Acceleration gradient limitations in room temperature

and superconducting accelerationstructures

Nikolay SolyakFermilab

RT Breakdown studies in recent years

• NLC/GLC and CLIC type TW and SW accelerating structures @ 11 GHz (SLAC) and 30 GHz (CERN)• Waveguide, Single cell cavities (SW, TW), • Special cavities (choke-cavity, TM01, PBG)

– Frequency dependence – Stored energy, Power and Group velocity– Pulse length dependence– Geometry (low/high impedance structure, slots, choke, phase advance, …)– Materials (Cu, CuZn, Au, Mo, Ti, AL, W, SS, …), coating– Surface preparation, Oxidation– HP processing

• Surface and data analysis, Simulations and Theoretical models (field emission models, plasma formation, pulse heating, etc…)

Aggressive R&D program to understand gradient limitations and physics of breakdown (collaboration of many institutions: SLAC,CERN,KEK,...)

Still we have no theory which can explain all data for all set of parameters and make a prediction (example: T53vg3MC (good) T26vg3 (not good))

(J.P.Delahaye)

Test Results at 11 GHz

power fit = E30

T53VG3MC-11GHzHDS60S

Summary on gradient scaling (CLIC)

consttE pa 6/130~ aEBDR

For a fixed pulse length For a fixed BDR

constBDR

tE pa 530

New constrain - modified Poynting vector which is geometry independent and can be scaled among all (TW and SW) Cu structures.

SS ImRe0000 cSW

TW

SWTW

c gHEC

CHES

• A model of the breakdown trigger based on the pulsed heating of the potential breakdown site by the field emission currents

• describes both TW and SW accelerating structure experimental results• Good agreement experiments with analytical estimate

A.Grudiev, W.Wuensch

TD18vg2.4 T18vg2.4 T28vg3 TD28vg3 CLIC_G

Sc=6.2MW/mm2 @ tpP=100ns

Average unloaded gradient [MV/m]

109 106 105 103 120

Prediction of average unloaded gradient at rect. pulse length of 100ns and BDR=1e-6 based on the results achieved in T53vg3MC: 102.3MV/m at 100ns and BDR=1e-6: Sc=6.2MW/mm2@100ns.

Predictions for test structures

Best tested CLIC type structure TD18vg2.4: Eacc ~ 105 MV/m at 100 ns

From Alexei Grudiev

“Practical” Gradient limitations in RT accelerating structure

• Stored energy (SW), RF power and group velocity (TW)– fraction of energy dissipated in tip vs. geometry?– < 1J energy dissipated in breakdown site– In TWS power absorbed in spot, before plasma built

• Breakdown triggered by field emission - Es

– Fatigue, modification of boundaries and grains

• Magnetic pulse heating – T~ (Hs2)()1/4 (plastic

deformation limit: T~50ºC for annealed Cu and ~130ºC)

• Possible other limitations (dark current, Multipactoring, couplers, windows, etc…)

• Growing tips

Theory of Breakdown

• A lot of experimental data is available, more will come soon– Need more statistics for each tested geometry

• Few models are presented, but non of them is able to explain all experiments and predict gradient limits – Do we need one theory to explain everything?– Few breakdown mechanisms few theories– Clear criteria to define which mechanism is dominated for

chosen parameters space

• Facts, which should be explained by theory– Breakdown in TM01 cavity (no surface electric field)– Factor of ~2 difference in electric field in low/ high impedance

structures– Gradient limitation in cavity with small gaps (choke cavity)– ….

• Let’s try to scale gradient limit results obtained for room temperature structures to SC cavity:

RT: Eacc~100 MV/m (Epk~200 MV/m) at 100ns.

SC: Eacc~ 20 MV/m (Epk~40 MV/m) @ 1ms or Eacc~ 7 MV/m (Epk~14 MV/m) @ 1sec

• BDR – usually not considered for SC cavity• Frequency of SC cavity is lower stored energy higher

• Conclusion: Physics of breakdown in SRF is different

Scaling results from NC to SC structures

consttE pa 6/1

Achieved Limit of SRF electric field•No known theoretical limit•1990: Peak surface field ~130 MV/m in CW and 210 MV/m in 1ms

pulse. J.Delaen, K.Shepard,”Test a SC rf quadrupole device”, Appl.Phys.Lett,57 (1990)

•2007: Re-entrant cavity: Eacc= 59 MV/m (Epk=125 MV/m,Hs=206.5mT). (R.L. Geng et. al., PAC07_WEPMS006) – World record in accelerating gradient

CW 4.2 K

EP

1.3 GHz cavities – 4 production batch

“Practical” gradient limitations for SC cavities

• Surface magnetic field – 200 mT (absolute limit?)

• Field emission, X-ray, starts at ~ 20 MV/m

• Thermal breakdown (strong limits for F>2GHz)

• Multipactoring (in cavity or couplers)

• Medium and high field Q-slopes (cryogenic losses)

• Lorentz detuning and microphonics (frequency change)

• Quality of surface treatment and Assembly

From Alex Gurevich

Energy gap is a function: = (T,Es)

Magnito-optical Imaging – A.Polyanskii & P.Lee

Limit in gradient and Magnetic field

Re-entrant single-cell

K. Saito

Pushing up SRF gradient limit

New Geometry

J.Sekutowicz, V.Shemelin, K.Saito

Traveling Wave Accelerating Structure

~ 25% (max 42%) higher accelerating gradient (vs. TESLA cavity)

• Shorter cells (105 deg phase advance) to improve transit-time factor

• No limitation (up to 10m) in cavity length

• Need tuning to cancel reflected wave

A.Kanareykin, N.Solyak, V.Yakovlev, P.Avrakhov

Field Emission

•Caused by macro-particles at the surface

•Starts typically >20 MV/m

•Exponential Q-drop

•X-ray

•Dark current

•Disappear after HPR

•Clean assembly

Thermal breakdown

Particle and hole on thesurface

Sub-mm size defects lead to quench Temperature Mapping

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/m

Large-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)

Conclusion

0,..,

xHTPx

TT

x cmdiss

2,2

1RFRFmsdiss HHTRP

bsbsKRFcms TTTThHHTR ,,..,2

1 2

hk

b

Thickness

Thermal Breakdown in pure Niobium

Solution for k=const (thermal conductivity)

Test result of 3.9 GHz accelerating cavity

1.E+09

1.E+10

1.E+11

0 2 4 6 8 10 12 14 16 18 20

Eacc, MV/m

Q

1.45 K

1.8 K

2.0 K

1

10

100

1000

10000

2.0 3.0 4.0 5.0 6.0 7.0

Tc/T

nOhm

s

RS

BCS

3-cell cavity built at FNAL

Final cavity preparation done at FNAL (BCP,HPWR)

Residual resistance R_res ~ 6 n Achieved: H_peak = 103 mT, E_acc = 19 MV/m

(Goal: Hpeak= 68 mT, Eacc=14 MV/m)

Magnetic field is likely limited by thermal breakdown

No Field Emission Q ~ 8*10 9 at E_acc =15 MV/m

Maximum accelerating field not depend on Temp

9-cell/3.9GHz: Eacc= 21 MV/m

TESLA: Eacc= 24 MV/mH = 103 mT

No Temperature dependence – why?

resc

RFlinBCSs

nlins R

H

HTRBTR

2

0,

2

,1

481,

resRF

cH

HT

linBCSs

nlins R

H

H

T

eTRBTR

c

RF

2/7

0,

27,2

2)(

4,

0,

0

232 ,/

,c

RF

b H

Hh

TkT

For 2 cases: β·h<<1 and β·h>>1 it gives:

1. Non-linear Surface Resistance

res

linbcss

nlins Rdt

thT

thTtTRBTR

0

2, cos

cossinhsin

2,

The non-linear BCS resistance (RF pair breaking) in the clean limit is given by

Where:

*P.Bauer et.all, “Discussion of possible evidence for non-linear BSC resistance in SRF cavity”, SRF 2005 workshop

TklinBCSs

beT

fATR

2

,

Δ ~1.5 meV is the superconducting energy gap

Surface Resistance used in models

Non-linear case - blue

- Linear BCS:

Measured values of Kapitza conductance and values used in model

2. Kapitza Resistance of Niobium surface

*A. Aizaz, “Improved Heat Transfer in SRF Cavities, NSCL, MSU, 2006** Bousson et. All, “Kapitza Conductance and Thermal Conductivity of Materials Used For SRF Cavities Fabrication”, SRF workshop, 1999

Summarized parameters a, n obtained experimentally**.

Kapitza Conductance

0,01

0,1

1

1,4 1,6 1,8 2 2,2Tb (K)

hk (W

/cm

^2/K

)Used in Model 3cellsStrained*As recieved RRR232*Bousson data**after Ti*after HT*Used in Model 9cells

Measured values of thermal conductivity* (left) and values used in models for 3 cells and 9 cells case(right)

* A. Aizaz, “Improved Heat Transfer in SRF Cavities, NSCL, MSU, 2006

3. Thermal Conductivity of the bulk Niobium

Thermal Conductivity Model

10

100

1000

1 10T(K)

W/mK

3 cells Cavity9 cells Cavity

Cavity #4 QvsE July 10 2008

1.E+08

1.E+09

1.E+10

0 2 4 6 8 10 12 14 16 18 20 22 24

Eacc, MV/m

Q

1.8K

2K

1,E+09

1,E+10

1,E+11

0 2 4 6 8 10 12 14 16 18 20Eacc, MV/m

Q

1.45 K

1.8 K

2.0 K

Example 1: 3-cell 3.9 GHz cavity

1,E+09

1,E+10

0 20 40 60 80 100 120

H surface (mT)

Q

T=1.44 KT=1.8 KT=2 K

Example 2: 9-cell 3.9 GHz cavity

1,E+09

1,E+10

0 20 40 60 80 100 120

H surface (mT)

Q

T=1.8 KT=2 K

Results for 3.9 GHz cavity

3-cell cavity

9-cell cavity

I. Gonin - FNAL

Multipacting

MP-exponential increase of electrons under certain resonance conditions

M.Liepe

Multipacting in ICHIRO cavity

Example: MP in HOM coupler

Gradient Limitation due to MP in HOM coupler.

Result: Coupler was destroyed

Reproducibility problem – Scattering of gradient

0

5

10

15

20

25

30

35

40

45

Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06

Eac

c[M

V/m

]

BCP

EP

10 per. Mov. Avg. (BCP)

10 per. Mov. Avg. (EP)

Gradients achieved over time in DESY cavities

ILC Goal

K.Saito

ILC Word-wide S0 / S1 R&D program

• Aim to develop procedure which will provide 85 % acceptance yield in cavity performance– S0 – single cavity– S1 – 9-cell cavity

– Fabricate 4x4 cavities and demonstrate yield at 4x4 test each (statistics)

S0 - Single-cell R&D program (KEK)

ILC BCD receipt

• final EP (~20μm) after a heavy material removal ( > 100μm).

• HPR (70kg/cm2,1hr)• Bake (120OC,48hr).

Eacc = 46.5± 8.0 MV/m (17%)

EP FLASH receipt

• EP(20μm)+EP(3μm, fresh acid)

• HPR (70kg/cm2,1hr)• Bake (120OC,48hr)

Eacc = 46.7±1.9 MV/m (4%)

20 25 30 35 40 45 50 55 600.98

1

1.02

1.04

1.06

1.08

1.1

1.12

1.14

1.16

1.18

1.2

Project Cost -vs- Linac Gradient

Gradient ( MV/m)

Rela

tive C

ost

ILC Main Linac

~ Eacc2 - Cryogenic losses

~ Eacc Linac/Tunnel length

Summary

• Gradient limitation in RT structure is not fully understood yet, but…– mechanisms are identified– multi-parametric studies underway – better understanding of physics– Modeling and prediction of gradient limits in real

structure

• Gradient limitation in SC cavities– Few mechanisms are studied

• Critical magnetic field• FE, thermal breakdown, MP, Q-slope, …

– New ideas to push gradient limits• Geometry optimization (LL, Re-entrant, TW structure)• multilayer

– Surface treatment and Cleaning are the key problems in SC cavity to reach stable performance.