A two-film laboratory comparison of NbTiN found that the film deposited at 7 mTorr had wider magnetization loops, a larger reported critical current density and a more uniform, locally ordered arrangement of vortices than the film deposited at 5 mTorr. The comparison also found different crystallographic textures and field-dependent resistance behavior between the films. Because the experiment used one film at each pressure, the results show an association in these samples rather than a general causal effect.
Both films were approximately 160 nm thick, and all other stated deposition conditions were kept fixed. The team used X-ray diffraction, electrical and magnetic measurements, and cryogenic magnetic-force microscopy after field cooling to 1.7 K. The applied magnetic fields were perpendicular to the film surface.
The films had different textures and field responses
X-ray diffraction showed comparable (111) and (200) reflections in the 5 mTorr film. In the 7 mTorr film, (111) texture dominated and the (200) reflection was strongly suppressed. The scans did not resolve individual grain-boundary angles, so effective coupling was inferred primarily from texture and transport.
At zero field, reported resistance-onset temperatures were approximately 13.96 K for 5 mTorr and 14.13 K for 7 mTorr. These are onset temperatures defined by a resistance criterion, not independent thermodynamic transition measurements. When a field was applied, the 5 mTorr film showed separate transition features associated with superconductivity within grains and between grain boundaries, plus an isolated-superconducting regime. The 7 mTorr film showed one dominant transition boundary. The low-temperature shoulder was assessed qualitatively, and the lack of a comparable feature at 7 mTorr was limited by experimental resolution.
The paper also reports model-derived zero-temperature upper critical fields of 9.13 T for 5 mTorr and 16 T for 7 mTorr. The corresponding coherence lengths were 6.0 nm and 4.53 nm. These figures came from a dirty-limit Werthamer-Helfand-Hohenberg extrapolation, so they are estimates within that model; the proposed microstructural explanation remains plausible rather than established.
The 7 mTorr film had the stronger magnetic response
Magnetization measurements showed wider hysteresis loops, a stronger irreversible response and a second magnetization peak in the 7 mTorr film. The same film had a larger reported critical current density throughout the measured field range, while critical current density fell as temperature rose in both films. The measurements did not establish a unique microscopic origin for the second peak or a specific elastic-to-plastic crossover.
A separate analysis of vortex pinning gave crossover exponents of about 1.5 for the 5 mTorr film and 1.8 for the 7 mTorr film. Both fitted responses were closer to the delta-l limit, with the 7 mTorr value closer. The authors interpreted that pattern as more like pinning linked to mean-free-path fluctuations, especially in the 7 mTorr film. Because the fitted components and exponents depend on the model, they do not identify a unique microscopic defect.
The vortex maps showed a more even pattern
The magnetic-force-microscopy images showed a similar contrast. At 1.7 K, after field cooling in perpendicular fields from 10 to 80 Oe, the 5 mTorr film showed heterogeneous, cluster-like distributions, including large clusters and sparse regions. The 7 mTorr film showed a more homogeneous distribution without comparably large clusters or sparse areas. At higher fields, vortex identification became more difficult in the 5 mTorr images, which limited the comparison.
Local geometry showed a smaller but consistent difference. Over most fields, approximately 40% of the vortices in the 5 mTorr film were sixfold coordinated, compared with approximately 50% in the 7 mTorr film. That measure captures local six-sided order, not a perfect lattice across the sample. Neither film showed a defect-free, long-range triangular vortex lattice.
Coordinate-based estimates of local force imbalance and vortex interaction energy showed broader distributions and stronger spatial fluctuations in the 5 mTorr film. The 7 mTorr film was more uniform and closer to the ideal triangular-lattice reference. The calculations were relative, semiquantitative descriptors rather than direct measurements of absolute forces or energies, and they used assumed penetration-depth and freezing-temperature parameters plus finite-field-of-view approximations.
A close comparison with clear limits
Taken together, the structural, transport, magnetic and microscopy results corresponded, in the authors' interpretation, to better effective grain-boundary connectivity, less weak-link behavior and more homogeneous, locally ordered vortex matter in the 7 mTorr film. The authors propose sputtering pressure as a practical tuning parameter for NbTiN vortex pinning and magnetic-field tolerance. The measurements show that these features appeared together in this comparison, but they do not establish that pressure alone produced a particular grain-boundary structure or pinning defect.
The study's reach is narrow. It compares one film at each pressure. X-ray diffraction did not resolve individual grain-boundary angles, while the structural and vortex images were not spatially registered, so the vortex maps were not direct maps of individual boundaries. The pinning fits and second magnetization peak also do not identify unique microscopic defects or a specific pinning transition. Further replication and direct boundary-to-vortex mapping would be needed to test how broadly the pattern holds.
Paper data and sources
Original title: Grain Boundary Engineering Effect on Vortex Matter in Superconducting Films
Authors: Qun Wang, Ting Chen, Ya-Xun He et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-26
DOI: Not available
Original paper · Full text