Last week technicians at Argonne National Laboratory began to disassemble a particle accelerator known as the Advanced Photon Source (APS), a ring 1.1 kilometers around that since 1995 has shone as one of the world’s brightest sources of x-rays. It’s hardly the end for the facility, which annually serves nearly 6000 scientists from myriad fields. Within a year, workers will replace the electron accelerator with a new one that will boost the intensity of the APS’s output x-ray beams by a factor of 500. A major scientific facility will be rejuvenated. That’s not unusual.

For me personally, however, the dismantling of the original APS evokes strong emotions. My father, Yanglai Cho, was an accelerator physicist who spent his entire career at Argonne, a Department of Energy (DOE) laboratory outside of Chicago. Forty years ago, he led the small team that hammered out the conceptual design for the machine. In my mind, it was his baby. When dad died in 2015 at age 82—4 years after a devastating stroke—I took comfort in the thought that he lived on in that accelerator. Now, it, too, will be gone.

I was a teenager when, in the early 1980s, my dad started musing about the accelerator. I loved him dearly, but, as many people do, I had a complicated relationship with my father. He could be tyrannical and demanding, self-centered and remote. “I don’t care what you do just as long as you’re the best at it,” he would pronounce to me or one of my two brothers and then leave us to flounder on our own. Back then, the APS was this mysterious thing that occupied his time and his mind.

I did follow my father into physics, eventually grinding out a Ph.D. However, my path led me into science journalism. Over the past 20 years, I have written about many big scientific facilities, ranging from atom smashers and gravitational-wave detectors to x-ray lasers and neutron sources. I have never built anything, but I have learned a few things about what it takes to create these often-astounding machines. And that has helped me better understand my father.

“He was a superb and visionary accelerator physicist, and he transformed many large machines at Argonne and elsewhere,” says one former DOE official who still consults for the agency and therefore asked not to be named. “He was also a wonderful colleague and teacher.” Having locked horns with my father so many times, I marvel at that last assessment. Yet, thinking about his work, I’ve come to appreciate how a South Korean immigrant with a thick accent and a fiery temper could flourish in an unusual and demanding field.

A revolutionary tool

Like the other 70 x-ray synchrotrons around the world, the APS turns what was a nuisance into a powerful resource for studying materials. It accelerates electrons within a long vacuum tube to high energy and near–light-speed, while magnets steer them around the ring. The circulating electron beam radiates x-rays, just as a wet washcloth twirled overhead flings droplets of water. That synchrotron radiation saps the electrons’ energy, so when accelerators were built just for experiments in particle physics, it was an unavoidable waste.

In the 1960s, scientists began to siphon the x-ray radiation from electron accelerators to study materials by, say, measuring their absorption spectra. The first major dedicated sources emerged in the following decade. The APS led a wave of larger, higher energy third-generation sources, along with the European Synchrotron Radiation Facility (ESRF) in Grenoble, France, and the SPring-8 facility in Hyogo, Japan. Compared with previous sources, the Argonne machine produced more-compact electron beams that generated far more intense x-rays. It also pushed into the regime of hard x-rays, those with wavelengths shorter than 0.1 nanometers, which are ideal for probing a material’s atomic-scale structure. It replenished its electrons not every 12 hours, but every 30 seconds, keeping the intensity of the x-ray beams rock steady.

Most practically, the APS helped revolutionize the reliability of x-ray sources, says David Moncton, a physicist at the Massachusetts Institute of Technology who was Argonne’s associate lab director for the APS from 1987 to 2001. Earlier, more persnickety machines would operate between 50% and 75% of the available time, vexing officials trying to schedule a facility’s users. The APS pushed that reliability factor to 99%, Moncton says. “If you just buy equipment, put it together, and cross your fingers, you will not wind up with a machine that performs 99% of the time.”

Such attributes have made the APS a font of discovery. Perhaps most strikingly, it and other third-generation x-ray synchrotrons have revolutionized the study of the structure and function of proteins and other biomolecules, says Helen Berman, a structural biologist at Rutgers University and a co-founder of the global Protein Data Bank (PDB). Before probing molecules with x-rays, structural biologists must crystallize them, an arduous task. Berman says the APS and other third-generation sources provided “the ability to take data with a very intense x-ray source and use much smaller crystals.”

Of the 201,000 protein structures in the PDB, 72% come from x-ray synchrotrons. Of those, 30,466 come from the APS—51% of the yield from U.S. synchrotrons. Data from the APS helped win two Nobel Prizes in Chemistry—in 2009, for studies of the function and structure of the ribosome, the cell’s proteinmaking machinery; and in 2012, for studies of cell membrane proteins called G protein-coupled receptors. The APS helped determine the structure of the SARS-CoV-2 virus, which causes COVID-19, and develop Paxlovid, a drug used to treat it.

The APS supports many other types of work, as I saw last month when I walked around its expansive, tunnel-like experimental hall. Within the facility’s 68 experimental end stations, scientists are analyzing the quantum properties of magnetic materials, developing biologically inspired adhesives, and even studying how the atomic-scale structure of lead-acid batteries changes as they run, a study made possible by the intensity of the APS’s x-rays.

A vision and a quest

All of this was a gleam in scientists’ eyes when my dad started to think about the accelerator—which he wanted to name Phoebus, for the Greek god of the Sun. In 1983, he was helping fix a troubled smaller x-ray synchrotron called Aladdin at the University of Wisconsin-Madison when a review panel released a report arguing for a larger hard x-ray source. Sitting in the Aladdin control room, my dad read the report and then dashed back to Argonne to urge lab officials to fund R&D on the machine and to push for Argonne to host it, Moncton says.

The lab badly needed such a project. It had once had a thriving particle physics program, which is what had attracted my father. But in 1979, Argonne shut down its proton accelerator, which had been superseded by a much bigger, new one at Fermi National Accelerator Laboratory 50 kilometers away. “The lab was struggling for a mission,” Moncton says. “Yang immediately thought that this made a good potential project and was of a size to carry the lab into its future.”

The project also gave my father something he needed personally. Like most of us, he was a jumble of mismatched puzzle pieces. He could be tetchy one moment, and ridiculously overindulgent the next. My parents had divorced when I was young, yet he was a constant presence, letting himself into our mother’s house as he pleased. He had contracted polio as a child and had a withered leg. Nevertheless, he liked to take us bowling, even if he would sometimes fall. He loved to go out for lunch and, oddly, liked John Wayne movies. But, overall, after the divorce he seemed unhappy.

The clubby, intense effort of designing the new machine revitalized him. The team consisted of my father; Gopal Shenoy, a material scientist at Argonne who died in 2017; and a dozen others. On a choice table in the Argonne cafeteria, my father posted a sign, “Reserved for APS staff”—and replaced it as cafeteria workers repeatedly removed it. In 1985, the Chicago Bears football team stormed to a championship. dad brought in a TV so researchers could keep tabs on the games while working on Sundays.

It took the team 3 years to complete the conceptual design. Exactly what my father did remains a bit of a mystery to me. As an accelerator physicist, he understood how electrons surf radio waves to gain energy, magnetic fields focus those particles, resonances can obliterate a beam, and synchrotron radiation itself kicks the electrons around. But he had to turn that knowledge into a workable design. His team specified the myriad parameters that defined the APS: the beam energy, the radius of the ring, the number of bunches of electrons in the beam, the arrangements of the magnets, the frequency of the radio waves, etc.

The first goal was to make the most compact electron beam possible, which would then radiate the brightest x-ray beams, says John Galayda, an accelerator physicist who worked on the APS. The beam also could not move, he says. A tiny electron beam radiates a tiny x-ray beam, which can probe minuscule samples—provided it consistently hits the target. Finally, the machine had to run as reliably as possible.

Machine designers must strike a delicate balance. The design cannot be so ambitious that the machine can’t be built. But it can’t be so cautious that it merely replicates what already exists. So, a design invariably contains elements that builders do not yet know how to make. “Every facility that I’ve been involved with—that’s been lots—was one of a kind, first of a kind,” the former DOE official says. “And that means there are enormous technical problems that haven’t yet been resolved.”

Apparently, my father was good at identifying what, with effort, could be achieved. “He would look at what other projects did and use it and make it better,” says Marion White, an accelerator physicist at Argonne and my father’s widow. “He was incredibly good at that.”

Of course, a project leader must also manage people. And that’s where my father struggled. His autocratic style worked early on, when the project staff consisted of a self-selected few. It became less effective as the effort became more formal and ballooned to hundreds of people. “He’d hold a meeting and afterwards I’d have people coming into my office and saying, ‘I can’t take it anymore,’” Moncton recalls. So in 1991, as construction ramped up, a physicist named Ed Temple replaced my father as project director.

My father remained deeply involved in the project, however. He chaired the committee that had to approve any modification to the final design. “He’d be pretty rough about that,” Galayda says. “I think he viewed it as an adversarial process.” As with any machine, some changes, more or less painful, had to be made. Nevertheless, the APS came in on budget at $467 million and ahead of schedule.

To me, it seemed that my father had his fingers in nearly every aspect of the facility. For example, the APS’s 90,000-square-meter concrete floor has no expansion seams. Contractors had urged including them to keep the floor from cracking, Moncton says, but the design team insisted that the stability of the floor was more important than superficial cracks. I remember my father talking the finer points of concrete floors over lunch.

The optimist

Now, workers are dismantling dad’s machine to replace it with a “fourth-generation” design. The new machine will double the current in the ring to 200 milliamps. More important, its electron beam will be even more compact, says Jim Kerby, a mechanical engineer at Argonne and director of the $815 million project. The original APS’s beam measured 10 micrometers high and 275 micrometers wide. The new APS’s beam will measure 3 micrometers high and 15 micrometers wide—less than the width of a human hair.

That subtle shrinking depends on a key difference between the two machines, Kerby says. In the old APS, the beam always bent inward, to the right. In the new design, it will occasionally bend outward, to the left. These kinks give rise to new dynamics that shrink the beam—an approach pioneered at the MAX IV facility in Sweden and deployed in a rebuild of the ESRF completed in 2020.

The scheme requires an almost entirely new accelerator. Workers will replace the original APS magnets with 1321 new ones, and change the entire vacuum system. “We are swapping out essentially the whole ring,” Kerby says. The transformation will take just 1 year. “It’s always been a deliverable of the project that the downtime would be as short as humanly possible,” Kerby says. By then my father’s machine will be a memory.

But my father himself was thinking of new machines even before the APS turned on. In the late 1990s, Oak Ridge National Laboratory began building the Spallation Neutron Source (SNS), which slams a proton beam into a mercury target to generate neutrons for studying materials. The project struggled and DOE nearly canceled it, says Thom Mason, director of Los Alamos National Laboratory, who was SNS project director from 2001 to 2008. Moncton, White, my father, and others went to Oak Ridge to help.

My father led the team that made a key design change, Mason says. The original plan for the SNS called for a conventional linear accelerator made of copper accelerating cavities. The team switched to a novel design with cavities made of a superconducting metal, which promised to be more energy efficient, reliable, and flexible. “As a result, we wound up building the first superconducting proton accelerator instead of the last normal one,” Mason says.

My father consulted on accelerator projects in South Korea, Germany, Japan, and elsewhere, finding his niche in the odd community of scientific machine builders. He had grown up dirt poor in what is now South Korea, then occupied by Japan. He came to the United States when he was 24 and didn’t return home for 17 years. Whether because of cultural differences, his disability, or his temperament, often he was an outsider.

Not when he was among his colleagues, however. The happiest I ever saw my father was when he was playing with his grandchildren. A close second was when he was hobnobbing with his colleagues. At least some of them enjoyed his company, too. “To me, working with your father was a wonderful experience,” says Giorgio Margaritondo, a physicist at EPFL, formerly the Swiss Federal Institute of Technology Lausanne, who teamed up with him on Aladdin.

In fact, my father managed to find a community in which he could succeed not in spite of his prickly personality, but, to some degree, because of it. “To build an accelerator is a very complex task with many subtasks and a lot of coordination and so on, so you need to run the thing almost the military way,” Margaritondo says. “There is one element that is absolutely necessary for somebody to be a leader which is to be respected. Your father really commanded the respect of the collaborators.”

Thinking about my father’s work, I also realize how he and I differed in an important respect. I couldn’t cut it as a physicist in part because of my reflexive pessimism. Confronted with some complex scheme, I tend to respond, “That will never work.” In contrast, my father had the confidence to break a barely conceivable technical proposal into parts, identify the obstacles, and devise ways to overcome them. “He was the most optimistic person I ever met,” White says.

By virtue of that optimism, my father helped create facilities that have enabled thousands of scientists to explore the natural world, to the benefit of us all. That legacy is far less concrete, but far more important than any particular accelerator. So, that’s what I’ll hold on to now.