物理化学学报 >> 2020, Vol. 36 >> Issue (12): 2005033.doi: 10.3866/PKU.WHXB202005033

所属专题: 神经界面

论文 上一篇    下一篇

轻掺杂硅基神经电极的光噪声消减

魏春蓉1,2,4, 王飞1,2,4, 裴为华1,2,4,*(), 刘智多1,2,4, 毛旭瑞1, 赵宏泽1,2,4, 王思凯1,2,4, 王毅军1,2,4, 杨晓伟1,2,4, 刘媛媛3,4, 赵姗姗5, 归强1,2,4, 陈弘达1,2   

  1. 1 中国科学院半导体研究所,集成光电子学国家重点实验室,北京 100083
    2 中科院脑科学与智能技术卓越中心,北京 100083
    3 半导体集成技术工程研究中心,北京 100083
    4 中国科学院大学,北京 100049
    5 天津职业技术师范大学信息技术与工程学院,天津 300222
  • 收稿日期:2020-05-12 录用日期:2020-06-12 发布日期:2020-06-15
  • 通讯作者: 裴为华 E-mail:peiwh@semi.ac.cn
  • 基金资助:
    国家重大研究发展计划项目(2017YFA0205903);国家重大研究发展计划项目(2017YFA0701100);国家重大研究发展计划项目(2016YFB0402405);国家自然科学基金(61634006);国家自然科学基金(61671424);中国科学院战略性先导科技专项(XDB32030100);中国科学院战略性先导科技专项(XDB32040200);中国科学院前沿科学重点研究项目(QYZDY-SSW-JSC004)

Light-Induced Noise Reduction of Lightly Doped Silicon-based Neural Electrode

Chunrong Wei1,2,4, Fei Wang1,2,4, Weihua Pei1,2,4,*(), Zhiduo Liu1,2,4, Xurui Mao1, Hongze Zhao1,2,4, Sikai Wang1,2,4, Yijun Wang1,2,4, Xiaowei Yang1,2,4, Yuanyuan Liu3,4, Shanshan Zhao5, Qiang Gui1,2,4, Hongda Chen1,2   

  1. 1 The State Key Laboratory on Integrated Optoelectronics, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, P. R. China
    2 CAS Center for Excellence in Brain Science and Intelligence Technology, Beijing 100083, P. R. China
    3 Engineering Research Center for Semiconductor Integrated Technology, Beijing 100083, P. R. China
    4 University of Chinese Academy of Sciences, Beijing 100049, P. R. China
    5 School of Information Technology and Engineering, Tianjin University of Technology and Education, Tianjin 300222, P. R. China
  • Received:2020-05-12 Accepted:2020-06-12 Published:2020-06-15
  • Contact: Weihua Pei E-mail:peiwh@semi.ac.cn
  • Supported by:
    the National Key Technologies Research and Development Program of China(2017YFA0205903);the National Key Technologies Research and Development Program of China(2017YFA0701100);the National Key Technologies Research and Development Program of China(2016YFB0402405);the National Natural Science Foundation of China(61634006);the National Natural Science Foundation of China(61671424);the Strategic Priority Research Program of Chinese Academy of Sciences Pilot Project(XDB32030100);the Strategic Priority Research Program of Chinese Academy of Sciences Pilot Project(XDB32040200);the Key Research Programs of Frontier Sciences, CAS(QYZDY-SSW-JSC004)

摘要:

硅基神经电极是记录神经细胞放电活动的一种实用工具。使用标准的集成电路加工技术,在宽度仅为70 μm的单个硅基针上能排布上千个电极记录位点。光遗传学的发展使控制神经元活动更加精确,通过在给予光刺激的同时记录神经元的电活动,可以获取更丰富的脑活动信息。当使用黄光或蓝光刺激神经元时,光子的能量大于硅衬底的禁带宽度,价带电子被激发到导带,从而生成电子-空穴对。因此,在光刺激下使用硅基神经电极时,硅基板中的光生载流子将严重干扰电极的信噪比。为满足在光刺激同时记录电活动的应用需求,必须减少光对硅基神经电极的噪声干扰。传统的降噪方法是使用重掺杂硅作为衬底材料,通过增加杂质浓度来降低载流子寿命,从而降低硅电极的光学噪声。但是,重掺杂的硅衬底比轻掺杂的硅衬底具有更多的晶格缺陷,这使得硅基电极更加脆弱,并且该方法与标准的集成电路加工技术不兼容。通过分析在轻掺杂硅衬底上制造电极的光致噪声机理,我们发现由光激发产生的载流子的不均匀分布将使轻掺杂硅衬底极化。由光致极化引起的电势将影响在其上制造的电极。将轻掺杂硅衬底金属化和接地将有效降低极化电位。使用这种方法,由光诱发的噪声幅度将下降到原始值的0.87%。为了确保神经元的放电率,将光刺激脉冲频率选择为20 Hz。在1 mW·mm-2的光照下,电极的背景噪声可控制在45 μV以下,可以满足一般光遗传学应用的需求。经过上述方式改造后的轻掺杂硅衬底将满足光遗传学应用对神经探针的要求。与传统的通过重掺杂整个衬底降低光噪声方法不同,该方法与标准的集成电路加工技术兼容,为利用标准集成电路加工技术制备高密度、高通量硅电极提供了噪声消除方法。

关键词: 光噪声, 光生载流子, 光遗传学, 轻掺杂, 硅基神经微探针

Abstract:

Silicon-based neural probes are practical tools for recording neural cell firing. A single silicon-based needle with a width of only 70 μm, prepared using the standard complementary metal-oxide-semiconductor (CMOS) process technology, can contain thousands of electrode-recording sites. Optogenetics has made control over neuronal activity more precise. By recording the electrical activity of neurons stimulated by light, more information about brain activity can be recorded and analyzed. When yellow light or blue light is used to stimulate neurons, the photon energy is greater than the forbidden bandwidth of the silicon substrate, and the valence-band electrons are excited to the conduction band, generating electron-hole pairs. The photoinduced carrier in the silicon substrate severely disrupts the probe's signal-to-noise ratio. Decreasing the disturbance caused by light is a pragmatic way to execute recording and stimulating simultaneously. The traditional noise reduction method involves using heavily doped silicon as the substrate, reducing the carrier life by increasing the impurity concentration, and then reducing the noise of the silicon electrode under illumination. However, the heavily doped silicon substrate has more lattice defects than its lightly doped counterparts, which makes the silicon electrode fragile, and this method is not compatible with the standard CMOS process technology. On analyzing the photoinduced noise mechanism of manufacturing electrodes on lightly doped silicon substrates, we found that the inhomogeneous distribution of carriers generated by light excitation polarizes lightly doped silicon substrates. The potential caused by photoinduced polarization will affect the electrodes fabricated on it. Metalizing and grounding the lightly doped silicon substrate will effectively decrease the polarization potential. On using this method, the noise amplitude caused by the illumination can drop to 0.87% of the original value. To ensure an appropriate firing rate of neurons, the photo-stimulation frequency was chosen to be 20 Hz. Under the illumination of 1 mW·mm-2, the background noise of the electrode could be controlled below 45 μV, which meets the needs for general optogenetics applications. Modification of the lightly doped silicon substrate will meet the requirements of the neural electrode for optogenetics applications. Unlike the traditional method of reducing light-induced noise by heavily doping the entire substrate, the noise reduction method of lightly doped silicon substrate is compatible with the standard CMOS process technology. It provides a noise cancellation method for the preparation of silicon-based neural microelectrodes with dense recording sites and high channel count using standard CMOS processes.

Key words: Light-induced noise, Photoinduced carrier, Optogenetics, Lightly doped, Silicon-based neural probe

MSC2000: 

  • O646