C−Pb. Sandbh (talk) 01:39, 8 December 2020 (UTC)

@Sandbh: Over here I'll just answer the question about Jensen as it is a question that can be answered from publicly available RS without editorialising on my part: presumably Jensen would support Lu for the reason he publicly expressed in 2017, which is also the reason I support it: a consideration of their available excited-state configurations (2, 4, 12)... reveals that both La and Ac have low-lying empty f-orbitals and that these are implicated in several low-lying excited configurations, whereas Lu and Lr have no available empty f-orbitals. This implies, in turn, that both La and Ac – like 25% of the other d- and f-block elements (including their nearest neighbor Th) – are best viewed as having anomalous ground-state valence configurations, which, in their case, take the place of an idealized (n-2)fns valence configuration. Since such an interpretation is not possible for Lu/Lr, it pretty much determines that this pair should be assigned to the d-block whereas the La/Ac pair should be assigned to the beginning of the f-block. He cites among other sources 10.1364/JOSA.61.001112. I don't think Scerri referred to this argument when he mentioned that Jensen's arguments were too selective, although I may be wrong. Regarding the other points, I'll write on your talk page. Double sharp (talk) 01:49, 8 December 2020 (UTC)

Jensen's 2017 paper

Double sharp, you've honed in on a key consideration.

As I read it, Jensen's 2017 unpublished and non-peer reviewed paper is a mixed bag:

On the one hand he wrote:

"To tip the scales in favor of one pair of elements or the other requires instead a consideration of the additional chemical and physical properties discussed in my original paper of 1982 and, above all, a consideration of their available excited-state configurations.

This is bad. We showed that nearly all of the chemical and physical properties he referred to in his 1982 paper were inconclusive. Not to mention Scerri's view.

To his credit, Jensen brilliantly goes on to write:

"Though there are many misconceptions concerning the nature and function of the periodic law and table, perhaps the most prevalent among modern chemists is the belief that the periodic table is nothing more than an electron configuration table. While there is certainly a significant correlation between electron configurations and chemical periodicity, the correlation is, as already noted, far from perfect. The increasing prevalence of irregular configurations among the d- and fblock elements, the increasing lack of correlation between minor irregularities in these configurations and actual chemical behavior, and the ever present empirical question of how to properly divide an atom’s configuration into the chemically relevant categories of valence versus core, all require a careful balancing of both chemical and physical evidence rather than an appeal to authority and a naive, and apparently arbitrary, freshman chemistry application of spectroscopic atomic ground states."

Exactly. The PT is not a simple electron configuration block based table. Electron configuration considerations, in isolation, do not tip the balance. Sandbh (talk) 05:59, 8 December 2020 (UTC)

PS: Of course, they can tip the balance if that is the context of interest, in which case you have the beautiful He-Be + Sc-Y-Lu-Lr table. And that is a perfectly fine way to approach it, among others. Sandbh (talk) 06:03, 8 December 2020 (UTC)

WP:NOTAFORUM

Are there any thoughts about this WP policy, given EdChem has mentioned it a few times?

Everything I post here I do so with the aim of sharing information and enriching perspectives, in support of improving articles. Good writers are well-informed writers, as I see it. --- Sandbh (talk) 01:13, 8 December 2020 (UTC)

Please be specific. -DePiep (talk) 01:45, 8 December 2020 (UTC)

The context is these posts by EdChem:

• "As a Wikipedian, I think the discussion is wandering into WP:NOTAFORUM territory."
• "Double sharp, if you are going to leave aside any action on group 3 until there is movement from IUPAC, then why put a lot of time into discussion of the topic, especially in light of WP:NOTAFORUM?"
• "Double sharp, I mentioned WP:NOTAFORUM to both you and Sandbh as there are some pretty long posts appearing and they are not clearly directed to article-space issues. It was meant as a reminder and an invitation for reflection, so please don't take it as more than was intended."

--- Sandbh (talk) 02:20, 8 December 2020 (UTC)

I find that the number and volume of posts makes it difficult for me to concentrate on the forest rather than the trees, and so it distracts me from the main purpose: building article space. I also note that when I see substantial references to primary sources and non-mainstream secondary and tertiary sources, I lose confidence in my ability to distinguish between due and undue WP:WEIGHT. So while I don't think we have violated the letter of WP:NOTAFORUM, I think the project would better served by having fewer and smaller posts. YBG (talk) 16:50, 8 December 2020 (UTC)

@YBG and Sandbh: Well, how about this as a halfway house: we put here whatever is directly describing and discussing what we're planning in the articlespace, and write the sort of thing Sandbh describes to each others' user talk pages? There aren't that many of us, so it shouldn't be hard to find (since mostly the long conversations happen between me and Sandbh), and this talk page becomes easier to follow. Double sharp (talk) 17:08, 8 December 2020 (UTC)

This seems to be a very workable solution. And when a topic starts to gel into something that might directly affect article space in the immediate future, a new thread can be started on this page, with a summary of the user talk thread and a link to the complete discussion. Yes, this might work out very well indeed. YBG (talk) 17:42, 8 December 2020 (UTC)

@YBG: in you post you mention two issues: use of primary/secondary sources and long posts. While reducing post and thread lengths by moving preliminary work to userspace by itself is an improvement indeed, that does not solve the sources-item you mentioned. And if I am correct, that is what EdChem is pointing to by mentioning NOTAFORUM. This issue then will pop up again when a usertalk thread arrives on this page, and so introduces the issue for article space. (IOW, in userspace talk is more free, but when a conclusion/proposal is posted here to go live somehow, the sources-problem you mention is activated). ping @EdChem and Sandbh: -DePiep (talk) 21:10, 8 December 2020 (UTC)

My experience of our talk page comes in four flavours:

1. Someone posts something here that I don't have much interest in so I don't follow it, so much.
2. Someone posts something here that I find interesting, so I join in the discussion.
3. I post something I find interesting to see if anybody else is interested. Nobody else is interested. End of thread. Collapse into a hat note after a short while.
4. I post something I find interesting to see if anybody else is interested. Somebody else is interested. I join in the discussion.

Any of these may impact the article space. Personally, this has worked fine for me. I pay as much attention as I feel I need to.

Primary, secondary, tertiary sources
Here's what our policy says:

"Articles should be based on reliable, published secondary sources and, to a lesser extent, on tertiary sources and primary sources. Secondary or tertiary sources are needed to establish the topic's notability and to avoid novel interpretations of primary sources. All analyses and interpretive or synthetic claims about primary sources must be referenced to a secondary or tertiary source, and must not be an original analysis of the primary-source material by Misplaced Pages editors."

For primary sources, these do sometimes appear in the article space, as e.g. "In 2020 it was reported that".

For secondary sources, there seems to be two kinds: mainstream, and higher-level. Mainstream = e.g. textbooks e.g. those that show polonium as a metalloid. Polonium is normally excluded from metalloids by higher-level texts, and for an element that is hard to study those will be more reliable. This example comes from Double sharp.

Tertiary sources (of which WP is one) seem unremarkable to me, provided you keep your wits about you. They can provide a global summary of what secondary sources say.

The beauty of posting things here is the convenience of exposing it to all of us, and the ease of searching. Since we've started moving some of our discussions to each other's talk pages it's notably harder to find things i.e. I may now need to search my talk page, YBG's talk page, Double sharp's talk page, and this talk page, before I can find whatever it was I was looking for.

The things we've talked about here recently are: our periodic table article; categories; noble metals; group 3; colouring. There are plenty of reliable secondary sources addressing these. All of our discussions are related to associated content appearing in the article space. Where we have got to now seems to provide fair coverage. --- Sandbh (talk) 23:38, 8 December 2020 (UTC)

An informal poll

It seems to me that the general WP philosophy about what types of sources we should pay the greatest attention to is:

• WP:TERTIARY > WP:SECONDARY > WP:PRIMARY > WP:OR

Q1 This general assessment is admittedly overly simplified. But is it completely wrong-headed? Why or why not?

However, my general impression of the volume of material discussed on this page is:

• WP:OR > WP:PRIMARY > WP:SECONDARY > WP:TERTIARY

Q2 This general impression is admittedly overly subjective. But is it completely wrong-headed? Why or why not?

I would appreciate my fellow project editors answering these questions and one of the following two:

Q3 If you agree with me on these points, what is your overall reaction to this unfortunate contrast?

Q4 If you disagree on one or both points, do you have any idea why I might have gotten such wrong-headed ideas?

YBG (talk) 09:14, 14 December 2020 (UTC)

from Double sharp

@YBG: My answers.

Q1. I think you are correct, though I'd rather nuance it to TERTIARY > SECONDARY >> PRIMARY >>> OR. And I think that when a basic topic is being discussed like chemical periodicity, it becomes even more important to follow this order. In this case I would say that tertiary sources (i.e. standard textbooks) are the way to go, with secondary sources only called in when the tertiary sources don't agree or where secondary sources generally agree that the tertiary ones are wrong. Primary sources, being generally reports of new research, are hardly the way to go here, when we're supposed to provide a standard view on the topic: if something is really standard, it should not be hard to find tertiary sources giving that view. And there is always the danger of interpreting primary sources to say something that they do not really say or cherrypicking primary sources, when actually an analysis of tertiary and secondary sources would better indicate to us how common a view expressed in primary sources really is and whether and how we should describe it per WP:DUE. Finally, OR ought to be kept to a minimum, because its only conceivable purposes are (1) fun diversions or (2) announcements that something is soon to appear in a primary source.

Q2. I also think that you are correct.

Q3. I think it is not a good situation. I have contributed to it in the past, and I regret having done so. I would also like it to stop, and have since tried to do so. Double sharp (talk) 10:26, 14 December 2020 (UTC)

@Double sharp: Yes, I agree that my first ordering should be more nuanced, and I generally agree with your reordering. YBG (talk) 05:45, 15 December 2020 (UTC)

from DePiep

• We should also consider related policies like WP:UNDUE and WP:FRINGE. -DePiep (talk) 22:08, 14 December 2020 (UTC)

from Sandbh

@YBG: Some more answers.

Q1. The general WP philosophy is "Articles should be based on reliable, published secondary sources and, to a lesser extent, on tertiary sources and primary sources."

Q2. An example would be helpful.

--- Sandbh (talk) 00:34, 15 December 2020 (UTC)

@Sandbh:

re Q1: Yes, it appears my general impression needs to be a bit more nuanced. See Double sharp's better arrangement listed above.

re Q2: What brought this subject to the forefront was the extended discussions of correlation coefficients elsewhere on this page, which appeared to me to have huge swaths of OR. But this only crystalized my general impression of our discussions here over the years. Double sharp acknowledges above that he has contributed to this trend in the past and I is now seeking to take a different tack.

--- YBG (talk) 05:45, 15 December 2020 (UTC)