Superconducting qubits have arisen as a leading technology platform for quantum computing, which is on the verge of revolutionizing the world's calculation capacities. Nonetheless, the fabrication of computationally reliable qubit circuits requires increasing the quantum coherence lifetimes, which are predominantly limited by the dissipations of two-level system defects present in the thin superconducting film and the adjacent dielectric regions. In this paper, we demonstrate the reduction of two-level system losses in three-dimensional superconducting radio frequency niobium resonators by atomic layer deposition of a 10 nm aluminum oxide Al2O3 thin films, followed by a high vacuum heat treatment at 650 °C for few hours. By probing the effect of several heat treatments on Al2O3-coated niobium samples by x-ray photoelectron spectroscopy plus scanning and conventional high resolution transmission electron microscopy coupled with electron energy loss spectroscopy and energy dispersive spectroscopy, we witness a dissolution of niobium native oxides and the modification of the Al2O3-Nb interface, which correlates with the enhancement of the quality factor at low fields of two 1.3 GHz niobium cavities coated with 10 nm of Al2O3.

Superconducting radio frequency (SRF) resonators, historically used to accelerate particles and routinely achieving very high quality factors Q > 1010–1011, are finding a different use in the quantum regime (at temperatures below 1 K) whether to be integrated into 3D quantum processing units1,2 or to be used as quantum sensors to search for dark matter and gravitational waves.3 The motivation behind these recent applications is that SRF cavities benefit from a 1000-fold higher coherence lifetime than other 2D qubit architectures and can offer sensitivities orders of magnitude higher, which would greatly enhance many fundamental physics experiments.4 Nonetheless, SRF cavities still suffer from dielectric losses arising from two-level systems (two-state defects) dissipations occurring in the native oxide layers that form once the superconductor is exposed to air. Regardless of their chemical composition, amorphous solids exhibit universal behavior at low temperature5 caused by the presence of two-level system (TLS) defects within the material. Because of their low energy, such defects are saturated at high temperature. However, once the material is cooled to few kelvins, these additional degrees of freedom become available and are a major source of noise and decoherence in superconducting quantum devices.5 Niobium, which is a commonly used material for these resonators, is known to grow an amorphous oxide layer once exposed to air. Recent work on three-dimensional 3D niobium (Nb) resonators in the low field regimes has shown that the niobium native oxide is a major source of TLS losses4,6 responsible for the degradation in the quality factors from 6.1010 to 2.1010 in 1.3 GHz cavities. The mitigation of TLS dissipations in 3D superconducting resonators is a fresh subject of research. So far, the only reported improvement in the quality factor of an SRF cavity in the quantum regime has been achieved by applying a high vacuum (HV) annealing step at 340 °C for 5 h and keeping the cavity under vacuum to prevent its re-oxidation.7 While this annealing step is effective, the necessity to sustain a vacuum environment makes it unpractical for quantum computing applications or quantum sensing. Another approach has been investigated by Bal et al8 in which a metallic niobium film in a qubit structure has been encapsulated with an in situ deposited tantalum film. This approach improved the coherence times by a factor of 2 to 5 but requires an oxide-free niobium surface to begin with.

In this paper, we investigate an approach combining an atomic layer deposition (ALD) coating on air-exposed niobium surfaces, followed by a subsequent thermal treatment in HV to dissolve the initially present niobium native oxides. Such an approach was initially proposed by Proslier et al.9,10 in order to get rid of magnetic impurities present in niobium sub-oxides believed to limit the performances of accelerating cavities; the inner surface of the Nb cavity was coated with a thin protective layer of alumina, followed by a subsequent annealing at 450 °C for 24 h in HV. This resulted in the reduction of Nb2O5 into niobium sub-oxides by oxygen diffusion into the bulk Nb, while the Al2O3 layer protected the metal layer from further oxidation. The corresponding RF tests showed an increase in Q due to the improvement of the superconducting surface properties, but no investigation was made at that time of the low-field regime.

In this study, we opt for higher annealing temperatures, up to 650 °C, in order to promote further reduction of the initial native oxide and reduce the presence of sub-oxides. To this end, x-ray photoelectron spectroscopy (XPS) measurements and TEM analysis were performed on air-exposed cavity-grade niobium coupons coated with a Al2O3 layer deposited by ALD. ALD is a self-limiting, sequential surface chemistry that has the unique ability to achieve uniform atomic-scale thickness control on complex-shaped substrates.11 Pieces of polycrystalline Nb were cut from larger sheets used to fabricate SRF cavities. These pieces were electro-polished (EP) in a manner similar to that done on SRF cavities12 with an etching of at least 400 μm insuring the removal of the surface damaged layer,13 cleaned in an ultrasonic bath and dried in air. They were later introduced into the ALD reactor and coated with 10 nm of Al2O3 at a temperature of 250 °C under a flow of ultra-pure nitrogen gas using a standard ALD recipe of 100 cycles of alternating trimethylaluminium (TMA) and water H2O.14,15 Both precursors were pulsed for 1 s and purged for 15 s. These deposition parameters resulted in a growth rate of 0.12 nm per cycle on niobium, in agreement with literature values.15 More details on the ALD system used in this work can be found in Ref. 16. The coated samples were then baked in HV (pressure <10−6 mbar) at temperature of 650 °C for several hours.

In order to assess this approach for the low-field performances of 3D Nb resonators, a 1.3 GHz niobium cavity was coated with a 10 nm Al2O3 film using the same deposition parameters used on the samples and post-annealed in HV at 650 °C for 4 h. As a second test, we electropolished the niobium cavity to remove the Al2O3 layer and reset the niobium surface. We then tested the cavity to have a baseline and performed the same Al2O3 coating; however, this time, we annealed the cavity at 650 °C for 10 h. Prior to each RF test, the cavity was subject to a high pressure rinsing (HPR) with ultra-pure water.17 The RF tests were conducted in the synergium vertical-testing facility at CEA. The low-field region quality factor has been measured following the cavity ringdown procedure after shutting off RF power.18 These tests are shown in Fig. 1.

FIG. 1.

RF tests results showing the quality factor Q vs the accelerating gradient E at 1.5 K in the low field regime on a Nb cavity coated with Al2O3 (10 nm) and post-annealed at 650 °C for (a) 4 h and (b) 10 h. The baseline of EP cavity is in green and the curve of the Al2O3 coated cavities in blue. The red curves represent the TLS fit using Eq. (7) from Ref. 19.

FIG. 1.

RF tests results showing the quality factor Q vs the accelerating gradient E at 1.5 K in the low field regime on a Nb cavity coated with Al2O3 (10 nm) and post-annealed at 650 °C for (a) 4 h and (b) 10 h. The baseline of EP cavity is in green and the curve of the Al2O3 coated cavities in blue. The red curves represent the TLS fit using Eq. (7) from Ref. 19.

Close modal

The baseline RF tests (green curves) show typical performances for air-exposed EP 1.3 GHz niobium cavities without post thermal treatments with Q values at low field ranging between 1.5 and 2.5 × 1010 that are consistent with values obtained in Refs. 6 and 7. Such variations of the quality factor at low fields can be due to slightly different niobium oxide composition and thickness. The blue curves represent the RF tests after deposition of 10 nm of Al2O3 and subsequent thermal treatments. It is clear that for both tests, the quality factor has been improved by a factor of two at low fields. The red lines are fits using the interacting and non-interacting two level system (TLS) model described in Ref. 19. The fitting and the extracted TLS parameters are summarized in Table I. The fit parameters for the baseline EP are consistent with the ones obtained from Ref. 19 for a niobium EP cavity with very similar RF performances.

TABLE I.

Fitting parameters from Eq. (7) of Ref. 19 and TLS parameters extracted from the fits.

Fitting parameters TLS parameters
Treatments c (C2/J) ξ EC (V/m) 1/Qnon-TLS √(T1 T2) (s) σTLS (cm−2) tan(δTLS), ε, d(nm)
Baseline EP  9.0 ± 0.1 × 10−24  140 ± 10  2 ± 1 × 104  2.7 ± 0.1 × 10−11  9 × 10−10  2.5 × 1011  1.5 × 10−3, 30, 5 
Al2O3 10 nm + 650 °C-4 h  3.8 ± 0.2 × 10−24  500 ± 50  9 ± 1 × 102  2.1 ± 0.1 × 10−11  1.4 × 10−8  8.5 × 1010  7.7 × 10−4, 10, 10 
Baseline EP  4.8 ± 0.2 × 10−24  40 ± 2  3 ± 1 × 104  2.5 ± 0.1 × 10−11  4 × 10−10  1.1 × 1011  6.9 × 10−4, 30, 5 
Al2O3 10 nm + 650 °C-10 h  1.6 ± 0.1 × 10−24  750 ± 50  4 ± 1 × 102  1.7 ± 0.1 × 10−11  3.2 × 10−8  3.5 × 1010  3.2 × 10−4, 10, 10 
Fitting parameters TLS parameters
Treatments c (C2/J) ξ EC (V/m) 1/Qnon-TLS √(T1 T2) (s) σTLS (cm−2) tan(δTLS), ε, d(nm)
Baseline EP  9.0 ± 0.1 × 10−24  140 ± 10  2 ± 1 × 104  2.7 ± 0.1 × 10−11  9 × 10−10  2.5 × 1011  1.5 × 10−3, 30, 5 
Al2O3 10 nm + 650 °C-4 h  3.8 ± 0.2 × 10−24  500 ± 50  9 ± 1 × 102  2.1 ± 0.1 × 10−11  1.4 × 10−8  8.5 × 1010  7.7 × 10−4, 10, 10 
Baseline EP  4.8 ± 0.2 × 10−24  40 ± 2  3 ± 1 × 104  2.5 ± 0.1 × 10−11  4 × 10−10  1.1 × 1011  6.9 × 10−4, 30, 5 
Al2O3 10 nm + 650 °C-10 h  1.6 ± 0.1 × 10−24  750 ± 50  4 ± 1 × 102  1.7 ± 0.1 × 10−11  3.2 × 10−8  3.5 × 1010  3.2 × 10−4, 10, 10 

The data and the fitting parameters reveal reproducible trends after Al2O3 deposition and annealing as compared to bare niobium with its native oxides, which are explained as follows:

  1. The saturating electrical field Ec is decreased by more than an order of magnitude. Using the relation E c = 3 2 p 1 T 1 T 2, where p e Å is the electric dipole moment and T1 and T2 are the TLS energy relaxation and homogenous broadening times, we can deduce that T 1 T 2 is increased by 15–80 times.

  2. The spectral diffusion parameter, ξ, that describes the TLS impurities' coupling strength increases significantly.

  3. The c parameter that describes the saturating value of the Q at very low electrical fields is also consistently reduced by a factor of 2–3 as compared to the baselines. Using the equations c = π 12 p 2 tanh Ω 0 k b T ρ and tan δ TLS = 4 c ϵ d 2 k b T Ω 0 in the limit E , T 0 and following the procedure of reference,19 the area density of TLS σ TLS and the loss tangent tan δ TLS extracted from the fitting parameter c, the thickness of the oxide d, and the assumed dielectric constants ϵ values19,20 listed in Table I, are systematically reduced by a factor 2 as compared to the baselines.

These reproducible trends point toward a different nature of the TLS impurities in ALD coated and annealed Nb cavities as compared to native niobium oxides present in the cavity baselines. In order to investigate the microscopic origin of these changes in the RF performances and TLS losses, we performed STEM, high-angle annular dark-field imaging (HAADF), high resolution transmission electron microscopy (HRTEM), electron energy loss spectroscopy (EELS), and energy dispersive spectroscopy (EDX) spectral imaging and XPS measurements on niobium samples that underwent the same processes as the cavities. The details on the apparatus, measurement conditions, analysis, and fitting procedures can be found in the supplementary material.

Figure 2 summarizes the results obtained on a cavity-grade electro-polished Nb sample coated with 100 cycles of Al2O3 by ALD without any thermal post-treatment. The high-resolution STEM images show an amorphous 12 nm thick Al2O3 film (light gray) on top of an intermediate 5 nm amorphous layer (darker gray hue) on crystalline Nb [Fig. 2(a)].16,21 The EELS analysis [Fig. 2(b)] reveals that this 5 nm interfacial layer is composed of NbOx (see the supplementary material references for details on how the NbOx component is isolated). The XPS allows for a more detailed chemical analysis and the Nb-3d core levels spectrum and shows that the NbOx is composed mostly of NbO2 (7%), NbO (24%), and Nb2O5 (23%) [in Fig. 2(c) bottom]. This Nb oxide composition differs significantly from a typical un-coated electro-polished Nb sample with a native oxide composition dominated by Nb2O5 (57%) in agreement with previous work22 and displayed in Fig. 2(b) top for comparison. The partial reduction of the Nb2O5 to sub-oxides is caused by the ALD deposition temperature of 250 °C applied for 2–3 h in agreement with previous works.23 

FIG. 2.

(a) Bright field STEM imaging and local FFT analyses, (b) HAADF(top left) and EELS analysis of an as-deposited Al2O3-coated Nb, and (c) XPS spectrum of Nb-3d core levels of a reference EP niobium sample (top). XPS spectrum of Nb 3d core level of as deposited Al2O3-coated Nb (bottom).

FIG. 2.

(a) Bright field STEM imaging and local FFT analyses, (b) HAADF(top left) and EELS analysis of an as-deposited Al2O3-coated Nb, and (c) XPS spectrum of Nb-3d core levels of a reference EP niobium sample (top). XPS spectrum of Nb 3d core level of as deposited Al2O3-coated Nb (bottom).

Close modal

After annealing the Al2O3-coated Nb sample at 650 °C for 4 h, the high resolution STEM and EELS measurements [Fig. 3(a)] reveal an amorphous 10 nm Al2O3 layer on top of an amorphous 1.5–2 nm thick Nb oxide interface. The composition of this oxide measured by XPS [Fig. 3(b)] is NbO (14%) and Nb2O (5%). The thickness reduction of the Nb oxide layer from 5 to 1.5–2 nm along with a strong decrease in the total Nb oxide composition from 75% to 19% of the Nb-3d spectrum after annealing indicates a further reduction and a diffusion of the oxygen from the Nb oxide layer into the bulk Nb.

FIG. 3.

(a) Bright field STEM imaging and local FFT images, (b) HAADF (top left) and EELS analysis, and (c) XPS spectrum of Nb-3d core levels of an Al2O3-coated Nb sample after annealing at 650 °C during 4 h.

FIG. 3.

(a) Bright field STEM imaging and local FFT images, (b) HAADF (top left) and EELS analysis, and (c) XPS spectrum of Nb-3d core levels of an Al2O3-coated Nb sample after annealing at 650 °C during 4 h.

Close modal

Upon increasing the annealing time from 4 to 10 h at 650 °C, the TEM and EDX analyses [Fig. 4(a)] show that the Al2O3 thickness remains unchanged at 10 nm. The Nb oxide layer, however, becomes discontinuous (see S3 in the supplementary material), keeping a thickness of about 2 nm in regions where it is still present. The XPS data analysis reveals an increased Nb2O concentration from 5% to 37%, whereas the NbO decreases from 14% to 7%. The decrease in NbO and increase in Nb2O concentrations indicate a further reduction of NbO into Nb2O upon longer annealing time, but it cannot account, however, for the total concentration of Nb2O measured. We suspect that the Al2O3 interface with the Nb could start releasing partially its oxygen atoms to the Nb underneath under the combined influence of prolonged annealing and high Nb reactivity with O.

FIG. 4.

(a) HRTEM and local FFT images, (b) EDX analyses, and (c) XPS spectrum of Nb-3d core levels of an Al2O3-coated Nb sample after annealing at 650 °C during 10 h.

FIG. 4.

(a) HRTEM and local FFT images, (b) EDX analyses, and (c) XPS spectrum of Nb-3d core levels of an Al2O3-coated Nb sample after annealing at 650 °C during 10 h.

Close modal

The HRTEM images display the crystalline structure of bcc Nb with a strong structural elongation (2%–3%) perpendicular to the surface and indicate local areas with larger d-spacing (0.238–0.241 nm) in comparison with the bulk Nb (0.229–0.236 nm). This distortion could be associated with a local increase in the crystal parameter, related to the inclusion of oxygen into the niobium unit cell, as supported by the XPS data and seen previously.24 

The baseline RF performances at low field are typical for air exposed EP Nb cavities with quality factor values ranging from 1 to 2.5 1010,6,7 and it is, therefore, reasonable to assume that the Nb oxide thickness and structure are similar to what was found in Ref. 24, i.e., a 5 nm thick amorphous Nb2O5 layer on top of a ∼1.5–2 nm sub-oxide NbOx layer. Our microscopic analyses of the Al2O3-coated and annealed niobium surface show a similar NbOx thickness and composition, whereas a 10 nm amorphous Al2O3 film replaces the amorphous Nb2O5. The Nb2O5 removal can, therefore, explain the reproducible improvement of the quality factor at low RF field amplitudes upon annealing at 650 °C for 4 or 10 h and the corresponding reduction by a factor of 2 of the TLS losses tan δ TLS listed in Table I. This result is in agreement with previous works4,6 that emphasize the strong contribution and dependence of RF TLS losses on the thickness of Nb2O5 in SRF cavities. Furthermore, the predominant presence of the amorphous ALD Al2O3 layer also explains the notable changes in the nature of the previously mentioned TLS defects that entail very different values as compared to Nb2O5 for the critical saturation field E c and the measured coupling strength ξ. In order to further identify and separate the TLS losses contribution from the NbOx interface and the Al2O3 layer to the overall quality factor at low RF fields, one could vary the thickness of the Al2O3 film with the same post annealing treatments. Assuming a similar NbOx composition and thickness for different Al2O3 layer thickness, the total cavity losses at low field should depend linearly on the Al2O3 layer thickness with a constant contribution from the NbOx TLS losses. A similar procedure can also be repeated with different capping layers properties, to study, for instance, how the crystallographic properties (amorphous vs crystalline) can affect the TLS losses.

Previous measurements of the loss tangent of ALD-deposited Al2O3 films on superconducting resonators20 give values of ∼2 to 3 × 10−3 for films ranging from 30 to 100 nm, whereas our values are about one order of magnitude lower: ∼3 × 10−4. The HRTEM analysis reveals that the Al2O3 film thickness changes from 12 to 10 nm upon annealing, indicating a densification (in the absence of measurable Al diffusion into the Nb) from 3 ± 0.1 g/cm3 as measured by x-ray reflectivity (XRR) on as-deposited films on Si samples (supplementary material), to an estimated 3.5 g/cm3 on niobium after the thermal treatments. In-depth investigation of ALD synthesized Al2O3 films showed that a significant concentration of hydroxyl groups remains in the film during the growth, contributing to the low film density,25 that are fully eliminated by post-annealing above 1000 °C.26 The hydroxyl groups have been proposed, and in some cases identified,27–30 as a potential source of TLS losses. Their concentration decrease during the annealing step in HV at 650 °C could provide a microscopic origin for the lower loss tangent in our annealed films as compared to the as-deposited Al2O3 measured in Ref. 20.

In order to reduce further the niobium sub-oxide presence at the interface, we have increased the post-annealing temperature to 800 °C for few hours. Subsequent depth-profile XPS measurements revealed the presence of Nb and O (at the surface) without Al, indicating that the Al2O3 film was no longer present at the surface and that it did not diffuse deeper into the Nb (not shown). We suspect that the oxygen diffusion mechanism proposed earlier that emerges at 650 °C during 10 h becomes more severe at 800 °C and the progressive reduction of the Al2O3 film to metallic aluminum causes its evaporation into the vacuum chamber due to its low melting temperature of 660 °C.

In conclusion, we have investigated the effect of Al2O3 deposition by ALD and post-annealing on the RF performances at low fields of Nb 1.3 GHz resonators. This approach resulted in an enhancement of the low-field quality factor caused by a modification of the nature of TLS defects at the surface and a reduction of the loss tangent associated with important modifications of the niobium native oxide chemical composition and structure, as measured by STEM-EELS, HRTEM, and XPS. The creation of a protective Al2O3 layer at the surface enables air-stable performance improvements and facilitates greatly 3D superconducting resonator handling and characterizations. The combined ALD deposition method and thermal treatment approach provide a platform to study the effects of passivation layers with different chemical and structural (crystalline, amorphous, and thickness) properties on the TLS losses and a pathway toward identifying the nature of the layers and interface superconductor/dielectric TLS losses. Due to the unique properties of ALD, this approach can also be applied to different resonator geometries such as co-planar resonators typically used in quantum technology hardware and therefore presents a critical step toward real world applications.

See the supplementary material for experimental details on the x-ray reflectivity measurements (S1), x-ray photoelectron spectroscopy measurements and fitting parameters (S2), and transmission electron microscope analysis.

The authors would like to thank Enrico Cenni from Commissariat de l'Energie Atomique (CEA) for providing the electrical field distribution in a 1.3 GHz cavity for the TLS numerical simulations, Mohammed Fouaidy and Thierry Pepin Donat from IJCLab for providing a HV thermal treatment and Claire Antoine from CEA for the insightful discussions. This work was partly supported by the French RENATECH network (Focused Ion Beam TEM thin foils preparation by David Troadec at IEMN Lille) and by the CNRS-CEA METSA French network (FR CNRS 3507) for the STEM experiments at LPS. This project has received funding from the region Ile de France project SESAME AXESRF, the European Union's Horizon 2020 Research and Innovation programme under Grant Agreement Nos. 101004730 and 730871.

Authors Y.K. and T.P. have patent No. PCT/FR2023/051937 pending.

Y. Kalboussi: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Resources (equal); Validation (equal); Visualization (equal); Writing – original draft (equal). B. Delatte: Resources (equal). S. Bira: Resources (equal). K. Dembele: Data curation (equal); Formal analysis (equal); Investigation (equal); Writing – review & editing (equal). X. Li: Data curation (equal); Formal analysis (equal); Investigation (equal). F. Miserque: Investigation (equal); Writing – review & editing (equal). N. Brun: Data curation (equal); Formal analysis (equal); Investigation (equal); Writing – review & editing (equal). M. Walls: Data curation (equal); Formal analysis (equal); Investigation (equal); Writing – review & editing (equal). J.-L. Maurice: Data curation (equal); Formal analysis (equal); Investigation (equal). D. Dragoe: Investigation (equal). J. Leroy: Investigation (equal). D. Longuevergne: Resources (supporting). A. Gentils: Investigation (equal). S. Jublot-Leclerc: Investigation (equal). G. Jullien: Resources (equal). F. Eozénou: Resources (equal). M. Baudrier: Investigation (equal). L. Maurice: Investigation (equal). T. Proslier: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (lead); Investigation (equal); Methodology (equal); Supervision (lead); Validation (equal); Visualization (equal); Writing – original draft (equal).

The data that support the findings of this study are available from the corresponding authors upon reasonable request.

1.
S.
Chakram
,
A. E.
Oriani
,
R. K.
Naik
,
A. V.
Dixit
,
K.
He
,
A.
Agrawal
,
H.
Kwon
, and
D. I.
Schuster
, “
Seamless high-q microwave cavities for multimode circuit quantum electrodynamics
,”
Phys. Rev. Lett.
127
,
107701
(
2021
).
2.
M. S.
Alam
,
S.
Belomestnykh
,
N.
Bornman
,
G.
Cancelo
,
Y.-C.
Chao
,
M.
Checchin
,
V. S.
Dinh
,
A.
Grassellino
,
E. J.
Gustafson
,
R.
Harnik
et al, “
Quantum computing hardware for hep algorithms and sensing
,” arXiv:2204.08605 (
2022
).
3.
A.
Berlin
,
S.
Belomestnykh
,
D.
Blas
,
D.
Frolov
,
A. J.
Brady
,
C.
Braggio
,
M.
Carena
,
R.
Cervantes
,
M.
Checchin
,
C.
Contreras-Martinez
et al, “
Searches for new particles, dark matter, and gravitational waves with SRF cavities
,” arXiv:2203.12714 (
2022
).
4.
A.
Romanenko
,
R.
Pilipenko
,
S.
Zorzetti
,
D.
Frolov
,
M.
Awida
,
S.
Belomestnykh
,
S.
Posen
, and
A.
Grassellino
, “
Three-dimensional superconducting resonators at T< 20 mK with photon lifetimes up to τ= 2 s
,”
Phys. Rev. Appl.
13
,
034032
(
2020
).
5.
C.
Müller
,
J. H.
Cole
, and
J.
Lisenfeld
, “
Towards understanding two-level-systems in amorphous solids: Insights from quantum circuits
,”
Rep. Prog. Phys.
82
,
124501
(
2019
).
6.
A.
Romanenko
and
D.
Schuster
, “
Understanding quality factor degradation in superconducting niobium cavities at low microwave field amplitudes
,”
Phys. Rev. Lett.
119
,
264801
(
2017
).
7.
D.
Bafia
,
A.
Grassellino
, and
A.
Romanenko
, “
Magnetic suboxides as a source of two-level system losses in superconducting niobium
,” arXiv:2108.13352 (
2021
).
8.
M.
Bal
,
A. A.
Murthy
,
S.
Zhu
,
F.
Crisa
,
X.
You
,
Z.
Huang
,
T.
Roy
,
J.
Lee
,
D.
van Zanten
,
R.
Pilipenko
et al, “
Systematic improvements in transmon qubit coherence enabled by niobium surface encapsulation
,” arXiv:2304.13257 (
2023
).
9.
T.
Proslier
,
J.
Zasadzinski
,
J.
Moore
,
M.
Pellin
,
J.
Elam
,
L.
Cooley
,
C.
Antoine
,
J.
Norem
, and
K. E.
Gray
, “
Improvement and protection of niobium surface superconductivity by atomic layer deposition and heat treatment
,”
Appl. Phys. Lett.
93
(
19
),
192504
(
2008
).
10.
T.
Proslier
,
Y.
Ha
,
J.
Zasadzinski
,
G.
Ciovati
,
P.
Kneisel
,
C.
Reece
,
R.
Rimmer
,
A.
Gurevich
,
L.
Cooley
,
G.
Wu
et al, “
Atomic layer deposition for SRF cavities
,” in
Proceedings of the 23rd Particle Accelerator Conference
(
IEEE
,
2009
) pp.
803
805
.
11.
S. M.
George
, “
Atomic layer deposition: An overview
,”
Chem. Rev.
110
,
111
131
(
2010
).
12.
H.
Diepers
,
O.
Schmidt
,
H.
Martens
, and
F.
Sun
, “
A new method of electropolishing niobium
,”
Phys. Lett. A
37
,
139
140
(
1971
).
13.
C. Z.
Antoine
, “
Materials and surface aspects in the development of SRF niobium cavities
,” in
EuCARD Editorial Series on Accelerator Science and Technology
, edited by
J. P.
Koutchouk
and
R. S.
Romaniuk
(
Varsaw
,
2012
).
14.
S.
George
,
A.
Ott
, and
J.
Klaus
, “
Surface chemistry for atomic layer growth
,”
J. Phys. Chem.
100
,
13121
13131
(
1996
).
15.
M.
Groner
,
F.
Fabreguette
,
J.
Elam
, and
S.
George
, “
Low-temperature Al2O3 atomic layer deposition
,”
Chem. Mater.
16
,
639
645
(
2004
).
16.
Y.
Kalboussi
, Ph.D. thesis,
Université Paris-Saclay
,
2023
, https://theses.hal.science/tel-04116992.
17.
P.
Bernard
,
D.
Bloess
,
W.
Hartung
,
C.
Hauviller
,
W.
Weingarten
,
P.
Bosland
, and
J.
Martignac
, “
Superconducting niobium sputter-coated copper cavities at 1500 MHz
,” in
Proceedings of 3rd European Particle Accelerator Conference
(
IEEE
,
1991
), pp.
1269
1271
.
18.
H.
Padamsee
,
J.
Knobloch
, and
T.
Hays
,
RF Superconductivity for Accelerators
(
John Wiley & Sons
,
2008
).
19.
N.
Gorgichuk
,
T.
Junginger
, and
R.
de Sousa
, “
Modeling dielectric loss in superconducting resonators: Evidence for interacting atomic two-level systems at the Nb/oxide interface
,”
Phys. Rev. Appl.
19
,
024006
(
2023
).
20.
M. R.
Otto
, Ph.D. thesis,
Waterloo University
,
Ontario, Canada
,
2015
.
21.
S.
Bira
, Ph.D. thesis,
Université Paris-Saclay
,
2021
, https://theses.hal.science/tel-03485344.
22.
T.
Hryniewicz
,
K.
Rokosz
, and
H. R.
Zschommler Sandim
, “
SEMP/EDX and XPS studies of niobium after electropolishing
,”
Appl. Surf. Sci.
263
,
357
361
(
2012
).
23.
Q.
Ma
and
R.
Rosenberg
, “
Surface study of niobium samples used in superconducting rf cavity production
,” in
PACS2001. Proceedings of the 2001 Particle Accelerator Conference (Cat. No. 01CH37268)
(
IEEE
,
2001
), Vol.
2
, pp.
1050
1052
.
24.
J.-S.
Oh
,
X.
Fang
,
T.-H.
Kim
,
M.
Lynn
,
M.
Kramer
,
M.
Zarea
,
J. A.
Sauls
,
A.
Romanenko
,
S.
Posen
,
A.
Grassellino
et al, “
In-situ transmission electron microscopy investigation on surface oxides thermal stability of niobium
,”
Appl. Surf. Sci.
627
,
157297
(
2023
).
25.
C.
Guerra-Nuñez
,
M.
Dobeli
,
J.
Michler
, and
I.
Utke
, “
Reaction and growth mechanisms in al2o3 deposited via atomic layer deposition: Elucidating the hydrogen source
,”
Chem. Mater.
29
,
8690
8703
(
2017
).
26.
L.
Gosset
,
J.-F.
Damlencourt
,
O.
Renault
,
D.
Rouchon
,
P.
Holliger
,
A.
Ermolieff
,
I.
Trimaille
,
J.-J.
Ganem
,
F.
Martin
, and
M.-N.
Semeria
, “
Interface and material characterization of thin Al2O3 layers deposited by ALD using TMA/H2O
,”
J. Non-Cryst. Solids
303
,
17
23
(
2002
).
27.
S.
De Graaf
,
A.
Adamyan
,
T.
Lindström
,
D.
Erts
,
S.
Kubatkin
,
A. Y.
Tzalenchuk
, and
A.
Danilov
, “
Direct identification of dilute surface spins on Al2O3: Origin of flux noise in quantum circuits
,”
Phys. Rev. Lett.
118
,
057703
(
2017
).
28.
M. V. P.
Altoé
,
A.
Banerjee
,
C.
Berk
,
A.
Hajr
,
A.
Schwartzberg
,
C.
Song
,
M.
Alghadeer
,
S.
Aloni
,
M. J.
Elowson
,
J. M.
Kreikebaum
et al, “
Localization and mitigation of loss in niobium superconducting circuits
,”
PRX Quantum
3
,
020312
(
2022
).
29.
J. M.
Martinis
,
K. B.
Cooper
,
R.
McDermott
,
M.
Steffen
,
M.
Ansmann
,
K.
Osborn
,
K.
Cicak
,
S.
Oh
,
D. P.
Pappas
,
R. W.
Simmonds
et al, “
Decoherence in Josephson qubits from dielectric loss
,”
Phys. Rev. Lett.
95
,
210503
(
2005
).
30.
S.
de Graaf
,
S.
Un
,
A.
Shard
, and
T.
Lindström
, “
Chemical and structural identification of material defects in superconducting quantum circuits
,”
Mater. Quantum Technol.
2
,
032001
(
2022
).

Supplementary Material