Point defects in Ce-doped Y 3 Al 5 O 12 crystal scintillators C. L. Wang,* D. Solodovnikov, and K. G. Lynn Center for Materials Research, Washington State University, Pullman, Washington 99164-2711, USA Received 28 December 2005; revised manuscript received 16 May 2006; published 16 June 2006 Defect properties of Ce-doped and undoped Y 3 Al 5 O 12 YAGcrystals were studied by Doppler broadening of positron annihilation rays and thermoluminescence TLas a function of temperature 25–300 °C. The positron diffusion length L + was evaluated mainly from the S parameter versus positron energy. Compared with undoped YAG, Ce-doped YAG has a smaller positron diffusion length, due to its higher density of defects. L + in Ce-doped YAG decreases with increasing temperature up to 100 °C, and then increases with temperature. The TL intensity in Ce-doped YAG shows the opposite behavior to L + . The results indicate that point defects probed by positrons may be responsible for the energy transfer to Ce ions and TL intensity. Possible defects detected by positrons are negatively charged or neutral defects related to impurity antisites, cation vacancies, and vacancy complexes. DOI: 10.1103/PhysRevB.73.233204 PACS numbers: 78.70.Bj I. INTRODUCTION Recently there has been intense interest in searching for scintillation materials for radiation detection and medical imaging. 1–3 Optical materials, e.g., halides, oxides, and chal- cogenides, were traditionally used as -ray and x-ray scintillators. 1 Undoped crystals such as alkali-metal halides were often applied; while metal-ion-doped such as Ce 3+ , Eu 3+ crystals are more favored as fast scintillators due to their shorter luminescent decay times of 10– 100 ns. Yttrium aluminum garnet Y 3 Al 5 O 12 , YAGdoped with rare-earth or transition metals makes excellent laser and op- tical crystals. Cerium-doped YAG Ce:YAGhas been used as electron imaging sensors on scanning electron microscopes 4 and radiation scintillators. 5 It was speculated that the luminescence intensity at 550 nm in Ce-doped YAG increases with the concentration of defects such as oxygen vacancies or F centers. 6 These defects also lead to a change in the rise and decay transient profiles of luminescence. 6,7 Positron annihilation spectroscopy PASis a sensitive method for studying point defects in materials. In this tech- nique, a positron in a solid material is rapidly thermalized and diffuses until it annihilates with an electron, producing two quanta that are almost collinear. The energy spectrum of the rays is Doppler shifted from 511 keV the rest mass of electrons, due to momentum conservation in the annihi- lation processes. If positrons are efficiently trapped at defects such as vacancies and voids, they annihilate with electrons with lower momenta and produce a narrower energy spec- trum of rays compared with positron annihilation in the bulk. Therefore, the sharpness parameter S can be used to characterize the defects. 8,9 On the other hand, S versus the wing parameter Wcan provide independent information on the number of layers with different defect properties, 10,11 where W is due to positron annihilation with high- momentum electrons core electrons. Furthermore, the pos- itron energy dependence of S or W can give us the positron diffusion length and therefore information on defect proper- ties in materials. In this work, defects in cerium-doped YAG were studied by PAS and thermoluminescence TL. The correlation be- tween TL intensity and defect concentration is discussed. It was shown that controlling defect concentration is crucial for improving luminescence intensity and probably scintillation efficiency. II. EXPERIMENTS Three YAG samples grown by the Czochralski Cz method were obtained from VLOC a subsidiary of II-VI, Inc: sample R9 with 0.15 at. % Ce, sample R20 with 1.0 at. % Ce and 1.0 at. % Er, and an undoped YAG sample R0. Before positron and TL measurements, they were etched in phosphoric acid 85% concentrationat 200 °C to reduce surface damage, and were radiated by a Xe lamp so that the TL intensity was enhanced. 12 The TL measurement was performed on a heat stage with a heating rate of 12 °C/min after the sample was radiated with a Xe lamp for 15 min. A thermocouple monitored the temperature of heating stage. A photospectrometer Ocean Optics, Inc.was used to collect the thermal luminescence. The luminescence spectrum in the region of 500– 700 nm shows a peak at around 550 nm. The integrated intensity between 500 and 700 nm was calculated as a function of temperature. Doppler broadening spectra of positron annihilation quanta were measured with a variable-energy positron beam with a flux of about 5 10 5 / s cm 2 , a diameter of 5 mm, and an energy range of 0.1– 12 keV. Thermocouples monitored temperatures at the surface of the sample and the surface and bottom of the heater. Only the temperature on the surface of the samples is used in the positron results. The shape param- eters S and W are used in the analysis. 8 As an example, the measured SEand SWrelations 10,11 at room temperature for three samples are shown in Fig. 1. In Fig. 1b, the good linear relation between S and W at room temperature indicates that a single layer is adequate to describe the positron behavior, 10 as do the SWlinear rela- tions at higher temperatures. We used the VEPFIT program 13 to fit SEby a single-layer model with an epithermal posi- tron state in the surface. The diffusion lengths had large er- rors after fixing the bulk S or epithermal parameters at cer- PHYSICAL REVIEW B 73, 233204 2006 1098-0121/2006/7323/2332044©2006 The American Physical Society 233204-1