1.State Key Laboratory of Functional Materials for Informatics, Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, Shanghai 200050, China 2.Center for Excellence in Superconducting Electronics, Chinese Academy of Sciences, Shanghai 200050, China 3.University of Chinese Academy of Sciences, Beijing 100049, China
Fund Project:Project sponsored by the Natural Science Foundation of Shanghai, China (Grant No. 18ZR1447400)
Received Date:10 March 2021
Accepted Date:20 April 2021
Available Online:07 June 2021
Published Online:05 September 2021
Abstract:Superconducting quantum interference device (SQUID) has been used as an extremely sensitive flux sensor up to now. Series SQUID array (SSA) is made up of several identical element-SQUIDs in series, in which each element-SQUID is coupled with the same set of input coils by mutual inductance to realize the amplified output of the input current. From the noise viewpoint, each element-SQUID in SSA is independent of each other, resulting in the total voltage noise across the array rising linearly with the square root of the number of element-SQUIDs. From the perspective of input signals, since the signals come from the same set of input coils, the voltage output of the array is enlarged with the proportion of element-SQUID number, N. Taken together, the signal-to-noise ratio of SSA is increased by $\sqrt{ {N}}$ times, or the flux noise of SSA is reduced by 1/$\sqrt{ {N}}$ times compared with that of an element-SQUID ideally. However, with the increase in the number of element-SQUIDs in series, the chip design of SSA becomes more complicated, which puts forward higher requirements for the consistency and stability of its fabrication process. Besides, there exists a certain flux coherence between element-SQUIDs in SSA, whose normal operation depends on the working state of each element-SQUID in the array. In this paper, the fabrication of series SQUID array is carried on the autonomous superconducting micro-nano process platform, with a yield rate reaching over 80% on a 4-inch standard silicon wafer. Two kinds of SSAs with 200 and 800 element-SQUIDs, respectively, are integrated in a meandering way on a chip in a millimeter area. Home-made directly-coupled readout circuit is used to obtain the characteristics of SSA. The experimental results reveal that the flux noise at best working point is as low as 0.5μ$\varPhi _{\text{0}}/\sqrt{\text{Hz}}$ and the current sensitivity is about 35 μA/Φ0, thus, the equivalent input current noise is achieved at a level of 18 pA/$ \sqrt{\text{Hz}} $. Additionally, the dependence of relevant parameters in array on the number of element-SQUIDs is verified, which is consistent with theoretical expectation basically. These show that the reliability of device design and the consistency of fabrication process perform well, thus laying the technical foundation for developing the low-noise SQUID amplifier and the multiplexed readout of low-impedance detectors. Keywords:superconducting quantum interference device/ flux noise/ superconducting fabrication process
图 1 串联SQUID阵列(SSA)示意图. Ii和If分别表示通过输入线圈和反馈线圈的电流, Vo表示在偏置电流为Ib时SSA的输出电压 Figure1. Schematic diagram of series SQUID array (SSA). Ii and If represent the current through the input coil and feedback coil, respectively. Vo is the output voltage of SSA biased at the current Ib.
为了研究SSA与SQUID数量的关系, 在200个SQUID基础上, 增加SQUID数量达到800, 并可以抽出不同数目的SQUID构成小的SSA, 由此来验证器件性能与串联个数之间的关系. 将SSA的正常态电阻Rn、最大电压调制幅度Vpp和最大VΦ与SQUID数量的关系汇总到图6中. 从线性拟合结果看, 以上参数与SQUID的数量呈现较好的线性关系, 这也说明了SSA设计中的SQUID之间保持了一定的独立性, 没有因串扰或设计缺陷造成SSA性能恶化[21]. 同时, 也说明了我们的制备工艺保持了很好的稳定性和一致性. 但是, 通过对SQUID数量从20到200不等的SSA的噪声测试来看, 虽然随着SQUID数量的增加, SSA的磁通噪声呈下降趋势, 但仍然未达到理论值. 也就是说, SSA中的SQUID还没有达到理想的非相关. 这启示我们, 现有的SSA设计和测试系统仍然有优化的空间, 同时可以进一步提升输入电流灵敏度, 以获得更低的输入电流噪声水平. 图 6 (a) SSA的正常态电阻Rn(方点)和最大电压调制幅度Vpp(星点)与SQUID数量之间的关系, 实线和虚线均为测试数据的线性拟合; (b) SSA的最大磁通-电压转换系数VΦ与SQUID数量之间的关系, 其中点为测试值, 线为线性拟合的结果 Figure6. (a) Normal resistance Rn (square dot) and maximum voltage swing Vpp (star dot) of SSA dependence on the SQUID-element number, in which the solid and dashed line are linear fittings, respectively (b) maximum flux-to-voltage coefficient VΦ dependence on the SQUID-element number, where the dots are experimental data and the line is the result of linear fitting.