It was almost six o’clock in the early evening of Thursday, May 17, when electrical engineer Sergio Marques took the opportunity to stretch his legs and look for more energy in yet another cup of coffee. Then, he would resume taking the measurements his team had been working on since the beginning of the week, together with Brazilian physicist Liu Lin’s research group, sometimes for 24 hours at a stretch. Marques and Lin, both researchers at the Brazilian Synchrotron Light Laboratory (LNLS) in Campinas, in central São Paulo State, had been testing the components of a linear electron accelerator purchased for US$6 million from the Institute of Applied Physics in Shanghai, China. Installed during the previous weeks in a 32-meter tunnel with concrete walls, every half a second the device propels microscopic packets of trillions of negatively charged particles at close to the speed of light. The accelerator will feed the largest, most complex and versatile research instrument ever built in the country: Sirius, a state-of-the-art source of synchrotron radiation, which is a special type of light that allows researchers to investigate the structure of matter at the scale of atoms and molecules.
Sirius has been under construction since 2014 at the Brazilian Center for Research in Energy and Materials (CNPEM), 15 kilometers from Campinas. It should be ready for an initial test by the end of this year, if the requested funds approved by the federal government months ago are released soon. The new synchrotron light source is a particle accelerator comprising three parts. It is installed in a 68,000 square-foot building that must remain as isolated as possible from temperature changes and external vibrations, especially those generated by truck traffic on the highway connecting Campinas to Mogi-Mirim, which is two kilometers away.
![](http://revistapesquisa.fapesp.br/wp-content/uploads/2019/04/018_Sirius08_269.jpg)
Léo Ramos Chaves
Detail of an wiggler, a series of magnets that make the electrons snake inside the storage ring and release energy in the form of synchrotron lightLéo Ramos ChavesDesigned by the LNLS teams, Sirius will replace the UVX, the first source of synchrotron light in the Southern hemisphere. Built in the 1990s, today the UVX is no longer competitive. Approximately 90% of Sirius’s components were developed at the LNLS workshops or designed there and produced by Brazilian high-tech companies. The linear accelerator is an exception. “Due to time concerns, we commissioned a machine with high-level specifications from researchers who had completed a third-generation synchrotron light source in Shanghai, one generation prior to Sirius, and they provided us data on almost every part of the accelerator,” Marques explains. He began working at the UVX in 1997 at the age of 16 and now leads the LNLS diagnostic group, which monitors the electron beam and the quality of synchrotron light that arrives at its experimental stations.
When it goes into full operation, Sirius will be—for a limited time—the most advanced source of synchrotron light in the world, in addition to being the brightest X-ray spectral source in its energy class (see article). Put simply, this means that the accelerator will allow researchers to extract very concentrated beams of light from electron streams traveling at almost 300,000 kilometers per second. These beams that can penetrate deep into dense materials such as rock and will produce clear images of objects only a few nanometers (millionths of a millimeter) apart. The intense brightness of the beams will reduce the image acquisition time from samples from hours to seconds, which is crucial in the study of biological materials, which degrade rapidly. The reduction in the time needed to produce each image will allow a greater number of images to be obtained per second and enable scientists to reconstruct the movement of very fast phenomena at the level of atoms and molecules, such as interactions between two compounds, or the movement of ions in charging and discharging batteries.
• The race for the best light
Sirius’s resolving power will outpace third-generation synchrotron light sources such as the current technology at the European Synchrotron Radiation Facility (ESRF) in France. Israeli researcher Ada Yonath performed some of the experiments that defined the three-dimensional structure of the ribosome—the protein-producing organelle in cells—at the ESRF, which earned her the 2009 Nobel Prize in Chemistry. Images from Sirius are also expected to achieve a resolution up to 1,000 times better than that of the UVX, which is a second-generation light source. Even though it is out of date, the UVX allowed physicist and professor Glaucius Oliva and his team at the University of São Paulo (USP) in São Carlos to identify the three-dimensional structure of the NS5 protein, which is essential to Zika virus reproduction (see Pesquisa FAPESP issue no. 254).
With the new synchrotron in Campinas, researchers expect to go even further and identify the three-dimensional structures of larger, more complex proteins of interest in biology and pharmaceuticals and study materials of interest to industry (see infographic). “Sirius is very close to the limit of what engineering can currently build, and will be able to produce internationally competitive science for at least a decade,” says physicist Antônio José Roque da Silva, director of the LNLS and the Sirius project. A professor at USP and a specialist in the mathematical modeling of materials at the atomic scale, Silva arrived at the LNLS in 2009 with two missions: first, to improve the UVX, which as an aging technology was beginning to lose users and researchers to institutions abroad, and second, to carry out the project of building its replacement. The name Sirius would come later, borrowed from the brightest star in the night sky.
From the beginning, Silva sought the help of two former LNLS collaborators: civil engineer Antonio Ricardo Droher Rodrigues, one of the three Brazilians who led the construction of the UVX from 1987 to 1997, and French physicist Yves Petroff, who directed synchrotron light labs in France and participated in the first Brazilian light source project. “The UVX no longer had the ability to compete, so we opted to improve in niches where we could produce relevant research using infrared and ultraviolet radiation,” Silva says. At the same time, the trio perfected a third-generation light source project developed by the team led by physicist José Antônio Brum, who directed the Brazilian Association of Synchrotron Light Technology (ABTLuS), now CNPEM, from 2001 to 2009. Three years later, with a mature project in hand, Silva and his team submitted their project to an international scientific committee.
In their final report, the committee members said the design of the new light source was excellent according to the current standards, but they recommended that the team strive for a level of brightness that would remain competitive into the future. “There wasn’t a machine with the characteristics they were suggesting in operation anywhere in the world,” recalled Silva on the morning of May 17, in his office at the LNLS. “It was our chance to get out in front of the United States, Japan, and the European countries, and stay there for a while.”
The LNLS teams returned to the design table and resumed equipment testing. Responsible for accelerator physics at the LNLS, Liu Lin and her group redesigned Sirius’s magnetic lattice so that its brightness would surpass that of the existing machines. Six months later, the committee approved the new project, budgeted at US$585 million (R$1.3 billion at the time). Obtaining stable financing was critical but only one part of the problem. “We had to acquire a location for construction and define the building’s characteristics while at the same time we were redesigning the machine and looking for a way around technological challenges,” Silva recalls. “There were times when we were juggling 20 plates in the air.”
![](http://revistapesquisa.fapesp.br/wp-content/uploads/2019/04/018_Sirius04_269.jpg)
Glaucius Oliva/IFSC-USP
The three-dimensional structure of the NS5 protein of the Zika virus, defined atom by atomGlaucius Oliva/IFSC-USPThe first R$9 million for the preproject were disbursed in 2009 and 2010 by the then Brazilian Ministry of Science and Technology (MCT), under the management (2005–2010) of physicist Sergio Rezende, who had first encountered Brum’s project in 2008. However, a definitive source of funding was missing, which would initially be provided by the MCT (now MCTIC, after incorporating Innovations and Telecommunications), together with the Brazilian Development Bank (BNDES) and other development agencies. Two other underwriters succeeded the ministry to lead the funding and contributed R$77 million to the project. Finally, in 2014, engineer Clélio Campolina Diniz was able to give the green light to begin construction with a proposed 2015 budget of R$240 million. The following year, Sirius was included in the second edition of the Growth Acceleration Program (PAC), which today is a Programa Avançar project.
Fluctuations in the dollar, inflation, and improvements to the light source and structural design elevated Sirius’s cost to R$1.8 billion. “It’s the only Brazilian project on this scale going forward without major delays,” says electronics engineer and physicist Rogério Cezar de Cerqueira Leite, chairman of the CNPEM board of directors, an NGO linked to MCTIC, the managing agency of the LNLS.
Pedro Wongtschowski, a chemical engineer who chaired the CNPEM board from 2010 to 2015, attributes the project’s adherence to its schedule and its low number of budget changes to the adoption of a governance model used by the private sector on large-scale projects. “Project execution only began once a detailed implementation plan was completed; the contracting of work was done through a careful bidding process, and the equipment that required longer delivery times was acquired first,” he recalls. “We also took advantage of the deployment of Sirius to develop components through Brazilian suppliers, a move that received support from FAPESP,” says Wongtschowski, current chair of the board of directors at Ultra (Ultrapar Participações) and a member of FAPESP’s governing council.
![](http://revistapesquisa.fapesp.br/wp-content/uploads/2019/04/018_Sirius01_269.jpg)
Renan Picoreti – LNLS Publicity/CNPEM
Aerial image of the Sirius building, taken in mid-June in 2018Renan Picoreti – LNLS Publicity/CNPEMOf the total estimated cost, R$1.16 billion has already been delivered by the MCTIC, of which R$760 million was spent under the management of Gilberto Kassab, as noted by Cerqueira Leite, who played a fundamental role in the implementation of the UVX during the 1980s. In Leite’s view, Sirius only survived the recent economic slowdown because the project managed to gradually interest “authorities and politicians in Brasília,” in addition to its creators and the scientific community.
A similar conclusion was reached years ago by two researchers who analyzed the process of creation and implementation of the UVX. Léa Velho, a professor at the Department of Science and Technology Policy at the University of Campinas (UNICAMP), and Osvaldo Frota Pessoa Junior, a professor at the Department of Philosophy at USP, evaluated the reasoning that motivated the construction of the first Brazilian synchrotron and the negotiations that took it from design to reality. In a 1998 article in the journal Social Studies of Science, they stated that support for the project came more from science policy sectors than from researchers and potential users. They added that the political skills of the few scientists involved were crucial to the project’s implementation.
![](http://revistapesquisa.fapesp.br/wp-content/uploads/2019/04/018_Sirius06_269.jpg)
Léo Ramos Chaves
Engineer Rafael Seraphim tests the vacuum system of the chambers that will propel the electronsLéo Ramos Chaves“Sirius represents an attempt to leap to a new level of quality in Brazilian science,” observes Argentine physicist Aldo Craievich. At the age of 79, retired from USP, he’s still doing research using the UVX. Together with physicist Cylon Gonçalves da Silva and Ricardo Rodrigues, Craievich was the third member of the trio who coordinated the construction of the first Brazilian synchrotron.
The first large-scale research equipment project in Brazil—i.e., “Big Science,” such as that which began in the United States during World War II with the nuclear bomb project—was initiated at the Brazilian Center for Physics Research (CBPF) in Rio de Janeiro in the early 1980s by physicist Roberto Leal Lobo and Silva Filho. With support from Lynaldo Cavalcanti de Albuquerque, then president of the National Council for Scientific and Technological Development (CNPq), Lobo guided the project until the beginning of the democratic government in 1985. With the creation of the MCT, he was replaced by Cylon, who had the support of the new agency’s minister, Renato Archer.
![](http://revistapesquisa.fapesp.br/wp-content/uploads/2019/04/018_Sirius07_269.jpg)
Léo Ramos Chaves
Quadrupole magnets, one of the components of the storage ringLéo Ramos Chaves“When we decided to build Brazil’s first synchrotron light source, the only operational model that made sense was that of a national lab along the lines of US facilities, open to users from research institutions and companies in Brazil and abroad,” Cylon notes. “The construction of the machine was merely an excuse to educate people who would be qualified to generate technology in Brazil, and capable of producing science at the frontiers of knowledge. We opted on designing and building as much as we could here nationally, which gave us the expertise used to create Sirius.”
Building equipment to do science on a large scale demands a continuous flow of funds and technical and scientific expertise, and it almost always generates controversy. This was the case with the UVX project, and on a smaller scale, with Sirius. Soon after Brazil’s first synchrotron light source project was approved, the directors of the Brazilian Society of Physics published a manifesto condemning the effort. It stated that there was not enough technical competence within the country to build it, that there would not be any users for it and that the UVX would drain resources from other areas of science and technology. “None of these predictions came true,” says Rodrigues, coordinator of the Sirius accelerators. “We built the machine, the researchers came—today there are 6,200 registered users—and the level of funding has increased in every area.”
![](http://revistapesquisa.fapesp.br/wp-content/uploads/2019/04/018_Sirius05_269.jpg)
Léo Ramos Chaves
The hall where some of Sirius’s experimental stations will be installedLéo Ramos Chaves“Large facilities like Sirius are expensive anywhere in the world, but they pay for themselves over time,” says Fernanda De Negri, an economist at the Institute for Applied Economic Research (IPEA). Its cost represents 0.05% of the overall Brazilian national budget (government revenue), approximately R$3.5 trillion. “In many areas, infrastructure like this is necessary to producing quality science capable of generating innovation and making the country more economically competitive,” the researcher says. In Negri’s book Novos caminhos para a inovação no Brasil (New pathways to innovation in Brazil; Editora Wilson Center), she mentions Sirius as a rare example of long-term scientific planning in Brazil, launched in June.
“Since the atomic bomb project and the Apollo mission, science is no longer done only on small investments and short-term vision,” says Glauco Arbix, a professor in the Department of Sociology at USP. Arbix is a former president (2011–2015) of the federal innovation promotion agency FINEP (Brazilian Funding Authority for Studies and Projects). He states that “It’s necessary to have medium- and long-term vision, and to irrigate the system in such a way as to nourish smaller labs and create research projects with scientific, economic, and social relevance that are capable of raising the level of Brazilian science and increasing its impact. Without this, the country will continue to slip behind.”
Scientific article
VELHO, L. and PESSOA JR., O. The decision-making process in the construction of the Synchrotron Light National Laboratory in Brazil. Social Studies of Science. v. 28, i. 2, p. 195–219. April, 1998.