上海交通大学：《探索物质微观世界》课程教学资源（补充材料）引力波_The advanced LIGO input optics.pdf_大学文库

rossMark REVIEW OF SCIENTIFIC INSTRUMENTS 87.014502 (2016) The advanced LIGO input optics Chris L.Mueller,1.a)Muzammil A.Arain,1.b)Giacomo Ciani,1 Ryan.T.DeRosa,2 Anamaria Effler,2 David Feldbaum,1 Valery V.Frolov,3 Paul Fulda,1 Joseph Gleason,1 Matthew Heintze,1 Keita Kawabe,4 Eleanor J.King,5 Keiko Kokeyama,2 William Z.Korth,6 Rodica M.Martin,1 Adam Mullavey,3 Jan Peold,7 Volker Quetschke,8 David H.Reitze,1.c) David B.Tanner,1 Cheryl Vorvick,4 Luke F.Williams,1 and Guido Mueller1 University of Florida,Gainesville.Florida 32611,USA Louisiana State University,Baton Rouge,Louisiana 70803,USA 3LIGO Livingston Observatory,Livingston,Louisiana 70754,USA LIGO Hanford Observatory,Richland,Washington 99352,USA SUniversity of Adelaide,Adelaide.SA 5005,Australia LIGO,California Institute of Technology,Pasadena,California 91125,USA Max-Planck-Institut fuir Gravitationsphysik,30167 Hannover.Germany University of Texas at Brownsville,Brownsville,Texas 78520,USA (Received 27 July 2015:accepted 19 November 2015;published online 22 January 2016) The advanced LIGO gravitational wave detectors are nearing their design sensitivity and should begin taking meaningful astrophysical data in the fall of 2015.These resonant optical interferometers will have unprecedented sensitivity to the strains caused by passing gravitational waves.The input optics play a significant part in allowing these devices to reach such sensitivities.Residing between the pre-stabilized laser and the main interferometer,the input optics subsystem is tasked with preparing the laser beam for interferometry at the sub-attometer level while operating at continuous wave input power levels ranging from 100 mW to 150 W.These extreme operating conditions required every major component to be custom designed.These designs draw heavily on the experience and understanding gained during the operation of Initial LIGO and Enhanced LIGO.In this article,we report on how the components of the input optics were designed to meet their stringent requirements and present measurements showing how well they have lived up to their design.2016 AlP Publishing LLC.[http://dx.doi.org/10.1063/1.4936974] I.INTRODUCTION inject it into the main IFO.The PSL consists of a master laser, an amplifier stage,and a 200 W slave laser which is injection A worldwide effort to directly detect gravitational radi- locked to the amplified master laser.The 200 W output beam ation in the 10Hz to a few kHz frequency range with large is filtered by a short optical ring cavity,the pre-mode cleaner, scale laser interferometers (IFOs)has been underway for the before it is turned over to the IO(see Figure 1).The PSL pre- past two decades.In the United States the Laser Interferometer stabilizes the laser frequency to a fixed spacer reference cavity Gravitational-Wave Observatories (LIGO)in Livingston,LA using a tunable sideband locking technique.The PSL also (LLO)and in Hanford,WA,(LHO)have been operating provides interfaces to further stabilize its frequency and power. since the early 2000's.Initial and Enhanced LIGO (eLIGO) The IFO is a dual-recycled,cavity-enhanced Michelson produced several significant upper limits but did not have the interferometer-as sketched in Figure 2.The field enters the sensitivity to make the first direct detection of gravitational 55 m folded power recycling cavity(PRC)through the power waves.During this time of operation a significant amount recycling mirror(PRM).Two additional mirrors(PR2.PR3) of effort was invested by the LIGO Scientific Collaboration within the PRC form a telescope to increase the beam size to research and design Advanced LIGO (aLIGO),the first from ~2 mm to ~50 mm (Gaussian beam radius)before the major upgrade of Initial LIGO.In 2011 the Initial LIGO large beam is split at the beam splitter and injected into the detectors were decommissioned and installation of these up- two 4 km arm cavities formed by the input and end test masses. grades started.The installation was completed in 2014 and The reflected fields recombine at the BS and send most of the the commissioning phase has begun for many of the upgraded light back to the PRM where it constructively interferes with subsystems at the LIGO observatories.This paper focuses on the injected field.3 This leads to a power enhancement inside the input optics (IO)of aLIGO. the power recycling cavity and provides additional spatial,fre- The main task of the IO subsystem is to take the laser beam quency,and amplitude filtering of the laser beam.The second from the pre-stabilized laser system(PSL)and prepare and output of the BS sends light into the 55 m long folded signal recycling cavity(SRC)which also consists of a beam reduc- ing telescope(SR2,SR3)and the partially reflective signal aElectronic mail:cmueller@phys.ufl.edu b)Present address:KLA-Tencor,Milpitas,California 95035.USA. recycling mirror(SRM). )Present address:LIGO Laboratory.California Institute of Technology, This paper is organized as follows:Section II gives an Pasadena,California 91125.USA. overview of the IO;its functions,components,and the 0034-6748/2016/87(1)/014502/16/$30.00 87,014502-1 2016 AIP Publishing LLC Reuse of AlP Publishing cor subject fo the terms at:https://publishing.aip.org/authors/rights-and-permissions.Download to IP 183195.2516 On:Fri.22Ap1 20160051:35

REVIEW OF SCIENTIFIC INSTRUMENTS 87, 014502 (2016) The advanced LIGO input optics Chris L. Mueller, 1,a) Muzammil A. Arain, 1,b) Giacomo Ciani, 1 Ryan. T. DeRosa, 2 Anamaria Effler, 2 David Feldbaum, 1 Valery V. Frolov, 3 Paul Fulda, 1 Joseph Gleason, 1 Matthew Heintze, 1 Keita Kawabe, 4 Eleanor J. King, 5 Keiko Kokeyama, 2 William Z. Korth, 6 Rodica M. Martin, 1 Adam Mullavey, 3 Jan Peold, 7 Volker Quetschke, 8 David H. Reitze, 1,c) David B. Tanner, 1 Cheryl Vorvick, 4 Luke F. Williams, 1 and Guido Mueller1 1University of Florida, Gainesville, Florida 32611, USA 2Louisiana State University, Baton Rouge, Louisiana 70803, USA 3LIGO Livingston Observatory, Livingston, Louisiana 70754, USA 4LIGO Hanford Observatory, Richland, Washington 99352, USA 5University of Adelaide, Adelaide, SA 5005, Australia 6LIGO, California Institute of Technology, Pasadena, California 91125, USA 7Max-Planck-Institut für Gravitationsphysik, 30167 Hannover, Germany 8University of Texas at Brownsville, Brownsville, Texas 78520, USA (Received 27 July 2015; accepted 19 November 2015; published online 22 January 2016) The advanced LIGO gravitational wave detectors are nearing their design sensitivity and should begin taking meaningful astrophysical data in the fall of 2015. These resonant optical interferometers will have unprecedented sensitivity to the strains caused by passing gravitational waves. The input optics play a significant part in allowing these devices to reach such sensitivities. Residing between the pre-stabilized laser and the main interferometer, the input optics subsystem is tasked with preparing the laser beam for interferometry at the sub-attometer level while operating at continuous wave input power levels ranging from 100 mW to 150 W. These extreme operating conditions required every major component to be custom designed. These designs draw heavily on the experience and understanding gained during the operation of Initial LIGO and Enhanced LIGO. In this article, we report on how the components of the input optics were designed to meet their stringent requirements and present measurements showing how well they have lived up to their design. C 2016 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4936974] I. INTRODUCTION A worldwide effort to directly detect gravitational radiation in the 10 Hz to a few kHz frequency range with large scale laser interferometers (IFOs) has been underway for the past two decades. In the United States the Laser Interferometer Gravitational-Wave Observatories (LIGO) in Livingston, LA (LLO) and in Hanford, WA, (LHO) have been operating since the early 2000’s. Initial and Enhanced LIGO (eLIGO) produced several significant upper limits but did not have the sensitivity to make the first direct detection of gravitational waves. During this time of operation a significant amount of effort was invested by the LIGO Scientific Collaboration to research and design Advanced LIGO (aLIGO), the first major upgrade of Initial LIGO. In 2011 the Initial LIGO detectors were decommissioned and installation of these upgrades started. The installation was completed in 2014 and the commissioning phase has begun for many of the upgraded subsystems at the LIGO observatories. This paper focuses on the input optics (IO) of aLIGO. The main task of the IO subsystem is to take the laser beam from the pre-stabilized laser system1 (PSL) and prepare and a)Electronic mail: cmueller@phys.ufl.edu b)Present address: KLA-Tencor, Milpitas, California 95035, USA. c)Present address: LIGO Laboratory, California Institute of Technology, Pasadena, California 91125, USA. inject it into the main IFO. The PSL consists of a master laser, an amplifier stage, and a 200 W slave laser which is injection locked to the amplified master laser. The 200 W output beam is filtered by a short optical ring cavity, the pre-mode cleaner, before it is turned over to the IO (see Figure 1). The PSL prestabilizes the laser frequency to a fixed spacer reference cavity using a tunable sideband locking technique. The PSL also provides interfaces to further stabilize its frequency and power. The IFO is a dual-recycled, cavity-enhanced Michelson interferometer2 as sketched in Figure 2. The field enters the 55 m folded power recycling cavity (PRC) through the power recycling mirror (PRM). Two additional mirrors (PR2, PR3) within the PRC form a telescope to increase the beam size from ∼2 mm to ∼50 mm (Gaussian beam radius) before the large beam is split at the beam splitter and injected into the two 4 km arm cavities formed by the input and end test masses. The reflected fields recombine at the BS and send most of the light back to the PRM where it constructively interferes with the injected field.3 This leads to a power enhancement inside the power recycling cavity and provides additional spatial, frequency, and amplitude filtering of the laser beam. The second output of the BS sends light into the 55 m long folded signal recycling cavity4 (SRC) which also consists of a beam reducing telescope (SR2, SR3) and the partially reflective signal recycling mirror (SRM). This paper is organized as follows: Section II gives an overview of the IO; its functions, components, and the 0034-6748/2016/87(1)/014502/16/$30.00 87, 014502-1 © 2016 AIP Publishing LLC Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 183.195.251.6 On: Fri, 22 Apr 2016 00:51:35

014502-3 Mueller et al. Rev.Sci.Instrum.87,014502(2016) reference for many of the required measurements;this will be range of the beam (~13 m)is too large to allow easy access discussed in Section VI. to the second degree of freedom with a second piezo-actuated mirror. Between the lenses is a wedge to pick off a small fraction II.OVERVIEW OF THE INPUT OPTICS of the laser beam for diagnostic purposes.A fast photode- Figure 3 shows a sketch of the first part of the input optics. tector monitors the residual amplitude modulation at the phase This part is co-located with the PSL on the same optical table modulation frequencies while a second photodetector moni- inside the laser enclosure,outside of the vacuum system.It tors the DC power.A fraction of the main beam also transmits prepares the laser beam for the injection into the vacuum sys- through the bottom periscope mirror and is used to monitor the tem.The beam from the PSL is first routed through a half-wave power going into the vacuum system as well as the size,shape, plate and a polarizing beam splitter.These two elements form and quality of the beam. a manual power control stage which is used mainly during Following the periscope,the main beam is sent through a alignment processes on the optical table.The following mirror metal tube which includes a mechanical shutter and through transmits 2.5%of the light.This light is used by the arm length HAM16 into HAM2;all in-vacuum IO components are stabilization (ALS)system during lock acquisition of the main mounted on seismically isolated optical tables inside HAM2 interferometer.3 and HAM3.As shown in Figure 4,the beam passes over the Most of the light is sent through an electro-optic modu- Faraday isolator to a second periscope which lowers the beam lator which modulates the phase of the laser field with three to the in-vacuum beam height.The next element in the IO different modulation frequencies.Two of these frequencies are is the suspended IMC,a 33 m long (round-trip)triangular used by the interferometer sensing and control (ISC)system cavity.The two flat input and output mirrors,named MCI to sense most of the longitudinal and alignment degrees of and MC3,respectively,are located in HAM2 while the third freedom of the mirrors inside the IFO and to stabilize the curved mirror,MC2,is located in HAM3.Following MC3 are laser frequency and the alignment of the laser beam into the two suspended mirrors,IMI and IM2.which steer the beam interferometer.The third frequency is used to control the input through the Faraday isolator.IM3 and IM4 are used to steer mode cleaner.The two lenses LI and L2 mode match the beam the beam into the PRC.IM2 and IM3 are curved to mode to the in-vacuum input mode cleaner (IMC).The next steering match the output mode of the IMC to the mode of the main mirror directs the beam through another half-wave plate inside interferometer. a motorized rotation stage in front of two thin film polarizers. Two of the steering mirrors,IMI and IM4,transmit a This second power control stage is used during operations to small fraction of the light creating three different auxiliary adjust the power to the requested level.The periscope raises beams which are used to monitor the power and spatial mode the height of the beam and steers it into the vacuum system. of the IMC transmitted beam,of the beam going into the IFO, The top mirror is mounted on a piezo-actuated mirror mount to and of the beam which is reflected from the IFO.The latter fine tune the alignment of the beam into the vacuum chamber. two beams are routed to IOT2R,7 an optical table on the right A single piezo-actuated mirror is used because the Rayleigh side of HAM2,while the first beam and the field which is to ALS EOM PBS HWP RWP DCPD from PSL REPD Diag. BD DCPD to Vacuum Periscope ALS:Arm length stabilization system L1,L2:Lenses BD:Beam dump RFPD:Fast photo-detector DCPD:Photo-detector RWP:Rotating half-wave plate Diag:to Diagnostics TFP:Thin film polarizer EOM:Electro-optic modulator ∠：Water cooled beam dump HWP:Half-wave plate FIG.3.The IO on the in-air PSL table modulates the phase of the laser beam with the EOM,mode matches the light into the input mode cleaner (located inside the vacuum system),and controls the power injected into vacuum system. Reuse of AlP Publishing content is subject to the terms at:https://publishing.aip.org/authors/rights-and-permissions.Download to IP:183.195.251.6 On:Fri.22 Apr 20160051:35

014502-3 Mueller et al. Rev. Sci. Instrum. 87, 014502 (2016) reference for many of the required measurements; this will be discussed in Section VI. II. OVERVIEW OF THE INPUT OPTICS Figure 3 shows a sketch of the first part of the input optics. This part is co-located with the PSL on the same optical table inside the laser enclosure, outside of the vacuum system. It prepares the laser beam for the injection into the vacuum system. The beam from the PSL is first routed through a half-wave plate and a polarizing beam splitter. These two elements form a manual power control stage which is used mainly during alignment processes on the optical table. The following mirror transmits 2.5% of the light. This light is used by the arm length stabilization (ALS) system during lock acquisition of the main interferometer.5 Most of the light is sent through an electro-optic modulator which modulates the phase of the laser field with three different modulation frequencies. Two of these frequencies are used by the interferometer sensing and control (ISC) system to sense most of the longitudinal and alignment degrees of freedom of the mirrors inside the IFO and to stabilize the laser frequency and the alignment of the laser beam into the interferometer. The third frequency is used to control the input mode cleaner. The two lenses L1 and L2 mode match the beam to the in-vacuum input mode cleaner (IMC). The next steering mirror directs the beam through another half-wave plate inside a motorized rotation stage in front of two thin film polarizers. This second power control stage is used during operations to adjust the power to the requested level. The periscope raises the height of the beam and steers it into the vacuum system. The top mirror is mounted on a piezo-actuated mirror mount to fine tune the alignment of the beam into the vacuum chamber. A single piezo-actuated mirror is used because the Rayleigh range of the beam (∼13 m) is too large to allow easy access to the second degree of freedom with a second piezo-actuated mirror. Between the lenses is a wedge to pick off a small fraction of the laser beam for diagnostic purposes. A fast photodetector monitors the residual amplitude modulation at the phase modulation frequencies while a second photodetector monitors the DC power. A fraction of the main beam also transmits through the bottom periscope mirror and is used to monitor the power going into the vacuum system as well as the size, shape, and quality of the beam. Following the periscope, the main beam is sent through a metal tube which includes a mechanical shutter and through HAM16 into HAM2; all in-vacuum IO components are mounted on seismically isolated optical tables inside HAM2 and HAM3. As shown in Figure 4, the beam passes over the Faraday isolator to a second periscope which lowers the beam to the in-vacuum beam height. The next element in the IO is the suspended IMC, a 33 m long (round-trip) triangular cavity. The two flat input and output mirrors, named MC1 and MC3, respectively, are located in HAM2 while the third curved mirror, MC2, is located in HAM3. Following MC3 are two suspended mirrors, IM1 and IM2, which steer the beam through the Faraday isolator. IM3 and IM4 are used to steer the beam into the PRC. IM2 and IM3 are curved to mode match the output mode of the IMC to the mode of the main interferometer. Two of the steering mirrors, IM1 and IM4, transmit a small fraction of the light creating three different auxiliary beams which are used to monitor the power and spatial mode of the IMC transmitted beam, of the beam going into the IFO, and of the beam which is reflected from the IFO. The latter two beams are routed to IOT2R,7 an optical table on the right side of HAM2, while the first beam and the field which is FIG. 3. The IO on the in-air PSL table modulates the phase of the laser beam with the EOM, mode matches the light into the input mode cleaner (located inside the vacuum system), and controls the power injected into vacuum system. Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 183.195.251.6 On: Fri, 22 Apr 2016 00:51:35

014502-4 Mueller et al. Rev.Sci.Instrum.87,014502(2016) HAM2 to IOT2L HAM 3 IMI QPD to HAM ISC-Sled o HAM rom PSL R PR2 to BS MC:Mode Cleaner mirror(suspended) IM:Input Optics Mirror (suspended) QPD:Quadrant Photo Detector PSL-PD FI:Faraday Isolator to IOT2R BS:Beam Splitter (Core Optics) FIG.4.A sketch of the in-vacuum components and beam directions within the input optics in HAM 2 and HAM 3.The red beam is the forward going main beam while the green beams are auxiliary beams.The main items in the in-vacuum input optics are the input mode cleaner(IMC)which is formed by the three mirrors MC1,MC2,and MC3;the Faraday isolator (FD);and the four suspended steering mirrors IM1-4 of which IM2 and 3 match the spatial mode of the IMC into the main interferometer.The recycling cavity mirrors PRM,PR2,and PR3 are not part of the input optics.The ISC sled in HAM3 belongs to the interferometer sensing and control subsystem and provides alignment signals for the recycling cavity. reflected from MC1 are routed to IOT2L on the left side of III.INPUT OPTICS REQUIREMENTS HAM2.The position of the forward going beam through IM4 The aLIGO interferometer can be operated in different is monitored with an in-vacuum quadrant photodetector while a large fraction of this beam is also sent to an in-vacuum modes to optimize the sensitivity for different sources.10 These photodetector array which is used to monitor and stabilize the modes are characterized by the input power and the micro- laser power before it is injected into the IFO.Most of the IFO scopic position and reflectivity of the signal recycling mirror. reflected field goes back to the Faraday isolator where it is The requirements for the aLIGO input optics are specified separated from the incoming beam.This field is routed into to simultaneously meet the requirements for all anticipated HAMI where it is detected to generate length and alignment science modes and address all degrees of freedom of the laser field.Requirements in aLIGO are defined for three distinct sensing signals. frequency ranges:DC,the control band up to 10 Hz,and the In HAM3,a small fraction of the intra-mode cleaner field transmits through MC2 onto a quadrant photodetector to signal or detection band from 10 Hz to a few kHz.The require- ments in the detection band are defined in terms of linear monitor the beam position on MC2.The forward and back- wards traveling waves inside the PRC partly transmit through spectral densities and include a safety factor of ten such that all PR2 and are routed into HAMI and to an optical breadboard technical noise sources are an order of magnitude less than the sum of the fundamental limiting noise sources.To first order. inside HAM3,respectively.These beams are used by ISC for sensing and control of the interferometer and for diagnostic a perfectly symmetric Michelson interferometer is insensitive to all input noise sources which is an often overlooked reason purposes.The breadboard uses a lens to image the beam with for its use in the first place.However,all degrees of freedom orthogonal Gouy phases onto two quadrant photodetectors to of the injected laser field couple via some asymmetry to the monitor beam position and pointing inside the power recy- output signal.This drives the requirements in the control band cling cavity.IOT2R and IOT2L host photodetectors and digital cameras to monitor the power and beam sizes in each of the which are usually defined as RMS values.The more critical picked-off beams.IOT2L also hosts the photodetectors which requirements for the IO are as follows. are used by the interferometer sensing and control system to generate length and alignment sensing signals for the input A.Power mode cleaner. While the figure shows all key components in the correct The high power science modes require to inject 125 W of sequence,we intentionally left out the detailed beam routing, mode matched light into the interferometer with less than an the baffles used to suppress scattered light and protect all additional 5%in higher order modes.The PSL has to deliver components from the laser beam in case of misalignments,and 165 W of light in an appropriate TEMoo mode.Consequently, the beam dumps to capture all ghost beams. the net efficiency of TEMoo optical power transmission from A complete document tree which contains all design and the PSL output to the main interferometer has to be above as-built layouts as well as drawings of all components is avail- 75%.This sets limits on accumulated losses in all optical able within the LIGO Document Control Center(DCC)under components but also limits the allowed thermal lensing in document number E1201013.9 the EOM,the Faraday isolator,and the power control stages; Reuse of AlP Publishing content is subject to the terms at:https://publishing.aip.org/authors/rights nd-pem sions.Download to 183195.251.60nFi22A1 20160051:35

014502-4 Mueller et al. Rev. Sci. Instrum. 87, 014502 (2016) FIG. 4. A sketch of the in-vacuum components and beam directions within the input optics in HAM 2 and HAM 3. The red beam is the forward going main beam while the green beams are auxiliary beams. The main items in the in-vacuum input optics are the input mode cleaner (IMC) which is formed by the three mirrors MC1, MC2, and MC3; the Faraday isolator (FI); and the four suspended steering mirrors IM1-4 of which IM2 and 3 match the spatial mode of the IMC into the main interferometer. The recycling cavity mirrors PRM, PR2, and PR3 are not part of the input optics. The ISC sled in HAM3 belongs to the interferometer sensing and control subsystem and provides alignment signals for the recycling cavity. reflected from MC1 are routed to IOT2L on the left side of HAM2. The position of the forward going beam through IM4 is monitored with an in-vacuum quadrant photodetector while a large fraction of this beam is also sent to an in-vacuum photodetector array which is used to monitor and stabilize the laser power before it is injected into the IFO. Most of the IFO reflected field goes back to the Faraday isolator where it is separated from the incoming beam. This field is routed into HAM1 where it is detected to generate length and alignment sensing signals. In HAM3, a small fraction of the intra-mode cleaner field transmits through MC2 onto a quadrant photodetector to monitor the beam position on MC2. The forward and backwards traveling waves inside the PRC partly transmit through PR2 and are routed into HAM1 and to an optical breadboard inside HAM3, respectively. These beams are used by ISC for sensing and control of the interferometer and for diagnostic purposes. The breadboard uses a lens to image the beam with orthogonal Gouy phases onto two quadrant photodetectors to monitor beam position and pointing inside the power recycling cavity. IOT2R and IOT2L host photodetectors and digital cameras to monitor the power and beam sizes in each of the picked-off beams. IOT2L also hosts the photodetectors which are used by the interferometer sensing and control system to generate length and alignment sensing signals for the input mode cleaner. While the figure shows all key components in the correct sequence, we intentionally left out the detailed beam routing, the baffles used to suppress scattered light and protect all components from the laser beam in case of misalignments, and the beam dumps to capture all ghost beams. A complete document tree which contains all design and as-built layouts as well as drawings of all components is available within the LIGO Document Control Center8 (DCC) under document number E1201013.9 III. INPUT OPTICS REQUIREMENTS The aLIGO interferometer can be operated in different modes to optimize the sensitivity for different sources.10 These modes are characterized by the input power and the microscopic position and reflectivity of the signal recycling mirror. The requirements for the aLIGO input optics are specified to simultaneously meet the requirements for all anticipated science modes and address all degrees of freedom of the laser field. Requirements in aLIGO are defined for three distinct frequency ranges: DC, the control band up to 10 Hz, and the signal or detection band from 10 Hz to a few kHz. The requirements in the detection band are defined in terms of linear spectral densities and include a safety factor of ten such that all technical noise sources are an order of magnitude less than the sum of the fundamental limiting noise sources. To first order, a perfectly symmetric Michelson interferometer is insensitive to all input noise sources which is an often overlooked reason for its use in the first place. However, all degrees of freedom of the injected laser field couple via some asymmetry to the output signal. This drives the requirements in the control band which are usually defined as RMS values. The more critical requirements for the IO are as follows. A. Power The high power science modes require to inject 125 W of mode matched light into the interferometer with less than an additional 5% in higher order modes. The PSL has to deliver 165 W of light in an appropriate TEM00 mode. Consequently, the net efficiency of TEM00 optical power transmission from the PSL output to the main interferometer has to be above 75%. This sets limits on accumulated losses in all optical components but also limits the allowed thermal lensing in the EOM, the Faraday isolator, and the power control stages; Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 183.195.251.6 On: Fri, 22 Apr 2016 00:51:35

014502-5 Mueller et al. Rev.Sci.Instrum.87,014502(2016) the reflective optics and fused silica lenses are much less The IO does not provide any active element to change susceptible to thermal lensing.Efficient power coupling is or stabilize the laser power within the control or the detec- also dependent on good mode matching between the recycling tion band.The PSL uses a first loop which stabilizes the cavities and the arm cavities in the main interferometer. laser power measured with a photodetector on the PSL table to 2x 10-8/VHz between 20 and 100 Hz and meeting the aforementioned requirements above 100 Hz.The PSL further B.Power control stabilizes the injected power in the 20 Hz-100 Hz band with The injected power into the interferometer has to be the photodetector array shown(PSL-PD)in Figure 4 which adjustable from the control room from minimum to full power is placed after the IMC.The IO has to supply the auxiliary for diagnostic and operational purposes,to acquire lock of the beam for this array and maintain a sufficiently high correlation main interferometer,and to operate between different science with the injected beam and minimize the chances of additional modes.The rate of power change(dP/dt)has to be sufficiently power fluctuations within any of these two beams. small to limit the radiation pressure kick inside the IMC and the main interferometer to a level that can be handled by the length and alignment control system.It has to be sufficiently D.Frequency fluctuations fast to not limit the time to transition to full power after lock In the detection band,the laser frequency will ultimately acquisition,i.e.,it should be possible to change from minimum be stabilized to the common mode of the two arm cavities to maximum power within a few seconds. which are the most stable references available in this frequency Note that minimum power here cannot mean zero power range.At lower frequencies the arm cavities are not a good because of the limited extinction ratio of polarizers.Going to reference and are made to follow the frequency reference zero power requires actuation of the aforementioned mechan- inside of the PSL.The input mode cleaner acts as a frequency ical shutter which can only be accessed manually between the reference during lock acquisition and as an intermediate fre- laser enclosure and HAM1.The emergency shutter is part of quency reference during science mode.It is integrated into the PSL laser system and cuts the laser power at the source. the complex and nested laser frequency stabilization system. Furthermore,the power control system within the IO is not Based on the expected common mode servo gain the frequency used for actively stabilizing the laser power within the control noise requirements for the IMC are set to or the detection band. oy(f=10Hz)<50 mHz/VHz. C.Power fluctuations 6v(f 100Hz)<1 mHz/VHz. Fluctuations in the laser power can couple through many These requirements can be expressed equivalently as different channels to the error signal used to detect the differ- length fluctuations of the IMC: ential length of the interferometer arms,i.e.,the gravitational wave detection signal.The noise scales with the asymmetries 6(f=10Hz)<3.10-15m/H五， in the interferometer.Two different mechanisms are expected to dominate the susceptibility of the interferometer to power 6(f≥100Hz)<6.10-17m/VHz. fluctuations.The optical power inside the arm cavities will push the test masses outwards.Any change in power will E.RF modulation frequencies cause fluctuations in that pressure which can lead to displace- ment noise at low frequencies.The susceptibility to radiation The main laser field consists of a carrier and multiple pressure noise scales with differences in the power build up pairs of sidebands.The carrier has to be resonant in the arm inside the arm cavities and it is assumed that these differences cavities and the power recycling cavity:the resonance condi- are below 1%.At high frequencies,direct coupling of power tion in the signal recycling cavity depends on the tuning and fluctuations to the gravitational wave signal limits the allowed specific science mode.One pair of sidebands must be resonant power fluctuations.When the interferometer is held at its oper- in the power recycling cavity,while the second pair must ating point the two arm cavities are detuned by a few pm which resonate in both the power and signal recycling cavity.The causes some light to leak out to the dark port.Gravitational modulation signals of f1=9.1 MHz and f2 =45.5 MHz are waves will modulate these offsets causing the light power at provided by the interferometer sensing and control system. the dark port to fluctuate.Obviously,power fluctuations in the A third modulation frequency of f3=24.1 MHz is required laser itself.although highly filtered by the interferometer.will to sense and control the input mode cleaner.The last pair of cause similar fluctuations.The relative intensity noise in the sidebands should be rejected by the input mode cleaner so as detection band has to be below 2 x 10-/VHz at 10 Hz increas- not to interfere with the sensing and control system of the main ing with f to 2 x 10-8/VHz at 100 Hz and remaining flat after interferometer. this.Furthermore,the expected seismically excited motion of the test masses limits the allowed radiation pressure noise in the control band to 10-2/VHz below 0.2 Hz.Above 0.2 Hz,the F.RF modulation depth requirements follow two power laws;initially f-7 then f-3, The required modulation depths depend on the final length before connecting with the detection band requirement at and alignment sensing and control scheme.This scheme is 10Hz. likely to evolve over the commissioning time but the current Reuse of AlP Publishing content is subject to the terms at:https://publishing.aip.org/authors/rights -and-perm ssions.Download to 183195.251.60nFri22A1 2016005135

014502-5 Mueller et al. Rev. Sci. Instrum. 87, 014502 (2016) the reflective optics and fused silica lenses are much less susceptible to thermal lensing. Efficient power coupling is also dependent on good mode matching between the recycling cavities and the arm cavities in the main interferometer. B. Power control The injected power into the interferometer has to be adjustable from the control room from minimum to full power for diagnostic and operational purposes, to acquire lock of the main interferometer, and to operate between different science modes. The rate of power change (dP/dt) has to be sufficiently small to limit the radiation pressure kick inside the IMC and the main interferometer to a level that can be handled by the length and alignment control system. It has to be sufficiently fast to not limit the time to transition to full power after lock acquisition, i.e., it should be possible to change from minimum to maximum power within a few seconds. Note that minimum power here cannot mean zero power because of the limited extinction ratio of polarizers. Going to zero power requires actuation of the aforementioned mechanical shutter which can only be accessed manually between the laser enclosure and HAM1. The emergency shutter is part of the PSL laser system and cuts the laser power at the source. Furthermore, the power control system within the IO is not used for actively stabilizing the laser power within the control or the detection band. C. Power fluctuations Fluctuations in the laser power can couple through many different channels to the error signal used to detect the differential length of the interferometer arms, i.e., the gravitational wave detection signal. The noise scales with the asymmetries in the interferometer. Two different mechanisms are expected to dominate the susceptibility of the interferometer to power fluctuations. The optical power inside the arm cavities will push the test masses outwards. Any change in power will cause fluctuations in that pressure which can lead to displacement noise at low frequencies. The susceptibility to radiation pressure noise scales with differences in the power build up inside the arm cavities and it is assumed that these differences are below 1%. At high frequencies, direct coupling of power fluctuations to the gravitational wave signal limits the allowed power fluctuations. When the interferometer is held at its operating point the two arm cavities are detuned by a few pm which causes some light to leak out to the dark port.11 Gravitational waves will modulate these offsets causing the light power at the dark port to fluctuate. Obviously, power fluctuations in the laser itself, although highly filtered by the interferometer, will cause similar fluctuations. The relative intensity noise in the detection band has to be below 2 × 10−9 / √ Hz at 10 Hz increasing with f to 2 × 10−8 / √ Hz at 100 Hz and remaining flat after this. Furthermore, the expected seismically excited motion of the test masses limits the allowed radiation pressure noise in the control band to 10−2 / √ Hz below 0.2 Hz. Above 0.2 Hz, the requirements follow two power laws; initially f −7 then f −3 , before connecting with the detection band requirement at 10 Hz. The IO does not provide any active element to change or stabilize the laser power within the control or the detection band. The PSL uses a first loop which stabilizes the laser power measured with a photodetector on the PSL table to 2 × 10−8 / √ Hz between 20 and 100 Hz and meeting the aforementioned requirements above 100 Hz. The PSL further stabilizes the injected power in the 20 Hz–100 Hz band with the photodetector array shown (PSL-PD) in Figure 4 which is placed after the IMC. The IO has to supply the auxiliary beam for this array and maintain a sufficiently high correlation with the injected beam and minimize the chances of additional power fluctuations within any of these two beams. D. Frequency fluctuations In the detection band, the laser frequency will ultimately be stabilized to the common mode of the two arm cavities which are the most stable references available in this frequency range. At lower frequencies the arm cavities are not a good reference and are made to follow the frequency reference inside of the PSL. The input mode cleaner acts as a frequency reference during lock acquisition and as an intermediate frequency reference during science mode. It is integrated into the complex and nested laser frequency stabilization system. Based on the expected common mode servo gain the frequency noise requirements for the IMC are set to δν( f = 10 Hz) < 50 mHz/ √ Hz, δν( f ≥ 100 Hz) < 1 mHz/ √ Hz. These requirements can be expressed equivalently as length fluctuations of the IMC: δℓ( f = 10 Hz) < 3 · 10−15 m/ √ Hz, δℓ( f ≥ 100 Hz) < 6 · 10−17 m/ √ Hz. E. RF modulation frequencies The main laser field consists of a carrier and multiple pairs of sidebands. The carrier has to be resonant in the arm cavities and the power recycling cavity; the resonance condition in the signal recycling cavity depends on the tuning and specific science mode. One pair of sidebands must be resonant in the power recycling cavity, while the second pair must resonate in both the power and signal recycling cavity. The modulation signals of f1 = 9.1 MHz and f2 = 45.5 MHz are provided by the interferometer sensing and control system. A third modulation frequency of f3 = 24.1 MHz is required to sense and control the input mode cleaner. The last pair of sidebands should be rejected by the input mode cleaner so as not to interfere with the sensing and control system of the main interferometer. F. RF modulation depth The required modulation depths depend on the final length and alignment sensing and control scheme. This scheme is likely to evolve over the commissioning time but the current Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 183.195.251.6 On: Fri, 22 Apr 2016 00:51:35

014502-6 Mueller et al. Rev.Sci.Instrum.87,014502 (2016) assumption is that a modulation index of 0.4 for a 10 Vpp signal for a long time without a Faraday isolator between the mode driving the EOM is more than sufficient.Note that this only cleaner and the main interferometer and encountered prob- applies to the two modulation frequencies which are used for lems due to the uncontrolled length between the IMC and sensing and control of the main interferometer:the modulation IFO5(a parasitic interferometer).Initial and Enhanced LIGO index for the third frequency needs only to be large enough to never encountered any major problems with insufficient op- control the IMC. tical isolation in the Faraday isolator.The requirements of The classic phase modulation/demodulation sensing 30 dB for the optical isolation in the Faraday isolator were set scheme for a single optical cavity measures how much the based on the experience in Initial LIGO,taking into account cavity converts phase modulation into amplitude modulation the higher injected power. when near resonance.Unfortunately all phase modulators also modulate the amplitude of the laser field.This amplitude J.Additional requirements modulation can saturate the RF amplifiers and mixers in the detection chain and generate offsets in the error signals which It is well known that parasitic interferometers and scat- have to be compensated.aLIGO requires that the amplitude tered light together with mechanically excited surfaces can add modulation index is less than 10-4 of the phase modulation frequency and amplitude noise to a laser beam.The IO adopted index.12 a policy to limit the added noise to 10%of the maximum allowed noise (based on the main interferometer sensitivity); note that the allowed frequency and amplitude noise prior to G.RF modulation noise the input mode cleaner is significantly higher than after the Changes in the amplitude and phase of the RF modulation mode cleaner.This drives requirements on the residual motion signals can pollute the gravitational wave detection signal by of the optical components,the surface quality of all optical changing the power buildup of the carrier in the arm cavities components and their coatings,and on the placement and or through cross coupling in the length and alignment sensing efficiency of the optical baffes.The requirement to align the IO and control schemes.These effects were analyzed by the ISC drives requirements on actuation ranges for all optics and,last group.10 The analysis uses specifications from a commercial but not the least,the IO has to meet the stringent cleanliness crystal oscillator manufacturer produced by Wenzel Asso- and vacuum requirements of aLIGO.These requirements are ciates,Inc.as the expected oscillator phase and amplitude discussed throughout the paper when relevant. noise.These specifications for phase noise are 10-5 rad/VHz at 10 Hz falling with 1/f3/2 to3 x 10-7rad/VHz at 100 Hz and then a little faster than 1/f to 2x 10-rad/VHz at a kHz above IV.INPUT OPTICS COMPONENTS which they stay constant.The specifications for amplitude AND FINAL LAYOUT noise are 10-7/VHz at 10 Hz falling with 1/f between 10 and This section will first discuss the individual components 100 Hz and then with 1/f until 1 kHz above which they and their measured performance.This will be followed by a stay constant at 3x 10-9/VHz.These specifications have been description of the optical layout which includes a discussion adopted as requirements although the analysis shows that they of beam parameters and mode matching between the various could be relaxed at higher frequencies. areas. H.Beam jitter A.Electro-optic modulators Changes in the location and direction of the injected beam The electro-optic modulators must use a material capable can be described as scattering light from the TEMoo into a of withstanding CW optical powers of up to 200 W and TEMio mode.This light scatters back into the TEMoo mode intensities up to 25kW/cm2.At these power levels the induced inside a misaligned interferometer and creates noise in the thermal lensing,stress induced depolarization,and damage gravitational wave signal.13 This is an example where noise in threshold of the electro-optic material must be taken into the detection band,here beam jitter,couples to noise in the con- consideration.Rubidium titanyl phosphate (RTP)was chosen trol band,here tilt of the input test masses.It is expected that many years ago over other electro-optic materials,such as the test masses will all be aligned to better than 2 nrad RMS rubidium titanyl arsenate (RbTiOAsO4 or RTA)and lithium with respect to the nominal optical axis of the interferometer. niobate (LiNb03),as the most promising modulator material Under this assumption,the relative amplitude of the injected after a literature survey,discussions with various vendors, 10-mode has to stay below 10-/VHz at 10 Hz falling with and corroborating lab experiments.16.17 RTP has a very high 1/f2 until 100 Hz above which the requirement stays constant damage threshold,low optical absorption,and a fairly high at 10-8/VHz. electro-optical coefficient.Enhanced LIGO allowed for testing of the material and design over a one-year period at 30 W input I.Optical isolation power.18 The aLIGO EOM uses a patented design19 which is The Faraday isolator isolates the IMC from back reflected very similar to the one used in eLIGO:both consist of a light from the main interferometer.The requirements for the 4 x 4 x 40 mm long wedged RTP crystal(see Figure 5).The isolation ratio are based on experience gained during the initial 2.85 wedges prohibit parasitic interferometers from building years of operating LIGO and also VIRGO.4 Virgo operated up inside the crystal and allow for separation of the two Reuse of AlP Publishing conte nt is subiect to the temms at:httos //oub hing.aip.org/authors/rights-and-p Download to 183.195.251.60nFi22A1 20160051:35

014502-6 Mueller et al. Rev. Sci. Instrum. 87, 014502 (2016) assumption is that a modulation index of 0.4 for a 10 Vpp signal driving the EOM is more than sufficient. Note that this only applies to the two modulation frequencies which are used for sensing and control of the main interferometer; the modulation index for the third frequency needs only to be large enough to control the IMC. The classic phase modulation/demodulation sensing scheme for a single optical cavity measures how much the cavity converts phase modulation into amplitude modulation when near resonance. Unfortunately all phase modulators also modulate the amplitude of the laser field. This amplitude modulation can saturate the RF amplifiers and mixers in the detection chain and generate offsets in the error signals which have to be compensated. aLIGO requires that the amplitude modulation index is less than 10−4 of the phase modulation index.12 G. RF modulation noise Changes in the amplitude and phase of the RF modulation signals can pollute the gravitational wave detection signal by changing the power buildup of the carrier in the arm cavities or through cross coupling in the length and alignment sensing and control schemes. These effects were analyzed by the ISC group.10 The analysis uses specifications from a commercial crystal oscillator manufacturer produced by Wenzel Associates, Inc. as the expected oscillator phase and amplitude noise. These specifications for phase noise are 10−5 rad/ √ Hz at 10 Hz falling with 1/ f 3/2 to 3 × 10−7 rad/ √ Hz at 100 Hz and then a little faster than 1/ f to 2 × 10−8 rad/ √ Hz at a kHz above which they stay constant.10 The specifications for amplitude noise are 10−7 / √ Hz at 10 Hz falling with 1/f between 10 and 100 Hz and then with 1/  f until 1 kHz above which they stay constant at 3 × 10−9 / √ Hz. These specifications have been adopted as requirements although the analysis shows that they could be relaxed at higher frequencies. H. Beam jitter Changes in the location and direction of the injected beam can be described as scattering light from the TEM00 into a TEM10 mode. This light scatters back into the TEM00 mode inside a misaligned interferometer and creates noise in the gravitational wave signal.13 This is an example where noise in the detection band, here beam jitter, couples to noise in the control band, here tilt of the input test masses. It is expected that the test masses will all be aligned to better than 2 nrad RMS with respect to the nominal optical axis of the interferometer. Under this assumption, the relative amplitude of the injected 10-mode has to stay below 10−6 / √ Hz at 10 Hz falling with 1/ f 2 until 100 Hz above which the requirement stays constant at 10−8 / √ Hz. I. Optical isolation The Faraday isolator isolates the IMC from back reflected light from the main interferometer. The requirements for the isolation ratio are based on experience gained during the initial years of operating LIGO and also VIRGO.14 Virgo operated for a long time without a Faraday isolator between the mode cleaner and the main interferometer and encountered problems due to the uncontrolled length between the IMC and IFO15 (a parasitic interferometer). Initial and Enhanced LIGO never encountered any major problems with insufficient optical isolation in the Faraday isolator. The requirements of 30 dB for the optical isolation in the Faraday isolator were set based on the experience in Initial LIGO, taking into account the higher injected power. J. Additional requirements It is well known that parasitic interferometers and scattered light together with mechanically excited surfaces can add frequency and amplitude noise to a laser beam. The IO adopted a policy to limit the added noise to 10% of the maximum allowed noise (based on the main interferometer sensitivity); note that the allowed frequency and amplitude noise prior to the input mode cleaner is significantly higher than after the mode cleaner. This drives requirements on the residual motion of the optical components, the surface quality of all optical components and their coatings, and on the placement and efficiency of the optical baffles. The requirement to align the IO drives requirements on actuation ranges for all optics and, last but not the least, the IO has to meet the stringent cleanliness and vacuum requirements of aLIGO. These requirements are discussed throughout the paper when relevant. IV. INPUT OPTICS COMPONENTS AND FINAL LAYOUT This section will first discuss the individual components and their measured performance. This will be followed by a description of the optical layout which includes a discussion of beam parameters and mode matching between the various areas. A. Electro-optic modulators The electro-optic modulators must use a material capable of withstanding CW optical powers of up to 200 W and intensities up to 25 kW/cm2 . At these power levels the induced thermal lensing, stress induced depolarization, and damage threshold of the electro-optic material must be taken into consideration. Rubidium titanyl phosphate (RTP) was chosen many years ago over other electro-optic materials, such as rubidium titanyl arsenate (RbTiOAsO4 or RTA) and lithium niobate (LiNb03), as the most promising modulator material after a literature survey, discussions with various vendors, and corroborating lab experiments.16,17 RTP has a very high damage threshold, low optical absorption, and a fairly high electro-optical coefficient. Enhanced LIGO allowed for testing of the material and design over a one-year period at 30 W input power.18 The aLIGO EOM uses a patented design19 which is very similar to the one used in eLIGO; both consist of a 4 × 4 × 40 mm long wedged RTP crystal (see Figure 5). The 2.85◦ wedges prohibit parasitic interferometers from building up inside the crystal and allow for separation of the two Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 183.195.251.6 On: Fri, 22 Apr 2016 00:51:35

014502-7 Mueller et al. Rev.Sci.Instrum.87,014502(2016) 898 LGO EOM2 FIG.5.Images of the electro-optic modulator.The housing uses two modules:the crystal and the electrodes are placed in the lower module while the upper module houses the coils for the three resonant circuits.The left picture shows the inside of the lower module:the aLIGO EOM consists of a wedged RTP crystal with three pairs of electrodes.The two 15 mm electrodes on the outside are used for the main modulation frequencies f=9.1 MHz and f2=45.5 MHz.The 7 mm electrodes in the middle are used for f3=24.1 MHz.The crystal and the electrodes are clamped between two macor pieces.The right picture shows the final modulator (both modules)on a five axis alignment stage. polarizations of the injected laser field with an extinction ratio part of a resonant circuit in the form of a m network where the of better than 105.This separation avoids polarization rotation additional inductor and capacitor are used to simultaneously which could otherwise convert phase modulation to amplitude match the resonance frequency and create the required 50 modulation.The AR coated surfaces have a rest reflectivity of input impedance. less than 0.1%.For aLIGO we use two 15 mm long pairs of After installation and alignment at both sites,initial tests electrodes for the two main modulation frequencies and one confirmed that the RTP crystals do not produce a signifi- 7 mm long pair for the auxiliary frequency used to control cant thermal lens.An optical spectrum analyzer was used to the IMC.A three electrode scheme was chosen over a single measure the modulation index as a function of modulation electrode design in order to simplify the circuitry design frequency for each of the three resonant circuits.The results especially with regard to the fact that the resonant frequencies are shown in Figure 6.The modulation indices for fi and f2 must be precisely tuned after the interferometer lengths are meet the requirements at both sites while the f3 modulation determined.Each electrode pair forms a capacitor which is index is still a little low,especially at LLO.However,early EOM Resonances 9.1 MHz 24 MHz 45.5 MHz LLO 0.4 +LHO 0.35 0.3 0.25 0.2 0.15 0.1 0.05 9 10 22 23 24 25 44 454647 Frequency(MHz) Frequency(MHz) Frequency(MHz) FIG.6.The measured modulation indices for both the Livingston and Hanford EOM with a 24 dBm drive.The data is shown together with a best fit to the expected circuit response. Reuse of AlP Publishing content is subject to the terms at:https://publishing.aip.org/authors/rights-and-permissions.Download to IP:183.195.251.6 On:Fri.22 Apr 2016005135

014502-7 Mueller et al. Rev. Sci. Instrum. 87, 014502 (2016) FIG. 5. Images of the electro-optic modulator. The housing uses two modules; the crystal and the electrodes are placed in the lower module while the upper module houses the coils for the three resonant circuits. The left picture shows the inside of the lower module: the aLIGO EOM consists of a wedged RTP crystal with three pairs of electrodes. The two 15 mm electrodes on the outside are used for the main modulation frequencies f1 = 9.1 MHz and f2 = 45.5 MHz. The 7 mm electrodes in the middle are used for f3 = 24.1 MHz. The crystal and the electrodes are clamped between two macor pieces. The right picture shows the final modulator (both modules) on a five axis alignment stage. polarizations of the injected laser field with an extinction ratio of better than 105 . This separation avoids polarization rotation which could otherwise convert phase modulation to amplitude modulation. The AR coated surfaces have a rest reflectivity of less than 0.1%. For aLIGO we use two 15 mm long pairs of electrodes for the two main modulation frequencies and one 7 mm long pair for the auxiliary frequency used to control the IMC. A three electrode scheme was chosen over a single electrode design in order to simplify the circuitry design especially with regard to the fact that the resonant frequencies must be precisely tuned after the interferometer lengths are determined. Each electrode pair forms a capacitor which is part of a resonant circuit in the form of a π network where the additional inductor and capacitor are used to simultaneously match the resonance frequency and create the required 50 Ω input impedance. After installation and alignment at both sites, initial tests confirmed that the RTP crystals do not produce a significant thermal lens. An optical spectrum analyzer was used to measure the modulation index as a function of modulation frequency for each of the three resonant circuits. The results are shown in Figure 6. The modulation indices for f1 and f2 meet the requirements at both sites while the f3 modulation index is still a little low, especially at LLO. However, early FIG. 6. The measured modulation indices for both the Livingston and Hanford EOM with a 24 dBm drive. The data is shown together with a best fit to the expected circuit response. Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 183.195.251.6 On: Fri, 22 Apr 2016 00:51:35

014502-8 Mueller et al. Rev.Sci.Instrum.87,014502(2016) commissioning experience indicates that the modulation The softer calcite polarizers and the deuterated potassium indices are sufficient for the aLIGO length and alignment dihydrogen phosphate(DKDP)crystal were procured from the sensing scheme and it was decided to use the EOM as is manufacturers with their standard polishings and coatings.The for now and potentially improve the resonant circuits later if two calcite polarizers each have a thickness of~5 mm and necessary. are wedged at 8.5 to allow the orthogonally polarized beams The residual amplitude modulation20(RFAM)produced to separate sufficiently.The calcite wedges have an extinction by the EOM was also characterized.The AM/PM ratio for ratio of at least 105 and more than 99%optical efficiency. each of the three sidebands was measured to be 1.0.10-4, The magnetic field is created by a stack of seven magne- 1.2.10-5.and4.1.10-5 for the9.1MHz.24.0MHz.and45.5 tized Fe-Nb magnetic disks25 each having a bore of 24 mm MHz sidebands.All three measurements come out to be at or and a thickness of 19.7 mm.This stack produces a maximum below the requirement of 10-derived by Kokeyama et al.12 axial field of 1.16 T (LLO)and 1.55 T(LHO)near its center Temporal variation of the RFAM generation was found to which falls off towards the end.The difference in the magnetic be due primarily to temperature dependence which is able to field is caused by the selection of the magnetic materials and push the AM/PM ratio at 9.1 MHz as high as 3.10-4.This the thermal treatment of the individual magnets.26 The TGG may need to be addressed with a temperature stabilization crystals and quartz rotator are installed about 3 cm apart from servo in the future if RFAM is found to be an issue during each other before being fine tuned to produce 22.5 of rotation detector commissioning,but the design of the modulator was by adjusting their depth in the magnet.The entire FI is mounted left unchanged until such an issue arises. on a 648 mm x 178 mm breadboard for convenient transfer Detailed design drawings,assembly instructions,and into the horizontal access module (HAM)chamber after out- test reports are available under LIGO document number of-vacuum optimization. T1300084.21 After undergoing a thorough cleaning procedure,the FI was assembled and aligned with the main PSL beam in the laser enclosure.The optical table in the enclosure is made B.Faraday isolator from stainless steel while the optical table in HAM2 is made The Faraday isolator is a much more complicated optical from aluminum.The differences in magnetic susceptibility are device compared to the EOM.It is more susceptible to thermal significant enough to require the FI to be raised with an~11 cm lensing and its location after the mode cleaner amplifies the thick granite block visible in the bottom picture in Figure 7. requirement to maintain a good spatial mode.The FI has to The bottom periscope mirror in Figure 3 was removed and the handle between 20 and 130 W of laser power without signifi- beam was sent via several mirrors through the Faraday isolator. cantly altering the beam profile or polarization of the beam. This setting ensured that the beam parameters,beam size, Like the EOM,the aLIGO FI is also very similar to the FI and divergence angle,are very similar to the ones expected used ineLIGO.18 Both were designed to minimize and mitigate in-vacuum. thermal lensing and thermal stress induced depolarization by The thermal lensing of the Faraday isolator was deter- compensating these effects in subsequent crystals.22.23 mined from beam-scan measurements of a sample of the beam The aLIGO FI design consists of a Faraday rotator,a pair after it was transmitted through the isolator for incident powers of calcite-wedge polarizers,an element with a negative dn/dT as high as 120 W at LLO and 140 W at LHO.At both sites. for thermal-lens compensation,and a picomotor-controlled the diagnostic beam was focused with a lens of 1 m focal half-wave plate for restoring the optical isolation in-situ.In length and the beam profile was recorded with CCD or rotating addition,a heat sink is connected to the holders of the magneto- slit beam scans as a function of power for different DKDP optical crystals to drain excess heat into the FI breadboard. crystals.The thermal lens at the location of the Faraday was The Faraday rotator is based on an arrangement developed by then computed using an ABCD matrix algorithm. Khazanov et al.,24 that uses a pair of~1 cm long Terbium Figure 8 shows the thermal lensing measurements for the Gallium Garnet(TGG)crystals as magneto-optical elements, TGG crystals and different DKDP crystals at LHO and LLO. each nominally producing a 22.5 rotation of the electric field The magnitude of the thermal lensing in the DKDP is a nearly when placed in a magnetic field of about IT.They are separated linear function of the incident power if all other parameters are by a ~1 cm long piece of quartz that rotates the polarization equal.In reality the absorption varies from sample to sample field reciprocally by67.5°±0.6°.This arrangement(shown and causes the selection of the DKDP to be somewhat stochas- schematically at the top of Figure 7)allows thermally induced tic,a fact which is evident in the small difference between the birefringence produced in the first magneto-optical element to 3.0 mm and 3.5 mm measurements at LLO. be mostly compensated in the second one.The HWP is a zero- The length of the DKDP crystal was chosen to compen- order epoxy-free quartz half-wave plate.It is set to rotate the sate the a priori unknown thermal lensing in the TGG crys- polarization by an additional 22.5 to have 0 net rotation in tals.Based on experience from Initial and Enhanced LIGO. the forward going and 90 in the backward going direction. the expectation was that DKDP crystals between 3.5 mm and All crystals were selected to minimize absorption,ther- 5.5 mm would be needed to compensate the thermal lensing mal beam distortion and surface roughness.Those made of in the TGG crystals.However,the absorption in the newly harder and non-hygroscopic materials,the half-wave plate, purchased TGG crystals was lower than expected and even our quartz rotator,and TGG crystals,are all super-polished (sur- shortest crystals overcompensated.While the low absorption face roughness below 0.5 nm)and received a custom low loss in TGG is obviously good,it required to shorten the originally IBS AR coating with a rest reflectivity of less than 300 ppm. ordered DKDP.We choose 3.5 mm for both isolators instead Reuse of AlP Publishing c tent is subject to the te at:https//publis .aip.org/authors/rights wnload to IP. 183.195.251.60nFi22A 2016005135

014502-8 Mueller et al. Rev. Sci. Instrum. 87, 014502 (2016) commissioning experience indicates that the modulation indices are sufficient for the aLIGO length and alignment sensing scheme and it was decided to use the EOM as is for now and potentially improve the resonant circuits later if necessary. The residual amplitude modulation20 (RFAM) produced by the EOM was also characterized. The AM/PM ratio for each of the three sidebands was measured to be 1.0 · 10−4 , 1.2 · 10−5 , and 4.1 · 10−5 for the 9.1 MHz, 24.0 MHz, and 45.5 MHz sidebands. All three measurements come out to be at or below the requirement of 10−4 derived by Kokeyama et al.12 Temporal variation of the RFAM generation was found to be due primarily to temperature dependence which is able to push the AM/PM ratio at 9.1 MHz as high as 3 · 10−4 . This may need to be addressed with a temperature stabilization servo in the future if RFAM is found to be an issue during detector commissioning, but the design of the modulator was left unchanged until such an issue arises. Detailed design drawings, assembly instructions, and test reports are available under LIGO document number T1300084.21 B. Faraday isolator The Faraday isolator is a much more complicated optical device compared to the EOM. It is more susceptible to thermal lensing and its location after the mode cleaner amplifies the requirement to maintain a good spatial mode. The FI has to handle between 20 and 130 W of laser power without significantly altering the beam profile or polarization of the beam. Like the EOM, the aLIGO FI is also very similar to the FI used in eLIGO.18Both were designed to minimize and mitigate thermal lensing and thermal stress induced depolarization by compensating these effects in subsequent crystals.22,23 The aLIGO FI design consists of a Faraday rotator, a pair of calcite-wedge polarizers, an element with a negative dn/dT for thermal-lens compensation, and a picomotor-controlled half-wave plate for restoring the optical isolation in-situ. In addition, a heat sink is connected to the holders of the magnetooptical crystals to drain excess heat into the FI breadboard. The Faraday rotator is based on an arrangement developed by Khazanov et al., 24 that uses a pair of ∼1 cm long Terbium Gallium Garnet (TGG) crystals as magneto-optical elements, each nominally producing a 22.5 ◦ rotation of the electric field when placed in a magnetic field of about 1T. They are separated by a ∼1 cm long piece of quartz that rotates the polarization field reciprocally by 67.5 ◦ ± 0.6 ◦ . This arrangement (shown schematically at the top of Figure 7) allows thermally induced birefringence produced in the first magneto-optical element to be mostly compensated in the second one. The HWP is a zeroorder epoxy-free quartz half-wave plate. It is set to rotate the polarization by an additional 22.5 ◦ to have 0◦ net rotation in the forward going and 90◦ in the backward going direction. All crystals were selected to minimize absorption, thermal beam distortion and surface roughness. Those made of harder and non-hygroscopic materials, the half-wave plate, quartz rotator, and TGG crystals, are all super-polished (surface roughness below 0.5 nm) and received a custom low loss IBS AR coating with a rest reflectivity of less than 300 ppm. The softer calcite polarizers and the deuterated potassium dihydrogen phosphate (DKDP) crystal were procured from the manufacturers with their standard polishings and coatings. The two calcite polarizers each have a thickness of ∼5 mm and are wedged at 8.5 ◦ to allow the orthogonally polarized beams to separate sufficiently. The calcite wedges have an extinction ratio of at least 105 and more than 99% optical efficiency. The magnetic field is created by a stack of seven magnetized Fe-Nb magnetic disks25 each having a bore of 24 mm and a thickness of 19.7 mm. This stack produces a maximum axial field of 1.16 T (LLO) and 1.55 T (LHO) near its center which falls off towards the end. The difference in the magnetic field is caused by the selection of the magnetic materials and the thermal treatment of the individual magnets.26 The TGG crystals and quartz rotator are installed about 3 cm apart from each other before being fine tuned to produce 22.5 ◦ of rotation by adjusting their depth in the magnet. The entire FI is mounted on a 648 mm × 178 mm breadboard for convenient transfer into the horizontal access module (HAM) chamber after outof-vacuum optimization. After undergoing a thorough cleaning procedure, the FI was assembled and aligned with the main PSL beam in the laser enclosure. The optical table in the enclosure is made from stainless steel while the optical table in HAM2 is made from aluminum. The differences in magnetic susceptibility are significant enough to require the FI to be raised with an ∼11 cm thick granite block visible in the bottom picture in Figure 7. The bottom periscope mirror in Figure 3 was removed and the beam was sent via several mirrors through the Faraday isolator. This setting ensured that the beam parameters, beam size, and divergence angle, are very similar to the ones expected in-vacuum. The thermal lensing of the Faraday isolator was determined from beam-scan measurements of a sample of the beam after it was transmitted through the isolator for incident powers as high as 120 W at LLO and 140 W at LHO. At both sites, the diagnostic beam was focused with a lens of 1 m focal length and the beam profile was recorded with CCD or rotating slit beam scans as a function of power for different DKDP crystals. The thermal lens at the location of the Faraday was then computed using an ABCD matrix algorithm. Figure 8 shows the thermal lensing measurements for the TGG crystals and different DKDP crystals at LHO and LLO. The magnitude of the thermal lensing in the DKDP is a nearly linear function of the incident power if all other parameters are equal. In reality the absorption varies from sample to sample and causes the selection of the DKDP to be somewhat stochastic, a fact which is evident in the small difference between the 3.0 mm and 3.5 mm measurements at LLO. The length of the DKDP crystal was chosen to compensate the a priori unknown thermal lensing in the TGG crystals. Based on experience from Initial and Enhanced LIGO, the expectation was that DKDP crystals between 3.5 mm and 5.5 mm would be needed to compensate the thermal lensing in the TGG crystals. However, the absorption in the newly purchased TGG crystals was lower than expected and even our shortest crystals overcompensated. While the low absorption in TGG is obviously good, it required to shorten the originally ordered DKDP. We choose 3.5 mm for both isolators instead Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 183.195.251.6 On: Fri, 22 Apr 2016 00:51:35