What does metal gas look like microscopically?

[SpY]]

I was thinking of a substance I could use to describe particle arrangement in solid/liquid/gas phases to school kids, and after realising water would be bad to use (since the liquid is denser than solid), I thought of using metal elements since they can exist in all phases.

Basically the question is, what does metal gas look like microscopically?

For example Mercury - with relatively low boiling/melting points and high vapour pressure - exists as a liquid at room temperature because its 6s electrons don't mix very well (something like that). But what does Hg vapour look like microscopically - are the gas particles just Hg atoms, or does the air mix with grains of Hg (clumped atoms) or dare I say Hg forms some bond with itself? I'm quite sure metallic bonding breaks down for gases, and that metals do not generally bond covalently. I know the Mercury cation is diatomic (Hg_2 2+), but is Mercury gas then just diatomic ions? Diatomic metal gases...

Lastly, because of anomalies like hydrogen bonding in water and expansion under freezing (Beryllium i think), is it incorrect to say particles are more tightly packed in a solid than liquid?

 

[snorkack]

Basically the question is, what does metal gas look like microscopically?

For example Mercury - with relatively low boiling/melting points and high vapour pressure - exists as a liquid at room temperature because its 6s electrons don't mix very well (something like that). But what does Hg vapour look like microscopically - are the gas particles just Hg atoms, or does the air mix with grains of Hg (clumped atoms) or dare I say Hg forms some bond with itself? I'm quite sure metallic bonding breaks down for gases, and that metals do not generally bond covalently. I know the Mercury cation is diatomic (Hg_2 2+), but is Mercury gas then just diatomic ions? Diatomic metal gases...

That varies. Mercury vapour is certainly monoatomic at low pressures. I think that alkali metal molecules are known (diatomic) but am not sure how widespread they are above the boiling point.

Lastly, because of anomalies like hydrogen bonding in water and expansion under freezing (Beryllium i think), is it incorrect to say particles are more tightly packed in a solid than liquid?

Yes. But there is, I think, a relatively simple explanation for the anomalies.

Look at how you can pack spheres densely.
There is only one densest way to pack circles on a plane. Hexagonal packing, where each circle touches 6 equally nearest neighbours. Now, the densest way to pack spheres in space is to stack such hexagonal layers above each other so that each sphere rests between 3 neighbouring spheres on the layer below, and supports 3 spheres on the layer above. Depending on how the layer below and layer above are placed relative to each other, this is either hexagonal or cubic structure; either way, each sphere has 12 nearest neighbours.

This is fine for crystals held together by the non-directional metallic bonds or van der Waals bonds. But not for covalent bonds. No covalent compounds have an atom with 12 bonds.

For example, look at graphite. Also hexagonal layers. But carbon does not form 6 covalent bonds to 6 carbon atoms - it forms only 3 bonds to 3 nearest neighbours, so the other 3 directions, in the middle of the hexagons, are holes. Also the distance between graphene sheets is much bigger than the distance between neighbours along bond. So graphite is full of holes.

Diamond has 4 bonds, and is much denser than graphite, but still not 12 bonds. Diamond is still full of holes, and so are silicon, germanium and gray tin.

If a substance which already is densely packed, like most metals, is melted, the atoms need space to slide past each other, and the only way to make such space is by expanding the packing. If a network of directional bonds is melted, however, then because that network is full of holes, breaking the bonds allows to fill some of the holes, so the atoms/molecules can slide past each other, and still contract because filling the holes leaves more free space than is needed for the liquid to flow.

Does this expanation make sense?

 

[gadong]

Students often store the idea that vaporization is a consequence of boiling or sublimation at a specific temperature and pressure, even though this conflicts with their day to day experience.

All metals have finite vapour pressures, and thus exist to some extent as gases at room temperature.

The vaporization processes is determined by energy/temperature considerations. It is easier to extract an atom from the surface of a metal than a molecule, so sublimation is mainly by evaporation of atoms. However, if two metal atoms meet in the gas phase they will bond with a high probability to form dimers, trimers etc.

In a closed container, a dynamic equilibrium is eventually reached, whereby for every metal atom that evaporates or sublimes from the condensed phase, another one is condensed from the gas phase (if the partial pressure is high enough).

At low enough partial pressures, the metal will exist purely as a gas. Mercury in an energy-saving bulb is an example of this (full of mercury, yet no liquid condenses out because the partial pressure of mercury is too low).

Real metal samples do not behave as the pure elements, due to surface contamination by the atmosphere. One atomic layer of oxide etc. is enough to modify behaviour completely.

In the lab, metal atoms (and clusters of metal atoms) are routinely produced and used under vacuum conditions for a range of experimental purposes. The easiest method is to bombard a metal target with energetic particles (e.g. argon ions). The ion beam erodes the target in a process known as sputtering, and ejects metal atoms and clusters from the target (a process known as sputtering).

 

[SpY]]

Thanks for the replies. snorkack, I understand what you mean about densest sphere packing of particles. My question was how does this change, and differ, in the transition from solid to gas? My original understanding (or rather misconception) is that particles in the solid move away from each other, creating tiny gaps - which is obviously wrong since liquids are generally incompressible. It's different to say that melting is the creation of defects in the sphere packing, because in that process the spheres are sliding past each other out of their rigid packing into an irregular arrangement. That irregularity is the only difference in the arrangement of particles of a solid and gas. Am I right so far?

gadong, I follow your post on vapour pressure and appreciate your info on how it's done in a lab. My question goes back to what snorkack said about Mercury vapour being monoatomic - is gaseous Mercury above its boiling point also monoatomic? What about alkali and Earth metals? When sodium vaporises in a vapour lamp we get beams of monochromatic light, which implies excitation cycles (and thus ionisation) - are the gaseous sodium particles then just monoatomic ions?

At high enough temperatures and low enough pressures hydrogen is monoatomic, can the same reasoning be applied to metals? Even though hydrogen is ordinarily diatomic by covalent bonds and metals are networked by metallic bonds.

 

[gadong]

All atoms can associate to some extent as dimers (diatomic molecules) in the gas phase. Most of these interactions are strong enough to be classified as chemical bonds, but there are many cases of van der Waals molecules that involve weak, long-range forces and only have appreciable populations at cryogenic temperatures. By population, I mean the fraction of the material that adopts that form (e.g. does it exist as Hg2 or Hg atoms?).

I compiled a list of metal dimer binding (dissociation) energies known in 2001, at which time all alkali metals Li-Cs were thought to form stable dimers with BEs of 0.22 eV per atom or greater. This is quite large and means that the dimer form will predominate in the vapour.

The alkaline Earth metals are either van der Waals molecules, or borderline cases. I had to check on Hg (http://pubs.acs.org/doi/abs/10.1021/jp952807x?journalCode=jpchax) which turns out to have a BE of 0.02 eV per atom (another van der Waals). These elements will consist of a mixture of dimers and atoms in the gas phase (with the atoms becoming more prevalent as temperature increases).


The above comments assume that thermodynamic has been reached, but in most practical applications (e.g. sodium arc lamp) this is unlikely.

At the high temperature limit (which will differ for each dimer), the atomic and dimer forms will be equally probable, meaning that there will be two atoms in the gas phase for every dimer.

 

[snorkack]

All atoms can associate to some extent as dimers (diatomic molecules) in the gas phase.

Not true. Helium 3 atoms cannot.

 

[gadong]

Not true. Helium 3 atoms cannot.

Apparently that is correct - thanks for pointing it out.