10/4/06workshop on high qe milan diamond amplified photocathodes preparation and capsule formation...

10/4/06 Workshop on High QEMilan

Diamond Amplified Photocathodes Preparation and Capsule Formation

John Smedley


Collaborators

I. Ben-Zvi

A. Burrill

X. Chang

J. Grimes

T. Rao

Z. Segelov

Q. Wu

R. Hemley, CI, Washington

Z. Liu, CI, Washington

S. Krasnicki CI, Washington

K. Jenson, NRL, Washington

J. Yater, NRL, Washington

R. Stone, Rutgers U

J. Bohon, Case Western U

P. Stephens, SUNY SB


Overview

• The DAP concept & capsule cathodes

• Diamond options

• Diamond preparation

• Secondary Yield Measurements

• Transparent cathode preparation

• Capsule fabrication

• Thinning diamond

Diamond Amplified Photocathode

3 – 10 kV

TransparentConductor

(Sapphire w/ITO or thin metal )

Photocathode(K2CsSb)

LaserPrimary

electronsSecondaryelectrons

DiamondThin Metal

Layer(10-30 nm)

Diamond(30 μm)

HydrogenTermination

Primaryelectrons

Secondaryelectrons

a


Advantages

• Secondary current can be >300x primary current– Lower laser power– Higher average currents

• Diamond acts as vacuum barrier– Protects cathode from cavity vacuum and ion

bombardment– Protects cavity from cathode (prevents Cs migration)– Should improve cathode lifetime

• e- thermalize to near conduction-band minimum– Minimize thermal emittance


Challenges• Electrons must escape diamond

– Diamond must be 15-30 microns for 700 MHz RF– Surface must have Negative Electron Affinity (NEA)

• Diamond must not accumulate charge– Material must have a minimum of bulk traps– Metal layer required to neutralize holes– Metal layer must avoid surface traps

• Field in the diamond is a critical parameter– Field should be high enough for ve to saturate

– Field should be low enough to minimize e- energy– Modeling suggests 3 MV/m – good for SRF injector


Jewelry Store• Natural Type IIa

– Single crystal – few grain boundaries– Nitrogen content, variable, expensive

• Polycrystalline Chemical Vapor Deposition– Available in large sizes, low impurity content, cheap,

can be grown to promote desired crystal orientation– Grain boundaries (weaken structure & charge traps)

• Single-crystal CVD– Single crystal, low impurities– Unavailable in large sizes, few options for crystalline

orientation


Diamond Preparation

• Acid etch– Remove graphite and surface contaminants– Strip metal layer– Provide C-O bond

• Hydrogenation– Bake to >450C in nTorr vacuum to break C-O bonds– Expose sample to ~1 μTorr of atomic H for 20 min– Hydrogen cracked by 1800C Tungsten filament– Hydrogen termination is robust versus atmospheric

exposure

IR absorption spectra

0

2

4

6

8

10

12

14

800180028003800

Wavenumber (cm-1)

Ab

. C

oef

. (c

m-1

)

as deliveredafter etchafter hydrogenation


Diamond Preparation - Metallization

• Initial layer must bond well to diamond– Titanium forms Ti-C layer

• Subsequent layers – Protect titanium from oxidation

– High conductivity

– Low electron stopping power (Low Z)

• Ti-Pt sputtering is a common choice for diamond detectors; we have had some success with gold

• We may use Ti-Pt-Al, with 5nm of Ti, 5 nm Pt and 20 nm Al

• Possibly thicker layer near the perimeter


Experimental Arrangement

Two types of measurementsTransmission – both sides metallizedEmission – one side metallized,

one side hydrogenated

Prim

ary

Sec

onda

ry

Kapton

Diamond

A

V2

I2 A

V1

I1


Ep=4keV

0

50

100

150

200

250

300

350

400

450

500

0 0.5 1 1.5

Field (MV/m)

Gai

n t

ran

smis

sio

n m

od

e

10 μA primary100 nA primary

Gain for CVD Detector Grade

Transmission Mode


Gain for CVD Detector Grade

Emission Mode

Ip=100nA

0

50

100

150

200

250

300

350

0 0.2 0.4 0.6 0.8 1

Field (MV/m)

Gai

n e

mis

sio

n m

od

e

4keV3keV2keV1.5keV


Gain for CVD Electronic Grade, Emission Mode

Ip=100nA, 300K

0

10

20

30

40

50

60

0 1 2 3 4 5 6

Gradient (MV/m)

SE

Y e

mis

sio

n m

od

e

2keV3keV4keV5keV

Large emission threshold implies charge trapping

IP=100nA

0

10

20

30

40

50

60

70

0 0.1 0.2 0.3 0.4 0.5Gradient (MV/m)

SE

Y e

mis

sio

n m

od

e

4KeV

5KeV

3KeV

2KeV

Change process orderThreshold field reduced dramatically


Transmission Photocathode

• Metal coating and refractive index of diamond make standard “reflection” illumination less attractive

• Concept – K2CsSb on a transparent conductor

– Model suggests optimal thickness is 10-30 nm

– Resistivity is ~20Ωm (109 Ω/sq for t=20 nm)• Fisher, McDonie and Sommer, J. Appl. Phys. 45 487 (1974)

– For I>100μA, a conducting substrate is required

10 nm Copper

Indium Tin Oxide

Resistivity (Ω/sq) 1.7 8-12

Trans. (532 nm) 55% >85%

Trans. (355 nm) 35% >75%

Is ITO compatible with K2CsSb?


Diamond brazed on to Nb

Successfully vacuum tested

Polished from 200 to 70 micron thickness

Chemically treated brazed Nb

Ceramic brazed to Nb

Cold weld Cu to Cu w/ indium

Diamond Capsule Fabrication


Thinning Diamond

• Diamonds will be brazed to support structure prior to final polish

• Mechanical polishing is extremely slow, and prone to breakage

• Reactive Ion Etching may also be an option– Etch rates of 40 nm/min have been demonstrated with

O2 plasmas • Sandhu & Chu, Appl. Phys. Lett. 55 437 (1989)

– Surface finish and effect on polycrystalline surfaces yet to be determined

– Graphite formation may be a problem


Conclusions• A functioning DAP injector will provide a path forward for

ampere-class linacs and ERLs• For sufficiently pure diamonds, transmission gains of >300 have

been demonstrated. ~14 eV of primary energy = 1 e-hole • With proper preparation, emission gain of >300 with small

emission threshold.• H termination can provide a robust, NEA surface for electron

emission. e- affinity is affected by surface crystalline structure.• Synthetic diamonds have desirable properties, although some

development is still needed.• The energy spread of the emitted beam is expected to be <1 eV• With proper choice of the electric field in the diamond, v = vsat

while energy spread is nearly unchangedStill a lot to do!


Still to do• Test other diamonds, including some 30μm thick• Understand differences between transmission

gain and emission gain• Demonstrate gain with Ip = 10mA (larger area)• Temporal and energy profile of emitted electrons• Characterize impurities and crystalline

orientation • Determine the proper method of metallization• Develop transparent photocathode• Capsule fabrication and thinning diamond to

30μm are significant “engineering” challenges


Secondary Electron Yield Measurements

• Ongoing work a NRL– J. Yater and A. Shih, J. Applied Physics 87

(2000) 8103, 97 (2005) 093717

• Primarily reflection mode

• Boron-doped diamonds, Natural and CVD

• H or Cs termination

• Measured Secondary Electron Yield (SEY) and Energy Distribution Curves (EDC)


Deposition System

Laser access from both sides of cathode


Natural Diamond, 100 orientation Natural Diamond, 111 orientation

SEY

EDC

SEY

EDC


CVD Diamond, polycrystalline

SEY

EDC

Implications

• Crystalline orientation affects NEA surface

• Diamond quality affects Secondary Electron Yield

• Electron energy at emission is <1 eV

• SEY of 100+, with 3 keV primaries


Modification of 1.3 GHz gun for testing diamond capsules


Quarter wave Superconducting Choke Cavity at 1300 MHz F = 1299.367 MHz

C:\LANL\ZHAOWORK\AES1.3GCHOKE\CHOKE4.AF 8-18-2004 14:17:10

0

1

2

3

4

5

6

7

8

9

10

11

12

0

1

2

3

4

5

6

7

8

9

10

11

12

-4 -2 0 2 4 6 8 10

SRF Gun modified to accept Choke joint with diamond


Measurements with Type II a Natural diamond


High Average Current Linacs• Applications:

– Linac-based light sources– Free-electron lasers– Electron-cooling for RHIC

• Challenges – Energy recovery– Large laser power required, even for high QE

Cathodes:

WVA

e

h

QE

IPower 233.2

1.0

1


Time

Ele

ctr

ic F

ield

Time

Ele

ctr

ic F

ield

PhotoinjectorSW, TM010f = 1298.07726 MHzQ0 = 7.07x109

Pd = 5.1WBmax/Emax = 2.2 mT/(MV/m)Emax/Ecathode = 1.048

SUPERFISH simulationSW, TM010f = 1298.07726 MHzQ0 = 7.07x109

Pd = 5.1WBmax/Emax = 2.2 mT/(MV/m)Emax/Ecathode = 1.048

SUPERFISH simulation


BNL SEEP Project

• Goals– Identify suitable diamond- tested 4 types of

diamond: natural, synthetic optical, electronic, single crystal Best so far- natural and electronic grade

– Determine optimal metallic coating– Measure gain in transmission mode >200– Measure Temperature dependence of gain No effect– Measure gain in emission >30 – Measure time behavior and EDCs– Capsule fabrication– Test in RF injector

10/4/06workshop on high qe milan diamond amplified photocathodes preparation and capsule formation...

Documents

nm diamond

high qe milan gain

high qe milan collaborators

high qe milan conclusions

diamond detectors

perimeter slide

problem slide

refractive index of