THE SCIENCE OF LIGHT-PRODUCTION

In previous chapters much of the historical development of artificial lighting has been presented and several subjects have been traced to the modern period which marks the beginning of an intensive attack by scientists upon the problems pertaining to the production of efficient and adequate light-sources. Many historical events remain to be touched upon in later chapters, but it is necessary at this point for the reader to become acquainted with certain general physical principles in order that he may read with greater interest some of the chapters which follow. It is seen that from a standpoint of artificial lighting, the "dark age" extended well into the nineteenth century. Oil-lamps and gas-lighting began to be seriously developed at the beginning of the last century, but the pioneers gave attention chiefly to mechanical details and somewhat to the chemistry of the fuels. It was not until the science of physics was applied to light-sources that rapid progress was made.

All the light-sources used throughout the ages, and nearly all modern ones, radiate light by virtue of the incandescence of solids or of solid particles and it is an interesting fact that carbon is generally the solid which emits light. This is due to various physical characteristics of carbon, the chief one being its extremely high melting-point. However, most practicable light-sources of the past and present may be divided into two general classes: (1) Those in which solids or solid particles are heated by their own combustion, and (2) those in which the solids are heated by some other means. Some light-sources include both principles and some perhaps cannot be included under either principle without qualification. The luminous flames of burning material such as those of wood-splinters, candles, oil-lamps, and gas-jets, and the glowing embers of burning material appear in the first class; and incandescent gas-mantles, electric filaments, and arc-lamps to some extent are representative of the second class. Certain "flaming" arcs involve both principles, but the light of the firefly, phosphorescence, and incandescent gas in "vacuum" tubes cannot be included in this simplified classification. The status of these will become clear later.

It has been seen that flames have been prominent sources of artificial light; and although of low luminous efficiency, they still have much to commend them from the standpoints of portability, convenience, and subdivision. The materials which have been burned for light, whether solid or liquid, are rich in carbon, and the solid particles of carbon by virtue of their incandescence are responsible for the brightness of a flame. A jet of pure hydrogen gas will burn very hot but with so low a brightness as to be barely visible. If solid particles are injected into the flame, much more light usually will be emitted. A gas-burner of the Bunsen type, in which complete combustion is obtained by mixing air in proper proportions with the gas, gives a hot flame which is of a pale blue color. Upon the closing of the orifice through which air is admitted, the flame becomes bright and smoky. The flame is now less hot, as indicated by the presence of smoke or carbon particles, and combustion is not complete. However, it is brighter because the solid particles of carbon in passing upward through the flame become heated to temperatures at which they glow and each becomes a miniature source of light.

A close observer will notice that the flame from a match, a candle, or a gas-jet, is not uniformly bright. The reader may verify this by lighting a match and observing the flame. There is always a bluish or darker portion near the bottom. In this less luminous part the air is combining with the hydrogen of the hydrocarbon which is being vaporized and disintegrated. Even the flame of a candle or of a burning splinter is a miniature gas-plant, for the solid or liquid hydrocarbons are vaporized before being burned. Owing to the incoming colder air at this point, the flame is not hot enough for complete combustion. The unburned carbon particles rise in its draft and become heated to incandescence, thus accounting for the brighter portion. In cases of complete combustion they are eventually oxidized into carbon dioxide before they are able to escape. If a piece of metal be held in the flame, it immediately becomes covered with soot or carbon, because it has reduced the temperature below the point at which the chemical reaction—the uniting of carbon with oxygen—will continue. An ordinary flat gas-flame of the "bats-wing" type may vary in temperature in its central portion from 300°F. at the bottom to about 3000°F. at the top. The central portion lies between two hotter layers in which the vertical variation is not so great. The brightness of the upper portion is due to incandescent carbon formed in the lower part.

When scientists learned by exploring flames that brightness was due to the radiation of light by incandescent solid matter, the way was open for many experiments. In the early days of gas-lighting investigations were made to determine the relation of illuminating value to the chemical constitution of the gas. The results combined with a knowledge of the necessity for solid carbon in the flame led to improvements in the gas for lighting purposes. Gas rich in hydrocarbons which in turn are rich in carbon is high in illuminating value. Heating-effect depends upon heat-units, so the rating of gas in calories or other heat-units per cubic foot is wholly satisfactory only for gas used for heating. The chemical constitution is a better indicator of illuminating value.

As scientific knowledge increased, efforts were made to get solid matter into the flames of light-sources. Instead of confining efforts to the carbon content of the gas, solid materials were actually placed in the flame, and in this manner various incandescent burners were developed. A piece of lime placed in a hydrogen flame or that of a Bunsen burner is seen to become hot and to glow brilliantly. By producing a hotter flame by means of the blowpipe, in which hydrogen and oxygen are consumed, the piece of lime was raised to a higher temperature and a more intense light was obtained. In Paris there was a serious attempt at street-lighting by the use of buttons of zirconia heated in an oxygen-coal-gas flame, but it proved unsuccessful owing to the rapid deterioration of the buttons. This was the line of experimentation which led to the development of the lime-light. The incandescent burner was widely employed, and until the use of electricity became common the lime-light was the mainstay for the stage and for the projection of lantern slides. It is in use even to-day for some purposes. The origin of the phrase "in the lime-light" is obvious. The luminous intensity of the oxyhydrogen lime-light as used in practice was generally from 200 to 400 candle-power. The light decreases rapidly as the burner is used, if a new surface of lime is not presented to the flame from time to time. At the high temperatures the lime is somewhat volatile and the surface seems to change in radiating power. Zirconium oxide has been found to serve better than lime.

Improvements were made in gas-burners in order to obtain hotter flames into which solid matter could be introduced to obtain bright light. Many materials were used, but obviously they were limited to those of a fairly high melting-point. Lime, magnesia, zirconia, and similar oxides were used successfully. If the reader would care to try an experiment in verification of this simple principle, let him take a piece of magnesium ribbon such as is used in lighting for photography and ignite it in a Bunsen flame. If it is held carefully while burning, a ribbon of ash (magnesia) will be obtained intact. Placing this in the faintly luminous flame, he will be surprised at the brilliance of its incandescence when it has become heated. The simple experiment indicates the possibilities of light-production in this direction. Naturally, metals of high melting-point such as platinum were tried and a network of platinum wire, in reality a platinum mantle, came into practical use in about 1880. The town of Nantes was lighted by gas-burners using these platinum-gauze mantles, but the mantles were unsuccessful owing to their rapid deterioration. This line of experimentation finally bore fruit of immense value for from it the gas-mantle evolved.

A group of so-called "rare-earths," among which are zirconia, thoria, ceria, erbia, and yttria (these are oxides of zirconium, etc.) possess a number of interesting chemical properties some of which have been utilized to advantage in the development of modern artificial light. They are white or yellowish-white oxides of a highly refractory character found in certain rare minerals. Most of them are very brilliant when heated to a high temperature. This latter feature is easily explained if the nature of light and the radiating properties of substances are considered. Suppose pieces of different substances, for example, glass and lime, are heated in a Bunsen flame to the same temperature which is sufficiently great to cause both of them to glow. Notwithstanding the identical conditions of heating, the glass will be only faintly luminous, while the piece of lime will glow brilliantly. The former is a poor radiator; furthermore, the lime radiates a relatively greater percentage of its total energy in the form of luminous energy.

The latter point will become clearer if the reader will refresh his memory regarding the nature of light. Any luminous source such as the sun, a candle flame, or an incandescent lamp is sending forth electromagnetic waves not unlike those used in wireless telegraphy excepting that they are of much shorter wave-length. The eye is capable of recording some of these waves as light just as a receiving station is tuned to record a range of wave-lengths of electromagnetic energy. The electromagnetic waves sent forth by a light-source like the sun are not all visible, that is, all of them do not arouse a sensation of light. Those that do comprise the visible spectrum and the different wave-lengths of visible radiant energy manifest themselves by arousing the sensations of the various spectral colors. The radiant energy of shortest wave-length perceptible by the visual apparatus excites the sensation of violet and the longest ones the sensation of deep red. Between these two extremes of the visible spectrum, the chief spectral colors are blue, green, yellow, orange, and red in the order of increasing wave-lengths. Electromagnetic energy radiated by a light-source in waves of too great wave-length to be perceived by the eye as light is termed as a class "infra-red radiant energy." Those too short to be perceived as light are termed as a class "ultraviolet radiant energy." A solid body at a high temperature emits electro-magnetic energy of all wave-lengths, from the shortest ultra-violet to the longest infra-red.

Another complication arises in the variation in visibility or luminosity of energy of wave-lengths within the range of the visible spectrum. Obviously, no amount of energy incapable of exciting the sensation of light will be visible. The energy of those wave-lengths near the ends of the visible spectrum will be inefficient in producing light. That energy which excites the sensation of yellow-green produces the greatest luminosity per unit of energy and is the most efficient light. The visibility or luminous efficiency of radiant energy may be ranged approximately in this manner according to the colors aroused: yellow-green, yellow, green, orange, blue-green, red, blue, deep red, violet.

Newton, an English scientist, first described the discovery of the visible spectrum and this is of such fundamental importance in the science of light that the first paragraph of his original paper in the "Transactions of the Royal Society of London" is quoted as follows:

In the Year 1666 (at which time I applied my self to the Grinding of Optick Glasses of other Figures than Spherical) I procured me a Triangular Glass-Prism, to try therewith the celebrated Phaenomena of Colours. And in order thereto, having darkened my Chamber, and made a small Hole in my Window-Shuts, to let in a convenient Quantity of the Sun's Light, I placed my Prism at its Entrance, that it might be thereby refracted to the opposite Wall. It was at first a very pleasing Divertisement, to view the vivid and intense Colours produced thereby; but after a while applying my self to consider them more circumspectly, I became surprised to see them in an oblong Form; which, according to the receiv'd Law of Refractions, I expected should have been circular. They were terminated at the Sides with streight Lines, but at the Ends the Decay of Light was so gradual, that it was difficult to determine justly what was the Figure, yet they seemed Semicircular.

Even Newton could not have had the faintest idea of the great developments which were to be based upon the spectrum.

Now to return to the peculiar property of rare-earth oxides—namely, their unusual brilliance when heated in a flame—it is easy to understand the reason for this. For example, when a number of substances are heated to the same temperature they may radiate the same amount of energy and still differ considerably in brightness. Many substances are "selective" in their absorbing and radiating properties. One may radiate more luminous energy and less infra-red energy, and for another the reverse may be true. The former would appear brighter than the latter. The scientific worker in light-production has been searching for such "selective" radiators whose other properties are satisfactory. The rare-earths possess the property of selectivity and are fortunately highly refractory. Welsbach used these in his mantle, whose efficiency is due partly to this selective property. Recent work indicates that much higher efficiencies of light-production are still attainable by the principles involved in the gas-mantle.

Turning again to flames, another interesting physical phenomenon is seen on placing solutions of different chemical salts in the flame. For example, if a piece of asbestos is soaked in sodium chloride (common salt) and is placed in a Bunsen flame, the pale-blue flame suddenly becomes luminous and of a yellow color. If this is repeated with other salts, a characteristic color will be noted in each case. The yellow flame is characteristic of sodium and if it is examined by means of a spectroscope, a brilliant yellow line (in fact, a double line) will be seen. This forms the basis of spectrum analysis as applied in chemistry.

Every element has its characteristic spectrum consisting usually of lines, but the complexity varies with the elements. The spectra of elements also exhibit lines in the ultra-violet region which may be studied with a photographic plate, with a photo-electric cell, and by other means. Their spectral lines or bands also extend into the infra-red region and here they are studied by means of the bolometer or other apparatus for detecting radiant energy by the heat which it produces upon being absorbed. Spectrum analysis is far more sensitive than the finest weighing balance, for if a grain of salt be dissolved in a barrel of water and an asbestos strip be soaked in the water and held in a Bunsen flame, the yellow color characteristic of sodium will be detectable. A wonderful example of the possibilities of this method is the discovery of helium in the sun before it was found on earth! Its spectral lines were detected in the sun's spectrum and could not be accounted for by any known element. However, it should be stated that the spectrum of an element differs generally with the manner obtained. The electric spark, the arc, the electric discharge in a vacuum tube, and the flame are the means usually employed.

The spectrum has been dwelt upon at some length because it is of great importance in light-production and probably will figure strongly in future developments. Although in lighting little use has been made of the injection of chemical salts into ordinary flames, it appears certain that such developments would have risen if electric illuminants had not entered the field. However, the principle has been applied with great success in arc-lamps. In the first arc-lamps plain carbon electrodes were used, but in some of the latest carbon-arcs, electrodes of carbon impregnated with various salts are employed. For example, calcium fluoride gives a brilliant yellow light when used in the carbons of the "flame" arc. These are described in detail later.

Following this principle of light-production the vacuum tubes were developed. Crookes studied the light from various gases by enclosing them in a tube which was pumped out until a low vacuum was produced. On connecting a high voltage to electrodes in each end, an electrical discharge passed through the residual gas making it luminous. The different gases show their characteristic spectra and their desirability as light-producers is at once evident.

However, the most general principle of light-production at the present time is the radiation of bodies by virtue of their temperature. If a piece of wire be heated by electricity, it will become very hot before it becomes luminous. At this temperature it is emitting only invisible infra-red energy and has an efficiency of zero as a producer of light. As it becomes hotter it begins to appear red, but as its temperature is raised it appears orange, until if it could be heated to the temperature of the sun, about 10,000°F., it would appear white. All this time its luminous efficiency is increasing, because it is radiating not only an increasing percentage of visible radiant energy but an increasing amount of the most effective luminous energy. But even when it appears white, a large amount of the energy which it radiates is invisible infra-red and ultra-violet, which are ineffective in producing light, so at best the substance at this high temperature is inefficient as a light-producer.

In this branch of the science of light-production substances are sought not only for their high melting-point, but for their ability to radiate selectively as much visible energy as possible and of the most luminous character. However, at best the present method of utilizing the temperature radiation of hot bodies has limitations.

The luminous efficiencies of light-sources to-day are still very low, but great advances have been made in the past half-century. There must be some radical departures if the efficiency of light-production is to reach a much higher figure. A good deal has been said of the firefly and of phosphorescence. These light-sources appear to emit only visible energy and, therefore, are efficient as radiators of luminous radiant energy. But much remains to be unearthed concerning them before they will be generally applicable to lighting. If ultra-violet radiation is allowed to impinge upon a phosphorescent material, it will glow with a considerable brightness but will be cool to the touch. A substance of the same brightness by virtue of its temperature would be hot; hence phosphorescence is said to be "cold" light.

An acquaintance with certain terms is necessary if the reader is to understand certain parts of the text. The early candle gradually became a standard, and uniform candles are still satisfactory standards where high accuracy is not required. Their luminous intensity and illuminating value became units just as the foot was arbitrarily adopted as a unit of length. The intensity of other light-sources was represented in terms of the number of candles or fraction of a candle which gave the same amount of light. But the luminous intensity of the candle was taken only in the horizontal direction. In the same manner the luminous intensities of light-sources until a short time ago were expressed in candles as measured in a certain direction. Incandescent lamps were rated in terms of mean horizontal candles, which would be satisfactory if the luminous intensity were the same in all directions, but it is not. Therefore, the candle-power in one direction does not give a measure of the total light-output.

If a source of light has a luminous intensity of one candle in all directions, the illumination at a distance of one foot in any direction is said to be a foot-candle. This is the unit of illumination intensity. A lumen is the quantity of light which falls on one square foot if the intensity of illumination is one foot-candle. It is seen that the area of a sphere with a radius of one foot is 4p or 12.57 square feet; therefore, a light-source having a luminous intensity of one candle in all directions emits 12.57 lumens. This is the satisfactory unit, for it measures total quantity of light, and luminous efficiencies may be expressed in terms of lumens per watt, lumens per cubic foot of gas per hour, etc.

Of course, the efficiencies of light-sources are usually of interest to the consumer if they are expressed in terms of cost. But from a practical point of view there are many elements which combine to make another important factor, namely, satisfactoriness. Therefore, the efficiency of artificial lighting from the standpoint of the consumer should be the ratio of satisfactoriness to cost. However, the scientist is interested chiefly in the efficiency of the light-source which may be expressed in lumens per watt, or the amount of light obtained from a given rate of consumption or of emission of energy. This method of rating light-sources penalizes those radiating considerable energy which does not produce the sensation of light or which at best is of wave-lengths that are inefficient in this respect. That radiant energy which is wholly of a wave-length of maximum visibility, or, in other words, excites the sensation of yellow-green, is the most efficient in producing luminous sensation. Of course, no illuminants are available which approach this theoretical ideal and it is not likely that this would be a practical ideal. Under monochromatic yellow-green light the magical drapery of color would disappear and the surroundings would be a monochrome of shades of this hue. Having no colors with which to contrast this color, the world would be colorless. This should be obvious when it is considered that an object which is red under an illuminant containing all colors such as sunlight would be black or dark gray under monochromatic yellow-green light. The red under present conditions is kept alive by contrast with other colors, because the latter live by virtue of the fact that most of our present illuminants contain their hues. It is assumed that the reader knows that a red object, for example, appears red because it reflects (or transmits) red rays and absorbs the other rays in the illuminant. In other words, color is due to selective absorption reflection, or transmission.

Perhaps the ideal illuminant, which is most generally satisfactory for general activities, is a white light corresponding to noon sunlight. If this is chosen as the scientific ideal, the illuminants of the present time are much more "efficient" than if the most efficient light is the ideal.

The luminous efficiency of the radiant energy most efficient in producing the sensation of light (yellow-green) is about 625 lumens per watt. That is, if energy of this wave-length alone were radiated by a hypothetical light-source, each watt would produce 625 lumens. The luminous efficiency of the most efficient white light is about 265 lumens per watt; in other words, if a hypothetical light-source radiated energy of only the visible wave-lengths and in proportions to produce the sensation of white, each watt would produce 265 lumens. If such a white light were obtained by pure temperature radiation—that is, by a normal radiator at a temperature of 10,000°F., which is impracticable at present—the luminous efficiency would be about 100 lumens per watt. The normal radiator which emits energy by virtue of its temperature without selectively radiating more or less energy in any part of the spectrum than indicated by the theoretical radiation laws is called a "black-body" or normal radiator. Modern illuminants have luminous efficiencies ranging from 5 to 30 lumens per watt, so it is seen that much is to be done before the limiting efficiencies are reached.

The amount of light obtained from various gas-burners for each cubic foot of gas consumed per hour varies for open gas-flames from 5 to 30 lumens; for Argand burners from 35 to 40 lumens; for regenerative lamps from 50 to 75 lumens; and for gas-mantles from 200 to 250 lumens.

In the development of light-sources, of course, any harmful effects of gases formed by burning or chemical action must be avoided. Some of the fumes from arcs are harmful, but no commercial arc appears to be dangerous when used as it is intended to be used. Gas-burners rob the atmosphere of oxygen and vitiate it with gases, which, however, are harmless if combustion is complete. That adequate ventilation is necessary where oxygen is being consumed is evident from the data presented by authorities on hygiene. A standard candle when burning vitiates the air in a room almost as much as an adult person. An ordinary kerosene lamp vitiates the atmosphere as much as a half-dozen persons. An ordinary single mantle burner causes as much vitiation as two or three persons.

In order to obtain a bird's-eye view of progress in light-production, the following table of relative luminous efficiencies of several light-sources is given in round numbers. These efficiencies are in terms of the most efficient (yellow-green) light.

The luminous efficiency of a light-source is distinguished from that of a lamp. The former is the ratio of the light produced to the amount of energy radiated by the light-source. The latter is the ratio of the light produced to the total amount of energy consumed by the device. In other words, the luminous efficiency of a lamp is less than that of the light-source because the consumption of energy in other parts of the lamp besides the light-source are taken into account. These additional losses are appreciable in the mechanisms of arc-lamps but are almost negligible in vacuum incandescent filament lamps. They are unknown for the firefly, so that its luminous efficiency only as a light-source can be determined. Its efficiency as a lighting-plant may be and perhaps is rather low.