Ultra-Fast Science Succeeds at SLAC
By Heather Rock Woods
Pulse Source (SPPS) collaboration has published data from the first
experiments ever using a linear accelerator-based femtosecond x-ray
source, and has developed an important tool for ultra-fast science. SPPS
makes the world’s shortest bunches of electrons in the SLAC linear
accelerator and turns them into very bright pulses of x-ray light one
thousand times shorter than those made in synchrotron rings like SPEAR3.
bunch alters the electro-optic crystal (vertical line) when it
passes by the crystal. Laser light (diagonal line) that shines
through the altered crystal is also affected, recording the
electron bunch length and arrival time.
Image by Adrian Cavalieri (Univ. Michigan)
Pulses of such short duration—lasting some 80 femtoseconds, or 80
millionths of billionths of a second—shine a lightning-fast strobe light
on the swift movements of the microscopic milieu.
“Because of the amazing properties of the x-ray source, we were able to
answer a long-standing problem in condensed matter physics, concerning
how solids transform into liquids on ultra fast time scales,” said Aaron
Lindenberg was the lead author on the SPPS paper published April 15 in
Science. The other paper, headed by Adrian Cavalieri (University of
Michigan), appeared in the March 25 Physical Review Letters; the
researchers used SPPS to develop and test a new timing technique, which
will be essential for many SPPS and Linac Coherent Light Source (LCLS)
experiments. Both papers involved collaborators from multiple
institutions around the world.
“Since SPPS has so many similarities to future free electron lasers like
LCLS, currently being built at SLAC, these experiments lay the
groundwork for the next generation of ultra-fast experiments,”
The new Ultra-fast Science Center at SLAC, Stanford and SSRL will
provide world leadership in ultra-fast research (including experiments
at SPPS and LCLS) and the development of experimental techniques.
In the first SPPS experiment, researchers shone laser light to melt a
room temperature crystal of semiconductor material, and sent x-ray
pulses to probe the material. The scattered x-rays provided a glimpse of
the first step in the transition from solid to liquid. In those first
few hundred femtoseconds between solid and liquid, the atom positions
had on average the crystalline (regular, repeated) structure of a solid,
yet the atoms had moved far from their initial positions, with a
disordered structure like a liquid.
“It’s the first time we’ve been able
to watch the pathways the atoms follow in the first femtoseconds as the
material transitions from solid to liquid,” he said.
Initially, atoms randomly move small distances as they vibrate, but are
kept in position by chemical bonds to other atoms. The laser light
instantaneously broke the bonds, allowing atoms to continue moving in
the random direction they were headed just before the bonds broke. This
takes place before the atoms heat up because the time scale is faster
than the time it takes to transfer energy from the laser to atoms in the
crystal. The result is a very unusual, intermediate state of matter.
Researchers learned the transition state is governed by inertial
dynamics, simply stated by Newton’s First Law as: an object in motion
continues in motion (in the same direction). Understanding the
transition steps of ultra-fast melting may have technical applications,
for example in micro machining and laser eye surgery.
Clocking Femtosecond X-rays
In ultra-fast experiments, timing is everything. The other SPPS
experiment solved a major issue by borrowing ideas from ultra-fast laser
technology. Many SPPS and LCLS experiments will require a laser to pump,
or start, a process in the system under investigation. To put data in
order chronologically—important for seeing chemical or other reactions
over time—researchers need to time-stamp the arrival of the laser pulse
and the arrival of the x-ray pulse that probes, or observes, the system.
light (bottom image) initiates the transition of a material from
solid to liquid while the scattered x-rays provide a glimpse of
the first transitional step. The upper diffraction pattern (top
image) results when no laser light shines on the sample. The
lower diffraction pattern shows the sample at various times
before and after being struck by laser light.
Image by Aaron Lindenberg (SSRL)
Cavalieri and his collaborators used electro-optic sampling to measure
the arrival time of the x-ray pulses in relation to the arrival of the
laser pulses. The laser pulses travel a few feet to the experimental
sample, while the x-ray pulses originate as electron bunches two miles
away at the start of SLAC’s accelerator. And while laser pulses can be
put out in steady, reliable intervals, it’s tricky to perfectly time the
electron/x-ray beam, so there is an intrinsic time jitter, where each
x-ray pulse arrives at a slightly different time relative to the laser
before an electron bunch gets converted into an x-ray pulse, it speeds
past an electro-optic crystal placed next to the beam. The strong
electric field generated by each electron bunch alters the properties of
the crystal, but only at the instant the electrons pass by.
Experimenters then use an ultra-fast laser pulse to probe this change.
The characteristics of the laser light exiting the crystal reveal the
electron bunch length and arrival time, which in turn indicates the
arrival time of the corresponding x-ray pulse.
“The angle of the laser pulse sweeping through the electro-optic crystal
changes space into time. The geometry fixes the sweep rate and the time
window,” Cavalieri said.
To confirm the technique’s reliability, scientists plotted their
electro-optic timing data against the timing data from the SPPS melting
experiment—a rare case where the signal strength allowed data collection
at all time intervals in a single shot—and found good agreement.