Introduction

There are numerous factors that effect the quality and reliability of parameters derived from a single crystal structural analysis. Some of the obvious factors are instrument alignment, calibration of the pulse counting system, stability of the X-ray source, and the temperature stability of the crystal during data collection.

Assuming the above are all within acceptable limits, the most important remaining variable is the crystal itself. Crystal quality is of course of paramount importance, and a crystal which is twinned, disordered, has lost solvent, etc. can only lead to less than desired results. Assuming that "well behaved" crystals are available with none of the above mentioned maladies, the remaining factors which must be taken into account are the effects of the crystal shape and quality on the measured intensity data. Extinction can be a significant problem, especially with nearly "perfect" crystals, but most modern least squares programs will provide a reasonable correction.

The size of the crystal is important in several respects. If the crystal is too large (even in one direction), there may be serious problems with the beam uniformity since it is assumed that the crystal is bathed in a uniform beam at all times. Indeed beam uniformity may account for most of the systematic errors in many studies. With "heavy atom" molecule problems (assuming beam uniformity can be ignored), the single most important remaining factor is absorption. There are many ways to correct for absorption, including empirical, semi-empirical, and analytic techniques. The validity of using various techniques has been the basis of a discussion in the sci.techniques.xtallography news group, and the results are available.

While most crystallographers agree that the analytical procedure (i.e. carefully indexed and measured crystal faces) should yield the best results, the small size of the crystals used (usually less than 0.5mm) make accurate measurements difficult.

The availability of an ophthalmic examination station and an inexpensive video camera system have allowed us to develop techniques to significantly increase the accuracy of the crystal measurements.

The Nikon FS-2 ophthalmic instrument, shown above, consists of a binocular microscope with an integral Nikon FE2 35mm camera back. The system is designed so that when the object is placed in the viewing field the base can be locked so that the microscope and camera can be rotated about a horizontal axis through the sample. The system has been modified to hold goniometer heads, sample vials, nmr tubes, etc. The advantage of the system is that stereo graphic images can be taken with ease. Examples are given for a well formed inorganic complex in an nmr tube, several minerals, and a fibrous material in a vial are shown.

While the above system provides excellent images for "large" samples (i.e. crystals ca. .5 mm or larger), typical samples which have heavy atoms are usually much smaller. In addition, the samples are often air sensitive and must be maintained in a nitrogen stream on the diffractometer. For these samples a CCD camera (Sony CCD-IRIS SSC-C350) with a microscope adapter (Edmond Scientific) is held in a fixture on the goniostat detector dovetail (see illustration).

The goniostat is adjusted so that the camera lens is perpendicular to the chi circle of the four-circle instrument as shown. The crystal is then oriented using the chi and phi axes so that the crystal faces can be easily identified and measured. A micrometer eyepiece on the goniostat microscope is used to determine the distances initially. Images are recorded using an inexpensive video capture board (Computer Eyes). Since the microscope is perpendicular to the phi axis, a rotation of 6 degrees in phi will allow a stereo graphic image to be recorded. By recording the image of a 1/32" diameter steel ball bearing, it is possible to use the images to obtain measurements on the crystal.

Once a preliminary assignment of the indices of the faces of the crystal is completed, the Cerius2 Morphology module is used to generate a computer representation of the crystal morphology. By placing the recorded stereo graphic image of the crystal above a stereo graphic image of the crystal as generated by Cerius, slight errors become immediately apparent. The distances of the faces can be adjusted until the two images are identical. The figures shown below illustrate this clearly.

Several other techniques have been used to help insure the proper measurements of the crystal. One of the easiest ways to have a "scale" present in the figures is to use the glass mounting fiber. The diameter can be readily determined to good precision using a standard micrometer (the diameter should be measured in several directions to insure it is uniform).

Examples

To test the described techniques, a sample of Wulfenite (lab sample, origin unknown) was examined. The mineral, PbMoO4, has a linear transmission coefficient of 507.3 cm-1. Three earlier studies have been reported, the latest being 1965. Well formed transparent crystals were present on the surface of a large sample as shown. The illustration was obtained using the ophthalmologist photo microscope system. A purposely large crystal was removed from the matrix and affixed to the end of a glass fiber using hot shellac. The space group and cell parameters were determined using reflections located via. a systematic search of a hemisphere of reciprocal space, and agreed with the earlier studies. Data were collected using a standard moving crystal-moving detector continuous scan. The crystal was then carefully characterized using the techniques described above. The final residual dropped from 0.11 to 0.03 when the data were corrected for absorption. The minimum and maximum transmission coefficients were 0.002 and 0.094. The residuals are remarkably low considering the magnitude of the absorption in the sample. Complete crystallographic data are available as MSC95488.

Conclusions

The ability to view a stereo graphic image of the crystal while trying to "construct" an image using imaging software has proven invaluable. Errors and missed faces are obvious. The method is also valuable when the crystal does not have well defined faces, such as a fragment cleaved from a larger sample. The modeling software can be used to try different indices for the faces as well as to adjust the distances of the faces. The ability to have multiple windows open on a workstation proved to be especially beneficial, since a stereo pair of the recorded image could be placed directly above the modeled image. Discrepancies were immediately obvious.

Future Work

The disadvantages of using the ophthalmic station are the relatively low magnification and the time delay required to obtain the photographic prints. The advantage is the high quality of the prints compared to the CCD images possible with the Sony system in use. The IUMSC has received an internal grant to purchase a high resolution CCD camera to replace the Sony CCD-IRIS on the goniostat. Modifications will be made to allow the new CCD camera to be used with the Nikon ophthalmic system.