The rapid advance and falling costs of computing power and data storage during the last 40 years have greatly enhanced the data reduction speed and analytical capacity of energy dispersive spectroscopy (EDS) systems. At the same time, the solid state X-ray detector [until recently, the Si(Li) diode] has seen performance increases due to the use of advanced materials and processing techniques. In addition, the performance of the electronic components (field effect transistor, preamp, and pulse processor) of an EDS system has been constantly improved. These technology advances have resulted in improved spectral quality, excellent light element detection, and increased count rate performance. The results have been truly remarkable and driven by the needs of the analyst. This article will summarize the progress made in these areas in the last 40 years and make a brief reference to prospects for future development in EDS system performance.

On February 2, 1968, R. Fitzgerald, K. Keil, and K.F.J. Heinrich published a seminal paper in Science (159, 528–530) in which they described a solid-state Si(Li) energy dispersive spectrometer (EDS) for electron probe microanalysis (EPMA) with, initially, a resolution of 600 eV. This resolution was much improved over previous attempts to use either gas-filled proportional counters or solid-state devices for EDS to detect X-rays and was sufficient, for the first time, to make EDS a practically useful technique. It ushered in a new era not only in EPMA, but also in scanning electron microscopy, analytical transmission electron microscopy, X-ray fluorescence analysis, and X-ray diffraction. EDS offers many advantages over wavelength-dispersive crystal spectrometers, e.g., it has no moving parts, covers the entire X-ray energy range of interest to EPMA, there is no defocusing over relatively large distances across the sample, and, of particular interest to those who analyze complex minerals consisting of many elements, all X-ray lines are detected quickly and simultaneously.

The energy-dispersive Si(Li) X-ray spectrometer, introduced 30 years ago into electron probe mi-croanalysis (EPMA) by R. Fitzgerald et al., has profoundly affected the development of microanalysis. It offers many advantages over the wavelength-dispersive crystal spectrometer. It has no moving parts and covers the full energy range of interest in EPMA. There is no defocusing over large distances on the specimen, the efficiency of the device is high, varies slowly and continuously with atomic number, and can be predicted fairly accurately, and, most importantly, all emission lines are detected and can be observed simultaneously. The one remaining disadvantage of the Si(Li) spectrometer is its poorer energy resolution. Solid-state detection devices now under development promise to achieve resolution comparable to that of the crystal spectrometer.