Drafting the RFC
Should the default composition of group 3 on Wikipedia be changed from Sc-Y-La-Ac to Sc-Y-Lu-Lr?
- Summary
Currently, on WP, group 3 on the periodic table is by default drawn as Sc-Y-La-Ac, with lanthanum (La) and actinium (Ac) as the heavier congeners of yttrium (Y). This was decided in the previous RFC, based on it being interpreted as the dominant form in the literature, as well as some science-based arguments. The previous default form had been Sc-Y-Lu-Lr, with lutetium (Lu) and lawrencium (Lr) in those positions.
I argue that this was a mistaken decision because:
- its status as the dominant form in the literature is actually quite doubtful;
- the most relevant sources (the ones focusing on the group 3 question, rather than just baldly presenting a periodic table) for the most part support the Sc-Y-Lu-Lr arrangement.
- the science-based arguments for Sc-Y-La do not, in fact, stand up to scrutiny.
As a party formerly on the other side of this issue, who then proceeded to change his mind once new evidence came in, I think I have some claim to being science-based in this holy war. ;)
I am writing this conversationally, both because I want to avoid having tempers run high (this is not the time to give Fëanor's speech before the Noldor, it's supposed to be scientific and civil discussion), and also because I don't want to lose people who are not 100% conversant with the chemistry involved. The beginning of each heavy d block row is somewhat of a backwater, speaking chemically. So, come with me and we will go, where the rocks with thorium glow. Follow me! ;)
- Historical
When Mendeleev first drew a periodic table in 1869, and left the rare earths in their own groups for lack of a better way to accommodate them, perhaps he did not quite imagine the length and scale of the holy war that has been going on ever since lanthanum was first placed below yttrium.
Almost from the start, though, there has been an opposing campaign maintaining that this placement is wrong, and that lutetium should take that place.
And this is far more than a fringe campaign. Many glorious names from the past have gotten behind it:
(quote Landau and Lifschitz)
But that doesn't look like the periodic table from my textbook!
Actually, those textbooks often have a little inconsistency.
Firstly, many of them show the asterisk within the cell [explain as DePiep did]. Therefore they're literally speaking not actually Sc-Y-La tables.
Secondly, there is a more insidious self-contradiction within these. You see, the Sc-Y-La table makes the statement that after Cs (caesium) and Ba (barium) fill the 6s orbital, La (lanthanum) fills the 5d orbital. (On such a table, it is a d block element). And then Ce (cerium) through Lu (lutetium) intervene and fill the 4f orbital, before a return to normalcy where Hf (hafnium) through Hg (mercury) fill the rest of the 5d orbital. So instead of "6s, 4f, 5d, 6p", the Madelung rule corresponding to such a table should be "6s, one 5d, 4f, nine 5d, 6p". That's not something I made up, some Sc-Y-La-Ac authors say it outright.
(Silberberg)
You'll look in vain for a Madelung rule graphic that is altered to fit that statement. So we have a bizzarro situation where a lot of authors show a Sc-Y-La table, say that the Madelung rule is the idealised sequence of electron filling that the table reflects, and present that rule in a form that corresponds to the Sc-Y-Lu table instead. Go figure.
In fact Silberberg is even inconsistent with his group 3 diagram, in which both the La-Ac and the Sc-Y-Lu-Lr column have a 3. What he means by this is not clear. Maybe he thinks that Sc-Y-La/Lu-Ac/Lr are all group 3 elements. Or perhaps he thinks Sc-Y-La-Ac are the only ones, and he's just duplicating the "3" because those are scattered across two columns. In any case, this is hardly grounds for another group 3 option with bifurcation, especially since in his other tables he shows another normal Sc-Y-La*-Ac** that presumably is meant to imply that La and Ac complete group 3 but literally implies that the rest of the lanthanides and actinides belong there too.
It is, as an aside for now, not even true chemically that there is a 5d electron hanging up. It certainly makes a pedagogically easy explanation for why the lanthanides would like to be trivalent, with 5d6s2 above a [Xe]4fn core. The only trouble is that it is dead wrong:
[Glenn T. Seaborg]
But that's another story which we will get to in the section on scientific arguments. My point is just: does it make sense to cite textbooks as our main source for why group 3 should be Sc-Y-La-Ac, when:
- from the perspective of a first course in chemistry, these elements are actually pretty far down the radar, and not the focus of those sources (*); and
- they're not even being totally self-consistent?
(*) One astonishing exception is Wulfsberg, whose general approach tries to underscore the generalisations that work throughout most of inorganic chemistry, and as such has no qualms giving f block elements in his basic exercises. But he uses Sc-Y-Lu. ^_^
But this is the periodic table, the icon of chemistry! Surely IUPAC has an opinion?
Well...yes and no.
The latest version of the IUPAC periodic table dates from 2018 and has a "compromise" group 3, where the spaces below yttrium are filled with placeholders reading "57-71 lanthanoids" and "89-103 actinoids". That implies that all those elements are in group 3.
This is very much not a majority view, and in fact IUPAC has an ongoing project to decide this very question. It lists two alternatives:
"This project will deliver a recommendation in favor of the composition of group 3 of the periodic table as consisting either of
the elements Sc, Y, Lu and Lr, or the elements Sc, Y, La and Ac."
These are the two self-same alternatives that are being considered here; Sc-Y-*-** (placeholders) is not one of them, and so far no recommendation or partial result has been delivered. So we are now in a weird position where the official IUPAC table makes a statement about group 3 that we know for sure is not going to be the option selected.
In any case, we certainly can and do ignore IUPAC when most relevant sources do that. Witness how we're calling the elements 57-71 "lanthanides", not "lanthanoids", because that's what everyone in rare earth chemistry does.
So what do people think now?
Well, the thing is, if you look at the sources that consider the group 3 issue, they overwhelmingly support Sc-Y-Lu. And they've been doing that for decades, always with strong scientific arguments.
It is true that they have not gotten a total victory. But the fact that 1 in 6 tables now have followed suit is not nothing. This includes some well-known textbooks and resources, such as
- Clayden (Organic Chemistry)
- Wulfsberg (Inorganic Chemistry)
- WebElements [which, in the 2000s, was basically the periodic table on the WWW]
As a sobering comparison, one should think about the status of explanations of hypervalence. It has been known since the 1990s that compounds like SF6 really do not have d orbital hybridisation going on on the sulfur atom, and that there are better ways to explain such molecules that fit the facts. So, of course, all the textbooks followed suit—
[quotes]
Oh. Hmm.
So, do we want to follow the beginning-level textbooks, for which the group 3 issue is very far down the list of priorities (how many beginning chemistry courses cover the lanthanides and actinides in detail, anyway?), or follow the most relevant sources
- Scientific arguments
Delayed collapse
But ground-state configurations in the gas phase aren't relevant!
Yes, they are not. I agree!
[quote Seaborg, Jorgensen]
Chemical similarity