Kumar Priyank V, Rossi Tuomas P, Marti-Dafcik Daniel, Reichmuth Daniel, Kuisma Mikael, Erhart Paul, Puska Martti J, Norris David J
Optical Materials Engineering Laboratory , ETH Zurich , 8092 Zurich , Switzerland.
Department of Physics , Chalmers University of Technology , 41296 Gothenburg , Sweden.
ACS Nano. 2019 Mar 26;13(3):3188-3195. doi: 10.1021/acsnano.8b08703. Epub 2019 Feb 28.
Plasmon-induced hot-carrier transfer from a metal nanostructure to an acceptor is known to occur via two key mechanisms: (i) indirect transfer, where the hot carriers are produced in the metal nanostructure and subsequently transferred to the acceptor, and (ii) direct transfer, where the plasmons decay by directly exciting carriers from the metal to the acceptor. Unfortunately, an atomic-level understanding of the direct-transfer process, especially with regard to its quantification, remains elusive even though it is estimated to be more efficient compared to the indirect-transfer process. This is due to experimental challenges in separating direct from indirect transfer as both processes occur simultaneously at femtosecond time scales. Here, we employ time-dependent density-functional theory simulations to isolate and study the direct-transfer process at a model metal-acceptor (Ag-CdSe) interface. Our simulations show that, for a 10 fs Gaussian laser pulse tuned to the plasmon frequency, the plasmon formed in the Ag-CdSe system decays within 10 fs and induces the direct transfer with a probability of about 40%. We decompose the direct-transfer process further and demonstrate that the direct injection of both electrons and holes into the acceptor, termed direct hot-electron transfer (DHET) and direct hot-hole transfer (DHHT), takes place with similar probabilities of about 20% each. Finally, effective strategies to control and tune the probabilities of DHET and DHHT processes are proposed. We envision our work to provide guidelines toward the design of metal-acceptor interfaces that enable more efficient plasmonic hot-carrier devices.
（i）间接转移，即热载流子在金属纳米结构中产生，随后转移到受体；（ii）直接转移，即等离子体激元通过直接将载流子从金属激发到受体而衰变。不幸的是，尽管据估计直接转移过程比间接转移过程更有效，但对其直接转移过程的原子级理解，尤其是在定量方面，仍然难以捉摸。这是由于在飞秒时间尺度上，直接转移和间接转移同时发生，在将两者分离方面存在实验挑战。在这里，我们采用含时密度泛函理论模拟来分离和研究模型金属 - 受体（Ag - CdSe）界面处的直接转移过程。我们的模拟表明，对于调谐到等离子体激元频率的10 fs高斯激光脉冲，Ag - CdSe系统中形成的等离子体激元在10 fs内衰变，并以约40%的概率诱导直接转移。我们进一步分解直接转移过程，并证明电子和空穴都直接注入受体，即直接热电子转移（DHET）和直接热空穴转移（DHHT），各自发生的概率约为20%。最后，提出了控制和调节DHET和DHHT过程概率的有效策略。我们期望我们的工作为设计能够实现更高效等离子体热载流子器件的金属 - 受体界面提供指导。
