A century after the prediction of the existence of gravitational waves by A. Einstein and after over fifty years of experimental efforts, gravitational waves have been detected at Earth directly. This result is a major achievement and opens new prospectives for the exploration of our universe. Gravitational waves carry different and complementary information about the source with respect to electromagnetic signals. In particular the first detection demonstrated the existence of stellar-mass black holes, binary systems of black holes and their coalescence. The detection was made by the LIGO instruments which are twin kilometer-scale Michelson interferometers in the US. These detectors represent the second generation of gravitational wave interferometers and, for the first time, they achieved the outstanding strain sensitivity of 10^(-23) Hz^(-1/2) between 90Hz and 400Hz. In the next months the LIGO network will be joined by another second generation detector: Advanced Virgo located near Pisa, Italy. The sensitivity of these advanced detectors is set by different noise sources. In particular, in the low frequency range (below 100Hz) major contributions come from thermal noises, gravity gradient noise and radiation pressure noise; instead, the high frequency band (above 100-200Hz) is dominated by shot noise. Quantum noise (radiation pressure and shot noise) is expected to dominate the detector sensitivity in the whole frequency band at the final target laser input power. To decrease the shot noise while increasing the radiation-pressure noise, or vice-versa, Caves \cite{Caves1981} proposed in 1981 the idea of the squeezed-state technique. The LIGO collaboration demonstrated for the first time in 2011 that the injection of a squeezed vacuum state into the dark port of the interferometer can reduce the shot noise due to the quantum nature of light. This result was achieved with the German-British interferometer GEO600 and was replicated in 2013 with the LIGO interferometer at Livingston. After these results, the LIGO collaboration have pursued further the research in the squeezed-state technique which is considered mandatory for third generation of ground based interferometric detectors. In 2013, the Virgo collaboration started developing the squeezed-state technique. The subject of my thesis is the realization of a prototype of frequency independent squeezed vacuum state source to be injected in Advanced Virgo. This prototype is developed in collaboration with other Virgo groups.
A century after the prediction of the existence of gravitational waves by A. Einstein and after over fifty years of experimental efforts, gravitational waves have been detected at Earth directly. This result is a major achievement and opens new prospectives for the exploration of our universe. Gravitational waves carry different and complementary information about the source with respect to electromagnetic signals. In particular the first detection demonstrated the existence of stellar-mass black holes, binary systems of black holes and their coalescence. The detection was made by the LIGO instruments which are twin kilometer-scale Michelson interferometers in the US. These detectors represent the second generation of gravitational wave interferometers and, for the first time, they achieved the outstanding strain sensitivity of 10^(-23) Hz^(-1/2) between 90Hz and 400Hz. In the next months the LIGO network will be joined by another second generation detector: Advanced Virgo located near Pisa, Italy. The sensitivity of these advanced detectors is set by different noise sources. In particular, in the low frequency range (below 100Hz) major contributions come from thermal noises, gravity gradient noise and radiation pressure noise; instead, the high frequency band (above 100-200Hz) is dominated by shot noise. Quantum noise (radiation pressure and shot noise) is expected to dominate the detector sensitivity in the whole frequency band at the final target laser input power. To decrease the shot noise while increasing the radiation-pressure noise, or vice-versa, Caves \cite{Caves1981} proposed in 1981 the idea of the squeezed-state technique. The LIGO collaboration demonstrated for the first time in 2011 that the injection of a squeezed vacuum state into the dark port of the interferometer can reduce the shot noise due to the quantum nature of light. This result was achieved with the German-British interferometer GEO600 and was replicated in 2013 with the LIGO interferometer at Livingston. After these results, the LIGO collaboration have pursued further the research in the squeezed-state technique which is considered mandatory for third generation of ground based interferometric detectors. In 2013, the Virgo collaboration started developing the squeezed-state technique. The subject of my thesis is the realization of a prototype of frequency independent squeezed vacuum state source to be injected in Advanced Virgo. This prototype is developed in collaboration with other Virgo groups.
The era of gravitational-wave astronomy has begun, with the detection of 5 confirmed binary black holes and a binary neutron star coalescence by the Advanced Laser Interferometer Gravitational- wave Observatory (aLIGO), and later with the advanced Virgo detector. These detections have already revealed a wealth of discoveries across the fields of nuclear physics, general relativity and astrophysics. The work presented in this thesis is part of the ongoing effort to improve the sensitivity of ground-based interferometric gravitational-wave detectors. The sensitivity of aLIGO, and other interferometric detectors, is broadly limited by quantum noise. Improving on the quantum noise will increase the astrophysical range of the detectors, and improve the source parameter estimation. One way to reduce quantum noise is to inject audio- band squeezed vacuum states into the detection port. This technique has been demonstrated on the initial LIGO and GEO600 detectors. A squeezed light source for aLIGO must meet stringent requirements in terms of optical loss, phase noise, and scattered light. The squeezer must produce high levels of audio-band squeezing and operate under vacuum, to take advantage of the excellent existing isolation systems and to minimise optical loss. At design sensitivity, squeezed states whose quantum noise depends on frequency will be required. We have demonstrated an ultra-stable glass-based squeezed light source, the first experiment of this kind to operate under vacuum. The squeezer cavity is constructed quasi-monolithically, with optics and nonlinear crystal oven optically contacted to a breadboard base. The cavity is designed to have extremely low length noise, and to produce high levels of audio-band squeezing. We have measured 8.6 ± 0.9 dB of squeezing and infer the generation of 14.2 ± 1.0 dB after accounting for all known losses. The squeezer has demonstrated record phase noise performance of 1.3 mradRMS, dominated by sources other than cavity length noise. This exceeds the phase-noise requirement for a squeezer for aLIGO. A copy of this squeezer is currently being installed in a squeezing- ellipse rotation experiment to demonstrate frequency-dependent squeezing for aLIGO. Lessons learnt during the construction and operation of the in-vacuum squeezer have helped inform the design of a frequency-independent squeezed light source currently being installed at the LIGO sites. Future gravitational-wave detectors will continue to use interferometric techniques, and will be limited by quantum noise for the foreseeable future. To improve on thermal noise limits and interferometer power handling, future detectors look to cryogenic silicon as a test mass material. To take advantage of the desirable properties of silicon, including low scatter and absorption, a longer operating wavelength is required. The proposed LIGO Voyager upgrade has an operating wavelength in the 2 μ m region, with the specific wavelength to be determined. LIGO Voyager will require a squeezed light source in the 2 μ m region to reach its design sensitivity. We present the design, characterisation, and results of the first squeezed light source in the 2 μ m region. Laser and detector technologies at 2 μ m are less developed than their 1064 nm counterparts, causing significant technical challenges. We have measured 4.0 ± 0.2 dB of squeez- ing at 1984 nm, limited by loss due to detector quantum efficiency. Accounting for known losses in the system, we infer the generation of 10 dB of squeezing. This is an important demonstra- tion of quantum noise reduction for future detectors, and a pathfinder technology for the design choices of LIGO Voyager. So far we have found no reason why a 2 μ m interferometer should not be feasible.
Direct detection of gravitational waves will require
earth based detectors to measure strains of the order 10-21, at
frequencies of 100 Hz, a sensitivity that has been accomplished
with the initial generation of LIGO interferometric gravitational
wave detectors. A new generation of detectors currently under
construction is designed improve on the sensitivity of the initial
detectors by about a factor of 10. The quantum nature of light will
limit the sensitivity of these Advanced LIGO interferometers at
most frequencies; new approaches to reducing the quantum noise will
be needed to improve the sensitivity further. This quantum noise
originates from the vacuum fluctuations that enter the unused port
of the interferometer and interfere with the laser light. Vacuum
fluctuations have the minimum noise allowed by Heisenberg's
uncertainty principle, [Delta]X1 [Delta]X2 >/=1, where the two
quadratures X1 and X2 are non-commuting observables responsible for
the two forms of quantum noise, shot noise and radiation pressure
noise. By replacing the vacuum fluctuations entering the
interferometer with squeezed states, which have lower noise in one
quadrature than the vacuum state, we have reduced the shot noise of
a LIGO interferometer. The sensitivity to gravitational waves
measured during this experiment represents the best sensitivity
achieved to date at frequencies above 200 Hz, and possibly the
first time that squeezing has been measured in an interferometer at
frequencies below 700 Hz. The possibility that injection of
squeezed states could introduce environmental noise couplings that
would degrade the crucial but fragile low frequency sensitivity of
a LIGO interferometer has been a major concern in planning to
implement squeezing as part of baseline interferometer operations.
These results demonstrate that squeezing is compatible with the low
frequency sensitivity of a full scale gravitational wave
interferometer. We also investigated the limits to the level of
squeezing observed, including optical losses and fluctuations of
the squeezing angle. The lessons learned should allow for
responsible planning to implement squeezing in Advanced LIGO,
either as an alternative to high power operation or an early
upgrade to improve the sensitivity. This thesis is available at
DSpace@MIT and has LIGO document number P1300006.
Direct detection of gravitational waves will require earth based detectors to measure strains of the order 10−21, at frequencies of 100 Hz, a sensitivity that has been accomplished with the initial generation of LIGO interferometric gravitational wave detectors. A new generation of detectors currently under construction is designed improve on the sensitivity of the initial detectors by about a factor of 10. The quantum nature of light will limit the sensitivity of these Advanced LIGO interferometers at most frequencies; new approaches to reducing the quantum noise will be needed to improve the sensitivity further. This quantum noise originates from the vacuum fluctuations that enter the unused port of the interferometer and interfere with the laser light. Vacuum fluctuations have the minimum noise allowed by Heisenberg’s uncertainty principle, ∆X1∆X2 ≥ 1, where the two quadratures X1 and X2 are non-commuting observables responsible for the two forms of quantum noise, shot noise and radiation pressure noise. By replacing the vacuum fluctuations entering the interferometer with squeezed states, which have lower noise in one quadrature than the vacuum state, we have reduced the shot noise of a LIGO interferometer. The sensitivity to gravitational waves measured during this experiment represents the best sensitivity achieved to date at frequencies above 200 Hz, and possibly the first time that squeezing has been measured in an interferometer at frequencies below 700 Hz. The possibility that injection of squeezed states could introduce environmental noise couplings that would degrade the crucial but fragile low frequency sensitivity of a LIGO interferometer has been a major concern in planning to implement squeezing as part of baseline interferometer operations. These results demonstrate that squeezing is compatible with the low frequency sensitivity of a full scale gravitational wave interferometer. We also investigated the limits to the level of squeezing observed, including optical losses and fluctuations of the squeezing angle. The lessons learned should allow for responsible planning to implement squeezing in Advanced LIGO, either as an alternative to high power operation or an early upgrade to improve the sensitivity. This thesis is available at DSpace@MIT and has LIGO document number P1300006. Thesis Supervisor: Nergis Mavalvala Title: Professor of Physics
The detection of the binary neutron star inspiral GW170817 by the Advanced Laser Interferometer Gravitational Wave Observatory (LIGO) and Advanced Virgo, followed by counterpart observations across a broad electromagnetic spectrum, marks the dawn of multi-messenger astronomy. The discovery rate and potential of gravitational-wave astronomy is limited by the detector's sensitivity. A key challenge in improving the sensitivity of gravitational-wave detectors is to reduce quantum noise which arises due to the quantum nature of light. A well established technique to reduce the quantum noise in the detector is by replacing the coherent vacuum fluctuations entering the readout port with a squeezed vacuum state. This technique has been successfully demonstrated at the GEO-600 and LIGO Hanford detector, and is now in routine operation in current gravitational wave detectors. This thesis presents research from three experiments to develop quantum noise reduction techniques for next-generation gravitational wave detector using non-classical quantum states such as squeezed states. Current proposals for future generation gravitational wave detectors, such as LIGO Voyager, plan to utilise cryogenic silicon instead of fused silica as the test mass material. This is to improve the power handling of the detector and reduce thermal noise. The switch to silicon test masses requires a change of laser wavelength to the 2 micron region. In the first experiment, we built and operated the first squeezed vacuum source at a wavelength of 1984 nm. With a stable squeezing ellipse phase control employed, 3.9 dB of squeezing was measured, predominately limited by losses due to the low quantum efficiency of the photodiode. The squeezed vacuum system presented is compatible with future detectors and is a key pathfinder experiment to inform the design of cryogenic silicon interferometers. The quantum nature of light results in quantum radiation pressure noise that limits the low frequency sensitivity of gravitational wave detectors. In collaboration with Louisiana State University, the second experiment demonstrates the reduction of quantum radiation pressure noise by injecting bright squeezed light into a centimetre-scale optomechanical cavity at room temperature. The experiment involves the development and integration of a low-frequency bright squeezed source to reduce the quantum radiation pressure noise by 1.2 dB at 20 kHz. The result also represents the lowest frequency bright squeezed state measured from a crystal-based squeezer. Quadrature rotation of the injected squeezed state is required to achieve broadband quantum noise reduction in gravitational wave detectors due to the quantum back-action effect. This requires the squeezed state to interact with a low loss optical filter with a low frequency bandwidth prior to injection and represents a significant technical challenge. The final experiment presented in this thesis demonstrates an alternate method where frequency-dependent squeezing with Einstein-Podolsky-Rosen (EPR) entangled states is generated. When integrated into a gravitational wave detector, the alternate method allows the detector to simultaneously serve as the optical filter, eliminating the need for the additional apparatus.
Current gravitational wave detectors, such as LIGO, Virgo and KAGRA, are laser interferometers. This technique consists of injecting a laser beam through an input port, which is then separated into two arms several kilometers long by a semi-reflecting mirror. At the end of each arm, a mirror reflects thebeam back to the semi-reflecting mirror. The two beams then interfere and the power of the resulting beam, measured at the detection port, contains information about the relative length of the two arms. When a gravitational wave passes through the detector, this relative length varies in time. Several noise sources degrade the quality of the recorded signal, which affects the number of gravitational waves that can be observed. Currently, this sensitivity is limited at high-frequency by the quantum noise.An electromagnetic wave described by quantum mechanics has a classical amplitude and quantum fluctuations in amplitude and phase. A state with a null classical amplitude is called a vacuum state, but still has these quantum fluctuations. Thus, in the detector, although no light beam is introduced through the detection port, the vacuum state is present. It is then the quantum fluctuations of this vacuum state that are responsible for the quantum noise of the interferometer. In particular, the phase and amplitude fluctuations are respectively responsible for the quantum noise at high and low frequencies; the separation between these two regions occurs at about 100 Hz.The work presented in this thesis focuses on the injection of squeezed vacuum states into the Virgo detector via the detection port in order to lower the quantum noise. These states are characterized by the fact that their phase (amplitude) fluctuations are reduced, which allows reducing the quantum noise at high (low) frequency. As the Virgo detector is currently limited only by high-frequency quantum noise, only phase squeezed vacuum states were used during the last observation period O3. It was first necessary to ensure that these squeezed vacuum states had the proper geometric and phase parameters with respect to the interferometer beam such that they were not degraded. Control loops allowed these parameters to be maintained at their nominal values during O3, a year long period. The decrease in high-frequency quantum noise, but also the increase in low-frequency quantum noise, have been measured for the first time with the Virgo detector. These observations were in agreement with the measurements of the losses that affect the squeezed vacuum states.From O4 , quantum noise will also be limiting at low frequency. It will then be possible to use squeezed vacuum states having their amplitude fluctuations reduced at low frequency and their phase fluctuations reduced at high frequency. Thus, the total quantum noise can be reduced. Such states can be generated using an optical cavity called a filter cavity. The second part of the work presented in this thesis consisted in defining the optical parameters of the filter cavity that will be installed on Virgo for O4. To do this, it was first necessary to study the sources of degradation of the squeezed states of the vacuum. This then allowed defining the length of the filtering cavity and the set of parameters of the mirrors that compose it. Finally, the implementation of this filter cavity within the Virgo infrastructure was studied. This has shown that the use of compressed states will effectively reduce quantum noise over the entire sensitivity spectrum of Virgo.
Current gravitational wave detectors, such as LIGO, Virgo and KAGRA, are laser interferometers. This technique consists of injecting a laser beam through an input port, which is then separated into two arms several kilometers long by a semi-reflecting mirror. At the end of each arm, a mirror reflects the beam back to the semi-reflecting mirror. The two beams then interfere and the power of the resulting beam, measured at the detection port, contains information about the relative length of the two arms. When a gravitational wave passes through the detector, this relative length varies in time. Several noise sources degrade the quality of the recorded signal, which affects the number of gravitational waves that can be observed. Currently, this sensitivity is limited at high-frequency by the quantum noise.An electromagnetic wave described by quantum mechanics has a classical amplitude and quantum fluctuations in amplitude and phase. A state with a null classical amplitude is called a vacuum state, but still has these quantum fluctuations. Thus, in the detector, although no light beam is introduced through the detection port, the vacuum state is present. It is then the quantum fluctuations of this vacuum state that are responsible for the quantum noise of the interferometer. In particular, the phase and amplitude fluctuations are respectively responsible for the quantum noise at high and low frequencies; the separation between these two regions occurs at about 100 Hz.The work presented in this thesis focuses on the injection of squeezed vacuum states into the Virgo detector via the detection port in order to lower the quantum noise. These states are characterized by the fact that their phase (amplitude) fluctuations are reduced, which allows reducing the quantum noise at high (low) frequency. As the Virgo detector is currently limited only by high-frequency quantum noise, only phase squeezed vacuum states were used during the last observation period O3. It was first necessary to ensure that these squeezed vacuum states had the proper geometric and phase parameters with respect to the interferometer beam such that they were not degraded. Control loops allowed these parameters to be maintained at their nominal values during O3, a year long period. The decrease in high-frequency quantum noise, but also the increase in low-frequency quantum noise, have been measured for the first time with the Virgo detector. These observations were in agreement with the measurements of the losses that affect the squeezed vacuum states.From O4 , quantum noise will also be limiting at low frequency. It will then be possible to use squeezed vacuum states having their amplitude fluctuations reduced at low frequency and their phase fluctuations reduced at high frequency. Thus, the total quantum noise can be reduced. Such states can be generated using an optical cavity called a filter cavity. The second part of the work presented in this thesis consisted in defining the optical parameters of the filter cavity that will be installed on Virgo for O4. To do this, it was first necessary to study the sources of degradation of the squeezed states of the vacuum. This then allowed defining the length of the filtering cavity and the set of parameters of the mirrors that compose it. Finally, the implementation of this filter cavity within the Virgo infrastructure was studied. This has shown that the use of compressed states will effectively reduce quantum noise over the entire sensitivity spectrum of Virgo.