Faster than a Speeding Femtopulse
When the Linac Coherent Light Source (LCLS) begins delivering x-rays in 2009, it will be one of the most powerful x-ray instruments ever built. From this power will come one of the LCLS's most intriguing applications: atomic-scale imaging.
That's not easy. For researchers to produce x-rays with the right characteristics for gathering information on the atomic scalewhere events are measured in quadrillionths of a secondthe LCLS must first produce pulses of electrons that are extremely fast, faster in fact than typical instruments can measure. Regulation of the length and timing of these electron pulses will be crucial to any type of investigation of ultra-fast phenomena. But how do you go about measuring something that fast?
Oscilloscopes are the standard signal measuring devices used to characterize laser pulses and are precise down to around 20 picoseconds. Although this is extremely fast (light travels a mere quarter inch in 20 picoseconds), pulses from the LCLS will last only about 100 femtoseconds (one quadrillionth of a second—the amount of time light takes to travel a hair's breadth). To measure and control phenomena that fast, specialized instruments and techniques will be required. "Around 20 picoseconds is the boundary when you have to start being clever," said SLAC physicist Joe Frisch, who works on pulse length diagnostics for the LCLS.
Scientists at the Deutsches Elektronen-Synchrotron (DESY), in collaboration with SLAC researchers, have refined a type of tool called a transverse deflection cavity that can measure pulses down to 25 femtoseconds. DESY physicists used a spare section from SLAC's original linac to construct the cavity, which measures an ultrafast electron pulse by deflecting its lengthwise orientation as it travels. This deflection happens much the same way as if an arrow, traveling toward a target, were to tilt upward as it flew through the air and strike the target at an angle. By measuring the impression left in the target, one could determine the length of the arrow.
The LCLS will employ transverse deflection cavities to help calibrate pulse length, but there is a downside to this technique. Deflection cavities can accurately measure pulses well within the timescale of the LCLS, but the pulse itself is destroyed when it is measured.
To avoid this, the LCLS will also use what Frisch calls coherent radiation monitors. These devices consist of tiny waveguides that capture a minute burst of radiation emitted by a passing electron pulse and channel it to a detector, leaving the pulse itself intact. Because the radiation signature of a passing pulse of electrons changes as the pulse length changes, LCLS operators will know when pulses of the desired duration are being produced. But coherent radiation monitors cannot measure absolute pulse length by themselvesthey can only indicate when the pulse length changes. These monitors will therefore need constant calibration against measurements from the transverse deflection cavities.
The two-pronged scheme of using deflection cavities to set up the machine and radiation monitors to indicate when those settings need adjustment will provide accurate pulse length measurements, but the approach is still less than ideal. The invasive nature of deflection cavities will require users to temporarily sacrifice access to the beam as the laser pulses are recalibrated. In the future, researchers hope to refine a pulse-length measurement scheme that is accurate and completely non-invasive.
One such promising measurement scheme is called electro-optic pulse length monitoring. With this approach, the electron pulse passes near a special crystal whose transparency changes in the presence of a field of radiation. At the same time as the electron pulse passes near the crystal, an optical laser beam is fired through the crystal. When timed correctly, the optical laser beam then becomes imprinted with the signature of the field of radiation that has grazed the crystal. The optical laser pulse is then processed in such a way so that the dimension of the original electron pulse shows up within the length of the optical laser beam.
"This turns measuring an electron pulse into an optics problem, which is a problem that has already been solved," said Frisch.
But for now, researchers are unsure if the technique can be made to work for pulses faster than 100 femtoseconds.
"We're going to try using electro-optic monitors, but it's not been reliably demonstrated to the right timescale." Frisch said. "The technology is not quite ready yet."
Above image: Joe Frisch holds a miniature waveguide for use on the LCLS coherent radiation monitors. The pinhole-size opening is the same size as the wavelength of radiation emitted by a passing electron pulse. (Click on the image for a larger version.)