1. Surface Resistance of a bulk-like Nb Film Sarah Aull, Anne-Marie Valente-Feliciano, Tobias Junginger and Jens Knobloch

2. Sample Preparation OFHC copper substrate:
mechanically polished
Electron beam welded to Nb ring (EBW 1)
12 μm electro polishing
Rinsing with ultra pure water at 6 bar
Shipped to Jefferson Lab for coating
Shipped back to CERN, EBW to support structure (EBW 2)
Mounted in the Quadrupole Resonator
EBW 1
EBW 2

3. Deposition Conditions
Cu substrate
Final sulfamic acid rinse for cu passivation
Deposition Conditions
ECR
Bake & coating temperature: 360 °C
Total coating time: 60’
Dual ion energy:
184 eV for nucleation/early growth
64 eV for subsequent growth
Hetero-epitaxial film Nb on OFHC Cu
Typical Cu substrate

4. Film characterization
Witness sample Nb/(11-20) Al2O3
Tc= 9.36 ± 0.12 K
RRR = 179
Diffraction on Nb/Cu witness sample:
EBSD IPF map and XRD pole figure show very good crystallinity and grain sizes in the range of the typical Cu substrate

5. Penetration Depth and R(T) Measurement
λ(0K) [nm]
Rres [nΩ]
Δ [K]
400 MHz
40 ± 2
46.6 ± 0.8
14.2 ± 0.3
800 MHz
38 ± 1
79 ± 2
14.8 ± 0.2
1200 MHz
156 ± 11
15.1 ± 1
ℓ* [nm]
RRR
144 ± 20
53 ± 7
* with λL = 32 nm and ξ0 = 39 nm
Bulk-like film
in the clean limit

6. RRR is unlikely the cause for the strong Q-slope of Nb films.
Q-Slope: film vs. bulk
2.5 K
Q-Slope of Nb film is linear for B > 5 mT for temperatures up to 4 K.
Q-Slope of the Nb film is significantly stronger than for bulk Nb (1 order of magnitude)
RRR is unlikely the cause for the strong Q-slope of Nb films.
4 K

7. Thermal cycling does not affect the (low field) BCS contribution.
Thermal cycling: warm up the sample to the normal conducting state and cool down under different conditions.
Thermal cycling does not affect the (low field) BCS contribution.

8. Influence of the Cooling Conditions
Influence on the surface resistance: Slow uniform cooling increases RS by more than a factor 2.
Nb Film, 400 MHz, 2K, 5 mT
Bulk Nb, 400 MHz, 2.5K, 15 mT

9. Influence of the Cooling Conditions
400 MHz, 2 K
Thermal cycling acts on the Q-slope: The faster the cooling the flatter the slope.

10. Conclusions for the ECR film
This bulk-like Nb film shows significantly different behaviour than bulk Nb with the same RRR:
In contrary to bulk Nb: cooling fast and with a high temperature gradient leads to lower surface resistance.
Lowest surface resistance was achieved by quenching.
The Q-Slope of the film is much more severe than the one of bulk Nb. Therefore the RRR is unlikely the cause for strong Q-slopes in Nb film cavities.
The cooling conditions act on the Q-Slope, leading to better performance after fast cooling.

11. Comparison with HIPIMS coating
Single cell 1.3 GHz Cu cavity + EP
Coating by Giovanni Terenziani
RF Cold test by Tobias Junginger
For more RF results of this cavity, see: HIPIMS Development for Superconducting Cavities, G. Terenziani et al. (TFSRF14)
Lower RS for fast cooling and smaller temperature gradient.
Thermal cycling influences the Q-Slope as well.

12. Comparison with HIE Isolde
Quarterwave, 100 MHz
For more RF results, see The influence of cooldown conditions at transition temperature on the quality factor of niobium sputtered quarter-wave resonators, P. Zhang et al. (TFSRF14)
Lower surface resistance for small temperature gradients.
Cooling rate has no significant influence on RS.
Courtesy of Pei Zhang

13. Comparison between QPR, 1.3 GHz and HIE Isolde
RRR
Geometry
Cooling
Grain size
Quadrupole Resonator:
ECR
Lower RS for fast cooling with T gradient 53
disc
conduction
tens of microns
1.3 GHz: HIPIMS
Lower RS for fast cooling with small T gradient 21
elliptical
Bath cooled
30 nm
HIE Isolde:
Diode sputtering
Lower RS for small T gradients 15
QWR
200 nm – 1 μm depending on thickness
Unknown
Influence of grain size
Influence of geometry
Thermal currents
Influence of stress
Oxidation
Roughness …

14. Conclusions for Nb films
As for bulk Nb: The cooling conditions, speed and/or spatial gradient, influence the RF performance.
Different film projects are difficult to compare due to different coating techniques and geometries.
Optimum cooling procedure to minimize the low field RS is accompanied by a flattened Q-Slope.
Further conclusions require dedicated experiments, where spatial and temporal gradients and thermal currents can be controlled independently.

15. Questions for the TTC meeting
For bulk Nb cavities which perform better after fast cooling:
How do the cavities behave after quench?
For bulk Nb cavities with regular Q-slope:
Do you see an effect on the slope depending on the cooling conditions?
Concerning the mechanism:
If the Q change with cooling conditions is due to flux expulsion, why is the effect in the same order of magnitude for films although they are much less sensitive to trapped flux? (0.1∗ nΩ μT vs. 3.5 nΩ μT )
*Benvenuti et al, SRF97B05
(recent HIE Isolde results
agree well after scaling to 100 MHz)