The News: Oak Ridge National Laboratory. Friday, April 2, 1965, p.1.

Laboratory Scientist Draws "Atoms-In-Depth" Using Computer-Oriented Graphic Technique


A picture is worth a thousand words - particularly to an Oak Ridge National Laboratory scientist who has revived an old turn-of-the century household entertainment device to help visualize complex crystal structures.

Remember grandma's stereoscope and cabinet of stereoscopic pictures? They were found in nearly every parlor during the early 1900's. C. K. Johnson of ORNL's Chemistry Division has developed a modern computer program to utilize "automated graphic arts" in the drawing of similar stereoscopic (three dimensional) figures. But, instead of seeing Niagara Falls or a circus elephant, the drawings are of highly complex chemical and biological structures.

The precision of the computer and mechanical plotter used makes it feasible to produce accurate stereoscopic line drawings with great detail in just a fraction of the time normally required to handdraw figures.

The layman might ask, what is a stereo figure? It's a pair of drawings with one member for each eye. The drawings are perspective projections relative to the "view points" of the eyes. The brain fuses the separate figures automatically and the perception of depth results. This permits the true three-dimensional appearance of a crystal structure model to be visualized rapidly by the reader.

Johnson's research project began when he attempted to illustrate a technical paper with a picture of the thermal motion in magnesium citrate. A crude program was prepared to see if this could be better accomplished with a computer. The results were encouraging enough to warrant further research and the present, rather elaborate general computer program, ORTEP (Oak Ridge Thermal Ellipsoid Plot Program), evolved. It was largely a case of trial and error method to find a graphical technique applicable to automation and numerical techniques suitable for graphics.

The normal method for viewing a stereo figure is the "stereoscope", a simple combination of lenses or prisms. However, if the stereomates are close enough together, the stereo viewing can be performed by placing a divider, such as a piece of thin cardboard, between the eyes. (This can easily be accomplished with the line drawing accompanying this article.)

In the past, the optimal way of studying the three-dimensional character of a crystal structure has been to build a model (see accompanying picture). Scientists have been using such methods since the study of crystallography began. Another method has been to hand-drawn stereoscopic figures. This, however, requires very time consuming draftsmanship. For this reason, extensive use of such graphic aids techniques have not been made to any extent since the 1920's and 1930's when some work along these lines was done by pioneers in the field of X-ray diffraction crystallography.

A first application of stereoscopic drawing to crystal structures was made by the German crystallographer Max von Laue, who in 1926 edited an atlas of stereoscopic crystal structure drawings. This was followed by a second volume in 1936.

Automated graphic arts techniques are particularly well suited to the field of crystallography because the crystal structure models are defined in numerical terms. The parameters which define the model are normally obtained from an X-ray, neutron or electron diffraction study of crystalline material. Structure models are used in fields such as chemistry, solid state physics, biology, mineralogy, and metallurgy.

At present, the program for drawing the crystal structure illustrations is complete and its documentation in press. Soon, in answer to many requests, it will be distributed to various scientific personnel across the country. Among those soliciting the program is a staff member of an architectural school who wants to use automated graphics as a way of introducing students to computers. He feels that by having them produce graphical results rather than purely numerical results, the students will develop greater insight into computer use.

Several stereo-pair illustrations of crystal structures drawn with Johnson's program have been accepted for publication in scientific literature. It is hoped that the use of such computer-drawn stereoscopic illustrations will help alleviate the bottleneck in information transfer that results from the fact that the three-dimensional character of the crystallographic results must be communicated on the two-dimensional printed page. Stereoscopic viewers costing from 35 cents to $10 can be used to view the figures in stereo directly from a technical journal.

Another use of stereoscopic figures is in the presentation of crystal structure information to an audience. A special projector and polaroid viewing spectacles are needed for this visual aid technique. In order to explore the potential of this medium in an actual crystallographic conference, the ORNL Chemistry Division, in cooperation with the Mathematics Division, has offered to prepare a computer-drawn stereo slide for any crystal structure paper presented at the annual meeting of the American Crystallographic Association in Gatlinburg this June.