User:Double sharp/Metallicity

I have been thinking about Sandbh's proposed new categorisation on and off, so here are some disorganised thoughts in the process of blossoming.

Let's look at period 3, because it is pretty much the representative set of elements at the head of each group (proceeding with caution, since period 2 is anomalous and period 1 even more so). If a categorisation is going to make sense anywhere, it had better do so here. What do we see across the period?

Of course, effective nuclear charge decreases from Na to Ar, and hence electronegativity and first ionisation energies increase. The melting points go up in the metals depending on how many valence electrons are delocalised (hence Na < Mg < Al), and up in the oligoatomic nonmetals (you know what I mean) depending on the size of the electron cloud and hence magnitude of dispersion forces (S8 > P4 > Cl2 > Ar). Si is intermediate, having a giant covalent lattice (but before we jump on that as a property of metalloids, the same is true for C and Se on the "nonmetallic side", and α-Sn on the "metallic side"). The place where the metalloid Si is clearly distinguished from the others is in electrical conductivity: Na, Mg, and Al are metals with mobile electrons, and P, S, and Cl have negligible conductivities due to the lack of mobile charge carriers, but Si is a semiconductor. Now let's see what this implies for the compounds.

Because of the electronegativity differences decreasing as we go to the right, the bonding of the oxides change from ionic to covalent. Now the melting point trend of the oxides nevertheless starts by going up from Na2O to MgO, because of the increased lattice energy afforded by the smaller ionic size and higher charge of Mg2+ compared to Na+. But this cuts both ways, because this also polarises the oxide anion and causes increased covalent character. So the trend then goes down through Al2O3, and we reach a giant covalent structure at SiO2 (sort of "intermediate" between covalent molecules and ionic lattices), before falling down to the more and more low-melting P4O10, SO3, and Cl2O7. (To be "fair" I am comparing the elements in their group oxidation states.)

The "amphotericity" line cuts through period 3 at Al (reacting with HCl and NaOH), one step to the left of the physical "semiconductor" line at Si. To this end the general principle that metallicity drops off earlier in chemistry than in physics is sound. Nevertheless I have a nagging feeling that Sandbh's proposal focuses too much on chemistry.

For one, most aquated ions are actually noticeably acidic; the ones that aren't are for truly huge cations with very low charge densities (the champion of course being Cs+). Even [Mg(H2O)6]3+ is a noticeably weak acid; a solution of that is slightly acidic (pH 6.5 if you make it from dissolving MgCl2, which has noticeable covalent character too), grabbing a proton from nearby water molecules. [Al(H2O)6]3+ hydrolyses much more (pH 3 from AlCl3). But you can look at this as hydrolysis becoming more and more complete as oxidation state increases – which is also true in the transition metals. One can analyse SiO2·2H2O as "silicon hydroxide" coming from the complete hydrolysis of "Si4+", and this happens even more when we go to PCl5 forming H3PO4 outright (don't think I need to mention sulfur here, because it messes up the group oxidation state comparison). So already not only Si, but also Al and Mg, are displaying weakly nonmetallic behaviour. Aluminium in particular is quite clearly amphoteric and the pH value you end up with doing this is closer to that you get when doing it with the chlorides of Si and P than for Na and Mg. But I think calling aluminium a weak nonmetal is really going too far.

And the reason? Al is actually highly, highly electropositive. Sc is much the same, hydrolysing significantly in water as well, despite being a row further down. It is also highly, highly electropositive. Not only that, its physical properties are those of a first-class metal!

Now what saves polonium, for instance? And what about some of those transitions? (case study: I dunno, tungsten?)