ANYONE WHO HAS TRIED TO construct a homemade gas laser that operates in the visible spectrum knows how difficult and expensive it can be to carry out the ambitious project. Even the popular helium-neon laser can easily be beyond the reach of an amateur. The laser tube requires precise glasswork, for example, and the dielectric mirror that reflects light back through the tube to maintain lasing is often beyond the budget of an amateur.

Now Martin Gosnell of Charmhaven, near Sydney in Australia, has sent me plans and instructions that will enable an adept amateur to build a copper chloride laser without great expertise and at a reasonable cost. The lasing element is copper vapor, which is generated when the first of a pair of electrical discharges is sent through the laser; the second discharge, about 150 microseconds later, causes the vapor to lase with a pulse of green and yellow light. Gosnell is able to fire the laser at a rate of up to 50 times per second, so that the output appears to be continuous.

Although the project is certainly quite challenging, little specialized glasswork is required, and the light production is so strong that an aluminum-coated microscope slide can be substituted for the dielectric mirror. (A strong word of caution: the high-voltage discharges involved in the lasing are lethal, and so no one should consider building the laser who has not had substantial experience with high-voltage circuits.)

Before I get to the details of construction, I should explain how it is that a copper vapor can be made to lase. The simplified chart in the upper illustration at the left indicates some of the energy levels that a copper atom is permitted by quantum mechanics to occupy. The atom is initially in its "ground state," or at the lowest energy level. When a discharge runs through the vapor, an electron in the current collides with the atom and transfers enough energy for the atom to jump to a higher energy level to be in an "excited state." Two pairs of excited states are shown on the chart. (The levels in each pair differ slightly in energy, for reasons that I shall not consider here; I have labeled each pair arbitrarily rather than with reference to their true spectroscopic designations.)

An excited atom can "de-excite" to a lower level by spontaneously emitting a photon, which carries away the energy the atom loses in making the jump. Because the energy levels that are allowed an atom are preset, the energy of the photon is restricted to certain fixed values. The light that is emitted is often pictured as being a wave instead of photon, in which case the wavelength associated with the wave is restricted to certain fixed values. For example, if a copper atom jumps from B2 to A1 in the chart, it emits light with a wavelength of 510.6 nanometers, which is green. A jump from B1 to A2 releases light with a wavelength of 578.2 nanometers, which is yellow.

An excited atom can also de-excite in a process known as stimulated emission. Suppose that a copper atom in the B2 state is passed by a wave with a wavelength of 510.6 nanometers- just the wavelength the atom would emit with a spontaneous jump to A1 [see Figure 2]. The passing wave interacts with the atom, forcing it into the jump as if on command. In this process the passing light is said to stimulate the emission, preventing the atom from making some other jump, such as to the ground state, or even from making the same jump later on. The wave emitted by the atom reinforces the original wave because they are identical in wavelength, travel in the same direction and are "coherent," or locked in step, and so the resultant light is brighter than the original light. Similarly, an atom that is initially in the B1 state can be stimulated into jumping to the A2 state by light with a wavelength of 578.2 nanometers. In each case the pair of levels involved in the stimulated jump is referred to as the laser pair, because they are the basis of the laser's emission of light.

In a copper-vapor laser the idea is to excite the atoms into the B states by bombarding them with the electrons in an electrical discharge. Some of the excited atoms will then happen to jump spontaneously down to the A states and fortuitously emit waves (or photons) along the tube. That emitted light is quickly reinforced by the chain reaction of stimulated emission it sets up when it passes other atoms in the B state. In the wave picture of light, the wave grows stronger; in the photon picture, the number of photons increases. The light that reaches a mirror at one end of the tube is largely reflected back through the atoms for another go at stimulating jumps. The light that reaches the opposite end of the tube, which lacks a mirror, escapes as the laser beam.

When gas lasers first appeared in the early 1960's, a copper-vapor laser was an intriguing prospect because it promised to be more efficient than other gas lasers. One reason for the expected efficiency is the fact that what I have called the B levels are not high above the ground state, so that not much energy should be required for atoms to reach them. Research soon revealed a major drawback, however: a copper vapor had to be heated to a temperature of about 1,500 degrees Celsius if it was to lase.

In 1973 researchers discovered that when a copper halide such as copper chloride was substituted for the pure copper in the original design and a series of double-pulse discharges was sent through the tube, the required temperature was a more attainable 400 degrees C. The success was attributable to the role of the pair of discharges. The first discharge dissociates the molecules while also exciting and ionizing some of the released atoms. If the second discharge is delayed long enough for the copper atoms to settle back to the ground state but not long enough for them to recombine with the chloride, it excites the copper atoms into the B states, just as in the original design with pure copper as the source of the vapor. The double-discharge technique was an excellent idea, but it required expensive switches, a double-pulse signal generator and other costly electronics. What Gosnell managed to do was to build a copper chloride laser with inexpensive and more readily available parts.

The core of Gosnell's laser is a quartz tube that passes through a furnace fashioned from alumina-silica furnace bricks [see Figure 3]. The tube, 55 centimeters long and with a one-centimeter bore, extends into brass posts at each end of the furnace. The posts, each standing about four centimeters beyond the oven, are 2 5 millimeters square and 12 centimeters high The tube is sealed into the holes in the posts with a silicone sealant.

On the opposite side of each post, the laser is extended with a 13-centimeter-long aluminum tube; it cools and condenses the internal vapors so that they do not reach the optical elements, which are at the far ends of each aluminum tube. At one end of the laser an aluminum-coated microscope slide serves as a mirror. (The mirror is mounted with its reflecting surface on the exterior.) At the other end an uncoated glass slide allows the laser beam to escape from the tube. (Were the vapors to condense on either slide, the lasing action would be eliminated.) To mount each slide so that it could later be adjusted, Gosnell devised an assembly of plates that are separated by an O ring and held together by three screws. He bored a hole through the assembly and then, with the sealant, glued the assembly up against the end of the aluminum tube and fixed the slide over the exterior of the hole.

When the laser is fired, the discharge through the tube runs between the brass posts, which are electrically connected to the power supply. Current is delivered to each post by a strip of thick aluminum foil, which is connected to the post by a large, spring-loaded clamp of the kind that normally clips together sheets of paper. Vertical holes drilled into the posts serve as ports through which one end of the tube is linked to a vacuum pump and the other end to a tank of helium. The pump is needed to remove air from the tube and to draw in the helium. The helium has two functions. In the segment of the tube between the oven and a post, where the vapor may not be abundant, the helium helps to conduct the electrical discharge. The helium also promotes the condensation of the vapors in the outer regions of the tube by colliding with vapor atoms

and removing their energy. Gosnell favors the alumina-silica -bricks because they are easy to cut, but he suggests that other high-temperature confinement materials might be tested. The essential characteristic of any confinement is that when the furnace is heated, the temperature along the confined length of the tube should be as uniform as possible. Gosnell heats the furnace with a common electric heating element that can be purchased from a supplier of home electrical parts. The heating element is inserted into a quartz tube that runs through the oven, parallel to the lasing tube and about 25 millimeters away from it. (Closer spacing would invite arcing between the two tubes.) Insulating fiber is used to seal off the holes in the bricks where the tubes enter the oven.

The helium is pulled into the laser by an ordinary single-stage vacuum pump, but refrigerator compressors working in tandem might be adequate substitutes. A valve inserted into the hose connecting the laser and pump allows one to close off the pump. A mercury manometer is also inserted to monitor the helium pressure, which was kept at about two torr.

The reader will recall that the lasing action depends on a pair of closely timed discharges through the vapor. Rather than purchase expensive switches and a pulse generator, Gosnell built a mechanical switching device that he calls a pulser [see illustration below]. On one side of the device two bars serve as electrodes. They are separated by a short gap from a plastic disk, into which an aluminum strip is sunk. The disk is mounted on an aluminum shaft, which is rotated at about 6,000 revolutions per minute by a belt and motor at the opposite side of the device. Each bar electrode is connected to a charged capacitor, which is also connected to one of the brass posts supporting the laser tube. The aluminum strip and the shaft and its mount are electrically connected to the other brass support post.

As the disk rotates and the sunken strip approaches the tip of one of the bars, the capacitor connected to that bar discharges across the gap-and so also through the laser tube. Another discharge takes place when the strip approaches the other bar, which is connected to the other capacitor. The time between the discharges is set by the relative location of the bars and the rotational speed of the disk. The time should be about 150 microseconds, but the optimum value depends on the temperature of the copper vapor and related parameters.

Gosnell suggests that the strip should be flush with the disk face to eliminate the possibility of its catching on a bar's tip during rotation. He also advises that the pulley belt that connects the motor to the shaft should not be conducting, as many common belts are. (They are designed that way so that they bleed off any electrostatic buildup, but such a belt will short out the pulser.) The spacing between the tips of the bars and the disk is usually a few millimeters, but the optimum spacing can be determined only experimentally.

The power-supply circuit for the laser is shown in the illustration on the opposite page. At the left, plugged directly into the house electrical supply, are two identical neon-sign transformers rated at 15 kilovolts AC at 60 milliamps. (One transformer may well provide enough current.) The current from the transformers is rectified by a series of high-voltage diodes and fed to two storage capacitors, C1 and C2. Positioned along the way are two optional capacitors that smooth the current supply and allow current to be drawn regardless of the particular phase of the alternating current. Gosnell says the laser will fire well enough without these extra capacitors.

The resistance represented by R1 and R2 in the illustration consists of 1,500-ohm resistors rated at two watts and connected in a series-parallel combination. The C1 and C2 capacitors are made of flat, alternating layers of aluminum foil and plastic sheets and have a face area of about 1,500 square centimeters; Gosnell used polyester or polyethylene for the plastic component to attain a capacitance of about 15 nanofarads.

In Gosnell's setup the capacitors are placed on a support just above the laser in order to minimize electrical problems that greater distance would create. (Each of the optional capacitors was similarly constructed of foil and plastic sheets but then was rolled up and inserted into PVC pipe, one meter long and 10 centimeters in diameter, that was mounted below the laser.) The pulser sits on a rigid, insulating piece of thick plastic just above the capacitors. The strips of thick aluminum foil that connect the capacitors, pulser and laser posts are all about five centimeters wide. The inductance of the capacitors and of the circuit between them and the laser tube must be low so that the current in the discharge increases sharply, dissociating the copper chloride molecules and exciting the atoms abruptly.

To guard against accidental rupture of the rotating elements of the pulser, Gosnell erected a thick plastic shield in front of the pulser. To decrease the danger of electric shock, he connected high-voltage "bleeder" resistors across each capacitor to drain their charge when the system was turned off. (Here again I must warn of the danger implicit in the lethal currents that run through the power supply and laser, which can cause trouble if a charged capacitor is touched even after the system is turned off.)

The current for the heating element was controlled by a Variac, a variable transformer. For an alternative control, Gosnell connected a second heating element outside the furnace to the internal one and then attached one of the leads from the electrical source to the second element with an alligator clamp. By varying the location of the lead along the second heating element, he could control how much of the second element was in the circuit, thereby varying the resistance in the circuit and consequently the heat within the oven. With either technique of control, he usually heated the furnace to 390 degrees C, as read with a thermocouple he placed within it.

To align the mirror on the laser tube so that it reflected directly back along the tube, Gosnell sighted through the opposite end from a distance of about a meter while an assistant adjusted the mounting screws on the mirror platter. When Gosnell spotted a reflection of his eye at the center of the mirror, the alignment was correct. (It should go without saying that he never looked into the laser when it was firing. Laser bursts can cause severe damage to the retina.)

To check for an air leak in the tube, Gosnell disconnected the electric circuit and then connected a neon-sign transformer between the brass posts. After pumping down the tube and flushing it with helium several times, he filled it with helium to a pressure of about 10 torr and plugged in the transformer. When the system was free of leaks, the discharge through the tube was whitish gray; a pink tint indicated a leak.

Gosnell prepared his copper chloride by heating about 1/4 teaspoon (roughly one milliliter) of the crystals in a chemical hood to yield a greenish-brown liquid. (The hood is mandatory because breathing the vapor is harmful.) After the material cooled and solidified, he hammered it into a fine powder, which he sealed in a desiccating container. Both the heating and the desiccation serve to remove water and excess halide.

When he was ready to operate the laser, he put the powder into the central part of the tube through an opened end with a long, thin "spoon" he had fashioned. (He points out that the transfer would be easier if the tube were outfitted with a vertical section through which the powder could be poured. The extra section would extend out of the furnace and could be sealed with a rubber bung.)

The furnace was heated for about an hour before laser operation to stabilize the temperature. During that period Gosnell pumped down the system and flushed it with helium several times. After a final check on the pressure and the spacing between the bar electrodes and the plastic disk in the pulser, he turned on the motor that drives the pulser. In his initial trials the bursts of laser light were not continuous, but some experimentation with the gas pressure, the furnace temperature and the locations of the bar electrodes in the pulser eventually produced a more reliable output of bright green and yellow light on a card placed in the beam.

Gosnell, ever modest, suggests that someone who is particularly skilled in the construction of homemade lasers might well improve on his design.

Bibliography

EFFICIENT PULSED GAS DISCHARGE LASERS. W. T. Walter, N. Solimene, M. Piltch and G. Gould in IEEE Journal of Quantum Electronics, Vol. QE-2, No. 9, pages 474-479; September, 1966.

DOUBLE-DISCHARGE COPPER VAPOR LASER WITH COPPER CHLORIDE AS A LASANT. C J. Chen, N. M. Nerheim and G. R. Russell in Applied Physics Letters, Vol. 23, No. 9, pages 514-515; November 1, 1973.

A PARAMETRIC STUDY OF THE COPPER CHLORIDE LASER Noble M. Nerheim in Journal of Applied Physics, Vol. 48, No. 3, pages 1186-1190; March, 1977.

RESONANCE RADIATION TRAPPING EFFECTS IN COPPER AND MANGANESE LASERS. K. Stigouri, S. Ramaprabhu and T. A. Prasada Rao in Journal of Applied Physics, Vol. 61, No. 3, pages 859-863; February 1, 1987.

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.