The research on amorphous coatings for future GW detectors - PowerPoint PPT Presentation

The research on amorphous coatings for future GW detectors Francesco Piergiovanni University of Urbino - INFN Firenze on behalf of the Virgo Collaboration

Thermal Noise • It characterizes each dissipative mechanical system at thermal equilibrium • It’s produced by irreversible processes with typical time constants and activation energies • Same mechanism produces energy dissipation and thermal fluctuation Fluctuation-Dissipation Theorem Mechanical dissipation is quantified by the loss angle 𝑇 𝑦 𝜕 = 4𝑙 𝐶 𝑈 Dissipation 𝐹 𝑚𝑝𝑡𝑢 𝑞𝑓𝑠 𝑑𝑧𝑑𝑚𝑓 𝜕 2 𝑆𝑓[𝑍 𝜕 ] 𝜚 𝜕 = 2𝜌𝐹 𝑡𝑢𝑝𝑠𝑓𝑒 ( Callen, Welton 1951) Fluctuation 𝜚 𝜕 = 1/𝑅 @ resonances Thermal noise can be reduced: • low temperature • low dissipation material 2 TAUP 2019 - TOYAMA

The origin of the TN amorphous materials • In Two Level System (TLS): metastable states are separated 𝜐 ∝ 𝜐 0 Δ exp 𝑊/ 𝑙 𝑐 𝑈 Gilroy & Phillips by an energy barrier 𝑊 : barrier high Philosophical Magazine B (1981) • Transitions between the two levels explain the anelastic behavior of amorphous materials • Only transitions with a relaxation time of the same order of the period of the strain wave propagating in the material produce mechanical losses. At room temperature ( 300 𝐿 ) only TLS with 𝑊 ≈ 0.5𝑓𝑊 are relevant 𝜐 Too fast relaxation 𝜐 Too slow relaxation 𝜐 Relaxation producing losses Dove et al. Mineralogical Magazine (2000) 3 TAUP 2019 - TOYAMA

Coating In GW detectors SiO2 thinner layer (cap) • Alternate layers of transparent materials with different Ta2O5-TiO2 layer N doublets index of refraction (Bragg reflection). ~6 µm SiO2 layer Ta2O5-TiO2 thicker layer • Impedance mismatch and interference produce high coefficient of reflectivity. 35 cm • Its structure is not compact as the substrate. • 5 𝜈𝑛 of coating produces more thermal noise than 20 cm 20 𝑑𝑛 of substrate. 40 kg 2 2𝑂 𝑜 𝐼 − 1 Aasi J et al. 2015 Classical and Quantum Gravity 32 𝑜 𝑀 𝑆 = 𝑠 2 = Acernese F et al. 2015 Classical and Quantum Gravity 32 2𝑂 𝑜 𝐼 + 1 𝑜 𝑀 4 TAUP 2019 - TOYAMA

Coating thermal noise Advanced Virgo Coating Thermal Noise (CTN) limits the detection band in the « bucket » (middle frequencies) which is the most sensitive region of the Advanced and the future Advanced+ GW detectors 5 TAUP 2019 - TOYAMA

Coating thermal noise (CTN) Coating thermal noise (CTN) contribution goes like: 𝑇 𝑦 𝜕 ∝ 𝑈 𝑒 𝑥 2 𝜚 𝐷 Research to reduce CTN involves: • enlarging the laser beam size 𝑥 ; • minimizing the overall coating thickness ( 𝑒 ) increasing the contrast between the high and the low refractive index materials in Bragg reflector; • finding new materials-techniques for reducing the coating loss angle 𝜚 𝐷 . 6 TAUP 2019 - TOYAMA

The Virgo Coating R&D activities Sysnthesis: Modeling: ✓ Coating deposition ✓ Structure ✓ Heat treatments ✓ TLS relaxations Macroscopic Microscopic characterization: characterization: ✓ Loss angle measurements ✓ TEM, SEM ✓ Optical characterization ✓ Raman, Brillouin ✓ Dielectric response ✓ XRD, XPS, XAS ✓ Elastic constants ✓ AFM ✓ Density 7 TAUP 2019 - TOYAMA

The Virgo Coatings R&D Formed on January 2017 Collaboration ● LMA ♦ IBS HighT, IAD ● PADOVA ● GENOVA ♦ GeNS [300-10] K ♦ FEA ♦ Mag. Sputtering ♦ Ellipsometry ♦ Optical metrology ♦ XRD High T ♦ Optical properties ♦ AFM, XPS LMA ● g-MAG ♦ Raman ● URBINO ♦ Pulsed Laser Dep. ♦ Rapid Th. Annealing ♦ GeNS Cryo GENOVA PADOVA g-MAG ♦ Raman, Brillouin ♦ FEA ♦ Physics of Glasses URBINO ♦ Molecular Dynamics PISA ● PERUGIA-CAMERINO ● PISA ♦ Cantilevers & GeNS Cryo PERUGIA ♦ Physics of Glasses VIRGO ♦ Study of the crystallization processes ♦ Brillouin, Raman ROMA 2 ♦ Physics of deposition and ● ROMA 2 ♦ SEM, XRD, XAS ultrastable glasses ♦ Laser Polishing ♦ Molecular Dynamics and Modelling ♦ GeNS 300K 1’’ ♦ Calorimetry and Dielectric response ROMA 1 ♦ FEA and AFM ♦ XPS SALERNO ♦ Sample production ♦ Ellipsometry SANNIO ♦ Characterization ● SALERNO/SANNIO ● ROMA 1 ♦ IAD ♦ Structural Other groups (from KAGRA and ♦ SEM,TEM,AFM and XRD characterization ♦ nanolayered composites ♦ Thermobalance Belgium) are interested to join Credits: G. Cagnoli and Mie-metamaterials 8

The Virgo Coating R&D research lines Oxides ● Materials Deposition Mixing ● ● High Temperature High Index ● ● Nano-layering Silica Glasses ‒ High index ‒ Low index Nitrides ● BETTER ‒ HR stack Fluorides ● COATING High Coordination ● Number Glasses SiN, GaN, SiC, etc Post-deposition Origin of absorption ● ● Annealing Absorption and treatments Loss measurement ● Metrology ● Outgassing and protocol chemical status Thermo-elastic ● ● Controlled effect crystallization O5 horizon, CRD project accepted by funding agencies Beyond O5 9 TAUP 2019 - TOYAMA

Updated measurement of current coatings 𝑈𝑏 2 𝑃 5 𝑈𝑗𝑃 2 : 𝑈𝑏 2 𝑃 5 𝜚 𝑑 (𝑔) = 𝑏 𝑔 𝑐 𝜚 𝑑 𝑔 = 𝑏 𝑔 𝑐 + 𝜗 𝑒 𝜚 𝑓 𝑇𝑗𝑃 2 𝜚 𝑑 = 𝑏 𝑔 𝑐 Cagnoli et al. PLA (2018), M. Granata, OIC 2019 Refractive index: ቊ 𝑜 𝐼 = 2.09 𝑜 𝑀 = 1.45 Extinction coeff.: 𝑙 ≈ 10 −7 10 TAUP 2019 - TOYAMA

Oxide mixtures recap High refractive index 𝑂𝑐 2 𝑃 5 500°C 𝑈𝑏 2 𝑃 5 500°𝐷 𝑈𝑗𝑃 2 : 𝑈 𝑏2 𝑃 5 500°C 𝑈𝑗𝑃 2 : 𝑈𝑏 2 𝑃 5 400 °C 𝑎𝑠𝑃 2 : 𝑈𝑏 2 𝑃 5 700 °C Post-deposition annealing temperature are reported 𝑇𝑗𝑃 2 500°𝐷 𝑇𝑗𝑃 2 900°𝐷 Low refractive index Granata et al. Optical Interference Coatings Conference (OIC) 2019 11 TAUP 2019 - TOYAMA

Annealing Amato et al. J. Phys Conf. Ser. 957 (2018) Granata et al Phys. Rev. Mater. 2 (2018) SiO 2 𝑇𝑗𝑃 2 IBS GC – not annealed – IBS SPECTOR – annealed – fused silica – bulk – 12 TAUP 2019 - TOYAMA

Deposition parameters (IBS coating) as deposited annealed 𝐻𝑠𝑏𝑜𝑏𝑢𝑏 𝑓𝑢 𝑏𝑚, 𝑗𝑜 𝑞𝑠𝑓𝑞𝑏𝑠𝑏𝑢𝑗𝑝𝑜 𝐵𝑛𝑏𝑢𝑝 𝑓𝑢 𝑏𝑚, 𝐾. 𝑄ℎ𝑧𝑡. 𝐷𝑝𝑜𝑔. 𝑇𝑓𝑠. 957 (2018) erasing effect SPECTOR 𝑈𝑏 2 𝑃 5 ✓ Films deposited with different DIBS coaters (at different rates) show GC different loss angles structural limit? ✓ Slower is the deposition, lower the dissipation as deposited annealed ✓ A sort of erasing effect of the SPECTOR annealing is visible, which is more evident for the tantala 𝑇𝑗𝑃 2 GC coating 13 TAUP 2019 - TOYAMA

Absorption vs mechanical losses Urbach tail: broadening in the band-gap absorption edge especially visible in poor crystallin ✓ A strong correlation and amorphous material, related to thermal and between 𝐹 𝑉 and structural disorder mechanical losses has 𝛽 𝐹 ∝ 𝑓 𝐹/𝐹 𝑉 been found Urbach energy Absorption coefficient ✓ Different materials show the same behavior Ellipsometry ✓ Rapid estimation of mechanical losses Amato et al. ✓ Extend the range of structural analysis arXiv:1903.06094 Photon energy (eV) 14 TAUP 2019 - TOYAMA

Modelling: Molecular Dynamics MD-DMS: Molecular Dynamics simulation of Dynamical Mechanical Spectroscopy • A theory-independent method: the only ingredient is the interatomic potential of the specific glassy system ( 𝑈𝑏 2 𝑃 5 , 𝑇𝑗𝐷 in progress) • Mechanical losses are computed by the dephasing btw applied oscillating strain and the resulting stress • Simulation frequency range is GHz to THz , but the frequency power law is compatible with what has been experimentally found in acoustic band • MD-DMS and experimental results, 𝑈𝑏 2 𝑃 5 are in good agreement MD-DMS makes possible a rapid evaluation of mechanical properties of new materials Puosi et al. in preparation Puosi 10th Einstein Telescope Symposium 2019 15 TAUP 2019 - TOYAMA

New materials Granata et al. Optical Interference Coatings Conference (OIC) 2019 IBS coatings Amato et al. Journal of Physics: Conf. Series 957 (2018) Nitrides: 𝑇𝑗𝑂 𝑦 is a promising high index material 𝑈𝑏 2 𝑃 5 • Mechanical losses about 3x better than 𝑈𝑏 2 𝑃 5 issues • Refractive index slightly lower than 𝑈𝑏 2 𝑃 5 𝑇𝑗𝑂 𝑦 • Extinction coefficient more than 10x larger!!! 𝑁𝑕𝐺 2 as deposited Fluorides: 𝑁𝑕𝐺 2 and 𝐵𝑚𝐺 3 𝑁𝑕𝐺 2 400°C • lower refractive index than silica Bischi et al. poster GWADW 2019 issues • Higher mechanical losses than silica 𝑇𝑗𝑃 2 500°𝐷 (at least at room T) 𝑇𝑗𝑃 2 900°𝐷 16 TAUP 2019 - TOYAMA

Nano-layered coating deposition • Post-deposition annealing largely improves coating mechanical characteristics. • Maximum annealing temperature is limited by the beginning of the film crystallization (i.e. T max ≈ 300 °𝐷 for 𝑈𝑗𝑃 2 ) • In nano-layered films crystallization is frustrated by the size of the layers, higher annealing temperature is allowed TiO 2 /SiO 2 stack TiO 2 = 1.8 𝑜𝑛 𝑦 38 SiO 2 = 3.6 𝑜𝑛 𝑦 37 TiO 2 = 7.36 𝑜𝑛 𝑦 10 SiO 2 = 4.32 𝑜𝑛 𝑦 9 Kuo et al., Opt. Lett.44, 247-250 (2019); Chao et al., 41 st PIERS 2019; Principe, Opt. Express 23, 10938-10956 (2015) 17 TAUP 2019 - TOYAMA