Few amateurs own spectrographs, because instruments of reasonably good quality cost several hundred dollars. It is now possible, however, for an amateur to build an inexpensive spectrograph. Sam Epstein, chief chemist of the Federated Metals Division of the American Smelting and Refining Company in Los Angeles, has recently designed an instrument of excellent quality that can be built at home for less than $100. With it experimenters can readily identify approximately 70 chemical elements listed in the periodic table, sometimes even if their presence in a mixture of substances amounts to no more than a few parts per million. In addition the apparatus can be adapted for use with telescopes in analyzing phenomena on the sun, although its size limits its application to permanently mounted telescopes.

Epstein writes: "Physically all spectrographs consist essentially of three elements: a narrow slit, through which passes the light that is to be analyzed; a dispersing element, which may be either a glass prism or a grating that consists of a pattern of closely and uniformly spaced lines ruled on either a transparent or a reflecting surface, and a camera. For a number of years all spectrographs were based on an optical principle first described by Isaac Newton. When a narrow beam of sunlight passes through a glass prism in a darkened room, a pattern of rainbow colors forms on the opposite wall. In the instrument based on this principle a narrow beam is formed by passing light rays through a slit. Diverging rays from the slit are made parallel by a lens; after passing through the prism they are focused on a screen by a second lens. The screen can be replaced by photographic film or the colors can be observed directly with a small telescope. The resulting spectrum consists of a band of adjacent colored lines that are images of the slit.

"Good instruments of this type are difficult to build because they require lenses of high optical quality that focus all colors equally. Moreover, the resolving power, or ability of the instrument to separate the closely spaced lines, increases with the size of the prism, and so does the cost. Glass is opaque to ultraviolet rays-those wavelengths shorter than 3,300 angstrom units constituting the part of the spectrum that is most useful for identifying unknown atoms.

"For these reasons gratings made by ruling lines on the surface of a concave mirror have largely replaced prisms in spectrographs. Gratings used in general-purpose spectrographs may have 30,000 or more evenly spaced rulings cut on a surface of polished metal only an inch wide and two inches long. Such gratings are costly. In recent years, however, methods have been developed for making plastic duplicates of the rulings. These replicas are then mounted on the concave surface of aluminized glass mirrors. As assembled in the instrument the grating receives light either directly from the slit or after reflection by a mirror and focuses the lines of color on a photographic film or plate. All rays are reflected without significant absorption The resulting spectrum spans some 30 octaves, including the single octave of light.

"An easily constructed spectrograph that can serve as a powerful analytical tool in many fields of experimentation is based on a discovery of the physicist Henry A. Rowland. In the l9th century he observed that if a concave grating, a slit and photographic film are placed on a circle equal in diameter to the radius of curvature of the grating, the diffracted rays come to a focus on the film.

"Unlike the prism, the grating presents the spectrum simultaneously at a number of positions. At one point on the circle the undispersed image of the slit appears. This narrow image is flanked on both sides by a series of spectra, the ends of which may overlap more or less depending on the design of the grating and the angle at which the incoming rays impinge on the grating. These images are known as spectral orders, the undispersed image being designated as the "zeroth" order and the flanking spectra as the first, second and third orders and so on. In general the spectrum of the first order is the shortest and brightest. By cutting the rulings at a certain angle, however, it is possible to construct gratings that reflect most of the light into a particular order. The spacing between the spectral lines increases in proportion to the increased length of the higher orders. "If the spacing between the rulings of the grating is known, as well as the angle at which incoming light rays fall on the grating, the angular position at which light of any color will come to a focus on the circIe can be computed simply. An example can be given for a grating ruled with 15,000 lines per inch. A wave of light 6,300 angstroms in length falling on the grating at an angle of 19 degrees with a line perpendicular to the face of the grating will be diffracted at an angle of 2.9 degrees from the perpendicular and on the same side of it. A wavelength of 2,300 angstroms striking the grating at an angle of 19 degrees will be diffracted at 10.9 degrees on the opposite side of the perpendicular. "If the 15,000-line grating is sup ported on the face of a concave mirror with a radius of curvature of 100 centimeters, these two lines, which spar 4,000 angstroms, will be separated by a distance of 25.4 centimeters at the plane of the film. That is about 16 angstroms per millimeter. This resolving power is adequate for the analysis of most metallic substances except those that emit a large number of closely spaced lines; examples are iron and the rare-earth elements. A replica grating of this size can be obtained from the Edmund Scientific Co., Barrington, N.J. 08001. Th catalogue number is 50,220. "Begin the construction by drawing on a flat surface a circle with a radius of 53 centimeters. On this 'Rowland circle' locate the exact positions of the slit, grating and camera or film holder as specified in the accompanying illustration. The outline of the spectrograph housing, also shown in the illustration, should be superposed on the circle. The dimensions of the housing are not critical but construction problems will be minimized if they are followed closely. The light baffle not only prevents scattered light from entering the camera and fogging the film but also serves as a support for the top of the housing. All interior parts are painted black to minimize unwanted reflections from the various surfaces.

"The camera consists of a lightproof box that encloses a film holder curved to match the circumference of the Rowland circle [see Figure 5]. The curved members of the box are wooden arcs of slightly smaller radius than the Rowland circle. This difference in radius compensates for the thickness of the film holder and thus ensures that the emulsion of the film follows precisely the curvature of the Rowland circle.

"Each of the wooden arcs is attached to the camera support by wood screws, and additional support is provided by the steel angles. The machine screws needed to attach the film holder are fastened in place with epoxy cement. The several brass strips that constitute the film holder are also assembled with epoxy cement. Coat the surfaces as indicated and mount the assembly on the camera by means of the nuts. When the cement has hardened, the holder will have assumed permanently the shape of the Rowland circle. Do not neglect to install black felt or velvet for blocking light at the indicated points where the removable slides provide access to the film.

"The camera slides up and down on ways. Light is admitted to the film through the thin horizontal slot in the mask of sheet metal. Altering the vertical position of the camera in its ways enables the experimenter to record several spectrograms on a single sheet of 35-millimeter film. An indicator on the racking mechanism marks the position of the film with sufficient accuracy so that successive exposures can be separated by .5 millimeter. An internal shutter, which can consist of a hinged flap of sheet metal, must be installed in order to close the entrance slit through which light is admitted to the spectrograph. The shutter can be operated either electromagnetically or by a mechanical shaft that extends through the side of the housing.

"The slit consists of a pair of safety-razor blades [see Figure 6] Two slit assemblies should be made, one with a spacing of about 50 microns for use with intense light sources and another of 150 microns for observing flames. To set the 50-micron spacing loosen the screws that clamp the razor blades, slip a sheet of notepaper between the cutting edges, press the edges snugly against the paper and tighten the screws. Use a stack of three or four sheets for adjusting the wider slit.

"As I have indicated, spectral lines shorter than 4,000 angstroms are of most interest in spectroscopic analysis. They must be photographed because the eye is insensitive to this part of the spectrum. The Eastman Kodak Company manufactures two special emulsions for spectrographic work: Spectrum Analysis No. 1, SA 421-1, and Spectrum Analysis No. 3, Sp 421-1. Both are sensitive down to about 2,200 angstroms. The upper limit of the No. 1 emulsion is about 4,500 angstroms (deep blue); that of No. 3, about 5,300 angstroms (green) Panchromatic film must be used for recording the complete visible spectrum The special spectrographic emulsions are sold in 100-foot rolls that cost about $12 each. Single rolls can be bought from Spex Industries, Inc., 3880 Park Avenue Metuchen, N.J. 08841.

"Unless the experimenter insists on achieving the best possible results the extreme portion of the ultraviolet range, ordinary black-and-white emulsions can be used. The usual precaution should be taken to avoid accidental exposure when loading the film holder. The films should be developed for high contrast. I use Kodak D-l9 developer. The negative is used for analysis. No positive print is required. When making an exposure, pull the opaque slide out just far enough to clear the film aperture of the holder and remember to push it back again before removing the holder from the camera.

"Atoms emit bright-line spectra only when they are excited in the gas or vapor state. At relatively low temperature, however, such as the temperature of an ordinary gas flame, this applies to only a few elements, including potassium, lithium, calcium, barium, strontium and copper.

"Some 25 additional metals can be added to the list by substituting a flame of acetylene and compressed air for natural or manufactured gas. The results of flame excitation can be improved by preparing the specimens in the form of a chloride salt solution. The solution is sprayed into the flame with an atomizer during the exposure.

"The most widely used sources of excitation are electric arcs and sparks. The arc is struck between rods of highly purified graphite specially manufactured for spectrographic work and is sustained by a direct current of about eight amperes. Electrodes of spectrographic grade can be bought from the Ultra Carbon Corporation, Post Office Box 747, Bay City, Mich. 48706. An inexpensive power supply for operation from a 110-volt, 60-cycle power line can be constructed by connecting as a full-wave rectifier four silicon diodes of 10-ampere capacity that are designed for 200-volt operation. These can be bought from most dealers in radio supplies. A resistor must be connected in series with one side of the power line to limit the current. It can consist of a replacement unit of the type used in radiant heaters. A 750-watt unit will do, as will a group of seven 100-watt incandescent lamps connected in parallel.

"The carbon electrodes of the arc must be supported in vertical alignment by a pair of insulated jaws that can be moved together for striking the arc and then separated as soon as the carbon rods become hot enough to maintain an arc. A rack-and-pinion mechanism is the preferable means of moving the jaws. Small used arc lamps easily modified for supporting the carbons vertically can be obtained from dealers in theatrical supplies. The arc should be fully enclosed except for an exit window that will allow a beam about an inch in diameter to fall on the slit of the spectrograph. An experimenter must protect his eyes by wearing welder's goggles when the arc is in operation.

"The material to be analyzed is applied to the tip of the lower carbon rod, which is made the cathode. An axial hole half the diameter of the carbon rod can be drilled to a depth of about an eighth of an inch in the tip of the cathode for admitting powdered specimens. A shallow cut by a saw can also be used. Fluid specimens can be applied to the rod by dipping the tip of the cathode into the solution and letting it dry. Highly volatile specimens should be mixed with powdered graphite to retard the rate at which they evaporate in the arc. Some highly refractory specimens that volatilize at a rate too slow for analysis can be mixed with ammonium chloride. Heat decomposes this salt, which then carries the unknown material into the arc.

"To adjust the spectrograph for operation remove the slit from the instrument and substitute a slit about one millimeter wide made by placing two strips of masking tape over the slit aperture of the housing. Cut a strip of white paper 32 centimeters long and 10 centimeters wide. This strip is used as a temporary screen. Bisect it lengthwise with a heavy black line and make two vertical lines to indicate the ends of the spectrum aperture. Tape the strip to the inside of the camera section. Center the horizontal line with respect to the opening. Darken the room and direct a flashlight beam into the slit. Adjust the grating so that the resulting spectrum is centered vertically and positioned at the long-wavelength end of the screen. Remove the paper.

"Place a few large crystals of rock salt on a metal screen above the flame of a gas burner and focus the resulting yellow light on the slit with a magnifying lens. The burner should be about 18 inches from the slit. Tape a piece o translucent wax paper three inches wide across the camera opening so that it covers the S,893-angstrom point, which is about 2.6 centimeters from the long wavelength end of the opening. A broad yellow line will appear on the screen. Adjust the grating until this line occupies the 5,893-angstrom position.

"Remove the paper and install the permanent slit. Load the camera and place it in position on the instrument. Rack the film holder to either the top or the bottom. Charge the bottom electrode with iron filings. Put the arc about 12 inches from the slit, aligned so that the beam transmitted by the slit floods the grating. Make exposures by striking the arc and opening the shutter. The correct exposure is the one that results in the greatest range of line density. It must be determined experimentally.

"The line images may be fuzzy. If so, the instrument should be focused. Set the grating at one of its limits of travel, either toward or away from the film. Then make a series of exposures, advancing the grating in equal increments toward the limit of its travel after each exposure. Keep a written record for correlating the successive positions with the exposures.

"After developing and examining the film, set the grating at the position that yields the sharpest image. If close examination discloses that each spectral line actually consists of a closely spaced group of fine lines, the slit is not parallel to the rulings of the grating. This can be corrected by rotating the slit in its oversized mounting holes. The instrument is now ready for use.

"Analysis of spectrograms requires the precise measurement of the line positions. The measurements can be made conveniently by projecting an image of the lines on a screen along with that of a scale calibrated in angstroms. A standard 3S-millimeter projector can be modified for projecting the spectrogram and scale simultaneously.

"The standard used for calibrating the scale can be a spectrogram of carbon. Strike an arc between a pair of clean spectroscopic carbons and make an exposure. The most prominent features on the resulting spectrogram will be the two cyanogen (CN) bands starting at 3,883 and 3,590 angstroms and the 2,479-angstrom carbon line [see Figure 1].

"Lay the film on a flat surface and measure the distance between the 2,479-angstrom line and the head of the band beginning at 3,883 angstroms. Assume that it is 88 millimeters. If so, each millimeter represents 16 angstroms (3,883 - 2,479/88 = 16). One hundred angstroms will occupy 100/16, or G.25, millimeters, and a scale that is 25 centimeters long will span 4,000 angstroms.

"Draw the scale on white cardboard, divide it into 40 increments of 100 angstroms each, photograph the drawing and make a positive transparency with an enlarger. The final image should be precisely 25 centimeters long. Draw an auxiliary scale 12.5 centimeters long and divide it into 20 equal increments. The ratio of this scale to the one previously drawn is 20 :1; when the scale previously drawn is magnified 20 diameters by projection, the distance between adjacent graduations is 12.5 centimeters and spans 100 angstroms. Line positions can be measured to within one angstrom by placing the auxiliary scale between a pair of projected graduations.

"To determine positions of the lines emitted by an unknown chemical element, make a spectrogram of carbon and on the same film a spectrogram of the unknown element. Insert the film in its compartment in the holder and move the spectrogram with respect to the scale until the 3,883-angstrom point of the scale coincides in position with the 3,883-angstrom cyanogen band head. The distance between the projector and the screen must of course be adjusted for a magnification of 20 diameters.

"Using the auxiliary scale at the screen, measure in angstroms the spectral positions of the lines you believe to be those emitted by the unknown element. Always look for the most prominent lines of an element when trying to establish its presence in the sample. Extensive tables of wavelengths can be found in The Handbook of Chemistry and Physics. At least two lines of an element must be detected before it can be considered as being definitely present. Not even the roughest estimate of its presence can be made, however, unless the spectrogram can be compared with one of a similar type of sample material containing a known amount of the element in question.

"Imperfections in a grating can produce 'ghost,' or spurious, lines. Ghosts are easily spotted. They always occur in telltale pairs spaced symmetrically on opposite sides of the parent, or true, line, which is usually of high intensity. Ignore the ghosts.

"The spectrograph can be used for the analysis of many different types of material such as glasses, soils, ceramic materials and ashed substances of all kinds. It is particularly applicable for detecting fractions in concentrations of 1 percent to .0001 percent and even less in some cases. Amateur prospectors, for example, can readily check rocks for metals in amounts far below those required for ordinary chemical or blowpipe analysis.

Bibliography

CHEMICAL SPECTROSCOPY. W. R. Brode. John Wiley & Sons, Inc., 1943.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.