Department of Chemistry, University of California, Berkeley, California 94609, USA.
J Chem Phys. 2021 Jan 7;154(1):014107. doi: 10.1063/5.0038652.
Using large deviation theory and principles of stochastic optimal control, we show that rare molecular dynamics trajectories conditioned on assembling a specific target structure encode a set of interactions and external forces that lead to enhanced stability of that structure. Such a relationship can be formulated into a variational principle, for which we have developed an associated optimization algorithm and have used it to determine optimal forces for targeted self-assembly within nonequilibrium steady-states. We illustrate this perspective on inverse design in a model of colloidal cluster assembly within linear shear flow. We find that colloidal clusters can be assembled with high yield using specific short-range interactions of tunable complexity. Shear decreases the yields of rigid clusters, while small values of shear increase the yields of nonrigid clusters. The enhancement or suppression of the yield due to shear is rationalized with a generalized linear response theory. By studying 21 unique clusters made of six, seven, or eight particles, we uncover basic design principles for targeted assembly out of equilibrium.
利用大偏差理论和随机最优控制原理，我们表明，在组装特定目标结构的条件下，罕见的分子动力学轨迹编码了一组相互作用和外力，这些相互作用和外力导致该结构的稳定性增强。这种关系可以被表述为变分原理，为此我们开发了一种相关的优化算法，并利用它来确定非平衡稳态下靶向自组装的最佳力。我们在胶体团簇在线性剪切流中组装的模型中说明了这种反向设计的观点。我们发现，使用可调复杂度的特定短程相互作用，可以以高收率组装胶体团簇。剪切降低了刚性团簇的收率，而小剪切值增加了非刚性团簇的收率。用广义线性响应理论可以解释由于剪切而导致的产率的增强或抑制。通过研究由六个、七个或八个粒子组成的 21 个独特的团簇，我们揭示了非平衡状态下靶向组装的基本设计原则。
