Characterizing low-energy neutrons (<; 10 MeV)-induced single-event upsets (SEUs) and multiple cells upsets (MCUs) is essential to validate the current standard for terrestrial soft error rate (SER) and investigate its enhancement. Following our preliminary analysis on the contribution of the low-energy neutrons to the total terrestrial SER at the nominal operating voltage of 1.0 V by Liao et al. (2020), this article newly presents and analyzes the data measured at the low operating voltage of 0.4 V. The dependence of SEU cross section on the neutron energy is similar between the operating voltages of 0.4 and 1.0 V, including onset energy of around 6 MeV. The existence of MCUs at 4.1-MeV neutrons was also confirmed at both the operating voltages. Based on the measurement, we approximate the dependence of SEU and MCU cross sections as Weibull functions of the neutron energy. The terrestrial SER of SEUs and MCUs was calculated by folding the Weibull function and the flux spectrum. The calculated result indicates that the SER originating from the low-energy neutrons is less than 6% in the terrestrial environment at New York and Tokyo City. We confirm that disregarding the flux of neutrons below 10 MeV in the acceleration factor calculation at accelerated neutron tests, which follows the current standard defined in JESD89, could give a reasonable SER estimation accuracy for both SEUs and MCUs. On the other hand, for covering the beams having an extremely high proportion of low-energy neutrons, considering the flux of neutrons above 6 MeV would be an option for better SER estimation.

中文翻译：

---

表征低能中子诱发 SEU 和 MCU 的能量依赖性及其对 65 nm Bulk SRAM 中地面 SER 估计的影响


表征低能中子 (<; 10 MeV) 引起的单粒子扰乱 (SEU) 和多细胞扰乱 (MCU) 对于验证当前的地面软错误率 (SER) 标准并研究其增强至关重要。 Liao 等人对标称工作电压 1.0 V 下低能中子对地面总 SER 的贡献进行了初步分析。 (2020)，本文新提出并分析了在 0.4 V 低工作电压下测得的数据。SEU 截面对中子能量的依赖性在 0.4 和 1.0 V 工作电压之间相似，包括 6 左右的起始能量兆伏。 4.1-MeV 中子下的 MCU 的存在也在两种工作电压下得到了证实。根据测量结果，我们将 SEU 和 MCU 截面的依赖性近似为中子能量的威布尔函数。 SEU 和 MCU 的地面 SER 是通过折叠威布尔函数和通量谱来计算的。计算结果表明，在纽约和东京市的陆地环境中，源自低能中子的SER小于6%。我们确认，在加速中子测试的加速因子计算中忽略低于 10 MeV 的中子通量（遵循 JESD89 中定义的当前标准），可以为 SEU 和 MCU 提供合理的 SER 估计精度。另一方面，为了覆盖具有极高比例的低能中子的束流，考虑高于 6 MeV 的中子通量将是更好的 SER 估计的一个选择。