Ultra-narrow Bandpass Coatings For Deep Space Optical .

Omega Optical, Inc.tr: 9/13/2017Ultra-narrow Bandpass Coatings for Deep Space OpticalCommunications (DSOC)Thomas Rahmlow, Timothy Upton, Markus Fredell, Terry Finnell, Stephen Washkevich, KirkWinchester, Tina Hoppock and Robert JohnsonOmega Optical, Inc.21 Omega DriveBrattleboro, VT, 05301thoppock@omegafilters.com(802) 251-7390NASA Phase 1 SBIR: NNX17CP58PProgram Monitor: Michael Peng, PhD1

Omega Optical, Inc.tr: 9/13/2017Problem Statement Deep Space Optical Communication has the potential for higher data rates and information density. Once developed, the protocol can handle a large number of channels in parallel.Ground Terminal:Mars 1550.1 nm, 0.175 nm FWHM ‘Flat Top’ultra-narrow bandpass filter 3o operating angle Tilt tunable Well collimated (F#48) beam OD 12 off-band rejectionSunEarthFlight Terminal: 1064 nm FWHM ‘Flat Top’ bandpass filter Space qualified, thermally stable Signal strength is very low: photon counting Very high background noise: can be within 5% of the sun when Earth and Mars are inopposition.2

Omega Optical, Inc.tr: 9/13/2017Ground-Based Terminal Filter DesignFigure 1: Modeled transmission for the ground baseterminal filter. The design is a multi-cavity, flat topdesign to maximize throughput and signal to noiseParameterCenter WavelengthTemperature Shift: 1oCOD 350 nm to 4000 nmAverageGoalDesign1550.1 nmat 3o AOI 0.011550.1 nm at3o AOI0.0212 8 no add’lblockerFigure 2: Modeled transmission for the same design aspresented in the previous figure is plotted on a logplot to highlight off band optical density.Phase 1Goal - 0.01 nmPhase 2Goal - 0.01 nm0.02 0.01 10 123CommentCenter wavelength can beangle tuned - 0.5 degreesSubstrate selection (CTE of0.92)We propose an additionalblocker element

Omega Optical, Inc.tr: 9/13/2017Flight Terminal Filter DesignFigure 3: Modeled transmission for the flight terminalultra-narrow band-pass filter. The design is a multicavity, flat top design to maximize throughput andsignal-to-noise.ParameterCenter WavelengthWavelength Shift: 1oTemperature Shift: 1oCOD 350 nm to 4000 nmAverageGoalDesign1064 nm at3o AOI1064 nm at3o AOI 0.0112Figure 4: Modeled transmission for the same design aspresented in the previous figure is plotted againstoptical density (OD -Log10 T). Both the filter on S1and blocker design on S2 is modeled.Phase 1Goal - 0.01 nmPhase 2Goal - 0.01 nm0.020.02 0.0110 10 124CommentCenter wavelength can beangle tuned - 0.5 degreesSubstrate selection (CTE of0.92)Alternate blocker schemesare being considered

Omega Optical, Inc.tr: 9/13/2017State of the ArtFigure 5: Measured transmission for two 2.5 nm, multi-cavitybandpass filters at 0 and 5º AOI are overlaid with the target laserwavelengths. The application is for free space laser communication.The operational angle of incidence is 0 to 5º. The laser wavelengthsare 1552.3 and 1548.7.Figure 6: The same measured transmission data presented in theprevious figure are plotted on a log scale to highlight filter slope andrejection of the corresponding adjacent laser line. The filters providehigh in-band transmission at 0 to 5º and OD 4 rejection of theadjacent laser bands.Figure 7: Measured transmission of three ultra-narrow notch filters fabricated at Omega Optical, Inc. is presented. These are laser wavelengthscanning data for 1.0nm wide, 0.65nm wide, and 0.3nm wide bandpass filters. Ultra-narrow notch bandpass filters can be reliably fabricated,but spectral shift with angle and temperature need to be matched to system requirements. (SPIE Paper 9612-21: Sub-nanometer band passcoatings for LIDAR and astronomy)5

Omega Optical, Inc.tr: 9/13/2017State of the Art Hardware and ProcessesFigure 8: The Helios cleanroom provides a cleanenvironment for final cleaning and inspection ofsubstrates prior to coating. Tight controls limit surfacecontamination and particles that can lead to pinholesin the surface.Figure 9: The Helios multi-target high volume reactivesputtering coater provides high volume capability andreliable performance for the most challengingdesigns. Even so – uniformity of 0.25% across a 200mm plate and 5 to 20 ang layer thickness errors limitcapability for the manufacture of sub-nm multi cavitydesigns.6

Omega Optical, Inc.tr: 9/13/2017Significance of the Innovation: Why the State of Art Doesn’t Meet the NeedFilter DesignSEM of a Multi-cavity Interference Filter7

Omega Optical, Inc.tr: 9/13/2017Single Cavity versus Multi-Cavity DesignFigure 10: The transmission of a single cavity (red) andmulti-cavity (olive dashed) bandpass filter designed tothe same bandwidth is overlaid. The multi-cavitydesign gives a flat top response and a sharper edgeand deeper shirt.Figure 11: The same filter designs presented in theprevious figure are plotted on a log scale. The multicavity design drives down to an optical density of 6within 1 nm of the CW. Signal to noise for the multicavity is estimated to be 7.4x better than the singlecavity.Signal to noise for the multi-cavity is estimated to be7.4x better than the single cavity 8

Omega Optical, Inc.tr: 9/13/2017But Making the Multi-cavity comes with some very strong challengesSingle CavityFigure 12: Transmission for 100 trials of a Monte Carlosimulation of a 0.5 nm single cavity design assuming a1% error in layer thickness is overlaid.Figure 13: Relative sensitivity of thickness layer errorsby layer. The impact of error is greatest for errors inthe thickness of the central cavity layer. Thicknesserrors in the outer reflector are small.Random errors (1%) in layer thickness for a single cavity Fabry Perotdesign do not significantly distort the shape, band width ortransmission, only the center wavelength.9

Omega Optical, Inc.tr: 9/13/2017Multi-CavityFigure 14: Transmission for 10 trials of a MonteCarlo simulation of a 0.175 multi-cavity designfor a 0.005% error in layer thickness is overlaid.Random errors in layer thickness for the multicavity design distort the bandpass shape anddepress in band transmission.Figure 15: Relative sensitivity of thickness layererrors by layer. The design is comprised ofthree cavities. Thickness errors in any of thethree cavity layers drives the filter out ofcoherence and destroys the filter’s in bandtransmission.Random errors of only 0.005% in layer thickness for the multi-cavity designdistort the bandpass shape and depress in band transmission.10

Omega Optical, Inc.tr: 9/13/2017Uniformity, Layer Thickness Control and Plate to Plate Variability We reliably meet uniformity across a 200 mm plate of /-0.25% of thecenter wavelength.o At 1550 nm, 0.25% non-uniformity translates to a gradient error of0.75 nm across a 25.4 mm aperture.o This non-uniformity is 100x too high to produce the 0.175 nm targetbandwidth filters using state of the art technology. Similarly, layer thickness accuracy using the turning point monitor on theHelios coater is estimated from measured scans to be in the range of 8 to20 angstroms (depending on the algorithm used)o This error can push the multiple cavities apart by as much as 2 nmand thereby destroy cavity-to-cavity coherence.11

Omega Optical, Inc.tr: 9/13/2017Work in Progress Filter design studies using the current material system were completed Four test deposition runs were completed demonstrating the limits of theHelios systems monitor algorithms for making ultra-narrow bandpassfilters Annealing studies demonstrate an effective method for annealing filterslocally to correct for inhomogeneity across the filter sample High confidence in measuring filter performance was gained frommeasurements of test filters with and without AR coating In-situ process monitoring development is continuing12

Omega Optical, Inc.tr: 9/13/2017Filter Fabrication We are currently doing a number of filter test runs. These test runs provide test samples for process characterization as wellas provide an ultra-narrow band metrology test leCavityTwoCavityMonitor2 layers – rest byroundsFirst ordermonitoringSecond ordermonitoringSecond ordermonitoringResultLow %T (60%) and off wavelength: 1598nmMonitor failed to correctly count turning pts comingout of cavityRan well – high %T (97%) with AR and 2x750nmmonitor placed CW at 1496 nm.Ran well – Low %T (OD4) – cavities not aligned13

Omega Optical, Inc.tr: 9/13/2017Metrology – In SituFigure 16: A circulator and scrambler were added to the laser sourcepath but no improvement is signal stability was noted.Figure 17: Screen display of the optical monitor stability using thelaser alone, with the laser and circulator and with the laser, circulatorand scrambler. No significant differences were noted.Metrology – Post Process14

Omega Optical, Inc.tr: 9/13/2017Metrology – Filter CharacterizationFigure 18: Laser measured transmission of the filterfabricated in test run 3 using 2 order monitoring. Therewas no AR on the second surfaceFigure 19: Laser measured transmission of the filterfabricated in test run 3 using 2 order monitoring.There was no AR on the second surfaceFigure 20: Laser measured transmission of the filterfabricated in test run 3 using 2 order monitoring. Thereis an AR on the second surfaceFigure 21: Laser measured transmission of the filterfabricated in test run 3 using 2 order monitoring.There is an AR on the second surface15

Omega Optical, Inc.tr: 9/13/2017UniformityFigure 22: Distribution of center wavelengthFigure 23: Distribution of bandwidth (FWHM)Figure 24: Distribution of peak transmissionFigure 25: Filter transmission and OD16

Omega Optical, Inc.tr: 9/13/2017Annealing StudiesExposure power Percent change CWL Percent Change -7.68%17

Omega Optical, Inc.tr: 9/13/2017Optical DensityOptical Density is high by design – but several factors limit what can be realized Pin holes and inclusion in the film Material absorption ScatterThe impact of Pin Holes on Optical DensityDensity Dia: 0.001 mm Dia: 0.01 mm Dia: 0.1 E-093.14E-0718

Omega Optical, Inc.tr: 9/13/2017In Conclusion We are working with state of the art tools to extend the our capability tofabricate ultra-narrow band high performance optical filters Current work quantifies process and metrology capability and definesareas of measureable improvement We are developing in-process and post process annealing techniques toprecisely tune the center wavelength of each cavity and of the completedfilter We are developing micro-mapping tooling to precisely measureperformance across the apperature Work is in process – stay tuned.19

ultra-narrow band-pass filter. The design is a multi-cavity, flat top design to maximize throughput and signal-to-noise. Figure 4: Modeled transmission for the same design as presented in the previous figure is plotted against optical density (OD -Log 10 T). Both the filter on S1 and block