Suggestions for setting up an Optics teaching laboratory
In Physics and Engineering, it is usually difficult to provide practical experience that will supplement and illustrate the theoretical side of studies while fitting in the available time and budget. The theory generally works so well that labs become 'verification of the formula' and give a thoroughly inaccurate impression of both applications of the theory and methods of research. Electronics and Optics are exceptions to the rule, since valuable, and even exciting, experiences can be provided with reasonable time and cost, and will show that there is more to the subjects than mere application of formulas. I want to give a few suggestions here on how to create an instructional optics laboratory at a professional, that is, university, level. The ideas can also be used, appropriately modified, for secondary education. In this case, I would warn against oversimplification of difficult pieces, and encourage retention of a rigorous presentation of those that are retained.
The first step is to collect a library of optical experiments. There are many sources, and the old ones are not inferior to the newer ones. These include, for example, the books by Hilton, Strong and B. K. Johnson. When you are familiar with the experiments, select the ones that seem the best for your purpose, and actually perform them, noting all points of difficulty, how they are affected by your particular apparatus, what has to be known in advance, important observations, and details of technique. Then, write up your own laboratory manual, modularized to suit your timetable, that includes necessary data and instructions, questions to exercise the understanding, and some alternatives to suit different interests. The difficulty of the procedures should be graded, and both easy and hard should be included. Later experiments should depend on fundamental earlier ones, but other than this should be independent. None of the experiments should be intellectually pointless, and all should emphasize observation rather than calculation. Balance is the key. Organize the laboratory book into modules so that it can easily be revised and changed. One four-hour period a week is better than two two-hour periods, and the laboratory should be available at other hours. Do not expect students to make up their own experiments from suggestions; it is too much of a wilderness to encounter for the first time without a guide.
The laboratory itself should be in a room from which all external light can be excluded. It need not be a darkroom, only dark enough that the eye can become somewhat dark-adapted when necessary. A basement room is excellent. Small 'desk' lamps should be provided for making notes, and occasionally as light sources. There should be a locked cabinet for equipment.
An optical bench is necessary to study optical elements and centred systems, polarization, and diffraction and interference patterns. It ought to be some 2 metres in length, with an accurate millimeter scale along it. A home-made bench can be satisfactory and much less expensive than a commercial one. Such a bench can be made by any metal shop from rod, channel, or angle section. The construction of carriers will be the most troublesome aspect of do-it-yourself. Basically, these should slide smoothly along the bench, able to be clamped without tilting them, with an index pointer to measure displacements on the scale, and accepting a standard post on the top, so that commercial holders can be used if desired. The bench should be supported on three feet that can be varied in height to level it.
Essential accessories to the bench are: (1) a low-power telescope (f.l. about 250 mm) with crosshairs and (2) a micrometer eyepiece; (3) a collimator with variable slit and reticle screen; (4) a low-power He-Ne laser with holder; (5) a beam expander and spatial filter; (6) a milk-glass screen for observing real images; (7) a pinhole, 0.5mm diameter; (8) an adjustable diaphragm; (9) several lens holders, allowing as much alignment freedom as possible. On at least two, the lens should have two degrees of rotational freedom and two degrees of translational freedom. There does not have to be much range of adjustment, which should be done by turning a screw with a screwdriver. (10) a holder for flat screens or reticles serving as objects; (11) an incandescent source of small size, such as the Pointolite of past times. A low-voltage automotive bulb may be acceptable. These items are generally useful, and some can be successfully improvised. You may require others for special experiments.
Workhorse light sources are a low-pressure Na arc, which gives bright, almost monochromatic light in the middle of the spectrum, and the low-pressure Hg discharge, which gives a few lines distributed over the spectrum. In addition to the laser, and the incandescent sources already mentioned, these will prove sufficient. The Na lamp is all but essential to a good laboratory.
The next most important major piece of equipment is a good prism spectrometer. This must be equipped with divided circles and verniers, a rotating prism table with levelling screws, a telescope with Gauss eyepiece, and a collimator with variable slit. The angular positions of telescope and collimator should be able to be read on the circles to an accuracy of 1' of arc. Many instruments are furnished that do not meet this specification, in a search for economy, but the economy is a false one. The major purpose of this spectrometer is, of course, not to observe spectra, but to make accurate angular measurements, and to teach the proper method of using an accurate instrument. The first experiment with the spectrometer should be how to adjust it so that the telescope and collimator are focused at infinity, and so that the faces of the prism are perpendicular to the direction of view. The refracting angle of the prism is then measured, and then the angles of minimum deviation for the lines in the Hg spectrum, from which a dispersion curve is plotted. The value of this experiment is evident. This value is seriously compromised if the experiment is not done correctly. It must be recognized that young eyes will never focus a telescope properly unless some other procedure than simply the sharpest image is adopted.
Third in importance is a Michelson interferometer. The problem here is the very accurate alignment that is necessary, which can be provided only by an expertly manufactured instrument. The Michelson gives a graphic illustration of interference, several kinds of patterns that beg for explanation, and a good feeling for the sensitivity of interferometry in measurement. Again, proper procedures of alignment are important. The classic experiment of calibrating a millimeter scale is, unfortunately, somewhat dull. Finding the separation of the Na doublet is more interesting, and involves a lot of exciting counting of fringes.
If you still have any money left for expensive equipment, a good travelling microscope is the next buy. This should be low power, and a micrometer eyepiece should be available. With this, you can measure slits, gratings and interference patterns. Finally, a 35mm camera is handy for recording images and for producing reticles, test patterns, and Fresnel screens. Such a camera is usually otherwise available, and there is no need having one specially for the optics laboratory. The same may be said for a photographic darkroom, which is very nice, but not absolutely essential. Common optical instruments, such as binoculars, telescopes and biological microscopes are useful as items of study, but not for general experiments.
The three items of bench, spectrometer, and interferometer provide the essentials of a teaching optics laboratory that will reward students with experiences they will not forget. Now we need to consider the smaller pieces, the optical components. First of all, we need a good 60° crown glass prism for the spectrometer, as well as a replica grating of a few hundred lines per mm. A prism of a different kind of glass, as well as an assortment of diffraction gratings, gives some selection. Screens are available with various arrangments of slits; at least a double slit should be obtained, perhaps a set with different spacings and widths. Optical flats, perhaps 75mm by 125mm are very useful; at least two should be on hand. One of these, and a slit, are all that is needed for Lloyd's Mirror, an excellent interference experiment. A small box of microscope cover glasses or slides is handy. A first-surface plane mirror is useful. Several achromatic lenses should be on hand, one perhaps 50mm diameter and 250mm focal length, one 25mm diameter and 50mm focal length, for example. They can be found as surplus offerings, and whatever is available is fine. Somewhere in the laboratory you should bring out clearly the superiority of achromats in practical applications over simple lenses. Sometimes projection lenses for copying machines, or similar, are offered; anything like this should be acquired. A number of simple lenses of various diameters, shapes, and focal lengths should be purchased. Reading lenses are a source of large simple lenses that have magnificent aberrations. At least one rather weak plano-convex lens should be available for Newton's Rings. A glass cell with parallel sides is useful for experiments on liquids, or for liquid filters.
Some materials that make interesting experiments, if you can find them, include calcite, mica and lycopodium powder. Calcite cleaves naturally into rhombohedra, and well-cleaved clear specimens of large size are available at mineral and gem shows. Its main appeal is double refraction. The mica should be clear muscovite in reasonably large sheets (25mm minimum dimension). It has the interesting property of being able to be split into very thin flakes. Lycopodium powder is the spores of a fern with the interesting property of being all about the same size, and so giving very clear diffraction patterns. It was used by pharmacists for coating pills, but this is no longer done. One should be on the lookout for such interesting materials that may become available.
If you wish to study polarization, a good deal can be done with inexpensive Polaroid filter material, which comes in sheets that can be cut as needed. Sheet plastics are often birefringent, since the long polymer molecules are aligned in the manufacturing process. Such materials can be mounted in 2x2 slide mounts for convenient handling and labelling. Plastics become birefringent when strained. For more than qualitative observations, it is necessary to have filters that can be rotated about the optic axis. The orientation need only be measured to about a degree, so these may be easy to improvise. This permits the study of rotation of the plane of polarization, polarization by reflection, double refraction, and polarization colours. Polarizing prisms are free from the colour distortion of Polaroid, but are very expensive. A Nicol prism of reasonable aperture is a real treasure for polarization studies. Quarter-wave plates for Na or He-Ne light are a good acquisition.
The quantitative measurement of intensity can be sacrificed in the teaching laboratory, and commercial equipment for it is rather costly. However, the availability of solid-state detectors (CdS cells, phototransistors, solar cells) and electronics opens a wide door to do-it-yourself apparatus, and makes this much cheaper and easier than it ever has been in the past. One might even think of motor-driven scanning and recording, using a computer for control and data acquisition. This is very much easier with the early DOS machines than with the current range of computers, which are extremely ill-suited. I would never throw away old PC's no longer up to Windows, but which have all the advantages of mass storage and an easy-to-use peripheral interface. Aside from time variation studies (as in fluorescence) and automatic recording, computers have little application in the optics teaching laboratory.
Any given list of experiments would not be suitable in all cases, and might include less important experiments while omitting more important ones, in the view of the user. Nevertheless, I want to give here a list I certainly would consider adequate for a first course in optics, consisting of traditional experiments of proved value. The patent aim of an experiment, expressed in its title, is not usually the real aim of the experiment, of course, only the vehicle for the actual aim, which is the development of the understanding and skills of the student. Each experiment requires about four hours work in the laboratory.
The home laboratory presents a somewhat different aspect, but can be very successful and entertaining, like the home electronics shop. The expensive instruments -- spectrometer, Michelson and so forth -- cannot be afforded without the chance of amortization over a number of students, but a few choice items can be acquired. An optical bench can be improvised from a metre stick. Carriers can be fashioned with hand tools, and Plasticine or Blu-Tack used to hold things in place. Most of the smaller items -- lenses, flats, mirrors, small prisms, gratings, magnifiers and so on -- are affordable. A little money can be spent on a laser, such as a surplus one from a supermarket scanner, which I use, and have provided with an optical bench made from aluminium channel. Any considerable amounts of money available should be reserved for things that measure (e.g., micrometer eyepiece) and monochromatic light sources (e.g., sodium lamp). Perhaps you might find a university physics department that has gone over to computer simulation and wants to sell the things it no longer knows how to use.
Composed by J. B. Calvert
Created 12 April 2000
Last revised 15 April 2000