It's not unusual to see small blue and green flames when watching a campfire or burning waste paper, and many people assume that somehow copper contaminated the fuel to produce the colors. Although that can happen, the real reason is a lot more interesting.
Perhaps the most surprising thing about fire is that solids and liquids don't burn. Only gases burn. When you touch a match to a piece of paper, the paper heats up until flammable gases are given off, either from being vaporized or from the heat of the flame breaking down molecules to produce flammable gases. Once those gases are emitted the match flame causes them to combust. The fire is now lit. Radiant heat from the flames cooks still more flammable gases out of the fuel and the fire is sustained as long as there's enough heat and fuel. The process of breaking down the fuel into combustible gases is called pyrolysis.
The radiant heat from a fire can do weird stuff. I have a friend whose home burned down years ago. I helped him with the salvage. I found a piece of charred paper on a table, moved it, and found completely intact finish underneath. Every exposed part of the table was charred. He had a stack of records that I expected were a total loss. All I could see of the outside was a charred cube. I peeled the albums apart to find that the albums inside were intact. Not only were the records undamaged but the plastic wrap on the album covers was perfectly preserved. Clearly, all the damage was cause by radiant heat and even a sheet of paper was enough to shield materials from damage. That's why a thin silvered Mylar blanket can provide emergency fire protection. The environment isn't hot. Only surfaces exposed to radiant heat are damaged. (At the conclusion of the salvage, my fried said we could take anything else we wanted. I saw some singed books I thought might be useful. After I rounded them up, I went to the back door only to find it had been nailed shut! Edgar Allen Poe would have loved it.)
Complete combustion produces the blue flame of a gas burner. The combustion products are carbon dioxide and water vapor, both non-toxic. (Yes, too much carbon dioxide is toxic, but if you have enough ventilation to give a good flame, the carbon dioxide won't build up to dangerous levels. For that matter, a lung full of water will kill you, too) But if combustion is incomplete, some of the carbon in the fuel only burns to carbon monoxide (CO) which is toxic because it binds to the hemoglobin in the blood. If you're using a stove and your lips turn cherry red (usually accompanied by headache) get into fresh air ASAP. Since there isn't enough oxygen to combine with all the carbon, some of it forms microscopic particles of carbon (soot) which are heated to incandescence by the flame, giving the familiar yellow or orange color of flames. This is why an excessive amount of yellow in your gas furnace flame is a sign to call the repairman. The unburned carbon collects on any surface it hits leaving a film of fine carbon particles familiar to everyone as soot. Yes, there is carbon monoxide any time there's a yellow flame. No, it's not enough to be dangerous in the case of candles or campfires. It can be lethal in house fires.
People who know just enough chemistry to be dangerous are aware that burning copper gives off green flames, and they claim that any green flames are due to copper. I've even seen claims that the blue of a gas flame is due to the copper tubing used in the burner. All I can say is, if your furnace is emitting vaporized copper into your living space, you'd better find a safer heating system like, say, setting your sofa on fire. To be fair, blue and green flames can be due to copper. Flares, fireworks, and decorative fireplace logs use copper to create green colors. And stray copper can cause green and blue flames in other settings, for example, wood with copper-based fungicides or paper with copper in the ink or copper staples. But if you're burning yard waste or plain waste paper, copper is very unlikely to be present to cause a colored flame. And copper is absolutely not the cause of blue gas flames on a kitchen stove or the blue inner core of a candle flame. From now on, any mention of blue or green flames will refer only to flames not involving copper or barium, or any other element that gives off blue or green wavelengths. (Actually, that's not quite true. The substance that is giving off the colors is just one that nobody would suspect.)
Somewhat more savvy writers, who know just enough physics to be dangerous, know about black body radiation. Black body radiation is the radiation given off by objects that are hot, solely because of heat. There is no reflected light (hence the term "black") or specific wavelengths emitted by atoms. Fires, incandescent light bulbs, atomic fireballs, and the Sun emit pretty pure black body radiation; lasers, fireworks and sodium vapor lights do not. The hotter something is, the farther the peak of radiation lies toward the blue end of the spectrum. A clothes iron doesn't give off visible light at all. Its peak is in the infrared. A bar of iron glows red, then orange, then yellow as it gets hotter and its emission peak shifts through the spectrum. But it never glows green. By the time something is hot enough for its emission peak to be in the green (like the Sun) there is such a broad range of wavelengths present that it appears white. The very hottest stars emit far more blue, violet and ultraviolet than red, and they actually appear blue-white.
Black body radiation is a hugely important tool in science because the color of the light depends solely on the temperature. This means we can view a lava flow or a star from a distance and measure its temperature. What about the Moon? The Moon and Venus shine by reflected light but since they're pretty neutral in color, they reflect the black body radiation of the Sun. Mars, on the other hand, reflects red light more strongly and its reflected light is not close to black body. But the visible light reflected from planets indicates the temperature of the Sun, not the planet. On the other hand, the planets absorb energetic radiation and re-emit infrared radiation, which does indicate their temperature. So, people who know about black body radiation sometimes claim that yellow flames are cool and blue flames are hot. They are, but black body radiation has nothing to do with it. Rigel, one of the bluest stars in the sky, has a surface temperature of 12,000 degrees K, twice as hot as the Sun or the vaporization point of tungsten, and Rigel is nowhere near as blue as the flame on a kitchen gas burner. If your kitchen stove is giving off flames hotter than 12,000 K, you won't have to worry about "a watched pot never boils." What's in the pot will boil instantly. The pot will vaporize soon afterward, followed quickly by the stove itself.
Most of the colors we associate with flames are black body radiation, whether the red of glowing coals, or the yellow and orange of a campfire. That's why blue and green flames are so startling.
If you spread the light from a flame, an incandescent light bulb, or the Sun (1) into a spectrum, you'll get a continuous rainbow of color. The Sun's spectrum will be bright in all wavelengths; the flame or light bulb will be a lot brighter on the red end. But all colors will be present. This is called a continuous spectrum. If you do the same to a fluorescent light or mercury or sodium vapor lights, you'll only see bands of certain colors with nothing in between. This light is given off by specific atoms or molecules at specific wavelengths, and no others. This is called an emission spectrum. Blue and green flames, it turns out, have an emission spectrum. But they don't contain copper, barium, cesium, or any of the other elements that burn to give off blue or green light. So what's burning to create the colors? The answer, surprisingly, is carbon and hydrogen. But not individual atoms. These colors are given off by molecules.
|At left is the spectrum of a fluorescent light, taken through an inexpensive hand-held spectroscope. At left in the picture is the slit that admits the light from the bulb. At right is the spectrum. The lamp itself emits mostly ultraviolet light, but the phosphor coating of the bulb fluoresces in visible light. The red, green and violet lines are due to mercury, which is why mercury vapor lights look purplish-white but give everything a green cast. Additional elements are put into the lamp phosphor to fill in gaps in the spectrum. The blue is most likely due to antimony and the orange to manganese.|
Atoms and molecules give off radiation in many ways. At the atomic level, all energy is quantized or restricted to specific levels. Molecules can emit radiation when their rotational energy changes. This energy is generally weak and emitted as microwaves. Then there are relative motions between atoms in molecules. Atoms can oscillate back and forth toward and away from each other, groups of atoms can twist relative to other atoms, and angles between atomic bonds can change. Many of these oscillations are caused by infrared radiation and are the cause of the heat-absorbing properties of water vapor, carbon dioxide, and methane (the well known "greenhouse effect"). But they can also involve visible light as well. Water really is blue, not in a glass, but in swimming pool sized batches, because some of its oscillations extend weakly into the red part of the spectrum. Methane also absorbs red light, which is while the methane rich atmospheres of Uranus and Neptune are blue. But generally, speaking, molecular vibrations are too weak to absorb or emit visible light.
Many molecules, when pumped up to high energies, emit some of that
energy as visible light. Individual atoms give off or absorb light by exciting
electrons to higher energy levels (absorption) and having them
fall back to lower levels (emission). When multiple atoms are
involved, colors can be created in a variety of ways. Electrons
can be kicked back and forth between atoms. This mechanism,
called charge transfer, produces the red color of rust. In many
cases, electrons aren't bound to a single atom but are shared among
several atoms. These electrons are said to occupy molecular orbitals,
and molecular orbitals produce the colors of leaves, flowers, and
synthetic dyes. Molecular orbitals make the world colorful.
One common molecule that has an emission spectrum is nitrogen. It emits spectral lines in the red and blue part of the spectrum and is responsible for the red seen in energetic auroral displays. So how come nitrogen doesn't absorb red and blue in the air, leaving the air with a greenish tint? Because the nitrogen molecules have to be pumped up to fairly high energies first. They have to be excited by ultraviolet light or energetic particles, then they give back some of that energy as visible light. An analogy might help here. It's not the casual weekend gamblers in Las Vegas who give $100 tips. It's the high rollers. Casual gamblers just don't have enough money, but the high rollers won't miss it. Ordinary visible light and atomic collisions aren't energetic enough to get nitrogen molecules to levels where they can emit visible light. The molecules have to be at much higher energies before they can give back any energy as visible light.
The cause of the colors of green and blue flames are molecular emissions known as Swan Bands. The first stage in pyrolysis is to break complex molecules into simpler, usually electrically charged fragments known as radicals (Detecting radicals is how smoke alarms work. That's why they occasionally give false alarms from excessive humidity or other causes). One of the most common radicals consists of two carbon atoms with the formula C2. The carbon atoms in the C2 molecule merge some of their orbitals to create molecular orbitals. C2 molecules in a flame, given energy by the heat of the flame, emit radiation at definite frequencies. The emission comes not from electrons around a single carbon atom, like a copper flame is due to individual atoms, but to the energy levels in the molecular orbitals. These frequencies occupy bands of the spectrum from orange to blue but become progressively stronger in the green and blue. When the electron drops to a lower energy state, it also generally perturbs the rotational and vibrational energies of the molecule as well. These changes are small compared to the energy of the visible light, and the effect is to create many closely-spaced emission lines rather than a single line. The result is a band of absorption rather than a sharp line. Far from being a nuisance, band spectra allow chemists to deduce the rotational energy and mass of the emitting molecule.
|Not only do Swan bands show up in some flames, but surprisingly enough in the colors of comets. Comets emit a lot of green and blue light and much of it is Swan Band emission. This is Comet Holmes, which had been plodding uneventfully through the inner Solar System since its discovery in 1892, then suddenly brightened by a factor of half a million in 2007 to become visible to the unaided eye.|
C2 gas emits green and blue light, but what happens when light passes through it? As a general rule, when light passes through a gas, the gas absorbs the very wavelengths it emits. So if C2 gas emits strongly in the green and blue, it ought to absorb strongly as well. Thus, a star with C2 gas in its outer layers will lose much of its green and blue and transmit mostly red and violet. Carbon rich stars, already cool and very red to begin with, have their red enhanced by Swan Band absorption.
1. If you pass sunlight through a narrow slit, you won't quite get a continuous spectrum. There will be thin dark lines caused by the cool outer layer of the Sun absorbing some wavelengths. Strictly speaking, the bright disk of the Sun (the photosphere) has a continuous spectrum. The cool outer layer (the chromosphere) absorbs light when the photosphere shines through it, but displays an emission spectrum when seen in isolation, as in a total eclipse of the Sun.
1. The authoritative sources for the spectroscopy of all diatomic molecules (up to about 1979) are the two books: G. Herzberg, Molecular Spectra and Molecular Structure I. Spectra of Diatomic Molecules, (2nd edition), Van Nostrand, New York, 1950 and K.P. Huber and G. Herzberg, Molecular Spectra and Molecular Structure IV. Constants of Diatomic Molecules, Van Nostrand Reinhold, New York, 1979.
2. J.I. Steinfeld, Molecules and Radiation, An Introduction to Modern Molecular Spectroscopy, (2nd edition), MIT Press, Cambridge, 1986, pages 171-174.
Created March 12, 2013, Last Update April 15, 2013
Not an official UW Green Bay site