Attosecond Strong-Field Physics allows us to investigate electron wave packet dynamics both during the strong-field ionization process and during the subsequent propagation of the photoelectron. Ionized photoelectrons can take part in many different phenomena, such as creating highly excited Rydberg states, or contributing to High Harmonic Generation.
This fundamental research helps to understand ultrafast charge transport in a variety of process, for example during chemical reactions or in photovoltaic cells.
Recently, I got interested in bound state dynamics of electron wave packets under the influence of strong driving fields, between the ground and excited states and light-induced features.

Strong-Field Tunnel Ionization

Photoelectron Momentum Distribution Analysis

In strong-field tunnel ionization, the photoelectron momentum distribution recorded on a detector reflects the geometry and subcycle dynamic of the strong field. This is because

• the ionization probability depends exponentially on the field amplitude,
• and the final momentum of an electron essentially reflects the vector potential of the laser field at the moment of ionization (ignoring corrections and perturbations)
  

In consequence, elliptical polarization of the ionizing field leads to an elliptical geometry of the photoelectron momentum distribution. Therefore, an analysis restricted to cartesian or spherial symmetries will in principle introduce spurious features which are not related to the physics of the process itself.

An analysis matched to the correct ellipticity, on the other hand, yields nicely behaved physical observables, such as Gaussian distributions for longitudinal momentum spreads
.

Relevant publication: [03]

Initial longitudinal momentum

In most well-established tunnel ionization theories, the momentum of the photoelectron parallel to the electric field immediately after the tunnelling process, when it exits the potential barrier and enters the continuum, is assumed to be negligibly small.

However, more recently this assumption has been questioned, or calculations predicted even a finite most probable longitudinal momentum outwards (see Li2016, Camus2017 for example).

Comparing classical trajectory monte carlo simulations (CTMC) to experimental data revealed that an initial longitudinal momentum spread in the calculation leads to a better match with the measurement.

Relevant publications: [03], [05], [15]

Non-adiabatic effects

Non-adiabatic effects refer to effects influencing the tunnel ionization process due to the time-dependent nature of the acting force on the system. Strong-field ionization is analytically much easier to treat when one can approximate the electric field during the relevant time period as quasi-static. However, most typical experiments are performed in a regime where, at least mathematically, this limit is not satisfied.

Unfortunately, it is still a challenge to clearly identify non-adiabatic effects in experimental measurements, because doing so in most cases requires

• knowledge of the absolute field strength,
• calibration of this field strength by matching the measurement with predictions from either adiabatic or non-adiabatic theories.

Despite that, the importance of accounting for non-adiabatic effects is more and more being recognized. 
In a recent publication, we have shown that Rydberg states offer an alternative approach to identifying non-adiabatic effects without this same calibration caveat.

Relevant publications: [05], [06], [08], [10], [12,13], [15]

Attoclock Revisited

The question of how long does it take for a quantum particle to tunnel through a potential barrier has been around since the advent of quantum mechanics. With the development of ultrashort strong-field laser pulses, we now have an option to experimentally test various theoretical predictions.

The attoclock is a method to do just that, by exploiting a rotating electric field. Therefore, the moment of ionization is, in principle, encoded in the angle of the final electron momentum. 
However, there are many challenges in interpreting and analysing the experimental results. And similarly, theoretical predictions often use different sets of approximations and assumptions, leading to difficulties comparing their values. 

Nevertheless, we have found that the helium attoclock measurement is in agreement with finite tunnelling time models.

Relevant publications: [14], [15], [18]
Highlight article on the NCCR MUST website.

Post-Ionization Propagation

Rydberg state creation

Photoelectrons which have tunnelled through the potential barrier are subsequently propagating in the superposed field of the laser and the parent ion. However, there is a chance that an electron will not gain enough kinetic energy by the end of the pulse to completely escape. Instead, it is recaptured into an excited bound state.

The efficiency of this so-called Frustrated Tunnel Ionization [Nubbemeyer2008] depends on various factors, such as

• ellipticity of the laser field,
• intensity,
• and pulse duration.

Relevant publications: [2], [12,13], [17]

High Harmonic Generation: Long Trajectory Suppression & Symmetry breaking

Photoelectrons returning back to the vicinity of their parent ion can be involved in a number of different processes. For example:
gain energy and produce the high-energy tail of the photoelectron momentum distribution on the detector, kick a second electron out of its bound state, …
Or it can recombine into the ground state, and the energy difference is radiated off as a high harmonic photon.

This highly nonlinear upconversion of driving IR photons into sub-femtosecond pulses of XUV photons is the bases for most experiments with attosecond temporal resolution, and table-top XUV sources.
Through phase matching conditions during the propagation of the beams through the entire generation target, long trajectories are suppressed and contributions from short trajectories dominate the resulting XUV spectrum.

Combining two-colour counter-rotating circularly polarised driving field with a spatial inhomogeneity modifies these simple recombining photoelectron trajectories significantly. The combination of these two effects offers detailed sub-cycle control of the electron dynamics, and therefore also allows the modification of the polarisation of individual attosecond pulses in a pulse train.

Relevant publication: [11], [16]

Publications in Preparation

List of Publications