New directions in piezoelectric phononic integrated circuits: exploiting field confinement (SOUNDMASTER)

在芯片级平台上操纵光和千兆赫（GHz）频率声波之间的类比已经被广泛探索，硅光子学的思想，主要是强几何约束和路由，被应用于声波来开发声子集成电路（PnICs）。值得注意的是，虽然光声类比通常应用于应变场，但压电材料中的声波也具有共传播电磁场（EM）。这种电磁场振荡频率为GHz，但仅限于声波波长（约小10^5），支撑了压电谐振器和射频（RF）设备中滤波器的主导地位。尽管PnICs在过去十年中取得了进展，但大多数基于体波和表面波的压电器件仍然依赖于弱横向约束和准平面声波的操纵。这个项目试图回答这样一个问题：如果一个人能够主动控制和操纵这些在波导几何形状中具有强横向约束的共传播电磁场，那么我们可以实现什么样的定性新传感和信息处理范式？我们表明，通过利用PnIC平台上的强场增强，可以设计用于电子自旋共振实验的共振磁场近场发生器，可以将自旋探测灵敏度提高~10^7，直至热噪声极限。此外，通过设计波导几何形状中的声电相互作用，声学移相器和模式选择，可以实现对声波传播施加主动控制的非互易放大器，并将有源无源器件集成推向其极限，使射频信号处理的新一类器件成为可能。为了确保这些设备的性能符合预期，我们还解决了一个重要的问题：我们能否以接近统一的效率将GHz频率的声波输入和输出微米级设备？

The analogy between manipulating light and gigahertz (GHz) frequency acoustic waves in chip-scale platforms has been extensively explored, with ideas from silicon photonics, mainly strong geometric confinement and routing, being applied to acoustic waves to develop phononic integrated circuits (PnICs). It is important to note that while the light-sound analogy is generally applied to the strain field, acoustic waves in piezoelectric materials have a co-propagating electromagnetic (EM) field as well. This EM field, which oscillates at GHz frequencies, but is confined to acoustic wavelengths (~10^5 smaller), underpins the dominance of piezoelectric resonators and filters in radio frequency (RF) devices. Despite the advances made in PnICs in the past decade, the majority of piezoelectric devices, both bulk and surface wave based, still rely on weak transverse confinement and manipulation of quasi plane acoustic waves.This project seeks to answer the question: if one could actively control and manipulate these co-propagating EM fields in waveguide geometries with strong transverse confinement, what qualitatively new sensing and information processing paradigms can one enable?We show that by exploiting strong field enhancement in a PnIC platform, one can design resonant magnetic near field generators for electron spin resonance experiments that can improve the spin detection sensitivity by ~10^7, down to the thermal noise limit. In addition, by engineering acousto-electric interactions in waveguide geometries, acoustic phase shifters and mode-selective, non-reciprocal amplifiers can be realized that exert active control on acoustic wave propagation, and push active passive device integration to its limit, enabling a new class of devices for RF signal processing. To ensure these devices perform as expected, we also address the important question: can we get GHz frequency acoustic waves into and out of micrometre-scale devices with near-unity efficiency?