Heloise Chomet,Samuel Plesnik,Constantin Nicolae,Jack Dunham,Lesley Gover,Timothy Weaving,Carla Figueira de Morisson Faria
Abstract:We study non-classical pathways and quantum interference in enhanced ionisation of diatomic molecules in strong laser fields using machine learning techniques. Quantum interference provides a bridge, which facilitates intramolecular population transfer. Its frequency is higher than that of the field, intrinsic to the system and depends on several factors, for instance the state of the initial wavepacket or the internuclear separation. Using dimensionality reduction techniques, namely t-distributed stochastic neighbour embedding (t-SNE) and principal component analysis (PCA), we investigate the effect of multiple parameters at once and find optimal conditions for enhanced ionisation in static fields, and controlled ionisation release for two-colour driving fields. This controlled ionisation manifests itself as a step-like behaviour in the time-dependent autocorrelation function. We explain the features encountered with phase-space arguments, and also establish a hierarchy of parameters for controlling ionisation via phase-space Wigner quasiprobability flows, such as specific coherent superpositions of states, electron localisation and internuclear-distance ranges.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to control the non - classical paths and quantum interference phenomena in the enhanced ionization process of diatomic molecules in a strong laser field through machine - learning techniques. Specifically, the researchers use machine - learning methods, such as t - Distributed Stochastic Neighbor Embedding (t - SNE) and Principal Component Analysis (PCA), to explore and optimize the influence of multiple parameters on enhanced ionization. These parameters include the localization degree of the initial wave packet, the relative phase, the internuclear distance within the molecule, and the intensity and frequency of the external electric field, etc. Through these methods, the researchers hope to find the optimal conditions to achieve enhanced ionization in a static field and controlled ionization release in a two - color driving field.
### Research Background and Objectives
1. **Research Background**:
- Enhanced ionization is a phenomenon observed in stretched molecules, and its ionization rate is at least one order of magnitude higher than that of atoms with similar ionization potentials. This phenomenon is considered to be caused by the narrowing of the upper - field potential barrier due to the adjacent lower - field potential wells, as well as the coupled charge - enhanced resonance states.
- Quantum interference provides a "bridge" that promotes electron transfer within the molecule. Its frequency is higher than that of the applied field and depends on the intrinsic properties of the system, such as the state of the initial wave packet or the internuclear distance.
2. **Research Objectives**:
- **Understand the Mechanism**: Through phase - space analysis and Wigner quasi - probability distribution, deeply understand the roles of quantum interference and non - classical paths in enhanced ionization.
- **Optimize Control**: Use machine - learning techniques to explore the influence of multiple parameters on enhanced ionization, find the optimal conditions, and achieve precise control of the enhanced ionization process.
- **Practical Applications**: The research results are helpful for developing new experimental techniques and theoretical models in the attosecond science field, such as controlling ultrafast molecular dissociation, creating quantum switches, etc.
### Main Research Contents
1. **Model Establishment**:
- Use a one - dimensional single - electron molecular model to describe the dynamics of the electron wave function by solving the time - dependent Schrödinger equation.
- The molecular binding potential consists of two asymmetric potential wells, and the shape and depth of the potential wells can be changed by adjusting parameters such as the internuclear distance and nuclear charge.
2. **Key Parameters**:
- **Initial Wave Packet**: By adjusting the localization degree, width, and relative phase of the initial wave packet, study the influence of different initial conditions on enhanced ionization.
- **Molecular Parameters**: Including the internuclear distance, the weights of the left and right potential wells, etc. These parameters directly affect the electron dynamics within the molecule.
- **External Electric Field**: It can be a static field or a time - dependent two - color driving field. Control the ionization process by adjusting the intensity and frequency of the electric field.
3. **Data Analysis**:
- **Ionization Rate**: Defined as the reduction in the probability density of the wave function from \(t = 0\) to \(t = T\).
- **Autocorrelation Function**: Used to evaluate the dynamic behavior of the system in a time - dependent electric field, showing step - like, constant, or other shapes.
- **Wigner Quasi - probability Distribution**: Used to analyze the dynamic behavior in phase space, especially the influence of quantum interference and non - classical paths.
### Research Methods
1. **Dimensionality Reduction Techniques**:
- **t - SNE**: Project high - dimensional data onto a two - dimensional space to reveal the overall trends and patterns.
- **PCA**: Used to verify the accuracy of t - SNE results and retain global characteristics.
2. **Data Generation and Processing**:
- Randomly generate data points of multiple parameters, and obtain the time - evolution information of the electron wave function by solving the time - dependent Schrödinger equation.
- Use the shape of the ionization rate or autocorrelation function to characterize each data point.
### Results and Discussions
1. **Results in Static Field**:
- The ionization rate has a linear relationship with the localization degree of the initial wave packet, and the wave packet localized at the center of the upper field has the highest ionization rate.
- By adjusting molecular parameters (such as the internuclear distance and nuclear charge), the ionization rate can be significantly affected, especially in heteronuclear molecules.
2. **Results in Time - Dependent Field**:
- The shape of the autocorrelation function can be divided into step - like, constant, and other types, reflecting different ionization behaviors.
- Through t - SNE and PCA analysis, the key parameter combinations that affect the shape of the autocorrelation function are found, providing a theoretical basis for achieving controlled ionization.