Fe techn) do with the nature of the lens. oI ‘Astronomy is based primarily oa ate ae cag CSN eee eae ead ope thatwill:make visible By Roger C. Baker Jt Photograph by George H Mayer ae ‘Apr, 1991, Svence PROBE! 53The Homemade MICROSCOPE forming the front lens of a micro scope objective. Anton van Leeuwenhoek (1632-1723), the fa ‘mous Dutch maker of microscopes, was the first to dis: cover a means of grinding even ‘smaller high-power lenses, but ne kept his techniques secret. Con temporary optical artisans are nonetheless able to make similar lenses with simple, traditional equipment. Given this fact, itis strange that few investigators have sought to duplicate the optical technology that allowed Leeuwenhoek to dis: cover bacteria with a simple hand-held microscope. The fol lowing discussion attempts to help fill this conspicuous tech: nical gap. True, it is probably not practical to handcraft a lens with the full capabilities of a good color-corrected, oil-immersion ob- jective. One can, however, come reasonably close for only a few dollars and end up with some: thing that can be slipped in the pocket and is also fun to build The path of exploration de- scribed in the pages that follow is, based on both simple and com: pound objective lenses used without an eyepiece lense. such Tenses, when they are held close to the eye, are the basis for com: pact, high-resolution microscopes of the sort that Leeuwenhoek use. Lenses used in this way give a very bright and clear image, Superior in some ways to that of @ conventional microscope. Micro scopics should strive for better resolution rather than greater magnification. A large but dim and fuzzy image Is no improve: ment over a small bright clear image The Holy Grail of amateur mi croscopy might be considered Leeuwenhoek's feat of observing well-defined bacteria darting about in a sample of water. The limit of resolution of a simple spherical or biconvex lens is» about one micrometer, just suffi cient to reveal many bacteria as fuzzy moving specks. (One micron is one millionth of a meter or one thousandth of a millimeter. One inch is 25.4 millimeters.) The reso: lution needed to observe bacteria with some clarity, about one-half micron, is not possible with a single lens. Instead, two lenses 54 Science PROBE!, Apri 1991 are required. They form what is known as the compound objective lens. The Options in Lens Design Jo understand why compound lenses are necessary for the best resolution, we must first under Stand that the resolution of even the most perfectly shaped micro Scope lens is limited by the angle Of the cone of light (Figure 1) that itis able to gather from the object being viewed. The larger the angle of the cone at its apex, the more light gathered. This rela tionship only became clearly Understood after 1873 when Ernst Abbe, a German optician, pub- lished the theory linking numerical aperture, the measure Of this angle, and the theoretical resolution of microscope objec: tives, The larger the angle, the greater the aperture and the bet ter the resolution. The theory states that the maximum resolu- tion of a lens is proportional to its numerical aperture divided by its light wavelength. To improve reso- lution further, we have a choice of two directions to follow. We can either increase the numerical ap: erature or shorten the wavelength of light we use ‘an increase in the numerical aperture can be accomplished with a pair of lenses, known as a ‘compound-lens objective. Ama: teurs can achieve good results with a combination of two plano: convex lenses, further assisted by the use of the immersion princi: ple described below. Compound objective lenses with high-numerical apertures generally take the form of a hemi spherical plano-convex front lens mounted with its flat side toward the object to be viewed and one or more lenses of lower power mounted above, toward the eye Even if the front lens is incapable of completely bending the cone of light completely inward to create a focused image under these conditions, the second lens mounted close above it can com: plete the job. Such a combination acts to shorten the effective focal length of the front lens, thereby allowing it to capture a wider cone of light. This increases the resolving power of the comb! tion. When the greatest magnifica tion is not required, single lenses are appropriate. These may be spherical, biconvex or plano-con: vex. It Is physically impossible tor a single lens to resolve details Smaller than about one micron. Eitfieien t observe ving bacte fia but not very clearly When they sre ophertcal, Such single lenses tend to nove a narrow field of View in other words a clear zone Inthe center of the viewing area Surrounded by a blurry peripheral region. Nevertheless, 9 pocket mr Cfescope with such single spherical lens is very easy £0 thake snd isa good frst step for the amateur For many purposes, itis useful toenlarge the eid of view more than fully spherical lens wl allow. One way to achieve this goals to start witha glass capil {ary and biow a biconvex lens. A Biconver liquid glycerin lens is another faily eSsy way fo give a wide Feld Sucha lone Gan also be Sealed to give a more or less permanent plano-convex glycerin fons with similar advantages The most sophisticated route forthe amateur isto grind glass Tenees in much the seme way that eeuwenhosk might have done te according to 2 method | have recreated ficonvex and plano- ewvex Ifiees ground andl pol 'shed in this way have the most accurate curvatures and thus Give the best results of any meth: a when lenses with radius ofcurvature of less than about one millimeter are not required, Sim ple lenses of these types are very suitable for general purposes due to the clear image and wide field Whenever the very greatest reso lution is not required. The illumination requirements are easily met with a number of sim: ple arrangements, The most highly curved lenses, which give the highest magnifica tion, are easiest to prepare using the inherent surface tension on a droplet of glycerin or on a tiny bead of molten glass. The surface tension acting on a droplet of fluid forces the droplet into a sphere. The sphere would be per. fect if made in space where there Js no gravitational pull, or near perfect when the droplet is so small that the surface tension is extremely large for the mass it has to confine. Tiny raindrops are spherical because of surface ten The reader may wonder what role the conventional compound microscope plays in this discus: sion. In addition to the objective lens, this type of microscope in corporates an eyepiece lens and sometimes a field lens mounted along a common axis in a tube Viewing the image Is easy but the instrument is bulky. Our various types of objective lenses could be used in such a microscope, but the image quality ultimately de. pends on the objective lens, which is our major topic. How Compound Lenses Are Made and Used ‘As we have seen, the oreatest resolving power is obtained with a compound lens. Unfortunately, if such a lens is used dry, part of the cone of light entering the lens arrives at too great an angle (the critical angle) is reflected away from the surface of the lens and lost For this reason, the very great. est resolving power is attained in high-resolution, compound micro- scope objectives by the ‘immersion principle’ in which an optical coupling layer of liquid is, placed between the object being viewed and the glass surface of the front lens. This significantly increased the numerical aperture of the lens, If the liquid is an oll with the same refractive properties as glass, the end result is identical to embedding the observed ob. ject inside the glass of the lens, enabling the top surface of this front lens to refract and bend inward all or most of the entire cone of light that reaches the surface of the lens. Liquids other than immersion oil, such as water, will work almost as well ‘50 much for theory. The best results | have achieved so far are from the combination of a small radius, nearly hemispherical, plano-convex front lens, of either the glass or liquid variety, in we contact with a microscope slide A second ground and polished plano-convex glass lens of lesser curvature is mounted imme: ein ~« SPACER ssembly nomemade compound lens Giately above it with the curved side of both lenses facing the eye. Figure 1 shows this arrangement. The two lenses are mounted so close that they nearly touch. With the help of a thin metal spacer and a few dabs of silicone rubber, the spacing is maintained. It ap: pears possible to approximately duplicate Leeuwenhoek’ best ob- servations with such a lens combination. The practical bene: fits of a water meniscus beneath the front lens and the slide are due to a complex combination of a decrease in light losses from reflection, a reduction in spher. ical aberration, and an increase in numerical aperture. If the plano-convex lenses are both made of glass ground and polished according to a method ‘similar to that Leeuwenhoek used (described below), the radius of curvature of the hemispherical front lens, and thus the magnify. ing power of the combination, cannot be made so great as when the front lens is a drop ofF glycerin. itis not hard to make a glycerin hemisphere about one half millimeter in diameter, whereas a corresponding glass lens cannot be made less than about two millimeters in diame- ter. Therefore objects such as bacteria are easier to observe