Báo cáo toàn văn Kỷ yếu hội nghị khoa học lần IX Trường Đại học Khoa học Tự nhiên, ĐHQG-HCM II-P-1.10 EFFICIENCY CALCULATION OF AN N-TYPE HPGE DETECTOR USING MCNP5 CODE Trần Nguyễn Thùy Ngân1, Trương Thị Hồng Loan1,2, Trương Hữu Ngân Thy2, Huỳnh Thị Yến Hồng2, Vũ Ngọc Ba2 Nuclear Physics Department, Faculty of Physics and Engineering Physics, University of Science, VNU-HCMC Nuclear Technique Laboratory, VNU-HCMC Email: tntngan@hcmus.edu.vn ABSTRACT The peak efficiencies of multi-gamma lines from eight point sources (Co-60, Co-57, Cs-137, Cd109, Ba-133, Mn-54, Na-22, Zn-65) placed at various source-to-detector distances are determined via a low background gamma spectroscopy using an n-type HPGe detector The authors propose a geometrical modelling of the n-type detector based on the manufacturer parameters by using MCNP5 code Effectiveness of the geometrical modelling is examined by comparing theoretical efficiency curve generated by Monte Carlo technique and experimental efficiency curve Since there is considerable disagreement between experimental and calculated results, a detail study of the geometrical configuration to find optimum configuration is carried out The optimum geometrical model of the detector can be employed in conducting experiments to investigate radioactivity of environmental samples Keywords: efficiency, HPGe detector, MCNP5 simulation INTRODUCTION A brand new low-background gamma spectroscopy using an n-type High Purity Germanium (HPGe) detector model GMX35P4-70 has been established since July 2013 at Nuclear Technique Laboratory, University of Science VNU-HCMC There is a requirement to examine working characteristics of detector before applying this spectroscopy into practical analysis of environmental samples M Schlager [9] developes geometry configuration of two HPGe detectors (a reverse electrode REGe and an extended range XtRa) by conducting experiments and MCNP4C2 simulation The discrepancy between experimental efficiencies and simulating efficiencies decreases from 10 – 20 % to 3% by scanning detector with collimated point source F.P Cabal [5] developes a procedure to optimize detector configuration by adjusting physical parameters of an n-type HPGe detector through MCNPX 2.6 and GEANT4 9.2 Therefore, the discrepancy between experimental efficiencies and simulating efficiencies decreases from 18% to 4% In addition, Ngo Quang Huy [8] studies the effect of detector dead layer by MCNP5 and results indicate that the dead layer thickness increased from 0.35mm to 1.46mm after 13 years of operating A Elanique [6] studies effect of HPGe detector dead layer on efficiency at low energy region by experiment and simulation MCNPX and then the dead layer thickness is adjusted from 0.4m to 7.5m In this study, the geometry configuration of an n-type HPGe detector is modelled by MCNP5 simulation and then detector efficiency is calculated Since there is considerable disagreement between experimental and calculated results, a detail study of the geometrical configuration to find optimum configuration is carried out The optimum geometrical modelling of the detector can be employed in conducting experiments to investigate radioactivity of environmental samples MCNP5 SIMULATION AND RESULT Detector modelling An initial modelling of the gamma spectroscopy including GMX35P4-70 detector, shielding and a point source for MCNP5 code is established based on the physical dimensions which is provided by Ortec manufacturer In order to check the validation of simulated modelling, some main properties of detector such as energy resolution (FWHM), peak to Compton ratio (P/C) and relative efficiency at 1332.5keV of Co-60 placed at 25cm from detector window are calculated from simulation results and then compared with values from manufacturer The spectroscopy geometry consisting of detector, shielding and source is described with 24 cells and 56 surfaces MCNP5 is used in mode P, pulse height spectra simulations are obtained by using tally F8 Illustration including physical parameters of GMX35P4-70 detector and schematic drawing of detector, point source and shielding by MCNP5 are presented via figure and figure ISBN: 978-604-82-1375-6 162 Báo cáo toàn văn Kỷ yếu hội nghị khoa học lần IX Trường Đại học Khoa học Tự nhiên, ĐHQG-HCM Table The comparison of some detector’s main properties given by manufacturer (warranted and experimental) and MCNP5 simulation for initial configuration At E = 1332.5keV of Co-60 Warranted Experimental Simulated FWHM 1.95keV 1.80keV 1.85keV P/C 55:1 61:1 74:1 Relative efficiency 35% 38% 42% In table 1, the stimulated FWHM (1.85keV) is adequate However, the stimulated P/C and relative efficiency are significantly different from values of manufacturer Particularly, discrepancy between stimulated relative efficiency (42%) and experimental relative efficiency (38%) is 10.52%; discrepancy between stimulated P/C (74:1) and experimental P/C (61:1) is 21.31% A: 55.8 mm – germanium crystal diameter B: 78.1 mm – germanium crystal length C: 8.6 mm – hole diameter D: 69.6 mm – hole depth E: mm – hole round radius F: 94 mm – aluminum cup length G: mm - space H: 0.03 mm aluminum + 0.03 mm mylar I: 0.05 mm- beryllium window J: mm – crystal rounding edge radius K: 0.8 mm aluminum L: mm aluminum M: 0.3 m boron N: 0.7 mm lithium O: mm aluminum Figure Illustration and physical parameters of GMX35P4-70 detector Figure Schematic drawing of GMX35P4-70 detector, point source and shielding by MCNP5 In addition, full-energy peak efficiencies at gamma rays (81.0keV, 88.0keV, 122.1keV, 136.5keV, 276.4keV, 302.9keV, 356.0keV, 661.7keV, 834.8keV, 1115.5keV, 1173.2keV, 1274.5keV and 1332.5keV) of standard point sources Co-57, Co-60, Na-22, Cd-109, Ba-133, Zn-65, Mn-54 and Cs-137 placed at 25.55cm form detector window are investigated by experiment and simulation and are shown in table The discrepancy between experimental and simulated efficiency is up to 17.30% ISBN: 978-604-82-1375-6 163 Báo cáo toàn văn Kỷ yếu hội nghị khoa học lần IX Trường Đại học Khoa học Tự nhiên, ĐHQG-HCM Table Comparison between experimental and simulated full-energy peak efficiencies of standard point sources E (keV) experiment Error (%) simulation Error (%) Discrepancy (%) 81.0 0.0022069 3.14 0.0023036 0.12 4.38 88.0 0.0021419 3.09 0.0022551 0.23 5.29 122.1 0.0020868 3.00 0.0022068 0.07 5.75 136.5 0.0019236 3.32 0.0021215 0.20 10.29 276.4 0.0013096 3.09 0.0014617 0.33 11.61 302.9 0.0011963 3.09 0.0013298 0.22 11.16 356.0 0.0010838 3.02 0.0012020 0.12 10.91 661.7 0.0006733 3.00 0.0007621 0.09 13.18 834.8 0.0005726 3.01 0.0006505 0.22 13.59 1115.5 0.0004585 3.02 0.0005378 0.22 17.30 1173.2 0.0004467 3.01 0.0005134 0.31 14.94 1274.5 0.0004222 3.01 0.0004881 0.36 15.60 1332.5 0.0004060 3.01 0.0004754 0.32 17.11 Modifying configuration Some physical parameters of detector are adjusted as follows: diameter of germanium crystal decreases 0.84mm, length of germanium crystal decreases 0.8mm, rounding edge radius decreases 1.0mm and the distance between aluminum cup front end and front surface of germanium crystal increases 0.5mm [3], [4], [9] Table The comparison of some detector’s main properties given by manufacturer (warranted and experimental) and MCNP5 simulation for modifying configuration At E = 1332.5 keV of Co-60 Warranted Experimental Simulated FWHM 1.95keV 1.80keV 1.85keV P/C 55:1 61:1 72:1 Relative efficiency 35% 38% 40% According to table 3, the stimulated FWHM is 1.85keV; stimulated relative efficiency is 40% and discrepancy between stimulated relative efficiency and experimental relative efficiency is 5.26%; stimulated P/C is 72:1 and discrepancy between stimulated P/C and experimental P/C is 18.03% The discrepancy between experimental and stimulated full-energy peak efficiency of standards point sources placed at 25.55cm from detector window is up to 11.81% These results indicate that revised modelling is not suited to the real spectroscopy Modifying configuration The dead layer thickness (lithium layer covering the crystal hole) is changed from 0.7mm to 2.0mm, the rest of physical parameters are unchanged [1], [6], [8] According to table 4, the stimulated FWHM is 1.84keV; stimulated relative efficiency is 39% and discrepancy between stimulated relative efficiency and experimental relative efficiency is 2.63%; stimulated P/C is 68:1 and discrepancy between stimulated P/C and experimental P/C is 11.48% The discrepancies between experimental and stimulated full-energy peak efficiency of standards point sources placed at 25.55cm from detector window are under 10% at all observed gamma rays According to this results, revised modelling is more suited to the real spectroscopy and is used in the next study ISBN: 978-604-82-1375-6 164 Báo cáo toàn văn Kỷ yếu hội nghị khoa học lần IX Trường Đại học Khoa học Tự nhiên, ĐHQG-HCM Table The comparison of some detector’s main properties given by manufacturer (warranted and experimental) and MCNP5 simulation for modifying configuration At E = 1332.5 keV of Co-60 Warranted Experimental Simulated FWHM 1.95keV 1.80keV 1.84keV P/C 55:1 61:1 68:1 Relative efficiency 35% 38% 39% Studying detector efficiency of environmental cylindrical samples Full-energy peak efficiencies at gamma rays (59.5keV, 63.3keV, 88.0keV, 122.1keV, 295.2keV, 351.9keV, 661.7keV, 834.8keV, 1173.2keV and 1332.5keV) of cylindrical samples which have different matrices (dirt, dry dirt, grass powder and milk powder) with observed density (0.5, 0.8, 1.0, 1.2 and 2.0g/cm3) are investigated by using MCNP5 [2], [7] Simulated full-energy peak efficiencies of different-material samples at different densities are shown in table 5, table 6, table 7, table and table In the case of samples having the same matrix, full-energy peak efficiency decreases significantly when the sample’s density increases gradually from 0.5g/cm3 to 2.0g/cm3 In the case of samples having the same density, effect of matrix to full-energy peak efficiency is clear at low energy region (less than 100keV) and can be negligible at higher energy region Table Simulated full-energy peak efficiencies of different-material samples at density = 0.5g/cm3 E (keV) 59.5 63.3 88.0 122.1 295.2 351.9 661.7 834.8 1173.2 1332.5 Dry dirt 0.0823209 0.0839190 0.0851022 0.0816147 0.0447375 0.0382427 0.0223618 0.0185172 0.0144719 0.0131685 = 0.5 g/cm3 Dirt 0.0832072 0.0841250 0.0863753 0.0807908 0.0442393 0.0381693 0.0220967 0.0183789 0.0142830 0.0130176 Milk powder 0.0857681 0.0869509 0.0881656 0.0823650 0.0447831 0.0381367 0.0223331 0.0186093 0.0142840 0.0131716 Grass powder 0.0880841 0.0881740 0.0882529 0.0812662 0.0447711 0.0381802 0.0220993 0.0186527 0.0144643 0.0130184 Table Simulated full-energy peak efficiencies of different-material samples at density = 0.8g/cm3 E (keV) 59.5 63.3 88.0 122.1 295.2 351.9 661.7 834.8 1173.2 1332.5 Dry dirt 0.0741017 0.0756762 0.0790402 0.0769136 0.0430841 0.0368766 0.0217369 0.0182117 0.0141911 0.0129149 = 0.8 g/cm3 Dirt 0.0750591 0.0770066 0.0808191 0.0770080 0.0428221 0.0366922 0.0213579 0.0179128 0.0139795 0.0127471 Milk powder 0.0823035 0.0829883 0.0834473 0.0782801 0.0447904 0.0368314 0.0217156 0.0182193 0.0141764 0.0129038 Grass powder 0.0832264 0.0835378 0.0830197 0.0780032 0.0429042 0.0367077 0.0216296 0.0181892 0.0141349 0.0128037 As sample’s density increasing, the full-energy peak efficiency discrepancy at low energy region between samples having different matrix is more and more important Particularly, at sample’s density 2.0g/cm3, the full-energy peak efficiency discrepancy between dry dirt sample and grass powder sample is up to 31.4% at 59.5keV and the full-energy peak efficiency discrepancy between dry dirt sample and milk powder sample is up to 31.5% at 59.5keV ISBN: 978-604-82-1375-6 165 Báo cáo toàn văn Kỷ yếu hội nghị khoa học lần IX Trường Đại học Khoa học Tự nhiên, ĐHQG-HCM Table Simulated full-energy peak efficiencies of different-material samples at density = 1.0g/cm3 E (keV) 59.5 63.3 88.0 122.1 295.2 351.9 661.7 834.8 1173.2 1332.5 Dry dirt 0.0696201 0.0715814 0.0745304 0.0740366 0.0419315 0.0355378 0.0211170 0.0178659 0.0139739 0.0127002 = 1.0 g/cm3 Dirt 0.0706251 0.0727437 0.0774679 0.0742483 0.0417267 0.0358571 0.0210853 0.0176666 0.0137814 0.0125343 Milk powder 0.0800232 0.0803788 0.0804686 0.0756726 0.0419792 0.0359763 0.0213266 0.0179186 0.0139773 0.0127388 Grass powder 0.0772379 0.0796933 0.0802789 0.0755133 0.0419185 0.0358926 0.0212901 0.0178083 0.0139609 0.0125904 Table Simulated full-energy peak efficiencies of different-material samples at density = 1.2g/cm3 E (keV) 59.5 63.3 88.0 122.1 295.2 351.9 661.7 834.8 1173.2 1332.5 Dry dirt 0.0656305 0.0671462 0.0714767 0.0702202 0.0406100 0.0351207 0.0209714 0.0176175 0.0137802 0.0124398 =1.2 g/cm3 Dirt 0.0671488 0.0689439 0.0741751 0.0710747 0.0407234 0.0349680 0.0208976 0.0175533 0.0135060 0.0124519 Milk powder 0.0768408 0.0772352 0.0775650 0.0731645 0.0409248 0.0351268 0.0209404 0.0176251 0.0137829 0.0125716 Grass powder 0.0764337 0.0765606 0.0767284 0.0720314 0.0405958 0.0350849 0.0208678 0.0175162 0.0136727 0.0124925 Table Simulated full-energy peak efficiencies of different-material samples at density = 2.0g/cm3 E (keV) 59.5 63.3 88.0 122.1 295.2 351.9 661.7 834.8 1173.2 1332.5 Dry dirt 0.0500773 0.0540718 0.0597848 0.0617058 0.0370767 0.0320664 0.0195292 0.0165505 0.0130390 0.0119001 = 2.0 g/cm3 Dirt 0.0536537 0.0559865 0.0634380 0.0624329 0.0369182 0.0319067 0.0192566 0.0162648 0.0128545 0.0118313 Milk powder 0.0658518 0.0663517 0.0673939 0.0642489 0.0370733 0.0320137 0.0194811 0.0165126 0.0130384 0.0119322 Grass powder 0.0657949 0.0660624 0.0669200 0.0637982 0.0369268 0.0319763 0.0194141 0.0162648 0.0129997 0.0118346 CONCLUSION In this thesis, initial modelling and two revised modellings of gamma spectroscopy are established by MCNP5 Full-energy peak efficiencies at observed gamma rays of point sources are investigated and full-energy peak efficiencies at observed gamma rays of cylindrical samples with different matrices at different densities (0.5 – 2.0g/cm3) are evaluated as well Acknowledgement: The authors appreciate Nuclear Technique Laboratory – University of Science VNUHCMC for technique assistance and good cooperation ISBN: 978-604-82-1375-6 166 Báo cáo toàn văn Kỷ yếu hội nghị khoa học lần IX Trường Đại học Khoa học Tự nhiên, ĐHQG-HCM TÍNH TOÁN HIỆU SUẤT CỦA ĐẦU DÒ BÁN DẪN HPGE LOẠI N BẰNG CHƯƠNG TRÌNH MCNP5 Trần Nguyễn Thuỳ Ngân1, Trương Thị Hồng Loan1,2, Trương Hữu Ngân Thy2, Huỳnh Thị Yến Hồng2, Vũ Ngọc Ba2 Bộ môn Vật lý Hạt nhân, Khoa Vật lý – Vật lý Kỹ thuật, Trường Đại học Khoa học Tự nhiên, VNU-HCMC Phòng thí nghiệm Kỹ thuật hạt nhân, Trường Đại học Khoa học Tự nhiên, VNU-HCMC Email: tntngan@hcmus.edu.vn TÓM TẮT Trong nghiên cứu này, hiệu suất đỉnh lượng từ tám nguồn chuẩn (Co-60, Co57, Cs-137, Cd-109, Ba-133, Mn-54, Na-22, Zn-65) khoảng cách khác nguồn đầu dò xác định hệ phổ kế gamma phông thấp sử dụng đầu dò bán dẫn germanium siêu tinh khiết loại n Đồng thời, cấu hình hệ đo xây dựng chương trình mô MCNP5 dựa thông số cung cấp nhà sản xuất hiệu suất đỉnh lượng từ nguồn chuẩn tính toán thông qua phương pháp Monte Carlo Do có sai khác hiệu suất thực nghiệm hiệu suất mô phỏng, thông số đầu đò từ nhà sản xuất khảo sát để xây dựng cấu hình phù hợp cho hệ đo Cấu hình hệ đo sau hiệu chỉnh dùng đề tính toán mô thí nghiệm xác định hoạt độ mẫu phóng xạ môi trường Từ khóa: đầu dò HPGe, hiệu suất, mô MCNP5 REFERENCE [1] Võ Xuân Ân, Nghiên cứu hiệu suất ghi nhận detector bán dẫn siêu tinh khiết (HPGe) phổ kế gamma phương pháp Monte Carlo thuật toán di truyền, Doctor Dissertation, University of Science, VNU-HCMC, 2008 [2] Trương Thị Hồng Loan, Áp dụng phương pháp mô Monte Carlo để nâng cao chất lượng hệ phổ kế gamma sử dụng đầu dò bán dẫn HPGe, Doctor Dissertation, University of Science, VNU-HCMC, 2009 [3] Berndt R., Mortreau P., Monte Carlo modelling of a N-type coaxial high pure germanium detector, Nuclear Instruments and Methods in Physics Research A 694 (2012) 341 – 347 [4] Boson J., Agren G., Johansson L., A detailed investigation of HPGe detector response for improved Monte Carlo efficiency calculations, Nuclear Instruments and Methods in Physics Research A 587 (2008) 304 – 314 [5] Cabal F.P., Lopez-Pino N., Bernal-Castillo J.L., Martinez-Palenzuela Y., Aguilar-Mena J., D’Alessandro K., Arbelo Y., Corrales Y., Diaz O., Monte Carlo based geometrical model for efficiency calculation of an n-type HPGe detector, Applied Radiation and Isotopes 68 (2010) 2403 – 2408 [6] Elanique A., Marzocchi O., Leone D., Hegenbart L., Breustedt B., Oufni L., Dead layer thickness characterization of an HPGe detector by measurements and Monte Carlo simulations, Applied Radiation and Isotopes 70 (2012) 538 – 542 [7] Mostajaboddavati M., Hassanzadeh S., Faghihian H., Abdi M.R., Kamali M., Efficiency calibration and measurement of self-absorption correction for environmental gamma-spectroscopy of soil samples using Marinelli beaker, Journal of Radioanalytical and Nuclear Chemistry, 268(3) (2006) 539 – 544 [8] Ngo Quang Huy, The influence of dead layer thickness increase on efficiency decrease for a coaxial HPGe p-type detector, Nuclear Instruments and Methods in Physics Research A 621 (2010) 390 – 394 [9] Schlager M., Precise modelling of coaxial germanium detectors in preparation for a mathematical calibration, Nuclear Instruments and Methods in Physics Research A 580 (2007) 137 – 140 ISBN: 978-604-82-1375-6 167