The semi-metals and Silicon

Following John's update on the Periodic Table two months ago, Peter told us about one class of chemical elements, the Rare Earths.

I'm going to talk about Silicon, touching slightly on the group to which it belongs, the semi-metals or metalloids. Perhaps in the coming months others members of the group may focus on some other elements or groups of elements.

Silicon has lots of fascinating aspects. And it has two quite distinct personalities. Here's one of course - the use of wafers pure monocrystalline silicon as the starting point for silicon chips. But there's far more, and here are the topics I shall cover.

So let's start with the Periodic Table. Generally speaking, leaving out Peter's Rare Earths, most elements are classified as metals. Non metals are found in the top right-hand corner. Sulphur for example, is quite definitely a non-metallic solid.

But some elements are not quite sure what they are. They lie along the boundary between metals and non-metals and are known as semi-metals or metalloids. They have a mixture of metallic and non-metallic characteristics, and they either are, or are a basis for semi-conductors. Back to that later.

A little now on atomic models. In the classical atomic model, an atom has a nucleus, surrounded by orbiting electrons. The number of electrons is the same as the atomic number and the electrons are grouped into shells As we move along the Periodic Table, every time we reach an inert gas, a shell has become full, and a new outer shell begins to be filled. And it is the number of electrons in the outermost shell which strongly influences the properties of the element. When that shell is only partly filled, electrons are easily detached - making for the good electrical and thermal conductivity of metals. When the shell is very nearly full, then electrons are not easily detached, especially with the smaller atoms where the outer shell is closest to the nucleus - and indeed the atom is hungry for electrons. Hence chemical bonding readily occurs between Sodium and Chlorine say.

Silicon is, in a sense, right in the middle, having just 4 electrons in an outer shell that has room for 8. Nowadays the first application of Silicon which comes to mind is the silicon chip, and this all stems from the atomic structure. So I'll be coming back to this in the second half of my talk where I will also touch briefly upon some of the other semi-metals and their compounds.

I said that Silicon has two personalities. The first derives from its membership of the semimetals club, leading to its applications in electronics in nearly pure elemental form. In that case it is the applications as a semiconductor, and in those applications the neighbours B, Ga, Ge, As and P are particularly important. The second derives from its closeness to Carbon and therefore to its participation in chemical compounds which can be highly complex - even polymers. And these divide into two groups. There are inorganic minerals, especially feldspars which are aluminosilicates. The minerals are often crystalline network solids. And there are organsilicon compounds where the molecules contain silicon-carbon pairings

Silicon is of course everywhere. Unlike Peter's Rare Earths, Silicon is among the most abundant elements - 27% of the earth's crust. We won't run out! However pure silicon is rare or non-existent in nature. All the silicaceous rocks such as sand, quartz and clay are compounds. Man has used these materials for thousands of years and learnt how to make glass in about 3000 BC. Silicates are an essential part of cement and concrete and of ceramics since kaolinite is a silicate mineral. Many gemstones such as opal and amethyst are forms of silica and many others are more complicated compounds of silicon. And asbestos is another silicate. Other silicon compounds which are mined in quite large quantities are zeolites which I shall come back to. And there is obsidian which is a natural glass.

Silicon was first prepared in pure form in 1824 and given it's name very soon after that. Silicon could now become the basis for new chemical compounds. But that had to wait until the 1890s. First of all Acheson accidentally produced Silicon Carbide which he called carborundum when he was actually trying to produce artificial diamonds. Silicon Carbide continues o be highly important as an abrasive. Then the invention of the electric arc furnace in 1899 allowed industrial production of silicon itself to begin. The result is metallurgical grade silicon. This has huge application in its own right as well as being a step in the production of the much purer electronic grade silicon.

Metallurgical grade silicon is used a) in huge quantities alloyed with aluminium in aluminium casting production and b) in the chemical industry as a basis for other chemicals.

As one might expect from its position in the periodic table above carbon, Silicon is quite good at combining with other elements - so there are many compounds. In the naming of compounds, confusion reigns supreme. We have fumed silica - and silica fume. Fumed silica is produced in huge quantities and is a basis for cosmetics and for thickening agents, in milk shakes for instance. Then we come to silicones - hugely important materials. They resulted from research into organosilicon compounds - silicon compounds which are analogous to organic compounds based on carbon.

The first of such compounds were discovered in the 1860s. But great impetus was given by the work of Frederick Kipping, a professor at Nottingham University. He began this in about 1900 and discovered the silicones and gave them their name. They are inert often rubber-like polymers which have found lots of applications as seals and insulators, oils and greases and of course in silicone bakeware. The main trouble is that the words silicone and silicon are so often confused.

The zeolites are another remarkable group of Silicon compounds and zeolite chemistry, like silicone chemistry, results from the work of a british chemist, Professor Richard Barrer in the 1930s. Unlike silicones zeolites occur naturally associated with volcanic lava and their odd behaviour was recognised as early as the mid-18th century. Zeolites are microporous crystalline solids. Within the crystal, the individual molecules are assembled into open frameworks which include microscopic pores. That then gives them remarkable properties. Uses: catalysts, washing powder surfactants, air conditioning and water purification. They have also been taken up and aggressively marketed by the complementary medicine community - especially for detoxification, but also to cure most other ills! Hence this rejuvenation cream for just £30 a jar. I'll take orders later.

Let's now move on to semi-conductors. The real story of began with the invention of the transistor in 1949. Prior to this some solid-state semiconductor devices such as selenium and copper oxide rectifiers did exist. Those were the days when smell was a good method of fault diagnosis in electronic equipment - Selenium rectifiers are memorable for the really foul smell they made when they failed !

The 1949 transistor was based on Germanium and was a point-contact device analogous to the crystal detectors in early crystal radios. Point-contact devices were very soon supplanted by junction transistors, and germanium was largely swept aside by silicon from the mid 1950s - and it is the silicon junction devices which I'll try to explain.

Extremely pure silicon can be grown into crystals which have this kind of molecular structure. This is monocrystalline silicon. All the electrons are tightly bound into the structure, and this intrinsic silicon is a pretty good insulator - not a perfect one because temperature effects shake out a few electrons.

To be useful, the silicon must be doped by diffusing in small amounts of other elements which supplant silicon atoms here and there. Those elements are ones with either one more, or one less electron in the outer shell.

If there's one more electron resulting from doping with Phosphorus, then the extra electron becomes an unbound conduction electron, and the material is called n-type. n-type silicon conducts electricity through the movement of the free electrons.

If there's one less electron, resulting from doping with Boron, then there is a hole. p-type silicon conducts through the movement of holes resulting from electrons shuffling their way along using the holes.

Something interesting happens when we place join n-silicon to p-silicon like this. It will conduct electricity in one direction but not in the other.

We have a diode.

Now something even more useful happens when we combine two diodes in a suitable way. Here there can be no conduction in either direction. But if we build a sandwich like this - and make the middle layer very thin - then we have a transistor, and it can behave as a switch in which a signal applied to the central layer will control larger current flows across the entire device

Transistors in the 1950s and 60s were like that, but different structures, still involving p and n regions were developed more suited to large scale integrated circuits. The production of a silicon chip involves photolithographic processes which define the regions where p and n regions are produced in the silicon wafer, together with selective deposition of conducting and insulating layers on the substrate.

More and more silicon is now being used in solar panels. The photovoltaic effect has been known for a very long time, and semiconductor photodiodes have been available for over 50 years. What happens is that photons of light deliver energy which produces free electrons and in a diode they can only go in one direction, thus delivering electricity. In early years, photovoltaics used electronics grade monocrystalline silicon - and that would be far too expensive for large solar panels. It was realised that Panels would work very nearly as well with lower grade silicon - often so-called polysilicon where the crystalline structure has fault lines within it. That, and UMG silicon, is far cheaper than monocrystalline silicon.

Finally, one other new form of silicon - porous silicon. Lots of interest a) because of optical properties, and b) suitability for cell culture.

So that's silicon. Isn't it remarkable that such a very common element has such extraordinary properties!