Machine learning enhances light-beam performance at the ALS

Successful demonstration of algorithm by Berkeley Lab-UC Berkeley team shows technique could be viable for scientific light sources around the globe.

Synchrotron light sources are powerful facilities that produce light in a variety of “colors,” or wavelengths – from the infrared to X-rays – by accelerating electrons to emit light in controlled beams.
Synchrotrons like the Advanced Light Source at the Department of Energy’s Lawrence Berkeley National Laboratory (Berkeley Lab) allow scientists to explore samples in a variety of ways using this light, in fields ranging from materials science, biology, and chemistry to physics and environmental science. Researchers have found ways to upgrade these machines to produce more intense, focused, and consistent light beams that enable new, and more complex and detailed studies across a broad range of sample types. But some light-beam properties still exhibit fluctuations in performance that present challenges for certain experiments.

Image: This image shows the profile of an electron beam at Berkeley Lab’s Advanced Light Source synchrotron, represented as pixels measured by a charged coupled device (CCD) sensor. When stabilized by a machine-learning algorithm, the beam has a horizontal size dimension of 49 microns (root mean squared) and vertical size dimension of 48 microns (root mean squared). Demanding experiments require that the corresponding light-beam size be stable on time scales ranging from less than seconds to hours to ensure reliable data.
Credit: Lawrence Berkeley National Laboratory

Translation of ‘Hidden’ Information Reveals Chemistry in Action

New method allows on-the-fly analysis of how catalysts change during reactions, providing crucial information for improving performance.

Chemistry is a complex dance of atoms. Subtle shifts in position and shuffles of electrons break and remake chemical bonds as participants change partners. Catalysts are like molecular matchmakers that make it easier for sometimes-reluctant partners to interact.

Now scientists have a way to capture the details of chemistry choreography as it happens. The method—which relies on computers that have learned to recognize hidden signs of the steps—should help them improve the performance of catalysts to drive reactions toward desired products faster.

The method—developed by an interdisciplinary team of chemists, computational scientists, and physicists at the U.S. Department of Energy’s Brookhaven National Laboratory and Stony Brook University—is described in a new paper published in the Journal of Physical Chemistry Letters. The paper demonstrates how the team used neural networks and machine learning to teach computers to decode previously inaccessible information from x-ray data, and then used that data to decipher 3D nanoscale structures.

RAW Power!

MacCHESS software brings synchrotron-level data processing to the laptop and home laboratory

Since its introduction by Søren Skou (Nielsen) in 2010, the BioXTAS RAW software has been a familiar interface to the many biomedical scientists collecting data at CHESS beamlines in recent years. From the start, RAW was designed specifically with novice users in mind: when scientists arrive at the beamline, they need something fast and easy to learn in the very limited time available … often late at night.

The program was literally designed by looking over the shoulders of beamline users as they collected data. But rather than simply create an automated data processing pipeline, we opted to give people the power to fully process data on their own computers at home, if they choose. This allows them to use the same software at other beamlines and even on their own home X-ray sources: from initial raw data reduction to final publication. Indeed, with over 4000 downloads in 2017, RAW is now the primary processing software at several other beamlines and lab source facilities worldwide.

>Read More

Picture: Richard Gillilan, Jesse Hopkins, and Soren Skou at the annual Amrican Crystallographic Association meeting where they conducted a tutorial in the BioXTAS RAW software.

Emergent magnetism at transition-metal-nanocarbon interfaces

Researchers have shed light on the origin of the magnetism arising at carbon/non-magnetic 3d,5d metal interfaces

These results may allow the manipulation of spin ordering at metallic surfaces using electro-optical signals, with potential applications in computing, sensors, and other multifunctional magnetic devices.

Interfaces are key in solid state and quantum physics, controlling many fundamental properties and enabling emergent interfacial, bi-dimensional like phenomena. Therefore they offer potential opportunities for designing hybrid materials that profit from promising combinatory effects.

In particular, the fine-tuning of spin polarization at metallo–organic interfaces opens a realm of possibilities, from the direct applications in molecular spintronics and thin-film magnetism to biomedical imaging or quantum computing. This interaction at the interface can control the spin polarization in magnetic field sensors, generate magnetization spin-filtering effects in non-magnetic electrodes or even give rise to magnetic ordering when non-magnetic elements such as diamagnetic copper or paramagnetic manganese are put in contact with carbon/fullerenes at such interfaces.

 

>Read More