[Lee¡¡¡]

Morning people, I just have something that is troubling me. I was wondering what happens at the end of a carbon in diamond or graphite. How does the diamond get its shape if carbon needs all four bonds bonded when in a diamond there contains only carbon atoms.

Also, the same goes to polymers when they are formed from monomers. What happens at both the ends of the polymer? Basically, how does these compounds come to a stop forming objects that we use?

Thank you for your time...

[Arkcon]

This is a common question, I used to wonder about it myself, even into college.  Simply put, the end of the crystal, or polymer are bound to something else.  Every crystal has flaws, even the tiniest, purest diamond crystal on Earth has millions of atoms, many of them other than carbon.  Eventually, inclusions and flaws add up so that the crystal can end, otherwise it wouldn't.  And it has to.

[jeffmoonchop]

It ends if there is no more carbon to fuel growth. The Gibbs free energy of atoms in the bulk crystal is a lot lower than atoms on the surface due to the interior bonds being fully satisfied with its 4 bonds. The atoms on the surface are under stress because its bonds are not satisfied. Yes there will be defects within the bulk crystal but in order to satisfy the bonds on the surface it will form a layer of whatever it can find to neutralise its instability. The surface tension between the diamond surface and another phase is quite high, a few layers form to stabilise it.

[Corribus]

Most solid materials have an oxide or similar layer on the surface at the atmospheric interface. Diamond is no exception.

http://www.sciencedirect.com/science/article/pii/0301751682900291

[Lee¡¡¡]

Thank you for the information.

[Enthalpy]

This is a fundamental question in microelectronics because most semiconductor devices operate within the first tens on nm of the solid material - and one thing that desesperate process engineers because their needs are so uncommon that help isn't available elsewhere.

"Pending bonds" at the crystal end are so reactive that they last for a few seconds, at most minutes, in the best vacuum achievable by humans. About every molecule of gas remaining in the chamber that impinges gets trapped by the reactive surface to make a bond, and very soon, the surface is passivated by anything that was present, including H₂ or Ar - until O₂ becomes available.

So a bare crystal surface exists for very short between the moment it is created and the moment it's spoiled. This is one explanation to very inconsistent data about the electron work function (the energy to rip an electron from the crystal), because it depends fundamentally on the crystal's surface - the other explanations being that the work function depends on crystal edges, crystal orientation, and for polycrystals, on the location.

Very noble materials like Au or Ir tend to adsorb gas weakly in few atomic layers. More reactive materials like Si or Al make strong chemical bonds with oxygen, not broken by reasonable heat and irreversible, in a layer that grows until it's thick enough to stop O or Si atoms diffuse through. Thicker layers are obtained by heating or with electricity to transport the ions through the layer - that's called anodization, common for aluminium.

Pending bonds exist at interfaces between solids too. One reason is that both solids can't match their atom spacing, so some atoms have no neighbour at the other material. There, charges can be stored since lone electrons are easier to add or remove. This happens naturally in MOS transistors at the Si-SiO₂ interface as a shift in the threshold voltage and as a "1/F" (=at low frequency) noise when ambient heat randomly adds and removes electrons. The exceptionally good quality of the Si-SiO₂ interface was the basic reason to use Si for semiconductor components, and still now, MOS or MIS transistors are very difficult to make with other semiconductors; Mesfets (metal+semiconductor, without an insulator layer) dominate the other semiconductor components, say on GaAs, but the poor interface gives them a strong 1/F noise. One means of action is to passivate the interface's pending bonds by a hydrogen or deuterium plasma, where the small hydrogen atoms diffuse to the interface and complete the pending bonds.

Interfaces are also a headache to components' models because they aren't well defined like single-crystal semiconductors are. Especially, metal-semiconductor "Schottky" contacts don't behave at all as one would expect from materials bandgap and work function, because the interface spoils all potentials and theories. Components and process designers must work with strictly experimental and badly understood data here - when available...

The charges stored at interfaces (then between insulators, like Si₃N₄ and SiO₂) have good sides too: they are the information in Flash memory, put there and removed by tunnelling through an insulator thin enough. Many of the brought charges reside in pending bonds.