Would >50 MV/m be Possible with Superconducting RF Cavities• Peter Kneisel, Grigory Eremeev,...

Operated by Los Alamos National Security, LLC for the U.S. Department of Energy’s NNSA

U N C L A S S I F I E D Slide 1

Would >50 MV/m be Possible with

Superconducting RF Cavities ?

Tsuyoshi Tajima

LANL

PAC2011, New York, 28 March – 01 April 2011

LA-UR-11-01937


U N C L A S S I F I E D

Outline

Background / motivation

An innovative idea to increase the accelerator gradient

using thin film superconductors

Results of Hc1 measurements using SQUID

magnetometry at LANL

Results of RF surface resistance and quench field

measurements at SLAC

Discussion of issues

Conclusion

Slide 2

Nb1.3 GHz 9-cell cavity to be

used for European XFEL



While experimental results at LANL and SLAC are

mostly discussed in this talk, a lot of people have been

involved in this study. Many thanks to them!!

LANL• Nestor Haberkorn and Leonardo Civale (MPA-STC): DC magnetization measurements,

• Ray DePaula and Isaiah Apodaca (MPA-STC): Alumina coatings

• Roland Schulze and his student (MST-6): AES/XPS analyses

• Dave Devlin (MST-7): discussions on cavity coating techniques.

• Marilyn Hawley (MST-8): AFM analyses and discussions

External collaborations• Jiquan Guo, Sami Tantawi and their co-workers (SLAC): high-power RF testing at SLAC

• Thomas Proslier and Mike Pellin (ANL): some ALD coatings

• Brian Moeckly and Chris Yung (STI): preparation of MgB2 samples using reactive co-evaporation

technique

• Peter Kneisel, Grigory Eremeev, Binping Xiao (Jlab): providing Nb materials and Rs measurements

• Akiyoshi Matsumoto, Hideki Abe and Minoru Tachiki (NIMS, Tsukuba, Japan): discussions

• Eiichiro Watanabe, Daiju Tsuya and Hirotaka Ohsato (NIMS, Tsukuba, Japan): ALD of alumina layers

• Toshiya Doi, Takafumi Nishikawa, Tomoaki Nagamine and Kazuki Yoshihara (Kagoshima

University, Japan): preparation of some MgB2 samples using E-beam co-evaporation technique

• Hitoshi Inoue (KEK, Tsukuba, Japan): vacuum baking of some Nb samples for coating

Slide 3



Background / motivation

Niobium (Nb) Superconducting RF (SRF) cavities

have been replacing Cu cavities for the last ~30 years

due to better energy efficiency and other benefits.

The highest accelerating gradient (Eacc) of existing

accelerators as user facilities is ~ 20 MV/m.

In the last few years, the technology to achieve >35

MV/m with 1.3 GHz 9–cell cavities has been maturing

But, it is doubtful that we can get >50 MV/m SRF

cavities in a practical sense using Nb technology (a

traveling wave structure might have a chance, but it

will not be discussed in this talk.)

Slide 4



What sets the fundamental limit of Nb SRF cavities?

The fundamental limit that prevents niobium cavities

from reaching >50 MV/m is the thermodynamic critical

magnetic field of Nb, Hc 2000 Oe (Bc µ0Hc = 200 mT)

Slide 5

(Hc 2000 Oe)



Hc1 sets practical limit of SRF technology

Slide 6

Cornell 1.3 GHz Re-entrant cavity

[Geng et al., Proc. PAC’07, p. 2337]

50

(1700 Oe)

(2000 Oe)



1.E+09

1.E+10

1.E+11

0 10 20 30 40 50 60

Q0

Eacc (MV/m)

LANL 805 MHz single-cell cavity

[Tajima et al., LINAC2010]

2 K

Hc1 (1700 Oe)

Hc (2000 Oe)

Hc1 sets practical limit of SRF technology


U N C L A S S I F I E DSlide 8

Concept of field enhancement

with a single layer of MgB2

Insulating layers

Gurevich’s innovative idea: to overcome the fundamental limit of Nb

cavities by using multilayers of superconducting/insulating thin films

Problem: the thinner the layer, the smaller the field screening

Solution: multilayers to produce a “cascade” field screening effect

Gurevich only considered the limit d<<l

Strongest Hc1 enhancement, but

• many layers may be needed

• coating curved walls with uniform multilayers

of very thin layers is challenging

Each layer reduces

the field further,

until it drops below

the Hc1 of Nb

SC layer

H = 3550 OeH = 1700 Oe



The key of Gurevich’s proposal is that Hc1 in parallel

with the material surface can be increased using thin

films if the thickness d ~ l (penetration depth) or thinner

Slide 9

[Gurevich, APL 88 (2006) 012511]

l

ln4 2

01

cH (d >> l)

~ln

22

01

d

dH c

(d << l)

Hc1(d

/l)/

Hc1(

)

1

d/l1

L. Civale et al.,

PRB 48 (1993) 7576

(general expression)

H

d

Same as in the cavity : coherence length

l



The walls of a superconducting cavity must remain free of

vortices. If not: RF field vortex oscillations dissipation

drastic Q decrease

d l

d

H

B(x)

B=0

l

d >> l

Energy cost of expelling field: H

Energy of a vortex: H independent

Hc1: vortex nucleation becomes

thermodynamically favorable

Superconducting film with H//surface

B≠0

Cost of expelling field decreases

Hc1 increases (thermodynamics)

Cavities must operate below Hc1

Possibility: coat the inside of the cavity with a superconducting thin film



0 10 20 30 400

500

1000

1500

2000

MgB2 film 300nm

MgB2 film 500nm

Nb bulk

Hc1 [O

e]

T [K]

0.0

12.5

25.0

37.5

50.0

Ea

cc [M

V/m

]

N. Haberkorn et al.,

unpublished

Slide 11

Hc1 of 300 nm MgB2 film shows higher Hc1 than Nb by ~25 % !

At 4.5 K, the lowest measured temperature, Hc1 > 2000 Oe

Hc1(~300 nm MgB2 film)> Hc1(bulk Nb)

MgB2 films

from STI

Converted

Using

Hpeak /Eacc = 40

500 nm

SQUID magnetometry

at LANLT (K)

0 10 20 30 40

2000

1500

1000

500

0

Hc

1(O

e)

Solid Nb

0

25

50

Ea

cc

(MV

/m)



0 500 1000 1500 2000 2500

-0.0005

-0.0004

-0.0003

-0.0002

-0.0001

0.0000

6K

10K

12K

20K

25K

27.5K

30K

35K

37K

m [e

mu

]

H [Oe]0 5 10 15 20 25 30 35 40 45

-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

dm

/dH

[1

0-6 e

mu

/Oe

]

T [K]

Preliminary results of 200 nm MgB2 film: SQUID

magnetometry (L. Civale of LANL)

M (

em

u)

H (Oe)

0 1000 2000500 1500 2500

dM

/dH

(1

0-6

em

u/O

e)

Straight line

Meissner slope

Point where a large number of vortices start to enter

6 K

37 K

T (K)

0 10 20 30 40



“raw data”(only input parameters are L & W)

0 5 10 15 20 25 30 35 40

0

50

100

150

200

250

300

350

400

47 nm

113 nm

65 nml

ab [

nm

]

T [K]

d=300 nm

d=200 nm

two-fluid model

MgB2 film from STI

97 nm

40 nm

135 nm

0 5 10 15 20 25 30 35 40 45

0

20

40

60

80

100

120

effe

ctive

th

ickn

ess d

eff [n

m]

T [K]

Penetration depth increases with higher temperature!

This causes the reduction of effective thickness!!

(L. Civale of LANL)

Eff

ec

tive

th

ick

ne

ss

(n

m)

0

20

40

60

80

100

120

0 10 20 30 40

T (K)

Pe

ne

tra

tio

n d

ep

th (

nm

)

0 10 20 30 40

T (K)

400

300

200

100

0

200 nm film

300 nm film



Magnetization measurements for <100 nm films have been difficult. We

are considering a new method to detect field of vortex penetration into

very thin films & multilayers based on ac transport techniques

Standard method:

SQUID magnetometer to detect

deviation from Meissner response

Meissner state:

m=-(H/4)Veff deff

deff d-2ltanh(d/2l)H

d

Problem:Very small signal.Present resolution:d~100nm (for l~100nm) V

oscillations

Proposed alternative method:

Lock-in detection of voltage due to vortex

oscillations driven by ac electric currents

vortices

Sensitivity does not decrease for small d



RF measurements of 2-inch (~5 cm) diameter wafers (~1

mm thick) have been carried out at SLAC using 11.4

GHz pulsed power (~1.6 µs) [J. Guo, S. Tantawi et al.]

Hemi-spherical TE013–

mode cavity with magnetic

fields in parallel with the

sample surface

Sample: <1.5 mm thick

Cold headTemperature sensor

15

Typical distribution of

superconducting and normal-

conducting regions after quench

Radial H profile



Some results from Nb samples have shown that surface

condition affects the surface resistance (Rs) and quench

field [J. Guo et al., poster TUP102 in this conference]

Slide 16



Q0 vs. T of LANL single-grain sample, polished to Ra <1 nm



QL vs. Hpeak of a FNAL fine-grain sample, indicating

thermal quench due to high surface resistance

Slide 18

Thermal quench



A LANL single-grain, polished and baked sample, has

shown the highest quenching field (170-180 mT) so far

Slide 19

3 K

Magnetic quench



So, what about MgB2 ?

Slide 20

11.4 GHz

MgB2 alone shows very low resistance

But, not when coated on Nb with an insulator

BCS resistance at 11.4 GHz at 5 K ~ 1E-4, at 3 K ~ 1.5E-5 ohm



300 nm MgB2 films have shown low quenching field (250

Oe) at SLAC tests, which is puzzling, considering Hc1

>2000 Oe from DC magnetization measurements

Slide 21

Upper limit of Q0 due to the copper host cavity ~ 3.5E+5

MgB2(300nm)/Sapphire (430µm)

Is this quench thermal

or magnetic?



Issues on coating

Degradation of Nb surface due to decomposition of

stable natural Nb2O5 layer into a thick lossy NbOx

layer during coating processes at elevated

temperatures (e.g., ALD Alumina at 300-345 ˚C and

reactive co-evaporation MgB2 at 550 ˚C)

Inter-diffusion of elements (e.g., O, Al, Mg, etc.) that

degrade the quality of MgB2 and increase RF losses

Slide 22



Actual

structure

Sample M10061 (ALD Al2O3 10nm /

MgB2 50nm) x4 / sapphire

0 50 100 150 200 250 300 3500

10

20

30

40

50

60

70

Sputter Depth (nm)

Ato

mic

Concentr

ation (

%)

B1

O1

Mg2

Al2

Sapphire

Alumina (10nm) x 4 with Atomic Layer Deposition at 345 ˚C

Intended

structure

B

O

Mg

Al

MgB2 (50 nm) x 4 with reactive co-

evaporation at 550 ˚C

10

0

20

30

40

50

60

70

Depth from surface (nm)

0 50 100 150 200 250 300 350

B BB

Mg

Mg

10

0

20

30

40

50

60

70

Inter-diffusion

is a problem



0

5 104

1 105

1.5 105

2 105

2.5 105

3 105

3.5 105

0 5 10 15 20 25 30 35

S31-multilayer MgB2

Q vs T, low power and high power

Low power test8mT6mT

Q0

T

Q0 of single layer MgB2

SLAC test results of

[Alumina(10nm)/MgB2(50nm)]x4/sapphire structure

showed degradation of MgB2 and quench field

Slide 24

1 105

1.5 105

2 105

2.5 105

3 105

3.5 105

6 7 8 9 10 11

S31-Multilayer MgB2

Q vs H

T=3K

Ql Q0

Qs

Hpeak (mT)

Q0 vs. T at low power Q0, QL vs. Hpeak

Quench at ~93 Oe3.5E+5

Q0

QL

3.5E+5

Q0

Hpeak (Oe)

600

T (K)70 80 90 100 11010 20 30

Tc lowered from 39 K

~2 mΩ vs.

<100 µΩ



Conclusion

Hc1 , a practical limit of SRF cavities, can be increased

by thin superconductor films. A >25 % higher Hc1

(>2000 Oe) than bulk Nb with 300 nm MgB2 films at

4.5 K was demonstrated. Found that effective

thickness must be considered for a layered structure!

High-power RF tests at SLAC have shown low quench

field of 250 Oe with 300 nm MgB2 films, which needs

to be understood. (>300 Oe has been reported

elsewhere. )

Preventing the increase of RF losses due to the

degradation of Nb surface and coated

superconductor will be the next step of development.

Slide 25



Thanks to our sponsors!Defense Threat Reduction Agency (past)

DOE Office of Science/Nuclear Physics (current)

Thanks for your attention!

Slide 26

Would >50 MV/m be Possible with Superconducting RF Cavities• Peter Kneisel, Grigory Eremeev,...

Documents

Transcript of Would >50 MV/m be Possible with Superconducting RF Cavities• Peter Kneisel, Grigory Eremeev,...