[en] The High Luminosity Upgrade of the Large Hadron Collider (HL-LHC) at CERN will replace the main ring inner triplet quadrupoles, identified by the acronym MQXF, adjacent to the main ring intersection regions. For the past decade, the U.S. LHC Accelerator R&D Program, LARP, has been evaluating conductors for the MQXFA prototypes, which are the outer magnets of the triplet. Recently, the requirements for MQXF magnets and cables have been published in P. Ferracin et al., IEEE Trans. Appl. Supercond., vol. 26, no. 4, 2016, Art. no.4000207, along with the final specification for Ti-alloyed Nb3Sn conductor determined jointly by CERN and LARP. This paper describes the rationale beneath the 0.85 mm diameter strand’s chief parameters, which are 108 or more sub-elements, a copper fraction not less than 52.4%, strand critical current at 4.22 K not less than 631 A at 12 T and 331 A at 15 T, and residual resistance ratio of not less than 150. This paper also compares the performance for ~100 km production lots of the five most recent LARP conductors to the first 163 km of strand made according to the HL-LHC specification. Two factors emerge as significant for optimizing performance and minimizing risk: a modest increase of the sub-element diameter from 50 to 55 μm, and a Nb:Sn molar ratio of 3.6 instead of 3.4. Furthermore, the statistics acquired so far give confidence that the present conductor can balance competing demands in production for the HL-LHC project.

[en] A model for the onset of the reduction in superconducting radio-frequency (SRF) cavity quality factor, the so-called Q-drop, at high accelerating electric fields is presented. Since magnetic fields at the cavity equator are tied to accelerating electric fields by a simple geometric factor, the onset of magnetic flux penetration determines the onset of Q-drop. We consider breakdown of the surface barrier at triangular grooves to predict the magnetic field of first flux penetration H_pen. Such defects were argued to be the worst case by Buzdin and Daumens (1998 Physica C 294 257), whose approach, moreover, incorporates both the geometry of the groove and local contamination via the Ginzburg-Landau parameter κ. Since previous Q-drop models focused on either topography or contamination alone, the proposed model allows new comparisons of one effect in relation to the other. The model predicts equivalent reduction of H_pen when either roughness or contamination were varied alone, so smooth but dirty surfaces limit cavity performance about as much as rough but clean surfaces do. Still lower H_pen was predicted when both effects were combined, i.e. contamination should exacerbate the negative effects of roughness and vice versa. To test the model with actual data, coupons were prepared by buffered chemical polishing and electropolishing, and stylus profilometry was used to obtain distributions of angles. From these data, curves for surface resistance generated by simple flux flow as a function of magnetic field were generated by integrating over the distribution of angles for reasonable values of κ. This showed that combined effects of roughness and contamination indeed reduce the Q-drop onset field by ∼ 20%, and that contamination contributes to Q-drop as much as roughness. The latter point may be overlooked by SRF cavity research, since access to the cavity interior by spectroscopy tools is very difficult, whereas optical images have become commonplace. The model was extended to fit cavity test data, which indicated that reduction of the superconducting gap by contaminants may also play a role in Q-drop.

[en] Many modern and proposed future particle accelerators rely on superconducting radio frequency cavities made of bulk niobium as primary particle accelerating structures. Such cavities suffer from the anomalous field dependence of their quality factors Q₀. High field degradation—so-called ‘high field Q-slope’—is so far unexplained even though an empirical cure is known. Here, we propose a mechanism based on the presence of proximity-coupled niobium hydrides that can explain this effect. Furthermore, the same mechanism can be present in any surface-sensitive experiments or superconducting devices involving niobium. (paper)

[en] Mechanical techniques for polishing the inside surface of niobium superconducting radio-frequency (SRF) cavities have been systematically explored. By extending known techniques to fine polishing, mirror-like finishes were produced, with <15 nm RMS (root mean square) roughness over 1 mm² scan area. This is an order of magnitude less than the typical roughness produced by the electropolishing of niobium cavities. The extended mechanical polishing (XMP) process was applied to several SRF cavities which exhibited equator defects that caused quench at <20 MV m⁻¹ and were not improved by further electropolishing. Cavity optical inspection equipment verified the complete removal of these defects, and minor acid processing, which dulled the mirror finish, restored performance of the defective cells to the high gradients and quality factors measured for adjacent cells when tested with other harmonics. This innate repair feature of XMP could be used to increase manufacturing yield. Excellent superconducting properties resulted after initial process optimization, with quality factor Q of 3 × 10¹⁰ and accelerating gradient of 43 MV m⁻¹ being attained for a single-cell TESLA cavity, which are both close to practical limits. Several repaired nine-cell cavities also attained Q > 8 × 10⁹ at 35 MV m⁻¹, which is the specification for the International Linear Collider. Future optimization of the process and pathways for eliminating requirements for acid processing are also discussed. (paper)

[en] A series of small fine-grained and single-crystal bars, with strain from 0% (recrystallized) to 50%, were given different amounts of chemical polishing. Four-point resistivity (ρ) data was used to characterize the electron scattering from dislocations, hydrogen, and any other trace contaminants. As noted by previous studies, annealed Nb displayed a weak linear increase of ρ (11 K) with polishing time due to hydrogen absorption, and bulk hydrogen concentration did not exceed 15% for 200 μm metal removed. Cold-worked samples displayed steeper slopes with polishing time (after subtracting resistivity due to strain alone), suggesting that dislocations assist the absorption of hydrogen during polishing. Absorption accelerated above 30% strain and 100 μm material removal, with room-temperature hydrogen concentration rising rapidly from 2% up to 5%. This threshold is significant, since superconducting radio-frequency (SRF) cavities are usually polished as-formed, with >35% strain, and polishing removes >150 μm of metal. Resistance jumps between 40 and 150 K, which signal the formation of hydride precipitates, were stronger in cold-worked samples, suggesting that dislocations also assist precipitate nucleation. High-vacuum anneals at 800 "°C for 2 h, which are known to fully recrystallize cavity-grade niobium and de-gas hydrogen, removed the 40–150 K jumps and recovered the resistivity increase due to chemical polishing entirely. But, about 30% of the resistivity increase due to cold work remained, possibly due to residual dislocation clusters. Continued annealing only facilitated the diffusion of surface impurities into the bulk and did not recover the initial 0% state. Strain, polishing, and annealing thus appear to combine as irreversible paths that change the material. Bearing this in mind, the significant difference in hydrogen uptake between annealed and cold-worked samples suggests that annealing SRF cavities prior to chemical polishing could greatly reduce hydrogen uptake and storage in the metal, reducing risk of quality-factor loss. This inverts key steps of the present widely-used cavity processing sequence. (paper)

[en] A primary difference between pure MgB₂ and its alloyed forms is that the former is a line compound and, once formed, has the same composition everywhere, whereas the latter is a solid solution and requires diffusion to move alloying elements. Since defect energies are high, this opens up the possibility that alloying elements might not be distributed homogeneously, which could have important consequences for the observed superconducting properties. To address this issue, two sets of Mg_1-xAl_xB₂ samples, with 0≤x≤0.45, were prepared from elements using reaction temperatures and times at opposite extremes of those typically reported in the literature. Sample set A was given a reaction of 1 h at 850 deg. C, which stopped just short of completion, while sample set B was reacted at temperatures as high as 1200 deg. C and thoroughly annealed for over 80 h. The trace reactants remaining after reaction A indicated that Al is taken up more slowly than Mg, thereby making compositional gradients likely. Indeed, Williamson-Hall analyses of x-ray diffraction peaks showed that set A had higher crystalline strain than set B when x>0 but not when x = 0. Since the presence of Al correlated with increased strain only for set A, it was concluded that reaction A produced substantial Al gradients across the individual grains while reaction B did not. Magnetization and heat capacity measurements indicated good bulk superconducting properties for all samples despite their structural differences, and consistent trends were observed when each sample set was considered alone. However, when both sets were considered together, their behaviour was distinct when plotted versus x (e.g. two T_c(x) curves), with trends for set A being shifted toward higher x than for set B. On the other hand, all of the data merged (e.g. one T_c(v) curve) when analysed in terms of the unit cell volume v. Thus, while the first analysis might suggest that the different reactions produced different superconducting behaviour, the second analysis, which captures the average Al content actually present inside the grains, shows that the samples have common behaviour intrinsic to the addition of Al. Moreover, these analyses show that it is important to coordinate structural and property characterizations to remove artifacts of composition gradients and uncover the intrinsic trends. Because the standard characterizations of the superconducting properties above gave no clear indication that the two sample sets had different homogeneities, the structural information was vital to make a correct assessment of the effects of Al doping on superconductivity. Since many investigations have used reactions similar to reaction A and did not analyse data in terms of structural changes, previous results should be interpreted cautiously

[en] Formation of MgB₂ by reactions of Mg with B₆Si and Mg with B were compared, the former also producing Mg₂Si as a major product. Compared to the binary system, the ternary reactions for identical time and temperature were more complete at 750 deg. C and below, as indicated by higher diamagnetic shielding and larger x-ray diffraction peak intensities relative to those of Mg. MgB₂ could be produced at temperatures as low as 450 deg. C by the ternary reaction. Analyses by electron microscopy, x-ray diffraction and of the upper critical field show that Si does not enter the MgB₂ phase

[en] Recent coordination of thermometry with optical images has shown that obvious defects at specific locations produce heat or even quench superconducting radio-frequency (SRF) cavities, imposing a significant limit on the overall accelerating gradient produced by the cavity. Characterization of the topography at such locations provides clues about how the defects originated, from which schemes for their prevention might be devised. Topographic analyses also provide understanding of the electromagnetic mechanism by which defects limit cavity performance, from which viability of repair techniques might be assessed. In this paper we discuss how a variety of two-component silicone-based room-temperature vulcanizing agents can be routinely used to make replicas of the cavity surface and extract topographic details of cavity defects. Previously, this level of detail could only be obtained by cutting suspect regions from the cavity, thus destroying the cavity. We show 3D profiles extracted from several different 1.3 GHz cavities. The defect locations, which were all near cavity welds, compelled us to develop extraction techniques for both equator and iris welds as well as from deep inside long 9-cell cavities. Profilometry scans of the replicas yield micrometre-scale information, and we describe various curious features, such as small peaks at the bottoms of pits, which were not apparent in previous optical inspections. We also discuss contour information in terms of electromagnetic mechanisms proposed by others for local cavity heating. We show that production of the replica followed by high-pressure rinsing does not adversely affect the cavity RF performance.

[en] High-purity niobium rods were cold-worked by wire-drawing, followed by various combinations of chemical polishing and high-vacuum baking at 120 °C or annealing at 800 °C in order to better understand changes to the surface superconducting properties resulting from typical superconducting radio-frequency cavity processing. AC susceptibility measurements revealed an enhanced upper transition T_c at ∼ 9.3–9.4 K in all samples that was stable through all annealing steps, a value significantly above the accepted T_c of 9.23 K for pure annealed niobium. Corresponding elevations were seen in the critical fields, the ratio of the surface critical field H_c_3 to the bulk upper critical field H_c_2 rising to 2.3, well above the Ginzburg–Landau value of 1.695. Orientation imaging revealed an extensive dislocation rich sub-grain structure in the as-drawn rods, a small reduction of the surface strain after baking at 120 °C, and a substantial but incomplete recrystallization near the surface after annealing at 800 °C. We interpret these changes in surface superconducting and structural properties to extensive changes in the near-surface interstitial contamination produced by baking and annealing and to synergistic interactions between H and surface O introduced during electropolishing and buffered chemical polishing. (paper)

[en] The costs of superconducting magnet strands are compared by calculating a 'production scaling factor' P that relates purchase data to the cost of raw materials. Using a consistent method, we normalize for different conductor geometries and strand diameters to arrive at cost indices in $ kg^-1, $ m^-1, and $ kA^-1 m^-1. Analyses of Nb47Ti conductors taken from the past 25 years of high-field magnet projects reveal that the price of raw materials and, to a lesser extent, finished strands, have tracked the price of niobium pentoxide. Performance gains during the 1980s produced $ kA^-1 m^-1 indices that fell with time ahead of strand cost in $ m^-1, a situation that may reflect the present status of Nb₃Sn magnet conductors. Analyses of present materials show that P decreases systematically with billet mass. While production strands in 200-500 kg billets have costs ∼3 times the cost of raw materials, the 20-50 kg billet size for internal-tin Nb₃Sn composites drives P up to 8-10. Thus, in contrast to LHC-type Nb47Ti strands that cost $150 kg^-1, $0.60 m^-1, and $1.00 kA^-1 m^-1 at 5 T, 4.2 K, Nb₃Sn strands required for the next generation of accelerator magnets are $1000 kg^-1, $4.00 m^-1, and >$5.75 kA^-1 m^-1 at 12 T, 4.2 K (where J_c is comparable to that for Nb47Ti at 5 T, 4.2 K). This high cost might be reduced by a factor of ∼3 if a large-scale internal-tin or powder-in-tube Nb₃Sn process can be found. Replacing expensive components with functionally equivalent but cheaper materials can produce 20% changes in $ kg^-1 and $ m^-1, but this might come at a performance penalty and no net savings in $ kA^-1 m^-1. Removing stabilizer from the strand cross-section and replacing it elsewhere in the cable can reduce the cost for a given length of cable significantly, but only if the processing cost for the strand remains unchanged after reduction of the stabilizer area. Emerging powder-in-tube composites show promise: Nb₃Sn strands could reach $6 kA^-1 m^-1 at 12 T, 4.2 K, while Bi-2212 strands could fall below $10 m^-1. MgB₂ superconducting strands could have very low raw materials cost at $0.20 m^-1, which translates to $1.00 m^-1 for finished strands and ∼ $1.00 kA^-1m^-1 at 2 T, 4.2 K based on published data. (topical review)