INTRODUCTION

The Powder Diffraction File (PDF) is a continuing compilation of diffraction patterns gathered from many sources. Produced and published by the JCPDS--International Centre for Diffraction Data, 1 the PDF is used for identification of crystalline materials by matching d-spacings and diffraction intensity measurements. Under the partial sponsorship of the JCPDS, the program at the National Bureau of Standards contributes new or improved data to the PDF. Our work also aids in the development of diffraction techniques. This report presents information for 58 experimental patterns, and is the twenty-eighth of the series of Standard X-ray Diffraction Powder Patterns.2

EXPERIMENTAL POWDER PATTERNS

Names. The nonmenclature follows the current practice of the PDF.

CAS registry number. The Chemical Abstracts Service Registry Number is included, when available, to help identify the sample. This number forms the basis for computer-aided searching of Chemical Abstracts.

Sample. The samples used to make NBS patterns were obtained from a variety of sources or were prepared in small quantities in our laboratory.

1JCPDS--International Centre for Diffraction Data, 1601 Park Lane, Swarthmore, PA 19081. This Pennsylvania non-profit corporation functions in cooperation with the American Ceramic Society, the American Crystallographic Association, the American Society for Testing and Materials, The Clay Minerals Society, The Institute of Physics, the Mineralogical Association of Canada, the Mineralogical Society of America, The Mineralogical Society of Great Britain and Ireland, the National Association of Corrosion Engineers, and the Société Francaise de Minéralogie et de Cristallographie.

2See previous page for other published volumes.

Appropriate annealing or recrystallization of the samples improved the quality of many of the patterns. A check of phase purity was provided by indexing the x-ray pattern and by optical examination.

Optical data. When reported, optical measurements were made by grain immersion methods, in white light, using oils standardized in sodium light, in the refractive index range 1.49 to 2.1 [Hartshorne and Stuart, 1970].

The names of the sample colors are selected from the ISCC-NBS Centroid Color Charts [1965].

Interplanar spacings. All spacing determinations were made using an internal standard mixed with the sample, packed in a shallow holder. Choice of the standard was determined by the need for low angle and unobstructed reflections. The amount of standard was estimated so that the intensity of its strongest peak would be about equal to the intensity of the strongest peak of the sample. The internal standards used were of high purity (99.99%). The lattice constants used for them at 25 °C are given in Table 1; the 20 angles were computed using cell dimensions uncorrected for index of refraction.

The internal standard Si powder is available as Standard Reference Material 640 [1974]. The lattice constant for the Si was refined from multiple powder data measurements made with tungsten as an internal standard [Swanson et al., 1966]. Single crystal cell parameter data were also collected. The lattice parameters from the two methods agreed within three parts in 105 [Hubbard et al., 1975]. D-spacing results using SRM 640 will be in agreement with patterns recorded in this series of Monographs since 1966.

All patterns were recorded at 25 ± 1 °C on a diffractometer equipped with a focusing graphite or

Structure, lattice constants. The space group symbols are given the short Hermann-Mauguin notation. Also given are the space group numbers listed in the International Tables for X-ray Crystallography, Vol. I [1965].

Orthorhombic cell dimensions are arranged according to the Dana convention b>a>c [Palache et al., 1944]. Monoclinic and triclinic lattice constants are transformed if necessary in order to follow the convention of Crystal Data [1973]. The lattice constant ratios, a/b, c/b, and c/a, also follow the conventions used for the determinative ratios in Crystal Data [1973].

In most cases, preliminary lattice constants were available in the literature, and were used for the initial indexing and refinement. In cases where such data were not available, other methods were tried. If suitable single crystals were available, the lattice constants were obtained by use of a four-circle diffractometer. Axial ratios and densities from Groth [1908] were sometimes useful. Cell constants were also found in some instances by use of the Visser computer program [1969].

A least squares program [Evans et al., 1963] assigned hkl's and refined the lattice constants. In indexing cubic patterns, for a given reflection multiple hkl's were not utilized in the refinement or reported. Instead, the single appropriate index having the largest h was used. Cell refinement was based only upon 20 values which could obs be indexed without ambiguity. The program minimized the value (obs-0, 2. The estimated standard deviations (e.s..s) of the reciprocal cell parameters were determined from the inverse matrix of the normal equations. The program calculated the e.s.d.'s of the direct cell constants by the method of propagation of errors. Since 1973, the e.s.d.'s derived by the computer program have been increased by 50% in order to reflect more truly the uncertainty in the lattice constants. A similar increase should also be applied to all lattice constants published in this series prior to 1973. The e.s.d.'s in the least significant figures are given in parentheses following the lattice constants.

For each d-value, the number of significant figures was derived from the average error in 120 - 20 calc and the equation Ad/d = -(cote)40. With these conditions, the rounded value of d agrees with its appropriate 20 within the average error in 20. The value of 40 varies with the summetry and crystallinity of each sample.

Densities. These were calculated from the specified lattice constants, the Avogadro number 6.0220943 x 1023 [Deslattes et al., 1974] and 1977 atomic weights published by the International Union of Pure and Applied Chemistry [1979].

Figure of merit. Several figures of merit ratings are available for assessing indexed powder data. M20 [de Wolff, 1968] is a criterion for the reliability of the unit cell and indexing. A value of M20 > 10 will guarantee the essential correctness of the indexing provided there are not more than 2 spurious lines (X202) [de Wolff, 1968]. In general, patterns reported in this publication had M20 > 20 and X = 0. M20 is reported if a cell was derived only through computer indexing from powder data, without further

Intensity measurements. The intensities of the diffraction lines were measured as peak heights above background and were expressed in percentage of the strongest line. It has been found that samples which give satisfactory intensity patterns usually have an average particle size smaller than 10 μm, as recommended by Alexander et al. [1948]. In order to avoid the orientation effects which occur when powdered samples are packed or pressed, a sample holder was made that had in its top face a rectangular cavity which extended to one end of the holder. To prepare the sample, a glass slide was clamped over the top face to form a temporary cavity wall (see Figure 1), and the powdered sample was allowed to drift into the end opening while the holder was held in a vertical position. With the sample holder returned to a horizontal position, the glass slide was carefully removed so that the sample could be exposed to the x-ray beam (see Figure 2). As a general practice, approximately 50 volume percent of finely ground silica-gel was added as a diluent. Occasionally,

a rotating sample holder was used.

Where conversion of intensities for effects of theta-compensating slits was required, each o was multiplied by the conversion factor