Have you ever wondered what causes all those wonderful colors in your favorite specimens? Many have, and the causes have been the subject of many a study. Years ago, the explanations were simple – there were three causes.  Idiochromatic minerals (“self-colored”) were colored by some essential element (the “chromophore”), such as copper or iron. Allochromatic minerals (“other-colored”) were colored by trace impurities (either chemical or mechanical in nature), or to defects in the crystal structure. Pseudochromatic minerals obtained their colors from the diffraction or scattering of light by structures within the mineral.

Simple, but the definitions satisfied many. But these definitions did not stop mineralogists and physicists from delving deeper into the causes of color. Current thoughts on the cause of color in minerals were summarized in an article in the American Mineralogist by Kurt Nassau (1978). He describes twelve separate causes, which can be placed in four main groups: 1) crystal field effects; 2) molecular orbital effects; 3) band gap theory; and 4) physical optics effects. More on each group later.  Before elaborating on the ideas above, we need to look at some color theory. Generally, the color we see for an object is the result of the interaction of white light with that object. White light is made up of red, orange, yellow, green, blue, indigo, and violet light (the spectrum, or “rainbow”). The color of an object is the result of that object absorbing certain colors of light; the color(s) seen is (are) the color(s) not absorbed. For example, if the blue/green/violet colors are absorbed, the color seen is red/orange. A final bit here: From physics, we know that each color has a frequency and a wavelength (actually, each has a narrow range). They also have an associated energy. Keeping things simple, the highest energy is in the violet end of the spectrum, decreasing towards the red end (part of the reason why the ultraviolet light [higher energy than violet light] in sunlight causes sunburn).  Theoretically, white light is made up of equal amounts of the aforementioned seven colors. However, we have a number of sources of “white” light, with minor differences that can be discerned using special instruments called spectroradiometers. Sunlight is actually a bit richer in the greenish-yellow part of the spectrum (our sun is a “yellow” star). “Daylight”, the light obtained from a clear, northern, blue sky at noon (the photographic definition), is enhanced in the blue part. Incandescent bulbs are rich in the red end of the spectrum. Fluorescent bulbs tend to have “spikes” in the orange, green, and violet, due to the excitation of mercury atoms in the bulb; the actual “white” light comes from fluorescing phosphors. (The “spikes” result from the excitation of atoms, and add an overprint to the general white light.) Metal halide (halogen) bulbs have “spikes” in several color areas, depending on the elements used in the bulb. So why do they all appear “white”? Our brain processes what we see to tell us it is white; if you compare several side by side, you can perceive the ever-so-slight differences. However, the world of minerals gives us an example that shows the differences among light sources.  Alexandrite, the gem variety of chrysoberyl, appears red under incandescent light (richer in red), but green under daylight or fluorescent light (richer in green-blue). The subtle differences in light sources results in a major change in appearance of the mineral.

Now we go back to the causes. The first main group of causes is due to crystal field effects. Within this group are two main causes – transition metal ions and color centers (“defects”). The metal ions include only a few of the 92 naturally-occurring elements, consisting of a few in Period 4 of the Periodic Table (vanadium, chromium, manganese, iron, cobalt, nickel, and copper), and some of the Lanthanide and Actinide series of elements. The cause of color due to these metals is related to the filling of the d- or f-shells of electrons (check your chemistry book….).  Generally, electrons like to move about in pairs. When these metals form ions, a single, lone electron may be left in the orbital; in order to satisfy its need for companionship, it may absorb energy (i.e. color) from any incident light.  The energies (wavelengths, frequencies) not absorbed causes the color seen. And, the “typical” colors seen may depend on the valence state (the electrical charge) of the ion. For example, copper with a +1 charge is typically colorless, whereas when it is in the +2 state it is green or blue. Iron in the +2 state is typically colorless to green, and in the +3 state it is yellow to orange. The color variations are explained by “coordination effects”, that is, how the negatively-charged ions pack themselves around the metal ions (see, for example, Hurlbut and Klein for an explanation).

Visible Light Wavelength and Perceived Color Color centers are also referred to as defects. There are two types of color centers. An electron color center, also called an F-center (from Farbe – German for color), is due to unpaired electrons not on metal ions. Surprisingly, these electrons occupy positions within the crystal lattice with a missing ion; instead of an anion (a negatively-charged ion), there is only that poor, lonely electron. This source is often invoked to explain the colors in fluorite. The second type is called a “hole center”, whereby an electron is missing from a metal ion (often an impurity metal) which usually has a pair of electrons. This type of color center appears to be formed mainly by the effects of radiation; an impurity metal having a different electrical charge than the “main” one is required in the structure to keep the electrical charges balanced. Such effects are thought to cause the colors of smoky quartz, amethyst, and blue topaz, to name a few. It should be noted that colors caused by this effect can be lost by heating, but reintroduced by irradiating the specimen.
The next group, molecular orbital effects, involves electrons not located on a single ion, but rather involved with a group of ions. And, there are three types: Metal-metal, metalnonmetal, and nonmetal-nonmetal. In the metal-metal effect, there are two (generally transition, see above) metal ions, each of which can exist in two valence states (electrical charges). The absorption of light energy (the colors you don’t see) causes an electron to transfer from one ion to the other; it then returns to the original ion, releasing heat (a very tiny amount). For example, on absorption of light, the pair iron +2/titanium +4 becomes iron +3/titanium +3, resulting in the blue color in sapphire.  The metal-nonmetal effect involves a multiple-ion anion (a negatively charged group), in which the ions are covalently bonded (again, see your chemistry text). The most obvious group here is the chromate group, consisting of a chromium ion surrounded by four oxygen ions. The electron transfers between the chromium and oxygen cause absorption of the blue end of the spectrum, yielding yellow to orange colors. As with the crystal field effect, the coordination of ions may also affect coloration.  The nonmetal-nonmetal effect involves, you guessed it, nonmetals, which are covalently bonded. In the mineral world, this is best exemplified by sulfur; the absorption of the blue end results in the yellow color. Interestingly, as you heat and melt sulfur, the coordination of the atoms changes, causing the color to change to orange, to red, then finally to black (the sulfur is absorbing all light at that point, assuming it hasn’t burst into flames). This effect probably also contributes to the colors of those nasty halogen elements: chlorine and fluorine (both greenish gases), bromine vapor (red), and iodine vapor (violet); you would not want to try to collect these guys, even if they existed in nature.

The next beast, band gap theory, is the hardest to explain in simple terms. In the above two groupings, the electrons belong to either a single ion, or to a discrete grouping of a few ions. In this group, the electrons belong to the crystal (i.e., the mineral) as a whole; they are not constrained to a single atom or ion. This group includes conductors (metals), semiconductors (most sulfides, sulfosalts, oxides), and semiconductors with impurities (e.g., colored diamonds). This group gets into the ever-so-nasty quantum effects, wherein the energy levels occupied by the outermost electrons are stretched into “bands”. There is a valence band, which is the normal energy level of the atom (the ground state). There is also a conduction band, which is a higher energy level where the electrons are more mobile. The energy difference between the two is termed the “band gap”; an electron within the valence band can absorb light energy and move into the conduction band.  This is simply put. In the metals, the outer electrons are in a common pool, that is, the valence band merges with the conduction band, allowing them to move rather freely throughout the crystal structure. This is the cause of the metallic luster, the high electrical conductivity, and other traits found in metals. The surface electrons can absorb light of any energy, but they reemit it; the efficiency of reemission produces the colors (i.e., copper, silver, gold).

Within the next subgroup, the minerals are covalently bonded, in which the average number of bonding electrons is four per atom. Light with energy greater than the band gap will be absorbed, moving electrons from the valence band into the conduction band. For example, if the ions absorb all the visible light energy, the color we see is gray or black (e.g., galena). If only the high-energy light (blueviolet) is absorbed, the color seen is red to yellow (cinnabar, cuprite, realgar, orpiment). 

Semiconductors with impurities are just that…..a semiconductor (such as carbon) with an impurity. In the mineral world, diamond is the best example, although this group is important in the electronics industry (but not for pretty, colored materials). For example, traces of nitrogen in diamond cause a yellow color, whereas traces of boron cause a blue color. Basically, this is similar to the color center effect, except that the electronic bonding between atoms/ions is different. The impurities lead to extra electrons, or “holes”, within the structure; in this case electrons bandy about with the conduction band.  The final group includes the physical optics effects. Here, light is interacting with structures similar to the wavelength of light. These effects include scattering, dispersion, diffraction, and interference. Scattering involves the reflection of light off small particles, and includes chatoyancy (as in tigers eye), asterism (star sapphire), the luster of pearls and fibrous minerals, aventurescence
(sunstone, some schiller), and adularescence (moonstone).  

For more details on these, consult your favorite reference, please.

Dispersion has to do with a material’s index of refraction – simply put, different wavelengths (colors) bend differently when passing through a crystal at an angle; again, the explanation can be found elsewhere, and is not given here.  This is what happens when white light is passed through a prism to produce a “rainbow” spectrum. This effect is used in faceting – you essentially are creating a complicated prism, which breaks the light into the colors (the “fire”).  This is best seen in materials such as diamond, (pure) rutile, zircon, and cubic zirconia, to name a few.  Diffraction occurs when you have particles or layers with the same general dimension or thickness as the wavelength of light (or, at least of some of the colors). For example, precious opal has been examined under electron microscopes, and has been found to be composed of regular arrays of small spheres, all of roughly equal size; the size determines the color(s) of the “fire” seen.  Labradorite, a variety of feldspar, has numerous very thin layers of differing composition, with differing optical properties; when oriented correctly, light diffracts through the layers producing incredible flashes of color.  Interference has to do with light passing through thin layers of differing refractive index. Examples include the tarnish layers on bornite and chalcopyrite (the “peacock coppers”), and the iridescent hematite (or whatever it is) found at Graves Mountain.

Thus I conclude my article on the causes of color in minerals. Is this simple and complete? No. I have deliberately kept a lot of stuff out. Hopefully, your interest is piqued enough to get you to dig a bit deeper, to learn more. You can do it. Just don’t rush into it…. Even with a scientific background, many struggle with a lot of it (the band-gap, for example), but they work with it. That’s how they learn.

Below is a list of some coloring elements and the color they produce in at least one mineral:
 

• Cobalt, Co, produces the violet-red color in erythrite, (cobalt arsenate).
• Chromium, Cr, produces the color orange-red color of crocoite, (lead chromate).
• Copper, Cu, produces the azure blue color of azurite, (copper carbonate hydroxide).
• Iron, Fe, produces the red color of limonite, (hydrated iron oxide hydroxide).
• Manganese, Mn, produces the pink color of rhodochrosite, (manganese carbonate).
• Nickel, Ni, produces the green color of annabergite, (hydrated nickel arsenate).
• Uranium, U, produces the yellow color of zippeite, (hydrated potassium uranyl sulfate hydroxide).
• Vanadium, V, produces the red-orange color of vanadinite, (lead vanadate chloride).

From http://mineral.galleries.com/minerals/property/color.htm