We outline a method to control time-varying effusion cell flux in molecular beam epitaxy. This can be used to grow 2D nanostructures with precise continuously graded alloys, and we explore the specific case of controlling an Al cell for the growth of continuously graded Al${}_x$Ga${}_{1-x}$As parabolic quantum wells. We develop a simple model of the Al cell’s thermal dynamics, including the effects of shutter transients. This model’s parameters can be empirically derived using a beam flux monitoring ion gauge. Though simple, this model can already be used to generate near-arbitrary time-varying fluxes (within the physical limits of the cell). For growths that require higher precision than the simple model can achieve, we introduce an iterative correction scheme to eliminate any lingering errors. Using this method, we control the composition in a 3 THz Al${}_x$Ga${}_{1-x}$As parabolic quantum well array to a root mean squared error of just $\pm 0.18$ %Al. This accuracy is achieved at the standard growth rates necessary for thick structures (0.15–0.25 nm/s). The approach is quite generally applicable to other composition profiles or other situations where time-dependent flux control is necessary, and it could be extended to other effusion cells.

中文翻译：

---

分子束外延中时变渗出细胞通量的精确控制

我们概述了一种控制分子束外延中随时间变化的渗出细胞通量的方法。这可用于生长具有精确连续梯度合金的 2D 纳米结构，我们探索了控制 Al 电池以生长连续梯度 Al 的特定情况${}_X$嘎${}_{1-X}$作为抛物线量子阱。我们开发了一个简单的铝电池热动力学模型，包括快门瞬变的影响。该模型的参数可以使用束流监测离子规凭经验推导出来。尽管很简单，但该模型已经可以用于生成近乎任意的时变通量（在单元的物理限制内）。对于需要比简单模型更高的精度的增长，我们引入了迭代校正方案来消除任何挥之不去的错误。使用这种方法，我们控制了 3 THz 铝中的成分${}_X$嘎${}_{1-X}$作为抛物线量子阱阵列的均方根误差仅为 $\pm 0.18$ %铝。这种精度是在厚结构所需的标准生长速率 (0.15–0.25 nm/s) 下实现的。该方法非常普遍适用于其他成分分布或其他需要时间依赖性通量控制的情况，并且可以扩展到其他积液细胞。