What is a crystal?
The definition I use in the course I teach on X-ray Crystallography is this:
"A crystal is homogenous solid exhibiting a high degree of internal order and a definite although not necessarily stoichiometric overall chemical composition."
For simplicity, I usually only talk about systems which exhibit 3 dimensional order. There are, however, materials which display a high degree of internal order, but whose structures cannot be described by a 3 dimensional lattice.
By high degree of internal order means that the material is ordered over many atomic dimensions. Short range order means that order exists only a few atomic dimensions. For example, Silica (SiO2) can exist as quartz (a crystal), or as glass (like in your windows, and has only short range order).
In addition, Tony Linden (alinden@oci.unizh.ch) adds:
A crystal can consist of any virtually pure single chemical compound (small impurities can be present and often give rise to colors in minerals, and mixtures of compounds can co-crystallize, but are less common). The compound may be inorganic, as in minerals and salts, e.g. SiO2 (quartz) or NaCl (salt), or organic, such as sugar. In fact, just about any pure organic substance can be crystallized given the right conditions. Chemists actually use this process to purify their compounds as traces of impurity generally remain in solution when the crystals are formed. While classic compounds such as CuSO4 can produce huge crystals and are the common classroom example because of this, most compounds crystallize as much smaller crystals and the solvents and techniques needed to obtain well formed crystals can vary from one compound to the next (these techniques are the primary focus of our web site). Even proteins can be encouraged to crystallize nicely under appropriate conditions and this is very important for scientists to be able to determine the structure of proteins and thereby understand their function. Protein crystallography is a very hot field in the biosciences these days.
What types of crystals are there?
There are a couple of ways to answer this question. I generally think in terms of crystal systems and lattice types. There are 7 crystal systems:
triclinic, monoclinic, orthorhombic, tetragonal, trigonal, hexagonal and cubic.
Lattices can either be primitive (only one lattice point per unit cell) or non-primitive (more than one lattice point per unit cell).
If you combine the 7 crystal systems with the 2 different types of lattices, you end up with 14 Bravais Lattices (named after Auguste Bravais who figured all this out in 1850).
Another way to answer your question is to catagorize crystals by their physical/chemical properties. In this classification you have four types of crystals:
Covalent Crystals: This is a crystal which has real chemical covalent between all of the atoms in the crystal. So really a single crystal of a covalent crystals is really just one big molecule. An example of this is a crystal like diamond or zinc sulfide. Covalent crystals can have extremely high melting points.
Metallic Crystals: Individual metal atoms sit on lattice sites while the outer electrons from these atoms are able to flow freely around the lattice. Metallic crystals normally have high melting points and densities.
Ionic Crystals: This is a crystal where the individual atoms don't have covalent bonds between them, but are held together by electrostatic forces. An example of this type of crystal is sodium chloride (NaCl). Ionic crystals are hard and have relatively high melting points.
Molecular Crystals: This is a crystal where there are recognizable molecules in the structure and the crystal is held together by non-covalent interactions like van der Waals forces or hydrogen bonding. An example of this type of crystal would be sugar. Molecular crystals tend to be soft and have lower melting points.
Of course, this classification scheme has some ambiguous areas. For example, what of you have a crystal like [(CH3)2NH2]+ (CH3CO2)- is that a molecular crystal or an ionic crystal?
How do crystals form and how do they grow?
Crystals start growing be a process called "nucleation". Nucleation can either start with the molecules themselves (we'll call this unassisted nucleation), or with the help of some solid matter already in the solution (we'll call this assisted nucleation). I'll write about both. Before I do that, here is a general explanation from Tony Linden:
Once a solution is saturated, or a melt nears the solidification point, solid material starts to form. If the molecules come together in a random arrangement, they do not occupy the closest packed space. However, if the molecules come together in an ordered array, they pack together in a much smaller space, like in a properly assembled jigsaw puzzle (see how much more space an unassembled puzzle occupies when the pieces are randomly positioned to touch each other, but not overlap). Thus, the proper packing uses less space and is also of lowest energy, which is always the most stable condition. As it happens, this ordered array pattern repeats itself regularly in 3 dimensions, and the crystal is the macroscopic object we see as a result. The nice faces of a crystal result from the fact that certain directions in this array are more accessible to the attachment of new molecules, so the crystals grow uniformly in these directions. However, the incoming molecules need a little time to align themselves properly at the surface of a growing crystal in order for the crystal to continue to grow nicely. Hence the need for slow crystallization. If the solution becomes over-saturated so that solid forms quickly, the incoming molecules do not have time to align properly with the result that one obtains small crystals that are usually poorly formed.
Now we can talk a bit about nucleation, a very important step in crystallization.
Unassisted nucleation:
When molecules of the "solute" (the stuff of which you want to grow crystals) are in solution, most of the time they see only solvent molecules around them. However, occaisionally they see other solute molecules. If the compound is a solid when it is pure, there will be some attractive force between these solute molecules. Most of the time when these solute molecules meet they will stay together for a little while, but then other forces eventually pull them apart. Sometimes though, the two molecules stay together long enough to meet up with a third, and then a fourth (and fifth, etc.) solute molecule. Most of the time when there are just a few molecules joined together, they break apart. However, once there becomes a certain number of solute molecules, a so-called "critical size" where the combined attractive forces between the solute molecules become stronger than the other forces in the solution which tend to disrupt the formation of these "aggregates". This when this "protocrystal" (a sort of pre-crystal) becomes a nucleation site. As this protocrystal floats around in solution, it encounters other solute molecules. These solute molecules feel the attractive force of the protocrystal and join in. That's how the crystal begins to grow. It continues growing until eventually, it can no longer remain "dissolved" in the solution and it falls out (as chemists like to say) of solution. Now other solute molcules begin growing on the surface of the crystal and it keeps on getting bigger until there is an equilibrium reached between the solute molecules in the crystal and those still dissolved in the solvent.
Assisted Nucleation:
Pretty much the same thing happens as in unassisted nucleation, except that a solid surface (like a stone, or brick) acts as a place for solute molecules to meet. A solute molecule encounters the surface of a stone, it adsorbs to this surface, and stays on it for a certain time before other randomizing forces of the solution knock it off. Solute molecules will tend to adsorb and aggregate on the surface. This is where the protocrystal forms, and the same process as described above happens.
You can probably see from what I wrote above, why solution in which the concentration is near saturation, that crystals grow fastest. If there are more solute molecules in a given volume, then there is more of a chance they will meet one another. You also don't want to heat up the solution because that acts as the major randomizing force in solution which causes the aggregates of molecules to break up.
Why do different crystals have different shapes and sizes?
This depends on 2 factors: 1) The internal symmetry of the crystal, and 2) The relative growth rates along the various directions in the of the crystal. For example, suppose you have mutally perpendicular axes, a, b, and c. Suppose the crystal grows at equal rates along a, b, and c, then the crystal shape will be a cube. Now suppose a different crystal grows fast in the a and b direction, but very slowly in the c direction. The crystal will then grow as thin plates with the face of the plate being perpendicular to c. These are only simple examples. More complicated cases (and shapes) happen when the crystal doesn't have mutually perpendicular axes, and when the fastest directions of growth correspond to face or body diagonals (or even other directions) in the crystal.
Why crystals grow at different rates in different directions is a very complicated question. If there is a highly attractive interaction (energetically speaking) along a certain direction of a crystal, then that direction will probably grow fast. However, it could also grow slowly, if that direction interacted strongly with the solvent; having strongly absorbed solvent on the surface of the crystal could block growth along that face.
What do you consider a perfect crystal?
A perfect crystal is one which has a single lattice (i.e. not a twinned crystal) and is completely regular, free of defects and dislocations. Most single crystals, however, are imperfect in the sense that they are composed on regions of slight relative misalignment (about 0.1 to 0.2 degrees). For X-ray diffraction (which is what I do), this slight imperfection is desirable because the diffracted intensities from such a crystal are higher than those whose lattices are "perfect". When a crystal is too perfect, the X-ray diffraction pattern suffers from what is called "extinction". Crystallographers then talk about an "ideally perfect" and "ideally imperfect" crystals. We use these terms to talk about real crystals which behave in accordance with simple theories of X-ray diffraction.
How does light affect the color of a crystal?
The color of any compound (whether or not it is a crystal) depends on how the atoms and or molecules absorb light. Normally white light (what comes out of light bulbs) is considered to have all wavelengths (colors) of light in it. If you pass a white light through a colored compound some of the light is absorbed (we don't see the color which is absorbed, but we see the rest of the light) as it is reflected off the surface. This gives rise to the idea of "Complementary Colors". If a compound absorbs light of a certain color the compound appears to be the complimentary color. Here is a table of colors and thier compliments:
| Color | Compliment | Wavelength (of color nm) |
| violet | green-yellow | 400-424 |
| blue | yellow | 424-491 |
| green | red | 491-570 |
| yellow | blue | 570-585 |
| orange | green-blue | 585-647 |
| red | green | 647-700 |
So if you have a crystal which absorbs red light, it will appear green. Conversely, if the crystal absorbs green light, it will appear red.
With regard to the optical properties of crystals, Tony Linden, adds:
One interesting feature of some crystals is their effect on an image viewed through them. A calcite crystal placed over a cross on a page will make the cross look doubled because of total internal reflection within the crystal.