Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Abe, Shinichiro; Hashimoto, Masanori*; Liao, W.*; Kato, Takashi*; Asai, Hiroaki*; Shimbo, Kenichi*; Matsuyama, Hideya*; Sato, Tatsuhiko; Kobayashi, Kazutoshi*; Watanabe, Yukinobu*
IEEE Transactions on Nuclear Science, 70(8, Part 1), p.1652 - 1657, 2023/08
Times Cited Count：0 Percentile：0.01(Engineering, Electrical & Electronic)Single event upsets (SEUs) caused by neutrons is a reliability problem for microelectronic devices in the terrestrial environment. Acceleration tests using white neutron beam provide realistic soft error rates (SERs), but only a few facilities can provide white neutron beam in the world. If single-source irradiation applicable to diverse neutron source can be utilized for the evaluation of the SER in the terrestrial environment, it contributes to solve the shortage of beam time. In this study, we investigated the feasibility of the SER estimation in the terrestrial environment by any one of these measured data with the SEU cross sections obtained by PHITS simulation. It was found that the SERs estimated by our proposed method are within a factor of 2.7 of that estimated by the Weibull function. We also investigated the effect of simplification which reduce the computational cost in simulation to the SER estimation.
Abe, Shinichiro
Nihon Genshiryoku Gakkai-Shi ATOMO
, 65(5), p.326 - 330, 2023/05
Non-destructive faults (the so-called soft errors) in microelectronics caused by environmental radiation such as neutrons and muons have been recognized as a serious reliability problem. The number of microelectronics requiring high reliability increases with the growth of the information society. Therefore, it is not realistic to evaluate the soft error rate (SER) of all microelectronics by measurement. Moreover, the evaluation of SER in the pre-manufacturing stage is sometimes required. As a result, the evaluation of SER by simulation become more important. We have developed the soft error simulation method with PHITS code. We have also simulated the neutron- and muon-induced soft errors. These results will be reported in the journal of the Atomic Energy Society of Japan (AESJ) as the explanatory article.
Liao, W.*; Ito, Kojiro*; Abe, Shinichiro; Mitsuyama, Yukio*; Hashimoto, Masanori*
IEEE Transactions on Nuclear Science, 68(6), p.1228 - 1234, 2021/06
Times Cited Count：2 Percentile：31.78(Engineering, Electrical & Electronic)Secondary cosmic-ray neutron-induced single event upset (SEU) is a cause of soft errors on micro electronic devices. Multiple cell upsets (MCUs) are particularly serious problems since it is difficult to recover MCUs. In this study, we have performed irradiation tests of neutrons on 65-nm bulk SRAM at the national metrology institute of Japan (NMIJ) in Advanced Industrial Science and Technology (AIST) and measured SEU cross sections and MCU cross sections to investigate the effect on neutrons with the energies below 10 MeV on soft errors. It was found that SEU cross sections change drastically around 6 MeV. The proportion of MCU to total events does not change very much over the wide range of neutron energy. We also analyzed the total soft error rate (SER) of SEU and MCU by folding the neutron energy-dependent cross section and the flux spectra of the terrestrial neutron at New York and Tokyo. The calculated result indicates that the SER originating from the low-energy neutrons below 10 MeV is mostly negligible in the terrestrial environment.
Kuroda, Junya*; Manabe, Seiya*; Watanabe, Yukinobu*; Ito, Kojiro*; Liao, W.*; Hashimoto, Masanori*; Abe, Shinichiro; Harada, Masahide; Oikawa, Kenichi; Miyake, Yasuhiro*
IEEE Transactions on Nuclear Science, 67(7), p.1599 - 1605, 2020/07
Times Cited Count：4 Percentile：45.45(Engineering, Electrical & Electronic)Soft errors induced by terrestrial radiation in semiconductor devices have been of concern from the viewpoint of their reliability. Generally, to evaluate the soft error rates (SERs), neutron irradiation tests are performed at neutron facility. We have performed SER measurement for the 65-nm bulk SRAM and the FDSOI SRAM at RCNP in Osaka University and CYRIC in Tohoku University. In this study, we performed SER measurement for the same devices at BL10 in J-PARC MLF. The increasing rate of SER by reducing the supply voltage at J-PARC BL10 is larger than those obtained at RCNP and CYRIC. From PHITS simulation, the cause of this difference can be explained by the influence of the protons generated by neutron elastic scattering with hydrogen atoms in the package resin.
Abe, Shinichiro; Sato, Tatsuhiko; Kuroda, Junya*; Manabe, Seiya*; Watanabe, Yukinobu*; Liao, W.*; Ito, Kojiro*; Hashimoto, Masanori*; Harada, Masahide; Oikawa, Kenichi; et al.
Proceedings of IEEE International Reliability Physics Symposium (IRPS 2020) (Internet), 6 Pages, 2020/04
Times Cited Count：2 Percentile：64.23(Engineering, Electrical & Electronic)Single event upsets (SEUs) caused by neutrons have been recognized as a serious reliability problem for microelectronic devices on the ground level. In our previous work, it was found that hydride placed in front of the memory chip has considerably impact on SEU cross sections because H ions generated via elastic scattering of neutrons with hydrogen atoms are only emitted in a forward direction. In this study, the effect of components neighboring transistors on neutron-induced SEUs was investigated for 65-nm bulk SRAMs by using PHITS. It was found that the shape of the SEU cross section around few MeV comes from the thickness and the position of components placed in front of transistors when that components do not contains hydrogen atoms. By considering components adjoin memory cells in the test board used in the simulation, measured data at J-PARC BL10 were reproduced well. In addition, it was found that the effect of components neighboring transistors on neutron-induced SERs does not negligible in terrestrial environment.
Hashimoto, Masanori*; Kobayashi, Kazutoshi*; Furuta, Jun*; Abe, Shinichiro; Watanabe, Yukinobu*
Integration, 69, p.161 - 179, 2019/11
Times Cited Count：10 Percentile：62.95(Computer Science, Hardware & Architecture)Soft error originating from cosmic ray is a serious concern for reliability demanding applications. Device miniaturization and lower voltage operation degrade the immunity of SRAM and flip-flops, and then soft error countermeasures will be demanded in more and more products. This paper characterizes and discusses soft error rates of SRAM and flip-flops in the terrestrial environment with the results of investigation for soft error phenomena by measurements and simulations.
Abe, Shinichiro; Liao, W.*; Manabe, Seiya*; Sato, Tatsuhiko; Hashimoto, Masanori*; Watanabe, Yukinobu*
IEEE Transactions on Nuclear Science, 66(7, Part 2 ), p.1374 - 1380, 2019/07
Times Cited Count：7 Percentile：61.94(Engineering, Electrical & Electronic)Single event upsets (SEUs) caused by secondary cosmic-ray neutrons have recognized as a serious reliability problem for microelectronic devices. Acceleration tests at neutron facilities are convenient to validate soft error rates (SERs) quickly, but some corrections caused from measurement conditions are required to derive realistic SERs at actual environment or to compare other measured data. In this study, the effect of irradiation side on neutron-induced SEU cross sections was investigated by performing neutron transport simulation using PHITS. SERs for 65-nm bulk CMOS SRAMs are estimated using the sensitive volume model. It was found from simulation that SERs for the sealant side irradiation are 30-50% larger than those for the board side irradiation. This difference comes from the difference of production yield and angular distribution of secondary H and He ions, which are the main cause of SEUs. Thus the direction of neutron irradiation should be reported when the result of acceleration tests are published. This result also indicates that SERs can be reduced by equipping device with sealant side facing downward.
Manabe, Seiya*; Watanabe, Yukinobu*; Liao, W.*; Hashimoto, Masanori*; Abe, Shinichiro
IEEE Transactions on Nuclear Science, 66(7), p.1398 - 1403, 2019/07
Times Cited Count：8 Percentile：66.68(Engineering, Electrical & Electronic)Cosmic ray-induced soft errors have been recognized as a major threat for electronics used at ground level. Recently, cosmic-ray muon-induced soft errors have received much attention due to the reduction of soft error immunity on SRAMs. In the previous studies, muon-induced soft error rates (SERs) for various technology devices were predicted with only the positive muon irradiation tests and simulation. In this paper, the muon-induced SEU rates for the 65-nm bulk and UTBB-SOI SRAMs are estimated by using the experimental data of both negative and positive muons. The experimental results showed that the negative muon SEU cross sections for the bulk SRAM are significantly larger than those for the UTBB-SOI. Estimation of muon-induced SEU rates at ground level was performed using PHITS with the experimental results. The muon-induced SER on the first floor of the building was estimated to be at most 10% of the neutron-induced SER on the same floor.
Abe, Shinichiro; Sato, Tatsuhiko
Journal of Nuclear Science and Technology, 53(3), p.451 - 458, 2016/03
Times Cited Count：9 Percentile：64.88(Nuclear Science & Technology)Secondary cosmic-ray neutron-induced soft errors have been recognized as a serious problem affecting the reliability of microelectronic devices. Our developed Multi-scale Monte Carlo simulator called PHYSERD is a reliable code for analysis of soft error. However, PHYSERD takes long computational time to calculate processes of collecting charges because event-by-event technology computer-aided design (TCAD) simulation is time consuming. In this study, the multiple sensitive volume (MSV) model is adopted to estimate collected charge. Secondary cosmic-ray neutron-induced soft errors in an NMOSFET are analyzed based on the MSV model using PHITS. The results are compared with those obtained by PHYSERD based on event-by-event TCAD simulation and by single sensitive volume (SSV) model using PHITS. It is found that PHITS+MSV provides approximate SERs in a shorter time than PHYSERD. Furthermore, PHITS+MSV reproduces SERs and collected charges more accurately than PHITS+SSV by considering the spatial dependence of the charge collection efficiencies.
Abe, Shinichiro; Sato, Tatsuhiko
Proceedings of Radiation Effects on Components and Systems Conference 2016 (RADECS 2016) (Internet), 5 Pages, 2016/00
Secondary cosmic-ray neutron have been recognized as a major cause of soft error for microelectronics on ground. Recently, the effect of secondary cosmic-ray muon on soft error is concerned as semiconductor devices become sensitive to radiation. In this study, we investigate the shielding effect on secondary cosmic-ray neutron- and muon-induced SER. Transports of secondary cosmic-ray neutrons and muons in a building are simulated by PHITS with considering distributions of energy and zenith-angle of secondary cosmic-rays by PARMA 4.0 which is the analytical model for estimating cosmic-ray fluxes nearly anytime and anywhere in the world. The calculated neutron fluxes are in surprisingly good agreement with measured data. The soft errors in 25-nm bulk-design NMOSFET caused by neutrons and muons in the open air and on the first floor of the building are analyzed based on the MSV model using PHITS. It is found that the shielding effect of the building only contributes to decrease neutron-induced SER. As a result, muon-induced SER is about 20% of that caused by neutron on the first floor. Therefore it is necessary to consider the secondary cosmic-ray muon-induced SER when reliability of devices used in the building is evaluated.
Abe, Shinichiro; Sato, Tatsuhiko; Matsuba, Hirotaka*; Watanabe, Yukinobu*
Proceedings of 11th International Workshop on Radiation Effects on Semiconductor Devices for Space Applications (RASEDA-11) (Internet), p.45 - 48, 2015/11
Secondary cosmic-rays have been recognized as a cause of soft errors for microelectronics in terrestrial environment. Recently, the contribution of terrestrial muons to soft errors is concerned for advanced microelectronics because it becomes small and sensitive to radiation. Muons generate energetic secondary through photonuclear interaction and negative muon capture. In the present work, we investigate the effect of these interactions on terrestrial muon-induced soft errors. The analysis of soft error rate (SER) in the 25-nm design rule NMOSFET is performed based on the multiple sensitive volume (MSV) model using PHITS. It is clarified that the terrestrial muon-induced SER is a few or less of neutron-induced SER and it is mainly caused though negative muon capture while the effect of muon photonuclear interaction is small. It is also found that direct ionization only affects soft errors with extremely low critical charge.
Fukui, Hironobu*; Hamaguchi, Masafumi*; Yoshimura, Hisao*; Oyamatsu, Hisato*; Matsuoka, Fumitomo*; Noguchi, Tatsuo*; Hirao, Toshio; Abe, Hiroshi; Onoda, Shinobu; Yamakawa, Takeshi; et al.
Proceedings of 2005 Symposia on VLSI Technology and Circuits, p.222 - 223, 2005/00
no abstracts in English
m CMOS LSIsFukui, Hironobu*; Hamaguchi, Masafumi*; Yoshimura, Hisao*; Oyamatsu, Hisato*; Matsuoka, Fumitomo*; Noguchi, Tatsuo*; Hirao, Toshio; Abe, Hiroshi; Onoda, Shinobu; Yamakawa, Takeshi; et al.
Proceedings of the 6th International Workshop on Radiation Effects on Semiconductor Devices for Space Application (RASEDA-6), p.169 - 172, 2004/10
no abstracts in English
Nashiyama, Isamu
Materiaru Raifu, 9(2), p.69 - 72, 1997/04
no abstracts in English
Hirao, Toshio; Nashiyama, Isamu; Kamiya, Tomihiro; Nishijima, Toshiji*
Nuclear Instruments and Methods in Physics Research B, 104, p.508 - 514, 1995/00
Times Cited Count：29 Percentile：91.66(Instruments & Instrumentation)no abstracts in English
Nashiyama, Isamu; Hirao, Toshio
Hoshasen To Sangyo, 0(61), p.37 - 41, 1994/00
no abstracts in English
Abe, Shinichiro; Sato, Tatsuhiko
no journal, ,
Neutron-induced soft error in a NMOSFET is analyzed based on the multiple sensitive volume (MSV) model using the PHITS code. The result is compared with those obtained by PHYSERD based on event-by-event TCAD simulation and by simple sensitive vollume (SSV) model using the PHITS code. The SER calculated by PHITS-MSV is in good agreement with the SER obtained by PHYSERD. From the comparisons of the collected charges between TCAD simulation and SV model calculations for each single event, it is found that the accuracy of the SER calculation based on the MSV model is improved by considering the spatial dependence of the charge collection efficiency.
Abe, Shinichiro; Sato, Tatsuhiko; Ogawa, Tatsuhiko
no journal, ,
In this study, we have analyzed the impact of the new version of event generator mode (e-mode ver. 2) implemented in PHITS code on evaluations of neutron-induced single event upset (SEU) for semiconductor devices. Single event effect is one of the radiation effects on the semiconductor device. The data stored in the device are flipped by the noise charge created by radiation (so-called SEU), and temporary malfunction is happened in the microelectronic device (so-called soft error). Since neutrons have no charge, they create the noise charge via a nuclear reaction. Therefore the accuracy of the nuclear reaction model is important for validating SEU cross section by simulation. The improvement of e-mode ver. 2 has been validated by comparing the product cross sections of secondary ions with those stored in the JENDL-4.0. From the comparison of SEU cross sections by e-mode ver. 2 with those obtained by e-mode ver. 1 (existing version), it is found that the improvement of e-mode have an impact on the SEU analysis.
Abe, Shinichiro; Sato, Tatsuhiko; Kato, Takashi*; Matsuyama, Hideya*
no journal, ,
Radiation-induced charges in semiconductor device cause temporary and non-destructive faults (the so-called soft errors) in microelectronic devices. The model which estimates charges collected in the storage node of a memory element quickly and accurately is necessary to evaluate soft error rate by simulation. It is predicted that charge collection process in FinFET is differs from that in planar process due to differences in their structures. In this study, we perform systematic investigation of charge collection caused by radiation using 3-D TCAD simulator HyENEXSS in order to construct the charge collection model for FinFET. As a result of that investigation, it is found that temporal conduction between source node and drain node arises at few psec. It is also clarified that the temporal conduction has a major impact on charge collection process when charges deposited only near fin. Meanwhile, it is found that diffusion process is the main cause of charge collection when charges deposited in the substrate.
Abe, Shinichiro; Sato, Tatsuhiko; Kato, Takashi*; Matsuyama, Hideya*
no journal, ,
Radiation-induced charges in semiconductor device cause temporary and non-destructive faults (the so-called soft errors) in microelectronic devices. The model which estimates charges collected in the storage node of a memory element quickly and accurately is necessary to evaluate soft error rate by simulation. In our previous study, it was found that charge collection efficiency in FinFET change with the position and the amount of deposited charge. In this study, we construct the multiple sensitive volume model to estimate collected charge for FinFET. We obtained arbitrary charge deposition events by PHITS and analyzed charge collection of these events for by 3-D TCAD simulator HyENEXSS, the simple sensitive volume model and the multiple sensitive volume model. From the comparisons of the collected charges, it is found that the accuracy of collected charge calculation is improved by considering the variations of charge collection efficiency accompanying the position and the amount of deposited charge.
