WorksPhysics and Astronomy
Characterizing Niobium Nitride Superconducting Microwave Coplanar Waveguide Resonator Array for Circuit Quantum Electrodynamics in Extreme Conditions
Abstract
The high critical magnetic field and relatively high critical temperature of niobium nitride (NbN) make it a promising material candidate for applications in superconducting quantum technology. However, NbN-based devices and circuits are sensitive to decoherence sources such as two-level system (TLS) defects. Here, we numerically and experimentally investigate NbN superconducting microwave coplanar waveguide resonator arrays, with a 100 nm thickness, capacitively coupled to a common coplanar waveguide on a silicon chip. We observe that the resonators' internal quality factor (Qi) decreases from Qi ~ 1.07*10^6 in a high power regime (< nph > = 27000) to Qi ~ 1.36 *10^5 in single photon regime at temperature T = 100 mK. Data from this study is consistent with the TLS theory, which describes the TLS interactions in resonator substrates and interfaces. Moreover, we study the temperature dependence internal quality factor and frequency tuning of the coplanar waveguide resonators to characterise the quasiparticle density of NbN. We observe that the increase in kinetic inductance at higher temperatures is the main reason for the frequency shift. Finally, we measure the resonators' resonance frequency and internal quality factor at single photon regime in response to in-plane magnetic fields B||. We verify that Qi stays well above 10^4 up to B|| = 240 mT in the photon number < nph > = 1.8 at T = 100 mK. Our results may pave the way for realising robust microwave superconducting circuits for circuit quantum electrodynamics (cQED) at high magnetic fields necessary for fault-tolerant quantum computing, and ultrasensitive quantum sensing.
Introduction
Superconducting coplanar waveguide resonators, or CPWs, play an essential role in circuit quantum electrodynamics (cQED) [1, 2, 3], superconducting quantum sensing [4, 5, 6], quantum metrology [7, 8], and quantum computing [9, 10]. The electric field in CPWs can be confined inside the resonator, making them an appropriate choice for coupling between quantum devices. A further benefit of CPWs is their simple geometry, which makes them a fundamental component of quantum circuits. In this regard, it is crucial to maintain a high internal quality factor () of CPWs in quantum circuits to minimise unwanted losses that affect qubit gate fidelities and qubit relaxation times [11, 12] to achieve robust and fault-tolerant quantum computers [10, 13].
Recently, attention has been on identifying the loss mechanisms contributing to the quality factor of the CPWs [14]. Recent results show that two-level systems (TLSs) and quasiparticle losses are mostly responsible for quantum circuits' decoherence [15, 16, 17, 18]. TLS loss appears when the defects in amorphous material (such as the substrate of quantum circuits and oxides growth on the superconductor) are coupled with the electric field of CPWs. These TLS defects are mostly confined to the ground states at millikelvin temperatures and low powers. They absorb energy from CPWs and dissipate heat to the resonator when defects jump to excited energy states [19]. Accordingly, circuit engineering techniques have been developed on topics of finding low-loss substrate and superconductor materials [20, 21], optimising circuits geometry [22, 23, 24] and improving micro and nanofabrication techniques [25, 26]. Several research groups have pursued these methods to improve the quality factor in the CPWs and achieved a high internal quality factor's CPWs in the quantum devices [27, 28, 29, 30, 31].
Moreover, there has been considerable interest in the characterisation of the quasiparticle loss [32, 33, 34, 35], another cause of loss in CPW. An active area of research is to find the dynamic of the quasiparticle effect on superconducting quantum devices [36, 37, 38], which can host Andreev bound states (ABSs) [39, 40, 41] and Majorana bound states (MBSs) [39, 42]. Emerging MBSs in hybrid superconductor-semiconductor devices require an application of magnetic fields while maintaining the superconductivity [43]. Therefore, topological superconductivity can potentially appear in a hybrid circuit with superconductors, such as niobium titanium nitride (NbTiN) [44, 45], tantalum (Ta) [46], lead (Pb) [47], and niobium nitride (NbN) with their relatively high critical magnetic fields, and also for this purpose, it will be important to characterise the microwave superconducting circuits losses in the presence of the magnetic field.
A main challenge in superconducting circuits operating in extreme conditions, e.g., in the presence of magnetic fields, is achieving high resonance since changes in the magnetic flux density in superconductors cause suppression of superconductivity or a phase transition from superconductor states to normal states. Moreover, in type II superconductors, the superconductor enters to the mixed states when the magnetic field is applied, leading to vortices pinning and flux creep. The following issues need to be addressed to preserve high resonators based on the type II superconductors [30]. First, the becomes more vulnerable to even small magnetic fields due to the creation of Abrikosov vortices: supercurrent regions that circulate around a non-superconducting core [48]. Second, type II superconductors have a lower coherence length in comparison with type I superconductors; therefore, energy losses have a higher value at single photon regimes. The possible solutions to overcome these challenges are to reduce the geometry of the resonator to below magnetic field penetration depth or to define lithographically artificial defect sites to pin the vortices [48, 49, 50, 51].
In this work, we first report the design and fabrication of NbN CPWs arrays on two silicon chips whose frequency resonance window is between 4 GHz and 8 GHz. Following this, we discuss characterisation and analyses of the frequency shift and internal quality factor of the CPWs in response to sweeping temperature, microwave power and magnetic field. We further verify that the of CPW in one of the chips increases from in single photon regime to at photon number , at mK. Indeed, we find that the of the resonator changes when the magnetic field is applied at different temperatures. Moreover, we show that stays above at , and when parallel magnetic fields mT are applied to the chips at mK. These results show that the NbN resonators are promising superconducting material candidates for hybrid quantum electrodynamic circuits requiring magnetic fields.
Method
Design and Fabrication
Our superconducting microwave circuits consist of an array of microwave resonators capacitively coupled to a common coplanar waveguide and fabricated on a 100 nm thin NbN film sputtered on a silicon chip. The silicon substrate is dipped in Buffered Oxide Etchant (BOE) to strip the native SiO from the substrate's surface to achieve a clean substrate-metal interface. Immediately after removing the SiO layer, the chip was placed in the load lock of an MP 600 S Plassys sputter system to prepare for NbN metal deposition [52]. After the load lock pumps down and reaches a low pressure, the sample was transferred to the main chamber to reactively sputter 100 nm of NbN. In the main chamber, by sputtering niobium (Nb) for 15 minutes with 25 sccm of Argon (Ar) and 3.0 sccm of Nitrogen at the base current of 0.85 A, 100 nm NbN film was formed on top of the 525 µm silicon substrate. After removing samples from the load lock of the sputtering machine, ZEP resist was spun and baked at 180 °C on hotplates for four minutes. The resonators were patterned with standard e-beam lithography followed by /Ar anisotropic dry etch. The resonator length was chosen to have a fundamental frequency mode between 4 – 8 GHz.
Two different types of samples are fabricated and measured, each containing three resonator arrays with 4 µm width, and 2 µm gap, multiplexed to a single feedline. Figure 1(a), and (Figure 1b) show a wire-bonded superconducting CPW chip, and the packaged chip, respectively. A top-view schematic of one resonator is shown in Figure 1(c). Moreover, a zoom-in view of the resonator (R1) and the magnified area showing the gap and width position are schematically depicted in Figure 1(d) and Figure 1(e), respectively. More details of the fabricated superconducting CPW circuits in this study are provided in Table 1.
Figure 1. NbN based superconducting microwave coplanar waveguide resonators. (a) Optical image of sample 1 which is wire-bonded in the sample holder. (b) Packaged chip for cryogenic microwave spectroscopy. (c) A topview schematic of the superconducting circuit used in the experiments shows three capacitively coupled resonators multiplexed to a common coplanar waveguide (brown dashed line area in Fig. (a)). The length of resonators from left to right are mm, mm, mm. Here, the thickness of NbN film, and silicon substrate is 100 nm, and 525 µm, respectively. (d) Magnified schematic of resonator 1 (blue dashed line in Figure 1(c)). (e) The magnified area shows the resonator's gap and width position (red dashed line in Figure 1(d)).
| Sample | Resonator Width | Resonator Gap | R1 Length | R2 Length | R3 Length |
| 1 | 4 m | 2 m | 4.688 mm | 6.250 mm | 7.9 mm |
| 2 | 4 m | 2 m | 4.5 mm | 5.5 mm | 6.5 mm |
Table 1. NbN superconducting microwave coplanar waveguide resonator array design detail.
Experimental setup
The sample box (Figure 1(b)) was mounted in two different cryogenic setups to collect the data presented in this paper: 1) an adiabatic demagnetization refrigerator (ADR) with a base temperature of mK (see the fridge setup in FIG.S1. (a)). 2) a dilution refrigerator (DR) with a base temperature of mK (see FIG.S1. (b)). The complex microwave transmission spectroscopy () was performed with a MS4642B Keysight Vector Network Analyzer (VNA) by sending signals to ports on the NbN transmission line. In the ADR setup, the input line of the VNA is heavily attenuated (-40 dB) at room temperature. Then, VNA signals were attenuated by -20 dB at K, 4K, and 0.5 K stages before reaching the transmission line of the sample, suppressing thermal noise and allowing the resonator to reach into the single photon regime. At the K stage, after the signal passes through the sample box, an isolator is placed to prevent contamination of transmission signals from spurious reflections back to the transmission line. Then, the signal was amplified by a high electron mobility transistor (HEMT) which provided +40 dB amplification. Moreover, an additional +45 dB amplification is provided with a room-temperature amplifier. RF signals generated by the VNA propagate through the ADR before reaching the CPW transmission line, where they excite the resonators. The signal power levels range from -30dBm to 15dBm (see FIG.S1. (a)).
We note that, for magnetic field measurements, we mounted our sample inside a Solenoid (NbTi superconducting magnets) in the ADR such that the NbN transmission line is parallel to the applied magnetic fields. The current in the Solenoid sweeps between -10 A to 10 A, leading to a magnetic field sweep between -800 mT to 800 mT [51]. In this work, we present the microwave power and temperature sweep of sample 2 which are measured in a dilution refrigerator (DR) and the ADR refrigerator, and the magnetic field sweep of sample 1 measured in the ADR refrigerator. The DR cryogenic setup detail is shown in FIG.S1. (b).
Numerical methodology
The extraction of resonator parameters such as resonance frequency (), internal quality factor (), coupling quality factor (), and loaded quality factor () from measurements, is a crucial tasks in the characterisation and design of resonate circuit. To pursue this goal different types of RLC circuit models have been used in previous research based on amplitude or phase of [28, 53, 54, 55, 56]. Modern methods typically utilize the full complex scattering data provided by VNA as opposed to focusing on the power or amplitude to extract resonator parameters [28, 57, 58]. Using entire complex data results in a better signal-to-noise ratio (SNR) and a more accurate determination of resonator parameters. This article uses a notch-type resonator model to extract resonance frequency and quality factors of the NbN microwave superconducting resonators. The of notch-type resonators can be evaluated with the following equation [58]:
The second part of Eq. (1) characterises an ideal notch-type resonator where indicates the probe frequency, and quantifies the impedance mismatch. In more detail, indicates the difference between input and output impedance at the ports of the resonator. In an ideal notch-type RLC circuit, the circle (Im() and Re() circle plot) intersects the real axis at unity, which corresponds to a probe frequency at called off-resonant point.
Note that the diameter () of the circle comes from . However, in experimental measurements from VNA, a rotation in the phase and a change in the amplitude of the circle have been observed. To overcome the above issue, the first term in Eq. (1) has been added to describe the environmental noise. First, an amplitude shows the cable damping effect in , a phase shift shows an extra rotation in the phase of coming from cable length, and finally, an electronic delay which changes the initial phase in signal input to a different value from zero. The length of the cable and the finite speed of light cause an electronic delay. Besides, with a circuit model we also can drive the photon number inside the resonator with the below formula [59]:
Where is the photon number, is the input power, is the transmitted power through the feedline, is the source of losses, is plank constant and is resonance angular frequency of resonator. A typical result of such a fit is shown in Figure 2, where we show the measured and calculated frequency spectrum of sample 2 at fundamental frequency GHz at mK. Indeed, we investigate relation with power in the DR cryogenic setup to confirm the reproducibility of measurements taken in the ADR and cover a higher range of photon numbers.
Figure 2. (a) Parametric plot and fitted Im () and Re (), (b) Measured and calculated frequency spectrum of sample 2 at fundamental frequency GHz at mK, same data in (a).
Figure 3(a) shows the frequency spectrum comparison, confirming a good agreement between measurements taken in two different DR and ADR cryogenic setups. In Figure 3(b), we compare extracted and taken in the ADR and DR cryogenic setup at power range -100 dBm -150 dBm at mK. The results confirm that both measurement setups are in accordance with each other during power sweep measurements. In the following chapter, we characterise kinetic inductance, TLS loss, and quasiparticle loss.
Figure 3. (a) Frequency spectrum of sample 2 measured in two different cryogenic setups: the ADR (red dashed-line), and DR (blue dashed-line). A good agreement between the two measurements can be seen at mK. (b) Measured at fundamental frequency GHz in the power range of dBm to -100 dBm at mK in two different setups.
Results
Kinetic inductance characterisation
In the simplest case of capacitively coupled superconducting quantum circuits, modelling the resonator as an circuit, the coupling strength between the feedline and resonator will be proportional to the impedance of the quantum circuits. In the quarter wavelength resonators impedance and resonance frequency of the circuit can be extracted by the equation and where and . Here, , and are geometrical inductance, kinetic inductance, and geometrical capacitance per unit length of transmission line, respectively. Moreover, emphasizes on the resonance frequency number ( for the fundamental mode) [28].
Therefore, characterising the kinetic inductance of superconducting circuits to identify coupling strength and resonance frequency is an essential step toward robust quantum technologies. Moreover, the resonators' kinetic inductance helps us to evaluate quasiparticle loss and frequency shift in superconducting circuits. More details are discussed in § 3.3.
For obtaining kinetic inductance, we first use conformal mapping techniques [60] to extract the geometrical inductance H/m and the capacitance F/m. We obtain kinetic inductance H/m, which shows lower kinetic inductance compared to previous reports on NbN coplanar waveguide resonators and thin NbN sheets [28, 61, 62]. The following section investigates the effects of power, temperature, and magnetic field on NbN's CPW.
Power dependence
Knowing how sweeping power and temperature affect material defects' behaviour can help us make more robust quantum circuits. To achieve this, we study the resonator’s to an average photon number inside the resonator at three different temperatures: at (i) mK, (ii)
200 mK, and (iii) mK. This is done in order to identify the sources of energy loss of the resonator at different temperatures. We split measured losses into three terms, as shown in equation (5). Here, describes TLS loss, is quasiparticle loss, is vortex and magnetic field loss, and is power and temperature independent loss [33].
TLS defects in the ground state are the dominant cause of increasing loss at sub-millikelvin temperatures since TLS defect's jumping to the excited states induces heating dissipation in the resonator [19]. Meanwhile, the superconducting energy gap is larger than the thermal noise of the system at temperatures , where is superconducting gap, is bolt zman constant, so Cooper pair breaking is less likely to happen. Therefore, quasiparticle density becomes negligible. In this case, the loss rates can be characterised [21]:
where is resonance frequency, is the average photon number, is critical photon number, and is TLS loss at zero power and temperature.
As shown in Figure 4, in mK, saturates at in single photon regime, and as power rises it goes to surpasses . We find an agreement between TLS theory and measurement results, as shown in Figure 4(b). We find that , and are , 13 and 0.35, respectively. In comparison with previous measurements [27, 28, 29, 30, 63] we achieve a of around one million at only , at mK, which suggests a good improvement in the development of high- superconducting CPWs based on thin NbN films.
Figure 4. (a) Measured at fundamental frequency GHz in the power range dBm to -100 dBm at mK, 200 mK and 650 mK. The shaded area corresponding to deviation/error from the actual data. (b) versus photon number at mK fitted with the theoretical model of TLS. These data are measured for sample 2 in the DR.
In comparison, at higher temperatures, TLS defects saturate in a mixture containing equal proportions of ground and excited states; therefore, an energy exchange happens between energy states, reducing the energy loss from TLS defects. However, the energy gap is near the required Cooper pair-breaking energy in this temperature range. Decreasing the superconducting energy gap leads to an increase in quasiparticle density in the resonator, which is the primary cause of energy loss at this temperature range. The quasiparticle loss can be characterised by the below model [64, 65]:
Where is the ratio between kinetic inductance and total inductance , (see § 3.1), is NbN superconducting energy gap, is density states of Fermi level, and is quasiparticle density. Therefore, as expected from equation (7), the loss of the resonator will be mainly dependent on the temperature, and it will be inconsistent with sweeping power, which is what we see in our measurement at mK and depicted in Figure 4(a).
Table 2 compares the highest measured of both samples at millikelvin temperatures during power sweep measurements. It shows that the values are compatible with each other in single photon and high power regimes, even though each sample has been fabricated in a different timeline. These results confirm that thin NbN superconductor film and silicon substrate quality is approximately maintained regardless of the environmental situation and measurement setup.
Temperature dependence of and
In the next step, we extracted the and at single photon regime in the temperature ranges between mK and 3K. The results show that at around K, the value is at its highest value, as shown in Figure 5 (a). This behaviour is in good agreement with previous research on TiN and MoRe based microwave resonators [66, 67]. The same investigation has been done for the frequency shift of the resonator versus temperature; the data is shown in Figure 5 (b). We observe that at K, the resonance frequency experiences a noticeable red shift, the same temperature when we extracted the highest . In more detail, when energy loss is mainly due to TLS defects, the resonance frequency shift also happens due to TLS loss (Figure 5) which can be described by the below model [65]:
Figure 5. (a) versus temperature at single photon regime. The shaded area corresponds to deviation/ error from the actual data GHz. (b) Resonance frequency shift versus temperature. The blue and red circles show and calculated quasiparticle loss frequency shift, and total loss frequency shift, respectively. These results are measured data of sample 2 in the ADR.
where is the digamma function. We can plot this function using data extracted from the previous fit in Figure 5 (b). At higher temperatures, due to an increase in quasiparticle density, kinetic inductance increases, which leads to the below frequency red shift [65]:
In Figure 5 (b), we plot the measured relation with temperature and calculate frequency shifts obtained by equation (9) and equation (10). In this equation, we use the kinetic inductance ratio obtained from section 3.1. This theory agrees well with the measured data extracted from the resonator samples; the plot has been shown in Figure 5 (b):
We see that in higher temperatures; therefore, the main reason for resonance shifts in high temperatures is quasiparticle defects. Table 3 shows the highest and lowest in a single photon regime for both samples measured during the temperature sweeps. It shows that remains higher than , although we measured our samples at high temperatures which are suspected to have high Cooper pair breaking rates.
| Photon number | Max () | Min () | |||
| Sample 1 | 2418 | 2314 | |||
| mK | K | ||||
| Sample 2 | 2142 | 2118 | |||
| K | K |
Table 3. The highest and lowest internal quality factors of each sample in single photon regime measured during temperature sweep. Both of these measurements were performed in the ADR setup.
Note that and values of sample 1 differ from the results in Table 2 because we did the wire bonding again due to the destruction of Aluminum wires during temperature sweep measurements. Wire bonding is mainly responsible for the impedance value of superconducting CPW. Therefore, as equation (1) shows, impedance mismatch affects the and .
| (mK) | (GHz) | (high power) | (single photon) | |||
| Sample 1 | 100 | 5.303 | 6378 | 6340 | ||
| Sample 1 | ||||||
| Sample 2 | 26 | 5.952 | 2308 | 2252 | ||
| Sample 2 |
Table 2. Comparing the measured highest internal quality factor of each sample.
Evaluation of the CPWs under parallel magnetic fields
Finally, we examine the behaviour of the resonators when a static magnetic field is applied. The sample box (sample 1) is mounted in a cryogenic setup in the way that the NbN transmission line is positioned parallel to magnetic field . We perform three measurements to characterise the magnetic field effect on the resonators. At first, we upsweep from 0 to 720 mT with 80 mT steps (see Figure 6 (a)). The results of these measurements indicate that as the magnetic field is increased up to 720 mT, the resonance frequency experiences a redshift, and the dip value decreases because of Cooper pair breaking and emerging Abrikosov vortices.
Figure 6. (a) The versus frequency spectrum between T and 720 mT with 80 mT steps at mK. The shaded area corresponds to frequency shift in magnetic field sweep in the range of . (b) Extracted frequency shift as a function of (mT) with , 6 and 1.8. A fit of the measurement data with the equation which gives at mK and , 6 and 1.8. (c) as a function of at mK and , 6 and 1.8. The remains above until mT at all the mentioned photon numbers. These are measured data of sample 1 in the ADR cryogenic setup.
Trapped flux inside the NbN, as a type II superconductor, strongly depends on sample preparation and geometry. Moreover, existing and stability of these fluxes inside the superconductor, which affect resonator loss, come from two different forces, which appear due to the Meissner effect and Lorentz forces. First forces, due to the Meissner effect, a vortex is pushed toward the edges of the superconductor to expel from it. The second force happened when the vortex is attracted to the centre of the superconductor due to weak interactions of the Meissner screening current. The second one is proportional to the , and at a high enough field, the vortex will exist and pin in a stable condition in a line of superconductor thin films [68, 69]. Based on theoretical investigation [68, 69], vortices in thin superconducting film for experience expulsion force, resulting in a more stable condition mT.
Then vortices begin to align in a straight line when and the edges of the thin film become vortices free mT. For , the density of vortices increases in a way where vortices-vortices interaction should be taken into serious consideration where leading to a higher loss (Figure 6 (a)). Here, and exhibit flux quantum and thickness, respectively. In the next step, we measure the effect of applying on the and of CPW. As increases, experiences a redshift, as shown in Figure 6 (b), due to the increased rate of Cooper pairs breaking. The frequency shift can be modelled as a parabolic decrease [70] as , with , which depends on resonator thickness , critical temperature and the electron diffusion constant in NbN. The electron diffusion constant is approximated to , consistent with the previous experiments [28, 71, 72].
Figure 6 (c) demonstrates the effect of in-plane magnetic fields on the resonators at mK, and in , and . It shows that by increasing the magnetic fields, decreases. However, remains above up to mT, which shows the resonator's effectiveness under applying magnetic fields. In superconducting resonators, the resonance frequency is determined by the RF surface impedance of the superconductor, which can be affected by flux creep and vortices pinning [68]. As magnetic vortices move and generate electrical resistance and heat, the surface impedance of the superconductor changes, which in turn can lead to a shift in the resonance frequency. The degree of resonance shift depends on several factors, including the strength of the applied magnetic field, the temperature, and the structure of the superconductor, including any defects and impurities or material disorder.
To further study the effect of in-plane magnetic fields on resonance frequency , we reduced the steps in the magnetic field measurement to 8 mT. We performed an upsweep measurement in the range between 0 mT 80 mT (Figure 7). We observed that experiences butterfly effects at low magnetic fields 0 mT 32 mT, showing the mobility of the vortices which have a hysteresis effect in the conductivity of the NbN superconductors. Figure 7(a) shows that the changes in the complex conductivity of planar superconducting circuits due to dissipative vortices motions lead to a jump in resonance frequency [73]. Indeed, this jump which occurs at magnetic fields ranging from 16 mT 32 mT suggests a sudden movement or depinning of the vortices in the NbN superconductor. As the magnetic field increases, vortices tend to pin in the superconducting material. As shown in Figure 7(b), we observed that, in a higher magnetic field range between 32 mT 80 mT, the resonance frequencies decrease as the magnetic field increases implying that NbN thin film conductivity wasn’t affected by the mobility of the vortices since no butterfly effect of the resonance frequency is detected [69] [73].
Figure 7. (a) Measured frequency spectrum for an upsweep measurement of between mT to 32 mT with a step of 8 mT at temperatures mK. In this magnetic field range, we observe a resonance frequency jump at mT. (b) Measured frequency spectrum for an upsweep measurement of the magnetic field between mT to 80 mT, the resonance frequency shift as expected.
Moreover, to understand the effect of the solenoid on the of our sample, we increased the magnetic field's range to -160 mT 160 mT. We expected that the plot should have a symmetrical shape around mT where . However, the maxima () shift towards negative fields mT. This shift may happen due to the remaining magnetic field from the ADR magnet which is responsible for the adiabatic demagnetisation procedure. Besides, we observe a very high jump at mT) at mK (see Figure 8). To identify the jump further, we performed temperature dependent measurements at mK, 500 mK, 1 K and 1.5 K to examine the effect of temperatures. As shown in Figure 8, we observe a similar jump at all temperatures, which we presume is due to vortices trapped in the NbTi solenoid.
Figure 8. Measured frequency spectrum with magnetic field sweep raging between with a field step of 8 mT at (a) 100 mK, (b) 500 mK, (c) 1 K, (d) 1.5 K. Arrows indicate sweep direction.
To further study the progress on the fabrication procedure of NbN coplanar waveguide resonator, we compare our techniques with the most recent reports on NbN superconductor resonators in Table 4. We fabricated our CPW by NbN sputter deposition without heating, which causes a simpler and more robust fabrication procedure compared to atomic layer deposition (ALD) and heating sputtering. Note that the recent goal in quantum circuits research is to reduce the chip footprint, which we also move towards by design and fabrication of resonators. To reduce the oxidation of NbN, we used CF and Ar gases since they are less susceptible to introducing unwanted defects and material added to our superconducting thin film. However, more studies on how the etching gas influences the internal quality factors will be worth investigating.
Table 4. Comparison of NbN-based microwave coplanar waveguide resonator design, and fabrication.
Regarding design parameters, we choose 100 nm film thickness for two reasons: not only does 100 nm NbN superconducting thin film have a lower kinetic inductance, and is more robust against noises sources in the cryogenic measurement setup in comparison with thinner films, but it also gives a more robust internal quality factor after applying magnetic fields in comparison to bulk superconducting film [27, 28]. Our resonator has stayed in the middle range of film thickness . However, finding a sweet spot in NbN film thickness will be worth more investigation. Moreover, we compare our resonator internal quality factor in single photon and high power regimes to recent works to further understand the performance of NbN superconducting resonators.
The results are shown in Table 5. We find that at single photon regime, the is in line with the latest research on superconducting NbN coplanar waveguide resonators, which confirms the reliability of the fabrication process and measurements at different magnetic fields and temperatures.
Table 5. Comparison of NbN-based coplanar waveguide resonator microwave response detail.
In terms of at high power regime, our resonator has the second highest value, even though we measured at far lower photon number and temperature than Carter, Faustin W., et al work. Our work in NbN based circuits is the first research demonstrating resonance frequency shift and internal quality factor behaviour as a function of temperature from the TLS loss to the quasi-particle loss ranges.
Meanwhile, although we obtained a high internal quality factor, we have not used artificial pinning in our circuit design, leading to a quicker and easier e-beam lithography procedure.
These comparisons confirm that our approach in designing, fabricating, and characterising NbN superconducting CPW is a step towards developing high internal quality factor resonators useful for quantum technology from quantum computing to quantum sensing and metrology.
Conclusion
In summary, we reported the successful design, fabrication and microwave spectroscopy of arrays of CPWs based on superconducting NbN on silicon chips in the cryogenic environment. We studied resonator parameters as a function of microwave power, temperature, and magnetic fields. We studied TLS, quasiparticles, magnetic fields, and temperature-independent losses to identify the loss mechanism in our superconducting circuits. At low temperatures ( mK), the range in which TLS loss is the dominant reason for the rising loss in CPWs, the quality factor increases from in the single photon regime to at high powers. Sweeping temperature from mK to 3 K in the single photon regime also confirms that TLS loss decreases when the TLS defects saturate in the energy bands. Moreover, in higher temperatures, the range in which quasiparticle loss is dominant, reduces from at K to at K. To examine the suitability of such superconducting circuits for cQED at high magnetic fields, we studied the microwave response of our CPWs to in-plane magnetic fields . Our results showed that the peak of is stable in the magnetic field ranges between mT to 240 mT at mK. We extracted at single photon when mT. Our results shall help the development of superconducting microwave quantum circuits compatible with high magnetic fields necessary for applications in fault-tolerant quantum computing, quantum sensing, and quantum metrology.
Acknowledgment
This work was supported in part by the Royal Society Research under Grant RGS/R2/222168, the Research Fellowship from the Royal Society of Edinburgh (K Delfanazari), the Innovate UK (QTools, grant No 79373 and FABU, grant No 50868), the EPSRC Quantum Computing Hub EP/T001062/1, and by the FET Open initiative from the European Union’s Horizon 2020 program under N° 899561. The discussion with Dr A. Sultanov is greatly appreciated.
Appendix
Microwave spectroscopy setup
Figure S.1. (a) Adiabatic demagnetisation refrigerator (ADR) experimental setup schematic used to perform microwave and magnetic field measurements with NbTiN superconducting solenoid, Cean is the current source of NbTi solenoid. (b) Schematic of the dilution refrigerator used to perform microwave measurements.
References
- [1]Clerk, A., et al., Hybrid quantum systems with circuit quantum electrodynamics. Nature Physics, 2020. 16(3): p. 257-267.
- [2]Blais, A., et al., Quantum-information processing with circuit quantum electrodynamics. Physical Review A, 2007. 75(3): p. 032329.
- [3]Frunzio, L., et al., Fabrication and characterization of superconducting circuit QED devices for quantum computation. IEEE transactions on applied superconductivity, 2005. 15(2): p. 860-863.DOI
- [4]Danilin, S., N. Nugent, and M. Weides, Quantum sensing with tuneable superconducting qubits: optimization and speed-up. arXiv preprint arXiv:2211.08344, 2022.arxiv.org/abs/2211.08344
- [5]Degen, C.L., F. Reinhard, and P. Cappellaro, Quantum sensing. Reviews of modern physics, 2017. 89(3): p. 035002.
- [6]Lita, A.E., et al., Development of Superconducting Single-Photon and Photon-Number Resolving Detectors for Quantum Applications. Journal of Lightwave Technology, 2022. 40(23): p. 7578-7597.DOI
- [7]Earnest, C.T., et al., Substrate surface engineering for high-quality silicon/aluminum superconducting resonators. Superconductor Science and Technology, 2018. 31(12): p. 125013.DOI
- [8]McRae, C., et al., Thin film metrology and microwave loss characterization of indium and aluminum/indium superconducting planar resonators. Journal of Applied Physics, 2018. 123(20): p. 205304.DOI
- [9]de Leon, N.P., et al., Materials challenges and opportunities for quantum computing hardware. Science, 2021. 372(6539): p. eabb2823.DOI
- [10]Arute, F., et al., Quantum supremacy using a programmable superconducting processor. Nature, 2019. 574(7779): p. 505-510.DOI
- [11]Koch, J., et al., Charge-insensitive qubit design derived from the Cooper pair box. Physical Review A, 2007. 76(4): p. 042319.DOI
- [12]Harvey, S., et al., Coupling two spin qubits with a high-impedance resonator. Physical Review B, 2018. 97(23): p. 235409.
- [13]Behera, B.K., et al., Designing quantum router in IBM quantum computer. Quantum Information Processing, 2018. 18(11): p. 1-13.
- [14]McRae, C.R.H., et al., Materials loss measurements using superconducting microwave resonators. Review of Scientific Instruments, 2020. 91(9): p. 091101.DOI
- [15]Burnett, J., et al. Noise and loss of superconducting aluminium resonators at single photon energies. in Journal of Physics: Conference Series. 2018. IOP Publishing.DOI
- [16]Wisbey, D.S., et al., Effect of metal/substrate interfaces on radio-frequency loss in superconducting coplanar waveguides. Journal of Applied Physics, 2010. 108(9): p. 093918.DOI
- [17]Higginbotham, A.P., et al., Parity lifetime of bound states in a proximitized semiconductor nanowire. Nature Physics, 2015. 11(12): p. 1017-1021.DOI
- [18]Burnett, J., L. Faoro, and T. Lindström, Analysis of high quality superconducting resonators: consequences for TLS properties in amorphous oxides. Superconductor Science and Technology, 2016. 29(4): p. 044008.DOI
- [19]Oliver, W.D. and P.B. Welander, Materials in superconducting quantum bits. MRS bulletin, 2013. 38(10): p. 816-825.
- [20]O’Connell, A.D., et al., Microwave dielectric loss at single photon energies and millikelvin temperatures. Applied Physics Letters, 2008. 92(11): p. 112903.
- [21]Macha, P., et al., Losses in coplanar waveguide resonators at millikelvin temperatures. Applied Physics Letters, 2010. 96(6): p. 062503.DOI
- [22]Woods, W., et al., Determining interface dielectric losses in superconducting coplanar-waveguide resonators. Physical Review Applied, 2019. 12(1): p. 014012.
- [23]Wang, H., et al., Improving the coherence time of superconducting coplanar waveguide resonators. Applied Physics Letters, 2009. 95(23): p. 233508.DOI
- [24]Lahtinen, V. and M. Möttönen, Effects of device geometry and material properties on dielectric losses in superconducting coplanar-waveguide resonators. Journal of Physics: Condensed Matter, 2020. 32(40): p. 405702.DOI
- [25]Sage, J.M., et al., Study of loss in superconducting coplanar waveguide resonators. Journal of Applied Physics, 2011. 109(6): p. 063915.
- [26]Verjauw, J., et al., Investigation of microwave loss induced by oxide regrowth in high-Q niobium resonators. Physical Review Applied, 2021. 16(1): p. 014018.DOI
- [27]Carter, F.W., et al., Low-loss single-photon NbN microwave resonators on Si. Applied Physics Letters, 2019. 115(9): p. 092602.DOI
- [28]Yu, C.X., et al., Magnetic field resilient high kinetic inductance superconducting niobium nitride coplanar waveguide resonators. Applied Physics Letters, 2021. 118(5): p. 054001.DOI
- [29]Sheagren, C., et al., Atomic layer deposition niobium nitride films for high-Q resonators. Journal of Low Temperature Physics, 2020. 199(3): p. 875-882.
- [30]Foshat, P., et al., Nitride-based superconducting microwave coplanar waveguide resonators with internal quality factors above one million for circuit quantum electrodynamics. Bulletin of the American Physical Society, 2023.
- [31]Li, H., et al., Experimental demonstrations of high-Q superconducting coplanar waveguide resonators. Chinese Science Bulletin, 2013. 58(20): p. 2413-2417.
- [32]De Visser, P., et al., Evidence of a nonequilibrium distribution of quasiparticles in the microwave response of a superconducting aluminum resonator. Physical review letters, 2014. 112(4): p. 047004.
- [33]Valenti, F., et al., Interplay between kinetic inductance, nonlinearity, and quasiparticle dynamics in granular aluminum microwave kinetic inductance detectors. Physical review applied, 2019. 11(5): p. 054087.DOI
- [34]Goetz, J., et al., Loss mechanisms in superconducting thin film microwave resonators. Journal of Applied Physics, 2016. 119(1): p. 015304.
- [35]Amin, K.R., et al., Loss mechanisms in TiN high impedance superconducting microwave circuits. Applied Physics Letters, 2022. 120(16): p. 164001.
- [36]Nsanzineza, I. and B. Plourde, Trapping a single vortex and reducing quasiparticles in a superconducting resonator. Physical review letters, 2014. 113(11): p. 117002.
- [37]Gustavsson, S., et al., Suppressing relaxation in superconducting qubits by quasiparticle pumping. Science, 2016. 354(6319): p. 1573-1577.DOI
- [38]Taupin, M., et al., Tunable quasiparticle trapping in Meissner and vortex states of mesoscopic superconductors. Nature communications, 2016. 7(1): p. 1-7.DOI
- [39]Aguado, R., A perspective on semiconductor-based superconducting qubits. Applied Physics Letters, 2020. 117(24): p. 240501.DOI
- [40]Delfanazari, K., et al., On-Chip Andreev Devices: Hard Superconducting Gap and Quantum Transport in Ballistic Nb–In0.75Ga0.25As-Quantum-Well–Nb Josephson Junctions. Advanced Materials, 2017. 29(37): p. 1701836.DOI
- [41]Delfanazari, K., et al., On-chip hybrid superconducting-semiconducting quantum circuit. IEEE Transactions on Applied Superconductivity, 2018. 28(4): p. 1-4.DOI
- [42]Aasen, D., et al., Milestones toward Majorana-based quantum computing. Physical Review X, 2016. 6(3): p. 031016.DOI
- [43]Prada, E., et al., From Andreev to Majorana bound states in hybrid superconductor–semiconductor nanowires. Nature Reviews Physics, 2020. 2(10): p. 575-594.
- [44]Chen, J., et al., Ubiquitous non-Majorana zero-bias conductance peaks in nanowire devices. Physical Review Letters, 2019. 123(10): p. 107703.
- [45]Zuo, K., et al., Supercurrent interference in few-mode nanowire Josephson junctions. Physical review letters, 2017. 119(18): p. 187704.DOI
- [46]Estrada Saldaña, J., et al., Charge localization and reentrant superconductivity in a quasi-ballistic InAs nanowire coupled to superconductors. Science advances, 2019. 5(7): p. eaav1235.DOI
- [47]Kanne, T., et al., Epitaxial Pb on InAs nanowires for quantum devices. Nature Nanotechnology, 2021. 16(7): p. 776-781.DOI
- [48]Bothner, D., et al., Improving the performance of superconducting microwave resonators in magnetic fields. Applied Physics Letters, 2011. 98(10): p. 102504.DOI
- [49]Borisov, K., et al., Superconducting granular aluminum resonators resilient to magnetic fields up to 1 Tesla. Applied Physics Letters, 2020. 117(12): p. 120502.DOI
- [50]Bothner, D., et al., Improving superconducting resonators in magnetic fields by reduced field focussing and engineered flux screening. Physical Review Applied, 2017. 8(3): p. 034025.DOI
- [51]Baity, P.G., et al., Strong magnon–photon coupling with chip-integrated YIG in the zero-temperature limit. Applied Physics Letters, 2021. 119(3): p. 033502.DOI
- [52]Banerjee, A., et al., Optical properties of refractory metal based thin films. Optical Materials Express, 2018. 8(8): p. 2072-2088.
- [53]Petersan, P.J. and S.M. Anlage, Measurement of resonant frequency and quality factor of microwave resonators: Comparison of methods. Journal of applied physics, 1998. 84(6): p. 3392-3402.
- [54]Deng, C., M. Otto, and A. Lupascu, An analysis method for transmission measurements of superconducting resonators with applications to quantum-regime dielectric-loss measurements. Journal of Applied Physics, 2013. 114(5): p. 054504.DOI
- [55]Khalil, M.S., et al., An analysis method for asymmetric resonator transmission applied to superconducting devices. Journal of Applied Physics, 2012. 111(5): p. 054510.DOI
- [56]Gao, J., The physics of superconducting microwave resonators. 2008: California Institute of Technology.
- [57]Sanchez, M., E. Martin, and J. Zamarro. New vectorial automatic technique for characterisation of resonators. in IEE Proceedings H (Microwaves, Antennas and Propagation). 1989. IET.DOI
- [58]Probst, S., et al., Efficient and robust analysis of complex scattering data under noise in microwave resonators. Review of Scientific Instruments, 2015. 86(2): p. 024706.DOI
- [59]Bruno, A., et al., Reducing intrinsic loss in superconducting resonators by surface treatment and deep etching of silicon substrates. Applied Physics Letters, 2010. 106(18): p. 182601.
- [60]Göpl, M., et al., Coplanar waveguide resonators for circuit quantum electrodynamics. Journal of Applied Physics, 2008. 104(11): p. 113904.
- [61]Annunziata, A.J., et al., Tunable superconducting nanoinductors. Nanotechnology, 2010. 21(44): p. 445202.DOI
- [62]Niepce, D., J. Burnett, and J. Bylander, High kinetic inductance Nb N nanowire superinductors. Physical Review Applied, 2019. 11(4): p. 044014.DOI
- [63]Wei, X.-Y., et al., Compact NbN resonators with high kinetic inductance. Chinese Physics B, 2020. 29(12): p. 128401.
- [64]Barends, R., et al., Minimizing quasiparticle generation from stray infrared light in superconducting quantum circuits. Applied Physics Letters, 2011. 99(11): p. 113507.DOI
- [65]Scigliuzzo, M., et al., Phononic loss in superconducting resonators on piezoelectric substrates. New Journal of Physics, 2020. 22(5): p. 053027.DOI
- [66]Götz, K., et al., Co-sputtered MoRe thin films for carbon nanotube growth-compatible superconducting coplanar resonators. Nanotechnology, 2016. 27(13): p. 135202.DOI
- [67]Alexander, A., C.G. Weddle, and C.J. Richardson, Power and temperature dependent model for High Q superconductors. arXiv preprint arXiv:2205.06291, 2022.DOI
- [68]Kuit, K., et al., Vortex trapping and expulsion in thin-film YBa 2 Cu 3 O 7− δ strips. Physical Review B, 2008. 77(13): p. 134504.
- [69]Stan, G., S.B. Field, and J.M. Martinis, Critical field for complete vortex expulsion from narrow superconducting strips. Physical review letters, 2004. 92(9): p. 097003.
- [70]Samkharadze, N., et al., High-kinetic-inductance superconducting nanowire resonators for circuit QED in a magnetic field. Physical Review Applied, 2016. 5(4): p. 044004.DOI
- [71]Engel, A., et al., Temperature-dependence of detection efficiency in NbN and TaN SNSPD. IEEE transactions on applied superconductivity, 2013. 23(3): p. 2300505-2300505.DOI
- [72]Knehr, E., et al., Nanowire single-photon detectors made of atomic layer-deposited niobium nitride. Superconductor Science and Technology, 2019. 32(12): p. 125007.DOI



