Carefully-shaped femtosecond pulses of light can now enhance the
yield of chemical reactions and synthesize novel molecules. Gustav
Gerber, one of the pioneers of laser chemistry, talks to Rob van den
Berg about his work.

Coherent control
For hundreds of years chemists have been heating, stirring and
pressurizing their reaction mixtures to encourage molecules to
undergo specific changes. Unfortunately, such "shake and bake"
techniques are not always particularly selective or efficient,
especially when it comes to the synthesis of complicated molecules
which may require many processing steps.

Recent optical research by chemists and physicists around the world
looks set to change that situation. For several years, scientists
have been quietly developing ways to directly manipulate the
electronic structure of the molecule using laser pulses, forcing
individual bonds to break and reattach.

The ultimate aim is to establish optical synthesis techniques that
can deliver a higher yield with less unwanted by-products, or produce
molecules that are hard to make by other means. The idea is that
laser pulses aligned to the exact frequency of a chemical bond can
deliver just enough energy to cause selective breakage of bonds.

In initial experiments, unfortunately, this selectivity was lost due
to the rapid redistribution of energy within the molecule. As a
result, laser excitation became simply an expensive method of heating
up the entire molecule. However, the use of femtosecond Ti:sapphire
lasers, pulse-shaping devices and sophisticated feedback algorithms
now looks set to change the situation.

Coherent control

Today's state-of-the-art approach to laser chemistry is called
coherent control. It involves making an initial guess at the shape of
the optical pulse that will excite the molecules and start the
reaction. The pulse is generated and the reaction products that form
as a result are detected and analysed by a learning algorithm. It
then adjusts the pulse's duration, phase and amplitude to increase
the yield of the reaction. After several iterations the best pulse
shape for the reaction is found. In a typical experiment, it takes a
matter of minutes to test thousands of different laser pulses and
find the most suitable pulse parameters.


Compounds
Gustav Gerber, a professor of experimental physics at the University
of Würzburg, Germany, is one of the technique's pioneers. Gerber
thinks that the technique offers many possibilities - creating novel
stable or metastable molecules, obtaining better image-quality in
microscopy and improving optical data transfer, for instance. "This
scheme was first applied to increase the fluorescence of a large dye
molecule, but already, one year later, we have used it to optimize
the outcome of a dissociation reaction," Gerber said.

In the coherent control method, two diffraction gratings and a liquid-
crystal multi-channel modulator (LCM) are used to adjust the shape of
the optical pulse. The first grating separates the individual
wavelengths within the pulse, while the multi-channel modulator gives
each wavelength a specific phase delay. Finally, the wavelengths are
recombined into a single pulse by the second grating. The process not
only reshapes the pulse but also reorders the wavelengths within it
so that they strike molecules in a predetermined sequence.

But because the LCM has 128 channels that can each be set at 4000
different levels, an astronomical number of different laser pulse
shapes can be generated. To find the optimal shape, an evolutionary
algorithm is used. Each pulse is labelled by the setting of the
individual modulator channels. This setting works as a kind of
genetic code and forms the basis for a process similar to natural
selection.


Schematic
Using a random setting of the individual modulator channels, the
chemical reaction in question is initiated, and its products are
detected. The reaction is then performed again using slightly
different settings. A computer analyses the results and determines
which are the most "successful" pulses - that is, which have
generated the largest quantity of the desired product. Only the most
successful pulses are used to generate offspring, becoming
the "parents" for the next generation of pulses. After a surprisingly
small number of generations, the optimal pulse shape is reached.

The system acts rather like a very efficient analogue computer,
solving the problem depending on the conditions set by the
experimenter. "Basically, it is the system itself that determines
what it needs at each instant of time," said Gerber."It is as if the
molecule effectively solves the Schrödinger equation, and determines
at what time it needs what frequencies."

On the experimental side, in the last four years Gerber's group has
discovered numerous examples of reactions in which the product yield
may be enhanced using the coherent control technique.


Pulse shaping
"For instance, in the molecule CH2ClBr it is very difficult to break
only the C-Cl bond and leave the C-Br bond intact," said Gerber. "Our
method proved to be four times better than conventional synthetic
chemistry." His group has obtained a similar enhancement in the
removal of a carbonyl ligand (CO) from organometallic complexes like
Fe(CO)5.

"In the pharmaceutical industry synthetic chemists use acid groups to
protect certain parts of a molecule during reactions," said
Gerber. "After that you have to get rid of these acid groups, but
unfortunately, other molecular groups often break off at the same
time and are lost."

Gerber has succeeded in decreasing this unwanted loss by a factor of
five. He was also the first to make the method work in liquids by
selectively exciting complex dye molecules that have overlapping
absorption spectra. It had been thought that the complex interactions
between the excited molecule and surrounding solvent molecules would
cause problems, but this has not turned out to be the case.

Theoretical success

Researchers now want to take a step further, and find out what it is
that determines the optimal laser field (pulse shape). For relatively
simple processes, like the ionization of a calcium atom, Gerber
succeeded in finding out why the laser pulse came out as it did. But
he admitted: "If we want to try to understand what the molecules are
telling us, we need the help of theoreticians."

A group from the Free University of Berlin led by Ludger Wöste and
Jörn Manz has recently managed to decipher the reaction dynamics
underlying the optimal laser field as determined by coherent
control. "With ab initio quantum calculations and simulations of wave
packet dynamics we were able to decode the optimal femtosecond pulse
generated by adaptive learning techniques," explained Wöste. "That
was our main goal: to understand exactly what the light pulse does to
the molecule. The first part of the pulse is for excitation and the
second for ionization."


Pulse shaping (2)
Gerber also explains that a common criticism of laser chemistry -
that it is a technique that will never be suitable for generating
macroscopic quantities of a product - is also no longer an issue. His
approach is as follows:

"We first determine the optimal laser field with a 1 kHz laser in the
gas phase, but then copy this on a different system with a much
higher repetition-rate and perform the same experiment in the liquid
phase at a much higher density," said Gerber. "In this way we are
able to produce the same microgram to milligram quantities in 24
hours as 'classic' synthesis of a complicated substance, which
requires many separate steps. And the operating costs of a laser are
much lower."

Gerber and his team are now looking into using LCMs to change the
polarization state of each laser pulse. In this way they hope to be
able to address the 3D properties of molecules and pull a chemical
bond in a specific direction. Ultimately, this may enable synthesis
of either the left or right-handed form of molecules - a capability
that can be crucial for making pharmaceutical compounds, where often
only one of the two forms is biologically active.

Ultimately, the pulse-shaping techniques developed for laser
chemistry could also have important applications in other fields. In
telecommunications, the shape of a light pulse travelling along an
optical fibre tends to distort over long distances. The effect, known
as dispersion, might be preventable by pre-compensating the light
pulses with a pulse shaper developed for laser chemistry.

In biological imaging, multi-photon laser microscopes use femtosecond
pulses to generate 3D images of a sample. The images from deeper
layers tend to be poor-quality, because the laser light is distorted
as it passes through the upper layers.

"With a pre-compensated pulse, we might be able to correct for this,
increasing the resolution and contrast ratio of the images," said
Gerber. He sees similar benefits in the fields of materials
micromachining with lasers and laser surgery.

About the author

Rob van den Berg is a freelance science and technology journalist
based in the Netherlands.