By simply turning up the energy of an ultrashort laser pulse, researchers can steer a niobium–titanium multilayer surface through a complete sequence of self-organized transformations, from molten clusters to rippled nanostructures and finally to ablation. A new study published in Applied Physics B systematically maps how pulse energy governs the formation of laser-induced periodic surface structures, known as LIPSS, on thin films of alternating niobium and titanium layers, and shows that the same processing step leaves behind an ultra-thin oxide layer rich in oxygen.
The team, led by Suzana Petrović with colleagues including Vladimir Rajić, David Canteli, Cristina Munoz Garcia, Jelena Potočnik, Dubravka Milovanović and Carlos Molpecres, deposited their multilayers on single-crystal silicon substrates using DC ion sputtering. Each sample contained fifteen alternating niobium–titanium bilayers with a total thickness of 440 nanometres, with titanium deposited first onto the silicon and niobium as the topmost layer. The titanium layers were roughly 15 nanometres thick and the niobium layers about 16 nanometres, deposited at rates of 0.13 and 0.16 nanometres per second respectively. The researchers report that the terminal niobium layer showed good stability, with no peeling or delamination observed either during storage or after laser modification.
For the laser processing, the team used a picosecond Nd:YVO4 laser operating at a wavelength of 355 nanometres, a repetition rate of 10 kilohertz and a pulse duration of just 15 picoseconds. The beam was focused normally onto the sample to a spot roughly 30 micrometres across, and a computer-driven translation stage with 0.1-micrometre resolution scanned the surface to inscribe lines at speeds of 1 to 5 millimetres per second. Because the pulses overlap heavily along the scan direction, the effective dose delivered to any point on the surface is far higher than the single-pulse fluence, and the authors quantify this cumulative exposure in their parametric study, in which the pulse energy was varied while all other laser parameters were held constant.
At a scanning speed of 5 millimetres per second, with pulse energies ranging from 1.5 to 2.4 microjoules corresponding to effective fluences of 12.7 to 20.3 joules per square centimetre, the surface morphology evolved in a clear sequence. At the lowest energy of 1.5 microjoules, partial surface melting at the centre of the line produced densely packed clusters of resolidified material, 50 to 100 nanometres in size and expected to be niobium-rich since niobium caps the structure. Raising the energy to 1.8 and then 1.9 microjoules produced elongated resolidified regions that arranged themselves periodically, marking the onset of high-spatial-frequency LIPSS with an average periodicity of about 110 nanometres and ripple lengths from 100 nanometres up to 1 micrometre.
The next threshold brought a striking new feature. At 2.0 microjoules, discrete droplets of roughly 200 nanometres emerged from the ripples and aligned into columns spaced approximately 370 nanometres apart, close to the 355-nanometre laser wavelength, which the authors interpret as the beginning of low-spatial-frequency LIPSS formation. The droplet columns ran perpendicular to the laser polarization, while the high-spatial-frequency ripples filling the spaces between them ran parallel to it. At 2.2 microjoules the two structure types coexisted, with droplet density increasing and the ripples persisting only between columns. At the highest energy tested at this speed, 2.4 microjoules, the periodic structure became disrupted, and sharp boundaries spaced about 350 nanometres apart appeared between the droplet columns, attributed to elevated stress in the resolidified material.
Slowing the scan to 1 millimetre per second, which increases the number of overlapping pulses per location, and raising the energies to between 2.0 and 2.8 microjoules, corresponding to effective fluences of 84.9 to 118.8 joules per square centimetre, pushed the surface through further regimes. At 2.0 microjoules, long cracks formed in the line centre, separating droplet columns into a well-defined low-spatial-frequency structure with a periodicity of about 400 nanometres oriented perpendicular to the polarization. At 2.2 microjoules the cracks densified and the droplets gradually disappeared, and at 2.5 microjoules smaller particles of 100 to 140 nanometres with sharp edges appeared, attributed to recrystallization of molten material or redeposition of ejected material. At 2.8 microjoules the accumulated energy raised surface temperatures enough to trigger material ablation, while rapid solidification of the molten layer promoted crystallization with defects acting as nucleation sites.
Profilometric measurements at a fixed pulse energy of 2 microjoules revealed an instructive asymmetry. Changing the scanning speed from 5 to 1 millimetre per second changed the ablation depth only slightly, from roughly 31 to 36 nanometres, but widened the modified line dramatically, from about 26 to about 45 micrometres. The authors conclude that increasing pulse overlap primarily enhances lateral energy redistribution rather than vertical material removal, a behaviour they link to the many titanium–niobium interfaces hindering efficient energy transport into deeper layers. The cross-sections also changed shape, from U-shaped profiles at the faster scan to W-shaped profiles at the slower one, reflecting non-uniform energy deposition and the differing thermophysical responses of the two metals.
The periodicity of the low-spatial-frequency structures itself proved tunable. At the slower scanning speed, the spacing in the centre of the lines decreased from approximately 400 nanometres at 2.0 microjoules to about 180 nanometres at 2.8 microjoules, while at the line periphery the values fell from 400 to 310 nanometres over the same range. The authors interpret this shrinking period as a shift in the dominant formation mechanism, from interference between incident and reflected laser beams at lower fluence toward ablation-dominated structuring at higher fluence. Fourier-transform analysis corroborated the picture quantitatively: sharp anisotropic maxima at 2.0 and 2.2 microjoules signalled well-ordered, highly aligned structures, while broadened and diffuse maxima at 2.5 and 2.8 microjoules documented a progressive loss of periodic order into irregular groove-like features.
Compositional analysis added a chemical dimension to the story. Energy-dispersive X-ray spectroscopy line profiles across a line written at 1 millimetre per second and 2.0 microjoules showed a pronounced decrease in niobium concentration in the central modified zone, a smaller change in titanium, and a noticeable increase in oxygen, indicating enhanced laser-induced surface oxidation associated with melting of the upper niobium layer. Because titanium melts at 1941 kelvin, well below niobium’s 2750 kelvin, melting of the surface niobium layer can propagate into the underlying titanium through heat transfer, facilitating niobium–titanium interdiffusion and, under ambient conditions, oxidation. The authors suggest the transient molten state may promote titanium oxide and niobium oxide phases, possibly mixed oxides at the interfaces, though they caution that without chemical-state analysis the specific oxide phases cannot be conclusively identified from the present data, and they propose X-ray photoelectron spectroscopy as the natural next step.
The study also carries honest limitations. The estimated heat-affected zone of about 22.4 nanometres suggests modification is confined mainly to the upper two bilayers under 15-picosecond irradiation, and because the EDS information depth at 10 kilovolts exceeds the modified surface region, the measured signal averages over the entire 440-nanometre multilayer plus a contribution from the silicon substrate, so the technique evidences overall elemental redistribution rather than resolving the chemistry of the modified surface layer itself. Even so, the work demonstrates that picosecond laser parameters critically govern the morphological, structural and chemical transformations of titanium–niobium multilayers, offering predictive control over LIPSS formation. Given the well-known biocompatibility and corrosion resistance of titanium- and niobium-based systems, and the stable passive oxides these metals form, the ability to write ordered nanostructures and oxide layers in a single laser pass points toward tailored surface functionalization for advanced biomaterial applications.
Subject of Research: Pulse-energy-dependent formation of laser-induced periodic surface structures and oxidation in niobium–titanium multilayer thin films processed with picosecond laser irradiation.
Article Title: Pulse energy effects on surface structuring and oxidation of Nb/Ti multilayers
Article References: Petrović, S., Rajić, V., Canteli, D., Garcia, C. M., Potočnik, J., Milovanović, D., & Molpecres, C. (2026). Pulse energy effects on surface structuring and oxidation of Nb/Ti multilayers. Applied Physics B, 132(11), Article 116. https://doi.org/10.1007/s00340-026-08715-z
Image Credits: AI Generated
DOI: 10.1007/s00340-026-08715-z
Keywords: LIPSS, picosecond laser, niobium, titanium, multilayer thin films, surface structuring, oxidation, laser ablation, sputtering, surface morphology, biomaterials, effective fluence
News Source: Denise Maddox. (October 29, 2026). Picosecond Laser Pulse Energy Controls Nanostructure and Oxide Growth on Niobium–Titanium Multilayers. Scienmag.
