1.School of Optoelectronic Engineering and Instrumentation Science, Dalian University of Technology, Dalian 116024, China 2.School of Physics, Dalian University of Technology, Dalian 116024, China 3.Department of Electrical and Computer Engineering, National University of Singapore, Singapore 119260, Singapore
Fund Project:Project supported by the National Natural Science Foundation of China (Grant No. 51607022) and the Fundamental Research Funds for the Central Universities, China (Grant No. DUT17LK13).
Received Date:23 March 2019
Accepted Date:23 May 2019
Available Online:01 August 2019
Published Online:05 August 2019
Abstract:Magnetic fields are generally sensed by a device that makes use of the Hall effect. Hall-effect sensors are widely used for proximity switching, positioning, speed detecting for the purpose of control and condition monitoring. Currently, the Hall sensor products are mainly based on the narrow-bandgap Si or GaAs semiconductor, and they are suitable for room temperature or low temperature environment, while the novel wide-bandgap GaN-based Hall sensors are more suitable for the application in various high-temperature environments. However, the spatial structure of the GaN-based sensor is mainly horizontal and hence it is only able to detect the magnetic field perpendicular to it. To detect the parallel field on the sensor surface, the vertical structure device is required despite encountering many difficulties in technology, for example reducing the vertical electric field in the two-dimensional electron gas (2-DEG) channel. The vertical Hall sensor has not been reported so far, so it is technically impossible to realize three-dimensional magnetic field detection on single chip. To address the mentioned issues, in this paper we propose a design of the vertical Hall sensor based on the wide-bandgap AlGaN/GaN heterojunction material, which adopts a shallow etching of 2-DEG channel barrier to form a locally trenched structure. The material parameters and physical models of the proposed device are first calibrated against real device test data, and then the key structural parameters such as device electrode spacing ratio, mesa width and sensing electrode length are optimized by using technology computer aided design, and the device characteristics are analyzed. Finally, the simulation results confirm that the proposed Hall sensor has a higher sensitivity of magnetic field detection and lower temperature drift coefficient ($\sim $600 ppm/K), and the device can work stably in a high-temperature (greater than 500 K) environment. Therefore, the vertical and horizontal devices can be fabricated simultaneously on the same wafer in the future, thus achieving a three-dimensional magnetic field detection in various high-temperature environments. Keywords:magnetic sensor/ AlGaN/GaN heterojunction/ two-dimensional electron gas/ high temperature stability
图 4 霍尔电压(或2-DEG电子浓度)与势垒层剩余厚度的关系 Figure4. Hall voltage (or 2-DEG concentration) vs. AlGaN barrier thickness.
图5展示了感测电极下方势垒层剩余厚度d = 7 nm时, 在无外加磁场和外加磁场B = 1 T情况下, 传感器电流密度空间分布对比情况. 从图5可以看出, 当器件表面平行方向不存在磁场时, 两侧的感测电极下方电流密度呈空间对称分布, 因此感测电极S1和S2之间电势差为零, 如图5(a)所示; 而当垂直纸面向里方向存在磁场时, 感测电极S1下方的电子在磁场中受到洛伦兹力作用发生向下偏移, 而S2下方的电子运动向上偏移, 两边感测电极处纵向束缚电场的减弱间接加速了电子在纵向的偏移过程, 稳定后感测电极S1和S2之间将产生明显的电势差, 如图5(b)所示, 图中虚线清晰地标明两侧电流空间分布差值情况. 图 5 当d = 7 nm时, 传感器电流密度空间分布对比 (a)无外加磁场; (b)外加磁场B = 1 T Figure5. Comparisons of current density distribution in vertical Hall sensor with d = 7 nm under the conditions of (a) B = 0 and (b) B = 1 T.
霍尔传感器中的结构尺寸, 特别是L2/L1比值、感测电极长度l2和器件台面宽度w等参数对传感器灵敏度影响较大, 本文在仿真过程重点对这些参数进行优化, 仿真中采用的激励电流为0.5 mA, 外加磁场强度B = 0.5 T. 图6展示了霍尔传感器电流相关敏感度SI与L2/L1比值的关系, 图中数据点为经过对数十个不同尺寸器件仿真模拟并经过计算得到. 仿真中设置感测电极的长度l2 = 1 μm, 器件的宽度w = 10 μm. 图6中虚线是对仿真所得数据点分布进行线性拟合得出的, 其数学关系式已列在图中. 从仿真结果可以看出, 电流相关敏感度随着L2/L1比值的增加而增加. 其原因是: 随着L2/L1比值增加, 感测电极相对两侧主电极的位置发生改变, 当感测电极靠近器件两侧时, 器件的几何因子G增加[21,26], 由前面表达式(2)可知电流相关敏感度也随之增加; 另外, 随着L2/L1比值增加, 靠近两侧主电极处的载流子受洛伦兹力的影响增强, 感测电极之间的电势差逐渐增加, 因此霍尔电压增加, 器件的电流相关敏感度也随之增加. 因此在实际器件设计制作过程中, 为了增大传感器的灵敏度, 可以适当减小主电极C0与C1(或C2)之间的间距L1, 或者适当增加C0与S1 (或S2)之间的间距L2. 图 6 电流相关敏感度SI与L2/L1比值的关系 Figure6. Current-related sensitivity as a function of the ratio of L2/L1.
图7展示了电流敏感度以及输入电阻仿真结果, 与感测电极长度的关系, 其中感测电极l2的长度变化范围为0.50—2 μm, 而主电极间距L1 = 4.05 μm, 感测电极与主电极间距L2 = 2 μm以及器件宽度w = 10 μm. 由仿真结果可知, 器件的敏感度随着感测电极长度的增加有所增加, 当感测电极长度为1.25 μm时, 器件的峰值电流相关敏感度为22.8 V/(A·T). 随着感测电极长度继续增加, 器件的输入电阻呈线性增加, 而敏感度逐渐下降. 其原因是: 器件的电流相关敏感度的大小与器件导电沟道中的载流子浓度成反比, 随着感测电极长度的增加, 其下方势垒层经过刻蚀后, 异质结界面处的2-DEG浓度平均值有所降低[31], 因此电流敏感度逐渐增加; 而当感测电极的长度过大时, 两侧有效电势差部分抵消, 因此霍尔电压开始下降[25]. 综合考虑, 本文选取其最优值为l2 = 1.25 μm. 图 7 电流相关敏感度SI(或输入电阻Rin)与感测电极长度l2的关系 Figure7. Current-related sensitivity and input resistance as a function of the l2.
接下来进一步优化霍尔传感器台面宽度w, 仿真中w值分别设置为2, 4, 6, 8, 10 μm. 器件的其他参数如下: 主电极间距L1 = 4.05 μm, 感测电极与主电极间距L2 = 2 μm, 感测电极长度l2 = 1.25 μm. 仿真结果如图8所示, 随着器件台面宽度的缩减, 电流敏感度和输入电阻逐渐上升, 其变化规律符合(2)式. 器件的霍尔电压VH及电流相关敏感度SI与器件的宽度w成反比关系, 即w越大, 器件的霍尔电压及电流相关敏感度均降低. 因此理论上器件宽度w应当减小[21], 但是电流敏感度的提高将以牺牲器件输入电阻为代价. 此外, 考虑实验室中利用常规光刻技术制作传感器, 线宽越小曝光和剥离过程难度越大, 成品率更低, 因此在实际器件设计中应当折衷考虑, 同时结合电路匹配情况进行选择. 图 8 电流相关敏感度(或输入电阻)与器件宽度w的关系 Figure8. Current-related sensitivity and input resistance as a function of the w.
图9展示了不同工作温度下, GaN材料垂直型霍尔传感器的霍尔电压随外加磁场强度的变化关系. 仿真中, 选取主电极间距L1 = 4.05 μm、感测电极的长度l2 = 1.25 μm、感测电极与中心主电极的距离L2 = 2 μm以及器件的宽度w = 3 μm. 传感器工作中激励电流选取为0.3 mA, 外加磁场强度变化范围为–0.5 T ≤ B ≤ 0.5 T, 温度变化范围为300—525 K. 从图9可以看出, 无论正反磁场条件下或者不同环境温度下, 输出霍尔电压与磁场大小均呈现良好的线性关系. 进一步利用(2)和(3)式可以计算得到电流相关敏感度和温漂系数. 图10展示了传感器敏感度随工作温度的变化关系, 其中虚线由线性回归拟合得到, 图中列出了利用该线性关系计算得到的温漂系数. 从仿真结果可知, 室温下(T = 300 K)器件的电流敏感度为75.7 V/(A·T), 在300—375 K温度范围内器件的温度漂移系数为637.4 ppm/K, 在375—525 K温度范围内器件的温度漂移系数为–613.0 ppm/K, 图10电流敏感度变化趋势反映真实传感器工作过程中的物理规律[32]. 其原因是, 材料迁移率主要受电离杂质散射和晶格散射这两种机制的影响. 当传感器处于相对较低工作温度时, 载流子迁移率主要受电离杂质散射影响, 随着温度上升, 该散射影响减弱, 因此电流敏感度增加; 而当传感器处于高温工作情况时, 载流子迁移率主要受晶格振动散射影响, 以声学散射为主且晶格散射随着温度的升高而增强, 因此电流敏感度开始下降. 但本文宽带隙GaN材料霍尔传感器相对其他窄带隙材料, 晶格散射对其敏感度影响明显更小. 图 9 器件输出电压随磁场和工作温度的变化 Figure9. Temperature dependence of output Hall voltage as a function of magnetic induction.
图 10 电流相关敏感度随工作温度的变化 Figure10. Current-related sensitivity as a function of temperature.