A bibliometric study in crystallography

Abstract

This is an application of the mathematical and statistical techniques of bibliometrics to the field of crystallography. This study is, however, restricted to inorganic compounds. The data were taken from the Inorganic Crystal Structure Database, which is a well defined and evaluated body of literature and data published from 1913 to date. The data were loaded in a relational database system, which allows a widespread analysis. The following results were obtained: The cumulative growth rate of the number of experimentally determined crystal structures is best described by a third-degree polynomial function. Except for the upper end of the curve, Bradford's plot can be described well by the analytical Leimkuhler function. The publication process is dominated by a small number of periodicals. The probability of the author productivity in terms of publications follows an inverse power law of the Lotka form and in terms of database entries an inverse power law in the Mandelbrot form. In both cases the exponent is about 1.7. For the lower tail of the data an exponential correction factor has to be applied. Multiple authorship has increased from 1.4 authors per publication to about four within the past eight decades. The author distribution itself is represented by a lognormal distribution.