SLAC Today logo

Fourth LCLS Instrument Captures Its First X-ray Image

(Photo - detector team members)
SLAC contributors in the detector project stand in the LCLS XPP hutch just after the detector was installed there. From left: Jeff Tice, Brian Duda, Jack Pines, Lupe Salgado, Miguel Pinillos, Matt Weaver, Yolanda Casas, Philip Hart, Chris Kenney, Gunther Haller, John Morse, Ryan Herbst, Tom Nieland and Leo Manger. Not pictured: Matt Swift, Garth Williams, and Martin Nordby, and the team's collaborators from Cornell. (Photo by Brad Plummer.)

The Coherent X-ray Imaging instrument at the Linac Coherent Light Source achieved the first X-ray image from its newly installed detector last Tuesday. The detector was installed inside the instrumentís vacuum-sealed experimental chamber in January. Through this week, CXI instrument scientists will test and adjust the new device in preparation for arrival of the first experimental users this Sunday.

The CXI hutch is the fourth end station to receive X-rays since LCLS turned on in 2009. Researchers hope to use the instrument to turn patterns of X-rays scattered off experimental samples into faithful images of single molecules. The detector, which will collect these scattered photons, is a key component in the setup. Because the LCLS X-ray pulses are unique in the world, the detector needed to have the sensitivity, dynamic range and data processing speed to match.

"This detector can process about half a gigabyte of data per second," said software engineer Matt Weaver, who wrote analysis software that will allow researchers to see live displays of whatís going on inside the experimental chamber. "This is about 4000 times as fast as your typical Internet connection."

A team of scientists, engineers and technicians from across the lab began working on the detector design several years ago when LCLS was still under construction. At the time, the CXI instrument existed only on paper, "so we just had a very general idea what the requirements for the detector were," said mechanical engineer Brian Duda, who worked on the mechanical design.

SLAC contracted with a detector development lab led by Sol Gruner at Cornell University to design and build a prototype X-ray detector. The Cornell group concentrated on the X-ray sensor chips and front-end electronics, culminating with a prototype single-chip detector. A group of SLAC engineers, mainly from the Particle Physics and Astrophysics directorate, took responsibility for building a large, multi-chip detector using the Cornell chips. They designed the data-management electronics and mechanical housing for a detector optimized for the CXI instrument. SLAC's Manufacturing Department stepped in to construct the detectorís complex steel skeleton.

"With so many parts, it was imperative that the detector matched our model perfectly," Duda said. "The Manufacturing Department did a great job with the precision cuts that the detector required."

Step inside the CXI instrument hutch in this slideshow of the detector installation. You can also download the video as a Quicktime file (48 MB). (Video by Brad Plummer. Time: 2:47 minutes).

The circular detector is about a foot across, with an array of 32 silicon sensors divided among four quadrants. The quadrants can move relative to each other, creating an adjustable opening in the middle of the detector to let unscattered photons from LCLS's intense X-ray beam pass through. Inside the detector, X-rays that are scattered by the sample are collected in the silicon sensors.

The entire device has about 2.3 million pixels, each with its own data processing circuitry. Each pixel is sensitive enough to detect a single X-ray photon, yet capable of recording thousands of photons at once. A custom integrated circuit chip converts the charge signals from the incoming photons into digital signals, which are rapidly transferred to data storage. The detector can download data, reset and be ready for a new picture in just a few milliseconds, fast enough to keep up with LCLS's 120 X-ray pulses per second. Once the data is stored, software takes over, interpreting the signals to generate a picture of the experimental sample.

"This detector is unique not only because of its speed and X-ray sensitivity, but because itís made up of 32 different sensor chips," Weaver said. While other detectors have just one rectangular sensor to absorb photons, the CXI detector is effectively 32 parallel cameras. "Aligning all the chips is important for reconstructing the data. Our software knows the positions and spaces between the chips and can stitch them all together to make one image."

Although the detector was originally contracted for the CXI instrument, LCLS administrators asked whether it could be modified to work on the X-ray Pump Probe instrument. The third of LCLS's six planned experimental instruments, XPP was scheduled to come online for its first user experiments last fall, several months before the CXI installation.

Time lapse video shows the CXI instrument coming together over a period of about nine weeks. You can also download the video as a Quicktime file (50 MB). (Video by Brad Plummer. Time: 1:12).

"They needed the detector ahead of schedule [in order to use it] for the XPP," said detector physicist Chris Kenney. "But we finished it, and the detector was installed in the experimental hutch in time for commissioning. This was a great example of one lab working together."

After the success with XPP, the team had only a few months to complete a second unit for the CXI instrument. The XPP detector, which operates in ambient air, had a simplified mechanical design compared to the CXI requirements. The CXI detector would operate under vacuum, making it difficult to remove heat produced by the detector's fast electronics.

The team needed to find a way to cool the detector while maintaining freedom of movement for its quadrants. Since the heat can't be drawn away from the CXI detector by convection through air, as it can in the XPP hutch, the team decided to remove heat by running cooled water across flexible copper foils attached to the detector. The flexible straps allow each quadrant of the detector to move about six millimeters, creating a maximum aperture of 10 mm at the center of the detector.

The team has now turned their focus to testing the instrument and its electronics, with the goal of being ready for first user experiments next week. Over the next several months, working with users from around the world, the CXI detector will continue to refine and enhance the machine's capabilities, with the aim of adding an additional detector layer for enhanced image capture as early as this summer.

óLauren Rugani
SLAC Today, February 7, 2011