Two atoms gave two orbitals — what does a trillion give?
Everything in this rung has rested on one promise of the LCAO recipe: combine N atomic orbitals and you get back exactly N molecular orbitals — no more, no fewer. Two hydrogen 1s orbitals gave one bonding level (sigma) and one antibonding level (sigma*). The conservation of orbital count is not a quirk of small molecules; it is a law that holds no matter how many atoms you throw together. So let us be bold and throw together a lot.
Picture a chain of identical lithium atoms, each donating one 2s orbital. With two atoms you get two levels, one bonding and one antibonding, split apart by some energy gap. Add a third atom and you get three levels: one fully bonding (every neighbor in phase), one fully antibonding (every neighbor out of phase), and one in between that is neither. With four atoms, four levels. The pattern is relentless: N atoms give N levels, all squeezed into roughly the same energy window between the all-bonding bottom and the all-antibonding top.
When the rungs get too close to see
Now do the arithmetic that turns a molecule into a crystal. A speck of lithium metal you can barely see contains on the order of 10^20 atoms, so its 2s orbitals give 10^20 molecular orbitals — and they all have to fit inside that same finite energy window. Divide the width of the window by 10^20 levels and the spacing between neighboring levels is so vanishingly small that no experiment could ever resolve one rung from the next. The discrete staircase of molecular orbitals has, for all practical purposes, become a smooth continuous ramp. We call that ramp a band.
atoms (N) levels picture
2 2 sigma | sigma* (clear gap)
4 4 . . | . . (rungs visible)
8 8 :::: | :::: (getting crowded)
~10^20 ~10^20 [######### BAND #########] (a continuum)
bottom of band = all neighbors in phase (most bonding)
top of band = all neighbors out of phase (most antibonding)Notice that nothing new was invented here. A band is not a different kind of physics from a molecular orbital — it is just an enormous family of molecular orbitals, ranging continuously from the most bonding combination at the bottom to the most antibonding one at the top. The bottom of the band is exactly the all-in-phase orbital you would draw for a tiny molecule; the top is exactly the all-out-of-phase one. The middle is densely filled with intermediate mixtures. This is the single most important sentence in the guide: a band is MO theory at scale.
Filling the band: why lithium is a metal
A band is only a stack of empty seats until you pour in the electrons. The rule is the one you already know from filling a molecular orbital diagram: electrons enter from the bottom up, two per level (opposite spins), lowest energy first. Each lithium atom brings one 2s electron, so N atoms bring N electrons into a band of N levels. Two electrons fit per level, so the electrons fill the band exactly halfway up. The energy of the highest filled level at absolute zero has a name worth knowing — the Fermi level.
Here is why a half-filled band makes a metal conduct. Just above the Fermi level sit empty orbitals separated from the filled ones by essentially zero energy — remember, the rungs are spaced 10^-20 of a window apart. So the faintest nudge from an electric field can promote an electron into an empty level where it is free to drift through the crystal. The electrons at the top of the filled region are mobile precisely because there is somewhere infinitesimally close for them to go. This delocalized sea is the honest quantum version of the metallic bond you first met as 'a sea of electrons' many rungs ago.
Gaps between bands: insulators and semiconductors
A real solid does not grow just one band. Each kind of atomic orbital spawns its own band — the 2s orbitals make an s-band, the 2p orbitals make a p-band, and so on — exactly as the 2s and 2p atomic orbitals gave separate sets of MOs in a diatomic. Sometimes these bands overlap; sometimes a clean stretch of forbidden energy, a band gap, separates a lower filled band (the valence band) from an upper empty band (the conduction band). The size of that gap, set by how strongly the atoms interact, decides almost everything about how the material behaves. This is the core idea of the band theory of solids you will meet head-on in the next rung.
The three great classes of solid fall straight out of this picture. A conductor has a partly filled band (lithium) or two bands that overlap (magnesium), so there is always somewhere for electrons to move. An insulator has a full valence band and a wide band gap — diamond's gap is so large (about 5.5 eV) that ordinary heat or voltage can never lift an electron across, so no current flows. A semiconductor like silicon has the same arrangement but a modest gap (about 1.1 eV): a little heat or light can kick a few electrons across into the conduction band, leaving behind mobile vacancies called holes. The whole difference between a wire, a window, and a microchip is the width of one energy gap.
Semiconductors become useful when you deliberately spoil their purity. Replacing a few silicon atoms (four valence electrons) with phosphorus (five) drops spare electrons into empty levels just below the conduction band; swapping in boron (three) leaves hungry holes just above the valence band. This is doping, and tuning it is how every transistor, solar cell, and LED is engineered. Doping is the band-theory cousin of delocalized pi systems in molecules: in both, adding or removing a few electrons from a shared, spread-out set of orbitals dramatically changes the electrical behavior of the whole structure.
One picture, from H2 to the whole crystal
Step back and admire what you have. The very same idea — atomic orbitals overlapping in or out of phase to make spread-out, shared orbitals filled from the bottom up — carries you across an astonishing range of scale. Two hydrogen atoms make H2. Six carbon p orbitals make the delocalized pi system of benzene. A trillion-trillion silicon atoms make a wafer with a band gap tuned to switch transistors. There is no boundary where 'molecule physics' stops and 'solid physics' starts; there is one continuous theory of bonding, and the molecule is simply its smallest case.
Be honest, though, about where the simple sketch is a sketch. The neat one-dimensional chain of atoms is a cartoon; a real crystal is three-dimensional, and the proper accounting tracks not just energy but the direction electrons travel (a quantity called the wavevector), which is why band diagrams in textbooks look like wiggly curves rather than fat bars. Bands are also not always perfectly full or perfectly empty: thermal energy smears the Fermi level, which is exactly why a semiconductor conducts a little at room temperature and more when warmed. And the picture assumes electrons mostly ignore each other — a good approximation for many solids, but it breaks down for strongly correlated materials, some of which become surprises like insulators where the simple band count predicts a metal.
That is the perfect place to leave the molecular orbital rung. You arrived able to draw the levels of a diatomic and explain why oxygen is magnetic; you leave able to see those same levels multiplied a trillionfold into the bands that make a copper wire conduct and a silicon chip compute. Hold onto the unifying thread — overlap, phase, fill from the bottom — because the next rung takes the crystal seriously, packing real ions into real lattices and measuring the energy that binds them.