Evaluating the Performance of a Commercial Silicon Drift Detector for X-ray Microanalysis
Edward A. Kenik Materials Science & Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831
kenikea@ornl.gov
Introduction Te development of large-area silicon driſt detectors
(SDDs) provides a significant improvement in X-ray micro- analysis, especially in scanning electron microscopes (SEMs) and electron microprobes, where high incident probe currents are possible. Te resultant improved detection limits and/ or speed of elemental mapping and analysis make the SDD the detector of choice for microanalysis [1]. With the larger geometric collection efficiency and faster response times of these detectors, the higher input count rates can place significant demands on the performance and speed of the signal processing electronics. Tere are numerous benefits of using SDD technology for
X-ray microanalysis when compared with conventional Si(Li) detectors, including: 1) larger geometric collection efficiency as a result of the much larger detector area that is possible without loss of energy resolution
2) higher count rate capability (300–800 kc/s) from the lower capacitance of the SDD versus conventional Si(Li) detectors
3) the possibility of improved energy resolution (some detectors offer ~125 eV)
4) no liquid nitrogen cooling required (thermoelectric cooler with passive or active secondary cooling) and rapid cool down to operating temperature (~30 minutes)
5) light weight (relates to column stability) and physical size (low acoustic interaction) of the SDD in part as a result of the elimination of the liquid nitrogen dewar One drawback of the SDD is the thinner active volume
(450 mm for the detector being discussed compared to the nominal 3-mm thickness of most Si(Li) crystals). Tis results in a decreased detection efficiency for X-rays with energies >10 keV because a greater fraction of such X-rays pass through the SDD. Tere could, however, be an advantage in using the thinner SDDs for microanalysis of radioactive materials that emit high-energy gamma rays as they decay. Tese gamma rays cause higher background counts , detector reset rates, and attendant higher dead time for conventional (that is, thicker) Si(Li) detectors. Gamma rays would deposit less energy into the thinner SDD detector, possibly with reduced deleterious effects on microanalysis. Te performance of the present SDD on a JEOL 6500F
SEM will be discussed here as a function of detector time constant (TC) and input count rate [2, 3]. Te stability of energy resolution and peak energy will be addressed along with pulse pile-up effects, including sum peaks and energy tailing.
Materials and Methods Te detector was mounted on one of the standard EDS ports of a JEOL 6500F SEM (Figure 1). Te JEOL 6500F SEM
40
Figure 1: Silicon drift detector mounted on a JEOL 6500F SEM. Note the plastic lines on the left back of the detector for non-aqueous cooling of the Peltier thermoelectric cooler stack.
doi:10.1017/S1551929511000241
www.microscopy-today.com • 2011 May
was designed with a high-brightness Schottky field emission gun (FEG) that is placed in the magnetic field of the first condenser lens. Tis design provides high beam currents (up to 200 nA for some conditions) at small probe diameters. Data collection and analyses were performed using a commercial X-ray microanalysis system with a digital pulse processor, which is designed for improved pulse processing and signal discrimination. Te SDD is an older style 40-mm2 Ketek chip with a 30-mm2 aperture to collimate the X-rays to the center of the chip. Tis chip has been replaced by newer chips in currently available SDDs. Te detector system manufacturer indicates that this design results in spectra with improved peak-to-background ratio (P/B). Te SDD is cooled to –60ºC by a double stack of Peltier thermoelectric coolers, which are subsequently cooled by an external, non-aqueous chiller. In general, spectra were acquired at 20 kV and 100-second clock time (unless otherwise specified). A pure manganese elemental standard was used for many of the performance tests. Calibration of the absorbed specimen current for this Mn standard versus the incident probe current was performed with a Faraday cup and a Keithley 6485 picoammeter. A brief definition of some of the terms used in this paper
is included to avoid confusion. Te input count rate is the number of electronic pulses per second that are sent to the detector electronics. Te input count rate quoted in this paper is that indicated by the soſtware package during spectrum acquisition. In general, each pulse should represent a single X-ray detected. Te throughput, or throughput count rate,
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76