Q-Slope Studies at Fermilab: New Insight From Cavity and Cutouts - - PowerPoint PPT Presentation

Q-Slope Studies at Fermilab: New Insight From Cavity and Cutouts Investigations

• A. Romanenko

Fermilab

Outline

• New experimental findings on Q slopes

– Decomposition of the components of surface resistance (RBCS and Rres)

• Shows which Q slope is due to what component
• New superconducting measurements

– Low energy muon spin rotation

• Baked/unbaked cutouts
• N doped
• New proximity effect model of the high field Q slope

– Evidence from cryogenic TEM investigations in cutouts

• New model of the 120C baking

– Vacancy-based 120C baking mechanism and supporting evidence from cutouts – Suppression of the second phase of hydrides in direct observations

• Conclusions

Decomposition of Rs into components

• Using different temperature dependence to

deconvolute the components of average surface resistance at ALL fields

September 30, 2013 Alexander Romanenko 3

Rs(T) = RBCS(T) + Rres

Due to thermally excited quasiparticles

Non-T-dependent, saturation value at T -> 0

Rs(B) decomposition

September 30, 2013 Alexander Romanenko 4

5 10 15 20 25 30 35 0.0 2.0x10

10

4.0x10

10

6.0x10

10

Q0 Eacc (MV/m)

Measure Q(Eacc,T) at many different T<2.17K and Eacc

1.5 1.6 1.7 1.8 1.9 2.0 4 6 8 10 12 14 16 18 20

3 5 7 9 11 13 15 17 19 21 23 25 27 29

Rs (nOhm) Temperature (K)

Can be fitted using both approximate formula RBCS(T)=A/T exp(-⊗/kT), and by more precise BCS calculation based on Halbritter’s program – virtually no difference in the results

Fit a set of Rs(T) curves to extract Rres at each Eacc

• A. Romanenko and A. Grassellino, Appl. Phys. Lett. 102, 252603 (2013)

Residual resistance

September 30, 2013 Alexander Romanenko 5

5 10 15 20 25 30 1 10

BCP EP BCP+120C EP+120C

R0 Eacc (MV/m)

High field Q slope is clearly a residual resistance effect

Contributes to the medium field Q slope

For some treatments decreases at lower fields

• A. Romanenko and A. Grassellino, Appl. Phys. Lett. 102, 252603 (2013)

BCS resistance

September 30, 2013 Alexander Romanenko 6

5 10 15 20 25 30 6 8 10 12 14

BCP EP BCP+120C EP+120C

Rbcs2K (nOhm) Eacc (MV/m)

Typically cited effect

• f 120C

baking on the BCS surface resistance

A strong change in the field dependence due to 120C bake

Unbaked Baked at 120C More on the medium field Q slope – hot topic session on Thursday

• A. Romanenko and A. Grassellino, Appl. Phys. Lett. 102, 252603 (2013)

SC gap change with field

September 30, 2013 Alexander Romanenko 7

Field dependence

• f RBCS may be

explained by the expected changes

• f pairing potential

Δ=Δ (H) in clean (unbaked) and dirty (120C baked) limits

mfp << ξ mfp >> ξ

Role of thermal “feedback”

September 30, 2013 Alexander Romanenko 8

20 40 60 80 100 120 1 2 3 4 5 6 7

EP+120C, 2K - heating effect EP+120C, 1.66K - heating effect BCP+120C, 2K EP+120C, 2K EP+800C+EP+120C, 2K

∆RBCS (nΩ) B (mT)

Negligible effect on RBCS at T <= 2K

Instead of modeling the full temperature transfer with only Rs=G/Q0 as an input use temperature mapping to measure the outside wall temperature More – hot topic session on Thursday

Correlation between medium and high field Q slopes in unbaked cavities

September 30, 2013 Alexander Romanenko 9

More info – please see [A. Romanenko et al, TUP101]

T-map data shows that local surface resistance in HFQS regime is highly correlated to Rs at lower fields (MFQS)

New cavity data allows to “filter” models

• High field Q slope is due to residual

– Not SC gap closing, thermal feedback etc.

• Medium field Q slope is a combination of RBCS

and Rres

– Not due to the difference in Trf and Tbath – Correlation between high and medium fields in unbaked cavities

• Low field Q slope is likely due to residual

September 30, 2013 Alexander Romanenko 10

New superconducting measurements

• Bulk muon spectroscopy

– A. Grassellino et al, TUP031

• Low energy muon spectroscopy

– A. Romanenko et al, TUP038

• Bitter decoration

– F. Barkov et al, TUP016

September 30, 2013 Alexander Romanenko 11

Muon spin rotation

Bµ

Contains physics

aG(t) ~ Muon Spin Polarization

Frequency – field amplitude Damping – field non-uniformity

Muon spin rotation – measure B(z)

1 2 3 4 5 6 7 8 9 10

• 1.0
• 0.8
• 0.6
• 0.4
• 0.2

0.0 0.2 0.4 0.6 0.8 1.0

Muon Spin Polarisation Time (µs)

1 2 3 4 5 6 7 8 9 10

• 1.0
• 0.8
• 0.6
• 0.4
• 0.2

0.0 0.2 0.4 0.6 0.8 1.0

Muon Spin Polarisation Time (µs)

1 2 3 4 5 6 7 8 9 10

• 1.0
• 0.8
• 0.6
• 0.4
• 0.2

0.0 0.2 0.4 0.6 0.8 1.0

Muon Spin Polarisation Time (µs)

B(z) z

Superconductor in the Meissner State

Bext 

LEM – data on EP baked/unbaked

20 40 60 80 0.0 0.5 1.0

EP 120 um + BCP 10 um finish EP 120 um EP 120 um + 120C bake Nitrogen treatment

B/Ba Average depth (nm)

Ba = 25 mT

10 20 30 40 50 60 70 80 90 100 110 120 130 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07

25.3 keV 20 keV 17.5 keV 15 keV 12.5 keV 10 keV 7.5 keV 5 keV Normalized stopping distribution (nm

• 1)

Depth (nm) 3.3 keV

Use variable energy muons, which stop in the first ~100nm

Fit by Gaussian model for the field at the muon site – approximate, qualitative comparison

BCP and EP unbaked -> strong screening, excellent fit provided by the clean limit Pippard/BCS model EP+120C bake-> strongly suppressed m.f.p., gradient of the m.f.p. from the surface, dirty limit N-doped -> intermediate m.f.p., no gradient

mfp ~ 2 nm at the surface, increasing deeper

~15 nm - no screening

mfp~40 nm mfp > 400 nm

New model of the HFQS

• Main element: presence of small proximity

effect coupled nanohydrides within the penetration depth

– Q disease “in miniature”

• Consistent with all experiments, provides

quantitative description

• Falsifiable

– Testable predictions

September 30, 2013 Alexander Romanenko 15

• A. Romanenko, F. Barkov, L. D. Cooley, A. Grassellino, Supercond. Sci. Technol. 26 (2013) 035003

Neither standard 800C degassing nor “fast” cooldown help

September 30, 2013 Alexander Romanenko 16

Integrate the H diffusion over the time spent in the precipitation temperature range T < 160K => L > 1 um

All free near-surface H will precipitate into hydrides

Typical fast cooldown of a cavity (FNAL)

• C. Antoine et al, SRF’01
• T. Tajima et al, SRF’03

Near-surface H-rich layer is still there after standard H degassing treatments

Nanohydrides upon cooldown

Not 120C baked sample

“fast” cooldown

T= 300K

Interstitial hydrogen

~50 nm

Oxide

T= 2K

Niobium hydrides Oxide

Note drastic change in the hydrogen-related m.f.p.

Proximity effect model

September 30, 2013 Alexander Romanenko 18

FH

Cumulative distribution function of proximity- controlled breakdown fields of hydrides

Shape is determined by the distribution of hydride critical fields Hb

Rs~ R0 + Rn * FH(Ha)

• Normal conducting

hydrides of size d are superconducting by proximity effect up to the field Hb ~ 1/d

Q disease High field Q slope

• Excellent

fits

H

• A. Romanenko, F. Barkov, L. D. Cooley, A. Grassellino, Supercond. Sci. Technol. 26 (2013) 035003

• So what happens with 120C bake?

September 30, 2013 Alexander Romanenko 19

• A. Romanenko, C. J. Edwardson, P. G. Coleman, P. J. Simpson, Appl. Phys. Lett. 102, 232601 (2013)

Positron annihilation on cavity cutouts

BCP EP

Large grain Fine grain

• Positron annihilation spectroscopy: 120C baking results in “doping” of the

first ~50 nm from the surface with defects, most likely vacancies

– EP itself introduces some vacancies in ~1 um – may be the reason for more efficient 120C baking in EP cavities

Effect of 120C baking

September 30, 2013 Alexander Romanenko 21

T= 300K

Free interstitial hydrogen

~50 nm

Oxide 120C baking

T= 300K

Hydrogen is trapped by vacancies

Oxide

• A. Romanenko, C. J. Edwardson, P. G. Coleman, P. J. Simpson, Appl. Phys. Lett. 102, 232601 (2013)

Effect of 120C baking

September 30, 2013 Alexander Romanenko 22

“fast” cooldown

Cooling down of 120C baked niobium

T= 2K

No/smaller hydrides are formed due to significant portion of hydrogen trapped

Oxide

T= 300K

Oxide

Note no change in the hydrogen-related m.f.p. – remains low

TEM evidence for nanohydrides

• Direct imaging of the cross-sections of cavity cutouts in cryo-TEM [see Y. Trenikhina

et al, TUP043]

September 30, 2013 Alexander Romanenko 23

Look at this area with subnanometer resolution in TEM at room AND T<100K temperatures

See also R. Tao et al, J. Appl. Phys. 114, 044306 (2013) and TUP042 for cryoimaging of H-reach Nb samples

TEM

Direct evidence for nanohydrides

September 30, 2013 Alexander Romanenko 24

• Y. Trenikhina et al, TUP043

Direct observation of large hydrides

t=0

• F. Barkov et al, TUP014

1 min 2 min

5 min

15 min 45 min 100 min

3 hr

Growing of hydrides at T=160K in a mechanically polished sample

Further evidence: 100K and 120C baking effect

• Second phase (lower concentration, lower

temperature) forms at 100K

– NOT observed on 120C baked samples

September 30, 2013 Alexander Romanenko 26

T=110K T=100K

Summary

• Both residual and BCS surface resistances carry a field

dependence

– Analysis of Q slopes should only be done on components

• Mean free path/ Meissner screening is lowest, depth-

dependent in 120C baked material, highest in unbaked, N- doping leads to the “intermediate” situation

• Nanohydrides may be an omnipresent entity not appreciated

before

– May be THE cause of the high field Q slope

• Proximity-induced superconductivity breaks down at lower fields than

host (Nb)

– May be related to the residual resistance field dependence

• Dominant source of the medium field Q slope in unbaked cavities

– Absence of nanohydrides may be behind the effect of doping – Plausible mechanism of 120C baking -> trapping of hydrogen by vacancies -> preventing/decreasing size of nanohydrides

September 30, 2013 Alexander Romanenko 27

Acknowledgments

September 30, 2013 Alexander Romanenko 28

• FNAL: F. Barkov, A. Grassellino, A.

Crawford, D. Sergatskov, O. Melnychuk, R. Pilipenko

• IIT/FNAL: Y. Trenikhina
• IIT: J. Zasadsinski
• Univ. of Chicago: S. Antipov
• Cornell Univ.: H. Padamsee