User talk:Sandbh/Group 3

R8R comments

I've been invited by Sandbh to provide a comment on what is written in the proposal.

Before I begin, I want to note that despite my long support for -Lu-Lr, I cannot find myself as supportive towards the idea. First of all, the idea itself has become less clearly important to me. Also, quite recently, DS introduced me to some good reasons why -La-Ac is a valid option to at very least discuss.

I read the article prior to being introduced to comment on it. I remember having some general problems with it, but I want to leave that to when I come close to an end of this review. A more thorough read is a good thing to do before saying so anyway.

My sincere intent is not to push for either option, but only to help finalize a non-contradictary set of arguments. As a mathematician and a citizen of the modern media world, I have set importance of logical consistency in an argument high. Apart from inconsistecies within theory, should I find any, I will also mention the inconsistencies in logical reasoning. This will have to do not as much with the science as with the correct presentation. Aforehand, I'll note I won't focus much on what I find to be correct.

I assume the authors claim to aim to provide a review with no aforegiven opinion.

Cover sheet

Fine. One very minor picking is that you introduced Sandbh as a "Wikipedia editor," while DS is a "Wikipedia Editor."

Fixed Sandbh (talk) 00:06, 16 December 2016 (UTC)

Abstract

Finely summarizes the content of this page as it currently is.

Ack. Sandbh (talk) 02:33, 17 December 2016 (UTC)

Lu-Lr arguments

After reading the whole section, I can't help the feeling it was meant to show there is no good argument for -Lu-Lr. As it turns out, there is not a single argument you could simply agree with! Those arguments that were found to be inconclusive are included here (of course). Also an argument that even the proponent found inconclusive (Ionization energy of Lr). Why not include it too among those bad arguments, among which there is not a single good one?

This creates, alongside some missing logical steps described below and eagerness to look at group 2 but only once, for unclear reason dismissively, at group 4, the feeling that the winner was already determined before the analysis.

Ack. Yes, Double sharp and I had discussed aspects of the -La-Ac v Lu-Lr question before we decided to put in a submission. And we had agreed that -La-Ac was our preferred option. I don't believe either of us meant to show there was no good argument for -Lu-Lr. That was just the way things ended up once we'd listed and analysed all the -Lu-Lr arguments we were aware of. However, the fact that it came over that way to you has given me pause for thought and I'll look again at our submission to see if needs to be adjusted to avoid giving this unintended impression. Sandbh (talk) 03:11, 16 December 2016 (UTC)

I did not, implicitly or explicitly, claim the opposite; but I felt obliged to share the feeling.--R8R (talk) 19:28, 16 December 2016 (UTC)

Before this submission I had supported -Lu-Lr, including so as to avoid a split d block: something which I thought was highly objectionable (in hindsight there was no real basis for my discomfort). Upon looking closer at things I switched back to La-Ac, then reverted to Lu-Lr, and have now come back to settle on La-Ac. I have no interest either way. So far, I've found the Lu-Lr arguments to be like a shotgun discharge, with no cohesion whereas the La-Ac arguments, few as they are, are like a closely grouped bunch. (Mind you, it's hard to know if I still have any unconscious bias colouring my thoughts). Sandbh (talk) 03:57, 18 December 2016 (UTC)

I couldn't think you were unbiased; we all humans are. My personal bias is beginning to grow back on -Lu-Lr as I begin to find something I don't like in your reasoning. In my mind, it's currently an interesting idea that could be worth investigating. I don't make any conclusions prior to the investigating (it took me a long time of what they call "intellectual athleticism" to naturally begin not to do so, but now I am confident in saying this), however, and I hope to avoid the need of that.

I'd want to suggest a few ideas for consideration. The biggest problem with the review as it is is that it only deals with the arguments you have seen elsewhere. Even the best possible analysis based on that could provide the following conclusion: "the arguments mentioned in the literature that we are aware of suggest X" and not "X". This is why I suggest you reach beyond what has been said in the literature: see my remarks on groups 4--10 vs. 4--5, for example. The "not the full story" remark is close to this problem.

Comment: We did this for the La-Ac Blocks and Chemical behaviour arguments, which are our own, and we've addressed your comments about 4--10 v. 4--5 (but we're waiting for at least one reply to a ping). I attribute the lack of non-literature based Lu-Lr arguments to the fact that there are so many literature based ones that the well of original ideas has run dry. Sandbh (talk) 09:57, 19 December 2016 (UTC)

Also, I would want you to reflect somehow (if I'm guessing your intentions correctly) that there are multiple possible definitions of a block. The theoretical Seaborgian rectangles, the more chemisty-dependent ones, the ones per filling shells, etc. Only recently did I realize there was some reasoning for the other definitions; you say your case was similar.

Comment: I could be wrong but I think the standard accepted defintions of the s-, p-, d- and f-blocks are so fundamental to the theoretical underpinnings of the periodic table that I wouldn't see any profit in considering alternative definitions, at least not for the purpose of this submission. Sandbh (talk) 09:45, 19 December 2016 (UTC)

I want to read all of your comments, assess them, and reply to them. Unfortunately, I am currently short on time. Possibly in the late December if my comments are still useful then.--R8R (talk) 12:48, 18 December 2016 (UTC)

Comment: Late December will be fine. I recall Double sharp only has access via a mobile phone until 24 Dec, so that works too. As well, there are still a few items to finalise. Sandbh (talk) 21:55, 18 December 2016 (UTC)

I'm going through and adding some clarifications (mostly from memory, because of the current problem that you so concisely and elegantly summarised). Please do not hesitate to correct me if I mess anything up! Double sharp (talk) 06:20, 19 December 2016 (UTC)

Comment: Roger that! Sandbh (talk) 09:45, 19 December 2016 (UTC)

1958: Electron configurations

This critique of the provided argument seems meaningless or at least far too picky to me. As I read the argument, it said La was [Ba] + 5d1, while Lu was [Ba] + 4f14 + 5d1. I geniuinely believe the authors did mean to imply that the overall number of electrons did not play a role on which group to assign to an element.

"The argument that a filled 4f shell in Lu forces its placement in the 5d transition metals is weak given that Yb also has a filled 4f shell" -- given the above, this logic is flawed.

I don't say this is a deciding factor, even though quite strong, but I think the critique at its pickiness is absolutely not.

Done. I looked again at Landau and Lifshitz (web link added in References section) and have updated the analysis. They still don't help. Sandbh (talk) 00:06, 16 December 2016 (UTC)

I still disagree. This is actually a strong argument. It's not a deal-breaker (though it could be for some), but it is strong as an auxiliary fact. All other 5d elements have a filled 4f shell. I would think there has to be a strong reason to overcome this (I have not assessed the relative strength of -La-Ac arguments, and don't plan to do so, just noting that some of them are good). I have no immediate access to the book you quote, but I believe that if you look one step beyond what is written in this page, you'll see more.

As a wider comment, I genuinely do think that you're limiting yourselves to the comments you're aware of and not all possible comments. Taking a look back is important. You mentioned elsewhere you wanted to keep the chronology; I share this feeling. Yet now the chronology is a main player, and it limits the analysis, which is not good. A better format overall would be to describe the chronology to give a historical perspective and then describe the quotes and look beyond them. Or maybe even only mention that somebody ever suggested another fact. I don't see why you cannot develop the thought. Also, you actaully do that for at least one -La-Ac argument: While aspects of this argument were previously considered in our analysis of Lavelle's (2008) "pair out of place" argument, we consider the concept of a block merits further exploration. You're free to do the same here.--R8R (talk) 19:26, 16 December 2016 (UTC)

Comment: Here's the link to the book: Quantum Mechanics (Non-relativistic Theory) Sandbh (talk) 00:43, 17 December 2016 (UTC)

More broadly, I think you're suggesting we should consider the outcome of L&L's argument i.e. that all period 6 d-block elements would then have a filled 4f shell. Jensen (1982) actually discusses this:

The argument that the total (i.e., core plus valence) electronic configurations of lanthanum and actinium are closer to those of lutetium and lawrencium (due to their filled (n – 2) + f14 subshells) is misleading. One must consider intraperiod as well as intragroup analogs. Thus, the remaining nine d-block elements of period 6 (Hf–Hg) all have the complete [Xe] + 4f14 core like lutetium and not just the [Xe] core of lanthanum. Likewise, in passing down the columns of the d block from Ti–Zn one always encounters the addition of the 4f14 subshell on passing from period 5 to period 6. In short, if one wishes to use analogies based on trends in the configurations of the cores, eighteen of these analogies favor the assignment of lutetium and lawrencium to Group IIIB and only one favors the assignment of lanthanum and actinium.

Comment: But I can't now remember why I left this argument of his out of our submission. Sandbh (talk) 01:45, 17 December 2016 (UTC)

I have added one more paragraph, noting that such an argument has two sticking points:

1. It does not hold consistently through a period. When moving from period 5 to period 6, groups 4-18 (leaving out 3 as contentious for the moment) add 32 to the atomic number, but 1-2 only add 18. The pre-transition metals look like they belong to the previous period, like they really would in the Janet table, but that goes against the chemical sense of periodicity where a period should start with a reactive alkali metal and end with an inert noble gas.
2. There is already a precedent in ignoring this argument for chemistry in the placement of helium. Hydrogen over lithium has the expected change in atomic number of 2, but helium over neon has 8. Now I am not saying that La or Lu have such striking chemical differences from Y like He does to Be, but this shows that chemical resemblances can and do indeed overrule this sort of argument - because they already have! Double sharp (talk) 06:25, 19 December 2016 (UTC)

Comment: A nice piece of work, that extra para! Sandbh (talk) 09:13, 19 December 2016 (UTC)

In all honesty, this thinking is doubtful. (Considering no other data is available, just looking at consistency of this argument, since you break them apart anyway.) We place He above Ne because the difference is huge. What follows from this is not "chemical resemblances can and do indeed overrule this sort of argument"; what follows is "chemical resemblances of that huge strength can and do indeed overrule this sort of argument. It does not say that the far more moderate difference in group III does. (Also it does not say that the difference is insufficient to overrule or even that there is anything to overrule to begin with. We don't get anything from this argument.) This argument as such is irrelevant per pure logical reasoning.--R8R (talk) 18:34, 22 December 2016 (UTC)

Done. I've edited our analysis to make the relevance of the He over Ne clearer, and less emphatic. Sandbh (talk) 04:44, 23 December 2016 (UTC)

While I am still unable to sit down and read your comments and write mine, I want to note I reconsidered my thinking on this one. I see your point now; but as I see it, you cannot treat this as an argument, it is rather a starting point for thinking on this issue. I declared my starting point as "no alignance with either -La-Ac or -Lu-Lr," which implies this argument won't be reasonable. There could be other arguments, for which it could be reasonable, but it is not for mine.

I don't object and see your point now. Just calling on you to consider noting this simply for the purpose of writing a more reader-friendly text. (Or provide comments on why not, that's also always acceptable.) It took me a week (I figured this some time ago) to think of this. You probably want to make it clear to your readers immediately. --R8R (talk) 18:17, 22 December 2016 (UTC)

• (the quoted phrased is to be found at pp.256--257, not 243--245)
• The comment re the table is a picking. They clearly show lutetium in the row of platinum and lanthanum above it. This makes some, noted, sense given that they admit most other authors do it differently and this book is not Jensen of some kind and is not aimed at the group 3 revolution.
• But I otherwise agree.--R8R (talk) 16:02, 24 December 2016 (UTC)

Done. Edited to address your comments. Thank you. Sandbh (talk) 23:14, 24 December 2016 (UTC)

1965: Similarity of Lu with Sc and Y

I understand the general idea behind the comment of the first paragraph. Later, it becomes more complicated and consistency becomes questionable.

The devil is in the details. "Shall we place Al in the same group as Be, or Mg in the same group as Li?" -- no, but it's a no based on the different numbers of valence electrons. The reason why no is more important and basic for the whole PT than boiling points or so on and thus beats the idea of having Be and Al in one column. There is no similarly strong argument to be applied here. What this argument provides, given that you've written later essentially the same, is that you debunk the Be-Al and Li-Mg groupings (to which anyone would agree) and directly after that debunk the Y-Lu grouping. This is a far more controversial move. In fact, you do admit that: "to do so would not be quite as bad since they share the same valence." But if it's not the same, why did you even mention those first two unrelated cases apart from making the main one look better? I genuinely see no reason for doing so.

"Rather, one expects increasing atomic radius, increasing basicity, and increasing electropositivity." -- does one? The lanthanide contraction is not obvious, but can be expected. Actually, why do you think that the periodic table should be built on expectations of trends (you will refer to it later as well)? And if so, why do you think the standard linear trends should be the trends? (see below for more on that.) I am left to assume this, but I don't yet see why.

Comment: Well, since Scerri and colleagues comprise an educated audience, I don't see the need to explain to them the linear trends that tend to characterise the periodic table. Sure, there are anomalies but we don't start with them (that's why they're called anomalies or exceptions); we start with the generally observed trends and expectations, we cross-check our conclusions, and refine our findings as we go. That's the way, as I understand it, that science works: start with the general observations and see how well they hold up. Sandbh (talk) 04:03, 20 December 2016 (UTC)

Also, see this quote: "No, we daresay it is better to consider the similarity of Y^III to Lu^III as just an effect of the lanthanide contraction, like the astonishingly similar size of Zr^IV and Hf^IV." And yet we do group Zr and Hf together. The same is true for Nb and Ta, etc. This is an important factor in all other early d block groups. And here, it is implicitly called a side flaw. What I would suggest to at least do that explicitly and explain why "an effect of the lanthanide contraction" is "just an effect of the lanthanide contraction."

Comment: Double sharp has further adjusted the wording of our analysis in response to your concerns. Sandbh (talk) 02:27, 17 December 2016 (UTC)

Comment: I'm no longer sure we need this argument. Will look closer but I think we can say, like we've said elsewhere, that trends for La-Ac tend to match those of groups 1–2, whereas trends for Lu-Lr tend to match 4–5 or 4–10, so really we need to look at other arguments to decide. Sandbh (talk) 11:57, 22 December 2016 (UTC)

@Double sharp: I've simplified and changed our analysis, in order to address some of R8R's comments. Could you have a look at the new analysis and let me know if it looks OK to you. It is less judgemental and more questioning of the proposition that similarity connotes group membership. I'll make some knock-on amendments to the rest our submission. Sandbh (talk) 23:07, 23 December 2016 (UTC)

I like this very much! ^_^ Double sharp (talk) 15:25, 24 December 2016 (UTC)

So do I (mostly). I mostly agree with the outcome, but I don't like the presentation.

"We think these authors fail to explain" -- the interesting thing is, you found another author who extends the thinking mentioned under the title in the previous section. If no author does that, why don't you try yourselves? You later explore an argument all by yourselves. Besides, I don't even thibk that's the case. here's a phrase you quote: "the chemistry of Lu is more like the other nine 5d metals than is the chemistry of La." A solid argument, as for me. You may even say it's not a deal-breaker, but it is solid. (What he says is too wide a statement, sure, that I could see you argue against. I would limit myself on "the trend in chemistry of the shift from Y to Lu is more like the other nine 4d to 5d trends than is the Y to La trend." Still, I don't think the phrase "fail to explain" is correct.)

"In some other parts" --in Wikipedia, we have an appropriate template for that: {{weasel}}. But let's go on thinking. "In some other parts of the periodic table [I read, "in groups 1 and 2." I checked groups 13 and 14 against ionization energies trends (granted, that's where my analysis halted), and values for Tl and Pb slightly exceeded those of In and Sn, so even the label "main group elements" would be at least questionable] we see a continuation of trends." -- let's just put it there: group 1 and 2 in one side, groups 4 and 5 in the other.

I will also ignore the "tipping point" cliche for now. If it is the case, we can conclude that later. I remember thinking about that that cliche would fit, but is not found in, some of La-Ac arguments.

"We note that while IUPAC classifies the group 3 elements as transition metals they behave chemically like main group elements, a phenomenon we discuss later in this submission." -- the problem is that you are working with wide statements. The term "transition metals" spans over (~?)30 elements in periods 4--6. The term "main group elements" spans over.. what, by the way? Do we include group 13 here? Anyway, I think that wide statements like this can make you (could make me) lose focus in the borderline case.

Also, it's a different argument. Since you deal with arguments one apart from another, keep it that way. That "later" will come and we will talk about it there. It's not adding much to the discussion now anyway except that feeling that a -Lu-Lr argument just can't end on a good note. I do admit there are good -La-Ac arguments, and there is even one for which there is no -Lu-Lr counterpart (though if I recall correctly, it also has its flaws). But the thing is, I have thought about it and it seems I find myself in a situation in which I argue for -Lu-Lr arguments becase they are not given the same optimistic treatment as -La-Ac arguments. I will provide examples later if I haven't yet unless they're gone.

Long story short: you analyzed one argument, you made conclusions from that one argument. There is no reason to interrupt the discussion with a different argument since you deal with them separately.

Nevertheless, I do agree with this principal statement: "if group 3 were treated as main group elements we would expect more of a linear trend going down the group. On the other hand, if group 3 was treated as a transition metal group, it would be reasonable to expect more of a convergence properties on going from period 5 to period 6"--R8R (talk) 16:44, 24 December 2016 (UTC)

Thank you!

(2) "We think these authors fail to explain"—I've edit this expression to try and make our intended meaning clearer. I'm not happy about suggesting Landau and Lifshitz's position could be made stronger if they had extended it along the line of Jensen's "consistent addition of 32 to the atomic number" argument. So I'll move Jensen's argument into its own section. [Done.] Re extending the thinking about Lu-Lr generally, I expect we would if we could think of any new arguments. But I suspect all of the major positions in support of Lu-Lr have already been published. Wether or not we further explore or comment on any particular argument depends, I suspect, on if we see any merit in doing so (@Double sharp: feel free to chime in here). Some arguments are worth exploring or commenting further on; some aren't. The phrase, ""the chemistry of Lu is more like the other nine 5d metals than is the chemistry of La" was made by a chemist with expertise in Ln chemistry, and he made that comment based on Jensen's separation groups argument, as noted in our submission. I don't think it requires any further commentary since we only listed it as an example of the kind of thinking which equates similarity in properties with group membership. Sandbh (talk) 00:16, 25 December 2016 (UTC)

(3) "In some other parts"—I think I've fixed this one with some further editing but I'd like to hear if @Double sharp: agrees. Sandbh (talk) 11:22, 25 December 2016 (UTC)

Agreed by Double sharp. Sandbh (talk) 00:59, 2 January 2017 (UTC)

(5–7) Group 3 as main group—I agree and have removed these comments. Sandbh (talk) 21:59, 25 December 2016 (UTC)

1967: f-character of La

Generally, seems to be okay. One picking is this: "We found evidence for the latter effect (Gschneidner 1993, p. 8) to be inconclusive" -- I'd want you to mention that you found the former effect was valid to give the idea both were given equal treatise.

Done. Sandbh (talk) 01:22, 17 December 2016 (UTC)

"We note precedents elsewhere in the periodic table, in which Ca, Sr, and Ba are universally regarded as s-block elements despite having low-lying d-bands that contribute to their spectra, thereby showing the inconclusiveness of this argument." That is because their configurations are [Ng]ns2 elements that they are located in the s block. Both La and Lu, both [Xe](4f14)5d1. Both could be located in the d block if we decided the block could stretch behind 10 columns ("block" is human-made term, after all). We decided we have to choose one. Small arguments begin to gain ground once we don't have the big arguments (electron configs, for example). The analogy is not great here assuming our starting position is "we don't know if it's Sc-Y-La or Sc-Y-Lu." (If our starting position is "Sc-Y-La," then sure. if out starting position is "Sc-Y-Lu," then this sentence probably doesn't matter much.)

Done. I deleted the sentence. Sandbh (talk) 01:54, 17 December 2016 (UTC)

Looks good to me. (I will assess all "tipping point" statements later. Probably I'll agree with this one.)--R8R (talk) 16:47, 24 December 2016 (UTC)

Thank you! Sandbh (talk) 23:19, 24 December 2016 (UTC)

1982: Separation groups

"The fact that Sc, Y and Lu occur in the so called yttrium group, and that La and Ac occur in the "cerium" group does not imply anything particularly significant; it is simply a reflection of the increasing basicity of these elements as atomic radius increases. Taking the alkaline earth metals as another example, Mg (less basic) belongs in the "soluble group" and Ca, Sr and Ba (more basic) occur in the "ammonium carbonate group" (Moeller et al. 1989, pp. 955–956, 958)" -- had you stopped there, I wouldn't have anything to comment on. "Arguing that Lu should go under Y simply because they occur in the same chemical separation group ignores periodic trends. Moeller (1961, p. 10) in fact gives Sc, Y, La and Ac as the first members of the four "transition" series." -- you have sections devoted to that. I suggest you leave this to them.

Not done. I haven't acted on your suggestion as I think our analysis is accurate, thorough, and appropriately worded as a counter-argument. OTOH, I haven't made up my mind up yet about the factual statements v counter-arguments issue. My current thinking is that if Jensen presents a series of arguments then any analyses of them, as well as assessing factual accuracy, should consider any counter-arguments. Sandbh (talk) 03:02, 17 December 2016 (UTC)

Update: As noted in the Discussion section below, item 2, I believe Double sharp and I are now comfortable with the way we have addressed Jensen's "arguments". Sandbh (talk) 22:27, 18 December 2016 (UTC)

My thinking is that periodic trends is a different problem than the problem in question, separation groups. This is not very appropriate as a counter-argument to go under the title 1982: Separation groups.

"Similarly, Li is usually found with Mg and not with the heavier alkali metals Na and K; and Be does not occur together with Mg, Ca, Sr, and Ba." -- this is a far better counter-argument which I find absolutely relevant.

Otherwise, the argument seems okay now.--R8R (talk) 17:29, 24 December 2016 (UTC)

Thank you. I edited what we said about periodic trends to make it fit better. Sandbh (talk) 00:10, 26 December 2016 (UTC)

I still don't follow. The problem is not that the argument is poor nor that it is breaks the flow of the text. It is different.

"Making an argument that Lu should go under Y because they occur in the same chemical separation group fails to consider separation group patterns elsewhere in the periodic table and, in this context, does not demonstrate why Lu under Y is a superior outcome. Moeller (1961, p. 10) in fact gives Sc, Y, La and Ac as the first members of the four "transition" series." -- this reads like it's supposed to be an antithesis, but it isn't one. Belonging to another transition series and belonging to a another separation groups are different problems. You keep separate problems separately.

In fact, I believe that the first sentence in the quote I provided is in itself essence of a good counter-argument that absolutely diminishes the original one. I don't see why you would need anything that doesn't even belong here anyway.--R8R (talk) 19:01, 30 December 2016 (UTC)

Thanks again. I've edited this section to try make its meaning clearer. And I've separated out the discussion of the four transition series. So the structure of the analysis is now: opening statement addressing Jensen's argument as supported by his reference (1); paragraph addressing Jensen's supporting reference (2), noting that this reference is partly lacking in support for, and not consistent with, Jensen's argument; closing paragraph addressing Jensen's reference (3). Sandbh (talk) 05:51, 31 December 2016 (UTC)

Also, "If anything, this would seem to suggest that since modern periodic tables generally show Be-Mg-Ca despite these resemblances they should analogously show Sc-Y-La." -- I still think too much is left to me to think why. Because Be-Mg-Ca is not the best pick and we should choose Sc-Y-La because it's not the best pick, either? I don't understand what you mean.

Done. Copy-edited to make meaning and intent clearer. Sandbh (talk) 03:02, 17 December 2016 (UTC)

Nevertheless, I think the original argument was weak if not useless. Yet I don't approve of the current way of saying that as it doesn't stay focused on the argument itself, but rather attempts to draw a conclusion on the general group 3 question right now, with its flaws.

Noted, as per above. Sandbh (talk) 03:02, 17 December 2016 (UTC)

Irregular electron configurations

More of how you look at it, but I see no logical problems here.

Thank you! Sandbh (talk) 02:28, 17 December 2016 (UTC)

Electron configuration analogy

— New section placeholder: I moved Jensen's extension argument to Landau and Lifshitz's argument, here. Sandbh (talk) 00:05, 26 December 2016 (UTC)

Probably not how I would write about it, but I find this fine.--R8R (talk) 18:39, 28 December 2016 (UTC)

Thank you! Sandbh (talk) 23:46, 28 December 2016 (UTC)

Ionization potentials

Two potentials are indeed not as important. Suggest dropping.

Not done given this is one of Jensen's arguments. Sandbh (talk) 04:01, 18 December 2016 (UTC)

I see your point but I'm still calling on you to reconsider. Even though Jensen mentions this, you are not bound to do a counter-analysis. I do not mean to limit you in a way and I think that this could be mentioned. But that is, mentioned. You get essentially same results for 2 IEs and 3 IEs, so I suggest focusing on the more relevant stat. Probably I would follow your explanation if the results were different, but they are similar.--R8R (talk) 18:45, 28 December 2016 (UTC)

As for "not the full story": you see, this set of arguments is collectively called "Lu-Lr arguments". It would make sense if you included a similar argument among La-Ac arguments instead of criticizing this one.

Not done, for the reason explained in the Discussion section below, item 2. Sandbh (talk) 03:54, 17 December 2016 (UTC)

I'll take it.--R8R (talk) 18:45, 28 December 2016 (UTC)

As for closer look: "We thought his reference to "ns-electrons" explained why he was using the sum of the first two IE's rather than the first three (since it was group 3 that was the anomaly) but this does not work since, for example, Nb, Cr and Mo only have one s electron, so he is not really comparing like with like as, to get the sum of the first two IE's for these metals he has to include one d electron each" -- how would you work around it then? Maybe you would suggest not including these arguments? That's what I'm suggesting you do. The two-electrons argument is not well reasoned enough. "We think this explains why Jensen left out the group 11 and 12 metals in his article." -- could be. yet I'll add that group 9 already isn't too relevant here. All in all, it's better to focus on categorization than on categorizers.

Not done. Our closer look was intended to show that we were interested in trying to work out the puzzle of what Chistyakov and Jensen may have seen as the relevance of the first two IP's, without dismissing this argument completely. Sandbh (talk) 03:54, 17 December 2016 (UTC)

Atomic radii

See the remark above re "not the full story."

Not done, for the reason explained in the Discussion section below, item 2. Sandbh (talk) 06:11, 18 December 2016 (UTC)

Ionic radii

This argument boils down to "Sc-Y-La is a linear trend, and that's the sort of thing we expect to see in the periodic table." Similar reasoning has been made in some sections above, to all of which I'll try to answer here. While the linearity part is out of question, why is it taken that that's the sort of thing we want? This has not been explained.

Comment: I don't believe we're saying here that we necessarily expect to see linearity. I think we are saying, or trying to say, two things:

1. Jensen says or implies that the trend going down Sc-Y-Lu for atomic radii; sum of the first two ionisation potentials; melting points; electronegativity; ionic radii and redox potentials matches the trend seen in groups 4–8 or 4–10 better than is the case for Sc-Y-La;
2. We say that the trend for Sc-Y-La matches the trend seen in groups 1 and 2 better than is the case for Sc-Y-Lu.

Done. Sandbh (talk) 23:32, 19 December 2016 (UTC)

Perhaps we should adjust our analysis of this argument. Sandbh (talk) 00:34, 18 December 2016 (UTC)

Done. Sandbh (talk) 01:04, 21 December 2016 (UTC)

(The two examples of that occurring before:

• "similarity of YIII to LuIII as just an effect of the lanthanide contraction" -- why just an effect and not the effect? Other early d block period 6 elements have this effect as well. As does gallium.

Comment: Double sharp has addressed this in response to your concerns about the "1965: Similarity of Lu with Sc and Y" argument, above. Sandbh (talk) 00:34, 18 December 2016 (UTC)

• "Arguing that Lu should go under Y simply because they occur in the same chemical separation group ignores periodic trends." -- hmm, even that is not as straightforward as you put it. More on that just below.)

Addressed: see the 1982: Separation groups section, above. Sandbh (talk) 00:34, 18 December 2016 (UTC)

Group 4 has a similar trend, and so does group 5. It cannot be taken for granted that we want linearity to go into period 6.

Clarification requested. @R8R Gtrs: The trend going down groups 4 and 5 is similar to what? Sandbh (talk) 00:34, 18 December 2016 (UTC)

Similar to the Sc-Y-Lu trends, I presume. I agree, but would suggest again that these are due to the lanthanide contraction just before. This doesn't really hold as a criticism of Sc-Y-La because La precedes the 4f collapse implicated in the lanthanide contraction, though. So I would argue that since we have a choice, we should choose the more normal-looking trend. Double sharp (talk) 06:33, 19 December 2016 (UTC)

I am okay with your replies up to this point. But the section is still bad, and I'll try to expalin why I think so.

Let's follow the order: Jensen gives a pro-Lu-Lr argument -- "No data nor a reference is given" -- Sc-Y-Lu is similar to groups 4 onward, Sc-Y-La is similar to groups 1 and 2. _{(so Jensen's argument was not based on nothing, just incomplete. why start off with such negativity then and attack the presentation rather than the argument itself?)} -- A data comparsion by Atkins argues for -La-Ac. This last one is a lengthy argument that does not provide any information we have not said by this point. What does it add rather than the length of text arguing for -La-Ac?

I am left with the feeling that at least the authors put -La-Ac over -Lu-Lr despite having explicitly said than it depends on what trends you focus on.--R8R (talk) 19:23, 28 December 2016 (UTC)

Addressed. I've deleted the comment about Jensen not providing a reference nor data. I've edited and retained the Atkins argument as counterpoint to Jensen. So the structure of this section now goes: (1) Jensen makes a statement saying trends in ionic radii support Lu-Lr; (2) The data we can find suggests Lu-Lr is similar to the trend seen in groups 4–10 whereas La-Ac is similar to the trend seen in 1-2; (3) Another argument in the literature suggests that ionic radii data supports La-Ac.

The argument by Atkins provides new information: (1) it mentions the importance of ionic radii in influencing chemical properties; and (2) it finds in favour of -La-Ac on this basis.

I've added a further closing comment about Atkins that clarifies that they haven't really chosen any particular configuration and that it seems that their ionic radius criterion was intended as no more than an item for discussion. Sandbh (talk) 01:53, 29 December 2016 (UTC)

Much better now on the issue of balance. Agree on that it's not a decisive factor anyway.--R8R (talk) 19:04, 30 December 2016 (UTC)

Thank you! Sandbh (talk) 05:53, 31 December 2016 (UTC)

Redox potentials

Sure. Yet consider that this argument is equally valid for -La-Ac and appears only here.

NFA intended. I'd also list this argument in the -La-Ac section if I could find it in a reference, or if I thought it had some relevance. Since neither of these criteria were satisfied, I didn't. Since Jensen thought there was some merit to this argument in support of -Lu-Lr, I listed it here. Sandbh (talk) 04:28, 17 December 2016 (UTC)

Not really happy about your reply (How does this have any less relevance for -La-Ac than for -Lu-Lr?), but let's leave it there.--R8R (talk) 19:26, 28 December 2016 (UTC)

Thank you! Opening attack removed. Sandbh (talk) 01:56, 29 December 2016 (UTC)

Electronegativities

Again, see the comment on "not the full story."

Not done, for the reason explained in the Discussion section below, item 2. Sandbh (talk) 04:02, 18 December 2016 (UTC)

"The Pauling scale instead clearly favours Sc-Y-La. Groups 1, 2, 4, and 5 have the period 6 element somewhat more electropositive than the period 5 element; this works with La (1.1) under Y (1.22) but not with Lu (1.27) under Y." not as clear to me. see the comment in the previous section.

Addressed. Edited our analysis and added another example. Sandbh (talk) 05:07, 17 December 2016 (UTC)

Looks good to me now.--R8R (talk) 19:28, 28 December 2016 (UTC)

Thank you! Sandbh (talk) 23:48, 28 December 2016 (UTC)

Melting points

Again, see the comment on "not the full story."

Not done, as per above. Sandbh (talk) 04:03, 18 December 2016 (UTC)

Crystal structures (elements)

Agree on "relevance is questionable."

"We note that in groups 7–9 we have a first-row anomaly between the 3d metal and the heavier two members whereas Sc-Y-La-Ac has the heavier elements be inconsistent, which is more akin to the situation in group 2." -- You yourselves would tear a Lu-Lr argument of similar strength.

Done. I think; Double sharp deleted this sentence. Sandbh (talk) 05:12, 17 December 2016 (UTC)

Crystal structures (oxides, chlorides, various intermetallics)

OK.

Thank you! Sandbh (talk) 05:13, 17 December 2016 (UTC)

Excited state spectra

"So, we tend to find Hamilton inconclusive as he does not say anything about the spectra of Ca, Sr, and Ba." -- uuuh, we're talking about group 3, right? Wonder why you are okay with him not mentioning anything about group 4. (Or even worse, what if he does and you just ignore group 4 but ask for group 2?) This is one the strongest points why I wrote what I did in the header of this level 3 section.

"We would also draw some links between the situation in groups 2 and 3." -- could we also have some links to group 4?

Good. You've picked up on some unconscious bias in the way we expressed our analysis.

I guess there might be some p electrons in the excited energy levels of group 4? For Ca, Sr and Ba in group 2 I don't know for sure but I presume there would be d electron orbits in their excited energy levels whereas this would not be the case for Be and Mg. I think we should adjust our analysis but would like to hear from Double sharp on this point. I'm not sure whether to change our status marker to "tipping point" or whether to leave it as "meaningless" (which now seems a little harsh). Sandbh (talk) 05:47, 17 December 2016 (UTC)

Analysis adjusted. Sandbh (talk) 04:41, 20 December 2016 (UTC)

Superconductivity

OK.

Thank you! Sandbh (talk) 05:47, 17 December 2016 (UTC)

Conduction band structures

Not exactly sure what troubles me here, so let's say this is OK, too.

Thank you! Sandbh (talk) 05:48, 17 December 2016 (UTC)

But something did trouble me. I gave it another look.

"In any event, the significance of such an atypical structure is not clear to us since La still has the high density of states that are characteristic of transition metals" -- this contradicts, without any explanation that I can recall, your general line of La being a group 3 pre-transition metal. Later, you say, "Group 3 shows chemical behaviour that is manifestly uncharacteristic of the transition metals proper." Your reasoning appears to be inconsistent.

Comment: Thanks for noticing this. Conduction band structure is not something I'd regard as chemical behaviour, so I see no contradiction. Sandbh (talk) 09:17, 31 December 2016 (UTC)

This whole talk is difficult to keep track of without any numeric data. Some quick comments:

• "since we generally expect the conduction bands of metals at or towards the end of the s, d and f blocks, and some of those of the p block metals (e.g. Sn, Bi, Po) adjacent to the metalloids, to be atypical for their block." -- a) do we? b) why "at or towards the end" and not the beginning and, most importantly, c) why at all? can a full explanation shed more light?

Comment: I've presumed that, since Scerri and colleagues comprise an educated audience, I don't need to elaborate. Elsewhere in our submission we mention the influence of the empty d-bands on the heavy alkaline earth metals. We discuss the fact that even though group 12 have full d-shells, these still affect their properties. It's common knowledge that we see strange things happening to band structures when we get close to the metal-nonmetal dividing lines e.g. the semimetals (electronically speaking) As, Sb and Bi.

@Double sharp: How do you see this?

I think we should strive to make the submission self-contained, so we should probably add a (hopefully brief) explanation. Double sharp (talk) 11:45, 1 January 2017 (UTC)

Done. Simplified and added citations. Sandbh (talk) 00:56, 2 January 2017 (UTC)

• "We have commented elsewhere that similarity in properties does not necessarily connote group membership, and we think this is another example." -- this is confusing. So it could be a reason, but we sort of already decided that whatever does not support our -La-Ac hypothesis may be just data falling out of line, that happens, never mind. Please rethink the text so it does not allow for this interpretation.

Comment: I've edited this to try and address your points. Sandbh (talk) 09:17, 31 December 2016 (UTC)

"Sc-Y-La-Ac further shows simple trends of increasing basicity and in this respect are much more like their leftward neighbours Ca-Sr-Ba than their rightward neighbours Ti-Zr-Hf." but "we have commented elsewhere that similarity in properties does not necessarily connote group membership, and we think this is another example"? why would it not work this way?--R8R (talk) 20:39, 30 December 2016 (UTC)

Good: I've removed words in the second quote after "but". Sandbh (talk) 09:17, 31 December 2016 (UTC)

1985: Seven properties

"As a general observation we think Holden falls into the trap of thinking that similarity, rather than periodic trends" -- this, again, is in need of being proven.

Edited accordingly. Sandbh (talk) 01:07, 20 December 2016 (UTC)

"We further note that the density of Ba (in group 2) is about the average of the density of Cs (group 1) and that of La (group 3). This relationship is reinforced by analogies seen in period 5 for Rb-Sr-Y, and in period 4 for K-Ca-Sc." what is this supposed to mention? Personally, I attribute this to a group I-II-III trend, which is not relevant to the discussion. If you mean anything else, say so.

Comment: We are saying that the logic of his argument is flawed. We then provide a rebuttal. When he says the average of Ca + Ti ~ Sc, and the average of Sr + Zr ~ Y, these are OK but when he says Ba + Hf ~ Lu, he isn't comparing like with like. Our counter-example is K + Sc ~ Ca; Rb + Y ~ Sr; and Cs + La ~ Ba. Sandbh (talk) 01:01, 18 December 2016 (UTC)

2005: Physical properties

Have not read the source, but if I found what is stated here, I would agree.

"When this is done for Lu and La, as set out elsewhere in this submission, we find a sound case for La, rather than Lu, as a homologue of Y." the case is becoming less clear to me as we proceed.

Noted. Hopefully this will get better as we consider and address your comments. Sandbh (talk) 01:04, 18 December 2016 (UTC)

2006: Characteristic electron configurations of ions

OK to me.--R8R (talk) 14:17, 1 January 2017 (UTC)

Thank you! Sandbh (talk) 01:21, 2 January 2017 (UTC)

2011: Triads

This one is hard to comment on as I don't know the context and don't know what Scerri defined as a triad. I am positive Scerri can do basic arithmetic.

NFA required. He's aware of this particular argument but hasn't shared his thoughts about it with me. Sandbh (talk) 01:06, 18 December 2016 (UTC)

2012: Split d block

This is both rhetoric from you and from him, so I won't comment.

NFA required. Sandbh (talk) 01:06, 18 December 2016 (UTC)

2013: Carbonyls

OK. The original argument was also weak.

Thank you! Sandbh (talk) 01:07, 18 December 2016 (UTC)

2014: Lu as a transition metal

This one made me think if the structure of this article was chosen correctly. You have a source advocating for -La-Ac and a source advocating for -Lu-Lr. I would think these are in similar positions and should be described similarly, but rather you put the former as a counterweight to the latter.

NFA required. We cannot describe each source similarly since we don't have access to the mathematical analysis techniques used by the authors—this is not our job in any case. All we have done is point out the flaw in their logic and provide some counter-examples. Which is a pity as it is otherwise a good article. Sandbh (talk) 01:16, 18 December 2016 (UTC)

Name matches position

"In any event we consider that this argument is almost too weak to bother with." -- true. But why did you even bother to mention and write a para opposing it, then?

Comment: It was an "argument" made on a Scerri blog by someone with professional expertise in the Ln. It isn't a strong argument but we thought it was (just) worth noting. Sandbh (talk) 01:54, 18 December 2016 (UTC)

2015: Dimer spectroscopy

Let's say OK, but note that the other lanthanides are irrelevant.

Thank you!/Not done. The argument of Fang et al. is incomplete, as explained in our analysis. Sandbh (talk) 02:43, 18 December 2016 (UTC)

I didn't question its incompleteness. (You having noted that they had not mentioned groups 2 and 1 was the argument for me.) But you're probably right.--R8R (talk) 20:44, 30 December 2016 (UTC)

Thank you! Sandbh (talk) 09:19, 31 December 2016 (UTC)

Relativistic contraction of 6s shell

Agree.

Thank you! Sandbh (talk) 02:44, 18 December 2016 (UTC)

Aluminide dimers

Sure. Why do you even mention such insignificant points when there are more important ones?

Thank you! We mention it because Jensen did. Sandbh (talk) 02:45, 18 December 2016 (UTC)

Heat of vapourisation

Sure

Thank you! Sandbh (talk) 02:46, 18 December 2016 (UTC)

Ionization energy of Lr

Sure

Thank you! Sandbh (talk) 02:46, 18 December 2016 (UTC)

La-Ac arguments

Six properties

The title reads, "Six properties." Why are only five properties listed?

Comment: There are six properties listed. Some of them are multifacted; others are are overlapping, so we grouped, renumbered and translated his observations into five arguments. I've edited our summary to make this clearer. Sandbh (talk) 04:27, 18 December 2016 (UTC)

One by one:

1. OK.
2. OK.
3. The general idea is OK.
4. OK.
5. Again, eager to see in the review why that stepping out of the ordinary for groups 4 and 5 is okay and why this is not the ordinary for the early d-block.

Response pending your response to my earlier ping. Sandbh (talk) 04:27, 18 December 2016 (UTC)

Sorry. I've seen the ping a while ago. Only now am I able to finish the second go-over on this review. I missed it now; anyway, DS got it right.--R8R (talk) 20:56, 30 December 2016 (UTC)

Thank you! Sandbh (talk) 09:21, 31 December 2016 (UTC)

The added historical note is somewhat confusing as its presence has not been explained. See this quote from this very review: "the historical record does not necessarily flag ongoing relevance." Why is this the case, and is it even?

Comment: I've edited the note so that it's relevance is clearer. I believe the point about the historical record is that there are some things in it that are no longer relevant, such as the placement of Be and Mg in group 12. OTOH, there are some things in the historical record that still ring true, and that we appreciate for their farsightedness. I thought 1927 was a good effort. Sandbh (talk) 04:27, 18 December 2016 (UTC)

(I want to add another comment on this historical note but cannot formulate it correctly. Will take some time to think if it's because my comment is irrelevant or not; putting a note for now.)--R8R (talk) 20:56, 30 December 2016 (UTC)

Noted. Sandbh (talk) 09:21, 31 December 2016 (UTC)

I've sat for 20 mins trying to write it down. Still can't get it, so I'll give it a pass. Would probably write something if there was something to write.

One idea I've had for a long time, but am not sure to have mentioned: take all my critical comments as comments and thinking you may induce in people in the committee. I am sure that you won't want to receive a letter beginning with "Dear Sandbh & DS, we were happy to receive your letter but you're missing something:" that would lay down ideas similar to mine. I'm not saying that would happen; just want to be sure you've considered that and agreed the the risk is minimal or that you'd ready to take it if this does happen.--R8R (talk) 16:56, 6 January 2017 (UTC)

Yes. I'd be comfortable if that would happen; I suspect Double sharp would be too, as we'd both be delighted to read the task group's reasoning. Sandbh (talk) 02:10, 7 January 2017 (UTC)

I think this sounds somewhat unwelcoming; let me rephrase that. The thing is, I wanted (I guess I've already said what I wanted to say by now and can put this in past tense) to expose you to what could've been said in response to you especially given that the chairmen support the other option; this way, you could deal with it before there is an actual chance to receive such a reply. I haven't had this thought consistently throughout the review, but it did influence my responses many times. I think that in many ways, my review wasn't for naught since it did change the letter in some aspects. I also hope it did induce more thinking not only about how to compose a good piece, but also about how it will be received and if it will get through.--R8R (talk) 22:30, 6 January 2017 (UTC)

Yes. Your input has been most welcome and has reminded of forging and tempering. What Double sharp and I have forged has benefitted considerably from your tempering review! I'm not fussed about Scerri's preference, if indeed he still has one. He's changed his mind on other things and was the one that got me going on questioning Jensen's arguments. I expect Lavelle will help in any event. As I said, it's an impressive task group. Sandbh (talk) 02:10, 7 January 2017 (UTC)

Thank you and DS for your kind words and the opportunity to contribute with my review in first place. I wish you luck and that all goes well and the thing breaks through. Glad to learn there's a good chance of that.--R8R (talk) 16:31, 11 January 2017 (UTC)

1974: Integrity of the f block

@R8R Gtrs: Could you please look up this reference: Shchukarev SA 1974, Neorganicheskaya khimiya, vol. 2. Vysshaya Shkola, Moscow (in Russian), p. 118. I believe the author argues for La(Ac?) on the basis of the sequence of filling the electronic subshells in Sm–Eu–Gd and in Tm–Yb–Lu. Thank you. Sandbh (talk) 07:00, 27 December 2016 (UTC)

@Double sharp: I'll have a go at guessing this argument but if you see this before I get to it feel free to give me any thoughts. Thank you. Sandbh (talk) 07:00, 27 December 2016 (UTC)

This sounds a little like Michael Laing's periodic table with La and Lu flanking Gd vertically, since all three of them have a differentiating 5d electron. But I don't think such an argument would prove anything by itself, so there is probably more to it in the source (which I wouldn't be able to read). Double sharp (talk) 07:17, 27 December 2016 (UTC)

It may have something to do with where the f block is deemed to end. So, the secondary periodicity that we see in the Ln goes from Ce to Gd (the cerium group) and Tb to Lu (the yttrium group) and you can see this reflected in the way the 4f shell fills over Sm-Eu-Gd as it gets to half way, and the way the pattern similarly maxes out in Tm-Yb-Lu. So I think you could say that because the effects of filling the f shell recur from Ce to Gd; and Tb to Lu, most visibly(?) in the +2 states for Eu and Yb, this would suggest that the f-block should be considered to end at Lu. Hence La goes under Y. Something like that? Sandbh (talk) 09:42, 27 December 2016 (UTC)

R8R Gtrs comment: The book indeed argues for -La-Ac. Here's what I found: "If we [...] considered the latter [Lu and 103] not as 4f and 5f elements but rather as members of 5d and 6d series, the d-electron prevention [probably you find that this word doesn't really fit in the context of English. It doesn't look any better in Russian -- R8R] determining fillng f vacancies as stable would be lost as well as the correctness of placing of imitators before Gd and Cm as well as Lu and 103. The exceptional uniqueness of Gd and Cm, akin to that of Mg and Ca, would also be unclear."

That book reminded me of the specific language of Soviet scientific literature. (Senior scientists still use it; the younger ones often prefer a more casual language, though the shift is slow.) --~~~~

Wow! That was quick, thank you. The libraries here are closed until Tue 3 Jan :(
When they talk about the d-electron prevention I guess they are talking about the fact that in lanthanum, the 5d shell is more stable than the 4f. So the 4f shell doesn't start filling until cerium, so if you put Lu-Lr in group 3, it looks like there are only 13 4f elements. And I guess when they talk about imitators before Gd and Cm, and Lu and 103, they are referring to the fact that Eu and Am have half-filled f shells (for Am this applies only to the gas phase; as a solid it is 5f6 6d1 7s2) just like Gd and Cm, and that Yb and No have full f shells, just like Yb and No. I don't understand their reference to Mg and Ca having a uniqueness akin to Gd and Cm.
Overall, I suppose their argument is that the filling of the 4f shell is periodic, with two periods. Each period corresponds to the first or second half of the 4f shell being filled by electrons. And if you treat Lu as a 6d metal, you lose this 4f periodicity. Sandbh (talk) 04:54, 28 December 2016 (UTC)

@Double sharp: Does my attempt to turn Shchukarev into an argument look OK? Sandbh (talk) 07:44, 28 December 2016 (UTC)

R8R Gtrs comment: For the record, I have not found anything regarding the group 3 issue beyond what I translated. Didn't see anything about two periods (but actually, what if those imitators would be it? don't know, really). I am not entirely sure if you interpeted his words correctly (because I'm not sure if I understand them correctly).

Comment: I've edited our explanation to make it clearer that this is our interpretation of what Shchukarev appears to be saying.

"We agree with Shchukarev" -- not really sure what you agree with. Can't comment now. Could you go over the original quote phrase by phrase (and what imitators are we talking about)?

Comment: As above. Sandbh (talk) 23:40, 31 December 2016 (UTC)

"noting the most important periodic property of the Ln and An is their valency" is this a different argument? why is it here?

Comment: No it's not a different argument. It's a statement reinforcing the significance of what Shchukarev appears to have said. Sandbh (talk) 23:40, 31 December 2016 (UTC)

I see. No argument then.--R8R (talk) 13:52, 3 January 2017 (UTC)

"Thus, in the Ln, we see the analogous +2 ions of Eu and Yb, and the +4 ions of Ce and Tb" -- but wait, this is a sound argument for -Lu-Lr, is it not? If we assume that La is an f-block element, then the idealized configurations would be: Eu²⁺ = [Xe]4f7, Yb²⁺ = [Xe]4f14. If we assume that La is a d-block element, then the idealized configurations would be: Eu²⁺ = [Xe]4f6 5d1, Yb²⁺ = [Xe]4f13 5d1. (There is no difference in the +4 state.) We would rather expect Gd and Lu, rather than Eu and Yb, to form 2+ if La was a d-block element?--R8R (talk) 21:12, 30 December 2016 (UTC)

Comment: We have a separate argument discussing the fact that gas phase configurations support -Lu-Lr, in that there are less discrepancies between idealised and actual electron configurations than is the case for -La-Ac. Sandbh (talk) 23:40, 31 December 2016 (UTC)

Not quite. You do have an argument that says something like "trivalent ions show filling order of 0 through 14 for La through Ac, and we should take the 1--14 elements to name them the f block." That argument is logically consistent. However, as you discuss a wider picture of an argument within the initial analysis, there's something I've missed there. We assign positions in the d block per differeing electrons in the neutral state. We never talk about oxidation states and oxidized elements when doing so for 3d elements, for instance. Why is a different approach "more relevant" in the f block? That would at very least require an explanation, though I'm leaning towards a complete rethinking.

Overall, I've been uneasy about those "we agree that" parts in which you accept some initial conditions to accept -La-Ac. (Just to note, I assume it was those conditions that made you choose -La-Ac and not the other way around.) But you completely miss an opportunity to discuss the possibility of other initial conditions, which may be a reason to reject your argument for a person who considers different initial conditions (what we should look at in the f block, in this case. There were other such points elsewhere, I may look them up if you want me to).

Also, what you say there does not reply to the issue raised here. You discuss trivalent configs only there, while I pointed you at deviations from them. The issue are interconnected but are different. Again, I do think this raises a point for -Lu-Lr and cannot be just dismissed (can be overweighted by other arguments, though) as this is absolutely consistent in the -Lu-Lr picture and is not in the -La-Ac one, where we could expect +2 at Gd and Lu instead. (In the 3d series, copper is the penultimate deviant but it still is predominantely +2 in compounds, which is consistent with its idealized, rather than actual, electronic configuration, so the Cu--Yb analogy within -La-Ac doesn't work.)--R8R (talk) 13:52, 3 January 2017 (UTC)

Comment: You are confusing two separate arguments. We have one argument discussing gas phase v solid phase electron configurations. We have a separate argument looking at the electron configuration of ions.

About ions, I thought Wulfsberg explained it well enough. Important ions have no anomalous configurations. So, in the d block there is thus little reason to worry students about anomalous electron configurations. In the f block (so called because the elements concerned have f electrons) the characteristic electron configurations are most easily arrived at by considering the f-block to start with Ce and Th, noting that Ce 3+ is [Xe]4f1 and Th3+ is [Rn]5f1. (You could argue the point by saying that +3 is not really a characteristic oxidation state for Th but even Wiberg (2001, p. 1645), in comparing the lanthanides and actinides, says that trivalency is a characteristic of both Ln and An).

About Eu and Yb as +2 v Gd and Lu as +2, as I've noted previously at the end of this section we do not base arguments to do with actual chemical behaviour on idealised electron configurations. We have one argument that says that idealised configurations match better with actual configurations found in the gas phase for -Lu-Lr whereas the match is better for with actual configurations found in the solid phase for -La-Ac (the solid phase being arguably more relevant to actual chemistry). We have a separate argument that says the actual chemical periodic behaviour of the f bock is better captured with -La-Ac. I do not see the validity of mixing these two arguments. Sandbh (talk) 03:23, 4 January 2017 (UTC)

@Double sharp: Hi Double sharp. Happy end of 2016/start of 2017!! Did the above ping get to you? And could you also have a look at the new Lu-Lr argument by Wulfsberg, called 2006: Characteristic electron configurations of ions? This is a strange one as he says it justfies Lu-Lr but when I looked closer at it, it supports La-Ac, quite elegantly and strongly, to my astonishment. Thank you. I'll post a sit rep a bit later it terms of what still needs to be done overall. Sandbh (talk) 01:16, 31 December 2016 (UTC)

It probably did, but was drowned in a sea of other pings. And yes, it's still 31 December - actually I'm pretty sure it's not 2017 yet anywhere. I like both arguments, but certainly Wulfsberg's one (repurposed for -La-Ac) is by far the more breathtaking! Double sharp (talk) 01:49, 31 December 2016 (UTC)

"Placing Lu and Lr under Y obscures the start and end of the f block and truncates its double periodicity." -- this is a strong claim. And such, it needs a good rationalization. There is none at this point? Moreover, those "+2 ions of Eu and Yb, and the +4 ions of Ce and Tb" do support -Lu-Lr.--R8R (talk) 14:22, 1 January 2017 (UTC)

Comment: I've expanded the first sentence to make it clearer what we mean about double periodicity etc. Re the +2 ions of Eu and Yb etc I think you are applying idealised configurations to real life, whereas we know that the f block metals don't necessarily comply with their idealised configurations. Regardless of whether we have -Lu-Lr or La-Ac, it is Eu and Yb that show +2 and this works for their gas phase configurations and their solid phase configurations. We do however have a separate argument showing that gas phase configurations match -Lu-Lr idealised configurations much better than is the case for La-Ac (and, of course, this extends to a match for Eu and Yb). Sandbh (talk) 02:47, 2 January 2017 (UTC)

2004: La-Ac remains the most popular form

OK.--R8R (talk) 21:13, 30 December 2016 (UTC)

2008: A pair out of place

The first part of this argument is purely rhetorical. "He says that placing Lu and Lr in the d block, and La and Ac in the f block leads to a worse outcome than leaving La and Ac in the d block since this would represent, 'the only case where a pair of elements [i.e. La-Ac] is placed such that they are part of block [i.e., the f block] with no outer electrons in common with that block.'" Following similar logic, I can say, "placing La and Ac in the d block, and Lu and Lr in the f block leads to a worse outcome than leaving Lu and Lr in the d block since this would represent the only case where a pair of elements [i.e. Lu-Lr] is placed such that they are part of block [i.e., the f block] with all outer electrons in common with that block 'covered' by another electron." These arguments are of equal strength.

Comment: I don't believe you're comparing like with like. It's one thing to have a pair of elements in a block with zero outer electrons in common with that block. It is a pale imitation of that argument to compare it to a pair of elements, both having f electrons in common with the rest of the f block, and each having these 'covered' by a few d and s electrons. So what? It is like comparing a rose to an onion. Same thing happens with the Gd–Cm pair. Sandbh (talk) 05:16, 18 December 2016 (UTC)

I would not provide that argument if the change from Yb to Lu was adding the fourteenth 4f electron rather than a 5d one to a full 4f shell. "It's one thing to have a pair of elements in a block with zero outer electrons in common with that block." I don't see how an element with the differentiating electron being added past a complete shell is less strange.--R8R (talk) 21:19, 30 December 2016 (UTC)

Comment: I think it's less strange because in group 12 we already have a triad in which a differentiating electron is added past a complete shell; and with the Gd-Cm pair we have the bizarre situation of a 5d or 6d differentiating electron being added in the middle of the f block. There are no other examples of a pair of elements [i.e. La-Ac] placed such that they are part of block [i.e., the f block] with no outer electrons in common with that block. Sandbh (talk) 00:02, 1 January 2017 (UTC)

Likewise, there is no situation elsewhere in the PT similar to Lu and Lr being put into the f-block. The analogy with the Cu-Zn is not an exact match: Compare [Ng]3d10 4s1 -> [Ng]3d10 4s2 with [Ng]4f14 6s2 -> [Ng]4f14 5d1 6s2. Cu has a complete 3d shell only at the expense of a 4s electron and idealized Cu has an incomplete shell at [Ar]3d9 4s2. Yb, while featuring that in -La-Ac, does not feature that in -Lu-Lr. While you should note the former, you should also note the latter.

"La has no f electrons and can't be put into the f block" works if you go -La-Ac and doesn't, being just a deviation (an unprecedented one, sure), if you go -Lu-Lr. But absolutely analogously, "Lu has adds an electron past a complete f shell and can't be put into the f block" works if you go -Lu-Lr and is just a deviation (an unprecedented one as well) if you go -La-Ac.

An absolute analogy to me.--R8R (talk) 22:51, 2 January 2017 (UTC)

I retract my comments re unprecedented. Still, I keep thinking this is an absolute analogy if you do look both ways.--R8R (talk) 13:25, 3 January 2017 (UTC)

Comment: I've expanded our analysis, accordingly. Thank you! Sandbh (talk) 13:41, 3 January 2017 (UTC)

"He also relies on the fact that several well-known reference books show La and Ac in the d block." -- this one is more solid. Current usage is an argument for making it official.

Comment: We were being conservative. There is no book that I know of that justifies the placement of La-Ac under Y (maybe with the one exception of the Shriver & Atkins Q&A). Perhaps we should clarify our reasoning. Sandbh (talk) 05:16, 18 December 2016 (UTC)

Done. Sandbh (talk) 06:32, 19 December 2016 (UTC)

@Double sharp: In this section we say, "it is not a weakness of Lavelle's argument to allow Th ([Rn]6d27s2) in the f block but not Ac ([Rn]6d17s2), since there are low-lying 5f orbitals present in Th but not (as far as we know) in Ac." Writing here, Jensen says (p. 3): "both La and Ac have low-lying empty f-orbitals…" and gives these sources: L. Brewer, “Energies of the Electron Configurations of the Lanthanides and Actinides,” J. Opt. Soc. Am., 1971, 61, 1101-1111; and L. J. Nugent, K. L. Vander Sluis, “Theoretical Treatment of the Energy Differences Between fqd1s2 and fq+1s2 Electron Configurations for Lanthanide and Actinide Atomic Vapors,” J. Opt. Soc. Am., 1971, 61, 1112-1115. Are you able to check these out to see if they address the low-lying empty f-orbital situation in Ac? Sandbh (talk) 10:11, 2 January 2017 (UTC)

@Double sharp: I have these two papers if there is some way I could get them to you. Sandbh (talk) 05:08, 6 January 2017 (UTC)

PS I also have the "Spectra and electronic structures of free actinide atoms and ions" chapter, from The chemistry of the actinide and transactinide elements, 3rd ed., vol. 1 book if that would help.

PPS Looking through some of those papers it seems that fs² for La is ≤ 15.1 × 10³ cm^–1 whereas Ac fs² is 30.0. But I don't know where the cut off is between low-lying and not low-lying. Sandbh (talk) 06:18, 6 January 2017 (UTC)

PPPS So, if my physics is right, 30 × 10³ cm^–1 = 3.7 eV which seems a little high to me. At least La (≤ 1.87) has the prospect of being within the visible light range. Sandbh (talk) 06:47, 6 January 2017 (UTC)

PPPPS I see Th fds² is 7.8 × 10³ cm^–1 (0.967 eV), which is impressive. Sandbh (talk) 07:58, 6 January 2017 (UTC)

Right. I retract all of the above. I had a close look at Hamilton DC 1965, "Position of lanthanum in the periodic table", American Journal of Physics, vol. 33, no. 8, pp. 637–640, to see how he worked what constituted a low-lying f orbital, supplemented by the papers I mentioned above and the NIST tables. La does have low a lying f orbital; actinium not nearly so low lying but still low enough to be non-hydrogenic, and Th has the lowest of the three. I'll make some adjustments to our submission. Stop me if it sounds like I've gone feral. Sandbh (talk) 10:38, 6 January 2017 (UTC)

There should be an option for "Email this user" on the sidebar of my userpage, now that I finally remembered to confirm my email address. ^_^ The trouble is that low-lying is rather subjective as a term. In the case of Ac, from what I read about the actinides (you may even have shown me the papers) it isn't low-lying enough to cause substantial 5f contributions to the metallic bonding, while it is for Th, which would seem like a reasonable qualitative boundary. Double sharp (talk) 13:23, 6 January 2017 (UTC)

Comment: What Hamilton (1965) did was to work out the binding energy of the 4f level by subtracting the energy of the lowest level containing a 4f electron from the energy of the ionised atom. So, for H, and going by the NIST tables, H is ionised at 109678 minus 102823 for the lowest 4f level = 6855 cm^-1. In his article Hamilton does the same thing for Al, In, Cs and Ba, and their 4f level binding energies range from 6961 to 7430. This range covers what he calls "hydrogenic" f levels. For La he says ionisation occurs at 45293; NIST gives 44981. The fs² level for La according to Brewer (1971, p. 1106) is ≤ 15100 so the binding energy is ~30000. The same approach applied to Th gives a figure of ~50000, and ~ 11000 for Ac. So, Ac is still counted as non-hyrodgenic but there is not that much in it. I suppose we could say something like the following: In Th, at 50000, we can see a 5f contribution; in La, at 30000, things are getting a little ephemeral (although we have acknowledged their may be an impact), while in Ac at 11000, which is closer to being hydrogenic rather than non-hydrogenic, we are not aware of any discernible influence? Sandbh (talk) 01:57, 7 January 2017 (UTC)

Comment 2: Perhaps anothet way to look at this is to note that the 4f/5f level lies about 33% above the ground state in La, 71% for Ac, and 15% for Th. If low-lying = up to 33%; mid-way = 34 to 66%; and hydrogenic = 67%+ then La could only be regarded as marginally non-hydrogenic, Ac not at all, and Th definitely non-hydrogenic. Hamilton does not even use the term "low lying". He just says the binding energy in La is appreciably larger than in the other examples he gave. Sandbh (talk) 10:08, 7 January 2017 (UTC)

2016: Condensed- v gas-phase electron configurations

Generally a good point. Some comments, though:

"There are four exceptions." and "with just two exceptions, every last one of the lanthanides [italics present in the review]" are very different emotionally. Why so? This is meant to be a scientific paper and it should not play with emotions. The less contradictory "only two [no italics] lanthanides are exceptional" would suit better.

Done. Sandbh (talk) 05:20, 18 December 2016 (UTC)

"It also supports their chemistry" from what I've read during your discussions, the reality is, "if we assume that this version is correct, then things are non-contradictory." But we apparently do not exactly know or at least there is no unversal agreement on why this is their chemistry. Thus it cannot be used to prove the point. "If we pick -La-Ac, then chem is okay" is what we know. From this does not follow an opposite implication: "the chem looks like this could be an explanation, so it's -La-Ac."

Good. I've changed this to "It is also consistent with." I agree with you. We are not trying to prove the point; we are only observing that it appears to be consistent. I did check with an established and published chemistry professor and they agreed it was a plausible explanation. Sandbh (talk) 05:28, 18 December 2016 (UTC)

I guess it would be better if you explicitly said that. I read this part and I just about to complain at it, but then I saw this phrase in the review.

Comment: It now says, "It is also appears [italics added] to be consistent with their chemistry". Sandbh (talk) 00:43, 1 January 2017 (UTC)

"(In the case of Ac, the [Rn]5f17s2 configuration would be so high in energy that ionisation would actually happen first.)"

-- really? I have been shown the source, I just found it incomplete as it had little data on actinium. Does the idea of this not even raise your eyebrows?

@Double sharp: Thoughts? Sandbh (talk) 00:43, 1 January 2017 (UTC)

The NIST page I got this from gives levels for Ac⁺ as well, when 5f finally appears (it does not appear for Ac, apparently). Both are from the same source (Moore 1958). Double sharp (talk) 11:48, 1 January 2017 (UTC)

Thank you! Citation added. Sandbh (talk) 00:57, 2 January 2017 (UTC)

"If 4f is supposed to, by the Aufbau principle, be lower in energy than 5d, then why is 5d filled preferentially in La, with the [Xe]4f16s2 configuration only appearing at very high energies? (In the case of Ac, the [Rn]5f17s2 configuration would be so high in energy that ionisation would actually happen first.)

Here's info in that source for Lu I (i.e., Lu⁰). It mentions configurations of up to 6s²35f because the authors were able to measure it. As you can see, those last levels are quite close to complete ionization energetically, with difference for the last level being 100 cm^-3.

Here's the info that source has on Ac I. The lowest difference listed is about 8,000 cm^-3. Overall, this data should be difficult to obtain for actinium at great precision, as radioactiity comes into play.

Also, I still don't see how an system Ac⁺ + 5f¹ would be less thermodynamically favorable than Ac⁺ + an infinitely distant e.--R8R (talk) 15:37, 1 January 2017 (UTC)

Yet Rn⁰ has configurations like 6p⁵12d and 6p⁵10f listed too, from the very same author, even though it is even more radioactive, and the difference for the last level is even less than 100 cm⁻¹. Double sharp (talk) 16:05, 1 January 2017 (UTC)

Yes, indeed; just note that if you can run two difficult experiments, then of course you can run a third one, too. But it's still going to be extra effort and extra money put into it. (I don't know the technical problems of those exact experiments the author performed, granted.)

In the original document, which I checked some time ago and which you can check out now here, actinium was the last element and was the only one to have so few levels listed. The entry gives a sense that the data on actinium is incomplete:

The observation, description, and analyses of the emission spectra of actinium have been undertaken especially for inclusion here. The analysis is by Meggers, who has measured sad interpreted are and hollow-cathode spectrograms made at the Argonne National Laboratory by Fred and Tomkins. Thus far, 87 Ac i lines, with wavelengths ranging from 2968.82 A to 7880.10 A, have been classified as combinations among 40 energy levels. An estimated value of s PVi is entered in brackets in the table. This term has not yet been found because of the lack of infrared observations.--R8R (talk) 22:31, 2 January 2017 (UTC)

Why is it that, even with the sudden contraction of the 4f orbitals after La, we cannot avoid 5d occupancy in Ce even in the gas phase?

Additionally, chemical arguments also call into question the predictions of a Sc-Y-Lu-Lr table. We know that the 4f subshell is buried deeply in the core due to its low principal quantum number. So why are the lanthanides not predominantly divalent (with the exception of Eu and Yb), losing only the 6s electrons?

To answer these questions, we should look at the condensed-phase electron configurations, since they tend to be more chemically relevant than the gas-phase ones."

--the funny thing is, we have not got explicit answers to any of these questions. The gas-phase config questions are ignored as gas phase is not important anyway (so why did you even bother raising them?); and we only get "a plausible explanation" for oxidation states.

Comment: We include the gas phase questions because in thinking about the constitution of group 3 they occurred to us. To myself and Double sharp, the answers seem to lie in solid phase configurations. Sandbh (talk) 00:43, 1 January 2017 (UTC)

Still, it's not fair.

"Yet La and Ce are a little odd here." -- you see, I feel that "seemed good far but." Then I hear a couple of questions set to diminish the importance of -Lu-Lr version. To which there won't be answers: the thing that we don't find as too important is also imperfect. We won't discuss it, because it's not important anyway, we'll just note it's still imperfect. No precise answers, though: not important.

You see, I do think that having solid phase as more important than gas phase is at least one way to look at it and thus an argument. However, you're trying to give the argument the strengths it does not have.

"It also appears to be consistent with their chemistry" -- but we don't know if that's the reason, right?

Comment: I'd like to hear from @Double sharp: on this point. My thoughts about this are that if there was a black and white answer to the group 3 question the decision would've been made years ago.

In the absence of black and white information and theory we are trying to fit the data coming from lots of different sources and to blend this into some kind of consistent narrative. Think of Double sharp and I like symphony directors rather than deductive logicians. Bohr had to apply the same kind of multifaceted non-linear thinking in developing his thinking of the periodic table. He had his doubts, and if I recall correctly, he felt sorry for the lack of black and white reasoning underpinning his work but, in the end, he was shown to be right (and periodic table science benefitted enormously).

I want to write a reply to this.

First of all, I cannot help but feel this is familiar, as I have a similar feeling when writing element articles. I sympathize that.

I also understand your desire for a consistent story.

In fact, I am not all too well aware of your incentive towards this. If you, say, think that -La-Ac supporters are going to be underrepresented in the IUPAC comittee, then of course. (I remember seeing that the committee is headed by Jensen and Scerri, both -Lu-Lr supporters, but I don't know how many peoplу are even there in total and what opinions they have. I've decided to think that the two are gentlemen and wouldn't want to push their version over a committee that wouldn't question their words and that they would be interested at getting a good contest for their version, as that would strengthen legacy of their idea. Do you know anything I don't?)

I believe (please correct me if you have a slightest reason to think I could be wrong) that the committee will assemble and then the parties will begin to present their arguments. But if you will provide not a convincing but a consistent pro-La-Ac story, then surely someone can reply to that by providing a consistent pro-Lu-Lr story.

If I was to write this or to provide the answer as if I was that committee, I would focus on those point that don't have an immediate analog for the other version: consistent +3 state for lanthanides; comparably stable +2 for Eu and Yb and +4 for Ce and Tb; electron configurations in gas phase and solid phase; popularity of the -La-Ac version; (maybe a few more I forgot). Then I would say both version have merits given some preconditions and think which set of preconditions would be more suitable to display in the PT.

(all that said, by no means do I intend to prevent you from submitting the document, quite the opposite.)--R8R (talk) 22:00, 2 January 2017 (UTC)

Comment: Normally I respond to your edits in chronological order. That is to say I look at your latest comments and respond to them in the same order you posted them. I make an exception today because I want to think more about your earlier new comment.

I expect Jensen will support Lu-Lr. Scerri has supported Lu-Lr but has questioned the basis of some of Jensen's arguments. Scerri thinks there is an underlying truth to the organisation of the periodic table but so far this underyling truth seems to be too subtle for people to uncover (including him). Lavelle I expect will support La-Ac. Ball is respected writer for Chemistry World. I don't know which way he will go but, judging by his other work, I think he will look at the question with an open mind. Restrepo is, as far as I know, a mathematician who will let the numbers tell the story. Ohrström is the IUPAC Vice-President of the Inorganic Chemistry Division, and I don't know anything else about him. Overall I have no concerns about the task group and in fact think it looks very strong.

My motivation for writing a submission to the task group is to give them something to think about, hopefully in addition to what they themselves have been thinking about or have received from others. I wasn't going to do this originally because I couldn't see a way of spotting the difference between the two options. Like I said earlier I have switched back and forth between the two more than once. I think I changed my mind about writing a submission after reading about gas phase v solid phase electron configurations (partly inspired by Scerri) and some subsequent discussions I had with Double sharp. It then seemed to me that we could discuss and analyse all of the arguments we were aware of and that by weighing all of these up, reach a position. I was actually looking for as many arguments as we could find for each option, as my aim was to smash all the arguments together and see what was left. I don't care which option is stronger. Unfortunately, the La-Lr arguments, of which there are very many, were mainly very disappointing. I started with them hoping that we could find an answer in such a large amount of literature. I think if I had started with the La-Ac argument I would have given up writing the submission for lack of content. Starting with Lu-Lr was good as it prompted the thinking about the La-Ac arguments.

More to follow. Sandbh (talk) 10:29, 3 January 2017 (UTC)

Let me see if I can answer your earlier and later arguments, then I'll come back here. Sandbh (talk) 10:51, 3 January 2017 (UTC)

In about 12 hours or so. 13:50, 3 January 2017 (UTC)

Double sharp and I are attempting to provide a submission to the task group that considers both options. Sure, before we started, we both thought that La-Ac was a better option. But, instead of writing the La-Ac story, just as Jensen only provided the Lu-Lr story, we thought we would write about both stories. I was actually hoping that we would be wrong and that one or more of the Lu-Lr arguments would tip the balance the other way, but it hasn't turned out that way. There is no smoking gun that we can point to in support of one option or the other. There is only a contest of ideas. There are lots of Lu-Lr arguments but they don't tell an integrated story. The arguments don't withstand scrutiny. There are only a few La-Ac arguments but they tell an integrated story. They withstand scrutiny, as far as we can see (noting we still need to finish addressing some of your comments). When our submission is ready to be turned into a word document, if you feel that it does not provide a fair summary of the merits of each option then I would encourage you to consider writing to the task group about this. Actually, it may be better and easier for you to continue with the current review process, give Double sharp and I an opportunity to consider all of your comments, and then stand back and see what still concerns you and if that would be worth writing to the task group about. It's up to you. 01:40, 4 January 2017 (UTC)

Thank you for your response. I will reply only partially because I'm (agaij) short on time. I will be more time-flexible in late January, but it may be too late then. In fact, I'm currently writing from the subway.

I won't complain about your analysis to the working group. First of all, this is not very polite; politeness aside, still, I don't think I could figure something they couldn't (then there would come thw actual talk on whether your analysis deserves to be complained at, but I see no point in doing that for reasons I've just described). The goal to keep reviewing wasn't originally, but is now, about people around me in Wiki and critical thinking of theirs. As I said, I don't have a strong sympathy for either version because to make one, I would need to do a good analysis on a way I have described. I have not made up my choice and O acknowledge that, despite admitting I like -Lu-Lr better intuitively. However, I do think that you're looking one way but not the other, and I do want you try botg and see tge legitimacy of the other version. I do think I've made clear why the otger version is legit as a contestant at least. So is yours, of course.

Back to the issue. I would rather want you to get a picture that -Lu-Lr is fair to try. I'm not interested in converting you to this option---as I said, I'm not even sure if that would be right or wrong; I am genuinely not---I want you to recognize it as a (even though not necessarily the) valid version in part because you not doing so in part relies, as I see it, in you missing some ideaa to consider. And if you excluded the many minor points (dialumnides?) that blur out your vision, maybe you could see that -Lu-Lr is similar to -La-Ac, in a way at least.

As I said, a lot depends on your starting points. You can conclude that d block being a solid rectangle is a starting point you would need a strong force to overcome. Or that none is enougg, bam, it's -Lu-Lr. And that's not right nor wrong; that's a point that relies on an axiom. There are different outcomes for different axioms; your best job would be ti rationalize tge use of yours.

Anyway, I think that summarizes my thinking on tge problem nicely.--R8R (talk) 10:26, 5 January 2017 (UTC)

Thank you R8R. That was a good reply. I intend to now go into print but will think again about -Lu-Lr, in light of your comments, to see if we've missed anything. Sandbh (talk) 11:05, 6 January 2017 (UTC)

Is this not what the advancement of classification science is about? As long as a classification decision is beneficial to economy of description, to structuring knowledge and to our understanding, then make it. If our classification is subsequently shown to be less than optimal, then scrap it and replace it with a classification based on different shared characteristics. Sandbh (talk) 03:40, 2 January 2017 (UTC)

I don't think there is a "perfect" periodic table. There may be an "optimal" one based on what data we have but that will keep changing as more knowledge comes in. My idea of what "optimal" means for a periodic table is how much of the chemistry it fits. It doesn't have to logically explain it from first principles; that would be asking for too much, because we cannot even get the Aufbau ordering from first principles. Double sharp (talk) 04:05, 2 January 2017 (UTC)

"We further note that—in compounds—the trivalent state is rare in europium and limited for ytterbium (Gschneidner 1968, p. 13), consistent with their irregular electron configurations." this is also consistent with their regular configurations at mid-4f and end of 4f if we accept -Lu-Lr. What's your point?

Comment. At least in aqueous solution, +3 is the most stable oxidation state for all of the Ln including Eu and Yb. Sandbh (talk) 03:40, 2 January 2017 (UTC)

Generally, I think this is a big obstacle towards a decision. It seems to me that this oxidation states thing could be a very strong argument for -La-Ac. Indeed, it makes sense: the idealized configuration is [Xe]4f^n-15d¹6s², so they easily lose their three outermost electrons. It indeed is a plausible explanation, I absolutely agree. However, a "plausible explanation" does not imply that "the case for Sc-Y-La-Ac is quite clear." If this was the opinion shared by the chemical community as a whole, oxidation states would be very interesting. But now, I assume they are not (feel free to correct me) and thus do not yet serve as a strong argument. I also assume that this is an understudied topic in general?

Comment: Rather than saying "the case for Sc-Y-La-Ac is quite clear" we say that, "we submit [italics added] that the case for Sc-Y-La-Ac is quite clear." In other words, this is our opinion based on the observations we have discussed and and we leave it to the task group to see what they make of our submission. My impression of the literature is that it is (a) dominated by gas phase configurations; and (b) main group and transition metal chemistry. While the solid state configurations of the lanthanides have been known for quite a while, what relatively few treatments there are of the Ln focus on their gas phase configurations, and their configurations as ions (with the latter supporting La-Ac, as per the new Lu-Lr argument called, "2006: Characteristic electron configurations of ions").

I genuinely don't see how "the case for Sc-Y-La-Ac is quite clear" differs from "we submit that the case for Sc-Y-La-Ac is quite clear" given the context of a discussion and building arguments, rather than listing facts. Nevertheless, I don't see why you submit that the case for Sc-Y-La-Ac is quite clear. There is no argument in this subsection that wouldn't allow for a different interpretation.--R8R (talk) 15:37, 1 January 2017 (UTC)

Comment: Please see the comment I made above (about conducting a symphony v deductive logicians), in response to your concerns about fairness. We are making a judgement by weighing up or reconciling all the ‘for and against’ arguments and, on this basis, it is our collective opinion that the case for Sc-Y-La-Ac is quite clear. The IUPAC task group can make to this what they please, as I know they will! There is an argument in this section that we made which would allow for a different interpretation and that is the gas phase configurations. In our opinion, however, the weight of the discussion in the whole section favours Sc-Y-La-Ac. Sandbh (talk) 03:40, 2 January 2017 (UTC)

This actually made me think if it's not too early to call for a final decision to the IUPAC.--R8R (talk) 21:49, 30 December 2016 (UTC)

Comment: That is something the task group can make a call on, noting our other La-Ac arguments. Sandbh (talk) 00:43, 1 January 2017 (UTC)

Yes, sure; just sharing my thoughts.--R8R (talk) 15:37, 1 January 2017 (UTC)

Cool. Always welcome. Sandbh (talk) 03:40, 2 January 2017 (UTC)

Blocks

This is a good argument. I don't see why you have to dilute these good arguments with the bad ones I found in the previous section.

Thank you! I hope some of the the other arguments have become less bad. Sandbh (talk) 05:31, 18 December 2016 (UTC)

"Thus, eka-actinium [...] has more of a familial relationship to Ac than it has to Nh." Why are you skipping the possibility that 121 starts a new column?

Response: Because the radial collapse of the 5g orbital only happens later, at E125. (Did we say that? It not we should add it.) Double sharp (talk) 07:10, 16 December 2016 (UTC)

Done, by Double sharp. Sandbh (talk) 22:09, 19 December 2016 (UTC)

"In eka-thorium, the ground state is predicted to be [Og]7d18s28p1, but low-lying configurations involving the 6f level are predicted for the +1, +2, and +3 cations. [...] So, while the beginning of the 5g series remains as yet unexplored, the first four undiscovered elements do not appear to necessitate any revision of the concept of blocks at the moment, instead taking their places quite naturally below Fr, Ra, Ac, and thorium" this is inconsistent in relation to element 122 with "We think it plausible that the low-lying 4f levels in La may influence some of its properties. [...] Overall, we think the presence of any 4f influence, as a relatively low-order phenomenon, would only be a 'tipping point' argument."

Comment and question: There's two levels of f-involvement we are trying to distinguish here:

1. f-orbitals occupied in atom or ion (e.g. Th³⁺ is [Rn]5f¹, and 122³⁺ is similar);
2. f-orbitals unoccupied in all atoms or ions, and only weakly used in hybridisation, mostly exceptionally (e.g. La, or d-orbitals used in Ca, Sr, Ba)

How would you suggest we clarify this? Double sharp (talk) 07:10, 16 December 2016 (UTC)

Clarified, by Double sharp. Sandbh (talk) 22:09, 19 December 2016 (UTC)

I remember not wanting to go too harsh on -La-Ac arguments during my first go-over. I had heavily criticized -Lu-Lr comments; surprised to find myself in that position, I wanted to change the attitude. Apparently, there was lost of stuff I missed.

"After all, an f-block element must have non-hydrogenlike f orbitals that have been filled for it to deserve the name."

-- I want to raise your awareness on this: according to yourselves, "This may be associated with the presence in La of a low-lying nonhydrogenic (unoccupied) f orbital." I don't really follow. Why have you said that while arguing for -La-Ac?

Fixed: I removed the first mention of non-hydrogenlike. Sandbh (talk) 01:34, 1 January 2017 (UTC)

@Double sharp: Could you please address R8R Gtrs' comments in the rest of this section? Sandbh (talk) 05:04, 1 January 2017 (UTC)

"At the end of the block we have Cu, [Ar]3d104s1, with a filled 3d shell that can still be ionised."

-- this move of yours is called "substitution of notions." That's not how you look for the end of a block. (s-block never ends?) A block ends when that last electron of that block is added to the idealized configuration. That's the definition (at least the one I've been taught in school). Copper's idealized config is [Ar]3d9 4s2, zinc's is [Ar]3d10 4s2.

True, that was bad phrasing on my part. Yours is the one I was looking for and messed up in expressing. -La-Ac would argue for Yb having idealised [Xe]4f¹³5d¹6s² and Lu [Xe]4f¹⁴5d¹6s², while -Lu-Lr would argue for Yb having idealised [Xe]4f¹⁴6s² and Lu [Xe]4f¹⁴5d¹6s². Double sharp (talk) 16:13, 1 January 2017 (UTC)

"Just as 3d is only completed along with its upper 4s after Zn and thus the p block begins at the element adding electron to that full shell, at Ga, 4f is only completed along with its upper 6s after Yb and thus the d block begins at Lu. The fact that the element preceding Lu---Yb---at the end of the f block can use its f electrons for bonding is immaterial. So can Ba at the end of the s block, reflecting the not so great stability of a full shell configuration in an early block. The important thing is that the f block must end before a new electron is added to the filled idealized f orbitals with 14 f electrons and thus it has to have 14 columns. Since it ends at Yb, it must begin at La.":

-- see, you may get different results if you stray from the definition.

"Classification of an element in the periodic table is based on four steps:

1. Assignment to a major block based on the kinds of available valence electrons (i.e., s, p, d, f, etc.).
2. Assignment of the elements within each block to groups based on the total number of available valence electrons.
3. Verification of the validity of the resulting block and group assignments through the establishment of consistent patterns in overall block, group, and period property trends.
4. Verification that the elements are arranged in order of increasing atomic number as required by the periodic law."

"Jensen claims that criteria 1 and 2 do not lead to an unambiguous assignment in the case of La and Ac, but we find that they actually do." -- perhaps I missed it. I'm with Jensen on criteria 1 and 2.

"La is easily assigned to the d block because it only has 5d and 6s orbitals available for bonding. Lu, however, has a newly filled 4f shell, similar to Zn with a newly filled 3d shell" which is also not available for bonding. Jensen clearly said, "valence electrons." Why are you even implying there are 4f valence electrons at Lu? are 3d actually valence electrons for Zn?

See, this is the problem with his criteria that we point out. 3d are not actually valence electrons for Zn, and neither are 4d for Cd and 5d for Hg, but that seems to imply that group 12 is in the s-block. Even worse we find that 1s and 2p are not actually valence electrons for He and Ne, so they cannot seem to be allocated to any block. Clearly, that's not what the definition of a block means. Hence we added the caveat that when a shell is completed, its electrons may not actually be ionisable, and in fact usually are not (group 12 won't break the d-subshells and the noble gases need a lot of coaxing to surrender electrons). Double sharp (talk) 16:13, 1 January 2017 (UTC)

No, wait. You said you were not fighting his criteria ("See, this is the problem with his criteria that we point out"), that could make sense (maybe). You were saying that Jensen's arguments were actually implying a positive feedback for -La-Ac ("Jensen claims that criteria 1 and 2 do not lead to an unambiguous assignment in the case of La and Ac, but we find that they actually do"). They don't.

As for comparison with group 12: it only works if you already decided you go -La-Ac. If you're trying to judge without an aforehand opinion, then you must note the comparison is invalid if you go -Lu-Lr.--R8R (talk) 16:48, 1 January 2017 (UTC)

That's why we added our caveat after quoting him: "In the case of the last column of each block, i.e. He, Ne, Ar, Zn, Cd, Hg, Lu, and Lr, we would note that criterion 1 would have to be slightly amended to accommodate the fact that some "theoretically available" valence electrons may not actually be ionisable, but we do not think that this seriously affects our argument."

The comparison with group 12 works both ways. In the p-block and d-block, it is a normal thing that when the p- or d-subshell fills, it will be very difficult to remove electrons for it. This matches with the behaviour of Lu and Lr but not with that of Yb and No. Indeed, the latter pair acts more like Cu and Ag. Double sharp (talk) 15:38, 5 January 2017 (UTC)

"The trends obtained are quite valid, even more so than for Sc-Y-Lu-Lr, thus fulfilling criterion 3." -- at very best, questionable. You have mentioned many times that Sc-Y-Lu trends are often similar with those in most of the d-block. I find this somewhat wishful thinking.

"Indeed Fricke and McMinn (1976) wrote" -- that 1976 was 40 years ago. The word "indeed" doesn't really fit.

"which have received complete calculations by Eliav et al. (1998, 2002)"

-- this at least sounds like poor (or underdescribed) reasoning. what makes those calculations complete? are they more complete than the more recent Nefedov, Trzhaskovskaya and Yarzhenskii (2006) which was not described as complete? Surely many sources did elements 121 and 122 if at least four did 164. That multitude as a whole would be a valid source.

I don't like how you're assigning positions to 121 and 122 without even revieweing the other possibilities. This is not "we analyzed and, having consider all possibilities, we can say it's this one"; this is "we assumed this position for the topic in question and the theory seems not to contradict it." Sure, you don't have data as precise for elements beyond 122. But making the decision for elements 57 and 71 included reviewing some information on elements 56, 72, 58--70.

Did we not already explain why E121 and E122 can't be in the g-block – namely, the 5g shell is still hydrogen-like and unoccupied, similar to our main argument for not letting La and Ac in the f-block? Double sharp (talk) 16:13, 1 January 2017 (UTC)

You see, the whole argument "La has no f electrons and thus can't go to the f-block" is bad to me.

If we assume that we don't know (or as I would put, have not yet decided) what group 3 is made up of, then we only get the following facts: 1) La has no f electrons, and the 4f shell is not open yet; 2) the 4f shell ended at the ytterbium (as [Yb] = [Xe]6s2 4f14), and Lu adds an electron past a complete shell. Thus it cannot belong to the f block.

Both narratives are equally strong. Neither has preference.

Same would work here.

But then why doesn't it work this way: (1) Ca has no d electrons, and the 3d shell is not open yet; (2) the 3d shell ended at copper (as [Cu] = [Ar]3d¹⁰4s¹); and (3) Zn adds an electron past a complete shell and can't belong to the d-block? Double sharp (talk) 02:23, 2 January 2017 (UTC)

The difference is that we already decided that copper is the ninth 3d element becuase the other elements give a clear picture of it. The situation is not as clear with 4f and we don't know if Yb is thirteenth or fourteenth (you wouldn't be writing the submission otherwise). If we assume that our starting point is -La-Ac, then sure, the analogy works. However, I assume that our starting point is "we don't know," which leads me to the question: "what if we try -Lu-Lr, since that's one of the two possibly decisions?" And the analogy breaks.

I don't like the argument because while it is valid, it is incomplete but you sometimes turn a blind eye to that. Even incomplete analysis is valid; I've replied to it with that a few times (namely, in the beginning of this very subsetion) because I think it's okay to build arguments from both sides and then compare if we, rather than having a table and then fitting our reality into the narrative of that table, have a reality and only then invent a table as a way to characterize what we have.

However, you do call for analyses to be complete in the prevoius section so I assume the requisite should hold consistently. To summarize what I've said, your analogy here works only for -La-Ac and you should make clear that it does not for -Lu-Lr to keep your consistency.

(Similarly, its analog only works if you assume -Lu-Lr as your starting point. In general, I don't see anything in this argument that gives either version preference if you consider all of its analogs.)--R8R (talk) 21:20, 2 January 2017 (UTC)

"the first four undiscovered elements do not appear to necessitate any revision of the concept of blocks at the moment, instead taking their places quite naturally instead looking like they could take places below Fr, Ra, Ac, and thorium, but we're not [at least I am certainly not] sure if they will because we can't do full-phledged analysis because we don't have enough data yet."

that would be more fair.--R8R (talk) 22:57, 30 December 2016 (UTC)

That argument actually makes it reasonable to take out the whole section, because in the absence of more data about what the following elements are expected to do any placement of E121 and E122 would seem to be tentative. Double sharp (talk) 16:13, 1 January 2017 (UTC)

Chemical behaviour

"The chemical behaviour of Group 3, which is like that of pre-transition metals, favours Sc-Y-La." I will note that you are planning a submittion to IUPAC, and group 3 elements are transition metals under their definition.

Good. Yes, they are transition metals according to IUPAC but they behave chemically like pre-transition metals. Sandbh (talk) 06:00, 18 December 2016 (UTC)

No real opinion for this phrase: this is subjective anyway.--R8R (talk) 16:57, 1 January 2017 (UTC)

Noted. Sandbh (talk) 01:30, 2 January 2017 (UTC)

"Group 3 does not show the complex coordination chemistry that is characteristic of transition metals; they do not show multiple oxidation states; and they are more reactive and electropositive than any other transition metals, approaching the s-block metals in both properties." As for multiple oxidation states, group 4 is also dominantly +4. This weakens the oxidation states argument. Would love you to elaborate on those statements.

Noted. On group 4, Earnshaw and Harrington (1973 p. 52) say, "…this is the first group in which the really characteristic transitional properties of variable oxidation state, colour and paramagnetism are encountered." I think this suffices to distinguish group 3 from group 4. Sandbh (talk) 06:00, 18 December 2016 (UTC)

IMHO, Ti^III is so common that calling +4 dominant in group 4 the way +3 is almost unique in group 3 is not so true. Even though Zr(III) and Hf(III) reduce water, they are stable in simple salts like ZrCl₃ and HfCl₃. Double sharp (talk) 10:49, 18 December 2016 (UTC)

re ionicity, it is important to remember an ionic compound is not 100% ionic, and a covalent one is rarely 100% covalent. I found this post of mine in the DS's talk page archives: "I tried to check myself what the degree of ionicity of various oxides was and on a quick search I found one article suggesting that the ionicity in the K--Ca--Sc (same groups for periods 5 and 6) series falls quite linearly: 66--58--50 percent (other values for periods 5 and 6), with ionicity of TiO2, as I found elsewhere, being equal to ~0.5, so no real reason to say that ionicity falls so suddenly after group 3. Moreover, I've found this old Soviet document which says the big drops in ionicity come after groups 4 and 5 (for example, period 5, groups 2 through 6: 93--87--82--63--39 percent). Have you mentioned some data contradicting this information I see, or am I misinterpreting it?" I can't find corresponding data for the chlorides. Moreover, even if true, wouldn't that be a "tipping point" argument?

Comment: This one warrants a closer look. Sandbh (talk) 06:00, 18 December 2016 (UTC)

Ionicity may fall linearly, but it is theoretical. You can get a pretty accurate picture of SF₆ even if you assume 100% charge seperation, but at some point you get a discontinuous change from what you expect of an ionic compound (e.g. high mp and bp) and a covalent one (e.g. much lower mp and bp), no? Double sharp (talk) 11:45, 18 December 2016 (UTC)

Is it theoretical? It's real, or least that's how I am used to think about it (or I am wrong, I would want to be corrected).

You're going into vague terms. The change happens for many reasons and is not straightforward. For example, even CsCl doesn't have great the highest mp and bp. (I am not denying the idea, but I cannot take it for granted.) And this reasoning is poor: if mp and bp stem from ionicity, then you can't look and mp and bp and talk about ionicity. "A -> B" doesn't mean "B -> A."

Here's what Kleinberg, Argersinger and Griswold (1960, pp. 490–492) said about group 3 and 4 ionicity:

"The simple compounds [of the scandium group] are all ionic in nature. Thus, the oxides, which are all insoluble, are completely basic…As the ionization potentials [of the titanium family] indicate, the energy required for the removal of the four valence electrons is extremely large, and there is little doubt that in anhydrous binary compounds of the type MX₄…the binding is essentially covalent. Although the bonds between metal and oxygen in the dioxides…are considered to be largely ionic, these substances do not behave as completely basic oxides. Rather, the dioxides…have amphoteric characteristics…"

• Kleinberg J, Argersinger WJ & Griswold E 1960, Inorganic chemistry, DC Heath and Company, Boston

R8R, I read the old Soviet document the same way you do. I'm troubled by a few things. Y and La are found to have the same ionicity, yet La₂O₃ is experimentally found to be more significantly more basic than Y₂O₃. The CN of ZnO₂ is listed as 8 but is in fact 7 (at least that's what our article says). In any event, the ionicity of ZrO₂ appears to pass through some kind of amphoterism gate, as do the other group 4 metals, but this is not the case for group 3. The value given for Nb of 0.63 is questionable given it was based on CN 6 whereas Nb₂O₅ is largely CN 8.

I did some more research on ionicity (in particular, I found one article that has data for some (including La but not Lu) elements: 10.1021/jp063846j). Your concerns are valid. I retract that argument.--R8R (talk) 16:57, 1 January 2017 (UTC)

Thank you! Sandbh (talk) 01:32, 2 January 2017 (UTC)

In light of the above, what we say about ionicity seems to be fine, and I reckon it draws a nice distinction between group 3 and 4. Sandbh (talk) 05:01, 19 December 2016 (UTC)

"Given, as previously examined, that Sc-Y-La-Ac matches better with group 1 and 2 trends, whereas Sc-Y-Lu has been shown several times (including by Jensen), to more closely parallel trends in groups 4 to 10, we contend that the choice of Sc-Y-La-Ac over Sc-Y-Lu-Lr is clear." -- currently not as clear because the other section is poorly reasoned in some cases.

Comment: I hope things are getting better now, noting there are still a few o/s items. Sandbh (talk) 06:00, 18 December 2016 (UTC)

It's strange that a point that important (argueably the most important of all) is described in so few words. I would want those theses on coord chem and oxidation states and so on explained. We instead get "and so on," which is great to put after a few arguments and stop it there once you've run out of them.

Comment: I tend to think that since we're pitching to a highly knowledgeable project team no further explanation is warranted or required (and I want some of my RL back, not to mention time to ce Lead!). Sandbh (talk) 06:00, 18 December 2016 (UTC)

I tend to think that a human in general is weak and is prone to erring whenever you don't expect, so it's best to keep the review self-contained, as DS pointed above. Also, th _{(I wonder what I wanted to write here. My point is, if I can not understand it immediately, some others also can, even though I don't question their qualification on the issue over mine, and it's best to try to avoid that when the cost is not too high.)}--R8R (talk) 16:57, 1 January 2017 (UTC)

Comment: I've added some illustrative quotes to the end of this section. Sandbh (talk) 12:54, 3 January 2017 (UTC)

I don't imply to say there are no further arguments. But I explicitly state that more arguments are needed since chemistry is so important.

Noted. Sandbh (talk) 06:00, 18 December 2016 (UTC)

"Sc-Y-La-Ac further shows simple trends of increasing basicity and in this respect are much more like their leftward neighbours Ca-Sr-Ba than their rightward neighbours Ti-Zr-Hf." also "Sc-Y-Lu-Lr further shows a trend of basicity changing in its way and in this respect are much more like their rightward neighbours Ti-Zr-Hf than their leftward neighbours Ca-Sr-Ba," so what was the point?

Comment: I seem to remember that the oxide basicity sequence for Sc-Y-Lu is Sc < Y > Lu. Whereas the sequence for Ti-Zr-Hf is Ti < Zr < > Hf. Because the two sequences don't match in periods 5 and 6 I didn't say anymore. Sandbh (talk) 04:41, 2 January 2017 (UTC)

"The +4 state is too high to be ionic in group 4" -- now that we are explicitly dealing with the wide terms ("pre-transition", "transition"), we must note that +4 is not actually related to transition-ness. X₂O₃ are ionic among transition metals (I picked a few at {{Oxides}}; they all were).

Comment: We are simply comparing the chemistry of the elements, as guided by their common oxidation states, which is the normal practice. Even Jensen did this with his comparison of EN and ionisation energies. Sandbh (talk) 04:41, 2 January 2017 (UTC)

"While KCl, CaCl2, and ScCl3 are ionic compounds, the group 4 tetrahalides (Cl, Br, I) are volatile covalent liquids or solids." -- again, this is related to the value of the oxidation states rather than transition-ness.

Noted. We are attempting to demonstrate that the chemistry of group 3 is more like that of their leftward neighbours, than those to their right. Sandbh (talk) 04:41, 2 January 2017 (UTC)

In general, I do advise keeping it up to some neighobring groups, usually the range is 1 through 5. (I do think it's the smartest thing to do in our group 3 talk.) Now, however, you're extending the talk as you claim that group 3 are chemically pre-transition metals and prove that point. To do that, we need to commonly operate with terms "pre-transition" and "transition" as a whole as well as in particular. Your point I am currently responding to mostly depends on oxidation states, rather than transition-ness, at least it seems to me so, and thus it is not suitable for our current discussion re transition. To eliminate that contamation, I would suggest comparing group 3 to other +3 elements, perhaps iron/cobalt/etc. on one side and groups 13/15 on the other and see if there will be significant difference between the two and, if so, where the group 3 elements lean.

The difference between the current discussion, in which I suggest you consider other +3 elements, rather than typical groups 1 and 2 vs. 4 and 5 comparison is that you previously talked about periodic trends and now you talk about properties. Oxidation states greatly affect properties.

If I were you, I would probably not go into this vague "transition vs. pre-transition" topic. _{(Not suggesting you shouldn't, just giving my approach to this.)}--R8R (talk) 20:28, 2 January 2017 (UTC)

Comment: General chemistry is replete with references to the differences between main group elements and transition elements, and our use of the terms pre-transition and transition metals is simply a more focussed variation of this.

The chemical behaviour argument is based on coordination chemistry, reactivity and electropositivity (not necessarily related to oxidation states), increasing basicity and other periodic trends and, if you like, oxidation states. And the argument for the overall behaviour of group 3 is supported by two citations. (I have expanded one of these, and added two more quotes.) I see complementary similarity and periodic trend arguments in our analysis, which is fine. A good periodic table show both trends and similarities (Scerri E 1991, "The electronic periodic chart of the elements,'" Journal of Chemical Education, vol. 68, no. 8, p. 712).

We have not based the chemical behaviour argument mostly on oxidation states. Even if we did, so what? Oxidation states underpinned Mendeleev's extraordinary work in developing the periodic table, making predictions on this basis, and stimulating the further development of the periodic table. Sandbh (talk) 11:52, 3 January 2017 (UTC)

"While one can obtain aquated La3+, hydrolysis proceeds so far for Hf that HfO2+ is obtained instead." -- and this is related to the value of the OS, too!

Ditto, noting I've now mentioned Lu³⁺. Sandbh (talk) 04:41, 2 January 2017 (UTC)

But isn't it significant if the divide is between 3 and 4 instead of 2 and 3? Double sharp (talk) 12:12, 3 January 2017 (UTC)

"And so on." -- No argument among those listed gave me the reaction of "hmmm, there is some point it it," so I don't see what is hidden under these words.--R8R (talk) 16:57, 1 January 2017 (UTC)

Removed! Sandbh (talk) 04:41, 2 January 2017 (UTC)

Arguments compared

(I didn't check the tables, my possible comments are already written.)

"A pivotal consideration is that chemical behaviour of group 3, as either -Lu(Lr) or -La-Ac, generally resembles that of groups 1–2." the funny thing is that I have not drawn that conclusion from your text. I have for -La-Ac, but not for -Lu-Lr. I'll regard anything that follows from this as invalid unless I am corrected.

Thank you, and a comment: We'll have another look at what you said about the chemical behaviour of -Lu-Lr. Sandbh (talk) 06:04, 18 December 2016 (UTC)

Didn't we mention that all the lanthanides (not just La), and by extension also the late actinides from Cm to Lr, behave like trivalent pre-transitions (almost like the heavy alkaline earths Ca, Sr, Ba, and Ra)? If not, I mentioned this with a good source in Talk:Periodic table (R. Bruce King; may have been archived). Double sharp (talk) 11:48, 18 December 2016 (UTC)

Not sure if we mention it but the resemblance of the Ln to the pre-TM's is well recognised. Yes, King (1995, p. 289) is a very good example:

"[since] the f orbitals are relatively uninvolved in chemical bonding, lanthanide chemistry is predominantly the chemistry of highly electropositive metals in the +3 oxidation state, just as the chemistry of the alkali metals and the alkaline earth metals is the chemistry of highly electropositive metals in the +1 and +2 oxidation states, respectively.... In fact, in many respects the trivalent Ln³⁺ ions...functions like a trivalent version of the heavier alkaline earths, namely strontium and barium."

King RB 1995, Inorganic chemistry of main group elements, Wiley-VCH, New York, p. 289

I've read other sources that effectively say the same thing. Sandbh (talk) 03:13, 19 December 2016 (UTC)

I'll also add general comments re reasoning here. One is that you compare -La-Ac to groups 1 and 2, while -Lu-Lr is compared with groups 4 to 10. This is not very fair since the latter set is wider and the middle-point balance is shifted further from group 3 than is the case for -La-Ac. Consider limiting yourselves to groups 4 and 5.

Noted. We blame Jensen (and Chistyakov). Sandbh (talk) 06:04, 18 December 2016 (UTC)

To be fair, if or when groups 6-10 are different from 4 and 5, we should take R8R's point into account. Double sharp (talk) 11:47, 18 December 2016 (UTC)

I did a scan of this. We do it for Ionization potentials; Electronegativities; and 1985: Seven properties, and it doesn't make a difference. In other cases, groups 4 and 5 appear to behave the same as groups 6–10. (It's worth checking again though). Sandbh (talk) 05:28, 19 December 2016 (UTC)

Conclusion

The set of -La-Ac arguments is good. I have some pickings regarding presentation, however, but generally, the arguments themselves are valid.

Thank you! Sandbh (talk) 06:10, 18 December 2016 (UTC)

The set of -Lu-Lr arguments, however, suffers from unfair judgment and incomplete reasoning.--R8R (talk) 20:29, 14 December 2016 (UTC)

Thank you R8R! I hope I've explained the "substantiated facts" v "arguments up for debate" approach, and improved our reasoning, or shown promise to do so. Sandbh (talk) 06:10, 18 December 2016 (UTC)

Discussion

1. @Double sharp: R8R raises some good points re our presumption of a linear trend going down group 3; and our mentions of "not the full story" in some of the -Lu-Lr analyses.

On the first point, when it comes to group 3, is it fair to presume a continuation of periodic trends down the group rather than the kind of pattern given by -Lu-Lr?

Maybe the answer to this question is "no", it's not fair, presuming we start from the position that we don't know which one of La or Lu is a better fit under Y. We can however note that La fits better with the trends seen in groups 1 and 2 and that Lu fits better with the trends seen in groups 4–8+. Does my thinking seem right here? Sandbh (talk) 00:42, 18 December 2016 (UTC)

Given that a linear trend is the most common thing you will find in the table, it does not appear unreasonable to me to say that we expect one. When we talk about the lanthanide contraction (which starts at Ce with the 4f collapse, as I keep saying), we are talking about it as a second-order correction to the normal trends, which you see unadulterated in groups 1 and 17 most of all. Further, while in group 4 we have no choice (you can't put Ce under Zr, it's not even really tetravalent, so it must be the similar Hf), we have one in group 3 (La and Lu look similarly plausible as eka-Y at first glance).

Otherwise I agree with you, and have used that argument (namely, since group 3 behaves like a main group, it should probably follow the main groups). Double sharp (talk) 11:53, 18 December 2016 (UTC)

2. On the second point, if we were to leave out the "not the full story" content, Jensen's arguments would become a series of substantiated facts, rather than arguments up for debate. For example, Jensen argues that the trend in atomic radii going down Sc-Y-Lu is similar to that occurring in groups 4 to 8—which is a fact—and unlike that of Sc-Y-La, thereby supporting Sc-Y-Lu. Our analysis currently says: "True, but not the full story. The trend in atomic radii going down Sc-Y-La is instead similar to the trend seen in the group 2 and 1 metals, -Ca-Sr-Ba- and -K-Rb-Cs-." I tend to think that if Jensen presents his facts as arguments in support of -Lu-Lr then we're obliged to present our responses as fact-based counter-arguments, in the way that we've done. How do you see this? (I'm still thinking about this observation by R8R).

I've changed the language so that instead of saying, "not the full story" it now says, "Factually correct but incomplete as an argument". I think this is a less emotive way of saying the same thing. Googling around, it seems to me that a good argument is one that makes its point and considers alternative points. See, for example, Damer TE 2013, Attacking faulty reasoning (p.  235): "A good argument should provide an effective rebuttal to all serious challenges to the argument and the positions it supports. This is the most neglected feature of most arguments." I don't think it's necessary to "attack" an argument, but noting that it doesn't consider the other side, sure. And I guess the fact that Jensen didn't do this contributed to a low take up rate of his proposal for group 3 as -Lu-Lr. Sandbh (talk) 23:22, 17 December 2016 (UTC)

I like this! Double sharp (talk) 11:53, 18 December 2016 (UTC)

Situation report

1 Jan

@Double sharp and R8R Gtrs: On day one of 2017 here is where things are up to:

R8R Gtrs feedback
I've addressed all feedback by R8R Gtrs, or pinged Double sharp.

Items outstanding for Double Sharp (echoes from the sea of pings :)
1. 1965: Similarity of Lu with Sc and Y---need to hear from you if you're OK with my "In some other parts" edit. Look for the ping at the end of the talk page section.

I added a brief mention that groups 1 and 2 were most germane among the other groups with linear trends (17 and 18 have only nonmetals, at least among the elements you can see, so not including tennessine and perhaps astatine too). Other than that it is fine. Double sharp (talk) 11:44, 1 January 2017 (UTC)

2. Conduction band structures---need to hear from you as to how you see my comment in response to R8R's feedback on atypical conduction band structures. There is a ping about half way done the talk page section.

Responded. Double sharp (talk) 11:45, 1 January 2017 (UTC)

3. 2016: Condensed- v gas-phase electron configurations---need to hear from you re R8R Gtrs' comment on actinium. There is a ping about half way done the talk page section.

Responded. Double sharp (talk) 11:49, 1 January 2017 (UTC)

Note 1: Follow on discussion between R8R Gtrs and Double sharp ongoing. Sandbh (talk) 01:28, 2 January 2017 (UTC)

::Note 2: Request for comment re fairness, symphony directors v deductive logicians, etc Sandbh (talk) 03:46, 2 January 2017 (UTC)
4. Blocks: A ping I added a little while ago asking you to please address R8R Gtrs' comments here.

Still thinking about how to address this one most effectively; probably will be sorted tomorrow. Double sharp (talk) 12:11, 1 January 2017 (UTC)

I've been thinking about this one too but am no closer to a solution. I keep defaulting to Lavelle's pair out of place argument as a determinant of the start of the f-block. Sandbh (talk) 03:31, 4 January 2017 (UTC)

@Double sharp: I've had a go at adjusting the first three paragraphs of the analysis to see if that may help you. Sandbh (talk) 12:49, 4 January 2017 (UTC)

5. A pair out of place. A ping to see if you can check out a couple of refs on the low-lying f orbital status of Ac. Sandbh (talk) 10:13, 2 January 2017 (UTC)

Items outstanding for R8R Gtrs

1. 1984: Formation enthalpies of Lu and Lr halides---new request for information about what a Russian source says.
2. 2006: Characteristic electron configurations of ions---please let us know what you think of this new Lu-Lr argument.

Thank you Double sharp and R8R Gtrs. Sandbh (talk) 05:40, 1 January 2017 (UTC)

6 Jan

I believe we're good to go now. There'll still be time for any last minute updates but, subject to one more think about -Lu-Lr, I'll go to print shortly. Sandbh (talk) 11:15, 6 January 2017 (UTC)

Item	Status	Comments
Cover sheet	Ready	Recently updated
Lu-Lr arguments	Ready
1958: Electron configurations	Ready
1965: Similarity of Lu with Sc and Y	Ready
1967: f-character of La	Ready	R8R to assess tipping point arguments? (Not required R8R unless you have time)
1982: Separation groups	Ready
Irregular electron configurations	Ready
Electron configuration analogy	Ready
Ionization potentials	Ready
Atomic radii	Ready
Ionic radii	Ready
Redox potentials	Ready
Electronegativities	Ready
Melting points	Ready
Crystal structures (elements)	Ready
Crystal structures (oxides, chlorides, various intermetallics)	Ready
Excited state spectra	Ready
Superconductivity	Ready
Conduction band structures	Ready
1985: Seven properties	Ready
2005: Physical properties	Ready
2006: Characteristic electron configurations of ions	Ready
2011: Triads	Ready
2012: Split d block	Ready
2013: Carbonyls	Ready
2014: Lu as a transition metal	Ready
Name matches position	Ready
2015: Dimer spectroscopy	Ready
Relativistic contraction of 6s shell	Ready
Aluminide dimers	Ready
Heat of vapourisation	Ready
Ionization energy of Lr	Ready
La-Ac arguments
Six properties	Ready	R8R to add comment on historical note? (Not required R8R, unless you have time)
1974: Integrity of the f block	Ready	Any response to last post by Sandbh 4 Jan
2004: La-Ac remains the most popular form	Ready
2008: A pair out of place	Ready
2016: Condensed- v gas-phase electron configurations	Ready	(1) Actinium discussion R8R/DS to be finalised (I think I can resolve this one now) (2) Any response to last post by Sandbh 4 Jan
Blocks	Ready	Discussion R8R/DS to be finalised? (I think this one is fine now)
Chemical behaviour	Ready	Any response to last post by Sandbh 3 Jan
Arguments compared	Ready
Conclusion	Ready

Sandbh (talk) 02:21, 5 January 2017 (UTC)

8 Jan

Hi @Double sharp: I've attempted to expand/address (a) the rectangle v asymmetry thing in 2012: Split d block; and (b) Jensen's criticism of Lavelle's support for Th in the f block but not Ac, in 2008: A pair out of place. Comments welcome. My plan ^_^ is to go into print by NLT tomorrow. Sandbh (talk) 00:58, 8 January 2017 (UTC)

9 Jan

Going to print, now (o' happy days) ^_^ Sandbh (talk) 02:11, 9 January 2017 (UTC)

10 Jan

Proof copy of the Word.doc printed off just now (34pp + 1 cover page). Sandbh (talk) 02:07, 10 January 2017 (UTC)