10/4/06workshop on high qe milan diamond amplified photocathodes preparation and capsule formation...
TRANSCRIPT
10/4/06 Workshop on High QEMilan
Diamond Amplified Photocathodes Preparation and Capsule Formation
John Smedley
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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?
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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!
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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)
10/4/06 Workshop on High QEMilan
Deposition System
Laser access from both sides of cathode
10/4/06 Workshop on High QEMilan
Natural Diamond, 100 orientation Natural Diamond, 111 orientation
SEY
EDC
SEY
EDC
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
Modification of 1.3 GHz gun for testing diamond capsules
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
Measurements with Type II a Natural diamond
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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
10/4/06 Workshop on High QEMilan
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