1 Optimized cavity-enhanced compact inverse-Compton X- ray source for semiconductor metrology Jeremy Kowalczyk (BS Cornell '00 ECE) ([email protected]) University of Hawaii at Manoa
Dec 24, 2015
1
Optimized cavity-enhanced compact inverse-Compton X-ray source for
semiconductor metrology
Jeremy Kowalczyk(BS Cornell '00 ECE)
([email protected])University of Hawaii at Manoa
2
The need for compact X-ray sources Shorter wavelength extreme UV light sources not
good enough for next generation semiconductors [1].
Instead industry using multi-patterning approach [2] Same feature size as previous generation Stack multiple layers of circuits to increase density
New manufacturing challenges Mis-shapen vertical structures Misalignment of layers
Need in-line metrology to detect and correct
3
Metrology technique SAXS:small angle x-ray scattering
4
What can SAXS do?
Plots courtesy of Joseph Kline, NIST, Department of Commerce.
5
Ideal source for SAXS
Divergence angle ~ 1 mrad Spot size < 100 micron Energy > 20 keV (low absorption) 1010 photons/sec
Conventional sources 106 photons/sec Synchrotron meets spec
Fits in a fab Affordable High reliability
6
SAXS needs:compact, bright X-ray source
inverse-Compton
Eelectron,i
= γmc2 Ephoton,i
e-
Eelectron,i
= γmc2 - (Ephoton,f
-Ephoton,i
) Ephoton, f
≈ 4γ2Ephoton,i
e-
Ephoton, f
≈ 4γ2Ephoton,i
7
Design Philosophy
Use off the shelf parts Partner with vendors
Linac, gun, undulator Minimize engineering effort Minimize cost
Ebeam = expensive Laser = cheap
8
Design Philosophy (2)
Maximize cheap laser power Sufficient expensive ebeam current Total average X-ray power
Linear in laser power, ebeam current
9
HawaiiInverse-Compton X-ray source
Image courtesy of Eric Szarmes
10
Why GHz rep rate?
Allows inexpensive optical storage cavity Short length
GHz rate allows stacking on every pass
Small mirrors Near confocal cavity Very tight tolerance Large mirrors with tight tolerance = $$$$
Keeps thermal load manageable CW laser is a non-starter mirror distortion at high average power
11
Limitation: Iave
from thermionic gun
Back-bombardment limits Iave
< 50 μA
“Laser pre-pulse” technique increases Iave
by ~10X
Photocathode gun solves back-bombardment but... Only most advanced research photocathode guns
can do GHz rep rates (Cornell)
12
TE
MP
TIME
TE
MP
TIME
RF on
TE
MP
TIME
RF off
E-field
e-
e-
Back-bombardment heating
Short RF timeShort current pulseLow I
ave
Cathode assembly
CathodeTungsten heater
13
TE
MP
Laser pre-pulse cancels back-bombardment heating
RF on
E-field
e-
e-
TE
MP
RF offTE
MP
Laser pulse TE
MP
Laser pulse
Long RF timeLong, stable pulseHigh I
ave
14
Laser pre-pulse increases Iave
~8X increase in Iave
~26X increase in Iave
Status: preliminary experiments donemodest I
ave increase, but temp. too high
waiting long pulse (10 μs) laser
15
Expected Specs Enable SAXS
16
UH source more cost effective Lyncean Technologies
Similar specifications Focus on
ebeam hardware, small storage ring Low rep rate
~$10 millon UH source
Focus on inexpensive laser hardware high rep rate
Partners: KLA-Tencor, Boeing, Wenbing Yun (Xradia) ~$2 to 3 million
17
Thank you!!
Questions/Discussion
18
References
J. M. D. Kowalczyk and J. M. J. Madey. Back-bombardment compensation in microwave thermionic electron guns.
Physical Review Special Topics - Accelerators and Beams, 17(12):120402, Dec. 2014. doi:10.1103/PhysRevSTAB.17.120402.
J. M. J. Madey, E. B. Szarmes, M. R. Hadmack, B. T. Jacobson, J. M. D. Kowalczyk, and P. Niknejadi. Optimized cavity-enhanced x-ray sources for x-ray microscopy.
In Proc. SPIE 8851, X-Ray Nanoimaging: Instruments and Methods, pages 88510W–1 – 88510W–9, Sept. 2013. doi:10.1117/12.2027193.