Study of scattering points on LIGO mirrors Lamar

Study of scattering points on LIGO mirrors Lamar Glover (Cal State LA) Riccardo De. Salvo (Cal State LA) Innocenzo Pinto (Unisannio) LVC meeting 03/15/16 LIGO-G 1600430 Related doc: LIGO-G 1600431 1

Brief Summary • At Cal State LA, we are analyzing scattering points detected in beam illuminated photos of Advanced LIGO mirrors • In LIGO mirrors, A large number of scatterers were observed where none would be expected, through the depths of the coating layer stack • Perhaps the problem of scatterers can now be better understood LVC meeting 03/15/16 LIGO-G 1600430 2

Method • Individual scatterers were identified using astronomical algorithms for stars in galaxies 1. Extract apparent amplitude and position of each scatterer 2. Fit the stored beam position and profile to determine the illumination power on each scatterer. LVC meeting 03/15/16 LIGO-G 1600430 3

Daophot mechanism: Point Spread Templates • “daophot” creates a point spread template from selected sources within the image. • The template is used to search scatterers • Magnitude data is derived for each identified source. Point Spread Template for an exposure time of 1. 25 x 10 -4 s LVC meeting 03/15/16 LIGO-G 1600430 4

Identifying/subtracting scatterers Subtracted Photo Original Photo ignored 256 100 • pixel scale is expanded Dark pixels 256 • Black dots mark the place where bright scatterers have been excised • “dirt” does not fit template and is ignored by daophot and is less bright LVC meeting 03/15/16 LIGO-G 1600430 5

Daophot effectiveness and resolution Pixel amplitude histogram of original Image 100 200 • Residual pixel amplitude histogram after scatterer extraction -50 In a good image Daophot is very effective in extracting all scatterers • Width of residuals (2 -3 pixels FWHM) illustrates Daophot’s good resolution 50 LVC meeting 03/15/16 LIGO-G 1600430 6

Scatterer light intensity distribution @ Exposure Time = 0. 0125 sec The number of reconstructed scatterers grows rapidly at low amplitude 70 0 00 LVC meeting 03/15/16 LIGO-G 1600430 20 50 00 40 7

Number of Scatterers vs. Exposure Time • • Exposure @ 0. 125 ms @ 1 ms exposure @ 400 ms exposure scatterers 134 11, 806 363, 789 • Number increases almost linearly with exposure • Exploring smaller scatterers • But also deeper in the dielectric coating layers ! ! LVC meeting 03/15/16 LIGO-G 1600430 8

The apparent size of scatterer correctable in photos: Local illumination level not correctable: 1. Physical scatterer size inside host layer 2. Depth of host layer inside dielectric coating stack Beam profile (Note: all scatterers are nanometer scale, diffraction limited) LVC meeting 03/15/16 LIGO-G 1600430 9

Growth direction 1: Crystallite Formation in a single evaporated layer Side View Top View Growth Direction • As mirror layers are formed, crystallites appear at different depths. • Those that start earlier in the depositing process grow more and create larger scatterers LVC meeting 03/15/16 LIGO-G 1600430 10

2: Depth in stack effect • There are two depth attenuation components: 1. Attenuation of impinging light due to reflections through depth of coating layers 2. Attenuation of Scattered light from back-reflection on outer layers Scattered light detected by CCD • Complex effect to simulate LVC meeting 03/15/16 LIGO-G 1600430 11

Discussion of crystallite number and distribution intensity • How can so many scatterers exist, and be so uniformly distributed? Probable dirt smear Unresolved by daophot • Cannot be only “dirt” ! • Only a thermodynamical source can explain the observed large number LVC meeting 03/15/16 LIGO-G 1600430 12

What’s Next • Two steps proposed: 1. Take more images at Hanford/Livingston and study more photos 2. Study spare ad-LIGO mirrors with Caltech’s setup to determine scatterer’s depth distribution through layers and size distribution • How to? LVC meeting 03/15/16 LIGO-G 1600430 13

X, Y Axis micropositioning White light LED 50% Reflector @ 45° 1 3 2 Z Axis piezo focusing CCD Camera Microscope lens Ad. LIGO Multi-Layered Mirror LVC meeting 03/15/16 LIGO-G 1600430 14

X, Y Grid scan of Ad. LIGO Mirror X 1 X 2 … Xn Y 1 Y 2 … Defocused Scatterers Raster scan X-Y of microscope frames on mirror surface to identify rough scatterer position Yn LVC meeting 03/15/16 LIGO-G 1600430 15

Depth determination by Z- focussing • No scatterer present: Ø No reflected light • Scatterer present out of f-plane Ø Low illumination density Ø Defocused scatterer image Defocused Scatterer +Z Focused Scatterer Maximum Luminosity Defocused Scatterer -Z • Z-zoom with piezo-movement to bring focal plane on scatterer: 1. Brighter illumination 2. Focused airy disk image LVC meeting 03/15/16 LIGO-G 1600430 16

Depth determination by Z- focussing fits Depth of field 0. 6 µm / n Ø Sub-wavelength Z-position determination from Z-feedback fit Defocused Scatterer +Z Focused Scatterer Maximum Luminosity Defocused Scatterer -Z Nikon lens calculator http: //www. microscopyu. com/ tutorials/java/depthoffield/ LVC meeting 03/15/16 LIGO-G 1600430 17

X-Y determination by profile fitting Ø Sub-wavelength X-Y position determination from Airy disk profile fitting D = 0. 61λ/NA LVC meeting 03/15/16 LIGO-G 1600430 18

Scatterer strength Within Mirror Layers • In photo images, the signal amplitude depends on scatterer depth, due to reflectivity on layers 1 • White light and large angle focusing mitigate the effect 2 • Complex but probably soluble or mappable function LVC meeting 03/15/16 LIGO-G 1600430 19

Scatterers strength Within Mirror Layer 1 Independent of illumination/trans-reflection function: 2 at same depth, information of scatterer’s size is extracted from amount of light collected LVC meeting 03/15/16 LIGO-G 1600430 20

Conclusions • Further measurements on actual mirrors should nail down the observation that millions of scatterers are present throughout the layers of the Ad-LIGO dielectric coatings • Such large number of scatterers require a thermodynamic origin (see next presentation) LVC meeting 03/15/16 LIGO-G 1600430 21