R EVIEW ON Q - D ROP M ECHANISM B ernard V ISENTIN International Workshop on Thin Films 9 th - 12 th October 2006.

RREVIEW ONEVIEW ON

QQ--DDROP ROP MMECHANISMECHANISM

Bernard VISENTIN

International Workshop on Thin FilmsInternational Workshop on Thin Films

9th - 12th October 2006

2Bernard Visentin International Workshop on Thin Films - Legnaro - October 2006

Definition of Used ParametersDefinition of Used Parameters

Q0 Quality Factor (figure of merit) G Geometry Factor RS Surface Resistance

Eacc Accelerating Field,average electric field seen by particlecrossing cavity gap L

L

zacc dzEL

E0

1

Scycledissipated

stored

R

GQ

/0 2

C1-10 (O1) 1,3 GHz - BCP

1E+09

1E+10

1E+11

0 10 20 30Eacc (MV/m)

Q0

pas d' électronspas rayons X

limitation :- puissance RF- quench thermique

Baking : ~ 110 -120 °C / 2 days ( under UHV )

Annealing : ~ 800 °C - remove Hydrogen from Nb bulk

~ 1350 °C ( +Ti ) - remove Oxygen and improve thermal conductivity (bulk)


Thin Film Cavities & Q-DropThin Film Cavities & Q-Drop

Wuppertal – Nb3Sn / Nb – 1.5 GHz

Vapor Deposition Technique

G. Müller et al.- EPAC (1996)

Saclay – Nb / Cu – 1.5 GHz

Magnetron Sputtering in Ar

P. Bosland et al.- ASC (1998)

( no field emission, no quench only RF power limitation )

CERN – Nb / Cu – 1.5 GHz

Magnetron Sputtering in Kr

V. Arbet-Engels et al.- NIMA (2001)

Advantages to use Thin Film Technology for SRF Cavities :Reduced Cost – New Superconducting Material (higher Tc & Hsh)

severe Q-drop limits High Gradient Performances Eacc < 25 MV/m

Eacc=15 MV/m


1E+09

1E+10

1E+11

0 10 20 30 40

Eacc (V/m)

Qo

S110 - Thin Film Nb/Cu @ 1,5 GHz ............... ( scaled to 1,3 GHz )

C110 - Bulk Nb Cavity - BCP or EP - @ 1,3 GHz

PowerLimitationPower

Limitation

Thin Film & Bulk CavitiesThin Film & Bulk Cavitiesvery steep Q-drop exists on Bulk Cavities (BCP or EP)


very steep Q-drop exists on Bulk Cavities (BCP or EP)

Thin Film & Bulk CavitiesThin Film & Bulk Cavities

It can be cured by baking : limitations in Q0 and Eacc can be exceeded

reason why R&D has been more extended for Nb bulk cavities

1E+09

1E+10

1E+11

0 10 20 30 40

Eacc (V/m)

Qo

S110 - Thin Film Nb/Cu @ 1,5 GHz ............... ( scaled to 1,3 GHz )C110 - Bulk Nb Cavity - BCP or EP - @ 1,3 GHz

C110 - Buffered Chemical Polishing + Baking

C103 - ElectroPolishing + Baking

PowerLimitation

QuenchThermalQuench

PowerLimitation

B. Visentin et al .- EPAC (1998) & SRF (1999)


Thin Film R Thin Film R & & D on CavitiesD on Cavities

R&D gave up now at CERN and at Saclay since 2001

but still continues in Europe (CARE program) and USA (JLab, Cornell)

IPJ Poland / INFN – Nb / Cu – 1.3 GHz

Cylindrical UHV Arc Discharge

magnetic filter –droplet)

J. Langner - CARE Report (2005)

JLab – Nb / Cu – 500 MHz

E-beam evaporation

+ ECR plasma (Nb ionization)

G. Wu - Argonne Workshop (2004) & SRF (2005)


Application of Thin Film CavitiesApplication of Thin Film CavitiesCERN Technology

Large Size (< 700 MHz) with Thick Wall ( 6 mm )Specifications for Low Gradients ( < 15 MV/m)

LHC : 400 MHz

S 3rd H C : 1500 MHz

SOLEIL : 352 MHz

: 200 MHz - fact. - coll.

SLS


•Not a fundamental limitation : improve cleanness during process (substrate, sputtering,…)

•Granular Superconductor Theory : Josephson fluxon penetrationin weak links (grain boundaries oxidized sputter island)

•Thermal resistance at superconductor-substrate interface•Energy Gap dependence (H)

1E+09

1E+10

1E+11

0 10 20 30 40

Eacc (V/m)

Qo

PowerLimitationPower

Limitation

Lot of Theories and Experiments have been

performed on Nb Bulk cavities

Situation Review in bulk case( past + latest results )

Where do we standto understand Q-Drop origin ?

Hope to clear upthe Thin Film issue ???

J. Halbritter - Workshop of the Eloisatron Project (1999) B. Bonin - Supercond. Sci. Technol. 4,257 (1991)

V. Arbet - Engels et al.- NIMA (2001)

not enough data on Thin Film Cavities

Q-Drop Origin ( Thin Film )Q-Drop Origin ( Thin Film )

V. Palmieri - SRF (2005)


Three different slopes in bulk Nb Cavity at

Low Medium High Field

Q-Drops for Bulk CavityQ-Drops for Bulk Cavity

Associated Theories

NbOX Clusters Surface Heating - RS(T) I.T.E , M.F.E. … etc. …

C1-16 ( EP )F1 - no baking

1E+09

1E+10

1E+11

0 10 20 30Eacc ( MV/m )

Q0


D1-22 ( EP )

1E+09

1E+10

1E+11

0 10 20 30Eacc ( MV/m )

Q0

B1 - EP

B2 - Baking

C1 - HF 50'

C2 - Baking

quenchfield

emission

RFpower

Theory : NbOx Clusters in Nb

localized states inside energy gap (Rs)

Baking : Q-Slope enhancement Additional Clusters ( O Diffusion )

HF Rinse (10%) : initial Q-slope restored Phenomenon localized

at Ox./Nb Interface

Nb2O5 + 10 HF → 2 H2NbOF5 + 3 H2O

C1-03 ( EP @ KEK - Tests @ Saclay)I1 = E5 + air exposure 46 months

+ 20' HF + HPR1E+09

1E+10

1E+11

0 10 20 30 40Eacc ( MV/m )

Q0

E5 : ( 120°C / 60 h )

I1 : HF chemistry ( 20 mn )

quench

Low Field Q-DropLow Field Q-DropJ. Halbritter – SRF Workshop (2001)

B. Visentin – Argonne Workshop (2004)


Theory

quadratic dependence :

linear dependence : hysteresis losses due to Josephson fluxons in weak links (oxidation of grain boundaries)

Experimental Checking: quadratic and linear dependence at JLab & DESY (x-cells)

only quadratic dependence at Saclay (1-cell)

Medium Field Q-DropMedium Field Q-Drop

G. Ciovati - Argonne Workshop (2004)

J. Halbritter – 38th INFN Eloisatron Project Workshop (1999) & SRF (2001)

)()1(2

22

2

0 KBCSCC

PSS R

d

kTRB

B

BRR

PS bBaR


( Medium + High ) Field Q-Drop( Medium + High ) Field Q-Drop

)()()1(2

222

2

0 KBCSCC

PSS R

d

kTRB

TC

B

BRR

Thermal Model Refinement non linear correction due to RF pair breaking

Experimental Checking

Thermal Feedback Model

with linear or non linear RBCS

before and after baking.

P. Bauer et al. – SRF (2005)

A. Gurevich – Argonne Workshop (2004)

better fit with non linear model but not enough to explain

the high field Q-drop


High Field Q-Drop ( 6 theories )High Field Q-Drop ( 6 theories )

H. Safa - SRF (2001)

J. Halbritter – Eloisatron Workshop (1999)

B. Bonin - SRF (1995)

J. Knobloch - SRF (1999)

E. Haebel – TTF Meeting (1998)

A. Didenko – EPAC (1996)

Diffusion (O, Imp.) : “ Interface Tunnel Exchange ”

“ Bad Superconducting Layer

”

“ Granular Superconductivity ”

Surface Roughness : “ Magnetic Field Enhancement

”

High Field (T, Hpeak) : “ Thermal Feedback ”

“ Energy Gap Dependence

H”


High Field Q-Drop ( 6 theories )High Field Q-Drop ( 6 theories )

H. Safa - SRF (2001)

J. Halbritter – Eloisatron Workshop (1999)

B. Bonin - SRF (1995)

J. Knobloch - SRF (1999)

E. Haebel – TTF Meeting (1998)

A. Didenko – EPAC (1996)

Diffusion (O, Imp.) : “ Interface Tunnel Exchange ”

“ Bad Superconducting Layer

”

“ Granular Superconductivity ”

Surface Roughness : “ Magnetic Field Enhancement

”

High Field (T, HP) : “ Thermal Feedback ”

“ Energy Gap Dependence

H”


J. Halbritter - SRF (2001)

& IEEE Trans. on Appl. Supercond. 11, (2001)

RF field on

metallic surface

metalcleanfornegligibleZemissionelectroncausesE

seriesTaylorHHRRZHE

CHH ...1 22*0

Dielectric oxide layer on metal enhancement of ZE by I.T.E.

( localized states of Nb2O5-y and density of state of Nb )

with electron diffusion at NbOx - Nb2O5-y interface

I.T.E. quantitative description of Q-slope

E

C

E eR*

80:

:* ERRbyfittednallyconvention

factortenhancemenfieldelectricE

ITE reduction by :

•smoothening surface ( EP )

( * and E° )

•baking : Nb2O5-y vanished - better interface

( reduction of localised states )

valueonsetEatstarting

1E+09

1E+10

1E+11

0 10 20 30Eacc (MV/m)

Q0

C1-16 ( 1.3 GHz )

no electronsno X-rays

RFpower

RH

RE

E°

Interface Tunnel ExchangeInterface Tunnel Exchange


Magnetic Field EnhancementMagnetic Field Enhancement

J. Knobloch - SRF (1999) microstructure on RF surface

( surface roughness - step height 10 m )

magnetic field enhancement

normal conducting region if

factor ( BCP )

Hm

Cm HH

5.26.1 m

K. Saïto - PAC (2003)

electromagnetic code + thermal simulation Q0(Eacc)

Q-slope origin

the most dissipative G.B. quench (equator)

EP : ( HC/m= 223 mT ) m=1

BCP : ( HC/m = 95 mT ) m=2.34


HF Rinse

- used to suppress field emission -

does not affect baking benefit

Experimental High LightsExperimental High Lights

B. Visentin - Argonne (2004) & SRF(2005)

1E+09

1E+10

1E+11

1E+12

0 10 20 30 40Eacc (MV/m)

Q0

D1-22 ( EP - Saclay A1 )C1-03 ( EP - KEK 9 )C1-15 ( BCP - Saclay I1 )C1-16 ( BCP - Saclay P1 )C1-10 ( BCP - Saclay N1 )

Q-slopes before baking ( BCP and EP cavities )

RF PowerLimit


+ 20' HF + HPR1E+09

1E+10

1E+11

0 10 20 30 40Eacc ( MV/m )

Q0

E5 : ( 120°C / 60 h )


quench

Baking: Definitive Treatment

air exposure for 4 years - HPR

High Field Q-drop

Similarity between BCP and EP cavities (before baking) In contradiction with M.F.E.

theory

Baking: Universal Treatment

fine, large, single crystal, clad, shape

w / wo annealing @ 800 or 1350 °C

EP (>40 MV/m) or BCP chemistry In contradiction with I.T.E. theory


Eacc > 40 MV/m

( TESLA like shape - fine grains)

BCP chemistry instead of EP chemistry

Some Exceptional OccurrencesSome Exceptional Occurrences

P. Kneisel - SRF (1995)

Baking Resistance

B. Visentin - EPAC (2006)

48 h baking


C1 19

1E+09

1E+10

1E+11

0 10 20 30Eacc ( MV/m )

Q0

B1 : BCP - no HTC1 : Sdt UHV Baking (110°C / 60h in cryostat)E1 : BCP after HT 800°C / 2hF1 : Fast Argont Baking + HF (145°C / 3h in stove)F2 : Sdt UHV Baking (110°C / 60h in cryostat)

T. Saeki – TTC Meeting @ KEK (2006)

W. Singer - SRF (2001)

Nb/Cu clad cavity

(after baking) Niobium Cavity C1-15

1E+09

1E+10

1E+11

1E+12

0 10 20 30 40

Eacc ( MV/m )

Q0

before baking

after baking @ 110 °C / 60 h

QuenchRF Pow erLimitation


Theories / Experiments ConfrontationTheories / Experiments ConfrontationB. Visentin - SRF (2003) – updated at Argonne Workshop (2004)

Y / N = theory in agreement / contradiction with experimental observation N+ = undisputable disagreement with experiment


G. Eremeev, H. Padamsee - EPAC (2006)

Fine and large grain

cavities @ 1.5 GHz / BCP

Large grain (G.B.= white lines)

Fine grain

Global heating - Large spread out ( fine grain )

Hot spots for large grain cavity

Grain boundaries not involved in Q-drop

L. Lilje - SRF (1999)

Where do we stand now ?Where do we stand now ?

Fine grain


Hot spots ( large grain cavity )

Reduced after baking

Q-slope restored by 40 V anodization

1E+09

1E+10

1E+11

0 5 10 15 20 25 30Eacc (MV/m)

Q0

Baseline 40 V anodization1st 120C 12h bake 2nd 120C 12h bake

T = 1.7 K

1

4

7

10

13

16

19

22

25

28

31

34

S1 S

3 S5 S

7 S9 S11 S

13 S15

0

0.02

0.04

0.06

0.08

0.1

0.12

T (K)

AzimuthBottom Iris

Top Iris

Equator

Q0 = 6.1 109

Eacc = 30.3 MV/mLarge-grain single-cell

1

4

7

10

13

16

19

22

25

28

31

34

S1 S

3 S5 S

7 S9 S1

1 S1

3 S1

5

0

0.03

0.06

0.09

0.12

0.15

0.18

T (K)

AzimuthBottom Iris

Top Iris

Equator

Q0 = 2.9 109

Eacc = 24 MV/m

Large Grain : Hot SpotsLarge Grain : Hot Spots

G. Ciovati - LINAC (2006)


Hot Spot TheoryHot Spot Theory

A. Gurevich - SRF (2005)Localized sources of dissipation

caused by defects:

• grain boundaries (vortex penetration)

• precipitates

• non uniform surface oxide layer

Cavity surface with hotspots (dark grey)caused by smaller defects of radius r0 (black)

Hot spots consequences:

• non linear effect

• reduce breakdown magnetic field HC

• increase high field Q-drop


Open Issue : Oxygen RoleOpen Issue : Oxygen RoleCorrelated problem to the Q-drop existence , why baking suppress it ?

Cavity Baking Interstitial Oxygen diffusion

),,(2/

0

tTC

eDD

Dt

xerfcCC

SRTE

S

a

Oxygen Diffusionin Niobium

0,0

0,2

0,4

0,6

0,8

1,0

0 20 40X ( nm )

C/CS

100 °C / 3h

100°C / 60h

analytic solutions2nd Fick's law

semi infinite solid : C(0,t) = CS

Improved modelwith decomposition of oxide layer

G. Ciovati - SRF (2005)S. Calatroni - SRF (2001)

0

0.2

0.4

0.6

0.8

1

1.2

100 120 140 160 180 200

c(0, 48 h)u(0, 48 h)v(0, 48 h)

Oxy

gen

conc

entr

atio

n (a

t. %

)T (°C)

From oxide decomposition

Initial interstitial oxygen

minimum of O for 140°C at x=0near of optimum baking parameter 120°C


Oxygen Oxygen involvedinvolved in Q-slope in Q-slope

SIMS measurements on samples.

After baking,

O concentration is modified:

• increased for multiple grain

• reduced for large grain

J. Kaufman, H. Padamsee - SRF (2005)

Q-slope restored after oxide layer thickening

( anodization 40 V)

1E+09

1E+10

1E+11

0 5 10 15 20 25 30Eacc (MV/m)

Q0

Baseline 40 V anodization1st 120C 12h bake 2nd 120C 12h bake

T = 1.7 K

G. Ciovati - LINAC (2006)

Already observed at Cornell (30 V – 60 V)

H. Padamsee - Argonne (2004)

Some observations are in agreement…


Oxygen Oxygen involvedinvolved in Q-slope ( cont. ) in Q-slope ( cont. )

Fast Baking :

Based on a equivalence in terms of interstitial oxygen diffusion:

110 °C / 60 h ↔ 145 °C / 3 h

B. Visentin - SRF (2005)

O Diffusion( Semi Infinite Solid )

C/CS = erfc ( … )

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0

0 10 20 30 40 50 60

X ( nm )

C/CS

110°C/60h

145°C/3h

C1-09 ( BCP cavity )fast UHV baking

1E+09

1E+10

1E+11

1E+12

0 10 20 30Eacc ( MV/m )

Q0

V2 : baking 145 °C / 3 hours

V1 : no baking

quench


Oxygen Oxygen not involvednot involved in Q-slope in Q-slope

SIMS measurements

any noticeable difference before (A) and after UHV baking (8 & 3)

( multiple grain samples)

(only for baking at high temperature in air)


NbO/Nb

0,00

0,01

0,10

1,00

10,00

0 20 40 60 80 100

Abrasion Depth (nm)

Inte

nsi

ty (

a.u

.)

A Reference

8 UHV baking 110°C/60h

3 UHV Fast Baking 145°C/3h

5 Air Baking 145°C/3h

B. Visentin - Argonne (2004)

Q-slope not restored after formation of new

oxide layer ( HF rinse )


+ 20' HF + HPR1E+09

1E+10

1E+11

0 10 20 30 40Eacc ( MV/m )

Q0

E5 : ( 120°C / 60 h )


quench

Or after anodization 5 V – oxide thickness x2 ( 10 nm )

H. Padamsee - Argonne (2004)

But controversy exists in experimental results arguing for the non involvement…


SIMS Analysis and BakingSIMS Analysis and BakingEfficient Q-slope improvement by Baking ( Q0 vs. Eacc )

if there is any noticeable O diffusion in Nb ( SIMS analyses )- UHV or Argon atmosphere : oxygen free - no diffusion from surface- T / (110 °C / 60 h ↔ 145 °C / 3 h) upper limit before diffusion from NbOx

C1-09 ( BCP cavity )fast baking with IR heaters

1E+09

1E+10

1E+11

1E+12

0 10 20 30Eacc(MV/m)

Q0

V1 - No BakingV2 - UHV Baking 145°C/3hY2 - Air Baking 145°C/3h

quench

C1 09 (BCP cavity)fast baking in stove

1E+09

1E+10

1E+11

1E+12

0 10 20 30Eacc(MV/m)

Q0

AE1-No BakingAE2-Argon Baking 145°C/3hAF1-Air Baking 145°C/3h quench

Nb2O+/Nb2

+

1E-04

1E-03

1E-02

1E-01

1E+00

1E+01

0 10 20 30 40 50

Abrasion Depth (nm)

Inte

nsit

y (a

.u.)

A - Reference3 - UHV Baking 145°C/3h5 - Air Baking 145°C/2h

Nb2O+/Nb2

+

1E-04

1E-03

1E-02

1E-01

1E+00

1E+01

0 10 20 30 40 50

Abrasion Depth (nm)

Inte

nsit

y (a

.u.)

A - Reference

D - Argon Baking 145°C/3h stove

C - Air Baking 145°C/3h stove

B. Visentin – TTC Meeting @ KEK (2006)

Debate is still open :

If oxygen is not involved, which is responsible ?

Strong correlations

between

RF results and

SIMS analyses


CONCLUSIONCONCLUSIONBulk Cavity :

no substrate (only Nb) - BCP (reproducibility) – (EP) - Baking (cure)

lot of experimental data (worldwide) and theories since 1998

not sufficient to understand HF Q-Drop origin

noose is tightening (theoretical explanations rejected) - in progress

Thin Film (Nb/Cu) : more difficult

Substrate + Thin Film

Lot of parameters to adjust or different process for coating

(pressure and gas, bias, atomic Nb, ionic assistance, Nb ions… )

Very important for the SRF future (new superconducting material)

Split-up between substrate and thin film issues is necessary

Nb substrate (BCP) - no matter what the absolute RF performances are -

Coating parameters Optimization in terms of relative RF performances.

RF tests on Nb substrate before coating same substrate

Annealing possible ( hydrogen, oxygen contributions to Q-drop )

R EVIEW ON Q - D ROP M ECHANISM B ernard V ISENTIN International Workshop on Thin Films 9 th - 12 th October 2006.

Documents

films legnaro

bulk nb cavity

bulk cavity

halbritter workshop

nb bulk cavitiesb

bulk cavities bcp

film bulk cavitiesvery

bulk case