The Determination of the Efficiency of Energy DispersiveX-Ray Spectrometers by a New Reference MaterialMarco Alvisi,1Markus Blome,2Michael Griepentrog,3Vasile-Dan Hodoroaba,3Peter Karduck,2Marco Mostert,4Michele Nacucchi,1Mathias Procop,3,* Martin Rohde,5Frank Scholze,6Peter Statham,7Ralf Terborg,5and Jean-Francois Thiot41ENEA, Research Centre of Brindisi, S.S. 7 Appia km 713.7, I-72100 Brindisi, Italy2HDZ Herzogenrather Dienstleistungszentrum GbR, Kaiserstrasse 100, D-52134 Herzogenrath, Germany3Bundesanstalt fuer Materialfors chung und –pruefung (BAM), D-12200 Berlin, Germany4SAMx, 1554 route de la Roquette, F-06670 Levens, France5Roentec, Schwarzschildstrasse 12, D-12489 Berlin, Germany6Physikalisch-Technische Bundesanstalt, Abbestraße 2-12, 10587 Berlin, Germany7Oxford Instruments Analytical Limited, Halifax Road, High Wycombe, Bucks HP12 3SE, UKAbstract: A calibration procedure for the detection efficiency of energy dispersive X-ray spectrometers ~EDS!used in combination with scanning electron microscopy ~SEM! for standardless electron probe microanalysis~EPMA! is presented. The procedure is based on the comparison of X-ray spectra from a reference material~RM! measured with the EDS to be calibrated and a reference EDS. The RM is certified by the line intensities inthe X-ray spectrum recorded with a reference EDS and by its composition. The calibration of the reference EDSis performed using synchrotron radiation at the radiometry laboratory of the Physikalisch-Technische Bunde-sanstalt. Measurement of RM spectra and comparison of the specified line intensities enables a rapid efficiencycalibration on most SEMs. The article reports on studies to prepare such a RM and on EDS calibration andproposes a methodology that could be implemented in current spectrometer software to enable the calibrationwith a minimum of operator assistance.Key words: energy dispersive X-ray spectrometry, standardless analysis, X-ray detectors, detection efficiency,spectrometer calibrationINTRODUCTIONElectron probe X-ray microanalysis ~EPMA! is applied todetermine the concentration of chemical elements in amicroscopic volume of a specimen. In many cases analysis isperformed by means of an energy dispersive X-ray spectrom-eter ~EDS! attached to a scanning electron microsc ope ~SEM!.Physical equations relate the element concentrations to thenumber of emitted photons. A long-term goal of X-raymicroanalysis is to calculate the concentrations without theadditional need to measure spectra from standard speci-mens. The accuracy of so-called standardless analysis hasbeen repeatedly discussed at EMAS workshops ~Pouchou,1994; Wernisch & Röhrbacher, 1998; Joy, 2002!.However,little attention has been paid to the spectrometer efficiency,which is the critical factor relating the counts measured bythe EDS to the number of photons emitted toward thedetector entrance window. This relation is defined as spec-trometer efficiency and must be known for standardlessanalysis.The spectrometer efficiency depends strongly on pho-ton energy. It is usually calculated from X-ray absorption ofthe individual detector components ~window, front contact,crystal material!. Some layer thicknesses are not known withsufficient accuracy to calculate exactly the efficiency for anindividual spectrometer. In addition, the efficiency maychange during spectrometer operation due to buildup ofcontamination layers on the detector or by degr adation.Hence, when spectrometer efficiency is calculated from as-sumed data for the detector design, considerable errors canbe involved. This study explores the feasibility of measuringspectrometer efficiency by means of an easy methodology.The basic idea is to find the unknown spectrometer effi-ciency by comparison of two spectra from a reference mate-rial ~RM!. One spectrum is measured with the spectrometerof unknown efficiency, the other one with a spectrometer ofknown efficiency, that is, with a calibrated reference spec-trometer. In the following, the notation “unknown spectrom-eter” stands for the spectrometer with unknown efficiency,and “efficiency transfer” means the procedure to get the un-known efficiency by comparison of the RM X-ray spectra.Received December 8, 2005; accepted June 2, 2006.*Corresponding author. E-mail: [email protected]. Microanal. 12, 406–415, 2006DOI: 10.1017/S1431927606060557MicroscopyANDMicroanalysis© MICROSCOPY SOCIETY OF AMERICA 2006The present work describes the selection of the RM,which is specified mainly by its X-ray spectrum. The prepa-ration as a thick film and its characterization concerningchemical composition, structure, and stability are given. Ashort outline of the determination of the efficiency of thereference spectrometer using synchrotron radiation ~SR! isfollowed by the explanation of the efficiency transfer proce-dure to get the efficiency for an unknown spectrometer. Amethodology that can be applied to most SEM/EDS sys-tems is proposed. Finally, examples of the efficiency transferare presented, and its precision is estimated.MATERIALS AND METHODSRequirements for the Reference MaterialThe reference material should meet the following require-ments w ith respect to X-ray spectroscopy:1. It should contain different chemical elements emittingcharacteristic X-ray lines in a wide energy range fromabout 0.2 to 20 keV. X-ray lines should be preferablypositioned at energies close to ~below or above! absorp-tion edges in the efficiency curve of the EDS detector todistinguish the effects of different absorbing materials onefficiency. The X-ray lines should be separated at least bytwo times the full width at half maximum ~FWHM! toenable a robust background subtraction method and toavoid an uncertainty in peak area determination causedby deconvolution.2. The peak-to-background ratio for characteristic peaksshould be high to maximize the precision that can beobtained in an acceptable measurement time.3. The intensities of the strongest X-ray lines should besimilar to balance the precision of measurement through-out the energy range.In addition, the requirements related to materials prop-erties and stability are:1. The surface should be smooth and flat to have a well-defined take-off angle ~TOA!. The RM should be large indiameter ~at least 20 mm! to enable accurate adjustmentof surface tilt in the SEM.2. The RM must be laterally homogeneous in order tomake the result of the efficiency calibration independentof the position on the specimen surface.