MANY AMATEURS have built a cloud chamber and used it to detect the subatomic particles emitted by the isotopes in a speck of radium paint. Few, however, have undertaken other atomic experiments, primarily because of the hazards associated with radioactive materials. Radioactive substances are indeed hazardous. On the other hand, some of the most fascinating experiments in the field of atomic physics can be conducted with harmless substances. For example, molecules of common table salt can be split apart and formed into high-velocity beams that can be manipulated in interesting ways by electric and magnetic fields. Atoms in the beam can be separated into isotopes. The isotopes can be graded according to their mass and identified.

The apparatus required for doing such experiments is reasonably simple and can be built at home, much of it from cast-off odds and ends. For example, Roger Flood of Natick, Mass., recently devised an apparatus for generating a slender beam of sulfur molecules and detecting it chemically. His device was based on an apparatus developed in 1911 by the French physicist Louis Dunoyer for generating a beam of sodium molecules.

From a piece of glass tubing about eight inches long Dunoyer made a vessel in the form of three interconnecting bulbs. The vessel resembled an hourglass with three bulbs instead of two. A bulb at one end contained a small lump of sodium. When Dunoyer pumped the air from the vessel and heated the sodium, a dark spot appeared on the wall of the bulb at the opposite end. The spot was in perfect alignment with the small openings that interconnected the three bulbs. The position and size of the spot could be explained only on the assumption that molecules of evaporated sodium had traveled in straight lines through the openings to form a beam of particles that condensed on the glass.

Flood's version of the generator was made largely of metal. "My apparatus," he writes, "is identical in principle with Dunoyer's. It consists of a brass tube about half an inch in diameter and four inches long that is divided into three compartments by two brass disks that serve as partitions. A pinhole in the center of each disk interconnects the three compartments. The disks fit snugly in the tube and are soldered in place half an inch apart near the middle of the tube. The ends of the tube are closed by plastic plugs removed from a pair of UG-680/U coaxial-cable connectors. O rings inserted between the shoulders of the plugs and the ends of the pipe make the assembly vacuum-tight.

"The compartment at one end serves as an oven for evaporating sublimed sulfur. The perforated disks and the space between them function as a collimator: only the molecules that follow a straight path through the perforations are transmitted. The beam is detected in the third compartment, where molecules of sulfur react with a plate of-hot copper to form a black deposit of cupric sulfide. This method of detecting a molecular beam is approximately a million times more sensitive than mechanical techniques of depositing a spot of sulfur thick enough to be visible to the unaided eye.

"The sulfur is evaporated by an electrically heated coil inside a tube of Pyrex glass. The tube is supported by one of the coaxial connectors [see illustration at left]. The coil is made from 30-gauge Nichrome wire, whose ends are crimped to copper leads. The outer end of the glass tube is sealed with epoxy cement and the interior is partly filled with a paste of water glass and talc that serves as insulation. The inner end of the tube is open. It is charged with a few milligrams of sulfur just before the apparatus is assembled for operation.

"A similar tube heats the copper target, which is a small piece of polished copper foil crimped in the form of a cup over the end of the glass. The proper operating temperatures are determined by experiment. Sufficient voltage should be applied to the vaporizing coil to melt sulfur in air within one minute. Similarly, sufficient voltage should be applied to the target coil to produce a thin coating of copper oxide in air within three minutes.

"The apparatus must be exhausted to a pressure of at least 10 torr. This pressure is just above the capability of mechanical air pumps. For the mechanical pump I used the compressor of a discarded home refrigerator, which I modified according to instructions in The Scientific American Book of Projects for the Amateur Scientist (Simon and Schuster, 1960). In series with the compressor I connected an inexpensive oil diffusion pump that is available from Morris & Lee Company, 1685 Elmwood Avenue, Buffalo, N.Y. 14207. The combination exhausts the apparatus to about 10 torr in 10 minutes.

"I apply a thin coating of vacuum grease to the O-ring seals. After the system has been exhausted to the limit of the pumps I heat the target for about five minutes and turn on the oven heater to vaporize the sulfur. A perceptible spot should form on the target within 15 minutes. I then turn off the oven heater but keep heating the target for another five minutes to ensure that the decaying beam continues to react with the copper. Air is admitted after the apparatus has cooled to room temperature, a procedure that prevents the target from oxidizing. The target is now examined. If all has gone well, a black spot with sharp edges is seen on the target. A fuzzy spot or random sputtering usually indicates that air is leaking into the system or that the pumps are not working properly.

"Molecular beams of many chemical elements can be detected by their reaction with appropriate target materials, but I have experimented with only two other combinations. A beam of hydrogen reduces yellow molybdenum trioxide to blue molybdenum dioxide. Oxygen reacts with lead monoxide, which is pale yellow, to form lead dioxide, which is brown. I admit the gases to the heater tube through a small leak and operate the pumps continuously to maintain the desired low pressure."

BY replacing the collimating apertures of Flood's apparatus with a pair of slits and passing the beam first through an electric field and then through a magnetic field, atoms constituting the beam can be separated into isotopes. The resulting apparatus is appropriately called a mass spectrometer because it sorts and separates isotopes according to their mass. In the December 1963 issue of the American Journal of Physics the late John W. Dewdney of Harvard University described the construction of a mass spectrometer that he assembled from copper tubes, darning needles, razor blades, a miniature incandescent lamp and other inexpensive parts. In l963 the complete instrument cost less than $5, not counting the magnet and vacuum pumps.

The substance to be analyzed is vaporized by the incandescent filament of a small lamp bulb. Adjacent to the filament is a pair of metal plates and a pair of facing razor blades spaced to form a narrow slit. A potential difference of about 100m volts is applied to the plates and the filament. The resulting electric field accelerates the ions and focuses them on the slit between the blades.

Ions pass through the slit as a diverging beam and enter a magnetic field placed at a right angle across the beam. The field creates a deflecting force that always acts at right angles to the direction in which the ions move. The force continues to act at right angles even though the path of the ions curves continuously. As a result the ions move in a circular arc. Overlapping arcs of equal radius intersect at two points. Hence the diverging beam of ions, which escape through a slit and enter a uniform magnetic field, converges at a distant point. In effect the magnetic field acts as a lens that focuses ions from a source slit onto an object slit [see illustration at right].

In Dewdney's mass spectrometer the magnet bends the path of the ions through an arc of 90 degrees and brings them to focus on an object slit. The object slit is also made of razor blades. Ions that pass through the object slit strike a collector electrode, to which they transfer charge. A continuous stream of ions that reaches the collector electrode induces a continuous current in the output circuit. The intensity of the current is a measure of the number of ions that impinge on the collector.

The magnetic field does not deflect all ions equally. Ions of low mass curve more sharply than those of higher mass. (In much the same way, although for different reasons, a simple lens bends short waves of light more sharply than long waves.) Moreover, ions that move at low velocity are deflected more than those that move at high velocity. The velocity of ions emerging from the source slit is determined in part by the potential difference between the filament and the focusing plates. By changing this potential one can alter the influence of the magnetic field on the path of the ions. In effect the focal length of the "magnetic lens" can be adjusted by adjusting the accelerating voltage.

Dewdney altered the accelerating voltage continuously by connecting an alternating-current transformer in series with the source of accelerating voltage. This arrangement automatically focuses ions of differing mass sequentially on the object slit and causes corresponding currents to appear sequentially in the output current. The current can be applied, after appropriate amplification, to the vertical plates of an oscilloscope. The same continuously varying potential can be applied to the horizontal plates of the oscilloscope. The combination causes the oscilloscope to display a trace in the form of a jagged line, each peak of which corresponds to a species of isotope in the beam. With this technique Dewdney made a graph of the isotopes of potassium [see illustration at left]. The graph displays two peaks. The shorter one shows the relative abundance in the beam of potassium of mass 41, and the larger peak displays the more abundant isotope of mass 39.

The vacuum chamber of the spectrometer and the ports through which air is exhausted are assembled from copper plumbing parts. The chamber consists of a pair of 1/2-inch copper T's; the crossarms are coupled by a 90-degree elbow of 1/4-inch diameter through reducers that were sawed off flush with the end of the T's. The legs of the T's are coupled to the vacuum pumps by 1/2 inch copper tubing [see below right]. All joints are soldered.

The functional parts of the instrument, which include the filament, focusing plates, slits and collector electrode, are supported by No. 3 steel darning needles pushed through No. 1 rubber stoppers that fit the open ends of the T's. Dewdney warned that the installation of the darning needles is the most difficult part of making the instrument, and I agree! The needles must be positioned as accurately as possible and parallel to the axes of the stoppers. Dewdney made a jig for guiding the needles into the rubber. It is a thick cap of metal drilled with holes that make a sliding fit with the needles. The small end of the stopper is pushed into the cap. The needles are pushed through the holes and the rubber. One must hope they remain parallel.

I did the job by chucking the needles in a drill press, standing the stoppers on a block of wood that rested on the bed of the press and lowering the chuck until the points emerged from the rubber and penetrated the wood. Each needle was then pushed farther through the rubber by hand. Eight needles evenly spaced in a circle 3/8 inch in diameter are needed for the source assembly. Three needles evenly spaced across the diameter of the second stopper support the object slit and the collector electrode [see illustration at lower left].

The filament used for evaporating specimen materials is removed from a low-voltage incandescent lamp, such as a No. 41 pilot lamp or a No. 63 automobile lamp. File a scratch about halfway around the glass at the upper edge of the base and, while wearing rubber gloves, grip the bulb with one hand, the base with the other and pull. Do not attempt to "bend" the glass. The filament will remain intact when the glass breaks if the lamp has not been used. With pliers crack the remainder of the bulb from the leads, but do not remove the colored glass bead that ties the leads together.

To mount the filament to the stopper assembly, thread the connecting leads through the eyes of the needles and pull the needles into the rubber until the tips of the eyes are flush with the surface. Clip off the excess wire and press the colored bead into contact with the rubber at the center of the stopper, being careful to preserve the shape of the filament. The focusing electrodes can be made of any thin sheet metal, such as brass shim stock. Mount the electrodes as shown in the accompanying illustration, lower right.

Make the slits out of a double-edged razor blade broken into four rectangular pieces, each with a cutting edge. Clean both sides of each piece, coat the sides with soldering flux and sandwich each pair of pieces, with the cutting edges facing, between No. 12 brass washers of 1/2-inch diameter. Clamp the sandwiches with alligator clips. Adjust the space between the cutting edges to form a slit .005 inch wide. Use another razor blade as a feeler gauge to make the adjustment. Solder the assembly while it is clamped. Break off portions of the soldered blades that extend beyond the edge of the washers.

Drill a hole that makes a snug fit with the needles in the center of a 1/8-inch copper pipe cap, turn the cap so that the open end faces away from the small end of the stopper and solder the tip of the middle needle into the hole. This completes the collector electrode. In the same way drill a pair of holes spaced 3/8 inch apart near the edges of one of the slits and solder it to the remaining pair of needles. This completes the objective slit.

The space between the collector electrode and the slit can be adjusted readily by sliding the needles through the rubber. The dimension is not critical but should be approximately 1/8 inch. The source assembly is similarly mounted. I the needles are numbered clockwise from 1 through 8, needles 1 and 5 sup port the filament; 3 and 7, the source slit; 2 and 4, one focusing electrode, and 6 and 8, the remaining focusing electrode.

The magnetic field is supplied by permanent alnico magnet provided wit pole pieces in the shape of right cylinders. The pole pieces can be cut from mild-steel shafting and should be face squarely by a lathe. They should have a diameter of 7/8 inch. The air gap between the poles should be made as narrow as possible, just wide enough to admit the 90-degree elbow of the vacuum chamber. The radius of the circular magnetic field should equal the radius of the 90-degree elbow.

The magnetic field spreads outward as it emerges from the magnet, and its radius is accordingly longer than the radius of the pole pieces. Dewdney found experimentally that the diameter of the pole pieces is about 65 percent less than the diameter of the effective field. This information should be kept in mind if the dimensions of the spectrometer are modified.

Any alnico magnet can be used that develops a magnetic field in the air gap of 3,000 gauss or more. The magnet should be positioned as illustrated. Magnets that can be modified for this application are sold by the Edmund Scientific Co., Barrington, N.J. 08007.

Electrical connections to the source are made by pushing a miniature eight hole vacuum-tube socket over the needles. If you wish, you can force the large end of the rubber stopper into a 1/2-inch copper nipple, so that 5/8 inch of the small end of the stopper protrudes. The nipple provides a base for supporting the socket mechanically. The stopper that supports the object slit and collector can be equipped with a 1/2-inch to 1/4-inch reducer to serve as a mechanical base for a Type BNC connector that couples the collector electrode to the output circuit.

The electrical power supply provides a six-volt alternating current for heating the lamp filament and a pulsating direct voltage at a peak potential of up to 425 volts to the filament, the focusing plates and the horizontal plates (x axis) of the oscilloscope. The pulsations are developed by connecting the secondary winding of a 115-to-25-volt transformer in series with an external 400-volt direct-current power supply [see illustration at left]. The potential applied to both the oscilloscope and the focusing plates is controlled by potentiometers that also determine the polarity of the focusing plates with respect to each other and to the filament.

The procedure for adjusting and operating the instrument is determined in part by the nature of the substances to be analyzed. Alkaline metals ionize readily and are convenient for introductory experiments. This group of metals includes lithium, sodium, potassium, rubidium and cesium. Dewdney used the chlorides of these elements, particularly potassium chloride.

For an initial experiment dissolve as much potassium chloride as possible in distilled water at room temperature, place a drop on the filament and let it dry. Apply a film of vacuum grease to both rubber stoppers and insert them in the T's. Rotate the stoppers so that the slits are approximately parallel to the lines of magnetic force in the air gap. Exhaust the vacuum chamber to the limit of the pumps, at least to 10 torr. Turn on the power supply and adjust the filament to a potential of approximately 150 volts and the focusing electrodes to about 50 volts. The filament is now at a positive potential of approximately 100 volts with respect to the focusing plates and is of the same polarity as the ions.

Substances such as the alkali metals, which have small ionization potentials with respect to the work function of tungsten, emit positive ions. Substances of ionization potential higher than the work function of tungsten emit negative ions. The polarity of the filament must correspond to the polarity of the ions. (The like charges repel the ions, thereby speeding them toward the slit.)

Doubtless the oscilloscope will display a straight or nearly straight horizontal trace. Rotate the object slit back and forth through a few angular degrees. Stop when one or more small peaks appear in the horizontal trace and then increase and decrease the potential of the focusing plates to maximize the peaks. Interesting salts for experimentation include lithium sulfate, sodium sulfate, barium chloride and cesium chloride.

Bibliography

UNDERGRADUATE MASS SPECTROMETER. John W. Dewdney in American Journal of Physics, Vol. 28, No. 5, pages 452-456; May, 1960.

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.