Brief History of the Atlas of Microscopic Particles
By John Gustav Delly
Microscopists have been identifying particulate matter since the seventeenth century. Reference sets of study slides, and identification keys and even atlases to specific groups of substances were prepared throughout the eighteenth and nineteenth centuries, particularly during the latter half of the nineteenth century and the early twentieth century, but none of these attempted to be comprehensive. Over a half century ago, Dr. Walter C. McCrone, namesake of McCrone Associates, knew that the microscope could be used more extensively in air pollution control – a public health concern, which was then drawing considerable attention. Airborne particles, of course, consist of every conceivable type of substance, necessitating some knowledge of mineralogy, biology, chemistry and industrial processes. An attempt to identify all airborne particles would also require an identification scheme embracing as many optical, physical, and chemical characteristics as possible, as well as carefully-prepared data-retrieval methods for presentation of the results. In 1950, Dr. McCrone managed to get the support of MAPPA, the Midwest Air Pollution Prevention Association, to test the possibility of using the microscope for the general identification of all particle types. A final report to the Association in 1952 confirmed the practicality of the idea applied to air pollution particles, but also indicated that the results could have far wider application in other areas as well, including industrial hygiene, manufacturing, pharmaceuticals, cleanroom monitoring, and criminalistics (trace evidence analysis).
The Particle Atlas, First Edition, one volume, 1967
During the next several years, a variety of industrially sponsored projects at Armour Research Foundation (later, Illinois Institute of Technology Research Institute, IITRI) provided more background in both technique development, and additional substances characterized. In 1958, the United States Public Health Service (USPHS) began a three-year program at McCrone Associates, during which samples of known atmospheric particles (coal and petroleum flyash, etc.) were submitted by members of the Air Pollution Control Association. These known particulate substances were added to the growing collection of reference samples, and considerable characterization work was done on these and other substances. USPHS support ended in 1962 with the work well advanced, but not ready for publication. In the next four years, various additional government and industrial sponsors made it possible to complete and publish The Particle Atlas in 1967 (W.C. McCrone, R.G. Draftz, and J.G. Delly, Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan). This one-volume, 406-page photomicrographic atlas described and illustrated 404 particulate substances. The computer-assisted analytical instruments we have today were not available then; we had to rely more on optical crystallography, microchemistry, and experience. The Atlas consisted of photomicrographs and descriptions, together with analytical tables.
A word about the photomicrographs is relevant in contrasting the illustrative methods used in the present Atlas. Originally, Dr. McCrone wanted the photomicrographs to be stereo pairs, in which one of the pair would be the particle with top light; together the pair would show the particles in 3-D, and by alternately closing each eye, one would see the particle in top light, or transmitted light alone. The cost of over 800 four-color separation illustrations put an end to that idea; for economic reasons, we were pretty much restricted to one photomicrograph – rarely two – for each particle.
What lighting condition should be chosen? This was not a trivial consideration when one realized that some samples could contain isotropic, anisotropic, and opaque components: photograph in ordinary transmitted light, and you can’t tell isotropic from anisotropic; photograph using crossed polarizers, and you can’t see isotropics, opaques, and anisotropic particles at extinction; and in neither case will you see the surface characteristics of opaque components. It was decided that the best compromise was to use slightly uncrossed (9°-12°) polarizers. Every effort was made to maintain image fidelity to the sample. First of all, I installed a constant-voltage transformer at the wall outlet and all microscope lamps were operated from this source – it would take voltage fluctuations of ± 15%, yet maintain constant output voltage. The rheostats on the microscope illuminators were set to 12 volts, and then never touched; the lamps were all turned on and off by foot switches; constant voltage was thus assured to maintain correct color temperature. Quartz-halogen lamps were not then installed in microscopes; Kodak light balancing filters were carefully selected for use with fresh tungsten lamps. Plan-apochromatic objectives were just being introduced, and these were employed; for top light, where more working distance was needed, plan-fluorite objectives were used. I extensively tested every 35-mm color film available at that time, hoping to use the slowest-speed film, with lighting conditions of slightly uncrossed polarizers. Unfortunately, the film I wanted to use recorded the neutral gray background of slightly uncrossed polarizers as green – and not a very pleasant green. We ended up calling the exact shade of green “sick green”; it took a CC20 magenta color-correction filter to just make up for the green, but then all whites in the sample were magenta. The best film at the time turned out to be Agfa CK-20. A custom professional processor was selected to develop the transparencies, to ensure that all of the developing solutions were fresh and tested.
The Particle Analyst – an interim solution
The 1967 first edition of The Particle Atlas sold out rather quickly, as it proved beneficial at a time when the pharmaceutical industry finally was persuaded to become more concerned about particles in parenteral solutions, and other products. The same applied to the food processing industry. Newly designed integrated circuits and satellite components were found to be particle sensitive, and cleanrooms were introduced into all of these industries. Demand for extensive expansion of The Particle Atlas followed, and the interim solution was the publication The Particle Analyst, in which articles and particle descriptions continued every two weeks during 1968. At the same time, work was started on a multi-volume second edition. Scanning electron microscopes, electron microprobes, and energy and wavelength detectors were being introduced, and it was decided to include these. The new work would be in four volumes: Volume 1 would be a theory/technique book; Volume 2 would be a photomicrographic atlas, with descriptions; Volume 3 would be the same coverage of particles as in Volume 2, but all samples examined by scanning or transmission electron microscopy, with elemental data; Volume 4 would contain the analytical tables, glossary, bibliography, and index.
The Particle Atlas, Second Edition, six volumes volumes 1-4, 1973, volume 5, 1979, volume 6, 1980
For the second edition of The Particle Atlas, we decided to change the lighting conditions, because anisotropic particles at extinction could not be distinguished from isotropic particles using slightly uncrossed polarizers [for those interested in all of the considerations, see “Photomicrographic Documentation of Mixed Isotropic/Anisotropic Specimens,” John Gustav Delly, Journal of Biological Photography 54 (4) 157-160 (October 1986)]. The only way to show everything in one-color photomicrograph was to use modified circularly-polarized light. To produce circularly-polarized light, the polarizers are crossed, and then two quarter-wave plates are placed in crossed position, at 45° to the crossed polarizers; in practice, one quarter-wave plate is placed in the usual accessory slot, and the second is introduced beneath the specimen, but above the polarizer, 90° from the upper quarter-wave plate – the microscope used at the time had a compensator slot in the sub-stage position. Under these lighting conditions, there is no extinction, and, thus, all anisotropic particles show interference colors, which do not go to extinction on rotating the stage. If the quarter-wave plates are matched, however, the background is black, in which case isotropics and opaques cannot be seen. To lighten the background, different quarter-wave material was tried in the sub-stage location until one was found to mismatch by just enough to make the background gray. By this time, the Agfa CK-20 film used for the first edition, was no longer available in this country. Another round of extensive film tests was made, with no film found to be ideal for recording using the new lighting conditions. With the pressure of publication deadline, the film closest to best was selected, a slow-speed Kodachrome Type A. The gray background was recording too dark, however, and the solution was found to be a very short exposure of ordinary light superimposed (double exposure) on the circularly-polarized light exposure. For opaque samples, and samples with opaque components, top light needed to be added as an additional third exposure on the same frame; the top light was achieved through the use of two American Optical Company Universal Illuminators, with a carefully selected light-balancing filter pack. All of the samples in the first edition were rephotographed, and their number extended to 609.
Even before the 1973 publication of this 1138 page, four volume Atlas, it was realized that continuing demand made two additional volumes imperative; the proposed Volume 5 was to be an extension of Volume 2, and Volume 6 was to be an extension of Volume 3. Volume 5, pages 1145-1454, was published in 1979, and Volume 6, pages 1457-1703, was published in 1980. The number of particles described now numbered 1022. The six volumes comprising The Particle Atlas were sold individually for $90.00 per volume. Volume 2, then Volume 5, the two photomicrographic volumes, were the first to go out of print. More than 4000 copies of the six volumes were sold before the last were sold in 1990. It was when another print run was to be made that we learned that the printing plates had been “lost,” along with the original photomicrographs. Today, on the Internet, these volumes, used, are being offered for $400 each.
The Particle Atlas, Electronic Edition, on CD-ROM, 1992
In about 1990, Steve Shaffer of MicroDataware offered to republish all six volumes of The Particle Atlas on CD-ROM. Fortunately, at Dr. McCrone’s original request, I had made several duplicate exposures while photographing each sample, so these were available for producing the electronic edition, PAE2, which appeared in late 1992. This electronic edition filled the immediate demand, at a time when CD-ROMs were just being introduced, and gaining in popularity. Dr. McCrone reviewed this electronic version favorably [see “The Particle Atlas, Electronic Edition,” Walter C. McCrone, American Laboratory, pp. 39-40, 42-44 (April 1993)], citing particularly the improvement in rendition of the sample. The printed version, by the nature of the color-printing process, had color decisions applied uniformly over several pages, and it seems, at times, as though achieving a uniform gray background took precedence over high-order interference colors, so that these pale colors were over-exposed on, for example, nylon fiber, which appears white. The same photomicrograph on the CD-ROM version shows the correct fifth- and sixth-order interference colors. Additionally, the electronic version, seen on a computer screen, appears to be lighted by transmission, rather than reflection from the printed page, and so more closely resembles the image seen when looking through a microscope.
McCrone Atlas of Microscopic Particles, an online database, 2005
Both the printed editions and the electronic edition of The Particle Atlas have proven themselves useful to analytical microscopists for 36 years. Nevertheless, the restriction to a single photomicrograph per particle because of the costs involved has always been viewed as a severe restriction; besides, in the working laboratory, circularly-polarized light is rarely used for particle identification – that was a presentation expedient. What would the optimum presentation consist of, if there were no color printing cost restrictions? That question brings us to the present edition, whose name has been changed from The Particle Atlas to the Atlas of Microscopic Particles to more accurately indicate the true nature of the Atlas.
There are now no restrictions as to the number or kind of illustrations used to accompany the new descriptions. This means that every image can be recorded in every way in which it is actually seen in real life analysis. In the past, I would have been restricted to one photomicrograph in circularly-polarized light, and a description length no longer than the side dimension of the photomicrograph. The dark circularly-polarized light was especially unsatisfactory for the isotropic components. Now, however, we can start with plane-polarized light, and record the very beautiful shades of yellow/orange/amber in the case of, for example, the Cochineal sample. Then, we can cross the polarizers and see which, if any, component is anisotropic – and record the results. Are the whites high or low order? Put in the first-order red compensator, and record the results. Is some feature better seen with higher magnification? Try it, and record the results. Fluorescence is no longer restricted to an entry in a table in another volume; try it, and record the results. Does the color bleed if water is added? Try it, and record the results. There are two components to this Cochineal sample; can any additional features be revealed by looking at the sample with the scanning electron microscope? Try it, and record the results. While we are at it, let us determine the elements that are present through energy dispersive spectroscopy (EDS), but now, instead of recording the results in a different volume (to keep costs down, the black-and-white images are printed in a different volume), we can add them directly to our on-going characterizations in the same sample file. Seeing that the results of the EDS indicate that we are dealing with two organic components, we can examine the sample next, using Fourier transform infrared spectroscopy (FTIR), and record the resulting spectra right along with the other data. In a similar way, we can keep adding data until we feel we have sufficient data to characterize the sample and to distinguish it from all other particle types, plus the option to make additions, deletions, corrections, etc. at any time.
There are so many other advantages to an on-line atlas. One big advantage is that we no longer have to wait for a specific number of samples to be completed before publishing; completed sample characterizations can be posted, as they are completed; corrections can be made instantly; additional facts or illustrations can be added or deleted; references can be cited within a description – it’s all so wonderful. And the best part is that the Atlas of Microscopic Particles has the potential of being perpetual! As new materials are manufactured, or new drugs introduced, they can be added; visitors to the on-line atlas may wish to submit particles that have not yet been included. Over time, and as needed, samples from the printed and CD-ROM editions will be replaced with better examples, and/or updated with new characterizations.
There is an IMPORTANT note that must be inserted here on the proper way to view the photomicrographs in this Atlas. To begin with, the particles included are being characterized using an Olympus BX-51 Polarizing Microscope, equipped with Plan-Apochromatic objectives; strain-free Plan-Fluorites are used for quantitative conoscopy; Long-Working-Distance objectives are used where top light is needed; condensers are both strain-free type, and achromatic-aplanatic; an epi-fluorescence attachment has been added for examinations using 365 nm ultraviolet light from a high-pressure mercury-vapor lamp. Top light is achieved by using a Volpi Intralux® 6000-1 halogen light source, with bifurcated fiber-optic bundles, terminating in a focusing lens. The microscopical images are being recorded digitally with carefully chosen components.
The resolving power of the objective with the highest numerical aperture relative to its magnification, and the resolving power of the digital camera’s CCD at the primary image have been calculated, and we find the camera to be adequate for the microscope optics. The images are recorded and saved as TIFF images; when printed out on photo-grade paper at maximum resolution settings of a good printer, the images will be faithful representations of the microscopical image. The severest limitation will occur when the atlas images are viewed on your monitor screen – this happens to me too when I am focusing an image on the screen of my laptop. When looking at a sample with 1 µm anisotropic specks embedded in an isotropic matrix, the image on the monitor does not show the individual “pin-points” of light – it can’t. Try this: take the eyepiece out of your microscope, turn it upside down, and place it close to your eye; move in to your monitor screen and examine a white area; you are using your eyepiece as a highly-corrected magnifier, and you will see the phosphor-deposited areas making up your screen; you will see red, green, and blue dots, or squares, or rectangles, or vertical bars depending on your monitor type. A 72 dots-per-inch screen will look coarser than a 92 dpi screen, etc. Ideally, you need the highest quality, highest resolution monitor you can afford, and you need to carefully adjust your monitor’s settings for color temperature and color corrections; the Brightness and Contrast settings need to be optimized. Due to the settings of different quality monitors, with different phosphor types and designs, the color images may only depict approximate representations of what should be seen through your microscope. This limitation will be well known to experienced computer users; the prints, however, should be fine.