Suppression of magnetic vortex losses in submicron NbN coplanar waveguide resonators Authors: Ariel Roitman, Leonid Burlachkov, Amos Sharoni, Avner Shaulov, Yosef Yeshurun Published: 2024 ABSTRACT AbstractWe present a method for improving the performance of microwave coplanar resonators in magnetic fields, by using narrow superconducting strips of width close to the London penetration depth. In a range of low fields, the narrow strips inhibit the presence of magnetic vortices, thus preventing the generation of losses caused by their motion, leading to enhanced resistance to magnetic fields. Our method provides a more straightforward solution compared to previously proposed techniques designed to restrict vortex motion, holding potential for the development of improved devices based on microwave resonators. Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-024-76166-7.Keywords: Coplanar resonator, Superconducting NbN, Quality factor, Magnetic vorticesSubject terms: Superconducting devices, Applied physics Introduction Superconducting microwave resonators are employed in a broad range of applications, including microwave kinetic inductance detectors (MKIDs), circuit quantum electrodynamics, and quantum bits (qubits)1–9. The resilience of these resonators to magnetic fields is pivotal for stable and consistent performance in environments where magnetic fields are present. For example, the internal quality factor, , of MKIDs is significantly deteriorated by magnetic fields10,11. This has been attributed to losses caused by motion of magnetic vortices. Various strategies have been explored in earlier studies to minimize these losses by restraining vortex motion12–22. This was achieved, e.g., by incorporating artificial pinning centers13–19. However, these methods fall short of preventing vortex entry into the resonators. While pinning centers significantly reduce losses due to vortex motion, the presence of vortices leads to an increase in the density of quasi-particles, giving rise to losses and alterations in the resonance frequency. In this paper, we propose a method to eliminate vortices altogether by using thin superconducting strips of width smaller than the London penetration depth, . In this geometry, the Pearl penetration depth23, , (d is the film thickness) is significantly larger than the strip width, w, preventing the presence of any vortices within the resonator below a certain critical field24. As demonstrated later in this paper, the resonance frequency and the quality factor of such a resonator exhibit significantly higher resilience to magnetic fields compared to resonators with strips of width larger than This approach not only inhibits vortex formation but also enables a more compact, high-impedance resonator. Experimental Here we compare two types of resonators, made of NbN strips of widths 10 μm and 200 nm. Half-wavelength spiral coplanar waveguide resonators coupled to a 50 Ω transmission line were designed using the SONNET electromagnetic simulation software, see Fig. 1. In zero magnetic field and they exhibited resonance frequencies of 3.120 and 3.631 GHz, respectively. The NbN strips were fabricated from a 50 nm thick NbN layer with that was deposited onto a sapphire substrate using DC reactive magnetron sputtering (for further details, see supplementary). The lithography was conducted using the Elionix ELS 100 keV e-beam lithography machine. Both resonators were fabricated simultaneously on the same NbN film to ensure consistency and eliminate material and fabrication differences. A resist mask, 495PMMA A6, was employed, and Reactive Ion Etching (RIE) with Cl2/BCl3 gas was used to pattern the resonators. Subsequently, the resist was removed using acetone to eliminate any potential impact on the resonators. After fabrication, the transition temperatures of the narrow and wide resonators reduced to 9.3 and 11.5 K, respectively. SEM picture of the resonator. White and black regions present the superconductor and the bare substrate, respectively. We utilized the “Quantum Design” Physical Properties Measurement System (PPMS) to cool down to 2 K and apply magnetic fields up to 4 T perpendicular to the sample. Measurements of the resonance frequency (), and the internal quality factor () were conducted using a Keysight Vector Network Analyzer with 50 dB attenuators at the input of the resonators and a high-electron-mobility transistor (HEMT) amplifier at the output. The internal quality factor () was estimated using the formula25, where represents the value of at the resonance frequency, and , where Δf is the half-power bandwidth. The transmission line was connected to a 50 Ohm dedicated card using wire bonding. In Zero Field Cooled (ZFC) measurements, the resonator was cooled in zero field to the measurement temperature and then the field was applied and ramped up to 0.4 T, then ramped down to -0.4T and back to zero field. In Field Cooled (FC) measurements, the field was applied at a temperature above Tc and then the resonator was cooled down to the measurement temperature with the field on. Heating the device to above Tc and cooling it down to the desired temperature with the field on, was repeated for each field. Results Figure 2 shows the temperature dependence of the resonance frequency, measured in zero field, for the 200 nm and the 10 resonators, (main frame and inset, respectively). As expected, in both resonators decreases with increasing temperature, however, as a result of the lower Tc in the narrow resonator, (9.3 K compared to 11.5 K in the 10 resonator, as found from the fit to Eq. (1)), in this resonator decreases more rapidly with temperature as compared to the wider resonator. Indeed, plots of vs. T/Tc for both resonators overlap, as shown in Figure S1 in the supplementary section. Temperature dependence of the resonance frequency, , for the 200 nm resonator (main frame) and the 10 μm resonator (inset) in zero magnetic field. Solid lines are fits to Eq. 1 with =3.9. Figure 3 shows the quality factor, , for the two resonators as a function of Tc-T. Apparently, as a function of the distance from their respective critical temperatures, the narrow resonator exhibits higher quality factor. However, as a function of (Tc-T)/Tc, (see Figure S2 in the supplementary section) the wide resonator exhibits better , an advantage which diminishes as the temperature is lowered. In the following we demonstrate that a clear superiority of the narrow resonator is revealed when an external magnetic field is applied. Quality factor Qi for the two resonators as a function of Tc-T, with Tc = 9.3 and 11.5 K for the narrow and wide resonators, respectively. Figure 4a and b show the field dependence of in the two resonators, under zero-field-cooling (ZFC) condition. Evidently, the wide resonator exhibits a sharp drop of near H = 0 followed by an approximately linear decrease. This sharp drop is absent in the narrow resonator exhibiting a mild, approximately parabolic decrease of at low fields. Notably, upon decreasing the field back to zero, the wide resonator exhibits a significant hysteresis while the narrow resonator exhibits a reversible behavior. Magnetic field dependence of the resonance frequency, , at 2.1 K for (a) the 200 nm and (b) the 10 resonators. In ZFC measurements, the field was ramped up from zero to 4000 Oe (blue circles). It was then ramped down to -4000 Oe (yellow) and back to zero (red). In FC measurements (crosses), the field was applied at a temperature above Tc and then the device was cooled down to 2.1 K with the field on. Inset: Fit (solid line) to for field ramp up, based on Eq. 6. Figure 5a and b show the field dependence of . The high exhibited by the wide resonator at zero field, decays fast as the field increases and then continues to decrease with a decreasing rate. In contrast, in a logarithmic scale, in the narrow resonator decreases approximately linearly with the field, giving rise to a significantly lower rate of change of with the field. For example, upon application of a magnetic field of 2000 Oe, the of the wider resonator reduces by more than 4 orders of magnitude, from 3 × 106 to approximately , while the of the narrow resonator decreases by less than an order of magnitude, from 9 × 105 to only . Evidently, a resilience to magnetic fields is significantly pronounced in the 200 nm resonator. Magnetic field dependence of the internal quality factor, , at 2.1 K for the 200 nm and the 10 resonators. The symbols and color coding of the data are the same as in Fig. 4. Insets: Field dependence of the losses due to magnetic field. The crosses in Figs. 4a and b and 5a and b present results of field-cooled measurements of and . Evidently, in the narrow resonator the field-cooled and the zero-field-cooled data coincide for both and up to ~ 1100 Oe. In contrast, in the wide resonator the FC measurements show higher and somewhat lower than that of the ZFC data for the whole range. Discussion As mentioned above, the apparent difference between the temperature dependence of the resonance frequency in the wide and narrow resonators (see Fig. 2) is related to the difference between their critical temperatures, . The later can be derived from the temperature dependence of the resonance frequency, which stems from the temperature dependence of the kinetic inductance per unit length, where is the London penetration depth, is the density of the Cooper pairs, and and t are the width and the thickness of the strip, respectively. With , the resonance frequency of the half wavelength resonator becomes11: Here, , where denotes the magnetic inductance and (0) is the resonance frequency at T = 0. From SONNET simulations and measurements of, we estimated being and for the 200 nm and 10 , respectively. Fitting Eq. (1) to the data of Fig. 2, we obtained critical temperatures of and for the 10 m and 200 nm resonators, respectively, and for both resonators, in reasonably good agreement with the value of predicted theoretically26. Inspection of the effect of magnetic field on and show remarkable differences between the narrow and the wide resonators. In the following we show that the origin of these differences is the presence of magnetic vortices in the wide resonator but not in narrow one. In addition, in contrast to the wide resonator, the magnetic field in the narrow one is approximately uniform due to the fact that its width is close to and much smaller than . Vortices in thin superconducting strips in a perpendicular magnetic field were discussed in several works17,27–32. According to these studies, in the FC process there is a threshold field for trapping vortices near the center of the strip, being inversely proportional to the square of the width strip29: The threshold field for penetration of vortices in the ZFC process is much larger due to effect of the Bean-Livingston barrier33,34, which is especially effective in superconductors with large values of , and can increase up to28: For the 200 nm NbN resonator surface barriers are significant as , thus Eq. (3) is applicable, yielding , close to our maximum measurement field. Thus, our assumption of absence of vortices in ZFC measurements in the 200 nm resonator below this field is justified. Based on this assumption, we explain the observed differences between the narrow and wide resonators, as discussed below. The most important difference between the resonators is demonstrated in the field dependence of , (Fig. 5). For example, as mentioned above, in the range of fields between zero and 2000 Oe, in the wide resonator decreases by approximately 4 orders of magnitude, whereas the drop in of the narrow resonator in the same range is ~ 1 order of magnitude. Clearly, the presence of vortices in the wide resonator introduces significant losses due to their motion. These losses are not present in the narrow resonator due to the absence of vortices. Less sensitivity to magnetic field is also exhibited in the resonance frequency of the narrow resonator. As shown in Fig. 4, in ZFC measurements, in the narrow resonator exhibits an approximate parabolic decrease with the field, while the wide resonator exhibits a more complex behavior: a sharp initial drop followed by approximately a linear decrease. To explain the parabolic-like decrease of in the narrow resonator, we note that under uniform field, H, the density of Cooper pairs, ns, in narrow strips, depends quadratically on H35: where is a thermodynamic critical field, and are the density of Cooper pairs in presence and absence of the external magnetic field, respectively. According to Eq. (4) becomes: thus, The solid curve in the inset to Fig. 4a is a fit of Eq. (6) to the experimental data of fr(H) measured in the 200 nm resonator, using , , , and as a fitting parameter. An excellent fitting is obtained using which reasonably agrees with published values for NbN36. The complex behavior of in the wide resonators was observed previously and explained as resulting from the combined effect of vortices and screening currents11. The data of Fig. 5 also show a clear difference between the field dependence of the quality factor of the narrow and wide resonators. In a logarithmic scale, in the narrow resonator decreases approximately linearly with the field, while the wide resonator exhibits initially a sharp drop and continuous to decrease with a decreasing rate. This behavior can be explained along the same lines as outlined above, namely a decrease in ns with the field in the narrow resonator and an increase in the number of vortices, N, with field in the wide one. As the dependence of N on H is linear30, the losses due to the vortices, 1/Qv, in the wide resonator, must be also linear with H. Indeed, calculating as – using the data of Fig. 5b, yields linear dependence in H as shown in the inset to Fig. 5b. Performing similar calculations of for the narrow resonator (Fig. 5a) yields a different dependence that looks more like a parabola. This behavior can be explained as follows. The decrease in the density of Cooper pairs with the field, , gives rise to an increase in the density of the quasiparticles, , which at low enough temperature is approximately equal to , the latter being proportional to (Eq. 4), and, therefore, in Fig. 5a is expected to show a parabolic-like dependence on H as, indeed, shown in the inset to Fig. 5a. A clear difference between the behavior of the narrow and wide resonators is also demonstrated when comparing the increasing and decreasing of the external field, (see Fig. 4). In contrast to the wide resonator, the narrow one shows no hysteresis in both and upon increasing and decreasing the field. The hysteresis in the wide resonator is due to pinning of vortices causing different distribution of vortices on increasing and decreasing the field, as previously discussed by Bothner et al.11,13. The absence of vortices in the narrow resonator gives rise to reversible behavior without hysteresis. Finally, the narrow resonator exhibits no distinction between FC and ZFC data up to ~ 1100 Oe (see inset to Fig. 4a), after which they diverge. As previously discussed, the Bean-Livingstone barrier is absent in FC measurements, and, therefore, according to Eq. 2 the initial vortex penetration in FC is theoretically expected at ≅850 Oe, in a reasonable agreement with our experimental results. In contrast, the wide resonator displays a notable difference between ZFC and FC measurements, since in both cases vortex penetration occurs at fields close to 0. Such a difference between the ZFC and FC measurements in a wide resonator was observed previously and attributed to the combined effect of vortex pinning and screening currents resulting from field gradients in the ZFC, whereas in FC such currents are absent11. Conclusion We have successfully fabricated a microwave coplanar resonator, employing narrow superconducting with width close to the London penetration depth. The magnetic field dependence of both the resonance frequency and the Q-factor in this resonator differ markedly from that observed in the resonator with wide strips, exhibiting a considerably enhanced resistance to magnetic field. This improvement is ascribed to the presence of a uniform field and the absence of vortices within the narrow strip. The results of this research may significantly contribute to the development of improved devices that exploit microwave resonators, such as MKIDs and qubits. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements We thank Vladimir Kogan and Vadim Geshkenbein for very helpful discussions. Author contributions A.R., Av.S., and Y.Y., conceived the concept and designed the experiment and prepared the manuscript. Am.S. fabricated the NbN thin films. A.R., Av.S., Y.Y. and L.B., analyzed the data. Data availability The datasets used and analyzed during the current study are also available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Ariel Roitman, Email: roitmaa@biu.ac.il. Yosef Yeshurun, Email: yeshurun@biu.ac.il. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The datasets used and analyzed during the current study are also available from the corresponding author on reasonable request.