Upload
makana
View
21
Download
0
Embed Size (px)
DESCRIPTION
Overview of Coating Research Sheila Rowan University of Glasgow On behalf of coating subgroup. Program Overview. Goal: Get the best coating for Advanced LIGO and still be prepared for future challenges. Development for Advanced LIGO – ongoing Research for future - PowerPoint PPT Presentation
Citation preview
1
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Overview of Coating Research
Sheila Rowan University of Glasgow
On behalf of coating subgroup
2
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Program Overview
Goal: Get the best coating for Advanced LIGO and still be prepared for future challenges
• Development for Advanced LIGO – ongoing
• Research for future•Crucial to increase coating sample throughput
• Add vendors beyond CSIRO and LMA• Existing LSC coating labs prepared for more samples• Scatter, Q measuring, other experiments• Collaborate with others outside field
3
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Coating Sensitivity Impact – Advanced LIGO
Multiple noise sources:• Brownian – internal friction• Thermo-optic
• dL/dT - thermoelastic• dn/dT - thermorefractive
Wideband Advanced LIGO Sensitivity
Narrowband Advanced LIGO Sensitivity
Assuming Advanced LIGO Baseline Values
Coating Thermal Noises Limiting noise source in Adv LIGO
4
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Brownian noise
Experiments suggest» Ta2O5 is the dominant source of dissipation
in current SiO2/Ta2O5 coatings» Doping the Ta2O5 with TiO2 can reduce the
mechanical dissipation by ~40% for overall coating
Mechanism responsible for the mechanical loss in Ta2O5 as yet not clearly identified
Research ongoing to:» Identify and directly reduce dissipation
mechanisms in the coatings» ‘optimise’ coating designs by minimising
volume of Ta2O5 present in the coatings
Mirror suspension, with HR coating on front face.
5
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Single layer coating samples (I. Martin et al)
0.5m single layer of Ta2O5 doped with TiO2 ((14.5 ± 1)% by cation) deposited by ion beam sputtering
Thin silicon substrates used for coating» Loss of silicon decreases at low temperature» Coating will dominate the loss of thin
samples
Samples etched from silicon wafers (Stanford)
34mm
500m48m Coating applied here
Titania doped tantala coatedsilicon cantilever in clamp
Uncoated silicon cantilever in clamp
6
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Mechanical loss of coated and uncoated samples
A clear increase in the measured loss is observed due to the addition of the single-layer of doped tantala coating to the silicon cantilever.
0 50 100 150 200 250 300
0.0
5.0x10-6
1.0x10-5
1.5x10-5
2.0x10-5
2.5x10-5
3.0x10-5
3.5x10-5
3rd mode uncoated 3rd coated
Me
cha
nic
al l
oss
Temperature (K)
Low T peak present for all modes of sample studied
Temperature at which the dissipation peaks is a function of frequency
7
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Comparison of un-doped Ta2O5 and Ta2O5 doped with TiO2 (14.5% by cation)
Third mode, ~990 Hz. Tail of the peak is higher in un-doped case.
0 50 100 150 200 250 3005.0x10-6
1.0x10-5
1.5x10-5
2.0x10-5
2.5x10-5
3.0x10-5
3.5x10-5
3rd mode undoped Ta2O
5
3rd mode, TiO2 doped Ta
2O
5
Me
cha
nic
al l
oss
of c
oa
ted
sili
con
ca
ntil
eve
r
Temperature (K)
0 50 100 150 200 250 300
1.0x10-5
1.5x10-5
2.0x10-5
2.5x10-5
3.0x10-5
3.5x10-5 4th mode undoped Ta
2O
5
4th mode, TiO2 doped Ta
2O
5
Mec
hani
cal l
oss
of c
oate
d si
licon
can
tilev
er
Temperature (K)
Fourth mode,
~ 1940 Hz, un-doped peak is higher than doped peak.
8
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
In 3 of the 4 modes measured, doping has reduced the peak height
0 5 10 15 20 25 30 351.0x10-5
1.5x10-5
2.0x10-5
2.5x10-5
3.0x10-5
3.5x10-5
4.0x10-5
Mode 2 undoped Ta2O
5
Mode 2 TiO2 doped Ta
2O
5
Me
cha
nic
al l
oss
of c
oa
ted
sili
con
ca
ntil
eve
r
Temperature (K)
Second mode, ~340 Hz.
Comparison of un-doped Ta2O5 and Ta2O5 doped with TiO2 (14.5% by cation)
9
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Relation of mode frequency to temperature at which loss peaks
Can calculate an activation energy associated with the dissipation peak of (42 ± 2) meV
The low temperature dissipation peak in fused silica has a similar activation energy (44 meV)
» Oxygen atoms can undergo thermally activated transitions between two possible energy states in a double well potential
» Results suggest this dissipation mechanism in Ta2O5 may be similar
0.046 0.048 0.050 0.052 0.054 0.056 0.0585
6
7
8
9
10
11
Linear fit y = 492x + 33.4
ln(
0)
1/Tpeak
10
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Work in progress – Brownian noise
Studies of the underlying material physics by means of high-resolution TEM, Electron Energy Loss Spectroscopy and Electron Energy Loss Near Edge Structure to investigate short and medium range order and bonding in doped and undoped coatings
Heat treatment of bulk SiO2 can significantly improve its mechanical dissipation :
» alters its bonding structure; narrows distribution of potential barrier heights in the double well systems.
» effects of heat treatment can be clearly identified through the behaviour of the low temperature dissipation peak,
Research ongoing at HWS (S. Penn) on effects of heat treatment of single layers of thin film silica (and of multi-layer coatings) – preliminary results show loss decreasing with increasing ‘annealing’ T.
Research ongoing at Glasgow on effects of annealing on single layers of Ta2O5
11
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Coating Research at Sannio University
• ‘Optimization’ of coating layer thicknesses:• Goal: lowest noise at prescribed transmittance;
• Preliminary study based on genetic optimization led to focus further study on stacked-doublet (SD) geometries with tweaked end layers;
• MATHEMATICA code for SD coating optimization and characterization developed;
• Plain-tantala-based optimized coating mirror prototypes manufactured at LMA (fall 2006), scheduled for testing at TNI.
12
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Equilibrium fluctuations of the temperature of the test mass coatings cause fluctuations in physical parameters of the coating. Coupling parameters:
» Thermal expansion coefficient, Thermoelastic noise.» Thermorefractive coeff. Thermorefractive noise.
Update on Thermo-optic noise measurements at Embry-Riddle (A. Gretarsson)
Thermorefractive contribution somewhat higher than thermoelastic contribution but same order of magnitude.
Thermo-optic noise
= (coherent) sum of thermoelastic and thermorefractive contributions.
dTdn
13
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
The setup
Obtain:
(Ptrans / Pinput) & (Prefl / Pinput)
Versus Tsample
14
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Result for
Sili
ca V
alue
15103.5 K
Mean of allmeas. at 543 nm(LMA plain Ta2O5/SiO2):
New 1064nm meas. (LMA 5**)
15100.7 K
15
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Thermo-optic noise in Advanced LIGO
15103.552
K Using OTa
Contributes 25% of total noise floor in 70-100 Hz region.
Is the limiting noise source in narrowband mode.
Measurements are ongoing at ERAU. Ellipsometer measurements on single
layer recently completed at SOPRA. Results fed to Univ Sannio to optimise
doped-Tantala coatings - total (B+TR+TE) noise included;
Porting of optimization code to BENCH scheduled.
16
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Thermal Noise Interferometer (TNI):Recent Coating Results (E. Black)
Broadband measurement of doped-tantala/silica
Homogeneity experiment: Coating thermal noise vs. spot position
Lab move:» Reassembly/optical-alignment complete» Servo recommissioning in progress
natural cavity axis
natural cavity axis with weights on
17
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
TNI Future Coating Plans
Optimized silica/tantala coating measurement
» In progress
Doped, optimized coating measurement
» Funding approved, contingent on un-doped, optimized results
Photo-thermal measurements of thermo-physical properties in advanced coatings
Direct measurement of thermo-optic noise (image from Andri Gretarsson)
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
UV Irradiation Effects on Initial LIGO-type Mirror Coating
Nick Leindecker, Sei Higuchi, Ashot Markosyan, Ke-Xun Sun, Bob Byer, Sacha Buchman, Marty Fejer, Roger Route, Stanford
In Collaboration with LIGO Charging Work Group
19
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
UV Irradiation Geometry
~15 mm~2.5 x 3 mm2
Initial LIGO Mirror by REOHR/AR OX936/OX937
Ball lens UV LED
Irradiation by UV LED at 255 nmOperated with average power 3.3 WIrradiation intensity 15 W/cm2
Irradiation for 48 hoursTotal UV energy deposition ~2.6 J/cm2
Removable mask formsa small aperture
UV LED with a Ball Lens
20
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Fig. 1. Loss measurement prior to UV irradiation. A transversal scan across the surface over 12 mm length. The average value of ~ 0.6 ppm.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0 2 4 6 8 10 12
Distance scanned [mm]
Ab
sorp
tion
[pp
m]
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0 2 4 6 8 10 12
Distance scanned [mm]
Ab
so
rpti
on
[p
pm
]
Fig. 2. In the exposed area (~2.5 mm window) = 1.7- 1.8 ppm. continuous increase of while approaching the exposed area.
Optical Loss Measurement Results
• Possibility that loss increase is due to contaminants? (which may become worse in the UV region because of disassociation?)• Plan to increase the dose of UV irradiation and try to parameterize the phenomena. Meanwhile continue to check the vacuum cleanliness.
21
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Visits by Eric Gustafson/Helena Armandula + others to talk to possible vendors (for R&D)
MLD (US) - Very Interested in working with us.
OCLI (US) - Very Interested in working with us. Want to concentrate on what causes stress in the coatings. Fabulous measurement capability. They can do any coating technique except for CVD and have any ISB source you can imagine.
REO (US) - Interested but no particular suggestion on what they might work on with us.
ATF (US) - Very interested in working with us. Really want to find ways to improve thermal noise in coatings for frequency standard applications
LMA (France) – Ongoing program on coatings for Advanced LIGO.
Solicit formal bids Coating Workshop at December or March LSC Meeting – (Gregg Harry)
Enhanced effort on coatings planned with additional vendors
22
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
SummarySummary
• Coating thermal noise crucial limit to gravitational wave detector sensitivity for foreseeable future• Optical and thermal properties also important and demanding• Progress has been made in understanding and improving coating thermal noise
• Titania-doped tantala reduces mechanical loss• Optimization of thickness improves thermal noise• We have the first handle on a dissipation mechanism through temperature dependence of dissipation
• Need continuing research• Development of Advanced LIGO coating• Research on improved coating for future detectors
• Plan developed to increase on-going LSC coating research• Need to involve more vendors than current research• Lab payment of coating costs crucial• Existing team (including Virgo) with experience • Allows for both Advanced LIGO and beyond work
23
L-V meeting, Hannover, Germany, 22-25 Oct, 2007
Comparison of un-doped Ta2O5 and Ta2O5 doped with 14.5% TiO2
Fifth mode, ~ 3190 Hz.
0 5 10 15 20 25 30
1.0x10-5
1.5x10-5
2.0x10-5
2.5x10-5
3.0x10-5
3.5x10-5
Me
cha
nic
al l
oss
of c
oa
ted
sili
con
ca
ntil
eve
r
Temperature (K)
5th mode undoped Ta2O
5
5th mode TiO2 doped Ta
2O
5