We became aware that some of the calculation results in our article(1) are wrong by a factor of π⁴. This error affects the following parts of the article:

• Paragraph following eq 7: The efficiency of the inelastic process in vacuum is of the order of 10^–8 eV^–1 and not 10^–6 eV^–1.
• Figure 2b: The vertical axis needs to be divided by π⁴.
• Figure 4b: The vertical axis needs to be divided by π⁴.
• Last paragraph on page 4198: The probability for a tunneling electron to excite an MIM mode is of the order of 10^–3 eV^–1 and not 10^–1 eV^–1.

Paragraph following eq 7: The efficiency of the inelastic process in vacuum is of the order of 10^–8 eV^–1 and not 10^–6 eV^–1. Figure 2b: The vertical axis needs to be divided by π⁴. Figure 4b: The vertical axis needs to be divided by π⁴. Last paragraph on page 4198: The probability for a tunneling electron to excite an MIM mode is of the order of 10^–3 eV^–1 and not 10^–1 eV^–1. This error does not affect the main conclusion of the article, namely, that the conversion efficiency of MIM modes to SPPs is similar to the conversion efficiency of MIM modes to photons. This article references 1 other publications.

中文翻译：

---

校正隧道尽头的光线

我们意识到，一些在我们的文章（1）的计算结果是错了π的因素⁴。此错误影响文章的以下部分：

• 等式7之后的段落：真空中的非弹性过程的效率约为10 ^–8 eV ^–1，而不是10 ^–6 eV ^–1。
• 图2b：垂直轴需要由π被划分⁴。
• 图4b：垂直轴需要由π被划分⁴。
• 第4198页的最后一段：隧穿电子激发MIM模式的概率约为10 ^–3 eV ^–1，而不是10 ^–1 eV ^–1。

等式7之后的段落：真空中的非弹性过程的效率约为10 ^–8 eV ^–1，而不是10 ^–6 eV ^–1。图2b：垂直轴需要由π被划分⁴。图4b：垂直轴需要由π被划分⁴。第4198页的最后一段：隧穿电子激发MIM模式的概率约为10 ^–3 eV ^–1，而不是10 ^–1 eV ^–1。此错误不会影响本文的主要结论，即MIM模式到SPP的转换效率类似于MIM模式到光子的转换效率。本文引用了其他1个出版物。