Time filter

Source Type

Home > Press > Scientists boost catalytic activity for key chemical reaction in fuel cells: New platinum-based catalysts with tensile surface strain could improve fuel cell efficiency Abstract: Fuel cells are a promising technology for clean and efficient electrical power generation, but their cost, activity, and durability are key challenges to commercialization. Today's fuel cells use expensive platinum (Pt)-based nanoparticles as catalysts to accelerate the reactions involved in converting the chemical energy from renewable fuels--such as hydrogen, methanol, and ethanol--into electrical energy. Catalysts that incorporate less expensive metals inside the nanoparticles can help reduce cost and improve activity and durability, but further improvements to these catalysts are required before these fuel cells can be used in vehicles, generators, and other applications. Now, scientists from the U.S. Department of Energy's (DOE) Brookhaven National Laboratory, California State University-Northridge, Soochow University, Peking University, and Shanghai Institute of Applied Physics have developed catalysts that can undergo 50,000 voltage cycles with a negligible decay in their catalytic activity and no apparent changes in their structure or elemental composition. As described in a paper published online in the December 16 issue of Science, the catalysts are "nanoplates" that contain an atomically ordered Pt and lead (Pb) core surrounded by a thick uniform shell of four Pt layers. To date, the most successful catalysts for boosting the activity of the oxygen reduction reaction (ORR)--a very slow reaction that significantly limits fuel cell efficiency--have been of the Pt-based core-shell structure. However, these catalysts typically have a thin and incomplete shell (owing to their difficult synthesis), which over time allows the acid from the fuel cell environment to leach into the core and react with the other metals inside, resulting in poor long-term stability and a short catalyst lifetime. "The goal is to make the ORR as fast as possible with catalysts that have the least amount of platinum and the most stable operation over time," said corresponding author Dong Su, a scientist at Brookhaven Lab's Center for Functional Nanomaterials (CFN), a DOE Office of Science User Facility, who led the electron microscopy work to characterize the nanoplates. "Our PtPb/Pt catalysts show high ORR activity and stability--two parameters that are key to enabling a hydrogen economy--placing them among the most efficient and stable bimetallic catalysts reported for ORR." Lattice strain for enhanced catalytic performance In previous studies, scientists have shown that ORR activity can be optimally enhanced in core-shell catalysts by compressing the Pt atoms on one specific lattice surface plane called Pt(111). This compressive strain is induced by adding metals smaller in size than Pt, such as nickel, to the shell's core, and has the effect of weakening the binding of oxygen to the Pt surface, where the catalytic reaction takes place. "The ideal ORR catalyst needs to help break bonds (between oxygen molecules) and form bonds (between oxygen and hydrogen), so oxygen can't be too strongly or too weakly bound to the platinum surface," explained Su. "Scientists have focused their research on the compressively strained Pt(111) surfaces, in which Pt atoms are squeezed across the surface, because the oxygen binding energy is optimized. In general, scientists thought that tensile strain on the same surface plane would result in overly strong binding of oxygen and thus hinder the ORR reaction." But Su and his collaborators showed that introducing a large tensile strain along one direction of a different surface plane, Pt(110), could also improve ORR catalytic activity. They added Pb (which is larger than Pt) to the core of the Pt shell, causing the Pt atoms to stretch across the surface. Nanoplate characterization and durability testing After the research group led by Xiaoqing Huang, corresponding author from Soochow University, synthesized the nanoplates, Su characterized their structure and elemental composition at the CFN. Using electron diffraction patterns and images from high-resolution scanning transmission electron microscopy (STEM), both of which reveal the relative positions of atoms, he confirmed the core-shell structure and the composition and sequence of the atoms. To verify that the core contained Pt and Pb and that the shell contained Pt, he measured the change in energy of the electrons after they interacted with the nanoplates--a technique called electron energy-loss spectroscopy. With this information, the team distinguished how the nanoplates formed with the individual Pt and Pb atoms. To their surprise, the surface planes were not Pt(111) but Pt(110), and these Pt(110) planes were under biaxial strain--compressive strain in one direction and tensile strain in the other--originating from the PtPb core. In durability tests simulating fuel cell voltage cycling, Su's collaborators found that after 50,000 cycles there was almost no change in the amount of generated electrical current. In other words, the nanoplates had minimal decay in catalytic activity. After this many cycles, most catalysts exhibit some activity loss, with some losing more than half of their original activity. Microscopy and synchrotron characterization techniques revealed that the structure and elemental composition of the nanoplates did not change following durability testing. "The electron microscopy work at CFN was critical in explaining why our nanoplates showed such high catalytic activity and stability," said Huang. Compared to commercial Pt-on-carbon (Pt/C) catalysts, the team's PtPb/Pt nanoplates have one of the highest ORR activities to date, taking the amount of Pt used into account, and excellent durability. The team's nanoplates also showed high electrocatalytic activity and stability in oxidation reactions of methanol and ethanol. "We believe the relatively thick and complete Pt layers play an important role in protecting the core," said Su. To understand how the high ORR activity originates in the nanoplates, the scientists calculated the binding energy between oxygen atoms and Pt atoms on the surface. Their calculations confirmed the tensile strain on the Pt(110) surface was responsible for the enhanced ORR activity. "This work opens up a new way to introduce large tensile strain on the stable Pt(110) plane to achieve very high activity for oxygen reduction catalysis. We believe that our approach will inspire efforts to design new nanostructured catalysts with large tensile strain for more efficient catalysis," said corresponding author Shaojun Guo of Peking University. Eventually, the laboratory-level electrocatalysts will need to be tested in a larger fuel cell system, where real-world variables--such as pollutants that could impact surface reactivity--can be introduced. ### In addition to the electron microscopy work performed at CFN, supported by DOE's Office of Science, this work involved theoretical calculations by Gang Lu's group at California State University-Northridge and synchrotron characterization by Jing-Yuan Ma at Shanghai Synchrotron Radiation Facility. This work was further supported by the National Basic Research Program of China, the National Natural Science Foundation of China, the Ministry of Science and Technology of the People's Republic of China, Soochow University, Peking University, the Young Thousand Talents Program, the Priority Academic Program Development of Jiangsu Higher Education Institutions, and the U.S. Army Research Office. About Brookhaven National Laboratory Brookhaven National Laboratory is supported by the Office of Science of the U.S. Department of Energy. The Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit science.energy.gov. For more information, please click If you have a comment, please us. Issuers of news releases, not 7th Wave, Inc. or Nanotechnology Now, are solely responsible for the accuracy of the content.


News Article | December 16, 2016
Site: www.eurekalert.org

UPTON, NY--Fuel cells are a promising technology for clean and efficient electrical power generation, but their cost, activity, and durability are key challenges to commercialization. Today's fuel cells use expensive platinum (Pt)-based nanoparticles as catalysts to accelerate the reactions involved in converting the chemical energy from renewable fuels--such as hydrogen, methanol, and ethanol--into electrical energy. Catalysts that incorporate less expensive metals inside the nanoparticles can help reduce cost and improve activity and durability, but further improvements to these catalysts are required before these fuel cells can be used in vehicles, generators, and other applications. Now, scientists from the U.S. Department of Energy's (DOE) Brookhaven National Laboratory, California State University-Northridge, Soochow University, Peking University, and Shanghai Institute of Applied Physics have developed catalysts that can undergo 50,000 voltage cycles with a negligible decay in their catalytic activity and no apparent changes in their structure or elemental composition. As described in a paper published online in the December 16 issue of Science, the catalysts are "nanoplates" that contain an atomically ordered Pt and lead (Pb) core surrounded by a thick uniform shell of four Pt layers. To date, the most successful catalysts for boosting the activity of the oxygen reduction reaction (ORR)--a very slow reaction that significantly limits fuel cell efficiency--have been of the Pt-based core-shell structure. However, these catalysts typically have a thin and incomplete shell (owing to their difficult synthesis), which over time allows the acid from the fuel cell environment to leach into the core and react with the other metals inside, resulting in poor long-term stability and a short catalyst lifetime. "The goal is to make the ORR as fast as possible with catalysts that have the least amount of platinum and the most stable operation over time," said corresponding author Dong Su, a scientist at Brookhaven Lab's Center for Functional Nanomaterials (CFN), a DOE Office of Science User Facility, who led the electron microscopy work to characterize the nanoplates. "Our PtPb/Pt catalysts show high ORR activity and stability--two parameters that are key to enabling a hydrogen economy--placing them among the most efficient and stable bimetallic catalysts reported for ORR." In previous studies, scientists have shown that ORR activity can be optimally enhanced in core-shell catalysts by compressing the Pt atoms on one specific lattice surface plane called Pt(111). This compressive strain is induced by adding metals smaller in size than Pt, such as nickel, to the shell's core, and has the effect of weakening the binding of oxygen to the Pt surface, where the catalytic reaction takes place. "The ideal ORR catalyst needs to help break bonds (between oxygen molecules) and form bonds (between oxygen and hydrogen), so oxygen can't be too strongly or too weakly bound to the platinum surface," explained Su. "Scientists have focused their research on the compressively strained Pt(111) surfaces, in which Pt atoms are squeezed across the surface, because the oxygen binding energy is optimized. In general, scientists thought that tensile strain on the same surface plane would result in overly strong binding of oxygen and thus hinder the ORR reaction." But Su and his collaborators showed that introducing a large tensile strain along one direction of a different surface plane, Pt(110), could also improve ORR catalytic activity. They added Pb (which is larger than Pt) to the core of the Pt shell, causing the Pt atoms to stretch across the surface. After the research group led by Xiaoqing Huang, corresponding author from Soochow University, synthesized the nanoplates, Su characterized their structure and elemental composition at the CFN. Using electron diffraction patterns and images from high-resolution scanning transmission electron microscopy (STEM), both of which reveal the relative positions of atoms, he confirmed the core-shell structure and the composition and sequence of the atoms. To verify that the core contained Pt and Pb and that the shell contained Pt, he measured the change in energy of the electrons after they interacted with the nanoplates--a technique called electron energy-loss spectroscopy. With this information, the team distinguished how the nanoplates formed with the individual Pt and Pb atoms. To their surprise, the surface planes were not Pt(111) but Pt(110), and these Pt(110) planes were under biaxial strain--compressive strain in one direction and tensile strain in the other--originating from the PtPb core. In durability tests simulating fuel cell voltage cycling, Su's collaborators found that after 50,000 cycles there was almost no change in the amount of generated electrical current. In other words, the nanoplates had minimal decay in catalytic activity. After this many cycles, most catalysts exhibit some activity loss, with some losing more than half of their original activity. Microscopy and synchrotron characterization techniques revealed that the structure and elemental composition of the nanoplates did not change following durability testing. "The electron microscopy work at CFN was critical in explaining why our nanoplates showed such high catalytic activity and stability," said Huang. Compared to commercial Pt-on-carbon (Pt/C) catalysts, the team's PtPb/Pt nanoplates have one of the highest ORR activities to date, taking the amount of Pt used into account, and excellent durability. The team's nanoplates also showed high electrocatalytic activity and stability in oxidation reactions of methanol and ethanol. "We believe the relatively thick and complete Pt layers play an important role in protecting the core," said Su. To understand how the high ORR activity originates in the nanoplates, the scientists calculated the binding energy between oxygen atoms and Pt atoms on the surface. Their calculations confirmed the tensile strain on the Pt(110) surface was responsible for the enhanced ORR activity. "This work opens up a new way to introduce large tensile strain on the stable Pt(110) plane to achieve very high activity for oxygen reduction catalysis. We believe that our approach will inspire efforts to design new nanostructured catalysts with large tensile strain for more efficient catalysis," said corresponding author Shaojun Guo of Peking University. Eventually, the laboratory-level electrocatalysts will need to be tested in a larger fuel cell system, where real-world variables--such as pollutants that could impact surface reactivity--can be introduced. In addition to the electron microscopy work performed at CFN, supported by DOE's Office of Science, this work involved theoretical calculations by Gang Lu's group at California State University-Northridge and synchrotron characterization by Jing-Yuan Ma at Shanghai Synchrotron Radiation Facility. This work was further supported by the National Basic Research Program of China, the National Natural Science Foundation of China, the Ministry of Science and Technology of the People's Republic of China, Soochow University, Peking University, the Young Thousand Talents Program, the Priority Academic Program Development of Jiangsu Higher Education Institutions, and the U.S. Army Research Office. Brookhaven National Laboratory is supported by the Office of Science of the U.S. Department of Energy. The Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit science.energy.gov.


« Continental enters strategic partnership with DigiLens for augmented reality head-up displays | Main | FCA delivers 100 uniquely built Chrysler Pacifica PHEVs to Waymo for autonomous driving test fleet » Scientists from the US Department of Energy’s (DOE) Brookhaven National Laboratory; California State University–Northridge; Soochow University; Peking University; and Shanghai Institute of Applied Physics have developed new catalysts for the oxygen reduction reaction (ORR) in fuel cells that can undergo 50,000 voltage cycles with a negligible decay in their catalytic activity and no apparent changes in their structure or elemental composition. In a paper published in Science, the team reports on a class of platinum-lead/platinum (PtPb/Pt) core/shell nanoplate catalysts that exhibit large biaxial strains. (Modifying the electronic structure of catalysts can improve their performance; lattice strain (either compressive or tensile) modifies the distances between surface atoms and hence modifies catalytic activity. Earlier post.) The stable Pt (110) facets of the nanoplates have high ORR specific and mass activities that reach 7.8 milliampere (mA) per cm2 and 4.3 ampere per milligram of platinum at 0.9 volts versus the reversible hydrogen electrode (RHE), respectively. Hydrogen enters a fuel cell at the anode, which is coated with a platinum catalyst. When the hydrogen molecules come into contact with the platinum, they split into hydrogen ions and free electrons in a reaction called hydrogen oxidation. The hydrogen ions travel through the electrolyte to the cathode; the electrons travel through an external circuit to generate electricity before arriving at the cathode. In the oxygen reduction reaction, oxygen molecules combine with the electrons that have completed their circuit and the hydrogen ions to produce water. The rate of ORR is very slow because molecular oxygen has a very strong double bond that requires significant energy to be broken. As a result, the fuel cell suffers from a reduction in voltage (overpotential) that limits its efficiency. To date, the most successful catalysts for boosting the activity of the oxygen reduction reaction (ORR)—a very slow reaction that significantly limits fuel cell efficiency—have been of the Pt-based core-shell structure. However, these catalysts typically have a thin and incomplete shell (owing to their difficult synthesis), which over time allows the acid from the fuel cell environment to leach into the core and react with the other metals inside, resulting in poor long-term stability and a short catalyst lifetime. In previous studies, scientists have shown that ORR activity can be optimally enhanced in core-shell catalysts by compressing the Pt atoms on one specific lattice surface plane called Pt(111). This compressive strain is induced by adding metals smaller in size than Pt, such as nickel, to the shell’s core, and has the effect of weakening the binding of oxygen to the Pt surface, where the catalytic reaction takes place. Su and his collaborators showed that introducing a large tensile strain along one direction of a different surface plane, Pt(110), could also improve ORR catalytic activity. They added Pb (which is larger than Pt) to the core of the Pt shell, causing the Pt atoms to stretch across the surface. After the research group led by Xiaoqing Huang, corresponding author from Soochow University, synthesized the nanoplates, Su characterized their structure and elemental composition at the CFN. Using electron diffraction patterns and images from high-resolution scanning transmission electron microscopy (STEM), both of which reveal the relative positions of atoms, he confirmed the core-shell structure and the composition and sequence of the atoms. To verify that the core contained Pt and Pb and that the shell contained Pt, he measured the change in energy of the electrons after they interacted with the nanoplates—a technique called electron energy-loss spectroscopy. With this information, the team distinguished how the nanoplates formed with the individual Pt and Pb atoms. To their surprise, the surface planes were not Pt(111) but Pt(110), and these Pt(110) planes were under biaxial strain—compressive strain in one direction and tensile strain in the other—originating from the PtPb core. Microscopy and synchrotron characterization techniques revealed that the structure and elemental composition of the nanoplates did not change following the durability testing. Compared to commercial Pt-on-carbon (Pt/C) catalysts, the team’s PtPb/Pt nanoplates have one of the highest ORR activities to date, taking the amount of Pt used into account, and excellent durability. The team’s nanoplates also showed high electrocatalytic activity and stability in oxidation reactions of methanol and ethanol. To understand how the high ORR activity originates in the nanoplates, the scientists calculated the binding energy between oxygen atoms and Pt atoms on the surface. Their calculations confirmed the tensile strain on the Pt(110) surface was responsible for the enhanced ORR activity. Eventually, the laboratory-level electrocatalysts will need to be tested in a larger fuel cell system, where real-world variables—such as pollutants that could impact surface reactivity—can be introduced. In addition to the electron microscopy work performed at CFN, supported by DOE’s Office of Science, this work involved theoretical calculations by Gang Lu’s group at California State University–Northridge and synchrotron characterization by Jing-Yuan Ma at Shanghai Synchrotron Radiation Facility. This work was further supported by the National Basic Research Program of China, the National Natural Science Foundation of China, the Ministry of Science and Technology of the People’s Republic of China, Soochow University, Peking University, the Young Thousand Talents Program, the Priority Academic Program Development of Jiangsu Higher Education Institutions, and the U.S. Army Research Office.


News Article | October 26, 2016
Site: www.nature.com

No statistical methods were used to predetermine sample size. The experiments were not randomized and the investigators were not blinded to allocation during experiments and outcome assessment. The genes encoding TRIC-B1 and -B2 from Caenorhabditis elegans (UniProt codes: Q9NA75, Q9NA73; GI: 290457497, 71998474) were synthesized (Genscript) with optimized codon usage for protein expression in Pichia pastoris. The target cDNA was inserted between the EcoR1/Xho1 sites of the pPICZ-A (or pPICZ-C for TRIC-B2) vector (Invitrogen), yielding a construct with a C-terminal fusion polypeptide containing the c-Myc epitope and a polyhistidine (6 × His) tag. To improve the crystallizability of the target proteins, 48 and 61 amino acid residues at the flexible C-terminal regions of TRIC-B1 and -B2 as well as those of the Myc epitope sequence from the vector were truncated, yielding the expression products (residues 1–247 of TRIC-B1 and 1–252 of TRIC-B2) covering the transmembrane domain of the full-length protein. All point mutations were introduced through the Quikchange site-directed mutagenesis. The expression vectors were linearized with Pme I and transformed into P. pastoris GS115 strain by electroporation using the Micropulser Electroporator (BioRad). The transformants were selected by plating on the YPD agar plates with 0.1 and 1.0 mg/ml zeocin. For the large-scale protein expression, the colony with the highest protein expression level was used to inoculate 250 ml Minimal Glycerol Media containing Histidine (MGYH). When the OD reached 2.0, the culture was used to inoculate 1 l MGYH media at 1:40 (v:v) ratio in a 5 l baffled flask. The cells were grown in the shaking incubator for 24 h at 29 °C and 280 r.p.m. with the feeding of glycerol as the carbon source. When the OD reached 4.0, the cells were spun down and resuspended in the Minimal Methanol + Histidine (MMH) media. After the media exchange, the cells were grown at 27 °C (or 24 °C) and 260 r.p.m. The induction of protein expression was initiated by adding methanol to the culture at 0.5% (v:v) final concentration. Protein expression continued for ~48–60 h and during the period, methanol was added every 24 h. The cells were harvested by centrifugation at 8,983g in JLA-8.1000 rotor (Beckman) and were stored at −80 °C after being frozen instantly in liquid nitrogen. For protein purification, the frozen cell pellets were suspended in a lysis buffer (50 mM TRIS-HCl pH 8.0, 150 mM KCl) at 1:10 (m:v) ratio and then homogenized by using the T10 basic homogenizer (IKA). The cells were lysed by passing through a high-pressure homogenizer (ATS Engineering, Shanghai) at 1,300 bar 4 times. To solubilize the membrane proteins, Triton X-100 was added to the cell lysate at 1.5% (v/v) final concentration. The mixture was stirred at room temperature (RT) for 2 h to extract the target proteins from the membrane. The insoluble cell debris was removed by centrifugation at 37,044g in JA-25.50 rotor (Beckman) for 1 h. The supernatant was collected and mixed with the cobalt affinity beads (Talon, BD Science) at 1 ml resin/30 g cell pellet ratio. The resin was pre-equilibrated with a solution containing 150 mM KCl, 25 mM HEPES, pH 7.5, 10 mM imidazole, 0.5% n-decyl-β-d-maltopyranoside (β-DM, Anatrace). After 1-h incubation at RT, the mixture was loaded on a column, washed with five column volumes of buffer A (150 mM KCl, 25 mM HEPES, pH 7.5, 10 mM imidazole, 0.5% β-DM) and then five volumes of buffer B (150 mM KCl, 25 mM HEPES, pH 7.5, 20 mM imidazole, 0.4% β-DM). The target protein was eluted with buffer C (150 mM KCl, 25 mM HEPES, pH 7.5, 300 mM imidazole, 0.4% β-DM). The fractions with protein concentration above 0.3 mg/ml were combined and then diluted by adding two volumes of buffer D (150 mM KCl, 25 mM HEPES, pH 7.5, 0.4% β-DM) to lower imidazole concentration and prevent protein precipitation. The protein was concentrated to 10–15 mg/ml in the 50 kDa cut-off Amicon Ultra-4 centrifugal filter unit (Millipore). The concentrated protein sample was further purified through a Superdex-200 10/300 GL (GE Healthcare) gel filtration column in buffer E (150 mM KCl, 10 mM HEPES pH 7.5, 0.3% β-DM). The fractions between 12 and 13 ml were collected, concentrated to ∼10 mg/ml, frozen in liquid nitrogen as small aliquots and stored at −80 °C for further use in crystallization or functional assays. TRIC-B1(CΔ48) was crystallized through the hanging-drop vapour diffusion method at 16 °C. The initial crystallization conditions were identified via the sparse-matrix screening method using the MemGold I and II kits (Molecular Dimensions). Plate crystals (space group C222 ) grew within drops prepared by mixing the concentrated protein solution (10–13 mg/ml) with the well solution (22% PEG550 MME, 0.2 M NaCl, 0.1 M HEPES, pH 7.0) in 1:1 (v:v) ratio. Another form, namely tetragonal bipyramid crystals (in P4 2 2 space group), was grown with the well solution containing 20–24% PEG400, 10% glycerol, 50 mM ADA buffer (pH 6.5). For the purpose of phasing, the TRIC-B1 crystals in C222 space group were soaked in 22% PEG550 MME, 15% glycerol, 0.2 M NaCl, 0.1 M HEPES, pH 7.0, 0.4% β-DM with 2 mM CH HgCl for 16 h and yielded diffraction data to 3.9 Å resolution. For the P4 2 2 crystals used for solving the structure with Ca2+ bound, they were co-crystallized with 4 mM CaCl by combining 1.0 μl protein-CaCl mixture with 1.0 μl well solution (18.4% PEG400, 4.4% PEG550 MME, 8% glycerol, 40 mM ADA buffer, pH 8.0, 40 mM NaCl, 20 mM HEPES, pH 7.0). For derivatizing the P4 2 2 crystals with Rb+ or Cs+, the TRIC-B1(CΔ48) protein was co-purified and co-crystallized with 150 mM RbCl or CsCl and the crystals were soaked for 1–3 min in solutions containing 0.5 M RbCl or 1 M CsCl, 22% PEG400, 4.4% PEG550 MME, 10% glycerol, 40 mM ADA buffer (pH 8.0), 40 mM NaCl, 20 mM HEPES, pH 7.0 and 0.4% β-DM. The BaCl derivatives were obtained by growing the C222 crystals in a solution with 50 mM BaCl , 22% PEG550 MME, 0.2 M NaCl, 0.1 M HEPES, pH 7.0. TRIC-B2(CΔ61) was crystallized through the sitting or hanging-drop vapour diffusion method at 16 °C. The well solution contains 20% PEG400, 50 mM NaAc buffer (pH 4.4), 50 mM MgAc and 10 mM betaine hydrochloride. Rhombohedron-shaped crystals usually appeared in two weeks and matured in 2–3 months. Before data collection, a crystal was soaked in situ for 15 h at 16 °C in a solution with 20% PEG400, 50 mM NaAc (pH 4.4), 50 mM BaCl , 150 mM KCl, 0.5% β-DM and 4% 2,2,2-trifluoroethanol. The cryoprotection was achieved by raising PEG400 concentration to 30% while the other components remain constant. Although the soaking solution contains BaCl , it is evident that Ba2+ did not bind to TRIC-B2 under the acidic soaking condition, as no Ba2+ signals can be detected in the anomalous difference Fourier map, presumable due to the low affinity of metal binding at pH 4.4. On the other hand, the TRIC-B1 crystal with BaCl bound was prepared in a solution at pH 7.0, more favourable for the binding of Ba2+. The diffraction data were collected at BL17U of the Shanghai Synchrotron Radiation Facility (SSRF), or at BL1A, BL5A or NW12A beamlines of the Photon Factory (Tsukuba, Japan). The data processing was carried out by using iMosflm or HKL2000 programs. The initial experimental phases were solved by using the CH HgCl-derivatized TRIC-B1 data (collected at 1.00731 Å wavelength near the L-III edge of Hg) through the single-wavelength anomalous diffraction (SAD) method by using the Autosol program of the Phenix suite31. The initial phases solved from a total of 11 Hg atoms have a figure-of-merit (FOM) at 0.36 and after density modification, the FOM is improved to 0.59. An initial model of TRIC-B1 with seven transmembrane helices was manually built in Coot32. The anomalous difference Fourier signals of the two Hg atoms bound to Cys61 on the M2 helix and Cys197 on the M6 helix were used to verify the sequence registration on the transmembrane helices. Moreover, the relative topological relationship of M1, M3–5 and M7 transmembrane helices with respect to M2 and M6 was deduced according to the secondary structure prediction result from the PSIPRED web server. The structural model of TRIC-B1 was completed through iterative manual model building and refinement in CNS program (versions 1.2 and 1.3)33 with the maximum likelihood target function using amplitudes. Each asymmetric unit of the TRIC-B1 crystal in P4 2 2 or C222 space group contains a homotrimer of the TRIC-B1–PIP complex. The TRIC-B2 structure was solved through the molecular replacement method using the initial poly-Ala model of a partial TRIC-B1 monomer (with only seven transmembrane helices). After a solution was found in the Phaser program34 (TFZ = 7.7, LLG = 39), the model was completed by running the automatic model building programs with the 2.3 Å high-resolution data in PHENIX (AutoBuild)31 first, leading to an improved model with R  = 35.08% and R  = 38.76%. A second round of automatic model building was carried out in ARP/wARP35, resulting in a more complete model with R  = 22.9% and R  = 28.6%. Further model building and adjustment was performed manually with the Coot program32, and the structure refinement was carried out with the CNS program33. Lipid, detergent and water molecules were added manually in the model for refinement at later stages when their electron densities were well defined. Each asymmetric unit of the TRIC-B2 crystal in R32 space group contains a TRIC-B2 monomer in complex with one PIP molecule. The crystallographic threefold axis coincides with the C3 axis of the homotrimer of TRIC-B2–PIP complex. The presence of PIP in TRIC-B1 and -B2 is supported by the following crystallographic evidence. The SigmaA-weighted 2F − F map and simulated annealing (SA) omit map of TRIC-B2 both show strong electron densities that fit well with the PIP structural model (Extended Data Fig. 2a, b). Moreover, the diffraction data collected with the TRIC-B1 crystals at 3 Å wavelength yielded three anomalous difference Fourier peaks matching well with the three phosphorus atoms on the PIP head group (Extended Data Fig. 2d), confirming the presence of PIP in TRIC-B1 protein. The statistics of data analysis, phasing and structure refinement are summarized in Extended Data Table 1a, b. For the structures of TRIC-B1/B2, 91.8%/90.8% amino acid residues have their main chain dihedral angles in preferred regions of the Ramachandran plot; 8.2%/9.2% were in the allowed regions; and 0.0%/0.0% were in the outliers. The molecular graphics were produced with PyMOL36. To extract lipids in the purified TRIC-B1 protein samples, the wild-type or K129A/R133L mutant protein (200 μl at ∼5 mg/ml) was mixed with 180 μl of chloroform/methanol/concentrated HCl (1:2:0.02, v/v/v) solution. Subsequently, 60 μl of chloroform and 60 μl of 2 M KCl (sigma) were added to each tube. The tubes were vortexed and then centrifuged for 5 min at 2,000g to separate the organic phase from the aqueous phase. The organic phase was pipetted out and applied to the PVDF membrane in small dotted area, and the membrane was dried in air. The PIP lipid was detected through incubating the blotted membrane with the Mouse anti-PIP antibody 2C11 (Abcam) as primary antibody, and then with a secondary antibody of goat anti-mouse IgG*HRP (Zsbio). The signals were developed by adding the western lightning ultra ECL horseradish peroxidase substrate (Perkin-Elmer). Images of blots were captured on a chemiluminescence CCD imaging system (ChemiScope 3500 mini imager, Clinx Science Instruments). To further verify the presence of PIP in the TRIC-B2 protein sample, mass spectrometry analysis was performed on the lipid extracts of the sample. To extract the lipid, 1.8 mg purified TRIC-B2 protein (at 1.5 mg/ml) was treated with 18 mg SM2 Biobeads (Bio-rad) at 4 °C overnight to remove the detergent (β-DM). After the detergent-free sample was separated from the Biobeads, it was split into two aliquots and then dried to 10 μl under vacuum. Subsequently, a mixture of 300 μl methanol and 1 ml methyl tertiary butyl ether (MTBE) and 13 μl 0.1 N HCl was added to every 10 μl protein sample and then the tubes were vortexed once every 2 min (for 20 min). Then, 250 μl 0.1 N HCl was added to the above mixture and the tubes were shaken to homogenize the samples. The mixture was centrifuged for 5 m at 16,200g and the upper phase was kept as ‘Extract I’. To further process the sample, the mixture of 300 μl methanol, 1 ml MTBE and 13 μl 0.1 N HCl was centrifuged and the lower organic-phase solvent was taken and mixed with ‘Extract I’. This mixture was centrifuged again and the upper layer was kept as ‘Extract II’. One half of ‘Extract II’ was treated with trimethylsilyl diazomethane (the other half was stored at −20 °C as a backup sample) and mass spectrometry was performed according to the protocol described in ref. 37. The PIP molecules were resolved as methylated derivatives. The result confirms that the TRIC-B2 protein sample contains 34:0, 34:1, 34:2, 34:3, 36:1, 36:2, 36:3 and 36:8 PIP molecules (for the spectrum of 34:1-PIP , see Extended Data Fig. 2c). For the preparation of small unilamellar vesicle (SUVs) samples for fluorescence-based K+ flux assay, 10 mg/ml Azolectin lipid mixture dissolved in chloroform was aliquoted and dried under vacuum in a CentriVap Concentrator (Labconco). The dried lipid film was solubilized to 10 mg/ml in the dialysis buffer (10 mM HEPES pH7.0, 150 mM KCl, 0–5 mM CaCl ) plus 8% n-octyl-β-d-maltopyranoside (β-OM). The sample was then sonicated in a bath sonicator three times (10 s on, 10 s off for 1 min). Subsequently, the purified TRIC-B1(CΔ48) protein was added to the mixture at a protein:lipid ratio of 1:100 (w:w). The sample was rotated gently for 1 h at RT before being injected into a 10 kDa-cut-off Slide-A-Lyzer cassette (Pierce) for dialysis at 4 °C. During dialysis, the removal of detergent from the sample was facilitated by adding 40 mg/ml SM2 biobeads (Bio-rad) in the dialysis buffer. After being dialysed for 5–7 days with buffer exchange every 12 h, the sample with reconstituted SUV was retrieved from the cassette with a syringe, aliquoted in 20-μl batches and then stored at −80 °C after being flash frozen in liquid nitrogen. The control vesicles without protein added were prepared as above, while the protein sample was replaced with the blank buffer containing 10 mM HEPES-KOH (pH 7.0), 150 mM KCl and 0.3% β-DM. The protocol for the K+ flux assay was based on the published methods in refs 38,39. For each reaction, 5 μl frozen vesicle sample was thawed, briefly sonicated, and diluted 20 fold into a flux-assay solution containing 150 mM NMDG-Cl (pH 7.0), 10 mM HEPES (pH 7.0), 0.5 mg/ml BSA, 0–5 mM free Ca2+ (in the form of CaCl ), 0.2 mM EGTA, 2 μM 9-amino-6-chloro-2-methoxyacridine (ACMA) and MgCl . EGTA was added to remove background Ca2+, and the concentration of total exogenous Ca2+ added in the assay solution was estimated through an online calculator (MAXCHELATOR: http://maxchelator.stanford.edu/CaMgATPEGTA-TS.htm) by inputting the designated free [Ca2+]. MgCl was added in the assay solution at the concentration of ([Ca2+] − [Ca2+] ) to balance the membrane potential change arising from the difference of internal and external [Ca2+]. The fluorescence (excitation: 410 nm; emission: 490 nm) was monitored in the infinite M1000 PRO plate reader (TECAN) every 30 s. After the fluorescence signal stabilized, K+ efflux was coupled to the influx of proton by adding 2 μM carbonyl cyanide m-chlorophenyl hydrazone (CCCP, a proton ionophore) into the assay buffer. The increase of proton concentration within the liposome led to protonation and fluorescence quenching of the ACMA dye. At the end of experiment, the K+-selective ionophore valinomycin was added at 2 μM final concentration to dissipate the potassium gradient. For the flux assay results shown in Fig. 3 and Extended Data Fig. 7, each set of the assays were conducted in parallel in a 96-well black plate (Costar), and each flux assay was repeated four to eight times. The fluorescence data were normalized to the mean value of the initial reading before being plotted against time. For preparing the SUV samples to be transformed into the giant unilamellar vesicles (GUV) for electrophysiology, a lipid mixture containing 90% 1,2-diphytanoyl-sn-glycero-3-phosphocholine (DPhPC, Avanti) and 10% cholesterol (w:w) was dried under vacuum for 4 h. The dry lipid sample was then suspended at 10 mg/ml concentration in a low-salt buffer with 1 mM HEPES-KOH (pH 7.2) and 5 mM KCl, and then subjected to tip sonication to clarity (50 Hz, 1 s on, 5 s off for 2.5 min). The sample was solubilized by adding β-DM to a final concentration of 10 mM, and incubated for 30 min at RT. The purified TRIC-B1(CΔ48) protein was then added to the solubilized lipid mixture to achieve a protein:lipid ratio of 1:100 (wt:wt). More β-DM was then added to reach a final concentration of 17.5 mM and the resulting mixture was gently agitated for 1 h at 25 °C. (Alternatively, Triton X-100 at a final concentration of 7.8 mM can be used to replace β-DM during the above sample-preparing steps.). The detergent was removed by dialyzing the protein–lipid mixture against a low-salt buffer (1 mM HEPES pH 7.2, 5 mM KCl) with the addition of 40 mg/ml washed SM2 biobeads (Bio-rad) in the dialysis buffer at 16 °C. The external buffer was changed every 12 h and the dialysis lasted for 6–7 days. After dialysis, the resulting SUV sample was aliquoted, flash-frozen in liquid nitrogen and then stored at −80 °C. The GUV samples used in the single-channel electrophysiological studies were prepared through the electroformation technique by using the Nanion Vesicle Prep Pro device (Nanion). To protect the protein during the partial dehydration process before electroformation, trehalose was added to the preformed SUV sample to a final concentration of 10 mM and then the mixture was sonicated in water bath for 30 s. Subsequently, about 10 μl SUV solution was pipetted in small droplets (∼0.2 μl/droplet) on the indium tin oxide (ITO)-coated glass slide. The droplets were exposed under room atmosphere for approximately 15 min to let them dry. After partial dehydration, the lipid films were rehydrated by adding 270 μl of 1 M sorbitol solution and then a cassette sandwiching the sample in the middle was assembled. The ITO layers of the slides face and contact the sample during the cassette assembling. The electroformation was first run at 0.1 to 1.0 V and 12 Hz frequency for 3 h. Afterwards, the frequency was lowered to 4 Hz and the voltage was increased to 2 V. The procedure was continued for 30 min to release the GUVs from the glass slides. Throughout the electroformation process, the temperature was maintained at 36 °C. The single-channel activity recordings were performed with the inside-out mode. The data were recorded under symmetrical solutions with 210 mM KCl and 10 mM HEPES (pH 7.2). The data were acquired at 50 kHz with a 0.5-kHz filter and 50 Hz notch filter, using an EPC-10 amplifier (HEKA). The Clampfit Version 10.0 (Axon Instruments) was used for data analysis, Excel Version 2010 (Microsoft) and OriginPro 8 were used for statistical analysis and Igor Pro 6.37A (WaveMetrics) was used for making the graphs. The single-channel conductance was obtained through linear fitting of the data recorded under different voltage values. The open probability data shown in Fig. 3i, j were based on P  = t /T, where t is the total time that the channel was observed in the open state at the ith level and T is the total recording time. Analyses on the potential cooperative/independent gating behaviours among the three pores within the TRIC-B1 trimer were performed according to the methods reported by Ding and Sachs40 (See Supplementary Discussion for details). For the cysteine accessibility assay experiments, the mutant proteins (M38C, A126C and S166C on a Cys-less template) were purified and reconstituted on the GUVs (protein:lipid = 1:200–250, w-w; lipid: 95% azolectin + 5% cholesterol, w-w) through a modified sucrose method41. The data were recorded on the inside-out patches from the reconstituted GUVs under ±40 mV at 50 kHz by using an EPC-10 amplifier (HEKA) with a 0.5-kHz filter and 50 Hz notch filter. The bath and pipette buffers both contained 210 mM KCl and 10 mM HEPES (pH 7.2). The experiments were performed at room temperature (21–24 °C). During the experiments, freshly prepared stock solution of MTSET (1 M in water) was added to the bath solution to a final concentration of 2 mM. The same patch of membrane containing active TRIC channels was used for collecting the data before and after MTSET was added. Tryptophan fluorescence spectra were measured on a Spectrofluorometer F7000 (Hitachi). All samples were excited at 295 nm and the fluorescence emission spectra were scanned in the range between 300–400 nm. During the measurements, 90° and 0° polarizers were used for excitation and emission, respectively. The slit size was 5 nm × 5 nm and the voltage was set at 650 V. To minimize the influence from background Ca2+ for the starting sample, EGTA was added to the TRIC-B1 or W180A mutant protein samples (at 6.9 μM protein concentration in 10 mM HEPES, 150 mM KCl, 0.3% β-DM, pH 7.5 buffer) at a final concentration of 10 mM. The addition of 10 mM EGTA in the sample reduced the background free Ca2+ to about 10 nM. The sample buffer with 10 mM EGTA was used as the blank, providing background data to be subtracted from the overall fluorescence spectroscopic data of the sample. Stock solution of 1 M CaCl was titrated into 200 μl protein sample to yield final [Ca2+] at 1 μM, 10 μM, 100 μM, 1 mM and 10 mM respectively. Experiments with the wild-type TRIC-B1 and W180A mutant were repeated 10 times. The difference spectra were derived by subtracting the data of the W180A mutant from the data of the wild-type sample. The range of fluorescence signals used for calculating the integrated intensity is from 305 to 345 nm. The fluorescence below 305 nm may include some reflection from incident light, while the background noise increases at wavelength larger then 345 nm. The formula used for normalization of the integrated fluorescence signals is I /I , where the fluorescence intensity under conditions with Ca2+ added (I ) was normalized to the sample with minimal Ca2+ (10 nM Ca2+ in the presence of 10 mM EGTA, I ). Excel Version 2010 (Microsoft), Origin 8.0 and Prism 6.0 (GraphPad) programs were used for statistical analysis and making the graphs. To verify the conformational change of M5–6 loop region, a double cysteine mutants, namely A49C/N185C, have been designed and used for disulphide cross-linking experiment under conditions with or without Ca2+. The two Cys mutation sites are located at the monomer–monomer interface of TRIC-B1. One of the mutation sites (N185C) is located on M5–6 loop, and the other (A49C) is in an invariable region adjacent to M5–6 loop. To eliminate the potential background signal from endogenous cysteine residues during cross-linking, all the endogenous Cys residues in TRIC-B1 were mutated to Ser, and then the A49C/N185C mutation was introduced to such a Cys-less template. The A49C/N185C mutant protein was expressed and purified according to the same protocol used for purifying wild-type TRIC-B1 protein. Before the cross-linking reactions were started, the protein sample was treated with 100 μM Tris(2-chloroethyl) phosphate (TCEP) for 10 min at room temperature to reset the Cys residues in a maximally reduced state. Subsequently, EGTA or CaCl was added to the sample at 10 mM final concentration and the sample was incubated for 10 min at room temperature. To induce cross-linking between two Cys residues, diamide was titrated into the samples at various concentrations (0, 0.1, 0.3 and 0.5 mM) and the reactions were conducted at 25 °C for 10 min. To quench the reaction, N-ethylmaleimide (NEM) was added to the sample at 50 mM final concentration to block the remaining free sulfhydryls. The samples were mixed with the non-reducing SDS–PAGE loading buffer containing 150 mM NEM and then the products of cross-linking reaction are separated through SDS–PAGE. The gels were stained with Coomassie brilliant blue R-250. As a control, A49C/N185C/W180A triple mutant was produced and cross-linked under the same conditions as the A49C/N185C mutant. Three repeats were performed for each mutant and the ImageJ was used for quantifying the intensities of bands on the SDS–PAGE gel. The Excel Version 2010 (Microsoft) and Prism 6 (GraphPad) programs were used for analysing and plotting the data. To analyse the oligomeric state of TRIC-B1 protein (wild type and K129A/R133L mutant), chemical cross-linking was performed with the purified protein sample in detergent solution or directly with those on the cellular membranes. For cross-linking of purified protein in detergent solution, the wild-type or K129A/R133L mutant TRIC-B1 protein was diluted to 1 mg/ml in a reaction buffer consisting of 10 mM HEPES (pH 7.5), 150 mM KCl and 0.3% β-DM. The protein was cross-linked by 0–10 mM glutaraldehyde for 20 min at room temperature and the reactions were stopped by adding 100 mM TRIS-HCl (pH 7.5). The cross-linked samples were loaded on SDS–PAGE and the gels were stained with Coomassie brilliant blue R250. For cross-linking of target proteins on the membrane, the cells containing wild-type or K129A/R133L mutant protein were re-suspended in a buffer consisting of 50 mM TRIS-HCl, pH 8.0, and 150 mM KCl. After the cells were lysed by passing through high-pressure homogenizer (ATS Engineering), the cell debris was removed through low-speed centrifugation at 10,000g for 30 min and the membrane fraction was further collected through ultracentrifugation at 100,000g for 1 h at 4 °C. The membrane pellets were re-suspended in a buffer consisting of 100 mM HEPES (pH 7.5) and 150 mM KCl, and then treated with disuccinimidyl suberate (DSS) at 0–10 mM final concentration for 30 min at room temperature. The reactions were quenched by adding 100 mM TRIS-HCl (pH 7.5). The cross-linked samples were solubilized by adding 5 × SDS–PAGE loading buffer, and then loaded on SDS–PAGE. The target protein bands were detected through western blot by using the HRP-conjugated His-tag antibody (Genscript). After being developed with the western lightning Ultra ECL horseradish peroxidase substrate (Perkin–Elmer), the western blots were imaged on a chemiluminescence CCD system (ChemiScope 3500 mini imager, Clinx Science Instruments).


News Article | November 4, 2015
Site: www.nature.com

No statistical methods were used to predetermine sample size, the experiments were not randomized and the investigators were not blinded to allocation during experiments and outcome assessment. The pcDNA3.1-Piezo1-pp (PPase, PreScission protease cleavage site) -GST-IRES-GFP construct was subcloned by inserting the coding sequence of the PreScission protease cleavage site between Piezo1 (E2JF22, UniprotKB entry) and GST coding sequences in the parental construct of pcDNA3.1-Piezo1-GST-IRES-GFP3. Piezo1-Cterm-Flag-IRES-GFP was subcloned by inserting the synthesized double-stranded DNA fragment encoding Flag between the Piezo1-coding sequence and IRES-GFP using the restriction enzymes AscI and SacII. Piezo1(A2419)-Flag-IRES-GFP was constructed using a one-step cloning kit (Vazyme Biotech) by introducing the Flag-tag coding sequence after the residue Piezo1(A2419) into the Piezo1-GST-IRES-GFP construct and the Piezo1(Δ2219–2453) construct was generated by deleting amino acids 2219–2453 from the Piezo1-pp-GST-IRES-GFP construct. The coding sequence of the CED of Piezo1 (residues 2214–2457) was cloned into a pET22b (Novagen) vector with a C-terminal 6×His tag using the restriction enzymes NdeI and XhoI. All the constructs were confirmed by sequencing. HEK293T cells were grown in DMEM (basic) with 10% FBS. When the density of cells cultured in 150 mm × 25 mm dishes reached 80–90%, the expression plasmids were transiently transfected with polyethylenimines (Polysciences). The protein purification procedure was slightly modified from similar previously described methods3. After 48 h, the transfected cells were collected, washed twice with PBS and homogenized in buffer A, containing 25 mM Na-PIPES, pH 7.2, 140 mM NaCl, 2 mM dithiothreitol (DTT), detergents CHAPS (1%) and C12E9 (0.1%), 0.5% (w/v) l-α-phosphatidylcholine (Avanti) and a cocktail of protease inhibitors (Roche) at 4 °C for 1 h. After centrifugation at 100,000g for 40 min, the supernatant was collected and incubated with glutathione–sepharose beads (GE Healthcare) at 4 °C for 3 h. The resin was washed extensively with buffer B, containing 25 mM Na-PIPES, pH 7.2, 140 mM NaCl, 2 mM DTT, 0.1% (w/v) C12E9 and 0.01% (w/v) l-α-phosphatidylcholine. The GST-free or GST-tagged Piezo1 was cleaved off by PreScission Protease (Amersham-GE) in buffer B at 4 °C overnight or directly eluted from the protein-loaded resin with buffer B plus 10 mM GSH, respectively, and applied to size-exclusion chromatography (Superpose-6 10/300 GL, GE Healthcare) in buffer C (25 mM Na-PIPES, pH 7.2, 140 mM NaCl, 2 mM DTT) plus 0.026% (w/v) C12E10 or other detergents in the final concentration of 2× critical micelle concentration. For amphipol-bound Piezo1, amphipols were substituted for detergents as described34, after which the protein was loaded on a Superpose-6 column in buffer C. Proteins with different kinds of detergents or amphipols were examined by both gel filtration and negative staining. Peak fractions representing oligomeric Piezo1 were collected for electron microscopy analysis. Protein in C12E10 was used for final cryo-electron microscopy structure determination. All detergents and amphipols used in this project were purchased from Anatrace. Overexpression of Piezo1 CED was induced in Escherichia coli BL21 strain by 0.5 mM isopropyl-β-d-thiogalactoside when the cell density reached an optical density of ~0.8 at 600 nm. After growing at 18 °C for 12 h, the cells were collected, washed, resuspended in buffer D, containing 25 mM Tris-HCl, pH 8.0, 500 mM NaCl and 20 mM imidazole, and lysed by sonication. The lysates were clarified by centrifugation at 23,000g for 1 h and the supernatant was collected and loaded onto Ni2+-nitrilotriacetate affinity resin (Ni-NTA, Qiagen). The resin was washed extensively with buffer D and eluted with buffer D plus 280 mM imidazole. The eluate was concentrated and subjected to gel filtration (Superdex-200, GE Healthcare) with buffer E, containing 25 mM Tris-HCl, pH 8.0, 200 mM NaCl, 2 mM DTT, or buffer F, containing 25 mM Tris-HCl, pH 8.0, 25 mM NaCl and 2 mM DTT (Extended Data Fig. 6e). The purified Piezo1 proteins were subjected to 3–12% NativePAGE Novex Bis-Tris gel for native electrophoresis according to the manufacturer’s protocol at 150 V for 2 h. The native gel was transferred to a positively charged nylon/nitrocellulose membrane at 100 V for 1.5 h. After incubating in 8% (v/v) acetic acid to fix the proteins, air drying and rewetting with methanol, the membrane was blocked with 5% (w/v) milk in TBS buffer with 0.1% (w/v) Tween-20 (TBST buffer) at room temperature (~26 °C) for 1 h. The membrane was then incubated with the anti-Piezo1 antibody (1:1,000) (custom generated using the peptide YIRAPNGPEANPVK) at room temperature for 1 h, followed by washing with TBST buffer and further incubated with anti-rabbit IgG antibody (1:10,000) at room temperature for 1 h. Proteins were detected with the SuperSignal West Pico Chemiluminescent Substrate (Thermo). For live-cell labelling, cells grown on coverslips were incubated with the anti-Flag antibody (1:100, Sigma) diluted in prewarmed culture medium at room temperature for 1 h. After three washes, cells were incubated with the Alexa Fluor 594 donkey-anti-mouse IgG secondary antibody (1:200, Life Technologies) at room temperature for 1 h and then washed and fixed with 4% (w/v) paraformaldehyde. For permeabilized staining, cells were first fixed with 4% (w/v) paraformaldehyde and permeabilized with 0.2% (w/v) Triton X-100, then incubated with the anti-Flag antibody (1:200, Sigma) or the anti-GST antibody (1:200, Millipore) at room temperature for 1 h. Cells were washed and then incubated with the Alexa Fluor 594 donkey-anti-mouse IgG (1:200, Life Technologies) or Alexa Fluor 594 donkey-anti-rabbit IgG (1:200, Life Technologies) secondary antibody at room temperature for 1 h. After washing, coverslips were mounted and imaged using a Nikon A1 confocal microscope with a 60× oil objective (N.A. = 1.49) at either the GFP (488-nm exciting wavelength) or the TRITC channel (561-nm exciting wavelength). Crystals of CED proteins were obtained at 18 °C using the sitting-drop method by mixing 1 μl protein (15 mg ml−1) with 1 μl reservoir solution (0.1 M HEPES, pH 7.5, 0.2 M MgCl and 25% w/v PEG3350). Crystals appeared after 2–3 weeks and reached full size in about a month. The crystals were cryo-protected in reservoir solution containing 15–20% glycerol and flash frozen in liquid nitrogen before data collection. Native data of CED crystals were collected at beamline BL17U of the Shanghai Synchrotron Radiation Facility (SSRF). Single-wavelength anomalous dispersion data were collected at 100 K using a MAResearch M165 charge-coupled device (CCD) detector at the Beijing Synchrotron Radiation Facility (BSRF), with the crystals soaked in 2 M NaI for 1 min. All diffraction data were processed with HKL2000 (ref. 49). Further processing was carried out using programs from the CCP4 suite (Collaborative Computational Project)50. The heavy-atom positions in the iodine-soaked crystal were determined using SHELXD51. Heavy-atom parameters were then refined and initial phases were generated in the program PHASER52 using the single-wavelength anomalous dispersion experimental phasing module. The real-space constraints were applied to the electron density map in DM53. The resulting map was of sufficient quality for building the model of the CED in Coot54. The structures were refined with the PHENIX packages55. Full data collection and structure statistics are summarized in Extended Data Table 1. An aliquot of 4 μl Piezo1 (0.05 mg ml−1) was applied to glow-discharged carbon-coated copper grids (200 mesh, Zhongjingkeyi, Beijing). After the grids were incubated at room temperature for 1 min, excessive liquid was absorbed by filter paper. Grids containing the specimen were stained by applying droplets of 2% uranyl acetate for 30 s and air dried. Micrographs were generated on a T12 microscope (FEI) operated at 120 kV, using a 4k × 4k CCD camera (UltraScan 4000, Gatan). Images of Piezo1 purified with C12E10, C12E8 and amphipol A8-35 were recorded at a nominal magnification of 68,000× and with a pixel size of 1.59 Å (Extended Data Fig. 2). Images of Piezo1(ΔCED) in C12E10 were recorded at a nominal magnification of 49,000× and with a pixel size of 2.21 Å. Micrographs of random conical tilt (RCT) pairs were taken at 50° and 0° tilt angles at a nominal magnification of 49,000×. The detergent C12E10 was chosen for cryo-electron microscopy analysis because it produced slightly better micrographs (Extended Data Fig. 2). Aliquots of 4 μl detergent-solubilized (C12E10) Piezo1 at a concentration of 0.2 mg ml−1 were applied to glow-discharged 300-mesh Quantifoil R2/2 grids (Quantifoil, Micro Tools GmbH, Germany) coated with a self-made continuous thin carbon. After 15 s of waiting time, grids were blotted for 1.5 s and plunged into liquid ethane using an FEI Mark IV Vitrobot operated at 4 °C and 100% humidity. Grids were examined using a TF20 microscope (FEI) operated at 200 kV with a nominal magnification of 62,000× and images were captured on a CCD camera (Gatan) under low-dose conditions. High-resolution images were captured on a Titan Krios microscope, operated at 300 kV, with a K2 Summit direct electron detector (Gatan) in counting mode. Data acquisition was performed using UCSF-Image4 (X. Li and Y. Cheng), with a nominal magnification of 22,500×, which yields a final pixel size of 1.32 Å at object scale and with defocus ranging from –1.7 μm to –2.9 μm. The dose rate on the detector was about 8.2 counts per pixel per second, with a total exposure time of 8 s. Each micrograph stack consists of 32 frames. The data sets of negative-staining electron microscopy were processed with EMAN2.1 (ref. 56) and RELION57. Reference-free 2D classification was performed with RELION. The numbers of Piezo1 particles in the presence of C12E10, C12E8 and amphipol A8-35 are 7,279, 14,045 and 7,565, respectively. For RCT58 data processing, particle picking and classification were performed with EMAN2.1 (ref. 56) and reconstruction of RCT classes and structural refinement from all untilted particles were performed with SPIDER59. The final number of particles used in generating the initial model is 5,670. The initial 3D reference created using the RCT method is shown in Extended Data Fig. 3. For cryo-electron microscopy (TF20) data processing, 505 micrographs were processed with SPIDER59 and RELION57. Particles were picked using SPIDER, manually screened (39,555 in total) and subjected to reference-free 2D classification using RELION. A final number of 16,729 particles were used for 3D refinement using the RCT model as initial reference. To validate the 3D model, 3D refinement was also performed with a Gaussian density ball as initial reference. During refinement, both the symmetry-free (C1) and symmetry-imposed (C3) reconstructions were tested (Extended Data Fig. 3d). For processing K2 micrographs, motion correction was applied at the micrograph level using the dosefgpu_driftcorr program (developed by X. Li) to produce average micrographs across all frames60. Micrograph screening, particle picking and normalization were performed with SPIDER. The program CTFFIND3 (ref. 61) was used to estimate the contrast transfer function parameters. The 2D and 3D classification and refinement were performed with RELION exclusively to avoid potential structural overfitting. Classification of raw cryo-electron microscopy particles resulted in well-resolved 2D class averages, with many secondary structural features clearly discernable. In particular, on class averages of typical side views, many pieces of rod-like densities arranged in parallel fashion could be readily identified, raising the possibility that they were transmembrane helices (Fig. 2c). A total of 179,805 particles from 1,042 micrographs were subject to a cascade of 2D and 3D classification (Extended Data Fig. 5a). During 3D classification, no symmetry was imposed. Different combinations of particles from these classes were tested in refinement. After two rounds of 3D classification, a set of adequately homogeneous particles (30,021), which best matched the C3 symmetry, was subjected to a third round of 3D classification. This resulted in generally similar class structures, with no detectable improvement on particle homogeneity. Consequently, this set of particles was used for final refinement, with the RCT model low-pass filtered to 60 Å as initial reference. Applying the C3 symmetry in the refinement resulted in an overall structure at a resolution of 10.24 Å. After the first refinement, we noted that translation parameters of particles (OriginX and OriginY in RELION) were rather large, with many particles having x or y shifts of more than 15 pixels. Particles were rewindowed from original micrographs by applying their x and y shifts. Rewindowed particles were subjected to a second round of refinement using RELION, which only marginally improved the density map. A third round of refinement was performed by applying an enlarged soft mask (Extended Data Fig. 5a) of the Piezo1 channel, which improved the overall resolution to 6.03 Å. Last, particle-based beam-induced movement correction was performed by statistical movie processing in RELION, using movie frames 2–15. This yielded a final 3D density map with an overall resolution of 5.9 Å, with regions defined by the soft mask being 4.8 Å (Extended Data Fig. 5b). All reported resolutions are based on the gold-standard FSC = 0.143 (ref. 62) and the final FSC curve (4.8 Å) was corrected for the effect of a soft mask using high-resolution noise substitution63. In addition, subregion refinements, as previously described for ribosomal complex structural determination64, 65, 66, 67, were applied to improve the local densities of interest, by using a soft mask of the cap domain, the lower central pore region and a single subunit. The subsequent reported resolutions were still in the range of 4.8–5.5 Å, but with much-improved densities for these masked regions. This led to a separation of secondary structural elements in the cap and transmembrane regions. However, in all cases, the densities at the distal ‘blade’ domain are fragmented and limited our further quantitative analysis. Final density maps were sharpened by a B-factor of –100 Å2 using RELION. A local resolution map was calculated using ResMap68. UCSF Chimera69 was used to fit the crystal structure of the CED to the density map of the cap domain. Main-chain tracing and building a poly-alanine model were done manually using Coot70. Sequence alignment was performed using Clustal W2 (ref. 71). Secondary structures were predicted with PredictProtein72 using the full-length Piezo1 sequence. Transmembrane segments were predicted using multiple prediction web servers, including Topcons73, TMHMM2 (ref. 74), HMMTOP75 and Phobius76, with their results shown as green, blue, orange and pink lines, respectively, in Extended Data Fig. 9. Sequence alignment and secondary structure prediction of Piezo1 from different species were used to aid the assignment of structural elements in the density map. Multiple rounds of model rebuilding in Coot were performed for model optimization.


Shen C.,Zhejiang University | Chen H.,Zhejiang University | Wu S.,Zhejiang University | Wen Y.,Zhejiang University | And 4 more authors.
Journal of Hazardous Materials | Year: 2013

Metal-biopolymer complexes has recently gained significant attention as an effective adsorbent used for the removal of Cr(VI) from water. Unfortunately, despite increasing research efforts in the field of removal efficiency, whether this kind of complex can reduce Cr(VI) to less-toxic Cr(III) and what are the mechanisms of detoxification processes are still unknown. In this study, despite the highly adsorption efficiency (maximum adsorption capacity of 173.1. mg/g in 10. min), the significant improvement of Cr(VI) reduction by chitosan-Fe(III) complex compared with normal crosslinked chitoan has been demonstrated. In addition, the structure of chitosan-Fe(III) complex and its functional groups concerned with Cr(VI) detoxification have been characterized by the powerful spectroscopic techniques X-ray absorption fine structure (XAFS) and X-ray photoelectron spectroscopy (XPS). The XPS spectra indicated that the primary alcoholic function on C-6 served as an electron donor during Cr(VI) reduction and was oxidized to a carbonyl group. The X-ray adsorption near edge spectra (XANES) of the Cr(VI)-treated chitosan-Fe(III) complex revealed the similar geometrical arrangement of Cr species as that in Cr(III)-bound chitosan-Fe(III). Overall, a possible process and mechanism for highly efficient detoxification of Cr(VI) by chitosan-Fe(III) complex has been elucidate. © 2012 Elsevier B.V.


Li C.,Yantai Wanhua Polyurethanes Company | Han J.,East China University of Science and Technology | Huang Q.,Yantai Wanhua Polyurethanes Company | Xu H.,Yantai Wanhua Polyurethanes Company | And 2 more authors.
Polymer | Year: 2012

The microstructure development of 4,4′-diphenylmethane diisocyanate (MDI) and 1,4-butanediol (BDO) based thermoplastic polyurethane (TPU) under compression was investigated. Influential factors on the permanent compression deformation were discussed in detail. Two types of samples with the same chemical compositions while different aggregation structures were prepared by altering the stirring speed during the synthesis process. The difference in the first-order structure of these samples was studied by 13C NMR. The aggregation structures and the corresponding development under compression were characterized by DSC, WAXD and SAXS, respectively. These observed results indicated that difference of microstructures lead to different deformation behaviors, and the disruption of hard segment domains or reorganized structures by less ordered hard segments played significant roles in the permanent deformation under compression. The viscoelastic behaviors of these samples were described by DMA and simulated by Rouse model. The derived terminal viscosity and relaxation times were used to explain the different permanent deformations of these samples. Finally, an optimized micro-crosslink structure was introduced in TPUs, and a better deformation resistance property was obtained. © 2012 Elsevier Ltd. All rights reserved.


He Z.,Jiangsu University | Zhong W.,Jiangsu University | Wang Q.,Jiangsu University | Jiang Z.,Jiangsu University | Fu Y.,Shanghai Synchrotron Radiation Facility
Applied Thermal Engineering | Year: 2013

In diesel engines, the cavitating flow in nozzles greatly affects the fuel atomization characteristics and then the subsequent combustion and exhaust emissions. In this paper, with the needle lift curve on the basis of injection rate experimental data, a moving mesh generation strategy was applied for 3D simulation of the nozzle cavitating flow. Based on the third-generation synchrotrons of Shanghai Synchrotron Radiation facility (SSRF), a high-precision three-dimension structure of testing nozzle with detailed internal geometry information was obtained using X-ray radiography for a more accurate simulation. A flow visualization experiment system with a transparent scaled-up vertical multi-hole injector nozzle tip was setup. The experimental data was obtained to make a comparison to validate the calculated results and good qualitative agreement was shown between them. Afterward, the effects of needle movement on development of the cavitating flow and flow characteristics parameters were investigated. Finally, the influence of fuel temperature on development of the cavitating flow was also studied. Research of the flow characteristics for the diesel and biodiesel revealed that the flow characteristics of the biodiesel with a temperature rise of between 50 K and 60 K in injector nozzles will be similar to those of the diesel fuel. © 2013 Elsevier Ltd. All rights reserved.


Hu X.,University of Chinese Academy of Sciences | Wang L.,University of Chinese Academy of Sciences | Xu F.,University of Chinese Academy of Sciences | Xiao T.,Shanghai Synchrotron Radiation Facility | Zhang Z.,CAS National Center for Nanoscience and Technology
Carbon | Year: 2014

In contrast with traditional methods of observation, synchrotron radiation X-ray computed tomography (SR-CT) is an advanced technique that allows direct three-dimensional (3D) and non-destructive observation of microstructures in materials. High-resolution in situ observations (0.7 μm/pixel) of fractures in short carbon fiber/epoxy composites are achieved using the SR-CT technique, and the mechanical load response of short carbon fibers treated with oxidation and those untreated are compared. By the quantitative extraction and analysis of microstructure parameters in high-resolution 3D images, the failure mechanisms of the two materials were studied. The proportion of broken fibers to other types of fiber damage in the sample with oxidation-treated fibers increases by about 6%. Also, the oxidation treatment is able to reduce the ineffective length of the fibers by about 20%, thereby improving the mechanical properties of these composites. The results show that computed tomography can promote characterization of the internal microstructures in carbon fiber-reinforced polymer composites, which will facilitate further theoretical research on the failure mechanisms of these composites. © 2013 Elsevier Ltd. All rights reserved.


News Article | October 10, 2016
Site: www.spie.org

Advantages such as high resolution and throughput, along with insensitivity to misalignment, make extreme-UV interference lithography a powerful enabling technology for academic and industrial research. Extreme-UV interference lithography (EUV-IL), at a wavelength of 13.5nm, has proved to be a powerful technique due both to its relative simplicity and record-high-resolution patterning capabilities. In an EUV-IL setup, a mask with transmission-diffraction gratings is illuminated by a spatially coherent beam of EUV light from an undulator synchrotron source. Periodic images are then created by the interference of two or more diffracted coherent beams (see Figure 1). The sinusoidal aerial image produced by two-beam IL has a period that is half of the original mask grating period when first-order diffracted beams are used. Consequently, this method provides a demagnification of grating patterns that are written by electron beam lithography (EBL). Moreover, versatile periodic patterns and quasi-periodic patterns can be obtained by using multiple beams and by controlling their phases.1, 2 The ultimate resolution (i.e., half-pitch, HP) that can be achieved with EUV-IL, however, is limited by light diffraction. Although it is therefore theoretically possible to resolve features to below 4nm with EUV-IL,3 the achievable resolution is limited by the grating resolution and by other factors. Figure 1. Schematic diagram of an extreme-UV interference lithography (EUV-IL) setup, in which first-order diffraction is used to create an aerial image on a resist, by interference. θ, θ , θ : Diffraction angles. m : Diffraction orders. (Adapted from Mojarad et al., 2015. Schematic diagram of an extreme-UV interference lithography (EUV-IL) setup, in which first-order diffraction is used to create an aerial image on a resist, by interference. θ, θ, θ: Diffraction angles. m: Diffraction orders. (Adapted from Mojarad et al., 2015. 7 In addition to high resolution and throughput, EUV-IL offers many other advantages, such as achromaticity, insensitivity to misalignment, and infinite depth of focus, making this lithography technique extremely useful.4 For example, with EUV-IL, it is possible to pattern high-resolution periodic images to create highly ordered and dense nanostructures. Such structures can be difficult or time-consuming to pattern by EBL, but are interesting for a wide range of applications, such as nanocatalysis,5 electronic6 and photonic devices,7 and fundamental materials analysis.8 In this work,9 we discuss the long-term performance and capabilities of EUV-IL activities at the Paul Scherrer Institute (PSI), Switzerland. We also describe some of the state-of-the-art research that has been conducted at PSI. In particular, we focus on the X-ray Interference Lithography (XIL) beamline at PSI, which is illuminated by the co-located Swiss Light Source. We have been able to achieve resolution down to 10nm HP by using masks that consist of silicon oxide diffraction gratings.10, 11 These gratings are directly written on silicon nitride membranes (<100nm thick), via EBL from hydrogen silsesquioxane. Below 10nm HP, however, the use of silicon oxide gratings drastically reduces the EUV diffraction efficiency (≪1%).7, 12 Record-high, sub-10nm resolution dense patterning down to 7nm HP has nevertheless been accomplished by EUV-IL at PSI through the use of novel resists based on tin oxide and hafnium oxide (see Figure 2).3,13 These negative-tone, metal-oxide-based resist materials can easily be spin-coated and directly patterned when they are exposed to an electron or photon beam. This enables the direct fabrication of gratings with high diffraction efficiencies, which are theoretically sufficient to pattern images with a resolution down to 5nm HP (∼3%).3, 10 Furthermore, using a period-doubling fabrication method, we produced a mask with iridium gratings (high diffraction efficiency) capable of lithographically patterning an impressive 6nm HP: see Figure 2(d) and (e).14 We have also used a pattern transfer onto a silicon-on-insulator substrate to fabricate well-ordered arrays of ultrathin, suspended silicon nanowires down to 6.5nm linewidths: see Figure 2(f). We plan to study these arrays as field-effect transistors, biosensors, or thermoelectric devices. Figure 2. Scanning electron microscropy (SEM) images of line/space (L/S) patterns with 9–6nm half-pitch (HP) resolution. These patterns were obtained by EUV-IL on hydrogen silsesquioxane (with a high-resolution negative-tone electron beam or EUV resist), and with the use of (a, b) tin oxide and (c) hafnium oxide gratings directly patterned by electron beam lithography, and (d, e) iridium gratings fabricated by a pitch-doubling method. Scanning electron microscropy (SEM) images of line/space (L/S) patterns with 9–6nm half-pitch (HP) resolution. These patterns were obtained by EUV-IL on hydrogen silsesquioxane (with a high-resolution negative-tone electron beam or EUV resist), and with the use of (a, b) tin oxide and (c) hafnium oxide gratings directly patterned by electron beam lithography, and (d, e) iridium gratings fabricated by a pitch-doubling method. 14 (f) Top-view SEM micrographs of silicon nanowires (SiNWs), with 6.5nm linewidths, directly transferred by reactive ion etching from patterns created by EUV-IL and thinned down by performing consecutive cycles of piranha solution and buffered hydrofluoric acid dips. V1, V2: Indicators of scale. Another application of EUV-IL, owing to its high resolution and relatively high throughput, is the development of model systems with precise nanoparticle size and position over large areas. In particular, we have shown that EUV achromatic Talbot lithography (ATL)—another simple but powerful interference scheme that uses a single grating and is based on the Talbot effect—yields well-defined dot sizes.15, 16 Using step-and-repeat EUV-ATL exposures, we obtained 15nm dots that had a pitch of 100nm and were spread over an area of several square centimeters with high uniformity.5 This IL technique offers substantial opportunities in the field of nanocatalysis, where single-particle-level studies are necessary to gain deeper insight into chemical mechanisms.17 We have also developed a novel method for the direct measurement of the absorption and Dill parameters of photoresists at the EUV wavelength of 13.5nm.18 These are key parameters needed for the accurate simulation and modeling of the dose response of a photoresist, but that are not widely studied because of the lack of bright EUV sources. EUV-IL is also extensively used for EUV photoresist testing and development, enabling research before industrial exposure tools are available.11, 19,20 The well-defined and high-resolution aerial imaging of EUV-IL, as well as its relatively low cost, make it an essential tool for the evaluation of EUV photoresists and their timely development. This is particularly important today because it is predicted (according to the semiconductor industry roadmap) that the introduction of EUV lithography (EUVL) to high-volume manufacturing, at the 7nm node, will occur in the near future.21 We have thus evaluated hundreds of state-of-the-art EUV materials from different vendors all around the world.11, 20,22–24 Among these, we have identified several promising chemically amplified resist (CAR) candidates for future testing. These candidates simultaneously meet sensitivity (best energy), linewidth roughness, and exposure latitude requirements for the successful introduction of EUVL into high-volume manufacturing at the 7nm logic node (16nm HP).19, 22 Although several resists at 13nm HP (5nm node) and below are well resolved, pattern collapse and pinching still limit the exposure latitude of CARs. We show a high-performance CAR, however, in Figure 3. This example is well resolved down to 12 and 11nm HP with minimal pattern collapse and bridging. Although this CAR example is remarkable, it is clear that there is still an impending need for alternative resist solutions at 13nm resolution and below. Figure 3. SEM L/S images showing one of the highest-performing chemically amplified resists at (a) 12nm and (b) 11nm HP resolution. Minimal pattern collapse and bridging are observed. (Adapted from Buitrago et al., 2016. SEM L/S images showing one of the highest-performing chemically amplified resists at (a) 12nm and (b) 11nm HP resolution. Minimal pattern collapse and bridging are observed. (Adapted from Buitrago et al., 2016. 22 In summary, we have presented a number of examples that showcase EUV-IL as a powerful enabling technology for academic and industrial research. We believe that, in the future, the interest in EUV-IL will increase as EUVL enters into the high-volume-manufacturing phase. Consequently, we will continue working with industry to develop the resist technology for the 5nm node and beyond. We will also continue to enable the fabrication of high-resolution periodic structures that are only possible with EUV-IL. Paul Scherrer Institute (PSI) Elizabeth Buitrago is a research scientist at PSI working in collaboration with ASML on the extension and development of EUV resists using the EUV-IL system at the Swiss Light Source. During her PhD at the École Polytechnique Fédéral de Lausanne, Switzerland, she developed a 3D vertically stacked silicon nanowire–field-effect transistor for attomolar sensing of biotin. 3. N. Mojarad, M. Hojeij, L. Wang, J. Gobrecht, Y. Ekinci, Single-digit-resolution nanopatterning with extreme ultraviolet light for the 2.5 nm technology node and beyond, Nanoscale 7, p. 4031-4037, 2015. 4. V. Auzelyte, C. Dais, P. Farquet, D. Grützmacher, L. J. Heyderman, F. Luo, S. Olliges, et al., Extreme ultraviolet interference lithography at the Paul Scherrer Institut, J. Micro/Nanolith. MEMS MOEMS 8, p. 021204, 2009. doi:10.1117/1.3116559 7. N. Mojarad, J. Gobrecht, Y. Ekinci, Interference lithography at EUV and soft x-ray wavelengths: principles, methods, and applications, Microelectron. Eng. 143, p. 55-63, 2015. 8. R. Fallica, J. K. Stowers, A. Grenville, A. Frommhold, A. P. G. Robinson, Y. Ekinci, Dynamic absorption coefficients of CAR and non-CAR resists at EUV, Proc. SPIE 9776, p. 977612, 2016. doi:10.1117/12.2219193 9. E. Buitrago, R. Fallica, D. Fan, M. Vockenhuber, Y. Ekinci, From powerful industrial platform for EUV photoresist development, to world record resolution by photolithography: EUV interference lithography at the Paul Scherrer Institute, Proc. SPIE 9926, p. 9926OT, 2016. doi:10.1117/12.2238805 10. E. Buitrago, R. Fallica, D. Fan, T. S. Kulmala, M. Vockenhuber, Y. Ekinci, SnOx high-efficiency EUV interference lithography gratings towards the ultimate resolution in photolithography, Microelectron. Eng. 155, p. 44-49, 2016. 11. E. Buitrago, O. Yildirim, R. Fallica, et al., The road towards single digit nanometer resolution patterning in mass production: state-of-the-art EUV resists platforms compared, Proc. Int'l Symp. EUVL, 2015. 12. S. Prezioso, P. De Marco, P. Zuppella, S. Santucci, L. Ottaviano, A study of the mechanical vibrations of a table-top extreme ultraviolet interference nanolithography tool, Rev. Sci. Instrum. 81, p. 045110, 2010. 15. D. Fan, E. Buitrago, S. Yang, W. Karim, Y. Wu, R. Tai, Y. Ekinci, Patterning of nanodot-arrays using EUV achromatic Talbot lithography at the Swiss Light Source and Shanghai Synchrotron Radiation Facility, Microelectron. Eng. 155, p. 55-60, 2016. 17. W. Karim, A. Kleibert, U. Hartfelder, A. Balan, J. Gobrecht, J. A. van Bokhoven, Y. Ekinci, et al., Size-dependent redox behavior of iron observed by in-situ single nanoparticle spectro-microscopy on well-defined model systems, Sci. Rep. 6, p. 18818, 2016. 18. R. Fallica, J. K. Stowers, M. Vockenhuber, Absorption coefficient and Dill parameters of CAR and non-CAR resists at EUV, Proc. Int'l Symp. EUVL, p. P-RE, 2015. 19. E. Buitrago, S. Nagahara, O. Yildirim, H. Nakagawa, S. Tagawa, M. Meeuwissen, T. Nagai, et al., Sensitivity enhancement of chemically amplified resists and performance study using EUV interference lithography, Proc. SPIE 9776, p. 97760Z, 2016. doi:10.1117/12.2220026 20. E. Buitrago, O. Yildirim, C. Verspaget, N. Tsugama, R. Hoefnagels, G. Rispens, Y. Ekinci, Evaluation of EUV resist performance using interference lithography, Proc. SPIE 9422, p. 94221S, 2015. doi:10.1117/12.2085803 22. E. Buitrago, S. Nagahara, O. Yildirim, H. Nakagawa, S. Tagawa, M. Meeuwissen, T. Nagai, et al., Sensitivity enhancement of chemically amplified resists and performance study using extreme ultraviolet interference lithography, J. Micro/Nanolith. MEMS MOEMS 15, p. 033502, 2016. doi:10.1117/1.JMM.15.3.033502 23. T. S. Kulmala, M. Vockenhuber, E. Buitrago, R. Fallica, Y. Ekinci, Toward 10 nm half-pitch in extreme ultraviolet lithography: results on resist screening and pattern collapse mitigation techniques, J. Micro/Nanolith. MEMS MOEMS 14, p. 033507, 2015. doi:10.1117/1.JMM.14.3.033507

Loading Shanghai Synchrotron Radiation Facility collaborators
Loading Shanghai Synchrotron Radiation Facility collaborators