Interview with S Hadjiloucas

11 October 2012

Dr Sillas Hadjiloucas from the University of Reading in the UK talks about the research behind his Letter ‘Femtosecond pulse shaping using differential evolutionary algorithm and wavelet operators'

Why are you interested in femtosecond pulse shaping?

 Much of my and my co-authors’ work concentrates on developing novel instrumentation and measurement solutions, and the use of signal processing and control algorithms for fast pulse spectrometry is a direct extension. Femtosecond (fs) pulses are used for resolving fast physicochemical processes and map potential energy surfaces when molecules interact. Pulse shaping techniques are used to modulate fs pulses in the time or frequency domains. Coherent excitation of a superposition of quantum eigenstates results in the formation of a wave packet whose shape and evolution can be controlled by adjusting the relative amplitude and phase of the superposition states. This way one can control the optical, chemical and electrical properties of matter.

There are exciting prospects in generating tailored light pulses. They can provide additional energy to a chemical reaction at specific resonance frequencies and at precise times facilitating chemical bond formation and breaking. By controlling proton-electron transfer processes they can act as catalysts enabling the formation of desirable products while supressing undesirable ones. Pulse sequences can thus be designed to control the yield of chemical reactions.

What you have presented in your Electronics Letters paper?

By combining evolutionary algorithms with computer-controlled pulse shapers it is possible to optimally excite molecules without the full knowledge of their potential energy surfaces or Hamiltonians. In most cases the potential energy landscape is complex and the goal is to efficiently search the evolutionary landscape of pulse shapes that would lead to the optimal ‘dressed states’ of light without getting stuck in a local minimum. Our Letter describes a methodology that uses efficient evolutionary algorithms to achieve convergence of the population of discretised laser pulse amplitudes and phases.

Sonograms recorded by SPIDER or FROG based techniques are examples of signals represented simultaneously in both the time and frequency domains. Multi-rate dynamics may be efficiently approximated and controlled through parsimonious parametrisations in the wavelet domain. This motivated us to develop the new wavelet operators.

How will this methodology be of benefit

Other application areas that would benefit include optimisation of NMR and ESR experiments, new FLIM and 2-photon microscopy modalities of improved contrast and specificity, efficient super-continuum generation for optical metrology applications, generation of single cycle optical pulses for atto-second spectroscopy at the XUV part of the spectrum, control of charge oscillations in quantum wells, the tailoring of superposition states in Rydberg atoms with applications to quantum computing, adaptive dispersion pre-compensation and channel equalization for UWB spread-spectrum photonics communications and unique opportunities for exploring new pulse management solutions with next generation photonic crystal fibre networks.

By selecting molecules exhibiting both bi-stability and ultrafast optical transitions, dressed light states can be used to perform classical Boolean logical operations implementing (AND) and (OR) gates at fs timescales. This could lead to molecular processors orders of magnitude faster than FPGAs. Optimally shaped pulses are also useful to accelerator based experiments and of relevance to the National Ignition Facility in the US and EU’s Extreme Light Infrastructure.

How do you plan to continue this work?

The current work establishes that wavelet operators can lead to efficient meta-algorithms for quantum control with fast convergence rates. This enables a larger population of amplitudes and phases to be considered, leading to high fidelity pulse shaping. This is necessary as complex molecules have much larger degrees of freedom and complex energy landscapes when interacting.  

In collaboration with Prof. Becerra at Reading, we have developed a state dependent Riccati equation methodology for sub-optimal tracking control of nonlinear systems. He is the developer of the PSOPT optimal control software, which we plan to use for closed-loop quantum control simulations (in collaboration with Prof. Galvão from ITA, Brazil). Such work will complement Prof. Rabitz’s activities at Princeton where nonlinear input-output maps are used to manipulate a system’s Hamiltonian. The software could also provide spatio-temporal synchronisation in neuronal cultures; such work lies within the neuroscience initiative at Reading.

How do you see this field progressing?

With lab-on-a-chip technologies integrating microfluidics with fs systems and pulse shapers there are new opportunities across different research disciplines: Control, Measurement Science, Femtochemomerics, Systems Biology, Synthetic Biology and Optogenetics. If tailored pulses could precisely modulate enzyme-substrate interactions or phosphorylation processes, this would lead to the development of a new discipline: that of Biomolecular Control.

Journal content

Cover of Electronics Letters, volume 50, issue 15

View all content

Browse or search all papers in the latest or past issues of Electronics Letters on the IET Digital Library.